Essential amino acid residues in the central transmembrane ... - Core

3 downloads 0 Views 2MB Size Report
Four residues in cytoplasmic loop e were essential for substrate hydrolysis and ...... H+-pyrophosphatase: necessity of Cys634 for inhibition by maleimides.
Biochimica et Biophysica Acta 1767 (2007) 930 – 939 www.elsevier.com/locate/bbabio

Essential amino acid residues in the central transmembrane domains and loops for energy coupling of Streptomyces coelicolor A3(2) H + -pyrophosphatase Megumi Hirono, Yoichi Nakanishi, Masayoshi Maeshima ⁎ Laboratory of Cell Dynamics, Graduate School of Bioagricultural Sciences, Nagoya University, Nagoya 464-8601, Japan Received 22 January 2007; received in revised form 20 March 2007; accepted 29 March 2007 Available online 14 April 2007

Abstract The H+-translocating inorganic pyrophosphatase is a proton pump that hydrolyzes inorganic pyrophosphate. It consists of a single polypeptide with 14−17 transmembrane domains, and is found in a range of organisms. We focused on the second quarter region of Streptomyces coelicolor A3(2) H+-pyrophosphatase, which contains long conserved cytoplasmic loops. We prepared a library of 1536 mutants that were assayed for pyrophosphate hydrolysis and proton translocation. Mutant enzymes with low substrate hydrolysis and proton-pump activities were selected and their DNAs sequenced. Of these, 34 were single-residue substitution mutants. We generated 29 site-directed mutant enzymes and assayed their activity. The mutation of 10 residues in the fifth transmembrane domain resulted in low coupling efficiencies, and a mutation of Gly198 showed neither hydrolysis nor pumping activity. Four residues in cytoplasmic loop e were essential for substrate hydrolysis and efficient H+ translocation. Pro189, Asp281, and Val351 in the periplasmic loops were critical for enzyme function. Mutation of Ala357 in periplasmic loop h caused a selective reduction of proton-pump activity. These low-efficiency mutants reflect dysfunction of the energy-conversion and/or proton-translocation activities of H+-pyrophosphatase. Four critical residues were also found in transmembrane domain 6, three in transmembrane domain 7, and five in transmembrane domains 8 and 9. These results suggest that transmembrane domain 5 is involved in enzyme function, and that energy coupling is affected by several residues in the transmembrane domains, as well as in the cytoplasmic and periplasmic loops. H+-pyrophosphatase activity might involve dynamic linkage between the hydrophilic and transmembrane domains. © 2007 Elsevier B.V. All rights reserved. Keywords: Energy coupling; H+-pyrophosphatase; Random mutagenesis; Proton pump

1. Introduction Proton pumps have two physiological roles: pH regulation and the formation of proton-motive forces across biomembranes. They convert the energy of chemical bonds into the active transport of protons. The simplest of these pumps is the H+-translocating inorganic pyrophosphatase (H+-PPase), which consists of a single polypeptide of ∼80 kDa [1]. H+-PPases are

Abbreviations: DTT, dithiothreitol; EGTA, ethylene glycol bis(2-aminoethylether)-N,N,N′,N′-tetra-acetic acid; H+-PPase, H+-translocating inorganic pyrophosphatase; IPTG, isopropyl-1-thio-β-galactopyranoside; LB, Luria Broth; PPi, pyrophosphate; ScPP, Streptomyces coelicolor A3(2) H+-PPase; sScPP, synthetic ScPP gene; TM, transmembrane domain; WT, wild type ⁎ Corresponding author. Tel./fax: +81 52 789 4096. E-mail address: [email protected] (M. Maeshima). 0005-2728/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbabio.2007.03.014

found in plants, parasitic and free-living protozoa, and some eubacteria and archaebacteria [2–7]. In prokaryotes, such as Rhodospirillum rubrum [2,8,9], Pyrobaculum aerophilum [10], and Agrobacterium tumefaciens' [11], the enzyme generates a proton gradient across the plasma membrane and the membranes of acidocalcisomes. Its physiological role has also been investigated in plants [12–16] and other organisms [8,17–19]. The H+-PPase enzyme is an excellent model for research into the coupling between pyrophosphate [PPi] hydrolysis and active H+ transport, because it consists of a single protein and has a simple substrate. Several studies have examined the overall membrane topology of H+-PPases and the functional residues of their catalytic sites. Site-directed mutagenesis revealed several functional motifs in the H+-PPases of Arabidopsis thaliana [4,20], Vigna radiata [21–23], R. rubrum [24,25], Carboxydothermus

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

hydrogenoformans [26], and Streptomyces coelicolor A3(2) [27–29]. Membrane topological analysis showed that all the common functional motifs for Mg-PPi binding and PPi hydrolysis are exposed to the cell cytosol, which is where the substrate is generated [27]. In the present study, we investigated how H+-PPase couples the hydrolysis of PPi with the active transport of protons across the membrane. As a model enzyme we used S. coelicolor A3(2) H+-PPase (ScPP), which comprises 794 amino acids and 17 transmembrane domains (TMs) [27]. ScPP was efficiently expressed in Escherichia coli. To examine the coupling mechanism and the structural–functional relationship, we separated the primary structure into four parts and constructed ScPP mutant libraries using random mutagenesis. We focused on the second quarter, which consists of TM5 to TM9, the primary PPi-binding site, and a few conserved motifs. We prepared a random mutant library of H +-PPase, and surveyed mutants with a low coupling efficiency between PPi hydrolysis and proton pumping. We also determined the effect of mutating residues in the five TMs and hydrophilic loops on enzyme activity and energy coupling. The structural and functional significance of these mutations are discussed in relation to the energy-transducing mechanism.

