SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

Objective

Physiotherapies are the most widely recommended conservative treatment for arthritic diseases. The present study was undertaken to examine the molecular mechanisms underlying the effects of gentle treadmill walking (GTW) on various stages of monoiodoacetate-induced arthritis (MIA) to elucidate the basis for the success or failure of such therapies in joint damage.

Methods

Knees were obtained from untreated control rats, rats with MIA that did not undergo GTW, rats with MIA in which GTW regimens were started 1 day post–MIA induction, and rats with MIA in which GTW regimens were started after cartilage damage had progressed to grade 1 or grade 2. The cartilage was examined macroscopically, microscopically, and by microfocal computed tomography imaging. Transcriptome-wide gene expression analysis was performed, and microarray data were assessed by Ingenuity Pathways Analysis to identify molecular functional networks regulated by GTW.

Results

GTW intervention started on day 1 post–MIA induction significantly prevented the progression of MIA, but its efficacy was reduced when implemented on knees exhibiting close to grade 1 cartilage damage. GTW accelerated cartilage damage in knees with close to grade 2 damage. Transcriptome-wide gene expression analysis revealed that GTW intervention started 1 day post–MIA inception significantly suppressed inflammation-associated genes and up-regulated matrix-associated gene networks. However, delayed GTW intervention after grade 1 damage had occurred was less effective in suppressing proinflammatory genes or up-regulating matrix synthesis.

Conclusion

The present findings suggest that GTW suppresses proinflammatory gene networks and up-regulates matrix synthesis to prevent progression of cartilage damage in MIA-affected knees. However, the extent of cartilage damage at the initiation of GTW may be an important determinant of the success or failure of such therapies.

Exercise is the most widely recommended and used conservative therapeutic approach to improve joint function in arthritis (1–3), and recent guidelines published by the Osteoarthritis Research Society International suggest that exercise is in general beneficial for patients with osteoarthritis (OA) (3). Such recommendations are supported by the results of large cohort studies demonstrating that adults engaging in minimal to no physical activity have a higher incidence of radiographically diagnosed OA (4). Additionally, older adults engaged in moderate physical activity have a reduced risk of arthritis-related disability (5). However, in some patients physical activity is either associated with greater risk or has no effect on knee joints, making it difficult to discern which patients will and which patients will not benefit from physical therapies (6–9).

Chondrocytes, mechanoresponsive cells within the cartilage, perceive and respond to mechanical stimuli by altering their biosynthetic ability, morphology, and cartilage extracellular matrix (10, 11). These cells interpret mechanical signals in a magnitude-dependent manner (12, 13). Excessive loading of joints is injurious and activates proinflammatory signaling cascades, similar to those implicated in the etiology of OA (14–17). In contrast, physiologic loading has been shown to be antiinflammatory and to induce interleukin-10 (IL-10) production in the synovium and up-regulate synthesis of glycosaminoglycans in the cartilage of patients at increased risk of OA (18–21). In fact, exercise is vital for cartilage homeostasis and its lack leads to atrophy of cartilage (22), implicating distinct roles of physiotherapies in both preventing damage and improving joint function.

The aim of this study was to determine the molecular mechanisms underlying the effectiveness of exercise in the form of gentle treadmill walking (GTW) on well-defined stages of cartilage damage, using a rat knee model of monoiodoacetate-induced arthritis (MIA) (23). We and others have shown previously that low or physiologic levels of compressive/tensile forces have an antiinflammatory effect on chondrocytes in vitro. These forces suppress several biomarkers of inflammation, such as IL-1β, tumor necrosis factor α (TNFα), matrix metalloproteinases (MMPs), and aggrecanases (12, 13, 24, 25). In the present study we systematically examined the efficacy of moderate exercise on the progression of MIA as assessed macroscopically, microscopically, and by microfocal computed tomography (micro-CT) imaging. Transcriptome-wide microarray analysis was also conducted to track the changes in gene expression subsequent to GTW therapy, in order to gain key insights into the basis of success or failure of such therapies (1, 3, 4, 26).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

Arthritis induction and GTW regimens.

MIA was induced in the right knees of 12–14-week-old female Sprague-Dawley rats (Harlan) via single intraarticular injection of monoiodoacetate (2 mg in 50 μl saline/knee) (23). A well-established model of MIA in rat knees was used, which yielded pathologies similar to those described by Guzman et al (23). Typically, monoiodoacetate-administered knees exhibited close to grade 1 cartilage damage on the condylar surface on day 5, close to grade 2 damage on condyles by day 9, and close to grade 3–3.5 cartilage and bone damage on day 21, according to the grading system described by Pritzker et al (27). Sham controls were injected with 50 μl saline in the right knees. The MIA-affected rats showed no signs of limping, pain, or resistance during GTW. One day post–MIA induction, prior to the start of GTW regimens, the pH of the synovial fluid was between 7.5 and 8.5, suggesting likely dissemination of monoiodoacetate from joints.

