Characterization of the Escherichia coli SecA Signal Peptide-Binding ...

5 downloads 11050 Views 1MB Size Report
Sep 8, 2011 - tion of the SecA signal peptide-binding site to design and isolate signal ... make up the cytoplasmic membrane, periplasm, or outer mem- brane. A majority ... helix finger region of the helical scaffold domain of SecA has been.
Characterization of the Escherichia coli SecA Signal Peptide-Binding Site Lorry M. Grady, Jennifer Michtavy, and Donald B. Oliver Department of Molecular Biology and Biochemistry, Wesleyan University, Middletown, Connecticut, USA

SecA signal peptide interaction is critical for initiating protein translocation in the bacterial Sec-dependent pathway. Here, we have utilized the recent nuclear magnetic resonance (NMR) and Förster resonance energy transfer studies that mapped the location of the SecA signal peptide-binding site to design and isolate signal peptide-binding-defective secA mutants. Biochemical characterization of the mutant SecA proteins showed that Ser226, Val310, Ile789, Glu806, and Phe808 are important for signal peptide binding. A genetic system utilizing alkaline phosphatase secretion driven by different signal peptides was employed to demonstrate that both the PhoA and LamB signal peptides appear to recognize a common set of residues at the SecA signal peptide-binding site. A similar system containing either SecA-dependent or signal recognition particle (SRP)-dependent signal peptides along with the prlA suppressor mutation that is defective in signal peptide proofreading activity were employed to distinguish between SecA residues that are utilized more exclusively for signal peptide recognition or those that also participate in the proofreading and translocation functions of SecA. Collectively, our data allowed us to propose a model for the location of the SecA signal peptide-binding site that is more consistent with recent structural insights into this protein translocation system.

A

pproximately one-third of the proteome of Gram-negative Eubacteria like Escherichia coli is comprised of proteins that make up the cytoplasmic membrane, periplasm, or outer membrane. A majority of these proteins are transported by the general secretion (Sec) pathway although many integral cytoplasmic membrane proteins are targeted to the translocon as nascent chain-ribosome complexes by the signal recognition particle (SRP)/signal recognition particle receptor system. The Sec pathway is minimally composed of the SecA ATPase nanomotor that drives protein transport through a protein-conducting channel formed by the SecYEG complex (14, 38). Preproteins are initially targeted to SecA either cotranslationally or posttranslationally by signal peptide recognition. SecA also binds to a SecYEG dimer, where one protomer appears to serve as a SecA receptor by binding the SecA DEAD motor region, while active protein translocation occurs at the second SecYEG protomer utilizing the preprotein cross-linking and helical scaffold domains of SecA to drive the protein translocation step(s) (10, 30, 34). In particular, the twohelix finger region of the helical scaffold domain of SecA has been proposed to serve as a translocation ratchet (15, 50). Finally, since SecA has been reported to function as a homodimer (20, 22, 27), other higher-order SecA-SecYEG structures may also be required for the translocation of proteins with more complex final topologies. SecA is a complex protein with a conserved protomer fold consisting of six domains: two nucleotide-binding domains (NBD-I and NBD-II), a preprotein cross-linking domain (PPXD), the helical scaffold domain (HSD), the helical wing domain (HWD), and a carboxyl-terminal linker domain (CTL) (see Fig. 1A below) (19, 36, 40, 46, 49). The NBD-I and NBD-II domains comprise the DEAD motor region of SecA that contains a high-affinity nucleotide-binding site (29). The PPXD domain is connected to NBD-I by two ␤-strands, and both of these regions have been implicated in SecA signal peptide binding (3, 32, 37). The HSD domain can be further divided into two subdomains: a long central helix that forms a backbone and organizational template for the other SecA domains and the two-helix finger region that has