931

random mutant library (total transformant number = 9.8 × 104). The mutation frequency of the library was determined by DNA sequencing as 0.9 nucleotide changes per clone. The amplified library DNA was purified and introduced into an E. coli expression host BLR(DE3)pLysS K128I [27].

2.3. Site-directed mutagenesis of ScPP Mutant derivatives of ScPP were generated from sScPP using a QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) according to the method of Kirsh and Joly [30] as described previously [21,27]. The mutants were confirmed by DNA sequencing.

2.4. ScPP expression and preparation of E. coli membranes Each ScPP construct was transformed into E. coli strain BLR(DE3)pLysS K128I and transformants were selected on Luria Broth (LB) plates containing 50 μg/ml ampicillin and 34 μg/ml chloramphenicol. Cells derived from a single transformant colony were grown in LB medium supplemented with antibiotics for 16 h at 37 °C, and then diluted 100-fold in SB medium (1.2% tryptone, 2.4% yeast extract, 0.5% glycerol, 0.38% KH2PO4, and 1.25% K2HPO4). After 3.5 h at 37 °C, isopropyl-1-thio-β-galactopyranoside (IPTG) was added at a final concentration of 0.5 mM. Cells were incubated for a further 4 h, harvested by centrifugation at 3500×g for 10 min, and suspended in a tenfold volume of 50 mM Tris–Mes (pH 7.3), 1 mM ethylene glycol bis(2-aminoethylether)-N,N,N′,N′-tetra-acetic acid (EGTA), 2 mM dithiothreitol (DTT), 75 mM KCl, 0.15 M sucrose, and 1 mM phenylmethanesulfonyl fluoride. Then, 40 μg/ml DNase I and 0.2 μg/ml lysozyme were added, and the suspension was incubated at 4 °C for 10 min. The cells were

2. Materials and methods 2.1. Cloning of the ScPP gene and expression in E. coli A laboratory strain of S. coelicolor A3(2) was cultivated on agar plates containing 0.1% yeast extract, 0.1% beef extract, 0.2% NZ-amine, 1% glucose, and 2% agar at 30 °C. DNA was isolated from the cells by standard procedures, and was used as a template for PCR amplification of the DNA sequence of the H+-PPase gene. As the native H+-PPase gene of S. coelicolor A3(2) has repeat sequences in its open reading frame and a high GC content it is unsuitable for gene manipulation by the polymerase chain reaction (PCR). We therefore designed and constructed a synthetic ScPP gene (sScPP) as described previously [27], in which the amino-acid sequence encoded by the gene remains unchanged. Plasmid pYN309 was derived from pET23b (Novagen, Madison, WI) by modifying the PstI site, and was used to express the ScPP proteins. A hexahistidine (His6) tag was added to the carboxy-terminus to enable detection of the ScPP protein by immunoblotting. ScPP constructs in pYN309 were introduced into E. coli strain BLR (DE3) pLysS (Novagen), and transformants were selected with 50 μg/ml ampicillin and 34 μg/ml chloramphenicol.

2.2. Random mutagenesis of the second quarter of sScPP sScPP was inserted into the NdeI and XhoI sites of pYN309 to obtain the expression plasmid pYN316. An sScPP variant, in which the Ser794 codon (AGC) is replaced by GAG, was inserted into the Eco RV site of pZErO2.1 (Invitrogen, Carlsbad, CA) to obtain the mother plasmid, pYN025, used for random mutagenesis. A random mutant library of sScPP was constructed using PCR-based random mutagenesis (error-prone PCR) according to the method described previously [21] with a few modifications. Briefly, the second quarter region of ScPP (amino acids Val183 to Gln383) was subjected to random mutagenesis. PCR was carried out in a 50-μl solution containing 10 ng pYN025 plasmid, 10 mM Tris–HCl (pH 8.3), 50 mM KCl, 0.001% gelatin, 2.5 mM MgCl2, 0.1 mM MnCl2, 0.1 mM dATP, 0.1 mM dCTP, 1 mM dGTP, 1 mM dTTP, 2.5 U AmpliTaq DNA polymerase (Perkin Elmer, Wellesley, MA), and 250 nM primers (forward primer, 5′-GGCATGTTTACAGTGGGCTTAGGAC3′; reverse primer, 5′-CGGACCGGTGAGACTTGATTTGC-3′). The PCR conditions were 18 cycles of 30 s at 95 °C, 60 s at 55 °C, and 3 min at 72 °C. The resulting PCR fragments were digested and inserted into the SacII–MunI site of pYN316. The plasmids generated were amplified in E. coli DH5α and used as a

Table 1 Summary of Random Mutants of S. coelicolor A3(2) H+-PPases Mutant

Mutation

Result

Mutant

Mutation

Result

P189L

566C → T 1014C → T 569A → G 654T → C 774A → G 572T → C 927T → C 583T → C 598G → A 605T → C 648T → C 873T → C 614T → C 653A → G 655A → G 668T → C 686A → G 698A → G 774A → G 725T → C 731A → G 736G → A 753A → G 737T → C 743A → G 1110T → C 755A → G 807A → G 990A → G

P→L silent K→R silent silent V→A silent F→L A→T I→T silent silent F→S D→G D→G L→P Q→R E→G silent I→T D→G V→I silent V→A D→G Silent D→G silent silent

M256T

601T → C 767T → C 776A → G 601T → C 1089T → C 781T → C 753A → G 756T → C 833C → A 842A → G 877A → G 988T → C 880G → A 753A → G 958T → C 585C → T 964T → C 597T → C 975C → G 983T → C 708A → G 995T → C 1052T → C 1125T → C 1070C → G 1089T → C 1130T → C 1137A → G 1145A → G

silent M→T D→G silent silent F→L silent silent A→V D→G I→V silent G→R silent R→C silent F→L silent S→R I→T silent L→P V→A silent A→G silent L→P silent Q→R

K190R

V191A F195L A200T I202T

F205S D218G D222G L223P Q229R E233G I242T D244G V246I V246A D248G D252G

D259G F261L

A278V

D281G I293V G294R R320C F322L S325R I328T L332P V351A A357G L377P Q382R

Random mutagenesis generated 1536 mutants of ScPP. In total, 34 mutants with low energy-coupling efficiency (no or low PPase activity) were selected from the random mutants (see text for details). DNAs for the mutants were sequenced individually. Mutation sites including silent mutations are listed.

