The immunosuppressant drug azathioprine restrains adipogenesis of muscle Fibro/Adipogenic Progenitors from dystrophic mice by affecting AKT signaling

Fibro/Adipogenic Progenitors (FAPs) define a stem cell population playing a pro-regenerative role after muscle damage. When removed from their natural niche, FAPs readily differentiate into adipocytes or fibroblasts. This digressive differentiation potential, which is kept under tight control in the healthy muscle niche, contributes to fat and scar infiltrations in degenerative myopathies, such as in Duchenne Muscular Dystrophy (DMD). Controlling FAP differentiation by means of small molecules may contribute to delay the adverse consequences of the progressive pathological degeneration while offering, at the same time, a wider temporal window for gene therapy and cell-based strategies. In a high content phenotypic screening, we identified the immunosuppressant, azathioprine (AZA) as a negative modulator of FAP adipogenesis. We show here that AZA negatively affects the adipogenic propensity of FAPs purified from wild type and mdx mice by impairing the expression of the master adipogenic regulator, peroxisome proliferator-activated receptor γ (PPARγ). We show that this inhibition correlates with a decline in the activation of the AKT-mTOR axis, the main pathway that transduces the pro-adipogenic stimulus triggered by insulin. In addition, AZA exerts a cytostatic effect that has a negative impact on the mitotic clonal process that is required for the terminal differentiation of the preadipocyte-committed cells.


Muscle mononuclear cell differentiation
After four days from plating, muscle mononuclear cells were refreshed with new GM and cultured for additional two days before the induction of differentiation.
Fibrogenic differentiation was performed by exposing cells to GM supplemented by 5 ng/ml of human recombinant Transforming Growth Factor-β (TGF-β, PeproTech, catalog 100-21). Cells were treated for three days before fixing. Similarly, osteogenic differentiation was obtained by exposing cells to a fresh GM supplemented with 1 µg/ml of human recombinant Bone Morphogenetic Protein-2 (BMP-2, PeproTech, catalog #120-02). Treated cells were fixed after three days after morphogen exposure. Adipogenic differentiation was induced by exposing cells to the Adipocyte Differentiation Medium (ADM) obtained by adding 1 µg/ml human recombinant insulin (Sigma-Aldrich, catalog I9278), 0.5 mM 3-isobutyl-1-methylxanthine (IBMX, Sigma-Aldrich, catalog I5879), 1 µM Dexamethasone (Dexa, Sigma-Aldrich, catalog D4902) to GM ® . After 48 hours, the ADM was removed and replaced by the Adipocytes Maintenance Medium (AMM) obtained by adding 1 µg/ml insulin to GM. Muscle mononuclear cells were cultured in AMM for three additional days before fixing. Adipogenic differentiation was further stimulated by supplementing the ADM with 1 µM Rosiglitazone.

3T3-L1 preadipocytes culture and differentiation
3T3-L1 preadipocyte cell line was purchased from American Type Culture Collection ® (ATCC ® ) and maintained in high glucose Dulbecco's Modified Eagle Medium (DMEM) (ATCC ® 30-2002 ™ ) supplemented with 10% Iron Fortified Calf Serum (ATCC ® , catalog 30-2030 ™ ). Differentiation assays were performed in 24-well plate (BD Falcon, catalog 353226) plating cells at the cell density of 3×10 4 cell/cm 2 . Adipogenic differentiation was induced following ATCC ® guidelines. Briefly, three-days confluent cells were exposed to Adipocyte Differentiation Medium (ADM) (day 0 of differentiation). Two-days later, ADM was replaced with Adipocyte Maintenance Medium (AMM). AMM was refreshed every two days until complete cell differentiation (day 8 of differentiation). For the effects of drugs on the adipogenic differentiation, azathioprine (AZA) and trichostatin A (TSA) were added at the appropriate concentrations at day 0 or continuously for the whole differentiation period. For the effects on the signaling pathways perturbed by the drugs, a twelve-hours pre-treatment exposure was performed in growth medium and then followed by the treatment in ADM. Other experimental procedures on 3T3-L1 were performed with the same treatment protocols in 6-well plate (BD Falcon, catalog 353046) maintaining the relation cells/surface area.
Live/death cells were discriminated using propidium iodide (Pi). For each sample 10,000 3 events were acquired using BD FACSCALIBUR (BD Biosciences).
For cell cycle analysis, 3T3-L1 preadipocytes were cultured as described and the DNA content assessed by flow cytometry after Pi staining (BD Biosciences, BD FACSCALIBUR). Briefly, 3T3-L1 were induced to differentiate in ADM supplemented with AZA and TSA and the cells harvested at the indicated time points. Trypsinized cells were washed twice in PBS 1X, fixed and permeabilized for 30' in ice with a cold solution of methanol:acetone (4:1) in ratio 1:1 with PBS 1X. Fixed and permeabilized cells were centrifuged at 900×g and each pellet resuspended in a solution of 100 µg/ml ribonuclease A (Sigma-Aldrich, catalog R6513) in PBS 1X for 30 minutes at RT. Samples were stained with propidium iodide for 20 minutes at RT at the final concentration of 0.5 mg/ml. For each sample 10,000 3 events were acquired. All cell cycle data were analyzed using FlowJo 10 (treestar).

