If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address correspondence and reprint requests to: Y. Yazici, Samumed LLC, 9381 Judicial Drive, Suite 160, San Diego, CA 92121, USA. Tel: 1-1858-926-2962.
Osteoarthritis (OA) is a degenerative disease characterized by loss of cartilage and increased subchondral bone within synovial joints. Wnt signaling affects the pathogenesis of OA as this pathway modulates both the differentiation of osteoblasts and chondrocytes, and production of catabolic proteases. A novel small-molecule Wnt pathway inhibitor, SM04690, was evaluated in a series of in vitro and in vivo animal studies to determine its effects on chondrogenesis, cartilage protection and synovial-lined joint pathology.
Design
A high-throughput screen was performed using a cell-based reporter assay for Wnt pathway activity to develop a small molecule designated SM04690. Its properties were evaluated in bone-marrow-derived human mesenchymal stem cells (hMSCs) to assess chondrocyte differentiation and effects on cartilage catabolism by immunocytochemistry and gene expression, and glycosaminoglycan breakdown. In vivo effects of SM04690 on Wnt signaling, cartilage regeneration and protection were measured using biochemical and histopathological techniques in a rodent acute cruciate ligament tear and partial medial meniscectomy (ACLT + pMMx) OA model.
Results
SM04690 induced hMSC differentiation into mature, functional chondrocytes and decreased cartilage catabolic marker levels compared to vehicle. A single SM04690 intra-articular (IA) injection was efficacious in a rodent OA model, with increased cartilage thickness, evidence for cartilage regeneration, and protection from cartilage catabolism observed, resulting in significantly improved Osteoarthritis Research Society International (OARSI) histology scores and biomarkers, compared to vehicle.
Conclusions
SM04690 induced chondrogenesis and appeared to inhibit joint destruction in a rat OA model, and is a candidate for a potential disease modifying therapy for OA.
. The disease is characterized by breakdown of articular cartilage and growth of subchondral bone causing pain, decreased mobility, limitation of function and failure of synovial joints
. There is therefore an unmet need for disease modifying OA drugs (DMOADs). Mesenchymal stem cells (MSCs) in synovium and subchondral bone are capable of differentiation into cartilage forming chondrocytes, bone forming osteoblasts, or adipocytes
, suggesting the failure to regenerate articular cartilage may not be due to insufficient stem cell supply, but rather their inappropriate differentiation when attempting to restore healthy cartilage homeostasis.
Wnt proteins interact with stem and differentiated cells to orchestrate organogenesis, cell differentiation, morphogenesis and tissue remodeling. In adults, the role of the Wnt pathway in tissue formation is extended to homeostatic control through the tightly regulated differentiation of resident stem cells to replenish and repair adult tissues
. Activation of Wnt signaling is initiated by Wnt proteins binding to their cognate receptors, and is modulated by endogenous antagonists such as Dickkopf, Wnt-inhibitory signaling protein, secreted Frizzled related protein and Cerberus. In the canonical Wnt/β-catenin-signaling pathway, binding of Wnt proteins to cell surface receptors leads to stabilized β-catenin levels by inhibition of β-catenin phosphorylation and proteasome-mediated degradation. Stabilized β-catenin translocates to the nucleus and interacts with the T-cell factor/lymphoid enhancer factor (TCF/LEF) family of transcription factors to activate Wnt target genes
. Wnt signaling is subject to complex modulation at multiple levels, resulting in signaling acting more as rheostats, than binary switches in most tissues. Modulating this pathway is an attractive therapeutic approach for regenerative medicine; however, success has been limited due to lack of potent, safe therapeutic agents
Recently, increased understanding of cartilage growth mechanisms during OA development and aging have identified new molecular mechanisms and targets. Amongst these, Wnt signaling has been shown to play a critical role in OA pathogenesis, particularly cartilage and subchondral bone remodeling
. A single nucleotide polymorphism analysis demonstrated an association of both proximal femur bone shape and risk of hip OA with an Arg324Gly substitution mutation in FrzB, a Wnt antagonist, implicating the Wnt pathway as a potential regulator of the disease
. At a molecular level, the balance of Wnt signaling maintains osteoblast and chondrocyte lineage fates and their homeostasis. Increased mechanotransductive force increases Wnt signaling in the joint, which results in osteoblast formation and release of proteases that remodel articular cartilage into an osteoconductive matrix, while decreased Wnt signaling stimulates chondrogenesis
. Wnt signaling has also been implicated in the induction of protease production, especially matrix metalloproteinases (MMP1, MMP3 and MMP13) by chondrocytes and synovial tissue in response to mechanical stress and pro-inflammatory cytokines
. In the context of a joint with abnormal loading, this has the potential to lead to a progressive disease such as OA, with Wnt playing a central role in its pathogenesis.
In this study, SM04690, a small-molecule Wnt pathway inhibitor being developed as a potential DMOAD, was evaluated for its effects on chondrocyte differentiation, cartilage regeneration and protection, and prevention of joint destruction in a preclinical model of knee OA.
Methods
Cell culture and differentiation
SW480 cells (ATCC) cultured in Dulbecco's modified Eagle's medium (DMEM, ThermoFisher, Carlsbad, CA, USA) with 10% fetal bovine serum (FBS) were transduced with the TCF/LEF-luciferase lentivirus (Qiagen, Germantown, MD, USA) and stable clones were selected using puromycin. Primary bone-marrow-derived human mesenchymal stem cells (hMSCs) (Lonza Inc., Basel, Switzerland) were grown in MSCGM™ mesenchymal stem cell growth medium (Lonza), and used between passages 2 and 6.
