Ventral hippocampus mediates inter-trial responding in signaled active avoidance

The hippocampus has a central role in regulating contextual processes in memory. We have shown that pharmacological inactivation of ventral hippocampus (VH) attenuates the context-dependence of signaled active avoidance (SAA) in rats. Here, we explore whether the VH mediates intertrial responses (ITRs), which are putative unreinforced avoidance responses that occur between trials. First, we examined whether VH inactivation would affect ITRs. Male rats underwent SAA training and subsequently received intra-VH infusions of saline or muscimol before retrieval tests in the training context. Rats that received muscimol performed significantly fewer ITRs, but equivalent avoidance responses, compared to controls. Next, we asked whether chemogenetic VH activation would increase ITR vigor. In male and female rats expressing excitatory (hM3Dq) DREADDs, systemic CNO administration produced a robust ITR increase that was not due to nonspecific locomotor effects. Then, we examined whether chemogenetic VH activation potentiated ITRs in an alternate (non-training) test context and found it did. Finally, to determine if context-US associations mediate ITRs, we exposed rats to the training context for three days after SAA training to extinguish the context. Rats submitted to context extinction did not show a reliable decrease in ITRs during a retrieval test, suggesting that context-US associations are not responsible for ITRs. Collectively, these results reveal an important role for the VH in context-dependent ITRs during SAA. Further work is required to explore the neural circuits and associative basis for these responses, which may be underlie pathological avoidance that occurs in humans after threat has passed.


Introduction
Proactive avoidance strategies can be highly adaptive to the extent that avoidant behavior prevents a harmful outcome [1,2].However, an avoidance response that continues after danger has passed does not provide the same adaptive benefit and can even become maladaptive if it interferes with the resumption of preferred patterns of behavior [3][4][5].Indeed, persistent forms of avoidance that disrupt normal life activities are a unifying symptom of disorders related to fear, anxiety, and obsessive thinking.Relatively little is known about the neural and behavioral processes underlying avoidant behaviors that persist despite being uncoupled from their protective/preventative effects.To explore this phenomenon, we used a signaled active avoidance (SAA) procedure in which persistent inter-trial responses (ITRs) are elicited by discrete conditioned stimuli (CSs).
Previous studies suggest that ITRs in SAA are driven by aversive associations with the environment in which training occurs [6].Indeed, a similar role for context in the production of ITRs has been demonstrated for appetitive instrumental behavior [7].The ventral hippocampus (VH) and associated cortices are important substrates for aversive contextual processes, such as the acquisition and expression of aversive contextual associations [8][9][10][11], the contextually-mediated renewal of previously-extinguished aversive CSs [12][13][14][15], and context-dependent instrumental avoidance [16][17][18].VH is also required for classic anxiety-like behaviors, in which animals organize their exploratory behavior to minimize contact with threatening aspects of an experimental apparatus [19][20][21][22][23].These results are consistent with the proposal that VH contains a map that links interior representations of aversive states with environmental stimuli [24].We hypothesized that an aversive association with the SAA training context engages VH to produce ITRs.
Consistent with this hypothesis, our results demonstrate that pharmacological inactivation of VH suppresses ITRs while having no effect on avoidance responses (ARs) in the training context, and that chemogenetic excitation of VH neurons facilitates the expression of ITRs but not ARs in both the training and alternate contexts.Though these data reveal the contribution of VH to ITR expression, we also show that extinction of the SAA training context had no significant effect on ITRs, suggesting that aversive associations with the training environment do not mediate ITR expression.This result indicates that an alternative associative process, such as a residual CS-elicited aversive state that lingers beyond performance of the avoidance response [25], engages VH to drive the expression of ITRs in SAA.

