Ageing affects DNA methylation drift and transcriptional cell-to-cell variability in muscle stem cells

Age-related tissue alterations have been associated with a decline in stem cell number and function1. Although increased cell-to-cell variability in transcription or epigenetic marks has been proposed to be a major hallmark of ageing2–5, little is known about the molecular diversity of stem cells during ageing. Here, by combined single-cell transcriptome and DNA methylome profiling in mouse muscle stem cells, we show a striking global increase of uncoordinated transcriptional heterogeneity together with context-dependent alterations of DNA methylation with age. Importantly, promoters with increased methylation heterogeneity are associated with increased transcriptional heterogeneity of the genes they drive. Notably, old cells that change the most with age reveal alterations in the transcription of genes regulating cell-niche interactions. These results indicate that epigenetic drift, by accumulation of stochastic DNA methylation changes in promoters, is a substantial driver of the degradation of coherent transcriptional networks with consequent stem cell functional decline during ageing.

Interestingly, we observed a negative correlation between changes in methylation levels 138 and changes in methylation heterogeneity (Promoters: Pearson's coefficient= -0.35, P < 2.2e-139 16, Fig. 4B). Regions becoming more homogeneous showed an increase in methylation, 140 suggesting that de novo methylation enzymes (Dnmt3a,b) are recruited to specific sites and add 141 methylation in a coordinated manner between cells. In contrast, regions becoming more 142 heterogeneous showed a decrease in their methylation levels. Despite the low proliferative 143 history of these cells, this pattern could reflect errors in DNA methylation maintenance during 144 DNA replication, or an active demethylation mechanism via TET enzymes (Fig. S5).

145
Epigenetic changes may contribute to the age-associated pattern of transcriptional 146 heterogeneity. To explore this possibility, we analysed the association between promoter DNA 147 methylation and gene expression. We calculated a correlation coefficient for each cell and 148 confirmed the expected negative correlation for methylation and transcription (Fig. 4C). 149 Interestingly, old cells that were most transcriptionally different from young cells showed 150 lower levels of correlation (Mann-Whitney-Wilcoxon test; P < 0.05, Fig. 4C). Furthermore, we 151 calculated changes in transcriptional variability between young and old cells (see Methods) 152 and observed that promoters with increased methylation heterogeneity tended to have increased 153 transcriptional heterogeneity (Mann-Whitney-Wilcoxon test; P <0.001) (Fig. 4D). It appears 154 therefore that deterioration of transcriptional coherence during ageing is associated with increased promoter methylation heterogeneity and with decreased connectivity between the 156 epigenome and the transcriptome. 157 In summary, we report transcriptional and epigenetic signatures associated with ageing 158 in a deeply quiescent population of muscle stem cells. Previous studies have investigated 159 transcriptional heterogeneity changes with age in mixed cell populations 4 which are affected 160 by differences in cellular composition, such as an increase in senescent cells 4 . In contrast, our 161 study is focused on a specific population of cells in which known stemness, activation and 162 senescent markers were not affected by ageing. Even in this restricted population, we observe 163 a global increase of uncoordinated transcriptional variability with age, indicating an intrinsic 164 mechanism of cellular ageing. Interestingly, mouse muscle stem cells were shown to maintain 165 clonal diversity during homeostatic ageing by lineage-tracing 18 , however, our study uncovers 166 a dramatic underlying molecular heterogeneity in these stem cells that extends beyond 167 maintenance of clonal homogeneity. We also observe that cells that have acquired more 168 differences with age showed alterations in multiple extracellular matrix related genes 169 potentially affecting cell-niche interactions.

170
Elevated transcriptional variability with age has been reported in several studies 2-4 , 171 however the underlying causes remain largely unknown. The accumulation of somatic 172 mutations only partially accounts for the increased cell-to-cell transcriptional variability 4 , 173 suggesting that epigenetic mechanisms might be a contributing factor 5 . In this study, by 174 applying for the first time a combined single cell method for DNA methylation and the 175 transcriptome, we show that epigenetic drift, or the uncoordinated accumulation of methylation 176 changes in promoters, contributes to the increased transcriptional variability with age (Fig. 4E).

177
Due to the deep quiescent state of the homeostatic cells chosen for study, our data highlight the 178 possibility that the observed epigenetic patterns could be independent of extensive cell 179 proliferation. We propose that this variability is detrimental due to uncoordinated transcription, thereby affecting the ability of stem cells to maintain quiescence or activate coherently upon 181 injury. Future studies of different stem cell populations integrating multiple layers of molecular 182 information will be highly informative for a more complete understanding of the underlying 183 molecular mechanisms of ageing and age-related diseases.
Animals were handled according to national and European Community guidelines, and an 187 ethics committee of the Institut Pasteur (CETEA) in France approved protocols. Young (2 188 months-old) and old (24 months-old) Tg:Pax7-nGFP 21 mice were used in this study.   We prepared scM&T-seq libraries 10 by isolating mRNA on magnetic beads and separating from expression (SCDE) was used to calculate differential expression analysis between young and 235 old cells (Table S1) 35 . 236 Cell-to-cell correlation analyses were performed using the top 500 most variable genes 237 within each individual and using Spearman's correlation as the measure of similarity between  (Table S4) were used for GO enrichment analysis 36 . 249 250 DNA-methylome 251 We discarded cells that had less than 1 million paired-end alignments or less than 500,000 CpG 252 sites covered (Fig. S1). To avoid biases that might occur due to different sequencing depths or 253 number of cells between individuals, we down-sampled the data to 1 million reads for each cell Here p is the frequency of pairs of methylation values. 277 We validated our approach by applying the method in simulated data with increasing 278 levels of methylation heterogeneity (Fig. S3). We also observed that our algorithm is highly 279 robust to missing data (Fig. S3). 280 We applied this method across multiple genomic regions for each individual

Isolation of satellite cells
independently and then computed the average of young and old samples. Pairwise comparisons 282 with fewer than 4 CpG sites were not considered in the analysis. Furthermore, to avoid 283 misinterpretations because of poor coverage depth we excluded regions with: i) less than 284 20CpG sites, ii) less than an average of 2 CpG sites covered per cell, iii) less than 100 cell-to-285 cell pairwise comparisons. We also excluded regions with high coverage differences between 286 ages (more than an average of 10 CpG sites or more than 200 cell-to-cell pairwise  Table S5). Same approach was used to calculate differences between 298 young and old transcriptional heterogeneity (mean distance to the median) ( Fig S9 and