mRNA nuclear export regulates itself like a big bowl of spaghetti

OK here's a refresher ...nuclear export of mRNA in yeast:

i-7acfb25a307fa45ad0670da38c01e61b-export.jpg

Nucleoplasm is on the bottom, cytoplasm on the top. The bilayered nuclear envelope is represented by the two black lines. The top line is the outer nuclear envelope the bottom line is the inner nuclear envelope. The big red/blue thing with the purple basket is the nuclear pore complex (NPC).

In green is the TREX mRNA nuclear export system (TREX = transcription, export). These proteins are loaded sequentially during transcription in yeast and during splicing in higher eukaryotes. In most cases the Tho complex, which is comprised of four proteins, and Sub2p are loaded onto the transcript during transcription . Sub2p then recruits Yra1 which in turn gets Mex67/Mtr2 to join the gang. Mex67 can associate with the FG repeats, a mesh work found inside the NPC and thus allows passage across this barrier (for more on this click here). Since this is biology and nothing in biology is simple, many RNA processing events such as capping, splicing and 3' processing can also stimulate the association of the proteins to the newly made transcript.

But how is the level of TREX components regulated? You might think that the cell would like to produce the correct amount of each component. Having too little or to much of these components might be bad. For example making too much Yra1p is toxic. If export is hyperactive, then the TREX system will overwhelm the machine that blocks the export of unspliced transcripts (the Mpl1/2-RES system) and export unspliced mRNA. In addition you may want to balance the protein export/import machinery regulated by Ran, with the mRNA export machinery so that one doesn't overwhelm the other. And you may want to take into account how fast you're splicing machinery is acting. You would want a big balancing act. Well we are begging to figure out how it all works ... welcome to spaghetti science. (Watch out or you may trip over a lose noodle and become part of the main course.)

Take for example Yra1p, a key player in the whole process. It turns out that the levels of Yra1p are regulated in part at the mRNA level. Here's a diagram about the latest knowledge on how the Yra1 mRNA is regulated:

i-1eee21f8f88c07b269dd4da9931aab5d-yra1.jpg

To orient you, all proteins are in purple, the Yra1 mRNA has two exons ( the orange boxes) Remember, an exon is a part of the mRNA that will be translated into protein. The two Yra1 exons are separated by an intron - quite rare for budding yeast. Only ~5% of genes have introns in yeast, most of which have something to do with ribosome biosynthesis. The blue cylinders are NPCs. Overall the above picture look strange. Lots of mRNA export proteins are performing secondary activities. There are feedback loops ... and lots of holes in our knowledge.

- The first thing that you'll notice is that Yra1 protein inhibits the splicing of its own mRNA. In fact the splicing of Yra1 transcript is relatively slow, and so this would be a good way for the cell to asses the levels of Yra1p and the activity of the cell's splicing ability.

- Second you'll notice that the unspliced Yra1 mRNA is exported by the Crm1p karyophilin and not by the TREX complex. Crm1p is the major exportin for proteins and is part of the Ran protein nuclear import/export system. I'm not sure why the cell would integrate the Ran system into Yra1p levels at this point. It could be that the cell just needs a way to export the unspliced mRNA using a strong export system to overcome the Mlp1/2-RES system and thus uses Crm1p instead of the TREX complex.

- Thrid you'll notice that Mex67p affects where the message is processed at the 3' end. When you have lots of Mex67p the 3' untranslated region becomes longer. It'[s unclear what the implications are.

- Fourth when the unspliced Yra1p message is exported it is degraded via the decapping pathway. Edc3p (enhancer of decapping 3) recognizes the Yra1 intron and stimulates the degradation of this transcript by stimulating the decapping enzymes (dcp1/2p). Interestingly Edc3p only acts on two types of mRNAs, unspliced Yra1 message and unspliced mRNA that encodes a particular ribosomal protein. It would also seem that Mex67p levels affect the rate of Yra1 mRNA decapping and destruction. Why would the cell have such a complicated procedure. Is the unspliced message producing some protein? Is this degradation just another way of integrating an mRNA decay pathway into the cellular computation that determines Yra1p levels? We'll have to see.

So there you have it, this is our current knowledge of Yra1 mRNA processing. I bet that other steps are equally regulated. And how about in higher eukaryotes? There are indications that other bizarre things are happenning. The orthologue of Mex67. called TAP, also has some weird mRNA export/mRNA splicing issues. I'm sure that the final picture will look like a tangled mess.

Ref:
Dong S, Li C, Zenklusen D, Singer RH, Jacobson A, He F.
YRA1 autoregulation requires nuclear export and cytoplasmic Edc3p-mediated degradation of its pre-mRNA.
Mol Cell. 2007 Feb 23;25(4):559-73.

More like this

Yeah good question.

It turns out that many of the enzymes that are responsible for producing IP6 show export defects when they are knocked out in combination with other genes (I don't remember off hand which combos are lethal). IP6 then promotes Gle1p-Dbp5p association and Dbp5p's RNA helicase activity. This helicase has been proposed to either pull the RNA out of the nuclear pore or strip proteins off of the exiting RNA. There are many weird aspects to this part of the mRNA export pathway. I wrote about this sometime last year:

http://scienceblogs.com/transcript/2006/09/ip6_and_mrna_export.php

The one thing I don't get is that if IP6 is a signalling molecule, it should get turned over, yet to date no one has found any enzyme capable of doing this. There are kinases that add an extra phosphate onto one of the preexisting phosphates on IP6 to form IP7, so that could potentially regulate IP6 levels.

I'm trying to put my brain into work so reading your post was good. Regarding your second point: the cell uses the protein export machinery to export Yra1 mRNA. Couldn't this be a left over (evolutionary speaking) from the mRNA world and the beginning of the protein world? Why that, I don't know.

If there was an RNA world it would have predated the first prokaryote, in contrast the mRNA nuclear export machinery dates back to the origin of the eukaryotic cell, so it is is much more recent developement than the RNA world and thus likely not to be a remnant.