A phyletic distribution of eukaryotic translation components has been proposed by Koonin and Aravind by recognizaing conserved predicted globular domains within each factor. An interesting feature is that many of these factors are shared between eukarya and archea. Few are seen to have any counterparts within bacteria, with the exception of eIF1A, eIF2, and eIF5A, which are universally highly conserved. These three factors are believed to have maintained a great proportion of their functions from a common ancestor in all life forms. As mentioned above, IF1 is the bacterial counterpart of eIF1A. Both of these factors contain a conserved nucleic-acid-binding oligomer-binding (OB)-fold domain, that is believed to mediate initiator codon recognition in all translation systems (Settle et al., 1997; Battiste et al., 2000).
Horizontal gene transfer has been implicated for the transfer of genes from archea to bacteria, however the functions of many of these factors seems not to be conserved between archael and eukaryotic and their bacterial orthologs. Archea and eukarya share a significant number of initiation factors, including eIF2 and eIF2C and eIF6. These factors are believed to form the core initiation system that was part of the common ancestor of archea and eukaryotes (Aravind and Koonin, 2001).
In addition to many eIFs containing alpha helices, many domains of eIF subunits are RRM. These are highly seen within eukaryotes and the folds within these proteins are ancient and predate the radiation of the three divisions of life. This distribution of structural features has stimulated the idea that eukaryotes have 'snatched' domains for translation factors from novelprotein families that were evolving by rapid sequence changes of pre-existing folds and therefore propagayed an expansion early in the evolution of eukaryotes. Acceptance of these novel domains resulted in their fixation due to their critical roles.
In conclusion, it seems that all three of the domains of life share some similarities in initiation factors to one another, as well as with ribosomes, depicting that the underlying mechanism of translation intiation 'should' be fairly similar. However, differential divergence and conservation of selected domains have entitled eukaryotes with a more complex and versatile system of translation initiation. I believe that subsequent studies to further advance our knowledge on the evolutionary basis of initiation factors should prove useful in our understanding of the translation initiation system and also to view the different proposed mechanisms, the scanning model and IRES initiation, in an evolutionary perspective to help clarify their differences and purposes.