Skip to main content
NCBI home page
Genetics logoLink to Genetics
. 2007 Mar;175(3):1465–1477. doi: 10.1534/genetics.106.065771

Natural Genetic Variation in Cuticular Hydrocarbon Expression in Male and Female Drosophila melanogaster

Brad Foley *,†,1, Stephen F Chenoweth *, Sergey V Nuzhdin †, Mark W Blows *
PMCID: PMC1840094  PMID: 17194783

Abstract

Cuticular hydrocarbons (CHCs) act as contact pheromones in Drosophila melanogaster and are an important component of several ecological traits. Segregating genetic variation in the expression of CHCs at the population level in D. melanogaster is likely to be important for mate choice and climatic adaptation; however, this variation has never been characterized. Using a panel of recombinant inbred lines (RILs) derived from a natural population, we found significant between-line variation for nearly all CHCs in both sexes. We identified 25 QTL in females and 15 QTL in males that pleiotropically influence CHC expression. There was no evidence of colocalization of QTL for homologous traits across the sexes, indicating that sexual dimorphism and low intersex genetic correlations between homologous CHCs are a consequence of largely independent genetic control. This is consistent with a pattern of divergent sexual and natural selection between the sexes.

CUTICULAR hydrocarbons (CHCs) are important for diverse functions in insects and have been studied extensively for their roles in mate and species recognition and ecology (Tillman et al. 1999). Of particular interest in Drosophila has been the role of CHCs in sexual signaling, and in Drosophila melanogaster the biosynthesis, genetic regulation, and sex specificity of CHC expression has been well described. Much work has been conducted describing the role of CHCs in sex recognition (Savarit and Ferveur 2002a) and the genetic basis of CHC-mediated sexual isolation between closely related species in the D. melanogaster species subgroup (e.g., Coyne 1996) and between intraspecific races of D. melanogaster (e.g., Fang et al. 2002). Although segregating genetic variation for CHC expression has been demonstrated in other species of Drosophila and is associated with variation in mate choice (Blows and Higgie 2002) and traits involved in ecological adaptation such as desiccation resistance (Rouault et al. 2004), genetic variation for CHC expression within populations of D. melanogaster has not yet been characterized.

In many Drosophila, including D. melanogaster, the mate choice system is composed of several distinct elements, including a courtship dance, wing song, and the assessment of the CHCs of a potential mate through olfaction, gustation, and chemoreceptors in the front legs (for a general review, see Greenspan and Ferveur 2000). Most of the interracial and interspecies work in Drosophila mate recognition has focused on the importance of CHCs as a signal (Ferveur 2005), although the genetics of wing song and its role in mate choice has also been studied (for instance, Ritchie and Kyriacou 1996). CHCs have been implicated in mediating assortative mating between the “countryside” and “urban” strains of D. melanogaster in the Congo (Haerty et al. 2002). They are also correlated with behavioral isolation between the two major cosmopolitan and African races of D. melanogaster (Takahashi and Ting 2004), although the precise causative role of CHCs in this isolation is unclear (Coyne et al. 1999). At the among-species level, variation in CHC expression appears to contribute to reproductive isolation between D. melanogaster and its sister clade of D. simulans, D. sechellia, and D. mauritiana (Coyne et al. 1994). CHC differences between species within the D. melanogaster species subgroup also vary geographically in a manner that suggests the operation of reinforcement (Cobb and Jallon 1990).

CHCs are highly sexually dimorphic in D. melanogaster, with many of the compounds present in one sex absent in the other, while shared compounds often differ between the sexes (Jallon and David 1987). Sexual dimorphism, such as that observed in the CHCs of D. melanogaster, is expected to result from sex-specific selection. When considering sexual display traits, the classic quantitative genetic model describing the evolution of sexual dimorphism regards sex-specific sexual selection as the primary mechanism driving evolutionary divergence between the sexes (Lande 1980). Several studies have used mapping approaches to investigate within-species patterns of genetic variation in sexual traits of D. melanogaster other than CHCs (Wayne et al. 2001; Gleason et al. 2002; McMahon et al. 2002; Drnevich et al. 2004; Moehring and Mackay 2004), but these studies have generally utilized established laboratory stocks or lines derived from laboratory stocks, rather than examining standing genetic variation within populations.

The large differences between races of D. melanogaster and between other closely related species have often been studied using mutant approaches (Ferveur and Jallon 1993; Savarit et al. 1999; Marcillac et al. 2005) or mapping and introgression techniques (Coyne 1996; Takahashi et al. 2001; Greenberg et al. 2003). It remains to be seen whether these genes are important in mediating population-level variation. Of the genes and genomic regions identified by these studies, the genetic factors identified have been generally expressed in a single sex. For instance, desat2 in females (Fang et al. 2002) corresponds to a major difference in CHC expression between the African and cosmopolitan races of D. melanogaster, and smoq plays a similar role in males (Ferveur and Jallon 1996). While these sex-specific genetic factors act to modify the products of CHC biosynthesis and result in strong qualitative sexual dimorphism in D. melanogaster, the presence of these loci does not indicate to what extent genetic variance for CHC is sex specific and whether shared compounds are highly genetically correlated between the sexes.

Work in natural populations of a closely related species, D. serrata, of the montium subgroup, indicates that CHCs are sexually selected as a consequence of mate choice within populations when measured on either lab-reared (Blows et al. 2004) or field-collected phenotypes (Hine et al. 2004; Petfield et al. 2005). The consequence of sexual selection acting on CHCs in Drosophila can also be seen in the patterns of variation in CHC expression between species and races of Drosophila (Jallon and David 1987), which have been shown to be important in mate recognition (Cobb and Jallon 1990), but naturally occurring genetic variation in CHCs at the population level has yet to be characterized in D. melanogaster. Given the importance of CHCs in generating sexual isolation between races of D. melanogaster and other species, and their importance in sexual selection in the closely related D. serrata, we focus here on the genetic basis of CHC expression in D. melanogaster recently derived from a natural population.

First, we assay naturally occurring genetic variation for the entire D. melanogaster CHC phenotype (31 compounds in males and 53 compounds in females) using a panel of recombinant inbred lines (RILs). Second, after finding substantial genetic variation for >95% of compounds, we mapped QTL for CHC expression in both sexes. Finally, segregating variation in CHC expression was compared with previous descriptions of genetic differences among races of D. melanogaster and between D. melanogaster and closely related species.

