Several evolutionary models of linked selection (e.g., genetic hitchhiking, background selection, and random environment) predict a reduction in polymorphism relative to divergence in genomic regions where the rate of crossing over per physical distance is restricted. We tested this prediction near the telomere of the Drosophila melanogaster and D. simulans X chromosome at two loci, erect wing (ewg) and suppressor of sable [su(s)]. Consistent with this prediction, polymorphism is reduced at both loci, while divergence is normal. The reduction is greater at ewg, the more distal of the two regions. Two models can be discriminated by comparing the observed site frequency spectra with those predicted by the models. The hitchhiking model predicts a skew toward rare variants in a sample, while the spectra under the background-selection model are similar to those of the neutral model of molecular evolution. Statistical tests of the fit to the predictions of these models require many sampled alleles and segregating sites. Thus we used SSCP and stratified DNA sequencing to cover a large number of randomly sampled alleles (50) from each of three populations. The result is a clear trend toward negative values of Tajima's D, indicating an excess of rare variants at ewg, the more distal of the two loci. One fixed difference among the populations and high F_ST values indicate strong population subdivision among the three populations at ewg. These results indicate genetic hitchhiking at ewg, in particular, geographically localized hitchhiking events within Africa. The reduction of polymorphism at su(s) combined with the excess of high-frequency variants in D. simulans is inconsistent with the hitchhiking and background-selection models.

---

SEVERAL evolutionary models of linked selection have been proposed to explain the patterns of DNA sequence variation observed in natural populations. Genetic hitchhiking is a model of strong directional selection in which the fixation of favorable variants removes linked neutral variation (MAYNARD SMITH and HAIGH 1974). This hitchhiking effect is expected to be strongest in genomic regions where crossing over is restricted per physical distance (KAPLAN et al. 1989). The background-selection model also predicts a reduction in polymorphism that is due to what essentially amounts to a decrease in effective population size, caused by selection's removal of linked deleterious mutants (CHARLESWORTH et al. 1993). Neither model predicts a reduction in interspecific divergence. A chief difference between the models is whether a skew toward rare polymorphisms is expected; the hitchhiking model predicts such a skew (AGUADé et al. 1989; BRAVERMAN et al. 1995), while such a skew is not expected in a practically sized sample of sequences under background selection (HUDSON and KAPLAN 1994; CHARLESWORTH et al. 1995). The pseudohitchhiking model also yields reduced polymorphism and a skew in the frequency spectrum in regions of restricted recombination (GILLESPIE 2000). Finally, random-environment models involving linked selection can also produce reduction in polymorphism (relative to divergence) and a skew toward rare variants (GILLESPIE 1997). All these models of linked selection predict that the effect(s) on selectively neutral polymorphism will be most apparent in regions of the lowest crossing over.

The distal tip of the X chromosome of Drosophila melanogaster (and its close relatives) offers an excellent opportunity to test models of linked selection, since the rate of crossing over per physical distance decreases to zero at the gene-rich euchromatic region at the telomere. For example, AGUADé et al. (1989) found a reduction in polymorphism using RFLP in the yellow-achaete-scute (y-ac-sc) region of D. melanogaster. BEGUN and AQUADRO (1991) and MARTíN-CAMPOS et al. (1992) studied y-ac using greater sample sizes from additional geographic locations and extended the investigation to the sister species D. simulans. All three studies found a reduction in polymorphism in both species and an excess of rare variants. When the site frequency spectrum was quantified with Tajima's D (TAJIMA 1989), observed values were negative, indicating a skew toward rare variants, although not always significantly so. Divergence data from D. melanogaster and D. simulans permitted a test of the neutral prediction that polymorphism and divergence are correlated; the levels of divergence observed were normal, thus ruling out a reduction in the neutral mutation rate or the exclusive action of genetic drift as an explanation for the data from these regions. Hence genetic hitchhiking appeared to explain the data from these studies.

Additional work on the X telomere extended the surveys to samples from Africa (BEGUN and AQUADRO 1993, 1995b). Polymorphism was reduced in the telomeric genes y and ac. The levels of polymorphism were higher in Africa than on other continents, and population subdivision between African and non-African populations was detected. These results supported the theory that D. melanogaster originated in sub-Saharan Africa and migrated to Europe and North America (DAVID and CAPY 1988; LACHAISE et al. 1988). D. simulans is thought to have a similar history. Thus demographic phenomena and/or local adaptation affect genetic variation in D. melanogaster, not unlike what was already known in D. ananassae (STEPHAN and MITCHELL 1992). Yet sample sizes were generally limited and Tajima's D was not statistically different from zero, raising questions about statistical power and the applicability of the hitchhiking model.

More recent surveys of genes near the X chromosome's telomere consider regions with intermediate levels of crossing over and larger samples sizes. The studies of AGUADé et al. (1994) and LANGLEY et al. (2000) investigated two loci, suppressor of sable [su(s)] and suppressor of white apricot [su(w^a)], which are (centromere) proximal to y-ac-sc. Crossing over is still reduced at these loci, but less so than at y-ac-sc. These authors found that the hitchhiking model could explain their data, according to the reduction in polymorphism, and a general trend of the skew in the site frequency spectrum toward rare variants, but again Tajima's D was not always significantly negative. In the North American sample, D was large and positive. Simulation analysis of the data found a better fit between that data and the hitchhiking model than between that data and the background-selection model, but neither model fit well. Further work is needed to examine these questions in a genomic region with even lower recombination using the same or similar samples. In such regions of extremely low crossing over, the impacts of both the hitchhiking and the background-selection models should be greater. The expected further reduction in polymorphism also means fewer segregating sites per base pair with which to evaluate the frequency spectrum, which thus requires greater survey effort.

One of the goals of the present study is to increase the statistical power of the tests for neutrality, such as Tajima's D, by using large sample sizes. We surveyed 50 lines per population to find additional variation, especially rare variations. An additional reason for our generous sample sizes is to make informative comparisons among different Drosophila populations. We sampled from three continents, Africa, Europe, and North America.

Another goal of this article is to use interspecific divergence to gain insight into the evolutionary forces at work. Thus we surveyed both D. melanogaster and its sister species D. simulans. A normal level of divergence, for example, would rule out a low local neutral mutation rate and/or mutagenic recombination in regions of normal crossing over. In addition, we can test the generality of the phenomena by comparing data from the same genes experiencing similar but not identical genetic and population conditions in more than one species. Although the rate of crossing over per physical distance is restricted at the telomere of both species, crossing over in D. simulans is thought to increase faster when moving away from the tip (STURTEVANT et al. 1929). Also, the effective population size may differ between these two species. The greater heterozygosity, greater codon bias, and fewer nonsynonymous polymorphisms observed in D. simulans has been interpreted as evidence that D. simulans has a larger population size than D. melanogaster (AQUADRO 1992; MORIYAMA and POWELL 1996; IRVIN et al. 1998).

We surveyed two genes located near the telomeres of D. melanogaster and D. simulans. The gene erect wing (ewg) codes for a transcription factor and is located at polytene chromosome band position 1A1 (KOUSHIKA et al. 2000; DRYSDALE et al. 2005), distal to yellow. In this first region, excluding insertion-deletions (indels), we surveyed 3166 bp in D. melanogaster and 3193 bp in D. simulans. The gene su(s) encodes an RNA-binding protein and is located at position 1B13 (GEYER et al. 1991; VOELKER et al. 1991; DRYSDALE et al. 2005). In this second region, excluding indels, we surveyed 2832 bp in D. simulans. The two loci are separated by 360 kb. Our D. simulans su(s) data complement a previously published survey of the su(s) region in D. melanogaster (LANGLEY et al. 2000).

