The structures of release factors also show molecular mimicry (Figure BE). Mutations in Release Factor (RF) genes reduce the efficiency of termination and allow some read through and frameshifting (see below). Over expression of RF genes increases termination efficiency. RF-1 recognizes UAA and UAG; RF-2 recognizes UAA and UGA. RF-3 (together with GTP hydrolysis) stimulates both RF-1 and RF-2. Defects in RF-3 cause misreading of all three stop codons. Both RF-1 and RF-2 are essential genes, but RF-3 is not (that is, loss of the RF-3 gene is not lethal). There are two eukaryotic termination factors: eRF-1 is equivalent to both prokaryotic RF-1 and RF-2, and eRF-3 is a GTP-dependent counterpart of RF-3. eRF-1 recognizes all three stop codons. Both eRF-1 and eRF-3 are essential genes.
(See Table 27.4 on page 1044)
E. coli Release Factors |
||
| Factor | Binds GTP | Role |
| RF-1 | No | Recognizes UAA and UAG |
| RF-2 | No | Recognizes UAA and UGA |
| RF-3 | Yes | GTPase, stimulates RF-1 and RF-2 |
Many, but not all, E. coli genes terminate with tandem stop codons; that is, two stop codons next to each other. The three stop codons have inherently different efficiencies. In E. coli UAA is the most efficient and is most frequently represented in highly expressed genes. UAG is less efficient, and UGA is the least efficient being naturally leaky at a level of 1-3%.
Termination is probably ruled by a combination of four, not just three, nucleotides. The first base 3' to the stop codon has a large effect on the efficiency of termination:
The decreasing order of efficiency for the first base 3' to the stop codon in E. coli: U > A > G > C.
The decreasing order of efficiency for the first base 3' to the stop codon in mammals: A or G >> C or U.
Other structural elements also affect the efficiency of termination, but we will not consider them further in this class.
The process of termination begins when a stop codon on mRNA is encountered in the A site (Figure BD). In bacteria, recognition of the stop codon involves two release factors, RF1 and RF2. Both factors recognize UAA; however, UAG is recognized by RF1 while UGA is recognized by RF2. In eukaryotes, a single factor, eRF1, recognizes all three stop codons. Another release factor, RF3, is a GTPase and binds to RF1 or RF2 in the A site of the ribosome. RF3 is not essential in bacteria, and does not bind to RF1 or RF2 outside the ribosome. (Note: eRF1 and eRF3 form a complex on as well as off the ribosome.)
The binding of RF1 or RF2 to a ribosome with the appropriate stop codon in the A site triggers the hydrolysis and release of the peptide chain from tRNA in the P site. It is not clear whether RF1 or RF2 participates directly in catalysis or whether it induces catalysis by the ribosome. RF3 promotes rapid dissociation of RF1 and RF2. Originally, it was thought that the binding of RF3 to the ribosome triggered its GTPase activity with concomitant release of RF1 or RF2. However, more recent work shows that the hydrolysis of peptidyl tRNA by RF1 or RF2 is required for binding GTP to RF3 on the ribosome. This in turn leads to a conformation of RF3 witb high affinity for ribosomes and the dissociation of RF1 or RF2. The hydrolysis of GTP is required for subsequent dissociation of RF3.
There is an additional crucial step in translation termination: recycling of ribosomes through decomposition of the termination complex. In E. coli this process requires a ribosome recycling factor (RRF, originally called ribosome releasing factor). The structure of RRF (Figure BE) is the closest in shape and charge distribution to tRNA of any of the factors determined so far, supporting suggestions that it mimics tRNA in the A site. The process of releasing the ribosome after protein synthesis has terminated is fundamental because the gene for RRF is essential.
After release of the peptide chain, the ribosome is left with mRNA and a deacylated tRNA in the P site (Figure BD). This complex needs to be disassembled to prepare the ribosome for a new round of protein synthesis. Ribosome recycling factor (RRF) along with EF-G is required for this process. RRF and EF-G lead to the dissociation of ribosomes into subunits upon GTP hydrolysis. Subsequently, initiation factor IF3 is required for removal of the deacylated tRNA from the 30S subunit.
