At A Glance
Name: Shiv Grewal
Position: Senior investigator, National Cancer Institute
Background: Associate professor, Cold Spring Harbor Laboratory — 2002-2003; Assistant professor, Cold Spring Harbor Laboratory — 1998-2002; Postdoc, National Cancer Institute — 1993-1998; PhD, molecular biology, University of Cambridge — 1992; BS, biology, Punjab Agricultural University —1986
After spending several years at Cold Spring Harbor Laboratory, an institution known for being home to groundbreaking RNAi research, Shiv Grewal moved on to the National Cancer Institute, where he is investigating the relationship between RNAi and heterochromatin. Recently, he spoke to RNAi News about his research.
Could you give an overview of what you do at NCI?
The main focus of my lab is to study higher-order chromatin assembly using the fission yeast Schizosaccharomyces pombe as a model system. There are basically two areas that we focus on: one is the role of histone-modifying enzymes and the chromatin proteins in the formation of heterochromatin. The second area is the role of small RNAs in heterochromatin formation.
Could you tell me about your research accomplishments?
We started working on heterochromatin assembly when I was still a postdoc; we identified several of the proteins involved in the modifications of histones as factors involved in heterochromatin assembly, and we also showed that centromeric repeat elements play an important role in the establishment of heterochromatic structures.
I continued to work on the role of histone-modifying activities and centromeric repeats in heterochromatin assembly after I moved to Cold Spring Harbor Laboratory as an independent research investigator. In 2001, we published two key papers where we worked out the sequence of molecular events with regard to histone modifications leading to the assembly of heterochromatic structures. Our studies revealed that histone-modifying enzymes such as deacetylases and methyltransferases cooperate with each other in setting up a specific histone-modification pattern that is recognized by factors involved in higher-order chromatin assembly such as heterochromatin protein 1 and others. Moreover, we showed that distinct site-specific histone H3 methylation patterns dictate the organization of chromosomes into discrete structural and functional domains. Histone H3 methylated at lysine 9 is strictly localized to silent heterochromatic regions whereas H3 methylated at lysine 4, only a few amino acids away, is specific to the surrounding active euchromatic regions.
We also showed in a paper published in Cell that what we call a gene in eukaryotic systems is not always just a piece of DNA — we showed that genes sometimes comprise DNA plus some of the chromatin proteins that are associated with it. So if you change the definition [of gene], it brings in a completely [new] dimension; simply move those protein sequences without altering the sequence of DNA and that will lead to certain changes in gene expression and obviously would have consequences including increased risk for diseases and phenotypic variations within populations.
More recently, we provided a key connection between heterochromatin assembly and RNAi, whereby double-stranded RNAs silence cognate genes. We started working on that around 1999. We had knocked out a gene called Argonaute in S. pombe in collaboration with another lab at Cold Spring Harbor Laboratory. We knocked it out … because we wanted to study its effect on stem cell lineage pattern of mating-type switching … but it didn’t have any effect on that. In the process I noticed, however, that there were some chromosome-segregation defects associated with the mutant. That later on led us to test if there were any heterochromatin-assembly problems at the centromeres, because centromeres are obviously important for proper segregation of chromosomes.
We tested that and found that there were, indeed, some problems with the heterochromatin at centromeres. That led to further testing using this technique called chromatin immunoprecipitation to find out if there are any defects in establishment of heterochromatin-specific histone modification patterns. A key observation was that deletion of the factors involved in RNAi such as Argonaute, Dicer, or RNA-dependent RNA polymerase in S. pombe resulted in defects in heterochromatin-specific histone modifications such as methylation of histone H3 at lysine 9, and the targeting of HP1 homologue Swi6 to heterochromatic loci. That work was chosen as breakthrough of the year by Science magazine, and later on led to the Newcomb Cleveland prize for our research group.
We have continued working on the role of small RNAs. In collaboration with a lab at Harvard Medical School, we identified a complex that directly linked small RNAs to heterochromatin assembly — that’s called RITS.
The RITS complex has an Argonaute protein, a protein now known to be associated with RNAi effector complexes, which is the one that initially we knocked out to study its effect on mating-type switching and later on for heterochromatin assembly, and two other proteins. One of them we knew already was associated with heterochromatin at centromeres — that’s called Chp1. The third one is called Tas3.
So there’s this protein complex of three proteins that also contains small RNAs generated by the RNAi pathway. That provided a direct link — here is a complex that we know is associated with centromeres and it has small RNAs. If it didn’t have small RNAs, it couldn’t associate with centromeres, and that led to a direct connection between small RNA and heterochromatin assembly.