GTW regimens were started on day 1 after MIA induction (when there was no apparent cartilage damage and <2% cell death observed on the surface of cartilage [as determined using a Live/Dead cell kit; Invitrogen]), on day 5 (close to grade 1 cartilage damage), or on day 9 (close to grade 2 cartilage damage) post–intraarticular monoiodoacetate injection (Figures 1A and B). Rats were randomly assigned to 1 of 5 groups (n = 15 rats per group) (Figure 1A). The control group consisted of saline-injected nonexercised sham controls. The MIA21 group consisted of rats in which MIA was induced on day 0 and no exercise was performed between day 0 and day 21. The MIA + GTW1–21 group consisted of rats in which MIA was induced on day 0 and GTW was performed daily from day 1 through day 21. The MIA + GTW5–21 group consisted of rats in which MIA was induced on day 0 and GTW was performed daily from day 5 through day 21. The MIA + GTW9–21 group consisted of rats in which MIA was induced on day 0 and GTW was performed daily from day 9 through day 21 (Figure 1A). GTW was performed on a small animal treadmill (Columbus Instruments) at a speed of 12 meters/minute for 45 minutes/day (∼0.5 km). This regimen was based on earlier studies (28), but was gentler to avoid pain and resistance to walking due to MIA.

thumbnail image

Figure 1. Effects of gentle treadmill walking (GTW) on the progression of monoiodoacetate-induced arthritis (MIA). A, Graphic depiction of the 5 experimental protocols. All rats were injected in the right knee on day 0 and killed 21 days later. Rats in the sham control group (Cont) were injected with 50 μl saline. Rats in the MIA21 group were injected with 2 mg monoiodoacetate in 50 μl saline. Rats in the MIA + GTW1–21 group, the MIA + GTW5–21 group, and the MIA + GTW9–21 group were injected with 2 mg monoiodoacetate in 50 μl saline and performed GTW on days 1–21, days 5–21, or days 9–21, respectively. B, Cartilage section showing dead cells (red) (arrow) and live cells (green) on day 1 post–MIA inception. Original magnification × 50. C, Macroscopic appearance, microscopic appearance (with hematoxylin and eosin staining), and bone imaging (by microfocal computed tomography [μCT] [360° μCT projections are shown in Supplementary Figures 1–5, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131]) of cartilage and bone from a–d, a rat in the healthy sham control group, showing smooth cartilage surface, normal histologic features, and no bone lesions on the femoral condyles and patellar grove; e–h, a rat in the MIA + GTW1–21 group, showing no surface abrasions on the condyles, some cartilage lesions on the patellar groove and ridges, near-normal histologic features, and no bone involvement; i–l, a rat in the MIA + GTW5–21 group, showing some abrasions on condyles, cartilage damage on the patellar groove and ridges, focal matrix condensation, cell clustering and disorganization, and fibrocartilage formation seen histologically, and some bone lesions; m–p, a rat in the MIA + GTW9–21 group, showing extensive cartilage lesions on condyles and the patellar groove, severe cartilage loss and denuded bone seen histologically, and excessive bone lesions on femoral condyles and the patellar groove; and q–t, and a rat in the MIA21 group, showing cartilage matrix loss, delamination of superior surface, and excavation and matrix loss in the superficial and midzones. Arrows indicate cartilage damage. Within each treatment group, all samples exhibited similar characteristics; representative images are shown. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Download figure to PowerPoint

On day 21, all rats were killed 2 hours after the last exercise session and their knees harvested. Cartilage damage in rats subjected to GTW was compared to cartilage damage in nonexercised rats on day 21 post–MIA inception. In each group, femurs from 5 rats were snap-frozen in liquid nitrogen for molecular analysis, and femurs from 10 rats were fixed in 10% buffered formalin for macroscopic, microscopic, or micro-CT imaging analyses. Cartilage damage was graded as described by Pritzker et al (27).

All protocols were approved by the Institutional Animal Care and Use Committee at The Ohio State University.

RNA extraction and microarray analysis.

Cartilage from the distal end of individual femurs was examined under a stereomicroscope (Zeiss). Using a scalpel blade, cartilage was carefully sliced off into small chips while maintained in a frozen state, avoiding the areas immediately around lesions to exclude tissue ingrowth in the lesions. The cartilage chips from individual femurs (∼10 mg/femur) were separately collected, and pulverized in a Mikrodismembrator-S (Sartorious) for 30 seconds at 2,500 revolutions per minute. RNA was extracted with TRIzol reagent (Invitrogen) (29) and analyzed with a 2100 Bioanalyzer (Agilent) to ensure integrity.

RNA (300 ng) from 3 independent samples per group was used for complementary DNA (cDNA) synthesis and labeling, using a GeneChip Whole Transcript cDNA Synthesis and Amplification Kit and a GeneChip WT Terminal Labeling Kit, respectively (both from Affymetrix). The labeled cDNA samples were hybridized on a GeneChip Rat Gene 1.0 ST Array (Affymetrix) and scanned at the Microarray Shared Resource Facility at The Ohio State University.

The intensity scans from 3 independent GeneChips per treatment group were subjected to gene expression analysis using Partek Genomic Suite, version 6.4. The significance of differences among the groups was calculated by analysis of variance (ANOVA), and only significantly differentially regulated transcripts (P < 0.05) were considered for further analysis. Variations among the samples in each group were examined by principal components analysis, and subjected to hierarchical and partition clustering with the Partek Genomic Suite.

Functional gene network analysis.

The gene expression data derived from microarray analysis were subjected to Ingenuity Pathways Analysis (IPA; Ingenuity Systems) to generate functional molecular networks. A fold change cutoff of 2.0 was set to identify and assign the molecules to the Ingenuity Knowledge Base. The gene expression changes were considered in the context of physical, transcriptional, or enzymatic interactions of the gene/gene products and then grouped according to interacting gene networks.

