As a grad student at Columbia, I once saw a talk by Joachim Frank at Rockefeller. Siting in the audience, I was wowed as Frank described the cryo-EM structure of the ribosome in many different conformations, each representing one step of the polypeptide chain elongation cycle. Compiling them together, he produced a little movie of a ribosome manufacturing a protein. While I was taking in this movie I remember thinking - wow, this is how the machine works!
Recently a few other labs have published the cryo-EM structures of the ribosome with every freakin' Ribosome cofactor. The latest structures are the SRP-Ribosome Nascent Chain complex (RNC) in eubacteria and in eukaryotes. As a bonus, the newly made signal sequence is visible in one of the eubacteria structures. Read about the structures in these two papers.
(I know, people are deserting the archea, eubacteria, eukaryotic three kingdom division ... but since we're talking about the ribosome, and the three way division is based on rRNA, the three branched phylogeny should apply here.)
BACKGROUND (for the rest skip this section and just gaze at the pretty pictures):
mRNAs are translated into proteins by the ribosome which is the most conserved enzyme in biology. All of life, as we know it, depends on ribosome based protein synthesis.
Certain mRNAs encode what is know as a "signal sequence" or leader peptide. These short polypeptide chains bind to the sequence recognition particle (SRP) as soon as they emerge from the ribosome. SRP then turns off further synthesis of the protein and targets the nascent polypeptide chain, the ribosome that is manufacturing the new protein, and the associated mRNA, to the surface of the ER in eukaryotes and to the plasma membrane in prokaryotes. The ribosome studded ER is called rough ER. The signal sequence is then threaded through a pore called the translocon. The translocon is also found in all living organisms from bacteria to man. Once SRP has fallen off and protein insertion into the translocon has begun, protein synthesis resumes. The rest of the newly synthesized protein is pumped through this pore and thus into the lumen of the ER to generate a new secreted or transmembrane associated protein. This process is called co-translational insertion. Now in most cells there is a second process to get prefabricated proteins into the translocon, this is called post-translational insertion. Depending on the cell type, this second process uses different machines.
OK here are some structures of ribosomes with their many cofactors:
Previously there was a ribosome-translocon structure. This structure was very odd in the sense that the ribosome had a second translocon stuck to it's side. This non-physiological translocon turned out to be better ordered than the proper translocon. In the diagram below, the ribosome is in pale yellow and blue, the non-physiological translocon (or protein conducting channel "PCC") is in red and the proper translocon is in dark blue. In addition, there was some uncertainty in the field about how many copies of the translocon were at the real peptide exit site, but this structure seemed consistent with 1-2 copies. In fact I think that they stated that there were 1.5 copies of the translocon in this structure ... I'll have to go back and check. Recent evidence from structural and cross linking experiments supports the model that a single copy of the translocon (the SecYEG heterotrimer) forms the pore and not a tetramer. It must also be said that although a monomer acts as a pore, several pores maybe found under each ribosome.
OK the new structures ...
Here is the cryo-EM structure of the eubacteria ribosome-SRP complex (paper #1):
SRP sits exactly where the translocon (SecYEG) is. Of course this is no surprise. If we go on and read paper #2 we find out that:
Notably, additional density is present in the hydrophobic groove of the SRP54 M domain (Fig. 2d), which has previously been suggested as the main signal-sequence-binding site of SRP19. This density, thus very probably representing the signal sequence, was observed in a position that in unbound SRP is occupied by a part of the finger loop of the SRP54 M domain. This may serve as a pseudo-substrate in the absence of a signal sequence.
So they can actually see the signal sequence emerging from the ribosome and bound to SRP.
