|Accueil | Aperçu | Revues | Soumettre | Nous Contacter | English|
Dopamine was first identified as a neurotransmitter localized to the midbrain over 50 years ago. The dopamine transporter (DAT; SLC6A3) and the vesicular monoamine transporter 2 (VMAT2; SLC18A2) are two regulators of dopamine homeostasis in the presynaptic neuron. DAT transports dopamine from the extracellular space into the cytosol of the presynaptic terminal. VMAT2 then packages this cytosolic dopamine into vesicular compartments for subsequent release upon neurotransmission. Thus, DAT and VMAT2 act in concert to move transmitter efficiently throughout the neuron. The accumulation of dopamine in the neuronal cytosol can trigger oxidative stress and neurotoxicity, suggesting that the proper compartmentalization of dopamine is critical for neuron function and risk of disease. For decades, studies have examined the effects of reduced transporter function in mice (e.g. DAT-KO, VMAT2-KO, VMAT2-deficient). However, we have only recently been able to assess the effects of elevated transporter expression using BAC transgenic methods (DAT-tg, VMAT2-HI mice). Complemented with in vitro work and neurochemical techniques to assess dopamine compartmentalization, a new focus on the importance of transporter proteins as both models of human disease and potential drug targets has emerged. Here we review the importance of DAT and VMAT2 function in the delicate balance of neuronal dopamine.
Dopamine was first identified in 1910, but the compound was thought to be merely an intermediary chemical in the synthesis of norepinephrine for decades (Hornykiewicz, 2002). In 1958, dopamine was identified in the brain (Carlsson et al., 1958), and the massive depletion of dopamine was linked to Parkinson’s disease in the 1960s (Ehringer and Hornykiewicz, 1960; Hornykiewicz, 1966). Dopamine neuron populations and pathways were soon identified as responsible for reward, cognition, and motor function (Haber, 2014; Swerdlow and Koob, 1987). Dopamine receptors are not saturated upon typical neurotransmission, suggesting that the dopamine system is capable of a large range of signaling intensities, the uppermost of which are not tapped at baseline. Thus, the dopamine system is not acting at its maximum function, leaving room for novel manipulations that could augment dopaminergic neurotransmission (Rice and Cragg, 2008). Here we discuss transgenic mouse models of altered dopamine output achieved by modifying dopamine compartmentalization and storage within the neuron terminal (Table 1). These mice provide novel insights into the neurochemical, behavioral, and protective or pathological effects of modified dopamine dynamics at the synapse.
Dopamine is synthesized via the conversion of the amino acid tyrosine into L-dihydroxyphenylalanine (L-DOPA) by the rate-limiting enzyme tyrosine hydroxylase (TH) (Axelrod and Weinshilboum, 1972). L-DOPA is then converted into dopamine by amino acid decarboxylase (AADC) in the neuronal cytosol (Kuhar et al., 1999). Several studies have shown that accumulation of dopamine in the cytosolic space can lead to deleterious consequences (Caudle et al., 2007; Chen et al., 2008; Masoud et al., 2015; Mosharov et al., 2009). In a unique study, Mosharov and colleagues directly measured cytosolic dopamine in cultured neurons using intracellular patch electrochemistry (Mosharov et al., 2009). They demonstrated that manipulations that increased cytosolic dopamine levels (such as inhibiting dopamine metabolism) exacerbate toxicity, whereas reducing cytosolic dopamine levels (by inhibiting conversion of L-DOPA to dopamine or by increasing VMAT2 levels) is protective and improves cell survival. These results suggest that cytosolic dopamine levels can directly determine toxic outcomes in dopaminergic cells.
Cytosolic dopamine is highly reactive and causes toxicity by two mechanisms: deamination by cytosolic enzymes or autooxidation. Dopamine can be deaminated by mitochondrial monoamine oxidase (MAO), which converts cytosolic dopamine to DOPAL, a reactive aldehyde intermediate, and hydrogen peroxide (Eisenhofer et al., 2004; Goldstein et al., 2013; Rees et al., 2009). Luckily, DOPAL is most often converted to DOPAC, a relatively benign metabolite, via aldehyde dehydrogenase. However, our understanding of DOPAL toxicity in disease pathology, like that seen in Parkinson’s disease, is growing. DOPAL itself can be oxidized, creating reactive oxygen species (Eisenhofer et al., 2004). DOPAL has been shown to induce neurotoxicity in numerous culture systems and in a mouse model lacking cytosolic and mitochondrial aldehyde dehydrogenases (Kristal et al., 2001; Mattammal et al., 1995; Rees et al., 2009; Wey et al., 2012). Additionally, DOPAL is increased in the post mortem brains of individuals with Parkinson’s disease, suggesting a dysfunctional compartmentalization or breakdown of dopamine in those affected by the disease (Goldstein et al., 2011).
Cytosolic dopamine can be also autooxidized to form reactive oxygen species including hydroxyl radicals, superoxide, and hydrogen peroxide (Graham, 1978; Jenner, 2003). Oxidized dopamine can then be converted to dopamine-quinones and protein function-altering cysteinyl adducts (LaVoie and Hastings, 1999; Lotharius and O’Malley, 2001; Rabinovic et al., 2000). Additionally, dopamine synthesis proteins have been reported to interact with the vesicular monoamine transporter 2 (VMAT2), helping to accumulate dopamine in synaptic vesicles from the cytosolic space perhaps as a protective mechanism. For example, TH and AADC have been immunoprecipitated with VMAT2, and vesicular filling is reduced following the disruption of these interactions (Cartier et al., 2010). Based on the detrimental effects of cytosolic dopamine, it has been speculated that the relative distribution of dopamine within the neuronal terminal, possibly mediated by vesicular storage and plasma membrane uptake, may be responsible for a neuron’s vulnerability to injury (Guillot and Miller, 2009; Miller et al., 1999b; Uhl, 1998).