932

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

thoroughly homogenized using a bath-type sonicator (Cosmobio, Tokyo, Japan), and centrifuged at 2000×g for 5 min. The supernatant was further centrifuged at 150,000×g for 30 min. The resulting precipitate was suspended in 10 mM Tris– Mes (pH 7.3), 1 mM DTT, 1 mM MgCl2, 100 mM KCl, 0.15 M sucrose, and 0.5 mM phenylmethanesulfonyl fluoride.

2.5. Protein and enzyme assays The protein content was quantitated using the Bradford method [31] with the Bio-Rad assay reagent (Bio-Rad, Hercules, CA). PPi hydrolysis was measured at 30 °C as described previously [5] in a modified reaction mixture

Fig. 1. Expression levels and enzyme activities of random mutants of ScPP. (A) Membrane vesicles were prepared from E. coli cells expressing WT ScPP and mutants with single-residue replacements. Membranes (5 μg per lane) were subjected to immunoblotting with anti-His6 antibodies (upper panel). The intensities of the immunostained bands were quantified and are expressed relative to that of WT ScPP (lower panel). (B) Crude membranes were prepared from E. coli cells expressing ScPP mutants and assayed for PPi hydrolysis and PPi-dependent H+-pump activity. PPi hydrolysis by the mutants is expressed relative to that of the WT (170 nmol PPi/ min/mg protein). (C) H+-pump activities of WT and mutants in E. coli membrane vesicles were measured by fluorescence quenching of acridine orange, and are expressed relative to that of the WT enzyme (60%ΔF/min/mg protein). (D) Coupling efficiency was calculated as the ratio of H+-pump activity to PPi hydrolysis activity, and is expressed relative to that of the WT enzyme. Mutant ScPPs are categorized as “no activity” (black circles), “pump-less” (open circles), and “loosecoupling” (grey circles; see text for details).

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

933

(5 mM Bicine–NaOH [pH 8.0], 100 mM KCl, 1 mM MgCl2, 0.15 M sucrose, 0.32 mM Na4PPi, 1 mM sodium molybdate, and 0.5 mM NaF). Actual substrate for H+-PPase is the Mg2PPi complex [32] and the concentration of Mg2PPi under the assay condition was calculated to be 71 μM. PPi-dependent H+-transport activity was measured as the initial rate of fluorescence quenching of acridine orange at 25 °C with a Shimazu (Kyoto, Japan) RF-5000 fluorescence spectrophotometer set at 493 nm for excitation and 540 nm for emission as described previously [5,33,34] in a modified reaction medium (5 mM BicineNaOH [pH 8.0], 100 mM KCl, 1 mM MgCl2, 0.15 M sucrose, and 1 μM acridine orange). The enzyme reaction was initiated by adding 0.4 mM Na4PPi.

2.6. Immunoblotting Proteins were separated by 12% SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Millipore, Billerica, MA) with a semidry blotting apparatus (Bio-Rad) using standard procedures. Immunoblotting was carried out using horseradish peroxidase-linked protein A and ECL Western blotting detection reagents (GE Healthcare Bio-sciences, Piscataway, NJ). Polyclonal antibodies against the hexa-histidine tag (anti-His6 antibodies; MBL Co., Nagoya, Japan) were used to quantitate ScPP mutant proteins in the E. coli membrane fractions.

Fig. 2. Effect of substrate concentration on the activity of mutant enzymes. Crude membranes prepared from E. coli expressing WT and mutant ScPPs were assayed for PPase activity at different concentrations of PPi. The concentration of MgCl2 was fixed at 1.0 mM. Mutation sites are in the TMs (I293Vand F322L) and cytoplasmic loop e (Q229R and M256T).

3. Results 3.1. Screening of the ScPP random mutant library We examined the effect of replacing amino acids in the second quarter of ScPP, which comprises 201 residues from TM4 to TM9. The accumulation of ScPP protein and enzyme activity was assessed in 1536 ScPP mutants. Membrane fractions prepared from all mutants were individually assayed for PPi hydrolysis (PPase) and PPi-dependent H + -pump activities. Most mutants had full PPase activity, although some had low or negligible activity. We selected mutants that lacked PPi hydrolysis activity or had a decreased energy-coupling efficiency (that is, the relative ratio of H+-pump activity to PPase activity) for functional analysis. The latter included mutants with normal PPase activity but decreased H+-pump activity, decreased PPase and H+-pump activities, and decreased PPase activity and no H+-pump activity. DNA sequencing of 102 mutants revealed that many had mutations at multiple sites (data not shown) and so were unsuitable for functional analysis. We therefore selected 34 single mutants for further analysis, some of which contained silent mutations (Table 1). 3.2. Mutant ScPP expression and enzyme activity The protein levels of mutant ScPPs were determined by immunoblotting with an anti-His6 antibody. Addition of the C terminus His6 tag to wild-type (WT) ScPP has previously been shown to have no effect on PPase and H+-pump activities [27]. All of the mutants, except for G294R, accumulated in E. coli membranes at a similar level to the WT enzyme (Fig. 1A). Bands of 80 kDa were detected for all of the mutant and WT enzymes, suggesting that no proteolytic cleavage or covalent modification had occurred in the mutants (data not shown). We next examined the substrate affinity of these mutants, in order to determine the effect of the mutations on maximal enzyme activity. PPi hydrolysis (Fig. 2) revealed samples of

