Professional Documents
Culture Documents
(Methods in Molecular Biology 1781) Qing Yan - Psychoneuroimmunology-Springer New York - Humana Press (2018)
(Methods in Molecular Biology 1781) Qing Yan - Psychoneuroimmunology-Springer New York - Humana Press (2018)
Psychoneuro-
immunology
Methods and Protocols
Second Edition
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, AL10 9AB, UK
Second Edition
Edited by
Qing Yan
PharmTao, Santa Clara, CA, USA
University of Maryland University College, Adelphi, MD, USA
Editor
Qing Yan
PharmTao
Santa Clara, CA, USA
University of Maryland University College
Adelphi, MD, USA
This Humana Press imprint is published by the registered company Springer Science+Business Media, LLC part of
Springer Nature.
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
Preface
v
vi Preface
signaling networks to analyze the information flow in the bio-behavioral circuitry and to
support behavioral stress management therapy (see Chapter 8).
In addition, studies in PNI such as those focusing on systemic inflammation and gut
microbiota may help elucidate the neuroimmune mechanisms of comorbid disorders includ-
ing depression in patients with heart failure (see Chapter 9). In systemic disorders such as
cancer and autoimmune diseases, the imbalances in cytokine networks, the hypothalamic-
pituitary-adrenal axis, and circadian rhythms can be critical markers for prognosis and
disease control (see Chapter 10).
Part II of this book introduces various cutting-edge technologies and methods for PNI
studies. These technologies include the utilizations of mouse models, the chromium release
whole blood assay, imaging techniques, as well as vaccine models.
For example, novel tools including optogenetics and chemogenetics may empower the
studies of the brain-immune interactions in PNI (see Chapter 11). Natural killer (NK) cells
are sensitive barometers of the effects of stressors on the immune system. A chromium
(51Cr) release whole blood bioassay can be used to examine the target cell killing capacity of
NK cells (see Chapter 12).
In addition, mouse models have extensive applications in PNI studies. Immunobeha-
vioral phenotyping is a first-line method for exploring the neuroimmune system. Behavioral
tests are frequently used to examine neuroimmune activation in mice (see Chapter 13). The
murine MRL model with high validity in analyzing principal pathogenic circuits has been
considered indispensable in understanding the brain-immune links and autoimmune dis-
eases (see Chapter 14).
Positron emission tomography (PET) imaging is a tool for measuring brain metabolism
and target molecules. By detecting brain variables, PET imaging can be combined with
other experimental and clinical model systems for PNI studies (see Chapter 15).
Furthermore, vaccination models are very useful for analyzing the effects of psychosocial
factors on immunity (see Chapter 16). Such protocols can help elucidate the association
between stress and vaccination responses. The modern multiplex techniques would
empower PNI research for promoting vaccine responses among at-risk populations (see
Chapter 17).
Moreover, the approaches for analyzing the fetal cholinergic signaling on systems and
cellular scales are very useful for the studies of PNI phenotypes (see Chapter 18). The
protocols for the research in prenatal stress and postnatal brain development would contrib-
ute to the development of perinatal PNI (see Chapter 19).
By covering topics from systems-based models to advanced technologies, this book can
be used by biomedical students and professionals at all levels who are interested in integra-
tive studies in psychology, psychiatry, neuroscience, immunology, molecular biology, genet-
ics, bioengineering, physiology, pathology, microbiology, systems biology, and clinical
medicine. Written by leading experts in the field, this book intends to provide a practical,
state-of-the-art, and holistic view for the translation of PNI into better preventive and
personalized medical practice.
I would like to thank all of the authors for sharing their profound thoughts and
experiences, and for making valuable contributions to this exciting new field. I also thank
the series editor, Dr. John Walker, for his help with the editing.
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
vii
viii Contents
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
Contributors
ix
x Contributors
Abstract
Studies in psychoneuroimmunology (PNI) would provide better insights into the “whole mind-body
system.” Systems biology models of the complex adaptive systems (CASs), such as a conceptual framework
of “Yin-Yang dynamics,” may be helpful for identifying systems-based biomarkers and targets for more
effective prevention and treatment. The disturbances in the Yin-Yang dynamical balance may result in stress,
inflammation, and various disorders including insomnia, Alzheimer’s disease, obesity, diabetes, cardiovas-
cular diseases, skin disorders, and cancer. At the molecular and cellular levels, the imbalances in the cytokine
pathways, mitochondria networks, redox systems, and various signaling pathways may contribute to
systemic inflammation. In the nervous system, Yin and Yang may represent the dynamical associations
between the progressive and regressive processes in aging and neurodegenerative diseases. In response to
the damages to the heart, the Yin-Yang dynamical balance between proinflammatory and anti-inflammatory
cytokine networks is crucial. The studies of cancer have revealed the importance of the Yin-Yang dynamics
in the tumoricidal and tumorigenic activities of the immune system. Stress-induced neuroimmune imbal-
ances are also essential in chronic skin disorders including atopic dermatitis and psoriasis. With the
integrative framework, the restoration of the Yin-Yang dynamics can become the objective of dynamical
systems medicine.
Key words Complex adaptive systems, Dynamical medicine, Immune, Inflammation, Mind-body,
Psychoneuroimmunology, Stress, Systems biology, Systems medicine, Yin-Yang
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_1, © Springer Science+Business Media, LLC, part of Springer Nature 2018
3
4 Qing Yan
Table 1
Yin-Yang dynamics and neuroimmune imbalances in health and diseases
(continued)
Stress and Systemic Inflammation: Yin-Yang Dynamics in Health and Diseases 7
Table 1
(continued)
Sleep is essential for health. Sleep problems such as short sleep and
sleep disturbances have been closely associated with cellular altera-
tions and aging-associated diseases [4]. Inflammation has been
identified as the biological link between sleep problems and the
elevated risks of depression, pain, and infectious diseases. In addi-
tion, the connections between chronic stress and obesity may
8 Qing Yan
Stressful life events have been closely correlated with the onset of
inflammatory skin diseases [4]. Depression is common among
patients with dermatological problems. PNI evidences have
revealed the stress-induced neuroimmune imbalances in chronic
skin disorders including atopic dermatitis, psoriasis, malignant mel-
anoma, scleroderma, lichen sclerosus, and eosinophilic fasciitis
[4, 46].
The dynamical balances in the psycho-neuro-immuno-endo-
crine-cutaneous networks rely on the interactions among various
signaling pathways of neuropeptides, hormones, and immune mes-
sengers. For example, stress-induced alterations in the CNS may
disturb the balance between cell-mediated (Th1) and humoral
(Th2) immune functions, leading to the development of various
skin disorders ([47]; also see Table 1).
In itches and other allergic reactions, environmental factors
including allergens and psychosocial stress may stimulate the gen-
eration of neuropeptides and activate mast cells, leading to the
stress responses from peripheral organs including the skin
[48]. The Yin-Yang dynamical balances in neuropeptides and neu-
rotrophins may affect the promotion or inhibition of tissue regen-
eration and inflammation.
For instance, among patients with allergic dermatitis, perceived
stress may lead to higher levels of the neurotrophin nerve growth
factor (NGF), leading to elevated levels of proallergic cytokines
([49]; also see Table 1). Such immune imbalance may result in
cutaneous inflammation.
In the heterogeneity disease of psoriasis, the imbalanced
miRNA axis has been identified between miR-31/miR-203 and
hsa-miR-99a/miR-125b with the Yin-Yang features of opposite
yet complementary, interconnected, and interdependent correla-
tions [50]. The higher levels of the pair of miR-31/miR-203 and
lower levels of the pair of hsa-miR-99a/miR-125b may be involved
Stress and Systemic Inflammation: Yin-Yang Dynamics in Health and Diseases 17
8 Conclusion
Cytokine Pathways
Neurotransmitters
Interactions Hormones
at Molecular/ Metabolic Pathways
Cellular Levels Redox Systems Imbalances Systemic
Yin-Yang
Epigenetics Inflammation
Dynamics
Mitochondrial Functions
Circadian Networks
Fig. 1 The conceptual framework of “Yin-Yang dynamics” in health and diseases at various systems levels
18 Qing Yan
References
1. Yan Q (2014) From pharmacogenomics and 8. Selfridge JM, Gotoh T, Schiffhauer S et al
systems biology to personalized care: a frame- (2016) Chronotherapy: intuitive, sound, foun-
work of systems and dynamical medicine. ded. . .but not broadly applied. Drugs
Methods Mol Biol 1175:3–17 76:1507–1521
2. Yan Q (2017) Toward dynamical systems med- 9. Yan Q (2012) The role of psychoneuroimmu-
icine: personalized and preventive strategies. nology in personalized and systems medicine.
Pers Med 14:377–380 Methods Mol Biol 934:3–19
3. Goldman AW, Burmeister Y, Cesnulevicius K 10. Blackman MA, Yates JL, Spencer CM et al
et al (2015) Bioregulatory systems medicine: (2014) The Yin and Yang of inflammation.
an innovative approach to integrating the sci- Curr Mol Med 14:1238–1243
ence of molecular networks, inflammation, and 11. Lee SJ, Park SH, Chae H (2016) Biopsycho-
systems biology with the patient’s autoregula- logical structure of Yin-Yang using Cloninger’s
tory capacity? Front Physiol 6:225 temperament model and Carver and White’s
4. Yan Q (2016) Psychoneuroimmunology: sys- BIS/BAS scale. PeerJ 4:e2021
tems biology approaches to mind-body medi- 12. Poupot M, Turrin C-O, Caminade A-M et al
cine. Springer, New York (2016) Poly(phosphorhydrazone) dendrimers:
5. Elisei C, de Castro AP (2017) Insight into role Yin and Yang of monocyte activation for
of microbiota-gut-brain peptides as a target for human NK cell amplification applied to immu-
biotechnology innovations. Front Biosci (Elite notherapy against multiple myeloma. Nanome-
Ed) 9:76–88 dicine 12:2321–2330
6. Vida C, González EM, De la Fuente M (2014) 13. Torterolo P, Chase MH (2014) The hypocre-
Increase of oxidation and inflammation in ner- tins (orexins) mediate the “phasic” compo-
vous and immune systems with aging and anxi- nents of REM sleep: a new hypothesis. Sleep
ety. Curr Pharm Des 20:4656–4678 Sci 7:19–29
7. Sturmberg JP, Bennett JM, Picard M, Seely 14. Kirszenblat L, van Swinderen B (2015) The Yin
AJE (2015) The trajectory of life. Decreasing and Yang of sleep and attention. Trends Neu-
physiological network complexity through rosci 38:776–786
changing fractal patterns. Front Physiol 6:169
Stress and Systemic Inflammation: Yin-Yang Dynamics in Health and Diseases 19
15. Wright KP, Drake AL, Frey DJ et al (2015) 29. Sun Z, Lazar MA (2013) Dissociating fatty
Influence of sleep deprivation and circadian liver and diabetes. Trends Endocrinol Metab
misalignment on cortisol, inflammatory mar- 24:4–12
kers, and cytokine balance. Brain Behav 30. Almanza-Perez JC, Alarcon-Aguilar FJ,
Immun 47:24–34 Blancas-Flores G et al (2010) Glycine regulates
16. Motivala SJ, Tomiyama AJ, Ziegler M et al inflammatory markers modifying the energetic
(2009) Nocturnal levels of ghrelin and leptin balance through PPAR and UCP-2. Biomed
and sleep in chronic insomnia. Psychoneuroen- Pharmacother 64:534–540
docrinology 34:540–545 31. Mysona BA, Matragoon S, Stephens M et al
17. Madonna R, Novo G, Balistreri CR (2016) (2015) Imbalance of the nerve growth factor
Cellular and molecular basis of the imbalance and its precursor as a potential biomarker for
between vascular damage and repair in ageing diabetic retinopathy. Biomed Res Int
and age-related diseases: as biomarkers and tar- 2015:571456
gets for new treatments. Mech Ageing Dev 32. Lieb DC, Parson HK, Mamikunian G, Vinik AI
159:22–30 (2012) Cardiac autonomic imbalance in newly
18. Bai G, Pfaff SL (2011) Protease regulation: the diagnosed and established diabetes is associated
Yin and Yang of neural development and dis- with markers of adipose tissue inflammation.
ease. Neuron 72:9–21 Exp Diabetes Res 2012:878760
19. Jiang J, Dingledine R (2013) Prostaglandin 33. Scherer T, Buettner C (2011) Yin and Yang of
receptor EP2 in the crosshairs of anti- hypothalamic insulin and leptin signaling in
inflammation, anti-cancer, and neuroprotec- regulating white adipose tissue metabolism.
tion. Trends Pharmacol Sci 34:413–423 Rev Endocr Metab Disord 12:235–243
20. Baudry M, Bi X (2016) Calpain-1 and calpain- 34. Berglund LM, Kotova O, Osmark P et al
2: the Yin and Yang of synaptic plasticity and (2012) NFAT regulates the expression of
neurodegeneration. Trends Neurosci AIF-1 and IRT-1: Yin and Yang splice variants
39:235–245 of neointima formation and atherosclerosis.
21. Baker DJ, Jin F, van Deursen JM (2008) The Cardiovasc Res 93:414–423
Yin and Yang of the Cdkn2a locus in senes- 35. Kubli DA, Gustafsson ÅB (2012) Mitochon-
cence and aging. Cell Cycle 7:2795–2802 dria and Mitophagy: the Yin and Yang of cell
22. Czeh M, Gressens P, Kaindl AM (2011) The death control. Circ Res 111:1208–1221
Yin and Yang of microglia. Dev Neurosci 36. Brar SS (2016) The Yin and Yang of long-term
33:199–209 dual antiplatelet therapy. J Am Coll Cardiol
23. Owens R, Grabert K, Davies CL et al (2017) 67:1155–1157
Divergent neuroinflammatory regulation of 37. Fett JD, Ansari AA (2010) Inflammatory mar-
microglial TREM expression and involvement kers and cytokines in peripartum cardiomyopa-
of NF-κB. Front Cell Neurosci 11:56 thy: a delicate balance. Expert Opin Ther
24. Shen Y, Meri S (2003) Yin and Yang: comple- Targets 14:895–898
ment activation and regulation in Alzheimer’s 38. Carmichael ST (2012) Brain excitability in
disease. Prog Neurobiol 70:463–472 stroke. Arch Neurol 69:161–167
25. Cilia R, van Eimeren T (2011) Impulse control 39. Thompson PA, Khatami M, Baglole CJ et al
disorders in Parkinson’s disease: seeking a (2015) Environmental immune disruptors,
roadmap toward a better understanding. inflammation and cancer risk. Carcinogenesis
Brain Struct Funct 216:289–299 36:S232–S253
26. Falk T, Gonzalez RT, Sherman SJ (2010) The 40. Khatami M (2014) Chronic inflammation: syn-
Yin and Yang of VEGF and PEDF: multiface- ergistic interactions of recruiting macrophages
ted neurotrophic factors and their potential in (TAMs) and Eosinophils (Eos) with host mast
the treatment of Parkinson’s disease. Int J Mol cells (MCs) and tumorigenesis in CALTs.
Sci 11:2875–2900 M-CSF, suitable biomarker for cancer diagno-
27. Serviddio G, Romano AD, Greco A et al sis! Cancers (Basel) 6:297–322
(2009) Frailty syndrome is associated with 41. Neufeld TP (2012) Autophagy and cell
altered circulating redox balance and increased growth—the Yin and Yang of nutrient
markers of oxidative stress. Int J Immuno- responses. J Cell Sci 125:2359–2368
pathol Pharmacol 22:819–827 42. Budanov AV (2014) The role of tumor sup-
28. Chen Y-J, Lin C-L, Li C-R et al (2016) Asso- pressor p53 in the antioxidant defense and
ciations among integrated psychoneuroimmu- metabolism. Subcell Biochem 85:337–358
nological factors and metabolic syndrome. 43. Spolski R, Leonard WJ (2008) The Yin and
Psychoneuroendocrinology 74:342–349 Yang of interleukin-21 in allergy,
20 Qing Yan
autoimmunity and cancer. Curr Opin Immunol 47. Tausk F, Elenkov I, Moynihan J (2008)
20:295–301 Psychoneuroimmunology. Dermatol Ther
44. Kong SYJ, Bostick RM, Flanders WD et al 21:22–31
(2014) Oxidative balance score, colorectal ade- 48. Peters EMJ (2012) The neuroendocrine-
noma, and markers of oxidative stress and immune connection regulates chronic inflam-
inflammation. Cancer Epidemiol Biomark matory disease in allergy. Chem Immunol
Prev 23:545–554 Allergy 98:240–252
45. Salazar-Ramiro A, Ramı́rez-Ortega D, Pérez 49. Peters EMJ, Liezmann C, Spatz K et al (2011)
de la Cruz V et al (2016) Role of redox status Nerve growth factor partially recovers inflamed
in development of Glioblastoma. Front Immu- skin from stress-induced worsening in allergic
nol 7:156 inflammation. J Invest Dermatol 131:735–743
46. Peters EMJ (2016) Stressed skin?—a molecular 50. Wang M-J, Xu Y-Y, Huang R-Y et al (2016)
psychosomatic update on stress-causes and Role of an imbalanced miRNAs axis in patho-
effects in dermatologic diseases. J Dtsch Der- genesis of psoriasis: novel perspectives based on
matol Ges 14:233–252; quiz 253 review of the literature. Oncotarget
8:5498–5507
Chapter 2
Abstract
Multidiscipline-based research holds promise toward revealing complex mechanisms that determine health
and disease. For decades, scientists have conducted studies defining the relationships between neuroendo-
crine and immune function culminating into the discipline of psychoneuroimmunology (PNI). In addition,
the discipline of microbial endocrinology has similarly enhanced our understanding of disease processes.
With an increase in genetic-based sequencing technologies, the convergence of neuroendocrine-
immunological-microbial research is expected to significantly further such knowledge needed for medical
discoveries. In this chapter, we provide a review of the current findings that support the conceptual
framework linking microbiota, immunity, and neuroendocrine disciplines.
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_2, © Springer Science+Business Media, LLC, part of Springer Nature 2018
21
22 Colette G. Ngo Ndjom et al.
30
Citations per year
20
10
17
19 5
19 6
87
19 8
89
19 0
19 1
19 2
19 3
94
19 5
19 6
19 7
19 8
20 9
20 0
01
20 2
03
20 4
20 5
06
20 7
08
20 9
20 0
20 1
20 2
13
20 4
20 5
20 6
8
8
9
9
9
9
9
9
9
9
9
0
0
0
0
1
1
1
1
1
1
19
19
19
19
20
20
20
20
20
Year
Fig. 1 Number of citations between 1985 and 2017. Bar graph represents the number of citations found in
Pubmed database (https://www.ncbi.nlm.nih.gov/pubmed/advanced) using key words, “Psychoneuroimmu-
nology” and “Disease”
2 Microbial Endocrinology
The term “stress” has been broadly defined, but its meaning is
largely accepted as the neuronal activation and release of neurohor-
monal intermediaries from the sympathetic (SNS), parasympathetic
(PNS), and autonomic nervous (ANS) systems that mediate host
behavioral and physiological functions.
Exposure to various forms of stress (physical, psychological,
social, and infectious) activates the central nervous system, leading
to a change in neuronal tone of SNS, PNS, and ANS networks.
Dr. Lyte speaks of his experience in addressing the question of why
should one consider neuroendocrine hormones as part of a micro-
bial in vitro culture. His response, “because we do not have tryptic
soy broth and brain heart infusion media flowing through our veins
and arteries and until we use media that reflect the same environ-
ment that bacteria must survive in, then we will never fully under-
stand the mechanisms underlying the ability of infectious agents to
cause disease.” The implications of his response mirror that of PNI
research and reinforce the value of a multidiscipline framework
toward understanding mechanisms related to the pathogenesis of
human disease [6].
Initial studies determining the effects of stress on bacterial
resistance were instrumental in raising the notion of the potential
for neuroendocrine factors to affect bacterial physiology [7]. Such
studies emerged primarily from PNI research which defined stress-
induced influences on immunity as a determinant of susceptibility
Intersections Between Neuroimmune and Microbiota 23
20
0
82
04
05
07
08
09
10
11
12
13
14
15
16
19
20
20
20
20
20
20
20
20
20
20
20
20
Year
Fig. 2 Number of citations between 1982 and 2016. Bar graph represents the
number of citations found in Pubmed database (https://www.ncbi.nlm.nih.gov/
pubmed/advanced) using key words, “Microbiota and Neurohormones”
Catecholamine
Release Scavanges Fe From
Bacterial Species
Fig. 3 Catecholamine’s role in facilitating Fe uptake by bacterial species. Commensals depend on nutrients for
survival. Acquiring iron (Fe) in the blood through high-affinity ferric-binding proteins such as lactoferrin and
transferrin. However, when certain endocrine factors such as norepinephrine are introduced, the iron can
become available to the bacteria. Norepinephrine and other catecholamines can bind to these ferric-binding
proteins resulting in the coordinated reduction of Fe (III) to Fe (II), an iron valency for which the proteins have
low affinity
3.3 Cholinergic The autonomic nervous system (ANS) controls visceral activity
Effects within the body through three main divisions, all of which have
preganglionic neurons in the central nervous system (CNS) that
synapse with ganglionic neurons and that release acetylcholine
28 Colette G. Ngo Ndjom et al.
4.1 Maturation As a first line of defense, the immune system is required to arm the
of the Immune System: host through a balanced mechanism of defining “self” versus “non-
Role of the Microbiota self” antigenic peptides. This complex mechanism requires mutual
mediation with the trillions that inhabit host tissues. This rational is
best demonstrated in experimental studies, utilizing germ-free mice
(e.g., devoid of microbes at birth) as a model. Without a functional
microbiome, the immune system of germ-free mice is severely
immature compared to conventional mice [94, 95]. In the gut,
Intersections Between Neuroimmune and Microbiota 29
A.
Neuroendocrine B.
Response
C.
Bacterial Immune
Response Response
D.
Health
Outcomes
Acknowledgments
References
1. Glaser R, Kiecolt-Glaser J (2005) How stress immunization targeting beta-hemolysin and
damages immune system and health. Discov CAMP factor. J Invest Dermatol 131:401–409
Med 5:165–169 16. Dashper SG, Seers CA, Tan KH, Reynolds EC
2. Kiecolt-Glaser JK (2009) Psychoneuroimmu- (2011) Virulence factors of the oral spirochete
nology: psychology’s gateway to the biomed- Treponema denticola. J Dent Res
ical future. Perspect Psychol Sci 4:367–369 90:691–703
3. Straub RH, Cutolo M (2017) 17. Zhou Y, Lin P, Li Q, Han L, Zheng H, Wei Y,
Psychoneuroimmunology-developments in Cui Z, Ni Y, Guo X (2010) Analysis of the
stress research. Wien Med Wochenschr microbiota of sputum samples from patients
4. Dinan TG, Cryan JF (2017) Microbes, immu- with lower respiratory tract infections. Acta
nity, and behavior: psychoneuroimmunology Biochim Biophys Sin Shanghai 42:754–761
meets the microbiome. Neuropsychopharma- 18. Walk ST, Blum AM, Ewing SA, Weinstock JV,
cology 42:178–192 Young VB (2010) Alteration of the murine
5. Lyte M (2016) Microbial endocrinology: an gut microbiota during infection with the par-
ongoing personal journey. Adv Exp Med Biol asitic helminth Heligmosomoides polygyrus.
874:1–24 Inflamm Bowel Dis 16:1841–1849
6. Freestone PP, Lyte M (2008) Microbial endo- 19. Stecher B, Chaffron S, Kappeli R,
crinology: experimental design issues in the Hapfelmeier S, Freedrich S, Weber TC,
study of interkingdom signalling in infectious Kirundi J, Suar M, McCoy KD, von
disease. Adv Appl Microbiol 64:75–105 Mering C, Macpherson AJ, Hardt WD
7. Lyte M (1993) The role of microbial endocri- (2010) Like will to like: abundances of closely
nology in infectious disease. J Endocrinol related species can predict susceptibility to
137:343–345 intestinal colonization by pathogenic and
commensal bacteria. PLoS Pathog 6:
8. Ashley NT, Demas GE (2017) e1000711
Neuroendocrine-immune circuits, pheno-
types, and interactions. Horm Behav 20. Sokol H, Vasquez N, Hoyeau-Idrissi N,
87:25–34 Seksik P, Beaugerie L, Lavergne-Slove A,
Pochart P, Marteau P (2010) Crypt abscess-
9. Schubert C (2014) Psychoneuroimmunology associated microbiota in inflammatory bowel
of the life span: impact of childhood stress on disease and acute self-limited colitis. World J
immune dysregulation and inflammatory dis- Gastroenterol 16:583–587
ease in later life. Psychother Psychosom Med
Psychol 64:171–180 21. O’Keefe SJ (2010) Tube feeding, the micro-
biota, and Clostridium difficile infection.
10. Freestone P, Lyte M (2010) Stress and micro- World J Gastroenterol 16:139–142
bial endocrinology: prospects for ruminant
nutrition. Animal 4:1248–1257 22. Mallozzi M, Viswanathan VK, Vedantam G
(2010) Spore-forming Bacilli and Clostridia
11. Lyte M (1992) The role of catecholamines in in human disease. Future Microbiol
gram-negative sepsis. Med Hypotheses 5:1109–1123
37:255–258
23. Endt K, Stecher B, Chaffron S, Slack E,
12. Lyte M (2004) Microbial endocrinology and Tchitchek N, Benecke A, Van Maele L, Sirard
infectious disease in the 21st century. Trends JC, Mueller AJ, Heikenwalder M, Macpher-
Microbiol 12:14–20 son AJ, Strugnell R, von Mering C, Hardt
13. Lyte M (2014) The effect of stress on micro- WD (2010) The microbiota mediates patho-
bial growth. Anim Health Res Rev gen clearance from the gut lumen after
15:172–174 non-typhoidal Salmonella diarrhea. PLoS
14. Lofgren JL, Whary MT, Ge Z, Pathog 6:e1001097
Muthupalani S, Taylor NS, Mobley M, 24. Koyanagi T, Sakamoto M, Takeuchi Y,
Potter A, Varro A, Eibach D, Suerbaum S, Ohkuma M, Izumi Y (2010) Analysis of
Wang TC, Fox JG (2011) Lack of commensal microbiota associated with peri-implantitis
flora in Helicobacter pylori-infected INS-GAS using 16S rRNA gene clone library. J Oral
mice reduces gastritis and delays intraepithe- Microbiol 2
lial neoplasia. Gastroenterology 140:210–220 25. Montagner F, Gomes BP, Kumar PS (2010)
15. Lo CW, Lai YK, Liu YT, Gallo RL, Huang CM Molecular fingerprinting reveals the presence
(2011) Staphylococcus aureus hijacks a skin of unique communities associated with paired
commensal to intensify its virulence:
32 Colette G. Ngo Ndjom et al.
samples of root canals and acute apical 36. Everest P (2007) Stress and bacteria: micro-
abscesses. J Endod 36:1475–1479 bial endocrinology. Gut 56:1037–1038
26. Bokulich NA, Rideout JR, Mercurio WG, 37. Moreira CG, Sperandio V (2016) The
Shiffer A, Wolfe B, Maurice CF, Dutton RJ, epinephrine/norepinephrine/autoinducer-3
Turnbaugh PJ, Knight R, Caporaso JG Interkingdom signaling system in Escherichia
(2016) mockrobiota: a public resource for coli O157:H7. Adv Exp Med Biol
microbiome bioinformatics benchmarking. 874:247–261
mSystems 1(5) 38. Dowd SE (2007) Escherichia coli O157:H7
27. Han M, Yang P, Zhou H, Li H, Ning K gene expression in the presence of catechol-
(2016) Metagenomics and single-cell omics amine norepinephrine. FEMS Microbiol Lett
data analysis for human microbiome research. 273:214–223
Adv Exp Med Biol 939:117–137 39. Nakano M, Takahashi A, Sakai Y, Kawano M,
28. Lyte M, Ernst S (1992) Catecholamine Harada N, Mawatari K, Nakaya Y (2007)
induced growth of gram negative bacteria. Catecholamine-induced stimulation of
Life Sci 50:203–212 growth in Vibrio species. Lett Appl Microbiol
29. Asano Y, Hiramoto T, Nishino R, Aiba Y, 44:649–653
Kimura T, Yoshihara K, Koga Y, Sudo N 40. Waldor MK, Sperandio V (2007) Adrenergic
(2012) Critical role of gut microbiota in the regulation of bacterial virulence. J Infect Dis
production of biologically active, free cate- 195:1248–1249
cholamines in the gut lumen of mice. Am J 41. Sandrini S, Alghofaili F, Freestone P, Yesilkaya
Physiol Gastrointest Liver Physiol 303: H (2014) Host stress hormone norepineph-
G1288–G1295 rine stimulates pneumococcal growth, biofilm
30. Mittal R, Debs LH, Patel AP, Nguyen D, formation and virulence gene expression.
Patel K, O’Connor G, Grati M, Mittal J, BMC Microbiol 14:180
Yan D, Eshraghi AA, Deo SK, Daunert S, 42. Gonzales XF, Castillo-Rojas G, Castillo-Rodal
Liu XZ (2017) Neurotransmitters: the critical AI, Tuomanen E, Lopez-Vidal Y (2013) Cat-
modulators regulating gut-brain Axis. J Cell echolamine norepinephrine diminishes lung
Physiol 232:2359–2372 epithelial cell adhesion of Streptococcus
31. Ney DM, Murali SG, Stroup BM, Nair N, Pneumoniae by binding iron. Microbiology
Sawin EA, Rohr F, Levy HL (2017) Metabo- 159:2333–2341
lomic changes demonstrate reduced bioavail- 43. Baldwin HE, Bhatia ND, Friedman A, Eng
ability of tyrosine and altered metabolism of RM, Seite S (2017) The role of cutaneous
tryptophan via the kynurenine pathway with microbiota harmony in maintaining a func-
ingestion of medical foods in phenylketon- tional skin barrier. J Drugs Dermatol
uria. Mol Genet Metab 121:96–103 16:12–18
32. Hegde M, Wood TK, Jayaraman A (2009) 44. Maguire M, Maguire G (2017) The role of
The neuroendocrine hormone norepineph- microbiota, and probiotics and prebiotics in
rine increases Pseudomonas aeruginosa PA14 skin health. Arch Dermatol Res 309:411–421
virulence through the las quorum-sensing 45. Zouboulis CC, Picardo M, Ju Q, Kurokawa I,
pathway. Appl Microbiol Biotechnol Torocsik D, Biro T, Schneider MR (2016)
84:763–776 Beyond acne: current aspects of sebaceous
33. Nietfeld JC, Yeary TJ, Basaraba RJ, Schauen- gland biology and function. Rev Endocr
stein K (1999) Norepinephrine stimulates Metab Disord 17:319–334
in vitro growth but does not increase patho- 46. Ali MF, Soto A, Knoop FC, Conlon JM
genicity of salmonella choleraesuis in an (2001) Antimicrobial peptides isolated from
in vivo model. Adv Exp Med Biol skin secretions of the diploid frog, Xenopus
473:249–260 tropicalis (Pipidae). Biochim Biophys Acta
34. Anderson MT, Armstrong SK (2006) The 1550:81–89
Bordetella bfe system: growth and transcrip- 47. Beasley FC, Marolda CL, Cheung J, Buac S,
tional response to siderophores, catechols, Heinrichs DE (2011) Staphylococcus aureus
and neuroendocrine catecholamines. J Bacter- transporters Hts, Sir, and Sst capture iron
iol 188:5731–5740 liberated from human transferrin by
35. Cogan TA, Thomas AO, Rees LE, Taylor AH, Staphyloferrin A, Staphyloferrin B, and cate-
Jepson MA, Williams PH, Ketley J, Hum- cholamine stress hormones, respectively, and
phrey TJ (2007) Norepinephrine increases contribute to virulence. Infect Immun
the pathogenic potential of campylobacter 79:2345–2355
jejuni. Gut 56:1060–1065
Intersections Between Neuroimmune and Microbiota 33
48. Brotman RM, Ravel J, Cone RA, Zenilman 60. Freestone PP, Haigh RD, Williams PH, Lyte
JM (2010) Rapid fluctuation of the vaginal M (2003) Involvement of enterobactin in
microbiota measured by Gram stain analysis. norepinephrine-mediated iron supply from
Sex Transm Infect 86:297–302 transferrin to enterohaemorrhagic Escherichia
49. Cocco JF, Antonetti JW, Burns JL, Heggers coli. FEMS Microbiol Lett 222:39–43
JP, Blackwell SJ (2010) Characterization of 61. Choi JM, Jo JY, Baik JW, Kim S, Kim CS,
the nasal, sublingual, and oropharyngeal Jeong SM (2017) Risk factors and outcomes
mucosa microbiota in cleft lip and palate indi- associated with a higher use of inotropes in
viduals before and after surgical repair. Cleft kidney transplant recipients. Medicine (Balti-
Palate Craniofac J 47:151–155 more) 96:e5820
50. Fardini Y, Chung P, Dumm R, Joshi N, Han 62. Freestone PP, Al-Dayan N, Lyte M (2016)
YW (2010) Transmission of diverse oral bac- Staphylococci, catecholamine inotropes and
teria to murine placenta: evidence for the oral hospital-acquired infections. Adv Exp Med
microbiome as a potential source of intrauter- Biol 874:183–199
ine infection. Infect Immun 78:1789–1796 63. Lyte M, Ernst S (1993) Alpha and beta adren-
51. Spear GT, Sikaroodi M, Zariffard MR, Landay ergic receptor involvement in catecholamine-
AL, French AL, Gillevet PM (2008) Compar- induced growth of gram-negative bacteria.
ison of the diversity of the vaginal microbiota Biochem Biophys Res Commun
in HIV-infected and HIV-uninfected women 190:447–452
with or without bacterial vaginosis. J Infect 64. Freestone PP, Haigh RD, Lyte M (2007)
Dis 198:1131–1140 Blockade of catecholamine-induced growth
52. Tada A, Hanada N (2010) Opportunistic by adrenergic and dopaminergic receptor
respiratory pathogens in the oral cavity of antagonists in Escherichia coli O157:H7, Sal-
the elderly. FEMS Immunol Med Microbiol monella enterica and Yersinia enterocolitica.
60:1–17 BMC Microbiol 7:8
53. Freestone PP, Sandrini SM, Haigh RD, Lyte 65. Freestone PP, Haigh RD, Williams PH, Lyte
M (2008) Microbial endocrinology: how M (1999) Stimulation of bacterial growth by
stress influences susceptibility to infection. heat-stable, norepinephrine-induced autoin-
Trends Microbiol 16:55–64 ducers. FEMS Microbiol Lett 172:53–60
54. Salton MR (1964) Requirement of dihydrox- 66. Voigt B, Schweder T, Sibbald MJ, Albrecht D,
yphenols for the growth of micrococcus lyso- Ehrenreich A, Bernhardt J, Feesche J, Maurer
deikticus in synthetic media. Biochim Biophys KH, Gottschalk G, van Dijl JM, Hecker M
Acta 86:421–422 (2006) The extracellular proteome of Bacillus
55. Shearer N, Walton NJ (2016) Dietary cate- licheniformis grown in different media and
chols and their relationship to microbial under different nutrient starvation condi-
endocrinology. Adv Exp Med Biol tions. Proteomics 6:268–281
874:101–119 67. Reissbrodt R, Rienaecker I, Romanova JM,
56. Neilands JB (1984) Siderophores of bacteria Freestone PP, Haigh RD, Lyte M,
and fungi. Microbiol Sci 1:9–14 Tschape H, Williams PH (2002) Resuscita-
57. Lee JY, Janes BK, Passalacqua KD, Pfleger BF, tion of salmonella enterica serovar typhimur-
Bergman NH, Liu H, Hakansson K, Somu ium and enterohemorrhagic Escherichia coli
RV, Aldrich CC, Cendrowski S, Hanna PC, from the viable but nonculturable state by
Sherman DH (2007) Biosynthetic analysis of heat-stable enterobacterial autoinducer. Appl
the petrobactin siderophore pathway from Environ Microbiol 68:4788–4794
Bacillus anthracis. J Bacteriol 189:1698–1710 68. Roberts A (2005) Bacteria in the mouth.
58. Koppisch AT, Browder CC, Moe AL, Shelley Dent Update 32:134–136, 139–140, 142
JT, Kinkel BA, Hersman LE, Iyer S, Ruggiero 69. Lyte M, Frank CD, Green BT (1996) Produc-
CE (2005) Petrobactin is the primary side- tion of an autoinducer of growth by norepi-
rophore synthesized by Bacillus anthracis str. nephrine cultured Escherichia coli O157:H7.
Sterne under conditions of iron starvation. FEMS Microbiol Lett 139:155–159
Biometals 18:577–585 70. Simard M, Hill LA, Lewis JG, Hammond GL
59. Hickford SJ, Kupper FC, Zhang G, Carrano (2015) Naturally occurring mutations of
CJ, Blunt JW, Butler A (2004) Petrobactin human corticosteroid-binding globulin. J
sulfonate, a new siderophore produced by Clin Endocrinol Metab 100:E129–E139
the marine bacterium Marinobacter hydrocar- 71. Simard M, Hill LA, Underhill CM, Keller BO,
bonoclasticus. J Nat Prod 67:1897–1899 Villanueva I, Hancock RE, Hammond GL
34 Colette G. Ngo Ndjom et al.
(2014) Pseudomonas aeruginosa elastase dis- activity and lipid peroxidation in the brain
rupts the cortisol-binding activity of and spinal cord of rats infected with Trypano-
corticosteroid-binding globulin. Endocrinol- soma evansi. Vet Parasitol 175:237–244
ogy 155:2900–2908 82. Wolkmer P, da Silva CB, Paim FC, Da Silva
72. Verbrugghe E, Boyen F, Van Parys A, Van AS, Tavares KC, Lazzarotto CR, Palma HE,
Deun K, Croubels S, Thompson A, Thome GR, Miletti LC, Schetinger MR,
Shearer N, Leyman B, Haesebrouck F, Pas- Lopes ST, Mazzanti CM (2012) Biochemistry
mans F (2011) Stress induced Salmonella detection of acetylcholinesterase activity in
Typhimurium recrudescence in pigs coincides Trypanosoma evansi and possible functional
with cortisol induced increased intracellular correlations. Exp Parasitol 132:546–549
proliferation in macrophages. Vet Res 42:118 83. Bolino CM, Bercik P (2010) Pathogenic fac-
73. Morris DJ, Ridlon JM (2017) Glucocorti- tors involved in the development of irritable
coids and gut bacteria: “the GALF hypothe- bowel syndrome: focus on a microbial role.
sis” in the metagenomic era. Steroids Infect Dis Clin N Am 24:961–975, ix
125:1–13 84. Craig OF, Quigley EM (2010) Bacteria,
74. Ngo Ndjom CG, Kantor LV, Jones HP genetics and irritable bowel syndrome. Expert
(2017) CRH affects the phenotypic expres- Rev Gastroenterol Hepatol 4:271–276
sion of sepsis-associated virulence factors by 85. Fox JG, Feng Y, Theve EJ, Raczynski AR,
streptococcus pneumoniae serotype 1 in vitro. Fiala JL, Doernte AL, Williams M, McFaline
Front Cell Infect Microbiol 7:263 JL, Essigmann JM, Schauer DB, Tannenbaum
75. Weinstein LI, Revuelta A, Pando RH (2015) SR, Dedon PC, Weinman SA, Lemon SM, Fry
Catecholamines and acetylcholine are key reg- RC, Rogers AB (2010) Gut microbes define
ulators of the interaction between microbes liver cancer risk in mice exposed to chemical
and the immune system. Ann N Y Acad Sci and viral transgenic hepatocarcinogens. Gut
1351:39–51 59:88–97
76. Stanaszek PM, Snell JF, O’Neill JJ (1977) 86. Fujimura KE, Slusher NA, Cabana MD,
Isolation, extraction, and measurement of Lynch SV (2010) Role of the gut microbiota
acetylcholine from Lactobacillus plantarum. in defining human health. Expert Rev Anti-
Appl Environ Microbiol 34:237–239 Infect Ther 8:435–454
77. Pandey S, Sree A, Sethi DP, Kumar CG, 87. Ganal-Vonarburg SC, Fuhrer T, Gomez de
Kakollu S, Chowdhury L, Dash SS (2014) A Aguero M (2017) Maternal microbiota and
marine sponge associated strain of Bacillus antibodies as advocates of neonatal health.
subtilis and other marine bacteria can produce Gut Microbes 8(5):479–485
anticholinesterase compounds. Microb Cell 88. Smith PD, Smythies LE, Shen R, Greenwell-
Factories 13:24 Wild T, Gliozzi M, Wahl SM (2011) Intestinal
78. Horiuchi Y, Kimura R, Kato N, Fujii T, macrophages and response to microbial
Seki M, Endo T, Kato T, Kawashima K encroachment. Mucosal Immunol 4:31–42
(2003) Evolutional study on acetylcholine 89. Okada H, Kuhn C, Feillet H, Bach JF (2010)
expression. Life Sci 72:1745–1756 The ‘hygiene hypothesis’ for autoimmune and
79. Yamada T, Fujii T, Kanai T, Amo T, allergic diseases: an update. Clin Exp Immu-
Imanaka T, Nishimasu H, Wakagi T, nol 160:1–9
Shoun H, Kamekura M, Kamagata Y, 90. Martin R, Nauta AJ, Ben Amor K, Knippels
Kato T, Kawashima K (2005) Expression of LM, Knol J, Garssen J (2010) Early life: gut
acetylcholine (ACh) and ACh-synthesizing microbiota and immune development in
activity in Archaea. Life Sci 77:1935–1944 infancy. Benef Microbes 1:367–382
80. Costa MM, Silva AS, Paim FC, Franca R, Dor- 91. Belkaid Y, Harrison OJ (2017) Homeostatic
nelles GL, Thome GR, Serres JD, Schmatz R, immunity and the microbiota. Immunity
Spanevello RM, Goncalves JF, Schetinger 46:562–576
MR, Mazzanti CM, Lopes ST, Monteiro SG 92. Luo A, Leach ST, Barres R, Hesson LB,
(2012) Cholinesterase as inflammatory mar- Grimm MC, Simar D (2017) The microbiota
kers in a experimental infection by Trypano- and epigenetic regulation of T helper 17/reg-
soma evansi in rabbits. An Acad Bras Cienc ulatory T cells: in search of a balanced
84:1105–1113 immune system. Front Immunol 8:417
81. da Silva AS, Monteiro SG, Goncalves JF, 93. Lloyd CM, Marsland BJ (2017) Lung homeo-
Spanevello R, Oliveira CB, Costa MM, Jaques stasis: influence of age, microbes, and the
JA, Morsch VM, Schetinger MR, Mazzanti immune system. Immunity 46:549–561
CM, Lopes ST (2011) Acetylcholinesterase
Intersections Between Neuroimmune and Microbiota 35
94. Costa MC, Santos JR, Ribeiro MJ, Freitas GJ, 98. Chu H, Mazmanian SK (2013) Innate
Bastos RW, Ferreira GF, Miranda AS, Arifa immune recognition of the microbiota pro-
RD, Santos PC, Martins Fdos S, Paixao TA, motes host-microbial symbiosis. Nat Immu-
Teixeira AL, Souza DG, Santos DA (2016) nol 14:668–675
The absence of microbiota delays the inflam- 99. McDermott AJ, Huffnagle GB (2014) The
matory response to Cryptococcus gattii. Int J microbiome and regulation of mucosal immu-
Med Microbiol 306:187–195 nity. Immunology 142:24–31
95. Chung H, Pamp SJ, Hill JA, Surana NK, 100. Hansen JD, Vojtech LN, Laing KJ (2011)
Edelman SM, Troy EB, Reading NC, Villa- Sensing disease and danger: a survey of verte-
blanca EJ, Wang S, Mora JR, Umesaki Y, brate PRRs and their origins. Dev Comp
Mathis D, Benoist C, Relman DA, Kasper Immunol 35:886–897
DL (2012) Gut immune maturation depends 101. Salzman NH (2011) Microbiota-immune sys-
on colonization with a host-specific micro- tem interaction: an uneasy alliance. Curr Opin
biota. Cell 149:1578–1593 Microbiol 14:99–105
96. Fransen F, van Beek AA, Borghuis T, 102. Brown RL, Clarke TB (2017) The regulation
Meijer B, Hugenholtz F, van der Gaast-de of host defences to infection by the micro-
Jongh C, Savelkoul HF, de Jonge MI, Faas biota. Immunology 150:1–6
MM, Boekschoten MV, Smidt H, El Aidy S,
de Vos P (2017) The impact of gut microbiota 103. Bailey MT (2016) Psychological stress,
on gender-specific differences in immunity. immunity, and the effects on indigenous
Front Immunol 8:754 microflora. Adv Exp Med Biol 874:225–246
97. Geva-Zatorsky N, Sefik E, Kua L, Pasman L, 104. Bailey MT, Dowd SE, Parry NM, Galley JD,
Tan TG, Ortiz-Lopez A, Yanortsang TB, Schauer DB, Lyte M (2010) Stressor expo-
Yang L, Jupp R, Mathis D, Benoist C, Kasper sure disrupts commensal microbial popula-
DL (2017) Mining the human gut microbiota tions in the intestines and leads to increased
for immunomodulatory organisms. Cell 168 colonization by Citrobacter rodentium.
(928–943):e911 Infect Immun 78:1509–1519
Chapter 3
Abstract
Accumulating evidence has made clear that experience—the knowledge an individual acquires during a
lifetime of sensing and acting—is of fundamental biological relevance. Experience makes an impact on all
adaptive systems, including the endocrine, immune, and nerve systems, and is of the essence, not only for
the unfolding of an organisms’ healthy status, but also for the development of malfunctional traits.
Nevertheless, experience is often excluded from empirical approaches. A variety of complex interactions
that influence life histories are thereby neglected. Such ignorance is especially detrimental for psychoneu-
roimmunology, the science that seeks to understand how the exquisite and dynamic interplay between
mind, body, and environment relates to behavioral characteristics. This chapter reviews claims for incorpor-
ating experience as a member of good explanatory standing in biology and medicine, and more specifically
claims that experiential knowledge is required to enable meaningful and relevant explanations and predic-
tions in the psychoneuroimmunological realm.
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_3, © Springer Science+Business Media, LLC, part of Springer Nature 2018
37
38 Elling Ulvestad
the experiential dimension and thus lose touch with the very indi-
viduals they represent.
Even a science that acknowledges the experiential dimension
and thus attempts to incorporate experiential parameters in the
form of major life events—e.g., divorce, death of a spouse, or
serious disease—often fails the task. As highlighted in a critical
review, individuals rarely apply the same meaning to a life event
[5]. Different subjects interpret events differently, and this differ-
ential interpretation is a determinant of how individuals respond
psychoendocrinologically. The “objective” characteristics of an
event, which are foundational in group-based investigations, are
thus not objective in a strict sense—they are rather interpreted in an
idiosyncratic manner by each different participator.
Knowledge of group characteristics is highly relevant but nev-
ertheless insufficient for understanding the individual, and it thus
appears that science needs a theory of the organism that also allows
for the emergence of “private” responses. This does not, however,
imply that science should become subjective and so comply with
the slogan “anything goes.” Science’s ultimate task is to give objec-
tive accounts of nature, also of individual experience. To stay true to
its ideal, science should therefore give well-grounded accounts of
subjective experience. Subjective accounts of experience, although
important for individual behavior, are nevertheless, at least for the
time being, outside the scientific realm.
In the following I provide an exploration of the challenge posed
by Plaut and Friedman. In so doing I will invoke the age-old
dilemma between the one and the many, and investigate how this
applies to the objective perspective taken by an external observer
and to the subjective perspective of a participant. The importance
of the subject’s environment, exemplified by the microbial com-
munities of our intestines, will also be highlighted, as will the role of
subjective interpretations of environmental stimuli over the life
cycle. These deliberations will hopefully reveal the complexity of
the experiential challenges facing psychoneuroimmunology.
Fig. 1 Each individual is the result of the lineage’s adaptive processes during evolutionary time and adaptive
processes during developmental time. These two adaptive processes are integrated in every organism. The
concave lens depicts the evolutionary resources of the organism, and includes its genetic makeup. The convex
lens, in contrast, depicts the developmentally shaped resources of the organism, including the epigenetic
makeup and the wirings of the central nervous system, the immune system, and the endocrine system. The
umwelt makes up the context in which the experiences make their impact. Dependent upon how these
systems integrate within the organism, the response may turn out as functional or dysfunctional. And, since
the organism’s responses are in many ways instructive of latter responses, the responses are depicted to feed
back to alter both the umwelt and latter responses to the same or dissimilar challenges
Human beings come into the world with naturally selected coping
mechanisms. And since these mechanisms have evolutionary pre-
conditions, it is a task for science to ask whether or not such
preconditions interfere with man’s perception of the umwelt in a
The Experiential Dimension 43
true manner. There are two reasons why this question is impor-
tant—first, a science that aims to understand humans has to know
how humans experience the world, and second, a science that aims
to understand the world must have an idea of science’s own foun-
dations for knowledge acquisition. Exactly such preconditions and
preconceptions have been given critical attention by researchers in
the phenomenological tradition [20] and their investigations are
thus in many ways supplementary to the Uexküllian tradition.
The phenomenological tradition has made huge efforts to
understand the human experiential dimension and thus make it
accessible to investigation. As the phenomenologist sees it, any
biological individual accesses the umwelt through a first-person
perspective. For humans, this is the world as they know it, imbued
with meaning and emotions. Any human being has access to a wide
range of historically situated knowledge that helps him to respond
to external challenges in a meaningful way. The knowledge of each
generation is different, and knowledge also differs in different parts
of the world. Such knowledge is therefore spatiotemporally
restricted.
But human beings can also access knowledge that is true irre-
spective of time and place. To obtain such valued knowledge, man
has to “bracket” his first-person perspective on the world. He has to
take a God-like perspective, be the spectator who takes a view from
nowhere. The third-person perspective, which is the foundational
view of science, is not easily achieved. To reach the goal of true
knowledge, scientists have to act as disinterested, emotionless, and
neutral observers, and so have to undergo a long and arduous
training to achieve control over their inborn perceptual capabilities.
This is a complicated task, as they have to erase some of their
developmentally learned presuppositions.
The degree to which the third-person perspective can be
achieved varies widely between the sciences. While mathematicians
and logicians can be trained to master their subject in a true
“disinterested” manner, it is more questionable whether biologists,
social scientists, and humanists can achieve the same degree of
perfection. And for a simple reason—biological entities, including
human beings, are historically situated; they have a history that
matters as to what and who they are. That biological entities,
including the nervous, immune, and endocrine systems, have a
history does not, however, imply that they elude investigation by
a science with universalizing ambitions. But it does imply that
science should make more precise which aspects of the historical
entities it can reach firm conclusions about and which it cannot.
Although an arduous task, especially since science constantly
develops new concepts and exploratory technologies that push the
line of demarcation between knowledge and ignorance, it does
appear evident, at least for the time being, that science cannot
reach the innermost experiences of an individual. Science can for
44 Elling Ulvestad
example explore the general effects of major life events, but how
each individual experiences a divorce or the loss of a beloved one is
a private matter. It thus appears necessary to make an analytic
distinction between a public first-person perspective, which is ame-
nable for scientific investigations, and a private first-person perspec-
tive which is not. The private perspective is a specific characteristic
of each individual, be it a human being or a perceptual system. And
as such, it has no characteristics that can be generalized—it is thus
located beyond the realms of science. The public first-person per-
spective is, on the other hand, accessible from the outside. It
includes perceptual traits that are specific for a given species, and
is as such co-extensional with the animal’s umwelt.
While the public first-person aspect can be made explicit by
means of genetic, environmental, and developmental investiga-
tions, the private first-person aspect is an experiential dimension
and as such not accessible for scientific investigation. Such experi-
ence does not lend itself easily to standardized interpretation; it is
always an experience of something for someone, in a unique con-
text. There is thus a gap between what science can explain and what
it cannot—and the gap goes straight through the individual,
between the private and public aspects of the first person. None-
theless, exactly where the line of accessibility should be drawn is a
matter on which science should have a saying. The explanatory gap
should be made more precise, but prospects for its closing are for
the time being dim.
colonize our mucous membranes and skin have made evident that it
is no longer possible to conceptualize microbes as simply “exter-
nal.” They are internal and cooperative as well. The sequencing of
the human genome made clear that our chromosomes are teemed
with microbially derived elements. The genome consists of 45% of
transposons—DNA sequences that are able to copy and move
within chromosomes—of which approximately 8% are retrovirus-
like [21]. Some of these retroviral integrations have been of great
importance for vertebrate physiological development. Although
most transposons that accumulate in the genome have no known
function, they contribute a large potential substrate for the evolu-
tion and development of regulatory networks [22, 23].
The genome also contains bacterially derived DNA, some of
which regulates the interaction between the eukaryotic cells and
their bacterially derived mitochondrial symbionts. The mitochon-
dria, which have evolved to become an integral part of the host’s
cells, have transferred some of their genes to the cell’s nucleus. And
in so doing, they lost the ability to reproduce freely. This loss has,
however, been matched by a comparative gain in survival capacity—
mitochondria, by their very location, have become shielded off
from immune destruction. The importance of keeping the interac-
tion between the eukaryotic cell and its mitochondria tightly regu-
lated is dramatically spelled out during debilitating physical trauma
in which mitochondria relocate or become destroyed. This leads to
a breakdown of the conditions for cooperation between the host
cell and the symbiont, and the host may thus develop a dangerous
systemic inflammatory response. The response includes fever, low
blood pressure, and increased heart rate [24], and is thus analogous
to the inflammatory process observed as a result of contaminating
bacteria during sepsis.
Another surprising observation that came out of the sequenc-
ing of the human genome was the relative paucity of genes. Based
on complexity estimates, man was thought to have about 100,000
genes prior to the sequencing. But only about 20,000 genes were
detected. Man was as complex as before, so how could the com-
plexity be accounted for by so few genes? One answer has to do
with the way the DNA is used for making proteins and regulatory
factors, and it has turned out that this process is far more efficient
than first thought [25]. But this is not the whole story; additional
data have since revealed that humans also have access to a plethora
of genes not coded for in the genome. And these genes, which are
located within bacteria and viruses on the skin and the mucous
membranes, by far outnumber the genes in the cellular nucleus.
Estimates have indicated that an adult human being can be
described as a superorganism consisting of 50% prokaryotic and
50% eukaryotic cells. Since every bacterium may be infected with as
many as 100 bacteriophages, thus giving an estimate of ten billion
46 Elling Ulvestad
8 Summing Up
References
1. Ader R (1981) Psychoneuroimmunology. Aca- 4. Wittgenstein L (1960) The blue and brown
demic Press, New York books. Harper & Row, New York
2. Good RA (1981) Foreword: interactions of the 5. Biondi M, Picardi A (1999) Psychological
body’s major networks. In: Ader R stress and neuroendocrine function in humans:
(ed) Psychoneuroimmunology. Academic the last two decades of research. Psychother
Press, New York, pp xvii–xxix Psychosom 68:114–150
3. Plaut SM, Friedman SB (1981) Psychosocial 6. Kull K (2001) Jakob von Uexküll: an introduc-
factors in infectious disease. In: Ader R tion. Semiotica 134:1–59
(ed) Psychoneuroimmunology. Academic 7. Uexküll J (1982) The theory of meaning
Press, New York, pp 3–30 (Original 1940). Semiotica 42:25–82
52 Elling Ulvestad
8. Ulvestad E (2007) Defending life. The nature 25. Sholtis SJ, Noonan JP (2010) Gene regulation
of host-parasite relations. Springer, Dordrecht and the origins of human biological unique-
9. Hoffmeyer J (2008) Biosemiotics. An exami- ness. Trends Genet 26:110–118
nation into the signs of life and the life of signs. 26. Dethlefsen L, McFall-Ngai M, Relman DA
University of Scranton Press, Scranton (2007) An ecological and evolutionary per-
10. Hill K, Kaplan H (1999) Life history traits in spective on human-microbe mutualism and
humans: theory and empiricial studies. Annu disease. Nature 449:811–818
Rev Anthropol 28:397–430 27. Pennisi E (2011) Going viral: exploring the
11. Craig AD (2009) How do you feel—now? The role of viruses in our bodies. Science 331:1513
anterior insula and human awareness. Nat Rev 28. Spor A, Koren O, Ley R (2011) Unravelling
Neurosci 10:59–70 the effects of the environment and host geno-
12. Craig AD (2004) Human feelings: why are type on the gut microbiome. Nat Rev Micro-
some more aware than others? Trends Cogn biol 9:279–290
Sci 8:239–241 29. Turnbaugh PJ, Hamady M, Yatsunenko T et al
13. Frith C (2008) Making up the mind. How the (2009) A core gut microbiome in obese and
brain creates our mental world. Blackwell Pub- lean twins. Nature 457:480–484
lishing, Oxford 30. Reyes A, Haynes M, Hanson N et al (2010)
14. Odling-Smee FJ, Laland KN, Feldman MW Viruses in the faecal microbiota of monozy-
(2003) Niche construction. The neglected gotic twins and their mothers. Nature
process in evolution. Princeton University 466:334–338
Press, Princeton 31. Gonzalez A, Clemente JC, Shade A et al
15. Sterelny K (2007) Social intelligence, human (2011) Our microbial selves: what ecology
intelligence and niche construction. Philos can teach us. EMBO Rep 12:775–784
Trans R Soc Lond Ser B Biol Sci 362:719–730 32. Nicholson JK, Holmes E, Wilson ID (2005)
16. Laland KN, Odling-Smee J, Myles S (2010) Gut microorganisms, mammalian metabolism
How culture shaped the human genome: and personalized health care. Nat Rev Micro-
bringing genetics and the human sciences biol 3:431–438
together. Nat Rev Genet 11:137–148 33. Musso G, Gambino R, Cassader M (2011)
17. Beauregard M (2007) Mind does really matter: Interactions between gut microbiota and host
evidence from neuroimaging studies of emo- metabolism predisposing to obesity and diabe-
tional self-regulation, psychotherapy, and pla- tes. Annu Rev Med 62:361–380
cebo effect. Prog Neurobiol 81:218–236 34. Ivanov II, Littman DR (2011) Modulation of
18. Getz L, Kirkengen AL, Ulvestad E (2011) The immune homeostasis by commensal bacteria.
human biology—saturated with experience. J Curr Opin Microbiol 14:106–114
Nor Med Assoc 131:683–687 35. Lee YK, Mazmanian SK (2010) Has the micro-
19. Henrich J, Heine SJ, Norenzayan A (2010) biota played a critical role in the evolution of
The weirdest people in the world? Behav the adaptive immune system? Science
Brain Sci 33:61–83 330:1768–1773
20. Gallagher S, Zahavi D (2008) The phenome- 36. Heijtz RD, Wang S, Anuar F et al (2011) Nor-
nological mind. An introduction to philosophy mal gut microbiota modulates brain develop-
of mind and cognitive science. Routledge, ment and behavior. Proc Natl Acad Sci U S A
London 108:3047–3052
21. International human genome consortium 37. Boulanger LM (2009) Immune proteins in
(2001) Initial sequencing and analysis of the brain development and synaptic plasticity. Neu-
human genome. Nature 409:860–921 ron 64:93–109
22. Kurth R, Bannert N (2010) Beneficial and det- 38. Van Der Zanden EP, Boeckxstaens GE, de
rimental effects of human endogenous retro- Jonge WJ (2009) The vagus nerve as a modu-
viruses. Int J Cancer 126:306–314 lator of intestinal inflammation. Neurogas-
23. Venner S, Feschotte C, Biémont C (2009) troenterol Motil 21:6–17
Dynamics of transposable elements: towards a 39. Tracey KJ (2010) Understanding immunity
community ecology of the genome. Trends requires more than immunology. Nat Immunol
Genet 25:317–323 11:561–564
24. Zhang Q, Raoof M, Chen Y et al (2010) Cir- 40. Gareau MG, Silva MA, Perdue MH (2008)
culating mitochondrial DAMPs cause inflam- Pathophysiological mechanisms of stress-
matory responses to injury. Nature induced intestinal damage. Curr Mol Med
464:104–107 8:274–281
The Experiential Dimension 53
41. Söderholm JD, Yang PC, Ceponis P et al cancer and the prospects for the human race.
(2002) Chronic stress induces mast cell- Princeton University Press, Princeton
dependent bacterial adherence and initiates 53. Gluckman P, Hanson M (2005) The fetal
mucosal inflammation in rat intestine. Gastro- matrix: evolution, development and disease.
enterology 123:1099–1108 Cambridge University Press, Cambridge
42. Padgett DA, Glaser R (2003) How stress influ- 54. Osmond C, Barker DJ (2000) Fetal, infant,
ences the immune response. Trends Immunol and childhood growth are predictors of coro-
24:444–448 nary heart disease, diabetes, and hypertension
43. Li J, Cowden LG, King JD et al (2007) Effects in adult men and women. Environ Health Per-
of chronic stress and interleukin-10 gene poly- spect 108(Suppl 3):545–553
morphisms on antibody response to tetanus 55. World Health Organization (2008) The global
vaccine in family caregivers of patients with burden of disease 2004 update. WHO Press,
Alzheimer’s disease. Psychosom Med Geneva
69:551–559 56. Troen BR (2003) The biology of aging. Mt
44. McEwen BS, Gianaros PJ (2011) Stress- and Sinai J Med 70:3–22
allostasis-induced brain plasticity. Annu Rev 57. Williams GC (1957) Pleiotropy, natural selec-
Med 62:431–445 tion, and the evolution of senescence. Evolu-
45. Hamburger V (1980) Embryology and the tion 11:398–411
modern synthesis in evolutionary theory. In: 58. De S (2011) Somatic mosaicism in healthy
Mayr E, Provine WB (eds) The evolutionary human tissues. Trends Genet 27:217–223
synthesis. Harvard University Press, Cam-
bridge, pp 96–112 59. Pennington R, Gatenbee C, Kennedy B et al
(2009) Group differences in proneness to
46. Mayr E (1961) Cause and effect in biology. inflammation. Infect Genet Evol 9:1371–1380
Science 134:1501–1506
60. Licastro F, Candore G, Lio D et al (2005)
47. Reik W (2007) Stability and flexibility of epige- Innate immunity and inflammation in ageing:
netic gene regulation in mammalian develop- a key for understanding age-related diseases.
ment. Nature 447:425–432 Immun Ageing 2:8
48. Fraga MF, Ballestar E, Paz MF et al (2005) 61. Shonkoff JP, Boyce WT, McEwen BS (2009)
Epigenetic differences arise during the lifetime Neuroscience, molecular biology, and the
of monozygotic twins. Proc Natl Acad Sci U S childhood roots of health disparities: building
A 102:10604–10609 a new framework for health promotion and
49. Rook GA (2010) 99th Dahlem conference on disease prevention. JAMA 301:2252–2259
infection, inflammation and chronic inflamma- 62. Danese A, Pariante CM, Caspi A et al (2007)
tory disorders: darwinian medicine and the Childhood maltreatment predicts adult inflam-
‘hygiene’ or ‘old friends’ hypothesis. Clin Exp mation in a life-course study. Proc Natl Acad
Immunol 160:70–79 Sci U S A 104:1319–1324
50. Okada H, Kuhn C, Feillet H et al (2010) The 63. Dube SR, Fairweather D, Pearson WS et al
‘hygiene hypothesis’ for autoimmune and aller- (2009) Cumulative childhood stress and auto-
gic diseases: an update. Clin Exp Immunol immune diseases in adults. Psychosom Med
160:1–9 71:243–250
51. Fitzsimmons CM, Dunne DW (2009) Survival 64. Pawelec G, Akbar A, Caruso C et al (2004) Is
of the fittest: allergology or parasitology? immunosenescence infectious? Trends Immu-
Trends Parasitol 25:447–451 nol 25:406–410
52. Cairns J (1997) Matters of life and death. Per- 65. Miller RA (1996) The aging immune system:
spectives on public health, molecular biology, primer and prospectus. Science 273:70–74
Chapter 4
Abstract
There is considerable research interest overlap between biological anthropology and psychoneuroimmu-
nology (PNI), particularly given recent anthropological interest in endocrine and immune system func-
tioning over the life span and in different environmental contexts. In this chapter, I argue that conducting
research on non-WEIRD populations and applying an anthropological, evolutionary approach to PNI can
greatly strengthen our understanding of immune-endocrine-behavior connections. This chapter reviews
population-level variation in the human immune and endocrine systems, as well as genetic and environ-
mental contributions to this variation. The effects of culture on shaping health outcomes and stress
responses are also considered. Finally, this chapter discusses some noninvasive sampling methodologies
appropriate to field research and alternatives to laboratory-based research designs. By confronting variable
social and environmental contexts, PNI can greatly expand on its existing contributions to the treatment
and understanding of depression, mood disorders, stress, and other aspects of health and well-being.
