Horizontal Gene Transfer: A Force for Bacterial Speciation?

Bacterial evolution is one of those topics usually sidelined to a footnote in introductory biology classes. We all learn about Mendel's peas, Darwin's finches, and peacock sexual selection, but bacteria are just too foreign. They're subject to the same force of natural selection as these macroscopic examples, but since they don't have sex things get more complicated: how do you define what constitutes a species? Oh, the teacher says, but actually they do have sex (in the form of horizontal gene transfer), and that's even more complicated because the genomes of your non-species get all mixed up!

If you couldn't tell, some clear thinking on this subject is definitely in order, and I was pleased to come across an innovative paper¹ that does just that. Before delving in, let's step back and get a sense of the bigger picture. There's a big difference between a human genome and a bacterial genome. Like all sexually reproducing species, our genome is a conglomeration of genes from our two parents (and thus our four grandparents, eight grandparents, and so on. Eventually this power-of-two rule falls apart because two or more of your 2ⁿ n-grandparents are one in the same!).

So we're not clones of our parents, and our genes are more or less mixed together, so that you don't have huge chunks of DNA from a single parent (biology students might know this as "independent assortment.") But in bacteria, offspring are genetic clones of their single parent. In fact, it's a bit dubious to use the word parent and offspring, since in order to reproduce, a bacterial cell divides into two more or less equivalent halves. The important thing is that (according to conventional wisdom) in sexual species, genes get mixed together across the generations, and in asexual species, they don't.

However, we know that there are benefits to sexuality (it is all but ubiquitous among us multicellular life forms). Without recombination, the genomes of asexual species are all or nothing. Sure a good mutation might arise now and then, but it's stuck with its incidental compatriot genes. A good mutation paired with an equally bad one won't be visible to natural selection. Alternatively, in sexually reproducing species, the good mutation will by chance find itself in genomes with other good genes, and that individual will have higher fitness than its competitors. In other words, it's much easier to rid bad genes from the gene pool.

But the basis of this paper is that that picture is incomplete. Bacteria do exchange genes, and we want to understand the consequences of this. How does horizontal gene transfer compare to and differ from plain old sexual reproduction? At first glance they're pretty similar, as both are gene mixing processes that make genetic novelties more accessible to selection. But, due to the fact that horizontal gene transfer occurs between different species, this is oversimplifying things, as a perspective piece on the article² notes:

. . . imagine that acacia trees could exchange DNA with lions and that the resulting new tree developed "limbs" that allowed them to attack grazing giraffes. This is in a sense what prokaryotes do all the time.

Bacteria are a really ancient and diverse group (taxonomically at the same level as all eukaryotes), so this analogy isn't as much of a stretch as you might think. The lions and acacia trees of this study are 20 strains of Vibrio cyclitrophicus, a marine bacterium. When the team filtered seawater to isolate the bacteria, some were found in larger particle size fractions, and others in smaller particle size fractions. Divergent evolution is happening-right now-between the large particle strains and the small particle strains. But what does divergent evolution mean for a bacterium? The old picture of bacterial populations as co-existing clones is fading.

Instead, as populations become ecologically separated, gene flow through horizontal gene transfer decreases (see figure at top). The solid lines represent lineages of bacteria. The ancestral bacterium of this lineage had traits symbolized as purple. Gene transfer between coexisting bacteria occurred, shown as skinny arrows. However, random mutation independently produced a green and a red type bacterium suited for a different niche (thick arrow). These traits spread through a subset of the population through horizontal gene transfer. As the two populations become ecologically separated, the rate of horizontal gene transfer within each population far exceeds the rate between them. This is what the authors refer to as "ecological differentiation."

To my eye, the clearest representation of this model within their data is a comparison between flexible genome sequences across strains. The flexible genome is part of the overall genome that varies a lot between different strains. As you can see in the heat map, when comparing strains isolated from small particles (S) against those isolated from large particles (L), there is a striking difference in genome content. Warmer colors mean that similarity is higher, so strains with similar ecology are swapping genes more frequently amongst themselves than with unlike strains. The idea is that horizontal gene transfer fosters genetic homogeneity within strains sharing similar ecology, but creates a genetic barrier that over time increases diversity between strains of dissimilar ecologies.

This is tantalizingly similar to how sexually reproducing populations speciate, where geographical (or ecological) partitioning of one population into two creates two distinct gene pools. It's just that bacteria are likely more willing to continue exchanging genes with their old friends once the partition gives way.

Image Credit: Figs. 3B and 2B from Shapiro et al. (2012)

References:

1. Shapiro, B. J. et al. Population Genomics of Early Events in the Ecological Differentiation of Bacteria. Science 336, 48-51 (2012).

2. Papke, R. T. & Gogarten, J. P. How Bacterial Lineages Emerge. Science 336, 45-46 (2012).

3. Brehm, D. Study Shows Unified Process of Evolution in Bacteria and Sexual Eukaryotes. MIT News. 5 April 2012.