Gaba

GABA neurocircuits controlling the Hypothalamic-
Pituitary-Adrenal Axis
Masters Thesis
By
Andreas Søderman
Supervisors
Dr. Jens Damsgaard Mikkelsen PhD*
Henning F Bjerregaard PhD**
Dr James P Herman PhD***
Roskilde University Centre**
NeuroSearch A/S*
Genome Research Institute, University of Cincinnati***
2005
GABA NEUROCIRCUITS CONTROLLING JANUARY 2006
THE HYPOTHALAMIC-PITUITARY-ADRENAL AXIS NEUROSEARCH
PREFACE
This thesis comprises work performed at the Department of Functional Neuroanatomy and Biomarkers,
NeuroSearch AS in the period from September 2004 to January 2005. Part of the study period September to
December 2005 was spent at the Laboratory of Stress Neurobiology at the Cincinnati University, Genome
Research Institute.
During the little more than 1 year and 3 months I have worked on this project I have become familiar with,
different areas of pharmacology, molecular physiology, histology and neuroanatomy.
The project has evolved from being a purely pharmacological project, to cross over borders into the areas of
histology and physiology. During the project I have had the pleasure and challenge of learning a wide array of
techniques including: radioimmunoassay, immunocytochemistry, in situ hybridization and neurotracing techniques.
In order to lawfully conduct animal procedures I was required by Danish law to take a 3 week course in laboratory
animal science at the Royal Danish Veterinary and Agricultural University (KVL), this was done in January of 2005.
The project has during the entire period been carried out as an integrated part of Neurosearch’s anxiety research,
with special emphasis on GABA inhibitory regulation of the hypothalamic pituitary adrenal axis (HPA axis).
The work done in the thesis has resulted in one published paper; the article has been accepted by The European
journal of Pharmacology (Mikkelsen et al., 2005) presented in appendix E. I have also had the pleasure of
presenting my work as a poster at the Society of Neuroscience annual meeting in November 2005. The abstract
from the meeting is presented in appendix D.
Acknowledgements
First I would like to thank my internal supervisor at Roskilde University Henning F. Bjergaard, you have been an
excellent help and have given me good and constructive criticism.
Secondly I would like give a warm thanks to the entire technical staff at the Department of Functional
Neuroanatomy and Biomarkers at NeuroSearch, especially Pia Rosholm and Tine Engelbeck, you have been
extremely helpful and good teachers, you have also been hard on me, when that was needed and I have learned
very much, thanks to you.
Next I would like to thank Dr. James P. Herman and the entire staff at “The Laboratory of Stress Neurobiology,
Genome Research institute, University of Cincinnati”, I had a great time at your lab where I really learned some
advanced histology and poker.
Last but definitely not least I would like to express my deepest gratitude to my advisor at NeuroSearch
Dr. Jens Damsgaard Mikkelsen without whom this project could never have been, your great patience and
knowledge have helped me immensely throughout the entire project.
−i−
Abstract
Previous studies have demonstrated that classical benzodiazepines decrease hypothalamic–pituitary–adrenocortical
cortex (HPA) axis activity. Paradoxically, high doses of benzodiazepines also stimulate basal circulating
corticosterone levels in some conditions. Because benzodiazepine agonists display little selectivity to any of the α
subtypes of the γ-amino butyric acid GABAA receptor to which they bind, we propose that the unequivocal results
are due to an a subtype-dependent modulation of the hypothalamic–pituitary–adrenocortical cortex axis output. To
test this, basal hormonal output and induction of FOS in the hypothalamic paraventricular nucleus were measured
after administration of various benzodiazepine ligands in mice. Zolpidem, a selective α1 subtype agonist, produced
a very strong increase in plasma adrenocorticotropic hormone and corticosterone
The non-selective full agonist’s diazepam and zopiclone induced a lower increase in circulating corticosterone than
after zolpidem. In contrast, the α2,3,5 selective benzodiazepine agonist and α1 antagonist L-838,417 had no effect on
corticosterone levels. Strong induction of FOS in the paraventricular nucleus was found in response to zolpidem,
diazepam, and zopiclone, but not L-838,417. Finally, pre-administration of L-838,417 prior to zolpidem strongly
inhibited the effect of zolpidem on corticosterone. Likewise, the non-selective agonist’s diazepam and zopiclone at
a dose that alone had no effect on corticosterone also inhibited the effect of zolpidem. Taken together, these
results suggest that benzodiazepine ligands modulate the hypothalamic–pituitary–adrenocortical axis through partly
opposite mechanisms; and that the net effect is dependent on the composition of the GABAA receptor subunits to
which they bind.
In order to further explain the biological mechanism by which the different benzodiazepins activate the
hypothalamic–pituitary–adrenocortical axis, we performed in situ hybridization to clarify the GABAA subunit
composition in the paraventricular nucleus and sub paraventricular area, we show in agreement with the available
literature that the paraventricular nucleus is comprised of neurons with the α2 phenotype. We also find that the sub
paraventricular nucleus is made up of neurons with the α1 phenotype this has not been shown elsewhere.
In this study we also conclude that the neurons in the paraventricular nucleus that are activated after zolpidem
administration do not co-express oxytocin to any significant degree.
The study also makes an attempt to map afferent connection to the paraventricular nucleus, this is done using
Cholera toxin subunit B as a retrograde tracer. We focuses the projections origination in the bed nucleus stria
terminalis, and in agreement with the available literature find that the bed nucleus stria terminalis has numerous
projections to the paraventricular nucleus, we also find that most of these projections originate in either the lateral
division or in the medial division, anterior part of the bed nucleus stria terminalis.
− ii −
Resume
Tidligere studier har vist at terapeutiske doser af de klassiske benzodiazepiner hæmmer aktiviteten af
hypothalamus-hypofyse-binyrebark aksen. Pardoksalt er det dog at høje doser af benzodiazepiner samtidigt
stimulere produktionen af corticosterone. De klassiske benzodiazepiner udviser ingen selektivitet over for de
forskellige α subunits, der er en del af γ-amino butyric acid GABAA receptoren. Vores hypotese er at, de i
litteraturen modsigende resultater skyldes en α subunit afhængig modulation af hypotalamus-hypofyse-binyrebark
aksens produktion af corticosterone (cortisol i mennesker). For at teste dette blev plasma hormon niveauerne
(ACTH og corticosterone), samt induktion af FOS i den paraventrikulære hypothalamiske kerne målt i mus, efter
dosering af benzodiazepiner med forskellig affinitet for α subunits.
Zolpidem som er en α1 selektiv agonist, var årsag til en kraftig stigning i plasma corticosterone hos mus. De to non-
selektive agonister diazepam og zopiclone induerede også en stigning i plasma cortocosterone, denne var dog for
både effekt og potens lavere end efter behandling med zolpidem. I skarp kontrast til disse resultater inducerede
den α2,3,5 selective benzodiazepin agonist og α1 antagonist L-838,417 ingen effekt på hypotalamus-hypofyse-
binyrebark aksen. Svarende til disse resultater, sås en kraftig induktion af FOS i den paraventrikulære
hypothalamiske kerne efter behandling med zolpidem, diazepam og zopiclone men ikke efter L-838,417.
Ved at forbehandle dyrene med L-838,417 før indgift af zolpidem var vi i stand til kraftigt at hæmme zolpidems
effekt på corticosterone. Det samme var tilfældet hvis vi forbehandlede med enten diazepam eller zopiclone i doser
der i dem selv ikke stimulere et corticosterone respons.
Dette har ledt til en hypotese hvor vi foreslår at benzodiazepiner modulere hypotalamus-hypofyse-binyrebark aksen
via en modsat rettet mekanisme hvor netto effekten afhænger af sammensætningen af GABAA receptorens
underenheder. Vi fastslår også i dette studie at de neuroner, der er aktiveret i den paraventrikulære hypothalamiske
kerne efter behandling med zolpidem, og derfor er FOS positive ikke samtidigt udtrykker oxytocin.
For at forklare den biologiske mekanisme i nærmere detaljer, var det nødvendigt at bestemme i hvordan GABAA α
enheder var udtrykt i den paraventrikulære hypothalamiske kerne samt de omkringliggende områder, dette blev
gjort ved at benytte in situ hybridzering. Vi bestemte i overensstemmelse med litteraturen at neuronerne i PVN
udtrykker GABAA α2 noget nyt er at vi her viser at neuronerne i det sub i paraventrikulære hypothalamiske kerne
område udtrykker α1.
I en del af studiet forsøger vi at kortlægge afferente nervebaner til PVN, i studiet benytter vi os af den retrograde
tracer cholera toxin subunit B. I studiet bliver der fokuseret på nervebaner der har deres udspring i bed nucleus
stria terminalis, resultaterne viser at der er adskillige nerveceller i bed nucleus stria terminalis med projektioner til
den paraventrikulære hypothalamiske kerne, dette er i god overensstemmelse med hvad der findes i litteraturen, vi
viser dog at de fleste af disse projektioner oprinder i enten den laterale eller mediale, anterior del af bed nucleus
stria terminalis.
− iii −
TABLE OF CONTENTS
1. Introduction .............................................................................................................................................................................. 1
1.1. Summary of the project aims:......................................................................................................................................... 3
2. Background theory................................................................................................................................................................... 4
2.1.1. GABAergic Neurotransmission ............................................................................................................................. 4
2.1.2. GABAA receptor biology......................................................................................................................................... 6
2.2. The hypothalamus............................................................................................................................................................ 7
2.2.1. Development of the hypothalamus ....................................................................................................................... 8
2.2.2. Anatomical Boundaries............................................................................................................................................ 8
2.2.3. Blood supply............................................................................................................................................................ 10
2.3. The Paraventricular Nucleus ........................................................................................................................................ 12
2.3.1. Anatomy of the PVN............................................................................................................................................. 14
2.3.2. The subparaventricular Area................................................................................................................................. 15
2.4. Biological stress .............................................................................................................................................................. 17
2.4.1. The HPA axis and the circadian rhythms ........................................................................................................... 18
2.4.2. Negative feedback regulation of the HPA axis.................................................................................................. 19
2.5. C-FOS as a marker of activity in neuroendocrine systems...................................................................................... 19
2.6. Retrograde neurotracing using Cholera toxin subunit B.......................................................................................... 20
2.7. Benzodiazepines ............................................................................................................................................................. 21
2.7.1. Basic Chemistry....................................................................................................................................................... 22
2.7.2. Zolpidem.................................................................................................................................................................. 23
2.7.3. Diazepam ................................................................................................................................................................. 23
2.7.4. Zopiclone................................................................................................................................................................. 25
2.7.5. L-838,417 ................................................................................................................................................................. 26
3. Methods and materials. ......................................................................................................................................................... 27
3.1. Compounds..................................................................................................................................................................... 27
3.2. Animals ............................................................................................................................................................................ 27
3.3. Single dosing of drugs.................................................................................................................................................... 27
3.4. Combination of drugs.................................................................................................................................................... 27
3.5. Radioimmunoassay......................................................................................................................................................... 28
3.6. FOS studies ..................................................................................................................................................................... 28
3.7. Double labelling FOS / oxytocin ................................................................................................................................ 29
3.8. Data analysis.................................................................................................................................................................... 29
3.8.1. FOS........................................................................................................................................................................... 29
3.8.2. Double labelling FOS / oxytocin......................................................................................................................... 30
3.9. Retrograde tracing with cholera toxin subunit B (ChB)........................................................................................... 30
3.9.1. Status ChB injected animals.................................................................................................................................. 32
3.10. In situ hybridization, locating the anatomical distribution of the GABAA α1, α2 and α3 subunits. .............. 33
3.11. In situ hybridization followed by immunocytochemistry (Conducted at the Genome Research Institute,
University of Cincinnati)....................................................................................................................................................... 34
3.11.1. Probe preparation protocol................................................................................................................................. 34
3.11.2. In situ hybridization followed by immunocytochemical staining ................................................................. 34
3.12. Electronic picture processing ..................................................................................................................................... 35
3.13. Statistics ......................................................................................................................................................................... 35
4. Results...................................................................................................................................................................................... 36
4.1. Effect of acute treatment with benzodiazepines / GABAA positive modulators. ............................................... 36
4.2. Interaction inhibition studies........................................................................................................................................ 38
4.3. Effects of benzodiazepine ligands on the induction of FOS in the paraventricular nucleus............................. 40
4.4. Effects of benzodiazepine ligands on FOS expression in the paraventricular nucleus....................................... 40
4.5. Identifying FOS expressing neurons in the PVN ..................................................................................................... 42
4.6. Results from in situ hybridization................................................................................................................................ 43
− iv −
4.7. Detailed analysis ............................................................................................................................................................. 43

4.8. GABAA alpha 1 Distribution in the PVN and SubPVN area ................................................................................ 44
4.9. GABAA alpha 2 distribution in the PVN and SubPVN area.................................................................................. 44
4.10. GABAA alpha 3 Distribution in the PVN and SubPVN area............................................................................... 45
4.11. Results from ChB retrograde neuro-tracing............................................................................................................. 45
4.11.1. 6.11.1. Projections of the BSNT to the SubPVN area ................................................................................... 48
4.12. Dual in situ hybridization / immunocytochemical stain........................................................................................ 49
5. Discussion ............................................................................................................................................................................... 51
5.1. GABA neurocircuits connecting the Bed Nucleus Stria terminalis to the Paraventricular Nucleus................. 53
5.2. Methodological considerations..................................................................................................................................... 53
5.2.1. FOS........................................................................................................................................................................... 53
5.2.2. Drugs ........................................................................................................................................................................ 54
5.2.3. Combination of drugs............................................................................................................................................ 54
5.2.4. Animal models ........................................................................................................................................................ 54
5.2.5. In situ hybridization ............................................................................................................................................... 54
5.2.6. Retrograde tracing using Cholera Toxin subunit B (ChB)............................................................................... 55
5.2.7. Dual in situ hybridization/immunocytochemical stain .................................................................................... 55
6. Conclusion .............................................................................................................................................................................. 56
7. References ............................................................................................................................................................................... 57
8. Appendices.............................................................................................................................................................................. 65
8.1. Apendix A (Preparation of mediums, buffer and solutions: Immunocytochemistry.)...........................................I
8.2. Appendix B (Preparation of mediums, buffer and solutions: In situ hybridization )......................................... III
8.3. Apendix C Sequence of the oligonucleotid probes used for in situ hybridization................................................ V
8.4. Apendix D (abstract presentet at the society for neuroscience annual meeting november 2005) ....................VI
8.5. Apendix E European Journal of Pharmacology 519 (2005) 223 – 230...............................................................VII
−v−
ABBREVIATIONS
AC Anterior commissure
ACA Anterior commissure, anterior part
ACTH Adrenocorticotrophic hormone
AVP Arginine vasopressin
BNST Bed nucleus of the stria terminalis
BSTIA Bed nucleus of the stria terminalis, intraamygdaloid division
BSTL Bed nucleus of the stria terminalis, lateral division
BSTLD Bed nucleus of the stria terminalis, lateral division, dorsal part
BSTLI Bed nucleus of the stria terminalis, lateral division, intermediate part
BSTLJ Bed nucleus of the stria terminalis, lateral division, juxtacapsular part
BSTLP Bed nucleus of the stria terminalis, lateral division, posterior part
BSTLV Bed nucleus of the stria terminalis, lateral division, ventral part
BSTMA Bed nucleus of the stria terminalis, medial division, anterior part
BSTMPI Bed nucleus of the stria terminalis, medial division, posterointermediate part
BSTMPL Bed nucleus of the stria terminalis, medial division, posterolateral part
BSTMPM Bed nucleus of the stria terminalis, medial division, posteromedial part
BSTMV Bed nucleus of the stria terminalis, medial division, ventral part
BSTS Bed nucleus of stria terminalis, supracapsular part
ChB Cholera toxin subunit B
CRH Corticotrophin releasing hormone
DEPC Diethylpyrocarbonate
GABA γ aminobutyric acid
GABAA γ aminobutyric acid receptor A
GAD Glutamic acid decarboxylase
GLU Glutamic acid
GR Glucocorticoid receptor
HPA axis Hypothalamic-pituitary-adrenal axis
ICC Immunocytochemistry
I.P. Intraperitoneal
KPBS Potassium phosphate buffered saline
LH Luteinizing hormone
LV Lateral ventricle
MNOP Median preoptic nucleus
MR Mineralocorticoid receptor
OX Optic chiasm
PBS Phosphat buffered saline
POMC Pro-opiomelanocortin
SCC Podium cacodylate buffer
SCN Suprachiasmatic Nucleus
SubPVN Sub paraventricular nucleus
VIP Prolactine releasing hormone releasing factor
− vi −
JANUARY 2006
INTRODUCTION NeuroSearch
1. INTRODUCTION
Stress is a phenomenon we see all around us, but seemingly the term is too well known and too little understood.
Exposure to acute or chronic stress will induce immediate as well as long term changes in the central nervous
system (CNS). During the last 30 years, it has become increasingly clear that stress serves as one of the main
triggers for psychiatric as well as non-psychiatric disorders, including depression, anxiety, psychosis, drug abuse and
dementia (Able, 2005).
Thus the topic of stress and understanding the underlying biological principles governing the stress response is of
the outmost importance to the scientific and medical community.
This thesis will focus on the role of the inhibitory neurotransmitter γ-aminobuturic acid, receptor A (GABAA) and
the receptors role in controlling specific neurocircuits of the hypothalamic pituitary adrenal cortical axis (HPA axis)
which is a key component of the biological stress response.
GABA is the major inhibitory neurotransmitter in the mammalian brain and GABA exerts its effect through two
distinct receptor types the GABAA and GABAB receptors, this thesis will focus on the GABAA receptor.
The GABAA receptor is the site of action of benzodiazepines in the brain. Multiple isoforms of GABAA receptor
exist; each receptor is comprised of a pentameric complex, formed by the co-assembly of subunits selected from 16
genes (α1-6, β1-3, γ1-3, δ, ε, π, and θ) surrounding a central chloride ion-channel (Sieghart, 1995). Agonist activation of
the GABAA receptor is augmented by the anxiolytic benzodiazepine ligands, causing a parallel shift of the GABA
concentration-response curve (Barnard et al., 1998; Sieghart, 1995). The α subunit isoform present within an
individual GABAA receptor subtype is the primary determinant of benzodiazepine ligand recognition (Fleck, 2002;
Mehta and Ticku, 1999; Pritchett and Seeburg, 1990). GABAA receptors containing α1, α2, α3, or α5 together with β,
and a γ2 subunit are all recognised by the classical benzodiazepines. Selective benzodiazepines as well as other drugs
distinguish the GABAA receptors on the basis of their α subunit composition. For example, the sedative-hypnotic
imidazolpyridine zolpidem has higher affinity for α1 than α2, α3, or α5-containing GABAA receptors (Squires et al.,
1979). More recently new anxiolytic compounds with subtype selection for α2, α3, and α5 over α1 containing
GABAA receptors such as L-838,417 have been developed (McKernan et al., 2000). It is therefore plausible that
distinct α subtype-specific GABAA receptors mediate various effects of classical benzodiazepines.
Disturbances in the hypothalamic–pituitary–adrenocortical (HPA) axis are well known in human depression and
anxiety, as well as in experimental animal models of anxiety and depression (Arborelius et al., 1999; Holsboer,
2000; Jensen et al., 2001; Van de Kar and Blair, 1999). The abnormal hypersecretion of cortisol that occurs in a
variety of psychiatric disorders including major depression, sleep disturbances, and anxiety is suppressed by
administration of benzodiazepines (De Boer et al., 1990). Neurons in the medial parvocellular part of the
hypothalamic paraventricular nucleus contain corticotrophin-releasing hormone (CRH), and these neurons
integrate excitatory and inhibitory signals and behavioural responses to stress into appropriate secretion of CRH
that leads to adrenocorticotropic hormone (ACTH) release to the general circulation (Van de Kar and Blair, 1999;
Carrasco and Van de Kar, 2003). GABA is essential as a negative regulator of neuronal excitability in the
paraventricular nucleus, thus mediating the amplitude and the duration of the stress response (Kovacs et al., 2004).
GABA is also contained in neurons comprised in the negative feedback loop to the paraventricular nucleus
(Bowers et al., 1998), and it is plausible that GABAA receptors may act on converging projections to the
paraventricular nucleus. GABAergic afferents in the paraventricular nucleus originating from a number of forebrain
areas have been reported and ultrastructural studies have shown that GABAergic axons, to a large extent, terminate
on CRH neurons (Sawchenko et al., 1996; Miklos and Kovacs, 2002), and in situ hybridisation and in vitro
autoradiography have detected α1, α2, and α3 subtypes in the paraventricular nucleus (Cullinan, 2000; Duncan et al.,
1995; Fenelon et al., 1995).
The role of benzodiazepines on basal corticosteroid release is complex, and even though it has been extensively
studied over the last three decades, there is still a lot of controversy in the literature. Benzodiazepines have, under
basal conditions, been shown to stimulate (Vargas et al., 2001; Le et al., 1979; Calogero et al., 1990), to inhibit
(Owens et al., 1989; Pivac and Pericic, 1993), and to be ineffective (Matheson et al., 1988). The state of the animal
under which these drugs are tested should be considered
−1−
THE HYPOTHALAMIC-PITUITARY-ADRENAL AXIS NeuroSearch
with special care, because benzodiazepines interact with stress-related behaviours and cause a dose- dependent
inhibition of stress-induced rise in corticosteroid levels (De Souza, 1990), for review). In addition, the effect of
acute benzodiazepines on basal corticosterone levels might be related to several factors such as dose, compound
and gender of the animal (Wilson et al., 1996), and these features might underlie the unequivocal results. In
agreement with the literature, we show here that doses of diazepam higher than those producing anxiolytic effects
increase corticosterone levels in mouse blood (Lakic et al., 1986; Kalman et al., 1997). We propose that the
conflicting results reported earlier may also be a consequence of the net effect of non-selective benzodiazepines on
the various GABAA receptor α subtypes. In order to address this issue, the responsiveness of the murine HPA axis
output was examined after acute and combined treatments with a number of benzodiazepine receptor modulators
specific for the α1 and/or α2,3,5 containing GABAA receptors. In addition, we used induction of FOS as a marker to
identify the level of activation in the paraventricular nucleus after acute administration of benzodiazepine ligands.
After having established that neurons in the PVN expressed FOS after administration of benzodiazepines, we
made an effort to identify the phenotype of the FOS expressing neurons, utilizing double immunocytochemical
labelling for FOS and oxytocin.
The afferent projections to the PVN were also investigated; this was done using retrograde tracing techniques, with
a special emphasis on the neurocircuits originating in the Bed Nucleus Stria Terminalis, the results obtained in
these studies are in good agreement with the available literature (Herman et al., 2002; Ninan, 1999).
We also made attempts to determine, whether or not the PVN projecting neurons in the Bed Nucleus Stria
Terminalis expressed GAD65 m-RNA and therefore could be classified as being GABAergic.
This was done using a labour intensive novel histological technique; more specifically a combination of in situ
hybridization and immunocytochemistry, this last part of the project was carried out at the Genome Research
Institute, University of Cincinnati, in close collaboration with Dr. James P. Herman.
−2−
JANUARY 2006
INTRODUCTION NeuroSearch
1.1. SUMMARY OF THE PROJECT AIMS:

• Investigate why high concentrations of benzodiazepines increase the basal level of corticosterone in mice.
• Do neurons containing the GABAA receptor inhibit or activate the HPA axis, and what role do the
different α subunits play?
• Where in the brain do the different benzodiazepine ligands exert their effect.
• What type of neurons is activated in the PVN?
• To map afferent projections to the PVN and relate these to GABAergic neurons with emphasis on the
BNST.
−3−
2. BACKGROUND THEORY
2.1.1. GABAERGIC NEUROTRANSMISSION