From paper #2, here is the complex with a newly synthesized signal sequence in green:
Here is a scheme of the eubacterial complex:
The large subunit of the ribosome is in blue, the small subunit in yellow. The canal in the large subunit is where the newly synthesized peptide (protein) is pushed through (think of it as the birth canal of the ribosome. As new aminoacids are being added to the top where the green blob is (the green blob is a tRNA - I will not explain it here any further) the newly synthesized peptide (in green) is being pushed out, or down in this diagram, towards the exit. At the exit site, SRP (here in red) sits and in this diagram is holding the part of the polypeptide that encodes the signal sequence (the green cylinder). SRP is also making contacts to two subunits of the ribosome (the two orange blobs). When bound to the signal sequence SRP makes a total of four contacts to the ribosome. Some of these contacts are shared with another ribosome cofactor, trigger factor (TF), which acts as a chaperone for the newly emerging nascent chain. Since TF and SRP share binding sites they may be mutually exclusive. In fact when SRP holds on to a newly made signal sequence and engages the SRP receptor in the ER, TF is known to be released from the ribosome.
Now lets look at our SRP-ribosome ("our" meaning the eukaryotic complex). Here SRP has a long stretch ("SRP Alu")that reaches over and interacts with the t-RNA and halts further translation:
In comparison with the bacterial SRP54 subunit, the mammalian protein has a C-terminal extension of the M domain of about 70 amino acids, which is predicted to adopt a primarily -helical secondary structure. Using protein threading we created a molecular model of this C terminus that could be fitted into the remaining density with only very minor modifications (Fig. 4d). It wraps around the ribosomal rRNA helix 59 and reaches up towards the tunnel exit, also interacting with ribosomal rRNA helix 50.
This interaction may slow down translation while the SRP is engaged to the signal sequence. Once the signal sequence is delivered to the translocon, SRP falls off and thus translation
In summary, the machine and all it's accessories come more and more in focus ... Now can all you crystalographers get the crystal structure of these various complexes? ;)
- Log in to post comments
In grad school, we used to joke about the xtal people setting up trays to get a structure of the cell.
The joke got pretty funny toward the bottom of a keg of Sam Adams.
What about the Vault? I don't remember you posting about that one, Alex.
Oh man, I won't tell any of your friends that you gave so much credence to that first reference. My only other comment is that it doesn't really look like a signal sequence to me. It may be over interpreted. Still, cool structures. It is also kind of funny that you ignored the recent eubacterial 70S structure that contained tRNAs, mRNA and is at a resolution where they could identify all the magnesiums. That was something.
Yeah I saw the Noller and Ramakrishnan papers (were there any others?) There was also a splicesosome structure (cryo-EM). I just don't have enough time. Also I have an interest in SRP for my own stuff (as you know).
Tell you what, Mr. Rambling Prof, if you send me a synopsis of those other papers, which should not be so hard for you to do, I'll post it.
One quick comment on those papers you mentioned. The Noller paper was at much lower resolution and really didn't improve on what we already know. The Ramakrishnan structure was at a high enough resolution to see the peptidyl transferase center in detail and, essentially, confirmed the growing hypothesis that the ribosome is not a ribozyme. The RNA doesn't play a catalytic role and, in fact, there is a bit of essential protein right at the heart of the action.
I heard Frank give a talk earlier this year. I felt that he did a really poor job of explaining why anyone other than the ribosome cognescenti--which, to his credit, did compose a substantial fraction of the audience--should give a shit about these details of ribosome structure/function analysis.
Physioprof,
To a certain extent, many structure talks are really boring - as one famous scientist said "I would talk about the crystal structure that our collaborators solved, but it would just be 'Angstrom this, Angstrom that.'"
There are some good crystal talks though - BTM gives a pretty good one (I think). My guess is that the key to a "good talk" is to provide some greater insight into the process and to elaborate on the broad relevance of this insight - not an easy task for any speaker, but harder for those scientists that are more associated with a technique (crystallography, mass spec ...) than other researchers.
Thanks, I think. Man, I love it when people are hypercritical of talks outside their own field. I've been to plenty of bad cell bio talks. I think one of the biggest problems is that most biologists don't really appreciate the power of structure, or how to use it. I guess I'll just keep trying to make a difference teaching and then maybe even AP will come around.
By the way, most structural biologists (I don't even like that term - I propose structural biochemists) aren't associated simply with a technique. Frank is a great example of that. He is a world leader of cryo-EM but is generally remembered for his work on the ribosome. Not to say that this is true for everyone but then technique is what drives most of us, including the people trying to get the laser and the microscope to get the cells to make pretty pictures.