Fine Reading: Nematodes & HGT

by Elio

The larger the catalog of sequenced genomes, the greater the chance for seeing horizontal gene transfer (HGT) in action. Although reports of HGT between prokaryotes and eukaryotes have appeared sporadically, it seems that incontrovertible evidence is only beginning to appear now. Indeed, the genomes of nematodes reveal that genes have been acquired from bacteria a number of times. A recent, one-page review of this subject discusses examples of this phenomenon.

Evidence for inter-domain HGT seems to be on its firmest footing when it comes to nematodes. Given the intimate contact between bacteria, fungi, and many soil-dwelling worms (many of which are bacteriovores), this should not be surprising. Nor is it unexpected that the genes that have been transferred encode enzymes that can break down otherwise indigestible sugars, such as glycosides and cellulose, that are constituents of plant food that the worms ingest. Little is known as yet about how the foreign genes “adapt” to their new genomic context, although it is known, at least in some cases, that they are indeed expressed. I can predict with confidence that this subject will be fertile territory for many years to come

Given that HGT provides endless opportunities for evolution, the question begs: Why does it not take place more often between prokaryotes and eukaryotes? Aren’t prokaryotes and their viruses a nearly endless repository of potentially useful genes? This grandmother of Talmudic Questions brings up the stark fact that organisms must be highly selective in what foreign genes they allow into their genome, lest they lose their identity and potentially their survival skills. But even without getting into the thicket of mechanisms designed to stabilize genomes and keep them from going astray, I still wonder why HGT between prokaryotes and eukaryotes (in either direction) appears to be something that may happen with some frequency but that is only rarely triumphant.

Comments

[Nathan Myers]

Merry hints at my immediate reaction: when we encounter evidence a foreign inclusion in a euk, we are inclined to attribute it to viral activity. To distinguish them, I suppose we would have to sequence very many closely-related nematode species, and note wholesale gene insertions in one but not in clade neighbors. (I assume viruses that infect nematodes are not especially selective.)

But might not a retrovirus be the favored vehicle of a codon-sequence fragment from a bacterium to a nematode, anyway? Could the two be hard to distinguish because they are most frequently the same event?

[Ryan F]

In response to Merry's last point, there was a paper a year or year and a half ago from Hank Seifert's lab about the presence of a line element in Neiserria's genome? I guess it isn't a gene, just a genetic element so no splicing required.

[Merry]

Elio raised a most interesting question here, to which I could not resist posting an impromptu response. I do not find it surprising that demonstrable examples of HGT from prokaryotes (proks) to eukaryotes (euks) are few and far between. Here are a few reasons that come to mind.

First there is the matter of numbers and probability. Let’s consider nematodes in particular because they were featured in Elio’s discussion and because, with their estimated global population of 1022, they are the most abundant multicellular organism on Earth. That’s a lot of worms, but they’re outnumbered a hundred million to one by the 1030 proks. Moreover, what matters here is not just the number on the planet at any one moment, but the number of generations per year, where the proks also score higher. And consider that proks were swapping genes around for billions of years before there were nematodes. Lastly, whenever a useful gene transfers into a prok genome, it has the possibility of hitting the evolutionary jackpot and becoming fixed in a clonal population. In a nematode, it has a chance only if it happens to be in a germ cell, and only if that germ cell happens to give rise to a successful gamete.

As Elio noted, viruses are another rich source of possibly useful genes, but most of the time we wouldn’t know a viral gene when we saw it in a euk genome. Recognition requires similarity to some known gene, and most genes found in viruses collected from any natural environment are “novel.”

Suppose a segment of prok DNA containing an open reading frame (ORF) has safely traversed the cytoplasm of a euk cell, finagled its way into the nucleus, and recombined into a chromosome. Now what? A “gene” is not simply a string of codons that can be plugged into any cell and be expected to function. It also includes regulatory elements that govern the initiation and termination of transcription, and these mechanisms differ between proks and euks. At the level of translation, ribosomal binding sites differ. Even the ribosomes themselves differ. Supposing all these hurdles are surmounted and a functional prok protein is made in a euk, it must now work alone, isolated from its former operon or regulon. To spread in the population, it (usually) must confer some fitness benefit in a heterozygote. All of this makes success less and less likely.

What about transfer in the other direction, from euk to prok? This comment is already too long, so I’ll just mention one obvious issue, i.e., that most protein-encoding euk genes are interrupted by introns of a type that requires the spliceosome for processing, an apparatus that proks lack. And beside, metabolically speaking, most everything a euk can do, some prok already knows how to do…and likely can do it better. Proks have access to a large shared gene pool, so why bother with the hassle of importing euk genes?

I bet you have some other ideas on this, so please join the discussion by posting a comment.

A rejoinder from Elio:

Merry makes a thoughtful case for HGT being a rarity between proks and euks. The obstacles for establishing functional genes in a phylogenetically distant host are just too pervasive. Well, who am I to argue that. But given the huge number of possible encounters (Merry nicely points out that a humongous number of nematodes crawl around this Earth), the known obstacles, formidable though they are, ought to be counterbalanced by the selective advantage that transfer of certain genes would confer. Consider herbivory among nematodes alone. Wouldn't the acquisition of bacterial genes for cellulases and chitinases, among many others, provide a huge advantage to the host? Now it turns out that this actually happens. The question remain, why does it happen so so seldom? And, while I'm at it, why don't we have such enzymes in our digestive juices?