Key words Psychoneuroimmunology, Life history theory, Human biological variation, Ecoimmunol-
ogy, Hormones, Stress, Culture, Methods
1 Introduction
Over the past three decades or so, researchers working under the
umbrella of psychoneuroimmunology have made many fundamen-
tal contributions to our understanding of basic human physiology
and psychology. Foremost of these is the realization that the
immune and endocrine systems regularly cross-communicate, and
the discovery of the bidirectional effects of mood on immune
function. Although these advances have reshaped how we think
about immunity and stress, it must be acknowledged that this
research has almost uniformly been conducted in Western countries
utilizing Western study participants. This is, of course, a matter of
convenience and all discoveries must begin somewhere before find-
ings can be generalized across a range of conditions. Yet what can
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_4, © Springer Science+Business Media, LLC, part of Springer Nature 2018
55
56 Eric C. Shattuck
lungs [7], while populations that have lived at high altitude for
generations, such as native Tibetans, show evidence of natural
selection on genes, including EPAS1 and PPARA, that affect eryth-
ropoiesis and vasculogenesis, among other effects [8].
Similar evolutionary logic has been applied to PNI as well.
Chronic inflammatory diseases have been attributed to a “mis-
match” between ancestral human states and current lifestyles [9];
depression has been put forth as an evolved mechanism to limit
contact with, and improve recovery from, infectious disease [10];
stress reactions, via glucocorticoids, drive leukocyte migration to
peripheral tissues in anticipation of wounding and subsequent
pathogen contact [11]. What is often missing from this adaptation-
ist approach is a consideration of short-term, variable ecological
contexts.
Life history theory, a currently booming topic in biological
anthropology and ecoimmunology, provides just such a theoretical
framework. A major focus of life history theory is an individual’s
partitioning of limited resources (i.e., time and energy) into
growth, somatic maintenance (including immune function), stor-
age, or reproduction. Resources allocated to one category cannot
be used in another which gives rise to a number of life history trade-
offs, such as stunted growth following infection due to temporary
overinvestment in immune responses [12]. Optimal life history
strategies to maximize reproductive fitness are numerous and
highly context dependent, varying based on life stage and local
ecologies. Anthropology adds another constraint to this model,
namely culture [13], which sets boundaries for appropriate beha-
viors. With this perspective, human behavior is seen as tremen-
dously variable across both time and space, which correctly
implies that findings from one population (or even at different
time points in the same population) are not likely to be relevant
to others. This simple truth encourages researchers to not overgen-
eralize, though this is unfortunately sometimes forgotten. It is
certainly true that many biological traits have evolved to meet a
universal need, but the assumption that these traits operate the
same everywhere should be questioned, particularly in the case of
those traits that have significant effects on health or are intimately
involved with internalizing external environmental cues, as is the
case for both the endocrine and immune systems.
4.1 Population With the advent of cheaper and less technologically burdensome
Differences endocrinological assays such as enzyme immunoassays (EIAs), and
in Hormone Levels particularly with the development and validation of salivary EIAs,
biological anthropologists have cemented their interests in hor-
mones as a mechanistic link between health outcomes and several
related domains, particularly energetics and growth/development.
A critically important anthropological contribution to the wider
field of endocrinology has been the discovery of population-
dependent variation in hormone levels and the exploration of pos-
sible determinants of this variation.
Ecological Context and Human Variation: Applying the Principles of. . . 59
4.3 Sources One possible explanation for population differences is genetic dif-
of Variation: ferences. Single-nucleotide polymorphisms in cytokine promoter
Population Genetics regions such as IL6-174G and TNF-308A have been shown to
increase cytokine concentrations (for extensive review of cytokine
Ecological Context and Human Variation: Applying the Principles of. . . 61
4.4 Sources Both nutritional status and pathogen exposure during childhood
of Variation: can have significant effects on the development and action of several
Nutritional Status important physiological systems, including the endocrine and
and Pathogen immune systems. In addition to negative effects on growth, high
Exposure pathogenic environments during childhood may “prime” the
immune system differently. Temporary fasting results in a dysfunc-
tional HPA stress response that can be counteracted with glucose
replacement [48], for instance, and protein-energy malnutrition
62 Eric C. Shattuck
cortisol levels after the Trier social stress test than their Canadian
counterparts, although a number of different factors, including
socioeconomic status, could contribute to this finding [69]. Fur-
thermore, perceived racial discrimination during adolescence pre-
dicted flatter diurnal cortisol slopes in both African-American and
Caucasian-American adults, as well as lower waking and average
cortisol levels in African-American adults only [70].
5.2 Cultural “Health,” “disease,” and allied concepts are known to vary between
Differences groups as well. Although they were speaking specifically about
in Conceptions mental illness, Rao and colleagues note that diagnoses are made
of Health based on deviations from sociocultural and behavioral norms
[71]. The same reasoning could apply to other health states as
well. Physical illness or disability is defined in relation to normative
states that certainly differ between individuals; cultural norms and
practices can shape what is defined as “normal.”
Different conceptions of sickness and health may translate into
different subjective illness experiences and, subsequently, variation
in “objective” clinical measures or other patient assessment out-
comes. For instance, there is a considerable literature on ethnic
differences in pain responses. A recent meta-analysis found that
African-American individuals reported lower cold pain thresholds
and tolerances, as well as higher cold pain intensity than Caucasians
[72]. African-Americans also exhibited lower thresholds and toler-
ances for pain elicited through pressure and electrical stimuli.
Biological mechanisms that may account for these differences
include genetic [73, 74] and hormonal [75] contributions to pain
tolerance and regulation. There is also a considerable role for
sociocultural and psychological factors in shaping pain responses.
These include spiritual or religious beliefs, socioeconomic status,
socialization/learned responses, mood, coping strategies, and the
like [72]. Some groups, including Mexicans, Mexican-Americans,
and Quichuas, traditionally value stoicism in the face of pain
[76, 77]. Other factors, such as the belief that illness or pain is a
test from God and must therefore be endured, have been shown to
affect treatment seeking [78].
6 Field Methods
6.1 Sample As mentioned above, biological assay technologies have been devel-
Collection Outside oping rapidly, making research on hormones and immunological
the Lab factors far more feasible than in the past. Although blood (i.e.,
plasma or serum) is still the gold standard for biological samples,
it is now possible to quantify many biomarkers in less invasive
matrices, such as saliva and urine. This has several practical benefits,
including ease of collection, reduced supplies cost, increased par-
ticipant compliance, and reduced exposure to potentially
66 Eric C. Shattuck
References
1. McDade TW (2001) Lifestyle incongruity, Applied evolutionary anthropology: Darwin-
social integration, and immune function in ian approaches to contemporary world issues.
Samoan adolescents. Soc Sci Med Springer, New York, pp 1–11
53:1351–1362 14. Henrich J, Heine SJ, Norenzayan A (2010)
2. McDade TW, Stallings JF, Worthman CM The weirdest people in the world? Behav
(2000) Culture change and stress in Western Brain Sci 33:61–83
Samoan youth: methodological issues in the 15. Gurven M, Stieglitz J, Trumble B et al (2017)
cross-cultural study of stress and immune The Tsimane health and life history project:
function. Am J Hum Biol 12:792–802 integrating anthropology and biomedicine.
3. Demas GE, Carlton ED (2015) Ecoimmu- Evol Anthropol 26:54–73
nology for psychoneuroimmunologists: con- 16. Marlowe FW (2010) The Hadza: hunter-
sidering context in neuroendocrine-immune- gatherers of Tanzania. University of California
behavior interactions. Brain Behav Immun Press, Berkeley
44:9–16 17. Blurton Jones N (2016) Demography and
4. Segerstrom SC (2010) Resources, stress, and evolutionary ecology of Hadza hunter-
immunity: an ecological perspective on gatherers. Cambridge University Press,
human psychoneuroimmunology. Ann Behav Cambridge
Med 40:114–125 18. Hill K, Hurtado AM (1996) Ache life history:
5. Boas F (1912) Changes in the bodily form of the ecology and demography of a foraging
the descendants of immigrants. Am Anthro- people. Aldine de Gruyter, New York
pol 14:530–562 19. Flinn MV (2008) Why words can hurt us:
6. Little M (2010) History of the study of social relationships, stress, and health. In:
human biology. In: Muehlenbein M Trevathan W, Smith EO, McKenna J (eds)
(ed) Human evolutionary biology. Cam- Evolutionary medicine and health. Oxford
bridge University Press, New York, pp 29–48 University Press, Oxford, pp 247–258
7. Frisancho AR (1977) Developmental adapta- 20. Stieglitz J, Trumble BC, Thompson ME et al
tion to high altitude hypoxia. Int J Biome- (2015) Depression as sickness behavior? A test
teorol 21:135–146 of the host defense hypothesis in a high path-
8. Simonson TS, McClain DA, Jorde LB et al ogen population. Brain Behav Immun
(2012) Genetic determinants of Tibetan 49:130–139
high-altitude adaptation. Hum Genet 21. Blackwell AD, Trumble BC, Maldonado
131:527–533 Suarez I et al (2016) Immune function in
9. Straub RH, Del Rey A, Besedovsky HO Amazonian horticulturalists. Ann Hum Biol
(2007) Emerging concepts for the pathogen- 43:382–396
esis of chronic disabling inflammatory dis- 22. Blackwell AD, Gurven MD, Sugiyama LS et al
eases: neuroendocrine-immune interactions (2011) Evidence for a peak shift in a humoral
and evolutionary biology. In: Ader R response to helminths: age profiles of IgE in
(ed) Psychoneuroimmunology. Elsevier, San the Shuar of Ecuador, the Tsimane of Bolivia,
Diego, CA, pp 217–232 and the US NHANES. PLoS Negl Trop Dis
10. Anders S, Tanaka M, Kinney DK (2013) 5:e1218
Depression as an evolutionary strategy for 23. Flinn MV, England BG (2003) Childhood
defense against infection. Brain Behav stress: endocrine and immune responses to
Immun 31:9–22 psychosocial events. In: Wilce JM (ed) Social
11. Dhabhar FS (2002) Stress-induced augmen- and cultural lives of immune systems. Routle-
tation of immune function—the role of stress dge Press, London, pp 107–147
hormones, leukocyte trafficking, and cyto- 24. Ellison PT, Bribiescas RG, Bentley GR et al
kines. Brain Behav Immun 16:785–798 (2002) Population variation in age-related
12. McDade TW, Reyes Garcı́a V, Tanner S et al decline in male salivary testosterone. Hum
(2008) Maintenance versus growth: investi- Reprod 17:3251–3253
gating the costs of immune activation among 25. Rohrmann S, Nelson WG, Rifai N et al
children in lowland Bolivia. Am J Phys (2007) Serum estrogen, but not testosterone,
Anthropol 136:478–484 levels differ between black and white men in a
13. Gibson MA, Lawson DW (2014) Applying nationally representative sample of Americans.
evolutionary anthropology to a changing J Clin Endocrinol Metab 92:2519–2525
world. In: Gibson MA, Lawson DW (eds)
Ecological Context and Human Variation: Applying the Principles of. . . 73
26. Lopez DS, Peskoe SB, Joshu CE et al (2013) 38. Doyle WJ, Casselbrant ML, Li-Korotky H-S
Racial/ethnic differences in serum sex steroid et al (2010) The interleukin 6 -174 C/C
hormone concentrations in US adolescent genotype predicts greater rhinovirus illness. J
males. Cancer Causes Control 24:817–826 Infect Dis 201:199–206
27. Carlton ED, Demas GE (2014) Leptin med- 39. Bull SJ, Huezo-Diaz P, Binder EB et al (2008)
iates seasonal variation in some but not all Functional polymorphisms in the interleukin-
symptoms of sickness in Siberian hamsters. 6 and serotonin transporter genes, and
Horm Behav 66:802–811 depression and fatigue induced by inter-
28. Lilja M, Rolandsson O, Shaw JE et al (2010) feron-α and ribavirin treatment. Mol Psychia-
Higher leptin levels in Asian Indians than try 14:1095–1104
Creoles and Europids: a potential explanation 40. Piraino B, Vollmer-Conna U, Lloyd AR
for increased metabolic risk. Int J Obes Relat (2012) Genetic associations of fatigue and
Metab Disord 34:878–885 other symptom domains of the acute sickness
29. Bribiescas RG (2005) Serum leptin levels in response to infection. Brain Behav Immun
Ache Amerindian females with normal adipos- 26:552–558
ity are not significantly different from Ameri- 41. Vollmer-Conna U, Piraino BF, Cameron B
can anorexia nervosa patients. Am J Hum Biol et al (2008) Cytokine polymorphisms have a
17:207–210 synergistic effect on severity of the acute sick-
30. Baerwald CG, Mok CC, Fife MS et al (1999) ness response to infection. Clin Infect Dis
Distribution of corticotropin-releasing hor- 47:1418–1425
mone promoter polymorphism in different 42. Berger FG (2004) The interleukin-6 gene: a
ethnic groups: evidence for natural selection susceptibility factor that may contribute to
in human populations. Immunogenetics racial and ethnic disparities in breast cancer
49:894–899 mortality. Breast Cancer Res Treat
31. Kelley-Hedgepeth A, Lloyd-Jones DM, Col- 88:281–285
vin A et al (2008) Ethnic differences in 43. Shattuck EC, Muehlenbein MP (2015)
C-reactive protein concentrations. Clin Human sickness behavior: ultimate and prox-
Chem 54:1027–1037 imate explanations. Am J Phys Anthropol
32. Gurven M, Kaplan H, Crimmins E et al 157:1–18
(2008) Lifetime inflammation in two epide- 44. Russcher H, Smit P, van den Akker ELT et al
miological worlds: the Tsimane of Bolivia and (2005) Two polymorphisms in the glucocor-
the United States. J Gerontol A Biol Sci Med ticoid receptor gene directly affect
Sci 63A:196–199 glucocorticoid-regulated gene expression. J
33. Dujuardi Y, Sartono E, Wibowo H et al Clin Endocrinol Metab 90:5804–5810
(2010) A longitudinal study of BCG vaccina- 45. Russcher H, van Rossum EFC, de Jong FH
tion in early childhood: the development of et al (2005) Increased expression of the glu-
innate and adaptive immune responses. PLoS cocorticoid receptor-a translational isoform as
One 5:e14066 a result of the ER22/23EK polymorphism.
34. Lalor MK, Floyd S, Gorak-Stolinska P et al Mol Endocrinol 19:1687–1696
(2011) BCG vaccination induces different 46. Derijk RH, De Kloet ER (2008) Corticoste-
cytokine profiles following infant BCG vacci- roid receptor polymorphisms: determinants
nation in the UK and Malawi. J Infect Dis of vulnerability and resilience. Eur J Pharma-
204:1075–1085 col 583:303–311
35. Kollmann TR (2013) Variation between 47. Derijk RH (2009) Single nucleotide poly-
populations in the innate immune response morphisms related to HPA axis reactivity.
to vaccine adjuvants. Front Immunol 4:81 Neuroimmunomodulation 16:340–352
36. Bucasas KL, Franco LM, Shaw CA et al 48. Rohleder N, Kirschbaum C (2007) Effects of
(2011) Early patterns of gene expression cor- nutrition on neuro-endocrine stress
relate with the humoral immune response to responses. Curr Opin Clin Nutr Metab Care
influenza vaccination in humans. J Infect Dis 10:504–510
203:921–929 49. Ritz BW, Gardner EM (2006) Malnutrition
37. Gallagher G, Eskdale J, Bidwell JL (2003) and energy restriction differentially affect viral
Cytokine genetics: polymorphisms, functional immunity. J Nutr 136:1141–1144
variations, and disease associations. In: 50. Gurven M, Kaplan H, Winking J et al (2009)
Thomas AW, Lotze MT (eds) The cytokine Inflammation and infection do not promote
handbook. Academic Press, London, pp arterial aging and cardiovascular disease risk
19–57
74 Eric C. Shattuck
factors among lean horticulturalists. PLoS Americans: is there evidence for differential
One 4:e6590 item functioning of depressive symptoms?
51. McDade TW, Rutherford J, Adair L et al Psychol Med 44:2567–2578
(2010) Early origins of inflammation: micro- 64. Abdullah T, Brown TL (2011) Mental illness
bial exposures in infancy predict lower levels stigma and ethnocultural beliefs, values, and
of C-reactive protein in adulthood. Proc Biol norms: an integrative review. Clin Psychol Rev
Sci 277:1129–1137 31:934–948
52. Lisciandro JG, Prescott SL, Nadal-Sims MG 65. Gravlee CC, Dressler WW (2005) Skin pig-
et al (2012) Neonatal antigen-presenting cells mentation, self-perceived color, and arterial
are functionally more quiescent in children blood pressure in Puerto Rico. Am J Hum
born under traditional compared with mod- Biol 17:195–206
ern environmental conditions. J Allergy Clin 66. Gravlee CC, Non AL, Mulligan CJ (2009)
Immunol 130:1167–1174 Genetic ancestry, social classification, and
53. McDade TW, Hoke M, Borja JB et al (2013) racial inequalities in blood pressure in South-
Do environments in infancy moderate the eastern Puerto Rico. PLoS One 4:e6821
association between stress and inflammation 67. Lewis M, Ramsay DS, Kawakami K (1993)
in adulthood? Initial evidence from a birth Differences between Japanese infants and
cohort in the Philippines. Brain Behav Caucasian American infants in behavioral and
Immun 31:23–30 cortisol response to inoculation. Child Dev
54. McDade TW, Borja JB, Adair LS et al (2012) 64:1722–1731
Depressive symptoms are not associated with 68. Doan SN, Tardif T, Miller A et al (2017)
inflammation in younger and older adults in Consequences of ‘tiger’ parenting: a cross-
the Philippines. Evol Med Public Health cultural study of maternal psychological con-
2013:18–23 trol and children’s cortisol stress response.
55. Lochmiller RL, Deerenberg C (2000) Trade- Dev Sci 20:e12404
offs in evolutionary immunology: just what is 69. Souza-Talarico JN, Plusquellec P, Lupien SJ
the cost of immunity? Oikos 88:87–98 et al (2014) Cross-country differences in basal
56. Cordain L, Eaton SB, Sebastian A et al (2005) and stress-induced cortisol secretion in older
Origins and evolution of the Western diet: adults. PLoS One 9:e105968
health implications for the 21st century. Am 70. Adam EK, Heissel JA, Zeiders KH et al
J Clin Nutr 81:341–354 (2015) Developmental histories of perceived
57. Maes M, Christophe A, Bosmans E et al racial discrimination and diurnal cortisol pro-
(2000) In humans, serum polyunsaturated files in adulthood: a 20-year prospective study.
fatty acid levels predict the response of proin- Psychoneuroendocrinology 62:279–291
flammatory cytokines to psychologic stress. 71. Rao D, Feinglass J, Corrigan P (2007) Racial
Biol Psychiatry 47:910–920 and ethnic disparities in mental illness stigma.
58. Kiecolt-Glaser JK, Belury MA, Porter K et al J Nerv Ment Dis 195:734–744
(2007) Depressive symptoms, omega-6: 72. Rahim-Williams B, Riley JL 3rd, Williams
omega-3 fatty acids, and inflammation in AKK et al (2012) A quantitative review of
older adults. Psychosom Med 69:217–224 ethnic group differences in experimental pain
59. Fabris N, Mocchegiani E, Muzzioli M et al response: do biology, psychology, and culture
(1991) The role of zinc in neuroendocrine- matter? Pain Med 13:522–540
immune interactions during aging. Ann N Y 73. Fillingim RB, Kaplan L, Staud R et al (2005)
Acad Sci 621:314–326 The A118G single nucleotide polymorphism
60. Samartın S, Chandra RK (2001) Obesity, of the mu-opioid receptor gene (OPRM1) is
overnutrition and the immune system. Nutr associated with pressure pain sensitivity in
Res 21:243–262 humans. J Pain 6:159–167
61. Singer MK, Dressler W, George S et al (2016) 74. Nielsen CS, Stubhaug A, Price DD et al
Culture: the missing link in health research. (2008) Individual differences in pain sensitiv-
Soc Sci Med 170:237–246 ity: genetic and environmental contributions.
62. Deisenhammer EA, Çoban-Başaran M, Man- Pain 136:21–29
tar A et al (2011) Ethnic and migrational 75. Mechlin B, Morrow AL, Maixner W et al
impact on the clinical manifestation of depres- (2005) African Americans show alterations in
sion. Soc Psychiatry Psychiatr Epidemiol endogenous pain regulatory mechanisms and
47:1121–1129 reduced pain tolerance to experimental pain
63. Kalibatseva Z, Leong FTL, Ham EH (2014) procedures. Psychosom Med 62:517–523
A symptom profile of depression among Asian
Ecological Context and Human Variation: Applying the Principles of. . . 75
76. Calvillo ER (2013) Insights on the pain expe- 89. Lima DP, Diniz DG, Moimaz SAS et al
rience in Mexican Americans. In: (2010) Saliva: reflection of the body. Int J
Incayawar M, Todd K (eds) Culture, brain, Infect Dis 14:e184–e188
and analgesia. Oxford University Press, 90. Ryan D, Robards K, Prenzler PD et al (2011)
New York, pp 49–61 Recent and potential developments in the
77. Incayawar M, Maldonado-Bouchard S (2013) analysis of urine: a review. Anal Chim Acta
We feel pain too: asserting the pain experience 684:17–29
of the Quichua people. In: Incayawar M, 91. Meyer JS, Novak MA (2012) Hair cortisol: a
Todd K (eds) Culture, brain, and analgesia. novel biomarker of hypothalamic-pituitary-
Oxford University Press, New York, pp 61–75 adrenocortical activity. Endocrinology
78. Gonzalez-Swafford MJ, Gutierrez MG 153:4120–4127
(1983) Ethno-medical beliefs and practices 92. Wosu AC, Valdimarsdóttir U, Shields AE et al
of Mexican-Americans. Nurse Pract 8:29–34 (2013) Correlates of cortisol in human hair:
79. Pacheco CM, Daley SM, Brown T et al (2013) implications for epidemiologic studies on
Moving forward: breaking the cycle of mis- health effects of chronic stress. Ann Epide-
trust between American Indians and research- miol 23:797–811
ers. Am J Public Health 103:2152–2159 93. Gao W, Stalder T, Foley P et al (2013) Quan-
80. Rathnayake N, Åkerman S, Klinge B et al titative analysis of steroid hormones in human
(2013) Salivary biomarkers of oral health: a hair using a column-switching LC–AP-
cross-sectional study. J Clin Periodontol CI–MS/MS assay. J Chromatogr B 928:1–8
40:140–147 94. Stalder T, Kirschbaum C (2012) Analysis of
81. Teles RP, Likhari V, Socransky SS et al (2009) cortisol in hair: state of the art and future
Salivary cytokine levels in chronic periodonti- directions. Brain Behav Immun
tis and periodontally health subjects. A cross- 26:1019–1029
sectional study. J Periodontal Res 95. Gao W, Kirschbaum C, Grass J et al (2016)
44:411–417 LCMS based analysis of endogenous steroid
82. McCullough ME, Churchland PS, Mendez hormones in human hair. J Steroid Biochem
AJ (2013) Problems with measuring periph- Mol Biol 162:92–99
eral oxytocin: can the data on oxytocin and 96. Yang HZ, Lan J, Meng YJ et al (1998) A
human behavior be trusted? Neurosci Biobe- preliminary study of steroid reproductive hor-
hav Rev 37:1485–1492 mones in human hair. J Steroid Biochem Mol
83. McDade TW (2013) Development and vali- Biol 67:447–450
dation of assay protocols for use with dried 97. Cohen S, Doyle WJ, Skoner DP et al (1997)
blood spot samples. Am J Hum Biol 26:1–9 Social ties and susceptibility to the common
84. McDade TW, Williams S, Snodgrass JJ (2007) cold. JAMA 277:1940–1944
What a drop can do: dried blood spots as a 98. Cohen S, Doyle WJ, Turner R et al (2003)
minimally invasive method for integrating Sociability and susceptibility to the common
biomarkers into population-based research. cold. Psychol Sci 14:389–395
Demography 44:899–925 99. Cohen S, Doyle WJ, Skoner DP (1999) Psy-
85. Chiappin S, Antonelli G, Gatti R et al (2007) chological stress, cytokine production, and
Saliva specimen: a new laboratory tool for severity of upper respiratory illness. Psycho-
diagnostic and basic investigation. Clin Chim som Med 61:175–180
Acta 383:30–40 100. Cohen S (2004) Childhood socioeconomic
86. Kirschbaum C, Hellhammer DH (1994) Sali- status and host resistance to infectious illness
vary cortisol in psychoneuroendocrine in adulthood. Psychosom Med 66:553–558
research: recent developments and applica- 101. Phillips AC (2012) The vaccination model in
tions. Psychoneuroendocrinology 19:313–333 psychoneuroimmunology research: a review.
87. Muehlenbein MP, Prall SP, Chester E (2011) In: Yan Q (ed) Psychoneuroimmunology:
Development of a noninvasive salivary mea- methods and protocols. Humana Press,
sure of functional immunity in humans. Am J Totowa, NJ, pp 355–370
Hum Biol 23:267–268 102. Williamson S, Munro C, Pickler R et al
88. Demas GE, Zysling DA, Beechler BR et al (2012) Comparison of biomarkers in blood
(2011) Beyond phytohaemagglutinin: asses- and saliva in health adults. Nurs Res Pract
sing vertebrate immune function across eco- 2012:246178. https://doi.org/10.1155/
logical contexts. J Anim Ecol 80:710–730 2012/246178
76 Eric C. Shattuck
103. Lim PW, Garssen J, Sandalova E (2016) 115. Charteris-Black J, Seale C (2009) Men and
Potential use of salivary markers for longitu- emotion talk: evidence from the experience of
dinal monitoring of inflammatory immune illness. Gender Lang 3:81–113
responses to vaccination. Mediat Inflamm 116. Pletzer B, Petasis O, Ortner TM et al (2015)
2016:6958293 Interactive effects of culture and sex hor-
104. Shattuck EC, Muehlenbein MP (2017) mones on the sex role self-concept. Front
Human sickness behavior not expressed in Neurosci 9:240
response to the rabies vaccine. Am J Phys 117. Willer R, Rogalin CL, Conlon B et al (2013)
Anthropol 162(S64):294 Overdoing gender: a test of the masculine
105. McDade TW, Borja JB, Kuzawa CW et al overcompensation thesis. Am J Sociol
(2015) C-reactive protein response to influ- 118:980–1022
enza vaccination as a model of mild inflamma- 118. Garcı́a AR, Gurven M, Blackwell AD (2017)
tory stimulation in the Philippines. Vaccine A matter of perception: perceived socio-
33:2004–2008 economic status and cortisol on the island of
106. Muehlenbein MP, Alger J, Cogswell F et al Utila, Honduras. Am J Hum Biol 34:e23031
(2005) The reproductive endocrine response 119. Sorensen MV, Snodgrass JJ, Leonard WR et al
to Plasmodium vivax infection in Hondurans. (2009) Lifestyle incongruity, stress and
Am J Trop Med Hyg 73:178–187 immune function in indigenous Siberians:
107. Gravlee CC (2005) Ethnic classification in the health impacts of rapid social and eco-
Southeastern Puerto Rico: the cultural nomic change. Am J Phys Anthropol
model of “color”. Soc Forces 83:949–970 138:62–69
108. Alcántara C, Chen C-N, Alegrı́a M (2014) Do 120. Heald AH, Patel J, Anderson SG et al (2007)
post-migration perceptions of social mobility Migration is associated with lower total, but
matter for Latino immigrant health? Soc Sci not free testosterone levels in South Asian
Med 101:94–106 men. Clin Endocrinol 67:651–655
109. Cohen S, Alper CM, Doyle WJ et al (2008) 121. Lagana K (2003) Come bien, camina y no se
Objective and subjective socioeconomic sta- preocupe—eat right, walk, and do not worry:
tus and susceptibility to the common cold. selective biculturalism during pregnancy in a
Health Psychol 27:268–274 Mexican American community. J Transcult
110. Sturgeon JA, Arewasikporn A, Okun MA et al Nurs 14:117–124
(2016) The psychosocial context of financial 122. Rosenberg N, Daviglus ML, DeVon HA et al
stress. Psychosom Med 78:134–143 (2017) The association between parity and
111. Liu B (2017) Many facets of sentiment analy- inflammation among Mexican-American
sis. In: Cambria E, Das D, Bandyopadhyay S, women of reproductive age varies by accultur-
Feraco A (eds) A practical guide to sentiment ation level: results of the National Health and
analysis. Springer, New York, pp 11–41 Nutrition Examination Survey (1999–2006).
112. Pennebaker JW, Mehl MR, Niederhoffer KG Womens Health Issues 27:485–492
(2003) Psychological aspects of natural lan- 123. Fleuriet KJ, Sunil TS (2014) Perceived social
guage use: our words, our selves. Annu Rev stress, pregnancy-related anxiety, depression
Psychol 54:547–577 and subjective social status among pregnant
113. Dunn KC, Neumann IB (2016) Undertaking Mexican and Mexican American women in
discourse analysis for social research. Univer- South Texas. J Health Care Poor Under-
sity of Michigan Press, Ann Arbor, MI served 25:546–561
114. Courtenay WH (2000) Constructions of 124. Garruto RM, Little MA, James GD et al
masculinity and their influence on men’s (1999) Natural experimental models: the
well-being: a theory of gender and health. global search for biomedical paradigms
Soc Sci Med 50:1385–1401 among traditional, modernizing, and modern
populations. Proc Natl Acad Sci U S A
96:10536–10543
Chapter 5
Abstract
Evidences from psychoneuroimmunology (PNI) and systems biology studies support a conceptual frame-
work of “Yin-Yang dynamics” for understanding the “whole mind-body system.” The Yin-Yang dynamical
balances in the stress response networks may be critical for health and diseases, especially mental health and
psychiatric disorders. Specifically, the neuroimmune imbalances have been found as the important features
and potential biomarkers of stress, anxiety, depression, and systemic inflammation. At the system levels,
factors such as psychosocial stress and obesity, especially a leaky gut, may result in the imbalance between
regulatory and proinflammatory T cells. At the molecular and cellular levels, the imbalances in multiple
networks including the cytokine and redox pathways, immune-kynurenine networks, HPA axis, and
synaptic plasticity in the hypothalamus are the key factors in depression. The recognition of the neuroim-
mune imbalances and the restoration of the Yin-Yang dynamical balances need to become a high priority
toward the development of dynamical systems medicine for psychiatric diseases including depression and
schizophrenia.
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_5, © Springer Science+Business Media, LLC, part of Springer Nature 2018
77
78 Qing Yan
Table 1
Yin-Yang dynamics and neuroimmune imbalances in psychiatric disorders
2.2 The Imbalances Multiple pathways have been associated with the development of
in the Immune- depression, especially those of monoamine metabolism and neuro-
Kynurenine Networks endocrine functions. For example, the immune-kynurenine net-
works have the essential roles in the communications between
stress and neuroimmune systems in MDD ([3, 22]; also see
Table 1).
Chronic stress may elevate proinflammatory cytokine levels and
activities of IDO. The imbalances in the downstream kynurenine
pathway may have neurotoxic results in the brain with damages in
the glial-neuronal network, leading to higher risks of depression
[22]. As the important connections among stress, inflammation,
and depression, the kynurenine networks have been proposed as
the potential antidepressant therapeutic targets [3, 23].
In addition to stressful events, sleep disturbance has been
closely related to depression. Studies have connected sleep distur-
bance with higher levels of the inflammatory marker C-reactive
protein (CRP) and lower ratios of kynurenic acid (KynA)/quino-
linic acid (QA) among patients with depression [24]. With the
82 Qing Yan
2.3 The Imbalances T cells have the key roles in maintaining normal behaviors and
in the T-Cell Functions immune homeostasis. The imbalance between adaptive and mal-
adaptive functions of T cells, together with the imbalance between
neurodegenerative and regenerative repair activities, has been iden-
tified as the features of the neuroimmune pathways in depression
([31]; also see Table 1).
Specifically, peripheral naı̈ve T cells are involved in the regula-
tion of neural plasticity. T lymphocytes have the protective roles
against maladaptive behavioral responses linked to depression
[31]. However, psychogenic stress may affect the T cells’ activities,
and alter the productions of neurotrophic factors and depression-
Neuroimmune Imbalances and Yin-Yang Dynamics in Stress, Anxiety, and Depression 83
3 Conclusion
References
1. Yan Q (2014) From pharmacogenomics and related to interferon and interleukins treat-
systems biology to personalized care: a frame- ment. Metab Brain Dis 24:55–68
work of systems and dynamical medicine. 14. Haroon E, Raison CL, Miller AH (2012)
Methods Mol Biol 1175:3–17 Psychoneuroimmunology meets neuropsycho-
2. Yan Q (2017) Toward dynamical systems med- pharmacology: translational implications of the
icine: personalized and preventive strategies. impact of inflammation on behavior. Neurop-
Pers Med 14:377–380 sychopharmacology 37:137–162
3. Yan Q (2016) Psychoneuroimmunology: sys- 15. Grimm S, Beck J, Schuepbach D et al (2008)
tems biology approaches to mind-body medi- Imbalance between left and right dorsolateral
cine. Springer, New York prefrontal cortex in major depression is linked
4. Peters EMJ, Müller Y, Snaga W et al (2017) to negative emotional judgment: an fMRI
Hair and stress: a pilot study of hair and cyto- study in severe major depressive disorder. Biol
kine balance alteration in healthy young Psychiatry 63:369–376
women under major exam stress. PLoS One 16. Szelényi J, Vizi ES (2007) The catecholamine
12:e0175904 cytokine balance: interaction between the brain
5. Head KA, Kelly GS (2009) Nutrients and bota- and the immune system. Ann N Y Acad Sci
nicals for treatment of stress: adrenal fatigue, 1113:311–324
neurotransmitter imbalance, anxiety, and rest- 17. Ge J-F, Qi C-C, Zhou J-N (2013) Imbalance
less sleep. Altern Med Rev 14:114–140 of leptin pathway and hypothalamus synaptic
6. Hou R, Garner M, Holmes C et al (2017) plasticity markers are associated with stress-
Peripheral inflammatory cytokines and induced depression in rats. Behav Brain Res
immune balance in generalised anxiety disor- 249:38–43
der: case-controlled study. Brain Behav Immun 18. Bardelli C, Amoruso A, Manzetti E et al (2013)
62:212–218 Recurrent major depressive disorder: imbal-
7. Salim S, Chugh G, Asghar M (2012) Inflam- ance of neurokinin (NK)-1 and NK-2 receptor
mation in anxiety. Adv Protein Chem Struct expression in monocytes. Pharmacol Res
Biol 88:1–25 68:24–30
8. Han HW, Ohn JH, Moon J, Kim JH (2013) 19. Neumann ID, Landgraf R (2012) Balance of
Yin and Yang of disease genes and death genes brain oxytocin and vasopressin: implications for
between reciprocally scale-free biological net- anxiety, depression, and social behaviors.
works. Nucleic Acids Res 41:9209–9217 Trends Neurosci 35:649–659
9. R K, D M A, C N, S N W, C D (2014) Oxida- 20. Depciuch J, Sowa-Kućma M, Nowak G et al
tive imbalance and anxiety disorders. Curr (2017) The role of zinc deficiency-induced
Neuropharmacol 12:193–204 changes in the phospholipid-protein balance
10. Kara MI, Yanık S, Keskinruzgar A et al (2012) of blood serum in animal depression model by
Oxidative imbalance and anxiety in patients Raman, FTIR and UV-Vis spectroscopy.
with sleep bruxism. Oral Surg Oral Med Oral Biomed Pharmacother 89:549–558
Pathol Oral Radiol 114:604–609 21. Monteagudo PT, Falcão AA, Verreschi ITN,
11. Wood SK (2014) Cardiac autonomic imbal- Zanella M-T (2016) The imbalance of
ance by social stress in rodents: understanding sex-hormones related to depressive symptoms
putative biomarkers. Front Psychol 5:950 in obese men. Aging Male 19:20–26
12. Stein MB, Paulus MP (2009) Imbalance of 22. Won E, Kim Y-K (2016) Stress, the autonomic
approach and avoidance: the Yin and Yang of nervous system, and the immune-kynurenine
anxiety disorders. Biol Psychiatry pathway in the etiology of depression. Curr
66:1072–1074 Neuropharmacol 14:665–673
13. Myint AM, Schwarz MJ, Steinbusch HWM, 23. Christmas DM, Potokar J, Davies SJ (2011) A
Leonard BE (2009) Neuropsychiatric disorders biological pathway linking inflammation and
Neuroimmune Imbalances and Yin-Yang Dynamics in Stress, Anxiety, and Depression 85
Abstract
The brain maintains homeostasis in part through a network of feedback and feed-forward mechanisms,
where neurochemicals and immune markers act as mediators. Using a previously constructed model of
biobehavioral feedback, we found that in addition to healthy equilibrium another stable regulatory program
supported chronic depression and anxiety. Exploring mechanisms that might underlie the contributions of
subjective well-being to improved therapeutic outcomes in depression, we iteratively screened 288 candi-
date feedback patterns linking well-being to molecular signaling networks for those that maintained the
original homeostatic regimes. Simulating stressful trigger events on each candidate network while main-
taining high levels of subjective well-being isolated a specific feedback network where well-being was
promoted by dopamine and acetylcholine, and itself promoted norepinephrine while inhibiting cortisol
expression. This biobehavioral feedback mechanism was especially effective in reproducing well-being’s
clinically documented ability to promote resilience and protect against onset of depression and anxiety.
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_6, © Springer Science+Business Media, LLC, part of Springer Nature 2018
87
88 J. Tory Toole et al.
positively with improved immune function [2, 3]. Its protective and
therapeutic benefits are numerous, and an empirically validated
treatment manual for well-being therapy was produced within the
last year [4]. As can often be the case, interventions based on
subjective well-being have found their way into clinical practice
yet little is known of the underlying physiological mechanisms
involved in its regulation.
In this work, we construct a causal model from literature
describing regulatory associations linking key neurotransmitters,
hormones, and behavioral constructs. We use this model as a basis
for reverse engineering the possible biobehavioral mechanisms
involved in regulating subjective well-being by working back from
known outcomes such as the documented reduction in recidivism
rate in depression [2]. In the current study, we consider subjective
well-being as a single entity for the sake of simplicity, and iteratively
propose stimulatory and inhibitory control circuitry linking subjec-
tive well-being to a model neurotransmission network. Each set of
proposed mechanisms is then tested for its ability to instill increased
resilience to the onset of depression through repeated simulations.
This computer-based exploration of mechanisms supporting the
role of subjective well-being in depression suggests that the latter
may interact directly with norepinephrine, cortisol, acetylcholine,
and dopamine as part of a biobehavioral regulatory loop.
2 Materials
3 Methods
The analysis that serves as the focus of this chapter is divided into
three discrete steps. The first step (Subheading 3.1) involves apply-
ing a discrete ternary logic analysis to the biobehavioral regulatory
model. Using this logic to dictate allowable departures from a given
state, a Monte Carlo simulation framework is applied to determine
the homeostatic steady states of the system (Subheading 3.2). In an
attempt at reverse engineering documented clinical outcomes of
improved subject well-being, augmented models are created which
include a subjective well-being node. In each of these models,
candidate mechanisms are proposed whereby subjective well-
being directly modulates two molecular targets and in turn is
directly promoted or inhibited by two members of this same molec-
ular network as part of a biobehavioral feedback loop. The behavior
supported by each of these candidate regulatory feedback loops is
simulated to determine (1) which possible connection patterns
continue to support established steady states, and (2) which of
these also reproduce the observed protective effects of increasing
subjective well-being (Subheading 3.3).
3.1 Discrete Ternary The discrete ternary logical network analysis used in the present
Logical Signal work is an extension of a methodology proposed by Mendoza and
Processing Xenarios (2006) [53] and Thomas (1991) [54], and has been
reported previously by our group [55, 56]. We encode documented
feedback mechanisms within the neuropsychological system using
only the direction (source and target) and type (activator or inhibi-
tor) of interaction. As data describing the relative magnitude of
cellular responses of these various signals remain limited, we con-
sider model components to be equally sensitive to all signals.
Computational Discovery of Resilience Mechanisms in Depression 91
Fig. 1 A basic logic model of behavior and neurotransmission. A simple causal regulatory network model
linking 15 soluble mediators in the brain and associated behavioral constructs informed by published literature
where green connectors with arrowhead terminators indicate stimulatory actions whereas red connectors with
dot terminators indicate inhibitory actions. Also illustrated is a potential biobehavioral feedback circuit derived
from simulations of documented therapeutic outcomes and describing the regulatory interactions of subjective
well-being and neuro-signaling
92 J. Tory Toole et al.
3.2 Using a Monte Evolution of the system through the sequence of state transitions
Carlo Environment supported by the model is simulated by applying a Monte Carlo
to Simulate Response algorithm. From any initial starting state, allowable state transitions
are determined based on Eq. (2). Applying the latter to each
variable in the model for the mth state of the system, ~ x m ðt Þ, defines
m th
the image vector ~ x ðt þ 1Þ for the m state. With ~ x m ðt þ 1Þ
defined, the system is updated asynchronously following the
generalized logical analysis of [54] Thomas (1991). According to
this method the ith variable of the mth state vector x im ðt Þ is moved
one step toward its preferred image x im ðt þ 1Þ (e.g., If x im ðt Þ ¼ 1
and x im ðt þ 1Þ ¼ 1, then x im ðt þ 1Þ is set to 0). Thus, for each
current state of the system there are potentially several subsequent
states toward which it may asynchronously evolve. From the allow-
able transitions a target state is chosen at random using a uniform
equal distribution then used to generate the next set of allowable
target states. Executing the simulation multiple times gives a distri-
bution of paths that is used to determine the behavior of the system
from any given start state. For additional details the reader is
referred to our previous work [55, 56].
3.3 Identifying As described above, homeostatic states are defined as those states to
Stable Regulatory which the system naturally returns following a perturbation. Even
Behaviors simple biological circuits often support more than one stable equilib-
rium state. These were determined here by enumerating all possible
states that may be occupied by the model shown in Fig. 1. In the
current example, this exhaustive search pointed to two steady homeo-
static states (SS) where neurotransmitter and psychobehavioral con-
structs converged to the levels described in Fig. 2. In addition to a
typical healthy resting state (SS0) we found that this biobehavioral
model could also accommodate an alternate equilibrium state where a
persistent depressive mood and increased anxiety are accompanied by
reduced physical fatigue and impaired attention. Serotonin is regu-
lated to chronically low levels while levels of glutamate, cortisol,
GABA, and epinephrine are maintained high. This finding would
suggest that a persistent depressive mood might be perpetuated at
least in part by normal regulatory drive (see Note 1).
3.4 Discovering Given the current scarcity of data linking subjective well-being to
Biobehavioral physiological markers, we propose a reverse engineering analysis
Regulatory where we iteratively construct possible biobehavioral feedback
Mechanisms mechanisms and simulate their ability to mimic observed clinical
outcomes. Specifically, we propose circuits where well-being
3.4.1 A Circuit Model directly modulates two molecular targets and is itself modulated
for Subjective Well-Being by two feedback signals. These candidate circuits are refined
94 J. Tory Toole et al.
4 Notes
0.5
0.45
0.4
0.35
Cook's distance
0.3
0.25
0.2
0.15
0.1
0.05
0
0 20 40 60 80 100
Network sorted rank
Fig. 3 Ranking biobehavioral feedback mechanisms for subjective well-being. Constraining the circuit
architecture for subjective well-being to two incoming and two outgoing control actions, simulations show
that three candidate feedback circuits perform significantly better than average ( p < 0.05, dotted line) in
reproducing observed clinical outcomes, as measured by Cook’s outlier distance. One circuit in particular that
outperformed the others in reproducing the protective effect of subjective well-being consisted of direct
downregulation of cortisol and upregulation of norepinephrine, with positive feedback from acetylcholine and
dopamine
Acknowledgments
References
1. Ryff CD, Keyes CLM (1995) The structure of 13. El Mansari M, Guiard BP, Chernoloz O,
psychological well-being revisited. J Pers Soc Ghanbari R, Katz N, Blier P (2010) Relevance
Psychol 69(4):719 of norepinephrine–dopamine interactions in
2. Wood AM, Joseph S (2010) The absence of the treatment of major depressive disorder.
positive psychological (eudemonic) well-being CNS Neurosci Ther 16(3):e1–e17
as a risk factor for depression: a ten year cohort 14. Calabresi P, Picconi B, Parnetti L, Di Filippo M
study. J Affect Disord 122(3):213–217 (2006) A convergent model for cognitive dys-
3. Ryff CD, Singer BH, Love GD (2004) Positive functions in Parkinson’s disease: the critical
health: connecting well-being with biology. dopamine–acetylcholine synaptic balance. Lan-
Philos Trans R Soc Lond B Biol Sci 359 cet Neurol 5(11):974–983
(1449):1383 15. Biello SM, Golombek DA, Harrington ME
4. Fava GA (2016) Well-being therapy. In: The (1997) Neuropeptide Y and glutamate block
Wiley handbook of positive clinical psychology. each other’s phase shifts in the suprachiasmatic
Wiley, Hoboken, p 395 nucleus in vitro. Neuroscience 77
5. Wurtman RJ, Pohorecky LA, Baliga BS (1972) (4):1049–1057
Adrenocortical control of the biosynthesis of 16. Serfozo P, Bartfai T, Vizi ES (1986) Presynap-
epinephrine and proteins in the adrenal tic effects of neuropeptide Y on [3 H] nor-
medulla. Pharmacol Rev 24(2):411–426 adrenaline and [3 H] acetylcholine release.
6. Janssen SA, Arntz A, Bouts S (1998) Anxiety Regul Pept 16(2):117–123
and pain: epinephrine-induced hyperalgesia 17. Dryden S, McCarthy HD, Malabu UH,
and attentional influences. Pain 76 Ware M, Williams G (1993) Increased neuro-
(3):309–316 peptide Y concentrations in specific hypotha-
7. Piazza PV, Rougé-Pont F, Deroche V, lamic nuclei of the rat following treatment with
Maccari S, Simon H, Le Moal M (1996) Glu- methysergide: evidence that NPY may mediate
cocorticoids have state-dependent stimulant serotonin’s effects on food intake. Peptides 14
effects on the mesencephalic dopaminergic (4):791–796
transmission. Proc Natl Acad Sci 93 18. Herman JP, Renda A, Bodie B (2003) Norepi-
(16):8716–8720 nephrine–gamma-aminobutyric acid (GABA)
8. Yuen EY, Liu W, Karatsoreos IN, Feng J, McE- interaction in limbic stress circuits: effects of
wen BS, Yan Z (2009) Acute stress enhances reboxetine on GABAergic neurons. Biol Psy-
glutamatergic transmission in prefrontal cortex chiatry 53(2):166–174
and facilitates working memory. Proc Natl 19. Wahlestedt CLAES, Hakanson ROLF, Vaz CA,
Acad Sci 106(33):14075–14079 Zukowska-Grojec ZOFIA (1990) Norepi-
9. Tafet GE, Toister-Achituv M, Shinitzky M nephrine and neuropeptide Y: vasoconstrictor
(2001) Enhancement of serotonin uptake by cooperation in vivo and in vitro. Am J Phys
cortisol: a possible link between stress and Regul Integr Comp Phys 258(3):R736–R742
depression. Cogn Affect Behav Neurosci 1 20. Bremner JD, Krystal JH, Southwick SM, Char-
(1):96–104 ney DS (1996) Noradrenergic mechanisms in
10. Bergstrom DA, Walters JR (1984) Dopamine stress and anxiety: II. Clinical studies. Synapse
attenuates the effects of GABA on single unit 23(1):39–51
activity in the globus pallidus. Brain Res 310 21. Southwick SM, Bremner JD, Rasmusson A,
(1):23–33 Morgan CA, Arnsten A, Charney DS (1999)
11. Daw ND, Kakade S, Dayan P (2002) Oppo- Role of norepinephrine in the pathophysiology
nent interactions between serotonin and dopa- and treatment of posttraumatic stress disorder.
mine. Neural Netw 15(4):603–616 Biol Psychiatry 46(9):1192–1204
12. Gillard ER, Dang DQ, Stanley BG (1993) Evi- 22. Bymaster FP, Katner JS, Nelson DL, Hemrick-
dence that neuropeptide Y and dopamine in the Luecke SK, Threlkeld PG, Heiligenstein JH,
perifornical hypothalamus interact antagonisti- Morin SM, Gehlert DR, Perry KW (2002)
cally in the control of food intake. Brain Res Atomoxetine increases extracellular levels of
628(1):128–136 norepinephrine and dopamine in prefrontal
cortex of rat: a potential mechanism for efficacy
Computational Discovery of Resilience Mechanisms in Depression 99
association and interaction with cortisol. Mol 54. Thomas R (1991) Regulatory networks seen as
Psychiatry 12(6):544–555 asynchronous automata: a logical description. J
46. Connor KM, Davidson JR (1998) The role of Theor Biol 153(1):1–23
serotonin in posttraumatic stress disorder: neu- 55. Craddock TJ, Fritsch P, Rice MA Jr, del
robiology and pharmacotherapy. CNS Spectr 3 Rosario RM, Miller DB, Fletcher MA, Klimas
(S2):42–51 NG, Broderick G (2014) A role for homeo-
47. Charney DS, Woods SW, Goodman WK, static drive in the perpetuation of complex
Heninger GR (1987) Serotonin function in chronic illness: gulf war illness and chronic
anxiety. Psychopharmacology 92(1):14–24 fatigue syndrome. PLoS One 9(1):e84839
48. Owens MJ, Nemeroff CB (1994) Role of sero- 56. Fritsch P, Craddock TJ, del Rosario RM, Rice
tonin in the pathophysiology of depression: MA, Smylie A, Folcik VA, de Vries G, Fletcher
focus on the serotonin transporter. Clin MA, Klimas NG, Broderick G (2013) Suc-
Chem 40(2):288–295 cumbing to the laws of attraction: exploring
49. Roelands B, Meeusen R (2010) Alterations in the sometimes pathogenic versatility of discrete
central fatigue by pharmacological manipula- immune logic. Systems Biomedicine 1
tions of neurotransmitters in normal and high (3):179–194
ambient temperature. Sports Med 40 57. Müller N, Schwarz MJ (2007) The immune-
(3):229–246 mediated alteration of serotonin and gluta-
50. Czermak C, Staley JK, Kasserman S, Bois F, mate: towards an integrated view of depression.
Young T, Henry S, Tamagnan GD, Seibyl JP, Mol Psychiatry 12:988–1000
Krystal JH, Neumeister A (2008) β2 Nicotinic 58. Soeteman DI, Miller M, Kim JJ (2012) Mod-
acetylcholine receptor availability in post- eling the risks and benefits of depression treat-
traumatic stress disorder. Int J Neuropsycho- ment for children and young adults. Value
pharmacol 11(3):419–424 Health 15(5):724–729
51. Botly LC, De Rosa E (2008) A cross-species 59. Gruwez B, Poirier MF, Dauphin A, Olié JP,
investigation of acetylcholine, attention, and Tod M (2007) A kinetic-pharmacodynamic
feature binding. Psychol Sci 19 model for clinical trial simulation of antidepres-
(11):1185–1193 sant action: application to clomipramine–-
52. Daniel JM, Dohanich GP (2001) Acetylcholine lithium interaction. Contemp Clin Trials 28
mediates the estrogen-induced increase in (3):276–287
NMDA receptor binding in CA1 of the hippo- 60. Holca-Lamarre R, Lücke J, Obermayer K
campus and the associated improvement in (2017) Models of Acetylcholine and Dopa-
working memory. J Neurosci 21 mine Signals Differentially Improve Neural
(17):6949–6956 Representations. Front. Comput Neurosci
53. Mendoza L, Xenarios I (2006) Theoretical 11:54
biology and medical Modelling. Theor Biol 61. Cipresso P, Immekus JC (2017) Back to the
Med Model 3:13 Future of Quantitative Psychology and Mea-
surement: Psychometrics in the Twenty-First
Century. Front Psychol 8:2099
Chapter 7
Abstract
Complex disorders like Gulf War illness (GWI) often defy diagnosis on the basis of a single biomarker and
may only be distinguishable by considering the co-expression of multiple markers measured in response to a
challenge. We demonstrate the practical application of such an approach using an example where blood was
collected from 26 GWI, 13 healthy control subjects, and 9 unhealthy controls with chronic fatigue at three
points during a graded exercise challenge. A 3-way multivariate projection model based on 12 markers of
endocrine and immune function was constructed using a training set of n ¼ 10 GWI and n ¼ 11 healthy
controls. These groups were separated almost completely on the basis of two co-expression patterns. In a
separate test set these same features allowed for discrimination of new GWI subjects (n ¼ 16) from
unhealthy (n ¼ 9) and healthy control subjects with a sensitivity of 70% and a specificity of 90%.
Key words Cytokine profile, Co-expression patterns, Exercise response, Gulf War illness, Regression
model, Diagnostic classification, Partial least squares, Batch PLS
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_7, © Springer Science+Business Media, LLC, part of Springer Nature 2018
101
102 Gordon Broderick et al.
2 Materials
2.1 A Standardized In this example, subject inclusion criteria were derived from Fukuda
Case Definition et al. (1998) [17], and consisted in identifying veterans deployed to
and Psychometric the theater of operations between August 8, 1990, and July
Tools 31, 1991, with one or more symptoms present after 6 months
from at least two of the following: fatigue; mood and cognitive
complaints; and musculoskeletal complaints (joint pain, stiffness, or
muscle pain). The suitability of the Fukuda definition for identify-
ing GWI patients is supported by the work of Collins et al.
(2002) [18].
In support of the case definition psychometric questionnaires
used to assess illness severity included the Multidimensional
Fatigue Inventory (MFI) [19], a 20-item self-report instrument
designed to measure fatigue, and the Medical Outcomes Study
36-item short-form survey (SF-36) [20] assessing health-related
quality of life. Typically all subjects also receive a physical examina-
tion and provided a medical history including the GWI symptom
checklist as per the case definition.
2.3 Standard Assays The data used here to illustrate the identification of biomarkers
for Measuring Immune co-expression patterns is based on measurements of immune cell
Markers in Blood signaling. Peripheral blood mononuclear cells (PBMCs) were
recovered from heparinized samples of whole blood and cultured
for 48 h with phytohemagglutinin at 37 C, 5% CO2. Following
incubation, culture supernatants were collected and frozen at
70 C until analyzed for concentrations of IL-1α, IL-5, IL-6,
IL-10, TNFα, and IFNγ. Levels of IL-6, IL-10, and TNFα were
also determined in plasma samples that were separated with 2 h,
stored at 70 C, and not thawed prior to analysis. In all cases
measurement was performed using commercial ELISA kits (Immu-
notech, Miami, FL), quality-controlled using National Institute for
Biological Standards and Control/World Health Organization
cytokine reference standards. Plasma sCD26 was also measured by
104 Gordon Broderick et al.
n ½X
n
k ¼ n ½T n A A ½P 0 A k þ n ½E n k ð1Þ
Immune Response Patterns in Gulf War Illness 105
n ½Y
n
m ¼ n ½U n A A ½C 0 A m þ n ½F n m ð2Þ
n ½U A ¼ n ½T n A þ n ½H n A ð3Þ
n
2.4.3 Statistical Tools Classification performance statistics and receiver operating charac-
for Assessing teristic (ROC) curves were calculated using functions found in the
the Diagnostic SPSS Statistics software package (IBM, Armonk, NY) and the
Performance MatLab Statistics Toolbox software (MathWorks, Natick, MA).
3 Methods
3.1 Recruitment of a In this example, a first subset of 10 GWI and 11 control subjects
Study Cohort. described previously [9] were recruited from the Miami Veterans
Administration Medical Center and studied. Subjects were male
and ranged in age between 30 and 55. Subjects were in good health
prior to 1990, and had no current exclusionary diagnoses that
could reasonably explain the symptoms and their severity. Subjects
taking medications that may impact immune function were
excluded (e.g., steroids, immunosuppressive agents). Control sub-
jects consisted of Gulf War-era sedentary veterans and were
matched to GWI subjects by age, gender, body mass index
(BMI), and ethnicity. Controls were also selected and matched to
patients on the basis of their general fitness level, estimated a priori
from their history of activity, and confirmed during exercise chal-
lenge. Consistent with these criteria and specific to this analysis a
second group was subsequently recruited consisting of an addi-
tional 16 male and 10 female GWI subjects as well as 9 male CFS
subjects and 2 new male control subjects. This new group of sub-
jects was used as a separate validation set in this example.
3.2 Applying The McArdle protocol [28] for graded exercise was used to elicit an
Standard Exercise immune-endocrine response. Subjects pedal at an initial output of
Challenge 60 W for 2 min, followed by an increase of 30 W every 2 min until
the subject (1) reaches a plateau in maximal oxygen consumption
(VO2); (2) reaches a respiratory exchange ratio above 1.15; or
(3) stops the test.
3.3 Sample Prior to the exercise challenge a first blood draw was conducted
Collection after subjects sat quietly for 30 min. Second and third blood draws
and Analysis were conducted upon reaching peak effort (VO2 max) and at 4 h
post-exercise, respectively. At each blood draw, five 8 mL tubes of
blood were collected in CPT vacutainers (BD Biosciences, San Jose,
CA). Importantly, diurnal variations in this and other indicators
were controlled by conducting assessments at the same time of day
for all subjects. Blood samples were processed and markers
recorded as described in Subheading 2.3.
Table 1
Changes in the expression of signaling molecules in GWI
3.4.2 Isolating Patterns The nervous, immune, and endocrine systems are generally recog-
of Biomarker nized as being highly integrated regulatory networks. As a result,
Co-expression even though there were relatively few differences in the absolute
expression of individual markers, there may still be co-regulatory
imbalances. Because the basic PLS algorithm described in Subhead-
ing 2.4.2 can be applied to all data points independently of sample
time the dynamics of the immune response are not captured. To
address this we applied a two-step analysis commonly referred to as
three-way PLS or batch PLS [29, 30]. This model is described in
Fig. 1.
1. First PLS was applied to all measurements across all individual
markers using sample time t as a response Y in level 1 of the
model (steps 1–2). This was done to enforce the sequence of
values when identifying the features that best described the
patterns of co-expression linking the 12 markers (X in level
1). In a pre-processing step all regressor variables in the X block
were log transformed. Each log-transformed variable was then
scaled by dividing with the corresponding standard deviation
computed about zero instead of the mean. The response vari-
able Y describing the progression of each batch or exercise
challenge was not transformed nor was it centered. To capture
the unequal spacing in time of the three observations made
during each trial we used an approximate time vector of [t0, t1,
t2] ¼ [0, 30, 270] in minutes. In addition, the training set used
here contains roughly 8% missing array elements. Missing data
was handled implicitly by the NIPALS algorithm using an
extension to the original computation proposed by Christof-
fersson (1970) [31] that consists essentially in iteratively sub-
stituting missing values with their model predictions. This
approximate method performs well with up to 10–20% missing
Immune Response Patterns in Gulf War Illness 109
Fig. 1 PLS-DA classification model structure. Diagram of the input and output relationships defined at each
level of the PLS model for classification of subjects to the GWI or healthy control class
Table 2
PLS model fit and structure
Component
Model a R2X R2X(cum) Eigenvalues R2Y R2Y(cum) Q
2
Q2(cum)
Level 1 Model
Observations (N ) ¼ 69 1 0.83 0.83 9.91 0.40 0.40 0.40 0.40
K ¼ 12 Biomarkers 2 0.07 0.89 0.81 0.06 0.46 0.04 0.42
vs. Time
Level 2 Model
Observations (N ) ¼ 21 1 0.53 0.53 3.19 0.38 0.38 0.33 0.33
2 Level-1 features at 2 0.32 0.85 1.93 0.09 0.46 0.04 0.36
3 times (K ¼ 6)
R2x and R2y are Pearson correlation coefficients or total sum of squared regression (SSR) captured by the partial least
squares (PLS) model for the regressor X (molecular signals) and predicted variable Y (diagnostic class) spaces, respec-
tively. The value Q2 ¼ 1-PRESS/SST expresses overall model performance in terms of reduction in the predicted residual
sum of squared errors (PRESS) in leave-one-out cross-validation
Fig. 2 Structure of composite features. The relative contribution pka of each biomarker k to each of the two
level 1 model features (a ¼ 1, 2) that optimally separate the subjects into their respective diagnostic groups.
The first feature captures a broad-spectrum disturbance across all biomarkers. The second feature captures a
pattern of co-expression whereby increases in IL-1a are offset by a corresponding increase in cortisol, IL-6,
and IL-10. Whiskers represent the standard error in estimation for each weight pka. All level 1 feature
1 weights are significant at p 0.001. Only the feature 2 weights for cortisol ( p 0.001), IL-1a ( p ¼ 0.024),
IL-6 ( p ¼ 0.008), and IL-10 ( p 0.001) were significant at p < 0.05
Fig. 3 Separation of subjects in feature space. Distribution of GWI and control subjects in a feature space
defined by the co-expression of 12 biomarkers of neuroendocrine-immune status. Each axis is defined by a
specific linear combination of biomarkers and the evolution in these patterns over the course of a graded
exercise challenge. Each point captures the entire time course recorded for that subject. In this feature space
4 of 11 control subjects appears to group with the 10 GWI subjects
Fig. 4 Evolution of feature scores in time for matched pairs. Median expression of level 1 feature 1 (panel A)
and feature 2 (panel B) prior to exercise (t0), at peak effort (t1), and 4 h post-exercise (t2) is shown in panels A
and B for GWI subject CS 201 (blue diamonds) and case-matched control subject PC 105 (red squares).
Panels C and D show data for GWI subject VR 203 (blue diamonds) and control subject CC 104 (red squares).
Error bars display the median absolute deviation from median (MADM) as computed from repeated cross-
validation for these observations. In panels A and B time and illness effects can be seen in both level 1 feature
1 and feature 2. Significant deviations at peak effort in feature 2 illustrate why this time point is highly
weighted in the level 2 model. In panels C and D we observe significant time effects in level 1 feature 1 and
while illness results in a persistent offset in the expression of feature 2
3.4.3 Exercise Response To assess classifier performance the continuous scale for class mem-
as a Basis for Classification bership can be transformed to a discrete class assignment by apply-
ing a membership threshold. Detection sensitivity and assignment
specificity values can then be computed using these binary assign-
ments. This is repeated over a range of threshold values to produce
a receiver-operator characteristic (ROC) curve [33].
