Joint Effects of Natural Selection and Recombination on Gene Flow Between Drosophila ananassae Populations

Corresponding author: Wolfgang Stephan, Department of Biology, University of Rochester, Rochester, NY 14627-0211., stephan{at}troi.cc.rochester.edu (E-mail)

Communicating editor: C.-I WU

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We estimated DNA sequence variation in a 5.7-kb fragment of the furrowed (fw) gene region within and between four populations of Drosophila ananassae; fw is located in a chromosomal region of very low recombination. We analyzed gene flow between these four populations along a latitudinal transect on the Indian subcontinent: two populations from southern, subtropical areas (Hyderabad, India, and Sri Lanka) and two from more temperate zones in the north (Nepal and Burma). Furthermore, we compared the pattern of differentiation at fw with published data from Om(1D), a gene located in a region of normal recombination. While differentiation at Om(1D) shows an isolation-by-distance effect, at fw the pattern of differentiation is quite different such that the frequencies of single nucleotide polymorphisms are homogenized over extended geographic regions (i.e., among the two populations of the northern species range from Burma and Nepal as well as among the two southern populations from India and Sri Lanka), but strongly differentiated between the northern and southern populations. To examine these differences in the patterns of variation and differentiation between the Om(1D) and fw gene regions, we determine the critical values of our previously proposed test of the background selection hypothesis (henceforth called F_ST test). Using these results, we show that the pattern of differentiation at fw may be inconsistent with the background selection model. The data depart from this model in a direction that is compatible with the occurrence of recent selective sweeps in the northern as well as southern populations.

DURING the past decade studies of genetic variation in Drosophila have focused on the detection of natural selection at the DNA sequence level by comparing patterns of variation in gene regions of low and high recombination rates. Most of these studies have found that levels of average nucleotide diversity in low-recombination regions are reduced (AGUADE et al. 1989 ; STEPHAN and LANGLEY 1989 ). In contrast, divergence between closely related species is not affected by recombination (BERRY et al. 1991 ; BEGUN and AQUADRO 1992 ). This lack of correlation between levels of variation and divergence is not consistent with a constant-mutation-rate neutral model (KIMURA 1983 ), but can be explained by models invoking natural selection. In particular, two models have been suggested: (i) the hitchhiking model, which considers the effect of rare, strongly advantageous substitutions on linked, neutral polymorphisms (MAYNARD SMITH and HAIGH 1974 ; KAPLAN et al. 1989 ; STEPHAN et al. 1992 ), and (ii) the background selection model, which assumes that the driving mutations are frequent and strongly deleterious (CHARLESWORTH et al. 1993 ; HUDSON and KAPLAN 1995 ; CHARLESWORTH 1996 ).

Properties of these selection models have been explored for panmictic populations, but little work has been done for substructured populations. In a substructured population, background selection against deleterious mutations has been shown to increase F_ST, a relative measure of differentiation between subpopulations, in chromosomal regions of low recombination because the effective size of local demes is reduced relative to that of high-recombination regions (CHARLESWORTH et al. 1997 ). On the other hand, directional selection and genetic hitchhiking associated with the fixation of advantageous alleles may lead to greater homogeneity among populations if the selected allele causing the hitchhiking event in one deme migrates to other demes and causes a hitchhiking event in these demes as well. This scenario, however, is expected only in regions of zero or extremely low recombination. For larger recombination rates, different neutral variants may become linked to the selected allele in the population in which the advantageous mutation arose. In this situation limited migration of the selected allele may lead to temporarily increased differentiation at linked neutral loci between populations (SLATKIN and WIEHE 1998 ). Yet another case is that of local adaptation, in which the selected allele causing the hitchhiking event is locally adapted. Here hitchhiking events are assumed to be restricted to single demes or parts of the species range and, as a consequence, may cause substantial genetic differences between populations (STEPHAN and MITCHELL 1992 ; BEGUN and AQUADRO 1993 ).

To test these ideas about the migration behavior of selected genes and to compare it to that of neutral genes, we utilize the facts that (1) in some well-characterized sexually reproducing species, such as Drosophila (LINDSLEY and SANDLER 1977 ), recombination rates vary drastically along chromosomes and (2) the evolutionary dynamics of genes in regions of high recombination rates do not generally deviate from a (nearly) neutral model, whereas genes in regions of reduced recombination exhibit footprints of natural selection due to linkage to selected loci. As a species showing extensive population structure, we use Drosophila ananassae (STEPHAN et al. 1998 ). D. ananassae is a largely tropical species, which exists in many semi-isolated populations around the equator, particularly on mainland Southeast Asia and on the islands of the Pacific Ocean (TOBARI 1993 , chap. 3).

