Professional Documents
Culture Documents
2020 Osteoporosis and Osteoarthritis.-humana
2020 Osteoporosis and Osteoarthritis.-humana
2020 Osteoporosis and Osteoarthritis.-humana
Osteoporosis
and Osteo-
arthritis
Second Edition
METHODS IN MOLECULAR BIOLOGY
Series Editor
John M. Walker
School of Life and Medical Sciences
University of Hertfordshire
Hatfield, Hertfordshire, UK
Edited by
This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer
Nature.
The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.
Preface
v
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
vii
viii Contents
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Contributors
ix
x Contributors
Abstract
Osteocytes are thought to be the mechanosensors of bone by sensing mechanical loads imposed upon the
bone and transmitting these signals to the other bone cells to initiate bone modeling and remodeling. The
location of osteocytes deep within bone is ideal for their function. However, this location makes the study
of osteocytes in vivo technically difficult. There are several methods for obtaining and culturing primary
osteocytes for in vitro experiments and ex vivo observation. In this chapter, several proven methods are
discussed including the isolation of avian osteocytes from chicks and osteocytes from calvaria and long
bones of young mice. A detailed protocol for the isolation of osteocytes from hypermineralized bone of
mature and aged animals is provided. In addition, a modified version of this protocol that can be used to
isolate osteocytes from human trabecular bone is described.
1 Introduction
Osteocytes are the most abundant of the bone cells and recently
found to be multifunctional [1, 2]. They serve as orchestrators of
bone remodeling and regulators of mineral homeostasis. They are
the mechanosensors of bone, sensing imposed bone loads, and
translating these mechanical signals into biological signals of bone
modeling and remodeling. They are housed in cave-like voids
within the bone called lacunae. Their location deep within the
mineralized bone matrix is ideal for their cellular functions, but
makes their observation and study difficult. Methods to isolate
these bone matrix-embedded cells have been developed through-
out the years and vary by the species, state, and extent of minerali-
zation of the bone.
In 1992, the group of Peter Nijweide was the first to describe
the isolation of osteocytes from 18-day-old chick embryos. Their
approach yielded a relatively pure population of osteocytes based on
morphology [3]. In 1995, Kumegawa and colleagues published a
method for isolating primary avian osteocytes from the parietal
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_1, © Springer Science+Business Media, LLC, part of Springer Nature 2021
3
4 Matthew Prideaux et al.
2 Materials
3 Methods
3.1 Isolation 1. Aseptically dissect the long bones (femurs, tibiae, and humeri)
of Osteocytes from from the mice using surgical scissors or scalpel. Be sure not to
Skeletally Mature break any of the bones at this point and also try to keep the
Mouse Bone abdomen intact during the dissection to reduce contamination
potential (see Note 4).
2. After dissection of bones and removal of as much soft tissue as
possible, place them in 100 mm petri dishes containing αMEM
with 1% penicillin and streptomycin (and gentamicin (25 μg/
mL)—optional) (see Note 4).
3. Remove any remaining muscle and connective tissue from the
bones and scrape away the periosteum using a scalpel (see
Note 4).
8 Matthew Prideaux et al.
3.2 Isolation 1. Under sterile conditions, use the bone cutters to remove tra-
of Osteocytes from becular bone pieces from the surgical samples (see Note 14).
Human Trabecular 2. Further dissect the bone pieces into 1–2 mm fragments using
Bone the bone cutters or a scalpel.
3. Place the bone pieces (1–1.5 g, wet weight) into a 50 mL
culture tube.
4. Wash the bone pieces 3 times with 20 mL HBSS containing
Pen/Strep.
5. Digest the bone pieces in 20 mL pre-warmed collagenase type
II (2 mg/mL in α-MEM containing Pen/Strep) for 25 min
with gentle shaking at 37 C.
6. Remove the collagenase from the bone pieces. Wash the bone
pieces 3 times with 10 mL HBSS (see Note 15).
7. Incubate the bone pieces in a second digest of 20 mL collage-
nase for 25 min. Collect the supernatant and save. Wash the
bone pieces 3 times in 10 mL HBSS and add the rinsate to the
saved collagenase solution.
8. Repeat step 7 (third collagenase digest) and combine the
digest solutions and the rinsates. Centrifuge at 500 g for
5 min and resuspend the pelleted cells in 3 mL of osteoblast
media. Divide between 2–3 wells of a collagen-coated 12-well
plate (see Notes 16 and 17).
10 Matthew Prideaux et al.
4 Notes
References
1. Bonewald LF (2011) The amazing osteocyte. J genes. Biochem Biophys Res Commun 246
Bone Miner Res 26(2):229–238 (2):404–408
2. Dallas SL, Prideaux M, Bonewald LF (2013) 15. Mikuni-Takagaki Y (1999) Mechanical
The osteocyte: an endocrine cell and more. responses and signal transduction pathways in
Endocr Rev 34(5):658–690 stretched osteocytes. J Bone Miner Metab 17
3. van der Plas A, Nijweide PJ (1992) Isolation (1):57–60
and purification of osteocytes. J Bone Miner 16. Mikuni-Takagaki Y et al (1996) Distinct
Res 7(4):389–396 responses of different populations of bone
4. Tanaka K et al (1995) Time-lapse microcine- cells to mechanical stress. Endocrinology 137
matography of osteocytes. Miner Electrolyte (5):2028–2035
Metab 21(1–3):189–192 17. Gu G et al (2005) Estrogen protects primary
5. Aarden EM et al (1996) Adhesive properties of osteocytes against glucocorticoid-induced apo-
isolated chick osteocytes in vitro. Bone 18 ptosis. Apoptosis 10(3):583–595
(4):305–313 18. Kato Y et al (2001) Establishment of an osteoid
6. Ajubi NE et al (1996) Pulsating fluid flow preosteocyte-like cell MLO-A5 that spontane-
increases prostaglandin production by cultured ously mineralizes in culture. J Bone Miner Res
chicken osteocytes—A cytoskeleton- 16(9):1622–1633
dependent process. Biochem Biophys Res 19. Kato Y et al (1997) Establishment of an
Commun 225(1):62–68 osteocyte-like cell line, MLO-Y4. J Bone
7. Kamioka H, Honjo T, Takano-Yamamoto T Miner Res 12(12):2014–2023
(2001) A three-dimensional distribution of 20. Woo SM et al (2011) Cell line IDG-SW3 repli-
osteocyte processes revealed by the combina- cates osteoblast-to-late-osteocyte differentia-
tion of confocal laser scanning microscopy and tion in vitro and accelerates bone formation
differential interference contrast microscopy. in vivo. J Bone Miner Res 26(11):2634–2646
Bone 28(2):145–149 21. Zhao S et al (2002) MLO-Y4 osteocyte-like
8. Kamioka H et al (2007) Primary cultures of cells support osteoclast formation and activa-
chick osteocytes retain functional gap junctions tion. J Bone Miner Res 17(11):2068–2079
between osteocytes and between osteocytes 22. Kramer I et al (2010) Osteocyte Wnt/beta-
and osteoblasts. Microsc Microanal 13 catenin signaling is required for normal bone
(2):108–117 homeostasis. Mol Cell Biol 30(12):3071–3085
9. Kamioka H et al (2006) Fluid shear stress 23. Nakashima T et al (2011) Evidence for osteo-
induces less calcium response in a single pri- cyte regulation of bone homeostasis through
mary osteocyte than in a single osteoblast: RANKL expression. Nat Med 17
implication of different focal adhesion forma- (10):1231–1234
tion. J Bone Miner Res 21(7):1012–1021 24. Stern AR et al (2012) Isolation and culture of
10. Klein-Nulend J et al (1995) Pulsating fluid flow primary osteocytes from the long bones of skel-
increases nitric oxide (NO) synthesis by osteo- etally mature and aged mice. BioTechniques 52
cytes but not periosteal fibroblasts--correlation (6):361–373
with prostaglandin upregulation. Biochem 25. Jahn K et al (2012) Skeletal muscle secreted
Biophys Res Commun 217(2):640–648 factors prevent glucocorticoid-induced osteo-
11. Klein-Nulend J et al (1995) Sensitivity of cyte apoptosis through activation of beta-
osteocytes to biomechanical stress in vitro. catenin. Eur Cell Mater 24:197–209. discus-
FASEB J 9(5):441–445 sion 209-10
12. Westbroek I et al (2000) Differential stimula- 26. Kalajzic I et al (2013) In vitro and in vivo
tion of prostaglandin G/H synthase-2 in approaches to study osteocyte biology. Bone
osteocytes and other osteogenic cells by pulsat- 54(2):296–306
ing fluid flow. Biochem Biophys Res Commun 27. Jilka RL (2013) The relevance of mouse mod-
268(2):414–419 els for investigating age-related bone loss in
13. Mikuni-Takagaki Y et al (1995) Matrix miner- humans. J Gerontol A Biol Sci Med Sci 68
alization and the differentiation of osteocyte- (10):1209–1217
like cells in culture. J Bone Miner Res 10 28. Prideaux M et al (2014) SaOS2 osteosarcoma
(2):231–242 cells as an in vitro model for studying the tran-
14. Kawata A, Mikuni-Takagaki Y (1998) Mechan- sition of human osteoblasts to osteocytes. Cal-
otransduction in stretched osteocytes—Tem- cif Tissue Int 95(2):183–193
poral expression of immediate early and other
Isolation of Murine and Human Osteocytes 13
29. Bodine PV, Vernon SK, Komm BS (1996) 31. Qing H et al (2012) Demonstration of osteo-
Establishment and hormonal regulation of a cytic perilacunar/canalicular remodeling in
conditionally transformed preosteocytic cell mice during lactation. J Bone Miner Res 27
line from adult human bone. Endocrinology (5):1018–1029
137(11):4592–4604 32. Skottke J, Gelinsky M, Bernhardt A (2019) In
30. Prideaux M et al (2016) Isolation of osteocytes vitro co-culture model of primary human
from human trabecular bone. Bone 88:64–72 osteoblasts and osteocytes in collagen gels. Int
J Mol Sci 20(8):1998
Chapter 2
Abstract
Human bone marrow-derived mesenchymal stem/stromal cells (BM-MSC) are adult multipotent progeni-
tor cells that can be isolated from bone marrow. BM-MSCs have the ability to be expanded and differ-
entiated into the chondrogenic lineage in vitro. Here we describe a standardized method to expand and
chondrogenically differentiate human BM-MSCs, highlighting how to overcome technical challenges and
indicating the most common readout parameters to evaluate the chondrogenic differentiation capacity.
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_2, © Springer Science+Business Media, LLC, part of Springer Nature 2021
15
16 Roberto Narcisi et al.
2 Materials
2.1 Expansion of 1. Complete expansion medium (CEM; see also Note 1):
BM-MSCs (See (a) MEM-α, nucleosides.
Table 1)
(b) 10% Fetal calf serum (FCS; see Note 2).
(c) 1 ng/mL Fibroblast growth factor-2 (FGF2).
(d) 104 M Ascorbic acid-2-phosphate.
(e) 1.5 μg/mL Fungizone (Amphotericin B).
(f) 50 μg/mL Gentamicin.
2. CEM for enhanced expansion and chondrogenic differentia-
tion capacity (CEM+):
(a) CEM.
(b) 250 ng/mL WNT3A or 2.5 μM CHIR99021, both
canonical WNT agonists (see Note 3).
Expansion and Chondrogenic Differentiation of Human Bone Marrow-Derived. . . 17
Table 1
Summary of the materials for the expansion and passaging of BM-MSC
3. Passaging of BM-MSCs:
(a) CEM.
(b) Trypsin-EDTA (0.25%) phenol red (Gibco).
(c) Dulbecco’s phosphate-buffered saline, no calcium, no
magnesium (DPBS).
Table 2
Summary of the materials for the chondrogenic differentiation of BM-MSC
3 Methods
After determining the total volume of the aspirate, the cells are
counted by mixing 20 μL of bone marrow aspirate with 380 μL of
3% acetic acid with methylene blue (or comparable volumes in a
1:20 ratio). After 2 min the membrane of white and red blood cells
will be lysate and the remaining white blood cell nuclei can be
counted, for example, with a Bürker counting chamber. The cells
are then seeded at a density of approximately 75,000 25,000
cells/cm2 in CEM (see Subheading 2.1) for a total volume of 20 mL
in a T175 flasks or 8.5 mL in a T75 flasks (see Notes 6 and 7). In
this way BM-MSCs are isolated by their ability to adhere to plastic
culture flasks. After 24 h, nonadherent cells are removed by 3
washing with DPBS+2% FCS and adherent cells cultured in stan-
dard conditions (5% CO2 at 37 C) for up to 14 days. During this
time, cell colonies appear and grow at a speed that depend on
individual BM-MSC donor (Fig. 1a). Medium is renewed twice a
week. When BM-MSCs neared confluence (Fig. 1b) or after a
maximum of 14 days, they are detached and passaged for further
expansion (see Subheading 2.1, step 3 and Subheading 3.1). At this
point we have Passage-1 (P1) BM-MSC.
3.1 Expansion and P1 BM-MSC are 2 washed with DPBS by adding around 50% of
Passaging of BM- the volume used for the expansion media, then Trypsin-EDTA is
MSCs added (3 mL for a T175 flask and 1.25 mL for a T75) for 3–4 min
in the culture incubator (37 C and 5%CO2) till all the cells look
detached after a check with the microscope. The cell suspension is
harvested in appropriate sized tubes by adding CEM (containing
FCS) in order to neutralize the effect of the Trypsin-EDTA solu-
tion. For the harvesting, it is suggested to use a Trypsin-EDTA/
CEM ratio of at least a 1:4. The cell suspension is then centrifuged
at 300 g for 6–8 min in 50 or 15 mL tubes, the resulting cellular
pellet resuspended in CEM or CEM+ and the cell counted, for
example, with a Bürker counting chamber. Next, the BM-MSC are
reseeded at the density of 2300 cells/cm for further expansion (see
Note 6). Generally, BM-MSC take 4–6 days to reach
sub-confluence (75–85% confluence; Fig. 1b) after seeding (see
Note 8). In order to keep as much as possible the BM-MSC in
continuous proliferation, we do not grow them till confluence
(Fig. 1c).
Every passage we monitor the BM-MSC for their morphology
(Fig. 2) and expansion speed. These parameters are known to be
influenced over time in vitro and have an impact on BM-MSC
differentiation capacity, and one reason may be cellular senescence
[18–20]. For example, when during expansion BM-MSC take
>50% more time to reach the sub-confluence compared to the
time needed to go from P1 to P2, we stop the culture and do not
proceed with the differentiation assay (see Note 9). By default, we
would then no longer perform surface marker analysis on our
plastic-adherent-selected BM-MSC (see Note 10).
20 Roberto Narcisi et al.
Fig. 1 Representative images of BM-MSC forming a colony 2 days after isolation (a), at 75–85% confluence (b)
and over-confluent (c). Scale bar ¼ 200 μm
Fig. 2 Representative images of BM-BMCs showing their classical morphology during the expansion (left
panel) and a more enlarged morphology (right panel, black arrow) usually associate with aging and low
expansion rate (right panel). Scale bar ¼ 100 μm
Fig. 3 Representative images of pellet cultures (200,000 BM-MSCs/tube) immediately after centrifugation (a),
after 24 h in the presence of CCM (b), and after 10 days of chondrogenic induction in CCM (c). Arrows indicate
the pellets on the bottom of the tubes
3.3 Evaluation of In this section you will find the general information about our
Chondrogenic standard assays to evaluate chondrogenic differentiation. This sec-
Differentiation tion is not meant to provide all the details necessary to perform the
Capacity of BM-MSC assays, but more to guide you toward the minimum set the analysis
we suggest to perform to verify chondrogenic maturation of the
tissue.
3.3.2 Transcript Analysis After chondrogenic differentiation we extract the mRNA by first
(See Note 13) mechanically disrupting the pellet with a micro pestle (Eppendorf,
cat.: 0030 120.973) in a 1.5 mL Eppendorf tube containing
350 μL of ice-cold RNA-STAT60 solution (Gentaur, cat:
CS-111-200; see Note 14). At this stage the samples can be stored
at 80 C till the moment of the RNA extraction (see Note 15),
which is done following the manufacturer’s instructions.
We generally perform RT-PCR analysis on the following genes
1. Chondrogenic genes.
COL2A1 and ACAN (optional: COL2A1-IIB, SOX9)
2. Hypertrophic genes.
COL10A1 and ALPL (optional: RUNX2).
We have a panel of several housekeeper/reference genes and
among them the most used are GAPDH, RPS27a, β-ACTIN, 18S,
and HTPR1. We select our reference gene based on its stability
across the conditions and/or time-points, with the intention to
select one stable gene applicable for all the experiments. However,
in case of high variations between the conditions of the same
experiments (generally 1 Ct value) we apply the “best house-
keeper index” (BHI [21]). We run the BHI by using at least
3 housekeeper/reference genes.
4 Notes
Acknowledgements
References
1. Tavassoli M, Crosby WH (1968) Transplanta- their use in cell therapy. Exp Hematol
tion of marrow to extramedullary sites. Science 28:707–715
161:54–56 13. Gronthos S, Zannettino AC, Hay SJ, Shi S,
2. Friedenstein AJ, Petrakova KV, Kurolesova AI, Graves SE, Kortesidis A et al (2003) Molecular
Frolova GP (1968) Heterotopic of bone mar- and cellular characterisation of highly purified
row. Analysis of precursor cells for osteogenic stromal stem cells derived from human bone
and hematopoietic tissues. Transplantation marrow. J Cell Sci 116:1827–1835
6:230–247 14. Sacchetti B, Funari A, Michienzi S, Di
3. Pittenger MF, Discher DE, Peault BM, Phin- Cesare S, Piersanti S, Saggio I et al (2007)
ney DG, Hare JM, Caplan AI (2019) Mesen- Self-renewing osteoprogenitors in bone mar-
chymal stem cell perspective: cell biology to row sinusoids can organize a hematopoietic
clinical progress. NPJ Regen Med 4:22 microenvironment. Cell 131:324–336
4. Owen M, Friedenstein AJ (1988) Stromal stem 15. Digirolamo CM, Stokes D, Colter D, Phinney
cells: marrow-derived osteogenic precursors. DG, Class R, Prockop DJ (1999) Propagation
Ciba Found Symp 136:42–60 and senescence of human marrow stromal cells
5. Caplan AI (1991) Mesenchymal stem cells. J in culture: a simple colony-forming assay iden-
Orthop Res 9:641–650 tifies samples with the greatest potential to
6. Bianco P, Gehron RP (2000) Marrow stromal propagate and differentiate. Br J Haematol
stem cells. J Clin Invest 105:1663–1668 107:275–281
7. Dominici M, Le Blanc K, Mueller I, Slaper- 16. Martin I, Muraglia A, Campanile G,
Cortenbach I, Marini F, Krause D et al Cancedda R, Quarto R (1997) Fibroblast
(2006) Minimal criteria for defining multipo- growth factor-2 supports ex vivo expansion
tent mesenchymal stromal cells. The Interna- and maintenance of osteogenic precursors
tional Society for Cellular Therapy position from human bone marrow. Endocrinology
statement. Cytotherapy 8:315–317 138:4456–4462
8. Cleary MA, Narcisi R, Focke K, van der 17. Sivasubramaniyan K, Ilas DC, Harichandan A,
Linden R, Brama PA, van Osch GJ (2016) Bos PK, Santos DL, de Zwart P et al (2018)
Expression of CD105 on expanded mesenchy- Bone marrow-harvesting technique influences
mal stem cells does not predict their chondro- functional heterogeneity of mesenchymal
genic potential. Osteoarthr Cartil 24:868–872 stem/stromal cells and cartilage regeneration.
Am J Sports Med 46:3521–3531
9. Johnstone B, Hering TM, Caplan AI, Gold-
berg VM, Yoo JU (1998) In vitro chondrogen- 18. Narcisi R, Cleary MA, Brama PA, Hoogduijn
esis of bone marrow-derived mesenchymal MJ, Tuysuz N, ten Berge D et al (2015) Long-
progenitor cells. Exp Cell Res 238:265–272 term expansion, enhanced chondrogenic
potential, and suppression of endochondral
10. Pittenger MF, Mackay AM, Beck SC, Jaiswal ossification of adult human MSCs via WNT
RK, Douglas R, Mosca JD et al (1999) Multi- signaling modulation. Stem Cell Rep
lineage potential of adult human mesenchymal 4:459–472
stem cells. Science 284:143–147
19. Gruber HE, Somayaji S, Riley F, Hoelscher
11. Muraglia A, Cancedda R, Quarto R (2000) GL, Norton HJ, Ingram J et al (2012)
Clonal mesenchymal progenitors from human Human adipose-derived mesenchymal stem
bone marrow differentiate in vitro according to cells: serial passaging, doubling time and cell
a hierarchical model. J Cell Sci 113 senescence. Biotech Histochem 87:303–311
(Pt 7):1161–1166
20. Banfi A, Bianchi G, Notaro R, Luzzatto L,
12. Banfi A, Muraglia A, Dozin B, Cancedda R, Quarto R (2002) Replicative
Mastrogiacomo M, Cancedda R, Quarto R aging and gene expression in long-term cul-
(2000) Proliferation kinetics and differentia- tures of human bone marrow stromal cells.
tion potential of ex vivo expanded human Tissue Eng 8:901–910
bone marrow stromal cells: implications for
28 Roberto Narcisi et al.
21. Pfaffl MW, Tichopad A, Prgomet C, Neuvians Hypertrophy in mesenchymal stem cell chon-
TP (2004) Determination of stable drogenesis: effect of TGF-beta isoforms and
housekeeping genes, differentially regulated chondrogenic conditioning. Cells Tissues
target genes and sample integrity: Best- Organs 192:158–166
Keeper—Excel-based tool using pair-wise cor- 26. Cals FL, Hellingman CA, Koevoet W, Baaten-
relations. Biotechnol Lett 26:509–515 burg de Jong RJ, van Osch GJ (2012) Effects
22. Chen HH, Decot V, Ouyang JP, Stoltz JF, of transforming growth factor-beta subtypes
Bensoussan D, de Isla NG (2009) In vitro on in vitro cartilage production and minerali-
initial expansion of mesenchymal stem cells is zation of human bone marrow stromal-derived
influenced by the culture parameters used in mesenchymal stem cells. J Tissue Eng Regen
the isolation process. Biomed Mater Eng Med 6:68–76
19:301–309 27. Barry F, Boynton RE, Liu B, Murphy JM
23. Willert K, Brown JD, Danenberg E, Duncan (2001) Chondrogenic differentiation of mes-
AW, Weissman IL, Reya T et al (2003) Wnt enchymal stem cells from bone marrow:
proteins are lipid-modified and can act as stem differentiation-dependent gene expression of
cell growth factors. Nature 423:448–452 matrix components. Exp Cell Res
24. Narcisi R, Arikan OH, Lehmann J, Ten 268:189–200
Berge D, van Osch GJ (2016) Differential 28. Cleary MA, Narcisi R, Albiero A, Jenner F, de
effects of small molecule WNT agonists on Kroon LMG, Koevoet W et al (2017) Dynamic
the multilineage differentiation capacity of regulation of TWIST1 expression during chon-
human mesenchymal stem cells. Tissue Eng drogenic differentiation of human bone
Part A 22:1264–1273 marrow-derived mesenchymal stem cells.
25. Mueller MB, Fischer M, Zellner J, Berner A, Stem Cells Dev 26:751–761
Dienstknecht T, Prantl L et al (2010)
Chapter 3
Abstract
Bone marrow mesenchymal stem cells (MSCs) are promising therapeutic tools for tissue repair and
treatment of a number of human diseases. As a result, there is substantial interest in characterizing and
expanding these cells to uncover their therapeutic potential. Bone marrow mesenchymal progenitors,
containing both MSCs and their proliferative progeny, are commonly isolated from the central region of
rodent long bones. However, challenges exist in expanding these central mesenchymal progenitors in
culture. We have designed an enzymatic digestion protocol to isolate mesenchymal progenitors within
rodent long bones that resides close to the bone surface, which we termed endosteal mesenchymal
progenitors. These cells are more metabolically active and more responsive to external stimuli compared
to central mesenchymal progenitors. Therefore, they represent a biologically important target for MSC
research. This chapter describes the approach in detail how to isolate and culture endosteal mesenchymal
progenitors as well as their central counterparts from rodent long bones.
Key words Mesenchymal stem cells, Endosteal mesenchymal progenitors, Bone marrow, Enzymatic
digestion, Colony forming unit-fibroblast
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_3, © Springer Science+Business Media, LLC, part of Springer Nature 2021
29
30 Leilei Zhong et al.
2 Materials
2.1 Animals This protocol was developed from experiments performed with
Sprague-Dawley rats and C57Bl/6 mice (see Note 1). It is also
compatible with other strains of mice we have tested, such as 129.
2.2 Instruments 1. Class II biological safety cabinet/cell culture hood and a hori-
zontal laminar flow clean bench: both should be equipped with
a UV light for decontamination.
A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors 31
3 Methods
3.2 Isolation 1. Under a sterile tissue culture hood, use forceps and a scalpel to
of Central remove all of the soft tissue surrounding the bones. After
and Endosteal Bone removal of the soft tissue, place the long bones into a new
marrow Cells 100 mm Petri dish filled with 10 mL of flushing medium.
2. Cut off both ends of the tibia and femur at the growth plate
with a scalpel.
3. Fill a syringe with 5 (mouse) or 10 (rat) mL of flushing medium
per animal (2 tibiae and 2 femurs). Attach a 25- (rat) or
27-gauge needle (mouse) to the syringe.
4. With the prefilled syringe, drill a hole at each end of the bone
and press down the plunger at one end to force medium
through the bone. This will flush the bone marrow out of the
bone through the opposite end of the bone.
5. Reverse the bone and repeat the flushing from the other end of
the bone. Use 1 (mouse) or 2 (rat) mL of flushing medium to
flush out each bone. Bone marrow cells released by flushing
mainly come from the central part of the diaphyseal shaft, and
hence are central bone marrow cells (Fig. 1 step 1, see Note 4).
The bone marrow cells that are located in close proximity to the
A Digestion Method for Isolating Bone Marrow Mesenchymal Progenitors 33
Fig. 1 Representative images of a rat femur during the isolation of endosteal bone marrow. Step 1: flush out
central bone marrow cells from a bone that is free of its surrounding soft tissue and with both ends removed at
the growth plates; Step 2: predigest the whole bone to remove the periosteal progenitors and then
longitudinally cut the bones into two halves; Step 3: gently wash the bones to remove loosely attached
bone marrow; Step 4: digest bone fragments to collect endosteal bone marrow cells. BM: bone marrow.
Reprinted from Bone, 53(2), Siclari et al., Mesenchymal progenitors residing close to the bone surface are
functionally distinct from those in the central bone marrow. 575–86, Copyright (2013), with permission from
Elsevier
3.3 CFU-F Assays 1. Seed 1 106 mononuclear endosteal bone marrow cells per
of Endosteal 25 cm2 flask in the growth medium and incubate the culture at
Mesenchymal 37 C in 5% CO2 in a humidified tissue culture incubator (see
Progenitors Note 10).
2. Typically, after 5 (rat) or 7 (mouse) days of incubation, most
colonies should contain more than 50 fibroblastic cells (Fig. 2a,
b, see Note 11). At this point, remove medium and wash flasks
twice with PBS.
3. Stain the colonies with 3% crystal violet in methanol for at least
1 hour at RT.
4. Rinse flasks thoroughly with tap water to remove unbound
stain.
5. Air-dry the flasks completely.
6. Count the number of CFU-F colonies under an inverted micro-
scope with a 4 objective. We recommend drawing lines on the
bottom of the flask to divide the surface into 8 regions in order
to facilitate counting. Only count colonies larger than 50 cells.
The expected CFU-F frequency of endosteal bone marrow cells
is about 80–150 CFU-Fs per 1 106 mononuclear cells from
both mouse and rat (Fig. 2c, d). The size of CFU-F colonies is
normally larger in endosteal bone marrow compared to central
bone marrow (Fig. 2c, d, see Note 12).
Fig. 2 CFU-F assays of rodent endosteal bone marrow cells compared to their central counterparts. (a)
Representative images of 25 cm2 flasks with mouse central and endosteal CFU-F colonies after staining. Note
that the initial seeding densities are 3 106 and 1 106 cells per flask for central and endosteal bone
marrow, respectively. (b) Representative images and morphologies of mouse CFU-F colonies at low (top) and
high (bottom) magnification. (c, d) Quantification of CFU-F frequency and diameter of endosteal bone marrow
cells from mouse (c) and rat (d). **: p < 0.01 vs. central
Fig. 3 Endosteal mesenchymal progenitors are capable of multi-lineage differentiation. (a) In vitro osteogenic
differentiation as detected by von Kossa staining. (b) In vitro adipogenic differentiation as detected by oil red O
staining. (c) In vitro chondrogenic pellet differentiation as detected by alcian blue staining
4 Notes
Acknowledgments
References
1. Friedenstein AJ, Chailakhjan RK, Lalykina KS 9. Ellis SL, Grassinger J, Jones A et al (2011) The
(1970) The development of fibroblast colonies relationship between bone, hemopoietic stem
in monolayer cultures of Guinea-pig bone mar- cells, and vasculature. Blood 118:1516–1524
row and spleen cells. Cell Tissue Kinet 10. van Gastel N, Torrekens S, Roberts SJ et al
3:393–403 (2012) Engineering vascularized bone: osteo-
2. Friedenstein AJ, Gorskaja JF, Kulagina NN genic and proangiogenic potential of murine
(1976) Fibroblast precursors in normal and periosteal cells. Stem Cells 30:2460–2471
irradiated mouse hematopoietic organs. Exp 11. Morikawa S, Mabuchi Y, Kubota Y et al (2009)
Hematol 4:267–274 Prospective identification, isolation, and sys-
3. Lindner U, Kramer J, Rohwedel J et al (2010) temic transplantation of multipotent mesen-
Mesenchymal stem or stromal cells: toward a chymal stem cells in murine bone marrow. J
better understanding of their biology? Transfus Exp Med 206:2483–2496
Med Hemother 37:75–83 12. Xu S, De Becker A, Van Camp B et al (2010)
4. Wang S, Qu X, Zhao RC (2012) Clinical appli- An improved harvest and in vitro expansion
cations of mesenchymal stem cells. J Hematol protocol for murine bone marrow-derived
Oncol 5:19.5 mesenchymal stem cells. J Biomed Biotechnol
5. Short B, Brouard N, Occhiodoro-Scott T et al 2010:105940
(2003) Mesenchymal stem cells. Arch Med Res 13. Nakamura Y, Arai F, Iwasaki H et al (2010)
34:565–571 Isolation and characterization of endosteal
6. Krishnappa V, Boregowda SV, Phinney DG niche cell populations that regulate hemato-
(2013) The peculiar biology of mouse mesen- poietic stem cells. Blood 116:1422–1432
chymal stromal cells--oxygen is the key. 14. Ohishi M, Ono W, Ono N et al (2012) A novel
Cytotherapy 15:536–541 population of cells expressing both hemato-
7. Banfi A, Muraglia A, Dozin B et al (2000) poietic and mesenchymal markers is present in
Proliferation kinetics and differentiation poten- the normal adult bone marrow and is aug-
tial of ex vivo expanded human bone marrow mented in a murine model of marrow fibrosis.
stromal cells: implications for their use in cell Am J Pathol 180:811–818
therapy. Exp Hematol 28:707–715 15. Short BJ, Brouard N, Simmons PJ (2009) Pro-
8. Siclari VA, Zhu J, Akiyama K et al (2013) Mes- spective isolation of mesenchymal stem cells
enchymal progenitors residing close to the from mouse compact bone. Methods Mol
bone surface are functionally distinct from Biol 482:259–268
those in the central bone marrow. Bone 16. Zhu H, Guo ZK, Jiang XX et al (2010) A
53:575–586 protocol for isolation and culture of mesenchy-
mal stem cells from mouse compact bone. Nat
Protoc 5:550–560
Chapter 4
Abstract
Cells isolated from the intervertebral disc are often used for in vitro experimentation. Correctly separating
the intervertebral disc tissue in annulus fibrosus and nucleus pulposus is particularly challenging when
working with surplus material from surgery or specimens from donors with an advanced age. Moreover,
lineage controls are only sparsely reported to verify tissue of origin. Here we describe an approach to
intervertebral disc cell isolation from human and bovine origin.
Key words Intervertebral disc, Cell isolation, Nucleus pulposus, Annulus fibrosus, Collagenase, Cell
characterization
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_4, © Springer Science+Business Media, LLC, part of Springer Nature 2021
41
42 Guus G. H. van den Akker et al.
IVD cells are isolated from tissues for cellular studies. Multiple
types of (sequential) enzymatic digestions at different concentra-
tions and durations have been reported. Digestion time, type of
enzyme, and its concentration are known to affect cell viability
[6]. In addition, the time post-mortem at which cells are isolated
may affect cell viability and/or phenotype. For human post-mortem
IVD tissue, maximum sampling times of 24–72 h have
been reported [1, 7]. Surgical surplus material and animal tissues
are typically obtained within hours of dissection or death of the
animal. Surgical surplus material from scoliosis correction surgery
often consists of smaller fragments; this demands careful selection,
labeling, orientation, and dissection to obtain NP and AF cell
populations. In case of surplus material from IVD herniation sur-
gery it is recommended to verify tissue of origin by histological
analyses (NP/AF). Histological scoring of IVD degeneration in
human and certain animal samples can be done by various systems,
e.g., Mankin [8], Thompson [9], or modified Mankin [10]. Based
on our earlier work, we recommend a pre-isolation comparison
and/or histological analyses of NP and AF tissue of the same
donor to ensure correct isolation of the cell type of interest [11].
Following enzymatic digestion of the tissues, cells are strained,
washed, and plated at a high cell density in monolayer cultures or
brought into a 3D culture environment. A plethora of (pseudo) 3D
culture options are available, i.a. pellets, micromasses, alginate,
Matrigel, etc. This choice is mainly guided by the prevailing
research question, e.g., to expand cells, limit chondrocyte dediffer-
entiation, or mimic the microenvironment [3, 12–14]. Importantly,
different cell culture media were shown to affect marker gene
expression of IVD cells [15]. DMEM or mixtures of DMEM with
alpha-MEM or Ham’s F-12 are recommended for expansion of
IVD cells. Finally, it is recommended to perform post-isolation
measurements of IVD marker expression to confirm the successful
isolation of NP or (outer) AF cell populations [11, 16, 17].
The ORS/Spine Research Interest group released a consensus
definition of young healthy human NP cells [18]. The recom-
mended markers are: stabilized expression of HIF-1, GLUT-1,
aggrecan/collagen II ratio >20, Shh, Brachyury, KRT18/19,
CA12, and CD24 to assess variation between NP cells. However,
it may not be practical to perform all recommended measurements
for each individual donor. Moreover, NP cells from aged or degen-
erated tissue are known to have different expression profiles
[19]. Nevertheless, we advocate the systematic measurement of a
limited marker subset of NP and AF phenotypic marker genes as
standard practice to acquire a reference for and record on common
characteristics of isolated types. This is in particular valuable when
both NP and AF cells are isolated from one donor. These markers
provide an internal reference for the donor and the experimenter. A
subset of marker genes that we have successfully used in situ and
Isolation of Cells from the Intervertebral Disc 43
2 Materials
2.1 Medical Ethical Prior to performing the procedures described below, it is obligatory
Approval for Human to obtain approval from a local medical ethical committee for the
Tissues use of tissues from deceased individuals or surgical leftover material
for research purposes prior to performing the procedures described
below.
2.2 Cell Culture 1. Fetal calf serum (FCS): Stock is thawed overnight at 4 C.
Decomplement FCS for 30 min at 56 C in a water bath. Fill
out 25 mL FCS per sterile polypropylene tube. Store at 20 C
in upright position.
2. Anti-Anti, Antibiotic Antimycotic (A/A), Gibco 15240-062:
Stock (100) is thawed at 37 C in a water bath. Aliquot
5.5 mL per tube, store at 20 C in upright position.
3. Non-essential amino acids (NEAA), Gibco 11140-035: Stock
(100) is thawed at 37 C in a water bath. Stock is stored at
4 C.
4. Sodium chloride (0.9%): Dissolve 9 g sodium chloride (suitable
for cell culture, >99% purity, Mw ¼ 58.44 g/mol) in 1 L
MQ-water. Sterilize the solution by autoclaving (121 C,
15 min). Store at room temperature.