Validation of salient genes differentially expressed in the cluster analysis.

Expression of selected genes from cluster analysis was confirmed by real-time polymerase chain reaction (PCR) (13). Briefly, first-strand cDNA was synthesized from RNA using a Superscript III Reverse Transcriptase Kit (Invitrogen). Gene expression was assessed by amplifying the cDNA with custom-designed primers using the iCycler iQ Real-Time PCR System (Bio-Rad). The primers used were as follows: Rps18 sense 5′-GCGGCGGAAAATAGCCTTCG-3′, antisense 5′-GGCCAGTGGTCTTGGTGTGCTG-3′; Fcgr1a sense 5′-AGCGGCATCTATCACTGCTCA-3′, antisense 5′-TCAGCACTGGTGTGGCAAATA-3′; Aspn sense 5′-CAAAGAGCCAGTGAACCCCTT-3′, antisense 5′-TCAGAACAGTGGACGACTCGA-3′; Mmp12 sense 5′-CCAGGAAATGCAGCAGTTCTTT-3′, antisense 5′-GCTGTACATCAGGCACTCCACAT-3′; Alox5 sense 5′-TTCTCCGCACACATCTGGTGT-3′, antisense 5′-GGCAATGGTGAACCTCACATG-3′; Vcam1 sense 5′-GCCGGTCATGGTCAAGTGTTT-3′, antisense 5′-CATGAGACGGTCACCCTTGAA-3′; Cilp sense 5′-TGTGAAGTCCAAGGTCACCCA-3′, antisense 5′-GTAGAAGGAGTTGGTGGCATTCTG-3′; Sox9 sense 5′-ATCTGAAGAAGGAGAGCGAG-3′, antisense 5′-CAAGCTCTGGAGACTGCTGA-3′; Col9a1 sense 5′-TGATGGCTTTGCTGTGCTG-3′, antisense 5′-TGACTGGCAGTTCATGGCA-3′; Frzb sense 5′-TGCCCTCCCCTCAGTGTTAAT-3′, antisense 5′-CAAGCCGATCCTTCCACTTCT-3′; Col2a1 sense 5′-ATGAGGGCCGAGGGCAACAG-3′, antisense 5′-GATGTCCATGGGTGCAATGTCAA-3′.

Statistical analysis.

The significance of differences in the microarray data among the experimental groups was tested by ANOVA using Partek Genomic suite (n = 3 independent samples per group). ANOVA with Tukey's honest significant difference post hoc test, applied using SPSS version 17, was used to determine the significance levels of real-time PCR data, which were derived using 2 additional independent samples per group (n = 5 per group). P values less than 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

GTW prevents progression of cartilage destruction when implemented during the early stages of MIA.

Comparison of the anatomy/morphology of femoral cartilage from rats in the MIA + GTW1–21 group versus rats in the nonexercised MIA21 group revealed that exercise significantly prevented progression of MIA. Cartilage from rats in the MIA + GTW1–21 group exhibited a smooth surface with minimal aberrations or lesions. Histologic examination of cartilage from this group revealed slight thinning of cartilage in small areas, whereas cartilage and subchondral bone was preserved, with no signs of bone changes (Figure 1C, parts e–h). Micro-CT imaging confirmed that early intervention with exercise prevented the bone erosion seen in cartilage from nonexercised rats in the MIA21 group (Figure 1C, parts h and t).

We next investigated whether exercise could prevent or delay progression of MIA in rats with close to grade 1 cartilage damage. Analysis of femurs from rats in the MIA + GTW5–21 group revealed that GTW appeared to delay the progression of MIA, as evidenced by the relatively smooth condylar surface (Figure 1C, parts i–l). Histologic analysis showed close to grade 1.5–2 damage, whereas cartilage from rats in the MIA21 group had grade 3–3.5 damage. In parallel, micro-CT imaging showed reduced severity of bone erosions in the femurs of rats in the MIA + GTW5–21 group (Figure 1C, part l). In contrast, initiation of GTW 9 days after inception of MIA (MIA + GTW9–21 group) resulted in close to grade 4 or greater damage, with denuded cartilage and sclerotic subchondral bone that covered the femoral condyles, patellar groove, and ridges. Imaging by micro-CT also confirmed greater bone loss on both the femoral condyles and the patellar groves (Figure 1C, part p), compared to bone damage in the MIA21 rats (Figure 1C, part t).

Extent of cartilage damage at the inception of GTW critically influences the expression of catabolic and anabolic genes.

