Neuronal cicuits and reinforcement mechanisms underlying feeding behaviour

Change log
Huang Cao, Zhen Fang 

Animal survival depends on the brain’s ability to detect the energetic state of the body and to alter behaviour in order to maintain homeostasis. Current research in the control of food consumption stresses the importance of identifying and establishing the specific roles of homeostatic neurons, which sense the body’s energetic state and elicit complex and flexible food seeking behaviours. Recent developments in optogenetics, molecular genetics, and anatomical techniques have made these investigations possible at the resolution of specific cell types and circuits. These neurons are of particular interest because they serve as key entry points to the identification of downstream circuits and reinforcement mechanisms that control feeding behaviour. This dissertation probes the role of two kinds of homeostatic neurons— agouti-related peptide (AGRP) in the arcuate nucleus (ARC) and leptin receptor (LepRb) neurons in the lateral hypothalamic area (LHA)—in the control of food intake. First, I examined the role of LepRb neurons in the LHA in feeding. Results from electrophysiological studies indicate that these neurons consist of a subpopulation of homeostatic sensing LHA γ-aminobutyric acid (GABA) expressing neurons. In addition to their response to leptin, these neurons are capable of modulating their activity in response to changes in glucose levels, further substantiating their role as homeostatic sensing neurons. Behavioural studies using optogenetic activation of these neurons show that their elevated activity is capable of reducing body weight, although their role in modulating feeding remains unclear. Second, I investigated the reinforcement mechanisms employed by AGRP neurons to elicit voracious food consumption and increased willingness to work for food. Conditioned place avoidance studies under optogenetic activation of AGRP neurons reveal that their increased activity has negative valence and is avoided. In addition, imposition of elevated AGRP neuron activity in an operant task reduced instrumental food seeking with particular sensitivity under high effort requirements. Taken together, these results suggest that AGRP neurons employ a negative reinforcement teaching signal to direct action selection during food seeking and consumption. Third, I systematically analyzed the contribution of specific AGRP neuron projection subpopulations in AGRP neuron mediated evoked-feeding behaviour. Optogenetic activation studies of AGRP neuron axons in downstream projection regions indicate that several, but not all, subpopulations are capable of independently evoke food consumption. This work reveals a parallel and redundant functional circuit organization for AGRP neurons in the control of food intake. Interestingly, all AGRP neuron subpopulations examined displayed similar modulation by states of energy deficit and signals of starvation, despite their apparent divergence in function. As a whole, this dissertation extends our understanding of the role of homeostatic neurons in food consumption and uncovers previously unappreciated functional organization and reinforcement mechanisms employed by neuronal circuits that control feeding behaviour.

Feeding behaviour, AGRP, Leptin, Reinforcement, Neuronal circuits
Doctor of Philosophy (PhD)
Awarding Institution
University of Cambridge
This work was supported by the Howard Hughes Medical Institute Graduate Scholar program.