IL-3 Decreases Cartilage Degeneration by Downregulating Matrix Metalloproteinases and Reduces Joint Destruction in Osteoarthritic Mice

Osteoarthritis (OA) is a degenerative disease of joints, characterized by progressive loss of cartilage and subchondral bone. Genetic predisposition and environmental factors such as sex, obesity, misalignment, and injuries increase the risk for development of OA (1). These factors cause repeated abnormal mechanical stress at load-bearing joints, leading to its altered biomechanics (2). The origin of OA from cartilage or subchondral bone is not very clear; however, both are interdependent and damage from one subsequently suffuses to the other (3–5).

Cartilage at articular surface aids in smooth mobility of joints with least friction because of tensile strength of matrix secreted by cartilage cells (6). Cartilage is composed of a sparse population of chondrocytes distributed in collagen- and proteoglycan-rich extracellular matrix. Chondrocytes are the first skeleton-specific cell type to appear during embryonic development. They form growth plate in long bones and are of foremost importance in the process of bone formation as they lay the basic framework for longitudinal growth of bones (7, 8). Chondrocytes participate in the synthesis, as well as the degradation, of cartilage matrix and are highly sensitive to pathological changes in the joint microenvironment. Any injury that leads to irreversible physical damage to cartilage induces inflammatory microenvironment in the joints (9, 10). Hypertrophic chondrocytes lose their proliferative potential as well as the property of matrix synthesis, and hence cannot compensate for the cartilage damage (11, 12).

The inflammatory microenvironment at the affected joints attracts immune cells leading to enhanced degeneration of cartilage matrix and bone (13). The proinflammatory cytokines IL-1β and TNF-α are key players in the pathophysiology of OA (14). These cytokines promote the onset of disease and enhance the degenerative processes by stimulating apoptotic and matrix-degrading pathways in affected cartilage (15). Matrix metalloproteinases (MMPs) triggered by proinflammatory cytokines assist in cartilage destruction during arthritic conditions (16).

IL-3, a cytokine secreted by activated T lymphocytes, stimulates proliferation, differentiation, and survival of pluripotent hematopoietic stem cells. It is a broadly acting hematopoietic-regulatory protein, which acts on a number of cell lineages including macrophages, mast cells, neutrophils, eosinophils, and megakaryocytes. Previously, we have documented that IL-3 irreversibly inhibits in vitro osteoclast differentiation induced by receptor activator of NF-κB ligand and TNF-α in both mice and human osteoclast precursors and directs the cells to macrophage and dendritic cell lineages (17–19). We have also shown that IL-3 is a potent inhibitor of pathological bone resorption induced by TNF-α and other proinflammatory cytokines such as IL-1α, TGF-β1, TGF-β3, IL-6, and PGE₂ (20). Recently, we demonstrated that IL-3 has anti-inflammatory activity in vivo and indirectly protects cartilage and bone damage in murine models of inflammatory and rheumatoid arthritis (20, 21). However, the role of IL-3 on chondrocyte differentiation and cartilage degeneration in OA is not yet known.

In this study, we investigated the effect of IL-3 on cartilage degeneration under both in vitro and in vivo conditions. We found that IL-3 upregulates the expression of chondrocyte genes important for matrix synthesis, and downregulates the expression of MMPs under inflammatory conditions in both mouse and human chondrocytes. Interestingly, IL-3 reduces the degeneration of articular cartilage and subchondral bone microarchitecture in mouse model of human OA. Moreover, IL-3 decreases IL-1β–induced matrix degradation in micromass pellets of human mesenchymal stem cells (MSCs). Thus, to our knowledge, we provide the first evidence that IL-3 has a chondroprotective role in OA.

BALB/c mice pups (2–4 d old) and C57BL/6 male mice (8–10 wk old) were obtained from the Experimental Animal Facility of National Centre for Cell Science, Pune, India. Water and food were provided ad libitum. All the protocols involving animal use were approved by an Institutional Animal Ethics Committee.

Nondegenerated human cartilage tissues were harvested from femoral condyles after joint replacement surgeries at Armed Forces Medical College, Pune, India. All the protocols followed for harvesting, handling, processing, and disposal of human samples were approved by an Institutional Ethics Committee.

