Signature of Selective Sweep Associated With the Evolution of sex-ratio Drive in Drosophila simulans

Corresponding author: Catherine Montchamp-Moreau, CNRS, Avenue de la Terrasse, 91198 Gif-sur-Yvette Cedex, France., catherine.montchamp{at}pge.cnrs-gif.fr (E-mail)

Communicating editor: S. W. SCHAEFFER

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In several Drosophila species, the XY Mendelian ratio is disturbed by X-linked segregation distorters (sex-ratio drive). We used a collection of recombinants between a nondistorting chromosome and a distorting X chromosome originating from the Seychelles to map a candidate sex-ratio region in Drosophila simulans using molecular biallelic markers. Our data were compatible with the presence of a sex-ratio locus in the 7F cytological region. Using sequence polymorphism at the Nrg locus, we showed that sex-ratio has induced a strong selective sweep in populations from Madagascar and Réunion, where distorting chromosomes are close to a 50% frequency. The complete association between the marker and the sex-ratio phenotype and the near absence of mutations and recombination in the studied fragment after the sweep event indicate that this event is recent. Examples of selective sweeps are increasingly reported in a number of genomes. This case identifies the causal selective force. It illustrates that all selective sweeps are not necessarily indicative of an increase in the average fitness of populations.

A seminal article by MORGAN 1910 on sex-linked inheritance in Drosophila showed that the sex ratio of the progeny in fruit flies results from a simple XY Mendelian segregation. The observation of sex-ratio X chromosomes in natural populations of Drosophila obscura (GERSHENSON 1928 ) and D. pseudoobscura (STURTEVANT and DOBZHANSKY 1936 ), however, revealed that genes sometimes circumvent Mendel's law. The sex-ratio trait has since been reported in males of a dozen Drosophila species and more recently in Diopsidae (reviewed by JAENIKE 2001 ). It is caused by X-linked meiotic drive factors that induce the loss of Y-bearing sperm, resulting in the production of female-biased progeny. The spread of X-linked drivers induces biases in the population sex ratio and confers an advantage to parents reducing the bias in their offspring (FISHER 1930 ). This can be viewed as a genetic conflict between the X chromosome and the rest of the nuclear genome, which induces a selective pressure in favor of drive suppressors on the Y chromosome and the autosomes (HAMILTON 1967 ). It has been proposed that distorters and suppressors occur as balanced polymorphisms in natural populations (discussed in CARVALHO and VAZ 1999 ). An alternative is that sex-ratio systems evolve through waves of recurrent invasions, making it necessary to characterize a model species where sex-ratio events can be easily followed in natural populations.

Drosophila simulans is a potential model for addressing this question. Sex-ratio chromosomes are widespread in the African range of this cosmopolitan species, from Sao Tome Island, on the west coast, to Réunion Island, on the east coast, suggesting that the system was established long ago (JUTIER et al. 2004 ). An efficient suppression system has evolved, thus maintaining an 1:1 sex ratio, even in populations where sex-ratio chromosomes are at a high frequency (ATLAN et al. 1997 ). There are virtually no inversions in D. simulans, making it easier to map sex-ratio factors and to interpret DNA variation than in D. pseudoobscura (BABCOCK and ANDERSON 1996 ; KOVACEVIC and SCHAEFFER 2000 ) where a characteristic "sex-ratio" chromosome arrangement occurs. When segregation distortion is completely suppressed, sex-ratio chromosomes are expected to be lost or fixed by random drift if neutral and to be eliminated by selection if deleterious. Their distorting ability may also degenerate. This may explain why distorting X chromosomes are rare in some African locations and frequent in others. About one-half of X chromosomes from Madagascar and Réunion are sex-ratio. This is the highest frequency ever recorded in a Drosophila species (JUTIER et al. 2004 ) and suggests that these populations experienced a recent spread of drive factors.

No D. simulans sex-ratio factors have been characterized at the molecular level, making it difficult to study their dynamics in natural populations. Therefore, characterizing the genotype of individuals involves genetic tests, through counting and sexing the progeny of individual males in a standard background. This is achieved after several generations of crosses in the laboratory. These techniques cannot be applied on a large scale even though it would be worthwhile to distinguish the frequency of distorting elements from the distribution of their strength. In addition, they cannot provide information about past events. The molecular characterization of distorter loci would allow us to study selective events associated with the evolution of sex-ratio X chromosomes in populations by using the sweeping effect of selection on DNA neutral variation (AGUADE et al. 1989 ; KAPLAN et al. 1989 ).

Here, we present the first molecular genetic study attempting to characterize sex-ratio in D. simulans and to trace its history. First, we used a collection of recombinant X chromosomes to fine-map the sex-ratio candidate region using molecular markers. Second, we looked for the molecular signature of selective sweeps within the candidate region in natural populations from Madagascar and Réunion. We used the vermilion gene as a neutral reference marker (HAMBLIN and VEUILLE 1999 ) to distinguish between local and genome-wide departure from neutral equilibrium model.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks and genetic study:
The stocks used were previously described by MONTCHAMP-MOREAU et al. 2001 and MONTCHAMP-MOREAU and CAZEMAJOR 2002 . The ST8 standard stock provides a reference genetic background free of distorters and drive suppressors. All crosses mentioned below were carried out using lines with an ST8 background obtained through repeated backcrossing to the ST8 stock. Males from the SR6 line carry a sex-ratio X chromosome (X^SR6). The sn,lz/C(1)RM line is a stock in which females carry the compound X chromosomes from the lz[sp]/C(1)RM,y,w stock (Bloomington Stock Center), whereas males carry a standard (nondriving) X chromosome with the singed and lozenge mutations. The X^sn,lz chromosome was obtained by recombination of chromosomes from the lz[sp]/C(1)RM, y w stock with the sn stock (Bloomington Stock Center).

