Patterns of Nucleotide Polymorphism and Divergence in the Odorant-Binding Protein Genes OS-E and OS-F: Analysis in the Melanogaster Species Subgroup of Drosophila

Corresponding author: Julio Rozas, Facultat de Biologia, Universitat de Barcelona, Diagonal 645 08028, Barcelona, Spain., jrozas{at}ub.edu (E-mail)

Communicating editor: M. VEUILLE

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The Olfactory Specific-E and -F genes (OS-E and OS-F) belong to the odorant-binding protein gene family, which includes the general odorant-binding proteins and the pheromone-binding proteins. In Drosophila melanogaster, these genes are arranged in tandem in a genomic region near the centromere of chromosome arm 3R. We examined the pattern of DNA sequence variation in an 7-kb genomic region encompassing the two OS genes in four species of the melanogaster subgroup of Drosophila and in a population sample of D. melanogaster. We found that both the OS-E and the OS-F gene are present in all surveyed species. Nucleotide divergence estimates would support that the two genes are functional, although they diverge in their functional constraint. The pattern of nucleotide variation in D. melanogaster also differed between genes. Variation in the OS-E gene region exhibited an unusual and distinctive pattern: (i) a relatively high number of fixed amino acid replacements in the encoded protein and (ii) a peak of nucleotide polymorphism around the OS-E gene. These results are unlikely under the neutral model and suggest the action of natural selection in the evolution of the two odorant-binding protein genes.

THE olfactory system of terrestrial animals has an extreme sensitivity and specificity. It can detect and discriminate a large number of olfactory signals, the odorants. Olfactory perception is accomplished by specialized bipolar sensory neurons that extend their dendrites into an aqueous medium: the olfactory mucus in vertebrates and the sensillar fluid in insects (STERN and MARX 1999 ). Hence, the airborne molecules must traverse the aqueous space that separates neuronal cells from the external air and stimulate the odorant receptors (STEINBRECHT 1969 , STEINBRECHT 1996 ). These receptors are located on the dendritic membrane of the sensory neurons (BUCK and AXEL 1991 ; VOSSHALL et al. 1999 ).

The odorant-binding proteins (OBPs) are abundant low-molecular-weight proteins that bind and solubilize hydrophobic odorants (or pheromones) in the vertebrate olfactory mucus and in the insect sensillar lymph. These small globular proteins are synthesized and secreted by some accessory cells surrounding the sensory neurons. In insects, the OBP family includes the general odorant-binding proteins (GOBPs) and the pheromone-binding proteins (PBPs), which are not homologous to vertebrate odorant-binding proteins (VOGT and RIDDIFORD 1981 ; PELOSI and MAIDA 1995 ).

Despite the low sequence similarity among different insect OBPs, most of these proteins exhibit a similar distribution of conserved hydrophobic residues with a nearly identical predicted secondary structure. Most proteins of this family contain six highly conserved cysteines located in similar positions of the protein (PIKIELNY et al. 1994 ). In Lepidoptera, these cysteines are involved in disulfide bridges in both PBPs and GOBPs (SCALONI et al. 1999 ). The similar distribution of cysteine residues in both groups of OBPs suggests that the disulfide-bridge pairing might be a general feature of this family of molecules in insects.

Although the specific function of OBPs in olfaction is still unknown, they seem to play an important role in olfactory coding. It has been shown that several OBPs have different odorant specificities and are present in distinct subsets of antennal sensilla (PELOSI and MAIDA 1995 ). Additionally, genes encoding olfactory receptors with different binding specificities are also expressed in specific areas of the olfactory organ (VOSSHALL et al. 2000 ). These observations suggest that these proteins might participate in odor detection by restricting the spectrum of odorants accessible to the underlying receptors. In addition to the established functions of OBPs as carrier molecules and in concentrating hydrophobic odorants in the aqueous medium, it has also been proposed that these proteins could participate in the deactivation of the odorant stimulus (PELOSI and MAIDA 1995 ).

In Drosophila melanogaster, 51 putative members of the OBP family have been identified (HEKMAT-SCAFE et al. 2002 ; VOGT et al. 2002 ). Two of theses proteins, OS-E (olfactory-specific E) and OS-F (olfactory-specific F), colocalize at the same restricted area of the ventro-lateral region of the antenna. The OS-E and OS-F genes (named Obp83a and Obp83b in HEKMAT-SCAFE et al. 2002 ), which have a similar intron-exon organization, are arranged in tandem in cytological band 83CD of the third chromosome. The two encoded proteins are highly conserved (72% amino acid identity in the mature protein; PELOSI and MAIDA 1995 ) except for a region in the C-terminal domain of the proteins, which has been named the heterogeneous region (hr; HEKMAT-SCAFE et al. 2000 ). The close physical proximity and the high degree of sequence similarity of the two coding regions seem to reflect a recent gene duplication event (MCKENNA et al. 1994 ; see, however, HEKMAT-SCAFE et al. 2000 ). Two hypotheses have been proposed to explain the low amino acid conservation of the hr region: (i) the hr region might form a putative binding site for the odorant molecule and (ii) the hr region might be a putative contact site for the olfactory receptor (HEKMAT-SCAFE et al. 2000 ).

