Fig 2: Diabetes-induced RUNX2 loss is mediated through the mTORC1/p-70S6K/GSK3β axis.Murine MSCs (C3H10T1/2) were subjected to adipogenic differentiation in the presence of either low (5.5 mM) or high (25 mM) levels of glucose in the medium and subjected to A, B Western blot analysis as indicated. C Quantification of alizarin red-S staining by spectrophotometry and D Alizarin red-S-stained images of murine MSCs (C3H10T1/2), which were induced to adipogenic differentiation with either low (5.5 mM) or high (25 mM) levels of glucose and with or without LiCl (0.5 mM) treatment, as indicated. E Oil red O-stained murine MSCs (C3H10T1/2), which were subjected to adipogenic differentiation in the presence of either low (5.5 mM) or high (25 mM) levels of glucose and with or without LiCl (0.5 mM) treatment, as indicated, and F quantification of oil red O staining by spectrophotometry. G Bone marrow-derived MSCs (BM-MSCs) were subjected to adipogenic differentiation in the presence of either low (5.5 mM) or high (25 mM) levels of glucose, and cell lysates were collected and subjected to Western blot analysis as indicated. Murine MSCs (C3H10T1/2) treated with LiCl (0.5 mM) with or without treatment with high glucose (25 mM) and H Western blot analysis, as indicated. Mean ± S.E.M.; N = 3, *p < 0.1 versus control, **p < 0.01 versus control, NSp > 0.1 versus control. AM adipogenic medium, LG low glucose, HG high glucose, LiCl lithium chloride, BM-MSCs bone marrow-derived mesenchymal stromal cells freshly isolated from mice, NS not significant.

Fig 3: In silico prediction of AMPK phosphorylation site in RUNX2 protein.a List of corresponding putative motif-specific kinases along with AMPK predicted by phospho motif finder in the N-terminal region (DNA-binding domain) of RUNX2. b The predicted corresponding putative AMPK substrate motif of RUNX2 significantly aligned with all the important residues of known and validated AMPK substrate motifs. c The ClustalW2 sequence alignment showing cross-species conservation with predicted optimal AMPK substrate motif of RUNX2

Fig 4: Interaction of GAPDH and L1CAM. The single and double asterisks indicate p < 0.05 and 0.01, respectively. (A) Expression levels of Hsp90ab1 and GAPDH after their silencing or overexpression in C28/I2 chondrocytes. (B) Upregulation of GAPDH in response to recombinant Hsp90ab1proteins, and downregulation of Hsp90ab1 by silencing GAPDH in C28/I2 chondrocytes. (C) Elevation of p-p38 and TNFα by silencing Hsp90ab1 in C28/I2 chondrocytes, while their reduction in response to recombinant Hsp90ab1 protein. Elevation of p-p38 and TNFα by silencing GAPDH in C28/I2 chondrocytes, while their reduction in GAPDH-overexpressing C28/I2 chondrocytes. (D,E) Reduction in MTT-based viability and transwell invasion by silencing GAPDH in SW1353 cells and the opposite effect in GAPDH overexpressed SW1353 cells (n = 6). (F) Downregulation of Snail by silencing GAPDH in SW1353 cells and elevation of Snail in GAPDH overexpressed SW1353 cells. (G) Co-immunoprecipitated L1CAM with GAPDH. (H,I) Suppression of GAPDH-mediated inhibition of MTT-based viability and EdU-based proliferation of SW1353 cells by RNA silencing of L1CAM (n = 6). (J) Downregulation of Runx2 and Snail in response to recombinant GAPDH in SW1353 cells, and its suppression by silencing L1CAM. Full Western blot images are available in Figure S8.

Fig 5: SIRT7 regulates myeloid cell calcification in high glucose. (a) Effects of incubating the myeloid cell line THP-1 cells with osteogenic medium on calcification assessed by von Kossa staining. (b) Calcification, as assessed by von Kossa staining, by THP-1 cells grown in osteogenic medium and transfected with siRNA against SIRT7 or with a SIRT7 plasmid to obtain overexpression. (c) Quantification of von Kossa-stained area (left) and quantitative measure of calcium content (right) in the four culture conditions of THP-1 cells shown in panels (a) and (b). (d) Gene expression of RUNX2 and OSX/SP7 in the four conditions shown in (a) and (b). (e–g) Effects of exposing THP-1 cells to high (20 mmol/l) glucose (Glu) compared with normal glucose (5 mmol/l) and high mannitol (Mann) on gene expression of SIRT7 (e), and pro-calcific genes RUNX2 and OSX/SP7 (f). (g) Representative microphotographs of THP-1 cells grown in osteogenic medium and exposed to normal (5 mmol/l) or high (20 mmol/l) glucose with or without overexpression of SIRT7 (over-SIRT7). (h) Quantification of calcification in the three conditions shown in (g) using von Kossa staining (left) and quantitative measure of calcium content (right). Histograms show means with SEM bars and superimposed circles indicate individual data points. For the indicated comparisons, statistical significance was as follows: *p<0.05; **p<0.01; ***p<0.001. Ctrl, control