Imbalanced Activity in the Orbitofrontal Cortex and Nucleus Accumbens Impairs Behavioral Inhibition

Abstract

Contemporary models of behavioral regulation maintain that balanced activity between cognitive control areas (prefrontal cortex, PFC) and subcortical reward-related regions (nucleus accumbens, NAC) mediates the selection of appropriate behavioral responses, whereas imbalanced activity (PFC < NAC) results in maladaptive behavior [1-6]. Imbalance can arise from reduced engagement of PFC (via fatigue or stress [7]) or from excessive activity in NAC [8]. Additionally, a concept far less researched is that an imbalance can result from simultaneously low PFC activity and high NAC activity. This occurs during adolescence, when the maturation of PFC lags behind that of NAC and NAC is more functionally active compared to adulthood or pre-adolescence [2, 5, 9, 10]. Accordingly, activity is disproportionately higher in NAC than in PFC, which may contribute to impulsivity and risk-taking exhibited by adolescents [5, 6, 10-12]. Despite having explanatory value, support for this notion has been solely correlational. Here, we causally tested this using chemogenetics to simultaneously decrease neural activity in the orbitofrontal cortex (OFC) and increase activity in NAC in adult rats, mimicking the imbalance during adolescence. We tested the effects on negative occasion setting, an important yet understudied form of inhibitory learning that may be particularly relevant during adolescence. Rats with combined manipulation of OFC and NAC were impaired in learning to use environmental cues to withhold a response, an effect that was greater than that of either manipulation alone. These findings provide direct evidence that simultaneous underactivity in OFC and overactivity in NAC can negatively impact behavioral control and provide insight into the neural systems that underlie inhibitory learning.

Keywords: DREADDs; adolescence; associative learning; chemogenetics; conditioning; learning; negative occasion setting; prefrontal cortex.

Figure 1. Configuration of stimuli in the negative occasion setting procedure and schematic of the balance model of behavioral control

A) Red and green lines indicate inhibitory and excitatory relationships in the behavioral procedure, respectively (US=unconditioned stimulus). The feature stimulus acts to gate, or ‘set the occasion’ for the meaning of the target stimulus and indicates that a response should be withheld during the subsequent presentation of the target [21]. B) Balanced activity (left side of panel, and indicated by dashed line) is necessary for appropriate behavioral control and is present in preadolescence and adulthood. An imbalance can result from impairing the function of OFC (middle left [7]), increasing the influence of NAC (middle right [8]), or by simultaneously disrupting and potentiating activity in OFC and NAC, respectively [2,5,9,10].

Figure 2. Example of conditioned food cup behavior during training in the negative occasion setting procedure

A) Conditioned responding during presentation of the tone on reinforced (R) and non-reinforced trials (NR), and B) the difference in responding between trial types, exhibited by the vehicle-treated control group in Experiment 1. * refers to the first of three consecutive sessions during which the group’s difference score between R and NR trials was significantly different from zero, defined as a Z score ≥2.325 (i.e., p<0.01; see Supplemental Experimental Procedures). Data are means±SEM.

Figure 3. Summary data indicating the number of sessions until each group in the study consistently exhibited a significant difference in responding to the tone on reinforced and non-reinforced trials

(i.e., Z-scores of ≥2.325, p<0.01 for 3 consecutive sessions; see Supplemental Experimental Procedures). Data in black are from Experiment 1 (combined manipulations of OFC and NAC), data in blue are from Experiment 2 (manipulation of OFC alone), and data in red are from Experiment 3 (manipulation of NAC alone). Rats with combined manipulation of OFC and NAC required more training to consistently exhibit a significant difference in responding to the tone on reinforced versus non-reinforced trials compared to either manipulation alone (see Results). Dotted line indicates the mean number of sessions for all control groups in the study (the individual control groups in each experiment exhibited comparable learning (see Results), thus only the controls that received the DREADDs virus and vehicle treatment are shown). All groups in Experiments 1 and 2 exhibited low levels of baseline responding during the Pre-CS epoch (Ps>0.3; see Supplemental Experimental Procedures). In Experiment 3 there was a significant group difference [F(3,40)=4.74, p<0.01] in that the NAC-hM3Dq-CNO group had higher levels of baseline responding than rats in the NAC-GFP-CNO group, but there were no differences compared to either NAC-hM3Dq-vehicle or NAC-GFP-vehicle rats. Similarly high levels of feeding behavior during the Post-CS epoch were observed each experiment (Ps>0.1). There was no effect of group on time spent rearing during the light in any experiment (Ps>0.1). In addition, food cup behavior during the light was low (on average ~0.75s) and did not differ between groups in Experiments 1 or 2 (P’s>0.7). In Experiment 3, a main effect of group was observed [F(3,40)=4.24, p<0.05], which was driven by lower levels of responding by rats in the NAC-GFP-CNO group relative to rats in the NAC-hM3Dq-CNO and NAC-hM3Dq-VEH groups and thus cannot explain the primary impairment observed in the NAC-hM3Dq-CNO group. Furthermore, although all groups showed a slight increase in responding during the first few sessions (likely attributable to a decrease in rearing during across these same sessions), we did not observe any other changes in food cup responding during the light across training. Abbreviations: OFC, orbitofrontal cortex; NAC, nucleus accumbens; CNO indicates treatment with clozapine-N-oxide; VEH indicates vehicle-treated control group.

Figure 4. Expression of the DREADD reporter in OFC and NAC

Fluorescent labeling of A) OFC neurons expressing hM4Di and C) NAC neurons expressing hM3Dq (20X magnification). Schematic diagram of representative minimum (dark gray) and maximum (light gray) expression of B) hM4Di in OFC and D) hM3Dq in NAC. Expression of the reporter was comparable and evident along the rostrocaudal extent of each region. Few fluorescent neurons were observed outside of the target region. Expression of the reporter in the GFP control groups was comparable to that observed in the hM4Di groups (data not shown). Expression of the reporters for the viruses was very similar across experiments.