PVC bacteria are not half-way to being eukaryotes

Over the past few years, manuscripts have been appearing in the literature that have been suggesting that PVC (Planctomycetes, Verrucomicrobia, Chlamydia) bacteria are in many ways like Eukaryotes and consequently, they are an excellent system for understanding eukaryote biology and indeed this group of bugs actually gave rise to modern eukaryotes.

This is complete nonsense.

So, where does this idea come from?  Well, some PVC bacteria have their DNA enclosed in a membrane and under a microscope this looks a bit like a eukaryote nucleus, they have ‘compartments’ in their cells and so do eukaryotes and they lack some things that eukaryotes lack -peptidoglycan, for instance.

However, the key to understanding evolution is the understanding of the difference between homology and analogy.

Homology means one specific thing – descent from a common ancestor.  It does not mean similarity.  Famously, Walter Fitch said that homology was like pregnancy: you cannot be 80% pregnant and two evolving entities cannot be 80% homologous.  They are either homologous or they are not.

Now that we have a good definition of homology, we have the seriously difficult problem of identifying homologs.

For morphological characters, making homology statements is usually a matter of having a really careful look at the organs in question.  Getting a better microscope, for instance, if you need a greater level of detail.  We expect that morphological characters that are homologous will display similarity in many of their details – their structure, their function etc.

For molecular characters, it can turn out to be more difficult. We only have four nucleotides or twenty amino acids.  We usually try to make statements of homology by using a similarity search of databases and deciding that if two sequences showed a higher level of similarity than you might expect to see by random chance, then we make a leap of faith and infer that they are homologs.  However, our homology statements can be confounded by a number of issues.  Some sequences might manifest high levels of similarity as a consequence of simple repetition of residues.  They might really be homologues, but sequence divergence has been so great that sequence similarity is not longer apparent.  They might also not be homologues, but convergent evolution has resulted in compositional similarity that makes them look homologous.

In order to correctly infer homology for molecular sequence data we need careful and rigorous analysis of the sequences and maybe their structures and when no other explanation can be offered for their similarity, then and only then might we agree that two sequences are homologous.

Our next task when we discover homologous sequences is to ask how they came to being in the organisms in which they are seen.

For many genes, the reason why two organisms have homologous in their genomes is because the gene copies were passed down from one generation to the next at cellular or organismal reproduction time.  However, for other organisms, the reason why they have homologs is because these homologues were acquired via horizontal gene transfer.

Therefore, if we are to say that the PVC bacteria are of special interest for studying eukaryotic origins, then we need to be able to say that they are the closest relatives of extant eukaryotes and that they share genes and structures because of vertical acquisition of genes.

Ultimately, when you look at the data, you can see that PVC bacteria do not share a lot of homologous structures exclusively with eukaryotes and are therefore no more useful in the study of eukaryotes than any other bacterium would be.

If you want more details on this, please see our recent paper on the issue:

McInerney, J. O., Martin, W. F., Koonin, E. V., Allen, J. F., Galperin, M. Y., Lane, N., Archibald, J. M. and Embley, T. M. (2011), Planctomycetes and eukaryotes: A case of analogy not homology. BioEssays. doi: 10.1002/bies.201100045. [link]