Epigenetic function during heroin self-administration controls future relapse-associated behavior in a cell type-specific manner
Edited by Yasmin Hurd, Icahn School of Medicine at Mount Sinai, New York, NY; received June 25, 2022; accepted January 6, 2023
Significance
Opioid use disorder (OUD) is a chronic, relapsing disease that impacts public health substantially. Understanding how active opioid use creates future relapse vulnerability could lead to treatments for people that suffer from OUD. Here, we show how an epigenetic factor, HDAC5, limits the vulnerability to relapse-associated behavior in a rodent model of OUD. We also show that HDAC5 acts in different classes of neurons to regulate different triggers of heroin seeking. Finally, we discovered that HDAC5 regulates a genetic program linked to neuron excitability, and it reduces the firing rate of critical neurons in a major brain-reward region, the nucleus accumbens, suggesting that HDAC5 reduces future relapse-associated behavior by limiting neuronal excitability during active opioid use.
Abstract
Opioid use produces enduring associations between drug reinforcement/euphoria and discreet or diffuse cues in the drug-taking environment. These powerful associations can trigger relapse in individuals recovering from opioid use disorder (OUD). Here, we sought to determine whether the epigenetic enzyme, histone deacetylase 5 (HDAC5), regulates relapse-associated behavior in an animal model of OUD. We examined the effects of nucleus accumbens (NAc) HDAC5 on both heroin- and sucrose-seeking behaviors using operant self-administration paradigms. We utilized cre-dependent viral-mediated approaches to investigate the cell-type–specific effects of HDAC5 on heroin-seeking behavior, gene expression, and medium spiny neuron (MSN) cell and synaptic physiology. We found that NAc HDAC5 functions during the acquisition phase of heroin self-administration to limit future relapse-associated behavior. Moreover, overexpressing HDAC5 in the NAc suppressed context-associated and reinstated heroin-seeking behaviors, but it did not alter sucrose seeking. We also found that HDAC5 functions within dopamine D1 receptor-expressing MSNs to suppress cue-induced heroin seeking, and within dopamine D2 receptor-expressing MSNs to suppress drug-primed heroin seeking. Assessing cell-type–specific transcriptomics, we found that HDAC5 reduced expression of multiple ion transport genes in both D1- and D2-MSNs. Consistent with this observation, HDAC5 also produced firing rate depression in both MSN classes. These findings revealed roles for HDAC5 during active heroin use in both D1- and D2-MSNs to limit distinct triggers of drug-seeking behavior. Together, our results suggest that HDAC5 might limit relapse vulnerability through regulation of ion channel gene expression and suppression of MSN firing rates during active heroin use.
Sign up for PNAS alerts.
Get alerts for new articles, or get an alert when an article is cited.
Individuals with substance use disorders (SUDs), which include opioid use disorder (OUD), typically struggle to remain abstinent despite repeated attempts to discontinue drug use. The current opioid crisis (1) has highlighted the significant challenges to help individuals with a SUD mitigate relapse vulnerability. Active use of addictive substances, like heroin and cocaine, produces potent memory traces that indelibly link external and internal drug-associated cues with the rewarding/euphoric effects of the drugs. These drug memories often serve as powerful triggers of relapse in abstinent individuals recovering from SUDs, but the cellular processes that create and maintain these relapse-promoting memories remain poorly understood. Epigenetic enzymes, such as histone deacetylases, histone methyltransferases, and DNA methylating/demethylating enzymes, have enduring effects on cellular function by modifying chromatin structure, enhancer accessibility, and basal and adaptive gene expression programs (2). Moreover, addictive drugs cause numerous epigenetic changes in the nucleus accumbens (NAc), a region within the mesolimbic dopamine pathway and an important area for drug reward (3).
One epigenetic enzyme in the NAc is the class IIa histone deacetylase 5 (HDAC5) (4–6). HDAC5 shuttles between the nucleus and cytoplasm through a phosphorylation-dependent process that is governed by calcium and cyclic adenosine monophosphate (cAMP) signaling (7), and in the cell nucleus, HDAC5 appears to function primarily as a repressor of gene expression. NAc HDAC5 is regulated by several classes of addictive drugs. For example, ethanol downregulates nuclear HDAC5 protein (8) and cocaine alters HDAC5’s phosphorylation and subcellular localization (5). Interestingly, human postmortem studies show that HDAC5 messenger ribonucleic acid (mRNA) is downregulated in the NAc of heroin-using individuals (9). Consistent with its regulation by cocaine, we found previously that overexpression of nuclear HDAC5 in the NAc suppresses both cue- and drug prime-induced cocaine seeking in the extinction-reinstatement paradigm following operant cocaine self-administration, a gold-standard model of relapse-associated behavior in rodents (4).
In the current study, we discovered that endogenous HDAC5 in the NAc limited relapse-associated behavior following operant heroin self-administration and forced abstinence. Consistent with this result, overexpression of nuclear HDAC5 suppressed context-associated and reinstated heroin seeking, but without altering natural reward (sucrose) seeking behaviors. This suggests a drug-selective function for NAc HDAC5 in relapse-associated memory formation. We also found that overexpressed HDAC5 functioned during active heroin taking to influence future relapse-associated behaviors, including heroin seeking in the drug-taking context and in response to the presentation of the drug-associated cues or drug priming in the extinction-reinstatement model of relapse. Using cell-type–specific approaches, we found that HDAC5 functioned in D1-medium spiny neurons (MSNs) to suppress cue-induced heroin seeking, whereas it functioned in D2-MSNs to block heroin-primed drug-seeking behavior. Cell-type–specific analysis of HDAC5’s impact on D1- or D2-MSN gene expression revealed enrichment of downregulated genes that regulate cellular ion transport. Consistent with this, we observed that HDAC5 significantly reduced intrinsic excitability of both D1 and D2-class MSNs in the NAc, without altering basal glutamatergic synaptic transmission. Together, our findings reveal that HDAC5 functions during active heroin use to regulate MSN neuronal excitability and limit the formation of potent relapse-triggering drug memories in a D1 vs. D2 cell-type–specific manner.
Results
NAc HDAC5 Expression Bidirectionally Regulates Three Distinct Forms of Heroin-Seeking Behaviors.
Since NAc Hdac5 mRNA is reduced in postmortem samples from humans that used heroin (9), we tested the influence of reducing endogenous NAc HDAC5 on intravenous self-administration of heroin and on future heroin-seeking behaviors (Fig. 1A). To this end, we bilaterally infused into the adult rat NAc a neuron-specific adeno-associated virus (AAV2) (10) expressing an HDAC5 short hairpin RNA (shRNA) (AAV-shHDAC5) (SI Appendix, Fig. S1A) (11). AAVs often require at least 2 wk to approach peak expression, but then continually express for years. Compared to the control luciferase shRNA condition (AAV-shLuc), the reduction of endogenous NAc HDAC5 had no obvious effects on the acquisition or stable intake of heroin self-administration (Fig. 1B and SI Appendix, Fig. S1B). Following 7 d (1 wk) of forced abstinence, the rats were placed back into the operant chambers and nonreinforced lever-pressing behavior was measured. We observed that reduction of NAc HDAC5 augmented lever pressing on the lever formerly paired with heroin delivery (Fig. 1C), suggesting that NAc HDAC5 limits context-associated heroin-seeking. Following extinction training (SI Appendix, Fig. S1C), we observed that reduction of NAc HDAC5 increased both cue-induced (Fig. 1D) and heroin-primed reinstatement of drug seeking (Fig. 1E). Together, our findings indicate that endogenous NAc HDAC5 does not influence active heroin-taking behavior, but it limits multiple forms of relapse-associated behavior.