Our results can be summarized as follows. First, the ewg region has an extreme reduction in polymorphism and a negative Tajima's D in both D. melanogaster and D. simulans, which is consistent with the hitchhiking model. Second, the pattern of variation across populations of D. melanogaster could be the result of geographically localized hitchhiking events, similar to what has been found in D. ananassae (STEPHAN and MITCHELL 1992; STEPHAN et al. 1998; BAINES et al. 2004) and in other regions of the D. melanogaster X telomere (BEGUN and AQUADRO 1993). Third, variation at su(s) is reduced in D. simulans, but Tajima's D is positive; neither the hitchhiking model nor the background-selection model can explain results at that gene in that species.

Samples:

D. melanogaster flies were obtained from the following sites: North America (Raleigh, NC; same collection and extraction as for MIYASHITA et al. 1993), Europe [14 from the Canary Islands, Spain, 17 from Groningen, Holland, and 21 from Requena, Spain; same collection and extraction as MARTíN-CAMPOS et al. 1992 (see their Figure 1)], and Africa (collected in September 1990 in the Sengwa Wildlife Preserve, Zimbabwe; same collection as BEGUN and AQUADRO 1993). The following collections of D. simulans were studied: North America (25 collected in September 1995 from the Noble Apple Orchard, Paradise, CA, and 25 collected in July 1995 from the Wolfskill Orchard, Winters, CA, and extracted in 1995 in the laboratory of M. Aguadé using the attached-X strain kindly provided by J. Coyne); Europe (collected in 1993 in Montblanc, Spain, by M. Aguadé and extracted in her laboratory using the attached-X strain); and Africa (collected about 1993 in Harare, Zimbabwe, and extracted using the attached-X strain in the laboratory of C. H. Langley). We refer to these samples by their continent of origin.

View larger version (6K): In this window In a new window Download PPT slide   FIGURE 1.— The SSCP fragments of ewg, shown as small horizontal lines below the gene. Only part of the entire gene is shown, and it is oriented with the 3'-end on the left in contrast to the standard orientation to illustrate the fragment positions relative to the physical location of the su(s) fragments. The solid boxes are exons, the shaded boxes are alternatively spliced exons, and the thin lines connecting the solid boxes are the introns. The scale is indicated with a bar 200 nucleotides long.

 
The same samples were used for both the ewg and su(s) studies. The study of su(s) in D. melanogaster was reported by AGUADé et al. (1994) for North America and by LANGLEY et al. (2000) for Europe and Africa. Line numbers in the figures in those publications are the same as those in supplementary Tables S1–S9 at http://www.genetics.org/supplemental/. The following lines were not represented in all three studies. For the D. melanogaster sample from Africa, lines 51, 52, and 53 were present only in the ewg study. For the D. melanogaster sample from Europe, line 46 was absent from the su(s) study. For the D. melanogaster sample from North America, line 13 was absent from the ewg study while lines 51 and 52 were absent from the su(s) study. For the D. simulans study of Europe, line 10 was absent in the su(s) study. The sample sizes are presented in Table 1.

View this table: In this window In a new window   TABLE 1 Polymorphism statistics for ewg and su(s)

 

SSCP and sequencing:

The single-strand conformation polymorphism (SSCP) protocol of AGUADé et al. (1994) was used to bin sequence fragments (ranging in size from 136 to 345 bp) into allelic classes. The protocol of AGUADé et al. (1994) was modified in that the fragments were labeled with ³³P instead of being silver stained. The locations of the fragments are depicted in Figures 1 and 2. Representative alleles of each SSCP class were sequenced to identify underlying nucleotide polymorphisms. DNA sequencing was carried out on an ABI 377 automated sequencer using standard protocols.

View larger version (5K): In this window In a new window Download PPT slide   FIGURE 2.— The SSCP fragments of su(s). These are similar but not identical to fragments in LANGLEY et al. (2000). The open box is the 5'-UTR. See the Figure 1 legend for more information.

 

Data analysis:

We report , the average number of pairwise differences per nucleotide. When direct sequencing revealed polymorphism undetected by SSCP, the procedures of AGUADé et al. (1994) were followed to estimate *, the average number of pairwise differences per nucleotide, which incorporates an estimate of the amount of hidden variation. The 95% confidence intervals associated with and * were calculated by bootstrapping over alleles for 1000 replications. Calculations of the HKA test (HUDSON et al. 1987) and Tajima's D (TAJIMA 1989) assumed that sequences within SSCP classes were identical to the sequenced subsample. DnaSP 4.0 (ROZAS et al. 2003) was used for the HKA test, the calculation of F_ST and the permutation test (HUDSON et al. 1992a,b; HUDSON 2000), R_M (HUDSON and KAPLAN 1985), and the estimation of the number of silent sites (NEI and GOJOBORI 1986).

Gene regions:

We annotated our ewg data from both D. melanogaster and D. simulans according to GenBank entry no. AE003417, which was prepared as part of the D. melanogaster genome annotation release 3.1 (CELNIKER et al. 2002). The D. melanogaster ewg study included introns (2200 bp, excluding polymorphic indels) and exons (966 bp). Also excluding polymorphic indels, the D. simulans survey covered 2252 bp of noncoding DNA (introns) and 941 bp of exons. The total number of silent sites (noncoding + synonymous coding; NEI and GOJOBORI 1986) studied was 2587.16 in D. melanogaster and 2472.16 in D. simulans. For D. simulans su(s), we followed the GenBank entry no. M57889 (D. melanogaster) for our annotation of this gene, and accordingly 2832 noncoding bp were surveyed (excluding polymorphic gaps); this includes introns, a 5'-untranslated sequence, and a 5'-flanking sequence.

Computer simulations:

First, neutral coalescent simulations (HUDSON 1990) were used to estimate confidence intervals for Tajima's D. We also ran these simulations (10,000 iterations) to estimate exact P-values for the observed Tajima's D's. Second, the recurrent simulation method of BRAVERMAN et al. (1995) was used to assess the probability of obtaining the observed D values (D_o) or greater under a model of recurrent, strong directional selection at linked sites. That probability is labeled Prob {D D_o|H.H.}, where H.H. stands for hitchhiking. Next we followed the logic that background selection can be modeled by a neutral coalescent simulation in which the effective population size is appropriately reduced (CHARLESWORTH 1996; STEPHAN et al. 1998; LANGLEY et al. 2000). Those simulations were used to calculate either Prob {D D_o|N.T. and D_o > 0} or Prob {D < D_o|N.T. and D_o < 0}, where N.T. stands for neutral theory. These simulations are conditioned on the observed number of segregating sites, and thus population size is not a factor.

The hitchhiking simulations require calibration. A rate of hitchhiking was chosen to produce, on average, the observed reduction in from a value from a region of normal crossing over. It is important to choose a value matching the population source and the type of sequence (e.g., silent sites). For D. melanogaster, we set the level of normal variation to be 0.023 in Africa and 0.0081 in North America and Europe. These numbers were obtained from the DNA sequencing study of vermilion (BEGUN and AQUADRO 1995a).

