Cognition in an Insect Brain

ongoing

There is a common perception that larger brains mediate higher cognitive capacity. Social insects, however, demonstrate that sophisticated cognition is possible with miniature brains. Honeybees display higher-order learning such as categorization, non-linear discriminations, concept learning and numerosity, which are unique among insects. These capacities are mediated by a miniature brain with only 950 000 neurons. Despite extensive behavioral analyses, no study has attempted to elucidate the neural mechanisms underpinning the higher-order learning of bees. Our current breakthrough establishing virtual-reality protocols for tethered honeybees offers a unique opportunity to uncover the minimal circuits that mediate higher-order forms of cognitive processing in the brain of a behaving bee. We have recently shown that bees learn to solve elemental and non-elemental problems in this experimental context, which allows integrating behavioral, neurobiological and computational approaches to unravel the neural mechanisms underlying non-elemental learning in the honeybee. I will combine behavioral recordings of bees learning non-linear discriminations and relational rules in a virtual reality environment, with access to their brain via multi-photon calcium imaging and multielectrode recordings of neural populations. I will determine the neural circuits of elemental and non-elemental visual learning along the visual circuits of the bee brain, and the necessity and sufficiency of these circuits for these capacities via selective knockdown and rescuing via wavelength-selective multi-photon uncaging of neurotransmitters. Data will be fed into computational models to test hypotheses about minimal neural architectures for visual cognition, working towards whole-brain modeling. This project will expand the information available on the neurobiology of insect learning, and will provide the first integral characterization of the mechanisms underlying cognition in a miniature brain.