Address correspondence and reprint requests to Kristen A. Keefe - 30 South 2000 East, Room 201 L.S. Skaggs Hall, Salt Lake City, UT 84112, USA. E-mail: email@example.com
The immediate-early gene Arc (activity-regulated cytoskeleton-associated protein) is provocative in the context of neuroplasticity because of its experience-dependent regulation and mRNA transport to and translation at activated synapses. Normal rats have more preproenkephalin-negative (ppe-neg; presumed striatonigral) neurons with cytoplasmic Arc mRNA than ppe-positive (ppe-pos; striatopallidal) neurons, despite equivalent numbers of these neurons showing novelty-induced transcriptional activation of Arc. Furthermore, rats with partial monoamine loss induced by methamphetamine (METH) show impaired Arc mRNA expression in both ppe-neg and ppe-pos neurons relative to normal animals following response-reversal learning. In this study, Arc expression induced by exposure to a novel environment was used to assess transcriptional activation and cytoplasmic localization of Arc mRNA in striatal efferent neuron subpopulations subsequent to METH-induced neurotoxicity. Partial monoamine depletion significantly altered Arc expression. Specifically, basal Arc expression was elevated, but novelty-induced transcriptional activation was abolished. Without novelty-induced Arc transcription, METH-pre-treated rats also had fewer neurons with cytoplasmic Arc mRNA expression, with the effect being greater for ppe-neg neurons. Thus, METH-induced neurotoxicity substantially alters striatal efferent neuron function at the level of Arc transcription, suggesting a long-term shift in basal ganglia neuroplasticity processes subsequent to METH-induced neurotoxicity. Such changes potentially underlie striatally based learning deficits associated with METH-induced neurotoxicity.
activity-regulated cytoskeleton-associated protein
caged control rats
extracellular signal-regulated kinase
Methamphetamine (METH) is a highly addictive psychostimulant that can induce substantial forebrain dopamine (DA) and serotonin depletions when administered at high doses (Wagner et al. 1980; Woolverton et al. 1989). Although METH-induced monoamine depletions are significantly less than those observed in Parkinson's disease, these partial depletions lead to impairments in basal ganglia-mediated behavioral and cognitive abilities in both rodents and humans (Chapman et al. 2001; Volkow et al. 2001; Kalechstein et al. 2003; Johanson et al. 2006). Furthermore, recent data show that individuals with a history of methamphetamine or other amphetamine abuse are more likely to develop Parkinsonism (Callaghan et al. 2012), suggesting that METH exposure may be associated with a significant pre-clinical period of partial striatal DA loss and associated consequences.
The bases for the cognitive and behavioral sequelae of METH-induced partial DA loss remain unknown. Approximately, 95% of neurons in striatum are medium spiny neurons, and these efferent neurons can be divided into two equal subpopulations: striatonigral ‘direct’ pathway neurons and striatopallidal ‘indirect’ pathway neurons (Wichmann and DeLong 1996; Obeso et al. 1997; Grillner et al. 2005). These efferent neurons can be phenotypically differentiated at the cell body level by their selective expression of neuropeptides: striatopallidal neurons express preproenkephalin (ppe-pos), whereas striatonigral neurons do not [ppe-neg (Gerfen and Young 1988)]. Striatonigral neurons predominately express D1-DA receptors, whereas striatopallidal neurons predominately express D2-DA receptors. Prior work suggests that METH-induced functional impairments are substantially more pronounced in striatonigral neurons (Chapman et al. 2001; Johnson-Davis et al. 2002; Daberkow et al. 2008), consistent with findings reported for rats with partial DA loss induced by 6-hydroxydopamine (Nisenbaum et al. 1996). Such changes may be because, in part, of impairments in phasic DA signaling associated with the METH-induced neurotoxicity (Howard et al. 2011), as the D1-DA receptor may be particularly sensitive to such changes (Floresco et al. 2003; Dreyer et al. 2010).
