Limits...
Linkages between changes in the 3D organization of the genome and transcription during myotube differentiation in vitro

View Article: PubMed Central - PubMed

ABSTRACT

Background: The spatial organization of eukaryotic genomes facilitates and reflects the underlying nuclear processes that are occurring in the cell. As such, the spatial organization of a genome represents a window on the genome biology that enables analysis of the nuclear regulatory processes that contribute to mammalian development.

Methods: In this study, Hi-C and RNA-seq were used to capture the genome organization and transcriptome in mouse muscle progenitor cells (C2C12 myoblasts) before and after differentiation to myotubes, in the presence or absence of the cytidine analogue AraC.

Results: We observed significant local and global developmental changes despite high levels of correlation between the myotubes and myoblast genomes. Notably, the genes that exhibited the greatest variation in transcript levels between the different developmental stages were predominately within the euchromatic compartment. There was significant re-structuring and changes in the expression of replication-dependent histone variants within the HIST1 locus. Finally, treating terminally differentiated myotubes with AraC resulted in additional changes to the transcriptome and 3D genome organization of sets of genes that were all involved in pyroptosis.

Conclusions: Collectively, our results provide evidence for muscle cell-specific responses to developmental and environmental stimuli mediated through a chromatin structure mechanism.

Electronic supplementary material: The online version of this article (doi:10.1186/s13395-017-0122-1) contains supplementary material, which is available to authorized users.

No MeSH data available.


Related in: MedlinePlus

Distribution of transcript levels of genes residing within the confines of the negatively correlated interacting regions in pairwise condition comparisons. Transcript levels (expressed as FPKM) of genes whose a TSSs had reduced PC1 values and b TSSs showed increases in PC1 values within negatively correlated interacting regions in myotubes vs myoblasts comparison. c Transcript levels of genes whose TSSs showed increases in PC1 values within negatively correlated interacting regions in AraC-treated myotubes vs myotubes comparison. Genes which had FPKM values equal to 0 in both conditions are excluded from the analyses; however, genes having FPKM value equal to 0 in one of the conditions and detectable FPKM value in the other condition are included (p values–paired Wilcoxon test, p < 0.001***, p = 0.01**, NS not significant). Outliers not plotted
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC5382473&req=5

Fig3: Distribution of transcript levels of genes residing within the confines of the negatively correlated interacting regions in pairwise condition comparisons. Transcript levels (expressed as FPKM) of genes whose a TSSs had reduced PC1 values and b TSSs showed increases in PC1 values within negatively correlated interacting regions in myotubes vs myoblasts comparison. c Transcript levels of genes whose TSSs showed increases in PC1 values within negatively correlated interacting regions in AraC-treated myotubes vs myotubes comparison. Genes which had FPKM values equal to 0 in both conditions are excluded from the analyses; however, genes having FPKM value equal to 0 in one of the conditions and detectable FPKM value in the other condition are included (p values–paired Wilcoxon test, p < 0.001***, p = 0.01**, NS not significant). Outliers not plotted

Mentions: Correlation analyses can be used to identify genomic regions that change their interaction profiles and not just their PC1 values [55]. Central to this analysis is the fact that negatively correlated regions (i.e. r < 0) interact with different partners in two conditions. Fifty-five genomic regions (400 kb) were negatively correlated following development of myotubes from myoblasts (Additional file 10). The transcription start sites (TSS) falling within the boundaries of these negatively correlated regions were identified using HOMER and divided into two groups: (1) TSSs whose PC1 values reduced from one condition to the next or (2) TSSs whose PC1 values increased from one condition to the next. The transcript levels of the genes that corresponded to the TSSs in each pool were determined from the transcription data. The differentiation of myoblasts to myotubes resulted in a significant decrease (p < 0.001 paired Wilcoxon test) in the transcript levels of 90 genes that were associated with 199 TSSs that had reduced PC1 values (Fig. 3a). In contrast, the 59 genes that were associated with the 112 TSSs that increased their PC1 values (became more open) also significantly increased their transcript levels (i.e. they were upregulated; p < 0.01 paired Wilcoxon test; Fig. 3b; Additional file 11). GO analysis identified enrichment for ‘nucleosome assembly’ and ‘chromatin assembly’ (p < 0.01; Additional file 1: Table S7) within 11 of the 90 downregulated genes which, with the exception of Cebpg (chr7:35046422–35056573), encoded histone proteins and were located within patch three of the HIST1 cluster on chr 13 (23,600,000–24,000,000 bp). Notably, the Hist1h2bc, Hist1h1c and Histih1a genes within patch three of HIST1 were upregulated upon myotube differentiation (Additional file 4 and Additional file 11). Treatment with AraC resulted in differential expression of specific replicative histone variant genes within the HIST1 cluster. The detection of transcripts that were not previously present (Additional file 1: Table S1) is consistent with the observed changes in expression occurring within the differentiated myotubes and not simply reflecting a change in the population structure (i.e. level of differentiation; Additional file 3: Figure S1). Consistent with the observations by Li et al. [67], the three patches within the HIST1 locus were spatially clustered (Additional file 3: Figure S10). Interestingly, there were only minimal differences in this inter HIST1 clustering between the myotubes and myoblasts indicating it was not responsible for the negative correlation for interactions in this region between the cell types.Fig. 3


