|Additional related information may be found at:|
|Neuropsychopharmacology: The Fifth Generation of Progress|
The Dopamine Transporter: Potential Involvement in Neuropsychiatric Disorders
The Dopamine Transporter: Potential Involvement in Neuropsychiatric Disorders
Michael J. Bannon , Paola Sacchetti and James G. Granneman
Although constituting only 1 of every million neurons in the mammalian brain, neurons utilizing the neurotransmitter dopamine (DA) have received disproportionate attention by medical researchers because of their known or suspected role in a variety of disorders ranging from Parkinson’s disease to attention deficit disorder, schizophrenia to drug abuse. This emphasis on DA neurotransmission is reflected in the present volume, which includes separate chapters on DA receptors and their signal transduction mechanisms, presynaptic control of DA synthesis and the DA transporter (DAT). The DAT is a plasma membrane transport protein thought to control extracellular DA concentrations through the recapture into DA nerve terminals of DA that has been released during the process of neurotransmission. As discussed below, the DAT is also an important target for a variety of therapeutic agents and drugs of abuse.
DA transport and its accumulation in nerve endings was first characterized in 1969, a number of years after the description of tissue uptake of the related neurotransmitter norepinephrine (NE; for reviews, see 45; 4). During the early 1980’s, the development of radiolabeled probes with some selectivity for the DAT advanced our understanding of the distribution, pharmacology and regulation of the DAT. More recently, brain imaging techniques have enabled us to determine DAT density and occupancy in vivo, while the cloning of the DAT and related neurotransmitter transporters has increased exponentially our knowledge about this important gene family.
Investigations into the cellular and molecular biology of the dopamine transporter (DAT) and its potential involvement in brain disorders have progressed at a breathtaking pace since this chapter was first written 4 years ago. This extensively revised review will provide a brief overview of the structure, function, pharmacology and regulation of the DAT, as revealed by recent analyses of the DAT gene and protein. For more in-depth discussions of these topics, the reader will be referred to a number of recently published, comprehensive reviews. The specific emphasis of this chapter will be an examination of DAT gene expression and regulation in normal and diseased human brain, as suggested by recent brain neuroimaging, postmortem and genetic studies.
DAT SEQUENCE AND STRUCTURE
The DAT is a member of a large family of Na+/Cl--dependent transporters, including the closely related NE transporter (NET) as well as transporters for serotonin, GABA, glycine, proline, creatine, betaine, taurine (for reviews, see 3; 33; 86; 77). Using sequences conserved between the GABA transporter and the NET, DAT cDNAs were cloned from rat (55; 34; 98), cow (107) and human (5; 35; 109) sources. The cloning data predicts that the DAT is a 619 (rat) or 620 (human) amino acid protein. Hydropathy analysis suggests that the DAT includes 12 transmembrane domains (TMDs), with both the amino- and carboxy-termini residing within the cytoplasm (figure 1), consistent with recent immunochemical data (13; 79). Monoamine transporter sequences are least conserved at these termini and a large extracellular loop occurring between TMD 3 and TMD 4 and most conserved in the putative TMDs.
The generation of chimeric transporters produced by exchanging similar domains between the highly related DAT and NET transporters has provided clues as to the functions associated with different DAT domains (9, 10; 37; see figure 1). Different studies have suggested that residues within TMDs 1-3, or alternatively TMDs 9-12, greatly influence substrate affinities. TMDs 1-3 also influence affinity for certain DAT ligands such as GBR12395 (or antidepressants in the case of the NET). TMDs 5-8 may also be critical in determining affinities for DAT and NET inhibitors (e.g. cocaine and desipramine). This observation is independently supported by studies with photoaffinity reagents, demonstrating labeling of amino terminal portions of DAT with a GBR derivative and midportions of the DAT with a cocaine analog (112). In addition, this region affects substrate Vmax. Finally, the affinity of the DA neurotoxin MPP+ for the DAT is influenced by residues within TMDs 11-12.
Studies using chimeric transporters have been complemented by site-directed mutagenesis experiments (53, 54; 73). Mutation of an aspartate (D79) in TMD 1 affects DA affinity and Vmax as well as affinity for a cocaine analog. Mutations within TMDs 7 and 11 affect MPP+ affinity and Vmax. Replacement of serine residues in TMD 7 affects DA transport Vmax more than antagonist binding affinity. Mutation of a specific tyrosine residue in TMD 11 confirms the importance of this region in determining interactions with MPP+.
