Is there a tree of life?: Gene transfer between fungal species

There is a group of what I call “postmodern evolutionists” who loudly claim that, for one reason or another, the modern synthetic theory of evolution has to be thrown out the window.  Epistasis, modularity, adaptive mutation, genetic assimilation—these are the buzzwords that will supposedly dismember modern Darwinism.  On closer examination, though, one always finds that the theory remains pretty much intact: that these phenomena, while interesting, hardly constitute a Kuhnian revolution in evolutionary thought.

One of the main claims of postmodern biology is that “horizontal gene transfer” (HGT), the movement of genes between distantly related species via ingestion, viral vectors, and so on, can not only serve as a previously unknown source of genetic variation, but will also, by blurring the true pattern of ancestry and descent, efface the tree of life.  The former claim is correct.  The latter claim, while probably wrong, is still quite popular, as evidenced by the cover of New Scientist last February:

Why was Darwin “wrong”? As I noted at the time, New Scientist claimed that HGT is so pervasive among species that “the tree of life is not a fact of nature.”  New Scientist’s claim was palpably wrong, since species do have ancestors, whether or not we can recover that ancestry, and their evolutionary history, via gene sequencing.  Too, there’s no evidence that HGT is pervasive in nature.  It’s fairly common in bacteria, and has been seen in a few multicellular species, including rotifers and even Drosophila, but it’s so rare in metazoans that in those groups it hardly constitutes a challenge to the tree of life.

HGT has now been found in fungi, previously thought to be largely immune to the phenomenon since fungi can’t digest entire cells—usually the way genes are passed among species. The new issue of Current Biology has a dispatch on the topic by Thomas H. Richards, “Genome evolution: horizontal movements in the fungi“, summarizing recent work in that group, including a nice new paper in the journal by Slot and Rokas (reference below).

As Richards notes, HGT is usually detected by reconstructing evolutionary trees using different genes.  Most of the trees will coincide, but occasionally the use of one gene gives a tree that’s wildly disparate, showing “relatedness” of species that we know from other data are pretty unrelated. That’s a clue that that gene has moved horizontally among the unrelated species, giving a false signal of close evolutionary ancestry.

Using this technique, Slot and Rokas found that an entire cluster of 23 genes, those involved in the metabolic pathway for making sterigmatocystin (a toxic compound that is a precursor for the deadly aflatoxins), had been horizontally transferred from Aspergillis to the distantly related Podospora.  That’s a lot of genes, and since genes involved in the same pathway are often physically linked on the DNA of fungi, there’s a potential for widespread transfer of entire metabolic pathways between species (metabolic genes appear to be horizontally transferred among all species much more often than “informational” genes involved in DNA replication and transcription).

How did this happen? Well, Aspergillis and Podospora often occupy the same niche: both are “saprotrophs“, or species that break down dead animals and plants.  Living cheek by jowl, some of the DNA of an Aspergillis could have been ingested by a Podospora in the stew of organic matter, and incorporated into its genome.  But we don’t know exactly how this happened.

Richards gives a nice diagram summarizing cases of HGT among fungal species. HGT events are shown as colored arrows superimposed on the family tree of species determined from sequencing other genes. (The Slot and Rokas finding is the blue arrow.)  Note that transfers often involve species that are quite distantly related (click to enlarge):

But Richards hastens to add that this transfer is not common enough to efface the fungal tree of life, which can easily be discerned by concordant patterns of ancestry among the many non-transferred genes:

The inventory of HGTs identified represents a relatively small fraction of any of the fungal genomes studied. Consequently, the scale of transfer is unlikely to prevent the accurate resolution of a fungal phylogeny, as some have suggested to be the case for the prokaryote phylogeny [1], but will provide an interesting source of future research as investigators compare trait evolution with the fungal species phylogeny.

Even with these fungi, then, Darwin wasn’t really “wrong,” But what is exciting is the manifestly non-orthodox idea that adaptive genetic change can involve not just mutations within a species, but the wholesale movement of genes between different species—often very different ones.  As Richards notes, fungi have acquired genes not just from other fungi, but from bacteria and even plants.

Look for more examples of bizarre and wide HGT as DNA sequencing becomes more common.  But don’t expect it to demonstrate that there’s no tree of life!


Richards, Thomas A.  Genome evolution: horizontal movements in the fungi. Current Biology 21:R166-R167.

Slot, J.C., and Rokas, A. 2011. Horizontal transfer of a large and highly toxic secondary metabolic gene cluster between fungi. Curr. Biol. 21, 134–139.

44 thoughts on “Is there a tree of life?: Gene transfer between fungal species

    1. How silly.
      Clearly the noble Aspen Grove is the only acceptable plant-growth-form metaphor for phylogeny!

  1. Another nice example of HGT (but which wouldn’t make you throw out the idea of common descent if you engaged brain for a second) is in aphids, which have picked up carotenoid genes FROM fungi… pretty cool stuff. Especially given how important carotenoids are for things like immunity and mating displays – at least in some sp.