Subjects
Subjects were 221 experimentally naïve, adult Sprague-Dawley rats (123 males, 98 females) weighing 200-300 g at time of arrival.Of these, 25 males contributed data to a previously published study (Oleksiak et al., 2021).The remaining 98 male and 98 female subjects comprised equally divided, mixedsex cohorts for all other experiments.Rats were purchased from Envigo for Experiment 1, from Envigo and Charles River for Experiment 2, from Charles River for Experiment 3, and Envigo for Experiment 4.
Multiple vendors were used to accommodate availability issues stemming from COVID-related changes in demand for experimental subjects.Rats were individually-housed in the Interdisciplinary Life Sciences Building vivarium and maintained on a 14h light: 10h dark cycle beginning at 7 AM.All subjects had ad libitum access to food and water.All experiments were conducted during the light cycle.Experimenters gently handled each rat for two minutes/day for four days prior to each experiment.All experiments were approved by the Animal Care and Use Committee from Texas A&M University.

Surgery
For experiments involving muscimol inactivation of ventral hippocampus (VH), surgical methods have been published previously (Oleksiak et al., 2021).This approach was adapted for experiments involving DREADD activation of VH neurons.Briefly, rats were anesthetized with isoflurane (5% for induction, slowly reduced throughout the procedure) and fixed in a stereotaxic apparatus.Lidocaine was injected subcutaneously at the incision site, and rimadyl was injected at the beginning of surgery for postsurgical analgesia.Then an incision was made in the scalp, and bore holes were placed to allow intracranial injections.Rats then received bilateral infusions of either AAV8-CAMKIIa-hM3D(Gq)-mCherry or AAV8-CaMKIIa-GFP (5x10 12 vg/mL for both viruses; Addgene) into the VH (either A/P: -5.25 or -5.65, M/L: ± 5, D/V: either -7.3 or -8.5 from bregma).Infusions were made at a rate of 0.1 µl/min for a total volume of 0.75 µl.Five minutes were allowed for diffusion after each injection.Injectors were then slowly removed at a rate of 0.1 mm/min for the next five minutes.Following surgery, subjects were allowed to recover for at least four weeks before tests.

Behavioral Apparatus
Six identical shuttle boxes (50.8 x 25.4 x 30.5 cm, LxWxH; Coulbourn Instruments) constructed of Plexiglas and metal were used for all training and testing.Each shuttle box was divided into two equal compartments (8 x 9 cm, WxH) between which subjects could pass freely through an aperture in the divider.The floors were made of conductive stainless-steel bars through which electric shock was delivered.Two speakers, one on either side of the length of each chamber, delivered a 2 kHz, 80 dB tone CS for a maximum of 15 seconds.A scrambled shocker delivered a 0.7 mA, 0.5 second footshock US through the floor.Each chamber was contained inside a larger sound-attenuating chamber.The subjects' motion was monitored by two infrared arrays on either side of the divider between compartments.CS and US presentation were controlled by GraphicState Software (Coulbourn Instruments), which also collected shuttling data via the infrared arrays in each compartment Two distinct shuttle-box contexts were created using methods published previously (Oleksiak et al., 2021).In all experiments, SAA training was conducted in the shuttle boxes with the house light off, black construction paper wall inserts with glow-in-the-dark stars, a 1% ammonia odor, and open doors of the sound-attenuating chamber during training (room lights also remained off).Rats were transported in black transport boxes without bedding.To create an alternate shuttle-box context, the house light in each compartment was illuminated (0.5 W light bulb), black-and-white striped wall inserts were placed behind the Plexiglas walls of the chamber, the box was scented with a 3% acetic acid solution, a black Plexiglas floor was placed over the conductive bars (no shocks were delivered in the alternate context), and the doors of the sound-attenuating chamber remained closed.Rats were transported to this alternate context in white transport boxes with bedding.
For locomotor activity testing (Experiment 2, see below) and two extinction-learning control conditions (Experiment 4, see below), five identical behavioral chambers (30 x 24 x 21 cm, L x W x H; Med-Associates, St Albans, VT) were used.Each box was a single compartment made of Plexiglas and metal.The floors were comprised of stainless-steel bars covered with a black Plexiglas floor insert.To fully distinguish this environment from the shuttle box training context, the alternate context cues were used in these single-chamber boxes (high-contrast striped backgrounds were placed behind the Plexiglas, the chamber was scented with acetic acid, and subjects were transported from the colony to the singlechamber boxes via white plastic carriers with bedding).A grating in one of the aluminum sidewalls allowed the delivery of tone CSs (2 kHz, 80 dB tone) from a speaker mounted on the outside of the box.
Each box sat on a load-cell platform that recorded how the chamber was displaced by the rat's locomotion.
The data was collected by Threshold Activity software (Med-Associates).Each chamber's load-cell voltages were digitized at 5 Hz, which provided one observation every 200 ms for locomotion measures.Between each trial was an inter-trial interval (ITI) that averaged 120 seconds.Shuttling during the ITI was defined as an inter-trial response (ITR).Rats were considered poor avoiders and excluded from further analysis if they avoided on £ 20% of trials averaged for the last 3 days of training.
Tests of SAA were conducted under extinction conditions in either the training or alternate shuttle box context described above.During these tests, ten 15-sec CSs were presented separated by a 2-min ITI.
CSs did not terminate with shuttling and no USs were presented.

