Individual differences in honey bee behavior enabled by plasticity in brain gene regulatory networks

Beryl M. Jones, Vikyath D. Rao, Tim Gernat, Tobias Jagla, Amy Cash-Ahmed, Benjamin E. R. Rubin, Troy J. Comi, Shounak Bhogale, Syed Sameed Husain, Charles Blatti, Martin Middendorf, Saurabh Sinha, Sriram Chandrasekaran, Gene E. Robinson

TL;DR

This study investigates the molecular underpinnings of behavioral plasticity in honey bees, revealing that individual differences in egg-laying and foraging behaviors are directly linked to variations in brain gene expression and chromatin accessibility. The research demonstrates that gene regulatory networks (GRNs) in the brain are highly predictive of an individual bee's behavior, offering a new perspective on how complex social behaviors evolve.

What Problem Does This Paper Address?

Understanding how genetic and molecular mechanisms translate into diverse behaviors within a population is a fundamental challenge in biology. While we know that gene regulatory networks (GRNs) play a crucial role in development and evolution, their specific contribution to behavioral plasticity at the individual level remains less understood. This paper tackles this gap by examining honey bee colonies where workers, normally sterile, can develop functional ovaries and engage in egg-laying, alongside their typical foraging duties. This "laying worker" scenario creates a natural experiment with a wide spectrum of behaviors, from specialized egg-layers and foragers to "generalists" performing both. The authors aim to connect this behavioral variation to the underlying molecular machinery in the bee brain.

Methods

The researchers employed a multi-pronged approach:

  • High-Throughput Behavioral Tracking: They used automated systems to monitor thousands of individual bees over several days, capturing their egg-laying and foraging activities with high resolution. This allowed for the quantification of individual behavioral specialization.
  • Brain Gene Expression (RNA-seq) and Chromatin Accessibility (ATAC-seq): Brain tissue from bees with distinct behavioral profiles was analyzed to measure gene expression levels and identify regions of accessible DNA (which can indicate regulatory activity).
  • Principal Component Analysis (PCA): This statistical technique was used to reduce the dimensionality of the behavioral and molecular data, identifying key axes of variation.
  • Gene Regulatory Network (GRN) Reconstruction: Using the ASTRIX algorithm, the authors built individualized GRN models for each bee, inferring interactions between transcription factors (TFs) and their target genes in the brain.
  • Transcription Factor (TF) Motif Analysis: They analyzed the enrichment of TF binding sites in regulatory regions associated with specific behaviors.
  • Machine Learning for Prediction: Random Forests classification was used to predict an individual bee's behavioral group based on its brain TF expression.

Datasets

The study utilized data from six laying worker honey bee colonies. Brain RNA-seq and ATAC-seq data were generated from a subset of these bees. The authors also drew upon existing genomic data for Drosophila melanogaster for TF motif characterization and compared their findings to published datasets on gene expression in other social insect species. No new benchmarks were explicitly introduced, but the study provides a rich dataset linking individual behavior to molecular profiles.

Key Insights and Ideas

  • Behavioral Continuum: Honey bee laying workers don't fall into neat categories; they exhibit a continuous spectrum of behaviors from dedicated egg-layers to specialized foragers, with a significant proportion of "generalists" doing both. This challenges the idea of strictly discrete behavioral roles.
  • Molecular Fingerprints: Distinct behavioral profiles are mirrored by distinct molecular profiles in the brain. Generalists, occupying the middle ground behaviorally, also show intermediate gene expression and chromatin accessibility patterns.
  • GRNs as Behavioral Predictors: The activity of transcription factors (TFs) within an individual bee's brain GRN is remarkably predictive of its behavior. This suggests that the regulatory logic of the brain is a key driver of behavioral specialization.
  • Plasticity is Key: The study highlights the plasticity of GRNs as a mechanism for generating individual behavioral differences, rather than solely relying on fixed genetic predispositions. This plasticity is crucial for adapting to changing social environments.
  • Evolutionary Echoes: Many of the TFs implicated in honey bee behavioral plasticity are also associated with social evolution across different bee species, suggesting that the molecular mechanisms underlying sociality are conserved and adaptable.

Key Results

  • Behavioral Specialization: Over 50% of laying worker bees engaged in egg-laying, and 28% in foraging, with 10.8% performing both.
  • Molecular Differences: Nearly half (46%) of all genes expressed in the brain showed differential expression between foragers and layers. Chromatin accessibility also differed significantly between these groups.
  • Predictive Power: A machine learning model based on brain TF expression could predict specialist behaviors (foragers and layers) with 97% accuracy, underscoring the strong link between TF activity and behavior. Generalists, however, were harder to predict, reflecting their intermediate and more variable molecular profiles.
  • TF Motif Enrichment: TF binding sites were significantly enriched in regulatory regions associated with specialized behaviors, particularly for foragers.

Limitations and Future Work

The authors acknowledge that while TF expression can predict specialist behavior, generalists were less accurately classified. This might be due to their more variable molecular profiles or the complexity of integrating multiple behavioral inputs. Future work could explore the precise mechanisms by which TFs integrate signals to produce these diverse behaviors, investigate the role of other tissues (like the ovary) in behavioral plasticity, and further explore the evolutionary implications of these GRN mechanisms across different social insect species. The study also suggests that the observed plasticity in laying worker colonies might represent a partially reversible or incomplete decoupling of reproductive and non-reproductive programs, offering a window into the evolution of eusociality.

Summary updated

Abstract

Understanding the regulatory architecture of phenotypic variation is a fundamental goal in biology, but connections between gene regulatory network (GRN) activity and individual differences in behavior are poorly understood. We characterized the molecular basis of behavioral plasticity in queenless honey bee (<i>Apis mellifera</i>) colonies, where individuals engage in both reproductive and non-reproductive behaviors. Using high-throughput behavioral tracking, we discovered these colonies contain a continuum of phenotypes, with some individuals specialized for either egg-laying or foraging and 'generalists' that perform both. Brain gene expression and chromatin accessibility profiles were correlated with behavioral variation, with generalists intermediate in behavior and molecular profiles. Models of brain GRNs constructed for individuals revealed that transcription factor (TF) activity was highly predictive of behavior, and behavior-associated regulatory regions had more TF motifs. These results provide new insights into the important role played by brain GRN plasticity in the regulation of behavior, with implications for social evolution.
BibTeX
@inproceedings{jones2020individual,
  author = {Beryl M. Jones and Vikyath D. Rao and Tim Gernat and Tobias Jagla and Amy Cash-Ahmed and Benjamin E. R. Rubin and Troy J. Comi and Shounak Bhogale and Syed Sameed Husain and Charles Blatti and Martin Middendorf and Saurabh Sinha and Sriram Chandrasekaran and Gene E. Robinson},
  title = {{Individual differences in honey bee behavior enabled by plasticity in brain gene regulatory networks}},
  booktitle = {eLife},
  year = {2020},
  doi = {10.7554/elife.62850},
}
Figure