Transport of dopamine across the plasma membrane is an important process that contributes to overall dopamine neurotransmission and compartmentalization. This process is mediated by the dopamine transporter (DAT, SLC6A3), a membrane protein located on dopaminergic cells. Similar to other monoamine transporters, DAT has 12 transmembrane domains with intracellular N- and C- termini and belongs to the SLC6A family of Na+/Cl−-dependent symporters (Gainetdinov and Caron, 2003). In particular, DAT couples the active transport of dopamine with the movement of one Cl− and two Na+ ions along the concentration gradient. This concentration gradient is created by the plasma membrane Na+/K+ ATPase and serves as the driving force for DAT-mediated dopamine uptake (Gether et al., 2006; Kanner and Schuldiner, 1987). Dopamine translocation across the plasma membrane occurs as a result of conformational changes in DAT. The uptake cycle begins when DAT is open to the extracellular space in an outward facing state (Reith et al., 2015). In this conformation, Na+ and Cl− ions bind to DAT and prepare the transporter for dopamine binding. Upon binding of dopamine, the extracellular gate closes, generating an occluded DAT state. Importantly, dopamine binding induces a conformational change allowing the transporter to open on the cytosolic side. In this inward facing state, dopamine and ions dissociate from DAT. Finally, the cycle is reset once DAT returns to the outward facing conformation (Reith et al., 2015).
The function of DAT is to rapidly transport dopamine from the extracellular space into the cytosol of the presynaptic neuron. At the plasma membrane, DAT is located peri-synaptically, where it removes extracellular dopamine and provides spatial and temporal control of the dopamine signal (Cragg and Rice, 2004; Hersch et al., 1997; Jones et al., 1998). In dopaminergic brain regions such as the striatum and nucleus accumbens, DAT provides the principal mechanism of clearing extracellular dopamine and terminating neurotransmission (Ciliax et al., 1995). Aside from modulating the dynamics of released dopamine, DAT is also responsible for recycling the neurotransmitter back into the dopaminergic cell, allowing it to be reused (Sotnikova et al., 2006). By loading the presynaptic neuron with dopamine, DAT directly contributes to the buildup of cytosolic dopamine and indirectly influences vesicular dopamine as well. Accumulation of cytosolic dopamine can produce neurotoxicity as previously discussed. Hence, DAT is a key player in dopamine compartmentalization that can have significant consequences for the presynaptic neuron. Collectively, DAT regulates the concentrations of both 1) extracellular dopamine at the synapse and 2) intracellular dopamine within the presynaptic neuron.
DAT is the primary target for many compounds including psychostimulants, medications and neurotoxicants (Miller et al., 1999b; Torres et al., 2003). Since uptake of dopamine is dependent on DAT, pharmacological manipulation of DAT can produce profound effects on dopamine neurotransmission. Two classical psychostimulants that operate by altering DAT function are cocaine and amphetamine. Cocaine binds to DAT and blocks the transport of dopamine from the extracellular space to the presynaptic neuron (Ritz et al., 1987). Cocaine is a competitive inhibitor of dopamine transport because its binding site overlaps with dopamine’s site of action, precluding the endogenous substrate from binding (Beuming et al., 2008). As a result, dopamine accumulates in the extracellular space where it can reinforce downstream signaling. Conversely, amphetamines (amphetamine, methamphetamine, MDMA) compete with dopamine to enter dopaminergic cells, acting as a substrate for DAT (Sulzer et al., 2005). Once inside the cell, amphetamine disrupts the proton gradient required for vesicular storage of dopamine (Sulzer et al., 1995). This leads to leakage of dopamine from the vesicles into the cytoplasm of the presynaptic neuron. Ultimately, accumulation of cytosolic dopamine in combination with the actions of amphetamine on DAT cause a reversal of the transporter, resulting in efflux of intracellular dopamine into the extracellular space. This DAT-mediated release of dopamine produces a surge in dopamine signaling in response to amphetamine. While cocaine and amphetamine can also produce other effects in the CNS, it is the manipulation of DAT function that directly enhances dopamine neurotransmission and is thought to underlie the reinforcing properties of these psychostimulants (Donovan et al., 1999; Howell and Kimmel, 2008). In addition to enhancing extracellular dopamine levels, these psychostimulants have also been shown to activate phasic dopamine signaling events causing release of dopamine that contributes to drug reinforcement (Aragona et al., 2008; Daberkow et al., 2013; Wanat et al., 2009).
Drugs that modulate DAT are also used for therapeutic purposes in diseases such as attention deficit hyperactivity disorder (ADHD), depression, narcolepsy, obesity and drug dependence. Indeed, dextroamphetamine and methylphenidate, both act by modifying DAT function and are the main components of the leading medications for ADHD, Adderall and Ritalin, respectively (Gether et al., 2006; Gowrishankar and Hahn, 2014). In particular, dextroamphetamine causes DAT reversal and dopamine release similar to amphetamine, while methylphenidate is a DAT inhibitor. These drugs relieve disease symptoms by increasing dopamine signaling which is disrupted in ADHD patients (Sagvolden et al., 2005; Volkow et al., 2010). In addition to dopamine, these drugs also elevate norepinephrine levels which may contribute to their therapeutic effects (Del Campo et al., 2011). Furthermore, some antidepressants like bupropion also block DAT-mediated dopamine uptake and derive at least part of their therapeutic benefit from this inhibition (Learned-Coughlin et al., 2003). Modafinil, prescribed for the treatment of narcolepsy, is a selective DAT inhibitor that elevates dopamine transmission and promotes wakefulness (Volkow et al., 2009). Since the dopamine system has several physiological roles (cognition, motivation, reward, motor control), inhibiting DAT function provides a unique therapeutic opportunity to enhance dopamine signaling and impact diverse behaviors.
In addition to these examples, several compounds can inhibit DAT function with varying levels of selectivity and potency. Initially, it was postulated that all DAT inhibitors would have cocaine-like stimulant and reinforcing properties (Ritz et al., 1987). However, over the past 10–15 years, accumulating evidence has challenged this notion, showing heterogeneity among DAT inhibitors (Schmitt et al., 2013). In fact, different compounds preferentially bind and stabilize distinct structural states of DAT. Typical DAT inhibitors such as cocaine and methylphenidate have been shown to stabilize the outward facing conformation and produce locomotor stimulation and behavioral reinforcement (Loland et al., 2007). However, atypical DAT inhibitors such as modafinil, buproprion and vanoxerine (GBR12909) tend to promote occluded/inward facing conformations (Schmitt et al., 2013). Interestingly, these compounds also lack cocaine-like behavioral effects and possess limited rewarding properties (Schmitt and Reith, 2011). Very recently, ligands that bind to allosteric sites on DAT have been identified and shown to block dopamine uptake as well (Janowsky et al., 2016). Hence, it seems that the specific pharmacological profile of a drug and its behavioral effects are heavily dependent on how the drug interacts with DAT and which structural conformation is favored. In general, DAT antagonists lock the transporter in a certain structural state, preventing the conformational transitions that are required to shuttle dopamine across the plasma membrane (Reith et al., 2015). Taken together, these data exemplify 1) the diversity of DAT ligands and 2) the responsiveness of DAT to different types of pharmacological manipulation.