0021-9193/12/$12.00

Journal of Bacteriology

p. 307–316

been implicated in driving ongoing protein translocation at the mouth of the protein-conducting channel (15, 50). Finally, the CTL domain has been found to bind both acidic phospholipids as well as the chaperone SecB (5, 16), and it has also been recently proposed to serve as a competitive inhibitor of SecA signal peptide binding by blocking portions of the signal peptide-binding groove (17). The interaction between SecA and signal peptides has been an ongoing investigation for the past 2 decades, and a variety of approaches have been taken to elucidate this matter. A number of studies have utilized genetically mutated, truncated, or proteolytically cleaved SecA proteins along with signal peptide-binding or cross-linking assays to map portions of the relevant ligandbinding site on SecA (3, 31, 32, 37). At best, such approaches are of limited utility because either they identify a small number of residues that may or may not directly contribute to signal peptide binding or they facilitate construction of a linear rather than three-dimensional map of the actual binding site since they also include irrelevant regions and exclude relevant ones. Recently, structural approaches have been taken to define the SecA signal peptide-binding site. A nuclear magnetic resonance (NMR) structure of Escherichia coli SecA bound to the KRR-LamB signal peptide was determined where the signal peptide was positioned within a peptide-binding groove along the HWD-PPXD interface, with a roughly perpendicular orientation of the bound signal peptide relative to the scaffold domain of SecA (17). Another study utilizing Förster resonance energy transfer (FRET) mapped the alkaline phosphatase signal peptide (PhoA)-binding domain of

Received 8 September 2011 Accepted 27 October 2011 Published ahead of print 4 November 2011 Address correspondence to Donald B. Oliver, [email protected]. Supplemental material for this article may be found at http://jb.asm.org/. Copyright © 2012, American Society for Microbiology. All Rights Reserved. doi:10.1128/JB.06150-11

jb.asm.org

307

Grady et al.

SecA to a region encompassing portions of PPXD, HSD, and NBD-I (2). While these two studies agree in the approximate location of the SecA signal peptide-binding domain, they do differ significantly in that the latter study implicated more extensively the two-helix finger region of SecA in signal peptide interaction; furthermore, that study was more consistent with a parallel rather than perpendicular orientation of the bound signal peptide relative to the scaffold domain of SecA. The SecA signal peptidebinding site proposed in the FRET-based study is supported by recent work showing that the two-helix finger region of SecA interacts with the preprotein during translocation and lies at the entrance to the protein-conducting channel, based on the recent SecA-SecYEG structure (4, 15, 50). Here, we have combined genetic, biochemical, and physiological approaches in order to further elucidate the SecA signal peptide-binding domain and, in particular, to investigate the differences between the NMR- and FRET-resolved signal peptidebinding sites. Our results pinpoint a number of amino acid residues that contribute to the SecA signal peptide-binding site, and they further distinguish between residues that appear to contribute solely to SecA signal peptide binding and those that also play a role downstream during translocation. Our study allows us to propose a model for the location of the SecA signal peptidebinding site that is more consistent with recent structural insights into this protein translocation system. MATERIALS AND METHODS Media and chemicals. LB (Miller) broth and agar were obtained from EMD Chemicals and Difco, respectively. Other chemicals were obtained from Sigma or a comparable supplier and were reagent quality or better. The alkaline phosphatase signal peptide (MKQSTIALALLPLLFTPVTKA C-CONH2) was synthesized (Biomolecules Midwest, Waterloo, IL), purified by high-performance liquid chromatography (HPLC), labeled at the C-terminal cysteine with IANBD ester (N-{[2-(iodoacetoxy)ethyl]-Nmethyl}-amino-7-nitrobenz-2-oxa-1,3-diazole) (Molecular Probes), and repurified as previously described (2). Strains, plasmids, and protein purification. Escherichia coli BL21.19 [secA13(Am) supF(Ts) trp(Am) zch::Tn10 recA::CAT clpA::kan] (29) and BL21.20 [secA13(Am) supF(Ts) trp(Am) zch::Tn10 recA635::kan] containing the pLysS plasmid are derived from BL21(␭DE3) (42) and were used as hosts for secA-containing plasmids. BL21.21 [BL21(␭DE3) secA13(Am) supF(Ts) trp(Am) zch::Tn10 prlA4-2 rpsE ⌬recA635::kan] is isogenic with BL21.20; it was constructed by P1 transduction, and the presence of the prlA4-2 and ⌬recA635::kan alleles was verified by PCR amplification and DNA sequencing of the relevant chromosomal regions utilizing appropriate oligonucleotide primers. BL21.22 was constructed from BL21.21 by transformation of the pLysS plasmid (42). Plasmid pT7secA-his contains the T7 promoter-driven secA gene fused to a C-terminal His tag in pET29b (Novagen) (51). Plasmid mutations were made by the QuikChange (Stratagene) method utilizing oligonucleotide primers (Integrated DNA Technologies) designed on the Agilent (Santa Clara, CA) website, and all mutations were verified by DNA sequence analysis (University of Pennsylvania DNA Sequencing Facility). Plasmids containing secA mutations were transformed into BL21.19 or BL21.20 and checked for secA complementation by comparing their plating efficiency at 42°C and 30°C as described previously (20). His-tagged SecA proteins were overproduced and purified utilizing a HisBind resin column (Novagen) according to the manufacturer’s protocol as described previously (51), with the following modification: cells were washed after being harvested with 10 mM Tris-HCl (pH 7.5), 50 mM KCl, and 10 mM Mg(OAc)2, and purified SecA protein was dialyzed against 25 mM TrisHCl (pH 7.5), 25 mM KCl, 0.5 mM EDTA, and 0.5 mM phenylmethane-