For D. simulans, we set the normal level of polymorphism to be 0.0347 for Africa, 0.0279 for Europe, and 0.0288 for North America. These were calculated from vermilion from corresponding populations (BEGUN and AQUADRO 1995a; HAMBLIN and VEUILLE 1999). In some cases, their data were reanalyzed to obtain estimates of for silent sites.

Polymorphism:

The results of the SSCP and sequencing study of ewg and su(s) are presented in supplementary Tables S1–S9 at http://www.genetics.org/supplemental/. A total of 34 variable sites were found in the ewg region of D. melanogaster. Of these, 15 were indel polymorphisms of 1–34 bp long. The 17 variable sites found in D. simulans ewg include 5 indels, each 1 bp long. The D. simulans su(s) region was found to have 19 variable sites, including 7 insertion-deletion polymorphisms ranging from 1 to 8 bp.

Hierarchical DNA sequencing of a subset of the SSCP fragments identified the variants; the results are presented in part (b) of supplementary Tables S1–S9 at http://www.genetics.org/supplemental/. In a few cases, sequencing identified variants within SSCP classes. For D. melanogaster ewg, sequencing found two single-nucleotide polymorphisms in one fragment not detected by SSCP (in the exon of fragment 6 in the sample from North America). D. simulans ewg had eight instances (three different single-nucleotide variants among three fragments, only within Africa) of hidden variation. In D. simulans su(s), there were five cases of the same hidden variant in fragment 9 in the African and North American samples.

A statistical analysis of polymorphism found by the survey of ewg and su(s) is located in Table 1. The two regions have different levels of polymorphism, with the values of and * for ewg consistently lower than those for su(s). The D. melanogaster ewg variation in the African sample, for example, was less than one-sixth that for su(s) (LANGLEY et al. 2000). According to coalescent simulations, the probability of obtaining the observed number of segregating sites in the African ewg under the neutral model and no intralocus recombination, assuming it has the same value of 3Nµ as for su(s), is <0.001 (HUDSON 1990). In D. simulans, variation at ewg is less than half that at su(s). While values differ enough that the bootstrap 95% confidence intervals do not overlap in this comparison, the ewg and su(s) regions do not have significantly different estimates of 3Nµ according to neutral simulations. The other populations also were compared with simulations but power was too low to reveal differences.

Comparing across species, two different trends emerge (Table 1). Within Africa, the level of polymorphism ( and *) is higher in D. simulans than in D. melanogaster at both ewg (0.00079 vs. 0.00035) and su(s) (0.00219 vs. 0.00182), although the confidence intervals overlap for the su(s) comparison. The same trend presents itself for su(s) of Europe and North America. The opposite trend appears in ewg of Europe and North America. A simulation analysis was conducted to test for a difference in the levels of genetic hitchhiking in the two species (see Simulation analysis below), but none was detected.

Compared to other X-linked loci from regions with normal levels of crossing over, ewg and su(s) have less variation. For this comparison, LANGLEY et al. (2000) used averages of values from the white and vermilion regions, studied in the same populations with RFLP (MIYASHITA and LANGLEY 1988; BEGUN and AQUADRO 1993). The averages for Africa and North America are 0.007 and 0.004, respectively. These numbers are well above all the values observed in this study. For example, the African white-vermilion average is 21 times greater than the African ewg average .

More recent data from DNA sequencing studies are available for such comparisons against genes from X-linked regions of normal levels of crossing over. The vermilion locus, for example, was studied using DNA sequencing in a number of populations and in both D. melanogaster and D. simulans (BEGUN and AQUADRO 1995a; HAMBLIN and VEUILLE 1999). Their data are an appropriate baseline for comparison because they did not reject the neutral model according to the HKA (HUDSON et al. 1987) and Tajima (TAJIMA 1989) tests for most of the cases. The study of D. simulans vermilion by HAMBLIN and VEUILLE (1999) focused on a region of the gene with the highest level of polymorphism. So its may not represent average levels in African and European populations. For comparison with our North American sample, we use the BEGUN and AQUADRO (1995a) vermilion data from North Carolina. All of these data were reanalyzed to give values of for silent sites (noncoding and synonymous sites combined).

Comparison of from vermilion with and * ewg and su(s) within D. melanogaster show remarkable reductions in variation. For example, the African sample (Table 1) exhibits a 65-fold reduction in polymorphism at ewg compared to vermilion (* = 0.00035 at ewg vs. = 0.0023 for the vermilion silent sites). D. melanogaster su(s) polymorphism is also reduced (e.g., >12-fold in the African sample; see LANGLEY et al. 2000).

In D. simulans, ewg also has much less variation than vermilion. We recalculated the statistics for silent sites using the data collected by HAMBLIN and VEUILLE (1999). The value of for vermilion from Africa (in Zimbabwe, but a different collection date), for example, is 0.035 for vermilion, but * is only 0.00079 for ewg (Table 1). The su(s) locus has a * of only 0.00219 in the African sample. Again, this is a major decrease in variation (>40-fold).

Divergence:

The interspecific divergences between D. melanogaster and D. simulans at ewg and su(s) for all sites studied by SSCP are 0.056 and 0.123, respectively (Table 2). When considering only silent sites, ewg divergence is 0.101. These are similar to the average value, 0.061, reported for noncoding regions by MORIYAMA and POWELL (1996). At vermilion, silent divergence is 0.185 (BEGUN and AQUADRO 1995a). The level of divergence at y-ac-sc ranged between 0.0695 and 0.0558, depending on the type of data (MARTíN-CAMPOS et al. 1992). The average of Jukes-Cantor divergences reported by BEGUN and WHITLEY (2000) for 21 X-linked loci in regions of normal crossing over is 0.112. Our divergence estimates for ewg and su(s) are comparable to these other values.

View this table: In this window In a new window   TABLE 2 Divergence at ewg and su(s)

 

Polymorphism and divergence:

We applied the HKA test (HUDSON et al. 1987) to test the null hypothesis that the level of polymorphism is proportional to divergence (data not shown). The ideal reference locus matches the sequence type (here, silent) and the population source. These criteria are met in vermilion (BEGUN and AQUADRO 1995a; HAMBLIN and VEUILLE 1999), except that a European sampling source was not available for vermilion from D. melanogaster, so that population sample was tested against North American data. The HKA using ewg and su(s) individually against vermilion was either highly significant (P < 0.01) or very highly significant (P < 0.001). The 5'-flanking region of Adh was also used (KREITMAN and HUDSON 1991), although the sample is a combination of 11 sequences from many global locales; the results were again always highly significant or very highly significant. Finally, we conducted the test comparing ewg and su(s), the two loci from this study. None of those tests was significant. These results (and the normal level of divergence) can be interpreted as strong evidence that the level of polymorphism is reduced at the ewg and su(s) loci. This reduction of polymorphism is not consistent with the neutral model of molecular evolution.

Frequency spectrum:

We used Tajima's D (TAJIMA 1989) to assess the deviations from a neutral expectation of the frequency spectrum of segregating sites. The results are presented in Table 1. The ewg region exhibits a number of significantly negative values of D, indicating a skew toward rare variants. For example, Africa ewg has a significant value (–1.67; P = 0.0228), even though a few variants in the African sample have intermediate frequencies (e.g., site 29,790; supplementary Table S1 at http://www.genetics.org/supplemental/). Meanwhile, for the same sampled chromosomes from Africa, su(s) and su(w^a) have negative but not significant values of Tajima's D: –1.28 and –1.04, respectively (LANGLEY et al. 2000).