RNA from the femoral cartilage of rats in the control, MIA21, MIA + GTW1–21, and MIA + GTW5–21 groups was subjected to microarray analysis, and gene expression was compared. Due to limited cartilage remaining on the condyles, samples from rats in the MIA + GTW9–21 group were excluded from this analysis. Principal components analysis revealed significantly distinct distributions of gene expression between treatment groups (n = 3 samples per group) (Figure 2A), as evidenced by the average signal-to-noise ratio of 15.5 in a total of 27,342 transcripts on the Affymetrix GeneChip array. Hierarchical clustering of the differentially regulated genes (>2-fold change; P < 0.05) (Figure 2B) demonstrated that 1) cartilage from control rats showed minimal active genes (red) and a maximal number of quiescent genes (blue); 2) cartilage from rats in the MIA21 group regulated a maximal number of transcripts (1,179; 4.31%); 3) cartilage from rats in the MIA + GTW1–21 group regulated 847 transcripts (3.10%), with a gene expression pattern closer to that seen in the control group; and 4) cartilage from rats in the MIA + GTW5–21 group regulated 1,103 transcripts (4.03%), with a gene expression pattern closer to that seen in the MIA21 group.

thumbnail image

Figure 2. Transcriptome-wide microarray analysis of femoral cartilage from rats in each treatment group. A, Principal components analysis (PCA), showing reproducible gene expression in articular cartilage from the right knee joint of 3 separate rats per group from the control group, the MIA + GTW1–21 group, the MIA + GTW5–21 group, and the MIA21 group. B, Overall gene expression profiles in articular cartilage from the right knee joint of 3 separate rats per group. The intensity plot with dendrogram represents the transcripts that were significantly (P < 0.05) and differentially up-regulated (red) or down-regulated (blue) by >2-fold. The analysis shows that the most differential changes in gene expression (as compared to expression in cartilage from rats in the MIA21 group) occurred in the MIA + GTW1–21 group, followed by the MIA + GTW5–21 group. C, Temporal regulation of MIA-associated gene clusters associated with acute and innate immune responses (cluster I), chronic inflammatory responses (cluster II), musculoskeletal disorders and inflammatory diseases (cluster III), and genetic disorders and skeletal and muscular disorders (clusters IV and V). D, Percentage of genes in clusters I, II, III, IV, and V that were significantly (P < 0.05) up-regulated or down-regulated by exercise (MIA + GTW1–21 group and MIA + GTW5–21 group in comparison to MIA21 group). See Figure 1 for other definitions. Color figure can be viewed in the online issue, which is available at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131.

Download figure to PowerPoint

MIA-affected knees exhibited a temporal pattern of gene regulation during the progression of cartilage damage (Figure 2C). The differentially expressed genes identified could be categorized into 5 clusters: cluster I, immune response and innate immunity genes, showing peak up-regulation in cartilage with grade 1 damage (day 5 after MIA inception); cluster II, chronic immune response and immune trafficking genes, showing peak up-regulation in cartilage with close to grade 2 damage (day 9); cluster III, chronic inflammatory disease and immunosuppression/adaptation genes, showing a gradual increase in expression until cartilage showed close to grade 3–3.5 damage (day 21); cluster IV, musculoskeletal development and function–associated genes, showing peak down-regulation in cartilage with grade 1 damage (day 5); and cluster V, genetic disorder and skeletal and musculoskeletal disease–associated genes, showing peak down-regulation in cartilage with grade 2 damage (day 9).

We next examined the effects of GTW on genes in each cluster in rats with MIA (Figure 2D). Intervention by GTW in the MIA + GTW1–21 group suppressed ∼52%, ∼0%, and ∼59% of the genes in inflammatory clusters I, II, and III, respectively, that were up-regulated in rats in the MIA21 group. In parallel, GTW up-regulated ∼33% and ∼31% of the genes in clusters IV and V, respectively, that were suppressed in rats in the MIA21 group. However, when exercise was initiated after grade 1 cartilage damage had occurred (MIA + GTW5–21 group), the suppression of genes in clusters I, II, and III was only 6%, 14%, and 32%, respectively. Similarly, <3% of the genes in cluster IV and 11% of the genes in cluster V were up-regulated in the MIA + GTW5–21 group (Figure 2D).

Table 1 lists salient genes that were differentially regulated in rats that underwent exercise regimens (MIA + GTW1–21 and MIA + GTW5–21) as compared to rats in the MIA21 group. Among rats in the MIA + GTW1–21 group, the major cluster I genes that were suppressed by GTW were Alox5ap (arachidonate 5-lipooxygenase–activating protein, required for Alox activation), Fcgr1a (Cd64) (high-affinity immunoglobulin γ Fc receptor I, regulates innate/specific immune responses), Hla-dmb (HLA class II antigen β-chain), Cd53 (surface molecule, regulates innate levels of TNFα), Aspn (asporin, negatively regulates transforming growth factor β [TGFβ]), Calcr (calcitonin receptor, involved in bone formation), Ctsg (cathepsin G, a peptidase), and Il1rl1 (IL-1 receptor–like 1). The cluster II genes that were down-regulated by GTW were Cd84 (adherence-associated molecule), Il18 (interleukin-18), Mmp12 (elastase), Mmp19 (involved in tissue remodeling), Adamts4 (aggrecanase), Adamts7 (degrades cartilage oligomeric protein), Ccr1 (chemokine receptor 1, chemoattracts cells), and Ccl9 (osteoclast activation through Ccr1). The suppressed genes in cluster III included Alox5, Clec4d (C-type lectin domain family 4, involved in antigen uptake), Vcam1 (vascular cell adhesion molecule 1), Adam23 (disintegrin and metalloproteinase domain–containing protein 23, involved in cell adhesion), Postn (periostin, involved in bone formation), and Crlf1 (cytokine receptor–like factor 1), all of which are involved in chronic inflammation.

