Evidence for Positive Selection on the Floral Scent Gene Isoeugenol-O-methyltransferase
Department of Biological Sciences, Western Michigan University, Kalamazoo, Michiganq/#s(, 百拇医药
Abstractq/#s(, 百拇医药
Isoeugenol-O-methyltransferase (IEMT) is an enzyme involved in the production of the floral volatile compounds methyl eugenol and methyl isoeugenol in Clarkia breweri (Onagraceae). IEMT likely evolved by gene duplication from caffeic acid-O-methyltransferase followed by amino acid divergence, leading to the acquisition of its novel function. To investigate the selective context under which IEMT evolved, maximum likelihood methods that estimate variable dN/dS ratios among lineages, among sites, and among a combination of both lineages and sites were utilized. Statistically significant support was obtained for a hypothesis of positive selection driving the evolution of IEMT since its origin. Subsequent Bayesian analyses identified several sites in IEMT that have experienced positive selection. Most of these positions are in the active site of IEMT and have been shown by site-directed mutagenesis to have large effects on substrate specificity. Although the selective agent is unknown, the adaptive evolution of this gene may have resulted in increased effectiveness of pollinator attraction or herbivore repellence.
Key Words: Clarkia breweri • evolution of floral scent • gene duplication • maximum likelihood • positive selectionl4!2h, 百拇医药
Introductionl4!2h, 百拇医药
Volatile compounds mediate many interactions between organisms, including interplant signaling in response to pathogen infection , plant-parasitoid signaling in response to herbivory , and plant-pollinator communication during flowering. As pollinator attractants, volatiles are important cues that help insects locate flowers and signal the presence of food or mates . The floral scent chemical compositions of hundreds of species have been enumerated; however, only recently has the molecular genetic basis of the biosynthesis of these compounds begun to be elucidated. Numerous studies in Clarkia breweri (Onagraceae) and Antirrhinum majus (Scrophulariaceae) have revealed the identity and numbers of genes involved in scent production and their expression patterns . Within Clarkia, C. breweri is the only species with strongly scented flowers, suggesting that floral fragrance production has recently evolved in this taxon . One possible function of floral scent in C. breweri may be the attraction of pollinating moths as suggested by electroantennogram studies showing that the floral volatiles elicit physiological responses in its primary hawkmoth pollinators . An additional role of floral fragrance may be to repel herbivores, because many of the same volatile compounds produced in flowers are also released from leaves in response to herbivore damage . A novel scent-producing gene, Isoeugenol-O-methyltransferase (IEMT), has been characterized in C. breweri and catalyzes the production of methyleugenol and methylisoeugenol from eugenol and isoeugenol using the methyl-group donor S-adenosyl-L-methionine (SAM). This gene has high levels of sequence similarity to caffeic acid-O-methyltransferase (COMT), which is a ubiquitous enzyme in plants that participates in the biosynthesis of lignin by converting caffeic acid to ferulic acid . IEMT, likely duplicated from COMT, is expressed only in floral tissue of C. breweri, suggesting a potential role in pollinator attraction . Elegant site-directed mutagenesis studies of IEMT and COMT from C. breweri have shown that very few amino acids in a small region of the coding sequences are responsible for substrate specificity . Although much is known about the molecular biology of IEMT, little attention has been paid to the selective context leading to its evolutionary divergence and acquisition of novel function.
The importance of positive Darwinian selection as a process shaping the evolution of protein coding genes has been suggested by numerous recent studies (and references cited therein; ). The study of selection on protein coding genes relies on estimates of the nonsynonymous/synonymous (dN/dS) rate ratio, {omega}gl\@, 百拇医药
. Cases in which {omega}gl\@, 百拇医药
= 1 suggest the genes in question evolve neutrally, whereas purifying, or negative, selection is inferred when {omega}gl\@, 百拇医药
< 1. Evidence for positive selection is obtained only in cases where {omega}gl\@, 百拇医药
> 1 . Recently, maximum likelihood methods have been developed for the detection of positive selection among lineages , among sites within a gene ), or among sites within specific lineages of a phylogeny . In addition to simply inferring whether positive selection has been an important force shaping protein evolution, a recently developed Bayesian method has been applied to identify the most likely sites under positive selection in cases where estimates of {omega}
> 1 . In this paper, these recently developed methods are utilized to investigate the evolution of the floral scent gene, IEMT, in Clarkia breweri.(a!4a6, http://www.100md.com
Materials and Methods(a!4a6, http://www.100md.com
A total of 13 sequences were analyzed from GenBank . These included IEMT from C. breweri and 12 closely related COMT sequences. Maximum likelihood estimates of the COMT/IEMT phylogeny were obtained assuming the HKY85 model of nucleotide substitution using PAUP*4.0 . The ti/tv ratio was estimated from the data and 100 random addition sequences were used during heuristic searches. Bootstrap proportions were obtained from 500 pseudoreplicates with the ti/tv ratio estimated during each search. Neighbor-joining (assuming the HKY 85 model of nucleotide substitution) and unweighted parsimony were also used to estimate phylogenetic relationships among the genes. In all cases, the trees estimated using various models of nucleotide substitution or methods were similar in branching pattern and level of bootstrap support to the ML tree shown in . The strongly supported relationships implied by this tree estimate are congruent with traditional taxonomic treatments and recent phylogenetic studies based on independent data sets . PAML 3.1 was used for all likelihood calculations and estimates of {omega}
