The Effect of Cocaine Conditioning on Lateral Preoptic Area Glutamatergic and GABAergic Activity

Cocaine is a commonly abused drug contributing to substance use disorders (SUDs), which affect a large portion of the population and have increased in recent years. Cocaine’s rewarding effects arise primarily through elevated extracellular dopamine (and other monoamines), with acute use disrupting dopamine reuptake and chronic use producing dopamine depletion and reduced responsiveness, thereby lowering hedonic set points and promoting drug-seeking behavior. Cocaine use yields an initial positive affective state followed by a negative affective state (opponent process), which can drive continued use via negative reinforcement. Neuroimaging and circuit models implicate limbic and prefrontal regions (e.g., amygdala, orbitofrontal cortex, anterior cingulate) in craving, reward, and impaired inhibitory control during addiction, suggesting the value of studying region-specific neural activity changes with cocaine.

The lateral preoptic area (LPO), anterior to the hypothalamus, participates in homeostatic and motivational processes (sleep, osmoregulation, locomotion, reward) and is anatomically connected to key reward-related regions, including the ventral tegmental area (VTA) and the lateral habenula (LHb). LPO activity has been shown to change with drug exposure (e.g., increased glucose utilization in morphine dependence; modulation with cocaine self-administration). LPO projections to VTA can reinstate cocaine and sucrose self-administration, and LPO provides robust inputs to LHb, dorsal raphe, rostromedial tegmental nucleus (RMTg), and locus coeruleus; it also receives inputs from nucleus accumbens. Notably, LPO provides both glutamatergic and GABAergic inputs to single LHb neurons; optogenetic activation of LPO→LHb glutamate is aversive, while activation of LPO→LHb GABA is rewarding, suggesting opponent processing roles potentially relevant to cocaine’s affective dynamics.

Research question and goals: This study asks how cocaine conditioning influences LPO glutamatergic and GABAergic neuron activity. Using fiber photometry of LPO cell bodies in VGlut2::Cre and VGat::Cre mice during a conditioned place preference (CPP) paradigm, we examined whether chamber preference changes (cocaine vs saline) relate to changes in calcium activity (AUC) of LPO glutamate and GABA neurons. We hypothesized that cocaine-paired chamber preference would correlate with increased LPO glutamate activity and that LPO GABA activity would be higher in the saline-paired (aversive) chamber.

Subjects: Male and female VGat::Cre (n=15; 7 males, 8 females) and VGlut2::Cre (n=14; 6 males, 8 females) mice on a C57BL/6J background (3–5 months old, 25–38 g) were group-housed (2–5/cage) under a 12 h light/dark cycle with ad libitum food and water. Behavioral testing occurred during the light phase.

Surgeries: Under isoflurane anesthesia (1–5% induction; 1–3% maintenance), mice received perioperative bupivacaine (2.5 mg/kg), carprofen (20 mg/kg), and baytril (5 mg/kg). After exposure of the skull and leveling bregma–lambda within 100 µm, 150 nL of Cre-dependent AAV-DIO-GCaMP7 was injected into LPO (AP 0.15–0.3; ML 0.8–0.9; DV −5.05 to −5.10). A 400 µm core optic fiber (0.48 NA; Doric) was implanted at the same coordinates and secured with three screws and dental acrylic. Postoperative care included carprofen for three days and a ≥2-week recovery for expression.

Conditioned Place Preference (CPP): A three-chamber acrylic apparatus comprised two large side compartments (one with vertical white stripes spaced 3 cm apart; one plain black with thin porous metal mesh floor) and a smaller central connecting chamber. Timeline: habituation (room exposure), pretest (15 min free exploration with fiber photometry and AnyMaze tracking), four consecutive pairing days, and posttest (15 min free exploration with recording). A biased design assigned saline to the initially preferred side and cocaine to the less preferred side based on pretest. Pairing days included a morning saline session and an afternoon cocaine session (15 mg/kg; i.p.) to avoid cocaine carryover into saline sessions. During each 30-min pairing session, mice were confined to the paired chamber using an acrylic divider. Position tracking (AnyMaze) recorded time spent per chamber to compute change scores (ΔS = posttest − pretest).

Fiber photometry: GCaMP7 was excited at 490 nm (calcium-dependent) and 405 nm (isosbestic control) using amplitude-modulated LEDs coupled via a fluorescence minicube (Doric). Emission passed a 500–550 nm filter to a Doric Fluorescence Detector, digitized at 1 kHz, and acquired by TDT RZ5D using Synapse software. Synchronized position signals from AnyMaze (via AMi-2) were recorded for alignment.

Histology: After CPP, mice were perfused (0.1 M phosphate buffer followed by 4% paraformaldehyde), brains post-fixed, cryoprotected in 18% sucrose, frozen, sectioned coronally at 40 µm, and imaged at 20x (Keyence BZX800). Only mice with fluorescent LPO cell bodies localized within 200 µm under the fiber tip (with majority expression in LPO if nearby regions showed expression) were included for analysis.

Photometry analysis: Data were processed using pMAT. The first and last 2 s of recordings were removed to eliminate LED power artifacts. ΔF/F was computed and peri-event time histograms (PETHs) were generated around chamber entries with windows −5–0 s (pre-entry), 0–5 s (early entry), and 5–10 s (late stay). For each subject and chamber (cocaine-, saline-, middle-paired), trace AUCs were computed for the 0–5 s post-entry window as the primary measure of neural activity after chamber entry.