High-throughput screening
A chemical library was screened in a cellular TCF/LEF-reporter-based assay using SW480 cells. Compounds were transferred into screening plates using ECHO 550 (Labcyte, San Jose, CA, USA). SW480 cells were plated (3000/well) in DMEM with 1% FBS. After 48 h, BrightGlo (Promega, Madison, WI, USA) luminescence was measured using Envision (Perkin Elmer, Waltham, MA, USA). For chondrogenic differentiation, hMSCs were plated (20,000/well) in incomplete chondrocyte differentiation medium (iCDM; Lonza) and treated with SM04690 or compounds – FH535, IWR-1, ICG001, iCRT14, KY02111, and, CX-4945 (Sigma–Aldrich) or Transforming Growth Factor β3 (TGFβ3; 20 ng/ml, Peprotech Inc., Rocky Hill, NJ, USA). On day 5, cells were fixed, stained with 1 μg/mL Rhodamine B and imaged using Cellomics CellInsight (ThermoFisher).
Chondrocyte differentiation
For chondrocyte differentiation, hMSCs were plated in either 48-well plates (40,000/well) or dispensed into 15 ml conical tubes (150,000/tube) in iCDM and treated with SM04690 in dimethyl sulfoxide (DMSO) or TGFβ3 (20 ng/ml). Cells were incubated for 21 days, with media changes every 5 days, fixed and immunostained with specific antibodies or with Alcian Blue (1% in acetic acid, pH 2.5) or Safranin O (0.1% aqueous solution). Cells were imaged using EVOS FL Microscope (ThermoFisher). Gene expression was measured by quantitative real time PCR (qRT-PCR) using SYBR Green based, gene specific primers.
Primary calvarial cells were isolated from mouse E13.5 embryos using collagenase (Sigma–Aldrich, St. Louis, MI, USA) and plated in poly-l-lysine coated plates. After 5 days cells were treated with DMSO or SM04690. On day 21, cells were fixed and stained with Alcian Blue.
hMSCs were differentiated into chondrocytes using TGFβ3 for 21 days followed by treatment with either tumor necrosis factor (TNF)α (20 ng/ml) + Oncostatin M (10 ng/ml) or IL1β (10 ng/ml) and SM04690 for 72 h. Chondrocytes were digested with papain (Sigma). GAG content was measured using the dimethylmethylene blue (DMMB) kit (Chondrex, Redmond, WA, USA) and absorbance at 535 nm was measured using Cytation 3 (Biotek, Winooski, VT, USA). NO was measured using Greiss reagent kit (Promega).
Pharmacokinetics
Following single intra-articular (IA) SM04690 injection into Sprague–Dawley (SD) rats (12 weeks old, male), knee joints were collected, flash frozen and stored at −70°C. Bone and cartilage samples were isolated from tibias, homogenized and SM04690 extracted with acidified organic solvent. Extracts were analyzed using Phenomenex PFP column and high performance liquid chromatography (HPLC) gradient method in tandem to a triple quadrupole mass spectrometer (API Triple Quad 3000) with a Turboionspray source for detection in positive ion mode.
Surgery-induced OA model
All animal studies were performed in accordance with approved Samumed, LLC Animal Committee protocols. Male SD rats were housed at Samumed LLC, and provided food and water ad libitum. At 10 weeks postnatal age, 24 rats were subjected to severing of the anterior cruciate, medial collateral and medial meniscotibial ligaments (anterior cruciate ligament transection and partial medial meniscectomy model; ACLT + pMMx). One-week post-surgery, all rats were randomized and given IA SM04690 (0.3 μg in 50 μl) or vehicle (n = 12 rats/group). Thirteen weeks post-surgery, knee joints were isolated, fixed in 10% formalin, decalcified and embedded in paraffin blocks. Frontal sections (5 μm thick, three sections to ensure redundancy) from different levels, each 100 μm apart, were obtained and stained with Safranin O/Fast Green. At least 12 sections were obtained from each rat and imaged using a light microscope (EVOS FL, Life Technologies).
Osteoarthritis Research Society International (OARSI) scoring, cartilage protection and regeneration
Histological evaluation was performed by two blinded observers. Images were scored based on the OARSI cartilage histopathology scoring system by evaluating damage based on grade (depth of progression into cartilage) and stage (extent of joint involvement)
. Briefly, femurs and tibias were evaluated separately, and each assigned a score for cartilage damage grade on a scale of 0–6 (0 – intact surface; 6 – extensive deformation) and cartilage damage stage on a scale from 0 to 4 (0 – normal joint; 4 – >50% involvement). Total score was the product of Grade × Stage (0 representing normal joint; 24 representing severe OA)
. Twelve sections were scored per rat and no rats were excluded. Following scoring, the study was unblinded and the four sections from each rat with lowest scores (representing least cartilage damage) were excluded from further analysis. Mean OARSI score for each rat was generated by averaging the scores from the two blinded observers. Additionally, histological evaluation was repeated by four independent blinded observers based on a modified objective quantitative histomorphometrical OARSI scoring system
Quantification of articular chondrocytes was performed using immunohistochemistry for doublecortin (Dcx), followed by imaging (Keyence, confocal mode), and analysis using Cellomics (ThermoFisher). Levels of circulating biomarkers were measured by ELISA on weeks 3, 4, and 6.
