Lampreys deep-six their genomes during development

August 26, 2012 • 5:43 am

One of the most bizarre phenomena uncovered since we’ve been able to sequence genomes is that of “programmed genome rearrangement” (PGR), whereby an animal starts its life as a zygote with a full genome, and then some of its genes are lost from the somatic (body) cells as development proceeds.   This has been seen in organisms as diverse as flies, hagfish, zebra finches, and, especially, ciliate protozoans, which extensively remodel their genomes during development, getting rid of repetitive DNA elements (“satellite DNA”).

It’s not entirely clear why this happens, but what is clear is that in organisms like us that whose somatic cells are segregated from the germ cells (cells that produce our sperm and eggs) gene loss doesn’t occur in germ cells. It couldn’t, for if those genes did have functions in the germ tissue, they’d be irrevocably lost in the next generation. In fact, if you find a gene present in germ cells but not somatic cells, because it’s lost in the latter, that gene almost certainly does something in germ cells.

Gene loss has also been described previously in the hagfish, a jawless vertebrate that, together with lampreys, make up the monophyletic group cyclostomes. Here’s a hagfish:

But the most comprehensive study yet of programmed genome reduction was just published by Jeremiah Smith and three colleagues in Current Biology (reference below). Building on previous but less comprehensive work suggesting that from development of egg to adult, the sea lamprey (Petromyzon marinus) lost about 20% of its genes in body tissues (while retaining them all in the germ line), Smith et al. sequenced DNA from both germ tissues and body tissues of single individuals. Here’s are two sea lampreys on a brown trout: they’re parasitic on fish and usually kill them by sucking blood:

The authors sequenced DNA from body tissues (and blood) as well as germline tissue (sperm), and also looked at the RNA transcripts.  What they found was this:

  • 13% of DNA sequences found in the germline were missing in body tissues, roughly consonant with the 20%  reported in previous work on this species.
  • The genes eliminated from the DNA during development—and we’re not quite sure how this happens— included not just repeated satellite DNA, but real, single-copy genes that have functions.  Eight genes were identified that were active in germ cells but not somatic cells, and there are undoubtedly more.
  • The functions of these genes in germ cells give one clue why they might be eliminated in body cells. (The genes include APOBEC-1 Complementation Factor, RNA Binding Motif 46 [cancer/testis antigen 68], and two “zinc finger” proteins.) The authors note that these genes do act very early in development to segregate the germ cells from the body cells, and have other unknown functions in the germline—they could, for example, be involved in crossing-over between chromosomes or the production of sperm and eggs themselves.
  • Why, then are those genes eliminated from body cells? The authors suggest that genes like those identified above have crucial functions either in germ cells or in segregating germ cells from body cells early in development, but might be deleterious within body cells, perhaps because their bad effects—in particular, in causing cancers—outweigh any good effects. In other words, there’s a conflict within the bodies of lampreys and hagfish between germ and body cells. The way evolution appears to have resolved this conflict is to simply get rid of the “bad” body genes during development.  An alternative strategy would be to “silence” those genes in the body tissues—prevent their expression in non-germ cells—and I’m not sure why they’re removed rather than silenced. (It might be evolutionarily “easier” to snip out genes than silence them, but yet many species, including ourselves, have ways of silencing different genes in different tissues without removing those genes from the DNA. The reason different tissues are different is because they express different sets of genes.)

Many questions remain.  Are those genes eliminated from body cells in fact deleterious if they remain in body cells? If so, why do they remain active in the bodies of non-jawless vertebrates, like ourselves?  Second, if they’re harmful in the body, how are they harmful? Do they cause disease, or do they simply impose a useless metabolic and somatic burden?  Third, is this phenomenon of PGR an ancestral condition, since jawless vertebrates are the descendants of the earliest vertebrates, or has it evolved secondarily in those lineages? As the authors note:

Notably, both extant lineages of jawless vertebrates (agnathans: lampreys and hagfish) are known to undergo PGR, which would seem to indicate that the phenomenon is common to all extant agnathans [jawless fish] and potentially represents an ancestral condition. Thus, PGR may represent an ancient mechanism for moderating genetic conflict between germline and soma that evolved within an ancestral vertebrate lineage (alternately, repeated evolution of PGR in lamprey, hagfish, and numerous invertebrate and protist lineages may reflect recurrent selective advantages for PGR).

