Transcriptomic characterization of the molecular mechanisms induced by RGMa during skeletal muscle nuclei accretion and hypertrophy

Abstract

Background: The repulsive guidance molecule a (RGMa) is a GPI-anchor axon guidance molecule first found to play important roles during neuronal development. RGMa expression patterns and signaling pathways via Neogenin and/or as BMP coreceptors indicated that this axon guidance molecule could also be working in other processes and diseases, including during myogenesis. Previous works from our research group have consistently shown that RGMa is expressed in skeletal muscle cells and that its overexpression induces both nuclei accretion and hypertrophy in muscle cell lineages. However, the cellular components and molecular mechanisms induced by RGMa during the differentiation of skeletal muscle cells are poorly understood. In this work, the global transcription expression profile of RGMa-treated C2C12 myoblasts during the differentiation stage, obtained by RNA-seq, were reported.

Results: RGMa treatment could modulate the expression pattern of 2,195 transcripts in C2C12 skeletal muscle, with 943 upregulated and 1,252 downregulated. Among them, RGMa interfered with the expression of several RNA types, including categories related to the regulation of RNA splicing and degradation. The data also suggested that nuclei accretion induced by RGMa could be due to their capacity to induce the expression of transcripts related to 'adherens junsctions' and 'extracellular-cell adhesion', while RGMa effects on muscle hypertrophy might be due to (i) the activation of the mTOR-Akt independent axis and (ii) the regulation of the expression of transcripts related to atrophy. Finally, RGMa induced the expression of transcripts that encode skeletal muscle structural proteins, especially from sarcolemma and also those associated with striated muscle cell differentiation.

Conclusions: These results provide comprehensive knowledge of skeletal muscle transcript changes and pathways in response to RGMa.

Keywords: Axon Guidance; Hyperplasia; Hypertrophy; Myogenesis; Skeletal muscle differentiation; Transcriptomic analysis.

Fig. 1

Quality and transcriptomic profile of RGMa-treated myoblasts during myogenic differentiation. A Experimental design. B Pearson Correlation Coefficient (PCC) analysis of normalized read-counts denoted a high internal consistency and reproducibility of treated and control replicates. C MA plot analysis showing the RNA-seq profile of the log₂ (fold change) distributions of all DETs in the average of normalised counts. Each point represents one transcript. Those dots marked in blue were detected as differentially expressed at a 5% FDR with log₂(FC) > 0 (upregulated) and log₂(FC) < 0 (downregulated) after RGMa treatment. Transcripts with similar expression levels are represented around the horizontal line (y = 0). Dots that are outside the window are plotted as triangles. D Heatmap analysis of DET with muscle-associated terms (‘cellular component,’ ‘biological process,’ and ‘molecular function’) of Gene Ontology (GO). Transcripts with the lowest expression values are marked in red, median expression values in black, and the highest expression values in green

Fig. 2

RNA biotypes modulated by RGMa treatment. RGMa could modulate the differential expression of 13 RNA biotypes, classified in six RNA categories according to Ensembl (https://m.ensembl.org/info/genome/genebuild/biotypes.html): (1) protein coding, (2) processed transcripts (lncRNA: antisense, bidirection-promoter-lncRNA, lincRNA, retained intron and ncRNA: snRNA and Mt-rRNA), (3) nonsense mediated decay, (4) pseudogenes (processed-pseudogenes, transcribed-processed-pseudogene, and unprocessed-pseudogene), and (5) Tec (to be experimentally confirmed)

Fig. 3

Functional analysis of the non-protein coding RNA differentially regulated by RGMa treatment. For this analysis, we considered upregulated DETs that do not encode proteins. Pie analysis of the GO enrichment, showing the most frequent terms, including cellular component, biological process, molecular function, and immune system process, and KEGG GO terms that were A upregulated and B downregulated. The right-sided hypergeometric test was used in statistical inference, and the Benjamini–Hochberg method was applied for a p-value correlation (p < 0.05). The analysis was conducted using the plugin ClueGO (v.2.5.4) for Cytoscape (v3.7.1)

Fig. 4

Functional analysis of the protein coding RNA upregulated by RGMa. For this analysis, we considered the DETs that encode proteins that were found to be upregulated (FC > 1) by the treatment with RGMa, compared to the control. A-C Pie chart analysis of the three GO categories used to classify the upregulated protein coding transcripts. The right-sided hypergeometric test was used in statistical inference, and the Benjamini–Hochberg method was applied for a p-value correlation (p < 0.0001). The analysis was conducted using the plugin ClueGO (v.2.5.4) for Cytoscape (v3.7.1)

Fig. 5

Functional analysis of the protein coding RNA downregulated by RGMa. For this analysis, we considered the DETs that encode proteins and were found to be downregulated (FC < 1) in the RGMa treated group, compared to the control one. A-C Pie chart analysis of the three GO categories for downregulated DETs. D Functionally grouped network of enriched categories for expressed transcripts, annotated for ‘biological process,’ ‘cellular component,’ and ‘molecular function’ GO terms. The right-sided hypergeometric test was used in statistical inference, and the Benjamini–Hochberg method was applied for a p-value correlation (p < 0.001). The analysis was conducted using the plugin ClueGO (v.2.5.4) for Cytoscape (v3.7.1)

Fig. 6

Muclei accretion and muscle-related enriched terms from the functional analysis of all DETs in response to RGMa. GO enrichment and the network analysis of DETs was performed using the software ClueGO. Terms were selected for network analysis related to nuclei accretion A and to muscle differentiation and structure B. The right-sided hypergeometric test was used in statistical inference, and the Benjamini–Hochberg method was applied for a p-value correlation (p < 0.001). The network was designed using the ForceAtlas2 algorithm and node size represents network centrality which was calculated using Eigenvector Centrality algorithm

Fig. 7

Validation of the RNA-seq expression profiles by qPCR. A subset of twelve DETs that were upregulated and downregulated by RGMa treatment during muscle differentiation were used to validate the obtained RNA-seq expression data. Transcripts were selected by their expression and their known association with muscle hyperplasic or hypertrophic phenotypes. Expression patterns indicate agreement between the two methods and *, significance of p-adj < 0.05