Regarding the European D. melanogaster sample, the values of Tajima's D are also negative. They are significant in the case of Europe for both ewg and su(s). For the same collection, su(w^a) also exhibits a negative but not significant Tajima's D. The North American D's are negative for ewg and su(w^a) but not su(s).

Turning to the results for D. simulans, Tajima's D for the ewg region has negative but not significant values (Table 1) in the African sample for both single-nucleotide and indel variation. Only one single-nucleotide variant was found in North America. The lack of polymorphism in the European sample precluded this analysis. The su(s) region of D. simulans, in contrast, did not have negative values at all, except for the indel variation; the North American sample actually had a significant positive value (+1.96; P = 0.0258). Likewise, the North American D. melanogaster su(s) had a large positive value. The European D. simulans sample has a large positive but not quite significant D at su(s).

Simulation analysis:

Simulations are a useful method for distinguishing the hitchhiking and background-selection models. They can provide probabilities of observing particular data sets under each model, which can then be compared.

For D. melanogaster, the simulation results (Table 1) can be interpreted as follows. First, Tajima's D from the ewg African and European samples can be explained better by the hitchhiking model than by the background-selection model. This is evident in the negative and significant values of Tajima's D's observed (–1.67 and –1.86). In particular, the hitchhiking simulations showed relatively large P-values (0.2126 and 0.5177), while the background-selection (neutral) model is significantly inconsistent with the observed data (P = 0.0228 and 0.0093).

Second, the background-selection model seems to explain the value of Tajima's D (–0.47) observed in the D. melanogaster ewg North American sample better than the hitchhiking model (Table 1). The background-selection P-value is 0.3825 while the hitchhiking P-value is only 0.0537.

Third, for su(s) from D. melanogaster, we repeated the simulations presented in Figure 3 of LANGLEY et al. (2000) (Table 1). Again, the hitchhiking model explains the observed D (–1.54; P = 0.0377) in Europe better than the background-selection model does. However, because we used different data (see MATERIALS AND METHODS) to calibrate the hitchhiking model, the results are different in the case of su(s) for Africa. The new values of from vermilion are much larger than the values used by LANGLEY et al. (2000). Thus the rate of recurrent hitchhiking required to achieve the observed relative reduction in is larger, and the simulated values of Tajima's D are smaller. Therefore, the observed value of Tajima's D (–1.28), while negative, occurs less often in the hitchhiking simulation runs. However, P = 0.0583, so the hitchhiking model is still not rejected. Meanwhile, the background-selection model has P = 0.0823, which is also not a significant rejection. Thus, both the hitchhiking model and the background-selection model are marginally consistent with the data, although neither produces a very good fit.

Fourth, the value of D from the D. melanogaster su(s) from Europe (–1.54), as suggested by LANGLEY et al. (2000), is explained better by hitchhiking, even with the new parameters (Table 1). The hitchhiking P-value is 0.2340, while background selection is significantly rejected by the data (P = 0.0377).

Fifth, the value of Tajima's D observed in the D. melanogaster su(s) sample from North America is explained better by the background-selection model. Because this sample has a large positive value of Tajima's D, we estimated the Prob {D D_o|N.T.} instead of Prob {D < D_o|N.T.} (Table 1). The results indicate where the observed value falls in the upper half of the simulated distribution under the neutral or background-selection models. The simulations show that this value could be accounted for by the background-selection model, but it is not very likely (P = 0.0812). Just as the background-selection model is not likely to produce strongly negative values of Tajima's D, it is not likely to produce large positive values. This positive D is also inconsistent with the hitchhiking model (P = 0.0012).

Turning to the D. simulans results, Tajima's D at ewg from Africa has a negative value (–1.23; Table 1), but neither the background-selection model nor the hitchhiking model is rejected under these simulations. The power to discriminate among models is reduced in this case due to the small number of segregating sites. Low polymorphism precludes these analyses entirely in North American and European D. simulans ewg samples.

The remaining three cases are from D. simulans su(s) (Table 1). All three had positive Tajima's Ds. The first case, su(s) from Africa, is explained better by the background-selection model (P = 0.8592). The final two cases had large positive values of Tajima's D. Their associated P-values, interpreted as Prob {D D_o} for both models, are very small. Consequently, neither the background-selection model nor the hitchhiking model is able to explain these cases very well.

Hitchhiking simulations were used to test for a difference in the rate of hitchhiking in the two species. In the case of the ewg sample from Africa, it appears that the rate of hitchhiking is greater in D. melanogaster than in D. simulans, since is smaller in the former species. We used the same rates of hitchhiking used above for D. melanogaster ewg with the D. simulans sample size and number of segregating sites and asked how often the observed reduction, or a smaller one, in the total size of the coalescent tree was obtained. The size of the coalescent tree is proportional to the amount of variation and an indicator of the strength of hitchhiking (BRAVERMAN et al. 1995). If, all other things being equal, the rate of hitchhiking were significantly greater in D. melanogaster, then the distribution of the relative reduction would be well beneath the observed relative reduction in D. simulans. The simulation results did not detect evidence of such a difference (P = 0.8586). The converse simulations (using D. simulans rate and D. melanogaster parameters) also did not detect a significant difference (P = 0.6516).

Population subdivision:

There is one fixed difference among the populations: a nonsynonymous change at site 28,218, fixed in the African D. melanogaster ewg sample as GCG (Ala) and as GGG (Gly) elsewhere (Table 3). There is one nearly fixed difference at site 27,501. All lines except one (no. 50, which has a T) in the African sample have an A. The non-African samples also have a T at this site. The D. simulans sequence at these two sites is the same as in the non-African populations, suggesting an African origin of the mutation subsequent to the species' colonization of the other locations.

View this table: In this window In a new window   TABLE 3 The haplotypes of D. melanogaster ewg for nonunique single-nucleotide polymorphism

 
Across species and genes, African populations stand out with the highest level of polymorphism (Table 1). The polymorphic sites in non-African D. melanogaster ewg populations are not a subset of those found in Africa. The only exception is one indel polymorphism, at which the rarer form is found only twice in the African sample, once in the European sample, and four times in the North American sample. Similarly, the polymorphisms at D. simulans ewg in Africa are not found in the non-African populations, as the latter have nearly no polymorphism.

For D. simulans su(s), the variation is evenly distributed across the three population samples. Of 14 nonunique segregating sites, 8 segregate in all three populations, many at high frequencies. Three are polymorphic only in the African sample. Three tightly linked indels segregate only in the European and North American samples. Thus the European and North American variation cannot be said to be a subset of the African variation.

To measure the level of differentiation among the three populations, we calculated F_ST according to HUDSON et al. (1992b) and applied the permutation test to various subdivision statistics (HUDSON et al. 1992a; HUDSON 2000). First, as a preliminary step, we calculated F_ST for ewg for comparisons of the three different locales from which the European D. melanogaster flies were collected; the values were very low and not statistically significantly different from zero subdivision. Thus we pooled these three groups. Second, we applied the same procedure to the two groups of D. simulans North American lines using data from su(s) (ewg had nearly no variation in these populations with which to detect any subdivision). The value was also not significant. This also justifies pooling these two locales.

Second, the estimates of F_ST for comparisons between Africa and North America and between Africa and Europe are reported in Table 4. All the comparisons exhibit statistically significant subdivision. The estimates for D. melanogaster range from 0.153 to 0.811. The subdivision at ewg is greater than at su(s). Subdivision is also present in D. simulans, with F_ST values ranging from 0.100 to 0.312 (Table 4). They are slightly higher at ewg than at su(s).

