Exploring the K⁺ binding site and its coupling to transport in the neurotransmitter:sodium symporter LeuT

Solveig G. Schmidt
Andreas Nygaard
Joseph A. Mindell
Claus J. Loland author has email address

Laboratory for Membrane Protein Dynamics, Department of Neuroscience, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, Denmark
Membrane Transport Biophysics Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland, USA

https://doi.org/10.7554/eLife.87985.2

Open access
Copyright information

Figures and data

K⁺ competitively inhibits Na⁺-mediated [³H]leucine binding in LeuT.
(A) Na⁺-mediated [³H]leucine (10x K_d) binding to purified LeuT in the absence (red triangles) or presence of 200 mM (light blue diamond) or 800 mM (blue squares) K⁺. Data are normalized to B_max and fitted to a Hill equation, yielding EC₅₀ values in 0, 200 and 800 mM K⁺ of 7.7 [7.3; 8.1], 19.4 [19.1; 19.8] and 48.5 [47.5; 49.5] mM, respectively. (B) K⁺-dependent displacement of Na⁺-mediated [³H]leucine binding in the presence of 7.7 mM Na⁺, which corresponds to the EC₅₀ determined in (A). Data are normalized to a control with 0 mM K⁺ and fitted to a Hill equation with an IC₅₀ value of 234.8 [224.8; 243.1] mM. All data points are shown as mean ± standard error of the mean (s.e.m.), n = 3–6 conducted in triplicates. Error bars often smaller than data points. EC₅₀ and IC₅₀ values are reported as mean [s.e.m. interval]. The ionic strength was maintained by substituting Na⁺ and K⁺ with Ch⁺.

K⁺ shifts the conformational equilibrium towards an outward-closed state.
(A) Left; cartoon representation of the superimposed structures of LeuT in the outward-open Na⁺ bound state (orange; PDB-ID 3TT1) and inward-open state (turquoise; PDB-ID 3TT3), viewed parallel to the plane of the membrane. Right; enlarged view of the top of TM10 and EL4a, carrying the F5M (yellow, shown adjacent to K398C) and the Ni²⁺-chelating His-X₃-His site (A313H-A317H). Dashed lines indicate for both structures the distance between the coordinated Ni²⁺ ion (green) and the side chain sulfur of K398C (10.7 Å in 3TT3 and 19.8 Å in 3TT1). (B) tmFRET efficiencies (1-F/F_{no site}) obtained for LeuT in 800 mM of the ions specified (± 50 µM leucine or 200 µM alanine when indicated) upon saturation of the His-X₃-His site with 10 mM Ni²⁺. Right vertical axis shows FRET efficiencies converted to distances by the FRET equation. ****p<0.0001 represents the significance levels from a Tukey multiple comparison-corrected one-way ANOVA. (C) tmFRET (F₀/F) as a function of Na⁺ (red triangles) or K⁺ (blue squares) performed on LeuT^tmFRET in the presence of 750 µM Ni²⁺. The K⁺ response fitted to a Hill equation yields a Hill slope of 1.15 ± 0.03 (mean ± s.e.m.) and an EC₅₀ of 182.6 [176.4; 189.1] mM, mean [s.e.m. interval]. All data points represent mean ± s.e.m. (error bars often smaller than data points), n = 3-5 conducted in triplicates.

K⁺ modulates LeuT mediated [³H]alanine transport.
(A) Schematic of LeuT reconstituted into liposome containing buffer with various cations (X) and Cl^- as corresponding anion. (B) Time-dependent [³H]alanine (2 µM) uptake in the presence of 200 mM Na⁺ into liposomes containing 200 mM of the various cations (colored as in (A)). Data are fitted to a non-linear regression fit. The maximum uptake predicted with NMDG⁺ is defined as 100% (supplementary table 1). (C) Concentration-dependent [³H]alanine uptake for 5 min (colored as in (A)). Lines are fits to Michaelis-Menten kinetics (supplementary table 2). None of the K_m values were significantly different (Tukey multiple comparison-corrected one-way ANOVA, p>0.05). The V_max value in Cs⁺ was not significantly different from NMDG⁺ (Tukey multiple comparison-corrected repeated-measures one-way ANOVA, p=0.5780), but K⁺ and Rb⁺ were significantly different from NMDG⁺ (p=0.0011 and 0.0132, respectively). (D) Time-dependent [³H]alanine uptake in the presence of 50 mM Na⁺ and absence of K⁺ (black squares) was defined as 100%. Addition of 150 mM K⁺ (blue triangles) did not significantly change the maximal uptake (unpaired t-test, p=0.92). With 150 mM intra-vesicular K⁺ (blue circles) the maximum uptake increased to 368 ± 15 % (mean ± s.e.m). This increase was not significantly (unpaired t-test, p=0.43) affected further when 150 mM K⁺ was also added to the extra-vesicular side (purple diamonds). The ionic strength was kept constant by substitution with NMDG⁺. (E) Time-dependent [³H]alanine uptake into liposomes containing 25 mM Na⁺ and 200 mM K⁺ in the presence of either 25 mM Na⁺ and 200 mM K⁺ (black line, no ion gradients) or 25 mM Na⁺ and 200 mM NMDG⁺ (blue line, outward directed K⁺ gradient). The data points from the two conditions are not significantly different (unpaired Mann-Whitney test, p=0.645). All data points are shown as mean ± s.e.m, n=3 performed in duplicates (B) or triplicates (C-E).