Fig. 1.
We previously showed that cocaine promotes enhanced nuclear localization of HDAC5 to limit drug-context associations (5), so we hypothesized that heroin might also regulate HDAC5 nucleocytoplasmic localization. To test this idea, we analyzed the subcellular localization of viral-mediated expression of a Flag-tagged HDAC5 in the NAc of adult rats following five daily infusions of heroin (2 mg/kg, intraperitoneal (i.p.)) or saline vehicle. Four hours after the last heroin injection, we observed a significant increase in HDAC5 nuclear localization (SI Appendix, Fig. S1E). We therefore hypothesized that enhancing NAc HDAC5 nuclear function might limit the heroin-seeking behavior. To this end, we infused AAV-HDAC5-3SA bilaterally into the adult rat NAc to overexpress a nuclear-enriched mutant of HDAC5 lacking three conserved serine residues (3SA) that promote cytoplasmic localization when phosphorylated (4, 5). The viral-mediated HDAC5 overexpression (SI Appendix, Fig. S1E) produced no obvious effects on acquisition or stable intake during heroin self-administration (Fig. 1F and SI Appendix, Fig. S1F). However, following 7 d (1 wk) of forced abstinence, HDAC5-3SA reduced non-reinforced pressing of the lever formerly paired with heroin (Fig. 1G). Following normal extinction training (SI Appendix, Fig. S1G), we found that NAc HDAC5-3SA suppressed both cue-induced (Fig. 1H) and heroin-primed reinstatement of drug seeking (Fig. 1I). Taken together with the HDAC5 shRNA data (Fig. 1 C–E) and our prior cocaine self-administration study (4), these data indicate that NAc HDAC5 limits multiple relapse-associated behaviors across multiple classes of addictive substances (i.e., opioids and stimulants).
NAc HDAC5 Does Not Alter Sucrose Seeking Behaviors.
Since the NAc HDAC5-mediated suppression of heroin- and cocaine-seeking behavior could be caused by a reduction of general reward-seeking behavior, we next tested the effects of HDAC5-3SA on sucrose self-administration in a similar study design (Fig. 2A). We detected no statistically significant effects of HDAC5-3SA on acquisition or stable intake of sucrose self-administration (Fig. 2B and SI Appendix, Fig. S2A), nor any effects on context-associated sucrose seeking (Fig. 2C), extinction training (SI Appendix, Fig. S2B), or reinstated sucrose seeking (Fig. 2 D and E), suggesting that NAc HDAC5 effects on heroin and cocaine relapse-associated behavior are selective to addictive substances.
Fig. 2.
NAc HDAC5 Functions During the Acquisition Phase to Limit Heroin-Seeking Behaviors.
Since the viral-mediated manipulations of NAc HDAC5 occurred throughout the duration of active heroin self-administration and subsequent heroin-seeking tests, we sought to determine if HDAC5 functions during the acquisition phase of heroin self-administration to suppress future drug-seeking behavior. To this end, we allowed rats to established normal heroin self-administration (Fig. 2 F and G and SI Appendix, Fig. S2C) and then we infused AAV-HDAC5-3SA into the NAc. To allow the AAV-HDAC5-3SA virus time to approach peak expression levels, we extended the home-cage abstinence phase to 14 d. However, post-acquisition expression of HDAC5-3SA did not suppress either context-associated heroin seeking (Fig. 2H), extinction learning (SI Appendix, Fig. S2D), or reinstated heroin seeking (Fig. 2 I and J), indicating that HDAC5-3SA needs to function during the acquisition phase of active heroin self-administration to limit future drug-seeking.
Nuclear HDAC5 Functions in D1-MSNs to Suppress Cued Heroin Seeking, but in D2-MSNs to Suppress Drug-Primed Heroin Seeking in a Doubly Dissociable Manner.
A vast majority (~85 to 90%) (12) of NAc neurons are MSNs expressing either dopamine D1 receptors or dopamine D2 receptors (D1-MSN or D2-MSN), and they typically have opposing effects on drug-associated behavior with D1-MSNs promoting drug-associated behavior (13–15). Based on this literature, we hypothesized that HDAC5’s suppression of heroin seeking would be mediated by its role in D1-MSNs and not D2-MSNs. Using bacterial artificial chromosome (BAC) transgenic rats expressing Cre under the control of either the Drd1 promoter (D1-Cre) or Drd2 promoter (D2-Cre) (16), we first examined the effects of heroin on nuclear localization of HDAC5 in these cell types. We found that heroin causes HDAC5 to localize to the nucleus in both D1- and D2-MSNs; however, the nuclear enrichment was significantly greater in the D1-Cre cells (Fig. 3A and SI Appendix, Fig. S3A). To test the effects of nuclear HDAC5 on heroin-seeking behaviors in D1- and D2-cells, we generated a new virus with double-floxed inverse orientation (DIO) HDAC5-3SA (AAV2-DIO-HDAC5-3SA) to allow for Cre-dependent expression of HDAC5-3SA (Fig. 3B). We infused AAV2-DIO-HDAC5-3SA or AAV2-DIO-GFP bilaterally into the adult NAc. We confirmed D1- or D2-cell–specific overexpression of Cre-dependent viral proteins (Fig. 3C), confirmed that HDAC5-3SA localizes to the nucleus in both D1- and D2-MSNs (SI Appendix, Fig. S3B), and showed that HDAC5-3SA was overexpressed to a similar extent in both cell types (SI Appendix, Fig. S3C). We then tested the effects of cell-type–specific HDAC5-3SA or green fluorescent protein (GFP) expression on heroin self-administration and subsequent drug-seeking behaviors (Fig. 3D). In both the D1-Cre and D2-Cre rats, AAV-DIO-HDAC5-3SA has no significant effects on acquisition or stable intake in heroin self-administration (Fig. 3 E and J and SI Appendix, Fig. S3 D and F) and, surprisingly, no effects on context-associated seeking (Fig. 3 F and J). Following extinction training (SI Appendix, Fig. S3 E and G), D1-specific NAc HDAC5-3SA significantly reduced cue-reinstated heroin seeking (Fig. 3G), but surprisingly, it had no effect on heroin-reinstated drug seeking (Fig. 3H). In contrast, the D2-specific NAc HDAC5-3SA had no effect on cue-induced reinstatement (Fig. 3K), but it suppressed heroin-reinstated drug seeking (Fig. 3L). Of note, in the control virus condition, we observed significant differences in cue-reinstated heroin seeking, with D1-Cre rats displaying an increase in paired lever presses compared to D2-Cre rats (unpaired t test, t(14) = 32.93, P = 0.0053), suggesting a possible influence of the Cre transgene insertion on cued responding (Fig. 3 G and K). In addition, there are some potential differences in heroin-primed reinstatement levels between Long Evans and the Sprague-Dawley rats, consistent with known strain effects in rodent self-administration studies (17, 18). However, in our studies we did not directly compare strains within the same experiment, so we cannot draw firm conclusions based on our findings reported here. These findings suggest that HDAC5 limits reinstated heroin-seeking behaviors through roles in distinct NAc cell populations in a doubly dissociable manner and that HDAC5’s suppression of context-associated heroin seeking cannot be explained by a selective role in either D1- or D2-MSNs.
Fig. 3.
Cell-Type–Specific Transcriptomic Analysis Reveals That HDAC5 Downregulates Genes Linked to Ion Transport Functions.