As a plasma membrane transporter, DAT also provides a site of entry into dopaminergic cells. The most potent dopaminergic toxicants, 6-hydroxydopamine and MPP+, are substrates of DAT (Gainetdinov et al., 1997; Miller et al., 1999b; Schober, 2004). These compounds are used to mimic symptoms of Parkinson’s disease in animal models because they cause robust degeneration of dopaminergic cells. Since 6-hydroxydopamine is a structural analog of dopamine, it can hijack the DAT-mediated uptake mechanism to access dopamine cells. It should be noted that 6-hydroxydopamine is also a substrate for the norepinephrine transporter and thus, must be administered specifically to dopaminergic regions to exert its toxicity. With regards to MPTP, after crossing the blood brain barrier, this compound is converted to its toxic metabolite MPP+ by monoamine oxidase. MPP+ is specifically translocated into dopaminergic neurons by DAT (Javitch et al., 1985). Once these toxins accumulate in dopaminergic cells, they cause oxidative stress and mitochondrial dysfunction, culminating in neurotoxicity (Abdulwahid Arif and Ahmad Khan, 2010; Miller et al., 1999a; Simola et al., 2007). Hence, DAT provides a molecular gateway for toxicants to selectively access and damage dopaminergic cells.
In summary, various compounds produce significant effects in the brain as a result of their actions on DAT. Given this rich pharmacology, manipulation of DAT function occurs in different ways: 1) by inhibiting DAT and causing buildup of extracellular dopamine, 2) by reversing DAT and causing release of dopamine and 3) by acting as substrate of DAT and using the transporter to access dopamine cells. Interestingly, compounds that increase DAT activity are currently lacking.
The critical role of DAT in maintaining appropriate dopaminergic function is clearly demonstrated by DAT knockout mice (DAT-KO). Genetic ablation of this plasma membrane transporter produces dramatic changes in extracellular and intracellular dopamine dynamics (Giros et al., 1996; Jaber et al., 1997; Jones et al., 1998). DAT-KO mice display 5-fold elevated extracellular dopamine levels due to lack of uptake. Additionally, dopamine remains in the extracellular space 300 times longer since diffusion is the only mechanism to clear the neurotransmitter in DAT-KO mice. Conversely, intracellular dopamine content is reduced by 95% demonstrating that DAT-mediated recycling of dopamine is chiefly responsible for maintaining presynaptic dopamine levels (Sotnikova et al., 2005). Due to depleted intracellular stores, evoked dopamine release is also diminished by 75% in DAT-KO mice (Jones et al., 1998). These neurochemical changes illustrate the vital role of DAT in balancing dopamine levels across different cellular compartments.
Furthermore, lack of DAT activity also triggers compensatory alterations in other pre- and post-synaptic markers of the dopamine system. Striatal post-synaptic D1 and D2 receptors are downregulated by 60% and 40% respectively, to adapt to high extracellular dopamine (Ghisi et al., 2009). Presynaptic D2 autoreceptors are also desensitized, disrupting regulatory negative feedback mechanisms (Giros et al., 1996; Jones et al., 1999). Levels of dopamine metabolites, HVA and 3-MT, are increased suggesting that dopamine degradation may be altered in these mice (Jones et al., 1998). Without DAT-mediated dopamine recycling, presynaptic dopamine levels in DAT-KO mice are solely dependent on synthesis by TH. Paradoxically, while TH expression is reduced, dopamine synthesis rates are doubled, highlighting major adaptive changes in attempts to stabilize dopamine levels in DAT-KO mice (Jaber et al., 1999; Jones et al., 1998). Behaviorally, these animals show spontaneous hyperlocomotion and impaired habituation as a result of increased extracellular dopamine (Giros et al., 1996). DAT-KO mice also display disturbances in cognition (Weiss et al., 2007) and sensorimotor gating (Barr et al., 2003; Ralph et al., 2001; Yamashita et al., 2006). Pharmacologically, DAT-KO mice are insensitive to the classical stimulant actions of cocaine and amphetamine but show paradoxical calming effects instead (Gainetdinov et al., 1999; Giros et al., 1996). In particular, when treated with cocaine or amphetamine, dopamine release and locomotor activity are not enhanced in DAT-KO mice, validating that transporter function is compulsory for psychostimulant effects (Giros et al., 1996). Also, DAT-KO mice are completely resistant to nigrostriatal damage induced by MPTP (Bezard et al., 1999; Gainetdinov et al., 1997) demonstrating that DAT-mediated uptake of MPP+ is required for neurotoxic effects. Collectively, mice lacking DAT show dramatic neurochemical, adaptive and behavioral changes in the dopamine system as summarized in Table 1. Investigation of DAT-KO mice has contributed essential knowledge on the physiological role of DAT as well as the importance of this transporter as a pharmacological target.
