Disinhibition of ventral tegmental area during initial punishment learning causes enduring punishment insensitivity

Disinhibition of ventral tegmental area during initial punishment learning causes enduring punishment insensitivity | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Disinhibition of ventral tegmental area during initial punishment learning causes enduring punishment insensitivity Philip Jean-Richard-Dit-Bressel, Shannen Tan, Michelle Shen, Luke Keevers, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7865029/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 17 Feb, 2026 Read the published version in Neuropsychopharmacology → Version 1 posted You are reading this latest preprint version Abstract Avoiding actions with negative consequences is fundamental to adaptive behavior. Traditional theories suggest GABAergic inhibition of midbrain dopamine neurons, including those within ventral tegmental area (VTA DA ), mediate suppression of actions that lead to aversive outcomes. However, the role of dopamine inhibition in punishment learning remains unclear. To examine this, we conducted fiber photometry, pharmacological, and chemogenetic experiments in rats to measure VTA DA activity and GABA input across punishment learning, and test their causal contribution to behavior. VTA DA activity and GABA input phasically increased to response-elicited outcomes, with VTA DA activity being more strongly activated by rewards, while GABA input being more strongly activated by shock punishers during initial punishment. Pharmacologically blocking GABA A receptors in VTA or chemogenetically activating VTA DA neurons during initial, but not later, punishment sessions produced enduring deficits in punishment avoidance. These findings suggest long-term avoidance depends upon a critical window of GABA-mediated VTA DA inhibition during punishment learning Biological sciences/Neuroscience/Learning and memory/Operant learning Biological sciences/Physiology/Neurophysiology Health sciences/Risk factors Figures Figure 1 Figure 2 Figure 3 Introduction Actions with positive consequences tend to be reinforced (i.e., repeated), whereas actions with negative consequences tend to be punished (i.e., suppressed) 1 . This fundamental adaptive function, known as instrumental conditioning, helps organisms dynamically adjust their behavior to maximize rewards and minimize harms. Dopaminergic neurons of the ventral midbrain, including those within the ventral tegmental area (VTA DA ), are widely considered to be critical for this learning 2–6 . These neurons exhibit phasic increases in activity to better-than-expected rewards, which are considered necessary and sufficient for reinforcing behaviors 7–10 . Conversely, VTA DA neurons exhibit phasic decreases in activity to worse-than-expected outcomes (including aversive events) 5,11,12 , and inhibition of VTA DA has been shown to be aversive 13–15 . Crucially, when specific actions cause brief optogenetic inhibition of VTA DA neurons, those actions (but not other actions) are suppressed 13 . This shows pauses in VTA DA neuron activity can function as punishment to produce selective instrumental avoidance. Endogenously, pauses in VTA DA activity are driven by GABA input to VTA DA neurons from local interneurons and long-range GABA inputs 5,16,17 , which act on GABA A receptors to suppress VTA DA firing 15,18,19 . However, patterns of GABA release onto VTA DA neurons during punishment and the necessity of VTA DA inhibition in punishment avoidance remain unclear. To examine this, we performed fiber photometry recordings of VTA DA calcium (Ca 2+ ; neural activity proxy) and GABA input across a punishment task to characterize how VTA DA and GABA dynamics relate to behavior under punishment. We tested the causal significance of VTA GABA input and VTA DA activity on punishment learning and choice via pharmacological (GABA A receptor antagonist) and chemogenetic manipulations (hm3D DREADD) across phases of punishment, and show preventing GABA A -mediated inhibition of VTA and direct excitation of VTA DA during initial punishment learning produces enduring punishment insensitivity. Methods Further details for Methods are supplied in Supplemental Materials . Subjects All experiments used experimentally-naive rats aged 8–24 weeks old. Photometry and chemogenetic experiments used heterozygous TH::Cre Sprague Dawley rats (SD-Th-cre tm1sage ; Sage Laboratories). TH::Cre + animals express Cre in tyrosine hydroxylase (TH; dopamine precursor enzyme) neurons; TH::Cre- animals (used in control experiment reported in Supplemental Materials) do not express Cre. Experiment 2 (pharmacological manipulation) used wild-type Sprague Dawley rats. Animals were group-housed (4) in plastic cages in a climate-controlled colony room maintained on a 12 hr light–dark cycle. Rats had ad libitum access to chow until 2 days before behavioral training, after which they received 10–15 g chow daily (after behavioral session) to maintain them at ~ 90% of their free-feeding weight. Rats had access to water in their homecages throughout experiments. All procedures were approved by the Animal Care and Ethics Committee at UNSW Sydney and conducted in accordance with the National Health and Medical Research Council Code for the Care and Use of Animals for Scientific Purposes in Australia (2013). Apparatus & Materials All operant behavior was assessed in MedAssociates operant chambers, each housed within light and sound-attenuating cabinets. Each chamber contained two retractable levers that flanked a magazine port where grain pellet rewards were delivered. The punisher was a 0.5sec footshock, delivered through the grid floor. Footshock intensity was 0.4mA for fiber photometry experiments, and 0.5mA for neural manipulation experiments. A lower intensity footshock was chosen for photometry experiments to avoid floor effects in responding that would undermine key analyses of peri-event dynamics. Locomotor tests were conducted in open field chambers that tracked movement via 16-beam infrared arrays located along X- and Y-axes. Fiber photometry recordings were conducted using Doric Lenses photometry components (465nm and 405nm LEDs, mini-cubes, photodetectors) and Tucker Davis Technologies photometry processor (RZ5P). Adeno-associated viruses (AAVs) were used to express Cre-dependent calcium sensor (AAV-CAG-DIO-GCaMP6f), GABA sensor (AAV-hSyn-DIO-iGABASnFR-F102G), or excitatory DREADD (AAV-hSyn-DIO-hm3D-mCherry) in VTA DA neurons of TH::Cre + animals. Microinfusions of GABA A antagonist bicuculline (0.1µg/µl; Tocris, Sydney, Australia) were used to prevent GABA-mediated inhibition in VTA 15 , 18 , 19 . Systemic injections of 3mg/kg clozapine- N -oxide (CNO; National Institute of Mental Health Chemical Synthesis and Drug Supply Program), dissolved in 5% DMSO and saline, were used to activate hm3D. Potential off-target effects of CNO 20 were addressed via TH::Cre- control subjects ( Supplemental Materials–Figure S4 ). Surgeries Rats were anaesthetized and placed into a flat skull position within a stereotaxic frame. Craniotomies were performed above VTA. For photometry and chemogenetic experiments, a 5µl 30-gauge microinfusion syringe (Hamilton; Reno, NV, USA) was used to inject 0.75µl AAVs (0.25µl/min) encoding Cre-dependent GCaMP6f (unilateral), iGABASnFR (unilateral), or hm3D (bilateral) into VTA (AP: -5.5, ML: ±0.8, DV: -8.2 from bregma) of TH::Cre rats. Following injections, the syringe remained at the injection site for an additional 5min for diffusion. For photometry experiments, a 400µm optic fiber was unilaterally implanted into VTA (AP: -5.5, ML: ±0.8, DV: -8.2 from bregma). For pharmacology experiments, a bilateral 26-gauge 11mm guide cannula (PlasticsOne) was implanted into VTA (AP: -5.8, ML: ±0.75, DV: -8.2 from bregma). Implants were anchored in position with dental cement and jeweller’s screws. Immediately following surgery, animals were given antibiotics and received post-operative monitoring and care for 1 week. Rats that received AAV injections were given an additional 3 weeks before behavioral training to allow sufficient transgene expression. Behavioral task All rats underwent a previously validated punishment task, which has been shown to elicit robust punishment avoidance with minimal contamination from Pavlovian fear 21 – 23 . Lever-press training Rats were first trained to press two levers (R1, R2) for food. For 2 sessions, both levers were presented concurrently, and each press on a lever was rewarded with a pellet (FR1 training). A lever remained extended until it received 25 presses or after 1 hour. Rats that failed to acquire lever-pressing were manually shaped in the second FR1 session. Rats then received 7–8 days of VI30s training (40min sessions). In these sessions, levers were presented individually for 5min blocks in alternating fashion (first lever randomized per day). Lever-presses were reinforced on a 30sec variable interval (VI30s) schedule, such that the first press after an average interval of 30 seconds led to pellet delivery. Punishment Subjects then received daily 40min punishment sessions. Lever-pressing on either lever continued to yield pellets (VI30s). However, every 10th press (FR10) on the punished R1 lever resulted in immediate footshock delivery. Presses on the unpunished R2 lever had no additional consequence. If a press was scheduled to deliver both footshock and pellet, both were delivered. Assignment of left vs. right levers as punished vs. unpunished was counterbalanced across (but not within) subjects. For pharmacology experiments, rats received intra-VTA infusions of 0.5µl GABA A antagonist bicuculline or control saline (0.25µl/min; 1min diffusion) immediately prior to the first two sessions of punishment (between-subjects), and bicuculline vs. saline on punishment days 6 and 7 (within-subjects, order counterbalanced). This design permits efficient interrogation of neural manipulation effects on acquisition and expression of punishment avoidance 21 , 22 . The same design was employed for DREADD manipulations, except rats received i.p. injections of CNO or vehicle (30mins before session start) instead of microinfusions, and expression tests were conducted on punishment days 7 and 8. Choice Test Rats were then given choice test(s) where both levers were presented concurrently. No shocks were delivered and presses on either lever delivered pellets on a shared VI60s schedule, so there was no advantage to pressing either lever exclusively or a combination of both levers. Photometry experiments only involved a single 15min choice test. For manipulation experiments, animals received within-subjects drug vs. control across two choice tests (order counterbalanced) (pharmacology experiment: 30min tests; DREADD experiment: 20min tests). Each choice test was preceded by a drug-free punishment session the day prior 21 , 22 . Locomotor tests Effects of VTA manipulation on locomotion were assessed following completion of the punishment task. Rats first received a 30 min habituation session, where they were placed into the open field chamber without any injections. On the following 2 days, rats received drug or control injection (within-subjects, counterbalanced order) before being placed into the chambers for 30 mins to assess distance travelled. Histology At the end of all experiments, brain tissue was examined to verify virus expression and/or implant locations. Data Analysis Rats that failed to acquire lever-pressing during lever training, or had inappropriate virus expression or implant placements, were excluded from all analyses. Behavior analysis The key behavioral dependent measures were self-normalized rates of responding on each lever (“suppression ratios”) 22 , and average latency to initially press each lever across trials (averaged per session). Suppression ratios normalize response rates per lever during punishment and choice sessions to pre-punishment (final VI30s) rates. This was calculated per lever as follows: $$\:Suppression\:ratio=\frac{Session\:LP\:rate}{(Session\:LP\:rate+Training\:LP\:rate)}$$ Suppression ratios can range from 0 to 1. Scores above 0.5 indicate greater lever-pressing relative to training, scores below 0.5 indicate less lever-pressing, while a score of 0.5 indicates no difference relative training. This was done to address any spurious difference in punished or unpunished response rates prior to punishment. Nonetheless, analyses of raw response rates are provided in Supplemental Materials. Behavioral data was analyzed using repeated measures ANOVA. Within-subjects factors were lever, session, and drug. Between-subjects factor was acquisition group (drug vs. control). For all analyses, Type 1 error was controlled at 0.05. Fiber photometry analysis 465nm (neural dynamic-related) and 405nm (isosbestic control) signals and event timestamps were extracted into MATLAB, and signals during logged disconnections were discarded. Each signal was low-pass (3 Hz) and notch (1.0322–1.0326, 2.547–2.55 Hz) filtered to remove high-frequency noise identified via Fast Fourier Transform. Filtered 405nm signals were fit to filtered 465nm signals via iteratively-reweighted least squares 24 to create fitted 405nm signals. A normalized fluorescence change score (dF/F) was calculated using the standard formula: $$\:dF/F=\frac{(465nm\:signal-fitted\:405nm)}{fitted\:405nm}$$ This motion-artifact-corrected dF/F was detrended via 60 sec moving median (5 sec mean smoothing window). Detrended signals were converted into standard deviation units by dividing session signals by their sum squared deviation from 0 (nullZ-score) 24 . All photometry analyses were derived from this normalized, artifact-corrected dF/F. The key dependent variable was change in VTA DA activity and GABA input around response-elicited outcomes (reward delivery, footshock) and actions (R1, R2). dF/F around pellets vs. footshocks, and R1 vs. R2 lever-presses alone (i.e., those not yielding footshock or pellets) were collated. Each trial was re-zeroed to pre-event baseline (-5:-3 sec) and averaged per subject; all analyses used mean peri-event transients per subject. Due to the scarcity of punished lever-presses and footshocks in late punishment sessions, late punishment data (Pun4 onwards) was combined to obtain more accurate peri-event activity traces per subject. Significant transients were identified via bootstrapped confidence intervals (CI) 25 . Bootstrapped means were obtained by randomly resampling from subject mean waveforms with replacement (1000 iterations). 