_____________________________________
O R I G I N . O F . H U M A N S
DID LANGUAGE REDUCE SEQUENCE VARIATION?
Andrew Gyles
___________________________Return to main index page
CLICK ON TITLE TO GO STRAIGHT TO ARTICLE
- Recent articles on genetic recombination and biased gene conversion
- Selective advantage of language might have driven low sequence variation in human genes
- The 'bushy' hominid tree, hominid extinction and human linkage disequilibrium
ARTICLES ARE ARRANGED BELOW BY DATE OF PUBLICATION, NEWEST AT TOP
________________________________CONTENTS
Recent articles on genetic recombination and biased gene conversion
In my two articles lower on this page I suggest that biased gene conversion might have been important in human evolution and might explain the low sequence variation in the nuclear genes of humans.
I admitted that there was little published information on biased gene conversion in many species of organism.
Two articles just published online before print, 27 November 2001, in the Proceedings of the National Academy of Sciences ( http://www.pnas.org/ ) provide some new data on genetic recombination and on biased gene conversion in simple eukaryotic organisms. They are:
High frequency mitotic gene conversion in genetic hybrids of the oomycete Phytophthora sojae, by Jureerat Chamnanpunt, Wei-xing Shan, and Brett M. Tyler.
These authors reported that "At many loci, conversion showed extreme disparity, with one allele always being lost, suggesting that conversion was initiated by allele-specific double-stranded breaks". In other words, there was biased gene conversion at many loci.
Meiotic recombination frequencies are affected by nutritional states in Saccharomyces cerevisiae, by Mohamad F. F. Abdullah and Rhona H. Borts.
One of the conclusions the latter authors drew from their work was that "the effects of metabolic state may be global and may account for some as yet unexplained features of recombination in higher organisms, such as the differences in map length between the sexes".
Possible relevance to my hypotheses
The organisms studied in the above two papers are simple ones. The first paper studied mitotic gene conversion, but of course only meiotic gene conversion could be important in human evolution. So the papers seem to have little direct relevance to my hypotheses. They may, however, point to similar mechanisms not yet discovered in higher organisms, which would be relevant.
There is no connection between the above authors and me.
Published 04 December 2001. © Andrew Gyles
________________________________CONTENTS
Selective advantage of language might have driven low sequence variation in human genes
The sequence variation in human genes is low compared with that in apes. (This is the variation in sequence between individuals in a species.) It is low in the mitochondrial chromosome and it is low in nuclear chromosomes.
The explanation given in the standard model of human evolution is that the human species has recently expanded from a small or 'bottleneck' population size. Indeed, four 'bottlenecks' have been proposed: a bottleneck in Africa, a bottleneck outside Africa, a bottleneck in the immediate ancestors of the northern Europeans and a bottleneck in the Neanderthals. The Neanderthal bottleneck was proposed after it was found that there was probably low sequence variation in Neanderthal mitochondrial chromosomes.
I have preferred to try to find other explanations for the low sequence variation in humans, partly because four bottlenecks seemed to me implausible (why did not other animals pass through bottlenecks recently?) and partly because I agree with the published remark by two 'diehard' palaeontologists that Mongoloids seem to have inherited some of the traits of hominid fossils found in China (see my page 'Multiregional evolution', accessible through the main index page).
Mitochondrial chromosome
I suggested in some of the articles in 'Multiregional evolution' that language could not be fully developed and refined unless the mitochondrial performance in the various classes of nerve cell that make up the brain remained highly standardised in each individual. Therefore the net mitochondrial mutation rate in somatic cells and female germline cells in humans was driven down by the selective advantage conferred by refinements in the use of language. I described how this could have been achieved by the elimination of any mitochondrial chromosome in which a mutation was physically detectable.
Nuclear chromosomes
I turn now to the low sequence variation in nuclear chromosomes in humans. I propose that the selective advantage given by refinements in the use of language drove down this rate too, because the performance of the various classes of cell involved had to be standardised between individuals. Language made each individual a part of a 'phone system' in which sentences had to be formed, spoken, heard and understood by all individuals in the same way. These tasks depend on brain cells, muscle cells, cartilage cells, and sensory cells in the ear. Cells that form the chest, larynx (the organ of voice), tongue, teeth, ears and brain are all involved in the use of spoken language.
Standardised performance of components
Telephone companies faced with the much simpler task of transducing, transmitting and reproducing sounds seek classes of electronic component with standardised performance so that the tasks of design and maintenance of a huge network can be based on predictable outcomes. It therefore seems likely that the evolution of the use of language required a great standardisation in the performance of all of the cells involved in the use of spoken language. This performance depends on many nuclear genes, so the use and refinement of language evolved most rapidly in groups of humans in which the variation of the sequence of each nuclear gene (comparing individual with individual in the group) was low and stayed low as generation followed generation.
Reduction of variation through biased gene conversion
What mechanism could have been naturally selected that would reduce the sequence variation in humans compared with other animals? I suggest that it was a biased form of 'gene conversion', in which mismatched pairs of bases in the heteroduplex segments formed during meiosis are repaired. One base of any mismatched pair is cut out and is replaced with the correctly fitting base. If this is done in a biased way the sequence variation in nuclear chromosomes will be gradually reduced at each meiosis. Perhaps this (hypothetical) kind of 'biased gene conversion' began in our hominid ancestors.
