Squeezing at Entrance of Proton Transport Pathway in Proton-translocating Pyrophosphatase upon Substrate Binding

Background: H⁺-PPase is a proton pump that hydrolyzes PP_i to transport protons for maintenance of pH homeostasis.

Results: The FRET efficiencies between two corresponding residue pairs at the pathway entrance were altered upon substrate binding.

Conclusion: Substrate mediates squeezing at the pathway entrance of H⁺-PPase.

Significance: Squeezing at the pathway entrance is essential for the initial step of proton translocation.

Abstract

Homodimeric proton-translocating pyrophosphatase (H⁺-PPase; EC 3.6.1.1) is indispensable for many organisms in maintaining organellar pH homeostasis. This unique proton pump couples the hydrolysis of PP_i to proton translocation across the membrane. H⁺-PPase consists of 14–16 relatively hydrophobic transmembrane domains presumably for proton translocation and hydrophilic loops primarily embedding a catalytic site. Several highly conserved polar residues located at or near the entrance of the transport pathway in H⁺-PPase are essential for proton pumping activity. In this investigation single molecule FRET was employed to dissect the action at the pathway entrance in homodimeric Clostridium tetani H⁺-PPase upon ligand binding. The presence of the substrate analog, imidodiphosphate mediated two sites at the pathway entrance moving toward each other. Moreover, single molecule FRET analyses after the mutation at the first proton-carrying residue (Arg-169) demonstrated that conformational changes at the entrance are conceivably essential for the initial step of H⁺-PPase proton translocation. A working model is accordingly proposed to illustrate the squeeze at the entrance of the transport pathway in H⁺-PPase upon substrate binding.

Introduction

H⁺-PPase² (EC 3.6.1.1) and H⁺-ATPase (EC 3.6.1.3) are the major primary proton pumps generating proton gradients to drive secondary transport of many substances across biological membranes (1–3). Unlike H⁺-ATPase hydrolyzing the “hard currency” of bio-energy ATP, H⁺-PPase efficiently uses the metabolically “wasting byproduct,” PP_i, as an energy source for H⁺ translocation (4, 5). H⁺-PPases are primarily present in higher plants, algae, protozoa, archaebacteria, and bacteria but not in animals (6–8). H⁺-PPases are classified into two functional subfamilies: K⁺-dependent (Type I) and K⁺-independent (Type II) H⁺-PPases (6, 9–10). Both types strictly require Mg²⁺ combining with PP_i as an actual substrate (Mg-PP_i) for function and structural stabilization of the H⁺-PPase (11).

H⁺-PPase is a homodimer containing a single type of polypeptide with a molecular mass ranging from 60 to 80 kDa among various species (12, 13). A three-dimensional structure of Vigna radiata H⁺-PPase (VrH⁺-PPase) binding with substrate analog, imidodiphosphate (IDP) exhibits a concentric structure with 6 core transmembrane helices (TMs; TM5, -6, -11, -12, -15, and -16) and 10 surrounding helices (14). The core TMs encompass hydrophilic regions accommodating the catalytic site facing the cytosolic side and hydrophobic portions embedding the proton transport pathway through the membrane (14). Acidic motifs, PP_i binding motif, and several highly conserved charged residues compose the major section of the catalytic site and play significant roles in PP_i hydrolysis reaction (11, 15, 16). Further trypsinolysis analysis and structural mapping for the catalytic sites as determined by single molecule FRET (smFRET) technique demonstrated that H⁺-PPase exists at least in three major conformations during the catalytic cycle (14, 17). In addition, site-directed mutagenesis and crystal structure studies suggested that four essential residues, Arg-242, Asp-294, Lys-742, and Glu-301 along the proton transport pathway in VrH⁺-PPase convey protons from the catalytic site toward the luminal side (14, 18–21). Furthermore, Arg-242 in the vicinity of the catalytic site has been considered to likely be the first proton-carrying residue for proton translocation (14, 16, 21, 22). This putative first residue along with several polar residues surrounding the entrance of the translocation pathway may play a crucial role at the initial step of proton translocation in H⁺-PPase (14).

