Unveiling the Secret Lives of Glutamate Transporters: VGLUTs Engage in Multiple Transport Modes

Abstract

Accumulation of glutamate in synaptic vesicles is mediated by vesicular glutamate transporters called VGLUTs. In the current issue of Neuron, Preobraschenski et al. (2014) show that the VGLUTs, in addition to transporting glutamate, also provide the conductances necessary to maintain the appropriate voltage and pH inside these vesicles.

During chemical synaptic transmission, neurotransmitters accumulate in and are subsequently released from intracellular compartments called synaptic vesicles (SVs). This process is a fundamental step in neuronal function as well as in nonneuronal systems. Glutamate is the major excitatory neurotransmitter, and therefore, the mechanism of its accumulation into SV is of major interest to neurobiochemists, neurophysiologists, and cell biologists. Dysregulation of glutamate accumulation in SV has been associated with schizophrenia and epilepsy (Omote et al., 2011). Our genome encodes for three vesicular glutamate transporters (VGLUT1–VGLUT3), which belong to the SLC17 family (Omote et al., 2011). These transporters utilize the electrochemical proton gradient (ΔμH) created by V-ATPases as the driving force for glutamate uptake in the SVs. The H⁺-pumping activity of the V-ATPases establishes a pH gradient (ΔpH), and in so doing, it also gives rise to a membrane potential (Δψ) (Figure 1). Interestingly, in addition to its function as a H⁺-driven glutamate transporter, VGLUT2 can also utilize a Na⁺ gradient to drive phosphate transport in a gluta-mate-independent manner (Juge et al., 2006). In the current issue of Neuron, a new paper from Preobraschenski et al. (2014) shows that VGLUTs have yet other transport function: they mediate uncoupled Cl⁻ movement as well as K⁺/H⁺ exchange.

Figure 1. The Many Functions of VGLUT.

Schematic representation of the energetics of glutamate accumulation in SVs mediated by VGLUT. The different operational modes of VGLUT are shown on the right.

Since the internal volume of SVs is small, the presence of uncompensated charges and unbuffered protons would generate Δψ unfavorable to transport and drop the vesicular pH below physiological levels, both factors which limit neurotransmitter accumulation. Therefore, pathways for the efflux of H⁺ and other counterions, mainly Cl⁻ and K⁺, must exist in SVs to maintain physiological levels of pH and charge balance. While several reports indicated that these conductances do exist (Goh et al., 2011; Stobrawa et al., 2001), their molecular identity remains debated (Schenck et al., 2009). Among these ions, Cl⁻ plays a key role in regulating neurotransmitter uptake, acting both as a counterion and as a regulator of VGLUT activity (Martineau et al., 2013; Schenck et al., 2009; Stobrawa et al., 2001). However, the molecular identity of Cl⁻ transport pathway of SVs remains unknown and controversial. An initial report suggested that the CLC-3 H⁺/ Cl⁻ exchanger might serve as the main pathway for Cl⁻ efflux from SVs (Stobrawa et al., 2001), but this idea was later refuted since SVs purified from CLC-3 knockout mice still retain WT-like Cl⁻-dependent activity (Martineau et al., 2013; Omote et al., 2011; Schenck et al., 2009).

Chloride regulates VGLUT activity in a complex manner: it is both a stimulator and an inhibitor of VGLUT-mediated glutamate uptake into SVs. In the absence of Cl⁻, transport is completely eliminated, indicating a strict requirement for this substrate. As the Cl⁻ concentration rises, transport activity increases sharply, peaking at ~4 mM, and subsequently decreases again in the presence of high Cl⁻ (Juge et al., 2010; Moriyama and Yamamoto, 1995; Omote et al., 2011; Schenck et al., 2009). Although this phenomenon has been well characterized functionally, its underlying mechanistic basis remains poorly understood and controversial. An initial report indicated that purified and reconstituted VGLUT1 directly transports Cl⁻, consistent with a model in which VGLUT plays a dual role. It provides an efflux pathway for the high Cl⁻ content of freshly endocytosed vesicles, therefore creating the initial driving force for glutamate uptake before the V-ATPase has been able to establish the ΔpH. Subsequently, VGLUT shunts the charge accumulation of glutamate by allowing Cl⁻ efflux and maintaining charge neutrality.