5. DMEM/F12 (1:1) Glutamax, Gibco 31331-028.
6. DMEM/F12 (1:1), HEPES buffered, Gibco 31330-038.
7. Collagenase II, Invitrogen 17101-015: 10 concentrated
stock ¼ 300 U/mL or 1.1%. Dissolve in HEPES-buffered
DMEM/F12 with 1% A/A. Aliquot in 5 mL portions in
50 mL tubes and store at 20 C until use. Thaw and dilute
to 1 (NP) or 2 (AF) in HEPES-buffered DMEM/F12 with
1% A/A.
44 Guus G. H. van den Akker et al.
2.3 Cell Counting 1. Bürker Turk cell counter: Depth 0.1 mm, surface per square
0.04 mm2.
2. Trypan blue (0.4%), Sigma-Aldrich Chemie T8154.
3. Phosphate-buffered saline (PBS, 20 concentrated stock):
87.52 g Sodium chloride (suitable for cell culture, >99%
purity, Mw ¼ 58.44 g/mol), 14.16 g disodium hydrogen
phosphate dihydrate (>99.5% purity, Mw ¼ 177.99 g/mol),
2.15 g potassium dihydrogen phosphate (>99.5% purity,
Mw ¼ 136.09 g/mol), fill up to 500 mL. pH is between the
required range of 7.2–7.4, no further pH-correction is needed.
Stock is diluted 20 in MQ water to obtain 1 stock. Sterilize
by autoclaving (121 C, 15 min).
2.4 IVD Marker Gene RNA was isolated according to standard protocols and cDNA
Analyses by RT-qPCR synthesis was done using random hexameres. RT-qPCR reactions
were set up in Takyon™ No Rox SYBR Master Mix dTTP blue
(Eurogentec) in a CFX96 Real-Time PCR Detection System
(Biorad) according to the MIQE guideline [20]. Primer pair
sequences for human and bovine were described before and are
provided in Tables 1 and 2 [11, 16, 17].
Isolation of Cells from the Intervertebral Disc 45
Table 1
RT-qPCR primer sequences for human NP (T & KRT19) and AF (SFRP2 & COL12A1) marker genes
Human
Table 2
RT-qPCR primer sequences for bovine NP (T & Krt19) and AF (Sfrp2 & Col12a1) marker genes
Bovine
3 Methods
3.1 Tissue 1. Tissues are obtained as fresh as possible from human or animal
Dissection and Cell donors (<24 h post-mortem).
Isolation It is recommended to transfer tissues as quickly as possible
to a sterile solution of PBS with 1 A/A (4 C).
2. A sterile metal plate or large sterile petri dish is used for sorting
tissue pieces and dissection with a sterile scalpel blade (size 20)
in a laminar flow cabinet (see Notes 1, 2 and Fig. 1 for tips on
recognizing the boundaries of tissues).
46 Guus G. H. van den Akker et al.
Fig. 1 Dissection of an intact bovine tail IVD. (a) The bovine tail IVD was separated from the vertebral body. The
border between NP and inner AF (iAF) is difficult to recognize by eye. A transversal cut from outer AF (oAF) and
the gradual loss of lamellae become obvious. (b) Pushing the AF toward an open position (indicated by arrows)
with a pincer reveals the demarcation of the inter zone and allows excision of the central NP. (c) Correct
isolation of amorphous proteoglycan-rich NP tissue was confirmed by formalin fixed, paraffin-embedded
histology (Safranin O staining)
Fig. 2 Use of the Bürker Turk counting chamber. The cells of a single field are
counted within the square including the cells overlapping with the top and left
sides of the small square and not those overlapping with the right and bottom
sides. Only cells enclosed in the large square are counted
48 Guus G. H. van den Akker et al.
3.4 Passaging IVD Passaging of IVD cells is usually done 1:4 and it may take 4–5 days
Cells in Monolayer for human NP cells to reach the same confluency after each passage.
Adjust this ratio accordingly; if less confluent, consider splitting 1:2
or 1:3 and if more confluent consider splitting 1:5 or 1:6. AF cells
usually reach confluency faster. An example of typical morphology
of human NP and AF cells is provided below (see Fig. 3).
1. Remove culture medium and wash with an appropriate amount
of 0.9% sodium chloride
Optional: wash twice to reduce the required trypsinization
time. This can be particularly useful at P0 when the cells have
generated a large amount of extracellular matrix.
2. Remove all sodium chloride without disturbing the cells and
add X mL Trypsin-EDTA according to the amounts indicated
below.
Fig. 3 Typical morphology of NP and AF cells in vitro. Cells were isolated from three human scoliotic IVD donors
and photographed using a Nikon Eclipse TE200 microscope. NP cells tend to prefer cell-cell contact and can be
more spindle shaped when compared to AF cells. D donor, pX passage number. Bar represents 20 μm
Isolation of Cells from the Intervertebral Disc 49
3.5 Freezing Down 1. Write down the donor number, cell type, passage number, and
IVD Cells date on the cryotubes.
2. Trypsinize and count the cells according to “Subheading 3.4.”
3. Centrifuge for 5 min at 316 RCF at room temperature (RT).
4. Remove supernatant and resuspend the IVD cells at a density of
2. 106 cells/mL.
5. Add cold 2 freeze down medium 1:1 drop-by-drop and mix
at the same time.
Note: be quick, DMSO at this concentration is toxic to the
cells.
6. Aliquot 1 mL cell suspension (1. 106 cells) per cryotube and
close the cap. Put cryotubes on ice immediately when freezing
multiple donors.
7. Put the vials in a refrigerated Mr. Frosty and store overnight at
80 C to freeze the cells.
Note: A maximum of 18 cryotubes fits in one Mr. Frosty.
8. The next day, transfer the cryotubes to a liquid nitrogen con-
tainer for long-term storage.
3.6 Thawing IVD 1. Transport tubes from the liquid nitrogen on ice to the cell
Cells from the Liquid culture laboratory.
Nitrogen 2. Release any liquid nitrogen that might have gotten into the
tube by slowly unscrewing the cap in the flow cabinet.
50 Guus G. H. van den Akker et al.
3.7 RT-qPCR In brief, RT-qPCR reactions are set up with 15 ng cDNA, 300 nM
Analyses of NP and AF forward and 300 nM reverse primers. The program was: 50 C for
Marker Genes 2 min, denaturation at 95 C for 10 min, 40 cycles of amplification
(15 s 95 C and 1 min 60 C) and followed by a melting curve
(95–50 C, 1 C/min). The standard curve method was used to
calculate gene expression and this was normalized to stable refer-
ence genes indicated in Tables 1 and 2.
4 Notes
Acknowledgements
References
1. Power KA, Grad S, Rutges JPHJ, Creemers 6. Molinos M, Almeida CR, Goncalves RM, Bar-
LB, van Rijen MHP, O’Gaora P, Wall JG, bosa MA (2015) Improvement of bovine
Alini M, Pandit A, Gallagher WM (2011) Iden- nucleus pulposus cells isolation leads to identi-
tification of cell surface–specific markers to tar- fication of three phenotypically distinct cell
get human nucleus pulposus cells: expression subpopulations. Tissue Eng Part A 21
of carbonic anhydrase XII varies with age and (15–16):2216–2227. https://doi.org/10.
degeneration. Arthritis Rheum 63 1089/ten.tea.2014.0461
(12):3876–3886. https://doi.org/10.1002/ 7. Thorpe AA, Binch ALA, Creemers LB,
art.30607 Sammon C, Le Maitre CL (2016) Nucleus
2. McCann MR, Séguin CA (2016) Notochord pulposus phenotypic markers to determine
cells in intervertebral disc development and stem cell differentiation: fact or fiction? Onco-
degeneration. J Dev Biol 4(1):3. https://doi. target 7(3):2189–2200. https://doi.org/10.
org/10.3390/jdb4010003 18632/oncotarget.6782
3. Pattappa G, Li Z, Peroglio M, Wismer N, 8. Mankin HJ, Dorfman H, Lippiello L, Zarins A
Alini M, Grad S (2012) Diversity of interverte- (1971) Biochemical and metabolic abnormal-
bral disc cells: phenotype and function. J Anat ities in articular cartilage from osteo-arthritic
221(6):480–496. https://doi.org/10.1111/j. human hips. II Correlation of morphology
1469-7580.2012.01521.x with biochemical and metabolic data. J Bone
4. Minogue BM, Richardson SM, Zeef LA, Free- Joint Surg Am 53(3):523–537
mont AJ, Hoyland JA (2010) Transcriptional 9. Thompson JP, Pearce RH, Schechter MT,
profiling of bovine intervertebral disc cells: Adams ME, Tsang IK, Bishop PB (1990) Pre-
implications for identification of normal and liminary evaluation of a scheme for grading the
degenerate human intervertebral disc cell phe- gross morphology of the human intervertebral
notypes. Arthritis Res Ther 12(1):R22–R22. disc. Spine (Phila Pa 1976) 15(5):411–415.
https://doi.org/10.1186/ar2929 https://doi.org/10.1097/00007632-
5. Hayes AJ, Isaacs MD, Hughes C, Caterson B, 199005000-00012
Ralphs JR (2011) Collagen fibrillogenesis in 10. Thomas CM, Fuller CJ, Whittles CE, Sharif M
the development of the annulus fibrosus of (2007) Chondrocyte death by apoptosis is
the intervertebral disc. Eur Cell Mater associated with cartilage matrix degradation.
22:226–241 Osteoarthr Cartil 15(1):27–34. https://doi.
org/10.1016/j.joca.2006.06.012
52 Guus G. H. van den Akker et al.
11. van den Akker GGH, Koenders MI, van de Loo 11(1):e0144497–e0144497. https://doi.org/
FAJ, van Lent PLEM, Blaney Davidson E, van der 10.1371/journal.pone.0144497
Kraan PM (2017) Transcriptional profiling dis- 17. van den Akker GGH, Surtel DAM, Cremers A,
tinguishes inner and outer annulus fibrosus from Rodrigues-Pinto R, Richardson SM, Hoyland
nucleus pulposus in the bovine intervertebral JA, van Rhijn LW, Welting TJM, Voncken JW
disc. Eur Spine J 26(8):2053–2062. https://doi. (2014) Novel immortal human cell lines reveal
org/10.1007/s00586-017-5150-3 subpopulations in the nucleus pulposus.
12. Tang X, Jing L, Richardson WJ, Isaacs RE, Arthritis Res Ther 16(3):R135. https://doi.
Fitch RD, Brown CR, Erickson MM, Setton org/10.1186/ar4597
LA, Chen J (2016) Identifying molecular phe- 18. Risbud MV, Schoepflin ZR, Mwale F, Kandel
notype of nucleus pulposus cells in human RA, Grad S, Iatridis JC, Sakai D, Hoyland JA
intervertebral disc with aging and degenera- (2015) Defining the phenotype of young
tion. J Orthop Res 34(8):1316–1326. healthy nucleus pulposus cells: recommenda-
https://doi.org/10.1002/jor.23244 tions of the spine research interest group at
13. Huang Y-C, Hu Y, Li Z, Luk KDK (2018) the 2014 annual ORS meeting. J Orthop Res
Biomaterials for intervertebral disc regenera- 33(3):283–293. https://doi.org/10.1002/
tion: current status and looming challenges. J jor.22789
Tissue Eng Regen Med 12(11):2188–2202. 19. Rutges J, Creemers LB, Dhert W, Milz S,
https://doi.org/10.1002/term.2750 Sakai D, Mochida J, Alini M, Grad S (2010)
14. Johnson WEB, Stephan S, Roberts S (2008) Variations in gene and protein expression in
The influence of serum, glucose and oxygen on human nucleus pulposus in comparison with
intervertebral disc cell growth in vitro: implica- annulus fibrosus and cartilage cells: potential
tions for degenerative disc disease. Arthritis Res associations with aging and degeneration.
Ther 10(2):R46–R46. https://doi.org/10. Osteoarthr Cartil 18(3):416–423. https://
1186/ar2405 doi.org/10.1016/j.joca.2009.09.009
15. Schubert AK, Smink JJ, Pumberger M, 20. Bustin SA, Benes V, Garson JA, Hellemans J,
Putzier M, Sittinger M, Ringe J (2018) Standar- Huggett J, Kubista M, Mueller R, Nolan T,
disation of basal medium for reproducible culture Pfaffl MW, Shipley GL, Vandesompele J,
of human annulus fibrosus and nucleus pulposus Wittwer CT (2009) The MIQE guidelines:
cells. J Orthop Surg Res 13:14. https://doi.org/ minimum information for publication of quan-
10.1186/s13018-018-0914-y titative real-time PCR experiments. Clin Chem
16. van den Akker GGH, Surtel DAM, Cremers A, 55(4):611. https://doi.org/10.1373/
Richardson SM, Hoyland JA, van Rhijn LW, clinchem.2008.112797
Voncken JW, Welting TJM (2016) Novel 21. Roth S (1974) Tissue Culture. Methods and
immortal cell lines support cellular heterogene- Applications. Paul F. Kruse, Jr. , M. K. Patter-
ity in the human annulus Fibrosus. PLoS One son, Jr. Q Rev Biol 49(2):150–151. https://
doi.org/10.1086/408024
Chapter 5
Abstract
Co-culture of chondrocytes and mesenchymal stromal cells (MSCs) has been shown to be beneficial in
engineering cartilage tissue in vitro. In these co-cultures, MSCs increase the proliferation and matrix
deposition of chondrocytes. The MSCs accomplish this beneficial effect by so-called trophic actions.
Thus, large cartilage constructs can be made with a relatively small number of chondrocytes. In this chapter,
we describe different methods for making co-cultures of MSCs and chondrocytes. We also provide detailed
protocols for analyzing MSC-chondrocyte co-cultures with cell tracking, proliferation assays, species-
specific polymerase chain reactions (PCR), rheological analysis, compression analysis, RNA-sequencing
analysis, short tandem repeats analysis, and biochemical examination.
Key words Chondrocytes, Mesenchymal stromal cells, Co-culture, Trophic effects, Cartilage engi-
neering, Matrix deposition
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_5, © Springer Science+Business Media, LLC, part of Springer Nature 2021
53
54 Yao Fu et al.
2 Materials
2.1 Cell Sources 1. Bovine primary chondrocytes (bPCs) are isolated from full-
thickness cartilage knee biopsies of female calves that are
approximately 6 months old. Cartilage is separated and
digested to extract primary chondrocytes (see Subheading 3.1).
2. Human primary chondrocytes (hPCs) are obtained from full-
thickness cartilage dissected from knee biopsies of a patient
undergoing total knee replacement (see Subheading 3.2).
3. Human MSCs (hMSCs) are isolated from bone marrow aspi-
rates of healthy donors (see Note 1).
Fig. 1 Schematic illustration of PTEE mold for hydrogel formation (a and b) and PDMS convex mold for micro-
aggregation (c)
3 Methods
3.1 Isolation of 1. Human cartilage tissue was obtained from total knee or hip
Human Articular joint replacement.
Chondrocytes 2. Cartilage tissue is cut from underlying bone and connective
tissue with scalpels and chopped into pieces of approximately
2 2 mm.
3. Digest cartilage pieces for 20–22 h in collagenase type II
(0.15%) in DMEM supplemented with penicillin (100 U/
mL) and streptomycin (100 mg/mL).
4. Seed the cell mixture onto the agarose chips resulting in aggre-
gates consisting of 100 cells. Immediately after seeding, centri-
fuge the agarose chips at 300 g for 3 min to allow the cells to
settle down in the microwells and then incubate at 37 C for
24 h to form micro-aggregates.
5. The micro-aggregates can be obtained by flushing the well
plates with medium and centrifugation.
11. Open green image of the same pellet; set threshold and area
restriction (see step 6) to count NUMBER of green cell.
12. Run “Image calculator” by selecting green image in Image
1 box and red image in Image 2 box, with AND in Operation
box to generate new image named “result of green.”
13. Run “Analyze particles” on new image “result of green” with
same setting as above to count NUMBER of green plus red cell.
14. Input all NUMBERs into an Excel spreadsheet and perform
the following calculations: Rate of EdU positive
Chondrocyte¼ NUMBER of green plus red cell NUMBER of
red cell 100%; Rate of EdU positive MSCs¼(NUMBER of green
cell – NUMBER of green plus red cell)(NUMBER of total cell – N
UMBER of red cell) 100%; Labeling efficiency ¼ NUMBER of
red cell NUMBER of total cell100% (see Note 8).
3.9 Quantitative GAG 1. Perform glycosaminoglycan (GAG) and DNA assay at the end
and DNA Assay of co-culture (i.e., 4 weeks).
2. Wash cell pellets (n ¼ 6) with PBS and freeze pellets overnight
at 80 C.
3. Digest pellets in 500 μL of proteinase K digestion buffer (see
Note 9) for more than 16 h at 56 C.
4. To prepare a standard curve, make dilution series of cysteine-
HCl, according to Table 1.
5. Add 5 μL of a 2.3 M NaCl solution and 25 μL of the samples or
the standard in one well of a 96-well non-tissue culture treated
plate.
6. Add 150 μL of the DMMB (1, 9-Dimethyl-Methylene Blue)
solution (see Subheading 2.2) and read the absorbance at
520 nm on an ELISA reader. Figure 2 gives an example of a
standard curve (see Subheading 2.2).
7. Determine cell number by quantification of total DNA using a
CyQuant DNA Kit, according to the manufacturer’s
instructions.
Table 1
Series dilution of GAG standards
Fig. 2 An example of standard curve for GAG quantification. The blank (25 μL
PBE, 5 μL 2.3 M NaCl and 150 μL DMMB solution) has an O.D. value of
0.18 0.03
3.10 Cell Tracking 1. Perform species-specific PCR to determine the ratio of MSCs
with Species-Specific and chondrocytes in xenogeneic co-culture (hMSCs and bPCs)
PCR pellets at the end of culture (i.e., 4 weeks).
2. Isolate DNA samples of pellets with a QIAamp DNA Mini Kit
according to the manufacturer’s protocols.
3. Extract RNA samples of pellets with an RNeasy Mini Kit (see
Note 10).
4. Reverse-transcribe one microgram of total RNA into cDNA
using the iScript cDNA Synthesis kit.
5. Perform species-specific quantitative PCR (qPCR) on genomic
DNA or cDNA samples by using the iQ SYBR Green
Supermix.
6. Carry out PCR Reactions on MyiQ2 Two-Color Real-Time
PCR Detection System under the following conditions: cDNA
is denatured for 5 min at 95 C, followed by 45 cycles consist-
ing of 15 s at 95 C, 15 s at 60 C, and 30s at 72 C.
7. Generate a melting curve for each reaction to test primer dimer
formation and nonspecific priming.
8. The primers for real-time PCR, either species specific or cross-
species specific, are listed in Tables 2 and 3.
9. For each gene, standard curves are obtained by serial dilutions
of cDNA (see Note 11). Figure 3 gives an example of standard
curve for qPCR.
62 Yao Fu et al.
Table 2
Forward (F) and Reverse (R) primers used for quantitative PCR on genomic DNA
Product
Gene name Primer sequence size Gene bank no.
0 0
Cross-species F: 5 GCATTGCCCTCAACGACCA 3 179 OR NC_000012 &
GAPDH R: 50 CACCACCCTGTTGCTGTAGCC 30 171a NC_007303
Human-specific F: 50 TTCCACCCATGGCAAATTCC 30 131 NC_000012
GAPDH R: 50 TTGCCTCCCCAAAGCACATT 30
Bovine-specific F: 50 AGCCGCATCCCTGAGACAAG 30 132 NC_007303
GAPDH R: 50 CAGAGACCCGCTAGCGCAAT 30
a
Product size of human genomic GAPDH is 179, of bovine genomic GAPDH is 171
Table 3
Forward (F) and Reverse (R) primers used for quantitative RT-PCR
Product
Gene name Primer sequence size Gene bank no.
Cross-species F: 50 GCGCAAGTACTCCGTGTGGA 30 123 NM_001101 &
β-actin R: 50 AAGCATTTGCGGTGGACGAT 30 NM_173979
Cross-species F: 50 AGCTCACTGGCATGGCCTTC 30 116 NM_002046&
GAPDH R: 50 CGCCTGCTTCACCACCTTCT 30 NM_001034034
Human-specific F: 50 CGCTCTCTGCTCCTCCTGTT 30 82 NM_002046
GAPDH R: 50 CCATGGTGTCTGAGCGATGT 30
Bovine-specific F: 50 GCCAT CACTG CCACC CAGAA 30 207 NM_001034034
GAPDH R: 50 GCGGCAGGTCAGATCCACAA 30
Human-specific F: 50 TTCCCATCGTGCCTTTCCA 30 121 NM_013227
aggrecan R: 50 AACCAACGATTGCACTGCTCTT 30
Bovine-specific F: 50 CCAAGCTCTGGGGAGGTGTC 30 98 NM_173981
aggrecan R: 50 GAGGGCTGCCCACTGAAGTC 30
Human-specific F: 50 GGCGGGGAGAAGACGCAGAG 30 129 NM_001844
collagen II R: 50 CGCAGCGAAACGGCAGGA 30
Bovine-specific F: 50 AGGTCTGACTGGCCCCATTG 30 101 NM_001001135
collagen II R: 50 CTCGAGCACCAGCAGTTCCA 30
Human-specific F: 50 GGCAGAAATGGCCGAGACG 30 150 NM_001851
collagen IX R: 50 CCCTTTGTTAAATGCTCGCTGA 30
Bovine-specific F: 50 GGACTCAACACGGGTCCACA 30 102 XM_601325
collagen IX R: 50 ACAGGTCCAGCAGGGCTTTG 30
Methods for Co-culturing of Chondrocytes and MSCs 63
Standard Unknown
35
30
Threshold Cycle
25
20
15
10
-0,5 0 0,5 1 1,5 2 2,5
Log Starting Quantity,copy number
Fig. 3 An example of standard curve for qPCR. SYBR, SYBR1, and SYBR2 stand for three different primer sets
Fig. 4 Example figures of mechanical properties analysis. (a) Storage modulus curve for rheological analysis.
(b) Stress–strain curve for compressive test
1X
n
RPKMaverage ¼ RPKMn > 0:3
n
i¼1
RPKMmonoculture
%RPKMmale ¼
RPKMcoculture
5. To determine the fold change, compare the real values with the
expected values:
RPKM Coculturereal
Fc ¼
RPKM Cocultureexpected
6. Values above 2 are considered highly upregulated and values
below 0.5 downregulated (see Note 16).
Fig. 5 Example figures of RNA-sequencing data analysis. (a) Heatmap of collagen-related genes comparison in
different conditions. (b) Inter-connecting cluster showing collagen-related genes
Methods for Co-culturing of Chondrocytes and MSCs 67
3.14 Short Tandem 1. Perform STR analysis to determine the ratio of MSCs and
Repeats (STR) Analysis chondrocytes in allogeneic co-cultures (hMSCs and hPCs)
pellets.
2. Extracts genomic DNA samples from pellets (n ¼ 6) with the
QIAamp DNA Mini Kit.
3. Amplify the sixteen loci of the kit PowerPlex 16 System, type
“sequence,” and analyze all loci according to manufacturer’s
protocol.
4. Compare monocultures of hMSCs or hPCs to find informative
alleles only present in either the hMSCs or the hPCs donor (see
Note 17).
5. Make electropherograms of the informative loci.
6. In the resulting electropherogram, calculate the area under the
peaks, which reflects the abundance of the alleles.
7. The sum of the area under the peak for the two donor-specific
alleles represents a relative amount of DNA for this donor.
8. Calculate the relative DNA amount for both the hMSC and the
hPC donor.
9. Calculate the ratio of hMSCs and hPCs in the pellet by
dividing through the total amount of relative DNA present
in the pellet.
4 Notes
References
biodegradable scaffolds for the repair of carti- 19. Levine B, Kroemer G (2008) Autophagy in the
lage in a rat osteochondral defect model. Bio- pathogenesis of disease. Cell 132:27–42
materials 35:7460–7469 20. Hara T, Nakamura K, Matsui M, Yamamoto A,
17. de Windt TS, Vonk LA, Slaper-Cortenbach IC, Nakahara Y, Suzuki-Migishima R et al (2006)
van den Broek MP, Nizak R, van Rijen MH Suppression of basal autophagy in neural cells
et al (2017) Allogeneic mesenchymal stem causes neurodegenerative disease in mice.
cells stimulate cartilage regeneration and are Nature 441:885
safe for single-stage cartilage repair in humans 21. Abramoff M, Magelhaes P, Ram S (2004)
upon mixture with recycled autologous chon- Image processing with ImageJ. Biophoton Int
drons. Stem Cells 35:256–264 11:8
18. Elmore S (2007) Apoptosis: a review of pro-
grammed cell death. Toxicol Pathol
35:495–516
Chapter 6
Abstract
Induced pluripotent stem cells (iPSCs) are generated from somatic cells that have been reprogrammed by
the ectopic expression of defined embryonic transcription factors. This technology has provided investiga-
tors with a powerful tool for modeling disease and developing treatments for human disorders. This chapter
will provide the researcher with some background on iPSCs and details on how to produce
MEF-conditioned medium, prepare mitotically arrested mouse embryonic fibroblasts (MEFs), create
iPSCs using viral vectors, passage iPSCs, and cryopreserve iPSCs. The methods offered here have been
used in many laboratories around the world and the reader can initially follow these methods. However, not
all cell types are easily transduced using viral vectors and other methods of delivering the reprogramming
transcription factors may need to be tested.
Key words Induced pluripotent stem cells, Pluripotency, Reprogramming, iPSC isolation, Mouse
embryonic fibroblasts, Cryopreservation, iPSC passaging
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_6, © Springer Science+Business Media, LLC, part of Springer Nature 2021
71
72 David R. Deyle
2 Materials
18. Polybrene.
19. Foamy viral reprogramming vector (ΔΦ53MOSKMETNW)
(see Note 3).
20. CytoTune™-iPS Reprogramming System (see Note 4).
(a) CytoTune® 2.0 KOS (human KLF4, OCT4, and SOX2).
(b) CytoTune® 2.0 hc-MYC (human cMYC).
(c) ©CytoTune® 2.0 hKLF4 (human KLF4).
21. Passage 2 (P2), DR4 mouse embryonic fibroblasts (MEFs) (see
Note 5).
22. DMEM low glucose.
23. L-Glutamine.
2.5 Working 1. bFGF solution: 10 μg bFGF (see Note 6), 1 mL PBS, and
Solutions 20 μL Knockout SR. Dispense into 8–125 μL aliquots and
store FGF solution and store at -20 C.
2. Human embryonic stem cell medium (HESCM): 500 mL
DMEM/F12 with GlutaMAX and sodium pyruvate, 100 mL
Knockout SR (final conc. ~17%), 6 mL Pen/Strep (final conc.
~1%), 6 mL NEAA (final conc. ~1%), bFGF (final conc. 2 ng/
mL) (see Note 6), 0.6 mL 0.1 M β-mercaptoethanol (0.35 mL
β-mercaptoethanol in 50 mL H2O). Pass medium through
0.22 μm filter into sterile bottle. The medium can be stored
at 4 C for 14 days. Before an experiment, warm the medium to
37 C.
3. Dispase: 100 mL PBS, 100 mg of dispase (final conc. 1 mg/
mL). Pass dispase through 0.22 μm filter into sterile bottle.
Dispase can be stored at 4 C for 7 days. Before an experiment,
warm dispase to room temperature.
4. MEF culture medium: 500 mL DMEM high glucose, 50 mL
FBS (final conc. ~10%), and 5 mL Pen/Strep (final conc. ~1%).
The medium can be stored at 4 C for 14 days. Before an
experiment, warm the medium to 37 C.
5. Mesenchymal Stem Cell Medium (MSCM): 500 mL DMEM
low glucose, 50 mL FBS (final conc. ~10%), 5 mL L-glutamine,
and 5 mL Pen/Strep (final conc. ~1%). The medium can be
stored at 4 C for 14 days. Before an experiment, warm the
medium to 37 C.
6. Gelatin stock: 100 gm of gelatin in 1 liter of double-distilled
water and autoclave. 0.5% gelatin stock solution can be stored
at room temperature.
7. Gelatin working solution: 100 mL of 0.5% gelatin stock solu-
tion plus 400 mL sterile water. 0.1% Gelatin working solution
can be stored at room temperature.
8. MEF freeze medium (FM): 50% MEF culture medium, 40%
FBS, and 10% DMSO.
Generation and Culture of Human iPSCs 75
3 Methods
17. The next day store cryogenic tubes in liquid nitrogen. Alter-
nately, MEF-conditioned medium can be produced immedi-
ately following the expansion of MEFs.
18. When cells are confluent, aspirate medium and wash 8 dishes
with 4 mL PBS. Save remaining dishes for generation of
MEF-conditioned medium.
19. Add 2 mL 0.05% trypsin to each dish and incubate at 37 C for
3–5 min.
20. Collect using MEF culture medium and pool the 8 dishes into
a 50 mL conical tube.
21. Centrifuge at 800 g for 5 min.
22. As cells are spinning, make FM.
23. Aspirate medium off cells and resuspend cells in 10 mL of FM.
24. Add 1 mL of cells to cryogenic tubes, place in Styrofoam
container, and freeze at 80 C.
25. The next day store cryogenic tubes in liquid nitrogen.
16. Add 9 mL MEF culture medium to each dish and gently rock
dish to disperse cells evenly.
17. Remaining 2 100 mm dishes aspirate medium and wash cells
with 4 mL PBS.
18. Add 2 mL 0.05% trypsin to each dish and incubate at 37 C for
3–5 min.
19. Collect using MEF culture medium and pool into a 50 mL
conical tube.
20. Centrifuge at 800 g for 5 min.
21. As cells are spinning, make fresh FM.
22. Aspirate medium off cells and resuspend cells in 10 mL of FM.
23. Place 1 mL of cell suspension in cryovial and put in room
temperature isopropanol freezing containers and place con-
tainers in 80 C overnight.
24. The next day store cryogenic tubes in liquid nitrogen.
25. Aspirate medium on cultured dishes and add 10 mL new MEF
culture medium every 3 days.
26. When cells are confluent, aspirate medium and wash cells with
4 mL PBS. (Collect 6 dishes at a time).
27. Add 2 mL 0.05% trypsin to each dish and incubate at 37 C for
3–5 min.
28. Use 8 mL of MEF culture medium to collect each dish and
pool 3 dishes into a 50 mL conical tube.
29. Use 10 mL MEF culture medium to rinse 3 100 mm dishes
and add to 50 mL conical tube.
30. Centrifuge both tubes at 800 g for 5 min.
31. Aspirate medium off 50 mL conical tubes and collect next
6 dishes.
32. Repeat steps 19 and 24 until all dishes have been harvested.
33. Resuspend both cell pellets in 10 mL MEF culture medium
and combine into one tube.
34. Count cells with hemocytometer and trypan blue.
35. Irradiate cells with 40 Gy (see Note 8).
36. Centrifuge at 800 g for 5 min and aspirate medium.
37. Prepare fresh FM.
38. Resuspend cells in freezing medium at 4 106 cells per mL (see
Notes 9 and 10).
39. Place 1 mL of cell suspension in cryovial and put in room
temperature isopropanol freezing containers and place con-
tainers in 80 C overnight.
40. The next day store cryogenic tubes in liquid nitrogen.
Generation and Culture of Human iPSCs 79
Table 1
Gelatin per culture dish size
3.4 Preparing 1. In sterile hood, add autoclaved 0.1% gelatin to tissue culture
Irradiated-MEF Plates dishes as indicated in Table 1.
for iPSCs 2. Place dishes containing gelatin in incubator at 37 C for a
minimum of 5 min or maximum of several hours.
3. Thaw MEFs (see Note 11).
4. Transfer cells from cryovial to 15 mL conical tube with 5 mL
pipet.
5. Add 4 mL MEF culture medium.
6. Centrifuge at 800 g for 5 min and aspirate medium.
7. Cells are resuspended as desired and plated (see Note 12).
8. Allow MEFs to attach for minimum 8 h, best to allow MEFs to
attach overnight.
9. Plated MEFs can be used for up to 7 days.
3.5 Generating iPSCs This protocol is based on two previously published papers for
from Fibroblasts/ lentiviral and foamy viral vector-mediated reprogramming [3, 7].
Mesenchymal Stem
1. Seed human fibroblasts/MSCs at 2 105 cells per well in
Cells Using Lenti and 6-well plate in MEF culture medium/MSCM.
Foamy Viral
2. Incubate at 37 C overnight.
Reprogramming
Vectors 3. Aspirate off medium and add 2 mL fresh medium to each well.
4. For lentiviral infections, add 4 μg/mL polybrene to each well
and infect each well with LV vectors at corresponding multi-
plicity of infection (MOI) (see Table 2) (see Note 13).
10. For excisable FV reprogramming vector transductions, infect
each well with foamy viral reprogramming vector (ΔΦ53M
OSKMETNW) at MOI 1.
5. Prepare 2 MEF-coated dishes (see Subheading 3.4) for each
infected well.
6. The next day, wash transduced wells with 2 mL PBS.
80 David R. Deyle
Table 2
Lentiviral reprogramming vectors
Fig. 1 Creating the cutting tool from a Pasteur Pipet. To create the cutting tool to pick iPSC colonies, heat
Pasteur pipette over Bunsen burner and make a 90 degree bend approximately 1/3 distance from the tip (a, b)
and allow to cool. Carefully holding the end, place pipette into flame again distal to the 90 degree bend, and as
the glass softens, quickly pull bend and end away from each other (c, d). Remove excess glass by tapping with
another pipet to generate cutting tool (e, f)
3.6 Generating iPSCs 1. Seed human fibroblasts/MSCs at 2 105 cells per well in
from Fibroblasts/ 6-well plate in MEF culture medium/MSCM.
Mesenchymal Stem 2. Incubate at 37 C overnight.
Cells Using
3. Aspirate off medium and add 2 mL fresh medium to each well.
Nonintegrating Sendai
Viral Reprogramming
4. Transduce cells using the CytoTune® 2.0 Sendai reprogram-
ming vectors corresponding multiplicity of infection (MOI)
Vectors
(see Table 3).
5. The next day, aspirate off medium and add 2 mL fresh MEF
culture medium/MSCM to each well.
6. Two days later, aspirate off medium and add 2 mL fresh MEF
culture medium/MSCM to each well.
82 David R. Deyle
Table 3
Sendai reprogramming vectors
7. Two days later, aspirate off medium and add 2 mL fresh MEF
culture medium/MSCM to each well.
8. Prepare 2 MEF-coated dishes (see Subheading 3.4) for each
infected well.
9. The next day, wash transduced wells with 2 mL PBS.
10. Add 0.5 mL 0.05% trypsin to each well and incubate at 37 C
for 3–5 min.
11. Collect each with 4.5 mL MEF culture medium/MSCM and
place in 50 mL conical tube.
12. Centrifuge at 800 g for 5 min.
13. Aspirate medium off conical tubes and MEF dishes blue.
14. Resuspend transduced cells in 20 mL of fresh MEF culture
medium/MSCM.
15. Count cells with hemocytometer and trypan.
16. Plate 50,000 viral transduced cells onto one 10 cm MEF dish
and 100,000 viral transduced cells on the other per each well
containing MEFs.
17. The next day change medium on all dishes to HESCM.
18. Change medium daily with HESCM for the next 8 days.
19. From days 18 to 21, scan dishes and monitor changing cell
morphology using inverted microscope.
20. Pick colonies as described in Subheading 3.5 (Steps 19–26).
3.8 Freezing iPSCs in This protocol is based on the published report by Ware et al. [8].
Straws
1. Prepare Bio-Cool III controlled-rate freezer by filling tank with
methanol, turning it On, setting the start temperature to –
10 C, setting the speed to 1 degree/minute, setting the end
temperature to –33 C, setting the hold minutes to 600 or
greater, and setting the alarm to 0 C. Press RUN to start
cooling to –10 C (see green light under “start”).
2. Put canes with buckets in freezer to cool.
3. Prepare fresh FMs (5 mL).
4. Collect 100 mm dish containing 70–80% confluent, healthy
appearing iPSCs (1 100 mm dish ¼ 5 straws).
5. Aspirate medium off iPSCs.
6. Add 4 mL of dispase to each dish.
7. Incubate at 37 C until colony edges start to curl (2–5 min).
8. Aspirate dispase and gently add 4 mL of wash medium to rinse
cells (see Note 18). Be careful not to dislodge colonies (see
Note 19).
9. Aspirate wash and add 4 mL of wash medium.
10. Detach colonies from dish using a 5 mL pipet. Try to keep the
colonies as large as possible.
