The sex-ratio trait, reported in a dozen Drosophila species, is a type of naturally occurring meiotic drive in which the driving elements are located on the X chromosome. Typically, as the result of a shortage of Y-bearing spermatozoa, males carrying a sex-ratio X chromosome produce a large excess of female offspring. The presence of sex-ratio chromosomes in a species can have considerable evolutionary consequences, because they can affect individual fitness and trigger extended intragenomic conflict. Here, I present the main results of the study performed in Drosophila simulans. In this species, the loss of Y-bearing spermatozoa is related to the inability of the Y chromosome sister-chromatids to separate properly during meiosis II. Fine genetic mapping has shown that the primary sex-ratio locus on the X chromosome contains two distorter elements acting synergistically, both of which are required for drive expression. One element has been genetically mapped to a tandem duplication. To infer the natural history of the trait, the pattern of DNA sequence polymorphism in the surrounding chromosomal region is being analysed in natural populations of D. simulans harbouring sex-ratio X chromosomes. Initial results have revealed the recent spread of a distorter allele.
The primary characterization of the sex-ratio meiotic drive dates back more than three-quarters of a century, when Gershenson  investigated the causes of a strong bias towards females in the progeny of certain Drosophila obscura males descending from wild-caught flies: “The factor which calls forth the appearance of female culture is transmitted by the X chromosome…this gene is absolutely sex-limited…it provokes a sharp preponderance of females by almost totally removing the spermatozoa with the Y chromosome from the fertilization process”. Since it was first described by Gershenson , the trait has been reported in a dozen Drosophila species and two other families of dipterans, giving rise to numerous theoretical and empirical studies, most of them in the field of evolutionary biology and population genetics  and dealing with: (i) the population dynamics of sex-ratio segregation distorters, in relation to their segregation advantage at the gamete stage and their deleterious effects at the zygotic stage; (ii) the evolution of drive suppressors on the Y chromosome and on the autosomes, which results from the intragenomic conflict triggered by the sex-ratio X chromosome; and (iii) sexual selection, the evolution of mating behaviour and speciation. Less attention has been paid to the molecular and cellular basis of sex-ratio drive. One of the reasons for this disaffection is that the distorter elements are usually associated with complex chromosome inversions , which impedes their genetic mapping.
In the present paper, I review investigations of the sex-ratio trait discovered 11 years ago in D. simulans . The lack of associated inversion in this species has made it possible recently to identify candidate distorter genes on the X chromosome, offering a new window through which to investigate the process of meiosis and providing molecular tools with which to study the evolution of the trait in populations and species.
The cellular mechanism of drive
Sex-linked distorters acting in males seem to differ from autosomal distorters with regard to the stage of spermatogenesis at which defects are first observed. In the two well-studied autosomal systems, Segregation Distorter in Drosophila melanogaster and the t-complex in mice, meiosis unfolds normally, the differential numbers of functional spermatozoa recovered being related to defects expressed at post-meiotic stages [4,5]. In contrast, changes in chromosome behaviour during meiosis usually occur in sex-linked distorters, including sex-ratio distorters in Drosophila pseudoobscura, Drosophila athabasca, Drosophila subobscura and D. simulans [6–8].
In D. simulans, meiosis has been investigated in males carrying a sex-ratio chromosome from a reference stock producing approx. 90% female progeny . The only defect observed was at anaphase II and this consisted of a failure of disjunction of the Y chromosome sister-chromatids. Most of the Y-bearing cells (∼95%) were affected. In most of the abnormal plates (∼63%), the pictures showed that sister-kinetochores had separated and moved towards each of the poles, whereas the sister-chromatids remained connected and stretched between the kinetochores. A very thin chromatin fibre could usually be seen between the separating Y parts, suggesting that the chromatids did not break as they stretched (Figure 1A). In the other plates, there was no evidence that the kinetochores separated. Both Y chromatids, lagging behind the autosomes, remained connected and moved together towards the same pole (Figure 1B). The spermiogenesis of sex-ratio males was characterized by the failure of some spermatids to undergo normal elongation . Fluorescence in situ hybridization, using DNA probes specific to each sex chromosome, demonstrated that the deficient spermatids came from spermatocytes II where the Y sister-chromatids had not segregated properly . Their nuclei appeared to be either Y-labelled or unlabelled, often paired and connected by a thin DNA bridge. Y-labelled and unlabelled nuclei were also found among the normal-looking spermatids. Some of them may also have been defective since the excess of females observed in the progeny of sex-ratio males was higher than would be predicted from the observed number of deficient spermatids. However, the identification of XO males in the progeny indicated that some nullo-XY spermatids have become functional spermatozoas.
Orcein-staining of late anaphase II in sex-ratio males of D. simulans showing disjunction failure of the Y chromosome
Consequences for male fertility
Based on the cytological findings described above, the loss of Y-bearing spermatozoa in sex-ratio males does not appear to be compensated for by the overproduction of X-bearing spermatozoa. They were therefore thought to produce less functional spermatozoa than wild-type males, a feature that could be disadvantageous in situation of spermatozoa limitation or spermatozoa competition. This was first suggested by the observation that a sex-ratio chromosome was quickly eliminated from experimental populations by a non-driving chromosome  and was proven conclusively by measuring male fertility under different mating conditions . Sex-ratio and wild-type males were found to be equally fertile when allowed to mate repeatedly and freely with one or two females per male during the whole period of egg laying. However, when only one mating per male was allowed, the number of progeny was significantly lower for sex-ratio males than for wild-type males. The disadvantage of sex-ratio males was still more marked when the males were offered multiple mates (ten females) for a limited time (24 h), the toll on male fertility outweighing the segregation advantage of the sex-ratio chromosome. Finally, they also showed a strong disadvantage in spermatozoa competition. The ability of sex-ratio distorters to spread in the wild is therefore expected to depend markedly on local mating conditions, with the population density and the operational sex ratio probably being key parameters.
