Whatever role gestational selectionism plays, it is probably fanciful to suggest it could counterbalance the interventions of medical science in allowing individuals with otherwise lethal deleterious mutations to lead full reproductive lives.

Given the average survival rate for fertilised eggs is 30%, how much 'flex' would a mother with high mutational load have to spontaneously abort a higher proportion of her eggs without reducing her own level of fitness in the process? I simply can't see gestational selection replacing natural or sex-based selection, unless you view things through group-selectionism.

For some evidence, I imagine there has been a large proportional increase in genetic diseases within populations which have had good quality long-term public health care.

Interesting to note that Hamilton himself considered natural abortions (mainly in cocunut embryos) in Narrow Roads of Gene Land Vol 2.

Natural selection is still operating. It's just not operating on hunter-gatherers. We are being adapted to an urban environment.

The 30% maximum rate of live births to conception rate is comparable to IVF success rates (one in three). This may be a natural rate for humans that IVF may not be able to improve on. Since IVF costs $14k per attempt, and an average of 3 attempts is necessary for pregnancy, IVF will remain an option for the well-off unless the cost drops dramatically. Since its the well-off who tend to have kids later in life (increase future time orientation, right?), this may not be such a big deal.

Is chromosomal abnormalities responsible for the high rate of spontaneous abortion? The article seems to infer this more than present direct evidence of this. In any case, human reproduction does appear to be kludgy compared to other mammals. Why would human embryos have higher levels of chromosomal abnormalities and kludgy reproduction, in general, compared to other mammals? Perhaps its related to our big brains and intelligence. If so, this is yet another "penalty" we pay for intelligence and big brains. Perhaps this is yet another explanation for the fermi paradox.

It would be nice to see similar research done on the various primates (chimpanzees, gorillas, etc.) to see if they have similar kludges in their reproduction. Is this problem limited to humans only?

I wonder if this has implications for anti-aging research. Perhaps these chromosomal abnormalities are the results of stochastic genomic DNA damage. Obviously, only the undamaged embryos are able to gestate and be successfully born, which make for the selection process. If so, this is something we have to deal with in order to cure aging.

Perhaps the problem is with mitochondria and not genomic DNA. The article does not specific exactly what "chromosomal abnormalities" exist in the "failed" embryos.

I can offer an explanation for this. Most mammal reproduce within 1-3 years of birth. We humans reproduce only once we reach sexual maturity, which is around 14 years old, at earliest. Most humans do not have kids until their 20's and the yuppies have kids in their 30's. Even the 20's is beyond the life span of the other mammals. So, enough time passes for the DNA and mitochondrial damage to show up and make human reproduction kludgy. If this is right, it really is our big brains at fault. Because its our big brains that make us take 15-20 years to reach full maturity as well as to require a long (9 month) gestation period followed by the birth of an infant that is competely helpless, compared to the new-born of other mammals.

We really do pay a price for our big brains. Since we're not in space yet, in any meaningful manner, it is not clear if big brains are a successful evolutionary trait (compared to flight, for example).

I certainly would not be surprised to learn that bonobos have low fertility, such that only the absolute sex maniacs successfully reproduce.

kurt9 said:

"So, enough time passes for the DNA and mitochondrial damage to show up and make human reproduction kludgy. If this is right, it really is our big brains at fault."

I think it nearer the case that our big brain is caused by DNA damage. The big human brain seems to have evolved rather rapidly in a pretty small population. How could such a small population generate the required number of mutations in so few generations without considerable chromosomal instability?

It seems to me that there is an inverse relationship between a species' potential for rapid evolution and the stability of its chromosomes. Kludgy reproduction, and other things like high cancer rates, may well be the price we pay for having the capacity for rapid evolution.

Okay, you tell me what I am misunderstanding.

I read the article at Science Daily.

I found the following section confusing:


"Ms Vanneste and her team studied each cell from 23 three or four day-old IVF embryos from young (less than 35 years old), fertile couples who had asked for preimplantation genetic diagnosis (PGD). PGD is carried out where one or both parents have a known genetic abnormality, in this case an X-linked disorder or the microdeletions (loss of a tiny piece of a chromosome) that can cause such disorders as the cancer predisposition syndrome neurofibromatosis type 1. The embryos are screened to avoid the implantation of one carrying that abnormality. Such embryos are the most representative of normal human embryogenesis, the process that begins once an egg has been fertilised."



It seems to be saying that under 35 is "young". That seems like a high cut off for "young". Then it goes on to imply the parents knew they had genetic abnormalities that prompted them to seek testing.

I wouldn't describe these people as young and healthy, or even fertile.