11. Spin 200 g for 5 min at room temperature.
12. Aspirate gently. Pellets may be loose, so you can use a 5 mL
pipette here.
13. Resuspend cells in 1.1 mL FMs.
14. Attach straw adaptor to syringe.
15. Attach sterile, plugged straw to the adaptor.
16. Draw up about 1/2 inch FMs, being careful to avoid cells.
17. Draw up about 1/8–1/4 inch of an air bubble.
18. Draw up most of straw with cells, about 0.2 mL (mix cells first),
but leave 1/2 inch to go.
84 David R. Deyle
19. Pull straw out of medium, then draw up last 1/2 inch to plug.
Leave 1/2 inch air at bottom.
20. Place back in sterile package.
21. Repeat steps 17–21 with the other 4 straws.
22. Briefly flame bottom of straw and use gloved hand to press
melted straw together to seal.
23. Repeat step 23 on top of straw.
24. Return to sterile package and bring to Biocool III freezer.
25. The time between cell resuspension to placing straws contain-
ing iPSCs into Biocool III freezer should be at least 15 min.
26. Bring liquid N2 tank near Bio-Cool III controlled-rate freezer.
Make sure canes are ready with buckets.
27. Place straws in buckets at –10 C for about 1 min.
28. Cool long forceps in N2, then grab straw(s) with forceps above
air bubble, and allow ice crystal to form above the bubble.
Then place straws back in bucket in Biocool III freezer.
29. Once all straws set, press RUN again on Biocool III freezer to
start controlled-rate freezing.
30. When frozen, load straws into liquid N2 tanks.
3.9 Freezing iPSCs in 1. Thaw the required amount of mFreSR, store on ice.
Cryovials 2. Prepare FMc, store on ice.
3. Label cryovials.
4. Prechill isopropanol freezing containers at 4 C.
5. Collect 100 dish of iPSCs (1 100 mm dish ¼ 5–6 cryovials).
6. Aspirate medium from iPSCs.
7. Add 4 mL of dispase to each dish.
8. Incubate at 37 C until colony edges start to curl (2–5 min).
9. Aspirate dispase and gently add 4 mL of wash medium to rinse
cells (see Note 18). Be careful not to dislodge colonies (see
Note 19).
10. Aspirate and add 4 mL of wash medium.
11. Detach colonies from dish using a 5 mL pipet. Try to keep the
colonies as large as possible.
12. Spin 200 g for 5 min at room temperature.
13. Aspirate gently (pellets may be loose, so you can use a 5 mL
pipette here).
14. Gently resuspend the cells in 6 mL cold FMc. Take care to leave
the cell aggregates larger than would normally be done for
passaging.
15. Transfer 1 mL of cells to each labeled cryovial.
Generation and Culture of Human iPSCs 85
16. Place vials into an isopropanol freezing container and place the
container in -80 C freezer overnight.
17. The next day store cryogenic tubes in liquid nitrogen.
3.10 Thawing iPSCs 1. Fill 15 mL conical tube with 5 mL HESM or wash solution.
from Straw 2. Take straw from liquid N2 and place in a beaker of room
temperature water (dH2O, tap).
3. Wipe straw with 70% ethanol.
4. Cut bottom of straw with sterile scissors over waste bin.
5. Cut top of straw over 15 mL tube, while holding little bit
with plug.
6. Let cells drain into tube.
7. Spin 200 g for 5 min at room temperature, resuspend as
desired and plate onto MEF-coated dish.
3.11 Thawing iPSCs 1. Rapidly thaw cells in a 37 C water bath by gently shaking the
from Cryovial cryovial continuously until only a small frozen pellet remains.
2. Remove cryovial from the water bath and wipe with 70%
ethanol.
3. Transfer the cells to a 15 mL conical tube.
4. Add 1.5 mL of warm HESCM dropwise to the tube over
5 min, gently mixing as the medium is added.
5. Add an additional 1.5 mL of warm HESCM dropwise to the
tube over 1–2 min.
6. Spin 200 g for 5 min at room temperature, resuspend as
desired, and plate onto MEF-coated dish.
4 Notes
References
1. Takahashi K, Yamanaka S (2006) Induction of 8. Ware CB, Nelson AM, Blau CA (2005)
pluripotent stem cells from mouse embryonic Controlled-rate freezing of human ES cells.
and adult fibroblast cultures by defined factors. BioTechniques 38(6):879–880. 882-3
Cell 126(4):663–676 9. Watanabe K et al (2007) A ROCK inhibitor
2. Takahashi K et al (2007) Induction of pluripo- permits survival of dissociated human embry-
tent stem cells from adult human fibroblasts by onic stem cells. Nat Biotechnol 25(6):681–686
defined factors. Cell 131(5):861–872 10. Gharwan H et al (2007) Transduction of
3. Yu J et al (2007) Induced pluripotent stem cell human embryonic stem cells by foamy virus
lines derived from human somatic cells. Science vectors. Mol Ther 15(10):1827–1833
318(5858):1917–1920 11. Fusaki N et al (2009) Efficient induction of
4. Chan EM et al (2009) Live cell imaging distin- transgene-free human pluripotent stem cells
guishes bona fide human iPS cells from par- using a vector based on Sendai virus, an RNA
tially reprogrammed cells. Nat Biotechnol 27 virus that does not integrate into the host
(11):1033–1037 genome. Proc Jpn Acad Ser B Phys Biol Sci
5. Robinton DA, Daley GQ (2012) The promise 85(8):348–362
of induced pluripotent stem cells in research 12. Seki T et al (2010) Generation of induced plu-
and therapy. Nature 481(7381):295–305 ripotent stem cells from human terminally dif-
6. Xu C et al (2001) Feeder-free growth of undif- ferentiated circulating T cells. Cell Stem Cell 7
ferentiated human embryonic stem cells. Nat (1):11–14
Biotechnol 19(10):971–974 13. Hogan B (1994) Manipulating the mouse
7. Deyle DR et al (2012) Normal collagen and embryo: a laboratory manual. Cold Spring
bone production by gene-targeted human Harbor Laboratory Press, Plainview, NY
osteogenesis imperfecta iPSCs. Mol Ther 20
(1):204–213
Chapter 7
Abstract
Recent work emphasizes that bone comprises numerous mesenchymal cell types each with different
biologic functions, and deconvoluting the functions of these cells requires technical approaches with
single-cell resolution, such as single-cell RNA sequencing (scRNA-seq). A critical step in conducting a
successful single-cell sequencing study of bone is generation of a single-cell suspension of skeletal cells while
preserving cell viability. Here we describe a method to prepare single-cell suspensions from skeletal tissue in
preparation for single-cell sequencing studies. Also included are optional steps for fluorescence-activated
cell sorting (FACS) of skeletal cells, which allows subsequent scRNA-seq studies to be focused on specific
populations of interest.
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_7, © Springer Science+Business Media, LLC, part of Springer Nature 2021
89
90 Shawon Debnath and Matthew B. Greenblatt
2 Materials
3 Methods
Fig. 1 Digestion of mouse femurs. Isolated 15-day-old mouse femur suspended in media in a 12-well plate (a).
Mouse femurs placed on a glass slab to be chopped by razor blades (b). Texture of a thoroughly minced mouse
femur sample (c). Slurry texture of an appropriately digested mouse femur that can be easily pulled up by
pipetting (d)
Isolation of Skeletal Cells for scRNA-seq 93
Table 1
Suggested digestion conditions based on mouse age and specimen site
3.2 FACS Staining Skeletal mesenchymal cells can be subjected to single-cell sequenc-
ing with or without sorting. However, sorting provides efficient
elimination of contaminating cell population such as hematopoietic
and endothelial cells and additionally facilitates focus on specific
mesenchymal populations if desired. The following steps need to be
carried out on ice (see Note 10).
1. Add 200 μL of freshly prepared blocking buffer to the samples.
Suspend the pellet thoroughly by using 1 mL pipet. Do not
vortex. Incubate the tubes on ice for 30 min (see Note 11).
2. Add 200 μL of freshly prepared primary antibody solution to
the tubes. Suspend the pellet thoroughly by using 1 mL pipet.
Do not vortex. Incubate the tubes on ice for 45 min to 1 h.
Protect the tubes from light. After adding equal volumes of
blocking buffer and primary antibody solution, the final work-
ing solution of the blocking buffer is 1:100 and primary anti-
body solution is 1:200 (see Note 12).
Isolation of Skeletal Cells for scRNA-seq 95
Fig. 2 Schematic representation of isolation of mesenchymal cells by FACS. DAPI-negative live cells are
analyzed for cell size and singlets are isolated from the doublets. Mesenchymal cells (Ter119, CD45,
CD31) can be sorted by exclusion of mature red blood cells (Ter119+), hematopoietic cells (CD45+), and
endothelial cells (CD31+)
3.3 Sample During FACS or after bone tissue digestion without further FACS
Collection isolation, samples must be collected or suspended using a method
for Single-Cell that both preserves cell viability and is compatible with the down-
Sequencing stream single-cell sequencing methodology. The approach will vary
depending on whether a droplet based (e.g., Drop-seq, 10 Geno-
mics) or index sorting based (e.g., SMART-seq, MARS-seq,
CEL-seq) methodology is subsequently employed [8].
1. For droplet based assays using Drop Seq or 10 Genomics
platforms, mesenchymal cells are sorted in bulk to conduct
single-cell sequencing. Sorted mesenchymal cells should be
collected in collection media using 1.5 mL Eppendorf tubes.
Collection media should be free of serum, and basal media
containing 0.5% bovine serum albumin is suitable for many
applications. Sorting cells into PBS should be avoided, since
it will cause significant loss of sensitive cell types such as mes-
enchymal stem cells. We recommended to follow the guidelines
for the specific sequencing platform utilized.
2. Index-based techniques (such as CEL-seq, Smart-seq) require
sorting single cells into individual wells of 96-well or 384-well
plates. In many cases, the collection plate will be provided by
the sequencing facility and are pre-loaded with unique bar-
codes, primers, and lysis buffer. Following sorting, the plates
are sealed and snap frozen on dry ice before submitting the
samples for sequencing. Validation experiments must be con-
ducted to determine that sorting produces as few empty wells
as possible while avoiding doublets. This validation can take the
form of sorting a test cell line with direct microscopic visuali-
zation of cell capture after colonies are grown from the sorted
cells.
4 Notes
References
1. Chan CKF, Gulati GS, Sinha R et al (2018) 6. Debnath S, Yallowitz AR, McCormick J et al
Identification of the human skeletal stem cell. (2018) Discovery of a periosteal stem cell med-
Cell 175:43–56.e21 iating intramembranous bone formation.
2. Mizuhashi K, Nagata M, Matsushita Y et al Nature 562:133–139
(2019) Growth plate borderline chondrocytes 7. Machado L, Esteves de Lima J, Fabre O, et al
behave as transient mesenchymal precursor (2017) In situ fixation redefines quiescence and
cells. J Bone Miner Res 34(8):1387–1392 early activation of skeletal muscle stem cells. Cell
3. Chan CKF, Seo EY, Chen JY et al (2015) Iden- Rep 21(7):1982–1993
tification and specification of the mouse skeletal 8. Greenblatt MB, Ono N, Ayturk UM et al (2019)
stem cell. Cell 160:285–298 The unmixing problem: a guide to applying
4. Tikhonova AN, Dolgalev I, Hu H et al (2019) single-cell rna sequencing to bone. J Bone
The bone marrow microenvironment at single- Miner Res 34:1207–1219
cell resolution. Nature 569:222–228 9. Chang C-Y, Lee Y-H, Jiang-Shieh Y-F et al
5. Baryawno N, Przybylski D, Kowalczyk MS et al (2011) Novel distribution of cluster of differen-
(2019) A cellular taxonomy of the bone marrow tiation 200 adhesion molecule in glial cells of the
stroma in homeostasis and leukemia. Cell 177 peripheral nervous system of rats and its modu-
(7):1915–1932.e16 lation after nerve injury. Neuroscience
183:32–46
Chapter 8
Abstract
Cytosine modifications can alter the epigenetic landscape of a cell, affecting the binding of transcription
factors, chromatin organizing complexes, and ultimately affecting gene expression and cell fate.
5-Hydroxymethylcytosine (5hmC) modifications are generated by the Ten-eleven-translocation (TET)
family of enzymes, TET 1, 2, and 3, through the oxidation of methylated cytosines (5mC). The TET
family is capable of further oxidizing 5hmC to 5fC and 5caC, leading to eventual DNA demethylation.
However, 5hmC marks can also exist stably in DNA. Stable 5hmC is enriched in the gene bodies of
activated genes in multiple tissues, as well as associated with regulatory regions such as enhancers. Altera-
tions to 5hmC patterns have now been found in multiple diseases including osteoarthritis. Here, we
describe a method to map 5hmC modifications by next-generation sequencing using a technique based
on the selective modification and enrichment of the 5hmC mark. We additionally provide a bioinformatic
analysis pipeline to interpret the resulting data.
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_8, © Springer Science+Business Media, LLC, part of Springer Nature 2021
101
102 Fiorella Carla Grandi and Nidhi Bhutani
2 Materials
2.1 DNA Shearing 1. Qubit dsDNA BR Assay Kit (Invitrogen) and Qubit
and Glucosylation Fluorometer.
Reaction 2. Covaris S220 Ultrafocused sonicator.
3. Covaris milliTUBE 1 mL AFA Fiber.
4. DNA cleanup columns.
5. PCR Tubes.
6. Active Motif HydroxymethylCollector Kit (see Note 1) or:
(a) T4 Phage β-glucosyltransferase.
(b) UDP-Azide-Glucose.
(c) Biotin solution.
2.3 DNA Library 1. NEB Ultra II DNA Library Prep with Sample Purification
Preparation Beads (NEB #E7103).
and Next-Generation 2. NEBNext Multiplex Oligos for Illumina.
Sequencing
3 Methods
3.1 DNA Shearing 1. Extract DNA using standard methods for your skeletal tissue of
and Glucosylation choice. Quantify DNA (see Note 2).
Reaction 2. Shear DNA: Place 2 μg of DNA in a Covaris tube and shear
DNA to 200–300 base pair fragments. Conditions should be
optimized for each sample type based on concentration and
purity. Ideally, you want 2 μg of DNA in a volume of less than
35.5 μL (see Notes 3–5).
3. Set aside 10% of your fragmented input DNA as your input
control for sequencing. This will be used to calculate the back-
ground for your assay. Freeze this sample at 20 C until
Subheading 3.3.
4. Glycosylation reaction: set up the reactions in a 200 μL PCR
tube as in Table 1. Control reactions: A negative control should
be set up without the addition of the β-glucosyltransferase. In
the skeletal system, ATDC5 cells at day 15 of differentiation
can serve as a positive control, as these cells are known to have
high levels of 5hmC [5].
5. Incubate the reaction at 37 C for 1 h in a PCR thermocycler.
6. Spin down the PCR tube to collect all the volume.
7. Add 20 μL of biotin solution (Active Motif) to each reaction.
8. Incubate the reaction for 1 h at 37 C.
9. Purify biotinylated DNA using standard DNA cleanup column.
Make sure that you are using a column that is compatible with
your input DNA amount.
3.2 DNA Fragment 1. Resuspend the streptavidin beads by vortexing and allow to
Enrichment come to room temperature.
2. Set up each enrichment reaction in a 1.5 mL DNA LoBind tube
(Eppendorf) a 200 μL according to Table 2. Incubate the
reaction for 2–3 h at room temperature with end-to-end rota-
tion (see Note 6).
3. After incubation, spin down the tube to collect the volume at
the bottom and place the tube in a magnetic stand. Allow the
beads to pellet on the side of the tube. This may take 1–3 min.
Remove the supernatant (see Note 7).
104 Fiorella Carla Grandi and Nidhi Bhutani
Table 1
β-glucosyltransferase
Reagent Volume
Buffer 5 μL
10 mM DTT 5 μL
UDP-Azide-glucose 2.5 μL
B-Gluctranserase 2 μL
Sheared DNA ______μL
dH2O To volume
Total volume 50 μL
Table 2
DNA fragment enrichment
Reagents Volume
Streptavidin beads 25 μL
Binding buffer 25 μL
Purified biotinylation reaction 50 μL
Total volume 100 μL
3.4 Data Analysis 1. The quality of your fastq files can be analyzed using
(See Note 11) FASTQC [12].
2. Trim the low quality ends identified using FASTQC using
Trimmomatic [13]. After trimming, run the fastq files back
through FASTQC to ensure that all low quality reads have
been removed.
3. Map reads to the correct species genome using HISAT2
[14]. For mouse and human, we suggest using mm10 and
hg19, respectively, as they have the most updated additional
epigenetic information from the ENCODE and ROADMAP
projects [15–17].
4. 5hmC peaks can be called using MACS [18], with the input
sample as a control for the background to the biotin pull-
down.
5. Differentially hydroxymethylated regions (dHMRs) can be
called using diffREPS [19]. In order to do this, you must first
convert your bam files to bed files using BEDOPS [20]
bamtobed tool.
6. Annotation of peaks or dHMRs can be done using BEDOPS
closest-feature tool.
7. Motif analysis of the DNA nearby your 5hmC can be useful to
determine if specific transcription factors are either being tar-
geted or being used to guide the TET enzymes. This analysis
can be done using HOMER [21], which will predict both
known and de novo motifs.
8. Visualize your 5hmC reads on gene bodies or other genes of
interest using ngsplot [22].
9. Additional downstream analysis options (see Note 12):
(a) Intersect genes with dHMRs or peaks with genes differ-
entially expressed in RNA-sequencing studies.
106 Fiorella Carla Grandi and Nidhi Bhutani
4 Notes
Acknowledgments
References
1. Kohli RM, Zhang Y (2013) TET enzymes, 4. Ramos YFM, Meulenbelt I (2017) The role of
TDG and the dynamics of DNA demethyla- epigenetics in osteoarthritis: current perspec-
tion. Nature 502:472–479. https://doi.org/ tive. Curr Opin Rheumatol 29:119–129.
10.1038/nature12750 https://doi.org/10.1097/BOR.
2. Klein K, Gay S (2015) Epigenetics in rheuma- 0000000000000355
toid arthritis. Curr Opin Rheumatol 27:76–82. 5. Taylor SE, Li YH, Smeriglio P et al (2015)
https://doi.org/10.1097/BOR. Stable 5-hydroxymethylcytosine (5hmC)
0000000000000128 acquisition marks gene activation during chon-
3. Letarouilly J-G, Broux O, Clabaut A (2019) drogenic differentiation. J Bone Miner Res Off
New insights into the epigenetics of osteopo- J Am Soc Bone Miner Res 31(3):524–534.
rosis. Genomics 111:793–798. https://doi. https://doi.org/10.1002/jbmr.2711
org/10.1016/j.ygeno.2018.05.001 6. Taylor SEB, Li YH, Wong WH, Bhutani N
(2015) Genome-wide mapping of DNA
108 Fiorella Carla Grandi and Nidhi Bhutani
Abstract
Here we show how to measure the mobility of transcription factors using fluorescence recovery after
photobleaching (FRAP). Transcription factors are DNA-binding proteins that, upon binding to specific
DNA motifs, regulate transcription of their target genes. FRAP is a simple, fast, and cost-effective method,
and is a widely used quantitative method to measure the dynamics of fluorescently labeled molecules in
solution, membranes, and inside living cells. Dynamics, specified by the immobile fraction, recovery half-
time, diffusion constant, and ratio of molecules contributing to different phases of FRAP recovery, can be
quantified by FRAP. This can be useful to understand their function in gene regulation. This tutorial is
intended to familiarize the reader with the FRAP procedure to quantify transcription factor dynamics using
a standard confocal microscope and analysis using MATLAB (MathWorks®). This article will guide the
reader through the preconditions of FRAP, and a detailed and step-by-step procedure of preparing cells,
bleaching protocol, data analysis in MATLAB, and visualization of the FRAP data.
Key words Protein dynamics, Transcription factor activity, Fluorescence recovery, FRAP , SOX9,
mGFP, CLSM
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_9, © Springer Science+Business Media, LLC, part of Springer Nature 2021
109
110 Kannan Govindaraj and Janine N. Post
1.1 Different FRAP is a widely used method to study the protein dynamics in live
Methods for cells due to its simple setup in a laser confocal microscope. Other
Measuring Protein methods, such as FLAP (fluorescence loss after photoactivation),
Dynamics FLIP (fluorescence loss in photobleaching), FCS (fluorescence
correlation spectroscopy), and SMM (single molecule microscopy)
are also used to study protein kinetics [3]. These methods have
their own advantages and disadvantages, as described in Table 1.
1.2 When to Use FRAP can be used to study the signaling mechanism or to map the
FRAP factors that regulate a cellular process, such as the activity of tran-
scription factors. For this, cells must be transfected with a fluores-
cent fusion protein, thus the protein of interest is tagged with a
fluorescent protein.
The choice of transfection method depends on the cell type and
the protein of interest. It is recommended to make a stably trans-
fected cell line using, for example, CRISPR/Cas9 [9], if a cell line
(which properties are not affected by passage number) is used in the
study, or if overexpression of the protein of interest is toxic to the
cells. This helps to study the protein of interest at the native
expression levels and helps to avoid the transfection step every
time before FRAP. If using primary cells or cells which tend to
(de)differentiate over a number of passages, transient transfection
would be the best option. If the cells are easy to transfect, lipid
mediated transfection may be used. Retroviral or other viral-
Protocol for Quantifying Transcriptional Activity by FRAP 111
Table 1
Advantages and disadvantages of various methods used to study protein dynamics
1.3 FRAP Principle In FRAP a small region of interest (ROI, can be circular, square, or
rectangular) is photobleached with a high-intensity laser, and the
fluorescent recovery is monitored in that ROI in a given timescale
(from seconds to minutes). Post-bleach imaging duration and
frame-rate is determined based on the mobility of the molecule of
interest. The higher the mobility, the lower the imaging time, the
higher the frame rate. For example, if the mobility of a transcription
is relatively high, with a stable plateau reached within a minute,
fluorescence recovery can be recorded for 60 s, at a speed of
4 frames per second (fps). Before photobleaching, a number of
pre-bleach images (usually ~10) are recorded for normalization of
the fluorescence recovery. This process can be translated into a
graph.
As an example, we measure the fluorescence recovery of the
transcription factor SOX9, the master regulator for cartilage for-
mation [10], see Fig. 1. In this graph, one can observe that the
fluorescence in the bleach spot does not reach the level of the
original fluorescence. The plateau level is lower than the
pre-bleach intensity because some of the FRAP-bleached molecules
are immobile within the region of interest that is bleached. Because
of their immobility, they do not contribute to the recovery, while at
the same time occupying binding sites for incoming unbleached
proteins. We therefore name the fraction of molecules that contri-
butes to the recovery the “mobile fraction” and the one that does
not contribute the “immobile fraction.”
1.4 Mapping Signal For studies of immediate response of transcription factors to extra-
Transduction cellular signals, cells can be stimulated with cytokines (before/
Pathways Regulating during FRAP experiments) in an imaging buffer (see below). For
Transcription Factor example, we transfected the cells with a fluorescently tagged tran-
Mobility scription factor, SOX9-mGFP and treated cells with predicted ago-
nists/antagonists to study their transcriptional regulation
[8]. Stimulation time depends on the mechanism of action of
cytokine and needs to be determined empirically. If the cytokine
exerts fast cellular responses, as we have seen for the SOX9 cata-
bolic factors, such as WNT3a and IL1β, changes were observed
within 20 mins after stimulation (Fig. 2, [8]). However, a slightly
longer stimulation time (60 min) was required for the SOX9 ana-
bolic factor BMP7 (see Fig. 2).
1.5 Quantitation of Two major factors affect the mobility of a molecule in a biological
Protein Mobility system: diffusion and chemical interactions. The rate of diffusion is
determined by the diffusion constant (D) of the molecule, which is
dependent on the size of the molecule, the viscosity of the sur-
rounding medium, including physical structures that hinder diffu-
sion, and the temperature. In the cell, proteins interact with other
molecules, including other proteins and DNA/RNA. The binding
constants of the molecular interactions, binding (kon) and dissocia-
tion (koff) with other molecules, will affect mobility of the protein
of interest.
Fig. 2 FRAP curves showing SOX9-mGFP mobility changes in C20/A4 cells after
treatment with cytokines. Treatment of C20/A4 cells with WNT3a at 10 ng/mL
increased SOX9-mGFP mobility at 20 mins (yellow curve) as compared to the
control (blue curve). In contrast, BMP7 at 10 ng/mL did not change SOX9-mGFP
mobility significantly at 20 min (orange curve) as compared to the control.
However, 100 ng/mL of BMP7 with 1 h treatment decreased the SOX9-mGFP
mobility significantly (gray curve). This indicates that concentration of cytokines
and stimulation time varies among cytokines and should be determined
empirically
114 Kannan Govindaraj and Janine N. Post
1.6 Analyze FRAP Proper fitting and modeling of the FRAP curve, which reflects the
Data actual biological process, is important to extract useful dynamics
information from the FRAP data. Fitting and modeling should
consider the number of reactions contributing to the FRAP recov-
ery, the timescales, and whether these reactions occur in similar or
distinct timescales. For example, untagged mGFP (OriGene) pro-
tein does not bind to any intracellular target and recovers
completely and within seconds after photobleaching (see Fig. 3a).
FRAP data of mGFP can be easily fit with a single exponential fit (see
Fig. 3b, Eq. 1).
In contrast, SOX9-mGFP binds to DNA and the FRAP recov-
ery curve is the result of two reactions, namely fast (weakly bound
to DNA, nonspecific binding) and slow diffusion (strongly bound
to DNA, specific binding). FRAP data of a transcription factor like
SOX9 should be fit with two (or more) component fits (see Fig. 3c,
Eq. 5). The presence of shoulders in the curve can be an indicator
of the number of processes contributing to the FRAP recovery.
After bleaching, mGFP quickly recovers and attains a plateau, while
SOX9-mGFP slowly recovers and continues to recover (due to
exchange at the binding sites) throughout the recovery timescale,
with a visible shoulder in the recovery curve (see Fig. 3a, red curve).
In our model, we use a two-component fit (see Eq. 5) for
transcription factor dynamics, as two processes at distinct timescales
116 Kannan Govindaraj and Janine N. Post
Fig. 3 Fitting FRAP curves. (a) FRAP curves of mGFP (gray) and SOX9-mGFP (red). (b) The mGFP FRAP curve is
fit with a one-component fit (blue circles are data-points, red line is the fit) and (c) the SOX9-mGFP curve is fit
with a two-component fit (gray and green lines are first- and second-component fit, respectively)
1.7 Explanation of Immobile fraction (IF, Eq. 3) refers to the fraction fluorescent
FRAP Parameters protein bound to its target, e.g., transcription factor bound to
DNA. For transcription factors there are two populations, namely
fast (A1) and slow (A2) diffusing populations that contribute to
the fluorescence recovery (see Fig. 4). The fraction of protein that
are weakly bound to DNA constitute the fast diffusion population,
whereas the population bound to DNA constitutes the slow diffus-
ing population. The ratio A1/A2 refers to the increase or the
decrease of fluorescent proteins in the A1 compared to A2 fraction
(i.e., the ratio of the fast diffusing population to the slow diffusing
population).
If the fast and the slow diffusing populations are equally present
inside the nucleus, the value of the A1/A2 ratio will be 1. If the
Protocol for Quantifying Transcriptional Activity by FRAP 117
Fig. 4 Schematic representation of a FRAP recovery curve and explanation of parameters. If fast and slow
diffusion occur at two different timescales, the FRAP curve can be split into two phases, as shown. FI: initial
intensity, F0: intensity at time point t0 (first post-bleach intensity), FE: end value of the recovered intensity, t½:
halftime of recovery, immobile fraction (IF) is the population of SOX9-mGFP bound to DNA. A1 is the amplitude
of fast diffusing population of SOX9-mGFP, which is not bound to DNA and contributes to quick recovery. A2 is
the amplitude of slow diffusing population of SOX9-mGFP, which interacts on the various binding sites in
the DNA
value is less than 1, this indicates that the number of fast diffusing
proteins is less than that of the slow diffusing proteins. The diffu-
sion constant (D) is the rate at which a molecule diffuses in a
specific area. In a cellular milieu, the term “effective diffusion
(Deff)” indicates the recovery that mimics diffusion, but at a rate
that is slowed by binding interactions. This is calculated using
Eq. 4. The recovery time is the length of time after bleaching,
required for the fluorescence recovery to reach a constant value.
Recovery half-time (t½, Eq. 3) refers to half of recovery time.
Recovery half-time of A1 and A2 indicates half of the time
required for the recovery of corresponding phase of the FRAP
curve. For detailed information, the reader can refer to [8].
2 Materials
2.1 Materials for Cell 1. Silicone gasket with 1.5 mm thickness (cat#: 70465-2R2,
Culture and EMS, USA).
Transfection 2. Fibreless tissue paper (such as Kimberly-Clark® Professional).
3. 24-Well plates.
118 Kannan Govindaraj and Janine N. Post
2.2.2 Confocal Laser 1. Laser scanning confocal microscope (We used a Nikon A1
Scanning Microscope confocal microscope) with suitable laser lines for excitation of
the fluorescent protein.
2. Option to maintain the physiological temperature, CO2
control.
2.4 Materials for 1. Data analysis and graphing software, we used Origin Pro
Data Visualization and (OriginLab®).
Statistical Analysis
3 Methods
3.1 Cell Culture and To study the cellular physiology, cells ideally should be maintained
Transfection in their native state. For example, primary chondrocytes should
maintain their chondrogenic potential during the experiments. To
3.1.1 Cell Culture
prevent dedifferentiation, expand the primary chondrocytes in phy-
sioxical (2.5–5% O2) conditions. Use low passage numbers (<4)
and low population doubling levels (PDL < 5). Culture the cells
without antibiotics especially during and after transfection. We
usually do not serum-starve the cells before FRAP experiments,
unless we stimulate with growth factors.
Culturing with serum or serum-starvation: Depending on the
nature of the study, cells can be cultured with or without serum. If
the dynamics of protein under study is influenced by cell cycle, it
would be appropriate to synchronize the cell cycle by serum starva-
tion. Also, if the study is intended to investigate the long-time
effect of an external growth factor/cytokine cells may need to be
cultured without or with inactivated serum to avoid interference
with the growth factors present in the serum.
3.1.4 Cell Stimulation Cell stimulation time and concentration of cytokines/growth fac-
tors are other important factors to be considered. Stimulation time
depends on the speed of cellular response, the concentration
depends on the efficacy of the cytokine/growth factor, and both
needed to be determined empirically. We stimulated C20/A4 cells
[20] with different concentrations (1 ng/mL, 10 ng/mL and
200 ng/mL) of WNT3a and found that 10 ng/mL was the mini-
mal concentration needed to destabilize SOX9-mGFP from DNA
[8]. In contrast, BMP7 promoted DNA binding of SOX9-mGFP at
100 ng/mL concentration and with a longer incubation time (see
Fig. 2).
3.1.5 Preparing Cells for Day 1: Seed the cells on a sterile, microscopic glass coverslip (12 or
FRAP 13 mm Ø) placed in a 24-well plate (see Note 3).
Day 2: Transfect the cells with the plasmid encoding the fluo-
rescent protein of interest.
Day 3: Perform FRAP using the protocol below. For FRAP
experiments, transfected cells grown on coverslip (on day 2) need
to be prepared in the following way (see Fig. 5).
1. Mount a silicone well divider (Electron Microscopy Sciences,
USA, with right well size) on to a microscopic glass slide (see
Fig. 5a, Note 4).
2. Fill the silicone well with imaging buffer (see Fig. 5b).
3. Place the coverslip in the middle of silicone well filled with
imaging buffer, with the transfected cells facing the imaging
buffer (see Fig. 5c, Note 5).
4. Excess water will be squeezed out and the coverslip will nicely
mount on to the silicone divider. Make sure that no air bubble
is introduced in the buffer. If there is an air bubble, carefully
remove the coverslip with the tweezer and start from step 1.
Once the coverslip with transfected cells is successfully
mounted on to the silicone well, it is ready for FRAP. For this
follow the FRAP protocol in Subheading 3.2.
3.2 FRAP There is no standardized universal protocol for FRAP since the
experimental design depends on bleaching and recovery character-
istics of the fluorescent molecule under study. However, all FRAP
experiments contain three phases, i.e., (1) pre-bleach image acqui-
sition, (2) photobleaching, and (3) post-bleach image acquisition.
Protocol for Quantifying Transcriptional Activity by FRAP 121
Fig. 5 Preparation of the cells grown on coverslip for FRAP. (a) The silicone divider is mounted on to a
microscopic glass slide. (b) Imaging buffer is filled in the silicone well. (c) Transfected cells grown on a
coverslip is placed on the imaging buffer (cells are facing the imaging buffer)
Fig. 6 Optimizing bleaching laser intensity. Bleaching efficiency of SOX9-mGFP in a fixed cell at different
percentages and μW (at the objective) of laser power as indicated in the image. Bleaching is inefficient at (a)
10% (6.8 μW) and (b) 25% (17.1 μW) laser power, whereas (c) 50% (34.3 μW) and (d) 100% (69.0 μW)
efficiently bleach the SOX9-mGFP and the bleach profile is shown in (e)
expressing the same fluorescent protein (see Fig. 6). The fol-
lowing guidelines can be helpful:
(a) Laser power: Usually, 25–100% laser power is used for
photobleaching. Higher laser power enables faster bleach-
ing, but is toxic to the cells [21]. It should be optimized to
the minimum laser power needed to achieve complete
photobleaching of the ROI (see Fig. 6).
(b) Size of the ROI: The size of the ROI is usually less than
10% of the total fluorescence distribution area. It also
depends on the kinetics of the fluorescent molecule. A
size of 1–2 μm (Ø) ROIs with higher magnification objec-
tives (60/100) give a good signal-to-noise ratio
(SNR). Smaller ROIs will increase the noise.
(c) Shape of the ROI: The ROI can be circular or a rectangle
or a square. Again, this depends on the shape/distribution
of the fluorescent protein. We use a circular ROI for
nuclear proteins, such as transcription factors. However,
the calculation formulae may need to be adjusted accord-
ing to the shape of the ROI [22].
Protocol for Quantifying Transcriptional Activity by FRAP 123
3.2.1 Optimal FRAP The optimal FRAP parameters for our study of SOX9-mGFP
Parameters dynamics in the nucleus are given below as an example. Although
we used a Nikon A1, the settings should be easily transferable to a
confocal microscope of a different manufacturer.
1. Pre-bleach images: 25 images.
2. Bleaching: 1 iteration of high-intensity laser pulse (50%,
488 nm laser).
3. Post-bleach images: 240 images (60 s, imaging time).
4. Objective: 60 (water immersion), 1.4 NA. Alternatively, a
silicone oil immersion objective can be used, as long as the
refractive index of the objective is close to that of the cell to
avoid aberrations.
5. Scan mode: Unidirectional.
6. Frame rate: 4 frames/s(fps, for both pre- and post-bleach
imaging). The frame rate can also be set based on the pixel
dwell time.
7. Frame size: 256 256 pixels.
8. Averaging: Normal (No averaging!).
9. Pinhole: 1.2 AU (Z-step size: 0.25 μm and Optical sectioning:
0.77 μm).
10. HV (Gain of the detector): ~80–90% (depends on the intensity
of the fluorescent protein).
11. Offset (of the detector): 1.
12. Laser power: 0.35% (0.12 μW at the objective).
13. Zoom: 7.09.
14. Pixel size: 0.12 μm.
15. Laser: 488 nm.
16. Bleaching: 50% (34.3 μW, at the objective) laser power, 16 fps
(the highest speed possible in the Nikon A1).
17. Export the fluorescent intensity values to a Microsoft Excel
document using the “Export” option and save it in a folder.
This Excel document will be used for data analysis in
MATLAB.
Protocol for Quantifying Transcriptional Activity by FRAP 125
18. Repeat the FRAP experiment for at least 50 cells per condition
and collect FRAP image files and Excel documents containing
FRAP data.
Fig. 7 “A1plus Compact GUI” provides options to control image acquisition. (a) While “Eye port” is selected
(red box 2), interlock is automatically activated (red box 3) and the rest of the options are automatically
disabled. Samples can be viewed through the binocular. (b) While “Scan” mode is on (red box 1), “Eye port” is
not selected (red box 2), “Interlock” is removed (red box 3), “Unidirectional” scan mode is selected (red box 4),
“Scan speed” is set to 4 fps, “image size” is set to 256 256 pixels (red box 5), “Averaging” is set to normal
(red box 6). “Pinhole” is set to 1.2 AU (red box 7). Needed laser line can be activated by the “laser settings”
option (red box 8), “HV” is set to 88, “Offset” is set to 1, “488 nm” laser line is selected and “laser power” is
set to 0.35% (red box 9)
15. If the nucleus does not appear in the window, the view area
(step 6) or the scan area (step 9) might not be centered
properly.
