Professor, neuroscience research, University of California, Santa Barbara
• Professor, neurology/neuroscience, Harvard Medical School — 1996-2005
• Associate professor, neurology, Harvard Medical School — 1988-1996
• Postdoc, McLean Hospital — 1980-1982
• Resident, neurology, Tufts University — 1977-1980
• MD, Medical College of Pennsylvania — 1876
• BA/MA, English literature, Case Western Reserve University — 1972
Researchers from the University of California, Santa Barbara, last month reported in RNA on the discovery of the small, non-coding RNAs known as piwi-interacting RNAs in the central nervous system, particularly the hippocampus, of rodents.
While the piRNAs identified had been found by other groups, the team further demonstrated that suppression of one particular piRNA caused a change in the animals' spine morphology, suggesting that the small RNAs have a role beyond their well-studied restriction of transposable elements.
Gene Silencing News spoke with the study's senior author, Kenneth Kosik, about the findings.
Can we start with a little background on piwi-interacting RNAs?
They are another category of small, non-coding RNAs that tend to be a little larger than microRNAs. While microRNAs are 21 to 24 nucleotides, the piRNAs are 26 to 31 nucleotides. In contrast to microRNAs, for which we have a pretty clear idea about the biogenesis pathway, it's not fully understood how they are generated. Some of the details of the piRNA biogenesis pathway are just beginning to emerge.
You have experience with microRNAs. Is that what led to your interest in piRNAs?
Like many groups, one of the approaches that we take to study the entire world of these small RNAs is deep sequencing, in our case with the [Applied Biosystems] SOLiD sequencer, to sequence all of the small non-coding RNAs in a brain tissue sample. Quite frankly, we weren't even looking for piRNAs — we were really interested in microRNAs — but we saw in our sample some small RNAs that were a little larger than microRNAs and did not align with microRNAs. When we tried to align those sequences with the piRNA databases, there were many matches. This was quite a surprise to us.
Was that because these are mainly associated with the germline?
That's right. It was a surprise that we found them in brain tissue. However, ours is not the first report of piRNAs being associated with a somatic cell. There are other somatic cells in reproductive organs — the ovaries and testes — that are known to have piRNAs, as well.
Can you provide a snapshot of the paper and its findings?
We began with a sequencing run from the rodent hippocampus. In doing the analysis of the sequence tags, we found sequences that matched piRNAs. But we still wanted to do further validation to show that what we were looking at was the real thing.
We took some of the highest expressing sequences and confirmed them with Northern blots, by real time PCR, and by immunoprecipitation with a piwi antibody. We then showed that they can be detected in primary neurons in culture by in situ hybridization.
What are your thoughts on the implications of these findings?
When piRNAs are spoken about in the broadest terms, they have been called the “guardians of the genome.” The reason for that is because they appear to have some role in controlling transposons, and this is a very important function in the germline.
You could then ask whether the piRNAs we've found have a similar function. Many of the highly abundant piRNAs identified in germline cells align with transposons. The piRNAs that we found in brain align with intergenic sequences and are not associated with repetitive elements. That has suggested to us, and others before us who found similar types of piRNAs in somatic cells and reproductive organs, that there may be another function for piRNAs that is unrelated to transposons.
There is a little bit of work on this category of piRNAs — those that derive from intergenic regions rather than repetitive elements — suggesting they may have a role in silencing, or at least suppressing in some way, mRNA translation. This is still largely in the realm of hypothesis, and the mechanism involved may or may not bear some similarity to what microRNAs do. But that was the direction we followed as we tried to find some functions for a few of the high-expressing piRNAs that we observed.
In the paper, we developed some bioinformatic arguments based on alignments of the piRNA sequences with the 3' untranslated regions of some genes involved in synaptic spines and neurons. These imperfect matches were a little bit suggestive of a function related to translation, but by no means proof. The final experiment used hippocampal neurons from rodents. When we suppressed one of the piRNAs that was particularly abundant we observed significant changes in synaptic spine morphology. While we don't have any definitive proof that these piRNAs are affecting translation, the paper does contain some suggestions that point in that direction.
What about follow-up work?
We would like to follow up. There is a lot more to learn here. The question is how important are they and what is their functionality? That is one direction that we might take these findings.