Note: no equivalent requirement for a eukaryotic RRF has been demonstrated. Eukaryotic genes with homology to RRF seem to encode proteins that are targeted to organelles, such as mitochondria.
Figure AQ shows a model of E. coli release factor 2 (RF2) regulation through frameshifting. If RF2 recognizes the stop codon UGA, translation is terminated after synthesis of a 25 amino acid peptide (which is degraded). If frameshifting occurs, the complete sequence of RF2 is synthesized.
The DNA sequence of the release factor 2 (RF2) gene revealed a stop codon at position 26 where the open reading frame switched to the +1 frame. Comparison of amino acid and mRNA sequences showed that ribosomes can shift reading frame at codon 25, avoid the stop codon, and decode the main portion of the message. Further work showed that this site-specific shift in reading frame is not just "noise" in translation but remarkably efficient: 30% of the ribosomes make complete RF-2 and 70% terminate at the zero frame UGA after 25 codons. The RF-2 frameshift site is a "slippery" codon where the mRNA slides within the ribosome complex one nucleotide by breaking codon-anticodon pairing with peptidyl tRNA in the ribosome P site and re-establishing pairing with an overlapping codon in the new frame. In this case tRNA-Leu with anticodon 3'-GAG-5' pairs with CUU in the first frame and then with UUU in the +1 frame (CUU.UGA), inserting one leucine for four nucleotides in the mRNA.
Competition between termination and frameshifting is over which process captures the U of the UGA sequence. Release factor 2, the recoding product, promotes ribosome termination at UGA (and UAA) and so competes with recoding by enhancing termination at UGA, the 26th codon in the original frame. Competition is tilted in favor of recoding in two ways. The presence of a C 3' of the UGA makes a poor termination context, and CUU is a particularly shift-prone codon. If extra RF2 is present, frameshifting decreases so that less RF2 is made. When RF2 is in short supply termination loses, frameshifting wins, and RF2 concentration is restored.
The stimulatory signal is a short sequence three nucleotides 5' of the shift site that pairs with 16S rRNA of the translating ribosome, just like the Shine-Dalgarno pairing that occurs 5' of the AUG start codon during ribosome initiation. Thus, RF-2 mRNA has two Shine-Dalgarno sequences, one for initiation and one in the coding sequence for promoting frameshifting. Spacing between the Shine-Dalgarno sequence and the site of action is crucial and varies for these cases. With initiation the optimal spacing to the AUG is 5 bases, although there is considerable latitude (3 to 12 bases). With RF-2 required spacing is 3 bases. The implication is that pairing between mRNA and rRNA at these sites distorts the complex to either put the ribosome in an "initiation mode", a "plus mode" to force it forward, or a "minus mode" to force it backward (see below).
Figure AU shows a model of E. coli dnaX regulation through frameshifting. The dnaX single message encodes both the tau and the gamma subunits of DNA polymerase III holoenzyme. Only the tau subunit is essential for growth, but the holoenzyme without gamma has slightly different kinetic properties. Standard translation makes the longer tau protein (71 kDa). A -1 frameshift two-thirds of the way through the tau coding region leads quickly to a -1 stop codon to make the smaller gamma protein (47 kDa). The shift site is a particularly slippery tandem site, A AAA AAG. The efficiency of frameshifting approximates 50%. Two stimulatory signals are required: a 5' Shine-Dalgarno sequence which pairs with 16S ribosomal RNA and a 3' stem loop that plays a role in pausing the ribosome at the shift site. Deletion of the Shine-Dalgarno sequence decreases the efficiency two-fold and deletion of the stem loop decreases it ten-fold. With the dnaX -1 shift, optimal spacing of the Shine-Dalgarno sequence is 10-13 bases.