Alex Palazzo is an Assistant Professor in the Department of Biochemistry at The University of Toronto.
I've been struggling with eIF4E. You see eIF4E is the major cytosolic cap binding complex. When mRNA is synthesized in the nucleus the cap is loaded with nuclear cap binding complex, CBP80/20. Then (the story goes) the mRNA is exported to the cytoplasm where the ribosome engages the transcript and thus the pioneer round of translation begins. If there are problems such as premature termination codons, the RNA is degraded by nonsense mediated mRNA decay (NMD). If all is well, CBP80/20 is exchanged for eIF4E and now full scale translational synthesis can begin.
How does eIF4E promote translation? (This and new stuff concerning mitosis are below the fold.)
It is part of the eIF4F complex, a collection of proteins includes helicases that can unwind RNA and scaffold proteins that bind to a plethora of other components such as other initiation factors, poly-A binding proteins and the 40S subunit of the ribosome. But of all the eIF4F contributors, eIF4E is apparently limiting, and so not all RNAs are blessed with it. Some RNAs seem to need eIF4E more than other transcripts, and some don't need it at all. So you have:
1) mRNAs that are highly dependent on eIF4E. These RNAs are thought to have long and highly structured 5' untranslated regions that need to be unfolded by the eIF4F complex. Many of these mRNAs encode proteins necessary for cell proliferation.
2) Regular mRNA that need some eIF4E.
3) mRNAs that have internal ribosome entry sites (IRES) that can recruit the ribosome without the assistance of the eIF4F complex.
eIF4E is also found in stress granules, large aggregates of mRNA that are stalled in translation initiation (for experts, these transcripts are thought to be bound to the 48s preinitiation complex), and P-bodies, smaller mRNA filled entities thought to be involved in degrading transcripts. P-bodies are related to mRNA storage granules in oocytes and RNA transport granules in dendrites. These two entities seem involved in mRNA storage and transport. What is eIF4E doing in P-bodies? I have no clue.
Now since eIF4E is limiting the cell by regulating how much of it is around can regulate which pools of mRNAs it translates. One way to regulate eIF4E activity is by producing eIF4E binding proteins (4E-BPs) that can prevent eIF4E's ability to recruit the rest of the eIF4F complex to the transcript.
Under normal circumstances, 4E-BPs are inactivated. They are phosphorylated by mTOR (target of rapamycin) and are unable to bind to eIF4E. But if cells are metabolically stressed, or if they are treated with rapamycin, 4E-BP phosphorylation goes down, and they now inhibit eIF4E function. Result: the first pool of mRNAs are no longer translated and cell proliferation goes down.
But are there other circumstances when eIF4E is inhibited? Apparently mitosis.
A paper in the March 15th edition of Nature describes how 14-3-3-sigma gets phosphorylated during mitosis allowing it to interact with the eIF4B subunit of the eIF4F complex. This basically prevents eIF4E from binding to eIF4B and inhibits cap dependent translation. In fact, cap dependent translation is so low that most mRNAs are no longer translated and ribosomes and other translation initiation factors can now focus their efforts on IRES containing transcripts.
When the researchers reduced 14-3-3-sigma levels, the cells couldn't finish mitosis, they got stuck in the final act of fission or cytokinesis. Incredibly rapamycin treatment partially rescue the cytokinesis defect!
The authors of the study made some guesses as to which IRES containing transcript may be important for cytokinesis, but the effect of overexpressing this particular mRNA (a cyclin dependent kinase) was small ... cytokinesisis is likely to involve the translation of many other transcripts. One surprising result is the profound effect 14-3-3-sigma has on translation as seen in the incredible data as reproduced on the right. Cells are synchronized and thus are proceeding through the cell cycle at the same time and hit mitosis together at about 16hrs. The blot shows newly made proteins as detected by feeding S35Met to the cells. Under each time point are control cells (Con) and cells where 14-3-3-sigma production is inhibited by RNAi (sigma). Look at the decrease in protein production in control cells at 16hrs. Look at the increase in protein production in cells lacking 14-3-3-sigma. Wow! Now that's an incredible difference.
So there you have it another twist in the continuing story of eIF4E. What's next?
Comments
14-3-3 sigma is a bit of a strange duck too. If i remember correctly, it has no introns so is likely to be a a retro-transposon event. Also, for a lot of 14-3-3's there seems to be a lot of redundancy, but for some reason sigma has a lot of specific roles (like this one)
Posted by: Dave | April 3, 2007 4:22 PM
My one problem with this paper is that although the two 16hr lanes (Con and sigma RNAi) are different in intensity, the band patterns are similar which suggests that the same proteins are being made, just at different levels.
Posted by: Acme Scientist | April 3, 2007 6:23 PM
AS,
Yeah someone mentioned that to me. Those bands are probably reflective of the major cellular translational products such as ribosomal and cytoskeletal proteins and metabolic enzymes. So all those are turned down. Now specialized genes that have IRES have so few transcripts that when the expression of those transcripts are upregulated, there isn't enough protein to see in total lysates. Remember that for the most part there will be an upregulation of kinases and other trigger proteins. You probably don't need much of those products.
Posted by: apalazzo | April 4, 2007 8:11 AM