for all of the models used assuming the tree shown in .57ito, 百拇医药
fig.ommitted57ito, 百拇医药
FIG. 1. Phylogenetic relationships among COMT sequences from multiple representative rosid families and IEMT from Clarkia breweri. GenBank accession number and family are listed next to each species. Tree was estimated using maximum likelihood (HKY85). Bootstrap proportions (>50) are listed above each branch57ito, 百拇医药
Results and Discussion57ito, 百拇医药
A series of likelihood ratio tests (LRTs) were performed to test hypotheses related to the molecular evolution of IEMT. The first test compared model M0 with the branch-specific model to detect whether selective pressures differ between the branch of interest (IEMT) and the other branches of the phylogeny. The M0 model of assumes a single {omega}57ito, 百拇医药
ratio for all nucleotide sites and lineages. The branch-specific model corresponds to that used by and estimates two {omega}57ito, 百拇医药
ratios for both a "foreground" and "background" lineage. In the case examined here, the foreground lineage is Clarkia breweri IEMT, and all of the COMT sequences comprise the background lineage . This a priori assignment of ratios to different lineages is based on the previous studies of IEMT, suggesting its recent evolution from COMT . The LRT statistic for this comparison was 2(diff. lnL) = 35.00 (P < 0.001, 1 df ). This significantly better fit of the branch-specific model to the data suggests that the IEMT lineage has experienced different levels of selective constraint than the COMT lineages. The fact that {omega}
is 40 times greater in the foreground lineage suggests that there is either relaxed selective constraint or positive selection on IEMT.0ng3%, 百拇医药
The second test compared model M0 with M3 (discrete model), the latter of which was presented in to investigate whether selective pressure varies among sites. Model M3 (with three site classes) assumes that there is some proportion of sites belonging to each of three classes with different {omega}0ng3%, 百拇医药
ratios . In this case, both the proportions and {omega}0ng3%, 百拇医药
ratios for three site classes were estimated from the data. The LRT statistic for this comparison was 2(diff. lnL) = 242.84 (P < 0.001, 4 df ). This significant result suggests that the {omega}0ng3%, 百拇医药
ratio is variable among sites in the COMT and IEMT genes, with {omega}0ng3%, 百拇医药
0 = 0.02 at 60% of the sites, {omega}0ng3%, 百拇医药
1 = 0.20 at 35% of the sites, and {omega}
2 = 0.68 at 5% of the sites. Although {omega}'#u#ds, 百拇医药
appears to be heterogeneous for these sequences, this test provides no evidence for sites under positive selection in these genes, because in no case was {omega}'#u#ds, 百拇医药
estimated to be greater than 1. However, because the same {omega}'#u#ds, 百拇医药
ratios were estimated for all lineages, any difference in selective constraint in IEMT only could not be detected.'#u#ds, 百拇医药
fig.ommitted'#u#ds, 百拇医药
Table 1 Likelihood Scores and Parameter Estimates Assuming Various Models for the Methyltransferase Data.'#u#ds, 百拇医药
The third test involved a comparison of model M3 (two site classes) with the branch-sites model (two ratios and M3 with two site classes) to test whether sites are under positive selection along the IEMT branch in . M3 with two site classes was defined in and estimates the proportion of sites that belong to each of two site classes . The branch-sites model was recently developed in and relies on the a priori specification of foreground and background lineages. Proportions of sites belonging to each of two site classes ({omega}
0 and {omega}9, 百拇医药
1) are estimated for all lineages, and a third site class ({omega}9, 百拇医药
2) is estimated for the foreground lineage alone . A significant LRT for this model comparison would indicate that sites in IEMT are under differential selection and if {omega}9, 百拇医药
2 is greater than 1, this test provides evidence for positive selection . The LRT statistic was 2(diff. lnL) = 50.06 (P < 0.001, 2 df). In the branch-sites model, 59% of sites were estimated with {omega}9, 百拇医药
0 = 0.03, 22% with {omega}9, 百拇医药
1 = 0.31, and 19% with {omega}9, 百拇医药
2 = 1.93. These results strongly suggest that there has been positive selection on some sites in the IEMT lineage since its divergence from COMT.9, 百拇医药
Assuming the branch-sites model, PAML 3.1 was used to calculate the posterior probabilities that a particular site belongs to the site class that has experienced positive selection. As shown in , 30 sites were assigned to the site class with {omega}
= 1.93 (posterior probability > 0.8). Of these substitutions, 11 were in the region of the coding sequence that has been shown to be important for the different substrate specificities of COMT and IEMT based on chimeric COMT-IEMT protein constructs (region B in ) . Furthermore, within this region, all of the sites shown by site-directed mutagenesis to have major effects on substrate discrimination of IEMT had high posterior probabilities . It should also be noted that two of the sites with the highest posterior probabilities (138 and 139) were shown to have some of the largest effects on substrate discrimination.@j[0p}, http://www.100md.com
fig.ommitted@j[0p}, http://www.100md.com
FIG. 2. A, Posterior probabilities (>0.8) that a particular site in IEMT is from the site class {omega}@j[0p}, http://www.100md.com