FITC anti-CD123 (IL-3Rα) and its isotype were from BD Biosciences. Abs for Sox9, collagen type IIa (Col2a), and aggrecan were from Santa Cruz. The fluorochrome-conjugated secondary Abs were obtained from Abcam, and HRP-conjugated secondary Abs were from Bangalore Genei. Abs for MMP-3 and MMP-13 and recombinant cytokines such as human IL-3, IL-1β, TNF-α, and TGF-β3 were obtained from R&D Systems. Recombinant mouse IL-3 and IL-1β were obtained from BD Biosciences. DMEM with high glucose (4.5 g/l), FBS, l-glutamine, TRIzol reagent, cDNA synthesis kit, and SYBR Green were obtained from Invitrogen. Chondrogenesis induction media plus bullet kit was from Lonza. Collagenase and dispase were purchased from MP Biomedicals. BrdU incorporation ELISA was obtained from Roche. Bicinchoninic acid kit and ECL picochemiluminescence substrate were purchased from Thermo-Pierce. Total MMP-3 ELISA kit was obtained from R&D Systems, and the assay was performed as per the manufacturer’s instructions.

Murine chondrocytes were isolated from knee balls (cartilaginous epiphyses of tibia and femur in the tibio-femoral joints) and sternum of mouse pups. Tissues were digested with collagenase (3 mg/ml) at 37°C for 1–2 h for soft tissue removal, followed by second enzyme digestion for 3–4 h for isolation of cells. Cells were washed and cultured in DMEM with high glucose supplemented with 2 mM l-glutamine and 10% FBS.

Human cartilage tissues were digested in medium containing 1% collagenase and 0.2% dispase for 10, 20, 30, 30, 15, and 10 min at 37°C in shaker water bath. The first cell fraction was discarded, whereas rest were pooled, washed, and cultured at a density of 5–10 × 10³ cells/cm². The cells were allowed to adhere to plastic surface and were fed every 2 d.

Expression of IL-3Rα, MMP-1, MMP-3, MMP-13, and chondrocyte-specific genes Sox9, Col2a, and aggrecan was assessed by RT-PCR and real-time PCR using primer sequences (IDT) listed in Table I. RNA was isolated from chondrocytes using TRIzol reagent. Total RNA (2 μg) was used to synthesize cDNA by reverse transcription (cDNA synthesis kit). cDNAs were amplified using PCR for 30–35 cycles. Each cycle consisted of 30 s of denaturation at 94°C, 30 s of primer annealing at 60°C, and 30 s of extension at 72°C. GAPDH and β-actin were used as endogenous controls.

For real-time PCR, 10 μl reaction mixture containing SYBR Green and 10 pmol of each primer were used. PCR was set using StepOnePlus system (Applied Biosystems). The PCR program was comprised of 1 cycle of denaturation at 95°C for 10 min and 40 cycles of denaturation at 95°C for 15 s, primer annealing, and extension at 60°C for 60 s, followed by melt curve analysis. Each reaction was run in duplicates. Data were analyzed for fold difference using comparative 2^−ddCT method.

Chondrocytes were lysed in proteinase and phosphatase inhibitor containing radioimmunoprecipitation assay–based cell lysis buffer. For isolation of proteins from tissues, operated whole-knee joints were crushed in liquid nitrogen and tissue lysates were collected in radioimmunoprecipitation assay lysis buffer. Supernatant obtained after centrifugation at 12,000 × g for 20 min was quantified for total cellular proteins using bicinchoninic acid kit. A total of 40 μg total protein was denatured, subjected to SDS-PAGE (10%), and transferred onto nitrocellulose membrane. Blots were blocked for 30 min in BSA and incubated with primary Ab (1:1000) for 3 h. After washing, the membranes were incubated with HRP-conjugated secondary Ab, and labeled proteins were detected using ECL reagents. Relative intensities for the expression of MMPs were analyzed by densitometry using ImageJ software (National Institutes of Health).

Chondrocytes were characterized by flow cytometry using primary Abs specific for Sox9, aggrecan, Col2a, and IL-3Rα. Cells were fixed in 4% paraformaldehyde, followed by blocking of nonspecific binding in BSA. Cells were then incubated with primary Ab (1:50) for 1 h at 4°C. After washing, cells were treated with fluorochrome-conjugated secondary Ab for 1 h in the dark. For intracellular staining, cells were permeabilized using 0.1% Triton X-100 for 15 min. Cells were washed, resuspended in sheath fluid, and acquired on BD FACS Canto II. Cells incubated with fluorochrome-conjugated secondary or isotype Ab were used as nonspecific controls. Data were analyzed using FACS DIVA software (BD Biosciences).

