Rolling down the road
Today’s post is from regular guest blogger:
In my idle moments (alas, too few these days!) I often try to think up lists of rock songs with a virus theme: you know, like “Cucumo” by the Beech Boys… “I got them ol’ burnin’, hurtin’ herpesvirus blues again, mama” and “Lord, won’t you buy me some Efivarenz” by Jannie Joplin and the Zoviraxes… “Mama, weer all crazee now” by Noddy Holder and the Bornavirus Collective… but I pretty quickly run out of viable examples, and move on to more productive behaviour. But because my daughter keeps making me watch “Absolutely Fabulous”, with “This wheel’s on fire” as the theme song, which I like a lot, and I work on viruses with single-stranded circular DNA genomes, I just had to work it in somewhere…and then my Honours student Guy Regnard did a fantastic journal club paper on geminiviruses replicating in E. coli, and I saw my chance.
Geminiviruses are probably the single most important group of viral plant pathogens right now (though there is some debate about potyviruses). What is more, their continuing emergence means that there is an ever-increasing number of them in genome databases – and in people’s fields, which lab virologists tend to forget all too easily. The viruses are fascinating for a number of reasons, one being their unique virion morphology, and the other the fact that they seem to recombine so readily. The most important geminiviruses in terms of total yield losses and sheer numbers are undoubtedly the begomoviruses (named for Bean golden mosaic virus). These viruses may have one or two similarly-sized genome components, with the optional B component dependent on the A for replication, but providing movement functions which are necessary for the viruses with a B component.
Typical genomic organization of begomoviruses. ORFs are on the virion (V) or complementary (C) strand. The +-strand ori TAATATTAC-containing stemloop is shown. TrAP = transcriptional activator protein; REn = replication enhancer, MP = movement protein, NS = nuclear shuttle. [A]V2 ORF (in grey) not present in New World begomoviruses.
Replication of geminiviruses is apparently mostly via a rolling-circle mechanism (RCR), similar to many ssDNA phages and not a few bacterial plasmids: in fact, it is pretty generally accepted that the replication machinery of geminiviruses and nanoviruses from plants, circoviruses, anelloviruses and even parvoviruses from animals, and the aforementioned phages and plasmids, all has a common origin – which may extend to the mobilisation mechanism used by bacteria like Agrobacterium tumefaciens. A general scheme for ssDNA virus replication can be seen here.
Geminiviruses, like other ssDNA entities, have a rep gene, producing a Rep protein: in plants, this is expressed from a double-stranded replicative intermediate form of the genome (RF), and cleaves the genome (+) strand at a specific ori sequence, binds to the free 5′ end, and then mediates ligation of the newly-displaced (+) strand. The accumulation of ssDNA seems to depend on the production of coat protein, which probably sequesters nascent ss(+)DNA, as CP- mutants produce no ssDNA.
There is a fair degree of specificity in all this, with viruses having a specified host range, and Reps having specificity for a narrow range of virus genomes. Thus, one might quite reasonably surmise that ssDNA viruses have been speciating with their hosts over aeons, and so would have evolved to be quite specific with regard to just what machinery they interact with.
And here’s where things get strange: apparently this is not necessarily true at all.
It has been known for many years that some geminivirus promoters work in E. coli; subsequently it has been found that some greater-than-unit-length begomovirus genomic clones seem to release viral RFs from their parent plasmids in E. coli and in A. tumefaciens, and in budding yeast cells. Lately, it is clear that the mammalian Porcine circovirus (PCV) also replicates in E. coli in the same way. Presumably, all that is required is (1) expression of functional Rep, (2) replicational or possibly recombinational release of a viral genome, (3) rolling circle replication. While this is academically fascinating, and potentially provides useful platforms outside of the natural host to study Rep-genome interactions, it is still homologous interactions (cognate Rep/DNA interactions) producing the result.
Now things have got more interesting: the Wu et al paper I started out with describes how an Ageratum yellow vein virus (AYVV) single-copy DNA A genome cloned in an M13-derived plasmid in E. coli – one RCR genome cloned into another one – produced single-stranded (+)- and (-)-sense AYVV genomes, presumably complexed with the M13 pV ssDNA-binding protein. Leaving aside the fact that this is impossible in terms of the standard models of replicational or recombinational release of unit genomes – and the models they present are largely hand-waving – it is a fascinating example of what amounts to cooperation between a prokaryote and a eukaryote virus sharing a very distant common ancestor, in a host of one of them. It is clear that M13 phage proteins are interacting with geminivirus DNA – and possibly even with the AYVV Rep – in order to effect a result not achievable with the latter alone.
The authors go on to speculate quite reasonably about what this could mean in the greater scheme of things, including whether or not it could mean that geminiviruses actually do replicate in the bacterial-descended chloroplasts, as has been claimed a couple of times: for example, in 1987, Holger Jeske’s group detected genomic ssDNA of Abutilon mosaic virus in chloroplasts of ornamental abutilon. It is also interesting that Maize streak virus (MSV) very thoroughly mangles the chloroplasts of infected maize, in the cells where virus is found.
Why this is all so interesting is that it shows that virus proteins separated by an evolutionary gulf of possibly billenia can still interact with similarly-replicating genomes. This opens up all sorts of possibilities for virus evolution and even virus survival: maybe plant viruses can survive inside endophytic bacteria; phages may be able to find refuge in plant or even animal cells; circoviruses may hide inside gut bacteria. Which means that the speculation by Mark Gibbs and Georg Weiler, that animal circoviruses derive from a complex recombination in animal cells between a plant nanovirus (ssDNA) and an animal calicivirus (ssRNA), may not be so far-fetched after all…except I think it all happened inside an insect, given that they have been associating with plants a lot longer than vertebrates have.
So what all this information means, possibly, is that there is the potential for some very deep interactions between different virus genomes and their proteins, in a very wide range of hosts. Which could make a lot of nonsense of a lot of what we think we know about how viruses and their hosts evolve.
So notify my next of kin, this wheel shall explode… current paradigms, hopefully!
Apologies to Bob Dylan and Rick Danko
If you would like to be a guest blogger on MicrobiologyBytes, click the Guests tab above.