Statistics: CPP change scores (ΔS) were analyzed via one-way repeated-measures ANOVA across chambers (black/striped/middle) with Geisser-Greenhouse correction. Posttest photometry AUCs (0–5 s) were compared between cocaine- and saline-paired entries using paired t-tests. Relationships between CPP change scores (cocaine or saline) and corresponding AUCs were assessed using simple linear regression (Prism). Significance threshold was p<0.05.

• The CPP paradigm produced a robust place preference for the cocaine-paired chamber. Across all mice, there was a significant effect of chamber on change scores ΔS (F(2,56)=48.40, p<0.0001). Mean ΔS (s): cocaine-paired 342.3 ± 27.84; saline-paired −202.5 ± 43.39. A small sex difference in cocaine ΔS was noted (t(12)=2.404, p=0.0333).
• VGat::Cre (GABA) mice: Significant chamber effect on ΔS (F(2,28)=35.50, p<0.0001). Mean ΔS (s): cocaine 288.0 ± 25.82; saline −169.2 ± 48.35; middle −73.05 ± 37.93. No sex difference in cocaine ΔS (t(6)=0.8134, p=0.4471). • Posttest AUC (0–5 s after entry): Cocaine-paired entries showed lower AUC than saline-paired entries (cocaine −21.63 ± 10.15 vs saline 13.75 ± 11.72); paired t-test t(14)=2.520, p=0.0245. • Regression: Cocaine ΔS positively correlated with cocaine-paired AUC (F(1,13)=5.782, p=0.0318; R^2=0.3079), suggesting higher LPO GABA activity predicts stronger cocaine CPP. Saline ΔS did not correlate with saline AUC (F(1,13)=0.7913, p=0.3899; R^2=0.05738).
• VGlut2::Cre (glutamate) mice: Significant chamber effect on ΔS (F(2,26)=21.90, p<0.0001). Mean ΔS (s): cocaine 400.4 ± 46.75; saline −238.1 ± 74.27; middle −155.2 ± 57.66. No sex difference in cocaine ΔS (t(5)=1.916, p=0.1135). • Posttest AUC (0–5 s after entry): No significant difference between cocaine- and saline-paired entries (cocaine 31.18 ± 20.19 vs saline 23.21 ± 20.79); t(13)=0.2406, p=0.8136. • Regression: No significant correlations between ΔS and AUC for either cocaine (F(1,12)=2.583, p=0.1340; R^2=0.1771) or saline (F(1,12)=1.123, p=0.3102; R^2=0.08556).

Cocaine conditioning reliably produced a place preference, indicating successful association of chamber context with cocaine’s positive affective state. LPO GABAergic activity (VGat::Cre) differentiated chamber associations: early post-entry AUC was significantly higher for saline-paired entries than cocaine-paired entries, yet greater GABA activity during cocaine-paired entries positively predicted stronger cocaine CPP across individuals. This suggests that while immediate early-entry GABA signals may be lower upon entry into the cocaine-paired context, overall variability in LPO GABA engagement relates to the magnitude of conditioned cocaine preference, potentially reflecting the role of LPO GABAergic neurons in broader reward circuitry beyond the LPO→LHb pathway.

In contrast, LPO glutamatergic activity (VGlut2::Cre) did not differ significantly between cocaine- and saline-paired entries and did not correlate with CPP strength, implying that bulk LPO glutamate population activity may not directly reflect the aversion/reward coding detectable at specific projection-defined subcircuits (e.g., LPO→LHb vs LPO→VTA). Given prior work showing LPO→LHb glutamate promotes aversion and LPO→LHb GABA promotes reward, the present cell body recordings likely pooled heterogeneous projection populations, potentially obscuring projection-specific effects.

Overall, these findings support a role for LPO GABAergic neurons in encoding aspects of cocaine-conditioned preference and highlight the need for projection-specific dissection of LPO circuits. They also raise questions about timing (e.g., delayed reward-related responses) and state-dependent dynamics (positive vs negative affective phases) that may differentially engage LPO neuron subtypes.

This thesis demonstrates that cocaine conditioning produces a robust place preference and that LPO GABAergic neuron activity relates to the strength of this conditioned preference, whereas bulk LPO glutamatergic activity shows no significant modulation under the same conditions. The results suggest LPO GABA neurons may provide predictive signals for cocaine-seeking behavior within the CPP framework, while projection-specific heterogeneity likely masks distinct roles of LPO glutamate neurons.

Future directions include: (1) projection-specific recording or manipulation (e.g., fiber photometry or optogenetics targeting LPO→LHb vs LPO→VTA) to dissociate reward vs aversion coding; (2) conditioning protocols emphasizing cocaine’s negative affective phase (e.g., delayed or backward conditioning) to probe aversion-related LPO activity; (3) anatomical quantification (e.g., AAV-hSyn-DIO-mCherry tracing, synapse counts) to relate circuit architecture to individual susceptibility in CPP or aversion paradigms; and (4) exploration of pharmacological targets that modulate negative affective states without engaging reward pathways, potentially mitigating relapse risk.

• Fiber photometry at LPO cell bodies aggregates signals from heterogeneous projection-defined subpopulations; projection-specific effects (e.g., LPO→LHb vs LPO→VTA) may be obscured.
• Early-entry analysis window (0–5 s) may miss delayed components of reward-related responses.
• CPP reflects conditioned context preference and may not capture all facets of cocaine’s positive and negative affective states; the negative phase may require alternative conditioning designs (e.g., delayed/backward paradigms).
• Potential variability in viral expression and fiber placement, despite histological verification, could introduce inter-subject variability.
• Sex effects were only briefly assessed; sample sizes for sex-specific analyses were limited.