Adolescent THC impacts on mPFC dopamine-mediated cognitive processes in male and female rats

Abstract

Rationale: Adolescent cannabis use is linked to later-life changes in cognition, learning, and memory. Rodent experimental studies suggest Δ⁹-tetrahydrocannabinol (THC) influences development of circuits underlying these processes, especially in the prefrontal cortex, which matures during adolescence.

Objective: We determined how 14 daily THC injections (5mg/kg) during adolescence persistently impacts medial prefrontal cortex (mPFC) dopamine-dependent cognition.

Methods: In adult Long Evans rats treated as adolescents with THC (AdoTHC), we quantify performance on two mPFC dopamine-dependent reward-based tasks-strategy set shifting and probabilistic discounting. We also determined how acute dopamine augmentation with amphetamine (0, 0.25, 0.5 mg/kg), or specific chemogenetic stimulation of ventral tegmental area (VTA) dopamine neurons and their projections to mPFC impacts probabilistic discounting.

Results: AdoTHC sex-dependently impacts acquisition of cue-guided instrumental reward seeking, but has minimal effects on set-shifting or probabilistic discounting in either sex. When we challenged dopamine circuits acutely with amphetamine during probabilistic discounting, we found reduced discounting of improbable reward options, with AdoTHC rats being more sensitive to these effects than controls. In contrast, neither acute chemogenetic stimulation of VTA dopamine neurons nor pathway-specific chemogenetic stimulation of their projection to mPFC impacted probabilistic discounting in control rats, although stimulation of this cortical dopamine projection slightly disrupted choices in AdoTHC rats.

Conclusions: These studies confirm a marked specificity in the cognitive processes impacted by AdoTHC exposure. They also suggest that some persistent AdoTHC effects may alter amphetamine-induced cognitive changes in a manner independent of VTA dopamine neurons or their projections to mPFC.

Keywords: THC; chemogenetics; cognition; dopamine; medial prefrontal cortex; ventral tegmental area.

Fig. 1. Adolescent THC history selectively impacts adulthood learning and cognition

a) Experimental timeline. b) Schematic of the visual cue discrimination task. c) AdoTHC rats (green wedges) were more likely than AdoVEH rats (black wedges) to acquire the visual cue discrimination task to criterion in only one training session (unfilled wedges), rather than requiring 2 sessions to acquire (filled wedges). Both sexes showed similar patterns. d, e) AdoTHC females took fewer trials to meet criterion, and made fewer errors when learning visual cue discrimination than AdoVeh females, no such effects were seen in males. f, g) AdoTHC females took longer to respond, and made fewer errors than AdoVEH females during cue discrimination training, without effects in males. h) Schematic of the subsequently tested strategy set-shifting task. i) AdoTHC did not alter the number of trials to learn the new rule to criterion in either sex. Likewise, AdoTHC did not alter in either sex j) the number of errors, k) the types of errors, l) latency to respond, or m) omitted trials during set shifting training. Individual rats shown as grey dots in each graph: AdoVEH (n = 14 males, 11 females), AdoTHC (n = 13 males, 10 females). χ² p* < 0.05 and repeated measure two-way ANOVA, Sidak post hoc: p* < 0.05. Data presented as mean + SEM.

Fig. 2. AdoTHC history does not affect baseline probabilistic discounting

a) Schematic of the probabilistic discounting task. Sessions consisted of 5 training blocks, in which 2 levers are presented, and rats must choose either a small reward lever that always delivers one pellet, or a large reward-delivering lever which becomes increasingly unlikely to deliver any reward as the session progresses. b) Data is shown for males and females, indicating that AdoTHC rats did not differ from their AdoVEH counterparts in probabilistic discounting, with both shifting from nearly exclusively preferring the large reward lever, but appropriately shifting away from it as it became less likely to deliver reward. AdoVEH (n = 14 males, 11 females), AdoTHC (n = 15 males, 8 females). Data represented as average within each probability block from three consecutive days of stable performance. Mean ± SEM.

Fig. 3. Amphetamine-induced ‘risky’ responding is potentiated after adolescent THC.

a) Data from saline (black line/bar), low dose (0.25mg/kg; pink line/bar) and high dose (0.5mg/kg; red line/bar) amphetamine tests in each sex are shown. When performance patterns were further interrogated, we found that high dose of AMPH in Males increased inflexibility across both AdoTX groups, however in b) females, the high dose of AMPH in AdoVEH rats reduced risky responding at high probability blocks not seen in AdoTHC females. c) AMPH did not alter win-stay,but reduced lose-shift in AdoTX males. d) In AdoVEH females, AMPH decreased win-stay and increased lose-shift, while in AdoTHC AMPH had no effect on win-stay but decreased lose-shift. AMPH = amphetamine. AdoVEH (n = 14 males, 10 females), AdoTHC (n = 15 males, 6 females). Repeated measure three-way ANOVA; Sidak post hoc: p* < 0.05, p** < 0.01. Data represented as mean + SEM, individual animals shown as grey dots.

Fig. 4. Stimulation of VTA dopamine neurons, or VTA dopamine projections to mPFC does not affect probabilistic discounting

a) Bilateral Cre-dependent hM3Dq DREADD AAV injections were made in VTA of TH:Cre rats, and of wildtype (WT) littermates. b) Example hM3Dq DREADD expression (red) is localized to tyrosine hydroxylase+ (TH; green) neurons within VTA (yellow=merge). Scale bar, 300 μm. c) For pathway-specific stimulation of VTA dopamine projections to mPFC, Cre-dependent hM3Dq DREADDs were injected into VTA as in Experiment 1, and cannulae targeting mPFC allowed CNO microinjection (1mM, 0.5 μl) upon DREADD-expressing dopamine neuron axons in this pathway. d) Cannula placements of each rat in pathway stimulation Experiment 2 is shown. e) Neither VTA dopamine neuron stimulation induced by systemic CNO in TH:Cre rats, nor f) stimulation of the VTA dopamine projection to mPFC induced by mPFC CNO microinjections in TH:Cre rats robustly altered probabilistic discounting in AdoTHC or AdoVEH rats. Likewise, lower panels of e,f) indicate that CNO did not have robust effects on WT rats without DREADDs. Data is represented as average % choice of the large reward lever in each probability block across the two CNO, and two VEH tests conducted in each rat, mean±SEM, p* < 0.05. TH:Cre+: VTA dopamine stimulation; AdoVEH (n = 8M, 3F) AdoTHC (n = 8M, 4F); VTA dopamine to mPFC: AdoVEH (n = 10M, 4F) AdoTHC (n = 9M, 5F). Wildtype: VTA dopamine stimulation; AdoVEH (n = 5M, 3F) AdoTHC (n = 5M); VTA dopamine to mPFC: AdoVEH (n = 9M, 4F) AdoTHC (n = 8M).