In a second study, 24 rats (male, 10-weeks old) were subjected to the ACLT + pMMx procedure. One-week post-surgery all rats were randomized and given an IA injection of SM04690 (0.3 μg) or vehicle (n = 12 rats/group). On week 5 after surgery, cartilage was isolated from the rats for gene expression measurement of chondrocyte markers and proteases.
Statistical analysis
Statistical analysis was performed using Prism 7 (Graphpad Inc., La Jolla, CA, USA). EC50 values were obtained using sigmoidal dose–response curve-fitting. For those outcomes in which an underlying normal distribution was assumed, an independent, two-tailed t test was utilized for two group comparisons, and for more than two groups one-way analysis of variance (ANOVA) was utilized. OARSI scores, for which the normal distribution was not assumed, were evaluated using Mann–Whitney U test. Data were represented as Mean ± 95% confidence interval (CI) or Mean ± standard deviation (SD) as noted in the figure legends, and significance values listed. Significance was set at P < 0.05. *P < 0.05, **P < 0.01, ***P < 0.001.
Results
Wnt pathway modulation – cellular high-throughput screening identified SM04690 as a potent and selective inhibitor of Wnt signaling
Small-molecule inhibitors of Wnt signaling were identified using a high-throughput TCF/LEF-reporter assay in SW480 colon cancer cells bearing a mutation in the adenomatous polypopis coli (APC) protein
, which leads to constitutively active canonical Wnt signaling. Hits were counter-screened against SW480 cells expressing a SV40-driven-luciferase reporter, to eliminate compounds that acted non-specifically. Based on the elucidated structure activity relationship from these compounds, iterative medicinal chemistry efforts generated SM04690 (N-(5-{3-[7-(3-fluorophenyl)-3H-imidazo [4,5-c] pyridin-2-yl]- 1H-indazol-5-yl} pyridin-3-yl)-3-methylbutanamide; CAS number: 1467093-03-3; [Supplementary Fig. 1(a)]), which inhibited the TCF/LEF reporter (EC50 = 19.5 nM) and did not affect the SV40 luciferase reporter [Fig. 1(a)]. The activity of SM04690 was then compared with known Wnt pathway inhibitors such as FH535
. SM04690 was found to be ∼150- to 500-fold more potent than the other compounds [Supplementary Fig. 1(b) and (c)]. Inhibition of Wnt signaling was confirmed with SM04690 treatment in SW480 cells by qRT-PCR, and Western blot for β-catenin and its target proteins [Fig. 1(b) and (c)]. Bone-marrow-derived hMSCs (CD29+, CD44+, CD166+, CD105+, CD45−), treated with SM04690, showed a dose-dependent decrease in the expression of Wnt pathway genes (ASCL1, LEF1, TCF7L2, TCF7, C-MYC and AXIN2), as measured by qRT-PCR [Fig. 1(d)] and Western Blot [Fig. 1(e)]. SM04690 also inhibited the expression of AXIN2, TCF7 and LEF1 in hMSCs [Supplementary Fig. 1(d)] and AXIN2 and LGR5 in IEC6 [intestinal stem cells sensitive to Wnt activation; Supplementary Fig. 1(e)] when the Wnt pathway was selectively activated using either Wnt3a or a GSK3β inhibitor, CHIR-99021. The inhibition of Wnt target genes by SM04690 (30 nM) was comparable to the activity of CX-4945
. SM04690 had minimal effects on the non-canonical Wnt pathway and the BMP pathway (Supplementary Figs. 2 and 3). Changes in the expression of genes in these pathways (e.g., decreased expression of DAAM1, RHOA, WNT4, BMP4 and increased expression of DKK1, NLK1, WNT16, MAPK13, SMAD7) correlated with Wnt pathway inhibition and the induction of chondrogenic differentiation
Fig. 1SM04690 was a potent and specific inhibitor of canonical Wnt signaling. SW480 cells and hMSCs treated with SM04690 or DMSO for 48 h. (a) Dose response of TCF/LEF promoter driven- or SV40 promoter driven-luciferase reporters in SW480 cells (n = 4, Mean ± 95% CI). (b) Expression of genes in the Wnt signaling pathway in SW480 cells as measured by qRT-PCR. Fold change relative to DMSO (n = 3, Mean ± 95% CI, Ascl1 ***P < 0.0001, LEF1 ***P = 0.0001, TCF7L2 **P = 0.0038, TCF7 ***P = 0.0005, c-myc ***P = 0.0003, Axin2 *P = 0.014, t test). (c) Expression of proteins in the Wnt signaling pathway in SW480 cells as measured by Western blot. (d) Expression of genes in the Wnt signaling pathway in hMSCs as measured by qRT-PCR. Fold change relative to DMSO (n = 3, Mean ± 95% CI, Ascl1 **P = 0.002, LEF1 *P = 0.017, TCF7L2 **P = 0.004, TCF7 **P = 0.0053, c-myc ***P < 0.0001, Axin2 ***P = 0.0008, t test). (e) Expression of proteins in the Wnt signaling pathway in hMSCs as measured by Western blot.