A number of sequences have been identified which may be subject to posttranslational modification, while other sequences may mediate protein interactions with the DAT. Consensus sequences for multiple (3-4) N-linked glycosylation sites are found in the large extracellular domain situated between TMD 3 and TMD 4 (figure 1). The DAT is extensively glycosylated, with a molecular mass of ~80 kDA prior to deglycosylation, but ~60 kDa after enzymatic treatment (for a review, see 83). Studies have indicated that the nature or extent of DAT glycosylation is somewhat cell-, tissue- and species-dependent, although the physiological significance of this observation is unclear. It has been suggested that differences in glycosylation may contribute to differences in Km between native DAT and DAT expressed in heterologous systems, as well as transporter targeting/stability (83). Cysteine residues present in the large extracellular domain may form disulfide bonds important for protein trafficking; leucine zipper-like motifs within TM2 and TM9 may mediate intra- or intermolecular protein interactions, although there is little experimental evidence in this regard. Radiation inactivation experiments to date have not clarified the potential multimeric nature of the DAT (7; 72).
The DAT also contains a number of potential phosphorylation sites for cAMP-dependent protein kinase, protein kinase C (PKC), and Ca++-, calmodulin-dependent protein kinase. Numerous studies have examined the effects of phosphorylating conditions on DAT activity. The evidence for DAT modulation by phosphorylation is most compelling in the case of PKC. Activation of PKC inhibits DAT-mediated uptake (Vmax) through the rapid sequestration/internalization of DAT protein (128; 87 and refs therein). Recent studies (43; 111) have demonstrated basal [32P]orthophosphate incorporation into the DAT which is increased by PKC activation; although one might assume that the direct phosphorylation of DAT at PKC consensus sequences mediates PKC-induced inhibition/sequestration of the DAT, this has not been verified experimentally.
THE DAT HAS PROPERTIES OF A CARRIER AND A CHANNEL
The DAT and other monoamine transporters utilize ion gradients across the plasma membrane as the driving force for intracellular accumulation of neurotransmitter (for reviews see 59; 90). The DAT exhibits an ion dependence distinct from the NET and serotonin transporter, with an apparent stoichiometry of transport of 2Na+:1Cl-:1DA, suggesting that DA transport is an electrogenic process. Electrophysiological analyses (100) have confirmed directly the electrogenic nature of DAT activity and demonstrated an increase in DAT velocity with hyperpolarization. Several groups (82; 69; 14) have reported that DA receptors located on DA nerve terminals (i.e. DA autoreceptors) modulate the activity of the DAT. Thus DA autoreceptor-mediated hyperpolarization would serve to simultaneously decrease DA release and increase DAT-mediated clearance of extracellular DA, sharpening DA neurotransmission (100). Under some conditions (e.g. drug treatment), DAT-mediated reverse transport of DA can occur (see below).
Electrophysiological studies have further revealed unanticipated electrical properties of transporters not accounted for solely on the basis of substrate and ion cotransport. The DAT exhibits channel-like properties, including substrate-activated currents uncoupled from substrate transport, substrate-independent but cocaine-inhibitable ‘leak’ currents, and ‘transient’ currents. It has been suggested that transport-coupled or -uncoupled currents may alter membrane potential and underlie a novel means of neural signaling (100).
DISTRIBUTION OF DAT GENE EXPRESSION
DAT distribution was determined first by Na+-dependent DA uptake and later by DAT ligand binding. Subsequent to the cloning of DAT cDNAs, DAT mRNA has been visualized and quantified using in situ hybridization, northern blot analysis and nuclease protection assays. DAT protein has been localized by immunochemical means using antisera raised against amino acid sequences deduced by cloning. These recent studies confirm and expand upon our previous understanding of the distribution of DAT gene expression.
The DAT gene is expressed only in the CNS within a small subset of neurons (i.e. DA-containing neurons) and not in glia. DAT expression is more restricted, for instance, than the expression of genes encoding DA biosynthetic enzymes (tyrosine hydroxylase [TH], aromatic amino acid decarboxylase) or DA receptors. DAT therefore provides an excellent marker for most DA neurons and their projections although, as reviewed below, the cellular abundance of DAT is regulated under certain conditions.
In the rodent, DAT mRNA is found in great abundance within midbrain DA neurons of the substantia nigra, with somewhat lower expression in the ventral tegmental nuclei and adjacent nuclei (93; 62 and refs. therein). Within the hypothalamus, DAT is expressed within the A13 (zona incerta) and, to a lesser extent, the A14 (periventricular ) and A12 (arcuate nucleus) cell groups, but not other TH-positive cell groups (70). Moderate DAT expression is also seen in the A16 cell group of the olfactory bulb. DAT mRNA is not found in regions devoid of DA cell bodies or within DA nerve terminals. This distribution has been confirmed using northern blot analysis and nuclease protection assays (125; 62). In human brain, DAT mRNA exhibits the same general distribution as seen in rodents (5). Within human midbrain DA cells, the abundance of DAT mRNA is greatest within the ventral tier of the substantia nigra, followed by the dorsal tier and the ventral tegmental area, with the lowest levels of DAT mRNA seen within the retrorubral field (6; see figure 2).
