NAME: Alex Stark
POSITION: Postdoc, Kellis lab, Massachusetts Institute of Technology
BACKGROUND: Postdoc, EMBL, Heidelberg — 2004-2005; PhD, biocomputing, University of Koln — 2004; BS, biochemistry, University of Tubingen — 2000
Earlier this month, a team of researchers from the Broad Institute of the Massachusetts Institute of Technology and Harvard University published a study in Genes & Development showing that the miRNA iab-4 locus in the Drosophila Hox cluster is transcribed convergently from both strands of certain DNA, giving rise to two distinct and functional miRNAs.
The paper also describes the existence of these so-called sense and antisense miRNAs in the mouse and antisense transcripts close to miRNA in both flies and mammals, indicating that more of these pairs exist in mammals.
This week, RNAi News spoke with Alex Stark, the paper’s first author, about the work.
Let’s start with the lab you work in and the work you do there.
I am very interested in gene regulation in general — both transcriptional regulation and post-transcriptional regulation, the latter mostly by microRNAs.
The [Manolis] Kellis lab [at MIT’s Computer Science and Artificial Intelligence Laboratory] is using comparative genomics as a tool to find and understand functional parts in genome sequences based on the fact that functional parts evolve differently, showing different patterns of sequence similarity across different but related species.
Pouya Kheradpour, Manolis Kellis — both of whom are co-authors on the [Genes & Development] paper — and I have been working together for about two years on understanding how small binding sites for either transcription factors or microRNAs evolve across the Drosophila lineage … and how we can use [that knowledge] to find functional instances of those.
In addition, we’re looking at whether we can find the regulators themselves, specifically for microRNAs. By looking across different species, can we find novel microRNAs? This is how we came across the fact that some of the microRNA hairpins show surprisingly symmetrical scores if you look at them in the sense and antisense transcribed way.
Can you talk a bit about the paper and the key findings?
The key findings were that, for several fly microRNAs, the hairpin that the precursor sequence adopts is very symmetrical, whether you take the actual transcribed strand or what was previously known to be transcribed and gives rise to the mature microRNA and its reverse complement — that would be the sequence that would arise if you transcribed the locus on the opposite DNA strand.
That led us to propose that if there was antisense transcription of these loci, we would also find processing of this antisense transcript into functional, mature microRNAs. This [microRNA] iab-4 locus [in the Drosophila Hox cluster] has been known for many years to be transcribed in both directions in a non-overlapping pattern in the fly embryo.
In collaboration with the [David] Bartel lab [at the Massachusetts Institute of Technology] and the [Greg] Hannon lab [at Cold Spring Harbor Laboratory], we found processed microRNAs that specifically came from this antisense transcript. We went on to look for potential targets of this antisense microRNA and … found that this antisense microRNA also regulated Hox genes. processed microRNAs that specifically came from this antisense transcript. We went on to look for potential targets of this antisense microRNA and … found that this antisense microRNA also regulated Hox genes.
This is quite an interesting pattern because it is known that iab-4, which is the sense microRNA of that locus, also regulates Hox genes, as do other Hox microRNAs. So finding that this antisense microRNA regulates Hox genes immediately suggested that there is something going on in vivo.
In collaboration with the Steven Cohen group at EMBL in Heidelberg, we confirmed this expression pattern in the embryo and were able to show that if we topically express this microRNA in places where it is not usually expressed, we get this homeotic transformation where the haltere balancing organ transforms to a wing — and that is a very typical homeotic transformation indicative of a reduction in Ubx activity.
That basically not only established [iab-4] as the first antisense microRNA, but also as a functional microRNA and a new Hox gene.
What are the implications of the existence of sense and antisense microRNAs?
This is very intriguing because, based on work by other labs, for example in yeast, we know that sense and antisense transcription of the same locus can block each other by transcriptional interference. You can imagine that this creates a situation where the two microRNAs are likely expressed in non-overlapping domains more generally.
We saw that for the iab-4 example, but the same logic and the same interference mechanism should work for other such pairs so that nature can now quite easily, by regulating those microRNAs independently and transcribing them convergently, create non-overlapping expression domains. If you imagine that the two microRNAs regulate different genes, which they almost certainly would given their slightly different sequences, you have a very attractive means of establishing and maintaining different expression domains.
Given the fact that we've found more examples in the fly, as well as in the mouse, we know this is probably a global phenomenon in animals.
So you’ve seen the existence of these in mammals?
Yes. Using small RNA sequencing data from Greg Hannon we found antisense microRNAs for known mouse microRNAs. We have cloned the first 10 antisense microRNAs, of which eight are very strongly supported, immediately extending this finding to mammals.
Are there plans to follow up on this work?
We don’t have experiments running at this time, but what we’re doing computationally is following up on these and looking, for example, if we can find functions or convincing target genes for those microRNAs.