Table 1. GTW-induced differential expression of salient genes in MIA-affected cartilage*
GeneMIA + GTW1–21 groupMIA + GTW5–21 group
  • *

    Values are the degree of up-regulation or down-regulation of each gene, expressed as the percentage in relation to expression of the gene on day 21 in rats with monoiodoacetate-induced arthritis (MIA) that were not subjected to gentle treadmill walking (GTW). Rats in the MIA + GTW1–21 group underwent GTW on days 1–21 after induction of MIA (on day 0); rats in the MIA + GTW5–21 group underwent GTW on days 5–21 after induction of MIA. See Results for description of the gene clusters.

Cluster I
 Alox5ap−206−119
 Calcr−183−136
 Fcgr1a−182−101
 Tlr7−179+115
 C3−178−104
 Hla-dmb−176+130
 Cd53−169−122
 Aspn−166+148
 Ctsg−126+106
 Il1rl1−124+156
Cluster II  
 Cd84−161−133
 Il18−149−102
 Mmp12−144+130
 Cd44−140−104
 Tnfsf13−140−103
 Adamts7−136−112
 Adamts4−118−185
 Ccl9−117+125
 Ccr1−112−117
 Mmp19−110−121
Cluster III  
 Clec4d−209+103
 Alox5−193−126
 Vcam1−171−102
 Adam23−138+103
 Crlf1−131−126
 Cdh13−125−139
 Postn−122+123
 C1s−120−142
 Serpine1−119+101
 Cd14−115−152
Cluster IV
 Cilp+445−104
 Cytl1+226−107
 Cilp2+210−156
 Hapln3+158+103
 Acan+152−110
 Sox9+150−123
 Gdf10+140−176
 Igf2+139+136
 Casr+139−219
 Chst3+114−128
Cluster V  
 Col9a1+1102+341
 Matn3+979+322
 Frzb+761+129
 Mia+381+229
 Col2a1+324+109
 Chad+287−101
 Hapln1+277+108
 Col11a2+264+126
 Vit+251−101
 Prg4+165−140

More importantly, in cartilage from rats in the MIA + GTW1–21 group, GTW up-regulated the expression of extracellular matrix–associated cluster IV and cluster V genes that were suppressed in cartilage from rats in the MIA21 group. Up-regulated genes in cluster IV included Cilp (cartilage intermediate-layer protein), Cilp2, Acan (aggrecan), Sox9 (transcription factor required for chondrocyte matrix proteins), Cytl1 (cytokine-like 1, promotes proteoglycan synthesis), Igf2 (insulin-like growth factor 2, chondrocyte growth and differentiation). Similarly, genes in cluster V that were up-regulated by GTW were collagens (Col2a1, Col9a1, Col11a2), Matn3 (matrilin 3), Frzb (Wnt signaling inhibitor), Mia (melanoma-derived growth regulatory protein), Chad (chondroadherin, mediates chondrocyte adhesion), Hapln1 (hyaluronan and proteoglycan link protein 1), Vit (vitrin, promotes matrix assembly), and Prg4 (lubricin). In cartilage from rats in the MIA + GTW5–21 group, in contrast, genes that were suppressed in cartilage from nonexercised rats with MIA either remained suppressed or were up-regulated to a lesser extent after GTW than was observed in cartilage from rats in the MIA + GTW1–21 group.

The extent of cartilage damage at the initiation of GTW determines the effectiveness of GTW through differential regulation of major intracellular pathways.

Among the 1,179 genes that were up- or down-regulated by >2-fold (P < 0.05) in the MIA21 group, 142 genes were included in the “human and experimental ‘arthritis’” disease category in the Ingenuity Knowledge Base. These 142 genes were clustered to show the general trends regulated by exercise (Figure 3). This analysis demonstrated that intervention by GTW in the MIA + GTW1–21 group suppressed 96% of the genes that were up-regulated in the MIA21 group and up-regulated 81% of the genes that were down-regulated in the MIA21 group (as evidenced by color shifting toward darker shades, i.e., closer to control levels [Figure 3A]). In contrast, when initiated 5 days after the onset of MIA (MIA + GTW5–21 group), GTW suppressed 80% of the genes that were up-regulated in the MIA21 group and up-regulated only 58% of the genes that were down-regulated in the MIA21 group (Figure 3A).

thumbnail image

Figure 3. Catabolic and anabolic networks regulated by GTW in MIA. A, Intensity plot of 142 genes that are a known set of arthritis-associated genes in the Ingenuity Knowledge Database and were up-regulated (red) or down-regulated (green) by >2-fold in the MIA21 group. (A list of the genes and the data on up- and down-regulation by treatment group are provided in Supplementary Table 1, available in the online version of this article at http://onlinelibrary.wiley.com/journal/10.1002/(ISSN)1529-0131) B–E, Ingenuity Pathways Analysis–generated networks depicting catabolic genes that were down-regulated by GTW in the MIA + GTW1–21 group, showing suppression of genes involved in acute and chronic inflammatory/immune responses via the suppression of NF-κB activity (B); anabolic genes that were up-regulated by exercise in the MIA + GTW1–21 group, showing induction of Sox9 and Tgfb that in turn may up-regulate the expression of matrix-associated genes (C); catabolic genes that were regulated by GTW in the MIA + GTW5–21 group, showing partial suppression of genes involved in acute and chronic immune responses (D); and anabolic genes that were regulated by GTW in the MIA + GTW5–21 group, showing suppression of Sox9, Tgfb, and the genes for a number of matrix proteins that may limit the ability of cartilage to undergo repair (E). Red, green, and white represent up-regulation, down-regulation, and no regulation, respectively. The shading of colors represents the degree of regulation, from greater changes (dark) to lesser changes (light). See Figure 1 for definitions.

