Evolutionary diversification of alternative splicing along a social complexity gradient in aculeata insects

The evolution of animal sociality is characterized by profound phenotypic differentiation originating from a shared genome. While differential gene expression is known to contribute to caste development, the evolutionary dynamics of post-transcriptional regulation, specifically alternative splicing, remain poorly quantified across the social spectrum. Here, we investigate the genome-wide alternative splicing landscape using bulk brain transcriptomes of adult females across 13 Aculeata species, representing a continuous gradient of social complexity. Utilizing Phylogenetic Generalized Least Squares regression to control for genomic structural constraints and sequencing depth variations, we show that increasing social complexity statistically correlates with a genome-wide reduction in the splicing specificity index, where lower values indicate more even isoform usage and thus higher splicing diversity. This indicates a genome-wide shift toward a more uniform distribution of transcript isoforms, reflecting increased evenness in tissue-level splicing patterns. However, competitive modeling revealed that the absolute and proportional frequencies of caste-biased differential alternative splicing events are statistically constrained by foundational genomic architecture and technical variance, independent of the social complexity gradient. Despite this quantitative constraint, cross-lineage comparisons identified a shared subset of orthologous genes consistently exhibiting caste-biased differential alternative splicing across independent eusocial origins. These shared loci are non-randomly concentrated within specific neural and behavioral pathways, indicating that independent lineages exhibit localized splicing variations within a conserved biological framework during caste differentiation. While these findings establish splicing diversification as a quantitative transcriptomic signature of sociality, current bulk RNA-seq approaches cannot distinguish whether this pattern represents enhanced intracellular regulatory plasticity or increased cellular heterogeneity inherent in complex brains. Resolving this distinction will require future single-cell transcriptomic analyses.