Dinner with the Transcription Crowd

Well last night I was invited to dine at Clio's with our Seminar Speaker, James Manley and some of the local transcription gurus, Kevin Struhl, Danesh Moazed, Steve Buratowski and Miriam Bucateli, a postdoc in the Buratowski lab. Unfortunately Dr. Manley had to leave early to catch a flight back to NY. But the rest of us had a nice conversation about that 70% of the genome being transcribed. (For past discussions click here and here).

So all the transcription guys firmly believed that the whole bit about 70% of the genome being transcribed is true. In fact Kevin Struhl pointed out a review he wrote for Nature Structural Biology where he made some calculations as to the feasibility of this hypothesis. In fact if you calculate the number of actively transcribing RNA Polymerase II molecules and the occupancy of each transcript in the yeast genome, you find that only about 10% of active Pol II are transcribing coding RNAs. Indeed by ChIP analysis it is clear that RNA Pol II is almost everywhere in the genome except for a few actively silenced spots. For highly expressed genes, RNA Pol II occupancy is high, but for most genes with mediocre promoters, RNA Pol II occupancy is barely above the "basal state". If you calculate the Boltzman distribution between Pol II binding at a moderate promoter and just any random piece of DNA, and then you factor in how much non-coding DNA is in a cell, you realize that random transcription throughout the genome is not only probable but almost inevitable.

The big question is why do cells have this?

Well the simplest idea is that all this "junk DNA" is actually used by the cell. So that instead of 1000 non-coding genes, most eukaryotes contain 100,000 non-coding yet useful RNAs. Probably the truth is that there are lots of non-coding RNAs but that the number of useful non-coding transcripts ain't so high.

Another idea is that RNA Pol II is actively probing the genome for transcripts and that this may facilitate any the expression of transcripts (coding or non-coding) that may be useful for the organism. Think of it as the cell experimenting with the potential transcriptional output that it can produce.

In many cases, RNA proofreading mechanisms such as RNA export, the Exon Junction Complex (through activating NMD), the exosome and other RNA degradation machines can act to filter out most aberrant transcripts ... but of course these processes are not full proof and so some of these extra transcripts will make it past the quality control checks. If these extra transcripts help the organism, evolution simply has to promote the stability of these transcripts. This process where experimentation within an organism's potentiality (in this case the potential transcripts in a genome) can lead to certain extreme phenotypes (in this case rare aberrant transcripts) that then can be enhanced by natural selection (in this case the stability or production of these "aberrant transcripts") is a type of facilitated evolution, as first described by James Baldwin (hence the Baldwin Effect). For a fuller description of evolvability and the Baldwin Effect read Kirschner and Gerhart's book The Plausibility of Life.

But perhaps it's not what is being produced. Perhaps the key lies in the act of transcription itself. For example, RNA Pol II activity may affect local chromatin structure, so that it's not important what is being produced but only that nucleosomes are being moved around and perhaps being modified as a consequence of active transcription.

Some other factoids:
- from 5'UTR caging experiments as much as 1/3 of all transcripts may have extended 5'ends probably due to leaky transcription occurring upstream of the gene's promoter.
- acording to James Manley, about 50% of genes have multiple 3'end termination sites. Since many regulatory elements, such as miRNAs binding sites, are found in the 3'UTRs, this has major implications for mRNA metabolism.
- by recent estimates, 70% of genes are alternatively spliced. Again this may act in certain cases as an increase in gene regulation but may actually have more to do with facilitated evolution as discussed above.

Ref:
Struhl K.
Transcriptional noise and the fidelity of initiation by RNA polymerase II.
Nature Structural & Molecular Biology - 14, 103 - 105 (2007)
doi:10.1038/nsmb0207-103