Since olfaction is essential for survival and reproduction, genes involved in olfactory perception have likely evolved by the action of positive natural selection. Indeed, recognition and discrimination of olfactory signals are critical for finding food sources and for the reproduction of individuals; furthermore, certain chemoreceptive processes, like pheromone perception, contribute to critical evolutionary processes such as reproductive isolation and speciation. In fact, positive natural selection has been proposed to be involved in the evolution of PBPs of the moth Chortstoneura (Lepidoptera; WILLETT 2000 ) and also in the evolution of the OBPs of fire ants and other closely related species, in which these proteins could control some aspects of social organization (KRIEGER and ROSS 2002 ). Here, we analyze DNA variation at the OS-E and OS-F genes in four species of the melanogaster subgroup of Drosophila (D. melanogaster, D. simulans, D. mauritiana, and D. erecta) and also in a natural population of D. melanogaster to infer the evolutionary history of these genes. We found that the two genes are present in all surveyed species and thus originated from an ancient duplication event; nevertheless, these genes differ in their functional constraint. We show that the OS-E gene region has very distinctive evolutionary patterns, specifically, (i) an accumulation of fixed amino acid replacements in the OS-E protein of D. melanogaster and (ii) an atypical pattern of nucleotide polymorphism. These results suggest that positive natural selection was likely involved in the evolution of this gene.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks:
Fourteen D. melanogaster isochromosomal strains for the third chromosome were used; these strains were obtained from flies collected in a natural population of Montemayor, Spain, with crosses with the TM6/MKRS balancer stock (CIRERA and AGUADE 1997 ; RAMOS-ONSINS and AGUADE 1998 ). A highly inbred D. simulans line (S40; from a natural population in Montblanc, Spain), obtained by 10 generations of sib mating, was also used (ROZAS et al. 2001 ). Additionally, one line of each D. mauritiana and D. erecta kindly provided by F. Lemeunier were included in the present study.

DNA extraction, PCR amplification, and DNA sequencing:
Genomic DNA from the D. melanogaster lines was CsCl purified (BINGHAM et al. 1981 ). DNA from D. simulans, D. mauritiana, and D. erecta was extracted from a single individual by using a modification of protocol 48 in ASHBURNER 1989 . In D. melanogaster, an 4.7-kb genomic region that includes the complete coding region of both the OS-E and OS-F genes, the intergenic region, and 174 bp of the OS-E 5' flanking region was amplified by PCR (SAIKI et al. 1988 ; Fig 1). An additional 2-kb region upstream of the OS-E gene was PCR amplified in 13 of the D. melanogaster lines (in all lines except line M47). The amplified fragments were purified with Qiaquick columns (QIAGEN, Chatsworth, CA) and subsequently sequenced by using several oligonucleotides designed at intervals of 400 nucleotides. The sequenced fragments were separated on ABI PRISM 377 and 3700 automated DNA sequencers. For each line, the DNA sequence was determined on both strands.

View larger version (11K): In this window In a new window Download PPT slide   Figure 1. Physical map of the OS region. Solid and shaded boxes indicate translated and untranslated exons, respectively. Numbers above the thick line indicate the position of the nucleotide polymorphisms detected in D. melanogaster. Nucleotide sites are numbered from the translation initiation site of the OS-E gene. The 2-kb region upstream of the OS-E gene is not included in the figure.

For D. simulans and D. mauritiana, the same 4.7-kb region was amplified and sequenced by using several of the primers designed for D. melanogaster and, for the more divergent DNA regions, by primer walking. In D. erecta, only the OS-E and OS-F genes were PCR amplified and sequenced. In this species, primers for amplification and sequencing were designed on the most conserved regions of the genes among the other three species and also by the primer walking technique.

Data analysis:
DNA sequences were assembled using the SeqEd version 1.0.3 program (Applied Biosystems, Foster City, CA). Sequences were multiply aligned with the ClustalW program (THOMPSON et al. 1994 ), and the initial alignment was optimized manually. The MacClade program, version 3.06 (MADDISON and MADDISON 1992 ), was used to edit the DNA sequences for further analyses. The secondary structure of the OS-E and OS-F proteins was inferred by using the PHD and PROF secondary structure prediction programs (ROST 2001 ). The DNA divergence among the studied species was estimated as K, the number of nucleotide differences per site corrected according to JUKES and CANTOR 1969 . Phylogenetic analysis was performed using the neighbor-joining algorithm (SAITOU and NEI 1987 ) implemented in the MEGA version 2 program (KUMAR et al. 2000 ) and by the maximum-likelihood method (FELSENSTEIN 1993 ). The bootstrap analysis was based on 1000 replicates. Coding DNA sequences at internal nodes of the phylogenetic tree were reconstructed by the maximum-likelihood ancestral reconstruction approach using codon substitution models (GOLDMAN and YANG 1994 ; YANG et al. 1995 ). From the ancestral sequences, we estimated the number of synonymous and nonsynonymous substitutions in each branch. All these analyses were performed using the codeml program included in the PAML 3.0 software (YANG 1997 ).

The DnaSP version 3.98 program (ROZAS and ROZAS 1999 ) was used for most intraspecific and some interspecific analyses. The level of DNA polymorphism was estimated as the per-site nucleotide diversity (; NEI 1987 ), the Watterson parameter (WATTERSON 1975 ), and the haplotype diversity (Hd; NEI 1987 ). Codon bias was measured as the effective number of codons (ENC; WRIGHT 1990 ), which measures the deviation from equal usage of synonymous codons.