Drug Infusions
Procedures for muscimol (GABAA agonist) infusions into VH have been described previously (Oleksiak et al., 2021).Clozapine-N-oxide (CNO) was purchased from RTI International.For DREADD experiments, CNO was dissolved in 2.5% DMSO until clear liquid was achieved and then diluted in sterile saline to a concentration of 3 mg/kg/mL immediately before injection.Approximately 30 minutes before start of test or locomotion session, rats received i.p.CNO or 2.5% DMSO saline (vehicle) injections in the vivarium and were placed back in their home cage until transport to the shuttle boxes or locomotor test boxes.

Experiment 1: Effect of pharmacological inactivation of VH on ITRs in the training context.
Rats received chronic guide cannula implanted in VH and were allowed to recover before undergoing four days of SAA training.Following training but before testing, all subjects were assigned to treatment-order conditions.A pair of counterbalanced SAA tests were then conducted, one in the training and one in the alternate shuttle-box context.Both were conducted under extinction conditions (see above).
Prior to each test, all subjects received intra-VH infusions of muscimol or vehicle.Here, we report novel ITR data in the training context from this experiment -AR data have been reported previously (Oleksiak et al., 2021).

Experiment 2. Effect of chemogenetic activation of VH principal neurons on ITRs in the training context.
Rats were injected with a virus bearing the gene construct for either the excitatory hM3Dq DREADD or GFP and were allowed to recover for at least three weeks.After recovery, rats underwent four days of SAA training.Treatment-order conditions for SAA testing were determined by balancing groups based on avoidance performance during training.Rats then underwent two days of SAA testing under extinction conditions.Both tests occurred in the training context.Thirty minutes prior to each, both hM3Dq and GFP groups received i.p.CNO or vehicle in a counterbalanced order.One week after testing, a subset of animals from both hM3Dq and GFP groups (7 hM3Dq and 6 blank-GFP rats) underwent an off-baseline test of locomotor activity in a novel environment (cubicle Plexiglas-and-metal chambers situated on load cells for the quantification of locomotion).These rats received two counterbalanced locomotor tests preceded by CNO or Vehicle, in which where they were allowed to move freely in the novel environment for 10 minutes.