γ-aminobutyric acid (GABA) is widely found in both vertebrates and invertebrates and in both neuronal and non-
neuronal tissues. It was first identified as the neurotransmitter for an inhibitory neuromuscular junction in the
walking leg of the lobster. Since its discovery over 50 years ago, numerous biochemical and neurophysical
observations have been made about brain GABA and GABA-systems that make a strong case for its role as a
neurotransmitter in the mammalian brain. (Cooper et al., 2003).
GABA is synthesised from glutamate by the enzyme glutamic acid decarboxylase (GAD). It is stored in vesicles
within the synaptic terminal of neurons. When an action potential travels down the axon and reaches the synaptic
terminal, calcium channels are activated, resulting in calcium influx. This leads to fusion of the neurotransmitter-
containing vesicles with the cell membrane and release of the neurotransmitter, in this case GABA, into the
synaptic cleft. Then GABA diffuses across the synaptic gap and binds to its receptor. GABA receptors are of two
types, ionotropic (GABAA) or metabotropic (GABAB), and their activation leads to hyperpolarization of the
postsynaptic membrane and a decrease in the excitability of the cell. The response is terminated by desensitization
of the receptors and by enzymatic degradation or transport of GABA from the synapse into the presynaptic
terminal or surrounding glial cells. The latter is achieved through specific transporter proteins referred to as GABA
transporters. Once taken up into a cell, two potential fates await it. The GABA can be taken up into mitochondria
and metabolized to succinic semialdehyde (by GABA transaminase), and then to succinic acid, and enter the
tricarboxylic acid pathway. This is referred to as the GABA shunt. Alternatively, in the presynaptic terminal, it can
be recycled into synaptic vesicles and, thereby, made available for subsequent release.
Figure 1 “A schematic presentation of the vertebrate GABAergic synapse. GABA (filled circles) is generated from glutamate by glutamic
acid decarboxylase (GAD). When released from presynaptic vesicles into the synaptic cleft, it diffuses across and binds to postsynaptic
GABAA and GABAB receptors. It may also bind to presynaptic GABAB receptors. GABA is removed from the synaptic cleft into
surrounding glial cells or the presynaptic terminal by GABA transporters. It is directly recycled into synaptic vesicles or taken up by
mitochondria, converted by GABA transaminase (filled triangle)” (Creigton, 1999)
Like other neurotransmitters GABA and its biosynthetic enzyme glutamine acid decarboxylase GAD, have a
discrete non-uniform distribution through out the mammalian brain.
Storage of GABA can be demonstrated in selected synaptosomal populations and the vesicular GABA transporter
has been cloned and sequenced (Cooper and Bloom, 2003).
−4−
JANUARY 2006
BACKGROUND THEORY NeuroSearch
Release of endogenous or radioactive labelled GABA can be exogenously accumulated; GABA release can be
evoked by appropriate experimental conditions. The presence of GABA-containing neurons has been mapped out
using in situ hybridization for GAD mRNA and immunocytochemical detection of GAD. The most compelling
proof, that GABA is a neurotransmitter in the brain, is that intracellular recording studies have shown that
synthetic GABA causes a hyperpolarization of neurons similar to that evoked by naturally occurring transmitter
substances, and that these inhibitory actions of synthetic GABA and GABA-containing pathways can be
antagonized by drugs selective for the GABA receptor, such as bicuculline. (Decavel and Van den Pol, 1990).
Three primary enzymes are involved in the catabolism of GABA before its final metabolite succinate enters the
Krebs cycle. The relative activity of enzymes involved in the degradation of GABA suggests that they play only a
minor role in the termination of the action of GABA.
Figure 2 outlines the metabolism of GABA and its relationship to the Krebs cycle and carbohydrate metabolism.
GABA is formed by the α-decarboxylation of L-glutamic acid, a reaction catalyzed by GAD, an enzyme that
occurs uniquely in the mammalian CNS and retinal tissue. The precursor of GABA, L-glutamic acid, can be
formed from α-oxoglutarate by transamination or reaction with ammonia. GABA is intimately related to the
oxidative metabolism of carbohydrates in the CNS by means of a “shunt” involving its production from glutamate,
its transanimation with α-oxoflutarate by GABA-α-oxoglutarate transaminase (GABA-T) yielding succinic
semialdehyde and regeneration GLU, and finally its entry into the Krebs cycle as succinic acid via oxidation of
succinic semialdehyde dehydrogenase. In essence then the shunt bypasses the normal oxidative metabolism
involving the enzymes α-oxoglutarate dehydrogenase and succinyl thiokinase.
From a metabolic standpoint the significance of the shunt is unknown; energetically, at least it is less efficient than
direct oxidation through the Krebs cycle. The most common precursor of GABA formation is glucose, although
pyrovate can also act as a precursor. (Cooper et al., 2003).
Figure 2 “GAD catalyzes the formation of GABA from glutamic acid. The synthesis of GABA is linked to the Krebs cycle. GAD requires
vitamin B6 (pyridoxal phosphate) as a cofactor, which can be used to regulate the levels of GABA. GABA can be metabolized by a
transamination reaction with α-ketoglutarate, catalyzed by GABA-transaminase. Compounds such as the competitive GAD inhibitor
allylglycine, inhibit GABA formation and cause convulsions due to the lack of GABA activity”(Messer, 2005) .
Glutamic acid decarboxylase (GAD) is the only synthetic enzyme responsible for the conversation of L-glutamic
acid to GABA and the reaction is irreversible. In mammals this relatively specific decarboxylase is found primarily
in the CNS, where it occurs in high concentrations in the gray matter. In general the localization of this enzyme in
mammals correlates quite well with the GABA content (Cooper et al., 2003). The brain enzyme has been purified
to homogeneity and its properties have been studied in detail. It has a pH optimum of about 6.5 and requires
pyridoxal phosphatate (vitamin B6) as a coenzyme. This purified enzyme is inhibited by structural analogs of Glu,
carbonyl (pyridoxal phosphate [PLP] traping, agents).
−5−
Two isoforms of GAD have been identified which are encoded by two distinct genes. These two isoforms are
designated GAD65 and GAD67 in accordant with their molecular weight. GAD65 and GAD67 differ in amino acid
sequence antigenicity, cellular and subcellular location (Cooper et al., 2003) .
2.1.2. GABAA RECEPTOR BIOLOGY
GABA is as mentioned above the predominant inhibitory neurotransmitter in the mammalian central nervous
system (CNS), with 30 % of all synapses classified as GABAergic. The GABA signal is transmitted by a family of
synaptic and extra synaptic hetero-oligomeric proteins referred to as the GABAA receptors.
When GABA binds to the GABAA receptor it acts by inducing conformational changes of the receptor, which
increases the permeability of the central pore to chloride ions. The resulting chloride flux hyperpolarizes the
neuron, thereby reducing is excitability and producing a general inhibitory effect on neural activity (Nutt and
Malizia, 2001).
Studies of the GABAA receptors molecular biology, done primarily using CDNA libraries have shown, that the
GABAA receptor is assembled from 19 related genes. These are denoted as 6α, 4β, 3γ, 1δ, 1ε, 1π and 3ρ, the
expressed proteins all weigh about ~50.000 Daltons, and each carry 4 transmembrane hydrophobic segments
(TM1-4) (Barnard et al., 1998).
The GABAA receptor found in the central nervous system is assembled primarily from the α1-6 and the β1-4 and
one more of the γ, δ or ε subunits types. The most likely stoichiometry of GABAA receptor subunits in the CNS is
2α, 2β and 1γ subunits (Tretter et al., 1997; Sieghart et al., 1999), arranged as a γ-β-α-β-α pentamer. The major
part 60% of GABAA receptor isoforms in the adult mammalian brain consists of α1,β2, and γ2 subunits. Receptors
with the α2 or α3 configuration constitute approximately 10- 20 % of the GABAA receptors in the brain.
In contrast, most other combinations are rarely expressed or have a unique anatomical distribution and it is
uncertain if they exist. For example the α4 subunit is concentrated in the thalamus and the hippocampal dentate
gyrus. The α5 subunit is primarily expressed in the hippocampal pyramidal cells, while α6 m-RNA is restricted to
cerbellar granule cells. Additional the π subunit is abundant in the uterus but is virtual absent in the brain(Hedblom
and Kirkness, 1997)
A B
Figure 3 (A) A schematic presentation of the GABAA ion channel with the different modulator binding sites indicated. (B) Shows yet
another schematic presentation showing the amino acids implicated in the binding of GABA and most of the classic benzodiazepines.
Pharmacological analysis of recombinant GABAA receptors has shown that the structural heterogeneity of the
GABAA receptor may determine the functional and pharmacological properties of the receptor expressed at the cell
−6−
JANUARY 2006
surface (Barnard et al., 1998; Sigel et al., 1990b). The α and β subunits participate in high affinity GABA/Muscimol
∗
binding and play a vital role in receptor assembly (Taylor et al., 1999).
The classical benzodiazepine ∗ binding site has been localized, at the interface of the α and the γ subunit (Pritchett
and Seeburg, 1990). Moreover α4 and α6 constitute a subfamily of receptors that are insensitive to classical, non-
selective benzodiazepine modulation. The γ is also essential to the modulatory action of benzodiazepines since the
binding site for the benzodiazepines is between the αx and γ2 subunit (Sigel et al., 1990a).
The δ subunit is involved in neurosteroid modulation, agonist affinity and desensitization (Wohlfarth et al., 2002).
The ρ-subunits is expressed primarily in the retina and confer a distinct pharmacology and are often termed
GABAC receptors (Bormann, 2000).
2.2. THE HYPOTHALAMUS

In order to fully understand and appreciate the basic concepts of biological stress, it is necessary to introduce one
of the major brain components, regulating the body’s endocrine, physical response to stress namely the
hypothalamus.
There is clinical as well as experimental evidence that the hypothalamus plays an important role in survival of both
the individual as well as the species. The common view is that the hypothalamus plays a pivotal role in integrating
a wide range of endocrine autonomic and behavioural responses, which guarantee homeostasis and reproduction.
Generally speaking, these responses are involved in regulating metabolism, and thereby providing an adequate
nutrient and water intake, defending the animal from predators and other threats, and allowing the generation and
care of offspring. The hypothalamus is involved in exploratory, digestive, thermoregulatory, aggressive-defensive,
sexual and parental behaviours and these expressions are tied in a fundamental way to the circadian rhythm of sleep
and wakefulness.
There is abundant evidence that experimental manipulation of specific hypothalamic regions profoundly affects the
expression of these behaviours. The complexity of the hypothalamus is immense and the attempts to understand
its functions are based on three types of hypothesis which are in themselves not necessarily mutually exclusive.
The first and earliest was laid down by Hughlings Jackson, Ramón y Cajal, and Sherrington and assumes that cell
groups in the hypothalamus are nothing more than components of extensive central neural circuits which are
particular important for the integration of somatomotor and viceromotor responses involved in homeostasis
(Jackson and Schnieden, 1968).
The second hypothesis was described as the stellar hypothesis and assumes that the hypothalamus contains an array
of centres each of which plays a critical role in the inhibition or cessation of eating, drinking, and sexual behaviour
(Stellar, 1954). The third and latest hypothesis is based on the conclusion that many behavioural effects of
hypothalamic manipulations are the result of tampering with ascending or descending fibbers that pass through the
hypothalamus (Morgan and Pfeil, 1979). The gap between structure and function is immense in the hypothalamus.
The picture is further diffused by the fact that, cell groups and fibre tracts in this part of the brain is relatively
poorly differentiated.
The hypothalamus is the neuroendocrine control centre in the brain, from which autonomic and homeostatic
regulation is achieved. The hypothalamus controls hormone secretion from the anterior and posterior pituitary. It
also controls or modifies a number of homeostatic processes ranging from body temperature regulation, respiration
and the ingestion of food and water.
The integration of somatomotor, autonomic and endocrine responses which characterizes the expression of
emotional behaviours, are often discussed in relation to the limbic system, and are also attributable in large part to
the hypothalamus. The hypothalamus is able to exert such profound controls over these functions through a rich
∗
5-(aminomethyl)-3-isoxazolol, potent GABAA agonist
∗
Non subunit selective benzodiazepines, ex diazepam, midazolam
−7−
array of connections with the rest of the central nervous system. Notably are the brainstem reticular formation and
autonomic zones, limbic forebrain (particularly amygdala) and cerebral cortex.
2.2.1. DEVELOPMENT OF THE HYPOTHALAMUS
The hypothalamus develops from the embryonic diencephalons, which is itself one division of the most anterior
part of the three primary brain vesicles, the posencephalon (the other divisions being the telencephalon which
becomes the cerebral hemisphere). It first appears as the lowest of the three swellings in the lateral wall of the third
ventricle and it becomes separated from the other two, the epithalamus and the thalamus, by a longitudial groove
called the hypothalamic sulcus. As the hypothalamus and thalamus develop in the lateral walls of the
diencephalons, the thalamus unlike the hypothalamus develop in close association with the cortex. The central
cavity between them is gradually transformed to the slit-like third ventricle.
The roof of the third ventricle is formed by a single layer of ependyma covered by a capillary-rich mesenchyme.
The local invagination of these capillaries together with the underlying ependyma forms the choroid plexus of the
third ventricle.
The hypophysis develops in close association with the hypothalamus, indeed the neurohypophysis develops as an
evagination of the diencephalons called the infundibulum. By contrast the adenohypophysis (anterior lobe) has a
non-neural origin developing as it does, from an ectodermal diverticulum of the primitive oral cavity called
Rathke’s pouch. This diverticulum elongates dorsally to contact the down growing infundibulum with which it
fuses, losing contract with the oral cavity. Cells from the adrenohypophysis extend along and around the pituitary
stalk to form the pars tuberalis. Thus the neurohypophysis is in neural communication with the overlying
hypothalamus. The adreno-hypophysis on the other hand has no direct contact with the hypothalamus but instead
there is a rich vascular communication between the two, which is called the portal system. It is by this means that
anterior pituitary secretion is controlled by the hypothalamus, that is, the two are in humoral contact (Hamilton and
Mossman, 1982; Kandel et al., 2000).
Figure 4 schematic drawing of the development of the hypothalamus
2.2.2. ANATOMICAL BOUNDARIES
From this developmental perspective it can be seen that the hypothalamus is situated at the base of the
diencephalon beneath the thalamus, from which it is separated by the longitudinally running hypothalamic sulcus,
which delimits the dorsal extent of the hypothalamus on each side of third ventricle. The rostral boundary of the
hypothalamus is conventionally taken to be the lamina terminalis (which appears to run ventrally from the region of
the anterior commissure to the base of the brain), while caudally the boundary is taken as an imaginary line, which
extends between the caudal limits of the mammalian body below to the commissure above. These caudal and
rostral boundaries are not well defined when examining the hypothalamus microscopically; it gives the expression
that it merges with the surrounding parts of the CNS. Thus rostrally the hypothalamus continues through the
preoptic area with the septal area and part of the anterior perforated substance. The anterior zone of the lateral
hypothalamus merges laterally with the substantia innominata while more caudally it abuts the internal capsule
(dorsolaterally) and globus pallidus (laterally and ventrally). Caudally, the hypothalamus is continuous with the
central grey matter and tegmentum of the mesencephalon. The basal surface of the hypothalamus is defined by the
optic chiasm rostrally and the mamillary bodies caudally. In between is a grey swelling called the turber cinereum
−8−
JANUARY 2006
which, together with infundibular part of the adenohypophysis forms the hypophyseal stalk. This infundibular
region at the base of the hypothalamus is also known as the median eminence. It is a very important site for
vascular communication between hypothalamus and pituitary. The lateral eminence found at the lateral margin of
the turber cinereum marks the site of the ventrally located lateral tuberal nuclei of the hypothalamus.
Figure 5 illustrates the crucial parts involved in regulating the hypothalamic neuroendocrine response in humans. (Psycheduation.org, 2005).
−9−
2.2.3. BLOOD SUPPLY
The blood supply of the basal hypothalamus, infundibulum, hypophyseal stalk and pituitary gland is derived from
the carotid arteries via the superior and inferior hypophyseal arteries. The more posterior part of the median
eminence is supplied by separate vessels derived from the circle of willis, while the most ventral part of the
hypphyseal stalk is supplied by the paired trabecular arteries.
The majority of the blood supply of the adenohypophysis circulation is derived not from the hypophyseal arteries
directly but indirectly through a capillary portal plexus. Thus the arteries supplying the median eminence, empty
into a rich capillary network. Some of these capillaries form long loops, which extend upwards penetrating the
palisade zone of the median eminence, almost the ventricular floor, while others are shorter and remain in a
“mantle” plexus. The portal vessels run downwards on both sides of and behind the hypophyseal stalk to reach the
adenohypophysis where the long portal vessels are joined by the short portal vessels, arising in the ventral-most
hypophyseal stalk and derived in part from the inferior hypophyseal arteries. Some 90 % of adenohypophyseal
blood supply arrives through these long portal vessels.
There are relative few venous channels that run between the anterior pituitary and cavernous sinus, although this
has been assumed to be the obvious route of venous drainage of the gland, ultimately leaving the skull via the
jugular veins. Conversely numerous veins run between the neurohypophyseal and cavernous sinus. It has been
suggested on the basis of such observations, that the major adrenohypophyseal venous drainage is through the
short portal veins into the neurohypophyseal capillary plexus. Thus the flow of blood containing the
adrenohypophyseal hormones might be regulated in part by those factors normally controlling the
neuropypophyseal capillary plexus. Hence the direction of venous blood flow may vary during different
physiological states, and may indeed reverse through the capillary plexus to reach the median eminence or through
the arteries that supply large areas of the mediobasal hypothalamus. It has however not yet been established
whether such retrograde flows of portal blood occurs under physiological circumstances nor if it does occur, what
significance this may have for the feedback regulation of the hypothalamic hormone secretion or for the purposed
central effects of the pituitary peptides. However there is a clear potential for dynamic, regulatory changes in the
pituitary capillary bed.
One additional important feature of the portal capillaries is that they share structural characteristics of peripheral
not central capillaries. That is to say, the capillary endothelium is fenestrated and not marked by tight junctional
complexes as in the case of cerebral capillaries.
Thus it can be said, that the median eminence region of the hypothalamus lies outside the blood-brain barrier, the
structural manifestations of which is provided by tight junctions between capillary endothelial cells. The
characteristics of fenestrated cells are shared by the circumventricular organs, with which the median eminence is
often grouped. (Krieger and Hughes, 1980; Palkovits, 1982).
− 10 −
JANUARY 2006
Figure 6 Arteries and Veins of the human hypothalamus and pituitary (Psycheduation.org, 2005)
− 11 −
2.3. THE PARAVENTRICULAR NUCLEUS

A crucial component of the stress response is the paraventrivular nucleus of the hypothalamus (PVN). The
following will in detail describe the histology as well as the anatomy of the PVN.
The paraventricular nucleus (PVN) of the rat lies adjacent to the third ventricle in the anterior hypothalamus and is
the most prominent of the group of anterior supra group of nuclei.
Figure 7 Schematic presentation of the location of the PVN in the rat barin (Bains and Ferguson, 1995)
Figure 8 Schematic presentation of the PVN anatomy (in humans) (Site de neuroanatomie et al., 2005)
The rat PVN is comprised of a cell community of about 1500 cells (Kiss et al., 1983) Based on their morphology
and efferent connections the majority of the paraventricular neurons seems to belong to three basic categories of
cells. Along with the supraoptic nucleus (SON) the PVN contains the majority of magnocellular neurosecretory
neurons in the hypothalamus, these relatively large cells secrete vasopressin the antidiuretic hormone and oxytocin,
a hormone involved in milk ejection, parturition as well as in ejaculation in man, the axons of these cells project to
the posterior lobe of the pituitary gland, where they secrete their hormonal content directly in to systemic
circulation.
− 12 −
JANUARY 2006
Secondly the PVN contains a large population of parvocellular neurosecretory cells. These neurons form part of
the hypothalamic hypophysiotropic systems that control the release of pituitary hormones. The axons of the
paraventricular hypophysiothropic cells project to the external zone of the median eminence and release their
neurohormonal content into the portal capillary system that “irrigates” the portal anterior lobe cells. These cells
provide CRH (corticotrophin releasing hormone) and regulate pituitary-adrenocortical activity.
Other hyphysiotrophic neurons in the PVN produce thyroid releasing hormone and regulate the release of thyroid
stimulating hormone thus the PVN participate in the regulation of the body’s temperature metabolism (Kiss, 1988).
Finally, paraventricular parvocellular neurons may provide VIP, the putative prolactine releasing hormone releasing
factor to the median eminence. Many other peptides have been localized in the parvocellular neurons including
cholecystokinin (CCK), substance P, neurotensin, atriopeptin), Galaning, SOM enkephami, growth hormone
releasing factor, angiotensin II, Oxytocin and VP (Kiss et al., 1983; Kiss, 1988).
The third major population of neurons, the “autonomic cells,” has been shown to send projections to the brain
stem and spinal cord. Descending fibres of these neurons travel to the dorsal motor nucleus of the vagal nerve and
the intermediolateral column of the spinal cord where they innervate parasympathetic and sympathetic
preganglionic cells respectively. In addition, these neurons innervate a number of other regions associated with the
autonomic nervous system including the nucleus of the solitary tract, the Edinger-Westphal nucleus, the
parabrachial nucleus, the locus coeruleus and the pedunculopontine nucleus. The periaqueductal grey, the sensory
nucleus of the trigeminal nerve and the substantial gelatinosa of the dorsal horn of the spinal cord is another target
for descending PVN projections. The neurons that give rise to these projections are small to medium size cells in
the dorsal and caudal portions of the nucleus. These cells contribute about 30% of all axons descending from the
hypothalamus. It is worth recalling that a significant number of the “autonomic” cells were shown to contain
oxytocin or vasopressin. Through these descending projections, the PVN may influence cardiovascular functions,
analgesia, control of food and water intake, gastrointestinal functions, as well as melatonin synthesis in the pineal
gland. The existence of neurons involved in the three homeostatic systems (the magnocellular neurosecretory, the
parvocellular neurosecretory and the “autonomic” cells) in one nucleus is unique in the hypothalamus. On the basis
of this anatomy, it has been proposed that the PVN has a pivotal role in integrating autonomic and neuroendocrine
functions.(Swanson and Kuypers, 1980).
Figure 9 “Schematic frontal view of the hypothalamic paraventricular nucleus of at the central level. Individual sub nuclei are depicted using
the nomenclature of (Swanson and Kuypers, 1980)” (Larsen et al., 1994) where mp is the medial parvocellular subdivision and pm being the
posterior magnocellular cellular part, pm is the posterior magnocellular part, and dp is the dorsal parvocellular part, lp is the lateral
parvocellular part.
− 13 −
Figure 10 “The three major efferent systems are shown together with the known neurotransmitter (neuropeptide) content of the different
pathways. For reference see the present review. The magnocellular population that projects to the posterior lobe contains vasopressin,
oxytocin and several coexisting peptides. The parvocellular neurosecretory division (2) projecting to the median eminence mainly contains
neuropeptides that are involved in hypophyseotropic functions. The descending projections to the brainstem are mostly oxytocinergic.
Abbreviations: AIIangiotensin II; CCK-cholecystokinin; CRF-coticotropin releasing factor; DYN-dynorphins; ENK-enkephalins; GAL-
galanin; OX-oxytocin; PHI-VIP related peptide; SOM-somatostatin; SP-substance P; TRH-thyrotropin releasing hormone; vasopressin;
MFB-median forebrain bundle; PVN-Pam ventricular nucleus; 3V-third ventricle; d-dorsal subdivision; p-medial parvocellular subdivision,
m-magnocellular subdivision.” (Kiss, 1988)
2.3.1. ANATOMY OF THE PVN
The paraventricular nucleus lies adjacent to the third ventricle in the anterior hypothalamus. The most striking
architectural characteristic of the nucleus is its division into various sub nuclei. Magnocellular neurosecretory,
parvocellular neurosecretory and “autonomic” cells tend to segregate and are localized within certain anatomical
boundaries. The most obvious demarcation, though not complete, is seen between the “compact” magnocellular
and the parvocellular divisions. In the rat, there are approximately 6000 magnocellular and 11000 parvocellular
neurons (Kiss et al., 1983; Kiss, 1988). The magnocellular division is constituted of magnocellular neurosecretory
cells, and can be further subdivided into at least two, topographically segregated, cytoarchitectonic subdivisions.
The parvocellular part appears to be formed by at least 4 parvocellular neurosecretory and 2 “autonomic”
subdivisions. It has become clear that cytoarchitectonic subdivisions comprise functionally distinct subgroups of
neurons. For example, separate populations of magnocellular neurons, distributed within the same magnocellular
subdivision, produce oxytocin or vasopressin. Similarly, discrete populations of parvocellular neurosecretory cells
appear to contain TRH, CRF, or VIP. It seems that neurons having different neurotransmitters are not randomly
mixed within an anatomical subdivision. Cells which have been found to synthesize the same substance have a
tendency to reside adjacent to each other. Anatomical subdivisions of the PVN can be further subdivided into
neurochemically specialized and topographically segregated “subunits” of neurons.
It is clear that paraventricular afferents preferentially innervate certain subdivisions and seem to exhibit
homogenous densities over subpopulations of cells. For example, the noradrenergic innervations from the A 1 cell
group in the brain stem is mainly directed at vasopressin-producing magnocellular neurons as opposed to those
that produce oxytocin. Similarly, ACTH-immunoreactive input appears to innervate periventricular and caudal
paffocellular regions while this input is only sparsely distributed in regions where CRF containing parvocellular
cells are distributed. In contrast, epinephrine-containing fibers seem specifically to terminate on neurons containing
− 14 −
JANUARY 2006
CRF and on neurons sending projections to the brain stem. This, of course, does not necessarily mean that CRF
neurons are the only target of epinephrine containing axons. In fact scattered epinephrine fibers and varicosities
can be found all over the PVN. However, it is possible to determine a distribution pattern of these axons, which
seems to exhibit high, even densities over certain subgroups of neurons. It is important to emphasize that subsets
of neurons specified by their neuromediator content are not strictly demarcated anatomical formations (Everitt and
Hokfelt, 1990; Kiss, 1988).
2.3.2. THE SUBPARAVENTRICULAR AREA
The immediate area surrounding the PVN is richly innervated by a number of limbic structures implicated in the
HPA axis regulation, including the ventral subiculum, prefrontal cortex, lateral septum, medial amygdala, PVN and
suprachiasmatic nucleus (SCN) (Canteras et al., 1995; Risold et al., 1997; Herman et al., 2002). Anterograde tracing
studies indicate that all of these regions innervate the SubPVN area while almost completely avoiding the medial
parvocellular are of the PVN (Herman et al., 2002). The medial parvocellular PVN neurons are very limited
dendritic trees largely contained within the PVN itself (Rho and Swanson, 1989) thus terminal boutons of limbic
afferents to the SubPVN and the region immediately adjacent to the lateral and dorsal limits of the PVN proper do
not appear to contact the parvocellular neurons, but rather act through local interneurons. A substantial proportion
of local neurons contacted by descending limbic projections are likely to be GABAergic in phenotype (Bowers et
al., 1998). Previous studies indicate that the vast majority of PVN-projecting neurons in the hypothalamus and the
bed nucleus of the stria terminalis (BNST) are glutamic acid decarboxylase (GAD) m-RNA -positive (Bowers et al.,
1998) The immediate surroundings of the PVN is particularly rich in GABAergic neurons that project into the
PVN. These cell populations are stress-responsive, showing upregulation of the GABA synthesizing enzyme GAD
(Bowers et al., 1998) as well as c-FOS induction following stress exposure (Cullinan et al., 1996)). As such, these
neurons are putative targets for descending information relevant to the HPA axis function.
The PVN-projecting cell populations co-express glutamate receptors and GAD mRNA, which indicate that
SubPVN GABAergic neurons are likely to integrate glutamatergic neurotransmission.
Similar to glutamate receptors, studies have revealed that GABA receptors are expressed in the parvocellular PVN
and adjacent territories (Cullinan, 2000). GABAA receptor mRNAs specific for the α1-2 β1-3 and γ1-2 subunits have
been found to be expressed within hypophysiotropic CRH neurons in the PVN, as well as within the SubPVN
zone (Kiss, 1988). These receptors are also variably expressed within additional intrahypothalamic regions that send
local projections to the parvocellular PVN (medial preoptic area, dorsomedial hypothalamic nucleus), indicating
that these areas serve as GABA-to-GABA relays capable of disinhibiting CRH cells.
Exposure to chronic (nonhabituating) stress results in a down-regulation of transcripts encoding the β1 and β2
subunits of the receptor in the parvocellular PVN without changes in the β3 subunit, suggesting the possibility that
GABAA receptors might be decreased in CRH-containing cells, or alternatively, that changes in subunit
composition may occur that alter GABAA receptor efficacy at this key regulatory site (Herman et al., 2002).
− 15 −
Figure 11 ”Three-dimensional rendering of the presynaptic GABAergic sites depicted in three planes” (Herman et al., 2002)
− 16 −
JANUARY 2006
2.4. BIOLOGICAL STRESS

