Ventral hippocampus to nucleus accumbens shell circuit regulates approach decisions during motivational conflict

Abstract

Successful resolution of approach-avoidance conflict (AAC) is fundamentally important for survival, and its dysregulation is a hallmark of many neuropsychiatric disorders, and yet the underlying neural circuit mechanisms are not well elucidated. Converging human and animal research has implicated the anterior/ventral hippocampus (vHPC) as a key node in arbitrating AAC in a region-specific manner. In this study, we sought to target the vHPC CA1 projection pathway to the nucleus accumbens (NAc) to delineate its contribution to AAC decision-making, particularly in the arbitration of learned reward and punishment signals, as well as innate signals. To this end, we used pathway-specific chemogenetics in male and female Long Evans rats to inhibit the NAc shell projecting vHPC CA1 neurons while rats underwent a test in which cues of positive and negative valence were presented concurrently to elicit AAC. Additional behavioral assays of social preference and memory, reward and punishment cue processing, anxiety, and novelty processing were administered to further interrogate the conditions under which the vCA1-NAc shell pathway is recruited. Chemogenetic inhibition of the vCA1-NAc shell circuit resulted in animals exhibiting increased decision-making time and avoidance bias specifically in the face of motivational conflict, as the same behavioral phenotype was absent in separate conditioned cue preference and avoidance tests. vCA1-NAc shell inhibition also led to a reduction in seeking social interaction with a novel rat but did not alter anxiety-like behaviors. The vCA1-NAc shell circuit is therefore critically engaged in biasing decisions to approach in the face of social novelty and approach-avoidance conflict. Dysregulation of this circuit could lead to the precipitation of addictive behaviors in substance abuse, or maladaptive avoidance in situations of approach-avoidance conflict.

Fig 1. Chemogenetic inhibition of ventral hippocampus CA1 (vCA1) to NAc shell projections.

(A) Schematic diagram of inhibitory DREADDS (hM4Di-mCherry) or empty vector (GFP) transduction in vCA1 and cannula implantation in NAc shell via which CNO or saline was microinfused for pathway-specific inhibition. (B) Representative images of somatic transduction of hM4Di-mCherry or empty vector GFP in vCA1 (left) and axonal transport of the viral load in the NAc shell (right). Scale bar, 1 mm. (C) Cannula placements in the NAc shell (DREADDs group only) and the minimum and maximum DREADDs expression in vCA1 and NAc shell, AP: Anterior-Posterior to bregma. (D) Representative images of c-Fos expression in the NAc following either saline (SAL) or CNO (1 mM) infusion, showing cannula tract, and tip localized within the shell subregion (arrow). Scale bar, 100 μm. Intracerebral CNO infusions attenuated c-Fos signal in areas of the NAc shell that received hM4Di positive inputs (bottom graph). c-Fos-positive cell density in the adjacent NAc core was not affected by CNO infusions. Data show means ± SEM. ***p < 0.001. The data underlying panel D are available at: https://osf.io/pwcer/. CNO, clozapine N-oxide; NAc, nucleus accumbens.

Fig 2. Effect of chemogenetic inhibition of vCA1-NAc shell on approach-avoidance conflict.

