1Biomolecular Structure and Modelling Group, Department of Biochemistry and Molecular Biology, University College London, London WC1E 6BT, UK.

2Department of Molecular Biology and Biochemistry, University of Malaga, Spain.

*e-mail: ranea@biochemistry.ucl.ac.uk

The idea that evolution is led by increasing genetic complexity has, over the years, annealed into the general opinion of geneticists. This increasing complexity is developed through the expensive and slow innovation cycle of gene duplication, mutation, and selection; and so, it seems contrary that a species’ fitness could be improved by losing hard-won genetic capabilities, as has been proposed by the “less is more” hypothesis (Olson, 1999). For that reason, genetic research has traditionally dealt with active genes instead of “broken” ones (or pseudogenes). The lack of a deep knowledge about the genomic significance of adaptive gene losses in mammalian genome evolution explains the important contribution of a recent study in this research area (Zhu et al, 2007). This recent work applied an ingenious method to systematically identify the losses of genes that had been long established in the human lineage over the last 75 million years. A total of 26 well-established genes, inactivated long after their birth, were identified by this analysis, with the identification of 16 previously uncharacterized human pseudogenes. This work completes former studies about pseudogene formation during human origin (Wang et al, 2006), and provides important insights for a better comprehension of this particular genetic phenomenon, in a field scarcely documented until now.

The scale of adaptive gene losses identified by this study is likely underestimate: the conservative filtering criteria and limitations in the methodology applied make the analysis far more prone to produce false negative than false positive predictions. The low sensitivity of the method reflects too difficulty of obtaining genetic evidence of advantageous inactivation events. Adaptive gene loss is difficult to demonstrate since an inactivated gene can become fixed by contrasting evolutionary causes: genetic drift following the relaxation of purifying selection in functionally redundant dispensable genes; or positive selection due to shifting environmental conditions. Distinguishing adaptive losses from the remaining pseudogene pool therefore requires additional evidence of directional selection. Such evidence is difficult to obtain, since inactivated genes favoured by adaptive selection will begin to accumulate further mutations at the background neutral rate.

Although the number of cases in which we identify losses of long-established genes is likely to rise as the sensitivity of analytical techniques improves or simply as more species are studied– we must expect that they will prove to be rare genetic events. Current data points to relaxation of selection as the main cause of a gene’s inactivation, since the huge majority of the detected pseudogenes are either “dead-on-arrival” or inactivated quickly after duplication (Wang et al, 2006; Lynch and Conery, 2000).

It may be that the majority of cases involving polymorphism for inactivation of veteran or long-established genes are explained by overdominance. Classical examples (Ringelhann et al, 1976) are the higher frequency observed for some hemoglobinophaties-related human recessive alleles in high-malaria environments where heterozygosis (recessive + functional allele copies) produces a slight advantage to the disease while homozygosis (two recessive copies) causes anemia.

At least 80 genes in the human lineage appear to have been inactivated during the last 6-7 million years since the separation from chimpanzees (Wang et al, 2006). Amongst other functions, in this set of inactivated human genes there is an over-representation of chemoreception and immunity functions, consistent with the differences observed in the human senses of smell, diet, behaviour, or susceptibility to pathogens compared to chimpanzee. Presumably gene-losses are behind other human changes occurring after the separation from chimpanzees, such as a bigger brain size, bipedalism or language capability. For example, it is speculated that the reduction of the masticatory muscles, realised by the inactivation of a human myosin gene, may have allowed the hominid brain size expansion (Stedman et al, 2004).

The study of the loss of long-established genes goes far deeper into the past, over 75 millions years, encompassing the common ancestor of apes, mice and dogs (Zhu et al, 2007). The set of functions recognised for these inactivated genes is correspondingly broader. For instance, there are genes involved in hormonal regulation, cerebellum or apoptosis, suggesting changes in the human linage through the pseudogenisation of such genes. Some of these veteran genes lost in humans are known to be still functional in mouse; others may be active elsewhere in the clade, including perhaps monkeys. This is of particular significance to when it comes to use animal models for human systems that may have been affected by the gene loss.

