Targeting the food reward system in preclinical models for eating disorders

Eating disorders, such as anorexia nervosa (AN) and binge eating disorder (BED), pose a significant health burden, yet current treatment options fail to reach and cure all patients. Furthermore, there is still much to be discovered about the underlying pathophysiology of eating disorders, knowledge crucial for developing more effective, biology-driven medications. Given the abnormal processing of food rewards in both AN and BED, this thesis aimed to explore potential targets within the food reward system in these disorders. In Chapter 2, we investigated the impact of different doses of olanzapine in the rat ABA model. Additionally, we explored intranasal administration as a rapid and effective delivery method for olanzapine in the ABA model. We found that olanzapine, both systemic and intranasal administration, decreased running wheel activity in the ABA model in rats. As intranasal administration mainly reduced running in the first hours after administration, we recommend further exploration of this approach in patients with AN. The microbiome is a potential future treatment for AN as alterations in microbiome composition are found in these patients. Whether these microbiome alterations also contribute to AN-related symptoms remains uncertain. In Chapter 3, we transplanted the microbiota of patients with AN in rats before they underwent the ABA model. We demonstrated that fecal microbiota transplantation from patients with AN did not contribute to weight loss or changes in activity, although limited microbiota successfully colonized the rat's gut. However, the only antibiotic-treated group decreased survival, body weight, and body temperature. In Chapter 4, we studied whether AN microbiota altered flexible value-based decision-making in the reversal learning task and related dopamine signaling and anxiety. Microbiota fecal transplantation did not alter flexible learning, ventral tegmental area (VTA) dopamine signaling and anxiety, although the transplantation efficiency was low. Furthermore, the antibiotic treatment decreased task engagement acutely but did not alter flexible learning or VTA dopamine activity. In Chapter 5, we investigated semaglutide’s effect on VTA dopaminergic neuron signaling during food-predicting cues and sucrose reward consumption. We found that the semaglutide decreased food-seeking behaviors in the sucrose conditioning task. At the same time, the VTA dopamine activity increased during sucrose consumption but was unchanged during the food-predictive cue. In Chapter 6, we explored the influence of ghrelin on VTA dopamine neuron signaling during a reward-seeking task. We discovered that ghrelin increased VTA dopamine neuron activity in response to a food-predictive cue and increased food-seeking behaviors. Additionally, prefeeding food-restricted mice with chow lowered endogenous ghrelin levels, decreased food-seeking behaviors and reduced VTA dopamine neuron activity during cue presentation. Besides dopamine neurons in the VTA, there is also a proportion of gamma-aminobutyric acid (GABA) neurons. The VTA GABA-neurons too are suggested to play a role in cue and reward signaling. In Amendment 1, we observed changes in VTA GABA neuron activity during the food-predictive cue and reward collection. Interestingly, the VTA neuronal activity during the food-predictive cue depended on the reward collection following the cue, both for the dopamine and GABA neurons.