The immediate-early gene Arc/Arg3.1 (activity-regulated cytoskeleton-associated protein) is a critical mediator of normal neuroplasticity processes and is induced by N-methyl-d-asparate (NMDA)-receptor activation before being trafficked to activated synapses, where it is locally translated (Lyford et al. 1995; Bramham et al. 2008). Arc mRNA expression is induced in an experience-dependent manner (Guzowski et al. 1999; Daberkow et al. 2007) within GABAergic striatal neurons (Vazdarjanova et al. 2006). Arc protein itself is critical to modulating excitatory glutamatergic inputs via endocytosis of α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA)-type glutamate receptors (Chowdhury et al. 2006; Rial Verde et al. 2006) and to maintaining synaptic plasticity via modulation of long-term potentiation (LTP) and long-term depression (LTD) (Bloomer et al. 2008; Bramham et al. 2008). Blocking Arc translation via antisense oligonucleotide infusions impairs consolidation of learning (Guzowski et al. 2000), including basal ganglia-mediated learning (Hearing et al. 2011; Pastuzyn et al. 2012). Thus, Arc is critical for consolidation of learning, including those processes mediated by dorsal striatum.
We have previously reported that despite equivalent numbers of neurons showing transcriptional activation of the Arc gene, there are more striatonigral neurons with Arc mRNA in the cytoplasm than striatopallidal neurons in normal animals following exploration of a novel environment (Daberkow et al. 2007) – the prototypical behavioral paradigm used to assess Arc expression (Guzowski et al. 1999; Chawla et al. 2005; Vazdarjanova et al. 2006). Furthermore, we have previously reported that rats with prior exposure to a neurotoxic regimen of METH show decreased numbers of striatal neurons with cytoplasmic Arc mRNA expression after engaging in a striatally-mediated, response-reversal learning task, with the greatest impairment being in the numbers of striatonigral neurons (Daberkow et al. 2008). Also, the correlation between Arc mRNA expression in dorsomedial (DM) striatum and trials to criterion on the response-reversal learning task observed in normal animals (Daberkow et al. 2007, 2008) is lost in METH-pre-treated animals (Daberkow et al. 2008). Furthermore, reversal learning in METH-pre-treated rats is no longer sensitive to antisense oligonucleotide-mediated disruption of Arc translation in DM striatum (Pastuzyn et al. 2012). However, the paradigms examined to date do not allow us to discern whether the decreased numbers of striatal neurons with Arc mRNA expression in the cytoplasm in METH-pre-treated rats arises because of deficits in transcriptional activation of the Arc gene or deficits in the trafficking of Arc mRNA into the cytoplasm. Thus, the goal of the present studies was to determine whether METH-induced neurotoxicity is associated with acute disruption of Arc transcriptional activation versus longer duration processes regulating Arc trafficking/cytoplasmic stability. Furthermore, an additional goal of the present work was to examine further whether there are phenotypic differences in the acute regulation of Arc mRNA transcriptional activation and cytoplasmic expression following METH-induced neurotoxicity so as to better define how partial DA depletion impacts upon striatal efferent neuron function. Clarifying the impact of METH-induced neurotoxicity on the transcriptional activation and cytoplasmic regulation of Arc mRNA in striatum is critical so as to better inform treatment strategies to ameliorate cognitive and behavioral impairments in human METH abusers (Volkow et al. 2001; Kalechstein et al. 2003; Johanson et al. 2006).
Male Sprague-Dawley rats (Charles River Laboratories, Raleigh, NC, USA; 275–300 g) were singly housed in tub cages in a room controlled for temperature and lighting (12 : 12 h). All animal care and experimental procedures conformed to the Guide for the Care and Use of Laboratory Animals (8th Ed.), followed the ARRIVE guidelines (Kilkenny et al. 2010) and were approved by the Institutional Animal Care and Use Committee at the University of Utah.