= 1.93 estimated with PAML. Regions A–C were used in domain swapping experiments between COMT and IEMT . Only region B had an effect on substrate specificity of IEMT. Regions I–III are highly conserved domains among plant methyltransferases and are known to be involved in binding of the methyl group donor SAM . B, Amino acids 131 to 177 representing the active site of IEMT, with posterior probabilities listed for those sites inferred to be under positive selection. The underlined amino acids represent IEMT sites mutated by site-directed mutagenesis studies . The sites were mutated in IEMT to match the amino acids present in the ancestral sequence shown. The percent reduction in the ratio of methyl isoeugenol and ferulic acid produced when using isoeugenol and caffeic acid as substrates for mutated IEMT sequences relative to wild-type IEMT is listed below the amino acids
The branch-sites model that allowed for sites to be under positive selection within IEMT provided a statistically better fit to the data than models without such a parameter. The fact that {omega}'a'%, http://www.100md.com
2 > 1 suggests that some substitutions in this enzyme are adaptive and should have positive effects on fitness. Interestingly, the overall spatial pattern of positively selected amino acid replacements in IEMT is not homogeneous. There are relatively few positively selected sites in the first and last 100 amino acids, whereas there are two to three times more substitutions between positions 131 to 177 and 222 to 274. This heterogeneous pattern of amino acid replacement would be expected for a gene that has evolved by duplication and functional divergence because most of the substitutions occur in the catalytic regions of IEMT. The fact that most of the positively selected sites in IEMT have significant effects on substrate specificity suggests that adaptive evolution of this gene has been driven by a selective agent that is sensitive to the activities of this protein.
If it is assumed that efficient biosynthesis of methyl eugenol and methyl isoeugenol is important for the fitness of Clarkia breweri, then the inference that multiple IEMT sites have evolved by positive selection is supported by the experimental studies on enzyme kinetics . Unlike proteins that have experienced Darwinian selection for increased proportions of charge changing residues (e.g., arginine in eosinophil cationic protein ), it appears that IEMT has been selected for increased substrate specificity. Although the order in which mutations accumulated is unknown, experimental studies do suggest that the many of the sites shown to be under positive selection could have had synergistic effects on the ancestral IEMT substrate specificity. For example,demonstrated that mutations of IEMT at positions 134 and 135 led to an enzyme with a 42% lower ability to discriminate substrates than wild-type . Combined mutations of sites 134 and 135 and 168 and 169 reduced the substrate specificity by 95% compared with wild-type . Finally, it was shown that mutations of 134 and 135, 168 and 169, and 137 and 139 resulted in an enzyme with almost no ability to catalyze the formation of methyl isoeugenol. If the selective agent were sensitive to varying levels of emissions, then the pattern of mutations in the active site of this enzyme is consistent with positive Darwinian selection for increased substrate specificity because the rate of volatile production is probably directly related to substrate preference. Further studies are needed to determine the role of IEMT because the volatiles it produces may attract pollinators, repel herbivores from the flowers, or have some other as of yet unknown function.
In a general context, the results of this study have implications for the study of genes involved in secondary metabolism. Much like IEMT, only a small number of amino acid changes were responsible for novel substrate specificities in another methyltransferase, eugenol-O-methyltransferase (EOMT), relative to its sister gene, chavicol-O-methyltransferase . If positive selection is localized to a few codons in one or a few lineages, the recently developed branch-sites model used here will likely become increasingly important in the study of adaptive evolution of genes involved in secondary metabolism.1n, http://www.100md.com
Acknowledgements1n, http://www.100md.com
Ziheng Yang, DeWayne Shoemaker, and two anonymous reviewers are thanked for comments and suggestions on earlier drafts of this manuscript.1n, http://www.100md.com
Literature Cited1n, http://www.100md.com
Angiosperm Phylogeny Group. 1998. An ordinal classification of the families of flowering plants. Ann. Mo. Bot. Gard. 85:531-553.
Bishop, J. G., A. M. Dean, and T. Mitchell-Olds. 2000. Rapid evolution in plant chitinases: molecular targets of selection in plant-pathogen coevolution. Proc. Natl. Acad. Sci. USA 97: 5322-5327.$u, 百拇医药
Cronquist, A. 1981. An integrated system of classification of flowering plants. Columbia University Press, New York.$u, 百拇医药
Dudareva, N., R. A. Raguso, J. Wang, J. R. Ross, and E. Pichersky. 1998. Floral scent production in Clarkia breweri. Plant Physiol. 116:599-604.$u, 百拇医药
Dudareva, N., L. M. Murfitt, C. J. Mann, N. Gorenstein, N. Kolosova, C. M. Kish, C. Bonham, and K. Wood. 2000. Developmental regulation of methyl benzoate biosynthesis and emission in snapdragon flowers. Plant Cell 12:949-961.$u, 百拇医药
Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791.$u, 百拇医药
Gang, D., N. Lavid, C. Zubieta, F. Chen, T. Beuerle, E. Lewinsohn, J. P. Noel, and E. Pichersky. 2002. Characterization of phenylpropene O-methyltransferases from sweet basil: facile change of substrate specificity and convergent evolution within a plant O-methyltransferase family. Plant Cell 14:505-519.