On the other end of the spectrum are mice that over-express DAT. DAT was selectively over-expressed in dopaminergic neurons of DAT-transgenic (DAT-tg) mice using bacterial artificial transgenesis (Salahpour et al., 2008). DAT-tg mice display approximately a 3-fold increase in total striatal DAT protein and a 30% increase specifically in the synaptic plasma membrane fraction. Functionally, this translates to a 40–50% increase in dopamine uptake. Due to increased dopamine clearance, DAT-tg mice also have about a 40% reduction in extracellular dopamine levels (Salahpour et al., 2008). While increased DAT expression is expected to enhance dopamine accumulation in the presynaptic neuron, both striatal dopamine tissue content and evoked dopamine release were reduced in DAT-tg mice (Masoud et al., 2015). These results reflect compromised integrity of dopamine neurons in DAT-tg mice as evidenced by 30–40% loss of midbrain dopamine cells (Masoud et al., 2015). Neuronal loss is accompanied by increases in dopaminergic markers of oxidative stress, cysteinyl-dopamine and cysteinyl-DOPAC. These detrimental consequences in DAT-tg mice are likely due to the neurotoxic effects of high cytosolic dopamine. Although it is difficult to assess cytosolic dopamine concentrations in an intact animal, increased dopamine-to-metabolite ratios in these mice indicate buildup of cytosolic dopamine. Behaviorally, DAT-tg mice display fine motor deficits that are reversed by L-DOPA treatment, suggesting that dopamine neuronal loss negatively impacts motor behavior that can be rescued by restoring dopaminergic tone (Masoud et al., 2015). Pharmacologically, DAT-tg mice are particularly vulnerable to MPTP toxicity and show enhanced response to amphetamine, as expected (Masoud et al., 2015; Salahpour et al., 2008). These outcomes reinforce the physiological importance of proper dopamine compartmentalization in the vulnerable nigrostriatal pathway. Characteristics of these mice are summarized in Table 1. Since DAT-tg mice display phenotypes related to Parkinson’s disease (loss of dopamine cells, reduced dopaminergic tone, oxidative stress, L-DOPA responsive motor deficits), characterization of these mice could inform potential disease mechanisms.
Although the physiological actions of DAT are well-elucidated, deciphering the role of this transporter in disease pathogenesis has been complex. In fact, the DAT protein sequence appears to be highly conserved, possibly as an evolutionary mechanism to preserve appropriate function of the dopamine system (Vandenbergh et al., 2000). Evidence surrounding the involvement of DAT in neurological and psychiatric diseases has only recently been uncovered and is still developing. The first genetic condition directly caused by DAT mutations was discovered in 2009 (Kurian et al., 2009). DAT deficiency syndrome (DTDS) is caused by autosomal recessive loss-of-function mutations in the DAT gene (SLC6A3) (Kurian et al., 2011, 2009). In vitro analyses indicate that the mutations associated with this disease obstruct maturation of the DAT protein, leading to reduced transporter expression at the plasma membrane. As a result, overall dopamine uptake is impaired triggering a complex motor disorder characterized by progressive parkinsonism-dystonia. Typically, the syndrome manifests in infancy and severely reduces life expectancy (Ng et al., 2014). The striking phenotypes in DTDS patients convincingly demonstrate the significance of DAT genetics in controlling motor behavior.
The most common neurodegenerative movement disorder in humans is Parkinson’s disease. Motor symptoms of Parkinson’s disease are characterized by muscle rigidity, postural instability, resting tremor and bradykinesia (Jankovic, 2008). These motor deficits are caused by degeneration of nigrostriatal dopamine neurons that lead to reduced dopaminergic tone in the basal ganglia (Dauer and Przedborski, 2003). Unlike DTDS, concrete evidence of a causal link between genetic DAT mutations and Parkinson’s disease is lacking. This is probably because the etiology of Parkinson’s disease is multifactorial and genetic mutations account for only a small proportion of cases (5–10%) (Dauer and Przedborski, 2003; Sulzer, 2007). However, neuroanatomical analyses indicate that regions of the human brain containing the highest levels of DAT protein – the caudate and putamen – are most sensitive to damage in PD (Miller et al., 1997), suggesting that DAT may act as risk factor. The potential role of DAT in enhancing vulnerability of dopamine neurons in PD is two-fold: first, it functions to increase the pool of cytosolic dopamine, which is highly reactive and second, it allows toxicants such as MPTP selective access to dopaminergic cells. Hence, DAT activity could sensitize dopamine neurons to both intrinsic oxidative stress as well as extrinsic environmental insult. In fact, a study by Ritz et al demonstrates that DAT genetic variants in combination with pesticide exposure can increase risk of PD by several fold (Ritz et al., 2009). These DAT variants include single nucleotide polymorphisms in the 5′ region as well as variable number tandem repeats (VNTR) at the 3′ region of the gene. Although the functional consequences of these DAT variants are unclear, these results highlight the importance of considering genetic and environmental interactions in PD (Kelada et al., 2006; Sulzer, 2007). In summary, while DAT mutations produce drastic childhood-onset motor syndromes like DTDS, in a progressive age-related disorder like PD, DAT is more likely to play a modulatory role in combination with other risk factors.
Another disorder that has received considerable attention with regards to DAT dysfunction is ADHD (Gowrishankar and Hahn, 2014). There are two reasons that spurred particular interest in ADHD: 1) DAT is fundamental in dopamine neurotransmission and a large body of work indicates that dopamine signaling is dysregulated in ADHD and 2) the most common pharmacological therapies for this disorder, dextroamphetamine and methylphenidate, directly target and antagonize DAT. Hence, genetic links between DAT and ADHD have been investigated. Initially, Cook et al. reported that the VNTR polymorphism in the 3′ untranslated region of the DAT-1 allele conferred significant risk for developing ADHD (Cook et al., 1995). Other independent studies including a meta-analysis confirmed this result (Chen et al., 2003; Gill et al., 1997; Yang et al., 2007). While the precise molecular basis for this association is unclear, it has been shown that the VNTR polymorphism can impact expression of the DAT gene (Michelhaugh et al., 2001; Mill et al., 2002). Aside from polymorphisms in the non-coding region, rare mutations in the coding sequence of the DAT gene have also been revealed in ADHD. In particular, a point substitution, Ala559Val was discovered in two male siblings diagnosed with ADHD (Mazei-Robison et al., 2008). When tested in vitro, this variant caused abnormal DAT-mediated efflux of dopamine (Mazei-Robison et al., 2008). In vivo analysis of DAT Val 559 knock-in mice confirmed in vitro findings by demonstrating elevated extracellular dopamine levels consistent with DAT-mediated leakage of cytosolic dopamine (Mergy et al., 2014). Basally, these mice demonstrate context-dependent hyperactivity when handled, which may represent phenotypes similar to ADHD patients. In response to amphetamine, these mice show diminished dopamine release and blunted locomotor activation (Mergy et al., 2014), suggesting reduced sensitivity to psychostimulant effects. Hence, the Ala559Val DAT mutation produces noteworthy changes in dopamine dynamics and behavior that may relate to ADHD pathogenesis. In another subject with ADHD, Sakrikar and colleagues described an additional DAT coding variant, Arg615Cys (Sakrikar et al., 2012). This mutation affects the C-terminus of the DAT protein and gives rise to a transporter that is constitutively recycled. Hence, DAT localization and trafficking are largely impacted by this mutation. Taken together, these findings suggest that genetic alterations in DAT structure, expression or function can impose an increased risk for the development of ADHD, a disorder that is linked to aberrant dopamine signaling (Sagvolden et al., 2005; Volkow et al., 2010).
