Discovery at the VICB
Exploring the Dynamics of Active Transport
By: Carol A. Rouzer, VICB Communications
Published: October 10, 2014
Double electron-electron resonance studies of two LeuT class transporters reveal distinct mechanisms despite highly similar structures.
The LeuT class of Na+-coupled transporters harnesses the electrochemical energy stored in Na+ gradients for the uphill transport of molecules across the plasma membrane. In humans, members of this class are responsible for neurotransmitter reuptake in the nervous system, thereby making them targets of antidepressants and some drugs of abuse. The importance of these proteins to neurophysiology and pharmacology led Vanderbilt Institute of Chemical Biology member Hassane Mchaourab and his laboratory to explore the dynamics of two members of the LeuT transporter class. His results both confirm and challenge prior views regarding the mechanisms by which these proteins execute their function (K. Kazmier et al., (2014) Nat. Struct. Mol. Biol., 21, 472, and K. Kazmier et al. (2014) Proc. Natl. Acad. Sci. U.S.A., published online September 29, DOI:10.1073/pnas.141043111.
Much of what is known about the structure and function of these proteins comes from studies of the prototypical leucine transporter (LeuT) from the bacterium Aquifex aeolicus. Crystal structures of LeuT reveal twelve transmembrane helices (TMs) linked by a series of intracellular and extracellular helices and loops (Figure 1). The core of the protein is notable for a characteristic pseudo two-fold symmetry along an axis within the plane of the membrane, which relates TM1 through TM5 with TM6 through TM10 (Figure 1D). The binding sites for two Na+ ions and one molecule of leucine are located between TM1 and TM6 near the center of the lipid bilayer. The two helices unwind partially to provide interaction sites for the substrate and its accompanying ions.
Figure 1. (A) Schematic of the LeuT structure, showing the twelve transmembrane (TM) helices, the sodium (blue spheres) and leucine (yellow triangle) binding sites, and the extracellular (EL) and intracellular (IL) loops. (B) Three-dimensional structure of LeuT as seen from within the plane of the membrane. (C) Three-dimensional structure of LeuT as seen from the cytoplasm. (D) Structure of TM1 through TM10, showing the pseudo two-fold axis (black oval). (E) Ribbon diagram of the LeuT dimer. Image reproduced by permission from Macmillan Publishers Ltd, from A. Yamashita, et al., (2005) Nature, 437, 215. Copyright 2005.
Crystal structures of LeuT obtained in the unliganded (apo) state, in the presence of Na+ alone, or in the presence of both Na+ and leucine suggest three major conformational states for the protein, designated inward-facing, outward-facing, and occluded, based on how the substrate can access the binding site. Homology modeling suggested a “rocking bundle” model to explain how LeuT converts between these conformations. This model envisions the LeuT structure in terms of a rigid four-helix scaffold (containing TMs 3, 4, 8, and 9) and an equally rigid four-helix bundle (containing TMs 1, 2, 6, and 7). Movement of the bundle relative to the scaffold is proposed to account for changes between conformations during binding and transport of Na+ and leucine.
Although the structural data provide considerable insight into the mechanism of action of LeuT, the Mchaourab group realized that dynamic studies were sorely needed. To accomplish this goal, they used double electron-electron resonance (DEER) studies of site-specifically spin-labeled LeuT. Normally, this approach requires that the researchers first construct a mutant protein in which all cysteine residues have been converted to alanine. This was not necessary for LeuT, which naturally contains no cysteines. So, the investigators could go immediately to the next step, which is to construct a series of mutant proteins, each containing two cysteine residues located at carefully selected sites that are predicted to move with respect to each other during a conformational transition. Once this was accomplished, they then labeled each of the mutant proteins with the sulfhydryl-targeted spin label MTSSL (1-oxyl-2,2,5,5-tetramethylpyrroline-3-methyl) methanethiosulfonate) and confirmed that each of the mutant proteins retained the ability to bind Na+ and leucine. This allowed them to use DEER measurements to monitor changes in the distance between the two labels as a result of Na+ and leucine binding across the molecular ensemble.
Because proteins are not rigid static structures, DEER data are actually distributions that provide the probability that the two labels will be at a particular distance from each other in a given state. A broad distribution suggests conformational flexibility between the two labeled sites, while a narrow distribution suggests rigidity. Data for LeuT suggested that, in the absence of Na+ and leucine, the protein samples all three major conformations (inward-facing, outward-facing, and occluded), with the equilibrium favoring inward-facing. Addition of Na+ alone shifts the equilibrium toward the outward-facing conformation, while in the presence of both Na+ and leucine, the occluded conformation predominates (Figure 2). At the extracellular surface, the investigators found that Na+- and leucine-dependent conformational changes most strongly affected the positions of TM1b, TM7b, TM6a, and extracellular loop (EL)4. At the intracellular surface, the most strongly affected structures were TM6b, TM7a, and the N-terminal segment. Although the TMs involved in these structural changes were those identified as the moving bundle in the rocking bundle model, the researchers found that these helices move independently in response to ligand binding and not as a single rigid structure as predicted by that model. Furthermore, the rocking bundle model predicted motions of TMs1 and 2 on the intracellular side and TM 2 on the extracellular side, which were not confirmed experimentally.
Figure 2. Example of DEER data measuring the probability distribution (P(r)) for the distance (r) between a label on TM6 and a second label on TM3 (left) or TM5 (right) at the external surface of the protein. Note that in the apo state (black), the equilibrium favors the outward-closed (C) conformation, but the outward-open (O) is also abundant. Binding of Na+ shifts the equilibrium to the O conformation, while addition of Na+ and leucine (red) produces a strong shift to the (C) conformation. Image reproduced by permission from Macmillan Publishers Ltd, from K. Kazmier et al., (2014) Nat. Struct. Mol. Biol., 21, 472. Copyright 2014.