MATERIALS AND METHODS

Line development and marker analysis:

The Winters lines are a panel of 144 lines derived from a cross of a single male progeny and a single female progeny of two field-caught, inseminated females. The F2 single-pair families were established, further isogenized by 25 generations of full-sib inbreeding, and then scored for marker state and retained for phenotypic analysis (described in detail by Kopp et al. 2003) although, due to attrition of inbred lines during transport and lab maintenance, not all traits were subsequently measured in all lines. The lines were genotyped for 152 variable roo inserts by in situ hybridization for the presence, absence, or polymorphism of a marker at a given cytological locus, giving a map coverage of ∼1/2 cM (Kopp et al. 2003). Any marker that was scored as segregating in a line was dropped from any further analyses. There was a single inversion on chromosome 3 (3R Payne, ∼89EF, 96A). Given the design of the cross, there were potentially four parental haplotypes for each of the autosomes and three for the X chromosome. However, when the probable parental haplotypes were reconstructed (Kopp et al. 2003), two of the parental third chromosomes appear to have been homologous, except for a small region near the tip. All linkage groups subsequently described represent these parental haplotypes.

CHC phenotyping:

Virgins were collected from the lines within 8 hr of eclosion (female line n = 124; male line n = 126) and aged in single-sex vials for 4 days. At least five flies of each sex per line were anesthetized with CO2, washed individually in 100 μl of hexane for 4 min, and vortexed for another minute before being removed (female n = 697; male n = 652) (Blows and Allan 1998). All samples were run on an Agilent 6890N gas chromatograph. Relative proportions of CHCs were transformed to logcontrasts to remove the unit-sum constraint associated with compositional data (Blows and Allan 1998).

Previous studies of Drosophila CHCs have focused primarily on the most abundant CHCs (e.g., Cobb and Jallon 1990; Coyne et al. 1994; Ferveur and Jallon 1996; Dallerac et al. 2000). While abundant CHCs may play a role in sexual isolation, less abundant CHCs are often implicated as being under sexual selection in Drosophila (Blows et al. 2004; Hine et al. 2004). Thus we took a comprehensive approach and analyzed all compounds that could be measured using gas chromatography. In females, 53 compounds were analyzed, while 31 were analyzed in males (Figure 1). For each individual, the area under each CHC peak in a chromatogram was integrated and expressed as a proportion of the total integrated area for all CHCs. These proportions were transformed into logcontrasts, an appropriate transformation when dealing with compositional data such as hydrocarbon blends as they break the unit-sum constraint associated with proportions and also have a covariance matrix that is nonsingular, unlike more commonly used log ratios (Atchison 1986; Blows et al. 2004). Calculation of logcontrasts reduces trait number to n − 1; therefore, further analyses of traits were conducted on 52 logcontrasts in females and on 30 in males. Of these, only 20 CHCs were shared between the sexes, and a great deal of sexual dimorphism was evident in the relative abundance of shared compounds as well (Figure 1). For the most part, the most abundant compounds in males are those with 23–25 carbons, whereas in females there is a greater proportion of compounds with 25–27 carbons. In this article, linear alkanes are referred to by the abbreviation Cn, where “n” is the carbon number; thus pentacosane is designated C25, for example. Methylated alkanes are referred to with a 2Me prefix, as all methyl groups are attached to the second carbon in the chain; thus 2-methyl-hexacosane is abbreviated 2MeC26. The position of double bonds is referred to before a colon. The compound 9-tricosene is referred to by the abbreviation 9:C23, and the compound 7, 11-heptacosadiene is abbreviated 7, 11:C27.

Figure 1.—

Figure 1.—

Open in a new tab

Gas chromatograph traces of cuticular hydrocarbons of D. melanogaster for (A) females and (B) males. (A) 1–9:C21; 2–C22; 3–7, 11:C23; 4–9:C23; 5–7:C23; 6–5, 9:C23; 7–5:C23; 8–C23; 9–9, 13:C24; 10–7, 11:C24; 11–9:C24; 12–7:C24; 13–5, 9:C24; 14–5:C24; 15–C24; 16–9,13:C25; 17–7,11:C25; 18–2MeC24; 19–9:C25; 20–6, 10:C25; 21–7:C25; 22–5, 9:C25; 23–5:C25; 24–C:25; 25–9, 13:C26; 26–7, 11:C26; 27–2MeC25; 28–9:C26; 29–7:C26; 30–5, 9:C26; 31–5:C26; 32–C26; 33–9, 13:C27; 34–7, 11:C27; 35–9:C27; 36–2MeC26; 37–7:C27; 38–5, 9:C27; 39–5:C27; 40–C27; 41–9, 13:C28; 42–7, 11:C28; 43–2MeC27; 44–9:C28; 45–C28; 46–9, 13:C29; 47–7, 11:C29; 48–9:C29; 49–7:C29; 50–C29; 51–7, 11:C31; 52–2MeC30; 53–C31. (B) 1–C21; 2–2MeC20; 3–9:C22; 4–7:C22; 5–5, 9:C22; 6–C22; 7–2MeC22; 8–9:C23; 9–7:C23; 10–5, 9:C23; 11–5:C23; 12–C23; 13–9:C24; 14–7:C24; 15–5, 9:C24; 16–C24; 17–2MeC24; 18–7, 11:C25; 19–9:C25; 20–7:C25; 21–5, 9:C25; 22–5:C25; 23–C25; 24–2MeC26; 25–7:C27; 26–C27; 27–2MeC28; 28–C29; 29–2MeC30; 30–2MeC32; 31–C33.

QTL analysis:

QTL analysis was performed on the line–mean logcontrasts of 50 CHC traits in females and 30 in males, which exhibited significant among-line variance using ANOVA. The association between individual markers and traits was estimated in SAS (SAS Institute, Cary, NC) using a GLM procedure (Doerge and Churchill 1996; Wang et al. 2004). The null hypothesis of no association between marker and trait was tested for each trait by randomly permuting the marker-trait associations1000 times and recording the most significant probability of association between marker and trait for each of the permutations. The statistic for marker-trait association is above the P = 0.05 threshold where it exceeds 95% of these highest permuted values (Doerge and Churchill 1996).

