Effect of interval-training exercise on subchondral bone in a chemically-induced osteoarthritis model

Accepted Manuscript Effect of interval training exercise on subchondral bone in a chemically-induced osteoarthritis model Boudenot Arnaud, Presle Nathalie, Uzbekov Rustem, Toumi Hechmi, Pallu Stéphane, Lespessailles Eric PII:

S1063-4584(14)01099-1

DOI:

10.1016/j.joca.2014.05.020

Reference:

YJOCA 3163

To appear in:

Osteoarthritis and Cartilage

Received Date: 23 October 2013 Revised Date:

20 May 2014

Accepted Date: 28 May 2014

Please cite this article as: Arnaud B, Nathalie P, Rustem U, Hechmi T, Stéphane P, Eric L, Effect of interval training exercise on subchondral bone in a chemically-induced osteoarthritis model, Osteoarthritis and Cartilage (2014), doi: 10.1016/j.joca.2014.05.020. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT EFFECT OF INTERVAL TRAINING EXERCISE ON SUBCHONDRAL BONE IN A CHEMICALLY-INDUCED OSTEOARTHRITIS MODEL

Stéphane1* and LESPESSAILLES Eric1,4*

RI PT

BOUDENOT Arnaud1ᶲ, PRESLE Nathalie2, UZBEKOV Rustem3, TOUMI Hechmi1, PALLU

EA 4708 I3MTO, University of Orléans, Orléans, France.

2

UMR 7365 CNRS – Universite de Lorraine, Vandoeuvre-les-Nancy, France.

3

Department of Microscopy, University of Francois Rabelais, Tours, France.

4

Service de rhumatologie, centre hospitalier régional d’Orléans, Orléans, France.

M AN U

SC

1

* both senior authors equally contributed to this work ᶲ Corresponding author:

EP

TE D

Dr BOUDENOT Arnaud [email protected] IPROS - EA 4708 I3MTO CHR ORLEANS - BP 2439 1 rue Porte Madeleine 45032 ORLEANS Cedex 1 France Tel: +33 (0)2 38 74 40 25 – Fax: +33 (0)2 38 74 40 24

This article contains supplementary materials (tables and figures). All authors state that they

AC C

have no conflicts of interest.

Subchondral bone in OA: exercise effects

ACCEPTED MANUSCRIPT

ABSTRACT: Objectives

RI PT

The role of subchondral bone in osteoarthritis (OA) development is well admitted. Cross-talk between subchondral bone and cartilage may be disrupted in OA, leading to altered subchondral bone remodeling. Osteocytes are involved in bone remodeling control and could

SC

play a key role in OA progression. Our purpose of this study was to evaluate the preventive effect of interval-training exercise on subchondral bone and osteocyte in monosodium

M AN U

iodoacetate (MIA) model of experimental OA.

Methods

At baseline, 48 male Wistar rats (8 weeks old) were separated into 2 groups: interval-training exercise or no exercise for 10 weeks. After this training period, each group was divided into 2

TE D

subgroups: MIA-injected knee (1 mg/100 µl saline) and saline-injected knee. Four weeks later, rats were sacrificed and carefully dissected. Evaluated parameters were: cartilage degeneration by OA scoring, bone mineral density (BMD) by Dual energy X-ray

EP

Absorptiometry, trabecular subchondral bone microarchitecture by micro-computed

AC C

tomography, cortical subchondral bone lacunar osteocyte occupancy (by Toluidine Blue staining) and cleaved caspase-3 positive apoptosis (by epifluorescence).

Results

Our results showed deleterious effects of MIA on cartilage. OA induced a decrease in proximal tibia BMD which was prevented by exercise. Exercise induced increase in full osteocyte lacunae surface and osteocyte occupancy (+60%) of cortical subchondral bone

Subchondral bone in OA: exercise effects

ACCEPTED MANUSCRIPT independently of OA. Osteocyte apoptosis (<1%) in cortical subchondral bone was not different whatever the group at sacrifice.

Conclusion

RI PT

Our results suggest that preliminary interval-training improved BMD and osteocytes lacunar occupancy in subchondral bone. Our interval-training did not prevent MIA-induced cartilage

SC

degeneration.

Key words: bone mineral density, microarchitecture, knee, interval-training, osteoarthritis,

AC C

EP

TE D

M AN U

subchondral bone

Subchondral bone in OA: exercise effects

ACCEPTED MANUSCRIPT

INTRODUCTION

2

It has been proposed that the subchondral bone may contribute to the initiation and

3

progression of OA1. It undergoes numerous changes during OA such as osteophyte formation,

4

sclerosis, bone marrow lesions and cysts, trabecular thinning and undermineralized bone

5

matrix due to hyper-remodeling2,3. Although OA is not solely a mechanically-induced joint

6

disorder, cartilage and subchondral bone interaction has been shown to play an important role

7

in OA pathology4. Crosstalk between cartilage and subchondral bone may be disrupted both

8

by specific lesions of subchondral bone5 but also by initial cartilage damage6.

SC

RI PT

1

M AN U

9

Dysregulations of the cells implicated in the subchondral bone homeostasis (osteoblast,

11

osteoclast and osteocyte) occur during OA. Pro-inflammatory cytokines and growth factors

12

including IL-1β, IL-6, IGF-1 and TGF-β are expressed in osteoblasts during the initiation

13

and/or the progression of OA10. Moreover, the expression and the activity of alkaline

14

phosphatase and osteocalcin levels are upregulated in OA subchondral osteoblasts10. Although

15

the role of osteoclasts in the pathogenesis of OA remains uncertain, the expression of RANK-

16

L in the joint could contribute to subchondral bone resorption2. Higher levels of TRAP

17

expression have been observed in subchondral bone of osteoarthritic knees compared to non-

18

OA knees11. The osteocyte has been qualified as an orchestrator of bone remodeling7 and

19

might be largely involved in OA-induced subchondral bone changes through dysregulation of

20

osteoblast and osteoclast. The osteocyte could have an important role in OA due to its

21

implication in the Wnt pathway in regulating bone formation8. In addition Wnt pathway

22

signaling has been shown to have a crucial implication in cartilage destruction in OA through

23

promotion of chondrocyte hypertrophy9. In 2009, van Hove et al. have shown a specific large

24

and elongating morphology for OA osteocytes12. More recently, it has been demonstrated that

25

there are more apoptotic subchondral osteocytes in severe OA patients than in control

AC C

EP

TE D

10

1

ACCEPTED MANUSCRIPT 26

subjects11. However, the paucity of data related to subchondral bone osteocytes biology limits

27

our understanding of OA mechanisms, particularly in the early stages of the disease.

28 The osteocyte is considered as a mechanosensing cell able to treat mechanical loading signals

30

by delivering biochemical response in order to control bone remodeling8. Indeed, in rat

31

osteopenia models, it has been shown that exercise reduced apoptosis13 and improved

32

osteocyte viability14. We hypothesize that physical exercise may regulate subchondral bone

33

homeostasis and may prevent OA initiation.

34

The main purpose of this study was therefore to evaluate the potential preventive effect of

35

interval-training exercise on the subchondral bone changes in the well-established

36

monosodium iodoacetate (MIA) model of experimental OA. It has been observed that both

37

medial and lateral compartments are distinct from each other in pathophysiology like OA15,16.

38

Therefore, we wanted to investigate the microarchitecture responses due to OA and/or

39

exercise in these both compartments.

40

MATERIALS AND METHODS

41

Animals

42

Forty eight male Wistar rats, aged 8 weeks were received from Janvier breeding (Le Genest St

43

Isle, France) and were randomly numbered and placed two per cage. They had 12h of

44

dark/light cycle, at constant temperature (21 ± 2°C). They were allowed to eat and drink ad

45

libitum. During all the experiment, rats were allowed to move freely in standardized cages.

