Genomic Analysis of Animal Personality: Scout Bees

1
Genomic Analysis of Animal Personality Scout
Bees

• Zhengzheng S. Liang
• Robinson Lab
• Neuroscience Program _at_ UIUC

2
What is a Scout Bee?

• Worker bees who seek new food sources or nest
  sites independently.
• Food scouts form a minor fraction of the total
  honey bee foragers, but can vary from 5-35 of
  total foragers depending on food availability and
  season.
• Scout bees are crucial for colony success.

Recruit
Scout
Seeley Visscher 1988
3
How to Find a Scout Bee?

• Novel-feeder preference testBees who abandon
  the old feeder but continuously pursue the
  novel-feeders.

• Hive-moving method
• Bees who successfully bring back forage in a new
  location after the entire hive was moved 4-5
  miles away from the old location.

4
A personality-like trait?

• Novelty-seeking hypothesis scouting behavior is
  analogous to novelty-seeking behavior in
  vertebrates.
• A tool to suggest candidate genes
• Brain neurochemical difference in the
  dopaminergic and octopaminergic systems.
• 2. A new way to test the behavior
  novel-feeder preference test to collect scouts.
• Scout bee is a fascinating research subject in
  her own right.

5
Candidate Gene Expression
6
Brain Transcriptome of Scouts A BeeSpace Project

• Experiments
• 58 microarrays in loop design were used to
  compare individual brain mRNAs from 20 scouts and
  20 recruits.
• Result Highlights
• 10 of the genome changed. 1219 genes, mostly
  up-down 5-15, differentially regulated in a
  fine-tuning manner.
• Mitochondrial energy metabolism pathway were GO
  enriched and down-regulated in scout brains.
• Glutamate and GABA related genes were all up-
  regulated in scout brains. The products of these
  genes are known to be involved in learning,
  memory and sensory processing in honey bee
  brains.
• Candidate gene AmdopR1 was regulated in the same
  direction and to the similar degree as was
  measured by qRT-PCR.

7
Significant Differentially Expressed Genes

• Volcano plot

• Size range

Down 2/3
FDRlt0.01 568 genes
Up 1/3
FDRlt0.05 1219 genes
8
Linear Discriminant Analysis
Top 40 genes Behavior Colony
S1- scouts in colony 1 S2- scouts in colony 2 S3-
scouts in colony 3 R1- non-scouts in colony 1 R2-
non-scouts in colony 2 R3- non-scouts in colony 3
9
Class Prediction Signature Genes

• Gene selection by t - test across 40 individuals
  suggested 26 genes as good predictors.
• Leave-one-out cross validation coupled with
  machine learning algorithm showed that these
  predictor genes can assign individuals into scout
  or non-scout groups with 75 correction rate.
• 11 Up and 15 Down (scouts to non-scouts).
• Myosin alkali light chain 1, meiotic 9, fusilli,
  genghis khan, WW45 protein, GABA neurotransmitter
  transporter-1A, Lipoamide acyltransferase,
  Dromyosuppressin, Exosome complex exonuclease
  RRP42, and Odorant binding protein ASP4, Matrix
  metalloproteinase 1.

10
Principal Component Analysis Comparing Two Gene
Lists
Green non-scouts Purple scouts
pc1
pc1
pc2
pc2
Top 100 genes by p-values
26 signature genes by class prediction
11
Further Research

• Causal Genes or Pathways? Can be tested by
  pharmacological treatments.
• Brain Regional Expression Identify the brain
  functional units that are associated with
  scouting behavior.
• Nest-site Scouts vs. Food Scouts Measuring the
  food scout signature genes in the nest-site
  scouts can allow us to compare scouting
  behaviors in two different contexts.

12
Acknowledgement

• Robinson Lab
• Gene Robinson
• Trang Nguyen,
• Tom Newman,
• Cedric Alaux,
• Seth Ament.
• Rodriguez-Zas Lab
• Sandra Rofriguez-Zas
• Heather Adams

• Thesis Committee
• Charlie Whitfield
• Becky Fuller
• Justin Rhodes
• David Clayton
• Funding
• NSF FIBR grant to BeeSpace Project
• (Bruce Schatz, PI)