Composite interval mapping (CIM) was performed on the same traits using the software QTL Cartographer (Wang et al. 2001–2004). As has been discussed elsewhere (Mezey et al. 2005), the standard QTL-modeling software is not designed to analyze crosses containing up to four parental haplotypes. However, the two-allele model in QTL Cartographer can be used to analyze the effects of each marker present in a linkage group against the mean of the trait for the other linkage groups (where the roo insert is absent). For each of the parental haplotypes (three for the X chromosome, four for the second chromosome, and three for the third chromosome) a separate likelihood function was derived, using recombination distances as calculated by Mezey et al. (2005). The location of QTL for expression of all CHCs in both sexes was estimated by QTL Cartographer, using the CIM option with a walk speed of 2 cM, a window size of 10 cM, and the significance threshold estimated with 1000 permutations.

To determine whether more CHC QTL are shared between the sexes than are expected by chance—which would indicate some level of shared genetic regulation of CHC expression—we performed a resampling test (Macdonald and Goldstein 1999). We transposed the position of each of the male CHC QTL onto the female LOD profile and summed the LOD scores of all the 19 homologous CHCs across each of these QTL, both for individual traits and for all homologous CHCs at once. We compared these values with the ranked sum of LOD scores obtained by randomizing QTL position 10,000 times. The significance level for QTL coincidence is >0.05 in cases for which the observed sum of LOD scores exceeds 95% of the randomized values.

Evaluating agreement in QTL location between traits:

For all significant individual-trait QTL, 95% confidence intervals were calculated by dropping down 2-LOD scores along the likelihood function (Van Ooijen 1992; Mangin et al. 1994; Lynch and Walsh 1998). As a conservative estimate, we assumed that putative QTL for different traits with overlapping 95% confidence intervals were potentially a single, pleiotropic QTL. Pleiotropic QTL were assumed when CIM found significant LOD maxima for multiple traits at identical locations. Where two single-trait QTL had nonoverlapping confidence intervals, a minimum of two QTL were assumed in the region. While this tends to overestimate the pleiotropy of individual QTL and underestimate the number of QTL, this parsing procedure was generally a straightforward means of resolving an estimate of QTL numbers in a chromosomal region.

For presumed pleiotropic QTL, the likelihood maximum was estimated as the mode of the maxima of contributing-trait QTL. If there were equal numbers of two or more local maxima, the map position was taken to be the mean of those maxima. All QTL were plotted on a centimorgan scale (Figure 2) for each parental haplotype. Confidence intervals were calculated for these QTL using a bootstrapping procedure (Lebreton and Visscher 1998), although these estimates are likely to be conservative (Manichaikul et al. 2006). For each sex and linkage group, RILs were sampled with replacement and CIM was performed on each bootstrapped replicate. The computational overhead associated with the bootstrapping procedure was significant; thus we selected a candidate trait from each QTL to perform the bootstrapping procedure. For each new QTL, the associated trait with the highest LOD score was retained, as confidence intervals estimated on strongly supported QTL have been shown to be more accurate and more tightly resolved (Visscher et al. 1996; Lebreton and Visscher 1998). For each QTL, 150 bootstrap replicates were used to estimate the confidence intervals, which is sufficient for accurate QTL confidence interval estimation.

Figure 2.—

Figure 2.—

Figure 2.—

Figure 2.—

Figure 2.—

Open in a new tab

Linkage maps of D. melanogaster showing the confidence intervals of individual-trait QTL as estimated by 2-LOD dropdown. The estimated locations of the pleiotropic CHC QTL as determined by CIM in QTL Cartographer are also shown, with confidence intervals estimated by bootstrapping. Likelihood ratio statistic maxima are indicated. The position of markers is noted by a broken line, and cytological positions are provided. A dot on the broken line indicates that a significant association between marker and CHC expression was found in the single-marker analysis.

Comparison with other studies:

Several QTL and deficiency mapping studies have implications for this analysis. Most directly relevant is another study of the Winters lines (Wang et al. 2004) examining starvation susceptibility and longevity in both sexes. A deficiency mapping study of differences in female CHC production between D. melanogaster and D. simulans (Coyne 1996) was considered, as well as a QTL study of female reproductive success and ovariole number (Wayne et al. 2001), given the association between ovariole production and CHC expression (Wicker and Jallon 1995). All published confidence intervals of QTL and deletion intervals were already expressed in a cytogenic scale. For comparative purposes, all LOD maxima and confidence interval endpoints of CHC QTL were plotted onto the cytological scale according to their relative position between flanking markers.

A number of genetic studies of D. melanogaster CHC expression have been conducted. These have focused mainly on mutagenesis of lab stocks and genetic differences between the cosmopolitan and Zimbabwean races of D. melanogaster (Dallerac et al. 2000; Wicker-Thomas and Jallon 2000, 2001; Savarit and Ferveur 2002a). We considered genes identified in these studies to be potential candidates. As well as examining genes that have already been identified as contributing to genetic variation in CHC expression in Drosophila, we also queried FlyBase for genes that might be directly involved in CHC biosynthesis, using the key words “fatty acid biosynthesis,” “very-long-chain-fatty acid biosynthesis,” “very-long-chain-fatty acid metabolism,” “regulation of fatty acid biosynthesis,” and “fatty acid desaturation” (supplemental Table 3 at http://www.genetics.org/supplemental/). Many of the latter genes have not been cited in other studies, but have had functions inferred through sequence similarity with known genes.

RESULTS

In males, all CHCs showed significant among-line variance with univariate analysis of variance (corrected for multiple tests using the sequential Bonferroni) except for 7C:22 and 2MeC32. In females, all CHCs, apart from six (C23; 9, 13:C24; 9C24; C24; 9C25; and 7, 11:C27), showed significant among-line variance after correcting for multiple tests.

