Activity-based anorexia alters hippocampal membrane-bound glucocorticoid receptor signaling and induces structural instability and spatial memory impairment in adolescent female rats

Background: Anorexia Nervosa (AN) is a multifactorial psychiatric disorder mainly a>ecting pubescent girls [1] characterized by body weight loss due to food avoidance and overexercise [2]. Despite a still unknown etiology, high cortisol plasma levels, index of a hyperactivated Hypothalamic-Pituitary-Adrenal (HPA) axis, were found in AN patients, suggesting its role in mediating neuroendocrine and cognitive alterations in AN [3]. In parallel, the increased susceptibility to stressful situations is also partly attributed to reduced volume of the hippocampus [4], a brain region critically involved in learning and memory. Glucocorticoids, the main molecular mediators of the HPA stress response, are known to modulate hippocampal structural organization via rapid non-genomic mechanisms [5]. Objective: To understand the impact of the AN phenotype on hippocampal neuroplasticity controlled by membrane-bound glucocorticoid receptor and on cognitive impairments, this study investigates the molecular, structural and behavioral signatures of AN induction in female rats by means of the Activity-Based Anorexia (ABA) model. Methods: At postnatal day (P)35 adolescent female rats were divided into: Controls (food ad libitum-sedentary), Food restricted (food restriction–sedentary), Exercise (food ad libitum-exercise) and ABA (food restricted- exercise). At P38, ABA rats had free access to a running wheel and limited food access (2h/day) till P42. Animals were sacrificed at the acute phase of the pathology (P42) or following 7-day recovery period with food ad libitum and no wheel access (P49). Brains were dissected to measure protein levels in homogenate and membrane fraction through Western Blot or processed to perform dendritic morphological analyses. Plasma was obtained by trunk blood from decapitation. At P42 and P49, another cohort of Controls and ABA rats performed the Spatial Order Object Recognition (SOOR) test and the spatial discrimination index (DI) was calculated. Statistical significance was assumed at p<0.05 and ascertained by repeated measures two-way ANOVA, two-way ANOVA, followed by Sidak or Tukey’s multiple comparisons tests, or t-tests as appropriate. Results: ABA rats, together with body weight loss and hyperactivity, showed increased food anticipatory activity (F (4, 88)=12,12, p<0,0001), a readout of their motivation to engage in intense physical activity. Corticosterone plasma levels were enhanced at P42 (+551±196 ng/mL vs Controls, F (1, 19)=11,38, p=0,0032) while reduced at P49 (-510±65 ng/mL, F (1,20)=5,697, p=0,0270) in ABA rats. At P42, ABA showed reduced levels of glucocorticoid receptor (-26±7% vs Controls, F (1, 20)=5,544, p=0.0289), caldesmon (-26±3%, F (1, 19)=29,12, p<0,0001) and neuroligin-1 (-17±7%, F (1, 20)=8,713, p=0,0079), molecular markers of cytoskeletal stability and glutamatergic homeostasis. Accordingly, reduced dendritic spine density (-3,15±0,36 spine/10µm vs Controls, F (1, 28)=4,706, p=0,0387) and number of mushroom-shaped spines (-11,8±4,36%, df=14, p=0,0345), together with an increased number of thin-shaped spines (+13,05±3,26%, df=14, p=0,0054) were found. These events were paralleled by impairment in spatial memory measured in the SOOR test (DI= -0,435±0,06 vs Controls, F (1, 34)=5,864, p=0,0209). These e>ects persisted even after bodyweight recovery. Conclusions: Our findings indicate that ABA induction orchestrates hippocampal maladaptive structural and functional plasticity contributing to cognitive deficits, providing a putative mechanism that could be targeted in AN patients.