Finally, are different genes eliminated in different body tissues, or do all body cells get rid of the same set of genes? The authors’ analysis was too coarse to answer this question.

Regardless, the phenomenon of eliminating some genes from body tissues but not from germ tissues appears to have an evolutionary advantage, for it has happened (or been retained) in many lineages.  What that advantage is remains to be seen. This is an example of one of the evolutionary questions that could only be studied properly once we became able to sequence DNA.


Smith, J.J., Baker, C., Eichler, E.E., and Amemiya, C.T. (2012). Genetic consequences of programmed genome rearrangement. Curr. Biol. 22, 1524–1529.

27 thoughts on “Lampreys deep-six their genomes during development

  1. PGR is involved in the development of lymphocytes, involving factors that we don’t see active in other cell types. It will be interesting to see the mechanism of the process in these organisms where it occurs more extensively.

  2. “For example, many of our olfactory receptor genes, no longer useful in humans who don’t rely much on smell, are “dead”: they’re in our genome but are not expressed since they’ve been inactivated by mutations. They are active in our relatives, like rodents, who rely heavily on smell.”

    This is a bad analogy. The human olfactory genes do not work in any stage of our development. They are not silenced, they are disabled and not by a development mechanism, but by their own mutations.

    The pushed out genes of these animals seems to be working genes.

    1. You make a good point, but what I was trying to say that that it is possible to silence genes that aren’t used. Now this may not be possible to differentially silence genes by mutation unless there in parts of the body, but they could be silenced by other means. Indeed, that is what has been done in body tissues, for different genes are expressed in different parts of the body. I’ve clarified that above now.

  3. Perhaps lampreys are in the early stages of evolution towards a more parasitic lifestyle? And, yes, I realise there there is no ‘direction’ in evolution – but if lampreys exist in a low energy niche then any change (such as offloading genes that are no longer required) that reduces energy requirements (but increases fitness) could fall through into parasite life-style.

    1. I think it could be argued that parasites are in a “high-energy” niche. What food is more rich than blood? And delivered directly into the mouth, without even the need to move? Eat and sleep at the same time! Sounds like an easy life once you are hooked up.

    2. Evolution doesn’t work that way. There is no teleology. Evolution isn’t directed toward a goal, it just happens and the result seems to be goal directed because humans have hyperactive agency detection and impute causes where there are none.

  4. This is fascinating. I wonder whether it has something to do with the algorithmic complexity of retreiving the data in the genome but I don’t know enough about genetics to get my head round that

  5. Amazing stuff. I’ll bet this system will produce interesting info for the next 20 years! I’d better start eating healthy.
    I’m a little uneasy with JC and the authors saying PGR has a ‘function’. It probably has a proximal function, in that if it didnt happen the organism would die but a relatively minor change in the promoters of a few genes might have achieved the same purpose. I suppose as lineages gain in complexity they have to make a ‘decision’ at some point: shut of genes with heterochromatin or complex regulatory networks or delete them from the somatic genome. The option thats ‘chosen’ might be somewhat random.
    Could this be a system for removing repetitive DNA from the genome that started to removed genes as well? If so I predict that many deleted genes could be retained in the somatic geneome with no ill effect

  6. This is something I haven’t been aware of at all. I knew of the silencing of genes, but the removal of genes is a fascinating idea. Thanks for posting this.

  7. Amazing phenomenon. I would think that genetically editing every somatic cell in the body would be hugely dangerous and prone to error, risking all sorts of mutations. I suppose that if one is doing that in somatic cells, and post-development, that gross effects of a single mutation may be mitigated, but I’d still think that cancers would be a concern.