In this study the smFRET technique was employed to investigate the residues surrounding the entrance of the transport pathway in Clostridium tetani H⁺-PPase (CtH⁺-PPase) (Fig. 1), where the initial event of proton translocation was proposed to occur. The efficiency of energy transfer from donor and acceptor labeling at two introduced cysteine residues on TM6 of both subunits (the 219–219 pair) was decreased, whereas that on TM16 (the 642–642 pair) was increased upon binding with Mg-IDP. These results suggested that the entrance of the proton transport pathway was squeezed upon substrate binding to H⁺-PPase. Moreover, smFRET measurement of R169K/G219C and R169K/S642C mutants demonstrated that the motion at the entrance is crucially relevant to the initial step of proton translocation. Accordingly, a working model is proposed delineating substrate mediating squeeze at the entrance of the proton transport pathway in H⁺-PPase.

FIGURE 1.

Essential residues for proton translocation and labeling positions for smFRET in the predicted structural models of CtH⁺-PPase. The homology model of CtH⁺-PPase was constructed based on the crystal structure of VrH⁺-PPase (PDB 4A01) by SWISS-MODEL workspace (31, 32). The secondary (A) and crystal (B) structures exhibit 16 TMs, 8 larger cytosolic loops with essential motifs forming the catalytic site (orange), and 7 smaller loops with N and C termini facing outside the membrane. Residues involved in proton translocation are shown as red circles in A. Residues substituted to cysteine and further labeled with maleimide-reactive probes are Val-170 (pink), Gly-219 (blue), Val-411 (green), Gly-454 (cyan), Ala-598 (brown), and Ser-642 (black). C, shown is sequence alignment of H⁺-PPases at essential residues near the entrance of the proton transport pathway. The alignments were obtained by the ClustalW program as in the previous study (15). The single letter abbreviations with a black background stand for highly conserved residues and with the gray background stand for moderately conserved residues with similarity exceeding 60%.

EXPERIMENTAL PROCEDURES

Expression and Preparation of Mutant CtH⁺-PPases

Single cysteine mutants were generated on the background of the cysteine-less mutant by the QuikChange site-directed mutagenesis method (23). Escherichia coli C43(DE3) strain (Lucigen, Middleton, WI) containing the plasmid with single cysteine substitution of CtH⁺-PPase was grown at 37 °C in Luria-Bertani medium supplemented with 50 μg/ml ampicillin. Overnight cell solutions were diluted 50-fold and incubated for 2–3 h at 37 °C until A₆₆₀ = 0.8; isopropyl β-d-thiogalactoside was subsequently added to a final concentration of 0.5 mm. Cultures were further grown for 4 h at 37 °C, and cells were finally harvested by centrifugation at 4000 × g for 15 min. The microsomal membrane fraction was isolated as described previously (17). CtH⁺-PPase was purified from microsomal membranes according to the previous protocol (17). Fluorescence pairs (donor/acceptor), Alexa Fluor 488/647 (maleimide-reactive probe; Molecular Probes, Invitrogen), and ATTO 495/680 (maleimide-reactive probe; ATTO-TEC GmbH, Siegen, Germany) were individually incorporated into the purified single cysteine CtH⁺-PPase mutants based on the protocols described previously (17). The labeled CtH⁺-PPase was then mixed at 1:500 to a 1:1000 molar excess of asolectin liposomes (24). The mixture was gently shaken at 4 °C for 12 h. The reconstituted CtH⁺-PPase mutants were collected by 75000 × g for 55 min at 4 °C.

Fluorescence Spectroscopy and smFRET Data Analysis

Single molecule fluorescence images for statistical analysis on FRET efficiencies were obtained by total internal reflection fluorescence microscopy. Olympus IX71 inverted microscopy (Olympus, Tokyo, Japan) used in the present study was equipped with argon (488 nm, 10 milliwatts) (Olympus) and red diode (638 nm, 35 millwatts) (Omicron Laserage, Rodgau, Germany) lasers for donor and acceptor excitation, respectively. The fluorescence emissions from donor (Alexa Fluor 488 and ATTO 495) and acceptor (Alexa Fluor 647 and ATTO 680) were separated by an emission splitter (U-SIP, Olympus) with a 600-nm dichroic mirror and band pass filters (Omega Optical, Brattleboro, VT) appropriate to the donor (520-DF-40) and acceptor (670-DF-40), respectively. The images were then recorded by a sensicam electron-multiplying charge coupled device camera (Cooke, Romulus, MI).