1. In this example separate receiver-operator characteristic (ROC)
curves were computed for classification using the broad-
spectrum disturbance alone (feature 1) as well as classification
including the inflammatory imbalance pattern (feature 2)
(Fig. 5, Table 3). Both classifiers performed similarly on the
training set of 10 male GWI and 11 male healthy control
subjects, each producing an area under the curve (AUC) sig-
nificantly better than random assignment ( p ¼ 0.001 Table 3).
Nonetheless, at 85% specificity the use of both features delivers
a sensitivity of 90% rather than 70% possible with one feature
alone.
2. Performance differences became very noticeable when these
models were applied to a test set. Our first validation set con-
sisted of 16 new male GWI subjects, 9 male CFS controls, and
2 new male healthy controls. The corresponding ROC curves
for the 1 and 2 feature classifiers are shown in Fig. 5. When
tested on these new male subjects, classification using only the
first feature does not differ significantly from a random assign-
ment ( p ¼ 0.23 AUC). Inclusion of the second feature deliv-
ered a significant increase in AUC ( p ¼ 0.04 Table 3) with 90%
specificity available at 70% sensitivity. Recall that the training set
contained only male GWI and control subjects. To test the
impact of gender we constructed a second validation set by
adding ten female GWI cases to the first set. Classification
performance decreased drastically. Neither model performed
better than random assignment although two features were still
preferable ( p ¼ 0.11) to one ( p ¼ 0.31) (Table 3).
3. We intentionally included CFS subjects in the validation set as a
disease control group since CFS is clinically similar to GWI.
Using the ROC analysis performed on the training set we
Immune Response Patterns in Gulf War Illness 115
Fig. 5 Linear classifier performance. Receiver-operator characteristic (ROC) curves for a linear discriminant
model constructed from cytokine, NPY, and cortisol co-expression in 10 GWI and 11 male control subjects.
Sensitivity and specificity of classification obtained at various membership threshold values are presented for
the original training set as well as for a validation test set consisting of 9 male CFS patients as well as 16 new
GWI and 2 new male control cases
Table 3
Receiver-operator characteristics (ROC) for PLS-DA classifiers
Asymptotic 95%
Area Std.
under Error Asymptotic Upper Lower
Model Data set Positive:Negative curve (a) Sig. (b) bound bound
Feature Training set 10:11 Ctrl 0.900 0.064 0.001 0.774 1.026
1 Only
Feature Training set 10:11 Ctrl 0.900 0.065 0.001 0.772 1.028
1 and 2
Feature Test set (male 16:9 CFS, 2 Ctrl 0.646 0.122 0.234 0.407 0.885
1 Only only)
Feature Test set (male 16:9 CFS, 2 Ctrl 0.750 0.110 0.042 0.534 0.966
1 and 2 only)
Feature Test set (male 26 (incl. 10 female): 0.615 0.117 0.308 0.387 0.844
1 Only and female) 9 CFS, 2 Ctrl
Feature Test set (male 26 (incl. 10 female): 0.679 0.113 0.113 0.458 0.901
1 and 2 and female) 9 CFS, 2 Ctrl
Detailed statistics for receiver-operator curves (ROC) describing the sensitivity as a function of 1 specificity obtained in
classification models based on partial least squares discriminant analysis (PLS-DA). These include the area under the
curve (AUC) with the associated standard error of estimation, the corresponding significance p-value, as well as the upper
and lower confidence limits at the 0.05 significance level. Statistics are reported for classification in the training set, a
male-only test set, and a mixed-gender set
4 Notes
Acknowledgments
References
1. Kang HK, Natelson BH, Mahan CM et al cortisol secretion in patients with fibromyalgia
(2003) Post-traumatic stress disorder and and/or chronic fatigue syndrome. Brain Behav
chronic fatigue syndrome-like illness among Immun 18:314–325
Gulf War Veterans: a population-based survey 4. Golier JA, Schmeidler J, Legge J et al (2006)
of 30,000 veterans. Am J Epidemiol Enhanced cortisol suppression to dexametha-
157:141–148 sone associated with Gulf War deployment.
2. Unwin C, Blatchley N, Coker W et al (1999) Psychoneuroendocrinology 31
Health of UK servicemen who served in Per- (10):1181–1189
sian Gulf War. Lancet 353:169–178 5. Golier JA, Schmeidler J, Legge J et al (2007)
3. Crofford LJ, Young EA, Engleberg NC et al Twenty-four hour plasma cortisol and adreno-
(2004) Basal circadian and pulsatile ACTH and corticotropic hormone in gulf war veterans:
Immune Response Patterns in Gulf War Illness 119
relationships to posttraumatic stress disorder and baseline characteristics. Control Clin Trials
and health symptoms. Biol Psychiatry 62 23(3):333–353
(10):1175–1178 19. Smets EM, Garssen B, Bonke B et al (1995)
6. Maher K, Klimas NG, Fletcher MA (2003) The multidimensional fatigue inventory (MFI)
Immunology. In: Jason LA, Fennell PA, Taylor psychometric qualities of an instrument to
RR (eds) Handbook of chronic fatigue syn- assess fatigue. J Psychosom Res 39(3):315–325
drome. Wiley, Hoboken, pp 124–151 20. Ware JE Jr, Sherbourne CD (1992) The MOS
7. Vojdani A, Thrasher JD (2004) Cellular and 36-item short-form health survey (SF-36). I
humoral immune abnormalities in Gulf War Conceptual framework and item selection.
veterans. Environ Health Perspect 112 Med Care 30(6):473–483
(8):840–846 21. Cadore E, Lhullier F, Brentano M et al (2008)
8. Siegel SD, Antoni MH, Fletcher MA et al Correlations between serum and salivary hor-
(2006) Impaired natural immunity, cognitive monal concentrations in response to resistance
dysfunction, and physical symptoms in patients exercise. J Sports Sci 26(10):1067–1072
with chronic fatigue syndrome: preliminary 22. Wold S, Trygg J, Berglund A et al (2001) Some
evidence for a subgroup? J Psychosom Res 60 recent developments in PLS modeling. Che-
(6):559–566 mom Intell Lab Syst 58:131–149
9. Whistler T, Fletcher MA, Lonergan W et al 23. Eriksson L, Antti H, Gottfries J et al (2004)
(2009) Impaired immune function in Gulf Using chemometrics for navigating in the large
War Illness. BMC Med Genet 2:12 data sets of genomics, proteomics, and meta-
10. Broderick G, Fuite J, Kreitz A et al (2010) A bonomics (gpm). Anal Bioanal Chem 380
formal analysis of cytokine networks in chronic (3):419–429
fatigue syndrome. Brain Behav Immun 24 24. Wold S, Esbensen K, Geladi P (1987) Principal
(7):1209–1217 component analysis. Chemom Intell Lab Syst
11. Broderick G, Kreitz A, Fuite J et al (2011) A 2:37–52
pilot study of immune network remodeling 25. Barker M, Rayens W (2003) Partial least
under challenge in Gulf War Illness. Brain squares for discrimination. J Chemom 17
Behav Immun 25(2):302–313 (3):166–173
12. Moss RB, Mercandetti A, Vojdani A (1999) 26. Efron B, Gong G (1983) A leisurely look at the
TNF-alpha and chronic fatigue syndrome. J bootstrap, the jackknife, and cross-validation.
Clin Immunol 19(5):314–316 Am Statist 37:36–48
13. Gaab J, Rohleder N, Heitz V et al (2005) 27. Krzanowski WJ (1987) Cross-validation in
Stress-induced changes in LPS-induced proin- principal component analysis. Biometrics
flammatory cytokine production in chronic 43:575–584
fatigue syndrome. Psychoneuroendocrinology 28. McArdle WD, Katch FI, Katch VL (2007)
30:188–198 Exercise physiology: energy, nutrition, and
14. Skowera A, Cleare A, Blair D et al (2004) High human performance. Lippincott Williams &
levels of type 2 cytokine-producing cells in Wilkins, London, p 954
chronic fatigue syndrome. Clin Exp Immunol 29. Wold S, Kettaneh N, Friden H et al (1998)
135(2):294–302 Modelling and diagnostics of batch processes
15. ter Wolbeek M, van Doornen LJ, Kavelaars A and analogous kinetic experiments. Chemom
et al (2007) Longitudinal analysis of pro- and Intell Lab Syst 44:331–340
anti-inflammatory cytokine production in 30. Antti H, Bollard ME, Ebbels T et al (2002)
severely fatigued adolescents. Brain Behav Batch statistical processing of H-1 NMR-der-
Immun 21(8):1063–1074 ived urinary spectral data. J Chemom
16. Light AR, White AT, Hughen RW et al (2009) 16:461–468
Moderate exercise increases expression for sen- 31. Christoffersson A (1970) The one component
sory, adrenergic, and immune genes in chronic model with incomplete data. Ph.D. thesis,
fatigue syndrome patients but not in normal Uppsala University
subjects. J Pain 10(10):1099–1112
32. Grung B, Manne R (1998) Missing values in
17. Fukuda K, Nisenbaum R, Stewart G et al principal component analysis. Chemom Intell
(1998) Chronic multisymptom illness affecting Lab Syst 42:125–139
Air Force veterans of the Gulf War. JAMA
280:981–988 33. Greiner M, Pfeiffer D, Smith RD (2000) Prin-
ciples and practical application of the receiver-
18. Collins JF, Donta ST, Engel CC et al (2002) operating characteristic analysis for diagnostic
The antibiotic treatment trial of Gulf War tests. Prev Vet Med 45(1–2):23–41
Veterans’ illnesses: issues, design, screening,
120 Gordon Broderick et al.
Abstract
We propose that the complexity of regulatory interactions modulating brain neurochemistry and behavior is
such that multiple stable responses may be supported, and that some of these alternate regulatory programs
may play a role in perpetuating persistent psychological dysfunction. To explore this, we constructed a
model network representing major neurotransmission and behavioral mechanisms reported in literature as
discrete logic circuits. Connectivity and information flow through this biobehavioral circuitry supported
two distinct and stable regulatory programs. One such program perpetuated a depressive state with a
characteristic neurochemical signature including low serotonin. Further analysis suggested that small
irregularities in glutamate levels may render this pathology more directly accessible. Computer simulations
mimicking selective serotonin reuptake inhibitor (SSRI) therapy in the presence of everyday stressors
predicted recidivism rates similar to those reported clinically and highlighted the potentially significant
benefit of concurrent behavioral stress management therapy.
Key words Computational modeling, Homeostatic regulation, Depression, SSRI, Stress, Neuro-
transmitters, Glutamate, Serotonin, Network complexity, Regulatory logic, Multi-stability
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_8, © Springer Science+Business Media, LLC, part of Springer Nature 2018
121
122 J. Tory Toole et al.
2 Materials
2.2 A Discrete To direct the flow of information through this model circuit we use
Signaling Logic a simple discrete ternary logic where each model behavioral con-
Framework struct and signaling molecule could be expressed at a low, nominal,
or high level. The specific formalism used in this work is an exten-
sion of a paradigm proposed by Mendoza and Xenarios (2006) [56]
and Thomas (1991) [57], and has been reported previously by our
group [58, 59]. We encode documented regulatory mechanisms
within the neuropsychological system using only the direction
(source and target) and type (activator or inhibitor) of interaction.
As data describing the relative magnitude of response to these
126 J. Tory Toole et al.
~
x ðt Þ ¼ ðx 1 ðt Þ; x 2 ðt Þ; . . . ; x N ðt ÞÞ ð1Þ
where xi(t) is the state of the ith variable in the N variable system
at time t. The image vector ~ x ðt þ 1Þ describes the preferred state
toward which the system evolves in the next time increment. The
state value of the image vector for the ith variable is determined
from its current state and a set of balanced ternary logic statements
based on the current value of each variable and the mode of action
(i.e., promote or inhibit) of the neighboring input variables. These
logic statements are expressed as follows (Eq. (2)):
8 A I
>
> x i1 ðt Þ∨x i2
A
ðt Þ∨ . . . ∨x iXA
ðt Þ ∇ x i1 ðt Þ∨x i2
I
ðt Þ∨ . . . ∨x iY
I
ðt Þ
< A
x i ðt þ 1Þ ¼ x i1 ðt Þ∨x i2A
ðt Þ∨ . . . ∨x iX
A
ðt Þ
>
> I
:
˜ x i1 ðt Þ∨x I
i2 ðt Þ∨ . . . ∨x I
iY ð t Þ
ð2Þ
Fig. 1 Basic logic model of behavior and neurotransmission. A basic causal regulatory network model representing the principal signaling mechanisms that
127
coordinate the expression of 15 soluble mediators in the brain and associated behavioral constructs. Green connectors with an arrowhead terminator represent
stimulatory actions whereas red connectors with a dot terminator represent inhibitory actions
128 J. Tory Toole et al.
2.3 A Monte Carlo The evolution of state transitions supported by the model was
Discrete Event simulated by applying the decision logic described in Subheading
Simulation 2.2 iteratively using a Monte Carlo algorithm. From any initial
Environment starting state, allowable state transitions are determined based on
Eq. (2). Applying the latter to each variable in the model for the mth
state of the system, ~ x m ðt Þ, defines the image vector ~
x m ðt þ 1Þ for
the m state. With ~
th
x ðt þ 1Þ defined, the system is updated
m
2.4 Defining The above simulation paradigm can serve as a basis for the in silico
an Idealized Virtual testing of strategies for rescuing the neurobehavioral system from a
Treatment stable but pathological regulatory program. Indeed, by extending
the paradigm in Subheading 2.3 to include externally applied state
changes it becomes possible to predict the outcome of therapeutic
strategies consisting of robust sequences of discrete treatments with
a select combination of drug actions (agonist or antagonist),
applied to specific targets at carefully timed intervals. In this case
simulated treatment is directed toward supporting the recovery of
normal homeostasis and delivering lasting remission in the absence
of any further therapy. Each treatment event applied to the system
at time t in the treatment course is represented by the vector T ~ ðt Þ
specifying the change in state applied to each of the N model state
variables, in this case behavioral constructs and signaling molecules
such that
~ ðt Þ ¼ ðT 1 ðt Þ; T 2 ðt Þ; . . . ; T N ðt ÞÞ
T ð3Þ
2.5 A Global Search Using the framework described in Subheading 2.4, treatment
Iterative Simulation courses can be simulated and the desirability of their outcomes
Engine assessed with respect to a specific treatment goal. Repeating this
trial-and-error process iteratively makes it possible to gradually
improve the candidate treatment but only if (1) we learn efficiently
from past attempts while (2) continuing to maintain a broad view of
potential intervention choices. These two features are often and
odds, with conventional methods typically favoring a very efficient
search which all too quickly becomes very narrow in scope leading
to the equivalent of numerical tunnel vision. In an attempt to avoid
settling prematurely on a mediocre design a family of so-called
global search algorithms prescribe and conduct broad sets of con-
current simulations, retaining not only the best design but also the
runner-up solutions. The best features of each of these competing
treatment solutions are combined to create a new generation of
candidate designs. With a basic form that naturally accommodates
discrete state models, we apply in this work a simple genetic algo-
rithm (GA) [60] to balance breadth with efficiency in this trial-and-
error process.
A treatment course vector C ~ with M treatment events is
therefore defined as
~¼ T
C ~ ðt 1 Þ; T
~ ðt 2 Þ; . . . ; T
~ ðt M Þ ð5Þ
2.6 Measures To explore factors that may increase vulnerability to falling into a
of Vulnerability persistent depressed state, we define and conduct sensitivity
and Impact analysis-based simulations of the regulatory logic supported by
the biobehavioral circuit model. Using a normal healthy regulatory
balance as a start point, an exogenous agonist or antagonist was
applied to each signaling molecule in turn such that corresponding
levels were maintained high or low while the response of the
regulatory network to a stressful triggering event was simulated.
This was repeated 1000 times for both abnormally high and low
levels of each signaling molecule and the results used to determine
the frequency with which response exceeded the tipping point and
degenerated into a persistent depressed state. This simulated fre-
quency of onset in the wake of an idealized triggering event was
used as a measure for describing the robustness of the biobehavioral
network to spurious excursions in the levels of individual signaling
neurotransmitters and hormones.
3 Methods
The relevance and value of these tools and the framework described
in Subheading 2 in creating new insight into complex illnesses are
best described using an example problem, in this case a model-
based analysis of regulatory dynamics in depression. First an analysis
of regulatory stability is applied (Subheading 3.1) to identify natu-
rally occurring regulatory programs that might promote persis-
tence of depression and resistance to conventional treatment.
Fluctuations in neurotransmitter and hormone levels that increase
vulnerability and facilitate access to pathologic programs, in this
case depression, are then identified (Subheading 3.2) using com-
puter simulations and the most direct course of onset associated
with such fluctuations is predicted based on the signaling logic
(Subheading 3.3). Extending this analysis of depression as a persis-
tent regulatory trap, the potential for delivering lasting remission
using conventional SSRI therapy (Subheading 3.4.1) is simulated
and results are compared to clinical observations of treatment
efficacy and recidivism rates. The same numerical protocol is
applied to assess the predicted efficacy of GABA inhibition (Sub-
heading 3.4.2) as well as behavioral intervention (Subheading
3.4.3) used in combination with conventional SSRI treatment.
The most synergistic way in which these individual intervention
components might be combined and scheduled was analyzed for-
mally using the model-based framework described in Subheading 2.
132 J. Tory Toole et al.
3.1 Identifying As described above, steady homeostatic states are defined as resting
Stable Regulatory states to which the system naturally returns after perturbation.
Programs Robust and adaptable, even relatively simple biological systems
are capable of supporting more than one such resting state. These
were determined computationally in this work by enumerating all
possible states that may be occupied by the biobehavioral circuit
model shown in Fig. 1 and the corresponding control logic. An
exhaustive search using the simulation paradigm described in Sub-
heading 2.2 identified two homeostatic states (SS) where neuro-
transmitter and psychobehavioral constructs converged to
equilibrium levels described in Fig. 2. The reference steady state
(SS0) described typical healthy equilibrium with all markers
expressed at their nominal levels encoded as zero. In addition to
this baseline equilibrium state (SS0) we found that this regulatory
circuitry can also accommodate a persistent depressive state also
characterized by increased anxiety accompanied by reduced physical
fatigue and impaired attention. This same regulatory program main-
tained chronically low levels of serotonin while promoting elevated
cortisol glutamate, GABA, and epinephrine levels. This result would
suggest that a persistent depressive mood might be perpetuated
at least in part by homeostatic regulatory action (see Note 1).
3.2 Sensitivity With the possibility that persistent depression may at least partially
Analysis coincide with a naturally occurring regulatory trap it is relevant to
and Predisposing ask what factors may promote the initiation of this pathologic
Factors of Onset biobehavioral program. Repeated simulations were conducted
using the protocol and measures described in Subheading 2.6 to
assess the sensitivity of normal regulatory programming to anoma-
lous excursions in the levels of individual signaling molecules. In
the current example, the results of this sensitivity analysis (Fig. 3)
indicated that under normal equilibrium conditions, individual
deviations in almost all of the neurotransmitters and hormones
provided no increased risk for falling into the depressed steady
state beyond random chance. There were, however, three signifi-
cant exceptions. Increased levels of NPY were protective, prevent-
ing the biobehavioral network from escaping normal homeostasis
and falling into a depressive regulatory mode in 776 out of 1000
simulations. Interestingly abnormally low levels of NPY did not
significantly affect vulnerability. Changes in glutamate, however,
produced significant effects when either abnormally high or low.
Spurious increases in glutamate levels facilitated migration from
Computing an Escape From Depression 133
3.3 Trajectory The sensitivity analysis in Subheading 3.2 suggested that abnor-
Analysis and Mapping mally elevated glutamate provides a high-risk environment in which
the Course of Onset a depressive illness state is most easily reached when the system is
put under stress. However, this analysis does not inform on how
gradual or precipitous the onset might be, nor does it highlight
telltale signs of how far this migration into depression has pro-
gressed. Knowledge of the expected course of illness has important
clinical implications. In order to explore this aspect a trajectory
analysis was conducted to determine the sequence of states that
134 J. Tory Toole et al.
To Health
To Illness
894
775 776
668
615
550 576 549
516 526 522 507 507
484 505
495 474 478 493 493
450 451
424
385
332
225 224
106
Fig. 3 Sensitivity analysis of key factors affecting vulnerability to depression. Each node was artificially held
high, and then low, one at a time. During each orientation, the system was allowed to naturally evolve under
stress 1000 times. Of those 1000 iterations for each orientation, the number of times the system settled in the
illness state versus the health state was recorded to determine which individual nodes represented the largest
risk or protective factor
1 1 1 1 1 1 1 1 1 1
0 0 0 0 0 0 0 0 0 0
1 2 3 4 5 6 7 8 9 10
-1 -1 -1 -1 -1 -1 -1 -1 -1
'Cortisol' 'PASAT'
Fig. 4 Sequential migration path of acute illness onset. Simulated sequence of state changes (1 low,
0 nominal, +1 elevated) along the shortest average transition path from normal homeostatic regulation to
persistent depression. This transition path is triggered by a simulated stressful event under initial conditions of
a spurious elevation in glutamate. Only five state transition steps separate normal from pathological regulation
under such conditions and the sequence of events adheres strictly to and is derived directly from the
regulatory signaling circuitry
3.4.2 Designing Optimal Treatment with GABA inhibitors is also often used as a second line
Combination Therapies of therapy. Once again, we applied the numerical optimization
with GABA Inhibition scheme of Subheading 2.5, this time allowing for the optimal
scheduling of both SSRI administration and GABA inhibition.
The optimal treatment schedules for this combination therapy
were identified for two up to eight cycles of treatment. As before
each optimal treatment course was simulated 100 times in each of
100 virtual subjects. Overall, we found that these optimal treat-
ment courses included GABA inhibition only very sparingly and
that the burden of treatment was carried mainly by SSRI adminis-
tration. Nonetheless, the inclusion of GABA blockade improved
the predicted remission rates slightly under simulated real-world
conditions of 30% stressful events. However, this improvement did
not reach statistical significance.
100
90
% Mean remission (SD)
80
70
60
40
30
2 3 4 5 6 7 8
Rounds of treatment"
Fig. 5 Simulated intervention time steps. Summary of simulated intervention time courses with the maximum
predicted frequency of remission as a function of repeated intervention events simulated in (A) SSRI with
behavioral stress management (blue line), and treatment in the presence of randomly occurring stressful
events (30% frequency) without behavioral intervention using (B) SSRI only (red line) and (C) SSRI with GABA
blockade (green line)
4 Notes
Table 1
Remission rates of simulated SSRI treatment under various conditions
Acknowledgments
References
1. Substance Abuse and Mental Health Services glutamatergic transmission in prefrontal cortex
Administration (2013) Results from the 2012 and facilitates working memory. Proc Natl
National Survey on Drug Use and Health: Acad Sci 106(33):14075–14079
Mental Health Findings, NSDUH Series 11. Tafet GE, Toister-Achituv M, Shinitzky M
H-47, HHS Publication No. (SMA) (2001) Enhancement of serotonin uptake by
13–4805. Substance Abuse and Mental Health cortisol: a possible link between stress and
Services Administration, Rockville, MD depression. Cogn Affect Behav Neurosci 1
2. American Psychiatric Association (2013) Diag- (1):96–104
nostic and statistical manual of mental disor- 12. Janssen SA, Arntz A, Bouts S (1998) Anxiety
ders: DSM-V. 5. American Psychiatric and pain: epinephrine-induced hyperalgesia
Association, Washington, DC and attentional influences. Pain 76
3. Greenberg PE, Fournier A, Sisitsky T, Pike CT, (3):309–316
Kessler RC (2014) The economic burden of 13. Bergstrom DA, Walters JR (1984) Dopamine
adults with major depressive disorder in the attenuates the effects of GABA on single unit
United States (2005 and 2010). J Clin Psychi- activity in the globus pallidus. Brain Res 310
atry 76(2):155–162 (1):23–33
4. Antonuccio DO, Danton WG, DeNelsky GY 14. Daw ND, Kakade S, Dayan P (2002) Oppo-
(1995) Psychotherapy versus medication for nent interactions between serotonin and dopa-
depression: challenging the conventional wis- mine. Neural Netw 15(4):603–616
dom with data. J Psychother Pract Res 26 15. Gillard ER, Dang DQ, Stanley BG (1993) Evi-
(6):574–585 dence that neuropeptide Y and dopamine in the
5. Sałat K, Podkowa A, Kowalczyk P, Kulig K, perifornical hypothalamus interact antagonisti-
Dziubina A, Filipek B, Librowski T (2015) cally in the control of food intake. Brain Res
Anticonvulsant active inhibitor of GABA trans- 628(1):128–136
porter subtype 1, tiagabine, with activity in 16. El Mansari M, Guiard BP, Chernoloz O,
mouse models of anxiety, pain and depression. Ghanbari R, Katz N, Blier P (2010) Relevance
Pharmacol Rep 67(3):465–472 of norepinephrine–dopamine interactions in
6. Cuijpers P, van Straten A, Andersson G, van the treatment of major depressive disorder.
Oppen P (2008) Psychotherapy for depression CNS Neurosci Ther 16(3):e1–e17
in adults: a meta-analysis of comparative out- 17. Calabresi P, Picconi B, Parnetti L, Di Filippo M
come studies. J Consult Clin Psychol 76 (2006) A convergent model for cognitive dys-
(6):909 functions in Parkinson’s disease: the critical
7. Burcusa SL, Iacono WG (2007) Risk for recur- dopamine–acetylcholine synaptic balance. Lan-
rence in depression. Clin Psychol Rev 27 cet Neurol 5(11):974–983
(8):959–985 18. Biello SM, Golombek DA, Harrington ME
8. Wurtman RJ, Pohorecky LA, Baliga BS (1972) (1997) Neuropeptide Y and glutamate block
Adrenocortical control of the biosynthesis of each other’s phase shifts in the suprachiasmatic
epinephrine and proteins in the adrenal nucleus in vitro. Neuroscience 77
medulla. Pharmacol Rev 24(2):411–426 (4):1049–1057
9. Piazza PV, Rougé-Pont F, Deroche V, 19. Serfozo P, Bartfai T, Vizi ES (1986) Presynap-
Maccari S, Simon H, Le Moal M (1996) Glu- tic effects of neuropeptide Y on [3 H] nor-
cocorticoids have state-dependent stimulant adrenaline and [3 H] acetylcholine release.
effects on the mesencephalic dopaminergic Regul Pept 16(2):117–123
transmission. Proc Natl Acad Sci 93 20. Dryden S, McCarthy HD, Malabu UH,
(16):8716–8720 Ware M, Williams G (1993) Increased neuro-
10. Yuen EY, Liu W, Karatsoreos IN, Feng J, McE- peptide Y concentrations in specific hypotha-
wen BS, Yan Z (2009) Acute stress enhances lamic nuclei of the rat following treatment with
142 J. Tory Toole et al.
methysergide: evidence that NPY may mediate hypertensive rats. Metab Brain Dis 15
serotonin’s effects on food intake. Peptides 14 (4):297–304
(4):791–796 33. Mathew SJ, Coplan JD, Smith EL, Schoepp
21. Herman JP, Renda A, Bodie B (2003) Norepi- DD, Rosenblum LA, Gorman JM (2001) Glu-
nephrine–gamma-aminobutyric acid (GABA) tamate—hypothalamic-pituitary-adrenal Axis
interaction in limbic stress circuits: effects of interactions: implications for mood and anxiety
reboxetine on GABAergic neurons. Biol Psy- disorders. CNS Spectr 6(07):555–564
chiatry 53(2):166–174 34. Carlsson A, Waters N, Holm-Waters S,
22. Wahlestedt CLAES, Hakanson ROLF, Vaz CA, Tedroff J, Nilsson M, Carlsson ML (2001)
Zukowska-Grojec ZOFIA (1990) Norepi- Interactions between monoamines, glutamate,
nephrine and neuropeptide Y: vasoconstrictor and GABA in schizophrenia: new evidence.
cooperation in vivo and in vitro. Am J Phys Annu Rev Pharmacol Toxicol 41(1):237–260
Regul Integr Comp Phys 258(3):R736–R742 35. Aultman JM, Moghaddam B (2001) Distinct
23. Bremner JD, Krystal JH, Southwick SM, Char- contributions of glutamate and dopamine
ney DS (1996) Noradrenergic mechanisms in receptors to temporal aspects of rodent work-
stress and anxiety: II. Clinical studies. Synapse ing memory using a clinically relevant task.
23(1):39–51 Psychopharmacology 153(3):353–364
24. Southwick SM, Bremner JD, Rasmusson A, 36. Paul IA, Skolnick P (2003) Glutamate and
Morgan CA, Arnsten A, Charney DS (1999) depression. Ann N Y Acad Sci 1003
Role of norepinephrine in the pathophysiology (1):250–272
and treatment of posttraumatic stress disorder. 37. Carrey NJ, MacMaster FP, Gaudet L, Schmidt
Biol Psychiatry 46(9):1192–1204 MH (2007) Striatal creatine and glutamate/
25. Bymaster FP, Katner JS, Nelson DL, Hemrick- glutamine in attention-deficit/hyperactivity
Luecke SK, Threlkeld PG, Heiligenstein JH, disorder. J Child Adolesc Psychopharmacol 17
Morin SM, Gehlert DR, Perry KW (2002) (1):11–17
Atomoxetine increases extracellular levels of 38. Van Praag HM, Kahn RS, Asnis GM, Wetzler S,
norepinephrine and dopamine in prefrontal Brown SL, Bleich A, Korn ML (1987) Deno-
cortex of rat: a potential mechanism for efficacy sologization of biological psychiatry or the
in attention deficit/hyperactivity disorder. specificity of 5-HT disturbances in psychiatric
Neuropsychopharmacology 27(5):699–711 disorders. J Affect Disord 13(1):1–8
26. Zhang K, Grady CJ, Tsapakis EM, Andersen 39. Aghajanian GK, Marek GJ (1999) Serotonin–-
SL, Tarazi FI, Baldessarini RJ (2004) Regula- glutamate interactions: a new target for anti-
tion of working memory by dopamine D4 psychotic drugs. Neuropsychopharmacology
receptor in rats. Neuropsychopharmacology 21:S122–S133
29(9):1648 40. MacDermot J, Higashida H, Wilson SP,
27. Delgado PL, Moreno FA (1999) Role of nor- Matsuzawa H, Minna J, Nirenberg M (1979)
epinephrine in depression. J Clin Psychiatry Adenylate cyclase and acetylcholine release
61:5–12 regulated by separate serotonin receptors of
28. Casey DE, Gerlach J, Christensson E (1980) somatic cell hybrids. Proc Natl Acad Sci 76
Behavioral aspects of GABA-dopamine interre- (3):1135–1139
lationships in the monkey. Brain Res Bull 41. Guiard BP, El Mansari M, Merali Z, Blier P
5:269–273 (2008) Functional interactions between dopa-
29. Moises HC, Woodward DJ (1980) Potentia- mine, serotonin and norepinephrine neurons:
tion of GABA inhibitory action in cerebellum an in-vivo electrophysiological study in rats
by locus coeruleus stimulation. Brain Res 182 with monoaminergic lesions. Int J Neuropsy-
(2):327–344 chopharmacol 11(5):625–639
30. Bankson MG, Yamamoto BK (2004) Seroto- 42. Korsgaard S, Gerlach J, Christensson E (1985)
nin–GABA interactions modulate MDMA- Behavioral aspects of serotonin-dopamine
induced mesolimbic dopamine release. J Neu- interaction in the monkey. Eur J Pharmacol
rochem 91(4):852–859 118(3):245–252
31. Edden RA, Crocetti D, Zhu H, Gilbert DL, 43. O’hara R, Schröder CM, Mahadevan R,
Mostofsky SH (2012) Reduced GABA concen- Schatzberg AF, Lindley S, Fox S, Weiner M,
tration in attention-deficit/hyperactivity disor- Kraemer HC, Noda A, Lin X, Gray HL, Hall-
der. Arch Gen Psychiatry 69(7):750–753 mayer JF (2007) Serotonin transporter poly-
32. Russell VA, Wiggins TM (2000) Increased morphism, memory and hippocampal volume
glutamate-stimulated norepinephrine release in the elderly: association and interaction with
from prefrontal cortex slices of spontaneously cortisol. Mol Psychiatry 12(6):544–555
Computing an Escape From Depression 143
44. Connor KM, Davidson JR (1998) The role of 55. Daniel JM, Dohanich GP (2001) Acetylcholine
serotonin in posttraumatic stress disorder: neu- mediates the estrogen-induced increase in
robiology and pharmacotherapy. CNS Spectr 3 NMDA receptor binding in CA1 of the hippo-
(S2):42–51 campus and the associated improvement in
45. Charney DS, Woods SW, Goodman WK, working memory. J Neurosci 21
Heninger GR (1987) Serotonin function in (17):6949–6956
anxiety. Psychopharmacology 92(1):14–24 56. Mendoza L, Xenarios I (2006) A method for
46. Owens MJ, Nemeroff CB (1994) Role of sero- the generation of standardized qualitative
tonin in the pathophysiology of depression: dynamical systems of regulatory networks.
focus on the serotonin transporter. Clin Theor Biol Med Model 3:13
Chem 40(2):288–295 57. Thomas R (1991) Regulatory networks seen as
47. Roelands B, Meeusen R (2010) Alterations in asynchronous automata: a logical description. J
central fatigue by pharmacological manipula- Theor Biol 153(1):1–23
tions of neurotransmitters in normal and high 58. Craddock TJ, Fritsch P, Rice MA Jr, del
ambient temperature. Sports Med 40 Rosario RM, Miller DB, Fletcher MA, Klimas
(3):229–246 NG, Broderick G (2014) A role for homeo-
48. Keely SL, Lincoln TM, Corbin JD (1978) static drive in the perpetuation of complex
Interaction of acetylcholine and epinephrine chronic illness: gulf war illness and chronic
on heart cyclic AMP-dependent protein kinase. fatigue syndrome. PLoS One 9(1):e84839
Am J Phys Heart Circ Phys 234(4): 59. Fritsch P, Craddock TJ, del Rosario RM, Rice
H432–H438 MA, Smylie A, Folcik VA, de Vries G, Fletcher
49. Dixon JS, Jen PY, Gosling JA (2000) The dis- MA, Klimas NG, Broderick G (2013) Suc-
tribution of vesicular acetylcholine transporter cumbing to the laws of attraction: exploring
in the human male genitourinary organs and its the sometimes pathogenic versatility of discrete
co-localization with neuropeptide Y and nitric immune logic. Sys Biomed 1(3):179–194
oxide synthase. Neurourol Urodyn 19 60. Whitley D, Garrett D, Watson JP (2003,
(2):185–194 January) Quad search and hybrid genetic algo-
50. Walker SW, Strachan MWJ, Lightly ERT, Wil- rithms. In: Genetic and evolutionary computa-
liams BC, Bird IM (1990) Acetylcholine stimu- tion—GECCO 2003. Springer, Berlin, pp
lates cortisol secretion through the M3 1469–1480
muscarinic receptor linked to a 61. Müller N, Schwarz MJ (2007) The immune-
polyphosphoinositide-specific phospholipase mediated alteration of serotonin and gluta-
C in bovine adrenal fasciculata/reticularis mate: towards an integrated view of depression.
cells. Mol Cell Endocrinol 72(3):227–238 Mol Psychiatry 12:988–1000
51. Kalsner S, Quillan M (1988) Presynaptic inter- 62. Niciu MJ, Henter ID, Luckenbaugh DA,
actions between acetylcholine and adrenergic Zarate CA Jr, Charney DS (2014) Glutamate
antagonists on norepinephrine release. J Phar- receptor antagonists as fast-acting therapeutic
macol Exp Ther 244(3):879–891 alternatives for the treatment of depression:
52. Leboulenger F, Benyamina M, Delarue C, ketamine and other compounds. Annu Rev
Netchitailo P, Saint-Pierre S, Vaudry H Pharmacol Toxicol 54:119–139
(1988) Neuronal and paracrine regulation of 63. Morgan CA, Rasmusson AM, Wang S, Hoyt G,
adrenal steroidogenesis: interactions between Hauger RL, Hazlett G (2002) Neuropeptide-
acetylcholine, serotonin and vasoactive intesti- Y, cortisol, and subjective distress in humans
nal peptide (VIP) on corticosteroid production exposed to acute stress: replication and exten-
by frog interrenal tissue. Brain Res 453 sion of previous report. Biol Psychiatry 52
(1):103–109 (2):136–142
53. Czermak C, Staley JK, Kasserman S, Bois F, 64. Thorsell A (2010) Brain neuropeptide Y and
Young T, Henry S, Tamagnan GD, Seibyl JP, corticotropin-releasing hormone in mediating
Krystal JH, Neumeister A (2008) β2 Nicotinic stress and anxiety. Exp Biol Med 235
acetylcholine receptor availability in post- (10):1163–1167
traumatic stress disorder. Int J Neuropsycho- 65. Schubert KO, Clark SR, Van LK et al (2017)
pharmacol 11(3):419–424 Depressive symptom trajectories in late adoles-
54. Botly LC, De Rosa E (2008) A cross-species cence and early adulthood: A systematic review.
investigation of acetylcholine, attention, and Aust N Z J Psychiatry 51(5):477–499
feature binding. Psychol Sci 19 66. Anderson IM (1998) SSRIs versus tricyclic
(11):1185–1193 antidepressants in depressed inpatients: a
144 J. Tory Toole et al.
Abstract
Heart failure (HF) is a major and costly public health concern, and its prognosis is grim—with high
hospitalization and mortality rates. HF affects millions of individuals across the world, and this condition
is expected to become “the epidemic” of the twenty-first century (Jessup et al., 2016). It is well docu-
mented that individuals with HF experience disproportionately high rates of depression and that those who
are depressed have worse clinical outcomes than their nondepressed counterparts. The purpose of this
chapter is to introduce the reader to the study of depression in HF, and how psychoneuroimmunologic
principles have been applied to further elucidate mechanisms (i.e., neurohormonal and cytokine activation)
linking these comorbid disorders.
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_9, © Springer Science+Business Media, LLC, part of Springer Nature 2018
145
146 Jessica A. Jiménez et al.
2 Heart Failure
2.1 Epidemiology In the United States, among those aged 20 years and older, the
of Heart Failure prevalence of HF is 2.2% [9, 10]. At 40 years of age, the lifetime risk
of developing HF is 1 in 5. HF incidence rates in men approxi-
mately double with each 10-year age increase from 65 to 85 years;
however, the HF incidence rate triples for women between ages
65 and 84 years [11]. Researchers from the cohort study Multi-
Ethnic Study of Atherosclerosis found that African-Americans had
the highest risk of developing HF, followed by Hispanic, White,
and Chinese-Americans (4.6, 3.5, 2.4, and 1.0 per 1000 person-
years, respectively). This higher risk may reflect racial and ethnic
disparities as evidenced by the prevalence of hypertension, DM, and
low socioeconomic status among various groups [12].
Despite significant advances in treatment, the prognosis for
patients remains grim: 29.6% of HF patients die within 1 year of
diagnosis and 50% die within 5 years [13]. Approximately three
million patients get hospitalized each year with a primary or sec-
ondary diagnosis of HF making it one of the most common causes
of hospitalization in the elderly population in the United States
[14]. HF costs approximately $31 billion, and is projected to cost a
total of $70 billion by 2030 [11].
2.2 Definition The American College of Cardiology (ACC) and American Heart
and Classification Association (AHA) define HF as a complex clinical syndrome that
of Heart Failure can result from any structural or functional cardiac disorder that
impairs the ability of the ventricles to fill with or eject blood
[15]. The upper chambers of the heart are composed of the right
and left atria, and the lower chambers include the right and left
ventricles. The ventricles are muscular chambers that contract to
pump blood (systole). After systole, the ventricle muscles normally
relax during diastole, allowing blood from the atria to fill the
ventricles. The heart’s ability to pump can be compromised via
two mechanisms: (1) reduction in the volume of oxygenated
blood ejected from the left ventricle as a result of diminished
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 147
Table 1
Functional classifications and disease progression stages of heart failure
Definition Examples
NYHA No limitation of physical activity Ordinary physical activity does not cause
Class I undue fatigue, palpitation, or dyspnea
(shortness of breath)
NYHA Slight limitation of physical activity Comfortable at rest, but ordinary physical
Class activity results in fatigue, palpitation, or
II dyspnea
NYHA Marked limitation of physical activity Comfortable at rest, but less than ordinary
Class activity causes fatigue, palpitation, or
III dyspnea
NYHA Unable to carry out any physical activity Symptoms of cardiac insufficiency at rest. If
Class without discomfort any physical activity is undertaken,
IV discomfort is increased
American College of Cardiology/American Heart Association stages of heart failure
Definition Examples
Stage A High risk for developing HF, but without Hypertension, diabetes mellitus, CAD, family
structural heart disease or symptoms of HF history of cardiomyopathy
Stage B Structural heart disease, but asymptomatic Previous myocardial infarction, left ventricular
dysfunction, valvular heart disease
Stage C Structural heart disease with previous or Structural heart disease, dyspnea and fatigue,
current symptoms, but managed with impaired exercise tolerance
medical treatment
Stage D Marked symptoms at rest despite maximal Advanced disease requiring hospital-based
medical therapy support, a heart transplant, or palliative care
3.1 Renin- The RAAS system maintains renal blood flow after the myocardium
Angiotensin- has sustained injury via its effects on remodeling the vasculature
Aldosterone System and increasing plasma volume. Decreased renal perfusion pressure
results in secretion of renin by juxtaglometer cells lining the affer-
ent renal arterioles. Specifically, renin cleaves angiotensinogen to
form decapeptide angiotensin I. The angiotensin-converting
enzyme (ACE) cleaves two C terminal amino acids to form angio-
tensin II, the primary effector of the system. Receptors for angio-
tensin II are divided into subtypes, AT-1 and AT2. AT-1 is the
predominate subtype in the vascular endothelium and a primary
target for pharmacologic blockade. Binding of angiotensin II to
AT-1 receptors results in increases in the release of intracellular
calcium from the sarcoplasma reticulum via activation of protein
kinase C. The binding of angiotensin II in the vasculature results in
an increase in systematic vasculature resistance and restoration of
blood pressure.
Prolonged compensatory actions of the RAAS in HF bring
adverse consequences, however, including increased vascular resis-
tance. Increases in resistance create undue myocardial burden and
decreased cardiac output, resulting in left ventricular hypertrophy.
Angiotensin II initiates apoptosis and interstitial fibrosis, which
contribute to the remodeling of the extracellular matrix in the
myocardium (e.g., myocyte hypertrophy). The effects of angioten-
sin II on the myocardium and peripheral vasculature result in
decreased cardiac output and renal perfusion. Angiotensin II is
also involved in the increase of plasma volume by initiating produc-
tion of mineralocorticoid aldosterone by the adrenal cortex. Aldo-
sterone acts on the distal tubules of the renal nephron and activates
a sodium-potassium exchange, which results in the retention of
sodium and water. The increased plasma volume exacerbates fluid
overload and peripheral edema. Chronic excess of aldosterone leads
to increased fibrosis in the atria, ventricles, kidneys, and
perivasculature.
4.1 Epidemiology Depression in HF has been extensively studied because of its high
of Depression in Heart prevalence in HF patients and its tendency to worsen medical
Failure Patients prognosis [1–6]. Studies show that depression is one of the most
important modifiable risk factors because it is responsible, in part,
for the high readmission rates among HF patients. The estimated
prevalence of depression in HF is 24–42%, which is 2–3 times
higher than the general population. The prevalence rate of minor
depression, as defined by the Beck Depression Inventory 10, is as
high as 35.5%. Although the majority of studies demonstrate that
depression worsens cardiac prognosis, the magnitude of the effects
has varied greatly depending on how depression is measured [2].
Depression has also been associated with incident HF. In a
community sample of 2501 patients (mean follow-up 14 years),
Williams and colleagues found that depression was an independent
predictor of developing HF in women (HR ¼ 1.96, 95% CI ¼ 1.11,
3.46, p ¼ 0.02), but not in men [7]. In a study of 4538 older adult
patients enrolled in the Systolic Hypertension in the Elderly Pro-
gram (SHEP), depressed patients were 2.82 times (95% CI ¼ 1.71,
4.67; p < 0.001) more likely to develop HF over a 4.5-year follow-
up period [8]. The association between depression and HF did not
significantly vary by sex. Although not all studies have found gen-
der effects, Williams et al.’s findings follow trends seen in
depressed, non-HF populations: the National Comorbidity Survey,
for example, reported a 1.7 greater odds of women developing
depression at some point in their lifetime compared to men [19].
4.3 Clinical Of great clinical significance, studies find that depression has
Outcomes in Heart adverse effects on the course and prognosis of HF. Increased psy-
Failure Patients chological surveillance of HF patients over the past 15 years has
with Depression highlighted the pivotal role of depression in HF [23]. Development
of depressive symptoms after the diagnosis of HF is associated with
a 1.5- to 2.2-fold increase in all-cause and cardiovascular mortality
[24]. Depression is an independent risk factor for hospital readmis-
sion in patients with HF [25–27]. In the Telemonitoring to
Improve Heart Failure Outcomes Trial, the 30-day readmission
rate was 17.1% [28]. Kato et al. found that depression was asso-
ciated with more cardiac mortality or HF hospitalization in both
HFrEF (55% vs. 12%, p < 0.01) and HFpEF (35% vs. 11%,
p < 0.05; [29]). Sherwood et al. [30] reported that depressive
symptomatology was associated with a 1.56 (95% CI; 1.07, 2.29;
p < 0.001) increased risk of death or hospitalization during a
median 3-year follow-up period. In a sample of 374 patients hospi-
talized for HF, Jiang and colleagues [31] found that HF patients
with major depression had 2.23 greater odds (95% CI 0.04, 4.77;
p ¼ 0.04) of mortality and 3.07 (95% CI 1.41, 6.66; p ¼ 0.005)
greater odds of readmission at 1 year compared to HF patients with
no depression. In a sample of 1006 hospitalized HF patients, Jiang
et al. [32] found that patients whose Beck Depression Inventory
(BDI) scores were 5 to 9, 10 to 18, and 19 were 21%, 53%, and
83%, respectively, more likely to die than patients whose BDI score
was 5 ( p < 0.001). Vaccarino et al. [33] also found that there was
a grade associated between the number of depressive symptoms and
increased risk of death or decline of daily living at 6 months. In this
prospective study of 391 patients with decompensated HF on
admission to the hospital, patients with 11 depressive symptoms,
compared with those with <6 depressive symptoms, had an 82%
higher risk of either functional decline or death. In a study of
longitudinal outcomes (mean follow-up 39 months) in HF patients
with comorbid atrial fibrillation, Frasure-Smith and colleagues [34]
found that elevated depressive symptoms significantly predicted
cardiovascular mortality (HR: 1.57; 95% CI 1.20, 2.07;
p < 0.001). The authors also commented that the increased risk
of death was similar to risks associated with not taking standard
medications to manage HF, such as anticoagulants and aldosterone
antagonists. Worsening depressive symptomatology over time is
also associated with increased risk of adverse outcomes. A study of
147 HF outpatients found that a 1 point change in BDI scores was
associated with 1.07 increased risk of death or cardiovascular hos-
pitalization (95% CI 1.02, 1.12, p ¼ 0.007) [24]. The results from
these studies indicate that depression is an independent predictor of
worse clinical outcomes in HF patients [23].
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 153
4.4 Quality of Life In addition to its cardiotoxic effects, depressed HF patients suffer
in Heart Failure from reduced physical functioning and of course worse quality of
Patients life: depressed patients report poorer quality of life and greater
with Depression functional impairment than nondepressed patients, even when
compared with patients of a higher (i.e., worse) NYHA functional
class, which may suggest that patient perceptions of physical func-
tioning, rather than the clinical status itself, predict quality-of-life
outcomes [35, 36]. In a small study (n ¼ 58) of associations
between disease severity, functional status, depression, and daily
quality of life, greater depression severity was positively associated
with worse self-reported physical and emotional quality of life in
HF patients. A recent study by Hallas et al. [37] conducted in
146 HF patients found that patients with more negative beliefs
about the consequences of HF, and less perceived control, were
more anxious and depressed compared to patients with more posi-
tive beliefs. Greater depression ratings also predicted poorer quality
of life. Patients with more negative beliefs also had more maladap-
tive behaviors and less coping resources, which may also have
downstream effects on quality of life. In a study of 155 HF patients,
Gottlieb et al. [21] found that depressed patients scored signifi-
cantly worse than nondepressed patients on all components of the
quality-of-life questionnaires. In a more recent study, Gottlieb and
colleagues [38] demonstrated that depression is minimally related
to objective assessments of HF severity, such as peak O2 consump-
tion, B-type natriuretic peptide (BNP) levels, or ejection fraction.
However, depression significantly affects subjective measurements
of HF severity, such as NYHA classification or 6-min walk test
[10]. Undoubtedly, depression negatively affects quality of life,
but there is building evidence to also suggest that depression alters
patient perceptions of physical functioning and disease severity,
which may result in poorer ratings on subjective measures.
4.5 Screening Given that depressive symptoms are a strong predictor of HF out-
for Depression in Heart comes, screening has become an important part of patient assess-
Failure Populations ment [20]. The American Heart Association (AHA) advocates a
two-step screening process using two versions of the Patient Health
Questionnaire (PHQ): the PHQ-2, which includes two items from
the PHQ-9 is administered as a first-line screening for depression.
Positive results on the PHQ-2, the PHQ-9 is subsequently admi-
nistered for further evaluation [39]. The usefulness of the PHQ-2
or the PHQ-9 results as a predictor of prognosis in patients with
HF has been examined in many studies [40–42]. However, it is
unclear if there is any advantage in using a two-step screening
process as opposed to using the PHQ-2 alone or the PHQ-9
alone [43].
154 Jessica A. Jiménez et al.
5.1 Lessons Learned One of the main lines of investigation in our laboratory has been
on Inflammatory the application of the cytokine model and other theoretical models
and Other Immune of broader immune activation in the context of depression in
Responses HF. We have conducted several studies that lend support to the
in Depression in HF theory that systematic inflammation, as well as broader dysregula-
tion of immune system, may underlie the relationship between
depressive symptoms and progression of HF. Our approach to the
study of these topics is the use of both cross-sectional and prospec-
tive studies. For the former, we have assessed a broad range of
inflammatory and cell adhesion biomarkers in patients with estab-
lished HF and with a range of depressive symptomatology. For the
latter, we are studying ACC/AHA Stage B patients, individuals
who were at risk for developing symptomatic HF but who were
not symptomatic. In these individuals, we are assessing a broad
range of inflammatory and cell adhesion biomarkers as well as
depression, but this time repeatedly over the course of several
years. The intention is to temporally model inflammation as it
relates to the onset and offset of depression and to the onset and
progression from non-symptomatic to symptomatic HF.
5.2 Assessing A potential explanation for the link between depression and HF
Depressive Symptoms may be inherent due to the diagnosis of depression, and measure of
its severity. The diagnostic criteria for major depressive disorder as
well as screening for depressive symptoms include both cognitive
and somatic symptoms. Thus, depressive symptomatology, such as
fatigue, loss of energy, problems concentrating, weight loss or gain,
and sleep disturbance, may be the result of underlying cardiac
dysfunction [5]. More recent studies have recognized this overlap,
and now it is favored to report somatic and cognitive symptomatol-
ogy separately as we have done in our studies.
In our studies we have primarily measured depression via the
Beck Depression Inventory (version–IA; BDI-IA), which is a
21-item self-administered assessment of extent to which patients
experience depressive symptoms [50]. Scores of 0–9 indicate mini-
mal or no depression, 10–18 mild-moderate depression, 19–29
moderate-severe depression, and 30–63 severe depression. The
reliability of this measure in our samples has been α > 0.90. A
BDI score of less than 10 is usually not considered to be clinically
significant depression, but yet studies have shown scores of 4–9 to
be associated with increased mortality in post-myocardial infarction
patients [2]. Since there has been some evidence to suggest that
cognitive/affective and somatic aspects of depression differentially
156 Jessica A. Jiménez et al.
5.3 Assessing Circulating TNF-α, IL-6, and IL-1β levels are determined in plasma
Cytokines and Cellular by commercial ELISA. For IL-6, the intra-assay CV (%) is 2.2, the
Adhesion Molecules inter-assay CV (%) is 3.9, and the assay sensitivity is <0.71 pg/mL;
for TNF-α, the intra-assay CV (%) is 8.0, the inter-assay CV (%) is
16.3, and the assay sensitivity is <0.18 pg/mL; for IL-1β, the intra-
assay CV (%) is 6.8, the inter-assay CV (%) is 8.3, and the assay
sensitivity is <0.1 pg/mL.
Circulating CRP levels are determined in plasma using com-
mercial ELISA. Intra- and inter-assay coefficients of variation are
<5%. Precision and sensitivity performance values are excellent:
intra-assay CV (%) < 1.0, inter-assay CV (%) ¼ 1.6, and sensitivity
<0.05 mg/L.
Soluble intercellular adhesion molecule-1 (sICAM-1, sCD54)
and sP-selectin (sCD62P) are too determined by commercial
ELISA. The precision and sensitivity performance values are as
follows: sICAM-1 (intra-assay CV (%) ¼ 4.6, inter-assay CV
(%) ¼ 6.6, sensitivity <0.35 ng/mL) and sCD62P (intra-assay
CV (%) ¼ 5.1, inter-assay CV (%) ¼ 8.8, sensitivity <0.5 ng/mL).
Proinflammatory cytokines, such as TNF, IL-6, and CRP, are
associated with cardiac dysfunction in both human and animal
models. However, in 2009 Wirtz et al. [52] in our group provided
the first study investigating whether depressive symptoms were
associated with exercise-induced increases in circulating levels of
adhesion molecules expressed on endothelial cells (sP-selectin and
soluble sICAM-1), leukocytes (sICAM-1), and platelets
(sP-selectin). Using data from 39 middle-aged male HF patients
and 19 male control subjects, the authors found that higher depres-
sion symptomatology moderated greater increases in sP-selectin
levels in response to an acute exercise challenge over time in HF
patients as compared with control subjects (F ¼ 3.25, p ¼ 0.05).
Post hoc testing revealed that in HF patients, higher depression
scores (BDI) were significantly associated with greater increases in
sP-selectin levels over time in response to the exercise (Fig. 1). Also,
in the HF patients, higher BDI scores were associated with higher
sP-selectin levels at pre-exercise and post-exercise time points (main
effect of BDI: F ¼ 4.86, p ¼ 0.035). These effects were not found
for the control subjects. These findings suggested that levels of
sP-selectin are higher before and after exercise and have greater
increases in response to exercise in male HF patients with increasing
depressive symptom severity. These findings could have
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 157
90
Low BDI score (n=23)
High BDI score (n=15)
70
60
50
40
30 Exercise
-5 0 5 10 15 20 25 30 35
Time in minutes
5.4 Assessing Until recently, immune cell migration, particularly chemotaxis, has
Chemotaxis been largely ignored in respect to depression symptoms and
and Cellular Immunity HF. Chemokines are essential for providing signaling to leukocytes
for extravasation from the blood and directing locomotion
[54, 55]. When overexpressed, recruitment and migration factors
are injurious to the cardiovascular system [56] and can generate
angiogenesis and fibrous tissue deposition, which can lead to myo-
cardial dysfunction in HF [57, 58]. Particularly, studying acute
physiologic responses to controlled challenges serves as a window
158 Jessica A. Jiménez et al.
0.20
0.15
0.10
Chemotaxis
Stimulation 0.05
Index (log
transformed) 0.00
-0.05
-0.10
-0.15
Pre- exercise Post- exercise
CHF- hi BDI
CHF- lo BDI
non CHF- hi BDI
non CHF- lo BDI
0.15
0.10
0.05
Chemotaxis
Stimulation 0.00
Index (log
transformed) -0.05
-0.10
-0.15
-0.20
-0.25
Pre- exercise Post- exercise
CHF- hi BDI
CHF- lo BDI
non CHF- hi BDI
non CHF- lo BDI
Fig. 2 Changes in a logarithmic transformed stimulation index (SI) of chemotaxis to fMLP (TOP PANEL) and the
chemokine SDF-1 (BOTTOM PANEL) in HF patients and non-HF controls with high (hi) and low (lo) Beck
Depression Inventory (BDI) scores in response to exercise. Data are expressed as means SEM
160 Jessica A. Jiménez et al.
0.20
0.15
0.10
Chemotaxis 0.05
Stimulation
Index (log
0.00
transformed)
-0.05
-0.10
-0.15
1 nM 10 nM 100 nM
Isoproteronol
non-HF lo BDI
non-HF hi BDI
HF lo BDI
HF hi BDI
Fig. 3 Change scores (pre- minus post-exercise) and chemotaxis to three concentrations of isoproterenol
(1 nM, 10 nM, and 100 nM/l) in HF patients and non-HF controls. High (hi) versus low (lo) depression are
determined by scores 10 and <10 on the Beck Depression Inventory (BDI). Data expressed as means
SEM
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 161
5.5 Th1/Th2 Ratio Although, as we have discussed, inflammatory cytokines have been
implicated as a possible mediator of psychological symptoms of
depression and HF, it has been unclear if systematic inflammation
represents a broader dysregulation of the immune system. Particu-
larly, cellular immunity is important for protection against infec-
tion. Th1 cells promote cellular immunity by rapidly producing a
range of cytokines such as IFN-γ that activate other Th1 cells to
fight infectious agents. Th1 cells also exert a negative regulatory
role on Th2 cells that produce cytokines such as IL-4 and IL-10.
Th2 cytokines, on the other hand, attenuate immune defenses if
they are locally over-expressed, by decreasing activities of major
effectors such as Th1 cells [63, 64]. A Th2 shift may have a
profound effect on the susceptibility of the organism to infection
[65], increase inflammation, and lead to dilated cardiomyopathy
and HF [66]. Maintaining Th1/Th2 homeostasis is important for
preserving health. Thus, examining Th1/Th2 ratios can provide
information on the balance of cellular immune activation versus
negative regulation of cellular immunity.
Redwine et al. [67] examined the relationship of depressive
symptoms with cellular immune activity measured by the
Th1/Th2 ratio and cardiac rehospitalization and/or death in
18 HF patients (mean age ¼ 62, NYHA classes II–IV). The authors
reported that higher baseline depression scores were associated
with a prospective increase in incidence of cardiac-related hospita-
lizations and/or death ( p ¼ 0.037). Lesser IFN-γ/IL-10-expres-
sing CD4+ T cell ratios were related to higher depressive symptom
scores at baseline ( p ¼ 0.005, Fig. 4) and a prospective increased
incidence of cardiac-related hospitalization or death over a 2-year
period ( p ¼ 0.05). The results suggest that a shift in the Th1/Th2
ratio may play a role in the association between depressive symp-
toms and morbidity and mortality in HF patients, suggesting
broader immune dysregulation.
162 Jessica A. Jiménez et al.
3.5
3.0
IFN gamma/IL-10 ratio
2.5
2.0
1.5
1.0
.5
0.0
0 10 20 30 40
HAM-D Score
Fig. 4 Relationship between IFN-γ/IL-10 ratios and Hamilton Depression Scores in HF patients
5.6 Alterations A burgeoning new area of interest in our laboratory is the role of
of the Gut, Gut the “heart-gut axis” and the “gut hypothesis” of HF, which are
Microbiota, recently emerging concepts that attempt to address the complex
and Metabolites pathophysiology of HF. HF is associated with altered gastrointesti-
in Heart Failure nal function due at least in part to ischemic conditions and conges-
tion within the gut. The host may then be affected by impaired gut
barrier function, resulting in systemic inflammatory responses and
oxidative stress. In the context of the microbiota, altered composi-
tion and/or function may influence metabolic processes in HF. For
example, the gut microbiota metabolizes dietary choline and
L-carnitine into trimethylamine which is then oxidized to trimethy-
lamine N-oxide (TMAO) in the liver [68]. Circulating TMAO is an
important predictive risk factor for cardiovascular disease and has
been observed as elevated in HF patients compared to age- and
sex-matched controls [69, 70]. Increased TMAO levels are also
associated with a greater-than-threefold increase in mortality risk
[70]. In addition, plasma concentrations of TMAO, choline, and
betaine have been observed as elevated in patients with chronic HF
compared to control subjects, and patients with New York Heart
Association classes III and IV have the highest levels [71]. More-
over, elevated TMAO has been observed as anticorrelated to long-
term survival in HF patients even after controlling for cardiorenal
parameters. As TMAO is detoxified through the kidneys, renal
functional status may affect TMAO levels. In HF patients, TMAO
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 163
5.7 Altered Plasma HF is associated with depression, and key differences have been
Metabolites in Patients identified in the metabolome of depressed compared to nonde-
with Heart Failure pressed HF patients [89]. The amino acids aspartate and serine
and Depression are significantly elevated in depressed HF patients compared to
controls. Aspartate, the main excitatory amino acid in the central
nervous system (CNS), can activate, while serine functions as
co-agonist, N-methyl D-aspartate (NMDA) receptors, which are
cation channels that mediate fast synaptic transmission in the brain
and are important in memory and behavioral functioning. Elevated
concentrations of these amino acids in the synaptic space may
promote selective neuronal loss and various chronic neurological
disorders [90]. The amino acids leucine and isoleucine are also
elevated in depressed compared to nondepressed HF patients.
These branch-chain amino acids have stress response and protein
regulatory functions. Leucine is also relevant to key brain metabolic
functions.