Experiment 3. Effect of chemogenetic activation of VH principal neurons on ITRs in an alternate shuttle-box context.
All rats received VH viral infusions for the expression of hM3Dq or GFP at least three weeks before the start of SAA training.Most of the animals from both hM3Dq and GFP groups (32 hM3Dq and 27 GFP) received a cannula implantation procedure following the VH viral injection but prior to SAA.
After four days of SAA training, these subjects received intracranial infusions and a pair of SAA tests (data not presented).Cannulated subjects then received an additional day of SAA training to re-establish baseline avoidance, followed by two SAA tests in the alternate shuttle-box context preceded by i.p.CNO or Vehicle in a counterbalanced order.For a subset of animals (7 hM3Dq and 8 GFP), testing in the alternate SAA context began following the four days of SAA training.All other elements of SAA testing were the same in both subsets of rats.Because ANOVA revealed that the additional procedure and training did not influence the systemic effects of CNO on ITR expression in hM3Dq and GFP subjects, we did not consider this variable in subsequent analyses.

Experiment 4. Effect of context extinction on ITRs.
Rats received 4 days of SAA training, as above.This was followed by three additional days under one of the following four conditions: 1) context extinction via exposure to the SAA training context in the absence of the CS or US, for the normal duration of a training session (73 minutes); 2) equivalent-duration exposure to single-chamber alternate context, also in the absence of the CS or US; 3) equivalent-duration exposure to single-chamber alternate context, with the delivery of 30 CSs/day separated by a 2-min ITI; and 4) equivalent-duration exposure to white 5-gallon buckets (handling control).Following three days of exposure to these conditions, all rats received a SAA test under extinction conditions, conducted in the SAA training context (as described above).

Histology
Rats in pharmacological and chemogenetic experiments were injected with a fatal dose of sodium pentobarbital (Fatal Plus; 100 mg/kg, Vortech Pharmaceuticals) and were transcardially perfused first with refrigerated saline followed by 10% formalin.The brains were collected and were left in 10% formalin for 24 hours.The brains were then transferred to 30% sucrose in PBS at 4°C until they sank.Brains were sliced coronally on a cryostat (-20°C) at 30 micron thickness.Slices were mounted on subbed slides and coverslipped with fluoromount to visualize the mCherry or GFP from the viral manipulation.For all experiments, a poor-avoider criterion was applied following SAA training but prior to test.
Animals averaging six or fewer ARs during the last three sessions of training were determined to be poor avoiders [26][27][28].Because these experiments were designed to explore neural substrates in animals that successfully express SAA responses, poor avoiders were removed from our analyses.to be poor avoiders and were excluded from the analysis (data not shown).ARs from SAA training were analyzed using a mixed-design, two-way ANOVA with a within-subjects factor of Day and a betweensubjects factor of Virus (hM3Dq or Blank-GFP).This analysis revealed a main effect of Day  To perform a more nuanced analysis of shuttling across each test session, we again broke each trial into 15-sec epochs and generated session-wide averages of shuttling within each epoch, as described under Experiment 2. These data were analyzed with a mixed-design, three-factor ANOVA with two withinsubjects factors of Drug (CNO or Vehicle) and Time Bin and a between-subjects factor of Virus (hM3Dq or Blank-GFP).This analysis revealed a main effect of Virus