Download figure to PowerPoint

During the progression of MIA, NF-κB and TGFβ play a key role in regulating expression of genes in the inflammatory clusters I, II, and III and the anabolic clusters IV and V. Consequently, the functional relationships among NF-κB, TGFβ, and 142 arthritis-related genes were analyzed using an IPA custom molecular network generation tool (Network Explorer) (Figures 3B–E). This analysis revealed that the catabolic gene networks induced during the progression of MIA were down-regulated by GTW in the MIA + GTW1–21 group, likely via suppression of NF-κB activity, a major node in this network (Figure 3B). The NF-κB in turn may regulate the genes in the arachidonate-lipoxygenase pathway (Alox5, Alox5ap), genes for adhesion molecules (Itgb1, Itgal, Vcam1, Cd44), cell cycle–associated genes (Ck6, Cdkn1a, Bcl2, Ank1), and genes for cytokines (Il18, Il15) and MMPs (Mmp12, Mmp14). Similarly, GTW in the MIA + GTW1–21 group up-regulated anabolic gene networks via Tgfb and Sox9. These growth factors and transcription factors have been shown to regulate the expression of collagen genes (Col2a1, Col9a1, Col9a2, Col9a3, Col11a2), Cilp, Cilp2, Mgp (matrix Gla protein), Acan, and other matrix proteins (Figure 3C and Table 1).

Studies of the same arthritis-associated networks in rats in the MIA + GTW5–21 group demonstrated that intervention by GTW, even after grade 1 cartilage damage had occurred, suppressed several genes associated with NF-κB activity (Figure 3D). In contrast, several proinflammatory genes, such as Aspn, Mmp12, Ccl9, Irf5 (interferon regulatory factor 5), Itgb2 (β2 integrin), Ctsg, and Postn, were not suppressed or up-regulated by GTW when the exercise was implemented after grade 1 cartilage damage had occurred (Figure 3E). Additionally, in rats in the MIA + GTW5–21 group, Sox9 was further suppressed and, together with Tgfb, likely led to the down-regulation of Acan, Alp (alkaline phosphatase), Cilp, Cilp2, and Mgp required for matrix assembly. Nevertheless, the expression of collagens (ColIXa1, ColIXa2, ColIXa3, ColIIa1, ColXIa2) was up-regulated in rats in the MIA + GTW5–21 group compared to the MIA21 group (Figure 3E).

Results of real-time PCR validated the microarray findings, showing down-regulation of the salient genes in clusters I, II, and III (Fcgr1a, Aspn, Mmp12, Alox5, Vcam1) and up-regulation of the genes in clusters IV and V (Cilp, Sox9, Col9a1, Frzb, Col2a1) by GTW in the MIA + GTW1–21 group (Figure 4A). IPA showed that the regulation of NF-κB may play a focal role in the antiinflammatory effects of GTW. Since NF-κB activity in the cells has been shown to be oscillatory (30, 31) and thus may not provide the true activation state in inflamed knees, we examined the expression of several signaling molecules in the NF-κB pathway. As shown in Figure 4B, GTW in rats in the MIA + GTW1–21 group suppressed the expression of Traf2 (TNF receptor–associated factor 2 [TRAF2]), Traf3, Traf6, Tank (TRAF family member–associated NF-κB activator), Irak4 (Il-1 receptor–associated kinase), Ripk1 (receptor [TNF receptor superfamily]–interacting Ser-Thr kinase), Ripk3, and Ikbkg (IκB kinase γ/IKKγ), which were up-regulated in MIA-affected knees. However, when GTW was initiated in rats that already had grade 1 cartilage damage, suppression of some of these MIA-induced genes, i.e., Traf3, Tank, Ripk1, Ripk3, and Ikbkg, was not observed (Figure 4B).

thumbnail image

Figure 4. A, Genes selected from each cluster to demonstrate their up- or down-regulation, as determined by quantitative real-time polymerase chain reaction analysis, in cartilage from rats in the MIA + GTW1–21 or MIA + GTW5–21 group as compared to cartilage from rats in the MIA21 group. Values are the mean ± SEM from analysis of RNA from 5 separate rats per group. ∗ = P < 0.05; ∗∗ = P < 0.01. B, Regulation of the genes required for activation of NF-κB by GTW. Expression of Traf2, Traf3, Traf6, Tank, Irak4, Ripk1, Ripk3, and Ikbkg was suppressed in rats in the MIA + GTW1–21 group as compared to the MIA21 group. Expression of some of these genes, i.e., Traf2, Traf6, and Irak4, was suppressed (to a lesser extent) in the MIA + GTW5–21 group as well. Up-regulation of gene expression in the MIA21 group was determined based on comparison to the control group without MIA. Values are the mean from 3 independent cartilage specimens. See Figure 1 for definitions.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

The present study documents the effects of GTW as a form of exercise on global gene regulation in the articular cartilage of the knee affected with MIA in various stages. In MIA, acute inflammation and chronic inflammation drive the destruction of the knee, whereas inhibition of matrix synthesis and its breakdown prevent repair, which worsens the joint damage (32–35). We showed that GTW suppresses inflammation and up-regulates repair to prevent active progression of cartilage damage. However, the maximal effects were observed when GTW was implemented on knees with grade 1 or lesser cartilage damage. In contrast, when knees with grade 2 damage were subjected to GTW, its effectiveness was compromised, and the cartilage damage was further intensified as compared to that in the knees of rats with MIA that had not undergone GTW. Benefits of exercise in the form of GTW are likely contingent upon many factors, including adherence to exercise regimens, frequency of exercise, speed of walking, range of motion, and actual loading of the symptomatic compartment. Furthermore, the efficacy of exercise in humans with arthritic lesions in various stages is less predictable, mainly due to limitations in detecting the extent of cartilage damage in humans (36–38). Our observations in the present MIA model indicate that the extent of cartilage damage may also play an important role in achieving the optimal effects of gentle exercise.

