The "Transport Specificity Ratio": a structure-function tool to search the protein fold for loci that control transition state stability in membrane transport catalysis

Abstract

Background: In establishing structure-function relationships for membrane transport proteins, the interpretation of phenotypic changes can be problematic, owing to uncertainties in protein expression levels, sub-cellular localization, and protein-folding fidelity. A dual-label competitive transport assay called "Transport Specificity Ratio" (TSR) analysis has been developed that is simple to perform, and circumvents the "expression problem," providing a reliable TSR phenotype (a constant) for comparison to other transporters.

Results: Using the Escherichia coli GABA (4-aminobutyrate) permease (GabP) as a model carrier, it is demonstrated that the TSR phenotype is largely independent of assay conditions, exhibiting: (i) indifference to the particular substrate concentrations used, (ii) indifference to extreme changes (40-fold) in transporter expression level, and within broad limits (iii) indifference to assay duration. The theoretical underpinnings of TSR analysis predict all of the above observations, supporting that TSR has (i) applicability in the analysis of membrane transport, and (ii) particular utility in the face of incomplete information on protein expression levels and initial reaction rate intervals (e.g., in high-throughput screening situations). The TSR was used to identify gab permease (GabP) variants that exhibit relative changes in catalytic specificity (kcat/Km) for [14C]GABA (4-aminobutyrate) versus [3H]NA (nipecotic acid).

Conclusions: The TSR phenotype is an easily measured constant that reflects innate molecular properties of the transition state, and provides a reliable index of the difference in catalytic specificity that a carrier exhibits toward a particular pair of substrates. A change in the TSR phenotype, called a Delta(TSR), represents a specificity shift attributable to underlying changes in the intrinsic substrate binding energy (DeltaGb) that translocation catalysts rely upon to decrease activation energy (Delta G(T)(++). TSR analysis is therefore a structure-function tool that enables parsimonious scanning for positions in the protein fold that couple to the transition state, creating stability and thereby serving as functional determinants of catalytic power (efficiency, or specificity).

Figure 1

Results from TSR analysis are valid across a broad range of competing-substrate concentration ratios The Transport Specificity Ratio (TSR) is calculated using results from a dual-label competitive uptake assay in which structurally distinct, labelled substrates are allowed to compete for transport at the same active site. Panel A: Mixtures of 10 μM [³H]NA (0.6 μCi/ml) and 10 μM [¹⁴C]GABA (0.2 μCi/mL) were prepared such that [NA] + [GABA] = 10 μM. E. coli strains SK105 (GabP-positive) and SK45 (GabP-negative) were exposed in parallel experiments for 10 seconds at 30°C to substrate mixtures containing the indicated concentrations of [³H]NA. The GabP-dependent (SK105 minus SK45) uptake of either [³H]NA (■) or [¹⁴C]GABA (▲) may be read from the left-side ordinate. The calculated TSR (Equation. 6) may be read from the right-side ordinate (○). Panel B: The substrate concentrations were varied in constant proportion such that the GABA concentration (ranging from 1.8–31.5 μM) was always 42.9 percent of the NA concentration (ranging from 4.2–73.4 μM). The radiochemical concentrations for [³H]NA and [¹⁴C]GABA were 0.23 μCi/ml and 0.03 μCi/ml, respectively. The indicated concentration ranges produce about 50 percent combined active site occupancy (bound GABA plus NA) – since the affinities for GABA and NA are 40 μM and 200 μM, respectively [25].

Figure 2

Results from TSR analysis are valid across a broad range of carrier expression levels E. coli strains SK11 (GabP-positive) and SK45 (GabP-negative) were grown to early logarithmic phase as described in Methods except that expression was induced by exposing cultures to the indicated IPTG concentrations. The cells were washed with 100 mM potassium phosphate buffer (pH 7.0), and dual-label competitive transport reactions were initiated by exposing the cells to 7 μM [³H]NA (0.42 μCi/ml) and 3 μM [¹⁴C]GABA (0.06 μCi/ml) for 10 seconds (initial rate) at 30°C. Error bars represent the S.E.M. (n = 3). Panel A. GabP-dependent uptake (SK11 signal minus SK45 signal) of either [³H]NA (■) or [¹⁴C]GABA (▲). Panel B. Transport Specificity Ratio (GABA/NA). Inset. Immunoblot of plasma membrane vesicle protein (2 μg per lane) probed with an anti-pentaHis mAb and developed with a chemiluminiscent alkaline phosphatase substrate (see Methods). Lane 1: Membranes from E. coli strain SK45 (GabP-negative). Lanes 2–10: Membranes from E. coli SK11 (GabP-positive) grown in the presence of 2, 5, 10, 20, 50, 100, 200, 500, or 1000 μM IPTG, respectively.

Figure 3

Results from TSR analysis are valid across a broad range of reaction times E. coli strain SK11 (GabP-positive) was exposed simultaneously to 6 μM [³H]NA (0.42 Ci/ml) and 4 μM [¹⁴C]GABA (0.06 Ci/ml) for the indicated times at 30°C. Parallel experiments were carried out in the presence of 2 mM GABA, which was included to block the GabP. Panel A shows the GabP-dependent component of competitive uptake (difference between the parallel experiments) over a 10-fold time range. The red arrow indicates a probable mechanical error, causing low uptake inconsistent with other points on the curve. The Panel B shows the GABA to NA mole ratio (left-side ordinate) calculated from data shown in the Panel A. The associated TSR values may be read from the right-side ordinate. The red arrow has the same meaning as in the Panel A, and serves here to emphasize the reliability of the TSR analysis, which has self-correcting properties that compensate for many routine sample processing problems that may cause inconsistency in times or volumes (see discussion).