Discussion
Here, we report that pharmacological inhibition of ventral hippocampus (VH) reduces the expression of inter-trial responses (ITRs -shuttling during the inter-trial interval) in a two-way signaled active avoidance (SAA) paradigm.In addition, we demonstrate that chemogenetic excitation of principal neurons in VH is sufficient to enhance the expression of ITRs when tested in both training and alternate shuttle-box contexts.Neither manipulation had an effect on avoidance responses (ARs -shuttling during the CS), and control experiments revealed no effect of VH activation on locomotion, suggesting that VH selectively facilitates ITRs when activated, instead of producing an overall increase in locomotor activity.
Though we hypothesized that context-shock associations would be a critical associative substrate for ITR expression, extinction of the training context via prolonged exposure had no effect on ITRs.We conclude that context-shock associations may still play some role, but that VH is primarily engaged in some other associative/motivational process to drive the expression of ITRs.
The hypothesized role for context-shock associations in VH-mediated ITRs was grounded in work demonstrating that VH neurons display place fields [29][30][31][32] and respond to presentation of a foot-shock US [11,33].Indeed, VH has been shown to mediate multiple aversive/associative contextual processes [8][9][10][11][12][13][14][15]17,18].Particularly relevant to this study is the demonstration that contextual extinction via prolonged exposure to the training environment reduces ITRs in a wheel-run signaled avoidance task [6].For all of these reasons, we were surprised to find that extinction of context-shock associations had no effect on VH-mediated ITRs in the two-way SAA paradigm.However, it has been demonstrated that context exposure reduces ITRs in Roman high-avoidance rats bred for augmented performance in SAA (i.e. enhanced expression of both ARs and ITRs), suggesting that an aversive association between the US and the training context may indeed account for the degree to which Roman high-avoidance rats differ from Sprague-Dawley rats (the strain employed here) in ITR expression [34].Thus, context-shock associations could function to boost ITRs in subjects with a genetic propensity for high avoidance, even if they are not comparably operative in subjects lacking the same inborn tendencies.Further, though superficially similar, behavior in distinct avoidance tasks (i.e.wheel-run vs two-way SAA) may rely on distinct forms of aversive memory.
Associative structures other than direct associations between the training context and shock may play a role in the generation of ITRs in two-way SAA.Contexts can serve as occasion setters and modulate both Pavlovian and instrumental behavior without direct associations with USs, CSs, or the responses they evoke [35].Several studies have shown that the hippocampus has an important role in occasion setting [36,37], including in procedures in which contexts serve as occasion setters [38,39].Although most studies have focused on the role of the dorsal hippocampus in occasion setting, VH does play a role in this process, including the context-dependent modulation of Pavlovian extinction [13].Hence, it is possible that the VH is recruited to produce ITRs via the occasion-setting properties of the shuttle-box context, rather than direct context-shock associations.
Though manipulations of context often involve the physical apparatus in which behavior is acquired and tested, a comprehensive consideration of contextual process should include any aspect of the background against which action occurs, including psychological processes intrinsic to the subject [37].
It has been argued that a residual aversive state triggered by CS presentation is an element of the interoceptive context that energizes ITR expression [25].Indeed, SAA extinction via presentation of the CS in the absence of the US leads to the suppression of both ARs and ITRs [40], demonstrating that CSevoked aversion may contribute to both responses.Our results may thus suggest that a lingering cueelicited aversive state can engage VH, which then transforms this interior state into the behavioral output of ITRs via the recruitment of the appropriate efferent pathway, consistent with recent conceptual frameworks describing VH function [24].From a preclinical perspective, these VH-mediated responses can be considered perseverative behaviors decoupled from any exterior reinforcement that provide a viable model of the behavioral compulsions observed in obsessive compulsive disorder (OCD; [41,42]).Indeed, recent brain imaging work in human patient populations links hippocampal volume and functional connectivity to OCD symptomology [43,44].VH-mediated ITRs may thus serve as a valuable research tool with which to generate mechanistic insight into compulsive behavior and thus multiple, related psychopathologies.
Other VH-mediated processes may explain our results.All SAA tests reported here were conducted under extinction conditions, meaning that, unlike training, the CS was presented for a full 15seconds regardless of whether the subject produced an AR.Previous work indicates that VH maintains behavioral responses when CSs are presented for longer than expected [45] and mediates behavioral responses under other forms of timing uncertainty [46,47].Furthermore, it has been shown that VH activation can increase context generalization [48], which may explain the observation that VH activation drives up ITR expression in both the training and alternate shuttle-box contexts.Finally, VH lesions have been linked to perseverative responding in other tasks [49,50].Because strong forms of neuronal activation can dysregulate hippocampal function [48,51], it is possible that DREADD excitation of VH principal neurons produced a deficit in VH function that released perseverative responding to the CS in the form of ITRs.