95% CI limits were derived from 2.5 and 97.5 percentiles of bootstrap distribution, expanded by a factor of √(n/(n-1)). A significant transient was identified as a period that CI limits did not contain 0 (baseline) for at least 1/3secs (low-pass filter window 25 ). Significant differences between event waveforms were similarly determined by bootstrapping the within-subject difference waveform (mean event1–mean event2 waveform) per subject 25 . Results Experiment 1: VTA DA neuron a ctivity and GABA input during punishment learning We first examined activity of VTA DA neurons and GABA input to VTA DA neurons across punishment learning and choice ( Figure 1 ). This was done by selectively expressing GCaMP6f (Ca 2+ sensor; neural activity proxy) or iGABASnFR (GABA sensor; GABA input proxy) in VTA DA neurons of TH::Cre+ rats and recording from VTA across a punishment task ( Figure 1a-c ). N = 14 rats had valid biosensor expression and fiber placements ( n = 7 GCaMP [ Figure 1f ]; n = 7 iGABASnFR [ Figure 1h ]) and were thus included in analyses. Task Behavior Animals first received lever-press training, where they could press two individually-presented levers (R1, R2) for food ( Figure 1b ). Across this training, rats acquired similarly high rates of responding on R1 and R2 (lever: F (1,12) =0.10, p =.763); this did not depend on which sensor animals expressed (group: F (1,12) =1.19, p =.297; group*lever: F (1,12) <0.01, p =.949) ( Figure S1 ). Rats then received punishment sessions, where lever-presses on R1 and R2 continued to yield food, but every 10 th press on R1 was punished with footshock ( Figure 1b ). Rats were sensitive to this punishment schedule, selectively suppressing punished R1 responding relative to unpunished R2 (lever : F (1,12) =46.40, p <.001; group: F (1,12) =0.92, p =.357) ( Figure 1d ). When given a choice test, where both levers were presented together and no shocks were delivered, rats showed a strong preference for the unpunished lever (lever: F (1,12) =65.71, p <.001; group: F (1,12) =0.04, p =.843) ( Figure 1d ). VTA DA neural dynamics around appetitive and aversive outcomes When examining activity of VTA DA neurons around response-elicited outcomes, VTA DA neurons exhibited pronounced excitatory Ca 2+ transients to reward deliveries across sessions ( Figure 1i ). More surprisingly, excitatory transients were also observed to the shock punisher across punishment sessions. Critically, this excitatory shock transient began during shock delivery, and not simply to shock offset. This contradicts canonical accounts of VTA DA as reward coding, but is consistent with existing reports of some VTA DA subpopulations being excited by aversive events 5,26,27 . There were also outcome-related fluctuations in GABA input to VTA DA neurons ( Figure 1j ). There were significant increases in GABA following reward deliveries. We also observed a sharp increase in GABA input in response to shocks during initial punishment sessions. Shock-related GABA transients during later punishment sessions were notably smaller and did not significantly deviate from baseline. Interestingly, in stark contrast to VTA DA neuron activity, phasic GABA signals were greater to shock than to reward. These observations conform with the idea that GABA input to VTA DA provides a negative prediction error signal, suppressing VTA DA neuron activity during expected rewards (i.e., rewards cued by the sound of pellet delivery) and unexpected aversive events 5,15,28 . VTA DA neural dynamics around punished versus unpunished actions To examine whether VTA neural dynamics tracked changing action values under punishment, we examined signals around punished vs. unpunished actions alone (i.e., actions not coinciding with outcome deliveries). VTA DA neurons exhibited punishment-related changes to activity around actions ( Figure 1k ). Prior to punishment, VTA DA neurons exhibited transient reductions in activity around each action relative to baseline. As punishment was learned, punished actions began eliciting excitatory transients, as previously reported 29 , whereas unpunished actions retained their inhibitory activity pattern across punishment sessions. Generally, we observed modest decreases in GABA signal in the lead up to actions ( Figure 1l ). Besides a modest unexpected difference in GABA signal to punished versus unpunished actions in training, GABA release around punished versus unpunished actions were not significantly distinguished across punishment and choice. Altogether, these findings indicate task-relevant fluctuations in VTA DA population activity and GABA input to VTA DA . In partial agreement with traditional reward prediction error accounts, VTA DA neurons were more strongly activated by rewards than aversive events, while GABA inputs to VTA DA were more pronounced to aversive events. These dissociated dynamics to motivationally relevant events are thought to contribute to the reinforcing vs. punishing effects of outcomes on antecedent actions. We next sought to causally examine how GABA signaling within VTA contributes to punishment. Experiment 2: Effects of GABA A blockade in VTA during punishment To examine this, we implanted bilateral guide cannulae into VTA of wild-type rats ( Figure 2a ), and blocked GABA A -mediated inhibition in VTA across phases of the punishment task ( Figure 2d ). Post-experiment histology confirmed 13 subjects had bilateral VTA placements ( Figure 2b ). Lever-press training and punishment acquisition Prior to punishment, rats acquired similarly high rates of responding on R1 and R2 across lever-press training (lever: F (1,11) =4.36 , p =.061) ( Figure S2 ). Rats then received punishment sessions, where R1 responses were punished with shock ( Figure 2c ). Overall, rats were sensitive to this punishment schedule, suppressing responding on punished R1 more than unpunished R2 ( F (1,11) =256.75, p <.001) ( Figure 2e ). Rats were also slower to initially press R1 relative to R2 across punishment ( F (1,11) =30.79, p <.001) ( Figure 2f ). To examine the role of GABA inhibition in VTA on this learning, rats received microinfusions of GABA A antagonist bicuculline (A-Bic group; n =6) or control saline (A-Sal group; n =7) into their VTA before the first 2 sessions of punishment. GABA A blockade in VTA, attenuated punishment avoidance during infusion days, such that A-Bic rats suppressed punished R1 responding less than A-Sal rats ( F (1,11) =9.05 , p =.012) ( Figure 2e ), significantly increasing the number of shocks incurred ( F (1,11) =10.60, p =.008) ( Figure S2c ). Bicuculline also attenuated the increase in latency to initially press the punished lever (group*session: F (1,11) =6.58 , p =.026) ( Figure 2f ). Crucially, bicuculline had no effect on unpunished R2 response ratios (group: F (1,11) =0.02 , p =.893) or latencies (group: F (1,11) =3.23 , p =.100; group*session: F (1,11) =0.15 , p =.707) during infusion days. Interestingly, this effect of bicuculline persisted in subsequent non-infusion sessions. Despite 3 additional non-infusion days to learn punishment avoidance, A-Bic rats continued to show less R1 suppression ( F (1,11) =6.72 , p =.025; Figure 2e ) and shorter latencies to press R1 ( F (1,11) =13.96 , p =.003; Figure 2f ) than A-Sal rats across remaining acquisition sessions. Groups did not differ in R2 suppression ( F (1,11) =0.05 , p =.822) or latencies ( F (1,11) <0.01, p =.983) during these sessions. This indicates VTA GABA A blockade during initial punishment produced enduring, consequential insensitivity to punishment. Punishment expression We then examined the effect of GABA A blockade on expression of learned punishment avoidance. All rats received bicuculline or saline across two punishment sessions (within-subjects, counterbalanced). The effect of acquisition infusions on punished responding persisted into expression tests; A-Bic rats pressed the punished lever more than A-Sal rats overall (group: F (1,11) =14.00, p =.003) ( Figure 2g ), incurring substantially more shock punishment ( F (1,11) = 16.15 , p =.002) ( Figure S2d ), without any significant group differences in R2 responding ( F (1,11) =0.06, p =.812). There was no acute effect of expression drug on punished responding ( F (1,12) =0.29 , p =.600) ( Figure 2g ). However, there was a modest decrease in unpunished responding ( F (1,12) =5.60 , p =.036). There was no interaction between acquisition group and expression drug on punished ( F (1,11) =0.31 , p =.588) or unpunished ( F (1,11) =0.01 , p =.920) response ratios. To further examine whether GABA A blockade during expression test impaired later punishment avoidance, as found for acquisition infusions, we compared punished responding in pre- versus post-bicuculline punishment sessions. Bicuculline had no effect on the subsequent day’s punishment suppression (session: F (1,11) =0.10 , p =.757; session[A-Sal]: F (1,6) =0.60 , p =.468) ( Figure S2b ). In terms of lever-press latencies, A-Bic rats continued to press R1 faster than A-Sal rats ( F (1,11) =5.61 , p =.037) ( Figure 2h ), with no acquisition group differences for R2 latencies ( F (1,11) =1.20 , p =.297). Expression drug did not significantly affect latencies to press R1 ( F (1,11) =0.13 , p =.724) or R2 ( F (1,11) =4.68 , p =.053), nor did it interact with acquisition group on lever-press latencies (drug*group: F (1,11) =0.012 , p =.915; drug*group*lever: F (1,11) =0.674 , p =.429). Taken together, this suggests the effects of VTA GABA blockade on punished behavior are not observed once punishment is already learned. However, there may be a modest role for GABA action in VTA in directing animals towards the unpunished lever. Choice test Rats were then given two unpunished choice tests (bicuculline vs. saline), each flanked by non-infusion punishment sessions to limit any carry-over effects of these tests ( Figure 2d ). Overall, rats preferred the unpunished lever over the punished lever during these tests ( F (1,11) =58.16, p<.001) ( Figure 2i ). There was no main effect of acquisition group ( F (1,11) =1.072 , p =.323) or choice infusion ( F (1,11) =0.531 , p =.481), but there was a significant interaction of group, choice infusion, and lever ( F (1,11) =7.586 , p =.019). Consistent with the persistent impairment in punishment avoidance, A-Bic rats pressed the punished lever more than A-Sal rats during saline choice tests ( F (1,11) =15.18 , p =.002). This was not observed during bicuculline choice tests ( F (1,11) =0.237 , p =.636) tests; bicuculline significantly increased selection of R1 in A-Sal ( F (1,6) =9.39 , p =.022) but not A-Bic ( F (1,5) =2.73 , p =.159) rats. Acquisition group did not interact with choice infusion for unpunished responding ( F (1,11) =1.57 , p =.236). Effects of VTA disinhibition on open field activity Rats then received bicuculline or saline infusions (within-subjects, counterbalanced) prior to an open field test. A-Bic rats were hyperactive compared to A-Sal rats ( F (1,11) =20.18 , p =.001) ( Figure 2j ).There was no acute effect of bicuculline ( F (1,11) =0.01 , p =.972), nor any interaction between acquisition group and open field infusion ( F (1,11) =2.02 , p =.183), on distance travelled. In summary, the findings of Experiment 2 indicate that preventing GABA A inhibition in VTA during initial punishment learning, but not already-learned punishment, drives an enduring impairment in punishment avoidance and hyperactivity. Experiment 3: Effects of chemogenetic activation of VTA DA neurons during punishment The findings of Experiment 2 broadly conform with the notion that inhibition of midbrain dopamine neurons mediates aversive learning. However, GABA also act on non-dopamine neurons within VTA 16 , so the effects of bicuculline into VTA could be mediated by effects on other VTA populations. We therefore tested whether direct upregulation of VTA dopamine neuron activity during punishment learning produces enduring impairments in punishment avoidance. To examine this, we expressed excitatory designer receptor hm3D in VTA DA neurons of TH::Cre+ rats ( Figure 3a-b ). Rats then underwent the same task described for Experiment 2, except rats received systemic injections of CNO or control vehicle instead of microinfusions; A-CNO group ( n =5) received CNO during acquisition injections whereas A-Veh group ( n =5) received vehicle. Lever-press training and punishment acquisition Prior to punishment, rats acquired similar rates of pressing on both levers (lever: F (1,8) =1.57, p =.246; lever*group: F (1,8) =0.08, p =.790) ( Figure S3a ). During punishment, R1 responding was suppressed (ratio: F (1,8) =138.39, p<.001; latencies: F (1,8) =14.22, p =.005), relative to unpunished R2 responses ( Figure 3e-f ). Chemogenetic activation of dopamine neurons during initial punishment sessions (A-CNO group) produced acute suppression of both punished (group: F (1,8) =10.69, p =.011) and unpunished responding (group: F (1,8) =28.91, p =.001) ( Figure 3e ). It is worth noting all A-CNO animals still made responses on both levers, with a non-significant trend towards more unpunished responding (lever [A-CNO only]: F (1,4) =5.95, p =.071). All animals pressed enough to receive shock(s) during initial punishment ( Figure S3c ). On following non-injection days, responding rebounded in A-CNO animals. A-CNO group pressed the unpunished lever at similarly high rates to A-Veh group ( F (1,8) =0.03, p =.876). However, A-CNO group suppressed punished responding significantly less ( F (1,8) =7.08, p =.029), engaged the punished lever significantly faster ( F (1,8) =5.85, p =.042), and received many more shocks ( F (1,8) = 10.33, p =.012) ( Figure S3c ) than A-Veh animals across non-injection days. This was not solely attributable to delayed learning due to initially reduced responding as A-CNO R1 responding (and shocks incurred) across these sessions remained higher than even the first session of punishment for A-Veh group. This indicates that activation of VTA DA during initial punishment learning produces enduring punishment insensitivity, as found for VTA disinhibition using GABA A blockade. Punishment expression Prior to days 7 and 8 of punishment, rats received CNO or vehicle injections (within-subjects, counterbalanced). A-CNO animals continued to show punishment insensitivity; they pressed the punished ( F (1,8) =6.46, p =.035) but not unpunished ( F (1,8) =0.78, p =.403) lever significantly more than A-Veh following vehicle injections ( Figure 3g ), . As observed during acquisition injections, CNO administration acutely suppressed responding (drug: F (1,8) =39.99, p <.001) ( Figure 3g ) and increased latencies to press levers (drug: F (1,8) =6.56, p =.034) ( Figure 3h ), without eliminating discriminated responding (lever[CNO]: F (1,8) =47.38, p <.001). Choice test Rats were then given two choice tests (CNO vs. Veh, counterbalanced). Overall, rats preferred the unpunished over punished lever ( F (1,8) =78.95, p <.001) and CNO broadly suppressed responding (drug: F (1,8) =23.52, p =.001) ( Figure 3i ). Acquisition group did not significantly interact with effects of lever or choice injection (all F (1,8) ≤1.881, p ≥.207). Open field activity VTA DA activation via CNO profoundly increased distance travelled in the open field test (drug: F (1,8) =38.18, p<.001) ( Figure 3j ). In contrast to Experiment 2, locomotor activity did not depend on acquisition group (group: F (1,8) =0.273, p =.615; group*drug: F (1,8) =0.259, p =.625). Discussion Avoiding punishment is a core component of adaptive behavior. The current study explored the role of VTA DA inhibition in punishment learning and choice. Using fiber photometry to record VTA DA dynamics (Experiment 1), we observed phasic increases in VTA DA neuron activity and GABA input around response-elicited appetitive and aversive events. VTA DA activity was more reward-biased, whereas GABA input was punisher-biased (at least during initial punishment). This generally conforms with traditional theories that GABA inhibition of VTA DA during adverse events drives punishment learning 4,5 . Testing this, we blocked GABA A inhibition in VTA (Experiment 2) or directly activated VTA DA neurons (Experiment 3) and showed disinhibiting VTA during initial punishment learning induced long-term impairments in punishment avoidance. This accords with previous studies that show chemogenetic activation of VTA DA promotes risky decision-making 30,31 . Interestingly, we found acute disinhibition of VTA after punishment was learned did not induce subsequent insensitivity. Together, these findings suggest long-term avoidance depends upon a critical window of GABA-mediated VTA DA inhibition during initial punishment learning. One explanation for the time-sensitive effect of GABA A blockade and hm3D activation on avoidance is that these manipulations prevented normal inhibitory prediction error signaling within VTA DA during the initially unexpected shock outcomes. In theory, this would undermine aversive learning about the antecedent action. Indeed, we found punisher-elicited GABA efflux onto VTA DA neurons was most pronounced during initial punishment, as predicted by aversive prediction error accounts of VTA DA 5 . Although parsimonious, this interpretation is speculative, as manipulations in this study were not restricted to the moment of shock delivery. However, previous studies have shown brief optogenetic inhibition of VTA DA , delivered in the same manner as shocks were in the current study, was sufficient to drive punishment avoidance 13 . Together, these findings suggest punisher-elicited inhibition of VTA DA is both sufficient and necessary for the acquisition of punishment avoidance. One observation that deviates from this punishment-driven inhibition account was that VTA DA population activity generally increased during the footshock punisher, despite concurrent increases in GABA input. This highlights the dissociation between VTA DA activity and its inhibitory inputs. An additional consideration here is the heterogeneity of signaling across VTA DA neuron subtypes. Seminal reports of VTA DA being broadly inhibited by aversive events were from neurons with a specific electrophysiological signature, which ignored VTA