Rats were treated with (±)-methamphetamine hydrochloride (NIDA, Research Triangle Park, NC, USA) as previously described (Daberkow et al. 2008; Son et al. 2011). This specific regimen was chosen to model METH-induced damage to striatal monoamine systems. Although METH self-administration models may more closely mimic human exposure by self-administration, such contingent models typically do not recapitulate long-lasting striatal DA toxicity (Schwendt et al. 2009; Brennan et al. 2010; McFadden et al. 2012; Reichel et al. 2012), perhaps because of the duration of the exposure, as one study (Krasnova et al. 2010), with much longer daily access (15 h/day for 8 days) did report DA neuron toxicity. Thus, the single day, non-contingent binge regimen implemented in this study serves as a more rapid and reproducible model with face validity to the striatal DA depletions present in individuals with a history of METH exposure and who may, in fact, be in a pre-clinical Parkinsonian state. On the METH treatment day, rats were housed five to six animals per cage. Rats received four injections of either (±)-METH (10 mg/kg free base, s.c.) or 0.9% saline (1 mL/kg, s.c.) at 2-h intervals, with rectal temperatures recorded hourly (Fig. 1a). If core temperature exceeded 41°C, the rat was removed and placed in a separate cage over ice to decrease hyperthermia. Twelve hours after the final injection, rats were returned to home cages and given free access to food and water until behavioral manipulations 3 weeks later.
Behavioral activation via novel environment exploration
Rats were divided into three experimental groups [‘5-min’, ‘30-min’, caged controls (CC)], each consisting of 6–8 saline- or 10–11 METH-pre-treated rats. After initially observing elevated basal Arc expression in CC METH rats, an additional seven rats were treated with METH as described above and killed as CC. Thus, the METH CC group consisted of 18 animals. Rats were removed from their home cage after having been isolated for 24 h. CC rats were killed immediately upon removal from the home cage. Rats in the ‘5-min’ group were exposed to a novel environment [50 × 40 × 40 cm tall Rubbermaid tub with pictures on two walls, as previously described (Daberkow et al. 2007; Guzowski et al. 1999)] for 5 min, then immediately killed. Rats in the ‘30-min’ group were exposed for 5 min to the novel environment, returned to the home cage for 25 min, and then killed. Animals were killed by CO2 exposure, decapitated, and brains immediately removed and flash-frozen in 2-methylbutane (Mallinckrodt Baker, Phillipsburg, NJ, USA) chilled on dry ice.
Twelve-micrometer cryosections of striatum [Cambridge Instruments, Bayreuth, Germany; Bregma: +1.60 to −0.8 mm (Paxinos and Watson, 1998)] were thaw-mounted onto SuperFrost Plus slides (VWR, Batavia, IL, USA), then post-fixed as previously described (Ganguly and Keefe 2001).
METH-induced DA depletions in dorsal striatum were analyzed using [I125]RTI-55 binding to the dopamine transporter (DAT), as described previously (Boja et al. 1992; Pastuzyn et al. 2012). DAT binding levels were normalized to percent of saline controls for DM and dorsolateral (DL) striatum (Fig. 1b).
Fluorescent in situ hybridization for Arc/ppe mRNAs
Expression of Arc mRNA in striatal efferent neurons was determined by double-label fluorescence in situ hybridization histochemistry (FISH) for Arc and ppe mRNAs, as previously described (Daberkow et al. 2007, 2008). Full-length ribonucleotide probes complementary to mRNAs for Arc (Lyford et al. 1995) and ppe (Yoshikawa et al. 1984) were synthesized from cDNAs using digoxigenin-UTP (DIG-UTP) and fluorescein-UTP (FITC-UTP) with T7 and SP6 RNA polymerases and DIG- and FITC-UTP RNA-labeling kits, respectively (Roche Applied Science, Indianapolis, IN, USA). Ribonucleotide probes were hybridized and detected as previously described (Daberkow et al. 2007, 2008).
DAT film autoradiograms were digitized, and four sections per animal in rostral (+1.6 mm from bregma), middle (+0.5 mm from bregma), and caudal (−0.8 mm from bregma) dorsal striatum analyzed with ImageJ (Pastuzyn et al. 2012).