(A) Experimental timeline showing mixed valence cue training (2 or 3 × 4 sessions), followed by a cue acquisition test (5 min), refresher training session, and finally the cue conflict test, prior to which animals received intracranial microinfusion of CNO (1 mM) or saline. (B) After 2 to 3 rounds of 4 conditioning sessions, 34 hM4Di-expressing animals (hM4Di Sal: n = 17 (6 females), hM4Di CNO: n = 17 (7 females), symbols with red border depict female data) successfully acquired cue-outcome associations, demonstrating a greater amount of time in the appetitive vs. the neutral arm (p < 0.001) and less time in the aversive vs. the neutral arm (p < 0.001). (C) In the conflict test, rats were given a choice to explore a neutral cued arm or an arm with the appetitive and aversive cues superimposed (conflict cue). Time spent in the conflict and neutral cued arm was significantly reduced in the vCA1-NAc shell inhibited group (hM4Di CNO) compared to control groups (hM4Di SAL). (D) CNO-infused hM4Di animals avoided the conflict cued arm relative to the control hM4Di SAL animals. (E) Amount of time spent in the hub (central compartment), and (F) risk assessment (checking) behavior were also significantly increased in the vCA1-NAc shell inhibited group, compared to the control SAL group. (G) Latency to enter the conflict cue arms for the first time was increased in the males but not females after vCA1-NAc shell inhibition. (H) When a decision-making composite z score was computed and compared across the 4 groups (including GFP-expressing CNO and SAL groups), the hM4Di-CNO group exhibited significantly higher scores, indicative of potentiated decision-making hesitancy. (I) Number of retreats in both conflict and neutral arms were more pronounced in the vCA1-NAc shell inhibited group compared to the control group, and (J) the males exhibited more retreat behavior overall than females. (K) vCA1-NAc shell inhibition did not alter arm entries; however, the females made more entries into the neutral than conflict arms. Data show means ± SEM. * p < 0.01, ** p < 0.01, ***p < 0.001. The data underlying this figure are available at: https://osf.io/pwcer/. CNO, clozapine N-oxide; NAc, nucleus accumbens.

Fig 3. No effect of vCA1-NAc shell inhibition on cue preference and avoidance tests.

(A) Experimental timeline showing the administration of a refresher conditioning session following the conflict test, then the administration of cue preference (appetitive vs. neutral cues) and cue avoidance (aversive vs. neutral cues) tests in counterbalanced order, prior to which animals were microinfused with CNO or saline. (B, C) All animals (hM4Di Sal: n = 16 (6 females), hM4Di CNO: n = 16 (7 females), symbols with red border depict female data) spent significantly more time in the appetitive vs. neutral arm in the cue preference test, and less time in the aversive vs. neutral arm in the cue avoidance test, irrespective of drug treatment or sex. (D, E) There was no group differences (drug treatment, sex) in the hub time during the cue preference test and cue avoidance tests. (F, G) Latency to enter the appetitive vs. neutral arms was lower in the males but not the females. Latency to enter the aversive cued arm for the first time was significantly higher compared to latency to enter the neutral arm in all groups. (H) The number of retreats observed in the aversively cued arm was significantly higher than in the neutral cued arms in the cue avoidance test. (I, J) There was no difference in risk assessment behavior between all groups in cue preference and avoidance tests. (K, L) All animals made more entries into the appetitive arm over the neutral arm in the cue preference test and made less entries into the aversive arm over the neutral arm in the cue avoidance test, irrespective of pathway inhibition or sex. Data show means ± SEM. * p < 0.01, ** p < 0.01, ***p < 0.001. The data underlying this figure are available at: https://osf.io/pwcer/. CNO, clozapine N-oxide; NAc, nucleus accumbens.

Fig 4. Effect of vCA1-NAc shell inhibition on assays of novelty processing, social seeking, and discrimination.

(A) Timeline of experimental procedures; following the completion of the cue preference/avoidance tests, animals underwent testing in: cue novelty test, social interaction and social novelty seeking tests. Prior to each of the tests, animals were microinfused with CNO or saline. (B) Animals (hM4DI CNO: n = 9 (5 females), GFP CNO: n = 9 (5 females), symbols with red border depict female data) spent more time engaging with novel over neutral cues in the Y-maze, regardless of drug treatment or sex. No significant drug treatment or sex differences were observed with (C) hub time, (D) risk assessment, (E) latencies to first entry into the novel vs. neutral cued arms, and (F) retreats in the arms. (G) vCA1-NAc shell inhibited animals spent significantly less time interacting with the novel conspecific over an empty cage compared to SAL control group and failed to show preference for social interaction (hM4Di Sal: n = 15 (7 females), hM4Di CNO: n = 11 (7 females)). (H, I) vCA1-NAc shell inhibited animals spent more time interacting with the familiar rat and significantly less time interacting with (i.e., avoidance of) the novel rat compared to SAL controls. Data show means ± SEM. * p < 0.01, ** p < 0.01, ***p < 0.001, ^## p < 0.01. The data underlying this figure are available at: https://osf.io/pwcer/. CNO, clozapine N-oxide; NAc, nucleus accumbens.