Impaired myogenic development, differentiation and function in hESC-derived SMA myoblasts and myotubes

Abstract

Spinal muscular atrophy (SMA) is a severe genetic disorder that manifests in progressive neuromuscular degeneration. SMA originates from loss-of-function mutations of the SMN1 (Survival of Motor Neuron 1) gene. Recent evidence has implicated peripheral deficits, especially in skeletal muscle, as key contributors to disease progression in SMA. In this study we generated myogenic cells from two SMA-affected human embryonic stem cell (hESC) lines with deletion of SMN1 bearing two copies of the SMN2 gene and recapitulating the molecular phenotype of Type 1 SMA. We characterized myoblasts and myotubes by comparing them to two unaffected, control hESC lines and demonstrate that SMA myoblasts and myotubes showed altered expression of various myogenic markers, which translated into an impaired in vitro myogenic maturation and development process. Additionally, we provide evidence that these SMN1 deficient cells display functional deficits in cholinergic calcium signaling response, glycolysis and oxidative phosphorylation. Our data describe a novel human myogenic SMA model that might be used for interrogating the effect of SMN depletion during skeletal muscle development, and as model to investigate biological mechanisms targeting myogenic differentiation, mitochondrial respiration and calcium signaling processes in SMA muscle cells.

Fig 1. Impaired SMN2 splicing and reduced SMN protein in myoblasts from SMA patients.

Molecular analysis of the two hESC unaffected and the two hESC SMA Type 1 lines revealed no SMN1 expression in the genomic PCR analysis, but 2 SMN2 copy numbers (A). SMN2FL (FL) and SMNΔ7 (Δ7) transcripts were significantly reduced (B). SMN protein levels in SMA Type 1 remained low compared to unaffected myoblasts (C), and were stable over 3 days after change to differentiation medium (D). Data represented as mean ± SEM of n = 6–12 individual samples and normalized to unaffected and endogenous controls (B: GAPDH, C: HSP90, D: Actin). *: p < 0.05; **: p < 0.01; ***: p < 0.001, two-tailed unpaired Student's t-test.

Fig 2. Altered myogenic marker expression in myoblasts from SMA patients.

Representative fluorescence microscopy showing myoblasts stained for PAX3, MYOD, Ki67 and MYH1E as well as nuclei (Hoechst) (A). Quantification of PAX3, Ki67, MYOD, MYH1E positive cells showed significant increase of early myogenic markers (PAX3) and proliferation marker (Ki67) and reduced late myogenic markers (MYOD and MYH1E) in SMA Type 1 cells (B). Data represented as mean ± SEM of n = 6 individual assessment. ***: p < 0.001, two-tailed unpaired Student's t-test. Scale bar: 300μm.

Fig 3. Alterations in the transcripts of myoblasts from SMA patients.

Genes from the NanoString assay panel were analyzed for direct pairwise interactions using MetaBase (www.metabase.com; ubiquitination and cleavage events were excluded). The constructed interaction network shown here includes “binding” (black lines), “positive transcription regulation” (red lines with arrow tips), “negative transcription regulation” (green lines with “T” arrows), and “phosphorylation” (blue lines with arrow tips). Nodes not connected to this major cluster were excluded. Nodes are colored by the log fold change values between healthy and diseased–blue means down in disease, red means up (Table 1). Transcription factors are depicted as rounded rectangles, and kinases as hexagons. All other nodes are circles. Eleven nodes that belong to the “Muscle_contraction” network from REACTOME are highlighted by a thicker node line. This network is significantly down-regulated as a whole. Graph generated using CytoScape (www.cytoscape.org). Please note that the assay used for the detection of SMN gene expression detects both SMN1 and SMN2. Hence, the origin of an expression signal can only be made in a context where one of the two genes is missing.

Fig 4. Impaired myogenic differentiation in myotubes from SMA patients.

Representative fluorescence microscopy showing myotubes stained for MYOG, MYH1E and TNNT as well as cell nuclei (Hoechst) and quantification of the positively stained cells showing significant decrease of the later developmental markers in SMA Type 1 cells (A). Altered transcription and translation of myogenic factors (MYOD, MYOG analyzed by RT-qPCR; MYH1/2/4/6, TNNT and Desmin analyzed by Western Blot) during 7 days of culturing observed in SMA Type 1 cells (B). Data represented as mean ± SEM of n = 6 individual samples and normalized to unaffected day 0. **: p < 0.01; ***: p < 0.001, two-tailed unpaired Student's t-test. Scale bar: 350μm.

Fig 5. Reduced cholinergic Ca2+ response in differentiated myotubes from SMA patients.

RT-qPCR of relative AChR subunit expression CHRNA1 (A) and CHRNG (B) decreased in SMA Type 1 cells. Fluorometric measurement of intracellular Ca2+ concentrations changes by carbachol treatment in ESC-SkM. Dose-response curve carbachol responses for WT and SMA ESC-SkM showed different sensitivity between genotypes (EC50 values, p < 0:05, sum-of-squares F-test). Baseline-subtracted amplitudes of carbachol responses normalized to baseline subtracted amplitudes of 4-Br-A23187 responses. These ratios were normalized to maximal values within an experiment. Data represented as mean ± SEM (A, B) and mean ± SEM of n = 6–8 individual samples from two independent experiments. *: p < 0:05, **: p < 0:01, ***: p < 0:001, two-tailed unpaired Student's t-test.

Fig 6. Impaired glycolytic function and metabolic activity in myoblasts from SMA patients.

Analysis of glycolysis with the glycolysis rate assay using the Seahorse revealed decreased levels of compensatory glycolysis in SMA Type 1 myoblasts (A, B). Mitochondrial stress test detected impaired oxidative phosphorylation in SMA Type 1 myoblasts ranging from reduced levels during basal respiration, ATP production, maximal respiration to spare capacity and proton leak (C, D). ECAR: extracellular acidification rate, OCR: oxidative consumption rate, Data represented as mean ± SEM of n = 3 individual samples. *: p < 0:05, **: p < 0:01, ***: p < 0:001, two-tailed unpaired Student's t-test.

Fig 7. Reduced ATP production in myotubes from SMA patients.

ATP levels were reduced in myotubes of SMA and unaffected cells compared to myoblasts (A). Challenge with Oligomycin and treatment with Cyclosporin A in myoblasts increased ATP levels only in SMA Type 1 derived myoblasts (B), while in myotubes Oligomycin challenge decreased in all cells the ATP levels, but also here Cyclosporin A had only a protective effect on SMA Type 1 cells (C). Mitochondrial DNA levels, an indicator for the number of mitochondrial, were not significantly reduced in SMA derived cells (D). Data represented as mean ± SEM of n = 3 individual samples. p < 0:05, **: p < 0:01, ***: p < 0:001, two-tailed unpaired Student's t-test, (A-D n = 3).