Matt Brauer posted Entry 1755 on December 14, 2005 08:20 PM.
Trackback URL:

Review of:
Marques et al., “Emergence of Young Human Genes after a Burst of Retroposition in Primates.”

PLoS Biology 3(11):1970-1979.(Synopsis on PLoS Biology)

November’s issue of PLoS biology
has several papers of evolutionary relevance.

  1. Richard Robinson gives a nice review of some current thinking about abiogenesis.
  2. The evolution of “genetic robustness” is explored in a paper from Paul Turner’s lab.
  3. A paper by Sabeti ET al. demonstrates that the evolution of a disease resistance locus in humans, thought to have been under strong recent selection, cannot actually be distinguished as non-neutral.
  4. Mating preferences in fruit flies were shown by Rundle ET al. to evolve as a side effect of selection in divergent environments.

In addition, this very elegant paper describes some surprising results relating to the evolution of new genes in humans.

Considered at even moderate timescales the genome is a dynamic structure. Chromosomes get rearranged, duplicated or lost, and genes pop in and out of existence like virtual particles. One of the mechanisms for the movement of genes is the so-called retrotransposable element, or retrotransposon. Along with their cousins the transposons, retrotransposons are parasites on the genome, and rely upon the cellular machinery to copy their sequences to new locations. (In some cases, transposable elements retain a structure that is suggestive of an origin from a viral genome. Just as tapeworms have lost the superfluous parts required for life outside the host, transposons appear to be viruses that have shed their corporeal existence in order to live as a string of bits in someone else’s genome.) Retrotransposons look a lot like degenerate retrovirus genomes. They also replicate similarly, passing first through an mRNA cycle (transcribed by the host’s polymerase) and then re-integrating as DNA synthesized with a retrotransposon-provided Reverse Transcriptase (RT).

The presence of the enzyme RT has an unusual side-effect: occasionally a molecule of the cell’s mRNA will act as a substrate, and the message will be converted back to DNA. There is an important difference between the original gene and its reverse-transcribed child, however: while the original gene might have had introns that interrupted the sequence, the retrotranscribed child molecule derives from a processed mRNA, and so is intron-free. Because of this, the parent gene can be distinguished from the child. (In addition, the child gene most likely will have integrated at a new place in the genome where it would not have been under useful cellular control, and it would have degenerated into a pseudogene as a result.)

One family of retrotransposons, known as LINE elements, appears to have gone through a phase of high activity in several mammalian lineages about 100 Myears ago. The activity was so high that up to 20% of some mammalian sequences derives from LINE element, in fact. This burst of activity presumably resulted in a lot of RT appearing in the nuclei of various mammalian cells at that time. So much so, in fact, that the incidence of “retroposition” – this copying of a gene to another location via mRNA – seems to have increased dramatically, most notably in the primates, including humans.

What became of these retrocopies? Like most copies, they degenerated into shadows of their former selves. But in some few cases, they retained activity. These retrocopies presumably landed in a portion of the genome where the the transcriptional regulation had a positive effect on the organism. This is expected to happen rarely, and indeed the event appears to be uncommon. Marques et al. found that only about 18% of identified retrocopies appeared to be intact. (Many more retrocopies, made invisible by extensive degradation, may have been missed, of course.) Based on the rate of neutral mutation, they estimated that most of these retrocopies originated at around the same time, about 40-50 MYA. This is after the primate-mouse split, and is consistent with a burst of retroposition occurring in primates at about that time.

The authors went on to examine the fate of intact (and presumably still functional) retrocopies. They found that the 38 genes they examined arose at various points during primate evolution. Seven of these genes were unlikely to have remained intact in all lineages without the action of some selection. These seven were therefore identified as candidate functional retrogenes.

What further can be said about these candidates? The authors looked at the tissues where the parents of these genes were expressed, and compared them to the location of the functional retrogenes.

They found that the parents of the seven genes were expressed in all tissues. But in every single case, the corresponding retrogene was expressed in testes only.

It’s fun to imagine the authors’ sense of glee as this realization dawned on them. For there is no way that they ought to have expected this to be the case. The genes were chosen by criteria that were independent of the location of expression. To find that all of the genes are expressed in the same tissue, and only in that tissue, is a phenomenal result.