Stress is a condition originally defined by Hans Selye, to describe the pathophysiological state associated with
specific physiological changes that could be induced by diverse physical or physiological stimuli (selye, 1936).
The research of Hans Selye in the 1940s lead to further characterisation of the “general adaptation syndrome” and
the proposal that the pituitary was critical for the control of endocrine secretion, including those of the adrenal
glands. He proposed that the adrenocorticotropic hormone (ACTH) released from the pituitary lobe, was
associated with increased production of glucocorticoids from the adrenal glands that excerted profound and
widespread effect on the metabolism and lymphoid organ activity.
In addition to activation of the pituitary-adrenocortical system, endocrine responses to stress include activation of
the symphathic nerves that innervates the adrenal directly. In response to stress the symphatic activation enhances
secretion of the adrenal catecholamines, principally the hormone adrenalin, which is important in the regulation of
the autonomic response to stress (Able, 2005).
The endocrine system encompasses the pituitary gland, the peripheral endocrine organs and the hormones the two
produce. The pituitary gland or the hypophysis is responsible for the production of a range of hormones that has
influence over a wide array of bodily functions.
The pituitary is subjected to afferent control via the actions of specific hypothamlamic releasing or inhibiting
factors that act on the pituitary cells in order to regulate synthesis and release of pituitary hormones.
Stress influences the neuroendocrine regulation of a number of pituitary hormones including ACTH, prolactin,
growth hormone, luteinising hormone (LH), tyroprotrophin, vasopressin and oxytocin.
Upon release into the hypophysial blood supply the hypothalamic-releasing factors; Corticotropin-releasing
hormone (CRH) and the nonapeptide arginine vasopressin (AVP) are transported to the adrenohypophysis where
they activate pituitary cortitrophs which in turn, synthesis and release ACTH in to the general circulation (Able,
2005).
In the blood ACTH passes to the adrenal gland where it binds to receptors on the cells of the zona fasciculata of
the adrenal cortex to promote the conversion of cholesterol esters into to free cholesterol, and stimulate the
steriogenic pathways. This causes a rapid enzymatic conversion of precursor cholesterol into steroid intermediates
ultimately resulting in the formation of the end product of the steroid pathway, the glucocortiticoids, cortisol in
man and corticosterone in rodents who lack the enzyme (C-17 hydroxylase) (Cooper and Bloom, 2003).
In conditions that is associated with chronic ACTH secretion cortisol release may increase several times resulting in
secretion rates of up to 200-250 mg/day. In the rat corticosterone is the principle HPA end product, and small
amounts of this steroid are also secreted in humans (1-4 mg/day). The amount of ACTH present in the plasma, is
the major factor regulating the glucocorticoid release. However, additional hormones from the adrenal medulla are
involved but to a much lesser extent (Bornstein et al., 2000).
The principle role of the glucocorticoids is to produce homeostatic adaptation to stress and this is achieved
through catabolic actions that mobilise energy resources, increasing blood pressure, at the same time suppressing
hunger and sexual behaviour. (Able, 2005)
A central component of the HPA axis is the parvocellular part of the paraventricular nucleus (PVN). It is here that
CRH and AVP is synthesized in the tuberoinfundibular parvocellular cells. CRH and AVP in turn evoke the release
of ACTH via their synergistic actions on the pituitary corticotrophs.
The axon terminals of the parvocellular neurons terminate in the external zone of the median eminence adjacent to
the capillaries of the hyperphysial portal blood supply, where they secrete their content into the portal blood
circulation. Parvocellular AVP, in contrast to magnocellular AVP is involved in regulation of pituitary ACTH
release and does not contribute to osmotic balance regulation. CRH is the primary source of ACTH secretagogue
while AVP colocalises with CRH in approximate 50 % of the CRH containing neurons of resting animals and
humans (Able, 2005; Whitnall, 1993). It has been shown in vitro by (Gillies et al., 1982) and in vivo by (Rivier and
Vale, 1983) that AVP potentiates the effect of CRH on the ACTH secretion. AVP alone has little or no effect on
− 17 −
ACTH secretagogue activity. In addition to inducing ACTH release CRH induces transcription of pro-
opiomelanocortin (POMC) mRNA, which encodes the ACTH precursor protein (Lightman and Young, III, 1988).
During stimuli that activate the HPA axis, synthesis and secretion of CRH and AVP is increased, this in turn has
profound effect on the amount of ACTH and glucosteroid secretion. Thus expression of CRH mRNA and AVP
mRNA in the PVN is increased and POMC mRNA expression in the adrenohypophysis in response to stress. In
addition to AVP several neuropeptides are co-localised with the parvocellular CRH neurons of the PVN including
enkephalin, neurotensin, cholecytokinin, vasoactive intestinal peptide and galanin. CRH containing CRH neurons
has also been reported to be co-localised with neurons expressing GAD (Meister et al., 1988). The co-localization
of CRH and GAD provides a way of subtle regulation of ACTH release.
Figure 12 ”The two major components of the pituitary gland” (Rosenzwig et al., 1996)
2.4.1. THE HPA AXIS AND THE CIRCADIAN RHYTHMS
Glucocorticoid secretion varies through out the day in a pulsating fashion, the same principle applies to ACTH,
and thus circulating levels of ACTH mirrors the levels of glucocorticoids. The maximum levels of glucocorticoids
peak approximately 15-20 minuets after a pulse of ACTH in normal human subjects, a major peak of ACTH
occurs in the early hours of the morning. Following this major pulse of ACTH and glucocorticoids, secretion is
stimulated by further pulses through the day.
The subparaventricluar zone (SubPVN) receives massive projection of vasopressin-containing neurons from the
suprachiasmatic nucleus (SCN). The dorsal and lateral compartments of the parvocellular part of the PVN also
receive projection from the SCN (Vrang et al., 1995).
The SCN is involved in control of the body’s circadian rhythms; however the SCN may also be involved in the
actual stress response, it has been shown that evening activation of the SCN, results in a surge of ACTH and in
turn glucocorticoids (Cascio et al., 1987).
− 18 −
JANUARY 2006
2.4.2. NEGATIVE FEEDBACK REGULATION OF THE HPA AXIS
In order to prevent excessive and potentially dangerous overproduction of glucocorticoids, the HPA axis functions
as a closed-looped system. The production of glucocorticoids is tightly regulated via a negative-feedback system.
Regulatory feedback occurs at several sites and involve both rapid feedback and delayed feedback in humans and
rats (Keller-Wood and Dallman, 1984; Able, 2005). Rapid feedback occurs immediately following a surge in
glucocorticoids and will last between 5 to 15 min whereas delayed feedback is initiated 1-2 hours later and usually
persists up to 5 hours and is dependent on the concentration of glucocorticoids. Delayed feedback can persist up
to 24 hours, this suggests that delayed feedback is a consequence of genomic action of the glucocorticoids whereas
the rapid feedback is exerted primarily via the direct inhibitory action of the glucocorticoids on the synthesis and
release of ACTH at the hypothalamic level by decreasing the level of mRNA expression of CRH and AVP (Able,
2005).
The actions of the glucocorticoids are mediated through two distinct types of receptor complexes, the
mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR).
The GR and MR receptors are located inside the cytoplasm, here they bind steroids that diffuse across the plasma
membrane, once bound, the receptor ligand-complex translocates to the nucleus and interacts with the palindromic
hormone response element on the DNA molecule. Thus activated steroid receptors function as transcriptions
factors, that influence transcription of target genes, which ultimately leads to changes in protein synthesis (Able,
2005).
The MR receptor is primarily distributed in the hippocampus, and in sensory and motor nuclei outside the
hypothalamus (Reul et al., 2000). The GR receptor is however more widely distributed in the hypothalamus, PVN
and brainstem, amygdala and hippocampus in addition to the pituitary gland. Especially the hippocampus is a very
important structure in the negative-feedback regulation of the neuroendocrine stress response both MR and GR
are present in this brain structure. Lesions of the hippocampus-PVN (the BNST) connections result in a higher
basal level of glucocorticoids and in abnormal feedback regulation (Herman et al., 1992).
2.5. C-FOS AS A MARKER OF ACTIVITY IN NEUROENDOCRINE SYSTEMS.
Large parts of the present study, is based on immunocytochemical studies using FOS, the protein product of the
immediate early gene C-FOS, as a marker of transient neuro-activity. The following section will try to justify using
FOS as a biomarker of neuroendocrine activity
In the late 1980s it was discovered that neurons transiently expressed C-FOS when stimulated either electrically,
with growth factors or with neurotransmitters. The protein gene product FOS is localized in the nucleus and can
easily be detected using immunocytochemistry.
The immediate-early gene, c-FOS, is rapidly expressed in several brain sites in response to various physiological and
pharmacological stimuli, including growth factors, neurotransmitters, electroconvulsive shock, osmotic challenge,
and stress (D'Costa et al., 1991; Curran and Morgan, 1985; Ceccatelli et al., 1989; Arnold et al., 1992).
The c-FOS gene encodes the nuclear regulatory factor, FOS, which forms heterodimeric complexes with protein
products of the jun proto-oncogene family (c-jun, jun-B, jun-D)(see figure 11); the resultant heterodimers regulate
gene transcription upon binding to genomic AP-1 sites (Curran and Franza, Jr., 1988; Cohen et al., 1989). FOS and
other regulatory proteins function as signalling molecules linking intracellular second messenger systems to target
genes. FOS is not normally expressed constitutively by most central neurons, but only expressed transiently
following stimulation (Curran and Morgan, 1985).
Thus, demonstrability of either C-FOS m-RNA, or FOS protein is an indicator of a transcriptional-activated cell
populations within the central nervous system (Sagar et al., 1988).
− 19 −
The hypothalamus functions to integrate neuroendocrine, autonomic, and behavioral responses to stress. Studies
have sought to characterize the distribution and mean numbers of neurons in the hypothalamus and the adjacent
preoptic area, that are activated via the FOS stimulus-transcription cascade in response to stress. There is
consistent evidence that parvocellular neurons in PVN, express FOS immunoreactivity in response to intense stress
stimuli, e.g., immobilization or foot shock (Arnold et al., 1992; Smith et al., 1992; Kononen et al., 1992).
C-FOS expression and neuron-firing is frequently linked, however depolarization and expression of c-FOS might
be dissociated. When using FOS to study neuronal systems it is essential to keep in mind, that it is the changes in
signal transduction pathways which induce c-FOS expression, not depolarization in itself. There are 3 major
regulatory pathways for induction of FOS; for a schematic presentation (see figure 13)
Figure 13 “Illustration of the principal signal transduction pathway that evoke FOS in the neurons” (Hoffman and Lyo, 2002).
One of the pathways involved in c-FOS induction results in either increased intracellular calcium or increased
levels of cyclic AMP acting through a calcium response/cyclic AMP responsive element (CaRE/CRE) on the c-
FOS promoter (Sheng et al., 1990; Sassone-Corsi et al., 1988).
A second transduction cascade operates via serum response factors (SRFs) which activate a DNA serum response
element (SRE) independent of cyclic AMP (Gilman, 1988).
A wide number of transmitter receptors, when occupied, increase intracellular cyclic AMP levels or activate voltage
sensitive calcium channels leading to the phosphorylation of CREB and c-FOS induction (Thompson et al., 1995).
These are frequently engaged by stimuli that evoke increased firing in the target neurone. This feature forms the
basis of using FOS as a marker for stimulated neurone activity. However, there can be changes in second
messenger systems without accompanying depolarization just as there can be depolarization without FOS
expression.
2.6. RETROGRADE NEUROTRACING USING CHOLERA TOXIN SUBUNIT B

Part of this study utilises Cholera toxin subunit B as a retrograde tracer, the following section will briefly give some
essential background on the tracer.
− 20 −
JANUARY 2006
Cholera toxin is an oligomeric protein of MW 84,000 daltons and consists of a single A subunit surrounded by five
B subunits (Gill, 1976; Finkelstein et al., 1974). The A subunit is a potent activator of adenylate cyclase and is the
pathogenic agent responsible for the symptoms of cholera. The B subunit (choleragenoid) is responsible for the
binding of the holotoxin to GM1 ganglioside receptors on mammalian cell surfaces (Holmgren and Lonnroth, 1975)
and facilitates entrance of the A subunit into the cell. The A subunit bears the ADP-ribosyl-transferase activity,
which deregulates the Gs protein causing activation of adenylate cyclase. Due to the ubiquitous occurrence of the
GM1 ganglioside receptor on eukaryotic cell membranes, cholera toxin activates adenylate cyclase in a wide variety
of systems.
The cholera toxin B subunit can be used for tract tracing in neurological research, taking advantage of GM1
ganglioside binding and retrograde transport, and since ChB is a protein it can easily be detected using
immunocytochemistry.
Figure 14 shows the principle of retrograde tracing where the tracer is transported towards the cell body.
2.7. BENZODIAZEPINES
Benzodiazepines are probably the most widely taken family of psychotropic drugs in the history of modern
medicine. Therefore it would be out of the scope for this project to describe every detail from synthesis to
pharmacology of the benzodiazepines.
In the 1950s meprobamate (Miltown®), a propanediol carbamate, was available for use as a “tranquilizer”,
meprobamate was more safe than the early barbiturates, but still had many problems.
The goal of the pharmaceutical industry was to produce a drug that selectively relieved anxiety without causing
overt intoxication or general sedation.
The separation of the ability to dull certain mental faculties from general sedation had already been demonstrated,
with the phenothiazines, notably chlorpromazine (Thorazine®). At the time, no one thought of these drugs as
"antipsychotic agents" – there probably would have been a fair amount of chortling, were someone to use a term
with such sweeping and entirely baseless implications. Instead, these agents were called neuroleptic (lobotomizing)
drugs, or simply major tranquilizers.
In the mid 1950s the Hoffman-LaRoche company with chemist Leo E. Sternbach and pharmacologists Lowell
Randall began a systematic search for minor tranquilizers. Chlordiazepoxide (RO 5-0690) was the result of a
mistaken synthesis on the part of one of Sternbach's bench chemists of the last of a series of analogs of a certain
dye compound (which they thought were heptodiazines, but which turned out to be quinazolone-3-oxides) they
had been making. Randall's pharmacologists wondered why RO 5-0690 alone showed all of these behavioural
− 21 −
properties (most notably, "taming" at vastly sub-ataxic doses), when the prior 39 analogs were all inactive. The
answer turned out to be that it wasn't an analog after all – it was a brand new compound, the first of the
benzodiazepines, which was initially called methaminodiazepoxide and then chlordiazepoxide (CDZ), which
remains its USAN name 1 . Chlordiazepoxide goes by the trade name Librium®.
Benzodiazepines act selectively on the GABAA receptor, which as stated earlier mediate fast inhibitory synaptic
transmission in the CNS. Benzodiazepines enhance the response of GABA, by facilitating the opening of GABA-
activated chloride channels. They bind specifically between the α and γ subunit of the GABAA receptor which is
distinct from the GABA binding site and here benzodiazepines act by allosterically increasing the affinity of GABA
for the receptor. (Rang et al., 2004)
2.7.1. BASIC CHEMISTRY
Pharmacology is a combination of physiology and organic chemistry, 1,4-Benzodiazepin is the basic structure of a
benzodiazepine from a chemical perspective.
Figure 15 1,4 Benzodiazepine
However from a pharmacological perspective, there are several additional features required before any drug activity
is seen.
Figure 16 1,4 Benzodiazepin-2-one
Almost all active benzodiazepines except those possessing a fused heterocyclic ring or a thionyl group have a
carbonyl group at position 2. Two additional features are required before we have a pharmacological active
benzodiazepine skeleton. The first of these is another benzene ring (phenyl group), the second is an electron-
withdrawing group, and there must be an electronegative group at position 7, for example chloride, fluoride, and
bromide.
1 United States Adopted Names Designation for nonproprietary names (for drugs) adopted by the USAN Council
in cooperation with the manufacturers concerned; the designation USAN is applicable only to nonproprietary
names coined since June 1961
− 22 −
JANUARY 2006
Elektron attracting
group
Figure 17 N-methyl-7-chloro-5-phenyl-1,4-Benzodiazepin-2-one (Diazepam)
2.7.2. ZOLPIDEM
Figure 18 Shows the chemical structure of Zolpidem (N,N,6-trimethyl-2-(4-methylphenyl)imides[1,2-a]pyridine-3-acetamide hemitartrate)
Zolpidem is an imidazopyridine which differs in structure from the benzodiazepines and zopiclone. It is a
strong sedative with only minor anxiolytic, myorelaxant and anticonvulsant properties, and has been
shown to be effective in inducing and maintaining sleep in adult patients suffering from insomnia (Salva
and Costa, 1995).
Zolpidem, like other benzodiazepines, displays a high affinity to GABAA receptors, containing the α1
subunit and intermediate affinity to α2- and α3-GABAA receptors and fails to bind to α5-GABAA
receptors (Crestani et al., 2000; McKernan et al., 2000j; Salva and Costa, 1995; Salva and Costa, 1995c)..
Zolpidem potentiates the effect of GABA at the GABAA receptor α1 complex with up to 400 % using
whole patch clamp on Purkinje and striatal neurons (Itier et al., 1996; Crestani et al., 2000).
2.7.3. DIAZEPAM
− 23 −
Figure 19 shows the chemical structure of Diazepam (N-methyl-7-chloro-5-phenyl-1,4-Benzodiazepin-2-one)
Diazepam (Valium) was introduced by a Swiss drug company called Roche Labs in 1963, where it quickly
became one of “mother's little helpers” for millions of housewives throughout the '60s and beyond.
Throughout the history of prescription medication, Valium was the first billion-dollar medicine and one
of the first brand-name drugs. Valium launched the era of blockbuster medicines. More prescriptions
were written for Valium than for any other drug between 1969 and 1982. However, today the drug is not
taken with the same frequency due to adverse effects such as amnesia, and the high risk of dependency.
Diazepam is a non selective GABAA agonist (Benson et al., 1998), Diazepam binds at the
benzodiazepine binding pocket between the γ2 and respective αx subunit, with equal affinity for the
GABAA α1, α2, α3 and α5 , not α4/α6 (Scott-Stevens et al., 2005). Studies in genetically modified mice
(McKernan et al., 2000) and transfected human embryonic kidney cells (HEK cells) using patch clamp
(Benson et al., 1998) have shown that a crucial element for diazepam binding is a His residue located at
position 101, when substituted with Arg there is a markedly reduced binding of diazepam to the GABAA
receptor (Benson et al., 1998).
− 24 −
JANUARY 2006
2.7.4. ZOPICLONE
Figure 20 Shows the Chemical structure of Zopiclone (4-methyl-1-piperazinecarboxylic acid 6-(5-chloro-2-pyridinyl)-6,7-dihydro-7-oxo-5H-

pyrrolo[3,4b]-pyrazin-5-yl ester)
Zopiclone, is a cyclopyrrolone derivative, the drug is used as a short-acting hypnotic. Although

structurally unrelated, it has a pharmacologic profile similar to that of benzodiazepines. Following oral
administration, the drug is rapidly absorbed, with a bioavailability of approximately 80% and an
elimination half-life that ranges from 3.5 to 6.5 hours. Zopiclone is well tolerated and effective and is a
suitable alternative to benzodiazepines for short-term treatment of insomnia. Rebound insomnia is
substantially less frequent with zopiclone compared with hypnotic benzodiazepines.
Several studies have shown that although zopiclone has high affinity for the GABAA receptor it does not
discriminate between the different GABAA α subunit compositions (Sanger, 2004; Goa and Heel, 1986).
− 25 −
2.7.5. L-838,417
Figure 21 Shows the chemical structure of L-838,417 (7-tert-Butyl-3-(2,5-difluoro-phenyl)-6-(2-methyl-2H-[1,2,4]triazol-3-ylmethoxy)-

[1,2,4]triazolo[4,3-b]pyridazine).
The anxiolytic compound L-838,417 is a selective GABA agonist with nanomolar affinity for the
benzodiazepine binding site on the GABAA receptor.
L-838,417 has high affinity for receptors with, α2β3γ2, α3β3γ2 and α5β3γ2 markedly lower for receptors
containing α1β3γ2, α4β3γ2 and α6β3γ2. (McKernan et al., 2000).
The efficacy of L-838,417 at the BZ site has been measured by whole-cell patch-clamp recording from
Ltk- cells expressing the same GABAA receptor subtypes. The degree of potentiation of a submaximal
GABA concentration (EC20) has demonstrated that L-838,417 has no efficacy at the α1 subtype (it is a
weak antagonist) and is a partial agonist with equal efficacy at the α2, α3 and α5 subtypes (McKernan et
al., 2000).
In contrast to diazepam L-838,417 does not impair the motor performance of wild type mice on the
rotarod at doses up to 30 mg/kg. With this in mind, it can be said that L-838,417 is a true anxiogenic
with very little of the normal negative side effect of the classical benzodiazepins.
− 26 −
JANUARY 2006
METHODS AND MATERIALS. NeuroSearch
3. METHODS AND MATERIALS.
3.1. COMPOUNDS
Zolpidem and zopiclone were purchased from Tocris Ltd. (Bristol, UK). Diazepam was purchased from Sigma-
Aldrich (St. Louis, USA). L-838,417 was synthesised by the Department of Medicinal Chemistry, NeuroSearch
A/S. All compounds were dissolved in 5% Chremophor and administered intraperitoneally (IP).
3.2. ANIMALS
Adult male NMRI mice weighing 30-35 grams were purchased from Harlan Inc (Hamburg, Germany). The animals
were received at the animal facility at Neurosearch, and housed 5 per cage under 12:12 light: dark cycle, humidity in
a temperature controlled room for at least 7 days before the experiment. Food and water were available ad libitum.
All procedures were conducted in accordance with the Danish National Guide for Care and Use of Laboratory
animals (justits ministeriet, 2003).
Adult male Wistar Rats 180-220 grams were purchased from Harlan Inc (Hamburg, Germany). The animals were
received at the animal facility at Neurosearch, and housed 2-3 per cage under 12:12 light: dark cycle, humidity in a
temperature controlled room for at least 7 days before the experiment. Food and water were available ad libitum.
All procedures were conducted in accordance with the Danish National Guide for Care and Use of Laboratory
animals (justits ministeriet, 2003).
3.3. SINGLE DOSING OF DRUGS

In order to analyse the dose-dependent effect of all benzodiazepine ligands on circulating corticosterone, mice (n =
5) were administered the drugs at doses 0,025, 0.1, 0.5, 2.5, 12.5, and 25 mg/kg ip. The mice were returned to their
home cages and sacrificed by decapitation 60 min after drug administration. Trunk blood was collected in
centrifuge tubes containing 2 mg EDTA and stored at -20°C.
Another experiment was carried out in order to determine the temporal profile of HPA axis activation by
zolpidem, these mice were treated with 12,5 mg/kg zolpidem and the animals were sacrificed and trunk blood was
taken at various time points after drug administration (t = 5, 10, 15, 30, 60, and 120 minutes). The blood was
collected in EDTA containing tubes, kept at 4°C and plasma was isolated by centrifugation and stored at - 20°C
until processed.
3.4. COMBINATION OF DRUGS

In order to investigate the relationship between activation of α1 and non– α1 containing GABAA
receptors, combinations of L-838,417 (12.5 mg/kg i.p.) or vehicle was given 15 min before a single dose
of zolpidem i.p at increasing concentrations (0.5, 2.5, and 12.5 mg/kg). L-838,417 acts at the
benzodiazepine binding site and is a functionally selective agonist at α2, α3, and α5 subtypes. In addition,
L-838,417 is an antagonist on the α1 site; it might be that any inhibition result from its antagonistic
properties. Therefore, combination of the non-selective benzodiazepine ligand diazepam and zopiclone
was tested together with zolpidem. In this case mice were treated with vehicle, diazepam (0.5 mg/kg i.p.)
or zopiclone (0.5 mg/kg i.p.) 15 min before administering the mice with 0.5 mg/kg zolpidem or 2.5
− 27 −
mg/kg. The animals were sacrificed 60 min after zolpidem and trunk blood was taken for corticosterone
measurements. All single and combined dosing experiments were carried out twice.
3.5. RADIOIMMUNOASSAY
All plasma Corticosterone measurements were carried out using a commercially available solid face
radioimmunoassay [125I] kit from DCP diagnostics. ACTH measurements were carried out using a [125I]
kit from Amersham.
The general principle of the two kits is the same and is briefly described below:
I125-labelled rat corticosterone or ACTH competes for a fixed time with corticosterone/ACTH in the
plasma sample for antibody sites. The antibody is immobilized to the wall of a polypropylene tube, so
decanting the supernatant is sufficient to terminate the competition and to isolate the antibody-bound
fraction of the radiolabelled corticosterone. Counting the tubes in a gamma counter, then yields a
number which converts by way of a calibration curve to a measurement of the corticosterone content of
the sample (Gwosdow-Cohen et al., 1982; Gomez-Sanchez et al., 1975)
3.6. FOS STUDIES

Adult male NMRI mice weighing 30-35 grams purchased from Harlan Inc (Hamburg, Germany) were
used for the C-FOS immunohistochemistry studies. The animals were administered intraperitoneally (i.p)
with 0,1, 0,5, 2,5 and 12,5 mg/kg diazepam, zopiclone, zolpidem, L-838,417 or vehicle (5% chremophor)
at 4 hours after the onset of the light phase and returned to their home cage. Sixty minutes after the
injections the animals were deeply anesthetised with mebumal and perfused transcardially with 0.9%
saline followed by 4% paraformaldehyde in 0.1M phosphate buffer (pH 7.4) for 5 min. The brains were
post-fixed over night in 4% paraformaldehyde and subsequently submerged in 20% sucrose in 0.01M
buffered saline (PBS) at 4ºC overnight. 40 μm serial coronal sections were cut through the hypothalamus
on a freezing sledge microtome in series of four and stored in a solution containing ethylene glycol
glycerol in 30% sucrose at -20ºC until further processing. Prior to the immunocytochemical steps, the
sections were rinsed for 3 x 10 min in 0.01M PBS, in 1% H2O2–PBS for 10 minutes, and in PBS with
0.3% Triton X-100, 5% swine-serum, and 1% bovine serum albumin (BSA) for 30 min. Then the
sections were incubated at 4°C for 24 hours in an antiserum against FOS (code #94012-5) diluted 1:4,000
in PBS with 0,3% Triton X-100 and 1% bovine serum albumin. The primary antiserum was raised in
rabbit against the N-terminal peptide similar to 2-17 of the rat FOS protein and characterised previously
(Mikkelsen et al., 1998b). The immunoreactive cells were identified by means of the avidin-biotin
protocol (Hsu and Raine, 1981) and diaminobenzidine was applied as chromogene as described by
(Mikkelsen et al., 1998a). The sections were then mounted into 1% of gelatine dissolved in 0.025M PBS,
air-dried, coverslipped with Permount, and examined under 40 × objective light microscope.
− 28 −
JANUARY 2006
3.7. DOUBLE LABELLING FOS / OXYTOCIN