In addition, numerous dicarboxylic acid (DCA) species, which
are produced in the context of fatty acid oxidation dysfunction or
saturation of mitochondrial β-oxidation, are elevated in depressed
compared to control HF patients. Decreased concentrations of the
ketone body 3-hydroxybutyrate, which are used as energy for the
CNS and generated through β-oxidation of fatty acids, are found in
the depressed subjects. During β-oxidation of lipids, the acetyl-
CoA generated may then enter the tricarboxylic acid cycle (TCA);
however excess concentrations of acetyl-CoA will saturate the TCA
cycle and contribute to the formation of ketone bodies such as
3-hydroxybutyrate. Thus, concentrations of 3-hydroxybutyrate
are decreased while DCA species are increased in depressed com-
pared to nondepressed HF patients, which suggests a metabolic
shift away from the primary route of fatty acid metabolism via
β-oxidation toward ω-oxidation of fatty acids in which the
ω-carbon is oxidized to carboxylic acid to form DCA. These results
are consistent with previous associations between altered amino
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 165
6 Conclusions
References
1. Shapiro PA (2017) Psychiatric aspects of heart 2. Ghosh RK, Ball S, Prasad V et al (2016)
disease and cardiac aspects of psychiatric disease Depression in heart failure: intricate relation-
in Critical Care. Crit Care Clin 33:619–634 ship, pathophysiology and most updated
166 Jessica A. Jiménez et al.
evidence of interventions from recent clinical for Heart and Lung Transplantation. J Am
studies. Int J Cardiol 224:170–177 Coll Cardiol 53:e1–e90
3. Kent LK, Shapiro PA (2009) Depression and 16. Inamdar AC (2016) Heart failure: diagnosis,
related psychological factors in heart disease. management and utilization. J Clin Med 5(7)
Harv Rev Psychiatry 17:377–388 17. Ponikowski P et al (2016) 2016 ESC guide-
4. Norra C, Skobel EC, Arndt M et al (2008) lines for the diagnosis and treatment of acute
High impact of depression in heart failure: and chronic heart failure. Eur Heart J
early diagnosis and treatment options. Int J 18:891–975
Cardiol 125:220–231 18. Bristow MR (2000) β-adrenergic receptor
5. Watson K, Summer KM (2009) Depression in blockade in chronic heart failure. Circulation
patients with heart failure: clinical implications 101:558–569
and management. Pharmacotherapy 29:49–63 19. Kessler RC, Berglund P, Demler O et al (2003)
6. Shapiro PA (2009) Treatment of depression in The epidemiology of major depressive disor-
patients with congestive heart failure. Heart der: results from the National Comorbidity
Fail Rev 14:7–12 Survey Replication (NCS-R). JAMA
7. Williams SA, Kasl SV, Heiat A et al (2002) 289:3095–3105
Depression and risk of heart failure among the 20. Rutledge T, Reis VA, Linke SE et al (2006)
elderly: a prospective community-based study. Depression in heart failure: a meta-analytic
Psychosom Med 64:6–12 review of prevalence, intervention effects, and
8. Abramson J, Berger A, Krumholz HM et al associations with clinical outcomes. J Am Coll
(2001) Depression and risk of heart failure Cardiol 48:1527–1537
among older persons with isolated systolic 21. Gottlieb SS, Khatta M, Friedmann E et al
hypertension. Arch Intern Med (2004) The influence of age, gender, and race
161:1725–1730 on the prevalence of depression in heart failure
9. Bui AL, Horwich TB, Fonarow GC (2010) patients. J Am Coll Cardiol 43:1542–1549
Epidemiology and risk profile of heart failure. 22. Havranek EP, Spertus JA, Masoudi FA et al
Nat Rev Cardiol 8:30–41 (2004) Predictors of the onset of depressive
10. Roger VL, Go AS, Lloyd-Jones DM et al symptoms in patients with heart failure. J Am
(2010) Heart disease and stroke statistics— Coll Cardiol 44:2333–2338
2011 update: a report from the American 23. Tousoulis D, Antonopoulos AS, Antoniades C
Heart Association. Circulation 123:18–209 et al (2010) Role of depression in heart fail-
11. Mozaffarian D, Benjamin EJ, Go AS et al ure—choosing the right antidepressive treat-
(2016) Executive summary: heart disease and ment. Int J Cardiol 140:12–18
stroke statistics—2016 update: a report from 24. Fan H, Yu W, Zhang Q et al (2014) Depression
the American Heart Association. Circulation after heart failure and risk of cardiovascular and
133:447–454 all-cause mortality: a meta-analysis. Prev Med
12. Bahrami H, Kronmal R, Bluemke DA et al 63:36–42
(2008) Differences in the incidence of conges- 25. Chaudhry S, McAvay G, Chen S et al (2013)
tive heart failure by ethnicity: the Multi-Ethnic Risk factors for hospital admission among older
Study of Atherosclerosis. Arch Intern Med persons with newly diagnosed heart failure:
168:2138–2145 findings from the cardiovascular health study.
13. Levy D, Kenchaiah S, Larson MG et al (2002) J Am Coll Cardiol 61:635–642
Long-term trends in the incidence of and sur- 26. Vader JM, LaRue SJ, Stevens SR (2016)
vival with heart failure. N Engl J Med Timing and causes of readmission after acute
347:1397–1402 heart failure hospitalization—insights from the
14. Go AS, Mozaffarian D, Roger VL et al (2013) heart failure network trials. J Card Fail
Heart disease and stroke statistics—2013 16:30059–30058
update: a report from the American Heart 27. Freedland KE, Carney RM, Rich MW et al
Association. Circulation 127:e6–e245 (2016) Depression and multiple
15. Hunt SA, Abraham WT, Chin MH et al (2009) re-hospitalizations in patients with heart fail-
2009 Focused update incorporated into the ure. Clin Cardiol 39:257–262
ACC/AHA 2005 guidelines for the diagnosis 28. Krumholz HM, Chaudhry SI, Spertus JA et al
and management of heart failure in adults: a (2015) Do nonclinical factors improve predic-
report of the American College of Cardiology tion of readmission risk? Results from the tele-
Foundation/American Heart Association Task HF study. JACC Heart Fail 15:653–658
Force on Practice Guidelines Developed in 29. Kato N, Kinugawa K, Shiga T et al (2012)
Collaboration with the International Society Depressive symptoms are common and
Neuroimmune Mechanisms of Depression in Adults with Heart Failure 167
associated with adverse clinical outcomes in 40. Rollman BL, Herbeck Belnap B, Mazumdar S
heart failure with reduced and preserved ejec- et al (2012) A positive 2-item Patient Health
tion fraction. J Cardiol 60:2330 Questionnaire depression screen among hospi-
30. Sherwood A, Blumenthal JA, Trivedi R et al talized heart failure patients is associated with
(2007) Relationship of depression to death or elevated 12-month mortality. J Card Fail
hospitalization in patients with heart failure. 18:238–245
Arch Intern Med 167:367–373 41. Lee KS, Lennie TA, Heo S, Moser DK (2012)
31. Jiang W, Alexander J, Christopher E et al Association of physical versus affective depres-
(2001) Relationship of depression to increased sive symptoms with cardiac event-free survival
risk of mortality and rehospitalization in in patients with heart failure. Psychosom Med
patients with congestive heart failure. Arch 74:452–458
Intern Med 161:1849–1856 42. Moraska AR, Chamberlain AM, Shah ND et al
32. Jiang W, Kuchibhatla M, Clary GL et al (2007) (2013) Depression, healthcare utilization, and
Relationship between depressive symptoms death in heart failure: a community study. Circ
and long-term mortality in patients with heart Heart Fail 6:387–394
failure. Am Heart J 154:102–108 43. Lee KS, Moser DK, Pelter M et al (2017)
33. Vaccarino V, Kasl SV, Abramson J et al (2001) Two-step screening for depressive symptoms
Depressive symptoms and risk of functional and prediction of mortality in patients with
decline and death in patients with heart failure. heart failure. Am J Crit Care 26:240–247
J Am Coll Cardiol 38:199–205 44. O’Connor CM, Jiang W, Kuchibhatla M et al
34. Frasure-Smith N, Lesperance F, Habra M et al (2010) Safety and efficacy of sertraline for
(2009) Elevated depression symptoms predict depression in patients with heart failure: results
long-term cardiovascular mortality in patients of the SADHART-CHF (sertraline against
with atrial fibrillation and heart failure. Circu- depression and heart disease in chronic heart
lation 120:134–140 failure) trial. J Am Coll Cardiol 56:692–699
35. Sullivan M, Levy WC, Russo JE et al (2004) 45. Jiang W, Krishnan R, Kuchibhatla M et al
Depression and health status in patients with (2011) Characteristics of depression remission
advanced heart failure: a prospective study in and its relation with cardiovascular outcome
tertiary care. J Card Fail 10:390–396 among patients with chronic heart failure
36. Jiang W, Kuchibhatla M, Cuffe MS et al (2004) (from the SADHART-CHF study). Am J Car-
Prognostic value of anxiety and depression in diol 107:545–551
patients with chronic heart failure. Circulation 46. Freedland KE, Carney RM, Rich MW et al
110:3452–3456 (2015) Cognitive behavior therapy for depres-
37. Hallas CN, Wray J, Andreou P, Banner NR sion and self-care in heart failure patients: a
(2011) Depression and perceptions about randomized clinical trial. JAMA Intern Med
heart failure predict quality of life in patients 175:1773–1782
with advanced heart failure. Heart Lung 47. Maes M (2001) Psychological stress and the
40:111–121 inflammatory response system. Clin Sci
38. Gottlieb SS, Kop WJ, Ellis SJ et al (2009) 101:193–194
Relation of depression to severity of illness in 48. Penninx BW, Kritchevsky SB, Yaffe K et al
heart failure (from Heart Failure and a Con- (2003) Inflammatory markers and depressed
trolled Trial Investigating Outcomes of Exer- mood in older persons: results from the health,
cise Training [HF-ACTION]). J Am Coll aging and body composition study. Biol Psy-
Cardiol 103:1285–1289 chiatry 54:566–572
39. Lichtman JH, Bigger JT, Blumenthal JA et al 49. Bremmer MA, Beekman AT, Deeg DJ et al
(2008) Depression and coronary heart disease: (2008) Inflammatory markers in late-life
recommendations for screening, referral, and depression: results from a population-based
treatment: a science advisory from the Ameri- study. J Affect Disord 106:249–255
can Heart Association Prevention Committee 50. Beck A, Steer RA (1987) Manual for the
of the Council on Cardiovascular Nursing, revised Beck Depression Inventory. Psycholog-
Council on Clinical Cardiology, Council on ical Corportation, San Antonio
Epidemiology and Prevention, and Interdisci- 51. Linke SE, Rutledge T, Johnson BD et al
plinary Council on Quality of Care and Out- (2009) Depressive symptom dimensions and
comes Research: endorsed by the American cardiovascular prognosis among women with
Psychiatric Association. Circulation suspected myocardial ischemia: a report from
118:1768–1775 the National Heart, Lung, and Blood Institute-
168 Jessica A. Jiménez et al.
78. Qin J et al (2012) A metagenome-wide associ- 86. Sandek A et al (2014) Intestinal blood flow in
ation study of gut microbiota in type 2 diabetes. patients with chronic heart failure: a link with
Nature 490:55–60 bacterial growth, gastrointestinal symptoms,
79. Cani PD et al (2009) Changes in gut micro- and cachexia. J Am Coll Cardiol
biota control inflammation in obese mice 64:1092–1102
through a mechanism involving GLP-2-driven 87. Levine B, Kalman J, Mayer L et al (1990)
improvement of gut permeability. Gut Elevated circulating levels of tumor necrosis
58:1091–1103 factor in severe chronic heart failure. N Engl J
80. Charalambous BM, Stephens RC, Feavers IM Med 323:236–241
et al (2007) Role of bacterial endotoxin in 88. Maeda K et al (2000) High levels of plasma
chronic heart failure: the gut of the matter. brain natriuretic peptide and interleukin-6
Shock 28L:15–23 after optimized treatment for heart failure are
81. Sandek A, Rauchhaus M, Anker SD et al independent risk factors for morbidity and
(2008) The emerging role of the gut in chronic mortality in patients with congestive heart fail-
heart failure. Curr Opin Clin Nutr Metab Care ure. J Am Coll Cardiol 36:1587–1593
11:632–639 89. Steffens DC et al (2010) Metabolomic differ-
82. Sandek A et al (2012) Studies on bacterial ences in heart failure patients with and without
endotoxin and intestinal absorption function major depression. J Geriatr Psychiatry Neurol
in patients with chronic heart failure. Int J 23:138–146
Cardiol 157:80–85 90. Meldrum BS (1986) Drugs acting on amino
83. Nagatomo Y, Tang WH (2015) Intersections acid neurotransmitters. Adv Neurol
between microbiome and heart failure: revisit- 43:687–706
ing the gut hypothesis. J Card Fail 21:973–980 91. Riegel B, Lee CS, Dickson VV (2011) Self care
84. Sandek A, Anker SD, von Haehling S (2009) in patients with chronic heart failure. Nat Rev
The gut and intestinal bacteria in chronic heart Cardiol 8:644–654
failure. Curr Drug Metab 10:22–28 92. Jaarsma T et al (2004) Design and methodol-
85. Niebauer J et al (1999) Endotoxin and ogy of the COACH study: a multicenter ran-
immune activation in chronic heart failure: a domised coordinating study evaluating
prospective cohort study. Lancet outcomes of advising and counselling in heart
353:1838–1842 failure. Eur J Heart Fail 6:227–233
Chapter 10
Abstract
The clinical management of patients affected by systemic diseases, including cancer and autoimmune
diseases, is generally founded on the evaluation of the only markers related to the single disease rather
than the biological immuno-inflammatory response of patients, despite the fundamental role of cytokine
network in the pathogenesis of cancer and autoimmunity is well known. Cancer progression has appeared to
be associated with a progressive decline in the blood levels of the main antitumor cytokines, including IL-2
and IL-12, in association with an increase in those of inflammatory cytokines, including IL-6, TNF-alpha,
and IL-1-beta, and immunosuppressive cytokines, namely TGF-beta and IL-10. On the other hand, the
severity of the autoimmune diseases has been proven to be greater in the presence of high blood levels of
IL-17, TNF-alpha, IL-6, IL-1-beta, IFN-gamma, and IL-18, in association with low levels of TGF-beta and
IL-10. However, because of excessive cost and complexity of analyzing the data regarding the secretion of
the single cytokines, the relation between lymphocyte-induced immune activation and monocyte-
macrophage-mediated immunosuppression has been recently proven to be expressed by the simple
lymphocyte-to-monocyte ratio (LMR). The evidence of low LMR values has appeared to correlate with a
poor prognosis in cancer and with a disease control in the autoimmune diseases. Moreover, since the in vivo
immunoinflammatory response is physiologically under a neuroendocrine modulation, for the evaluation of
patient biological response it would be necessary to investigate the function of at least the two main
neuroendocrine structures involved in the neuroendocrine modulation of the immune responses, consist-
ing of the hypothalamic-pituitary-adrenal axis and the pineal gland, since the lack of physiological circadian
rhythm of cortisol and pineal hormone melatonin has appeared to be associated with a worse prognosis in
the human systemic diseases.
Key words Autoimmunity, Biological response, Cancer, Cannabinoid system, Cytokine network,
Immunotherapy, Lymphocyte-to-monocyte ratio, Melatonin, Neuroimmunomodulation, Opioid sys-
tem, Pineal gland, Synchronization
1 Introduction to Psycho-Neuro-Endocrino-Immunology
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_10, © Springer Science+Business Media, LLC, part of Springer Nature 2018
171
172 Paolo Lissoni et al.
1.1 The Biological It is a general common opinion that the biological response is
Response consisting only of the acute and chronic immuno-inflammatory
reactions. On the contrary, since both the immune and the inflam-
matory responses are under a physiological psychoneuroendocrine
regulatory control, the investigation of host biological response in
the human systemic diseases would have to include at least the
evaluation of the immune functionless, the degree of the inflamma-
tory response, and the functional status of the main nervous and
endocrine structures involved in the neuroendocrine modulation
on the biological immuno-inflammatory responses, which have
appeared to be represented by the hypothalamic-pituitary-adrenal
(HPA) axis and by the pineal gland for the endocrine system and by
174 Paolo Lissoni et al.
1.3 Clinical The inflammatory response is the most ancient manner of living
Investigation organisms to biologically respond. All inflammatory response-
of the Inflammatory related subjective symptoms are the end result of several and com-
Response plex cytokine interactions. Fever itself has been proven to be due to
the action of inflammatory cytokines, including IL-1-beta, IL-6,
TNF-alpha, IL-2, and IL-12, at hypothalamic sites, while other
cytokines also provided by some inflammatory effects, such as
IL-18, do not induce fever. The inflammatory response may be
clinically identified and evaluated by determining several biomar-
kers; most of them are related to the macrophage response, includ-
ing ESR, CRP, neopterin, and soluble IL-2 receptor (sIL-2R),
which express the same biological significance, since their physio-
pathological and prognostic significance is similar, because of the
evidence that abnormally high values of one or more inflammatory
markers have been proven to predict a poor prognosis and a lack of
response to the various anticancer therapies in patients with
advanced neoplasms [22]. From a physiopathological point of
view, the inflammatory response may be generated from four essen-
tial origins, consisting of (1) the macrophage system through the
production of several inflammatory cytokines, namely IL-1-beta,
IL-6, and TNF-alpha [22, 23]; (2) the TH-1 lymphocytes through
the secretion of IL-2, which induces a general activation and prolif-
eration of lymphocytes [9]; (3) the TH-17 lymphocytes by secret-
ing IL-17, which plays a fundamental role in the induction of the
acute inflammatory response by either stimulating IL-2 production
from TH1-lymphocytes, or inhibiting the immunosuppressive
activity of T reg lymphocytes [24]; and (4) the endothelial cells,
in particular by the secretion of IL-18 [25], which also exerts
inflammatory effects by stimulating interferon (IFN)-gamma secre-
tion. Macrophage production of IL-6 is under a stimulatory role
played by the same IL-1-beta released from macrophages them-
selves. Moreover, IL-6 is the major stimulator of the hepatic pro-
duction of acute-phase inflammatory proteins. In any case, the
inflammatory response depends not only on the degree of the
same inflammatory response, but also on the antagonistic action
played by the anti-inflammatory cytokines. Then, according to their
major effect of the inflammatory response, the overall cytokines
may be subdivided into three major classes, consisting of (1) -
pro-inflammatory cytokines IL-1-beta, IL-6, IL-4, IL-5, IL-8,
IL-13, IL-17, and IL-18; (2) anti-inflammatory cytokines: IL-10
and TGF-beta; (3) modulatory cytokines with both pro- and anti-
inflammatory effects: IL-2 and IL-12. IL-18, which is mainly pro-
duced by endothelial cells, macrophages, and dendritic cells, may
176 Paolo Lissoni et al.
Fig. 1 The cytokine network and its relation to the main immune cells
not only of the main cytokines, but also of the main endogenous
inhibitors of the action of cytokines, which are represented by
soluble TNF-alpha receptor 1 and 2 (sTNF-R1, sTNF-R2) for
TNF-alpha, soluble IL-1-beta receptor 1 and 2 (sIL-1-R1, sIL-1-
R2) for IL-1-beta, and IL-18-binding protein (IL-18BP) for
IL-18 [25].
2.1 Clinical The clinical evaluation of the activity of brain opioid system may be
Investigation of Opioid performed by detecting both blood and liquoral concentrations of
System Functionless the main endogenous opioids, including beta-endorphin, enkepha-
lins, and dynorphins, which act as mu-, delta-, and kappa-opioid
agonists, respectively, or in a more clinically simple manner by
analyzing the endocrine response of cortisol and LH to the oral
administration of a mu-opioid antagonist, such as naltrexone
(NTX) [31]. In normal conditions, NTX at a dose of 50 mg induces
a peak in LH and cortisol secretions, whereas the lack of LH and
cortisol response to NTX would indicate the existence of an
enhanced brain opioid system activity. On the contrary, the evi-
dence of an endocrine response to low-dose NTX would be the
expression of a diminished brain opioid system activity. The most
evident immunomodulatory effects of mu-opioid agonists, such as
beta-endorphin and morphin, consist of the induction of an
immunosuppressive status due to the inhibition of TH1 and den-
dritic cell functions, with a consequent decline in IL-2 and IL-12
production, respectively, in association with a stimulatory effect on
T reg lymphocytes, TH2 cells, and macrophages, with a following
increase in IL-10 and TGF-beta secretion [12, 34]. Then, because
of the immunosuppressive activity of the mu-opioid system, its
block by the mu-opioid antagonist NTX may improve the immune
functionless [32]. In contrast, the immunomodulating effects of
delta- and kappa-opioids are still controversial, being dependent on
dosage and experimental conditions, since both immune stimula-
tory and inhibitory effects have been described [12].
2.3 Clinical The pineal gland may be clinically investigated in its endocrine
Investigation function by evaluating the light/dark circadian rhythm of its most
of the Pineal Gland investigated hormone, MLT, even though it is known that the
pineal may produce several other hormones; the most important
of them are represented by the 5-methoxytryptamine (5-MTT) and
the beta-carboline 6-methoxy-1,2,3,4-tetra-hydro-beta-carboline,
also called pinoline or pinealine [33], which are both provided by
anticancer and antidepressant activity. MLT circadian secretion may
be explored by measuring MLT blood levels at least at the four
main phases of the light/dark period, including morning, noon,
afternoon, and night, or in a more clinically simple way by evaluat-
ing the light and dark urinary excretion of the main MLT metabo-
lite, the 6-sulfatoxy-melatonin (6-MTS) [7]. The MLT rhythm may
be considered as normal when night blood levels of MLT or night
urinary excretion of 6-MTS are at least two times greater with
respect to those found during the light phase of the day.
2.4 Clinical Both cytokine and endocrine secretions are under a control realized
Investigation of Neuro- by several feedback mechanisms. The endocrine secretions are
Endocrino-Immune mainly regulated by negative feedback mechanisms, whereas the
Interactions secretion of cytokines is preferentially under both negative and
positive feedback circuits. Moreover, there are feedback circuits
operating between endocrine and immune systems. The existence
of cytokine-endocrine feedback mechanisms would suggest that
some endocrine secretions are not only under a control played by
other hormones, but also under a regulation exerted by cytokines,
and in the same way some cytokine secretions are controlled by
hormones and neuroactive molecules. Then, according to Descar-
tes’s method suggesting the need to reduce the complexity of a
phenomenon to only some essential evidences for its interpretation,
within the great number of simultaneous interactions between
endocrine and immunocytokine secretions, it is necessary to iden-
tify the fundamental mechanisms responsible for the regulation of
the link between endocrine and immune system, and at present it is
possible to identity four main cytokine-endocrine circuits [34]:
(1) inflammatory cytokine-HPA axis: most inflammatory cytokines,
including IL-1-beta, IL-6, TNF-alpha, IFN, as well as IL-2 and
IL-12, stimulated cortisol secretion by acting on both
hypothalamic-pituitary sites and directly at adrenal level; (2) inflam-
matory cytokine-pineal gland: most inflammatory cytokines,
How to Monitor the Neuroimmune Biological Response in Patients Affected by. . . 181
3.1 Cytokine Despite the controversial results, the major evidences have demon-
Network of Aging strated that aging is characterized by a progressive decline in IL-2
and an increase in IL-17 concentrations [38]. These two events
would be connected between them, since IL-2 decline induces an
increase in IL-17 levels because of the inhibitory effect of IL-2 on
IL-17 secretion. Age-related decline in IL-2 production would be
at least in part the consequence of the concomitant age-related
decrease in the pineal function, because of the stimulatory effect
of MLT on IL-2 secretion by TH1 cells [6, 7, 35]. The decrease
with age in IL-2 levels and the increase in those of IL-17 are already
sufficient to explain aging-related enhanced frequency of both
neoplastic and autoimmune diseases. On the contrary, as far as T
reg cell behavior with age is concerned, both high and low T reg
cell percentages have been reported in aged subjects [38, 39]. Since
low or high T reg cell percentage may predispose to the autoim-
mune diseases or to the neoplastic diseases, respectively, the differ-
ent behavior of T reg cell population in aged subjects may influence
the type of disease risk for each single subject.
How to Monitor the Neuroimmune Biological Response in Patients Affected by. . . 183
3.2 Cytokine The stress is one of the main causes, which predispose to the main
Network human diseases, including cancer and autoimmunity. The neuroim-
and Endocrine mune effects induced by stress and the different effects of the
Interactions of Stress different types of stress need to be further investigated and defined.
However, according to the knowledge available up to now, stress
induces three main neuroimmune effects, as follows [12, 40, 41]:
(1) activation of HPA axis, with a consequent enhanced production
of cortisol; (2) activation of brain opioid system; and (3) activation
of the sympathetic system, mainly induced by the hypothalamic
release of CRH. Both cortisol and catecholamines may induce
lymphocytolysis, inhibition of IL-2 and IL-12 production, and
stimulation of IL-10 secretion. Catecholamines inhibit lymphocyte
functions by acting on the beta-adrenergic receptor expressed by
the immune cells. Then, catecholamine-induced immunosuppres-
sion is a beta-adrenergic-mediated phenomenon, which may be
potentially abrogated by the administration of beta-adrenergic
antagonists. Mu-opioids would act by stimulating T reg cell and
by inhibiting TH-1-dependent functions, with a following increase
in IL-10 and decline in IL-2 and IL-12 levels. The stress has been
proven to predispose to both cancer and autoimmune diseases.
Then, the main question is to establish through which mechanisms
the same condition of stress may predispose to two different and
opposite immunobiological responses, as well as those involved in
cancer and in autoimmune diseases. To better understand stress-
related neuroimmune events, firstly it is necessary to evaluate the
functional status of the two major brain structures involved in the
psychoneuromodulation of the immune responses, consisting of
brain opioid and cannabinergic systems. Stress-induced enhanced
risk for cancer development has appeared to be abrogated by the
administration of mu-opioid antagonists, such as naltrexone, by
suggesting an opioid mediation of stress-related promoting effect
on tumor development, as well as the existence of an increased
brain opioid tone during cancer progression [40]. In contrast,
stress conditions predisposing to the onset of autoimmune diseases
would be characterized by a diminished brain opioid tone, with a
consequent reduced opioid-mediated immunosuppression to
counteract the possibility of an exaggerated immune response dur-
ing the activation of the immune system, such as that occurring
during the infective processes [35]. On the other hand, brain
cannabinergic system activity, which mediates both the perception
of pleasure and the spiritual sensitivity, would be reduced in both
cancer and autoimmune diseases, with a consequent potential
enhanced endogenous production of IL-17, its secretion being
physiologically under an inhibitory control played by the endoge-
nous cannabinoids [13]. IL-17-enhanced endogenous production
may predispose to both autoimmunity and cancer, because of its
fundamental role in inducing autoimmunity-related inflammatory
response and its promoting effect on cancer cell proliferation [24].
184 Paolo Lissoni et al.
3.4 Cytokine It is known that autoimmune diseases are due to a severe dysregula-
Network tion of the cytokine network, consisting of an exaggerated activa-
of the Autoimmunity tion of the inflammatory response. However, because of the great
number of in vivo cytokine interactions, it is difficult to establish the
role played by the single cytokine in relation to the different auto-
immune diseases [42–52]. IL-17 would play a major role in the
maintenance of autoimmunity-related inflammatory response, and
the severity of disease would be enhanced by a concomitant
increased secretion of IL-18. In any case, most systemic autoim-
mune diseases, including rheumatoid arthritis (RA), systemic lupus
erythematosus (SLE), inflammatory bowel disease, multiple sclero-
sis, psoriasis, and Sjogren’s syndrome, are characterized by a non-
specific increase in blood levels of the main inflammatory cytokines,
namely IL-1-beta, IL-6, TNF-alpha, IL-8, IL-13, IFN-gamma,
IL-17, and IL-18. The pathogenetic role of TNF-alpha would be
particularly relevant in the case of RA. On the contrary, the role of
IL-2 in the autoimmunity is still controversial, since it stimulates
both TH and T reg lymphocytes, and in any case it has been shown
that autoimmunity-related inflammatory status may be maintained
also in the absence of IL-2 [24]. IL-21 and IL-23 would be also
involved in the pathogenesis of the autoimmunity, because of the
How to Monitor the Neuroimmune Biological Response in Patients Affected by. . . 185
3.5 Cytokine The allergic diseases, namely those due to an enhanced IgE pro-
Network in Allergic duction, are generally characterized by low IFN-gamma and IL-12
Diseases blood levels in association with high concentrations of IL-18,
which has been proven to stimulate IgE production, even though
it may also play an inhibitory effect on IgE secretion in association
with IL-12 [25].
186 Paolo Lissoni et al.
5 Conclusions
positively correlated with IL-2 and IL-12 blood levels [22, 23], and
in the same way the production of inflammatory cytokines generally
correlates with macrophage activation, which is reflected by the
simple monocyte count. Then then the interaction between
immunosuppressive and immunostimulatory cytokines may be
sinthetized by the simple LMR value. The only cytokines, whose
blood detection may be useful to identify the main origin of the
inflammatory response occurring in the single patient, would be
represented by IL-10, because of its potential suppressive and
stimulatory immune effects, and IL-17, being an alternative source
of the inflammatory response with respect to the most ancient
inflammatory response mainly mediated by macrophages through
the release of IL-6 and IL-1-beta. In conclusion, today without a
clinical investigation and monitoring of cytokine network during
the clinical course of disease it is not possible to understand the real
pathogenesis of human systemic diseases and their prognosis. The
only problem is to identify which may be the fundamental bio-
marker of the biological response in relation to the different pathol-
ogies, in an attempt to avoid an excessive social medical cost, due to
the detection of several cytokine levels and lymphocyte subsets,
with the risk of having to analyze many and often controversial
immune parameters. Therefore, if it is true that the investigation of
cytokine network would require the detection of several cytokines,
on the basis of the fact that cytokine-related immune alterations of
the systemic diseases are essentially due to an imbalance between
inflammatory and anti-inflammatory cytokines, it is also true that
from a clinical point of view it could be sufficient to detect LMR
values, which may constitute an inexpensive routinary biomarker to
monitor the clinical course of human systemic diseases, since low
values of LMR would reflect a decreased T lymphocyte function in
association with an enhanced macrophage-T reg cell system activity,
as occurring in the neoplastic diseases, whereas the evidence of a
normal or enhanced LMR would be the expression of a concomi-
tant enhanced activation of both lymphocyte and macrophage
systems, as occurring in the autoimmune diseases [34]. Not only
this, but the monitoring of LMR values during the clinical course of
the neoplastic and autoimmune diseases, if its significance is
explained to each patient, could also contribute to improve the
medical and human relationship with patients, who become more
active and conscious in living their disease.
References
36. Maccarone M, Valensise H, Bari M et al (2001) systemic autoimmune disease. J Exp Med
Progesterone up-regulates anandamide hydro- 203:2785–2791
lase in human lymphocytes: role of cytokines 48. Paramalingam SS, Thumboo J, Vasoo S et al
and implication in fertility. J Immunol (2007) In vivo pro- and anti-inflammatory
166:7183–7189 cytokines in normal and patients with rheuma-
37. Candelari PV, Rampoldi A, Harbuzariv A et al toid arthritis. Ann Acad Med Singapore
(2017) Leptin signaling and cancer chemore- 36:96–99
sistance: perspectives. Word J Clin Oncol 49. Gold R, Luhder F (2008) Interleukin-17.
8:106–119 Extended features of a key player in multiple
38. Tesar BM, Shirali AC, Walker WE et al (2009) sclerosis. Am J Pathol 172:8–10
Aging augments IL-17 T -cell alloimmune 50. Katsifis GE, Rekka S, Moutsopoulos NM et al
responses. Am J Transplant 9:54–63 (2009) Systemic and local interleukin-17 and
39. Dejaco C, Duftner C, Schirmer M (2006) Are linked cytokines associated with Sjogren’s syn-
regulatory T cells linked with aging? Exp Ger- drome immunopathogenesis. Am J Pathol
ontol 41:339–345 175:1167–1177
40. Elenkov IJ, Papanicolaou DA, Wilder RL et al 51. Murugaiyan G, Saha B (2009) Protumor vs
(1996) Modulatory effects of glucocorticoids antitumor functions of IL-17. J Immunol
and cathecolamines on human interleukin-12 183:4169–4175
and interleukin-10 production: clinical implica- 52. Harrison OJ, Srinivasan N, Pott J, Schiering C
tions. Proc Assoc Am Physicians 108:374–381 et al (2015) Epithelial-derived IL-18 regulates
41. Zagon IS, Donahue RN, Bonneau RH et al Th17 cell differentiation and Foxp3+ T reg cell
(2011) T lymphocyte proliferation is sup- function in the intestine. Mucosal Immunol
pressed by the opioid growth factor met (5)- 8:1226–1236
enkephalin - opioid growth factor receptor 53. Lee J, Shin EK, Lee SY et al (2014) Oestrogen
axis: implication for the treatment of autoim- up-regulates interleukin-21 production by
mune diseases. Immunobiology 216:579–590 CD4+ T lymphocytes in patients with systemic
42. Aringer A, Smolen JS (2003) Complex cyto- lupus erythematosus. Immunology
kine effects in a complex autoimmune disease: 142:573–580
tumor necrosis factor in systemic lupus erythe- 54. Eid RE, Rao DA, Zhou J et al (1982)
matosus. Arthritis Res Ther 5:172–177 Interleukin-17 and interferon-gamma are pro-
43. Fujino S, Andoh A, Bamba S et al (2003) duced concomitantly by human coronary
Increased expression of interleukin-17 in artery-infiltrating T cells and act synergistically
inflammatory bowel disease. Gut 52:65–70 on vascular smooth muscle cells. Circulation
44. Arican O, Aral M, Sasmaz S et al (2005) Serum 119:1424–1432
levels of TNF-alpha, IFN-gamma, IL-6, IL-8, 55. Zheng C, Zhou XW, Wang JZ (2016) The dual
IL-12, IL-17, and IL-18 in patients with active role of cytokines in Alzheimer’s disease: update
psoriasis and correlation with disease severity. on interleukins, TNF-alpha, TGF-beta and
Mediators Inflamm 2005(5):273–279 IFN-gamma. Transl Neurodegener 5:7–18
45. Langrish CL, Chen Y, Blumenschein W et al 56. Ehrke MJ, Mihich E, Berd D et al (1982)
(2005) Divergent pro- and antiinflammatory Effects of anticancer drugs on the immune sys-
roles for IL-23 and IL-12 in joint autoimmune tem in humans. Semin Oncol 16:230–239
inflammation. J Exp Med 201:233–240 57. Lissoni P, Brivio F, Fumagalli L et al (2009)
46. Dinarello CA (2006) Interleukin 1 and inter- Effects of the conventional antitumor thera-
leukin 18 as mediators of inflammation and the pies: surgery, chemotherapy, radiotherapy and
aging process. Am J Clin Nutr 83:447S–455S immunotherapy on regulatory T lymphocytes
47. Lohr J, Knoechel B, Wang J et al (2006) Role in cancer patients. Anticancer Res
of IL-17 and regulatory T lymphocytes in a 29:1847–1852
Part II
Abstract
For many years, the complexity and multifactorial nature of brain-immune interactions limited our ability to
dissect their underlying mechanisms. An especially challenging question was how the brain controls
immunity, since the repertoire of techniques to control the brain’s activity was extremely limited. New
tools, such as optogenetics and chemogenetics (e.g., DREADDs), developed over the last decade, opened
new frontiers in neuroscience with major implications for neuroimmunology. These tools enable mapping
the causal effects of activating/attenuating defined neurons in the brain, on the immune system. Here, we
present a detailed experimental protocol for the analysis of brain-immune interactions, based on chemo-
genetic or optogenetic manipulation of defined neuronal populations in the brain, and the subsequent
analysis of immune cells. Such detailed and systematic dissection of brain-immune interactions has the
potential to revolutionize our understanding of how mental and neurological states affect health and
disease.
Key words Immune system, Brain, Neuroimmunology, Neuroscience, Central nervous system,
Immunity, DREADDs, Chemogenetics, Optogenetics
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_11, © Springer Science+Business Media, LLC, part of Springer Nature 2018
195
196 Ben Korin and Asya Rolls
1.1 Harnessing Both optogenetics and chemogenetics enable the control of neuro-
Advances nal activity in vitro and in vivo [6–12]. These two key technologies
in Neuroscience share many similarities, while the differences between them expand
to Study the range of their potential applications.
Neuroimmunology In optogenetics, light-activated membrane ion channels,
known as opsins (e.g., channelrhodopsin and halorhodopsin), are
expressed in neurons using site-specific viral vector delivery. The
expressed ion channel is activated by light of a defined wavelength,
using optic fibers inserted in the area of viral expression in vivo
[4]. The stimulation of each ion channel provides a specific out-
come in the target neuron, depending on the channel type. For
example, channelrhodopsin proteins induce neuronal excitation
due to the inflow of cations following exposure to blue light
(~480 nm), and halorhodopsin channels induce neuronal activity
attenuation by the inflow of chloride ions following exposure to
yellow-green light (~570 nm) [13]. The frequency of light trans-
mission can determine the frequency of excitation of the neuron,
thus providing highly accurate temporal control of neuronal activ-
ity [14]. Expression and manipulation of more than one type of
light-sensitive ion channel in each neuron are possible, by selecting
different excitation wavelengths, allowing an even more accurate
control over the neuronal state [11, 15].
In chemogenetics, a similar viral vector delivery method is
used, but the vector encodes cell receptors activated by small mole-
cules (and not by light) [16]. For example, designer receptors
exclusively activated by designer drugs (DREADDs) are genetically
modified G protein-coupled receptors (GPCRs), activated by a
specific ligand (e.g., clozapine N oxide; CNO) [12]. Once trig-
gered by their ligand, these modified GPCRs induce a chain of
intracellular processes, consequently altering neuronal activity. For
example, the Gq DREADD (hM3Dq) initiates a signaling pathway
that induces neuronal activation, and the Gi DREADD (hM4Di)
initiates a signaling pathway that results in neuronal silencing
[17]. Once expressed in neurons in vivo, DREADDs can be acti-
vated using systemic injection of the ligand, or by oral administra-
tion. Depending on the ligand’s half-life and available
concentration in vivo, its administration results in prolonged mod-
ification of neuronal activity (residual activity may last for hours
[18, 19]).
1.2 Experimental There are several levels at which neuronal pathways in the brain may
Considerations be defined:
in Optogenetics l Spatial (e.g., ventral tegmental area; VTA).
and Chemogenetics
l Neuronal cell type (e.g., dopaminergic neurons in the VTA).
l Neuronal cell projection (e.g., dopaminergic neurons that proj-
ect from the VTA to the nucleus accumbens).
Application of Chemogenetics and Optogenetics to Dissect Brain-Immune. . . 197
1.3 Immune System After selecting the neuronal manipulation technique, the
Characterization subsequent experimental design, and more specifically, the timing
Following Neuronal of immune analysis, should be considered. The effects of optoge-
Manipulation netics and chemogenetics on neuronal activation are immediate.
However, subsequent effects on immune cells in the periphery may
not be instantaneous, and will most likely occur on different time-
scales, depending on the evaluated component. When analyzing
mRNA levels, one may need to consider the time required for
mRNA synthesis and/or degradation. In the case of protein expres-
sion (e.g., intracellular and extracellular receptors) a longer lag
period (several hours) should be expected. Moreover, under an
experimental design dependent on accumulative neural transmis-
sion to the periphery, requiring prolonged and/or repeated neuro-
nal manipulation, several hours to several days from the beginning
of the experiment may be needed. In the case of chemogenetics, the
actual initiation time and duration of the chemogenetic receptor
activation (depending on ligand availability) may also affect the
timescale of the response.
One may also consider introducing an immunological stimulus;
this is expected to provide a more orchestrated and distinctive
response by the immune system, depending on the stimulus. For
example, when we performed reward system activation in naı̈ve
mice, we mainly saw small effects that we interpreted as priming
[1]. However, once we introduced a bacterial challenge, the effects
of reward system activation on the antibacterial response became
functionally evident [1]. Taken together, the features of the applied
neuronal manipulation, the timing of experimental analysis, and the
immunological context are especially important. These crucial vari-
ables should be chosen based on the proposed experimental
hypothesis.
Another crucial factor to consider in the experimental design is
the appropriate control group(s). The control group should
account for the effects of surgery, viral expression, and
Application of Chemogenetics and Optogenetics to Dissect Brain-Immune. . . 199
2 Materials
2.1 Animals 1. Cre-dependent transgenic mice: Age: 8–10 weeks. For exam-
ple, for targeting dopaminergic neurons: TH-Cre mice (e.g.,
B6.Cg-Tg(Th-Cre)1Tmd/J).
2. All experiments should be performed in accordance with the
institution’s guidelines for the care and use of laboratory
animals.
3 Methods
3.1 Site-Specific 1. Prepare the surgical environment in advance; make sure that all
Delivery of Viral Vector materials and instruments are readily available and properly
for Chemogenetics sterilized.
202 Ben Korin and Asya Rolls
22. After the needle is in the proper location, wait for 3 min for
tissue adjustment. Monitor for bleeding or any signs of stress.
23. During this time, keep skin tissue moist using a cotton swab
with sterile saline solution (NaCl 0.9%) or sterile PBS / .
24. Inject the viral vector at a rate of 0.1 μL/min (e.g., 0.7 μL of
viral vector over 7 min).
25. Following injection, keep the needle in place for an additional
5 min, to enable optimal cell transfection, and avoid backflow
of the viral suspension.
26. Slowly and carefully pull back the needle. Monitor for bleeding
or any abnormalities.
27. Disinfect the surgical area in the skull using a cotton swab with
povidone-iodine solution.
28. Using tissue glue (can also be done with stitches or staples),
reattach the two skin flaps while fully covering the skull. Make
sure to fully reattach the tissue.
29. To support postsurgical recovery, inject (subcutaneous, s.c.)
1 mL of pre-warmed (37 C) 0.9% NaCl.
30. Following the surgical procedure, let mice rest in a controlled
warm environment, and monitor until they are awaken.
31. Carefully monitor the condition of the mice (i.e., mobility, fur,
behavioral changes, weight) daily during the following week.
32. Provide analgesics (e.g., buprenorphine 0.05–0.1 mg/kg) as
required. Exclude animals with unresolved complications.
33. Allow recovery and effective viral expression for 21–30 days.
3.2 Protocol 1. Following viral injection, wait for 10 min for virus to fully
Adjustments diffuse.
for Optogenetics In transgenic mice that heritably express the opsin of choice,
(See Fig. 1) continue as follows (i.e., without the viral injection).
2. Using a stereotactic holder, slowly insert the optical fiber
implant based on the desired coordinates, i.e., in the same
location as drilled previously.
3. Secure doric/cannula implant location with Metabond dental
cement.
4. Release the optical fiber implant from the stereotaxic holder.
5. To the extent possible, reattach the two flaps of the skin using
tissue glue.
6. To support postsurgical recovery, inject (subcutaneous, s.c.)
1 mL of pre-warmed (37 C) 0.9% NaCl.
7. Following the surgical procedure, let mice rest in a controlled
warm environment, and monitor until they are awaken.
204
Ben Korin and Asya Rolls
Chemogenetics Optogenetics
CNO Blue Yellow
Promoter DREADD Reporter
Virus expression Recovery (30 days)
Promoter ChR2 Reporter
Ready for
Promoter Reporter (control)
neuronal manipulation
hM3Dq or hM4Di Na+ or Cl-
(excitation) (attenuation) (excitation) (attenuation)
Fig. 1 Viral delivery with spatial resolution. Schematic representation of the stereotaxic injection of a viral vector that encodes the chemogenetic/optogenetic
information, and a fluorescent reporter (or control–fluorescent reporter only). Following injection, mice are allowed to recover for at least 30 days
Application of Chemogenetics and Optogenetics to Dissect Brain-Immune. . . 205
3.4 In Vivo Activation 1. Following surgical recovery, habituate mice to the use of the
of Light-Sensitive optical fiber cord.
Channels (See Fig. 2) 2. On the day of the experiment, attach the optical cord extension
to the mouse implant, and secure tightly.
3. Connect the optical cord to the appropriate light source.
4. Provide the desired light pulse (manually or automatically
applied) to stimulate opsin-expressing cells.
5. Note pulse intensity, frequency, and duration of light pulses,
along with the overall number of trials.
206 Ben Korin and Asya Rolls
CNO
Chemogenetic Tissue extraction
stimulation and processing Flow cytometry
Immune
system analysis Mass cytometry
Light Proteomics
Optogenetic
stimulation RNA sequencing
4 Notes
Acknowledgments
References
1. Ben-Shaanan TL, Azulay-Debby H, Dubovik 20:700–707. https://doi.org/10.1038/nn.
T et al (2016) Activation of the reward system 4526
boosts innate and adaptive immunity. Nat Med 3. Pavlov VA, Tracey KJ (2017) Neural regulation
22:940–944. https://doi.org/10.1038/nm. of immunity: molecular mechanisms and clini-
4133 cal translation. Nat Neurosci 20:156–166.
2. Abe C, Inoue T, Inglis MA et al (2017) C1 https://doi.org/10.1038/nn.4477
neurons mediate a stress-induced anti-inflam- 4. Deisseroth K (2011) Optogenetics. Nat Meth-
matory reflex in mice. Nat Neurosci ods 8:26–29
208 Ben Korin and Asya Rolls
Abstract
Natural killer (NK) cells are an essential component of innate immunity. These lymphocytes are also
sensitive barometers of the effects of endogenous and exogenous stressors on the immune system. This
chapter describes a chromium (51Cr)-release bioassay designed to measure to the target cell killing capacity
of NK cells (NKCC). Key features of the cytotoxicity assay are that it is done with whole blood and that
numbers of effector cells are determined for each sample by flow cytometry and lymphocyte count. Effector
cells are defined as CD3-CD56+ lymphocytes. Target cells are the K562 erythroleukemia cell line. Killing
capacity is defined as number of target cells killed per effector cell, at an effector cell/target cell ratio of 1:1
during a 4-h in vitro assay.
Key words Natural killer cells, NK cells, Natural killer cell cytotoxicity, NKCC, Lymphocytes, K562
target cells, Chromium-release assay, Innate immunity, Flow cytometry
1 Introduction
The natural killer (NK) cell is a large, granular lymphocyte with the
ability to lyse tumor cells and virus-infected cells without prior
exposure and immunization [1]. These cells can prolong asymp-
tomatic states in HIV-infected persons with low CD4 counts [2]
and protect against malignancy [3]. The NK cell is a reliable marker
of neuroendocrine–immune interactions. Stressful life events that
trigger the fight or flight response, such as a natural disaster, can
alter lymphocyte trafficking and function, leading to elevated
( p < 0.001) natural killer cell cytotoxicity (NKCC) and number
of circulating NK cells ( p < 0.000) as were seen in post-Hurricane
Andrew samples [4]. According to a meta-analysis by Segerstrom
and Miller [5], the mobilization of NK cells during acute psycho-
logic stressors is one of the most replicated and robust findings in
human psychoneuroimmunology. For example, 45 first-time tan-
dem parachutists were examined for NK activity 2 h before,
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_12, © Springer Science+Business Media, LLC, part of Springer Nature 2018
209
210 Mary Ann Fletcher et al.
Table 1
Friedman test: effect of aerobic exercise on NK cell counts in Gulf War illness cases (T0 ¼ baseline;
T1 ¼ VO2; T2 ¼ 4 h)
Table 2
Friedman test: effect of aerobic exercise on NK cell counts in healthy controls (T0 ¼ baseline;
T1 ¼ VO2; T2 ¼ 4 h)
20
N = 40; p = .001
16
12
% NKCC
0
Baseline V02 max
Fig. 1 Effect of exercise challenge to V02 max on NKCC (% of target cells killed at a target cell ratio of 1:1) in
40 healthy controls
Fig. 2 Histogram of NKCC (% of target cells killed at a target cell ratio of 1:1) for 230 healthy controls
2 Materials
Fig. 3 Histogram of NKCC (% of target cells killed at a target cell ratio of 1:1) for 176 chronic fatigue syndrome
cases
3.1 Tissue Culture 1. Stock media (SM): In a flask mix 500 mL of RPMI, 5 mL of
phosphate-buffered saline (PBS), 5 mL MEM, 5 mL sodium
pyruvate. Filter through 0.2 μm pore size filter.
2. Assay media (AM): Mix and filter (0.2 μm pore size filter) 90 mL
of SM and 10 mL of FBS.
3. Culture media (CM): Mix and filter (0.2 μm pore size filter)
85 mL of SM and 15 mL of FBS.
4. Storage requirements: Chromium in solution, stock media,
assay media, nonessential amino acids, FBS, MEM, and sodium
pyruvate should all be kept in the refrigerator at 2–8 C for up to
1 week. L-Glutamine and Pen-Strep should be kept in the
freezer at 20 C.
5. Target cells: The usual target cell is the K562 erythroleukemic
cell line available from the ATCC http://www.atcc.org/. The
NK-sensitive K562 cell line is maintained in a humidified 5%
CO2 atmosphere at 37 C in stationary suspension in CM. Cells
are subcultured twice weekly to produce log-phase growth (see
Note 1; 4 106 cells in 20 mL of CM will yield approximately
16 106 cells).
4 Assay Procedure
4.1 Target Cell 1. Pour the entire volume of cultured cells in flask into a 50 mL
Preparation conical tube. Remove 100 μL of the cell suspension and com-
(in Biohazard Hood bine them with 100 μL of trypan blue for the initial cell count.
with Sterile Technique) Spin 50 mL conical tube at low speed (400 g) for 5 min.
2. Pour off the supernatant into waste, and add 10 mL of room-
temperature SM to resuspend the cells. Centrifuge for 5 min at
400 g.
3. Remove the supernatant, and add the chromium solution equal
to 100 μCi, for each 20 106 cell mix. The amount of 51Cr
depends on weeks of “age” (time elapsed since the date of
calibration of the chromium solution, see Note 2). Incubate
for 1 h at 37 C in a humidified 5% CO2 atmosphere, with gentle
shaking every 15 min.
4. In the same 50 mL tube, add 10 mL of 37 C AM to the
chromium/cell solution, mix gently, and spin at 400 g for
5 min. Pour off the supernatant into the radioactive waste
container. Repeat this washing four times.
5. After the fourth wash add 10 mL of 37 C AM. Remove 50 μL
of cell/assay media solution to a test tube; add 50 μL trypan blue
for the final cell count in order to determine the volume of assay
media needed to produce a cell count of 2 106 K562 cells.
With remaining solution in the 50 mL tube, centrifuge for 5 min
at 400 g.
4.2 NKCC Reaction 1. Pour off supernatant from E above, add the volume of room-
Setup temperature AM proscribed by the final cell count, and mix
gently. In separate tubes make graded dilutions to produce the
four target cell concentrations as shown in Table 3. An assay
with four blood samples requires a minimum of 2 mL of each
target cell dilution.
2. Add 150 μL well-mixed blood from EDTA tube to each well of
one row in culture plate.
3. For spontaneous-release control (SR), dispense 150 μL AM to
one row of 12 wells.
4. For total-release control (TR), dispense 150 μL 1% Triton
X-100 to one row of 12 wells.
5. Add 50 μL cell suspensions to each of the three wells for each of
the four dilutions (change tips if progressing from stronger to
weaker concentration; otherwise tips can be conserved for one
row).
6. Cover plate and centrifuge for 10 min at room temperature at
400 g.
216 Mary Ann Fletcher et al.
Table 3
Dilution of target cells
Final concentration Amount of initial cell suspension (2 106 cells/mL) Amount of media
2 10 cells/mL
6
2 mL 0 mL
1 10 cells/mL
6
1 mL 1 mL
0.5 106 cells/mL 0.5 mL 1.5 mL
0.25 10 cells/mL
6
0.25 mL 1.75 mL
4.4 Lymphocyte This information is required for calculation of NKCC using Eq. (1).
Count This should be done using an electronic hematology analyzer,
which will provide the hematocrit that is also needed for Eq. (1).
4.5 Flow Cytometry Lymphocytes that are CD45+, CD14, CD3, and CD56+ con-
stitute the bulk of the cells capable of NK cell cytotoxic activity.
Using four-color flow cytometry, determine the percent of cells in
the lymphocyte gate that meet this requirement. Multiply this by
the lymphocyte count to obtain the number of NK cells for Eq. (1).
4.6 Calculations Subtract the background counts (b) from all experimental release
(ER), SR, and TR and calculate the percent cytotoxicity for each
dilution of target cells according to Eq. (1):
2 n o 3
ðER b Þ V t ðVVb HCT Þ
ðSR b Þ
% cytotoxicity ¼ 4 5 100
t
ðTR b Þ ðSR b Þ
ð1Þ
4.7 Quality Control There is no commercial proficiency test available for this procedure.
for Assay For quality control, always perform the assay in triplicate and have
blinded duplicates submitted to the laboratory. The SR should be
<20% of TR and the correlation coefficient for the Michaelis–Men-
ten equation should be >90%. Each laboratory must determine an
expected range for healthy individuals. In our laboratory, the
230 healthy controls shown in Fig. 1 had a mean % cytotoxicity at
an effector-to-target cell ratio of 1:1 of 28%. Table 4 gives the
parameters of NKCC for controls and a patient group, CFS.
4.8 Performance Outlying individual scores when found in triplicate data sets (i.e.,
Parameters from three wells from the same tube) should be excluded, and the
scores which are closest to one another in the set should be aver-
aged for results.
4.9 Limitations This procedure is done with fresh blood samples. Blood that is <8 h
of the Procedure old is preferred. Samples that are <8 but >24 h can be run.
However, their results must be compared to the control range
determined for samples of that age.
218 Mary Ann Fletcher et al.
Table 4
The parameters of NKCC for controls and a patient group CFS
Natural killer cell cytotoxicity for chronic fatigue syndrome and healthy
controls Statistic Std. error
CFS Mean 15.94688 0.858119
95% Confidence interval for mean Lower bound 14.25328
Upper bound 17.64047
Median 12.05000
Variance 129.601
Std. deviation 11.384233
Minimum 1.200
Maximum 55.300
Range 54.100
Interquartile range 13.675
Skewness 1.392 0.183
Kurtosis 1.718 0.364
HC Mean 27.83922 0.808923
95% Confidence interval for mean Lower bound 26.24533
Upper bound 29.43310
Median 28.00000
Variance 150.502
Std. deviation 12.267920
Minimum 2.000
Maximum 69.000
Range 67.000
Interquartile range 17.025
Skewness 0.211 0.160
Kurtosis 0.033 0.320
5 Notes
Acknowledgments
References
1. Herberman RB, Holden HT (1979) Natural physicians in Japan: health assessment by
killer cells as antitumor effector cells. J Natl immune variables (CD4, CD8, CD56, and
Cancer Inst 62:441–445 NK cell activity) at the beginning of work. J
2. Solomon GF, Benton D, Harker JO, Occup Health 50:136–146
Bonavida B, Fletcher MA (1994) Prolonged 11. Fletcher MA, Zeng XR, Maher K, Levis S,
asymptomatic states in HIV-seropositive per- Hurwitz B, Antoni M, Broderick G, Klimas
sons with fewer than 50 CD4+ T cells per NG (2010) Biomarkers in chronic fatigue syn-
MM3. Psychoneuroimmunologic findings. drome: evaluation of natural killer cell function
Ann N Y Acad Sci 741:85–90 and dipeptidyl peptidase IV/CD26. PLoS One
3. Su DM, Vankayalapati R (2010) A new avenue 5(5):e10817
to cure cancer by turning adaptive immune T 12. Maher K, Klimas NG, Fletcher MA (2005)
cells to innate immune NK cells via reprogram- Chronic fatigue syndrome is associated with
ming. J Mol Cell Biol 2:237–239 diminished intracellular perforin. Clin Exp
4. Ironson G, Wyings C, Schneiderman N, Immunol 142:505–511
Baum A, Rodriquez M, Greenwood D, 13. Alter G, Malenfant JM, Altfeld M (2004)
Benight C, Antoni M, LaPerriere A, Huang CD107a as a functional marker for the identi-
H-S, Klimas NG, Fletcher MA (1997) Post- fication of natural killer cell activity. J Immunol
traumatic stress symptoms, intrusive thoughts, Methods 294:15–22
loss and immune function after Hurricane 14. Brunner K, Mauel TJ, Cerottini J-C, Chapuis B
Andrew. Psychosom Med 59:128–141 (1968) Queantitative assay of the lytic action of
5. Schedlowski M, Jacobs R, Stratmann G, immune lymphoid cells on Cr labeled alloge-
Richter S, H€adicke A, Tewes U, Wagner TO, neic target cells in vitro Inhibition by isoanti-
Schmidt RE (1993) Changes of natural killer body and by drugs. Immunology 14:181
cells during acute psychological stress. J Clin 15. Baron GA, Klimas NG, Fischl MA, Fletcher
Immunol 13:119–126 MA (1985) Decreased natural killer cell-
6. Segerstrom SC, Miller GE (2004) Psychologi- mediated cytotoxicity per effector cell. Diagn
cal stress and the human immune system: a Immunol 3:197–204
meta-analytic study of 30 years of inquiry. 16. Fletcher MA, Baron GA, Ashman MA, Klimas
Psychol Bull 130:601–630 NG (1987) Use of whole blood methods in
7. Klimas NG, Rosenthal M, Fletcher MA (2006) assessment of immune parameters in immuno-
Immune effects of an acute exercise challenge deficiency syndromes. Clin Diag Immunol
in Gulf War illness. Assoc Military Surgeons 5:69–81
US, San Antonio, TX 17. Bourin P, Mansour I, Doinel C, Roué R,
8. Harvey JM, Barnes Z, Zeng X, Sol C, Fletcher Rouger P, Levi F (1993) Circadian rhythms of
MA, Klimas NG (2011) Exercise effects on circulating NK cells in healthy and human
biomarkers of CFS/ME and GWI. Presented immunodeficiency virus-infected men. Chron-
to the International Conference on CFS/ME, obiol Int 10:298–305
Ottawa, Canada 18. Cleland WW (1967) The statistical analysis of
9. Cohen S, Tyrrell DA, Smith AP (1991) Psy- enzyme kinetic data. Adv Enzymol 29:1
chological stress and susceptibility to the com- 19. Callewaert DM, Mahle NH (1985) Kinetic
mon cold. N Engl J Med 325:606–612 models of natural killer cells and their use for
10. Okamoto H, Tsunoda T, Teruya K, Takeda N, studying activated NK cells. In: Herberman
Uemura T, Matsui T, Fukazawa S, Ichikawa K, RB, Callewaert DM (eds) Mechanisms of cyto-
Takemae R, Tsuchida K, Takashima Y (2008) toxicity by NK cells. Academic Press, Orlando,
An occupational health study of emergency FL, p 381
Chapter 13
Abstract
The field of psychoneuroimmunology (PNI) aims to uncover the processes and consequences of nervous,
immune, and endocrine system relationships. Behavior is a consequence of such interactions and manifests
from a complex interweave of factors including immune-to-neural and neural-to-immune communication.
Often the signaling molecules involved during a particular episode of neuroimmune activation are not
known but behavioral response provides evidence that bioactives such as neurotransmitters and cytokines
are perturbed. Immunobehavioral phenotyping is a first-line approach when examining the neuroimmune
system and its reaction to immune stimulation or suppression. Behavioral response is significantly more
sensitive than direct measurement of a single specific bioactive and can quickly and efficiently rule in or out
relevance of a particular immune challenge or therapeutic to neuroimmunity. Classically, immunobehavioral
research was focused on sickness symptoms related to bacterial infection but neuroimmune activation is
now a recognized complication of diseases and disorders ranging from cancer to diabesity to Alzheimer’s.
Immunobehaviors include lethargy, loss of appetite, and disinterest in social activity/surrounding environ-
ment. In addition, neuroimmune activation can diminish physical activity, precipitate feelings of depression
and anxiety, and impair cognitive and executive function. Provided is a detailed overview of behavioral tests
frequently used to examine neuroimmune activation in mice with a special emphasis on pre-experimental
conditions that can confound or prevent successful immunobehavioral experimentation.
Key words Mouse, Maze, Exploration, Brain based, Biobehaviors, Memory, Motor activity,
Anhedonia
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_13, © Springer Science+Business Media, LLC, part of Springer Nature 2018
221
222 Albert E. Towers et al.
between the nervous system and the immune system [2]. This is to
say that both neural-to-immune and immune-to-neural signaling
occur. Shared pathways exist between the nervous and immune
systems that use a repertoire of signaling molecules such as cyto-
kines and neurotransmitters [3] that are capable of interacting with
both immune and nervous system cells. These bioactives can con-
vey the state of peripheral immunity to the neuroimmune system,
communicate the status of neuroimmunity to the peripheral
immune system [3–5], and provide immunoactivating and deacti-
vating signals to immune cells throughout the body [6].
Observation of innate immune-mediated behavioral change
(immunobehaviors) is largely used as a method of measuring neu-
roimmune activation in response to pathogenic insults of infectious
[6] or non-infectious [7] etiologies. Sickness behavior, in a classical
sense, is a set of coordinated behavioral changes aimed at conserv-
ing and redirecting body energy stores toward combating illness
and promoting recovery [4, 8]. Immunobehaviors are best known
for their manifestation in association with bacterial infection [8],
but materialize in a spectrum of conditions and diseases including
cancer [8], autoimmune disorders [9], wounding [10], depression
[11], obesity [12], and neurodegenerative diseases [13]. In any
circumstance in which the innate immune system is activated,
peripheral inflammatory mediators can impact the brain, altering
normal function and causing symptoms of illness/loss of well-being
[4, 8]. Typical sickness behavior symptoms include reduction in
food intake, lethargy, malaise, loss of interest in social and/or
environmental surroundings, changes in sleep patterns, and
impaired cognition [4, 8, 11]. Furthermore, continued or dysre-
gulated activation of the neuroimmune system can progress beyond
acute sickness symptoms and transition to behaviors observed in
the anxious or depressed [11]. Fatigue is often a lingering compli-
cation of neuroimmune activation [6] and can present as purely
mental or physical but, most commonly, in a combinational form
[14]. In rodents, exercise behaviors like spontaneous wheel run-
ning (SWR) [15] are helping to unravel the complex biology of
physical fatigue. Tests examining memory formation (learning) and
memory recall, as well as those capable of measuring decision-
making, are being used to explore mental fatigue [16] and
impairment [17].
Finally, immunobehaviors are a powerful indicator of neuroim-
mune status offering insight into the pro-inflammatory milieu of
the brain. Altered behavior manifests prior to detectable changes in
brain-based bioactives and lingers past their resolution. Such con-
ditions indicate that the brain is very sensitive to small perturba-
tions and that traditional chemical bioassays are often not sensitive
enough or appropriately targeted to detect brain-based dysfunc-
tion. Hence, use of behavioral testing provides highly sensitive and
phenomenologically relevant information in regard to brain
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 223
2 Pre-experimental Considerations
2.1 Model and Strain A first consideration should be the animal model and strain chosen.
Choice The vast majority of PNI behavioral testing utilizes rats and mice.
Porcine models are becoming more prevalent [18–20], as have
other types of rodents, especially prairie voles [21]. In addition
zebra fish are emerging as a high-throughput model for many
behaviors, especially anxiety-like behaviors [22]. Mice are particu-
larly useful in neuroimmune and immunobehavioral research due
to their ability to reproduce and mature rapidly, and the relative
ease to which genetic modification can be applied through muta-
tional, transgenic, and knockout approaches [23]. Different behav-
ioral phenotypes exist between strains. Therefore, it is important to
be aware of and control for potential inter-strain and inter-sub-
224 Albert E. Towers et al.
2.3 Age Natural aging modulates immune function with subject variation
dependent on factors such as lifelong physiological stressor expo-
sure [30]. It comes with little surprise that immunobehaviors are
also affected by age. While sickness symptoms appears beneficial to
young mice, they may be maladaptive in older mice [31]. Age can
also negatively impact physical activity since aged mice run less
[32]. Young mice can be difficult to analyze in running tests, as
well, due to a progressive increase in distance traveled [33]. There-
fore, investigators need to consider confounds related to non-age-
matched experimental animals.
2.4 Transportation Environmental factors play a significant role in how mice behave
and respond to neuroimmune stimuli and immunobehavioral treat-
ments. Mice should be allowed a transition phase to acclimatize to a
new environment, whether a procedure/experimentation room
itself [34, 35], or the housing room following arrival from an
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 225
2.5 Light Cycle Alteration of light cycle impacts natural murine behavioral patterns
[36], and immunobehaviors (e.g., anxiety) [37]. Mice are under
control of a genetically driven circadian clock that serves to regulate
physiological and behavioral processes in a diurnal fashion
[38]. Therefore, it is important to ensure that experimental
rooms and housing rooms have similar light cycles. In addition, it
is advisable to initiate behavioral experiments at the same time of
“day,” especially when performing repeat testing. Mice are noctur-
nal; thus most testing should occur during the dark cycle. Reverse
light cycle housing is advantageous so as not to put undue burden
on personnel performing the behavioral tests. Mice are crepuscular
[39], with heightened activity during the early (dusk) and late
(dawn), and should be tested during these peak activity times.
Methods for determining the timing of these active periods are
described in Subheading 3.2.
2.7 Noise Noise has the potential to activate neuroimmune signaling path-
ways and alter behavior, as bell ringing [43] and noise produced by
vacuuming [44], as example, stress laboratory mice by activating
the hypothalamic-pituitary-adrenal (HPA) axis [45, 46]. White
noise generators that create a constant background are useful in
reducing behavioral responses to sudden loud noises [47].