mutants: those, such as M256T, which had increased activity; Q229R and others in its group, which had a mutation site located in cytoplasmic loop e; and those, such as F322L and I293V, which had mutation sites in the TM domains. Maximal activities of the WT and mutant ScPPs were recorded at a PPi concentration of 320 μM in the presence of 1 mM MgCl2. Therefore, we concluded that a PPi concentration of 320 μM was adequate for measuring the PPase activity of the mutants and WT like the purified ScPP from S. coelicolor A3(2) [35]. Fig. 1B and C shows the activities of the PPase and H+ pump of the mutants, respectively. We also calculated the ratio of H+ -pump activity to PPase activity, in order to evaluate the energy-coupling efficiency of the mutants, and expressed it as a percentage of the value for the WTenzyme (Fig. 1D). The mutants could be divided into three groups. The first group comprised 11 mutants (D218G, D222G, L223P, D244G, D248G, D252G, D259G, D281G, G294R, S325R, and L377P) that completely lacked PPase activity and had no H+-pump activity as assayed; we designated mutants with activity b 10% of that of the WT as “no activity” mutants. The second group comprised five mutants (P189L, F195L, F261L, L332P, and V351A) with 10–25% of the PPase activity of the WT enzyme, and no H+-pump activity; these were designated “pump-less” mutants. The third group comprised the remaining 18 mutants, which had relatively high PPase activity but H+-pump activity that did not parallel their PPase activity; the coupling efficiency of these mutants was 17–60% of that of the WT enzyme and they were designated as “loosecoupling” mutants. The three groups are separately marked in Fig. 1. 3.3. Analysis of site-directed ScPP mutants The position of each mutation site of the random mutants is mapped in the membrane topology model shown (see Fig. 5 for topological positions). The sites of 16 of the 34 single mutants were distributed throughout TM5 to TM9, and especially in TM5 and TM8. The other mutation sites were located in the

934

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

hydrophilic loops of ScPP (see Fig. 5). Several residues in the TMs of the second quarter of the ScPP sequence are conserved in the H+-PPases of various organisms (Fig. 3). Random mutagenesis revealed that the three highly conserved residues in this region, Phe205, Phe261, and Ser325, are essential for function. Our findings revealed that mutations at the other 13 non-conserved residues affect the coupling ratio. Thus, screening of random mutants is useful to identify unexpected residues that participate in enzyme functions. We further examined the effects of amino-acid replacements at single sites in the region from TM5 to TM9. We selected 16 sites for site-directed mutagenesis, which included charged residues in the TMs (Glu193, Arg207, and Glu262), conserved residues, especially glycine and serine (Gly194, Gly198, Gly210, Gly211, Ser263, and Gly374), and two serine residues in cytoplasmic loop g.

In general, glutamic-acid residues are thought to play essential roles in proton transport across biomembranes, and glycine is reportedly involved in the tight packing of the TM helices of membrane proteins [36]. Residues Leu368 and Ile381 were excluded from the random mutagenesis analysis, as the double mutants I293V/L368P and I381V/L332P lacked enzyme activity. In total, 13 of the residues selected for site-directed mutagenesis were in the TMs (see Fig. 5 for topological model), and three (Pro307, Ser310 and Ser313) were in loop g. In total, 16 residues were replaced individually by alanine or a related amino acid. We finally produced 29 site-directed mutant ScPPs and expressed them in E. coli. The mutant ScPP proteins were expressed in E. coli, as detected by immunoblotting (Fig. 4A). Although mutant G374A expressed a relatively low level of protein, it was sufficient for the enzyme assay.

Fig. 3. Multiple alignment of the second quarter (a region from Val183 to Gln383) of ScPP with H+-PPases from various organisms. The accession numbers of sequences are: Vigna radiata (AB009077, BAA23649), Cucurbita moschata (BAA33149), Nicotiana tabacum (Q43798), Arabidopsis thaliana (P31414), Beta vulgaris (AAA61609), Oryza sativa (BAA08232), Vitis vinifera (AAF69010), Chara corallina (AB018592, BAA36841), Chlamydomonas reinhardtii (CAC44451), Rhodospirillum rubrum (AAC38615), A. tumefaciens (AAK86977), and Streptomyces coelicolor A3(2) (Q9X913). Identical residues are boxed and shaded. Highly conserved residues (identity with more than 75%) are boxed.

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

Fig. 4B and C shows the effects of the amino-acid substitutions on PPase and H+-pump activities, respectively. In total, eight mutants (G198A, R207A, E262A, E262D, S263E, L368P, G374A, and I381P) lacked PPase activity, and the PPase activity of six (E193D, G194A, G210A, G211A, E262Q, and S263A) was reduced to b 25% of that of the WT. These results suggest that at least eight residues (Gly194 , Gly198, Arg207, Gly210, Gly211, Glu262 , Ser263 , and Gly374 ) are strictly required for PPi hydrolysis, even though they are located in the TMs. By contrast, the substitution of Pro291, Ser310, and Ser313