Here we report patterns of nucleotide variation in a 5.7-kb segment of furrowed (fw), a gene located in a region of very low recombination near the centromere on the right arm of the X chromosome of D. ananassae (STEPHAN and MITCHELL 1992 ), and compare it with Om(1D), a previously surveyed gene located in a region of intermediate recombination rates (STEPHAN et al. 1998 ). To explore the observed differences in the patterns of variation and differentiation between these two gene regions, we refine our recently proposed F_ST test of background selection (STEPHAN et al. 1998 ) and apply it to this new and much larger data set. As in our previous study, we estimate DNA sequence variation in four populations along a latitudinal transect on the Indian subcontinent: two from southern, subtropical areas (Hyderabad, India, and Sri Lanka), and two from more temperate zones in the north (Nepal and Burma).

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains and DNA preparation:
A total of 45 D. ananassae X chromosome lines that originated from four different collections were used in this survey: 9 lines from Mandalay (Burma), 10 lines from Hyderabad (India), 12 lines from four localities near Kathmandu in Nepal (Bharatpur, Godawari, Hetauda, and Kathmandu), and 14 from two localities in Sri Lanka (Beruwala and Colombo). The lines and the isolation of the X chromosomes are described in STEPHAN et al. 1998 . The fw sequence of D. ananassae (STD) was determined from three EcoRI subclones derived from a ca; px strain (STEPHAN and MITCHELL 1992 ). The D. pallidosa strain was obtained from the National Drosophila Stock Center (Bowling Green, OH) and put through four generations of brother-sister mating (STEPHAN et al. 1998 ).

Sequencing of fw clones:
The fw sequence (STD) of Fig 1 was determined using three EcoRI fragments (R1, R9, and R42) carried in pUC18 vectors and kindly provided by V. Corces. R1 is a 4.2-kb fragment of the fw gene region covering part of the 5' untranslated region (UTR) and exons 1–9; R9 is a 1.2-kb fragment containing exon 12 as well as part of the 3' UTR; and R42 is a 4.7-kb fragment encompassing the 3' flanking region. The GenBank accession number of R1 is AF185289 and that of R9 and R42 (combined) is AF185290. The R11 fragment, located between the R1 and R9 fragments and containing exons 10 and 11, was not sequenced because of technical difficulties (it contained long stretches of repetitive DNA). Both DNA strands of all three EcoRI fragments were sequenced. Sequencing was performed on an ABI377 automated sequencer. The total length of the region sequenced was 10,025 bp and includes most of the transcriptional unit (except for the two exons in the R11 fragment) and a large fraction of the 3' flanking region. The exon/intron structure was determined by comparison with the D. melanogaster fw sequence (LESHKO-LINDSAY and CORCES 1997 ; see also GenBank accession no. U70770).

View larger version (21K): In this window In a new window Download PPT slide   Figure 1. DNA polymorphisms of the fw gene region found in 45 lines of D. ananassae. The localities of the strains are indicated on the left. The nucleotides within the reference sequence STD are shown at the top. The numbers above the sequence represent the position numbers of each segregating site within the reference sequence. These coordinates refer to the EcoRI subclones used to determine the STD sequence: R1b, R9, and R42. Note that R1b covers most of the 3' half of the fw gene, including exon 6 (E6), exon 7 (E7), exon 9 (E9), and intron 9 (I9), while the other clones cover the 3' flanking region (F3). The homologous nucleotides within the D. pallidosa sequence (PAL) are at the bottom.

SSCP analysis and sequencing of the wild-caught fw alleles:
Single-strand conformation polymorphism (SSCP) analysis and stratified sequencing were used to identify polymorphisms because of the greater efficiency of this method, relative to direct sequencing, in regions with low levels of variation. SSCP analysis was run as described in STEPHAN et al. 1998 . A total of 34 pairs of primers were used to amplify overlapping segments of fw (numbers in parentheses are distances between primers in base pairs):

• R1 fragment: 2151–2177 and 2352–2377 (227), 2317–2338 and 2543–2563 (247), 2512–2536 and 2681–2699 (188), 2659–2678 and 2859–2878 (220), 2806–2830 and 3038–3057 (252), 2927–2946 and 3173–3193 (267), 3142–3165 and 3410–3430 (289), 3385–3406 and 3638–3658 (274), 3603–3626 and 3781–3802 (200), 3719–3740 and 3950–3974 (256), and 3928–3949 and 4124–4157 (230).

