To assess whether this coupling mechanisms might operate als
To assess whether this coupling mechanisms might operate also in a native system of untransfected cells, we select a cell line SH-SY5Y neuroblastoma that is known to express both the dopamine transporter and the M-currents (Jiang et al., 2004; Wickenden et al., 2008). These experiments were performed in the presence of a cocktail of inhibitors of dopamine receptors (1 μM prazosin plus 10 μM haloperidol plus 3 μM eticlopride) since dopamine autoreceptors might control the activation of additional compensatory currents (Sonders et al., 1997). Transport assays revealed a significant inhibitory effect of the potassium channel blocker XE-991 on [3H]-dopamine uptake (the uptake in the presence of 20 μM XE-991 was 76 ± 4 percent of the controls, n = 12, p < 0.005). This suggests that endogenous DAT and M-currents are also coupled because the lack of the clamping effect of Kv7 channels on the cell membrane potential affects negatively dopamine transport. To test this possibility, the dopamine substrate currents (IDA – IControl) were determined at different holding potentials (IV curves) in differentiated SH-SY5Y NMS-E973 (Fig. 8). The plot for SH-SY5Y cells exhibited a similar profile to transfected HT22 cells, with the transport-associated depolarizing conductance increasing at more hyperpolarized potentials (i. e., for holding potentials more negative than −50 mV). Blockade of endogenous M-channels with 20 μM XE-991 increased the dopamine substrate currents, mainly at more negative membrane potentials (Fig. 8), again supporting that endogenous DAT and M-currents are electrically coupled.
Discussion As GLT-1 and DAT are secondary active carriers that share some biochemical and biophysical properties, it was hypothesized that they might share some protein partners. In the current work, by using the BioID assay, we identified two membrane proteins, Tmem263 and Kv7.3, as potential members of the interactome of GLT-1 and DAT. Biochemical assays supported that only one of them, the potassium channel subunit Kv7.3, was tightly association, forming complexes by immunoprecipitation with both transporters. Previous studies have indicated that GLT-1 and DAT also share subcellular localization, since both are found on lipid rafts in different systems, an association that seem to be dynamic and therefore probably subject to regulatory interactions (Cremona et al., 2011). Consistently, Kv7.2/7.3 potassium channels are also known to be located on lipid rafts, unlike Kv7.5 (Roura-Ferrer et al., 2012; Zhang et al., 2013), which indeed did not coimmunoprecipitate with these transporters. Regarding Tmem263, we do not know whether this interaction has a role in the lateral mobility of the transporters within the plane of the membrane. It is a small protein (9 kDa), almost entirely embedded in the membrane, and software-based protein structure prediction suggested it consists of two membrane-spanning domains, an intracellular N and C-terminus and an extracellular loop (Rizzi et al., 2015). Interestingly, Tmem263 has also been reported as part of the interactome of the potassium channels Slick and Slack (Rizzi et al., 2015). Nevertheless, its physiological role in the nervous system remains uncertain, as no neurologic phenotype has been reported in animals or humans carrying mutations in this gene. Loss-of-function mutations in gene Tmem263 had consequences in the musculoskeletal system, influencing the deposition of bone mineral and thereby affecting skeletal development and body growth, as observed in autosomal dwarf chickens (Wu et al., 2018). Our observations supporting a tight association of DAT and GLT-1 with Kv7.2/7.3 potassium channels fit with the well-known dependence of these transporters on the ionic distribution through the membrane. Both transporters are electrogenic in nature, and the entrance of sodium ions into the cell during transport cycles might promote depolarization of the membrane. Therefore, the existence of counterbalancing mechanisms that instantaneously readjust membrane potentials in response to neurotransmitter uptake would not be surprising. M-channels play an important role in the control of neuronal excitability in the brain (Greene and Hoshi, 2017). The high density of these channels at the AIS is in consonance with their function as “brakes” on repetitive firing and membrane depolarization. But M-channels also localize along the entire axon (Chung et al., 2006) and in presynaptic terminals, where they are correctly positioned to control neurotransmitter release (Cooper et al., 2001; Martire et al., 2004; Sun and Kapur, 2012). Regarding the subcellular localization of DAT, although it is expressed on the entire surface of dopaminergic neurons, it is enriched on the surface of axons of the striatal system, especially in varicosities and terminals (with the exception of the active zone) (Block et al., 2015; Nirenberg et al., 1996). Therefore, DAT and Kv7.2/7.3 potassium channels might coincide along the axonal shafts and terminals. Our data supports the idea of DAT and Kv7.2/7.3 co-localization in multiple clusters along axonal tracts. Consistent with this, immunohistochemical studies have proven that the subunit Kv7.2 co-localizes with tyrosine hydroxylase on striatal nerve terminals (Martire et al., 2007). Additionally, our dopamine uptake assays and electrophysiological measurements corroborate the existence of a functional coupling between them, both in a heterologous expression system (transfected HT22 cells) and in a native system like the dopaminergic neuroblastoma (SH-SY5Y cells). This coupling would boost DAT performance, increasing the uptake of dopamine in neurons. Whether this functional coupling exists in native dopaminergic neurons and if the association of Kv7.2/7.3 with DAT is constitutive or under regulatory mechanisms remains to be investigated. In any case it might be relevant for the electrical properties of DAT, probably in collaboration with the uncoupled anion conductance through this transporter (Ingram et al., 2002), as well as for trafficking mechanisms that are dependent on membrane potential (Richardson et al., 2016), and perhaps other ionic currents controlled by dopamine autoreceptors (Sonders et al., 1997). To add a layer of complexity, dopamine could be negatively regulating M-channels, as our results suggest (see Fig. 5B). In those experiments, dopamine likely activates some type of Gq/11 protein-coupled receptor that, after phospholipase C (PLC) activation, would produce a cleavage of phosphatidylinositol 4,5-bisphosphate (PIP2), which is required for M-channel activation (Winks et al., 2005) (Brown et al., 2007). There are various reports of promiscuous interaction between dopamine and adrenergic receptors, particularly α1 (Cornil et al., 2002; Lin et al., 2008) (Zhang et al., 2004). α1-adrenergic receptors are coupled to Gq/11 proteins, making them conceivable candidates to mediate the inhibition that we detected in our preparation. Despite the fact that the intricate interplay between Kv7 channels and dopaminergic neurotransmission has been previously studied (Hansen et al., 2008), future studies should include the notion of an interaction between DAT and Kv7 channels.