935

did not markedly affect enzyme activity. The effect of a substitution depends on the type of amino-acid residue substituted. This is relevant for Glu193 (alanine or aspartic acid), Arg207 (alanine or lysine), Leu368 (proline or alanine), and Ile381 (proline, alanine, or valine), and is discussed further later. The energy-coupling efficiencies of the ScPP mutants were calculated, and are shown in Fig. 4D. In total, 21 of the sitedirect mutants could be divided into three groups: “no activity”, “pump-less”, and “loose-coupling” mutants, as described for the random mutants. A further six mutants, with PPase and H+ -pump activities that were not affected, were designated as “no effect” mutants. The remaining two mutants (E193A and I381V) had low PPase and H+ -pump activities, although their coupling efficiencies were normal. These were categorized as “low activity”. The results of the random and site-directed mutagenesis are summarized in Fig. 5, and the mutation sites are mapped on the helical wheels of the TMs shown in Fig. 6. 4. Discussion Elucidating the proton-transport pathway of H+-PPase, and the coupling mechanism between PPi hydrolysis and H+ active transport, led us to perform random mutagenesis of the H+-PPase of S. coelicolor A3(2). Site-directed mutagenesis has been previously used to determine the essential residues for the binding and hydrolysis of the substrate, and has provided critical information on the structure–function relationship of the cytoplasmic catalytic domain [4,20,22–27,37,38]. Site-directed mutagenesis is effective for evaluating specific residues, such as conserved charged residues, which might be involved in PPi hydrolysis. The present study revealed several unexpected residues that were involved in H+-active transport and energy transfer. These functional residues were found in the cytosolic loops, periplasmic (intravacuolar) loops, and TMs. We discuss them separately below.

Fig. 4. Expression levels and enzyme activities of site-directed mutants of ScPP. (A) Membrane vesicles were prepared from E. coli cells expressing WT ScPP and site-directed mutant ScPPs as indicated. Membranes (5 μg) were subjected to immunoblotting with anti-His6 antibodies (upper panel). The intensities of immunostained bands were quantified, and are expressed relative to that of the WT enzyme (lower panel). In many mutants, the endogenous residue was replaced by two or three residues as indicated. (B) PPase activities of WT and mutant ScPPs in E. coli membranes are expressed relative to that of the WT (155 nmol PPi/min/mg). (C) H+-pump activities of the WT and mutants in E. coli membrane vesicles are expressed relative to that of the WT (58%ΔF/min/ mg). (D) Coupling efficiencies were calculated as the ratios of H+-pump activities to PPi hydrolysis activities, and are expressed relative to that of the WT enzyme. Mutant ScPPs are categorized as “no activity” (black circles), “pumpless” (open circles), and “loose-coupling” mutants (grey circles) as for the random mutants shown in Fig. 2. Mutants with coupling efficiencies of b70% are categorized as “loose-coupling” mutants. Open squares indicate mutants with activities that were not affected by the amino-acid replacements. These are designated “no effect” mutants (open squares). The PPi hydrolysis and H+-pump activities of I381V were 75% of those of the control enzyme (“low activity”, open diamond).

936

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

Fig. 5. The topological arrangement of mutation sites and effects of the mutations on enzyme activities. The membrane topology model of ScPP is based on a previous report [27]. Mutant ScPPs are categorized as “no activity” (black circle), “pump-less” (open circles), “loose-coupling” (light green circles), “low activity” (open diamond), and “no effect” mutants (black diamonds). Highly conserved residues are shown in red, and the four functional motifs in loop e are boxed. The numbers beside the residues show the positions of mutation sites. Residues of the site-directed mutants are underlined in red (see text for the details).

4.1. Residues essential for energy conversion in the cytosolic loops The results of functional analysis on 63 mutants in 49 sites are summarized in Fig. 5. The loop e has four characteristic sequences conserved in H+ -PPases: GGGIFTK, ADVGADLVGKVE, EDDPRN, and IADNVGDNVGDCA [3,4,39] (Fig. 3). These motifs have been shown to form the catalytic site of H+-PPases [21,24]. Four conserved residues in loop e (Asp218 , Asp244, Asp248, and Asp252) were confirmed to be essential for PPi hydrolysis [21]. The present random mutagenesis study showed that Asp222 , Leu223 , and Asp259 were also essential for PPase activity (Fig. 2). We consider Asp259 to be the most likely candidate for the entrance residue for H+ in the cytoplasmic face, because the residue is highly conserved acidic residue at the interface between cytoplasmic loop and TM. By contrast, replacement of Ser310 and Ser313 by alanine, proline, or glutamic acid had no effect on enzyme activity. This suggests

that these serine residues are not part of the catalytic site. An interesting point is that two residues in cytoplasmic loops e (Val246 ) and g (Pro307) are critical for substrate hydrolysis and energy coupling. This is reasonable, as Val246 is part of the conserved motif, and substitution of P307A might alter the length of TM7, and the secondary structure of the joint region between TM7 and the loop g. It should be noted that at least four mutations (Q229R, E233G, I242T, and M256T) are critical for H+-pump activity but not for PPase activity (Fig. 1). These four loose-coupling mutants retained N 70% of PPi-hydrolysis activity and only 30% of proton-pumping activity. The PPi-hydrolysis reaction includes three steps: binding of the substrate PPi–Mg complex, hydrolysis of PPi, and release of Pi. This process must be linked to active proton transport, which consists of binding of H+ on the cytoplasmic side, translocation of H+, and release of H+ on the periplasmic or vacuolar lumen side. In these mutant ScPPs, the energy-transfer process from PPi hydrolysis might be impaired. Conformational changes around the catalytic site

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

937

Fig. 6. Helical wheel models of the TM5 to TM9 segments. Amino-acid residues located in TMs are plotted on helical wheels. Conserved residues are shown in red. Glycine residues are marked in light blue. Aromatic residues are marked in orange.

