Just yesterday I said (about the herpesvirus-encoded microRNA stories at the International Herpesvirus Workshop) that I would wait until publication before I talked about them. I know people were on tenterhooks waiting for the update, so fortunately it only took a day before something was published about them. (Image at right: Herpes simplex capsid, from this paper.)

One thing I didn’t mention about the microRNAs was how hard it seems to be to assign functions to them. Now, I’m not a microRNA expert by any means; I use siRNAs quite a bit, and I had (without thinking much about it) assumed that to find the function of a microRNA you simply look for the complementary sequence in the databases, and Bob’s your uncle. Of course, I knew that microRNAs differ from siRNAs in that they are not necessarily perfect matches, but I didn’t take the obvious next mental step and conclude that simply looking for a complementary sequence wouldn’t work.

The two main goals described in the IHW posters were to, first, identify microRNAs in herpesviral genomes. (As I said, there seem to be a bunch, up to 18 in some genomes.) But having identified them, how to find their targets? A fair bit of bioinformatic energy was turned to that aim, with a bunch of targets being identified. Many were at the preliminary stage of testing and confirmation, so it wasn’t at all clear to me how accurate these predictions are.

In today’s issue of Science, there was this article:

Host Immune System Gene Targeting by a Viral miRNA

Noam Stern-Ginossar, Naama Elefant, Albert Zimmermann, Dana G. Wolf, Nivin Saleh, Moshe Biton, Elad Horwitz, Zafnat Prokocimer1 Mark Prichard, Gabriele Hahn, Debra Goldman-Wohl, Caryn Greenfield, Simcha Yagel, Hartmut Hengel, Yael Altuvia, Hanah Margalit, Ofer Mandelboim

Science 20 July 2007:Vol. 317. no. 5836, pp. 376 – 381 DOI: 10.1126/science.1140956


This story was not reported at the Workshop, but what they find is that one of the microRNAs encoded by human cytomegalovirus (hcmv-miR-UL112) targets a member of the non-classical major histocompatibility complex family, MICB. MICB (picture at left) is a ligand for NK cells, and the conclusion is that this microRNA is an NK cell immune evasion molecule — it downregulates one of the molecules that NK cells use to recognize virus-infected cells. (This also ties in with my earlier post, where I commented on how NK cell immune evasion was flourishing these days, as we understand more about NK cell receptors.)

OK, what’s really interesting to me is this: This same microRNA was reported at the Workshop to have an entirely different target!1 At the Workshop, a different group reported that this microRNA targets some of the virus’s own regulatory proteins, and proposed that it’s involved in control of the virus’s genes.

The published paper uses “our newly developed target prediction algorithm, RepTar” — described in some detail in the supplementary information, but (I think) not publically accessible as such for testing; the abstract used a “bioinformatic approach” that may have been described in more detail on their poster, but which I don’t have here. In any case, it seems that there’s no widely-accepted technique to find valid microRNA targets.

Are both right? Both wrong? I believe there are examples of microRNAs that regulate multiple genes, but I thought they were mostly in gene families, so that it makes (teleological) sense to control them together. These are completely different genes with completely different functions. Viruses in general are very good at multi-tasking proteins (e.g. adenovirus E1B protein, which basically controls everything in an infected cell, washes dishes, and changes your oil too), so maybe they’re also good at multi-tasking microRNAs? Both groups have validated their results with other techniques, but I’d like to see third-party validation before I firmly accept either conclusion.

The most likely explanations, I think, are that (1) One of the groups is mistaken in their target identification, or that (2) I’ve mixed up two different miRNAs. I’ll have to wait for the other group’s publication to find out, if then.

  1. At least, I think it’s the same microRNA — the nomenclature is a little different, and the abstract from the workshop doesn’t offer a sequence I can use to confirm. At any rate, the Workshop version is called “UL-112-1” and published version is “hcmv-miR-UL112”; I think they’re the same.[]