These fertility defects show many similarities to those reported in other species . Although their impact on the population dynamics of the trait has been the subject of many theoretical investigations [2,12], we still know nothing about mating conditions in the wild, a point of crucial importance if we are to understand/predict the evolution of the trait.
Towards the molecular identification of sex-ratio distorters
As is usually the case for other meiotic drive systems, the sex-ratio trait in D. simulans results from the combined effects of several loci, one of which has a major impact . This primary locus was recently characterized on a reference sex-ratio X chromosome originating from the Seychelles . It is located in the cytological bands 7E-8A of the polytene chromosome and contains two distorter elements acting synergistically, both of which are required for drive expression (Figure 2).
Organization of the primary drive locus on the sex-ratio chromosome XSR6 of D. simulans, deduced from the genetic mapping experiment described in 
One element was genetically mapped to a duplication inserted in a direct orientation, at a short but still undetermined distance from the original fragment. Drive persists after recombination that changes the alleles within a very large part of the duplication, which suggests that it does not involve specific, functionally divergent alleles, but rather a difference in gene dosage where two copies confer drive. The duplication contains six annotated genes: Trf2, CG32712, CG12125, CG1440, CG12123 and org-1. Among them, Trf2 is the only one known to be involved in meiosis: certain defects in Trf2 cause chromosome non-disjunction in D. melanogaster . It encodes a member of the TATA-box-binding protein family . The complex utilizing Trf2 has numerous targets, including genes involved in DNA replication and in chromatin remodelling . The original and duplicated copies of Trf2 on the sex-ratio chromosome are both expressed in the testes of D. simulans sex-ratio males.
The second drive element was mapped to a candidate region located approx. 110 kb away, spanning seven annotated genes: CG11265, CG12111, CG2056, CG12065, CG12081, CG12659 and Crag. All these genes, apart from CG12081 and CG12659, are expressed in the testes. CG11265 is an attractive candidate because the product encoded by its yeast homologue is a DNA polymerase required for sister-chromatid cohesion .
What is now needed is a quantitative study of gene expression as well as transgenic experiments to identify the distorter elements among the candidate genes. However, it is formally possible that drive is caused by a rearrangement of non-coding material, rather than by gene duplication itself.
D. simulans, a model for studying the evolution of sex-ratio drive
As highlighted in the Introduction section, sex-ratio drive may have a considerable impact on the evolution of species . This makes it important to find out how common it is and how it evolves.
It has been postulated that sex-linked distorters may be widespread in species, but that they are usually undetectable because they are completely neutralized by suppressors, which are promoted by natural selection in order to restore equal sex proportions . Sex-ratio in D. simulans was the first case reported that supported this prediction [3,20]. The frequency of driving X chromosomes varies widely within this worldwide species and can reach high values (up to 60%), but an efficient system of Y-linked and autosomal suppression maintains an approx. 1:1 sex ratio in populations and bias towards females is rarely observed in the progeny of wild-caught flies .
The evolution of sex-ratio drive in species is a subject of debate, fed by the high variability of its characteristics among Drosophila . Because sex-ratio chromosomes seem to be stable at moderate frequency in natural populations of several species, it has been proposed that they are maintained in balanced polymorphism as a result of equilibrium between their segregation advantage and their deleterious effects on fitness . An alternative could be that non-steady state dynamics dominates the evolutionary history of sex-ratio systems and that waves of invasions of distorters and suppressors recurrently occur. On that point, the situation revealed by a wide geographic survey of D. simulans is appealing. There is a contrast between (i) a large African range where there are very variable frequencies of sex-ratio X chromosomes (from 0 to 60%) and suppression is always strong and (ii) the rest of the world, where sex-ratio chromosomes are usually rare if not totally missing and suppression is moderate or non-existent [21,22]. This may reflect either the existence of different balanced equilibria, depending on local demographic and ecologic conditions, or the current dissemination of a sex-ratio chromosome that originally arose somewhere in Africa. To address this question, we have started a population study of DNA sequence polymorphism in the area surrounding the primary drive locus on the X chromosome. The results obtained with a first polymorphic marker, located within the gene Nrg between the two candidate regions, showed the signature of a strong selective sweep induced by a distorter allele in Madagascar and Réunion, two islands in the Indian Ocean where approximately one-half of the X chromosomes are sex-ratio . The complete association of the sex-ratio trait in samples from both islands with a unique haplotype of the marker indicated that this allele must have spread recently. Its age was estimated to be less than 18000 years in Madagascar, an upper limit that may be greatly overestimated.
These first findings are very promising, suggesting that thorough investigations using additional markers in the surrounding chromosome region and analysing samples from other populations will allow us to trace the history of the trait in the species. This molecular population genetics-based approach should also help to identify the distorter alleles on the X chromosome.
Meiosis and the Causes and Consequences of Recombination: Biochemical Society Focused Meeting held at University of Warwick, U.K., 29–31 March 2006. Organized by R. Borts (Leicester, U.K.), D. Charlesworth (Edinburgh, U.K.), A. Eyre-Walker (Sussex, U.K.), A. Goldman (Sheffield, U.K.), G. McVean (Oxford, U.K.), D. Monckton (Glasgow, U.K.), G. Moore (John Innes Centre, U.K.), J. Richards (Roslin Biocentre, U.K.) and M. Stark (Glasgow, U.K.). Edited by D. Monckton.