Chondrogenesis occurs as a result of MSC and progenitor cell differentiation. To assess SM04690 effects on early chondrogenesis, hMSCs were treated for 5 days, and Rhodamine B stained chondrogenic nodules
were measured as an indicator of early cell condensation phenotype associated with chondrogenesis induction [Fig. 2(a)and (b)]. SM04690 promoted aggregation of hMSCs in a dose-dependent manner [EC50 = 10 nM, Supplementary Fig. 4(a)], increasing Rhodamine B stained colonies >40-fold (P < 0.0001), compared to DMSO treated control. TGFβ3
by immunocytochemistry; Fig. 2(c) and (d)] and Western blot [Supplementary Fig. 4(b)], compared to DMSO, indicating an early chondrocyte lineage phenotype.
Fig. 2SM04690 induced chondrocyte differentiation from hMSCs. (a) hMSCs treated with SM04690 (30 nM) or DMSO or TGFβ3 (20 ng/ml) for 5 days and stained with Rhodamine B (scale bars, 200 μm). (b) Quantification of the number of Rhodamine B stained colonies in (a) (n = 6, Mean ± 95% CI, ***P < 0.0001, t test). (c) hMSCs treated with SM04690 (30 nM) or DMSO or TGFβ3 (20 ng/ml) for 3 days and stained for Sox9 (scale bars, 200 μm). (d) Quantification of the number of Sox9 positive cells in (c) (n = 36 images, Mean ± 95% CI, ***P < 0.0001, t test). (e) hMSCs treated with SM04690 (30 nM) or DMSO in 3-D sphere culture for 21 days and stained with Safranin O for mature chondrocytes (scale bars, 200 μm). (f) hMSCs treated with SM04690 (30 nM) or DMSO for 21 days and stained for various markers for mature chondrocytes (scale bars, 200 μm). (g) Total sulfated GAG in hMSCs treated with SM04690 (30 nM) or DMSO for 21 days measured by DMMB assay. Total GAG levels relative to the weight of the cell aggregates shown (n = 3, Mean ± 95% CI, *P = 0.041, t test).
hMSCs treated with SM04690 in high density three dimensional, 21-day pellet culture conditions efficiently differentiated into mature chondrocytes, as evidenced by larger pellets with increased Safranin O staining relative to DMSO treated control [Fig. 2(e)]. Additionally, hMSCs treated with SM04690 for 21 days under monolayer conditions demonstrated chondrocyte differentiation, with the presence of mature chondrocyte specific proteins and cartilage matrix components (tissue inhibitor of metalloproteinase 1 (TIMP1), Type II collagen, aggrecan, Alcian Blue, Safranin O, Rhodamine B, Toluidine Blue and CD44)
as compared to DMSO treated cells [Fig. 2(f)]. Treatment with SM04690 also induced significant increases in the sulfated GAG content of differentiated chondrocytes
Following SM04690 treatment, qRT-PCR analysis of differentiated cells further confirmed increased expression of chondrocyte associated genes, including SOX9, cartilage oligomeric matrix protein (COMP), aggrecan, COL2A1, TGFβ1, TIMP1, CD44, and Type 10 collagen (COL10A1) [Fig. 3(a)], as compared to DMSO treatment, with expression levels greater than those induced by CX-4945 (10 μM; P < 0.05 for all genes except COL2A1) and KY02111 (10 μM; P < 0.05) [Supplementary Fig. 5(a)]. SM04690 treatment decreased expression of genes associated with osteoblast differentiation (BGLAP [osteocalcin], ALPL [Alkaline phosphatase], BMP4, RUNX2)
and tendon or ligament differentiation (COL1A1) [P < 0.01, Fig. 3(b)] with levels comparable to CX-4945 (10 μM) and KY02111 (10 μM) treatment [Supplementary Fig. 5(b)]. While increased COL10A1 expression was observed in vitro, a corresponding increase was not observed in the in vivo model treated with SM04690 [Supplementary Fig. 9(a)], suggesting increased COL10A1 expression may be a cell culture system result and not indicative of hypertrophic differentiation
Further, the ability of SM04690 to induce differentiation of mouse progenitor cells into chondrocytes was evaluated. SM04690 treatment for 21 days induced chondrocyte differentiation in ATDC5 cells (mouse chondrogenic cell line
and cultured in the presence of SM04690 (100 nM) for 21 days showed a marked increase in the number and size of Alcian Blue stained chondrocytes, increasing the area of field covered with chondrocytes to >30% compared to <3% in the DMSO treated control [Fig. 3(d) and (e)].
Since SM04690 inhibited Wnt signaling, several commercially available Wnt inhibitors were tested in the early chondrogenic aggregation and chondrocyte differentiation assays. In line with their activities in the TCF/LEF-reporter assay [Supplementary Fig. 1(b) and (c)], several Wnt pathway inhibitors were ∼50- to 500-fold less potent than SM04690 [Supplementary Fig. 6(a) and (b)]. SM04690 also induced ∼3- to 10-fold more Alcian Blue and Safranin O stained chondrocyte colonies, at ∼200-fold lower EC50 as compared to CX-4945, KY02111 and ICG001 [Supplementary Fig. 6(c)–(h)].