2.8 Odors Mice use odors for communication, marking territory, and in rec-
ognition signaling [36]. In addition to using patterns of urine
deposition for communication, mice produce specialized odors
via several glandular secretions [36]. Evidence indicates that neu-
roimmune activation alters odor production, which, in turn, pre-
cipitates behavioral change. As example, mice administered
lipopolysaccharide (LPS) generate olfactory cues to indicate that
they may have a transmissible pathogen. This causes healthy cage
mates to socially withdraw from “sick” mice [48]. A similar phe-
nomenon is seen in healthy mice housed with tumor-bearing cage
mates [49]. Finally, exposure to foreign/strange odors (e.g.,
human-associated odors) can result in stress responses [36]. Unfor-
tunately, no specific research has been performed examining the
duration of olfactory stress responses in mice, nor is there an
identified acclimation or exhaustion time for evocative scents. It
should also go without mention that eliminating as many olfactory
cues in/on rooms, cages, and experimental equipment (e.g.,
through use of 70% v/v ethanol) is best practice. Mouse handler
odors (i.e., perfumes and predator scents (e.g., cat odor)) should be
minimized and/or eliminated.
2.10 Housing Mice are social animals and should be housed, if possible, with
Method/Environmental other mice [40]. Several studies have investigated the impact that
Enrichment individual housing/social isolation has on neuroimmunity and
immunobehaviors [53]. In general, social isolation induces aggres-
sion in male mice [54]. Individually housed male mice also appear
prone to developing anxiety and depressive-like behaviors follow-
ing unpredictable chronic mild stress [55] despite their increased
propensity to explore [26, 35]. Group-housed mice order them-
selves into a social hierarchy, with 1–2 dominant mice and several
subordinate mice. Subordinates as well as dominant mice show
similar exploratory behavior in an open-field context [26] but this
seems dependent on the relatedness of the group-housed mice. As
evidence, an introduced non-sibling-dominant intruder mouse
evokes social stress and immune cell glucocorticoid resistance in
the group-housed mice [56]. Therefore, housing mice in groups at
a density of one 25–30 g mouse per 77.4–96.7 cm2 of cage floor
area by 12.7 cm of cage ceiling height is advantageous [40].
Unintuitively, clean cage environment [36, 57] and novel cage
construction materials can reduce mouse welfare and cause aberrant
behaviors [36, 40]. Introduction to a clean/novel cage increases
plasma corticosterone in mice, and increases physical activity in the
first 24 h of exposure. Metal cages, as opposed to plastic cages [40],
are colder to the touch, more conducive to noise generation, and
less permeable to light [36, 40]. Additionally, the use of solid
flooring with absorbent bedding is recommended by most institu-
tional care and use committees because wire mesh flooring can lead
to paw injury [40]. Such injury can confound behavioral experi-
ments by triggering the innate immune and pain systems [6].
Environmental enrichment is thought to enhance mouse well-
being by providing motor and sensory stimulation. Environmental
enrichment may include nesting material, structures, and/or shel-
ters within the cage [40]. Lack of environmental enrichment dam-
pens mouse reactivity and alertness in many behavioral tests
[58]. Environmental enrichment is, however, somewhat strain
dependent because the loss of reactivity and alertness mentioned
earlier observed in BALB/c mice but not in C57BL/6 mice. In
fact, Van de Weerd et al. concluded that male BALB/c mice housed
in enriched environments are anxious [58]. Olsson and Dahlborn
noted that changing a barren cage environment by placing objects
within it does not necessarily lead to “enrichment” [59]. Instead, it
is important to measure behavioral and physiological changes that
occur in enriched-housed animals, and if said enrichments result in
long-term improved health and/or well-being. Some “environ-
mental enrichers” are felt to result in stress and/or anxiety
[58, 59]. Thus, the environment within the cage may be as impor-
tant as the environment in which the cage is housed when establish-
ing appropriate pre-experimental procedures.
228 Albert E. Towers et al.
3 Sickness Behaviors
3.1 Social The paradigm of social exploration appears to have evolved from
Exploration of a Novel work done by Thor and Holloway showing that adult laboratory
Juvenile rats actively investigate and form social memories of conspecifics
[65], principally via anogenital sniffing, nosing, mutual grooming,
and close following [66]. Rodents do not show behavioral habitu-
ation during social investigation, provided that the conspecific
juvenile is novel at each presentation into the home cage of the
adult [67]. Furthermore, adult male rats do not exhibit aggression
toward prepubertal male juveniles but do toward unrelated post-
pubertal males. This is a function of androgen-related odors from
the postpubertal rat, which can elicit aggressive attacks from the
adult due to unfamiliarity [65]. These aspects of social recognition
first allowed Dantzer et al. to show that social memory could be
modulated by neurological peptides [61], and were likely the first
experiments that used social exploration as a tool to measure the
effects of neuroactive compounds. Social exploratory behavior was
adapted from social memory testing by using a different juvenile at
each observation time point [68]. Due to a lack of habituation
when using a novel juvenile, social exploration is routinely used as
a sensitive test of immunobehavioral perturbation. Bluthé et al.
adapted the rat-based Dantzer procedure to mice [66].
3.1.1 Procedure All observations should be made during the dark/active cycle of the
mice. Mice should be housed individually at least 24 h prior to the
initial measurement and be allowed to acclimatize to the procedure
room (if procedure room differs from housing room) for the time
period in which they are individually housed. In some studies,
infrared light is used with the aid of infrared or night-vision capable
cameras [66, 68], but use of red-tinted lighting is also acceptable, as
mice only have limited ability to detect light from the red portion of
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 229
3.2.1 Procedure All observations should be made during the dark/active cycle of the
mice. Mice should be housed individually at least 24 h prior to the
initial measurement and be allowed to acclimatize to the procedure
room (if procedure room differs from the housing room) for the
time period in which they are individually housed. As with social
exploration, infrared or red lights are used to provide illumination.
At each time point of interest, mouse movement is recorded for
5 min [12], with a camera placed over the center of a single cage or
a grouping of cages. If multiple mice are being recorded during a
given observation point, they should be shielded from one
another’s view. Movement including total distance traveled, veloc-
ity, and time spent moving is best determined from the video record
using automated tracking software. However, if tracking software is
not available, videos can be hand scored by a blinded trained
observer. To score manually, a thin-line grid comprised of six
equally sized rectangles is affixed to the television or monitor screen
directly over the cage and the blinded trained observer counts the
number of times the mouse crosses a line (line crossing) through-
out the 5-min designated investigational period. A mouse is only
considered to have line crossed if both fore and hind limbs cross a
line [69]. A more powerful method for assessing mouse locomotor
activity is through long-duration (hours–days) tracking [73]. This
is required when detailing mouse crepuscular movement. While
video recording can be used for such evaluation, the data collection,
storage, and interpretation can be burdensome due to video file
sizes. Therefore, biotelemetry is a useful method for this type of
testing [74]. In this method, a biotelemetric emitter is surgically
implanted within the peritoneum of a mouse and a receiver pad
linked to a PC running data collection software is placed directly
underneath the home cage of the implanted mouse (Mini Mitter,
Bend, OR). Mouse movement is tracked and recorded automati-
cally. Specific procedures and training for this method should be
provided by the manufacturer of the device chosen.
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 231
3.3.1 Food Mice should be individually housed as per social exploration at least
Disappearance Procedure 24 h prior to experimentation. With single housing, food should be
moved from the overhead cage food hopper and placed in an 8 cm
diameter 5 cm stainless steel bowl or glass dish in the cage. Steel
and glass are preferred over ceramic because ceramic containers can
absorb water if they are not completely glazed. In addition, steel
dishes can be magnetically secured to the cage bottom or side with
the use of a strong magnet. This prevents mice from tipping the
bowl and spilling food which can easily occur with plastic bowls.
After the acclimation period, and just prior to initiation of testing,
new food should be added to the bowl and the bowl weighed. This
process should occur at the very beginning of the dark/active cycle
of the mice in order not to disturb mice during their sleep cycle.
Food disappearance is measured by weighing the bowl plus food at
fixed intervals, such as every 24 h. For longer term experiments,
food can be re-added to the bowl and reweighed [62]. The term
food disappearance is used in place of food consumption because
not all food is ingested. Some food inadvertently falls in the cage
bedding [75]. Mice should be individually housed as per social
exploration at least 24 h prior to experimentation. 24 h prior to
testing mice should be fasted but allowed full access to water. The
empty 8 cm diameter 5 cm food bowl should be present in the
cage to reduce novelty effects. One hour prior to testing, mice
should be removed to similarly sized cage without bedding but
with full water access. Testing is initiated by cleaning the bottom of
the bedding-less cage and placing a pre-weighed bowl with food in
the cage. After 1 h, the food bowl is removed and weighed as are
any food remnants within the cage. The difference between food
bowl food disappearance and food collected from the cage floor is
considered food consumed [72]. A more powerful method for
assessing food disappearance and/or consumption is through use
of automated food and water intake measurement systems where
food and/or water intake initiated by the animal is evaluated by
computer-controlled electronics (BioDAQ, New Brunswick, NJ).
Specific procedures and training for this method should be
provided by the manufacturer of the device chosen.
3.4 Rotarod Testing Inducers of neuroimmune activation and sickness behavior impair
motor coordination and induce physical fatigue [63]. The rotarod
performance test can measure motor coordination [64] by
232 Albert E. Towers et al.
assessing how well mice avoid falling off a rotating rod [76]. Some
strains of mice progressively perform better on the rotarod test
during repeated trials at the same rotational speed indicating a
physical training or memory component to this procedure
[77, 78]. Rotarod apparatuses are available via commercial vendors
such as Med Associates (Fairfax, WT), with some variance in fea-
tures (number of lanes and/or rod diameter, for example). In
general, rotarod apparatuses have the same basic design featuring
a 3–9 cm diameter rod [64, 77] partitioned by plastic divider disks
spaced evenly longitudinally along the rod. The end point
measured is latency to fall from the rod [79]. Fall detection ranges
from pressure-sensitive pads located under the rod to infrared
beams that automatically stop an integrated timer when hit or
blocked. Rotarod performance can, however, be assessed manually
from a video record [79] by a blinded trained observer or by
automated tracking software.
3.4.1 Procedure Single housing prior to testing is not required but, like with social
exploration, mice need at least 24 h of acclimation to the procedure
room. Use of pre-experimentation acclimation to the rotarod is not
agreed upon. Some have exposed (trained) mice to the rotating
rod, by placing mice on the rod at a low speed (4 rpm [63] and
18 rpm [64]), while others have not [77]. Pre-experimentation
exposure to the rod has been done 1 week in advance of testing
[64] and immediately prior to testing [63]. Finally, the test itself
can be performed in one of the two ways. The rotarod performance
test measures the duration of time a mouse can remain on the
rotating rod at a single or several fixed speeds [79]. The accelerating
rotarod performance test measures the maximum speed of rotation
the mouse can tolerate before it falls from the rod in a fixed amount
of time [77, 79]. All pretest conditioning and testing should be
made during the dark/active cycle of the mice [63], although
testing has been performed during the light/inactive cycle [64].
3.4.2 Rotarod Rotarods should be calibrated such that they rotate at a constant
Performance Test speed. Rods should be kept clean and as odor free as possible
between trials, as urine and feces on the rod can affect performance
[79]. Testing is initiated by placing mice on the rotating rotarod
which rotates at a fixed speed. Mice are allowed to maintain them-
selves on the rod as long as they can and the test session continues
until they fall or a designated time point is achieved. At such time,
the latency to fall is recorded. Fixed-speed trials are best used after
significant validation testing on the strain of mouse chosen. Fixed-
speed trials are best used on mice with significant loss of coordina-
tion because small losses of coordination may not manifest at the
speed or time chosen. Some testing protocols investigate several
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 233
4 Depressive/Anxiety-Like Behaviors
4.1 Burrowing Rodents are well-known burrowers [81]. This behavior is related to
tunnel maintenance and possibly defense. Defensive burying is a
known indicator of anxiety and can, itself, be measured [81]. Bur-
rowing appears to be largely hippocampal driven but mouse strain
differences exist with C57BL/6 mice burrowing more than
129S2/Sv mice [81]. Burrowing is associated with depressive/
anxiety-like behavior where reduced burrowing reflects a
depressive/anxiety-like state [12]. As burrowing utilizes relatively
simple equipment and minimal labor, it is an easy and inexpensive
method for evaluating immunobehaviors [12, 81].
4.1.1 Procedure All observations should be made during the dark/active cycle of the
mice. Mice should be individually housed as per social exploration
at least 24 h prior to experimentation and, with single housing, a
clean empty burrowing tube should be placed in the cage. Burrow-
ing tubes can be constructed from standard white 6.8 cm diameter
PVC piping cut to 20 cm in length [81]. The open end of the
burrowing tube is elevated 3 cm by bolting two 50 mm machine
screws 1 cm from the open end, and spaced so that the tube
entrance is elevated. This elevation keeps burrowing substrate
from spilling out of the open end. The closed end is sealed with a
standard PVC end cap [12]. Testing should begin 3 h prior to the
onset of the dark/active phase of the light cycle and is initiated with
addition of burrowing substrate to the tube [12]. The burrowing
substrate used needs to be suitable to the mouse strain and experi-
mental treatment. Pelleted mouse chow, gravel, or sand is the
common material used for burrowing [81]. The burrowing tube
can be completely filled [81] or filled with a fixed amount of
substrate [12] (if ceiling effects are not a concern). Ceiling effects
arise with vigorous burrowers, because these mice will remove all
substrate from a tube in a rapid time frame obscuring any difference
in burrowing activity relative to time. After substrate is placed in the
burrow, the burrow plus substrate is weighed and returned to the
cage. If mouse chow is used as a substrate, food from the cage food
hopper should be removed for the duration of the burrowing test
[81]. Depending on anticipated mouse burrowing activity, experi-
mental observation time points can range from 1 to 24 h. Amount
burrowed (in grams) is calculated from the pre- and post-
burrowing weight of the tube plus substrate. Following a measure-
ment, the burrowing tube can be refilled, reweighed, and returned
to the cage for additional testing [81]. Alternatively, a single mea-
surement of burrowing can be utilized [12]. Occasionally, with
poorly burrowing mice, one or several training sessions may be
necessary [88]. Such practice runs improve burrowing activity and
reduce variability between animals [81], likely by mitigating the
novel object effect of the burrowing substrate.
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 235
4.2.1 Procedure All observations should be made during the dark/active cycle of the
mice. Single housing prior to testing is not required but, like with
social exploration, mice need at least 24 h of acclimation to the
procedure room. The maze can be made of a variety of materials but
those that can be easily wiped clean between each mouse tested are
recommended. Maze shape is that of a plus sign where the four
arms are spaced 90 apart, radiating from an open central
5 cm 5 cm platform. Arm length and width are 25 cm 5 cm,
respectively. Maze elevation should be at least 40 cm from the floor
[82, 89]. Arm wall height is 15 cm in the “closed” arms and there
are no side walls in the “open” arms. The central 5 cm 5 cm
platform has no walls. The open arms are at 180 from each other,
likewise with the closed arms. Unlike the tests described earlier, the
EPM should be well lit by overhead white light. Significant arm
wall-generated shadows, especially those confined to a single arm,
should be avoided. Testing is initiated by placing the mouse in the
open central 5 cm 5 cm platform. Each subject mouse is intro-
duced to the maze in a similar fashion and placed on the maze in the
same spot with analogous orientation [82]. Mouse exploration is
video recorded for 5 [82] to 10 min [90]. Time spent in open and
closed arms, number of entries between arms (defined as all four
paws of the mouse crossing the threshold of an arm), frequency of
head dips (downward movement of the mouse head toward the
floor from an open arm), rears, and stretch-attend postures [82] are
best determined from the video record using automated tracking
software [69]. If tracking software is not available videos can be
hand scored by a blinded trained observer [82].
4.3 Open-Field Test The OFT can be used to measure movement [84] and anxiety-like
behavior [83, 84]. OFT apparatuses are walled arenas that vary in
shape (square, rectangle, and circle) and size (250–2500 cm2)
[84]. OFT testing should not be used as a surrogate test for
spontaneous locomotor activity because the OFT uses a novel
environment [84]. Anxiety-like behavior in the OFT is evaluated
by examining mouse movement throughout the arena with a spe-
cial focus on the amount of time the mouse spends/moves next to
walls of the OFT apparatus (thigmotaxis). The novelty, size, and
236 Albert E. Towers et al.
4.3.1 Procedure All observations should be made during the dark/active cycle of the
mice. Single housing prior to testing is not required but, like with
social exploration, mice need at least 24 h of acclimation to the
procedure room. For testing of anxiety-like behaviors the arena
should be larger than 1600 cm2 [84]. Arena wall height should
be at least 35 cm so as to limit the ability of the mouse to see over
and above the arena. The arena can be made of a variety of materials
but those that can be easily wiped clean between each mouse tested
are recommended. Testing is initiated by placing the mouse in the
center of the arena, and each subject mouse needs to be introduced
to the arena in a similar fashion and placed in the same spot with
analogous orientation. Movement through the arena is video
recorded for 5–10 min and analysis of movement is best documen-
ted with automated tracking software because thigmotaxis is easily
appreciated with this method. Path tracing is a key aid so that
patterns of movement can be evaluated. These pattern tracings
supplement the usual measurements of time spent adjacent to the
arena walls, wall preferences, time spent not adjacent to the arena
walls, overall distance traveled, velocity, and time spent moving.
Videos can be manually examined using a line-crossing scoring
approach (similar to that described for spontaneous locomotor
activity) but this method should be carefully validated due to the
complex grid pattern needed to ascertain time spent close to the
arena walls. Due to this complexity, some have used the end point
of total distance moved plus time spent in the central 25% of the
arena when utilizing manual scoring [89]. Finally, OFT can be used
as a repeated measure to determine if therapeutics improves perfor-
mance over time [91].
4.4 Elevated Like the EPM, the zero maze measures anxiety-like behaviors in
Zero Maze mice [77]. Time spent in the open indicates a reduced level of fear/
anxiety as validated by use of anxiolytic agents that increase the time
mice spent in the open area of the zero maze [77]. The advantage of
the zero maze over the EPM is the elimination of the central
platform, which can complicate analysis of open/closed arm
comparisons [92].
4.4.1 Procedure All observations should be made during the dark/active cycle of the
mice. Single housing prior to testing is not required but, like social
exploration, mice need at least 24 h of acclimation to the procedure
room. Maze design varies but in general is comprised of a circular
track 30–45 cm in diameter that is 3–5 cm wide. Maze elevation
should be at least 40 cm from the floor [77, 93]. The track should
be divided into four quadrants with two quadrants having no side
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 237
4.5 Light/Dark Box Like the EPM and EZM the light/dark box measures anxiety-like
behaviors in mice by exploiting the innate avoidance behavior of
mice to bright lights [94]. In this test, increased time spent in the
light part of the arena is a marker of reduced anxiety-like behavior.
Importantly, mice administered benzodiazepines demonstrate
increased time in the lit section, indicating that the light/dark
box paradigm is responsive to anti-anxietal therapies.
4.5.1 Procedure All observations should be made during the dark/active cycle of the
mice. Single housing prior to testing is not required but, like social
exploration, mice need at least 24 h of acclimation to the procedure
room. Test boxes can range from 44 21 cm to 91 92 cm, and
vary in the proportion of light to dark areas. Generally, the light/
dark portions are 1:1 or 2:1 [95]. Increasing the light:dark ratio
exacerbates the anxiogenic properties of the arena by creating an
open-field-like space. Light and dark areas are typically separated by
either a simple wall with a small opening (usually no more than
7 cm) or small tunnel between the two sides (5 7 10 cm).
Arenas are typically constructed of translucent and opaque plastics
to create the light and dark areas. Fully automated computer-
connected arenas are helpful (Omnitech Electronics, Columbus,
Ohio), and can easily track mouse movement automatically within
the testing period. Animals should be introduced to the testing
arena in a standardized place and orientation, usually into the light
part of the arena [96]. Testing should be video recorded for
5–10 min. Time spent in the light, time spent in the dark, latency
238 Albert E. Towers et al.
4.6 Tail The TST is a commonly used behavioral test for assessing
Suspension Test depressive-like behavior in mice. It is thought to induce an escape
response [97]. With increased depressive-like behavior mice fail to
extricate themselves from the apparatus and become immobile.
Increased immobility indicates a greater degree of depressive symp-
toms [98]. Importantly, antidepressants shorten immobility offer-
ing a degree of validation to the test’s usefulness in measuring
depressive-like behaviors [97]. The TST can be automated through
use of commercially available apparatuses that utilize computer-
linked linear load cells and load cell filters to determine mouse
movement/struggle (Med Associates, St. Albans, VT). As with
any commercially purchased device, specific procedures and train-
ing should be provided by the manufacturer. However, certain
basic procedures should be followed and considered with use of
the TST including the difficulty in examining young mice (espe-
cially C57BL/6) due to their robust tail climbing behavior and
penchant for escaping.
4.6.1 Procedure Unlike all the previous behavioral tests described the TST should be
administered during the light/inactive cycle of the mice. Single
housing prior to testing is not required but, like social exploration,
mice need at least 24 h of acclimation to the procedure room.
Testing is initiated by affixing the mouse to the apparatus “hook”
with adhesive tape wrapped around the tail at three-quarters of its
length from the tail base. Attach the mouse to the apparatus
through the tape as close to the tail as possible. The tail should
remain straight so as not to injure the mouse. Mice should be
suspended as uniformly as possible, and if multi-mouse devices
are used the mice should be shielded from each other’s view
[98]. Immobility versus movement/struggle should be measured
for 6 min. Nonautomated devices can be constructed, which are
essentially chambers with hooks. Mouse behavior can be video
recorded and immobility determined from the video record by
automated tracking software [99] or a blinded trained observer
[100]. With any of the aforementioned analyses, time of immobility
is compared between control and experimental groups of
mice [98].
4.7 Forced The FST (also called the Porsolt test for the investigator who
Swim Test developed the test in rodents) like the TST is a tool for assessing
depressive-like behavior in mice. The FST is relatively easy to
administer [101] and felt to be the best validated test for depression
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 239
4.7.1 Procedure Like the TST, the FST should be administered during the light/
inactive cycle of the mice [102]. Single housing prior to testing is
not required but, like social exploration, mice need at least 24 h of
acclimation to the procedure room. As with most device requiring
tests, equipment design varies. A simple setup is to use clean white
or black cylindrical PVC containers 16 cm in diameter and 31 cm in
height (essentially 2 gallon pails) containing 20 cm of water main-
tained at 25 1 C [103]. The FST should be performed under
30 lux white light [104]. Testing is initiated by introducing the
subject mouse to the water-filled container. Mouse swimming is
video recorded for 6 min [103]. Immobility is determined from the
video record from the last 5 min of the FST using either automated
tracking software [105] or a blinded trained observer [103]. Immo-
bility scoring should not include movements necessary for the
mouse to maintain its head above water [98]. Like the TST, time
of immobility is compared between control and experimental
groups of mice.
4.8 Sucrose/ Anhedonia, or the inability to gain pleasure from otherwise enjoy-
Saccharin Preference able experiences, is one of the features of depression [106, 107]. In
mice, their preference for sweetened solutions has been exploited to
measure anhedonia. The decreased consumption of a sweet-tasting
solution is indicative of anhedonia and can be reversed with anti-
depressives [107]. Sucrose [107] and saccharin [12] solutions are
commonly used opposite normal tap water in a two-choice test. For
investigators concerned with mouse caloric intake, saccharin is the
recommended sweetener [12]. Advantages to the sucrose/saccha-
rin preference test are that it can be run continuously for many days
without significant concern of adaptation or learning.
4.8.1 Procedure Three days prior to testing mice should be singly housed in stan-
dard cages adapted for two-bottle water access (adaptation phase).
If the experimental design requires mice to be challenged with a
neuroimmune activator, each bottle should contain either saccharin
as a 0.4% sodium saccharin solution (1% for sucrose can be sub-
stituted for saccharin) or water. If the experimental design does not
require exogenous challenge as with a comparison of mice of
240 Albert E. Towers et al.
5 Cognitive Behaviors
5.1 Novel Object Novel object recognition is a test of working memory in mice. The
Recognition test exploits the innate tendency of mice to investigate a new entity
[110]. Novel object testing is one of the simplest of cognition tests,
but test variations are described that add significant complexity
through mixing of objects, object placement (novel location test-
ing), and testing arena conditions [12, 110–112]. The setup for
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 241
5.1.1 Memory Recall All observations should be made during the dark/active cycle of the
Procedure mice. Mice should be housed individually at least 24 h prior to the
initial measurement. The 24-h training phase is initiated by intro-
ducing two identical objects into the home cage (standard shoebox
cage size; 28 cm in length; 17 cm in width; 12.5 cm in height) of
the singly housed mouse. The objects are placed 10 cm apart at the
short-side wall end, 5 cm from the short-side wall, and 3.5 cm from
the long-side wall. Tall (3–5 cm in height) complex objects are
preferred because when a tall complex object is introduced during
the testing phase it provokes significantly more exploration time.
Tall complex objects can be constructed from Lego® blocks
(Enfield, CT). Magnets can be used to secure the structures to
the cage floor. All structures should be taken apart and cleaned
prior to reuse. After the 24-h training phase, the mouse is exposed
to the chosen neuroimmune activator. At relevant times after the
applied immunobehavioral challenge, the memory recall testing
phase is initiated by placing the mouse in a home-cage-like arena
(including bedding) which contains a similar object setup as in the
training phase, where one of the familiar objects has been replaced
by a novel object. The mouse should be introduced at the cage end
opposite the objects. Object exploratory behavior is video recorded
for 5 min and object investigation is determined from the video
record by either automated tracking software or a blinded trained
observer. Object exploration is considered as contact by mouth,
nose, or paw. Accidental contact such as bumping into an object
while passing should not be considered [111]. Mice with a memory
recall deficit should examine both the familiar and novel objects
equally [114]. Once recovered from neuroimmune activation, mice
should explore the novel object more frequently than the familiar
object. This test cannot be performed as a repeated measure and,
thus, requires separate groups of mice to determine at what time
after neuroimmune activation cognitive recovery occurs.
5.1.2 Memory Formation All observations should be made during the dark/active cycle of the
Procedure mice. Single housing prior to testing is not required, but, like with
social exploration, mice need at least 24 h of acclimation to the
procedure room. Memory formation testing differs from recall
testing in that training occurs after endogenous activation of the
242 Albert E. Towers et al.
5.2 Novel Location The novel location recognition test is the spatial memory analog to
Recognition the novel object recognition test. This test is performed in an
identical manner as the novel object recognition test with two
minor alterations. First, spatial cues are placed around the outside
of the testing arena; these cues can be simple tape markings on at
least three sides of the cage [115, 116]. The second alteration is
that rather than replacing one of the objects with a new, novel
object, one of the objects is simply moved to the opposite end of
the cage than where it was during training. Performed in this
manner the novel location recognition test can also be used for
repeated measure testing by moving the object to a new area within
the testing arena for each subsequent test.
5.2.1 Memory Recall The memory recall procedure for the novel location recognition
Procedure task is identical to the procedure described in Subheading 5.1.1,
combined with the following changes. The 24-h training phase is
initiated by introducing two identical objects into the home cage
(standard shoebox cage size; 28 17 12.5 cm), with spatial cues
on three of the four sides. Objects are placed 10 cm apart at the
short-side wall end, 5 cm from the short-side wall, and 3.5 cm from
the long-side wall. The memory recall testing phase is initiated by
placing the mouse in a home-cage-like arena which contains a
similar object setup as in the training phase where one of the objects
has been moved to the opposite end of the cage, 5 cm from the
short-side wall. The mouse should be introduced at the cage
between the two objects.
5.2.2 Memory Formation The memory formation procedure for the novel location recogni-
Procedure tion task is identical to the procedure described in Section 5.1.2,
combined with the following changes. The two familiar objects are
placed on one short end of a testing arena (28 17 12.5 cm)
with spatial cues on three of four sides for 30 min and then returned
to the home cage. After 2 h in the home cage, testing is initiated by
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 243
placing the mouse back in the testing arena where one of the
familiar objects has been moved to the opposite short end of
the cage.
5.3.1 Cued Fear- All observations should be made during the dark/active cycle of the
Conditioning Procedure mice. Mice should be single housed for this behavioral procedure
and, like social exploration, mice need at least 24 h of acclimation to
the procedure room. The training/testing should take place in a
separate area of the procedure room, as animals in the same space
may be able to hear auditory cues, unless using soundproof fear-
conditioning chambers. Automated commercially available fear-
conditioning apparatuses (San Diego Instruments, San Diego,
CA) are the easiest way to adapt this testing paradigm. General
apparatus parameters are fairly uniform. There is a shock generator
and scrambler that delivers a 0.1–1.0 mA foot shock through a wire
grid floor in concert with a sound generator that produces auditory
cues, all contained in a shoebox cage-sized chamber [117]. It is
recommended that sound meters and voltmeters are used to verify
and record stimulus intensities [117]. Prior to testing mice require
training. In the initial training session, mice are placed in the fear-
conditioning apparatus for 120 s (Phase A) before the presentation
of a 30-s sound cue (Phase B). A 2-s foot shock is delivered
immediately after the sound cue (Phase C). Mice are returned to
their home cages 30 s after the shock ends. Repeat training can be
utilized to reinforce the memory. As noted earlier, training relative
to neuroimmune stimulation determines whether memory forma-
tion or recall is being tested. Testing is usually initiated 24 h post-
training and consists of reintroducing the mouse to the fear-
conditioning apparatus and representing the sound cue. The
sound cue now lasts for 180 s. Mouse behavior during this 180-s
244 Albert E. Towers et al.
5.3.2 Contextual Fear- Contextual fear conditioning uses the same pre-experimental and
Conditioned Learning scoring procedures as cued fear conditioning. However, in contex-
Procedure tual fear conditioning, no sound cues are delivered. The mouse is
expected to associate the apparatus with the foot shock. Testing
time is 180 s. Complexity can be added by using an alterable
microenvironment within the fear-conditioning apparatus (altered
contextual fear conditioning). A variety of cues from visual to
olfactory can theoretically be utilized [117].
5.4.1 Procedure All observations should be made during the dark/active cycle of the
mice. Single housing prior to testing is not required but, like with
social exploration, mice need at least 24 h of acclimation to the
procedure room. As with all mazes described previously, the
Y-maze used for spontaneous alternation should be made of a
material that can be easily wiped clean between each mouse tested.
Clear plastics are preferred so that a different black-colored design
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 245
5.5 Barnes Maze The Barnes maze, like spontaneous alternation, is a test of spatial
memory. This test combines several aspects of the previously men-
tioned mazes including elevation, open/exposed illuminated space,
and a dark enclosed area [121]. Use of the Barnes maze was
popularized as an alternative to the Morris water maze (MWM)
(described in the next section) because swimming may produce
anxiety [122]. Removal of water favors testing of mice different
fat density due to elimination of the buoyancy effect. Importantly,
the Barnes maze appears to rely on the same hippocampal-
dependent memory function as the MWM [121]. As with any
maze designed to test spatial memory, extra- and intra-maze cues
serve as location reference points, and without these cues mice
perform less well [121]. Like spontaneous alternation, the Barnes
maze can be used to test memory formation and recall depending
on when neuroimmune activation is triggered relative to the train-
ing period. However, recall is significantly simpler to measure when
using transient memory impairment paradigms.
5.5.1 Procedure All observations should be made during the dark/active cycle of the
mice. Single housing prior to testing is not required but, like social
exploration, mice need at least 24 h of acclimation to the procedure
room. As with all mazes, construction materials should be easily
246 Albert E. Towers et al.
5.6 Morris The MWM is used for assessing spatial or place learning. Advan-
Water Maze tages of the MWM include no requirement for pre-training, high
reliability across different tank designs, and proved validity in mea-
suring hippocampal-dependent spatial and reference memory.
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 247
5.6.1 Spatial Acquisition All testing should take place during the dark/active cycle of the
Procedure mice. Single housing prior to testing is not required but, like social
exploration, mice need at least 24 h of acclimation to the procedure
room. The mouse is placed (not dropped) at the selected start
position in the maze, facing the tank wall. Video tracking should
start immediately at the placement of the mouse in the maze. The
video remains on until the mouse reaches (comes in contact with)
the platform. Standard trial limits of 1–2 min per trial are usually
used, and mice that have not reached the platform should be placed
on or guided to it by the investigator [123]. The animal should
remain on the platform for 15 s. This step helps mice orient their
position in space relative to the extra-maze cues [123]. Following
the inter-trial interval, the mouse is again placed in the maze, this
time in a different but predetermined location (most protocols start
mice from one of the four positions—south (the investigator’s
position), north (opposite the investigator’s position), east (to the
248 Albert E. Towers et al.
right of the investigator’s position)), and west (to the left of the
investigator’s position)). Trials are repeated four times per day for
5 days. Following training, the experimental treatment is adminis-
tered and time is allowed for the treatment to take effect, or in the
case of neuroimmune activators, for sickness behavior to resolve so
as not to confound the results [11]. The probe trial is run, during
which the platform is removed from the maze. The probe trial is
video recorded and lasts 60 s, after which the mouse is removed.
The objective of the probe trial is to determine whether or not the
mouse can recall memories of where the platform was during
training sessions based on the distal visual cues [123]. End points
measured in spatial acquisition include the number of platform site
crossovers, time and distance swam in the target quadrant relative
to the other quadrants, time in a predefined radius around the
original platform position (larger than the original platform itself),
average distance swam to target site, and latency to first target site
crossover. For investigators without automated tracking capabil-
ities, blinded trained observers should use a timer to calculate the
time spent in the aforementioned areas, as distance traveled is not
reliably measurable. Percent time spent in the target quadrant or
percent of distance swam in the target quadrant is the most com-
mon reported end points in MWM spatial acquisition
testing [123].
5.6.2 Spatial Reversal Spatial reversal testing determines the ability of mice to extinguish a
Testing Procedure particular memory in favor of a new one [123]. In this paradigm,
training procedures are the same as they were for spatial acquisition,
but the probe trial differs. During the reversal training probe trial,
the platform is moved, typically to the opposite side of the maze,
but cues remain in their same position as during training trials
[124]. Mice are placed first on the platform for 30 s to allow
them to gain some spatial cues as to where the new platform
location is. Mice are then given 1–3 trials to reach the platform,
starting from different locations if necessary [124]. The same end
points are used in spatial reversal training as with spatial acquisition
[123]. Since the platform remains in the maze, latency to reach the
platform, swim speed, and total distance swam are also useful end
points [124]. Some variations of the MWM include repeated
learning, latent learning, and cued learning [123].
5.7 5-Choice The 5-choice reaction time task (5CRT) is a useful test for assessing
Reaction Time Task a variety of cognitive functions, specifically attention, reaction time,
and executive function. The test was adapted from protocols for
testing attention in human subjects [125]. Typically, the test is
performed in a single chamber (25 25 cm) with one curved
wall containing five (2.5 2.5 cm) openings. In automated devices,
each opening is bisected by a photo beam capable of detecting
mouse movement [126], and contains a light cue that the animal
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 249
5.7.1 Procedure For the serial reaction time procedure the training phase is initiated
by placing mice into the testing chamber for a period of 1 min to
acclimate them to the chamber. Following acclimation, the first light
cue is activated in one of the openings selected at random. Beam
crosses into the opening are recorded following cue initiation and
the light is turned off following a beam cross into the correct
opening. Once mice select the correct opening a food reward is
given on the opposite end of the chamber. A new trial begins 2 s
after the mouse retrieves the food reward. A completion of this cue-
to-reward circuit is a single trial. Trials continue until a specific
amount of trials or specific time determined by the researcher is
met [129]. Computer software combined with human observers
can measure a variety of parameters such as percent of mice reaching
trial criterion, percent correct trials, percent premature responses,
percent incorrect responses, and percent omissions [129].
6 Physical Activities
6.1 Spontaneous A key advantage of SWR is that it can be assessed without moving
Wheel Running mice from their home cage and the length of examination time can
be very long. A disadvantage is that mice need to be singly housed.
Specialized caging is needed to accommodate the running wheel
and bedding must be correctly adjusted and monitored so as not to
interfere with the wheel. A basic running wheel may measure only
revolutions of the wheel and may need manual resetting at each
data collection point. Advanced running wheel systems (Mini Mit-
ter, Bend, OR) can obtain hourly, daily, or weekly run distance.
Regardless of the running wheel sophistication, wheels need to be
clean and well lubricated. Running wheel size and structure should
also accommodate the size of mouse used. In long-duration stud-
ies, cage cleaning and contact with animal facility and/or investiga-
tive personnel can result in an acute reduction in running. For the
below procedure, an automated, multichannel running wheel sys-
tem (Mini Mitter, Bend, OR) is utilized. Specific procedures and
training for any given wheel should be provided by the manufacture
of the device chosen.
6.1.1 Procedure Mice need to be individually housed for experiments with running
wheels, and need to acclimate to the procedure room for at least
24 h prior to experimentation. Groups of mice in cages with locked
running wheels and in cages with no wheels should be included for
proper experimental controls. Prior to experimentation (such as
neuroimmune activation), a baseline measurement is recorded in
case post hoc normalization of distance run is required. After
immunobehavioral stimulation, mice are immediately returned to
their cage and allowed access to the wheel (rotating or locked) or
cage environment (no wheel). A 10-day course of wheel running is
recommended. With automated running wheel systems, total dis-
tance run is reported. With manual wheels, wheel revolutions are
recorded and distance traveled calculated by multiplying the wheel
circumference by the number of revolutions. A limitation in SWR is
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 251
6.2 Forced Treadmill FTR better measures mouse fatigue [133]. Like running wheel
Running systems, mouse treadmills vary in sophistication with some allowing
both uphill and downhill running (IITC Inc. Life Science, Wood-
land Hills, CA). Treadmills coupled to oxygen consumption sys-
tems can be used to determine mouse “fitness.” Non-rodent
treadmills (Jog-A-Dog, Ottawa Lake, MI) divided into lanes with
plastic dividers allow for high-throughput studies of up to 20 mice.
Treadmills should contain a protective end (foam) to prevent mice
from being thrown from the device and to provide an impetus to
move forward should the mouse reduce its speed or stop running.
Mice will respond to the contact of their tail/hind portion to the
foam end. A ventilated cover is also recommended. Intra-
experimental prodding can be used if a mouse or mice appear to
predominantly “ride” the treadmill, but this encouragement can
lead to bias due to the difficulty of applying prodding evenly to
every subject mouse.
6.2.1 Procedure All observations should be made during the dark/active cycle of the
mice. Mice do not need to be single housed prior to this procedure
but should be allowed to acclimate for at least 24 h to the procedure
room. Prior to experimentation, mice should be trained daily for
3 days at speeds of 14–20 m/min (speed depends on mouse age
and strain). Mice that cannot learn the treadmill task should not be
included in experimental studies. Training sessions should last until
mouse exhaustion (1–2 h). Immunobehavioral stimulation is deliv-
ered 24 h after the final training session. Testing is initiated by
conducting a treadmill run to exhaustion. Exhaustion is considered
as a cease in running that is not motivated by protective end contact
with the foam end. Time to exhaustion is the measured end point.
Distance to exhaustion can be calculated from the time run and
velocity of the treadmill.
6.3 Forelimb Grip FGS is a reliable measure of muscle strength and motor function in
Strength mice. Measurement devices for FGS vary widely, from simple force
dynamometers to advanced computerized grip strength meters
such as the GPM-100 (Melquest, Toyama, Japan). During the
test, the animal is pulled by the tail while holding onto the dynam-
ometer with increasing force until it can no longer hold on. This
provides an accurate measurement of the amount of force a single
mouse can produce. In studies of young and old mice, older mice
produce less force during FGS testing, providing a validation of this
technique as a measure of muscle strength [134]. NF-kB blockade
has also been shown to increase FGS performance in the MDX
mouse model [135].
252 Albert E. Towers et al.
6.3.1 Procedure All observations should be made during the dark/active cycle of the
mice. Mice do not need to be individually housed prior to this
procedure, but should be allowed to acclimate to the testing room
for 24 h. The test can be performed either horizontally or vertically,
with vertical testing shown to be more consistent across multiple
tests [134]. Testing is initiated by allowing mice to grab hold of the
bars on the force gauge. Once the mouse has grabbed the rods, the
provided force gauge should be reset to 0. The investigator then
pulls the animal by the tail in a direction directly opposite the force
gauge until the mouse falls or is unable to hold itself onto the bars.
Force exerted immediately prior to the mouse releasing is recorded.
Testing should be repeated 3–6 times with approximately 1 min
between each successive test. Results are an average and reported as
force (kg) recorded on the force gauge.
7 Conclusion
Acknowledgments
References
1. Ader R (2000) On the development of muscle damage. Am J Physiol Regul Integr
psychoneuroimmunology. Eur J Pharmacol Comp Physiol 291:R1344–R1348
405(1–3):167–176 16. Rönnb€ack L, Hansson E (2004) On the
2. Maier SF, Watkins LR, Fleshner M (1994) potential role of glutamate transport in men-
Psychoneuroimmunology—the interface tal fatigue. J Neuroinflammation 1:22
between behavior, brain and immunity. Am 17. Tanila H (2017) Testing cognitive functions
Psychol 49(12):1004–1017 in rodent disease models: present pitfalls and
3. Kerschensteiner M, Meinl E, Hohlfeld R future perspectives. Behav Brain Res. https://
(2009) Neuro-immune crosstalk in CNS dis- doi.org/10.1016/j.bbr.2017.05.040
eases. Neuroscience 158:1122–1132 18. Warren EJ et al (1997) Coincidental changes
4. Kelley KW et al (2003) Cytokine-induced in behavior and plasma cortisol in unre-
sickness behavior. Brain Behav Immun 17: strained pigs after intracerebroventricular
S112–S118 injection of tumor necrosis factor-α. Endocri-
5. Irwin MR (2008) Human psychoneuroim- nology 138(6):2365–2371
munology: 20 years of discovery. Brain 19. Antonson AM, Radlowski EC, Lawson MA
Behav Immun 22:129–139 et al (2017) Maternal viral infection during
6. Dantzer R (2004) Cytokine-induced sickness pregnancy elicits anti-social behavior in neo-
behavior: a neuroimmune response to activa- natal piglet offspring independent of postna-
tion of innate immunity. Eur J Pharmacol tal microglial cell activation. Brain Behav
500:399–411 Immun 59:300–312. https://doi.org/10.
7. Johnson DR et al (2007) Acute hypoxia acti- 1016/j.bbi.2016.09.019
vates the neuroimmune system, which diabe- 20. Rytych JL, Elmore MRP, Burton MD et al
tes exacerbates. J Neurosci 27(5):1161–1166 (2012) Early life iron deficiency impairs spatial
8. Dantzer R (2001) Cytokine-induced sickness cognition in neonatal piglets. J Nutr
behavior: mechanisms and implications. Ann 142:2050–2056. https://doi.org/10.3945/
N Y Acad Sci 933:222–234 jn.112.165522
9. Gibertini M (1998) Cytokines and cognitive 21. Grippo AJ (2009) Mechanisms underlying
behavior. Neuroimmunomodulation altered mood and cardiovascular dysfunction:
5:160–165 the value of neurobiological and behavioral
research in animal models. Neurosci Biobehav
10. Wingfield JC et al (2006) Contexts and ethol- Rev 33(2):171–180
ogy of vertebrate aggression: implications for
the evolution of hormone-behavior interac- 22. Johnson A, Hamilton TJ (2017) Modafinil
tions. In: Nelson RJ (ed) Biology of aggres- decreases anxiety-like behaviour in zebrafish.
sion. Oxford University Press, New York PeerJ 5:e2994. https://doi.org/10.7717/
peerj.2994
11. Dantzer R et al (2008) From inflammation to
sickness and depression: when the immune 23. Crawley JN (2003) Behavioral phenotyping
system subjugates the brain. Nat Rev Neu- of rodents. Comp Med 53(2):140–146
rosci 9:46–56 24. Shigemura N et al (2004) Leptin modulates
12. Lavin DN et al (2011) Fasting induces an behavioral responses to sweet substances by
anti-inflammatory effect on the neuroim- influencing peripheral taste structures. Endo-
mune system which a high-fat diet prevents. crinology 145(2):839–847
Obesity 19(8):1586–1594 25. Rodgers RJ, Cole JC (1993) Influence of
13. Yi B, Sahn JJ, Ardestani PM et al (2017) Small social isolation, gender, strain and prior nov-
molecule modulator of sigma 2 receptor is elty on plus-maze behaviour in mice. Physiol
neuroprotective and reduces cognitive deficits Behav 54(4):729–736
and neuroinflammation in experimental mod- 26. Palanza P, Gioiosa L, Parmigiani S (2001)
els of Alzheimer’s disease. J Neurochem Social stress in mice: gender differences and
140:561–575. https://doi.org/10.1111/ effects of estrous cycle and social dominance.
jnc.13917 Physiol Behav 73:411–420
14. Paul RH et al (2000) Fatigue and its impact 27. Lightfoot JT et al (2004) Genetic influence on
on patients with Myasthenia Gravis. Muscle daily wheel running activity level. Physiol
Nerve 23(9):1402–1406 Genomics 19:270–276
15. Carmichael MD et al (2006) Role of brain 28. Basterfield L, Lumley LK, Mathers JC (2009)
IL-1β on fatigue after exercise-induced Wheel running in female C57BL/6J mice:
impact of oestrus and dietary fat and effects
254 Albert E. Towers et al.
on sleep and body mass. Int J Obes (Lond) behavior of male mice. Asian J Biol Sci
33:212–218 4:362–368
29. Clayton JA, Collins FS (2014) Policy: NIH to 43. Goshen I et al (2003) The role of endogenous
balance sex in cell and animal studies. Nature interleukin-1 in stress-induced adrenal activa-
509:282–283. https://doi.org/10.1038/ tion and adrenalectomy-induced adrenocorti-
509282a cotropic hormone hypersecretion.
30. Hawkley LC, Cacioppo JT (2004) Stress and Endocrinology 144:4453–4458
the aging immune system. Brain Behav 44. Naff KA et al (2007) Noise produced by
Immun 18(2):114–119 vacuuming exceeds the hearing thresholds of
31. Dilger RN, Johnson RW (2008) Aging, C57BL/6 and CD1 mice. J Am Assoc Lab
microglial cell priming, and the discordant Anim Sci 46(1):52–57
central inflammatory response to signals 45. Turnbull AV, Rivier C (1995) Regulation of
from the peripheral immune system. J Leukoc the HPA axis by cytokines. Brain Behav
Biol 84(4):932–939 Immun 9(4):253–275
32. Martin SA et al (2011) Voluntary-wheel exer- 46. Beishuizen A, Thijs LG (2003) Review: endo-
cise training attenuates the visceral adipose, toxin and the hypothalamo-pituitary-adrenal
but not central, inflammatory response to (HPA) axis. Innate Immun 9(1):3–24
LPS in aged C57BL/6J mice. Brain Behav 47. Pfaff J (1974) Noise as an environmental
Immun 25(S2):S217–S218. (Abstract) problem in the animal house. Lab Anim
33. Ma H et al (2010) Effects of diet-induced 8:347–354
obesity and voluntary wheel running on 48. Arakawa H, Cruz S, Deak T (2011) From
bone properties in young male C57BL/6J models to mechanisms: odorant communica-
mice. Calcif Tissue Int 86:411–419 tion as a key determinant of social behavior in
34. Obernier JA, Baldwin RL (2006) Establishing rodents during illness-associated states. Neu-
an appropriate period of acclimatization fol- rosci Biobehav Rev 35(9):1916–1928
lowing transportation of laboratory animals. 49. Alves GJ et al (2010) Odor cues from tumor-
ILAR J 47(4):364–369 bearing mice induces neuroimmune changes.
35. Tuli JS, Smith JA, Morton DB (1995) Stress Behav Brain Res 214:357–367
measurements in mice after transportation. 50. Conour LA, Murray KA, Brown MJ (2006)
Lab Anim 29:132–138 Preparation of animals for research—issues to
36. Jennings M et al (1998) Refining rodent hus- consider for rodents and rabbits. ILAR J 47
bandry: the mouse. Report of the Rodent (4):283–293
Refinement Working Party. Lab Anim 32 51. Balcombe JP, Barnard ND, Sandusky C
(3):233–259 (2004) Laboratory routines cause animal
37. Clénet F et al (2006) Light/dark cycle manip- stress. Contemp Top Lab Anim Sci 43
ulation influences mice behavior in the ele- (6):42–51
vated plus maze. Behav Brain Res 166 52. Hurst JL, West RS (2010) Taming anxiety in
(1):140–149 laboratory mice. Nat Methods 7
38. Ciarleglio CM et al (2009) Population encod- (10):825–826
ing by circadian clock neurons organizes cir- 53. Sun L, Min L, Zhou H et al (2017) Adoles-
cadian behavior. J Neurosci 29 cent social isolation affects schizophrenia-like
(6):1670–1676 behavior and astrocyte biomarkers in the PFC
39. Goulding EH et al (2008) A robust auto- of adult rats. Behav Brain Res 333:258–266.
mated system elucidates mouse home cage https://doi.org/10.1016/j.bbr.2017.07.
behavioral structure. Proc Natl Acad Sci U S 011
A 105(52):20575–20582 54. Koike H et al (2009) Behavioral abnormality
40. National Research Council of the National and pharmacologic response in social
Academies (2011) Guide for the care and isolation-reared mice. Behav Brain Res 202
use of laboratory animals. National Academy (1):114–121
of Sciences, Washington, DC 55. Ma X et al (2011) Social isolation-induced
41. Buchanan JB et al (2008) Cognitive and neu- aggression potentiates anxiety and
roinflammatory consequences of mild depressive-like behavior in male mice sub-
repeated stress are exacerbated in aged mice. jected to unpredictable chronic mild stress.
Psychoneuroendocrinology 33(6):755–765 PLoS One 6(6):e20955
42. Ajarem JS, Safar E, Ahmad M (2011) Effect of 56. Avitsur R, Stark JL, Sheridan JF (2001) Social
ethanol and thermal stresses on the social stress induces glucocorticoid resistance in
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 255
depression and anxiety. Nat Rev Drug Discov 96. Takao K, Miyakawa T (2006) Light/dark
4:775–790 transition test for mice. J Vis Exp:
84. Gould TD, Dao DT, Kovacsics CE (2009) e104–e104. https://doi.org/10.3791/104
The open field test. In: Gould T (ed) Mood 97. Cryan JF, Mombereau C, Vassout A (2005)
and anxiety related phenotypes in mice: char- The tail suspension test as a model for asses-
acterization using behavioral tests. Humana sing antidepressant activity: review of pharma-
Press, New York cological and genetic studies in mice.
85. Petit-Demouliere B, Chenu F, Bourin M Neurosci Biobehav Rev 29(4–5):571–625
(2004) Forced swimming test in mice: a 98. Porsolt RD et al (2001) Rodent models of
review of antidepressant activity. Psychophar- depression: forced swimming and tail suspen-
macology (Berl) 177:245–255 sion behavioral despair tests in rats and mice.
86. Avgustinovich DF, Lipina TV, Bondar NP Curr Protoc Neurosci:8.10A.1–8.10A.10
et al (2000) Features of the genetically 99. Lad HV et al (2007) Quantitative traits for the
defined anxiety in mice. Behav Genet tail suspension test: automation, optimiza-
30:101–109 tion, and BXD RI mapping. Mamm Genome
87. Moon ML, Joesting JJ, Blevins NA et al 18:482–491
(2015) IL-4 knock out mice display anxiety- 100. Steru L et al (1985) The tail suspension test: a
like behavior. Behav Genet 45:451–460. new method for screening antidepressants in
https://doi.org/10.1007/s10519-015- mice. Psychopharmacology (Berl)
9714-x 85:367–370
88. Muralidharan A, Kuo A, Jacob M et al (2016) 101. Cryan JF, Markou A, Lucki I (2002) Asses-
Comparison of burrowing and stimuli-evoked sing antidepressant activity in rodents: recent
pain behaviors as end-points in rat models of developments and future needs. Trends Phar-
inflammatory pain and peripheral neuropathic macol Sci 23(5):238–245
pain. Front Behav Neurosci 10:88. https:// 102. Cryan JF, Page ME, Lucki I (2005) Differen-
doi.org/10.3389/fnbeh.2016.00088 tial behavioral effects of the antidepressants
89. Tonelli LH et al (2009) Allergic rhinitis ¬reboxetine, fluoxetine, and moclobemide in
induces anxiety-like behavior and altered a modified forced swim test following chronic
social interaction in rodents. Brain Behav treatment. Psychopharmacology (Berl)
Immun 23:784–793 182:335–344
90. Komada M, Takao K, Miyakawa T (2008) 103. Moreau M et al (2008) Innoculation of Bacil-
Elevated plus maze for mice. J Vis Exp 22: lus Calmette-Guerin to mice induces an acute
e1088 episode of sickness behavior followed by
91. Fromm L et al (2004) Magnesium attenuates chronic depressive-like behavior. Brain Behav
post-traumatic depression/anxiety following Immun 22(7):1087–1095
diffuse traumatic brain injury in rats. J Am 104. Udo H et al (2008) Enhanced adult neuro-
Coll Nutr 23(5):529S–533S genesis and angiogenesis and altered affective
92. Shepherd JK et al (1994) Behavioural and behaviors in mice overexpressing vascular
pharmacological characterisation of the ele- endothelial growth factor 120. J Neurosci
vated “zero-maze” as an animal model of anx- 28(53):14522–14536
iety. Psychopharmacology (Berl) 116:56–64 105. Hédou G et al (2001) An automated analysis
93. Heisler LK et al (1998) Elevated anxiety and of rat behavior in the forced swim test. Phar-
antidepressant-like responses in serotonin macol Biochem Behav 70(1):65–76
5-HT1A receptor knockout mutant mice. 106. DSM-IV (1994) Diagnostic and statistical
Proc Natl Acad Sci U S A 95:15049–15054 manual of mental disorders, 4th edn. Ameri-
94. Ennaceur A (2014) Tests of unconditioned can Psychological Association, Washington,
anxiety — pitfalls and disappointments. DC
Physiol Behav 135:55–71. https://doi.org/ 107. Strekalova T, Steinbusch H (2009) Factors of
10.1016/j.physbeh.2014.05.032 reproducibility of anhedonia induction in a
95. Hascoë M, Bourin M (2009) The mouse chronic stress depression model in mice. In:
light–dark box test. In: Gould T (ed) Mood Gould T (ed) Mood and anxiety related phe-
and anxiety related phenotypes in mice, Neu- notypes in mice: characterization using
romethods, vol 42. Humana Press, Totowa, behavioral tests. Humana Press, New York
NJ. https://doi.org/10.1007/978-1- 108. Niemi MB et al (2008) Neuro-immune asso-
60761-303-9_11 ciative learning. In: Lajtha A, Galoyan A,
Besedovski HO (eds) Handbook of
Mouse Testing Methods in Psychoneuroimmunology 2.0: Measuring Behavioral. . . 257
neurochemistry and molecular neurobiology. 121. Harrison FE et al (2006) Spatial and nonspa-
Springer, New York tial escape strategies in the Barnes maze.
109. Brennan PA, Keverne EB (1997) Neural Learn Mem 13(6):809–819
mechanisms of mammalian olfactory learning. 122. O’Leary TP, Brown RE (2009) Visuo-spatial
Prog Neurobiol 51(4):457–481 learning and memory deficits on the Barnes
110. Bryan KJ et al (2009) Chapter 1: transgenic maze in the 16-month-old APPswe/PS1dE9
mouse models of Alzheimer’s disease: mouse model of Alzheimer’s disease. Behav
¬behavioral testing and considerations. In: Brain Res 201:120–127
Buccafusco JJ (ed) Methods of behavior anal- 123. Vorhees CV, Williams MT (2006) Morris
ysis in neuroscience, 2nd edn. CRC Press, water maze: procedures for assessing spatial
Boca Raton, FL and related forms of learning and memory.
111. Bevins RA, Besheer J (2006) Object recogni- Nat Protoc 1(2):848–858
tion in rats and mice: a one-trial non-match- 124. Rosczyk HA, Sparkman NL, Johnson RW
ing-to-sample learning task to study (2008) Neuroinflammation and cognitive
‘recognition memory’. Nat Protoc 1 function in aged mice following minor sur-
(3):1306–1311 gery. Exp Gerontol 43:840–846
112. Stefanko DP et al (2009) Modulation of long- 125. Carli M, Robbins TW, Evenden JL, Everitt BJ
term memory for object recognition via (1983) Effects of lesions to ascending norad-
HDAC inhibition. Proc Natl Acad Sci U S A renergic neurones on performance of a
106(23):9447–9452 5-choice serial reaction task in rats; implica-
113. Chiu GS, Chatterjee D, Darmody PT et al tions for theories of dorsal noradrenergic
(2012) Hypoxia/reoxygenation impairs bundle function based on selective attention
memory formation via adenosine-dependent and arousal. Behav Brain Res 9:361–380
activation of caspase 1. J Neurosci 126. Horner AE, Heath CJ, Hvoslef-Eide M et al
32:13945–13955. https://doi.org/10. (2013) The touchscreen operant platform for
1523/JNEUROSCI.0704-12.2012 testing learning and memory in rats and mice.
114. Win-Shwe TT, Fujimaki H (2012) Acute Nat Protoc 8:1961–1984. https://doi.org/
administration of toluene affects memory 10.1038/nprot.2013.122
retention in novel object recognition test 127. Bushnell PJ, Strupp BJ (2009) Assessing
and memory function-related gene expression attention in rodents. CRC Press/Taylor &
in mice. J Appl Toxicol 32(4):300–304 Francis, Boca Raton, FL
115. Gainey SJ, Kwakwa KA, Bray JK et al (2016) 128. Young JW, Light GA, Marston HM et al
Short-term high-fat diet (HFD) induced (2009) The 5-choice continuous perfor-
anxiety-like behaviors and cognitive mance test: evidence for a translational test
impairment are improved with treatment by of vigilance for mice. PLoS One 4:e4227.
glyburide. Front Behav Neurosci 10:156. https://doi.org/10.1371/journal.pone.
https://doi.org/10.3389/fnbeh.2016. 0004227
00156 129. Grissom NM, Herdt CT, Desilets J et al
116. Towers AE, Oelschlager ML, Patel J et al (2015) Dissociable deficits of executive func-
(2017) Acute fasting inhibits central caspase- tion caused by gestational adversity are linked
1 activity reducing anxiety-like behavior and to specific transcriptional changes in the pre-
increasing novel object and object location frontal cortex. Neuropsychopharmacology
recognition. Metabolism 71:70–82. https:// 40:1353–1363. https://doi.org/10.1038/
doi.org/10.1016/j.metabol.2017.03.005 npp.2014.313
117. Wehner JM, Radcliffe RA (2004) Cued and 130. Dishman RK et al (2006) Neurobiology of
contextual fear conditioning in mice. Curr exercise. Obesity 14:345–346
Protoc Neurosci Chapter 8:Unit 8.5C 131. Leasure JL, Jones M (2008) Forced a volun-
118. Deacon RMJ, Rawlins JNP (2006) T-maze tary exercise differentially affect brain and
alternation in the rodent. Nat Protoc 1 behavior. Neuroscience 156:456–465
(1):7–12 132. Garland TH Jr et al (2011) The biological
119. Lalonde R (2002) The neurological basis of control of voluntary exercise, spontaneous
spontaneous alternation. Neurosci Biobehav physical activity and daily energy expenditure
Rev 26:91–104 in relation to obesity: human and rodent per-
120. Bekker A et al (2006) Isoflurane preserves spectives. J Exp Biol 214(pt 2):206–229
spatial working memory in adult mice after 133. Noakes TD (2011) Time to move behind a
moderate hypoxia. Anesth Analg brainless exercise physiology: the evidence for
102:1134–1138
258 Albert E. Towers et al.
Abstract
The link between systemic autoimmunity, brain pathology, and aberrant behavior is still a largely unex-
plored field of biomedical science. Accumulating evidence points to causal relationships between immune
factors, neurodegeneration, and neuropsychiatric manifestations. By documenting autoimmunity-
associated neuronal degeneration and cytotoxicity of the cerebrospinal fluid from disease-affected subjects,
the murine MRL model had shown high validity in revealing principal pathogenic circuits. In addition,
unlike any other autoimmune strain, MRL mice produce antibodies commonly found in patients suffering
from lupus and other autoimmune disorders. This review highlights importance of the MRL model as a
useful preparation in understanding the links between immune system and brain function.
1 Introduction
1.1 Regulatory The basic principle of life is to adapt to the relentlessly changing
Metasystem environment, thus providing the basis for survival of an individual
and continuity of a species. When challenged by external and
internal stressors, functional homeostasis in mammals regulates
itself through a coordinated network of the regulatory metasystem,
which comprises diverse interactions between nervous, endocrine,
and immune systems (Fig. 1). While the nervous system is hard-
wired to endocrine glands and immune organs via autonomic
fibers, the immune system communicates with other tissues by
secreting various soluble messengers, such as cytokines, chemo-
kines, and proteins of the complement system [1]. These mediators
can affect brain function by activating the vagal and other nerves
[2], and the secondary messenger system of endothelial cells in
brain vasculature [3], or by diffusing into the brain parenchyma if
the blood–brain barrier becomes more permeable. These factors
can also alter behavior indirectly by changing hormonal activity in
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_14, © Springer Science+Business Media, LLC, part of Springer Nature 2018
259
260 Boris Šakić
Regulatory Metasystem
Behavior
Nervous System
HOMEOSTASIS
cytokines
Immune System Endocrine System
hormones (e.g., steroids)
• viruses/bacteria
external stressors
• toxins
• oncogens
internal stressors
• autoantigens
Fig. 1 The principal network of mammal regulatory metasystem. When challenged by various external and
internal stressors, functional homeostasis regulates itself through coordinated interactions between neuroen-
docrine and immune systems, which form the regulatory metasystem. While the CNS is hard-wired to
endocrine glands and immune organs via fibers of the autonomic nervous system, ANS, the immune system
communicates with other organs and tissues via release of various soluble messengers that affect the brain
either directly via neural pathways (e.g., nervous vagus) or indirectly, by affecting hormonal activity in major
endocrine pathways, such as hypothalamic-pituitary-adrenal axis. Imbalanced homeostasis (or allostasis)
ultimately results in increased production of steroids and altered behavior, which are part of the adaptive
responses to acute and chronic stressors. Acute sickness behavior and cortisol release (corticosterone in
animals) are examples of an adaptive mechanism to acute exposure to pathogens. Different forms of mental
and neurodegenerative CNS illnesses are proposed to ensue in genetically susceptible organisms when
disturbances in their regulatory metasystem become severe and chronic. Systemic autoimmune diseases
triggered by myriad of autoantigens (such as in SLE) are example of chronic activation in the regulatory
metasystem
The MRL Model 261
2.1 Development Animal models in psychiatry are not replicas of human mental
disorders. Nevertheless, they represent useful preparations where
analogous disease traits, underlying mechanisms, or drug effects
can be studied in a systematic and controlled manner. When asses-
sing the validity of a specific model it is important to identify its
limitations and characteristics that distinguish it from human
pathologies. It is equally important to then consider the choice of
a proper control group and the assessment of central nervous
system damage induced specifically by disease progress.