So what does it mean? There is some idea that retrogene formation may be favored in testes due to hyperactivity of transcription during male meiosis. But these are genes that have been under functional selection, so unless all retrogenes originated in testes, the odds of seeing this particular group of seven there are very low.

For five of these genes, the answer may be related to the fact that the parental genes lie on the X chromosome. Half of the male gametes lack the X chromosome, so the retrocopies might have originally been making up for this intrinsic lack. Both of the other retrogenes have parents involved in spermatogenesis, so it is reasonable to assume that the extra copies add functionality to the testes that carry them.

Many of the postulated mechanisms of rapid evolutionary change (sperm competition, reinforcement, varieties of sexual selection) relate to male reproductive function. Thus it is gratifying to see that the phenomenal source of genetic variation supplied by retrogenes is most apparent in the testes.

The next chapter in this story should be the identification of the new functions that these genes may have adopted in the testes. By doing this, we’ll have another very strong example of the purposeful arrangement of parts effected by a known, defined and purposeless mechanism.

Commenters are responsible for the content of comments. The opinions expressed in articles, linked materials, and comments are not necessarily those of See our full disclaimer.

Comment #62917

Posted by Mike Elzinga on December 14, 2005 8:33 PM (e)

Very interesting paper. Thank you for pointing it out.

Comment #62922

Posted by Mike Elzinga on December 14, 2005 9:07 PM (e)

In Box 1 of Richard Robinson’s review, the puzzle of chirality is discussed. Physicists have been wondering about this predominant handedness as well. As you may know from the famous confirmation of parity violation by C.S. Wu, physicists have since been wondering whether or not the chirality of positrons from these decays preferentially favored the chirality we find in biological systems as positrons from beta decays interacted with the electron of the molecules participating in early reactions leading to life. The effect would be extremely small, but would provide a slight bias. I don’t think there have been any definitive results of experiments trying to test this hypothesis. I recall that the experiments were jokingly referred to as “origin of life” experiments or “Vatican” experiments.

Comment #62924

Posted by Jeff McKee on December 14, 2005 9:17 PM (e)

I’ll have to read the paper carefully, but if there was a “positive effect on the organism,” then the 40-50 mya date, “Based on the rate of neutral mutation,” is rather meaningless.

Useful stuff, though.

Comment #62925

Posted by Matt on December 14, 2005 9:24 PM (e)

The “neutral rate of mutation” is the rate of change at neutral sites (i.e., 3rd base positions). Even if a protein is under functional constraint, neutral evolution can occur at these sites.

Comment #62926

Posted by Steviepinhead on December 14, 2005 9:24 PM (e)

The abiogenesis paper is also definitely worth a read–pretty cool stuff!

Comment #62927

Posted by Steve S on December 14, 2005 9:26 PM (e)

the review article by Richard Robinson is very good.

Comment #62934

Posted by Matt on December 14, 2005 10:40 PM (e)

If there is interest, I may start a thread on Robinson’s OOL review. (We could discuss it here as well, but giving it its own real estate might attract other commenters who aren’t interested in gene origins.)

Back to young genes:

As I finished writing up the summary of Marques et al., I found myself starting to question one of their assumptions. Namely, is it really all that unlikely that nearly all retrogenes might have originated in the testes? I haven’t yet looked up their references, but it seems to me that the biased origin idea might have more merit than they imply.

Comment #62953

Posted by Mike Elzinga on December 15, 2005 1:59 AM (e)

Question for Matt (from a confused physicist who may not be asking the right question):

Does the 1 – 1.3 x 10^-9 substitutions per site per year for the molecular clock rates in Old World monkeys, apes, and humans mean an average rate over the history of these species? As I understand it, the authors use this to locate the peak of a burst of retroposition about 38 to 50 MYA.

The reason I ask is that one might suspect that variations in radiation background affects the rates at which these events occur (e.g., the earth’s magnetic field has flipped many times, and this affects the amount of radiation coming from outer space). Is computing ratios KA/KS how one compensates for the very likely possibility that clock rates are varying for most of the molecules on the genome. In other words, is one more interested in this case in the relative rates against a background of silent (I assume that means relatively inactive) sites?

I guess what I really want to know is, what is the basic trick here in measuring these rates?

Comment #62959

Posted by Susan on December 15, 2005 5:17 AM (e)

Really interesting stuff. Thanks for posting that.