Coexpression of IL-3Rα and chondrocytic markers was carried out using fluorochrome-conjugated Abs. Chondrocytes cultured on glass coverslips were fixed in 4% paraformaldehyde for 10 min, followed by blocking in BSA. Cells were incubated with primary Ab (1:50) for 4 h, washed, and then incubated with fluorochrome-conjugated secondary Ab for 2 h in the dark. Coverslips were mounted onto glass slides using DAPI-containing mounting medium, and images were acquired using confocal microscope equipped with argon and helium lasers (Zeiss).

Chondrocytes (5 × 10³ cells/well) cultured in 96-well plate were treated for 72 h with different concentrations of IL-3. BrdU incorporation ELISA was performed as per the manufacturer’s instructions.

Cells were treated with different concentrations of IL-3. After 72 h, medium was replaced with MTT (0.5 mg/ml). Formazan crystals were dissolved in 100 μl acidified isopropanol, and absorbance was measured at 570 nm.

Human MSCs were differentiated into chondrocytes using pelleted micromass culture assay (22) with slight modification. In brief, MSCs (2 × 10⁵ cells/well) were seeded in U-bottom 96-well plate and centrifuged at 200 × g for 10 min. The cell pellets were incubated in chondrogenesis induction medium and 10 ng/ml TGF-β3. These cell pellets were half fed every 2 d. After 21 d, cell pellets were fixed in 10% formalin, and 5-μm-thick sections were examined microscopically for the presence of chondrocytes and lacunae by H&E staining and matrix deposition by toluidine blue staining. The images of H&E-stained sections were quantitated for percent matrix and percent matrix degradation using image analysis software ImageJ.

Mouse model of human OA was developed by surgical transection of anterior cruciate ligament (ACL) as described previously (23) with slight modification. In brief, C57BL/6 male mice 8–10 wk old were anesthetized. A para-median skin incision was made on the knee using medial para-patellar approach, and patella was retracted laterally. ACL was identified and transected using microsurgical instruments. The surgical wound was closed in layers by s.c. sutures.

Recombinant mouse IL-3 (100 ng/day in 10 μl injection volume) was injected intra-articularly in OA mice, whereas OA control mice received equal volume of PBS. Effect of IL-3 on OA-associated cartilage and subchondral bone degeneration was evaluated using four different treatment regimens, which were categorized into two different time points (days 15 and 43 postsurgery) and two different therapies (preventive and therapeutic). For day 15 studies, mice were injected with IL-3 from days 1 to 15 (as preventive therapy) and days 8 to 15 (as therapeutic therapy). For day 43 studies, mice were injected with IL-3 from days 1 to 43 (as preventive therapy) and days 8 to 43 (as therapeutic therapy). Mice were sacrificed on days 15 and 43, and limbs were preserved in 10% formalin. The microarchitecture of subchondral bone was evaluated by microcomputed tomography (μ-CT), and the changes in articular cartilage were assessed histologically.

μ-CT of excised bones was performed by a SkyScan 1076 CT scanner (SkyScan, Aartselaar, Belgium) as described previously (24, 25) and following the general guidelines for assessment of the bone microarchitecture in rodents using μ-CT (26). In brief, the knee joints of mice were scanned at 50 kV, 200 μA using a 0.5-mm aluminum filter and a resolution of 9 mm/pixel. Reconstruction of the sections was done by a modified Feldkamp cone-beam algorithm with beam hardening correction set to 50%. For morphometric quantification of subchondral bone, the region between articulating surface of condyles and growth plate in respective bones at knee joint was assessed. Various trabecular bone indices such as trabecular bone volume fraction (BV/TV; %), trabecular number (Tb. N.; 1/mm), connectivity density (Conn.D), structure model index (SMI), and trabecular pattern factor (Tb. Pf.; 1/mm) were evaluated using CT Analyzer software. The binary images of transaxial radiographs of subchondral bone microarchitecture for both femur and tibia were generated by Data Viewer software.