Most commonly studied spontaneous models of SLE include
the NZB, (NZB NZW)F1 hybrid, BXSB, and MRL mouse
strains. They are characterized by a wide spectrum of autoimmune
The MRL Model 263
MRL/MpJ substrain
(stock 486)
MRL/MpJ-Faslpr/J substrain
(stock 485)
Fig. 2 Genetic background of the MRL mouse strain. Four different strains of mice were used in mid-1970s to
produce the lupus-prone MRL (originally Murphy—Root Large) strain in The Jackson Laboratory. In compari-
son to the original stock 486, spontaneous Faslpr mutation accounted for accelerated disease development in
the stock 485 (MRL/lpr mice, shown on the picture). However, unexplained decline in autoimmune phenotype
was observed in this stock over the past decade. The subsequent stock was labeled 6825. As the stock 485, it
carries the Faslpr mutation, thus rendering this group a more adequate control in behavioral studies that test
causal links between systemic autoimmunity and brain damage
2.2 Autoimmunity- MRL/lpr and MRL+/+ mice are comparable in many respects
Associated Behavioral (e.g., appearance, size, reproductive age), except in the onset of
Syndrome autoimmunity and neurobehavioral dysfunction. Disease progres-
sion in MRL/lpr mice parallels emergence of behavioral deficits
often seen after exposure to chronic stressors [38]. The constella-
tion of behavioral differences between age- and sex-matched
MRL/lpr and MRL +/+ mice was operationally termed “autoim-
munity-associated behavioral syndrome” (AABS), which has been
largely characterized by progressive anxiety- and depressive-like
behaviors [39]. These constructs were supported by increased
thigmotaxis of MRL/lpr mice in large arenas, impaired exploration
of novel objects and spaces, performance deficits in the plus-maze
and step-down tests, excessive floating in the forced swim test
[56–58], reduced responsiveness to palatable stimulation
[51, 59], and reduced isolation-induced inter-male fighting
[60]. Moreover, their “cognitive inflexibility” and poor spatial
learning were noted in the Morris water maze [61, 62] and sponta-
neous alternation test [53]. In addition, diseased MRL/lpr animals
show lower nocturnal and open-field activity levels, as well as
significant deficiencies in neurological [62, 63] and psychomotor
tests [64].
2.3 Breached The blood–brain barrier (BBB) is formed by endothelial cells that
Blood–Brain Barrier tightly line capillaries and blood vessels of the brain. The BBB has
an important role in maintaining a well-regulated CNS microenvi-
ronment for reliable neuronal signaling. When its normal function
is disrupted, large molecules and cells can infiltrate into the brain
and lead to CNS damage. In many CNS SLE patients, the BBB
becomes transiently or permanently breached, as evidenced by an
increased albumin quotient [65]. Increased BBB permeability is an
important permissive condition which allows diffusion of cytokines
and anti-neuronal autoantibodies into cerebrospinal fluid (CSF)
and accounts for subsequent CNS manifestations [66].
Similar to patients, a breached BBB has been observed in
diseased MRL/lpr mice. This damage occurs at a very early age
266 Boris Šakić
2.5 Brain Pathology Clinical studies clearly demonstrate that NP manifestations are
accompanied by cerebral atrophy [93], progressive neuronal loss
[20, 28], and parenchymal lesions [94]. Similar to CNS SLE
patients and effects of chronic stress, MRL/lpr mice show brain
atrophy and ventricular enlargement alongside behavioral deficits
detected at the onset of SLE-like disease [53, 95, 96].
The MRL strain does not show a high incidence of inherited
neuroanatomical abnormalities [41], which minimize the possibil-
ity of congenital defects confounding the study of disease-induced
neurodegeneration. At the onset of autoimmune symptoms in
MRL/lpr mice, reports of reduced complexity of pyramidal neu-
rons, reduced brain weights [96], and selectively neurotoxic CSF
[97] provided indirect evidence of neuronal damage in diseased
animals. Direct evidence of neuronal death, however, was first
confirmed in MRL/lpr brains using the Fluoro Jade B (FJB) cyto-
chemical stain (specific for dying neurons). A small percentage of
these neurons were subsequently found to contain TdT-labeled
apoptotic nuclei, and co-localized with FJB [79] and anti-
neurofilament staining [76]. Moreover, while the size of hippocam-
pal fields and neuronal density are not reduced in young
Fas-deficient lpr mice [98], cell densities are reduced within the
hippocampus, cortex [53], and midbrain [90] of aged/diseased
lupus mice.
In addition to mature neurons, recent findings suggest that
progenitor cells also degenerate in MRL/lpr brains. More specifi-
cally, the subventricular zone [99], subgranular zone [78, 79], and
substantia nigra [90], all of which are known to contain prolifera-
tive progenitor cells capable of neurogenesis [100], show signs of
damage. CSF from diseased lupus mice is also cytotoxic to neurons
and neuronal progenitor cells in vitro [101], thus supporting a link
between toxic CSF IgG and neuronal/progenitor cell damage
[68]. If in vitro findings are predictive of in vivo events, then
autoimmune-induced lesions of germinal layers may reduce the
developmental and regenerative capacity of MRL/lpr brains. An
impairment in this process would likely exacerbate subsequent
autoimmune/inflammatory-mediated neuronal death and behav-
ioral deficits. For example, an impaired capacity for hippocampal
neurogenesis could account for the cognitive impairments
observed in these animals [53]. Stress hormones, known to be
chronically elevated in lupus mice [102], have also been shown to
inhibit cell proliferation and neurogenesis [103]. Therefore, one
may assume that such mechanisms account for impaired brain
growth and regeneration along the progression of autoimmune
disease.
Despite parallels between the emergence of behavioral dysfunc-
tion and systemic autoimmunity, there is no firm evidence that
brain pathology accounts for aberrant behavior in the MRL
model. However, significant correlations suggest principal links
268 Boris Šakić
2.6 Proposed The causative role of autoimmunity and inflammation in the path-
Neuropathogenic ogenesis of AABS has been supported by studies employing the
Factors and immunosuppressive drug cyclophosphamide (CY). Sustained treat-
Mechanisms ment with CY from an early age prevents several behavioral deficits
and brain pathology in MRL/lpr mice [51–53, 70]. More specifi-
cally, CY abolishes substrain differences in anxiety- and motivation-
related behaviors, as suggested by restored novel object explora-
tion, increased responsiveness to a palatable sucrose solution, and
normalized nocturnal activity. Although systemic autoimmunity
and inflammation have been proposed as key factors, the possibility
that subtle genetic dissimilarities, imbalanced hormonal produc-
tion, and peripheral tissue involvement contribute to certain
aspects of AABS could not be discounted. However, the use of
newly developed stock 6825 rejected the possibility that AABS is
entirely accounted for by FasR mutation in neuronal cells. Namely,
the constellation of differences between the lpr stocks 485 and
6825 confirmed the hypothesis that the lpr mutation per se does
not fully account for the brain pathology and altered behavior in
the stock 485 [106]. Together with significant correlations
between immunological and behavioral measures, this study sug-
gested that soluble immune factors play a key role in CNS patho-
genesis. In addition, combined use of immunoprecipitation with
homogenates of unaffected brains, 2-dimensional differential in-gel
electrophoresis, and mass spectrometry revealed strong binding of
CSF IgG antibodies to cytoskeletal antigens in brains of MRL/lpr
mice. This finding is consistent with the proposed pathogenic role
of brain-reactive autoantibodies (BRA) in the etiology of AABS.
CNS SLE is frequently accompanied by BRA cross-reactive
with diverse brain-specific and systemic antigens [12]. Most of
these autoantibodies have been identified on the basis of their
binding to tissues and cells, including neuroblastoma and glioblas-
toma cell lines [107]. There are also autoantibodies against lym-
phocytes, capable of being adsorbed by brain tissue
[108, 109]. Some antibodies can react specifically to CNS neurons
[110, 111], neuronal cytoplasm [112], and neuronal receptors
[113, 114]. A current literature review proposes approximately
20 pathogenic BRA [115], including anti-ribosomal, anticardioli-
pin, antiphospholipid, and more recently antibodies to an NMDA
The MRL Model 269
leukocyte entrapment
7.
ACTH
over-expression of CAM activation of nervous vagus
4.
3.
lymphocyte hyperactivity 6. 1.
SHs
5. 1. SHs
2.
Pro-inflammatory cytokines
leukocyte infiltration via CP, BRA to neurons and neural stem cells
microglial activation
9.
8.
Fig. 3 Proposed phases and pathways of the CNS damage during systemic autoimmune disease. Behavioral
dysfunction and brain damage in lupus-like disease may result from chronic stress-like response induced by
sustained autoimmunity and inflammation. In SLE patients and lupus-prone MRL/lpr mice, spontaneous onset
of systemic inflammation and autoimmunity are characterized by increased levels of pro-inflammatory
cytokines, which may activate pituitary-adrenal axis and promote sustained release of glucocorticoids. In
turn, steroid hormones suppress the immune system at multiple levels. Due to chronic nature of the disease,
glucocorticoids, cytokines, and other immune components remain elevated, thus compromising the integrity
of the blood–brain barrier and neuronal function. (A) The inflammatory phase is largely associated with early
272 Boris Šakić
Fig. 3 (continued) functional damage of the brain. The upregulation in circulating pro-inflammatory cytokines
(e.g., IL-1β, IL-6, TNF-α, and IFN-γ) is an initial serologic event that disturbs the activity of the immune
network. These cytokines can also activate the hypothalamic-pituitary-adrenal axis [1], which downregulates
peripheral inflammation via increased production of steroid hormones, SHs [2]. However, in addition to the
activation of major inhibitory signals from the hippocampus to hypothalamic paraventricular nucleus,
sustained binding of steroids to receptors in central neurons [3] induces stress-like manifestations (e.g.,
emotional disturbances, impaired mood) which are largely under the control of the limbic system. This effect
on brain function is further amplified by cytokine-induced activation of the nervous vagus and activation of
glial cells in the hypothalamus [4]. Moreover, activated lymphocytes [5] and cytokine-induced overexpression
of cell adhesion molecules on endothelial cells of the BBB and choroid plexus, CP [6], are conducive of
immunocyte entrapment [7]. (B) The autoimmune phase is largely characterized by structural damage, such
as neurodegeneration and brain atrophy. Chronic inflammatory responses increase the permeability of the
BBB and CP, thus leading to infiltration of immunocytes into perivascular spaces and cerebrospinal fluid
[8]. Structural brain damage can result from neurotoxic metabolites that accumulate after sustained activation
of microglia [9] and chronic binding of brain-reactive antibodies (BRA) to adult and immature neurons
[10]. Loss of periventricular and cortical mass may underlie psychosis, dementia, and seizures that frequently
accompany neuropsychiatric lupus. Note: Dashed lines represent inhibitory pathways
The MRL Model 273
neurons
B cells
antibodies
MAC
complement
T cells chemokines
cytokines MMP
ROS oligodendrocytes
macrophages
blood-brain barrier
PGE
microglia
astrocytes
Fig. 4 Summary of putative factors and cellular mechanisms underlying neuronal damage in CNS lupus. When
BBB is breached, various immune cells and mediators can compromise the viability of brain cells at different
stages of disease progress, age, and genetic deficits in affected individuals. These factors include cytotoxic T
cells, macrophages, brain-reactive antibodies to surface and intracellular receptors, the C5b-9 MAC, MMPs,
and reactive oxygen species, ROS
Genetics
Behavioral phenotype
(PNS + CNS function)
3 Summary
Acknowledgments
This work was supported by the grants from the Ontario Mental
Health Foundation and Canadian Institutes of Heath Research.
The MRL Model 277
References
1. Ader R, Felten D, Cohen N (2001) Psycho- manifestations. Clin Exp Rheumatol 14
neuroimmunology, vol 1-2, 3rd edn. Aca- (Suppl. 16):S23–S29
demic Press, New York 14. van Dam AP, Wekking EM, Oomen HA
2. Goehler LE, Gaykema RP, Hansen MK, (1991) Psychiatric symptoms as features of
Anderson K, Maier SF, Watkins LR (2000) systemic lupus erythematosus. Psychother
Vagal immune-to-brain communication: a vis- Psychosom 55(2–4):132–140
ceral chemosensory pathway. Auton Neurosci 15. Colamussi P, Giganti M, Cittanti C,
85(1–3):49–59 Dovigo L, Trotta F, Tola MR, Tamarozzi R,
3. Benveniste EN, Huneycutt BS, Shrikant P, Lucignani G, Piffanelli A (1995) Brain single-
Ballestas ME (1995) Second messenger sys- photon emission tomography with
tems in the regulation of cytokines and adhe- Tc-99m-HMPAO in neuropsychiatric sys-
sion molecules in the central nervous system. temic lupus erythematosus: relations with
Brain Behav Immun 9:304–314 EEG and MRI findings and clinical manifesta-
4. Dunn AJ (2000) Cytokine activation of the tions. Eur J Nucl Med 22:17–24
HPA axis. Ann N Y Acad Sci 917:608–617 16. Handa R, Sahota P, Kumar M, Jagannathan
5. McEwen BS (2000) Allostasis and allostatic NR, Bal CS, Gulati M, Tripathi BM, Wali JP
load: implications for neuropsychopharma- (2003) In vivo proton magnetic resonance
cology. Neuropsychopharmacology 22 spectroscopy (MRS) and single photon emis-
(2):108–124 sion computerized tomography (SPECT) in
6. Johnson EO, Kamilaris TC, Chrousos GP, systemic lupus erythematosus (SLE). Magn
Gold PW (1992) Mechanisms of stress: a Reson Imaging 21(9):1033–1037
dynamic overview of hormonal and behavioral 17. Huang WS, Chiu PY, Tsai CH, Kao A, Lee
homeostasis. Neurosci Biobehav Rev CC (2002) Objective evidence of abnormal
16:115–130 regional cerebral blood flow in patients with
7. Dantzer R, O’Connor JC, Freund GG, John- systemic lupus erythematosus on Tc-99m
son RW, Kelley KW (2008) From inflamma- ECD brain SPECT. Rheumatol Int 22
tion to sickness and depression: when the (5):178–181
immune system subjugates the brain. Nat 18. Lopez-Longo FJ, Carol N, Almoguera MI,
Rev Neurosci 9(1):46–56 Olazaran J, onso-Farto JC, Ortega A,
8. McEwen BS (2003) Mood disorders and allo- Monteagudo I, Gonzalez CM, Carreno L
static load. Biol Psychiatry 54(3):200–207 (2003) Cerebral hypoperfusion detected by
SPECT in patients with systemic lupus erythe-
9. McEwen BS, Biron CA, Brunson KW, matosus is related to clinical activity and
Bulloch K, Chambers WH, Dhabhar FS, cumulative tissue damage. Lupus 12
Goldfarb RH, Kitson RP, Miller AH, Spencer (11):813–819
RL, Weiss JM (1997) The role of adrenocor-
ticoids as modulators of immune function in 19. Komatsu N, Kodama K, Yamanouchi N,
health and disease: neural, endocrine and Okada S, Noda S, Nawata Y, Takabayashi K,
immune interactions. Brain Res Rev 23 Iwamoto I, Saito Y, Uchida Y, Ito H,
(1–2):79–133 Yoshikawa K, Sato T (1999) Decreased
regional cerebral metabolic rate for glucose
10. Cotman CW, Brinton RE, Galaburda A, in systemic lupus erythematosus patients
McEwen B, Schneider DM (1987) The with psychiatric symptoms. Eur Neurol 42
neuro-immune-endocrine connection. Raven (1):41–48
Press, New York
20. Brooks WM, Sabet A, Sibbitt WL, Barker PB,
11. Kapadia M, Sakic B (2011) Autoimmune and van Zijl PC, Duyn JH, Moonen CT (1997)
inflammatory mechanisms of CNS damage. Neurochemistry of brain lesions determined
Prog Neurobiol 95(3):301–333 by spectroscopic imaging in systemic lupus
12. Hanly JG (2005) Neuropsychiatric lupus. erythematosus. J Rheumatol 24
Rheum Dis Clin North Am 31(2):273–298 (12):2323–2329
13. Tincani A, Brey R, Balestrieri G, Vitali C, 21. Volkow ND, Warner N, McIntyre R,
Doria A, Galeazzi M, Meroni PL, Valentine A, Kulkarni M, Mullani N, Gould
Migliorini P, Neri R, Tavoni A, Bombardieri L (1988) Cerebral involvement in systemic
S (1996) International survey on the manage- lupus erythematosus. Am J Physiol Imaging
ment of patients with SLE .2. The results of a 3(2):91–98
questionnaire regarding neuropsychiatric
278 Boris Šakić
22. Gonzalez-Scarano F, Lisak RP, Bilaniuk LT, 32. Hirohata S, Hirose S, Miyamoto T (1985)
Zimmerman RA, Atkins PC, Zweiman B Cerebrospinal fluid IgM, IgA, and IgG
(1979) Cranial computed tomography in the indexes in systemic lupus erythematosus.
diagnosis of systemic lupus erythematosus. Their use as estimates of central nervous sys-
Ann Neurol 5(2):158–165 tem disease activity. Arch Intern Med 145
23. Kaell AT, Shetty M, Lee BC, Lockshin MD (10):1843–1846
(1986) The diversity of neurologic events in 33. Winfield JB, Shaw M, Silverman LM, Eisen-
systemic lupus erythematosus. Prospective berg RA, Wilson HA III, Koffler D (1983)
clinical and computed tomographic classifica- Intrathecal IgG synthesis and blood-brain
tion of 82 events in 71 patients. Arch Neurol barrier impairment in patients with systemic
43(3):273–276 lupus erythematosus and central nervous sys-
24. Miguel EC, Pereira RM, Pereira CA, Baer L, tem dysfunction. Am J Med 74(5):837–844
Gomes RE, de Sa LC, Hirsch R, de Barros 34. Yoshio T, Hirata D, Onda K, Nara H, Minota
NG, de Navarro JM, Gentil V (1994) Psychi- S (2005) Antiribosomal P protein antibodies
atric manifestations of systemic lupus erythe- in cerebrospinal fluid are associated with neu-
matosus: clinical features, symptoms, and ropsychiatric systemic lupus erythematosus. J
signs of central nervous system activity in Rheumatol 32(1):34–39
43 patients. Medicine 73(4):224–232 35. Greenwood DL, Gitlits VM, Alderuccio F,
25. Omdal R, Selseth B, Klow NE, Husby G, Sentry JW, Toh BH (2002) Autoantibodies
Mellgren SI (1989) Clinical neurological, in neuropsychiatric lupus. Autoimmunity 35
electrophysiological, and cerebral CT scan (2):79–86
findings in systemic lupus erythematosus. 36. Henn FA, McKinney WT (1987) Animal
Scand J Rheumatol 18(5):283–289 models in psychiatry. In: Meltzer HY
26. Ainiala H, Dastidar P, Loukkola J, (ed) Psychopharmacology: the third genera-
Lehtimaki T, Korpela M, Peltola J, Hietaharju tion of progress. Raven Press, New York, pp
A (2005) Cerebral MRI abnormalities and 687–695
their association with neuropsychiatric mani- 37. Dixon FJ, Andrews BS, Eisenberg RA, McCo-
festations in SLE: a population-based study. nahey PJ, Theofilopoulos AN, Wilson CB
Scand J Rheumatol 34(5):376–382 (1978) Etiology and pathogenesis of a spon-
27. Waterloo K, Omdal R, Jacobsen EA, Klow taneous lupus-like syndrome in mice. Arthritis
NE, Husby G, Torbergsen T, Mellgren SI Rheum 21:S64–S67
(1999) Cerebral computed tomography and 38. Szechtman H, Sakic B, Denburg JA (1997)
electroencephalography compared with neu- Behaviour of MRL mice: an animal model of
ropsychological findings in systemic lupus disturbed behaviour in systemic autoimmune
erythematosus. J Neurol 246(8):706–711 disease. Lupus 6(3):223–229
28. Sibbitt WL, Sibbitt RR (1993) Magnetic res- 39. Sakic B, Szechtman H, Denburg JA (1997)
onance spectroscopy and positron emission Neurobehavioral alteration in autoimmune
tomography scanning in neuropsychiatric sys- mice. Neurosci Biobehav Rev 21(3):327–340
temic lupus erythematosus. Rheum Dis Clin 40. Alexander JJ, Quigg RJ (2007) Systemic
North Am 19(4):851–868 lupus erythematosus and the brain: what
29. Sibbitt WL, Haseler LJ, Griffey RH, Hart BL, mice are telling us. Neurochem Int 50
Sibbitt RR, Matwiyoff NA (1994) Analysis of (1):5–11
cerebral structural changes in systemic lupus 41. Sherman GF, Galaburda AM, Behan PO,
erythematosus by proton MR spectroscopy. Rosen GD (1987) Neuroanatomical anoma-
AJNR Am J Neuroradiol 15:923–928 lies in autoimmune mice. Acta Neuropathol
30. Jennekens FG, Kater L (2002) The central (Berl ) 74:239–242
nervous system in systemic lupus erythemato- 42. Theofilopoulos AN (1992) Murine models of
sus. Part 2. Pathogenetic mechanisms of clini- lupus. In: Lahita RG (ed) Systemic lupus
cal syndromes: a literature investigation. erythematosus, 2nd edn. Churchill Living-
Rheumatology (Oxford) 41(6):619–630 stone, New York, pp 121–194
31. McLean BN, Miller D, Thompson EJ (1995) 43. Andrews BS, Eisenberg RA, Theofilopoulos
Oligoclonal banding of IgG in CSF, blood- AN, Izui S, Wilson CB, McConahey PJ, Mur-
brain barrier function, and MRI findings in phy ED, Roths JB, Dixon FJ (1978) Sponta-
patients with sarcoidosis, systemic lupus neous murine lupus-like syndromes. Clinical
erythematosus, and Behcet’s disease involving and immunopathological manifestations in
the nervous system. J Neurol Neurosurg Psy- several strains. J Exp Med 148:1198–1215
chiatry 58(5):548–554
The MRL Model 279
murine models of central nervous system dis- 82. Alexander EL, Alexander GE (1983) Aseptic
ease in connective tissue disorders. Ann meningoencephalitis in primary Sjogren’s
Neurol 14:242–248 syndrome. Neurology 33:593–598
70. Farrell M, Sakic B, Szechtman H, Denburg JA 83. Auer RN, Wieloch T, Olsson Y, Siesjo BK
(1997) Effect of cyclophosphamide on leuco- (1984) The distribution of hypoglycemic
cytic infiltration in the brain of MRL/lpr brain damage. Acta Neuropathol 64
mice. Lupus 6(3):268–274 (3):177–191
71. Ma X, Foster J, Sakic B (2006) Distribution 84. Fujioka M, Okuchi K, Hiramatsu KI,
and prevalence of leukocyte phenotypes in Sakaki T, Sakaguchi S, Ishii Y (1997) Specific
brains of lupus-prone mice. J Neuroimmunol changes in human brain after hypoglycemic
179(1–2):26–36 injury. Stroke 28(3):584–587
72. McHale JF, Harari OA, Marshall D, Haskard 85. Denenberg VH, Sherman GF, Rosen GD,
DO (1999) TNF-alpha and IL-1 sequentially Morrison L, Behan PO, Galaburda AM
induce endothelial ICAM-1 and VCAM-1 (1992) A behavior profile of the MRL/Mp
expression in MRL/lpr lupus-prone mice. J lpr/lpr mouse and its association with hydro-
Immunol 163(7):3993–4000 cephalus. Brain Behav Immun 6:40–49
73. Zameer A, Hoffman SA (2003) Increased 86. Alexander JJ, Zwingmann C, Quigg R (2005)
ICAM-1 and VCAM-1 expression in the MRL/lpr mice have alterations in brain
brains of autoimmune mice. J Neuroimmunol metabolism as shown with [(1)H-(13)C]
142(1–2):67–74 NMR spectroscopy. Neurochem Int 47
74. Tomita M, Holman BJ, Williams LS, Pang (1–2):143–151
KC, Santoro TJ (2001) Cerebellar dysfunc- 87. Alexander JJ, Bao L, Jacob A, Kraus DM,
tion is associated with overexpression of Holers VM, Quigg RJ (2003) Administration
proinflammatory cytokine genes in lupus. J of the soluble complement inhibitor, Crry-Ig,
Neurosci Res 64(1):26–33 reduces inflammation and aquaporin 4 expres-
75. Tomita M, Holman BJ, Santoro TJ (2001) sion in lupus cerebritis. Biochim Biophys Acta
Aberrant cytokine gene expression in the hip- 1639(3):169–176
pocampus in murine systemic lupus erythe- 88. Hampton DW, Seitz A, Chen P, Heber-
matosus. Neurosci Lett 302(2–3):129–132 Katz E, Fawcett JW (2004) Altered CNS
76. Alexander JJ, Jacob A, Bao L, Macdonald RL, response to injury in the MRL/MpJ mouse.
Quigg RJ (2005) Complement-dependent Neuroscience 127(4):821–832
apoptosis and inflammatory gene changes in 89. Lawson LJ, Perry VH, Dri P, Gordon S
murine lupus cerebritis. J Immunol 175 (1990) Heterogeneity in the distribution and
(12):8312–8319 morphology of microglia in the normal adult
77. McIntyre KR, Ayer-LeLievre C, Persson H mouse brain. Neuroscience 39(1):151–170
(1990) Class II major histocompatibility com- 90. Ballok DA, Earls AM, Krasnik C, Hoffman
plex (MHC) gene expression in the mouse SA, Sakic B (2004) Autoimmune-induced
brain is elevated in the autoimmune MRL/ damage of the midbrain dopaminergic system
Mp-lpr/lpr strain. J Neuroimmunol in lupus-prone mice. J Neuroimmunol 152
28:39–52 (1–2):83–97
78. Ballok DA, Ma X, Denburg JA, Arsenault L, 91. Hess DC (1997) Cerebral lupus vasculopathy.
Sakic B (2006) Ibuprofen fails to prevent Mechanisms and clinical relevance. Ann N Y
brain pathology in a model of neuropsychia- Acad Sci 823:154–168
tric lupus. J Rheumatol 33(11):2199–2213 92. Baraczka K, Nekam K, Pozsonyi T, Jakab L,
79. Ballok DA, Millward JM, Sakic B (2003) Szongoth M, Sesztak M (2001) Concentra-
Neurodegeneration in autoimmune MRL-lpr tion of soluble adhesion molecules (sVCAM-
mice as revealed by Fluoro Jade B staining. 1, sICAM-1 and sL-selectin) in the cerebro-
Brain Res 964(2):200–210 spinal fluid and serum of patients with multi-
80. Del Bigio MR (1993) Neuropathological ple sclerosis and systemic lupus erythematosus
changes caused by hydrocephalus. Acta Neu- with central nervous involvement. Neuroim-
ropathol 85(6):573–585 munomodulation 9(1):49–54
81. Gerber J, Raivich G, Wellmer A, Noeske C, 93. Chinn RJS, Wilkinson ID, Hallcraggs MA,
Kunst T, Werner A, Bruck W, Nau R (2001) A Paley MNJ, Shorthall E, Carter S, Kendall
mouse model of Streptococcus pneumoniae BE, Isenberg DA, Newman SP, Harrison
meningitis mimicking several features of MJG (1997) Magnetic resonance imaging of
human disease. Acta Neuropathol 101 the brain and cerebral proton spectroscopy in
(5):499–508
The MRL Model 281
patients with systemic lupus erythematosus. 104. Anderson KK, Ballok DA, Prasad N,
Arthritis Rheum 40(1):36–46 Szechtman H, Sakic B (2006) Impaired
94. Rocca MA, Agosta F, Mezzapesa DM, response to amphetamine and neuronal
Ciboddo G, Falini A, Comi G, Filippi M degeneration in the nucleus accumbens of
(2006) An fMRI study of the motor system autoimmune MRL-lpr mice. Behav Brain
in patients with neuropsychiatric systemic Res 166:32–38
lupus erythematosus. Neuroimage 30 105. Brey RL, Cote S, Barohn R, Jackson C,
(2):478–484 Crawley R, Teale JM (1995) Model for the
95. Sled JG, Spring S, van Eede M, Lerch JP, neuromuscular complications of systemic
Ullal S, Sakic B (2009) Time course and lupus erythematosus. Lupus 4(3):209–212
nature of brain atrophy in the MRL mouse 106. Loheswaran G, Stanojcic M, Xu L, Sakic B
model of central nervous system lupus. (2010) Autoimmunity as a principal patho-
Arthritis Rheum 60(6):1764–1774 genic factor in the refined model of neurop-
96. Sakic B, Szechtman H, Denburg JA, sychiatric lupus. Clin Exp Neuroimmunol
Gorny G, Kolb B, Whishaw IQ (1998) Pro- 1:141–152
gressive atrophy of pyramidal neuron den- 107. How A, Dent PB, Liao SK, Denburg JA
drites in autoimmune MRL-lpr mice. J (1985) Antineuronal antibodies in neuropsy-
Neuroimmunol 87(1–2):162–170 chiatric systemic lupus erythematosus. Arthri-
97. Maric D, Millward JM, Ballok DA, tis Rheum 28:789–795
Szechtman H, Barker JL, Denburg JA, Sakic 108. Williams GW, Bluestein HG, Steinberg AD
B (2001) Neurotoxic properties of cerebro- (1981) Brain-reactive lymphocytotoxic anti-
spinal fluid from behaviorally impaired auto- body in the cerebrospinal fluid of patients
immune mice. Brain Res 920(1–2):183–193 with systemic lupus erythematosus: correla-
98. Kovac AD, Grammig J, Mahlo J, Steiner B, tion with central nervous system involvement.
Roth K, Nitsch R, Bechmann I (2002) Com- Clin Immunol Immunopathol 18
parison of neuronal density and subfield sizes (1):126–132
in the hippocampus of CD95L-deficient 109. Bluestein HG, Zvaifler NJ (1976) Brain-
(gld), CD95-deficient (lpr) and nondeficient reactive lymphocytotoxic antibodies in the
mice. Eur J Neurosci 16(1):159–163 serum of patients with systemic lupus erythe-
99. Sakic B, Maric I, Koeberle PD, Millward JM, matosus. J Clin Invest 57(2):509–516
Szechtman H, Maric D, Denburg JA (2000) 110. Bresnihan B, Hohmeister R, Cutting J, Tra-
Increased TUNEL-staining in brains of auto- vers RL, Waldburger M, Black C, Jones T,
immune Fas-deficient mice. J Neuroimmunol Hughes GR (1979) The neuropsychiatric dis-
104(2):147–154 order in systemic lupus erythematosus: evi-
100. Yamashita T, Ninomiya M, Hernandez AP, dence for both vascular and immune
Garcia-Verdugo JM, Sunabori T, mechanisms. Ann Rheum Dis 38(4):301–306
Sakaguchi M, Adachi K, Kojima T, Hirota Y, 111. Golombek SJ, Graus F, Elkon KB (1986)
Kawase T, Araki N, Abe K, Okano H, Sawa- Autoantibodies in the cerebrospinal fluid of
moto K (2006) Subventricular zone-derived patients with systemic lupus erythematosus.
neuroblasts migrate and differentiate into Arthritis Rheum 29(9):1090–1097
mature neurons in the post-stroke adult stria- 112. Quismorio FP, Friou GJ (1972) Antibodies
tum. J Neurosci 26(24):6627–6636 reactive with neurons in SLE patients with
101. Sakic B, Kirkham DL, Ballok DA, neuropsychiatric manifestations. Int Arch
Mwanjewe J, Fearon IM, Macri J, Yu G, Allergy Appl Immunol 43:740–748
Sidor MM, Denburg JA, Szechtman H, 113. DeGiorgio LA, Konstantinov KN, Lee SC,
Lau J, Ball AK, Doering LC (2005) Prolifer- Hardin JA, Volpe BT, Diamond B (2001) A
ating brain cells are a target of neurotoxic CSF subset of lupus anti-DNA antibodies cross-
in systemic autoimmune disease. J Neuroim- reacts with the NR2 glutamate receptor in
munol 169(1–2):68–85 systemic lupus erythematosus. Nat Med 7
102. Lechner O, Dietrich H, Oliveira dos SA, Wie- (11):1189–1193
gers GJ, Schwarz S, Harbutz M, Herold M, 114. Tanaka S, Matsunaga H, Kimura M,
Wick G (2000) Altered circadian rhythms of Tatsumi K, Hidaka Y, Takano T, Uema T,
the stress hormone and melatonin response in Takeda M, Amino N (2003) Autoantibodies
lupus-prone MRL/MP-fas(Ipr) mice. J Auto- against four kinds of neurotransmitter recep-
immun 14(4):325–333 tors in psychiatric disorders. J Neuroimmunol
103. Mirescu C, Gould E (2006) Stress and adult 141(1–2):155–164
neurogenesis. Hippocampus 16(3):233–238
282 Boris Šakić
and the brain in CNS lupus. Lupus 12 and nitric oxide metabolites in neuropsychia-
(12):919–928 tric lupus erythematosus. Ann Rheum Dis 60
137. Tang B, Matsuda T, Akira S, Nagata N, (4):372–379
Ikehara S, Hirano T, Kishimoto T (1991) 149. Hayley S, Merali Z, Anisman H (2003) Stress
Age-associated increase in interleukin 6 in and cytokine-elicited neuroendocrine and
MRL/lpr mice. Int Immunol 3:273–278 neurotransmitter sensitization: implications
138. Singh AK, Lebedeva TV (1994) Interleukin-1 for depressive illness. Stress 6(1):19–32
contributes to high level IgG production in 150. Hu Y, Dietrich H, Herold M, Heinrich PC,
the murine MRL/lpr lupus model. Immunol Wick G (1993) Disturbed immuno-
Invest 23:281–292 endocrine communication via the hypotha-
139. Boswell JM, Yui MA, Endres S, Burt DW, lamo- pituitary-adrenal axis in autoimmune
Kelley VE (1988) Novel and enhanced IL-1 disease. Int Arch Allergy Immunol
gene expression in autoimmune mice with 102:232–241
lupus. J Immunol 141:118–124 151. Rivier C, Rivest S (1993) Mechanisms med-
140. Sakic B, Szechtman H, Braciak TA, Richards iating the effects of cytokines on neuroendo-
CD, Gauldie J, Denburg JA (1997) Reduced crine functions in the rat. In: Chadwick D,
preference for sucrose in autoimmune mice: a Marsh J, Ackrill K (eds) Corticotropin-
possible role of interleukin-6. Brain Res Bull releasing factor, vol Clayton Foundation
44(2):155–165 Laboratories for Peptide Biology, Salk Insti-
141. Kwant A, Sakic B (2004) Behavioral effects of tute, San Diego, CA 92128, Ciba Foundation
infection with interferon-gamma adenovec- Symposium, vol 172. John Wiley & Sons,
tor. Behav Brain Res 151(1–2):73–82 Chichester, pp 204–225
142. Sakic B, Szechtman H, Gauldie J, Denburg JA 152. Spangelo BL, Gorospe WC (1995) Role of
(2001) Behavioral effects of infection with the cytokines in the neuroendocrine-immune
IL-6 adenovector. Brain Behav Immun 15 system axis. Front Neuroendocrinol 16
(1):25–42 (1):1–22
143. Marshall D, Dangerfield JP, Bhatia VK, Larbi 153. Miller AH, Maletic V, Raison CL (2009)
KY, Nourshargh S, Haskard DO (2003) Inflammation and its discontents: the role of
MRL/lpr lupus-prone mice show exagger- cytokines in the pathophysiology of major
ated ICAM-1-dependent leucocyte adhesion depression. Biol Psychiatry 65(9):732–741
and transendothelial migration in response to 154. Del Rey A, Besedovsky HO (2000) The
TNF-alpha. Rheumatology (Oxford) 42 cytokine-HPA axis circuit contributes to pre-
(8):929–934 vent or moderate autoimmune processes. Z
144. Banks WA, Kastin AJ, Gutierrez EG (1994) Rheumatol 59(Suppl 2):II/31–II/35
Penetration of interleukin-6 across the 155. Lorton D, Lubahn C, Bellinger DL (2003)
murine blood-brain barrier. Neurosci Lett Potential use of drugs that target neural-
179(1–2):53–56 immune pathways in the treatment of rheu-
145. Banks WA, Kastin AJ, Gutierrez EG (1993) matoid arthritis and other autoimmune dis-
Interleukin-1 alpha in blood has direct access eases. Curr Drug Targets Inflamm Allergy 2
to cortical brain cells. Neurosci Lett 163 (1):1–30
(1):41–44 156. Spangelo BL, Judd AM, Call GB, Zumwalt J,
146. Gutierrez EG, Banks WA, Kastin AJ (1993) Gorospe WC (1995) Role of the cytokines in
Murine tumor necrosis factor alpha is trans- the hypothalamic-pituitary-adrenal and
ported from blood to brain in the mouse. J gonadal axes. Neuroimmunomodulation 2
Neuroimmunol 47(2):169–176 (5):299–312
147. Tsai CY, Wu TH, Tsai ST, Chen KH, 157. Spangelo BL, Judd AM, Isakson PC,
Thajeb P, Lin WM, Yu HS, Yu CL (1994) MacLeod RM (1989) Interleukin-6 stimu-
Cerebrospinal fluid interleukin-6, prostaglan- lates anterior pituitary hormone release
din E2 and autoantibodies in patients with in vitro. Endocrinology 125(1):575–577
neuropsychiatric systemic lupus erythemato- 158. Shanks N, Moore PM, Perks P, Lightman SL
sus and central nervous system infections. (1999) Alterations in hypothalamic-pituitary-
Scand J Rheumatol 23(2):57–63 adrenal function correlated with the onset of
148. Svenungsson E, Andersson M, Brundin L, murine SLE in MRL +/+ and lpr/lpr mice.
van Vollenhoven R, Khademi M, Brain Behav Immun 13(4):348–360
Tarkowski A, Greitz D, Dahlstrom M, 159. Sakic B, Laflamme N, Crnic LS,
Lundberg I, Klareskog L, Olsson T (2001) Szechtman H, Denburg JA, Rivest S (1999)
Increased levels of proinflammatory cytokines Reduced corticotropin-releasing factor and
284 Boris Šakić
CNS lupus. Clin Exp Neuroimmunol 3 disease in murine lupus. Clin Immun
(3):116–128 (Orlando, Fla) 185:100–108. https://doi.
180. Kapadia M, Zhao H, Ma D, Sakic B (2017) org/10.1016/j.clim.2016.08.019
Sustained immunosuppression alters olfac- 183. Kim A, Feng P, Ohkuri T, Sauers D, Cohn ZJ,
tory function in the MRL model of CNS Chai J, Nelson T, Bachmanov AA, Huang L,
lupus. J Neuroimmune Pharmacol 12 Wang H (2012) Defects in the peripheral
(3):555–564. https://doi.org/10.1007/ taste structure and function in the MRL/lpr
s11481-017-9745-6 mouse model of autoimmune disease. PLoS
181. Mahajan SD, Tutino VM, Redae Y, Meng H, One 7(4):e35588
Siddiqui A, Woodruff TM, Jarvis JN, 184. Kapadia M, Stanojcic M, Earls AM,
Hennon T, Schwartz S, Quigg RJ, Alexander Pulapaka S, Lee J, Sakic B (2012) Altered
JJ (2016) C5a induces caspase-dependent olfactory function in the MRL model of
apoptosis in brain vascular endothelial cells CNS lupus. Behav Brain Res 234(2):303–311
in experimental lupus. Immunology 148 185. Stanojcic M, Burstyn-Cohen T, Nashi N,
(4):407–419. https://doi.org/10.1111/ Lemke G, Sakic B (2009) Disturbed distribu-
imm.12619 tion of proliferative brain cells during lupus-
182. Chalmers SA, Wen J, Shum J, Doerner J, like disease. Brain Behav Immun 23
Herlitz L, Putterman C (2017) CSF-1R inhi- (7):1003–1013
bition attenuates renal and neuropsychiatric
Chapter 15
Abstract
Positron-emission tomography (PET) imaging is a valuable research tool that enables in vivo quantification
of molecular targets in the brain or of a physiologic process. PET imaging can be combined with various
experimental and clinical model systems that are commonly used in psychoneuroimmunology research. As
PET imaging can be used in animals and humans, promising results can therefore often be translated from
an animal model to human disease.
1 Introduction
1.1 The Basics Nuclear imaging refers to in vivo imaging modalities that measure
of Positron-Emission radioactivity in a tissue or an organ after injection of a radioactive
Tomography pharmaceutical (radiopharmaceutical) that binds specifically to a
molecular target in that tissue or that is incorporated in a physio-
logic process. The radioactivity measured in the tissue being stud-
ied can be used to quantify the density of a target or the activity of a
physiologic process. Radiopharmaceuticals are interchangeably
called radioligands or radiotracers; the latter refers to the very
small mass dose (usually less than 10 μg) of radiopharmaceutical
administered in PET studies. Such “trace” amounts rarely have any
measurable pharmacologic effects because the radiotracer only
binds to 1–5% of the available target molecules. Because of this,
the regulatory requirements for first-in-human studies of radio-
tracers are less strict than for first-in-human studies of new com-
pounds administered at pharmacological doses. For instance, a
single-dose toxicity study in a single species is sufficient for regu-
latory approval to administer a new radiotracer in humans. Radio-
tracers are designed to measure a certain physiologic process (e.g.,
glucose metabolism, oxygen uptake, blood flow) or to bind to a
specific extra- or intracellular molecular target (e.g., a receptor or
transporter, or a molecule that is a marker of a specific cell type); the
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_15, © Springer Science+Business Media, LLC, part of Springer Nature 2018
287
288 Jonas Hannestad
Table 1
Definitions of commonly used names and PET imaging
Radiotracer A pharmaceutical or similar molecule in which a radionuclide has been substituted for an
original atom
Synonyms: Radioligand; radiopharmaceutical
Radionuclide A nuclide that decays and emits radiation
Synonym: Radioisotope
Decay Emission of radiation due to an unstable configuration of the nucleus
Synonym: Transition
Activity The amount of disintegrations per minute of a specific radionuclide
Physical half- The time it takes for the activity to be halved
life
Biological half- The time it takes for the body to clear half of the radiotracer molecules
life
Effective half- The combination of physical and biological half-life
life
Attenuation The process of a gamma photon interacting with tissue and not reaching the detectors
Scatter The process of a photon interacting with tissue and giving rise to false counts
Specific Binding to the target molecule
binding
Binding Binding to other molecules
potential
Target The density of target molecules that are available to bind the radiotracer
availability
Input function The changes in radioactivity in the blood
Table 2
Two commonly used radionuclides in PET imaging
Half-life
Radionuclide (min) Advantages Disadvantages
F-18 110 Can be synthesized elsewhere and Can only be used once in a day
transported to a PET facility
C-11 ~20 Allows for multiple scans in 1 day Requires a cyclotron and a
radiochemical lab on-site for
synthesis
1.4 PET Detection PET systems detect annihilation photons using a solid scintillation
Systems detector, which is made of a crystal material with which the photon
interacts to produce visible-light photons (scintillations). The most
common solid scintillation crystals in PET imaging are bismuth
germanate (BGO), lutetium oxyorthosilicate with cerium (LSO),
and yttrium oxyorthosilicate with cerium (YSO). These crystals
vary in some important properties including (1) stopping power
(the distance the annihilation photon travels in the crystal before it
interacts with an atom and deposits its energy); (2) scintillation
decay time (the time, in nanoseconds, it takes from when the
annihilation photon excites an atom in the detector crystal until
this atom emits a visible-light photon); (3) the light output (how
much light is emitted when an annihilation photon excites a crystal
atom); and (4) the energy resolution (the fraction of the 511 keV
energy that is “converted” to visible light). Also, these crystals
should have a high ratio of photoelectric to Compton interactions
(the opposite of what a tissue has) because photoelectric interac-
tions will deposit the energy in the crystal, whereas Compton
scattering results in photons entering adjacent detectors, contribut-
ing to misreads. The visible-light photons that are produced in the
detector crystals are in turn detected by a photomultiplier tube,
PET in Psychoneuroimmunology 293
Fig. 2 PET detector pairs detect photon pairs that emerge from a positron
annihilation. Source: Atlas of Cardiac Imaging (Volume 1, Chapter 10,
Published January 23, 2002. Authors: Brunken R, Wong CO, Chen E, Go R,
Lee R, Braunwald E). Copyright 1998 by Current Medicine. Reproduced with kind
permission from Springer Science+Business Media B.V
1.5 Data Correction As mentioned above, several corrections must be applied to ensure
that each voxel value represents (as close as possible) the real tissue
radioactivity concentration. Because each detector pair may have
different detection efficiencies, a calibration process called normal-
ization is used. A phantom (a cylinder that emits positrons in a
homogeneous manner) is placed in the PET camera and any differ-
ences in measured radioactivity will be used to calibrate (normalize)
the detector pairs before the PET camera is used for research
subjects. As described above, annihilation photons undergo atten-
uation when traversing a tissue. A transmission scan, in which an
external radioactive beam goes through the organ to be imaged
(e.g., the head), is used to measure the degree to which different
parts of the organ attenuate radioactivity. This is often done before
the emission scan (the scan which measures radioactivity emitted
from the radiotracer in the organ). If no direct measurement of
attenuation is performed (i.e., a transmission scan), attenuation
correction can be done using a mathematical model. Several
PET in Psychoneuroimmunology 295
1.7 Specific, To disentangle how much of total binding is due to specific binding
Nonspecific, and Non- to the target and how much is due to nonspecific binding, one can
displaceable Binding use several approaches. The most robust approach is to do a block-
ing study, which is the administration of a high dose of a “cold”
(nonradioactive) ligand that binds to and blocks access to the same
binding site as the radiotracer (this ligand can be the same molecule
as the radiotracer without the positron-emitting isotope or a differ-
ent molecule). Blocking reduces radioactivity because the binding
sites are occupied and the radiotracer cannot bind to as many sites.
Any residual radioactivity (non-displaceable binding) is due to non-
specific binding. The ratio of specific to nonspecific binding is
important because that is the radiotracer’s signal-to-noise ratio.
296 Jonas Hannestad
2 Basics of Psychoneuroimmunology
3 PET Modalities
3.1 Functional PET Functional imaging refers to a brain imaging modality that is used
Imaging: FDG to obtain an estimate of cellular (neuronal) activity. If a disease state
or an experimental intervention is associated with changes in neu-
ronal activity, functional imaging can measure this to identify which
brain regions are involved. The measures are obviously proxies for
cellular activity, and the assumption is that these proxy measures
correlate strongly with changes in the activity of neurons. An
important caveat is that functional imaging cannot distinguish
between neuronal activity and that of other cells in the brain (e.g.,
astrocytes). One commonly used modality of functional imaging is
functional MR imaging (fMRI), in which MR is used to measure
changes in oxygenated blood in the brain. The assumption is that
an increase in regional blood flow (which is under tight physiologic
control) is indicative of increased metabolic demand and therefore
increased cellular activity in a brain region. Functional imaging can
PET in Psychoneuroimmunology 299
3.2 Molecular PET Molecular or receptor imaging refers to the use of PET imaging to
Imaging: TSPO measure the availability of specific receptors or other molecules of
interest (e.g., neurotransmitter transporters, cell-specific markers).
Radiotracers are designed to bind to a specific molecule so that its
distribution and density can be measured in vivo. The development
of new radiotracers is a laborious process, the description of which
is beyond the scope of this chapter. Although some radiotracers are
commercially available, the majority of radiotracers that have been
developed at PET facilities around the world require on-site
synthesis.
There is considerable interest in imaging neuroinflammation
in vivo. Most PET imaging of neuroinflammation uses radiotracers
that bind to the translocator protein (TSPO; see below); however,
there have been several promising new developments in this field
that are worth summarizing.
1. PET tracers that bind to the cannabinoid receptor type
2 (CB2) have been developed as markers of neuroinflamma-
tion. Examples include 11C-A836339 [37, 38], and 11C-RS-
PET in Psychoneuroimmunology 301
4 Summary
The few examples given in this chapter are only a small subset of the
types of experiments that can be designed to incorporate PET
imaging as a research tool. The two main limitations of using
PET imaging in psychoneuroimmunology research are cost and
tracer/PET center availability. If those two limitations can be
addressed, there are a variety of PET tracers that can be used to
address myriad questions in this field. For instance, PET can be
used to measure changes in striatal dopamine levels, and this
approach was recently combined with the human endotoxin admin-
istration model, showing that endotoxin increased the amount of
synaptic dopamine [69].
304 Jonas Hannestad
5 Notes
References
1. Morris ED, Constantinescu CC, Sullivan JM, 4. Iwashyna TJ, Ely EW, Smith DM, Langa KM
Normandin MD, Christopher LA (2010) (2010) Long-term cognitive impairment and
Noninvasive visualization of human dopamine functional disability among survivors of severe
dynamics from PET images. Neuroimage 51 sepsis. JAMA 304:1787–1794
(1):135–144 5. DellaGioia N, Hannestad J (2010) A critical
2. Innis RB, Cunningham VJ, Delforge J, Fujita review of human endotoxin administration as
M, Gjedde A, Gunn RN et al (2007) Consen- an experimental paradigm of depression. Neu-
sus nomenclature for in vivo imaging of rever- rosci Biobehav Rev 34:130–143
sibly binding radioligands. J Cereb Blood Flow 6. Raison CL, Capuron L, Miller AH (2006)
Metab 27:1533–1539 Cytokines sing the blues: inflammation and
3. Semmler A, Hermann S, Mormann F, the pathogenesis of depression. Trends Immu-
Weberpals M, Paxian SA, Okulla T et al nol 27:24–31
(2008) Sepsis causes neuroinflammation and 7. Bucks RS, Gidron Y, Harris P, Teeling J,
concomitant decrease of cerebral metabolism. Wesnes KA, Perry VH (2008) Selective effects
J Neuroinflammation 5:38 of upper respiratory tract infection on
PET in Psychoneuroimmunology 305
cognition, mood and emotion processing: a 20. Qin L, Wu X, Block ML, Liu Y, Breese GR,
prospective study. Brain Behav Immun Hong JS et al (2007) Systemic LPS causes
22:399–407 chronic neuroinflammation and progressive
8. Janicki-Deverts D, Cohen S, Doyle WJ, Turner neurodegeneration. Glia 55:453–462
RB, Treanor JJ (2007) Infection-induced proin- 21. Moreau M, Andre C, O’Connor JC, Dumich
flammatory cytokines are associated with SA, Woods JA, Kelley KW et al (2008) Inocu-
decreases in positive affect, but not increases in lation of Bacillus Calmette-Guerin to mice
negative affect. Brain Behav Immun 21:301–307 induces an acute episode of sickness behavior
9. Hermann DM, Mullington J, Hinze-Selch D, followed by chronic depressive-like behavior.
Schreiber W, Galanos C, Pollmacher T (1998) Brain Behav Immun 22(7):1087–1095
Endotoxin-induced changes in sleep and sleep- 22. Hannestad J, DellaGioia N, Ortiz N, Pittman
iness during the day. Psychoneuroendocrinol- B, Bhagwagar Z (2011) Citalopram reduces
ogy 23:427–437 endotoxin-induced fatigue. Brain Behav
10. Reichenberg A, Yirmiya R, Schuld A, Kraus T, Immun 25:256–259
Haack M, Morag A et al (2001) Cytokine- 23. Gosselin D, Rivest S (2008) MyD88 signaling
associated emotional and cognitive distur- in brain endothelial cells is essential for the
bances in humans. Arch Gen Psychiatry neuronal activity and glucocorticoid release
58:445–452 during systemic inflammation. Mol Psychiatry
11. Wright CE, Strike PC, Brydon L, Steptoe A 13:480–497
(2005) Acute inflammation and negative 24. Ching S, Zhang H, Belevych N, He L, Lai W,
mood: mediation by cytokine activation. Brain Pu XA et al (2007) Endothelial-specific knock-
Behav Immun 19:345–350 down of interleukin-1 (IL-1) type 1 receptor
12. Eisenberger NI, Inagaki TK, Rameson LT, differentially alters CNS responses to IL-1
Mashal NM, Irwin MR (2009) An fMRI depending on its route of administration. J
study of cytokine-induced depressed mood Neurosci 27:10476–10486
and social pain: the role of sex differences. 25. Dauphinee SM, Karsan A (2006) Lipopolysac-
Neuroimage 47(3):881–890 charide signaling in endothelial cells. Lab
13. Capuron L, Fornwalt FB, Knight BT, Harvey Invest 86:9–22
PD, Ninan PT, Miller AH (2009) Does cyto- 26. Magder S, Neculcea J, Neculcea V, Sladek R
kine-induced depression differ from idiopathic (2006) Lipopolysaccharide and TNF-alpha
major depression in medically healthy indivi- produce very similar changes in gene expres-
duals? J Affect Disord 119(1–3):181–185 sion in human endothelial cells. J Vasc Res
14. De La Garza R 2nd (2005) Endotoxin- or pro- 43:447–461
inflammatory cytokine-induced sickness behav- 27. Kapas L, Hansen MK, Chang HY, Krueger JM
ior as an animal model of depression: focus on (1998) Vagotomy attenuates but does not pre-
anhedonia. Neurosci Biobehav Rev vent the somnogenic and febrile effects of lipo-
29:761–770 polysaccharide in rats. Am J Physiol 274:
15. Dunn AJ, Swiergiel AH, de Beaurepaire R R406–R411
(2005) Cytokines as mediators of depression: 28. Opp MR, Toth LA (1998) Somnogenic and
what can we learn from animal studies? Neu- pyrogenic effects of interleukin-1beta and lipo-
rosci Biobehav Rev 29:891–909 polysaccharide in intact and vagotomized rats.
16. Pecchi E, Dallaporta M, Jean A, Thirion S, Life Sci 62:923–936
Troadec JD (2009) Prostaglandins and sickness 29. Konsman JP, Luheshi GN, Bluthe RM, Dant-
behavior: old story, new insights. Physiol Behav zer R (2000) The vagus nerve mediates beha-
97(3-4):279–292 vioural depression, but not fever, in response to
17. Anisman H (2009) Cascading effects of stres- peripheral immune signals; a functional anato-
sors and inflammatory immune system activa- mical analysis. Eur J Neurosci 12:4434–4446
tion: implications for major depressive 30. Reyes TM, Coe CL (1996) Interleukin-1 beta
disorder. J Psychiatry Neurosci 34:4–20 differentially affects interleukin-6 and soluble
18. Larson SJ, Dunn AJ (2001) Behavioral effects interleukin-6 receptor in the blood and central
of cytokines. Brain Behav Immun 15:371–387 nervous system of the monkey. J Neuroimmu-
19. Frenois F, Moreau M, O’Connor J, Lawson M, nol 66:135–141
Micon C, Lestage J et al (2007) Lipopolysacchar- 31. Raison CL, Borisov AS, Majer M, Drake DF,
ide induces delayed FosB/DeltaFosB immunos- Pagnoni G, Woolwine BJ et al (2009) Activa-
taining within the mouse extended amygdala, tion of central nervous system inflammatory
hippocampus and hypothalamus, that parallel pathways by interferon-alpha: relationship to
the expression of depressive-like behavior. Psy- monoamines and depression. Biol Psychiatry
choneuroendocrinology 32:516–531 65(4):296–303
306 Jonas Hannestad
32. Bierhaus A, Wolf J, Andrassy M, Rohleder N, 43. Pichika R, Taha A, Gao F, Kotta K, Cheon Y,
Humpert PM, Petrov D et al (2003) A mecha- Chang L, Kiesewetter D, Rapoport S, Eckel-
nism converting psychosocial stress into mono- man W (2012) The synthesis and in vivo phar-
nuclear cell activation. Proc Natl Acad Sci U S macokinetics of fluorinated arachidonic acid:
A 100:1920–1925 implications for imaging neuroinflammation. J
33. Pavlov VA, Tracey KJ (2005) The cholinergic Nucl Med 53(9):1383–1391
anti-inflammatory pathway. Brain Behav 44. Di Virgilio F (2015) P2X receptors and inflam-
Immun 19:493–499 mation. Curr Med Chem. 22(7):866–877
34. Vida G, Pena G, Deitch EA, Ulloa L (2011) 45. Sáez-Orellana F, Godoy P, Silva-Grecchi T,
alpha7-cholinergic receptor mediates vagal Barra K, Fuentealba J (2015) Modulation of
induction of splenic norepinephrine. J Immu- the neuronal network activity by P2X receptors
nol 186:4340–4346 and their involvement in neurological disor-
35. Wienhard K (2002) Measurement of glucose ders. Pharmacol Res. 101:109–115
consumption using [(18)F]fluorodeoxyglu- 46. Gao M, Wang M, Green M, Hutchins G,
cose. Methods 27:218–225 Zheng Q (2015) Synthesis of [(11)C]
36. Capuron L, Pagnoni G, Demetrashvili MF, Law- GSK1482160 as a new PET agent for targeting
son DH, Fornwalt FB, Woolwine B et al (2007) P2X(7) receptor. Bioorg Med Chem Lett 25
Basal ganglia hypermetabolism and symptoms of (9):1965–1970
fatigue during Interferon-alpha therapy. Neu- 47. Janssen B, Vugts D, Funke U, Spaans A, Schuit
ropsychopharmacology 32:2384–2392 R, Kooijman E, Rongen M, Perk L, Lam-
37. Horti A, Gao Y, Ravert H, Finley P, Valentine mertsma A, Windhorst A (2014) Synthesis
H, Wong D, Endres C, Savonenko A, Dannals and initial preclinical evaluation of the P2X7
R (2010) Synthesis and Biodistribution of receptor antagonist [11C]A-740003 as a
[11C]A-836339, a New Potential Radioligand novel tracer of neuroinflammation. J Label
for PET Imaging of Cannabinoid Type Compd Radiopharm 57(8):509–516
2 Receptors (CB2). Bioorg Med Chem 18 48. Antunes I, Doorduin J, Haisma H, Elsinga P,
(14):5202–5207 van Waarde A, Willemsen A, Dierckx R, de
38. Savonenko A, Melnikova T, Wang Y, Ravert H, Vries E (2012) 18F-FEAnGA for PET of b-
Gao Y, Koppe I, Lee D, Pletnikova O, Cho E, Glucuronidase Activity in Neuro- inflamma-
Sayyida N, Hiatt A, Troncoso J, Davies P, Dan- tion. J Nucl Med. 53(3):451–458
nals R, Pomper M, Horti A (2015) Cannabi- 49. Wang X, Tan M, Yu J, Tan L (2014) Matrix
noid CB2 receptors in a mouse model of Ab metalloproteinases and their multiple roles in
amyloidosis: immunohistochemical analysis alzheimer’s disease. Biomed Res Int
and suitability as a PET biomarker of neuro- 2014:908636
inflammation. PLoS One 10(6):e0129618 50. Yang Y, Rosenberg G (2015) Matrix metallo-
39. Slavik R, Herde A, Bieri D, Weber M, Schibli R, proteinases as therapeutic targets for stroke.
Kr€amer S, Ametamey S, Mu L (2015) Synthe- Brain Res 1623:30–38
sis, radiolabeling and evaluation of novel 4- 51. Wagner S, Breyholz H, Höltke C, Faust A,
oxo-quinoline derivatives as PET tracers for Schober O, Sch€afers M, Kopka K (2009) A
imaging cannabinoid type 2 receptor. Eur J New 18F-Labelled Derivative of the MMP
Med Chem 92:554–564 Inhibitor CGS 27023A for PET: Radiosynth-
40. Tietz O, Marshall A, Wuest M, Wang M, Wuest esis and Initial Small-Animal PET Studies. Appl
F (2013) Radiotracers for molecular imaging of Radiat Isot 67(4):606–610
cyclooxygenase-2 (COX-2) enzyme. Curr Med 52. Matusiak N, van Waarde A, Bischoff R, Olten-
Chem 30(35):4350–4369 freiter R, van de Wiele C, Dierckx R, Elsinga P
41. Wang C, Placzek M, Van de Bittner G, Schroe- (2013) Probes for non-invasive matrix metal-
der F, Hooker J (2016) A novel radiotracer for loproteinase-targeted imaging with PET and
imaging monoacylglycerol lipase in the brain SPECT. Curr Pharm Des 19(25):4647–4672
using positron emission tomography. ACS 53. Tambuyzer BR, Ponsaerts P, Nouwen EJ
Chem Neurosci. 7(4):484–489 (2009) Microglia: gatekeepers of central ner-
42. Esposito G, Giovacchini G, Liow J, Bhattachar- vous system immunology. J Leukoc Biol
jee A, Greenstein D, Schapiro M, Hallett M, 85:352–370
Herscovitch P, Eckelman W, Carson R, Rapo- 54. Block ML, Zecca L, Hong JS (2007) Micro-
port S (2008) Imaging neuroinflammation in glia-mediated neurotoxicity: uncovering the
Alzheimer’s disease with radiolabeled arachido- molecular mechanisms. Nat Rev Neurosci
nic acid and PET. J Nucl Med 49 8:57–69
(9):1414–1421 55. Maeda J, Higuchi M, Inaji M, Ji B, Haneda E,
Okauchi T et al (2007) Phase-dependent roles
PET in Psychoneuroimmunology 307
of reactive microglia and astrocytes in nervous 64. Owen D, Yeo A, Gunn R, Song K, Wadsworth
system injury as delineated by imaging of G, Lewis A, Rhodes C, Pulford D, Bennacef I,
peripheral benzodiazepine receptor. Brain Res Parker C, StJean P, Cardon L, Mooser V, Mat-
1157:100–111 thews P, Rabiner E, Rubio J (2012) An 18-kDa
56. Li L, Lu J, Tay SS, Moochhala SM, He BP translocator protein (TSPO) polymorphism
(2007) The function of microglia, either neu- explains differences in binding affinity of the
roprotection or neurotoxicity, is determined by PET Radioligand PBR28. J Cereb Blood
the equilibrium among factors released from Flow Metab 32(1):1–5
activated microglia in vitro. Brain Res 65. Brown AK, Fujita M, Fujimura Y, Liow JS,
1159:8–17 Stabin M, Ryu YH et al (2007) Radiation
57. Hanisch UK, Kettenmann H (2007) Micro- dosimetry and biodistribution in monkey
glia: active sensor and versatile effector cells in and man of 11C-PBR28: a PET radioligand
the normal and pathologic brain. Nat Neurosci to image inflammation. J Nucl Med
10:1387–1394 48:2072–2079
58. Pinteaux E, Rothwell NJ, Boutin H (2006) 66. Imaizumi M, Briard E, Zoghbi SS, Gourley JP,
Neuroprotective actions of endogenous inter- Hong J, Fujimura Y et al (2008) Brain and
leukin-1 receptor antagonist (IL-1ra) are whole-body imaging in nonhuman primates
mediated by glia. Glia 53:551–556 of [(11)C]PBR28, a promising PET radioli-
59. Cosenza-Nashat M, Zhao ML, Suh HS, Mor- gand for peripheral benzodiazepine receptors.
gan J, Natividad R, Morgello S et al (2009) Neuroimage 39:1289–1298
Expression of the translocator protein of 18 67. Redl H, Bahrami S (2005) Large animal mod-
kDa by microglia, macrophages and astrocytes els: baboons for trauma, shock, and sepsis stud-
based on immunohistochemical localization in ies. Shock 24(Suppl 1):88–93
abnormal human brain. Neuropathol Appl 68. Hillmer AT, Sandiego CM, Hannestad J,
Neurobiol 35:306–328 Angarita GA, Kumar A, McGovern EM,
60. Batarseh A, Papadopoulos V (2010) Regula- Huang Y, O’Connor KC, Carson RE, O’Mal-
tion of translocator protein 18 kDa (TSPO) ley SS, Cosgrove KP (2017) In vivo imaging of
expression in health and disease states. Mol translocator protein, a marker of activated
Cell Endocrinol 327:1–12 microglia, in alcohol dependence. Mol Psychi-
61. Venneti S, Lopresti BJ, Wiley CA (2006) The atry 22(12):1759–1766
peripheral benzodiazepine receptor (Translo- 69. Petrulli JR, Kalish B, Nabulsi NB, Huang Y,
cator protein 18kDa) in microglia: from Hannestad J, Morris ED (2017) Systemic
pathology to imaging. Prog Neurobiol inflammation enhances stimulant-induced
80:308–322 striatal dopamine elevation. Transl Psychiatry.
