The interplay of anxiety, hormones and the GABAergic system in adolescent females: Insights into the cellular mechanisms underlying behavior in an animal model of anorexia nervosa

I used an animal model of anorexia nervosa (AN), activity-based anorexia (ABA) to study the biobehavior and possible underlying molecular mechanisms in AN. ABA mimics the core phenotypic features of AN, namely, starvation, weight loss and excessive exercise (i.e., hyperactivity). Anxiety is a common symptom in AN that correlates with the severity of exercise. I have strengthened the face validity of the animal model by showing a strong correlation between anxiety-like behavior and wheel exercise in ABA-induced mice. This was important because the high co-morbidity between AN and anxiety may be due to a shared pathology in neural circuitry which can be investigated using the animal model. One region that could co-regulate behavior in anxiety and AN is the amygdala. There is considerable evidence pointing to the roles of the amygdala and the GABAergic inhibitory signaling system in the brain in anxiety behavior. Therefore, I investigated whether the GABAergic signaling pathway was changed as a result of ABA induction, thereby contributing to the hyperactivity. Using electron microscopic immunocytochemistry (EM-ICC), I found that a particular subunit, namely alpha4, of the GABAA receptor (GABAAR) is increased in putatively inhibitory neurons in the caudal basal amygdala of ABA-induced rats. This could lead to an increased outflow from the amygdala and contribute to increased anxiety and wheel activity. This study has implications for pharmacotherapy in AN because alpha4-containing GABAARs are insensitive to benzodiazepine class of anxiolytic drugs but are modulated strongly by the metabolite of the female hormone progesterone, allopregnanolone.;AN is much more common in females, with the onset most frequently in adolescence, when progesterone level fluctuates widely, and has a high frequency of relapse. The epidemiological features suggest a role for female sex hormones in the etiology of AN. Therefore, I investigated the effect of progesterone administration on the response to a second ABA induction, as measured by level of anxiety and hyperactivity. While progesterone did not change the level of anxiety, it exacerbated the hyperactivity response to a second ABA induction, specifically among the population of mice that had attained resilience during the recovery phase between the 1st and 2nd ABA inductions. Progesterone had no effect on the behavior of mice that were more vulnerable to the repeat episode of ABA. In our previous work, vulnerability to ABA was reliably reflected by the running activity in the period preceding the ABA induction. These data help to predict individuals whose response to a repeat ABA stressor will be exacerbated by progesterone treatment.;On investigating the cellular mechanism underlying the individual variability in the development of vulnerability/resilience to a repeat episode of a stressor such as starvation in ABA, I found an association between resistance developed before the repeated episode and high expression of the alpha4 subunit of the GABAAR in the hippocampus. Yet, a sensitized response to the repeated episode of ABA was associated with a high level of the alpha4 subunit expression, especially with progesterone administration. This provides further evidence of the interaction of hormones with the history of stress response.;Having found a potential role for alpha4-containing GABAARs in the response to ABA, I subjected mice with a reduced (partial knockout (KO)) or obliterated (complete KO) expression of the alpha4 subunit to ABA and compared their response to their littermates with full expression of the subunit. I found no differences in response to ABA between animals with full, reduced or zero expression of the alpha4 subunit. This indicates that other types of GABAARs may compensate for the global KO of alpha4-containing GABAARs. Alternatively, KO in different regions may have canceled out effects of each other. For example, if the alpha4 containing GABA AR in the hippocampus played a protective role and the alpha4 containing GABAAR in the basolateral amygdala played an exacerbating role, removal of this receptor from both locations will not affect the overall severity with which the animal will respond to ABA.;Overall, my findings (1) establish the face validity of the ABA model in studying the co-morbidity of anxiety and AN, (2) question the therapeutic potential of progesterone in AN, (3) demonstrate the importance of taking individual variability into consideration in analysis of behavior and molecular mechanisms, (4) suggest that hormones may not have a uniform effect on all individuals but may interact with the history of each individual, and (5) caution that a whole brain genetic manipulation may not be insightful to the etiology of ABA and AN. Further studies are needed to establish the utility of targeting GABAergic signaling in specific brain regions to ameliorate the maladaptive behavior in ABA and AN. (Abstract shortened by UMI.).