The recombination parameter c (in Drosophila, c = 2Nr, where N is the effective population size and r is the recombination rate per generation between adjacent sites) was estimated using three different methods. The HUDSON 1987 method estimates c from the variance of the average number of nucleotide differences between pairs of sequences, while the HUDSON and KAPLAN 1985 method estimates c from the minimum number of recombination events in the sample (R_M). The latter method requires the use of coalescent simulations to estimate c. An estimate of c based on the D. melanogaster recombination map was also obtained, assuming (i) that the recombination rate of the OS region (which is located on band 83CD) is 0.16 x 10^-8 (see COMERON et al. 1999 ) and (ii) that N is 10⁶ for D. melanogaster.

Statistical tests:
The TAJIMA 1989 and the FU and LI 1993 tests were used to contrast whether the polymorphism frequency distribution (frequency spectrum) conforms to neutral expectations. The overall genetic association between polymorphic sites was determined by the Z_nS (KELLY 1997 ), Wall's B and Q (WALL 1999 ), and Z_A (ROZAS et al. 2001 ) statistics. The confidence intervals of these statistics were obtained by computer simulations (10,000 replicates) on the basis of the coalescent process assuming a large constant population size (KINGMAN 1982 ; HUDSON 1983 , HUDSON 1990 ; ROZAS and ROZAS 1999 ). Coalescent simulations were conducted using different values of the recombination parameter and conditioning on the number of segregating sites.

The Hudson-Kreitman-Aguadé (HKA) test (HUDSON et al. 1987 ) was conducted to assess whether the levels of polymorphism and divergence between species correlated, as expected under neutrality. The test was carried out using the 5' Adh gene region (KREITMAN and AGUADE 1986 ) as a neutral evolving region.

The Kolmogorov-Smirnov statistic (D_KS) was used to test for heterogeneity in the ratio of polymorphism to divergence along the surveyed DNA region. The test is based on the maximum absolute difference between the observed and the expected cumulative number of polymorphic sites (SOKAL and ROHLF 1995 ; MCDONALD 1998 ), and it is generally the most powerful test for regions with two areas with very different levels of variation (MCDONALD 1998 ). The WU and LI 1985 relative-rate test was used to test for heterogeneity in the nucleotide substitution rate among lineages. This method is based on the standardized difference of the corrected estimates of the number of substitutions per site between two lineages. The K2WULI program (JERMIIN 1996 ) was used to perform this analysis. For nonsynonymous sites we used the ² test due to the small number of substitutions.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Interspecific analysis:
We have identified the OS-E and OS-F genes in the four species studied (D. melanogaster, D. simulans, D. mauritiana, and D. erecta). Moreover, both the intron-exon structure and the physical distance between genes are maintained across these species. Our results contrast with the previous report of HEKMAT-SCAFE et al. (2000), where the OS-E gene was not detected in either D. simulans or D. mauritiana. Nevertheless, the methodology used in their survey, a restriction-enzyme-based analysis, likely precluded its detection. Fig 2 shows the multiple alignment of the amino acid sequences encoded by the OS-E and OS-F genes. The six highly conserved cysteines of the OBP family are present in all OS proteins, except for the second cysteine of the OS-E protein in D. erecta that was replaced by a tryptophan. Moreover, the PHD and PROF programs predicted that all OS proteins are helical rich. To obtain clues on the function of specific parts of these proteins, the predicted structure of OS-E and OS-F was compared with that obtained for the pheromone-binding protein of Bombix mori (BmPBP). This protein is also a member of the OBP family and its three-dimensional (3D) structure has been determined by X-ray crystallography (SANDLER et al. 2000 ). The distribution of the predicted -helices along the OS proteins (Fig 2) is nearly identical to that found for the BmPBP.

View larger version (30K): In this window In a new window Download PPT slide   Figure 2. Amino acid sequence alignment of the OS-E and OS-F proteins. Amino acids identical to the first sequence are indicated by a dot. Shaded residues indicate the six cysteines highly conserved in the OBP family. The Cys-to-Trp replacement (OS-E protein of D. erecta) is also indicated. The line underneath the aligned sequences indicates the predicted signal peptide. The location of the predicted -helices is indicated by open boxes below the sequences. The box in the sequences delimits the hr region (see the Introduction). Dmel, D. melanogaster; Dsim, D. simulans; Dmau, D. mauritiana; Dere, D. erecta.

Nucleotide divergence between species was estimated using only the DNA sequence fragment clearly alignable among all species (Table 1); the D. melanogaster line M2 was used for this analysis. In general, nucleotide divergence was higher in the OS-E than in the OS-F region. Despite this difference, both genes have a very similar and quite low codon bias, with an average ENC value for all species equal to 50.55 for OS-E and to 45.28 for OS-F. In both genes, higher K_S than K_A values were detected. In the OS-E gene, divergence estimates were higher at synonymous than at noncoding sites.

