May 01, 2014
When a character is asked to show his badge in the movie The Treasure of the Sierra Madre, he famously says something along the lines of, “Badges? We don’t need no stinkin’ badges!*” If histones in yeast heterochromatin could talk they might say something similar, except instead of badges they’d bring up modifications. Maybe something along the lines of, “Modifications? We don’t need no stinkin’ modifications for activation!” At least, they’d say this if a new study by Zhang and coworkers holds up.
In this study, the authors show that two different genes in the yeast S. cerevisiae are activated in heterochromatin in the absence of any significant changes to the surrounding chromatin. This result is surprising because most researchers think activation and changes in chromatin always go hand in hand. Apparently, in at least some situations they do not.
This isn’t to say that chromatin didn’t do anything here…it most certainly did. It served as a general damper on transcription. But in this study chromatin was by no means the major player; it had a relatively small influence on the levels of basal and activated gene expression. The authors suggest that this may be true for other genes in the more transcriptionally active euchromatin as well.
In the first set of experiments, Zhang and coworkers used a model system where the heat inducible gene HSP82 is flanked by the HMRE silencer from the HMR mating type cassette. These silencers cause a 30-fold reduction in transcription of this hsp82-2001 transgene.
Using chromatin immunoprecipitation (ChIP) the authors show that their transgene is indeed embedded in heterochromatin. They see a lot of Sir3p around the promoter, a high density of histones that lack any of the telltale modifications of euchromatin, and very little RNA polymerase II (Pol II) or the mRNA capping enzyme Cet1p around the promoter. These are all hallmarks of heterochromatin in yeast.
Things change when the yeast is subjected to heat shock. Consistent with the observed 200-fold increase in transcription, they suddenly see lots of Pol II and Cet1p around. But there is not a big change in the number of histones around the gene nor in their modifications.
When HSP82 is in its normal place in the genome, its activation is accompanied by specific acetylation and methylation of H3 and H4 histones. In heterochromatin, despite significant induction, there is none of this. The histones remain looking the same whether there is significant transcription or not.
One trivial explanation for this might be that the chromatin is unaffected because the levels of transcription are lower than normal. In other words, the lower final activity in the induced state is affecting histone modification.
Zhang and coworkers rule this out by using a TATA-less HSP82 gene in euchromatin and show that all the appropriate histone modifications still happen. This is true even though the damaged gene has 5-fold less activity compared with their transgene. The low level of transcription does not appear to explain activation in the absence of histone modification.
Of course another reason for this unexpected observation might be that this pretty artificial construct isn’t representative of natural genes. This doesn’t change the fact that its transcription is activated in the absence of histone modification, but it does question its relevance in the real world.
To address this issue, the authors looked for an inducible gene in natural heterochromatin and with a little bit of detective work, found the subtelomeric YFR057W gene. No one knows what this gene does, but a close look showed a possible Stb5p binding site in its promoter.
When Stb5p heterodimerizes with Pdr1p, the resulting dimer activates genes involved in pleiotropic drug resistance. Indeed the authors found that YFR057w was induced 150-fold with a small amount of cycloheximide. And when they used ChIP to compare the induced and uninduced states, they again found almost no changes in the chromatin around this gene despite an increase in the amount of Pol II and Cet1p.
Taken together these results suggest that activation doesn’t always have to come with chromosomal changes. Which, while a bit surprising today, wouldn’t have turned any researchers’ heads a few decades ago.
In the old days (1980’s and 1990’s), a lot of focus was on how transcriptional activators might affect the ability of Pol II to load onto the DNA and to pry it open and start transcribing. A lot of this was based on prokaryotic work where there really isn’t very much in the way of chromatin and a lot of activation depends on improving the ability of the polymerase to transcribe.
These days when people think about turning up a gene, they think about changing nearby chromatin. Various enzymes work to modify histones at specific places, which both loosens up the chromatin to allow access by Pol II and serves as a way for various coactivators to recognize the DNA.
As usual, reality is probably a combination of the two. Activators can activate transcription in lots of different ways, some of which probably include chromatin changes while in others chromatin changes are simply a consequence of activation. Not all transcription activation needs stinkin’ histone modifications.
* This is actually a misquote that may have come from the Mel Brooks film Blazing Saddles.
by D. Barry Starr, Ph.D., Director of Outreach Activities, Stanford Genetics
Categories: Research Spotlight
Tags: heterochromatin , histone modification , Saccharomyces cerevisiae , transcription