Charles W. Whitfield


Associate Professor

Department of Entomology

Neuroscience Program

Institute for Genomic Biology

Program in Ecological and Evolutionary Biology


B.S., 1994, University of California, Davis

Ph.D., 2000, Stanford University


Phone: 217-244-2889; FAX 217-244-3499; Email:


Molecular mechanisms and molecular evolution of social behavior, using social insects as a model. Natural variation in individual behavior. High throughput genotyping and gene expression analyses.


My lab studies the molecular mechanisms that underlie social behavior. We are attempting to identify the genes and gene regulatory elements that have been gained, lost, or modified in the evolution of socially organized behavior. An important feature of social organization is division of labor, which involves variation in role (or behavioral repertoire) between individuals in a group. In social insects (bees, ants, wasps and termites) and in some mammals (mole-rats), division of labor involves both morphological differentiation (castes) and behavioral specialization in the absence of morphological differentiation. In the honey bee colony (Apis mellifera), individuals within the same morphological caste (workers) specialize on different tasks such as cleaning cells, feeding larvae (nursing), building comb, storing food, ventilating the hive, guarding the hive, removing dead bees (undertaking), or foraging for food and water. These individual behavioral phenotypes are semi-stable, lasting for hours, days, or weeks, and are a product of intrinsic factors (age and genotype) and extrinsic factors (social cues and the physical environment). Because these factors can be manipulated in a naturalistic social context in the colony, the honey bee provides an ideal model system to study how nature and nurture shape complex behavior, underlie variation between individuals, and contribute to structure in a highly organized animal society.


1) Mechanistic and evolutionary approaches to genes and social behavior. We have identified many genes that exhibit expression differences in the honey bee brain associated with long-term, socially regulated behavioral phenotype (Whitfield et al. Science. 2003. Oct 10; 302:296-99; see figure below). This association is so robust that mRNA profiles in single dissected brains can correctly predict behavior phenotype (nurse or forager) for 57 out of 60 individual bees in this study (and 56 out of 60 bees in a subsequent, unpublished study). Many or perhaps most of these genes are regulated also by treatment with a Juvenile Hormone analog (Whitfield, Ben-Shahar, Brillet, LeConte, Rodregues-Zas, Robinson, unpublished) and/or pheromones (Grozinger, Sharabash, Whitfield, and Robinson. PNAS. 2003. Nov 25; 100:14519-25) in the absence of task-related experience, suggesting that these genes may act downstream of important causal signals that pace the onset of foraging behavior. These data provide a wealth of testable hypotheses. For example, behavioral phenotype was associated with expression levels of a particular MAP kinase of unknown function (an ortholog of ERK8 in humans and hypothetical genes in D. melanogaster and C. elegans) but not other MAP kinases (e.g., orthologs of ERK1/2). Is behavioral phenotype associated with post-translational modifications specifically in ERK8, such as protein phosphorylation or nuclear translocation? Can pharmacological manipulation of MAP kinase signaling modify bee behavior and social organization? My lab will investigate this and other candidate behavior genes using a variety of approaches and model systems. Where in the brain are these genes expressed and where are the protein products? Are certain brain regions hot spots for behavior-associated gene expression? (Note that microarray analyses of brain sub-regions is feasible using current methods.) How are translation and post-translational modifications coupled to environmental cues (including social cues) that regulate behavioral phenotype? What are the functions of these genes in the genetic model systems, D. melanogaster and C. elegans (which are solitary in behavior and do not exhibit socially regulated division of labor)? Have these genes acquired novel translational or post-translational regulation in the evolution of sociality in the honey bee? The latter question can be addressed in the family Apidea by comparison of species that are solitary (orchid bees) primitively social (bumble bees) and eusocial (honey bees and stingless bees).



Figure from Whitfield et al. Science. 2003. Oct 10; 302:296-99. Columns represent single dissected brains from 60 individual bees. Rows indicate gene expression measured for 17 microarray cDNAs (corresponding to 14 different genes) expressed in the brain. Bees from typical colonies were collected while performing age-typical behaviors. YN = young nurses; OF = old foragers. Bees typically shift from hive tasks such as nursing to foraging at about 2 3 weeks of age. However, in colonies established with only young workers (single-cohort colonies), some bees accelerate this transition becoming precocious young foragers (YF). Later, as the entire population ages in the single-cohort colony, some bees remain as overaged old nurses (ON) despite advancing age. Young = 7 2 days of age; old = 30 2 days of age.


