I finally got off of my butt and read the latest paper from the Hegde group at the NIH. They answered a fundamental problem in membrane biology: how do you insert tail-anchored proteins into the membrane.
When proteins are synthesized, any newly made signal sequence or transmembrane domain that pops out of the ribosome will be recognized by the signal recognition particle (SRP; click here for some beautiful SRP-signal sequence-ribosome images). This large complex will then inhibit further translation of the protein and target the newly made signal sequence or transmembrane domain and the translating ribosome to the surface of the endoplasmic reticulum (ER) due to the action of the SRP receptor. Once the entire complex is at the membrane, SRP transfers the signal sequence or transmembrane domain to the translocon, a channel that allows proteins to be inserted into the lumen of ER. Next the ribosome positions itself so that it sits on top of the translocon and then resumes protein synthesis. As a result, the ribosome feeds the rest of the newly synthesized polypeptide directly into the channel.
But what happens if the protein has a single transmembrane domain at the very end of the protein (i.e. at the carboxy-terminal)?
Since this transmembrane domain is at the very end, once it is synthesized the whole newly made protein becomes detached from the ribosome. So then what? SRP is in contact with the ribosome and does not recognize free floating transmembrane domains.
Using some nice in vitro translocation assays, the Hegde group demonstrates that some cytosolic factor can bind to these free carboxy-terminal transmembrane domains and actively insert these proteins into ER derived membranes. Furthermore this cytosolic factor, named TRC40, is an ATPase protein that can bind to ER derived membranes. It also appears that inhibiting the translocon (i.e. the major protein conducting channel in the ER) does not inhibit this insertion reaction.
So TRC40 looks like a soluble ATPase that can bind to a receptor on the ER and insert proteins either directly into the ER or through some yet unidentified pore that is likely not the translocon (i.e. the Sec61 complex).
- Yeast have a TRC40 homolog, Get3, but the phenotype of the yeast deletion mutant seems very disconnected to TRC40's role as a insertion machine.
- In worms, Asna1 (TRC40's original name) has been reported to stimulate insulin secretion - does this mean that TRC40 could also work on signal sequences to promote post-translational insertion?
Obviously a lot more has to be done but we now have some insight into another fundamental biological activity that up until recently had been a mystery.
*** UPDATE ***
So over coffee with a bunch of Rapoport-ites we were discussing the TRC40 - insulin data and someone suggested that perhaps this reflected a defect in SNARE insertion into the membrane during synthesis. But if SNAREs insertion is compromised, why are the yeast Get3 deletions viable? Perhaps all C-terminal anchored proteins can insert into membranes, but the whole process is more efficient with TRC40???