DA neuron subtypes that do not share this signature (and are excited by aversive events) 32,33 . Measurement from the broader population of genetically-defined VTA DA neurons, as done here, often report excitatory VTA DA transients to aversive events 5,29,34,32 . The current study does not provide insight into whether manipulation effects were mediated by specific VTA DA subtypes or circuits 32,33,35 . It is plausible the manipulation effects were specifically due to actions on subpopulations that receive increased GABA input during punishers, but further exploration of the cell-type and circuit basis of effects are needed. Indeed, VTA DA neurons project to several regions strongly implicated in punishment avoidance, such as nucleus accumbens (NAc) and basolateral amygdala 4,16,32,36 . Elevated dopamine in nucleus accumbens is associated with increased risk-taking under punishment 31,37 , suggesting disinhibition within the VTA-NAc circuit could mediate the effects observed in the current study. Another open question is whether the enduring insensitivity found here represents a broad behavioral deficit that would carry over to new punishing scenarios, or is instead specific to the punished action, punisher, and/or context in which VTA DA disinhibition occurred. For example, VTA disinhibitions may have specifically altered the motivational value of the experienced shock (e.g., via counterconditioning 38 ). Alternatively, VTA disinhibitions may have undermined normal Action-Punisher association learning (a common locus for naturally-occurring punishment insensitivity 39,40 ). This latter idea accords with newer theories of dopamine which argue dopamine signals do not simply compute model-free prediction errors, but instead help build cognitive maps of relationships between actions, cues and outcomes 41,42 . Our observation that disinhibition-induced insensitivity was accompanied by locomotor hyperactivity in a different context suggests the perturbation extends beyond the punishment scenario in which VTA signaling was disrupted, but it will be important to examine whether insensitivity is observed with other actions, punishers, or contexts. A broader implication of the current findings is that brief perturbations of dopamine systems can cause long-lasting, selective impairments in avoiding harm. This has relevance for substance addictions, which are diagnostically characterized by the persistence of drug-seeking and -taking despite negative consequences 43 . Addictive substances across drug classes are known to artificially elevate dopamine and/or disrupt inhibitory input to dopamine neurons 44 . The current study highlights a potential mechanistic connection between these substances and their tendency to drive compulsive (i.e., punishment insensitive) drug-taking. Substance-induced deficits in appropriately learning about the negative consequences of drug-seeking may coalesce with addictive substances’ other effects on cognition, motivation, and neural circuit functioning 45–50 to drive the complex and difficult-to-treat nature of drug addiction. A key question that follows is whether deficits in harm avoidance observed here can be reversed or, in the case of anticipated hyperdopaminergic states (e.g., pharmacotherapies 51 ), ameliorated. In summary, our findings identify inhibitory input to VTA DA as a critical mechanism for adaptive punishment avoidance. Disrupting inhibition within VTA or directly upregulating VTA DA activity during initial punishment learning caused long-term deficits in avoidance. Further investigation is needed to identify the psychological nature of these deficits, the specific circuits and plasticity mechanisms mediating these effects, and how they might be reversed to restore adaptive choice. Declarations Funding and Disclosure This work was supported by grants from the Australian Research Council to PJRDB and SK (DP220102317) and GPM (DP220100040). Funders had no role in study design, data collection, analysis, decision to publish, or preparation of the manuscript. Author contributions SYST : Investigation, Analysis, Writing – original draft. MHS : Investigation, Analysis, Writing – review & editing. LJK: Investigation, Analysis, Writing – review & editing. MWS: Investigation, Analysis, Writing – review & editing. GPM : Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review & editing. SK: Funding acquisition, Supervision, Writing – review & editing. PJRDB : Conceptualization, Investigation, Analysis, Funding acquisition, Project administration, Supervision, Writing – original draft, Writing – review & editing. References Mackintosh, N. J. Conditioning and Associative Learning . (Clarendon Press [u.a.], Oxford, 1983). Schultz, W. Neuronal Reward and Decision Signals: From Theories to Data. Physiol. Rev. 95 , 853–951 (2015). Lerner, T. N., Holloway, A. L. & Seiler, J. L. Dopamine, Updated: Reward Prediction Error and Beyond. Curr. Opin. Neurobiol. 67 , 123–130 (2021). Jean-Richard-Dit-Bressel, P., Killcross, S. & McNally, G. P. Behavioral and neurobiological mechanisms of punishment: implications for psychiatric disorders. Neuropsychopharmacology 43 , 1639–1650 (2018). Bromberg-Martin, E. S., Matsumoto, M. & Hikosaka, O. Dopamine in Motivational Control: Rewarding, Aversive, and Alerting. Neuron 68 , 815–834 (2010). Pyon, W. S., Bizon, J. L. & Setlow, B. Neural Mechanisms of Decision Making Under Risk of Punishment: Insights From Rodent Models. WIREs Cogn. Sci. 16 , e70012 (2025). Adamantidis, A. R. et al. Optogenetic Interrogation of Dopaminergic Modulation of the Multiple Phases of Reward-Seeking Behavior. J. Neurosci. 31 , 10829–10835 (2011). Zweifel, L. S. et al. Disruption of NMDAR-dependent burst firing by dopamine neurons provides selective assessment of phasic dopamine-dependent behavior. Proc. Natl. Acad. Sci. 106 , 7281–7288 (2009). Kim, K. M. et al. Optogenetic Mimicry of the Transient Activation of Dopamine Neurons by Natural Reward Is Sufficient for Operant Reinforcement. PLOS ONE 7 , e33612 (2012). Fraser, K. M., Pribut, H. J., Janak, P. H. & Keiflin, R. From Prediction to Action: Dissociable Roles of Ventral Tegmental Area and Substantia Nigra Dopamine Neurons in Instrumental Reinforcement. J. Neurosci. 43 , 3895–3908 (2023). Mileykovskiy, B. & Morales, M. Duration of Inhibition of Ventral Tegmental Area Dopamine Neurons Encodes a Level of Conditioned Fear. J. Neurosci. 31 , 7471–7476 (2011). Matsumoto, M. & Hikosaka, O. Two types of dopamine neuron distinctly convey positive and negative motivational signals. Nature 459 , 837–841 (2009). Peng, C. Y., Jean-Richard-dit-Bressel, P., Gilchrist, S., Power, J. M. & McNally, G. P. Phasic inhibition of dopamine neurons is an instrumental punisher. Behav. Neurosci. 135 , 415–425 (2021). Danjo, T., Yoshimi, K., Funabiki, K., Yawata, S. & Nakanishi, S. Aversive behavior induced by optogenetic inactivation of ventral tegmental area dopamine neurons is mediated by dopamine D2 receptors in the nucleus accumbens. Proc. Natl. Acad. Sci. 111 , 6455–6460 (2014). Tan, K. R. et al. GABA Neurons of the VTA Drive Conditioned Place Aversion. Neuron 73 , 1173–1183 (2012). Morales, M. & Margolis, E. B. Ventral tegmental area: cellular heterogeneity, connectivity and behaviour. Nat. Rev. Neurosci. 18 , 73–85 (2017). Eshel, N. et al. Arithmetic and local circuitry underlying dopamine prediction errors. Nature 525 , 243–246 (2015). Lobb, C. J., Wilson, C. J. & Paladini, C. A. A Dynamic Role for GABA Receptors on the Firing Pattern of Midbrain Dopaminergic Neurons. J. Neurophysiol. 104 , 403–413 (2010). Ji, H. & Shepard, P. D. Lateral Habenula Stimulation Inhibits Rat Midbrain Dopamine Neurons through a GABAA Receptor-Mediated Mechanism. J. Neurosci. 27 , 6923–6930 (2007). Gomez, J. L. et al. Chemogenetics revealed: DREADD occupancy and activation via converted clozapine. Science 357 , 503–507 (2017). Jean-Richard-Dit-Bressel, P. & McNally, G. P. The Role of the Lateral Habenula in Punishment. PLOS ONE 9 , e111699 (2014). Jean-Richard-Dit-Bressel, P. & McNally, G. P. The role of the basolateral amygdala in punishment. Learn. Mem. 22 , 128–137 (2015). Jean-Richard-dit-Bressel, P., Tran, J., Didachos, A. & McNally, G. P. Instrumental aversion coding in the basolateral amygdala and its reversion by a benzodiazepine. Neuropsychopharmacology 47 , 1199–1209 (2022). Keevers, L. J. & Jean-Richard-dit-Bressel, P. Obtaining artifact-corrected signals in fiber photometry via isosbestic signals, robust regression, and dF/F calculations. Neurophotonics 12 , 025003 (2025). Jean-Richard-dit-Bressel, P., Clifford, C. W. G. & McNally, G. P. Analyzing Event-Related Transients: Confidence Intervals, Permutation Tests, and Consecutive Thresholds. Front. Mol. Neurosci. 13 , 14 (2020). Brischoux, F., Chakraborty, S., Brierley, D. I. & Ungless, M. A. Phasic excitation of dopamine neurons in ventral VTA by noxious stimuli. Proc. Natl. Acad. Sci. 106 , 4894–4899 (2009). de Jong, J. W. et al. A Neural Circuit Mechanism for Encoding Aversive Stimuli in the Mesolimbic Dopamine System. Neuron 101 , 133-151.e7 (2019). Eshel, N. et al. Arithmetic and local circuitry underlying dopamine prediction errors. Nature 525 , 243–246 (2015). Jacobs, D. S., Allen, M. C., Park, J. & Moghaddam, B. Learning of probabilistic punishment as a model of anxiety produces changes in action but not punisher encoding in the dmPFC and VTA. eLife 11 , e78912 (2022). Hynes, T. J. et al. Win-Paired Cues Modulate the Effect of Dopamine Neuron Sensitization on Decision Making and Cocaine Self-administration: Divergent Effects Across Sex. Biol. Psychiatry 95 , 220–230 (2024). Verharen, J. P. H. et al. A neuronal mechanism underlying decision-making deficits during hyperdopaminergic states. Nat. Commun. 9 , 731 (2018). Lammel, S., Lim, B. K. & Malenka, R. C. Reward and aversion in a heterogeneous midbrain dopamine system. Neuropharmacology 76 , 351–359 (2014). Lammel, S. et al. Unique Properties of Mesoprefrontal Neurons within a Dual Mesocorticolimbic Dopamine System. Neuron 57 , 760–773 (2008). Verharen, J. P. H., Luijendijk, M. C. M., Vanderschuren, L. J. M. J. & Adan, R. A. H. Dopaminergic contributions to behavioral control under threat of punishment in rats. Psychopharmacology (Berl.) 237 , 1769–1782 (2020). Kim, C. K. et al. Simultaneous fast measurement of circuit dynamics at multiple sites across the mammalian brain. Nat. Methods 13 , 325–328 (2016). Piantadosi, P. T., Halladay, L. R., Radke, A. K. & Holmes, A. Advances in understanding meso-cortico-limbic-striatal systems mediating risky reward seeking. J. Neurochem. 157 , 1547–1571 (2021). Freels, T. G., Gabriel, D. B. K., Lester, D. B. & Simon, N. W. Risky decision-making predicts dopamine release dynamics in nucleus accumbens shell. Neuropsychopharmacology 45 , 266–275 (2020). Pearce, J. M. & Dickinson, A. Pavlovian countercondition: Changing the suppressive properties of shock by association with food. J. Exp. Psychol. Anim. Behav. Process. 1 , 170–177 (1975). Jean-Richard-dit-Bressel, P., Ma, C., Bradfield, L. A., Killcross, S. & McNally, G. P. Punishment insensitivity emerges from impaired contingency detection, not aversion insensitivity or reward dominance. eLife 8 , e52765 (2019). Jean-Richard-dit-Bressel, P. et al. Punishment insensitivity in humans is due to failures in instrumental contingency learning. eLife 10 , e69594 (2021). Sharpe, M. J. et al. Dopamine transients do not act as model-free prediction errors during associative learning. Nat. Commun. 11 , 106 (2020). Jeong, H. et al. Mesolimbic dopamine release conveys causal associations. Science 378 , eabq6740 (2022). American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders . (Washington, DC, 2013). Pierce, R. C. & Kumaresan, V. The mesolimbic dopamine system: The final common pathway for the reinforcing effect of drugs of abuse? Neurosci. Biobehav. Rev. 30 , 215–238 (2006). Kalivas, P. W. & Volkow, N. D. The Neural Basis of Addiction: A Pathology of Motivation and Choice. Am. J. Psychiatry 162 , 1403–1413 (2005). Koob, G. F. & Volkow, N. D. Neurobiology of addiction: a neurocircuitry analysis. Lancet Psychiatry 3 , 760–773 (2016). Ramey, T. & Regier, P. S. Cognitive Impairment in Substance Use Disorders. CNS Spectr. 24 , 102–113 (2019). McNally, G. P. & Jean-Richard-dit-Bressel, P. A Cognitive Pathway to Persistent, Maladaptive Choice. Eur. Addict. Res. 30 , 233–242 (2024). McNally, G. P., Jean-Richard-dit-Bressel, P., Millan, E. Z. & Lawrence, A. J. Pathways to the persistence of drug use despite its adverse consequences. Mol. Psychiatry 28 , 2228–2237 (2023). Smith, R. J. & Laiks, L. S. Behavioral and neural mechanisms underlying habitual and compulsive drug seeking. Prog. Neuropsychopharmacol. Biol. Psychiatry 87 , 11–21 (2018). Cools, R. Dopaminergic modulation of cognitive function-implications for L-DOPA treatment in Parkinson’s disease. Neurosci. Biobehav. Rev. 30 , 1–23 (2006). Additional Declarations There is NO conflict of interest to disclose. Supplementary Files SupplementaryMaterials251010.docx Supplemental Material Cite Share Download PDF Status: Published Journal Publication published 17 Feb, 2026 Read the published version in Neuropsychopharmacology → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7865029","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":530607298,"identity":"7f3eafea-c210-467b-b487-0ec672a68282","order_by":0,"name":"Philip Jean-Richard-Dit-Bressel","email":"data:image/png;base64,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","orcid":"https://orcid.org/0000-0002-0898-8987","institution":"UNSW Sydney","correspondingAuthor":true,"prefix":"","firstName":"Philip","middleName":"","lastName":"Jean-Richard-Dit-Bressel","suffix":""},{"id":530607299,"identity":"68c178d7-c67b-4099-9810-eae3ebd900f9","order_by":1,"name":"Shannen Tan","email":"","orcid":"","institution":"UNSW Sydney","correspondingAuthor":false,"prefix":"","firstName":"Shannen","middleName":"","lastName":"Tan","suffix":""},{"id":530607300,"identity":"cb5a3fbe-d698-4090-b1ee-727fc004fd20","order_by":2,"name":"Michelle Shen","email":"","orcid":"","institution":"UNSW Sydney","correspondingAuthor":false,"prefix":"","firstName":"Michelle","middleName":"","lastName":"Shen","suffix":""},{"id":530607301,"identity":"ae856484-1468-47cf-bce4-1e523c65eab8","order_by":3,"name":"Luke Keevers","email":"","orcid":"","institution":"UNSW Sydney","correspondingAuthor":false,"prefix":"","firstName":"Luke","middleName":"","lastName":"Keevers","suffix":""},{"id":530607302,"identity":"6ed35ecc-67f3-4aa9-82a8-fd9ced986ba9","order_by":4,"name":"Matthew Williams-Spooner","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"","lastName":"Williams-Spooner","suffix":""},{"id":530607303,"identity":"33d89ec0-6c4c-4c3e-a8bb-054aa80abe39","order_by":5,"name":"Gavan McNally","email":"","orcid":"https://orcid.org/0000-0001-9061-6463","institution":"UNSW","correspondingAuthor":false,"prefix":"","firstName":"Gavan","middleName":"","lastName":"McNally","suffix":""},{"id":530607304,"identity":"69661b04-ef3e-4963-8882-880c26e3c6ef","order_by":6,"name":"Simon Killcross","email":"","orcid":"","institution":"Cardiff University","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Killcross","suffix":""}],"badges":[],"createdAt":"2025-10-15 07:36:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7865029/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7865029/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41386-026-02368-4","type":"published","date":"2026-02-17T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":93743774,"identity":"86fd77d3-742d-4e0b-9a62-5320b59d0a6d","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"tif","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5814812,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1.tif","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/6d59c19ec06e45603185de7c.tif"},{"id":93743771,"identity":"c9db586b-27d1-4c68-9e1c-692226d5c30f","added_by":"auto","created_at":"2025-10-17 