Cartilage protection – Wnt pathway inhibitor SM04690 protected chondrocytes from catabolic breakdown in vitro
In addition to regeneration of cartilage, long-term disease modification in OA requires inhibited degradation of newly formed and existing cartilage. Therefore, the effects of SM04690 on chondrocytes and hMSCs and cartilage catabolism under pathophysiological OA-like conditions were evaluated. Chondrocytes and hMSCs were treated either with a combination of TNFα (20 ng/ml) and Oncostatin M (10 ng/ml), or IL1β (10 ng/ml) to mimic cytokine-induced cartilage degeneration during OA progression. When treated with these cytokines, cultured chondrocytes and hMSCs upregulated the expression of MMP-1, MMP-3, MMP-13 and Indian hedgehog (IHH)
[Fig. 4(d)], as measured by the Greiss reaction, compared to untreated chondrocytes. Treatment with SM04690 inhibited cytokine-induced expression of matrix-degrading enzymes (MMP-1, MMP-3, MMP-13, IHH) [Fig. 4(a) and (b), Supplementary Fig. 7(a)–(d)], with potency similar or improved (P < 0.05) compared to CX-4945 (10 μM) and KY02111 (10 μM). Consistent with this effect, SM04690 treatment also significantly decreased breakdown and release of GAGs [Fig. 4(c)] and NO [Fig. 4(d)] compared to controls (P < 0.05).
Fig. 4SM04690 protected chondrocytes from catabolic breakdown in vitro. Chondrocytes treated with cytokines TNFα (20 ng/ml) + Oncostatin M (10 ng/ml) or IL1β (10 ng/ml) and SM04690 (30 nM) or DMSO control for 72 h. (a, b) Gene expression of proteases (MMP1, MMP3, MMP13, IHH) in chondrocytes measured by qRT-PCR. Fold change relative to unstimulated control (n = 3, Mean ± 95% CI, TNFα + Oncostatin M: MMP1 *P = 0.04, MMP3 *P = 0.014, MMP13 **P = 0.0046, IHH *P = 0.022; IL1β: MMP1 *P = 0.019, MMP3 *P = 0.017, MMP13 *P = 0.015, IHH *P = 0.026, t test). (c) Levels of secreted GAG expressed as a ratio of intracellular GAG, measured by the DMMB assay (n = 6, Mean ± 95% CI, TNFα + OM ***P < 0.0001, IL1β *P = 0.032, t test). (d) Levels of secreted nitric oxide (NO) measured using the Griess reagent assay (n = 6, Mean ± 95% CI, TNFα + OM ***P < 0.0001, IL1β **P = 0.005, t test).
Pharmacokinetics – IA SM04690 demonstrated sustained local and minimal systemic exposure
Pharmacokinetic evaluation of a single IA SM04690 injection in rats demonstrated knee joint residence time >180 days [Supplementary Fig. 8(a)]. Systemic exposure of drug was below quantifiable plasma levels (lower limit of quantification = 10 nM) and no obvious adverse effects (weight loss, significant swelling, signs of pain or distress) were observed in the treated rats.
Cartilage regeneration – IA SM04690 promoted cartilage growth and improved joint health in a rat model of knee OA
The ACLT + pMMx rat OA model causes severe joint destabilization, leading to OA-like disease with cartilage degradation within 1–2 weeks
Characterization of articular cartilage and subchondral bone changes in the rat anterior cruciate ligament transection and meniscectomized models of osteoarthritis.
. The efficacy of SM04690 was evaluated in the ACLT + pMMx rat knee OA model, following a single IA injection of SM04690 (0.3 μg) or vehicle control. Histological examination of the joints 12 weeks post vehicle injection (13 weeks post-surgery) showed loss of smooth cartilage articular surface and decreased intercondylar space between the femur and tibia in these rats [Fig. 5(a)]. By contrast, SM04690-treated animals showed increased articular cartilage and smooth cartilage surface as indicated by significantly increased Safranin O–Fast Green staining intensity [P < 0.0001; Supplementary Fig. 8(c)] and cartilage thickness measurements [P = 0.0042; Supplementary Fig. 8(d)]. Blinded OARSI histopathology analysis
of the most severely affected joint areas, revealed a significant decrease (P = 0.008) in treatment group scores [Fig. 5(b)] compared to vehicle controls. Additionally, blinded scoring performed using a modified objective quantitative histomorphometrical OARSI scoring method
also showed a significant decrease (P = 0.0005) in the OARSI score with SM04690 treatment [Supplementary Fig. 8(b)], indicating improvement in joint morphology.
Fig. 5SM04690 promoted cartilage repair and protection in the rat ACLT + pMMx OA model. ACLT + pMMx rats treated with either vehicle or SM04690 (0.3 μg). (a) Representative images of medial tibial plateau of the knee joint stained with Safranin O–Fast Green from naïve or vehicle-treated or SM04690 (0.3 μg) treated rats 12 weeks from treatment. (b) The medial joint score based on the OARSI scoring system (n = 12 rats, Mean ± 95% CI, **P = 0.008, Mann–Whitney U test). (c) Circulating PIIANP levels following treatment (n = 12 rats, Mean ± 95% CI, week 3 *P = 0.018, week 5 P = 0.13, t test). (d) Gene expression of chondrocyte markers in the cartilage of rats 4 weeks from treatment measured by qRT-PCR (n = 7 for vehicle, n = 8 for treatment, Mean ± 95% CI, Col2a1 *P = 0.024, COMP ***P = 0.0007, aggrecan ***P = 0.0014, t test). (e) Representative images of the superficial zone of articular cartilage from the ACLT + pMMx model stained for Dcx. (f) Quantification of Dcx expressing chondrocytes in (e) (n = 9 rats/group for vehicle, n = 12 rats/group for treatment, Mean ± 95% CI, **P = 0.001, t test). (g) Gene expression of proteases in the cartilage of rats 4 weeks from treatment measured by qRT-PCR (n = 7 for vehicle, n = 8 for treatment, Mean ± 95% CI, MMP1 ***P = 0.0006, MMP3 ***P = 0.0006, MMP13 **P = 0.0046, ADAMTS5 *P = 0.016, t test). (h) Circulating COMP levels following treatment (n = 12 rats, Mean ± 95% CI, week 2 P = 0.19, week 5 **P = 0.0064, t test).