Goldman, N., and Z. Yang. 1994. A codon-based model of nucleotide substitution for protein-coding DNA sequences. Mol. Biol. Evol. 11:725-736.g%q@4[w, 百拇医药
Kessler, A., and I. T. Baldwin. 2001. Defensive function of herbivore-induced plant volatile emissions in nature. Science 291:2141-2144.g%q@4[w, 百拇医药
Knudsen, J. T., L. Tollsten, and G. Bergstrom. 1993. Floral scents—a checklist of volatile compounds isolated by head-space techniques. Phytochemistry 33:253-280.g%q@4[w, 百拇医药
Kolosova, N., N. Gorenstein, C. M. Kish, and N. Dudareva. 2001. Regulation of circadian methyl benzoate emission in diurnally and nocturnally emitting plants. Plant Cell 13:2333-2347.g%q@4[w, 百拇医药
Nam, K. H., N. Dudareva, and E. Pichersky. 1999. Characterization of benzylalcohol acetyltransferases in scented and non-scented Clarkia species. Plant Cell Physiol. 40:916-923.g%q@4[w, 百拇医药
Nielsen, R., and Z. Yang. 1998. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 148:929-936.
Raguso, R. A., and D. M. Light. 1998. Electroantennogram responses of male Sphinx perelegans hawkmoths to floral and ‘green-leaf volatiles.’ Entomol. Exp. Appl. 86:287-293.^:$fq, 百拇医药
Raguso, R. A., D. M. Light, and E. Pichersky. 1996. Electroantennogram responses of Hyles lineata (Sphingidae: Lepidoptera) to volatile compounds from Clarkia breweri (Onagraceae) and other moth-pollinated flowers. J. Chem. Ecol. 22:1735-1766.^:$fq, 百拇医药
Raguso, R. A., and E. Pichersky. 1995. Floral volatiles from Clarkia breweri and C. concinna (Onagraceae): recent evolution of floral scent and moth pollination. Plant Syst. Evol. 194:55-67.^:$fq, 百拇医药
Ross, J. R., K. H. Nam, J. C. D'Auria, and E. Pichersky. 1999. S-Adenosyl-L-Methionine: salicylic acid carboxyl methyltransferase, an enzyme involved in floral scent production and plant defense, represents a new class of plant methyltransferases. Arch. Biochem. Biophys. 367:9-16.^:$fq, 百拇医药
Shulaev, V., P. Silverman, and I. Raskin. 1997. Airborne signalling by methyl salicylate in plant pathogen resistance. Nature 385:718-721.
Swanson, W. J., Z. Yang, M. F. Wolfner, and C. F. Aquadro. 2001. Positive Darwinian selection drives the evolution of several female reproductive proteins in mammals. Proc. Natl. Acad. Sci. USA 98:2509-2514.{4d|, http://www.100md.com
Swofford, D. L. 1998. PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4.0. Sinauer Associates, Sunderland, Mass.{4d|, http://www.100md.com
Turlings, T. C. J., J. H. Tumlinson, and W. J. Lewis. 1990. Exploitation of herbivore-induced plant odors by host-seeking parasitic wasps. Science 250:1251-1253.{4d|, http://www.100md.com
Wang, J., and E. Pichersky. 1998. Characterization of S-adenosyl-L-methionine: (iso)eugenol O-methyltransferase involved in floral scent production in Clarkia breweri. Arch. Biochem. Biophys. 349:153-160.{4d|, http://www.100md.com
Wang, J., and E. Pichersky. 1999. Identification of specific residues involved in substrate discrimination in two plant O-methyltransferases. Arch. Biochem. Biophys. 368:172-180.{4d|, http://www.100md.com
Wang, J., N. Dudareva, S. Bhakta, R. A. Raguso, and E. Pichersky. 1997. Floral scent production in Clarkia breweri (Onagraceae) II. Localization and developmental modulation of the enzyme S-adenosyl-L-methionine: (iso)eugenol O-methyltransferase and phenylpropanoid emission. Plant Physiol. 114:213-221.
Yang, Z. 1998. Likelihood ratio tests for detecting positive selection and application to primate lysozyme evolution. Mol. Biol. Evol. 15:568-573.psj, 百拇医药
Yang, Z. 2000. Phylogenetic analysis by maximum likelihood (PAML). Version 3.0. University College, London.psj, 百拇医药
Yang, Z., and J. P. Bielawski. 2000. Statistical methods for detecting molecular adaptation. Trends Ecol. Evol. 15:496-503.psj, 百拇医药
Yang, Z., and R. Nielsen. 2002. Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol. Biol. Evol. 19:908-917.psj, 百拇医药
Yang, Z., R. Nielsen, N. Goldman, and A.-M. K. Pedersen. 2000. Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 155:431-449.psj, 百拇医药
Zanotto, P. M. de A., E. G. Kallas, R. F. de Souza, and E. C. Holmes. 1999. Genealogical evidence for positive selection in the nef gene of HIV-1. Genetics 153:1077-1089.psj, 百拇医药
Zhang, J., H. F. Rosenberg, and M. Nei. 1998. Positive Darwinian selection after gene duplication in primate ribonuclease genes. Proc. Natl. Acad. Sci. USA 95:3708-3713.psj, 百拇医药
Accepted for publication September 24, 2002.(Todd J. Barkman)
Abstractq/#s(, 百拇医药
Isoeugenol-O-methyltransferase (IEMT) is an enzyme involved in the production of the floral volatile compounds methyl eugenol and methyl isoeugenol in Clarkia breweri (Onagraceae). IEMT likely evolved by gene duplication from caffeic acid-O-methyltransferase followed by amino acid divergence, leading to the acquisition of its novel function. To investigate the selective context under which IEMT evolved, maximum likelihood methods that estimate variable dN/dS ratios among lineages, among sites, and among a combination of both lineages and sites were utilized. Statistically significant support was obtained for a hypothesis of positive selection driving the evolution of IEMT since its origin. Subsequent Bayesian analyses identified several sites in IEMT that have experienced positive selection. Most of these positions are in the active site of IEMT and have been shown by site-directed mutagenesis to have large effects on substrate specificity. Although the selective agent is unknown, the adaptive evolution of this gene may have resulted in increased effectiveness of pollinator attraction or herbivore repellence.