Fig 3 shows the neighbor-joining trees reconstructed for the OS-E and OS-F genes (the same topology is obtained using the maximum-likelihood approach). In the OS-E tree, the branch leading to the D. melanogaster lineage was rather long. We conducted a relative-rate test (WU and LI 1985 ), using D. erecta as the outgroup, to determine whether the numbers of substitutions in the D. melanogaster and in the D. mauritiana (or D. simulans) lineages were significantly different. This test revealed that the OS-E region (including coding and noncoding sites) evolves faster in the D. melanogaster than in the D. mauritiana and D. simulans lineages (z = 2.052, P = 0.021 for D. mauritiana; z = 1.395, P = 0.081 for D. simulans). Equivalent z results were obtained when only the coding region was used (z = 2.006, P = 0.024 for D. mauritiana; z = 1.994, P = 0.023 for D. simulans). In fact, the significantly higher number of substitutions accumulated in the D. melanogaster lineage was mainly due to nonsynonymous substitutions (P = 0.058). All amino acid replacements fixed in the D. melanogaster lineage are conservative, i.e., with very low physicochemical distance values (GRANTHAM 1974 ).

View larger version (11K): In this window In a new window Download PPT slide   Figure 3. Phylogenetic trees of the OS-E and OS-F genes. The trees were built using 651 sites (423 coding sites) and 805 sites (462 coding sites) for OS-E and OS-F, respectively. The numbers of nucleotide substitutions and the number of nonsynonymous changes (in parentheses) are indicated on each branch. The percentages of bootstrap replicates supporting the different nodes are in italics.

Nucleotide polymorphism in D. melanogaster:
Fig 1 and Fig 4 show the distribution of DNA polymorphic sites along the 4.7-kb region surveyed. A total of 25 nucleotide polymorphisms (9 of them with singleton variants) and 1 indel polymorphism (6 bp) were detected. All polymorphisms were silent: 1 synonymous polymorphism at site 30 of the OS-E coding region and the rest at noncoding positions. Notably, polymorphism at site 540 results in two different stop codons (TAG and TAA) of the OS-E gene. Ten different haplotypes (Hd = 0.956) were detected in the 14 lines analyzed.

View larger version (23K): In this window In a new window Download PPT slide   Figure 4. Nucleotide polymorphisms detected in the OS region of D. melanogaster. Nucleotides identical to the first sequence are indicated by a dot. For length polymorphisms, the nucleotide position refers to the first site affected. d, deletion; d#, deletion of #bp. The last row gives, for the polymorphic positions in D. melanogaster, the nucleotide information in D. mauritiana and D. simulans. E1, exon 1. Information for the additional analysis of the upstream region is also shown.

Estimates of the per-site recombination parameter [c = 0.0026 from HUDSON 1987 , c = 0.0024 from HUDSON and KAPLAN 1985 , and c = 0.0032 from comparison of the physical and recombination maps] clearly indicate that the OS genomic region shows a reduction from the normal recombination levels (COMERON et al. 1999 ; ANDOLFATTO and PRZEWORSKI 2000 ). Table 2 summarizes the levels of nucleotide variation estimated separately for the different functional parts of the OS region. Estimates of nucleotide diversity for the complete 4.7-kb region ( = 0.0018, silent = 0.0021) were rather low. However, levels of nucleotide diversity varied considerably along the OS region. Silent variation was highest in the OS-E gene ( = 0.0081) and lowest in the OS-F gene ( = 0.0013). Putative heterogeneity in the distribution of polymorphic to fixed silent sites along the OS region was tested by means of the D_KS test. Significant heterogeneity (P = 0.03) along the region studied was detected using the most conservative c value (see MCDONALD 1998 ). We also compared, by means of an HKA test, silent polymorphism and divergence in the total OS region (as well as separately for each functional part) and in the 5' Adh region (HUDSON et al. 1987 ). A significant deviation from the neutral prediction was detected for the total OS region (P = 0.027). Interestingly, only variation in the OS-F region departed significantly from neutral predictions (OS-E region, P = 0.276; OS-F region, P = 0.017; intergenic region, P = 0.231).

We also conducted Tajima's D and Fu and Li's D and F tests separately for the three functional regions (OS-E, intergenic, and OS-F) to determine whether the frequency distribution of nucleotide variants departs from that expected under neutrality. For this analysis we used D. mauritiana as the outgroup. Under no recombination (which is the most conservative assumption for these tests), the analysis showed a significant deviation only in the OS-E region (Table 3). The significantly positive values of the Tajima's D and Fu and Li's F statistics in this region reflect an excess of nucleotide variants at intermediate frequencies. For the OS-F and intergenic regions, the test statistic values substantially increased when line M20 (which accounts for seven of the nine singletons found in the sample) was removed.

No overall significant association between polymorphic sites (linkage disequilibrium) was detected by the Z_nS statistic either in the whole region (Z_nS = 0.257; P = 0.541) or in its different functional parts (results not shown). Nevertheless, a significant association was detected between polymorphic sites in the OS-E region (4 of the 10 pairwise comparisons were significant even after the conservative Bonferroni correction). The OS-E region also showed significant values of the Z_A and Wall's B and Q statistics even using the conservative no-recombination assumption (Table 3). These results indicate that nucleotide variation at the OS-E region is highly structured (Fig 4).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The presence of both the OS-E and OS-F genes in the four Drosophila species studied, and also in species of the obscura group (A. SÁNCHEZ-GRACIA, M. AGUADÉ and J. ROZAS, unpublished results), indicates that the DNA duplication event is relatively old. Nucleotide divergence estimates among copies and among species and the phylogenetic trees clearly indicate that the two genes have evolved independently since their origin by gene duplication (i.e., there is no evidence for gene conversion between paralogous copies).