The major limitation of the MIA model used in the present study is that monoiodoacetate induces aggressive cartilage destruction, which progresses to grade 3–3.5 damage within 21 days. Thus, these lesions may not depict cartilage damage caused by trauma/insult that develops over extended periods of time in human OA (23). This is an important limitation that must be considered when trying to extend/translate these findings to arthritis in humans. Nevertheless, it is important to note that even during this aggressive progression of cartilage destruction, GTW could suppress progression of inflammation and cartilage loss, as evidenced by macroscopic examination and microscopic and micro-CT imaging. In support of the notion that exercise mediates joint function, it was recently reported that the extent of physical activity and better joint performance were proportionally related, in a cohort of 2,589 OA patients (39).

Both resistance and gentle exercises are currently prescribed for rehabilitation of injured or arthritic knees. However, the optimal duration and physical loading necessary for achieving the beneficial effects of exercise are unclear. In the present study, we subjected knees to 45 minutes of GTW, at a rate of 12 meters/minute. Considering the aggressive nature of the disease in the MIA model used, the selected speed and duration of GTW were gentler than those used in earlier experimental models, e.g., 16 meters/minute for 1 hour (28). The effects of longer or shorter duration or different speeds of GTW are yet to be determined. For example, a different exercise regimen in the MIA + GTW5–21 group may have potentially prevented cartilage damage more effectively. Interestingly, even a single session of exercise (90° knee bending, 250 times) has been shown to increase IL-10 and suppress release of cartilage oligomeric protein and aggrecan levels in intraarticular and perisynovial fluid from osteoarthritic knees (20). These findings further support the notion that exercise may act as an antiinflammatory and reparative signal in inflamed knees.

The information gained from the transcriptome-wide gene expression analysis demonstrated that counteracting the MIA-induced gene induction or suppression appears to be a primary mechanism underlying the GTW-mediated inhibition of cartilage damage. GTW suppressed expression of a significant number of innate and chronic immunity–related genes (clusters I, II, and III), demonstrating the potential of such exercise in suppressing inflammation. Additionally, GTW promoted repair by inducing the expression of genes related to musculoskeletal development and function (clusters IV and V) in rats in the MIA + GTW1–21 group.

It is likely that the partial success of GTW in limiting the existing cartilage damage in the MIA + GTW5–21 group may be related to its inability to inhibit/induce expression of some of the genes in the catabolic and anabolic clusters. Specifically, NF-κB controls induction of proinflammatory genes and has a critical role in cartilage inflammation (24, 40–42). Based on IPA network exploration, exercise intervention in the MIA + GTW1–21 group may have suppressed NF-κB activity and thus genes associated with the NF-κB networks, such as those involved in apoptosis and cell cycle (Cdkn1a and Bcl2), cell adhesion (Itgb1, Itgb2, and Vcam1), complement (C3), matrix breakdown (Mmp12 and Mmp14), and proinflammatory responses (Il15 and Il18) (43–45). Interestingly, GTW suppressed fewer genes in the NF-κB signaling cascade when implemented in the later stages of cartilage damage, as was also reflected in the lesser extent of suppression and number of proinflammatory genes suppressed in the MIA + GTW5–21 group.

During inflammation, activation of proinflammatory cytokine receptors leads to sequential activation of receptor-associated kinases, adaptor proteins, TRAFs, and IKK complex (IKKα and IKKβ kinases and IKKγ), which ultimately determines NF-κB activity (13). Figure 5 shows differential regulation of signaling molecules and gene expression when GTW was implemented 1 day or 5 days after the onset of MIA. For example, Ikbkg (IKKγ), Traf2, Traf3, Traf6, Tank, and Ripk, essential for activation of the IKK complex, were significantly suppressed by GTW initiated on day 1 post–MIA induction. Expression of Irak4, which is required for recruitment of TRAF6 into the signaling complex, was also suppressed by exercise in the MIA + GTW1–21 group. Additionally, receptor (TNF receptor superfamily)–interacting Ser-Thr (RIP) kinases, which activate RIP to bind IKKγ and recruit it to the TNFR1 signaling complex independent of TRAF (30), were also suppressed in the MIA + GTW1–21 group. These findings further suggest that exercise may collectively suppress gene expression required for NF-κB activity and thus inflammation. In this context, the inability of GTW to suppress Traf3, Tank, Ripk1, Ripk3, and Ikbkg in the MIA + GTW5–21 group may be responsible for only partial prevention of the progression of MIA (Figure 5). Previous in vitro studies showing that antiinflammatory actions of mechanical signals are mediated by suppression of NF-κB activation via TGFβ-activated kinase 1, IKK, and IκB (13, 24, 40, 42) also support the present findings.