16. Another way to center the desired nucleus (expressing SOX9-
mGFP) is to right click on the scan area (green square) and
select the option “Scan full field of view” in the “A1plus Scan
Area” window. The image acquisition parameters, such as
frame rate, frame size, and zoom, will change. Again, set the
zoom to 7.09 and scan area (red square) will appear again.
Drag and position it around the nucleus of interest and right
click on the red square to confirm the position (the red square
will turn green). Adjust the image acquisition parameters to
optimal conditions and click “Scan” button in the “A1plus
Compact GUI” window and return to step 13.
17. Draw three ROIs in the frozen window. To draw ROIs, right
click and drag inside the frozen window. Options to choose the
ROI shape will appear on the frozen window and choose
“Circular ROI” option. The mouse pointer will turn into a
“+” sign; click and drag on the frozen window, a red circular
ROI will be drawn.
18. Right click on the “red ROI” and click “ROI properties”
option. “ROI properties” window will appear and set the ROI
dimensions. After setting the values, close the “ROI properties”
window.
19. To draw two more ROIs, either repeat steps 17 and 18 or
simply right click on the current “red ROI” and select “Dupli-
cate Selected ROIs” option.
20. Assign a function to each ROI as mentioned in the order
below. Click on a ROI to select it, to assign a function, right
click on it, a window will appear, and select the function to be
assigned. Keep the below order throughout the FRAP mea-
surements, to get uniform datasets.
(a) “Stimulation ROI” (red, for FRAP measurements).
(b) “Reference ROI” (green, to measure the acquisition
photobleaching).
(c) “Background ROI” (blue, to measure the background
signal).
21. Once functions are assigned, ROIs in the frozen image will be
labeled accordingly and position the ROIs as per following
guidelines. Position the “Stimulation ROI S1:1” (red) in a
representative region close to the center of the nucleus. Posi-
tion the “Reference ROI R:2” (green) as much as away from
the stimulation ROI (but the fluorescent intensity levels should
be identical to stimulation ROI). Position the “Background
ROI B:3” at the outside of the cell.
Protocol for Quantifying Transcriptional Activity by FRAP 129
Fig. 8 ND Stimulation settings. (a) FRAP duration and file saving options are set in the “ND stimulation”
window. Check the box next to “Save to File” (red box 1). Select the file saving path and write the file name
(red box 2and 3). A FRAP experiment has three phases: (1) pre-bleach, (2) photobleaching (Stimulation), and
(3) post-bleach image acquisition. Add those phases using “Add” button (red box 4). Three phases are shown
(red box 5). Set the parameters for each phase as follows: Phase #1: Select Acquisition, Set the number of
“Loops” to 25 (25 pre-bleach images), Phase #2: Select Stimulation, Select “S1” ROI, Set the number of
“Loops” to 1 (1 iteration of high-intensity laser), Phase #3: Select Acquisition, Set the number of “Loops” to
240 (240 post-bleach images). Set the interval to “No delay” for all three phases to acquire images
continuously. Once number of loops were set, the duration will be set automatically. Check the box next to
“Perform Time Measurement” (red box 6). Select “Apply Stimulation Settings” (red box 7) and start the FRAP
experiment by clicking ‘Run now’ button (red box 8)
130 Kannan Govindaraj and Janine N. Post
3.3 Data Analysis It is necessary to check the FRAP data for “XY drift” before data
analysis. Even a small drift can significantly alter the FRAP dynam-
3.3.1 FRAP Data
ics data.
Validation Before Analysis
1. “XY drift” can be corrected by image registration using Fiji or
ImageJ. We used Fiji as it contains all plugins needed to read
Nikon’s “.nd2” file format. However, image registration plu-
gins need to be installed manually in Fiji. We used the “Tem-
plate Matching” plugin for image registration. This can be
downloaded from this link [23, 24]. Follow the instructions
given in the site to install the plugin in Fiji and for image
registration.
2. After image registration, play the FRAP image stack in Fiji to
check for the proper alignment. Still, if the alignment is not
proper, discard this FRAP data.
3. If the alignment is proper, go to the first post-bleach image
(26th image in the stack), draw a “circular ROI” on the bleach
area using Fiji “Circle” tool.
4. Select “Plot Z-axis Profile” option in Fiji (Image ! Stacks !
Plot Z-axis Profile). A new window showing a fluorescent
intensity graph will appear. Extract fluorescent intensity values
from this window and replace the respective column in the
Excel file containing FRAP data.
5. Draw another “Circular ROI” to collect reference and back-
ground fluorescent intensities as well and follow the previous
step (see Note 8).
3.3.2 FRAP Data There are many MATLAB scripts and free software, such as easy-
Analysis in MATLAB FRAP [25] and ImageJ plugins [26], available for FRAP data
analysis. However, we used our own MATLAB script for FRAP
analysis, and this session will guide the user through our script. The
MATLAB script and sample data sets are available on request. Only
the “Microsoft Excel” documents containing FRAP data are
needed for data analysis.
1. Download the “Data Analysis” folder from the above link and
the folder contains necessary MATLAB scripts for FRAP data
analysis and sample data sets.
(a) It has two folders: “Data Files” and “Effective Radius.”
(b) It has two MATLAB scripts FRAPAnalysis and findHT.
2. Place both the MATLAB scripts in the MATLAB home direc-
tory folder.
3. Open the MATLAB and “FRAPAnalysis” script in the
MATLAB.
4. The effective radius of the bleach spot should be determined
before the FRAP analysis. Due to the point spread function of
132 Kannan Govindaraj and Janine N. Post
Fig. 9 Using the bleach profile to calculate the effective radius. (a) “Straight line” tool is selected in Fiji (red
box 1). A straight line is drawn across the bleach ROI of the first post-bleach image (red box 2). (b) Profile of
the first post-bleach image. (c) Profile of the last pre-bleach image. Click on the “More>>” (red box 3) button
and select “Copy All Data” option. Paste it in a separate “.txt” file. (d) Averaged bleach profile and the
calculation of effective radius. Effective radius can be calculated from the center axis (red box 4) and using the
“μm” scale of the “X”-axis. Effective radius of this bleach profile is 2.2 μm
the light, in practice, the size of the bleached ROI will be always
higher than the actual size of the ROI. So, finding the actual
bleaching size and effective radius is necessary to calculate
precise protein dynamics. To know more about effective radius
and nominal radius, refer to [27]. The effective radius of the
bleach spot can be calculated as described below.
(a) Import the “FRAP image stack” to Fiji.
(b) Go to the first post-bleach image (26th image) and draw a
straight line across the bleach spot as shown in the Fig. 9a.
Center of the straight line should be at the center of the
bleach spot.
(c) Click on “Plot Profile” option (Analyze ! Plot Profile) to
get the fluorescent intensity profile along the line.
(d) A new window as shown in Fig. 9 (b or c) will appear.
Copy the fluorescent intensity profile data and paste it in a
new “.txt” file.
(e) Scroll back to the previous image (25th image or the last
image of the pre-bleach series). The “straight line” will
stay on the image stack. Repeat the steps (iii and iv).
(f) MATLAB script recognizes pre-bleach and post-bleach
line intensity profiles by the “file name” of the text files
and they should be named accordingly. The text file con-
taining post-bleach and pre-bleach fluorescent intensity
profile should be suffixed with “ps” and “pe,” respectively.
Protocol for Quantifying Transcriptional Activity by FRAP 133
Fig. 10 Variables to be entered in the “FRAPAnalysis” Matlab script. FRAP data files folder location is
mentioned the “datapath” (red box 1). The frame number of the first post-bleach image and the frame
number at which photobleach correction should be started are mentioned in line 20 and 21, respectively (red
box 2). The diameter of the bleach ROI is 25 pixels and the pixel size is 0.12 (red box 3). The effective radius is
2.2 as calculated in step 5 (red box 4)
134 Kannan Govindaraj and Janine N. Post
Table 2
Abbreviations in the “FitResults” Microsoft Excel sheet are explained (scripts available upon request)
3.4 Data FRAP measures a variety of dynamics data at the single cell level
Visualization and resolution. Proper graphical representation of the FRAP data
Statistics would effectively communicate underlying biological information,
such as different aspects of protein dynamics and presence of differ-
3.4.1 Data Visualization ent cell populations, etc. We use Origin® (93E) software (Origi-
nLab, Northampton, Massachusetts, USA) for making boxplots,
interval plots, scatterplots, and statistical analysis. Fluorescence
intensity values are plotted against the function of time in the
X-axis and shown in the scatterplot (see Fig. 11a). Interval plots
or boxplot with individual data points is a good choice to visualize
protein dynamics data (see Fig. 11b, c).
3.4.2 Statistics Our data does not show a normal distribution because of the
cellular heterogeneity. We therefore used a Mann-Whitney U-test
to calculate statistical significance.
136 Kannan Govindaraj and Janine N. Post
Fig. 11 Visualizing protein dynamics data as measured by FRAP. (a) Scatter plot showing fluorescence
recovery curves. (b) Boxplot and (c) interval plot showing immobile fraction data
4 Notes
Acknowledgements
References
(ed) Biophysical aspects of transmembrane sig- reduce cell contraction levels. Lab Chip 11
naling. Springer, Berlin, Heidelberg, pp 33–70. (13):2231–2240. https://doi.org/10.1039/
https://doi.org/10.1007/3-540-26511-2_2 C0LC00641F
15. Zacharias DA, Violin JD, Newton AC, Tsien 25. Koulouras G, Panagopoulos A, Rapsomaniki
RY (2002) Partitioning of lipid-modified MA, Giakoumakis NN, Taraviras S, Lygerou
monomeric gfps into membrane microdomains Z (2018) EasyFRAP-web: a web-based tool
of live cells. Science 296(5569):913. https:// for the analysis of fluorescence recovery after
doi.org/10.1126/science.1068539 photobleaching data. Nucleic Acids Res 46
16. Shaner NC, Steinbach PA, Tsien RY (2005) A (W1):W467–W472. https://doi.org/10.
guide to choosing fluorescent proteins. Nat 1093/nar/gky508
Methods 2:905. https://doi.org/10.1038/ 26. Blumenthal D, Goldstien L, Edidin M, Gheber
nmeth819 LA (2015) Universal approach to FRAP analy-
17. Cardarelli F, Tosti L, Serresi M, Beltram F, sis of arbitrary bleaching patterns. Sci Rep
Bizzarri R (2012) Fluorescent recovery after 5:11655. https://doi.org/10.1038/
photobleaching (FRAP) analysis of nuclear srep11655
export rates identifies intrinsic features of 27. Day CA, Kraft LJ, Kang M, Kenworthy AK
nucleocytoplasmic transport. J Biol Chem 287 (2012) Analysis of protein and lipid dynamics
(8):5554–5561. https://doi.org/10.1074/ using confocal fluorescence recovery after
jbc.M111.304899 photobleaching (FRAP). Curr Prot Cytom C
18. Wu B, Piatkevich KD, Lionnet T, Singer RH, HAPTER:Unit2.19-Unit12.19. https://doi.
Verkhusha VV (2011) Modern fluorescent pro- org/10.1002/0471142956.cy0219s62
teins and imaging technologies to study gene 28. Lidke DS, Nagy P, Heintzmann R, Arndt-Jovin
expression, nuclear localization, and dynamics. DJ, Post JN, Grecco HE, Jares-Erijman EA,
Curr Opin Cell Biol 23(3):310–317. https:// Jovin TM (2004) Quantum dot ligands pro-
doi.org/10.1016/j.ceb.2010.12.004 vide new insights into erbB/HER receptor–-
19. Nissim-Rafinia M, Meshorer E (2011) Photo- mediated signal transduction. Nat Biotechnol
bleaching assays (FRAP & FLIP) to measure 22:198. https://doi.org/10.1038/nbt929
chromatin protein dynamics in living embry- 29. Mueller F, Mazza D, Stasevich TJ, McNally JG
onic stem cells. J Vis Exp 52:2696. https:// (2010) FRAP and kinetic modeling in the anal-
doi.org/10.3791/2696 ysis of nuclear protein dynamics: what do we
20. Goldring MB, Birkhead JR, Suen LF, Yamin R, really know? Curr Opin Cell Biol 22
Mizuno S, Glowacki J, Arbiser JL, Apperley JF (3):403–411. https://doi.org/10.1016/j.
(1994) Interleukin-1 beta-modulated gene ceb.2010.03.002
expression in immortalized human chondro- 30. Beaudouin J, Mora-Bermúdez F, Klee T,
cytes. J Clin Invest 94(6):2307–2316. Daigle N, Ellenberg J (2006) Dissecting the
https://doi.org/10.1172/JCI117595 contribution of diffusion and interactions to
21. Post JN, Lidke KA, Rieger B, Arndt-Jovin DJ the mobility of nuclear proteins. Biophys J 90
(2005) One- and two-photon photoactivation (6):1878–1879. https://doi.org/10.1529/
of a paGFP-fusion protein in live Drosophila biophysj.105.071241
embryos. FEBS Lett 579(2):325–330. 31. Fritzsche M, Lewalle A, Duke T, Kruse K,
https://doi.org/10.1016/j.febslet.2004.11. Charras G (2013) Analysis of turnover dynam-
092 ics of the submembranous actin cortex. Mol
22. Siggia ED, Lippincott-Schwartz J, Bekiranov S Biol Cell 24(6):757–767. https://doi.org/10.
(2000) Diffusion in inhomogeneous media: 1091/mbc.E12-06-0485
theory and simulations applied to whole cell 32. Oh D, Zidovska A, Xu Y, Needleman Daniel J
photobleach recovery. Biophys J 79 (2011) Development of time-integrated multi-
(4):1761–1770. https://doi.org/10.1016/ point moment analysis for spatially resolved
S0006-3495(00)76428-9 fluctuation spectroscopy with high time resolu-
23. Tseng Q (2011) Template matching and slice tion. Biophys J 101(6):1546–1554. https://
alignment - ImageJ plugins. https:// doi.org/10.1016/j.bpj.2011.08.013
sitesgooglecom/site/qingzongtseng/tem 33. Kim SA, Heinze KG, Schwille P (2007) Fluo-
plate-matching-ij-plugin. Accessed Jan 2020 rescence correlation spectroscopy in living
24. Tseng Q, Wang I, Duchemin-Pelletier E, cells. Nat Methods 4:963. https://doi.org/
Azioune A, Carpi N, Gao J, Filhol O, Piel M, 10.1038/nmeth1104
Théry M, Balland M (2011) A new micropat- 34. Herrick-Davis K, Grinde E, Cowan A, Mazur-
terning method of soft substrates reveals that kiewicz JE (2013) Fluorescence correlation
different tumorigenic signals can promote or spectroscopy analysis of serotonin, adrenergic,
Protocol for Quantifying Transcriptional Activity by FRAP 139
muscarinic, and dopamine receptor dimeriza- cycle and G-actin regulation of polymer resto-
tion: the oligomer number puzzle. Mol Phar- ration. Proc Jpn Acad Ser B Phys Biol Sci 86
macol 84(4):630–642. https://doi.org/10. (1):62–83. https://doi.org/10.2183/pjab.
1124/mol.113.087072 86.62
35. Mendoza MC, Besson S, Danuser G (2012) 37. Watanabe N, Mitchison TJ (2002) Single-
Quantitative fluorescent speckle microscopy molecule speckle analysis of actin filament turn-
(QFSM) to measure actin dynamics. Curr Pro- over in lamellipodia. Science 295
toc Cytom Chapter 2:Unit2.18. https://doi. (5557):1083–1086. https://doi.org/10.
org/10.1002/0471142956.cy0218s62 1126/science.1067470
36. Watanabe N (2010) Inside view of cell locomo-
tion through single-molecule: fast F/G-actin
Chapter 10
Abstract
Computational modeling of biological networks is increasing in popularity due to the increased demand for
understanding biological processes. This understanding requires integration of a variety of experimental
data that allows understanding of complex mechanisms regulating cell and tissue function. However, the
mathematical complexity of many modeling tools have thusfar prevented broad adaptation and effective use
by molecular biologists. In this chapter, we show by example how one can start building a model in
ANIMO and how to adapt the model to experimental data. We show how this model can be used for
simulating network activities, testing hypotheses, and how to improve the model using wet-lab data.
Key words Modeling tool, Computational model, Signal transduction network, Signaling pathways,
Signaling cross talk, WNT, NFkB, TGF beta
1 Introduction
1.1 The Need of Over the past decades, advancement in high-throughput technol-
Modeling for Biological ogies have led to a rapid accumulation of genomic and proteomic
Networks data. It has provided a great wealth of information by studying the
gene regulation and gene-protein interactions in health, as well as
in pathological states. However, not all data are utilized in an
efficient manner, and changes in cellular signaling are often incom-
pletely explored. This necessitates a shift from the central dogma of
“hypothesis to experiment to models” toward “big data to models
to hypothesis to new experiment to additional data and more
inclusive models” [1]. One effective way of utilizing information
obtained from genomic and proteomic data is by building compu-
tational models. These models can be an option for storing large
and complex data as well as for hypothesis generation.
Models can be used as a tool to combine data from existing
literature. For example, data obtained on a protein studied in
context of one disease can be used to build a model. This model
can then be used to predict the mechanism of this protein in the
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_10, © Springer Science+Business Media, LLC, part of Springer Nature 2021
141
142 Sakshi Khurana et al.
1.2 Building a Model Computational models can graphically represent biological net-
works based on information obtained from literature, in combina-
tion with a formalization of the network in terms of activity. In
contrast to small networks, for large and complex networks it is
impossible to add all the components of the network. For complex
networks, it would be ideal to include at least the most important
molecules. However, it is not always easy to figure out which
molecules play key roles in the functioning of the network. Many
biological events are, in effect, changes in activity, comparable to
switches / dimmers. For example, changes in concentration, post-
translational modifications (i.e., phosphorylation, methylation) or
localization of a protein, or changes in gene expression are causal
Modeling with ANIMO 143
1.4 Comparison of Various computational tools are available for building models that
some Modeling define the dynamics of biological networks. In this part, we describe
Formalisms and Tools the formalism that is underlying some of the most-used modeling
tools. Tools based on formalisms described by Boolean and fuzzy
logic models are often used. Tools that belong to this category,
such as CytoCopteR [12], GINsim [13], and BooleSim [14], are
based mainly on discrete transitions and can be used to model
interactions among a large variety of proteins. These tools make it
easier to perform model building and model validation, and can be
used for predictions. On the other side of the tool-spectrum,
ordinary differential equations (ODEs) make a purely continuous
model and can efficiently represent biochemical reactions for large
and complex networks where mass-action approximations are
appropriate [10, 11, 15]. Examples of tools using ODEs are
Odefy [16], COPASI [17], and CellDesigner [18].
An actively growing field for computing graphical representa-
tions of biological networks through literature or high-throughput
data has been gaining interest in the last few years [10]. In this field,
biological networks are typically represented with the help of
graphs, where each component, e.g., gene or protein, is repre-
sented by nodes and potential interactions by edges. Edges can
typically represent a wide range of stimulatory or inhibitory inter-
action modes from direct physical binding to correlated gene
expression or phosphorylation by kinase, etc. For the purpose of
making computational modeling of dynamic biological networks
available to researchers without extensive mathematical training, we
previously developed ANIMO: Analysis of Networks with Interac-
tive Modeling [8, 11, 19].
ANIMO is a tool that helps a biological researcher to build and
analyze a protein interaction network, providing a visual represen-
tation of the activity profile over time. In order to formalize and
correctly analyze the models, ANIMO networks are based on a
Timed Automata model, that can be described as piecewise linear
approximation of ODE models. More information about ANIMO
and the comparison with other tools, as well as example signal
transduction models, can be found in Schivo et al. [11].
In this chapter, we want to give a working example of how to
build a computational model. In this model one can add data and
test hypotheses. Wet-lab experiments can be used to validate and
improve the model, as also described in this chapter.
Modeling with ANIMO 145
2 Materials
2.1 Computational To run ANIMO, a desktop or laptop computer is needed with the
Materials following software installed:
1. Java (java.com).
2. UPPAAL (www.uppaal.org) (see Note 1).
3. Cytoscape (www.cytoscape.org) (see Note 2).
An extensive user manual for installation and for creating a
network can be found here: http://fmt.cs.utwente.nl/tools/
animo/
The software to run Java-based programs is provided for free by
Oracle. More information is available on the java.com website.
Cytoscape is an open source project released under the terms of
the GNU Lesser General Public License. UPPAAL is developed by
a collaboration by the universities of Uppsala (Sweden) and Aal-
borg (Denmark), and is free for noncommercial applications in
academia only. All software works under Windows, Mac, and the
most common GNU/Linux OS (see Note 3).
Models used in this chapter, as well as raw data files, can be
found online via:
https://www.utwente.nl/en/tnw/dbe/research/additional-
materials/
2.2 Wet-Lab 1. 96-Well plates (BD biosciences, with white walls and a clear
Materials bottom such that they are suitable for cell culture and allow for
luciferase activity measurements).
2. Cells that are being studied (C2C12, mouse myoblasts, ATCC
®
CRL-1772™).
3. Cell culture media (DMEM), Thermo Fisher.
4. Fetal bovine serum (FBS), Thermo Fisher.
5. Penicillin-Streptomycin (Pen/Strep), Thermo Fisher.
6. Phosphate-buffered saline (PBS), Thermo Fisher.
7. Trypsin, Thermo Fisher.
8. Lipofectamine™ 2000 Transfection Reagent, Thermo Fisher
Scientific.
9. Reporter DNA (16xTOPFlash (TCF Reporter Plasmid
21-170, Sigma Aldrich).
10. Renilla DNA (pRL-CMV, Promega,for normalization of trans-
fection efficiency).
11. Opti-MEM reduced serum medium (ThermoFisher).
12. Lipofectamine LTX (ThermoFisher).
146 Sakshi Khurana et al.
3 Methods
3.1 Preliminary 1. Determine the pathways that need to be studied. In this exam-
Model and Hypotheses ple, the cross talk between the canonical WNT, NFκB, and
TGF-β pathways will be determined. These pathways play an
important role in various cellular processes, such as cell survival,
differentiation, and growth, but their cross talk is not yet fully
understood. The most important molecules and interactions in
these pathways are shown in Fig. 1 [20–22].
2. In this example, the model building was initiated by placing
nodes in ANIMO for the compounds that can be used to
stimulate each of these pathways (see Fig. 2a). The shape of
the nodes corresponds to the type of molecule they represent
(cytokine, receptor, transcription factor, gene, etc.) while the
Fig. 1 Cross talk between the canonical WNT, NFκB, and TGF-β pathways as found in literature
[20–22]. These pathways play an important role in various cellular processes, such as cell survival,
differentiation, and growth
Modeling with ANIMO 147
Fig. 2 Flow of generating an ANIMO model to simulate pathway interactions. (a) Start with compounds that can
be used to stimulate each of these pathways. (b) Add transcription factors and genes that are up- or
downregulated when each of these compounds is added to the cell. (c) Add interactions that describe the
cross talk between the studied pathways
6. After completing the model with all nodes and edges, a simula-
tion can be run to generate hypotheses (see Fig. 3). In our
example, we are interested in the influence of the stimuli on
the activation of TCF/LEF and subsequent target gene expres-
sion. Depending on the chosen scenario and reaction con-
stants, different hypotheses will be generated (see Note 9).
The rate constants can, at this stage, be manipulated until the
results roughly correspond to what is described in literature.
However, it is recommended to do the fine-tuning of the
model after the wet-lab experiments.
Modeling with ANIMO 149
Fig. 3 Hypotheses of the activation of the TCF/LEF luciferase reporter with different stimulation factors
generated using the ANIMO model as shown in Fig. 2c
Table 1
TCF/LEF-luciferase expression in C2C12 mouse myoblasts for different stimulation factors (10 ng/mL
TGF-β3, 1 μM BIO, and 10 ng/mL IL-1β) at different time points (17 h, 22 h, 41 h, and 48 h after
stimulation). N ¼ 4
Fig. 4 Luciferase expression under control of the TCF/LEF promoter in C2C12 mouse myoblasts for different
stimulation factors (TGF-β3, BIO, and IL-1β) at different time points (17 h, 22 h, 41 h, and 48 h after
stimulation). N ¼ 4
3.3 Validation and 1. Identify the differences between the wet-lab results and the
Adjustment of Model hypotheses generated with the preliminary model. These can
be differences in expression (e.g., the wet-lab results show an
increase in expression for a specific gene while this gene is
downregulated in the preliminary model) or in time scale (see
Note 15). In our example, the expression levels correspond
fairly well with the hypotheses, except that peak behavior is
seen in the wet-lab results (see Fig. 4, expression is downregu-
lated after 17 h for all conditions), which was not yet included
in the preliminary model.
2. Add missing nodes and/or connections to the model. For
example, adding a specific node might explain certain behavior
as observed in the wet-lab experiments or a specific inhibition
dynamic can solve problems in the simulated expression. This
completely depends on the specific model and might not always
be necessary. An additional literature search might help identi-
fying missing nodes or connections.
3. Incorporate peak behavior in the model if this is not already
done. In our example, peak behavior was observed in the
experimental results, but was not yet incorporated in the
model. This peak behavior can be explained by the fact that
luciferase fluorescent intensity is measured, which has a fairly
short half-life. By adding auto-inactivation/ inhibitory loops to
each node, this peak behavior could be added to the model to
better fit the experimental data, as described before [8]. In
Fig. 5, the optimized model is shown for our example, which
is based on Fig. 2c and improved using wet-lab data.
4. Optimize the rate parameters k to fit the wet-lab data. This can
be done manually by changing each and every rate parameter
until the simulations match the data from the wet-lab experi-
ment. However, for larger models this can get very time con-
suming and complicated. To make this process a little easier,
Modeling with ANIMO 153
Fig. 5 Optimized ANIMO model based on the model as shown in Fig. 2c. Wet-lab results are used to improve
the model. Auto-inactivation/inhibitory loops were added to each node to incorporate peak behavior in the
model
Fig. 6 Results of optimization of rate parameters k to obtain a model behavior (red) that corresponds with the
obtained wet-lab data (blue) for various stimuli
4 Notes
k-
From node To node Influence Scenario value
TGF-β3 Type 1/2 receptor Activation 1 0.004
Type 1/2 receptor Smad2/3 Activation 1 0.004
Smad2/3 Smad2/3/4 Activation 1 0.004
Smad2/3/4 SBE transcription Activation 1 0.002
factor
SBE transcription SBE luciferase Activation 1 0.001
factor
WNT β-catenin Activation 1 0.004
β-catenin TCF/LEF Activation 1 0.002
transcription factor
TCF/LEF TCF/LEF luciferase Activation 1 0.001
transcription factor
TCF/LEF luciferase CCND1 gene Activation 1 0.001
(continued)
Modeling with ANIMO 157
k-
From node To node Influence Scenario value
IL-1β NFκB Activation 1 0.004
NFκB NFκBia gene Activation 1 0.001
Smad2/3 Smad7 Activation 1 0.004
Smad2/3/4 β-catenin Activation 1 0.004
Smad7 NFκB Inhibition 1 0.004
Smad7 β-catenin Inhibition 1 0.004
Smad7 Type 1/2 receptor Inhibition 2 0.004
β-catenin Type 1/2 receptor Inhibition 1 0.004
NFκB TCF/LEF Activation 1 0.002
transcription factor
NFκB Smad7 Activation 1 0.004
NFκB CCND1 gene Activation 1 0.001
NFκB β-catenin Inhibition 1 0.004
9. The choice for the scenario and reaction constant k have a large
effect on the results of the simulation. For the initial model,
most edges are set to scenario 1 (unless described differently in
literature). For the reaction constant it generally holds true that
complex steps such as transcription of a gene are slow and thus
have a low reaction constant k, while other processes such as the
binding of a stimulation factor to a receptor or a transcription
factor to a gene are generally fast and have a high reaction
constant k. For each edge, an initial approximation of the
reaction constant k needs to be made, if possible based on
literature, which can be optimized using results from wet-lab
experiments [8].
10. It is important to note that measuring a reporter is not identical
to measuring the protein directly. In a reporter assay, the
measured signal is a result of a transcription factor (in this
case TCF/LEF) binding to a promotor region in a gene,
resulting in transcription of a protein bound to a reporter
such as luciferase. The luciferase then produces a fluorescent
signal after addition of a substrate, of which the intensity can
then be measured. Since this process includes many additional
steps that sometimes have a limited life-time, the measured
response can differ from the result that would be obtained
when directly measuring the transcription factor TCF/LEF.
11. C2C12 mouse myoblasts is an immortalized mouse myoblast
cell line that undergoes rapid proliferation and can be differ-
entiated into myoblasts and osteoblasts. They are a popular cell
158 Sakshi Khurana et al.
References
1. Liang Y, Kelemen A (2017) Dynamic model- for the analysis of cell signaling networks. Bio-
ing and network approaches for omics time chemistry 49(15):3216–3224. https://doi.
course data: overview of computational org/10.1021/bi902202q
approaches and applications. Brief Bioinform 11. Schivo S, Scholma J, van der Vet PE,
19(5):1051–1068. https://doi.org/10.1093/ Karperien M, Post JN, van de Pol J, Langerak
bib/bbx036 R (2016) Modelling with ANIMO: between
2. Du Q, Zhang X, Liu Q, Zhang X, Bartels CE, fuzzy logic and differential equations. BMC
Geller DA (2013) Nitric oxide production Syst Biol 10(1):56. https://doi.org/10.
upregulates Wnt/β-catenin signaling by inhi- 1186/s12918-016-0286-z
biting Dickkopf-1. Cancer Res 73 12. Terfve C, Cokelaer T, Henriques D,
(21):6526–6537. https://doi.org/10.1158/ MacNamara A, Goncalves E, Morris MK,
0008-5472.CAN-13-1620 Mv I, Lauffenburger DA, Saez-Rodriguez J
3. Zhong L, Schivo S, Huang X, Leijten J, (2012) CellNOptR: a flexible toolkit to train
Karperien M, Post JN (2017) Nitric oxide protein signaling networks to data using multi-
mediates crosstalk between interleukin 1β and ple logic formalisms. BMC Syst Biol 6(1):133.
WNT signaling in primary human chondro- https://doi.org/10.1186/1752-0509-6-133
cytes by reducing DKK1 and FRZB expression. 13. Chaouiya C, Remy E, Mossé B, Thieffry D
Int J Mol Sci 18(11):2491. https://doi.org/ (2004) Qualitative analysis of regulatory
10.3390/ijms18112491 graphs: a computational tool based on a dis-
4. Schivo S, Khurana S, Govindaraj K, Scholma J, crete formal framework. In: Benvenuti L, De
Kerkhofs J, Zhong L, Huang X, van de Pol J, Santis A, Farina L (eds) Positive systems. Lec-
Langerak R, van Wijnen AJ, Geris L, ture Notes in Control and Information Sci-
Karperien M, Post JN (2019) ECHO, the exe- ence, vol 294. Springer, Berlin. https://doi.
cutable CHOndrocyte: a computational model org/10.1007/978-3-540-44928-7_17
to study articular chondrocytes in health and 14. Bock M, Scharp T, Talnikar C, Klipp E (2013)
disease. Cell Signal 68:109471. https://doi. BooleSim: an interactive Boolean network sim-
org/10.1016/j.cellsig.2019.109471 ulator. Bioinformatics 30(1):131–132.
5. Luechtefeld T, Marsh D, Rowlands C, Har- https://doi.org/10.1093/bioinformatics/
tung T (2018) Machine learning of toxicologi- btt568
cal big data enables read-across structure 15. Schivo S, Scholma J, Karperien M, Post JN,
activity relationships (RASAR) outperforming Van De Pol J, Langerak R (2014) Setting para-
animal test reproducibility. Toxicol Sci 165 meters for biological modelsWith ANIMO. In:
(1):198–212. https://doi.org/10.1093/ Electronic Proceedings in Theoretical Com-
toxsci/kfy152 puter Science, EPTCS, pp 35–47. doi:https://
6. Grimm D (2019) U.S. EPA to eliminate all doi.org/10.4204/EPTCS.145.5
mammal testing by 2035. https://www. 16. Krumsiek J, Pölsterl S, Wittmann DM, Theis
sciencemag.org/news/2019/09/us-epa-elimi FJ (2010) Odefy - from discrete to continuous
nate-all-mammal-testing-2035 models. BMC Bioinformatics 11(1):233.
7. Grimm D (2019) EPA plan to end animal test- https://doi.org/10.1186/1471-2105-11-
ing splits scientists. Science 365(6459):1231. 233
https://doi.org/10.1126/science.365.6459. 17. Mendes P, Hoops S, Sahle S, Gauges R, Dada J,
1231 Kummer U (2009) Computational modeling
8. Scholma J, Schivo S, Urquidi RA, Pol JVD, of biochemical networks using COPASI. In:
Karperien M, Post JN (2014) Biological net- Maly IV (ed) Methods in molecular biology
works 101 : Computational modeling for (methods and protocols)-systems biology, vol
molecular biologists. Gene 533(1):379–384. 500. Vol systems biology. Humana Press,
https://doi.org/10.1016/j.gene.2013.10. New York, NY. https://doi.org/10.1007/
010 978-1-59745-525-1_2
9. Brodland GW (2015) How computational 18. Matsuoka Y, Funahashi A, Ghosh S, Kitano H
models can help unlock biological systems. In: (2014) Modeling and simulation using cellde-
Seminars in cell & developmental biology. signer. In: Miyamoto-Sato E, Ohashi H,
Elsevier, pp 62–73. https://doi.org/10. Sasaki H, Nishikawa J-I, Yanagawa H (eds)
1016/j.semcdb.2015.07.001 Methods in molecular biology (methods and
10. Morris MK, Saez-Rodriguez J, Sorger PK, protocols)- Transcription factor regulatory
Lauffenburger DA (2010) Logic-based models networks., vol 1164. Humana Press,
160 Sakshi Khurana et al.
New York, NY. https://doi.org/10.1007/ 30. Frederick JP, Liberati NT, Waddell DS, Shi Y,
978-1-4939-0805-9_11 Wang X-F (2004) Transforming growth factor
19. Schivo S, Scholma J, Wanders B, Camacho β-mediated transcriptional repression of c-myc
RAU, van der Vet PE, Karperien M, is dependent on direct binding of Smad3 to a
Langerak R, van de Pol J, Post JN (2014) novel repressive Smad binding element. Mol
Modeling biological pathway dynamics with Cell Biol 24(6):2546. https://doi.org/10.
timed automata. Ieee J Biomed Health 18 1128/MCB.24.6.2546-2559.2004
(3):832–839. https://doi.org/10.1109/Jbhi. 31. Higuera GA, Hendriks JA, van Dalum J, Wu L,
2013.2292880 Schotel R, Moreira-Teixeira L, van den
20. Eisenmann DM (2005) Wnt signaling. Doel M, Leijten JC, Riesle J, Karperien M,
https://doi.org/10.1895/wormbook.1.7.1. van Blitterswijk CA, Moroni L (2013) In vivo
Accessed 19 Dec 2019 screening of extracellular matrix components
21. Murphy AM, Wong AL, Bezuhly M (2015) produced under multiple experimental condi-
Modulation of angiotensin II signaling in the tions implanted in one animal. Integr Biol 5
prevention of fibrosis. Fibrogenesis Tissue (6):889–898. https://doi.org/10.1039/
Repair 8:7–7. https://doi.org/10.1186/ c3ib40023a
s13069-015-0023-z 32. Jonk LJC, Itoh S, Heldin C-H, ten Dijke P,
22. Morlon A, Munnich A, Smahi A (2005) TAB2, Kruijer W (1998) Identification and functional
TRAF6 and TAK1 are involved in NF-κB acti- characterization of a Smad binding element
vation induced by the TNF-receptor, Edar and (SBE) in the JunB promoter that acts as a
its adaptator Edaradd. Hum Mol Genet 14 transforming growth factor-β, activin, and
(23):3751–3757. https://doi.org/10.1093/ bone morphogenetic protein-inducible
hmg/ddi405 enhancer. J Biol Chem 273
(33):21145–21152. https://doi.org/10.