View this table: In this window In a new window   TABLE 4 Population structure at ewg and su(s)

 

Linkage disequilibrium:

We conducted Fisher's exact test on all pairs of polymorphisms present in at least two lines (i.e., excluding unique polymorphisms) to test for nonrandom associations. Each population was treated separately. Table 5 summarizes the percentage of formally significant (P < 0.05) linkage disequilibria among polymorphic sites. Table 5 also presents R_M, the inferred minimum number of recombination events in a sample (HUDSON and KAPLAN 1985), and average r², the squared correlation coefficient. In most cases, R_M > 0, evidence for occasional recombination in the history of these sampled alleles in both species. At the same time, the proportions of "significant" tests and the average r² indicate substantial linkage disequilibrium. Interlocus linkage disequilibrium estimates (average r²) in D. melanogaster in general are of the same order of magnitude as intralocus estimates, except in the North American sample, where both the ewg and su(s) intralocus values are higher than the interlocus values and, for su(s), in the European sample. In D. simulans, the sample from Africa was the only one with enough polymorphic sites in both genes for this analysis. The average r² is the greatest within su(s) and is an order of magnitude lower in ewg and between the two loci. Average r² is even higher within D. simulans su(s) from North America and Europe.

View this table: In this window In a new window   TABLE 5 Recombination and linkage disequilibrium

 

The data and analysis presented here consider the telomere-proximal region of low crossing over per physical length where the impact of linked selection is most apparent. The evidence for skewed frequency spectra at ewg in the African populations of both D. melanogaster and D. simulans points toward strong positive selection shaping neutral (and more mildly selected) variation at the tip of the X chromosome in these two species. The remainder of this DISCUSSION considers other forces that may have shaped our data from ewg and su(s). We also compare our data to previous publications.

Background selection:

Several lines of reasoning argue against the background-selection model as an explanation for the data at ewg and su(s). First, HUDSON and KAPLAN (1995) note that extremely high rates of deleterious mutation are required to obtain the large reductions observed at genes such as those at the telomere. Second, background selection cannot account for significant negative values of Tajima's D observed in practical sample sizes (HUDSON and KAPLAN 1994; CHARLESWORTH et al. 1995). Our data include several cases of significantly negative Tajima's D's. The case [D. simulans su(s) of North America] of Tajima's D that is large and significantly positive also does not fit the background-selection model. The large nonsignificant values of Tajima's D (D. simulans of Europe and D. melanogaster of North America) are not easily explained by the background-selection model according to our simulation analysis (Table 1). Third, KIM and STEPHAN (2000) compared the two models and found that in general the hitchhiking model better explains polymorphism in regions of very restricted crossing over.

Recombination:

Another issue raised by our results is the unexpected evidence for recombination in our sample, indicated by R_M > 0. It seems unlikely that crossing over is responsible for these nonzero values of R_M because observed crossing over is very low in this region. In addition, crossing over should reduce the hitchhiking effect, yet polymorphism is in fact low. Another process, gene conversion, could result in R_M > 0, which we observed in both genes and in both species (Table 5). Because the population genetic consequence of unbiased gene conversion is effectively short-range double exchange, its impact on linkage disequilibrium is qualitatively different from that of crossing over (ANDOLFATTO and NORDBORG 1998; FRISSE et al. 2001). For pairs of polymorphic sites less than a gene conversion-track length apart, gene conversion augments the decay of linkage disequilibrium with distance. In contrast, for pairs of polymorphic sites that are more widely separated, gene conversion reduces nonrandom association at a distance-independent rate. For example, LANGLEY et al. (2000) noted a lack of long-distance linkage disequilibria and the presence of short-distance disequilibria on the scale of gene conversion, and thus they interpreted the inferred recombination in their samples as gene conversion, not crossover, events.

Before considering any linkage disequilibrium in our data, it is important to note that not much power is available to discern patterns. Not only is there low variation, but also, when there is a skew toward rare variants, the number of nonsingleton sites available for LD analysis is even fewer. Hence it is best to focus on the African sample, which has the highest amount of variation in these regions, and because the African population is probably closest to equilibrium. Two observations from the African D. melanogaster data are relevant. First, the average r² is of the same order of magnitude within both su(s) and ewg (0.083 and 0.035, respectively; Table 5), as well as between the loci in the intergenic comparisons [0.034 between ewg and su(s)]. Thus we did not detect a decrease in the magnitude of linkage disequilibrium over large genomic distances. Second, the proportion of intralocus comparisons with nominally significant linkage disequilibria (17.99% at su(s) and 6.67% at ewg; Table 5) is not greater at the more distal ewg despite the clear reduction in the level of polymorphism. While there is clear evidence of recombination in the history of the sampled alleles at both ewg and su(s), the lack of any correlation with distance is consistent with gene conversion being the dominant form of recombination in this genomic region.

In D. simulans, the pattern of linkage disequilibrium is difficult to interpret. In African D. simulans, the order of magnitude of the r² is almost three times higher in su(s) than in ewg (Table 5). This difference between intralocus average r² and proportion of statistically significant associations may be ascribed to the strong skew in the frequency spectrum at ewg. On the other hand, the lack of significant interlocus associations between sites in ewg and su(s) suggests that the crossing over does contribute to recombination in this genomic region in D. simulans.

Little is known about the rate of gene conversion. Whether the few polymorphisms in these regions are those building up after a massive selective sweep or the equilibrium variation under background selection, the appearance of clear recombinants indicates that recombination (probably gene conversion) occurs at a rate comparable to (or larger than) that of neutral mutation. As new neutral mutations accumulate, they are recombined. A gene conversion rate of, for example, 10^–8/bp and a neutral mutation rate of 10^–9 may be sufficient to accommodate the observations.

Our data are similar but not identical to those from surveys of DNA sequence polymorphism on the fourth chromosome that found long-distance disequilibria as well as evidence for some form of recombination on respective regions of the D. melanogaster fourth chromosome (JENSEN et al. 2002; WANG et al. 2002). WANG et al. (2002) found Tajima's D to be –0.9745 (P = 0.1739) for all regions pooled, and JENSEN et al. (2002) found Tajima's D to be +0.47 in D. melanogaster and –0.68 in D. simulans for single-nucleotide variation at the ankyrin gene. To contrast, we had large positive values of Tajima's D. They also found two haplotypes present over long distances. Thus their results do not immediately offer insight into our data.

Random-environment models:

Linked selection models such as those studied by GILLESPIE (1997) might explain some of our results. He investigated random-environment-selection models and observed negative values of Tajima's D when selection reduces polymorphism at linked neutral sites. However, relevant sample properties of this statistic and/or appropriate parameter estimates under these models with which to conduct a statistical test on our data are not available.

Levels of polymorphism:

A number of studies have measured polymorphism at other telomeric genes in the D. melanogaster X chromosome. A comparison of our ewg and su(s) data to previous results follows. The yellow (y) gene (and its proximal neighbors ac and sc), for example, is important because it is located between ewg and su(s). As crossing over increases from ewg to su(s), it would be interesting to see how polymorphism is affected. For Zimbabwe collections of the X-linked yellow and ac, the values of were estimated as 0.0017 and 0.0012 using RFLP data (BEGUN and AQUADRO 1993). A DNA sequencing study of yellow from fly collections from Africa (Zimbabwe) estimated as 0.0003 (ANDOLFATTO and PRZEWORSKI 2001), and an expansion of that survey's sample size (n = 49) in more base pairs (2017 bp) yields a of 0.000658 (recalculated from data reported by ANDOLFATTO and WALL 2003). Meanwhile, the value of reported for ewg is 0.00035. This number and its upper bootstrap confidence limit are lower than the last value reported for yellow. A value of 0.00182 for the su(s) D. melanogaster Africa population (Zimbabwe) lies above the yellow numbers (LANGLEY et al. 2000). Thus the levels of polymorphism at these three loci in the African populations are consistent with their relative distances from the telomere and presumed relative rates of crossing over.

