Cocaine regulates MEF2 to control synaptic and behavioral plasticity

Abstract

Repeated exposure to cocaine causes sensitized behavioral responses and increased dendritic spines on medium spiny neurons of the nucleus accumbens (NAc). We find that cocaine regulates myocyte enhancer factor 2 (MEF2) transcription factors to control these two processes in vivo. Cocaine suppresses striatal MEF2 activity in part through a mechanism involving cAMP, the regulator of calmodulin signaling (RCS), and calcineurin. We show that reducing MEF2 activity in the NAc in vivo is required for the cocaine-induced increases in dendritic spine density. Surprisingly, we find that increasing MEF2 activity in the NAc, which blocks the cocaine-induced increase in dendritic spine density, enhances sensitized behavioral responses to cocaine. Together, our findings implicate MEF2 as a key regulator of structural synapse plasticity and sensitized responses to cocaine and suggest that reducing MEF2 activity (and increasing spine density) in NAc may be a compensatory mechanism to limit long-lasting maladaptive behavioral responses to cocaine.

Figure 1. MEF2A and MEF2D are highly expressed in the adult striatum

(A) Immunohistochemistry for MEF2A and MEF2D demonstrate strong nuclear staining throughout the adult striatum. MEF2A and MEF2D co-localize in the nucleus of most striatal neurons. (B) Electrophoretic mobility shift assays (EMSA) were performed on 10 µg of NAc lysate and ³²P-labeled MRE duplex oligos. The shifted MRE band (open arrow) was competed away with excess, unlabelled MRE. Pre-incubation with anti-MEF2A or anti-MEF2D antibodies results in supershifted MRE bands (SSBs, closed arrows) that migrate more slowly in the native gel. (C) Cultured striatal neurons transfected with an MRE-luciferase reporter plasmid were co-transfected with either plasmids expressing MEF2A and MEF2D specific shRNAs or vector alone. Reduction of MEF2A and MEF2D significantly reduces both basal and membrane depolarization (60 mM KCl)-induced MEF2 activity (***p<0.001, n=9, three independent experiments).

Figure 2. Chronic cocaine upregulates inhibitory MEF2 phosphorylation at Ser408/444

(A) Western blots using a MEF2A/2D phospho-S408/444-specific antibody demonstrate that chronic cocaine administration significantly increases MEF2 P-S408/444 phosphorylation in striatum 4 hours after the last injection (*p<0.05, n=3). (B) Quantification of MEF2 P-S408/444 western blotting at 24 hours after the last injection reveal that both acute and chronic cocaine significantly increase striatal P-MEF2 levels (***p<0.001, n=4–5). P-MEF2 levels returns to control levels by 48 hours after the final dose (p>0.05, n=4–5). (C) RNAi-based protein replacement assays comparing wild-type and S444A MEF2D activity. Expression of MEF2D S444A results in significantly elevated basal and KCl-induced MEF2-dependent transcription in MEF2-luciferase assays (***p<0.001 and ** p<0.01, respectively, n=3) (top). Similar effects are observed for the MEF2A phospho-mutant (S408A) (Supplemental Fig. S3A). Anti-MEF2D western blots of HEK-293T total cell lysates of cultures transfected with equal amounts of expression plasmids (bottom).

Figure 3. Cocaine regulation of MEF2 activity is necessary and sufficient to regulate NAc dendritic spine density in vivo

(A) Representative images of MEF2A (top) and MEF2D (bottom) immunostaining 28 days after stereotactic delivery of control AAV-shRNAs (left) or AAV-shRNAs against MEF2A/2D into the NAc. Coronal sections through striatum reveal dramatic knockdown of both MEF2A and MEF2D within the NAc (right). (B) RNAi-mediated reduction of MEF2A and MEF2D in the NAc significantly increases dendritic spine density in saline-treated mice (8.35+/−0.20 vs. 12.24+/−0.33, ***p<0.001). Representative confocal scans of NAc medium spiny neurons infected with either control or MEF2A/2D shRNAs (bottom). Dendritic spines were counted manually from confocal z-stacks of optically sectioned secondary and tertiary dendrites. Data represent spine density analysis from 4 saline-injected mice (4 weeks). (C) Representative image of AAV-MEF2-VP16 infection in the NAc 19 days after stereotactic delivery of virus. The bicistronically co-expressed GFP was visualized by immunohistochemistry. (D) Expression of MEF2-VP16 in the NAc significantly blocked cocaine-induced increases in dendritic spine density. Repeated cocaine injections (4 weeks at 20 mg/kg) induced a significant increase in NAc dendritic spine density compared to chronic saline in mice infected with the control MEF2ΔDBD-VP16 virus in their NAc (11.31 +/−0.34 vs. 8.36 +/−0.11; cocaine vs. saline, ***p<0.001). Expression of constitutively-active MEF2 (MEF2-VP16) significantly blocked the cocaine-induced increase in dendritic spine density in the NAc (11.31 +/−0.34 vs. 9.37 +/−0.24; control cocaine vs. MEF2-VP16 cocaine, ***p<0.001), but did not affect basal NAc dendritic spine density in saline-treated mice (8.36 +/−0.11 vs. 8.46 +/−0.17; control saline vs. MEF2-V16 saline, p>0.05). Representative confocal scans of NAc medium spiny neurons infected with either AAVs expressing control MEF2ΔDBD-VP16 or wild-type MEF2-VP16 (bottom). Data represent spine density analysis from 4 mice per condition (cocaine vs. saline).

Figure 4. MEF2 activity in the NAc modulates behavioral responses to cocaine

(A) Viral-mediated expression of MEF2-VP16 in the NAc significantly increases sensitivity to repeated cocaine administration. Mice expressing MEF2-VP16 in the NAc have normal locomotor response to saline and the first cocaine injection (15 mg/kg), but are significantly more sensitive to the subsequent dose compared to the MEF2-VP16 DNA binding mutant control (MEF2ΔDBD-VP16) (*p<0.05, n=9–10, Student's t-test on day 2). (B) Mice expressing MEF2-VP16 in their NAc remain significantly more sensitive to a challenge dose of cocaine (15 mg/kg) after one week of withdrawal (*p<0.05, n=9–10, Student's t-test). (C) Mice expressing MEF2-VP16 in the NAc spend more time in a cocaine-paired (8 mg/kg) environment as measured by conditioned place preference (*p<0.05, n = 13, Student's t-test). (D) Viral-mediated knockdown of MEF2A/2D in the NAc significantly reduces sensitivity to repeated cocaine administration. Mice expressing shRNAs against MEF2A/2D in the NAc have normal locomotor responses to saline and the first cocaine injection (15 mg/kg) but are significantly more sensitive to the subsequent dose compared to mice expressing control shRNAs (*p<0.05, n=9–11, Student's t-test on day 2). (E) Mice expressing shRNAs against MEF2A/2D in the NAc show significantly less cocaine-induced locomotor activity in response to a challenge dose (15 mg/kg) given two weeks after the acquisition of cocaine sensitization (**p<0.01, n=9–11, Student's t-test). RNAi-mediated reduction of MEF2A/2D in the NAc has only a slight trend towards reducing locomotor responses to a challenge dose of cocaine (15 mg/kg) one week after acquisition of cocaine sensitization (p>0.05, n=9–11).

Figure 5. Dopamine D1 receptor signaling and cAMP reduces calcium-dependent activation of MEF2 in cultured striatal neurons

(A) Dopamine D1 receptor stimulation (SKF81297, 10 µM) significantly reduces calcium-dependent activation of MEF2-luciferase activity in cultured striatal neurons (*p<0.05; n=15, five independent experiments, Student's t-test). (B) Forskolin (forsk) treatment (10 µM) of cultured striatal neurons significantly attenuates basal and KCl-induced MRE-luciferase activity. The inset shows the effect of forskolin on basal MRE-luciferase activity over a smaller scale (***p<0.001; n=21, seven independent experiments, Student's t-test). (C) Membrane depolarization (60 mM KCl) of cultured striatal neurons significantly increases CRE-luciferase activity. Treatment with dopamine D1 receptor agonist (SKF81297, 10 µM) significantly increases basal and KCl-induced CREB activity in cultured striatal neurons (***p<0.001; n=9, three independent experiments). (D) Constitutively-active calcineurin (CaNΔCT) blocks the inhibitory effect of forskolin on KCl-induced MEF2 activity (***p<0.001 or N.S. (p>0.05); n=6, two independent experiments, Student's t-test).

Figure 6. RCS mediates cAMP-dependent suppression of MEF2 activity

(A) Overexpression of RCS significantly enhances forskolin-induced inhibition of MEF2 activity. Cultured striatal neurons were transfected with a Flag-tagged RCS expression plasmid or vector and stimulated with either vehicle or forskolin (10 µM). Overexpression of RCS did not alter KCl-induced MRE-luciferase activity, but significantly potentiated the repressive effects of forskolin on KCl-induced activity (***p<0.001, n=12, four independent experiments, Student's t-test). (B) Phosphorylation of RCS at its protein kinase A (PKA) site is necessary for forskolin-induced inhibition of MEF2 activity. Cultured striatal neurons were transfected with either wild-type RCS or a point mutant of RCS (S55A) that cannot be activated by PKA. Neurons were then treated with forskolin alone or forskolin + 60 mM KCl. The enhanced suppression of MEF2 by wild-type RCS expression is not observed by expression of RCS S55A (***p<0.001, n=6, two independent experiments, Student's t-test). Anti-Flag western blots showing equal expression of wild-type RCS and mutant RCS in HEK-293T cell lysates transfected with equal amounts of the respective plasmids (bottom). (C) Adult striatal slices treated with forskolin (50 µM) or SKF81297 (10 µM) for 10 minutes significantly increased P-Ser55 RCS levels (*p<0.05 and **p<0.01, respectively, Student's t-test). (D) Repeated cocaine administration (7 days of daily IP injections of 20 mg/kg cocaine) significantly increased RCS Ser55 phosphorylation in the striatum at 4 hours after the last injection (cocaine vs. saline, **p<0.01, n=6, Student's t-test).

Figure 7. ChIP-chip analysis of MEF2 in the NAc reveals potential target genes involved in dendritic plasticity and cocaine sensitization

(A) MEF2 binding in the NAc of chronic cocaine-treated mice is displayed (log2 ratio) along chromosome 17. One region is magnified to display the MEF2 target genes, GABAB receptor and the ubiquitin protein, Ubd. (B) Venn diagram illustrating the overlap between the genes on which MEF2 is significantly enriched and the genes whose expression is downregulated >1.2 fold by chronic cocaine in the nucleus accumbens. (C) The location of 3 MEF2-response element (MRE)-like regions (labeled 1–3) are shown on the gene promoter of pik3cg, a gene both significantly bound by MEF2 and downregulated >1.2 fold by cocaine. Semi-quantitative ChIP demonstrates enrichment of MEF2 at each of these regions over the IgG control. (D) Confirmation that Pik3cg mRNA is significantly downregulated by cocaine in an independent set of mice (Student's t-test, *P < 0.05, n=6). (E) PC12 cells transfected with a plasmid expressing MEF2-VP16-ER^tm, which allows the inducible expression of MEF2-VP16 upon addition of 4-hydroxytomaxifen (4OHT), were treated with 4OHT or DMSO for 2hrs, 4hrs, or 8hrs, and Pik3cg mRNA levels were quantified by qRT-PCR. Activation of MEF2-VP16 expression significantly upregulated Pik3cg mRNA at 4hrs and 8hrs (*p<0.05, n=3, Student's t-test) compared to its DMSO control. (F) Chronic cocaine (7 days × 20 mg/kg daily IP injections) significantly reduced Akt phosphorylation (Ser473) in the NAc at 4 hours after the final injection (**p<0.01, n=8–11, Students t-test).

Figure 8

Model for how cocaine regulates MEF2 in the NAc to alter dendritic spine density and behavioral responses.