8 adult male NMRI mice weighing 30-35 grams were used for the C-FOS/oxytocin
immunohistochemical studies. The animals were administered intraperitoneally with 12.5 mg/kg
zolpidem, or vehicle (5% chremophor) at 4 hours after the onset of the light phase and returned to their
home cage. Sixty minutes after the injections the animals were deeply anesthetised with mebumal and
perfused transcardially with 0.9% saline followed by 4% paraformaldehyde in 0.1M phosphate buffer (pH
7.4) for 5 min. The brains were post-fixed over night in 4% paraformaldehyde and subsequently
submerged in 20% sucrose in 0,01M buffered saline (PBS) at 4 ºC overnight. 40 μm serial coronal
sections were cut through the hypothalamus on a freezing sledge microtome in series of four and stored
in a solution containing ethylene glycol glycerol in 30% sucrose at -20 ºC until further processing.
Prior to the immunocytochemical steps, the sections were rinsed for 3 x 10 min in 0.01M PBS, in 1%
H2O2–PBS for 10 minutes, and in PBS with 0.3% Triton X-100, 5% swine-serum, and 1% bovine serum
albumin (BSA) for 30 min. Then the sections were incubated at 4°C for 24 hours in an antiserum against
FOS (code #94012-5) diluted 1:4,000 in PBS with 0,3% Triton X-100 and and 1% bovine serum
albumin.
After several rinses in PBS, the sections were incubated with biotinylated goat-anti rabbit IgG (1:1000,
VectorStain Elite ABC, Vector Lab., Burlingame, CA, USA) for 1 h at room temperature. Next, PBS
rinsings were followed by incubation with the avidin–biotin peroxidase complex (1:500) for 1 h at room
temperature. Intensive PBS washing was followed by washing in Tris HCL buffer (SAB, pH 7.4). Finally,
the FOS antigenic sites were visualized by processing with TRIS HCL containing 0.0625% 3,3-
diaminobenzidine tetrahydrochloride (DAB, Sigma), 0.375% hydrogen peroxide, 3.0% nickel sulphate
hexahydrate (Sigma), for 15–20 min. The heavy metal-intensification of DAB produced black staining of
the labelled nuclei. After several washings in PBS, the FOS-positive sections were incubated with
oxytocin antibodies using the same procedure as described above. The oxytocin, (diluted 1:4000)
immunoreactions were visualized in a DAB chromogen solution in Tris HCl (pH 7.4). To reach the
appropriate light brown difference between the FOS and oxy-neuropeptide stainings, the developing time
was monitored under the light microscope. Finally, the sections were mounted into 1% of gelatine
dissolved in 0.025M PBS, air-dried, coverslipped with Permount, and examined under a light microscope.
3.8. DATA ANALYSIS
3.8.1. FOS
To quantify the effects of the benzodiazepine ligands tested, the slides were blinded, and the treatment was not
revealed until all specimens had been counted.
The number of positive cells was counted in each of the two paraventricular nuclei at its middle portion in the
same tissue section cut at (bregma -1,40) according to the atlas of (Paxinos and Watson, 1997). The number of
FOS expressing cells is given as the average of these two counts.
− 29 −
3.8.2. DOUBLE LABELLING FOS / OXYTOCIN
The same blinding procedure was used as in the FOS study, in order to quantify the effect of the benzodiazepine
ligands. The number of positive FOS, Oxytocin and double labelled FOS/oxytocin was counted each of the two
paraventricular nuclei at its middle portion in the same tissue section. The number of FOS/Oxytocin reactive cells
is given as the average of these two counts.
3.9. RETROGRADE TRACING WITH CHOLERA TOXIN SUBUNIT B (CHB)

Tracing using the retrograde tracer cholera toxin subunit B (ChB), was preformed in order to map the
afferent connections of the PVN, in this study rats was used over mice, because of the larger brain,
which gives a higher resolution when mapping neurocircuits. It is also much easier to make stereotactic
injections in the rat compared to the mouse.
Cholera toxin subunit B (ChB) was purchased from (List biological), 30 adult male wistar rats breed by
the Panum animal facility weighing between 180–250 gram were housed two animals per cage under
12:12 light: dark cycle, humidity in a temperature controlled room for at least 7 days before the
experiment.
The animals were anesthetized using a combination of (fentanyl / fluanisone) 100 mg/kg mixed by the
Panum animal facility. The animals were placed in a stereotaxic frame se figure 21. After having drilled an
appropriate hole in the rat’s skull, one glass capillary microelectrode having a tip diameter less than 25 μm
was filled with the solution of ChB using a Hamilton micro pipette and inserted into the region of the
PVN. Coordinates were -1.4 mm caudal and -0.5 mm lateral to the bregma, and -7.6 mm below the dura,
according to the atlas of (Paxinos and Watson, 1997) ChB was injected iontophoretically by applying a
pulsatile current of 10 pA (0.5 seconds on; 0.5 seconds off) for 5-10 minutes.
Figure 22 shows a wistar male rat mounted on a stereotaxic frame
− 30 −
JANUARY 2006
After a survival period of 7 days the animals were re-anesthetized with mebumal and perfused
transcardially, as described above. Immunohistochemical visualization of ChB was carried out on free-
floating 40 μm sections cut in series of 6 from BNST through PVN, the sections were stored in a
solution containing ethylene glycol glycerol in 30% sucrose at -20 ºC until further processing. The ChB
tracer was visualised using a primary goat antibody from List Biological inc., the visualisation protocol
was carried out using the avidin-biotin protocol as described for the FOS study in section (5.3.). Finally,
the sections were mounted in 1% gelatine dissolved in 0.025M PBS, air-dried, coverslipped with
Permount. and examined under a light microscope.
− 31 −
3.9.1. STATUS CHB INJECTED ANIMALS
A number of animals were selected for further investigations this was done on the basis of the results achieved in
the single ChB labelling. The animals were selected on the basis of the position of the injections site.
Identification number Position of injection site Further analysis
A138 No hit No further analysis

A139 Sub PVN No further analysis
A140 PVN dorsal, caudal hit Analysis
A142 PVN hit Analysis
A147 PVN caudal hit Analysis
A148 Sub PVN No further analysis
A149 PVN caudal hit No further analysis
A151 Direct PVN Hit Analysis
A153 euthanatized No further analysis
A207 Weak dorsal rostral hit No further analysis
Table 1 list the number of rates used in the ChB retrograde neuro-tracing studies, the animals selected for further investigations were chosen
on the basis of the results archived in the single ICC (immunocytochemistry) of ChB. 4 animals were in such a condition after the operation
that they had to be euthanatized before they could be perfused and could therefore not be used for ICC.
− 32 −
JANUARY 2006
3.10. IN SITU HYBRIDIZATION, LOCATING THE ANATOMICAL DISTRIBUTION OF

THE GABAA α1, α2 AND α3 SUBUNITS.
10 naïve male wistar rats weighing 180-220 g were decapitated and the brains rapidly removed, frozen on
dry ice and stored at -80ºC.
Brains were cut through the PVN using a Leica cryostat CM3050, and a series of twenty 12 μM thick
coronal sections were obtained and subsequently mounted on Superfrost Plus slides (VWR International
ApS, Albertslund, Denmark).
In situ hybridization was performed using S35 labelled RNA probes, directed against the GABAA α1, α2
and α3 subunits (see appendix D for details). The hybridization was kindly performed by technician Pia
Rovsing Sandholm according to a protocol previously described by (Bomholt et al., 2005; Mikkelsen and
Larsen, 1993)
For each brain, a total of twenty sections were obtained, Slides were allowed to dry and stored at -80 C
until in situ hybridisation (ISH) experiments were performed using one series of sections for
hybridization of each transcript.
After in situ hybridization, the sections were then exposed to an x-ray film for 3 weeks and analysed using
a standard light table.
For additional cellular localisation of GABAA α1, α2 and α3 mRNA, the slides were subsequently dipped
in nuclear emulsion (Ilford Imaging UK Ltd, Cheshire, UK) and exposed for 4 week. In order to
distinguish the anatomical boundaries the sections were thionin counterstained according to the protocol
described by (Wittekind et al., 1979), and analysis of GABAA transcript in the PVN, SubPVN and in the
was examined using a 60 × light microscope with a bright field condenser. All analysis and development
using nuclear emulation was performed by myself.
− 33 −
3.11. IN SITU HYBRIDIZATION FOLLOWED BY IMMUNOCYTOCHEMISTRY

(CONDUCTED AT THE GENOME RESEARCH INSTITUTE, UNIVERSITY OF
CINCINNATI)
3.11.1. PROBE PREPARATION PROTOCOL
The following procedure is to my best knowledge novel, and has not been previously described, using free floating
sections.
Antisense rat GAD65 probe was produced by in vitro transcription using [35S] UTP. Plasmids containing the
GAD65 inserts (Courtesy of J.P Herman, University of Cincinnati) were linearized with restriction enzyme StuI and
transcribed using T3 RNA polymerase.
The following was mixed in an Rnase-free 1.5 µl screw-top eppendorf tube:
3.0 µl 5× transcription buffer

1.0 µl 1,0 M DTT
1.5 µl 10 M NTP Mix
2.5 µg linearized GAD65 dissolved in 2.5 µl DEPC H2O
1.0 µl DEPC treatet water
5.0 µl 35S UTP
The mixture was incubated for 1 hour at 37 ºC, then 2.0 µl of Dnase was added followed by an additional 15 min
incubation at 37 ºC. The mixture was removed from the incubation oven and 4.0 µl of tRNA (15mg/ml) was
added and was allowed to sit for 10 minutes at room temperature (RT) then an additional 79 µl of DEPC H2O and
50 µl (7.5M) NH4Ac was added.
The mixture was resuspended in 350 µl of 100% cold ethanol and stored at -80ºC for 1 hour for ethanol
precipitation.
The mixture was spun in a micro centrifuge (12000 RPM) for 20 minutes at 4ºC, the ethanol was aspired and the
resulting pellet was allowed to dry. The pellet was then resuspended in 100 µl of DEPC treated water. 1 µl was
removed and the activity counted using a scintillation counter.
3.11.2. IN SITU HYBRIDIZATION FOLLOWED BY IMMUNOCYTOCHEMICAL STAINING
Day 1: A Series of tissue sections from the ChB treated animals were taken from the -20°C freezer, and transferred
to 6 well culture plates from PGC Scientific. Sections were rinsed twice in 50mM potassium PBS (KPBS) for 5× 5
min, sections were then incubated in blocking buffer (50mM KPBS, 0.3% TX-100, 0.2 % BSA) for 30 minutes.
The sections were then incubated in the hybridization solution (see appendix d) overnight at 37 ºC in a
hybridization oven.
Day 2: The sections were rinsed 2×10 minutes in 2 × SCC at RT, they were then incubated in RNAse A (200
µg/ml) in RNAse buffer 3 hrs at 37 ºC. Next the sections were rinsed in 2×SCC, 1×SCC 0,5×SCC for 10 minutes
at RT, and then transferred to blocking buffer and incubated for 1 hour at 45 ºC the sections were next incubated
overnight at 4 ºC in blocking buffer with primary anti sera: anti cholera toxin B raised in goat diluted 1:6000.
Day 3: The sections were rinsed in 50mM KPBS containing 0.02% TX-100 for 5×5 minutes RT, then the sections
were incubated in biotinylated secondary antibody (donkey anti goat) in 50mM KPBS + 0.02 % Tx-100 for 1
hour at RT. Next the tissue was rinsed 5×5 minutes in 50 m KPBS containing 0.02 % Tx-100 at RT, and incubated
− 34 −
JANUARY 2006
in ABC solution at 1:500 in 50mM KPBS +0.1 BSA for 1 hour at RT. The sections were then rinsed for 5×5
minutes in 50mM KPBS + 0.02% Tx-100 at RT.
ChB staining was visualized using TRIS HCL containing 0.0625% 3.3-diaminobenzidine tetrahydrochloride (DAB,
Sigma), 0.375% hydrogen peroxide.
The sections were then mounted in 1% gelatine dissolved in 50mM KPB, and air-dried. In order to verify the
GAD65 hybridization the glass slides were exposed on a phosphor screen for 3 days, the film was developed using a
storm dynamics Model 400E Phosphorimager. After having verified, that the GAD65 hybridization had been
successful, and that the glass slides were subsequential, under dark room conditions the slides were dipped in
Kodak Ntb nuclear emulsion and left to expose for at total of 30 days.
At the end of the 30 days the emulsion was developed using KODAK D-19 Developer and KODAK Fixer. Then
the glass slides were coverslipped with DPX mounting solution, and examined under a light microscope.
3.12. ELECTRONIC PICTURE PROCESSING

All pictures were taken using a standard light microscope, with a digital camera attached. The following levels of
magnification were available: 5 ×, 10 ×, 20 ×, 40 ×, 60 × (×=102). All pictures taken with 20 × or higher were
taken using a bright field or dark field condenser.
The digital pictures were saved to a local computer as high resolution TIFF files. Most pictures used in this rapport
were processed using Adobe Photoshop 8.0. This was done by adjusting, colour or contrast in order to achieve the
best illustrations possible.
3.13. STATISTICS
The standard error on the graphs in the following chapters is presented with standard error mean (SEM) this has
Std .Error
been calculated as standard deviation divided by the square root of the sample size (n). .
n
The difference between groups in the following chapters has been measured using Analysis of Variance ANOVA.
The rational being, that a factorial ANOVA can examine data that are classified on multiple independent variables.
For example, a two-way ANOVA (two factors ANOVA) can measure both the difference among treatments and
among groups simultaneously. You can use more than two independent variables. A factorial ANOVA can show
whether there are significant main effects of the independent variables, and whether there are significant interaction
effects between independent variables in a set of data. Interaction effects occur when the impact of one
independent variable depends on the level of the second independent variable. A crucial component when using
ANOVA is that your data are normalized and fitting a bell shaped curve, when this has not been the case, the data
have been log transformed in order to normalize it.
All calculations has been done using SigmaStat computational software from SysStat software inc.
− 35 −
4. RESULTS
4.1. EFFECTOF ACUTE TREATMENT WITH BENZODIAZEPINES / GABAA

POSITIVE MODULATORS.
Figure 23 Time course profile of ACTH and corticosterone responses to a single dose of zolpidem (12,5 mg/kg) in mice (n = 5). The figure
illustrates the relative rise of plasma ACTH (open circles) and corticosterone (closed circles) levels at various time points after zolpidem. The
time course illustrates that release of ACTH occurs immediately after administration of zolpidem and that it precedes the release of
corticosterone.
Figure 23 shows ACTH and corticosterone responses after acute administration of zolpidem. Zolpidem
significantly and dose-dependently stimulated the peripheral components of the mouse HPA axis. Plasma ACTH
was found to be about 15-fold above baseline levels and corticosterone increased about 10-fold. The time course
illustrates a significant effect of zolpidem on ACTH release within 5 min after administration which reached its
maximal level at 10 min. Thereafter the ACTH level declined over time and reached basal levels after 120 min (Fig.
23). The time course for corticosterone was slightly delayed compared to ACTH, peaked about 60 min after
zolpidem and returned to baseline after 120 min.
This indicates that the peripheral components of the HPA axis are activated almost solely through the release of
ACTH from the pituitary, and not through direct input from the sympathetic nervous system to the adrenal cortex.
− 36 −
JANUARY 2006
RESULTS NeuroSearch
Figure 24 The effect of increasing doses of the non-selective benzodiazepine ligands diazepam (closed squares) and zopiclone (closed
triangles) on plasma corticosterone levels in mice compared to the α1 selective zolpidem (closed circles) and the α2,3 selective agonist L-
838,417 (open circles). The four drugs are given as a single dose and the mice are sacrificed 60 min after the treatment. The data represent
mean ± S.E.M. of minimum 5 mice per group. There was a significant (p < 0.05) difference between the effect of non-selective and selective
benzodiazepine receptor ligands. #p < 0.05 vs zolpidem and *p < 0.05 vs vehicle. Two-way ANOVA and Student-Newman-Keuls test.
It is demonstrated that the rise in plasma corticosterone is drug and dose dependent. Treatment with zolpidem
which is an α1 selective agonist increased plasma corticosterone at a minimum effective dose of 0.5 mg/kg reaching
a maximum at 12.5 mg/kg. Zolpidem was effective and increased circulating corticosterone up to about 800
ng/ml. (see figure 24).
The potencies and efficacies of diazepam and zopiclone were significantly lower than for zolpidem (Fig. 24). Thus,
a rise in plasma corticosterone was seen after 2.5 mg/kg and the maximal level of corticosterone was in the range
of 400 ng/ml.
Notably, while zolpidem induced a slightly higher level of corticosterone at 25 mg/kg compared to 12.5 mg/kg, the
two non-selective benzodiazepine ligands produced a slightly lower level of corticosterone. The two non-selective
benzodiazepine ligands produced a different dose-response curve with zopiclone being slightly stronger than
diazepam.
The α2,3,5 selective agonist L-838,417 was ineffective in activating the peripheral component of the mouse HPA axis
at up to 25 mg/kg where nonspecific effects probably did induce a slight stress hormone response.
− 37 −
4.2. INTERACTION INHIBITION STUDIES

In order to investigate the relationship between activation of α1 and non– α1 containing GABAA receptors,
combinations of L-838,417 (12.5 mg/kg i.p.) or vehicle was given 15 min before a single dose of zolpidem i.p at
increasing concentrations (0.5, 2.5, and 12.5 mg/kg) L-838,417 clearly inhibited the effect of zolpidem at all doses
(see figure 25).
In order to determine that the inhibitory action of L-838,417 is not solely due to the compound’s antagonistic
properties at the GABAA α1 receptor, a combination of the two non-selective benzodiazepine ligands diazepam and
zopiclone was tested together with zolpidem. In this case mice were treated with vehicle, diazepam (0.5 mg/kg i.p.)
or zopiclone (0.5 mg/kg i.p.), 15 min before administering the mice with 0.5 mg/kg zolpidem (see figure 26) or 2.5
mg/kg zolpidem (see figure 26). The animals were sacrificed 60 min after zolpidem administration and trunk blood
was taken for corticosterone measurements. All single and combined dosing experiments were carried out twice. At
0.5 mg/kg diazepam or zopiclone before 0.5 mg/kg zolpidem, both diazepam and zopiclone inhibited the effect of
zolpidem, although diazepam was more potent than zopiclone. Neither diazepam nor zopiclone were able to
inhibit the effect of 2.5 mg/kg zolpidem (see figure 27).
1200
1000
Corticosterone [ng/ml]
800
**
600
**
400
**
200
0
Veh 0.5 2.5 12.5 Zolpidem (mg/kg)
Figure 25 The animals pre-treated with vehicle followed by Zolpidem are shown as closed circles and the animals pre-treated with L-
838,417 are shown as open circles. The values are given as mean (n = 5) ± SEM. Significance between groups was determined by two-way
ANOVA and **p < 0.01. Two-way ANOVA and Student-Newman-Keuls test
− 38 −
JANUARY 2006
RESULTS
ng corticosterone / ml plasma NeuroSearch
*
**
**
Figure 26 Pre-treatment with 0.5 mg/kg diazepam, zopiclone or 12.5 mg/kg L838,417 15 min before administration of 0.5 mg/kg zolpidem
(all compounds were administered i.p.) inhibit release of corticosterone. As negative control mice were treated with vehicle twice, and as
positive control zolpidem was administered 15 min after vehicle. ** p < 0.01; * p < 0.05 vs zolpidem. The data represent mean ± S.E.M. of
minimum 5 mice per group.
It is seen that pre-administration of both diazepam as well as zopiclone at a relatively low dose 0.5 mg/kg (i.p)
significantly inhibits the effect of zolpidem, compared to animals pre-administered only with vehicle. At this dose
of zolpidem, it seems that diazepam and to a lesser extent zopiclone is almost as effective as L-838,417 in inhibiting
the effect of zolpidem on the murine HPA axis.
ng corticosterone / ml plasma
**
Figure 27 Pre-treatment with 0.5 mg/kg diazepam, zopiclone or 12.5 mg/kg L838,417 15 min before administration of of 2.5 mg/kg
zolpidem does not inhibit the release of corticosterone compared to vehicle + zolpidem. L-838,417 still significantly inhibits the effect of
zolpidem even at this high dose. As negative control mice were treated with vehicle twice, and as positive control zolpidem was
administered 15 min after vehicle. ** p < 0.01 vs. zolpidem, Two-way ANOVA and Student-Newman-Keuls test . The data represent mean
± S.E.M. of minimum 5 mice per group.
− 39 −
4.3. EFFECTS
OF BENZODIAZEPINE LIGANDS ON THE INDUCTION OF FOS IN
THE PARAVENTRICULAR NUCLEUS.
Figure 28 shows the distribution of FOS in the paraventricular region in naive mice (A), and after administration with vehicle (B), L-838,417
(C), zopiclone (D), diazepam (E), and zolpidem (F). All drugs are given in a dose of 12.5 mg/kg 60 min before sacrifice. Scale bar =100 μm.
All pictures were taken using a digital camera mounted on a light microscope, at a 20 × resolution.
4.4. EFFECTS
OF BENZODIAZEPINE LIGANDS ON FOS EXPRESSION IN THE
PARAVENTRICULAR NUCLEUS
FOS induction was studied in male mice treated with zolpidem, L-838,417, zopiclone, or diazepam all at 12.5 mg pr
kg i.p., as controls a number of vehicle treated animals were also investigated.
Only a small number of FOS expressing cells was observed in the PVN of these control groups see (figure 28 A
and B). As for zopiclone, diazepam and zolpidem all compounds produced a strong increase of FOS containing
nuclei in the PVN. The activated cells were present in an area overlapping with the CRH containing medial
parvocellular subportion of the paraventricular nucleus (Kiss, 1988), but also in the magnocellular portion of
nucleus, se figure 28.
L-838,417 did not produce any FOS response in the treated animals compared to vehicle, furthermore diazepam
produced a significantly lower response compared to that of zolpidem. This finding corresponds well to the data
observed on the peripheral components of the HPA axis (corticosterone data shown in figure 24).
− 40 −
JANUARY 2006
RESULTS NeuroSearch
***
Figure 29 Quantitative analysis of FOS expression in the paraventricular nucleus after acute treatment with benzodiazepine ligands. The
effect of FOS induction was analysed 60 min after single administration of drugs at 12.5mg/kg i.p., the same dose that produces a maximal
increase of corticosterone in plasma. The values are given as the mean number of cells expressing FOS in one section of the middle part of
the paraventricular nucleus (mean-±S.E.M.). ***P <0.001; *P <0.05 compared to zolpidem. ANOVA nonparametric Tukey’s multiple
comparison test.
Figure 29 clearly show that zolpidem, diazepam and zopiclone all induce FOS in the PVN while there is no
induction after vehicle and L-838,417.
− 41 −
4.5. IDENTIFYING FOS EXPRESSING NEURONS IN THE PVN

An effort was made to investigate whether the FOS stained neurons in the PVN was co-localized with the
oxytocin producing neurons (Wotjak et al., 2001).
This was done using a double immunocytochemical protocol with nickel enhancement as described in section
(9.4.).
Oxytocin stained
cell
FOS positive cell
Figure 30 Shows a 40 μm section of the lateral part of the PVN cut at bregma (-1,40) from a male NMRI mouse treated with 12,5 mg/kg
zolpidem 1 hour before sacrifice. The slide has been stained for FOS using nickel enhancement and for Oxytocin using a standard ABC,
DAB protocol, as described in section 3.8.2.
Figure 31 Dual-labeling for FOS and oxytocin; after administration of 12,5 mg/kg Zolpidem, or vehicle 1 hour before sacrifice. The values
are given as the mean number of cells expressing FOS, oxytocin or co-localization in the middle part of the paraventricular nucleus (mean-
±S.E.M.). of minimum 8 mice per group. All brain sections were counted using a light microscope with a bright field condenser, FOS and
oxytocin and FOS/oxytocin was distinguished due to the colour difference achieved during immunocytochemistry.
− 42 −
JANUARY 2006
RESULTS NeuroSearch
Figure 31 shows that there seem to be no significant co-localization between oxytocin and FOS in the animals that were treated with 12.5
mg/kg zolpidem.
4.6. RESULTS FROM IN SITU HYBRIDIZATION

In situ hybridization using radioactive probes directed against the GABAA α1 α2 and α3 was performed on 12 μm
sections cut through the PVN as described in section 3.10.
It would be out of the scope of this project to make a detailed map of the global distribution of the different
GABAA subunits, for review see (Fritschy and Mohler, 1995), and the following sections will therefore only
concentrate on the PVN and SubPVN area.
A B C
α1 α2 α3
Figure 32 All tissue sections are from mice and are cut at the level of the PVN (bregma -1.40 mm). (A) shows a picture of the distribution of
radioactive labelled GABAA α1 m-RNA in the PVN, as it is seen; there is a very limited amount of α1 m-RNA expression in the direct
surrounding of the PVN, while there is a very heavy staining in the cortex and the BNST. (B) Shows the distribution of the GABAA α2
receptor in the PVN, It is clearly seen that there is a very distinct and heavy staining in the medial parvocellular section of the PVN as
indicated by the arrow. (C) Shows the distribution of the α3 subunit there is little or no distribution in the PVN but there is heavy staning in
the amygdale as is indicated by the arrow.
Although the pictures shown her have a relatively poor resolution, they still show that there is a very different
distribution of the GABAA α1, α2 and α3 subunit in the different brain areas and it is clear that the only subunit that
is expressed to any significant extent in the PVN itself, is the α2 while the α1 subunit is expressed manly in the
limbic areas and the α3 in the amygdala.
4.7. DETAILED ANALYSIS

Analysis using standard x-ray film does not allow for a detailed cellular resolution after in situ hybridization, so for
additional information on the cellular localisation of GABAA α1, α2 and α3 mRNA, tissue sections were
subsequently dipped in nuclear emulsion as described in section 3.10.
The sections were then thionin counterstained and analysis of GABAA transcript in the PVN and SubPVN area
was examined using a light microscope 60 × objective with and a bright field condenser.
− 43 −
4.8. GABAA ALPHA 1 DISTRIBUTION IN THE PVN AND SUBPVN AREA
1 2 3
A
Figure 33 (1) is a 10 × magnificationof the PVN of a naïve wistar rat cut at (bregma -1.40 mm). The slide has been treated with a probe
directed at the GABAA α1 subunit and dipped in nuclear emulsion, before counter staining with thionin. The PVN borders has been drawn
using Adobe Photoshop , and area A and B represent respectively the medial parvocellular part of the PVN, and the sub paraventricular
area. (2) Shows a 60 × picture of area A from 33 (1), the black spots represents labelled GABAA α1 m-RNA. It seen that there is a very
limited amount of black spots, and that the ones that are present seems dispersed and not located over the thionin stained neurons. Figure
33 (3) shows a 60 × picture of area B from figure 32(1), the black spots represents labelled GABAA α1 M-RNA. It is seen that there is heavy
staining located over the counter stained neurons. The arrows point to magnocellular neurons containing significant amounts of GABAA α1
m-RNA.
Figure 33 (1,2 and 3) shows that the PVN itself seems to be deprived of GABAA α1 expressing neurons while the
immediate surroundings (the SubPVN) seems to be heavily innervated by neurons containing the α1 subunit. This
is in good agreement with earlier observations reviewed by (Herman et al., 2002) that the PVN is surrounded by a
GABAA “ring”, although the subunit composition of the GABAA receptor in the SubPVN area has not earlier been
investigated.
1 2 3
A
Figure 34 is a10 × magnification of the PVN of a naïve wistar rat cut at (bregma -1.40 mm). The slide has been treated with a probe
directed at the GABAA α2 subunit and dipped in nuclear emulsion, before counter staining with thionin. The PVN borders has been drawn
using Adobe Photoshop, and area A and B represent respectively the medial parvocellular part of the PVN, and the sub paraventricular area.
(2) Shows a 60 × picture of area A from 34 (1), the black spots represents labelled GABAA α2 m-RNA. It seen that there is a very intense
staining over the magnocellular neurons in the PVN. Figure 34 (3) shows a 60 × picture of area B from figure 34(1) the black spots
represents labelled GABAA α2 M-RNA. It is seen that there is no significant staining in the SubPVN section
− 44 −
JANUARY 2006
RESULTS NeuroSearch
Figures 32 through 34 shows that the medial part of the PVN seems to consist primarily of neurons that contain
GABAA receptors with the α2 subtype, and the area surrounding the PVN (the SubPVN area) is populated by
neurons containing the GABAA α1 subtype.
The distribution of the GABAA α3 subunit was also investigated in the PVN and SubPVN area, α3 m-RNA did not
seem to be distributed to any significant degree neither in the PVN nor in the SubPVN area, and the role of the α3
subunit in the control of the PVN, can not be determined with the experimental data presented in this report.
4.11. RESULTS FROM CHB RETROGRADE NEURO-TRACING.