46

For in vivo analysis, rats were anesthetized with pentobarbital diluted in sodium chloride

47

(50/50% v/v and 0.1 ml per 100 g of body weight). At the end of the study, rats were

48

anesthetized with pentobarbital sodium (0.1 ml per 100 g of body weight) and then sacrificed

49

by cardiac exsanguinations. The procedure for the care and euthanizing of the animals was in

50

accordance with the European Community standards on the care and use of laboratory

AC C

EP

TE D

M AN U

SC

RI PT

29

2

ACCEPTED MANUSCRIPT animals. The study was approved by a board institution and an ethics committee (agreement

52

n°C45-234-9 and 2011-11-2) from the French Institute INSERM (Institut National de la Santé

53

et de la Recherche Médicale) and from the agriculture council (Ministère de l’agriculture,

54

France, approval ID: INSERM45-001).

55

Experimental schedule

56

Rats were acclimated for 2 weeks and were then randomly separated into 2 groups: Exercise

57

(Ex) and Non-Exercise (NEx) for 10 weeks. During these 10 weeks, Ex group was submitted

58

to a treadmill running interval training procedure. At the end of these 10 weeks, each group

59

(Ex and NEx) was divided into 2 subgroups i.e. control (NEx-Ctrl and Ex-Ctrl) and chemical

60

induced OA rats (NEx-MIA and Ex-MIA) (Fig 1). An additional week was devoted to the

61

achievement of both densitometry scans and knee injection. After 4 weeks, rats were

62

sacrificed (week 15).

63

Exercise protocol

64

Rats in the Ex group (n=24) were subjected to interval-training treadmill exercise for 10

65

weeks, 5 days/week, 1 h/day. Training protocol has been previously described for its positive

66

effects on bone, Maximal Aerobic Speed (MAS) determination has been also described17.

67

During the first week, session duration progressively increased from 20 to 60 min. Briefly, the

68

training program included 7 cycles as follows: 5 min of moderate speed at 50% of MAS (from

69

8 to 15 m/min), 2min of intensive running at 80% of MAS (from 14 to 24 m/min) and 1 min

70

of passive recovery. This program was repeated 7 fold and the training session ended with an

71

active recovery phase (4 min at 50% of MAS).

72

Induction of OA

73

OA was induced by a single intra-articular injection into the right knee joint of 1 mg of MIA18

74

dissolved in 100 µl of NaCl. Control rats received an injection 100 µl of NaCl into the right

AC C

EP

TE D

M AN U

SC

RI PT

51

3

ACCEPTED MANUSCRIPT 75

knee. NaCl was 0.9% solution. Left knee remained un-injected. Rats were sacrificed four

76

weeks after injection.

77 We have allocated different depths of the subchondral bone collected from the right knee joint

79

to evaluate changes in cartilage and bone as mentioned in Fig 2. Transversal sections were

80

employed from the upper part (proximal) to the downer part (distal) for: 1)cartilage

81

macroscopy and histology (1500 to 2000 µm), 2)bone histology (500 µm) and

82

3)immunostaining detected by epifluorescence (500 µm).

83

Dual energy X-ray absorptiometry

84

Body composition and bone mineral density

85

All animals were scanned at baseline, after training and at the end of the protocol. Scans were

86

analyzed using small animal mode on Hologic Discovery (Hologic, Waltham, MA, USA)

87

(pixel area resolution: 640 µm²) at the whole body and lower limbs for body composition

88

(total mass, lean mass and fat mass) and for BMD in order to control the exercise effect (Fig

89

3A).

90

Epiphyseal bone mineral density

91

Dual energy X-ray Absorptiometry (DXA) was achieved by Hologic Discovery at baseline, at

92

the end of the 10weeks training period and at sacrifice 4 weeks after OA induction. All animal

93

scans were analyzed with small animal software using high resolution specific software (pixel

94

area resolution: 311 µm²) to measure the epiphyseal subchondral BMD of proximal tibia (PT)

95

and distal femur (DF) (Fig 3B). The analysis method was developed in our lab by separating

96

regions of interest (ROIs) particularly involved in knee osteoarthritis.

97

The ROIs were set as follows:

AC C

EP

TE D

M AN U

SC

RI PT

78

4

ACCEPTED MANUSCRIPT 98

-

PT: a pentagon was drawn manually. Two sides were above the tibial plateaus with the

99

top located on inter-condylar spines. Two other sides were around the trabecular

100

subchondral bone beyond growth plate located near the fibula head. The mean area

101

(95% confidence intervals) was 32.9 (31.4, 34.4) mm². -

DF: a second pentagon was drawn manually. Two sides were above the femoral

RI PT

102

condyles. A third side separated the femoral condyles from patella. The most proximal

104

line made the straight line in the continuity of the posterior condyles and the last line.

105

The mean area was 34.1 (32.5, 35.7) mm².

SC

103

The intra-observer (AB) reproducibility expressed as root mean square coefficient of variation

107

(RMS CV) was 2.01% and 1.90%, respectively for PT and DF.

108

Data for non injected knees are not shown due to similarity with NEx-Ctrl group.

109

Trabecular subchondral bone micro-architecture

110

At sacrifice, tibiae were carefully dissected and placed in formalin (4% v/v) for 48 h. Then,

111

samples were placed in ethanol (50% v/v). In order to determine trabecular subchondral bone

112

microarchitecture, each bone was scanned by micro-computed tomography (µCT) from

113

Skyscan (Skyscan 1072, Skyscan, Kontich, Belgium). The X-ray source was set at 80 kV and

114

100 µA. Step of 0.45° on a total angular range of 180° with a time exposure set at 3.6 s

115

provided a 2h30 min acquisition (pixel size: 12.02 µm).

116

We have created two specific ROIs in the trabecular subchondral bone: medial and lateral

117

(Fig 2).

118

We have used cylinders adaptated to the work by McErlain et al.30, in order to obtain a

119

maximal volume of interest and we optimized the circle surface with 1.76mm of diameter.

120

The trabecular subchondral bone thickness (µm) was measured in each cylinder. Bone volume

121

fraction BV/TV (%), the trabecular thickness Tb.Th (mm), the trabecular spacing Tb.Sp (mm)

122

and the trabecular number Tb.N (1/mm) were calculated19.

AC C

EP

TE D

M AN U

106

5

ACCEPTED MANUSCRIPT

Cartilage examination

124

Macroscopy

125

Macroscopic views of the tibial cartilage were recorded using Epson scan software (version

126

3.04F) on Epson Perfection 4870 photo (Seiko Epson Corporation, Nagano, Japan). The

127

resolution was set at 2400 DPI. Macroscopy was only used for qualitative information on the

128

cartilage degeneration.

129

Histology and OA scoring

130

Cartilage samples (n=28) were fixed for 24h in 4% paraformaldehyde immediately after

131

removal, decalcified in rapid bone decalcifier (RDO, Apex, Canada) for 8h and further fixed

132

in 4% paraformaldehyde. Specimens were then dehydrated in a graded serie of alcohol and

133

embedded in paraffin. Safranin-O-fast green staining was performed on 5µm serial sagittal

134

sections for histologic grading. Images were acquired with Axioplan (Zeiss, Germany)

135

microscope equipped with 11.2. Color Mosaic camera using x10 and x40 magnification.

136

The severity of OA lesions was graded using two OA scores: Mankin20 and OARSI21 by two

137

blinded and independent observers (AB and SP).

138

Cortical subchondral bone histology

139

Bone explants preparation was done as previously described22 and adapted for epiphyseal

140

proximal tibia. We employed the Toluidine bleu staining method described by our team23.

141

Images were acquired with Axioplan (Zeiss, Germany) microscope equipped with 11.2. Color

142

Mosaic camera using x40 magnification.

143

Images were analyzed using Image J software (v1.46, National institutes of health, USA).

144

Characteristic semi thin sections of cortical subchondral bone stained with Toluidine blue are

145

presented in Fig 4. The lacunae density was calculated as the total number of lacunae per

AC C

EP

TE D

M AN U

SC

RI PT

123

6

ACCEPTED MANUSCRIPT mm². The lacunae surface and the percentage of osteocyte lacunar occupancy were also

147

calculated.