Key Words: Clarkia breweri • evolution of floral scent • gene duplication • maximum likelihood • positive selectionl4!2h, 百拇医药
Introductionl4!2h, 百拇医药
Volatile compounds mediate many interactions between organisms, including interplant signaling in response to pathogen infection , plant-parasitoid signaling in response to herbivory , and plant-pollinator communication during flowering. As pollinator attractants, volatiles are important cues that help insects locate flowers and signal the presence of food or mates . The floral scent chemical compositions of hundreds of species have been enumerated; however, only recently has the molecular genetic basis of the biosynthesis of these compounds begun to be elucidated. Numerous studies in Clarkia breweri (Onagraceae) and Antirrhinum majus (Scrophulariaceae) have revealed the identity and numbers of genes involved in scent production and their expression patterns . Within Clarkia, C. breweri is the only species with strongly scented flowers, suggesting that floral fragrance production has recently evolved in this taxon . One possible function of floral scent in C. breweri may be the attraction of pollinating moths as suggested by electroantennogram studies showing that the floral volatiles elicit physiological responses in its primary hawkmoth pollinators . An additional role of floral fragrance may be to repel herbivores, because many of the same volatile compounds produced in flowers are also released from leaves in response to herbivore damage . A novel scent-producing gene, Isoeugenol-O-methyltransferase (IEMT), has been characterized in C. breweri and catalyzes the production of methyleugenol and methylisoeugenol from eugenol and isoeugenol using the methyl-group donor S-adenosyl-L-methionine (SAM). This gene has high levels of sequence similarity to caffeic acid-O-methyltransferase (COMT), which is a ubiquitous enzyme in plants that participates in the biosynthesis of lignin by converting caffeic acid to ferulic acid . IEMT, likely duplicated from COMT, is expressed only in floral tissue of C. breweri, suggesting a potential role in pollinator attraction . Elegant site-directed mutagenesis studies of IEMT and COMT from C. breweri have shown that very few amino acids in a small region of the coding sequences are responsible for substrate specificity . Although much is known about the molecular biology of IEMT, little attention has been paid to the selective context leading to its evolutionary divergence and acquisition of novel function.
The importance of positive Darwinian selection as a process shaping the evolution of protein coding genes has been suggested by numerous recent studies (and references cited therein; ). The study of selection on protein coding genes relies on estimates of the nonsynonymous/synonymous (dN/dS) rate ratio, {omega}gl\@, 百拇医药
. Cases in which {omega}gl\@, 百拇医药
= 1 suggest the genes in question evolve neutrally, whereas purifying, or negative, selection is inferred when {omega}gl\@, 百拇医药
< 1. Evidence for positive selection is obtained only in cases where {omega}gl\@, 百拇医药
> 1 . Recently, maximum likelihood methods have been developed for the detection of positive selection among lineages , among sites within a gene ), or among sites within specific lineages of a phylogeny . In addition to simply inferring whether positive selection has been an important force shaping protein evolution, a recently developed Bayesian method has been applied to identify the most likely sites under positive selection in cases where estimates of {omega}
> 1 . In this paper, these recently developed methods are utilized to investigate the evolution of the floral scent gene, IEMT, in Clarkia breweri.(a!4a6, http://www.100md.com
Materials and Methods(a!4a6, http://www.100md.com
A total of 13 sequences were analyzed from GenBank . These included IEMT from C. breweri and 12 closely related COMT sequences. Maximum likelihood estimates of the COMT/IEMT phylogeny were obtained assuming the HKY85 model of nucleotide substitution using PAUP*4.0 . The ti/tv ratio was estimated from the data and 100 random addition sequences were used during heuristic searches. Bootstrap proportions were obtained from 500 pseudoreplicates with the ti/tv ratio estimated during each search. Neighbor-joining (assuming the HKY 85 model of nucleotide substitution) and unweighted parsimony were also used to estimate phylogenetic relationships among the genes. In all cases, the trees estimated using various models of nucleotide substitution or methods were similar in branching pattern and level of bootstrap support to the ML tree shown in . The strongly supported relationships implied by this tree estimate are congruent with traditional taxonomic treatments and recent phylogenetic studies based on independent data sets . PAML 3.1 was used for all likelihood calculations and estimates of {omega}
for all of the models used assuming the tree shown in .57ito, 百拇医药
fig.ommitted57ito, 百拇医药
FIG. 1. Phylogenetic relationships among COMT sequences from multiple representative rosid families and IEMT from Clarkia breweri. GenBank accession number and family are listed next to each species. Tree was estimated using maximum likelihood (HKY85). Bootstrap proportions (>50) are listed above each branch57ito, 百拇医药
Results and Discussion57ito, 百拇医药
A series of likelihood ratio tests (LRTs) were performed to test hypotheses related to the molecular evolution of IEMT. The first test compared model M0 with the branch-specific model to detect whether selective pressures differ between the branch of interest (IEMT) and the other branches of the phylogeny. The M0 model of assumes a single {omega}57ito, 百拇医药
ratio for all nucleotide sites and lineages. The branch-specific model corresponds to that used by and estimates two {omega}57ito, 百拇医药
