Molecular profiling of midbrain dopamine regions in cocaine overdose victims

Abstract

Chronic cocaine use in humans and animal models is known to lead to pronounced alterations in neuronal function in brain regions associated with drug reinforcement. To evaluate whether the alterations in gene expression in cocaine overdose victims are associated with specific dopamine populations in the midbrain, cDNA arrays and western blotting were used to compare gene and protein expression patterns between cocaine overdose victims and age-matched controls in the ventral tegmental area (VTA) and lateral substantia nigra (l-SN). Array analysis revealed significant up-regulation of numerous transcripts in the VTA, but not in the l-SN, of cocaine overdose victims including NMDAR1, GluR2, GluR5 and KA2 receptor mRNA (p < 0.05). No significant alterations between overdose victims and controls were observed for GluR1, R3 or R4 mRNA levels. Correspondingly, western blot analysis revealed VTA-selective up-regulation of CREB (p < 0.01), NMDAR1 (p < 0.01), GluR2 (p < 0.05), GluR5 (p < 0.01) and KA2 (p < 0.05) protein levels of cocaine overdose victims. The present results indicate that selective alterations of CREB and certain ionotropic glutamate receptor (iGluR) subtypes appear to be associated with chronic cocaine use in humans in a region-specific manner. Moreover, as subunit composition determines the functional properties of iGluRs, the observed changes may indicate alterations in the excitability of dopamine transmission underlying long-term biochemical and behavioral effects of cocaine in humans.

Neuropharmacological evidence in animal models and imaging studies in humans indicate that the mesocorticolimbic dopamine pathway is regulated by cocaine administration and that the functional integrity of this pathway is essential for the reinforcing and euphoric effects of cocaine (Ritz et al. 1987; Wise and Bozarth 1987; Di Chiara and Imperato 1988; Koob and Bloom 1988; Volkow et al. 1999). The mesocorticolimbic pathway originates in the ventral tegmental area (VTA) and projects to several forebrain regions, most notably the nucleus accumbens (NAc), a region implicated in cocaine reinforcement. Cocaine administration produces significant elevations in NAc and VTA extracellular dopamine concentrations in animal models (Pettit and Justice 1989, 1991; Iyer et al. 1995; Hemby et al. 1997a, 1999; Czoty et al. 2000) and alters metabolic function and blood flow in terminal regions following chronic cocaine use in humans (Volkow et al. 1988; Holman et al. 1991, 1993; Volkow et al. 1991; Strickland et al. 1993; Levin et al. 1994), effects that contribute to the abuse liability of cocaine (Ritz et al. 1987; Bergman et al. 1989).

In addition to the direct interaction between cocaine and the dopamine transporter and indirect agonist effects on dopamine receptor subtypes, neuroadaptive processes are induced by cocaine that may lead to persistent or even permanent alterations in neuronal function (Nestler 1993; White et al. 1995b; Nestler 1997; Pierce and Kalivas 1997; White and Kalivas 1998). Alterations in the expression of genes associated with dopaminergic neurotransmission have been an area of investigation into the neuroadaptations induced by chronic cocaine administration. The intracellular cascade mediating such events include a generalized up-regulation of the cyclic adenosine 3′,5′-monophosphate (cAMP) pathway (Nestler et al. 1990; Terwilliger et al. 1991; Striplin and Kalivas 1992; Miserendino and Nestler 1995; Carlezon et al. 1998; Self et al. 1998; Pliakas et al. 2001) and activator protein 1 family members (Hope et al. 1992; Couceyro et al. 1994; Nye et al. 1995; Hiroi et al. 1997; Pich et al. 1997; Haile et al. 2001). For example, chronic cocaine administration increased the formation of cAMP-dependent protein kinase A (PKA) and adenylate cyclase in the nucleus accumbens of rats (Terwilliger et al. 1991), as well as mRNA and protein levels of the α catalytic subunit of PKA and other potential transcriptional regulators in the NAc of rhesus monkeys following chronic cocaine administration (Freeman et al. 2001). Interestingly, stimulation of PKA or over-expression of cAMP responsive element binding protein (CREB) in the NAc attenuate the rewarding effects of cocaine (Carlezon et al. 1998; Self et al. 1998; Pliakas et al. 2001), suggesting that adaptive changes induced by cocaine on this intracellular cascade can alter subsequent responsiveness of the drug.

In addition to dopamine, several studies indicate an involvement of glutamate transmission in the VTA and NAc in the mediation of the behavioral and neurochemical effects of cocaine as well as neuroadaptations induced by chronic cocaine administration. AMPA and kainate subunits contribute to fast neurotransmission and all three ionotropic glutamate receptor (iGluR) subtypes are thought to play roles in long-term potentiation induced by cocaine (Nestler 2001; Ungless et al. 2001). iGluRs are classified as NMDA (NR1, NR2A-D, NR3) (+/−)-α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA; GluR1-4), and kainate (GluR5-7, KA1-2) receptor subunits based on their pharmacological characteristics and sequence information (Hollmann and Heinemann 1994; Borges and Dingledine 2002). Previous studies have shown cocaine administration increased extracellular glutamate concentrations in the NAc and VTA (Pierce et al. 1996; Reid et al. 1997; Kalivas and Duffy 1998), produced behavioral sensitization (Pierce et al. 1996) and resulted in increased responsivity of glutamate receptor stimulation in the NAc and VTA (White et al. 1995a; Zhang et al. 1997) with as little as one exposure to cocaine (Ungless et al. 2001). Increased AMPA and NMDA receptor expression in the VTA has been proposed as a possible mechanism for increased excitability of VTA dopamine neurons and behavioral sensitization to cocaine (White et al. 1995a; Zhang et al. 1997). Indeed, GluR1 and NR1 protein levels were increased in the VTA following chronic cocaine administration (Fitzgerald et al. 1996) and remained elevated following one day, but not three weeks, of withdrawal in rats that manifest behavioral sensitization to cocaine (Churchill et al. 1999). In contrast, others have reported elevated VTA NR1 protein levels after protracted withdrawal periods (3 and 14 days; (Loftis and Janowsky 2000) and no change in VTA GluR1 levels following 16–24 h of withdrawal (Lu et al. 2002). The discrepant results may be caused by several factors including dosing regimen, time since last cocaine injection, or the protein analysis procedures. In contrast to the protein data, NR1 mRNA levels were decreased in the VTA following acute, but not chronic, cocaine administration (Ghasemzadeh et al. 1999), whereas GluR1 and GluR2/3 subunit mRNAs in the VTA were not altered by cocaine administration in rats (Ghasemzadeh et al. 1999; Lu et al. 2002). Alterations in iGluRs in the VTA may represent another adaptive mechanism by which chronic cocaine can exert long-term effects on mesocorticolimbic dopamine function which, in turn, may affect subsequent drug intake.

