This is a fascinating article in The Scientist which has caused me to scratch my head a lot. Rajiv McCoy, Dmitri Petrov, and colleagues, at Stanford University, have been exploring the phenomenon of aneuploidy - departures from normal chromosome number - in embryos. Aneuploidy, and a range of other chromosomal errors, are surprisingly common in human embryos, accounting for substantial pregnancy loss. Genetic screening of a great number of embryos, including a genome wide association study which interrogates the whole genome for minor contributions from genes toward the effect under study, revealed a striking link, says the article, between mitotic aneuploidy (that occurring as the zygote begins cell division to form the early embryo or blastocyst) and a single nucleotide variant (SNP rs2305957) on chromosome 4 of maternal genomes. They singled out the gene PLK4 from this chromosome region because it is responsible for organizing the so-called mitotic spindle, the structure that forms in dividing cells along which sister chromatids move to opposite poles to form the nuclear genetic component of the two daughter cells. Any gene variant that caused the gene to make this process faulty could result in whole chromosomes, or bits of chromosomes, failing to arrive at their destinations in the sister nuclei - therefore causing aneuploidy. In other words, to their shock, they found a major potential cause of female infertility was due to a gene variant from the mother's genome! This gene variant does not exist in genome scans of Neanderthals or Denisovans and, they estimate, must have arisen, and then been subject to positive selection, between 400,000 and 100,000 years ago. What extraordinary selection pressure, operating on human ancestor females, could possible equate with the evolution of a gene that scrambles chromosomes and causes many embryos to perish - vastly reducing reproductive fitness? It doesn't make sense. The authors have flown a kite. They note, from the evolutionary anthropological literature, that concealed ovulation and continual sexual receptivity have evolved in the females of human ancestors in order to confuse males over paternity - making it difficult for males to track whether or not they are the likely father of any children born to women to which they have had sexual access. This paternity confusion may improve chances of females eliciting material support from men while weaning infants, because they have reasons to suspect the child might well be theirs, or, alternatively, may reduce the risk of infanticide from males wanting to rid females of all progeny they suspect are not theirs. Could it be, ask the authors, that reduced fertility through a mechanism that causes aneuploidy in the embryos they carry, also acts to sow confusion over paternity? The reduced fertility might then compute as a cost worth bearing if it equated to large amounts of male support for those infants they do produce.
This is too much for me to swallow and I'm very skeptical that they have discovered a reproductive adaptation for the reasons they posit. There are a number of weaknesses in the story. Is reduced fertility in reality capable of causing paternity confusion? Was it really selectively advantageous for females in this period of human pre-history to evolve a mechanism for paternity confusion that so dramatically lowered their fertility - reproductive fitness? Humans are one of the least fertile species on the planet and there are several causes of aneuploidy, and other reasons, why there is such wastage of human embryos. There is no clear path from the gene variant of PLK4 they have identified to the phenotype (infertility through aneuploidy) they are investigating. Also, if the maternal variant that confers a high risk of aneuploidy is present at high frequencies in the population - sometimes up to 30% - it must be being maintained in balanced polymorphism with the major variant - the original version of the gene. If the high-risk variant confers real evolutionary advantages to females through paternity confusion, it is not clear why it has not swept to fixation, displacing the original gene. In other words, it is not clear what is maintaining the original gene in the population - how is the polymorphism balanced? They are also assuming that PLK4 has in fact been the target of selection, rather than a gene that has hitchhiked to substantial frequency in human populations on the coat-tails of a neighbouring gene. It is obvious why they went for PLK4 - because it is so intimately involved in cell division - but are we sure that they have been investigating the right gene? Joris Vermeesch, at the University of Leuven, has shown, in perfectly normal embryos from normal, fertile women, that chromosomal abnormalities like aneuploidy occur in over 90% human embryos - and often all blastomeres are afflicted. Yet, although the attrition among human embryos is very high - at over 70% - it is nowhere nearly as high as the percentage - over 90% - of genetically chaotic embryos. So, while chromosomal instability like aneuploidy may be fatal for embryos, or the majority of blastomeres within an embryo, it is by no means a death sentence. Finally, PLK4 is not the only gene in this so-called haplogroup (chunk of DNA sequence) that has been under positive selection. The authors mention a handful of other genes but then ignore them. The peril lies in the perfectly logical choice of a "culprit" gene which then leads to casting round for an evolutionary context that seems to fit the story. A story which will, I am sure, unfold with time, and I will be watching.