A 3D adipogenesis platform to study the fate of fibro/adipogenic progenitors in muscular dystrophies

Abstract

In human dystrophies, progressive muscle wasting is exacerbated by ectopic deposition of fat and fibrous tissue originating from fibro/adipogenic progenitors (FAPs). In degenerating muscles, the ability of these cells to promote successful healing is attenuated, and FAPs aberrantly expand and differentiate into adipocytes and fibroblasts. Thus, arresting the fibro/adipogenic fate of FAPs, without affecting their physiological role, represents a valuable therapeutic strategy for patients affected by muscle diseases. Here, using a panel of adipose progenitor cells, including human-derived FAPs, coupled with pharmacological perturbations and proteome profiling, we report that LY2090314 interferes with a genuine adipogenic program acting as WNT surrogate for the stabilization of a competent β-catenin transcriptional complex. To predict the beneficial impact of LY2090314 in limiting ectopic deposition of fat in human muscles, we combined a poly-ethylene-glycol-fibrinogen biomimetic matrix with these progenitor cells to create a miniaturized 3D model of adipogenesis. Using this scalable system, we demonstrated that a two-digit nanomolar dose of this compound effectively represses adipogenesis at higher 3D scale, thus indicating the potential for LY2090314 to limit FAP-derived fat infiltrates in dystrophic muscles.

Keywords: Adipogenesis; Fibro/adipogenic progenitors; LY2090314; Muscular dystrophies; Tissue engineering; β-catenin.

Fig. 1.

A functional β-catenin transcriptional complex suppresses adipogenesis of 3T3-L1 cells. (A) Schematic representation of the drug perturbation experimental plan. (B) Representative immunofluorescence analysis showing Oil Red O (ORO) staining after the drug perturbation experiment described in A. Cells were treated with the reported compounds for 72 h in growth medium and then cultured for six additional days in a compound-free medium. Nuclei (blue) were revealed using Hoechst 33342. (C) Dot plot reporting the fraction of ORO-positive cells in the experiment in B. (D) Representative immunofluorescence analysis showing ORO staining after the drug perturbation experiments with different pairwise combinations. (E) Dot plot reporting the fraction of ORO-positive cells in the experiment in D. (F) Schematic model summarizing the results presented in this figure. All micrographs were captured at 20× magnification. Scale bars: 50 μm; 25 μm (insets). Statistical significance was determined by one-way ANOVA. Images and data are representative of at least three independent biological repeats. All data are presented as mean±s.e.m. **P<0.01; ***P<0.001.

Fig. 2.

LY2090314 stabilizes β-catenin intracellular amount and represses adipogenesis of 3T3-L1 cells. (A) Schematic model of action of LY2090314. IBMX, 1-methyl-3-isobutyl xanthine. (B) Representative western blot showing the expression of β-catenin and PPARγ proteins during the first 3 days of differentiation of 3T3-L1 cells. Vinculin was used as loading control. (C) Representative immunofluorescence analysis showing differentiated 3T3-L1 cells in the presence/absence of 20 nM LY2090314 (left). Cells were cultured for 3 days in adipocyte differentiation medium (ADM) in the presence/absence of LY2090314 and then cultured for six additional days in a compound-free maintenance medium. Nuclei (blue) were revealed using Hoechst 33342. The bar plot (right) reports the percentage of ORO-positive cells (red) in both conditions. (D) Representative western blots showing the expression of perilipin-1 and PPARγ proteins at the end point of the differentiation program of 3T3-L1 cells, with or without exposure to 20 nM LY2090314. Vinculin was used as loading control. (E) Representative western blots showing the expression of β-catenin and non-phospho active β-catenin proteins at the end point of the differentiation program of 3T3-L1 cells, with or without exposure to 20 nM LY2090314. Vinculin was used as loading control. (F) Representative immunofluorescence analysis showing differentiated 3T3-L1 cells in the presence/absence of a combinatorial treatment of 20 nM LY2090314 and 10 or 20 µM iCRT-3 (left). Nuclei (blue) were revealed using Hoechst 33342. The bar plot (right) reports the percentage of ORO-positive cells (red) in all the considered conditions. (G) Schematic model summarizing mechanistic details of LY2090314 and iCRT-3 in the adipogenic pathway. All micrographs were captured at 20× magnification. Scale bars: 50 μm; 25 μm (insets). Statistical significance was determined by paired two-tailed Student’s t-test (C) and one-way ANOVA (F). Figures and data are representative of at least three independent biological repeats. All data are presented as mean±s.e.m. **P<0.01; ***P<0.001.