Na1 site mutants of LeuT display altered ion selectivity.
(A) Cartoon representation of the Na1 site from the outward-occluded LeuT structure (PDB ID: 2A65) with the endogenous Na1 site coordinating residues along with the substrate leucine (yellow) and E290 labeled and shown as sticks. The coordination of Na⁺ is depicted with dashed lines. Structural elements not involved in Na1 site are omitted for clarity. (B) Na⁺-mediated [³H]leucine (10x K_d) binding for Na1 site mutations, A22V (green circles), A22S (purple rhombi), N27Q (red squares), T254S (blue triangles) and N286Q (orange triangles), with WT shown for reference (black dashed line). Data are normalized to B_max and fitted to a Hill equation. (C) K⁺-dependent displacement of Na⁺-mediated [³H]leucine binding for the Na1 site mutants (colored as in (B)). Na⁺ concentrations equivalent to the EC₅₀ values (determined in (B)) and 10x K_d of [³H]leucine for each mutant were used. Data points are normalized to a control without K⁺ and modelled by the Hill equation. All data points are mean ± s.e.m., n = 3-6. The ionic strength was maintained upon substitution with Ch⁺. EC₅₀ and IC₅₀ values for (B) and (C) are summarized in Table 2.

Na⁺-mediated [³H]leucine binding and the effect of K⁺ addition, and K⁺-dependent Na⁺/[³H]leucine displacement in Na1 site mutants.
For LeuT Na1 site mutants (and K398C), Na⁺-mediated [³H]leucine binding was assayed using 10x K_d of [³H]leucine in the presence of 0, 200, and 800 mM K⁺. For K⁺-dependent displacement of Na⁺-mediated [³H]leucine binding, 10x K_d of [³H]leucine and Na⁺ equivalent to the EC₅₀ value for each of the mutants were used. Na⁺ and K⁺ were substituted for Ch⁺ to preserve the ionic strength. Data were fitted to a Hill equation, yielding EC₅₀ and IC₅₀ values reported here as mean [s.e.m. interval], n = 3–6 determined in triplicates. Comparing the Na⁺ EC₅₀ obtained for the mutants with that of WT showed significance difference for all (p<0.0001) except for E290Q (p>0.05), and for the K⁺ IC₅₀ all showed significant differences from WT (p<0.0001) expect for T254S (p>0.05) (Dunnett multiple comparison-corrected one-way ANOVA). N.d. indicates that the value is not determined. WT data are also reported in Figure 1.

LeuT mutant affinities for leucine and alanine.
Affinities were determined by inhibition of [³H]leucine with either leucine or alanine. The IC₅₀ values, obtained by non-linear regression fit, were converted to K_d and K_i values by the Cheng-Prusoff equation. For ^A-marked K_d’s, affinities were obtained by [³H]leucine saturation binding. All experiments were assayed in 200 mM Na⁺ and data are reported as mean [s.e.m. interval].

The Na1 site LeuT mutants exhibit different conformational equilibria.
(A-E) tmFRET efficiencies (1-F/F_{no site}) for N286Q^tmFRET (orange, A), N27Q^tmFRET (red, B), A22V^tmFRET (green, C), A22S^tmFRET(purple, D), and T254S^tmFRET (blue, E) incubated in 800 mM of the indicated ions and 50 µM for leucine. The His-X₃-His site was saturated with 10 mM Ni²⁺. The corresponding tmFRET efficiencies for LeuT^tmFRET (white bars) are shown for reference. All data points are mean ± s.e.m., n = 4, performed in triplicates. In (B), **p<0.001; ****p<0.0001 represent the significance levels from a Tukey multiple comparison-corrected one-way ANOVA, comparing the mean obtained in K⁺ with that in NMDG⁺, and the mean in Na⁺ and Na⁺ ± leucine with that in NMDG⁺. In (C) and (E), ns, not significant and ****p<0.0001 using a one-way ANOVA with Bonferroni multiple comparison correction.