Recombinant chromosomes between the sex-ratio X^SR6 (producing 95% females in the ST8 background) and the standard X^sn,lz were obtained according to the crossing scheme given in Fig 1. Phenotypically [+lz] recombinants were maintained in male lineages through repeated backcrosses with C(1)RM,y,w females. Each recombinant chromosome was tested for its segregation ratio in the drive-sensitive ST8 background, following the protocol of MONTCHAMP-MOREAU and CAZEMAJOR 2002 .

View larger version (14K): In this window In a new window Download PPT slide   Figure 1. Crossing scheme used to obtain recombinant X chromosomes in mapping experiments. All lines are in an ST8 background.

The X^SR6 sex-ratio chromosome had been extracted from a laboratory stock originating from the Seychelles, and no population sample was available. For the population study, we used X chromosomes from the same area (islands off the eastern coast of Africa). These chromosomes came from random samples of flies collected in December 1996 at La Saline (Réunion) by C. Montchamp-Moreau and in 2000 at Antananarivo (Madagascar) by M. Veuille. Iso-X lines were started in the laboratory by crossing single wild-caught males with females carrying compound C(1)RM,y,w X chromosomes. Line maintenance and the characterization of sex-ratio and standard phenotypes are described in MONTCHAMP-MOREAU and CAZEMAJOR 2002 and ATLAN et al. 2003 . Males from each iso-X line were frozen and stored at –80° until use.

DNA polymorphism survey and data analysis:
Extraction of single male genomic DNA, PCR amplification, and sequencing reactions were carried out using classical protocols (see, e.g., HAMBLIN and VEUILLE 1999 ). The PCR amplification primers for the Nrg, otd, otu, and rdgA genes are shown in Table 1. For the v gene, we used formerly designed primers (HAMBLIN and VEUILLE 1999 ) and sequenced a region overlapping positions 654–1350 of GeneBank accession no. U27204. For the single-strand conformation polymorphism (SSCP) study, PCR products were denatured 2 min at 95° in 40% formamide and loaded on 8% polyacrylamide/7.5% glycerol nondenaturing gels in 1x TBE. Migration was run for 9 hr at 24 W at room temperature. Sequencing was carried out using an ABI-310 automated sequencer. Sequences were aligned using the Bioedit program, and the analysis of molecular variation was carried out using Allelix (MOUSSET et al. 2003 ), DNAsp 3.53 (ROZAS and ROZAS 1999 ), and MEGA version 2.1 (KUMAR et al. 1994 ).

We checked departure from a standard neutral mutation-drift equilibrium model using several tests. The Hudson-Kreitman-Aguadé (HKA) test (HUDSON et al. 1987 ) considers selection in one locus compared to a reference locus on the basis of population variation and divergence from an outgroup for the two loci. TAJIMA's (1989) D and FU and LI's (1993) D statistics consider departure from neutrality in the frequency pattern of a sample of polymorphic sites by examining the consistency of several estimates of the neutral mutation parameter that are based on different predictions of the infinite site model. The H and K haplotype tests (DEPAULIS and VEUILLE 1998 ) consider departure from neutrality in the frequency distribution of haplotypes on the basis of information from polymorphic sites. Haplotype tests were run using a constant number of mutations (S) for a sample of n chromosomes (DEPAULIS et al. 2001 , DEPAULIS et al. 2003 , DEPAULIS et al. 2004 ; WALL and HUDSON 2001 ). In practice, we will report results of the H-test, which uses haplotype heterozygosity and was run using 10,000 iterations. The K-test always gave results consistent with the H-test.

These tests may yield different results depending on whether intragenic recombination is taken into account or not. We thus ran them under both models. For tests with recombination, we used an experimentally derived recombination rate for D. melanogaster in Nrg (9.96 x 10^–9 event/bp/generation) as calculated using the method of AQUADRO et al. 1994 . This value was multiplied by 2/3 (thus assuming a 1:1 sex ratio), as our marker is X linked. We obtained a recombination rate r = 6.63 x 10^–9/site/generation. This recombination rate was conservative for the tests being used, as the value obtained from genetic experiments in D. simulans, 5.8 x 10^–8, was higher (see RESULTS). For vermilion, we used the recombination rate estimated from D. melanogaster (r = 3.60 x 10^–8 events/generation/bp). This value was higher than that for Nrg. However, as v was used as a reference neutral locus to compare with Nrg, using this estimate was conservative.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Characterizing the sex-ratio genomic region:
CAZEMAJOR et al. 1997 previously showed that a major sex-ratio effect is caused by one or more genetic factors located between the nipped wing (np) and lozenge (lz) genes. Further genetic experiments using the intervening locus sn (located between np and lz) showed that this or these factors are located within the sn-lz interval (data not shown). The recombination rate between sn and lz was 7.7 ± 2.6 cM. These experiments yielded 37 [+lz] X chromosomes derived from a cross between the [++] X^SR6 sex-ratio chromosome and the [sn lz] marker stock. Following the definition of sex-ratio phenotype (MONTCHAMP-MOREAU and CAZEMAJOR 2002 ), 14 [+lz] chromosomes were sex-ratio (yielding 77–96% female progeny), and 23 were standard (producing 46–66% female progeny).