FRET efficiency from probes on each single protein molecule was calculated according to Equation 1 (25, 26),

where I_D and I_A are the background-subtracted fluorescence intensities of donor and acceptor as FRET occurs. ψ_D and ψ_A are the quantum yields of donor and acceptor fluorophores, respectively. The ratios for Alexa Fluor 488/647 and ATTO 495/680 detection efficiencies (η_A/η_D) are 1.01 and 0.41, respectively, according to specification provided by manufacturers (Sensicam electron-multiplying charge coupled device camera, Cooke; Omega Optical). Energy transfer efficiencies (E) acquired from single protein molecules were converted into a histogram. The distance between donor and acceptor fluorophores was then determined based on Förster theory, as shown in Equation 2 (27),

For determining Förster distance (R_o), the orientation factor κ² was appropriately taken as the isotropically averaged value of 2/3 (28, 29). Förster distances between Alexa Fluor 488/647 and ATTO 495/680 fluorescent pairs were thereby obtained as 56 and 42 Å, respectively (17, 30).

RESULTS AND DISCUSSION

Characterization of CtH⁺-PPase Mutants

The CtH⁺-PPase gene with a His₆ was constructed into pET23d vector and heterologously expressed in E. coli strain C43(DE3) under the control of the T7/lac promoter. Microsomal CtH⁺-PPase was then isolated, and its PP_i hydrolysis and proton translocation activities were determined. Like VrH⁺-PPase, the enzymatic function of CtH⁺-PPase could be stimulated by K⁺ or Rb⁺ but only slightly or negligibly changed by Li⁺, Na⁺, and most divalent ions, except for Mg²⁺ (data not shown). These results convincingly demonstrate that CtH⁺-PPase belongs to type I H⁺-PPase (6).

An essential arginine at TM5 (Arg-169 in CtH⁺-PPase; Arg-242 in VrH⁺-PPase) was proposed to form a hydrogen bond with a water molecule at the junction between the catalytic site and the entrance of proton transport pathway (14) (Fig. 1C). Site-directed mutagenesis analysis definitively demonstrated that Asp-252, Met-256, and Asp-259 at TM6 in Streptomyces coelicolor H⁺-PPase, corresponding to Asp-214, Met-218, and Asp-221 of CtH⁺-PPase, are crucial for proton pumping activity (16) (Fig. 1C). In addition, highly conserved Asp-412 in TM11 specifically binds the proton translocation inhibitor, N,N′-dicyclohexylcarbodiimide (21). Moreover, mutation at Lys-541 in TM12 of VrH⁺-PPase (Lys-453 in CtH⁺-PPase) provoked serious impairment in both PP_i hydrolysis and H⁺ translocating activities (15) (Fig. 1C). According to the predicted three-dimensional structure of CtH⁺-PPase, these residues (Arg-169, Asp-214, Met-218, Asp-221, Asp-412, and Lys-453) and several polar residues in TM15 and TM16 reside at/near the entrance of transport pathway. To investigate the action of the initial event of proton translocation at the pathway entrance, a series of single cysteine mutants was generated under a cysteine-less background for further smFRET analysis. A single cysteine was individually incorporated into the positions at Ala-168, Val-170, Gly-217, Gly-219, Val-411, Ala-413, Gly-452, Gly-454, Asn-597, Ala-598, Gly-640, and Ser-642, all of which are in the vicinity of the pathway entrance in CtH⁺-PPase (Fig. 1). All the above mutants except A413C were successfully expressed in E. coli strain C43(DE3) (Fig. 2A). PP_i hydrolysis and proton translocation activities of these mutants were determined, and the coupling efficiencies were subsequently calculated as shown in Fig. 2, B–D. Substitution at Ala-168 brought an apparent increase (33%) in PP_i hydrolysis as compared with the cysteine-less variant. However, G217C, G454C, and G640C mutants showed dramatic aberrations of PP_i hydrolysis and proton translocation activities, suggesting these H⁺-PPase mutants might presumably undergo serious variations in their structures upon substitution. Furthermore, V170C, G219C, V411C, G452C, N597C, A598C, and S642C mutants displayed similar functional properties as the wild type or the cysteine-less mutant (Fig. 2, B–D). These single cysteine mutants could, therefore, be used as valuable subjects for the following fluorescent labeling.

FIGURE 2.

