Assistant professor, computational biology, University of Pittsburgh School of Medicine & Cancer Institute
• Research fellow, Memorial Sloan-Kettering Cancer Center — 2003-2005
• Graduate fellow, Rockefeller University — 2000-2003
• PhD, biomedical science, Rockefeller University — 2003
• MSc, physical chemistry, Indian Institute of Technology — 2000
Earlier this month, University of Pittsburgh School of Medicine researcher Bino John and colleagues reported in the Journal of Virology on their discovery that a so-called unusually small RNA — around 17 nucleotides long — can down-regulate a microRNA target, suggesting a gene-regulation role for the RNA.
The team further found that "human miRNA-derived usRNAs preferentially match to 5' ends of miRNAs, and are … more likely to associate with the siRNA effector protein Ago2, than Ago1," according to the study's abstract.
This week, RNAi News spoke with John about the findings.
Let's begin with some background on your lab and the research there.
My lab is a basic molecular biology lab, although we combine computational techniques with biology because I was actually trained in computational biology. I switched to molecular biology — and it’s still an ongoing change — because I wanted to focus on cancer because it affects my family.
The main focus of the lab is to look at non-coding RNAs. Protein-coding genes correspond to only three percent of the whole genome, and the rest of the 97 percent of the genome used to be called "junk DNA" for a long time. My hope is to look into this "junk DNA" to find diagnostic and prognostic markers, and, hopefully, therapeutic targets for cancer.
Could you give some background on these unusually small RNAs, or usRNAs?
We used to call everything junk DNA when we didn't have any clue of what it was doing for a long time, and the term has gone by, but now there are people calling things junk RNA, [particularly] RNAs smaller than microRNAs in the cell.
People have ignored the RNAs that are smaller than 20 nucleotides, primarily for two reasons: One, on the unfounded notion that they are simply transient degradation products and we shouldn't be looking at them, and two, the myth that these RNAs are so short that they can never be mapped back to the genome precisely, which is not true.
In fact, if you give me a set of 15-mer RNAs, I can find a unique match for 16 percent of those in the human genome. This is because, out of the approximately 415 possible 15-mers in the genome, 170 million 15-mers map to unique locations in the genome. So it is a common misconception in the field that they cannot be mapped.
This is where our research comes in. When we found a very small RNA that was associated with a microRNA, we started following it up, and surprisingly found thousands of such unusually small RNAs.
Can you give an overview of the study that was published?
We were studying the Kaposi sarcoma-associated herpesvirus microRNA K12-1 and doing a Northern blot, and we saw that this microRNA was consistently associated with another very distinct smaller band around 16 nucleotides. We did next-generation sequencing to identify the sequence and, to our surprise, we found that there is a much larger number of RNAs that are below 20 nucleotides than there are microRNAs and their peers — approximately four-fold more.
The apparent abundance of unusually small RNAs led us to further investigate the original viral microRNA-derived RNA. We found that, although it is very small, it can actually regulate a human gene like a microRNA. So if you just put [into a cell] a double-stranded RNA that is as small as 17 nucleotides long, it would actually go into the Ago-protein complex and regulate genes.
Is the regulation process similar to the way microRNAs work?
That is certainly one pathway … but we went on to find that these usRNAs have very position-specific motifs that are not present in any other type of small RNAs that we know like Piwi-interacting RNAs and [transcription start site-associated]-RNAs. The other surprise was that, using another data set that we could get from a paper published by Gunter Meister's lab [at the Max Planck Institute of Biochemistry], we found that these usRNAs are preferentially associated with Ago2 proteins, [rather] than Ago1.
On top of that, when we looked at another tissue, breast, we found that 25 percent of the usRNAs are [totally] identical to the one set we identified in the virus-infected cell line, [which showed] that these RNAs are consistently produced and are accurately generated in other tissues.
Cumulatively, [our data] show that the cell is putting an effort into making these RNAs consistently and accurately across multiple tissues. The data that we have in regard to the Ago association, the motifs that we don't see in other RNAs, and the fact that double-stranded RNA as small as 16 nucleotides can regulate gene expression, show that we should not be ignoring these RNAs as junk.
The other important point is that all gene-expression screening in all diseases so far has completely ignored these RNAs. Even if, with the extreme assumption that these are simple degradation products, the fact that they are made in multiple cell types in a precise manner shows that they may be useful as diagnostic and prognostic markers. One thing to note is that these RNAs are so small that they are perfect molecules for cell-to-cell transfer. So things like serum and other bodily fluid profiling would be the perfect next step for [using] these RNAs.
Is it possible that these could just be byproducts of a very precise degradation?
It is quite possible that there is a very specific degradation pathway in the cell that we have never known, and that produces [these usRNAs]. The point is, they are present in the cell in reasonable quantities, and double-stranded 15-mers can actually regulate genes. These RNAs also have highly position-specific motifs that are not present in other RNA products. An argument is that these motifs are simply stabilization motifs ― but then you would have seen them evolve in other classes of RNA products such as TSSa-RNAs ― we don't see that.
So the central point is that since these RNAs are produced, even if they are degradation products of a very specific mechanism that we don't know, they could have evolved some type of function, just like degraded peptide fragments used by the MHC complex for multiple reasons, particularly cell defense.
The message is, do not ignore these RNAs. People are doing massive sequencing these days, and the [technologies] are perfect for identifying these RNAs. But we continue to ignore them and we shouldn't because [they have] a lot of characteristics that are really interesting.
Is there work underway at your lab following up on these findings?
We are looking into a specific sub-class of these RNAs that has one of these motifs. Those experiments are ongoing, and we still have to figure out a lot. The preliminary results are interesting, but any detailed comment on this would be premature.