For FISH, a 0.18 mm2 montage from DM striatum was captured with an FV1000 confocal laser-scanning microscope (Olympus, Center Valley, PA, USA) with motorized stage (Prior Scientific, Rockland, MA, USA) using a 60x, 1.45 NA oil-immersion lens (plan APO) and 488-nm Ar and 543-nm and 633-nm HeNe lasers. Areas of analysis were z-sectioned in 1-μm-thick optical sections (10 z-sections per image) (Daberkow et al. 2007, 2008). The numbers of ppe-pos and ppe-neg neurons with Arc mRNA staining in the nucleus and in the cytoplasm were determined, as previously described (Guzowski et al. 1999; Daberkow et al. 2007, 2008).
Analysis of DAT binding was accomplished by two-sample t-tests. METH and saline effects on body temperature were analyzed by manova. Expression of Arc mRNA in each subcellular compartment was first compared across saline- and METH-pre-treated rats using manova (treatment × time × phenotype), although for clarity, Arc expression data are shown in separate, side-by-side graphs for saline- and METH-pre-treated rats (Figs 3, 4). Post hoc analyses of significant interactions were achieved via Tukey–Kramer test for between-subjects factors (treatment, time) or post hoc t-tests for the within-subjects factor (phenotype). For all analyses, p-values < 0.05 were significant.
METH-induced partial dopamine depletions
Treatment with METH resulted in hyperthermia necessary to produce long-term neurotoxicity (Ali et al. 1994). METH-treated rats reached maximum temperatures of 39.7°C (± 0.8°C; Fig. 1a), whereas saline-treated rats achieved maximum temperatures of 36.8°C (± 0.5°C). There were significant main effects of treatment (F(1,54) = 192.9, p < 0.0001) and time (F(3,54) = 102.1, p < 0.0001) and a significant time × treatment interaction (F(3,54) = 113.0, p < 0.0001). At all time points after baseline, core temperatures in METH-treated rats were significantly elevated over saline (baseline, p = 0.3; temp 1-temp 7, p < 0.0001). Three weeks later, DAT binding in DM striatum was 57.2 ± 4.2% of control (mean ± SEM; n = 40 METH, n = 23 saline; t(1,26) = −9.51, p < 0.0001), and that in DL striatum was 65.5 ± 3.7% of control (t(1,26) = −10.14, p < 0.0001; Fig. 1b).
METH-induced neurotoxicity is associated with impaired transcriptional activation of Arc mRNA
Using catFISH (Guzowski et al. 1999; Daberkow et al. 2007, 2008), we determined the subcellular localization of Arc mRNA in striatal efferent neurons following spatial exploration of a novel environment (Fig. 2a–f). As previously reported in normal animals (Guzowski et al. 1999; Daberkow et al. 2007), exposure to the novel environment increased nuclear Arc mRNA expression within 5 min (Figs 2 and 3a). manova revealed significant main effects of time (F(2,54) = 6.0, p < 0.005) and treatment (F(1,54) = 5.96, p < 0.02) and a significant time x treatment interaction (F(2,54) = 4.99, p < 0.02). There was no main effect of phenotype (F(1,54) = 2.72, p = 0.1) and no phenotype × time (F(2,54) = 0.7, p = 0.5), phenotype × treatment (F(1,54) = 0.5, p = 0.5) or phenotype × time × treatment (F(2,54) = 0.6, p = 0.6) interactions. Post hoc analysis of the time × treatment interaction revealed that in saline-pre-treated rats the number of cells with nuclear Arc mRNA expression was significantly greater in the 5-min group than in the CC (p < 0.01) or 30-min (p < 0.001) groups. There were no such effects in METH-pre-treated rats, as the numbers of cells with intranuclear foci of Arc mRNA in the METH-pre-treated groups were not different from each other (p > 0.1). However, overall (main effect of treatment) there were more cells with nuclear Arc mRNA in the METH- versus saline-pre-treated rats (p < 0.02; Fig. 3a vs. b). Thus, rats with METH-induced neurotoxicity have higher basal Arc transcription, but fail to induce Arc transcription in response to behavioral activation.
METH-induced neurotoxicity is associated with impaired cytoplasmic Arc mRNA expression
As previously reported (Daberkow et al. 2007), saline-pre-treated rats in the 30-min group had a significant increase in the number of striatal efferent neurons with Arc mRNA in the cytoplasm (Figs 4, 5). Furthermore, as we have previously reported (Daberkow et al. 2007, 2008), normal animals had more striatonigral (ppe-neg) than striatopallidal (ppe-pos) neurons with cytoplasmic Arc mRNA expression (Figs 4 and 5a). Specifically, manova revealed a significant main effect for time (F(2,54) = 16.9, p < 0.0001) and significant time × treatment (F(2,54) = 3.3, p < 0.05) and phenotype × time × treatment (F(2,54) = 3.8, p < 0.03) interactions. There was no significant main effect of treatment (F(1,54) = 0.7, p = 0.4) or phenotype (F(1,54) = 3.1, p = 0.08) and no significant phenotype × time (F(2,54) = 0.6, p = 0.5) or phenotype × treatment (F(1,54) = 0.8, p = 0.4) interactions. Post hoc analysis of the time × treatment interaction revealed greater numbers of neurons with cytoplasmic Arc mRNA expression in saline-pre-treated, 30-min rats relative to saline-pre-treated, CC (p = 0.002) and 5-min (p < 0.001) groups, consistent with previous reports regarding the time course of normal Arc trafficking (Guzowski et al. 1999; Daberkow et al. 2007). Conversely, there were no significant differences between the numbers of cells with cytoplasmic Arc mRNA expression in the METH-pre-treated, 30-min group relative to the METH-pre-treated, CC (p = 0.6) and 5-min (p = 0.1) groups, again indicating a general overall lack of Arc mRNA induction associated with spatial exploration in METH-pre-treated rats.
To examine the phenotypic differences, post hoc t-tests were performed on the numbers of ppe-neg and ppe-pos neurons with cytoplasmic Arc mRNA in saline- and METH-pre-treated rats. Paired t-tests confirmed (Daberkow et al. 2007) greater numbers of both ppe-neg (t = 3.9, p = 0.003) and ppe-pos (t = 3.9, p = 0.001) neurons with cytoplasmic Arc in the saline-pre-treated, 30-min group relative to saline-pre-treated, CC group (Fig. 4a) and that there were more ppe-neg than ppe-pos neurons with cytoplasmic Arc mRNA expression in the saline-pre-treated, 30-min group (Figs 4a and 5a; t = 2.5, p = 0.02). In METH-pre-treated, CC rats, there was no significant difference from saline-pre-treated, CC rats in the numbers of ppe-neg neurons with cytoplasmic Arc mRNA (Fig. 4; t = −0.28, p > 0.1). Additionally, in METH-pre-treated rats, there was no significant increase in numbers of ppe-neg neurons with cytoplasmic Arc at 30 min (Fig. 4b; t = 0.44, p = 0.31) relative to METH-pre-treated, CC group, whereas there was a significant increase in the numbers of ppe-pos neurons (Fig. 4b; t = 2.2, p = 0.02). Thus, in the METH-pre-treated, 30-min group, there was no longer a phenotypic difference between the numbers of ppe-neg and ppe-pos neurons with cytoplasmic Arc (Figs 4b and 5a; t = −0.95, p = 0.82), but there were significantly fewer ppe-neg neurons with cytoplasmic Arc mRNA expression relative to the saline-pre-treated, 30-min group (Figs 4b and 5a; t = 3.0, p = 0.004). Interestingly, there were more ppe-pos neurons with cytoplasmic Arc mRNA expression in the METH-pre-treated, CC group than in the saline-pre-treated, CC group (Fig. 4; t = −2.4, p = 0.01), and also slightly more ppe-pos neurons with cytoplasmic Arc mRNA in METH-pre-treated, 30 min relative to METH-pre-treated, CC rats (Fig. 4b; t = 1.9, p = 0.04). However, there was no difference between the saline- and METH-pre-treated 30-min groups in the numbers of ppe-pos neurons with cytoplasmic Arc mRNA expression (Figs 4 and 5a; t = 0.18, p = 0.43).
Upon observing impaired cytoplasmic Arc expression in ppe-neg neurons in the METH-pre-treated, 30-min group, we assessed whether this impairment was uniformly distributed across METH-pre-treated rats or whether this difference correlated to the DA depletion in each animal. Using only cell counts from the METH-pre-treated, 30-min group, we correlated the numbers of striatal efferent neurons with Arc mRNA in the cytoplasm with the percent DA depletion as assessed by DAT autoradiography (Fig. 5b and c). The numbers of ppe-pos neurons with cytoplasmic Arc mRNA were not significantly correlated with DA loss (Fig. 5b; R2 = 0.015, p = 0.7). Conversely, the numbers of ppe-neg neurons with cytoplasmic Arc mRNA expression was significantly inversely correlated to the DA depletion (Fig. 5c; R2 = 0.518, p = 0.019), indicating that greater DA loss is associated with fewer ppe-neg neurons with cytoplasmic Arc mRNA expression. In the METH-pre-treated, 5-min group, there was no significant correlation between cells with nuclear Arc and DA depletion (ppe-pos R2 = 0.005, p > 0.5; ppe-neg R2 = 0.05, p > 0.05).
Herein, we report novel findings that basal Arc transcription in striatal efferent neurons is enhanced, but Arc mRNA transcriptional activation and cytoplasmic localization in response to behavioral activation is severely impaired, in rats with METH-induced neurotoxicity. Specifically, behavioral activation associated with spatial exploration of a novel environment – the classic paradigm used to examine Arc mRNA induction and trafficking (Guzowski et al. 1999; Chawla et al. 2005; Vazdarjanova et al. 2006; Daberkow et al. 2007) – resulted in no concomitant increase in cells with nuclear Arc expression above the elevated CC levels in METH-pre-treated groups. There was also no significant overall increase in total number of neurons with cytoplasmic Arc mRNA in the METH-pre-treated, 30-min group, particularly with respect to ppe-neg neurons, similar to the suppressed cytoplasmic expression previously reported following a striatally mediated response-reversal task (Daberkow et al. 2008). Interestingly, the degree of DA depletion was significantly negatively correlated with the numbers of ppe-neg cells with cytoplasmic Arc, suggesting that the loss of this expression may arise as a consequence of disrupted DA signaling in striatum. Taken together, these data reveal a profound dysregulation of Arc transcriptional activation and subcellular trafficking. As Arc mRNA, and the resulting Arc protein, is critical to normal neuroplasticity processes (Chowdhury et al. 2006), such disruption may contribute to disrupted Arc-mediated memory processes of the basal ganglia subsequent to METH pre-treatment (Daberkow et al. 2008; Pastuzyn et al. 2012).
We presently demonstrate an intriguing increase in basal (CC) Arc mRNA transcription following METH-induced neurotoxicity relative to saline-pre-treated controls. Persistent changes in basal gene expression are not uncommon after repeated exposure to psychostimulants (c.f. (Unal et al. 2009), including reports from our lab showing long-term decreases in preprotachykinin mRNA expression after METH-induced neurotoxicity (Chapman et al. 2001; Johnson-Davis et al. 2002). However, to our knowledge, there are no prior reports of long-term changes in basal immediate-early gene expression in striatal neurons as a consequence of METH-induced neurotoxicity. Given that prior work suggested a relatively preferential impact of METH-induced neurotoxicity on striatonigral neuron function (Chapman et al. 2001; Johnson-Davis et al. 2002; Daberkow et al. 2008), the equivalent increases in basal Arc transcription observed in ppe-pos and ppe-neg neurons were surprising and suggest global alterations in striatal efferent neuron function following METH-induced neurotoxicity. This up-regulation may reflect alterations in signals regulating Arc transcription, including brain-derived neurotrophic factor or glutamate. Although both of these molecules are elevated 24–72 h after exposure of animals to a neurotoxic regimen of METH (Thomas et al. 2004; Mark et al. 2007), whether they remain elevated in striatum weeks later remains unknown. Clearly, further studies are needed to examine the mechanisms underlying long-term dysregulation of basal Arc transcription and also the extent to which the expression of other plasticity-related genes is similarly altered.
The present studies also reveal a profound impairment of acute, novelty-induced transcription of Arc in rats with METH-induced neurotoxicity. The basis for this observation remains presently unknown, but changes in phasic DA release seem a likely contributing factor. Partial monoamine loss is associated with impaired phasic-like, but not tonic-like, DA release (Bergstrom and Garris 2003; Howard et al. 2011). Modeling data suggest that D1-DA receptors should be particularly sensitive to changes in phasic DA amplitude (Dreyer et al. 2010). Consistent with this model, phasic-like stimulation of DA transmission increases gene expression in and electrophysiological activity of striatonigral, but not striatopallidal, neurons (Chergui et al. 1997; Gonon 1997; Onn et al. 2000). Finally, we presently report that in METH-pre-treated rats, the impairment in numbers of ppe-neg neurons with cytoplasmic Arc mRNA correlates with the extent of DA loss, which is further known to correlate with the degree of impairment of phasic-like DA neurotransmission (Bergstrom and Garris 2003; Howard et al. 2011). Thus, impaired phasic DA neurotransmission may ultimately contribute to the observed deficits in Arc expression subsequent to METH-induced neurotoxicity.
Decreased phasic DA neurotransmission and consequent decreases in D1-DA receptor activation may also impair Arc transcription by altering upstream signaling cascades. Activation of D1-DA receptors enhances NMDA receptor-mediated currents in striatonigral neurons (Cepeda et al. 1993; Andre et al. 2010; Jocoy et al. 2011) and increases activation of extracellular signal-regulated kinase (ERK1/2) (Valjent et al. 2005; Pascoli et al. 2011). Arc transcription is regulated by NMDA receptor and ERK1/2 activation (Korb and Finkbeiner 2011). Thus, decreased D1-DA receptor activation because of attenuated phasic DA transmission likely results in decreased NMDA receptor and ERK1/2 activation and, thus, decreased activity-dependent Arc induction in ppe-neg neurons.
Although the above-delineated effects can explain the basis for the impairment of acute, novelty-induced Arc transcription in ppe-neg neurons, the basis for the impaired transcriptional activation in ppe-pos neurons is less apparent. One possibility is that it also arises as a consequence of impaired phasic DA signaling and decreased D1-DA receptor activation in striatum. Disruption of striatal D1-DA receptor activation decreases cortical excitability (Steiner and Kitai 2000; Yano et al. 2006; Gross and Marshall 2009). Additionally, cortical immediate-early gene induction by DA receptor agonists is blunted in rats with METH-induced neurotoxicity (Belcher et al. 2009). Furthermore, novelty-induced c-fos mRNA expression in striatopallidal neurons is dependent on corticostriatal transmission (Ferguson and Robinson 2004), sensitive to decreases in D1-DA receptor activation (Ferguson et al. 2003), and sensitive to blockade of NMDA receptors or ERK1/2 signaling (Ferguson et al. 2003; Ferguson and Robinson 2004). These data thus suggest a model, wherein loss of D1-DA receptor activation in striatum because of disrupted phasic signaling secondary to METH-induced monoamine toxicity may lead to decreased cortical excitability and, consequently, decreased gene expression in ppe-pos neurons.
An alternative explanation for the overall loss of novelty-induced Arc transcription is that chronic elevations in basal Arc mRNA expression suppress further transcriptional activation. Massed exposures to a spatial environment in a single day impair Arc transcription (Guzowski et al. 2006) and α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptor activation can inhibit Arc transcription (Rao et al. 2006). Although there are DNA regulatory elements that repress Arc transcription (Pintchovski et al. 2009), what binds to those regulatory sites and the situations under which they are engaged are currently undefined. Therefore, the extent to which transcriptional repression of the Arc gene versus impaired activation of post-synaptic receptors (NMDA, D1) or intracellular signaling cascades contributes to the loss of evoked Arc expression in rats with METH-induced neurotoxicity remains to be determined.
The present results confirm prior observations (Daberkow et al. 2007, 2008) that normal, saline-pre-treated animals have more ppe-neg neurons than ppe-pos neurons with Arc mRNA in the cytoplasm. At present, the basis for this phenotypic difference in normal animals remains unknown, but it may be related to differences in Arc trafficking or cytoplasmic mRNA stability, given that Arc mRNA is a substrate for translation-dependent decay (Giorgi et al. 2007). The lack of transcriptional activation of Arc in rats with METH-induced neurotoxicity makes it somewhat difficult to discern whether DA contributes to this phenotypic difference; however, the present results suggest that the trafficking/stability of Arc mRNA in ppe-neg neurons in normal animals is dependent on DA signaling. First, despite the lack of novelty-induced transcriptional activation, METH-pre-treated, CC rats have more ppe-pos neurons with cytoplasmic Arc than do saline-pre-treated, CC rats, suggesting that METH-pre-treated rats still traffic basally transcribed Arc mRNA into the cytoplasm in ppe-pos neurons. Conversely, despite having high basal numbers of ppe-neg neurons with transcriptional activation of Arc, METH-pre-treated, 30-min rats do not have greater numbers of ppe-neg neurons with cytoplasmic Arc expression compared with saline-pre-treated, 30-min rats. Thus, the elevated basal transcription of Arc mRNA in ppe-neg neurons of METH-pre-treated rats does not translate into greater numbers of neurons with cytoplasmic Arc, suggesting some disruption of Arc trafficking/stability in ppe-neg neurons. Second, whereas there is a slight, but significant, increase in the numbers of ppe-pos neurons with cytoplasmic Arc mRNA in the METH-pre-treated, 30-min group relative to the numbers in the METH-pre-treated, CC group, there is no such increase in the numbers of ppe-neg neurons. Taken together, these observations suggest that METH-induced neurotoxicity is associated with impaired acute transcriptional activation of Arc in both subtypes of striatal efferent neurons, as well as impaired trafficking/stability specifically in ppe-neg neurons.
We presently report that a neurotoxic regimen of METH has significant, long-term impact on the regulation of transcriptional activation and subcellular expression of Arc mRNA. This METH-induced effect on Arc mRNA transcription and expression may thus impair Arc protein translation, with Arc protein synthesized within 60 min of neuronal stimulation (Lyford et al. 1995; Vazdarjanova et al. 2006; Baez et al. 2011). Importantly, this disrupted activity-dependent striatal Arc expression, and hypothesized impact on protein synthesis, may underlie previously reported changes in basal ganglia-mediated learning and memory processes in METH-pre-treated rats, including response-reversal (Daberkow et al. 2008; Pastuzyn et al. 2012), sequential motor (Chapman et al. 2001), and stimulus-response versus action-outcome (Son et al. 2011) learning. Our present results therefore suggest Arc mRNA or the factors regulating it as potential novel therapeutic targets in the treatment of METH-induced cognitive impairments (Volkow et al. 2001; Kalechstein et al. 2003; Johanson et al. 2006), as Arc is a critical mediator underlying normal neuroplasticity processes (Chowdhury et al. 2006; Rial Verde et al. 2006). Future work will thus need to examine whether therapeutically restoring partial DA loss will resolve the molecular and cellular changes that occur, as well as improve cognitive functionality compromised with METH addiction and abuse.
The authors declare no conflict of interest. This work was supported by NIDA DA024036 (KAK), the American Foundation for Pharmaceutical Education (MBH) and the GPEN Program (KO).