308

jb.asm.org

sulfonyl fluoride. Protein concentration was determined using a Bradford assay (Bio-Rad) with bovine serum albumin (BSA) as the standard. Construction of pCDF-BAD-phoA, pCDF-BAD-dsbA-phoA, and pCDF-BAD-KRR-lamB-phoA plasmids. Plasmid pCDF-BAD-phoA contains the E. coli phoA gene under the control of the araBAD promoter and araC regulator with a replication origin derived from plasmid CloDF13 along with a streptomycin/spectinomycin resistance gene. It was constructed as follows: pCDF-1b (Novagen) was cleaved with PstI and XbaI, and the 2.0-kb PstI-XbaI DNA fragment was gel purified and ligated to a 2.7-kb NsiI-XbaI DNA fragment derived from pBAD33 (18). The resulting plasmid, pCDF-BAD-phoA, was confirmed by restriction enzyme mapping as well as arabinose-dependent expression of alkaline phosphatase activity. Plasmids pCDF-BAD-dsbA-phoA and pCDF-BADKRR-lamB-phoA are derived from pCDF-BAD-phoA, whereby the phoA signal sequence was first deleted from the latter plasmid and then replaced with either the dsbA or KRR-LamB signal sequences, respectively, utilizing the QuikChange method and appropriate oligonucleotide primers. Introduction of the entire dsbA or KRR-LamB signal sequence required three consecutive rounds of QuikChange mutagenesis where 18 to 30 nucleotides of the desired signal sequence were introduced each round, and all engineered plasmids were verified by DNA sequence analysis. Alkaline phosphatase assay. Overnight cultures of BL21.20 or BL21.22 containing both the pT7SecA-his wild-type or mutant secA allele and pCDF-BAD-phoA, pCDF-BAD-dsbA-phoA, or pCDF-BADKRRlamB-phoA were subcultured at a 1:50 dilution and grown in LB broth supplemented with 100 ␮g/ml ampicillin, 25 ␮g/ml chloramphenicol, and 50 ␮g/ml streptomycin at 30°C to an A600 of 0.2. The culture was subsequently diluted 3-fold with fresh medium and grown at 39°C for 100 min or 200 min, after which alkaline phosphatase production was induced by addition of arabinose to 0.5%. Aliquots (1.4 ml) were removed at 0, 10, 20, 30, and 40 min after induction and placed on ice until assayed. Cells were washed with 1.0 M Tris-HCl (pH 8.0), resuspended in this buffer, and assayed at 37°C for alkaline phosphatase activity as described previously (6). Fluorescence anisotropy. Anisotropy experiments were performed as described previously (2) with the following modifications. Fluorescence anisotropy was recorded on a FluoroMax-4 spectrofluorometer (Horiba Jobin Yvon) with a programmable water bath (Thermo scientific). Data were fit using ORIGIN, version 8.0, with the following equation:



y ⫽ A0 ⫹ (Ai ⫺ A0)