ratios for both a "foreground" and "background" lineage. In the case examined here, the foreground lineage is Clarkia breweri IEMT, and all of the COMT sequences comprise the background lineage . This a priori assignment of ratios to different lineages is based on the previous studies of IEMT, suggesting its recent evolution from COMT . The LRT statistic for this comparison was 2(diff. lnL) = 35.00 (P < 0.001, 1 df ). This significantly better fit of the branch-specific model to the data suggests that the IEMT lineage has experienced different levels of selective constraint than the COMT lineages. The fact that {omega}
is 40 times greater in the foreground lineage suggests that there is either relaxed selective constraint or positive selection on IEMT.0ng3%, 百拇医药
The second test compared model M0 with M3 (discrete model), the latter of which was presented in to investigate whether selective pressure varies among sites. Model M3 (with three site classes) assumes that there is some proportion of sites belonging to each of three classes with different {omega}0ng3%, 百拇医药
ratios . In this case, both the proportions and {omega}0ng3%, 百拇医药
ratios for three site classes were estimated from the data. The LRT statistic for this comparison was 2(diff. lnL) = 242.84 (P < 0.001, 4 df ). This significant result suggests that the {omega}0ng3%, 百拇医药
ratio is variable among sites in the COMT and IEMT genes, with {omega}0ng3%, 百拇医药
0 = 0.02 at 60% of the sites, {omega}0ng3%, 百拇医药
1 = 0.20 at 35% of the sites, and {omega}
2 = 0.68 at 5% of the sites. Although {omega}'#u#ds, 百拇医药
appears to be heterogeneous for these sequences, this test provides no evidence for sites under positive selection in these genes, because in no case was {omega}'#u#ds, 百拇医药
estimated to be greater than 1. However, because the same {omega}'#u#ds, 百拇医药
ratios were estimated for all lineages, any difference in selective constraint in IEMT only could not be detected.'#u#ds, 百拇医药
fig.ommitted'#u#ds, 百拇医药
Table 1 Likelihood Scores and Parameter Estimates Assuming Various Models for the Methyltransferase Data.'#u#ds, 百拇医药
The third test involved a comparison of model M3 (two site classes) with the branch-sites model (two ratios and M3 with two site classes) to test whether sites are under positive selection along the IEMT branch in . M3 with two site classes was defined in and estimates the proportion of sites that belong to each of two site classes . The branch-sites model was recently developed in and relies on the a priori specification of foreground and background lineages. Proportions of sites belonging to each of two site classes ({omega}
0 and {omega}9, 百拇医药
1) are estimated for all lineages, and a third site class ({omega}9, 百拇医药
2) is estimated for the foreground lineage alone . A significant LRT for this model comparison would indicate that sites in IEMT are under differential selection and if {omega}9, 百拇医药
2 is greater than 1, this test provides evidence for positive selection . The LRT statistic was 2(diff. lnL) = 50.06 (P < 0.001, 2 df). In the branch-sites model, 59% of sites were estimated with {omega}9, 百拇医药
0 = 0.03, 22% with {omega}9, 百拇医药
1 = 0.31, and 19% with {omega}9, 百拇医药
2 = 1.93. These results strongly suggest that there has been positive selection on some sites in the IEMT lineage since its divergence from COMT.9, 百拇医药
Assuming the branch-sites model, PAML 3.1 was used to calculate the posterior probabilities that a particular site belongs to the site class that has experienced positive selection. As shown in , 30 sites were assigned to the site class with {omega}
= 1.93 (posterior probability > 0.8). Of these substitutions, 11 were in the region of the coding sequence that has been shown to be important for the different substrate specificities of COMT and IEMT based on chimeric COMT-IEMT protein constructs (region B in ) . Furthermore, within this region, all of the sites shown by site-directed mutagenesis to have major effects on substrate discrimination of IEMT had high posterior probabilities . It should also be noted that two of the sites with the highest posterior probabilities (138 and 139) were shown to have some of the largest effects on substrate discrimination.@j[0p}, http://www.100md.com
fig.ommitted@j[0p}, http://www.100md.com
FIG. 2. A, Posterior probabilities (>0.8) that a particular site in IEMT is from the site class {omega}@j[0p}, http://www.100md.com
= 1.93 estimated with PAML. Regions A–C were used in domain swapping experiments between COMT and IEMT . Only region B had an effect on substrate specificity of IEMT. Regions I–III are highly conserved domains among plant methyltransferases and are known to be involved in binding of the methyl group donor SAM . B, Amino acids 131 to 177 representing the active site of IEMT, with posterior probabilities listed for those sites inferred to be under positive selection. The underlined amino acids represent IEMT sites mutated by site-directed mutagenesis studies . The sites were mutated in IEMT to match the amino acids present in the ancestral sequence shown. The percent reduction in the ratio of methyl isoeugenol and ferulic acid produced when using isoeugenol and caffeic acid as substrates for mutated IEMT sequences relative to wild-type IEMT is listed below the amino acids
The branch-sites model that allowed for sites to be under positive selection within IEMT provided a statistically better fit to the data than models without such a parameter. The fact that {omega}'a'%, http://www.100md.com
2 > 1 suggests that some substitutions in this enzyme are adaptive and should have positive effects on fitness. Interestingly, the overall spatial pattern of positively selected amino acid replacements in IEMT is not homogeneous. There are relatively few positively selected sites in the first and last 100 amino acids, whereas there are two to three times more substitutions between positions 131 to 177 and 222 to 274. This heterogeneous pattern of amino acid replacement would be expected for a gene that has evolved by duplication and functional divergence because most of the substitutions occur in the catalytic regions of IEMT. The fact that most of the positively selected sites in IEMT have significant effects on substrate specificity suggests that adaptive evolution of this gene has been driven by a selective agent that is sensitive to the activities of this protein.
