Cdk5 modulates cocaine reward, motivation, and striatal neuron excitability

Abstract

Cyclin-dependent kinase 5 (Cdk5) regulates dopamine neurotransmission and has been suggested to serve as a homeostatic target of chronic psychostimulant exposure. To study the role of Cdk5 in the modulation of the cellular and behavioral effects of psychoactive drugs of abuse, we developed Cre/loxP conditional knock-out systems that allow temporal and spatial control of Cdk5 expression in the adult brain. Here, we report the generation of Cdk5 conditional knock-out (cKO) mice using the alphaCaMKII promoter-driven Cre transgenic line (CaMKII-Cre). In this model system, loss of Cdk5 in the adult forebrain increased the psychomotor-activating effects of cocaine. Additionally, these CaMKII-Cre Cdk5 cKO mice show enhanced incentive motivation for food as assessed by instrumental responding on a progressive ratio schedule of reinforcement. Behavioral changes were accompanied by increased excitability of medium spiny neurons in the nucleus accumbens (NAc) in Cdk5 cKO mice. To study NAc-specific effects of Cdk5, another model system was used in which recombinant adeno-associated viruses expressing Cre recombinase caused restricted loss of Cdk5 in NAc neurons. Targeted knock-out of Cdk5 in the NAc facilitated cocaine-induced locomotor sensitization and conditioned place preference for cocaine. These results suggest that Cdk5 acts as a negative regulator of neuronal excitability in the NAc and that Cdk5 may govern the behavioral effects of cocaine and motivation for reinforcement.

Figure 1.

Generation of conditional Cdk5 knock-out mouse model. A, Diagram depicting targeting strategy for generation of conditional Cdk5 allele. The targeted fCdk5 allele contains loxP sites (arrowheads) flanking exons 2–5. B, Utilization of transgenic CaMKII-Cre to mediate excision of Cdk5 in adult brain. Representative in situ hybridization of sagittal brain sections from control and Cdk5 cKO mice using Cdk5 riboprobe. The graph depicts normalized Cdk5 mRNA levels ± SEM in cortex (cor), hippocampus (hip), striatum (str), and cerebellum (cer) in control mice (Con) and Cdk5 cKO mice (***p < 0.001 compared with control, Student's t test, n = 5–7 per group). C, Quantitative immunoblots of Cdk5 and phospho-Thr75 DARPP-32 (pThr75 D32) in striatal homogenates from control and cKO mice. Representative immunoblots (top) and quantitation (bottom) are shown for each (*p < 0.05 compared with control, Student's t test, n = 5–8 per group). D, Immunostaining for EYFP (green) and the neuronal marker, NeuN (red), in CaMKII-Cre/R26R-EYFP reporter mice. Cre activity is shown in EYFP-positive cells (green) in the caudoputamen (CPu), NAc, and cortex at low magnification (left). EYFP/NeuN double-positive cells (yellow) are shown throughout the CPU and NAc at high resolution (right). Scale bars: 500 μm (left, low magnification), 20 μm (right, high magnification).

Figure 2.

Effect of transgenic Cdk5 cKO on cocaine-induced locomotor activity. A, Locomotor responses of control (Con) and CaMKII-Cre Cdk5 cKO mice to saline and cocaine injections in Latin-square dose–response experiment. Mice were habituated to saline injections for 3 consecutive days (S1–S3), followed by cocaine injections in the Latin-square dose–response (gray background). The graph depicts mean locomotor counts ± SEM for 60 min after injection with saline or the drug dose indicated (*p < 0.05 compared with control, post hoc Bonferroni, n = 10 per group). B, Time course of locomotor activity in response to administration of 20 mg/kg cocaine. The graph on the left depicts mean locomotor counts ± SEM over 60 min before (white background) and after (gray background) injection of control or cKO groups. Arrow denotes time of injection in each graph. Cumulative locomotor counts ± SEM over 60 min are displayed in histograms at right (*p < 0.05, Student's t test, n = 10 per group). C, Locomotor activity response to saline injection over the same time course as B with data summarized in histograms (right).

Figure 3.

Effect of Cdk5 cKO on instrumental responding for reward. A, Performance for control (Con) versus cKO mice for food reinforcement under a PR schedule. The graphs depict active (left) and inactive (right) nose poke responses ± SEM for 3 consecutive days of PR testing (*p < 0.05, **p < 0.01, ***p < 0.005 compared with control, post hoc Tukey, n = 16–20 per group). B, Breakpoints achieved for control versus cKO mice during 3 d of PR responding. The graphs show mean final ratios achieved ± SEM (*p < 0.05, **p < 0.01, ***p < 0.001 compared with control, post hoc Tukey, n = 16–20 per group). C, Performance in progressive ratio instrumental responding after prefeeding.

Figure 4.

Effect of Cdk5 cKO on intrinsic membrane excitability of NAc neurons. A, Current–voltage graphs in NAc neurons from control (Con) and cKO mice showing no effect on passive membrane properties. B, Analysis of intrinsic excitability of NAc neurons from control versus cKO mice. The graph shows mean spike number ± SEM in response to different current injections (*p < 0.05, **p < 0.01, ***p < 0.001 compared with control, post hoc Bonferroni, n = 7–8 per group). C, Representative traces of action potential firing evoked by depolarizing current pulses (140–220 pA; 500 ms) in NAc neurons from control and cKO mice. Calibration: 20 mV, 100 ms.

Figure 5.

Generation of NAc-specific Cdk5 knock-out mouse model. A, Diagram depicting recombinant viral vectors to mediate excision of Cdk5 in NAc. Viral constructs contain CMV promoter (CMV), human growth hormone intron (hGH), SV40 polyadenylation sequence (pA), and inverted terminal repeats (open squares). The control vector contains an EGFP cassette, whereas the KO vector expresses a chimeric EGFPCre protein. B, Representative in situ hybridization analysis of rAAV-mediated NAc-specific knock-out of Cdk5. Arrows indicate corresponding NAc regions of rAAV transduction in adjacent sections hybridized with Cdk5 (left) and Cre (right) riboprobes after rAAV-EGFPCre delivery. C, Quantitative immunoblotting of rAAV-mediated NAc-KO. Representative immunoblots of GFP and Cdk5 in homogenates from mice transduced with rAAV expressing EGFP (left) or EGFPCre (right) are shown in the top panels. GFP immunoblots show detection of EGFP (∼28 kDa) in control samples and EGFPCre (∼70 kDa) in NAc-KO samples. Quantitation of Cdk5 levels in NAc lysates transduced with EGFP (Con) versus EGFPCre (NAc-KO) is shown at bottom (*p < 0.05, Student's t test; n = 4 per group).

Figure 6.

Effect of NAc-specific Cdk5 KO on locomotor sensitization to repeated cocaine exposure. A, Locomotor activity in response to saline injection (S), followed by repeated 10 mg/kg cocaine injection over 5 consecutive days (1–5). Mean locomotor counts ± SEM in 10 min bins for 60 min immediately after injection for each day are shown for control (Con) versus NAc-KO groups (left). Data for the first 30 min after injection are summarized in histograms (right, *p < 0.05, **p < 0.01 compared with saline, post hoc Tukey; n = 17 per group). B, Locomotor sensitization to 15 mg/kg cocaine in control (Con) and NAc-KO mice. Time course locomotor activity and data summaries are shown as in A (***p < 0.001 compared with saline, post hoc Tukey; n = 16–19 per group).

Figure 7.

Effect of NAc-specific Cdk5 KO on cocaine-conditioned place preference. A, Time spent in drug-paired compartment for control and NAc-KO mice. Top graph depicts performance of control mice before (gray) and after (white) cocaine training at the indicated cocaine doses. Bottom graph depicts Cdk5 NAc-KO mice performance before (gray) and after (black) cocaine training. Mean time spent in drug compartment ± SEM is indicated for each group at each training dose of cocaine (*p < 0.05, **p < 0.01, ***p < 0.001, paired t test; n = 9–19 per group). B, Analysis of conditioning scores calculated from performance before and after training for each dose indicated. The graph depicts mean conditioning score ± SEM in control (Con, ^#p < 0.05, ^##p < 0.01 compared with saline, post hoc Newman–Keuls; n = 9–19 per group) versus Cdk5 NAc-KO (°p < 0.05, °°p < 0.01 compared with saline, post hoc Newman–Keuls; n = 12–18 per group) mice for the cocaine dose indicated.