2) Genetics of social behavior and genetics of gene expression in the brain. The goal is to identify the genes (and ultimately DNA polymorphisms) that cause heritable variation in behavior, particularly for behaviors that contribute to social organization in the bee colony. Genetic differences for many such behaviors exist between natural honey bee populations (subspecies), and these subspecies exhibit extensive differences in gene expression in the brain when co-fostered in the same colony (Whitfield, Brillet, LeConte, Robinson, unpublished). Because these subspecies can be hybridized, a genetic mapping and positional cloning approach is possible. This approach can be applied to both heritable differences in behavior and heritable differences in gene expression in the brain. Brain gene expression is, itself, a large set of heritable phenotypic traits that are influenced by cis-acting polymorphisms (i.e., in regulatory regions of the genes themselves) or trans-acting polymorphisms (i.e., in regulatory or coding regions of other genes, such as transcription factors that can influence the expression of many genes). Positional cloning efforts will benefit from genomic sequence and a large set of candidate genes identified in many brain/behavior gene expression studies. Combined with approaches described above, these studies will address several questions: What genes cause heritable differences in behavior? What genes cause heritable differences in gene expression in the brain? Are these the same genes? Are the genes that shape behavior over evolutionary time-scale (via DNA polymorphisms) the same genes as those that modify behavior in the life of an individual (via regulated expression changes)?


My lab will develop a large, SNP-based genotyping resource for the honeybee using Illumina bead array technology. This will allow genotyping at up to 1500 makers per individual in 100s of individuals, at costs of less than $0.10 per genotype. Initial studies will focus on the drone (male) to take advantage of increased mapping resolution and simplified genetics in haploids (see below). Mapping of gene expression traits (eQTLs) will be facilitated by an inexpensive, in-house gene expression platform for the bee that allows analyses of single dissected brains. Genetic studies will benefit from other advantages in the honey bee, including the highest recombination rate known in any higher eukaryote.


Behavioral genetics in the honey bee

The good:

  • Controlled mating via artificial insemination (not feasible in most social insects).
  • Production of up to 2000 progeny per day from a single mated queen, which can live from one to several years.
  • Haplodiploidy. Males are haploid (derived from unfertilized eggs); females are diploid. (Mating a virgin queen to her sons is effectively a self-cross.)
  • Highest recombination rate known in any higher eukaryote (1 cM = ~55 kb)
  • Relatively small genome (~270 Mb).
  • Genome sequencing project well underway (sequence generated from 2 haplotypes provides virtually unlimited SNP and microsatellite markers).
  • Due to a long history of study, the bee is perhaps the most well-understood single species at the level of behavior. Well established methods including marking, observation, and collection of individuals in a naturalistic or experimentally manipulated social context.
  • Characterized differences in behavior between natural populations (subspecies) that can be hybridized.

The bad:

  • The sex-determination system in hymenoptera (by genetic complementation) makes inbreeding difficult, although not impossible. 97% inbred lines have been created in the past but none exist at this time.
  • Generation time is short in theory (< 20 days egg to egg), but in practice using current beekeeping methods is limited to about 3 generations per year.

The ugly:

  • Strain maintenance is problematic. Several selected strains and closed populations are currently maintained, but maintenance of large numbers of strains (e.g., mutant, transgenic, or recombinant inbred lines) will not be feasible using current beekeeping methods. One possible workaround for this problem is collection and storage of sperm, which is routinely collected for artificial insemination, can be stored indefinitely, and is isogenic due to haploidy in males.