Our analysis has revealed a higher number of synonymous than nonsynonymous substitutions in all phylogeny branches, suggesting that both proteins are under purifying selection. The strength of natural selection, however, differs in the two genes. Indeed, the lower nucleotide substitution rates of the OS-F gene indicate an overall higher functional constraint (Table 1 and Table 2; Fig 3) and, therefore, would support that the idea that these genes have been functionally diverging since their origin. The detection of the two OS proteins in the D. melanogaster sensillar lymph (MCKENNA et al. 1994 ) also supports the active action of natural selection in the evolution of these genes.

It could be argued, nevertheless, that the different OS substitution rates were caused by local mutation rate differences and not by differential functional constraint. We found that divergence estimates at noncoding positions in the OS-F region are slightly lower than those in OS-E (Table 2). This fact is likely caused by the presence of an extremely conserved DNA fragment in the second intron of OS-F; this conserved region had been already identified in other species as distant as D. virilis (HEKMAT-SCAFE et al. 2000 ), although its functional significance is unknown. Levels of nucleotide divergence across noncoding regions, nevertheless, are distributed more homogeneously than in coding regions. This analysis does not support, therefore, that the different evolutionary rates found between the two OS genes were caused by putative differences in the silent mutation rate along the OS region.

Most D. melanogaster OBP family members (in addition to OS-E and OS-F) are located in gene clusters (GALINDO and SMITH 2001 ). This suggests that gene duplication is an important mechanism to increase diversity in this gene family. In fact, the high sequence divergence among functional members of the family suggests the contribution of positive selection to the rapid evolution and functional diversification of these genes (GALINDO and SMITH 2001 ). It has been also shown that insect OBPs can form dimers in physiological conditions (CAMPANACCI et al. 1999 ; DANTY et al. 1999 ; SANDLER et al. 2000 ). Since the OS-E and OS-F genes are coexpressed in the same cells, the encoded proteins would be able to form homodimers and also heterodimers. Although the formation of such heterodimers has not been demonstrated, this possibility is suggestive since it might be involved in the evolution of these genes: if heterodimers were more efficient than homodimers, selection might have favored the differentiation and maintenance of the two genes. Certainly, information on the quaternary structure of these OS proteins would be relevant to ascertain the role of putative dimers on the molecular evolution of these genes.

We also found an excess of substitutions at the OS-E coding region in the D. melanogaster lineage. This excess, largely due to a high number of nonsynonymous changes, could be explained by either a relaxation of natural selection or the action of positive directional selection. Although a reduction in the selective pressure could increase the fixation probability of weakly selected mutations (OHTA 1973 , OHTA 1992 ), the reduced levels of nucleotide polymorphism in the OS-E coding region would not support this hypothesis. Hence, it seems likely that positive directional selection would have driven these amino acid changes to fixation (see below).

Currently, the 3D structure of the OS-E and OS-F proteins has not been determined. However, two lines of evidence suggest that this structure could be similar to that of the BmPBP obtained by X-ray crystallography. First, the secondary structures predicted for OS-E and OS-F show a remarkable similarity to that previously predicted for the BmPBP, in which the predicted location of -helices has been confirmed by the 3D structure. Second, there are five highly conserved phenylalanines in BmPBP, with two of them (Phe12 and Phe118) involved in the general (i.e., not specific) binding hydrophobic surface (SANDLER et al. 2000 ). In the OS-E and OS-F proteins, there are also five conserved phenylalanines in all surveyed species. Two of these residues (Phe18 and Phe108) are also found in similar positions on the predicted BmPBP hydrophobic surface. Probably the residues constituting the odorant-binding pocket and those involved in the specific binding site of the OS-E and OS-F proteins are in locations equivalent to those described in BmPBP.

A preliminary analysis of OBPs (PLETTNER et al. 2000 ; SANDLER et al. 2000 ; PENG and LEAL 2001 ) has shown that some conservative amino acid changes observed across a number of Lepidoptera species might alter the protein-binding specificity. In particular, replacements among residues such as valine, leucine, isoleucine, or methionine would be responsible for this change in specificity. Remarkably, all amino acid replacements found in the four Drosophila species studied (except the Cys-to-Trp change in D. erecta; Fig 2) are conservative, with low physicochemical distance values (GRANTHAM 1974 ), and are found in protein locations similar to the changes observed in Lepidoptera OBPs. Furthermore, four of the six OS-E amino acid changes fixed in the D. melanogaster lineage also involve valine, leucine, isoleucine, and methionine residues. These replacements are located either in the heterogeneous region (HEKMAT-SCAFE et al. 2000 ) or close to it. Some of these replacements might have been beneficial due to plausible changes in the binding specificity of the protein. Directional positive selection would thus have driven them to fixation.

The action of positive selection should have also left a fingerprint on intraspecific polymorphism and on the ratio of polymorphism to divergence. We have shown that levels of silent nucleotide polymorphism in D. melanogaster were reduced, which is consistent with expectations for a low-recombining genomic region (BEGUN and AQUADRO 1992 ). However, the level of silent nucleotide polymorphism clearly differs between gene copies (Table 2). In fact, the observed number of segregating sites and the results of the Kolmogorof-Smirnov test (P = 0.03) clearly define two regions with distinct levels of variation: a left part including the OS-E and intergenic regions and a right part that includes the OS-F gene region. While the OS-F portion has a low level of variation, as expected in regions of reduced recombination, the level of nucleotide variation at the left part is concordant with the average level of silent variation in D. melanogaster (0.011; MORIYAMA and POWELL 1996 ). The analysis of the ratio of polymorphism to divergence also shows a clear drop in about the middle of the surveyed region. The results of the HKA test are significant only in the OS-F region; in this part, silent nucleotide polymorphism is likely reduced due to the effects of linked selection in concordance with its low recombining environment. In contrast, no significant HKA results were obtained at the left part of the OS region, reflecting a much higher level of intraspecific variation than expected in a region of low recombination. In particular, there was a local peak of variation, with most polymorphisms at noncoding regions (Fig 5).