Figure 4

Variants of the E. coli GabP that exhibit Δ(TSR) phenotypes Using data analogous to Figure 1, the concentrations of competing substrates were adjusted empirically such that the initial rates of label accumulation were superimposed for E. coli strains expressing the "control" gab permease (GabP). As a result, any separation between initial rate uptake curves for [¹⁴C]GABA (▲) and [³H]NA (■) provides a highly intuitive visual representation of a Δ(TSR) phenotype. Panel N302C shows TSR analysis of the single-Cys GabP variant, N302C. Compared to the Cys-less GabP control (TSR = 8) for which the initial label accumulation rates are superimposed [5], the N302C shows a relative increase in the specificity for NA with a calculated TSR of 2.5. The Panel INS Ala 320 shows TSR analysis of the GabP variant, INS Ala 320, which has an extra alanine residue inserted at position 320. Compared to the wild type GabP control (TSR = 4) for which the initial label accumulation rates are superimposed [6], the INS Ala 320 exhibits a relative increase in specificity for GABA (i.e., opposite of the Panel N302C) with a calculated TSR of 16.

Figure 5

Changes in catalytic specificity (k_cat/K_m) reflect underlying changes in transition state binding energy (ΔG_b) In this description of catalysis, (i) the magnitude of the non-catalysed activation energy () does not depend on a favourable protein-substrate interaction in the transition state, (ii) the catalysed translocation energy barrier is taken as the Gibbs Energy difference () between the free reactants (C + S) and the transition state complex (CS^‡), and (iii) intrinsic substrate binding energy is recognizable as the decisive factor in transition state stabilization. Thus, translocation catalysts (C) will use intrinsic substrate binding energy (ΔG_b) to stabilize the transition state (CS^‡). The role of ΔG_b in lowering the transition state energy barrier compared to a non-catalyzed reaction () may be deduced with aid from the accompanying energy diagrams, which emphasize several instances wherein the thermodynamic distance represented by one coloured arrow equals the summed distance represented by two shorter arrows of the same colour. The illustrated thermodynamic relationships are restated (with proper attention to sign convention) in equations A (red), B (green), and C (blue). Substituting A and C into B yields the fundamental relationship, (boxed), which says that the uncatalysed activation energy (, algebraically positive) is diminished by intrinsic substrate binding energy, ΔG_b (algebraically negative), which is the underlying parameter that TSR analysis probes (Eqn. 9). Note: These energy diagrams compare non-catalytic (dots and dashes) and catalytic (solid line) proteins. Imposition of a binding-averse interaction (ΔG_R) is seen to de-stabilize the Michaelis complex (CS, red arrows) in the catalytic protein. Subsequent attainment of favourable transition state complementarity (i.e., via conformational transitions that relieve ΔG_R , blue arrows) results in use of binding energy to stabilize the transition state complex (CS^‡). This internal ''give-and-take,'' involving ΔG_R is reflected in its algebraic cancellation when equations A, B, and C are combined to yield the boxed equation (text Eqn. 1), which says that intrinsic substrate binding energy decreases the energy barrier () for a translocation reaction carried out from solution (i.e., directly from the free carrier and substrate (C + S) to the transition state). When C and S are free in solution, the effective second-order rate constant associated with is k_cat/K_m, the specificity parameter compared in the dual-substrate TSR analysis (Equation.5). That k_cat/K_m should be associated with the free reactants may be appreciated by considering the Michaelis-Menten Equation when S << K_m, and CS complexes do not exist in appreciable amounts (see Discussion).

Figure 6

Comparison of equilibrium binding versus TSR analysis Envisage a catalytic protein interacting with two substrates (or substrate analogs), one exhibiting high-affinity binding (dashed RED line), and the other low-affinity binding (solid BLUE line). Equilibrium binding to the stable Michaelis complex (LEFT, Panels A and B) would produce concentration-dependent saturation of the binding site (Panel B). From the observed affinity difference (ΔK_d) between the two substrates, one can calculate a corresponding difference in binding energy, ΔΔG_S (Panel A), for the two substrates interacting with the stable Michaelis complex at the bottom of the reaction coordinate. In contrast, information on the interaction of substrates at the reaction coordinate peak would require a study of binding to the unstable transition state (RIGHT, Panels C and D). Unfortunately, due to the high energy-level and transient nature and of the transition state (denoted by ‡), the relevant binding experiment (Panel D) is technically impossible. However, TSR analysis allows direct calculation (Equation 9) of the transition state binding energy difference, ΔΔG_b (Panel C, yellow) between two competing substrates, A and B. A change in the TSR phenotype, or Δ(TSR), thus provides evidence for a change in the graphical separation distance, (Panel D, yellow), for the "impossible experiment" on substrate binding to the unstable transition state. Thus, observation of a Δ(TSR) phenotype reflects underlying structural changes that affect binding discrimination between substrates A and B in the transition state, which are of interest because transition state binding interactions create transport catalysis [2–4, 7] by lowering the activation energy, , and increasing k_cat/K_m. In summary, the equilibrium binding experiment depicted on the left does not address catalysis per se, whereas the TSR experiment depicted on the right does.