Fig 2: MiR-146a targets Tgif1, Camk2d and Ppp3r2. (a) Total RNA of mouse articular chondrocytes transfected with miR-146a inhibitor or control (NC) inhibitor was used for microarray analysis. Twelve genes were validated based on the seed region of miR-146a in 3'-UTR of candidate gene that is evolutionarily conserved in mammals. The gradient colors and numbers indicate fold change of gene expression in mouse chondrocytes transfected with the miR-146a inhibitor compared with control, as determined by microarray. Fold 146a/NC, fold change of gene expression in mouse chondrocytes transfected with miR-146a inhibitor compared with control inhibitor. (b) qPCR analysis of mRNA expression of candidate target genes in mouse articular chondrocytes transfected with miR-146a mimic or control mimic plasmid (n=3 for each treatment). (c and d) Luciferase activity analysis in HEK293T cells that were co-transfected with the indicated 3'-UTR reporter (c) or with the indicated WT or mutant 3'-UTR reporter (d) along with miR-146a mimic or control mimic plasmid (n=3 for each treatment). The results were normalized by internal firefly luciferase readout. (e-g) qPCR analysis of mRNA expression of Tgif1 (e), Camk2d (f) and Ppp3r2 (g) in articular cartilage of mice (WT, n=6; miR-146a KO, n=8) collected at 4 weeks after DMM surgery. (h and i) Western blot analysis of endogenous Sox9 and Col2a1 protein expression in mouse chondrocytes that were transfected with indicated expression vector constructs (h) or with gene-specific siRNAs as well as a scrambled control siRNA (siRNA-NC). For quantification, protein expression was normalized by the protein amount in the first lane. (i) Three independent experiments were performed with similar results in b-d. Relative mRNA expression was normalized to the expression of GAPDH. Data are mean±S.D. *P<0.05, **P<0.01 versus control group (Student's t-test) in b, c and d and Mann–Whitney test in e-g. NS, not significant between the indicated groups

Fig 3: Verification of CAMK2D (CA) and CNN3 (CN) but Not CACNB4 and CACNG7 as Cognate Targets of miR-7-5p(A) TargetScan-predicted binding sites of targets with miR-7-5p. (B) Effects of miR-7-5p on the expression levels of CACNB4, CACNG7, CAMK2D, and CNN3. *p < 0.05, **p < 0.01 versus Nor + NC; #p < 0.05, ##p < 0.01 versus Hyp + NC. (C) Diagrammatic sketch of the binding sites for the WT and MUT binding sites of CAMK2D and CNN3 associated with miR-7-5p. (D) Luciferase reporter assay for the luciferase activity of HPASMCs cotransfected with scrambled miRNA (NC) and VE/CAMK2D and CNN3 3' UTR-WT/CAMK2D and CNN3 3' UTR-MUT/complementary sequences of miR-7-5p (miR-7-5p-PC) or miR-7-5p mimics with VE/CAMK2D and CNN3 3' UTR-WT/CAMK2D and CNN3 3' UTR-MUT/complementary sequences of miR-7-5p (miR-7-5p-PC). **p < 0.01 compared with NC + WT; ##p < 0.01 compared with NC + miR-7-5p-PC. (E and F) HPASMCs were labeled with antibodies against CNN3 (E) and CAMK2D (F), and nuclei were stained with DAPI and merged to represent an overlay figure. Scale bars, 25 µm. All tests were performed at least three times, and the values are presented as the mean ± SEM.

Fig 4: CDR1as/miR-7-5p Aggravates HPASMC Calcification via CNN3 and CAMK2D(A) HPASMCs were incubated with CNN3- or CAMK2D-overexpressing plasmids with or without siCDR1as or miR-7-5p agomir, and calcium mineral deposits were assessed. (B) Transfection of siCDR1as or miR-7-5p agomir decreased the Runx2 level in Hyp HPASMCs, whereas the opposite phenomenon was observed after CNN3 or CAMK2D overexpression. Scale bars, 50 µm. All tests were performed at least three times.

Fig 5: CNN3 and CAMK2D Are Targets of CDR1as, which Are Required for Osteogenic Transition(A) The CNN3 and CAMK2D protein levels in the Hyp with the siCDR1as group were evidently downregulated compared with those in the Hyp with NC group. (B and C) siCDR1as reduced the level of CNN3 (B) and CAMK2D (C) in HPASMC cells induced by Hyp, and this effect was reversed by cotransfecting cells with antagomir. Scale bars, 25 µm. (D) Verification of the effect of knockdown and overexpression on CNN3 and CAMK2D. We used siCAMK2D-3 and siCNN3-1 to knock down CAMK2D and CNN3 in the subsequent experiments. (E) Effects of CNN3 and CAMK2D siRNA and overexpression plasmid transfection on the expression levels of SM22a and BMP2. (F) ALP activity was assayed in siNC-, siCNN3-, siCAMK2D-, VE-, CNN3-, and CAMK2D-overexpressing plasmid-transfected HPASMCs. (G) Quantification of the calcium content in Hyp-induced HPASMCs with or without silencing or overexpression of CNN3 and CAMK2D. *p < 0.05, **p < 0.01 compared with Nor + NC; #p < 0.05, ##p < 0.01 compared with Hyp + NC; &p < 0.05, &&p < 0.01 compared with Nor + VE; $p < 0.05, $$p < 0.01 compared with Hyp + VE. All tests were performed at least three times, and the values are presented as the mean ± SEM.