Since heterozygosity in D. simulans intron DNA is in the 1–2 x 10^–2 range, the probability of the two parental stocks used in the mapping experiment showing different SSCP alleles in an 200-bp PCR fragment was very high. Association between these alleles and the sex-ratio phenotype in recombinant lines was recorded for mapping sex-ratio genes. Results are shown in Fig 2. We used PCR fragments from the otu, Nrg, otd, and rdgA loci that had been cytologically characterized before the first release of the Drosophila Genome Project. We were able to characterize a sex-ratio region located between markers from the otu and otd loci. Mapping ended when recombination spots in our recombinant lines were separated by the intervening Nrg locus. Distortion in this region would thus involve a DNA fragment from this interval.

View larger version (10K): In this window In a new window Download PPT slide   Figure 2. Selecting a genetic marker for the sex-ratio phenotype: 37 [sn+ lz] recombinant X chromosomes obtained from a cross between an sn-lz line and the XSR6 sex-ratio chromosome are ranked according to a series of diallelic molecular markers; gene locations and intergenic distances are those from the standard map of D. melanogaster. Open boxes represent the part of the recombinant chromosomes that comes from the parental sex-ratio chromosome.

Selecting a marker for population studies:
A random sample of 41 X chromosomes from Réunion was surveyed for an association between the sex-ratio phenotype and SSCP variation for a 214-bp fragment (see Table 1) overlapping the second intron of the Nrg locus. In this sample, 21 X chromosomes were sex-ratio and 20 were standard (MONTCHAMP-MOREAU and CAZEMAJOR 2002 ). Results are shown in Table 2. We found six alleles, including a major allele (slow) at a frequency of 0.83. Heterozygosity among distorter X chromosomes was low (H = 0.09), all but one carrying the slow allele. Heterozygosity among nondistorter chromosomes was higher (H = 0.51). The slow allele thus showed a strong association with the sex-ratio trait. Statistical association, D' = 0.72 [calculated after LEWONTIN's (1964) coefficient of linkage disequilibrium], was significant (P = 0.04, Fisher's exact test after pooling all nonmajor alleles). The substantial difference in heterozygosity between the two allelic classes suggests that a selective sweep event associated with the spread of a distorter has affected polymorphism at the Nrg locus. This gene thus appears to be an appropriate marker of the history of sex-ratio in natural populations.

Nucleotide polymorphism at Nrg in natural populations:
Samples of 15 X chromosomes were taken at random from each of the Madagascar and Réunion populations. Among these, 7 chromosomes from Réunion and 10 chromosomes from Madagascar were sex-ratio.

Sequence polymorphism was recorded for an 802-bp fragment of the Nrg locus overlapping two exons and three introns. Pooled data for the two populations are summarized in Table 3 and Fig 3. After discarding positions overlapping an insertion/deletion, we obtained a 774-bp aligned sequence. Of 33 nucleotide polymorphisms, 26 occurred in introns (including a three-state polymorphism), and 7 in exons (all synonymous).

View larger version (31K): In this window In a new window Download PPT slide   Figure 3. Nucleotide polymorphism in a 774-bp fragment at Nrg in 30 X chromosomes from Madagascar and Réunion. Female percentage of the progeny of males was estimated in an ST8 background (from MONTCHAMP-MOREAU and CAZEMAJOR 2002 ); sex-ratio chromosomes are in boldface type.

The most striking feature of the data set was that 22 of the 30 chromosomes were strictly identical. They included the 10 sex-ratio chromosomes from Madagascar and 6 of the 7 sex-ratio chromosomes from Réunion. The seventh chromosome was that previously found to cause heterogeneity in the sex-ratio class using SSCP. It differed from the other 6 by a single nucleotide substitution. This suggests that this chromosome derives from the other sex-ratio chromosomes by a recent mutation or by recombination since this change (A G) matches a polymorphism that is also present in the rest of the sample. Below we will consider it as belonging to the same haplotype family as the others, that is, to the "major haplotype." Five standard chromosomes, all from Réunion, also belonged to the major haplotype. All other haplotypes were very different from each other. The association between the major haplotype and sex-ratio was complete and highly significant in Madagascar (D' = 1, P = 3.3 x 10^–4, Fisher's exact test). Considered in addition to SSCP data, the major haplotype appeared to be in association with sex-ratio in the two populations. A neighbor-joining tree of Nrg sequences is shown in Fig 4.

View larger version (10K): In this window In a new window Download PPT slide   Figure 4. Neighbor-joining tree of Nrg sequences for Madagascar and Réunion X chromosomes, using D. melanogaster as an outgroup. M, Madagascar; R, Réunion. , sex-ratio X chromosomes.