The DEER data revealed additional discrepancies from predictions based on the crystal structures of LeuT. The magnitude of some of the changes observed by the Mchaourab lab were >5 Å larger than those predicted from the static structural data. It became clear that the crystal structure obtained in the apo form, which required mutation of several residues to lock the protein in an putatively inward-facing mode, represented a conformation that was not sampled by the wild-type enzyme in solution. Finally, some discrepancies between the DEER data and the occluded crystal structure were explained on the basis of the detergent used to solubilize the protein, demonstrating the importance of experimental conditions in the exploration of protein structure and dynamics.
Once they had obtained distance distributions for a large number of sites on the protein, the Mchaourab group used these data to obtain a model for the Na+-dependent transport of leucine by LeuT. Their model (Figure 3) is characterized by conformational flexibility, in contrast to the rigid movements predicted by the rocking bundle model.
Figure 3. Proposed mechanism for Na+-dependent leucine transport by LeuT. (a) The apo form of the protein begins in equilibrium between a number of conformations, but the inward-facing one predominates. (b) Binding of Na+ shifts the equilibrium to the outward-facing conformation. (c) Leucine binds to the outward-facing conformer, shifting the equilibrium to the occluded conformation. (d) Although the occluded conformation is now favored, an equilibrium continues to exist that includes the inward-facing conformation. (e) Na+ is released from the inward-facing conformation. (f) Release of Na+ reduces the affinity for leucine, which also exits the binding site. This returns the protein to the apo form. Image reproduced by permission from Macmillan Publishers Ltd, from K. Kazmier et al., (2014) Nat. Struct. Mol. Biol., 21, 472. Copyright 2014.
Having raised serious questions regarding the rocking bundle model, the Mchaourab lab turned their attention to another LeuT family transporter, the Na+-dependent Mhp1 symporter from Microbacterium liquefaciens. This protein, which transports benzyl-hydantoin (BH) against a concentration gradient, was the first LeuT family transporter for which crystal structures captured inward-facing, outward-facing, and occluded conformations. In fact, it was structural data from Mhp1, and not LeuT, that supported the rocking bundle model. Although Mhp1 shares overall structural similarity with LeuT, it differs from LeuT in that there is only one binding site for Na+, designated Na2, whereas LeuT has two binding sites, designated Na1 and Na2. This led the Mchaourab lab to wonder if the two binding sites in LeuT were functionally distinct, and if the absence of Na1 in Mhp1 would result in an altered role for Na+ in that protein as compared to its role in LeuT. Specifically, the DEER data obtained with LeuT demonstrated that Na+ binding results in a shift from the inward-facing to the outward-facing conformation. Prior data had revealed that Na+ increases Mhp1’s affinity for BH binding, but no data addressed whether or not Na+ directly affects Mhp1’s conformational equilibria.
To answer these questions and to explore the dynamic mechanism of Mhp1, the Mchaourab lab conducted DEER studies very similar to the ones they had successfully executed using LeuT. The results showed that, in contrast to LeuT, addition of Na+ to Mhp1 has no effect on protein conformation. Rather, both Na+ and BH were required to see significant movements of the protein, which were characterized by large changes (7 to 10 Å) in the position of bundle helices relative to scaffold helices, consistent with the rocking bundle model. Also consistent with the model was the finding that helices within the scaffold moved very little relative to each other, and the same was true for helices within the scaffold. In general, the distance changes observed by DEER agreed well with those predicted from the available crystal structures.
A finding not predicted by the crystal structure was the high degree of flexibility of TM5, particularly in the apo state. Upon Na+ and BH binding, this flexibility decreases markedly, placing TM5 in the outward-facing conformation. This finding suggests that TM5 serves as a gate on the intracellular side of the protein. The opposite pattern was observed for TM10, consistent with its role as a gate on the extracellular side. Interestingly, the extracellular portion of TM9 exhibited similar dynamics to those of TM10, to which it is attached by a short helix.
The investigators noted that, despite the relatively concerted movements of the bundle helices relative to the scaffold helices, Mph1 exists in an equilibrium that samples multiple conformations in both the apo and bound state. This is important, because it allows a transition of the bound transporter to the inward-facing conformation, a necessary step for release of Na+ and BH into the cell. The researchers used the DEER data and molecular dynamics to develop a model of Na+-dependent BH transport by Mph1 (Figure 4). Notably for this transporter, Na+ acts to stabilize BH binding, but has no direct effect on protein conformation. These findings suggest the possibility that the Na1 site, which is present in LeuT but not Mph1, is required for Na+-dependent modulation of of ligand-binding but not conformational stabilization, a novel insight that will be explored through future work.
Figure 4. Proposed mechanism of Na+-dependent BH transport by Mph1. In the absence of BH and Na+, Mph1 exists in a equilibrium between inward facing (A) and outward facing (B) conformations, although the inward-facing conformation is favored. Binding of Na+ and BH stabilizes the outward-facing conformation (C); however, this conformation remains in equilibrium with the inward facing conformation (D). A switch to the inward facing conformation provides an opportunity for release of Na+ and BH, returning the transporter to the apo state (E). Image reproduced by permission from K. Kazmier et al. (2014) Proc. Natl. Acad. Sci. U.S.A., published online September 29, DOI:10.1073/pnas.141043111. Copyright 2014, K. Kazmier et al.