06:06:35","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2329521,"visible":true,"origin":"","legend":"","description":"","filename":"VTApunishment251014.docx","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/0aa7739ec004f137ba429ae3.docx"},{"id":93743776,"identity":"d6a3ef9a-91fc-41d1-aa8b-a907f2d410d3","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2360908,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2.tif","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/a769540c4720fe10fc086e8f.tif"},{"id":93743786,"identity":"af33d22e-db43-44f2-9506-7502bb4a640c","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14911500,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3.tif","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/1f87d4fae39bace1c44a6f63.tif"},{"id":93743781,"identity":"7b16bf6c-0885-46f3-9f04-600cb6a538ec","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"json","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7947,"visible":true,"origin":"","legend":"","description":"","filename":"NPP251493.json","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/f904ac2e555383c541039264.json"},{"id":93743780,"identity":"5fad5e92-cb65-4b00-aea5-05c03713e055","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1140706,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterials251010.docx","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/2ff2d59982997bff2d50faa1.docx"},{"id":93743789,"identity":"671f2f88-fa12-4919-9cb6-289eb577a85a","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"xml","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":139504,"visible":true,"origin":"","legend":"","description":"","filename":"NPP2514930enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/cd9f00a1fc928f07c32b9cef.xml"},{"id":93743791,"identity":"5c9a7dd3-4c70-4f29-96d6-3449fa276d67","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"tif","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5814812,"visible":true,"origin":"","legend":"","description":"","filename":"Fig1.tif","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/eb247fa952ffaa379773fbd7.tif"},{"id":93743778,"identity":"d131caf2-ae3d-443d-8dd0-2d8f7688570a","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"tif","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2360908,"visible":true,"origin":"","legend":"","description":"","filename":"Fig2.tif","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/13ac0b10647da5fcb22913a3.tif"},{"id":93744457,"identity":"323de7f4-b930-4344-bb98-e37f47f36797","added_by":"auto","created_at":"2025-10-17 06:14:36","extension":"tif","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14911500,"visible":true,"origin":"","legend":"","description":"","filename":"Fig3.tif","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/b73ac65b2687899bc12d5000.tif"},{"id":93744456,"identity":"d10b4e00-6926-4ce6-bc0e-d82c38396b5f","added_by":"auto","created_at":"2025-10-17 06:14:36","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":701149,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/9787d1337e43629857e5b022.png"},{"id":93743775,"identity":"fa502504-67c9-40e5-8c34-f4edae8d9a1e","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":420312,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/daa945198f47f2a4d351270e.png"},{"id":93743792,"identity":"d97cd86d-dcd6-4df8-baad-712569061794","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1092792,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/f679bf6f4b954b781ce758e4.png"},{"id":93743793,"identity":"e3e186b0-1973-4834-a27f-721e95298e68","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":590247,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig1.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/816e819d28b0b88363d171ad.png"},{"id":93743794,"identity":"f9b4bb41-6ac2-4701-9a1d-c56061954785","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":210704,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig2.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/ecae94962549b330001d1fdf.png"},{"id":93743784,"identity":"ec23429e-5306-45b3-8468-82d9fdd237db","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1558770,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFig3.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/3efdb824dc999e8311770708.png"},{"id":93743783,"identity":"83de83a3-220b-4b79-b916-19a3e8036f81","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":157442,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/27cba22ce9aa515710b5e80d.png"},{"id":93744459,"identity":"e7471563-14be-4d8e-ab34-98b4497b9efc","added_by":"auto","created_at":"2025-10-17 06:14:36","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":106864,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/fec8fc232d9f6d64630ad068.png"},{"id":93744458,"identity":"a01d6fdd-4062-4de1-872c-a104a77ca6f8","added_by":"auto","created_at":"2025-10-17 06:14:36","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":193593,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/d2838fbed38df0b5dc119195.png"},{"id":93743790,"identity":"c8c7bef4-9105-4dda-9725-7dc8b8fe3228","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"xml","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":135927,"visible":true,"origin":"","legend":"","description":"","filename":"NPP2514930structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/6a9675bad47565ee05ec966e.xml"},{"id":93743787,"identity":"10ef1618-7b72-4823-bcea-3fec1cf526ab","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"html","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":152285,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/eecd197231504a5116920753.html"},{"id":93743777,"identity":"f270d539-0832-4c9f-a1f3-716336d9b85f","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":10521469,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eVTA\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eDA\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e activity and GABA input during punishment learning.\u003c/strong\u003e \u003cstrong\u003e[A]\u003c/strong\u003e TH::Cre+ rats received Cre-dependent calcium or GABA sensors and fiber implant into VTA. \u003cstrong\u003e[B]\u003c/strong\u003e Punishment task. Rats could press two levers (R1, R2 [5min alternating blocks]) for food. In punishment sessions, R1 responses also yielded footshock (FR10 schedule). During choice test, both levers were presented to assess lever preference. \u003cstrong\u003e[C] \u003c/strong\u003eTimeline of task sessions. \u003cstrong\u003e[D]\u003c/strong\u003e Mean ± SEM lever-press ratios for the last session of lever training (T), punishment sessions (1-6) and choice (C) across subjects (\u003cem\u003eN\u003c/em\u003e=14). Punishment led to robust, selective suppression of R1 responding, and a strong preference for unpunished R2 over previously-punished R1 during choice test. \u003cstrong\u003e[E]\u003c/strong\u003e Example GCaMP expression and fiber placement. \u003cstrong\u003e[F]\u003c/strong\u003e Placement map for rats with valid GCaMP and fiber tip locations (\u003cem\u003en\u003c/em\u003e=7). \u003cstrong\u003e[G]\u003c/strong\u003e Example iGABASnFR expression and fiber placement. \u003cstrong\u003e[H]\u003c/strong\u003e Placement map for rats with valid iGABASnFR and fiber tip\u003c/p\u003e","description":"","filename":"Fig1.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/a3b719f01c7d9fcc34609bf1.png"},{"id":93744455,"identity":"22162b11-476c-4055-9888-de2fb55e817d","added_by":"auto","created_at":"2025-10-17 06:14:35","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3006476,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffect of VTA GABA\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eA\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e blockade on punished behavior.\u003c/strong\u003e\u0026nbsp; \u003cstrong\u003e[A]\u003c/strong\u003e Bilateral guide cannulae were implanted into VTA of wild-type rats [\u003cstrong\u003eB]\u003c/strong\u003e Cannulae placements for animals with valid placements, according to acquisition drug group (A-Sal [\u003cem\u003en\u003c/em\u003e=7], A-Bic\u003cem\u003e \u003c/em\u003e[\u003cem\u003en\u003c/em\u003e=6]). [\u003cstrong\u003eC]\u003c/strong\u003e Punishment task design. [\u003cstrong\u003eD] \u003c/strong\u003eTimeline of task sessions, with arrows indicating when subjects received intra-VTA infusions of GABA\u003csub\u003eA\u003c/sub\u003e antagonist bicuculline (Bic) and/or control saline (Sal). A-Bic vs. A-Sal groups received Bic vs. Sal (respectively) before the first 2 sessions of punishment. All groups received Bic vs. Sal (within-subjects, order counterbalanced) in subsequent punishment expression, choice, and locomotor tests. \u003cstrong\u003e[E]\u003c/strong\u003e Mean ± SEM lever suppression ratios by acquisition group across last day of training (T) and punishment acquisition sessions. Grey shaded area indicates infusion sessions. Rats that received Bic (A-Bic) exhibited a persistent deficit in punishment avoidance. \u003cstrong\u003e[F]\u003c/strong\u003e Mean ± SEM lever-press latencies by acquisition group across last day of training (T) and punishment acquisition sessions. A-Bic rats were quicker to press the punished lever relative to A-Sal rats.\u0026nbsp; \u003cstrong\u003e[G]\u003c/strong\u003e Mean ± SEM lever suppression ratios during punishment expression tests per acquisition group. A-Bic rats continued to press punished R1 more than A-Sal rats; Bic infusions during expression tests had no effect on R1 responding.\u003cstrong\u003e [H]\u003c/strong\u003e Mean ± SEM lever-press latencies during punishment expression tests. A-Bic rats continued to press punished R1 more quickly than A-Sal rats; Bic infusions during expression tests had no effect on R1 responding.\u003cstrong\u003e [I]\u003c/strong\u003e Mean ± SEM suppression ratios during choice tests. Bic selectively increased R1 responding in A-Sal group.\u003cstrong\u003e [J]\u003c/strong\u003e Mean ± SEM distance travelled during locomotor tests. A-Bic rats travelled further than A-Sal animals. Bic infusions during these tests had no significant effect on this. *\u003cem\u003ep\u003c/em\u003e\u0026lt;.05\u003c/p\u003e","description":"","filename":"Fig2.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/3eb3112a602ff9ef12c2ca8e.png"},{"id":93743773,"identity":"96b9167d-0ca7-446b-8ebd-42f2e0cfcabd","added_by":"auto","created_at":"2025-10-17 06:06:36","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":28817757,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEffects of chemogenetic activation of VTA\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eDA\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e neurons during punishment.\u003c/strong\u003e\u0026nbsp; \u003cstrong\u003e[A]\u003c/strong\u003e TH::Cre+ rats received Cre-dependent excitatory hm3D DREADD bilaterally into the VTA. \u003cstrong\u003e[B] \u003c/strong\u003ehm3D expression across animals included in analyses (\u003cem\u003eN\u003c/em\u003e = 10). \u003cstrong\u003e[C] \u003c/strong\u003eExample expression of hm3D within VTA.\u0026nbsp; \u003cstrong\u003e[D] \u003c/strong\u003eTimeline of task sessions, with arrows indicating when subjects received i.p. injections of CNO and/or vehicle control (Veh). A-CNO vs. A-Veh groups received CNO vs. Veh (respectively) before the first 2 sessions of punishment. All groups received CNO vs. Veh (within-subjects, order counterbalanced) in subsequent punishment expression, choice, and locomotor tests. \u003cstrong\u003e[E]\u003c/strong\u003e Mean ± SEM lever suppression ratios by acquisition group across last day of training (T) and punishment acquisition sessions. Grey shaded area indicates injection sessions. CNO acutely (but incompletely) reduced responding. On subsequent non-injection days, A-CNO rats exhibited a persistent deficit in punishment avoidance.\u0026nbsp; \u003cstrong\u003e[F]\u003c/strong\u003e Mean ± SEM lever-press latencies by acquisition group across last day of training (T) and punishment acquisition sessions. CNO administration did not significantly affect lever-press latencies acutely. However, A-CNO rats pressed the punished R1 lever significantly faster than A-Veh rats on subsequent non-injection days. \u003cstrong\u003e[G]\u003c/strong\u003e Mean ± SEM lever suppression ratios during punishment expression tests. CNO injections acutely (but incompletely) reduced responding across acquisition groups. A-CNO rats continued to press punished R1 more than A-Veh rats during control injections. \u003cstrong\u003e[H]\u003c/strong\u003e Mean ± SEM lever-press latencies during punishment expression tests.\u003cstrong\u003e [I]\u003c/strong\u003e Mean ± SEM suppression ratios during choice tests.\u003cstrong\u003e [J]\u003c/strong\u003e Mean ± SEM distance travelled during locomotor tests. CNO acutely increased distance travelled across groups. *\u003cem\u003ep\u003c/em\u003e\u0026lt;.05\u003c/p\u003e","description":"","filename":"Fig3.png","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/9afc48aa65538cdcd38f0cbe.png"},{"id":102822607,"identity":"ad700410-2c70-4296-88b3-b0efb8419530","added_by":"auto","created_at":"2026-02-17 08:08:14","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":40575373,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/8d740c04-bccb-4d86-9dec-fb8a6094ecb2.pdf"},{"id":93743770,"identity":"1b6fd0da-c148-4ead-b4da-4938eb8e0681","added_by":"auto","created_at":"2025-10-17 06:06:35","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1140706,"visible":true,"origin":"","legend":"Supplemental Material","description":"","filename":"SupplementaryMaterials251010.docx","url":"https://assets-eu.researchsquare.com/files/rs-7865029/v1/f6b93f063b581bbc89fbe900.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose.","formattedTitle":"Disinhibition of ventral tegmental area during initial punishment learning causes enduring punishment insensitivity","fulltext":[{"header":"Introduction","content":"\u003cp\u003eActions with positive consequences tend to be reinforced (i.e., repeated), whereas actions with negative consequences tend to be punished (i.e., suppressed)\u003csup\u003e1\u003c/sup\u003e. This fundamental adaptive function, known as instrumental conditioning, helps organisms dynamically adjust their behavior to maximize rewards and minimize harms.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDopaminergic neurons of the ventral midbrain, including those within the ventral tegmental area (VTA\u003csub\u003eDA\u003c/sub\u003e), are widely considered to be critical for this learning\u003csup\u003e2\u0026ndash;6\u003c/sup\u003e. These neurons exhibit phasic increases in activity to better-than-expected rewards, which are considered necessary and sufficient for reinforcing behaviors\u003csup\u003e7\u0026ndash;10\u003c/sup\u003e. Conversely, VTA\u003csub\u003eDA\u003c/sub\u003e neurons exhibit phasic decreases in activity to worse-than-expected outcomes (including aversive events)\u003csup\u003e5,11,12\u003c/sup\u003e, and inhibition of VTA\u003csub\u003eDA\u003c/sub\u003e has been shown to be aversive\u003csup\u003e13\u0026ndash;15\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCrucially, when specific actions cause brief optogenetic inhibition of VTA\u003csub\u003eDA\u003c/sub\u003e neurons, those actions (but not other actions) are suppressed\u003csup\u003e13\u003c/sup\u003e. This shows pauses in VTA\u003csub\u003eDA\u003c/sub\u003e neuron activity can function as punishment to produce selective instrumental avoidance. Endogenously, pauses in VTA\u003csub\u003eDA\u003c/sub\u003e activity are driven by GABA input to VTA\u003csub\u003eDA\u003c/sub\u003e neurons from local interneurons and long-range GABA inputs\u003csup\u003e5,16,17\u003c/sup\u003e, which act on GABA\u003csub\u003eA\u003c/sub\u003e receptors to suppress VTA\u003csub\u003eDA\u003c/sub\u003e firing\u003csup\u003e15,18,19\u003c/sup\u003e. However, patterns of GABA release onto VTA\u003csub\u003eDA\u003c/sub\u003e neurons during punishment and the necessity of VTA\u003csub\u003eDA\u003c/sub\u003e inhibition in punishment avoidance remain unclear.\u003c/p\u003e\n\u003cp\u003eTo examine this, we performed fiber photometry recordings of VTA\u003csub\u003eDA\u003c/sub\u003e calcium (Ca\u003csup\u003e2+\u003c/sup\u003e; neural activity proxy) and GABA input across a punishment task to characterize how VTA\u003csub\u003eDA\u003c/sub\u003e and GABA dynamics relate to behavior under punishment. We tested the causal significance of VTA GABA input and VTA\u003csub\u003eDA\u003c/sub\u003e activity on punishment learning and choice via pharmacological (GABA\u003csub\u003eA\u003c/sub\u003e receptor antagonist) and chemogenetic manipulations (hm3D DREADD) across phases of punishment, and show preventing GABA\u003csub\u003eA\u003c/sub\u003e-mediated inhibition of VTA and direct excitation of VTA\u003csub\u003eDA\u003c/sub\u003e during initial punishment learning produces enduring punishment insensitivity.