Neutrality tests:
Neutrality tests are shown in Table 4. When the sex-ratio and standard X chromosomes were considered together, the HKA test was nonsignificant for both samples (data not shown). However, this test should be significant only when a drastic reduction in the number of polymorphic sites has occurred. This was not the case in our samples, where most standard chromosomes were very polymorphic. Under the conservative assumption of no recombination, Tajima's test was significant only for Réunion; Fu and Li's test was nonsignificant for both samples. However, when the conservative recombination rate (r = 6.63 x 10^–9/site/generation) was used, Tajima's test became significant for both samples. The haplotype test (H-test; DEPAULIS and VEUILLE 1998 ) was always significant when no recombination was assumed. Significance was increased when recombination was assumed. This result is consistent with evidence from data inspection that one haplotype (the major haplotype) is at a high frequency, most likely due to a selective sweep. When only standard chromosomes were considered, all tests were nonsignificant with two exceptions when recombination was considered: Fu and Li's test became highly significant for Madagascar, probably because of the excess of singletons in this population; and the H-test for Réunion became significant, probably because of the presence of the major haplotype in six of eight chromosomes.

Comparison with polymorphism at vermilion:
This was the first sequence variation study for D. simulans in Madagascar and Réunion. We thus had to confirm that the observed deviation from the neutral equilibrium model at Nrg resulted from selection at a single locus and not from a genome-wide deviation in this population. A survey of several D. simulans populations by HAMBLIN and VEUILLE 1999 showed that molecular variation at the vermilion gene (v) does not depart from a neutral equilibrium in Africa. The vermilion gene thus appeared to provide a reliable neutral reference marker. We aligned 697 bp of the vermilion sequence, including a 165-bp intron, in 15 chromosomes from each of the two populations. Sequence alignments are shown in Fig 5, and summary statistics are shown in Table 5. The nucleotide variation in the sample from Madagascar was higher than that in the sample from Réunion, and both values were at least twice as high as those found at Nrg for the same populations. Each sequence in Madagascar belonged to a different haplotype. There was no significant deficit of haplotypes using the H-test. Indeed, there was an excess of haplotypes, even when assuming a recombination rate of r = 3.60 x 10^–8 (P value of bidirectional H-test <10^–4). This was due to an excess of singletons. In agreement with this, Tajima's D and Fu and Li's D were negative and became significant in simulations using recombination. In Réunion, the 15 sequences contained nine haplotypes. However, an H-test using recombination indicated a significant deficit of haplotypes. Tajima's D and Fu and Li's D were close to zero and were never significant.

View larger version (48K): In this window In a new window Download PPT slide   Figure 5. Nucleotide polymorphism at v (vermilion gene) in a 697-bp fragment from 30 X chromosomes of Madagascar (M) and Réunion (R).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identifying a selective sweep:
This study involved three steps to characterize genes causing the sex-ratio phenotype in D. simulans and to study their dynamics in natural populations.

In a first step, using a distorting X chromosome from the Seychelles, we showed the presence of a sex-ratio factor in subdivision 7F, between otu and otd. In D. melanogaster the otu-otd region overlaps a 170-kb fragment, including 15 annotated genes. The organization of this genomic region in D. simulans should not be very different, since the X's of the two species are homosequential (LEMEUNIER and ASHBURNER 1976 ).

In a second step, we focused on the otu-otd candidate region using as a marker the Nrg gene, which is in the middle of this interval. A sex-ratio factor would be within 100 kb from Nrg, assuming a monofactorial genetic determinism. Using SSCP, we found a significant association between sex-ratio and Nrg in a population sample from Réunion. Moreover, there was a striking contrast in heterozygosity between the sex-ratio X chromosomes and the standard ones, due to the presence of a major allele that was especially predominant among distorting chromosomes. Together, the results suggested that the distorting region found for the Seychelles was also active in Réunion and that the same gene was involved.

In a third step, we recorded sequence variation at Nrg in Madagascar and Réunion and found that a major allele was shared by the two populations. An examination of variation at the sequence level showed that sex-ratio chromosomes constitute a homogeneous class of haplotypes (the major haplotype), since heterogeneity in this class involved only a singleton in one chromosome, which could be due to either a mutation or recombination. Overall, the statistical association, along with the significance of the haplotype test, confirmed that a selective sweep related to the spread of sex-ratio chromosomes has affected Nrg sequence variation. The homogeneity of the sex-ratio class strongly suggests that the same allele is responsible for the trait in both populations, but we cannot say at this stage whether the selective sweep occurred once in an ancestral population or in each population separately. It also indicates that the Nrg locus maps very close to the selective sweep factor on the chromosome or that the sweep events are recent. However, the fact that the major haplotype was also found among standard chromosomes shows that the fragment sequenced at Nrg is not itself the distorting factor, assuming a monofactorial genetic determinism.

Alternative explanations:
The above interpretations are the most probable ones. We must, however, consider alternative explanations.