• R9 and R42 fragments: 6–34 and 248–266 (261), 215–235 and 480–495 (281), 441–460 and 590–607 (167), 480–495 and 719–741 (262), 590–607 and 814–835 (246), 719–741 and 934–957 (239), 814–835 and 1046–1066 (253), 934–957 and 1180–1200 (267), 1251–1285 and 1447–1467 (217), 1422–1444 and 1636–1655 (234), 1610–1629 and 1837–1860 (251), 1803–1828 and 2043–2065 (263), 1987–2009 and 2205–2224 (238), 2180–2202 and 2432–2452 (273), 2391–2414 and 2598–2617 (227), 2557–2584 and 2757–2788 (232), 2684–2703 and 2872–2891 (208), 2850–2871 and 3055–3079 (230), 3025–3050 and 3229–3246 (222), 3175–3201 and 3384–3407 (233), 3353–3373 and 3588–3609 (257), 3517–3535 and 3743–3765 (249), and 3685–3708 and 3925–3944 (260).

For each primer pair, all 45 wild-caught lines were scored by mobility class. Then the alleles were retested by grouping samples into mobility classes and electrophoresing similar classes side by side. Two randomly chosen lines were subsequently sequenced for each mobility class (unless only one variant existed). Both strands were sequenced. Sequencing did not reveal any undetected polymorphism; thus we proceeded assuming all variation was detected. Primers constructed for the SSCP analysis were used to sequence both fw strands of D. pallidosa.

F_ST test of the background selection model:
The basic idea of this test of the background selection model in a substructured population was described in STEPHAN et al. 1998 . For small samples, background selection generates genealogies that are approximately identical to those produced by a strict neutral model if the effective population size is adjusted such that the effects of recombination and background selection on the locus of interest are taken into account (HUDSON and KAPLAN 1995 ). The slight distortions of the allele frequency spectrum produced by background selection (CHARLESWORTH et al. 1993 ; FU 1997 ) are neglected because these can only be observed in rather large samples (not used here). The effect of background selection on neutral variation in a substructured population can thus be analyzed by simulating a neutral coalescent in an appropriate model of population structure. We use the finite island model (CROW 1986 , chap. 3.4), with background selection incorporated, as the general framework of our simulations (STEPHAN et al. 1998 ). In these simulations, the following parameters need to be specified: the per-locus nucleotide diversity _s, the migration rate M_s, the recombination rate R_s at the locus of interest, and the number of subpopulations k. The simulations were run using a modified version of the coalescent method of HUDSON et al. 1992 . The program with instructions can be downloaded from http://www.rochester.edu/college/BIO/StephanLab/YingChen.html.

The following modifications are introduced compared to the original version of the test. Instead of using Equation 1 of the previous article, the migration parameter of the locus being tested for the effect of selection, M_s, is now estimated from the data as

(1)

where M_o is the migration rate of the reference locus (located in a region of high or normal recombination), which is estimated for each pair of subpopulations of interest (rather than the whole set of subpopulations as in the previous version), –_s and _o are the arithmetic means of the per-site nucleotide diversities in the two subpopulations at the locus to be tested for the effects of selection and at the reference locus, respectively, and f_so is the ratio of the neutral mutation rate at the reference locus to that of the other locus. The factor f_so is newly introduced here to take into account that the neutral mutation rates (to be estimated from levels of divergence) may be different at the two loci. The new method of estimating M_o makes the test more conservative (see DISCUSSION). Furthermore, note that the subscript indicating the locus that may be under the effects of selection has been changed to s compared to the previous article.

Critical values of the F_ST test:
We produced empirical distributions of the test statistic F_ST as a function of the migration rate M_s for specified values of the parameters _s, R_s, k, and the sample sizes of the two subpopulations. Using the above-mentioned simulation program, 10,000 independent samples were generated for each parameter set. After these empirical distributions were obtained, the critical value at the 5% significance level was determined as the maximum value F^c_ST of F_ST such that the proportion of samples of the independently simulated samples with F_ST F^c_ST is equal to 5%. Examples of critical values for some parameter values are provided in Fig 2. To facilitate further analysis, the critical values obtained from simulation were fitted in interval [0, 1] by a function of the form

(2)

where the coefficients, a, b, c, d, and e were determined by least-squares analysis for each set of parameter values using Mathematica (WOLFRAM 1991 , Chap. 3.8).