To examine the cellular mechanism(s) by which HDAC5, a transcriptional repressor, limits cue-induced reinstatement of heroin seeking in D1-MSNs, we infused in the NAc of D1-Cre or D2-Cre rats a mixture of AAV2-DIO-HDAC5-3SA together with AAV2-viral translating ribosomal affinity purification (vTRAP), which allows for Cre-dependent expression of L10a-GFP and subsequent enrichment of actively translating mRNAs in a cell-type–specific manner (19). Since coinfusion of the two viruses produced >90% coexpression of GFP and HDAC5 (SI Appendix, Fig. S4A), this allowed us to assess cell-specific effects of HDAC5-3SA on gene expression. Following qPCR validation of vTRAP enrichment from NAc tissues from the D1-Cre or D2-Cre rats (Fig. 4A), we isolated ribosome-associated transcripts from both cell types in the presence or absence of AAV2-DIO-HDAC5-3SA after at least 3 wk of AAV expression. Due to the relatively low yields of mRNAs isolated from Cre-expressing cells in the vTRAP pulldown approach, we analyzed differentially expressed genes (DEGs) using gene microarrays. We found that HDAC5-3SA produced 157 DEGs (Fig. 4B and SI Appendix, Fig. S4B and Table S1) in D1-Cre animals, with the majority (76%) being downregulated transcripts as expected from overexpression of a transcriptional repressor. From D2-Cre rat NAc tissues, we found that HDAC5-3SA produced 1993 DEGs (Fig. 4C and SI Appendix, Fig. S4C and Table S1), and the majority (55%) were downregulated. Analysis of HDAC5 3SA DEGs commonly regulated between the two cell types revealed 56 genes, which was a significant overlap (Fischer's exact test; Odds Ratio: 3.9, P = 6.73e-14), and a vast majority showed a similar magnitude of change and direction of regulation (i.e., 91% were downregulated) (Fig. 4D). Functional enrichment and gene network analysis revealed that the common HDAC5 DEGs were significantly enriched for ion transport functions, including sodium- and potassium-ion transport (Fig. 4 E and F). Notable DEGs in this enrichment included Kcnv1, Kcnt2, Akap5, Scn3b, Scn4b, Akt3, and many other genes encoding ion channels and their regulators. Of note, AMPAR and NMDAR subunit genes were not significantly regulated by HDAC5 in either D1- or D2-cells (SI Appendix, Table S1). Similar functional enrichment was observed for D1-specific DEGs (SI Appendix, Fig. S4D), whereas D2-specific DEGs revealed additional enrichments for mitochondrial functions (SI Appendix, Fig. S4E). To validate our vTRAP DEGs, we independently confirmed using qRT-PCR a subset of targets, including Akap5 and Kcnv1 (Fig. 4G). Together, these data suggest that HDAC5-3SA functions predominantly to reduce target gene expression, and that the common DEGs between D1- and D2-MSNs show a strong functional enrichment for genes regulating sodium and potassium ion transport and neuronal excitability.
Fig. 4.
Nuclear HDAC5 Reduces Intrinsic Excitability in Both D1- and D2-MSNs in the NAc.
Since HDAC5-3SA produced a strong enrichment of ion transport-related DEGs in the shared set of both D1- and D2-MSNs (Fig. 4 E and F), we next examined in D1- or D2-MSNs the effects of HDAC5-3SA on intrinsic excitability, evoked α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) mediated and N-methyl-D-aspartate (NMDA) mediated glutamatergic synaptic transmission, and paired-pulse ratio (PPR) of AMPA-mediated excitatory postsynaptic currents using ex vivo acute slice preparations after at least 3 wk of in vivo AAV2 expression. In both D1- and D2-MSNs, we observed that HDAC5-3SA reduced the number of action potentials produced by increasing steps of current injection (Fig. 5 A and D), suggesting a reduction in intrinsic excitability. However, using 20 pA current steps, HDAC5-3SA did not alter rheobase in D1- or D2-MSNs (SI Appendix, Fig. S5 A and B), suggesting that HDAC5 does not alter the threshold for eliciting the first action potential, but rather it limits the number of action potentials produced by stronger depolarization. In contrast to the reduction in current-spike relationship, HDAC5-3SA had no effect on the electrically evoked AMPA/NMDA ratio (Fig. 5 B and E) or PPR (Fig. 5 C and F), suggesting that HDAC5 does not influence basal excitatory synaptic transmission and presynaptic function of glutamatergic inputs onto D1- and D2-MSNs in the NAc. Taken together, our findings reveal that HDAC5 reduces expression of numerous voltage-gated ion channel genes and suppresses the firing rate of NAc D1- and D2-MSNs, likely by altering the intrinsic electrical properties of D1- and D2-MSNs, but without altering their excitatory synaptic properties.
Fig. 5.
Discussion
Our findings strongly suggest that NAc HDAC5 functions during the acquisition phase of active heroin taking to limit multiple forms of relapse-associated behaviors following abstinence but without altering natural reward (sucrose) seeking. We also found that HDAC5 functions in NAc D1-MSNs to limit cue-reinstated heroin seeking, but in D2-MSNs to limit heroin-reinstated drug seeking. Our cell-type–specific transcriptomic analysis revealed that HDAC5 in the NAc MSNs reduces expression of numerous genes linked to ion transport, including voltage-gated sodium and potassium channels across both cell populations. Moreover, HDAC5 reduces MSN intrinsic excitability, suggesting that reducing the MSN firing rate of either D1- or D2-MSNs during the acquisition phase of heroin self-administration blocks the formation of potent heroin-associated drug-pairings that trigger drug seeking, but without altering active heroin-reinforced operant learning or extinction learning. As such, HDAC5 could control MSN firing rates necessary to form and/or consolidate the stable memories linking active drug reinforcement with the diffuse and discreet environmental cues associated with drug taking. Taken together with prior studies (4, 5), our current findings suggest that HDAC5’s effects are selective to addictive drugs (i.e., cocaine and heroin, but not sucrose) and that HDAC5 functions in different NAc neuronal populations to limit different triggers of drug seeking.