    1. Or it could be done during the later stage of development (these genes were only from an early stage the article claims, indicating the timing), where plasticity could cover programmed cell death of failures?

  8. Off the top of my head, many, if not most, species of lamprey are not parasitic, and do not feed as adults.

  9. Do humans have genes that have a function in the growth from the zygote that end up in later years causing cancer and other deleterious effects? If that is the case how does our body attack this problem?

  10. Looking at this posting, it’s just another example of science not seeing the deeper truth. No one has ever compared how these species, with their diminished genome, taste.

    I understand how difficult an experiment would be because of the difficulty of getting not only one but two blindfolds onto a cat.

    However, isn’t it obvious that Ceiling Cat has created these adaptations to improve the species taste. To cats.

  11. Wow. I’d never heard of this and it’s fascinating indeed.
    If agnathans develop like zebrafish, the very few (4) primordial germ cells don’t start differentiating until there are already hundreds of cells, so the programmed removal has to happen in hundreds of cells more or less simultaneously.

    If I may quibble with the phylogenetic stuff:

    the hagfish, a jawless vertebrate that, together with lampreys, make up the monophyletic group cyclostomes

    Maybe. I think this is how recent molecular studies are leaning, but there are serious long-branch attraction issues and they may be separate lineages (i.e. ‘cyclostomes’ may be massively paraphyletic, with lampreys the sister group to all other vertebrates and hagfish the most basal extant branch of the ‘craniates’.

    jawless vertebrates are the descendants of the earliest vertebrates

    as are we all.

    PGR…is common to all extant agnathans [jawless fish] and potentially represents an ancestral condition.

    If the cyclostomes are</i. monophyletic, then it's (slightly) more parsimonious to hypothesie that PGR evolved early in their lineage and is not ancestral to jawed vertebrates.

    many, if not most, species of lamprey are not parasitic, and do not feed as adults.

    yes, the brook lampreys in particular. Lampreys get a bad rap.

  12. Fascinating. As Tulse says, you’d think that this sort of genetic editing would be so error-prone that it would be selected out, but apparently not.

    FWIW, I go for the hypothesis of the genes being actively harmful. If metabolic/somatic burden were the driver, surely we’d expect things like Alu sequences being the first victims of the genetic pruning shears? It seems simpler and more effective. (Says he with the confidence of someone who doesn’t really know a great deal about the subject.)

    1. Although the notion of a metabolic burden for producing DNA is often raised, I wonder if anyone has tried to quantify that? Genomes seem to routinely have vast amounts of non-coding, non-functional sections, and chromosome doubling seems to be an extremely common occurrence, at least in plants. Surely if there were any significant detriment to carrying large swathes of non-functional genetic material, these examples wouldn’t exist.

      1. Sadly, the only time I’ve heard an estimate mentioned was as an un-cited statement, in this paper.

        If anyone sees a better citation on metabolic loads, please post it.

  13. The fish in the photo burdened by lampreys is the “lake trout” (technically a char) Salvelinus namaycush rather than a brown trout.

  14. Is this another facet by which we might try to tackle the ‘c-value enigma’, as described by T. Ryan Gregory here? According to Gregory, variation in genome size is one of the factors that affect developmental rates in many kinds of organisms. There is no simple relationship between development rate and c-value, but its clear that organisms that need to grow quickly have constraints on their genome size, because it affects cell division rates etc. Perhaps this somatic-line genome reduction is an adaptation to earlier genome enlargement that is no longer necessary, or is now detrimental.

  15. Deep memory: I recall discussions back in graduate school (late ’60s) of “chromosome diminution” in parasitic roundworm called Ascaris. The process was discovered by cytological methods, but a google search uncovers more recent molecular characterizations. It is interesting also to note that some of the foundational experiments of experimental embryology (e.g., isolation of blastomeres by folks like Hans Spemann and Hans Driesch)were designed to test the hypothesis that segregation of “determinants” (genes in today’s terminology)accounted for cell differentiation.

Leave a Reply