English
Understanding how the brain achieves higher-order cognitive processing, determining the basic neural architecture underlying such processing, and achieving simultaneous recordings of behavioral decisions and neural activity during complex learning tasks are, so far, ambitious and difficult goals for scholars in the field of neurosciences, psychology, and ethology. Recent studies in an insect model, the honeybee Apis mellifera, have uncovered unsuspected higher-order cognitive processing beyond simple associative learning. Despite their miniature brain, freely-flying bees are capable of non-elemental learning, in the form of object categorization and concept learning of varying complexity. The neural substrates mediating these capacities can be thoroughly investigated by means of a variety of invasive techniques if the insects are immobilized. Yet, the missing link between these two levels (behavior and neurobiology) resides in the fact that neural analyses are difficult in freely moving animals. The present project attempts to fill this gap and to provide a comprehensive view of the neural bases of cognitive processing in the miniature brain of honeybees. Our specific goal is to understand how a relatively ‘simple’ brain forms concepts and learn abstract rules. In the last years, several investigations by the applicant have shown that freely-flying bees can learn abstract concepts (e.g. sameness, above of, larger than, etc) in the visual domain. We will use a combination of behavioral and neurobiological methods in order to quantify and manipulate brain activity in tethered bees that, despite movement restrictions, will be able to solve conceptual discriminations. To this end, we will use a locomotion compensator designed for the analysis of visual orientation of bees, i.e. an experimental setup in which tethered bees will be placed on a hollow Styrofoam sphere whose movements, induced by the walking bee, are recorded by optic sensors thereby allowing reconstruction of the walking trajectory of the bee. Visual stimuli surrounding the setup will be reinforced or non-reinforced so that the bee will have to solve visual tasks that will be arranged in the form of elemental or conceptual discriminations (i.e. rule learning). At the same time, we will access the brain of the tethered bees to study the neural correlates of these two learning forms. Neuroanatomical analyses of neuropile volume and synaptic branching will be performed at the level of the optic lobes, central body and mushroom bodies, both in naïve and in conditioned animals subjected either to elemental or conceptual visual discriminations. In this way, we will determine when and where plastic changes occur in the honeybee brain as a consequence of elemental and higher-order visual discriminations. Reversible and selective inactivation of visual neuropiles (optic lobes and central complex) and mushroom bodies will be achieved by local injection of the anesthetic procaine. In this way, the necessity and sufficiency of these structures for elemental and conceptual visual learning will be determined. Multielectrodes will be implanted in the optic lobes and the mushroom bodies to record neural activity upon elemental and conceptual visual learning. Changes in the neural signatures of the stimuli/rules to be discriminated will be quantified in these regions of the bee brain. The combination of behavioral and neurobiological methods will allow determining 1) the neural structures that are necessary and sufficient for solving elemental and conceptual problems in the visual domain, 2) the dynamics of neural activity underlying conceptual learning in bees and 3) the minimal neural circuit required for achieving such higher-order learning. The knowledge gained in this project will be implemented in a neural-based model of conceptual learning that will constitute a fundamental contribution for scholars in the fields of psychology, neurosciences and informatics.