Proposed role of K⁺ in the translocation cycle.
Here exemplified for alanine (Ala) as substrate. LeuT apo-form will likely reside in an equilibrium between its outward- (T_out) and inward-facing (T_in) conformation. After Na⁺ and Ala are bound (1) and released (3), LeuT can either i) rebind Na⁺ and Ala, which will promote efflux (2); ii) transition to the outward-open conformation in its apo-form (4), or iii) bind K⁺ (5). K⁺ binding will promote an inward-facing LeuT state (T*) which is unable to bind Na⁺, either by competitive inhibition or by promoting a state that does not allow Na⁺ binding. From here, K⁺ could either be released again to the intracellular environment (6) or be counter-transported (7).

Analysis of [³H]alanine uptake data. For time-dependent uptake, specific uptake data, normalized to the amount of active protein (see section above), was fitted to a one-phase association using GraphPad Prism 9.0.

K⁺ inhibition is preserved for LeuT in low ionic strength.
(A) Displacement of Na⁺-mediated [³H]leucine (10x K_d) binding to LeuT by 0-200 mM K⁺, using Ch⁺ (red squares) or NMDG⁺ (blue triangles) as the counter ion. Displacement with 0-1600 mM K⁺, using Ch⁺ as counter ion, is shown for reference (black dashed line, from Figure 1B). The assay is performed in the presence of 7.7 mM Na⁺. Data are normalized to a control with 0 mM K⁺ and modelled by the Hill equation. Data are shown as mean ± s.e.m. (error bars often smaller than data points), n= 3–6 conducted in triplicates. The IC₅₀ values obtained using up to 0.2 M Ch⁺ or NMDG⁺ as the counter ions were not significantly different from that obtained in a total ionic strength of 1.6 M (Dunnett’s multiple comparison test (one-way ANOVA), p = 0.84 and p = 0.99, respectively).

TmFRET principle and characterization of LeuT^tmFRET.
(A) Theoretical FRET efficiencies (1-F/F_{no site}) (left axis), converted to distances by the FRET equation (right axis), as a function of the relative population of the outward closed state. This is according to R_obs = (p_OC R_OC^-6 + p_OO R_OO^{- 6})^-6, where p and R refer to the fractions and distances, respectively, for the outward-closed (OC) and outward-open (OO) states of LeuT³⁹. The plot is based on the distances determined in (Figure 2A) between the sulfur atom of K398C (TM10) and the coordinated Ni²⁺ ion (EL4a) and relies on the assumption of a static distance distribution. (B) tmFRET efficiency (1-F/F_{no site}) as a function increasing Ni²⁺ for LeuT^tmFRET incubated in 800 mM of K⁺ (blue squares, EC₅₀ = 25.6 [24.8; 26.5] µM), K⁺ + 50 µM leucine (purple diamonds, EC₅₀ = 24.6 [24.1; 25.1] µM), Na⁺ (red triangles, EC₅₀ = 30.9 [29.1; 32.8] µM), Na⁺ + 50 µM leucine (green triangles, EC₅₀ = 26.8 [23.1; 30.4] µM) and NMDG⁺ (black circles, EC₅₀ = 85.1 [81.4; 88.8] µM). Data points are fitted to a Hill equation and EC50 values are given as mean [s.e.m. interval]. (C) tmFRET (F₀/F) as a function of increasing Rb⁺ (purple) with Na⁺ (dotted line) and K⁺ (dashed line) for reference. Data points are fitted to a Hill equation. Data (A-C) are shown as means ± s.e.m., n = 3-4 in triplicates. (D) tmFRET (F₀/F) as a function of increasing K⁺, performed on LeuT^K398C (lacking the His-X₃-His site) in the presence of 750 µM Ni²⁺. Data points are shown as mean ± standard deviation (s.d.), n = 1 in triplicates. Error bars often smaller than data points.

[³H]alanine uptake into proteoliposomes.
(A) Binding of a saturating concentration (20x K_d) of [³H]leucine to LeuT solubilized from the proteoliposomes. The amount of binding in counts per minute (c.p.m.) was taken as a measure of the relative number of active transporters in each reconstitution condition and used to adjust the uptake data in Figure 3B and C for relative reconstitution efficiency. Bar plots show the mean ± s.d. from a representative reconstitution performed in triplicates. (B) Concentration-dependent [³H]alanine uptake for 5 min in the presence of 200 mM Na⁺ into proteoliposomes containing increasing concentrations of K⁺. Lines are fits to Michaelis-Menten kinetics. The V_max in 0 mM K⁺ was defined as 100% (supplementary table 3). V_max increased with increasing intra-vesicular K⁺ (one-way ANOVA followed by a test for a linear trend, p=0.0001). None of the estimated K_m values were significantly different (Tukey multiple comparison-corrected one-way ANOVA). (C) [³H]alanine uptake for 5 min in the presences of varying Na⁺ concentration. Data are fitted to Michaelis-Menten kinetics. The Na⁺ concentration required to reach half-maximum uptake was 25 ± 8 mM (mean ± s.e.m.). Data in (B) and (C) are shown as mean ± s.e.m., n=3 in triplicates. (D) SDS-PAGE analysis of LeuT proteoliposomes following time-dependent thrombin digestion of accessible C-terminals (reducing the mass of LeuT by ∼1.3 kDa). The reaction was terminated by the addition of PMSF at the specified time points. The lanes corresponding to the time-dependent proteolysis are flanked by lanes containing proteoliposomes without thrombin (left, 0 min) or digested in the presence of DDM (right, 180 min+DDM). Arrows indicate bands of full-length (top) and cleaved (bottom) LeuT.