62. Rupprecht R, Papadopoulos V, Rammes G, 7(3):e1076
Baghai TC, Fan J, Akula N et al (2010) Trans- 70. Fujimura Y, Kimura Y, Simeon FG, Dickstein
locator protein (18 kDa) (TSPO) as a thera- LP, Pike VW, Innis RB et al (2010) Biodistri-
peutic target for neurological and psychiatric bution and radiation dosimetry in humans of a
disorders. Nat Rev Drug Discov 9:971–988 new PET ligand, (18)F-PBR06, to image
63. Vivash L, O’Brien T (2016) Imaging micro- translocator protein (18 kDa). J Nucl Med
glial activation with TSPO PET: lighting up 51:145–149
neurologic diseases? J Nucl Med. 57
(2):165–168
Chapter 16
Abstract
This chapter explores the reasoning behind using the vaccination model to examine the influence of
psychosocial factors on immunity. It then briefly discusses the mechanics of the vaccination response and
the protocols used in psychoneuroimmunology vaccine research, before giving examples from the research
literature of the studies examining relationships such as the association between stress and vaccination
response. It also explores the ways the vaccination model can be used to answer key questions in psycho-
neuroimmunology, such as the following: “Does it matter when stressful life events occur relative to when
the vaccine is received?” “What are the effects of prior exposure to the antigen?” “Do other psychosocial
factors influence vaccine response besides stress?” Finally, it briefly considers the mechanisms underlying
psychosocial factors and vaccination response associations and the future research needed to understand
these better, and indeed to use current and future knowledge to improve and enhance vaccine responses in
key at-risk populations.
1.1 Alternative There are many methods for examining the effects of psychological
Approaches: factors on immunity. Early work concentrated on the influence of
Enumerative Measures psychosocial stress on enumerative measures of immunity. For
example, individuals exposed to chronic stress showed reduced
numbers of certain immune cells including reduced numbers of
B-lymphocytes [1, 2], helper T-lymphocytes [1, 3, 4], cytotoxic
T-lymphocytes [1, 5], natural killer (NK) cells [1, 5], and lowered
concentrations of secretory immunoglobulin A in saliva [6–10],
compared to matched controls. However, it is difficult to determine
the clinical significance of such enumerative changes, given that
they lie within the normal range for healthy participants [11] and
may simply reflect cell migration and recirculation rather than
increased production or better function [11]. Additionally, cell
number changes could be a consequence of shifts in plasma volume
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_16, © Springer Science+Business Media, LLC, part of Springer Nature 2018
309
310 Anna C. Whittaker
1.2 In Vitro Measures In vitro measures of immune function, such as cell proliferation to
stimulation with an antigen (foreign material, e.g., bacteria), or cell
cytotoxicity (killing ability), have been argued to provide a better
indication of the functional capacity of the immune system
[11]. These measures have been demonstrated to be susceptible
to impairment by chronic stress in many studies, e.g., [12–15]. For
example, recently unemployed individuals showed poorer lympho-
cyte proliferation to antigen than those in employment [16]. Fur-
ther, compared to non-bereaved controls, individuals who have
suffered bereavement showed lower neutrophil superoxide produc-
tion, one of their key cytotoxic capacities through which they
eradicate bacteria such as pneumonia [17]. Nevertheless, the
isolated testing of any particular network of immune cells provides
only limited information about the overall status of what is a highly
integrated and complex system [11], and an imperfect understand-
ing of the relationship between psychosocial factors and vulnerabil-
ity to disease [18].
2.1 Benefits of the A clinically relevant model which examines the impact of psychoso-
Vaccination Model cial factors on the integrated response of the immune system to a
challenge would avoid these disadvantages. The antibody response
to vaccination provides us with such a model. Vaccines act as real
immune system challenges, although they are altered in such a way
so as not to induce disease either by being inactivated or killed, or
only a component of the actual pathogen, so are really “imitation
infections.” Therefore, by measuring the antibody levels in
response to vaccination we can assess directly how well the immune
system responds to infectious challenge. It is also clinically relevant
in that antibody levels or titers are directly related to susceptibility
and resistance to infectious disease.
2.2 The Vaccination The vaccination response involves the coordination of a wide vari-
Response ety of immune cells. Antigen is initially recognized and presented
by professional antigen-presenting cells, such as dendritic cells.
Vaccine Research in PNI 311
stress, may alter both the quantity and quality of antibody present
at different times after immunization, meaning that individuals
suffering from higher levels of stress are more at risk of infectious
disease.
3.1 Stress The most common psychosocial factor examined in the context of
Questionnaires the vaccination response is stress. This is usually assessed via life
event checklists or perceived stress measures. Life event checklists
consist of a list of major and minor life events, e.g., bereavement
and moving house, and usually require participants to indicate
which have occurred during the past month or year [24]. Some
also ask participants to indicate how stressful each event was on a
rating scale. Life events have been shown to predict a variety of
important physical health outcomes, including infectious disease
[25], and mortality, particularly in the context of little emotional
support [26]. In contrast, perceived stress scales measure indivi-
duals’ feelings about how stressful their lives are rather than the
direct occurrence of events [27]. Thus these measures are more
susceptible to subjective bias, and are better predictors of subjective
health outcomes, such as angina, rather than objective outcomes,
such as myocardial infarction [28].
3.2 Caregiver Control Another common way of assessing stress in the context of vaccina-
Models tion is to examine antibody responses among those with a key
chronic stressor versus a sociodemographically matched control
group, for example, older adults caregiving for a spouse with
dementia. The stress of caregiving has been shown to relate to
poor health and mortality [29], and can thus be considered an
important source of ongoing psychological stress. Other stress
studies compare matched controls to other groups subject to
chronic stressors such as bereavement, marital separation/divorce,
or unemployment.
3.3 Protocol for In order to fully test the impact of psychosocial factors on the
Stress and Vaccination response to vaccination, both pre-vaccination and post-vaccination
Studies blood samples are required for assessment of antibody levels. This is
due to the impact that prior vaccination or environmental exposure
to the infectious agent can have on pre-vaccination antibody levels,
and consequently post-vaccination levels. Without taking a
pre-vaccination baseline, it is difficult to state whether stress is
affecting the antibody response to a vaccination administered dur-
ing a research study or simply the maintenance of previous antibody
levels. For example, in 37 nursing-home residents, those who
reported higher levels of perceived stress had lower pre-vaccine
antibody titers to two influenza vaccine components
Vaccine Research in PNI 313
3.4 Key Stress and One of the most common vaccinations studied in the context of
Vaccination Findings stress and antibody response is the influenza vaccination, particu-
larly in undergraduate student and older caregiver samples. The
influenza vaccination is a commonly utilized vaccine and consists of
three components or strains, usually two A strains and one B strain,
which change each year depending on the key circulating varieties.
A meta-analysis of 13 studies of psychological stress and influenza
vaccination concluded that there is a significant negative relation-
ship between psychological stress and antibody titer following
influenza vaccination [33]. These studies included five caregivers
and eight assessing the impact of stressful life events or perceived
stress. The meta-analysis concluded that psychological stress, how-
ever measured, had a similar negative impact on influenza vaccine
response, but that antibody responses to A/H1N1 and B-influenza
types were more sensitive to the influence of stress [33]. However,
it is difficult at this stage to explain why antibodies against influenza
strains are differentially associated with stress. One possibility is that
strain novelty influences the associations observed [11, 34], with
more novel strains being more susceptible to stress effects.
The impact on certain A-strains and on B-strains is clearly
shown in several studies of students. For example, those reporting
higher stressful life event exposures and/or higher perceived stress
314 Anna C. Whittaker
controls [45]. This raises the issue of whether the poor antibody
response observed in older caregivers is, to an extent, a function of
an interaction between chronic stress exposure and
immunosenescence [46].
There is an alternative explanation for the discrepancy in out-
comes among the caregiver vaccination studies. Rather than
immune ageing, perhaps it is the intensity of the stress experienced
that determines whether caregiving becomes an issue for immunity
[45]. Dementia is a disease characterized by much more severe
cognitive and behavioral disturbances than multiple sclerosis
[47–50], and older spousal caregivers of dementia patients have
been found to report greater distress than younger multiple sclero-
sis caregivers [45]. Further, the results of two recent meta-analyses
indicate that caregivers of dementia patients generally experience
greater burden and report more symptoms of depression than those
caring for non-dementia, e.g., cancer, patients, [51, 52]. Thus, it
might be hypothesized that, irrespective of the caregiver’s age,
caring for someone with severe cognitive and behavioral problems
will compromise immunity.
We have been able to test this hypothesis recently using a
caregiving model in younger adults, young parents caring for chil-
dren with developmental disabilities. Dealing with severe cognitive
difficulties and behaviors that are problematic and distressing is the
main challenge of such caring [53–56]. In our own studies of
30 caregivers for a child with a developmental disability (mainly
Autism) versus matched controls, we have demonstrated that care-
givers report high levels of stress, anxiety, depression, child problem
behaviors, and low levels of social support. These caregivers also
exhibited a poorer antibody response to a pneumonia vaccination
than parents caring for typically developing children at both 1 and
6 months post-vaccination [57]. Of the psychological variables
considered, child problem behaviors mediated this effect. In addi-
tion, within the caregivers, parents reporting more child conduct
problems, a component of the child problem behavior measure,
mounted a poorer antibody response at 1 month than parents
reporting less conduct problems [57]. Similarly, these parents
mounted a poorer antibody response to the B/Malaysia strain of
an influenza vaccine at 1 and 6 months post-vaccination, which
again appeared to be mediated by differences in child problem
behaviors [58].
These recent findings in younger caregivers reinforce the
hypothesis that an ageing immune system is not a prerequisite for
a poor response to medical vaccination in caregivers. Nevertheless,
among our parental caregivers, older caregivers tended to have a
poorer antibody response to B/Malaysia at 1 month, suggesting
that we cannot dismiss the hypothesis that chronic stress and
immunosenescence may have synergistic effects [46].
316 Anna C. Whittaker
4.1 Timing of Stress This issue of the timing of stress assessment has been developed in
Measurement studies of various vaccinations including hepatitis B, which is useful
in this context, as the vaccination schedule consists of three inocu-
lations over a 6-month period. The largest of these studies exam-
ined the association between life event stress and final antibody titer
in students, vaccinated either in the past 12 months or at least
13 months previously [31]. Whereas life event exposure was not
related to antibody response in the recently vaccinated cohort,
participants in the earlier vaccinated cohort who reported higher
life events over the past year were over twice as likely to show an
inadequate antibody titer as those with lower life event exposure.
This finding suggests that the immunogenicity, the ability to induce
a strong vaccination response, of hepatitis B vaccination may initi-
ally override the influence of life event stress, although there was
also more power to detect effects in the earlier vaccinated cohort as
more participants exhibited inadequate antibody titers [31]. Never-
theless, this study provides some evidence that psychosocial stress in
the period following vaccination can have effects on the rate of
deterioration of antibody protection [18].
In a study where a low dose of hepatitis B vaccine was adminis-
tered, a higher stress index, comprising life event exposure and
psychological symptoms, measured at 2 months post-vaccination
(thus considering the period post-vaccination) was associated with
a poorer final 6-month antibody response, and the stress index at
6 months also tended to relate negatively to antibody response
[59]. However, as only the final antibody titer was measured, it is
difficult to determine whether, in this instance, stress predomi-
nantly influenced initial formation or maintenance of antibody
levels. Also, the inclusion of psychological symptoms in the com-
posite stress index makes it difficult to ascribe this finding to any
specific aspect of stress [18]. A similar study using the full-dosage
hepatitis B vaccination did not yield any significant stress effects
[60], although it is possible that this was due to the absence of a
2-month assessment of stress, which was the main predictor of
antibody response in the previous study by this group. In a study
measuring perceived stress and anxiety during the vaccination
period, i.e., post-vaccination, these were not associated with the
final antibody response to hepatitis B [40]. Further, life event stress
Vaccine Research in PNI 317
4.2 Primary and Vaccination with an antigen to which the participant has not been
Secondary Exposure to previously exposed induces a primary antibody response whereas
Vaccine Antigens vaccination against more common pathogens such as influenza
induce a secondary immune response. By examining the effect of
stress on both primary and secondary immune responses, we can
begin to determine which aspects of the immune response are most
susceptible to stress-induced modulation.
Hepatitis B vaccination has been used in this context due to the
vaccination schedule and the low likelihood of prior naturalistic
exposure to this pathogen. In an earlier study, individuals reporting
higher mean perceived stress and anxiety over the vaccination
period were less likely to have seroconverted (produced a protective
antibody level) by the time of the second inoculation [40]. How-
ever, an emotional disclosure intervention group did not differ
from controls in antibody levels at the time of the second inocula-
tion [64]. However, psychological stress levels were not measured,
making it difficult to interpret these data. More recently, we have
used hepatitis A as a primary antigen. Students who reported a
higher number and severity of life events had a poorer antibody
response to hepatitis A at the 18-week, but not 4-week, follow-up,
suggesting that stress can impact upon the maintenance of antibody
levels [65]. Early studies using the vaccination model used novel
nonpathogenic antigens to examine the antigen-specific antibody
response. Keyhole limpet hemocyanin (KLH), a protein, has been
318 Anna C. Whittaker
4.3 Thymus- A further advantage to the vaccination model is that there are
Dependent Versus different types of vaccination, which can be used to help elucidate
Thymus-Independent which cells involved in the vaccination response are influenced by
Vaccines psychological factors. Most vaccinations, which consist of inacti-
vated or dead viruses like influenza, induce a thymus-dependent
antibody response, as described above. A few vaccinations, how-
ever, protect against bacterial infections or toxins, like meningo-
coccal A or tetanus, respectively, which do not require T-cell help.
There are also conjugate vaccines, in which substances that elicit a
T-cell response are conjugated to a thymus-independent pathogen,
such as a protein, in order to boost the efficiency of the antibody
response against the thymus-independent pathogen. If psychologi-
cal factors are consistently associated with the response to thymus-
dependent and conjugate vaccinations but not with thymus-inde-
pendent response, this would imply that it is T cells that are partic-
ularly liable to psychological influence.
Indeed, there is evidence to suggest that stress may exert its
effects mainly on T cells; we showed that higher frequency and
intensity of stressful life events were associated with a poorer
response to influenza and meningococcal C (following previous
conjugate meningococcal C vaccination), but not to thymus-
independent meningococcal A [36]. Similarly, no association was
found between stress and antibody response to a thymus-indepen-
dent pneumonia vaccination in preschool children [68]. However,
as older caregivers have been reported to show poorer maintenance
of antibody levels over time following pneumonia vaccination than
controls [41], it is possible that other factors such as age and
severity of stress may interact to impair antibody-mediated immu-
nity more generally than just the T-cell response.
It should be noted that in the study of caregivers and pneumo-
nia vaccination, perceived stress did not differ between the caregiver
and controls, but there was a significant difference in social support.
This might suggest that thymus-dependent vaccinations are sus-
ceptible to the effects of stressful life events, but that thymus-
independent vaccinations are more vulnerable to other psychosocial
Vaccine Research in PNI 319
factors such as lower social support. There is some evidence for this
suggestion. In our own laboratory, we found that social support,
but not life event stress, was positively associated with the response
to a thymus-independent pneumococcal vaccine in young healthy
students [65, 69].
The comparison between thymus-dependent and -independent
vaccination responses suggests that both types of response are suscep-
tible to psychosocial influence, but that there are key variables which
influence whether an effect on vaccination response is observed.
These include the type of psychosocial factor studied (i.e., stress
versus social support), and the age of the population sampled.
4.4 Acute Versus Following on from the discussion above in Subheading 4.1 regard-
Chronic Stress ing when stress is measured, such that moderate or less severe stress
at the time of vaccination might actually have a beneficial effect, it
has been suggested in recent years that acute (minutes or hours)
stress may be immune enhancing when experienced close to the
immune challenge. Such immune enhancement by acute stress
would be an adaptive mechanism, and might be regarded as an
integral component of the fight or flight response, and circum-
stances that elicit such a response are likely to also involve exposure
to antigens and, therefore, a robust immune response would be
adaptive for survival [63]. Our laboratory examined the effect of
acute psychological stress on antibody response to vaccination in
humans. Participants completed a 45-min time pressured, socially
evaluated mental arithmetic task, or a resting control period, imme-
diately prior to influenza vaccination. An enhancement of the anti-
body response to one of the influenza viral strains was found in
women in the psychological stress group compared to control
[70]. Similarly, in men, the antibody response to a meningococcal
A vaccination was enhanced by acute psychological stress
[71]. That these effects emerged for only one gender or the other
in these studies might be explained by examining the antibody
responses for each gender. In each case, stress was associated with
the antibody response in those with the poorest increase in anti-
bodies in response to vaccination: women for the influenza A/Pan-
ama strain and men for the meningococcal A vaccine. This latter
study [71] also provides further evidence that both thymus-
dependent and -independent vaccinations are responsive to the
impact of stress, as discussed above. Further, although not a psy-
chological stressor per se, acute eccentric exercise (arm contrac-
tions) was also shown to enhance the antibody response to
influenza vaccination in women, and the cell-mediated IFNγ pro-
duction response to stimulation with the influenza vaccine in men
[72]. This and similar studies have also lent weight to the conten-
tion that effects of behavioral factors on vaccination responses are
most likely to be observed in groups with the poorest antibody
response, or to vaccine strains which are not very immunogenic,
i.e., they engender lower antibody titers [72, 73].
320 Anna C. Whittaker
4.5 Timing of As well as the timing and duration of stress measurements asso-
Vaccination ciated with the vaccine response, other behavioral factors have been
found to impact upon the antibody response. One such factor is
that of the time of day of vaccination. In our study of the effects of
psychological stress on vaccination responses in the 184 older
adults [38], we observed that the time of day of vaccine administra-
tion significantly influenced antibody titer [74]. Men responded
better in the morning than the afternoon; 41% of men showed a
twofold response when vaccinated in the morning versus 24% of
men vaccinated in the afternoon. This effect was independent of
current illnesses, medication, vaccination history, and our reported
findings of the effects of bereavement and marital quality. Women
tended to show the reverse pattern. We also observed the same
pattern in a study of younger adults’ antibody response to the
hepatitis A vaccination [74]. However, these studies were not
fully randomized, and there was little opportunity to examine the
biological mechanisms, such as cytokine and stress hormone levels.
Consequently, there was a clear and pressing need for a randomized
controlled trial of the impact of time of influenza vaccination on
antibody response and vaccine efficacy in older adults in a National
Health Service (NHS) setting. We were able to conduct such a
cluster-randomized trial within the National Health Service in
Birmingham, UK, and confirmed that a simple manipulation of
the time of vaccination can improve the immune response against
influenza in older adults. Two hundred and seventy-six participants
were randomized to have the annual influenza vaccination at one of
24 general practice surgeries in either the morning (9–11 a.m.) or
the afternoon (3–5 p.m.). This trial provided some evidence that
morning vaccination produced higher antibody levels, at least for
the H1N1 A-strain with a trend for the same effect for the B
influenza strain [75, 76]. Consequently, although this requires a
larger scale trial for confirmation, we believe that morning vaccina-
tion could be employed as an easy-to-adopt intervention within the
health services, at little or no added cost. The potential benefits
would be a decreased incidence of influenza infection and
influenza-related mortality in older adults, although a multicenter
very large trial following up on verified influenza incidence and
health outcomes would be essential to fully prove this.
5.1 Social Support The support of friends and loved ones is an important determinant
of immune health, and is relatively easily measured in vaccination
studies via validated questionnaires. Studies have assessed both
functional social support, a measure of the quality and availability
of social resources a person has, and structural social support, the
number of friends a person can call on, in the context of
Vaccine Research in PNI 321
7 Conclusion
References
induced natural killer cell activity in older 28. Macleod J, Davey Smith G, Heslop P et al
adults. Psychosom Med 58:264–272 (2002) Psychological stress and cardiovascular
14. Irwin M, Patterson T, Smith TL et al (1990) disease: empirical demonstration of bias in a
Reduction of immune function in life stress and prospective observational study of Scottish
depression. Biol Psychiatry 27:22–30 men. BMJ 324:1247–1251
15. Schleifer SJ, Keller SE, Camerino M et al 29. Schulz R, Beach SR (1999) Caregiving as a risk
(1983) Suppression of lymphocyte stimulation factor for mortality: the Caregiver Health
following bereavement. JAMA 250:374–377 Effects Study. JAMA 282:2215–2219
16. Arnetz BB, Wasserman J, Petrini B et al (1987) 30. Moynihan JA, Larson MR, Treanor J et al
Immune function in unemployed women. Psy- (2004) Psychosocial factors and the response
chosom Med 49:3–12 to influenza vaccination in older adults. Psy-
17. Khanfer R, Lord JM, Phillips AC (2011) Neu- chosom Med 66:950–953
trophil function and cortisol:DHEAS ratio in 31. Burns VE, Carroll D, Ring C et al (2002)
bereaved older adults. Brain Behav Immun 25 Stress, coping and hepatitis B antibody status.
(6):1182–1186. https://doi.org/10.1016/j. Psychosom Med 64:287–293
bbi.2011.03.008 32. Burns VE, Drayson M, Ring C et al (2002)
18. Burns VE, Carroll D, Ring C et al (2003) Perceived stress and psychological well-being
Antibody response to vaccination and psycho- are associated with antibody status following
social stress in humans: relationship and meningitis C conjugate vaccination. Psycho-
mechanisms. Vaccine 21:2523–2534 som Med 64:963–970
19. Allsup S, Haycox A, Regan M et al (2004) Is 33. Pedersen AF, Zachariae R, Bovbjerg DH
influenza vaccination cost effective for healthy (2009) Psychological stress and antibody
people between ages 65 and 74 years? A rando- response to influenza vaccination: a meta-
mised controlled trial. Vaccine 23:639–645 analysis. Brain Behav Immun 23:427–433.
20. Patriarca PA (1994) Editorial: A randomized https://doi.org/10.1016/j.bbi.2009.01.004
controlled trial of influenza vaccine in the 34. Gulati U, Keitel WA, Air GM (2007) Increased
elderly: Scientific scrutiny and ethical responsi- antibodies against unfolded viral antigens in
bility. JAMA 272:1700–1701 the elderly after influenza vaccination. Influ-
21. Goodwin K, Viboud C, Simonsen L (2006) enza Other Respi Viruses 1:147–156. https://
Antibody response to influenza vaccination in doi.org/10.1111/j.1750-2659.2007.00017.
the elderly: a quantitative review. Vaccine x
24:1159–1169. https://doi.org/10.1016/j. 35. Burns VE, Carroll D, Drayson M et al (2003)
vaccine.2005.08.105 Life events, perceived stress and antibody
22. Nichol KL, Margolis KL, Wuorenma J et al response to influenza vaccination in young
(1994) The efficacy and cost effectiveness of healthy adults. J Psychosom Res 55:569–572
vaccination against influenza among elderly 36. Phillips AC, Burns VE, Carroll D et al (2005)
persons living in the community. N Engl J The association between life events, social sup-
Med 331:778–784 port and antibody status following thymus-
23. Simonsen L, Reichert TA, Viboud C et al dependent and thymus-independent vaccina-
(2005) Impact of influenza vaccination on sea- tions in healthy young adults. Brain Behav
sonal mortality in the US elderly population. Immun 19:325–333
Arch Intern Med 165:265–272 37. Miller GE, Cohen S, Pressman S et al (2004)
24. Brown GW, Harris TO (1989) Social origins of Psychological stress and antibody response to
depression: a study of psychiatric disorder in influenza vaccination: when is the critical
women. Routledge, London period for stress, and how does it get inside
the body? Psychosom Med 66:215–223
25. Cohen S, Tyrell DAJ, Smith AP (1991) Psy-
chological stress and susceptibility to the com- 38. Phillips AC, Carroll D, Bums VE et al (2006)
mon cold. N Engl J Med 325:606–612 Bereavement and marriage are associated with
antibody response to influenza vaccination in
26. Rosengren A, Orth-Gomer K, Wedel H et al the elderly. Brain Behav Immun 20:279–289.
(1993) Stressful life events, social support, and https://doi.org/10.1016/j.bbi.2005.08.003
mortality in men born in 1933. BMJ
307:1102–1105 39. Talbot HK, Coleman LA, Zhu Y et al (2015)
Factors associated with maintenance of anti-
27. Cohen S, Kamarck T, Mermelstein R (1983) A body responses to influenza vaccine in older,
global measure of perceived stress. J Health Soc community-dwelling adults. BMC Infect Dis
Behav 24:385–396. https://doi.org/10. 15:195. https://doi.org/10.1186/s12879-
2307/2136404 015-0926-8
Vaccine Research in PNI 325
40. Glaser R, Kiecolt-Glaser JK, Bonneau RH et al 53. Floyd FJ, Gallagher EM (1997) Parental stress,
(1992) Stress-induced modulation of the care demands, and use of support services for
immune response to recombinant hepatitis B school-age children with disabilities and behav-
vaccine. Psychosom Med 54:22–29 ior problems. Fam Relat 46:359–371
41. Glaser R, Sheridan JF, Malarkey W et al (2000) 54. Hastings RP, Daley D, Burns C et al (2006)
Chronic stress modulates the immune response Maternal distress and expressed emotion:
to a pneumococcal pneumonia vaccine. Psy- cross-sectional and longitudinal relationships
chosom Med 62:804–807 with behavior problems of children with intel-
42. Vedhara K, Cox NK, Wilcock GK et al (1999) lectual disabilities. Am J Mental Retard
Chronic stress in elderly carers of dementia 111:48–61
patients and antibody response to influenza 55. Higgins DJ, Bailey SR, Pearce JC (2005) Fac-
vaccination. Lancet 353:627–631 tors associated with functioning style and cop-
43. Segerstrom SC, Schipper LJ, Greenberg RN ing strategies of families with a child with an
(2008) Caregiving, repetitive thought, and autism spectrum disorder. Autism 9:125–137
immune response to vaccination in older 56. Maes B, Broekman TG, Dosen A et al (2003)
adults. Brain Behav Immun 22:744–752. Caregiving burden of families looking after
https://doi.org/10.1016/j.bbi.2007.11.004 persons with intellectual disability and beha-
44. Wong SY, Wong CK, Chan FW et al (2013) vioural or psychiatric problems. J Intellect Dis-
Chronic psychosocial stress: does it modulate abil Res 47:447–455
immunity to the influenza vaccine in Hong 57. Gallagher S, Phillips AC, Drayson MT et al
Kong Chinese elderly caregivers? Age (Dordr) (2009) Parental caregivers of children with
35:1479–1493. https://doi.org/10.1007/ developmental disabilities mount a poor anti-
s11357-012-9449-z body response to pneumococcal vaccination.
45. Vedhara K, McDermott MP, Evans TG et al Brain Behav Immun 23:338–346. https://
(2002) Chronic stress in non-elderly care- doi.org/10.1016/j.bbi.2008.05.006
givers: psychological, endocrine and immune 58. Gallagher S, Phillips AC, Drayson MT et al
implications. J Psychosom Res 53:1153–1161 (2009) Caregiving for children with develop-
46. Graham JE, Christian LM, Kiecolt-Glaser JK mental disabilities is associated with a poor
(2006) Stress, age, and immune function: antibody response to influenza vaccination.
toward a lifespan approach. J Behav Med Psychosom Med 71:341–344. https://doi.
29:389–400 org/10.1097/PSY.0b013e31819d1910
47. Keegan BM, Noseworthy JH (2002) Multiple 59. Jabaaij L, Grosheide PM, Heijtink RA et al
sclerosis. Annu Rev Med 53:285–302. https:// (1993) Influence of perceived psychological
doi.org/10.1146/annurev.med.53.082901. stress and distress on antibody response to
10390953/1/285 low dose rDNA hepatitis B vaccine. J Psycho-
48. Poser CM, Paty DW, Scheinberg L et al (1983) som Res 37:361–369
New diagnostic criteria for multiple sclerosis: 60. Jabaaij L, Van Hattum J, Vingerhoets AJJM
guidelines for research protocols. Ann Neurol et al (1996) Modulation of immune response
13:227–231. https://doi.org/10.1002/ana. to rDNA hepatitis B vaccination by psycholog-
410130302 ical stress. J Psychosom Res 41:129–137
49. Gregory CA, Hodges JR (1996) Clinical fea- 61. Marsland AL, Cohen S, Rabin BS et al (2001)
tures of frontal lobe dementia in comparison to Associations between stress, trait negative
Alzheimer’s disease. J Neural Transm Suppl affect, acute immune reactivity, and antibody
47:103–123 response to hepatitis B injection in healthy
50. Neary D, Snowden JS, Gustafson L et al young adults. Health Psychol 20:4–11
(1998) Frontotemporal lobar degeneration: a 62. Petry J, Weems LB, Livingstone JNI (1991)
consensus on clinical diagnostic criteria. Neu- Relationship of stress, distress, and the immu-
rology 51:1546–1554 nologic response to a recombinant hepatitis B
51. Pinquart M, Sorensen S (2003) Associations of vaccine. J Fam Pract 32:481–486
stressors and uplifts of caregiving with caregiver 63. Dhabhar FS, McEwen BS (1996) Stress-
burden and depressive mood: a meta-analysis. J induced enhancement of antigen-specific cell-
Gerontol B Psychol Sci Soc Sci 58:P112–P128 mediated immunity. J Immunol
52. Pinquart M, Sorensen S (2003) Differences 156:2608–2615
between caregivers and noncaregivers in psy- 64. Petrie KJ, Booth RJ, Pennebaker JW et al
chological health and physical health: a meta- (1995) Disclosure of trauma and immune
analysis. Psychol Aging 18:250–267. https:// response to a hepatitis B vaccination program.
doi.org/10.1037/0882-7974.18.2.250 J Consult Clin Psychol 63:787–792
326 Anna C. Whittaker
Abstract
Vaccine studies in psychoneuroimmunology (PNI) provide an insight into biopsychosocial interactions and
their role in infectious diseases. How to measure vaccine responses is therefore of critical importance for
PNI researchers. In this chapter, traditional and modern immunoassays for the assessment of vaccine
responses are discussed, highlighting how modern multiplex techniques provide researchers with greater
capacity and opportunity for novel research relating to vaccine outcomes.
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_17, © Springer Science+Business Media, LLC, part of Springer Nature 2018
327
328 Kieran Ayling et al.
Detection
Antibody
Serum
Antibody IgG
Antigen Antibody
Antigen “A” coated well Antigen “B” coated well Antibody coated control
well
Fig. 1 Schematic overview of an ELISA. Wells are coated on a 96-well plate with antigens of interest or a serial
dilution of human IgG (to generate a standard curve). After overnight incubation, the plate is washed and then
blocked, before addition of serum samples. After further incubation, the plate is again washed and a detection
antibody is added, followed by addition of substrate. The plate is read on a plate reader to obtain optical
density values for both the standard curve and the specific antibody responses
330 Kieran Ayling et al.
3.1.2 Advantages ELISAs are relatively simple to perform, low cost, and the only
and Limitations of ELISA specialist equipment required is the optical scanner that is common
to most laboratories. The assay produces a continuous measure of
antibody levels, through colorimetric change, and can be easily
adapted to measure any antibody isotype (e.g., IgA, IgM). Higher
antibody levels are indicative of greater clinical protection,
although there are no accepted thresholds at which antibody levels
as measured by ELISA are sufficient to be considered protective.
Further, in ELISA, absorbance is measured objectively by an optical
scanner and requires fewer consumables per sample than other
traditional assays (such as the HAI assay discussed below), as titra-
tion is not performed.
There are, however, multiple limitations of ELISA. To begin,
ELISA relies on colorimetric change, which has a comparatively
small dynamic range (values at which there is a linear relationship
Measuring Vaccine Responses in the Multiplex Era 331
3.2 Hemagglu- The HAI assay is a quasi-functional assay primarily used in the
tination Inhibition measurement of antibody response to influenza vaccination. It
Assay (HAI) relies on the characteristic property of red blood cells to form a
lattice-like suspension structure in the presence of the influenza
virus [16]. This process, known as hemagglutination, occurs
because receptors on the surface of red blood cells bind to hemag-
glutinin glycoproteins that are present on the surface of the influ-
enza virus. Importantly, if present in sufficient quantities,
antibodies specific to the hemagglutinin on the influenza viruses
(e.g., those produced following vaccination) can inhibit or prevent
this hemagglutination, because they bind to the same glycoproteins
on the viral surface, preventing them from interacting with the red
blood cells.
Practically, these characteristics of red blood cells, influenza
virus, and hemagglutination-inhibiting antibodies are manipulated
in the HAI assay (see Fig. 2).
1. Nonspecific agglutinins are removed by incubating samples
overnight with a receptor-destroying enzyme.
2. Serum samples are serially diluted to increasing levels (e.g., 1:4,
1:8, 1:16, 1:32) and then incubated in a microtiter plate with a
standardized quantity of influenza viral antigen (see Note 3).
3. A standardized quantity of red blood cells are then added to the
plate and if antibodies are present in sufficient quantities this
will inhibit the hemagglutination process, with red blood cells
settling in a small pellet at the bottom of the well. Antibody
titers are then read as the reciprocal of the highest dilution at
which hemagglutination is inhibited.
332 Kieran Ayling et al.
No Virus Control
1:10240
1:1280
1:2560
1:5120
1:160
1:320
1:640
1:10
1:20
1:40
1:80
Sample A
Sample B
Sample C
Settled Red
Sample D Blood Cells
Sample E No Reaction
Sample F
(Control Well)
Sample G
Sample H
No Serum Control
Serum Antibody
bound to virus
Lattice Suspension of
Red Blood Cells and
Settled Red Virus
Blood Cells
Hemagglutination Hemagglutination
Inhibited
Fig. 2 Schematic overview of a hemagglutination inhibition assay. Viral antigens and serially diluted sera are
incubated together in a nonreactive 96-well plate, alongside controls. A standardized quantity of red blood
cells are then added. If antibodies are present in sufficient quantities this will inhibit the hemagglutination
process, with red blood cells settling in a small pellet at the bottom of the well. Antibody titers are read by eye
as the reciprocal of the highest dilution at which hemagglutination is fully inhibited
3.2.2 Advantages The main advantages of the HAI assay are that it is relatively of low
and Limitations of HAI cost, is simple to perform, and does not require much specialist
equipment. Further, the HAI assay has been examined in multiple
viral challenge and epidemiological studies, meaning that estimated
thresholds for clinical protection have been established
[17–19]. This allows researchers to categorize samples according
to whether they reach “protective” levels.1
However, the HAI assay has many limitations. First, results of
the assay are read visually by the researcher, potentially adding a
1
Note that due to the variability in the real-world exposure to viral particles, it is not possible to truly say that an
individual is protected from infection based on antibody levels. Rather, protection estimates are probabilistic.
Measuring Vaccine Responses in the Multiplex Era 333
4 Multiplex Assays
4.1 Bead-Based Bead-based multiplex assays are characterized by their use of mini-
Assays ature polymer or glass beads (typically <10 μm in diameter) with
identifiable fluorescent signatures that can be detected using flow
cytometry techniques. Multiple sets of beads, each with different
fluorescent signatures, can be coated with specific capture antibo-
dies or antigens of interest and added together to a single sample
allowing the quantification of multiple analytes simultaneously.
Currently, there are two main types of bead-based assays:
334 Kieran Ayling et al.
Multiple
Bead Conjugated
Bead Sets
with Antigen
Detection
Antibody
Sample
Serum Antibody
Fig. 3 Schematic overview of bead-based assays. Separate bead sets are conjugated with antigens of interest,
before being added to and incubated with serum samples. Detection antibodies are then added and left to
incubate further. Beads are then washed and are read using an analytical flow chamber supporting individual
bead separation. Signals for each bead set can then be read using laser fluorescence techniques and
compared to a standard/calibration curve
Measuring Vaccine Responses in the Multiplex Era 335
4.1.2 Advantages As with all multiplex assays, the central advantage of bead-based
and Limitations of Bead- assays is that multiple analytes of interest (e.g., different antibody
Based Assays strains, isotypes) can be measured simultaneously in a given sample.
This is particularly valuable if sample volumes are limited and/or a
wide variety of analytes are required to be measured. As bead-based
assays rely on fluorescence signals, the assay also typically has a large
dynamic range. While these advantages are significant, it is impor-
tant to note that these assays require a greater level of expertise than
the traditional assays described above (e.g., use of flow cytometry
equipment). Further, unless already available to researchers, the
equipment costs associated with bead-based assays can be
prohibitive.
4.2 Microarray Microarray assays [21] have been widely used in genetic research,
Assays most frequently to examine the expression of multiple gene
sequences by simultaneously “printing” DNA or RNA in minute
quantities to a reactive glass slide [22, 23]. More recently, this
technology has been adapted to allow antibodies against specific
antigens to be measured in multiple, miniaturized assays (so-called
336 Kieran Ayling et al.
Detection
Slide Holder Antibody
Serum
Antibody
Printed Antigen
Glass Slide
Slide
Fig. 4 Schematic overview of microarray assays. Antigens of interest and serial dilutions of human IgG
(to generate a standard curve) are printed onto a reactive slide surface by an arrayer. The remainder of the
slide surface is then blocked to prevent further nonspecific binding. Diluted serum samples are then added to
the surface. After washing, a detection antibody is added, followed by a fluorescent dye that binds to the
detection antibody. Slides are read using a laser scanner to obtain fluorescence values for the standard curve
and specific antibody responses
4.2.2 Advantages One of the most significant advantages of antigen microarray assays
and Limitations is that they are miniaturized and multiplex, meaning that many
of Microarray Assays types of antibodies can be detected within a single processing of a
sample [24]. This vastly reduces the quantities of sera, antigen,
reagents, and time required to perform the assay [15]. Large parts
of the process are automated and resultant fluorescence signals are
read through laser scanning techniques, which produce a highly
specific continuous outcome with a large dynamic range. Further,
antigen microarrays can be made increasingly robust as a large
number of replicates can be performed simultaneously and they
can be adapted to include multiple internal quality control mea-
sures [24]. The most limiting factor of this assay is that it requires
338 Kieran Ayling et al.
5 Conclusion
6 Notes
References
1. Vedhara K, Cox NK, Wilcock GK et al (1999) 10. Vedhara K, Bennett PD, Clark S et al (2003)
Chronic stress in elderly carers of dementia Enhancement of antibody responses to influ-
patients and antibody response to influenza enza vaccination in the elderly following a
vaccination. Lancet 353:627–631 cognitive-behavioural stress management
2. Glaser R, Sheridan J, Malarkey WB et al (2000) intervention. Psychother Psychosom
Chronic stress modulates the immune response 72:245–252
to a pneumococcal pneumonia vaccine. Psy- 11. Siegrist C (2012) Vaccine immunology. In:
chosom Med 62:804–807 Plotkin S, Orenstein W, Offit P (eds) Vaccines,
3. Hayney MS, Coe C, Muller D et al (2014) Age 6th edn. Elsevier Inc., Amsterdam, pp 17–36
and psychological influences on immune 12. McElhaney JE, Ewen C, Zhou X et al (2009)
responses to trivalent inactivated influenza vac- Granzyme B: correlates with protection and
cine in the meditation or exercise for prevent- enhanced CTL response to influenza vaccina-
ing acute respiratory infection (MEPARI) trial. tion in older adults. Vaccine 27:2418–2425
Hum Vaccin Immunother 10:2759–2767 13. McElhaney JE, Xie D, Hager WD et al (2006)
4. Ayling K, Fairclough L, Tighe P et al (2018) T cell responses are better correlates of vaccine
Positive mood on the day of influenza vaccina- protection in the elderly. J Immunol
tion predicts vaccine effectiveness: a prospec- 176:6333–6339
tive observational cohort study. Brain Behav 14. Lange T, Dimitrov S, Bollinger T et al (2011)
Immun 67:314–323. https://doi.org/10. Sleep after vaccination boosts immunological
1016/j.bbi.2017.09.008 memory. J Immunol 187:283–290
5. Phillips A (2012) The vaccination model in 15. Leng SX, McElhaney JE, Walston JD et al
psychoneuroimmunology research: a review. (2008) ELISA and multiplex technologies for
In: Yan Q (ed) Psychoneuroimmunology: cytokine measurement in inflammation and
methods and protocols. Springer Science & aging research. J Gerontol A Biol Sci Med Sci
Business Media, LLC, Santa Clara, CA, pp 63:879–884
355–370 16. Pedersen J (2014) Hemagglutination-
6. Burns VE, Gallagher S (2010) Antibody Inhibition Assay for Influenza Virus Subtype
response to vaccination as a marker of in vivo Identification and the Detection and Quantita-
immune function in psychophysiological tion of Serum Antibodies to Influenza Virus.
research. Neurosci Biobehav Rev 35:122–126 In: Spackman E (ed) Animal influenza virus,
7. Chen W, Kozlovsky B, Effros R (2009) Vacci- methods in molecular biology. Springer Sci-
nation in the elderly: an immunological per- ence & Business Media, LLC, New York, pp
spective. Trends Immunol 30:351–359 11–25
8. Woods JA, Keylock KT, Lowder T et al (2009) 17. Hobson D, Curry RL, Beare AS et al (1972)
Cardiovascular exercise training extends influ- The role of serum haemagglutination-
enza vaccine seroprotection in sedentary older inhibiting antibody in protection against chal-
adults: the immune function intervention trial. lenge infection with influenza A2 and B viruses.
J Am Geriatr Soc 57:2183–2191 J Hyg (Lond) 70:767–777
9. Edwards KM, Burns VE, Reynolds T et al 18. de Jong JC, Palache AM, Beyer WEP et al
(2006) Acute stress exposure prior to influenza (2003) Haemagglutination-inhibiting anti-
vaccination enhances antibody response in body to influenza virus. Dev Biol (Basel)
women. Brain Behav Immun 20:159–168 115:63–73
340 Kieran Ayling et al.
19. Coudeville L, Bailleux F, Riche B et al (2010) 23. Jiang H, Deng Y, Chen H et al (2004) Joint
Relationship between haemagglutination- analysis of two microarray gene-expression data
inhibiting antibody titres and clinical protec- sets to select lung adenocarcinoma marker
tion against influenza: development and appli- genes. BMC Bioinfomatics 5:1–12
cation of a bayesian random-effects model. 24. Negm OH, Hamed MR, Dilnot EM et al
BMC Med Res Methodol 10:18 (2015) Profiling humoral immune responses
20. Segerstrom S, Smith G (2012) Methods, vari- to clostridium difficile-specific antigens by pro-
ance, and error in psychoneuroimmunology tein microarray analysis. Clin Vacc Immunol
research: the good, the bad, and the ugly. In: 22:1033–1039
Segerstrom S (ed) The Oxford handbook of 25. Selvarajah S, Negm OH, Hamed MR et al
psychoneuroimmunology. Oxford University (2014) Development and validation of protein
Press, New York, pp 421–432 microarray technology for simultaneous
21. Chang T (1983) Binding of cells to matrixes of inflammatory mediator detection in human
distinct antibodies coated on solid surface. J sera. Mediators Inflamm 2014:820304
Immunol Methods 65:217–223 26. Tighe P, Ryder RR, Todd I et al (2015) ELISA
22. Shena M, Shalon D, Davis RW et al (1995) in the multiplex era: potentials and pitfalls.
Quantitative monitoring of gene expression Proteomics Clin Appl 9:406–422
patterns with a complimentary DNA microar-
ray. Science 270:467–470
Chapter 18
Abstract
The non-neuronal, immunological effects of the cholinergic signaling are exerted on the system’s scale of
observation via the vagus nerve and on the cellular scale via α7 nicotinic acetylcholine receptor (nAChR)
signaling in myeloid cells of the periphery or brain’s microglia and astrocytes. The developmental effects of
such multi-scale signaling can be conceived of as an example of psychoneuroimmunological (PNI) home-
okinesis and, while reported in the literature, are not yet systematically well studied. To be better under-
stood, the intricacy of the multi-scale interactions requires relevant preclinical animal models. Chronically
instrumented non-anesthetized fetal sheep model comes with a strong track record of bench-to-bed
translation and a large body of evidence for its strong resemblance to and relevance for human physiology
on various scales of organization. Recently, there has been growing interest in pleiotropic effects of vagus
nerve stimulation (VNS) on various organ systems such as innate immunity, metabolism, and emotion with
implications for programming of PNI phenotype. Here we describe the procedures required to record and
manipulate the vagus nerve activity in this large pregnant mammalian organism. Extending this in vivo
model to in vitro, on the cellular scale, we present the method to manipulate the cholinergic signaling in
ovine fetal microglia and astrocytes and analyze their responses on protein and RNA levels. Together these
models can provide multi-scale-level mechanistic insights into the effects of cholinergic signaling on PNI
phenotype.
Key words Vagus nerve, Vagotomy, Nerve stimulation, ENG, Animal model, Physiology, Surgery,
Chronic experimentation, Multivariate data acquisition, Neuroscience
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_18, © Springer Science+Business Media, LLC, part of Springer Nature 2018
341
342 Martin G. Frasch et al.
Vagus nerve in
Fetal Sheep – Electrocardiogram (ECG)
– Vagus electroneurogram (VENG)
– Heart rate variability (HRV)
Analysis
– Complex signals bioinformatics
Data Acquisition
Astrocytes: – ELISA
– RNAseq
α7nAChR Analysis
Microglia:
– Bioinformatics
Fig. 1 The in vivo–in vitro approach to cholinergic manipulation in a large mammalian developing organism.
α7nAChR, α7 nicotinic acetylcholine receptor. In vivo, unanesthetized fetal sheep instrumentation allows for
chronic multimodal recording and stimulation of the cervical vagus nerve unilaterally or bilaterally. This is
followed by a set of data analyses to tie together the findings. Mirroring the cholinergic manipulation on the
systems scale, the in vitro primary glia (astrocytes or microglia) cultures can be derived from any of the four
described in vivo experimental groups. This results in creation of naı̈ve (no in vivo exposure, i.e., control) and
second-hit in vitro experimental conditions. Protein- and transcriptome-level responses to manipulation of
cholinergic signaling via the α7nAChR can be tested under conditions of exposure to endotoxin (lipopolysac-
charide) with the presence of a α7nAChR antagonist (equivalent of an in vivo vagotomy) or α7nAChR agonist
(equivalent of an in vivo vagus nerve stimulation). This is again followed by bioinformatics analyses. Creating
one cohesive PNI-type physiological framework from the systems-scale and cellular-scale bioinformatics
results and models represents an exciting challenge for the immediate future
2 In Vivo Protocol
2.1 Cholinergic Pregnant time-dated ewes are instrumented at the gestational age
Signaling of choice, usually starting at around 105 days of gestation (dGA,
Manipulation In Vivo 145 dGA term) with arterial, venous, and amniotic catheters and
Via Vagus Nerve ECG electrodes [8]. Ovine singleton fetuses of mixed breed are
Stimulation surgically instrumented with sterile technique under general anes-
thesia (both ewe and fetus). In case of twin pregnancy, the larger
2.1.1 Anesthesia and fetus is chosen based on palpating and estimating the outer inter-
Surgical Procedure orbital distance. The total duration of the surgical procedure (i.e.,
skin to skin and not counting preparation and anesthesia induction)
is about 2 h. Antibiotics are administered preoperatively to the
mother intravenously (trimethoprim sulfadoxine 5 mg/kg) as well
as to the fetus intravenously and into the amniotic cavity (ampicillin
250 mg) at the time when fetus is returned into the amniotic cavity
and the uterus has been closed. Amniotic fluid lost during surgery is
replaced with warm saline which in our experience averages
120 mL. The catheters exteriorized through the maternal flank
were secured to the back of the ewe in a plastic pouch. For the
duration of the experiment the ewe is returned to the metabolic
cage, where she can stand, lie, and eat ad libitum while the
non-anesthetized fetus is being monitored without sedating the
mother. During postoperative recovery, antibiotic administration
is continued for 3 days. Arterial blood is sampled for evaluation of
maternal and fetal condition and catheters are flushed with hepa-
rinized saline to maintain patency (1 U/mL, 2 mL maximum at a
time and no more than 10 mL daily).
Bilateral cervical vagotomy is performed in sheep fetuses as
follows. The head is extended and the neck is kept straight and
stable while approaching and exposing the vagal nerves. A bilateral
cervical vagotomy (Vx) is then performed and the VNS/VENG
electrodes are installed around the nerve. Vx step can be omitted.
Vx has the advantage that a more mechanistic dissection of efferent
and afferent VNS can be undertaken. The skin is then closed in a
continuous pattern. The animals not receiving Vx or instrumenta-
tion with VNS/VENG probe are subjected to a sham neck surgery
with all steps except Vx.
During surgery (once the first fetal arterial catheter was in place
and before returning the fetus to the uterus) a 3 mL fetal arterial
blood sample is taken for blood gas, lactate, and cytokine
measurements.
2.2 Vagus Nerve ECG, FHR, and arterial blood pressure can be monitored continu-
Electroneurogram ously (for example using the 1902 amplifier and micro3 1401 ADC
Recordings (VENG) and by CED, Cambridge, UK, and NL108A, NeuroLog, Digitimer,
Nerve Stimulation Hertfordshire, UK) and sampled at 1000 and 256 Hz, respectively.
(VNS) VNS can be applied via NeuroLog’s NL512/NL800A using pulse
Sculpting the Sculptors: Methods for Studying the Fetal Cholinergic. . . 345
2.3 Materials and Data acquisition system can be used such as a 1902 amplifier and
Methods micro3 1401 ADC by CED, Cambridge, UK. VENG should be
recorded at 20,000 Hz sampling rate. Care should be taken in
designing the VNS protocol, as this area is far from being well
defined and understood [2]. Nerve stimulation system can be
used such as NL108A, NeuroLog, Digitimer, Hertfordshire, UK.
The general framework of the pregnant sheep approach has
been published in [8]. This paper reports the precise materials
and steps needed to implement the approach.
3 In Vitro Protocol
3.1.1 In Vitro Microglia Fetal sheep brain tissues are obtained during sheep autopsy after
Culture Protocol completion of the in vivo experiment to conduct the in vitro study.
The non-instrumented, untreated twins were designated “naı̈ve”
(no LPS exposure in vivo). Instrumented animals that received LPS
in vivo were used for second-hit LPS exposure in vitro.
Fetal sheep microglia culture protocol was adapted from an
established human adult and fetal microglia culture protocol that
was modified to include a myelin removal step following the high-
speed centrifugation [30].
Briefly, fetal sheep cells were plated on poly-L-lysine (PLL)-
coated tissue culture flasks at a concentration of 2 106 cells/
mL in DMEM with 5% heat-inactivated fetal bovine serum (Gibco,
Canada Origin), 1% penicillin/streptomycin, and 1% glutamine (5%
DMEM), in which microglia grow best [30]. Cells were allowed to
incubate for 7 days at 37 C, 5% CO2, followed by a media change
by centrifugation and the addition of resuspended cells back to the
culture flask. Cells were continued to incubate for seven more days
with 5% DMEM at 37 C, 5% CO2, before the floating cells were
collected. After carefully collecting the floating microglia to avoid
contamination with astrocytes and oligodendrocytes, the cells were
incubated in 24-well plate at 1 105 cells/mL with 5% DMEM for
another 4–5 days, and then treated with or without LPS (100 ng/
mL, Sigma L5024, from E. coli O127:B8) for 6 h. Cell-conditioned
media were collected for cytokine analysis, and 0.5 mL TriZol per
well was added for RNA extraction.
To verify microglia purity, a portion of floating cells was
cultured in 24-well plate under the above conditions for flow
cytometry analysis. The cell morphology was documented with
light microscopy. Another portion of floating cells was plated
onto Lab-Tek 8-well chamber glass slide (Thermo Scientific) and
treated with or without LPS for immunocytochemistry (ICC)
analysis.
3.1.2 In Vitro Astrocyte Astrocytes are the major adherent cell population in flask. Astro-
Culture Protocol cytes are purified by passage into a new T75 flask (no PLL coated is
required) for 4–5 times before any manipulations and treatments.
After floating microglia collection, detach the adherent cells by
trypsinization (trypsin 0.25% + EDTA 0.1%, Wisent Cat. No
325-043-EL) and replate into a new flask; culture for another
7 days with 10% all-dressed DMEM. Repeat the above step 2.6.2
for 4–5 times; the cell passage procedure takes 4–5 weeks until
purified astrocytes can be used for treatment.
Plate pure astrocytes into 24-well plate at 1 105 cells/mL
with 10% DMEM for another 7 days, and then treat with or
Sculpting the Sculptors: Methods for Studying the Fetal Cholinergic. . . 347
3.1.3 Cell Culture and Prior to exposure to LPS, cells are pretreated for 1 h with either
Treatment 10 nM AR-R17779 hydrochloride (Tocris Bioscience Cat# 3964),
a selective α7nAChR agonist, or 100 nM α-bungarotoxin (Tocris
Bioscience Cat# 2133), a selective α7nAChR antagonist. Optimal
dose of AR-R17779 (A) or α-bungarotoxin (B) was chosen based
on a dose-response experiment with LPS; AR-R17779 hydrochlo-
ride was dissolved in DMSO into a stock solution. α-Bungarotoxin
was reconstituted with culture media into a stock solution and
underwent serial dilutions. AR-R17779 and α-bungarotoxin pre-
parations are added well by well; the same volume of vehicle (either
DMSO or cell culture media) was added in control wells. There-
fore, in a complete cell culture experiment, we have four experi-
mental groups: control (naı̈ve control or NC), LPS (naı̈ve LPS or
NL), LPS + B (naı̈ve LPS + B or NB), and LPS + A (naı̈ve LPS + A
or NA). If the ovine fetuses were preexposed to LPS (or another
experimental stimulus of choice) in vivo systemically, this constitu-
tes a second hit in vitro. Second-hit cell cultures are designed with
the same pattern and divided into four experimental groups: con-
trol (SHC), LPS (SHL), LPS + B (SHB), and LPS + A (SHA).
3.1.4 Measurements of The inflammatory responses are assessed on protein level with
Inflammatory Responses ELISA and on transcriptome level with RNAseq methods which
are described below in detail.
Measurement of Cytokines Cytokine concentrations in cell culture media (IL-1β) are deter-
in Plasma and Cell Culture mined by using an ovine-specific sandwich ELISA reported in detail
Media in [24, 26] which is available as in open access.
RNAseq Approach The overall experimental design can be divided into three phases:
sequencing, quantification, and discovery.
3.1.5 RNA Extraction and Total RNA is extracted from cultured microglia using TRIzol
RNA Quantification Reagent (Life Technologies). RNA quantity and quality (RNA
integrity number, RIN) are established by using a RNA Nano
Chip (Agilent RNA 6000 Nano Chips) with Agilent 2100
BioAnalyzer.
RNAseq libraries can be prepared using Illumina TruSeq RNA
Sample Preparation v2 kit (Illumina) and quality control can be
348 Martin G. Frasch et al.
3.1.6 RNAseq Data Read alignment to the reference genome. To maximize the number
Analysis of genes covered, raw data are mapped to the reference genome of
the sheep Ovis aris v3.1 from Ensembl (GCA_000298735.1) as the
reference genome. Index of the reference fasta file is built with
Bowtie2 [31]. Trim the adaptor of the fastQ files with TrimGalore
and map reads to the reference genome with Tophat2 [32]. From
the aligned reads from Tophat2, the number of reads aligned is
counted with HTseq and assembled into a matrix containing the
read count of each gene per sample [33].
3.1.7 Normalization and In order to find differentially expressed genes we used DESeq2 to
Transcriptome Analysis normalize the dataset, and generate Log2-fold changes and adja-
cent p values (padj) [34]. Customarily, a gene is considered differ-
entially expressed if its adjacent p-value is strictly lower than 0.1.
Pools of up- and downregulated differentially expressed genes are
clustered and visualized into heat maps, generated in R using the
log2-normalized counts and the heatmap.2 method of the gplots
library [35].
3.1.8 Gene Selection and The sheep genome is not yet supported by most gene ontology
Gene Ontology (GO) platforms; therefore, downstream analyses are performed with
orthologs in the human genome Homo sapiens. To select the rele-
vant genes among the upregulated and downregulated genes, per-
form gene enrichment analysis for biological process and molecular
function with ToppGenes and FDR <0.05 [36, 37]. Bar diagram of
significant GO terms ( p < 10–3) is presented on a –Log (P) scale.
Protein-protein interaction networks are generated with the
STRING database [38]. Gene Ontology can also be performed in
parallel with PantherDB and only biological processes be presented
in the pie charts [39].
3.2 Materials and Conduct tissue preparation and cell culture under aseptic condi-
Methods tions; every material used needs to be sterile.
1. Tissue culture T75 or T175 flask.
3.2.1 In Vitro Culture
Protocol 2. 24-Well cell culture plate.
3. Poly-L-lysine (PLL) solution (working concentration of 10 μg/
mL): Dilute PLL stock (1 mg/mL) by 1:100 with ultrapure
water. Prepare and use fresh. Prepare all the solutions and
buffers with ultrapure water (distilled and deionized water to
attain a sensitivity of 18 MΩ-cm at 25 C).
4. Microglia culture media (5% DMEM): DMEM with 5% heat-
inactivated fetal bovine serum (Gibco, Canada Origin), 1%
penicillin/streptomycin, and 1% glutamine. Microglia grow
preferably in DMEM with 5% heat-inactivated fetal bovine
serum. In the flask, microglia are cells of the floating portion;
care should be taken to collect those cells to avoid contamina-
tion with other cell types.
5. Astrocyte culture media (10% DMEM): DMEM with 10%
heat-inactivated fetal bovine serum (Gibco, Canada Origin),
1% penicillin/streptomycin, and 1% glutamine. Astrocytes
grow better in DMEM with 10% heat-inactivated fetal bovine
serum. Astrocytes are the adherent cells that need to be
detached by trypsinization after several passages and washes.
6. Lipopolysaccharides (LPS, Sigma L5024, from E. coli O127,
B8): Reconstitute the lyophilized LPS powder with sterile
saline, keep the stock at +4 C, and prepare LPS working
solution fresh at a concentration of 100 ng/mL.
7. Cell trypsinization solution: Trypsin 0.25% + EDTA 0.1%,
Wisent Cat. No 325-043-EL.
8. Lab-Tek 8-well chamber glass slide (Thermo Scientific).
9. TriZol reagent (Life Technology) or Qiagen RNeasy Mini Kit
(Cat no. 74104).
10. Microglia marker Iba1polyclonal antibody (Wako Chemicals,
Cat# 019-19741).
11. Astrocyte maker glial fibrillary acidic protein (GFAP) antibody
(Sigma).
12. Hoechst 33342 (Sigma): For immunocytochemistry (ICC)
nuclear count staining.
13. AR-R17779 hydrochloride (Tocris Bioscience Cat# 3964), a
selective α7nAChR agonist.
350 Martin G. Frasch et al.
3.2.2 RNAseq Approach: 1. RNA Nano Chip (Agilent RNA 6000 Nano Chips).
RNA Extraction and 2. Agilent 2100 BioAnalyzer (Agilent).
Quantification, RNAseq
3. Illumina TruSeq RNA Sample Preparation v2 kit (Illumina).
Library Preparation
4. Illumina HiSeq2500 (Illumina).
4 Conclusions
Acknowledgments
References
MG (2016) VNS in inflammation: systematic autonomic nervous system. Cleve Clin J Med
review of animal models and clinical studies. 76(Suppl 2):S86–S90
Bioelectron Med 3:1–6 16. Que CL, Kenyon CM, Olivenstein R, Macklem
3. Wang X, Rousset CI, Hagberg H, Mallard C PT, Maksym GN (2001) Homeokinesis and
(2006) Lipopolysaccharide-induced inflamma- short-term variability of human airway caliber.
tion and perinatal brain injury. Semin Fetal J Appl Physiol (1985) 91:1131–1141
Neonatal Med 11:343–353 17. Clancy JA, Mary DA, Witte KK, Greenwood
4. Gotsch F, Romero R, Kusanovic JP, Mazaki- JP, Deuchars SA, Deuchars J (2014)
Tovi S, Pineles BL, Erez O, Espinoza J, Hassan Non-invasive vagus nerve stimulation in
SS (2007) The fetal inflammatory response healthy humans reduces sympathetic nerve
syndrome. Clin Obstet Gynecol 50:652–683 activity. Brain Stimul 7:871–877
5. Svedin P, Kjellmer I, Welin AK, Blad S, Mallard 18. Liu RP, Fang JL, Rong PJ, Zhao Y, Meng H,
C (2005) Maturational effects of lipopolysac- Ben H, Li L, Huang ZX, Li X, Ma YG, Zhu B
charide on white-matter injury in fetal sheep. J (2013) Effects of electroacupuncture at auricu-
Child Neurol 20:960–964 lar concha region on the depressive status of
6. Nitsos I, Rees SM, Duncan J, Kramer BW, unpredictable chronic mild stress rat models.
Harding R, Newnham JP, Moss TJ (2006) Evid Based Complement Alternat Med
Chronic exposure to intra-amniotic lipopoly- 2013:789674
saccharide affects the ovine fetal brain. J Soc 19. Ylikoski J, Lehtimaki J, Pirvola U, Makitie A,
Gynecol Investig 13:239–247 Aarnisalo A, Hyvarinen P, Ylikoski M (2017)
7. Yan E, Castillo-Melendez M, Nicholls T, Non-invasive vagus nerve stimulation reduces
Hirst J, Walker D (2004) Cerebrovascular sympathetic preponderance in patients with
responses in the fetal sheep brain to low-dose tinnitus. Acta Otolaryngol 137(4):426–431
endotoxin. Pediatr Res 55:855–863 20. Frangos E, Ellrich J, Komisaruk BR (2015)
8. Burns P, Liu HL, Kuthiala S, Fecteau G, Non-invasive Access to the vagus nerve central
Desrochers A, Durosier LD, Cao M, Frasch projections via electrical stimulation of the
MG (2015) Instrumentation of near-term external ear: fMRI evidence in humans. Brain
fetal sheep for multivariate chronic Stimul 8:624–636
non-anesthetized recordings. J Vis Exp (105): 21. Diz-Chaves Y, Pernia O, Carrero P, Garcia-
e52581 Segura LM (2012) Prenatal stress causes altera-
9. Carmel E, Burns P, Durosier D, tions in the morphology of microglia and the
Duchatellier C, Cao M, Desrochers A, inflammatory response of the hippocampus of
Fecteau G, Frasch M (2012) Fetal brain MRI adult female mice. J Neuroinflammation 9:71
- experiences in the ovine model of cerebral 22. Diz-Chaves Y, Astiz M, Bellini MJ, Garcia-
inflammatory response. Reprod Sci 19 Segura LM (2013) Prenatal stress increases
(3):347A–348A the expression of proinflammatory cytokines
10. Pavlov VA, Tracey KJ (2012) The vagus nerve and exacerbates the inflammatory response to
and the inflammatory reflex—linking immunity LPS in the hippocampal formation of adult
and metabolism. Nat Rev Endocrinol male mice. Brain Behav Immun 28:196–206
8:743–754 23. Slusarczyk J, Trojan E, Glombik K,
11. Pavlov VA, Tracey KJ (2015) Neural circuitry Budziszewska B, Kubera M, Lason W,
and immunity. Immunol Res 63:38–57 Popiolek-Barczyk K, Mika J, Wedzony K,
12. Hosoi T, Okuma Y, Nomura Y (2000) Electri- Basta-Kaim A (2015) Prenatal stress is a vulner-
cal stimulation of afferent vagus nerve induces ability factor for altered morphology and
IL-1beta expression in the brain and activates biological activity of microglia cells. Front
HPA axis. Am J Physiol Regul Integr Comp Cell Neurosci 9:82
Physiol 279:R141–R147 24. Cortes M, Cao M, Liu HL, Burns P, Moore C,
13. Farrokhi CB, Tovote P, Blanchard RJ, Blan- Fecteau G, Desrochers A, Barreiro LB, Antel
chard DC, Litvin Y, Spiess J (2007) Cortagine: JP, Frasch MG (2017) RNAseq profiling of
behavioral and autonomic function of the primary microglia and astrocyte cultures in
selective CRF receptor subtype 1 agonist. near-term ovine fetus: a glial in vivo-in vitro
CNS Drug Rev 13:423–443 multi-hit paradigm in large mammalian brain.
