How do you make an bioinformatician with particular interest in protein functional evolution happy? By testing their functional predictions experimentally. Well, that's that's what makes this particular bioinformatician happy anyway!
In 2008, I had a paper out with Vasili Hauryliuk on the evolution of the Hbs1/eRF3/Ski7 and eRF1/Dom34 protein families, which was published in BMC Evolutionary Biology. eRF3 and Hbs1 belong to the EF1 family of translational GTPases. They orginated from a duplication of eukaryotic elongation factor eEF1A (also found as aEF1A in archaea) before the last common ancestor of all extant eukaryotes.
In 2008, I had a paper out with Vasili Hauryliuk on the evolution of the Hbs1/eRF3/Ski7 and eRF1/Dom34 protein families, which was published in BMC Evolutionary Biology. eRF3 and Hbs1 belong to the EF1 family of translational GTPases. They orginated from a duplication of eukaryotic elongation factor eEF1A (also found as aEF1A in archaea) before the last common ancestor of all extant eukaryotes.
 |
| Schematic diagram of the phylogenetic relationships among the major families of trGTPases. Subfamily names are coded by their taxonomic distribution as follows; green and underlined: bacteria, red and italic: eukaryotes, blue with names prefixed with e/a: subfamilies present in both eukaryotes and archaea. Figure from my conference abstract published in the journal of Molecular Structure and Dynamics |
A bit of background is probably in order. eEF1A is responsible for binding and delivering aminoacyl-tRNA (aa-tRNA) to the A site of the ribosome during the elongation stage of protein synthesis. Elongation factor eRF3 acts in an analogous manner during the termination stage, binding and delivering eRF1 to the A site. eRF1 is the stop codon-recognising factor, and is a structural mimic of aa-tRNA. eRF1/3 and participate in an mRNA surveillance machanism called nonsense mediated decay (NMD), promoting mRNA decay in response to premature stop codons. Hbs1 is the paralogue of eRF3 and binds Dom34, itself a paralogue of eRF1. Together, Hbs1 and Dom34 (also called pelota) function in another mRNA surveillance mechanism no-go decay, onset when a ribosome is stalled on the mRNA. Then there's Ski7, which is only found in Saccharomycetale fungi and functions in non-stop decay, where a stop codon fails to be interpreted, and translation runs on through to the poly-A tail of the mRNA. Here's the schematic family tree of these characters:
 |
| Evolution of mRNA quality control mechanisms. A schematic representation of a proposed scenario for the origin and divergence of components of trGTPase-associated mRNA decay mechanisms in archaea and eukaryotes |
And the actual phylogenetic trees from Bayesian phylogenetic analyses with MrBayes:
 | |
| eRF1/Dom34p |
 |
| eRF3/Hbs1p/Ski7p |
One of the most interesting things about this family is that Archaea don't have Hbs1 or eRF3. We proposed two possible explanations for how translation termination, nonsense-mediated and no-go decay work in this domain of life. 1: eRF1 and Dom34 can carry out their functions without the need of a GTPase to deliver them to the ribosome. 2. Another GTPase is doing it, the most likely candidate being the closest relative in archaea: elongation factor 1A (aEF1A). Recently, two exciting papers were published in PNAS that confirmed that scenario number 2 is the case.
In Kobayashi et al., the crystal structure of the aDom34 and GTP-bound aEF1A complex is presented, and biochemical and genetic analyses of Saito et al. show evidence of binding between aRF1 and aEF1A. Using the Kobayashi et al structure, a docking model of the aRF1·aEF1A·GTP complex has also been constructed.
So, with aEF1A being multifunctional in these mechanisms, this suggests that the original eukaryotic eEF1A was too (before the duplications that gave rise to eRF3 and Hbs1p). The modern eEF1A has a whole range of additional "moonlighting" functions (see Ejiri, 2002). Perhaps the evolution of Hbs1 and eRF3 freed up eEF1A from its ancestral functional constraints, and enabled it to diversify into its current roles.
There is another huge and exciting question that needs answering. Archaea isn't the only taxonomic group lacking Hbs1 - shock horror - some Apicomplexan protists also don't carry it! So what's the mechanism in those guys? Has eukaryotic EF1A reacquired the ability to deliver eRF1 and Dom34p? That would be another amazing story!
Atkinson GC, Baldauf SL, & Hauryliuk V (2008). Evolution of nonstop, no-go and nonsense-mediated mRNA decay and their termination factor-derived components. BMC evolutionary biology, 8 PMID: 18947425
In Kobayashi et al., the crystal structure of the aDom34 and GTP-bound aEF1A complex is presented, and biochemical and genetic analyses of Saito et al. show evidence of binding between aRF1 and aEF1A. Using the Kobayashi et al structure, a docking model of the aRF1·aEF1A·GTP complex has also been constructed.
So, with aEF1A being multifunctional in these mechanisms, this suggests that the original eukaryotic eEF1A was too (before the duplications that gave rise to eRF3 and Hbs1p). The modern eEF1A has a whole range of additional "moonlighting" functions (see Ejiri, 2002). Perhaps the evolution of Hbs1 and eRF3 freed up eEF1A from its ancestral functional constraints, and enabled it to diversify into its current roles.
There is another huge and exciting question that needs answering. Archaea isn't the only taxonomic group lacking Hbs1 - shock horror - some Apicomplexan protists also don't carry it! So what's the mechanism in those guys? Has eukaryotic EF1A reacquired the ability to deliver eRF1 and Dom34p? That would be another amazing story!
