The present report shows that the activation and/or overexpression of PKG in the CPu strikingly reduced the expression levels of the epigenetic parameters, MeCP2 and HDAC2, in dopaminergic projection areas of cocaine-treated rats. Both proteins were reduced maximally in about 30% of the cells, regardless of the percentage of cells that expressed each protein in control conditions. Studies from several laboratories, including ours, have shown that cocaine-induced modulation in gene expression is achieved, at least partially, via epigenetic regulation (Kumar et al. 2005
; Cassel et al. 2006
). For instance, DNA methylation and histone acetylation have been implicated in stimulant-related behavioral and molecular adaptations. We found that cocaine increased MeCP2 and HDAC2 nuclear expression, in response to repeated experimenter- or self-administered exposure (Cassel et al. 2006
; Host et al. 2011
), supporting the hypothesis that epigenetic regulation plays an important role in the development and maintenance of drug addiction. In this study, PKG was found to reduce the expression of these epigenetic factors, suggesting that activators of the cGMP pathway may be used as general pharmacological tools for downregulating the MeCP2/HDAC2 complex. This is comforted by the fact that PKG activation also reduced MeCP2 and HDAC2 expression in rats that were not injected with cocaine, although to a lesser extent.
In previous studies, we showed that activation of the cGMP pathway was sufficient to attenuate the increase in extracellular dopamine, immediate early gene expression, and locomotor activation produced by cocaine (Thiriet et al. 2001
; Jouvert et al. 2004
). As we used a technique to overexpress the PKG in which highest kinase activity was produced 24 h after plasmid delivery (Jouvert et al. 2004
), only effects of an acute injection of cocaine could be determined. Surprisingly, the simple activation of the endogenous PKG in the CPu was sufficient to decrease MeCP2 and HDAC2 protein expression in the dorsal CPu, NAc shell, and PFCx in this short timescale. This may result from a decreased protein expression, possibly due to the phosphorylation, and subsequent inactivation of an element of the transcriptional machinery. Alternatively, it may result from the PKG-induced degradation of the proteins. It is also possible that phosphorylation by PKG of the proteins or of some partner proteins may have induced the formation of alternate complexes that prevented accessibility of MeCP2 and HDAC2 to their respective antibodies. This latter possibility is particularly attractive as both proteins were similarly regulated. The effect on both proteins was abolished by the injection of the selective PKG inhibitor, KT5823, confirming that the effect was due to PKG-dependent phosphorylation.
The effect on MeCP2 expression was greatly potentiated in the three structures examined by the addition and subsequent activation of exogenous overexpressed PKG. Surprisingly, the latter effect was not observed when HDAC2 expression was measured. It is noteworthy that other HDACs, such as HDAC5 or HDAC11, which belong to different classes of histone deacetylases that are also regulated by cocaine (Host et al. 2011
), were not modulated by PKG (data not shown), suggesting a unique role of HDAC2 in this process.
In contrast, injection of the PKG vector or Br-cG into the VTA failed to alter MeCP2 or HDAC2 levels in any of the brain sites. The data probably underscore a differential action of PKG at the pre- and postsynaptic levels. Effectively, the observation that only intra-CPu infusions produced the effect probably denotes that it results from a PKG-modified signal transduction pathway taking place in neurons postsynaptic from the dopaminergic neuron. Indeed, in an electron microscopy study, we showed that PKG was localized mainly in postsynaptic sites (data not shown). These findings are consistent with our previous results that also failed to find an effect of intra-VTA infusions of Br-cG on cocaine-induced egr-1
expression in the CPu (Jouvert et al. 2004
When the PKG vector was injected into the CPu, overexpression of the kinase was found primarily in GABAergic medium-spiny neurons (Jouvert et al. 2004
). Activation of the cGMP pathway in these neurons probably modulates GABA release in many projection areas and therefore disrupts the striato-nigro-thalamic loop. Modulation of neurotransmitter release by cGMP has been demonstrated in many systems; cGMP activators like natriuretic peptides inhibit various secretory responses such as aldosterone and catecholamines (Samson et al. 1993
; Babinski et al. 1995
), whereas endogenous NO modulates the release of several neurotransmitters, including catecholamines, excitatory and inhibitory amino acids, and serotonin (Guevara-Guzman et al. 1994
; Prast and Philippu 2001
). Accordingly, disturbance of the striato-nigro-thalamic loop will have an impact on cortical neurons. One of these consequences would be the degradation or downregulation of the MeCP2/HDAC2 complex. The precise mechanism of this proposed effect is currently unknown.
HDAC2 was chosen because it is part of a complex composed also of Sin3A and MeCP2, after the latter binds to methylated DNA. The resulting HDAC activity thus carries transcriptional silencing to the corresponding genes (Yang and Seto 2008
). The enzyme is also highly expressed in the mesolimbic pathway (Cassel et al. 2006
; Broide et al. 2007
). HDAC2 has been reported to regulate memory formation and synapse plasticity in mature neurons (Grissom and Lubin 2009
; Guan et al. 2009
; Pastor et al. 2011
). Similarly, MeCP2 is highly expressed in mature neurons where it is required for modulating dynamic functions of the adult brain and mutations within the gene are known to be associated with Rett syndrome (Nelson et al. 2006
; Zhou et al. 2006
). The fact that PKG was able to downregulate the expression of both MeCP2 and HDAC2 proteins when injected into the CPu suggests that the cGMP pathway affects cognitive processes through a mechanism that comprises the MeCP2/HDAC2 complex and the gene silencing that it controls. Interestingly, egr-1
may be one of the genes silenced by this mechanism, as levels of AcH3 and AcH4 were increased in the egr-1
promoter in HDAC2 KO mice (Guan et al. 2009
). The fact that activation of PKG reduced both HDAC2 levels and egr-1
induction suggests that the MeCP2/HDAC2 complex regulates egr-1
expression, at least to some extent.
Phosphodiesterases have recently been suggested as potential new targets for cognition enhancement (Reneerkens et al. 2009
). Results of this study are consistent with this idea and suggest that amplification of the intracellular availability of the second messenger cGMP by phosphodiesterase inhibitors have therapeutic potential for the treatment of neuropsychiatric disorders involving disturbances of mood, emotion, and cognition, including drug addiction.