You'd Prefer An Argonaute

I didn’t make it to this year’s Keystone conference, RNA Silencing: Mechanism, Biology and Application, held last month in Colorado. I know, however, a bunch of people that did go, and they said it was pretty good. Pre-tay, pre-tay, pre-tay, pretay good.

Some things I heard (thanks kindly to my comrades):

A theme that first splashed at last year’s Keystone was a major interest this year: CLIP experiments. One surprise observed in these datasets is much more Argonaute binding in open reading frames than was previously considered. Some extend this to say that there is much more functional binding in ORFs than was previously thought. I imagine, however, that much of this binding is non-specific, or transient, analogous to what’s seen for transcription factor binding genome wide (e.g. see Li et al. 2008). Prudent calibration of potential background/noise should be required before the outstanding potential of these immense datasets can be exploited to make fresh, solid conclusions about miRNA targeting. One group is extending their CLIP method beyond just Argonautes, and looking for the RNA binding signatures of many other RNA binding proteins.

One speaker revealed a triumphant result that nicely explains why knockouts of mammalian Ago2, but not the other Argonautes, are embryonic lethal. This was previously very confounding because while mammalian Ago2 is the only Ago with slicer activity, slicing activity hasn’t been widely demonstrated in mammals. All four mammalian Agos obviously facilitate miRNA seed directed repression, but now we have a thoroughly satisfying explanation for why slicer activity is also necessary in the developing embryo.

A couple groups showed some nice new structural and biochemical data that further refines our understanding of which domains of GW182/TNRC6 and PABP bind each other. Early looks show some predicted binding differences in two model systems studied, fly and mammal. The data is revealing more about the mechanism of repression of mRNAs by miRNAs.

Lastly, another speaker convincingly showed why some sRNAs are methylated and some are not. In plants, for example, the vast majority of target sites form near perfect complementarity with miRNAs, and all the miRNAs are methylated. The same goes for endo-siRNAs in flies. But in mammals, the vast majority of miRNAs do not have perfect site targets, nor are they methylated. Methylation appears to provide protection from degradation of the sRNA when it pairs with a perfect site target. In contexts where many perfect sites for sRNAs are present, sRNAs will be methylated. In contexts where such sites are very rare, the sRNAs aren’t methylated. A very pleasant resolution.