If it is assumed that efficient biosynthesis of methyl eugenol and methyl isoeugenol is important for the fitness of Clarkia breweri, then the inference that multiple IEMT sites have evolved by positive selection is supported by the experimental studies on enzyme kinetics . Unlike proteins that have experienced Darwinian selection for increased proportions of charge changing residues (e.g., arginine in eosinophil cationic protein ), it appears that IEMT has been selected for increased substrate specificity. Although the order in which mutations accumulated is unknown, experimental studies do suggest that the many of the sites shown to be under positive selection could have had synergistic effects on the ancestral IEMT substrate specificity. For example,demonstrated that mutations of IEMT at positions 134 and 135 led to an enzyme with a 42% lower ability to discriminate substrates than wild-type . Combined mutations of sites 134 and 135 and 168 and 169 reduced the substrate specificity by 95% compared with wild-type . Finally, it was shown that mutations of 134 and 135, 168 and 169, and 137 and 139 resulted in an enzyme with almost no ability to catalyze the formation of methyl isoeugenol. If the selective agent were sensitive to varying levels of emissions, then the pattern of mutations in the active site of this enzyme is consistent with positive Darwinian selection for increased substrate specificity because the rate of volatile production is probably directly related to substrate preference. Further studies are needed to determine the role of IEMT because the volatiles it produces may attract pollinators, repel herbivores from the flowers, or have some other as of yet unknown function.
In a general context, the results of this study have implications for the study of genes involved in secondary metabolism. Much like IEMT, only a small number of amino acid changes were responsible for novel substrate specificities in another methyltransferase, eugenol-O-methyltransferase (EOMT), relative to its sister gene, chavicol-O-methyltransferase . If positive selection is localized to a few codons in one or a few lineages, the recently developed branch-sites model used here will likely become increasingly important in the study of adaptive evolution of genes involved in secondary metabolism.1n, http://www.100md.com
Acknowledgements1n, http://www.100md.com
Ziheng Yang, DeWayne Shoemaker, and two anonymous reviewers are thanked for comments and suggestions on earlier drafts of this manuscript.1n, http://www.100md.com
Literature Cited1n, http://www.100md.com
Angiosperm Phylogeny Group. 1998. An ordinal classification of the families of flowering plants. Ann. Mo. Bot. Gard. 85:531-553.
Bishop, J. G., A. M. Dean, and T. Mitchell-Olds. 2000. Rapid evolution in plant chitinases: molecular targets of selection in plant-pathogen coevolution. Proc. Natl. Acad. Sci. USA 97: 5322-5327.$u, 百拇医药
Cronquist, A. 1981. An integrated system of classification of flowering plants. Columbia University Press, New York.$u, 百拇医药
Dudareva, N., R. A. Raguso, J. Wang, J. R. Ross, and E. Pichersky. 1998. Floral scent production in Clarkia breweri. Plant Physiol. 116:599-604.$u, 百拇医药
Dudareva, N., L. M. Murfitt, C. J. Mann, N. Gorenstein, N. Kolosova, C. M. Kish, C. Bonham, and K. Wood. 2000. Developmental regulation of methyl benzoate biosynthesis and emission in snapdragon flowers. Plant Cell 12:949-961.$u, 百拇医药
Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:783-791.$u, 百拇医药
Gang, D., N. Lavid, C. Zubieta, F. Chen, T. Beuerle, E. Lewinsohn, J. P. Noel, and E. Pichersky. 2002. Characterization of phenylpropene O-methyltransferases from sweet basil: facile change of substrate specificity and convergent evolution within a plant O-methyltransferase family. Plant Cell 14:505-519.