23. MacDonald BT, Tamai K, He X (2009) Wnt/ 1074/jbc.273.33.21145
beta-catenin signaling: components, mechan-
isms, and diseases. Dev Cell 17(1):9–26. 33. Ranganathan P, Agrawal A, Bhushan R, Cha-
https://doi.org/10.1016/j.devcel.2009.06. valmane AK, Kalathur RKR, Takahashi T, Kon-
016 daiah P (2007) Expression profiling of genes
regulated by TGF-beta: differential regulation
24. Tran FH, Zheng JJ (2017) Modulating the wnt in normal and tumour cells. BMC Genomics
signaling pathway with small molecules. Pro- 8:98–98. https://doi.org/10.1186/1471-
tein Sci 26(4):650–661. https://doi.org/10. 2164-8-98
1002/pro.3122
34. Hyun Hwa C, Hye Joon J, Ji Sun S, Yong
25. Verstrepen L, Bekaert T, Chau TL, Tavernier J, Chan B, Jin Sup J (2008) Crossregulation of
Chariot A, Beyaert R (2008) TLR-4, IL-1R β-catenin/Tcf pathway by NF-κB is mediated
and TNF-R signaling to NF-κB: variations on by lzts2 in human adipose tissue-derived mes-
a common theme. Cell Mol Life Sci 65 enchymal stem cells. Biochim Biophys Acta
(19):2964–2978. https://doi.org/10.1007/ 1783(3):419–428. https://doi.org/10.1016/
s00018-008-8064-8 j.bbamcr.2007.08.005
26. Massagué J (1998) TGF-β signal transduction. 35. Du Q, Geller DA (2010) Cross-regulation
Annu Rev Biochem 67(1):753–791. https:// between Wnt and NF-κB signaling pathways.
doi.org/10.1146/annurev.biochem.67.1.753 For Immunopathol Dis Therap 1(3):155–181.
27. Cadigan KM, Waterman ML (2012) https://doi.org/10.1615/For
TCF/LEFs and Wnt signaling in the nucleus. umImmunDisTher.v1.i3
Cold Spring Harb Perspect Biol 4(11): 36. Wang J, Zhao J, Chu ESH, Mok MTS, Go
a007906. https://doi.org/10.1101/ MYY, Man K, Heuchel R, Lan HY, Chang Z,
cshperspect.a007906 Sung JJY, Yu J (2013) Inhibitory role of Smad7
28. Wu B, Crampton SP, Hughes CCW (2007) in hepatocarcinogenesis in mice and in vitro. J
Wnt signaling induces matrix metalloprotei- Pathol 230(4):441–452. https://doi.org/10.
nase expression and regulates T cell transmigra- 1002/path.4206
tion. Immunity 26(2):227–239. https://doi. 37. Guo X, Wang X-F (2009) Signaling cross-talk
org/10.1016/j.immuni.2006.12.007 between TGF-beta/BMP and other pathways.
29. Heather G, Chun P (2006) Activin receptor- Cell Res 19(1):71–88. https://doi.org/10.
like kinases: structure, function and clinical 1038/cr.2008.302
implications. Endocr Metab Immune Disord 38. Huang N, Li W, Wang X, Qi S (2018) Micro-
Drug Targets 6(1):45–58. https://doi.org/ RNA-17-5p aggravates lipopolysaccharide-
10.2174/187153006776056585 induced injury in nasal epithelial cells by
Modeling with ANIMO 161
Abstract
Our laboratories have used genetically engineered mouse models (GEMMs) to assess genetic contributions
to skeletal diseases such as osteoporosis and osteoarthritis. Studies on the genetic contributions to OA are
often done by assessing how GEMMs respond to surgical methods that induce symptoms modeling
OA. Here, we will describe protocols outlining the induction of experimental OA in mice as well as detailed
descriptions of methods for analyzing skeletal phenotypes using micro-computerized tomography and
skeletal histomorphometry.
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_11, © Springer Science+Business Media, LLC, part of Springer Nature 2021
165
166 Gabrielle E. Foxa et al.
2 Materials
2.1 Surgical 1. Experimental mice of the same strain divided into DMM sur-
Destabilization gery group (n 5) and sham surgery control group (n 5).
of the Medial Age of mice should be decided according to the goal of experi-
Meniscus (DMM) ments, but generally mice at or older than 3 months of age are
optimal as their skeletal maturity is reached. All mice should be
housed in a gnotobiotic facility, according to the National
Research Council’s Guide and Institutional Animal Care and
Use Committee-approved protocols for the care and use of
experimental animals (see Note 1).
2. Isoflurane machine: VetEquip Tabletop laboratory animal
anesthesia system (VetEquip # 901820).
3. Isoflurane (Piramal # NDC 66794–017-25).
4. Surgical microscope (Leica # M651).
5. Veterinary operating table heated mat (Peco #
Mediheat V500).
6. 9 mm Wound forceps and clips (Reflex # RS-9260, RS-9262).
7. Wound clip remover (Reflex # RS-9263).
8. Hair removal lotion.
9. Alcohol antiseptic wipes.
10. Micro-iris scissors.
11. Micro-surgical knife (#3 handle and #10/ #11 blade).
12. Sharppoint 15 5 mm blade micro-surgical knife.
13. Epinephrine (1:1000 dilution).
14. Amoxicillin (20 mg/kg).
2.3 Microcomputed All standards, equipment, and software for scanning and analysis are
Tomography produced by Bruker microCT, Kontich, Belgium and Micro Pho-
tonics Inc., Allentown, PA.
1. Neutral buffered formalin (NBF) 10%.
2. Deionized (DI) water.
3. Ethanol (EtOH) 70%.
4. Gauze cut to length of the sample.
5. Plastic conical.
6. Bruker Skyscan 1172.
7. Skyscan 1172 (Scanning software) Version 1.5.26.0.
8. Software NRecon Version 1.7.4.6.
9. GPUReconServer Version 1.7.4.2.
10. Data Viewer Version 1.5.6.3 software.
2.4 Bone Mineral 1. Bone mineral density standards 0.25 and 0.75.
Density Calibration 2. CT Analyzer Version 1.18.8.0 software.
3 Methods
3.1 DMM Surgery The methods described below that we utilize in the laboratory have
been developed and optimized based on several previous studies
[3–6]. Operate the anesthesia machine according to the VetEquip
user guide and operating manual (http://www.vetequip.com/
pdfs/LAAS%20
Manual.pdf). DMM surgery should be performed on both
knees of each individual mice of the DMM group, and sham
surgery will be performed to another group of mice for controls.
Surgery should be carried out with standard sterile techniques, i.e.,
preoperative instrument sterilization and disinfection, and cleans-
ing of incision area.
1. Anesthetize the mice with 4% isoflurane and 1 L/min oxygen
in the VetEquip tabletop laboratory animal anesthesia chamber.
2. Once the mice are anesthetized, transfer them to a mask and
maintain with 3% isoflurane and 1 L/min oxygen.
3. Place the mice on a veterinary operating table heated mat to
maintain body temperature during surgery.
4. To bare the incision area, apply hair removal lotion on the
medial side of knee for 3 min. Then wipe with alcohol antisep-
tic wipes.
5. In order to expose the medial side of stifle joint capsule, use a
micro-iris scissor or a micro-surgical knife to open a 3 mm
longitudinal incision through the skin and subcutaneous tissue
over the distal patella to proximal tibial plateau.
6. Expose medial meniscotibial ligament (MMTL), which
anchors medial meniscus (MM). During the process, avoid
destabilizing patellar tendon, medial collateral ligament, and
articular cartilage.
172 Gabrielle E. Foxa et al.
3.2 Osteoarthritis At 4 to 6 weeks (or at specific time point according to the experi-
(OA) Histological mental design) post DMM surgery, OA phenotypes can be graded
Grading or assessed by histological analysis, gene expression assay, microCT,
pain behaviors, or gait analysis. Here we focus on the histological
grading.
1. Collect knees from mice and arrange in a histological tissue
cassette case (see Note 7).
2. To fix the tissue, place the tissue cassettes with mouse knees in
10% NBF for 48 h at 4 C. Rinse in DI water 3 for 10 min
each time and store in 70% ethanol at 4 C.
3. To decalcify the samples, place the fixed tissues with cassettes in
a beaker containing at least 15 volumes of 5% formic acid on a
swirling/rocking platform for 7 days (see Note 8).
4. Rinse the tissues in DI water 4 for 10 min each time.
5. Dehydrate the tissues in 70%, 95%, and 100% ethanol for
multiple rounds 1–2 h each (see Note 9).
6. Clear tissues in xylene for 1 h 2.
7. Infiltrate tissues with paraffin (Paraplast X-tra) at 58 C for 1 h 2.
8. Embed tissues into wax blocks at 62 C.
9. Rehydrate tissues for histology using xylene, absolute alcohol,
95% ethanol, 80% ethanol, then rinse in distilled water (see
Note 10).
10. Cut 5 mm sections of the tissues embedded in the blocks (see
Note 11).
Generation and Characterization of Mouse Models for Skeletal Disease 173
3.3 Imaging by These methods have been adapted and optimized for use in our
Microcomputed laboratories from several sources [7–9]. Methodology is specific to
Tomography femurs from mice at 3–6 months of age. All microCT methods were
performed under the recommended guidelines of the SkyScan
software Standard Operating Procedures by Bruker MicroCT
(Kontich, Belgium).
1. Fix limbs in 10% NBF for 48 h. Rinse limbs in deionized
(DI) water and store in 70% ethanol.
2. Prepare the limbs by wrapping them in gauze saturated in 70%
ethanol. Load the limbs into a conical submerged in 70%
ethanol while preventing air bubbles from entering. Samples
should not be overlapping and should be snug in the container.
3. Turn the Skyscan 1172 on by turning key to “START” posi-
tion. The key will automatically snap back into the “ON”
position.
4. Open Skyscan 1172 software and start the X-ray tube by push-
ing the X-ray icon. A pop-up window will open with a progress
bar as the tube ages. Once the X-ray is on, an indicator window
will open.
5. Press the “grab image” icon to obtain a continuous preview
from the X-ray camera and establish scan settings (see
Note 13).
174 Gabrielle E. Foxa et al.
3.4 Bone Mineral 1. Scan, reconstruct, and orient the bone mineral density stan-
Density Calibration dards following the steps under “Imaging by Microcomputed
Tomography 3.3.”
2. Open the CTan software and load one of the density standards
by choosing the “open image or dataset” icon.
Generation and Characterization of Mouse Models for Skeletal Disease 175
3.5 Trabecular 1. In the CTan software, select the “open image or dataset” icon.
Analysis Choose a dataset within a newly created VOI folder.
of Microcomputed 2. Identify the reference point for the ROI, use a 0.25 mm offset,
Tomography Images and create an ROI 2.5 mm in height (see Note 31).
3. Create a region of interest to include trabecular bone only, by
manually drawing an ROI (see Note 32) or use an automated
program and save the regions. If manually drawing ROIs,
continue to step 6 after this step. To make an automated
program to create an ROI, continue to step 4.
4. Begin the automated trabecular ROI process by, saving the
ROI defined in step 2 and determining the thresholding scale
(see Note 33).
5. Open the ROI saved in step 4 by choosing the “open an
existing image or dataset” icon and choosing an image within
the ROI folder for the sample. Build a task list to create,
analyze, and save an automated trabecular ROI (see Note 34).
After building the task list, skip to step 9.
6. Open a sample’s manual ROI by choosing the “open an exist-
ing image or dataset” icon and choosing an image within the
trabecular ROI folder for the sample.
7. Apply the manual ROI to the open file and determine the
thresholding scale (see Note 35).
176 Gabrielle E. Foxa et al.
3.6 Cortical Analysis 1. In the CTan software, select the “open image or dataset” icon.
of Microcomputed 2. Identify the reference point for the ROI and set the height to
Tomography Images 0.6 mm (see Note 38).
3. Create a region of interest to include the entire area of the bone
(see Note 39).
4. Save the cortical ROI (see Note 40).
5. Open a sample ROI by choosing the “open an existing image
or dataset” icon and choosing an image within the cortical ROI
folder for the sample.
6. Apply the ROI to the open file and determine the thresholding
scale (see Note 41).
7. Build a 2D analysis task list using a combination of functions
until the cortical bone that remains in the “binary selection
preview” closely matches the raw images (see Note 42).
8. Build a histogram Task List by constructing a task list the same
as the “2D analysis Task List” except for the 2D analysis and
3D model functions. Add ROI shrink-wrap, reload, histogram
functions to the end of the list (see Note 43).
9. Run the task list on multiple samples using the batch manager.
Repeat for each task list (see Note 44).
3.7 Modeling Several modifications can be made when generating visual repre-
Skeletal Phenotypes sentations of scanned samples. The guidelines below are for a basic,
gray-scale model.
1. Using CTvol, open a 3D file by selecting the “open 3D object”
icon and choosing the p3g file within the ROI folder of the
desired sample.
2. Use the wheel on the computer mouse to zoom in and out on
the sample.
3. Select “Options” and choose “Objects” to open up the
“Objects” window. Adjust color, emission, diffusion, reflec-
tion, and opacity as desired within this window.
4. Select the “move camera” or “move objects” icon and use your
mouse to adjust the sample to the desired position and angle.
5. Save the Image by selecting “Save Image” from the “File”
dropdown menu.
Generation and Characterization of Mouse Models for Skeletal Disease 177
3.9 Infiltration 1. Fix the femurs in 10% NBF for 48 h, followed by storage in 70%
ethanol.
2. Remove excess tissue from femurs and move to individual
histocassette tissue cassettes.
3. Using the Dremel hand tool, notch each bone ¼ to ½ of the
way through in the midshaft region. Return each bone to its
cassette.
4. Dehydrate and clear the samples for 2–4 h in 70% ethanol, 95%
ethanol, 100% ethanol, and xylene (see Note 46).
5. Transfer the specimens to xylene until infiltration.
6. Within the RIP unit infiltrate the samples under 15 psi using
100% MMA for 24 h (infiltration 1), the infiltration solution
for 24 h (infiltration 2), followed by the infiltration solution for
at least 2 days (infiltration 3).
7. Transfer bone tissue to individual, prefabricated embedding
molds for coronal sectioning or glass culture tubes for cross-
sectioning.
3.10 Femoral When processing tissues that are light sensitive due to fluoro-
Embedding chrome labeling, keep in the dark in between processing steps.
and Cross-Sectioning
1. Add a few drops of glass culture tube embedding solution into
a labeled glass culture tube and insert the femur, ensuring the
head of the femur is closest to the open end of the tube (see
Note 47).
178 Gabrielle E. Foxa et al.
2. Cap the glass tube with a rubber stopper and transfer the
beaker into the bead bath at 25 C for optimal polymerization.
3. Check the glass tubes once daily until the polymerization pro-
cess is complete (see Note 48).
4. Remove the polymerized rod containing the femur from the
glass culture tube by gently shattering the surrounding glass.
5. Load the resulting plastic rod into the specimen holder of the
Extech Labcut 150. Secure the rod into the specimen holder by
tightening down the adjustment bolts with an Allen wrench.
6. Line up the plastic rod with the diamond wafering blade to the
midshaft of the femur and move the specimen arm so it sits at
the back of the Labcut 150.
7. Place either the unattached reed switch magnet or the canopy
in place so that the reed switch is covered.
8. Turn the instrument on, set the speed to 166.2 rpm, select the
“rev” button to start the blade.
9. Turn on the lubricant pump and slowly pull the specimen arm
forward.
10. Sit the rod atop the saw blade and begin to section the bone at
the midshaft or slightly more distal (see Note 49).
11. Save the distal end for MMA embedding and coronal cross-
sectioning or discard it.
12. Adjust the micrometer on the Extech Labcut 150 with the
diamond wafering blade to the desired thickness for the first
wafer. Ensure the rod is advancing over the blade (see Note
50).
13. Section desired number of wafers.
14. Remove any plastic dust particles from the wafers by dipping
each in a sonicator filled with deionized water for 3–5 s. Dab
off any excess water.
15. Place a nickel size pool of Eukitt mounting media onto the
back of a labeled glass slide and mount the wafers in the pool.
16. Cover the glass slide with 2–3 layers of 1–2 ply paper towel cut
into slide size strips.
17. Place a blank glass slide on top of the paper towel strips and
fasten both slides together with 2–3 binder clips and allow the
slide to air dry overnight.
18. Remove the binder clips, blank slide, and paper towel strips
from the surface of the slide with mounted wafers to be sanded.
19. Using a table micrometer, zero out the slide thickness by
placing a portion of the glass slide that does not contain a
wafer between the clamps. Tighten the clamp on the slide
and clear the counter.
Generation and Characterization of Mouse Models for Skeletal Disease 179
20. Open the micrometer clamp and move the slide so one of the
mounted wafers is measurable. Close the clamp down onto the
wafer and obtain a starting measurement of the wafer. Repeat
this step for all wafers.
21. Using 600 grit sandpaper, gently sand the surface of the wafers
until each measures 40–50 μm. Remeasure each wafer using
the micrometer throughout the sanding process in order to
determine when the desired thickness has been obtained.
22. Once all wafers have been sanded to the desired thickness, wipe
the sanding remnants from the surface of the slide with a damp
paper towel.
23. Add a thin layer of Eukitt mounting media to the surface of the
wafers and add a glass coverslip.
24. Add 2–3 small weights to the top of the coverslip to ensure any
air bubbles are removed and allow the slides to air dry
overnight.
3.11 MMA When processing tissues that are light sensitive due to fluoro-
Embedding chrome labeling, keep in the dark in between processing steps.
and Coronal
1. Add a thin layer of prefabricated mold embedding solution to
Sectioning of Femurs each prefabricated mold and allow the reagent to become
sticky/tacky.
2. Orient the tissue for coronal sectioning and fill remainder of
the mold with prefabricated mold embedding solution.
3. Add a paper label containing the specimen ID, and a paper clip
bent to a 45 –90 angle to each mold.
4. Cap the top of the mold with a silicone stopper and transfer
each mold to the bead bath at 25 C, ensuring the containers
are covered.
5. Remove molds from the bead bath daily and check until poly-
merization is complete (see Note 51).
6. Remove fully polymerized molds to grind and polish using the
MetaServ 250 Single Grinder/Polisher (see Note 52).
7. Apply a thin layer of the coating solution to a glass slide.
8. Place the embedded tissue mold in the microtome chuck and
face into the mold until the desired tissue is exposed.
9. Brush the exposed face of the block with a fine tip paint brush,
soaked in the cutting solution, and initiate the automatic
microtome to begin slicing a section from the block at 5 μm.
10. Cut a small rectangle from a Kimwipe, and soak in 70% ethanol.
11. Apply the Kimwipe to the surface of the mold and initiate
cutting on the microtome at a very slow speed.
180 Gabrielle E. Foxa et al.
12. As the Kimwipe and plastic section are slowly cut, grasp them
together with a pair of forceps. Continue to carefully pull the
section off the block as it is cut away by the microtome.
13. Transfer the section and wipe into a petri dish of 75% ethanol.
Carefully stretch the section if wrinkles are visible using two
pairs of forceps.
14. Once flat, place a piece of plastic film, just larger than the
section, on top of the section to keep it free of wrinkles.
15. Transfer the section and plastic film combination to the coated
slide.
16. Cover the slide with a thin sheet of plastic and lightly squeeze
out any excess ethanol using a gentle wiping motion with a
Kimwipe.
17. Cover the section with paper towel and roll flat with rubber
roller.
18. Each day sectioning is completed, stack the slides together with
a blank slide on the top and bottom of the stack and place in a
clamp. Transfer each stack to an oven set at 60 C, overnight.
19. Remove slides from the oven and inspect for damage.
20. Begin deplasticizing slides by soaking them in cellusolve for
1 h.
21. Place slides in acetone for 2–5 min.
22. Rinse slides in DI water for 2–5 min.
23. Continue to “Goldner’s Trichrome Staining 3.12” step 1 for
staining or step 10 to only coverslip for fluorochrome imaging.
3.13 Paraffin 1. Fix the femurs in 10% NBF for 48 h, followed by storage in 70%
Embedding ethanol.
and Sectioning 2. Remove excess tissue from femurs and move to individual
of Femurs histocassette tissue cassettes.
3. Decalcify samples in 10% EDTA pH 7.4, stirring for
10–14 days at 4 C (see Note 53).
4. Dehydrate, clear, and infiltrate the samples in cassettes within a
tissue processor using 70% ethanol for 60 min, 80% ethanol for
60 min, 95% ethanol for 60 min 2, 100% ethanol for 60 min
3, xylene for 30 min 2, paraffin for 30 min, and paraffin for
45 min.
5. Remove the cassettes from the tissue processor and move them
to the embedding center.
6. Set the heated areas of the embedding center between 60 C
and 70 C.
7. Remove the tissue from the cassette and place it in an appropri-
ately sized heated mold. Orient all the samples the same way
depending on the type of cut desired (sagittal or coronal).
8. Hold the tissue specimen down with forceps while partially
filling the mold with molten paraffin. Secure the tissue by
quickly cooling the base of the mold (see Note 54).
9. Place the cassette bottom atop the mold and fill with paraffin.
10. Cool the blocks in a cooling area to set the paraffin until it has
solidified.
11. Remove the blocks from the mold by pulling upward on the
cassette.
12. Rough cut/face each block to expose the desired surface area
of the tissue(s). Precool paraffin blocks with the tissue side
down in a tray of ice water to facilitate sectioning.
13. Using a steel microtome knife or disposable blade, cut the
appropriate sections (see Note 55).
14. For histological sections, place the sections on the pre-labeled
glass slides with the IDs that correspond to each block being
sectioned. Mount sections by floating a paraffin ribbon into a
water bath set between 36 C and 42 C.
15. Dry paraffin sections in a lab oven set to 60 C for 10–20 min.
16. Remove the sections from the oven and allow the slides to cool
at room temperature.
17. Deparaffinize slides by soaking in fresh xylene for 3 min 3,
hand-dip slides into 100% ethanol 10, repeat in fresh 100%
ethanol, hand-dip slides in 95% ethanol 10, repeat in fresh
95% ethanol, and rinse in DI water.
18. Store slides in DI water until TRAP staining.
182 Gabrielle E. Foxa et al.
3.14 TRAP Staining 1. Use the Sigma-Aldrich product 387-A: Acid Phosphatase, Leu-
kocyte (TRAP) kit and protocol to stain deparaffinized
samples.
2. Coverslip the slides using an aqueous mounting media and
glass coverslips.
3.15 Imaging Slides 1. Image the trichrome stained samples at 20 magnification
using a brightfield microscope (Leica Aperio AT2—High Vol-
ume, Digital Whole Slide Scanning).
2. Image fluorochrome labeled samples at 20 magnification
(Zeiss Axio Imager.A2, Lumen Dynamics X-Cite Series
120 Q, QImaging QIClick Digital CCD Camera) using objec-
tive filter ET-DAPI/FITC/Texas Red (Chroma product #
69002). Use BIOQUANT OSTEO 19.2.6 software and follow
the BIOQUANT OSTEO 19.2.6 “Image Region and Image
Menu Introduction.”
3.16 Histo- The analysis steps are specific to BIOQUANT OSTEO 19.2.6
morphometry Manual.
1. To quantify trabecular bone parameters, open BIOQUANT
OSTEO 19.2.6 software and create a new dataset following
“The Data Region: Introduction.”
2. Measure trabecular parameters (Goldner’s Trichrome stained,
Fluorochrome labeled, TRAP stained slides) by following the
“Rodent Trabecular Bone Protocol.”
3. To quantify cortical bone parameters, start by creating new
dataset following “The Data Region: Introduction.”
4. Measure cortical bone parameters by following the “Basic Cor-
tical Bone Protocol.”
4 Notes
19. In the “Output” tab, right click and drag over a nonporous
part of a sample. This will open an absorbance vs. distance
graph. This works best when using cortical bone. Record the
shape of the graph (cupping indicates beam-hardening, arching
indicates over-correction of beam-hardening, and flat indicates
little to no beam hardening). Repeat two or three times in
different proximal and distal regions of the sample. Click on
the “Settings” tab and perform the necessary corrections for
each graph type (increase the beam-hardening correction per-
centage for cupping, decrease the beam-hardening correction
percentage for arching, and no further correction needed for
flat graphs). Adjust beam-hardening settings until the graphs
become flat. The same beam-hardening settings are used for
the rest of the samples in a project. For a mouse femur, this
number typically ranges from 20% to 40%.
20. In the “Start” tab, select an area in the sample and click the
“Fine Tuning” button and select “Post-alignment.” Set num-
ber of trials to 5 and set parameter steps to 0.5 or 1. Click
“Start” and refer to the “Output” tab. Toggle through the
choices using the arrow buttons and choose an option that
minimizes changes in pores, edges, and dense particles. If the
fine tuning does not work in the range selected, manually
change the misalignment compensation in the “Settings” tab.
Repeat fine tuning for each part of the sample (there should be
3 or more segments for a mouse femur).
21. In the “Settings” tab, check “Ring Artifacts Reduction,” and
set to 10. In the “Start” tab, select an area in the sample, click
the “Fine Tuning” button, and select “Ring Artifacts Reduc-
tion.” Set number of trials to 5 and set parameter steps to
5. Click “Start” and refer to the “Output” tab. Toggle through
the choices using the arrow buttons and choose an option that
best reduces ring artifacts. This will only be done once per
sample.
22. This application should only be used if there is noise in the
background of the scan, which can be viewed in the “Output”
tab. To turn on smoothing, click the “Settings” tab and check
the “Smoothing” box. Adjust the levels until the noise in the
background is canceled. To see the effect of smoothing, choose
the “Start” tab, select an area in the sample and click “Pre-
view.” Use the same smoothing settings for the rest of the
samples in a project.
23. In the “Start” tab, select an area in the sample to view and click
“Preview.” In the “Output” tab, check “Use ROI,” and choose
a shape. Move and adjust the size of the ROI in the preview
window so that the sample is inside the ROI. Expand the ROI
186 Gabrielle E. Foxa et al.
Acknowledgements
References
1. Youlten SE, Baldock PA (2019) Using mouse 7. Bouxsein ML et al (2010) Guidelines for
genetics to understand human skeletal disease. assessment of bone microstructure in rodents
Bone 126:27–36 using micro-computed tomography. J Bone
2. Williams BO, Warman ML (2017) CRISPR/ Miner Res 25(7):1468–1486
CAS9 technologies. J Bone Miner Res 32 8. Parfitt AM et al (1987) Bone histomorphome-
(5):883–888 try: standardization of nomenclature, symbols,
3. Kamekura S et al (2005) Osteoarthritis devel- and units. Report of the ASBMR Histomor-
opment in novel experimental mouse models phometry nomenclature committee. J Bone
induced by knee joint instability. Osteoarthr Miner Res 2(6):595–610
Cartil 13(7):632–641 9. Kedlaya R et al (2013) Sclerostin inhibition
4. Glasson SS, Blanchet TJ, Morris EA (2007) reverses skeletal fragility in an Lrp5-deficient
The surgical destabilization of the medial mouse model of OPPG syndrome. Sci Transl
meniscus (DMM) model of osteoarthritis in Med 5(211):211ra158
the 129/SvEv mouse. Osteoarthr Cartil 15 10. Ma HL et al (2007) Osteoarthritis severity is
(9):1061–1069 sex dependent in a surgical mouse model.
5. Welch ID et al (2009) The retinoic acid bind- Osteoarthr Cartil 15(6):695–700
ing protein CRABP2 is increased in murine 11. Glasson SS et al (2010) The OARSI histopa-
models of degenerative joint disease. Arthritis thology initiative - recommendations for histo-
Res Ther 11(1):R14 logical assessments of osteoarthritis in the
6. Fang H et al (2018) Early changes of articular mouse. Osteoarthr Cartil 18(Suppl 3):
cartilage and subchondral bone in the DMM S17–S23
mouse model of osteoarthritis. Sci Rep 8
(1):2855
Chapter 12
Abstract
Our understanding of the mechanisms underlying fracture healing is rapidly developing and is contributing
to new therapeutic strategies to enhance repair. To gain new insights, animal models must also evolve. From
initially imprecise, uncontrolled bone defects we now have precise injury models that still capture all of the
stages and phases of bone repair yet do so in a highly reproducible manner. The simple mono-cortical defect
model allows assessment of bone repair through a cartilage intermediate, e.g., endochondral ossification, as
well as direct bone repair, e.g., intramembranous healing. Cellular contributions of the periosteum can be
distinguished from contributions originating in the bone marrow. In this chapter, we focus on the
advantages of this bone repair model, as well as its limitations.
1 Introduction
Bone fractures are one of the most common traumatic injuries and
their etiologies typically involve falls and/or accidents [1]. Fractures
can also occur because of underlying disease processes, e.g., cancer
and osteoporosis that weaken the bone structure and increase the
likelihood of fracture. Fracture severity is typically classified accord-
ing to whether it is closed or open (compound), simple or com-
minuted, and infected or not [2, 3]. The more severe the fracture,
the longer is the healing time. Despite the known importance of
these variables, few of the conditions are replicated in animal
models.
Standardized animal models of bone repair typically simplify
fracture severity, in hopes of producing reproducible injuries. For
example, one commonly model using blunt trauma created by
ophthalmic forceps or three-point bending equipment that is deliv-
ered to the tibia or femur and which creates a closed compound
fracture [4, 5]. These fractures can then be allowed to heal with or
without stabilization. In stabilized scenarios, an intramedullary
“nail” (e.g., an insect pin) is typically inserted into the femur to
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_12, © Springer Science+Business Media, LLC, part of Springer Nature 2021
193
194 Zhijun Li and Jill A. Helms
Fig. 1 (a) General morphological view of fibular, (b) sagittal view, (c) coronal view of 2D/3D reconstruction of
intact fibular, red arrow points the surgical site. (d) Safranin O stained images show the monocortical defect in
196 Zhijun Li and Jill A. Helms
Fig. 2 (a) Pentachrome stained images show the intact bone and healing site on post-injury (PID), (b) day
3, (e) 7, and (f) 28. (c) The PCNA expressing levels and (d) Collagen type II shows by immunostaining on PID3.
Abbreviations: gp growth plate, cb cortical bone, bm bone marrow
Fig. 1 (continued) the fibular. (e) Schematics shows gap between the bone ends constitutes the fracture.
Pentachrome stained images show healing site on post-injury (f) day 2 and (i) 21. Safranin O stained image
indicates heals via intramembranous ossification on post-injury (g) day 6; intramembranous healing show by
(h) TRAP stained image at the same time point. Abbreviations: gp growth plate, cb cortical bone, bm bone
marrow
Drill Hole Models to Investigate Bone Repair 197
Fig. 3 (a) Schematics and (b, c) Aniline blue stained images show the drill hole defects filled with a scaffold
with BMP2 and contribution to the healing process can be readily assessed. (d) Drill hole defects can receive
osteogenic cells that co-expressed luciferase, the enzyme that cleaves luciferin to produce photons of light,
were placed into a tibial defect, (e) then in vivo live imaging was used to monitor the cells as a function of time.
(f) Drill hole defects can be with growth factors(L-Wnt3a), pentachrome stained images show healing site with
(g) L-Wnt3a and with (h)L-PBS on PID 7. (i) The model can be adapted to reflect an osteonecrotic bone lesion,
(j) Aniline blue stained image shows the healing site on PID1. (k) DAPI staining show osteocytes at the cut bone
ends at the same time point. (l) The adapted drill holes model can be filled with a biomaterial, a biologic, a
scaffold, and/or cells, and their contribution to the healing process can be readily assessed (m, n). Abbrevia-
tions: gp growth plate, cb cortical bone, bm bone marrow
198 Zhijun Li and Jill A. Helms
2 Materials
2.1 Instrument As a survival surgery, sterile instruments must be used for each
Preparation animal. Therefore, all surgical instruments are placed in an envelope
and autoclaved prior to use. When performing multiple animal
surgeries at one time, instruments may be sterilized between ani-
mals using a glass bead sterilizer (30 s stay at 240–270 C). Take
care to ensure that instruments are no longer hot before touching
any tissues. Gloves should be changed between animals, and per-
sonal protective equipment (PPE) including a head cover, face
mask, and shoe covers should always be worn.
Drill Hole Models to Investigate Bone Repair 199
Fig. 4 (a) Skeletal staining show calvarial anatomy. (b) Drill holes can also be produced at the midline
overlying the sagittal sinus or in the center of the frontal or parietal bones. (c, d, e) 3D and 2D reconstruction of
calvarial defects model. (f) Schematics and (g) Aniline blue stained images show the case of calvarial defects.
(h) Calvarial defects can filled with a biomaterial, a biologic, a scaffold, and/or cells, (i) aniline blue stained
image and (k) BrdU immunoassayed image show their contribution to the healing process. Abbreviations:
P parietal bone, F frontal bone, IP intraparietal bone, N nasal bone, MN mandible
Table 1
Anesthetic agents
Table 2
Analgesic agents
Dosage
Agent name (mg/kg) Route Duration and frequency of administration
Buprenorphine 0.1 mg/kg Subcutaneous Every 6–12 h for 3 days post-op, then as needed.
(SQ)
Carprofen 5–10 mg/kg Subcutaneous Every 24 h for 3 days, then as needed
(SQ)
Buprenorphine 0.5–1.0 mg/ Subcutaneous Buprenorphine SR will be given before the start of
SR kg (SQ) surgery, and will last up to 72 h post-op.
Table 3
Anti-infective agents
Dosage
Agent name (mg/kg) Route Duration and frequency of administration
Cefazolin 25 mg/kg Subcutaneous The first dose intraoperatively and second dose 2 h
(SQ) postoperatively
Enrofloxacin 10 mg/kg Subcutaneous Only one dose intraoperatively, continue for a total of
(SQ) 2 days
Buprenorphine 0.5–1.0 mg/ Subcutaneous Buprenorphine SR will be given before the start of
SR kg (SQ) surgery, and will last up to 72 h post-op.
3 Methods
3.1 Animal 1. All animals are weighed prior to surgery and pre-surgical
Preparation weights are recorded on the surgical record.
2. Ophthalmic lubricant is placed in each eye to avoid corneal
drying during surgery.
3. Anesthetize the animal as per established regimens, such as
indicated in Table 1.
4. Remove hair with an electric clipper, then clean loose hairs with
ethanol pad.
5. At this stage, analgesics including Buprenorphine or Buprenor-
phine SR are given.
6. Next, the shaved skin is aseptically prepared and draped accord-
ing to APLAC Guidelines for Rodent Survival Surgery. Sterili-
zation of the area is performed using cotton-tipped applicators,
alternating between a Betadine solution and 70% ethanol for a
total of 3 times.
7. The animal is gently placed on top of a clean, absorbent surface
positioned on a validated heat source (e.g., circulating water
Drill Hole Models to Investigate Bone Repair 201
3.2.2 Injectable 1. Injectable anesthetic volume depends on sex, age, strain, body
Anesthesia weight, and the body condition of the animal.
2. Prepare the cocktail solution the day before surgery. Immedi-
ately before use, make sure that the solution is thoroughly
mixed.
3. In mice, injectable anesthetics are best administered via IP and
IV routes.
4. Duration of anesthesia is approximately 20–30 min.
3.3 Long Bone 1. After adequate anesthesia, an 8–10 mm skin incision is made in
Cortical Hole Drilling anterior surface of the tibia or on the lateral surface of the
femur.
2. A full-thickness flap is raised at the incision sites.
3. The overlying muscle is gently dissected off the tibia or femur.
4. The site is gently irrigated with sterile saline. Immediately
thereafter, a small hole is created using a dental drill running
at 1000–2000 rpm. The drill must be new and sharp; otherwise
too much pressure must be exerted to advance the drill and the
risk of touching the far cortex is increased.
5. The hole created in the anterior tibial plateau or in the
mid-shaft of femur should have an outer diameter of approxi-
mately 0.8–1.0 mm.
6. The drill hole should penetrate through a single tibial/femur
cortex only. No irrigation is required after creating the drill
hole. Bleeding usually stops spontaneously.
7. After creating the skeletal injury, soft tissues are closed with
absorbable suture.
8. The skin incision is closed with nonabsorbable sutures.
9. The nonabsorbable sutures are removed 7–10 days later.
202 Zhijun Li and Jill A. Helms
3.4 Calvaria Cortical 1. Once adequate anesthesia has been verified, a 3 mm skin inci-
Bone Hole Drilling sion is made in the mid-line at the caudal portion of the skull
(see Note 2).
2. The periosteum is gently elevated and reflected.
3. A drill hole is created using of a micro-dissecting trephine and
sterile saline irrigation.
4. Take extreme care to leave the underlying dura mater
undisturbed.
