Activity-Based Anorexia Dynamically Dysregulates the Glutamatergic Synapse in the Nucleus Accumbens of Female Adolescent Rats

Abstract

Intense physical activity and dieting are core symptoms of anorexia nervosa (AN). Their combination evolves into compulsivity, leading the patient into an out-of-control spiral. AN patients exhibit an altered activation of nucleus accumbens (NAc), revealing a dysfunctional mesocorticolimbic reward circuitry in AN. Since evidence exists that a dysregulation of the glutamate system in the NAc influences reward and taking advantage of the activity-based anorexia (ABA) rat model, which closely mimics the hallmarks of AN, we investigated the involvement of the glutamatergic signaling in the NAc in this experimental model. We here demonstrate that food restriction causes hyperactive and compulsive behavior in rodents, inducing an escalation of physical activity, which results in dramatic weight loss. Analysis of the glutamate system revealed that, in the acute phase of the pathology, ABA rats increased the membrane expression of GluA1 AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid) receptor subunits together with its scaffolding protein SAP97. Recovery of body weight reduced GluN2A/2B balance together with the expression of their specific scaffolding proteins, thus suggesting persistent maladaptive neurotransmission. Taken together, AMPA and NMDA (N-methyl-D-aspartate) receptor subunit reorganization may play a role in the motivational mechanisms underlying AN.

Keywords: activity-based anorexia; adolescence; female rats; glutamatergic receptors; nucleus accumbens; reward circuit.

Figure 1

(a) Schematic representation of the experimental paradigm performed in adolescent female rats to induce the activity-based anorexia (ABA) phenotype; (b) average daily body weight; and (c) food intake, in control (CTRL), food-restricted (FR), exercise (EXE) and ABA rats. Results are presented as the mean ± SEM. ^### p < 0.0001 FR and ABA vs. CTRL and EXE; * p < 0.05, ** p < 0.01 CTRL vs. FR; ^° p < 0.05, ^°° p < 0.01, ^°°° p < 0.001 ABA vs. CTRL; ^$ p < 0.05, ^$$ p < 0.01, ^$$$ p < 0.001 ABA vs. EXE; ^& p < 0.05 FR vs. ABA; ^§ p < 0.05 ABA and EXE vs. CTRL and FR. (two-way analysis of variance (ANOVA) with repeated measures followed by Bonferroni’s multiple comparisons test).

Figure 2

Running activity on the wheel in terms of (a) distance travelled in meters, (b) mean speed in meters/minute, (c) maximum speed in meters/minute, (d) brief- and (e) long-exercise sequences expressed as the total number of access lower and higher than 2 meters each respectively exhibited by EXE and ABA rats. Results are presented as the mean ± SEM. ^$ p < 0.05, ^$$ p < 0.01, ^$$$ p < 0.001 ABA vs. EXE (two-way ANOVA with repeated measures followed by Bonferroni’s multiple comparisons test). FR = food-restricted; EXE = exercise; ABA = activity-based anorexia.

Figure 3

Effect of the ABA induction on the AMPA receptor subunits composition in the whole homogenate (left) and in the crude membrane fraction (right) of the Nucleus Accumbens (NAc) measured in the acute phase of the pathology (postnatal day PND 42) and after a 7-days recovery period (PND 49). Protein levels of GluA1 and GluA2 subunits expressed as GluA1/A2 ratio are shown in the (a) homogenate and (b) crude membrane fraction at PND 42, and (c) in the homogenate and (d) crude membrane fraction at PND 49. Results are expressed as percentages of controls and represent the mean ± SEM of five-six rats per group. Panel (e) shows representative immunoblots for GluA1 and GluA2. *** p < 0.001 vs. Food ad libitum-sedentary; ^$ p < 0.05, ^$$ p < 0.01 vs. Food restriction-sedentary, ^# p < 0.05, ^### p < 0.001 vs. Food ad libitum-exercise (two-way ANOVA followed by Tukey’s multiple comparisons test). CTRL = control; FR = food-restricted; EXE = exercise; AMPA: (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid); ABA = activity-based anorexia.

