In recent years, it has become evident that substantial portions of mammalian genomes are actively transcribed as non-coding RNA, including thousands of cis
-natural antisense transcripts (cis
-NATs are transcripts produced from within the protein-coding loci, but from the opposite strand, and are thus complementary to the sense mRNA transcript (a). Cis
-NATs may play important regulatory roles via transcriptional interference caused by collisions of RNA polymerase complexes moving in opposite directions across the same locus (3
) or through the formation of double-stranded RNA leading to post-transcriptional silencing through RNA interference (7
). However, the extent to which non-coding RNAs in general, and cis
-NATs in particular are biologically functional remains a matter of debate. Some studies have suggested that the majority of non-coding RNA transcripts are non-functional and simply represent transcriptional noise (9
), while others have found evidence in support of the function for numerous non-coding RNAs (11–13
Figure 1. Delineation and analysis of cis-NAT promoters. Cis-NATs are initiated from protein-coding gene loci and transcribed in the opposite (antisense) direction. (a) Example of a protein-coding gene locus with a genic promoter that drives transcription in the (more ...)
Previously, investigators have interrogated the functional potential of novel non-coding RNA transcripts by evaluating the chromatin environment in-and-around their promoters (12
). These studies were motivated by the fact that the promoters of well-characterized human genes have characteristic chromatin properties, including distinct protein binding and histone-modification profiles, and these particular chromatin environments give indications as to the biological mechanisms, both genetic and epigenetic, by which the genes are regulated (12
). For example, chromatin immunoprecipitation (ChIP-seq) studies have revealed that the promoters of actively transcribed genes are occupied by RNA Pol II and marked with a suite of specific histone tail modifications, such as acetylation of the lysine at position 9 of histone H3 (H3K9Ac) (16–18
), whereas silent gene promoters are depleted for RNA Pol II and enriched for known repressive modifications such as trimethylation of lysine 27 of histone H3 (H3K27Me3). On the other hand, it has been shown that the promoters of many novel non-coding transcripts that have been characterized by high-throughput sequencing methods, but for which there is no additional supporting information, do not show enrichment for histone modifications or an active chromatin environment (12
). Thus, chromatin can be used to discriminate between the promoters of actively regulated genes versus putative transcription start site (TSS) that probably represents transcriptional noise.
In this study, we evaluated the chromatin environment surrounding hundreds of thousands of human cis-NATs across six different ENCODE cell types for 10 RNA isolation conditions. We sought to establish whether or not cis-NAT promoters show patterns of activity and chromatin modifications that are consistent with epigenetic regulation. We found that active cis-NAT promoters are enriched with active histone modifications and occupied by RNA Pol II, whereas silent cis-NAT promoters are depleted for both active modifications and RNA Pol II and enriched for the repressive modification of H3K27Me3. These data provide evidence for the epigenetic regulation of numerous human cis-NATs, presumably a prerequisite of their potential function as gene regulators.