For histological analysis of mouse articular cartilage, formalin-fixed knee joints were decalcified, embedded in paraffin, and sectioned at a thickness of 5 μm. These sections were stained with H&E and structural integrity of cartilage was assessed microscopically. The histological sections were evaluated using OsteoArthritis Research Society International (OARSI) scoring system (27) with modifications: in brief, grade 0 = intact articulating surface and zones of cartilage, and appropriate orientation of chondrocytes; grade 1 = intact superficial zone, hypocellular cartilage, hypertrophic cells, and clonal clusters of chondrocytes near articulating surface; grade 2 = discontinuous articulating surface, degraded matrix, disoriented cells, clonal clusters with four cells, and hypertrophic chondrocytes in hypocellular cartilage; grade 3 = fissured cartilage along with degraded matrix and articulating surface, hypocellular cartilage with hypertrophic cells and clonal clusters with four to six chondrocytes; grade 4 = erosion of superficial zone and excavation of cartilage, clonal clusters with six to eight chondrocytes, hypocellular cartilage with hypertrophic cells and degraded matrix; grade 5 = denuded articulation, presence of fibrocartilage, cysts, and initiation of osteophyte formation; and grade 6 = deformations, osteophyte formation, bone remodeling, and fibrocartilaginous and osseous repair extending above the previous surface.

Results are represented as mean ± SEM. Statistical significance was calculated using one-way ANOVA with a subsequent post hoc Tukey’s test for multiple comparisons. The significance values are defined as *^,$,#p ≤ 0.05, **^,$$,##p ≤ 0.01, ***^,$$$,###p ≤ 0.001, and ****^,$$$$,####p ≤ 0.0001.

To evaluate the effect of IL-3 on chondrocytes, we first examined the expression of IL-3Rα on both mouse and human chondrocytes. Primary cultures of both articular and sternal chondrocytes isolated from mouse knee balls and sternum, respectively, showed expression of Sox9 and IL-3Rα mRNA by RT-PCR (Fig. 1A). The protein expression of Sox9 and IL-3Rα on these chondrocytes was confirmed by Western blotting (Fig. 1B). We observed that in a homogenous population of chondrocytes, >90% of cells showed surface expression of IL-3Rα by flow cytometry (Fig. 1C). Fig. 1D represents the median fluorescence intensity showing expression of IL-3Rα and chondrocyte-specific molecules: aggrecan, Col2a and Sox9. The expression of IL-3Rα and chondrocyte-specific molecules was further confirmed by immunofluorescence microscopy (Fig. 1E). The gene expression of IL-3Rα was also confirmed on human cartilage tissues (Fig. 1F) and cultured human chondrocytes at passage 5 (Fig. 1G) along with the expression of chondrocyte-specific genes such as Sox9, Col2a, and aggrecan. The protein expression of IL-3Rα and Sox9 was also confirmed on human chondrocytes by Western blotting (Fig. 1H), and coexpression of IL-3Rα with Sox9 was observed by immunofluorescence microscopy (Fig. 1I). These results confirm that both mouse and human chondrocytes express IL-3Rα at gene and protein levels. The homogenous populations of mouse and human chondrocytes were used for all further studies.

Because chondrocytes exhibited strong expression of IL-3Rα, we examined the effect of rIL-3 on proliferation and gene expression of chondrocytes. Mouse (Fig. 2A, 2B) and human (Fig. 2C, 2D) chondrocytes were treated with varying concentrations of species-specific IL-3 for 72 h, and the proliferation of chondrocytes was analyzed by MTT assay (Fig. 2A, 2C). The dose-dependent effect of IL-3 on proliferation was also assessed after 72 h by BrdU incorporation ELISA (Fig. 2B, 2D). We observed that IL-3 does not affect the proliferation of both mouse and human chondrocytes, and showed no adverse effect on these cells. Further, we examined the effect of IL-3 on chondrocyte-specific gene expression by incubating mouse chondrocytes for 72 h with different concentrations of IL-3. IL-3 did not alter the expression of Sox9, Col2a, and aggrecan at all the concentrations. Also, IL-3 did not show any significant change in the expression of cartilage matrix-degrading enzymes, MMP-3 and MMP-13 (Fig. 2E). These results indicate that IL-3 does not alter the proliferation and gene expression of chondrocytes.

