Fig 1.
(A) Protein was collected from the articular cartilage of hind limb joints of 18-month-old mice. DNIIR protein levels were determined by Western blot. An antibody specific to the extracellular domain of TGFβR2 was used. A band corresponding to the truncated protein appeared at 34 kDa. Wild-type mice exhibited no DNIIR protein, while DNIIRTg and DNIIRTg;Prg4Tg mice exhibited a band representing the truncated TGFβR2 receptor. (B-G) Prg4 protein was identified via immunohistochemistry in 2-month-old mice. The articular cartilage of wild-type mice (B, E) exhibited Prg4 staining in the superficial zone and faint Prg4 staining in the middle zone. The articular cartilage of DNIIRTg mice (C, F) exhibited almost no Prg4 staining. The articular cartilage of DNIIRTg;Prg4Tg mice (D, G) exhibited staining in the superficial zone and faint staining in the middle zone, similar to the staining profile of the wild-type cartilage. Representative images are shown from wild-type mice (n = 3), DNIIRTg mice (n = 3), and DNIIRTg;Prg4Tg mice (n = 2).
Fig 2.
Prg4 prevents foot misplacement in DNIIRTg mice.
(A) Mice at 18 months of age were placed in an apparatus that recorded the number of missed steps while walking. The following results are reported as “mean (lower bound of the confidence interval, upper bound of the confidence interval).” (B) Wild-type mice missed 0.4 (0.0, 0.8) steps (n = 8; 4 female, 4 male), DNIIRTg mice missed 1.7 (0.6, 2.7) steps (n = 9; 4 female, 5 male), and DNIIRTg;Prg4Tg mice missed 0.4 (-0.1, 0.8) steps (n = 8; 5 female, 3 male). There were the following statistically significant differences (reported as “mean difference (lower bound of confidence interval, upper bound of confidence interval)”): wild-type and DNIIRTg mice (-1.3 (-2.1, -0.4)) and DNIIRTg and DNIIRTg;Prg4Tg mice (1.3 (0.4, 2.1)).
Fig 3.
Prg4 prevents fibrillation of articular cartilage in DNIIRTg mice.
μCT imaging was used to obtain 3D images of the surface of femoral articular cartilage in 18-month-old mice. (A-B, D-E) Wild-type mice exhibited mostly smooth cartilage, while DNIIRTg mice exhibited more fibrillation, especially in load-bearing regions. (C and F) DNIIRTg;Prg4Tg mice exhibited smoother cartilage than DNIIRTg mice. (D and F) Wild-type cartilage and DNIIRTg;Prg4Tg cartilage appeared similarly smooth. The medial condyle is shown up-close in D-F. To quantify the fibrillation on the medial condyle, we developed a scoring system. (G) A score of 0 corresponds to no fibrillation or very minimal fibrillation; (H) a score of 1 corresponds to fibrillation that partially covers the medial condyle; (I) a score of 2 corresponds to fibrillation that fully covers the medial condyle and consists of mostly shallow fissures; (J) and a score of 3 corresponds to fibrillation that fully covers the medial condyle and consists of mostly deep fissures. Images were scored by an observer blinded to genotype. The following results are reported as “mean (lower bound of the confidence interval, upper bound of the confidence interval).” (K) The fibrillation score was 0.3 (-0.5, 1.2) for wild-type mice (n = 6; 2 female, 4 male), 2.0 (-0.5, 4.5) for DNIIRTg mice (n = 3; 1 female, 2 male), and 0.3 (-0.5, 1.0) for DNIIRTg;Prg4Tg mice (n = 4; all 4 male). There were the following statistically significant differences (reported as “mean difference (lower bound of confidence interval, upper bound of confidence interval)”): wild-type and DNIIRTg mice (-1.7 (-2.9, -0.4)) and DNIIRTg and DNIIRTg;Prg4Tg mice (1.8 (0.4, 3.1)).
Fig 4.
Prg4 prevents thinning of articular cartilage in DNIIRTg mice.
H&E staining was used to visualize general histological characteristics of the articular cartilage in 18-month-old mice. The double arrows mark the thickness of the articular cartilage. (A-B, D-E) Wild-type tibial cartilage was thicker than DNIIRTg tibial cartilage. (C and F) DNIIRTg;Prg4Tg tibial cartilage was thicker than DNIIRTg tibial cartilage and similar in thickness to wild-type tibial cartilage. DNIIRTg cartilage exhibited fibrillation, whereas wild-type and DNIIRTg;Prg4Tg cartilage exhibited little, if any, fibrillation (n = 3 for each group). A needle indenter was used to confirm cartilage thickness. The following results are reported as “mean (lower bound of the confidence interval, upper bound of the confidence interval).” (G) The thickness was 92.3 (28.5, 156.2) μm for wild-type mice (n = 3; 1 female, 2 male), 55.8 (35.6, 76.0) μm for DNIIRTg mice (n = 5; 2 female, 3 male), and 84.0 (52.9, 115.1) μm for DNIIRTg;Prg4Tg mice (n = 3; all 3 male). There were the following statistically significant differences (reported as “mean difference (lower bound of confidence interval, upper bound of confidence interval)”): wild-type and DNIIRTg mice (36.5 (8.8, 64.2)) and DNIIRTg and DNIIRTg;Prg4Tg mice (-28.2 (-55.9, -0.5)).
Fig 5.
Prg4 prevents loss of proteoglycans and loss of superficial zone cartilage.
Safranin O was used to stain for proteoglycans in the articular cartilage of 18-month-old mice. Also, changes in the superficial zone morphology were observed between groups. (A-B, D-E) Wild-type cartilage exhibited a normal superficial zone with cells aligned parallel to the articular surface (arrows point to aligned cells), whereas DNIIRTg cartilage exhibited large regions with no superficial zone as evidenced by a lack of cells aligned parallel to the articular surface. (C and F) DNIIRTg;Prg4Tg cartilage exhibited a normal superficial zone with cells aligned parallel to the articular surface, similar to the wild-type cartilage. OARSI scoring was performed by an observer blinded to genotype. The following results are reported as “mean (lower bound of the confidence interval, upper bound of the confidence interval).” (G) The OARSI score was 6.7 (-8.5, 21.9) for wild-type mice (n = 3), 16.0 (16.0, 16.0) for DNIIRTg mice (n = 5), and 9.2 (1.4, 17.0) for DNIIRTg;Prg4Tg mice (n = 5). There were the following statistically significant differences (reported as “mean difference” (lower bound of confidence interval, upper bound of confidence interval)): wild-type and DNIIRTg mice (-9.3 (-18.2, -0.5)) and DNIIRTg and DNIIRTg;Prg4Tg mice (6.8 (-0.9, 14.5)).
Fig 6.
Prg4 does not affect pSmad3 protein levels.
pSmad3 and total Smad2/3 protein levels were determined by Western blot. DNIIRTg and DNIIRTg;Prg4Tg cartilage both exhibited reduced pSmad3 protein levels compared to wild-type cartilage.