View larger version (11K): In this window In a new window Download PPT slide   Figure 5. Sliding window of silent polymorphism in D. melanogaster () and silent divergence between D. melanogaster and D. mauritiana (K) for the complete OS region (including the additional upstream region). Step size, 250 silent sites; window size, 1000 silent sites.

To know whether the increase of variation in OS-E is really a peak of variation (i.e., whether it decays also in the upstream region), we sequenced an 2-kb region upstream of the OS-E gene. The level of variation in this 5' flanking region ( = 0.002) was similar to that detected in the OS-F region and therefore lower than that in the OS-E region. The sliding window analysis of the entire region surveyed (6.7 kb) reveals a peak of variation in the OS-E region (Fig 5). Clearly, these results are unlikely not only under the neutral model, but also under simplistic selective models [such as the genetic hitchhiking (MAYNARD SMITH and HAIGH 1974 ) or the background selection (CHARLESWORTH et al. 1993 ) models], or demographic scenarios. Heterogeneity in the recombination rate or in the silent mutation rate along the surveyed region could explain the different pattern of variation observed in the OS-E and OS-F regions. However, it is unrealistic that two closely located genes could differ strongly in their recombination rate. On the other hand, the similar estimates of the silent nucleotide divergence along the OS region also do not support the silent mutation rate heterogeneity hypothesis.

The anomalous levels of nucleotide variation found in OS-E could be explained by the action of some form of balancing selection. Nevertheless, it seems difficult to envisage the target of selection since there is no replacement polymorphism. Yet, selection might act on the RNA stability or on some regulatory elements present at noncoding positions. Furthermore, several other features of the data are consistent with the balancing selection hypothesis. This kind of selection is expected to increase the levels of nucleotide variation and consequently it can skew the frequency spectrum toward intermediate frequencies. The significantly positive Tajima's D and Fu and Li's F values (Table 3) observed in the OS-E region are in agreement with this prediction. Nevertheless, a simple balancing selection model would not easily explain the high number of amino acid changes accumulated in the D. melanogaster lineage. A version of the hitchhiking model, the "traffic" model (KIRBY and STEPHAN 1996 ), might account for both the positive Tajima's D values observed in the OS-E gene and the increase of nucleotide diversity in the target region. This model considers that the fixation of a favorable mutation might be retarded by the fixation process of another closely located favorable mutation. Under this model, neutral variants could increase in frequency near the selected sites, reaching intermediate frequencies. Eventually, recombination could generate haplotypes with a more favorable combination of mutations that would be driven to fixation. The high structure of genetic variation in the OS-E region revealed by the significant linkage disequilibrium values and the Z_A and Wall's B and Q tests is consistent with the traffic hypothesis. In conclusion, the pattern of nucleotide sequence variation in the OS genes is unlikely under the neutral model of molecular evolution and suggests the action of positive natural selection. Further surveys of variation at genes of the olfactory family might contribute to establishing which specific mode of natural selection is acting and thereby to an understanding of the evolutionary meaning and fate of these duplicated genes.

	FOOTNOTES

Sequence data from this article have been deposited in the EMBL/GenBank Data Libraries under accession nos. AJ574644, AJ574762, AJ574763, AJ574764, AJ574765, AJ574766, AJ574767, AJ574768, AJ574769, AJ574770, AJ574771, AJ574772, AJ574773, AJ574774 (D. melanogaster), AJ563750 (D. mauritiana), AJ567753 (D. simulans), and AJ574775, AJ574776 (D. erecta).

	ACKNOWLEDGMENTS

We thank D. Hekmat-Scafe and colleagues for sending us a copy of their manuscript before publication and Serveis Científico-Tècnics, Universitat de Barcelona, for the use of automated sequencing facilities. A.S. was a predoctoral fellow of Universitat de Barcelona. This work was supported by grants PB97-0918 and BMC2001-2906 from Comisión Interdepartamental de Ciencia y Tecnología, Spain, and by grant 2001SGR-00101 from Comissió Interdepartamental de Recerca i Innovació Tecnològica, Spain, to M.A.