148

Immunolabelling of cleaved caspase-3 in cortical subchondral bone

149

Immunolabelling of cleaved caspase-3 epifluorescence microscopy method23,24 was used to

150

visualise nuclei in live osteocytes in situ, the bone explants were also incubated with DAPI

151

(Santa Cruz Biotechnology, Santa Cruz, USA).

152

Epifluorescence images of osteocytes were obtained by using a Motic AE21 video camera

153

attached to a Microvision microscope. Then, bright field was employed for morphologic

154

osteocytes identification in each region of interest. The total magnification (ocular lens and

155

objective) was x100 (Supplementary Fig 1). The percentage and the absolute density

156

(n/mm²) of apoptotic osteocytes stained with cleaved caspase-3 for each sample were

157

determined by 2 observer agreement (AB and SP).

158

Statistical analysis

159

The normality of the distributions was assessed with the Shapiro-Wilk test using XLStat-2009

160

and the homogeneity of the variances with Fisher F-test using Statview 5.0 (SAS Institute

161

Inc). When the distribution of the data for each group followed the Gaussian distribution and

162

the variances of both groups were homogeneous, parametric tests were used. When the

163

distribution of at least one group did not follow the normal law or variances homogeneity,

164

non-parametric tests were used. As parametric tests, ANOVA followed by Fisher’s PLSD

165

post hoc tests were employed on some parameters (OARSI and Mankin scores, lacunae

166

density, osteocyte occupancy, percentage of osteocyte apoptosis). Two-way ANOVA was

167

employed for histology lacunar surface using occupancy factor and group factor. Because we

168

found an interaction, one-way ANOVA was employed for each factor. The non parametric

169

Kruskall-Wallis test followed by Mann-Withney U test (inter-group) and Wilcoxon test (intra-

170

group) were employed for all DXA measurements, for microarchitecture parameters, for

AC C

EP

TE D

M AN U

SC

RI PT

146

7

ACCEPTED MANUSCRIPT osteocytes apoptosis density. Time effect on whole body DXA measurements were observed

172

using non parametric Friedman test followed by Wilcoxon test. Data were presented as mean

173

and 95% confidence interval. The reproducibility was evaluated using a specific equation for

174

2 measures of root mean square coefficient of variation (RMS CV). Significance level was

175

fixed at p<0.05.

RI PT

171

176

RESULTS

178

Dual energy X-ray absorptiometry

179

Body composition

180

At baseline, all whole body composition parameters were similar between groups

181

(Supplementary Table 1). At the end of exercise, weight and fat mass at both whole body

182

and right lower limb were statistically lower in exercised groups than non-exercised groups.

183

The whole body lean mass was similar between groups. However, the right lower limb lean

184

mass was significantly higher in exercised groups compared with non-exercise groups

185

(p=0.0354). Four weeks after MIA or saline knee injection, only fat mass was still different

186

between groups in both whole body and right lower limb.

187

Increase in whole body composition and BMD over time due to the normal growth of the rats

188

was observed (Supplementary Table 1 and Table 1, respectively).

189

Bone mineral density

190

At sacrifice, we have found a loss of BMD in the NEx-MIA group at the right side proximal

191

tibia compared to NEx-Ctrl. Ex-MIA had a higher BMD compared to NEx-MIA (Table 1).

192

Trabecular subchondral bone microarchitecture

193

Comparing medial and lateral plateaus, our data indicated for each group that the medial part

194

of the knee joint exhibited approximately 10% higher values of Tb.Th than the lateral part

AC C

EP

TE D

M AN U

SC

177

8

ACCEPTED MANUSCRIPT (Table 2). Similarly, we observed the same significant effect on BV/TV (+15 to +26% in

196

medial) for all groups except Ex-Ctrl group. Tb.N was not significantly different between

197

medial and lateral plateaus for all groups. Medial Tb.Sp was 10% lower compared to lateral

198

Tb.Sp only in NEx-Ctrl. Trabecular subchondral bone thickness was significantly higher in

199

NEx-MIA lateral plateau compared to medial plateau (+13%).

200

Comparing groups in the medial plateau, microarchitecture parameters were not different

201

among exercise and OA conditions.

202

Comparing groups in the lateral plateau, significant lower BV/TV, Tb.N and Tb.Th values

203

were found in Ex-MIA vs Ex-Ctrl (-22%, -16% and -9%, respectively). We did not evidence

204

such differences on Tb.Sp. We did not find any difference between NEx-MIA and Ex-MIA

205

for all parameters (Table 2).

206

Cartilage examination

207

All MIA-injected rat knees showed articular cartilage degradation on tibial plateau (Fig 5).

208

Proteoglycan loss and fibrillation of the cartilage surface were increased in MIA-injected rats

209

when compared to saline-injected control rats. These observations were confirmed regarding

210

OA scores.

211

Both NEx-MIA and Ex-MIA exhibited significant higher Mankin and OARSI scores than

212

both NEx-Ctrl and Ex-Ctrl (supplementary Table 2).

213

Cortical subchondral bone histology

214

Cortical subchondral bone lacunae density was similar among the four groups (Table 3). Full

215

osteocyte lacunae surface and occupancy levels were significantly higher in Ex-Ctrl and Ex-

216

MIA when compared to NEx-Ctrl and NEx-MIA (Table 3). NEx-MIA empty osteocyte

217

lacunae surface was smaller than other groups. NEx-Ctrl empty lacunae surface was also

218

significantly lower than Ex-MIA. Finally, whatever the group, empty osteocyte lacunae had

219

significantly smaller surface than full osteocyte lacunae at sacrifice (Table 3).

AC C

EP

TE D

M AN U

SC

RI PT

195

9

ACCEPTED MANUSCRIPT

Cortical subchondral bone immunolabelling of cleaved caspase-3

221

There was no statistical difference between groups regarding the cleaved caspase-3 labelling,

222

as an early marker of apoptosis, on cortical subchondral bone of tibia (Supplementary Fig 1).

223

Indeed, at sacrifice, the percentage of apoptotic osteocyte was 0.35% (0, 0.85%) for NEx-Ctrl,

224

0.52% (0, 1.14%) for NEx-MIA, 0.76% (0.09, 1.43%) for Ex-Ctrl and 0.59% (0, 1.19%) for

225

Ex-MIA. The absolute apoptotic density (n/mm²) was not different between groups: 8.9 (0,

226

26.6) for NEx-Ctrl, 14.8 (0, 35.8) for NEx-MIA, 17.4 (0, 39.6) for Ex-Ctrl and 11.2(0, 24.5)

227

for Ex-MIA.

228

DISCUSSION

229

We demonstrated for the first time that an aerobic interval-training exercise may prevent the

230

decrease in proximal tibia BMD induced by OA condition. In addition, an increase of

231

osteocyte full lacunae surface and osteocyte lacunar occupancy of cortical subchondral bone

232

was observed in exercise groups with or without OA.

SC M AN U

TE D

233

RI PT

220

It has been reported that an exercise performed immediately after either chemical25 or

235

mechanical26 OA could potentially protect the knee cartilage from OA progression. These last

236

studies neither investigated the subchondral bone mineral density nor the subchondral

237

trabecular bone microarchitecture. Herein, we reported that a 10-week interval training (12

238

m/min for moderate intensity and 20 m/min for high intensity, 60 min/day) before knee injury

239

had no preventive effect on cartilage but protected the sBMD. These divergences compared to

240

the 2 last studies25,26 may be due to 1)protocol difference with, on one hand, an exercise

241

performed immediately after knee injury25,26 and, on the other hand, an exercise performed

242

before injury; 2)duration difference, we imposed a running lasting 60 min/day while 30

243

min/day seems to produce better cartilage protection26; 3)modality difference, we employed

244

an interval-training protocol (which produce positive effects on bone17,27, but its effects on

AC C

EP

234

10

ACCEPTED MANUSCRIPT 245

cartilage were not previously investigated) whereas the other teams employed continuous

246

training25,26. Recently, Siebelt et al. have shown that moderate continuous training in Wistar

247

rats does not protect OA progression but further enhance papain-induced OA progression28.