The distribution of DAT immunoreactivity is largely consistent with other indices of DAT gene expression and the density of DA innervation (17; 25). Thus, the striatum and nucleus accumbens are densely labeled, with labeling also apparent within the globus pallidus, cingulate cortex, olfactory tubercle and amygdala. DAT immunoreactivity is also seen in the perikarya, dendrites and axonal processes of midbrain DA neurons. Regional differences in somatodendritic DAT immunoreactivity covary with DAT mRNA levels and correlate with the susceptibility of subgroups of DA neurons to neurotoxins and idiopathic disease processes (see below). Electron microscopic immunocytochemical studies (78, 79) have revealed that DAT protein is localized primarily on extrasynaptic plasma membranes (boutons en passant) near aggregates of synaptic vesicles, consistent with DAT playing a key role in limiting the spatial domain of DA neurotransmission. Within the dendrites of midbrain DA cells, DAT is localized to plasma membranes and smooth endoplasmic reticulum, consistent with DAT-mediated dendritic DA release and/or DAT modulation of DA cell activity through its channel-like properties (see above). Within perikarya, DAT is localized primarily to tubulovesicles, which may represent DA-containing membranes in transit.
DAT GENE ORGANIZATION
The hDAT gene has been localized to chromosome 5p15.3 (35; 108), cloned and characterized (21; 50; 56). The gene spans 64 kb and is divided into 15 exons separated by 14 introns, with predicted intron-exon junctions. The hDAT coding region begins within exon 2 and extends partially into exon 15. A single transcriptional start site has been identified. There is no evidence for DAT RNA splice variants or the use of multiple polyadenylation sites. The overall exon-intron structure of the hDAT gene closely parallels that of the hNET (and to a lesser extent the serotonin and GABA transporter ) gene. In general, each hDAT exon encodes a functional domain such as the N- or C-terminus or a putative TMD and adjacent hydrophilic loop (figure 1). A number of Alu and other sequence repeats have been located within intronic portions of the gene. Of greater interest is a 40 bp variable tandem nucleotide repeat (VNTR) polymorphic sequence (108; 12; 94) found in the 3’ untranslated region just upstream of the polyadenylation site which may be associated with human diseases. A number of restriction fragment length polymorphisms have also been examined for such associations (see below).
The 5’ flanking sequences (21; 50; 56) controlling transcription of the hDAT gene are interesting in a number of regards. Neither a canonical ‘TATA’ box nor a ‘CAAT’ box are evident, prompting the suggestion that the DAT is a TATA-less gene. In this case, high local GC content and several putative SP1 sites might serve to direct DAT gene transcription. However, evidence of a single hDAT transcriptional start site may be more consistent with the behavior of a TATA-containing promoter. Sequence conservation is evident in a 180 bp region immediately 5’ to the human and murine DAT transcription start site. Within this region, a conserved TAAGA sequence positioned -32 bp relative to the start site may serve as a TATA box (figure 3), although this has not been confirmed experimentally. Aside from several potential SP1 sites, a rather limited number of potential transcription factor response elements (e.g. Egr-1, E-box, AP-2) may reside in the proximal hDAT promoter, although individual elements have not been assessed functionally. Nevertheless, in vitro experiments have demonstrated that proximal hDAT sequences constitute a strong promoter lacking in cell specificity (56; 91).
Recently, over 8 kb of hDAT 5’ flanking sequence has been cloned (91; GenBank accession #AF115382). Within this span of sequence, numerous potential regulatory sequences have been tentatively identified but, as with the core promoter, incompletely characterized. The development of midbrain DA cells is critically dependent on the expression of the nuclear receptor transcription factor nurr1, as proven by targeted disruption ("knockout") of the nurr1 gene (127; 95; 15). Nurr1 gene expression persists in rodent midbrain DA cells through adulthood, suggesting post-developmental functions as well. Interestingly, multiple nurr1 binding sites have been identified in 5’flanking sequences of the hDAT gene (figure 3), and hDAT promoter constructs are activated by nurr1 co-transfection in vitro (91). Similar data have been reported for the TH gene promoter (92). Furthermore, the human homologue of nurr1 (NOT-1) is expressed at high levels in human midbrain DA neurons (91). Hypothalamic DA neurons express much lower levels of both nurr1 and DAT. Thus it is likely that the nuclear receptor nurr1/NOT-1 plays an important role in the maintenance, as well as development, of the strong DAergic phenotype seen in midbrain.