Whereas animal models have advanced our understanding of the neurobiological basis of drug addiction, the evaluation of similar questions in human tissue are few, yet essential. Parallel investigations in human post-mortem tissue are important to determine whether biochemical changes observed and characterized in animal models are relevant to human drug abuse, as well as to identify novel changes that may be indicative of the human condition. The present study was undertaken to evaluate differences in gene expression in the VTA of cocaine overdose victims and age-matched, drug-free controls. In addition, we evaluated gene expression in the l-SN to provide a measure of brain region specificity. To this end, custom-designed macroarrays were used to simultaneously assess 81 genes and test the hypothesis that functional classes of genes were differentially expressed between cocaine overdose victims and controls in the VTA. Computational analyses were used to partition the data into groups of clones with similar function in order to facilitate the interpretation of these data to the relevance of cocaine addiction. Select genes that were differentially expressed were evaluated at the protein level by western blot analysis.

Experimental procedures

Subjects and tissue

Post-mortem human brain tissue was obtained at autopsy from 11 age-matched, drug-free control individuals [eight white males, one hispanic male, one black male and one white female; aged 35.0 ± 2.3 years; post-mortem interval (PMI) = 13.2 h ± 0.6 h] and 10 cocaine overdose victims (eight white males and two black males; aged 35.5 ± 2.2 years; PMI = 12.3 h ± 0.9 h) by the University of Miami Brain Endowment Bank (Miami, FL, USA) (Table 1). For protein analysis, tissue punches were dissected from the VTA and l-SN of the contralateral hemisphere and used for RNA analysis in eight cocaine overdose victims (5WM, 3BM; aged 34.4 ± 2.7 years; PMI = 12.5 ± 1.1 h) and seven age-matched, drug free controls (6WM, 1WF; aged 37 ± 2.7 years; PMI = 12.6 ± 0.7 h), representing a subset of subjects from whom tissue was used in the RNA experiments. Gross and microscopic diagnostic neuropathologic examinations, which included examination of multiple cortical and subcortical regions, were performed in all cases and no neuropathological abnormalities relevant to mental status were found. All cases were retrospectively accrued based on toxicological data and circumstances surrounding the death, including review of prior arrest records and treatment admissions, as well as pathological indications (e.g. perforation of the nasal septum and needle track marks), and were reviewed carefully before classifying a cocaine intoxication case. All cases were evaluated for common drugs of abuse and alcohol, and positive urine screens were confirmed by quantitative analysis of blood. Cocaine and benzoylecgonine concentrations in brain and blood were assessed using gas chromatography/mass spectroscopy as described previously (Hernandez et al. 1994).

Table 1.

Clinical information of study subjects

						Blood		Brain
Subject	Race/ Gender	Age (yrs)	PMI (h)	Cause of death	Toxicology	Coc (mg/L)	BE (mg/L)	Coc (mg/kg)	BE (mg/kg)	Cocethylene (blood/brain)
COD1	W/M	39	12.5	Cocaine intoxication	COC, CE	0.05	0.16	0.056	1.224	NA
COD2	W/M	36	13	Cocaine intoxication Atherosclerotic heart disease	COC, CE, EtOH, LC	0.05	0.11	0.136	0.276	0.204
COD3	W/M	41	12	Cocaine intoxication; Cardiac arrest	COC, EtOH	0.07	0.22	NA	NA	ND
COD4	B/M	32	18	Cocaine intoxication	COC, NIC	0.2	0.31	1.11	0.89	–
COD5	W/M	40	11.5	Cocaine intoxication	COC, DPHA	0.05	0.22	0.119	< 0.05	ND/< 0.05
COD6	B/M	34	14	Cocaine intoxication	COC, ME	0.049	2.4	–	–	0.09/ND
COD7	W/M	41	12	Cocaine intoxication	COC, ME	6.7	10	22.8	2.8	ND
COD8	W/M	23	8	Cocaine intoxication	COC, ME	0.05	0.28	0.11	0.15	0.1/0.14
COD9	W/M	27	12	Cocaine intoxication	COC	–	–	2.241	1.801	–
COD10	W/M	42	10.5	Cocaine intoxication	COC, CE	0.05	1.3	0.1	0.73	0.05/0.20
CTR1	W/M	35	16	Calcific aortic stenosis	ND
CTR2	W/F	39	11	Sharp/blunt force injuries	ND
CTR3	W/M	44	11.5	Acute myocardial infarction	ND
CTR4	W/H/M	26	14	Idiopathic cardiac conduction system disease	ND
CTR5	W/M	24	15.5	Occlusive coronary artery disease	ND
CTR6	W/M	46	11	Aortic aneurysim	ND
CTR7	W/M	34	14	Coronary arteriosclerosis	ND
CTR8	W/M	41	12	Atherosclerotic heart disease	ND
CTR9	W/M	34	11.5	Asthmatic bronchitis	ND
CTR10	W/M	37	14.5	NA	ND
CTR11	B/M	33	14	Surgical complications	ND

Abbreviations: BE, benzylecognine; CE, cocethylene; COC, cocaine; COD, cocaine overdose; CTR, control; DHPA, diphenhydramine; EtOH, ethanol; LC, lidocaine; NA, not available; ND, not detectable; PMI, post-mortem interval.

Following removal, brains were photographed and cut into 1.5-cm coronal blocks. Brain tissue was cryopreserved using a procedure adopted from previous studies (Hardy et al. 1983; Dodd et al. 1986). Punches (100 mg) were dissected from the blocks containing the VTA (posterior region of slab) and SN (anterior region containing the pars lateralis and the pars medialis): one hemisphere for RNA and the contralateral hemisphere for protein analysis. Possible neuronal loss, ischemic cell changes, and reactive gliosis were assessed using semiquantitative ratings by the neuropathologists and found to be negligible in all cases used in the present study.

RNA isolation and amplification

Micro-FastTrack Kits (Invitrogen, Carlsbad, CA, USA) were used to isolate polyadenylated RNA from the VTA and l-SN. mRNA yields ranged from 1 to 2 μg/100 ng of frozen tissue. Because of the quantity of mRNA, samples from each subject were amplified using a modification of the aRNA amplification procedure described previously (Ginsberg et al. 2000; Hemby et al. 2002, 2003). Oligo(dT)-T7-primer/promoter (500 ng) [AAACGACGGCCAGTGAATTGTAATACGACTCACTATAGGCGC(T)24] was hybridized to poly(A+)mRNA in the presence of 500 μm each of dNTPS- and RNase-free water for 30 min at 65°C then quick cooled on ice for 5 min. Following this, 10 mm dithiothreitol (DTT), 1X first-strand buffer and 20 u RNAsin were added and samples were incubated at 42°C for 2 min (Van Gelder et al. 1990). Reverse transcriptase (1 μL Superscript II, RNase H-; Invitrogen; Carlsbad, California) was added to the solution and incubated at 42°C for 60 min. Following phenol/chloroform extraction and ethanol precipitation in the presence of 10 μg linear acrylamide, samples were re-suspended in RNase free water and heat denatured at 94°C for 5 min, then quick cooled. Next, 1X second-strand buffer, 250 mm each of dNTPs, 10 u Klenow fragment and 10 u RNase H were added to the sample and incubated at 16°C for 4–6 h (Van Gelder et al. 1990). At the end of the incubation, 2.5 mm β-nictotanimide adenine dinucleotide and 1000 u Escherichia coli DNA ligase were added to the solution and incubated at room temperature (22°C) of 15 min, followed by the addition of 5 u T4 DNA polymerase at room temperature for 15 min (Sambrook and Russell 2001). At the end of the incubation, 2.5 mm ethylenediaminetetraacetic acid (EDTA) was added to stop the enzymatic reaction followed by phenol/chloroform extraction and ethanol precipitation. Following drop dialysis against RNase/DNase-free water for 15 min, purified cDNA templates were used to generate antisense RNA (aRNA). aRNA was synthesized using T7 RNA polymerase (EpiCentre Technologies; Madison, WI, USA) in 1X transcription buffer (Epicentre Technologies), 10 mm DTT, 250 μm each of NTPs and 20 u RNAsin at 37°C overnight. Following phenol/chloroform extraction and ethanol precipitation, aRNA was reverse transcribed using Superscript II in the presence of 1X first-strand buffer, 250 mm each of dNTPs, 100 ng of random hexamers and 20 u RNAsin incubated at 37°C for 1 h. For second-strand cDNA synthesis, samples were incubated with 10 u Klenow fragment, 250 mm each of dNTPs, 500 ng of the aforementioned oligo dT-T7 primer-promoter, and 5 u T4 DNA polymerase at 16°C overnight. Samples were phenol/chloroform extracted, ethanol precipitated and drop dialyzed for 15 min. For the second round of amplification, reaction conditions were identical for the first round except for the inclusion of 30 μCi [α-³³P]UTP (2500 Ci/mmol; Amersham Pharmacia Biotech; Piscataway, NJ, USA), 10 μm of UTP and 250 μm each of CTP, GTP and ATP. The aRNA procedure is a linear amplification process with minimal change in the relative abundance of the mRNA population in the native state of the neuron. mRNA can be reliably amplified from small quantities of fixed tissue including individual neurons and neuronal processes (Van Gelder et al. 1990; Eberwine et al. 1992; Ginsberg et al. 1999; Hemby et al. 2002, 2003). Although there is a general decrease in mRNA levels in autopsied tissue, the relative abundance of gene expression remains unchanged (Castensson et al. 2000).

Macroarray procedures

Reverse northern blots were prepared on nylon membranes containing, but were not limited to, dopamine receptors (e.g. D1, D2, D4, D5 and DAT), G protein subunits (αi1, αi2, αs, αz, αq, αo, β, γ1 and γ2), glutamate receptor mRNAs (mGluR3, GRIA1-4, GRIK5, 7 and GRIN1), GABA_A receptor subunits (α1, α2, β1, β3, γ2, δ, ∊ and π), regulators of G protein signaling (RGS1–7, 9, 10, 12, 13 and 16) and other transcripts (cannabinoid 1 receptor, cocaine–amphetamine regulated transcript (CART), serotonin 2 and serotonin 3 receptors and tyrosine hydroxylase). Inserts were amplified in 96-well plates using PCR with GF200 primers under the following conditions: 95°C for 5 min (1 cycle); 95°C for 30 s, 52°C for 45 s and 72°C for 2 min (40 cycles); and 72°C for 7 min (1 cycle). PCR samples were purified (Multiscreen PCR filter plate; Millipore, Bedford, MA, USA) and aliquots were electrophoresed on a 1% agarose gel (1X TAE, 0.05% ethidium bromide) at 5 V/cm for PCR band size verification. Gel images were captured by digital camera and archived. PCR product concentration was determined by spectrofluorometry (Gemini; Molecular Devices, Sunnyvale, CA, USA) using a 1 : 5000 dilution of SYBR 1 Green/TE and an aliquot of the PCR product. Values were compared with concentrations of known DNA standards for quantitation. Each amplified insert (250 ng) was spotted on Nytran SuPerCharge^° nylon transfer membrane (Schleicher and Schuell, Keene, NH, USA) using a 96-well dot blot apparatus (Minifold I; Schleicher and Schuell). DNA was crosslinked to the membrane by ultraviolet radiation.

Arrays were prehybridized with UltraHyb solution (Ambion, Austin, TX, USA) in hybridization bottles for 1 h at 42°C. Next, ³³P-labeled aRNA probes from each subject were heat denatured for 5 min at 65°C, and then immediately added to their respective bottles and allowed to hybridize for 18–24 h at 42°C in a rotisserie hybridization oven. Following hybridization, membranes were washed twice with 2X SSC/0.1% sodium dodecyl sulfate (SDS) and 0.1X SSC/0.1% SDS for 20 min each at 42°C. Labeled hybridized products were detected using phosphoimager screens, and hybridization signal intensities were analyzed using IMAGEQUANT software (Amersham Biosciences, Sunnyvale, CA, USA).

Protein preparation and western blot analysis

Tissue samples were dounce homogenized in 10 mm HEPES, 10 mm NaCl, 1 mm KH₂PO₄, 5 mm NaHCO₃, 1 mm CaCl₂, 0.5 mm MgCl₂, 5 mm EDTA and the following protease inhibitors (PI): 1 mm PMSF, 10 mm benzamidine, 10 μg/mL aprotinin, 10 μg/mL leupeptin, and 1 μg/mL pepstatin and centrifuged at 9645 g for 5 min. Supernatant (cytosol and membrane) was removed and the pellet (nuclei and debris) was resuspended in 20 mm Tris HCl, 1 mm EDTA (pH 8.0) with PIs and centrifuged at 9645 g for 5 min. This procedure = was repeated twice and the pellet was resuspended in the solution and stored at −20°C (nuclear fraction). Supernatant was centrifuged at 107 170 g for 30 min at 4°C. Following, the supernatant containing the cytosolic fragment was removed and stored at −20°C (cytosolic fraction). The pellet was re-suspended in 10 mm Tris (pH = 7.5), 300 mm sucrose, 1 mm EDTA (pH = 8.0), 0.1% NP40 and PIs and centrifuged at 4287 g for 5 min at 4°C. The supernatant was discarded and the pellet was resuspended in the buffer and washed three times before re-suspension in the buffer and PIs and storing the samples at −20°C (membrane fraction).

Protein concentrations of samples were calculated using the bicinochoninic acid protein assay kit (Pierce, Rockford, IL, USA) and diluted in Laemmli sample buffer to achieve the same final protein concentration. Identical quantities of proteins were loaded into a gel electrophoresis apparatus, subjected to sodium dodecyl sulfate-polyacrylamide (10%) gel electrophoresis (Bio-Rad, Richmond, CA) and transferred to nitrocellulose by electroblotting (30 V, overnight at 4°C) in 1X transfer buffer (Bio-Rad). Nitrocellulose membranes were blocked in 0.5% non-fat dry milk and 0.1% Tween 20 in phosphate-buffered saline (pH 7.4, 0.12 m) for 1 h at room temperature prior to being incubated with primary antibodies in blocking buffer (Bio-Rad) overnight at 4°C followed by secondary antibody for 1 h at room temperature. Protein bands were visualized on a Kodak XAR-5 film with enhanced chemiluminescence (ECL plus; Amersham Pharmacia Biotech). Primary antibodies were as follows: mouse monoclonal antibodies directed against NMDAR1 (Chemicon International, Temecula, CA, USA) and GluR5 (Upstate Biotechnologies, Waltham, MA, USA); rabbit polyclonal antibodies directed against GluR1, GluR2/3, KA2, Gα_s, Gα_i1/2, Gβ, CREB, phospho-CREB (Upstate Biotechnologies) and fos-related antigen 2 (FRA2) (L-15; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Secondary antibodies were HRP-conjugated anti-rabbit IgG and HRP-conjugated antimouse IgG (Upstate Biotechnologies). Protein abundances were calculated by optical densitometry with a Scan Jet 2200C and imported into NIH Image 1.61 software (National Institutes of Health, Bethesda, MD, USA). Film background was subtracted from the optical density values to give the optical density value for a single subject. All assays were conducted under conditions in which densitometric signal intensity was linear with protein concentration as determined in preliminary experiments.

Data analysis

Specific signal (minus background) of probe bound to each PCR product was expressed as a ratio of the total hybridization intensity of spots of the custom-designed array, thereby minimizing variations resulting from differences in the specific activity of the probe and the absolute quantity of probe present (Ginsberg et al. 2000; Hemby et al. 2002, 2003). Array data were analyzed using a two-way anova with group and anatomical region as the fixed effects and normalized hybridization intensity as the dependent measure. Post-hoc analyses were conducted as needed using Tukey’s test and the null hypothesis was rejected when p < 0.05. Western blot data were analyzed using t-tests comparing the band intensities between cocaine overdose victims and controls for each respective protein. Data were graphically depicted as the percentage of the controls for each respective protein. Null hypotheses were rejected when p < 0.05.

Results

Subject data

There was no significant difference in age between cocaine overdose victims and age-matched, non-drug controls (t = − 0.156, d.f. = 19, p = 0.877), post-mortem interval (t = 0.847, d.f. = 19, p = 0.408), or brain pH (t = 0.431, d.f. = 17, p = 0.672) indicating that these factors did not significantly influence the observed changes in gene or protein expression. Furthermore, neuroadaptive changes in the human brain post-mortem reflect chronic cocaine abuse, as death in a naive user is a rare occurrence, and the cohort of post-mortem subjects have many surrogate measures of chronicity (Ruttenber et al. 1997).

Gene expression

Custom-designed macroarrays containing 81 cDNAs were used to evaluate gene expression changes in the VTA and l-SN of cocaine overdose victims and controls. Although limited in scope compared with high-density arrays, the arrays contained a variety of transcripts including iGluR subunits (n = 8), dopamine signaling (n = 7), GABA transcripts (n = 10), G protein subunits (n = 14), GTPase/RGS proteins (n = 16), mRNA processing transcripts (n = 8), cell growth/death transcripts (n = 5) and others (n = 13).

Receptors

Several groups have reported alterations in glutamate receptor subunit protein and mRNA levels following cocaine administration. Consistent with these findings, we report a significant effect of cocaine overdose on glutamate receptor subunit mRNA expression in the VTA (F_1,209 = 18.541; p < 0.001) with a significant interaction between transcript and group (F_1,9 = 2.311; p < 0.017). Post-hoc analysis revealed a significant increase in NMDAR1, GluR2, GluR5 and KA2, but not GluR1, GluR3 or GluR4 glutamate receptor subunits or the mGluR3 receptor mRNAs in the VTA of cocaine overdose victims (Table 2). mGluR3 and GluR3 were in low abundance in the VTA. Furthermore, no significant differences were observed in glutamate receptor mRNA expression in the l-SN between cocaine overdose victims and controls.

Table 2.

Statistical analysis of gene expression profiles for VTA and I-SN of cocaine overdose victims and controls

	Main effect		Interaction		Post-hoc analysis
Gene class	VTA	I-SN	VTA	I-SN	VTA	I-SN
Glutamate Receptor	0.001a	0.070	0.017	0.972	–	–
GRIN (encoding NMDAR1)	–	–	–	–	0.033a	–
GRIA1 (encoding GluR1)	–	–	–		–	–
GRIA2 (encoding GluR2)	–	–	–	–	0.002a	–
GRIA3 (encoding GluR3)	–	–	–	–	–	–
GRIA4 (encoding GluR4)	–	–	–	–	–	–
GRIK1 (encoding GluR5)	–	–	–	–	0.003a	–
GRIK5 (encoding KA2)	–	–	–	–	0.004a	–
GRM3 (encoding mGluR3)	–	–	–	–	-	–
GABA transcripts	0.051	0.161	0.239	0.812	–	–
GABA subunit α1	–	–	–	–	–	–
GABA subunit α2	–	–	–	–	–	–
GABA subunit β1	–	–	–	–	–	–
GABA subunit β3	–	–	–	–	–	–
GABA subunit γ2	–	–	–	–	–	–
GABA subunit δ	–	–	–	–	–	–
GABA subunit ∊	–	–	–	–	–	–
GABA subunit π	–	–	–	–	–	–
GAD65	–	–	–	–	–	–
GAD67	–	–	–	–	–	–
G protein subunits	0.590	0.033b	0.023	1.000	–	–
Gαi1	–	–	–	–	–	–
Gαi2	–	–	–	–	–	–
Gαo	–	–	–	–	–	–
Gαl	–	–	–	–	–	–
Gαs	–	–	–	–	–	–
Gαt	–	–	–	–	–	–
Gαq	–	–	–	–	–	–
Gαz	–	–	–	–	–	–
Gα11	–	–	–	–	–	–
Gα15	–	–	–	–	–	–
Gβ1	–	–	–	–	0.001b	–
Gγ1	–	–	–	–	–	–
Gγ2	–	–	–	–	–	–
Gγ4	–	–	–	–	–	–
Dopamine transcripts	0.673	0.714	0.531	0.760	–	–
D1 receptor	–	–	–	–	–	–
D2 receptor	–	–	–	–	–	–
D4 receptor	–	–	–	–	–	–
D5 receptor	–	–	–	–	–	–
Dopamine transporter	–	–	–	–	–	–
Tyrosine hydroxylase	–	–	–	–	–	–
Dopa decaroxylase	–	–	–	–	–	–
Regulators of G protein signaling (RGS) & GTPases	0.175	0.027b	0.002	1.000	–	–
RGS1	–	–	–	–	–	–
RGS2	–	–	–	–	–	–
RGS3	–	–	–	–	0.001a	–
RGS4	–	–	–	–	–	–
RGS5	–	–	–	–	–	–
RGS6	_	–	–	–	–	–
RGS7	–	–	–	–	–	–
RGS9	–	–	–	–	–	–
RGS10	–	–	–	–	–	–
RGS11	–	–	–	–	–	–
RGS12	–	–	–	–	0.002b	–
RGS13	–	–	–	–	–	–
RGS16	–	–	–	–	–	–
Rho guanine nucleotide exchange factor (GEF) 5	–	–	–	–	–	–
RAB7, member RAS oncogene family like 1	–	–	–	–	–	–
Ras homolog enriched in brain 2	–	–	–	–	–	–
mRNA processing	0.204	0.210	0.442	0.998	–	–
Fos related antigen 2	–	–	–	–	–	–
Early growth response 2	–	–	–	–	–	–
Early growth response 3	–	–	–	–	–	–
protein kinase, interferon-inducible RNA dependent	–	–	–		–	–
CUG triplet repeat, RNA binding protein 2	–	–	–	–	–	–
RNApII	–	–	–	–	–	–
TATA box binding protein-associated factor, 30 kDa	–	–	–		–	–
RNA polymerase II polypeptide C, 33 kDa	–	–	–	–	–	–
Cell growth/death	0.759	0.515	0.118	0.943	–	–
activity-regulated cytoskeleton-associated protein (arc)	–	–	–	–	–	–
Brain derived neurotrophic factor (BDNF)	–	–	–	–	–	–
growth associated protein 43 (GAP43); neuromodulin	–	–	–	–	–	–
α fodrin	–	–	–	–	–	–
BCL2-associated X protein (bax)	–	–	–	–	–	–
Miscellaneous	0.481	0.278	0.003	0.205	–	–
Cocaine-amphetamine related transcript	–	–	–	–	0.001b	–
5-hydroxytryptamine (serotonin) receptor 2 A	–	–	–	–	–	–
5-hydroxytryptamine (serotonin) receptor 2C	–	–	–	–	–	–
5-hydroxytryptamine (serotonin) receptor 3	–	–	–	–	–	–
Cannabinoid receptor 1	–	–	–	–	–	–

Normalized expression values for designated transcript classes were analyzed using a two-way anova (Group × Transcript). Post-hoc analyses were conducted as needed using Tukey’s test and the null hypothesis was rejected when p < 0.05. p-values are indicated for Main effect, Interaction and post-hoc analysis.

^aIndicates significant up-regulation in cocaine overdose victims

^bIndicates significant down-regulation in cocaine overdose victims.

There was no significant difference in mRNA abundance for the various dopamine signaling transcripts including receptor subtypes (D1, D4 and D5), the dopamine transporter, tyrosine hydroxylase and dopa decarboxylase between cocaine overdose victims and controls in either the VTA (F_1,146 = 0.179, p = 0.679) or the l-SN (F_1,139 = 0.135, p = 0.714). There was a trend towards significance in GABA signaling transcripts including GABA_A receptor subunits (α1, α2, β1, β3, γ2, δ, ∊ and π) and glutamic acid decarboxylases (GAD) 65 and 67 in the VTA (F_1,230 = 3.843, p = 0.051); however, there was no significant difference in these transcripts between the groups in the l-SN (F_1,219 = 1.980, p = 0.161). Analysis of other neurotransmitter signaling transcripts (CART, serotonin receptor subtype 2C and 3, and cannabinoid receptor 1) revealed a significant GROUP by TRANSCRIPT interaction (F_1,4 = 4.440, p = 0.003) that was attributable to a down-regulation of CART in the VTA of cocaine overdose victims compared with controls (p < 0.05).

Signaling cascade transcripts

Previous studies in rodents have indicated a role for G protein subunits in cocaine intake. Examination of a variety of G protein subunit mRNAs (G_α11, G_α15, G_αl, G_αi1, G_αi2, G_αs, G_αt, G_αz, G_αq, G_αo, G_β1 and G_γ1–4) revealed a significant group and transcript interaction (F_1,16 = 1.863, p < 0.023) in the VTA that was attributable to a down-regulation of the Gβ1 subunit in cocaine overdose victims (p < 0.05). Interestingly, there was a significant down-regulation of G protein subunits in the l-SN of cocaine overdose victims compared with controls (F_1,339 = 4.589, p < 0.033), although there was no interaction.

GTPases and RGS proteins play an important role in regulating G protein function. In the VTA, there was a significant interaction of group and transcript for RGS/GTPase mRNAs [RGS1–7, 9, 10, 12, 13, 16, Rho guanine nucleotide exchange factor 5 (tim1), RAB7, member RAS oncogene family like 1 (RAB7L1), and Ras homolog enriched in brain 2 (RHEB2)] (F_1,314 = 2.514, p = 0.002) that was attributable to a significant up-regulation of RGS3 (p < 0.05) and down-regulation of RGS12 (p < 0.05) in cocaine overdose victims. In contrast, there was a significant difference in the overall abundance of these transcripts between cocaine overdose victims and controls in the l-SN (F_1,298 = 4.951, p = 0.027), but no significant interaction (Table 2).

mRNA processing and cell growth/death related transcripts

Analysis of several transcript encoding proteins related to mRNA processing [FRA2; early growth response (EGR) 2 and 3; protein kinase interferon-inducible double stranded RNA dependent (PRKR); CUG triplet repeat RNA binding protein 2 (CUGBP2); RNA polymerase II-DNA directed- polypeptide C 33 kDa (POLR2C); suppressor of RNA polymerase B7 (RNApII); TATA box binding protein-associated factor, RNA polymerase II, H, 30 kDa (TBP-af30)] did not reveal any significant difference between cocaine overdose victims and controls in either the VTA or the l-SN. Likewise, there was no significant difference in the abundance of cell growth/death-related transcripts [activity-regulated cytoskeleton-associated protein (arc); brain derived neurotrophic factor (BDNF); growth associated protein 43 (GAP43); α-spectrin, nonerythrocytic 1 (α-fodrin); and BCL2-associated X protein (bax)] between the groups in either the VTA or l-SN (Table 2).

Protein expression

Western blot analysis was performed on protein homogenates from individual subjects to assess the correlation of changes in mRNA levels with protein levels. From the gene expression analysis, proteins for NR1, GluR1, GluR2, GluR5, KA2, Gαi1, Gαs and Gβ were selected based on the availability of selective antibodies. In addition, protein levels of CREB, phosphorylated CREB and FRA2 were evaluated based on their regulation in animal models of drug abuse (Carlezon et al. 1998; Self et al. 1998; Pliakas et al. 2001).

In membrane fractions, NR1 immunoreactive protein was increased 111.2% (p < 0.05), GluR2/3 by 69.1% (p < 0.05), GluR5 by 507.9% (p < 0.01) and KA2 by 72.1% (p < 0.05) in the VTA of cocaine overdose victims, whereas GluR1 showed a non-significant 50.3% increase in this region (Fig. 1). Comparatively, these proteins were slightly increased in the l-SN, albeit in a non-significant manner (NR1, 12.9%; GluR1, 56.1%; GluR2/3, 30.1%; GluR5, 13.4%; KA2, 50.4%). A previous study indicates that the aforementioned iGluR subunits are stable up to 18 h PMI (Wang et al. 2000), suggesting that the present results are not compromised by proteolysis.

Fig. 1.

(a) Ionotropic glutamate receptor subunit protein levels in VTA and l-SN of cocaine overdose victims. Membrane fractions were isolated as described in Experimental procedures and 10 μg was separated on 10% SDS–PAGE to assess glutamate receptor subunit imunoreactivity. Data are expressed as mean ± SEM of the percentage of control values per quantity of protein loaded; *p < 0.05; **p < 0.01. (b) Representative bands from two cocaine overdose victims (+) and two control subjects (−) for each subunit.

Protein levels were increased 23.3% for G_αi1/2 and 23.6% for G_αs whereas G_β levels were decreased 21.7% in the VTA of cocaine overdose victims compared with controls in cytosolic fragments. In the l-SN, G_αi1/2 was increased by 10%, whereas G_αs was decreased by 15.1% and G_β by 10.1% in cocaine overdose victims compared with controls, all in a statistically non-significant manner (Fig. 2).

Fig. 2.

(a) Levels of G protein subunits in VTA and l-SN of cocaine overdose victims. Cytoloplasmic fractions were isolated as described in the Experimental procedures and 10 μg was separated on 10% SDS-PAGE to assess G protein subunit immunoreactivity. Data are expressed as mean ± SEM of the percentage of control values per quantity of protein loaded. (b) Representative bands from two cocaine overdose victims (+) and two control subjects (−).

Previous studies have hypothesized that adaptive changes induced by cocaine on the cAMP intracellular cascade can alter subsequent responsiveness of the drug (Carlezon et al. 1998; Self et al. 1998; Pliakas et al. 2001). Evaluation of CREB immunoreactive protein revealed a significant up-regulation (69.6%; p < 0.05) of CREB in the cytosolic fraction from the VTA, but not l-SN, of cocaine overdose victims (Fig. 3). There was a slight but statistically nonsignificant increase in CREB immunoreactivity in the nuclear protein fraction from the VTA (34.3%), whereas levels in the l-SN were similar between cocaine overdose victims and controls. We were unable to reliably detect phospho-CREB immunoreactivity in either the cytosolic or the nuclear fractions in either brain regions of cocaine overdose victims or controls (data not shown). There was no significant difference in FRA2 protein levels between cocaine overdose victims and controls in either the VTA or the l-SN (data not shown).

Fig. 3.

(a) CREB protein levels in the cytosolic and nuclear fragments of VTA and l-SN in cocaine overdose victims. Cytoplasmic and nuclear fractions were isolated as described in Experimental procedures and 15 μg was separated on 10% SDS–PAGE to assess CREB imunoreactivity. Phospho-CREB immunoreactivity was not detectable in either the cytoplasmic or the nuclear fractions. Data are expressed as mean ± SEM of the percentage of control values; *p < 0.05. (b) Representative bands from two cocaine overdose victims (+) and two control subjects (−).

Discussion

In the present study, macroarrays and western blot analyses were used to examine the expression of select genes and proteins in the VTA and l-SN of cocaine overdose victims and controls. A membrane-based macroarray platform with radioactive hybridization was used to query 81 human cDNAs of interest. Significant alterations were observed for a variety of protein families including iGluR subunits, G protein subunits, RGS proteins, miscellaneous transcripts including CART, and a trend towards significance in GABA_A receptor subunits. Previously, we have employed similar approaches to identify molecular profiles of other diseases in human brain tissue including Alzheimer’s disease, schizophrenia, drug abuse, as well as normal aging (Ginsberg et al. 2000; Hasenkamp and Hemby 2002; Hemby et al. 2002, 2003). Western blot analyses were used to confirm changes in protein levels for some of the differentially expressed transcripts and revealed significant up-regulation of iGluR subunit and CREB protein levels in the VTA of cocaine overdose victims. The present results provide a composite assessment of cocaine-induced alterations in gene expression in a brain region commonly associated with the drug reinforcement.

The major finding of the present study was the up-regulation of NR1, GluR2, GluR5 and KA2 glutamate receptor subunit mRNA and protein levels specific to the VTA in human cocaine overdose victims. Previous studies in rodents indicated that up-regulation of iGluR subunits was associated with augmented dopamine–glutamate interactions in the mesolimbic pathway (White et al. 1995a,b; Zhang et al. 1997; White and Kalivas 1998; Wolf 1998; Vanderschuren and Kalivas 2000; Giorgetti et al. 2001). Results of the majority of cocaine administration studies in rodents contrast with the present findings. For example, NR1 mRNA levels were decreased in the VTA following acute cocaine administration whereas chronic administration failed to alter mRNA levels of GluR1-4, NR1 or mGlur5 in this region (Ghasemzadeh et al. 1999; Lu et al. 2002). Likewise, GluR1 protein levels are elevated in the VTA following chronic cocaine administration (Fitzgerald et al. 1996) but were no different than controls 16–24 h after the last cocaine injection (Lu et al. 2002). In agreement with the present results, NR1 protein levels were elevated in the VTA following chronic cocaine administration (Fitzgerald et al. 1996) and remain elevated 1, 3 and 14 days following the last administration of cocaine (Churchill et al. 1999; Loftis and Janowsky 2000). These data suggest that increased NR1 protein levels in cocaine overdose victims may be induced by acute or subacute cocaine exposure or possibly reflect the last acute cocaine binge prior to death. Up-regulation of mRNA and protein levels of GluR2, GluR5 and KA2 may be indicative of chronic cocaine use and/or cocaine overdose in humans.

In the VTA and SN, GluR1-4 and NR1 immunoreactivity was found to be localized almost exclusively on dopaminergic neurons (Paquet et al. 1997); however, less is known about the localization of kainate subunits in these regions. Glutamatergic afferents in the midbrain originate from the medial frontal cortex, subthalamic and pedunculopontine tegmental nuclei (Christie et al. 1985; Sesack et al. 1989; Sesack and Pickel 1992; Lu et al. 1997). A recent study suggested prefrontal glutamatergic projections selectively synapse on mesoprefrontal dopamine neurons and mesoaccumbens GABA neurons in the VTA in rats (Carr and Sesack 2000), although the topography in non-human primates and humans remains unknown. The presence of these subunits on dopamine neurons and the up-regulation reported in the present study may provide a means by which to influence dopamine excitability in this region. Up-regulation of the kainite-preferring subunits GluR5 and KA2 in cocaine overdose victims and the probable localization of these subunits on dopamine neurons represent additional means by which increased Ca²⁺ influx may lead to hyperexcitability of VTA dopamine neurons. Interestingly, GluR2 subunit protein levels were increased in the hippocampus of alcoholics (Breese et al. 1995) and GluR5 mRNA and protein levels were increased in the dorsal prefrontal cortex of cocaine-treated rats (Toda et al. 2002). Because extensive editing of GluR2 in the adult brain renders ionophores less permeable to Ca²⁺, the up-regulation observed in the present study may represent a molecular compensation for the effects of increased Ca²⁺ influx generated by the combined up-regulation of NR1, GluR5 and KA2 (Paschen et al. 1994). As a result of the influence of RNA editing for each of these glutamate receptor subunits, which confer different desensitization kinetics and other pharmacological properties (Borges and Dingledine 2002), the possibility remains that cocaine administration may differentially affect the expression of specific variants (Lomeli et al. 1994; Mosbacher et al. 1994). Inasmuch as splice variants were not discriminated using either cDNAs or antibodies in the present study, future studies are warranted to assess the regulation of splice variants as a result of cocaine exposure. In summary, up-regulation of the iGluR subunit mRNA and protein levels in the present study are likely attributable to changes in VTA dopaminergic neurons, suggesting a neuroadaptive response to cocaine in these subjects.

Dopamine receptor mRNAs were not differentially regulated as a function of cocaine overdose, in agreement with previous literature (Peris et al. 1990; Meador-Woodruff et al. 1993; Moore et al. 1998), opening the possibility that intracellular signaling mechanisms may also contribute to the altered function of these neurons. Neuroadaptations in the cAMP pathway have been implicated in mesolimbic brain regions as a function of cocaine exposure (Nestler et al. 1990; Terwilliger et al. 1991; Striplin and Kalivas 1992; Miserendino and Nestler 1995; Carlezon et al. 1998; Self et al. 1998; Pliakas et al. 2001). Assessment of various α, β and γ G protein subunits revealed only a significant decrease in G_β1 subunit mRNA, but not in protein, in the VTA of cocaine overdose victims. The discrepancy between G_β1 mRNA and protein levels may be reflective of post-transcriptional, translational or post-translational processing/degradation and/or trafficking of the protein outside of the regions studied. In addition, decreased G_β mRNA levels may manifest as down-regulated protein levels in axonal targets of the mesoaccumbens dopamine neurons, as shown previously in cocaine-treated rats (Wang et al. 1997). The present data contrast previous studies showing decreased ADP ribosylation and immunoreactivity G_αi and G_αo in the VTA of cocaine-treated rats (Nestler et al. 1990; Striplin and Kalivas 1993). However, a significant increase in CREB mRNA and protein levels in the cytosolic fraction and a trend towards significance in the nuclear fraction in the VTA of cocaine overdose victims was observed – the first demonstration of increased CREB protein levels in the VTA of either humans or animal models as a function of cocaine exposure. Unfortunately, the length of the post-mortem intervals for these subjects are likely to have contributed to the inability to detect phosphorylated CREB imunoreactivity; precluding speculation on the role of elevated CREB levels in the present study. Although CREB is almost exclusively expressed in the nucleus, CREB protein has been identified in the cytoplasm by light and electron microscopy (Ferrer et al. 1996; Hermanson et al. 1996; Suzuki et al. 1998; Shaywitz and Greenberg 1999) as well as in postsynaptic density fractions (Suzuki et al. 1998) and dendrites (Crino et al. 1998) leading to the hypothesis of nuclear translocation as a means of site-specific plasticity (Crino et al. 1998; Suzuki et al. 1998).

In relation to the present data, differential regulation of cytosolic vs. nuclear CREB should be addressed. In order to avoid overspeculation regarding this difference, statistical analysis indicated a trend towards significance in nuclear CREB (p < 0.07) in the VTA of cocaine overdose victims. Secondly, the ability of the antibody to detect the antigen in the nucleus may be impaired by conformational changes in the protein caused by dimerization, binding of the dimer to DNA, and/or masking of the antigen by CREB binding proteins or other transcriptional regulators. A more precise role of cytoplasmic CREB and the mechanism underlying elevated CREB levels remains to be determined in animal models of the disease. The relevance of altered CREB levels in the present study supports several studies in rodents that indicated CREB levels were inversely proportional to the rewarding effects of the drug (Carlezon et al. 1998; Walters and Blendy 2001). One manner in which CREB activity may be regulated is by Ca²⁺ influx via NMDA receptors, demonstrating a requirement for CREB–NMDA receptor interactions in dopamine-regulated gene expression (Konradi et al. 1996).

One of the transcripts regulated by CREB is CART (Dominguez et al. 2002). Discovered as a novel transcript in the striatum, the regulation of which was induced by acute cocaine (Douglass et al. 1995), CART peptides were shown to be abundant in the NAc, VTA and SN as well as other brain regions (Smith et al. 1997; Smith et al. 1999; Hurd and Fagergren 2000; Dallvechia-Adams et al. 2002). The present study is the first to demonstrate CART mRNA regulation by chronic cocaine use in humans. Future studies are warranted to explore the location, abundance and regulation of these CART peptides by cocaine administration.

The up-regulation of RGS3 and RGS12 mRNAs in the VTA of cocaine overdose victims represents another potential mechanism for cocaine-induced alterations in cell signaling. Previous studies have shown elevated RGS2, 3 and 5 mRNA levels in the striatum caused by acute amphetamine administration (Burchett et al. 1999) and RGS2 mRNA levels by cocaine in the hippocampus, cortex and striatum (Ingi et al. 1998). RGS proteins modulate/ integrate G protein and other intracellular signaling proteins, some of which reduce the duration of G protein activation by increasing the rate of GTP hydrolysis (Berman et al. 1996; Hunt et al. 1996; Chatterjee et al. 1997; Hepler et al. 1997; Berman and Gilman 1998). In vivo, RGS3 interacts with the G protein βγ complex to accelerate the rate of recovery of N-type Ca²⁺ inhibition from G_βγ thereby leading to increased Ca²⁺ conductance (Jeong and Ikeda 1998; Melliti et al. 1999; Hollinger and Hepler 2002) and potentially to increased excitability of VTA dopamine neurons. There is limited information regarding the functionality of RGS12 and how cocaine may regulate its function. The manner in which cocaine and other drugs of abuse may regulate RGS expression and, in turn, how this regulation affects subsequent drug responsiveness and intake remain unanswered. The lack of effective antisera for RGS3 and 12 proteins and their potentially low protein expression (Hollinger and Hepler 2002) preclude a more complete understanding of the regulation of the proteins at the present time; however, it is reasonable to speculate that RGS proteins appear to be located in dopamine pathways and may be involved in neuroadaptations in signal transduction mechanisms associated with chronic cocaine exposure.

Understanding the consequences of long-term cocaine abuse on post-mortem human brain requires rigorous investigation. The present study demonstrated that some of the adaptations that occur in animal models of drug abuse are applicable to the human brain. Discrepancies between the present data and the existing rodent literature may reflect differences in the (i) comparative complexities of chronic cocaine use, (ii) anatomical connectivity of the VTA (Alheid et al. 1990; Heimer et al. 1995), (iii) as well as differences in the response contingency of drug administration (Hemby et al. 1997b). With regard to the present data, whether the observed changes could be attributed to acute or chronic cocaine use or cocaine overdose remains to be addressed. Future studies should include comparison of gene and protein profiles with individuals diagnosed with cocaine addiction histories but who do not die of cocaine overdose as well as individuals who die during excited cocaine delirium (Ruttenber et al. 1997). More importantly, additional efforts should be directed towards the recapitulation of these findings in rodent and non-human primate models (e.g. self-administration) in order to develop a better understanding of the variables that may contribute to these alterations. Although there are many difficulties with post-mortem brain studies, it is one of the most promising ways to view biochemical changes that are relevant to human drug abusers and to educate the public about the consequences of cocaine abuse. Whereas animal studies have advanced our understanding of the neurobiological basis of drug addiction, the evaluation of similar questions in human tissue are so far few and yet are essential. By assessing changes in defined biochemical pathways in human post-mortem tissue, we can begin to ascertain the fundamental molecular and biochemical processes that are associated with long-term cocaine use.

Abbreviations used

AMPA

(+/−)-α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

aRNA

antisense ribonucleic acid

BE

benzylecognine

cAMP

cyclic adenosine 3′,5′-monophosphate

CART

cocaine–amphetamine regulated transcript

CE

cocethylene

COC

cocaine

COD

cocaine overdose

CREB

cAMP response element binding protein

CTR

control

DHPA

diphenhydramine

EtOH

ethanol

FRA2

fos-related antigen 2

GAD

glutamic acid decarboxylases

GluR

glutamate receptor

iGluR

ionotropic glutamate receptor

LC

lidocaine

l-SN

lateral substantia nigra

mGluR

metabotropic glutamate receptor

NAc

nucleus accumbens

ND

not detectable

PI

protease inhibitors

PKA

protein kinase A

PMI

postmortem interval

RGS

regulators of G protein signaling

SDS

sodium dodecyl sulfate

VTA

ventral tegmental area