2. 4
Signaled Active Avoidance (SAA) Training & Test SAA training and tests occurred in identical Coulbourn shuttle boxes.To ensure that all subjects associated tone with shock, the first trial of the first session of SAA training was comprised of a CS paired with a US that could not be avoided by shuttling.The rest of training consisted of avoidance trials, in which shuttling (crossing through the divider to the opposite compartment of the shuttle box) during the CS resulted in the immediate offset of the CS and the omission of the US.Shuttling during the CS was defined as an avoidance response (AR).Each SAA training session involved 30 such avoidance trials.

2. 8
Data Analysis ANOVAs followed by the appropriate post hoc tests were performed on all data.All statistical analyses were run in Statview version 5.0.1 (SAS Institute) in a MacOS 9 open-source emulator.
1: Pharmacological inactivation of VH reduces the expression of ITRs.We first assessed whether VH is necessary for the expression of ITRs in SAA.Rats acquired ARs over the course of SAA training, prior to being assigned to drug (n=13) or vehicle (n=12) groups.Then, all subjects received SAA tests under extinction conditions in the training and alternate SAA contexts, preceded by intra-VH infusions of muscimol or saline, depending on group (behavioral procedure schematized in FIG 1A, only data collected from the SAA test in the training context are depicted here; all other AR data shown in: Oleksiak et al., 2021).As previously reported, VH inactivation had no effect on ARs (capped 1/CS) during tests conducted in the SAA training context (Oleksiak et al., 2021; FIG 1B, left).We also performed a novel analysis of the effects of VH inactivation on ITRs during the same test.A two-way ANOVA of mixed design with a within-subjects factor of ITI and a between-subjects factor of Drug (Muscimol or Vehicle) revealed a significant ITI X Drug interaction [F(9,207) = 2.232, p = 0.0213].Thus, muscimol inactivation of VH suppresses ITRs while having no effect on AR expression.We speculate that vehicle and muscimol ITRs converge later in the session due to within-session extinction effects.

3. 2 Fig. 1 .
Fig. 1.VH inactivation decreases ITI shuttling over the course of a 2-way SAA test in the training context.A) Rats were trained in two-way signaled active avoidance and had either muscimol (n=13) or vehicle (n=12) infused into the VH immediately before a test in the training context.Created with BioRender.com.B) The effect of muscimol and vehicle on capped CS shuttles and total ITI shuttles throughout the training context test.Muscimol significantly decreased the number of ITI shuttles during this test (* indicates p<0.05).All data calculated as means ± SEM.Adapted from Oleksiak et al., 2021.

Fig. 2 .
Fig. 2. VH DREADD activation increases ITI shuttles in the training context without affecting locomotion.(A)Example virus expression and schematic viral expression (viral expression that is more common shows as darkest area) targeting vCA1 for the excitatory DREADD virus (n=14) and the blank virus (n=10).Atlas images are from Swanson, 1998 [63].(B) Rats received 4 days of SAA training prior to counterbalanced CNO and VEH tests conducted under extinction conditions in the training context.Drug was administered with an IP injection 30 minutes prior to each test.(C) Average number of avoidance responses per day of training for active hM3Dq DREADD virus animals and blank virus-GFP animals.(D) Rats received either CNO or vehicle in an I.P. injection 30 minutes before each test; (Left) The effect of CNO and vehicle on total avoidance responses during counterbalanced drug tests; (Middle) the effect of CNO and vehicle on the total ITI shuttles over the session; (Right) Each ITI broken down into 15 second bins for a total of 8 segments for the full 2 minutes with all 10 ITI 15 second bins averaged for each animal.hM3Dq animals show a significant increase in ITI shuttles from the Vehicle to the CNO condition while blank-GFP animals do not.(E) Drug was administered I.P. 30 minutes before each locomotion test in the alternate single-chamber context.(F) Locomotion measured over a 10-minute period to assess CNO effects on movement.VH DREADD activation increases ITI shuttling without affecting acclimation, CS shuttles, or general locomotion.(* indicates p<0.05).All data presented as means ± SEM.A, B, & E created with BioRender.com