      1. What sort of mechanism do you think would allow this? I can understand fungi and plants doing gene captures more than animal life since plants (and I assume fungi) can grow by germination, but for an aphid to get a new gene it would have absorb before generating germline cells in order to pass it down.

  2. I think one of the earliest cases of demonstrated HGT involving a eukaryotic gene was the identification of a bovine trypsin gene within the genome of the bacterium S.griseus – leading to the famous comment by Brian Hartley: “the bactium must have been infected by a cow!”.

  3. What it does is tells us that genes are being ‘selfish’ then – replicators ‘want’ to replicate, even if it is the context of another genome than that in which it originated?

  4. Taking the stance of “the dissenters” on the Tree of Life debate only from that New Scientist article would be wrong. I know of at least one of the scientists quoted in that article who wasn’t entirely pleased with how New Scientist handled their quotes and what they were saying, to the point of it being misleading.

    I’ve had an opportunity to work quite closely with some of these people, and what they are saying makes a lot of sense, and a lot of it boils down to what we really mean by the tree of life, and what we really mean by things like LUCA (Last Universal Common Ancestor).

    And none of them dispute how well the Tree of Life holds for say Animals, and probably even Opistokonts in general, but we have to keep in mind that that is a small fraction of diversity on this planet. Bacteria and Archea outnumber everything else in terms of species diversity, and microbial eukaryotes aren’t far behind in third place.

    HGT looks like it really does have the potential to obscure the vertical component of bacterial evolution, and only a small fraction of genes may come close to giving that actual signal with a minimal number of HGT events in their history (Bill Martin’s “Tree of 1%”), the question then becomes how meaningful that tree-like signal is compared to the rest (majority) of the data, which is decidedly not tree like, but full of reticulation.

    I personally think it is still important but it does change how we think about the tree-like/non-tree-like nature of evolution and whether the concept of LUCA in that context is even meaningful.

    1. There’s quite a difference between obscuration of the signal of vertical evolution and saying that it’s unimportant or doesn’t happen. In the murky and arcane world of microbiology, it’s hard to get a grip on what a species really is; that’s an unavoidable hazard of the reliance, by necessity, upon molecular genetics in microbiological research (particularly in bacteria and archaea). Our common picture of the Tree of Life is built upon gene trees, and as we know, the gene is not the species. There is a well-recognized failure sometimes to recognize this, seen in the assumption that a gene tree is also an accurate depiction of the species tree. I think, however, that there is another, underappreciated way to misunderstand the difference between a gene and a species, which we see when people who get a kick out ripping up the tree of life forget that, just because genes move horizontally and gene trees disagree doesn’t mean that there ISN’T a species tree and that species don’t evolve vertically. If you consider a highly abstracted concept of a species as being a sort of body or substrate within which genes are acted upon by selection, it should be clear that they do in fact have a clear, vertical history, with ancestors and descendants. The movement of those genes from one body to another will damage the signal of the history of those bodies, but it’s foolish to claim as a consequence that that history isn’t even there. Things certainly do get weird and murky down at the base of the tree of life, but I do think it’s most reasonable to believe that the tree is real, even if we are unable to recover it.

  5. I doubt that there is any one of us that hasn’t had some incorporation of foreign DNA into our cells at some point in our lives (the prevalence of viral infections virtually assures us of that one). If we reproduced like bacteria (if every cell could lead to a new organism) then HGT would be enormously more common. The difference is that we, and other metazoans, reproduce through a select group of tissues – the germ cells, and these are (rarely) infected compared to other cells.

  6. Small nit pick:

    …the cover of New Scientist last February…

    The link to your older post (from February 2009) shows the issue is from January 2009. Neither is last February.

  7. Of course, it’s quite clear what happened. God directed the genes to move. Why? “It’s a mystery”, of course!

    He’s quite the traffic cop, that god. Intensely interested in fungi, bacteria, and beetles. Not so much in humans, though.

  8. Are there any hypotheses about the distinction of HGT for it being primarily metabolic genes and not all genes from an absorbed, eaten, or otherwise assimilated creature?

    1. This is one of the “unanswered questions” that’s posed at the end of that Current Biology piece. Just speaking off the top of my head, it may be that metabolism is so conserved among taxa that a packet of new genes, or a single gene, could integrate well into a different species without disrupting much. But I don’t know whether metabolic genes are more conserved than, say, genes involved in transcription and translations.

      1. Wouldn’t it be fair to say that a functioning metabolic gene packet would provide an instant advantage to an organism?

        Instead of disruption?

        That being said, one could wonder why HGT is so rare. I suspect the story is still unfolding.

