This is another archival repost first written for the old blog in 2007.
On Monday I mentioned John Maynard Smith’s videos at People’s Archive. They really are marvelous, and you should watch them all. One of the topics he discusses is horizontal gene transfer in bacteria, a subject I discussed a few months ago in the context of the species concept. JMS discusses horizontal gene transfer in relation to the gene-centered view of evolution, or “selfish gene” model of evolution. This story starts with Jacob and Monad who, in 1961, discovered “operons” while looking at the mechanisms of gene expression in bacteria. Operons are a bunch of genes which participate in the same task, are situated next to each other on the bacterial chromosome, and share a single promoter (metadata telling the cell under what conditions the genes in that operon should be switched on). This was a very important discovery, but operons, it turns out, are not the default mechanism of gene expression. Other genes are solitary, with their own dedicated promoter, and even when genes cooperate in performing a task, they are often far apart on the chromosome, with their own identical copies of the relevant promoters. As is often the case in science, Jacob and Monad had answered one question, and discovered another: why are only some genes organised in operons?
From the point of view of gene expression, there are two main types of genes: the always-on (“constitutive”), and the only-on-when-needed (“regulated”). Constitutive genes are rarely, if ever, found in operons. The obvious story of operon evolution was therefore focused on gene expression: it is simpler to switch on a whole system in one go, than to switch on each component individually. But operons are only common in bacteria (though they have been reported in some eukaryotes). Eukaryotes also have regulated genes whose products work together in a system, they just don’t tend to bunch together or share promoters. So, if the gene-expression explanation is the correct one, it too raises yet more questions. Is this difference just a historical accident — unlikely, since operons have probably evolved many times independently in bacteria, but never in eukaryotes — or does it reflect some other difference between bacteria and eukaryotes?
Lawrence and Roth proposed another hypothesis for why some groups of genes formed operons, and while others did not — a hypothesis that happened to explain why eukaryotes do not have operons. Operons, they proposed, were optimised for facilitating horizontal transfer, a process that only occurs in bacteria, and which shuffles and shares genes, serving a purpose somewhat like that of sexual reproduction. The genes on a particular operon are all involved in the same task (or “selectable phenotype”), and most of the tasks which involve operons can not be performed efficiently without all of those genes being present. Horizontal transfer would therefore be pointless unless all of those genes were involved, and grouping of related genes would therefore be favoured by selection.
It is often difficult to determine the truth about how things evolved: creationists ask “were you there?”, and biologists are always accusing each other of making untestable “just-so story” explanations (at least, when they don’t like that particular explanation). Lawrence and Roth, however, believe that their model neatly explains several observations, while making testable predictions. For example, if the selective advantage of operons is in relation to horizontal transfer, we should expect them to have arisen for non-essential systems. All of the essential genes will already be present in every individual of every species, so little will be gained by transferring them. Additionally, the functions of the non-essential systems will, over time, be lost due to mutation (selection against them is, depending on the particular environment, “weak”), and once one of the genes is lost, all of the others will succumb, as their selective advantage depends upon all of the genes working together. This does not occur in essential systems — after all, any harmful mutation will be instantly lethal to that cell (selection against them is always “strong”). Thus, there may be an advantage to “renewing” the non-essential systems over time, but none for essential systems. This provides an alternative explanation for why we only see regulated genes in operons, as well as allowing us to make a prediction about the sort of systems we should see in operons: they will not be performing everyday tasks, but will be of only occasional use.
This turns out to be largely true. The operons that we observe are for systems that deal with unusual situations. The first operon to be discovered was the “Lac” operon, which allows the carrier to thrive on the sugar lactose — something that a bacterium may not encounter very often. Other operons carry systems that covey resistance to anti-bacterial agents — e.g. by breaking down the antibiotic, pumping it out of the cell, or producing alternative versions of the systems that the antibiotic targets. This is part of the reason why antibiotic resistance spreads so fast, and why we are currently faced with the problem of hospital “superbugs”.
Lawrence and Roth describe operons as “selfish”. If there were no horizontal transfer, non-essential functions would very likely be lost over time, as they acquired mutations during those phases when the environment meant that they possessed no immediate selective advantage. However, by participating in horizontal transfer, selection events in those few individuals which do encounter an unusual environment — e.g. rich in lactose — can ensure that functional allelesout-compete mutants in the population as a whole. The genes in the operons depend on transfer for their long term survival, but the benefits for the individual cells participating may be small — they may never encounter the relevant environment. The individual cells are just vehicles for the selfish replicators.
There is much more to say on this topic, including some interesting pieces of evidence for and against, and the objections and caveats that have been raised. But those must wait for a future post. For now, I’ll leave you with this last thought. Operons illustrate once again an important fact in biology: from the point of view of both development and evolution, the data contained in genes is a lot more than just a sequence of nucleotides. Previously, I have mentioned that there is data in gene-product concentrations and chemical modifications to DNA; to these we can add data in the form of genome organisation.
- J.G. Lawrence and J.R. Roth, 1996. Selfish Operons: Horizontal Transfer May Drive the Evolution of Gene Clusters. Genetics 143(4): 1843–1860.