Ethanol-induced conditioned place preference and aversion differentially alter plasticity in the bed nucleus of stria terminalis

Abstract

Contextual cues associated with drugs of abuse, such as ethanol, can trigger craving and drug-seeking behavior. Pavlovian procedures, such as place conditioning, have been widely used to study the rewarding/aversive properties of drugs and the association between environmental cues and drug seeking. Previous research has shown that ethanol as an unconditioned stimulus can induce a strong conditioned place preference (CPP) or aversion (CPA) in rodents. However, the neural mechanisms underlying ethanol-induced reward and aversion have not been thoroughly investigated. The bed nucleus of the stria terminalis (BNST), an integral part of the extended amygdala, is engaged by both rewarding and aversive stimuli and plays a role in ethanol-seeking behavior. Here, we used ex-vivo slice physiology to probe learning-induced synaptic plasticity in the BNST following ethanol-induced CPP and CPA. Male DBA/2 J mice (2-3 months old) were conditioned using previously reported ethanol-induced CPP/CPA procedures. Ethanol-induced CPP was associated with increased neuronal excitability in the ventral BNST (vBNST). Conversely, ethanol-induced CPA resulted in a significant decrease in spontaneous glutamatergic transmission without alterations in GABAergic signaling. Ethanol-CPA also led to a significant increase in the paired-pulse ratio at excitatory synapses, suggestive of a decrease in presynaptic glutamate release. Collectively, these data demonstrate that the vBNST is involved in the modulation of contextual learning associated with both the rewarding and the aversive properties of ethanol in mice.

Fig. 1

Ethanol-induced Pavlovian learning in male DBA2/J mice. a Experimental timeline for ethanol-induced CPP paradigm. b, c No difference in preference for CS + side was observed between mice assigned to the saline group (n = 24) and mice assigned to the ethanol group (n = 24). There was also no difference in mean velocity between the two groups. d, e Following four days of conditioning with two 5 min trials per day, mice injected with ethanol showed a significant preference for the CS + side on a drug-free test day when compared to saline-treated mice. Post conditioning, mice in the ethanol-paired group showed a significant reduction in mean velocity compared to saline-paired mice. f Experimental timeline for ethanol-induced CPA paradigm. g, h) Average pretest data showing mice assigned to the saline group (n = 14) did not differ in their preference for CS + side when compared to ethanol group (n = 16). No difference in mean velocity between the two groups was observed. i, j Mice injected with ethanol immediately after 5 min exposure to a distinct cue (grid or hole floor) expressed a significant aversion for CS + side on Test day and also showed a significant reduction in mean velocity compared to saline-paired mice following ethanol-induced aversion. Data expressed as Mean ± SEM.*p < 0.05

Fig. 2

Ethanol-induced CPP increases excitability in vBNST neurons via inhibition of inward rectifying potassium channel. a Representative data obtained from vBNST neurons in Sal and EtOH mice in response to a 120 pA/s current ramp while injecting a constant current to hold the cell at −70 mV. The minimum current required to elicit an action potential (rheobase) was reduced in EtOH mice (n = 24 cells from 9 mice) compared to Sal (c; n = 24 cells from 7 mice) without any changes (b) in resting membrane potential (RMP). d There was a negative correlation between average rheobase for each animal and percentage time spent on CS + side. e We also observed a significant decrease in action potential threshold. f, g Representative traces of a neuron in Sal and EtOH group, respectively firing action potentials in response to a step protocol of increased current steps . There was a significant increase in the number of spikes in response to a graded current injection (10 pA/250 ms) in the EtOH group. h Current-voltage relationship of vBNST neurons indicates a reduction in an inwardly rectifying current in the EtOH-treated mice. i-l) Ethanol-induced CPP also altered the action potential kinetics resulting in decreased latency (i) and a trend toward decreased action potential half-width (k) in EtOH mice. There was no difference between EtOH and Sal mice groups in average action potential height (j) and fast after-hyperpolarization potential (l). In a different cohort of mice, excitability experiments (m–w) were conducted in the presence of Tertiapin-Q (TPQ, 200 nM) to block inward-rectifying potassium channels. No change in RMP (n) was observed between the two groups in the presence of TPQ. Representative data (m) obtained from vBNST neurons showing no change in rheobase between saline-treated (n = 21 cells from 10 mice) and ethanol-treated (n = 13 cells from 5 mice) groups (o). p There was a trend toward a decrease in action potential threshold in the EtOH group compared to the Sal group. q, r There was no significant difference in the number of spikes in response to a graded current injection between the two groups in the presence of TPQ. s There was no difference in the current-voltage relationship of vBNST neurons between the two groups. t–w Ethanol-induced CPP decreased AP latency (t) but not average action potential height (u), action potential half-width (v) or fast after-hyperpolarization potential (w) in the presence of TPQ. Scale bar (10 mV, 250 ms), lines in dot plots represent mean ± SEM. *p < 0.05

