Clonal raider ant brain transcriptomics identifies candidate molecular mechanisms for reproductive division of labor

Abstract

Background: Division of labor between reproductive queens and workers that perform brood care is a hallmark of insect societies. However, studies of the molecular basis of this fundamental dichotomy are limited by the fact that the caste of an individual cannot typically be experimentally manipulated at the adult stage. Here we take advantage of the unique biology of the clonal raider ant, Ooceraea biroi, to study brain gene expression dynamics during experimentally induced transitions between reproductive and brood care behavior.

Results: Introducing larvae that inhibit reproduction and induce brood care behavior causes much faster changes in adult gene expression than removing larvae. In addition, the general patterns of gene expression differ depending on whether ants transition from reproduction to brood care or vice versa, indicating that gene expression changes between phases are cyclic rather than pendular. Finally, we identify genes that could play upstream roles in regulating reproduction and behavior because they show large and early expression changes in one or both transitions.

Conclusions: Our analyses reveal that the nature and timing of gene expression changes differ substantially depending on the direction of the transition, and identify a suite of promising candidate molecular regulators of reproductive division of labor that can now be characterized further in both social and solitary animal models. This study contributes to understanding the molecular regulation of reproduction and behavior, as well as the organization and evolution of insect societies.

Keywords: Brood care; Eusociality; Gene expression; Gene regulation; Reproduction; Social behavior; Social insects; Time course.

Fig. 1

Design of the brood-swap experiment. For each biological replicate, a large source colony in the brood care phase was used to establish two colonies of 250 1-month-old workers and 100 marked ≥ 3-month-old workers. One of these colonies received approximately 250 larvae. After a full colony cycle, each colony contained a complete cohort of brood and workers and was in either peak brood care phase (with larvae) or early reproductive phase (with eggs and pupae). On the day the first eggs were laid, the 1-month-old workers were subdivided in colonies of 45 workers each. One colony from each phase served as the control colony and was given brood from the mother colony. The remaining colonies received brood from the mother colony in the opposite phase of the cycle, triggering the transition toward the alternative phase. Colonies were subsequently collected 6, 12, 24, 48, or 96 h post treatment. BR workers transitioning from the brood care phase to the reproductive phase (after larvae were removed and pupae added), RB workers transitioning from the reproductive phase to the brood care phase (after pupae and eggs were removed and larvae added), BC workers from the brood care phase with larvae (brood care phase control), RC workers from the reproductive phase with pupae (reproductive phase control)

Fig. 2

Cluster analysis of samples based on the mean gene expression of each time point, for 596 differentially expressed genes (adjusted p value ≤ 0.05). a PCA plot of brood-swap and control samples. Percentages on each axis indicate the proportion of variance explained by the indicated principal component. The blue, brown, and green ellipses show the k-means cluster assignment. The color of each sample indicates the average ovary activation score as per [77]; 0 indicates no signs of ovary activation while 4 indicates fully developed eggs are present. Sample names are as per Fig. 1. b Heatmap showing Euclidean distances between all time points. The dendrogram was constructed using the average distances between time points. The blue and green color bar above the heatmap indicates average ovary activation score, as in a. Sample names are as per Fig. 1

Fig. 3

Number of genes in clusters (enriched for DEGs) with maximal change in expression for each time interval. The distribution of such numbers across time intervals differed significantly between transitions (χ² = 1217.5, p < 0.00001, Additional file 5). This suggests that the transition from reproduction to brood care (RB transition; blue) and the transition from brood care to reproduction (BR transition; green) are associated with distinct time dynamics of gene expression

Fig. 4

Select candidate genes for the regulation of the transitions between brood care and reproduction. The plots show the expression changes over time after adding larvae (RB transition; blue) or removing larvae (BR transition; green) from the colonies. a Transferrin. b ILP2 (insulin-like peptide 2). c LGR4 (leucine-rich repeat-containing G-protein-coupled receptor 4). d Neuroparsin-a. e Queen vitellogenin. Gene expression is shown as variance-stabilized transformed read counts (which approximate log2-transformed read counts)