In addition to ADHD, point mutations in the DAT gene have been associated with other neuropsychiatric diseases as well. The Ala559Val DAT variant that was detected in 2 ADHD patients, was also identified in subjects with bipolar disorder and autism spectrum disorder, suggesting that altered DAT function and consequently, modified dopamine signaling may represent a common pathological pathway for these diseases (Bowton et al., 2014; Grünhage et al., 2000; Mergy et al., 2014). Furthermore, a whole-exome sequencing study in families with autism spectrum disorder uncovered another missense DAT mutation, Thr356Met. Similar to the Ala559Val DAT variant, the Thr356Met DAT mutation also produced persistent efflux of dopamine, although the two variants differed in other characteristics (Gowrishankar and Hahn, 2014; Hamilton et al., 2013). Interestingly, transgenic expression of the mutated human Thr356Met DAT in drosophila produced hyperactivity (Hamilton et al., 2013), implying possible overlap in the mechanisms of ADHD and autism spectrum disorder. Indeed, a significant proportion of individuals with autism spectrum disorder concurrently display ADHD symptoms (Goldstein and Schwebach, 2004). Another systematic screening approach of the DAT coding sequence revealed a missense substitution, Glu602Gly in an individual with bipolar disease that was inherited from an affected parent (Grünhage et al., 2000). The role of this mutation in disease mechanisms is currently unknown since the Glu602Gly mutant shows similar surface DAT expression and dopamine transport as wild-type DAT in vitro (Mazei-Robison and Blakely, 2005). This raises the possibility that there may be other properties of DAT dynamics and regulation that have not yet been identified in this mutant. Finally, when DAT coding exons were sequenced in a cohort of patients with atypical movement disorders, it revealed 2 variants Ile312Phe and Asp421Asn (Hansen et al., 2014). Remarkably, both mutations were detected in an individual with comorbid adult-onset neurodegenerative parkinsonism and ADHD. DAT-Ile312Phe was inherited from one of the parents while DAT-Asp421Asn appeared to be a de novo mutation. In vitro, both mutations result in reduced dopamine uptake capacity and the Asp421Asn DAT variant also shows anomalous efflux of dopamine (Hansen et al., 2014). The comorbidity in this individual suggests that DAT variants could provide a unifying mechanism between movement and psychiatric disorders.
In summary, mutations in the coding and non-coding regions of the DAT gene have been associated with various motor and neuropsychiatric disorders. While the mechanisms of these mutations in disease pathogenesis are yet to be fully elucidated, they have been shown to cause changes in dopamine signaling that is important for multiple brain functions. Identification of discrete DAT mutations in rare individuals is a significant finding given the genetic conservation of DAT and multifactorial nature of complex brain disorders.
Neurotransmission relies on the proper storage and release of neurotransmitters from synaptic vesicles. Without precise control of the quantal release of these molecules, the fundamental processes of effective signaling would be disrupted. Synaptic vesicles are small spherical lipid bilayers formed in the Golgi apparatus and transported to the synaptic terminal (Südhof, 2004; Zimmermann et al., 1993). In the terminal, vesicles are responsible for packaging and concentrating neurotransmitters for signaling events. The majority of the neurotransmitters are sequestered into synaptic vesicles at any given moment, making vesicular storage the primary determinant of transmitter capacity of the brain. Through active transport mechanisms, synaptic vesicles package high concentrations of neurotransmitters within their lumen. A vacuolar proton pump creates the electrochemical gradient in the vesicle (Maycox et al., 1988) It is estimated that the proton pump takes up about 10% of the vesicle’s volume (Stadler and Tsukita, 1984). Each type of neuron is defined by the vesicular transmitter transporter expressed on vesicles within the cell. For example, GABA, glutamate, and monoaminergic cells can be defined by their expression of VGAT, VGLUT1 and 2, or VMAT1 and 2, respectively.
The vesicular monoamine transporter 2 (VMAT2, SLC18A2) is responsible for the packaging of cytosolic monoamines (dopamine, serotonin, norepinephrine, histamine) into small synaptic and dense core vesicles in monoaminergic neurons throughout the brain (Eiden and Weihe, 2011; Erickson and Eiden, 1993; Erickson et al., 1992; Liu et al., 1992a). VMAT2 is a member of the SLC18 family of transporter proteins, where both VMAT1 and the vesicular acetylcholine transporter (VAChT) are members (Lawal and Krantz, 2013). VMAT1 is preferentially expressed in large dense core vesicles of neuroendocrine cells, including chromaffin and enterochromaffin cells; whereas VMAT2 is primarily expressed in monoaminergic neurons in the CNS, sympathetic nervous system, mast cells, and histamine containing cells in the gut (Peter et al., 1995; Schuldiner et al., 1995). Though its crystal structure is unknown, VMAT2 is characterized by 12 transmembrane helices (Peter et al., 1994). Both the N- and C-termini of VMAT2 are localized in the cytosol, and there is a large glycosylated intralumenal loop between the first and second transmembrane domains of the transporter (Wimalasena, 2011). VMAT2 is an H+-ATPase antiporter, which uses the vesicular electrochemical gradient to drive transport (Parsons et al., 1993; Reimer et al., 1998). The pH gradient across the vesicular membrane is established by the vacuolar H+-ATPase, which uses ATP hydrolysis to generate the energy required to move H+ ions into the vesicle (Kanner and Schuldiner, 1987; Moriyama and Futai, 1990). It is this movement of H+ ions that creates the vesicular proton gradient and establishes an acidic environment inside the vesicle (pH of ~5.5). This electrochemical gradient drives VMAT2-mediated transport (Rudnick et al., 1990). For every single monoamine molecule transported into the vesicular lumen via VMAT2, two protons are extruded into the cytosol. By driving the sequestration of neurotransmitter into the vesicle, VMAT2 is a critical mediator of dopamine dynamics in the neuronal terminal.