Population bottleneck: A population bottleneck could have caused a shift in haplotype distribution. HAMBLIN and VEUILLE 1999 studied sequence variation at the vermilion gene in D. simulans and found that non-African populations showed a decrease in both nucleotide and haplotype diversity compared to African ones. Here we found that v in Madagascar, as in other East African populations (Kenya, Tanzania, and Mayotte), shows no evidence of a bottleneck in D. simulans (range of the other African populations for the same 697-bp fragment, = 0.0119–0.0148, _W = 0.0143–0.0207, calculated from HAMBLIN and VEUILLE 1999 data). However, genetic variation at the v locus is lower in Réunion than in other African populations and close to that of derived populations from Europe and the Antilles (range = 0.0098–0.0106, _W = 0.0095–0.0105, calculated from HAMBLIN and VEUILLE 1999 data), suggesting the occurrence of a recent bottleneck or a partial selective sweep. The v and Nrg loci are located at some distance from each other on the X chromosome (at cytological positions 9F10-11 and 7F3-4, respectively) and are separated by the substantial genetic distance of 10 cM. In Madagascar, genetic variation at these two loci is in opposite directions, as v shows an excess of low-frequency variants, whereas Nrg shows a deficit of low-frequency variants. This confirms the selective sweep for Nrg in Madagascar. The conclusion is less firm for Réunion. We have independent evidence from other loci (E. BAUDRY, N. DEROME, M. HUET and M. VEUILLE, unpublished data) that the Madagascar population may be expanding and that the Réunion population may be the result of a recent immigration. However, the fact that the same Nrg haplotype is dominant in both populations and is associated with sex-ratio indicates that the selective sweeps observed at this locus in Madagascar and Réunion have the same cause.

Low local recombination rate: Segregation distorters are often associated with a low recombination rate due to a chromosomal inversion or a pericentromeric location, which may result in reduced polymorphism and statistical association over extended regions (HAMMER and SILVER 1993 ; BABCOCK and ANDERSON 1996 ; PALOPOLI and WU 1996 ). This cannot be the case for sex-ratio in D. simulans. No inversion was found on polytene chromosomes in the 7F region in both sex-ratio/standard heterozygous females and D. melanogaster/D. simulans hybrid females (F. LEMEUNIER, personal communication). The contrary would have been unexpected, since inversions are virtually absent from natural populations of D. simulans (ASHBURNER and LEMEUNIER 1975 ). The possibility of a minute inversion can also be ruled out, since the genetic distance observed between sn and lz in our genetic experiments was slightly higher for D. simulans (7.7 cM) than for D. melanogaster (6.7 cM), while the order of genes was apparently conserved.

One or two sweeps in Réunion?
There was clearly a selective sweep, and this sweep was statistically associated with a sex-ratio selective sweep. This is encountered in the two surveyed populations. However, the SSCP study in Réunion showed that 20/21 sex-ratio vs. 14/20 standard chromosomes belonged to the major haplotype. A noteworthy fact was the deficit of minor haplotypes among sex-ratio chromosomes [P(²) < 10^–4, 1 d.f.]. Following the selective sweep event, we would have expected some level of recombination to have occurred between Nrg and the distorter gene and to have imported the Nrg major haplotype into nondistorting chromosomes and, conversely, to have imported minor haplotypes into sex-ratio chromosomes. A first explanation is that the sex-ratio chromosome in Réunion has been close to fixation in the past and that the proportion of the major haplotypes moved to a standard chromosome through recombination was lower than that of other haplotypes moved to sex-ratio chromosomes. An alternative explanation is that a nearby selective sweep, independent of sex-ratio, occurred in Réunion and further increased the frequency of the major haplotype. We currently have no means to test these hypotheses. Note, however, that they do not contradict the main conclusion of a selective sweep linked to sex-ratio.

Age of the sweep:
The age of the selective sweep can be estimated from the fact that the major haplotype class is homogeneous or nearly so, allowing us to identify which chromosomes in the sample are descended from the selected haplotype. Assuming a star phylogeny of sex-ratio chromosomes, the expected number of mutations having occurred in this haplotype family after the selective sweep is E(S) = Lnt(/3N), where L is the length of the sequence alignment, n is the sample size of sex-ratio chromosomes, t is the age of the selective sweep in generations, q = 3N_eµ is the neutral mutation parameter (for an X-linked locus), N_e is the effective population size, and µ is the neutral mutation rate. This estimation is valid only for Madagascar, since an additional selective sweep is assumed to have affected the star phylogeny in Réunion. The expectation of having no mutation is e^–S. Assuming a 0.05 probability of obtaining this result yields the upper estimate t_0.05 < 0.181 N_e. It is often assumed that the effective population size of D. simulans is in the 10⁶ range and that there are 10 generations a year in the tropics. In this case, the age of the selective sweep would be <18,000 years. This upper limit is probably greatly overestimated, since the selective advantage of the distorting variant can be especially high as explained below.

Dynamics of the sex-ratio event:
Given the complete association between the major haplotype and the sex-ratio phenotype, the sex-ratio selective sweeps in Madagascar and Réunion must have been strong. The distorting X chromosomes were found to produce 80–96% female progeny in a background free of drive suppressors (MONTCHAMP-MOREAU and CAZEMAJOR 2002 ; ATLAN et al. 2003 ). The selective advantage of a standard chromosome provided by this segregation bias is so strong (s = 0.30–0.46) that, assuming an initial frequency of 0.001, sex-ratio chromosomes would be virtually fixed in 100 generations. It is, however, unlikely that these populations were sampled during an ongoing sex-ratio selective sweep, since segregation distortion was at the time inhibited by powerful suppressors. Given the low level of distortion expression (ATLAN et al. 2003 ), sex-ratio chromosomes could invade the populations only very slowly. Moreover, repeated sampling in Réunion during the last 10 years suggests that sex-ratio chromosome frequency has stabilized at an intermediate level of 50% (JUTIER et al. 2004 ). We can, therefore, suppose that a strong event occurred at some time in the past on the two islands.