\u0026nbsp;\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eFurther details for Methods are supplied in \u003cb\u003eSupplemental Materials\u003c/b\u003e.\u003c/p\u003e\n\u003ch3\u003eSubjects\u003c/h3\u003e\n\u003cp\u003eAll experiments used experimentally-naive rats aged 8\u0026ndash;24 weeks old. Photometry and chemogenetic experiments used heterozygous TH::Cre Sprague Dawley rats (SD-Th-cre\u003csup\u003etm1sage\u003c/sup\u003e; Sage Laboratories). TH::Cre\u0026thinsp;+\u0026thinsp;animals express Cre in tyrosine hydroxylase (TH; dopamine precursor enzyme) neurons; TH::Cre- animals (used in control experiment reported in Supplemental Materials) do not express Cre. Experiment 2 (pharmacological manipulation) used wild-type Sprague Dawley rats.\u003c/p\u003e\u003cp\u003eAnimals were group-housed (4) in plastic cages in a climate-controlled colony room maintained on a 12 hr light\u0026ndash;dark cycle. Rats had \u003cem\u003ead libitum\u003c/em\u003e access to chow until 2 days before behavioral training, after which they received 10\u0026ndash;15 g chow daily (after behavioral session) to maintain them at ~\u0026thinsp;90% of their free-feeding weight. Rats had access to water in their homecages throughout experiments. All procedures were approved by the Animal Care and Ethics Committee at UNSW Sydney and conducted in accordance with the National Health and Medical Research Council Code for the Care and Use of Animals for Scientific Purposes in Australia (2013).\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eApparatus \u0026amp; Materials\u003c/h2\u003e\u003cp\u003eAll operant behavior was assessed in MedAssociates operant chambers, each housed within light and sound-attenuating cabinets. Each chamber contained two retractable levers that flanked a magazine port where grain pellet rewards were delivered. The punisher was a 0.5sec footshock, delivered through the grid floor. Footshock intensity was 0.4mA for fiber photometry experiments, and 0.5mA for neural manipulation experiments. A lower intensity footshock was chosen for photometry experiments to avoid floor effects in responding that would undermine key analyses of peri-event dynamics.\u003c/p\u003e\u003cp\u003eLocomotor tests were conducted in open field chambers that tracked movement via 16-beam infrared arrays located along X- and Y-axes.\u003c/p\u003e\u003cp\u003eFiber photometry recordings were conducted using Doric Lenses photometry components (465nm and 405nm LEDs, mini-cubes, photodetectors) and Tucker Davis Technologies photometry processor (RZ5P).\u003c/p\u003e\u003cp\u003eAdeno-associated viruses (AAVs) were used to express Cre-dependent calcium sensor (AAV-CAG-DIO-GCaMP6f), GABA sensor (AAV-hSyn-DIO-iGABASnFR-F102G), or excitatory DREADD (AAV-hSyn-DIO-hm3D-mCherry) in VTA\u003csub\u003eDA\u003c/sub\u003e neurons of TH::Cre\u0026thinsp;+\u0026thinsp;animals.\u003c/p\u003e\u003cp\u003eMicroinfusions of GABA\u003csub\u003eA\u003c/sub\u003e antagonist bicuculline (0.1\u0026micro;g/\u0026micro;l; Tocris, Sydney, Australia) were used to prevent GABA-mediated inhibition in VTA\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e,\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. Systemic injections of 3mg/kg clozapine-\u003cem\u003eN\u003c/em\u003e-oxide (CNO; National Institute of Mental Health Chemical Synthesis and Drug Supply Program), dissolved in 5% DMSO and saline, were used to activate hm3D. Potential off-target effects of CNO\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e were addressed via TH::Cre- control subjects (\u003cb\u003eSupplemental Materials\u0026ndash;Figure S4\u003c/b\u003e).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eSurgeries\u003c/h3\u003e\n\u003cp\u003eRats were anaesthetized and placed into a flat skull position within a stereotaxic frame. Craniotomies were performed above VTA. For photometry and chemogenetic experiments, a 5\u0026micro;l 30-gauge microinfusion syringe (Hamilton; Reno, NV, USA) was used to inject 0.75\u0026micro;l AAVs (0.25\u0026micro;l/min) encoding Cre-dependent GCaMP6f (unilateral), iGABASnFR (unilateral), or hm3D (bilateral) into VTA (AP: -5.5, ML: \u0026plusmn;0.8, DV: -8.2 from bregma) of TH::Cre rats. Following injections, the syringe remained at the injection site for an additional 5min for diffusion.\u003c/p\u003e\u003cp\u003eFor photometry experiments, a 400\u0026micro;m optic fiber was unilaterally implanted into VTA (AP: -5.5, ML: \u0026plusmn;0.8, DV: -8.2 from bregma). For pharmacology experiments, a bilateral 26-gauge 11mm guide cannula (PlasticsOne) was implanted into VTA (AP: -5.8, ML: \u0026plusmn;0.75, DV: -8.2 from bregma). Implants were anchored in position with dental cement and jeweller\u0026rsquo;s screws. Immediately following surgery, animals were given antibiotics and received post-operative monitoring and care for 1 week. Rats that received AAV injections were given an additional 3 weeks before behavioral training to allow sufficient transgene expression.\u003c/p\u003e\n\u003ch3\u003eBehavioral task\u003c/h3\u003e\n\u003cp\u003eAll rats underwent a previously validated punishment task, which has been shown to elicit robust punishment avoidance with minimal contamination from Pavlovian fear\u003csup\u003e\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003eLever-press training\u003c/h3\u003e\n\u003cp\u003eRats were first trained to press two levers (R1, R2) for food. For 2 sessions, both levers were presented concurrently, and each press on a lever was rewarded with a pellet (FR1 training). A lever remained extended until it received 25 presses or after 1 hour. Rats that failed to acquire lever-pressing were manually shaped in the second FR1 session.\u003c/p\u003e\u003cp\u003eRats then received 7\u0026ndash;8 days of VI30s training (40min sessions). In these sessions, levers were presented individually for 5min blocks in alternating fashion (first lever randomized per day). Lever-presses were reinforced on a 30sec variable interval (VI30s) schedule, such that the first press after an average interval of 30 seconds led to pellet delivery.\u003c/p\u003e\n\u003ch3\u003ePunishment\u003c/h3\u003e\n\u003cp\u003eSubjects then received daily 40min punishment sessions. Lever-pressing on either lever continued to yield pellets (VI30s). However, every 10th press (FR10) on the punished R1 lever resulted in immediate footshock delivery. Presses on the unpunished R2 lever had no additional consequence. If a press was scheduled to deliver both footshock and pellet, both were delivered. Assignment of left vs. right levers as punished vs. unpunished was counterbalanced across (but not within) subjects.\u003c/p\u003e\u003cp\u003eFor pharmacology experiments, rats received intra-VTA infusions of 0.5\u0026micro;l GABA\u003csub\u003eA\u003c/sub\u003e antagonist bicuculline or control saline (0.25\u0026micro;l/min; 1min diffusion) immediately prior to the first two sessions of punishment (between-subjects), and bicuculline vs. saline on punishment days 6 and 7 (within-subjects, order counterbalanced). This design permits efficient interrogation of neural manipulation effects on acquisition and expression of punishment avoidance\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThe same design was employed for DREADD manipulations, except rats received i.p. injections of CNO or vehicle (30mins before session start) instead of microinfusions, and expression tests were conducted on punishment days 7 and 8.\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eChoice Test\u003c/h2\u003e\u003cp\u003eRats were then given choice test(s) where both levers were presented concurrently. No shocks were delivered and presses on either lever delivered pellets on a shared VI60s schedule, so there was no advantage to pressing either lever exclusively or a combination of both levers.\u003c/p\u003e\u003cp\u003ePhotometry experiments only involved a single 15min choice test. For manipulation experiments, animals received within-subjects drug vs. control across two choice tests (order counterbalanced) (pharmacology experiment: 30min tests; DREADD experiment: 20min tests). Each choice test was preceded by a drug-free punishment session the day prior\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e,\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eLocomotor tests\u003c/h3\u003e\n\u003cp\u003eEffects of VTA manipulation on locomotion were assessed following completion of the punishment task. Rats first received a 30 min habituation session, where they were placed into the open field chamber without any injections. On the following 2 days, rats received drug or control injection (within-subjects, counterbalanced order) before being placed into the chambers for 30 mins to assess distance travelled.\u003c/p\u003e\n\u003ch3\u003eHistology\u003c/h3\u003e\n\u003cp\u003eAt the end of all experiments, brain tissue was examined to verify virus expression and/or implant locations.\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eData Analysis\u003c/h2\u003e\u003cp\u003eRats that failed to acquire lever-pressing during lever training, or had inappropriate virus expression or implant placements, were excluded from all analyses.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eBehavior analysis\u003c/h2\u003e\u003cp\u003eThe key behavioral dependent measures were self-normalized rates of responding on each lever (\u0026ldquo;suppression ratios\u0026rdquo;)\u003csup\u003e22\u003c/sup\u003e, and average latency to initially press each lever across trials (averaged per session).\u003c/p\u003e\u003cp\u003eSuppression ratios normalize response rates per lever during punishment and choice sessions to pre-punishment (final VI30s) rates. This was calculated per lever as follows:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Suppression\\:ratio=\\frac{Session\\:LP\\:rate}{(Session\\:LP\\:rate+Training\\:LP\\:rate)}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eSuppression ratios can range from 0 to 1. Scores above 0.5 indicate greater lever-pressing relative to training, scores below 0.5 indicate less lever-pressing, while a score of 0.5 indicates no difference relative training. This was done to address any spurious difference in punished or unpunished response rates prior to punishment. Nonetheless, analyses of raw response rates are provided in Supplemental Materials.\u003c/p\u003e\u003cp\u003eBehavioral data was analyzed using repeated measures ANOVA. Within-subjects factors were lever, session, and drug. Between-subjects factor was acquisition group (drug vs. control). For all analyses, Type 1 error was controlled at 0.05.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003eFiber photometry analysis\u003c/h2\u003e\u003cp\u003e465nm (neural dynamic-related) and 405nm (isosbestic control) signals and event timestamps were extracted into MATLAB, and signals during logged disconnections were discarded. Each signal was low-pass (3 Hz) and notch (1.0322\u0026ndash;1.0326, 2.547\u0026ndash;2.55 Hz) filtered to remove high-frequency noise identified via Fast Fourier Transform. Filtered 405nm signals were fit to filtered 465nm signals via iteratively-reweighted least squares\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e to create fitted 405nm signals. A normalized fluorescence change score (dF/F) was calculated using the standard formula:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:dF/F=\\frac{(465nm\\:signal-fitted\\:405nm)}{fitted\\:405nm}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThis motion-artifact-corrected dF/F was detrended via 60 sec moving median (5 sec mean smoothing window). Detrended signals were converted into standard deviation units by dividing session signals by their sum squared deviation from 0 (nullZ-score)\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. All photometry analyses were derived from this normalized, artifact-corrected dF/F.\u003c/p\u003e\u003cp\u003eThe key dependent variable was change in VTA\u003csub\u003eDA\u003c/sub\u003e activity and GABA input around response-elicited outcomes (reward delivery, footshock) and actions (R1, R2). dF/F around pellets vs. footshocks, and R1 vs. R2 lever-presses alone (i.e., those not yielding footshock or pellets) were collated. Each trial was re-zeroed to pre-event baseline (-5:-3 sec) and averaged per subject; all analyses used mean peri-event transients per subject. Due to the scarcity of punished lever-presses and footshocks in late punishment sessions, late punishment data (Pun4 onwards) was combined to obtain more accurate peri-event activity traces per subject. Significant transients were identified via bootstrapped confidence intervals (CI)\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e. Bootstrapped means were obtained by randomly resampling from subject mean waveforms with replacement (1000 iterations). 95% CI limits were derived from 2.5 and 97.5 percentiles of bootstrap distribution, expanded by a factor of \u0026radic;(n/(n-1)). A significant transient was identified as a period that CI limits did not contain 0 (baseline) for at least 1/3secs (low-pass filter window\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e). Significant differences between event waveforms were similarly determined by bootstrapping the within-subject difference waveform (mean event1\u0026ndash;mean event2 waveform) per subject\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e\u003cem\u003eExperiment 1:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eVTA\u003csub\u003eDA\u003c/sub\u003e neuron a\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003ectivity and GABA input\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eduring punishment learning\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe first examined activity of VTA\u003csub\u003eDA\u003c/sub\u003e neurons and GABA input to VTA\u003csub\u003eDA\u003c/sub\u003e neurons across punishment learning and choice (\u003cstrong\u003eFigure 1\u003c/strong\u003e). This was done by selectively expressing GCaMP6f (Ca\u003csup\u003e2+\u003c/sup\u003e sensor; neural activity proxy) or iGABASnFR (GABA sensor; GABA input proxy) in VTA\u003csub\u003eDA\u003c/sub\u003e neurons of TH::Cre+ rats and recording from VTA across a punishment task (\u003cstrong\u003eFigure 1a-c\u003c/strong\u003e). \u003cem\u003eN\u003c/em\u003e = 14 rats had valid biosensor expression and fiber placements (\u003cem\u003en\u003c/em\u003e = 7 GCaMP [\u003cstrong\u003eFigure 1f\u003c/strong\u003e]; \u003cem\u003en\u003c/em\u003e = 7 iGABASnFR [\u003cstrong\u003eFigure 1h\u003c/strong\u003e]) and were thus included in analyses.