Is lateral gene transfer (LGT) Lamarckian?

There's an interesting discussion going on about lateral gene transfer (LGT) in eukaryotes. LGT is the process by which DNA from one species invades the genome of another species. It was apparently very common among primitive bacteria several billion years ago and it's still quite common in modern bacteria.

There are many reports of LGT in eukaryotes but some of them seem to be due to contamination from bacteria rather than true LGT. Many scientists are skeptical of these reports; notably Bill Martin (Heinrich Heine Universität, Düsseldorf, Germany) who suggests that almost all of them are artifacts and lateral gene transfer in eukaryotes is extremely rare [see Lateral gene transfer in eukaryotes - where's the evidence?].

Andrew J. Roger studies deep evolution at Dalhousie University in Halifax, Nova Scotia, Canada. He has vigorously defended the existence of LGT in eukaryotes (see his comments in my earlier post).

In addition to this debate about the existence of LGT in eukaryotes, there's a discussion about whether lateral gene transfer in eukaryotes, if it exists, is a fundamentally Lamackian process. The latest exchange took place with two letters published in the May issue of Nature Ecology & Evolution.

Eukaryote lateral gene transfer is Lamarckian, William F. Martin
[doi: 10.1038/s41559-018-0521-7]

Reply to 'Eukaryote lateral gene transfer is Lamarckian,' Andrew J. Roger
[doi: 10.1038/s41559-018-0522-6]

Martin is mostly concerned about adaptationist claims of LGT in eukaryotes. He references a number of papers that make such claims; for example, Hirooka et al. (2017) claim that the green alga, Chlamydomonas eustigma, recently adapted to an acidic environment by taking up genes from bacteria or other eukaryotes. These genes are not present in related species that have not adapted to an acidic environment. Here's how Martin describes such studies.

The core of eukaryotic LGT adaptation claims is that eukaryotes lack the genetic material required to survive in particular environments and acquire the genes needed in order to access those environments from those that already live there. Lamarckian? Yes. In eukaryote LGT adaptationism, the environment is the source of natural variation, not the evolving organism itself.

I see Martin's objection as essentially an attack on naive adapatationism invoking LGT as a mechanism for adaptive change. The fact that many of those claims of LGT are probably false is only part of the problem. The other part is the adaptationist claim to justify recent and abundant LGT.

It's true that there are Lamarckian characteristics behind these (probably false) claims. However, that's probably not the best way to criticize the hyper-adaptationism that's associated with false conclusions about LGT in eukaryotes. In addition, Martin clearly goes too far when he implies that all eukaryotic LGT is Lamarckian and false.

Andrew Roger takes up the challenge implied by Martin's exaggeration. The gist of his argument can be found in the first two sentences of his letter ...

Martin argues here and elsewhere that nearly all claims of lateral gene transfer (LGT) into eukaryotic genomes are untrue, and that accompanying narratives are fundamentally 'Lamarckian.' Some eukaryotic claims have proven false, but this does not mean that most are. Although rare, gene transfers have had a profound effect on the evolution of traits in eukaryotes.

He goes on to explain the "proper" view of LGT.

Chunks of DNA are accidentally incorporated into chromosomes creating genetic variation that is neutral, deleterious or, in rare cases, beneficial. If they enhance fitness, acquired genes are likely to be fixed in the population by natural selection. Any reasonable adaptive LGT claim has a similar etiological narrative that respects modern evolutionary principles.

Here's the problem. Martin knows the proper role of LGT but that's not what he was criticizing. He was criticizing many "unreasonable" adaptive LGT claims but he went too far by implying that all claims were of this type.

The real issue here is that a great many claims of LGT in eukaryotes are probably false—I suspect that most are false. We should not let bickering over Lamarck obscure that fact. I wish Bill Martin had not raised the issue about Lamarckian evolution.
The world is not inhabited exclusively by fools and when a subject arouses intense interest and debate, as this one has, something other than semantics is usually at stake.

Stephan Jay Gould (1982)
As Stephan Jay Gould once said, when scientists squabble over semantics, there's usually something more at stake. In this case, it's the origin of basic metabolic processes. Martin is one of a group of scientists who propose that primitive eukaryotes were facultative anaerobes. They were capable of growing and reproducing in the presence of oxygen and in its absence. They acquired this capability because the primitive mitochondrial endosymbiont had all of the enzymes necessary for both types of metabolism. In some lineages, the ability to carry out anaerobic metabolism has been lost. This hypothesis is sometimes called the "hydrogen hypothesis" because an important terminal electron acceptor is protons that can be reduced to form hydrogen.