Neuron restrictive silencing elements or related motifs within the promoter (21; 50; 56; 91; see figure 3) may contribute to the extraordinary cell specificity of DAT expression in vivo. DAT proximal promoter sequences are nonselectively active in vitro, yet native DAT expression is silenced in most cells (even within most central neurons), suggesting that a unique combination of positive and negative control elements govern DAT gene expression in vivo. Recent studies in our lab (P.S., J.G. and M.B., unpublished data) have demonstrated that the successive addition of DAT gene 5’ flanking sequences containing putative silencing elements results in profound (but incomplete) inhibition of gene expression in vitro. The relative paucity of response elements for activity-dependent factors (e.g. CREB, AP-1) may be in keeping with the general unresponsiveness of the DAT gene to a variety of stimuli. The lack of an authentic DAT expressing cell line (56) continues to hinder the molecular analysis of the DAT gene. The factors mediating striking disease-related changes (detailed below) remain enigmatic. Deciphering the control of hDAT gene expression remains an important task in the years ahead.
PSYCHOSTIMULANTS AND DAT KNOCKOUT ANIMALS
The DAT is a major target for psychostimulants such as cocaine, amphetamine and methamphetamine. The reinforcing properties of these drugs (which likely underlie their addictive properties) are strongly correlated with their affinities for the DAT (89), in keeping with evidence implicating mesolimbic DA neurons in drug abuse. Cocaine and related drugs bind to the DAT and prevent DA transport. The amphetamines gain access to DA nerve terminals by both lipophilic diffusion across the plasma membrane and DAT-mediated transport, releasing DA from vesicular stores and evoking DAT-mediated DA release. In each instance, psychostimulants raise the extracellular concentrations of DA, likely augmenting DA neurotransmission over longer distances and/or durations, and causing behavioral activation (increased locomotor activity and/or stereotyped behaviors).
Confirmation of these facts has been obtained by examination of DAT gene knockout animals (36; 48). These mice exhibit significant locomotor hyperactivity (but not stereotypies) but no further behavioral activation in response to cocaine or amphetamine. In the absence of DAT, DA clearance time from the extracellular space is markedly prolonged (100-300 fold) and stimulants fail to further augment extracellular DA concentrations. It is interesting that this phenotype is accompanied by only 5 fold increases in extracellular DA, presumably due to profound (75-95%) decreases in the DA biosynthetic enzyme TH, DA levels, and DA release. Reduced expression of basal ganglia D1 and D2 DA receptors is also evident. In spite of low basal expression of DAT in the hypothalamus, the development of anterior pituitary hypoplasia and dwarfism in DAT knockouts suggests an unanticipated importance of the DAT in regulating pituitary function. As would be predicted, the knockout mice are resistant to the DA neurotoxic effects of MPP+ and methamphetamine (28; 29). The principal findings have been independently replicated in a second DAT knockout strain (101).
As outlined above, several lines of evidence strongly suggest that the DAT is the primary locus mediating cocaine reward. It was contrary to predictions, therefore, to find that cocaine exhibits its normal reinforcing properties in DAT knockout mice in two different behavioral models, namely cocaine self-administration and place preference (88; 101). Although it has been suggested that cocaine’s reinforcing effects in DAT knockout mice may involve interactions with the serotonin transporter (88), other data argue against this hypothesis (101). The relevance of cocaine’s effects in DAT knockout mice to our understanding of the reinforcing and addictive properties of cocaine in normal mice (and man) is uncertain. It is possible that cocaine’s reinforcing effects are, in fact, not normally mediated through the DAT but instead a previously unappreciated target site. Alternatively, the DAT could be the primary mediator of cocaine reward in normal brain but, during development in the absence of DAT, another site might be targeted by cocaine due to neural adaptations. The importance of the DAT in mediating cocaine’s effects is supported by a recent PET study which demonstrated that cocaine doses used by human cocaine abusers resulted in 60-80% blockade of DAT sites. Furthermore, the degree of DAT blockade in vivo was highly correlated with the self-reported subjective effects (‘high’; 115). Further animal experiments, perhaps including inducible, cell-specific gene knockouts, may help to resolve the importance of the DAT in cocaine’s behavioral effects.
POTENTIAL ROLE OF DAT IN HUMAN DISORDERS: GENETIC ANALYSES
DA neurotransmission has been implicated in various neuropsychiatric disorders, including Parkinson’s disease, schizophrenia, Tourette’s syndrome, attention-deficit disorder, and substance abuse. While several DA receptor proteins have been identified, the DAT is the only DA-binding protein that functions to remove the neurotransmitter from the synaptic space. Given the nonredundancy of the gene, and its central role in controlling spatial and temporal aspects of DA neurotransmission, the DAT gene has received considerable attention as a candidate gene for DA-related neuropsychiatric disorders.