Drive suppressors have been found to occur together with sex-ratio chromosomes over a wide range of D. simulans populations, including East Africa, which is the likely source of emigrants to Indian Ocean islands. A first scenario is that migrants from the continent invaded sex-ratio-free island populations, importing both the segregation distorter and its antidote. At first, the two elements were at a low frequency in the invaded gene pool, with a low probability of occurring in the same male, allowing the distorter to benefit from its segregational advantage. An increased frequency in the suppressor secondarily occurred as a consequence of the increased frequency of sex-ratio chromosomes. Madagascar and Réunion may also have undergone secondary sweeps related to an arms race between suppressors and modifiers. Alternatively, a sex-ratio sweep may have occurred in a population ancestral to Madagascar and Reunion. However, available data do not favor this hypothesis, except in the scenario where the Réunion population is a recent invader established from Madagascar migrants.

It must be pointed out that unsuppressed sex-ratio chromosomes in D. simulans can cause detrimental effects at the individual level that exceed their segregation advantage at the gamete level, as evidenced by their rapid loss from experimental populations (CAPILLON and ATLAN 1999 ). Owing to the developmental failure of Y-bearing sperm (MONTCHAMP-MOREAU and JOLY 1997 ; CAZEMAJOR et al. 2000 ), sex-ratio males can suffer more than a twofold fertility loss relative to standard males through sperm competition in multiple mating situations (CAPILLON 2000 ). This suggests that a rapid spread of sex-ratio chromosomes is possible only when males have little opportunity to mate, that is, when population density is low. While the evolution of drive suppressors and the deleterious effects on male fertility are able to stop the spread of distorters in D. simulans, whether or not they can hold them in a balanced polymorphism remains to be assessed. Several models show that this is a theoretical possibility (JAENIKE 1996 ; CARVALHO et al. 1997 ; JAENIKE 1999 ; TAYLOR and JAENIKE 2002 ). However, the Madagascar and Réunion data do not show evidence of balancing selection.

Our data are consistent with a recent spread of distorters across D. simulans populations. Together with well-documented cases of meiotic drive, including Segregation Distorter in D. melanogaster, Sex-ratio in D. pseudoobscura, and the t-haplotype in mice (HAMMER and SILVER 1993 ; BABCOCK and ANDERSON 1996 ; PALOPOLI and WU 1996 ; KOVACEVIC and SCHAEFFER 2000 ), they suggest that nonequilibrium dynamics dominates the evolutionary history of meiotic drive systems. A useful property of the distorting system used in our study is that it is contained in a short fragment of the genome. Formerly studied classical cases involved large chromosome regions held in linkage disequilibrium through inversion systems or recombination inhibition by pericentromeric heterochromatin.

Since the revival of MAYNARD-SMITH and HAIGH's (1974) hitchhiking theory by KAPLAN et al. 1989 very few, if any, of the selective sweeps that have been characterized at the molecular level over small regions have, to our knowledge, been so close to a causal explanation of the selective force involved in the process. It is often assumed that all selective sweeps are the result of Darwinian selection. Interestingly, this case involves a selective advantage for the gamete, with no evidence of a selective advantage for the zygote. In fact, studies by CAPILLON (2000) rather suggest that local conditions determine whether or not sex-ratio chromosomes are at an advantage over standard chromosomes. Note also that segregation distorters are "selfish genes" (ORGEL and CRICK 1980 ) that are potentially detrimental to the survival of populations. This suggests that selective sweep data should be critically examined before they are used to estimate the level of Darwinian selection in natural populations.

	FOOTNOTES

Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under accession nos. AY321889, AY3211918 and AF486139, AF486140, AF486141, AF486142, AF486143, AF486144, AF486145, AF486146, AF486147, AF486148, AF486149, AF486150, AF486151, AF486152, AF486153, AF486154, AF486155, AF486156, AF486157, AF486158, AF486159, AF486160, AF486161, AF486162, AF486163, AF486164, AF486165, AF486166, AF486167, AF486168.
¹ These authors contributed equally to this work.

	ACKNOWLEDGMENTS

We thank Françoise Lemeunier for cytogenetic analysis and Matthew Cobb, Karen McCoy, and two anonymous referees for helpful comments. This work was supported by UPR9034, UMR 7625, Groupe de Recherche 1928 "Evolution des génomes dans les populations" of the Centre National de la Recherche Scientifique and Plan Pluri-Formations "Populations fractionnées et insulaires" of the Ecole Pratique des Hautes Etudes.

Manuscript received March 14, 2003; Accepted for publication December 18, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AGUADÉ, M., N. MIYASHITA, and C. F. LANGLEY, 1989  Reduced variation in the yellow-achaete-scute region in natural populations of Drosophila melanogaster.. Genetics 122:607-615.[Abstract/Free Full Text]

AQUADRO, C. F., D. J. BEGUN and E.C. KINDHAL, 1994 Selection, recombination, and DNA polymorphism in Drosophila, pp. 46–56 in Non-neutral Evolution. Theories and Molecular Data, edited by B. GOLDING. Chapman & Hall, New York.