Serum was evaluated during the 12-week post-treatment period for N-terminal propeptide of collagen IIA (PIIANP), a biomarker of cartilage matrix synthesis
. Levels of PIIANP in serum were significantly increased (P = 0.0186) at 3 weeks post-treatment in rats treated with SM04690 as compared to vehicle control and remained slightly elevated even at 5 weeks post-treatment [P > 0.05; Fig. 5(c)].
Additionally, the onset of chondrogenic differentiation in the SM04690-treated animals was confirmed by increased gene expression of Col2a1, aggrecan and COMP in the cartilage as compared to vehicle treatment [Fig. 5(d)]. However, no significant changes were observed in the expression levels of Col10a, a chondrocyte hypertrophy marker [Supplementary Fig. 9(a)], after 4 weeks from treatment with SM04690, as compared to vehicle. Further, quantification of chondrocytes in the superficial zone of the articular cartilage at 12 weeks post injection showed a significant increase (P = 0.001) in the number of Dcx positive chondrocytes
in SM04690-treated rats compared to vehicle [Fig. 5(e) and (f)].
Cartilage protection – single IA injection of SM04690 decreased cartilage breakdown in a rat model of knee OA
Local regulation of matrix degradation markers has previously been demonstrated by a significant increase in mRNA levels of ADAMTS4 and MMP-13 in animal models of OA
Characterization of articular cartilage and subchondral bone changes in the rat anterior cruciate ligament transection and meniscectomized models of osteoarthritis.
. MMP-1, MMP-3, MMP-13 and ADAMTS5 were highly expressed in the articular cartilage of anterior cruciate ligament (ACL)-transected, vehicle-treated animals. In SM04690-treated rats, significant (P < 0.05) reductions of protease levels (approximately 2- to 3-fold) were observed at 4 weeks post injection as compared to vehicle-treated animals [Fig. 5(g)]. Further, circulating COMP levels, which correlate with OA disease severity
, were slightly reduced in SM04690-treated rats as compared to control rats at 2 weeks post-treatment, and significantly reduced (P = 0.0064) at 5 weeks post-treatment [Fig. 5(h)].
Wnt pathway modulation – SM04690 inhibited the Wnt pathway in a rat model of knee OA
The expression of Wnt pathway genes in the cartilage from ACLT + pMMx rats was evaluated using a qRT-PCR panel comprising 84 Wnt pathway genes. Decreased expression (P < 0.05) of several Wnt pathway genes (e.g., GSK3β, Dvl1, Wnt3a, TCF7, Axin2, β-catenin), and upregulation in Wnt inhibitory genes (e.g., DKK1, WIF1) was observed in the cartilage of animals treated with SM04690 as compared to vehicle controls [Supplementary Fig. 9(b)–(d)]. Finally, decreased expression and nuclear localization of β-catenin was observed in articular chondrocytes in SM04690-treated rats as compared to vehicle-treated rats [Supplementary Fig. 10(a)–(c)].
Discussion
Wnt signaling has been shown to be a pivotal pathway in OA that modulates bone and chondrocyte lineage specification, protease production and joint homeostasis. SM04690 is being developed as a novel small-molecule inhibitor of the Wnt pathway. In this study, we used several in vitro assays for Wnt signaling, chondrogenesis and cartilage catabolism to test the effects of SM04690 on these processes. In vitro, SM04690 was ∼50- to 500-fold more potent than published Wnt inhibitors across multiple cellular assays. Functional inhibition of Wnt signaling in cartilage in the rat ACLT + pMMx model demonstrated by both gene expression and inhibited β-catenin nuclear localization, provided evidence for the ability of SM04690 to modulate the Wnt pathway in vivo, in the context of OA.
The appearance of Rhodamine B staining with SM04690-treated hMSCs and mouse stem cells in vitro indicated significant and dose-dependent induction of early chondrocyte condensation, followed by differentiation into mature chondrocytes that expressed both Type II collagen and Aggrecan that are required for the specialized extracellular matrix rich in highly sulfated GAGs necessary to provide elastic biomechanical properties for motion and weight bearing
. In the ACLT + pMMx model, SM04690 treatment led to increased cartilage as demonstrated by Safranin O staining, GAG measurement and serum biomarker PIIANP. An increased number of Dcx positive articular chondrocytes as well as increased expression of chondrogenic genes in the cartilage, as compared to vehicle controls, were consistent with induction of chondrocyte differentiation in SM04690-treated rats. These data demonstrated the ability of SM04690 to potentially regenerate functional chondrocytes and hence restore cartilage in vivo.
Enzymatic tissue degradation plays a significant role in the progression of OA and expression of proteases is increased in articular cartilage in response to joint injury
Characterization of articular cartilage and subchondral bone changes in the rat anterior cruciate ligament transection and meniscectomized models of osteoarthritis.