Alkaline phosphatase (ALP) activity assay
ALP activity was revealed by colorimetric assay using NBT/BCIP stock solution (Sigma-Aldrich, catalog 11681451001). Briefly, fixed cells were washed twice with PBS 1X and incubated for 15 minutes at room temperature with a staining solution in which the substrates 0.33 mg/ml of nitro-blue tetrazolium chloride (NBT) and 0.165 mg/ml of 5-bromo-4-chloro-3 ' -indolyphosphate p-toluidine salt (BCIP) were diluted in AP Buffer (100 mM TRIS, 10 mM NaCl, 1 mM MgCl2 ). Stained cells were washed twice with PBS 1X and counterstained using Hoechst 33342.

Multiplex gene expression analysis by QuantiGene® Plex Assay
The Adipogenic and fibrogenic fates of mdx FAPs were assessed at the transcriptional level by  Figure S1. Differentiation potential of muscle mononuclear cells purified from wild type mice. Muscle mononuclear cells were isolated from wild type mice and cultured for 6 days in growth medium (GM) prior to differentiation induction. For inducing osteogenic differentiation, muscle mononuclear cells were incubated for 3 days in GM supplemented with 1 µg/ml BMP-2 or vehicle alone. A. Representative immunofluorescence, taken at 20× magnification, of differentiated cells stained for alkaline phosphatase (ALP) with the NBT/BCIP solution (black). B. ALP stained wells (black) from three independent samples of muscle mononuclear cells exposed for three days to the vehicle or to 1 µg/ml BMP-2, respectively. C. The bar plot represents the average fraction of the ALP-positive area (expressed in pixel) per cell, in each field, in the two experimental conditions. D. a-SMA (E). The protein signals were normalized over vinculin signal. Densitometric analysis were performed using ImageJ. All experimental data are presented as mean of three independent experiments ± SEM. The statistical significance was estimated by one way ANOVA and defined as *p < 0.05; **p < 0.01; ***p < 0.001. F. Representative immunoblot in which, for each sample, 15 µg of the protein lysate was electrophoresed on a 4-15% gradient gel and the protein levels of both PPARg isoforms were revealed with a specific antibody. Vinculin serves as loading and normalizing control. The full-length blot is represented in Supplementary Figure S12. G, H. Bar plots of the densitometric analysis of PPARg1(G) and PPARg2(H) isoforms. Densitometric analysis were 8 performed using ImageJ. The experimental data is presented as mean of three independent experiments ± SEM. The statistical significance was estimated by Student's t-test and defined as *p < 0.05; **p < 0.01; ***p < 0.001. I. Freshly purified mdx FAPs were cultured and treated with 25 µM AZA and 50 nM TSA according to the protocol summarized in Figure 4 D to perform QuantiGene® plex analysis. Differentiated cells were lysed in 96-well and 4-fold dilutions were performed from the undiluted samples as recommended. Diluted cell lysates were incubated overnight at 54°C using the VorTemp 56 shaking incubator at 600 rpm in the presence of the capture beads that act as a support for the designed probe sets. The gene expression profile of mdx FAPs for adipogenic (C/ebpb, Pparg, Adipoq) and fibrogenic genes (S100a4, Col1a2, Col6a1, Fn1) was detected by the MagPIX ® system. The assay was run in technical triplicates for two independent samples (n=2). The relative expression level for each gene arises from the mean fluorescence intensity (MFI) of the coloured bead-oligo associated upon normalization over the MFI of the bead of the Actb gene. Statistical significance was estimated by Student's t-test and defined as *p < 0.05; **p < 0.01; ***p < 0.001. The full list of each oligo and its cognate bead is reported in Supplementary Table S2