As mentioned above, characterization of the hDAT cDNA and gene has identified a VNTR (108; 12). The role of VNTRs in gene function remains obscure; however, recent evidence suggests that these repetitive elements could play a role in transcriptional and post-transcriptional gene expression (75). There are currently no data indicating that the VNTR in the DAT gene affects its expression. Nevertheless, the polymorphic VNTR serves as a highly informative marker for association and linkage analyses.
The DAT VNTR is 40 bp sequence that is present in the 3’ nontranslated region of the DAT cDNA and is repeated 3-11 times. The most common allele contains 10 copies of the VNTR. In the US population, the frequency of the allele is 0.7 among Caucasians and Hispanics, and about 0.54 in African Americans (109; 22). Among Asians, the frequency of the 10 copy allele is about 0.9, making the degree of heterozygosity in these populations quite low (76; 58). The degree of ethnic heterogeneity in the DAT VNTR is an important consideration in interpreting disease association studies.
Perhaps the best evidence for involvement of the DAT with a disorder comes from work with attention-deficit hyperactivity disorder (ADHD). ADHD appears to be familial and heritable (8; 32), and is perhaps the most common childhood-onset behavioral disorder (122). It is well recognized that ADHD patients benefit from treatment with certain psychostimulants, such as methylphenidate and amphetamine, which directly interact with the DAT. This connection led Cook et al. (18) to evaluate the DAT as a candidate for susceptibility to ADHD using haplotype-based haplotype relative risk analysis. This study found a significant association of the 10 copy VNTR polymorphism with ADHD and undifferentiated ADHD. Similar results were independently obtained by Gill et al. (31) in an Irish population. More recently, Waldman et al. (117) utilized multiple approaches to confirm both association and linkage of the 10 copy VNTR with ADHD. Interestingly, the relation held for hyperactive-impulsive and combined hyperactive-impulsive/inattentive subtypes, but not for the purely inattentive subtype.
The DAT also has been proposed as a gene candidate for Parkinson’s disease (PD) owing to the potential ability of DAT to transport neurotoxins into the DA neurons that are destroyed in the disease. The rare 11 copy VNTR of the DAT gene was found to be disproportionately represented in a sample of Caucasian PD patients from Australia (57). These results, however, were not replicated in Chinese, Japanese or French populations (58; 41; 85). As mentioned above, case control association studies such as these should be interpreted cautiously, as they are prone to bias owing to sampling error arising from population stratification and founder effects.
With the possible exception of bipolar disorder (51), association analyses of the DAT gene with other DA-related disorders, including schizophrenia (11; 19; 52; 27), Tourette’s syndrome (30), and substance abuse (81; 84) have yielded largely negative results. It must be recognized, however, that there are numerous difficulties in establishing association and/or linkage when dealing with psychiatric disorders. First, diagnostic categories are complex and heterogeneous, and the lack of reliable biological markers is a major impediment in performing such studies. Moreover, the underlying biological basis for these diseases is likely to be polygenetic, and the contribution of certain genes is likely to vary significantly across various populations. With respect to DAT, there appears to be considerable ethnic variability in polymorphic markers. Furthermore, the TaqI and VNTR polymorphisms described for the DAT do not appear to be in linkage disequilibrium, suggesting that the available probes of DAT gene variation might be insufficient to rule out linkage to DA-related disorders (84).
POTENTIAL ROLE OF DAT IN HUMAN DISORDERS: IMAGING AND POSTMORTEM STUDIES
Despite the absence of genetic data convincingly linking the DAT gene to most brain disorders, the DAT could contribute to disease processes as a component of a polygenetic disorder (referred to above) or as a substrate for disease-inducing factors (e.g. DA neurotoxins, chronic stimulants). Alternatively, disease-related changes in DAT expression may reflect compensatory processes in response to DA neuron dysfunction. On occasion, intriguing discrepancies between measures of DAT in postmortem material and in vivo imaging studies have been identified. Such differences may reflect biological characteristics of the DAT (e.g. posttranslational modifications, internalization) or limitations of the techniques employed. Several illustrative cases will be reviewed.
DAT in Parkinson’s disease: PD is characterized by a substantial loss of midbrain DA neurons (particularly in the ventral tier of the substantia nigra; figure 2) with a consequent loss of DA innervation to forebrain structures (particularly the putamen). The vulnerability of certain subgroups of DA neurons in PD and MPTP-induced parkinsonism correlates with higher basal levels of DAT gene expression (93; 4; 6; 105 and refs therein). It is conceivable that avid transport of neurotoxins or even endogenous DA by the DAT may play a role in idiopathic PD.