5. The skin incision is closed with nonabsorbable suture.
6. The nonabsorbable sutures are removed 7–10 days later (see
Note 3).
3.5 Parameters 1. Animals are transferred into a clean cage which is partially over
Monitored After a 37 C heating pad. This allows animals, once ambulatory, to
Surgery choose whether to stay on the warm side or on the
non-heated side.
2. Separate awake animals from recovering animals.
3. Animals are allowed to ambulate freely but are closely moni-
tored for any evidence of pain as indicated by licking, biting,
crouching, vocalizations, teeth chattering, stiff walking, failure
to eat, self-imposed isolation/hiding, self-mutilation, rapid
breathing, open mouthed breathing, squinting, muscle rigidity,
lack of muscle tone, twitching/trembling, or abnormal
posture.
3.6 Early Euthanasia 1. The most likely complication of skeletal surgery is that the
Criteria injury site becomes infected. During the initial recovery period
(0–3 days post-surgery) closely evaluate each animal for signs of
inflammation/infection, such as redness and swelling, or signs
of exudate at the surgical site.
2. If signs of infection are found, consult with veterinarians about
the use of antibiotics. If signs of infection persist, the animal
should be euthanized.
3. The post-surgical weight of each mouse is monitored daily. If
any animal exhibits loss of weight greater than 10%, they are
typically removed from the study and euthanized.
Drill Hole Models to Investigate Bone Repair 203
4 Notes
Acknowledgements
References
1. Bell A, Templeman D, Weinlein JC (2016) 6. O’Neill KR, Stutz CM, Mignemi NA, Burns
Nonunion of the femur and tibia: an update. MC, Murry MR, Nyman JS et al (2012) Micro-
Orthop Clin North Am 47(2):365–375 computed tomography assessment of the pro-
2. Papakostidis C, Kanakaris NK, Pretel J, gression of fracture healing in mice. Bone 50
Faour O, Morell DJ, Giannoudis PV (2011) (6):1357–1367
Prevalence of complications of open tibial 7. Einhorn TA (1995) Enhancement of fracture-
shaft fractures stratified as per the Gustilo- healing. J Bone Joint Surg Am 77(6):940–956
Anderson classification. Injury 42 8. Tay BK, Le AX, Gould SE, Helms JA (1998)
(12):1408–1415 Histochemical and molecular analyses of dis-
3. Ktistakis I, Giannoudi M, Giannoudis PV traction osteogenesis in a mouse model. J
(2014) Infection rates after open tibial frac- Orthop Res 16(5):636–642
tures: are they decreasing? Injury 45 9. Choi P, Ogilvie C, Thompson Z, Miclau T,
(7):1025–1027 Helms JA (2004) Cellular and molecular char-
4. Cheng L, Ye F, Yang R, Lu X, Shi Y, Li L et al acterization of a murine non-union model. J
(2010) Osteoinduction of hydroxyapatite/ Orthop Res 22(5):1100–1107
beta-tricalcium phosphate bioceramics in mice 10. Leucht P, Jiang J, Cheng D, Liu B,
with a fractured fibula. Acta Biomater 6 Dhamdhere G, Fang MY et al (2013) Wnt3a
(4):1569–1574 reestablishes osteogenic capacity to bone grafts
5. Kayal RA, Siqueira M, Alblowi J, McLean J, from aged animals. J Bone Joint Surg Am 95
Krothapalli N, Faibish D et al (2010) (14):1278–1288
TNF-alpha mediates diabetes-enhanced chon- 11. Tanaka K, Tanaka S, Sakai A, Ninomiya T,
drocyte apoptosis during fracture healing and Arai Y, Nakamura T (2010) Deficiency of vita-
stimulates chondrocyte apoptosis through min A delays bone healing process in
FOXO1. J Bone Miner Res 25(7):1604–1615
204 Zhijun Li and Jill A. Helms
association with reduced BMP2 expression 23. Minear S, Leucht P, Miller S, Helms JA (2010)
after drill-hole injury in mice. Bone 47 rBMP represses Wnt signaling and influences
(6):1006–1012 skeletal progenitor cell fate specification during
12. Behr B, Leucht P, Longaker MT, Quarto N bone repair. J Bone Miner Res 25
(2010) Fgf-9 is required for angiogenesis and (6):1196–1207
osteogenesis in long bone repair. Proc Natl 24. Lorenzo J, Horowitz M, Choi Y (2008)
Acad Sci U S A 107(26):11853–11858 Osteoimmunology: interactions of the bone
13. Jawad MU, Fritton KE, Ma T, Ren PG, Good- and immune system. Endocr Rev 29
man SB, Ke HZ et al (2013) Effects of scler- (4):403–440
ostin antibody on healing of a non-critical size 25. Fan M, Peng J, Wang A, Zhang L, Liu B, Ren Z
femoral bone defect. J Orthop Res 31 et al (2011) Emu model of full-range femoral
(1):155–163 head osteonecrosis induced focally by an alter-
14. Yang F, Wang J, Hou J, Guo H, Liu C (2013) nating freezing and heating insult. J Int Med
Bone regeneration using cell-mediated respon- Res 39(1):187–198
sive degradable PEG-based scaffolds incorpor- 26. Allen MR (2009) Bisphosphonates and osteo-
ating with rhBMP-2. Biomaterials 34 necrosis of the jaw: moving from the bedside to
(5):1514–1528 the bench. Cells Tissues Organs 189
15. Christou C, Oliver RA, Pelletier MH, Walsh (1–4):289–294
WR (2014) Ovine model for critical-size tibial 27. Allen MR, Kubek DJ, Burr DB, Ruggiero SL,
segmental defects. Comp Med 64(5):377–385 Chu TM (2011) Compromised osseous heal-
16. Berner A, Reichert JC, Woodruff MA, ing of dental extraction sites in zoledronic acid-
Saifzadeh S, Morris AJ, Epari DR et al (2013) treated dogs. Osteoporos Int 22(2):693–702
Autologous vs. allogenic mesenchymal progen- 28. Velez R, Soldado F, Hernandez A, Barber I,
itor cells for the reconstruction of critical sized Aguirre M (2011) A new preclinical femoral
segmental tibial bone defects in aged sheep. head osteonecrosis model in sheep. Arch
Acta Biomater 9(8):7874–7884 Orthop Trauma Surg 131(1):5–9
17. Yuan X, Pei X, Zhao Y, Li Z, Chen CH, Tulu 29. Zalavras CG, Lieberman JR (2014) Osteone-
US et al (2018) Biomechanics of immediate crosis of the femoral head: evaluation and treat-
postextraction implant osseointegration. J ment. J Am Acad Orthop Surg 22(7):455–464
Dent Res 97(9):987–994. https://doi.org/ 30. Hernigou P, Flouzat-Lachaniette CH,
10.1177/0022034518765757 Delambre J, Poignard A, Allain J, Chevallier
18. Chen T, Li J, Cordova LA, Liu B, Mouraret S, N et al (2015) Osteonecrosis repair with bone
Sun Q et al (2018) A WNT protein therapeutic marrow cell therapies: state of the clinical art.
improves the bone-forming capacity of auto- Bone 70:102–109
grafts from aged animals. Sci Rep 8(1):119 31. Hamadeh IS, Ngwa BA, Gong Y (2015) Drug
19. Colnot C, Romero DM, Huang S, Helms JA induced osteonecrosis of the jaw. Cancer Treat
(2005) Mechanisms of action of demineralized Rev 41(5):455–464
bone matrix in the repair of cortical bone 32. Salmon B, Liu B, Shen E, Chen T, Li J, Gillette
defects. Clin Orthop Relat Res 435:69–78 M et al (2017) WNT-activated bone grafts
20. Leucht P, Kim JB, Currey JA, Brunski J, Helms repair osteonecrotic lesions in aged animals.
JA (2007) FAK-mediated mechanotransduc- Sci Rep 7(1):14254
tion in skeletal regeneration. PLoS One 2(4): 33. Cooper GM, Mooney MP, Gosain AK, Camp-
e390 bell PG, Losee JE, Huard J (2010) Testing the
21. Minear S, Leucht P, Jiang J, Liu B, Zeng A, critical size in calvarial bone defects: revisiting
Fuerer C et al (2010) Wnt proteins promote the concept of a critical-size defect. Plast
bone regeneration. Sci Transl Med 2 Reconstr Surg 125(6):1685–1692
(29):29ra30 34. Leucht P, Lam K, Kim JB, Mackanos MA,
22. Neagu TP, Tiglis M, Cocolos I, Jecan CR Simanovskii DM, Longaker MT et al (2007)
(2016) The relationship between periosteum Accelerated bone repair after plasma laser cor-
and fracture healing. Romanian J Morphol ticotomies. Ann Surg 246(1):140–150
Embryol 57(4):1215–1220
Chapter 13
Abstract
Fracture healing requires the integration of many cell types, growth factors, and cytokines that cannot be
adequately studied using in vitro and in silico models. This has prompted the development of highly
informative in vivo animal models to understand the complexities of fracture repair. Here, we describe a
modified procedure for mice, first developed for rats by Bonnarens and Einhorn, that does not require a
skin incision or suturing. This procedure involves boring a hole through the skin and articular surface of the
femoral condyle with a 25-gauge needle, fixation with a K-wire, and creation of a transverse mid-diaphyseal
fracture using a three-point bending fracture device. Fracture healing can be assessed using a variety of
techniques, including microcomputed tomography, torsion testing, histological and histomorphometric
analyses, and assessment of gene expression. There are many orthopedic trauma applications of this murine
femoral fracture model ranging from assessment of safety and efficacy of novel therapeutics to the influence
of specific genes on bone repair.
Key words Fracture, Murine, Bone repair, Healing, Closed fracture, micro-CT, Torsion testing
1 Introduction
The first report of the standard closed fracture in rats was described
by Bonnarens and Einhorn in 1984 [1]. It has since been adapted
to other model systems such as mice, which have become one of the
most widely accepted translational models to study fracture pheno-
types [2–4]. The availability of numerous transgenic mouse strains
coupled with the simplicity, speed, and reproducibility of this pro-
cedure have proven useful for studying both genetic and therapeu-
tic influences on healing [5]. Studies using this model have led to
significant advancements in our understanding of the mechanisms
by which fractures heal under both pathological and treatment
conditions.
After administration of anesthetics, this method involves the
creation of a hole in the femur medullary canal by perforating the
skin and articular surface of the femoral condyle with a 25-gauge
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_13, © Springer Science+Business Media, LLC, part of Springer Nature 2021
205
206 Joseph L. Roberts et al.
needle. The K-wire is then inserted into the medullary canal using a
retrograde approach and covered by the skin. Finally, the operated
limb is held perpendicular to the line of fracture, beneath a blunt
guillotine device, wherein a transverse fracture is created by three-
point bending. Administration of preoperative analgesics helps to
manage pain.
Fractured femurs are typically harvested at multiple timepoints
to evaluate the progression of healing. Fracture healing and quality
of newly formed bone can be quantitatively assessed using micro-
computed tomography (micro-CT) imaging and torsion testing,
respectively. Radiographic scoring is another useful qualitative mea-
sure of fracture healing. Typically, specimens are fixed in formalin as
needed and processed for histology. Histological and histomorpho-
metric assessments allow for the assessment of fracture callus matrix
composition through quantification of the distribution of cartilage,
bone, and fibrous tissue. Histological sections can be further uti-
lized for immunohistochemical analyses to detect cells and tissues
that express growth factors, cytokines, and other proteins of inter-
est during fracture healing. RT-qPCR from RNA isolated from
whole callus or from histological sections is also a powerful tool
to gain valuable insight into the relationship between different
experimental models and treatments and expression of genes dur-
ing fracture healing.
2 Materials
5. Nuclease-free water.
6. Forward and reverse primers for genes of interest and house-
keeping genes.
7. PCR plates or strips compatible with machine.
8. Centrifuge with plate holder attachments.
9. PCR tubes.
3 Methods
3.1 Closed Femoral Surgeries should be performed in a clean surgical suite or procedure
Fracture Model room on a surface disinfected with 70% ethanol or Quatricide (see
Note 1).
1. Record body weight of mouse prior to fracture (see Notes
2 and 3).
2. Anesthetize the mouse by placing it in a container with air flow
consisting of 2% isoflurane/oxygen mixture.
3. Once unresponsive to stimuli (e.g., pedal reflex), remove the
anesthetized mouse and place on a heating pad, coat eyes with
Puralube vet ointment, insert nose into nose cone with contin-
uous flow of 2% isoflurane/oxygen mixture, and administer
Burprenorphine SR (1 mg/kg subcutaneously) (see Note 4).
4. Shave the entire surgical leg and adjacent skin well with clip-
pers, wipe shaven limb with gauze soaked in chlorohexidine,
then gauze soaked in 70% isopropyl alcohol.
5. Transfer the mouse to the surgical table covered in sterile
drapes, place head inside a nose cone to deliver continuous
2% isoflurane/oxygen mixture in a supine position.
6. Cut drapes to expose a hole only large enough to slip the
surgically prepared limb through.
7. The surgeon holds the operated limb with the middle finger
positioned underneath the knee and the index finger resting on
the quadriceps muscle. The thumb is held against the tibia for
balance.
8. Using their other hand, the surgeon perforates the skin and
articular surface of the femoral condyle with a 25-gauge needle,
rotating (without applying pressure) gently in a single direction
until the medullary cavity is punctured (see Notes 5 and 6) (see
Fig. 1a).
9. The surgeon picks up a K-wire of an appropriate length (aver-
age 15–15.5 mm) using a hemostat.
210 Joseph L. Roberts et al.
Fig. 1 Creation of mid-diaphyseal femoral fracture. (a) Perforate the skin and articular surface of the femur
using a 25G needle to puncture the medullary canal. (b) Using a hemostat, a K-wire is inserted into the
medullary canal until it reaches the proximal end of the femur. (c) Excess K-wire is cut leaving 1–2 mm
exposed. (d) The exposed K-wire is bent to form a 90 angle using a hemostat, then (e) covered under the skin.
(f) The surgeon positions the pinned femur perpendicular to the blunt guillotine and pressure is applied to
create a mid-diaphyseal fracture
10. The needle is removed from the marrow cavity and the K-wire
is inserted into the medullary canal using a retrograde approach
until it reaches the proximal end of the femur (see Note 5).
Correct depth of insertion is felt with resistance upon meeting
cortical bone of the proximal femur (see Fig. 1b).
11. Excess K-wire is cut with wire cutters, leaving approximately
3–4 mm of K-wire that is bent to form a 90 angle, then
covered under the skin without suturing (see Fig. 1c–e).
12. The surgical assistant holds the mouse in a prone position,
while the surgeon positions the pinned femur perpendicular
to the blunt guillotine (see Fig. 1f).
13. With their free hand, the surgical assistant raises the lower stage
which applies pressure via three-point bending until an audible
break is noted.
14. Take an X-ray to confirm location, type, and severity of the
fracture.
15. The animal is placed in a clean cage that has food pellets located
at the bottom in addition to the normal food and water.
Generation of Murine Femoral Fracture 211
16. Monitor mice daily for the first 3 days after fracture to ensure
that they bear weight on their operated limbs and are eating
and drinking.
17. Take radiographs at intervals (e.g., 7, 14, 21 days post-
fracture) to ensure that the K-wire is maintaining fixation of
the fracture and track healing.
3.3 Torsion Testing Care must be taken to ensure the safety of both the individual
performing the test and those in the surrounding area when bio-
mechanical testing is performed. The ideal timepoint for torsion
212 Joseph L. Roberts et al.
Fig. 2 Microcomputed tomography (micro-CT) of fracture callus. (a) Manual contouring of the day 14 fracture
callus to exclude existing cortical bone. (b) Example of micro-CT reconstructions of a murine fracture callus at
day 14 post-fracture
Fig. 3 Potting and torsion mechanical testing of fractured femora. (a) The K-wire is carefully removed from the
fractured limb using forceps. (b) 1 cm2 aluminum cubes are positioned and (c) PMMA powder is added to the
cube until full (d). (e) PMMA liquid is added using a pipette. (f) The femur is carefully positioned inside of
PMMA using forceps and allowed to solidify. (g) The opposite end of the femur is then potted. The fully potted
femur is placed in a torsion machine (h) and torsion is applied until failure (i). (j) Typical data obtained from
torsion testing
9. Place the first specimen in the testing apparatus and tighten one
end into place (see Fig. 3h). Remove the gauze around the
bone.
10. Being careful not to damage the specimen, slowly tighten the
second end until locked. After the position is set, zero the
torque and run the test while recording a video. Record inner
and outer maximum diaphyseal diameters for all samples after
testing. Take pictures to confirm site of failure (see Fig. 3i).
11. Repeat this procedure for all samples.
12. Turn off all machinery.
13. Extract the data and analyze the torque curves to verify testing
accuracy (see Fig. 3j).
Category 0 1 2 3 4
Periosteal
and
No reaction Mild reaction Moderate reaction Marked reaction
Endosteal
Reaction
Heterogeneous Heterogeneous
with minimal with moderate
Callus No evidence of Confluent with the
mineralization mineralization and
Opacity mineralization cortices, uniform
and cortices well partially confluent
demarcated with the cortices
Complete
Minimal cortical
Partial cortical cortical union
Cortical union (3 cortical
All cortical edges union (1-2 cortical (no cortical
Remodeling No apparent edges visible)
seen but ill- edges visible) with edges visible)
and remodeling without
defined visible medullary with well
Bridging reformation of the
canal demarcated
medullary canal
medullary canal
Callus Opacity 1 2 3
Fig. 4 Radiographic scoring criteria and representative scores of three independent fractured femora
3.5 Histology The methodology outlined below describes general procedures for
histological fixation, sample preparation, sectioning, and histomor-
phometric quantification. Due to the wide variety of staining pro-
cedures, the experimenter is advised to consult relevant literature
specific to the staining procedures needed to visualize desired tissue
components (see Note 13).
3.5.1 Paraffin Embedding 1. After euthanasia, collect operative limbs and carefully remove
excessive soft tissue and muscle. Some soft tissue should be left
to maintain the structural integrity of the periosteum and the
fractured bone.
2. Fix bones in 10% neutral buffered formalin (20:1 fixative to
tissue ratio) at 4 C for 1 week.
3. Rinse fixed bones in PBS for 30 min on an orbital shaker.
4. Place bones in labeled embedding cassettes and submerge in a
beaker containing a sufficient volume of a decalcification agent
(e.g., 14% EDTA) (see Note 14). Place submerged bones on an
orbital shaker at room temperature. Change decalcification
solution every 2–3 days, for 2–3 weeks. Decalcification is com-
plete when bone is pliable and can be confirmed by X-ray.
5. Rinse decalcified bones twice by submerging in 1 PBS for
30 min with agitation.
6. To dehydrate bones, place cassettes in four separate changes of
70% ethanol and 95% ethanol for 20 min each. Followed by one
solution of 100% ethanol for 30 min, then overnight in 100%
ethanol with gentle shaking. The next day submerge the bones
in xylene for 10 min each for a total of 4 changes.
7. The bones are then placed in heated xylene:paraffin (1:1 ratio;
56 C) for 3 h, then allowed to solidify at room temperature
overnight. The following morning the xylene:paraffin contain-
ing the bones is melted, then the bones are placed in three
separate heated liquid paraffin (paraffin 1, 2, 3) solutions for
2 h each at 56 C.
8. Remove bones from cassettes and carefully retrieve the intra-
medullary pin from each fractured femur.
9. Turn-on embedding station to heat paraffin tank to 56 C and
to cool the cold plate.
10. Embed bones, with the femoral head pointing up, in hot
paraffin wax positioned in the center of a metallic mold on an
embedding station and allow the blocks to cool.
216 Joseph L. Roberts et al.
11. Remove any air bubbles that form in the center of the
wax mold.
12. Allow samples to cool on the embedding station for 30 min to
1 h before removing them from the metal molds and storing at
4 C or 20 C.
3.5.2 Paraffin Sectioning 1. Label all positively charged slides with specimen identification
(e.g., animal ID, genotype) and section number with a
chemical-resistant marker or pencil.
2. Turn on the hot water bath at least 30 min before sectioning.
3. Place a sharp microtome blade in the microtome.
4. Remove paraffin blocks from the refrigerator or 20 C
freezer and place on ice with paraffin side facing down toward
the ice.
5. Carefully position each block in the microtome and coarsely
section (15–20 μm) until the tissue is exposed.
6. Place the block on ice for 10–20 min.
7. Section the block at a setting of 5–7 μm until the mid-sagittal
plane is reached (see Note 15).
8. Carefully place sections on surface of the water in the water
bath to stretch the sections.
9. Allow paraffin sections to stand for 1–2 min in the water bath,
before carefully placing onto slides.
10. Repeat this procedure until the sample is fully sectioned or
until desired number of slides have been collected.
11. Place slides on a slide drying rack and allow slides to dry
overnight at room temperature.
3.5.3 Histological/IHC 1. Bake slides in an oven set at 56 C for 15–30 min or until
Staining paraffin is melted.
2. Rehydrate slides (generally 10 min in xylene, 5 min in xylene,
1 min in 100% ethanol, 1 min in 100% ethanol, 1 min in 95%
ethanol, 1 min in 95% ethanol, 5 min in deionized water).
3. Conduct histological staining/IHC procedure (differs depend-
ing on experiment: consult online literature) on rehydrated
slides (see Notes 16–18).
4. Dehydrate slides (generally 1 min in 95% ethanol, 1 min in
100% ethanol, 1 min in 100% ethanol, 1 min in xylene, 1 min in
xylene) (see Note 19).
5. Add appropriate mounting medium, carefully apply cover slip
and remove any bubbles by gently applying pressure to the
cover-slipped slide (see Note 20).
6. Allow slides to dry at room temperature overnight.
Generation of Murine Femoral Fracture 217
3.6 Gene Expression Special care must be taken to ensure that solutions and work area
are not contaminated with RNases/DNases, which can compro-
mise the validity of the experiment. It is recommended to spray and
wipe all laboratory benches, pipettes, and pipette tip boxes with
RNase Away prior to starting the experiment.
3.6.1 RNA Isolation from 1. Isolate fracture calluses following euthanasia by removing all
Whole Fracture Callus surrounding soft tissue and muscle, removing the pin, and
surrounding bone from the intact fracture callus. Wrap the
callus in labeled aluminum foil and flash freeze in liquid nitro-
gen. Store bones at 80 C until ready for further processing.
2. Homogenize calluses into a fine powder using either a tissue
homogenizer or a mortar and pestle, maintained at 80 C.
3. Place homogenized samples in a nuclease-free microcentrifuge
tube and add an appropriate volume of TRIzol. Incubate for
5 min at room temperature.
4. Add 0.2 mL of chloroform per 1 mL TRIzol reagent to the
samples and shake tubes vigorously by hand for 15 s.
5. Incubate samples at room temperature for 3 min and subse-
quently centrifuge them at 12,000 g for 15 min at 4 C.
6. Carefully isolate the upper clear aqueous phase and pipette it
into a new tube, taking care not to disturb the white interphase
or lower pink aqueous phase.
7. Add 0.5 mL of 100% isopropanol per 1 mL TRIzol used, mix
by pipetting, and incubate at room temperature for 10 min.
Centrifuge the mixture at 12,000 g for 10 min at 4 C.
8. Remove the supernatant from the tube, leaving only the RNA
pellet, which is washed in 1 mL 75% ethanol per 1 mL
TRIzol used.
9. Briefly vortex the mixture and centrifuge at 7500 g for 5 min
at 4 C.
218 Joseph L. Roberts et al.
10. Carefully aspirate the ethanol wash using a pipette and air-dry
the pellet for 10–15 min. It is important to prevent complete
drying of the pellet.
11. Depending on the size of the RNA pellet, resuspend with
appropriate volume of RNase-free water (approximately
10–30 μL).
12. Incubate the resuspended pellet in a heating block set at 65 C
for 5 min (or 55 C for 10 min).
13. Proceed with reverse transcription or store isolated RNA at
80 C for downstream applications.
3.6.2 RNA Isolation from 1. Cut fresh sections (7–10 μm) from paraffin-embedded frac-
Histological Sections tured femurs.
2. Carefully place sections on surface of the water in the water
bath to stretch the sections.
3. Allow paraffin ribbon sections to stand for 1–2 min in the bath,
before carefully placing onto slides.
4. Repeat this procedure until a total of 5–7 sections have been
collected.
5. Allow sections to dry on a drying rack that has been cleaned
with RNase-Away.
6. Spray microscope stage with RNase-Away.
7. Visualize fracture callus under microscope and using a sterile
10 μL pipette tip carefully scrape the callus (see Fig. 5). The
scraped tissue will adhere to the pipette tip via electrostatic
forces.
8. Place scraped tissue into a nuclease-free microcentrifuge tube.
9. Repeat until all slides have been scraped.
Fig. 5 Histological sections of a day 14 fracture callus before and after scraping the callus (red outline)
Generation of Murine Femoral Fracture 219
3.6.4 Quantification 1. Prepare a master mix for each gene of interest and
of Gene Expression housekeeping gene, containing forward primer, reverse primer,
and SYBR green mix (if SYBR green is used). Vortex tubes to
mix, and centrifuge briefly (see Note 23).
2. First add an appropriate volume of master mix to each well of
the PCR plate. Then add the cDNA obtained from each callus
to the corresponding wells of the PCR plate. There should be
at least two, but ideally three replicates per cDNA and gene of
interest. It is also recommended to include a no-template
control (NTC) which only contains master mix and no cDNA.
3. Seal plate with plastic cover tightly and centrifuge briefly.
4. Place plate into the RT-qPCR machine. Add labels to the
software and run PCR using the optimized conditions for the
sample of interest (see Note 24).
5. Record CT values and enter them into an excel spreadsheet for
data analysis.
6. Perform data analysis for obtained CT values, relating expres-
sion of genes of interest to housekeeping genes (see Note 25).
4 Notes
14. Many decalcification agents are available, and the most appro-
priate one is dependent on the desired downstream applica-
tion. 10%–14% EDTA is a slower decalcifier but works well for
most histological stains used to assess fracture healing. Acid-
decalcifiers work quicker but should not be used if there is an
interest in completing enzymatic staining like TRAP.
15. Occasionally the embedded bones contain residual mineral. If
this is observed, the block should be surface decalcified by
removing the block from the microtome and placing it face
down in an appropriate decalcification solution (e.g., 14%
EDTA) for 30–60 min. This will only remove mineral near
the surface, so additional surface decalcification may be
required as the paraffin block is further sectioned.
16. Different staining procedures require different materials and
dyes. To determine if additional materials are needed and to
familiarize the researcher with each specific staining procedure,
it is recommended to consult online literature before begin-
ning an experiment. Tissue processing can also affect tissue
processing. The staining time for each stain must be optimized
for each experiment.
17. The tissue of interest will determine the type of histological
staining used. Hematoxylin and Eosin is useful for gross mor-
phology and identifying callus bone, while Safranin O/Fast
Green and Alcian Blue/Hematoxylin staining is useful for
identifying cartilage.
18. IHC primary antibody dilutions should be verified in the
researcher’s lab with positive and negative controls prior to
any IHC staining.
19. Dehydration of stained slides should not be completed for
certain stains such as TRAP because it can weaken staining
intensity. Slides with these stains should be coverslipped using
an aqueous mounting medium and sealed using fingernail
polish by painting around the edge of the coverslip.
20. If the applied pressure does not remove air bubbles from under
the coverslip, soak the slide in xylene until the coverslip falls off.
Add mounting medium to the exposed tissue and apply a new
coverslip.
21. There are different techniques for quantifying the number of
cells after IHC staining. Some studies set a baseline level of
staining intensity and count positive cells above that level.
Other studies count all cells with moderate staining and above.
22. Ideally RNA should have a 260/280 ratio of 1.8–2, with a
260/280 ratio of 2 indicating pure RNA. An abnormal
260/280 ratio suggests contamination with ethanol, phenol,
222 Joseph L. Roberts et al.
Acknowledgements
References
1. Bonnarens F, Einhorn TA (1984) Production of 5. Marturano JE, Cleveland BC, Byrne MA,
a standard closed fracture in laboratory animal O’Connell SL, Wixted JJ, Billiar KL (2008) An
bone. J Orthop Res 2(1):97–101. https://doi. improved murine femur fracture device for bone
org/10.1002/jor.1100020115 healing studies. J Biomech 41(6):1222–1228.
2. Clifton KB, Paglia DN, Soung DY, Wolf JM, https://doi.org/10.1016/j.jbiomech.2008.01.
Moss IL, Drissi H (2014) Effects of Wnt5a hap- 029
loinsufficiency on bone repair. J Orthop Trauma 6. Cottrell J, O’Connor JP (2010) Effect of
28(8):e191–e197. https://doi.org/10.1097/ non-steroidal anti-inflammatory drugs on bone
bot.0000000000000041 healing. Pharmaceuticals (Basel) 3
3. Walia B, Lingenheld E, Duong L, Sanjay A, (5):1668–1693. https://doi.org/10.3390/
Drissi H (2018) A novel role for cathepsin K in ph3051668
periosteal osteoclast precursors during fracture 7. Bustin SA, Benes V, Garson JA, Hellemans J,
repair. Ann N Y Acad Sci 1415(1):57–68. Huggett J, Kubista M, Mueller R, Nolan T,
https://doi.org/10.1111/nyas.13629 Pfaffl MW, Shipley GL, Vandesompele J,
4. Hussein AI, Mancini C, Lybrand KE, Cooke Wittwer CT (2009) The MIQE guidelines: min-
ME, Matheny HE, Hogue BL, Tornetta P 3rd, imum information for publication of quantita-
Gerstenfeld LC (2018) Serum proteomic assess- tive real-time PCR experiments. Clin Chem 55
ment of the progression of fracture healing. J (4):611–622. https://doi.org/10.1373/
Orthop Res 36(4):1153–1163. https://doi. clinchem.2008.112797
org/10.1002/jor.23754
Chapter 14
Abstract
The surgical model of destabilization of the medial meniscus (DMM) has become a gold standard for
studying the onset and progression of post-traumatic osteoarthritis (OA). The DMM model mimics clinical
meniscal injury, a known predisposing factor for the development of human OA, and permits the study of
structural and biological changes over the course of the disease. In addition, when applied to genetically
modified or engineered mouse models, this surgical procedure permits dissection of the relative contribu-
tion of a given gene to OA initiation and/or progression. This chapter describes the requirements for the
surgical induction of OA in mouse models, and provides guidelines and tools for the subsequent histologi-
cal, immunohistochemical, and molecular analyses. Methods for the assessment of the contributions of
selected genes in genetically modified strains are also provided.
Key words Surgical model, Histology, Immunohistochemistry, RNA and DNA extraction
1 Introduction
Kirsty L. Culley, Purva Singh, Mary B. Goldring, and Miguel Otero contributed equally to this work.
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_14, © Springer Science+Business Media, LLC, part of Springer Nature 2021
223
224 Kirsty L. Culley et al.
2 Materials
2.4 Surgical All surgical tools must be sterilized prior to surgery using a PRI
Reagents MUS Sterilizer (LBR Scientific Inc., East Rutherford, NJ).
and Equipment
1. Zeiss surgical microscope, Super Lux 40.
DMM Mouse Model of Post-traumatic Osteoarthritis 227
2.5 Mouse Housing Housing of mice at 5 to 6 mice per cage in large cages, according to
NIH and IACUC guidelines, at least 2 weeks prior to surgery is
recommended to ensure that the mice are accustomed to one
another and will maintain a level of activity that will induce OA
postoperatively.
1. Thoren Ventilated Rack (Thoren Caging Systems Inc.,
Hazelton, PA).
2. Thoren weaning cages (model #2, polycarbonate, dimensions:
L12.125 W12.125 H5.625, Thoren Caging Systems Inc.,
Hazelton, PA).
3. Polished stainless steel wire for mice cages (Thoren Caging
Systems Inc., Hazelton, PA).
4. Nestpacks, Betachip, 100gr Bedding (Fisher and Son, NJ).
5. LabDiet 5053 irradiated, PicoLab Rodent Diet 20 (Fisher and
Son, NJ).
6. DietGel Recovery, Purified Dietary Supplement for Laboratory
Rodents (Fisher and Son, NJ).
7. Mouse Tunnels (BioServ Frenchtown, NJ).
228 Kirsty L. Culley et al.
2.6.2 Tissue 1. 20% Sodium citrate dihydrate: Dissolve 100 g of sodium citrate
Decalcification dihydrate in 350 mL of 1 PBS; add dH2O to a final volume of
500 mL.
2. 45% Formic acid: To prepare 500 mL dilute 225 mL of 95%
formic acid in 275 mL of dH2O.
3. 2 N Sodium hydroxide: Dissolve 4 g of NaOH pellets in 40 mL
dH2O; make to a final volume of 50 mL
4. 10% EDTA: Add 100 g of EDTA to 850 mL dH2O while
stirring. Adjust the final volume to 1 L with dH2O and pH to
7.4 using sodium hydroxide pellets.
5. Ammonium oxalate monohydrate (AO): Weigh 10 g of AO and
add to 150–200 mL of dH2O while stirring. Heat gently to
dissolve all the ammonium sulfate and continue to add AO
until the solution becomes saturated. Precipitates of ammo-
nium sulfate crystals should form and will indicate that the
solution is fully saturated. Allow the solution to cool to room
temperature before use.
2.8 Immuno- IHC and IF conditions (e.g., retrieval method, or primary and
histochemistry (IHC) secondary antibody concentration) may vary and require optimiza-
and Immuno- tion. The provided protocols are guidelines, and have been opti-
fluorescence (IF) mized for the specified antibodies utilizing the reagents indicated.
1. Xylene.
2. Ethanol (EtOH).
3. 10 Phosphate-buffered solution (PBS).
4. 2 mg/mL Hyaluronidase.
5. 3% H2O2.
6. Humid Chamber (BD Falcon BioDish XL Nontreated Surface,
BD Biosciences Discovery Labware, Bedford, MA).
230 Kirsty L. Culley et al.
2.9 RNA and DNA 1. Sterile phosphate-buffered saline (PBS) (Cell culture stan-
Extraction for Gene dard); ice-cold for dissecting mouse knees.
Expression and DNA 2. Dissecting scissors.
Methylation Analyses
3. Straight forceps (Hampton Research, Aliso Viejo, CA).
of Mouse Articular
Cartilage
4. Curved tipped forceps (Hampton Research, Aliso Viejo, CA).
5. Two star quality micro scissors 300 blade.
2.9.1 Isolation
of Articular Cartilage 6. Stereo zoom microscope.
7. Microscope stand with focus mount.
8. Nova 2000 fiber optic illuminator and optic guides (Nikon,
Melville, NY).
9. 2 Feather Scalpel handle # 3.
10. Surgical Blades; size 11 and size 15.
11. RNAlater (Ambion Life technologies, Grand Island, NY).
12. DNase/RNase-free Eppendorf tubes.
13. Filter pipette tips (nuclease-free).
2.9.2 Total RNA Isolation 1. TRIzol® Reagent (Invitrogen, Grand Island, NY).
from Cartilage Using 2. QIAshredder (Qiagen Inc., Valencia, CA).
a Modified mirVana™
3. Chloroform:Isoamyl alcohol (IAA) 24:1.
Protocol
DMM Mouse Model of Post-traumatic Osteoarthritis 231
4. Phenol:chloroform:IAA, 125:24:1.
5. Molecular-grade ethanol.
6. MirVana miRNA Isolation Kit, without phenol (Ambion Life
Technologies, Grand Island, NY).
7. DNA-free Kit (Ambion Life Technologies, Grand Island, NY).
8. 3 M Sodium acetate pH 5.5, nuclease-free.
9. UltraPure Glycogen, nuclease-free.
10. Nuclease-free water.
11. Handheld homogenizer (VWR Pellet Mixer) (VWR Interna-
tional, Radnor, PA).
12. Nuclease-free pestles (Argos Technologies, Elgin, IL).
13. NanoDrop 2000 (ThermoFisher Scientific Inc., Rockford, IL).
14. Bioanalyzer (Model 2100) (Agilent Technologies, Santa
Clara, CA).
2.9.3 Total RNA isolation 1. TRIzol® Reagent (Invitrogen, Grand Island, NY).
from Cartilage Using 2. QIAshredder (Qiagen Inc., Valencia, CA).
a Modified Rneasy® Mini
3. Chloroform:Isoamyl alcohol (IAA) 24:1 (Sigma Aldrich,
Protocol
St. Louis, MO).
4. Molecular-grade ethanol (American Bioanalytical Inc.,
Natick, MA).