I would describe as young, healthy and fertile, a group randomly selected from a pool of people under 25, with no family history of genetic defects who already have a healthy living child and no miscarriages as young, healthy and fertile. What am I missing in these criteria?

Bob,
Deleterious mutations arise continually, and are not desirable in any urban or primitive environment - they, likely, simply degrade complex traits like intelligence and beauty. That's the issue here. In both the forager period and the pre-industrial civilizational period, about 50% of people may have failed to have any children who reached adulthood. Those "failures" were probably richer on average in deleterious mutations and their failure probably helps remove deleterious mutations from the gene pool. If 80% (or whatever) of people have reproduced since 1850 (and all of them equally), then the mean load of mutations per individual may progress upward until reaching a new equilibrium.

Now, it's not strictly true that 50% of people failed to reproduce, because that only counts people who were born in the first place - the loss of fetuses may also be correlated to mutation load. However, it seems that many failing fetuses have these gross chromosome abnormalities - I don't see at a glance why those should necessarily correlate with mutation load.

For some evidence, I imagine there has been a large proportional increase in genetic diseases within populations which have had good quality long-term public health care.

You consider your imagination evidence? How very modest of you! I'd be interested in actual evidence. It looks to me like modern public health care is a phenomenon of recent vintage, which has tended to be available only in populations that have over the same timeframe experienced the benefits of hybrid vigor.
And even if we accept a longer timeframe, we would have to imagine a condition where the heterozygote carrier for some recessive genetic disease has reduced fitness, so that the lowered bar for survival to reproduction allows a greater spread of the gene than would otherwise occur. Of course plenty of such genetic variants exist; but the most common genetic diseases appear to have given heterozygotes a fitness advantage! Cystic fibrosis and sickle cell anemia, for example, confer increased disease resistance on carriers. So what happens when improved public health reduces the mortality from malaria or tuberculosis? The balance shifts against the variant and (very slowly) the incidence of the genetic disease falls.

sg, you are raising an issue of whether the sample is representative?

The individuals are going to be healthy themselves if all the variants in question are fully recessive. And I don't see why a modest mutation or deletion should impact the creation of gametes with chromosome abnormalities. Ergo, it seems like a fair sample as far as I can tell.

The central point is, or at least ought to be, that low fecundability in humans - which we know exists - is wildly anomalous.

An enigma.

Cochran beat me to it. The ancestors of humanity presumably didn't have such large hits to fertility due to chromosomal abnormalities. How did natural selection allow it to emerge for us?

Seems like logical consequence of having a high ratio (investment to raise one offspring:investment to produce and kick the tires on a zygote). If a human has to fertilize twenty eggs, lose some due to failure to implant and also suffer a miscarriage or two due to abnormalities, that investment is still small compared to what's needed to raise an infant to adulthood, so the hit to fitness is not huge. Compared this to (say) an octopus, which would suffer a loss in reproductive fitness in linear proportion to the number of lost zygotes...
In fact come to think of it, there would be an additional force at work - any mechanisms that the body has to reject abnormal embryos would need to be more sensitive in humans (to maximize fitness), because of the greater cost of raising abnormal offspring. I've seen cats eat some of their newborn kittens because there was something wrong with them - for humans, this sort of thing is off the table, so any earlier mechanisms should be more sensitive...

bbartlog, you're wrong. Sure, there's a reason for humans to care more about offspring quality: but there's no reason to have lower average quality. We're not talking maternal rejection for something a bit short of perfection, we're talking trisomies and aneuploidies that had zero chance of survival. We're talking >70% of zygotes totally ruined by chromosomal anomalies. I don't know of a similar pattern in any mammal - including mammals that live a long time, like elephants.

It is quite weird.

Among the traits subject to selection are those that determine the rate at which mutations occur within a species. For example, if you are a coelacanth living in a benthic marine environment that is extremely stable over hundreds of millions of years, you would evolve toward extreme genetic stability. I would wager coelacanths have excellent DNA repair, etc., because the downside of mutations would vastly outweigh the benefit they would derive from having the potential for rapid evolution.

The opposite would be the case for a species that stands to benefit from rapid evolution. If the benefits of instability are high enough, instability will be selected for despite the downside. Given that there are over six billion humans, it is obvious that we have in some way compensated for any fecundity problems we have.

The peculiarities of human evolution caused our lineage to evolve into a huge unexploited niche as the "technological animal." Our capacity for material culture has given us a dramatic advantage. Competition among our hominid, and later, human ancestors to exploit the new niche could have driven them not only to evolve better adaptations toward material culture, but to evolve adaptations toward higher rates of evolution. An evolutionary race into a bountiful unexploited niche might well explain the chromosomal instability described in the article.