Figure 4

Effect of the ABA induction on SAP97 scaffolding protein expression in the whole homogenate (left) and in the crude membrane fraction (right) of the NAc measured in the acute phase of the pathology (PND 42) and after a 7-days recovery period (PND 49). Protein levels of SAP97 are shown in the (a) homogenate and (b) crude membrane fraction at PND 42, and in the (c) homogenate and (d) crude membrane fraction at PND 49. Results are expressed as percentages of controls and represent the mean ± SEM of five-six rats per group. Panel (e) shows representative immunoblots for SAP97. ^$ p < 0.05, ^$$ p < 0.01, ^$$$ p < 0.001 vs. Food restriction-sedentary, ^# p < 0.05 vs. Food ad libitum-exercise (two-way ANOVA followed by Tukey’s multiple comparisons test). CTRL = control; FR = food-restricted; EXE = exercise; ABA = activity-based anorexia.

Figure 5

Effect of the ABA induction on GRIP scaffolding protein expression in the whole homogenate (left) and in the crude membrane fraction (right) of the NAc in the acute phase of the pathology (PND 42) and after a 7-days recovery period (PND 49). Protein levels of GRIP are shown in the (a) homogenate and (b) crude membrane fraction at PND 42, and in the (c) homogenate and (d) crude membrane fraction at PND 49. Results are expressed as percentages of controls and represent the mean ± SEM of five-six rats per group. Panel (e) shows representative immunoblots for GRIP. ^$ p < 0.05, ^$$$ p < 0.001 vs. Food restriction-sedentary, ^# p < 0.05 vs. Food ad libitum-exercise (two-way ANOVA followed by Tukey’s multiple comparisons test). CTRL = control; FR = food-restricted; EXE = exercise; ABA = activity-based anorexia.

Figure 6

Effect of the ABA induction on the NMDA receptor subunits composition in the whole homogenate (left) and in the crude membrane fraction (right) of the NAc in the acute phase of the pathology (PND 42) and after a 7-days recovery period (PND 49). Protein levels of GluN2A and GluN2B receptors expressed as GluN2A/2B ratio are shown in the (a) homogenate and (b) crude membrane fraction at PND 42, and in the (c) homogenate and (d) crude membrane fraction at PND 49. Results are expressed as percentages of controls and represent the mean ± SEM of five-six rats per group. Panel (e) shows representative immunoblots for GluN2A and GluN2B. * p < 0.05 vs. Food ad libitum-sedentary, ^$$ p < 0.01, ^$$$ p < 0.001 vs. Food restriction-sedentary, ^## p < 0.01, ^$$$ p < 0.001 vs. Food restriction-sedentary, ^### p < 0.001 vs. Food ad libitum-exercise (two-way ANOVA followed by Tukey’s multiple comparisons test). CTRL = control; FR = food-restricted; EXE = exercise; ABA = activity-based anorexia.

Figure 7

Effect of the ABA induction on PSD95 scaffolding protein expression in the whole homogenate (left) and in the crude membrane fraction (right) of the NAc in the acute phase of the pathology (PND 42) and after a 7-days recovery period (PND 49). Protein levels of PSD95 are shown in the (a) homogenate and (b) crude membrane fraction at the PND 42, and in (c) the homogenate and (d) crude membrane fraction at PND 49. Results are expressed as percentages of controls and represent the mean ± SEM of five-six rats per group. Panel (e) shows representative immunoblots for PSD95. ^$$ p < 0.01 vs. Food restriction-sedentary, ^# p < 0.05, ^## p < 0.01 vs. Food ad libitum-exercise (two-way ANOVA followed by Tukey’s multiple comparisons test). CTRL = control; FR = food-restricted; EXE = exercise; ABA = activity-based anorexia.

Figure 8

Effect of the ABA induction on SAP102 scaffolding protein expression in the whole homogenate (left) and in the crude membrane fraction (right) of the NAc in the acute phase of the pathology (PND 42) and after a 7-days recovery period (PND 49). Protein levels of SAP102 are shown in the (a) homogenate and (b) crude membrane fraction at PND 42, and in the (c) homogenate and (d) crude membrane fraction at PND 49. Results are expressed as percentages of controls and represent the mean ± SEM of five-six rats per group. Panel (e) shows representative immunoblots for SAP102. ^$ p < 0.05 vs. Food restriction-sedentary, ^# p < 0.05 vs. Food ad libitum-exercise (two-way ANOVA followed by Tukey’s multiple comparisons test). CTRL = control; FR = food-restricted; EXE = exercise; ABA = activity-based anorexia.

Figure 9

Pearson’s product–moment correlation (r) analyses between GluA1/A2 ratio and (a) distance travelled, (b) mean speed and (c) long-exercise sequences and between PSD95 and (d) long-exercise sequences and (e) maximum speed of ABA and EXE rats. EXE = exercise; ABA = activity-based anorexia.