Next, we examined the effect of IL-3 on chondrogenic differentiation of human bone marrow–derived MSCs using micromass pellet culture assay. Human bone marrow MSCs (2 × 10⁵ cells/ well) were centrifuged in U-bottom 96-well plates, and micromass cell pellets were cultured for 21 d in chondrogenesis induction media with different concentrations of human IL-3. Histological sections of micromass pellets were assessed for the presence of chondrocytes, matrix deposition, and lacunae formation. MSCs differentiated into chondrocytes located in lacunae and showed matrix deposition. We observed no difference in chondrocyte differentiation, lacunae formation, and matrix deposition upon IL-3 treatment (Fig. 2F). Chondrogenic differentiation includes proteoglycan-rich matrix deposition, which was evaluated by staining with toluidine blue. IL-3–treated chondrogenic pellets showed matrix deposition similar to untreated chondrogenic controls (Fig. 2F). Using image analysis software, we further quantitated the percent area covered by stained matrix and unstained white spaces in different regions of micromass pellet as a measure of matrix deposition and degradation, respectively. The percent matrix in IL-3–treated pellets was comparable with untreated chondrogenic pellets (Fig. 2G), and no degradation of matrix was observed upon IL-3 treatment (Fig. 2H). These data suggest that IL-3 does not change the functional properties of differentiating chondrocytes.

IL-1β and TNF-α are the key players in amplifying the inflammatory conditions and induces cartilage degeneration in OA joints (28, 29). Because IL-3 has an anti-inflammatory activity in vitro and in vivo (20, 21), we further investigated the effect of IL-3 on mouse chondrocytes in the presence of IL-1β. Mouse chondrocytes were incubated with IL-1β (5 ng/ml) and different concentrations of IL-3 for 24 h. IL-1β inhibited the expression of chondrocyte-specific genes including Sox9, Col2a, and aggrecan, and upregulated the expression of matrix-degrading enzymes, MMP-3 and MMP-13. In mouse chondrocytes, we observed that IL-3 significantly upregulated the expression of Sox9 and Col2a, which were downregulated by IL-1β. The expression of aggrecan was further downregulated in the presence of IL-3. Interestingly, IL-3 treatment significantly downregulated IL-1β–induced expression of matrix-degrading enzymes: MMP-3 at all the concentrations and MMP-13 at 50 and 100 ng/ml IL-3 (Fig. 3A). This suggests an anabolic role of IL-3 on cartilage matrix by promoting synthesis of matrix components and by downregulation of matrix-degrading enzymes secreted by chondrocytes. In human chondrocytes, both IL-1β and TNF-α individually, as well as in combination, downregulated the expression of Col2a and aggrecan, which was not restored upon IL-3 treatment. No change in the expression of Sox9 was observed in all the groups. IL-3 downregulated the expression of MMP-1, MMP-3, and MMP-13 induced by IL-1β or TNF-α alone. Interestingly, the synergistic effect of IL-1β and TNF-α on the increased expression of all these MMPs was downregulated by IL-3 (Fig. 3B). These data suggest the role of IL-3 in reducing the catabolic processes induced in chondrocytes by proinflammatory cytokines. It also suggests that under inflammatory conditions, IL-3 reduces cartilage damage by regulating matrix degradation mediated by MMPs along with neomatrix synthesis.

In joint articulation, articular cartilage is the prime tissue involved in OA. Because IL-3 prevented cartilage degeneration in vitro, we assessed the in vivo role of IL-3 on OA-associated cartilage damage. For this, we developed an OA model in C57BL/6 mice by surgical transection of ACL. Cartilage in knee joints was evaluated histologically for OA changes at day 8 after ACL transection. The H&E-stained sections of knee joints in OA mice showed disrupted and discontinuous (blue arrows) cartilage, hypocellular cartilage with clonal clusters (black arrows), and presence of hypertrophic chondrocytes or empty lacunae (red arrows) toward the articulating surface (Supplemental Fig. 1A). The superficial zone of resting cells was either eroded or differentiated to hypertrophic stage. Further, evaluation of these histological sections using a modified OARSI scoring system showed a significant degeneration of articular cartilage in OA mice (Supplemental Fig. 1B). These observations confirmed the degeneration of cartilage associated with OA, as well as the development of mouse model of human OA within 8 d of ACL transection.

Next, the effect of mouse IL-3 (100 ng/day by intra-articular injection) on cartilage degeneration was evaluated using four different treatment regimens, which were categorized into two different time points (days 15 and 43 postsurgery) and two different therapies (preventive and therapeutic). For day 15 studies, mice were injected with IL-3 from days 1 to 15 (as preventive therapy) and days 8 to 15 (as therapeutic therapy). For day 43 studies, mice were injected with IL-3 from days 1 to 43 (as preventive therapy) and days 8 to 43 (as therapeutic therapy). Mice were sacrificed on days 15 and 43. In H&E-stained sections of knee joints, we observed degeneration of articular cartilage in OA mice. As compared to disrupted and discontinuous cartilage (blue arrows) in OA mice, there was distinct cartilage articulation and reduced number of hypertrophic chondrocytes in different zones of cartilage in all groups of OA mice treated with IL-3. Also, there were reduced numbers of clonal clusters (black arrows) and hypertrophic cells (red arrows) near the articulating surface of cartilage in IL-3–treated OA mice (Fig. 4A). Further, OARSI-based histological evaluation of knee joint sections showed a significant degeneration of articular cartilage in OA mice. Interestingly, reduced disease severity was observed in all OA mice treated with IL-3 at days 15 and 43 in both preventive and therapeutic regimens as represented in box-and-whisker plot for histological score (Fig. 4B).

