Matrix produced by diseased cardiac fibroblasts affects early myotube formation and function

Abstract

The extracellular matrix (ECM) provides both physical and chemical cues that dictate cell function and contribute to muscle maintenance. Muscle cells require efficient mitochondria to satisfy their high energy demand, however, the role the ECM plays in moderating mitochondrial function is not clear. We hypothesized that the ECM produced by stromal cells with mitochondrial dysfunction (Barth syndrome, BTHS) provides cues that contribute to metabolic dysfunction independent of muscle cell health. To test this, we harnessed the ECM production capabilities of human pluripotent stem-cell-derived cardiac fibroblasts (hPSC-CFs) from healthy and BTHS patients to fabricate cell-derived matrices (CDMs) with controlled topography, though we found that matrix composition from healthy versus diseased cells influenced myotube formation independent of alignment cues. To further investigate the effects of matrix composition, we then examined the influence of healthy- and BTHS-derived CDMs on myotube formation and metabolic function. We found that BTHS CDMs induced lower fusion index, lower ATP production, lower mitochondrial membrane potential, and higher ROS generation than the healthy CDMs. These findings imply that BTHS-derived ECM alone contributes to myocyte dysfunction in otherwise healthy cells. Finally, to investigate potential mechanisms, we defined the composition of CDMs produced by hPSC-CFs from healthy and BTHS patients using mass spectrometry and identified 15 ECM and related proteins that were differentially expressed in the BTHS-CDM compared to healthy CDM. Our results highlight that ECM composition affects skeletal muscle formation and metabolic efficiency in otherwise healthy cells, and our methods to generate patient-specific CDMs are a useful tool to investigate the influence of the ECM on disease progression and to investigate variability among diseased patients. STATEMENT OF SIGNIFICANCE: Muscle function requires both efficient metabolism to generate force and structured extracellular matrix (ECM) to transmit force, and we sought to examine the interactions between metabolism and ECM when metabolic disease is present. We fabricated patient-specific cell derived matrices (CDMs) with controlled topographic features to replicate the composition of healthy and mitochondrial-diseased (Barth syndrome) ECM. We found that disease-derived ECM negatively affects metabolic function of otherwise healthy myoblasts, and we identified several proteins in disease-derived ECM that may be mediating this dysfunction. We anticipate that our patient-specific CDM system could be fabricated with other topographies and cell types to study cell functions and diseases of interest beyond mitochondrial dysfunction and, eventually, be applied toward personalized medicine.

Keywords: Barth syndrome; Cell-derived matrices; Extracellular matrix; Mitochondria; Myogenesis; Tissue engineering.

Fig. 1.

Fabrication and Characterization of aCDM and iCDM. PDMS (Sylgard 184) substrates with micromolded topographic features (20 μm wide × 20 μm deep) (A) and flat controls (B) were utilized as templates. A sacrificial monolayer of cardiac fibroblast was plated on the micromolded and flat PDMS (Ai-ii and Bi-ii). To increase production of ECM, Ficoll 400 was used as a macromolecular crowded (blue spheres). After 14 days, the CDMs were decellularized using Triton X-100 and DNase I (Aiii and Biii). CDMs were transferred to a 12 mm diameter glass coverslip treated with silane (APTMS) and 1% glutaraldehyde (GTA, Aiv-v, bottom row: NHS Ester-Stained images and picture of samples after transferring). After UV sterilization for 2 h, C2C12s were seeded on the aCDM and iCDM. Scale bars = 100 μm. AFM scan of the surface of a dried (C) Characterization of the preferred topographic orientation of aCDM from NHS-Ester-stained images confirmed higher orientation (OI) compared to iCDM (OI=1 perfect anisotropy, OI=0 perfect isotropy). Data represented as mean + s.d from n = 20 (4 ROI from 5 NIH Ester labeled images (shown in Fig. 1Av, Bv) per condition). Statistical differences between groups were analyzed using nonparametric Wilcoxon Test, ^∗∗∗p < 0.0001. (D) aCDM and (E) iCDM showing a 2D (left) and a 3D (right) representation illustrating the overall surface topography, color bars indicate height of dried sample.

