Restoration of myogenesis in ALS-myocytes through miR-26a-5p-mediated Smad4 inhibition and its impact on motor neuron development

Abstract

Amyotrophic lateral sclerosis (ALS) is the most common adult-onset paralytic disorder, characterized primarily by a progressive loss of motor neurons (MNs) in which degeneration skeletal muscle involvement has been demonstrated. Skeletal muscle is a plastic tissue that responds to insults through proliferation and differentiation of satellite cells. Skeletal muscle degeneration and regeneration are finely regulated by signals that regulate satellite cell proliferation and differentiation. It is known that satellite cell differentiation is impaired in ALS, but little is known about the involvement of microRNAs (miRNAs) and their role in intercellular communication in ALS. Here we demonstrated impaired differentiation of satellite cells derived from ALS mice related to the impairment of myogenic p38MAPK and protein kinase A (PKA)/pCREB signaling pathways that can be regulated by miR-882 and -134-5p. These miRNAs participate in autocrine signaling in association with miR-26a-5p that, secreted from wild-type (WT) and captured by ALS myoblasts, enhances ALS-related myoblast differentiation by repressing Smad4-related signals. Moreover, miR-26a-5p and -431-5p work in a paracrine way ameliorating motoneuron differentiation. These findings emphasize the need to better understand intercellular communication and its role in ALS pathogenesis and progression. They also suggest that miRNAs could be targeted or used as therapeutic agents for myofiber and MN regeneration.

Keywords: MT: non-coding RNAs; amyotrophic lateral sclerosis; cell communication; miRNA; myogenesis; neurogenesis; neuromuscular disorders; primary stem cells; skeletal muscle.

Graphical abstract

Figure 1

PCM derived from muscles of hSOD1(G93A) Tg mice show impaired in vitro differentiation compared to the hSOD1(WT) counterpart (A) The expression of EmbMyHC, Pax7, MyoD, and myogenin was evaluated by WB in hSOD1(WT) and hSOD1(G93A) PCM at different DIV. The left shows representative WBs and the corresponding Coomassie blue staining of the membrane, while the graphs on the right report the densitometric analysis for each protein. n = 3 for each DIV and hSOD1 genotype. (B) p38 (MAPK) phosphorylation (Thr180/Tyr182) and the expression of p21 were analyzed by WB at different DIV. The top shows representative WBs, and the graphs on the bottom show the corresponding densitometric analysis by reporting the ratio between phosphorylated and total protein amounts for p38 while the abundance of p21 was determined by normalizing the optical density of immunoreactive bands to that of the corresponding Coomassie blue-stained lanes. n = 4 (different PCM) for each DIV and hSOD1 genotype. (C) 4 DIV PCM protein extracts were analyzed for phosphorylation of PKA substrates by WB using α-pPKAs as antibody. Loading control with the Coomassie blue staining. n = 4 (different PCM) for each DIV and hSOD1 genotype. (D) CREB phosphorylation (Ser133) was analyzed by WB at different DIV. The left shows representative WBs, and the graph on the right show the corresponding densitometric analysis by reporting the ratio between phosphorylated and total protein amount. n = 4 (different PCM) for each DIV and hSOD1 genotype. ∗ indicates bands considered for protein quantification. In (A, B, and D), ∗p < 0.05 and ∗∗p < 0.01, hSOD1(WT) vs. hSOD1(G93A) in each DIV; §p < 0.05 and §§p < 0.01 with respect to hSOD1(WT) in 1 DIV; and #p < 0.05 and ##p < 0.01 with respect to hSOD1(G93A) in 1 DIV; two-way ANOVA followed by Sidak’s multiple comparison test. In (C), ∗p < 0.05, Mann-Whitney test.

