This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 19/3/779    most recent tpc.106.048694v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via ISI Web of Science (6) Citing Articles via Google Scholar Google Scholar Articles by Hoballah, M. E. Articles by Kuhlemeier, C. Search for Related Content PubMed PubMed Citation Articles by Hoballah, M. E. Articles by Kuhlemeier, C. Agricola Articles by Hoballah, M. E. Articles by Kuhlemeier, C.

Single Gene–Mediated Shift in Pollinator Attraction in Petunia^[W]

Institute of Plant Sciences, University of Bern, 3013 Bern, Switzerland

⁴ To whom correspondence should be addressed. E-mail cris.kuhlemeier{at}ips.unibe.ch; fax 41-31-631-4942.

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Animal-mediated pollination is essential in plant reproductive biology and is often associated with pollination syndromes, sets of floral traits, such as color, scent, shape, or nectar content. Selection by pollinators is often considered a key factor in floral evolution and plant speciation. Our aim is the identification and characterization of the genetic changes that caused the evolution of divergent pollination syndromes in closely related plant species. We focus on ANTHOCYANIN2 (AN2), a well-defined myb-type transcription factor that is a major determinant of flower color variation between Petunia integrifolia and Petunia axillaris. Analysis of sequence variation in AN2 in wild P. axillaris accessions showed that loss-of-function alleles arose at least five times independently. DNA sequence analysis was complemented by functional assays for pollinator preference using genetic introgressions and transgenics. These results show that AN2 is a major determinant of pollinator attraction. Therefore, changes in a single gene cause a major shift in pollination biology and support the notion that the adaptation of a flowering plant to a new pollinator type may involve a limited number of genes of large effect. Gene identification and analysis of molecular evolution in combination with behavioral and ecological studies can ultimately unravel the evolutionary genetics of pollination syndromes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Animals play an important role in the sexual reproduction of many flowering plants as vectors for pollen transfer between flowers. Specific sets of floral traits are frequently found to be associated with particular groups of pollinators. Such pollination syndromes have arisen independently in closely related plant species in many angiosperm families (Faegri and van der Pijl, 1979; Proctor et al., 1996; Fenster et al., 2004). The relationship between plants and their pollinators has been compared with a lock and key, where the lock has been fitted to a preexisting key (Grant and Grant, 1965). Although the evidence for adaptation between flowering plant species and their pollinators seems overwhelming (Fenster et al., 2004), aspects of the pollination syndrome concept have been criticized (Herrera, 1996; Waser et al., 1996; Johnson and Steiner, 2000; Thompson, 2001). One may describe plant pollinator systems as a continuum ranging from extreme adaptation of one plant species to a single pollinator, as found in some orchid species (Schiestl et al., 1999), via pollination syndromes, where a plant species is primarily pollinated by a morphologically or behaviorally defined class of pollinators (i.e., long-tongued nocturnal hawk moths) and the plant species features corresponding floral adaptations, to more generalized flower types that are visited by multiple pollinator types.

Many closely related plant species occur in sympatry or parapatry in nature without obvious gene flow among them. However, in many cases where isolation mechanisms, such as pollinator preference, limit interbreeding of sibling species in nature, hand pollination is often easily achieved in the laboratory. Such cross-compatibility provides an opportunity to identify the genetic basis of species-specific adaptations (Bradshaw et al., 1995, 1998; Fishman et al., 2002; Hodges et al., 2002; Stuurman et al., 2004; Galliot et al., 2006), to study their evolutionary history, and to assess the role of adaptive mutations on the genetic isolation between species (Schemske and Bradshaw, 1999; Bradshaw and Schemske, 2003).

Many studies have demonstrated the importance of flower color for pollinator preference (Waser and Price, 1981; Niovi Jones and Reithel, 2001; Bradshaw and Schemske, 2003). Recent studies report that natural variation in flower color can be due to differences both in the activity and expression of biosynthetic enzymes and their transacting regulators (Durbin et al., 2003; Zufall and Rausher, 2004; Schwinn et al., 2006; Whittall et al., 2006).

The genus Petunia comprises species with distinct hawk moth and bee pollination syndromes, respectively (Wijsman, 1983; Ando et al., 2001). In combination with the availability of genetic and molecular tools (Gerats and Vandenbussche, 2005), this makes Petunia an ideal model system to study the molecular genetic basis of pollination syndromes (Stuurman et al., 2004). Petunia axillaris (Lam.; Britton, Sterns, and Poggenb.) has white flowers with long and narrow corolla tubes, emits large amounts of volatiles at night, and contains large volumes of floral nectar (Ando et al., 1995a; Stuurman et al., 2004; Hoballah et al., 2005; Oyama-Okubo et al., 2005). Pollination primarily by nocturnal hawk moths has been observed in the field (Ando et al., 2001). Petunia integrifolia (Hook.; Schniz and Thell) has a violet-reddish flower with a short wide corolla tube containing low amounts of nectar and emitting small amounts of volatiles (Ando et al., 1995b; Stuurman et al., 2004; Hoballah et al., 2005); it is primarily visited by solitary bees (Ando et al., 2001). Although P. axillaris and P. integrifolia can be crossed in the laboratory, no hybrids were detected in sympatric populations (Ando et al., 2001). Recent DNA sequence analyses confirm that the species are closely related and that interspecific relationships are poorly resolved (Ando et al., 2005; Kulcheski et al., 2006). Further work with more polymorphic markers may soon resolve ambiguities and provide a robust phylogeny of the genus.

Quantitative trait loci analysis indicated that pollination syndrome–associated traits, such as morphology, scent, and nectar production, in P. axillaris and P. integrifolia are based on a few larger and some smaller quantitative trait loci (Stuurman et al., 2004; Galliot et al., 2006). However, none of the genes controlling natural variation in morphology, scent, and nectar production have been identified yet. By contrast, flower color pathways have been worked out in great detail (Holton and Cornish, 1995; Koes et al., 2005), and loci that specify color polymorphism between P. axillaris and P. integrifolia have been identified (Wijsman, 1983). This makes it feasible to study the genetics, ecology, and evolution of flower color as a pollination syndrome trait.

Anthocyanin biosynthesis of the corolla limb of these species is under the control of transcription factor ANTHOCYANIN2 (AN2) (Wijsman, 1983; Quattrocchio et al., 1999). The absence of anthocyanin pigmentation in the P. axillaris corolla limb correlates with loss of AN2 function through inactivating mutations in the AN2 coding region (Quattrocchio et al., 1999). This suggests that anthocyanin-colored petals, as in P. integrifolia, represent the ancestral and white flowers of the derived phenotype. This trend in flower color evolution was also suggested for angiosperms in general (Harborne, 1982).

Here, we analyze the role of AN2 in the shift of pollinator preference between P. integrifolia and P. axillaris. First, we provide a detailed analysis of the molecular evolution of the AN2 gene using a large collection of freshly collected wild material. Second, by employing genetic substitution and transgenic complementation coupled with pollinator choice experiments, we demonstrate that a color change due to a single-gene polymorphism causes major adaptive shifts in pollinator preference.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Analysis of Molecular Evolution
In a previous study (Quattrocchio et al., 1999), it was shown that in a few P. hybrida and three P. axillaris accessions, AN2 alleles are nonfunctional because of nonsense or frameshift mutations in the coding region. To ascertain that these mutations were not due to prolonged cultivation in botanical gardens, we collected new field material. Sequence data were obtained from the entire coding region of AN2 cDNAs derived from the corollas of 35 P. axillaris and P. integrifolia accessions (Table 1 ). Phylogenetic analysis showed that the species alleles fall into two divergent, monophyletic clades (Figure 1A ). While all P. integrifolia alleles encode putatively functional alleles, all P. axillaris alleles carry one of six frameshift or nonsense mutations (Figure 1A) at five different positions. In addition to the three previously found mutations (Quattrocchio et al., 1999), we found three additional and relatively recent loss-of-function mutations (Figure 1A). Because these inactivations were found in freshly collected field material, we conclude that they occurred under natural conditions and were not the result of prolonged cultivation.

View larger version (26K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. Molecular Analysis of AN2. (A) Bayesian tree of AN2 cDNA sequences. Five independent loss-of-function mutations (A to E) in P. axillaris (Pax) are color coded. Accession abbreviations are given in Table 1. Numbers above branches indicate Bayesian support values/maximum likelihood/maximum parsimony bootstrap support. (B) AN2 mRNA is present in all P. axillaris and P. integrifolia accessions. Semiquantitative RT-PCR of AN2 in top and middle panels (35 and 23 amplification cycles, respectively) and ACTIN1 (ACT1) control (27 cycles) in bottom panel in P. axillaris and P. integrifolia petals. All AN2 expression values were normalized to ACTIN1. The average of the P. integrifolia levels was taken as 100%.

The occurrence of multiple independent frameshift or stop codon alleles suggests that loss of color arose multiple times. The alternative scenario, however, is that it is a secondary event following selection for downregulation of the gene through inactivation of the promoter. If promoter inactivation were the primary event, one would expect that all P. axillaris alleles have very low expression levels. By contrast, if frameshift or stop codons were the primary event, then expression levels in at least some P. axillaris might be comparable to the putatively higher expression levels expected in the anthocyanin-producing flowers of P. integrifolia. To test these hypotheses, we performed semiquantitative PCR on cDNA of P. axillaris and P. integrifolia accessions. The results indicated that AN2 is expressed at high levels in all P. integrifolia accessions, while it is expressed at variable levels in P. axillaris accessions (Figure 1B). Six of the 17 P. axillaris accessions have an mRNA level that is at least 50% of the average P. integrifolia level (Figure 1B). Since the an2 mutation is recessive, we assume that 50% would be sufficient to give a wild-type color if the gene were otherwise functional. The fact that AN2 mRNA is present in all P. axillaris accessions and in some at comparable levels to P. integrifolia suggests that AN2 was first inactivated by mutations in the coding region and that variation in expression levels is secondary.

Analysis of molecular variation may provide support for past selection on a locus. Overall, Tajima's D (Tajima, 1989) for P. axillaris (D = –1.325) is only slightly more negative than the value for a linked reference gene (D = –1.012) and hence provides no strong evidence for directional selection (see Supplemental Figure 1 online). The McDonald-Kreitman test (McDonald and Kreitman, 1991), capable of detecting directional protein evolution, for the entire coding region of AN2 in P. integrifolia and P. axillaris indicated an excess of nonfixed, nonsynonymous mutations (G = 9.919; P = 0.00164), consistent with relaxed selection pressure in P. axillaris after alleles had been fixed for loss-of-function mutations. The distribution of mutations along the coding region of AN2 is not homogenous with an excess of fixed nonsynonymous mutations in the C terminus of the gene (² test; P < 0.01), while there is no significant difference between both regions for synonymous substitutions (² test, P > 0.1; see Supplemental Table 1 online), suggesting that amino acid substitutions in the C terminus have most likely been fixed before selection of loss-of-function alleles occurred. The fixed amino acid differences, however, did not cause a detectable change in protein function in transient expression assays (Quattrocchio et al., 1999).

Although this analysis of molecular evolution did not provide a clear indication of directional selection, the complete absence of functional alleles argues for selection as a likely factor driving the establishment of loss-of-function alleles in P. axillaris. The occurrence of multiple loss-of-function alleles could suggest that the selective advantage of these alleles was not so large that a single adaptive allele would have become fixed throughout the species populations. In geographically structured populations of Arabidopsis thaliana, multiple adaptive loss-of-function alleles have been well documented (Johanson et al., 2000; Michaels et al., 2003; Shindo et al., 2005). We did not observe comparable phylogeographic structure of AN2 alleles in P. axillaris (see Supplemental Figure 2 online). Hence, we suggest that loss-of-function alleles with a selective advantage may arise frequently and before selection fixes a single adaptive allele.

Pollinator Observations in the Wild
To determine which pollinators visit the flowers of P. axillaris and P. integrifolia in the natural habitat, we observed two populations of each species at different locations in Uruguay. P. axillaris flowers received some bee and beetle visits during the day and were exclusively visited by hawk moths at night (Figure 2A ). P. integrifolia flowers did not receive nocturnal visits and were visited by bees and diurnal butterflies during the day (Figure 2B). Thus, in nature, the two species attracted a wider spectrum of visitors in addition to the expected pollinator types. We identified several hawk moth, beetle, and bee species that visited wild Petunia species (see Supplemental Table 2 online). The most common bees visiting P. integrifolia were Leioproctus and Calliopsis, which were not observed on P. axillaris flowers. Pollinator exclusion experiments showed that both diurnal and nocturnal flower visitors can pollinate P. axillaris plants (Figure 2C), whereas P. integrifolia is exclusively pollinated by diurnal insects (Figure 2D).

View larger version (25K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Pollinator Visits to Petunia Species in Natural Habitats. (A) Mean (±SE) number of visits of pollinators to P. axillaris per flower per period (1, 8 AM to 12 AM; 2, 2 PM to 6 PM; 3, 8:30 PM to 9:30 PM; 4, 10 PM to 2 AM) in the localities José Ignacio and Carmelo (Uruguay) (n = 7). (B) Mean (±SE) number of visits of pollinators to P. integrifolia per flower per period (1, 10 AM to 2 PM; 2, 9 PM to 11 PM) in the localities Puerto Viejo and Rivera (Uruguay) (n = 7). Note that P. integrifolia was not visited by pollinators at night. (C) Pollinator exclusion shows that diurnal and nocturnal pollinators effectively pollinate P. axillaris in José Ignacio and Carmelo. There was no difference in capsule formation (mean ± SE) in day- or night-pollinated plants (Mann-Whitney test, Z = –0.711, P = 0.477 for José Ignacio; Z = –1.38, P = 0.167 for Carmelo, n = 4). (D) P. integrifolia is exclusively day-pollinated. Mean (±SE) percentage of capsule formation in Puerto Viejo and Rivera in day-pollinated (no capsule formation) and night-pollinated P. integrifolia plants (n = 4).

AN2 Introgressions Affect Pollinator Choice in the Field
To test the effect of floral anthocyanin pigmentation, controlled by AN2, on pollinator preference, we tested pollinator preference for two recombinant inbred lines derived from introgressing P. axillaris (Table 1, PaxS7) chromosome segments into a P. hybrida W138 background (Stuurman et al., 2004). Introgression line (IL) WP117 containing the inactive P. axillaris an2^– allele has very low amounts of anthocyanin (Figure 3A ), while WP119 carrying a P. integrifolia AN2 allele produces anthocyanins (Figure 3B). Both ILs are morphologically intermediate between both species (Figures 3A and 3B), produce virtually no odor (data not shown), and do not differ in nectar volume (Figure 3D) but differ in reflectance (Figure 3C).

View larger version (54K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 3. Petunia ILs WP117 and WP119 Are Polymorphic for the AN2 Locus and Affect Pollinator Behavior. (A) and (B) Petunia IL WP117 (A) carries a nonfunctional an2 allele and WP119 (B) carries an active AN2 allele (Stuurman et al., 2004). (C) Mean ± SE of percentage of reflectance relative to a white standard for different wavelengths. (D) Mean ± SE nectar content does not significantly (ns) differ between lines (Mann-Whitney test, Z = –0.662, P = 0.508, n = 6). (E) Mean (±SE) number of visits of diurnal butterflies in the wild habitat in José Ignacio (Uruguay) is higher for the line WP119 than for WP117 (n = 8). (F) Mean (±SE) number of visits of hymenopteran (Minusio, Switzerland) is higher for the line WP119 than for WP117 (n = 5). Asterisk over bars indicates the significance level of the statistical test (P = 0.01).

Field experiments with European hymenopteran pollinators in southern Switzerland showed a clear preference of for WP119 flowers compared with WP117 flowers. Hymenopteran landing rate per flower per hour was significantly higher on WP119 than on WP117 flowers (Figure 3F; Wilcoxon signed ranks test, Z = –2.023, P = 0.043). Hymenopterans observed landing on these flowers were bumblebees, common honeybees, and smaller hymenopterans. Thus, the field experiments show that diurnal pollinators prefer colored flowers. However, hawk moth preference could not be assessed.

Innate Pollinator Preference under Controlled Conditions
While field observations provide insight into the full complexity of pollination under natural conditions, they are less suitable for routinely assessing the effects of large numbers of genetic variants. Therefore, we set up a system with reduced complexity that tests the innate preference of two model pollinators under controlled greenhouse conditions. We first tested the system with the two wild Petunia species. In a scent-saturated greenhouse, the hawk moth Manduca sexta, a natural pollinator of P. axillaris (Figure 4A ) in Uruguay, showed significant preference for P. axillaris in first-choice feedings (26 versus 3, respectively; binomial test P < 0.0001) and number of feedings per time (Figure 4C; Wilcoxon signed-ranks test, Z = –4.579, P < 0.0001), while the bumblebee Bombus terrestris, as a representative hymenopteran pollinator, showed a significantly higher landing rate on P. integrifolia than on P. axillaris flowers (Figures 4B and 4D; Wilcoxon signed-ranks test, Z = –3.061, P = 0.002). Thus, the greenhouse system recapitulates important aspects of pollinator preference in the natural habitat and provides a practical means to assay large numbers of genetic variants.

View larger version (61K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 4. Pollinator Preference for Petunia Species under Controlled Greenhouse Conditions. (A) M. sexta feeding on P. axillaris. (B) B. terrestris feeding on P. integrifolia. (C) Mean (±SE) number of visits of M. sexta naïve females for feeding during 5 min to P. axillaris is significantly higher than to P. integrifolia flowers (n = 29). (D) Mean (±SE) number of landings per plant per flower in 2 h by B. terrestris on P. integrifolia is significantly higher than on P. axillaris (n = 12). (E) Mean (±SE) number of visits for feeding by M. sexta naïve females during 5 min in the greenhouse is significantly higher for WP117 than for WP119 (n = 34). (F) Mean (±SE) number of landings per flower per plant in 2 h of B. terrestris in the greenhouse is higher for the line WP119 than for the line WP117 (n = 12). Asterisks over bars in (C) to (F) indicate the significance level of the statistical tests. *, P = 0.01; **, P = 0.001; ***, P = 0.0001.

Thus, greenhouse experiments also showed that Petunia ILs that are polymorphic for AN2 are differentially visited by pollinators.

AN2 Has a Major Effect on Pollinator Preference
An important limitation of a genetic introgression approach is that the observed effects cannot strictly be ascribed to a single gene. To resolve this issue, we chose a transgenic approach and transformed P. axillaris with a functional AN2 P. integrifolia–type cDNA and tested pollinator preference in the greenhouse. The fact that species-specific AN2 alleles differ in their coding region makes AN2 an ideal gene for a transgenic approach in which the nonfunctional AN2 allele of P. axillaris is complemented by a transgenic allele of P. integrifolia. Transformed and nontransformed plants differed in the reflectance of the corolla limb (Figure 5C ) but neither in flower morphology (Figures 5A and 5B), emission of volatile compounds (Figure 5D; analysis of variance, P > 0.2, 1.762 > F > 0.334, n = 11 wild, n = 8 AN2-transformed), nectar volume (Figure 5E; analysis of variance, F = 0.154, P = 0.708, n = 4), or nectar sugar concentrations (Figure 5F; Mann-Whitney Test, –0.866 > Z > –1.732, P > 0.083, n = 4). We observed a higher feeding rate in a period of 5 min for wild P. axillaris than for AN2-transformed plants by the hawk moth M. sexta (Figure 5G; Wilcoxon signed-ranks test, Z = –3.946, P < 0.0001). First choice of these pollinators for flowers of nontransformed plants was significantly higher than for AN2-transformed wild plants (30 versus 6, binomial test P < 0.0001). When the same choice was offered to B. terrestris, a shift of preference in the opposite direction, toward AN2-transformed plants, was observed (Figure 5H; Wilcoxon signed-ranks test, Z = –2.393, P = 0.017). Thus, restoring AN2 function caused a major shift in pollinator preference.

View larger version (41K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 5. Effect of AN2 on Pollinator Preference. (A) and (B) Comparison of the wild type (A) and AN2-transformed P. axillaris (B) flowers. (C) Mean ± SE of percentage of reflectance relative to the standard for different wavelengths. (D) Mean (±SE) production of total volatiles (tot) and amount of single major compounds (1, benzaldehyde; 2, benzyl alcohol; 3, methyl benzoate; 4, methyl salicylate; 5, benzyl benzoate; 6, benzyl salicylate) (n = 11 wild; n = 8 transformed). (E) Mean (±SE) nectar amount (n = 4). ns, nonsignificant difference between treatments. (F) Mean (±SE) sugar concentrations (n = 4). (G) Mean (±SE) number of visits for feeding by M. sexta is significantly higher for wild P. axillaris than for AN2-transformed flowers (n = 36). (H) Mean (±SE) number of landings per flower per plant in 2 h of B. terrestris is higher on AN2-transformed than on wild plants (n = 12). Asterisks over bars indicate the significance of the statistical tests (*, P = 0.01; ***, P = 0.0001).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

Our study of AN2 provides a first step toward understanding the evolutionary steps involved in the transition between pollination syndromes in the genus Petunia. Using introgression and transgenic complementation, we have shown that AN2 is a major gene affecting both flower color per se and pollinator choice. Variation in AN2 homologues may also account for flower color variation in more distantly related taxa, such as snapdragon (Antirrhinum) species (Schwinn et al., 2006), suggesting that variation in highly specific transcription factors may be a major source of natural phenotypic variation and perhaps the favored target of natural selection in other species as well.

Changes in more general transcription factors, such as AN1, which also controls flower color (Spelt et al., 2002), would need to be subtle because of adverse pleiotropic effects. An example of such a scenario would be the TEOSINTE BRANCHED1 gene, which underwent a subtle change in expression during the domestication of maize (Zea mays; Doebley et al., 1997). The case described here is reminiscent of the multiple inactivating mutations in the vernalization genes FRIGIDA and FLOWERING LOCUS C in geographically structured populations of Arabidopsis (Johanson et al., 2000; Le Corre et al., 2002; Michaels et al., 2003; Shindo et al., 2005).

In light of the strong effect of AN2 alleles on pollinator behavior, it is surprising to find that loss-of-function alleles evolved at least five times independently in P. axillaris. Assuming absence of population structure, this suggests that either selection was not very strong, that these alleles arose at very high frequency, or both. This is in line with the marginal signature of selection, as indicated by Tajima's D statistics.

It has been suggested that the evolution of loss-of-function alleles in AN2 may represent a late step after Petunia species had become genetically isolated, as the species-specific alleles share a large number of fixed differences (Quattrocchio et al., 1999). These fixed differences are thought to have evolved before the occurrence of the loss-of-function alleles in P. axillaris. While our analysis supports the latter view and the notion that these alleles evolved relatively late, we do not see sufficient evidence for invoking reinforcement (Quattrocchio et al., 1999) (i.e., there is no evidence for an increase of phenotypic differences between the species in zones of sympatry).

Changes in flower color have been shown to reduce interspecific pollen flow and have been implicated in genetic isolation and speciation (Bradshaw et al., 1995, 1998; Clegg and Durbin, 2000; Hodges et al., 2002). The observed changes in pollinator preference associated with AN2 alleles in P. axillaris and P. integrifolia are comparable to the ones observed in Mimulus species carrying reciprocal introgressions of the YUP locus, which affects carotenoid pigmentation (Bradshaw and Schemske, 2003). It will be interesting to know the identity and evolutionary history of the YUP locus.

Accepting the hypothesis of a late major shift in flower color in the evolution of P. axillaris, a possible scenario to explain the transition from bee to nocturnal hawk moth pollination would be that an intermediate form would have been pollinated predominately by diurnal butterflies or hawk moths, which have an innate preference for colored flowers (Proctor et al., 1996; Kelber, 1997). Subsequent recruitment of nocturnal hawk moths might have involved selection for loss of AN2 function.

P. axillaris and P. intergrifolia display very distinct flower morphologies that conform to the traditional concepts of nocturnal hawk moth and bee pollination syndromes, respectively. However, in the natural habitat, these Petunia species are not visited by a single pollinator type but are pollinated by a wider range of pollinator types (Figure 2). Hence, adaptation to a new pollinator type may not leave a strong signature of selection.

The scenario suggested above is obviously speculative. Ultimately, the isolation of additional genes controlling traits such as morphology, scent, and nectar production and the analysis of their molecular evolution in combination with pollinator preference testing of isogenic or transgenic lines both under controlled conditions and in the wild will elucidate the transitions from one pollination syndrome to another.

	METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Molecular Analysis of AN2
The AN2 coding region was sequenced from 35 Petunia accessions collected in Uruguay or obtained elsewhere (Table 1). cDNA was obtained from total DNase-treated RNA from 2- to 3-d-old flower corollas from plants grown under greenhouse conditions. The primers AN2 start 5'-GCAGTGAGAACTATACATCATG-3' and AN2 stop 5'-TCTTCAATGGTCCCAATTAAC-3' were used to amplify the AN2 cDNA with Phusion polymerase (Finnzymes) at 98°C for 30 s and 35 cycles of 98°C for 10 s, 59°C for 30 s, and 72°C for 30 s, followed by 72°C for 5 min. The resulting cDNAs were sequenced by Fasteris. To obtain sequence variation of an additional locus on the same chromosome as AN2, we sequenced almost the entire coding region of the RT gene encoding rhamnosyl-transferase (Kroon et al., 1994). The primers RT forward 5'-GCTCGCAGTATTAAACAACAG-3' and RT reverse 5'-CAGCATTTTACAGCCACATTC-3' were used for amplification and sequencing. PCR conditions were as for AN2, but 61°C annealing temperature and an extension time of 40 s semiquantitative PCR on cDNA was performed with RedTaq (Sigma-Aldrich) primers AN2 qF 5'-GCCACATATAAAAAGAGGGGAC-3' and AN2 qR 5'-CAAGAAACATGATTCATTGCCG-3' with 94°C for 2 min and 23 or 35 cycles of 94°C for 30 s, 54°C for 30 s, and 72°C for 20 s, followed by 72°C for 5 min. For reference, semiquantitative PCR was performed with ACTIN1 with primers ACT F1 5'-TCCATGATTGGAATGGAAGC-3' and ACT R1 5'-GACCCACCACTGAGCACAA-3' with 94°C for 2 min and 27 and 30 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 15 s, followed by 72°C for 5 min. All controls without RNA were negative. All AN2 expression values were normalized to ACTIN1. The average of the Petunia integrifolia levels was taken as 100%.

Phylogenetic reconstruction was performed using maximum parsimony and maximum likelihood in PAUP version 4.0b 10 (Swofford, 1998) and Bayesian inference in MRBAYES (Huelsenbeck and Ronquist, 2001) under the Hasegawa-Kishino-Yano model (Hasegawa et al., 1985) with a gamma distribution shape parameter = 0.1527 and a transition-transversion ratio of 1.6061, the best-fitting model as determined with MODELTEST 3.7 (Posada and Crandall, 1998). Tajima's D (Tajima, 1989) and the McDonald-Kreitman test (McDonald and Kreitman, 1991) were calculated using DNASP4 (Rozas et al., 2003). For conducting the McDonald-Kreitman test, codons with nonsense or frameshift mutations were removed from the data set.

Petunia axillaris Transformation with AN2
The recombinant Ti plasmid pBCaMV, containing a P. integrifolia–derived AN2 cDNA under control of the 35S promoter (Quatrocchio et al., 1993), was transformed via Agrobacterium tumefaciens (strain LBA4404) into leaf discs of PaxS7 (Table 1).

Plants and Pollinators for Choice Experiments
Accessions of P. axillaris (PaxS7) and P. integrifolia (PintS6) (Table 1) were grown in the greenhouse for behavior experiments. The IL WP117 carries the complete chromosome VI, on which AN2 is located, from the wild PaxS7 in a Petunia hybrida W138 background. IL WP119 possesses half of chromosome VI from the wild species and the other half, including AN2, from P. hybrida W138. WP119 also contains a small fragment of the PaxS7 genome on chromosome III, which does not influence anthocyanin production.

Flower petal reflectance was measured with a DH2000 Micropack fitted with a UV-VIS-NIR light source (Tecan Rainbow Thermo), and data analysis was performed with the program OOIBase32. Nectar volume, nectar sugar concentration, and odor production were measured as previously described (Stuurman et al., 2004; Hoballah et al., 2005).

Four- to five-day-old, unmated Manduca sexta females, obtained as pupae from North Carolina State University Entomology Insectary (Raleigh) and reared under laboratory conditions were used in choice experiments. Bombus terrestris were obtained from Andermatt Biocontrol (Switzerland) as a minihive. B. terrestris were attracted to P. integrifolia flowers in the greenhouse (Figure 4) and could effectively pollinate them: 58.62% seed capsule formation for B. terrestris–pollinated flowers (n = 29) compared with 56.67% for hand-pollinated flowers (n = 30) and 0% for untouched control flowers (n = 30). Both M. sexta and B. terrrestris were naïve when used in choice experiments.

Pollinator Observations in the Wild
Pollinator visitation was observed in natural habitats in two allopatric populations of each species, P. axillaris (José Ignacio and Carmelo in January–February, 2004) and P. integrifolia (Puerto Viejo and Rivera in January–February, 2005). Pollinators were collected and sent to the Natural History Museum in London for identification.

To assess the effectiveness of diurnal and nocturnal pollinators, some plants were covered during the night (from 8 PM to 8 AM) and some plants during the day (from 8 AM to 8 PM) with insect tents from the time of flower opening until senescence. After flower senescence, the numbers of capsules per plant were counted.

The recombinant inbred lines WP117 and WP119 were tested in the wild in José Ignacio (Uruguay) between January 27 and February 2, 2005. Plants were grown in pots in the greenhouse and then transferred to the field and placed in a semirandomized design. Landings of the pollinators were observed for 4-h periods between 9.45 AM and 2:30 PM during the day (n = 4) and for 2-h periods between 9 PM and 12 PM during the night (n = 5). The same lines were tested in Minusio, southern Switzerland; observations were performed between 10 AM and 3 PM for 5 h in total. Each test was performed on 6 to 16 plants in a semirandomized design.

Pollinator Preference under Controlled Conditions
Experiments were conducted in a screen cage (144-cm height, 248 x 368 cm) in the middle of a Petunia scent–saturated greenhouse. We tested wild P. integrifolia versus P. axillaris (PaxS7), the recombinant inbred line WP117 versus WP119 (Table 1), and the wild versus the AN2-transformed PaxS7. Choice tests were conducted on 8 to 12 plants three to five times in semirandomization. M. sexta were allowed to fly and choose freely in the cage for 5 min. We recorded the first choice and the number of total flowers visited for feeding in 5 min. Hawk moths that did not feed in 5 min were discarded and annotated as "no choice"; 19.4 to 26.5% of hawk moths did not choose. For the B. terrestris experiments, a minihive was placed at one end of the cage and bumblebee landings on a flower in a period of 2 h were recorded. The experiments were repeated three to four times.

Accession Numbers
Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers EF423809 to EF423868. A reference cDNA sequence of AN2 was obtained from Genbank (accession number AF146704; Table 1).

Supplemental Data
The following materials are available in the online version of this article.

Supplemental Figure 1. Tajima's D Results.

Supplemental Figure 2. Phylogeographic Distribution of P. axillaris AN2 Loss-of-Function Alleles in Uruguay.

Supplemental Table 1. ² Test Results.

Supplemental Table 2. Species and Location of Pollinators Collected from P. axillaris and P. integrifolia Flowers in the Natural Habitat in Uruguay.

	Acknowledgments

 
We thank C. Ball and R. Alder (University of Bern) for plant care, E. Marchesi and D. Bayce (University Montevideo and Compania Forestal Uruguaya in Rivera) for help at the field site location in Uruguay, F. Quatrocchio (Free University, Amsterdam) for the AN2 construct and expert advice throughout the project, T. Turlings (University of Neuchâtel) for providing odor collection and analysis equipment, P. Heeb, S. Kulkarni, and N. Juillet (University of Lausanne) for reflectance measurements, S. Zeeman, T. Delatte, and M. Trevisan for sugar measurements (ETH Zurich), and L. Excoffier (University of Bern) for advice on population genetics. This project was funded by the National Centre of Competence in Research "Plant Survival," a research program of the Swiss National Science Foundation, and the University of Bern.

	Footnotes

 
¹ Current address: Keygene, PO Box 216, 6700AE Wageningen, The Netherlands.

² Current address: Laboratory of Plant Genetics, University of Geneva, Quai Ernest-Ansermet 30, 1211 Geneva 4, Switzerland.

³ Current address: Institute of Botany, University of Neuchâtel, Rue Emile-Argand 11, 2009 Neuchâtel, Switzerland.

^[W] Online version contains Web-only data.

www.plantcell.org/cgi/doi/10.1105/tpc.106.048694

Received November 2, 2006; Revision received January 31, 2007. accepted February 16, 2007.

	REFERENCES

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Ando, T., Iida, S., Kokubun, H., Ueda, Y., and Marchesi, E. (1995a). Distribution of Petunia axillaris sensu lato in Uruguay as revealed by discriminant analysis of the live plants. J. Japan. Soc. Hort. Sci. 64: 381–391.

Ando, T., Kokubun, H., Watanabe, H., Tanaka, N., Yukawa, T., Hashimoto, G., Marchesi, E., Suarez, E., and Basualdo, I.L. (2005). Phylogenetic analysis of Petunia sensu Jussieu (Solanaceae) using chloroplast DNA RFLP. Ann. Bot. (Lond.) 96: 289–297.[Abstract/Free Full Text]

Ando, T., Kurata, S., Sasaki, S., Ueda, Y., Hashimoto, G., and Marchesi, E. (1995b). Comparative morphological studies on intraspecific taxa of Petunia integrifolia (Hook.) Schinz et Thell. (Solanaceae). J. Japan. Bot. 70: 205–217.

Ando, T., Nomura, M., Tsukahara, J., Watanabe, H., Kokubun, H., Tsukamoto, T., Hashimoto, G., Marchesi, E., and Kitching, I.J. (2001). Reproductive isolation in a native population of Petunia sensu Jussieu (Solanaceae). Ann. Bot. (Lond.) 88: 403–413.[Abstract/Free Full Text]

Bradshaw, H.D., Otto, K.G., Frewen, B.E., McKay, J.K., and Schemske, D.W. (1998). Quantitative trait loci affecting differences in floral morphology between two species of monkeyflower (Mimulus). Genetics 149: 367–382.[Abstract/Free Full Text]

Bradshaw, H.D., and Schemske, D.W. (2003). Allele substitution at a flower colour locus produces a pollinator shift in monkeyflowers. Nature 426: 176–178.[CrossRef][Medline]

Bradshaw, H.D., Wilbert, S.M., Otto, K.G., and Schemske, D.W. (1995). Genetic mapping of floral traits associated with reproductive isolation in monkeyflowers (Mimulus). Nature 376: 762–765.[CrossRef]

Clegg, M.T., and Durbin, M.L. (2000). Flower color variation: A model for the experimental study of evolution. Proc. Natl. Acad. Sci. USA 97: 7016–7023.[Abstract/Free Full Text]

Doebley, J., Stec, A., and Hubbard, L. (1997). The evolution of apical dominance in maize. Nature 386: 485–488.[CrossRef][Medline]

Durbin, M.L., Lundy, K.E., Morrell, P.L., Torres-Martinez, C.L., and Clegg, M.T. (2003). Genes that determine flower color: The role of regulatory changes in the evolution of phenotypic adaptations. Mol. Phylogenet. Evol. 29: 507–518.[CrossRef][ISI][Medline]

Faegri, K., and van der Pijl, L. (1979). The Principles of Pollination Ecology. (Oxford, UK: Pergamon Press).

Fenster, C.B., Armbruster, W.S., Wilson, P., Dudash, M.R., and Thompson, J.D. (2004). Pollination syndromes and floral specialization. Annu. Rev. Ecol. Syst. 35: 375–403.[CrossRef][ISI]

Fishman, L., Kelly, A.J., and Willis, J.H. (2002). Minor quantitative trait loci underlie floral traits associated with mating system divergence in Mimulus. Evolution Int. J. Org. Evolution 56: 2138–2155.[CrossRef][ISI][Medline]

Galliot, C., Hoballah, M.E., Kuhlemeier, C., and Stuurman, J. (2006). Genetic control of flower size and nectar volume in Petunia pollination syndromes. Planta 225: 203–212.[CrossRef][ISI][Medline]

Gerats, T., and Vandenbussche, M. (2005). A model system for comparative research: Petunia. Trends Plant Sci. 10: 251–256.[CrossRef][ISI][Medline]

Grant, V., and Grant, K.A. (1965). Flower Pollination in the Phlox Family. (New York, London: Columbia University Press).

Harborne, J.B. (1982). Introduction to Ecological Biochemistry. (London: Academic Press).

Hasegawa, M., Kishino, H., and Yano, T.A. (1985). Dating of the human-ape splitting by a molecular clock of mitochondrial DNA. J. Mol. Evol. 22: 160–174.[CrossRef][ISI][Medline]

Herrera, C.M. (1996). Floral traits and plant adaptation to insect pollinators: A devil's advocate approach. In Floral Biology: Studies on Floral Evolution in Animal Pollinated Plants, D.G. Lloyd and S.C.H. Barrett, eds (New York: Chapman and Hall), pp. 65–87.

Hoballah, M.E., Stuurman, J., Turlings, T.C.J., Guerin, P.M., Connétable, S., and Kuhlemeier, C. (2005). The composition and timing of flower odour emission by wild Petunia axillaris coincide with the antennal perception and nocturnal activity of the pollinator Manduca sexta. Planta 222: 141–150.[CrossRef][ISI][Medline]

Hodges, S.A., Whittall, J.B., Fulton, M., and Yang, J.Y. (2002). Genetics of floral traits influencing reproductive isolation between Aquilegia formosa and Aquilegia pubescens. Am. Nat. 159: S51–S60.[CrossRef][ISI][Medline]

Holton, T.A., and Cornish, E.C. (1995). Genetics and biochemistry of anthocyanin biosynthesis. Plant Cell 7: 1071–1083.[CrossRef][ISI][Medline]

Huelsenbeck, J.P., and Ronquist, F. (2001). MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17: 754–755.[Abstract/Free Full Text]

Johanson, U., West, J., Lister, C., Michaels, S., Amasino, R., and Dean, C. (2000). Molecular analysis of FRIGIDA, a major determinant of natural variation in Arabidopsis flowering time. Science 290: 344–347.[Abstract/Free Full Text]

Johnson, S.D., and Steiner, K.E. (2000). Generalization versus specialization in plant pollination systems. Trends Ecol. Evol. 15: 140–143.[CrossRef][Medline]

Kauffman, S., and Levin, S. (1987). Towards a general theory of adaptive walks on rugged landscapes. J. Theor. Biol. 128: 11–45.[ISI][Medline]

Kelber, A. (1997). Innate preferences for flower features in the hawk moth Macroglossum stellatarum. J. Exp. Biol. 200: 827–836.[Abstract]

Koes, R., Verweij, W., and Quattrocchio, F. (2005). Flavonoids: A colorful model for the regulation and evolution of biochemical pathways. Trends Plant Sci. 10: 236–242.[CrossRef][ISI][Medline]

Kroon, J., Souer, E., de Graaff, A., Xue, Y., Mol, J., and Koes, R. (1994). Cloning and structural analysis of the anthocyanin pigmentation locus Rt of Petunia hybrida: Characterization of insertion sequences in two mutant alleles. Plant J. 5: 69–80.[CrossRef][ISI][Medline]

Kulcheski, F.R., Muschner, V.C., Lorenz-Lemke, A.P., Stehmann, J.R., Bonatto, S.L., Salzano, F.M., and Freitas, L.B. (2006). Molecular phylogenetic analysis of Petunia juss. (Solanaceae). Genetica 126: 3–14.[CrossRef][ISI][Medline]

Le Corre, V., Roux, F., and Reboud, X. (2002). DNA polymorphism at the FRIGIDA gene in Arabidopsis thaliana: Extensive nonsynonymous variation is consistent with local selection for flowering time. Mol. Biol. Evol. 19: 1261–1271.[Abstract/Free Full Text]

McDonald, J.H., and Kreitman, M. (1991). Adaptive protein evolution at the Adh locus in Drosophila. Nature 351: 652–654.[CrossRef][Medline]

Michaels, S.D., He, Y.H., Scortecci, K.C., and Amasino, R.M. (2003). Attenuation of FLOWERING LOCUS C activity as a mechanism for the evolution of summer-annual flowering behavior in Arabidopsis. Proc. Natl. Acad. Sci. USA 100: 10102–10107.[Abstract/Free Full Text]

Niovi Jones, K., and Reithel, J.S. (2001). Pollinator-mediated selection on a flower color polymorphism in experimental populations of Antirrhinum (Scrophulariaceae). Am. J. Bot. 88: 447–454.[Abstract/Free Full Text]

Oyama-Okubo, N., Ando, T., Watanabe, H., Marchesi, A., Uchida, K., and Nakayama, M. (2005). Emission mechanism of floral scent in Petunia axillaris. Biosci. Biotechnol. Biochem. 69: 773–777.[CrossRef][Medline]

Posada, D., and Crandall, K.A. (1998). Modeltest: Testing the model of DNA substitution. Bioinformatics 14: 817–818.[Abstract/Free Full Text]

Proctor, M., Yeao, P., and Lack, A. (1996). The Natural History of Pollination. (London: Harper Collins Publishers).

Quatrocchio, F., Wing, J.F., Leppen, H.T.C., Mol, J.N.M., and Koes, R.E. (1993). Regulatory genes controlling anthocyanin pigmentation are functionally conserved among plant species and have distinct sets of targets genes. Plant Cell 5: 1497–1512.[Abstract]

Quattrocchio, F., Wing, J., van der Woude, K., Souer, E., de Vetten, N., Mol, J.N.M., and Koes, R. (1999). Molecular analysis of the anthocyanin2 gene of petunia and its role in the evolution of flower color. Plant Cell 11: 1433–1444.[Abstract/Free Full Text]

Raguso, R.A., and Willis, M.A. (2002). Synergy between visual and olfactory cues in nectar feeding by naïve hawk moths, Manduca sexta. Anim. Behav. 64: 685–695.[CrossRef][ISI]

Rozas, J., Sanchez-DelBarrio, J.C., Messeguer, X., and Rozas, R. (2003). DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496–2497.[Abstract/Free Full Text]

Schemske, D.W., and Bradshaw, H.D. (1999). Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). Proc. Natl. Acad. Sci. USA 96: 11910–11915.[Abstract/Free Full Text]

Schiestl, F.P., Ayasse, M., Paulus, H.F., Löfstedt, C., Hansson, B.S., Ibarra, F., and Francke, W. (1999). Orchid pollination by sexual swindle. Nature 399: 421–422.[CrossRef]

Schwinn, K., Venail, J., Shang, Y., Mackay, S., Alm, V., Butelli, E., Oyama, R., Bailey, P., Davies, K., and Martin, C. (2006). A small family of MYB-regulatory genes controls floral pigmentation intensity and patterning in the genus Antirrhinum. Plant Cell 18: 831–851.[Abstract/Free Full Text]

Shindo, C., Aranzana, M.J., Lister, C., Baxter, C., Nicholls, C., Nordborg, M., and Dean, C. (2005). Role of FRIGIDA and FLOWERING LOCUS C in determining variation in flowering time of Arabidopsis. Plant Physiol. 138: 1163–1173.[Abstract/Free Full Text]

Spelt, C., Quattrocchio, F., Mol, J., and Koes, R. (2002). ANTHOCYANIN1 of petunia controls pigment synthesis, vacuolar pH, and seed coat development by genetically distinct mechanisms. Plant Cell 14: 2121–2135.[Abstract/Free Full Text]

Stuurman, J., Hoballah, M.E., Broger, L., Moore, J., Basten, C., and Kuhlemeier, C. (2004). Dissection of floral pollination syndromes in Petunia. Genetics 168: 1585–1599.[Abstract/Free Full Text]

Swofford, D.L. (1998). PAUP*: Phylogenetic Analysis Using Parsimony (*and Other Methods). (Sunderland, MA: Sinauer Associates).

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. (2001). How do visitation patterns vary among pollinators in relation to floral display and floral design in a generalist pollination system? Oecologia 126: 386–394.[CrossRef][ISI]

Waser, N.M., Chittka, L., Price, M.V., Williams, N.M., and Ollerton, J. (1996). Generalization in pollination systems, and why it matters. Ecology 77: 1043–1060.[CrossRef]

Waser, N.M., and Price, M.V. (1981). Pollinator choice and stabilizing selection for flower color in Delphinium nelsonii. Evolution Int. J. Org. Evolution 35: 376–390.[CrossRef][ISI]

Whibley, A.C., Langlade, N.B., Andalo, C., Hanna, A.I., Bangham, A., Thebaud, C., and Coen, E. (2006). Evolutionary paths underlying flower color variation in Antirrhinum. Science 313: 963–966.[Abstract/Free Full Text]

Whitlock, M.C., Phillips, P.C., Moore, F.B.G., and Tonsor, S.J. (1995). Multiple fitness peaks and epistasis. Annu. Rev. Ecol. Syst. 26: 601–629.[CrossRef][ISI]

Whittall, J.B., Voelckel, C., Kliebenstein, D.J., and Hodges, S.A. (2006). Convergence, constraint and the role of gene expression during adaptive radiation: Floral anthocyanins in Aquilegia. Mol. Ecol. 15: 4645–4657.[CrossRef][Medline]

Wijsman, H.J.W. (1983). On the interrelationships of certain species of Petunia II. Experimental data: Crosses between different taxa. Acta Bot. Neerl. 32: 97–107.

Wright, S. (1931). Evolution in Mendelian populations. Genetics 16: 97–159.[Free Full Text]

Zufall, R.A., and Rausher, M.D. (2004). Genetic changes associated with floral adaptation restrict future evolutionary potential. Nature 428: 847–850.[CrossRef][Medline]

This article has been cited by other articles:

				C. J. Majetic, R. A. Raguso, and T.-L. Ashman The Impact of Biochemistry vs. Population Membership on Floral Scent Profiles in Colour Polymorphic Hesperis matronalis Ann. Bot., December 1, 2008; 102(6): 911 - 922. [Abstract] [Full Text] [PDF]

				D. Kessler, K. Gase, and I. T. Baldwin Field Experiments with Transformed Plants Reveal the Sense of Floral Scents Science, August 29, 2008; 321(5893): 1200 - 1202. [Abstract] [Full Text] [PDF]

				Compiled by, F. Tooke, T. Chiurugwi, and N. Battey Flowering Newsletter bibliography for 2007 J. Exp. Bot., July 18, 2008; (2008) ern109v1. [Full Text] [PDF]

R. J. Waterman and M. I. Bidartondo Deception above, deception below: linking pollination and mycorrhizal biology of orchids J. Exp. Bot., March 2, 2008; (2008) erm366v1. [Abstract]