Fig. 3

No changes in excitability in vBNST neurons following either ethanol-induced CPA or unpaired-ethanol injections. EtOH mice (n = 7 mice) did not have a significant difference in RMP (a), rheobase (b), action potential threshold (c) or in the number of spikes in response to current steps (d) compared to Sal mice (n = 8 mice) following ethanol-induced CPA. Similarly, mice that received ethanol injections in their home cages (n = 4 mice) did not vary significantly across various excitability measures (e–h) from mice that received home cage saline injections (n = 4 mice). Lines in dot plots represent mean ± SEM

Fig. 4

Ethanol-induced CPA decreases spontaneous excitatory synaptic transmission in vBNST neurons. a Representative traces of spontaneous postsynaptic currents (sPSCs) from Sal-CPP (n = 6 mice; 19 cells) and EtOH-CPP mice (n = 6; 24 cells). There were no differences in sEPSC or sIPSC parameters (frequency, amplitude, and excitation-inhibition ratio) of vBNST neurons b–d between the two groups. e Representative data from Sal-CPA (n = 5) and EtOH-CPA mice (n = 5) showing a reduced excitatory drive on vBNST neurons following ethanol-induced aversion. There was a significant decrease in sEPSC frequency in the EtOH mice (f; n = 14 cells) without any alterations in sIPSC frequency when compared to Sal mice (n = 14 cells). No significant differences were observed in sEPSC or sIPSC amplitude (g). The excitation-inhibition ratio of vBNST neurons in EtOH mice was also significantly lower than Sal mice (h). i Representative traces of sPSCs from Sal-Home (n = 6; 15 cells) and EtOH-Home mice (n = 6; 12 cells) indicating an overall increase in both GABAergic and glutamatergic transmission following exposure to ethanol. There was a significant increase in both sEPSC and sIPSC frequency (j) without any alterations in amplitude (k) when compared to Sal mice. The excitation-inhibition ratio of vBNST neurons did not vary significantly between Sal mice and EtOH mice (l). m–p There were no differences in miniature postsynaptic currents (mPSCs) following either ethanol-induced CPP (n = 4 mice in each group) or CPA (n = 2 mice in sal group; n = 4 mice EtOH group). mPSCs were isolated by adding TTX (500 nM) to block action-potential mediated synaptic transmission. Scale bar (20 pA, 500 ms). Lines in dot plots represent mean ± SEM. *p < 0.05

Fig. 5

Ethanol-induced CPA selectively decreases presynaptic release probability at glutamatergic terminals without altering AMPA/NMDA ratio. a Representative traces of paired-pulse ratios (PPR with 50 ms inter-stimulus interval; amplitude of pulse 2/ amplitude of pulse 1) at glutamatergic synapses following ethanol-induced CPP. All traces are normalized to the first pulse for each graph. Scale bar (50 ms). No effect of ethanol-induced CPP on glutamatergic PPR was observed in the two groups (Sal: n = 21 cells from 6 mice; EtOH: n = 15 cells from 5 mice). b Representative traces of glutamatergic PPR following ethanol-induced CPA. There was a significant increase in PPR in the EtOH mice (n = 17 cells from 7 mice) when compared to Sal mice (n = 22 cells from 6 mice), indicative of a potential decrease in presynaptic release probability. c No changes were observed in glutamatergic PPR following home cage exposure to ethanol (n = 15 cells from 4 mice in each group). d–f Representative traces of paired-pulse ratios (PPR with 50 ms inter-stimulus interval; delta amplitude of pulse 2/ amplitude of pulse 1) at GABAergic synapses. Traces are normalized to the first pulse for each graph. Scale bar (100 ms). No changes were observed in PPR at GABAergic terminals following exposure to ethanol when compared to their respective saline controls. (n = 11 cells from 3 mice in Sal group and n = 12 cells from 3 mice for EtOH-CPP experiments; n = 10 cells from 3 mice in Sal-CPA and n = 12 cells from 4 mice in EtOH-CPA group; n = 7 cells from 2 mice in each group for unpaired ethanol exposure). g–i Representative traces of AMPA/NMDA ratio (AMPA current measured at −70 mV; NMDA current measured at + 40 mV after a delay of 50 ms) following exposure to ethanol-induced CPP. All traces are normalized to NMDA for each graph. Scale bar (100 ms). No effect of ethanol-induced CPP on the AMPA/NMDA ratio was observed between the two groups (g; n = 15 cells from 4 Sal mice; n = 14 cells from 5 EtOH mice). h Ethanol-induced CPA did not affect the AMPA/NMDA ratios between the Sal mice and EtOH mice (n = 17 cells from 6 Sal mice; n = 22 cells from 7 EtOH mice). i Similarly, home cage exposure to ethanol (n = 15 cells from 4 mice) did not significantly alter AMPA/NMDA ratios when compared to the saline control mice (n = 15 cells from 4 mice). j) Model depiction of ethanol-induced neuroadaptations in the vBNST following Pavlovian learning. Ethanol-CPP increases neuronal excitability by blocking inward-rectifying potassium channels on putative VTA-projection neurons which can be rewarding, while ethanol-CPA reduces the excitatory drive on vBNST neurons via modulation of presynaptic release probability at glutamatergic synapses which can potentially drive aversion. Lines in dot plots represent mean ± SEM. *p < 0.05