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eTask Behavior\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAnimals first received lever-press training, where they could press two individually-presented levers (R1, R2) for food (\u003cstrong\u003eFigure 1b\u003c/strong\u003e). Across this training, rats acquired similarly high rates of responding on R1 and R2 (lever: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=0.10, \u003cem\u003ep\u003c/em\u003e=.763); this did not depend on which sensor animals expressed (group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=1.19, \u003cem\u003ep\u003c/em\u003e=.297; group*lever: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e\u0026lt;0.01, \u003cem\u003ep\u003c/em\u003e=.949) (\u003cstrong\u003eFigure S1\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRats then received punishment sessions, where lever-presses on R1 and R2 continued to yield food, but every 10\u003csup\u003eth\u003c/sup\u003e press on R1 was punished with footshock (\u003cstrong\u003eFigure 1b\u003c/strong\u003e). Rats were sensitive to this punishment schedule, selectively suppressing punished R1 responding relative to unpunished R2 (lever\u003cem\u003e: F\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=46.40, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001; group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=0.92, \u003cem\u003ep\u003c/em\u003e=.357)\u0026nbsp;(\u003cstrong\u003eFigure 1d\u003c/strong\u003e). When given a choice test, where both levers were presented together and no shocks were delivered, rats showed a strong preference for the unpunished lever (lever: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=65.71, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001; group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=0.04, \u003cem\u003ep\u003c/em\u003e=.843)\u0026nbsp;(\u003cstrong\u003eFigure 1d\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eVTA\u003csub\u003eDA\u003c/sub\u003e neural dynamics around appetitive and aversive outcomes\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWhen examining activity of VTA\u003csub\u003eDA\u003c/sub\u003e neurons around response-elicited outcomes, VTA\u003csub\u003eDA\u003c/sub\u003e neurons exhibited pronounced excitatory Ca\u003csup\u003e2+\u003c/sup\u003e transients to reward deliveries across sessions (\u003cstrong\u003eFigure 1i\u003c/strong\u003e). More surprisingly, excitatory transients were also observed to the shock punisher across punishment sessions. Critically, this excitatory shock transient began during shock delivery, and not simply to shock offset. This contradicts canonical accounts of VTA\u003csub\u003eDA\u003c/sub\u003e as reward coding, but is consistent with existing reports of some VTA\u003csub\u003eDA\u003c/sub\u003e subpopulations being excited by aversive events\u003csup\u003e5,26,27\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThere were also outcome-related fluctuations in GABA input to VTA\u003csub\u003eDA\u003c/sub\u003e neurons (\u003cstrong\u003eFigure 1j\u003c/strong\u003e). There were significant increases in GABA following reward deliveries. We also observed a sharp increase in GABA input in response to shocks during initial punishment sessions. Shock-related GABA transients during later punishment sessions were notably smaller and did not significantly deviate from baseline. Interestingly, in stark contrast to VTA\u003csub\u003eDA\u003c/sub\u003e neuron activity, phasic GABA signals were greater to shock than to reward. These observations conform with the idea that GABA input to VTA\u003csub\u003eDA\u003c/sub\u003e provides a negative prediction error signal, suppressing VTA\u003csub\u003eDA\u003c/sub\u003e neuron activity during expected rewards (i.e., rewards cued by the sound of pellet delivery) and unexpected aversive events\u003csup\u003e5,15,28\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eVTA\u003csub\u003eDA\u003c/sub\u003e neural dynamics around punished versus unpunished actions\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eTo examine whether VTA neural dynamics tracked changing action values under punishment, we examined signals around punished vs. unpunished actions alone (i.e., actions not coinciding with outcome deliveries).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eVTA\u003csub\u003eDA\u003c/sub\u003e neurons exhibited punishment-related changes to activity around actions (\u003cstrong\u003eFigure 1k\u003c/strong\u003e). Prior to punishment, VTA\u003csub\u003eDA\u003c/sub\u003e neurons exhibited transient reductions in activity around each action relative to baseline. As punishment was learned, punished actions began eliciting excitatory transients, as previously reported\u003csup\u003e29\u003c/sup\u003e, whereas unpunished actions retained their inhibitory activity pattern across punishment sessions.\u003c/p\u003e\n\u003cp\u003eGenerally, we observed modest decreases in GABA signal in the lead up to actions (\u003cstrong\u003eFigure 1l\u003c/strong\u003e). Besides a modest unexpected difference in GABA signal to punished versus unpunished actions in training, GABA release around punished versus unpunished actions were not significantly distinguished across punishment and choice.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAltogether, these findings indicate task-relevant fluctuations in VTA\u003csub\u003eDA\u003c/sub\u003e population activity and GABA input to VTA\u003csub\u003eDA\u003c/sub\u003e. In partial agreement with traditional reward prediction error accounts, VTA\u003csub\u003eDA\u003c/sub\u003e neurons were more strongly activated by rewards than aversive events, while GABA inputs to VTA\u003csub\u003eDA\u003c/sub\u003e were more pronounced to aversive events. These dissociated dynamics to motivationally relevant events are thought to contribute to the reinforcing vs. punishing effects of outcomes on antecedent actions. We next sought to causally examine how GABA signaling within VTA contributes to punishment.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExperiment 2: Effects of GABA\u003csub\u003eA\u003c/sub\u003e blockade in VTA during punishment\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo examine this, we implanted bilateral guide cannulae into VTA of wild-type rats (\u003cstrong\u003eFigure 2a\u003c/strong\u003e), and blocked GABA\u003csub\u003eA\u003c/sub\u003e-mediated inhibition in VTA across phases of the punishment task (\u003cstrong\u003eFigure 2d\u003c/strong\u003e).\u0026nbsp;Post-experiment histology confirmed\u0026nbsp;13 subjects had bilateral VTA placements (\u003cstrong\u003eFigure 2b\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eLever-press training and punishment acquisition\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePrior to punishment, rats acquired similarly high rates of responding on R1 and R2 across lever-press training (lever: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=4.36\u003cem\u003e, p\u003c/em\u003e=.061) (\u003cstrong\u003eFigure S2\u003c/strong\u003e). Rats then received punishment sessions, where R1 responses were punished with shock (\u003cstrong\u003eFigure 2c\u003c/strong\u003e). Overall, rats were sensitive to this punishment schedule, suppressing responding on punished R1 more than unpunished R2 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=256.75, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001) (\u003cstrong\u003eFigure 2e\u003c/strong\u003e). Rats were also slower to initially press R1 relative to R2 across punishment (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=30.79, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001) (\u003cstrong\u003eFigure 2f\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo examine the role of GABA inhibition in VTA on this learning, rats received microinfusions of GABA\u003csub\u003eA\u003c/sub\u003e antagonist bicuculline (A-Bic group; \u003cem\u003en\u003c/em\u003e=6) or control saline (A-Sal group; \u003cem\u003en\u003c/em\u003e=7) into their VTA before the first 2 sessions of punishment. GABA\u003csub\u003eA\u003c/sub\u003e blockade in VTA, attenuated punishment avoidance during infusion days, such that A-Bic rats suppressed punished R1 responding less than A-Sal rats (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=9.05\u003cem\u003e, p\u003c/em\u003e=.012) (\u003cstrong\u003eFigure 2e\u003c/strong\u003e), significantly increasing the number of shocks incurred (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=10.60, \u003cem\u003ep\u003c/em\u003e=.008) (\u003cstrong\u003eFigure S2c\u003c/strong\u003e). Bicuculline also attenuated the increase in latency to initially press the punished lever (group*session: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=6.58\u003cem\u003e, p\u003c/em\u003e=.026) (\u003cstrong\u003eFigure 2f\u003c/strong\u003e). Crucially, bicuculline had no effect on unpunished R2 response ratios (group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.02\u003cem\u003e, p\u003c/em\u003e=.893) or latencies (group:\u003cem\u003e\u0026nbsp;F\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e=3.23\u003cem\u003e, p\u003c/em\u003e=.100; group*session:\u003cem\u003e\u0026nbsp;F\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e=0.15\u003cem\u003e, p\u003c/em\u003e=.707) during infusion days.\u003c/p\u003e\n\u003cp\u003eInterestingly, this effect of bicuculline persisted in subsequent non-infusion sessions. Despite 3 additional non-infusion days to learn punishment avoidance, A-Bic rats continued to show less R1 suppression (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=6.72\u003cem\u003e, p\u003c/em\u003e=.025; \u003cstrong\u003eFigure 2e\u003c/strong\u003e) and shorter latencies to press R1 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=13.96\u003cem\u003e, p\u003c/em\u003e=.003; \u003cstrong\u003eFigure 2f\u003c/strong\u003e) than A-Sal rats across remaining acquisition sessions. Groups did not differ in R2 suppression (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.05\u003cem\u003e, p\u003c/em\u003e=.822) or latencies (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e\u0026lt;0.01, \u003cem\u003ep\u003c/em\u003e=.983) during these sessions. This indicates VTA GABA\u003csub\u003eA\u003c/sub\u003e blockade during initial punishment produced enduring, consequential insensitivity to punishment.\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePunishment expression\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eWe then examined the effect of GABA\u003csub\u003eA\u003c/sub\u003e blockade on expression of learned punishment avoidance. All rats received bicuculline or saline across two punishment sessions (within-subjects, counterbalanced).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe effect of acquisition infusions on punished responding persisted into expression tests; A-Bic rats pressed the punished lever more than A-Sal rats overall (group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=14.00, \u003cem\u003ep\u003c/em\u003e=.003) (\u003cstrong\u003eFigure 2g\u003c/strong\u003e), incurring substantially more shock punishment (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=\u0026nbsp;16.15\u003cem\u003e, p\u003c/em\u003e=.002) (\u003cstrong\u003eFigure S2d\u003c/strong\u003e), without any significant group differences in R2 responding (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.06, \u003cem\u003ep\u003c/em\u003e=.812). There was no acute effect of expression drug on punished responding (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=0.29 \u003cem\u003e, p\u003c/em\u003e=.600) (\u003cstrong\u003eFigure 2g\u003c/strong\u003e). However, there was a modest decrease in unpunished responding (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,12)\u003c/sub\u003e=5.60\u003cem\u003e, p\u003c/em\u003e=.036). There was no interaction between acquisition group and expression drug on punished (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.31\u003cem\u003e, p\u003c/em\u003e=.588) or unpunished (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.01\u003cem\u003e, p\u003c/em\u003e=.920) response ratios. To further examine whether GABA\u003csub\u003eA\u003c/sub\u003e blockade during expression test impaired later punishment avoidance, as found for acquisition infusions, we compared punished responding in pre- versus post-bicuculline punishment sessions. Bicuculline had no effect on the subsequent day\u0026rsquo;s punishment suppression (session: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.10\u003cem\u003e, p\u003c/em\u003e=.757; session[A-Sal]: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,6)\u003c/sub\u003e=0.60\u003cem\u003e, p\u003c/em\u003e=.468) (\u003cstrong\u003eFigure S2b\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn terms of lever-press latencies, A-Bic rats continued to press R1 faster than A-Sal rats (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=5.61\u003cem\u003e, p\u003c/em\u003e=.037) (\u003cstrong\u003eFigure 2h\u003c/strong\u003e), with no acquisition group differences for R2 latencies (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=1.20\u003cem\u003e, p\u003c/em\u003e=.297). Expression drug did not significantly affect latencies to press R1 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.13\u003cem\u003e, p\u003c/em\u003e=.724) or R2 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=4.68\u003cem\u003e, p\u003c/em\u003e=.053), nor did it interact with acquisition group on lever-press latencies (drug*group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.012\u003cem\u003e, p\u003c/em\u003e=.915; drug*group*lever: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.674\u003cem\u003e, p\u003c/em\u003e=.429).\u003c/p\u003e\n\u003cp\u003eTaken together, this suggests the effects of VTA GABA blockade on punished behavior are not observed once punishment is already learned. However, there may be a modest role for GABA action in VTA in directing animals towards the unpunished lever.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eChoice test\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eRats were then given two unpunished choice tests (bicuculline vs. saline), each flanked by non-infusion punishment sessions to limit any carry-over effects of these tests (\u003cstrong\u003eFigure 2d\u003c/strong\u003e). Overall, rats preferred the unpunished lever over the punished lever during these tests (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=58.16, p\u0026lt;.001) (\u003cstrong\u003eFigure 2i\u003c/strong\u003e). There was no main effect of acquisition group (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=1.072\u003cem\u003e, p\u003c/em\u003e=.323) or choice infusion (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.531\u003cem\u003e, p\u003c/em\u003e=.481), but there was a significant interaction of group, choice infusion, and lever (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=7.586\u003cem\u003e, p\u003c/em\u003e=.019). Consistent with the persistent impairment in punishment avoidance, A-Bic rats pressed the punished lever more than A-Sal rats during saline choice tests (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=15.18\u003cem\u003e, p\u003c/em\u003e=.002). This was not observed during bicuculline choice tests (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=0.237\u003cem\u003e, p\u003c/em\u003e=.636) tests; bicuculline significantly increased selection of R1 in A-Sal (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,6)\u003c/sub\u003e=9.39\u003cem\u003e, p\u003c/em\u003e=.022) but not A-Bic (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,5)\u003c/sub\u003e=2.73\u003cem\u003e, p\u003c/em\u003e=.159) rats. Acquisition group did not interact with choice infusion for unpunished responding (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=1.57\u003cem\u003e, p\u003c/em\u003e=.236). \u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eEffects of VTA disinhibition on open field activity\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eRats then received bicuculline or saline infusions (within-subjects, counterbalanced) prior to an open field test. A-Bic rats were hyperactive compared to A-Sal rats (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=20.18\u003cem\u003e, p\u003c/em\u003e=.001) (\u003cstrong\u003eFigure 2j\u003c/strong\u003e).There was no acute effect of bicuculline (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e=0.01\u003cem\u003e, p\u003c/em\u003e=.972), nor any interaction between acquisition group and open field infusion (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,11)\u003c/sub\u003e=2.02\u003cem\u003e, p\u003c/em\u003e=.183), on distance travelled.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn summary, the findings of Experiment 2 indicate that preventing GABA\u003csub\u003eA\u003c/sub\u003e inhibition in VTA during initial punishment learning, but not already-learned punishment, drives an enduring impairment in punishment avoidance and hyperactivity. \u003cstrong\u003e\u003cem\u003e\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eExperiment 3:\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e\u003cem\u003eEffects of chemogenetic activation of VTA\u003csub\u003eDA\u003c/sub\u003e neurons during punishment\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe findings of Experiment 2 broadly conform with the notion that inhibition of midbrain dopamine neurons mediates aversive learning. However, GABA also act on non-dopamine neurons within VTA\u003csup\u003e16\u003c/sup\u003e, so the effects of bicuculline into VTA could be mediated by effects on other VTA populations. We therefore tested whether direct upregulation of VTA dopamine neuron activity during punishment learning produces enduring impairments in punishment avoidance. To examine this, we expressed excitatory designer receptor hm3D in VTA\u003csub\u003eDA\u003c/sub\u003e neurons of TH::Cre+ rats (\u003cstrong\u003eFigure 3a-b\u003c/strong\u003e). Rats then underwent the same task described for Experiment 2, except rats received systemic injections of CNO or control vehicle instead of microinfusions; A-CNO group (\u003cem\u003en\u003c/em\u003e=5) received CNO during acquisition injections whereas A-Veh group (\u003cem\u003en\u003c/em\u003e=5) received vehicle.