Fig 2: AMPK activation by metformin increases the biogenesis of let-7, which requires KSRP. (a, b) ARK2 cells were treated with AICAR (final concentration 0.5 mm), followed by RNA extraction and reverse transcriptase (RT)–quantitative PCR (qPCR) to determine RNA levels of let-7 (a) and H19 (b). (c) ARK2 cells were treated with or without metformin or AICAR, followed by western blot analysis for phosphorylated AMPK (p-AMPK) and total AMPK (AMPK) using β-actin (ACTB) as a loading control. Representative western blot gels of three independent experiments are shown. (d, e) ARK2 cells were transfected with siCon or Ksrp-specific siRNA (siKsrp), followed by RT–qPCR (d) and western blot analysis (e) to confirm KSRP knockdown at the mRNA and protein levels, respectively. (f, g) ARK2 cells were transfected with siCon or siKsrp, followed by treatment with or without metformin 24 h later. RNA levels of let-7 (f) and H19 (g) were determined by RT–qPCR 24 h post-transfection. Numbers are mean±s.d. (n=3). **P<0.01. *P<0.05. NS, no statistical difference.

Fig 3: Reduced H3K18LA/KSRP reduces the transcriptional expression of FSTL1. (A) qPCR analysis of glycolysis‐related gene mRNA in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (B) Fluorescent glucose uptake in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (C, D) ECAR analysis of osteoblasts from Usp26 cKO mice and their littermate controls. n = 5 in each group. (E) Representative serial micro‐PET imaging of 18F‐FDG uptake in the femur (tibia) of 24‐week‐old Usp26 cKO mice and their littermate controls. n = 5 in each group. Scale bars: 5mm. (F) ATP concentration in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (G) Lactate concentration in osteoblasts of Usp26 cKO mice and their littermate controls. n = 5 in each group. (H) Western blot and quantitative analysis of PAN‐KLA protein expression in osteoblasts of Usp26 cKO mice and their littermate controls.n = 3 in each group. (I)Western blot and quantitative analysis of H3K9LA, H3K14LA, H3K18LA, H4K8LA, and H4K12LA protein expression in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (J) Western blot and quantitative analysis of H3K18LA protein expression in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (K) qPCR analysis of Ksrp mRNA in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (L) Western blot and quantitative analysis of KSRP protein expression in osteoblasts of Usp26 cKO mice and their littermate controls. n = 3 in each group. (M, N) KSRP Immunofluorescence staining and quantitative analysis in femurs of Usp26 cKO mice and their littermate controls. n = 5 in each group. Scale bars: 25µm. (O, P) Western blot and quantitative analysis of PAN‐KLA and H3K18LA protein expression in osteoblasts of Usp26 cKO mice and their littermate controls treated with PBS or lactate. n = 3 in each group. (Q, R) Western blot and quantitative analysis of KSRP and FSTL1 protein expression in osteoblasts of Usp26 cKO mice and their littermate controls treated with PBS or lactate. (S) chip‐qPCR. (T) chip‐seq. Data are presented as mean ± SEM. Statistical significance was analyzed using two‐tailed Student's t‐test in (A), (B), (D), (F), (G), (H), (I), (J), (K), (L) and (N), one‐way ANOVA in (P) and (R), and two‐way ANOVA in (E) and (T). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig 4: Engineered EVs alleviate HFD‐induced IR and fibrosis in mice. (A) Western blot analysis and quantification of USP26 protein expression in osteoblasts treated with PBS or engineered EVs. n = 3 in each group. (B) Western blot analysis of PAN‐KLA and H3K18LA protein expression in osteoblasts from Usp26 cKO mice and their littermate controls treated with PBS or engineered EVs. n = 3 in each group. (C) Western blot analysis of USP26, KSRP, and FSTL1 protein expression in osteoblasts from Usp26 cKO mice and their littermate controls treated with PBS or engineered EVs. n = 3 in each group. (D) Representative micro‐CT images of femurs from WT, HFD‐induced mice, and HFD‐induced mice with EVs injection via the tail vein. Quantitative analysis of trabecular bone parameters: BV/TV, Tb. N, Tb. Th, Tb. Sp, and BMD. n = 5 in each group. (E, F) Glucose tolerance test (GTT) and area under the curve (AUC) in WT, HFD‐induced mice, and HFD‐induced mice with EVs injection. n = 5 in each group. (G, H) Insulin tolerance test (ITT) and AUC in WT, HFD‐induced mice, and HFD‐induced mice with EVs injection. n = 5 in each group. (I) Sirius Red staining of liver, spleen, lung, and kidney tissues from WT, HFD‐induced mice, and HFD‐induced mice with EVs injection. n = 5 in each group. Scale bars: 100µm. (J) Quantitative analysis of Sirius Red staining. n = 5 in each group. Data are presented as mean ± SEM. Statistical significance was analyzed using two‐tailed Student's t‐test in (A), one‐way ANOVA in (B), (C), (D), (F), (H), and (J), and two‐way ANOVA in (E) and (G). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig 5: Schematic of activating the osteoblastic USP26 pathway alleviates multiorgan fibrosis by decreasing IR. The loss of USP26 in osteoblasts results in decreased bone formation, as well as multi‐organ fibrosis associated with IR. Mechanistically, the absence of USP26 reduces glycolysis and lactate accumulation, leading to decreased H3K18LA in the promoter region of KSRP. This results in lower transcription of KSRP and diminished alternative splicing of FSTL1 mRNA by KSRP. Consequently, the elevated FSTL1 contributes to IR and elevated blood glucose levels. This leads to the buildup of AGEs in the blood, ultimately leading to multi‐organ fibrosis. Furthermore, activating the USP26 pathway specifically in osteoblasts through EV‐based bone‐targeting drug or compression loading can effectively prevent multi‐organ fibrosis induced by IR.