might be involved in the energy transfer associated with proton translocation in the TMs. A recent study revealed that Cys253 in loop e and Cys621 in loop m of ScPP forms a disulfide fond and are involved in redox regulation of the activity [29], suggesting the structural dynamics of the catalytic site. The present findings suggest that amino-acid substitution in loop e disturbs the interdomain interaction. This effect includes distorted tertiary structure. Previously, loose coupling of the E263D of mung bean H+-PPase, in which Glu263 corresponds to Glu228 in loop e of ScPP, was demonstrated by the patch-clamp technique [38]. Loop e is folded extensively, and is involved in both PPi hydrolysis and the conformational change associated with energy transfer to the TM region. 4.2. Residues essential for energy conversion in the periplasmic loops Random mutagenesis disclosed that five residues (Pro189, Lys190, Asp281, Val351, and Ala357) in the periplasmic loops, which correspond to the intra-vacuolar loops of plant H+-PPases, affect enzyme activity (Figs. 1 and 5). P189L, D281G, and V351A lacked both PPi hydrolysis and proton-pump activities. A357G retained 50% of PPi hydrolysis and only 18% of protonpump activity. Therefore, we conclude that Pro189, Asp281, Val351, and Ala357 are essential for energy conversion and/or proton translocation at a minimum. It is hard to imagine direct effects of residues in the periplasm loops on PPi hydrolysis.

In these mutant enzymes, impairment of the proton pump affects the PPi hydrolysis reaction on the cytoplasmic side because the both reactions are coupled. D281G had the most severe effect on enzyme function. It is possible that Asp281 forms an electrostatic interaction with a basic residue, such as Lys190. It has been recently reported that a salt bridge is formed between Glu427 and Lys461 to link two of the TMs of Arabidopsis H+-PPase [40]. Another possible function is that Glu281 is essential for stepwise proton translocation around the exit in the periplasm. 4.3. Functional residues in the TMs Amino-acid substitutions in the TMs impaired proton translocation and energy coupling. Six residues (Gly198 , Glu262, Gly294, Ser325, Gly374 , and Leu377 ) proved to be essential for PPi hydrolysis and proton-pump activity (Figs. 1 and 4). Glu262 corresponds to Glu231 of the R. rubrum enzyme, which is reportedly essential for enzyme function [25]. Glu262 is the only acidic residue on the cytoplasmic side of the TMs. Therefore, we consider Glu262 to be involved in mediation of H+ translocation in the TMs. The present study of five TMs revealed that the three glycine residues at positions 198, 294, and 374 are essential for enzyme activity. These glycines occupy equivalent positions in TM5, TM7, and TM9 (Fig. 5). Glycine is thought to be favorable for close packing between transmembrane helices [36,41]. Pairs of GXXXG sequences permit close packing of transmembrane

938

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

helices because hydrogen bonds are formed between backbone carbonyls and the Cα hydrogens of the glycine residues [36]. ScPP has a total of four GXXXG sequences in TM5 (GFGLG), TM10 (GISLG), TM13 (GLIAG), and TM14 (GFTLG). At least the two glycines of the GFGLG sequence in TM5 are essential for efficient coupling. The GFGLG sequence might therefore interact with one of the three other GXXXG sequences to permit close helix–helix packing. The first glycine of the conserved triple glycine motif in TM5 is reportedly essential for enzyme function [25]. We replaced the second and third glycine residues (Gly210 and Gly211) with alanine, and found that they were also essential for enzyme function. The smallest residue, glycine, makes polypeptide chains flexible; this might be essential for movement of the cytoplasmic catalytic loop e. Recently, a Gly229–Tyr230 –Gly231 motif found in mung bean + H -PPase was reported to be shared by most H+-PPases, and to be essential for enzyme function [23]. ScPP has a similar Gly194–Phe195 –Gly196 sequence, in which tyrosine is replaced with a similar residue, phenylalanine. The Gly194 and Phe195 residues are essential for proton pumping and efficient coupling (Figs. 2 and 4). We conclude that GY(F)G is a conserved motif in many H+-PPases, ranging from eubacteria to higher plants, and is essential for H+ translocation. In addition to the Phe195 in the GY(F)G motif, there are nine phenylalanine and two tyrosine residues in TM5 through to TM8 (Fig. 6). Phenylalanine and tyrosine tend to interact with other aromatic residues and might strengthen helix–helix interactions between the TMs. Interestingly most phenylalanine residues are located on the cytosolic or periplasmic side, not in the central region (Fig. 5). This allows ring–ring interactions between aromatic residues. We found that TM5 had 11 essential or critical residues (Fig. 5). TM5 is an anchoring component of functional loop e, and is also involved in proton translocation and/or energy coupling. The essential and conserved residues are clustered on the two sides of the helical wheel (Fig. 6, sides a and b), suggesting that TM5 interacts with the other membrane helices via sides a and b. ScPP has a single acidic residue Glu193 in TM5, and most other H+-PPases have aspartic acid (plants) [4,21,42] or glutamic acid (microorganisms such as R. rubrum) [3] at this position. Therefore, Glu193 might be an acidic residue essential for H+ translocation in the TMs, similar to Glu262 in TM6 (Fig. 5). TM5 might thus be the central pillar of H+-PPase in the H+-translocation pathway. A positive charged residue of Arg207 in TM5 seems to be essential, because a substitution of the residue with alanine (R207A) caused complete loss of the enzyme activity and R207K kept a partial activity (Fig. 4). This residue may be involved in electrostatic interaction with a negative charged residue in the other transmembrane helix or formation of hydrophobic space in the cytoplasmic portal. In conclusion, we identified several residues that were essential for efficient H+ translocation and energy coupling not only in the TMs but also in the hydrophilic loops. Furthermore, the present study revealed that substitution of the residues in the periplasmic loops affect on the substrate hydrolysis took place