248

RI PT

249

At sacrifice, in NEx-MIA, among all anatomical sites and time points, sole the specifically

251

localized proximal tibia sBMD at time of the sacrifice decreased. Age-related changes in bone

252

and cartilage might be a confounding factor in our OA animal model, we did observe an

253

increase in whole body weight and BMD over time due to normal growth of the rats.

254

However, since rats were of the same age, and as we compared age-matched groups in our

255

study, we cannot suspect a confounding role of the age. This sBMD decrease was observed

256

only in NEx-MIA group, suggesting, on one hand, a deleterious effect of MIA and, on the

257

other hand, a preventive effect of exercise since Ex-MIA did not present such sBMD loss. We

258

have observed a specific response of sBMD in the tibia with OA. Strassle et al. have studied

259

longitudinal in vivo sBMD in MIA induced-OA in growing rats18. Using DXA, they have

260

shown that rats submitted to higher doses of MIA (1.0 and 3.0 mg) had a significant lower

261

sBMD gain compared to lower dose (0.3 mg) and controls. However, distal femur and

262

proximal tibia were analyzed as only one ROI corresponding to the knee18. Here we present

263

discrimination between proximal tibia and distal femur responses.

M AN U

TE D

EP

AC C

264

SC

250

265

It has been observed in human OA that BV/TV, Tb.N and Tb.Th were higher in medial

266

plateau than lateral plateau29. In our work, we wanted to separate the two plateaus in order to

267

discriminate the same 3 parameters. The difference between medial and lateral trabecular

268

bone microarchitecture was observed thanks to a method modified from McErlain et al.15.

11

ACCEPTED MANUSCRIPT Comparing groups, differences were seen only in the lateral part. Alterations of the lateral

270

trabecular bone microarchitecture was observed in Ex-MIA when compared to Ex-Ctrl. NEx-

271

MIA and Ex-MIA trabecular microarchitecture were not different indicating no preventive

272

effect of exercise on this parameter. The location of differences in the lateral plateau was

273

unexpected since subchondral bone changes occurs more frequently and in a larger extent in

274

medial plateau in both dogs5 and human6,30-32. However, subchondral bone changes have also

275

been identified in both human plateaus33.

SC

276

RI PT

269

The preventive effect of exercise reported on the epiphyseal proximal tibia sBMD was not

278

associated with trabecular subchondral bone microarchitecture changes. Thus, the positive

279

effect of exercise observed by DXA on the tibial whole epiphysis might be explained by the

280

results on cortical subchondral bone. In OA animal models, changes have been shown on

281

subchondral plate thickness in tibial plateaus6,16,34. However, these last studies did not

282

investigate the preventive effect of exercise. In the present study, we found an exercise effect

283

on cortical subchondral bone with higher surface in full osteocyte lacunae and more than 60%

284

of lacunar osteocyte occupancy compared to NEx groups. The empty lacunae surface was

285

significantly lower in NEx-MIA versus all other groups. Moreover, NEx-MIA obtained

286

significantly lower tibial sBMD than other groups. Thus, we can suggest that such lower

287

surface of empty osteocyte lacunae could be involved in a local cortical subchondral bone

288

mineralization decrease (-5%). Indeed, osteocytes are known to regulate mineralization

289

around their lacunae7,35. We did not find literature regarding the discrepancy between full and

290

empty lacunae surface. Although enlargement of osteocyte lacunae was reported in models of

291

osteopenia23,36,37, to our knowledge, no study demonstrates an exercise effect on osteocyte

292

lacunae surface. Here, we report that osteocyte full lacunae were larger than empty lacunae in

293

all groups. Moreover, osteocyte full lacunae were larger in trained groups compared to

AC C

EP

TE D

M AN U

277

12

ACCEPTED MANUSCRIPT untrained groups. Differences in surface may represent osteocyte viability and its capacity to

295

modify the surrounding matrix. It has been postulated that osteocyte lacunae enlargement may

296

be an important factor of how cells perceive mechanical loadings38. The implication of

297

mechanical loadings on osteocyte response is plausible since we observed an exercise effect

298

on lacunae surface. However, we do not believe that large lacunae are systematically more

299

exposed to bone fragility and to an increase risk of crack. In fact, additionally to larger surface

300

in full lacunae than empty lacunae in all groups, we found higher osteocyte lacunae surface

301

and osteocyte occupancy in trained groups.

SC

RI PT

294

302

We can consider this increase in lacunar osteocyte occupancy as a positive effect since Qiu et

304

al. have reported that osteocyte occupancy was inversely correlated with bone microcracks39.

305

So, osteocyte occupancy was described as a determinant of bone quality39. In horses with

306

navicular syndrome, arthritic degeneration was associated to low bone mass, poor osteocyte

307

connectivity and low osteocyte density40. Moreover, Jaiprakash et al. observed that

308

mineralization was controlled by osteocyte in OA subchondral bone. These authors

309

characterized the tibial subchondral bone plate in patients with severe OA11. They did not

310

show any difference in the lacunar density between groups but significant higher empty

311

lacunae density in OA patients compared to controls. That led to approximately 40% and 60%

312

of lacunar osteocyte occupancy in OA and controls, respectively11. In our study, we did not

313

find OA effect on cortical subchondral bone. However, a beneficial effect of exercise on

314

occupancy was shown, as it was recently described in a model of osteopenia in rats on another

315

bone site14. In the latter work, voluntary running wheel exercise was performed in order to

316

counteract the bone loss induced by ovariectomy (OVX)14. Regarding femur cortical bone

317

diaphysis, authors have found higher lacunar density and osteocyte occupancy in trained

AC C

EP

TE D

M AN U

303

13

ACCEPTED MANUSCRIPT 318

groups compared to untrained groups. The osteocyte occupancy was approximately 40% in

319

untrained groups versus 60% in trained groups.

320 Our cortical subchondral bone histology results would suggest that osteocytes’ death might be

322

higher in NEx groups since lacunar osteocyte occupancy was significantly lower compared to

323

Ex groups. Surprisingly, in our study, osteocyte apoptosis was not different between groups,

324

suggesting that other cell death mechanisms may be involved such as autophagy41 or

325

osteonecrosis42,43.

326

Another potential explanation for the high lacunar osteocyte occupancy in Ex groups could be

327

a lower osteocyte apoptosis during the interval-training running leading to an elevated number

328

of osteocyte occupancy just before OA induction. This lower osteocyte apoptosis might have

329

lead to a low osteocyte death the following weeks. During the following weeks, osteocyte

330

death rate returns progressively to NEx-Ctrl values while training is stopped. Apoptosis is a

331

dynamic mechanism and is evaluated by determining the expression of cleaved caspase-3 at a

332

definite time point. Besides, the lacunar osteocyte occupancy measured upon histologic

333

analysis represents long-term consequence of cellular events. These two parameters have

334

different kinetics. Van Essen et al. have observed no difference in iliac bone apoptosis despite

335

higher lacunar osteocyte occupancy in osteoporosis treated women versus placebo44.

SC

M AN U

TE D

EP

AC C

336

RI PT

321

337

The strengths of this study were, on one hand, the multi-scale and the multi-modal analysis of

338

the subchondral bone and, on the other hand, the combined analysis of OA and exercise

339

effects. It is important to note that the absence of animal sacrifice at each time point and the

340

absence of synovial fluid analysis limit some interpretation.

341

14

ACCEPTED MANUSCRIPT 342

In summary, the major finding of this work was the positive effect of exercise on knee

343

subchondral bone mineral density and on osteocyte lacunar size and occupancy. It is

344

important to note that our interval-training running was not able to prevent cartilage

345

degeneration.

RI PT

346

ACKNOWLEDGEMENTS

348

Arnaud BOUDENOT obtained a young investigator award in Musculoskeletal Rehabilitation

349

in Patients with Osteoporosis working group at the American Society for Bone and Mineral

350

Research (ASBMR) 2010 Annual Meeting in Toronto, for the preliminary results.