Expression levels and characteristics of microsomal CtH⁺-PPase mutants. A, shown are expression levels of CtH⁺-PPase mutants. 20 μg of microsomal membrane fractions were prepared from E. coli transformed by wild type and mutant CtH⁺-PPase DNA. Immunoblotting was performed using the monoclonal anti-His antibody. B, PP_i hydrolysis activity of the mutants is shown. C, shown is proton translocation activity. D, shown is coupling efficiency. Coupling efficiency was calculated as the ratio of proton translocation activity to PP_i hydrolysis activity (ΔF%/μmol of PP_i hydrolyzed). ΔF% is the difference in relative fluorescence quenching of acridine orange. Values are the means ± S.D. from at least three independent experiments. *, not detectable.

Effects of Ligands on FRET Efficiency

For proper smFRET determination, each CtH⁺-PPase mutant was incubated simultaneously with both donor and acceptor fluorescent probes. The labeled CtH⁺-PPases were then immobilized on slides coated with l-polylysine through electrostatic interactions. When single molecule homodimeric CtH⁺-PPase was conjugated with both donor and acceptor, excitation by the 488-nm laser beam would generate Alexa Fluor 488 or ATTO 495 fluorescence signals in the donor channel (data not shown). Concomitantly, the emission corresponding to the Alexa Fluor 647 or ATTO 680 was observed in the acceptor channel (data not shown), indicating the energy transfer from donor to acceptor within a single CtH⁺-PPase molecule. To further confirm the existence of acceptor probe in this single molecule, a 638-nm laser was utilized to illuminate the labeled CtH⁺-PPase. The fluorescence from the acceptor probe was procured in the acceptor channel as anticipated (data not shown), demonstrating the coexistence of both donor and acceptor fluorophores on the same single homodimeric CtH⁺-PPase molecule. In addition, these dual-labeled CtH⁺-PPase mutants exhibited appropriate PP_i hydrolysis activities, presumably showing that the fluorophore labeling did not substantially perturb functionality and structure of the enzyme (data not shown). These results verified that the current system is definitely feasible for further smFRET measurements (cf. Refs. 17 and 25).

In the present study the G219C mutant labeled with Alexa Fluor 488 as FRET donor and Alexa Fluor 647 as FRET acceptor gave a single peak in the smFRET efficiency histogram at 55.2% (Fig. 3A, 219–219 pair). The corresponding distance between the two fluorescent probes on the 219–219 pair was calculated using Förster theory (Equation 2) as 54.1 Å (Table 1). Furthermore, the addition of Mg-IDP resulted in an increase of FRET efficiency (E* = 63.8%) (Fig. 3C, 219–219 pair), suggesting a decrease in distance between the two fluorescent probes on the 219–219 pair (Table 1). Similarly, the substrate analog-induced distance variation at TM6 of the pathway entrance was observed in the presence of K⁺ (E* = 53.1% in the presence of K⁺ only; E* = 61.3% in the presence of K⁺/Mg-IDP) (Fig. 3, B and D, 219–219 pair and Table 1). The increased energy transfer ratio (∼15%) for the G219C mutant indicated that two sites at the entrance of the proton transport pathway on both TM6 in homodimeric CtH⁺-PPase approach each other upon substrate binding. G219C CtH⁺-PPase mutant binding with hydrolyzed product, P_i, brought the FRET efficiency to 52.0% (Fig. 3E, 219–219 pair), revealing a similar FRET ratio as in the absence of ligands.

FIGURE 3.

Histograms of FRET efficiencies. Energy transfer efficiencies for mutant CtH⁺-PPases with fluorescence donor and acceptor probes labeled on 170–170, 219–219, 411–411, 452–452, 598–598, and 642–642 were determined without adding ligands (A) and in the presence of K⁺ (B), Mg-IDP (C), K⁺/Mg-IDP (D), and P_i (E).

TABLE 1.

FRET efficiencies and calculated distances between probes on cysteine pairs in mutants

The value of FRET efficiencies (E*, %) are the mean (± S.D.) from the indicated number of single molecules. Numbers of single molecule proteins analyzed are shown in parentheses. The distances (R) were obtained from E* using Equation 2.