A total of 30 Wistar male rats were one hour before the tracing studies commenced injected with 10 mg/kg
zolpidem i.p. They were then injected with ChB in the PVN as described in section 3.9.
A total of 30 animals were used for the experiment, of these four died or had to be euthanatized before they could
be perfused, due to post operative complications.
The 26 remaining animals were all perfused transcardially. Then immunocytochemical visualization of ChB was
carried out on free-floating 40 μm sections cut in series of 6 from BNST through PVN as described in section 3.9.
Of the 26 animals that were examined, there were only two direct injection sites in the PVN, the remaining
injections were located ether in the immediate surroundings of the PVN or in other hypothalamic regions, and in
rare cases it did not seem like any ChB had been released at all (see table 1).
The reason for the relatively low success rate can be ascribed to firstly the small size of the PVN, and secondly the
fact that there is a diversity between individual Wistar male rats. The stereotaxic coordinates of the PVN are based
on statistics and it is therefore not 100% useful when working with a outbred strain like Wistar (Paxinos and
Watson, 1997).
In the following section there will be a detailed analysis of one of the two animals, where we managed to put the
glass electrode directly in to the PVN. The projections of the two were the same. Again it would be beyond the
scope of this project to analyse all of the projections of the PVN, therefore it was chosen to focus on the bed
nucleus of stria terminalis (BNST). BNST is part of the extended amygdala that is involved in the anxiety/stress
response, and that has been the main reason for choosing this particular anatomical structure as an area of focus
(Weller and Smith, 1982).
− 45 −
A ChB hit in the PVN

visualized by B
immunocytochemical
stain
C D
Figure 35 A, B, C and D are all tissue sections from the same animal, A151 (see table 1). A shows a 4 × magnification of the injection site in
the PVN, visualized by immunocytochemistry. The tissue is cut at (bregma 1,40) according to the atlas of (Paxinos and Watson, 1997). It is
clearly seen that the ChB has been administered directly into the PVN. (B) and (C) shows the levels of the BNST the tissue has been cut at
(bregma -2,12 mm) and (Bregma -0,26 mm) respectively it is seen that the BSTMA, BSTL and BSTMV all have numerous projections to the
PVN. (D) Shows a 60 × magnification of the BSTMA area from figure 35 (B).
A151 and A142 represent the only true PVN hits I was able to make, but it is also interesting to further investigate
the hits in the immediate surroundings of the PVN.
In order to investigate the projection of the BNST in detail the slides were investigated using a light microscope
and a 60 × magnification. The stained neurons were then plotted using an electronic PDF version of the atlas of
(Paxinos and Watson, 1997). It would serve no purpose showing pictures from all relevant areas therefore the
following will only show selected pictures, and this only for the sake of illustration, since it is not possible to get
microscope resolution using digital pictures.
− 46 −
JANUARY 2006
RESULTS NeuroSearch
A B
Figure 36: (A),(B) and (C) are schematic presentation of the pictures shown in figure 35, drawn in red is the most prominent projections in
the area of which they occur, the schematic presentation is borrowed from the atlas of (Paxinos and Watson, 1997) and edited in Adobe
Photoshop. (A) represents the injection site in the PVN, while (B) and (C) represents the neurons in the different levels of the BNST that
project to the PVN.
Figure 35 and 36 clearly shows that most of the BNST have neurons that project to the PVN. Although it seems
like the BSTMA and BSTL are the areas that have the highest number of PVN projecting Neurons.
The results from A151 (see figure 35) were the same as those found using A142, the data from A-142 is therefore not
shown. This was the case even though the ChB injection in A-142 is more caudal than the hit in A-151. This
indicates that the projections of the BNST are well disbursed through the entire PVN.
− 47 −
4.11.1. 6.11.1. PROJECTIONS OF THE BSNT TO THE SUBPVN AREA
As described in the introduction to section (3.9) it was only possible to make 2 direct PVN hits with the ChB
tracer. As a consequence of this, there were several “misses” that hit the areas in the vicinity of the PVN, some of
these came very close and hit the SubPVN area (see area ”B” figure 34(1)). Since it seems that this area is of
particular importance, when it comes to the control of the PVN, the following section will therefore deal with the
BNST projection to the SubPVN area.
A B
C D
Figure 37 (A) Shows a 4 × magnification of a SubPVN ChB injection, the tissue section is cut at (Bregma -2,12) according to the atlas of
(Paxinos and Watson, 1997). This slide will hence forward be referred to as A-140 (se table 1). (B) Represents a schematic drawing of the
injection site. (D) and (E) are schematic presentation of the projections from the BNST to the SubPVN.
The data from the retrograde ChB tracing presented in figures 35 through 38 shows that the projections from the
different anatomical sub nuclei of the BNST, to the PVN and SubPVN areas respectively are quite similal, and it
seems that the projections of the BNST innervate these to area with almost equal intensity. Although not
quantifiable with the present data, it does appear that while the BSTMV and BSTLP have numerous projections to
the PVN, these areas only have a limited number of projections to the SubPVN area. While the BSTMA and the
BSTL are saturated with neurons projecting to both the PVN and the SubPVN area.
− 48 −
JANUARY 2006
RESULTS NeuroSearch
4.12. DUAL IN SITU HYBRIDIZATION / IMMUNOCYTOCHEMICAL STAIN

The purpose of this study was to identify which of the BNST, PVN projecting neurons are GABAergic, this was
done by doing double labelling for ChB and GAD65, according to the protocol described in section 3.11.
A number of sections from the different animals: A140, A142, A151 were used for the experiments (se table 1 for
details). Unfortunately I did not succeed in performing successful combined staining with tissue from neither A142
nor A151, which represented the animals with the best PVN hits. However I was successful in producing dual
staining on tissue from A140 see (figure 37(A)).
Figure 38 shows a 5 × magnification of A140, the section is cut at (bregma 0.20 mm) according to the atlas of (Paxinos and Watson, 1997)
the black surface of the tissue is do to massive staining of the emulation by radioactive labelled GAD65 m-RNA.
Figure 39 shows a 60 × magnification of dual ChB/GAD65 labelled cells in the BSTMA, the light brown cells are ChB positive cells and
dark stains in the emulation is radioactive labelled GAD65 mRNA. The arrows indicate possible dual labelled cells.
Figure 39 shows that there is a population of neurons in the BSTMA that has projection to the SubPVN. In this
area there is also an intense GAD65 m-RNA labelling, making it highly feasible that the SubPVN projecting
neurons in the BSTMA are GABAergic. The BSTL also had projections to the SubPVN, but the GAD65 staining
in this area was less intense than in the BSTMA. There was also some ChB staining in the BSTMV as shown in
section 6.11.
− 49 −
Figure 40 shows a 40 × magnification of the BSTMA and BSTL from A140 using dark field filter. The Tissue section is cut at (bregma 0.20
mm) according to the atlas of (Paxinos and Watson, 1997). The white areas areas are areas with intense GAD65 labelling.
Figure 40 is taken with a dark field filter using this filter it is clear to se that there is intense staining of radioactive
labelled GAD64 in the BSTMA and BSTL, making a good case for the assumption that a large portion of the
neurons in the BNST are GABAergic.
− 50 −
JANUARY 2006
DISCUSSION NeuroSearch
5. DISCUSSION
Activation of the HPA axis due to external or internal challenges is important to the survival of species as well as
the individual. Just as important as the stimulatory pathways are the inhibitory pathways, which limit the magnitude
and duration of the stress response.
The major mechanisms responsible for inhibition of stress are the steroid-dependent negative feedback via steroid
receptors in the hippocampus and the direct neural inhibition of paraventricular nucleus neurons driven by the
neurotransmitters GABA and substance P (Kovacs et al., 2004; Jessop et al., 2000).
The fact that high doses of benzodiazepines stimulate the HPA axis was first shown in 1970s by (Havard et al.,
1972), this has since been rediscovered by numerous others (McElroy et al., 1987; Vargas et al., 2001) .
But to date the seemingly controversial pharmacological mechanism by which this happens has not been clear. The
data, presented in this thesis, show that acute administration of GABAA receptor subtype selective and non-
selective drugs have different effects on the murine HPA axis.
The GABAA α1 selective agonist zolpidem induces a massive secretion of ACTH and corticosterone in the mouse
both of these hormones are key components of the biological stress response (Jacobson, 2005). Zolpidem also
upregulates the transcription factor FOS in the medial part of the PVN, this is a good indication that zolpidem
exerts its effect at the level of the brain (Hoffman and Lyo, 2002).
Administration of the two non selective benzodiazepine ligands diazepam and zopiclone also evoked a dose
dependent corticosterone response, although to a lesser extent than zolpidem. The data obtained from
administration of diazepam correlates well with those found in the literature (Vargas et al., 2001). Both of these
compounds also induced FOS in the PVN.
Meanwhile the GABAA α2,3,5 selective agonist and GABAA α1 antagonist L-838,417 (McKernan et al., 2000) was
ineffective in activating a corticosterone response, the same was true with regard to a FOS response in the PVN.
Pre-administration of L-838,417 15 minutes prior to a single dose of zolpidem clearly inhibits zolpidem’s effect on
the murine HPA axis compared to animals treated with vehicle prior to zolpidem. This effect could in theory be
ascribed to the fact that L-838,417 acts as an antagonist at the GABAA α1 receptor (McKernan et al., 2000). In
order to test whether this was the case, we also pre-administered the animals with either of the two non-selective
benzodiazepine ligands zopiclone or diazepam, prior to a single dose of zolpidem. Both diazepam and zopiclone
inhibited the effect of zolpidem on the murine HPA axis, compared to vehicle followed by zolpidem. This
indicates that the powerful inhibiting effect of L-838,417 is not solely due to the compound’s antagonistic
properties at the GABAA α1 receptor.
Since zolpidem is selective of the GABAA receptors containing the α1 over those containing the α2, α3 or α5 and
zopiclone and diazepam are non-selective (Fleck, 2002; Pritchett and Seeburg, 1990), it might be proposed that that
the α1 and non- α1 containing GABA neurons oppose each other in stimulating the CRH containing neurons of the
paraventricular nucleus.
It is most likely that the opposing GABAergic neurons in the PVN contain the α2 subunit, the rationale being that
the α2 subunit is highly expressed in the CRH synthesising neurons in the PVN. This was verified using in situ
hybridization, the same results have also been found by others (Wisden et al., 1992). This clearly indicates that the
PVN itself is comprised of neurons with GABAA α2 surrounded by a “zone” of neurons with the GABAA α1
receptor. That the PVN is surrounded by a zone of GABA neurons has been suggested earlier by (Cullinan, 2000;
Pirker et al., 2000) although the GABAA α subunit composition of this zone has not been clarified before this.
Since L-838,417 has almost equal affinity for the GABAA α2, α3 and α5,(McKernan et al., 2000) there is a theoretical
possibility that the opposing neurons in the PVN could contain the α5 subunit. According to (Pirker et al., 2000;
Fritschy and Mohler, 1995) the PVN does contain minute amounts of the α5 receptor shown by in situ
hybridization and immunocytochemistry. We have at our lab not been able to confirm this (data not shown), the
same being the case for works published by (Cullinan, 2000). If α5 is present in the PVN both (Fritschy and
Mohler, 1995; Pirker et al., 2000) agree that it is only in minute amounts compared to α2. This clearly supports the
proposed mechanism in where the apposing neurons are of the GABAA α2 phenotype.
− 51 −
Figure 41 Schematic illustration of the model proposed for the action of selective and non-selective benzodiazepine
ligands on the CRH neurons in the paraventricular nucleus. Non-selective benzodiazepine receptor ligands act on
at least two different sites. One is outside the paraventricular nucleus and is disinhibiting the CRH neurons. The
other is inside the nucleus leading to a direct inhibition of the HPA output. The balance between activation of
these two targets determines the release or inhibition of stress hormone secretion to the blood.
What the model suggests is that the CRH containing neurons in the PVN, that contain the α2 subunit, is under
tonic inhibitory control from GABAA α1 neurons lying adjacent to the PVN, in the SubPVN. When the neurons in
the SubPVN are inhibited with zolpidem, they are no longer able to exert their tonic inhibition of the neurons in
the PVN. This creates a disinhibiting effect causing the CRH containing neurons in the PVN to depolarize. This is
a possible explanation why high doses of benzodiazepines, causes the production of corticosterone in rodents and
cortisol in humans.
The setup presented in this thesis could also be used as an in vivo assay for a GABAA α2 effect, something that
could be very useful during drug discovery; one example could be when searching for anti anxiety drugs
(McKernan et al., 2000).
− 52 −
JANUARY 2006
5.1. GABA NEUROCIRCUITS CONNECTING THE BED NUCLEUS STRIA

TERMINALIS TO THE PARAVENTRICULAR NUCLEUS
The BNST is part of the extended amygdala and is involved in anxiety and fear conditioned behaviours (Davis,
1998a) this limbic forebrain structure links regions such as the amygdala and hippocampus with hypothalamic and
brainstem regions and thereby controls vital homeostatic functions (Weller and Smith, 1982; Moga et al., 1989;
Cullinan et al., 1993; Davis, 1998). It is therefore not surprising that this structure has numerous connections to the
PVN.
The attempt to map the afferent projections to the PVN, using the retrograde tracer ChB, verified what has been
found by others, namely that the BNST has numerous projections to the PVN (Coolen and Wood, 1998; Davis
and Shi, 1999). The results obtained show that much of the input to the PVN seem to originate in either in the
BSTMA, BSTMV or the BSTL sub nuclei of the BNST. The fact that the BSTMA has projections to the PVN has
previously been shown by (Okamura et al., 1990). But that the BSTL and BSTMV also have projections to the
PVN has to the best of my knowledge not previously been described.
Using a novel histological method, an attempt was made to investigate whether the PVN projecting neurons in the
BNST co-expressed GAD65. Unfortunately we ran out of tissue from the two animals with ChB injections directly
into the PVN, so we only managed to do the dual in situ hybridization/immunocytochemical stain described in
section 6.12. on an animal with a ChB injection in the SubPVN area. Using the tissue from this animal the only
conclusive results that can be made, is that the BSTMA and the BSTL have neurons that seem to express massive
amounts of GAD65 m-RNA, this correlates well with results from (Herman et al., 2002; Okamura et al., 1990) who
also agree that a great number of the PVN projecting neurons in the BNST are GABAergic.
The massive expression of GAD65 m-RNA in the BNST makes it close to impossible to tell whether neurons are
dual stained or not. With this in mind, it would be very interesting to do dual labelling for ChB and one of the
GABAA α subunit, and in this way determine the subunit composition of the PVN projecting neurons in the
BNST.
5.2. METHODOLOGICAL CONSIDERATIONS

5.2.1. FOS
The choice of FOS as a marker of neuroactivity has been rationalized in section 3.6. And zolpidem, diazepam and
zopiclone all induce FOS in the medial parvocellular part of the PVN, the activation of these neurons probably
happen through an increase of intracellular cyclic AMP levels or activation of voltage sensitive calcium channels
which lead to the phosphorylation of CREB and c-FOS induction (Thompson et al., 1995). This fits well with the
fact that these CRH containing neurons have projections to the median eminence where they release CRH in to the
hypothalamohypophyseal portal vascular system. CRH then stimulate the production of pituitary ACTH and the
subsequent secretion of corticosterone from the adrenal cortex.
There are other markers of one could have considered some examples are c-jun, jun-B, jun-D that lie upstream
from FOS (Curran and Franza, Jr., 1988a; Cohen et al., 1989), or Zif268 (Worley et al., 1991) but the fact that FOS
is more often associated with depolarization makes it a superior marker, especially when working with
neuroendocrine systems (Hoffman and Lyo, 2002).
Yet another thing that could have been considered is in vitro electrophysical recording studies, where one could
have used culture PVN slices, and in this way make direct electrophysical measurements after application of the
different compounds used in this study.
− 53 −
5.2.2. DRUGS
All drugs used except zopiclone and zolpidem were of the benzodiazepine family, the pharmacokinetics of a single
dose of diazepam at 20 mg/kg i.p. show a Cmax after approximately 10 minutes. The serum concentrations then
show a biexponential decline. The first rapid decline is characterized by a t½ (a) of 9.0 ± 2.1 min, the subsequent
decline (b) is characterized by a t½ of 72.0 min ± 10.5 min (Walker et al., 1998). Zolpidem also reaches a Cmax
within the first 15 minutes (Trenque et al., 1994), zopiclone although structurally unrelated to the benzodiazepines
do show similar pharmacokinetics (Goa and Heel, 1986b). Since L-838,417 belongs to the same chemical class as
diazepam it is expected to show similar pharmacokinetics but this was not confirmed.
The fact, that exact pharmacokinetic measurements were not conducted during the experiments, could
pose a problem if we had done the corticosterone measurements in as shorter timeframe. But after 60
minutes it is a reasonable assumption that all of the compounds have reached Cmax..
5.2.3. COMBINATION OF DRUGS
It could easily be argued, that the mechanism by which L-838,417 inhibits the effect of zolpidem on the
murine HPA axis, is simply due to L-838,417 antagonistic binding at the GABAA α1 receptors
(McKernan et al., 2000). However diazepam and zopiclone which are both full agonist at the α1 receptor
(McKernan et al., 2000; Sanger, 2004; Graham et al., 1996), for review see (Sieghart, 1995), at doses that
do not themselves stimulate any significant corticosterone response, also inhibit the effect of zolpidem,
making a strong case for the involvement of a non – α1 GABAA receptor.
5.2.4. ANIMAL MODELS
Adult NMRI male mice were used. Male mice were used in order to avoid any influence that the estrus cycle in
females might have on the corticosterone levels. Also all experiments were carried out 4 hours after the onset of
the light fase in order to avoid influence from the circadian rhythm that causes a natural variation in corticosterone
during the day (Able, 2005).
Whether or not the result can be transferred to higher mammals is hard to say. There is substantial proof in the
literature that the HPA axis is a highly conserved mechanism among mammals (Orchinik, 2005; Kandel et al.,
2000). And the fact that high doses of diazepam has been shown to increase basic levels of cortisol in human
subjects (Havard et al., 1972), indicate that a similar mechanism might be at play in higher mammals including
humans.
5.2.5. IN SITU HYBRIDIZATION
In situ hybridization was performed using radioactive antisense probes directed against the GABAA subunits α1, α2
and α3. in accordance with a protocol previously described by (Bomholt et al., 2005; Mikkelsen and Larsen, 1993).
Although nuclear emulation enables the detection of mRNA at a cellular level, it is not readily quantifiable, which
does pose a potential problem when comparing results from different brain areas. However the method is quite
useful as a detection method, when proving that a particular gene is expressed in a certain area. One could have
measured the intensity of staining on the x-ray film (Matsuoka et al., 1992), however this was not possible when
working with very restricted anatomical areas such as the SubPVN. But as seen with the α2 staining which were
localized exactly in the PVN there was no difficulty determining that this particular subunit is expressed in high
amounts in the PVN compared to other brain regions, and this has also been verified by others
− 54 −
JANUARY 2006
5.2.6. RETROGRADE TRACING USING CHOLERA TOXIN SUBUNIT B (CHB)
Even though the results achieved during the mapping of the afferent projections to the PVN, is in good agreement
with those found in the literature (Herman et al., 2002; Okamura et al., 1990), there are several methodological
concerns that need to be addressed when analysing the results.
The first and primary concern is that ChB may not be a true retrograde tracer. During the immunohistochemical
analysis of tissue from A142 and A151, the two animals with direct PVN injections, we found a small number of
ChB positive cells in the median eminence. We know from the literature that the magnocellular CRH neurons in
the PVN have several projections to the median eminence, and the median eminence does not have projections to
the PVN (Herman et al., 2003; Larsen et al., 1994; Swanson and Sawchenko, 1980). This indicates that the ChB
tracer may also to some degree be transported in an anterograde fashion, this of course dose pose some problems
when mapping afferent neurocircuits. This could perhaps be helped by choosing a different tracer, for example is
Flurogold (Nogueira et al., 2000; Ju et al., 1989).
Another concern that need to bee addressed is leakages, during the retraction of the glass pipette from the PVN.
During this retraction, a small amount of ChB leaked in to the tissue lying above the PVN, most probably the
reuniens thalamic nucleus (Paxinos and Watson, 1997) this could give some false positive cells elsewhere in the
brain. Again this is not very likely to be a problem due to the very small amount of leakage.
5.2.7. DUAL IN SITU HYBRIDIZATION/IMMUNOCYTOCHEMICAL STAIN
The method described in section 16.0. was used to identify neurons projecting to the PVN, that at the same time
express GAD65.
In the method two distinct techniques was combined namely immunocytochemistry and in situ hybridzation.
The technique is quite useful and clearly holds great potential, there are however some concerns when working
with this technique. One major concern is that the staining seen in the nuclear emulsion, actually lie in a layer above
the cells, and not on the cells themselves, this makes it hard to identify double labelled cells with 100 % certainty.
The specific probe used in this study the GAD65 probe, has a very high level of expression in the BNST, this makes
it hard to identify if the GAD65 staining over the ChB immunolabelled cells actually originate from the ChB
positive cells. One could solve this problem using a similar technique combining a fluorescent in situ hybridization
with an immunoreaction for ChB, in this way both markers would physically be on the cell.
− 55 −
6. CONCLUSION
It can be concluded that the different effects of benzodiazepines on the basal level of HPA activity in rodents are
due to the basal difference between GABAA α1 and non-α1-dependent (presumably α2) modulation of the CRH
neurons in the paraventricular nucleus and/or sub paraventricular area.
It can be concluded that the corticosterone induction, produced by the compounds used in this study, with the
outmost certainty has a central effect in the brain, and more specifically in the paraventricular nucleus of the
hypothalamus.
The results presented in this thesis does not conclusively determine what types of neurons are activated in the
paraventricular nucleus after administration of zolpidem, however the preliminary results suggest, that the FOS
expressing neurons does not co-express oxytocin.
It can be concluded that the bed nucleus stria terminalis has several projections to the paraventricular nucleus and
to the sub paraventricular area. It can also be concluded that the areas with the seemingly highest number of
projections are the bed nucleus of the stria terminalis, medial division, anterior part and the bed nucleus of the stria
terminalis, lateral division. This is however only an estimate.
On the basis of the results presented in this thesis, it is not possible to determine if the PVN projecting neurons in
the bed nucleus stria terminalis are GABAergic or not. The data do however show that the
bed nucleus of the stria terminalis, medial division, anterior part and the bed nucleus of the stria terminalis, lateral
division express high amounts of GAD65 making it highly likely that a large portion of the neurons in these
anatomical subdivisions could be classified as being GABAergic. This however is not proof, that the neurons
projecting to the PVN have the same phenotype.
− 56 −
JANUARY 2006
REFERENCES NeuroSearch
7. REFERENCES
Able KBeal, 2005. Handbook of stress and the brain, vol. 1.
Arborelius L, Owens MJ, Plotsky PM, Nemeroff CB, 1999. The role of corticotropin-releasing factor in depression
and anxiety disorders. J. Endocrinol. 160: 1-12.
Arnold FJ, De Lucas BM, Shiers H, Hancock DC, Evan GI, Herbert J, 1992a. Expression of c-fos in regions of the
basal limbic forebrain following intracerebroventricular corticotropin-releasing factor in unstressed or stressed male
rats. Neuroscience 51: 377-390.
Bains JS, Ferguson AV, 1995. Paraventricular nucleus neurons projecting to the spinal cord receive excitatory input
from the subfornical organ. Am J Physiol 268: R625-R633.
Barnard EA, Skolnick P, Olsen RW, Mohler H, Sieghart W, Biggio G, Braestrup C, Bateson AN, Langer SZ, 1998.
International Union of Pharmacology. XV. Subtypes of gamma-aminobutyric acidA receptors: classification on the
basis of subunit structure and receptor function. Pharmacol. Rev. 50: 291-313.
Benson JA, Low K, Keist R, Mohler H, Rudolph U, 1998. Pharmacology of recombinant gamma-aminobutyric
acidA receptors rendered diazepam-insensitive by point-mutated alpha-subunits. FEBS Lett. 431: 400-404.
Bomholt SF, Mikkelsen JD, Blackburn-Munro G, 2005a. Normal hypothalamo-pituitary-adrenal axis function in a
rat model of peripheral neuropathic pain. Brain Res. 1044: 216-226.
Bormann J, 2000. The 'ABC' of GABA receptors. Trends Pharmacol. Sci. 21: 16-19.
Bornstein SR, Tian H, Haidan A, Bottner A, Hiroi N, Eisenhofer G, McCann SM, Chrousos GP, Roffler-Tarlov S,
2000. Deletion of tyrosine hydroxylase gene reveals functional interdependence of adrenocortical and chromaffin
cell system in vivo. Proc. Natl. Acad Sci. U. S. A 97: 14742-14747.
Bowers G, Cullinan WE, Herman JP, 1998. Region-specific regulation of glutamic acid decarboxylase (GAD)
mRNA expression in central stress circuits. J Neurosci. 18: 5938-5947.
Calogero AE, Kamilaris TC, Bernardini R, Johnson EO, Chrousos GP, Gold PW, 1990. Effects of peripheral
benzodiazepine receptor ligands on hypothalamic-pituitary-adrenal axis function in the rat. J Pharmacol. Exp. Ther.
253: 729-737.
Canteras NS, Simerly RB, Swanson LW, 1995. Organization of projections from the medial nucleus of the
amygdala: a PHAL study in the rat. J Comp Neurol. 360: 213-245.
Carrasco GA, Van de Kar LD, 2003. Neuroendocrine pharmacology of stress. Eur. J. Pharmacol. 463: 235-272.
Cascio CS, Shinsako J, Dallman MF, 1987. The suprachiasmatic nuclei stimulate evening ACTH secretion in the
rat. Brain Res. 423: 173-178.
Ceccatelli S, Villar MJ, Goldstein M, Hokfelt T, 1989. Expression of c-Fos immunoreactivity in transmitter-
characterized neurons after stress. Proc. Natl. Acad Sci. U. S. A 86: 9569-9573.
Cohen DR, Ferreira PC, Gentz R, Franza BR, Jr., Curran T, 1989. The product of a fos-related gene, fra-1, binds
cooperatively to the AP-1 site with Jun: transcription factor AP-1 is comprised of multiple protein complexes.
Genes Dev. 3: 173-184.
− 57 −
Coolen LM, Wood RI, 1998. Bidirectional connections of the medial amygdaloid nucleus in the Syrian hamster
brain: simultaneous anterograde and retrograde tract tracing. J Comp Neurol. 399: 189-209.
Cooper JR, Bloom FE, Roth RH, 2003. The Biochemical Basis for of Neuropharmacology. Oxford University
press.
Creigton TE, 1999. Encyclopedia of Molecular Biology. Wiley-Interscience.
Crestani F, Martin JR, Mohler H, Rudolph U, 2000a. Mechanism of action of the hypnotic zolpidem in vivo. Br. J
Pharmacol. 131: 1251-1254.
Cullinan WE, 2000b. GABA(A) receptor subunit expression within hypophysiotropic CRH neurons: a dual
hybridization histochemical study. J. Comp Neurol. 419: 344-351.
Cullinan WE, Helmreich DL, Watson SJ, 1996. Fos expression in forebrain afferents to the hypothalamic
paraventricular nucleus following swim stress. J Comp Neurol. 368: 88-99.
Cullinan WE, Herman JP, Watson SJ, 1993. Ventral subicular interaction with the hypothalamic paraventricular
nucleus: evidence for a relay in the bed nucleus of the stria terminalis. J. Comp Neurol. 332: 1-20.
Curran T, Franza BR, Jr., 1988b. Fos and Jun: the AP-1 connection. Cell 55: 395-397.
Curran T, Morgan JI, 1985. Superinduction of c-fos by nerve growth factor in the presence of peripherally active
benzodiazepines. Science 229: 1265-1268.
D'Costa A, Breese CR, Boyd RL, Booze RM, Sonntag WE, 1991. Attenuation of Fos-like immunoreactivity
induced by a single electroconvulsive shock in brains of aging mice. Brain Res. 567: 204-211.
Davis M, 1998a. Are different parts of the extended amygdala involved in fear versus anxiety? Biol. Psychiatry 44:
1239-1247.
Davis M, Shi C, 1999. The extended amygdala: are the central nucleus of the amygdala and the bed nucleus of the
stria terminalis differentially involved in fear versus anxiety? Ann. N. Y. Acad Sci. 877: 281-291.
De Boer SF, Slangen JL, van der GJ, 1990. Effects of chlordiazepoxide and buspirone on plasma catecholamine
and corticosterone levels in rats under basal and stress conditions. Endocrinol. Exp. 24: 229-239.
De Souza EB, 1990. Neuroendocrine effects of benzodiazepines. J. Psychiatr. Res. 24 Suppl 2: 111-119.
Decavel C, Van den Pol AN, 1990. GABA: a dominant neurotransmitter in the hypothalamus. J. Comp Neurol.
302: 1019-1037.
Duncan GE, Breese GR, Criswell HE, McCown TJ, Herbert JS, Devaud LL, Morrow AL, 1995. Distribution of
[3H]zolpidem binding sites in relation to messenger RNA encoding the alpha 1, beta 2 and gamma 2 subunits of
GABAA receptors in rat brain. Neuroscience 64: 1113-1128.
Everitt BJ, Hokfelt T, 1990. Neuroendocrine anatomy of the hypothalamus. Acta Neurochir. Suppl (Wien. ) 47: 1-
15.
Fenelon VS, Sieghart W, Herbison AE, 1995. Cellular localization and differential distribution of GABAA receptor
subunit proteins and messenger RNAs within hypothalamic magnocellular neurons. Neuroscience 64: 1129-1143.
Finkelstein RA, Boesman M, Neoh SH, LaRue MK, Delaney R, 1974. Dissociation and recombination of the
subunits of the cholera enterotoxin (choleragen). J Immunol. 113: 145-150.
− 58 −
JANUARY 2006
Fleck MW, 2002. Molecular actions of (S)-desmethylzopiclone (SEP-174559), an anxiolytic metabolite of