View larger version (12K): In this window In a new window Download PPT slide   Figure 2. Critical values of the FST test at the 5% level vs. migration rate Ms. The values of the parameters k and Rs are shown; the remaining parameter values are those pertaining to the two southern populations from India and Sri Lanka (i.e., the sample sizes are 10 and 14, respectively, and the per-locus nucleotide diversity s is 1.685).

Achieved levels of significance:
When performing the F_ST test, we have to substitute estimates for several parameters. Therefore, the achieved levels of significance with critical values computed as described above may not be close to the nominal levels of significance (see the F_s test in FU 1997 ). To analyze the effect of parameter estimation on the achieved levels of the F_ST test, we followed FU's (1996) suggestion and simulated an additional 1000 samples for a given set of parameter values, independently of those used to obtain the critical points. The simulations were done in parallel for the locus of interest (denoted by index s) and the reference locus. For each simulated sample (consisting of a pair of genealogies for each locus), we estimated the values of F_ST and of nucleotide diversity and then, using Equation 1, obtained an estimate of M_s, _s. Finally, the value of the statistic F_ST was compared to the achieved critical value obtained from Equation 2 by substituting _s for M_s.

A full analysis of the properties of the F_ST test would require all parameters to be varied extensively. Such an analysis, however, is beyond the scope of this article. To reduce the number of parameters, we assumed that f_so, the ratio of the neutral mutation rate at the reference locus to that of the other locus, is accurately estimated from divergence data. The rest of the parameters were varied to some extent. Extensive simulations were done for parameter values close to those estimated from the data (see below).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

DNA polymorphism at fw:
Of the 10.0 kb of the fw clones sequenced, 5.7 kb were subjected to SSCP analysis and stratified DNA sequencing. These fragments comprise most of the 3' half of the fw transcriptional unit and a large part of the 3' flanking region. The complete polymorphism data (variations of the STD sequence) are shown in Fig 1. The estimates of nucleotide diversity (at silent + noncoding sites), and , are extremely low for each population (Table 1), ~20–50-fold lower than in regions of normal rates of crossing over in D. ananassae (STEPHAN and LANGLEY 1989 ; STEPHAN et al. 1998 ). However, they are comparable with the levels of nucleotide diversity found at vermilion (v), which is located in a chromosomal region of low recombination on the other side of the centromere on the X chromosome (STEPHAN et al. 1998 ). The values of TAJIMA's (1989) D statistic are negative in all four populations, reflecting the fact that single nucleotide polymorphisms are generally in low frequency (except for one polymorphism at coordinate 1070 in the Burma sample). Although the D values are consistently negative, they do not deviate significantly from zero.

Polymorphism and divergence:
Average divergence (at silent + noncoding sites) between D. ananassae and its sibling species D. pallidosa is very low compared to Om(1D) and v, the only other gene regions for which divergence data between D. ananassae and a close relative are available. While divergence at Om(1D) and v has been estimated as 0.032 and 0.022, respectively, divergence of all four subpopulations from the D. pallidosa sequence at fw was only 0.0055. Hudson, Kreitman, and Aguadé (HKA) tests (HUDSON et al. 1987 ) for each population, using Om(1D) as reference locus, revealed ² = 5.18 (P < 0.025) for Burma, 3.87 (P < 0.05) for Nepal, 1.89 (P < 0.17) for India, and 2.90 (P < 0.09) for Sri Lanka. Thus, there is a lack of correlation between levels of polymorphism and divergence in the two northern populations, rejecting a constant-rate, neutral model for these data. However, because of the relatively low rate of divergence between D. ananassae and D. pallidosa at fw and the lower levels of nucleotide diversity at Om(1D) in the two southern populations (see STEPHAN et al. 1998 ), the results of the HKA tests are only marginally significant for the populations from India and Sri Lanka.

