Inhibition of alpha7 nicotinic receptors in the ventral hippocampus selectively attenuates reinstatement of morphine-conditioned place preference and associated changes in AMPA receptor binding

Abstract

Recurrent relapse is a major problem in treating opiate addiction. Pavlovian conditioning plays a role in recurrent relapse whereby exposure to cues learned during drug intake can precipitate relapse to drug taking. α7 nicotinic acetylcholine receptors (nAChRs) have been implicated in attentional aspects of cognition and mechanisms of learning and memory. In this study we have investigated the role of α7 nAChRs in morphine-conditioned place preference (morphine-CPP). CPP provides a model of associative learning that is pertinent to associative aspects of drug dependence. The α7 nAChR antagonist methyllycaconitine (MLA; 4 mg/kg s.c.) had no effect on the acquisition, maintenance, reconsolidation or extinction of morphine-CPP but selectively attenuated morphine-primed reinstatement of CPP, in both mice and rats. Reinstatement of morphine-CPP in mice was accompanied by a selective increase in [³ H]-AMPA binding (but not in [³ H]-MK801 binding) in the ventral hippocampus that was prevented by prior treatment with MLA. Administration of MLA (6.7 μg) directly into the ventral hippocampus of rats prior to a systemic priming dose of morphine abolished reinstatement of morphine-CPP, whereas MLA delivered into the dorsal hippocampus or prefrontal cortex was without effect. These results suggest that α7 nAChRs in the ventral hippocampus play a specific role in the retrieval of associative drug memories following a period of extinction, making them potential targets for the prevention of relapse.

Keywords: [3H]-AMPA; [3H]-MK801; autoradiography; intracerebral drug delivery; methyllycaconitine.

Figure 1

Effect of MLA on stages of morphine‐CPP in C57BL/6 mice. (a) Acquisition. Mice were tested for any innate preference for a particular chamber (Habituation) and then pseudo‐randomly allocated to two groups with comparable mean preference scores. The MLA group received MLA (4 mg/kg) 20 minutes prior to a conditioning dose of morphine (10 mg/kg) or saline, paired with alternative compartments, on 4 consecutive days. The control group received saline instead of MLA. A preference test was conducted on the next day, giving the animals free access to both compartments of the CPP apparatus for 15 minutes. Preference scores indicate the time spent in the morphine‐paired compartment in seconds minus 450 (half of the total time). Data are expressed as the mean ± S.E.M. Both groups showed significant acquisition of morphine‐CPP (*saline: P < 0.05, **MLA: P < 0.005, n = 16 per group) indicating that MLA treatment had no effect on the acquisition of morphine‐CPP. (b) Maintenance. Mice were allocated to two groups and both groups were treated identically, to acquire morphine‐CPP as in (a). They were tested for expression of morphine‐CPP on 5 consecutive days and 1 week later. One group received MLA (4 mg/kg) 20 minutes prior to each preference test, the control group received saline instead of MLA. Both groups acquired morphine‐CPP (**P < 0.005, n = 12 per group). MLA treatment had no effect on the expression of morphine‐CPP during the maintenance phase. (c) Schematic for reconsolidation and reinstatement experiments. (d) Reconsolidation. Two groups of mice were treated identically, to acquire morphine‐CPP as in (a). Three days later mice received a further dose of morphine (10 mg/kg) in the drug‐paired compartment. Immediately afterwards, one group received MLA (4 mg/kg) while the control group received saline. The mice underwent a preference test the next day and one week later. Both groups acquired morphine‐CPP (**P < 0.005, n = 12 per group). Their preference for the drug‐paired compartment was consolidated by morphine and unaffected by MLA. E. Reinstatement. Two groups of mice were treated identically, to acquire morphine‐CPP as in (a), followed by 4 days of extinction, (saline injections only, paired with alternative compartments) on separate days. Both groups acquired morphine‐CPP (**P < 0.005, n = 20 per group) and this was attenuated following extinction training. On the following day, mice received a priming dose of morphine (5 mg/kg) prior to an extended preference test (30 minutes); one group received MLA (4 mg/kg) 20 minutes before morphine; the control group received saline instead of MLA. Only the control group of mice showed significant reinstatement of morphine‐CPP in contrast to MLA‐treated mice (saline: **P < 0.005). The time spent in the morphine‐paired compartment was significantly different between MLA and saline treatments (^† P < 0.005). MLA significantly inhibited reinstatement of morphine‐CPP