It had previously been postulated that a variety of factors may limit the packaging of neurotransmitters into monoaminergic vesicles (Edwards, 2007). These models include the set point, equilibrium, leak, and swelling models of vesicular storage. The set point model suggests that vesicles have a set maximum independent of the number of vesicular transporters on the membrane. This could be dictated by equilibrium across the membrane or even physical limitations of the biomechanics of the vesicular membrane. In the equilibrium model, the vesicular ATPase creates the electrochemical H+ gradient that is the driving force for filling via VMAT2. Once the H+ gradient is eliminated and equilibrium is reached between the cytosol and the vesicular lumen, monoamine packaging would stop. Finally, it is possible that biomechanics of the vesicular membrane may act as the limiting factors for filling. In vitro evidence has shown that L-DOPA application and VMAT2 inhibition has the potential to alter vesicle size (Bruns et al., 2000; Colliver et al., 2000; Gong et al., 2003; Pothos, 2002). This suggests that the concentration of transmitter inside of a vesicle does not increase; rather it is merely the volume of the vesicle increasing. These previous results also showed that limitations to the elasticity of membranes may be less stringent than previously thought. Finally, it was assumed that an increase in vesicular filling would result in compensatory changes to the rest of the dopamine system (e.g. increased extracellular dopamine clearance, decreased dopamine synthesis, altered receptor sensitivity) that would eventually downregulate dopamine output.
Vesicular dopamine is protected in the synaptic vesicle. However, dopamine can be left in the cytosol following its synthesis, uptake via the DAT, or after leak out of the synaptic vesicle. The correlation between vesicular function and neuronal vulnerability is well established. VMAT2 is phylogenetically related to a family of toxin extruding antiporters (TEXANs) in prokaryotes and eukaryotes and is also a member of the major facilitator superfamily (MFS) which remove toxins from the cytosol and sequester them into other intracellular compartments (Guillot and Miller, 2009; Schuldiner et al., 1995; Vardy et al., 2004). In fact, the cloning of VMAT2 and its initial characterization was based on the protein’s ability to confer protection from MPP+ toxicity in Chinese hamster ovary fibroblasts (CHO cells) (Liu et al., 1992a). In addition to its ability to protect from exogenous toxic insult, VMAT2 has also been shown to protect against native threats to neuronal health, including unpackaged cytosolic dopamine that can be neurotoxic (Guillot and Miller, 2009).
VMAT2 has two well-characterized inhibitors, reserpine and tetrabenazine. Reserpine, a compound isolated from Indian snakeroot Rauwolfia serpentine, was originally used in the treatment of hypertension (Freis, 1954). However, in addition to reserpine’s ability to inhibit dopamine uptake (Kirshner, 1962; Kirshner et al., 1963), it was also noted that reserpine treatment caused a variety of side effects including gastrointestinal dysfunction, movement deficits, and negative mood (Carlsson, 1976, 1972). It was these mood-altering effects of reserpine that supported the idea that monoamine neurotransmitters may be key in affective behaviors. While the pharmacological target of reserpine was not identified for many years (Erickson et al., 1992; Liu et al., 1992b), reserpine treatment in rats recapitulated many of the behavioral effects seen in parkinsonian humans, including profound akinesia (Colpaert, 1987). In this way, reserpine provided a Parkinson’s disease-like set of behavioral characteristics. VMAT2 inhibition via reserpine is also reversible by the application of a dopamine replacement therapy like L-DOPA (Carlsson, 1972). However, reserpine is not selective between VMAT1 and VMAT2 (Peter et al., 1994). Due to the differing distributions between VMAT1 and VMAT2 in the body, reserpine-mediated VMAT2 inhibition makes CNS-specific effects more difficult to assess. Tetrabenazine is a VMAT2-selective inhibitor (IC50 = 0.3 μM), since it has less of an inhibitory effect on VMAT1 (IC50 = 3 μM) (Chaudhry et al., 2008; Lawal and Krantz, 2013; Pothos, 2002; Wimalasena, 2011). Unlike reserpine, tetrabenazine acts at a binding site on VMAT2 distinct from the substrate binding site and allows for reversible VMAT2-specific inhibition of vesicular filling (Peter et al., 1994). Tetrabenazine and other derivate compounds are currently used clinically in the treatment of hyperkinetic disorders such as chorea associated with Huntington’s disease in Canada, Australia, the United Kingdom, and, more recently, in the United States (Wimalasena, 2011; Yero and Rey, 2008). Recent work has suggested that VMAT2 inhibition by lobeline may be therapeutic in the treatment of psychostimulant abuse (Dwoskin and Crooks, 2002; Nickell et al., 2010). Lobeline acts at the tetrabenazine binding site to decrease amphetamine-evoked dopamine release, suggesting that lobeline selectively inhibits amphetamine effects. It has been suggested that this reduction in dopamine signal occurs via increased metabolism and less dopamine available for reverse transport through the DAT (Miller et al., 2001). Thus, by taking a VMAT2 inhibitor, dopamine-releasing drugs such as amphetamine have less of a rewarding or euphoric effect.
Based on the importance of vesicular storage in neurotransmission and neuroprotection, an increase in VMAT2 function could be a novel clinical target for diseases characterized by loss of dopamine neurons and, subsequently, dopamine release. Despite an extensive history of studying the effects of pharmacological VMAT2 inhibition, there are no known positive modulators of VMAT2 function (Osherovich, 2014).
Previous data clearly show that disruption of VMAT2 function produces adverse effects (Caudle et al., 2008; Taylor et al., 2011). Pharmacological VMAT2 inhibition by reserpine or tetrabenazine results in monoamine depletion and negative behavioral consequences, including akinesia and depressive behaviors (Kirshner, 1962; Kirshner et al., 1963; Pettibone et al., 1984). Genetic knockout of VMAT2 is lethal, with animals dying a few days after birth (Mooslehner et al., 2001; Wang et al., 1997). Through serendipitous recombination events, mice with only 5% of VMAT2 levels were also created (Caudle et al., 2007; Mooslehner et al., 2001). These mice survive into adulthood, allowing for assessments of long-term vesicular deficits. The large reduction in VMAT2 levels in the VMAT2-deficient mice dramatically reduces vesicular filling (approximately 80% reduction) and also causes depletion of dopamine, norepinephrine, and serotonin levels throughout the brain (Caudle et al., 2007; Taylor et al., 2014). VMAT2-deficient mice also show progressive neurodegeneration in multiple monoaminergic regions, including the substantia nigra pars compacta and the locus coeruleus (Caudle et al., 2007; Fon et al., 1997; Taylor et al., 2014). Additionally, these animals show accumulation of α-synuclein, the signature protein component in Lewy body pathology (Caudle et al., 2007).