Goldman, N., and Z. Yang. 1994. A codon-based model of nucleotide substitution for protein-coding DNA sequences. Mol. Biol. Evol. 11:725-736.g%q@4[w, 百拇医药
Kessler, A., and I. T. Baldwin. 2001. Defensive function of herbivore-induced plant volatile emissions in nature. Science 291:2141-2144.g%q@4[w, 百拇医药
Knudsen, J. T., L. Tollsten, and G. Bergstrom. 1993. Floral scents—a checklist of volatile compounds isolated by head-space techniques. Phytochemistry 33:253-280.g%q@4[w, 百拇医药
Kolosova, N., N. Gorenstein, C. M. Kish, and N. Dudareva. 2001. Regulation of circadian methyl benzoate emission in diurnally and nocturnally emitting plants. Plant Cell 13:2333-2347.g%q@4[w, 百拇医药
Nam, K. H., N. Dudareva, and E. Pichersky. 1999. Characterization of benzylalcohol acetyltransferases in scented and non-scented Clarkia species. Plant Cell Physiol. 40:916-923.g%q@4[w, 百拇医药
Nielsen, R., and Z. Yang. 1998. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics 148:929-936.
Raguso, R. A., and D. M. Light. 1998. Electroantennogram responses of male Sphinx perelegans hawkmoths to floral and ‘green-leaf volatiles.’ Entomol. Exp. Appl. 86:287-293.^:$fq, 百拇医药
Raguso, R. A., D. M. Light, and E. Pichersky. 1996. Electroantennogram responses of Hyles lineata (Sphingidae: Lepidoptera) to volatile compounds from Clarkia breweri (Onagraceae) and other moth-pollinated flowers. J. Chem. Ecol. 22:1735-1766.^:$fq, 百拇医药
Raguso, R. A., and E. Pichersky. 1995. Floral volatiles from Clarkia breweri and C. concinna (Onagraceae): recent evolution of floral scent and moth pollination. Plant Syst. Evol. 194:55-67.^:$fq, 百拇医药
Ross, J. R., K. H. Nam, J. C. D'Auria, and E. Pichersky. 1999. S-Adenosyl-L-Methionine: salicylic acid carboxyl methyltransferase, an enzyme involved in floral scent production and plant defense, represents a new class of plant methyltransferases. Arch. Biochem. Biophys. 367:9-16.^:$fq, 百拇医药
Shulaev, V., P. Silverman, and I. Raskin. 1997. Airborne signalling by methyl salicylate in plant pathogen resistance. Nature 385:718-721.
Swanson, W. J., Z. Yang, M. F. Wolfner, and C. F. Aquadro. 2001. Positive Darwinian selection drives the evolution of several female reproductive proteins in mammals. Proc. Natl. Acad. Sci. USA 98:2509-2514.{4d|, http://www.100md.com
Swofford, D. L. 1998. PAUP*: phylogenetic analysis using parsimony (*and other methods). Version 4.0. Sinauer Associates, Sunderland, Mass.{4d|, http://www.100md.com
Turlings, T. C. J., J. H. Tumlinson, and W. J. Lewis. 1990. Exploitation of herbivore-induced plant odors by host-seeking parasitic wasps. Science 250:1251-1253.{4d|, http://www.100md.com
Wang, J., and E. Pichersky. 1998. Characterization of S-adenosyl-L-methionine: (iso)eugenol O-methyltransferase involved in floral scent production in Clarkia breweri. Arch. Biochem. Biophys. 349:153-160.{4d|, http://www.100md.com
Wang, J., and E. Pichersky. 1999. Identification of specific residues involved in substrate discrimination in two plant O-methyltransferases. Arch. Biochem. Biophys. 368:172-180.{4d|, http://www.100md.com
Wang, J., N. Dudareva, S. Bhakta, R. A. Raguso, and E. Pichersky. 1997. Floral scent production in Clarkia breweri (Onagraceae) II. Localization and developmental modulation of the enzyme S-adenosyl-L-methionine: (iso)eugenol O-methyltransferase and phenylpropanoid emission. Plant Physiol. 114:213-221.
Yang, Z. 1998. Likelihood ratio tests for detecting positive selection and application to primate lysozyme evolution. Mol. Biol. Evol. 15:568-573.psj, 百拇医药
Yang, Z. 2000. Phylogenetic analysis by maximum likelihood (PAML). Version 3.0. University College, London.psj, 百拇医药
Yang, Z., and J. P. Bielawski. 2000. Statistical methods for detecting molecular adaptation. Trends Ecol. Evol. 15:496-503.psj, 百拇医药
Yang, Z., and R. Nielsen. 2002. Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol. Biol. Evol. 19:908-917.psj, 百拇医药
Yang, Z., R. Nielsen, N. Goldman, and A.-M. K. Pedersen. 2000. Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 155:431-449.psj, 百拇医药
Zanotto, P. M. de A., E. G. Kallas, R. F. de Souza, and E. C. Holmes. 1999. Genealogical evidence for positive selection in the nef gene of HIV-1. Genetics 153:1077-1089.psj, 百拇医药
Zhang, J., H. F. Rosenberg, and M. Nei. 1998. Positive Darwinian selection after gene duplication in primate ribonuclease genes. Proc. Natl. Acad. Sci. USA 95:3708-3713.psj, 百拇医药
Accepted for publication September 24, 2002.(Todd J. Barkman)