zopiclone. J Pharmacol. Exp. Ther. 302: 612-618.
Fritschy JM, Mohler H, 1995. GABAA-receptor heterogeneity in the adult rat brain: differential regional and
cellular distribution of seven major subunits. J. Comp Neurol. 359: 154-194.
Gill DM, 1976. The arrangement of subunits in cholera toxin. Biochemistry 15: 1242-1248.
Gillies GE, Linton EA, Lowry PJ, 1982. Corticotropin releasing activity of the new CRF is potentiated several
times by vasopressin. Nature 299: 355-357.
Gilman MZ, 1988. The c-fos serum response element responds to protein kinase C-dependent and -independent
signals but not to cyclic AMP. Genes Dev. 2: 394-402.
Goa KL, Heel RC, 1986. Zopiclone. A review of its pharmacodynamic and pharmacokinetic properties and
therapeutic efficacy as an hypnotic. Drugs 32: 48-65.
Gomez-Sanchez C, Murry BA, Kem DC, Kaplan NM, 1975. A direct radioimmunoassay of corticosterone in rat
serum. Endocrinology 96: 796-798.
Graham D, Faure C, Besnard F, Langer SZ, 1996. Pharmacological profile of benzodiazepine site ligands with
recombinant GABAA receptor subtypes. Eur. Neuropsychopharmacol. 6: 119-125.
Gwosdow-Cohen A, Chen CL, Besch EL, 1982. Radioimmunoassay (RIA) of serum corticosterone in rats. Proc.
Soc. Exp. Biol. Med. 170: 29-34.
Hamilton WJ, Mossman HW, 1982. Human embryology. Heffer. Cambridge, pp. 437-535.
Havard CW, Saldanha VF, Bird R, Gardner R, 1972. The effect of diazepam on pituitary function in man. J.
Endocrinol. 52: 79-85.
Hedblom E, Kirkness EF, 1997. A novel class of GABAA receptor subunit in tissues of the reproductive system. J
Biol. Chem. 272: 15346-15350.
Herman JP, Cullinan WE, Young EA, Akil H, Watson SJ, 1992. Selective forebrain fiber tract lesions implicate
ventral hippocampal structures in tonic regulation of paraventricular nucleus corticotropin-releasing hormone
(CRH) and arginine vasopressin (AVP) mRNA expression. Brain Res. 592: 228-238.
Herman JP, Figueiredo H, Mueller NK, Ulrich-Lai Y, Ostrander MM, Choi DC, Cullinan WE, 2003. Central
mechanisms of stress integration: hierarchical circuitry controlling hypothalamo-pituitary-adrenocortical
responsiveness. Front Neuroendocrinol. 24: 151-180.
Herman JP, Tasker JG, Ziegler DR, Cullinan WE, 2002. Local circuit regulation of paraventricular nucleus stress
integration: glutamate-GABA connections. Pharmacol. Biochem. Behav. 71: 457-468.
Hoffman GE, Lyo D, 2002. Anatomical markers of activity in neuroendocrine systems: are we all 'fos-ed out'? J.
Neuroendocrinol. 14: 259-268.
Holmgren J, Lonnroth I, 1975. Oligomeric structure of cholera toxin: characteristics of the H and L subunits. J
Gen. Microbiol. 86: 49-65.
Holsboer F, 2000. The corticosteroid receptor hypothesis of depression. Neuropsychopharmacology 23: 477-501.
− 59 −
Hsu SM, Raine L, 1981. Protein A, avidin, and biotin in immunohistochemistry. J Histochem. Cytochem. 29: 1349-
1353.
Itier V, Depoortere H, Scatton B, Avenet P, 1996. Zolpidem functionally discriminates subtypes of native GABAA
receptors in acutely dissociated rat striatal and cerebellar neurons. Neuropharmacology 35: 137-145.
Jackson H, Schnieden H, 1968. Pharmacology of reproduction and fertility. Annu. Rev. Pharmacol. 8: 467-490.
Jacobson L, 2005. Hypothalamic-pituitary-adrenocortical axis regulation. Endocrinol. Metab Clin North Am. 34:
271-92, vii.
Jensen JB, Mork A, Mikkelsen JD, 2001. Chronic antidepressant treatments decrease pro-opiomelanocortin mRNA
expression in the pituitary gland: effects of acute stress and 5-HT(1A) receptor activation. J. Neuroendocrinol. 13:
887-893.
Jessop DS, Renshaw D, Larsen PJ, Chowdrey HS, Harbuz MS, 2000. Substance P is involved in terminating the
hypothalamo- pituitary-adrenal axis response to acute stress through centrally located neurokinin-1 receptors.
Stress. 3: 209-220.
Ju G, Han ZS, Fan LZ, 1989. Fluorogold as a retrograde tracer used in combination with immunohistochemistry. J
Neurosci. Methods 29: 69-72.
Justits ministeriet. Bekendtgørelse om forsøgsdyrs pasning og opstaldning og om anvendelse af udryddelsestruede

og vildtlevende dyr til forsøg mv. BEK nr 687 af 25/07/2003. 2003.
Ref Type: Internet Communication
Kalman BA, Kim PJ, Cole MA, Chi MS, Spencer RL, 1997. Diazepam attenuation of restraint stress-induced
corticosterone levels is enhanced by prior exposure to repeated restraint. Psychoneuroendocrinology 22: 349-360.
Kandel ER, Schwartz JH, Jessell TM, 2000. Principles of Neural Science. McGraw-Hill Companies Inc..
Keller-Wood ME, Dallman MF, 1984. Corticosteroid inhibition of ACTH secretion. Endocr. Rev. 5: 1-24.
Kiss JZ, 1988. Dynamism of chemoarchitecture in the hypothalamic paraventricular nucleus. Brain Res. Bull. 20:
699-708.
Kiss JZ, Palkovits M, Zaborszky L, Tribollet E, Szabo D, Makara GB, 1983a. Quantitative histological studies on
the hypothalamic paraventricular nucleus in rats: I. Number of cells and synaptic boutons. Brain Res. 262: 217-224.
Kononen J, Honkaniemi J, Alho H, Koistinaho J, Iadarola M, Pelto-Huikko M, 1992. Fos-like immunoreactivity in

the rat hypothalamic-pituitary axis after immobilization stress. Endocrinology 130: 3041-3047.
Kovacs KJ, Miklos IH, Bali B, 2004. GABAergic mechanisms constraining the activity of the hypothalamo-
pituitary-adrenocortical axis. Ann. N. Y. Acad Sci. 1018: 466-476.
Krieger DT, Hughes JC, 1980. Neuroendocrinology. Sinauer Associates Inc. Sunderland Massachussetts.
Lakic N, Pericic D, Manev H, 1986. Mechanisms by which picrotoxin and a high dose of diazepam elevate plasma
corticosterone level. Neuroendocrinology 43: 331-335.
Larsen PJ, Hay-Schmidt A, Mikkelsen JD, 1994b. Efferent connections from the lateral hypothalamic region and
the lateral preoptic area to the hypothalamic paraventricular nucleus of the rat. J Comp Neurol. 342: 299-319.
− 60 −
JANUARY 2006
Le FG, Guilloux F, Mitrani N, Mizoule J, Uzan A, 1979. Relationships between plasma corticosteroids and
benzodiazepines in stress. J Pharmacol. Exp. Ther. 211: 305-308.
Lightman SL, Young WS, III, 1988. Corticotrophin-releasing factor, vasopressin and pro-opiomelanocortin mRNA
responses to stress and opiates in the rat. J Physiol 403: 511-523.
Matheson GK, Gage D, White G, Dixon V, Gipson D, 1988. A comparison of the effects of buspirone and
diazepam on plasma corticosterone levels in rat. Neuropharmacology 27: 823-830.
Matsuoka I, Giuili G, Poyard M, Stengel D, Parma J, Guellaen G, Hanoune J, 1992. Localization of adenylyl and
guanylyl cyclase in rat brain by in situ hybridization: comparison with calmodulin mRNA distribution. J Neurosci.
12: 3350-3360.
McElroy JF, Miller JM, Meyer JS, 1987. Comparison of the effects of chlordiazepoxide and CL 218,872 on serum
corticosterone concentrations in rats. Psychopharmacology (Berl) 91: 467-472.
McKernan RM, Rosahl TW, Reynolds DS, Sur C, Wafford KA, Atack JR, Farrar S, Myers J, Cook G, Ferris P,
Garrett L, Bristow L, Marshall G, Macaulay A, Brown N, Howell O, Moore KW, Carling RW, Street LJ, Castro JL,
Ragan CI, Dawson GR, Whiting PJ, 2000. Sedative but not anxiolytic properties of benzodiazepines are mediated
by the GABA(A) receptor alpha1 subtype. Nat. Neurosci. 3: 587-592.
Meister B, Hokfelt T, Geffard M, Oertel W, 1988. Glutamic acid decarboxylase- and gamma-aminobutyric acid-like
immunoreactivities in corticotropin-releasing factor-containing parvocellular neurons of the hypothalamic
paraventricular nucleus. Neuroendocrinology 48: 516-526.
Messer WS. GABA synthesis. Neuroscience Research at The University of Toledo . 2005. 19-11-2005.
Mikkelsen JD, Larsen PJ, 1993. Substance P in the suprachiasmatic nucleus of the rat: an immunohistochemical
and in situ hybridization study. Histochemistry 100: 3-16.
Mikkelsen JD, Soderman A, Kiss A, Mirza N, 2005. Effects of benzodiazepines receptor agonists on the
hypothalamic-pituitary-adrenocortical axis. Eur. J Pharmacol. 519: 223-230.
Mikkelsen JD, Vrang N, Mrosovsky N, 1998. Expression of Fos in the circadian system following nonphotic
stimulation. Brain Res. Bull. 47: 367-376.
Miklos IH, Kovacs KJ, 2002. GABAergic innervation of corticotropin-releasing hormone (CRH)-secreting
parvocellular neurons and its plasticity as demonstrated by quantitative immunoelectron microscopy. Neuroscience
113: 581-592.
Moga MM, Saper CB, Gray TS, 1989. Bed nucleus of the stria terminalis: cytoarchitecture, immunohistochemistry,
and projection to the parabrachial nucleus in the rat. J. Comp Neurol. 283: 315-332.
Morgan WW, Pfeil KA, 1979. Evidence for a cholinergic influence on catecholaminergic pathways terminating in
the anterior and medial basal hypothalamus. Brain Res. 173: 47-56.
Ninan PT, 1999. The functional anatomy, neurochemistry, and pharmacology of anxiety. J Clin Psychiatry 60 Suppl
22: 12-17.
Nogueira MI, de Rezende BD, do Vale LE, Bittencourt JC, 2000. Afferent connections of the caudal raphe pallidus
nucleus in rats: a study using the fluorescent retrograde tracers fluorogold and true-blue. Ann. Anat. 182: 35-45.
− 61 −
Nutt DJ, Malizia AL, 2001. New insights into the role of the GABA(A)-benzodiazepine receptor in psychiatric
disorder. Br. J Psychiatry 179: 390-396.
Okamura H, Abitbol M, Julien JF, Dumas S, Berod A, Geffard M, Kitahama K, Bobillier P, Mallet J, Wiklund L,
1990. Neurons containing messenger RNA encoding glutamate decarboxylase in rat hypothalamus demonstrated
by in situ hybridization, with special emphasis on cell groups in medial preoptic area, anterior hypothalamic area
and dorsomedial hypothalamic nucleus. Neuroscience 39: 675-699.
Orchinik M. Mechanisms Mediating Rapid Stress Hormone Action in vertebrate Brains. 2005.
Ref Type: Hearing
Owens MJ, Bissette G, Nemeroff CB, 1989. Acute effects of alprazolam and adinazolam on the concentrations of
corticotropin-releasing factor in the rat brain. Synapse 4: 196-202.
Palkovits M, 1982. Neuropeptides in the median eminence: their sources and destinations. Peptides 3: 299-303.
Paxinos G, Watson C, 1997. The Rat Brain In steriotaxic Coordinates. Academic Press Limited.
Pirker S, Schwarzer C, Wieselthaler A, Sieghart W, Sperk G, 2000. GABA(A) receptors: immunocytochemical

distribution of 13 subunits in the adult rat brain. Neuroscience 101: 815-850.
Pivac N, Pericic D, 1993. Inhibitory effect of diazepam on the activity of the hypothalamic-pituitary-adrenal axis in
female rats. J Neural Transm. Gen. Sect. 92: 173-186.
Pritchett DB, Seeburg PH, 1990. Gamma-aminobutyric acidA receptor alpha 5-subunit creates novel type II
benzodiazepine receptor pharmacology. J Neurochem. 54: 1802-1804.
Psycheduation.org. hypothalamus. 2005.

Rang HP, Dale MM, Ritter JM, Moore PK, 2004. Pharmacology. Churchill Livingstone.
Reul JM, Gesing A, Droste S, Stec IS, Weber A, Bachmann C, Bilang-Bleuel A, Holsboer F, Linthorst AC, 2000.
The brain mineralocorticoid receptor: greedy for ligand, mysterious in function. Eur. J Pharmacol. 405: 235-249.
Rho JH, Swanson LW, 1989. A morphometric analysis of functionally defined subpopulations of neurons in the
paraventricular nucleus of the rat with observations on the effects of colchicine. J Neurosci. 9: 1375-1388.
Risold PY, Thompson RH, Swanson LW, 1997. The structural organization of connections between hypothalamus
and cerebral cortex. Brain Res. Brain Res. Rev. 24: 197-254.
Rivier C, Vale W, 1983. Interaction of corticotropin-releasing factor and arginine vasopressin on

adrenocorticotropin secretion in vivo. Endocrinology 113: 939-942.
Rosenzwig, Leiman, Breedlove, 1996. Biological psychology. Sinanser associates inc., p. Fig 7.11B.
Sagar SM, Sharp FR, Curran T, 1988. Expression of c-fos protein in brain: metabolic mapping at the cellular level.
Science 240: 1328-1331.
Salva P, Costa J, 1995. Clinical pharmacokinetics and pharmacodynamics of zolpidem. Therapeutic implications.
Clin. Pharmacokinet. 29: 142-153.
Sanger DJ, 2004. The pharmacology and mechanisms of action of new generation, non-benzodiazepine hypnotic
agents. CNS. Drugs 18 Suppl 1: 9-15.
− 62 −
JANUARY 2006
Sassone-Corsi P, Visvader J, Ferland L, Mellon PL, Verma IM, 1988. Induction of proto-oncogene fos
transcription through the adenylate cyclase pathway: characterization of a cAMP-responsive element. Genes Dev.
2: 1529-1538.
Sawchenko PE, Brown ER, Chan RK, Ericsson A, Li HY, Roland BL, Kovacs KJ, 1996. The paraventricular
nucleus of the hypothalamus and the functional neuroanatomy of visceromotor responses to stress. Prog. Brain
Res. 107: 201-222.
Scott-Stevens P, Atack JR, Sohal B, Worboys P, 2005. Rodent pharmacokinetics and receptor occupancy of the
GABAA receptor subtype selective benzodiazepine site ligand L-838417. Biopharm. Drug Dispos. 26: 13-20.
selye H, 1936. A syndrome produced by diverse nocuouse. Nature 138.
Sheng M, McFadden G, Greenberg ME, 1990. Membrane depolarization and calcium induce c-fos transcription via
phosphorylation of transcription factor CREB. Neuron 4: 571-582.
Sieghart W, 1995. Structure and pharmacology of gamma-aminobutyric acidA receptor subtypes. Pharmacol. Rev.
47: 181-234.
Sieghart W, Fuchs K, Tretter V, Ebert V, Jechlinger M, Hoger H, Adamiker D, 1999. Structure and subunit
composition of GABA(A) receptors. Neurochem. Int. 34: 379-385.
Sigel E, Baur R, Trube G, Mohler H, Malherbe P, 1990b. The effect of subunit composition of rat brain GABAA
receptors on channel function. Neuron 5: 703-711.
Sigel E, Baur R, Trube G, Mohler H, Malherbe P, 1990a. The effect of subunit composition of rat brain GABAA
receptors on channel function. Neuron 5: 703-711.
Site de neuroanatomie, http:webperso.easyconnect.fr/baillement/anatomie/hypothalamus.html. PVN. 2005.

Smith MA, Banerjee S, Gold PW, Glowa J, 1992. Induction of c-fos mRNA in rat brain by conditioned and
unconditioned stressors. Brain Res. 578: 135-141.
Stellar E, 1954. The physiology of motivation. Psychol. Rev. 61: 5-22.
Swanson LW, Kuypers HG, 1980b. The paraventricular nucleus of the hypothalamus: cytoarchitectonic
subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated
by retrograde fluorescence double-labeling methods. J Comp Neurol. 194: 555-570.
Swanson LW, Sawchenko PE, 1980. Paraventricular nucleus: a site for the integration of neuroendocrine and
autonomic mechanisms. Neuroendocrinology 31: 410-417.
Taylor PM, Thomas P, Gorrie GH, Connolly CN, Smart TG, Moss SJ, 1999. Identification of amino acid residues
within GABA(A) receptor beta subunits that mediate both homomeric and heteromeric receptor expression. J
Neurosci. 19: 6360-6371.
Thompson MA, Ginty DD, Bonni A, Greenberg ME, 1995. L-type voltage-sensitive Ca2+ channel activation
regulates c-fos transcription at multiple levels. J Biol. Chem. 270: 4224-4235.
Trenque T, Bustany P, Lamiable D, Legros S, Choisy H, 1994. Pharmacokinetics and brain distribution of
zolpidem in the rat after acute and chronic administration. J Pharm. Pharmacol. 46: 611-613.
− 63 −
Tretter V, Ehya N, Fuchs K, Sieghart W, 1997. Stoichiometry and assembly of a recombinant GABAA receptor
subtype. J Neurosci. 17: 2728-2737.
Van de Kar LD, Blair ML, 1999a. Forebrain pathways mediating stress-induced hormone secretion. Front
Neuroendocrinol. 20: 1-48.
Vargas ML, Abella C, Hernandez J, 2001b. Diazepam increases the hypothalamic-pituitary-adrenocortical (HPA)
axis activity by a cyclic AMP-dependent mechanism. Br. J. Pharmacol. 133: 1355-1361.
Vrang N, Larsen PJ, Mikkelsen JD, 1995. Direct projection from the suprachiasmatic nucleus to hypophysiotrophic
corticotropin-releasing factor immunoreactive cells in the paraventricular nucleus of the hypothalamus
demonstrated by means of Phaseolus vulgaris-leucoagglutinin tract tracing. Brain Res. 684: 61-69.
Walker MC, Tong X, Brown S, Shorvon SD, Patsalos PN, 1998. Comparison of single- and repeated-dose
pharmacokinetics of diazepam. Epilepsia 39: 283-289.
Weller KL, Smith DA, 1982b. Afferent connections to the bed nucleus of the stria terminalis. Brain Res. 232: 255-
270.
Whitnall MH, 1993. Regulation of the hypothalamic corticotropin-releasing hormone neurosecretory system. Prog.
Neurobiol. 40: 573-629.
Wilson MA, Biscardi R, Smith MD, Wilson SP, 1996. Effects of benzodiazepine agonist exposure on corticotropin-
releasing factor content and hormonal stress responses: divergent responses in male and ovariectomized female
rats. J Pharmacol. Exp. Ther. 278: 1073-1082.
Wisden W, Laurie DJ, Monyer H, Seeburg PH, 1992. The distribution of 13 GABAA receptor subunit mRNAs in
the rat brain. I. Telencephalon, diencephalon, mesencephalon. J Neurosci. 12: 1040-1062.
Wittekind D, Hilgarth M, Kretschmer V, 1979. [Simplification of the Papanicolaou stain which is easily
reproducable (author's transl)]. Geburtshilfe Frauenheilkd. 39: 969-972.
Wohlfarth KM, Bianchi MT, Macdonald RL, 2002. Enhanced neurosteroid potentiation of ternary GABA(A)
receptors containing the delta subunit. J Neurosci. 22: 1541-1549.
Worley PF, Christy BA, Nakabeppu Y, Bhat RV, Cole AJ, Baraban JM, 1991. Constitutive expression of zif268 in
neocortex is regulated by synaptic activity. Proc. Natl. Acad Sci. U. S. A 88: 5106-5110.
Wotjak CT, Naruo T, Muraoka S, Simchen R, Landgraf R, Engelmann M, 2001. Forced swimming stimulates the
expression of vasopressin and oxytocin in magnocellular neurons of the rat hypothalamic paraventricular nucleus.
Eur. J Neurosci. 13: 2273-2281.
− 64 −
JANUARY 2006
APPENDICES NeuroSearch
8. APPENDICES
− 65 −
JANUARY 2006
APPENDIKS NeuroSearch
8.1. APENDIX A (PREPARATION OF MEDIUMS, BUFFER AND SOLUTIONS:

IMMUNOCYTOCHEMISTRY.)
Cryoprotection solution
1L
Glycerine 250 ml
Ethylenglycol 250 ml
PBS 0.1 M 500 ml
Keep in refrigerator
Phosphate Buffer Solution (PBS) 0.1 M
• Acid solution:
NaH2PO4-H2O 2.73 g
NaCl 1.82 g
Distilled water 200 ml
• Alkaline solution:
Na2HPO4-H2O 14.25 g
NaCl 7.26 g
Distilled water 800 ml
Titrate the acid solution with with HCL or NaOH to reach pH 7.4
30 % Sucrose
Sucrose 30 g
PBS 0.1 M 100 ml
Tris-buffer (pH 7,4)
Tris 12.1 g
NaCl 17.5 g
Milli Q-water 1910 ml
1 N HCl 90 ml
-I-
Gelatine
Milli Q-vand 400 ml

Gelatine 2g
Blocking buffer 1 L
Normal Swine serum 5 ml

Bovine Serum Albumin (BSA) 1g
Triton 100 x (TX) 0.3 ml
PBS 994.7 ml
- II -
JANUARY 2006
8.2. APPENDIX B (PREPARATION OF MEDIUMS, BUFFER AND SOLUTIONS: IN

SITU HYBRIDIZATION )
Preparation of mediums, buffer and solutions: In situ hybridization
Hybridization solution
5× hybridization stock 1.125 ml/well

50 % dextran sulphate 1.125 ml/well
Deionized formamide 2.8125 ml/well
ssDNA 0.169 ml/well
1 M DTT (Dithiothreito) 0,133 ml/well
tRNA 0,056 ml/well
Probe 4.5 million counts/well
*This volume allows for 5 ml hybridization solution per well
DEPC Water
Mili-Q water 1L
DEPC (Diethylpyrocarbonate) 100 µl
Incubate for 1 hour shaking the bottle every 30 minutes

Store at room temperature.
1.0 M DTT
DTT (Dithiothreitol) 3.09 g

Sodium acetate 20 ml 0,01 M pH 5,2
Sterilize by filtration
Store at -20 °C
tRNA:
tRNA (Boehringer Mannheim) 15 mg