VMAT2-deficient mice and VMAT2 heterozygote mice also have exacerbated dopaminergic responses to toxicant exposure, including treatment with both MPTP and methamphetamine (Gainetdinov et al., 1998; Guillot et al., 2008; Lohr et al., 2016; Takahashi et al., 1997). Reduced VMAT2 modifies a variety of monoamine-mediated behaviors, including reduced locomotor activity (Caudle et al., 2007) as well as non-motor behaviors reminiscent of pre-clinical Parkinson’s disease, including depressive-like and anxiety-like behaviors, olfactory deficits, altered sleep latency, and gastrointestinal disturbances (Fukui et al., 2007; Taylor et al., 2009). Taken together, evidence from VMAT2-deficient mice has pointed out the numerous deficits arising from reduced vesicular function, while also highlighting the potential benefits of upregulating vesicular storage. Characteristics of these mice are summarized in Table 1.
Although the detrimental effects of reduced VMAT2 function are recognized, our understanding of the potential benefits of increased VMAT2 function in vivo had been limited to a Drosophila model (Chang et al., 2006; Lawal et al., 2010; Sang et al., 2007). Recently, BAC-mediated VMAT2 overexpression in mice (VMAT2-HI) was shown to enhance the capacity of the vesicle to store dopamine (Lohr et al., 2014). Due to this enhanced storage, VMAT2-overexpressing mice have increased total dopamine content and synaptic vesicle size. The elevated vesicular filling translates to elevated dopamine release and changes in monoamine-related behaviors, including locomotor activity, forced swim test, and the marble burying assay. As shown in Table 1, VMAT2-overexpressing mice are also resistant to neurotoxic insults, including MPTP and methamphetamine intoxication, based on VMAT2’s neuroprotective abilities (Lohr et al., 2016, 2015, 2014). Together, these findings suggest that elevated VMAT2 would be beneficial to enhance neuronal monoamine output, protect vulnerable dopamine neuronal populations, and improve monoamine handling at the synapse.
Genetic mutations linked to Parkinson’s disease often affect synaptic vesicle machinery, leading to deficits in trafficking, transmitter storage, and release. α-Synuclein, the first PARK locus identified, has long been known to bind to phospholipids on the vesicle membrane (Lotharius et al., 2002; Nemani et al., 2010; Volles and Lansbury, 2002). Protofibrillar α-synuclein can permeabilize the vesicular membrane, spilling dopamine into the cytosol (Volles and Lansbury, 2002). Oxidized dopamine stabilizes protofibrillar α-synuclein, creating a potentially vicious cycle of pathology (Conway, 2001; Mosharov et al., 2009; Ulusoy et al., 2012). While the biological function of α-synuclein is unknown, genetic ablation of the synuclein genes (α, β, γ) increases dopamine release, suggesting that the protein may act as a brake on transmitter release (Anwar et al., 2011).
However, α-synuclein, perhaps the best characterized Parkinson’s disease-related protein, is not the only link between Parkinson’s disease and vesicular function. Leucine-rich repeat kinase 2 (LRRK2) mutations, responsible for the majority of familial Parkinson’s disease, modify vesicle trafficking, localization of vesicular proteins, and release of dopamine (Belluzzi et al., 2012; Cirnaru et al., 2014; Piccoli et al., 2011). Reduced PTEN-induced putative kinase 1 (PINK1) function also decreases synaptic efficiency by immobilizing synaptic vesicles of the reserve pool (Morais et al., 2009). Protein deglycase DJ-1 interacts with synaptic vesicle proteins such as synaptophysin and Rab3A and modifies the expression of VMAT2 (Osherovich, 2014; Usami et al., 2011). DJ-1, PINK1 and parkin knock-out mouse models all show substantial presynaptic deficits in dopamine release (Cookson, 2012; Goldberg, 2005; Kitada et al., 2007). Most recently, synaptojanin, a vesicular protein integral for endocytotic traffic at synapses, was identified as a disease-causing gene of parkinsonism (Krebs et al., 2013). These results lend further support to the convergence of disrupted synaptic vesicle function and the pathogenesis of Parkinson’s disease.
There is also a growing body of evidence that genetic control of vesicular function, perhaps in the form of VMAT2 function or expression, is a modifier of Parkinson’s disease risk in human populations. Post-mortem Parkinson’s disease brains show dramatically reduced VMAT2-mediated vesicular filling, and this loss is greater than what could be explained by terminal loss alone (Pifl et al., 2014). These results suggest that impaired dopamine storage in vesicles influences the disease process. The Parkinson’s disease brain has also been shown to have higher dopamine turnover, suggesting a higher level of cytosolic dopamine (Goldstein et al., 2011, 2013). These findings could be due to an impaired storage of dopamine or non-specific dopamine leak. Alternatively, it is possible that the increased turnover is due to degradation of dopamine from other monoaminergic neurons or even glial cells. Single nucleotide polymorphisms (SNPs) or mutations within the coding regions of the VMAT2 gene (SLC18A2) are rare, likely due to the fundamental requirement of the protein for neurotransmission. A Saudi Arabian family with multiple members suffering from an infantile-onset parkinsonism with profound motor and cognitive impairments was the first known report of a mutation within the VMAT2 gene (Rilstone et al., 2013). When the mutation was introduced in vitro, it was shown to cause drastic reductions in vesicular filling. Although this familial mutation represents a pediatric neurotransmitter disease, the VMAT2-deficient mice recapitulated many of the neurochemical and behavioral deficits of the affected individuals (Caudle et al., 2007; Taylor et al., 2009).
Genetic evidence from human populations also suggests that increased VMAT2 level or function protects against the development of Parkinson’s disease. Gain-of-function VMAT2 promoter haplotypes protect against the development of Parkinson’s disease in women (Glatt et al., 2006). Two other SNPs in the promoter region of the VMAT2 gene have also been linked with reduced Parkinson’s disease risk, suggesting that increases in VMAT2 level protect against the disease (Brighina et al., 2013). Based on this growing evidence in humans, a better understanding of the neurochemical and toxicological effects of elevated vesicular function in a mammalian system may be of interest for future therapeutic pursuits.