3) Genomics and bioinformatics of social behavior. Although complex, gene expression is highly coordinated and can be partitioned into a small number of themes using bioinformatic methods such as clustering and principal component analysis. Analysis of gene clusters or gene expression components can provide a more holistic understanding of system function than traditional one-gene-at-a-time approaches. Genomic approaches can be used to test hypotheses in the global sense, such as whether age-related changes in brain gene expression are associated with age per se or with age-related and socially modulated changes in behavior (see Whitfield et al. Science. 2003. Oct 10; 302:296-99). Another goal of bioinformatics is identification of DNA sequence elements that regulate gene transcription. This is particularly important in understanding the evolution of social behavior because many (or perhaps most) of the relevant changes are likely to be in gene regulatory sequence rather than gene coding sequence. Computational identification of gene regulatory elements is possible from gene expression and genomic sequence information, using sophisticated mathematical approaches such as Bayesian network analysis. My lab will pursue such computational approaches, sampling a large set of gene expression data from the honey bee brain which was collected under experimental conditions known to perturb brain function (behavior being the ultimate function of the brain). The questions addressed by computational approaches are complementary to those addressed by mechanistic and comparative approaches above. What are the components of brain gene expression that are associated with behavioral phenotype, and how do these components respond to social and environmental cues that regulate behavior? What are the regulatory sequence elements that couple important social/environmental cues to transcription in the brain? Which kinds of behavioral phenotypes are associated with distinct gene expression signatures in the brain, and which are mediated by post-transcriptional mechanisms only? (I.e., What are the limits of genomic plasticity in behavior?)


Selected Publications:


Zayed A, Whitfield CW. A genome-wide signature of positive selection in ancient and recent invasive expansions of the honey bee, Apis mellifera. Proc Natl Acad Sci U S A. 2008. Mar 4; 105(9): 3421-6.

Sen Sarma M, Whitfield CW, Robinson GE. Species differences in brain gene expression profiles associated with adult behavioral maturation in honey bees. BMC Genomics. 2007. Jun 29;8:202.

Whitfield CW, Behura SK, Berlocher SH, Clark AG, Johnston JS, Sheppard WS, Smith DR, Suarez AV, Weaver D, Tsutsui ND. Thrice out of Africa: ancient and recent expansions of the honey bee, Apis mellifera. Science. 2006. Oct 27; 314(5799):642-5.

Honeybee Genome Sequencing Consortium. Insights into social insects from the genome of the honeybee Apis mellifera. Nature. 2006. Oct 26; 443(7114):931-49. (Leader, Population Genetics and SNPs section)

Whitfield CW, Ben-Shahar Y, Brillet C, Leoncini I, Crauser D, Leconte Y, Rodriguez-Zas S, Robinson GE. Genomic dissection of behavioral maturation in the honey bee. Proc Natl Acad Sci U S A. 2006. Oct 31; 103(44):16068-75.

Sinha S, Ling X, Whitfield CW, Zhai C, Robinson GE. Genome scan for cis-regulatory DNA motifs associated with social behavior in honey bees. Proc Natl Acad Sci U S A. 2006. Oct 31; 103(44):16352-7.

Rodriguez-Zas SL, Southey BR, Whitfield CW, Robinson GE. Semiparametric approach to characterize unique gene expression trajectories across time. BMC Genomics. 2006. Sep 13; 7:233.

Cash A, Whitfield CW, Ismail N, Robinson GE. Behavior and the limits of genomic plasticity: power and replicability in microarray analysis of honey bee brains. Genes Brain Behav. 2005. Jun; 4(4):267-71.

Robinson GE, Grozinger CM, Whitfield CW. Sociogenomics: social life in molecular terms. Nat Rev Gen. 2005. Apr; 6(4):257-70.

Grozinger CM, Sharabash NM, Whitfield CW, Robinson GE. Pheromone-mediated gene expression in the honey bee brain. PNAS. 2003. Nov 25; 100:14519-25.

Whitfield CW, Cziko AM, Robinson GE. Gene expression profiles in the brain predict behavior in individual honey bees. Science. 2003. Oct 10; 302(5643):296-99.

Whitfield CW, Band MR, Bonaldo MF, Kumar CG, Liu L, Pardinas JR, Robertson HM, Soares MB, Robinson GE. Annotated expressed sequence tags and cDNA microarrays for studies of brain and behavior in the honey bee. Genome Research. 2002. Apr; 12(6):555-66. (cover story)