5. Rneasy mini RNA Isolation kit (Qiagen Inc., Valencia, CA).
6. RNase-Free DNase set (Qiagen Inc., Valencia, CA).
7. TissueLyser II (Qiagen Inc., Valencia, CA).
8. 2.0 mL Eppendorf safe-Lock tubes (Eppendorf,
Hauppauge, NY).
9. Stainless steel beads, 5 mm (Qiagen Inc., Valencia, CA).
10. TissueLyser single bead dispenser, 5 mm (Qiagen Inc.,
Valencia, CA).
11. Handheld homogenizer (VWR Pellet Mixer) (VWR Interna-
tional, Radnor, PA).
12. Nuclease-free pestles (Argos Technologies, Elgin, IL).
13. NanoDrop 2000 (ThermoFisher Scientific Inc., Rockford, IL).
14. Bioanalyzer (Model 2100) (Agilent Technologies, Santa
Clara, CA).
3 Methods
3.2 Surgical The following steps describe DMM surgeries (see Note 3) per-
Resection of Mouse formed in 12-week-old male mice, as described previously by
Knee Joints Glasson et al. [37]. Age- and sex-dependent changes in OA severity,
including associated gene expression changes, have been reported
and should be considered when designing experiments
[11, 38]. Briefly, unilateral joint instability is induced by microsur-
gical transection of the medial meniscotibial ligament (MMTL),
234 Kirsty L. Culley et al.
which anchors the medial meniscus to the tibial plateau (see Note
4). DMM surgery is completed in the right knee, leaving the left
knee as a non-operated control or a sham-operated control, in
which the meniscotibial ligament is localized but not transected
(see Note 5).
1. Administer to mice subcutaneously 2 cc of lactated Ringer
solution prior to anesthesia.
2. Administer medications by intraperitoneal injection into the
right lower abdominal quadrant, with the anterior body tilted
down, via a ½ cc tuberculin syringe with a 27-gauge ½-in
needle. This single dose provides 15–20 minof surgical anes-
thesia. If necessary, anesthesia may be prolonged by adminis-
tration of isoflurane gas via a nose cone.
3. Use electrical hair clippers to remove hair from the surgical site.
Prepare the site by scrubbing twice with a 4% chlorhexidine
surgical brush/sponge, and then wiping with 70% isopropyl
alcohol.
4. Perform arthrotomy to expose the femoral tibial joint. Make a
longitudinal incision of approximately 5 mm over the distal
patellar tendon to the proximal tibial plateau.
5. Incise the joint capsule immediately medial to the patellar
tendon with a size 15 blade, and gently lift the tendon with
45 degree Dumont tweezers to allow access of the Tyrell hook
by technician. Use the Tyrell hook to move the tendon to one
side to allow access to the joint compartment.
6. Optional (see Note 6): Perform blunt dissection of the fat pad
over the intercondylar area using 45 degree Dumont tweezer
to expose the medial meniscotibial ligament (MMTL). Control
mild hemorrhaging from the fat pad by applying pressure with
Q-tips.
7. Identify the MMTL running from the cranial horn of the
medial meniscus laterally onto the anterior tibial plateau.
Take care to identify and avoid the lateral meniscotibial liga-
ment (LMTL), which is posterior and has fibers running in a
similar direction.
8. Section the MMTL with a size 11 blade, with the blade
directed proximo-laterally to destabilize the medial meniscus.
Use the 45-degree Dumont tweezer to check if the MMTL is
fully transected and the medial meniscus destabilized.
9. Suture close the joint capsule and subcutaneous layer with
Coated Vicryl Suture (8–0) and close the skin by application
of VETClose Surgical Adhesive.
10. Immediately after anesthetic recovery administer an initial dose
of buprenorphine (Buprenex 0.3 mg/mL) at 0.05 mg/kg
subcutaneously. Place mice in cages that are maintained on
DMM Mouse Model of Post-traumatic Osteoarthritis 235
3.3 Histological The animals are sacrificed (according to IACUC guidelines) (see
Assessment of OA Note 10) at appropriate time points post-DMM surgery and the
Pathology knee joints collected for histological assessment to determine the
effects of time or genotype on experimental OA. The pathology is
assessed using a modified Mankin scoring system recommended by
OARSI [40] based on Safranin O-stained histological sections.
Meniscus, subchondral bone, synovial changes, and osteophyte
formation may also be examined in the same sections to evaluate
the overall condition of the joints. Guidelines and examples for
sample processing in order to complete histological evaluation are
provided in the following sections.
236 Kirsty L. Culley et al.
Fig. 1 Trimming mouse knee joints for embedding in paraffin. Once the leg has been dissected (a) it is
important to remove as much muscle (M) as possible (b and c) to prevent the knee from bending during
processing. The tibia (T) and femur (F) should be cut to approximately 0.5 in with the patellar tendon
(PT) located in the center of the specimen
3.3.1 Fixation 1. Immediately after sacrifice dissect knees for histological analysis
and remove skin and excess muscle with dissection scissors.
Trim the femur and tibia so that they are ~0.5-in in length
(see Fig. 1), then place the dissected knees in individual falcon
tubes or biopsy cassettes for tissue fixation in 4% PFA. It is
advisable that the volume of PFA (or other fixative) is at least
15–20 the volume of tissue.
2. Fix samples for the desired length of time depending on your
antibody/staining to be completed downstream (duration of
fixation is based on previous standardized histology or immu-
nohistochemistry (IHC)/immunofluorescence (IF) and can be
as short as 6 h or as long as overnight). During fixation, place
samples on a rocker or shaker.
3. Following fixation, wash the samples with dH2O for 1 h at
room temperature on a shaker, changing the dH2O every
15 min.
DMM Mouse Model of Post-traumatic Osteoarthritis 237
3.3.2 Decalcification There are several available methods for decalcification and, while
the appropriate method should be selected based on the
subsequent histological/immunohistochemical analyses, it should
remain consistent within each experimental model studied.
Depending on the method selected, the time required for decalcifi-
cation and subsequent retrieval method for immunostaining will
change. In this section we will detail two methods for decalcifica-
tion (EDTA and formic acid:sodium citrate), both of which are
suitable for reliable Safranin O/Fast green staining and scoring.
Other available methods include 5–10% nitric acid or 10% HCl.
Decalcification Test
1. Take 1 mL of the decalcification solution from the samples that
you are decalcifying and add 5 mL of saturated AO solution.
Mix well, incubate for 30 min at room temperature, and check
for a white precipitate by holding against a black background.
2. If a white precipitate forms, remove the decalcification solution
from the samples and add fresh decalcification solution. Incu-
bate the samples for 2–3 more days at room temperature on a
shaker. Decalcification is complete only if no white precipitate
is formed.
3. Once decalcification is complete, wash samples in dH2O for
6 h. Change the water every 30 min or place samples (in biopsy
cassettes) under a gentle flow of clean running water.
238 Kirsty L. Culley et al.
3.3.5 Histological 1. Deparaffinize in Xylene 2 for 8 min each (in the fume hood).
Staining 2. Rehydrate in an ethanol series: 100% EtOH (2, 5 min each);
95% EtOH (5 min); 85% EtOH (4 min); and 70% EtOH
(4 min).
3. Incubate slides in filtered hematoxylin for 30 s.
4. Rinse slides in tap water 3 times for 5 min each until water is
clear.
5. Incubate in Scott’s Buffer for 2 min.
6. Rinse slides in tap water 3 times for 5 min each.
7. Incubate in 0.2% Fast green for 4 min.
8. Rinse quickly in 1% acetic acid solution (dip 3 times).
9. Rinse quickly in dH2O.
10. Stain slides in 0.5% Safranin O for 5 min.
11. Rinse slides in 95% EtOH (dip 3 times).
12. Rinse slides in 100% EtOH (dip 3 times).
13. Incubate slides in 100% EtOH for approximately 2 min. If the
ethanol is still pink after 2 min, place the slides in a fresh
ethanol bath for another minute.
14. Incubate slides in Xylene for 3 min followed by fresh Xylene for
10 min (in the fume hood).
15. Mount sections with Vectamount medium and SuperSlip
CoverGlass.
240 Kirsty L. Culley et al.
Fig. 4 Representative photomicrograph of Safranin O-stained coronal section (a) and schematic (b) detailing
the anatomical structures of the knee. Three quadrants are visualized: medial femoral condyle (MFC), medial
tibial plateau (MTP), and lateral tibial plateau (LTP). The growth plate (GP) is located at the proximal end of tibia
(T) and fibula (F). Skeletal muscle (SM) can be seen surrounding the knee. The medial meniscus (M) can be
identified between the MTP and MFC
Table 1
Mouse histological scoring system recommended by OARSI [40]
Fig. 5 Schematic (a) and Safranin O-stained coronal section (b). Representation
of the four quadrants of the knee, which can be graded post-DMM surgery,
including the medial tibial plateau (MTP), medial femoral head (MFH), lateral
tibial plateau (LTP), and lateral femoral head (LFH). The majority of cartilage
damage will be observed on the medial side of the joint
242 Kirsty L. Culley et al.
3.3.7 Statistical Analysis 1. Perform power analysis for the number of mice required per
group. Using a 2-point difference as the definition of a statisti-
cal difference, group sample sizes of 7 in each group achieve
85.8% power to detect a 2-point difference in the modified
mouse score after adjusting for nonparametric Mann-Whitney
U test and setting significance to 0.017 (to adjust for multiple
comparisons). To account for attrition due to surgery or
anesthesia-related deaths, the total number per group is
adjusted to 10.
2. For histological scoring, Mann-Whitney U tests and Kruskal-
Wallis with Dunn’s post-analysis (Prism® Graph-Pad) are
needed as the nonparametric equivalents for the independent
samples t-test and one-way ANOVA (see Note 13).
3.3.8 Osteophyte Scoring The progressive erosion of the cartilage observed post-DMM sur-
gery is accompanied by osteophyte formation and development.
The DMM model is a valuable tool to assess the contribution of
certain genes to the formation and development of osteophytes
during OA. Indeed, Loeser and colleagues [41] described in detail
the relationship between osteophyte development in OA and genes
involved in morphogenesis, differentiation, and development. A
histological scoring system was developed to score both osteophyte
size and maturity, as described by Little et al. [15] (see Tables 2 and
3). This system complements the modified Mankin scoring system
recommended by OARSI [40] for grading cartilage degradation,
and follows the same scoring guidelines for statistical analysis.
Briefly, osteophyte scoring is performed on the same slides/sec-
tions used to assess cartilage degradation and obtain an OA score
for each knee. The osteophyte size is highly dependent on cartilage
destruction, with joints exhibiting a high OA score often also
having large mature osteophytes (see Fig. 7). In addition, osteo-
phytes are often localized close to areas of cartilage degradation,
and thus are predominately located on the medial side of the tibial
plateau. Initially, osteophytes are composed of cartilage (see Fig. 7a,
e), and then transition to a combination of both cartilage and bone
(b) before they progress to a boney appearance as the osteophyte
matures (see Fig. 7b, f).
Fig. 6 Safranin O-stained coronal sections taken from the knee representing each OA grade of the modified
Mankin scoring system recommended by OARSI [40]. The grade presented represents the damage observed
on the medial tibial plateau. Graphs can be constructed based on scoring of multiple knee joints using Table 1
244 Kirsty L. Culley et al.
Table 2
Assessment of osteophyte size in mouse knee joints
Table 3
Assessment of osteophyte maturity in mouse knee joints
Fig. 7 Representative photomicrograph images of Safranin O-stained coronal sections taken from the knee
(wild type mouse at 8 weeks post-DMM surgery) with yellow outlines indicating osteophytes. The composition
of the osteophyte is dependent on its maturity: Early osteophytes are composed mainly of cartilage (a), and
then transition to a combination of both cartilage and bone (b) before they develop into mature osteophytes
that consist mainly of bone (c). The size and maturity of the osteophyte often correlates with the severity of OA
found in the joint post surgery, with small osteophytes consisting mainly of cartilage observed in joints with a
low histological score (d), and mature osteophytes consisting mainly of bone observed in highly damaged
joints (e). Osteophyte formation is modest in wild type mice subjected to DMM surgery, but may be enhanced
by certain gene modifications. Graphs can be constructed based on scoring of multiple knee joints using
Tables 2 and 3
3.5 Cartilage For nucleic acid isolation, cartilage is dissected from the femoral
Microdissection heads and tibial plateaus of the knee and homogenized in TRIzol
for RNA and DNA (Invitrogen). Below (see Subheadings 3.6 and 3.7) we describe
Extraction for Gene methods to isolate total RNA using mirVana™ miRNA isolation
Expression and DNA kits (Ambion), for work with mRNA and miRNA, or the RNeasy®
Methylation Analyses Mini (Qiagen) kits, following the manufacturer’s instructions with
additional modifications. Following these methods, an average of
50–80 ng of high-quality total RNA is obtained from the articular
cartilage isolated from one knee joint. This should serve as a guide
for the number of mice required to achieve a sufficient amount of
RNA for gene expression analyses.
For RNA isolation, it is critical to place each leg in RNAlater
(steps 1–5) as early as possible to obtain good RNA integrity; thus,
one person completes steps 1–7 and a second person completes
steps 8–12.
Person 1
1. Sacrifice mouse and immediately remove hind legs.
2. Use dissection scissors to quickly remove as much soft tissue
from the legs as possible.
3. Place the legs in separate Petri dishes and cover with ice-cold
1 PBS.
4. While working on one leg, keep the other leg in 1 PBS on ice,
complete steps 5–7 on the other leg.
5. Dissect remaining muscle and tendon with a scalpel and size
15 blade under the microscope. Remove dissected soft tissues
from the Petri dish, as these tissues may be sources of RNases.
6. Transfer cleaned leg to a fresh Petri dish and wash with ice-cold
1 PBS.
7. Transfer the washed leg into a fresh small Petri dish and cover
with RNAlater. Keep the leg on ice in RNAlater until ready to
complete steps 8–13.
Person 2: Keep legs in RNAlater at all times during the following
steps in the Petri dish placed under the dissection microscope, use a
scalpel with size 11 blade to separate the tibia and femur and expose
the articular surfaces.
248 Kirsty L. Culley et al.
3.6 RNA Isolation The following procedure is performed according to the manufac-
from Cartilage Using turer’s instructions (Ambion) (see Note 18), except that additional
a Modified mirVana phenol:chloroform steps have been introduced to help improve the
miRNA Isolation Kit 260/280 values of the RNA isolated. To obtain good RNA integ-
(Ambion) Protocol rity (RIN) values, all reagents, pipet tips, and Eppendorf tubes must
be certified nuclease-free. In addition, isolation should be com-
pleted on a clean work space treated with RNAse Zap (Ambion).
1. Place the Eppendorf tube containing cartilage sample on ice
and add 500 μL of TRIzol to the tube. Homogenize the
cartilage sample in TRIzol using a 1.5-mL RNAse-free pestle
DMM Mouse Model of Post-traumatic Osteoarthritis 249
3.6.1 Ethanol This precipitation step can be completed to concentrate the RNA if
Precipitation the concentration obtained above is not enough for required
analyses.
1. On ice, add 94 μL of ice-cold water to each 50 μL of RNA
sample (volumes should now be about 144 μL).
2. Add 16 μL of cold 3 M NaOAc, pH 5.0, buffer to each sample
to make give a concentration of around 0.3 M. Mix by pipet-
ting up and down. Final volume will be 160 μL.
DMM Mouse Model of Post-traumatic Osteoarthritis 251
3.7 RNA Isolation This modified RNA isolation protocol combines the TRIzol® RNA
from Cartilage Using extraction method with the Qiagen Rneasy® mini kit protocol. The
a Modified RNeasy protocol modifications are intended to improve RNA purity
Mini RNA Isolation Kit (A260/280), while preserving RNA integrity (RIN > 7). All
(Qiagen) Protocol reagents, pipet tips, and Eppendorf tubes must be certified
nuclease-free. In addition, isolation should be completed on a
clean workspace treated with RNAse Zap (Ambion).
1. Label 2 mL microcentrifuge tubes compatible with TissueLy-
ser and dispense 1 stainless steel bead per tube.
2. Place microdissected cartilage (see Note 22) in the 2 mL micro-
centrifuge tube along with 700 μL of Trizol and homogenize
the cartilage at 300 Hz for 2 min using the TissueLyser II. If
needed, repeat the step again once. See Note 23 for subsequent
homogenization.
(a) Alternatively, if no TissueLyser is available, homogenize
the cartilage using 1.5 mL RNAse-free pestle and hand-
held homogenizer, for approximately 5–10 min on ice, as
described in Subheading 3.6.
3. Transfer the homogenized sample in 700 μL of TRIzol to a
QIAshredder to further homogenize and disrupt cells: centri-
fuge for 2 min at 16,000 g.
252 Kirsty L. Culley et al.
3.8 DNA Isolation The Gentra® Puregene® DNA isolation protocol has been modified
from Cartilage Using to increase DNA yield and purity. All reagents, pipet tips, and
a Modified Gentra Eppendorf tubes must be certified nuclease-free, and all isolation
Puregene DNA steps should be completed in a clean workspace.
Isolation Kit (Qiagen) 1. Place freshly isolated or frozen knee cartilage in 1.5 mL
Protocol nuclease-free Eppendorf tube (see Note 22 for comment
regarding required amount of starting material).
2. Homogenize the cartilage using 1.5 mL pestle and handheld
motorized homogenizer.
3. Add 300 μL of cell lysis solution to the grounded cartilage, and
heat the sample at 65 C for 30 min.
4. Add 1.5 μL of proteinase K to the sample tube, shake, and
incubate at 55 C for 1 h. If cartilage is not completely dis-
solved, continue heating at 55 C for up to 3 hrs. Cartilage
must be completely digested before moving to the next steps.
5. Add 1.5 μL of RNase A solution, mix thoroughly by inverting
the tube, and incubate at 37 C for 45 min.
6. Quickly transfer the sample to ice and allow to cool for 1 min.
7. Add 100 μL of protein precipitation solution and vortex vigor-
ously for 20 s to mix, incubate on ice for 5 min.
8. Centrifuge sample tube for 3 min at 13000 g. A visible white-
colored precipitated protein pellet is formed at the bottom of
the tube.
254 Kirsty L. Culley et al.
4 Notes
allowing the mice time for complete gene ablation and recovery
from tamoxifen treatment before anesthesia. To observe the
effects of gene modification on OA development, comparisons
should be done in tamoxifen- versus vehicle-treated
littermates.
2. Different doxycycline concentrations have been used success-
fully, with no reported adverse effect [35, 36], but with differ-
ences in the time required for transgene activation to occur due
to the varying amount of time required for doxycycline to clear
from the mouse system. The concentrations achieved by the
doses administered are not sufficient to inhibit collagenase
activity. Adequate controls have to be used for comparison
after the DMM surgery [36].
3. All procedures must be approved by the Institutional Animal
Care and Use Committee (IACUC). Before live animals are
used, all personnel must obtain CLAS orientation and training,
including demonstration to veterinarians the surgical proce-
dure on cadaveric mice.
4. See Glasson et al. [37] for excellent photographs and sche-
matics to guide the surgery.
5. It is recommended that one surgeon perform all surgeries
involving each genetically modified mouse strain to reduce
variability. Comparisons between wild type and knockout or
transgenic strains are completed using littermates.
Non-operated or sham controls need to be checked to ensure
that there is no nonspecific effect of the genetic modification.
6. The fat pad may also be left in place and merely transected to
allow access and visualization of the MMTL.
7. Significant postoperative pain and debility are not anticipated.
Mice are monitored postoperatively by the Veterinary Staff, and
if there is evidence of pain, suffering, or illness, analgesia
and/or other treatment will be administered at their discretion.
Particular attention will be paid to ensure that animals are
ambulating normally after the procedure. For further informa-
tion on analgesic dugs, please see reviews [39, 42].
8. Since only male mice are used in the DMM model because of
the protection by estrogen in females [37], it is necessary to
establish the following procedures to avoid aggressive behavior.
Mice are housed together prior to surgery (completed as soon
as possible post weaning if males are not from the same litter),
and the same mice are housed together post-surgery in a cage
containing some of their original dirty bedding. Administra-
tion of analgesics is continued according to the protocol. Sur-
gical mice are monitored for the first few hours after surgery
and immediately the next morning. If a dominant male is
noticed, it is immediately separated from the group. The
256 Kirsty L. Culley et al.
remaining mice in the cage are monitored daily for fighting and
any new emerging dominant male is removed. A minimum of
3 mice must remain housed together post surgery to encourage
the level of activity required to promote OA initiation, devel-
opment, and progression.
9. Immediately after surgery, place one mouse tunnel per cage to
provide enrichment and promote the level of activity required
to promote OA initiation, development, and progression.
10. For histology, mice may be sacrificed by CO2 inhalation. For
RNA extraction, if a CO2 tank is not immediately available,
cervical dislocation may be required to avoid rigor mortis
before tissues can be dissected from the joints and placed in
extraction buffer.
11. Alternatively to test if the samples are decalcified, a needle can
be passed through the bone of the sample. If no resistance is
felt, the sample can be considered decalcified.
12. During sectioning check carefully every fifth slides to ensure
the knee is correctly orientated in the paraffin. The lateral
femoral condyle will usually appear in the first sections. Refer
to Figs. 4 and 5 to gain a good concept of the proper knee
orientation. If after 15 to 20 collected slides all four quadrants
are not identified, discard the knee.
13. Previous studies by Glasson et al. [37] found that the mean
maximum histological scores at 4 weeks postoperatively for the
unoperated, sham surgery, DMM, and ACLT groups, respec-
tively, were (S.E.M.) 1.1 0.1, 1.0 0.3, 3.7 1.5, and
4.3 0.4; and at 8 weeks the mean maximum scores were
1.2 0.3,1.2 0.2, 4.1 0.3, and 5.0 0.4.
14. The conditions for immunohistochemistry and immunofluo-
rescence (e.g., antigen retrieval method, primary and second-
ary antibody concentration) may vary and require
optimization. The provided protocols are guidelines and have
been optimized for applying the specified antibodies to mouse
knee joints utilizing the reagents indicated.
15. Hyaluronidase treatment is just one of many available antigen
retrieval methods. In general, the retrieval method depends
upon the antibody selected and the tissue processing, and
therefore requires optimization for the antibody used for
detection of specific antigen epitopes in a given tissue. It is
advisable to perform optimization steps comparing conditions
without antigen retrieval with one or two antigen retrieval
methods. Retrieval methods include:
(a) Heat retrieval in a sodium citrate buffer pH 6.0 (20 min at
95 C).
(b) Treatment with hyaluronidase (2 mg/mL 30 min at
37 C).
DMM Mouse Model of Post-traumatic Osteoarthritis 257
Acknowledgements
References
1. Poulet B, Ulici V, Stone TC, Pead M, and down-regulated in experimental osteoar-
Gburcik V, Constantinou E, Palmer DB, thritis. J Biol Chem 286(43):37758–37767.
Beier F, Timmons JA, Pitsillides AA (2012) https://doi.org/10.1074/jbc.M111.248039
Time-series transcriptional profiling yields 9. Yasuhara R, Ohta Y, Yuasa T, Kondo N,
new perspectives on susceptibility to murine Hoang T, Addya S, Fortina P, Pacifici M,
osteoarthritis. Arthritis Rheum 64 Iwamoto M, Enomoto-Iwamoto M (2011)
(10):3256–3266. https://doi.org/10.1002/ Roles of beta-catenin signaling in phenotypic
art.34572 expression and proliferation of articular carti-
2. Poulet B, Hamilton RW, Shefelbine S, Pitsil- lage superficial zone cells. Lab Investig 91
lides AA (2011) Characterizing a novel and (12):1739–1752. https://doi.org/10.1038/
adjustable noninvasive murine joint loading labinvest.2011.144
model. Arthritis Rheum 63(1):137–147. 10. Lodewyckx L, Cailotto F, Thysen S, Luyten FP,
https://doi.org/10.1002/art.27765 Lories RJ (2012) Tight regulation of wingless-
3. Ko FC, Dragomir C, Plumb DA, Goldring SR, type signaling in the articular cartilage - sub-
Wright TM, Goldring MB, van der Meulen chondral bone biomechanical unit: transcrip-
MC (2013) In vivo cyclic compression causes tomics in Frzb-knockout mice. Arthritis Res
cartilage degeneration and subchondral bone Ther 14(1):R16. https://doi.org/10.1186/
changes in mouse tibiae. Arthritis Rheum 65 ar3695
(6):1569–1578. https://doi.org/10.1002/ 11. Loeser RF, Olex AL, McNulty MA, Carlson
art.37906 CS, Callahan MF, Ferguson CM, Chou J,
4. Sato T, Konomi K, Yamasaki S, Aratani S, Leng X, Fetrow JS (2012) Microarray analysis
Tsuchimochi K, Yokouchi M, Masuko-Hongo- reveals age-related differences in gene expres-
K, Yagishita N, Nakamura H, Komiya S, sion during the development of osteoarthritis
Beppu M, Aoki H, Nishioka K, Nakajima T in mice. Arthritis Rheum 64(3):705–717.
(2006) Comparative analysis of gene expres- https://doi.org/10.1002/art.33388
sion profiles in intact and damaged regions of 12. Nuka S, Zhou W, Henry SP, Gendron CM,
human osteoarthritic cartilage. Arthritis Schultz JB, Shinomura T, Johnson J, Wang Y,
Rheum 54(3):808–817 Keene DR, Ramirez-Solis R, Behringer RR,
5. Aigner T, Fundel K, Saas J, Gebhard PM, Young MF, Hook M (2010) Phenotypic char-
Haag J, Weiss T, Zien A, Obermayr F, acterization of epiphycan-deficient and epiphy-
Zimmer R, Bartnik E (2006) Large-scale gene can/biglycan double-deficient mice.
expression profiling reveals major pathogenetic Osteoarthritis Cartilage 18(1):88–96.
pathways of cartilage degeneration in osteoar- https://doi.org/10.1016/j.joca.2009.11.006
thritis. Arthritis Rheum 54(11):3533–3544 13. Glasson SS, Askew R, Sheppard B, Carito B,
6. Glasson SS (2007) In vivo osteoarthritis target Blanchet T, Ma HL, Flannery CR, Peluso D,
validation utilizing genetically-modified mice. Kanki K, Yang Z, Majumdar MK, Morris EA
Curr Drug Targets 8(2):367–376 (2005) Deletion of active ADAMTS5 prevents
7. Little CB, Fosang AJ (2010) Is cartilage matrix cartilage degradation in a murine model of
breakdown an appropriate therapeutic target in osteoarthritis. Nature 434(7033):644–648
osteoarthritis--insights from studies of aggre- 14. Stanton H, Rogerson FM, East CJ, Golub SB,
can and collagen proteolysis? Curr Drug Tar- Lawlor KE, Meeker CT, Little CB, Last K,
gets 11(5):561–575. https://doi.org/10. Farmer PJ, Campbell IK, Fourie AM, Fosang
2174/138945010791011956 AJ (2005) ADAMTS5 is the major aggrecanase
8. Bernardo BC, Belluoccio D, Rowley L, Little in mouse cartilage in vivo and in vitro. Nature
CB, Hansen U, Bateman JF (2011) Cartilage 434(7033):648–652
intermediate layer protein 2 (CILP-2) is 15. Little CB, Barai A, Burkhardt D, Smith SM,
expressed in articular and meniscal cartilage Fosang AJ, Werb Z, Shah M, Thompson EW
DMM Mouse Model of Post-traumatic Osteoarthritis 259
Abstract
Immunostaining is the process of identifying proteins in tissue sections by incubating the sample with
antibodies specific to the protein of interest, then visualizing the bound antibody using a chromogen
(immunohistochemistry or IHC) or fluorescence (immunofluorescence or IF). Unlike in situ hybridization,
which identifies gene transcripts in cells, immunostaining identifies the products themselves and provides
information about their localization within cells (nuclear, cytoplasmic, or membrane) or extracellular
matrix. This can be particularly important in the context of bone and cartilage because they contain
many cell types as well as matrix components, each with distinct protein expression patterns. As the number
of antibodies continues to grow, this technique has become vital for research laboratories studying the
skeleton. Here, we describe a detailed protocol for antibody-based in situ analysis of bone and associated
tissues, addressing specific issues associated with staining of hard and matrix-rich tissues.
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_15, © Springer Science+Business Media, LLC, part of Springer Nature 2021
261
262 Crystal Idleburg et al.
1.1 Fixation Fixation is the process of treating tissue with solutions that preserve
gross morphology as well as molecular structures within the tissue
and should be started as soon as possible after harvest [2, 3]. Pene-
tration of fixative is determined by the size and nature of the tissue
of interest. Soft tissues and small pieces of tissue will fix faster than
larger or harder tissues. The standard fixative for paraffin embed-
ding is 10% neutral buffered formalin (NBF), while the most typical
for frozen sections is 4% paraformaldehyde (PFA). However, when
applied properly, either fixative can be used for each embedding
method. As with most fixatives, these solutions preserve tissue by
cross-linking the proteins. Therefore, tissues fixed in 10% NBF and
4% PFA usually require antigen retrieval before incubation with
primary antibody. Because of the cross-linking action, it is impor-
tant to avoid over-fixation as this can lead to excessive cross-linking,
which may mask the antigens, or dehydration, which may produce
an undulation artifact during sectioning [2]. Fixation in neutral
buffered zinc formalin, which prevents excessive cross-linking,
should also be considered for frozen sections, as it often eliminates
the need for antigen retrieval steps [4]. Although unfixed frozen
sections are useful in some situations such as reporter cell lines or
mice that express fluorescent proteins, fixation is almost always
required when performing antibody-based detection, and results
are generally better when tissue is fixed up front, rather than
dipping sections into fixative.
To ensure proper preservation when working with bone or
cartilage, it is necessary to clean away any unwanted soft tissue
such as skin and muscle. This allows for fixative penetration in a
timely manner and avoids under- or over-fixation. It also makes it
easier to orient the bone during embedding. Additionally, unde-
sired autofluorescence of blood cells can be removed by perfusing
the animal with phosphate-buffered saline (PBS) and 10% NBF or
4% PFA to rinse and fix the vasculature, followed by a post-fix of
immersion in 10% NBF or 4% PFA [5].
1.3 Antigen Retrieval Due to the cross-linking action of most fixatives, it is often neces-
sary to unmask antigens before staining [8]. The choice of retrieval
method will vary according to the antigens and antibodies used.
There are several methods of antigen retrieval but they fall into two
main categories, enzyme digestion and heat treatment. Each
retrieval method presents its own challenges and needs optimiza-
tion for different specimen types. Enzyme digestion requires preci-
sion in pH and duration of treatment because different tissues will
digest at different rates. The challenge in heat retrieval is in treating
the tissue long enough to ensure antigen retrieval without causing
it to lift off from the slide, which is a common problem when
working with cartilage and bone. For frozen sections, if retrieval
is necessary, enzymatic methods should be of primary consideration
due to the fragility of hydrated, aqueous-treated tissues. Antigenic-
ity in frozen sections can also be improved by use of a buffer
solution containing detergent (such as Tween 0.05–0.1%).
1.4 Data Analysis To accurately interpret staining, it is important to know the stan-
dard morphology and staining patterns in the tissue of interest.
Textbooks on histology, pathology, and developmental biology can
be a good resource for identifying the cells and structures. To
interpret the staining itself, the first priority is in determining
whether the signal (whether a chromogen or fluorescence) is spe-
cific or represents nonspecific background. Having both negative
and positive controls is crucial in making this determination. Nega-
tive control slides can be generated in two ways: no primary anti-
body or isotype- and species-matched immunoglobulin or serum
(which contains immunoglobulin) instead of primary antibody step
[8, 9]. While the no primary negative controls are usually accept-
able, the isotype control is the gold standard because it is possible to
264 Crystal Idleburg et al.
2 Materials
3 Methods
3.1 IHC on Paraffin 1. If perfusing for rodent studies, an approved dose of anesthetic
Sections should be administered as determined by your local animal
safety committee. When the animal is fully sedated, it can be
3.1.1 Tissue Preparation
perfused with an intracardial injection of 1 M PBS, followed by
10% NBF or 4% PFA (typically 10 mL to 25 mL each). Con-
stant flow rate can be applied with a peristaltic syringe pump or
other method to improve the quality of the perfusion
fixation [5].
2. Immediately after dissection, fix bones in 10% neutral buffered
formalin or 4% paraformaldehyde for 24–72 h at room temper-
ature or 4 C. Fixative volume should be 15–20 times tissue
volume. To ensure complete penetration, tissue should be
agitated on a shaker during fixation (see Note 2).
3. Rinse tissue in PBS or distilled deionized water (ddH2O)
6 times, 15 min each.
4. Decalcify in 14% free acid EDTA, pH 7.2–7.4, with rocking,
changing solution daily (on weekdays only is OK). The number
of days required for decalcification of mouse bones is as follows
(see Note 3):
(a) Embryo > E17.5: 1–2 days.
(b) Postnatal days (P) 1–4: 3 days.
(c) P5-P10: 4–5 days.
Immunostaining of Skeletal Tissues 267
3.2 IF on Frozen 1. Perfusion, fixation, decalcification, and washing are the same as
Sections Subheading 3.1.1, steps 1–5 above.
3.2.1 Tissue Preparation 2. Incubate tissue in 30% sucrose solution at 4 C for 3–5 days.
This helps to prevent formation of damaging ice crystals during
freezing/embedding. Tissue should sink to the bottom of the
container when ready for embedding.
3. Embed sucrose-permeated tissue in OCT Compound (Tissue-
Tek), paying close attention to orientation (see Note 17). Store
frozen blocks at –80 C until sectioning.
4. Tissue sections are then cut at the desired thickness (generally
from 10–100 μm) using a cryostat and placed on color frost
slides (see Note 18).
5. Store frozen slides at –80 C for long-term storage.
10. If the mounting media used contains DAPI, skip to step 12. If
not, add DAPI counterstain cocktail to slide and incubate for
5–7 min. Aspirate the liquid immediately to avoid overstaining.
For DAPI concentration suggestions, see Note 22.
11. Aspirate the liquid immediately to avoid overstaining. Rinse
with PBST 3 5 min.
12. Place 2–3 drops or a thin line of glycerol-based aqueous
mounting media along the edge of a coverslip. Carefully place
the slide (tissue down) on top, beginning on the edge with the
mounting media. This helps to prevent bubbles.
13. Seal slide edges with clear nail polish. This prevents the slides
from drying out and media from escaping, as aqueous media
does not fully set.
14. Keep slides at 4 C. Room temperature slides may dry out,
mold, or lose fluorescence more quickly (see Note 23).
4 Notes
References
1. Bord S (2003) Protein localization in 4. L’Hoste RJ Jr, Tourres MA (1995) Using Zinc
wax-embedded and frozen sections of bone Formalin as a Routine Fixative in the Histology
using immunohistochemistry. Methods Mol Laboratory. Lab Med 26(3):210–214
Med 80:237–247 5. Gage GJ, Kipke DR, Shain W (2012) Whole
2. Freida CL, Hladik C. Histotechnology.; A self animal perfusion fixation for rodents. J Vis Exp
instructional text , 2009 65
3. Sheehan DC, Hrapchak BB (1980) Theory and 6. Hao Z, Kalscheur VL, Muir P (2002) Decalcifi-
Practice of Histotechnology. 2nd ed. Bancroft cation of bone for histochemistry and immuno-
JD, Stevens J, editors. St. Louis, MO, The CV histochemistry procedures. J Histotechnol 25
Mosby Company (1):33–37
Immunostaining of Skeletal Tissues 273
7. Serowoky MA, Patel DD, Hsieh JW, Mariani FV 9. Watkins S (2003) Immunohistochemistry. In:
(2018) The use of commercially available adhe- Ausubel FM, Brent R, Kingston RE, Moore
sive tapes to preserve cartilage and bone tissue DD, Seidman JG, Smith JA et al (eds) Current
integrity during cryosectioning. BioTechniques protocols in molecular biology. John Wiley &
65(4):191–196 Sons Inc., Hoboken, NJ. p. Supplement 7 Unit
8. Battifora H (1999) Quality assurance issues in 14.6
immunohistochemistry. J Histotechnol 22
(3):169–175
Chaptert 16
Abstract
Quantification of cortical bone mass and architecture using μCT is commonplace in osteoporosis and
osteoarthritis research. Different groups often report substantially divergent mouse cortical bone responses
to nominally comparable interventions. In the case of studies assessing bones’ responses to externally
applied loading, these differences are commonly associated with methodological differences in the loading
regime. This chapter describes a widely published, standardized method of in vivo mouse tibia axial loading
to produce lamellar bone formation. Despite uniform application of axial loading, changes in bone mass are
highly site-specific within individual bones. For example, the mouse proximal tibia rapidly accrues new bone
following axial loading, but this osteogenic response tapers to produce undetectable differences distally.