Figure 2

hSOD1(G93A) co-cultured with hSOD1(WT) PCM recover the defective differentiation phenotype (A) Representative fluorescent images of the three different co-culture conditions (WT_WT, top; WT_G93A, middle; and G93A_G93A, bottom) (see Figure S2). At 4 DIV, target cells were co-stained with anti-desmin antibody (red signal) and the Hoechst 33342 dye (blue signal). Scale bars, 20 μm. (B) The graph reports the frequency distribution of the myotube area under the three different culturing conditions. (C) The diagram reports the average myotube area in the three co-culture settings normalized to the mean value of the WT_WT cultures. n = 21 for WT_WT and WT_G93A, while is 35 for G93A_G93A. (D and E) The bar diagrams show the differentiation index and the fusion index under the three co-culture conditions. (F) At 4 DIV, the EmbMyHC expression was assessed by WB in the three co-culture conditions. The left shows a representative WB and the corresponding Coomassie blue staining of the membrane, while graphs on the right report the densitometric analysis. (G) The activation of CREB and the expression of myogenin were assessed at 4 DIV co-cultures by WB as described in Figure 1. The left shows representative WBs and the corresponding Coomassie staining of the membrane, while the bar diagrams on the right show the corresponding densitometric analysis. n is reported inside the bars. For all panels, n.s., not significant, ∗p < 0.05 ∗∗p < 0.01, and ∗∗∗∗p < 0.0001. One-way ANOVA followed by Tukey’s multiple comparison test.

Figure 3

miRNA expression (A) Heatmap of differentially expressed miRNAs in the medium derived from the conditions described in the sample names (columns). The expression of each miRNA is indicated as relative to the average of the miRNA in the samples. Biological replicates cluster together and G93A_G93A samples cluster separated from WT_WT and G93A_WT. On the right, it is shown miRNA cluster with a low expression in the G93A_G93A sample (blue) and the one with high expression in G93A_G93A (violet). # and @ indicate miRNAs associated with striated muscle functions according to target function. (B) Relative expression of miRNAs within the blue cluster according to RT-qPCR experiments. (C) Relative expression of miRNAs within the violet cluster according to RT-qPCR. For both (B and C) ∗p < 0.05, ∗∗p < 0.005, and ∗∗∗p < 0.0005 calculated according to t test between indicated samples considering unequal variance. SD is indicated.

Figure 4

Luciferase reporter assay and miRNA activity in C2C12 cells Luciferase reporter assays were performed to demonstrate the direct interaction between miRNAs and their targets. Part of the 3′-UTR sequence containing the miRNA putative interaction site (or not containing; Ctrl) was cloned in pmirGLO vector. Firefly luciferase (reporter gene) and Renilla luciferase (control reporter for normalization) activities were measured after the transfection in C2C12 cells together with miRNA mimics or miRNA scramble sequence (scramble). (A–C) Targets of miR-134-5p (A), miR-26a-5p (B), and miR-882 (C) showed a decreased luciferase activity. (D) After cell culture for 24 h with the medium containing miRNA mimics and 24 h of recovery, miRNAs within the cells increased their expression more than two times in comparison with the cells cultured with scramble sequence. Expression of miRNA targets validated by luciferase reporter assays decreased when the expression of miRNAs increased. (E–G) The expression of Creb1 decreased about twice (E), MyoD1 and Pax7 by about 50 percent (F), and Igfbp5 showed a decrease in its expression by one-third (G). Data are expressed as the mean of at least four independent transfections. Error bars indicate SD. p values from t tests considering independent variance between two groups in analysis are indicated as follows: ∗p < 0.05, ∗∗p < 0.001, and ∗∗∗p < 0.0001.

Figure 5

Relative expression of miRNAs and targets within hSOD1(G93A) or hSOD1(WT) myoblasts co-cultured with hSOD1(WT) or hSOD1(G93A) myoblasts respectively and Smad activity (A) The bar diagram reports the normalized amount of miRNAs belonging to the blue (first three miRNAs) or violet (last three miRNAs) clusters in hSOD1(G93A) (G93A_G93A or G93A_WT) or hSOD1(WT) (WT_G93A or WT_WT) cells co-cultured as reported in Figure S2. miR-26a-5p is the most expressed miRNA in hSOD1(WT) cells, while miR-28a-3p is most expressed in hSOD1(G93A) cells. (B) The bar diagram reports the normalized amount of miRNA targets. The expression of Igfbp5 is decreased in G93A_WT condition, while Pax7, MyoD1, and Creb1 are more expressed both in the G93A_WT and WT_WT. Smad1 and Smad4 show an opposite behavior respect to Pax7, MyoD1, and Creb1. Specular expression of miRNA and their targets support miRNA activity. (C) The bar diagram reports the normalized amount of Smad4 target genes in hSOD1(G93A) PCM co-cultured with hSOD1(G93A) PCM (G93A_G93A) or hSOD1(WT) (G93A_WT). Expression of all Smad4 target genes decreased in hSOD1(G93A) PCM co-cultured with hSOD1(WT) PCM. hSOD1(WT) PCM secrete miR-26a-5p that modulate the expression of Smad4. (D) Relative DNA enrichment after Smad4 chromatin immunoprecipitation. Considering a DNA region not associated with Smad4 interaction and the immunoprecipitation with the isotype antibody, no differences were evidenced between G93A_G93A and G93A_WT co-cultures. We obtained opposite and expected results after the utilization of IgG against Smad4 confirming the low abundance of Smad4 caused by the upregulation of miR-26a-5p. ∗p < 0.05 and ∗∗p < 0.005, calculated according to t test between indicated samples considering unequal variance. SD is indicated considering at least four replicates per experiment.