The only bias that I can think of is the possibility that the repairing mechanism has a preference for removing certain bases from mismatched pairs of bases. I do not have specialist knowledge of genetics. Geneticists inform me that some forms of biased gene conversion have been observed in some organisms, but that detailed statistics on rates for many species are not yet available. So this article is speculative and hypothetical
(I sent this article to the internet discussion group 'talk.origins' on 07 September 2001, where it was subsequently published.)
Published on this site 07 September 2001. © Andrew Gyles
________________________________CONTENTS
The 'bushy' hominid tree, hominid extinction and human linkage disequilibrium
Fossil evidence indicates that there were many lines of hominid living in different parts of Africa millions of years ago, one of which led to the human species. The hominid tree was 'bushy'. A separate line (not a 'bushy' tree, as far as is known) led from the common ancestor of the hominids and the chimpanzees to the chimpanzees. (No fossils of chimpanzee ancestors have been found.)
Rapid evolution through loss of alleles
I propose that the common ancestor of the hominids adapted to a changing climate and environment by rapid genetic change. This change was not the gaining of alleles through random mutation (which could hardly be rapid without being fatal to the survival of the species). It was the rapid loss of alleles in the germ line through unusually vigorous, biased repairing of mispaired bases in heteroduplex segments during crossing over in meiosis.
In each of these repairing events an allele is lost from one strand of the heteroduplex and a different allele is gained by the other strand. If the repairing mechanism has a preference for removing certain bases of mismatched pairs the repairing will be biased, and the frequency of certain alleles will decrease.
The results of these repairing events are called gene conversions. My hypothesis depends on the possibility that the repairing events, and the resulting gene conversions, are biased. I do not know of any evaluation of biased gene conversions in human and other genomes. I hope that such work is being done.
Alternatively the length of heteroduplex segments might have increased in the hominids; this combined with the biased repairing of some of the mispaired bases would have increased the rate of loss of alleles.
I propose that as the hominids spread over a broader area of Africa and became isolated into several groups or 'gene pools' the vigorous biased repairing in heteroduplexes continued and the loss of alleles continued. The resulting frequencies of particular alleles in each of the gene pools might have begun to differ from pool to pool. This would have caused the average phenotype in each gene pool to differ from that in each of the other gene pools. Thus the hominid tree might have developed many branches and become 'bushy', as the hominid fossils found in different parts of Africa attest.
I propose that in the ape line of descent from the common ancestor of humans and chimpanzees the biased repairing of mispaired bases in heteroduplexes was not as vigorous as it was in the hominids; the result was the chimpanzee.
Rapid loss of alleles not necessarily fatal
The rapid loss of alleles from the gene pool of a species well stocked with them would not necessarily be fatal to the survival of the species. (This cannot be said of the rapid gain of alleles through random mutation.) There are, however, two important classes of allele that must not be lost from the gene pool of the species (or from an isolated group of interbreeding individuals within the species) if it is to survive.
The first class comprises the many alleles of the histocompatibility genes, which must be present in the gene pool of a species to ensure its survival in spite of the presence in its environment of any of various parasites.
The second class comprises beneficial dominant alleles that help an individual to survive and reproduce even if it bears a deleterious recessive allele at the same locus in the homologous chromosome. If the gene pool of a species lost too many beneficial dominant alleles and was left with too many deleterious recessive alleles the species might lose the struggle for survival.
Alleles in the first and second classes would be preserved in the gene pool by natural selection unless the overall rate of loss of alleles through biased repairing of mispaired bases in heteroduplex segments during meiosis left too few of them for natural selection to preserve. This might explain why all but one of the hominid lines became extinct: perhaps the line of hominids leading to humans was the only one lucky enough to retain all of the alleles needed for survival and to lose many deleterious alleles.
Linkage disequilibrium between widely separated markers increased by loss of alleles
I suggest that the great linkage disequilibrium between widely separated markers that has recently been measured in the chromosomes of some human groups is a reflection of the (hypothetical) increased rate of biased repairing of mispaired bases in heteroduplexes, or of increased length of heteroduplexes (or a combination of both of these) in hominids, including humans. In particular, an increase in length of the heteroduplex segments formed during meiotic crossing over in hominids and humans, combined with a greater rate of biased repairing of mispaired bases in these segments, would cause great linkage disequilibrium between widely separated markers.. This would be so because the different 'markers' by which linkage equilibrium is measured - variations in base sequence from one individual to another in a population - would become fewer at each marker position, and in some positions might be reduced to no variation.
The end result in the human species would have been remarkably little sequence variation between individuals as compared with other species.
Mechanism might have been used before
The (hypothetical) increase in the rate of loss of alleles during meiosis might have been the product originally of a few random genetic mutations in the ancestor of the hominids. However, it is worth considering the alternative possibility that this mechanism might be common to some other species of organism, might have been used several times before in the course of evolution and might be hundreds of millions of years old. It might be lying latent in these species waiting to be called into action whenever rapid environmental change threatened the survival of a species.
(This is a slightly revised version of an article I published on this site on 20 August 2001. The revision emphasises the fact that my hypothesis depends on gene conversion in hominids having been biased, so that the frequencies of certain alleles in the gene pool decreased.)
Published on this site 06 September 2001. © Andrew Gyles
________________________________CONTENTS