351

The authors thank Eric DOLLEANS from I3MTO laboratory, for animal experiment and

352

technical assistance and Cécile GUILLAUME from UMR 7365 CNRS for technical

353

assistance. We thank Gérard NOYER from University of Orléans for technical help on

354

treadmill for rats training and Philippe MOREAU from the laboratory of neurobiology in

355

Orléans for the housing of the rats. We also are grateful to Dr Sandrine VILLETTE from

356

Centre de Biophysique Moléculaire, CNRS, Orléans, France, for ELISA tests, and to Sonia

357

GEORGEAULT from Department of Microscopy, University of Francois Rabelais, Tours,

358

France, for Technical assistance in samples preparation. Finally, we thank master students

359

from University of Orléans for technical contribution: Anamarija MARKOTA, Amine

360

BOUCHAREB and Souria ATSI.

361

The authors would like to express their gratitude towards Dr Hugues PORTIER and Dr

362

Antonio PINTI from I3MTO laboratory for reviewing the manuscript.

AC C

EP

TE D

M AN U

SC

347

363

15

ACCEPTED MANUSCRIPT

CONTRIBUTORS

365

Study design: SP. Study conduct: AB. Data collection: AB, NP, RU, SP. Data analysis: AB.

366

Data interpretation: AB, EL, SP. Drafting manuscript: AB, SP. Revising manuscript content:

367

AB, EL, HT, NP, RU, SP. Approving final version of manuscript: AB, EL, HT, NP, RU, SP.

368

AB had full access to all data and takes responsibility for the integrity of the data analysis.

RI PT

364

369

FUNDING

371

Arnaud BOUDENOT is a Ph.D. student at the University of Orléans and is funded by Conseil

372

Général du Loiret, France, D3A Medical systems, Orléans, France, and University of Orléans,

373

France. This work was supported by grants from INSERM, France, from CNRS program

374

“longévité vieillissement” project, France, and IBIFOS 388N, Région Centre and FEDER

375

grant.

M AN U

SC

370

TE D

376

377

CONFLICT OF INTEREST STATEMENT

378

Authors have no conflict of interest.

EP

379

REFERENCES

381 382 383 384 385 386 387 388 389 390 391

1.

2.

3. 4.

AC C

380

Henrotin Y, Pesesse L, Sanchez C. Subchondral bone in osteoarthritis physiopathology: state-of-the art and perspectives. Biomed Mater Eng 2009; 19: 311316. Karsdal MA, Bay-Jensen AC, Lories RJ, Abramson S, Spector T, Pastoureau P, et al. The coupling of bone and cartilage turnover in osteoarthritis: opportunities for bone antiresorptives and anabolics as potential treatments? Ann Rheum Dis 2014; 73: 336348. Buckland-Wright C. Subchondral bone changes in hand and knee osteoarthritis detected by radiography. Osteoarthritis Cartilage 2004; 12 Suppl A: S10-19. Lories RJ, Luyten FP. The bone-cartilage unit in osteoarthritis. Nat Rev Rheumatol 2011; 7: 43-49.

16

ACCEPTED MANUSCRIPT

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

RI PT

10.

SC

9.

M AN U

7. 8.

TE D

6.

Blair-Levy JM, Watts CE, Fiorentino NM, Dimitriadis EK, Marini JC, Lipsky PE. A type I collagen defect leads to rapidly progressive osteoarthritis in a mouse model. Arthritis Rheum 2008; 58: 1096-1106. Intema F, Hazewinkel HA, Gouwens D, Bijlsma JW, Weinans H, Lafeber FP, et al. In early OA, thinning of the subchondral plate is directly related to cartilage damage: results from a canine ACLT-meniscectomy model. Osteoarthritis Cartilage 2010; 18: 691-698. Bonewald LF. The amazing osteocyte. J Bone Miner Res 2011; 26: 229-238. Bonewald LF, Johnson ML. Osteocytes, mechanosensing and Wnt signaling. Bone 2008; 42: 606-615. Luyten FP, Tylzanowski P, Lories RJ. Wnt signaling and osteoarthritis. Bone 2009; 44: 522-527. Lajeunesse D. Altered subchondral osteoblast cellular metabolism in osteoarthritis: cytokines, eicosanoids, and growth factors. J Musculoskelet Neuronal Interact 2002; 2: 504-506. Jaiprakash A, Prasadam I, Feng JQ, Liu Y, Crawford R, Xiao Y. Phenotypic characterization of osteoarthritic osteocytes from the sclerotic zones: a possible pathological role in subchondral bone sclerosis. Int J Biol Sci 2012; 8: 406-417. van Hove RP, Nolte PA, Vatsa A, Semeins CM, Salmon PL, Smit TH, et al. Osteocyte morphology in human tibiae of different bone pathologies with different bone mineral density--is there a role for mechanosensing? Bone 2009; 45: 321-329. Maurel DB, Boisseau N, Pallu S, Rochefort GY, Benhamou CL, Jaffre C. Regular exercise limits alcohol effects on trabecular, cortical thickness and porosity, and osteocyte apoptosis in the rat. Joint Bone Spine 2013; 80: 492-498. Fonseca H, Moreira-Goncalves D, Esteves JL, Viriato N, Vaz M, Mota MP, et al. Voluntary exercise has long-term in vivo protective effects on osteocyte viability and bone strength following ovariectomy. Calcif Tissue Int 2011; 88: 443-454. McErlain DD, Appleton CT, Litchfield RB, Pitelka V, Henry JL, Bernier SM, et al. Study of subchondral bone adaptations in a rodent surgical model of OA using in vivo micro-computed tomography. Osteoarthritis Cartilage 2008; 16: 458-469. Mohan G, Perilli E, Kuliwaba JS, Humphries JM, Parkinson IH, Fazzalari NL. Application of in vivo micro-computed tomography in the temporal characterisation of subchondral bone architecture in a rat model of low-dose monosodium iodoacetate induced osteoarthritis. Arthritis Res Ther 2011; 13: R210. Boudenot A, Pallu S, Lespessailles E, Jaffré C. Effects of an interval-training program on body composition and bone mineral density (BMD) in the rat. Sci sports 2012; 23: 188-191. Strassle BW, Mark L, Leventhal L, Piesla MJ, Jian Li X, Kennedy JD, et al. Inhibition of osteoclasts prevents cartilage loss and pain in a rat model of degenerative joint disease. Osteoarthritis Cartilage 2010; 18: 1319-1328. Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ, Muller R. Guidelines for assessment of bone microstructure in rodents using micro-computed tomography. J Bone Miner Res 2010; 25: 1468-1486. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and metabolic abnormalities in articular cartilage from osteo-arthritic human hips. II. Correlation of morphology with biochemical and metabolic data. J Bone Joint Surg Am 1971; 53: 523-537. Pritzker KP, Gay S, Jimenez SA, Ostergaard K, Pelletier JP, Revell PA, et al. Osteoarthritis cartilage histopathology: grading and staging. Osteoarthritis Cartilage 2006; 14: 13-29.

EP

5.

AC C

392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441

17

ACCEPTED MANUSCRIPT

27. 28.

29.

30. 31.

32.

33.

34.

35. 36.

37.

RI PT

26.

SC

25.

M AN U

24.

TE D

23.