. SM04690 inhibited the expression of catabolic enzymes in cytokine stimulated chondrocytes and hMSCs in vitro, as well as in the ACLT + pMMx animals, compared to vehicle controls, thus demonstrating specific control and regulation of catabolic enzymes in the context of OA. SM04690 also decreased cytokine-induced GAG breakdown and NO production in vitro, and levels of circulating COMP in vivo, together providing strong evidence that SM04690 prevented matrix breakdown by proteases under the pathological conditions of OA. While additional studies are needed to evaluate SM04690 effects on protease production and matrix catabolism in primary OA patient-derived chondrocytes, these data suggested that SM04690 may have potential cartilage protective effects.
Changes in gene expression, biomarkers and restored cartilage measured in the rat ACLT + pMMx model experiments by histological OARSI scores indicated improvements in joint morphology and attenuated OA phenotype with SM04690 treatment, despite the ongoing absence of joint ligaments responsible for inducing the model. The observed efficacy may be a result of a combination of mechanisms, including MSC differentiation into chondrocytes, and cartilage protective effects of SM04690, however, the relative contribution of each is unknown. In vivo pharmacokinetic data supported the use of a single SM04690 IA injection for this study and the demonstrated efficacy of the compound in human cells in vitro supported the rationale to pursue clinical development of SM04690. In a phase 1 first-in-human clinical trial (N = 61), SM04690 appeared safe and well-tolerated
A novel Wnt pathway inhibitor, SM04690, for the treatment of moderate to severe osteoarthritis of the knee: results of a 24-week, randomized, controlled, phase 1 study.
In the search for an effective treatment for OA, development of DMOADs has focused on either slowing down the progression of the disease by targeting either matrix-degrading enzymes (using matrix metalloproteinases (MMP)/a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) inhibitors)
. Expression of factors such as fibroblast growth factor (FGF)-18 and TGFβ to promote chondrocyte proliferation and maturation, or stem cell-based transplants have been tested, but challenges remain with translation of these techniques into the clinic
. The potential to stimulate cartilage repair via chondrocyte differentiation and reduce cartilage destroying proteases, could lead to a beneficial treatment effect with SM04690. These data support the further development of SM04690 as an injectable small-molecule potential DMOAD for the treatment of knee OA.
Author contributions
Conception and design: JH, CBa, SK, HH, VD, LD, JS, NL and YY.
Acquisition, analysis and interpretation of data: VD, HH, CBa, SK, LD, JS, KC, CBo, MI, MP, TS, JC, LD, SC, BT, JT, JH, NL and YY.
Drafting the article: VD, JH, NL and YY.
Final approval of the article: VD, HH, CBa, SK, LD, JS, KC, CBo, MI, MP, TS, JC, LD, SC, BT, JT, JH, NL and YY.
Financial support for this study and its publication was provided by Samumed LLC.
Acknowledgments
We thank Sarah Kennedy, PhD, David Herman, PhD, Timothy Phalen, PhD, Benoit Melchior, PhD and Hutch Humphreys for critical revision of the manuscript and Christopher Swearingen, PhD for assistance with statistical analysis.
Appendix A. Supplementary data
The following are the supplementary data related to this article:
Supplementary Fig. 1SM04690 was a potent and specific inhibitor of canonical Wnt signaling. (a) Chemical structure of SM04690. (b,c) SW480 cells and hMSCs treated with SM04690 or various Wnt pathway inhibitors or DMSO for 48 h. (a) Dose response of TCF/LEF promoter driven-luciferase reporter in SW480 cells (n = 4, Mean ± 95% CI). (b) EC50 values for Wnt inhibitors in (b). (d,e) Gene expression of Wnt pathway genes in (d) hMSCs and (e) IEC6 cells, stimulated either with Wnt3a (200 ng/ml) or CHIR-99021 (4 μM) and treated with SM04690 (30 nM) or CX-4945 (3 μM) or KY02111 (10 μM) for 16 h. Fold change relative to unstimulated control (n = 4, Mean ± 95% CI, *P < 0.05, ***P < 0.01, ***P < 0.001, one-way ANOVA).
Supplementary Fig. 2SM04690 was a potent and specific inhibitor of canonical Wnt signaling. (a, b) Expression of genes in the non-canonical Wnt pathway in hMSCs following treatment with DMSO or SM04690 (30 nM) or CX-4945 (10 μM) or KY02111 (10 μM) for 48 h in the (a) absence or (b) presence of Wnt5a (100 ng/ml), measured by qRT-PCR. Expression represented as ddCt relative to DMSO (n = 4, Mean ± 95% CI, *P < 0.05, **P < 0.01, one-way ANOVA).
Supplementary Fig. 3SM04690 was a potent and specific inhibitor of canonical Wnt signaling. Expression of genes in the BMP pathway in hMSCs following treatment with DMSO or SM04690 (30 nM, 100 nM) for 6, 40 or 72 h measured by qRT-PCR. Expression represented as ddCt relative to DMSO (n = 4, Mean ± 95% CI, *P < 0.05, **P < 0.01, ***P < 0.001, t test vs DMSO at each timepoint).
Supplementary Fig. 4SM04690 induced chondrocyte differentiation from hMSCs. (a) Dose response quantification of the number of Rhodamine B stained chondrogenic colonies from hMSCs treated with SM04690 or DMSO or TGFβ3 for 5 days (n = 4, Mean ± 95% CI, *P = 0.013, ***P < 0.0001, one-way ANOVA). (b) Expression of Sox9 in hMSCs following treatment with various doses of SM04690 or DMSO for 3 days as measured by Western blot. Tubulin binding protein (TBP) serves as a loading control.