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eLever-press training and punishment acquisition\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePrior to punishment, rats acquired similar rates of pressing on both levers (lever: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=1.57, \u003cem\u003ep\u003c/em\u003e=.246; lever*group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=0.08, \u003cem\u003ep\u003c/em\u003e=.790) (\u003cstrong\u003eFigure S3a\u003c/strong\u003e). During punishment, R1 responding was suppressed (ratio: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=138.39, p\u0026lt;.001; latencies:\u003cem\u003e\u0026nbsp;F\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=14.22, \u003cem\u003ep\u003c/em\u003e=.005), relative to unpunished R2 responses (\u003cstrong\u003eFigure 3e-f\u003c/strong\u003e).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eChemogenetic activation of dopamine neurons during initial punishment sessions (A-CNO group) produced acute suppression of both punished (group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=10.69, \u003cem\u003ep\u003c/em\u003e=.011) and unpunished responding (group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=28.91, \u003cem\u003ep\u003c/em\u003e=.001) (\u003cstrong\u003eFigure 3e\u003c/strong\u003e). It is worth noting all A-CNO animals still made responses on both levers, with a non-significant trend towards more unpunished responding (lever [A-CNO only]: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,4)\u003c/sub\u003e=5.95, \u003cem\u003ep\u003c/em\u003e=.071). All animals pressed enough to receive shock(s) during initial punishment (\u003cstrong\u003eFigure S3c\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003eOn following non-injection days, responding rebounded in A-CNO animals. A-CNO group pressed the unpunished lever at similarly high rates to A-Veh group (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=0.03, \u003cem\u003ep\u003c/em\u003e=.876). However, A-CNO group suppressed punished responding significantly less (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=7.08, \u003cem\u003ep\u003c/em\u003e=.029), engaged the punished lever significantly faster (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=5.85, \u003cem\u003ep\u003c/em\u003e=.042), and received many more shocks (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e= 10.33, \u003cem\u003ep\u003c/em\u003e=.012) (\u003cstrong\u003eFigure S3c\u003c/strong\u003e) than A-Veh animals across non-injection days. This was not solely attributable to delayed learning due to initially reduced responding as A-CNO R1 responding (and shocks incurred) across these sessions remained higher than even the first session of punishment for A-Veh group. This indicates that activation of VTA\u003csub\u003eDA\u003c/sub\u003e during initial punishment learning produces enduring punishment insensitivity, as found for VTA disinhibition using GABA\u003csub\u003eA\u003c/sub\u003e blockade.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003ePunishment expression\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003ePrior to days 7 and 8 of punishment, rats received CNO or vehicle injections (within-subjects, counterbalanced). A-CNO animals continued to show punishment insensitivity; they pressed the punished (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=6.46, \u003cem\u003ep\u003c/em\u003e=.035) but not unpunished (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=0.78, \u003cem\u003ep\u003c/em\u003e=.403) lever significantly more than A-Veh following vehicle injections (\u003cstrong\u003eFigure 3g\u003c/strong\u003e), . As observed during acquisition injections, CNO administration acutely suppressed responding (drug: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=39.99, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001) (\u003cstrong\u003eFigure 3g\u003c/strong\u003e) and increased latencies to press levers (drug: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=6.56, \u003cem\u003ep\u003c/em\u003e=.034) (\u003cstrong\u003eFigure 3h\u003c/strong\u003e), without eliminating discriminated responding (lever[CNO]: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=47.38, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eChoice test\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eRats were then given two choice tests (CNO vs. Veh, counterbalanced). Overall, rats preferred the unpunished over punished lever (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=78.95, \u003cem\u003ep\u003c/em\u003e\u0026lt;.001) and CNO broadly suppressed responding (drug: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=23.52, \u003cem\u003ep\u003c/em\u003e=.001) (\u003cstrong\u003eFigure 3i\u003c/strong\u003e). Acquisition group did not significantly interact with effects of lever or choice injection (all \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e\u0026le;1.881, \u003cem\u003ep\u003c/em\u003e\u0026ge;.207).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eOpen field activity\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eVTA\u003csub\u003eDA\u003c/sub\u003e activation via CNO profoundly increased distance travelled in the open field test (drug: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=38.18, p\u0026lt;.001) (\u003cstrong\u003eFigure 3j\u003c/strong\u003e). In contrast to Experiment 2, locomotor activity did not depend on acquisition group (group: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=0.273, \u003cem\u003ep\u003c/em\u003e=.615; group*drug: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,8)\u003c/sub\u003e=0.259, \u003cem\u003ep\u003c/em\u003e=.625).\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eAvoiding punishment is a core component of adaptive behavior. The current study explored the role of VTA\u003csub\u003eDA\u003c/sub\u003e inhibition in punishment learning and choice. Using fiber photometry to record VTA\u003csub\u003eDA\u003c/sub\u003e dynamics (Experiment 1), we observed phasic increases in VTA\u003csub\u003eDA\u003c/sub\u003e neuron activity and GABA input around response-elicited appetitive and aversive events. VTA\u003csub\u003eDA\u003c/sub\u003e activity was more reward-biased, whereas GABA input was punisher-biased (at least during initial punishment). This generally conforms with traditional theories that GABA inhibition of VTA\u003csub\u003eDA\u003c/sub\u003e during adverse events drives punishment learning\u003csup\u003e4,5\u003c/sup\u003e. Testing this, we blocked GABA\u003csub\u003eA\u003c/sub\u003e inhibition in VTA (Experiment 2) or directly activated VTA\u003csub\u003eDA\u003c/sub\u003e neurons (Experiment 3) and showed disinhibiting VTA during initial punishment learning induced long-term impairments in punishment avoidance. This accords with previous studies that show chemogenetic activation of\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e promotes risky decision-making\u003csup\u003e30,31\u003c/sup\u003e. Interestingly, we found acute disinhibition of VTA after punishment was learned did not induce subsequent insensitivity. Together, these findings suggest long-term avoidance depends upon a critical window of GABA-mediated\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e inhibition during initial punishment learning.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOne\u0026nbsp;explanation\u0026nbsp;for the time-sensitive effect of GABA\u003csub\u003eA\u003c/sub\u003e blockade and hm3D activation on avoidance is that these manipulations prevented normal inhibitory prediction error signaling within\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e during the initially unexpected shock outcomes. In theory, this would undermine aversive learning about the antecedent action. Indeed, we found punisher-elicited GABA efflux onto\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e neurons was most pronounced during initial punishment, as predicted by aversive prediction error accounts of\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e\u003csup\u003e5\u003c/sup\u003e. Although parsimonious, this interpretation is speculative, as manipulations in this study were not restricted to the moment of shock delivery. However, previous studies have shown brief optogenetic inhibition of\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e, delivered in the same manner as shocks were in the current study, was sufficient to drive punishment avoidance\u003csup\u003e13\u003c/sup\u003e. Together, these findings suggest punisher-elicited inhibition of VTA\u003csub\u003eDA\u003c/sub\u003e is both sufficient and necessary for the acquisition of punishment avoidance.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOne\u0026nbsp;observation\u0026nbsp;that deviates from this punishment-driven inhibition account was that\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e population\u0026nbsp;activity generally increased during the footshock punisher, despite concurrent increases in GABA input. This highlights the dissociation between\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e activity and its inhibitory inputs. An additional consideration here is the heterogeneity of signaling across VTA\u003csub\u003eDA\u003c/sub\u003e neuron subtypes.\u0026nbsp;Seminal reports of\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e being broadly inhibited by aversive events were from\u0026nbsp;neurons with a\u0026nbsp;specific electrophysiological signature, which ignored\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e neuron subtypes that do not share this signature (and are excited by aversive events)\u003csup\u003e32,33\u003c/sup\u003e. Measurement from the broader population of genetically-defined\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e neurons, as done here, often\u0026nbsp;report excitatory\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e transients to aversive events\u003csup\u003e5,29,34,32\u003c/sup\u003e. The current study does not provide insight into whether manipulation effects were mediated by specific\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e subtypes or\u0026nbsp;circuits\u003csup\u003e32,33,35\u003c/sup\u003e. It is plausible the manipulation effects were specifically due to actions on subpopulations that receive increased GABA input during punishers, but further exploration of the cell-type and circuit basis of effects are needed. Indeed,\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e neurons project to several regions strongly implicated in punishment avoidance, such as nucleus accumbens (NAc) and basolateral amygdala\u003csup\u003e4,16,32,36\u003c/sup\u003e. Elevated dopamine in nucleus accumbens is associated with increased risk-taking under punishment\u003csup\u003e31,37\u003c/sup\u003e, suggesting disinhibition within the VTA-NAc circuit could mediate the effects observed in the current study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAnother open question is whether the enduring insensitivity found here represents a broad behavioral deficit that would carry over to new punishing scenarios, or is instead specific to the punished action, punisher, and/or context in which VTA\u003csub\u003eDA\u003c/sub\u003e disinhibition occurred. For example, VTA disinhibitions may have specifically altered the motivational value of the experienced shock (e.g., via counterconditioning\u003csup\u003e38\u003c/sup\u003e). Alternatively, VTA disinhibitions may have undermined normal Action-Punisher association learning (a common locus for naturally-occurring punishment insensitivity\u003csup\u003e39,40\u003c/sup\u003e). This latter idea accords with newer theories of dopamine which argue dopamine signals do not simply compute model-free prediction errors, but instead help build cognitive maps of relationships between actions, cues and outcomes\u003csup\u003e41,42\u003c/sup\u003e. Our observation that disinhibition-induced insensitivity was accompanied by locomotor hyperactivity in a different context suggests the perturbation extends beyond the punishment scenario in which VTA signaling was disrupted, but it will be important to examine whether insensitivity is observed with other actions, punishers, or contexts.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA broader\u0026nbsp;implication\u0026nbsp;of the current findings is that brief perturbations of dopamine systems can cause long-lasting, selective impairments in avoiding harm. This has relevance for substance addictions, which are diagnostically characterized by the persistence of drug-seeking and -taking despite negative consequences\u003csup\u003e43\u003c/sup\u003e. Addictive substances across drug classes are known to artificially elevate dopamine and/or disrupt inhibitory input to dopamine neurons\u003csup\u003e44\u003c/sup\u003e. The current study highlights a potential mechanistic connection between these substances and their tendency to drive compulsive (i.e., punishment insensitive) drug-taking.\u0026nbsp;Substance-induced deficits in appropriately learning about the negative consequences of drug-seeking may coalesce with addictive substances\u0026rsquo; other effects on cognition, motivation, and neural circuit functioning\u003csup\u003e45\u0026ndash;50\u003c/sup\u003e to drive the complex and difficult-to-treat nature of drug addiction. A key question that follows is whether deficits in harm avoidance observed here can be reversed or, in the case of anticipated hyperdopaminergic states (e.g., pharmacotherapies\u003csup\u003e51\u003c/sup\u003e), ameliorated.\u003c/p\u003e\n\u003cp\u003eIn\u0026nbsp;summary, our findings identify inhibitory input to\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e as a critical mechanism for adaptive punishment avoidance. Disrupting inhibition within VTA or directly upregulating\u0026nbsp;VTA\u003csub\u003eDA\u003c/sub\u003e activity during initial punishment learning caused long-term deficits in avoidance. Further investigation is needed to identify the psychological nature of these deficits, the specific circuits and plasticity mechanisms mediating these effects, and how they might be reversed to restore adaptive choice.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding and Disclosure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by grants from the Australian Research Council to PJRDB and SK (DP220102317) and GPM (DP220100040). Funders had no role in study design, data collection, analysis, decision to publish, or preparation of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSYST\u003c/strong\u003e: Investigation, Analysis, Writing \u0026ndash; original draft. \u003cstrong\u003eMHS\u003c/strong\u003e: Investigation, Analysis, Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003eLJK:\u003c/strong\u003e Investigation, Analysis, Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003eMWS:\u003c/strong\u003e Investigation, Analysis, Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003eGPM\u003c/strong\u003e: Conceptualization, Funding acquisition, Project administration, Supervision, Writing \u0026ndash; review \u0026amp; editing.\u003cstrong\u003e\u0026nbsp;SK:\u0026nbsp;\u003c/strong\u003eFunding acquisition, Supervision, Writing \u0026ndash; review \u0026amp; editing. \u003cstrong\u003ePJRDB\u003c/strong\u003e: Conceptualization, Investigation, Analysis, Funding acquisition, Project administration, Supervision, Writing \u0026ndash; original draft, Writing \u0026ndash; review \u0026amp; editing.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eMackintosh, N. J. \u003cem\u003eConditioning and Associative Learning\u003c/em\u003e. (Clarendon Press [u.a.], Oxford, 1983).\u003c/li\u003e\n\u003cli\u003eSchultz, W. Neuronal Reward and Decision Signals: From Theories to Data. \u003cem\u003ePhysiol. Rev.\u003c/em\u003e \u003cstrong\u003e95\u003c/strong\u003e, 853\u0026ndash;951 (2015).