Manuscript received April 11, 2003; Accepted for publication July 25, 2003.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ANDOLFATTO, P. and M. PRZEWORSKI, 2000  A genome-wide departure from the standard neutral model in natural populations of Drosophila. Genetics 156:257-268.[Abstract/Free Full Text]

ASHBURNER, M., 1989 Drosophila: A Laboratory Handbook. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

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

BINGHAM, P. M., R. LEVIS, and G. M. RUBIN, 1981  Cloning of DNA sequences from the white locus of D. melanogaster by a novel and general method. Cell 25:693-704.[Medline]

BUCK, L. and R. AXEL, 1991  A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell 65:175-187.[Medline]

CAMPANACCI, V., S. LONGHI, P. NAGNAN-LE MEILLOUR, C. CAMBILLAU, and M. TEGONI, 1999  Recombinant pheromone binding protein 1 from Mamestra brassicae (MbraPBP1). Eur. J. Biochem. 264:707-716.[Medline]

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

CIRERA, S. and M. AGUADÉ, 1997  Evolutionary history of the sex-peptide (Acp70A) gene region in Drosophila melanogaster.. Genetics 147:189-197.[Abstract]

COMERON, J. M., M. KREITMAN, and M. AGUADÉ, 1999  Natural selection on synonymous sites is correlated with gene length and recombination in Drosophila. Genetics 151:239-249.[Abstract/Free Full Text]

DANTY, E., L. BRIAND, C. MICHARD-VANHEE, V. PEREZ, and G. ARNOLD et al., 1999  Cloning and expression of a queen pheromone-binding protein in the honeybee: an olfactory-specific, developmentally regulated protein. J. Neurosci. 17:7468-7475.

FELSENSTEIN, J., 1993 Phylogenetic Inference Package (PHYLIP), Version 3.5. University of Washington, Seattle.

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

GALINDO, K. and D. P. SMITH, 2001  A large family of divergent Drosophila odorant-binding proteins expressed in gustatory and olfactory sensilla. Genetics 159:1059-1072.[Abstract/Free Full Text]

GOLDMAN, N. and Z. YANG, 1994  A codon-based model of nucleotide substitution for protein-coding DNA sequences. Mol. Biol. Evol. 11:725-736.[Abstract]

GRANTHAM, R., 1974  Amino acid difference formula to help explain protein evolution. Science 185:862-864.[Abstract/Free Full Text]

HEKMAT-SCAFE, D. S., R. L. DORIT, and J. R. CARLSON, 2000  Molecular evolution of odorant-binding protein genes OS-E and OS-F in Drosophila. Genetics 155:117-127.[Abstract/Free Full Text]

HEKMAT-SCAFE, D. S., C. R. SCAFE, A. J. MCKINNEY, and M. A. TANOUYE, 2002  Genome-wide analysis of the odorant-binding protein gene family in Drosophila melanogaster.. Genome Res. 9:1357-1369.

HUDSON, R. R., 1983  Properties of a neutral allele model with intragenic recombination. Theor. Popul. Biol. 23:183-201.[Medline]

HUDSON, R. R., 1987  Estimating the recombination parameter of a finite population model without selection. Genet. Res. 50:245-250.[Medline]

HUDSON, R. R., 1990 Gene genealogies and the coalescent process, pp. 1–42 in Oxford Surveys in Evolutionary Biology, Vol. 7, edited by D. FUTUYMA and J. ANTONOVICS. Oxford University Press, New York.

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

JERMIIN, L. S., 1996 K2WuLi, Version 1.0. Australian National University, Canberra, Australia.

JUKES, T. H., and C. R. CANTOR, 1969 Mammalian Protein Metabolism. Academic Press, New York.

KELLY, J., 1997  A test of neutrality based on interlocus associations. Genetics 146:1197-1206.[Abstract]

KINGMAN, J. F. C., 1982  On the genealogy of large populations. J. Appl. Probab. 19A:27-43.

KIRBY, D. A. and W. STEPHAN, 1996  Multi-locus selection and the structure of the white gene of Drosophila melanogaster.. Genetics 144:635-645.[Abstract]

KREITMAN, M. and M. AGUADÉ, 1986  Genetic uniformity in two populations of Drosophila melanogaster as revealed by filter hybridization of four-nucleotide recognizing enzyme digests. Proc. Natl. Acad. Sci. USA 83:3562-3566.[Abstract/Free Full Text]

KRIEGER, M. J. and K. G. ROSS, 2002  Identification of a major gene regulating complex social behavior. Science 295:328-332.[Abstract/Free Full Text]

KUMAR, S., K. TAMURA, I. JAKOBSEN and M. NEI, 2000 MEGA, Molecular Evolutionary Genetics Analysis, Version 2.0.

MADDISON, W. P., and D. R. MADDISON, 1992 MacClade: Analysis of Phylogeny and Character Evolution, Version 3.0. Sinauer Associates, Sunderland, MA.

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

MCDONALD, J. H., 1998  Improved tests for heterogeneity across a region of DNA sequence in the ratio of polymorphism to divergence. Mol. Biol. Evol. 15:377-384.[Abstract]

MCKENNA, M. P., D. S. HEKMAT-SCAFE, P. GAINES, and J. R. CARLSON, 1994  Putative Drosophila pheromone-binding proteins expressed in a subregion of the olfactory system. J. Biol. Chem. 269:16340-16347.[Abstract/Free Full Text]

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

NEI, M., 1987 Molecular Evolutionary Genetics. Columbia University Press, New York.

OHTA, T., 1973  Slightly deleterious substitutions in evolution. Nature 246:96-98.[Medline]

OHTA, T., 1992  The nearly neutral theory of molecular evolution. Annu. Rev. Ecol. Syst. 23:263-286.