QTL analysis:

Of 30 CHCs in males, 28 were found to have significant QTL associations exceeding the significance threshold estimated by bootstrap, and of 52 CHCs in females where logcontrasts were calculated, 48 were associated with QTL (Figure 2). This slightly exceeds the number of CHCs found in females that have significant between-line variance after correction for multiple tests because, although all CHCs included in the analysis had significant between-line variance when tested by one-way ANOVA, the sequential Bonferroni was not used to exclude CHCs from QTL analysis. The additive effects for all QTL are given in supplemental Table 1 at http://www.genetics.org/supplemental/. The tight colocalization of CHC QTL in most cases (within the 2-cm limit of precision specified in QTL Cartographer) as well as overlapping 95% confidence intervals (Figure 2) is consistent with pleiotropic regulation of CHCs. The results for the CIM analysis for CHC expression largely agreed with the single-marker analysis as well, such that significant flanking marker-trait associations often supported local CIM maxima (supplemental Table 1 at http://www.genetics.org/supplemental/). As a result, most QTL reported are strongly supported by both analyses and are presumed to be significant for several traits. QTL were found for nearly all CHCs in both sexes. Only two CHCs with significant variance in expression in females (CHCs 9:C24 and C25) were not convincingly linked with any QTL.

A substantial number of QTL were found in each sex. In females, 25 CHC QTL were found, and 15 were found in males across all linkage groups (Table 1), although this might be an overestimate of the number of loci as QTL on one linkage group of a chromosome are potentially homologous with those on another linkage group on the same chromosome. While the extensive overlap of bootstrap confidence intervals suggests that several apparently adjacent QTL may be identical, this may be due to conservative bootstrap confidence intervals (Manichaikul et al. 2006) as many of the 2-LOD confidence intervals in these regions are nonoverlapping. QTL were distributed across all chromosomes and most linkage groups for both sexes; however, there was little direct evidence for shared genetic regulation of homologous traits. Considering homologous traits individually, only one, 7:C27, demonstrated significant coregulation (P = 0.021), a level that does not retain significance under a sequential Bonferroni correction for multiple testing. For all traits considered simultaneously, the summed LOD exceeds only 13.5% of the summed LOD scores of the simulations (P = 0.135). This level of independent genetic control of CHC expression between the sexes reflects the low correlations among line–means for the 19 homologous CHCs for which logcontrasts were calculated. Although 9 of these 19 CHCs displayed significant positive correlations between the sexes (data not shown), no homologous pair of traits displayed a line–mean (genetic) correlation of >0.5. No chromosome seems to have a disproportionate effect on CHC expression in either sex, given the percentage of the genome represented by each chromosome (Adams et al. 2000; Fitzpatrick 2004) (female χ2 = 1.456, d.f. = 2, P = 0.483; male χ2 = 2.340, d.f. = 2, P = 0.310) (Table 2).

TABLE 1.

Major pleiotropic QTL for CHC expression identified in male and female D. melanogaster

QTL Linkage group 95% C.I.a LOD maximum Maximum LOD scoreb Maximum traitc
Females
CHC 1 X-1 10A-17E 12E 3.975 7:C26
CHC 2 X-1 13B-17E 14F-15A 4.057 7, 11:C31
CHC 3 X-2 3BC-7E 4A 2.937 2MeC24
CHC 4 X-3 12A-12EF 12EF 3.678 7:C27
CHC 5 2-1 28A-34A 30B 4.501 9:C28
CHC 6 2-1 46D-57F 56C 3.459 7, 11:C29
CHC 7 2-2 32A-37C 35B 5.296 5:C26
CHC 8 2-2 55C-57B 56EF 5.941 7, 11:C31
CHC 9 2-3 28D-35A 31C 5.533 2MeC27
CHC 10 2-3 37A-41F/42A 41F 3.947 5, 9:C23
CHC 11 2-3 42A-44B 42B 2.924 7, 11:C29
CHC 12 3-1 64EF-69B 67B 3.951 C22
CHC 13 3-1 68A-70C 69D 7.219 9,13:C25
CHC 14 3-1 74A-79E 79C 9.58 2MeC25d
CHC 15 3-1 84C-84F 84E 9.206 2MeC25d
CHC 16 3-2&3 64B-73F 64DE 2.941 9:C25
CHC 17 3-2&3 66A-77B 70C 9.083 6, 10:C25
CHC 18 3-2&3 71C-84E 76F 3.786 7, 11:C26d
CHC 19 3-2&3 77C-90F 85C 3.931 7, 11:C26d
CHC 20 3-2&3 85C-90F 88D 4.341 7, 11:C26d
CHC 21 3-2&3 85E-91DE 90F 5.712 7, 11:C26d
CHC 22 3-4 64B-69C 64DE 3.274 9,13:C25
CHC 23 3-4 67C-86B 79AB 6.078 9, 13:C29
CHC 24 3-4 73A-88A 85C 5.492 C29
CHC 25 3-4 88B-99F 99F 4.935 7, 11:C26
Males
CHC 1 X-1 1E-4A 3E 5.28 5, 9:C25
CHC 2 X-1 9CD-13F 10C 5.058 2MeC24
CHC 3 X-1 7C-16D 9A 4.657 2MeC24
CHC 4 X-3 5F-7C 6E 2.205 5, 9:C23
CHC 5 X-3 10C-12F 12B 2.264 C25
CHC 6 2-1 22C-55C 28A 3.15 C29
CHC 7 2-2 28B-47D 31D 4.847 7:C25
CHC 8 2-3 27C-30B 29EF 9.585 7:C27
CHC 9 2-4 43BC-52A 49B 3.218 C24
CHC 10 3-1 85AB-87C 87A 5.739 C27
CHC 11 3-1 88F-98A 93D 2.509 7:C24
CHC 12 3-1 87F-98A 97B 2.832 C23
CHC 13 3-4 64B-79B 66B 2.299 C25
CHC 14 3-4 85D-90A 87A 5.225 2MeC26
CHC 15 3-4 87A-98F 94F 3.093 2MeC24
Open in a new tab
a

Determined by bootstrapping.

b

The highest single-trait LOD score associated with the QTL.

c

The trait with the highest LOD score associated with the QTL.

d

In two cases involving QTL in females, the same CHC returned the highest LOD score for adjacent QTL along a linkage group. In these cases, we used the trait with the second-highest LOD to calculate bootstrapped C.I.'s because it was impossible to verify homology of these multiple QTL among bootstrap replicates. For CHC 14 and CHC 15 on linkage group 3-1, 5:C23 and 9, 13:C26 were used, respectively, while along linkage group 3-2&3 the following substitutions were required: CHC 18–9,13:C25; CHC 19–9:C25; CHC 20–C26; CHC 21–C:25.