Linkages between changes in the 3D organization of the genome and transcription during myotube differentiation in vitro
Distribution of transcript levels of genes residing within the confines of the negatively correlated interacting regions in pairwise condition comparisons. Transcript levels (expressed as FPKM) of genes whose a TSSs had reduced PC1 values and b TSSs showed increases in PC1 values within negatively correlated interacting regions in myotubes vs myoblasts comparison. c Transcript levels of genes whose TSSs showed increases in PC1 values within negatively correlated interacting regions in AraC-treated myotubes vs myotubes comparison. Genes which had FPKM values equal to 0 in both conditions are excluded from the analyses; however, genes having FPKM value equal to 0 in one of the conditions and detectable FPKM value in the other condition are included (p values–paired Wilcoxon test, p < 0.001***, p = 0.01**, NS not significant). Outliers not plotted
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC5382473&req=5

Fig3: Distribution of transcript levels of genes residing within the confines of the negatively correlated interacting regions in pairwise condition comparisons. Transcript levels (expressed as FPKM) of genes whose a TSSs had reduced PC1 values and b TSSs showed increases in PC1 values within negatively correlated interacting regions in myotubes vs myoblasts comparison. c Transcript levels of genes whose TSSs showed increases in PC1 values within negatively correlated interacting regions in AraC-treated myotubes vs myotubes comparison. Genes which had FPKM values equal to 0 in both conditions are excluded from the analyses; however, genes having FPKM value equal to 0 in one of the conditions and detectable FPKM value in the other condition are included (p values–paired Wilcoxon test, p < 0.001***, p = 0.01**, NS not significant). Outliers not plotted
Mentions: Correlation analyses can be used to identify genomic regions that change their interaction profiles and not just their PC1 values [55]. Central to this analysis is the fact that negatively correlated regions (i.e. r < 0) interact with different partners in two conditions. Fifty-five genomic regions (400 kb) were negatively correlated following development of myotubes from myoblasts (Additional file 10). The transcription start sites (TSS) falling within the boundaries of these negatively correlated regions were identified using HOMER and divided into two groups: (1) TSSs whose PC1 values reduced from one condition to the next or (2) TSSs whose PC1 values increased from one condition to the next. The transcript levels of the genes that corresponded to the TSSs in each pool were determined from the transcription data. The differentiation of myoblasts to myotubes resulted in a significant decrease (p < 0.001 paired Wilcoxon test) in the transcript levels of 90 genes that were associated with 199 TSSs that had reduced PC1 values (Fig. 3a). In contrast, the 59 genes that were associated with the 112 TSSs that increased their PC1 values (became more open) also significantly increased their transcript levels (i.e. they were upregulated; p < 0.01 paired Wilcoxon test; Fig. 3b; Additional file 11). GO analysis identified enrichment for ‘nucleosome assembly’ and ‘chromatin assembly’ (p < 0.01; Additional file 1: Table S7) within 11 of the 90 downregulated genes which, with the exception of Cebpg (chr7:35046422–35056573), encoded histone proteins and were located within patch three of the HIST1 cluster on chr 13 (23,600,000–24,000,000 bp). Notably, the Hist1h2bc, Hist1h1c and Histih1a genes within patch three of HIST1 were upregulated upon myotube differentiation (Additional file 4 and Additional file 11). Treatment with AraC resulted in differential expression of specific replicative histone variant genes within the HIST1 cluster. The detection of transcripts that were not previously present (Additional file 1: Table S1) is consistent with the observed changes in expression occurring within the differentiated myotubes and not simply reflecting a change in the population structure (i.e. level of differentiation; Additional file 3: Figure S1). Consistent with the observations by Li et al. [67], the three patches within the HIST1 locus were spatially clustered (Additional file 3: Figure S10). Interestingly, there were only minimal differences in this inter HIST1 clustering between the myotubes and myoblasts indicating it was not responsible for the negative correlation for interactions in this region between the cell types.Fig. 3

View Article: PubMed Central - PubMed

ABSTRACT

Background: The spatial organization of eukaryotic genomes facilitates and reflects the underlying nuclear processes that are occurring in the cell. As such, the spatial organization of a genome represents a window on the genome biology that enables analysis of the nuclear regulatory processes that contribute to mammalian development.

Methods: In this study, Hi-C and RNA-seq were used to capture the genome organization and transcriptome in mouse muscle progenitor cells (C2C12 myoblasts) before and after differentiation to myotubes, in the presence or absence of the cytidine analogue AraC.

Results: We observed significant local and global developmental changes despite high levels of correlation between the myotubes and myoblast genomes. Notably, the genes that exhibited the greatest variation in transcript levels between the different developmental stages were predominately within the euchromatic compartment. There was significant re-structuring and changes in the expression of replication-dependent histone variants within the HIST1 locus. Finally, treating terminally differentiated myotubes with AraC resulted in additional changes to the transcriptome and 3D genome organization of sets of genes that were all involved in pyroptosis.

Conclusions: Collectively, our results provide evidence for muscle cell-specific responses to developmental and environmental stimuli mediated through a chromatin structure mechanism.

Electronic supplementary material: The online version of this article (doi:10.1186/s13395-017-0122-1) contains supplementary material, which is available to authorized users.

No MeSH data available.


Related in: MedlinePlus