Test of background selection model:
Two alternative models have been proposed to explain the reduction of DNA sequence polymorphism in regions of low rates of crossing over and the lack of correlation between polymorphism and divergence: the hitchhiking model (MAYNARD SMITH and HAIGH 1974 ; KAPLAN et al. 1989 ; STEPHAN et al. 1992 ) and the background selection model (CHARLESWORTH et al. 1993 ; HUDSON and KAPLAN 1995 ; CHARLESWORTH 1996 ). The first model assumes the hitchhiking of neutral (or nearly neutral) variants on chromosomes bearing rare, strongly selected, favorable mutations at closely linked loci that go rapidly to fixation. The second model involves the loss of neutral or nearly neutral variants as a result of steady elimination of linked deleterious mutations from the population. To test the background selection model, we used the method proposed by STEPHAN et al. 1998 with the modifications described in MATERIALS AND METHODS. This method compares genetic differentiation at a neutral marker gene with the pattern of differentiation at a locus that is to be tested for the possible effects of selection.

The prediction that background selection is expected to increase differentiation between subpopulations at fw was tested by generating the probability density of F_ST for the finite island model with k demes, a given migration rate M_s, a given mutation rate per locus _s, and a given recombination rate per locus R_s. We chose a range of reasonable values for the unknown parameters k and R_s; _s and M_s were estimated from the data [see Equation 1 and Equation 2 in STEPHAN et al. 1998 ]. The correction factor f_so of Equation 1 was estimated as 5.82, reflecting the observation that divergence at Om(1D) is ~5.82 times higher than at fw. The results are presented in Table 2. For all combinations of k and R_s values used in the simulations, the northern (Burma, Nepal) and southern (India, Sri Lanka) subsamples produced the lowest probabilities (P < 0.05). This may suggest that the F_ST values observed for the two northern subpopulations as well as for the two southern subpopulations are not compatible with those predicted by a finite island model of population structure that incorporates background selection.

Statistical properties of the F_ST test:
To interpret the P values in Table 2, we investigated the critical points of the test and the achieved levels of significance for a range of parameter values. Due to the multidimensionality of the parameter space, however, we had to concentrate on parameter values that are relevant for our data. Fig 2 shows the critical values of the F_ST test as a function of M_s for various values of the parameters k and R_s. The remaining parameter values are those used in the analysis of the two southern populations from India and Sri Lanka. The critical points as a function of M_s for other parameter sets look qualitatively very similar to those of Fig 2.

Next we examined the achieved levels of significance of the test using the critical values obtained for each parameter set. We found that parameter estimation had some influence on the test. For instance, the discrepancy between the achieved and nominal levels of significance increased with the number of subpopulations, k, and with nucleotide diversity at either locus. For the parameter sets examined, the critical points of the F_ST test (at 5% level) were the values corresponding approximately to the lower second to third percentile of its empirical distribution (similar to the F_s test; FU 1997 ). Therefore, in Table 2, adjusted P values are presented for the two northern and the two southern populations that take this bias of the test into account. Some of the adjusted P values for the two northern populations are no longer <0.05.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Variation within and between populations and divergence between species:
The estimates of nucleotide diversity, and , show extremely low levels of nucleotide polymorphism within each of the four populations from Burma, Nepal, India, and Sri Lanka. Nucleotide diversity at fw is lower than at v, which is located in a region of low recombination on the other side of the centromere of the X chromosome. This confirms earlier results from restriction map analyses and agrees with the hypothesis that recombination in the fw region is lower than at v (STEPHAN and MITCHELL 1992 ). In contrast to the vermilion data (STEPHAN et al. 1998 ), there is no evidence for a recombination event in the fw data set, based on the four-gamete rule (HUDSON and KAPLAN 1985 ).

One haplotype (henceforth called M), including the lines K8 (Nepal), H11, H81 (both from India), B8, and C6 (both from Sri Lanka), is of particular interest. The four lines from India and Sri Lanka harbor all the variation detected in the southern populations, and K8 contains much of the variation found in the northern populations. Removing this haplotype, which is distinctly different from the rest of the sample and is thus apparently a recent immigrant from a different (unknown) source population, reduces the estimates of nucleotide diversity within the four populations even more; in fact, nucleotide diversity falls to zero within each of the two southern populations.

Another interesting feature of the data is the between-population distribution of variation. Again, removing haplotype M shows that the remaining seven segregating sites are organized in a very simple way: either they are in low frequency in the total sample (three polymorphisms) or they occur as fixed differences between the northern and southern populations (three). The only exception is the polymorphism at position 1070 of fragment R9. These results confirm an earlier restriction map survey of the fw region, in which two fixed differences between the populations from Burma and India were found although the estimates of variation within these populations were zero (STEPHAN and MITCHELL 1992 ). While the northern and southern populations are strongly differentiated with respect to single nucleotide polymorphisms (but homogeneous within the northern as well as within the southern species range), it is noteworthy that the two northern populations from Burma and Nepal are differentiated with respect to two polymorphic microsatellites in intron 9.

