Behavioral Genetics: A Gene for Supersociality Dispatch

Behavioral genetics is an emerging field of biology. Many biologists have been reluctant to admit that behavior is under strong genetic control, because of the significant implications this may have on human social policy [1]. There is growing evidence, however, that most behaviors, be it the level of aggressiveness in mice, the foraging propensities of bees or the cognitive abilities of humans, are at least partially influenced by genes. Generally, many genes are involved and there is a strong interaction between genes and environment. But this need not always be so, as demonstrated by a new study [2] showing that a single gene determines a complex social behavior in the fire ant Solenopsis invicta, as well as in several other closely related species. The fire ant is a serious pest that has been inadvertently imported from South America to the southeastern United States and more recently to California and Australia. This species displays a fundamental social polymorphism that is under simple genetic control [3]. In the monogyne form, colonies invariably contain a single queen, whereas in the polygyne form colonies contain anywhere between 2 and 200 queens. This fundamental difference in social organization is completely associated with variation at the gene Gp-9. In the monogyne form, the queen and all workers invariably have the Gp-9BB genotype; by contrast, polygyne colonies always contain both Gp-9BB and Gp-9Bb workers. The queens of polygyne fire ant colonies are all heterozygotes. The lack of Gp-9BB queens in polygyne colonies has been shown to be caused by the selfish Gp-9b allele (or a closely linked locus) [4]. This allele induces workers carrying it selectively to kill all queens without a copy (Gp-9BB queens) immediately after the queens initiate reproduction. Both Gp-9bb queens and Gp-9bb workers are absent from polygyne colonies because Gp-9b is also a lethal recessive allele, causing the death of homozygous females soon after they eclose from the pupae stage [5]. Hence, the strong directional selection against Gp-9BB queens is exactly compensated by inviability of Gp-9bb queens, leading to a stable polymorphism of the two alleles in the polygyne form. Krieger and Ross [2] have now cloned and sequenced Gp-9. GenBank searches showed that the Gp-9 product most closely resembles moth odorant binding proteins. Although the absolute level of amino acid identity is low (26%), the conservation of six cysteine sites in Gp-9 and all known odorant binding proteins strongly suggests that Gp-9 is a novel odorant binding protein. These proteins are thought to be molecular components involved in chemical recognition, strongly supporting the view that Gp-9 is directly involved in the existence of the two social forms. Previous studies showed that Gp-9Bb queens have a distinctive odor compared to Gp-9BB queens, allowing Gp-9Bb workers to recognize and selectively eliminate Gp-9BB queens. Given the hypothesized function of odorant binding proteins, it is very likely that Gp-9BB and Gp-9Bb workers differ in how they perceive the odor of Gp-9BB and Gp-9Bb queens, hence explaining the different response of workers toward monogyne and polygyne queens. It remains to be determined why the presence of Gp-9Bb workers in polygyne colonies influences the number of queens accepted. Previous studies showed that the number of queens is probably influenced by the level of a specific ‘queen pheromone’ in the colony [6]. This pheromone is produced by queens when they initiate reproduction, and the colony level is presumably proportional to the number of reproductive queens [7]. In monogyne colonies, workers apparently start destroying queens when the pheromonal level in the colony is higher than the amount produced by a single queen. Gp-9Bb workers may have a higher pheromonal threshold than Gp-9BB workers, so that they accept several queens. Alternatively, Gp-9Bb queens may produce less of this pheromone than Gp9BB queens, allowing several queens of this genotype to be accepted in polygyne colonies. A remarkable finding of Krieger and Ross [2] is that the same gene also influences social organization in other, closely related species. A phylogenetic analysis of Gp-9 sequences of 10 species of the genus Solenopsis showed that, in the three other species of fire ants that display both monogyne and polygyne colonies, there is a perfect association between a queen genotype at Gp-9 and whether the queen was collected from a monogyne or polygyne colony (though sample sizes are small). In the three species, queens from monogyne colonies were homozygous for a Gp-9B allele very similar to the Gp-9B allele identified in S. invicta. By contrast, queens from the polygyne colonies all had one copy of a Gp-9b-like allele, again very similar to the Gp-9b alleles identified in S. invicta. Thus, the same gene influences social organization in four ant species. The Gp-9B-like allele of one of these three species (Solenopsis richteri) is the sister sequence to all other Gp-9band Gp-9B-like alleles, leading Krieger and Ross [2] to conclude that the ancestral Gp-9 allele for the socially polymorphic clade was of the Gp-9B type. This would imply that Current Biology, Vol. 12, R180–R181, March 5, 2002, ©2002 Elsevier Science Ltd. All rights reserved. PII S0960-9822(02)737-6