Honey bee aggression supports a link between gene regulation and behavioral evolution

Abstract

A prominent theory states that animal phenotypes arise by evolutionary changes in gene regulation, but the extent to which this theory holds true for behavioral evolution is not known. Because "nature and nurture" are now understood to involve hereditary and environmental influences on gene expression, we studied whether environmental influences on a behavioral phenotype, i.e., aggression, could have evolved into inherited differences via changes in gene expression. Here, with microarray analysis of honey bees, we show that aggression-related genes with inherited patterns of brain expression are also environmentally regulated. There were expression differences in the brain for hundreds of genes between the highly aggressive Africanized honey bee compared with European honey bee (EHB) subspecies. Similar results were obtained for EHB in response to exposure to alarm pheromone (which provokes aggression) and when comparing old and young bees (aggressive tendencies increase with age). There was significant overlap of the gene lists generated from these three microarray experiments. Moreover, there was statistical enrichment of several of the same cis regulatory motifs in promoters of genes on all three gene lists. Aggression shows a remarkably robust brain molecular signature regardless of whether it occurs because of inherited, age-related, or environmental (social) factors. It appears that one element in the evolution of different degrees of aggressive behavior in honey bees involved changes in regulation of genes that mediate the response to alarm pheromone.

Fig. 1.

PCA reveals effects of individual and colony genotype on aggression-related brain gene expression. Co- and cross-fostering produced the following four groups of soldiers, guards, and foragers: AA, AHB reared in AHB colony; EA, EHB in AHB colony; AE, AHB in EHB colony; and EE, EHB in EHB colony. Groups with similar expression coefficients are more similar to each other in terms of brain gene expression. The first pattern (PC1) reflects variation in brain gene expression that is similar across the four groups; PC2 is associated with differences in individual (AHB or EHB) genotype; PC3, with differences in colony genotype; and PC4, with differences in cross-fostered vs. non-cross-fostered bees.

Fig. 2.

GO functional analysis of genes associated with aggression: GO biological process and molecular function categories that were significantly enriched among the genes associated with aggression in all three experiments, i.e., as a function of heredity, alarm pheromone, and age (Table 4). Diagram represents GO categories hierarchically from top to bottom. Each box represents a GO category. Blue, up-regulation; red, down-regulation. Significantly enriched categories: (2) response to stimulus, (3) metabolic process, (6) electron carrier activity, (8) structural molecule activity, (13) oxidation reduction, (15) protein binding, (16) isomerase activity, (17) oxidoreductase activity, (18) structural constituent of ribosome, (21) detection of external stimulus, (29) inositol-3-phosphate synthase activity, (33) regulation of S phase, (35) visual perception, (36) inositol biosynthetic process, (37) regulation of S phase of mitotic cell cycle, (38) monovalent inorganic cation transmembrane transporter activity, (39) positive regulation of S phase of mitotic cell cycle, and (40) hydrogen ion transmembrane transporter activity. Other GO terms given in SI Text.

Fig. 3.

Aggression-related decrease in brain metabolism as a function of heredity, alarm pheromone exposure, and age. Results of assays of enzyme activity for Complex I (NADH dehydrogenase), IV (cytochrome c oxidase), and V (ATP synthase) performed on mitochondrial preparations from bee brains. n = 4 biological replicates for each of the six groups labeled on the x axes of the three graphs; two pools of five brains per colony per behavioral group from two colonies. n = 3 technical replicates per sample. Statistical analysis: 1-way ANOVA with Tukey HSD posthoc. EHB, EHB bees in EHB colonies; AHB, AHB bees in AHB colonies. See SI Text for methods. Means ± SD are shown.