TABLE 2.

Observed and expected number of CHC QTL per chromosome for male and female D. melanogaster

Proportion of genomic euchromatina Female
Male
Chromosome QTL found QTL expected QTL found QTL expected
X 0.184 4 4.60 5 2.76
2 0.374 7 9.35 4 5.61
3 0.442 14 11.05 6 6.63
Open in a new tab

The number of QTL expected is calculated from the proportion of genomic euchromatin on each of the major D. melanogaster chromosomes. There was no evidence for preferential location of CHC QTL on any chromosome in females (χ2 = 1.456, d.f. = 2, P = 0.483) or males (χ2 = 2.340, d.f.=2, P = 0.310).

a

Data in this column are from Table 1, Fitzpatrick (2004).

Comparison with other studies:

The comparison with the candidate gene and QTL/deficiency studies are shown in supplemental Table 3 after conducting the searches as detailed in supplemental Table 4 at http://www.genetics.org/supplemental/. There was some correspondence between CHC QTL and previously identified QTL for ovariole number (Wayne et al. 2001). Of the four QTL identified by Wayne et al. (2001), one of the ovariole QTL corresponds with chromosomal regions dense with CHC QTL. There was also a strong correspondence between starvation resistance/longevity (as measured in a separate study in the Winters RILs) (Wang et al. 2004) and CHC expression in both sexes.

Of the nine independent QTL identified by Wang et al. (2004) in both sexes, five fell within the confidence intervals of CHC QTL in this study. There is a suggestion of a trade-off between the production of long-chain CHCs (requiring a greater input of energy and carbon resources) and starvation resistance in females. QTL 7 and 10 in females correspond to a general decrease in long-chain CHC production and an increase in starvation resistance. Both QTL 13 in males and QTL 23 in females correspond with a QTL that has a negative relation with longevity in both sexes (Table 1). Some of these associations might be due to poorly resolved QTL and confidence intervals, particularly along female linkage groups 3-2&3 and 3-4 where nearly the entire extent of these linkage groups falls within the confidence intervals of one or more QTL. Apart from the association between the longevity QTL in females and CHC QTL 23, however, all of these associations remain even when these two poorly resolved haplotypes are excluded.

There is also some indication that genomic regions that are important for differences in CHC expression between species are highly variable within populations. The region that is most important for CHC differences between D. melanogaster and D. simulans females (Coyne 1996) is also dense with CHC QTL in the Winters lines, and four of these regions (between 65F and 72D), identified through deficiency mapping, span the maxima identified as probable QTL locations. Of 12 genes or gene complexes that have previously been identified as having effects on CHC expression or on the regulation of sexual dimorphism, 10 colocalize with QTL identified in this study. If we again disregard those associations found in female linkage groups 3-2&3 and 4, only the Antennapedia and Bithorax regions are excluded. Of the 20 genes identified as having metabolic functions related to the biosynthesis of long-chain fatty acids through searching the FlyBase database, 14 potentially contribute to variation in CHC expression in the Winters lines, disregarding the associations in the two poorly resolved female linkage groups.

DISCUSSION

Line–mean analysis of the Winters panel of D. melanogaster RILs indicated the presence of a substantial level of genetic variance in almost all male and female CHCs. A large number of genetic loci for CHC expression were found in both sexes (20 in males and 29 in females), many of which contributed to pleiotropic regulation of more than one CHC. QTL were found to be distributed among all linkage groups.

Candidate genes:

Despite the potential importance of metabolism and resource allocation to variation in CHC expression, most of the previously identified candidate genes and regions, apart from sept and ecdysoneless, were potentially associated with QTL (supplemental Table 3 at http://www.genetics.org/supplemental/). Some, like the desaturase 1 and 2 genes and smoq (Ferveur and Jallon 1996), have been demonstrated to be important for racial and species differences in CHC expression and mate recognition. Similarly, differences in CHC expression between D. melanogaster and D. simulans (Coyne 1996) match well with the locations of CHC QTL found here, and it is possible that genes associated with racial and species differences are implicated in within-population variation as well. It may be that these genes, which are important sources of variation at a population level, respond well to selection under situations favoring the evolution of isolating mechanisms between species.

Many of the CHC QTL detected in both sexes did not correspond with the location of candidate CHC biosynthesis genes identified through interracial or interspecific studies. The colocalization of an ovariole QTL (Wayne et al. 2001) with a QTL affecting CHC expression is consistent with previously identified correspondences between ovariole production and CHC expression (Schal et al. 1994; Wicker and Jallon 1995). The colocalization of several starvation and longevity QTL in both sexes (Wang et al. 2004) with CHC QTL suggests that a large proportion of variation in CHC expression may reflect variation in a number of biological processes that require the expenditure of energy or lipid resources. Genes identified through a search in FlyBase for genes involved in fatty acid metabolism or biosynthesis also correspond with loci identified in this study and may be candidates for common variation in CHC expression in D. melanogaster.

Biosynthesis and expression of CHCs:

The biosynthesis of CHCs in Drosophila is well characterized and results from a simple pathway, and several of the genes regulating CHC expression have been characterized by the effects of their alleles on ratios of different classes of CHCs, such as smoq (Ferveur and Jallon 1996) and desaturase 2 (Dallerac et al. 2000). In D. melanogaster, CHCs range in length from ∼21 carbons to 31. These are synthesized by chain lengthening with the serial addition of acetate onto myristate or palmitate (Howard and Blomquist 2005), after which double bonds are selectively created by various desaturases. Branched (methyl group) hydrocarbons are also present and are synthesized in an entirely different pathway by the addition of carbon to valine, an amino acid (Howard and Blomquist 2005), and males express more methylated alkanes than females (Savarit and Ferveur 2002b). Cosmopolitan females express more CHCs with double bonds at positions 7 and 11 (an abundance of 7:Cn, 11:Cn, and 7, 11:Cn alkenes and alkadienes) rather than with the single double bonds and monoenes (an abundance of 7:Cn, and Cn alkenes and alkanes) common in cosmopolitan males (Rouault et al. 2001).