These losses raise the intriguing question of how these long-established genes became inactive in the human lineage. Did it happen abruptly or gradually? Did it force the genetic system to compensate with a further major change? Or did a previous change in the genetic network lead to it becoming dispensable and lost? Have changes in the human environment, population size and connectedness affected the process?

References:

Lynch M, Conery JS (2000). Science 290: 1151-1155.

Olson MV (1999). Am J Hum Genet 64: 18-23.

Ringelhann B, Hathorn MK, Jilly P, Grant F, Parniczky G.A (1976). Am J Hum Genet 28: 270-279.

Stedman HH, Kozyak BW, Nelson A, Thesier DM, Su LT, Low DW, et al (2004). Nature 428: 415-418.

Wang X, Grus WE, Zhang J (2006). PLoS Biol 4: e52.

Zhu J, Sanborn JZ, Diekhans M, Lowe CB, Pringle TH, Haussler D (2007). PLoS Comput Biol 3: e247.

Howell et al. here provide a detailed response to Prof Bandelt's commentary, arguing that mtDNA evolution is not clock-like and that the evidence for time dependent rates should not be dismissed.

The trend can be used to infer the magnitude of gene flow.

François Rousset has written papers showing how to calculate such estimates but some researchers - especially plant geneticists - use rival methods based on the statistic "Moran's I". Rousset explains why his methods are an advance on these alternatives in a recent News and Commentary published in Heredity.

Click here to read the Heredity News and Commentary

Further information was obtained by enlarging the sample. Y chromosomes and surnames are both passed from father to son, so other men sharing the same rare east-Yorkshire surname as the original man were recruited for their study in the search for additional Y’s. One third of them were also found to carry the African chromosome. Conventional genealogical research was then used to link the participants to two family trees, both dating back to the 1780s in Yorkshire. Does this evidence pin down the date at which an African ancestor arrived in the UK?

Neil Bradman and Mark Thomas are doubtful. They have published a commentary in Heredity asking whether we should be surprised by the discovery of this Y haplotype in Yorkshire at all, given the accepted wisdom that Modern Man originated from Africa. Indeed, is it not more surprising that from a survey of 421 British males, only one carried the rare African Y chromosome?

Post a comment to share your thoughts.

Click here to read the European Journal of Human Genetics article

Click here to read the Heredity News and Commentary

Professor Marion Petrie and Dr Gilbert Roberts at Newcastle University provide another explanation for this paradox in their paper in the April issue of Heredity. The authors, using computer simulations, show that female choice can in fact lead to bursts of mutation affecting male attributes. The elevated mutation would throw up even more attractive males, sadly mutation is a blind and random process, so at the same time it produces the less attractive as well.

Click here to read the Heredity article

Click here to read Cotton and Pomiankowski's Heredity News and Commentary on this paper

Click here to read King and Kashi's Heredity News and Commentary on this paper

Click here to view the BBC story on this paper

Click here to view the Times comment on this paper

This month we find out if inbreeding spells doom for endangered populations, whether bottlenecks can lead to degeneration of genes, how X chromosome influences sperm length and much more.

Another model is to publish papers accompanied by referees’ comments, and correspondence columns offer an analogous forum to air legitimate differences of scientific opinion.

This blog will attempt to combine some of the better aspects of these approaches. It will evolve in the light of Heredity readers and authors’ recommendations, but I have in mind that we will use it to

1) Publish rapid feedback on papers that have appeared in Heredity: the paper is wrong because … , readers should also see paper x because … . i.e. something like a correspondence section of a journal, but with the merit of being very fast and brief. The more incisive or interesting comments could be published in print.
2) Publish comments provided for public consumption by the referees: the paper is controversial because … but I recommend publication because … .
3) To publish discussion on the editorial direction of the journal. In large part the content of Heredity is determined by what is submitted, but the News & Commentaries, Short Reviews and Special Issues are commissioned. Are we neglecting important areas?

Contributions will be screened before being posted. Please reply to the appropriate topics to contribute.