        Of course, I like the messiness of it all. If it were a straight-line, one-size-fits-all solution, that opens the door for the “same design, same designer, guided evolution” argument. Examples of HGT, epigenetics and all the rest just underscore how natural and unguided the whole process is.

        1. Only rare in large multi-cellular things. Looks fairly common in bacteria. A little less common in microbial eukaryotes.

      2. They’re not. Informational genes tend to be much more conserved than metabolic genes. There are a few good ideas floating around about why metabolic genes appear to be transfered more often, and they all seem to make some sense, and have good support.

        1) Many metabolic genes in bacteria are organised together as operons for whole pathways, so they are then easily transfered as a unit. Sometimes these metabolic islands are found in things like integrons, which are really easily transfered.

        2) Informational processing genes are often found in large supercomplexes of proteins (and sometimes rRNA as well of course) with large numbers of interactions. Transfered informational genes may not be able to bind properly in these complexes if they are too divergent. This is the Complexity Hypothesis.

  9. (1) It’s a tree of life. (2) The extent to which it isn’t tends to be insufficiently appreciated by biologists working with most major groups. (3) With respect to this, bacteriologists tend to smirk perhaps too much.

    1. I think it’s more like kudzu, actually.

      Lots of tendrils.

      Of course, ‘kudzu of life’ isn’t as elegant – doesn’t trip off the tongue the way “tree of life” does. Maybe English ivy? Ivy of life. Yes, I like that better.

  10. Any further examples of HGT are going to pale next to the animal that synthesizes chlorophyll.

    Another good example is the cellulose-synthase pathway in tunicates.

    Has anybody looked closely enough yet to test the hypothesis that chitin-making in animals came from fungi?

  11. I have two questions:

    1) Creationists dispute the validity of reconstructing evolutionary trees using genes, saying that just because there’s similarities between genes that doesn’t mean that evolution happened. So they don’t accept much of the genetic evidence for ToE to begin with. However, in order to say that HGT in some way debunks ToE, it would be necessary to accept the validity of reconstructing evolutionary trees using genes, since that is how HGT is usually discovered. So the creationists are contradicting themselves.

    Is this correct? Have I made any mistakes here?

    2) Does anyone have any links to places where creationists appropriated the Darwin Was Wrong article to disparage ToE? I searched past entries here, but couldn’t find any.

    1. Re 2: There was YouTube uproar back when the article was published. But I don’t follow the creationist sites (I like my brain cells functioning, thanks), so I can’t say whether after the initial flurry there was any lasting effect.

    2. In a letter by Jerry & the Gnu Review to New Scientist in response to that letter — — this example was cited:

      The accompanying editorial makes it clear that you knew perfectly well that your cover was handing the creationists a golden opportunity to mislead school boards, students and the general public about the status of evolutionary biology. Indeed, within hours of publication members of the Texas State Board of Education were citing the article as evidence that teachers needed to teach creationist-inspired “weaknesses of evolution”, claiming: “Darwin’s tree of life is wrong”.

      1. With all respect due our host, I thought that letter a load of self-righteous crap. Reminded me of Miss Shields. But those who designed this cover know their blame, and I’m sure that the guilt you feel is far worse than any punishment you might receive.

        Scientists and science journalists should not have to filter everything they say or write to ensure that it cannot possibly be distorted. Constantly asking ourselves What Would Jerry Do? is no way to enjoy biology nor to convey that joy to our students and the public.

  12. The “Super-eminent Name is Wrong” trope is popular because the media go nuts over it and a few scientists crave attention. Darwin was wrong, Mendel was wrong, Newton was wrong…When in fact, science & scientific theories just tend to evolve similarly–from broader or more basic ‘rules’ to the complicated, sort-of exceptions, corollaries, and expansions.

    The most memorable part of the Lynn Margulis lecture I heard a decade or so ago was her emphasis on the advantages of evolving by adopting ‘pre-existing modules.’ Seems to make good sense that modules involving valuable items like carotenoids, aflatoxin precursors, & photosynthetic capability would be esp. “selected for” in HGT, so to speak.

  13. The great thing is that now that people know it happens, people can investigate how so many genes related to metabolism can transfer from Aspergillis to Podospora.

    The increasingly common “Darwin Was Wrong” headline never fails to annoy me. Anyone remember the days when the news read “another line of investigation demonstrates that Darwin was right”? Darwin was also unaware of modern genetics, so he could not have ever stated that DNA can only come from an ancestor of the same species.

    1. I suppose it’s that “Darwin” is synonymous with “the theory of evolution” for most of the public, and so criticising his work or ideas is a backhanded way of implying that “evolution is false”.

  14. a Kuhnian revolution in evolutionary thought.

    About that; I hear Kuhn may have used up to 20 different concepts of “revolution”, and none seem to be testable. What is the reason to think it is anything but facile story-telling?

Comments are closed.