in the cytoplasmic loops. Dysfunction of H+-PPase caused by amino acid substitution includes direct and indirect effects. Lost of the proton pump activity by substitution of Asp259 and Glu262 may be direct effect. Substitution of glycine residues in the TMs may cause indirect effect through disorder of the helix–helix packing. Further detailed analysis of the remaining parts of the H+-PPase by random and site-directed mutagenesis will be needed to identify key residues and distinguish the direct and indirect role of each residue. Acknowledgements This work was supported by Grants-in-Aid for Scientific Research 18380064, 16085204, and 14COE02 (to M.M.) from the Ministry of Education, Sports, Culture, Science and Technology of Japan, the Program for Promotion of Basic Research Activities for Innovative Biosciences (PROBRAIN) and the Matsushita Electric Industrial Co. M.H. is a Japan Society for the Promotion of Science (JSPS) Research Fellow (16-5967). References [1] M. Maeshima, Vacuolar H+-pyrophosphatase, Biochim. Biophys. Acta 1465 (2000) 37–51. [2] M. Baltscheffsky, S. Nadanaciva, A. Schultz, A pyrophosphate synthase gene: molecular cloning and sequencing of the cDNA encoding the inorganic pyrophosphate synthase from Rhodospirillum rubrum, Biochim. Biophys. Acta 1364 (1998) 301–306. [3] M. Baltscheffsky, A. Schultz, H. Baltscheffsky, H+-PPases: a tight membrane-bound family, FEBS Lett. 457 (1999) 527–533. [4] Y.M. Drozdowicz, A. Rea, Vacuolar H+-pyrophosphatases: from the evolutionary backwaters into the mainstream, Trends Plant Sci. 6 (2001) 206–211. [5] M. Maeshima, S. Yoshida, Purification and properties of vacuolar membrane proton-translocating inorganic pyrophosphatase from mung bean, J. Biol. Chem. 264 (1989) 20068–20073. [6] Y. Nakanishi, N. Matsuda, K. Aizawa, T. Kashiyama, K. Yamamoto, T. Mimura, M. Ikeda, M. Maeshima, Molecular cloning of the cDNA for vacuolar H+-pyrophosphatase from Chara corallina, Biochim. Biophys. Acta 1418 (1999) 245–250. [7] D.A. Scott, W. de Souza, M. Benchimol, L. Zhong, G.G. Lu, S.N.J. Moreno, R. Docampo, Presence of a plant-like proton-pumping pyrophosphatase in acidocalcisomes of Trypanosoma cruzi, J. Biol. Chem. 273 (1998) 22151–22158. [8] C.R. Garcia, H. Celis, I. Romero, Importance of Rhodospirillum rubrum H+-pyrophosphatase under low-energy conditions, J. Bacteriol. 186 (2004) 6651–6655. [9] M. Seufferheld, C.R. Lea, M. Vieira, E. Oldfield, R. Docampo, The H+-pyrophosphatase of Rhodospirillum rubrum is predominantly located in polyphosphate-rich acidocalcisomes, J. Biol. Chem. 279 (2004) 51193–51202. [10] Y.M. Drozdowicz, Y.P. Lu, V. Patel, S. Fitz-Gibbon, J.H. Miller, P.A. Rea, A thermostable vacuolar-type membrane pyrophosphatase from the archaeon Pyrobaculum aerophilum: implications for the origins of pyrophosphate-energized pumps, FEBS Lett. 460 (1999) 505–512. [11] M. Seufferheld, M.C.F. Viera, F.A. Ruiz, C.O. Rodrigues, S.N.J. Moreno, R. Docampo, Identification of organelles in bacteria similar to acidocalcisomes of unicellular eukaryotes, J. Biol. Chem. 278 (2003) 29971–29978. [12] R.A. Gaxiola, J. Li, S. Undurraga, L.M. Dang, G.J. Allen, S.L. Alper, G.R. Fink, Drought- and salt-tolerant plants result from overexpression of the AVP1 H+-pump, Proc. Natl. Acad. Sci. U. S. A. 98 (2001) 11444–11449. [13] J. Li, H. Yang, W.A. Peer, G. Rchter, J. Blakeslee, A. Bandyopadhyay, B. Titapiwantakun, S. Undurraga, M. Khodakovskaya, E.L. Richards, B.