This model was later challenged by others who showed that purified and reconstituted VGLUT2 does not directly mediate Cl⁻ transport (Juge et al., 2010). Since purified VGLUT2 is not inhibited by high Cl⁻, these authors also suggested that the decreased glutamate accumulation seen in native vesicles arises from an effect of Cl⁻ on the primary V-ATPase rather than on VGLUT. The steep dependence of VGLUT activation on Cl⁻ suggests that multiple and highly cooperative binding sites are present on this transporter. Taken together, these observations led to the proposal of a competing model in which Cl⁻ is an allosteric regulator of VGLUT activity rather than a substrate.

The stark contrast between these results belies the biochemical simplicity of the reduced systems in which these experiments were carried out: purified transporters reconstituted in proteoliposomes. One obvious possibility accounting for these discrepancies is that different VGLUT isoforms could behave differently with regards to their ability to transport Cl⁻. Alternatively, the interpretation of the data might be confounded by the technical difficulties intrinsic to separating the indirect effects that H⁺, K⁺, and Cl⁻ transport have on VGLUT through changes in voltage and pH from their direct effects on VGLUT itself. In the current issue of Neuron, Preobraschenski et al. (2014) take on this controversy in a technical tour de force and investigate VGLUT-mediated transport using two systems: in the first, they use proteoliposomes reconstituted with purified VGLUT and the F₀/F₁ ATPase; in the second, they use the SNARE machinery to fuse purified SVs and proteoliposomes reconstituted with the F₀/F₁ ATPase. These systems allow Preobraschenski et al. (2014) to investigate VGLUT function in vesicles whose ionic composition is rigorously controlled and which either contain a full complement of correctly oriented native proteins or just VGLUT and the ATPase. Using these complementary approaches, they show that VGLUT alone is necessary and sufficient to dissipate an outwardly directed Cl⁻ gradient, unambiguously demonstrating that this protein directly mediates Cl⁻ transport. They show that VGLUT also has an additional activity: it can function as a K⁺/H⁺ exchanger. In both types of vesicles, the presence of K⁺ gradients leads to the depletion of the ΔpH established by the ATPase and to an increase in Δψ, indicating that VGLUTs mediate transport of both ions. Surprisingly, neither of these noncanonical functions of VGLUTs is coupled to glutamate transport, indicating that these transporters can operate in several distinct transport modalities.

While the coexistence of multiple “transport modes” by the same protein is not unprecedented (Adams and DeFelice, 2003; Artigas and Gadsby, 2003; Fairman et al., 1995), the degree of functional plasticity displayed by the VGLUTs is unique. In addition to their canonical function as glutamate transporters, these protean transporters mediate uncoupled Cl⁻ movement, K⁺/H⁺ exchange, and Na⁺-driven phosphate transport (Figure 1). How a single protein can adapt to and switch among such diverse modes is an open question that will require further work to be elucidated. In the current manuscript, Preobraschenski et al. (2014) propose a model in which the VGLUTs have three binding sites, one cationic and two anionic, that alternate exposure between the two sides of the membrane. The location of these sites is unknown, and their interaction is remains poorly understood. Evidence from mutagenesis studies showed that the glutamate and phosphate transport machineries are independent (Juge et al., 2006), suggesting that structurally separate pathways might exist. It will be interesting to see whether the VGLUTs can simultaneously support two or more of these different transport modes. The intertwined physiological roles of the various functions together with the independence of the glutamate and phosphate transport would be consistent with the idea that multiple activities could be carried out simultaneously. If the functions are mutually exclusive, then how do VGLUTs switch between the different transport modalities? Could it be as simple as being determined by which of the ions or substrate binds first? Alternatively, it is conceivable that a more complex conformational change might be required to place these transporters in the appropriate mode.

The physiological implication of these experiments is important and surprising: the VGLUTs alone can provide all the shunt conductances necessary for glutamate accumulation in SVs. This eliminates the need for any additional components, other than the V-ATPases, whose role is to provide the H⁺ driving force. The unusual multiplicity of activities carried out by the VGLUTs allows them to adapt to the varying ionic conditions encountered as the SVs change their ionic content while becoming loaded with glutamate molecules. The simplicity of this solution to such a complex problem is appealing and economical. While additional experiments will be needed to determine whether additional transporters and/or channels contribute to these shunt conductances, the results obtained by Preobraschenski et al. (2014) show that these additional partners are not necessary and that the VGLUTs are the Swiss army knives of transporters.