Initially we said in our papers, and other people have shown in other systems, that RNAi and small RNAs recruit heterochromatin. But now, in the paper published by my lab in Nature Genetics recently, we showed that it’s a self-enforcing loop. Small RNAs target heterochromatin to centromeres, but then heterochromatin recruits the RNAi machinery to centromeres as well as other heterochromatic loci — RITS is stabilized at the heterochromatic loci via Chp1 chromodomain binding to histone H3 methylated at lysine 9. RITS then recruits RNA-dependent RNA polymerase and presumably even Dicer to heterochromatic loci to process rare transcripts produced by repeat elements that escaped heterochromatin-mediated transcriptional silencing. So heterochromatin marks such as histone H3 methylated at lysine 9 stabilize RITS at heterochromatic loci, RITS then recruits RNA-dependent RNA polymerase, and these proteins could sense transcripts produced by centromeric repeat elements, and then process them and convert them into siRNAs, which can then go back to target more heterochromatin. So it’s a cycle where small RNAs target heterochromatin; and heterochromatin stabilizes the RNAi machinery on chromosomes; and if there are any transcripts produced by repeats and transposons at heterochromatic loci, those transcripts get processed by the machinery that converts them into siRNAs that target more heterochromatin.
That self-enforcing loop is very important because any epigenetic phenomenon has to sustain itself. To support that model, we have a paper appearing in the Proceedings of the National Academy of Sciences showing that, consistent with this loop model, heterochromatin stabilizes all the other RNA components is true. We have tested the loop model by different combinations of mutations and it seems like it is, indeed, a self-enforcing loop.
There are a number of other things we are doing — for example, [one thing] my lab has also been trying to focus on is: once you initiate formation of heterochromatin in one place, heterochromatin has the ability to spread along chromosomes. Once it starts spreading, you have to find a way to block it from spreading into important genes on chromosomes. We found some elements, which are also present in other species, that we call heterochromatin boundary elements. We have an effort underway to find out about how these elements function.
Those are basically the different areas we’re working on … and converge on the common theme of how heterochromatin is initiated, how it is maintained, and how do you protect genes that are near heterochromatin blocks from the oppressive effects of heterochromatin.
Where do you see your research going?
In complex genomes, including our own genome, there are large portions that are made up of repeats and transposons. Those repeats and transposons threaten our genome as far as genomic integrity is concerned; they could recombine with each other and create deletions and rearrangements. So the role of small RNAs in forming heterochromatin at those repeats is a really important one because heterochromatin not only silences genes, it also has the ability to prohibit intra- or intrachromosomal recombination. That’s one area that excites me a lot — the role of the RNAi machinery in forming heterochromatin, and protecting the integrity of our genome. That is one area I, personally, feel is not fully explored at the moment. We have explored the role of RNAi and heterochromatin in the segregation of chromosomes, and that is that heterochromatin formation at centromeres is very important for proper segregation.
Now, going beyond the gene silencing, there’s a genomic integrity issue that is obviously very important as far as my assignment here at the National Cancer Institute is concerned — cancer is linked to a number of different forms of genomic instability.
Then there is this issue that a single small RNA could have homology to many different targets in the cell. We’re learning with microRNA work people are doing that you can control different genes using microRNAs, and similar things could be happening at the genome level. By that I mean that you could a small RNA-based regulatory network where a small RNA that is produced from a repeat or transposon at one site on a chromosome could target heterochromatin at several different locations in the genome and could silence them simultaneously. You could view it as an immune system — you produce this small RNA and search for where in the cell this element that is homologous to this small RNA might be present, and you could shut all those elements off. Moreover, low levels of small RNAs in cell might provide a “memory” system for silencing of similar elements in future.
From a broader prospective, that same mechanism could not only be used to silence transposons and repeats, but could be used to control normal gene expression in cells. So, the one area I would like to see labs explore in higher eukaryotes is: Do you see different profiles for these small RNAs as cells differentiate? Are there any small RNA-based regulatory networks in cells? Clearly, miRNAs are able to do this, but when I talk about regulatory networks, I’m thinking more at the genome level. You want to shut off several genes in a cell, and you could either use DNA-binding proteins to target repressive chromatin complexes… but you could also use these small RNAs. We know this machinery exists in cells, so why couldn’t it be used by evolution? Maybe nature has exploited this important tool to shut genes off when they are not needed during normal development.
A little off topic, but do you have any plans for the holidays coming up?
For the new year, I would like to spend a little time with my family. As you can very well imagine, the field we are working in is extremely hot, and one of the things that has happened during the last few years for me is the significant increase in number of invitations to come talk about our work. It’s important for me to tell other people working in other areas about the work, and there’s a lot of traveling involved in that. I rarely get a chance where I can have a significant block of time dedicated to the family. I have two kids and a wife, and I often feel there is not enough time in a day, and because of my traveling I don’t get a chance to sit down with them as much I should.
So over the next two weeks I have no travel planned, and I purposely scheduled this so that I can spend time with them.