Given the simplicity of the system, we might expect to see genetic variants for CHC expression generally increasing the proportion of one class of compounds at the expense of others. There was, however, no evident simple correspondence between QTL and the relative amounts of different classes of compounds, such as 7, 11 dienes or methylated alkanes (supplemental Tables 1 and 2 at http://www.genetics.org/supplemental/). One possible exception might be QTL 7 and 10 in females, which seem to correspond with a trade-off between long and short chain CHCs and to colocalize with a QTL for starvation resistance (Wang et al. 2004). This could be suggestive of a negative correlation between long-chain CHC production and starvation resistance, perhaps due to the higher carbon and energy demands of synthesizing larger CHCs. Otherwise, the additive effects of QTL detected in the Winters RILs on CHC expression do not clearly indicate trade-offs between biochemical classes, perhaps indicating some level of control apart from biosynthesis of the blend of CHCs deposited on the cuticle.

Sexual dimorphism:

Consistent with the low correlations between homologous CHCs (none >0.5), there was little evidence for coregulation of homologous CHCs across the sexes. It should be noted, however, that it is difficult to assess which compounds to compare between the sexes, for while CHCs are produced from identical starting compounds in both sexes and share a similar biochemical pathway, there are genes for sex-specific processing of fatty acids (Gleason et al. 2005). Quite different compounds, varying in chain length or in the number and position of double bonds, may, apart from one or two processing steps late in biosynthesis, be products of identical genetic regulation while identical compounds might result from sex-specific enzymatic modifications of different substrates.

Mutant studies of CHC expression in D. melanogaster (Ferveur and Jallon 1993; Coyne et al. 1999; Dallerac et al. 2000; Wicker-Thomas and Jallon 2001; Fang et al. 2002) have also indicated independent genetic control of trait expression between the sexes (although see Labeur et al. 2002). While it might be tempting to assume that the qualitative sexual dimorphism in D. melanogaster, where almost entirely different compounds are expressed, has led to the low covariance between traits, in D. serrata, which does not exhibit qualitative differences in CHC expression between sexes or races (Howard et al. 2003), genetic covariance between males and females is likewise low although, in this case, the X chromosome plays a major role (Chenoweth and Blows 2003). In contrast, ecological traits such as longevity (Curtsinger and Khazaeli 2002; Mackay 2002; Forbes et al. 2004) and starvation resistance (Wang et al. 2004) do not seem to have quite the same degree of independent genetic regulation (but see Nuzhdin et al. 1997).

Sex-specific regulation of homologous traits is thought to be the result of sexually antagonistic selection (Lande 1980). The direction of sexual selection is expected to differ between the sexes (Johnstone et al. 1996; Bonduriansky 2001) due to their differing roles in reproduction and the differing costs of choosiness in the sexes. This difference in selection favors the evolution of independent genetic regulation of homologous traits, which in turn allows for the evolution of phenotypic sexual dimorphism (Lande and Arnold 1985). In D. melanogaster CHCs, a combination of (1) strong qualitative sexual dimorphism, (2) weak intersexual genetic correlations for homologous traits, and (3) little evidence for colocalization of QTL between the sexes all indicate low constraint on independent evolution of the sexes.

Conclusion:

Many of the loci important for interracial and between-species differences in D. melanogaster identified in previous studies colocalize with loci contributing to variation in CHC expression identified in this study. These loci are largely sex specific and lead to low intersex genetic correlations in CHC expression in this species. What remains to be determined are the selective processes responsible for the evolution of CHC sexual dimorphism in D. melanogaster.

Acknowledgments

We thank the anonymous reviewers of this article for their insightful comments. M.W.B. was supported by a grant from the Australian Research Council, and S.V.N. was supported by National Institutes of Health grant RO161773.