TMs (Pairs)	Ligand-free	Mg-IDP	K+	K+/Mg-IDP	Pi
5 (170–170)
E* (%; n)	83.4 ± 6.7 (80)	86.2 ± 5.5 (81)	84.4 ± 5.2 (77)	84.3 ± 5.7 (81)	85.0 ± 5.2 (73)
R (Å)	42.8	41.3	42.3	42.3	41.9
6 (219–219)
E* (%; n)	55.2 ± 9.8 (87)	63.8 ± 9.3 (91)	53.1 ± 9.2 (83)	61.3 ± 8.0 (81)	52.0 ± 8.0 (75)
R (Å)	54.1	51.0	54.9	51.9	55.3
11 (411–411)
E* (%; n)	80.8 ± 6.0 (83)	80.8 ± 5.4 (79)	81.1 ± 5.6 (77)	84.2 ± 5.5 (84)	84.1 ± 6.0 (83)
R (Å)	44.1	44.1	43.9	42.4	42.4
12 (452–452)
E* (%; n)	85.7 ± 5.2 (73)	86.9 ± 5.6 (72)	86.1 ± 6.1 (68)	84.9 ± 5.9 (75)	86.4 ± 5.7 (81)
R (Å)	41. 6	40.9	41.3	42.0	41.1
15 (598–598)
E* (%; n)	62.6 ± 6.2 (72)	66.1 ± 8.6 (83)	63.3 ± 7.3 (75)	61.4 ± 8.0 (75)	63.6 ± 8.2 (78)
R (Å)	38.5	37. 6	38.4	38.9	38. 3
16 (642–642)
E* (%; n)	90.1 ± 4.6 (87)	82.8 ± 5.9 (82)	90.2 ± 4.2 (87)	84.6 ± 5.0 (80)	91.2 ± 3.6 (74)
R (Å)	38.8	43.1	38.7	42.2	37.9

Residue 642 at the pathway entrance in both TM16 in the CtH⁺-PPase dimer were replaced with cysteine for fluorescent probes labeling (Fig. 1). The FRET efficiency of S642C CtH⁺-PPase was determined as 90.1% in the absence of ligands (Fig. 3A, 642–642 pair). The binding of the substrate analog with Mg²⁺, Mg-IDP, to the mutant CtH⁺-PPase further decreased smFRET efficiency to 82.8% (Fig. 3C, 642–642 pair), showing that the distance between donor and acceptor probes at both residues 642 on TM16 increased to 43.1 Å (Table. 1). Likewise, this substrate-induced action at the pathway entrance on TM16 occurred in the presence of K⁺ (E* = 90.2% in the presence of K⁺ only; E* = 84.6% in the presence of both K⁺ and Mg-IDP) (Fig. 3, B and D, 642–642 pair). These measurements indicated that the two corresponding residues at the entrance of the proton transport pathway in TM16 in homodimeric CtH⁺-PPase move apart from each other upon substrate analog binding. Nevertheless, the addition of P_i to the S642C mutant exhibited similar FRET efficiency and proximity as in the absence of ligands (Fig. 3E, 642–642 pair, and Table 1).

According to the predicted crystallographic structure of CtH⁺-PPase, TM16 is relatively close to the dimer interface of CtH⁺-PPase, whereas TM6 is relatively distant from it (Fig. 1) (14). Therefore, the increased proximity between the 642–642 pair at TM16 and the decreased distance between the 219–219 pair at TM6, as demonstrated above by substrate analog binding determined using smFRET, revealed a geometrical squeeze along the axis between TM6 and -16 at the entrance of the proton transport pathway in H⁺-PPase. The conformation at the sites on TM6 and -16 in the entrance of the proton transport pathway might be restored to its original state after the enzymatic reaction (in the presence of product, P_i).

Dual labeling of fluorescent donor and acceptor probes at position 170 on both TM5 at the pathway entrance of dimeric CtH⁺-PPase showed a FRET efficiency of 83.4% in the absence of ligands (Figs. 1 and 3A, 170–170 pair). Further addition of K⁺, Mg-IDP, K⁺/Mg-IDP, or P_i to the V170C CtH⁺-PPase mutant displayed smFRET histograms similar to those in ligand-free conditions (Fig. 3, B–E, 170–170 pair). Neither essential cofactors nor substrate analog and hydrolyzed product led to significant changes of the energy transfer ratio, suggesting that binding of these ligands might not induce conformational changes at the first proton-carrying residue (Arg-169) of the pathway entrance in homodimeric CtH⁺-PPase.

