A friend writes to draw our attention to these elements in the genome:
They have a mechanism by which they can mutate a specific part of the DNA, in a pattern and rate that’s different from random mutation.
How that happens: DGRs have a reverse transcriptase, or an enzyme that can synthesize DNA from RNA (as opposed to RNA from DNA, as the central dogma suggests). During this reaction, all the adenines of a specific part of their target protein gets mutated to some other nucleotide at a very high rate. To put some numbers on the phenomenon- the rate of mutation is 10,000 times higher than the RT enzyme of HIV, the (other) most error-prone enzyme known to man.
How it’s beneficial to the host: DGRs often target proteins that determine antigenicity, which require a lot of diversity for successful binding to their targets (cf. generation of antibodies in humans is such a diversity-generating process). So they act to provide diversity in a part of the protein that needs diversity. Some evidence suggests it might be a more general way for bacteria to generate diversity, regardless of functional context (see below).
How frequent is it: Research on DGRs have just begun, but they’ve been shown to be present in all bacterial phyla. A computational analysis in 2019 revealed the presence of about 1000 unique non-redundant DGRs in >80,000 bacterial genomes. The authors pointed out that there’s evidence to the effect that DGRs are not only involved in immunity or antigenicity, but also might be a universal, general process of generating diversity among bacteria that has consequences for many other biological functions. I thought this latter conclusion was somewhat weakly argued, however.
A 2017 paper in Nature Microbiology also argued that these elements were particularly enriched in uncultivated bacterial and archaeal phyla that contain minimal genomes and have deficiencies in their functional pathways, which further points to the importance of DGR in generating novelty in organisms that need it.
To sum up, Diversity-Generating Retroelements are quite frequent features in all sorts of bacteria that cause targeted, non-random mutations in very specific parts in the genome, leading to the host’s adaptive advantage. This element is a very obvious example of natural genetic engineering.
This sounds about as random as the Global Positioning Satellite system.