Given the extent of DA cell loss, it is not surprising that significant decreases in DAT ligand binding sites are detected in PD striatum postmortem (see 4; 71 for reviews). DAT binding is reduced equivalently in progressive supranuclear palsy, a disease involving global degenerative changes throughout the basal ganglia and associated nuclei including the substantia nigra (66; 16). Although initial estimates of DAT losses in PD were unexpectedly modest, this turned out to be (at least in part) an artifact of one of the DAT ligands commonly used (GBR 12935; 97). The recent development of DAT-specific antisera has facilitated the direct quantification and localization of DAT protein in PD brains. Miller et al (71) reported substantial DAT protein reductions in putamen (75%), caudate (64%) and nucleus accumbens (53%). Immunocytochemical staining reveals a similar gradient of DAT loss. Even within the most impacted regions, discrete islands of DAT immunoreactivity persist, in association with the matrix compartment.
One of the more interesting observations in PD relates to alterations in gene expression within the surviving midbrain DA neurons. In PD, the abundance of mRNAs encoding the DAT and vesicular monoamine transporter (VMAT) per DA cell are decreased, while TH mRNA levels per cell are increased, as compared to control brains (106; 40; 49). These data might be interpreted as evidence for the occurrence of compensatory mechanisms within surviving DA cells: an increased production of TH and diminished capacity to recapture and sequester DA may augment DA neurotransmission and partially offset DA cell loss.
A number of cocaine-related DAT ligands of high affinity and reasonable specificity have been developed for positron emission tomography (PET) and single photon emission computerized tomography (SPECT). Recent studies using these reagents have detected regional losses of DAT binding even in relatively early PD cases (24; 26; 38; 44; 104 and refs therein). The ability to monitor DAT (and therefore to some extent DA terminal) density in vivo may facilitate earlier diagnosis of PD and should provide a valuable tool for assessing the efficacy of new treatments aimed at slowing or reversing the disease process.
DAT in Alzheimer’s disease with parkinsonism: Signs of clinical parkinsonism occur in a sizable proportion (20-40%) of patients with Alzheimer’s disease in the absence of classical neuropathology of PD. Alzheimer’s subjects with parkinsonism exhibit a substantial loss of DAT in the caudate-putamen, albeit with a distribution which differs from the loss of DAT in idiopathic PD (2; 74). There is also a significant decline in the number of DAT mRNA-positive DA cells in the midbrain, as well as lower DAT mRNA levels per cell. The profound loss of midbrain DA cells which occurs in PD, however, is not seen in parkinsonian Alzheimer’s subjects, and TH expression is less impacted than the expression of DAT (74; 49). DA-related gene expression, therefore, is impacted differently in Alzheimer’s disease with parkinsonism than in idiopathic PD, although the underlying mechanisms are not understood. In nonparkinsonian Alzheimer’s subjects, DAT and DA systems in general are minimally impacted.
DAT in Wilson’s disease: Wilson’s disease is an autosomal recessive disorder involving mutations of the P-type copper ATPase ATP7B (10a), resulting in excess copper deposition. Liver and brain are most affected. Varied neurological symptoms include parkinsonism, dystonia, and psychosis. Structural changes are seen in numerous brain regions including the striatum, where a loss of D2 receptors occurs. DAT ligand binding in vivo is decreased in Wilson’s disease to the same extent as seen in PD, but without corresponding pathological changes in the substantia nigra (119; 46 and refs therein). These preliminary data are somewhat reminiscent of the changes seen in Alzheimer’s disease with parkinsonism and suggest a preferential loss of DA nerve terminals or a profound change in DAT biosynthesis, transport, or turnover.
DAT in Lesch-Nyhan disease: This X-linked infantile onset disease, resulting from the loss of hypoxanthine-guanine phosphoribosyl transferase activity, leads to compulsive self-injurious behavior and movement disorders such as dystonia and choreoathetosis. Self-injurious behavior can be elicited in rodents by neonatal manipulations of DA function, and DA systems seem to be significantly impacted in Lesch-Nyhan. Striatal DAT density (as determined by PET imaging) was reported to be decreased 50-75% in a small group of Lesch-Nyhan subjects; other measures of DA function and total striatal volume are also significantly affected in this severe disorder (123; 23). The extent of DA cell loss in Lesch-Nyhan is unknown. The comparability of a hypoxanthene-guanine phosphoribosyl transferase gene knockout animal is unclear, since it exhibits no self-injurious behavior and only modest changes in DA function (47).