[SP] ⫹ Kd ⫹ [P] ⫺ 冑([SP] ⫹ Kd ⫹ [P])2 ⫺ 4[SP][P] 2[SP]

where [SP] is the total concentration of the signal peptide, [P] is the total concentration of SecA protein, Kd is the equilibrium dissociation constant, A0 is the anisotropy of the signal peptide in the absence of SecA, and Ai is the anisotropy under saturating binding conditions. Signal peptide-stimulated SecA lipid ATPase assay. Small unilamellar vesicles (SUV) were prepared from E. coli phospholipids (Avanti Polar Lipids) as previously described (28), with the following modifications. Chloroform was evaporated from the phospholipid suspension under a stream of nitrogen gas, followed by drying in a lyophilizer for 24 h. The lipids were resuspended in 1 mM dithiothreitol, and SUV were formed by bath sonication (Special Ultrasonic Cleaner) until optical clarity was obtained (⬃40 min). Signal peptide-stimulated SecA lipid ATPase activity was measured in 50 mM HEPES-KOH (pH 8.0), 30 mM KCl, 30 mM NH4Cl, 0.5 mM Mg(OAc)2, 1 mM dithiothreitol, 4 mM ATP, 0.5 mg/ml BSA, 40 ␮g/ml SecA, and 320 ␮g/ml SUV in a total reaction volume of 25 ␮l. The reactions were initiated by the addition of SecA to samples containing increasing concentrations of the unlabeled PhoA signal peptide (0 to 30 ␮M), and incubation was at 37°C for 30 min. The concentration of inorganic phosphate (Pi) released was determined using the malachite green colorimetric method (23).

RESULTS

Isolation of SecA signal peptide-binding-defective mutants. We undertook the current study in order to further elucidate how

Journal of Bacteriology

SecA Signal Peptide Interaction

FIG 1 Location of mutated residues on the SecA-bound KRR-LamB NMR structure (Protein Data Bank code 2VDA). (A) The ribbon diagram is colored according to SecA domains or subdomains as follows: NBD-I is in dark blue, NBD-II is in light blue, PPXD is in orange, the central helix and two-helix finger subdomains of HSD are in dark green and pink, respectively, HWD is in light green, and the resolved portion of CTL is in red. The KRR-LamB signal peptide is shown in purple. (B and C) The KRR-LamB signal peptide is shown in purple, while the signal peptide-binding site identified by FRET is in yellow. The colored spheres represent the positions of the alpha-carbon atoms of the mutated residues. Orange and pink spheres indicate mutations located exclusively within the signal peptide-binding site determined by NMR and FRET, respectively, while blue spheres indicate mutations lying within the overlap between both sites, and black spheres indicate mutations lying outside either site (B). Aqua spheres indicate mutations that have 75% or less of the wild-type alkaline phosphatase level, while red spheres indicate mutations that are above this value (C).

SecA recognizes signal peptides and, in particular, to investigate the differences between the proposed signal peptide-binding sites determined by NMR and FRET approaches (2, 17). While SecA is a large multidomain protein (Fig. 1A), both studies agreed that signal peptide binding occurred at domain-domain interfaces within the central region of SecA (Fig. 1B). Accordingly, a total of 59 mutations were generated (Table 1) within the NMR- and FRET-proposed binding sites or within or proximal to the twohelix finger subdomain of SecA, given the importance of the latter region in SecA preprotein recognition and its proximity to the proposed signal peptide-binding sites (Fig. 1B) (15). Residues outside these three regions that may contribute to SecA signal peptide binding were not investigated here since they were beyond the scope of our study. SecA residues that showed the greatest shift in their methyl resonances, indicative of their proximity to the bound spin-labeled KRR-LamB signal peptide, or residues that were structurally close to the bound signal peptide were among those specifically selected to explore the NMR-proposed site (17). For both sites, phylogenetically conserved acidic and hydrophobic amino acid residues were preferentially chosen for substitution since SecA normally interacts with both the basic amino-terminal and hydrophobic core regions of signal peptides, where the latter interaction is essential for SecA signal peptide recognition (1, 7, 48). We utilized both alanine and arginine substitution mutagenesis since the former amino acid has the advantage of its small neutral side chain that is readily accommodated into a wide variety of structural contexts while the latter one is useful in potentially perturbing many of the hydrophobic or acidic amino acid residues that were selected in our study. In general, single amino acid residues were mutated, but in a few cases double or even triple substitutions of adjacent residues were constructed in order to create potentially stronger defects. The effect the mutations had on the essential function of SecA