In D. simulans, there are only three published studies of DNA sequence variation near the telomere of the X chromosome. MARTíN-CAMPOS et al. (1992) found no variation at y-ac in a sample of 103 non-African samples. BEGUN and AQUADRO (1991) found very low variation in non-African samples ( = 0.0001 at the same genes in a North American population; n = 36). SHELDAHL et al. (2003) surveyed variation among five lines of D. simulans at the same regions mentioned above for D. melanogaster. They found an average = 0.00116 for silent variation over two lines from Africa, two from North America, and one from the Seychelles Islands. While the species average is higher in D. simulans than in D. melanogaster for regions of normal crossing over (MORIYAMA and POWELL 1996; ANDOLFATTO 2001), these three studies and ewg and su(s) exhibit more reduced variation in D. simulans than in D. melanogaster at the X telomere.

In the region from the telomere to ewg where there is presumably even less crossing over, SHELDAHL et al. (2003) also surveyed three regions. In the African (Zimbabwe) collection (n = 4), the values of (silent) were 0, 0, and 0.00272, moving from the most distal to the most proximal. The trend stops just shy of the value reported in Table 1 for ewg Africa (Zimbabwe); thus these data from SHELDAHL et al. (2003) are consistent with those from larger samples.

Demography:

Our quantification of population structure (Table 4) can be compared to F_ST values from D. melanogaster su(w^a), which had an Africa-Europe F_ST of 0.291 and an Africa-North America F_ST of 0.343 (LANGLEY et al. 2000). The values at su(s) and su(w^a) are comparable to F_ST values for X-linked regions of normal crossing over, which have been reported for Africa-North America D. melanogaster (e.g., on the basis of RFLP data: white, 0.28; vermilion, 0.32; G6pd, 0.30; Pgd, 0.25; BEGUN and AQUADRO 1993). On the basis of DNA sequence data, vermilion has F_ST values of 0.370 for Africa vs. North America (BEGUN and AQUADRO 1995a). The values at su(s) and su(w^a) are slightly lower than other values of F_ST for regions of reduced recombination (BEGUN and AQUADRO 1993). For example, BEGUN and AQUADRO (1993) estimated F_ST as 0.56 for yellow and 0.54 for ac. CHARLESWORTH (1998) showed that estimates of F_ST may be inflated when using low levels of polymorphism, which was the case for yellow and ac, so there may be no real difference in F_ST between the different regions. To contrast, ewg has an enormous value of F_ST (0.811), which was calculated using a larger number of polymorphic sites than those for yellow and ac, although the values of at ewg are lower. The large geographic differentiation at ewg reflects the fixed difference and near-fixed difference (Table 3), and it is consistent with a geographically localized hitchhiking event(s). A single parameterization of a model of geographic differentiation by genetic drift and migration would not simultaneously account for this observation and data from the rest of the genome. Hitchhiking associated with strong selection, genomically localized to the X telomere and geographically differentiated, is proposed as an ad hoc explanation here but quantitatively documented elsewhere (e.g., BAINES et al. 2004).

IRVIN et al. (1998) studied population substructure in D. simulans using microsatellites and found a much lower level of substructure than that found in D. melanogaster, similar to the trends seen in our data (Table 4). These authors interpreted this trend as the result of a much less severe bottleneck in D. simulans than what occurred in D. melanogaster and/or a more recent colonization of non-African locales by D. simulans.

We now consider whether demographic forces can explain our results for the African sample of D. melanogaster. The significantly negative Tajima's D in the African sample (–1.67, P = 0.0228; Table 1) could be the result of bottleneck or expansion. For example, GLINKA et al. (2003) interpret their data as evidence of population expansion rather than hitchhiking. They studied many X-linked loci from the same population (Zimbabwe) and found many significantly negative Tajima's D's yet no significant HKA test results and only a weak correlation between recombination and polymorphism. However, our study contrasts to theirs in several ways, leading to a different conclusion. First, GLINKA et al. (2003) studied genes from regions of normal crossing over, while the two genes in the present study are from regions of highly restricted crossing over. GLINKA et al. (2003) treat regions of reduced crossing over as exceptions, while to further understand such regions is precisely the goal of our study. Second, we observe that the amount of polymorphism at ewg is lower than that at su(s), which does suggest a correlation between crossing over and polymorphism. A bottleneck or expansion alone could not explain this correlation. Third, our HKA test results are positive, indicating an extreme reduction in polymorphism, in contrast to those of GLINKA et al. (2003). Fourth, we observed a fixed difference at one site (28,218) and a near-fixed difference at another site (27,501), and the ancestral forms of these differences occur only in samples collected outside Africa. It is unknown whether the first site is itself the target of selection, but the difference at this site is nonsynonymous, making it a more likely target than the remaining synonymous and noncoding sites. Beyond GLINKA et al. (2003), ANDOLFATTO and PRZEWORSKI (2001) studied many genes from an African sample and concluded that hitchhiking is a better explanation than demographic explanations for that data. INNAN and STEPHAN (2003) applied a different method to the same data and also found hitchhiking to be the dominant force, although they were not considering demographic explanations.

Regarding demography and selection in the other cases of significant Tajima's D's from non-African populations in this study, namely, European D. melanogaster ewg and su(s), there is also reason to believe that hitchhiking played a role. In D. melanogaster, the large F_ST values and the greater variability in the African sample support a historical migration from Africa and subsequent restricted migration. This would indicate a demographic influence on non-African polymorphism. However, both GLINKA et al. (2003) and ORENGO and AGUADé (2004) found evidence of selection in European populations. ORENGO and AGUADé (2004) point out that this is an expected process during colonization of new environments.

Our results for D. simulans included negative but not significant Tajima's D's for Africa. Again, we view those results as consistent with a study by QUESADA et al. (2003), who surveyed a different African sample, measuring variation in regions with normal to high levels of crossing over and also finding evidence for hitchhiking in D. simulans. For non-African populations, WALL et al. (2002) reanalyzed the North American D. simulans polymorphism data from BEGUN and WHITLEY (2000), and the patterns observed were found to be explainable by a simple bottleneck. However, their model fits the data only if the ancestral X:autosome effective population sizes ratio is low and if the bottleneck is strong and recent. The authors did not know how reasonable those conditions were (WALL et al. 2002). Further, those interpretations are from smaller sample sizes and genomic regions of normal crossing over per physical length and so may not be applicable to our data.

Conclusion:

The excess of rare variants at ewg, the more distal of the two loci, and high F_ST values indicate strong population subdivision among the three populations at ewg. These results indicate genetic hitchhiking at ewg and perhaps geographically localized hitchhiking events within Africa. The reduction of polymorphism at su(s) combined with the excess of high-frequency variants in D. simulans is inconsistent with the hitchhiking and background-selection models. Although the D. simulans su(s) data are difficult to explain, our data from ewg can be explained by hitchhiking in the telomeric region of the X chromosomes of both D. melanogaster and D. simulans. While this mechanism may reasonably be extrapolated to other telomeres (and perhaps centromere-proximal euchromatic sequences), the extremely reduced crossing over in these regions and the unique functional aspects of telomeres (e.g., telomere capping and the telomere's role in mitotic and meiotic segregation) restrict generalization to the entire genome.

