Fig 2: MGP modulates the TGFβ pathway in C3H10T1/2 mesenchymal progenitor cells affecting adipogenesis and fibrosis in vitro. C3H10T1/2 cells transfected with lentiviral vectors were used to generate stable cell lines labeled Shmgp for Mgp knockdown or scrambled (Scr) control. The experiments were replicated 3 times. A. Schematic diagram of adipogenic induction, maintenance and sample collection in C3H10T1/2 cells. B. Expression of Mgp in Shmgp or Scr transfected cells. Shmgp suppressed expression of Mgp about 70% compared to Scr control on day 0. C. Time course expression of Adipoq, Col1a1, Tgfβ1, and Alk5 in Shmgp or Scr transfected cells, as determined by qPCR. D. Bodipy staining of Shmgp or Scr transfected cells after 12 days of treatment with different concentration of TGFβ1 (0–20 ng/ml). E, F. Expression of (E) SMAD4, SMAD2/3, pSMAD3, TGFβ1, and pSMAD1/5, and (F) Fibronectin (FN1), hormone-sensitive lipase (HSL), phospho(p)-HSL in Shmgp or Scr transfected cells, after 12 days of treatment with different concentrations of TGFβ1 (0–20 ng/ml), as determined by immunoblotting with densitometry (representative of 3 experiments). G. Expression of Adipoq, Col1a1, Tgfb1, and Alk5 after 12 days of treatment with different concentration of TGFβ1 (0–20 ng/ml), as determined by qPCR. H. Bodipy staining of Shmgp or Scr transfected cells after 12 days of treatment with 5 μM SB431542. Data are shown as mean ± SEM; One way ANOVA, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Fig 3: DPP4 inhibition limits adipose fibrosis caused by Pdgfrα Cre-mediated Mgp deletion. A. Schematic diagram of treatment for mice with sitagliptin (10 mg/kg) in 5% DMSO or control 5% DMSO. Mgpf/f (F/F) and Mgpf/f,PdCre (Pd–KO) mice were injected daily from P7 to P28. The experiments in panels b-f were performed using inguinal adipose tissue from F/F and Pd–KO mice after sitagliptin or control treatment. Each group has the same number of mice. B. Masson’s trichrome (MT) and Picrosirius Red staining, quantified by ImageJ (n = 3). C. Immunofluorescence for Perilipin-1 (green) and CD31 (red) in inguinal adipose tissue from F/F and Pd–KO mice at 4 weeks. DAPI (blue) was used to visualize nuclei. Bars, 25 μm. The LD size was quantified by ImageJ (n = 3). D. FACS analysis of the stromal vascular fraction (SVF) for detection of early adipose progenitor cells (PDGFRα+; DPP4+), after removal of immune and endothelial cells in inguinal adipose tissue from F/F and Pd–KO mice (n = 4 mice per group). Percent (%) cells is representative of 3 experiments. E. Expression of Pparγ1, Pparγ1, Perilipin-1, hormone-sensitive lipase (HSL), phospho(p)-HSL, and COL1A1, as determined by immunoblotting with densitometry (protein from n = 5; representative of 3 experiments). F. Expression of SMAD2, SMAD3, pSMAD3, pSMAD2, pSMAD1/5, TGFβ1, ALK5, and total SMAD, as determined by immunoblotting with densitometry (protein from n = 5; representative of 3 experiments). Data are shown as mean ± SEM; One way ANOVA, ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

Fig 4: DMSO does not protect mice from CDA–HFD-induced liver fibrosis. (A) Expression of Cyp7a1, Hsd3b7, Cyp8b1, Cyp27a1, and Cyp7b1 in WT mice kept on 1% or 2% (v/v) DMSO water for one week; n = 5 for each group. (B) Schematic of mouse treatment for panels (C–I): C57Bl/6J WT mice were all treated with CDA–HFD, but with or without 2% (v/v) DMSO in their drinking water. The treatment lasted for 6 weeks. (C) Hepatic expression of Cyp7a1, Hsd3b7, Cyp8b1, Cyp27a1, and Cyp7b1 in mice described in Panel (B). (D) Total BA levels in the liver. (E) Total BA levels of serum. (F) Representative H&E staining of liver sections from mice described in panel (B). Scale bar: 200 µm. (G) Hepatic expression of Col1a1 and Acta2. (H) Liver collagen levels in the mice described in panel (B). (I) Representative Sirius Red staining of liver sections from mice described in panel (B). Scale bar: 200 µm. For panels (C–I), n = 7 for the control group and n = 6 for the 2% DMSO group. All data are mean ± SEM; * p < 0.05, ** p < 0.01, relative to the control group.

Fig 5: Myeloid AMPK restricts fibrosis but is not critical for metformin-associated improvements. A and G: representative Picrosirius Red-stained liver sections (20x magnification, 100 μm scale bar) of male and female control Prkaa1/2fl/fl (Flox) and Prkaa1/2fl/fl/LysM-Cre+ (MacKO) mice fed a CDAHFD or CDAHFD + 250 mg/kg/d metformin. B and H: total collagen quantification. C–E and I–K: hepatic mRNA transcript levels of markers of extracellular matrix-regulating transcripts (Col3a1, Col1a1, Acta2). F and L: serum levels of liver injury markers (alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and total bilirubin (TBIL)). Transcript expression was normalized to the average expression of Actb and Hprt and expressed relative to LFD-fed Flox mice. Data were analyzed by Two-way ANOVA with Tukey’s test for multiple comparisons where ∗, ∗∗, ∗∗∗, and ∗∗∗∗ represent P < 0.05, P < 0.01, P < 0.001, and P < 0.0001, respectively. The geometric mean of LFD Flox controls is represented by a hashed line.