DAT in Gilles de la Tourette’s syndrome: This disorder is characterized
by symptoms including obsessions, compulsions, coprolalia and involuntary tics. Although the neuropathological mechanisms underlying Tourette’s syndrome are unknown, the therapeutic benefit of decreasing DA neurotransmission is cited as evidence for the involvement of DA systems. A provocative study (99) reported 37-50% increases in the density of DAT ligand binding in postmortem caudate-putamen from a small number of Tourette’s syndrome subjects compared to controls. The concentrations of striatal DA and DA metabolites and DA receptor density were reportedly unchanged. More recently, a SPECT imaging study involving a small number of Tourette’s subjects reported an equivalent (37%) increase in DAT density (64), although a small PET study (124) has not confirmed this observation. Several imaging studies have failed to identify other pre- or postsynaptic abnormalities in DA function which would account for the pharmacological responsiveness of the disorder (see 124). It is speculated that an undetermined perturbation of DA systems in Tourette’s syndrome may lead to a compensatory upregulation of striatal DAT.
DAT in schizophrenia: The longstanding suspicion that DA systems are somehow involved in the etiology, symptomatology and/or treatment response of schizophrenia has led to investigations of DAT expression in this disorder. A functional study (39) reported alterations in the Vmax and Km for DA uptake into cryopreserved nerve terminals from the striatum of schizophrenics relative to age- and sex-matched controls. Numerous subsequent studies have not detected differences in the affinity or density of DAT ligand binding in striatum (see 97). An age-related change in binding of the DAT ligand [3H]GBR 12935 in the prefrontal cortex of schizophrenics versus controls has been reported (42); given the relative paucity of DAT in cortical regions and concerns regarding the specificity of this ligand (see above), confirmatory data using complementary techniques are critical for interpretation of these data.
DAT in chronic stimulant abusers: As discussed above, the DAT represents a major target protein for cocaine and amphetamine-like drugs. Chronic exposure to these drugs might therefore be expected to elicit compensatory changes in DAT expression. The literature on stimulant modulation of DAT expression in rodent and human brain has, however, been confusing to say the least. Using the full complement of reagents available, recent studies are just beginning to clarify the effects of these drugs on DAT expression.
The VMAT is thought to provide a very good index of DA nerve terminal integrity, as it is unaffected by alterations in DA neurotransmission. In postmortem caudate, putamen and nucleus accumbens from chronic cocaine users, VMAT ligand binding and immunoreactivity are unaffected or only marginally decreased (121; 103). Thus it is likely that chronic cocaine abuse does not lead to a significant loss of midbrain DA cells or their nerve terminals. In contrast, a significant decrease in DAT protein has been reported (by quantitative immunoblotting; 121). It is interesting that in this same study DAT ligand binding was unchanged. Other studies have reported increases in DAT ligand binding (61; 102) which are dependent upon the ligand employed (60). In vivo imaging studies using different ligands have reported modest (20%) increases (65) or no change (118) in DAT binding. The discrepancy between DAT protein and binding could be related to posttranslational modifications or conformational states of the DAT (121). Importantly, the decrease in DAT protein immunoreactivity is consistent with cocaine-induced decreases in DAT mRNA in rodents (125) and, even more prominently, in human postmortem tissue (MJB, JGG et al., unpublished data). These data suggest that hDAT biosynthesis is decreased and its cellular disposition altered in response to chronic exposure to cocaine. One might speculate that decreased DAT expression is a homeostatic mechanism serving to reduce the number of cocaine target sites, thus limiting cocaine-induced fluctuations in extracellular DA levels. It is interesting in this regard that cocaine addicts show reduced extracellular DA in the striatum in response to the DAT blocker methylphenidate (116). Decreased DAT concentrations might also limit DA reuptake capacity and account for the somewhat diminished tissue levels of DA seen in chronic cocaine users, leading to dysphoria and placing these individuals at risk for the development of extrapyramidal system abnormalities (121).
In experimental animals (including nonhuman primates), repeated exposure to methamphetamine produces long-lasting depletions of striatal DA and other presynaptic DA markers, which are thought to reflect DA neurotoxicity (120; 67 and refs therein). A recent human postmortem study (120), however, found that the density of VMAT ligand binding (as well as the DA biosynthetic enzyme aromatic amino acid decarboxylase) was unchanged in chronic methamphetamine abusers. In contrast, in findings reminiscent of data seen in chronic cocaine users (above), methamphetamine abusers had substantially lower DA and DAT levels (39-55% and 25-53%, respectively), as well as slightly (20%) lower TH levels in caudate, putamen and nucleus accumbens (120). In this study, a better correspondence between changes in DAT protein and DAT ligand binding was evident. A recent PET study found reduced (23-25%) striatal DAT density in chronic methamphetamine abusers abstinent for months to years (67). These data suggest that chronic methamphetamine abuse does not cause a severe DA denervation although, as mentioned above, it may be linked to dysphoria and predisposition to subsequent DA-deficiency abnormalities. Although there are other possible interpretations of the data obtained from chronic cocaine and methamphetamine abusers, it seems plausible that stimulant-induced down-regulation of DAT gene expression underlies the observed changes in DA and TH levels, paralleling the sequellae of knocking out the DAT gene in mice (reviewed above).