In other words, it is possible that chromosomal instability isn't a bug, it's a feature.

U.S.,
Are you trying to argue this via individual selection or group selection?

The fitness gain of the mean mutation is negative. Individuals/lineages with high mutation rates are harmed on average by mutation, not helped. The rare individual with a helpful mutation would indeed be likely to occur in a lineage with an above-normal mutation rate. But in sexual species, the helpful mutation is very rapidly dissociated (by recombination) from the alleles responsible for the above-mean mutation rate.

Thus, it seems at least some authors on the topic believe the mutation rate is minimized in sexual species, and that it fails to go even lower only because of the energy cost of further reducing mutation.

see, eg, http://www.ncbi.nlm.nih.gov/pubm...pubmed/ 11084621

gcochran: thanks, that addresses my second point... not sure about the first, though it sounds like elephants really should be similar if it was just about maternal investment.

Slayton: it does seem likely that organisms can gradually evolve to mutate more or less slowly, but unless chromosomal instability is a side effect of an increase in other less radical mechanisms of mutation I don't think the idea holds water... almost all of our evolution is via SNPs, copy number changes and the like; chromosomal rearrangements that result in improved fitness just look like they'd be orders of magnitude less likely.

Maybe we're just poisoning ourselves? Our modern environment is surely full of a soup of substances that we have no evolutionary track record with...

> The rare individual with a helpful mutation would indeed be likely to occur

Well... /more/ likely.

"Our findings have shown that almost every cell of a human embryo carries a different genetic composition; consequently, the one cell that is analysed genetically is not representative of the rest of the embryo. If the tested cell is genetically abnormal, the embryo will not be transferred. But the rest of the embryo might be normal and develop into a healthy person. [...] The prevalent chromosomal instability in all early human IVF embryos explains the failure of PGS to improve the live birth rate per embryo transferred."

This seems to be suggesting that there is very early mosaicism for chromosomal problems in many people who are actually born! (Isn't it?) Was this known before now? Is there more such mosaicism in actually-born individuals in man than in other mammals?

It seems like this could be a possible variable in embyonic development, and potentially affect psychometric g and other fitness determinants. If human zygotes merely have a very high rate of fatal chromosome abnormalities compared to other mammals that would certainly be interesting enough. But if the chromosomal instability continues as the embryo develops and actually affects viable individuals (more than in other mammals), that would be of potentially great importance.

Consider this from wikipedia:

"It was recently discovered that mosaicism for aneuploid chromosome content may be part of the constitutional make-up of the mammalian brain.[3] [PNAS, 2001] This observation was then extended to normal human brain, where brain samples from six individuals ranging from 2–86 years of age were found to have mosaicism for chromosome 21 aneuploidy (average of 4% of neurons analyzed).[4] This low-level aneuploidy appears to arise from chromosomal segregation defects during cell division in neuronal precursor cells,[5] and neurons containing such aneuploid chromosome content reportedly integrate into normal circuits.[6] These results suggest the possibility that somatic mosaicism in the brain (and perhaps, by extension, other tissues) may contribute to the diversity between individuals."

Maybe we're just poisoning ourselves? Our modern environment is surely full of a soup of substances that we have no evolutionary track record with...Checking how domestic dogs are faring might shed some light on this.

Well, presumably the individuals wouldn't normally be severely mosaic for serious chromosomal abnormalities because the normal cells would outcompete the abnormal ones. If you cut an early stage embryo in two, you can get two babies, after all. Perhaps the embryo can adapt to "not dividing/expressing properly" cells just as well as it can adapt to missing cells.

Neziha,
I'm reading this free April 09 review right now:

http:// www.pubmedcentral.nih.gov...bmedid=19468329

While a portion of these aneuploid cells apparently die during development [7,23,24], aneuploid neurons have been identified in the mature brain in all areas assayed [3–8,11,25] indicating that aneuploidy does not necessarily impair viability [26]. Aneuploid neurons in the adult have been shown to make distant connections and express markers associated with neural activity which indicates that these neurons can be integrated into brain circuitry [18].

There may not yet be broad based data on whether this stuff is or is not more intense in man than in other animals, at least with respect to the brain (though it is known that human lymphocytes features some aneuploidy; it isn't brain-restricted):

The overall prevalence of aneuploidy in the normal adult mammalien brain is currently unclear [18], and might differ with respect to brain region, type of chromosome complement and species


Here's a bit of historical perspective:

While aneuploid cells have been typically associated with pathophysiological conditions such as [...], cells in normal individuals have basically been assumed to contain identical euploid genomes [18]. Still, earlier hypotheses suggested that a number of mammalian somatic tissues are populated by polyploid cells. Adult neurons of mammals were assumed to be postmitotic cells characterized to some extent by a polyploid chromosome complement. Testing this hypothesis in the past through histochemical methods, however, yielded controversial results through technical limitations [19,20]. However, with the recent development of molecular cytogenic techniques, aneuploid cells in the normal developing and mature brain have [in 2001 and afterward] clearly been identified, indicating that the maintenance of aneuploid neurons in the adult CNS is a widespread, if not universal, property of organisation [1–11].

