Evolutionary Significance

Both circular and linear retroplasmids show little direct relationship to other retroelements. As mentioned, the retroplasmid RTs are deeply rooted in RT phylogeny and are possible progenitors to a wide range of retroelements (Wang and Lambowitz 1993a; Walther and Kennell 1999; Eickbush and Malik

2002). Evidence in support of their putative ancient origin also derives from the distinct enzymatic properties of their RTs, such as the mechanistic similarity of the pMaur-RT to RNA-dependent RNA polymerases. The finding that the pFOXC retroplasmids have 3' terminal repeats similar to those of chromosomal telomeres and replicate via reverse transcription makes them leading contenders to be precursors of telomerase (Walther and Kennell 1999). In addition, retroplasmid clothespin DNAs contain a 5'-linked protein, suggesting a potential relationship to hepadnaviruses (e.g., hepatitis B virus) that have covalently attached terminal proteins that serve as primers for cDNA synthesis (reviewed in Ganem and Schneider 2001). These properties help to justify the position of retroplasmids at the base of retroelement phylogeny and the collective features of circular and linear retroplasmids suggest that they are related to primordial genetic elements that could have been a common ancestor of all known contemporary elements that replicate via reverse transcription. Further studies of this interesting group of genetic elements will likely shed more light on the origin of retroelements as well as events involved in the transition from the RNA world to the DNA/protein world.

It is clear that mitochondrial plasmids, and circular retroplasmids in particular, are able to interact directly with mitochondrial DNAs. Variant forms of Mauriceville and Varkud plasmids readily integrate into mtDNA and cause senescence (Akins et al. 1986, 1989; Chiang et al. 1994; D'Souza et al. 2005). Consequently, the Neurospora retroplasmids serve as excellent tools to understand nuclear-mitochondrial interactions as well as host mechanisms that developed to control or cope with invasive genetic elements. Interestingly, the degree and manner in which plasmids affect their hosts appears to be host-specific and the relative rates of senescence vary widely between the nearly identical Mauriceville and Varkud plasmids of N. crassa and N. intermedia, respectively (Fox and Kennell 2001). Variation in the behavior of plasmids and ways in which hosts respond to plasmid integration may partly explain why plasmid-induced senescence is not commonly detected outside of Neu-rospora spp.

Integration of the linear DNA plasmids has also been reported to linearize portions of mtDNAs of Podospora anserina (Hermanns et al. 1995; see also Chap. 8, in this volume), N. intermedia (Bertrand et al. 1985, 1986), Zea mays (Schardl et al. 1985) and Physarum polycephalum (Nomura et al. 2005). While there is currently no evidence that linear retroplasmids have integrated into mtDNAs, it cannot be discounted. Recent studies suggest that mtDNAs of most eukaryotes exist as various types of linear molecules (Nosek and Tomaska 2003). At least six different telomeric structures have been identified, several of which share features with the termini of linear plasmids. This has led to the hypothesis that mt plasmids played a major role in the development of linear mitochondrial genomes (Nosek and Tomaska 2003). Mitochondrial retroplasmids may not only have had a significant impact on mitochondrial genomes, but could have played a role in the development of eukaryotic cells as well. In a recent hypothesis paper, Martin and Koonin (2006) proposed that the invasion of mobile group II introns from the a-proteobacterial ancestor of mitochondria into the presumed archaebacterial genome of the primitive eukaryote created a selective pressure for the development of the nuclear membrane. They speculate that the nuclear membrane evolved as a mechanism to separate transcription from translation to ensure that only fully spliced mRNAs are translated. Building on this idea, it is possible that mitochondrial retroplasmids have interacted with nuclear DNAs and, over time, have had considerable influence on nuclear chromosome structure (Nosek et al. 2006) and the development of mechanisms used to exclude the transmission of invasive mitochondrial genetic elements (Ken-nell, submitted for publication).

Acknowledgements We thank Casey Dillman, Matthew Althage, and Rahul Malireddy for assistance in the phylogenetic analysis.

0 0

Post a comment