Results from our genomic comparisons indicate that several of the key proteins involved in the different RanGTP-dependent RNA export pathways are conserved across most eukaryotic lineages, and thus we infer that orthologs of them were highly likely to have been already present in the LECA. Examples of these are the exportins XPO1 and XPOt, Nmd3 and most of the proteins involved in the Ran cycle. Despite the relatively strong level of conservation, we also document how these export mechanisms vary in a lineage-specific manner as a consequence of the differential gene gains and losses, as documented by the secondary loss of XPOt in C. elegans
, and the emergence of Arx1 in Fungi and of PHAX in animals. In agreement with inferences drawn from studies of components of the cytoskeleton, endomembrane, NPC, and spliceosome [17
] a number of key RNA nucleocytoplasmic transport factors can be traced to the LECA. This would suggest that the different RanGTP-dependent RNA export mechanisms were already present in the LECA.
By contrast, our analyses indicate that several of the proteins involved in the RanGTP-independent export pathway from opisthokonts are lineage-specific innovations. The mRNA export pathway is the most complex and the least conserved among those examined in this study. We found CHS for few of the yeast proteins involved in mRNA export in most eukaryotic lineages, which suggests that mRNA export among them is different from what is observed in yeast. We document the acquisition of several lineage-specific innovations in the mRNA export mechanisms of opisthokonts relative to the other supergroups included in this study. It seems plausible that the evolution of a RanGTP-independent mRNA export pathway in opisthokonts might be related to the observed differences in the regulation of gene expression. In Amoebozoa, Chromalveolata and Excavata, most regulation is post-transcriptional [68
], whereas in human and yeast the presence of both transcriptional and post-transcriptional regulation is linked to a more refined control of gene expression. One possibility is that having an mRNA export pathway fully separated from the other RNA export pathways increases of the range possibilities of gene-specific control. This would be particularly important to allow both coordination and versatility in gene expression control [69
], as well as to open additional opportunities to connect active transcriptional sites to the NPC, allowing a fine control of gene expression in yeast and human [71
]. Functional data regarding mRNA export in excavates and chromalveolates is still limited, and despite the fact that some studies suggest if follows less complex routes [45
], the possibility that alternative specializations haven arisen in these groups cannot be discarded.
Taken together, our analyses suggest that the LECA possessed most of the basic components of the different RanGTP-dependent RNA export mechanisms, which are also well conserved among the different eukaryote supergroups included in this study. In addition, we also show that some of the major components of the RanGTP-independent mRNA export pathways can also be traced to the LECA, but that several of the proteins that play key roles in opisthokonts derive from lineage-specific innovations. It is likely that early in eukaryote evolution a single generalized ancestral exportin was probably responsible for nucleocytoplasmic transport of all RNA species (Figure ). Prior to the emergence of the LECA, orthologs of XPO1, XPO5, and XPOt would have already emerged and specialized in transporting a subset of the RNAs (Figure ). We speculate that in the LECA mRNA might have been exported by XPO1 in RanGTP-dependent manner, as in excavates in the genus Trypanosoma. The emergence of the RanGTP-independent mRNA export pathway in opisthokonts coincides with the origin of several lineage-specific innovations, and might be related to refinements in the regulation of gene expression in this supergroup (Figure ).
Figure 4 Model proposed of the evolution of the different RNA export pathways throughout eukaryotic history. (A) Early in the history of eukaryotes a single ancestral exportin similar to XPO1 was in charge of nucleocytoplasmic export of all RNAs in a RanGTP-dependent (more ...)