Comment #62975

Posted by Miguelito on December 15, 2005 11:10 AM (e)

I very much enjoyed the article on abiogenesis. Mainly because I’m a geologist and understood much of the geochemistry described. Alot of the detailed biochemistry and biology stuff posted here is far over my head.

Comment #62982

Posted by Katarina on December 15, 2005 11:50 AM (e)

Your “transposon” and “retrotransposon” links don’t work for me. How about anyone else? Think they may be broken.

Comment #62984

Posted by Justin on December 15, 2005 12:00 PM (e)


You bring up an interesting idea. Generally, mutation is thought to be a poisson process. However, it appears the variance is greater than the mean, giving the molecular clock an “overdispersed nature”. The reason for the overdispersed clock is unknown. However, your idea about the flipping of the magnetic field of the earth, as far as I know, hasn’t been proposed. It’s a really interesting idea - you might be onto something.


Comment #62986

Posted by Matt on December 15, 2005 12:34 PM (e)


I believe that the assumption is that mutation rates within a lineage are constant. This is a reasonable assumption, I think, over the fairly short timescales that the authors are working with.

But I think your idea is very interesting. And a changing mutation rate would certainly add some noise (maybe a lot) to estimates of gene ages dervied from measures of sequence divergence.

Comment #62987

Posted by CJ O'Brien on December 15, 2005 12:50 PM (e)

I was also intrigued by the origin article. A very good summary of the field to date.
Nice to know some money’s headed that way.

Comment #62989

Posted by BWE on December 15, 2005 1:13 PM (e)

I did a bunch of web searches and couldn’t find anything on this topic but in grad school (many years ago) I did some research on different asteroidea and one part involved mutations. (They exibit enormously different characteristics if exposed to certain toxins).

I created a timeline of species and I remember that the sudden shifts occured almost exactly when the magnetic reversals evidenced along ocean ridges occurred. A popular hypothosis at the time proposed that some sunspots were so powerful that they could knock the earth’s gravitational field clean off the earth and it would reorient itself however it did. Their was someone who claimed that the massive irradiation of the earth during the time of no magnetosphere could account for massive die offs and resulting mutations.

ANybody know if that hypothosis is still being thrown around?

Comment #62990

Posted by Henry J on December 15, 2005 1:15 PM (e)

Re “Namely, is it really all that unlikely that nearly all retrogenes might have originated in the testes?”

A thought just occurred to me - wouldn’t retrogenes that originate in other body parts be rather unlikely to get passed on to the next generation?


Comment #62993

Posted by Bayesian Bouffant, FCD on December 15, 2005 1:42 PM (e)

A thought just occurred to me - wouldn’t retrogenes that originate in other body parts be rather unlikely to get passed on to the next generation?

I think you are confusing ‘where did the mutation happen’ with ‘what tissue is the gene active in’.

Comment #62997

Posted by Henry J on December 15, 2005 2:04 PM (e)

Oh. In that case, never mind.

Comment #63000

Posted by Matt on December 15, 2005 2:43 PM (e)

Actually, I think that Henry is onto something that the paper glossed over a bit. It is true that a retrotransposition has to occur in the germ line, just as with any mutation that gets passed on.

If a particular tissue in the germ line has drastically increased rates of transcription, there will be a large increase in the number of messages in cells of that tissue. This means there are more opportunities for retrocopies to be made.

Also (and this is the key point), it is known that transcriptional activation is often accompanied by changes in chromatin structure. That is to say, the chromosomes “loosen up” in the region of genes that are actively being transcribed. This may make those regions of the chromosomes more open to integration of the retrocopies.

Thus, the most likely location for integration would be in the locale of genes that are under the control of testes-specific promoters. (It occurs to me that one test for this might be to map the genomic location of the seven new genes and see if they fall in the neighborhood of testes-specific non-retrogenes at some frequency greater than that expected by chance.)

I must say, though: as a microbiologist, I sometimes find this metazoan stuff unlikely, counterintuitive and just plain messy and unnecessary.

Comment #63001

Posted by Henry J on December 15, 2005 2:58 PM (e)

Re “I sometimes find this metazoan stuff unlikely, counterintuitive and just plain messy and unnecessary.”

I think I resemble that remark! :)

Comment #63003

Posted by Timothy Chase on December 15, 2005 3:05 PM (e)

Looks like the abiogenesis entry at EvoWiki needed a little updating with regard to the early atmosphere and the formation of ribose. Thanks for the link. Fixed now.