Fig. 3.

LY2090314 abrogates adipogenesis of murine and human fibro/adipogenic progenitors (FAPs). (A) Representative scheme showing the collection and utilization of human-derived FAPs (hFAPs). iST, in-StageTip. (B) Principal component analysis showing sample segregation of hFAP proteomes upon exposure to LY2090314 or vehicle. (C) Heatmap showing significantly [false discovery rate (FDR)<0.05; fold difference>±1] regulated proteins in the LY2090314-versus-vehicle comparison. LFQ, label-free quantitation. (D) Gene Ontology terms that are positively enriched (adjusted P<0.01) in the LY2090314-versus-vehicle comparison. GOBP, Gene Ontology biological process. (E) Volcano plot showing positively and negatively enriched proteins in the LY2090314-versus-vehicle comparison. (F) Representative western blot showing the expression of β-catenin in hFAPs in the absence/presence of 20 nM LY2090314 (left). The bar plot (right) reports the densitometric values of β-catenin in both conditions. (G) Representative immunofluorescence analysis showing differentiated hFAPs in the presence/absence of 20 nM LY2090314 (left). Cells were cultured for 3 days in ADM in the presence/absence of LY2090314 and then cultured for six additional days in a compound-free maintenance medium. Nuclei (blue) were revealed using Hoechst 33342. The bar plot (right) reports the percentage of ORO-positive cells in both conditions. (H) Network reporting the mode of action of LY2090314 in repressing the adipogenic program of hFAPs. Statistical significance was determined by paired two-tailed Student’s t-test. Images and data are representative of at least three independent biological repeats. All data are presented as mean±s.e.m. **P<0.01; ***P<0.001.

Fig. 4.

LY2090314 suppresses 3D adipogenesis of murine and human FAPs. (A) Representative scheme showing the generation of a miniaturized 3D model of fat infiltration using poly-ethylene-glycol-fibrinogen (PF) hydrogel. hASC, human adipose stromal cell; hFAP, human FAP; mFAP, murine FAP. (B) Components (i.e. the mold chamber and fork) for the generation of a 3D model of fat infiltration using PF hydrogel. A representative polymerized PF construct is also shown. Scale bar: 1 cm. (C) A previously transplanted PF construct containing hFAPs was explanted after 30 days (left). Scale bar: 1 cm. Representative cryosections of the explanted construct showing hFAPs expressing COL3A1 (red) and perilipin-1 (green) (right, top row), and COL1A1 (green) and ORO (red) (right, bottom row). Scale bars: 50 μm; 25 μm (insets). (D) PF hydrogel construct loaded with hASCs induced to differentiate into adipocytes using standard protocols. 3D differentiated adipocytes with evident lipid droplets are evident in brightfield microscopy (top). Scale bar: 100 µm; inset is an enlarged view. Representative whole-mount immunofluorescence showing perilipin-1 expression (green) in differentiated hASC-derived adipocytes (bottom). Scale bars: 50 µm; 10 µm (inset).

Fig. 5.

LY2090314 inhibits 3D adipogenesis of mFAPs and hFAPs. (A) Representative ORO staining of PF constructs previously loaded with mFAPs. Both constructs were exposed to ADM in the absence/presence of 20 nM LY2090314. Scale bar: 1 cm. (B) Representative brightfield micrographs showing mFAPs undergoing adipogenesis in the absence/presence of 20 nM LY2090314, at 3 and 6 days after ADM exposure. Scale bar: 15 µm. (C) Representative whole-mount immunofluorescence showing ORO labeling in mFAP-derived adipocytes (red) differentiated in the absence/presence of 20 nM LY2090314. Scale bars: 100 µm (top); 50 µm (middle); 10 µm (bottom). (D) Representative ORO staining of PF constructs previously loaded with hFAPs (top). Scale bar: 1 cm. Representative whole-mount immunofluorescence showing ORO labeling in hFAP-derived adipocytes (red) differentiated in the absence/presence of 20 nM LY2090314 (bottom). Scale bar: 50 µm. Images and data are representative of at least three independent biological repeats.