Using their new NET-seq technique in S. cerevisiae, Howard Hughes Medical Institute researchers recently found a new role for histone acetylation that adds yet another layer of complexity to RNA transcription regulation—local acetyl modifications apparently allow RNA polymerase to make anti-sense transcripts upstream of a gene, while the deacetylating enzyme complex Rpd3S keeps transcription going in the sense direction.
Along with previous findings, the experiments demonstrate that “the primary function of the Rpd3S histone deacetylase complex seems to be to enforce promoter directionality,” write HHMI’s Stirling Churchman and Jonathan Weissman in the Jan. 20 Nature. The paper also provides in vivo support of earlier observations that RNA polymerase tends to pause at histones along a DNA template, giving cellular machinery yet another opportunity to regulate the whole process.
Using the NET-seq technique, which involves halting RNA transcription and deep-sequencing nascent RNA molecules, the researchers sequenced transcripts of about 500 b.p. originating from promoters between tandem genes. When transcription goes “backward” into an upstream gene, that antisense RNA is easy to distinguish from the upstream gene’s sense transcript, the authors say.
For most of the promoters, NET-seq uncovered about eight times as many sense transcripts as antisense transcripts. And about 80 percent of these promoters showed a sense-to-antisense ratio of at least three to one.
Observing a strong correlation between H4 histone-protein acetylation and local antisense transcript levels, the Churchman-Weissman team decided to see what happens when they disable Rpd3S—an H4 deacetylation complex—by removing its gene from an experimental strain of yeast.
The levels of antisense transcripts in this deletion strain went through the roof, with NET-seq detecting about four times as many such transcripts as in wild-type yeast. The same thing happened when the team disabled a different Rpd3S gene, and neither situation resulted in higher levels of sense transcripts.
That finding has broad implications for transcriptional regulation, since sense transcription must be dependent upon histone acetylation, Rpd3S deacetylation, Rpd3S recruitment, and so on. Also, nobody fully understands the role of antisense RNA produced this way, or if it even has a role.
When Science in December 2008 published four papers reporting “divergent” eukaryotic transcription—sense and antisense RNA originating from the same promoters—it surprised even Harvard Medical School’s Stephen Buratowski, who wrote commentary on those findings for the journal. “It remains to be seen whether these RNAs have a function, but their existence challenges our simplistic models about how the DNA sequences known as ‘promoters’ define transcription start sites,” said Buratowski in that issue.
But it’s been difficult for transcription researchers to hone in on events at the 3’ end of nascent RNA molecules—partly due to low-resolution methods, and partly due to the rapid breakdown by cellular mechanisms of short RNAs while they’re being made or shortly thereafter.
NET-seq seems to be able to break through all that by literally freezing the complex of RNA polymerase, DNA, and nascent RNA in place. Using the method, which is short for “native elongating transcript sequencing,” researchers freeze active cells in liquid nitrogen, cryogenically lyse them, destroy the DNA, and immunoprecipitate the polymerase-DNA-RNA complex out of solution. They then use conventional methods to create a cDNA library from all the released RNA molecules.
In Churchman’s and Weissman’s work, deep-sequenced nascent RNAs reveal 3’ ends located deep in introns and far beyond poly-A tails. Along with other pieces of evidence, that suggests that the method works well to protect RNA until it can be analyzed down to the single-nucleotide level—and it reveals the exact position of the RNA polymerase at freezing time.
Regarding RNA polymerase pausing in vivo, Churchman and Weissman write that in metagene analysis, NET-seq reveals a higher density of polymerase within the first 700 bases of transcription, which is consistent with previous observations using lower-resolution nuclear run-on methods. With a large number of sequenced RNA molecules, metagene analysis shows peaks at nucleotides where polymerases sat when their host cells hit the liquid N2.
“The high density of pauses was not an artefact of library generation and sequencing biases, as we detected tenfold fewer spikes in data from messenger RNA lightly fragmented by alkaline hydrolysis,” the authors write.
Once the investigators compensated for polymerase backtracking, they found that the prime RNA polymerase pause sites corresponded neatly with the middle of nucleosome dyads early in a gene. That finding lines up neatly with previous in vitro evidence gleaned through optical trap methods.