\u003c/li\u003e\n\u003cli\u003eLerner, T. N., Holloway, A. L. \u0026amp; Seiler, J. L. Dopamine, Updated: Reward Prediction Error and Beyond. \u003cem\u003eCurr. Opin. Neurobiol.\u003c/em\u003e \u003cstrong\u003e67\u003c/strong\u003e, 123\u0026ndash;130 (2021).\u003c/li\u003e\n\u003cli\u003eJean-Richard-Dit-Bressel, P., Killcross, S. \u0026amp; McNally, G. P. Behavioral and neurobiological mechanisms of punishment: implications for psychiatric disorders. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 1639\u0026ndash;1650 (2018).\u003c/li\u003e\n\u003cli\u003eBromberg-Martin, E. S., Matsumoto, M. \u0026amp; Hikosaka, O. Dopamine in Motivational Control: Rewarding, Aversive, and Alerting. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e68\u003c/strong\u003e, 815\u0026ndash;834 (2010).\u003c/li\u003e\n\u003cli\u003ePyon, W. S., Bizon, J. L. \u0026amp; Setlow, B. Neural Mechanisms of Decision Making Under Risk of Punishment: Insights From Rodent Models. \u003cem\u003eWIREs Cogn. Sci.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, e70012 (2025).\u003c/li\u003e\n\u003cli\u003eAdamantidis, A. R. \u003cem\u003eet al.\u003c/em\u003e Optogenetic Interrogation of Dopaminergic Modulation of the Multiple Phases of Reward-Seeking Behavior. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 10829\u0026ndash;10835 (2011).\u003c/li\u003e\n\u003cli\u003eZweifel, L. S. \u003cem\u003eet al.\u003c/em\u003e Disruption of NMDAR-dependent burst firing by dopamine neurons provides selective assessment of phasic dopamine-dependent behavior. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e106\u003c/strong\u003e, 7281\u0026ndash;7288 (2009).\u003c/li\u003e\n\u003cli\u003eKim, K. M. \u003cem\u003eet al.\u003c/em\u003e Optogenetic Mimicry of the Transient Activation of Dopamine Neurons by Natural Reward Is Sufficient for Operant Reinforcement. \u003cem\u003ePLOS ONE\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, e33612 (2012).\u003c/li\u003e\n\u003cli\u003eFraser, K. M., Pribut, H. J., Janak, P. H. \u0026amp; Keiflin, R. From Prediction to Action: Dissociable Roles of Ventral Tegmental Area and Substantia Nigra Dopamine Neurons in Instrumental Reinforcement. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 3895\u0026ndash;3908 (2023).\u003c/li\u003e\n\u003cli\u003eMileykovskiy, B. \u0026amp; Morales, M. Duration of Inhibition of Ventral Tegmental Area Dopamine Neurons Encodes a Level of Conditioned Fear. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 7471\u0026ndash;7476 (2011).\u003c/li\u003e\n\u003cli\u003eMatsumoto, M. \u0026amp; Hikosaka, O. Two types of dopamine neuron distinctly convey positive and negative motivational signals. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e459\u003c/strong\u003e, 837\u0026ndash;841 (2009).\u003c/li\u003e\n\u003cli\u003ePeng, C. Y., Jean-Richard-dit-Bressel, P., Gilchrist, S., Power, J. M. \u0026amp; McNally, G. P. Phasic inhibition of dopamine neurons is an instrumental punisher. \u003cem\u003eBehav. Neurosci.\u003c/em\u003e \u003cstrong\u003e135\u003c/strong\u003e, 415\u0026ndash;425 (2021).\u003c/li\u003e\n\u003cli\u003eDanjo, T., Yoshimi, K., Funabiki, K., Yawata, S. \u0026amp; Nakanishi, S. Aversive behavior induced by optogenetic inactivation of ventral tegmental area dopamine neurons is mediated by dopamine D2 receptors in the nucleus accumbens. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e111\u003c/strong\u003e, 6455\u0026ndash;6460 (2014).\u003c/li\u003e\n\u003cli\u003eTan, K. R. \u003cem\u003eet al.\u003c/em\u003e GABA Neurons of the VTA Drive Conditioned Place Aversion. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e73\u003c/strong\u003e, 1173\u0026ndash;1183 (2012).\u003c/li\u003e\n\u003cli\u003eMorales, M. \u0026amp; Margolis, E. B. Ventral tegmental area: cellular heterogeneity, connectivity and behaviour. \u003cem\u003eNat. Rev. Neurosci.\u003c/em\u003e \u003cstrong\u003e18\u003c/strong\u003e, 73\u0026ndash;85 (2017).\u003c/li\u003e\n\u003cli\u003eEshel, N. \u003cem\u003eet al.\u003c/em\u003e Arithmetic and local circuitry underlying dopamine prediction errors. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e525\u003c/strong\u003e, 243\u0026ndash;246 (2015).\u003c/li\u003e\n\u003cli\u003eLobb, C. J., Wilson, C. J. \u0026amp; Paladini, C. A. A Dynamic Role for GABA Receptors on the Firing Pattern of Midbrain Dopaminergic Neurons. \u003cem\u003eJ. Neurophysiol.\u003c/em\u003e \u003cstrong\u003e104\u003c/strong\u003e, 403\u0026ndash;413 (2010).\u003c/li\u003e\n\u003cli\u003eJi, H. \u0026amp; Shepard, P. D. Lateral Habenula Stimulation Inhibits Rat Midbrain Dopamine Neurons through a GABAA Receptor-Mediated Mechanism. \u003cem\u003eJ. Neurosci.\u003c/em\u003e \u003cstrong\u003e27\u003c/strong\u003e, 6923\u0026ndash;6930 (2007).\u003c/li\u003e\n\u003cli\u003eGomez, J. L. \u003cem\u003eet al.\u003c/em\u003e Chemogenetics revealed: DREADD occupancy and activation via converted clozapine. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e357\u003c/strong\u003e, 503\u0026ndash;507 (2017).\u003c/li\u003e\n\u003cli\u003eJean-Richard-Dit-Bressel, P. \u0026amp; McNally, G. P. The Role of the Lateral Habenula in Punishment. \u003cem\u003ePLOS ONE\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, e111699 (2014).\u003c/li\u003e\n\u003cli\u003eJean-Richard-Dit-Bressel, P. \u0026amp; McNally, G. P. The role of the basolateral amygdala in punishment. \u003cem\u003eLearn. Mem.\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 128\u0026ndash;137 (2015).\u003c/li\u003e\n\u003cli\u003eJean-Richard-dit-Bressel, P., Tran, J., Didachos, A. \u0026amp; McNally, G. P. Instrumental aversion coding in the basolateral amygdala and its reversion by a benzodiazepine. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e47\u003c/strong\u003e, 1199\u0026ndash;1209 (2022).\u003c/li\u003e\n\u003cli\u003eKeevers, L. J. \u0026amp; Jean-Richard-dit-Bressel, P. Obtaining artifact-corrected signals in fiber photometry via isosbestic signals, robust regression, and dF/F calculations. \u003cem\u003eNeurophotonics\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 025003 (2025).\u003c/li\u003e\n\u003cli\u003eJean-Richard-dit-Bressel, P., Clifford, C. W. G. \u0026amp; McNally, G. P. Analyzing Event-Related Transients: Confidence Intervals, Permutation Tests, and Consecutive Thresholds. \u003cem\u003eFront. Mol. Neurosci.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 14 (2020).\u003c/li\u003e\n\u003cli\u003eBrischoux, F., Chakraborty, S., Brierley, D. I. \u0026amp; Ungless, M. A. Phasic excitation of dopamine neurons in ventral VTA by noxious stimuli. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e106\u003c/strong\u003e, 4894\u0026ndash;4899 (2009).\u003c/li\u003e\n\u003cli\u003ede Jong, J. W. \u003cem\u003eet al.\u003c/em\u003e A Neural Circuit Mechanism for Encoding Aversive Stimuli in the Mesolimbic Dopamine System. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e101\u003c/strong\u003e, 133-151.e7 (2019).\u003c/li\u003e\n\u003cli\u003eEshel, N. \u003cem\u003eet al.\u003c/em\u003e Arithmetic and local circuitry underlying dopamine prediction errors. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e525\u003c/strong\u003e, 243\u0026ndash;246 (2015).\u003c/li\u003e\n\u003cli\u003eJacobs, D. S., Allen, M. C., Park, J. \u0026amp; Moghaddam, B. Learning of probabilistic punishment as a model of anxiety produces changes in action but not punisher encoding in the dmPFC and VTA. \u003cem\u003eeLife\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, e78912 (2022).\u003c/li\u003e\n\u003cli\u003eHynes, T. J. \u003cem\u003eet al.\u003c/em\u003e Win-Paired Cues Modulate the Effect of Dopamine Neuron Sensitization on Decision Making and Cocaine Self-administration: Divergent Effects Across Sex. \u003cem\u003eBiol. Psychiatry\u003c/em\u003e \u003cstrong\u003e95\u003c/strong\u003e, 220\u0026ndash;230 (2024).\u003c/li\u003e\n\u003cli\u003eVerharen, J. P. H. \u003cem\u003eet al.\u003c/em\u003e A neuronal mechanism underlying decision-making deficits during hyperdopaminergic states. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 731 (2018).\u003c/li\u003e\n\u003cli\u003eLammel, S., Lim, B. K. \u0026amp; Malenka, R. C. Reward and aversion in a heterogeneous midbrain dopamine system. \u003cem\u003eNeuropharmacology\u003c/em\u003e \u003cstrong\u003e76\u003c/strong\u003e, 351\u0026ndash;359 (2014).\u003c/li\u003e\n\u003cli\u003eLammel, S. \u003cem\u003eet al.\u003c/em\u003e Unique Properties of Mesoprefrontal Neurons within a Dual Mesocorticolimbic Dopamine System. \u003cem\u003eNeuron\u003c/em\u003e \u003cstrong\u003e57\u003c/strong\u003e, 760\u0026ndash;773 (2008).\u003c/li\u003e\n\u003cli\u003eVerharen, J. P. H., Luijendijk, M. C. M., Vanderschuren, L. J. M. J. \u0026amp; Adan, R. A. H. Dopaminergic contributions to behavioral control under threat of punishment in rats. \u003cem\u003ePsychopharmacology (Berl.)\u003c/em\u003e \u003cstrong\u003e237\u003c/strong\u003e, 1769\u0026ndash;1782 (2020).\u003c/li\u003e\n\u003cli\u003eKim, C. K. \u003cem\u003eet al.\u003c/em\u003e Simultaneous fast measurement of circuit dynamics at multiple sites across the mammalian brain. \u003cem\u003eNat. Methods\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 325\u0026ndash;328 (2016).\u003c/li\u003e\n\u003cli\u003ePiantadosi, P. T., Halladay, L. R., Radke, A. K. \u0026amp; Holmes, A. Advances in understanding meso-cortico-limbic-striatal systems mediating risky reward seeking. \u003cem\u003eJ. Neurochem.\u003c/em\u003e \u003cstrong\u003e157\u003c/strong\u003e, 1547\u0026ndash;1571 (2021).\u003c/li\u003e\n\u003cli\u003eFreels, T. G., Gabriel, D. B. K., Lester, D. B. \u0026amp; Simon, N. W. Risky decision-making predicts dopamine release dynamics in nucleus accumbens shell. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e \u003cstrong\u003e45\u003c/strong\u003e, 266\u0026ndash;275 (2020).\u003c/li\u003e\n\u003cli\u003ePearce, J. M. \u0026amp; Dickinson, A. Pavlovian countercondition: Changing the suppressive properties of shock by association with food. \u003cem\u003eJ. Exp. Psychol. Anim. Behav. Process.\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 170\u0026ndash;177 (1975).\u003c/li\u003e\n\u003cli\u003eJean-Richard-dit-Bressel, P., Ma, C., Bradfield, L. A., Killcross, S. \u0026amp; McNally, G. P. Punishment insensitivity emerges from impaired contingency detection, not aversion insensitivity or reward dominance. \u003cem\u003eeLife\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, e52765 (2019).\u003c/li\u003e\n\u003cli\u003eJean-Richard-dit-Bressel, P. \u003cem\u003eet al.\u003c/em\u003e Punishment insensitivity in humans is due to failures in instrumental contingency learning. \u003cem\u003eeLife\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, e69594 (2021).\u003c/li\u003e\n\u003cli\u003eSharpe, M. J. \u003cem\u003eet al.\u003c/em\u003e Dopamine transients do not act as model-free prediction errors during associative learning. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 106 (2020).\u003c/li\u003e\n\u003cli\u003eJeong, H. \u003cem\u003eet al.\u003c/em\u003e Mesolimbic dopamine release conveys causal associations. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e378\u003c/strong\u003e, eabq6740 (2022).\u003c/li\u003e\n\u003cli\u003eAmerican Psychiatric Association. \u003cem\u003eDiagnostic and Statistical Manual of Mental Disorders\u003c/em\u003e. (Washington, DC, 2013).\u003c/li\u003e\n\u003cli\u003ePierce, R. C. \u0026amp; Kumaresan, V. The mesolimbic dopamine system: The final common pathway for the reinforcing effect of drugs of abuse? \u003cem\u003eNeurosci. Biobehav. Rev.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 215\u0026ndash;238 (2006).\u003c/li\u003e\n\u003cli\u003eKalivas, P. W. \u0026amp; Volkow, N. D. The Neural Basis of Addiction: A Pathology of Motivation and Choice. \u003cem\u003eAm. J. Psychiatry\u003c/em\u003e \u003cstrong\u003e162\u003c/strong\u003e, 1403\u0026ndash;1413 (2005).\u003c/li\u003e\n\u003cli\u003eKoob, G. F. \u0026amp; Volkow, N. D. Neurobiology of addiction: a neurocircuitry analysis. \u003cem\u003eLancet Psychiatry\u003c/em\u003e \u003cstrong\u003e3\u003c/strong\u003e, 760\u0026ndash;773 (2016).\u003c/li\u003e\n\u003cli\u003eRamey, T. \u0026amp; Regier, P. S. Cognitive Impairment in Substance Use Disorders. \u003cem\u003eCNS Spectr.\u003c/em\u003e \u003cstrong\u003e24\u003c/strong\u003e, 102\u0026ndash;113 (2019).\u003c/li\u003e\n\u003cli\u003eMcNally, G. P. \u0026amp; Jean-Richard-dit-Bressel, P. A Cognitive Pathway to Persistent, Maladaptive Choice. \u003cem\u003eEur. Addict. Res.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 233\u0026ndash;242 (2024).\u003c/li\u003e\n\u003cli\u003eMcNally, G. P., Jean-Richard-dit-Bressel, P., Millan, E. Z. \u0026amp; Lawrence, A. J. Pathways to the persistence of drug use despite its adverse consequences. \u003cem\u003eMol. Psychiatry\u003c/em\u003e \u003cstrong\u003e28\u003c/strong\u003e, 2228\u0026ndash;2237 (2023).\u003c/li\u003e\n\u003cli\u003eSmith, R. J. \u0026amp; Laiks, L. S. Behavioral and neural mechanisms underlying habitual and compulsive drug seeking. \u003cem\u003eProg. Neuropsychopharmacol. Biol. Psychiatry\u003c/em\u003e \u003cstrong\u003e87\u003c/strong\u003e, 11\u0026ndash;21 (2018).\u003c/li\u003e\n\u003cli\u003eCools, R. Dopaminergic modulation of cognitive function-implications for L-DOPA treatment in Parkinson\u0026rsquo;s disease. \u003cem\u003eNeurosci. Biobehav. Rev.\u003c/em\u003e \u003cstrong\u003e30\u003c/strong\u003e, 1\u0026ndash;23 (2006).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7865029/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7865029/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAvoiding actions with negative consequences is fundamental to adaptive behavior. Traditional theories suggest GABAergic inhibition of midbrain dopamine neurons, including those within ventral tegmental area (VTA\u003csub\u003eDA\u003c/sub\u003e), mediate suppression of actions that lead to aversive outcomes. However, the role of dopamine inhibition in punishment learning remains unclear. To examine this, we conducted fiber photometry, pharmacological, and chemogenetic experiments in rats to measure VTA\u003csub\u003eDA\u003c/sub\u003e activity and GABA input across punishment learning, and test their causal contribution to behavior. VTA\u003csub\u003eDA\u003c/sub\u003e activity and GABA input phasically increased to response-elicited outcomes, with VTA\u003csub\u003eDA\u003c/sub\u003e activity being more strongly activated by rewards, while GABA input being more strongly activated by shock punishers during initial punishment. Pharmacologically blocking GABA\u003csub\u003eA\u003c/sub\u003e receptors in VTA or chemogenetically activating VTA\u003csub\u003eDA\u003c/sub\u003e neurons during initial, but not later, punishment sessions produced enduring deficits in punishment avoidance. These findings suggest long-term avoidance depends upon a critical window of GABA-mediated VTA\u003csub\u003eDA\u003c/sub\u003e inhibition during punishment learning\u003c/p\u003e","manuscriptTitle":"Disinhibition of ventral tegmental area during initial punishment learning causes enduring punishment insensitivity","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-17 06:06:31","doi":"10.21203/rs.3.rs-7865029/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"796e75b4-c5f9-4ee0-b1b2-d1170531d853","owner":[],"postedDate":"October 17th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":56407889,"name":"Biological sciences/Neuroscience/Learning and memory/Operant learning"},{"id":56407890,"name":"Biological sciences/Physiology/Neurophysiology"},{"id":56407891,"name":"Health sciences/Risk factors"}],"tags":[],"updatedAt":"2026-02-17T08:07:47+00:00","versionOfRecord":{"articleIdentity":"rs-7865029","link":"https://doi.org/10.1038/s41386-026-02368-4","journal":{"identity":"neuropsychopharmacology","isVorOnly":false,"title":"Neuropsychopharmacology"},"publishedOn":"2026-02-17 05:00:00","publishedOnDateReadable":"February 17th, 2026"},"versionCreatedAt":"2025-10-17 06:06:31","video":"","vorDoi":"10.1038/s41386-026-02368-4","vorDoiUrl":"https://doi.org/10.1038/s41386-026-02368-4","workflowStages":[]},"version":"v1","identity":"rs-7865029","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7865029","identity":"rs-7865029","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}