Here's how Müller at al. (2012) explain the controversy over the origin of anaerobic metabolism in eukaryotes (e.g. protists).

For the origin of anaerobic energy metabolism in protists, the question is, Were the genes present in the single eukaryote common ancestor, or do they clearly reflect multiple origins, and if the former is true, does their single origin coincide with the origin of mitochondria? This has in turn given rise to two main competing alternative hypotheses for the origin of anaerobic energy metabolism in protists: (i) the enzymes were present in the eukaryote ancestor and were inherited vertically by modern groups, or (ii) they were lacking in the eukaryote ancestor (which would then implicitly have been a strict aerobe) and were acquired in different eukaryotes groups independently via lateral gene transfers (LGTs). Those views generated very different predictions with regard to the evolutionary patterns of the underlying genes.

The authors, including Bill Martin, conclude that the enzymes were present in the early mitochondrial ancestor although they don't preclude that 1-2% of the genes could have been acquired by LGT.

Andrew Roger has proposed that many of the genes required for anaerobic metabolism were acquired by LGT after the initial symbiotic event (Hug et al., 2010).

Thus, the two participants in the exchange of letters are on opposite sides of the bigger debate on the origin of genes for anaerobic metabolism in eukaryotes. Bill Martin favors the hydrogen hypothesis and the idea that the primitive bacterium giving rise to mitochondria was a facultative anaerobe carrying the genes necessary for anaerobic metabolism. These genes have been lost in many eukaryotes but the core genes all descend from a common ancestor. Andrew Roger is on the side of those who argue that anaerobic metabolism arose independently by LGT in many eukaryotic lineages. This is why he ends his letter in Nature Ecology & Evolution with ...

So why such resistance to LGT in eukaryotes? Endosymbiotic organelle origins and endosymbiotic gene transfer have been championed as dominant mechanisms in eukaryotic gene evolution. Indeed, the widely publicized 'hydrogen hypothesis' of eukaryogenesis depends heavily on assuming a mitochondrial ancestry of 'bacterial-like' enzymes of anaerobic energy metabolism in eukaryotes. Acknowledging LGT as an important mechanism provides an alternative explanation for such patchily distributed genes in eukaryotes that do not show the hallmarks of mitochondrial or plastid origin.

Keep in mind that if Andrew Roger is correct about how laterally transferred genes are eventually fixed in a lineage, then LGT must be very common because most events will not give rise to an adaptive advantage. They will be neutral or disadvantageous. Bill Martin believes that most claims of LGT in eukaryotes are false—he's probably right about this—and that LGT must be quite rare. If Bill Martinis is correct then it's unlikely that LGT can account for all the examples of anaerobic metabolism in eukaryotes.

I'm not a big fan of either explanation for anaerobic metabolism. The idea that it's the primitive condition that has been lost in may lineages seems a bit far-fetched given the patchy distribution in eukaryotes. On the other hand, using LGT to explain this patchy distribution of fundamentally similar enzymes activities seems equally unlikely.

The various tests of these hypotheses relay on sophisticated analyses of sequences that diverged billions of years ago. Both of these men (Roger and Martin) are experts in this field but they are pushing the boundaries of the field using algorithms that are incomprehensible to the average scientist.

---

Photos: The first photo is of Bill Martin and me having coffee at Tim Hortons in Toronto last year. The second one is Andrew Roger and me at the "Tree of Life" meeting in Halifax in 2009.

Hug, L.A., Stechmann, A., and Roger, A.J. (2009) Phylogenetic distributions and histories of proteins involved in anaerobic pyruvate metabolism in eukaryotes. Molecular Biology and Evolution, 27:311-324. [doi: 10.1093/molbev/msp237]

Müller, M., Mentel, M., van Hellemond, J.J., Henze, K., Woehle, C., Gould, S.B., Yu, R.-Y., van der Giezen, M., Tielens, A.G., and Martin, W.F. (2012) Biochemistry and evolution of anaerobic energy metabolism in eukaryotes. Microbiology and Molecular Biology Reviews, 76:444-495. [doi: 10.1128/MMBR.05024-11]