PELOSI, P. and R. MAIDA, 1995  Odorant-binding proteins in insects. Comp. Biochem. Physiol. B 111:503-514.[Medline]

PENG, G. H. and W. S. LEAL, 2001  Identification and cloning of a pheromone-binding protein from the oriental beetle, Exomala orientalis.. J. Chem. Ecol. 27:2183-2192.[Medline]

PIKIELNY, C. W., G. HASAN, F. ROUYER, and M. ROSBASH, 1994  Members of a family of Drosophila putative odorant-binding proteins are expressed in different subsets of olfactory hairs. Neuron 12:35-49.[Medline]

PLETTNER, E., J. LAZAR, E. G. PRESTWICH, and G. D. PRESTWICH, 2000  Discrimination of pheromone enantiomers by two pheromone binding proteins from the gypsy moth Lymantria dispar.. Biochemistry 39:8953-8962.[Medline]

RAMOS-ONSINS, S. and M. AGUADÉ, 1998  Molecular evolution of the Cecropin multigene family in Drosophila: functional genes vs. pseudogenes. Genetics 150:157-171.[Abstract/Free Full Text]

ROST, B., 2001  Protein secondary structure prediction continues to rise. J. Struct. Biol. 134:204-218.[Medline]

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

ROZAS, J., M. GULLAUD, G. BLANDIN, and M. AGUADÉ, 2001  DNA variation at the rp49 gene region of Drosophila simulans: evolutionary inferences from an unusual haplotype structure. Genetics 158:1147-1155.[Abstract/Free Full Text]

SAIKI, R. K., D. H. GELFAND, S. STOFFEL, S. J. SCHARF, and R. HIGUCHI et al., 1988  Primer-directed enzymatic amplification of DNA with a thermostable polymerase. Science 239:487-491.[Abstract/Free Full Text]

SAITOU, N. and M. NEI, 1987  The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406-425.[Abstract]

SANDLER, B. H., L. NIKONOVA, W. S. LEAL, and J. CLARDY, 2000  Sexual attraction in the silkworm moth: structure of the pheromone-binding-protein-bombykol complex. Chem. Biol. 7:143-151.[Medline]

SCALONI, A., M. MONTI, S. ANGELI, and P. PELOSI, 1999  Structural analysis and disulfide-bridge pairing of two odorant-binding proteins from Bombyx mori.. Biochem. Biophys. Res. Commun. 266:386-391.[Medline]

SOKAL, R., and F. J. ROHLF, 1995 Biometry. W. H. Freeman, New York.

STEINBRECHT, R. A., 1969 Comparative morphology of olfactory receptors, pp. 3–21 in Olfaction and Taste III, edited by C. PFAFFMANN. Rockefeller University Press, New York.

STEINBRECHT, R. A., 1996  Are odorant-binding proteins involved in odorant discrimination? Chem. Senses 21:719-727.[Abstract/Free Full Text]

STERN, D. and J. MARX, 1999  Making sense of scents. Science 286:703-728.

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

THOMPSON, J. D., D. G. HIGGINS, and T. J. GIBSON, 1994  CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap and weight matrix choice. Nucleic Acids Res. 22:4673-4680.[Abstract/Free Full Text]

VOGT, R. G. and L. M. RIDDIFORD, 1981  Pheromone binding and inactivation by moth antennae. Nature 293:161-163.[Medline]

VOGT, R. G., M. E. ROGERS, M. D. FRANCO, and M. SUN, 2002  A comparative study of odorant-binding protein genes: differential expression of the PBP1-GOBP2 gene cluster in Manduca sexta (Lepidoptera) and the organization of OBP genes in Drosophila melanogaster (Diptera). J. Exp. Biol. 205:719-744.[Abstract/Free Full Text]

VOSSHALL, L. B., H. AMREIN, P. S. MOROZOV, A. RZHETSKY, and R. AXEL, 1999  A spatial map of olfactory receptor expression in the Drosophila antenna. Cell 96:725-736.[Medline]

VOSSHALL, L. B., A. M. WONG, and R. AXEL, 2000  An olfactory sensory map in the fly brain. Cell 102:147-159.[Medline]

WALL, J., 1999  Recombination and the power of statistical test of neutrality. Genet. Res. 74:65-79.

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

WILLETT, C. S., 2000  Evidence for directional selection acting on pheromone-binding proteins in the genus Choristoneura.. Mol. Biol. Evol. 17:553-562.[Abstract/Free Full Text]

WRIGHT, F., 1990  The effective number of codons used by a gene. Gene 87:23-29.[Medline]

WU, C.-I and W.-H. LI, 1985  Evidence for higher rates of nucleotide substitution in rodents than in man. Proc. Natl. Acad. Sci. USA 82:1741-1745.[Abstract/Free Full Text]

YANG, Z., 1997  PAML: a program package for phylogenetic analysis by maximum likelihood. Comput. Appl. Biosci. 15:555-556.

YANG, Z., S. KUMAR, and M. NEI, 1995  A new method of inference of ancestral nucleotide and amino acid sequences. Genetics 141:1641-1650.[Abstract]

This article has been cited by other articles:

				I. Dworkin and C. D. Jones Genetic Changes Accompanying the Evolution of Host Specialization in Drosophila sechellia Genetics, February 1, 2009; 181(2): 721 - 736. [Abstract] [Full Text] [PDF]

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

				A. Takahashi and T. Takano-Shimizu A High-Frequency Null Mutant of an Odorant-Binding Protein Gene, Obp57e, in Drosophila melanogaster Genetics, June 1, 2005; 170(2): 709 - 718. [Abstract] [Full Text] [PDF]