Fig. 2.

Cell adhesion and alignment of hPSC-CFs cultured for 7 days on healthy and BTHS aCDM and iCDM. To characterize attachment of the cells and alignment of their cytoskeleton, samples were fixed and stained at Day 1, Day 5, and Day 7. A) Representative images of C2C12 skeletal myoblasts cultured for 7 Days on aCDM and iCDM produced by healthy hPSC-CFs. The myoblast were visualized using Phalloidin to stain for F-actin cytoskeleton (Green) and DAPI to label the nuclei (blue). Scale bar = 100 μm. B) Characterization of the orientation index of F-actin fibers in response to the aCDM and iCDMs. iCDM (OI=1 perfect anisotropy, OI=0 isotropic). Data represented as mean ± s.d from n = 12 (4 ROI from 3 replicates images per condition). Three-way ANOVA was used to analyze the effect on OI of CDM Anisotropy: p < 0.0001, Days: p < 0.0001, and CDMs Patient cell source: p = 0.1968 (interaction p = 0.8737), followed by a nonparametric comparison for each pair using Wilcoxon method as a post-hoc test to analyze differences aCDM and iCDM groups, ^∗∗∗p < 0.0001.

Fig. 3.

Myotubes after cultured for 7 Days on aCDM and iCDM from healthy and BTHS fibroblast. (A) Schematic illustrating the protocol used to differentiate into myotubes the C2C12 myoblasts cultured on healthy and BTHS aCDM and iCDMs. (B) Representative Immunofluorescent images from myoblast culture for 7 days on iCDM and aCDM fabricated by healthy and 3 different BTHS hiPSCs-CFs. IF images are showing MF20 (green) staining for myosin heavy chain and DAPI (blue) for cell nuclei. Scale bar = 100 μm. Data showing quantified Myotube Orientation Index (C) and Myotube Fusion index (D) at day 7. Data represented as mean + s.d from n = 12 images (4 ROI from 3 replicates) per condition. Two-way ANOVA was used to analyze the effect of multiple factors on Orientation Index (Alignment: p < 0.001, Patient cell source: p = 0.59, interaction: p = 0.6) and Myotube Fusion Index (Alignment: p < 0.001, Patient cell source: p < 0.001, interaction: p = 0.003), followed by a nonparametric comparison for individual factors using Wilcoxon method as a post-hoc test, significance was set at an alpha value of 0.05, ^∗∗∗p < 0.0001.

Fig. 4.

Mitochondrial function in response to healthy and BTHS CDMs during myogenic differentiation of C2C12s. (A) Schematic illustrating C2C12s cultured on two healthy and three BTHS CDMs for 24 h in standard medium (Day 0) and 7 Days in differentiation medium (Day 7). Representative images of α-actinin labeled myotubes differentiated on H1-CDM and B2-CDM for 7 days confirm equivalent differentiation stages of myoblast culture on healthy and BTHS CDMs. Scale bar 20 μm. (B) ATP generation, (C) ROS levels, and (E) TMRE levels were measured at the two timepoints. Data represented as mean ± s.d from N = 4 independent experiments (each data point represents an average of 4 to 6 wells). Statistical differences between groups were analyzed using non-parametric comparison for each pair using Wilcoxon Test, significance was set at an alpha value of p ≤ 0.05. ^∗p < 0.01, ^∗∗p < 0.001, ^∗∗∗p < 0.0001.

Fig. 5.

Nano-LC/MS/MS Analysis of B2-CDMs and H1-CDMs. (A) Flow chart depicting the protein screening process. (B) A volcano plot displaying log 2-fold change ratios of B2-CDMs as compared to H1-CDMs of the ECM and Associated Proteins. All proteins above the horizontal gray dashed line (p = 0.05) express significantly altered expression levels. (C) Western Blot confirming the upregulation of P3H1 and downregulation of HAPLN1 in B2-CDMs compared to H1-CDMs.(D) Table classifying the 15 differentially expressed proteins and summarizing some of their gene ontology in terms of associated molecular and biological processes.