Because IL-3 decreased the in vitro expression of MMPs in both mouse and human chondrocytes induced by proinflammatory cytokines, we further checked the effect of IL-3 on these MMPs in vivo in OA mice. Tissue lysates from knee joints of sham, OA, and IL-3–treated OA mice were examined for the presence of MMPs by Western blotting. We found that IL-3 decreased the OA-induced expression of MMP-3 and MMP-13 at day 15 in both preventive and therapeutic treatment regimens (Fig. 4C). However, at day 43, IL-3 decreased these MMPs only in the preventive treatment regimen (Fig. 4D). IL-3 also showed decreased trend, although nonsignificant, in total MMP-3 levels in some treatment regimens as examined by ELISA (Supplemental Fig. 2). All these results suggest that IL-3 plays an important role in reducing cartilage degeneration by regulating the expression of MMPs in OA conditions.

Articular cartilage and subchondral bone constitute the important components of joint articulation. In OA, the degenerative changes in articular cartilage gradually percolate into the subchondral region as both are interdependent (30). ACL transection alters the joint biomechanics and subsequently leads to the development of OA (23). In our mouse model of human OA, the binary images of transaxial radiographs generated by μ-CT of femoral subchondral bone showed destruction of trabecular microarchitecture within 8 d of ACL transection (Supplemental Fig. 3A). In the femoral subchondral bone, trabecular bone volume, connectivity, and geometry of connection were determined. In ACL transected mice, trabecular BV/TV was lower and SMI and Tb. Pf. were higher than the sham control, indicating reduced bone volume and poor geometry (Supplemental Fig. 3B). Topological parameters including Tb. N. and Conn.D were also determined to evaluate the connectivity of subchondral bone. Both Tb. N. and Conn.D are inversely related to Tb. Pf. (26). We observed a significant decrease in Tb. N. and Conn.D, and an increase in Tb. Pf. in OA mice compared with sham controls (Supplemental Fig. 3B). These data suggest the erosion of trabecular bone in the femoral subchondral region and confirmed the development of OA within 8 d of ACL transection in mice. However, no OA-associated changes were observed in tibial subchondral bone (Supplemental Fig. 3C, 3D).

Next, the subchondral bone of IL-3–treated OA mice was evaluated using four different treatment regimens as described earlier. The representative images of trabecular microarchitecture of femoral subchondral bone showed betterment upon IL-3 treatment in both preventive and therapeutic regimens at day 15 (Fig. 5A). The trabecular microarchitecture in IL-3–treated OA mice showed a significant increase in Conn.D in therapeutic treatment regimen and a trend toward improvement in SMI and Tb. Pf. in both preventive and therapeutic treatment regimens. However, no effect on BV/TV and Tb. N. was seen in IL-3–treated mice (Fig. 5B). At day 43, the representative images of trabecular microarchitecture of femoral subchondral bone also showed betterment upon IL-3 treatment in both preventive and therapeutic treatment regimens (Fig. 5C). IL-3 showed significant preservation of SMI and Conn.D in both treatment regimens, whereas Tb. Pf. was significantly protected in the therapeutic treatment regimen. A trend of improvement in BV/TV and Tb. N. was observed in both treatment regimens (Fig. 5D). Tibial subchondral bone showed milder symptoms of OA, and the effect of IL-3 was not evident in OA mice (Supplemental Fig. 4). These results suggest that IL-3 plays an important role in preserving femoral subchondral bone microarchitecture in OA stress.