ASHBURNER, M. and F. LEMEUNIER, 1975  Relationships within the melanogaster species subgroup of the genus Drosophila (Sophophora). I. Inversion polymorphisms in Drosophila melanogaster and Drosophila simulans. Proc. R. Soc. Lond. Ser. B 193:137-157.[Medline]

ATLAN, A., H. MERÇOT, C. LANDRÉ, and C. MONTCHAMP-MOREAU, 1997  The sex-ratio trait in Drosophila simulans: geographical distribution of distortion and resistance. Evolution 51:1886-1895.[CrossRef]

ATLAN, A., C. CAPILLON, N. DEROME, D. COUVET, and C. MONTCHAMP-MOREAU, 2003  The evolution of autosomal suppressors of sex-ratio drive in Drosophila simulans.. Genetica 117:47-58.[CrossRef][Medline]

BABCOCK, C. S. and W. W. ANDERSON, 1996  Molecular evolution of the Sex-Ratio inversion complex in Drosophila pseudoobscura: analysis of the Esterase-5 gene region. Mol. Biol. Evol. 13:297-308.[Abstract]

CAPILLON, C., 2000 Evolution du système sex-ratio chez Drosophila simulans. Ph.D. Thesis, Université Paris 6, Paris.

CAPILLON, C. and A. ATLAN, 1999  Evolution of driving X chromosomes and resistance factors in experimental populations of Drosophila simulans.. Evolution 53:506-517.[CrossRef]

CARVALHO, A. B. and S. C. VAZ, 1999  Are Drosophila SR drive chromosomes always balanced? Heredity 83:221-228.

CARVALHO, A. B., S. C. VAZ, and L. B. KLACZKO, 1997  Polymorphism for Y-linked suppressors of sex-ratio in two natural populations of Drosophila mediopunctata.. Genetics 146:891-902.[Abstract]

CAZEMAJOR, M., C. LANDRE, and C. MONTCHAMP-MOREAU, 1997  The sex-ratio trait in Drosophila simulans: genetic analysis of distortion and suppression. Genetics 147:635-642.[Abstract]

CAZEMAJOR, M., D. JOLY, and C. MONTCHAMP-MOREAU, 2000  Sex-ratio meiotic drive in Drosophila simulans is related to equational nondisjunction of the Y chromosome. Genetics 154:229-236.[Abstract/Free Full Text]

DEPAULIS, F. and M. VEUILLE, 1998  Neutrality tests based on the distribution of haplotypes under an infinite site model. Mol. Biol. Evol. 15:1788-1790.[Medline]

DEPAULIS, F., S. MOUSSET, and M. VEUILLE, 2001  Haplotype tests using coalescent simulations conditional on the number of segregating sites. Mol. Biol. Evol. 18:1136-1138.[Free Full Text]

DEPAULIS, F., S. MOUSSET, and M. VEUILLE, 2003  Powers of neutrality tests against bottleneck and hitchhiking. J. Mol. Evol. 57:s190-s200.

DEPAULIS, F., S. MOUSSET and M. VEUILLE, 2004 Detecting selective sweeps with haplotype tests, in Selective Sweep, edited by D. NURMINSKY. Landes Bioscience, Georgetown, TX (in press).

FISHER, R., 1930 The Genetical Theory of Natural Selection. Clarendon Press, Oxford.

FU, Y. X. and W. H. LI, 1993  Statistical tests of neutrality of mutations. Genetics 133:693-709.[Abstract]

GERSHENSON, S., 1928  A new sex ratio abnormality in Drosophila obscura.. Genetics 13:488-507.[Free Full Text]

HAMBLIN, M. T. and M. VEUILLE, 1999  Population structure among African and derived populations of D. simulans: evidence for ancient subdivision and recent admixture. Genetics 153:305-317.[Abstract/Free Full Text]

HAMILTON, W. D., 1967  Extraordinary sex ratios. Science 156:477-488.[Free Full Text]

HAMMER, M. F. and L. M. SILVER, 1993  Phylogenetic analysis of the alpha-globin pseudogene-4 (Hba-ps4) locus in the house mouse species complex reveals a stepwise evolution of t haplotypes. Mol. Biol. Evol. 10:971-1001.[Abstract]

HUDSON, R. R., M. KREITMAN, and M. AGUADÉ, 1987  A test of neutral molecular evolution based on nucleotide data. Genetics 116:153-159.[Abstract/Free Full Text]

JAENIKE, J., 1996  Sex-ratio meiotic drive in the Drosophila quinaria group. Am. Nat. 148:237-254.[CrossRef]

JAENIKE, J., 1999  Suppression of sex-ratio meiotic drive and the maintenance of Y-chromosome polymorphism in Drosophila. Evolution 53:164-174.[CrossRef]

JAENIKE, J., 2001  Sex chromosome meiotic drive. Annu. Rev. Ecol. Syst. 32:25-49.[CrossRef]

JUTIER, D., N. DEROME, and C. MONTCHAMP-MOREAU, 2004  The sex-ratio trait and its evolution in Drosophila simulans: a comparative approach. Genetica 120:87-89.