In this study, we investigated the D1- vs. D2-MSN cell-type–specific effects of nuclear HDAC5 and found a double dissociation of cued vs. drug-primed heroin reinstatement. In other words, we found that D1-MSNs exclusively mediated HDAC5’s ability to reduce cue-induced reinstatement, but D2-MSNs exclusively mediated the reduction of heroin-primed reinstatement. D1-MSNs in the NAc core are well known to play essential roles in cue-reinstated drug seeking (20), so our observation that HDAC5 functions predominantly in D1-MSNs to suppress cue-induced heroin seeking was not surprising. However, in much of the SUD literature, D1-MSNs primarily increase drug-related behaviors, while D2-MSNs oppose these same behaviors. For example, D1- vs. D2-MSN activation produces opposing effects on conditioned place preference (13), locomotor (13, 15) and cue-induced reinstatement behavior (21). Also, systemically administered D1- vs. D2-like agonists have opposing effects on cocaine-seeking behavior (22). In contrast to systemically administered agonists, intra-NAc administration of either D1- or D2-like agonists both induces reinstatement to drug-seeking (23–25) but has opposing effects on neuronal activation through differential activation of the cAMP pathway (26). The ability of HDAC5 in D2-expressing cells to suppress heroin-prime reinstated seeking was therefore unexpected. Within the SUD literature, most studies of D1 and/or D2 mechanisms examine only the extinction responding and/or reinstatement to cues and do not examine drug-primed reinstatement (20, 21, 27–31), so few insights exist. Outside the SUD literature, a similar cell-type dissociation between the role for D1-MSNs in cue conditioning for sucrose vs. a role for D2-MSNs in discrimination was reported (32). Combined, these results suggest that D1- and D2-MSNs might play separable, but complementary, roles in seeking behaviors. Perhaps D2-MSNs help to create stable associations of the discriminative effects of heroin to pressing of the drug-paired lever, which is necessary for drug-primed reinstatement. In contrast, D1-MSNs associate the discreet conditioned stimulus of the light/tone cues to the drug-paired lever. As such, it is possible that HDAC5’s cellular effects are the same in D1 and D2-MSNs, but each cell type plays a more critical role in each mode of reinstated seeking (Fig. 6). It is also important to note that HDAC5 does not appear to reduce reinstated drug seeking through a role following cessation of active drug taking. This suggests that HDAC5 in NAc MSNs limits maladaptive metaplasticity that occurs during the acquisition of heroin self-administration. Several other findings using nondrug rewards support a role for D2-MSNs in mediating rewarding effects in the NAc. Animals will self-administer D1- and D2-agonists into the NAc only when coadministered (33). Also, coactivation of both D1 and D2 receptors in the NAc is necessary for intracranial self-stimulation (34). Interestingly, selective activation of NAc D1- or D2-MSNs alone supports optogenetic self-stimulation, though at different rates (35). As such, our current findings suggest that HDAC5 might have similar effects on D1- and D2-MSN functions, but that each cell type plays selective roles in encoding discreet cue memories (D1) vs. interoceptive cue memories (D2), which are essential for the different modes of reinstated drug seeking. It is also worth noting that neither D1- nor D2-MSNs appear to underlie HDAC5’s ability to reduce context-associated heroin seeking during day 1 extinction responding. One possible explanation is that HDAC5’s role in both D1- and D2-MSNs is required, but neither is sufficient. However, it is also possible that HDAC5 functions within a NAc interneuron class to influence context-associated heroin seeking. It will be important to further examine these possibilities since nonextinguished, context-triggered drug seeking has face-validity for relapse in humans following voluntary or forced abstinence. Importantly, our data also show that altering the function of a single NAc cell type is insufficient to reduce relapse-associated behavior produced by different types of external or internal cues. As such, enhancing HDAC5 function in multiple NAc cell types might be needed to mitigate relapse potential.
Fig. 6.
Since HDAC5 is an epigenetic protein that represses numerous genes, we examined the effects of nuclear HDAC5 in D1-MSNs using a viral-mediated vTRAP assay to understand how HDAC5 may be exerting its actions on reinstatement behavior. Our transcriptomic analysis found many genes that were significantly downregulated in D1-MSNs. Gene ontology analyses showed that these genes were involved with electrophysiological properties like the regulation of ion transport, membrane potential regulation, and dopaminergic signaling. At least some of these genes are also downregulated in D2-MSNs, suggesting that nuclear HDAC5 may have a similar regulatory role in both D1- and D2-MSNs on cell excitability and synaptic or cellular adaptations engaged during active heroin use. Indeed, HDAC5-3SA reduced both D1- and D2-MSN intrinsic excitability, suggesting a neural activity-based mechanism for the reduced reinstatement behavior. It is known that optogenetic or chemogenetic suppression of MSN activity during the reinstatement test reduces cued drug seeking (36–38), but while HDAC5 suppresses the MSN firing rate at higher current steps, the rheobase is unchanged and reinstated drug seeking was unaffected when the HDAC5 was expressed following the completion of stable heroin self-administration, suggesting that HDAC5’s effects on MSN activity are relatively mild compared to activity suppression produced by optogenetic or chemogenetic tools. Instead, our findings might suggest that suppression of intrinsic excitability during active drug use is required for the plasticity mechanisms linking the ensemble of active neurons to the drug-taking context, discreet cues, and interoceptive cues. Moreover, enhancing HDAC5 function can prevent the formation of these learned relapse-associated behaviors, but it cannot reverse them.
The ability of NAc HDAC5 to limit future drug, but not sucrose, seeking is very interesting. Though effect sizes reported vary, heroin self-administration can increase dopamine levels in the NAc by 150 to 300% (39), whereas sucrose increases it only by about 20% in nondeprived rats (40). Sucrose taking and seeking produce different activity levels of NAc MSNs (41) and alter distinct neuronal ensembles compared to addictive drugs, like cocaine (29). The dopamine-sensitive action potentials produced by sucrose might fall below the threshold where HDAC5 is able to suppress firing rates, whereas longer duration and higher amplitude dopamine-modulated glutamatergic responses during heroin or cocaine self-administration might produce firing in a range where HDAC5’s effects are observed (i.e., >100 pA current; Fig. 5 A and D). This might suggest that sucrose reward memories engage a distinct plasticity mechanism and/or neural circuit than those engaged by addictive drugs; however, future experiments will be needed to test these speculative hypotheses. Future studies of the NAc circuit, cellular, and/or synaptic plasticity mechanism(s) comparing drug and natural reward memory formation may be important to fully understand the differential effects of HDAC5.
In conclusion, our findings reveal that the epigenetic enzyme, HDAC5, functions in both D1- and D2-MSNs during active heroin taking to limit the formation of drug, but not sucrose, memories that can trigger relapse-associated drug seeking following abstinence. As such, targeting HDAC5, or its relevant downstream target genes, might be beneficial to limit addiction liability during opioid use. Moreover, we’ve uncovered a potential mechanism in the NAc by which addictive substances, but not natural rewards, encode lasting memories linking the drug experience with diffuse and discreet cues in the drug-taking environment, which often act as powerful triggers of relapse in individuals suffering from opioid or stimulant use disorders. Our findings suggest that HDAC5 governs a program of gene expression, which is enriched for neuronal excitability genes, and that limiting the excitability of both D1- and D2-MSNs in the NAc during strong activity epochs, like drug use, limits the strength of drug-related associations. These data suggest that D1- and D2-MSN functions might play complementary roles in at least some drug-related behaviors. Finally, our data suggest that an epigenetic factor impinges on a common mechanism controlling drug-induced changes in NAc function that contribute to relapse vulnerability.
Materials and Methods
Animals and Animal Care.
Adult rats were singly housed in a climate-controlled environment (21 °C) on a 12-h light–dark cycle. Rats were either male Sprague-Dawley wild-type rats (Charles River) or both male and female D1-Cre/D2-Cre transgenic rats (Long-Evans background, National Institute on Drug Abuse (NIDA) transgenic rat program) bred at the Medical University of South Carolina (MUSC). All rats were habituated to the housing environment for at least 7 d prior to use in experiments and had food and water ad libitum except during the acquisition of heroin self-administration as detailed below. All behavioral experiments were performed during the dark cycle as described below and were approved by the MUSC Institutional Animal Care and Use Committee in facilities accredited by the American Association for the Accreditation of Laboratory Animal Care. All procedures were conducted in accordance with the guidelines established by the NIH and the National Research Council.
Viral Vector Cloning and Packaging.
HDAC5 mRNA was knocked down using an neurotropic adeno-associated vector serotype 2 containing a previously published short hairpin RNA (42) (AAV2-shHDAC5, sequence: GAAGGTTCTACAGAGAGCGAG, University of South Carolina (UofSC) vector core, titer: 1.2 × 1012) driven by a U6 promoter. The control was a short hairpin to luciferase with no known homology in the rat (AAV2-shLuc, CTTACGCTGAGTACTTCGA). Both viruses also expressed GFP under a cytomegalovirus (CMV) promoter. The same AAV2-HDAC5-3SA vector (promoter: CMV) which expresses a “nuclear HDAC5” as previously described (4) was used and compared to an AAV2-GFP control (promoter: CMV). A cre-dependent version of this mutated HDAC5-3SA viral vector (promoter: CMV) with an N-terminal Flag epitope-tag was made by plasmid cloning using the Addgene plasmid pAAV-Ef1a-DIO-enhanced GFP-WPRE-pA (RRID: Addgene_37084) as a template (WPRE = woodchuck hepatitis virus post-transcriptional regulatory element). The control was an AAV2-DIO-GFP (promoter: CMV). All viruses were packaged by the UofSC viral vector core.
Stereotaxic Surgery.
Rats underwent isoflurane-anesthetized survival surgery to microinject viral vectors bilaterally into the NAc (Anterior-Posterior (AP): +1.7, Dorsal-Ventral (DV): −7.7, Medial-Lateral (ML): ±1.0, no angle), and all rats in the study were allowed at least 5 d of recovery before any additional experimentation began. Ketophen (5 mg/kg) was used for post-surgical pain relief. For all behavioral studies, brains were fixed and sliced as described below to verify placements under blinded conditions. All AAVs still expressed at the end of the behavioral experiments as confirmed by the presence of GFP and/or HDAC5 expression as detailed below. Only rats with bilateral NAc viral expression were included in the analysis.
qPCR.
mRNA was extracted using QIAzol Lysis Reagent (Qiagen, cat:56008534) and the RNeasy Micro Kit (Qiagen, cat:74004). Copy deoxyribonucleic acid(cDNA) libraries were made with Superscript 3 kit (Invitrogen). qPCR was performed using a Biorad CFX96 using HDAC5 primers (forward: CTCTGGTCCAAAGAAGCATGATGG; reverse: CGGCCTCAACCATTCCCTCCCACAGC) and normalized to Glyceraldehyde-3-Phosphate Dehydrogenase (GAPDH) (forward: AGGTCGGTGTGAACGGATTTG; reverse: TGTAGACCATGTAGTTGAGGTCA).
Heroin Self-Administration.
Rats underwent anesthetized survival surgery to implant a chronic indwelling jugular vein catheter and all rats in the study were allowed at least 5 d of recovery before any additional experimentation began. Ketophen (5 mg/kg) was used for postsurgical pain relief. All self-administration experiments occurred in standard operant chambers with two retractable levers, a house light, and a cue light and tone generator (Med Associates). All paired lever presses, including those during the timeout, were recorded and are reported as “paired lever presses.” Heroin self-administered rats were fasted before each heroin self-administration session (food was restricted to maintain 85 to 90% of the rats' free-feeding weight, about 20 g of food per day, to allow for lever baiting with food pellets and/or lever extenders as needed during the early phases of acquisition). Rats were then put back on ad libitum food for the remainder of the experiment. During 3-h sessions, rats were trained to press the paired lever on a fixed ratio 1 schedule with 20 s timeout for an infusion of heroin-hydrochloride (100 µg/infusion for days 1 to 2, 50 µg/4 s infusion for days 3 to 4, 25 µg/infusion from day 5 on). Concurrent with the drug infusion, a cue tone and cue light immediately above the paired lever turned on. A rat reached criteria when they had self-administered at least 10 infusions of heroin for at least 12 d at any heroin dose. Patency was regularly ensured via methohexital infusions into the catheter and any rats that showed no response was removed from the study or given a second catheterization surgery and allowed to continue to self-administer heroin. Typically, after either a 7- or 14-d forced abstinence period depending on the experiment (see figure timelines), rats began extinction training for at least 6 d (a model of drug-seeking behavior) and until a criterion of fewer than 25 presses per session for 2 d was met. The D1-cre transgenic rats used in this study never reached this extinction criterion and were instead cut-off after 3 wk of extinction. Paired lever presses produced no drug infusion or light/tone cues. Next, cue-induced reinstatement behavior (a model of relapse-associated behavior) was tested on a subsequent day by allowing the rats to respond for light and tone cues without receiving heroin injections for a single 3 h session. Following at least three more extinction sessions and until responses were less than 25 presses per session for 2 d, rats were injected with a 0.25 mg/kg (i.p.) priming dose of heroin before an extinction session to test for prime-induced reinstatement (another model of relapse-associated behavior). These procedures were based on previous experiments from the Kalivas Lab (43) to maximize comparisons across labs within the MUSC Center for Opioid and Cocaine Addiction.
Sucrose Self-Administration.
Sucrose self-administration procedures were similar as the heroin self-administration experiments above except rodents received sucrose pellet rewards, underwent no intravenous catheter insertion surgery, and had no food restriction and all sessions were 1 h in length. All rats had 10 d of sucrose self-administration, 7 d of forced abstinence, at least 6 d of extinction and until responses were less than 25 presses per session for 2 d. During extinction, paired lever presses produced no sucrose pellets or light/tone cues. Next, cue-induced reinstatement behavior was tested on a subsequent day by allowing the rats to respond for light and tone cues for a single 1 h session. Following at least three more extinction sessions and until responses were less than 25 presses per session for 2 d, rats received five noncontingent sucrose pellets at the beginning of an extinction session to test for sucrose prime-induced reinstatement.
Cell Culture and Plasmid Transfection.
Human embryonic kidney cells (HEK293 cells) were cultured in Dulbecco’s modified Eagle medium containing 10% (v/v) Fetal Bovine Serum, penicillin-streptomycin (1×; Sigma), and L-glutamine (4 mM; Sigma). The plasmids pAAV-DIO-HDAC5-3SA and CAG-cre were transfected using the calcium phosphate method and harvested 1 d later for western blotting.
Western Blotting.
Tissue was lysed, immunoblotted according to previously published methods (4), and analyzed by western blot for HDAC5. In brief, the samples were homogenized using a tissue lysis buffer [5 mM HEPES (pH 7.4), 1% sodium dodecyl sulfate, 0.32 M sucrose, 10 mM NaF, and protease inhibitors in H2O], sonicated, and incubated at 98 °C for 10 min, and the protein concentration was estimated using the Bio-Rad detergent compatible Protein Assay. Then, 20 μg/μL of protein was diluted in 4× sample buffer (SAB), incubated at 98 °C for 10 min, and run on BioRad 4 to 20% Tris-Glycine eXtended (TGX) precast gels at 60 V for 10 min and then 250 V for 25 min. The gels were then transferred onto a BioRad Trans-Blot Turbo Transfer Pack (midi format, 0.20 μm polyvinylidene difluoride (PVDF), cat: 1704157) using the BioRad Trans-Blot Turbo Transfer System and run at 2.5 A, 25 V for 10 min and placed in Odyssey blocking solution for 1 h. Primary antibody: anti-HDAC5 (Research Resource Identifier (RRID): AB_880357, cat: ab50001, Abcam, mouse, 1:500). Secondary antibody: 800CW anti-mouse (RRID: AB_621842, cat: 926-32210, LI-COR, goat, 1:10,000). The use of fluorescent antibodies allowed for the quantification of HDAC5 and tubulin protein levels to all be performed on the same western blot to avoid variability between preparations. Blots were developed on a LI-COR Odyssey CLx and analyzed with ImageStudio.
Immunohistochemistry.
Brains from virus-infused rats were fixed in 4% paraformaldehyde. Following a 24 h post-fix, brains were cryoprotected with 30% sucrose then sliced at 60 μm. Tissue was blocked in buffer (3% bovine serum albumen, 1.5% normal donkey serum, 0.2% Triton-X, 0.2% Tween-20 in phosphate-buffered saline (PBS)) for at least 1 h and then transferred to new buffer with anti-GFP (RRID: AB_10000240, cat: GFP-1020, Aves, chicken, 1:4,000), anti-HDAC5 (Cell Signaling Technology Cat# 2082, RRID: AB_2116626, rabbit), anti-cre (Millipore MAB3120 clone 2D8, RRID: AB_2085748, mouse), anti-cre (Synaptic Systems Cat# 257 004, RRID: AB_2782969, guinea pig), or anti-Flag (Millipore, F1804, RRID: AB_262044, mouse). The next day, tissue was washed 3× 5 min, and anti-chicken, anti-guinea pig, or anti-mouse secondary was added for 1 h [(RRID: AB_2340375, cat: 703-545-155, 488 donkey anti-chicken, Jackson, 1:500) or (RRID: AB_2340854, 715-585-150, 594 donkey anti-mouse, Jackson, 1:500) or (anti-mouse 488, Thermo Fisher Scientific Cat# A-21202, RRID: AB_141607) or (anti-guinea pig Cy3, Jackson ImmunoResearch Labs Cat# 706-165-148, RRID: AB_2340460)]. Tissue was washed in bisbenzimide (1:5,000, Hoechst 33342, Invitrogen) for 2 min, followed by 2× 5 min PBS washes, and then mounted. Images were taken with a Nikon Eclipse 80i fluorescent microscope or the Leica Thunder 3-dimensional (3D) Tissue Imager and processed with ImageJ (RRID: SCR_002285, Fiji, NIH).
Nuclear/Cytoplasmic Ratio.
To quantify the nuclear/cytoplasmic ratio of HDAC5, HDAC5 antibody signal within the nucleus (as defined by the Hoechst stain) was compared to the signal in the cytoplasm using signal intensity in ImageJ. To test the effects of heroin on HDAC5 localization, a Flag-tagged AAV-HDAC5 wild-type virus was injected into the NAc of D1-Cre and D2-Cre rats. After 3 wk, rats were injected with either heroin (2 mg/kg, i.p.) or saline. Four hours after the final injection, the rats were perfused and coronal slices (60 μm thickness) were immunostained for anti-Flag, anti-Cre, and Hoechst (nuclear stain). D1- and D2-cells were identified based on the anti-Cre signal. We analyzed at least six neurons per brain (three cre-positive and three cre-negative) and averaged the nuclear/cytoplasmic ratio for each rat.
vTRAP.
AAV-DIO-L10aGFP was packaged by the UofSC viral vector core using the pAAV-DIO-L10aGFP plasmid (RRID: Addgene_98747). Rats were injected in the NAc with either AAV-DIO-L10aGFP or a 50–50 mixture of AAV-HDAC5-3SA and AAV-L10aGFP. After about 4 wk of expression, NAc tissue was harvested using native GFP fluorescence as a guide. vTrap was performed according to published methods (19).
Microarray Analysis.
RNA from the vTRAP assay was submitted to the Medical University of South Carolina Proteogenomics Facility for DNA microarray analysis (44–46). Briefly, total RNA samples were evaluated by Agilent 2100 Bioanalyzer, and samples were converted to fragmented and biotin-labeled cDNA using 90 ng of total RNA and the GeneChip™ WT PLUS Reagent Kit (ThermoFisher). Hybridizations overnight to Mouse Clariom D microarrays (ThermoFisher) and post-hybridization washing, staining, and scanning were done with GeneChip™ Hybridization, Wash, and Stain Kit, Fluidics Station 450, and 7G scanner according to the manufacturer’s protocols.
Transcriptomic Analysis.
Resulting microarray scan data (.CEL files) were analyzed with R using Oligo package (47). Probesets were background-corrected, and gene expression levels were calculated by Robust MultiArray Average. Protein coding genes were used for downstream analysis. Differential expression analysis was performed using limma (48) accounting for technical and biological confounding factors as follows:
Gene expression ~ Genotype + Batch + Sex.
We calculated DEGs using criteria of false discovery rate (FDR) ≤ 0.05 and log2 (fold change) ≥ |0.3|. The package clusterProfiler was used for the gene ontology analysis (49).
qPCR Validation.
Select targets were analyzed as above using the following primers: akap5 (forward: GACCGCAACGATCTGGGTAA, reverse: GGGGCTCAGTGCATCAGATT), kcnv1 (forward: GCGAGTACTAGAAGCAGTGGT, reverse: CCATAGCAGCAAGGCTTCATC) and normalized to GAPDH (forward: AGGTCGGTGTGAACGGATTTG; reverse: TGTAGACCATGTAGTTGAGGTCA).
Electrophysiology.
All acute-slice electrophysiological experiments were performed in D1-cre and D2-cre Long-Evans rats following an AAV surgery at least 3 wk before. Acute coronal slices (300-µm thickness) containing the NAc were prepared in a semifrozen artificial cerebrospinal fluid (ACSF) containing (in mM) 127 NaCl, 2.5 KCl, 1.2 NaH2PO4, 24 NaHCO3, 11 D-glucose, 1.2 MgCl2, and 2.40 CaCl2, 0.4 Na-ascorbate (pH 7.4, 315 to 320 mOsm). Kynurenic acid (5 mM) was added to ACSF to avoid overactivity of glutamatergic receptors during slicing. Slices were transferred to ACSF without kynurenic acid to recover at 37 °C for 30 min and then transferred to room temperature ACSF for an additional 30 min prior to recording. All solutions were continually equilibrated with 95% O2 and 5% CO2 prior to and during the slicing procedure.
Action potentials were recorded in whole cell current clamp mode and transmembrane current was clamped, at −70 mV. Current was injected at 20 pA steps from 0 to 500 pA. Each step was held for 1,000 ms, and the number of action potentials at each step was counted.
AMPA currents, NMDA currents, and PPR experiments were performed in whole cell voltage clamp mode using electrodes with a resistance of 4 to 6 MΩ pulled using a NARISHIGE puller (NARISHIGE, PG10) from borosilicate tubing (Sutter Instruments) and filled by an internal solution containing (in mM) 140 CsMetSO4, 5 KCl, 1 MgCl2, 0.2 egtazic acid (EGTA), 11 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), 2 Sodium adenosine triphosphate (NaATP), 0.2 Na2GTP, and 0.1 CaCl2 (pH 7.2, 290 to 295 mOsm). AMPA and NMDA receptor-mediated excitatory postsynaptic currents (AMPA- and NMDA-responses) were recorded, respectively, at −70 and +50 mV. Recordings were made in the presence of picrotoxin (50 µM) to block inhibitory postsynaptic currents mediated by gamma-Aminobutyric acid type A (GABA-A) receptors. The amplitude of AMPA responses was calculated at the maximum current value, and the amplitude of the NMDA response was calculated at 50 ms after stimulation. These values were used to calculate the AMPA/NMDA ratio. For PPR measurements, two excitatory postsynaptic potentials (EPSPs) were evoked at −70 mV with an interstimulus interval of 50 ms. The peak amplitude of the second EPSC (P2) was divided by the peak of the first amplitude (P1) to generate the PPR (P2/P1).
All data (Recordings) were acquired and analyzed by amplifier AXOPATCH 200B (Axon Instruments), digitizer BNC2090 (National instruments), and software AxoGraph v1.7.0, Clampfit v8.0 (pClamp, Molecular devices) and MiniAnalysis Program v6.0.9 (Synaptosoft). Data were filtered at 2 kHz by an AXOPATCH 200B amplifier (Axon Instruments) and digitized at 20 kHz via AxoGraph v1.7.0.
Statistics.
T tests and repeated measures two-way ANOVAs were used to analyze data where appropriate. Fishers least significant difference post hoc tests were used following significant interactions using ANOVAs. Grubbs’ test was used to identify single statistical outliers. All statistics were performed with GraphPad Prism 8, and P < 0.05 was considered significant. All statistics are reported in SI Appendix, Table S1.
Data, Materials, and Software Availability
Microarray data collected for this study have been deposited in NCBI Gene Expression Omnibus in accordance with MIAME conventions (50) under accession number GSE199985. For this manuscript draft, these data can be accessed here (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?&acc=GSE199985) using the token whydoisylnynfoh. Custom R codes and data to support the analysis, visualizations, functional, and gene set enrichments are available at https://github.com/BertoLabMUSC/AndersonEtAl_HDAC5.
Acknowledgments
We would like to thank Eric Dereschewitz, Deanna L. Miller, Jack Wolf, Mark Geesey, Kelsey Vollmer, Kyle Simpson, Danielle Dixon, Emma Russell, Carmela Reichel, and John Woodward for advice and technical assistance during the study. This work was supported by NIH grants, P50 DA046373 (M.T. and C.W.C.), R01 DA032708 (C.W.C.), K01 DA046513 (E.M.A.), R01 DA054589 (A.L.), and F32 DA047845 (S.Barry). The MUSC Proteogenomics Facility is supported by NIH Grant GM103499 and MUSC's Office of the Vice President for Research.
Author contributions
E.M.A., M.T., and C.W.C. designed research; E.M.A., E.T., A.G., D.D., L.M.M., D.W., S.Barry, and M.T. performed research; E.M.A., E.T., A.G., S.Berto, A.L., and M.T. analyzed data; and E.M.A. and C.W.C. wrote the paper.
Competing interests
The authors declare no competing interest.
Supporting Information
Appendix 01 (PDF)
- Download
- 16.56 MB
References
1
N. D. Volkow, C. Blanco, The changing opioid crisis: Development, challenges and opportunities. Mol. Psychiatry 26, 218–233 (2021).
2
C. D. Allis, T. Jenuwein, The molecular hallmarks of epigenetic control. Nat. Rev. Genet. 17, 487–500 (2016).
3
E. M. Anderson et al., It’s a complex issue: Emerging connections between epigenetic regulators in drug addiction. Eur. J. Neurosci. 50, 2477–2491 (2019).
4
M. Taniguchi et al., HDAC5 and its target gene, Npas4, function in the nucleus accumbens to regulate cocaine-conditioned behaviors. Neuron 96, 130–144.e136 (2017).
5
M. Taniguchi et al., Histone deacetylase 5 limits cocaine reward through cAMP-induced nuclear import. Neuron 73, 108–120 (2012).
6
W. Renthal et al., Histone deacetylase 5 epigenetically controls behavioral adaptations to chronic emotional stimuli. Neuron 56, 517–529 (2007).
7
A. C. W. Smith, P. J. Kenny, HDAC5 regulates the formation of drug memories. Trends Mol. Med. 24, 106–108 (2018).
8
E. A. Griffin Jr. et al., Prior alcohol use enhances vulnerability to compulsive cocaine self-administration by promoting degradation of HDAC4 and HDAC5. Sci. Adv. 3, e1701682 (2017).
9
G. Egervari et al., Striatal H3K27 acetylation linked to glutamatergic gene dysregulation in human heroin abusers holds promise as therapeutic target. Biol. Psychiatry 81, 585–594 (2017).
10
L. Haery et al., Adeno-associated virus technologies and methods for targeted neuronal manipulation. Front. Neuroanat. 13, 93 (2019).
11
X. Li et al., Role of dorsal striatum histone deacetylase 5 in incubation of methamphetamine craving. Biol. Psychiatry 84, 213–222 (2018).
12
J. Bertran-Gonzalez et al., Opposing patterns of signaling activation in dopamine D1 and D2 receptor-expressing striatal neurons in response to cocaine and haloperidol. J. Neurosci. 28, 5671–5685 (2008).
13
M. K. Lobo et al., Cell type-specific loss of BDNF signaling mimics optogenetic control of cocaine reward. Science 330, 385–390 (2010).
14
M. K. Lobo, E. J. Nestler, The striatal balancing act in drug addiction: Distinct roles of direct and indirect pathway medium spiny neurons. Front. Neuroanat. 5, 41 (2011).
15
H. S. Bateup et al., Distinct subclasses of medium spiny neurons differentially regulate striatal motor behaviors. Proc. Natl. Acad. Sci. U.S.A. 107, 14845–14850 (2010).
16
C. Garcia-Keller et al., Relapse-associated transient synaptic potentiation requires integrin-mediated activation of focal adhesion kinase and cofilin in D1-expressing neurons. J. Neurosci. 40, 8463–8477 (2020).
17
J. M. Leyrer-Jackson, P. F. Overby, A. Bull, J. A. Marusich, C. D. Gipson, Strain and sex matters: Differences in nicotine self-administration between outbred and recombinase-driver transgenic rat lines. Exp. Clin. Psychopharmacol. 29, 375–384 (2021).
18
K. T. Schmidt et al., The effects of strain and estrous cycle on heroin- and sugar-maintained responding in female rats. Behav. Brain Res. 409, 113329 (2021).
19
M. Heiman, R. Kulicke, R. J. Fenster, P. Greengard, N. Heintz, Cell type-specific mRNA purification by translating ribosome affinity purification (TRAP). Nat. Protoc. 9, 1282–1291 (2014).
20
J. M. Bossert, G. C. Poles, K. A. Wihbey, E. Koya, Y. Shaham, Differential effects of blockade of dopamine D1-family receptors in nucleus accumbens core or shell on reinstatement of heroin seeking induced by contextual and discrete cues. J. Neurosci. 27, 12655–12663 (2007).
21
J. A. Heinsbroek et al., Loss of plasticity in the D2-accumbens pallidal pathway promotes cocaine seeking. J. Neurosci. 37, 757–767 (2017).
22
D. W. Self, W. J. Barnhart, D. A. Lehman, E. J. Nestler, Opposite modulation of cocaine-seeking behavior by D1- and D2-like dopamine receptor agonists. Science 271, 1586–1589 (1996).
23
H. D. Schmidt, S. M. Anderson, R. C. Pierce, Stimulation of D1-like or D2 dopamine receptors in the shell, but not the core, of the nucleus accumbens reinstates cocaine-seeking behaviour in the rat. Eur. J. Neurosci. 23, 219–228 (2006).
24
P. I. Ortinski, L. A. Briand, R. C. Pierce, H. D. Schmidt, Cocaine-seeking is associated with PKC-dependent reduction of excitatory signaling in accumbens shell D2 dopamine receptor-expressing neurons. Neuropharmacology 92, 80–89 (2015).
25
R. K. Bachtell, K. Whisler, D. Karanian, D. W. Self, Effects of intra-nucleus accumbens shell administration of dopamine agonists and antagonists on cocaine-taking and cocaine-seeking behaviors in the rat. Psychopharmacology 183, 41–53 (2005).
26
F. W. Hopf, M. G. Cascini, A. S. Gordon, I. Diamond, A. Bonci, Cooperative activation of dopamine D1 and D2 receptors increases spike firing of nucleus accumbens neurons via G-protein betagamma subunits. J. Neurosci. 23, 5079–5087 (2003).
27
T. R. Pardo-Garcia et al., Ventral pallidum is the primary target for accumbens D1 projections driving cocaine seeking. J. Neurosci. 39, 2041–2051 (2019).
28
V. C. Chioma et al., Heroin seeking and extinction from seeking activate matrix metalloproteinases at synapses on distinct subpopulations of accumbens cells. Biol. Psychiatry 89, 947–958 (2021).
29
A. C. Bobadilla et al., Cocaine and sucrose rewards recruit different seeking ensembles in the nucleus accumbens core. Mol. Psychiatry 25, 3150–3163 (2020).
30
D. Roberts-Wolfe, A. C. Bobadilla, J. A. Heinsbroek, D. Neuhofer, P. W. Kalivas, Drug refraining and seeking potentiate synapses on distinct populations of accumbens medium spiny neurons. J. Neurosci. 38, 7100–7107 (2018).
31
V. Pascoli et al., Contrasting forms of cocaine-evoked plasticity control components of relapse. Nature 509, 459–464 (2014).
32
Y. Iino et al., Dopamine D2 receptors in discrimination learning and spine enlargement. Nature 579, 555–560 (2020).
33
S. Ikemoto, B. S. Glazier, J. M. Murphy, W. J. McBride, Role of dopamine D1 and D2 receptors in the nucleus accumbens in mediating reward. J. Neurosci. 17, 8580–8587 (1997).
34
E. E. Steinberg et al., Positive reinforcement mediated by midbrain dopamine neurons requires D1 and D2 receptor activation in the nucleus accumbens. PLoS One 9, e94771 (2014).
35
S. L. Cole, M. J. F. Robinson, K. C. Berridge, Optogenetic self-stimulation in the nucleus accumbens: D1 reward versus D2 ambivalence. PLoS One 13, e0207694 (2018).
36
M. T. Stefanik et al., Optogenetic inhibition of cocaine seeking in rats. Addict. Biol. 18, 50–53 (2013).
37
T. J. O’Neal, M. N. Nooney, K. Thien, S. M. Ferguson, Chemogenetic modulation of accumbens direct or indirect pathways bidirectionally alters reinstatement of heroin-seeking in high- but not low-risk rats. Neuropsychopharmacology 45, 1251–1262 (2020).
38
M. T. Stefanik, Y. M. Kupchik, R. M. Brown, P. W. Kalivas, Optogenetic evidence that pallidal projections, not nigral projections, from the nucleus accumbens core are necessary for reinstating cocaine seeking. J. Neurosci. 33, 13654–13662 (2013).
39
R. A. Wise, P. Leone, R. Rivest, K. Leeb, Elevations of nucleus accumbens dopamine and DOPAC levels during intravenous heroin self-administration. Synapse 21, 140–148 (1995).
40
W. M. Doyon, V. Ramachandra, H. H. Samson, C. L. Czachowski, R. A. Gonzales, Accumbal dopamine concentration during operant self-administration of a sucrose or a novel sucrose with ethanol solution. Alcohol 34, 261–271 (2004).
41
L. Canchy, P. Girardeau, A. Durand, C. Vouillac-Mendoza, S. H. Ahmed, Pharmacokinetics trumps pharmacodynamics during cocaine choice: A reconciliation with the dopamine hypothesis of addiction. Neuropsychopharmacology 46, 288–296 (2021).
42
X. Li et al., Role of dorsal striatum histone deacetylase 5 in incubation of methamphetamine craving. Biol. Psychiatry 84, 213–222 (2018), https://doi.org/10.1016/j.biopsych.2017.12.008.
43
H. Shen, P. W. Kalivas, Reduced LTP and LTD in prefrontal cortex synapses in the nucleus accumbens after heroin self-administration. Int. J. Neuropsychopharmacol. 16, 1165–1167 (2013).
44
J. L. Barth et al., Jarid2 is among a set of genes differentially regulated by Nkx2.5 during outflow tract morphogenesis. Dev. Dyn. 239, 2024–2033 (2010).
45
H. Lang et al., Neural stem/progenitor cell properties of glial cells in the adult mouse auditory nerve. Sci. Rep. 5, 13383 (2015).
46
C. H. Panganiban et al., Noise-induced dysregulation of quaking rna binding proteins contributes to auditory nerve demyelination and hearing loss. J. Neurosci. 38, 2551–2568 (2018).
47
B. S. Carvalho, R. A. Irizarry, A framework for oligonucleotide microarray preprocessing. Bioinformatics 26, 2363–2367 (2010).
48
M. E. Ritchie et al., limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).
49
G. Yu, L. G. Wang, Y. Han, Q. Y. He, clusterProfiler: An R package for comparing biological themes among gene clusters. OMICS 16, 284–287 (2012).
50
R. Edgar, T. Barrett, NCBI GEO standards and services for microarray data. Nat. Biotechnol. 24, 1471–1472 (2006).
Information & Authors
Information
Published in
Classifications
Copyright
Copyright © 2023 the Author(s). Published by PNAS. This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).
Data, Materials, and Software Availability
Microarray data collected for this study have been deposited in NCBI Gene Expression Omnibus in accordance with MIAME conventions (50) under accession number GSE199985. For this manuscript draft, these data can be accessed here (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?&acc=GSE199985) using the token whydoisylnynfoh. Custom R codes and data to support the analysis, visualizations, functional, and gene set enrichments are available at https://github.com/BertoLabMUSC/AndersonEtAl_HDAC5.
Submission history
Received: June 25, 2022
Accepted: January 6, 2023
Published online: February 6, 2023
Published in issue: February 14, 2023
Keywords
Acknowledgments
We would like to thank Eric Dereschewitz, Deanna L. Miller, Jack Wolf, Mark Geesey, Kelsey Vollmer, Kyle Simpson, Danielle Dixon, Emma Russell, Carmela Reichel, and John Woodward for advice and technical assistance during the study. This work was supported by NIH grants, P50 DA046373 (M.T. and C.W.C.), R01 DA032708 (C.W.C.), K01 DA046513 (E.M.A.), R01 DA054589 (A.L.), and F32 DA047845 (S.Barry). The MUSC Proteogenomics Facility is supported by NIH Grant GM103499 and MUSC's Office of the Vice President for Research.
Author contributions
E.M.A., M.T., and C.W.C. designed research; E.M.A., E.T., A.G., D.D., L.M.M., D.W., S.Barry, and M.T. performed research; E.M.A., E.T., A.G., S.Berto, A.L., and M.T. analyzed data; and E.M.A. and C.W.C. wrote the paper.
Competing interests
The authors declare no competing interest.
Notes
This article is a PNAS Direct Submission.
Authors
Metrics & Citations
Metrics
Altmetrics
Citations
Cite this article
Epigenetic function during heroin self-administration controls future relapse-associated behavior in a cell type-specific manner, Proc. Natl. Acad. Sci. U.S.A.
120 (7) e2210953120,
https://doi.org/10.1073/pnas.2210953120
(2023).
Copied!
Copying failed.
Export the article citation data by selecting a format from the list below and clicking Export.
Cited by
Loading...
View Options
View options
PDF format
Download this article as a PDF file
DOWNLOAD PDFLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginRecommend to a librarian
Recommend PNAS to a LibrarianPurchase options
Purchase this article to access the full text.