Because IL-3 reduces cartilage degeneration in OA mice, we determined the effect of IL-3 on matrix secreted during human chondrocyte differentiation in the presence of IL-1β. We first examined the preventative effect of IL-3 on matrix degradation by treating micromass pellets of human MSCs with IL-1β (5 ng/ml) and two concentrations of IL-3 (50 and 100 ng/ml) from days 1 to 21 (Fig. 6A). Histological evaluation showed that the matrix in IL-3–treated chondrogenic pellets was less degraded compared with IL-1β, as indicated by reduced white spaces and increased matrix (pink) in H&E-stained sections (Fig. 6B). Quantitation of percent matrix deposition (Fig. 6C) and percent matrix degradation (Fig. 6D) by image analysis confirmed that IL-3 (100 ng/ml) significantly reduces the matrix degradation induced by IL-1β.

Next, we wanted to determine whether IL-3 could also help in repairing IL-1β–induced matrix degradation. For this, the cell pellets were primed with IL-1β (5 ng/ml) for the first 10 d, and IL-3 (50 ng/ml) was added from days 10 to 21 (Fig. 7A). Histological analysis showed that IL-3 reduced matrix degradation induced by IL-1β (Fig. 7B), increased percent matrix (Fig. 7C), and decreased percent degradation of matrix (Fig. 7D). These results suggest that IL-3 shows a protective effect when the process of matrix degradation was initiated by IL-1β. Further, we examined the therapeutic effect of IL-3 on IL-1β–induced matrix degradation by incubating cell pellets with IL-1β from days 1 to 21, and IL-3 (50 and 100 ng/ml) was added from days 10 to 21 (Fig. 7E). Interestingly, we observed that IL-3 showed therapeutic effect and decreases further degradation of matrix under severe inflammatory conditions (Fig. 7F) as indicated by increased percent matrix (Fig. 7G) and decreased percent degradation (Fig. 7H). These results support that the downregulation of MMPs by IL-3 at transcript level is reducing the degradation of cartilage matrix in inflammatory conditions induced by OA stress.

Chondrocytes participate in anabolic-catabolic balance required for maintenance of cartilage by modulating synthesis and degradation of its matrix (31). Chondrocytes lose their proliferative potential upon terminal differentiation, rendering them incapable to compensate for the cartilage damage (11, 12). Also, any pathological stress that disturbs this anabolic-catabolic balance subsequently induces inflammation and alters the balance of anti-inflammatory and proinflammatory cytokines in the joint microenvironment (9). This aggravates the degenerative process, resulting in irreversible physical damage to cartilage, and gradually involves the whole joint with increase in the severity of OA (29). Besides this, inflammation attracts immune cells and osteoclast precursors, thereby increasing osteoclast formation and bone resorption in subchondral region (32). Various studies have shown that proinflammatory cytokines affect survival and phenotype of chondrocytes in several ways. These cytokines downregulate the expression of genes involved in maintenance of cartilage matrix, and promote hypertrophy and apoptosis of chondrocytes (14). In addition, they induce expression of MMPs, which degrade cartilage matrix, thereby aggravating the degeneration of joints (15). We have previously shown that IL-3 is a potent inhibitor of osteoclast formation and pathological bone resorption in vitro (17–20), and it protects cartilage and bone damage in murine models of inflammatory arthritis (20) and collagen-induced arthritis (21). In this study, we investigated the role of IL-3 in regulation of chondrocyte function in vitro and cartilage degradation in vivo in a mouse model of human OA.