Consequently, the bone sites selected for comparisons substantially influence the magnitude of differences
observed. Application of the freely available Site Specificity software allows site-specific responses to be
identified by rapidly quantifying cortical bone mass at each 1% site along the bone’s length. This high-
content screening tool has been informatively applied to study the local effects of changes in loading as well
as systemic interventions including hormonal treatment and aging. Automated multisite analyses of cortical
mass is increasingly identifying site-specific effects of “systemic” interventions such as global gene deletions.
Biological mechanisms underlying this apparent regionalization of cortical responses are largely unknown
but may start to be elucidated by increasingly widespread application of Site Specificity methods.
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_16, © Springer Science+Business Media, LLC, part of Springer Nature 2021
275
276 Sara H. Windahl et al.
2 Materials
2.1 Strain Gauging 1. Strain gauge, 120 Ω: Micro Measurements (Vishay Precision
and In Vivo Axial Tibial group, NC, USA), catalog number EA-06-031DE-120.
Loading 2. PCT-3 M gauge installation tape, Micro Measurements.
3. M-Flux, Micro Measurements.
4. Single conductor wire e.g.134 AWP, Micro Measurements.
5. M-Bond 200 adhesive and Catalyst, Micro Measurements.
6. Tin lead soldering wirer 60/40 0.8 mm.
7. Soldering station.
8. Multimeter, e.g., Digital multimeter, auto-range, VWR (Lut-
terworth, UK), catalog number 620–1920.
9. Dissection microscope with extra light.
10. Glass plate.
11. Cotton swabs.
12. Scalpels and forceps.
13. Isopropanol or 70% ethanol.
14. Polyurethane spray (can be purchased from local hardware
stores).
278 Sara H. Windahl et al.
2.2 Ex Vivo μCT 1. There are several μCT instruments. Methods and settings
Imaging described in Subheading 3.2 Ex vivo μCT imaging below is
optimized for the use of an 1172 model, Bruker microCT
(Aartselaar, Belgium).
2. BMD calibration phantoms, 0.25 and 0.75 g/cm3, Bruker
microCT (Aartselaar, Belgium).
3 Methods
3.1.1 Preparation 1. Place the gauge on the glass board (shining side up, Fig. 1).
of the GAUGES 2. Secure one of the long edges of the gauge to the board with
adhesive masking tape.
3. Cut three sides of the gauge as shown in Fig. 1.
4. Lightly scratch the terminals with the blade to improve solder
adherence.
5. Apply a drop of M-Flux to each terminal.
6. Apply solder to each terminal.
Fig. 1 Sites for trimming the strain gauge. The gauge is attached to a glass plate
at the bottom of the image. The gauge is then cut with a scalpel in the indicated
order 1–4
Mapping Regional Cortical Bone Responses to Local Changes in Loading. . . 279
Fig. 2 Photomicrograph of strain gauge with the attached wires before the final
cut. Solder points (arrows) with wire attached. The side cuts have been made
and the top cut is all that remains to be made. The top edge of the gauge is
immobilized by masking tape. The scale bar indicates 0.5 mm
3.1.2 Attach the Strain 1. Euthanize the mouse using an ethically approved method
Gauge to the Tibia of a which the experimenter is competent and confident
Recently Euthanized Mouse performing.
to Pre-gauge (See Fig. 3) 2. Locally dissect the area of the right tibia where the strain gauge
should be applied.
3. Wipe the bone with a cotton swab dipped in EtOH or isopro-
panol to degrease the bone and facilitate attachment of the
gauge.
3.1.3 Attachment 1. Place a drop of catalyst and a drop of glue on a piece of foil.
of the Gauge to the Mouse 2. Dip the gauge in catalyst.
Tibia
280 Sara H. Windahl et al.
Fig. 3 Attachment site of the gauge. The gauge is attached so that the center of
the gauge is located approximately at 37% of the bone’s length proximally. The
arrows indicate where the top of the gauge will be located
3. Immediately dip the gauge in glue and touch a clean bit of the
foil surface to remove excess glue. Too much glue will disable
the function of the gauge.
4. Hold the gauge steady on the bone with forceps to attach it to
the bone. See Fig. 3 for the correct site. The center of the
gauge should be where the tibia is the widest. The upper part of
the gauge should be positioned where there is an indention at
the medial side.
5. Test the gauge with a multimeter.
6. Secure with polyurethane (on gauge and bone if stored cold in
PBS before use).
(continued)
Mapping Regional Cortical Bone Responses to Local Changes in Loading. . . 281
6. Dwell Load 2s
7. Repeat 3–6 two times, recording the resistance value achieved each
time (in total three recordings per load).
8. Repeat 3–7 with increasing peak loads in step 3.
9. Ramp Displace 1.9 mm/s to 0.8 mm (unload)
Test each load at least three times on each bone (see table
above). The resulting load versus gauge resistance plot can then
be used to generate load:strain graphs by using the gauge factor to
convert resistance values into strain values.
3.1.5 Axially Load Once you know what loads need to be applied experimentally, place
the Tibia an anesthetized mouse in the loading device so that the knee is fixed
of Anesthetized Mice in the upper cup, and the foot is fixed in the lower cup. Make sure
that the foot and knee are aligned so that the joint is loaded axially
(see Note 2). The sequence below describes a standard rest-inserted
loading program applying 40 cycles at 11 N compressive load.
3.2 Ex Vivo μCT 1. Euthanize the mouse using an ethically approved method
Imaging which the experimenter is competent and confident
performing.
3.2.1 Dissect the Left
(Control) and Right 2. Using sharp-tipped scissors, make a skin incision over the
(Exogenously Loaded) Tibia femur. The skin can then be manually pulled distally over the
calcaneus (see Note 3).
3. Proximally, flex the knee (stifle) and insert the sharp tip of a
10A scalpel laterally to medially through the popliteus to sever
muscle and tendon attachments. Similarly, cut through the
straight patellar ligament and collateral ligaments on either
side of the knee. This should destabilize the joint, allowing
you to insert the scalpel into the joint to cut the collateral
ligaments and disarticulate the femur from the tibia (see
Note 4).
282 Sara H. Windahl et al.
4. Distally, flex the ankle (hock) and cut through the intertarsal
joints. It is not necessary to disarticulate the talus from the tibia
and doing so can occasionally damage the medial malleolus and
artifactually shorten the tibia, affecting bone measurement.
3.2.2 Fix and Dehydrate 1. Fix the bone and muscle in chilled 4% PFA for 2 days.
the Tibia 2. Dehydrate the tibia and muscle in 70% ethanol (see Note 5).
3.2.3 μCT Scan Each Methods and settings described here may need to be adapted to the
Bone specific μCT scanner available following the manufacturers’
instructions.
1. Prepare the bone by rolling the bone in non-PVC cling film
and place in the container (i.e., a straw). Attach the straw to the
μCT stage stand before screwing it into the scanner. If the bone
can be scanned dry, make sure the bone has dried out for
~10 min before the scan. Evaporation during the scan can
alter the density values. If the bone needs to remain moist,
then it can be scanned in 70% ethanol within a cut-down 1 mL
syringe using syringe stoppers above and below the bone to
keep the bone stable and avoid evaporation.
2. Select the parameters that best suit your bone. For mouse
bones, we use a 0.5 mm aluminum filter to reduce beam
hardening. Beam hardening is what happens when the bone
preferentially absorbs low energy photons and results in a dark
looking surface. The filter will decrease the low energy output
and improve the image. The whole tibiae and surrounding
muscles are typically imaged with a voxel size of 4.8 μm
(110 μm3). The applied X-ray voltage is 49–50 kV, current of
200 mA, with 0.5 mm aluminum filtration. Scans are obtained
over 180 degrees with a 0.6-degree rotation step. The images
can be reconstructed and binarized with global thresholding
(values: 1.000–1.160) using the NRecon Bruker software.
3. If the entire bone does not fit on one scanning run, an “over-
size scan” may be needed to scan larger bones, which can then
be merged digitally in post-processing to generate a single
bone series.
4. Once scanning is complete, replace bone in 70% ethanol solu-
tion and proceed with processing for additional tests (see
Note 6).
3.3 Site Specificity 1. Place the reconstructed cross-section μCT images for each
and Statistical bone in a single folder, henceforth the “Inpath.” Ensure
Analysis there are no blank (“spacer”) slices in the folder as bone length
is calculated from the number of slices in this folder.
2. Create a local copy of the SiteSpecificityV2.m script. This can
be downloaded from https://www.researchgate.net/publica
tion/295858230_SiteSpecificityV2 and is available with the
original manuscript [23].
3. Open MATLAB®, direct it to the directory containing SiteSpe-
cificityV2.m and open the script for editing.
4. Modify the Inpath line (line 13) to the folder containing the
bone to be analyzed (see Note 4). For example, change this line
to read:inpath ¼ ‘C:\Site specificity\Example\’
5. Save the script file and run it. The script will create an “Out-
put” folder within the Inpath. It will count the number of
images of the folder and identify the images corresponding to
each 1% site along the bone’s length (see Note 8). The
corresponding 100 images each undergo the following
processing (see Fig. 2a, b):
(a) Locally adaptive binarization to identify the bone and
exclude soft tissue.
(b) Identification of the largest continuous object (the tibia)
and exclusion of smaller objects (the fibula and trabecular
bone).
(c) Quantification of the binarized area. This is exported as
“Bone Area” and corresponds to Cortical Area (Ct.Ar) in
conventional μCT analysis. The output report will display
this as pixels2 and should be converted to mm2 using the
voxel size known from the μCT reconstruction settings.
(d) Quantification of empty space within the binarized area
(see Note 9). This is reported as “Marrow Area” (i.e., Ma.
Ar) in pixels2.
(e) The perimeter of the binarized area is also provided. If
required, Bone Area and Marrow Area can be summer to
calculate Total Tissue Area (Tt.Ar).
6. The Output folder will now contain 100 images saved as .gif
files corresponding to each 1% site along the bone’s length, .txt
result files with the cross-sectional measurements at each site,
and a single .csv file amalgamating the results (see Note 10).
The .gif files retain the same pixel dimensions as the original
image but are of much smaller size such that they can be easily
processed in additional software if needed.
284 Sara H. Windahl et al.
Fig. 4 Execution and application of the Site Specificity workflow. (a) Schematic representation of Site
Specificity analysis applied to the tibia of a 3-week-old mouse. The region of the bone analyzed by Site
Specificity are shown in the binarized reconstruction (right). Parts of the fibula (red highlight) not connected to
the tibia are identified and excluded. Images corresponding to each 1% site along the bone’s length are
identified. Proximal (including the growth plate, green) and distal sites are excluded from the analysis. (b)
Cross-section through the proximal tibia and fibula (red) at the 20% site from the proximal end. The binarized
tibia images are saved and can be exported into other analysis software such as BoneJ. Cortical area (Ct.Ar),
Marrow area (Ma.Ar), and external perimeter are provided as part of the in-built workflow. (c) Quantification of
Ct.Ar and Ma.Ar in the tibiae of five 3-week-old mice showing the expected pattern of cortical bone differences
in typical inter-mouse variability
Mapping Regional Cortical Bone Responses to Local Changes in Loading. . . 285
(a) Mixed models are used to test the effect of random effects
(e.g., cage number), fixed effects (e.g., loading), and fixed
covariates (e.g., % site along the bone’s length). Strengths
and limitations of mixed models are described elsewhere
(e.g., [27]).
(b) Set up a datasheet with the following headings: Animal
ID, Site (% bone length), and Treatment (e.g., loading,
drug, genotype). Enter all data in a single list under each
of these headings. Set the appropriate variable types in the
SPSS “Variable view” tab.
(c) In SPSS, select the linear mixed models option. Specify
the Animal ID as the “subject,” Site as the “repeated
measure” (see Note 11), and click continue. Set up your
model as needed on the next screen (see the SPSS help
function on your version of the software if you are not
familiar with how to do this).
(d) An important outcome of Site Specificity analysis is quan-
tification of whether the response to treatment is signifi-
cantly different between different sites. This is shown by a
significant Site by Treatment interaction (indicates the
response to treatment is significantly different between
different sites).
(e) To request a direct comparison between different sites
with a Bonferroni post hoc comparison, click “Paste” to
obtain the model syntax and add the following text at the
end of the syntax before the full stop:
/EMMEANS ¼ TABLES(Treatment*Site) COMPARE
(Treatment) ADJ(Bonferroni)
(f) Post hoc comparisons are generated in a table under the
subheading “Pairwise Comparisons” (see Notes 12 and
13).
(g) The same syntax can now be rerun for other bone para-
meters in the same experiment.
4 Notes
10. If the extremes of the bone have very faint signal which is of
soft-tissue equivalence, such that they binarize to produce a
black image, that slice will be excluded from the analysis. The
total number of output images and values will therefore be less
than 100. Additionally, the proximal and distal extremes of the
bone contain trabecular bone which cannot be analyzed using
Site Specificity. We therefore exclude the proximal and distal
10–15% of the tibia from our analyses. This may not be neces-
sary when analyzing less trabecular bone such as the
mouse ulna.
11. The structure of the model used will vary depending on the
endpoint. When testing whether an intervention enhances or
diminishes the response to loading, it may be more appropriate
to calculate the percentage difference between loaded and
control limbs for each mouse and use this loading response as
the dependent variable. This is especially true if the interven-
tion itself changes baseline bone mass (see Strain Gauging
methods above).
12. Displaying regions of significant differences between treatment
groups across multiple sites can be challenging. We and others
have adopted various methods to show this in previous
publications.
13. Calculating the statistical power to identify differences as sta-
tistically significant using a complex mixed model approach is
challenging. It is possible to simulate changes in bone mass by
resizing the Output images in Fiji by a known percentage. This
provides an indication of the proportion of sites at which that
percentage change would be detected as statistically significant.
In our initial validation of this analysis method [23], a 10%
change in Ct.Ar was detected as significant at 98.8% of sites
analyzed.
References
1. Hillam RA, Skerry TM (1995) Inhibition of 4. Galea GL, Meakin LB, Harris MA, Delisser PJ,
bone resorption and stimulation of formation Lanyon LE, Harris SE, Price JS (2017) Old age
by mechanical loading of the modeling rat ulna and the associated impairment of bones’ adap-
in vivo. J Bone Miner Res 10:683–689 tation to loading are associated with transcrip-
2. Skerry TM (2006) One mechanostat or many? tomic changes in cellular metabolism, cell-
Modifications of the site-specific response of matrix interactions and the cell cycle. Gene
bone to mechanical loading by nature and nur- 599:36–52
ture. J Musculoskelet Neuronal Interact 5. Skerry TM, Bitensky L, Chayen J, Lanyon LE
6:122–127 (1989) Early strain-related changes in enzyme
3. Meakin LB, Price JS, Lanyon LE (2014b) The activity in osteocytes following bone loading
Contribution of Experimental in vivo Models in vivo. J Bone Miner Res 4:783–788
to Understanding the Mechanisms of Adapta- 6. Moustafa A, Sugiyama T, Prasad J, Zaman G,
tion to Mechanical Loading in Bone. Front Gross TS, Lanyon LE, Price JS (2012)
Endocrinol (Lausanne) 5:154 Mechanical loading-related changes in
288 Sara H. Windahl et al.
osteocyte sclerostin expression in mice are relationship between bone formation and
more closely associated with the subsequent mechanical stimulus within cortical bone:
osteogenic response than the peak strains Combining 3D fluorochrome mapping and
engendered. Osteoporos Int 23:1225–1234 poroelastic finite element modelling. Bone
7. Robling AG, Niziolek PJ, Baldridge LA, Con- Rep 8:72–80
don KW, Allen MR, Alam I, Mantila SM, 17. Sugiyama T, Saxon LK, Zaman G, Moustafa A,
Gluhak-Heinrich J, Bellido TM, Harris SE, Sunters A, Price JS, Lanyon LE (2008)
Turner CH (2008) Mechanical stimulation of Mechanical loading enhances the anabolic
bone in vivo reduces osteocyte expression of effects of intermittent parathyroid hormone
Sost/sclerostin. J Biol Chem 283:5866–5875 (1-34) on trabecular and cortical bone in
8. Bergstrom I, Isaksson H, Koskela A, mice. Bone 43:238–248
Tuukkanen J, Ohlsson C, Andersson G, Wind- 18. Dodge T, Wanis M, Ayoub R, Zhao L, Watts
ahl SH (2018) Prednisolone treatment reduces NB, Bhattacharya A, Akkus O, Robling A,
the osteogenic effects of loading in mice. Bone Yokota H (2012) Mechanical loading, damp-
112:10–18 ing, and load-driven bone formation in mouse
9. Lionikaite V, Henning P, Drevinge C, Shah FA, tibiae. Bone 51:810–818
Palmquist A, Wikstrom P, Windahl SH, Lerner 19. Nakamura T, Imai Y, Matsumoto T, Sato S,
UH (2019) Vitamin A decreases the anabolic Takeuchi K, Igarashi K, Harada Y, Azuma Y,
bone response to mechanical loading by sup- Krust A, Yamamoto Y, Nishina H, Takeda S,
pressing bone formation. FASEB J Takayanagi H, Metzger D, Kanno J,
33:5237–5247 Takaoka K, Martin TJ, Chambon P, Kato S
10. Meakin LB, Todd H, Delisser PJ, Galea GL, (2007) Estrogen prevents bone loss via estro-
Moustafa A, Lanyon LE, Windahl SH, Price JS gen receptor alpha and induction of Fas ligand
(2017) Parathyroid hormone’s enhancement in osteoclasts. Cell 130:811–823
of bones’ osteogenic response to loading is 20. de Souza RL, Pitsillides AA, Lanyon LE, Skerry
affected by ageing in a dose- and time- TM, Chenu C (2005) Sympathetic nervous
dependent manner. Bone 98:59–67 system does not mediate the load-induced cor-
11. Svensson J, Windahl SH, Saxon L, Sjogren K, tical new bone formation. J Bone Miner Res
Koskela A, Tuukkanen J, Ohlsson C (2016) 20:2159–2168
Liver-derived IGF-I regulates cortical bone 21. Javaheri B, Carriero A, Wood M, de Souza R,
mass but is dispensable for the osteogenic Lee PD, Shefelbine S, Pitsillides AA (2018)
response to mechanical loading in female Transient peak-strain matching partially
mice. Am J Physiol Endocrinol Metab 311: recovers the age-impaired mechanoadaptive
E138–E144 cortical bone response. Sci Rep 8:6636
12. Meakin LB, Galea GL, Sugiyama T, Lanyon 22. Galea GL, Meakin LB, Williams CM, Hulin-
LE, Price JS (2014a) Age-related impairment Curtis SL, Lanyon LE, Poole AW, Price JS
of bones’ adaptive response to loading in mice (2014) Protein kinase Calpha (PKCalpha) reg-
is associated with sex-related deficiencies in ulates bone architecture and osteoblast activity.
osteoblasts but no change in osteocytes. J J Biol Chem 289:25509–25522
Bone Miner Res 29:1859–1871 23. Galea GL, Hannuna S, Meakin LB, Delisser PJ,
13. Sugiyama T, Meakin LB, Browne WJ, Galea Lanyon LE, Price JS (2015) Quantification of
GL, Price JS, Lanyon LE (2012) Bones’ adap- alterations in cortical bone geometry using site
tive response to mechanical loading is essen- specificity software in mouse models of aging
tially linear between the low strains associated and the responses to ovariectomy and altered
with disuse and the high strains associated with loading. Front Endocrinol (Lausanne) 6:52
the lamellar/woven bone transition. J Bone 24. Javaheri B, Herbert E, Hopkinson M,
Miner Res 27:1784–1793 Al-Jazzar A, Pitsillides AA (2019b) Sost hap-
14. Javaheri B, Bravenboer N, Bakker AD, van der loinsufficiency provokes peracute lethal cardiac
Veen A, de Souza RL, Saxon L, Pitsillides AA tamponade without rescuing the osteopenia in
(2019a) In vivo models of mechanical loading. a mouse model of excess glucocorticoids. Am J
Methods Mol Biol 1914:369–390 Pathol 189:753–761
15. Melville KM, Robling AG, van der Meulen MC 25. Orriss IR, Lanham S, Savery D, Greene NDE,
(2015) In vivo axial loading of the mouse tibia. Stanier P, Oreffo R, Copp AJ, Galea GL (2018)
Methods Mol Biol 1226:99–115 Spina bifida-predisposing heterozygous muta-
16. Carrieroa A, Pereirab AF, Wilson AJ, tions in planar cell polarity genes and Zic2
Castagno S, Javaheri B, Pitsillides AA, reduce bone mass in young mice. Sci Rep
Marenzana M, Shefelbine SJ (2018) Spatial 8:3325
Mapping Regional Cortical Bone Responses to Local Changes in Loading. . . 289
26. Doube M, Klosowski MM, Arganda-Carreras I, 27. Yang J, Zaitlen NA, Goddard ME, Visscher
Cordelieres FP, Dougherty RP, Jackson JS, PM, Price AL (2014) Advantages and pitfalls
Schmid B, Hutchinson JR, Shefelbine SJ in the application of mixed-model association
(2010) BoneJ: free and extensible bone image methods. Nat Genet 46:100–106
analysis in ImageJ. Bone 47:1076–1079
Chapter 17
Abstract
Musculoskeletal pain contributes significantly to chronic pain experienced by adults and to health care use.
This chapter details several methods to evaluate pain and physical activity in mice that can be applied to
preclinical orthopedic models. These methods include the von Frey filament assay that measures mechanical
allodynia, open-field activity assays for evaluation of ambulation, and incapacitance measurements to
determine static weight bearing.
Key words Functional assays, Mechanical allodynia, von Frey, Incapacitance, Evoked pain, Open-field
activity assays, Static weight bearing
1 Introduction
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7_17, © Springer Science+Business Media, LLC, part of Springer Nature 2021
291
292 David H. H. Molstad and Elizabeth W. Bradley
2 Materials
2.3 Static Weight 1. Incapacitance meter (Bioseb, Pinellas Park, FL, BIO-SWB-
Bearing Incapacitance TOUCH-M) or equivalent instrument.
Assays (a) Animal holder/restrainer (rat or mouse).
(b) Control unit.
(c) Platform with sensors.
(d) Footswitch to start/stop experiment hands free.
2. PC Station with appropriate hardware.
3. Lab tape.
Pain and Activity Measurements 295
3 Method
3.1.2 On Day of Testing 1. Assemble the animal enclosure/grid platform (see Fig. 2).
Ensure that each enclosure is fully accessible through the
underlying mesh of the grid platform. Note: Opaque material
should be used to separate mice so that they cannot see each
other during acclimation or testing periods (see Fig. 2).
2. Place animals in enclosure chambers. Let animals acclimate to
enclosure until they are no longer actively exploring. If animals
are properly acclimated to the enclosure chambers, this should
take under 15 min (see Note 1).
Fig. 2 Equipment for the von Frey Monofilament assays. (a) Assembly of the von Frey equipment includes
multiple animal enclosures placed above a suspended grid platform. Mice are placed within chambers so that
the plantar surface can be accessed through the mesh grid. An opaque wall separates chambers so that
animals cannot see each other during acclimation and evaluation periods. (b) A mouse placed within the
animal enclosure chamber demonstrating access of the hind limb plantar surface through the mesh grid. (c)
Monofilaments ranging in fiber bowing weight used to perform the von Frey assay. (d) The von Frey
monofilament is applied until the fiber bows to achieve a defined force
296 David H. H. Molstad and Elizabeth W. Bradley
3.2 Open-Field 1. Transport mice to the area where activity arenas are located.
Activity Measurements Allow mice 15 min in cages to recover from transport before
starting analyses.
2. Open the Opto-track software. Open a new experiment, select
the location to save the file, and name the file.
3. Select arenas to be used in each run by checking the appropriate
checkbox within the software. Adjust the total time for the
analysis to 20 min or desired alternate time.
4. Verify that the software is localizing objects in the three-
dimensional space by tracing along the arena surface. Place
one animal in each arena. Allow the animal to acclimate within
the arena for 2 min (see Note 2).
5. Hit the “Start All” button within the Opto-track software.
Ensure the software is tracking the animal accurately. Allow
time to elapse.
6. When the experiment is complete, return animals to cages.
Select “Analyze Experiment,” and then export all bin files.
This will save a .csv file to the location you selected that can
be viewed with Excel software.
Pain and Activity Measurements 297
7. Open the .csv file in Excel. To obtain the total distance and
time measurements, sum all measured intervals. The speed of
the animal can be obtained through the average of all speed
values obtained during the experiment.
8. Determine changes between experimental groups and or
over time.
3.3 Static Weight 1. Transport mice to the area where the incapacitance meter is
Bearing Incapacitance located. Allow mice 15 min in cages to recover from transport
Assays before starting analyses.
2. Place the animal within the animal enclosure (see Fig. 3). Tap-
ing down the animal’s tail to the base of the enclosure will help
the animal assume correct placement.
3. Ensure that the animal has both upper limbs placed equally on
the inclined portion of the animal enclosure and that both hind
feet are located on the force plate (see Fig. 3 and Note 3).
4. Use the footswitch to take a 5-s measurement.
5. Repeat for a total of three consecutive 5-s measurements. Use
the average of these three measurements.
6. Determine the percent weight bearing on the affected limb
using the following equation:
% weight on ipsilateral limb ¼ (weight placed on ipsilateral
limb/summed weight placed on both limbs) 100%
7. Determine changes between experimental groups and/or
over time.
Fig. 3 Incapacitance measurement equipment. The incapacitance assay measures static weight bearing on
each hind limb of a mouse. (a) Side view of a mouse placed in the incapacitance chamber. (b) Forward view of
a mouse within the incapacitance chamber. (c) Rear view of a mouse placed in an incapacitance chamber.
Note the foot on the left is not contacting the force plate within the chamber. Lab tape is also used to secure
the animal’s tail to prevent the animal from turning around within the chamber
298 David H. H. Molstad and Elizabeth W. Bradley
4 Notes
1. During the acclimation period of the von Frey assay, the reader
should be present. Acclimation of the experimental animals to
the enclosures will ensure that habituation occurs on test days.
Fiber bowing weights can be adjusted if a response is either not
elicited or is saturated (e.g., paw withdrawal is either 0% or
100%). During testing, be careful not to brush or scratch the
plantar surface with the monofilament. Contact with the outer
margins of the plantar surface should also be avoided. A posi-
tive response is evident if the animal actively withdraws its paw
from the stimulus.
2. The activity assay software should display the correct location
(lower left hand is 0,0 and the upper right is 30,30 within the
Cartesian plane, see (Fig. 4) and register that the z-plane is
broken. If software is not registering location correctly, ensure
that the arena walls or other items do not break infrared beams.
3. For static weight bearing assays, the animal must assume the
correct positioning within the enclosure before an accurate
reading can be taken. This includes the animal facing forwards
with both forelimbs placed on the incline and both hind feet
placed on the force plates. Likewise, it is difficult to determine if
an animal will not place its foot on the force plate due to pain or
due to non-compliance.
Fig. 4 Open-field activity assays. (a) An open-field activity assay arena. (b) Each activity arena tracks the
three-dimensional movement of a mouse within the arena. Cartesian coordinates for each corner of the arena
are shown and can be used to verify correct tracking of an animal by the infrared grid
Pain and Activity Measurements 299
Acknowledgements
This work was made possible by training and research grants from
the National Institutes of Health (AR065397 and AR072634) and
the University of Minnesota Stem Cell Institute and Board of
Regents. These contents are solely the responsibility of the authors
and do not necessarily represent the official views of the NIH.
References
1. Hawker GA (2017) The assessment of muscu- 8. Hwang SM, Feigenson M, Begun DL, Shull
loskeletal pain. Clin Exp Rheumatol 35(Suppl LC, Culley KL, Otero M, Goldring MB, Ta
107):8–12 LE, Kakar S, Bradley EW, Westendorf JJ
2. Carnide N, Hogg-Johnson S, Cote P, Irvin E, (2017) Phlpp inhibitors block pain and carti-
Van Eerd D, Koehoorn M, Furlan AD (2017) lage degradation associated with osteoarthritis.
Early prescription opioid use for musculoskele- J Orthop Res 36(5):1487–1497
tal disorders and work outcomes: a systematic 9. Xu M, Bradley EW, Weivoda MM, Hwang SM,
review of the literature. Clin J Pain Pirtskhalava T, Decklever T, Curran GL,
33:647–658 Ogrodnik M, Jurk D, Johnson KO, Lowe V,
3. Moshfegh J, George SZ, Sun E (2018) Risk Tchkonia T, Westendorf JJ, Kirkland JL (2017)
and risk factors for chronic opioid use among Transplanted senescent cells induce an
opioid-naive patients with newly diagnosed osteoarthritis-like condition in mice. J Geron-
musculoskeletal pain in the neck, shoulder, tol Ser A Biol Sci Med Sci 72:780–785
knee, or low back. Ann Intern Med 170 10. Majuta LA, Guedon JG, Mitchell SAT, Kus-
(7):504–505 kowski MA, Mantyh PW (2017) Mice with
4. Nicholson B (2006) Differential diagnosis: cancer-induced bone pain show a marked
nociceptive and neuropathic pain. Am J decline in day/night activity. Pain Rep 2:e614
Manag Care 12:S256–S262 11. Minville V, Laffosse JM, Fourcade O, Girolami
5. Deuis JR, Dvorakova LS, Vetter I (2017) JP, Tack I (2008) Mouse model of fracture
Methods used to evaluate pain behaviors in pain. Anesthesiology 108:467–472
rodents. Front Mol Neurosci 10:284 12. Lolignier S, Eijkelkamp N, Wood JN (2015)
6. Miller RE, Malfait AM (2017) Osteoarthritis Mechanical allodynia. Pflugers Archiv
pain: what are we learning from animal models? 467:133–139
Best Pract Res Clin Rheumatol 31:676–687 13. Mogil JS (2017) Laboratory environmental
7. Bradley EW, Carpio LR, McGee-Lawrence factors and pain behavior: the relevance of
ME, Castillejo Becerra C, Amanatullah DF, unknown unknowns to reproducibility and
Ta LE, Otero M, Goldring MB, Kakar S, Wes- translation. Lab Anim 46:136–141
tendorf JJ (2016) Phlpp1 facilitates post- 14. Pomonis JD, Boulet JM, Gottshall SL,
traumatic osteoarthritis and is induced by Phillips S, Sellers R, Bunton T, Walker K
inflammation and promoter demethylation in (2005) Development and pharmacological
human osteoarthritis. Osteoarthr Cartil 24 characterization of a rat model of osteoarthritis
(6):1021–1028 pain. Pain 114:339–346
INDEX
A E
Annulus fibrosus (AF)...............................................41–51 Endosteal mesenchymal progenitors ..........30, 33, 35–38
Antibodies .............................21, 91–93, 95, 96, 99, 100, Enzymatic digestions .............29–38, 42, 90, 98, 99, 271
196, 208, 221, 225, 229, 230, 236, 242, 244,
245, 256, 257, 261–266, 268, 270–272 F
Antigen retrieval................ 208, 244, 256, 262, 263, 267 Fixation ............................... 50, 168, 211, 214, 228–229,
236, 237, 244, 262, 266, 268, 270
B
Fluorescence activated cell sorting (FACS) ............89–91,
Bone marrow...........................15, 18, 19, 25, 29–38, 55, 93–97, 99, 100
57, 72, 194, 263 Fluorescence recovery ....... 112, 114–117, 121, 123, 134
Bone repair ........................................................... 193–203 Fluorescence recovery after photobleaching
Bones ...................................3–11, 15–26, 29, 30, 32, 33, (FRAP)...................................................... 109–137
35, 37, 38, 55, 57, 89, 90, 96, 97, 99, 154, 167, Fractures ............................ 193, 194, 205–222, 291, 292
174–178, 182, 185, 186, 188–190, 193, 194,
196, 198, 201–203, 206, 210, 212–214, 217, G
220, 221, 224, 234, 242, 248, 256, 262, 263,
Gene expression regulation ....................... 4, 50, 65, 110,
266, 267, 270, 275–287, 292, 293 142, 143, 172, 217–219, 222, 226, 230–233,
247, 264
C
Cartilage .................................... 21, 41, 53–69, 112, 142, H
154, 171, 183, 194, 206, 217, 221, 223–225,
Healing .............................193, 194, 196, 198, 203, 205,
230–232, 240, 242, 245, 247–254, 257, 262, 206, 210–212, 219–221, 225, 293
263, 292 Histology ........................37, 38, 50, 166, 172, 196, 206,
Cell isolation.......................................... 45–47, 51, 90, 97 207, 214–217, 236, 256, 263, 286
Chondrocytes ................................ 36, 41, 42, 48, 53–69,
Histomorphometry .............................165, 171, 182, 286
96, 111, 118, 119, 142, 254 Hydroxymethylcytosine ....................................... 101–107
Chondrogenesis ..................................15, 18, 20, 21, 102
Closed fracture ..................................................... 205, 206 I
Co-culture ................................53, 54, 58–61, 63, 65, 68
Collagenase................................ 4–11, 31, 37, 43, 46, 47, Immunofluorescence (IF).......................... 114, 116, 135,
50, 56, 57, 91, 232, 255 229, 230, 236, 239, 245, 247, 256, 261, 265,
Colony forming unit-fibroblast (CFU-F).............. 30, 33, 268–270
37, 38 Immunohistochemistry (IHC).................. 183, 208, 216,
Cryopreservation............................................................. 72 221, 229, 230, 236, 239, 242, 244–246, 256,
Cultures .............................. 5, 6, 8, 9, 11, 16, 19–21, 23, 261, 264, 266–268
25, 26, 29–38, 42–44, 46, 48–50, 54, 58–61, 63, Immunostaining........................... 21, 196, 237, 261–263
67, 68, 72, 74–78, 80–82, 86, 87, 109–137, 145, Incapacitance ............................................... 292, 294, 297
151, 168, 169, 177, 178, 224, 230 Induced pluripotent stem cells (iPSCs) ..................71–73,
78, 80–85
D Intervertebral disc (IVD) .........................................41–51
Isolation ........................ 3–11, 18, 29–38, 42, 43, 51, 57,
Decalcification .............................21, 183, 207, 214, 220, 89, 90, 96, 97, 202, 208–209, 217–219,
221, 228–229, 237, 244, 262–264, 266–268, 270 230–232, 247–254, 257
Andre J. van Wijnen and Marina S. Ganshina (eds.), Osteoporosis and Osteoarthritis, Methods in Molecular Biology, vol. 2221,
https://doi.org/10.1007/978-1-0716-0989-7, © Springer Science+Business Media, LLC, part of Springer Nature 2021
301
OSTEOPOROSIS AND OSTEOARTHRITIS
302 Index
M P
Matrix deposition......................................................53, 54 Pain ..........172, 202, 203, 206, 225, 255, 291–293, 298
Mechanical allodynia..................................................... 292 Protein dynamics.................................110, 121, 123, 132
Mechanical loading ..................................... 225, 275, 276
Mesenchymal stem cells .......................29, 54, 72, 74, 78, R
80–82, 96, 98, 111, 119 Regenerative medicine ................................................ v, 54
Mesenchymal stromal cells (MSCs) ................. 15–27, 29, Reprogramming ..............................71–73, 78, 80–82, 85
53–69, 78, 80
RNA and DNA extraction ..................230–232, 247, 248
Micro-computed tomography (micro-CT/μCT) ......206,
207, 210, 276 S
Mouse embryonic fibroblasts (MEFs) ...... 71, 73, 75–78,
80–82, 86 Single cell sequencing ......................................... v, 89–100
Site specificity ...................................... 276–278, 282–287
N Static weight bearing .................................. 292, 294, 298
Surgical models .................................................... 223–257
Next-generation sequencing ............................... 102–104
Nucleus pulposus (NP).............................................41–51 T
O Tissue engineering ................................................. 54, 154
Torsional testing............................................................ 207
Open field activity assays............................................... 298
Transcription factor activity.......................................... 110
Osteoarthritis ............................101, 102, 142, 165, 166, Trophic effects...........................................................54, 55
172, 173, 182, 183, 223–258, 291
Osteoblasts ................................4, 5, 7, 8, 10, 11, 30, 36, V
89, 96, 157, 276
Osteocytes ...........................................3–12, 97, 198, 276 Von Frey .......................................................292–296, 298