Uzbekov RE, Maurel DB, Aveline PC, Pallu S, Benhamou CL, Rochefort GY. Centrosome fine ultrastructure of the osteocyte mechanosensitive primary cilium. Microsc Microanal 2012; 18: 1430-1441. Maurel DB, Jaffre C, Rochefort GY, Aveline PC, Boisseau N, Uzbekov R, et al. Low bone accrual is associated with osteocyte apoptosis in alcohol-induced osteopenia. Bone 2011; 49: 543-552. Vatsa A, Semeins CM, Smit TH, Klein-Nulend J. Paxillin localisation in osteocytes-is it determined by the direction of loading? Biochem Biophys Res Commun 2008; 377: 1019-1024. Cifuentes DJ, Rocha LG, Silva LA, Brito AC, Rueff-Barroso CR, Porto LC, et al. Decrease in oxidative stress and histological changes induced by physical exercise calibrated in rats with osteoarthritis induced by monosodium iodoacetate. Osteoarthritis Cartilage 2010; 18: 1088-1095. Galois L, Etienne S, Grossin L, Watrin-Pinzano A, Cournil-Henrionnet C, Loeuille D, et al. Dose-response relationship for exercise on severity of experimental osteoarthritis in rats: a pilot study. Osteoarthritis Cartilage 2004; 12: 779-786. Chen X, Aoki H, Fukui Y. Effect of exercise on the bone strength, bone mineral density, and metal content in rat femurs. Biomed Mater Eng 2004; 14: 53-59. Siebelt M, Groen HC, Koelewijn SJ, de Blois E, Sandker M, Waarsing JH, et al. Increased physical activity severely induces osteoarthritic changes in knee joints with papain induced sulphate-glycosaminoglycan depleted cartilage. Arthritis Res Ther 2014; 16: R32. Bobinac D, Spanjol J, Zoricic S, Maric I. Changes in articular cartilage and subchondral bone histomorphometry in osteoarthritic knee joints in humans. Bone 2003; 32: 284-290. Buckland-Wright JC, Lynch JA, Dave B. Early radiographic features in patients with anterior cruciate ligament rupture. Ann Rheum Dis 2000; 59: 641-646. Dore D, Quinn S, Ding C, Winzenberg T, Cicuttini F, Jones G. Subchondral bone and cartilage damage: a prospective study in older adults. Arthritis Rheum 2010; 62: 19671973. Clarke S, Wakeley C, Duddy J, Sharif M, Watt I, Ellingham K, et al. Dual-energy Xray absorptiometry applied to the assessment of tibial subchondral bone mineral density in osteoarthritis of the knee. Skeletal Radiol 2004; 33: 588-595. Wolski M, Podsiadlo P, Stachowiak GW, Lohmander LS, Englund M. Differences in trabecular bone texture between knees with and without radiographic osteoarthritis detected by directional fractal signature method. Osteoarthritis Cartilage 2010; 18: 684-690. Botter SM, van Osch GJ, Clockaerts S, Waarsing JH, Weinans H, van Leeuwen JP. Osteoarthritis induction leads to early and temporal subchondral plate porosity in the tibial plateau of mice: an in vivo microfocal computed tomography study. Arthritis Rheum 2011; 63: 2690-2699. Bonewald LF. Osteocytes as dynamic multifunctional cells. Ann N Y Acad Sci 2007; 1116: 281-290. Lane NE, Yao W, Balooch M, Nalla RK, Balooch G, Habelitz S, et al. Glucocorticoidtreated mice have localized changes in trabecular bone material properties and osteocyte lacunar size that are not observed in placebo-treated or estrogen-deficient mice. J Bone Miner Res 2006; 21: 466-476. Sissons HA, Kelman GJ, Marotti G. Mechanisms of bone resorption in calciumdeficient rats. Calcif Tissue Int 1984; 36: 711-721.

EP

22.

AC C

442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490

18

ACCEPTED MANUSCRIPT

40.

41. 42.

43.

44.

RI PT

39.

Lane NE, Yao W. Glucocorticoid-induced bone fragility. Ann N Y Acad Sci 2010; 1192: 81-83. Qiu S, Rao DS, Fyhrie DP, Palnitkar S, Parfitt AM. The morphological association between microcracks and osteocyte lacunae in human cortical bone. Bone 2005; 37: 10-15. Bentley VA, Sample SJ, Livesey MA, Scollay MC, Radtke CL, Frank JD, et al. Morphologic changes associated with functional adaptation of the navicular bone of horses. J Anat 2007; 211: 662-672. Hocking LJ, Whitehouse C, Helfrich MH. Autophagy: a new player in skeletal maintenance? J Bone Miner Res 2012; 27: 1439-1447. Humphreys S, Spencer JD, Tighe JR, Cumming RR. The femoral head in osteonecrosis. A quantitative study of osteocyte population. J Bone Joint Surg Br 1989; 71: 205-208. McErlain DD, Ulici V, Darling M, Gati JS, Pitelka V, Beier F, et al. An in vivo investigation of the initiation and progression of subchondral cysts in a rodent model of secondary osteoarthritis. Arthritis Res Ther 2012; 14: R26. van Essen HW, Holzmann PJ, Blankenstein MA, Lips P, Bravenboer N. Effect of raloxifene treatment on osteocyte apoptosis in postmenopausal women. Calcif Tissue Int 2007; 81: 183-190.

SC

38.

M AN U

491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511

AC C

EP

TE D

512

19

ACCEPTED MANUSCRIPT

FIGURE LEGENDS

514

Fig 1.

515

Experimental schedule. DXA: Dual energy X-ray Absorptiometry. MIA: Monosodium

516

Iodoacetate. NEx-Ctrl: non exercised rats receiving right knee NaCl injection; NEx-MIA: non

517

exercised rats receiving right knee monosodium iodoacetate injection; Ex-Ctrl: exercised

518

receiving right knee NaCl injection; Ex-MIA: exercised rats receiving right knee monosodium

519

iodoacetate injection.

RI PT

513

SC

520 Fig 2.

522

Schematic representation of tibial subchondral bone sections based on Micro Computed

523

Tomography (µCT) ex vivo scan (resolution: 12.02 µm). Transversal sections were employed

524

from the upper part (proximal) to the downer part (distal) for: 1)cartilage macroscopy and

525

histology (1500 to 2000 µm), 2)bone histology (500 µm) and 3)immunostaining detected by

526

epifluorescence (500 µm). All the epiphyseal bone was scanned in-vivo by DXA.

527

Medial and lateral regions of interest (ROI) are drawn in white cylinders. ROIs were cylinders

528

of 1.76 mm of diameter drawn using with “CT analyser” software (Skyscan). The diameter

529

was obtained in order to maximize the volume of the cylinder. Height was anatomically

530

determined from the growth plate until the epiphyseal cortical subchondral plate. Each

531

cylinder started from the upper (proximal) part under cartilage (containing only trabecular

532

subchondral bone) to the downer (distal) part (one section directly above the growth plate).

AC C

EP

TE D

M AN U

521

533 534

Fig 3.

535

Regions of interest on Dual energy X-ray Absorptiometry. Small animal scan (A) allowed to

536

analyze the rat whole body and right and left lower limb. High resolution small animal scan

537

(B) allowed to analyze the proximal tibia (R1) and the distal femur (R2) on the knee.

20

ACCEPTED MANUSCRIPT 538 Fig 4.

540

Cortical subchondral bone Toluidine blue staining of the proximal tibia observed with bright

541

field microscopy. White arrow indicates lacunae containing a cell, whereas black arrow

542

indicates empty lacunae. For each rat, 400 different cortical subchondral lacunae were

543

counted. The global ROIs were chosen to provide the best representation of the whole bone

544

slice. Moreover, in order to improve the representativeness of each sample, at least 4 global

545

ROIs were drawn for a total surface of about 0.731 mm² per rat. The osteocyte lacunae areas

546

were determined with the Image J freehand tool by drawing their contour line in the ROI.

547

Lacunae density, lacunae surface and osteocyte lacunar occupancy were measured in non

548

exercised rats receiving right knee NaCl injection (A); non exercised rats receiving right knee

549

monosodium iodoacetate injection (B); exercised receiving right knee NaCl injection (C) and

550

exercised rats receiving right knee monosodium iodoacetate injection (D). We observed

551

higher osteocyte lacunae surface in full lacunae (contening cell) than in empty lacunae (no

552

cell) for all groups. Full lacunae surface were higher in both Ex-Ctrl and Ex-MIA groups

553

compared to both NEx-Ctrl and NEx-MIA. Moreover, the osteocyte lacunar occupancy was

554

60% higher in both Ex-Ctrl and Ex-MIA groups compared to both NEx-Ctrl and NEx-MIA.

555

EP

TE D

M AN U

SC

RI PT

539

Fig 5.