Fig. 3 .
Fig. 3. VH DREADD activation increases ITI shuttles and trends toward increasing CS shuttles in the alternate shuttle-box context.(A)Example virus expression and schematic viral expression (viral expression that is more common shows as darkest area) targeting vCA1 for the excitatory DREADD virus (n=39) and the blank virus (n=35).Atlas images are from Swanson, 1998 [63].(B) Rats received 4 days of SAA training.Drug was administered with an IP injection 30 minutes prior to each test in the alternate shuttle-box context.A & B created on BioRender.com.(C) Average number of avoidance responses per day of training for active hM3Dq DREADD virus animals and blank virus-GFP animals.(D) Rats were then tested in various measures in the alternate shuttle-box context (Left) The effect of CNO and vehicle on total avoidance responses during counterbalanced drug tests; (Middle) the effect of CNO and vehicle on the total ITI shuttles during tests; (Right) Each ITI broken down into 15 second bins for a total of 8 segments for the full 2 minutes with all 10 ITI 15 second bins averaged for each animal.hM3Dq animals show a significant increase in ITI shuttles from the Vehicle to the CNO condition while blank-GFP animals do not.Overall hM3Dq CNO significantly increases ITI shuttles and trends toward increasing CS shuttles.(* indicates p<0.05).All data presented as means ± SEM.

Fig. 4 .
Fig. 4. Context extinction does not significantly reduce ITRs.(A)Rats received 4 days of two-way SAA training.Rats were then assigned to groups where they were either exposed to the training context, the alternate chamber, the alternate chamber with the avoidance tones presented, or equivalent handling for the same period as typical avoidance training for 3 days.They finally received a 10-tone test under extinction conditions on day 8 to assess their avoidance and ITI shuttling behavior.Created with Biorender.com(B) Average avoidance responses performed each day of training.Rats were assigned to groups so that each groups would have similar levels of avoidance in the handling (n=14), training context (n= 13), tone exposure (n=14), and alternate context (n=14) groups across the days.(C) (Left) Total avoidance responses in the test for each group.All exposure groups showed similar levels of CS shuttling.(Right) Number of shuttles for each ITI throughout the 10-tone test.Training context extinction marginally but not significantly decreases ITI shuttling during the test compared to the other 3 control groups.All data are shown as mean ± SEM.
Drug interaction [F(1,72)=29.539,p<0.0001].Fisher's LSD post hoc tests confirmed that hM3Dq animals showed a significant overall increase when given CNO compared to vehicle [t(38)=0.061,p<0.0001],whereasBlank-GFPanimalsshowedequivalent responding in response to both CNO and vehicle [t(34)=0.044,p=0.977](FIG3D,right).Thus, CNO activation of VH neurons produced a uniform increase in shuttling that remained flat across the session, instead of selectively increasing shuttles at any point, including during the CS.These data are consistent with the interpretation that VH activation is sufficient to drive ITR expression in both training and alternate shuttle-box contexts.3.4Experiment4:ContextExtinctiondoes not significantly reduce ITI shuttling.Because we hypothesized that an aversive association between context and the US is a key associative structure underlying ITRs, we examined the effect of contextual extinction via prolonged exposure on ITR expression.All subjects received SAA training before being broken into one of four conditions that received either: exposure to the SAA context (n=13), exposure to an alternate chamber and manipulations influenced the robust expression of avoidance(FIG 4C, left).We also analyzed ITR expression across the 10 ITIs of the test session using a two-way, repeated-measures ANOVA with a within-subjects factor of ITI Period and a between-subjects factor of Exposure Group.andourcontrol manipulations(FIG 4C, right).Further, we note that the overall pattern and level of ITR expression across the SAA test session is roughly typical of what we observe in subjects that have not received any additional manipulations between training and test (see: FIG 1B, right for comparison).We conclude that context fear may contribute to ITR performance but is not the primary driver of this action.