J Neurosci Methods 276:23–32
14. Porges SW (1995) Cardiac vagal tone: a physi-
ological index of stress. Neurosci Biobehav Rev 25. Borovikova LV, Ivanova S, Zhang M, Yang H,
19:225–233 Botchkina GI, Watkins LR, Wang H,
Abumrad N, Eaton JW, Tracey KJ (2000)
15. Porges SW (2009) The polyvagal theory: new Vagus nerve stimulation attenuates the
insights into adaptive reactions of the
352 Martin G. Frasch et al.
systemic inflammatory response to endotoxin. 33. Anders S, Pyl TP, Huber W (2015) HTSeq — a
Nature 405:458–462 python framework to work with high-
26. Cao M, Cortes M, Moore CS, Leong SY, throughput sequencing data. Bioinformatics
Durosier LD, Burns P, Fecteau G, 31(2):166–169
Desrochers A, Auer RN, Barreiro LB, Antel 34. Love MI, Huber W, Anders S (2014) Moder-
JP, Frasch MG (2015) Fetal microglial pheno- ated estimation of fold change and dispersion
type in vitro carries memory of prior in vivo for RNA-Seq data with DESeq2. Genome Biol
exposure to inflammation. Front Cell Neurosci 15(12):550
9:294 35. Warnes GR (2008) Gplots: various R program-
27. Cortes M, Cao M, Liu HL, Moore CS, Duro- ming tools for plotting data
sier LD, Burns P, Fecteau G, Desrochers A, 36. Chen J, Bardes EE, Aronow BJ, Jegga AG
Barreiro LB, Antel J (2017) α7 nicotinic ace- (2009) ToppGene Suite for gene list enrich-
tylcholine receptor signaling modulates the ment analysis and candidate gene prioritiza-
inflammatory and iron homeostasis in fetal tion. Nucleic Acids Res 37:W305–W311
brain microglia. bioRxiv:097295 37. Kaimal V, Bardes EE, Tabar SC, Jegga AG,
28. Shytle RD, Mori T, Townsend K, Vendrame M, Aronow BJ (2010) ToppCluster: a multiple
Sun N, Zeng J, Ehrhart J, Silver AA, Sanberg gene list feature analyzer for comparative
PR, Tan J (2004) Cholinergic modulation of enrichment clustering and network-based dis-
microglial activation by alpha 7 nicotinic recep- section of biological systems. Nucleic Acids Res
tors. J Neurochem 89:337–343 38:W96–W102
29. Frasch MG, Szynkaruk M, Prout AP, 38. Franceschini A, Szklarczyk D, Frankild S,
Nygard K, Cao M, Veldhuizen R, Kuhn M, Simonovic M, Roth A, Lin J,
Hammond R, Richardson BS (2016) Minguez P, Bork P, von Mering C, Jensen LJ
Decreased neuroinflammation correlates to (2013) STRING v9.1: protein-protein interac-
higher vagus nerve activity fluctuations in tion networks, with increased coverage and
near-term ovine fetuses: a case for the afferent integration. Nucleic Acids Res 41:D808–D815
cholinergic anti-inflammatory pathway? J Neu- 39. Mi H, Poudel S, Muruganujan A, Casagrande
roinflammation 13:103 JT, Thomas PD (2016) PANTHER version 10:
30. Durafourt BA, Moore CS, Blain M, Antel JP expanded protein families and functions, and
(2013) Isolating, culturing, and polarizing pri- analysis tools. Nucleic Acids Res 44:
mary human adult and fetal microglia. Meth- D336–D342
ods Mol Biol 1041:199–211 40. Livak KJ, Schmittgen TD (2001) Analysis of
31. Langmead B, Salzberg SL (2012) Fast gapped- relative gene expression data using real-time
read alignment with Bowtie 2. Nat Methods quantitative PCR and the 2( Delta Delta C
9:357–359 (T)) Method. Methods 25:402–408
32. Kim D, Pertea G, Trapnell C, Pimentel H, 41. Untergasser A, Cutcutache I, Koressaar T, Ye J,
Kelley R, Salzberg SL (2013) TopHat2: accu- Faircloth BC, Remm M, Rozen SG (2012)
rate alignment of transcriptomes in the pres- Primer3—new capabilities and interfaces.
ence of insertions, deletions and gene fusions. Nucleic Acids Res 40:e115
Genome Biol 14:R36
Chapter 19
Abstract
Prenatal stress (PS) impacts early behavioral, neuroimmune, and cognitive development. Pregnant rat
models have been very valuable in examining the mechanisms of such fetal programming. A newer pregnant
sheep model of maternal stress offers the unique advantages of chronic in utero monitoring and manipula-
tion. This chapter presents the techniques used to model single and multigenerational stress exposures and
their pleiotropic effects on the offspring.
Key words DOHaD, Stress, Pregnancy, Animal models, Multi-hit, Generations, Rat, Sheep
1 Introduction
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1_19, © Springer Science+Business Media, LLC, part of Springer Nature 2018
353
354 Martin G. Frasch et al.
Notably, stress in early life programs health risks not only across
a single lifetime, but also across multiple generations [9, 20]. For
example, preterm birth risk was suggested to propagate through
generations [21], which is supported by our observation that stress
across generations programs higher vulnerability to preterm birth,
adverse offspring development, and epigenetic markers of human
preterm birth [22].
For example, David Olson and his team [23, 24] assessed
chronic, lifelong stressors and related these to spontaneous preterm
birth by designing a new questionnaire administered to postpartum
mothers who delivered preterm or at term. In these studies, the
combined score of childhood and adult abuse was significantly
associated with higher preterm birth risk [23, 24]. Moreover, the
proportion of women showing shortened gestation gradually
increased with a greater number of adverse childhood experiences.
These findings suggest a close association between adverse early-life
experiences and preterm birth risk, which represents a risk factor for
adverse health outcomes in later life, with psychoneuroimmunolo-
gical mechanisms thought to play a key mediating role [25]. Indi-
vidual psychoneuroimmunological traits are in turn influenced by
preterm birth which informs a transgenerational chain of events
[26, 27].
In the following, we present the pregnant rat and sheep models
of single-, trans-, and multigenerational prenatal stress.
2.1 Pregnant Rat For the last 20 years, the focus of our research has been centered in
Model of Prenatal the study of the outcomes of various neurotransmitter pathways
Stress and hormones in a rodent model of prenatal restraint stress trig-
gered by the hypothesis that stressful situations suffered prenatally
2.1.1 Overview are related to the propensity to develop psychiatric abnormalities in
the human adult life. Results obtained along these years showed
impairments at the biochemical, morphological, and behavioral
levels in the offspring of gestationally stressed dams and reversion
by a cross-fostering model.
In a brief summary of the results obtained along these years
with the single-generation rodent model of PS, we would like to
highlight the following results: (1) PS impairs dopamine metabo-
lism throughout development of the male offspring: In a model of
PS in pregnant rats, induced by restriction of movement and
subsequent cross fostering, PS increased DA D2 receptors in limbic
areas, decreased DA-stimulated release in cortical areas, and
impaired the expression of specific transcription factors during
development as well as the expression of TH and transporters. PS
altered the asymmetry in D2-type receptors in the nucleus accum-
bens, an area associated with impulsivity [28–33]. (2) PS impairs
356 Martin G. Frasch et al.
2.1.2 Model Design Vaginal smears were collected daily for 8 days before mating to
determinate the stage of the estrus cycle and the day of conception.
On the day of proestrus, sexually experienced male Wistar rats
weighing 250–300 g were introduced for mating. Vaginal smears
were taken on the following morning. The day on which sperma-
tozoa were found in the smear was designated day 1 of pregnancy.
A constant light/dark cycle (on at 06:00 h, off at 18:00 h) and the
temperature of 21–25 C were maintained.
Prenatal Stress. Pregnant dams are randomly assigned to either
the control (C) or the prenatal stress (PS) group and individually
housed with ad libitum access to standard rat chow and water. C
rats are left undisturbed in the home cage, while PS dams are
subjected to a restraint stress procedure, which involves rats being
transferred to an experimental room where the stressor is applied.
Pregnant females are individually placed into a transparent plastic
restrainer fitted closely to body size for three 45-min periods per
day (09:00, 12:00, and 16:00 h) between days 14 and 21 of preg-
nancy. The restrainer has ventilation holes, and dimensions appro-
priate for a pregnant rat of 350 g: internal diameter 64 mm, and an
adjustable length of 149–208 mm. This type of stressor is chosen
because it has an indirect influence on the fetuses via a direct stress
on the mother. The sessions are performed in a lit environment. No
other subjects should be present in the experimental room during
the stress exposure.
Postnatal Procedures. On the day of parturition, litter charac-
teristics are recorded and litters are culled to ten pups, maintaining
similar number of males and females, when possible. Weaning is
performed at postnatal day (PND) 21. The male and female off-
spring are housed in separated cages, with no more than five rats per
cage, and in standard housing conditions. To avoid litter effects,
1–2 pups from each litter (a female and a male) should be tested for
each experiment. Behavioral testing is conducted at that period
using the elevated plus maze and light/dark box tests for anxiety-
like behaviors and sucrose anhedonia test for depressive-like behav-
ior. Brains are collected after decapitation at the end of behavioral
tests at PND 35 or another date depending on the milestones to be
assessed in the given study.
6-OHDA Lesion Procedure. Control and prenatal stressed males
at postnatal day (PND) 75 of age (average weight * 320 g) are
injected with desipramine (25 mg/kg i.p.) 1 h prior to surgery to
358 Martin G. Frasch et al.
2.2 Pregnant Sheep The protocol has been established and validated by Schwab et al.
Model of Prenatal [10, 12, 42, 43]. The following protocol represents a minimalistic
Stress version for the specific experiments to be designed on its basis.
Time-dated pregnant sheep (singletons or twins) are delivered to
the research facility and allowed ~20 days for adaptation and quar-
antine (to check Q fever status as per local requirements—a condi-
tion that may vary across the jurisdictions). The acclimation time
also minimizes any maternal stress due to transportation and new
environment prior to commencing the experiments. During a ster-
ile instrumentation at ~102 dGA, the ewes are equipped with a
venous catheter while the fetus receives precordial ECG leads
sutured in the skin, two arterial and one venous catheters, and an
amniotic catheter exiting via the uterus to the maternal flank and
secured in a pouch together with telemetric transmitter, following a
protocol established in our laboratory [11]. Note that while telem-
etry is advantageous in behavioral studies and allows for longer
periods of data acquisition in social housing with animals undis-
turbed, a completely wired solution to instrumentation is also
possible. All tubing resides in a closed, sterile loop to prevent
bacterial contamination from the mother’s skin or the environ-
ment. The instrumentation procedure is followed by 3–4 days of
postoperative recovery.
2.2.1 Maternal Stress The animals are randomly assigned to control or maternal stress
Protocol groups. The stress group follows a repeated maternal stress proto-
col between 105 and 135 dGA [10]. Maternal stress is induced by
isolating pregnant ewes in a well-lit box (3.0 m 3.0 m 1.4 m,
w l h) with no visual, tactile, or auditory contact with flock-
mates. During isolation, ewes have no access to food or water. Ewes
362 Martin G. Frasch et al.
are isolated for 2 days per week (Monday to Friday) with at least
2 days’ recovery time between the particular isolation bouts. Dura-
tion of isolation is 3 h. Isolation is performed either between 07:00
and 10:00 h, 11:00 and 14:00 h, or 15:00 and 18:00 h. Day of the
week and time of isolation as well as isolation box are changed
randomly within the mentioned parameters to reduce habituation.
Two maternal and fetal blood samples are collected on the day
when the stress procedure is performed from the chronically
installed jugular vein and fetal arterial brachial catheters, respec-
tively: before and during isolation to estimate maternal and fetal
cortisol concentrations and degree of habituation. The sample
before the first isolation on the experimental day 1 (at 105 dGA)
serves as animal’s own baseline sample. There will be no additional
blood sampling on days between the isolations. Note that the
ability to sample fetal and maternal blood consecutively within
each animal reduces the sample size required or increases the
power of the study, respectively. This should be considered during
sample size estimations.
Blood samples in non-stressed (control) animals are taken at
the same times as in stressed animals. Maternal and fetal ECG, fetal
blood, and amniotic pressures are recorded telemetrically and con-
tinuously throughout the experiment.
2.2.2 Necropsy and Fetuses are delivered electively on 136 dGA. The ewe is killed with
Postmortem Analyses overdose of Euthanyl. The lamb can be either killed for histological
studies or weaned and monitored postnatally, as per specific proto-
col. The lamb brains are extracted and prepared into the regions of
interest for later processing. The approach will depend on the
specific study. One option is to proceed as follows. One hemi-
sphere’s brain regions of interest such as hypothalamus, hippocam-
pus, and prefrontal cortex tissues are split to be snap frozen to
perform DNA methylation analyses and for RNAlater storage to
conduct targeted or high-throughput RNA analyses. Another
hemisphere’s same brain regions are immersion-fixed and used to
perform immunohistochemistry. The remaining brain regions of
each hemisphere can be snap-frozen or immersion-fixed, respec-
tively, for future analyses.
2.2.3 Sample and Data The animals are kept together in dens with ad libitum access to food
Acquisition and water for 30 days with intermittent maternal and FHR record-
ing and sampling for maternal and fetal blood gases, metabolites,
and stress hormones. Maternal blood samples are collected from
the chronically installed jugular vein catheter regularly before and
during isolation to estimate maternal cortisol concentrations and
degree of habituation. Along with blood gas, glucose, and lactate
measurements, plasma samples can be stored at 80 (after centri-
fuging at 4000 rpm, 4 min, 4 Celsius) for later measurements of
Perinatal Psychoneuroimmunology: Protocols for the Study of Prenatal. . . 363
2.2.4 Heart Rate Analysis ECG-derived heart rate (HR) and HR variability (HRV) analyses of
as a Gateway to Measuring mother and the fetus may contain signatures of stress hidden in the
Stress Exposure subtle fluctuations of HRV on millisecond timescale. To reveal such
properties of HRV, a sampling rate of 1000 Hz is required [44, 45].
HRV classifies stressed versus non-stressed elephants [46]. The
mediating mechanism by which stress exposure alters HRV proper-
ties is thought be via an activation of autonomic nervous system
(ANS).
Corticosteroid administration during pregnancy has shown to
affect the autonomic balance in utero [47–50]. This effect is tran-
sient but repeated fetal administration of betamethasone alters
nervous system maturation. There was no association between
maternal cortisol and infant DNA methylation suggesting that the
effects of maternal depression may not be mediated directly by
glucocorticoids; instead, sympathetic nervous system activity, a
component of the fetal ANS, may be the mediating pathway [51].
Aside of the sympathetic nervous system, another key contrib-
utor to the ANS activity is the vagus nerve. Fetal HRV reflects
maturation and activation of the vagus branch of the ANS involved
in sensing and control of fetal acidemia, hypoxia, and inflammation
[52–54]. In human cohorts and in fetal sheep model of human
labor and fetal inflammation, multidimensional FHR variability
analysis predicts birth outcomes [45, 54–57].
The vagus nerve influences brain function and body metabo-
lism in a pleiotropic manner [58, 59]. A new field of bioelectronic
medicine is emerging. It aims to devise therapeutic approaches
using vagus nerve stimulation (VNS) to modify the endogenous
salutory signaling of the vagus nerve [60–62]. VNS reduced sym-
pathetic tone, stress-induced anxiety behaviors, and depression
symptoms in animal models and in clinical studies [63–69]. VNS
is thought to facilitate tonic inhibition of the basolateral amygdala
by the infralimbic region of the medial prefrontal cortex, which
results in reduced fear response [63]. VNS increases CRH expres-
sion in hypothalamus [70] and CRH receptor 1 agonism increases
vagal modulation of HRV [71–73]. This reciprocal CRH–vagus
nerve circuitry provides an important diagnostic and therapeutic
link between stress and ANS.
A more integrative dimension within which to gauge the phys-
iological role of HRV involves theories such as Porges’ polyvagal
theory [74, 75] and concepts whereby HRV is seen as a proxy
signature of nonlinearly coupled interacting homeostatic (or,
rather, homeokinetic [76]) physiological systems. Within such
364 Martin G. Frasch et al.
2.2.5 Prenatal Stress The general framework of the protocol has been published in [11]
Protocol in Pregnant Sheep and is freely accessible via the PubMed Central. This paper reports
the precise materials and steps needed for the pregnant sheep
model approach.
The stress behavioral component involves animal isolation, as
described, and is very straightforward. Two notes of caution are
worth sharing. First, as with all freely moving animal models, here it
is also important to ensure that all catheters and wires are properly
organized and secured around the animal to avoid accidental pull-
ing, chewing, or tearing of a catheter or wire.
For HRV analysis, we strongly recommend resisting the temp-
tation of using off-the-shelf software products without consulting
or, ideally, collaborating with HRV experts on the experimental
design, ECG data acquisition, and HRV analysis. Some papers fall
into this trap and produce data of doubtful validity. Two frequently
observed misconceptions should be stressed. First, HRV is not a
good source of biomarkers for sympathetic modulations of sinus
node activity. HRV is best suited to reflect more integrative, com-
plex fluctuations of ANS activity which can render excellent bio-
markers. Second, HRV does not reflect a sympathetic or vagal tone;
it may reflect fluctuations of vagal tone, but the complex mélange of
sympathetic and vagal rhythm fluctuations is difficult to untangle
through HRV alone, with some exceptions which require
specialized approach to HRV analysis not available in off-the-shelf
software [78].
VNS outside of epilepsy treatment is a nascent field with little
standardization, so a thorough literature search is advised prior to
deciding on a stimulation regimen. We refer the interested reader
to a recently published meta-analysis of the VNS approaches [62].
3.1 Overview Since epigenetic marks are potentially heritable, the use of transge-
nerational animal models is uniquely suited to reveal mechanisms of
inheritance of epigenetic, metabolomic, and phenotypic traits.
These models offer the unique opportunity to investigate chronic
Perinatal Psychoneuroimmunology: Protocols for the Study of Prenatal. . . 365
that provides them with toys and a diverse diet to generate rich
cognitive, social, motor, and sensory stimulation. EE was first
studied in the 1940s by the Canadian neuropsychologist Donald
O. Hebb who brought laboratory rats to his home. He noted that
rats housed as pets in his home performed better on memory tasks
compared with rats reared in standard laboratory conditions [94].
Moreover, when reared in an EE aging rats showed thicker frontal
and occipital cortices [95]. A recent study [96] expanded these
findings by showing that social enrichment has a significant impact
on neuroplasticity and stress resiliency particularly in females. Fur-
thermore, the benefits of maternal social enrichment could be
transmitted to their unexposed female descendants [96]. The
experiments raised two generations (F0 and F1) of male and female
rats in standard and social housing conditions which were examined
for neurohormonal and molecular alterations along with changes in
four behavioral modalities [96]. The findings revealed that in addi-
tion to higher neuronal density and thickness of the prefrontal
cortex, social experience reduced the hypothalamic-pituitary-adre-
nal (HPA) axis activity in F0 rats and their F1 nonsocial housing
offspring born to social mothers [96].
The trans- and multigenerational prenatal stress models in rats
as outlined above allow studies of epigenetic mechanisms at DNA
and miRNA levels as well as of downstream cell metabolic signa-
tures in organs such as brain and blood. The discovery of epigenetic
and metabolic signatures of ancestral stress may help identify the
mechanisms of disease programming. It is likely that adult complex
disease is programmed by transgenerational inheritance of adverse
life events, and associated epigenetic markers. Thus, finding epige-
netic linkages of inter- or transgenerational programming of disease
risk may lead to new prognostic markers and preventive strategies.
These strategies are central to the development of personalized,
preventive medicine based on epigenetic signatures.
3.2 Model Design Common rat models of transgenerational inheritance use gesta-
tional restraint stress during the second half of pregnancy. Behav-
3.2.1 Trans- and ioral and neuroendocrine assessments are usually performed in the
Multigenerational Stress parental generation (F0), their offspring (F1), grand-offspring
(F2), and great-grand-offspring (F3), for both males and females.
Truly heritable epigenetic marks can be found in the F2 generation
of the male lineage and in the F3 generation of the female lineage
[80, 97]. In this model, transgenerational stress refers to stress in
only the parental generation (S in F0, SN in F1, SNN in F2, SNNN
in F3; N: no stress), and multigenerational stress refers to continu-
ous experience of maternal stress through all generations (S in F0,
SS in F1, SSS in F2, SSSS in F3) [22, 98]. Controls will consist of
non-stressed, handled animals (N, NN, NNN, NNNN). To care-
fully standardize conditions across generations, personal breeding
colony room in a dedicated facility is advised.
368 Martin G. Frasch et al.
3.2.3 Behavior Through continuous cage-site recordings along with formal behav-
ioral assessments a wealth of behavioral data can be gained. Cage-
site observations may include prepartum maternal behavior, mater-
nal care (nest building, grooming, stereotypy, circadian rhythm,
pup retrieval), and maternal defense on lactational day (LD) 5
toward a virgin intruder female approaching the nest area. A com-
prehensive array of formal tests may assess offspring development
(P7 infantile, P21 juvenile) in sensorimotor and physical features
(skilled walking, vestibular and proprioceptive functions, body
weight, eye opening, play). Measures of maturation at P60 (adoles-
cence) and adult behavior (P90–18 months) may include five clas-
ses of behavior: (1) skilled movement (reaching, ladder rung
walking); (2) sensorimotor function (adhesive removal, von-Frey
hair test); (3) learning and memory (water maze, Ziggurat task,
Go-No Go decision-making); (4) social behavior (play fighting,
hierarchy, aggressive, defensive, and submissive behavior); and
(5) depression- and anxiety-like behavior (light/dark test, elevated
plus maze, open field, open table exploration).
References
1. van den Bergh BRH, van den Heuvel MI, mental health: the influence of maternal stress
Lahti M, Braeken M, de Rooij SR, in pregnancy. Neurosci Biobehav Rev
Entringer S, Hoyer D, Roseboom T, 2. Abbasi J (2017) The paternal epigenome
Raikkonen K, King S, Schwab M (2017) Pre- makes its mark. JAMA 317:2049–2051
natal developmental origins of behavior and 3. Mychasiuk R, Harker A, Ilnytskyy S, Gibb R
(2013) Paternal stress prior to conception
372 Martin G. Frasch et al.
alters DNA methylation and behaviour of prior and during pregnancy. Neurosci Biobe-
developing rat offspring. Neuroscience hav Rev
241:100–105 14. Barker DJ (1998) In utero programming of
4. Diz-Chaves Y, Pernia O, Carrero P, Garcia- chronic disease. Clin Sci (Lond) 95:115–128
Segura LM (2012) Prenatal stress causes 15. Gluckman PD, Hanson MA, Spencer HG
alterations in the morphology of microglia (2005) Predictive adaptive responses and
and the inflammatory response of the hippo- human evolution. Trends Ecol Evol
campus of adult female mice. J Neuroinflam- 20:527–533
mation 9:71 16. Gluckman PD, Hanson MA (2006) The
5. Diz-Chaves Y, Astiz M, Bellini MJ, Garcia- developmental origins of health and disease.
Segura LM (2013) Prenatal stress increases Early Life Origins Health Dis:1–7
the expression of proinflammatory cytokines 17. Van den Bergh BR (2011) Developmental
and exacerbates the inflammatory response to programming of early brain and behaviour
LPS in the hippocampal formation of adult development and mental health: a conceptual
male mice. Brain Behav Immun 28:196–206 framework. Dev Med Child Neurol 53(Suppl
6. Slusarczyk J, Trojan E, Glombik K, 4):19–23
Budziszewska B, Kubera M, Lason W, 18. Lahiri DK, Maloney B, Zawia NH (2009) The
Popiolek-Barczyk K, Mika J, Wedzony K, LEARn model: an epigenetic explanation for
Basta-Kaim A (2015) Prenatal stress is a vul- idiopathic neurobiological diseases. Mol Psy-
nerability factor for altered morphology and chiatry 14:992–1003
biological activity of microglia cells. Front
Cell Neurosci 9:82 19. Hertzman C (1999) The biological embed-
ding of early experience and its effects on
7. Boersma GJ, Tamashiro KL (2015) Individual health in adulthood. Ann N Y Acad Sci
differences in the effects of prenatal stress 896:85–95
exposure in rodents. Neurobiol Stress
1:100–108 20. Crews D, Gillette R, Scarpino SV,
Manikkam M, Savenkova MI, Skinner MK
8. Weinstock M (2017) Prenatal stressors in (2012) Epigenetic transgenerational inheri-
rodents: effects on behavior. Neurobiol Stress tance of altered stress responses. Proc Natl
6:3–13 Acad Sci U S A 109:9143–9148
9. Babenko O, Kovalchuk I, Metz GA (2015) 21. Porter TF, Fraser AM, Hunter CY, Ward RH,
Stress-induced perinatal and transgenera- Varner MW (1997) The risk of preterm birth
tional epigenetic programming of brain devel- across generations. Obstet Gynecol 90:63–67
opment and mental health. Neurosci
Biobehav Rev 48:70–91 22. Yao Y, Robinson AM, Zucchi FC, Robbins
JC, Babenko O, Kovalchuk O, Kovalchuk I,
10. Rakers F, Frauendorf V, Rupprecht S, Olson DM, Metz GA (2014) Ancestral expo-
Schiffner R, Bischoff SJ, Kiehntopf M, sure to stress epigenetically programs preterm
Reinhold P, Witte OW, Schubert H, Schwab birth risk and adverse maternal and newborn
M (2013) Effects of early- and late-gestational outcomes. BMC Med 12:121
maternal stress and synthetic glucocorticoid
on development of the fetal hypothalamus- 23. Christiaens I (2012) Chronic maternal stress
pituitary-adrenal axis in sheep. Stress and genetic variants in the etiology of sponta-
16:122–129 neous preterm birth. Department of Physiol-
ogy, University of Alberta, Edmonton, AB
11. Burns P, Liu HL, Kuthiala S, Fecteau G,
Desrochers A, Durosier LD, Cao M, Frasch 24. Christiaens I, Hegadoren K, Olson DM
MG (2015) Instrumentation of near-term (2012) Adverse childhood experiences are
fetal sheep for multivariate chronic associated with spontaneous preterm birth: a
non-anesthetized recordings. J Vis Exp 105: case-control study. Reprod Sci 19:183A
e52581 25. Shapiro GD, Fraser WD, Frasch MG, Seguin
12. Rakers F, Rupprecht S, Dreiling M, JR (2013) Psychosocial stress in pregnancy
Bergmeier C, Witte OW, Schwab M (2017) and preterm birth: associations and mechan-
Transfer of maternal psychosocial stress to the isms. J Perinat Med 41:631–645
fetus. Neurosci Biobehav Rev 26. Christian LM (2015) Stress and immune
13. Braun K, Bock J, Wainstock T, Matas E, function during pregnancy: an emerging
Gaisler-Salomon I, Fegert J, Ziegenhain U, focus in mind-body medicine. Curr Dir
Segal M (2017) Experience-induced transge- Psychol Sci 24:3–9
nerational (re-)programming of neuronal 27. Raju TNK, Buist AS, Blaisdell CJ, Moxey-
structure and functions: impact of stress Mims M, Saigal S (2017) Adults born pre-
term: a review of general health and system-
Perinatal Psychoneuroimmunology: Protocols for the Study of Prenatal. . . 373
specific outcomes. Acta Paediatr 38. Monteleone MC, Adrover E, Pallares ME,
106:1409–1437 Antonelli MC, Frasch AC, Brocco MA
28. Adrover E, Berger MA, Perez AA, Tarazi FI, (2014) Prenatal stress changes the glycopro-
Antonelli MC (2007) Effects of prenatal stress tein GPM6A gene expression and induces epi-
on dopamine D2 receptor asymmetry in rat genetic changes in rat offspring brain.
brain. Synapse 61:459–462 Epigenetics 9:152–160
29. Berger MA, Barros VG, Sarchi MI, Tarazi FI, 39. Baier CJ, Pallares ME, Adrover E, Katunar
Antonelli MC (2002) Long-term effects of MR, Raisman-Vozari R, Antonelli MC
prenatal stress on dopamine and glutamate (2014) Intrastriatal 6-OHDA lesion differen-
receptors in adult rat brain. Neurochem Res tially affects dopaminergic neurons in the ven-
27:1525–1533 tral tegmental area of prenatally stressed rats.
30. Barros VG, Berger MA, Martijena ID, Sarchi Neurotox Res 26:274–284
MI, Perez AA, Molina VA, Tarazi FI, Anto- 40. Baier CJ, Katunar MR, Adrover E, Pallares
nelli MC (2004) Early adoption modifies the ME, Antonelli MC (2012) Gestational
effects of prenatal stress on dopamine and restraint stress and the developing dopaminer-
glutamate receptors in adult rat brain. J Neu- gic system: an overview. Neurotox Res
rosci Res 76:488–496 22:16–32
31. Katunar MR, Saez T, Brusco A, Antonelli MC 41. Debeir T, Ginestet L, Francois C, Laurens S,
(2010) Ontogenetic expression of dopamine- Martel JC, Chopin P, Marien M, Colpaert F,
related transcription factors and tyrosine Raisman-Vozari R (2005) Effect of intrastria-
hydroxylase in prenatally stressed rats. Neuro- tal 6-OHDA lesion on dopaminergic innerva-
tox Res 18:69–81 tion of the rat cortex and globus pallidus. Exp
32. Carboni E, Barros VG, Ibba M, Silvagni A, Neurol 193:444–454
Mura C, Antonelli MC (2010) Prenatal 42. Rakers F, Bischoff S, Schiffner R, Haase M,
restraint stress: an in vivo microdialysis study Rupprecht S, Kiehntopf M, Kuhn-Velten
on catecholamine release in the rat prefrontal WN, Schubert H, Witte OW, Nijland MJ,
cortex. Neuroscience 168:156–166 Nathanielsz PW, Schwab M (2015) Role of
33. Silvagni A, Barros VG, Mura C, Antonelli catecholamines in maternal-fetal stress trans-
MC, Carboni E (2008) Prenatal restraint fer in sheep. Am J Obstet Gynecol 213(684):
stress differentially modifies basal and stimu- e681–e689
lated dopamine and noradrenaline release in 43. Dreiling M, Bischoff S, Schiffner R,
the nucleus accumbens shell: an ‘in vivo’ Rupprecht S, Kiehntopf M, Schubert H,
microdialysis study in adolescent and young Witte OW, Nathanielsz PW, Schwab M,
adult rats. Eur J Neurosci 28:744–758 Rakers F (2016) Stress-induced decrease of
34. Pallares ME, Baier CJ, Adrover E, Monte- uterine blood flow in sheep is mediated by
leone MC, Brocco MA, Antonelli MC alpha 1-adrenergic receptors. Stress
(2013) Age-dependent effects of prenatal 19:547–551
stress on the corticolimbic dopaminergic sys- 44. Durosier LD, Green G, Batkin I, Seely AJ,
tem development in the rat male offspring. Ross MG, Richardson BS, Frasch MG
Neurochem Res 38:2323–2335 (2014) Sampling rate of heart rate variability
35. Pallares ME, Adrover E, Baier CJ, Bourgui- impacts the ability to detect acidemia in ovine
gnon NS, Monteleone MC, Brocco MA, fetuses near-term. Front Pediatr 2:38
Gonzalez-Calvar SI, Antonelli MC (2013) 45. Li X, Xu Y, Herry C, Durosier LD, Casati D,
Prenatal maternal restraint stress exposure Stampalija T, Maisonneuve E, Seely AJ,
alters the reproductive hormone profile and Audibert F, Alfirevic Z, Ferrazzi E, Wang X,
testis development of the rat male offspring. Frasch MG (2015) Sampling frequency of
Stress 16:429–440 fetal heart rate impacts the ability to predict
36. Barros VG, Duhalde-Vega M, Caltana L, pH and BE at birth: a retrospective multi-
Brusco A, Antonelli MC (2006) Astrocyte- cohort study. Physiol Meas 36:L1–L12
neuron vulnerability to prenatal stress in the 46. Vezina-Audette R, Herry C, Burns P,
adult rat brain. J Neurosci Res 83:787–800 Frasch M, Chave E, Theoret C (2016) Heart
37. Adrover E, Pallarés ME, Baier CJ, Monte- rate variability in relation to stress in the Asian
leone MC, Giuliani FA, Waagepetersen HS, elephant (Elephas maximus). Can Vet J
Brocco MA, Cabrera R, Sonnewald U, 57:289–292
Schousboe A (2015) Glutamate neurotrans- 47. Dawes GS, Serra-Serra V, Moulden M, Red-
mission is affected in prenatally stressed off- man CW (1994) Dexamethasone and fetal
spring. Neurochem Int 88:73–87 heart rate variation. Br J Obstet Gynaecol
101:675–679
374 Martin G. Frasch et al.
48. Mulder EJ, Derks JB, Visser GH (1997) 58. Pavlov VA, Tracey KJ (2012) The vagus nerve
Antenatal corticosteroid therapy and fetal and the inflammatory reflex—linking immu-
behaviour: a randomised study of the effects nity and metabolism. Nat Rev Endocrinol
of betamethasone and dexamethasone. Br J 8:743–754
Obstet Gynaecol 104:1239–1247 59. Pavlov VA, Tracey KJ (2015) Neural circuitry
49. Senat MV, Minoui S, Multon O, and immunity. Immunol Res 63:38–57
Fernandez H, Frydman R, Ville Y (1998) 60. Pavlov VA, Tracey KJ (2017) Neural regula-
Effect of dexamethasone and betamethasone tion of immunity: molecular mechanisms and
on fetal heart rate variability in preterm clinical translation. Nat Neurosci 20:156–166
labour: a randomised study. Br J Obstet 61. Borovikova LV, Ivanova S, Zhang M, Yang H,
Gynaecol 105:749–755 Botchkina GI, Watkins LR, Wang H,
50. Derks JB, Mulder EJ, Visser GH (1995) The Abumrad N, Eaton JW, Tracey KJ (2000)
effects of maternal betamethasone administra- Vagus nerve stimulation attenuates the sys-
tion on the fetus. Br J Obstet Gynaecol temic inflammatory response to endotoxin.
102:40–46 Nature 405:458–462
51. Braithwaite EC, Kundakovic M, Ramchan- 62. Kwan H, Garzoni L, Liu HL, Cao M,
dani PG, Murphy SE, Champagne FA Desrochers A, Fecteau G, Burns P, Frasch
(2015) Maternal prenatal depressive symp- MG (2016) VNS in inflammation: systematic
toms predict infant NR3C1 1F and BDNF review of animal models and clinical studies.
IV DNA methylation. Epigenetics Bioelectron Med 3:1–6
10:408–417 63. Caliskan G, Albrecht A (2013) Noradrenergic
52. Frasch MG, Muller T, Wicher C, Weiss C, interactions via autonomic nervous system: a
Lohle M, Schwab K, Schubert H, Nathanielsz promising target for extinction-based expo-
PW, Witte OW, Schwab M (2007) Fetal body sure therapy? J Neurophysiol 110:2507–2510
weight and the development of the control of 64. Clancy JA, Mary DA, Witte KK, Greenwood
the cardiovascular system in fetal sheep. J JP, Deuchars SA, Deuchars J (2014)
Physiol 579:893–907 Non-invasive vagus nerve stimulation in
53. Frasch MG, Muller T, Weiss C, Schwab K, healthy humans reduces sympathetic nerve
Schubert H, Schwab M (2009) Heart rate activity. Brain Stimul 7:871–877
variability analysis allows early asphyxia detec- 65. George MS, Ward HE Jr, Ninan PT,
tion in ovine fetus. Reprod Sci 16:509–517 Pollack M, Nahas Z, Anderson B, Kose S,
54. Durosier LD, Xu A, Matushewski B, Cao M, Howland RH, Goodman WK, Ballenger JC
Herry C, Batkin I, Seely A, Ross M, Richard- (2008) A pilot study of vagus nerve stimula-
son BS, Frasch MG (2013) Neural signature tion (VNS) for treatment-resistant anxiety
of cerebral activity of the fetal cholinergic anti- disorders. Brain Stimul 1:112–121
inflammatory pathway derived from heart rate 66. Liu RP, Fang JL, Rong PJ, Zhao Y, Meng H,
variability. FASEB J 27 Ben H, Li L, Huang ZX, Li X, Ma YG, Zhu B
(1 Supplement):926–911 (2013) Effects of electroacupuncture at auric-
55. Liu D, Diorio J, Tannenbaum B, Caldji C, ular concha region on the depressive status of
Francis D, Freedman A, Sharma S, unpredictable chronic mild stress rat models.
Pearson D, Plotsky PM, Meaney MJ (1997) Evid Based Complement Alternat Med
Maternal care, hippocampal glucocorticoid 2013:789674
receptors, and hypothalamic-pituitary-adrenal 67. O’Keane V, Dinan TG, Scott L, Corcoran C
responses to stress. Science 277:1659–1662 (2005) Changes in hypothalamic-pituitary-
56. Durosier LD, Herry CL, Cortes M, Cao M, adrenal axis measures after vagus nerve stimu-
Burns P, Desrochers A, Fecteau G, Seely AJ, lation therapy in chronic depression. Biol Psy-
Frasch MG (2015) Does heart rate variability chiatry 58:963–968
reflect the systemic inflammatory response in 68. Pena DF, Childs JE, Willett S, Vital A, McIn-
a fetal sheep model of lipopolysaccharide- tyre CK, Kroener S (2014) Vagus nerve stim-
induced sepsis? Physiol Meas 36:2089–2102 ulation enhances extinction of conditioned
57. Lun Liu H, Garzoni L, Herry C, Durosier fear and modulates plasticity in the pathway
LD, Cao M, Burns P, Fecteau G, from the ventromedial prefrontal cortex to
Desrochers A, Patey N, Seely AJ, Faure C, the amygdala. Front Behav Neurosci 8:327
Frasch MG (2016) Can monitoring fetal 69. Ylikoski J, Lehtimaki J, Pirvola U, Makitie A,
intestinal inflammation using heart rate varia- Aarnisalo A, Hyvarinen P, Ylikoski M (2017)
bility analysis signal incipient necrotizing Non-invasive vagus nerve stimulation reduces
enterocolitis of the neonate? Pediatr Crit
Care Med 17(4):e165–e176
Perinatal Psychoneuroimmunology: Protocols for the Study of Prenatal. . . 375
effects of paternal stress. Proc Natl Acad Sci U 97. Skinner MK (2008) What is an epigenetic
S A 112:13699–13704 transgenerational phenotype?: F3 or F2.
94. Hebb DO (1947) The effect of early experi- Reprod Toxicol 25:2–6
ence on problem solving at maturity. Am 98. Ambeskovic M, Soltanpour N, Falkenberg
Psychol 2:737–745 EA, Zucchi FCR, Kolb B, Metz GAS (2017)
95. Diamond MC, Johnson RE, Protti AM, Ancestral exposure to stress generates new
Ott C, Kajisa L (1985) Plasticity in the behavioral traits and a functional hemispheric
904-day-old male rat cerebral cortex. Exp dominance shift. Cereb Cortex
Neurol 87:309–317 27:2126–2138
96. Faraji J, Soltanpour N, Lotfi H, Nejad- 99. Whishaw IQ, Kolb B (2004) The behavior of
Ghorban F, Moharrerie A, Roudaki S, Hos- the laboratory rat: a handbook with tests.
seini SA, Abdollahi AA, Soltanpour N, Oxford University Press, New York
Moeeini R, Metz GA S (Submitted) Social 100. Heijmans BT, Tobi EW, Stein AD, Putter H,
experience promotes brain health and stress Blauw GJ, Susser ES, Slagboom PE, Lumey
resiliency in mothers and their unexposed LH (2008) Persistent epigenetic differences
offspring associated with prenatal exposure to famine
in humans. Proc Natl Acad Sci U S A
105:17046–17049
INDEX
A Autophagosomes............................................................. 14
Autophagy ...................................................................7, 15
Adaptation ..................................................... v, 3–5, 8, 13,
14, 37, 38, 41, 46–49, 51, 56, 58, 59, 78, 79, 82, B
158, 201, 239, 240, 260, 261, 319, 359, 370
Aerobic exercise........................................... 154, 210, 211 Balance...................................................................... v, 4, 5,
Aging ........................................................ 4–11, 174, 181, 7–9, 11–17, 77, 79–81, 83, 84, 94, 118, 129, 130,
224, 250, 367, 370 145, 158, 177, 188, 357, 363
Allergic diseases ............................................................. 185 Batch PLS ...................................................................... 107
Allostasis ............................................................... 260, 261 Behavioral .................................................................. vi, 22,
Alzheimer’s disease ................................................ v, 6, 10, 64, 65, 69, 82, 88–92, 122, 126, 127, 130, 134,
17, 49, 186, 288, 301 136, 138, 154, 165, 205, 221–229, 231–252,
Amygdala ...............................................80, 363, 368, 371 263, 264, 267–271, 273–276, 356, 359, 368
Anhedonia .................................... 82, 239, 240, 297, 356 Bioactives ....................................................................... 222
Animal model ...................................................... 122, 150, Biobehaviors .................................................................. 252
156, 197, 223, 261, 262, 301, 342, 353, 356, Biological response........................................................ 172
364, 370 Biomarkers............................................................ v, 4, 7–9,
Anorexia................................................................ 228, 231 11–13, 16–18, 58, 65, 70, 78–80, 83, 102, 103,
Anthropology .............................................................. v, 55 106, 110–112, 155, 175, 177, 181–186, 357,
Antibiotic ...................................... 25, 26, 44, 46, 49, 344 364, 370
Antibody .............................................................. 176, 185, Biopsychosocial ..............................................................v, 4
201, 269, 310, 311, 313–323, 328–336 Blood-brain barrier (BBB)...........................................263,
Antidepressant ......................................................... 81, 83, 266, 269, 272, 273, 295–298
140, 154, 238 B-lymphocytes ...................................................... 309, 311
Antigen ............................................................15, 62, 117, Body mass index (BMI).................................59, 106, 112
176, 185, 210, 262, 268, 269, 310, 311, Brain........................................................................ v, 4, 22,
317–319, 322, 327–329, 331–334, 336, 338 41, 56, 79, 89, 122, 163, 172, 195, 222,
Anxiety .................................................................... v, 4, 18, 259–263, 265–270, 272–276, 291, 341, 353
46, 77, 78, 89, 90, 92, 223, 227, 228, 233–240,
C
245, 262, 263, 268, 275, 315–317, 356, 363, 368
Apoptosis ............................................................. 7, 10, 15, Cancer.................................................................. v, vi, 5, 7,
79, 149, 187, 263, 264, 272, 274 9, 11, 14, 17, 28, 49, 50, 70, 172, 174, 181, 183,
Astrocytes .................................7, 16, 298, 301, 342, 345 184, 186–188, 222, 315
Atherosclerosis .........................6, 9, 13, 50, 62, 146, 186 Cannabinoid system ............................................. 179, 181
Attention ..............................................................v, 5, 6, 8, Carcinogenesis................................................................. 15
89, 90, 94, 124, 125, 132–134, 165, 248, 354 Cardiovascular diseases ............................v, 8, 13, 17, 354
Attenuation.......................................................... 196, 200, Caregiving ................................................... 311, 314, 315
201, 274, 289, 291, 293 Catecholamine......................................... 24–27, 149, 183
Autoantigens ........................................................ 260, 261 Central nervous system (CNS)............................... 13, 16,
Autoimmune diseases ................................................. vi, 5, 25, 82, 89, 125, 163, 260, 262–264, 266–274,
181, 183–186, 188, 189, 260, 261 276
Autoimmunity ............................................................7, 15, Chemogenetics................................................ vi, 195–207
28, 48, 51, 172, 174, 181, 183–186, 259–263, Chemokines.......................................................... 172, 259
265–276 Chemotaxis.....................................................13, 157–161
Autonomic nervous system (ANS) ............................... 25, Cholinergic signaling ...................................... vi, 341–350
47, 260, 297, 342, 363 Chromium release assay ........................................... vi, 209
Qing Yan (ed.), Psychoneuroimmunology: Methods and Protocols, Methods in Molecular Biology, vol. 1781,
https://doi.org/10.1007/978-1-4939-7828-1, © Springer Science+Business Media LLC, part of Springer Nature 2018
377
PSYCHONEUROIMMUNOLOGY: METHODS AND PROTOCOLS
378 Index
Chronic experimentation...............................81, 227, 249 Endotoxin................................................... 163, 164, 297,
Circadian rhythms .............................. vi, 4, 174, 177, 179 298, 300, 301, 303, 304, 343
Co-expression patterns ................................................. 103 Environment.........................................................v, 16, 22,
Cognition ....................................... 58, 77, 222, 240, 297 29, 38–40, 44, 48, 51, 56, 58, 59, 61, 62, 92, 97,
Cognitive function ...........................................8, 240, 248 126, 133, 201–203, 223–225, 227, 235, 250,
Complement........................................................ 6, 10, 67, 252, 259, 261, 263, 356, 359, 366, 370
259, 266, 272, 328 Enzyme-linked immunosorbent assay
Complex adaptive systems (CAS) ...................... v, 3–7, 78 (ELISA).................................................... 103, 104,
Computational modeling ...................................... 96, 140 156, 329–331, 336–338, 347, 363
Coronary artery disease (CAD) ..................................... 13 Epidemiology ....................................................... 146, 151
Cortisol ................................................................ 8, 25, 47, Epilepsy................................................................. 341, 364
59, 61, 64, 66–68, 71, 88, 89, 92, 94, 96, 97, 102, Epinephrine (E)....................................................... 24, 25,
104, 107–109, 111, 115, 117, 124, 125, 130, 89, 92, 94, 124, 125, 132, 133, 138, 140
132–134, 138–140, 173, 174, 177, 179, 181, Escherichia coli ......................................................... 24, 25,
183, 260, 298, 362, 363 27, 28, 304, 346
C-reactive protein (CRP) ..........................................8, 12, Evolution ..................................................................45–48,
59, 62, 66, 67, 81, 156, 157, 175, 177 50, 58, 92, 112, 113, 126, 173, 223
Culture..................................................................... 22, 23, Exercise response ................................................. 103, 109
42, 51, 56, 57, 62, 69, 102, 103, 107–109, 117, Experience ............................................................. v, vi, 22,
212–215, 218, 219, 342, 345–347, 349 39–44, 48, 50, 62, 64, 65, 70, 121, 154, 155, 165,
Cyclooxygenase-2 (COX-2) ......................................... 301 228, 239, 262, 315, 344, 350, 353, 355, 365–367
Cytokine network................................................ 177, 181, Exploration .............................................................. 39, 58,
183–186, 189, 269 70, 88, 97, 228, 229, 231, 232, 234–239, 241,
Cytokine profiles .................................................. 117, 328 243–247, 263, 268, 338, 368, 370
Cytokines ............................................................... 4, 6–10,
12, 16, 47, 66, 78–80, 83, 101, 102, 106, 107, F
109, 117, 118, 145, 150, 156–158, 163, 164, Fatigue ..................................................................... 61, 79,
174–177, 180, 181, 184–186, 188, 189, 222,
82, 90, 92, 94, 101, 103, 114, 125, 132–134,
228, 259, 263, 266, 269, 271, 276, 297, 301, 148, 154–156, 210, 213, 218, 222, 228, 231,
311, 347 249, 251, 297, 300
Fear ........................................................................ 66, 236,
D
243, 244, 363
Dementia ....................................272, 311, 314, 315, 322 Feedback ................................................................... v, 4, 5,
Depression .............................................................. v, 4, 57, 18, 89, 91, 92, 94, 96, 97, 117, 122, 123, 138
77, 88, 121–141, 145, 174, 222, 261, 297, 315, Flow cytometry ................................................... 199, 212,
363 216, 333, 334, 346
Developmental Origin of Health and Disease Fluorine-18-labeled 2-deoxy-D-glucose
(DOHaD) ................................................. 354, 370 (FDG) ....................................................... 298–300
Diabetes ..................................................... v, 4–6, 8, 9, 11, Food consumption........................................................ 231
17, 18, 29, 46, 49, 50, 62, 66, 148, 365 Functional MR imaging (fMRI) ......................... 298, 299
Diagnostic classification ......................103, 110, 113, 363
Dimension .........................................................39, 43, 44, G
51, 87, 104, 356, 358, 363 Gamma-aminobutyric acid (GABA) .........................7, 14,
DNA methylation ........................................................... 48
89, 90, 92, 94, 122–125, 130, 132–134, 136, 138
Dopamine6, 11, 25, 78, 82, 88, 89, 94, 96, 97, 124, 296, Glutamate ..................................................... 7, 14, 89, 94,
303, 355 124, 125, 132, 134, 135, 138, 139, 272, 356
Drug-resistant ............................................................... 341
Glutathione ...............................................................11, 83
Dynamical medicine............................................. 3, 15, 18 G protein-coupled receptor(GPCR) ..................... 24, 196
Gulf war illness (GWI)............................. v, 101–118, 210
E
Gut microbiota and metabolites ......................... 162–164
Ecoimmunology.............................................................. 57
Emotional disturbances ................................................ 272 H
Endocrinology................................. v, 22, 23, 29, 58, 366
Heart failure (HF)............................vi, 14, 145, 148, 157
PSYCHONEUROIMMUNOLOGY : METHODS AND PROTOCOLS
Index 379
Heart-gut axis................................................................ 162 Interleukin-6 (IL-6).......................................... 13, 61–63,
Heart rate (HR) .........................................................4, 12, 66, 67, 79, 102, 103, 107, 109, 111, 117, 150,
13, 45, 79, 149, 151, 342, 363 155–157, 163, 175–177, 180, 181, 184, 185,
Hemagglutination inhibition assay (HAI)......... 330–333, 188, 269, 272, 297, 311
338, 339 Interventions .....................................................14, 80, 81,
Hepatitis B............................................................. 59, 313, 88, 89, 123, 125, 129, 130, 136–139, 154, 288,
316–318, 321, 322 296, 298, 299, 320, 322, 323, 327, 366
Hippocampus ........................................................ 83, 267,
272, 356, 368, 371 K
Homeostasisv, 4, 5, 10, 12–14, 24, 77, 82, 94, 122, 126, K562 target cells ........................................................... 214
132, 158, 259–261, 354 Kynurenine ......................................................... 78, 80, 81
Homeostatic regulation .................................90, 135, 139
Hormones...................................................................8, 13, L
16, 22, 23, 25, 26, 47, 58, 59, 65, 67, 68, 71, 81,
88, 89, 101, 105, 118, 123, 124, 130, 132, 138, Learning............................................................... 8, 9, 222,
139, 172, 173, 177, 180, 181, 261, 267, 269, 228, 233, 235, 239, 240, 243, 244, 246–248,
271, 272, 276, 320, 322, 355, 356, 362, 368, 370 263, 268, 273, 274, 368, 370
Host-microbiota.............................................................. 46 Life history theory .......................................................... 56
Human biological variation ............................................ 56 Life style.................................................................. 18, 137
Human Microbiome Project (HMP)............................. 23 Light cycle ................................................. 5, 12, 225, 234
Humidity .............................................................. 212, 225 Locomotor activity.............................................. 228, 229,
Hypothalamic-pituitary-adrenal (HPA) axis............ vi, 14, 235, 236, 244, 268, 274
18, 47, 61, 77, 81, 94, 117, 138, 155, 173, 175, Lupus ................................................................... 184, 261,
180, 181, 183, 226, 272, 366–368 263, 264, 266, 267, 269–276
Lymphocytes ........................................................... 24, 63,
I 107, 117, 158, 172, 173, 175–177, 179, 181,
183, 184, 186–189, 209, 211, 216, 217, 263,
Imbalance .................................................................v, vi, 3,
268, 310
6–9, 11–16, 18, 28, 77–83, 107, 109, 122, 356 Lymphocyte-to-monocyte ratio ..................................... 45
Immune ....................................................................v, vi, 8,
21, 37, 55, 79, 88, 101, 139, 148, 172, 173, 195, M
209, 221, 259, 288, 309, 327, 341, 364
Immunity ............................................................ vi, 22, 29, Macrophage............................................................. 10, 25,
49, 51, 55, 59, 67, 78, 79, 157, 158, 176, 184, 172, 173, 175–177, 179, 182, 184, 186–189,
187, 195, 222, 309, 311, 315, 318, 323, 328 266, 273, 301, 328, 342
Immunoactivation................................................ 177, 181 Macular degeneration ..................................................... 11
Immunoassays ......................................................... 58, 68, Magnetic resonance (MR) ..................294, 298, 368, 371
104, 328, 333, 338 Major depressive disorder (MDD)..........................80–83,
Immunobehaviors ............................................... 222, 224, 122, 140, 155
225, 227, 234, 250 Maze .................................................................... 224, 233,
Immunocompetence ....................................................... 47 235–237, 241, 244–248, 263, 356, 368, 370
Immunosenescence ....................................................... 315 Melatonin ............................................................. 172, 180
Immunosuppression ........................................... 172, 173, Memory ......................................................................9, 89,
177, 181, 183, 270 90, 124, 125, 163, 176, 222, 228, 232, 240–248,
Immunotherapy ............................................................ 187 268, 274, 311, 367, 368, 370
Infectious diseases ......................................................7, 38, Metabolic syndrome (MetS) ............................... 6, 11, 28
47, 49, 50, 57, 60, 69, 310, 311, 323, 328 Microarray ....................................................334, 336–338
Inflammation ..................................................... v, vi, 3–18, Microbiome ........................................................ 23, 28, 46
50, 59, 61, 62, 66, 67, 71, 77, 81, 117, 145, 150, Microbiota ................................................................v, vi, 4,
154, 155, 157, 158, 163, 164, 249, 261, 266, 21, 24, 30, 46, 162–164
268, 270–272, 274, 296–301, 304, 348, 363 Microbiota-gut-brain (MGB) axis .............................4, 18
Influenza ................................................... 49, 59, 66, 297, Microglia.......................................................................6, 7,
311, 313–315, 317–322, 328, 331, 333, 338 10, 16, 266, 272, 276, 301, 303, 342, 343,
Innate immunity ............................................................. 67 345–349
Insomnia .............................................................. v, 5–7, 17 Mind-body............................................................... v, 3, 77
PSYCHONEUROIMMUNOLOGY: METHODS AND PROTOCOLS
380 Index
miRNAs .................................................................. 16, 366 O
Mitochondria...................................................... 13, 14, 45
Molecular imaging ........................................................ 288 Obesity......................................................... v, 6, 7, 11–13,
Monte Carlo algorithm.......................................... 92, 126 17, 18, 28, 46, 63, 78, 80, 181, 222
Mood .................................................................55, 61, 62, Odors ...................................................226, 228, 232, 252
65, 66, 71, 79, 80, 89, 90, 92, 94, 103, 121, 132, Opioid system.............................................. 172, 179, 183
134, 136, 138, 145, 272, 297, 321 Opsins ..................................................196, 197, 202, 205
Motor activity.................................................................. 46 Opsonization ................................................................. 328
Mouse ......................................................................vi, 197, Optogenetics ................................................... vi, 195, 206
202, 205, 221–229, 231–252, 262–264, 272 Oxidant ............................................................... 15, 79, 83
Multi-hit ...................................................... 342, 345, 366
P
Multiplex ......................................................... vi, 327–339
Multi-stability ....................................................... 122, 133 Parkinson’s disease (PD) ................................................ 11
Multivariate data acquisition ........................................ 345 Partial least squares (PLS) .................................. 104, 105,
Myocardium ................................................ 147, 149, 150 107, 109–111, 116
Peripheral nervous system (PNS)........... 22, 89, 125, 262
N Personality ............................................................. 5, 6, 77,
Natural killer cell cytotoxicity (NKCC) ............. 209–212, 122, 314, 321, 322
215–218 Phagocytosis .................................................................. 328
Natural killer cells................................................. 209–219 Photons................................................288, 289, 291–294
Natural selection ................................................ 47–51, 57 Physiology .............................................................vi, 8, 11,
Nerve stimulation ...............................341, 343–345, 363 22, 25, 41, 45, 48, 50, 58, 59, 61, 67, 88, 92, 94,
Neurodegeneration ......................................................6, 9, 97, 103, 105, 122–124, 174, 188, 224–227, 269,
267, 269, 270, 272 342, 343, 353, 354, 363, 365, 366, 368, 370
Neurodegenerative diseases ........... 6, 8, 9, 186, 207, 222 Pineal gland .........................................172, 173, 179–181
Neuroendocrine .........................................................8, 15, Populations......................................................... vi, 24, 25,
21–30, 81, 82, 101, 102, 105, 106, 112, 118, 42, 50, 57–63, 66, 105, 129, 145, 146, 151, 153,
172–174, 177, 209, 260, 261, 276, 364, 365, 163, 176, 181, 197, 288, 310, 314, 319, 321,
368–370 322, 327, 346, 356, 365
Neuroimmune ..................................................... v, vi, 4–6, Positive psychology ...................................................87, 94
16, 17, 21–30, 77–84, 145–165, 171–189, Positron emission tomography (PET) ............................vi,
221–226, 231, 233, 239–241, 243–245, 248, 287–301, 303, 304
250, 345, 353 Prenatal stress (PS)..................................................vi, 342,
Neuroimmunology ............................................. 195, 196, 353–370
342, 346, 350 Prevention ....................................................................4, 7,
Neuroimmunomodulation (NIM)............................... 172 14, 15, 79, 83
Neuroinflammation................................................. 79, 83, Proinflammatory ........................................................8–10,
266, 272, 288, 299, 301, 343, 345 13, 14, 16, 78–81, 83, 145, 150, 156,
Neuropeptide Y (NPY) 79, 89, 102, 104, 107–109, 115, 157, 164, 272
124, 125, 132, 138, 181 Pseudomonas aeruginosa ................................................. 25
Neuroprotection ........................................................... 6, 9 Psoriasis............................................................... 7, 16, 184
Neuroscience .......................................... vi, 195, 196, 299 Psychiatric............................................................... v, 7, 13,
Neurotransmitters ......................................................4, 24, 78–80, 82–84, 174, 261, 276, 298, 355
25, 28, 78–80, 82, 83, 88, 89, 92, 101, 105, Psychology....................................................................... 55
122–126, 130, 132, 134, 138, 139, 165, 172, Psychoneuroimmunology (PNI).............................v, vi, 3,
222, 273, 276, 296, 299, 355 11, 13, 16, 17, 21, 22, 37–41, 55–71, 77, 83,
Neutralization ................................................................... 4 154–165, 209, 221–229, 231–252, 287–301,
Noise .................................................................... 226, 227, 303, 304, 309–311, 313–323, 327, 328, 333,
235, 252, 292, 296, 301, 303 338, 342, 343, 350, 353
Nonlinearity.......................................................... 4, 5, 363
Q
Normalization ............................................. 250, 293, 348
Nuclear imaging ................................................... 287, 288 Quality control ....................................103, 217, 336, 347
Quality of life............................................... 103, 145, 153
PSYCHONEUROIMMUNOLOGY : METHODS AND PROTOCOLS
Index 381
R Stroke......................................................... 7, 14, 149, 301
Surgery........................................................ 197, 200–202,
Redox............................................ 6, 7, 16, 18, 78, 79, 83 344, 356
Regression model ................................................. 111, 157 Sympathetic nervous system (SNS) ....................... 22, 24,
Regulatory logic .......................................... 123, 130, 134 147, 150, 155, 363
Regulatory metasystem........................................ 259–261 Synchronization ............................................................ 174
Renin-Angiotensin-Aldosterone System Systemic diseases .............................................. v, 172, 274
(RAAS)...................................................... 147, 149 Systemic lupus erythematosus (SLE)..........................184,
Reverse engineering .................................... 88, 89, 92, 94 185, 260–263, 265–269, 271, 274, 276
RNAseq........................................................ 347, 348, 350 Systems biology....................................... v, vi, 3, 9, 17, 77
Robustness................................................... 4, 5, 130, 338 Systems medicine .......................................v, 3, 15, 18, 84
RT-PCR ......................................................................... 348
T
S
T cells .................................................................51, 59, 78,
Salmonella typhimurium................................................. 25 80, 82, 158, 172, 173, 187, 210, 266, 273, 276,
Schizophrenia ..................................................... 78, 82, 83 311, 318
Selective serotonin re-uptake inhibitor Temperature ............................................................ 62, 66,
(SSRI) .................... 122, 123, 130, 134, 136–140 68, 200, 215, 225, 247, 356, 358
Self-organization ........................................................... 4, 5 Th1/Th2 ratio .............................................................. 158
Self-regulatory ................................................................... 4 Thymus .......................................172, 173, 263, 311, 318
Serotonin ................................................................. 80, 82, Toll-like receptors (TLR) ............................................... 15
89, 90, 92, 94, 122, 124, 125, 130, 132–134, Transcriptome .......................................79, 343, 347, 348
138, 139 Translocator protein (TSPO) .............299, 301, 303, 304
Sickness behavior................................................. 117, 222, Trauma.................................................................... 45, 301
228, 229, 231–233, 248, 260, 261, 297 Tumor ................................................................3, 4, 8, 10,
Signaling .............................................................. v, vi, 7, 9, 16, 150, 173, 174, 177, 183, 184, 187, 188, 209,
11, 12, 14–16, 18, 26, 59, 64, 78, 79, 82, 88, 226, 228, 297
90–92, 103, 107, 125–127, 130, 132, 135, 136,
157, 196, 222, 226, 263, 275, 341, 363 U
Single nucleotide polymorphisms (SNP).................59, 61
Skin7, 16–18, 23, 25, 29, 45, 64, 68, 69, 202, 203, 206, Umwelt .........................................................40–44, 47, 48
264, 344, 359, 360
V
Sleep......... 5–8, 12, 17, 79, 81, 121, 155, 222, 231, 297
Social support......................................313, 315, 318–322 Vaccination ................................................................ vi, 62,
Spatiotemporal ....................................4, 5, 9, 43, 77, 342 66, 304, 309–311, 313–323, 327
Stress ...................................................................v, 6, 7, 21, Vaccine ................................................................ vi, 44, 47,
46, 55, 78, 87, 122, 162, 181, 195, 210, 225, 49, 59, 66, 69, 310, 311, 313–320, 327–339
267, 309, 342, 353 Vagotomy ............................................................. 343, 344
Stress management..................................................vi, 123, Vagus nerve .................................. 46, 298, 341–344, 363
136, 138, 322
Stressors .................................................................vi, 5, 38, W
64, 78, 131, 136, 138, 209, 224, 239, 252,
Well-being ............................................................... 62, 71,
259–261, 263, 311, 319, 355, 356, 358, 359,
87–98, 222, 223, 227, 240
366, 369
Stress response......................................................v, 4, 7, 8, Y
10, 16, 18, 61, 77, 79, 89, 163, 210,
226, 368, 371 Yin-Yang ........................................................... v, 3, 77, 78