557

Histologic images of the tibial cartilage at sacrifice. Longitudinal histologic sections stained

558

with Safranin O and fast green of tibial plateaus. Non exercised rats receiving right knee NaCl

559

injection (A); non exercised rats receiving right knee monosodium iodoacetate injection (B);

560

exercised receiving right knee NaCl injection (C) and exercised rats receiving right knee

561

monosodium iodoacetate injection (D). NaCl injected rats (Ctrl) presented normal cartilage (A

562

& C). MIA injected rats (B & D) had loss of proteoglycans, fibrillation or vertical fissure in

AC C

556

21

ACCEPTED MANUSCRIPT 563

the mid zone and reduction of cartilage thickness. Magnification of histology images

564

presented was x40.

565

AC C

EP

TE D

M AN U

SC

RI PT

566

22

ACCEPTED MANUSCRIPT

Table 1 Bone mineral density of rats at the whole body, the right lower limb, the right lower limb distal femur and the right lower limb proximal tibia; at baseline, after training and at sacrifice

END W15

569 570 571 572 573 574 575 576 577

n= 12

n= 12

n= 12

WB BMD (g/cm²)

0.147 (0.145, 0.149)

0.147 (0.145, 0.149)

0.144 (0.141, 0.147)

0.145 (0.143, 0.146)

0.3107

RL BMD (g/cm²)

0.146 (0.141, 0.151)

0.146 (0.137, 0.154)

0.142 (0.135, 0.149)

0.150 (0.145, 0.153)

0.4061

RL FD BMD (g/cm²)

0.315 (0.308, 0.322)

0.310 (0.309, 0.318)

0.308 (0.304, 0.311)

0.302 (0.294, 0.310)

0.7411

RL TP BMD (g/cm²)

0.298 (0.283, 0.314)

0.292 (0.285, 0.299)

0.295 (0.281, 0.301)

0.290 (0.279, 0.301)

0.8588

WB BMD (g/cm²)

0.194 (0.189, 0.199)*

0.191 (0.187, 0.194)*

0.191 (0.188, 0.194)*

0.194 (0.191, 0.196)*

0.4576

RL BMD (g/cm²)

0.211 (0.205, 0.217)

0.202 (0.196, 0.209)

0.210 (0.205, 0.215)

0.206 (0.196, 0.215)

0.2876

RL DF BMD (g/cm²)

0.442 (0.422, 0.460)

0.451 (0.433, 0.470)

0.451 (0.435, 0.467)

0.436 (0.416, 0.455)

0.5867

RL PT BMD (g/cm²)

0.426 (0.406, 0.447)

0.429 (0.416, 0.443)

0.441 (0.422, 0.459)

0.429 (0.412, 0.445)

0.2524

WB BMD (g/cm²)

0.191 (0.188, 0.194)*,**

0.193 (0.190, 0.196)*

RL BMD (g/cm²)

0.202 (0.196, 0.209)

0.205 (0.199, 0.213)

0.209 (0.203, 0.216)

0.211 (0.204, 0.218)

0.4908

RL DF BMD (g/cm²)

0.451 (0.431, 0.469)

0.436 (0.421, 0.450)

0.446 (0.432, 0.451)

0.440 (0.420, 0.459)

0.6797

RL PT BMD (g/cm²)

0.429 (0.414, 0.443)

0.409 (0.399, 0.419)a, c, d

0.439 (0.423, 0.455)

0.435 (0.423, 0.437)

0.0309

RI PT

Ex-MIA

M AN U

SC

n= 12

Kruskal Wallis

Ex-Ctrl

TE D

AFTER TRAINING W10

NEx-MIA

EP

BASELINE W0

NEx-Ctrl

0.197 (0.192, 0.201)*,** 0.198 (0.195, 0.200)*,**

p-value

0.1349

WB: whole body; RL: right lower limb; DF: distal femur; PT: proximal tibia; BMD: bone mineral density; NEx-Ctrl: non exercised rats receiving right knee NaCl injection; NEx-MIA: non exercised rats receiving right knee monosodium iodoacetate injection; Ex-Ctrl: exercised receiving right knee NaCl injection; Ex-MIA: exercised rats receiving right knee monosodium iodoacetate injection. Data are presented as mean (95% confidence intervals). Significance level fixed at p<0.05.Inter-group differences were observed using the non-parametric Kruskal-Wallis test followed by Mann-Withney U tests. asignificantly lower than NEx-Ctrl; csignificantly lower than Ex-Ctrl; dsignificantly lower than Ex-MIA. Time effect on whole body was observed using non parametric Friedman test followed by Wilcoxon test: *significantly different versus baseline; **significantly different versus after training time point.

AC C

567 568

23

ACCEPTED MANUSCRIPT

Table 2 Medial and lateral tibial trabecular subchondral bone microarchitecture at sacrifice NEx-Ctrl

NEx-MIA

946.9 (902.9, 990.9)

915.5 (855.4, 975,7)

Medial

946.9 (865.6, 1028.2)

809.3 (722.8, 895.8)*

37.35 (33.12, 41.67)

34.19 (30.83, 37.51)c

Medial

42.79 (39.33, 46.37)*

Ex-Ctrl

Ex-MIA

862.2 (883.9, 947.1)

888.1 (855.6, 975.4)

831.6 (751.5, 911.7)

833.4 (765.1, 901.7)

BV/TV (%)

Lateral

42.21 (40.30, 44.18)

32.81 (29.21, 36.45)c

40.27 (36.48, 44.15)*

45.51 (42.32, 48.74)

41.51 (37.68, 45.42)*

Tb.Th (mm)

Lateral

0.111 (0.106, 0.116)

0.110 (0.107, 0.113)c

0.117 (0.113, 0.121)

0.107 (0.101, 0.113)c

Medial

0.124 (0.119, 0.129)*

0.125 (0.119, 0.131)*

0.128 (0.122, 0.134)*

0.127 (0.119, 0.135)*

Tb.N (1/mm)

Lateral

3.34 (3.06, 3.61)

3.09 (2.84; 3.33)c

3.61 (3.46, 3.75)

3.04 (2.82, 3.25)c

Medial

3.45 (3.21, 3.68)

3.23 (2.99, 3.46)

3.55 (3.40, 3.69)

3.25 (3.10, 3.39)

Lateral

0.216 (0.206, 0.226)

0.229 (0.217, 0.241)

0.203 (0.193, 0.213)

0.219 (0.202, 0.236)

Medial

0.196 (0.185, 0.207)*

0.209 (0.189, 0.229)

0.195 (0.180, 0.210)

0.209 (0.190, 0.228)

AC C

EP

TE D

Tb. SCB Th. (µm)

Lateral

Tb.Sp (mm)

M AN U

SC

RI PT

578

24

ACCEPTED MANUSCRIPT

EP

TE D

M AN U

SC

RI PT

Lateral plateau 3D volumes of interest are presented in the table. NEx-Ctrl: non exercised rats receiving right knee NaCl injection; NEx-MIA: non exercised rats receiving right knee monosodium iodoacetate injection; Ex-Ctrl: exercised receiving right knee NaCl injection; Ex-MIA: exercised rats receiving right knee monosodium iodoacetate injection. Tb. SCB Th.: Trabecular subchondral bone thickness. Data are presented as mean (95% confidence intervals). Significance level fixed at p<0.05. csignificantly lower than Ex-Ctrl; *significantly different from lateral plateau (intra-group comparison).

AC C

579 580 581 582 583 584 585

25

ACCEPTED MANUSCRIPT

586

Table 3 Osteocyte lacunae density, surface and occupancy in the cortical subchondral bone of proximal tibia at sacrifice NEx-MIA

Ex-Ctrl

RI PT

NEx-Ctrl

Ex-MIA

621.37 (532.59, 710.15)

539.82 (488.66, 590.98)

566.70 (504.02, 629.38)

671.44 (602.98, 739.89)

Full lacunae surface (µm²):

44.65 (34.28, 55.01)c, d

44.75 (34.04, 55.45)c, d

48.63 (37.11, 60.15)

47.09 (35.56, 58.61)

Empty lacunae surface (µm²):

33.53 (24.26, 42.79)d, *

32.48 (22.16, 42.79)a, c, d, *

34.08 (23.95, 44.20)*

35.57 (24.02, 47.11)*

OCY occupancy (%)

38.48 (26.63, 50.32)c, d

37.63 (30.00, 45.25)c, d

61.10 (56.41, 65.78)

63.03 (57.19, 68.86)

M AN U

SC

Density (lacunae/mm²)

587

Lacunae density, lacunae surface and percentage of osteocyte lacunar occupancy (OCY occupancy) were observed with bright field microscopy

589

on cortical subchondral bone of the tibia stained by Toluidine blue. Lacunae surfaces were measured in full osteocyte lacunae (cell present in

590

lacunae) and in empty osteocyte lacunae. NEx-Ctrl: non exercised rats receiving right knee NaCl injection; NEx-MIA: non exercised rats

591

receiving right knee monosodium iodoacetate injection; Ex-Ctrl: exercised receiving right knee NaCl injection; Ex-MIA: exercised rats receiving

592

right knee monosodium iodoacetate injection. Data are presented as mean (95% confidence intervals). Significance level fixed at p<0.05.