Divergence between D. ananassae and D. pallidosa was relatively low at fw, a factor of ~4–6 lower than at v and Om(1D), the only other genes that have been sequenced for these two species. Because of this small number of genes sequenced, it is difficult to evaluate the fw result. However, lower levels of divergence in regions of reduced recombination may not be unusual. Indeed, data from D. melanogaster suggest that the variance in levels of divergence may be quite large in regions of low recombination rates relative to that in regions of intermediate or high recombination (BEGUN and AQUADRO 1992 ). Finally, given the fact that differentiation between the northern and southern populations is strong because of the presence of three nearly fixed differences, it should be noted that all four populations show the same average level of divergence from D. pallidosa. Although these two observations may appear to be inconsistent, they may simply be due to the fact that differentiation between populations was measured throughout this study by F_ST, a relative measure of differentiation, whereas divergence between species was defined in absolute terms. Measuring differentiation between populations in absolute terms, as proposed by CHARLESWORTH 1998 , shows that the fw haplotypes of the northern and southern populations are much more similar to each other than either of them is to the D. pallidosa sequence.

Test of the background selection model:
To explain our results, a basic model, such as the nearly neutral model (OHTA 1972 ), is not sufficient, although some aspects of the data (e.g., the low levels of divergence at fw) are qualitatively in agreement with a hypothesis that invokes selective constraints due to slightly deleterious mutations. Adding more strongly selected mutations to this scenario, as postulated by the background selection model (CHARLESWORTH et al. 1993 ), does not suffice either. It may explain the strong reduction of nucleotide diversity in regions of low recombination and also the relative insensitivity of the rates of divergence; i.e., the sixfold lower level of divergence at fw relative to Om(1D) may be consistent with the background selection hypothesis. Furthermore, the background selection model may explain the strong differentiation between northern and southern populations. However, as our analysis seems to indicate, the background selection model cannot account for the observed homogeneity of single nucleotide polymorphism frequencies among the populations within the northern and southern species ranges.

There are two potential problems with our analyses. First, the F_ST test presumes that variation at the reference locus Om(1D) is (nearly) neutral, and the results of our test are based on the observation that migration rates between the four subpopulations at Om(1D) are low. As mentioned in STEPHAN et al. 1998 , there is no evidence that the observed patterns of variation at Om(1D) deviate from neutrality. We have further addressed this question by measuring genetic differentiation at Om(1D) between all pairs of populations. As predicted by the neutral model, we found an "isolation-by-distance" effect (WRIGHT 1943 ) in that the association between the geographic distance between populations and F_ST is statistically significant (P < 0.025, linear regression; see Fig 3). Furthermore, recent evidence from microsatellite studies (M. SCHUG, personal communication) suggests that the pattern of variation between the four subpopulations at Om(1D) is not locus specific. Our result that migration rates between these subpopulations are low at Om(1D) is rather in qualitative agreement with Schug's observation of strong differentiation at microsatellite loci.

View larger version (13K): In this window In a new window Download PPT slide   Figure 3. Genetic differentiation at Om(1D) as a function of geographic distance. Genetic differentiation (measured by FST) is plotted against the geographic distance (in kilometers) between each pair of populations. The FST values were estimated using the method of HUDSON et al. 1992 . The distances between Mandalay (Burma), Kathmandu (Nepal), Hyderabad (India), and Colombo (Sri Lanka) were estimated by choosing the most direct route across land. The distances are ranked as follows: Hyderabad-Colombo < Mandalay-Kathmandu < Kathmandu-Hyderabad < Mandalay-Hyderabad < Kathmandu-Colombo < Mandalay-Colombo.

Second, because the background selection model was incorporated into a specific model of population structure, the finite island model, the latter may be the problem instead of the background selection model itself. Is the finite island model a realistic model of migration for D. ananassae at the geographic scale used in this study? As mentioned above, we found an "isolation-by-distance" effect in that the association between the geographic distance between populations and F_ST is statistically significant (Fig 3). This suggests that the finite island model in which migration between each pair of populations is assumed to be identical may not be an adequate description of population subdivision for D. ananassae. At the same time, this result raises the question of how a more realistic, stepping-stone-type model in which migration rates between geographically more distant populations are lower would change our results.