To observe the action of sites at the pathway entrance on TM11 and 12 upon ligand binding, Val-411 and Gly-452 were selected for the smFRET measurement (Fig. 1). Although, the N, N′-Dicyclohexylcarbodiimide binding site, Asp-412, and K⁺ binding site, Ala-450, are flanked by the entrance of the proton transport pathway, neither essential cofactors (K⁺ and Mg²⁺) nor substrate analog (IDP) and hydrolyzed product (P_i) of CtH⁺-PPase induced any apparent change in the energy transfer efficiency for V411C and G452C mutants (Fig. 3, 411–411 and 452–452 pairs and Table 1). These results indicated that binding of ligands to CtH⁺-PPase does not elicit conformational variations at sites on TM11 and 12 on the entrance of the transport pathway.

To accurately measure the relatively shorter distance between donor and acceptor probes on two positions (the 598–598 pair) at TM15 in the pathway entrance on each subunit, another fluorescence donor and acceptor pair, ATTO 495 and 680, which has a Förster critical distance of 42 Å, was used. The major peak in the FRET efficiency histogram of the A598C mutant in the absence of ligands was 62.6% (Fig. 3A, 598–598 pair), yielding a distance of 38.5 Å for the 598–598 pair (Table 1). However, the addition of essential cofactors, substrate analog, or hydrolyzed product produced negligible changes on the energy transfer efficiency for the A598C mutant (Fig. 3, B–E, 598–598 pair), showing that ligands do not induce conformational changes at the site in TM15 on the pathway entrance of CtH⁺-PPase.

FRET Efficiencies on Mutant CtH⁺-PPases Impaired in Proton Translocation

A previous report demonstrated that Arg-242 in VrH⁺-PPase mutated to Ala provoked the impairment in both PP_i hydrolysis and proton translocating activities (19). In the present study the mutant in which Arg-169 of CtH⁺-PPase was replaced with the lysine residue possessed similar PP_i hydrolysis properties to the cysteine-less mutant. However, the R169K mutant showed a lower apparent proton translocating activity and coupling ratio (Fig. 2), suggesting that this highly conserved arginine residue at TM5 was involved in coupling of proton pumping and substrate hydrolysis functions (14, 21, 22). To further explore the association of the motion at the pathway entrance to the initial step of proton function in H⁺-PPase, 219–219 and 642–642 cysteine pairs were individually introduced into the R169K mutant under the cysteine-less template (R169K/G219C and R169K/S642C mutants). The R169K/G219C mutant labeled with the Alexa Fluor 488/647 pair displayed a single peak in the FRET efficiency histogram at 58.6% (Fig. 4A, 219–219 pair and Table 2). The corresponding distance between donor and acceptor probes on the 219–219 pair in the R169K/G219C mutant was calculated as 52.8 Å (Table 2), revealing a similar energy transfer efficiency to the G219C mutant (E* = 55.2%) (Fig. 3A, 219–219 pair). This smFRET measurement indicated that lysine replacement at position Arg-169 might not bring profound structural effects at the site in the pathway entrance on TM6 within homodimeric CtH⁺-PPase even though the substitution impaired proton translocation. Like the G219C mutant above, the addition of K⁺ alone generated no significant variation in the energy transfer efficiency between two fluorescent probes on the 219–219 pair in the R169K/G219C mutant (Fig. 4B, 219–219 pair, and Table 2). However, the CtH⁺-PPase mutant binding with substrate analog and Mg²⁺ yielded a FRET efficiency of 58.5 and 56.3% in the absence and presence of K⁺, respectively (Fig. 4, C and D, 219–219 pair, and Table 2), which differ from the FRET histogram results of the G219C mutant (E* = 63.8 and 61.3% with and without K⁺, respectively). The increase of fluorescence energy transfer efficiency between donor and acceptor probes on the 219–219 pair did not occur in the mutant with Arg-169 substitution. The smFRET measurement suggested that the motion at pathway entrance on TM6 is conceivably relevant to the function of the proton transport pathway in H⁺-PPase.

FIGURE 4.

Histograms of FRET efficiencies for CtH⁺-PPase mutants impaired in proton translocation. Energy transfer efficiencies between fluorescence probes labeled on 219–219 and 642–642 dye pairs for R169K mutant were obtained in the absence of ligands (A) and in the presence of K⁺ (B), Mg-IDP (C), K⁺/Mg-IDP (D), and P_i (E).

TABLE 2.

FRET efficiencies and calculated distances between probes on 219–219 and 642–642 cysteine pairs of mutant CtH⁺-PPases

The value of FRET efficiencies (E*, %) are the mean (± S.D.) from the indicated number of single molecules. Numbers of single molecule proteins analyzed are shown in parentheses. The distances (R) were obtained from E* using Equation 2.