DAT in aging: For some time, aging was thought to play an important contributory role in several neurodegenerative disorders such as PD and Alzheimer’s disease, but better characterization of underlying processes has led to a deemphasis of this relationship. Age-related changes in DAT gene expression warrant review based on the magnitude of the effects reported, even if the clinical significance of age-related changes in DA function is not understood at this time.
Although there has been some controversy on this point, histological and VMAT binding analyses generally indicate that DA cell number and the density of striatal DA innervation decline at a rate of approximately 6-7% per decade in human brain (68; 96; 63 and refs therein). In vivo imaging studies have reported similar age-related declines in DAT ligand binding (113, 114; 110). Interestingly, a more rapid and substantial age-related loss of DAT binding is seen in postmortem studies (1; 20; 80; 126).
An age-related loss of midbrain DAT mRNA abundance is also evident, with especially dramatic reductions occurring during the fifth decade of life (4, 5). Within the same subjects exhibiting a profound reduction of DAT mRNA, a more modest loss of striatal DAT ligand binding is evident (4). A positive correlation has been observed between initial levels of DAT mRNA in distinct subpopulations of DA cells and the subsequent age-related loss of DAT expression; cells of the ventral tier of the substantia nigra (figure 2) are most impacted (6). The decline in DAT mRNA during aging is due to a reduced DAT mRNA abundance per cell and the appearance of a high proportion of melanin- and TH mRNA-positive DA cells no longer exhibiting detectable DAT gene expression (6).
These provocative data have been extended by a recent analysis of DAT protein within DA cell perikarya. Ma et al. (63) used unbiased stereologic cell counting to quantify the number of DAT-immunoreactive and melanin-containing nigral neurons. During normal aging, the number of intensely labeled DAT-positive cells decreased by over 11% per decade, while the number of lightly DAT-stained and DAT-negative cells increased over time. The total number of DA cells decreased by less than 7% per decade overall. Beginning in the fifth decade of life, there was a marked reduction in midbrain DAT. Relative to younger (0-49 year old) subjects, there was a 75% and 88% reduction in intensely labeled cells in middle aged (50-69) and more aged (70-85) subjects, respectively (63).
These data suggest a complex age-dependent change in DAT gene expression. HDAT gene expression appears to be diminished significantly with age, as evidenced by decreased abundance of DAT mRNA. Perikarya DAT protein, mostly DAT protein in transit (see discussion above) is proportionately reduced. DAT ligand binding to striatal membrane preparations (reflecting both plasma membrane and intracellular sites) is impacted to a lesser extent. DAT ligand binding in vivo (reflecting surface [i.e. functional] DAT alone) decreases only in proportion to the age-related loss of DA cells and does not mirror the more severe losses seen in postmortem tissue analysis. These data suggest important compensatory changes in posttranslational modifications, trafficking, and/or recycling of the DAT. An overall diminution of DAT abundance may underlie other age-related changes in DA phenotype, as occurs in the DAT gene knockout mouse and may occur during chronic stimulant abuse (see above). Potential commonalities mean that insights gleaned from one clinical or experimental situation may elucidate more generally applicable mechanisms.
We have witnessed an explosion of new information about the DAT since the previous version of this chapter was published (4). Some advances, such as the channel-like properties of the DAT, were unanticipated. Other advances, such as the development of more routine procedures for in vivo imaging of the DAT, have proceeded more or less as expected. The exploitation of molecular biological techniques has facilitated the heterologous expression and methodical manipulation of DAT sequences but has not yet yielded DAT-binding cocaine antagonists. DAT gene knockout experiments have confirmed with breathtaking speed our hard-won knowledge of DAT physiology and pharmacology accumulated over a period of decades. Although attempts to directly link the DAT gene to neuropsychiatric diseases have been (by and large) disappointing, altered DAT expression in a variety of diseases suggests that the DAT plays a critical role in normal and abnormal DA neurotransmission. In short, although our new-found knowledge of the DAT has not been translated into major diagnostic or therapeutic advances to date, this is likely to occur in the not-so-distant future.