Supplementary Fig. 6SM04690 induced chondrocyte differentiation from hMSCs. (a) Dose response quantification of the number of Rhodamine B stained chondrogenic colonies from hMSCs treated with Wnt pathway inhibitors (n = 3, Mean ± 95% CI). (b) EC50 values for compounds in (a). (c) Dose response quantification of Alcian blue stained chondrocytes from hMSCs treated with DMSO or SM04690 or TGFβ3 or Wnt pathway inhibitors for 21 days (n = 6, Mean ± 95% CI). (d) EC50 values for compounds in (c). (e) Representative images of chondrocytes from (c). (f) Dose response quantification of Safranin O stained chondrocytes from hMSCs treated with DMSO or SM04690 or TGFβ3 or Wnt pathway inhibitors for 21 days (n = 6, Mean ± 95% CI). (g) EC50 values for compounds in (f). (h) Representative images of chondrocytes from (f).
Supplementary Fig. 7SM04690 protected hMSCs and chondrocytes from catabolic breakdown in vitro. (a, b) Gene expression of proteases (MMP1, MMP3, MMP13, IHH) in hMSCs treated with either (a) TNFα (20 ng/ml) + Oncostatin M (10 ng/ml) or (b) IL1β (10 ng/ml) and SM04690 (30 nM) or DMSO for 72 h measured by qRT-PCR. Fold change relative to unstimulated control (n = 3, Mean ± 95% CI, TNFα + Oncostatin M: MMP1 ***P = 0.0004, MMP3 P = 0.148, MMP13 **P = 0.01, IHH **P = 0.0075; IL1β: MMP1 P = 0.077, MMP3 P = 0.141, MMP13 P = 0.158, IHH *P = 0.014, t test). (c, d) Gene expression of proteases (MMP1, MMP3, MMP13) in human chondrocytes treated with either (c) TNFα (20 ng/ml) + Oncostatin M (10 ng/ml) or (d) IL1β (10 ng/ml) and SM04690 (30 nM) or CX-4945 (10 μM) or KY02111 (10 μM) or DMSO for 72 h as measured by qRT-PCR. Fold change relative to unstimulated control (n = 3, Mean ± 95% CI, *P < 0.05, **P < 0.01, ***P < 0.001, one-way ANOVA).
Supplementary Fig. 8SM04690 promoted cartilage repair and protection in the ACLT OA model in rats. (a) Pharmacokinetics of SM04690 in rat cartilage, bone and plasma following a single IA injection of SM04690 (0.3 μg). (b) The medial joint score in the ACLT + pMMx model, based on the OARSI scoring system (n = 12 rats, Mean ± 95% CI, ***P = 0.0005, Mann–Whitney U test). (c) Cartilage thickness measurement and (d) Safranin O staining intensity of knee sections from the ACLT + pMMx model (n = 12 rats, Mean ± 95% CI, **P = 0.0042, ***P < 0.0001, respectively, t test).
Supplementary Fig. 9SM04690 inhibited Wnt signaling in the ACLT + pMMx OA model in rats. ACLT + pMMx rats treated with either vehicle or single IA injection of SM04690 (0.3 μg) and cartilage analyzed 4 weeks from treatment for gene expression by qRT-PCR. (a) Col10a1 expression. (b) Expression of Wnt pathway genes, represented on a color scale as shown. Expression of Wnt pathway genes (c) Axin2 and (d) β-catenin (n = 7 for vehicle, n = 8 for treatment, Mean ± 95% CI, Col10a P = 0.48, Axin2 *P = 0.017, β-catenin *P = 0.04, t test).
Supplementary Fig. 10SM04690 inhibited Wnt signaling in the ACLT + pMMx OA model in rats. ACLT + pMMx rats treated with either vehicle or SM04690 (0.3 μg) and analyzed by immunohistochemistry (IHC) 12 weeks from treatment. (a) Representative images of the superficial zone of articular cartilage stained for β-catenin. (b, c) Quantification of staining intensity of β-catenin in (b) the total cell and in (c) the nucleus of the cells, from images in (a) (n = 6 rats/group, Mean ± 95% CI, **P = 0.0091 and *P = 0.044, respectively, t test).
Characterization of articular cartilage and subchondral bone changes in the rat anterior cruciate ligament transection and meniscectomized models of osteoarthritis.
A novel Wnt pathway inhibitor, SM04690, for the treatment of moderate to severe osteoarthritis of the knee: results of a 24-week, randomized, controlled, phase 1 study.
31167-6&id=gr1.jpg){kind=link}
31167-6&id=gr2.jpg){kind=link}
31167-6&id=gr3.jpg){kind=link}
31167-6&id=gr4.jpg){kind=link}
31167-6&id=gr5.jpg){kind=link}
31167-6&id=figs1.jpg){kind=link}
31167-6&id=figs2.jpg){kind=link}
31167-6&id=figs3.jpg){kind=link}
31167-6&id=figs4.jpg){kind=link}
31167-6&id=figs5.jpg){kind=link}
31167-6&id=figs6.jpg){kind=link}
31167-6&id=figs7.jpg){kind=link}
31167-6&id=figs8.jpg){kind=link}
31167-6&id=figs9.jpg){kind=link}
31167-6&id=figs10.jpg){kind=link}