Rate constants for time-dependent uptake of [³H]alanine by LeuT into proteoliposomes.
Rate constants from time-dependent [³H]alanine uptake into proteoliposomes with LeuT fitted to one phase association in GraphPad Prism 9.0 (see Figure 3B).

[³H]alanine-dependent uptake by LeuT into proteoliposomes.
Constants from [³H]alanine-dependent uptake into proteoliposomes with LeuT containing increasing intra-vesicular [K⁺] fitted to Michaelis-Menten kinetics in GraphPad Prism 9.0 (see Figure 3C).

[³H]alanine uptake by LeuT into proteoliposomes with increasing intra-vesicular [K⁺]
Constants from [³H]alanine-dependent uptake into proteoliposomes with LeuT containing various intra-vesicular cations fitted to Michaelis-Menten kinetics in GraphPad Prism 9.0(see Figure 3 – Figure supplement 1B).

[³H]alanine uptake by selected Na1 mutants.
(A) [³H]alanine activity over time by LeuT A22V in the presence of 200 mM Na⁺. LeuT A22V is reconstituted into proteoliposomes with either 200 mM Na⁺ (open diamond), NMDG⁺ (square) or K⁺ (open circle). Maximal transport obtained in NMDG⁺ was defined as 100%. Maximal [³H]alanine transport in K⁺ was 99 ± 2 %. The rate constant (k) was 1.55 ± 0.20 and 1.01 ± 0.15 min^-1 in NMDG⁺ and K⁺, respectively, which correspond to a half-time to reach the concentrative capacity of 0.45 and 0.67 min. (B) [³H]alanine uptake by LeuT A22S. The [³H]alanine transport with K⁺ increased to 756 ± 37%. The rate constant (k) was 0.071 ± 0.02 and 0.044 ± 0.005 min^-1 in NMDG⁺ and K⁺, respectively, corresponding to an half-time to reach the concentrative capacity of 9.8 and 15.8 min. (C) [³H]alanine uptake by LeuT T254S. Maximal uptake in K⁺ was 422 ± 51 %. The rate constant was 0.17 ± 0.04 and 0.04 ± 0.01 min^-1 in NMDG⁺ and K⁺, respectively, corresponding to a half-time of 4.1 and 17.3 min. All data are shown as mean ± s.e.m. n = 3, performed in triplicates.

Leucine and alanine binding to Na1 site mutants.
(A,C,E,I) [³H]leucine saturation binding for WT (A), A22V (C), A22S (E) and T254S (I). (B,D,F-H,J-N) Displacement of [³Hleucine by leucine for N27Q (G), N286Q (K) and E290Q (M), or by alanine for WT (B), A22V (D), A22S (F), N27Q (H), T254S (J), N286Q (L) and E290Q (N). Binding was assayed in 200 mM Na⁺. All data points are mean ± s.e.m., n = 3-4, and fitted to a one-site binding model (A, C, E, I) or non-linear regression fit. K_d and K_i values are summarized in Table 1.

Na⁺-mediated [³H]leucine binding and inhibition by K⁺ by the LeuT mutants.
(A-F) Na⁺-mediated [³H]leucine (10x K_d) binding for the Na1 site mutants performed in the absence (red triangles) and presence of 200 mM (light blue diamond) and 800 (blue squares) mM K⁺. (A) A22V, (B) A22S, (C) N27Q, (D) T254S, (E) N286Q, (F) E290Q and (G) K398C. LeuT K398C was included as negative control. WT is shown for reference for 0 mM (dashed line), 200 mM (dotted line) and 800 mM (dash-dotted line) of K⁺. Ionic strengths were maintained using Ch⁺. All data points are shown as mean ± s.e.m. normalized to B_max and fitted to a Hill model. EC₅₀ values are summarized in Table 2, n = 3-4.

Sign up for email alerts