DEPC treated water 1 ml
5 × hybridization stock: 10 ml
1 M Tris-Hcl pH 7,5 1 ml
0,5 M EDTA pH 8,0 in DEPC Water 100 µl
100 × Denhardt’s (sigma aldric) 100 µl
5 M NaCl in DEPC water 3.35 ml
DEPC water 4.55 ml
Keep in freezer
ssDNA ( short strained DNA)
Herring sperm (sigma aldric)
20 × SCC (4L)
- III -
NaCl 701.2 g
Sodium Citrate 352.8 g
Mili-q water 3200 ml
Adjust pH to 7,0 with NaOH

Autoclave
Store at room temperature
RNase A 100 µg/ml
NaCl 14,5 g
Tris pH 8,0 0,69 g
Rnase 50 µg
Autoclaved mili-Q 500 ml
Store in refrigerator
K-PBS (50 mM)
Mili-Q water 900 ml

0,5 M KH2PO4 (monobasic) 20 ml
0,5 M KH2PO4 (dibasic) 80 ml
NaCl 9g
DEPC (Diethylpyrocarbonate) 100 µl
1,0 M TEA (Triethanolamine)
Triethanolamine 33.18 ml
DEPC-treated water 150 ml
Adjust pH to 8,0 with concentrated HCL, then bring up to 250 ml with DEPC water, and then autoclave before
use.
- IV -
JANUARY 2006
8.3. APENDIX C SEQUENCE OF THE OLIGONUCLEOTID PROBES USED FOR IN

SITU HYBRIDIZATION
3 6 9 12 15 18 21 24 27 30 33 36 39 43 45
GABAA α1: 5’ GTT GTT TTT ATT GAG AGG ATC CTT CAC TTT CTT TGG CTT TTC TGG
3 6 9 12 15 18 21 24 27 30 33 36 39 43
GABAA α2: 5’ CAA CTT GGC TGG CCA CAC CAG GAG AAC AAG CAG CGG AAA CCA
3 6 9 12 15 18 21 24 27 30 33 36 39 43
GABAA α3: 5’ CAC TGT TGG AGT TGA AGA ACT GGG AGC AGC AGC CTT GGA CAT
-V-
8.4. APENDIX D (ABSTRACT PRESENTET AT THE SOCIETY FOR NEUROSCIENCE

ANNUAL MEETING NOVEMBER 2005)
Effects of Benzodiazepines Receptor Agonists on the Hypothalamic-Pituitary

Adrenocortical Axis
ANDREAS SØDERMAN*, ALEXANDER KISS, NAHEED MIRZA, AND JENS D. MIKKELSEN
Department of Functional Neuroanatomy and Biomarkers, and 3 Pharmacology, NeuroSearch A/S, Pederstrupvej
93, 2750 Ballerup, Denmark
Institute of Experimental Endocrinology, Slovak Academy of Sciences, Bratislava, Slovakia
Previous studies have demonstrated that classical benzodiazepines decreases hypothalamic-pituitary-

adrenocortical cortex (HPA) axis activity. Paradoxically, high doses of benzodiazepines also stimulate
basal circulating corticosterone levels in some conditions. Because benzodiazepine agonists display little
selectivity to any of the α subtypes of the GABAA receptor to which they bind, we propose that the
unequivocal results are due to an α subtype-dependent modulation of the HPA axis output. To test this,
basal hormonal output and induction of FOS in the hypothalamic paraventricular nucleus (PVN) were
measured after administration of various benzodiazepine ligands in mice. Zolpidem, a selective α1
subtype agonist, produced a very strong increase in plasma ACTH and corticosterone whereas the inverse
agonist FG7142 induced a small rise in plasma corticosterone. More surprisingly, the non-selective full
agonists diazepam and zopiclone induced a lower increase in circulating corticosterone than after
zolpidem. In contrast, the α2,3,5-selective benzodiazepine agonist and α1 antagonist L-838,417 had no
effect on corticosterone levels. Strong induction of FOS in the PVN was found in response to zolpidem,
diazepam, and zopiclone, but not after L-838,417. Finally, pre-administration of L-838,417 prior to
zolpidem strongly inhibited the effect of zolpidem on corticosterone. Likewise, the non-selective agonists
diazepam and zopiclone at a dose that alone had no effect on corticosterone also inhibited the effect of
zolpidem. Taken together, these results suggest that benzodiazepine ligands modulate the HPA axis
through partly opposite mechanisms; and that the net-effect is dependent on the composition of the
GABAA receptor subunits to which they bind.
- VI -
JANUARY 2006
8.5. APENDIX E EUROPEAN JOURNAL OF PHARMACOLOGY 519 (2005) 223 – 230
- VII -
European Journal of Pharmacology 519 (2005) 223 – 230
www.elsevier.com/locate/ejphar
Effects of benzodiazepines receptor agonists on the

hypothalamic–pituitary–adrenocortical axis
Jens D. Mikkelsen a,*, Andreas Søderman a, Alexander Kiss b, Naheed Mirza c
a
Department of Functional Neuroanatomy and Biomarkers, NeuroSearch A/S, Pederstrupvej 93, 2750 Ballerup, Denmark
b
Institute of Experimental Endocrinology, Slovak Academy of Sciences, Bratislava, Slovakia
c
Department of Pharmacology, NeuroSearch A/S, Pederstrupvej 93, 2750 Ballerup, Denmark
Received 24 March 2005; received in revised form 23 June 2005; accepted 30 June 2005
Available online 25 August 2005
Abstract
Previous studies have demonstrated that classical benzodiazepines decrease hypothalamic – pituitary – adrenocortical cortex (HPA) axis
activity. Paradoxically, high doses of benzodiazepines also stimulate basal circulating corticosterone levels in some conditions. Because
benzodiazepine agonists display little selectivity to any of the a subtypes of the g-amino butyric acid (GABA)A receptor to which they bind,
we propose that the unequivocal results are due to an a subtype-dependent modulation of the hypothalamic – pituitary – adrenocortical cortex
axis output. To test this, basal hormonal output and induction of Fos in the hypothalamic paraventricular nucleus were measured after
administration of various benzodiazepine ligands in mice. Zolpidem, a selective a1 subtype agonist, produced a very strong increase in
plasma adrenocorticotropic hormone and corticosterone whereas the inverse agonist FG7142 induced a small rise in plasma corticosterone.
More surprisingly, the non-selective full agonists diazepam and zopiclone induced a lower increase in circulating corticosterone than after
zolpidem. In contrast, the a2,3,5-selective benzodiazepine agonist and a1 antagonist L-838,417 had no effect on corticosterone levels. Strong
induction of Fos in the paraventricular nucleus was found in response to zolpidem, diazepam, and zopiclone, but not after L-838,417. Finally,
pre-administration of L-838,417 prior to zolpidem strongly inhibited the effect of zolpidem on corticosterone. Likewise, the non-selective
agonists diazepam and zopiclone at a dose that alone had no effect on corticosterone also inhibited the effect of zolpidem. Taken together,
these results suggest that benzodiazepine ligands modulate the hypothalamic – pituitary – adrenocortical cortex axis through partly opposite
mechanisms; and that the net effect is dependent on the composition of the GABAA receptor subunits to which they bind.
D 2005 Published by Elsevier B.V.
Keywords: ACTH (adrenocorticotropic hormone); Corticosterone; GABA; Paraventricular nucleus; Stress; L-838,417; Diazepam; Zolpidem; Zopiclone;
(Mouse)
1. Introduction activation of the GABAA receptor is augmented by the

anxiolytic benzodiazepine ligands, causing a parallel shift of
GABA is the major inhibitory neurotransmitter in the the GABA concentration –response curve (Barnard et al.,
mammalian brain and the GABAA receptor is the site of 1998; Sieghart, 1995). The a subunit isoform present within
action of benzodiazepines. Multiple isoforms of GABAA an individual GABAA receptor subtype is the primary
receptor exist; each receptor comprises of a pentameric determinant of benzodiazepine ligand recognition (Fleck,
complex formed by the co-assembly of subunits selected 2002; Mehta and Ticku, 1999; Pritchett and Seeburg, 1990).
from 16 genes (a1 – 6, h1 – 3, g1 – 3, y, q, k, and u) surrounding a GABAA containing a1, a2, a3, or a5 together with h, and a g2
central chloride ion-channel (Sieghart, 1995). Agonist subunit are all recognised by the classical benzodiazepines.
Other drugs distinguish the GABAA receptors on the basis of
their a subunit composition. For example, the sedative –
* Corresponding author. Tel.: +45 44608359; fax: +45 44608080. hypnotic imidazolpyridine zolpidem has higher affinity for
E-mail address: JDM@Neurosearch.dk (J.D. Mikkelsen). a1 than a2, a3, or a5 containing GABAA receptors (Squires et
0014-2999/$ - see front matter D 2005 Published by Elsevier B.V.
doi:10.1016/j.ejphar.2005.06.049
224 J.D. Mikkelsen et al. / European Journal of Pharmacology 519 (2005) 223 – 230
al., 1979). More recently, compounds with subtype selection increase corticosterone levels in the mouse blood (Kalman
for a2, a3, and a5 over a1 containing GABAA receptors such et al., 1997; Lakic et al., 1986). We propose that the
as L-838,417 are considered to be involved in anxiolytic conflicting results reported earlier may also be a conse-
properties (McKernan et al., 2000). It is therefore plausible quence of the net effect of non-selective benzodiazepines on
that distinct a subtype-specific GABAA receptors mediate the various GABAA receptor a subtypes. In order to address
various effects of classical benzodiazepines. this issue, the responsiveness of the murine HPA axis output
Disturbances in the hypothalamic – pituitary– adrenocor- was examined after acute and combined treatments with a
tical (HPA) axis are well known in human depression and number of benzodiazepine receptor modulators specific for
anxiety, as well as in experimental animal models of anxiety the a1 and/or a2,3,5 containing GABAA receptors. In
and depression (Arborelius et al., 1999; Holsboer, 2000; addition, we used induction of Fos as a marker to identify
Jensen et al., 2001; Van de Kar and Blair, 1999). The the level of activation in the paraventricular nucleus after
abnormal hypersecretion of cortisol that occurs in a variety acute administration of benzodiazepine ligands.
of psychiatric disorders including major depression, sleep
disturbances, and anxiety is suppressed by administration of
benzodiazepines (De Boer et al., 1990). Neurons in the 2. Materials and methods
medial parvocellular part of the hypothalamic paraventri-
2.1. Compounds
cular nucleus contain corticotrophin-releasing hormone
(CRH), and these neurons integrate excitatory and inhibitory Zolpidem and zopiclone were purchased from Tocris Ltd.
signals and behavioural responses to stress into appropriate (Bristol, UK). FG7142 (methyl-3-carboline-3-carboxamide), and
secretion of CRH that leads to adrenocorticotropic hormone diazepam was purchased from Sigma-Aldrich (St. Louis, USA).
(ACTH) release to the general circulation (Carrasco and Van L-838,417 (7-tert-butyl-3-(2,5-difluoro-phenyl)-6-(2-methyl-2H-
de Kar, 2003). GABA is essential as a negative regulator of [1,2,4]triazol-3-ylmethoxy)-[1,2,4]triazolo[4,3-b]pyridazine) was
neuronal excitability in the paraventricular nucleus, thus synthesised by the Department of Medicinal Chemistry, Neuro-
mediating the amplitude and the duration of the stress Search A/S. All compounds were dissolved in 5% chremophor
response (Kovacs et al., 2004). GABA is also contained in and administered intraperitoneally at 10 ml/kg.
neurons comprised in the negative feedback loop to the
2.2. Animals and single administration of benzodiazepine ligands
paraventricular nucleus (Bowers et al., 1998), and it is
plausible that GABAA receptors may act on converging Adult male NMRI mice weighing 30 – 35 g were purchased
projections to the paraventricular nucleus. GABAergic from Harlan Inc (Hamburg, Germany). The animals were received
afferents in the paraventricular nucleus originating from a at the animal facility, and housed 5 per cage under 12:12-h light/
number of forebrain areas have been reported and ultra- dark cycle, humidity in a temperature-controlled room for at least 7
structural studies have shown that GABAergic axons, to a days before the experiment. Food and water were available ad
large extent, terminate on CRH neurons (Miklos and libitum. All procedures were conducted in accordance with the
Kovacs, 2002; Roland and Sawchenko, 1993), and in situ Danish National Guide for Care and Use of Laboratory animals. In
hybridisation and in vitro autoradiography have detected a1, order to analyse the dose-dependent effect of all benzodiazepine
a2, and a3 subtypes in the paraventricular nucleus (Cull- ligands on circulating corticosterone, mice (n = 5) were adminis-
inan, 2000; Duncan et al., 1995; Fenelon et al., 1995). tered at doses 0, 025, 0.1, 0.5, 2.5, 12.5, and 25 mg/kg i.p. The
mice were returned to their home cages and sacrificed by
The role of benzodiazepines on basal corticosteroid
decapitation 60 min after drug administration. Trunk blood was
release is complex, and even though it has been extensively collected in centrifuge tubes containing 2 mg EDTA.
studied over the last three decades, there is still a lot of In order to examine the dynamics of the HPA axis, another
controversy in the literature. Benzodiazepines have, under experiment was carried out to determine the temporal profile of
basal conditions, been shown to stimulate (Calogero et al., HPA activation by zolpidem. These mice were treated with 12.5
1990; Le Fur et al., 1979; Vargas et al., 2001), to inhibit mg/kg zolpidem and the animals were sacrificed and trunk blood
(Owens et al., 1989; Pivac and Pericic, 1993), and to be was taken at various time points after drug administration (t = 5, 10,
ineffective (Matheson et al., 1988). The state of the animal 15, 30, 60, and 120 min). The blood was collected in EDTA
under which these drugs are tested should be considered containing tubes, kept at 4 -C, and plasma was isolated by
with special care, because benzodiazepines interact with centrifugation and stored at 20 -C until processed.
stress-related behaviours and cause a dose-dependent
inhibition of stress-induced rise in corticosteroid levels 2.3. Combination of drugs
(De Souza, 1990, for review). In addition, the effect of acute
In order to investigate the relationship between activation of a1
benzodiazepines on basal corticosterone levels might be and non-a1 containing GABAA receptors, combinations of L-
related to several factors such as dose, compound and 838,417 (12.5 mg/kg i.p.) or vehicle were given 15 min before a
gender of the animal (Wilson et al., 1996), and these single dose of zolpidem i.p. at increasing concentrations (0.5, 2.5,
features might underlie the unequivocal results. and 12.5 mg/kg). L-838,417 acts at the benzodiazepine binding site
In agreement with the literature, we show here that doses and is a functionally selective agonist at a2, a3, and a5 subtypes. In
of diazepam higher than those producing anxiolytic effects addition, L-838,417 is an antagonist on the a1 site; it might be that
J.D. Mikkelsen et al. / European Journal of Pharmacology 519 (2005) 223 – 230 225
2500
-C for 24 h in an antiserum against Fos (code #94012-5) diluted
1:4000 in PBS with 0.3% Triton X-100 and 1% bovine serum
2000
albumin. The primary antiserum was raised in rabbit against the N-
terminal peptide similar to 2 – 17 of the rat Fos protein and
1500
characterised previously (Mikkelsen et al., 1998; Woldbye et al.,
% increase
1996). The immunoreactive cells were identified by means of the

1000 avidin – biotin protocol and diaminobenzidine was applied as
chromogene as earlier described (Mikkelsen et al., 1998).
500
2.6. Quantifications and statistics

0
Time (min) To quantify the effects of the benzodiazepine ligands tested, an
observer blind to the treatment of the animals counted the number
0 20 40 60 80 100 120 140
of positive cells in each of the two paraventricular nuclei at its
Fig. 1. Time course profile of ACTH and corticosterone responses to a single middle portion in the same tissue section. The number of positive
dose of zolpidem (12.5 mg/kg) in mice (n = 5). The figure illustrates the cells is given as the average of these two counts. The middle
relative rise compared to naı̈ve animals of plasma ACTH (open circles) and portion of the paraventricular nucleus was delineated according to
corticosterone (closed circles) levels at various time points after zolpidem. Paxinos and Franklin (2004) to about 2.98 mm rostral to the
The time course illustrates that release of ACTH occurs immediately after interaural line. The number of immunoreactive nuclei was counted
administration of zolpidem and that precedes release of corticosterone.
within the total paraventricular nucleus at this level including both
any inhibition results from its antagonistic properties. Therefore, the the magnocellular and the parvocellular subportions of the nucleus.
non-selective GABA potentiating benzodiazepine ligands, diaze- The data were analysed by one- or two-way analysis of variance
pam and zopiclone, were tested together with zolpidem. In this case (ANOVA) followed by the Student – Newman – Keuls method. All
mice were treated with vehicle, diazepam (0.5 mg/kg i.p.), or data are represented as group means T standard error of mean
zopiclone (0.5 mg/kg i.p.) 15 min before administering the mice with (S.E.M.), and levels at p < 0.05 were considered significant.
0.5 mg/kg zolpidem. The animals were sacrificed 60 min after
zolpidem and trunk blood was taken for corticosterone measure-
ments. All single and combined dosing experiments were carried out 3. Results
twice. ED50 values were calculated by non-linear regression using
GraphPad Prism software (version 2.01, Gradpad Software Inc., San 3.1. ACTH and corticosterone responses after acute administration
Diego, CA). of benzodiazepine ligands
2.4. Radioimmunoassay Zolpidem significantly and dose-dependently stimulated the

mouse peripheral components of the HPA axis (Fig. 1). Plasma
Plasma aliquots were stored at 20 -C until hormone levels ACTH was found to be about 15-fold above baseline levels and
were determined. Plasma corticosterone and ACTH were measured
directly without prior extraction. In the case of corticosterone, it 1200
was assayed using a commercial [125I] radioimmunoassay kit from
DPC Diagnostics. Plasma ACTH concentrations were determined 1000
by a commercial radioimmunoassay kit from Amersham.
800
2.5. Immunohistochemistry for Fos

600 a,b
Sixty male mice were used for Fos studies. The animals were a,b
400 a,b
administered i.p. with 12.5 mg/kg diazepam (n = 9), zopiclone a,b
a
(n = 9), zolpidem (n = 12), L-838,417 (n = 9), or vehicle (5% 200 a,b
chremophor) (n = 11) at 4 h after the onset of the light phase and
returned to their home cage. Sixty minutes after the injection the a
0 a
a
animals were deeply anaesthetised with mebumal and perfused
transcardially with 0.9% saline followed by 4% paraformaldehyde 0 0,025 0,1 0,5 2,5 12,5 25 mg/kg i.p.
in 0.1 M phosphate buffer (pH 7.4) for 5 min. The brains were
post-fixed overnight and subsequently submerged in 20% sucrose Fig. 2. The effect of increasing doses of the non-selective benzodiazepine
in 0.01 M phosphate-buffered saline (PBS) at 4 -C overnight. ligands diazepam (closed circles) and zopiclone (closed triangles) on
Forty-micrometer serial coronal sections were cut through the plasma corticosterone levels in mice compared to the a1 selective zolpidem
(open circles) and the a2,3 selective agonist L-838,417 (open triangles). The
hypothalamus on a freezing sledge microtome in series of four and
four drugs are given as a single dose and the mice are sacrificed 60 min
stored in a solution containing ethylene glycol glycerol in 30%
after the treatment. The data represent mean T S.E.M. of minimum of 5 mice
sucrose at 20 -C until further processing. Prior to the per group. There was a significant ( P < 0.05) difference between the effect
immunocytochemical steps, the sections were rinsed for 3 10 of non-selective and selective benzodiazepine receptor ligands. #P < 0.05
min in 0.01 M PBS, in 1% H2O2 – PBS for 10 min, and in PBS vs. zolpidem and *P < 0.05 vs. vehicle. Two-way ANOVA and Student –
with 0.3% Triton X-100, 5% swine-serum, and 1% bovine serum Newman – Keuls test. The letters a and b denote a significant difference at a
albumin (BSA) for 30 min. Then the sections were incubated at 4 given dose to either zolpidem (a) or L-838,417 (b).
sterone up to about 800 ng/ml, a higher level comparable to that

seen in restraint-stressed rats (Kalman et al., 1997). Both diazepam
and zopiclone also produced a rise in plasma corticosterone.
However, the potencies and efficacies of diazepam and zopiclone
were significantly lower than for zolpidem (Fig. 2). Thus, a rise in
plasma corticosterone was seen after administration of these drugs
at 2.5 mg/kg and the level of corticosterone at this dose was in the
range of 400 ng/ml at maximal concentrations (Fig. 2). Notably,
while zolpidem induced a slightly higher level of corticosterone at
25 mg/kg compared to 12.5 mg/kg, the two non-selective
benzodiazepine ligands produced a slightly lower level of cortico-
sterone (Fig. 2). The two non-selective benzodiazepine ligands
produced a different dose – response curve, with zopiclone being
slightly stronger than diazepam. The two drugs were found to be
equally effective on the release of corticosterone. Based on the
Fig. 3. The effects of FG7142 (open circles) on plasma corticosterone
dose – response curves the ED50 were calculated to be 2.4 mg/kg
compared to zolpidem (closed circles). The compounds were given at 6
different doses 60 min before sacrifice. FG7142 significantly increased for zopiclone and 7.7 mg/kg for diazepam. In contrast, the a2,3,5
corticosterone at doses of 12.5 mg/kg and higher (P < 0.05). There was a subtype-specific agonist L-838,417 was ineffective at all doses
significant difference of FG7142 compared to zolpidem ( P < 0.01). Two- tested and produced only an insignificant increase in cortico-
way ANOVA and Student – Newman – Keuls test. sterone levels at the highest dose of 25 mg/kg (Fig. 2). Statistical
analysis revealed that both diazepam and zopiclone produced
corticosterone increased about 10-fold compared to naı̈ve animals significantly stronger effects on circulating stress hormones than
(Fig. 1) The time course illustrates a significant effect of zolpidem L-838,417, but significantly lower effect than zolpidem.
on ACTH release within 5 min after administration and reached its Treatment with a single dose of the a1 inverse agonist FG7142
maximal level at 10 min. The level of plasma ACTH 15 min after compared to zolpidem had relatively little effect on circulating
zolpidem was significantly higher than sham (zolpidem 332 T 70 corticosterone (Fig. 3). A significant release of corticosterone was
pg/ml vs. sham 27.6 T 6.8 pg/ml). Thereafter the ACTH level only observed at doses of 12.5 and 25 mg/kg, but even at these
declined over time and reached basal levels after 120 min (Fig. 1). doses the effects on corticosterone were much lower than after
The time course for corticosterone was slightly delayed compared zolpidem (Fig. 3).
to ACTH, and peaked about 60 min after zolpidem, and returned to
baseline after 120 min. 3.2. Effects of benzodiazepine ligands on Fos expression in the
Treatment with zolpidem increased plasma corticosterone at a paraventricular nucleus
minimum effective dose of 0.5 mg/kg reaching a maximum at 12.5
mg/kg (Fig. 2). The ED50 was calculated to be 1.5 mg/kg. In order to determine whether the effects of benzodiazepine
Zolpidem was effective and increased the circulating cortico- ligands occur at the level of the paraventricular nucleus and how
Fig. 4. a1 receptor agonists induce Fos in the mouse paraventricular nucleus. Distribution of Fos in the paraventricular region in naı̈ve mice (A), and after
administration with vehicle (B), L-838,417 (C), zopiclone (D), diazepam (E), and zolpidem (F). All drugs are given in a dose of 12.5 mg/kg 60 min before
sacrifice. Scale bar = 100 Am.
180 700
160 600
140
Corticosterine [ng/ml]
500
*
120
Fos cells/PVN
400
100
*
**
300
80
200
60
40 *** 100
20 0
Veh-Veh Veh-Zol DZ-Zol Zop-Zol
0
Vehicle Diazepam Zopiclone Zolpidem L-838,417
Fig. 7. Pre-treatment with diazepam or zopiclone (both 0.5 mg/kg) 15 min
Fig. 5. Quantificative analysis of Fos expression in the paraventricular before administration of 0.5 mg/kg zolpidem inhibited release of cortico-
nucleus after acute treatment with benzodiazepine ligands. The effect of Fos sterone. As negative control mice were treated with vehicle twice, and as
induction was analysed 60 min after single administration of drugs at 12.5 positive control zolpidem was administered 15 min after vehicle.
mg/kg, the same dose that produces a maximal increase of corticosterone in **P < 0.01; *P < 0.05 vs. zolpidem.
plasma. The values are given as the mean number of cells expressing Fos in
one section of the middle part of the paraventricular nucleus (mean - was found to be ineffective (Fig. 5). Furthermore, Fos induction
T S.E.M.). ***P < 0.001; *P < 0.05 compared to zolpidem. ANOVA non- after diazepam was significantly lower at a concentration of 12.5
parametric Tukey’s multiple comparison test. mg/kg, corresponding to the relative effect on corticosterone
levels (Fig. 2).
strong this input might be, Fos induction was studied in mice
treated with zolpidem, diazepam, zopiclone, and L-838,417 (all 3.3. Interaction/inhibition studies
drugs 12.5 mg/kg i.p.). A small number of Fos-immunoreactive
nuclei were observed in the medial paraventricular nucleus in Administration of 12.5 mg/kg L-838,417 immediately before
naı̈ve and vehicle treated animals (Fig. 4A and B). Similarly, in zolpidem (0.5, 2.5, or 12.5 mg/kg) produced a strong inhibition of
animals treated with L-838,417, a low number of Fos positive the release of corticosterone at all doses of zolpidem (Fig. 6). The
cells were observed in the paraventricular nucleus (Fig. 4C). maximal effects of the combined treatment of L-838,417 and
Zolpidem, diazepam, and zopiclone produced a strong increase of zolpidem were comparable to the maximal effect of the non-
Fos containing nuclei in the paraventricular nucleus (Figs. 4D – F selective benzodiazepines when given alone.
and 5). The activated cells were present in an area overlapping The level of corticosterone in mice pre-treated with 0.5 mg/kg
with the CRH containing medial parvocellular subportion of the diazepam before zolpidem was significantly lower than after
paraventricular nucleus, but also in the magnocellular portion of vehicle at a dose of 0.5 mg/kg zolpidem (Fig. 7). Also, pre-
nucleus (Fig. 4). Labelling was also found in the magnocellular treatment with 0.5 mg/kg zopiclone significantly inhibited the
supraoptic nucleus (not shown). Quantitative measurements of the effect of zolpidem at the same dose (Fig. 7).
number of Fos cells in the entire paraventricular nucleus revealed
a strong increase after all drugs tested except L-838,417, which
4. Discussion
1200
Activation of the HPA axis to threatening stressful
1000 environmental stimuli is essential, but accurate inhibitory
mechanisms are equally required to limit the magnitude in
800 activation of autonomic, neuroendocrine, and behavioural

