Hippocampus-sensitive and striatum-sensitive learning one month after morphine or cocaine exposure in male rats

Abstract

These experiments examined whether morphine and cocaine alter the balance between hippocampal and striatal memory systems measured long after drug exposure. Male rats received injections of morphine (5 mg/kg), cocaine (20 mg/kg), or saline for five consecutive days. One month later, rats were trained to find food on a hippocampus-sensitive place task or a striatum-sensitive response task. Relative to saline controls, morphine-treated rats exhibited impaired place learning but enhanced response learning; prior cocaine exposure did not significantly alter learning on either task. Another set of rats was trained on a dual-solution T-maze that can be solved with either place or response strategies. While a majority (67%) of control rats used place solutions, morphine treatment one month prior resulted in the exclusive use of response solutions (100%). Prior cocaine treatment did not significantly alter strategy selection. Molecular markers related to learning and drug abuse were measured in the hippocampus and striatum one month after drug exposure in behaviorally untested rats. Protein levels of glial-fibrillary acidic protein (GFAP), an intermediate filament specific to astrocytes, increased significantly in the hippocampus after morphine exposure, but not after cocaine exposure. Exposure to morphine or cocaine did not significantly change levels of brain-derived neurotrophic factor (BDNF) or a downstream target of BDNF signaling, glycogen synthase kinase 3β (GSK3β), in the hippocampus or striatum. Thus, exposure to morphine resulted in a long-lasting shift from hippocampal toward striatal dominance during learning, an effect that may be associated with lasting alterations in hippocampal astrocytes. Cocaine produced changes in the same direction, suggesting that use of a higher dose or longer duration of exposure might produce effects comparable to those seen with morphine.

Keywords: Astrocytes; BDNF; Decision-making; GFAP; GSK3β; Learning.

Fig. 1.

Experimental design. Rats were handled for five days prior to receiving once daily i.p. injections (1 mg / kg) of 0.9% saline, cocaine (20 mg / kg), or morphine (5 / mg / kg) for five days. After food restriction, rats were trained on one of three maze tasks one month after the last injection. A separate group of untrained rats was euthanized one month after the last drug injection and hippocampal and striatal tissue were collected and subsequently assayed for levels of glial fibrillary acidic protein, brain-derived neurotrophic factor, and glycogen synthase kinase 3β.

Fig. 2.

Training tasks. (A) Place learning involved finding the food reward in a specific room position. (B) Response learning used a fixed body turn to find the food reward. (C) A dual-solution task could be solved using either place or response strategies; after learning (9/10 correct trials), the rat was placed in the opposite start position to that used during training on each of three probe trials. The arm choice made on two of three or three of three probe trials identified the dominant learning strategy: place or response. Arrows indicate the rewarded arm(s).

Fig. 3.

Cocaine and morphine effects on place and response learning tested one month after drug exposure. Changes in trial accuracy throughout the training sessions are shown for (A) place and (B) response tasks. Accuracy collapsed across all trials (100 for place and 75 for response) is shown for (C) place and (D) response tasks. Dashed lines indicate chance levels (50%). The number of trials to criterion (TTC; 9/10 correct trials with at least 6 consecutively correct) is shown for (E) place and (F) response tasks. Note that prior morphine exposure resulted in reduced accuracy and a trend toward increased trials to criterion on the place task. In marked contrast, morphine resulted in increased accuracy and decreased trials to criterion on the response task. Prior cocaine exposure did not significantly affect place or response learning measures. * p < 0.05 vs. saline. Place: Saline n = 9, Cocaine n = 10, Morphine n = 10; Response: Saline n = 10, Cocaine n = 10, Morphine n = 10.

Fig. 4.

Cocaine and morphine effects on dual-solution strategies tested one month after drug exposure. (A) The percent of rats expressing place vs. response strategies on probe trials shifted from 67% place in the saline group to 100% response after prior exposure to morphine. The strategies expressed by rats after cocaine showed a modest shift in expression from place to response strategies that did not differ significantly from the strategies used by the saline group. (B) The number of trials to reach the 9/10 learning criterion (TTC) was comparable across conditions. * p < 0.05 morphine vs. saline. Saline n = 9, Cocaine n = 8, Morphine n = 9.

Fig. 5.

Glial fibrillary acidic protein (GFAP) levels (ng/mg wet weight) in the hippocampus and striatum one month after drug exposure as measured with ELISAs. Note that GFAP levels increased in the hippocampus but not in the striatum after morphine treatment. GFAP levels did not change significantly after cocaine treatment. * Hippocampus-saline vs hippocampus-morphine, p < 0.05. Overall, GFAP levels in the hippocampus were significantly higher than those in striatum (p < 0.0001). Hippocampus: Saline n = 9, Cocaine n = 9, Morphine n = 8; Striatum: Saline n = 9, Cocaine n = 9, Morphine n = 9.

Fig. 6.

Levels of brain-derived neurotrophic factor (BDNF) (pg/mg wet weight) in the hippocampus and striatum one month after drug exposure as measured with ELISAs. Levels of (A) mature BDNF and (B) pro BDNF did not differ significantly by drug treatment in either brain area. Both mature and pro BDNF content levels were significantly higher in the hippocampus than the striatum (p’s < 0.0001). Hippocampus: Saline n = 7, Cocaine n = 8, Morphine n = 6; Striatum: Saline n = 6, Cocaine n = 8, Morphine n = 7.

Fig. 7.

Levels of total and phosphorylated glycogen synthase kinase 3β (GSK3β) in the hippocampus and striatum one month after drug exposure as measured with western blots. (A) Mean levels of total GSK3β, phosphorylated (ser9) GSK3β, and the ratio of pGSK3β to total GSK3β in the (A) hippocampus and (B) striatum one month after drug treatments were not different across conditions. Representative bands are shown for each target. Hippocampus: Saline (S) n = 9, Cocaine (C) n = 9, Morphine (M) n = 9; Striatum: Saline (S) n = 9, Cocaine (C) n = 8, Morphine (M) n = 9.