As RA target genes need to be associated with a RARE, the DNA sequences within the RA-regulated H3K27ac/H3K27me3 ChIP-seq peaks we found near our list of 93 RA-regulated genes were searched for RARE sequences using the Homer transcription factor binding site program for the mm10 genome; we searched for 3 types of RAREs including those with a 6-bp direct repeat separated by either 5 bp (DR5), 2 bp (DR2), or 1 bp (DR1) [5], and the presence or absence of RAREs is summarized (S1 and S2 Tables). We found that 46 of these 93 genes contained at least 1 RARE in their nearby RA-regulated H3K27ac and/or H3K27me3 ChIP-seq peaks, thus narrowing down our list of candidate RA target genes to 49% of the genes originally identified. Our approach identified the 3 RAREs previously shown to have required functions during trunk development in vivo by knockout studies (RAREs for Hoxa1, Cdx1, Fgf8) plus several RAREs associated with known RA-regulated genes in the E8.5 trunk from Aldh1a2-/- studies (Rarb, Crabp2, Sox2, Dhrs3, Cdx2, Fst), thus validating our approach for identifying RA-regulated genes required for trunk development. The sequences of all the RAREs near these 46 RA target genes here are summarized; included are 65 RAREs near 34 RA-activated genes (we refer to these as RARE enhancers associated with increased H3K27ac and/or decreased H3K27me3 in the presence of RA) and 20 RAREs near 12 RA-repressed genes (we refer to these as RARE silencers associated with increased H3K27me3 and/or decreased H3K27ac in the presence of RA) (S3 Table).
The results here provide evidence that many of the RA-regulated H3K27ac and H3K27me3 marks are associated with regulation of the nearest genes. However, it is possible that some H3K27ac and H3K27me3 RA-regulated peaks may be related to RA-regulated genes located further away in the same topologically associated domain (TAD). In order to address this issue, we assigned each RA-regulated H3K27ac and H3K27me3 peak to a TAD using the 3D Genome Browser ( -c/view.php); TAD analysis has not been performed on mouse E8.5 trunk tissue, but as TAD domains are similar between different tissues [22], we used the TAD database for mouse embryonic stem (ES) cells, which is the closest biologically relevant database in the 3D Genome Browser. Then, the genes in each TAD containing an RA-regulated peak were searched in our RNA-seq database to identify genes whose mRNA levels are decreased or increased when RA is lost, and if at least 1 gene was found, we determined whether a RARE is present in the ChIP-seq peak. This analysis resulted in the identification of 82 additional RARE enhancers near RA-activated genes and 40 additional RARE silencers near RA-repressed genes, where the gene is not the gene nearest to the RARE in the TAD; in some cases, more than 1 RA-regulated gene was identified in a TAD (S3 Table).
The Silencers - So Be It 1995.rar
Although it possible that some RAREs that are conserved only in mammals perform mammal-specific functions, RAREs that are conserved from mammals to birds or lower may play fundamental roles in regulation of target genes. The candidate RARE enhancers and RARE silencers we identified here that are associated with RA-regulated epigenetic marks were searched for evolutionary conservation using the University of California Santa Cruz (UCSC) genome browser. Among the RAREs in which the nearest gene is RA-regulated, we found 6 RAREs that are conserved from mouse to zebrafish, 11 conserved to frog (Xenopus tropicalis), 18 conserved to reptile (lizard, painted turtle), 20 conserved to bird (chicken, turkey), 39 conserved to human, 65 conserved to rodent (rat), and 20 that are not conserved with rat (S3 Table). The large number of RAREs (i.e., 20) conserved beyond mammals to bird, lizard, frog, or fish demonstrate that our approach is able to identify highly conserved RAREs that point to excellent candidate genes required for development. Among the additional RAREs we found located farther away in the TAD from an RA-regulated gene, we identified only 4 more RAREs conserved beyond mammals to bird, lizard, frog, or fish, thus bringing the total to 24 highly conserved RAREs (S3 Table). Thus, most of the highly conserved RAREs we identified are located very close to an RA-regulated gene rather than further distant in the TAD. In addition, all these highly conserved RAREs are either identical to the RARE consensus or have only 1 mismatch. Here we summarize the 24 most highly conserved RAREs that point to 38 RA-regulated genes that may be required for development (Table 2).
Our epigenetic ChIP-seq studies combined with RNA-seq on wild-type versus Aldh1a2-/- RA-deficient trunk tissue provides a means for identifying new candidate RA target genes that may be required for development. By focusing on RA-regulated genes that also have changes in nearby RA-regulated H3K27ac and/or H3K27me3 epigenetic marks associated with highly conserved RARE enhancers or silencers, our approach can be used to identify excellent candidates for gene knockout studies to learn more about gene function. 2ff7e9595c
Comments