Abstract
Abstract 2367
Hematopoietic stem cell (HSC) Aging is a complex process linked to number of changes in gene expression and functional decline of self-renewal and differentiation potential. While epigenetic changes have been implicated in HSC aging, little direct evidence has been generated. DNA methylation is one of the major underlying mechanisms associated with the regulation of gene expression, but changes in DNA methylation patterns with HSC aging have not been characterized. We hypothesize that revealing the genome-wide DNA methylation and transcriptome signatures will lead to a greater understanding of HSC aging. Here, we report the first genome-scale study of epigenomic dynamics during normal mouse HSC aging. We isolated SP-KSL-CD150+ HSC populations from 4, 12, 24 month-old mouse bone marrow and carried out genome-wide reduced representative bisulfite sequencing (RRBS) and identified aging-associated differentially methylated CpGs. Three biological samples were sequenced from each aging group and we obtained 30–40 million high-quality reads with over 30X total coverage on ∼1.1M CpG sites which gives us adequate statistical power to infer methylation ratios. Bisulfite conversion rate of non-CpG cytosines was >99%. We analyzed a variety of genomic features to find that CpG island promoters, gene bodies, 5'UTRs, and 3'UTRs generally were associated with hypermethylation in aging HSCs. Overall, out of 1,777 differentially methylated CpGs, 92.8% showed age-related hypermethylation and 7.2% showed age-related hypomethylation. Gene ontology analyses have revealed that differentially methylated CpGs were significantly enriched near genes associated with alternative splicing, DNA binding, RNA-binding, transcription regulation, Wnt signaling and pathways in cancer. Most interestingly, over 579 splice variants were detected as candidates for age-related hypermethylation (86%) and hypomethylation (14%) including Dnmt3a, Runx1, Pbx1 and Cdkn2a. To quantify differentially expressed RNA-transcripts across the entire transcriptome, we performed RNA-seq and analyzed exon arrays. The Spearman's correlation between two different methods was good (r=0.80). From exon arrays, we identified 586 genes that were down regulated and 363 gene were up regulated with aging (p<0.001). Most interestingly, overall expression of DNA methyl transferases Dnmt1, Dnmt3a, Dnmt3b were down regulated with aging. We also found that Dnmt3a2, the short isoform of Dnmt3a, which lacks the N-terminal region of Dnmt3a and represents the major isoform in ES cells, is more expressed in young HSC. For the RNA-seq analysis, we focused first on annotated transcripts derived from cloned mRNAs and we found 307 genes were down regulated and 1015 gene were up regulated with aging (p<0.05). Secondly, we sought to identify differentially expressed isoforms and also novel transcribed regions (antisense and novel genes). To characterize the genes showing differential regulation, we analyzed their functional associations and observed that the highest scoring annotation cluster was enriched in genes associated with translation, the immune network and hematopoietic cell lineage. We expect that the results of these experiments will reveal the global effect of DNA methylation on transcript stability and the translational state of target genes. Our findings will lend insight into the molecular mechanisms responsible for the pathologic changes associated with aging in HSCs.
Disclosures:
No relevant conflicts of interest to declare.
Differential DNA Methylation and Gene Expression Between ALV-J-Positive and ALV-J-Negative Chickens