KAPLAN, N. L., R. R. HUDSON, and C. H. LANGLEY, 1989  The "hitchhiking effect" revisited. Genetics 123:887-899.[Abstract/Free Full Text]

KOVACEVIC, M. and S. W. SCHAEFFER, 2000  Molecular population genetics of X-linked genes in Drosophila pseudoobscura.. Genetics 156:155-172.[Abstract/Free Full Text]

KUMAR, S., K. TAMURA, and M. NEI, 1994  MEGA: molecular evolutionary genetics analysis software for microcomputers. Comput. Appl. Biosci. 10:189-191.[Abstract/Free Full Text]

LEMEUNIER, F. and M. A. ASHBURNER, 1976  Relationships within the melanogaster species subgroup of the genus Drosophila (Sophophora). II. Phylogenetic relationships between six species based upon polytene chromosome banding sequences. Proc. R. Soc. Lond. Ser. B 193:275-294.[Medline]

LEWONTIN, R. C., 1964  The interaction of selection and linkage. I. General considerations; heterotic models. Genetics 49:49-67.[Free Full Text]

MAYNARD-SMITH, J. and J. HAIGH, 1974  The hitch-hiking effect of a favorable gene. Genet. Res. 23:23-35.[Medline]

MONTCHAMP-MOREAU, C. and M. CAZEMAJOR, 2002  Sex-ratio drive in Drosophila simulans: variation in segregation ratio of X chromosomes from a natural population. Genetics 162:1221-1231.[Abstract/Free Full Text]

MONTCHAMP-MOREAU, C. and D. JOLY, 1997  Abnormal spermiogenesis is associated with the X-linked sex-ratio trait in Drosophila simulans.. Heredity 79:24-30.[CrossRef]

MONTCHAMP-MOREAU, C., V. GINHOUX, and A. ATLAN, 2001  The Y chromosomes of Drosophila simulans are highly polymorphic for their ability to suppress sex-ratio drive. Evolution 55:728-737.[CrossRef][Medline]

MORGAN, T. H., 1910  Sex limited inheritance in Drosophila. Science 32:120-122.[Free Full Text]

MOUSSET, S., L. BRAZIER, M.-L. CARIOU, F. CHARTOIS, and F. DEPAULIS et al., 2003  Evidence of a high rate of selective sweeps in African Drosophila melanogaster.. Genetics 163:599-609.[Abstract/Free Full Text]

ORGEL, L. E. and F. H. CRICK, 1980  Selfish DNA: the ultimate parasite. Nature 284:604-607.[CrossRef][Medline]

PALOPOLI, M. F. and C.-I WU, 1996  Rapid evolution of a coadapted gene complex: evidence from the segregation distorter (SD) system of meiotic drive in Drosophila melanogaster.. Genetics 143:1675-1688.[Abstract]

ROZAS, J. and R. ROZAS, 1999  DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis. Bioinformatics 15(2):174-175.[Abstract/Free Full Text]

STURTEVANT, A. H. and T. DOBZHANSKY, 1936  Geographical distribution and cytology "sex-ratio" in Drosophila pseudoobscura and related species. Genetics 21:473-490.[Free Full Text]

TAJIMA, F., 1989  Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123:585-595.[Abstract/Free Full Text]

TAYLOR, J. E. and J. JAENIKE, 2002  Sperm competition and the dynamics of X chromosome drive: stability and extinction. Genetics 160:164-174.

WALL, J. F. and R. H. HUDSON, 2001  Coalescent simulations and statistical tests of neutrality. Mol. Biol. Evol. 18:1134-1135.[Free Full Text]

This article has been cited by other articles:

				N. Derome, E. Baudry, D. Ogereau, M. Veuille, and C. Montchamp-Moreau Selective Sweeps in a 2-Locus Model for Sex-Ratio Meiotic Drive in Drosophila simulans Mol. Biol. Evol., February 1, 2008; 25(2): 409 - 416. [Abstract] [Full Text] [PDF]

				A. Sanchez-Gracia and J. Rozas Unusual Pattern of Nucleotide Sequence Variation at the OS-E and OS-F Genomic Regions of Drosophila simulans Genetics, April 1, 2007; 175(4): 1923 - 1935. [Abstract] [Full Text] [PDF]

				K. A. Dyer, B. Charlesworth, and J. Jaenike Chromosome-wide linkage disequilibrium as a consequence of meiotic drive PNAS, January 30, 2007; 104(5): 1587 - 1592. [Abstract] [Full Text] [PDF]

				C. Montchamp-Moreau, D. Ogereau, N. Chaminade, A. Colard, and S. Aulard Organization of the sex-ratio Meiotic Drive Region in Drosophila simulans Genetics, November 1, 2006; 174(3): 1365 - 1371. [Abstract] [Full Text] [PDF]

				L.-M. Chevin and F. Hospital The Hitchhiking Effect of an Autosomal Meiotic Drive Gene Genetics, July 1, 2006; 173(3): 1829 - 1832. [Abstract] [Full Text] [PDF]

				E. Baudry, N. Derome, M. Huet, and M. Veuille Contrasted Polymorphism Patterns in a Large Sample of Populations From the Evolutionary Genetics Model Drosophila simulans Genetics, June 1, 2006; 173(2): 759 - 767. [Abstract] [Full Text] [PDF]

				G. S. Wilkinson, E. G. Amitin, and P. M. Johns Sex-linked Correlated Responses in Female Reproductive Traits to Selection on Male Eye Span in Stalk-eyed Flies Integr. Comp. Biol., June 1, 2005; 45(3): 500 - 510. [Abstract] [Full Text] [PDF]