Epigenetics of amphetamine-induced sensitization: HDAC5 expression and microRNA in neural remodeling

BACKGROUND: Histone deacetylase (HDAC) activities modify chromatin structure and play a role in learning and memory during developmental processes. Studies of adult mice suggest HDACs are involved in neural network remodeling in brain repair, but its function in drug addiction is less understood. We aimed to examine in vivo HDAC5 expression in a preclinical model of amphetamine-induced sensitization (AIS) of behavior. We generated specific contrast agents to measure HDAC5 levels by in vivo molecular contrast-enhanced (MCE) magnetic resonance imaging (MRI) in amphetamine-naïve mice as well as in mice with AIS. To validate the MRI results we used ex vivo methods including in situ hybridization, RT-PCR, immunohistochemistry, and transmision electron microscopy. METHODS: We compared the expression of HDAC5 mRNA in an acute exposure paradigm (in which animals experienced a single drug exposure [A1]) and in a chronic-abstinence-challenge paradigm (in which animals were exposed to the drug once every other day for seven doses, then underwent 2 weeks of abstinence followed by a challenge dose [A7WA]). Control groups for each of these exposure paradigms were given saline. To delineate how HDAC5 expression was related to AIS, we compared the expression of HDAC5 mRNA at sequences where no known microRNA (miR) binds (hdac5AS2) and at sequences where miR-2861 is known to bind (miD2861). We synthesized and labeled phosphorothioated oligonucleic acids (sODN) of hdac5AS2 or miD2861 linked to superparamagentic iron oxide nanoparticles (SPION), and generated HDAC5-specific contrast agents (30 ± 20 nm, diameter) for MCE MRI; the same sequences were used for primers for TaqMan® analysis (RT-qPCR) in ex vivo validation. In addition, we used subtraction R2* maps to identify regional HDAC5 expression. RESULTS: Naïve C57black6 mice that experience acute exposure to amphetamine (4 mg/kg, by injection intraperitoneally) show expression of both total and phosphorylated (S259) HDAC5 antigens in GFAP(+) and GFAP(−) cells, but the appearance of these cells was attenuated in the chronic paradigm. We found that MCE MRI reports HDAC5 mRNA with precision in physiological conditions because the HDAC5 mRNA copy number reported by TaqMan analysis was positively correlated (with a linear coefficient of 1.0) to the ΔR2* values (the frequency of signal reduction above background, 1/s) measured by MRI. We observed SPION-mid2861 as electron dense nanoparticles (EDNs) of less than 30 nm in the nucleus of the neurons, macrophages, and microglia, but not in glia and endothelia. We found no preferential distribution in any particular type of neural cells, but observed scattered EDNs of 60–150 nm (dia) in lysosomes. In the acute paradigm, mice pretreated with miD2861 (1.2 mmol/kg, i.p./icv) exhibited AIS similar to that exibited by mice in the chronic exposure group, which exhibited null response to mid2861 pretreatment. Moreover, SPION-miD2861 identified enhanced HDAC5 expression in the lateral septum and the striatum after amphetamine, where we found neurprogenitor cells coexpressing NeuN and GFAP. CONCLUSIONS: We conclude that miD2681 targets HDAC5 mRNA with precision similar to that of RT-PCR. Our MCE MRI detects RNA-bound nanoparticles (NPs) in vivo, and ex vivo validation methods confirm that EDNs do not accumulate in any particular cell type. As HDAC5 expression may help nullify AIS and identify progenitor cells, the precise delivery of miD2861 may serve as a vehicle for monitoring network remodeling with target specificity and signal sensitivity after drug exposure that identifies brain repair processes in adult animals. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12929-016-0294-8) contains supplementary material, which is available to authorized users.