We address this latter question for the two southern populations from Hyderabad, India, and Sri Lanka, but a similar argument applies to the subsample of the two northern populations, for which the pattern of differentiation may also be inconsistent with the background selection hypothesis (Table 2). Consider a coalescent process for a stepping-stone-type model with k demes (i.e., populations that are assumed to be significantly differentiated among each other) with background selection incorporated such that the migration rates between populations are comparable (approximately equal) or smaller than the estimated rate of migration between the two populations from Hyderabad and Sri Lanka. This scenario is constructed based on the observation that the two southern populations under consideration (which are significantly differentiated; see STEPHAN et al. 1998 ) are separated by a smaller geographic distance than most other pairs of D. ananassae populations within the zoogeographic range of D. ananassae. (Note that this scenario includes all actual D. ananassae populations that are differentiated, not only those that have been sampled in this study.) Geography may imply that the migration rate between the Hyderabad and Sri Lanka populations is probably higher than the migration rates between most other pairs of populations within the zoogeographic range of D. ananassae. Thus, simulating a coalescent for this scenario under a background selection model would produce a distribution of F_ST values that is shifted to values larger than in the corresponding finite island model in which all migration rates are the same as between the populations from Hyderabad and Sri Lanka. As a consequence, the probability of obtaining the observed F_ST, given the background selection model incorporated into a stepping-stone-type population structure, is likely to be lower than the corresponding value shown in Table 2 for the finite island model.

Alternative explanations:
Our analyses indicate that, in addition to deleterious mutations, one may have to postulate the occasional occurrence of advantageous mutations. As with the v locus (STEPHAN et al. 1998 ), the observed pattern of differentiation at fw is consistent with recent selective sweeps, which homogenized single nucleotide polymorphism frequencies within the northern as well as within the southern populations. The sweep model seems also to be consistent with a large variance in between-species divergence rates in regions of low recombination (although no theoretical predictions are available at present). Furthermore, it appears that the fine-scale differentiation of two polymorphic microsatellite loci in intron 9 (between the Burma and Nepal populations) is in qualitative agreement with this explanation as length changes in microsatellites can occur more quickly than base substitutions (SCHUG et al. 1997 ).

Whether the data can be explained by a single-sweep model (SLATKIN and WIEHE 1998 ) or whether independent sweeps have to be postulated (STEPHAN and MITCHELL 1992 ) is difficult to decide. A single-sweep model seems to be more parsimonious than the other explanation. But for a single-sweep model to explain the data, at least two haplotypes of the fw locus must be linked to the advantageous allele. This would require that the (unknown) levels of nucleotide variation before the hitchhiking event are sufficiently high and that the sweep takes sufficiently long to spread through the northern and southern populations so that different haplotypes become associated with the advantageous mutation via recombination. The occurrence of (nearly) fixed differences between populations can be explained by both models. To distinguish between these models, low-frequency polymorphisms may be used. Most of the observed low-frequency polymorphisms, however, are probably due to recent mutations, and the variant C at position 970 of fragment R9 seems to be due to a recent migration event (discussed above).

Conclusions:
Our results suggest that the pattern of variation between populations and thus the migration behavior of nuclear genes depends critically on at least two factors (other than geography): the recombination environment in which a gene is located and natural selection. These results are important as they shed new light on inferences of the migration history of populations (including humans) from molecular data.

	FOOTNOTES

¹ Present address: Biology and Biotechnology Research Program, Lawrence Livermore National Laboratory, Livermore, CA 94550.

	ACKNOWLEDGMENTS

We thank two reviewers for their excellent comments on a previous version of this article. Furthermore, we thank Seth Bordenstein and Alex Mohseni for their help with the SSCP analysis. This work was supported in part by National Science Foundation grant DEB-9896179 to W.S. and an Ernst Caspari fellowship to Y.C.

Manuscript received August 19, 1999; Accepted for publication March 23, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

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].

BEGUN, D. J. and C. F. AQUADRO, 1992  Levels of naturally occurring DNA polymorphism correlate with recombination rates in Drosophila melanogaster.. Nature 356:519-520[Medline].

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[Medline].

BERRY, A. J., J. W. AJIOKA, and M. KREITMAN, 1991  Lack of polymorphism on the Drosophila fourth chromosome resulting from selection. Genetics 129:1111-1117[Abstract].

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, B., M. NORDBORG, and D. CHARLESWORTH, 1997  The effects of local selection, balanced polymorphism and background selection on equilibrium patterns of genetic diversity in subdivided populations. Genet. Res. 70:155-174[Medline].