References

  1. Adams, M. D., S. E. Celniker, R. A. Holt, C. A. Evans, J. D. Gocayne et al., 2000. The genome sequence of Drosophila melanogaster. Science 287: 2185. [DOI] [PubMed] [Google Scholar]
  2. Atchison, J., 1986. The Statistical Analysis of Compositional Data. Chapman & Hall, London.
  3. Blows, M. W., and R. A. Allan, 1998. Levels of mate recognition within and between two Drosophila species and their hybrids. Am. Nat. 152: 826–837. [DOI] [PubMed] [Google Scholar]
  4. Blows, M. W., and M. Higgie, 2002. Evolutionary experiments on mate recognition in the Drosophila serrata species complex. Genetica 116: 239–250. [PubMed] [Google Scholar]
  5. Blows, M. W., S. F. Chenoweth and E. Hine, 2004. Orientation of the genetic variance-covariance matrix and the fitness surface for multiple male sexually selected traits. Am. Nat. 163: E329–E340. [DOI] [PubMed] [Google Scholar]
  6. Bonduriansky, R., 2001. The evolution of male mate choice in insects: a synthesis of ideas and evidence. Biol. Rev. 76: 305–339. [DOI] [PubMed] [Google Scholar]
  7. Chenoweth, S. F., and M. W. Blows, 2003. Signal trait sexual dimorphism and mutual sexual selection in Drosophila serrata. Evolution 57: 2326. [DOI] [PubMed] [Google Scholar]
  8. Cobb, M., and J. M. Jallon, 1990. Pheromones, mate recognition and courtship stimulation in the Drosophila melanogaster species sub-group. Anim. Behav. 39: 1058–1067. [Google Scholar]
  9. Coyne, J. A., 1996. Genetics of differences in pheromonal hydrocarbons between Drosophila melanogaster and D. simulans. Genetics 143: 353–364. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Coyne, J. A., A. P. Crittenden and K. Mah, 1994. Genetics of a pheromonal difference contributing to reproductive isolation in Drosophila. Science 265: 1461–1464. [DOI] [PubMed] [Google Scholar]
  11. Coyne, J. A., C. Wicker-Thomas and J. M. Jallon, 1999. A gene responsible for a cuticular hydrocarbon polymorphism in Drosophila melanogaster. Genet. Res. 73: 189–203. [DOI] [PubMed] [Google Scholar]
  12. Curtsinger, J. W., and A. A. Khazaeli, 2002. Lifespan, QTLs, age-specificity, and pleiotropy in Drosophila. Mech. Ageing Dev. 123: 81–93. [DOI] [PubMed] [Google Scholar]
  13. Dallerac, R., C. Labeur, J. M. Jallon, D. C. Knippie, W. L. Roelofs et al., 2000. A Delta 9 desaturase gene with a different substrate specificity is responsible for the cuticular diene hydrocarbon polymorphism in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 97: 9449–9454. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Doerge, R. W., and G. A. Churchill, 1996. Permutation tests for multiple loci affecting a quantitative character. Genetics 142: 285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Drnevich, J. M., M. M. Reedy, E. A. Ruedi, S. Rodriguez-Zas and K. A. Hughes, 2004. Quantitative evolutionary genomics: differential gene expression and male reproductive success in Drosophila melanogaster. Proc. R. Soc. Lond. Ser. B Biol. Sci. 271: 2267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fang, S., A. Takahashi and C.-I Wu, 2002. A mutation in the promoter of desaturase 2 is correlated with sexual isolation between Drosophila behavioral races. Genetics 162: 781–784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ferveur, J. F., 2005. Cuticular hydrocarbons: their evolution and roles in Drosophila pheromonal communication. Behav. Genet. 35: 279. [DOI] [PubMed] [Google Scholar]
  18. Ferveur, J. F., and J. M. Jallon, 1993. Nerd, a locus on chromosome III, affects male reproductive behavior in Drosophila melanogaster. Naturwissenschaften 80: 474. [DOI] [PubMed] [Google Scholar]
  19. Ferveur, J. F., and J. M. Jallon, 1996. Genetic control of male cuticular hydrocarbons in Drosophila melanogaster. Genet. Res. 67: 211. [DOI] [PubMed] [Google Scholar]
  20. Fitzpatrick, M. J., 2004. Pleiotropy and the genomic location of sexually selected genes. Am. Nat. 163: 800. [DOI] [PubMed] [Google Scholar]
  21. Forbes, S. N., R. K. Valenzuela, P. Keim and P. M. Service, 2004. Quantitative trait loci affecting life span in replicated populations of Drosophila melanogaster. I. Composite interval mapping. Genetics 168: 301. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gleason, J. M., S. V. Nuzhdin and M. G. Ritchie, 2002. Quantitative trait loci affecting a courtship signal in Drosophila melanogaster. Heredity 89: 1–6. [DOI] [PubMed] [Google Scholar]
  23. Gleason, J. M., J. M. Jallon, J. D. Rouault and M. G. Ritchie, 2005. Quantitative trait loci for cuticular hydrocarbons associated with sexual isolation between Drosophila simulans and D. sechellia. Genetics 171: 1789–1798. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Greenberg, A. J., J. R. Moran, J. A. Coyne and C. I. Wu, 2003. Ecological adaptation during incipient speciation revealed by precise gene replacement. Science 302: 1754–1757. [DOI] [PubMed] [Google Scholar]
  25. Greenspan, R. J., and J. F. Ferveur, 2000. Courtship in Drosophila. Annu. Rev. Genet. 34: 205. [DOI] [PubMed] [Google Scholar]
  26. Haerty, W., J. M. Jallon, J. Rouault, C. Bazin and P. Capy, 2002. Reproductive isolation in natural populations of Drosophila melanogaster from Brazzaville (Congo). Genetica 116: 215–224. [PubMed] [Google Scholar]
  27. Hine, E., S. F. Chenoweth and M. W. Blows, 2004. Multivariate quantitative genetics and the lek paradox: genetic variance in male sexually selected traits of Drosophila serrata under field conditions. Evolution 58: 2754. [DOI] [PubMed] [Google Scholar]
  28. Howard, R. W., and G. J. Blomquist, 2005. Ecological, behavioral, and biochemical aspects of insect hydrocarbons. Annu. Rev. Entomol. 50: 371–393. [DOI] [PubMed] [Google Scholar]
  29. Howard, R. W., L. L. Jackson, H. Banse and M. W. Blows, 2003. Cuticular hydrocarbons of Drosophila birchii and D. serrata: identification and role in mate choice in D. serrata. J. Chem. Ecol. 29: 961–976. [DOI] [PubMed] [Google Scholar]
  30. Jallon, J. M., and J. R. David, 1987. Variations in cuticular hydrocarbons among the 8 species of the Drosophila melanogaster subgroup. Evolution 41: 294–302. [DOI] [PubMed] [Google Scholar]
  31. Johnstone, R. A., J. D. Reynolds and J. C. Deutsch, 1996. Mutual mate choice and sex differences in choosiness. Evolution 50: 1382. [DOI] [PubMed] [Google Scholar]
  32. Kopp, A., R. M. Graze, S. Z. Xu, S. B. Carroll and S. V. Nuzhdin, 2003. Quantitative trait loci responsible for variation in sexually dimorphic traits in Drosophila melanogaster. Genetics 163: 771. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Labeur, C., R. Dallerac and C. Wicker-Thomas, 2002. Involvement of desat1 gene in the control of Drosophila melanogaster pheromone biosynthesis. Genetica 114: 269–274. [DOI] [PubMed] [Google Scholar]