M. Hirono et al. / Biochimica et Biophysica Acta 1767 (2007) 930–939

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

Krizek, A.S. Murphy, S. Gilroy, R. Gaxiola, Arabidopsis H+-PPase AVP1 regulates auxin-mediated organ development, Science 310 (2005) 121–125. Y. Nakanishi, M. Maeshima, Molecular cloning of vacuolar H+-pyrophosphatase and its developmental expression in growing hypocotyls of mung bean, Plant Physiol. 116 (1998) 589–597. C. Bremberger, U. Lüttge, Dynamics of tonoplast poroton pumps and other tonoplast proteins of Mesembryanthemum crystallinum L. during the induction of crassulacean acid metabolism, Planta 188 (1992) 575–580. S. Park, J. Li, J.K. Pittman, G.A. Berkowitz, H. Yang, S. Undurraga, J. Morris, K.D. Hirschi, R.A. Gaxiola, Up-regulation of a H+-pyrophosphatase (H+-PPase) as a strategy to engineer drought-resistant crop plants, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 18830–18835. R.L. López-Marqués, J.R. Pérez-Castiñeira, M. Losada, A. Serrano, Differential regulation of soluble and membrane-bound inorganic pyrophosphatases in the photosynthetic bacterium Rhodospirillum rubrum provides insights into pyrophosphate-based stress bioenergetics, J. Bacteriol. 186 (2004) 5418–5426. K.J. Saliba, R.J.W. Allen, S. Zissis, P.G. Bray, S.A. Ward, K. Kirk, Acidification of the malaria parasite's digestive vacuole by a H+-ATPase and a H+-pyrophosphatase, J. Biol. Chem. 278 (2003) 5605–5612. G. Lemercier, S. Dutoya, S. Luo, F.A. Ruiz, C.O. Rodrigues, T. Baltz, R. Docampo, N. Bakalara, A vacuolar-type H+-pyrophosphatase governs maintenance of functional acidocalcisomes and growth of the insect and mammalian forms of Trypanosoma brucei, J. Biol. Chem. 277 (2002) 37369–37376. E.J. Kim, R.G. Zhen, P.A. Rea, Site-directed mutagenesis of vacuolar H+-pyrophosphatase: necessity of Cys634 for inhibition by maleimides but not catalysis, J. Biol. Chem. 270 (1995) 2630–2635. Y. Nakanishi, T. Saijo, Y. Wada, M. Maeshima, Mutagenesis analysis of functional residues in putative substrate-binding site and acidic domains of vacuolar H+-pyrophosphatase, J. Biol. Chem. 276 (2001) 7654–7660. Y.Y. Hsiao, R.C. Van, S.H. Hung, H.H. Lin, R.L. Pan, Roles of histidine residues in plant vacuolar H+-pyrophosphatase, Biochim. Biophys. Acta 1608 (2004) 190–199. R.C. Van, Y.J. Pan, S.H. Hsu, Y.T. Huang, Y.Y. Hsiao, R.L. Pan, Role of transmembrane segment 5 of the plant vacuolar H+-pyrophosphatase, Biochim. Biophys. Acta 1709 (2005) 84–94. A.M. Malinen, G.A. Belogurov, M. Salminen, A.A. Baykov, R. Lahti, Elucidating the role of conserved glutamates in H+-pyrophosphatase of Rhodospirillum rubrum, J. Biol. Chem. 279 (2004) 26811–26816. A. Schultz, M. Baltscheffsky, Properties of mutated Rhodospirillum rubrum H+-pyrophosphatase expressed in Escherichia coli, Biochim. Biophys. Acta 1607 (2003) 141–151. G.A. Belogurov, R. Lahti, A lysine substitute for K+: A460K mutation eliminates K+ dependence in H+-pyrophosphatase of Carboxydothermus hydrogenoformans, J. Biol. Chem. 277 (2002) 49651–49654.

939

[27] H. Mimura, Y. Nakanishi, M. Hirono, M. Maeshima, Membrane topology of the H+-pyrophosphatase of Streptomyces coelicolor determined by cysteine-scanning mutagenesis, J. Biol. Chem. 279 (2004) 35106–35112. [28] H. Mimura, Y. Nakanishi, M. Maeshima, Oligomerization of the H+-pyrophosphatase and its structural and functional consequences, Biochim. Biophys. Acta 1708 (2005) 393–403. [29] H. Mimura, Y. Nakanishi, M. Maeshima, Disulfide bond formation in the H+-pyrophosphatase of Streptomyces coelicolor and its implication in redox control and structure, FEBS Lett. 579 (2005) 3625–3631. [30] R.D. Kirsch, E. Joly, An improved PCR-mutagenesis strategy for two-site mutagenesis or sequence swapping between related genes, Nucleic Acids Res. 26 (1998) 1848–1850. [31] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizaing the principle of protein-dye binding, Anal. Biochem. 72 (1976) 248–254. [32] A.A. Baykov, N.P. Bakuleva, P.A. Rea, Steady-state kinetics of substrate hydrolysis by vacuolar H+-pyrophosphatase: a simple three-state model, Eur. J. Biochem. 217 (1993) 755–762. [33] A.B. Bennett, R.M. Spanswick, Optical measurements of ΔpH and Δψ in corn root membrane vesicles: kinetic analysis of Cl− effects on a proton– translocating ATPase, J. Membr. Biol. 71 (1983) 95–107. [34] E. Blumwald, P.A. Rea, R.J. Poole, Preparation of tonoplast vesicles: applications to H+-coupled secondary transport in plant vacuoles, Methods Enzymol. 148 (1987) 115–123. [35] M. Hirono, H. Mimura, Y. Nakanishi, M. Maeshima, Enzymatic and molecular properties of H+-pyrophosphatase of Streptomyces coelicolor expressed in Escherichia coli, J. Biochem. 138 (2005) 183–191. [36] J.U. Bowie, Solving the membrane protein folding problem, Nature 438 (2005) 581–589. [37] G.A. Belgurov, M.V. Turkina, A. Penttinen, S. Huopalahti, A.A. Baykov, R. Lahti, H+-pyrophosphatase of Rhodospirillum rubrum: high yield expression in Escherichia coli and identification of the Cys residues responsible for inactivation by mersalyl, J. Biol. Chem. 277 (2002) 22209–22214. [38] Y. Nakanishi, I. Yabe, M. Maeshima, Patch clamp analysis of H+ pump expressed in giant yeast vacuoles, J. Biochem. 134 (2003) 615–623. [39] M. Maeshima, Tonoplast transporters: organization and function, Annu. Rev. Plant Physiol. Plant Mol. Biol. 52 (2001) 469–497. [40] M. Zancani, L.A. Skiera, D. Sanders, Roles of basic residues and salt-bridge interaction in a vacuolar H+-pumping pyrophosphatase (AVP1) from Arabidopsis thaliana, Biochim. Biophys. Acta 1768 (2007) 311–316. [41] K. Liu, D. Kozono, Y. Kato, P. Agre, A. Hazama, M. Yasui, Conversion of aquaporin 6 from an anion channel to a water-selective channel by a single amino acid substitution, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 2192–2197. [42] M. Ikeda, E. Tanabe, M.H. Rahman, H. Kadowaki, C. Moritani, R. Akagi, Y. Tanaka, M. Maeshima, Y. Watanabe, Avacuolar inorganic H+-pyrophosphatase in Acetabularia acetabulum: partial purification, characterization and molecular cloning, J. Exp. Bot. 50 (1999) 139–140.