Mutants	Ligand-free	Mg-IDP	K+	K+/Mg-IDP	Pi
R169K/G219C
E* (%; n)	58.6 ± 9.0 (68)	58.5 ± 7.7 (65)	58.3 ± 8.1 (74)	56.3 ± 7.3 (70)	57.9 ± 8.7 (75)
R (Å)	52.8	52.9	53.0	53.7	53.1
R169K/S642C
E* (%; n)	90.1 ± 4.3 (68)	89.6 ± 5.2 (71)	90.3 ± 4.9 (67)	90.2 ± 4.3 (71)	89.4 ± 4.7 (79)
R (Å)	38.8	39.1	38.6	38.7	39.3

Similarly, the 642–642 pair was engineered into the R169K mutant (R169K/S642C) to examine the impact on structural transitions at the pathway entrance on TM16 in CtH⁺-PPase. The R169K/S642C mutant displayed similar PP_i hydrolysis activity to the cysteine-less CtH⁺-PPase mutant (110% of cysteine-less mutant) (Fig. 2). However, the coupling ratio of enzymatic activity and proton transport function was decreased to ∼45% that of the cysteine-less mutant. This result indicated that the R169K/S642C mutant showed impairment in its proton translocation but not catalytic functions. The following smFRET measurement yielded an energy transfer efficiency of as 90.1% for the R169K/S642C mutant in the absence of ligands (Fig. 4A, 642–642 pair, and Table 2). The corresponding distance between both probes on the 642–642 pair was thereby estimated to be 38.8 Å (Table 2), indicating that residue 169 replacement could not induce the apparent conformational variations at the position on the pathway entrance in dimeric CtH⁺-PPase. Furthermore, neither Mg-IDP nor K⁺/Mg-IDP binding to the R169K/S642C CtH⁺-PPase mutant caused considerable alteration in the energy transfer efficiency (Fig. 4, C and D, 642–642 pair, and Table 2), indicating that mutation at the initial proton carrier residue, Arg-169, might restrict the action at the pathway entrance on TM16. This result verified that the structural action at the entrance of the proton transport pathway on TM16 is presumably relevant to proton translocation of H⁺-PPase.

The present smFRET study is the first report to demonstrate the coupling between the proton transport pathway and catalytic site in the native state of H⁺-PPase on a structural basis. The fluorescence energy transfer efficiency between donor and acceptor fluorescent probes at the 219–219 pair on TM6 was increased, whereas that for the 642–642 pair on TM16 was decreased upon substrate analog binding to CtH⁺-PPase. The proximities between probes on cysteine pairs on TM5 and -12 showed negligible variation in the presence of ligands. The most reasonable explanation to reconcile this evidence is that the substrate mediates a geometrical squeeze along the axis between TM6 and -16 at the entrance of the proton transport pathway during the initial event without alteration in that between TM5 and -12 (Fig. 5). In other words, the structure at the entrance of the proton transport pathway was transformed from a relatively elliptical shape into more a spherical form (Fig. 5). Furthermore, according to the crystallographic structure of VrH⁺-PPase bound to the substrate analog, Arg-169 and Asp-221 were proposed to form hydrogen bonds with a water molecule near the entrance of the proton transport pathway (14). Arg-169 was also suggested to form salt bridges with Asp-214 and Asp-637 for deprotonation of the next residue (14). It is conceivable the conformation squeeze at the pathway entrance is presumably relevant to salt bridge or hydrogen-bond formation for proton translocation in the initial step. A working model is proposed for elucidating substrate mediated squeezing at the entrance of the proton pathway in H⁺-PPase. The significance of this initial event of proton transport in H⁺-PPase deserves further elucidations.

FIGURE 5.

Proposed model for the motion of proton transport pathway in CtH⁺-PPase. The working model shows the top view of the entrance of the proton transport pathway and illustrates its squeezing upon substrate binding, elucidated from smFRET analysis. The structure at the entrance of proton transport pathway is transformed from a relatively elliptical shape into a more spherical form. The numbers next to the residues indicate their relative positions in the amino acid sequence of H⁺-PPase. The gray dashed lines represent the interactions between residues and/or ligands. The block dashed circle shows the contour at the entrance of the proton transport pathway.