While the S site has been implicated in binding

While the S2 site has been implicated in binding, it protease inhibitors hiv is unclear whether the S2 site plays a role in the substrate transport process. Crystal structures have revealed that different drugs and detergents can bind to the S2 site of SERT [8] and LeuT [[13], [14], [15]]. However, there are no crystal structures of LeuT-fold transporters with substrate bound in the S2 site [16]. Recent NMR investigations of LeuT could only identify substrate binding at the S1 site [17]. Shi and co-workers proposed that the S2 site of LeuT plays an important allosteric role in the two-site substrate transport model, wherein substrate binding to S2 triggers the release of Na+ and substrate from the S1 site [9]. This model was computationally applied to DAT by Shan and colleagues [18] and supported by mutagenesis of DAT by Zhen and Reith [19]. Furthermore an allosteric relationship between the two sites has been shown for SERT via atomic force microscopy [20] and MD simulations [21]. Conversely, a recent investigation using a combination of MD simulations and experimental mutagenesis to examine the GlyT2 S2 site found that S2 does not influence substrate selectivity. Instead it is proposed that S2 may acts as a low-affinity, non-selective loading site [22]. Taken together, these studies indicate that the role of the S2 site may differ markedly between structurally related transporters. Structural and computational studies have revealed that SLC6 transporters shuffle between at least four states: outward-open (substrate acceptance/release, open to the extracellular space), outward-occluded (Fig. 2, substrate bound, closed to the extracellular space), inward occluded (substrate bound, closed to the intracellular space) and inward-open (Fig. 2, substrate acceptance/release, open to the intracellular space) [23]. However, the precise movements these transporters use to transition between the four main states are difficult to characterise using experimental techniques. Based on the available crystal structures for SLC6 transporters and computational studies, three alternating access transport models have been proposed: the rocking bundle model, the hinge model and the iris model. These models differ primarily in how they describe the movements of TM1 and TM6. The rocking bundle model, which was first devised by Forrest et al., is based on mutagenesis data and homology models of SERT in the outward-open and inward-open conformations, modelled on the LeuT structures [2]. Forrest et al. noticed that the LeuT crystal structure consists of two pseudo-symmetric halves, TM1–5 and TM6–10, which have inverted orientations relative to each other [2]. Computational analysis suggested that the rigid body rocking motion of TM1, 2, 6 and 7 (referred to as the four-helix bundle) within the scaffold formed by TM3–5 and TM 8–10 could provide a transition mechanism between the outward- and inward-facing conformations [2]. This conserved structural psuedosymmetry has been identified in other transporters from the SLC superfamily, including BetP, GltPh and Mhp1 [24]. Like the rocking bundle model, the hinge model also exploits the pseudosymmetry of the transporter architecture, but instead proposes that the kinked, central portions of TM1 and TM6 add subtle conformational flexibility. This precludes the rigid body motion of the four-helix bundle within the protein scaffold, that was proposed in the rocking bundle model [25]. The iris model, first proposed for the crystal structure of BetP—a bacterial sodium symporter that is structurally similar but sequence-unrelated to SLC6 neuotransmitter transporters—suggests that the pivot points for the movement of the four-helix bundle are located near the S1, Na1, and Na2 binding sites. In the iris model, the helices within the four-helix bundle are proposed to reorientate via a series of stepwise, anticlockwise rotations [26]. The three transport models are based on differences between the outward- and inward-facing crystallographic states. In each case, these crystallographic transitions represent static end points. Computational studies have played an integral role in identifying and characterising the dynamics associated with SLC6 transporters’ outward- to inward-facing transition, and vice versa, as recently reviewed by Bisha and Magistrato [27]. For instance, biased simulation techniques such as steered molecular dynamics simulations have been used to pull the substrate, leucine, from its occluded binding pocket in LeuT towards either the extracellular space or the cytoplasm [9]. While carried out over relatively short timescales (<30 ns), these simulations showed that pulling the substrate towards the cytoplasm induced hinging motions, particularly in TM1 and TM6, that the authors believed to be consistent with the hinge model [9]. A second simulation study used steered MD simulations on the outward-facing crystal structure of LeuT in conjunction with a homology model of the inward-facing LeuT (based on the vSGLT sodium/galactose transporter) to drive the conformation from an outward-facing to an inward-facing state [23]. While relatively high force constants were applied to drive the transitions, and the simulations were again conducted over short timescales, a hinging motion was noted in the intracellular halves of TM1 and TM6 of the four-helical bundle during the outward-facing to inward-facing transition. This motion was coupled to movements in the remaining helices in the four-helical bundle, and despite variations in the extent of the conformational coupling with each simulation, the authors deemed that the motions support the hinge model [23]. In fact, many other nonbiased and steered MD studies of the LeuT [28,29], and DAT crystal structures [30], and homology models of SERT [31], show increased flexibility in TM1 and TM6.