Monday, November 19, 2007

Holy Freaking Heterogeneous Ribosomes!

Once again I have Palazzo to thank for pointing out a cool paper - this one from Pam Silver's lab, showing that different paralogous (duplicated) yeast ribosomal proteins are involved in translating different messenger RNAs. Since there are some 60 or so duplicated ribsomal proteins, this suggests a huge possible amount of ribosomal heterogeneity in the cell. Ribosomes are not all the same - even in basic subunit structure -and this might be an important piece of the translational regulation puzzle. The authors go on to point out the parallels between this situation and that of histone proteins, which can also bind nucleic acid, are also highly duplicitous in yeast, and also highly subject to regulation through post-translational modification. Thus they propose that if there's a "histone code" that regulates transcription, there may also be a "ribosome code" that regulates translation. Personally, I don't see the evidence that either sort of code exists (vague speculation about codes usually strikes me as an easy way to score sexy points), but there's certainly something interesting going on here.

More interestingly, we recently had a great journal club about the evolution of paralogous histone-modifying enzymes in yeast. The paper examined the evolutionary balance between redundancy and diversity of function in duplicate gene pairs, and showed that functionally distinct (non-redundant) paralogues can compensate for their partner when, and only when, the other is deleted. Presumably then, a similar evolutionary dynamic could be at play in the evolution of ribosomes and the regulation of translation in general.

Speaking of histones, if there's no such thing as histone code, maybe it's more of a universe. The complexity of chromatin structure is unbelievable. What's even cooler is that it's dynamic. So not only is chromatin complex, it's ALIVE. At least that was the impression I took away from a recent talk by Ottawa chromatin mapper Marjorie Brand. Check out her lab's latest paper on how spreading of chromatin-associated MLL protein mediates communication between a distal upstream activator element and the B-globin promoter during differentiation.

Another random factoid that was mentioned in the first paper, and news to me: the yeast (S. cerevisiae) arose from a massive whole-genome duplication, but eventually all but 10% of duplicated genes were lost. Weird. Was the initial duplication event selection-neutral for the original yeast cell it occurred in? Hard to imagine, since I would think two genomes take longer to replicate than one, not to mention the structural instability associated with having all that homologous DNA around. So what's the advantage? Not much, if most duplications were eventually lost. Doesn't this say there was indeed a fitness cost to having extra copies of all those genes if they were ultimately weeded out by selection? The only thing that makes sense to me is if the cost/benefit of duplicate genes has varied over yeast evolutionary time. So it was advantagenous to have backup gene copies back sometime in ancient history, but then became unnecessary or cumbersome more recently. Maybe the environment or the yeast's response to it changed. For example, maybe at one point gene inactivation via radiation-induced DNA damage was a big problem for yeast, so it was often useful to have a backup genome copy around. Then, either environmental radiation diminished or yeast evolved a system for preventing or repairing DNA damage, and they no longer needed to keep the backup genes around. I suppose another possibility is that it has something to do with the role of sexual reproduction..........

Alas, I rant like a raving lunatic. Someone set me straight.


2 comments:

Lim Leng Hiong said...

Alas, I rant like a raving lunatic. Someone set me straight.

Birth-and-death model comes to mind.

Bayman said...

Sure, get all mathy on me...