Figure 6

Activity of miRNAs on motor neuron cells (A) Representative immunohistochemical images of motor neuron cultured with hSOD1(WT) (MCM WT) or hSOD1(G93A) PCM (MCM G93A). In blue, it is indicated cell nuclei and in red, motor neuron body and dendrites. (B) Calculation of neurite length. It decreases when hSOD1(G93A) PCM-derived culture medium was added to motor neurons. (C) Representative immunohistochemical images of motor neuron cells transfected with an empty pCMVmiR-GFP vector (empty) or a pCMVmiR-26a-GFP vector (miR-26a-5p) are presented. Green indicates transfected cells (GFP positive), blue indicates nuclei, and red represents the motor neuron body and neurites. (D) Only transfected cells (GFP-positive) were analyzed for neurite length, demonstrating an increase in neurite length when miR-26a is overexpressed. (E) Representative immunohistochemical images of motor neurons cultured in medium derived from C2C12 cells (MC C2C12) overexpressing miR-26a-5p, miR-134-5p, and miR-431-5p are presented. Red staining highlights the bodies and neurites of motor neurons, while blue staining depicts their nuclei. (F) Quantitative assessment of neurite length in motor neurons treated with culture medium derived from C2C12 cells overexpressing miR-26a-5p, miR-134-5p, and miR-431-5p. The miR-26a-5p and miR-431-5p miRNAs exhibit a stimulatory effect on neurite elongation. (G) Representative image of motor neuron neurite branching after miR-26a-5p overexpression. (H) Representative image of motor neuron neurite branching of motor neurons cultured in medium derived from C2C12 cells (MC C2C12) overexpressing miR-26a-5p. In both cases neurites appeared branched and several neurites outgrowth from cell body. (I) Quantification of neurites per body cell. The average number of neurites per 20 cells is represented. It increases when miR-26a is expressed from the cells (miR-26a modulation) or is present in the medium derived from C2C12 overexpressing miR-26a (MC C2C12). ∗∗p < 0.005 and ∗∗∗p < 0.0005, calculated according to t test between indicated samples considering unequal variance. SD is indicated considering at least three biological replicates and 15 cells analyzed per biological replicate.

Figure 7

Gene expression of miRNAs and Smad1 and Smad4 in skeletal muscle (A–E) SOD1 mouse model. Expression relative to the average expression of the gene in all samples. 1M, 3M, and 4M indicate 1 month, 3 months, and 4 months, respectively. Analyses were conducted on three biological replicates and four technical replicates for each sample. (F–I) FUS mouse model. 3W and 4W indicate 3 weeks and 4 weeks. GAS, gastrocnemius; TA, tibialis anterior; TC, triceps; F, FUS mouse; C, control. Analyses were conducted on three biological replicates and two technical replicates for each sample. ∗p < 0.05, according to Mann-Whitney test.

Figure 8

Schematic description of proposed mechanisms The downmodulation of significant MRFs prevents the proper differentiation of hSOD1(G93A) myocytes. Particularly, miR-882 and -134-5p downmodulate Pax7, MyoD, Creb, and p38. hSOD1(WT) PCM secretes miR-26a-5p that impacts on the expression and functionality of Smad4. It inhibits muscle differentiation; therefore, its downmodulation may stimulate muscle differentiation. miR-26a, in concert with miR-431, affects the differentiation of motor neuron by increasing the legth of their neurites..