** responses. The major mechanisms responsible for inhibition
600
of stress are the steroid-dependent negative feedback via
** steroid receptors in the hippocampus and the direct neural
400
** inhibition of paraventricular nucleus neurons driven by the
200 neurotransmitters GABA and substance P (Jessop et al.,
2000; Kovacs et al., 2004). Notably, part of the neuronal
0 inhibitory GABAergic input to the paraventricular nucleus
Veh 0.5 2.5 12.5 Zolpidem (mg/kg)
is a likely target of the corticoid-sensitive limbic forebrain
Fig. 6. Release of corticosterone by zolpidem (closed circles) is inhibited structures (Herman et al., 2003).
by pre-treatment with L-838,417 (open circles). Pre-administration of It has not been understood how anxiolytics like
L-838,417 (12.5 mg/kg) 15 min before one injection of increasing benzodiazepines can activate the HPA axis in high concen-
concentrations of zolpidem shows that L-838,417 inhibits the activation
by zolpidem at all doses tested. The values are given as mean trations, in particular whether it can be an effect independent
(n = 5) T S.E.M. Significance between groups was determined by two- of the GABAA receptor. We show here that acute adminis-
way ANOVA and **P < 0.01. tration of the GABAA receptor a1 subtype selective
benzodiazepine site agonist zolpidem strongly activates the A

HPA axis in mice. Furthermore, it is shown that zolpidem
induces expression of the transcription factor Fos in the α1 GABA α2 CRH CRH
medial part of the paraventricular nucleus, suggesting that
zolpidem is acting in the brain. Zolpidem stimulates rapidly
and robustly the secretion of both plasma ACTH and
corticosterone levels. The ED50 was found to be 1.5 mg/kg,
which is similar or lower than doses producing behavioural B
effects. The maximal response reached a level that has α1 GABA α2 CRH CRH
earlier been reported to occur in C57/B mice exposed to
restraint stress or forced swim (Droste et al., 2003). The
non-selective benzodiazepine site ligands, diazepam and
zopiclone, did not have the same maximal efficacy as
zolpidem on corticosterone levels. Finally, the a2,3,5 C
selective agonist L-838,417 was ineffective. Since zolpidem α1 GABA α2 CRH CRH
is selective for GABAA receptor complexes that contain a1
subunits over those containing a2, a3, or a5 subunits, and
zopiclone and diazepam are non-selective (Fleck, 2002;
Pritchett and Seeburg, 1990), it is suggested that activation Fig. 8. Schematic illustration of the model proposed for the action of
of a1 and non-a1 containing GABAA receptors may oppose selective and non-selective benzodiazepine ligands on the CRH neurons in
the paraventricular nucleus. Non-selective benzodiazepine receptor ligands
each other in stimulating CRH neurons in the paraven-
act on at least two different sites. One is outside the paraventricular nucleus
tricular nucleus. We consider the a2 subunit the main a and is disinhibiting the CRH neurons. The other is inside the nucleus
subunit opposing this action because it is highly expressed leading to a direct inhibition of the HPA output. The balance between
in all CRH synthesising neurons in the paraventricular activation of these two targets determines the release or inhibition of stress
nucleus (Cullinan, 2000). In contrast, a3 subunit mRNA is hormone secretion to the blood.
not expressed in the paraventricular nucleus and a5 subunit
mRNA expression is, to a great extent, limited to the selective action on a2, a3, and a5 subunits is also an
hippocampus (Tietz et al., 1999). antagonist at a1 receptors (McKernan et al., 2000), we
The proposed model illustrated in Fig. 8 illustrates that at cannot exclude the possibility that a competition between
least two GABAA receptors bearing at least two distinct a L838,417 and zolpidem at this receptor subtype explains our
subunits are involved in the regulation of the HPA axis findings. However, diazepam and zopiclone, both of which
output in vivo. Zolpidem activates the HPA axis by are full agonists at a1 receptors (Sieghart, 1995) at doses
inhibiting a GABAergic input onto CRH neurons. Several that do not themselves stimulate corticosterone release, also
lines of evidence support the presence of a disinhibiting inhibit zolpidem’s effect, emphasising the role of GABAA
GABAergic afferent input to the paraventricular nucleus. receptors containing an a subunit distinct from a1.
CRH neurons are under tonic inhibitory GABAergic There appeared to be a good correlation between the
control, because systemic administration of the GABAA relative rise in plasma corticosterone concentrations and the
receptor antagonist bicuculline both stimulates corticoster- number of Fos containing cells in the paraventricular
one secretion and mimics stress-induced ACTH release nucleus after administration of all drugs. However, this
(Ixart et al., 1983). Moreover, bicuculline administered does not mean that Fos induction is required to elicit a
directly into the paraventricular nucleus induces a rapid hormone output. The Fos signal in the paraventricular
increase in plasma ACTH (Cole and Sawchenko, 2002), and nucleus appears to be a good marker for neuronal
suppression of GABAergic transmission appears to play a excitability, emphasising, as illustrated in Fig. 8, that the
permissive role for inducing an increase in secretory activity net result of increased neuronal inhibition via activation of
of CRH in organotypic cultures (Bartanusz et al., 2004). a1 GABAA receptors outside the paraventricular nucleus is
Therefore, by disinhibiting the GABAergic input onto CRH an increased excitability inside the paraventricular nucleus.
neurons, zolpidem effectively stimulates the HPA axis (Fig. It is unlikely that zolpidem acts directly on the CRH
8). By contrast, direct inhibition of the paraventricular neurons, but rather on an inhibitory interneuron. Light and
nucleus likely via an a2 subunit containing GABAA electron microscopic studies indicate that nearly half of the
receptor expressed in CRH neurons inhibits the HPA axis. synapses within the paraventricular nucleus are GABAergic
Therefore, the balance between pharmacological activation (Decavel and Van den Pol, 1990) and CRH neurons express
of a1 and a2 containing GABAA receptors determines the GABAA receptor a and h subunits (Cullinan, 2000). This
response of benzodiazepine receptor ligands on the HPA raises the question: What are the inhibitory neurons
axis. This hypothetical relationship is demonstrated by the innervating the paraventricular nucleus that is modified by
antagonism of zolpidem-induced corticosterone release by zolpidem? Anatomical and functional studies suggest that
L-813,417 (Fig. 6). Since L-838,417 in addition to its sources of GABAergic inputs to the CRH neurons emanate
from the bed nucleus of the stria terminals, and these nucleus are important in this effect, and that this effect
neurons might be the same as those converging the negative represents a good and easy detectable biomarker for a2
steroid-dependent feedback to the paraventricular nucleus subtype efficacy in vivo.
(Herman and Cullinan, 1997). In addition, parvocellular In summary, despite the fact that diazepam, zolpidem,
neurons in the medial paraventricular subdivision receive and zopiclone all stimulate a1 containing GABAA receptors,
synaptic inputs from neurons in the lateral and dorsal their immediate effects on the HPA axis differ. We conclude
hypothalamus and the suprachiasmatic nucleus, all of which that the previous results reporting that various benzodiaze-
are considered to be GABAergic (Larsen et al., 1994; pines produce different effects on the basal level of HPA
Roland and Sawchenko, 1993; ter Horst and Luiten, 1986; activity in rodents are due to the balance between a1 and
Vrang et al., 1995). Glutamate microstimulation of neurons non-a1-dependent (presumably a2) modulation of the CRH
in the paraventricular nucleus shows that GABA producing neurons in the paraventricular nucleus.
neurons are abundant in this region, particularly those in a
region ventrolateral to the nucleus (Boudaba et al., 1996).
The same population of GABAergic neurons, in addition to Acknowledgements
those in the bed nucleus of the stria terminals, is considered
to be a part of the feedback inhibition to the paraventricular We would like to thank Tine Engelbrecht, Christina
nucleus, because the projections from the subiculum and Rasmussen, Pia Rovsing Sandholm, Torben Skov, and Pia
amygdala terminate rather around than inside the nucleus Weikop for skilful technical assistance.
(Canteras and Swanson, 1992; Herman and Cullinan, 1997;
Roland and Sawchenko, 1993). These GABAergic neurons
are the targets for zolpidem and remain to be established by
combining neural tracers and visualisation of a1 subunits. References
The available evidence suggests that the GABAergic
Arborelius, L., Owens, M.J., Plotsky, P.M., Nemeroff, C.B., 1999. The
neurons stimulated by stress are likely located in the role of corticotropin-releasing factor in depression and anxiety
paraventricular nucleus or at least in its close vicinity disorders. J. Endocrinol. 160, 1 – 12.
(Herman et al., 2003; Kovacs et al., 2004). It seems Atack, J.R., 2003. Anxioselective compounds acting at the GABA(A)
therefore plausible that the GABAergic neurons activated receptor benzodiazepine binding site. Curr. Drug. Target. CNS Neurol.
by stress are the same as those stimulated by benzodiazepine Disord. 2, 213 – 232.
Barnard, E.A., Skolnick, P., Olsen, R.W., Möhler, H., Sieghart, W., Biggio,
ligands. G., Braestrup, C., Bateson, A.N., Langer, S.Z., 1998. International
Significant effort has been given on developing drugs Union of Pharmacology: XV. Subtypes of gamma-aminobutyric acidA
which interact at the benzodiazepine recognition site but receptors: classification on the basis of subunit structure and receptor
which are devoid of some of the side effects, particularly in function. Pharmacol. Rev. 50, 291 – 313.
the tolerance and withdrawal (Atack, 2003). Separation of Bartanusz, V., Muller, D., Gaillard, R.C., Streit, P., Vutskits, L., Kiss,
J.Z., 2004. Local gamma-aminobutyric acid and glutamate circuit
the a1-dependent effect from the other a subunits is an control of hypophyseotrophic corticotropin-releasing factor neuron
important tool to select such new selective GABAA receptor activity in the paraventricular nucleus of the hypothalamus. Eur. J.
modulators. Thus, it is believed that development of a Neurosci. 19, 777 – 782.
subtype selective agents will provide better anxiolytics. Boudaba, C., Szabo, K., Tasker, J.G., 1996. Physiological mapping of
However, experimental models that can distinguish between local inhibitory inputs to the hypothalamic paraventricular nucleus.
J. Neurosci. 16, 7151 – 7160.
activities attributed to different a subtypes are required. Bowers, G., Cullinan, W.E., Herman, J.P., 1998. Region-specific regulation
Therefore, the delineation of the inhibitory inputs to the of glutamic acid decarboxylase (GAD) mRNA expression in central
paraventricular nucleus that are modulated by the a1 subunit stress circuits. J. Neurosci. 18, 5938 – 5947.
activation may provide insight into the mechanisms of Calogero, A.E., Kamilaris, T.C., Bernardini, R., Johnson, E.O., Chrou-
selective GABAA receptor ligands. Previous studies have sos, G.P., Gold, P.W., 1990. Effects of peripheral benzodiazepine
receptor ligands on hypothalamic – pituitary – adrenal axis function in
shown that benzodiazepine receptor ligands including L- the rat. J. Pharmacol. Exp. Ther. 253, 729 – 737.
838,417 inhibit the effects of various stressors (Le Fur et al., Canteras, N.S., Swanson, L.W., 1992. Projections of the ventral subiculum
1979; McKernan et al., 2000; Yagi and Onaka, 1996), and to the amygdala, septum, and hypothalamus: a PHAL anterograde tract-
our data suggest that this action could be mediated at the tracing study in the rat. J. Comp. Neurol. 324, 180 – 194.
Carrasco, G.A., Van de Kar, L.D., 2003. Neuroendocrine pharmacology of
level of the paraventricular nucleus. The doses of L-838,417
stress. Eur. J. Pharmacol. 463, 235 – 272.
and diazepam selected in the present study, exerting an Cole, R.L., Sawchenko, P.E., 2002. Neurotransmitter regulation of cellular
inhibitory effect on the HPA axis, correlate well with the activation and neuropeptide gene expression in the paraventricular
effective doses in models of anxiety (McKernan et al., nucleus of the hypothalamus. J. Neurosci. 22, 959 – 969.
2000). In line with the recent research emphasizing a role Cullinan, W.E., 2000. GABA(A) receptor subunit expression within
for a2 GABAA receptors in mediating the anxiolytic effect hypophysiotropic CRH neurons: a dual hybridization histochemical
study. J. Comp. Neurol. 419, 344 – 351.
of benzodiazepines (Löw et al., 2000; McKernan et al., De Boer, S.F., Van der Gugten, J., Slangen, J.L., 1990. Brain benzodiaze-
2000; Rudolph et al., 1999), it is tempting to speculate that pine receptor-mediated effects on plasma catecholamine and cortico-
a2 containing GABAA receptors in the paraventricular sterone concentrations in rats. Brain Res. Bull. 24, 843 – 847.
Decavel, C., Van den Pol, A.N., 1990. GABA: a dominant neurotransmitter Bristow, L., Marshall, G., Macaulay, A., Brown, N., Howell, O., Moore,
in the hypothalamus. J. Comp. Neurol. 302, 1019 – 1037. K.W., Carling, R.W., Street, L.J., Castro, J.L., Ragan, C.I., Dawson,
De Souza, E.B., 1990. Neuroendocrine effects of benzodiazepines. J. G.R., Whiting, P.J., 2000. Sedative but not anxiolytic properties of
Psychiatr. Res. 24 (Suppl. 2), 111 – 119. benzodiazepines are mediated by the GABA(A) receptor alpha1
Droste, S.K., Gesing, A., Ulbricht, S., Muller, M.B., Linthorst, A.C., Reul, subtype. Nat. Neurosci. 3, 587 – 592.
J.M., 2003. Effects of long-term voluntary exercise on the mouse Mehta, A.K., Ticku, M.K., 1999. An update on GABAA receptors. Brain
hypothalamic – pituitary – adrenocortical axis. Endocrinology 144, Res. Brain Res. Rev. 29, 196 – 217.
3012 – 3023. Mikkelsen, J.D., Vrang, N., Mrosovsky, N., 1998. Expression of Fos in the
Duncan, G.E., Breese, G.R., Criswell, H.E., McCown, T.J., Herbert, J.S., circadian system following nonphotic stimulation. Brain Res. Bull. 47,
Devaud, L.L., Morrow, A.L., 1995. Distribution of [3H]zolpidem 367 – 376.
binding sites in relation to messenger RNA encoding the alpha 1, beta Miklos, I.H., Kovacs, K.J., 2002. GABAergic innervation of corticotropin-
2 and gamma 2 subunits of GABAA receptors in rat brain. Neuro- releasing hormone (CRH)-secreting parvocellular neurons and its
science 64, 1113 – 1128. plasticity as demonstrated by quantitative immunoelectron microscopy.
Fenelon, V.S., Sieghart, W., Herbison, A.E., 1995. Cellular localization and Neuroscience 113, 581 – 592.
differential distribution of GABAA receptor subunit proteins and Owens, M.J., Bissette, G., Nemeroff, C.B., 1989. Acute effects of
messenger RNAs within hypothalamic magnocellular neurons. Neuro- alprazolam and adinazolam on the concentrations of corticotropin-
science 64, 1129 – 1143. releasing factor in the rat brain. Synapse 4, 196 – 202.
Fleck, M.W., 2002. Molecular actions of (S)-desmethylzopiclone (SEP- Paxinos, G., Franklin, K.B., 2004. The Mouse Brain in Stereotaxic
174559), an anxiolytic metabolite of zopiclone. J. Pharmacol. Exp. Coordinates, 2nd Edition Elsevier, Amsterdam.
Ther. 302, 612 – 618. Pivac, N., Pericic, D., 1993. Inhibitory effect of diazepam on the activity of
Herman, J.P., Cullinan, W.E., 1997. Neurocircuitry of stress: central control the hypothalamic – pituitary – adrenal axis in female rats. J. Neural
of the hypothalamo – pituitary – adrenocortical axis. Trends Neurosci. Transm.: Gen. Sect. 92, 173 – 186.
20, 78 – 84. Pritchett, D.B., Seeburg, P.H., 1990. Gamma-aminobutyric acid A receptor
Herman, J.P., Figueiredo, H., Mueller, N.K., Ulrich-Lai, Y., Ostrander, alpha 5-subunit creates novel type II benzodiazepine receptor pharma-
M.M., Choi, D.C., Cullinan, W.E., 2003. Central mechanisms of stress cology. J. Neurochem. 54, 1802 – 1804.
integration: hierarchical circuitry controlling hypothalamo – pituitary – Roland, B.L., Sawchenko, P.E., 1993. Local origins of some GABAergic
adrenocortical responsiveness. Front. Neuroendocrinol. 24, 151 – 180. projections to the paraventricular and supraoptic nuclei of the
Holsboer, F., 2000. The corticosteroid receptor hypothesis of depression. hypothalamus in the rat. J. Comp. Neurol. 332, 123 – 143.
Neuropsychopharmacology 23, 477 – 501. Rudolph, U., Crestani, F., Benke, D., Brunig, I., Benson, J.A., Fritschy,
Ixart, G., Cryssogelou, H., Szafarczyk, A., Malaval, F., Assenmacher, I., J.M., Martin, J.R., Bluethmann, H., Möhler, H., 1999. Benzodiazepine
1983. Acute and delayed effects of picrotoxin on the adrenocortico- actions mediated by specific gamma-aminobutyric acid(A) receptor
tropic system of rats. Neurosci. Lett. 43, 235 – 240. subtypes. Nature 401, 796 – 800.
Jensen, J.B., Mørk, A., Mikkelsen, J.D., 2001. Chronic antidepressant Sieghart, W., 1995. Structure and pharmacology of gamma-aminobutyric
treatments decrease pro-opiomelanocortin mRNA expression in the acid A receptor subtypes. Pharmacol. Rev. 47, 181 – 234.
pituitary gland: effects of acute stress and 5-HT(1A) receptor activation. Squires, R.F., Benson, D.I., Braestrup, C., Coupet, J., Klepner, C.A., Myers,
J. Neuroendocrinol. 13, 887 – 893. V., Beer, B., 1979. Some properties of brain specific benzodiazepine
Jessop, D.S., Renshaw, D., Larsen, P.J., Chowdrey, H.S., Harbuz, M.S., receptors: new evidence for multiple receptors. Pharmacol. Biochem.
2000. Substance P is involved in terminating the hypothalamo – Behav. 10, 825 – 830.
pituitary – adrenal axis response to acute stress through centrally located ter Horst, G.J., Luiten, P.G., 1986. The projections of the dorsomedial
neurokinin-1 receptors. Stress 3, 209 – 220. hypothalamic nucleus in the rat. Brain Res. Bull. 16, 231 – 248.
Kalman, B.A., Kim, P.J., Cole, M.A., Chi, M.S., Spencer, R.L., 1997. Tietz, E.I., Huang, X., Chen, S., Ferencak III, W.F., 1999. Temporal and
Diazepam attenuation of restraint stress-induced corticosterone levels is regional regulation of alpha1, beta2 and beta3, but not alpha2, alpha4,
enhanced by prior exposure to repeated restraint. Psychoneuroendocri- alpha5, alpha6, beta1 or gamma2 GABA(A) receptor subunit messenger
nology 22, 349 – 360. RNAs following one-week oral flurazepam administration. Neuro-
Kovacs, K.J., Miklos, I.H., Bali, B., 2004. GABAergic mechanisms science 91, 327 – 341.
constraining the activity of the hypothalamo – pituitary – adrenocortical Van de Kar, L.D., Blair, M.L., 1999. Forebrain pathways mediating stress-
axis. Ann. N. Y. Acad. Sci. 1018, 466 – 476. induced hormone secretion. Front. Neuroendocrinol. 20, 1 – 48.
Lakic, N., Pericic, D., Manev, H., 1986. Mechanisms by which picrotoxin Vargas, M.L., Abella, C., Hernandez, J., 2001. Diazepam increases the
and a high dose of diazepam elevate plasma corticosterone level. hypothalamic – pituitary – adrenocortical (HPA) axis activity by a cyclic
Neuroendocrinology 43, 331 – 335. AMP-dependent mechanism. Br. J. Pharmacol. 133, 1355 – 1361.
Larsen, P.J., Hay-Schmidt, A., Mikkelsen, J.D., 1994. Efferent connections Vrang, N., Larsen, P.J., Moller, M., Mikkelsen, J.D., 1995. Topo-
from the lateral hypothalamic region and the lateral preoptic area to the graphical organization of the rat suprachiasmatic – paraventricular
hypothalamic paraventricular nucleus of the rat. J. Comp. Neurol. 342, projection. J. Comp. Neurol. 353, 585 – 603.
299 – 319. Wilson, M.A., Biscardi, R., Smith, M.D., Wilson, S.P., 1996. Effects
Le Fur, G., Guilloux, F., Mitrani, N., Mizoule, J., Uzan, A., 1979. of benzodiazepine agonist exposure on corticotropin-releasing
Relationships between plasma corticosteroids and benzodiazepines in factor content and hormonal stress responses: divergent responses
stress. J. Pharmacol. Exp. Ther. 211, 305 – 308. in male and ovariectomized female rats. J. Pharmacol. Exp. Ther.
Löw, K., Crestani, F., Keist, R., Benke, D., Brunig, I., Benson, J.A., 278, 1073 – 1082.
Fritschy, J.M., Rulicke, T., Bluethmann, H., Möhler, H., Rudolph, U., Woldbye, D.P., Greisen, M.H., Bolwig, T.G., Larsen, P.J., Mikkelsen, J.D.,
2000. Molecular and neuronal substrate for the selective attenuation of 1996. Prolonged induction of c-fos in neuropeptide Y- and somatosta-
anxiety. Science 290, 131 – 134. tin-immunoreactive neurons of the rat dentate gyrus after electro-
Matheson, G.K., Gage, D., White, G., Dixon, V., Gipson, D., 1988. A convulsive stimulation. Brain Res. 720, 111 – 119.
comparison of the effects of buspirone and diazepam on plasma Yagi, K., Onaka, T., 1996. Chlordiazepoxide discriminates between the
corticosterone levels in rat. Neuropharmacology 27, 823 – 830. neural circuits mediating neuroendocrine responses to fear- and anxiety-
McKernan, R.M., Rosahl, T.W., Reynolds, D.S., Sur, C., Wafford, K.A., producing stimuli in the rat. Neurosci. Res. 24, 151 – 158.
Atack, J.R., Farrar, S., Myers, J., Cook, G., Ferris, P., Garrett, L.,

Gaba

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gaba

Uploaded by

Copyright:

Available Formats

GABA neurocircuits controlling the Hypothalamic-

4.7. Detailed analysis ............................................................................................................................................................. 43

1.1. SUMMARY OF THE PROJECT AIMS:

• What type of neurons is activated in the PVN?

2.1.1. GABAERGIC NEUROTRANSMISSION

2.1.2. GABAA RECEPTOR BIOLOGY

2.2. THE HYPOTHALAMUS

2.2.1. DEVELOPMENT OF THE HYPOTHALAMUS

Figure 4 schematic drawing of the development of the hypothalamus

2.2.2. ANATOMICAL BOUNDARIES

2.2.3. BLOOD SUPPLY

2.3. THE PARAVENTRICULAR NUCLEUS

2.3.1. ANATOMY OF THE PVN

2.3.2. THE SUBPARAVENTRICULAR AREA

2.4. BIOLOGICAL STRESS

2.4.1. THE HPA AXIS AND THE CIRCADIAN RHYTHMS

2.4.2. NEGATIVE FEEDBACK REGULATION OF THE HPA AXIS

2.5. C-FOS AS A MARKER OF ACTIVITY IN NEUROENDOCRINE SYSTEMS.

2.6. RETROGRADE NEUROTRACING USING CHOLERA TOXIN SUBUNIT B

2.7.1. BASIC CHEMISTRY

Figure 15 1,4 Benzodiazepine

Figure 16 1,4 Benzodiazepin-2-one

Figure 17 N-methyl-7-chloro-5-phenyl-1,4-Benzodiazepin-2-one (Diazepam)

Figure 18 Shows the chemical structure of Zolpidem (N,N,6-trimethyl-2-(4-methylphenyl)imides[1,2-a]pyridine-3-acetamide hemitartrate)

Figure 19 shows the chemical structure of Diazepam (N-methyl-7-chloro-5-phenyl-1,4-Benzodiazepin-2-one)

Figure 20 Shows the Chemical structure of Zopiclone (4-methyl-1-piperazinecarboxylic acid 6-(5-chloro-2-pyridinyl)-6,7-dihydro-7-oxo-5H-

Zopiclone, is a cyclopyrrolone derivative, the drug is used as a short-acting hypnotic. Although

Figure 21 Shows the chemical structure of L-838,417 (7-tert-Butyl-3-(2,5-difluoro-phenyl)-6-(2-methyl-2H-[1,2,4]triazol-3-ylmethoxy)-

3. METHODS AND MATERIALS.

3.3. SINGLE DOSING OF DRUGS

3.4. COMBINATION OF DRUGS

3.6. FOS STUDIES

3.7. DOUBLE LABELLING FOS / OXYTOCIN

3.8. DATA ANALYSIS

3.8.2. DOUBLE LABELLING FOS / OXYTOCIN

3.9. RETROGRADE TRACING WITH CHOLERA TOXIN SUBUNIT B (CHB)

Figure 22 shows a wistar male rat mounted on a stereotaxic frame

3.9.1. STATUS CHB INJECTED ANIMALS

Identification number Position of injection site Further analysis

A138 No hit No further analysis

3.10. IN SITU HYBRIDIZATION, LOCATING THE ANATOMICAL DISTRIBUTION OF

3.11. IN SITU HYBRIDIZATION FOLLOWED BY IMMUNOCYTOCHEMISTRY

3.11.1. PROBE PREPARATION PROTOCOL

The following was mixed in an Rnase-free 1.5 µl screw-top eppendorf tube:

3.0 µl 5× transcription buffer

3.11.2. IN SITU HYBRIDIZATION FOLLOWED BY IMMUNOCYTOCHEMICAL STAINING

3.12. ELECTRONIC PICTURE PROCESSING

4.1. EFFECTOF ACUTE TREATMENT WITH BENZODIAZEPINES / GABAA

4.2. INTERACTION INHIBITION STUDIES

4.5. IDENTIFYING FOS EXPRESSING NEURONS IN THE PVN

FOS positive cell

4.6. RESULTS FROM IN SITU HYBRIDIZATION

4.7. DETAILED ANALYSIS

4.8. GABAA ALPHA 1 DISTRIBUTION IN THE PVN AND SUBPVN AREA

4.9. GABAA ALPHA 2 DISTRIBUTION IN THE PVN AND SUBPVN AREA

4.10. GABAA ALPHA 3 DISTRIBUTION IN THE PVN AND SUBPVN AREA

4.11. RESULTS FROM CHB RETROGRADE NEURO-TRACING.

A ChB hit in the PVN

4.11.1. 6.11.1. PROJECTIONS OF THE BSNT TO THE SUBPVN AREA

4.12. DUAL IN SITU HYBRIDIZATION / IMMUNOCYTOCHEMICAL STAIN

5.1. GABA NEUROCIRCUITS CONNECTING THE BED NUCLEUS STRIA

5.2. METHODOLOGICAL CONSIDERATIONS

5.2.3. COMBINATION OF DRUGS