  34. Lande, R., 1980. Sexual dimorphism, sexual selection, and adaptation in polygenic characters. Evolution 34: 292. [DOI] [PubMed] [Google Scholar]
  35. Lande, R., and S. J. Arnold, 1985. Evolution of mating preference and sexual dimorphism. J. Theor. Biol. 117: 651. [DOI] [PubMed] [Google Scholar]
  36. Lebreton, C. M., and P. M. Visscher, 1998. Empirical nonparametric bootstrap strategies in quantitative trait loci mapping: conditioning on the genetic model. Genetics 148: 525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lynch, M., and B. Walsh, 1998. Genetics and Analysis of Quantitative Traits. Sinauer Associates, Sunderland, MA.
  38. Macdonald, S. J., and D. B. Goldstein, 1999. A quantitative genetic analysis of male sexual traits distinguishing the sibling species Drosophila simulans and D. sechellia. Genetics 153: 1683–1699. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mackay, T. F. C., 2002. The nature of quantitative genetic variation for Drosophila longevity. Mech. Ageing Dev. 123: 95–104. [DOI] [PubMed] [Google Scholar]
  40. Mangin, B., B. Goffinet and A. Rebai, 1994. Constructing confidence intervals for QTL location. Genetics 138: 1301–1308. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Manichaikul, A., J. Dupuis, S. Sen and K. W. Broman, 2006. Poor performance of bootstrap confidence intervals for the location of a quantitative trait locus. Genetics 174: 481–489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Marcillac, F., F. Bousquet, J. Alabouvette, F. Savarit and J. F. Ferveur, 2005. A mutation with major effects on Drosophila melanogaster sex pheromones. Genetics 171: 1617–1628. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. McMahon, C., G. Sureau and J. F. Ferveur, 2002. Mapping of genetic loci that change pheromone discrimination in Drosophila males. Genet. Res. 79: 149–159. [DOI] [PubMed] [Google Scholar]
  44. Mezey, J. G., D. Houle and S. V. Nuzhdin, 2005. Naturally segregating quantitative trait loci affecting wing shape of Drosophila melanogaster. Genetics 169: 2101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Moehring, A. J., and T. F. C. Mackay, 2004. The quantitative genetic basis of male mating behavior in Drosophila melanogaster. Genetics 167: 1249. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Nuzhdin, S. V., E. G. Pasyukova, C. L. Dilda, Z. B. Zeng and T. F. C. Mackay, 1997. Sex-specific quantitative trait loci affecting longevity in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 94: 9734–9739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Petfield, D., S. F. Chenoweth, H. D. Rundle and M. W. Blows, 2005. Genetic variance in female condition predicts indirect genetic variance in male sexual display traits. Proc. Natl. Acad. Sci. USA 102: 6045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Ritchie, M. G., and C. P. Kyriacou, 1996. Artificial selection for a courtship signal in Drosophila melanogaster. Anim. Behav. 52: 603. [Google Scholar]
  49. Rouault, J., P. Capy and J. M. Jallon, 2001. Variations of male cuticular hydrocarbons with geoclimatic variables: An adaptative mechanism in Drosophila melanogaster? Genetica 110: 117–130. [DOI] [PubMed] [Google Scholar]
  50. Rouault, J. D., C. Marican, C. Wicker-Thomas and J. M. Jallon, 2004. Relations between cuticular hydrocarbon (HC) polymorphism, resistance against desiccation and breeding temperature: a model for HC evolution in D. melanogaster and D. simulans. Genetica 120: 195–212. [DOI] [PubMed] [Google Scholar]
  51. Savarit, F., and J. F. Ferveur, 2002. a Genetic study of the production of sexually dimorphic cuticular hydrocarbons in relation with the sex-determination gene transformer in Drosophila melanogaster. Genet. Res. 79: 23–40. [DOI] [PubMed] [Google Scholar]
  52. Savarit, F., and J. F. Ferveur, 2002. b Temperature affects the ontogeny of sexually dimorphic cuticular hydrocarbons in Drosophila melanogaster. J. Exp. Biol. 205: 3241–3249. [DOI] [PubMed] [Google Scholar]
  53. Savarit, F., G. Sureau, M. Cobb and J. F. Ferveur, 1999. Genetic elimination of known pheromones reveals the fundamental chemical bases of mating and isolation in Drosophila. Proc. Natl. Acad. Sci. USA 96: 9015–9020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Schal, C., X. P. Gu, E. L. Burns and G. J. Blomquist, 1994. Patterns of biosynthesis and accumulation of hydrocarbons and contact sex-pheromone in the female German cockroach, Blattella germanica. Arch. Insect Biochem. Physiol. 25: 375. [DOI] [PubMed] [Google Scholar]
  55. Takahashi, A., and C. T. Ting, 2004. Genetic basis of sexual isolation in Drosophila melanogaster. Genetica 120: 273–284. [DOI] [PubMed] [Google Scholar]
  56. Takahashi, A., S. C. Tsaur, J. A. Coyne and C. I. Wu, 2001. The nucleotide changes governing cuticular hydrocarbon variation and their evolution in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 98: 3920–3925. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Tillman, J. A., S. J. Seybold, R. A. Jurenka and G. J. Blomquist, 1999. Insect pheromones: an overview of biosynthesis and endocrine regulation. Insect Biochem. Mol. Biol. 29: 481–514. [DOI] [PubMed] [Google Scholar]
  58. Van Ooijen, J. W., 1992. Accuracy of mapping quantitative trait loci in autogamous species. Theor. Appl. Genet. 84: 803. [DOI] [PubMed] [Google Scholar]
  59. Visscher, P. M., R. Thompson and C. S. Haley, 1996. Confidence intervals in QTL mapping by bootstrapping. Genetics 143: 1013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Wang, M. H., O. Lazebny, L. G. Harshman and S. V. Nuzhdin, 2004. Environment-dependent survival of Drosophila melanogaster: a quantitative genetic analysis. Aging Cell 3: 133. [DOI] [PubMed] [Google Scholar]
  61. Wang, S., C. J. Basten and Z. B. Zeng, 2001–2004 Windows QTL Cartographer 2.0. Department of Statistics, North Carolina State University, Raleigh, NC.
  62. Wayne, M. L., J. B. Hackett, C. L. Dilda, S. V. Nuzhdin, E. G. Pasyukova et al., 2001. Quantitative trait locus mapping of fitness-related traits in Drosophila melanogaster. Genet. Res. 77: 107–116. [DOI] [PubMed] [Google Scholar]
  63. Wicker, C., and J. M. Jallon, 1995. Influence of ovary and ecdysteroids on pheromone biosynthesis in Drosophila melanogaster (Diptera, Drosophilidae). Eur. J. Entomol. 92: 197–202. [Google Scholar]
  64. Wicker-Thomas, C., and J. M. Jallon, 2000. Role of Enhancer of zeste on the production of Drosophila melanogaster pheromonal hydrocarbons. Naturwissenschaften 87: 76–79. [DOI] [PubMed] [Google Scholar]
  65. Wicker-Thomas, C., and J. M. Jallon, 2001. Control of female pheromones in Drosophila melanogaster by homeotic genes. Genet. Res. 78: 235–242. [DOI] [PubMed] [Google Scholar]

Articles from Genetics are provided here courtesy of Oxford University Press

RESOURCES