Figure 2

Effect of morphine‐CPP reinstatement, in the presence and absence of MLA, on NMDA receptor binding in mouse brain. Animals underwent morphine‐CPP acquisition and extinction (Fig. 1c). A parallel set of animals followed the same procedure but received saline instead of morphine. Each group was then randomly divided into two sets that received either saline or MLA (4 mg/kg) 20 minutes before saline or a priming dose of morphine (5 mg/kg) and were then tested for reinstatement of morphine‐CPP. Immediately afterwards, mice were killed and brains were prepared for autoradiography as described in the Methods. Sections were labelled with 70 nM [³H]‐(+)‐MK801, in the absence and presence of 1 μM (+)‐MK801 to determine non‐specific binding. (a) Representative computer‐enhanced autoradiograms of [³H]‐(+)‐MK801 binding in coronal brain sections of mice taken at four levels corresponding to prefrontal cortex (bregma 1.94 mm), striatum (bregma 1.42 mm), dorsal hippocampus (bregma −1.22 mm) and the midbrain/ventral hippocampus (bregma −3.08 mm; Paxinos & Frankilin 2012). Binding levels are represented using a pseudocolour interpretation of black and white film images in fmol/mg tissue equivalent. Representative autoradiograms for non‐specific binding (NSB) are shown (far right column). (b) Quantitative NMDA receptor binding. Levels of [³H]‐(+)‐MK801 binding (fmol/mg tissue equivalent) are shown for prelimbic cortex (PrL), infralimbic cortex (IL), CA1 and CA2 regions of the ventral hippocampus combined (vCA1 + CA2), ventral hippocampus CA3 (vCA3), dorsal hippocampus CA1, CA2 and CA3 separately (dCA1, dCA2 and dCA3), for each of the treatment groups (saline + saline control, saline + morphine‐primed reinstatement, MLA + saline control, MLA + morphine‐primed reinstatement). Data are presented as mean ± S.E.M. (n = 4–7). There were no statistically significant differences between any treatment groups in any of the brain regions quantified

Figure 3

Effect of morphine‐CPP reinstatement, in the presence and absence of MLA, on AMPA receptor binding in mouse brain. Animals underwent morphine‐CPP acquisition and reinstatement, and their brains were prepared for autoradiography as described in the legend to Figure 2. Sections were labelled with 10 nM [³H]‐AMPA, in the absence and presence of 0.1 μM CNQX to determine non‐specific binding. (a) Representative computer‐enhanced autoradiograms of [³H]‐AMPA binding in coronal brain sections of mice taken at four levels corresponding to prefrontal cortex (bregma 1.94 mm), striatum (bregma 1.42 mm), dorsal hippocampus (bregma −1.22 mm) and the midbrain/ventral hippocampus (bregma −3.08 mm; Paxinos & Franklin 2012). Binding levels are represented using a pseudocolour interpretation of black and white film images in fmol/mg tissue equivalent. Representative autoradiograms for non‐specific binding (NSB) are shown (far right column). (b) Quantitative AMPA receptor binding. Levels of [³H]‐AMPA binding (fmol/mg tissue equivalent) are shown for prelimbic cortex (PrL), infralimbic cortex (IL), CA1 and CA2 regions of the ventral hippocampus combined (vCA1 + CA2), ventral hippocampus CA3 (vCA3), dorsal hippocampus CA1, CA2 and CA3 separately (dCA1, dCA2, dCA3), for each of the treatment groups (saline + saline control, saline + morphine‐primed reinstatement, MLA + saline control, MLA + morphine‐primed reinstatement). Data are presented as mean ± S.E.M. (n = 4–6). *Significantly different from saline control and from MLA‐pretreated, morphine‐primed condition, P < 0.05 (two‐way ANOVA)

Figure 4

Morphine‐CPP in Wistar rats and effect of MLA on reinstatement. (a) Rats were divided into two groups, both groups underwent acquisition and extinction of morphine‐CPP (5 mg/kg morphine) as described win the Methods. Following extinction, one group received a priming dose of morphine (2.5 mg/kg) while the other group received saline (arrow). They were tested for reinstatement of morphine‐CPP. Data are presented as mean ± S.E.M. preference scores (n = 26). Morphine‐primed group was significantly different from saline controls, **P < 0.01 (one‐way ANOVA and Benjamini–Hochberg test for multiple comparisons). (b) Rats were divided into two groups, both groups underwent acquisition and extinction of morphine‐CPP as aforementioned. Following extinction, both groups received a priming dose of morphine (2.5 mg/kg), but one group received MLA (4 mg/kg) prior to the priming dose while the other group received saline (arrow). They were tested for reinstatement of morphine‐CPP. Data are presented as mean ± S.E.M. preference scores (n = 26). Saline‐pretreated morphine‐primed group was significantly different from morphine‐primed group that received MLA, *P < 0.05 (one‐way ANOVA and Benjamini–Hochberg test for multiple comparisons)

Figure 5

Effect of intracerebral administration of MLA on reinstatement of morphine‐CPP in rats. Rats acquired morphine‐CPP that was extinguished prior to the implantation of bilateral intracerebral cannulae in the dorsal hippocampus (a), ventral hippocampus (b) or prefrontal cortex (c). Animals then received either saline or MLA (6.75 μg/hemisphere) prior to a priming dose of morphine (2.5 mg/kg) (arrows) and testing for reinstatement of morphine‐CPP. Data are presented as mean ± S.E.M. preference scores (n = 8). In ventral hippocampus, saline pretreated morphine‐primed group was significantly different from morphine‐primed group that received MLA, *P < 0.05 (one‐way ANOVA and Benjamini–Hochberg test for multiple comparisons)