We are grateful to the following people for their assistance: Kalpana White for providing a genomic DNA sequence of the erect wing locus; Chip Aquadro for providing Zimbabwe D. melanogaster and D. simulans flies; Michael Turelli for collecting and providing the California D. simulans flies; and Amanda Frank, Debbie Davis, Kristy Martinez, Marc Crepeau, Erja Heikkinen, and members of the Aguadé and Langley laboratories. We thank two anonymous reviewers for improving the manuscript with their comments. Funding for this research came from National Science Foundation (NSF) grant DEB 95-09548 to C.H.L. and grants PB97-0918 and BMC2001-2909 from Comisión Interdepartamental de Ciencia y Tecnología, Spain, and 2001SGR-101 from Comissió Interdepartamental de Recerca i Innovació Tecnològica, Catalonia, Spain, to M.A. J.M.B. was supported by postdoctoral fellowships from the NSF/Sloan Foundation and the Ministerio de Educación y Ciencia, Spain.

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

AGUADé, M., W. MEYERS, A. D. LONG and C. H. LANGLEY, 1994 Single-strand conformation polymorphism analysis coupled with stratified DNA sequencing reveals reduced sequence variation in the su(s) and su(w^a) regions of the Drosophila melanogaster X chromosome. Proc. Natl. Acad. Sci. USA 91: 4658–4662.[Abstract/Free Full Text]

ANDOLFATTO, P., 2001 Contrasting patterns of X-linked and autosomal nucleotide variation in Drosophila melanogaster and Drosophila simulans. Mol. Biol. Evol. 18: 279–290.[Abstract/Free Full Text]

ANDOLFATTO, P., and M. NORDBORG, 1998 The effect of gene conversion on intralocus associations. Genetics 148: 1397–1399.[Free Full Text]

ANDOLFATTO, P., and M. PRZEWORSKI, 2001 Regions of lower crossing over harbor more rare variants in African populations of Drosophila melanogaster. Genetics 158: 657–665.[Abstract/Free Full Text]

ANDOLFATTO, P., and J. D. WALL, 2003 Linkage disequilibrium patterns across a recombination gradient in African Drosophila melanogaster. Genetics 165: 1289–1305.[Abstract/Free Full Text]

AQUADRO, C. F., 1992 Why is the genome variable? Insights from Drosophila. Trends Genet. 8: 355–362.[Medline]

BAINES, J. F., A. DAS, S. MOUSSET and W. STEPHAN, 2004 The role of natural selection in genetic differentiation of worldwide populations of Drosophila ananassae. Genetics 168: 1987–1998.[Abstract/Free Full Text]

BEGUN, D. J., and C. F. AQUADRO, 1991 Molecular population genetics of the distal portion of the X chromosome in Drosophila: evidence for genetic hitchhiking of the yellow-achaete region. Genetics 129: 1147–1158.[Abstract]

BEGUN, D. J., and C. F. AQUADRO, 1993 African and North American populations of Drosophila melanogaster are very different at the DNA level. Nature 365: 548–550.[CrossRef][Medline]

BEGUN, D. J., and C. F. AQUADRO, 1995a Molecular variation at the vermilion locus in geographically diverse populations of Drosophila melanogaster and D. simulans. Genetics 140: 1019–1032.[Abstract]

BEGUN, D. J., and C. F. AQUADRO, 1995b Evolution at the tip and base of the X chromosome in an African population of Drosophila melanogaster. Mol. Biol. Evol. 12: 382–390.[Abstract]

BEGUN, D. J., and P. WHITLEY, 2000 Reduced X-linked nucleotide polymorphism in Drosophila simulans. Proc. Natl. Acad. Sci. USA 97: 5960–5965.[Abstract/Free Full Text]

BRAVERMAN, J. M., R. R. HUDSON, N. L. KAPLAN, C. H. LANGLEY and W. STEPHAN, 1995 The hitchhiking effect on the site frequency spectrum of DNA polymorphisms. Genetics 140: 783–796.[Abstract]

CELNIKER, S. E., D. A. WHEELER, B. KRONMILLER, J. W. CARLSON, A. HALPERN et al., 2002 Finishing a whole-genome shotgun: release 3 of the Drosophila melanogaster euchromatic genome sequence. Genome Biol. 3: research0079.0071–0079.0014.

CHARLESWORTH, B., 1996 Background selection and patterns of genetic diversity in Drosophila melanogaster. Genet. Res. 68: 131–149.[Medline]

CHARLESWORTH, B., 1998 Measures of divergence between populations and the effect of forces that reduce variability. Mol. Biol. Evol. 15: 538–543.[Abstract]

CHARLESWORTH, B., M. T. MORGAN and D. CHARLESWORTH, 1993 The effect of deleterious mutations on neutral molecular variation. Genetics 134: 1289–1303.[Abstract]

CHARLESWORTH, D., B. CHARLESWORTH and M. T. MORGAN, 1995 The pattern of neutral molecular variation under the background selection model. Genetics 141: 1619–1632.[Abstract]

DAVID, J. R., and P. CAPY, 1988 Genetic variation of Drosophila melanogaster natural populations. Trends Genet. 4: 106–111.[CrossRef][Medline]

DRYSDALE, R. A., M. A. CROSBY, and THE FLYBASE CONSORTIUM, 2005 FlyBase: genes and gene models. Nucleic Acids Res. 33: D390–D395 (http://flybase.org/[Abstract/Free Full Text]

FRISSE, L., R. R. HUDSON, A. BARTOSZEWICZ, J. D. WALL, J. DONFACK et al., 2001 Gene conversion and different population histories may explain the contrast between polymorphism and linkage disequilibrium levels. Am. J. Hum. Genet. 69: 831–843.[CrossRef][Medline]

GEYER, P. K., A. J. CHIEN, V. G. CORCES and M. M. GREEN, 1991 Mutations in the su(s) gene affect RNA processing in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 88: 7116–7120.[Abstract/Free Full Text]

GILLESPIE, J. H., 1997 Junk ain't what junk does: neutral alleles in a selected context. Gene 205: 291–299.[CrossRef][Medline]

GILLESPIE, J. H., 2000 Genetic drift in an infinite population: the pseudohitchhiking model. Genetics 155: 909–919.[Abstract/Free Full Text]

GLINKA, S., L. OMETTO, S. MOUSSET, W. STEPHAN and D. DE LORENZO, 2003 Demography and natural selection have shaped genetic variation in Drosophila melanogaster: a multi-locus approach. Genetics 165: 1269–1278.[Abstract/Free Full Text]

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

HUDSON, R. R., 1990 Gene genealogies and the coalescent process. Oxf. Rev. Evol. Biol. 7: 1–44.

HUDSON, R. R., 2000 A new statistic for detecting genetic differentiation. Genetics 155: 2011–2014.[Abstract/Free Full Text]

HUDSON, R. R., and N. L. KAPLAN, 1985 Statistical properties of the number of recombination events in the history of a sample of DNA sequences. Genetics 111: 147–164.[Abstract/Free Full Text]

HUDSON, R. H., and N. L. KAPLAN, 1994 Gene trees with background selection, pp. 140–153 in Non-neutral Evolution: Theories and Data, edited by G. B. GOLDING. Chapman & Hall, New York.

