Before we go on, lest we commit the all-too-common modern error of losing sight of the organism for the genes, do take a look at this astounding high-speed video of a bee taking off, by Sean McCann. For further photos and videos of Hymenoptera, see Sean's blog post here (h/t Ed Yong). They are gorgeous. Sean has kindly given us permission to post his video here.
Characteristics of eusocial insects, ants, bees and aculeate or stinging wasps of the order Hymenoptera, include reproductive division of labor, cooperative brood care, and overlapping generations. Simola et al. compared the genomes of seven ants, the honeybee and various solitary insects to try to identify genomic features that were common to the eusocial insects, even though these ants and the bee are from evolutionarily independent lineages and differ in various significant respects. The question was whether they had enough in common to elucidate the genomic basic of eusociality.
This paper is a goldmine of information about ant genomes, their structure and evolution. For example, the authors write that many 'taxonomically restricted genes' (TRGs) are found in ant genomes, but they did not find a "social toolkit" of conserved protein coding genes for eusociality. They found that Hymenoptera have a greater number of taxonomically restricted genes than do solitary insects, and that they evolved faster. They identified gene families that expanded or constricted depending on, e.g., the diet of a particular ant species, and they found fairly insignificant differences between the immune genes of ants and solitary insects. Distinct methylation patterns were found in ant genomes. And much more.
Did they find shared genomic features responsible for eusociality? They identified 64 genes conserved among the 7 ant lineages they analyzed, but these genes are not expressed in a caste-specific pattern, so can't explain the phenotypic plasticity of these insects. Nor do they encode any known protein domains. Thus, the authors write, "These results suggest that a broad "social toolkit" of conserved de novo protein-coding genes is not a requirement for eusociality."
...key changes in gene regulation (both in cis and in trans) may have convergently evolved in the early stages of eusocial evolution, whereas changes in gene composition may have been more important for lineage-specific social and ecological adaptations. Ant intergenic sequences are enriched for regulatory elements, including miRNAs, noncoding RNAs and [transcription factor]-binding sites that are conserved in both DNA sequence and homologous genomic position. Nearly 2000 genes share similar cis-regulatory changes in eusocial compared with solitary insects, and most of these changes are similar in both eusocial lineages of ants and honeybees.Many of these genes code for proteins with neuronal and hormonal functions, and changes in gene regulation seem to be associated with increasing complexity of social organization in ants. A handful of gene families that seem to be involved in regulating gene transcription have gained members during early ant evolution. The authors conclude that changes in gene regulation over evolutionary time are associated with the evolution of eusociality, and they cannot pinpoint any small set of genes or regulatory elements for eusociality or phenotypic plasticity.
They further suggest that
...organization of eusocial insect genomes appears to harbor sufficient degrees of freedom to allow convergence of higher-order complex traits, such as eusociality, from unique, lineage-cspeific evolutionary trajectories that involve distinct genes an modes of regulation. Such genomic complexity may be especially engendered by ants, where extreme reproductive divisions of labor resulting from a eusocial lifestyle may effectively reduce the strength of natural selection, thereby facilitating rapid sequence divergence among lineages.This is very good work and very interesting but one needs to be circumspect about it and avoid any over-interpretation. The paper reports numerous findings about the structure and evolution of the genomes of eusocial insects, and it's appropriately nuanced, but any report of genetic links with behavior in non-humans is taken by some as evidence that there must be adaptive genes that confer social behavior in humans. The usual view is that genes can be found that predict specific behavior.
However, sociality in humans heavily depends on language and symboling, is very different from what is seen in our closest relatives, and evolved very rapidly. This make it much less likely to be a major or simple genetic adaptation, and more likely the result of generic non-programming of our neural ware, or the result of whatever made our symboling mental behavior possible.
Sofware rather than hardware-dependent social behavior is what makes humans so unique, and that it can occur (if our surmises are right) without genetic programming makes us so interesting as well.