Despite the familial and spontaneous mutations linked to Parkinson’s disease, it is estimated that nearly 90–95% of disease cases are of unknown causes (Dauer and Przedborski, 2003; Hatcher et al., 2008). Parkinson’s disease incidence rates between monozygotic and dizygotic twins are similar and the concordance rate between twins sets is extremely low (Tanner et al., 1999; Wirdefeldt et al., 2011a, 2011b). These findings in human populations suggest that additional factors, beyond genetics, contribute largely to the development of Parkinson’s disease. There has been growing evidence that environmental toxicants, including organochlorine pesticides and polychlorinated biphenyl compounds, are capable of inhibiting VMAT2 activity and altering aspects of the nigrostriatal pathway (Bemis and Seegal, 2004; Guillot and Miller, 2009; Richardson and Miller, 2004). Additionally, it has been shown that chronic exposure of these toxicants and structurally-related compounds (polybrominated diphenyl ethers) modify transporter levels, such as DAT and VMAT2, and other dopamine-related proteins (Bradner et al., 2013; Caudle et al., 2012; Wilson et al., 2014). This may be particularly relevant in the pathophysiology of Parkinson’s disease since these compounds have been identified in postmortem Parkinson’s disease brains (Hatcher-Martin et al., 2012) and correlated with disease incidence (Cannon and Greenamyre, 2011; Hatcher et al., 2008).
Beyond Parkinson’s disease, alterations to VMAT2 function or level has been attributed to a variety of other disorders. For decades, monoamine neurotransmission has been strongly associated with affective disorders. The monoamine hypothesis of depression was largely based on the dramatic mood changes observed in patients taking the VMAT2 inhibitor, reserpine (Krishnan and Nestler, 2008). In humans, changes in VMAT2 level have been correlated with depression, bipolar disorder, and schizophrenia (Laufer et al., 2005; Simons and van Winkel, 2013; Zubieta et al., 2001, 2000; Zucker, 2002; Zucker et al., 2002). Single nucleotide polymorphisms upstream of VMAT2 have also been correlated to the severity of post-traumatic stress disorder (Solovieff et al., 2014). Additional studies have linked changes in VMAT2 level to the risk of Tourette’s syndrome (Ben-Dor et al., 2007), alcohol dependence (Fehr et al., 2013), ADHD symptoms in children (Toren et al., 2005), and even cognitive outcome following traumatic brain injury in adults (Markos et al., 2016). While other monoamines besides dopamine undoubtedly play an important role in these VMAT2-related disorders, all of these disorders have been connected to dopamine neurotransmission. In summary, while VMAT2 does not appear to be causative in disease, its interaction with the dopamine system is likely a contributing factor to complex disease pathogenesis.
DAT and VMAT2 act in concert to control the compartmentalization and movement of dopamine at the synapse. The interplay between DAT and VMAT2 activity may be best described by their importance in vulnerability to the neurotoxicant, MPTP. As mentioned, MPTP is metabolized into MPP+ by MAO outside of neurons (Ransom et al., 1987). After MPP+ enters dopaminergic neurons via the DAT, it then can be sequestered into the vesicle via VMAT2, which is neuroprotective. Based on the critical actions of both of these transporter mechanisms, the ratio of DAT to VMAT2 function has previously been suspected as a mediator of MPTP vulnerability (Masoud et al., 2015; Miller et al., 1999b; Uhl, 1998). Further, it has been speculated that differences in the ratio between DAT and VMAT2 in midbrain dopamine pathways may explain the preferential loss of nigrostriatal projections in Parkinson’s disease models. As seen in the immunohistochemical staining of DAT and VMAT2, dorsal striatum expresses greater levels of the DAT relative to VMAT2 compared to the ventral striatum (nucleus accumbens), which sees considerable sparing following MPTP administration and in Parkinson’s disease (Miller et al., 1999a, 1999b, 1997; Uhl, 1998). It has also been speculated that vesicular uptake of MPP+ in rats is approximately twice that of a mouse, presumably due to VMAT2 expression differences, suggesting a mechanism for the neuroprotection in these animals (Staal et al., 2000). Indeed, inhibition of VMAT2 in these rats increased the toxicity of MPTP.
In the absence of MPTP, DAT and VMAT2 continue to work to maintain the proper compartmentalization of dopamine within the synapse. The various mouse models of transporter function described by our laboratories and others (DAT-KO, DAT-tg, VMAT2-KO, VMAT2-deficient, VMAT2-HI; Table 1) allow for a neurochemical analysis of new combinations of transporter levels. We hope that genetic crosses of these animals will result in studies examining the contributions of these manipulations to dopamine release, presynaptic dopamine homeostasis, pharmacological responses, and behavioral paradigms ranging from locomotor activity to psychostimulant response.
We have summarized ways in which changes to transport at the vesicular and plasma membranes influence the synaptic compartmentalization of neuronal dopamine (Figure 1). Disruptions of dopamine storage that have the overall effect of increasing cytosolic dopamine, such as elevated DAT or decreased VMAT2 function, are of particular interest in hypodopaminergic disease states such as Parkinson’s disease. Conversely, manipulations meant to dampen excessive dopaminergic tone may be of use in disorders such as dyskinesias and ADHD. We have described the genetic, epigenetic, and environmental factors that may alter functional levels of DAT and VMAT2 and how these processes influence neuronal viability and susceptibility to neurodegeneration. The recent development of transgenic mouse models of varying levels of DAT and VMAT2 (Table 1) now allows the field to investigate dopamine neurochemistry on a continuum of transporter function and to carefully titrate the handling of dopamine content within the rodent brain. Finally, these mouse models provide insight into both the promise of therapies that modify the function of these transporters, and the pursuit of other potential players to manipulate the delicate balance of dopamine at the synapse.
This work was supported by The National Institutes of Health Grants R01ES023839 (G.W.M.), P30ES019776 (G.W.M.), F31NS084739 (K.M.L.), The Lewis Dickey Memorial Fund (G.W.M.) and by the Canadian Institutes of Health Research Grant 210296 (A.S).