HUDSON, R. R., and N. L. KAPLAN, 1995 Deleterious background selection with recombination. Genetics 141: 1605–1617.[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]

HUDSON, R. R., D. D. BOOS and N. L. KAPLAN, 1992a A statistical test for detecting geographic subdivision. Mol. Biol. Evol. 9: 138–151.[Abstract]

HUDSON, R. R., M. SLATKIN and W. P. MADDISON, 1992b Estimation of levels of gene flow from DNA sequence data. Genetics 132: 583–589.[Abstract]

INNAN, H., and W. STEPHAN, 2003 Distinguishing the hitchhiking and background selection models. Genetics 165: 2307–2312.[Abstract/Free Full Text]

IRVIN, S. D., K. A. WETTERSTRAND, C. M. HUTTER and C. F. AQUADRO, 1998 Genetic variation and differentiation at microsatellite loci in Drosophila simulans: evidence for founder effects in New World populations. Genetics 150: 777–790.[Abstract/Free Full Text]

JENSEN, M. A., B. CHARLESWORTH and M. KREITMAN, 2002 Patterns of genetic variation at a chromosome 4 locus of Drosophila melanogaster and D. simulans. Genetics 160: 493–507.[Abstract/Free Full Text]

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

KIM, Y., and W. STEPHAN, 2000 Joint effects of genetic hitchhiking and background selection on neutral variation. Genetics 155: 1415–1427.[Abstract/Free Full Text]

KOUSHIKA, S. P., M. SOLLER and K. WHITE, 2000 The neuron-enriched splicing pattern of Drosophila erect wing is dependent on the presence of ELAV protein. Mol. Cell. Biol. 20: 1836–1845.[Abstract/Free Full Text]

KREITMAN, M., and R. R. HUDSON, 1991 Inferring the evolutionary histories of the Adh and Adh-dup loci in Drosophila melanogaster from patterns of polymorphism and divergence. Genetics 127: 565–582.[Abstract]

LACHAISE, D., M. CARIOU, J. R. DAVID, F. LEMEUNIER, L. TSACAS et al., 1988 Historical biogeography of the Drosophila melanogaster species subgroup, pp. 159–225 in Evolutionary Biology, edited by M. K. HECHT, B. WALLACE and G. T. PRANCE. Plenum, New York.

LANGLEY, C. H., B. P. LAZZARO, W. PHILLIPS, E. HEIKKINEN and J. M. BRAVERMAN, 2000 Linkage disequilibria and the site frequency spectra in the su(s) and su(w^a) regions of the Drosophila melanogaster X chromosome. Genetics 156: 1837–1852.[Abstract/Free Full Text]

MARTíN-CAMPOS, J. M., J. M. COMERON, N. MIYASHITA and M. AGUADé, 1992 Intraspecific and interspecific variation at the y-ac-sc region of Drosophila simulans and Drosophila melanogaster. Genetics 130: 805–816.[Abstract]

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

MIYASHITA, N., and C. H. LANGLEY, 1988 Molecular and phenotypic variation of the white locus region in Drosophila melanogaster. Genetics 120: 199–212.[Abstract/Free Full Text]

MIYASHITA, N. T., M. AGUADé and C. H. LANGLEY, 1993 Linkage disequilibrium in the white locus region of Drosophila melanogaster. Genet. Res. 62: 101–109.[Medline]

MORIYAMA, E. N., and J. R. POWELL, 1996 Intraspecific nuclear DNA variation in Drosophila. Mol. Biol. Evol. 13: 261–277.[Abstract]

NEI, M., and T. GOJOBORI, 1986 Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol. Biol. Evol. 3: 418–426.[Abstract]

ORENGO, D. J., and M. AGUADé, 2004 Detecting the footprint of positive selection in a European population of Drosophila melanogaster: multilocus pattern of variation and distance to coding regions. Genetics 167: 1759–1766.[Abstract/Free Full Text]

QUESADA, H., U. E. RAMIREZ, J. ROZAS and M. AGUADé, 2003 Large-scale adaptive hitchhiking upon high recombination in Drosophila simulans. Genetics 165: 895–900.[Abstract/Free Full Text]

ROZAS, J., J. C. SANCHEZ-DELBARRIO, X. MESSEGUER and R. ROZAS, 2003 DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496–2497.[Abstract/Free Full Text]

SHELDAHL, L. A., D. M. WEINREICH and D. M. RAND, 2003 Recombination, dominance and selection on amino acid polymorphism in the Drosophila genome: contrasting patterns on the X and fourth chromosomes. Genetics 165: 1195–1208.[Abstract/Free Full Text]

STEPHAN, W., and S. J. MITCHELL, 1992 Reduced levels of DNA polymorphism and fixed between-population differences in the centromeric region of Drosophila ananassae. Genetics 132: 1039–1045.[Abstract]

STEPHAN, W., L. XING, D. A. KIRBY and J. M. BRAVERMAN, 1998 A test of the background selection hypothesis based on nucleotide data from Drosophila ananassae. Proc. Natl. Acad. Sci. USA 95: 5649–5654.[Abstract/Free Full Text]

STURTEVANT, A. H., C. B. BRIDGES, T. H. MORGAN, L. V. MORGAN and J. C. LI, 1929 Contributions to the genetics of Drosophila simulans and Drosophila melanogaster. Carnegie Inst. Wash. 399: 1–62.

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

VOELKER, R. A., W. GIBSON, J. P. GRAVES, J. F. STERLING, M. T. EISENBERG et al., 1991 The Drosophila suppressor of sable gene encodes a polypeptide with regions similar to those of RNA-binding proteins. Mol. Cell. Biol. 11: 894–905.[Abstract/Free Full Text]

WALL, J. D., P. ANDOLFATTO and M. PRZEWORSKI, 2002 Testing models of selection and demography in Drosophila simulans. Genetics 162: 203–216.[Abstract/Free Full Text]

WANG, W., K. THORNTON, A. BERRY and M. LONG, 2002 Nucleotide variation along the Drosophila melanogaster fourth chromosome. Science 295: 134–137.[Abstract/Free Full Text]

Communicating editor: L. HARSHMAN

This article has been cited by other articles:

				M. T. Levine and D. J. Begun Evidence of Spatially Varying Selection Acting on Four Chromatin-Remodeling Loci in Drosophila melanogaster Genetics, May 1, 2008; 179(1): 475 - 485. [Abstract] [Full Text] [PDF]

				J. A. Anderson, Y. S. Song, and C. H. Langley Molecular Population Genetics of Drosophila Subtelomeric DNA Genetics, January 1, 2008; 178(1): 477 - 487. [Abstract] [Full Text] [PDF]

				P. Andolfatto Hitchhiking effects of recurrent beneficial amino acid substitutions in the Drosophila melanogaster genome Genome Res., December 1, 2007; 17(12): 1755 - 1762. [Abstract] [Full Text] [PDF]

				W.-Y. Ko, S. Piao, and H. Akashi Strong Regional Heterogeneity in Base Composition Evolution on the Drosophila X Chromosome Genetics, September 1, 2006; 174(1): 349 - 362. [Abstract] [Full Text] [PDF]

International Access Link

Online ISSN: 1943-2361  Print ISSN: 0016-6731
Copyright 2009 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu