To b or not to b: a pheromone-binding protein regulates colony social organization in fire ants

Transcription

1 To b or not to b: a pheromone-binding protein regulates colony social organization in fire ants Michael J.B. Krieger Summary A major distinction in the social organization of ant societies is the number of reproductive queens that reside in a single colony. The fire ant Solenopsis invicta exists in two distinct social forms, one with colonies headed by a single reproductive queen and the other containing several to hundreds of egg-laying queens. This variation in social organization has been shown to be associated with genotypes at the gene Gp-9. Specifically, single-queen colonies have only the B allelic variant of this gene, whereas multiple-queen colonies always have the b variant as well. Subsequent studies revealed that Gp-9 shares the highest sequence similarity with genes encoding pheromone-binding proteins (PBPs). In other insects, PBPs serve as central molecular components in the process of chemical recognition of conspecifics. Fire ant workers regulate the number of egg-laying queens in a colony by accepting queens that produce appropriate chemical signals and destroying those that do not. The likely role of GP-9 in chemoreception suggests that the essential distinction in colony queen number between the single and multiple-queen form originates from differences in workers abilities to recognize queens. Other, closely related fire ant species seem to regulate colony social organization in a similar fashion. BioEssays 27:91 99, ß 2004 Wiley Periodicals, Inc. Introduction It is rather challenging to write an article under the heading of my favorite molecule for the reason that an enormous amount of knowledge has accumulated over the last two decades pertaining to the molecule s invertebrate host, but much less so with respect to the molecule itself. The host in question is the invasive pest species Solenopsis invicta, the red imported fire ant. In 1997, Ken Ross discovered that colony social organization of S. invicta was associated with different genotypes at the codominant protein locus general protein-9 (Gp-9 ). (1) At the time, it was not known whether the gene product of Center for Studies in Physics and Biology, Rockefeller University, New York, NY , USA. krieger@rockefeller.edu DOI /bies Published online in Wiley InterScience ( Gp-9 was directly involved in determining social organization nor was it known to what protein family it may belong. Having worked on the evolution of social systems myself, (2 4) I was fascinated by the possibility that a few genes, perhaps even a single gene could shape social organization in an ant species. In order to address this fascinating subject matter, I joined the Ken Ross laboratory to characterize the underlying genetic architecture of this protein, and if possible to understand its influence on social organization. To fully appreciate the findings on this interesting protein as well as the genetic basis of social organization in fire ants, it is first necessary to understand the basic features of its natural history and social biology. I will first describe the relevant features of the biology of S. invicta, detailing the major differences that characterize the two social forms of this species. Then I will continue with a summary of the genetic data showing that the gene Gp-9 is a major candidate gene influencing social organization in S. invicta and some other fire ant species. Background biology of S. invicta S. invicta, a South American native, was accidentally introduced to the US, probably around 1920 to the port of Mobile, Alabama. (5) Since its introduction, this ant has established itself throughout the southeastern US and more recently in California. Due to its high population densities, painful sting and negative impact on native wildlife and agriculture, S. invicta is considered a major economic and ecological pest in the US. A major distinction in the social organization of ant societies is the number of queens that co-exist in a colony. (6) Some species or populations have colonies that always contain a single queen whereas others have colonies that contain multiple queens. In S. invicta, both social forms exist and even occur at times in the same habitat. Colonies of the singlequeen form (monogyne form) are simple families headed by a single reproductive queen, whereas colonies of the multiplequeen form (polygyne form) contain several to hundreds of egg-laying queens. (7,8) The two social forms also differ in other important traits besides number of queens per nest, (9) including dispersal strategies, mode of colony founding and energy reserves of young queens. BioEssays 27:91 99, ß 2004 Wiley Periodicals, Inc. BioEssays

2 Young monogyne queens store large quantities of energy reserves in the form of fat and glycogen, which enables them to disperse over a relatively long distance. (10) A more important effect of these large energy reserves is however, the queen s ability to found new colonies without the assistance of workers (independently). After having mated with a single male in midair, monogyne queens land, dig a chamber in the soil and start laying eggs. Since queens do not feed during the initial stages of colony founding, they rear their first clutch of workers entirely on their own energy reserves. (11,12) In contrast, most polygyne queens do not accumulate sufficient energy reserves during their maturation to found new nests independently. Instead, they seek adoption into already existing polygyne nests where they initiate reproduction with the help of the existing worker force. General protein-9 (GP-9) The fact that colony social form was associated with different Gp-9 genotypes was quite surprising because the two forms strongly resemble one another genetically otherwise, typically harboring the same alleles and level of genetic variation at numerous genetic markers. (13,14) The genotypic pattern associated with each social form is remarkably simple: monogyne colonies harbor only the B allelic variant of Gp-9, whereas polygyne colonies always have the b variant as well. Accordingly, monogyne queens are always BB homozygous, and mate with a single, haploid male also bearing the B allele, resulting in female offspring that all bear the BB genotype. (1,15) In contrast, reproductive queens in polygyne colonies are always Bb heterozygotes but produce offspring with all three genotypes (BB, Bb, bb). However, only queens with the Bb genotype will become egg layers in polygyne colonies. This puzzling pattern arises through a complex interaction of queen phenotype, queen worker interaction and mode of colony founding (Fig. 1). The genotype at Gp-9 is strongly associated with the weight of young queens due to differential accumulation of fat during adult maturation. (16 18) Young queens with the homozygous BB genotype gain the most weight, regardless of whether they were raised in a monogyne or polygyne colony. Heterozygous queens gain intermediate weight and the bb homozygotes gain no weight at all (Fig. 2). The molecular basis for the differential weight gain is currently unknown. Perhaps GP-9 exerts this effect indirectly, perhaps currently unidentified, tightly linked genes are responsible for this effect. As mentioned above, the amount of energy reserves stored by queens determines their reproductive strategy. Relatively extensive reserves are necessary to carry a queen through independent colony founding, which can only be accomplished by queens with the BB genotype. In contrast, joining established polygyne colonies does not require extensive energy reserves and is easily accomplished by the lighter Bb queens (most bb queens die before they reach sexual maturity). The workers in a fire ant colony resolve which young queens will be accepted as new egg layers, which will be rejected by execution and how many queens are tolerated as permanent reproductives. Their decision is believed to depend on the presence or absence of specific queen pheromones. (19) In monogyne colonies, no additional reproductive queen of any genotype is ever tolerated besides the mother queen. (16,17,20,21) All intruder queens are seized and executed upon entering the colony, controlled by the queen s pheromonal signals. Identification of intruder queens might be further enhanced by the presence of colony-specific recognition cues derived from heritable and environmental (food, soil) sources (22) that are implicated in the highly aggressive behavior of monogyne workers towards non-nestmates. (23) In contrast, polygyne colonies display little or no intercolony discrimination. (23) Despite the tolerance towards non-colony members, only Bb queens are accepted in multiples as new reproductives into polygyne colonies, all BB queens are executed. (16,17,20,21) In short, monogyne colonies are initiated by a single queen carrying the BB genotype. All offspring carry the same BB genotype as their mother, providing the sexual offspring with extensive energy reserves that enables them to found new monogyne nests. Carrying this genotype however, makes it impossible to become reproductively active in any polygyne colony, as any attempt will lead inevitably to their execution by polygyne workers. Polygyne colonies contain only heterozygous reproductives but produce young queens of all three genotypes. However, the bb homozygotes die before sexual maturity, most of the BB queens approaching sexual maturity are executed and only the Bb heterozygotes mature in good numbers to become egg layers themselves. Most importantly however, only queens carrying the Bb genotype are ever accepted as new reproductives in polygyne colonies. Conversely, the lower energy reserve of this genotype prevents the Bb queens from founding nests independently (Fig. 1). What protein is encoded by Gp-9? Gp-9 was regularly used in our laboratory as one of many protein markers to assess population genetic structure in fire ant populations. (13,14) Each marker represents a protein that exists in distinct variants, differing in their electrophoretic mobility. These protein variants are first separated on starch gels by their charge differences and then stained in order to visualize the banding pattern. For known, enzymatic proteins, the protein stain typically contains the substrate and all the cofactors required by the enzyme to catalyze its reaction. In the case of non-enzymatic or unknown proteins, the visualization of the banding phenotypes is accomplished using nonspecific protein staining. GP-9 was in the latter category of unknown proteins, as might be hinted from its non-predicative name (general protein-9). 92 BioEssays 27.1

3 Figure 1. Colony cycle of the monogyne and polygyne social form. Monogyne BB homozygous queens mate with a single, haploid male also bearing the B allele, resulting in female offspring that all bear the BB genotype. After the mating flight, monogyne queens found new nests independently, and rear their first brood entirely on accumulated fat reserves. The BB genotype provides young queens with large energy reserves needed to found new monogyne nests. Carrying the BB genotype, however, makes it impossible to be accepted into polygyne colonies, as any attempt will lead to their inevitable execution. Polygyne colonies contain only Bb queens but produce young queens of all three genotypes (BB, Bb, bb). Yet, only the Bb queens survive to maturity and become egg layers by seeking adoption into existing polygyne colonies. Bb queens do not accumulate enough fat reserves to initiate new nests independently (see text for details). We determined the amino acid sequences of some of the peptide fragments isolated from the starch gel and subsequently used this information to recover the full-length mrna transcript of Gp-9. (24) Determination of the nucleotide sequence of Gp-9 and the predicted amino acid sequence of its protein product revealed that it shares the highest sequence similarity with genes encoding pheromonebinding proteins (PBP), a subclass of the odorant-binding protein (OBP) family. (25) OBPs are mainly expressed in the antenna and are characterized by six absolutely conserved cysteine residues located in similar positions. (26,27) They transport odorants from the porous sensillum wall to the receptors located on the dendritic membrane of the olfactory sensory neurons. (27) These neurons respond to a range of odors such as plant volatiles, food sources, (28) and conspecifics. (29) BioEssays

4 Figure 2. Weight gain of maturing queens of different Gp-9 genotypes and social organization. Mean weights are presented with their corresponding standard deviations. (16,18,20) Fire ant workers regulate the number and identity of egg-laying queens in a colony by accepting queens that produce appropriate chemical signals and destroying those that do not. (19,21,30) Thus, the core feature of colony social organization, the number of egg-laying queens, is mediated by worker recognition and subsequent discrimination among queens. The presumed role of GP-9 in chemoreception suggests that the essential distinction in colony queen number between the monogyne and polygyne forms is strongly connected to differences in workers abilities to recognize queens. The Gp-9 alleles in S. invicta Our first undertaking after the identification of Gp-9 as a PBP gene was to link the protein variation to the underlying nucleotide variation. Accordingly, we sequenced Gp-9 alleles from numerous individuals of both social forms collected throughout the introduced range in the US. We found two variants, and these corresponded in every case to the two alleles identified by protein electrophoresis (B and b). Remarkably, the two Gp-9 alleles differed by nine nucleotide substitutions in their coding regions, each of which is associated with an amino acid substitution. (24) This high ratio of nonsynonymous to synonymous substitutions suggests that positive selection has driven the divergence of these alleles, (31) consistent with behavioral studies implicating strong diversifying selection on Gp-9. (1,32) The correspondence of the sequence variants and electrophoretically determined alleles was further confirmed by identifying the chargechanging amino acid substitution at position 151, responsible for the different electrophoretic mobilities of the allelic proteins in starch gels. Thus, the allelic pattern that emerged from the nucleotide data corresponded unerringly with the pattern seen at the protein level: the B allele is retrieved from monogyne and polygyne populations, whereas the b allele is found exclusively in polygyne populations. This pattern led to the hypothesis that the b allele is required for the expression of the polygyne social form. However, previous protein electrophoretic studies of S. invicta from the native range in Argentina showed that, although the b allele is found only in the polygyne form, some nests of this form contain egg-laying queens scored electrophoretically as BB homozygotes. These findings, if confirmed at the nucleotide level, would have nullified our hypothesis that the b allele is required for the expression of the polygyne social form. By sequencing appropriate samples from the native range, we found that these polygyne queens were also heterozygotes at Gp-9, but possessed a cryptic, functionally b-like allele (b 0 ) that encodes a protein bearing the net charge of, and thus electrophoretically indistinguishable from, a B allele product. This cryptic b 0 allele is more similar to the b than the B allele over its entire coding sequence yet bears the same charge-conferring amino acid as the B allele at position 151 (Fig. 3A). Figure 3. Amino acid variation for the alleles of Gp-9. A: Socially polymorphic species from the South American fire ant clade. The alleles separate into two distinct groups, the B-like and the polygyny-permitting b-like alleles. Amino acid substitutions uniquely shared among all b-like alleles, are indicated by circles, the charge-changing amino acid substitution in the b allele of S. invicta is indicated by a star. B: Amino acid composition at the same positions in S. geminata, a species belonging to the North American fire ant clade. (24,25) 94 BioEssays 27.1

5 This confirmed link of allelic pattern and social form in the native range of S. invicta led to the possibility that the expression of polygyny in other species might be similarly associated with the genotypic pattern at Gp-9. To this end, we sequenced Gp-9 in nine other Solenopsis species, as well as in a species of a related genus, to establish the taxonomic range over which a homolog occurs. We were unable to amplify Gp-9 in the related genus, suggesting that if such a homolog exists, it must have undergone extensive sequence divergence at the primer-binding sites. Gp-9 alleles in other fire ant species We used sequence data from the ten Solenopsis species to reconstruct the evolutionary relationships of the Gp-9 variants (Fig. 4). The evolutionary relationships were found to be consistent with the classification of these ants, (33,34) forming two large groups the North American and the South American fire ant clade. (24) Most interestingly, Gp-9 sequences from the species in the South American clade known to display polymorphism in social organization are further divided in sister clades, one containing the close relatives of the polygyny-permitting b allele of S. invicta and the other containing the close relatives of the B allele of S. invicta. As expected if alleles in the b-like clade induce polygyny, queens from confirmed polygyne nests of three species (in addition to S. invicta) invariably carried such b-like alleles. The deduced phylogeny allowed us to infer that the ancestral Gp-9 allele for the socially polymorphic clade was of the B type, and, hence, that monogyne social organization preceded polygyny in the evolutionary history of South American fire ants. The two other South American species that are most closely related to the socially polymorphic species, are not known to exhibit polygyny. Their lack of a b-like allele is consistent with the hypothesis that b-like alleles are necessary for the expression of polygyny (Fig. 4). The implied single origin of b-like alleles in these ants apparently predated the origins of most of the species, suggesting that the expression of polygyny in each was made possible by survival of the descendants of an ancestral b-like allele through sequential speciation events. The availability of a gene phylogeny for Gp-9 also made it possible to investigate the role of selection in the evolutionary history of this gene. We tested the specific hypothesis that the b-like alleles, which are integral to the polygyne social system, are under different selective regimes than the B-like alleles, which must function in both social systems. As hypothesized, positive selection was statistically significant only on branches within the b-like clade. Despite positive selection having acted periodically on various b-like alleles to drive their divergence from their B-like counterparts, all b-like alleles uniquely share the three amino acids G42, I95 and I139 (Fig. 3A), suggesting that one or more of these residues are essential for the expression of polygyny. It is therefore of particular interest to investigate their relative position on the protein structure, perhaps allowing us to infer the molecular mechanism by which polygyny is induced. Unfortunately, we do not have the structure of GP-9 at our disposal, but computerized sequence alignment algorithms in combination with threading techniques (35,36) allows us to arrive at reasonably accurate structure prediction, assuming appropriate structural templates are supplied. Structural hypothesis Currently there are four OBP three-dimensional structures available on the Protein Data Bank: a silkmoth PBP, (37 39) a cockroach PBP, (40) a fruit fly OBP (41) and, most recently, a honey bee PBP. (42) While all these OBPs display similar folds, their binding pockets comprise a variety of shapes and binding affinities suggesting that the OBP fold is fairly versatile and able to bind a considerable range of organic compounds. Figure 4. Cladogram illustrating the phylogenetic relationship of the Gp-9 alleles from ten Solenopsis species. The species separate into two groups, the North American and the South American fire ant clade. Gp-9 sequences from the South American clade of fire ant species known to display polymorphism in social organization form sister clades, one containing the polygynypermitting b-like alleles, the other containing the B alleles. The tree is rooted using the sequence from the non-fire-ant species S. globularia littoralis. BioEssays

6 The silkmoth PBP undergoes a ph-dependent conformational change that results in a structure that is unable to bind the ligand. Specifically, the C-terminal tail of the protein forms an a-helix that folds into the binding pocket at low ph, blocking the ligand-binding site. (38) This has led to the hypothesis that ligands are unloaded through a conformational change of their carrier protein as they approach the acidic membrane. (39) However, ph-dependent conformational change does not seem to be an universal feature involved in unloading ligands in all OBPs, as only the silkmoth protein has such an extended C terminus that folds inside the protein at acidic ph. The other three OBP proteins either lack this C-terminal tail entirely (40) or carry only a short extended irregular structure, which, rather than forming a mobile helix, is part of the cavity wall. This stable structural configuration is unaffected at different physiological ph values. (41,42) To establish the significance of each of the three amino acid residues consistently associated with the polygyny-permitting b-like alleles, we mapped these residues to the GP-9 structure prediction obtained by GenTHREADER (35) and the Robetta server. (36) We found that two of the residues, Ile 95 and Ile, 139 are part of the cavity wall that surrounds the binding pocket. Residue Ile 139 even extends its side chain into the binding pocket, an indication that Ile 139 might be directly involved in ligand binding. The third residue, Gly 42 is located on a solvent exposed loop-like structure that is not part of the cavity wall. The change from a serine to a glycine on this loop probably has little effect on the protein structure as both amino acids are similar in size and exhibit similar hydrophobicities. However, the two substitutions located on the cavity wall (M95I, V139I), although not radically different, are more likely to have an effect on the binding of the ligand. The initial discovery of Gp-9 and its association with social form in S. invicta was due to the differential electrophoretic mobility of the two protein alleles, caused by a single chargechanging amino acid substitution at position 151 in the b allele. A second, cryptic b 0 allele was discovered in Argentina that encodes another functional b-like protein but bears the net charge of the B allele. The three additional South American polygyne species also invariably possess b-like alleles that lack the charge-changing amino acid substitution. Thus, only one allele, the b allele of S. invicta harbors the chargechanging amino acid substitution (Fig. 3A). This substitution maps to the C-terminal tail of the silkmoth PBP, the structure that is supposedly responsible for unloading the ligand or, in the case of the fruit fly or honey bee protein, to the irregular C-terminal structure that is a part of the cavity wall. The C- terminal tail of GP-9 is slightly shorter than the C-terminal tail of the silkmoth protein, yet longer than the fruit fly or honey bee protein. Hence, it is difficult to assess with certainty which of the two structures GP-9 more closely resembles at its C- terminal tail. However, in both cases, it is likely that a change from a basic lysine to an acidic glutamic acid will cause the PBP to lose its ability to bind or to release the ligand. A second possibility, as parts of the C-terminal tail of the silkmoth PBP are also involved in dimer formation, (37) is that the chargechanging substitution prevents proper dimer formation of GP-9 and, hence, renders the molecule biologically nonfunctional. Interestingly, these structural hypotheses implicating the lack of function of the b allele protein, is consistent with the differential weight gains of young maturing queens according to their Gp-9 genotypes (Fig. 2). According to these structural hypotheses, bb homozygotes do not express any functional GP-9 protein, consistent with the lack of weight gain during their maturation process. Heterozygous queens will generate 50% of their total GP-9 production in a functional form and they gain intermediate weight, and finally, the BB queens produce only functional GP-9 proteins and hence gain the most weight. While the details of such a direct relationship between weight gain and protein structure remain difficult to understand, it reveals that the charge-changing amino acid substitution occurring in the b allele of S. invicta and the three substitutions commonly shared among all b-like alleles have two distinct effects. According to this hypothesis, polygyny is induced by at least one of the three amino acid substitutions shared among all b-like alleles whereas the differential weight gains of young S. invicta queens and the associated low viability of the bb homozygotes is caused by the charge-changing amino acid substitution. This leads to a testable prediction, namely that polygyne colonies harboring b-like alleles that lack the chargechanging amino acid substitution should contain viable reproductive queens with a bb-like genotype. Preliminary data from a recent population screen in S. richteri from South America suggest that this hypothesis is correct (unpublished data). Are the molecular mechanisms underlying social organization in fire ants universal? Allelic determination of polygyny was found in four closely related species in the South American fire ant clade. The ancestral Gp-9 allele for the socially polymorphic clade was of the B type, implying that the monogyne social organization preceded polygyny in the South American fire ants and that the b-like alleles originated within that clade. Outside of this group, polygyny is unfortunately only well documented in a single North American species, S. geminata (43,44) Nevertheless, the occurrence of polygyny in S. geminata, makes it possible to address the question of whether allelic determination of social organization represents an universal mechanism that can be found in other ant species as well. We had already sequenced the Gp-9 sequence of a monogyne S. geminata specimen in our initial study (24) and this sequence features the characteristic B-like residues methionine and valine at positions 95 and 139, respectively (Fig. 3B). However, the sequence includes the b-like glycine residue at position 42, suggesting 96 BioEssays 27.1

7 that the amino acid at this position may not be essential to the function of GP-9 protein with respect to social organization. Interestingly, this is the same residue that was found on the loop-like structure of the GP-9 protein predicted by the protein structure prediction software. In our search for polygyne S. geminata specimens, Sanford Porter, a fire-ant researcher from Gainesville Florida, directed us to small, isolated polygyne population consisting of approximately twenty-five nests, nestled within a small strip of land among a collection of otherwise monogyne nests. Despite extensive search efforts, we were unable to locate any other polygyne population in the US. The sequence analysis of the polygyne Gp-9 sequences revealed that the amino acid replacements characteristic of all b-like alleles were not present in polygyne S. geminata (45) disproving our original hypothesis that one or both of these substitutions are necessary for the expression of polygyny in all fire ants. We found, however, that the polygyne form lost a great deal of genetic diversity at both their nuclear and mitochondrial genomes relative to the monogyne form. This led us to speculate that the polygyne form of S. geminata originated from a small, isolated founder population derived from the nearby monogyne population. In this view, the origin of polygyny in S. geminata was driven by a loss of genetic variation rather than by specific amino acid replacements at Gp-9, the evolutionary event that drove the origin of polygyny in the South American fire ant clade species. Loss of genetic variation has been previously invoked to account for shifts in colony social organization. For example, in the Argentine ant (Linepithema humile), a change in the ability to recognize nestmates, a feature well developed in its native South American range, but lost in the introduced ranges, has led to the formation of massive supercolonies in which individual ants mix freely among physically separated nests. (46 48) Reduced nestmate recognition in the introduced ranges coincides with loss of genetic diversity. These observations have led to the idea that loss of alleles encoding chemical recognition cues, caused by the founder events, (48,49) have eroded the nestmate discrimination abilities of Argentine ant workers in their introduced ranges, thereby inducing a shift in colony social organization. Future directions Our research on the molecular mechanism of social organization in fire ants revealed two possible routes leading to the polygyne social form. Both scenarios invoke changes in the molecular components of the chemoreception systems, although in a different manner. In the South American fire ant clade, polygyny evolved presumably by a change in the binding affinity of GP-9, altering recognition capabilities of workers that bear b-like alleles. Evolution of polygyny in S. geminata via loss of alleles at loci encoding recognition cues seems to involve a reduction in the diversity of chemical labels necessary for the proper functioning of a discrimination system that serves in regulating queen number. Several issues regarding Gp-9 and its role in determining social organization are still unresolved and need further attention. First, what is the molecular basis for the differential weight gain according to the Gp-9 genotypes in young S. invicta queens? Is GP-9 exerting this effect indirectly, caused by a change in the binding affinity of the b allele? For example, a change in the binding affinity could altered odor perception thereby causing queens to behave atypically. Since it is the workers that feed young queens, uncharacteristic queen behavior may lead to a neglect by workers, resulting in lower body weights of queens carrying the b allele. Alternatively, the differential weight gain is not a result of the different Gp-9 alleles but brought about by currently unidentified genes, tightly linked to Gp-9. To explore this idea, we intend to sequence an approximately 200-kb region around the B and b allele of the Gp-9 gene in order to find candidate genes that might affect weight gain. Differences in the DNA sequences of genes associated with the alternate Gp-9 alleles will point to such candidate genes. A second unresolved issue is the acceptance of only Bb queens into polygyne colonies. Why are all BB queens executed but most of the Bb queens entering polygyne colonies are left unharmed? The reason for this phenomenon is not known, but we speculate it is a combination of three factors. Two of the factors reduce worker aggression in general, and one relates specifically to Bb queens. The most important factor, we believe, is the presence of Bb workers in polygyne colonies. It has been shown that at least 5 10% of the worker force must be of the Bb genotype in order for Bb queens to be accepted as new reproductives. (21) We speculate that the recognition capability of workers that bear the b allele is altered, allowing the majority of Bb queens to pass undetected. In addition, the lack of intercolony discrimination in polygyne colonies further reduces aggression towards nonnestmate queens. The last factor is connected to queen pheromone production and may explain why only Bb queens are accepted into polygyne colonies. Fletcher and Blum (19) showed that the weight of a queen is positively correlated with the quantity of pheromone that she produces. It is therefore possible that the reduced pheromone signal of the lighter Bb queens (Fig. 2) is below the threshold that otherwise triggers execution of BB queens. Non-aggression towards Bb queens based on a weaker queen pheromone signal seems likely an important factor, but what is the exact role of the Bb workers that must be present in colonies that accept multiple Bb queens? How is the presence of the presumed smelling-impaired Bb workers preventing the rise of the otherwise aggressive collective? Most importantly, however, how does this phenomenon relate to the different allelic forms of GP-9? To address this issue, we aim to determine the three-dimensional structure of the BioEssays

8 three allelic protein variants in S. invicta. By determining the structure of the proteins, we will learn whether the differences in amino acid sequence translate into differences in protein structure, identify the residues involved in binding the pheromone ligand, and determine the shape of the binding pocket. This may lead to predictions about the nature of the unidentified ligand, as well as how these amino acid substitutions affect its binding affinity. This will be crucial for inferring the biochemical and behavioral mechanisms regulating colony queen number, and for illuminating how variation in Gp-9 genotype affects the process. Finally, to further investigate the molecular mechanism of social organization in the genus Solenopsis, we intend to sequence additional polygyne populations of S. geminata occurring elsewhere in its vast range to determine whether Gp- 9 sequence variation corresponds with polygyny in the manner that we initially hypothesized or if reduced genetic diversity in genes encoding recognition cues are consistently associated in the history of these polygyne populations. In a second approach, our aim is to examine Gp-9 genes from additional species throughout the genus Solenopsis, with the purpose of tracking the molecular evolutionary history of this fascinating molecule. Acknowledgments I thank Lara Carroll, Ken Ross, Adam Wilkins and two anonymous referees for helpful comments on the manuscript. References 1. Ross KG Multilocus evolution in fire ants: effects of selection, gene flow, and recombination. Genetics 145: Bernasconi G, Krieger MJB, Keller L Unequal partitioning of reproduction and investment between cooperating queens in the fire ant, Solenopsis invicta, as revealed by microsatellites. Proc Roy Soc Lond B 264: Krieger MJB, Billeter JB, Keller L Ant-like task allocation and recruitment in cooperative robots. Nature 406: Krieger MJB, Keller L Mating frequency and genetic structure of the Argentine ant Linepithema humile. Mol Ecol 9: Lofgren CS History of imported fire ants in the United States. In: Lofgren CS, Vander Meer RK, editors. Fire Ants and Leaf-cutting Ants: Biology and Management. Boulder, CO: Westview Press. p Hölldobler B, Wilson EO The number of queens: an important trait in ant evolution. Naturwissenschaften 64: Ross KG, Fletcher DJC Comparative study of genetic and social structure in two forms of the fire ant, Solenopsis invicta (Hymenoptera: Formicidae). Behav Ecol Sociobiol 17: Vargo EL, Fletcher DJC Effect of queen number on the production of sexuals in natural-populations of the fire ant, Solenopsis invicta. Physiol Entomol 12: Ross KG, Keller L Ecology and evolution of social-organizationinsights from fire ants and other highly eusocial insects. Annu Rev Ecol Syst 26: Keller L, Ross KG Phenotypic basis of reproductive success in a social insect: genetic and social determinants. Science 260: Markin GP, Dillier JH, Hill SO, Blum MS, Hermann HR Nuptial flight and flight ranges of the imported fire ant, Solenopsis saevissima richteri (Hymenoptera: Formicidae). J Ga Entomol Soc 6: Tschinkel WR, Howard DF Colony founding by pleometrosis in the fire ant, Solenopsis invicta. Behav Ecol Sociobiol 12: Ross KG, Krieger MJB, Shoemaker DD, Vargo EL, Keller L Hierarchical analysis of genetic structure in native fire ant populations: results from three classes of molecular markers. Genetics 147: Ross KG, Shoemaker DD, Krieger MJB, DeHeer CJ, Keller L Assessing genetic structure with multiple classes of markers: a case study involving the introduce fire ant Solenopsis invicta. Mol Biol Evol 16: Shoemaker DD, Ross KG Effects of social organization on gene flow in the fire ant Solenopsis invicta. Nature 383: Keller L, Ross KG Phenotypic plasticity and cultural transmission of alternative social organizations in the fire ant Solenopsis invicta. Behav Ecol Sociobiol 33: Ross KG, Keller L Genetic control of social organization in an ant. Proc Natl Acad Sci USA 95: DeHeer CJ, Goodisman MAD, Ross KG Queen dispersal strategies in the multiple-queen form of the fire ant Solenopsis invicta. Am Nat 153: Fletcher DJC, Blum MS Regulation of queen number by workers in colonies of social insects. Science 219: Keller L, Ross KG Major gene effects on phenotype and fitness: the relative roles of Pgm-3 and Gp-9 in introduced populations of the fire ant Solenopsis invicta. J Evolution Biol 12: Ross KG, Keller L Experimental conversion of colony social organization by manipulation of worker genotype composition in fire ants (Solenopsis invicta). Behav Ecol Sociobiol 51: Obin MS, Vander Meer RK Sources of nestmate recognition cues in the imported fire ant Solenopsis invicta Buren (Hymenoptera, Formicidae). Anim Behav 36: Morel L, Vander Meer RK, Lofgren CS Comparison of nestmate recognition between monogyne and polygyne populations of Solenopsis invicta (Hymenoptera, Formicidae). Ann Entomol Soc Am 83: Krieger MJB, Ross KG Identification of a major gene regulating complex social behavior. Science 295: Pelosi P, Maida R Odorant-binding proteins in insects. Comp Biochem Physiol B 111: Pikielny CW, Hasan G, Rouyer F, Rosbash M Members of a family of Drosophila putative odorant-binding proteins are expressed in different subsets of olfactory hairs. Neuron 12: Vogt RG Biochemical diversity of odor detection: OBPs, ODEs and SNMPs. In: Blomquist GJ, Vogt RG, editors. Insect Pheromone Biochemistry and Molecular Biology: The Biosynthesis and Detection of Pheromones and Plant Volatiles. London: Elsevier Academic Press. p Stensmyr MC, Giordano E, Balloi A, Angioy AM, Hansson BS Novel natural ligands for Drosophila olfactory receptor neurones. J Exp Biol 206: Vogt RG The molecular basis of pheromone reception: its influence on behavior. In: Prestwich GD, Blomquist GJ, editors. Pheromone Biochemistry. New York, NY: Academic Press. p Keller L, Ross KG Selfish genes: a green beard in the red fire ant. Nature 394: Li WH Molecular Evolution. Sunderland, MA: Sinauer Associates. 32. Ross KG, Keller L Joint influence of gene flow and selection on a reproductively important genetic-polymorphism in the fire ant Solenopsis invicta. Am Nat 146: Trager JC A revision of the fire ants, Solenopsis geminata group (Hymenoptera:Formicidae: Myrmicinae). J NY Entomol Soc 99: Pitts JP A cladistic analysis of the Solenopsis saevissima species group (Hymenoptera: Formicidae). Athens, GA: University of Georgia PhD dissertation. 35. McGuffin LJ, Jones DT Improvement of the GenTHREADER method for genomic fold recognition. Bioinformatics 19: Chivian D, Kim DE, Malmstrom L, Bradley P, Robertson T, et al Automated prediction of CASP-5 structures using the Robetta server. Proteins 53: Sandler BH, Nikonova L, Leal WS, Clardy J Sexual attraction in the silkworm moth: structure of the pheromone-binding-protein-bombykol complex. Chem Biol 7: BioEssays 27.1

9 38. Horst R, Damberger F, Luginbuhl P, Guntert P, Peng G, et al NMR structure reveals intramolecular regulation mechanism for pheromone binding and release. P Natl Acad Sci USA 98: Lee D, Damberger FF, Peng GH, Horst R, Guntert P, et al NMR structure of the unliganded Bombyx mori pheromone-binding protein at physiological ph. Febs Letters 531: Lartigue A, Gruez A, Spinelli S, Riviere S, Brossut R, et al The crystal structure of a cockroach pheromone-binding protein suggests a new ligand binding and release mechanism. J Biol Chem 278: Kruse SW, Zhao R, Smith DP, Jones DNM Structure of a specific alcohol-binding site defined by the odorant binding protein LUSH from Drosophila melanogaster. Nat Struct Biol 10: Lartigue A, Gruez A, Briand L, Blon F, Bezirard V, et al Sulfur single-wavelength anomalous diffraction crystal structure of a pheromone-binding protein from the honeybee Apis mellifera L. J Biol Chem 279: Adams CT, Banks WA, Plumley JK Polygyny in the tropical fire ant, Solenopsis geminata with notes on the imported fire ant, Solenopsis invicta. Fla Entomol 59: Mackay WP, Porter S, Gonzalez D, Rodriguez A, Armendedo H, et al A comparison of monogyne and polygyne populations of the tropical fire ant, Solenopsis geminata (Hymenoptera: Formicidae), in Mexico. J Kans Entomol Soc 63: Ross KG, Krieger MJB, Shoemaker DD Alternative genetic foundations for a key social polymorphism in fire ants. Genetics 165: Holway DA, Suarez AV, Case TJ Loss of intraspecific aggression in the success of a widespread invasive social insect. Science 282: Tsutsui ND, Case TJ Population genetics and colony structure of the argentine ant (Linepithema humile) in its native and introduced ranges. Evolution 55: Giraud T, Pedersen JS, Keller L Evolution of supercolonies: the Argentine ants of southern Europe. Proc Natl Acad Sci USA 99: Tsutsui ND, Suarez AV, Grosberg RK Genetic diversity, asymmetrical aggression, and recognition in a widespread invasive species. Proc Natl Acad Sci USA 100: BioEssays