I sho'ly don't mean to jack this thread from the avenue suggested above by gcochran. Those with access to his "10,000-year Explosion" can check pp 96-8. There you will find the hypothesis Cochran & Harpending put forth to explain the oddly high rates of inviability found in human zygotes due to chromosome abnormalities.

While they do give a complete and self-contained account of their idea, the concepts should be particularly familiar if you happen to have read Burt & Tivers _Genes in Conflict_. (Or at least the first half of Burt & Trivers, which I confess is all I've read so far.)

The hypothesis seems quite reasonable to me, for what that's worth -- and I think it predicts with moderate confidence that zygote inviability due to chromosomal aberrance should be measurably higher in populations with a longer history of agriculture. Data on that may well come forth in the future, if it hasn't already, provided there is growing interest in the study of in vitro fertilization.

'Genes in Conflict' is a good book.

I'd like to mention that I would really hate to have to bet my life on that hypothesis. "might have something to do with it" it as strong as I want to put it. I'm sure that low human fecundability is an enigma, though: maybe an important one.

What about another possible test for your idea: other taxa that also had population explosions. Surely wheat, corn, and pigs have exploded in population about as much as agri-humans. Have they also suffered zygote mayhem, compared to their respective wild allied taxa?

[Anyone who has no idea what I'm talking about, skip to the second section below.]

One potential pitfall there is that those organisms exist somewhat "in care" of man, whereas man himself is less in care of man. However, I don't see how this domesticity should particularly disfavor meiotic drive alleles, "cosset" aneuploid individuals, or speed the rise of anti-drive antidote alleles. So maybe it doesn't matter much?

There's a worse problem. These organisms certainly all have a much shorter generation time than man, which means antidote/suppressor alleles may have already evolved to tackle the problem (if indeed it did exist in prior millennia). If so, perhaps some drive/antidrive pairs of no net effect as a pair are fixed in these genomes (this might include some pseudogene pairs). These might become evident by experimental out-breeding, but it seems like an awful lot of work to show that they exist in an unusual number compared to what exists in organisms never domesticated.

Maybe, then, investigate more recent domestics -- ones that have only been domestic for the same ~200-400 generations as many human pops have been farming?

-----------------------------------

For those lacking the book, the idea is this. Hawks, Cochran, Harpending et al (see the book or the paper in PNAS) propose a general acceleration of human evolution during agricultural times. This is based on certain empirical data -- plus the idea that a sudden explosion of novel alleles (the raw material of evolution) should have happened over the last 10,000 years simply because way more people existed over that time than ever existed before. Meiotic drive alleles are noncooperative alleles that can "sneak" into a zygote with above-50% probability, and such alleles -- this idea being more speculative -- may also have proliferated during agricultural times. Such alleles have diverse means of subterfuge, and some of them get a leg up by tampering with chromosome segregation. It's very well-demonstrated that "antidote" alleles counteracting meiotic drive alleles can evolve in time -- and indeed this is probably very much the rule, since after all the world has innumerable different species that have ~10,000 different almost-totally-cooperative genes, and these species each exist for millions of years on average. However, if a "bloom" of a number of different drive alleles have only just become common, there ought not yet be much selection pressure to drive up the frequency of the corresponding antidote alleles. (The parry cannot be begun until the lunge is manifest, if you will.) The repair lags the break. Thus, a population explosion in an organism may lead to an "un-sweet" spot during which unopposed zygote-mangling drive alleles enjoy their transient heyday, and it's not impossible humans are currently in this "un-sweet spot."

Whatever causes the travails of human zygotes, in practice the costs are modest, at least from a cold scientific perspective -- one zygote dies of chromosome problems (very very early in most cases, I would imagine) and, unless you are a couple with very marginal fertility, you simply "try" another and soon succeed. There's also the issue of trisomy 21, AKA Down Syndrome. Perhaps other mammals have less of this sort of disease, although a second variable also figures into this, namely the viability of various anomolous chromosome sets. Trisomy 21 being the only severely anomolous chromosome set that is both common and viable in man.

Humanity is essentially a highly inbred species. We don't have a lot of genetic variability compared to other common species. This may be all there is too it, too much inbreeding. Even our interracial matches would be considered inbreeding if we applied the same standards to humans that we do to animals.