We first confirmed that chondrocytes are responsive to IL-3 by checking the expression of IL-3Rα on both mouse and human chondrocytes. The high affinity receptor for IL-3 consists of a heterodimer of IL-3–specific α-chain and a common β-chain, which is shared with GM-CSF and IL-5 (33). IL-3Rα is expressed by human hematopoietic stem cells (34), endothelial cells and monocytes (35), activated T cells (21), osteoclasts (17–19), osteoblasts, and MSCs (36). In this study, we report for the first time, to our knowledge, that both mouse and human chondrocytes express IL-3Rα at transcript and protein levels. When the effect of IL-3 was examined on the maintenance of cartilage by chondrocytes, we observed that IL-3 does not alter chondrocyte proliferation, gene expression, and matrix maintenance under nonpathological conditions. Interestingly, IL-3 significantly upregulated the expression of genes important for the synthesis of cartilage matrix such as Sox9 and Col2a in mouse chondrocytes, which were downregulated by IL-1β. Moreover, IL-3 significantly downregulated IL-1β–induced expression of matrix-degrading enzymes, MMP-3 and MMP-13. Interestingly, IL-3 downregulated the expression of matrix-degrading enzymes, MMP-1, MMP-3, and MMP-13, in human chondrocytes in the presence of proinflammatory cytokines, whereas no change in the expression of Sox9, Col2a, and aggrecan was observed. This suggests that IL-3 exerts its protective action by downregulating the expression of matrix-degrading enzymes induced by proinflammatory cytokines and also promotes neomatrix synthesis without increasing the proliferative potential of mouse chondrocytes. Whereas in humans, IL-3 mainly acts by downregulating matrix-degrading enzymes induced in the presence of proinflammatory cytokines. The components of cartilage matrix, collagens and proteoglycans, constitute the substrate for these MMPs. MMP-13, also known as collagenase-3, mainly degrades Col2a and, therefore, plays a central role in the progression of cartilage degeneration. Compared with other MMPs, MMP-13 has a higher catalytic activity for Col2a and gelatin, thereby making it the most potent peptidolytic enzyme among collagenase family (37). Also, MMP-13 overexpressing transgenic mice undergo spontaneous degeneration of articular cartilage characterized by excessive degradation of Col2a and aggrecan loss (38). We observed that IL-3 reduces cartilage damage mainly by downregulating the expression of MMPs and subsequent matrix degradation, as well as by promoting neomatrix synthesis in some cases and not by enhancing the proliferative potential of chondrocytes. To assess the translational implication of this observation, IL-1β–treated micromass pellets of human MSCs were cultured in the presence of IL-3 during the course of chondrogenic differentiation and were evaluated for matrix degradation. In agreement with other reports, we also observed that IL-1β hampers matrix deposition and induces its degradation (39, 40). Interestingly, IL-3 showed both preventive and therapeutic effect on IL-1β–induced matrix degradation. These observations supported that the decreased matrix degradation by IL-3 is a result of transcriptional downregulation of MMPs.

We further validated our in vitro studies by examining the in vivo effect of IL-3 in a mouse model of human OA. The present scientific understanding of OA suggests that the degeneration in cartilage parallels microarchitectural changes in subchondral bone (41). We evaluated both articular cartilage and subchondral bone in an ACL-transected mouse model to understand the role of IL-3 in the pathophysiology of OA. Microscopic evaluation showed hypertrophic chondrocytes toward the articulating surface of cartilage within 8 d of ACL transection. This suggests that the inflammatory conditions developed because of altered biomechanical stress in ACL-transected joints advances chondrocytes toward terminal differentiation. This process lays a foundation for osteophyte formation in OA joints as well as cartilage thinning by expansion of subchondral bone into cartilage (42, 43). In IL-3–injected OA mice, a proper cartilage zonation reflecting the structural integrity of cartilage was maintained along with the maintenance of zone-specific phenotype of chondrocytes. Also, the reduced number of clonal clusters in articular cartilage upon IL-3 treatment suggests the decreased severity of disease progression. This was confirmed by the decreased levels of OA-induced MMP-3 and MMP-13 in IL-3–treated OA joints.

Articular cartilage and subchondral bone are interdependent entities, and the tissue origin of OA is uncertain (3, 4). ACL transection alters biomechanics of the joint that attracts inflammatory mediators and osteoclast precursors in the affected joints. The newly formed osteoclasts from these precursors along with resident osteoclasts promote subchondral bone resorption. OA-associated damage reaches its peak in 15 d and remains persistent thereafter (23). Evaluation of subchondral bone in OA mice showed significant damage to trabecular microarchitecture within 8 d of ACL transection and was observed until day 43. The quality and trabecular microarchitecture of femoral subchondral bone was improved upon IL-3 treatment in OA mice at day 43. The protection of subchondral bone in the presence of IL-3 could be attributed to decreased osteoclast formation, and hence the bone resorption in the subchondral region (17–19). However, tibial subchondral bone showed milder symptoms of OA, and the protective effect of IL-3 was not evident. The in vitro and in vivo studies in mice strongly suggest that IL-3 has a multifaceted role in amelioration of cartilage and subchondral bone damage in addition to modulating the inflammatory response associated with OA. We also speculate that the downregulation of MMPs by IL-3 might be responsible for protection of cartilage in earlier models of arthritis (20, 21) by reducing the availability of cartilage-derived Ags in the joint microenvironment, leading to diminution of inflammatory conditions. Thus, our results suggest that IL-3 has a therapeutic potential for cartilage and bone degeneration associated with OA.

We thank Drs. Amod Kale and Pralhad Wangikar for histopathological assessment.

The authors have no financial conflicts of interest.