593

a

594

(intra-group comparison).

EP

TE D

588

AC C

significantly lower than NEx-Ctrl; csignificantly lower than Ex-Ctrl; dsignificantly lower than Ex-MIA, *significantly lower than full lacunae

595

26

ACCEPTED MANUSCRIPT

SUPPLEMENTARY MATERIAL

597

Supplementary Fig 1.

598

Epifluorescence microscopy of the cortical subchondral bone of proximal tibia. Bright field

599

was employed for morphologic osteocytes identification in each region of interest. We

600

imaged 5 regions of interest around the subchondral cortical bone per sample. The size of one

601

ROI was about 0.107 mm2 so that the total area analyzed per sample was 0.535 mm2. Finally,

602

we acquired images of cleaved caspase-3 for osteocyte apoptosis evaluation. The images were

603

analyzed with Image J software. Non exercised rats receiving right knee NaCl injection (A);

604

non exercised rats receiving right knee monosodium iodoacetate injection (B); exercised

605

receiving right knee NaCl injection (C) and exercised rats receiving right knee monosodium

606

iodoacetate injection (D). The total magnification (ocular and objective) was x100.

M AN U

SC

RI PT

596

AC C

EP

TE D

607

27

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Supplementary table 1. Rats body composition at the whole body and at the right lower limb at baseline, after training and at sacrifice NEx-Ctrl

RL lean mass (g)

AFTER TRAINING W10

WB weight (g) WB fat mass (g) WB lean mass (g) RL weight (g) RL fat mass (g) RL lean mass (g) WB weight (g)

END W15

WB fat mass (g) WB lean mass (g) RL weight (g) RL fat mass (g) RL lean mass (g)

n= 12

p-value

386.2 (376.2, 396.2)

0.7712

40.5 (36.9, 44.1)

39.5 (35.6, 43.4)

36.6 (32.8, 40.4)

40.0 (35.8, 44.2)

0.5566

335.4 (328.1, 342.7)

340.3 (331.8, 348.8)

336.7 (329.2, 344.2)

336.3 (327.8, 344.8)

0.8202

26.2 (25.2, 27.2)

26.1 (24.5, 27.7)

24.9 (24.2, 25.6)a

23.9 (22.6, 25.2)a

0.0218

2.1 (1.8, 2.4)

2.0 (1.6, 2.4)

1.8 (1.6, 2.0)

1.6 (1.3, 1.9)

0.1496

23.4 (22.6, 24.2)

23.5 (22.1, 24.9)

22.4 (21.8, 23.0)a

21.6 (19.6, 23.6)a, b

0.0448

615.2 (589.7, 640.7)*

611.8 (580.5, 643.1)*

556.6 (538.0, 575.2)a, b,*

559.5 (530.4, 588.6)a, b,*

0.0075

RI PT

n= 12

383.1 (374.3, 391.9)

SC

RL fat mass (g)

n= 12

M AN U

RL weight (g)

Kruskal-Wallis

389.8 (380.6, 399.6)

a, b

111.4 (93.3, 129.5)*

116.1 (99.0, 133.2)*

66.1 (55.2, 77.0)

485.7 (467.7, 503.7)*

477.7 (457.7, 497.7)*

473.7 (455.7, 491.7)*

43.5 (41.0, 46.0)

42.1 (39.6, 44.6)

5.2 (4.2, 6.2)

TE D

WB lean mass (g)

Ex-MIA

n= 12

4.5 (3.6, 5.4)

36.9 (35.0, 38.8)

36.2 (34.3, 38.1)

611.8 (580.5, 643.1)* ,

116.1 (99.0, 133.2)* **

,

629.8 (594.0, 665.6)* **

EP

BASELINE W0

WB fat mass (g)

Ex-Ctrl

386.2 (378.1, 394.3)

115.2 (98.5, 131.9)* ,

*

a, b

37.2 (35.2, 39.2)

a, b

71.9 (65.1, 78.7)

*

<0.0001

470.2 (445.7, 494.7)*

0.9030

a, b

38.0 (36.0, 40.0)

a, b

2.5 (2.0, 3.0)

3.1 (2.6, 3.6)

33.6 (31.7, 35.5) ,

618.8 (602.1, 635.5)* ** a, b, ,

* ** ,

0.0047 0.0002

a

33.5 (31.7, 35.3) 80.5 (69.7, 91.3)

a, b

0.0354 ,

0.8630

a, b, ,

* **

0.0037

,

627.6 (601.0, 654.2)* ** 89.8 (80.9, 98.7)

477.7 (457.7, 497.7)*

496.1 (473.0, 519.2)* **

519.9 (501.9, 537.9)* **

518.9 (497.5, 540.3)* **

0.9042

42.1 (39.6, 44.6)

42.6 (40.2, 45.0)

39.8 (38.1, 41.5)

40.4 (38.1, 42.7)

0.0956

4.5 (3.6, 5.4)

AC C

WB weight (g)

NEx-MIA

36.2 (34.3, 38.1)

4.5 (3.8, 5.2)

36.8 (34.7, 38.9)

a, b

a, b

2.8 (2.3, 3.3)

3.3 (2.8, 3.8)

0.0008

35.7 (34.1, 37.3)

35.8 (33.6, 38.0)

0.6960

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

WB: whole body; RL: right lower limb. NEx-Ctrl: non exercised rats receiving right knee NaCl injection; NEx-MIA: non exercised rats receiving right knee monosodium iodoacetate injection; Ex-Ctrl: exercised receiving right knee NaCl injection; Ex-MIA: exercised rats receiving right knee monosodium iodoacetate injection. Data are presented as mean and 95% confidence intervals. Significance level fixed at p<0.05. Inter-group differences were observed using the non-parametric Kruskal-Wallis test followed by Mann-Withney U tests. asignificantly lower than NEx-Ctrl; bsignificantly lower than NEx-MIA. Time effect on whole body was observed using non parametric Friedman test followed by Wilcoxon test: *significantly different versus baseline; **significantly different versus after training time point.

ACCEPTED MANUSCRIPT

Supplementary Table 2. OARSI and Mankin scores of right tbial articular cartilage of rats NEx-MIA

Ex-Ctrl

Ex-MIA

ANOVA pvalue

OARSI scoring

3.1 (1.1, 4.1)

5.7 (4.6, 6.8)a,c

3.1 (1.6, 4.6)

6.9 (5.0, 8.8)a,c

<0.0001

Mankin score

4.8 (3.5, 6.1)

7.6 (6.2, 9.0)a,c

4.9 (3.7, 6.1)

SC

RI PT

NEx-Ctrl

7.8 (5.9, 9.7)a,c

0.0027

AC C

EP

TE D

M AN U

NEx-Ctrl: non exercised rats receiving right knee NaCl injection; NEx-MIA: non exercised rats receiving right knee monosodium iodoacetate injection; Ex-Ctrl: exercised receiving right knee NaCl injection; Ex-MIA: exercised rats receiving right knee monosodium iodoacetate injection. Data are presented as mean and 95% confidence interval. Differences between groups were considered significant at p<0.05. asignificantly higher than NEx-Ctrl; csignificantly higher than Ex-Ctrl.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT