Introduction

At least four diseases have clear-cut linkage to abnormal dopaminergic (DAergic) synaptic transmission. In Parkinson's disease (PD), which is a prototypal degeneration of DAergic neurons, the loss of DAergic tone produces movement dysfunction and a subcortical type of dementia that includes impairment of executive function and working memory (Le Moal and Simon 1991; Robbins 2003). Schizophrenia and attention deficit hyperactivity disorder (ADHD) are conditions associated with executive dysfunction and loss of working memory that respond favorably to drugs that modulate DAergic tone in the prefrontal cortex (Kandel et al. 1991; Volkow et al. 2005). Drug addiction, especially the type involving the psychostimulant cocaine, is strongly linked to having excess DAergic tone in mesolimbic “reward circuits” that project through ventromedial striatum (nucleus accumbens) (Koob and Le Moal 1997; Volkow et al. 2004). People infected with human immunodeficiency virus (HIV-1) and/or those with acquired immunodeficiency syndrome (HIV/AIDS) often have HIV-associated subcortical dementia (HAD), motor slowing, neuropsychiatric disturbances including psychosis, and drug addiction (Heaton et al. 1995). For that reason, the interplay between HIV infection and DAergic dysfunction is a topic of some clinical relevancy (Nath et al. 2002; Koutsilieri et al. 2002; Grassi et al. 1997). Brain HIV infection might influence the basic neurobiology and clinical management of DAergic neurological dysfunction. For example, it has been suggested that HIV/AIDS increases the vulnerability to extrapyramidal side effects of antipsychotic medicines; these drugs act in part by blocking dopamine receptors, so altered DAergic tone is a possible mechanism (Berger and Nath 1997). Conversely, addictive behavior and neuropsychiatric problems driven by DAergic circuits can promote behaviors that can increase the risk of HIV transmission to other people (such as needle sharing or risky sexual practices) (Chiasson et al. 1991). Drug addiction and psychosis in HIV/AIDS can also decrease compliance to antiretroviral medication regimens and lead to increased morbidity and mortality rates. Drugs of abuse such as cocaine, morphine, and alcohol can alter brain immunity and/or HIV-1 replication rates in model systems, and could increase HIV neurovirulence or exacerbate HAD (Baldwin et al. 1998; Roth et al. 2002; Koutsilieri et al. 2002). These scenarios suggest that the interaction between HIV infection and DAergic systems has potential impact on disease transmission, progression of virus replication, neurological dysfunction, response to treatment, and comorbidity (Volkow et al. 2004, 2005). To determine more precisely the effect of HIV on DAergic function, we undertook a biochemical characterization of striatal DAergic synapses in people with HIV encephalitis (HIVE), which is the neuropathological substrate of HIV-associated dementia (HAD) in many cases.

Materials and methods

Patients

Twelve human brain samples were characterized biochemically (Table 1). The subjects were recruited into the National NeuroAIDS Tissue Consortium (NNTC) cohort. Subjects with HIV infection had various degrees of neurocognitive dysfunction documented prior to death by using the NNTC testing protocol (Morgello et al. 2001; Woods et al. 2004). When the subjects died, an autopsy was performed and brain specimens were frozen at −80°C and banked at the University of Texas Medical Branch in Galveston, TX (Gelman et al. 2004, 2005). The six people with HIV-1 infection died from end-stage AIDS and were selected for study based on the presence of HIV encephalitis (HIVE) as determined by neuropathological examination (Budka 1991). Specimens were excluded from consideration if they had a confounding neuropathological lesion. Six comparison brain specimens were selected from the Galveston NNTC brain repository that did not have HIV-1 infection and did not have encephalitis or neuropathological abnormality, and did not have a recorded history of neurocognitive dysfunction. Raw neurochemical data, clinical data, brain and other organ specimens, blood plasma, cerebrospinal fluid and other resources from this set of subjects, and other subjects are available to the public at no cost by contacting the primary author or the National Coordinating Office of the NNTC (http:\\www.hivbrainbanks.org).

Table 1 Summary of decedents
Full size table

Brain specimen processing

Frozen human postmortem brain hemispheres were dissected and stored at −80°C by using the protocol established by the NNTC (Gelman et al. 2004). A sample of the head of the caudate nucleus, weighing approximately 0.4 g, was excised from the frozen brain slice with a porcelain cutting tool (Dremel Corp., Racine, WI, USA). Samples were placed in 3 volumes of standard radio-immuno-precipitation (RIPA) homogenization buffer (0.05 M Tris–HCl pH 7.4, 0.15 M NaCl, 0.001 M EDTA, 1% IGEPAL, 0.1% SDS) containing protease inhibitor cocktail (Roche, Indianapolis, IN, USA) and phosphatase inhibitor cocktail set II (Calbiochem, San Diego, CA, USA) on ice. RIPA buffer was used to obtain optimal solubilization of membrane protein. Zirconia/silica 0.5-mm microbeads were added and samples were placed in a Minibeadbeater (BioSpec Products, Bartlesville, OK, USA) and pulsed for 20-s intervals and returned to ice for cooling. Pulsing was repeated three times, the samples were centrifuged at 13,000 × g at 4°C, and supernatant was used to extract protein. Protein concentration was determined by BCA Assay (Pierce, Rockford, IL, USA).

Western immunoblots and densitometry

Fifty micrograms of protein was mixed with an equal volume of 2× Laemmli Sample Buffer (Bio-Rad Laboratories, Hercules, CA, USA), boiled for 5 min, and then loaded onto 4–15% gradient Tris–HCl polyacrylamide gels. Samples were electrophoresed at 150 V for 3 h and tank-transferred to PVDF membranes at 20 V overnight at 4°C in Tris-glycine buffer (0.025 M Tris base, 0.192 M glycine, pH. 8.4). The membranes were blocked in TBST (0.05 M Tris–HCL, 0.15 M NaCl, and 0.1% Tween 20) containing 5% nonfat dry milk for 1 h. Membranes were incubated with primary antibody diluted in fresh block solution overnight at 4°C and were then washed three times at ambient temperature for 5 min each with TBST. Membranes were incubated with appropriate secondary antibodies diluted in TBST for 1 h, washed three times, and then developed with ECL Detection Reagent (Amersham Biosciences, Piscataway, NJ, USA). Membranes were exposed to X-ray film (Kodak, Cedex, France). Primary antibodies used were as follows: β-tubulin (Sigma-Aldrich T7816, St. Louis, MO), dopamine transport protein (DAT; Chemicon ab5802, Temecula, CA, USA), tyrosine hydroxylase (TH) (Sigma T1299), TH phosphorylated at serine 40 (pTH) (Sigma T9577), dopamine type 1 receptor (D1R) (RDI-D1RABX, Research Diagnostics, Concord, MA, USA), dopamine type 2 receptor (D2R) (Calbiochem, #324393), dopamine type 3 receptor (D3R) (RDI-DOPD3ABR, Research Diagnostics). β-tubulin, TH, and pTH were used at a dilution of 1:5,000; all other primary antibodies were used at a dilution of 1:1,000. Densitometry was performed by using the program 1D Scan (ScanAlytics, Rockville, MD, USA) to quantify band intensity and relative protein abundance.

Isolation of RNA, reverse transcription and PCR

RNA was isolated from caudate by using Tri-Reagent (Sigma) as recommended by the manufacturer. Total RNA was treated with 5 units of DNAse (Invitrogen, Carlsbad, CA, USA) and 2 units of RNAguard (Amersham) per microgram of RNA for 30 min at room temperature, phenol extracted, chloroform extracted, and ethanol precipitated at −80°C. cDNA synthesis was performed using the iScript cDNA Synthesis Kit (Bio-Rad). PCR was performed by using Taq DNA Polymerase (Fisher Scientific, Fairlawn, NJ) under the following conditions: 1 cycle of 95°C denaturation for 5 min followed by 95°C denaturation for 30 s, 55°C annealing for 30 s, and 72°C elongation for 30 s. Cycle numbers for GAPDH and Gag were 23 and 45, respectively. cDNA was normalized to the levels of GAPDH in each sample. Primer sequences for GAPDH and Gag were as follows:

  • GAPDH1 5′-TGATGACATCAAGAAGGTGGTGAA-3′

  • GAPDH2 5′-TCCTTGGAGGCCATGTGGGCCAT-3′

  • GAG1 5′-GTAATACCCATGTTTTCAGCAT-3′

  • GAG2 5′-TCTGGCCTGGTGCAATAGG-3′

Immunohistochemistry of dopamine transport protein

Brain specimens were fixed in 20% formalin and washed. Samples of anterior basal ganglia were excised and embedded in paraffin wax. Tissue blocks were sectioned at a thickness of 6 μm and placed on Superfrost Plus Gold glass slides (Fischer). Slides were baked overnight at 55°C, deparaffinized in three successive 100% xylene baths, and rehydrated with decreasing concentrations of ethanol. Endogenous peroxidase activity was quenched by using 3% H2O2 in methanol. Antigen retrieval was performed using microwave-heating in 0.01 M sodium citrate (pH. 6.0) containing 0.2% Triton X-100 for 20 min at 10% power in a 2-L water bath. Washed slides were blocked in 0.1% nonfat milk and 1% goat serum and were incubated overnight at 4°C with rabbit polyclonal antibody (Chemicon) against DAT diluted 1:200. Biotinylated antirabbit secondary antibody was applied for 1 h, followed by Vectastain ABC avidin–peroxidase complex (Vector, Burlingame, CA). Color was developed using diaminobenzidine (DAB kit Vector). Slides were counterstained in Mayer's hematoxylin and covered after dehydration using Permount (Fisher) containing 25% xylene.

Results

β-tubulin concentration was used as a general neuronal marker. It was not significantly different in the people with HIVE, and suggests that the two groups contained equivalent amounts of neuronal protein (Fig. 1a, b). Three different presynaptic DAergic markers were abnormal in people with HIVE. The dopamine reuptake transport protein (DAT) is a specific presynaptic DAergic marker; it controls the concentration of dopamine in the synaptic cleft by transporting it across the presynaptic membrane (Rudnick and Clark 1993). DAT band intensity was increased significantly in HIVE (by over 4-fold; p < 0.001) (Fig. 1a, c). TH is a presynaptic protein that is the rate-limiting enzyme of the dopamine synthetic pathway. The TH band intensity was decreased by 50% (p < 0.01) in HIVE (Fig. 1a, d). TH can be posttranslationally phosphorylated at serine 40 (pTH) to a catalytically more active form (Wolf and Roth 1990; Lewis et al. 1987; Lindgren et al. 2000, 2001; Kansy et al. 2004; Haavik et al. 1989). When phosphospecific antibody was used against pTH, these band intensities showed a sharp decrease (69%; p < 0.001) (Fig. 1a, e). The ratio of pTH to TH, which reflects the major posttranslationally regulated component of enzyme catalysis, was decreased significantly (p < 0.005) (Fig. 1f).

Fig. 1
figure 1

Presynaptic markers of dopaminergic synapses in the striatum of people with HIVE compared to HIV seronegatives. (a) Western blots of DAT, TH phosphorylated on serine 40, TH, and β-tubulin. (b–f) Band intensities in panel (a) were quantified by using densitometry. β-Tubulin concentration (b) was unchanged between controls and HIVE. DAT concentration was increased and TH was decreased significantly in HIVE (c, d). TH phosphorylated on serine 40 was significantly decreased in HIVE (e). The ratio of phospho-TH to total TH was decreased significantly (f), which reflects a decrease in posttranscriptional TH phosphorylation. The sequencing of subjects in panel (a), left to right, matches the order in Figs. 2 and 4, and Table 1.

Full size image

Altered presynaptic DAergic drive can produce changes in gene expression in striatal postsynaptic neurons (Burt et al. 1977; Creese and Snyder 1979; Seeman 1980). To determine if DAergic synapses were perturbed postsynaptically, the concentration was measured of three genomically distinct dopamine receptors, D1R, D2R, and D3R. D2R protein was significantly decreased in HIVE; D3R protein was increased significantly; D1R protein was not changed (Fig. 2). Two alternatively spliced D2R molecules were detected in the immunoblots. The long splice variant (D2L) band migrates at 59 kDa and contains 444 amino acids; the short form (D2S) migrates at 47 kDa and contains 415 amino acids (Dal Toso et al. 1989). D2L was strongly decreased (p < 0002); D2S was decreased to a lesser extent (p < 0.029). There is evidence that D2S is preferentially synthesized by presynaptic neurons and is the “dopamine autoreceptor” (D2Ra), whereas D2L is the prevalent isoform synthesized in the postsynaptic medium spiny neuron (Khan et al. 1998; Centonze et al. 2002).

Fig. 2
figure 2

Postsynaptic markers of dopaminergic synapses in the striatum of people with HIVE compared to HIV seronegatives. (a) Western blots of three dopamine receptors, D1R, D2R, and D3R. (b–d) Band intensities shown in panel (a) were quantified by using densitometry. The concentration of the D2R long isoform (D2L) was sharply and significantly decreased in HIVE (b). The D2R short isoform (D2S) was significantly decreased (b). The concentration of D3R was increased significantly (c). D1R was not changed (d). The sequencing of subjects in panel (a), left to right, matches the order in Figs. 1 and 4, and Table 1.

Full size image

To determine the anatomical distribution of abnormal striatal DAergic synapses, we performed immunohistochemistry to localize DAT, which was sharply increased in HIVE and is representative of these DAergic anomalies. Figure 3 illustrates punctate DAT deposits that are typical of its synaptic localization in striatum (Ciliax et al. 1995). Abnormal DAergic synapses in HIVE had increased DAT immunostaining that was broadly distributed in striatum. Increased DAT staining in HIVE was not restricted to foci that contained HIVE changes, such as microglial nodules or multinucleated cells (not illustrated).

Fig. 3
figure 3

Example of increased immunostaining of DAT in rostral neostriatum of a subject with HIVE (b, d) and a seronegative control (a, c). Whole mounts show a very diffuse pattern of increased DAT staining intensity in the subject with HIVE (b). High-power magnification shows punctuate deposits of DAT in a typical synaptic staining pattern (c, d). Slides were counterstained with hematoxylin. C = caudate nucleus, P = putamen, IC = internal capsule, NA = nucleus accumbens. Scale bar is 5 mm in (a) and (b), and 2 μm in (c) and (d).

Full size image

Finally, we asked whether having abnormal DAergic synapses was related to having increased HIV replication in striatum. Figure 4 shows that all of the subjects with HIVE had amplified HIV Gag transcripts and none of the uninfected controls did. The amount of HIV Gag transcript amplified in HIVE was correlated positively and significantly with DAT and D3R (Table 2). TH, pTH, and D2L were negatively, but not significantly, correlated with HIV Gag.

Fig. 4
figure 4

HIV Gag transcripts in human striatum were amplified by using RT-PCR. Seronegative subjects (HIV-negative, control) did not contain HIV Gag transcripts. All six HIVE patients had amplified HIV Gag transcripts at various concentrations. cDNA was normalized to the levels of Gapdh. The sequencing of subjects, left to right, matches the order in Figures 1 and 2, and Table 1.

Full size image
Table 2 Amplified HIV Gag mRNA and the composition of striatal DAergic synapses
Full size table

Discussion

These data provide broad biochemical evidence that striatal DAergic synapses are abnormal in NNTC subjects with HIVE. Four separate gene products enriched in DAergic synapses were abnormal. The panel of anomalies included examples of both pre- and postsynaptic neuronal markers. The fact that there were strong increases in some DAergic markers, with sharp decreases in others, implies that the changes are not likely to be the result of postmortem protein degradation, generalized neurotoxicity, or dropout of synapses. Abnormal presynaptic DAergic markers included a sharp increase in DAT concentration, and a reciprocal decrease in TH concentration. Down-regulation of TH synthesis leads to a long-term decrease in the rate of dopamine synthesis because TH activity is the rate-controlling step (Tank et al. 1986a, 1986b; Fossom et al. 1992; Meller et al. 1987). Moreover, there was a posttranslational decrease in catalytically active pTH(ser40), which rapidly suppresses the rate of dopamine synthesis further (Wolf and Roth 1990; Lewis et al. 1987; Lindgren et al. 2000, 2001; Kansy et al. 2004; Haavik et al. 1989). These results suggest that the rate of dopamine synthesis in presynaptic neurons is sharply decreased in HIVE. Another potential presynaptic anomaly in HIVE was the decrease in D2S; this splice variant of the D2R gene may be the presynaptic D2R “autoreceptor” (D2Sa) (Khan et al. 1998; Centonze et al. 2002). A decrease in D2Sa expression in HIVE is consonant with changes in presynaptic TH and DAT, because D2Sa modulates gene transcription and protein function of those markers (Meller et al. 1987; Dickinson et al. 1999; Lindgren et al. 2001; Kimmel et al. 2001; Zahniser and Doolen 2001; Mayfield and Zahniser 2001). The concentration of two different postsynaptic DAergic markers was perturbed in HIVE. D2L protein was significantly decreased and D3R was significantly increased. D1R was unchanged. Both postsynaptic changes are highly selective phenotypic shifts because D1R, D2L, and D3R arise from separate genes and undergo unique transcriptional and posttranslational regulation. Because all of the markers were differently regulated in HIVE (increased, decreased, or unchanged), a simple dropout of presynaptic DAergic neurons or postsynaptic medium spiny neurons is not a likely scenario. Instead, the panel suggests that functional adaptation occurred in viable pre- and postsynaptic neurons (as opposed to neuronal dropout or degeneration; Gelman et al. 2004). All told, the straightforward biochemical measurement of DAergic synaptic markers has produced convergent, self-reinforcing lines of evidence that striatal DAergic synapses are not normal in people with HIVE.

To determine the functional significance of abnormal DAergic synapses in HIVE, we compared changes in HIVE to those that occur when dopamine availability is manipulated experimentally (Fig. 5). The biochemical changes in HIVE resemble perturbations that occur when DAergic tone (synaptic dopamine availability) is increased chronically. For example, decreased TH and increased DAT expression in presynaptic neurons both occur in cocaine abuse when DAergic tone and dopamine receptor occupancy are high. This adaptation probably serves to decrease the synthesis of dopamine and increase reuptake from the synaptic cleft, leading to a compensatory dampening of DAergic tone (Little et al. 1998, 1999; Staley et al. 1994; Chen et al. 1999). Conversely, these shifts are opposite to what is observed in PD striatum and animal models in which DAergic tone and receptor occupancy are low. This adaptation probably serves to increase dopamine synthesis and decrease reuptake, to compensate for decreased DAergic tone (Joyce et al. 1997; Blanchard et al. 1994; Harrington et al. 1996; Uhl et al. 1994). A similar interpretation follows from results of the postsynaptic markers, because synaptic dopamine concentration also modulates dopamine receptor expression (Burt et al. 1977; Creese and Snyder 1979; Seeman 1980). Thus, in cocaine addicts with high DAergic drive and high receptor occupancy, striatal D2R expression is driven downward, while D3R is generally increased (Staley and Mash 1996; Mash and Staley 1999; Segal et al. 1997; Volkow et al. 2004; Nader and Czoty 2005). Conversely, striatal D2R is increased and D3R is decreased in people and animals with low receptor occupancy and tone (Gerfen et al. 1990; Herrero et al. 1996; Quik et al. 2000; Bezard et al. 2001; Wade et al. 2001; Thobois et al. 2004). All told, the panel of changes in HIVE indicates that striatal dopaminergic synapses underwent coordinated adaptations that resemble what happens when DAergic tone is increased, and are in contrast to what happens when tone is decreased. That surprising conclusion disagrees with some of the conclusions suggested by clinical observation alone. For example, patients with HIV/AIDS can exhibit Parkinsonian-like extrapyramidal signs, which can suggest that DAergic tone is decreased in dorsolateral striatal circuits involved with motor control (Hriso et al. 1991; Lopez et al. 1999; Mirsattari et al. 1998; Berger and Nath 1997; Berger and Arendt 2000). However, correlative neuroanatomical evidence in clinically examined subjects to support that suggestion is not yet available. Retrospective postmortem measurements suggested that DAergic tone may be blunted in AIDS because of the dropout of DAergic neurons in pars compacta of the substantial nigra, but these decedents did not have a Parkinsonian syndrome (Reyes et al. 1991; Marcario et al. 2004). Other evidence seems to agree with the suggestion that DAergic tone is increased in HIV infection. For example, increasing dopamine availability accelerated infection in the simian immunodeficiency virus (SIV) macaque model of HIV infection (Koutsilieri et al. 2002), and exacerbated the neuropathology of SIVE (Czub et al. 2001).

Figure 5
figure 5

Three synapses are illustrated that depict adaptations to experimental manipulation of DAergic tone. The size of the symbol reflects the concentration of each protein in striatal synapses. The synapse at right depicts changes produced when the concentration of dopamine in the synaptic cleft is increased experimentally. TH decreases and DAT increases in the presynaptic bouton; D2R decreases and D3R increases in the postsynaptic bouton. The synapse at left depicts changes in the opposite direction after DAergic tone is decreased, such as in Parkinson's disease. The changes observed in HIVE striatum suggested increased DAergic tone.

Full size image

An increase in striatal DAergic tone could have various clinical consequences. One highly important group of DAergic circuits to consider is the “reward circuitry.” Brainstem DAergic neurons in the ventral tegmental area (VTA) project to the ventromedial part of striatum, including the nucleus accumbens (NA). The NA has rich connections with the limbic lobe and coordinates reward-seeking and motivational behavior. There is evidence that when these circuits are “overloaded,” the increase in synaptic dopamine reinforces reward-seeking behavior, akin to addiction behavior or compulsive gambling in humans (Koob and Le Moal 1997; Esch and Stefano 2004). Addictive drugs such as cocaine produce permanent neural sensitization in brain mesolimbic systems that lead to a compulsive “wanting to take drugs” that can last for years (White and Kalivas 1998). Cocaine increases dopamine availability by directly blocking DAT and reinforces drug-seeking behavior in these reward circuits (Volkow et al. 2004). Hyperdopaminergic mutant mice lacking DAT gene function also exhibit a “high salience for reward” (Esch and Stefano 2004). In turn, chronically increased synaptic dopamine concentration is associated with both pre- and postsynaptic changes in gene expression. The premier example is striatal D2R transcription, which is driven downward when occupancy is high, and increases when receptor occupancy is low (Gerfen et al. 1990; Staley and Mash 1996; Herrero et al. 1996; Segal et al. 1997; Mash and Staley 1999; Quik et al. 2000; Bezard et al. 2001; Wade et al. 2001; Thobois et al. 2004; Nader and Czoty 2005). Decreased expression of striatal D2R is consistently observed in people who are addicted to drugs, particularly cocaine. Volkow et al. (2004) suggest that drug abuse increases D2R occupation, which then drives D2R expression downward. Having less available D2R “desensitizes” the reward circuit and causes the addicted person to seek more stimulation. The decreased striatal D2R expression that we observed in HIVE is therefore consistent with a postsynaptic response to chronic overstimulation, as occurs in cocaine abuse and therapies that stimulate dopaminergic transmission (Thobois et al. 2004; Nader and Czoty 2005). Low striatal D2R expression, for any reason, can increase vulnerability to the reinforcing effects of cocaine (e.g., self-administration in monkeys) (Nader and Czoty 2005). Therefore, decreased D2R in HIVE and the other convergent biochemical changes we have described [D3R, TH, pTH(ser40), DAT] are observed when DAergic tone is increased. These results imply that HIVE could increase vulnerability to situations in which striatal DAergic tone is increased pathologically, such as drug addiction. This suggestion raises several clinically relevant hypotheses for future investigation:

  1. (1)

    HIVE may exacerbate the rewarding properties of cocaine and other addictive drugs by stimulating DAergic circuits and driving down striatal D2R expression.

  2. (2)

    HIV infection could increase need for DAergic reinforcement, which, in turn, can lead to behavior that increases the risk of HIV transmission.

  3. (3)

    Aggressively treating drug addiction in people with HIV/AIDS, by reducing reward-seeking behavior, could decrease the rate of HIV transmission to others.

  4. (4)

    Suppressing HIV replication in the brain, by reducing DAergic tone, could improve the probability of successfully treating drug addiction.

  5. (5)

    Suppressing synaptic dopamine concentration pharmacologically might be useful to prevent HIV-induced overloading of DAergic synapses in some circuits.

These new neurochemical findings are comparable in some aspects with previously reported results using PET scanning. Striatal D2R binding availability was decreased in 10 people with HAD (Wang et al. 2004), but to a lesser extent than the decreased D2L concentration in HIVE. A potential reason for the difference of intensity is that [11C]raclopride binding, as measured in PET, is not specific to D2L. Raclopride binds to D2L, D2S, and D3R. Based on our biochemical quantification of all three of these binding partners, raclopride binding in HIVE would reflect a mixture of increased D3R binding, sharply decreased D2L binding, and modestly decreased D2S binding. The competing effect of increased D3R is one scenario that explains why decreased PET tracer binding is not as pronounced as the decrease in D2L concentration. PET scanning has also suggested that striatal presynaptic neurons are perturbed in HAD, i.e., binding availability of the presynaptic DAergic marker DAT was decreased (Wang et al. 2004). In seeming contrast, DAT protein was sharply increased in HIVE in our panel of changes. Experimentally, it is known that DAT binding availability is strongly influenced by factors other than the concentration of DAT protein or mRNA (Wilson et al. 1996). For example, DAT binding is sensitive to changes in the concentration of dopamine in the synaptic cleft, which competes with exogenous [11C]cocaine PET tracer molecules and limits available DAT binding sites (Gatley et al. 1995, 1997). Thus, an increase in the concentration of synaptic dopamine in HIVE (i.e., increased DAergic tone) could effectively compete with PET tracer molecules and sharply decrease the number of available binding sites. Another potential scenario is that the increased DAT reflects redistribution due to increased membrane internalization and trafficking of the protein, which decreases PET tracer binding due to a lack of access to the extracellular compartment (Melikian 2004; Zahniser and Doolen 2001). And finally, ligand affinity of DAT protein is sensitive to posttranslational modifications that include oligomerization and phosphorylation (Wilson et al. 1996). In sum, the interrelationship between DAT protein, its distribution within in the synapse, its conformational variants, and its binding affinities is very complicated. When comparing PET results to these postmortem protein measurements, it also is important to recognize that the two patient populations were not equivalent: PET was performed on people who were clinically classified to have HAD with unknown neuropathology (Wang et al. 2004), whereas DAT was measured in decedents with a confirmed neuropathological diagnosis of HIVE. HIVE is the neuropathological substrate of HAD in the majority of cases (Wiley and Achim 1994).

We conclude that biochemical measurements in autopsy brain specimens show that striatal DAergic synapses are abnormal in NNTC subjects with HIV encephalitis. Based on correlative changes that are described in the experimental literature, the overall pattern of abnormality in HIVE probably reflects adaptation to increased DAergic tone. More extensive multidisciplinary and translational studies are needed to determine whether abnormal synapses reflect altered physiological function of specific striatal DAergic circuits in the manner that we have suggested. The influence of HIV/AIDS on DAergic systems is ripe for a critical reappraisal.