thumbnail image

Figure 5. Schematic diagram showing the molecular basis of the protective effects of GTW on MIA-induced cartilage damage. GTW preserved cartilage integrity by preventing MIA-induced inflammation and matrix loss in rats in the MIA + GTW1–21 group (rectangles) to a greater extent than in rats in the MIA + GTW5–21 group (ovals). These differential effects stem from the ability of GTW to suppress (green) MIA-induced NF-κB signaling molecules if instituted 1 day after administration of monoiodoacetate, as compared to up-regulation (red) or minimal suppression of these molecules if GTW is instituted 5 days after administration of monoiodoacetate, when grade 1 cartilage damage had already occurred. This in turn may suppress several proinflammatory genes (arachidonate metabolites [Arach], receptors, proteases, chemokines [Chemo], and cytokines) in rats in the MIA + GTW1–21 group, but to a lesser degree in rats in the MIA + GTW5–21 group. Importantly, Aspn, an inhibitor of transforming growth factor β (TGFβ) that was up-regulated in rats with MIA that did not undergo GTW, was significantly suppressed in the MIA + GTW1–21 group. By suppressing asporin expression, GTW may up-regulate TGFβ complex and Sox9, required for synthesis of matrix proteins, in rats with MIA that begin the exercise on day 1. Simultaneous up-regulation of Frzb likely inhibits chondrocyte hypertrophy and mineralization (50). In contrast, the lack of Aspn suppression, and thus lack of expression of the molecules in TGFβ complex and Sox9 in rats not beginning GTW until day 5, may prevent MIA-induced matrix loss to a lesser extent, as evidenced by the suppression of several molecules required for matrix assembly. See Figure 1 for other definitions.

Download figure to PowerPoint

Additionally, the antiinflammatory effects of exercise have been demonstrated in arthritis and other conditions. For example, exercise has been shown to increase levels of the antiinflammatory cytokine IL-10 in the intraarticular and perisynovial fluid of patients with OA (19), to suppress low-grade systemic inflammation (20), to decrease inflammation in patients with diabetes (46), and to suppress IL-8, C reactive protein, and interferon-γ in patients with fibromyalgia (47). These findings demonstrate that in addition to local actions, exercise exerts systemic effects. Whether exercise also suppresses systemic markers associated with cartilage damage in MIA is yet to be determined.

In the context of GTW preventing matrix loss, Figure 5 shows the regulation of signaling molecules and matrix proteins by GTW in MIA-affected cartilage. MIA significantly up-regulates Aspn, a known inhibitor of TGFβ (48). Strikingly, GTW significantly suppressed Aspn expression with parallel up-regulation of molecules in the TGFβ complex in cartilage from rats in the MIA + GTW1–21 group. The anabolic networks of TGFβ may up-regulate Sox9 and together may serve as focal points for the significant up-regulation of many chondrocytic matrix-associated genes, such as aggrecan, collagens, Cilp, Cilp2, Matn3, and Vit, by exercise in the MIA + GTW1–21 group. In contrast, lack of asporin suppression in the MIA + GTW5–21 group may be responsible for the partial repair of MIA-affected cartilage. This dynamic down-regulation of TGFβ via Aspn induction in rats with MIA and counterregulation by GTW suggests an important role of Aspn in exercise-mediated anabolic responses in cartilage (48, 49). Interestingly, cartilage is believed to have very limited capacity to regenerate/repair. The present findings are novel in showing that exercise such as GTW can augment anabolic gene expression to prevent cartilage loss and matrix restructuring in inflamed cartilage.

The present study is the first to delineate the molecular basis for the efficacy of GTW in suppressing progression of cartilage destruction. We show that GTW is a robust promoter of anticatabolic and anabolic networks that suppress inflammation and up-regulate matrix synthesis, even during actively progressive cartilage destruction. Importantly, the effects of exercise appear to be inversely related to the extent of cartilage damage, i.e., its implementation at the earlier stages of cartilage damage may provide greater benefits. Further studies are needed to gain an understanding of how these robust therapeutic effects of exercise can be optimized to prevent cartilage destruction.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Agarwal had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Nam, Perera, Liu, Wu, Rath, Butterfield, Agarwal.

Acquisition of data. Nam, Perera, Liu, Rath, Agarwal.

Analysis and interpretation of data. Nam, Perera, Wu, Agarwal.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. REFERENCES
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
ART_30311_sm_supplementaryFigure1.mpg9245KSupplementary Figure 1
ART_30311_sm_supplementaryFigure2.mpg8462KSupplementary Figure 2
ART_30311_sm_supplementaryFigure3.mpg10359KSupplementary Figure 3
ART_30311_sm_supplementaryFigure4.mpg10799KSupplementary Figure 4
ART_30311_sm_supplementaryFigure5.mpg11358KSupplementary Figure 5
ART_30311_sm_supplementaryFig3.tif8752KSupplementary Figure 3
ART_30311_sm_supplementaryFig4.tif8536KSupplementary Figure 4
ART_30311_sm_supplementaryFig5.tiff9565KSupplementary Figure 5
ART_30311_sm_supplementaryTable1.doc128KSupplemental Data. A table showing GTW-driven regulation of 142 arthritic-related genes defined by IPA database that were regulated more than 2 fold by MIA

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.