CROW, J. F., 1986 Basic Concepts in Population, Quantitative, and Evolutionary Genetics. Freeman, New York.

FU, Y.-X., 1996  New statistical tests of neutrality for DNA samples from a population. Genetics 143:557-570[Abstract].

FU, Y.-X., 1997  Statistical tests of neutrality of mutations against population growth, hitchhiking and background selection. Genetics 147:915-925[Abstract].

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

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

KIMURA, M., 1983 The Neutral Theory of Molecular Evolution. Cambridge University Press, Cambridge, United Kingdom.

LESHKO-LINDSAY, L. A. and V. G. CORCES, 1997  The role of selectins in Drosophila eye and bristle development. Development 124:169-180[Abstract].

LINDSLEY, D. L. and L. SANDLER, 1977  The genetic analysis of meiosis in female Drosophila melanogaster.. Philos. Trans. R. Soc. Lond. B Biol. Sci. 277:295-312[Medline].

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

NEI, M. and W.-H. LI, 1979  Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc. Natl. Acad. Sci. USA 76:5269-5273[Abstract/Free Full Text].

OHTA, T., 1972  Evolutionary rate of cistrons and DNA divergence. J. Mol. Evol. 1:150-157.

ROZAS, J. and R. ROZAS, 1997  DnaSP version 2.0: a novel software package for extensive molecular population genetics analysis. Comput. Appl. Biosci. 13:307-311.

SCHUG, M. D., T. F. C. MACKAY, and C. F. AQUADRO, 1997  Low mutation rates of microsatellite loci in Drosophila melanogaster.. Nat. Genet. 15:99-102[Medline].

SLATKIN, M. and T. WIEHE, 1998  Genetic hitch-hiking in a subdivided population. Genet. Res. 71:155-160[Medline].

STEPHAN, W. and C. H. LANGLEY, 1989  Molecular genetic variation in the centromeric region of the X chromosome in three Drosophila ananassae populations. I. Contrasts between the vermilion and forked loci. Genetics 121:89-99[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., T. H. E. WIEHE, and M. W. LENZ, 1992  The effect of strongly selected substitutions on neutral polymorphism: analytical results based on diffusion theory. Theor. Popul. Biol. 41:237-254.

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].

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

TOBARI, Y. N., 1993 Drosophila ananassae: Genetical and Biological Aspects. Japan Scientific Societies Press, Tokyo.

WATTERSON, G. A., 1975  On the number of segregating sites in genetical models without recombination. Theor. Popul. Biol. 7:256-276[Medline].

WOLFRAM, S., 1991 Mathematica: A System for Doing Mathematics by Computer. Addison-Wesley Publishing Co., Reading, MA.

WRIGHT, S., 1943  Isolation by distance. Genetics 28:114-138[Free Full Text].

This article has been cited by other articles:

				M. D. Schug, S. G. Smith, A. Tozier-Pearce, and S. F. McEvey The Genetic Structure of Drosophila ananassae Populations From Asia, Australia and Samoa Genetics, March 1, 2007; 175(3): 1429 - 1440. [Abstract] [Full Text] [PDF]

				J. F. Baines, A. Das, S. Mousset, and W. Stephan The Role of Natural Selection in Genetic Differentiation of Worldwide Populations of Drosophila ananassae Genetics, December 1, 2004; 168(4): 1987 - 1998. [Abstract] [Full Text] [PDF]

				C. Vogl, A. Das, M. Beaumont, S. Mohanty, and W. Stephan Population Subdivision and Molecular Sequence Variation: Theory and Analysis of Drosophila ananassae Data Genetics, November 1, 2003; 165(3): 1385 - 1395. [Abstract] [Full Text] [PDF]

				A. D. Cutter and B. A. Payseur Selection at Linked Sites in the Partial Selfer Caenorhabditis elegans Mol. Biol. Evol., May 1, 2003; 20(5): 665 - 673. [Abstract] [Full Text] [PDF]

				E. Baudry, C. Kerdelhue, H. Innan, and W. Stephan Species and Recombination Effects on DNA Variability in the Tomato Genus Genetics, August 1, 2001; 158(4): 1725 - 1735. [Abstract] [Full Text] [PDF]

				P. Andolfatto and M. Przeworski Regions of Lower Crossing Over Harbor More Rare Variants in African Populations of Drosophila melanogaster Genetics, June 1, 2001; 158(2): 657 - 665. [Abstract] [Full Text] [PDF]