Skeletal Characterization of Smurf2-Deficient Mice and In Vitro Analysis of Smurf2-Deficient Chondrocytes

Abstract

Overexpression of Smad ubiquitin regulatory factor 2 (Smurf2) in chondrocytes was reported to cause spontaneous osteoarthritis (OA) in mice. However, it is unclear whether Smurf2 is involved in bone and cartilage homeostasis and if it is required for OA pathogenesis. Here we characterized age-related changes in the bone and articular cartilage of Smurf2-deficient (MT) mice by microCT and histology, and examined whether reduced Smurf2 expression affected the severity of OA upon surgical destabilization of the medial meniscus (DMM). Using immature articular chondrocytes (iMAC) from MT and wild-type (WT) mice, we also examined how Smurf2 deficiency affects chondrogenic and catabolic gene expressions and Smurf2 and Smurf1 proteins upon TGF-β3 or IL-1β treatment in culture. We found no differences in cortical, subchondral and trabecular bone between WT and MT in young (4 months) and old mice (16-24 months). The articular cartilage and age-related alterations between WT and MT were also similar. However, 2 months following DMM, young MT showed milder OA compared to WT (~70% vs ~30% normal or exhibiting only mild OA cartilage phenotype). The majority of the older WT and MT mice developed moderate/severe OA 2 months after DMM, but a higher subset of aged MT cartilage (27% vs. 9% WT) remained largely normal. Chondrogenic gene expression (Sox9, Col2, Acan) trended higher in MT iMACs than WT with/without TGF-β3 treatment. IL-1β treatment suppressed chondrgenic gene expression, but Sox9 expression in MT remained significantly higher than WT. Smurf2 protein in WT iMACs increased upon TGF-β3 treatment and decreased upon IL-1β treatment in a dose-dependent manner. Smurf1 protein elevated more in MT than WT upon TGF-β3 treatment, suggesting a potential, but very mild compensatory effect. Overall, our data support a role of Smurf2 in regulating OA development but suggest that inhibiting Smurf2 alone may not be sufficient to prevent or consistently mitigate post-traumatic OA across a broad age range.

Fig 1. Summary of experimental design for skeletal characterizations.

Fig 2. Smurf2 protein and gene expressions in WT (+/+) and Smurf2-deficient MT (T/T) skeletal tissues and primary cells.

(A) Protein expression of Smurf2 in various skeletal tissues from healthy 4 month old male WT and MT mice. Spleen, where Smurf2 is highly expressed, is included as a positive control. (B) Smurf2 mRNA expression in various skeletal tissues compared to spleen in 4 month old male WT and MT mice. (C) Protein expression of Smurf2 in bone marrow stromal cells (BMSC, passage 0) isolated from 4 month old WT and Smurf2-deficient MT mice and immature articular chondrocytes (iMAC, passage 0) isolated from WT and MT neonates.

Fig 3. Age- and gender-specific cortical bone and trabecular bone analyses of WT and Smurf2-deficient MT mice.

(A) Mid-shaft cortical bone analysis of male WT and MT femurs from 4 month (WT: n = 6; MT: n = 7) and 18 month old (WT and MT: n = 7) mice. (B) Gender-specific trabecular bone analysis of lumbar vertebrae from 16 month old WT and MT mice (Male: n = 7; Female: n = 5). (C) Representative 3D reconstruction of contoured vertebral trabecular bone. Scale bar = 500 μm.

Fig 4. Age-specific knee joint phenotypes of male WT (+/+) and Smurf2-deficient MT (T/T) mice.

(A) Normal and osteoarthritic joint articular cartilage histology scoring criteria modified over literature method[27] along with representative safranin O-stained cartilage sections. Yellow line indicates tidemark; Black arrows denote loss of staining, fibrillations, or erosions. (B) Combined histology scores of femoral and tibial articular cartilage of 4 month old WT (n = 13), 4 month old MT (n = 13), 21 month WT (n = 11), and 21 month old MT (n = 11). (C) Representative microCT images of bone mineral density color mappings of mid-frontal knee sections from young and old WT and Smurf2 MT mouse knees. Red indicates higher BMD while green indicates lower BMD. D) Quantitative comparisons of the lateral and medial subchondral bone analyses between 4 month (WT: n = 9; MT: n = 11) and 21 month (WT: n = 10; MT: n = 8) WT and MT mice. *p<0.05.

Fig 5. Differential severity of knee joint articular cartilage erosions in young (4 month) and old (21 ± 1.3 month) male WT and Smurf2-deficient MT mice after DMM surgery.

(A) Representative images of safranin O-stained articular cartilage sections from the medial compartment of WT and MT knees 2 months post-DMM surgery. (B) Semi-quantitative histology scores of the femoral and tibial articular cartilage of DMM knees vs un-operated controls (n = 13 for 4 month; n = 11 for 21 ± 1.3 month). The average femoral and tibial articular cartilage scores for each joint were plotted separately on the same graph. Groups with the same symbol are not statistically significant (p > 0.05) based on mean ranks. (C) Distribution of DMM knee scores for WT and MT based on OA severity.

Fig 6. Quantitative gene expression analyses of key anabolic and catabolic markers in WT and MT iMACs.

Cells were cultured as a monolayer on 2D tissue culture polystyrene for 4 days and treated with either 10 ng/mL of TGF-β3 or 5 ng/mL of IL-1β for 24 hours. Data are reported as the mean ± standard deviation of three separate chondrocyte isolations for each genotype. EM = Expansion Media, *p < 0.05 (p values approaching significance are noted on the graph).

Fig 7. Smurfs protein expression changes upon 24-h treatment of TGF-β3 and IL-1β.

A) WT chondrocytes show dose-dependent increase of Smurf2 protein with 24-h TGF-β3 treatment and dose-dependent decrease with 24-h IL-1β treatment. B) No compensatory increase of Smurf1 protein levels was detected between WT (+/+) and Smurf2-deficient MT chondrocytes (T/T) with and without treatment of TGF-β3 (10 ng/mL) or IL-1β (0.1 ng/mL). Type 2 collagen levels were also similar between WT and MT chondrocytes across all conditions.