January 2012 Volume 194 Number 2

was assessed by performing complementation experiments in two different hosts, BL21.19 and BL21.20. Both strains are BL21(␭DE3) derivatives that contain the secA13(Am) and supF(Ts) mutations so that chromosomal secA expression can be blocked by growth at 42°C, thus allowing for the sole expression or overexpression of the plasmid-borne secA mutant by the T7 promoter (11). BL21.19 has been shown to overexpress plasmidborne secA approximately 8-fold, while BL21.20 has been shown to express plasmid-borne secA at chromosomal levels due to the presence of the pLysS plasmid that reduces T7 RNA polymerase activity (42). Most mutations complemented secA function in vivo in both host strains (i.e., displayed plating efficiencies within ⬃50% of the wild-type level), with some notable exceptions (see Table S1 in the supplemental material). This result was not unexpected even if a number of these substitutions perturb SecA signal peptide binding since SecA interacts with signal peptides over a sizable surface area, where a single amino acid substitution would have only a modest effect on binding (2, 17). Furthermore, chaperones, such as SecB protein, that directly transfer preproteins to SecA may also help to suppress SecA signal peptide-binding defects (9). In contrast, mutations that severely affected or abolished secA function may affect another essential SecA-dependent step, may be the result of multiple defects (i.e., residues involved in multiple SecA-dependent protein translocation steps), or may simply affect SecA folding or stability. Given the speculative nature of these arguments, however, we retained an interest in both classes of mutations. In order to utilize a more precise screen for secA function, we monitored the rate of alkaline phosphatase secretion in our secA mutants. Alkaline phosphatase is an ideal enzyme to employ as a reporter here for the following reasons: (i) its secretion is normally SecB independent, thus simplifying our analysis (21); (ii) it needs to be secreted to the periplasm in a SecA-dependent fashion in

jb.asm.org 309

Grady et al.

TABLE 1 Alkaline phosphatase activities of secA mutants Vector or SecA proteina WT pET29b R188A D189A N190A D189A N190A N190A I224A I225A S226A I225R S226A S233A M235A M235R V239R I291R M292R I304A M305A L306A I304A L306A L306A L306R V310R T371A T371R L372R V651R A658R D668A D668R D671A D671R E708A E708R N712A N712E E757A E757R M758R F762R V766R M767R L771R L774A K776A A781R L785R I789R D798A D798R E802A E802R E806A E806R F808R M810R A812R M814R L815R E821A E821R V822R L826R

NMR and/or FRET detection of binding siteb NA NA Both Both Both Both Both Both Both Both Both Both Both Both Both Both Both Both Both FRET FRET Both FRET FRET Neither Neither Neither Neither NMR NMR NMR NMR Neither Neither Neither NMR Neither Neither FRET Both FRET FRET FRET FRET FRET FRET FRET FRET FRET FRET Both Both Both FRET FRET FRET Neither Neither

Alkaline phosphatase ratioc 1.00 0.09 0.68d 0.42d 0.90 0.63d 0.73d 0.82 1.15 0.99 1.40 0.47d 1.00 0.76 0.85 0.78 0.95 0.85 0.10d 1.00 0.65d 1.13 0.75d 0.55d 1.15 0.60d 1.15 1.12 1.18 0.61d 1.12 1.12 0.61d 0.90 0.74d 0.87 1.26 0.58d 0.90 0.98 0.77 0.61d 0.06d 0.71d 0.90 0.65d 0.06d 0.84 0.19d 0.52d 1.06 1.20 1.06 0.65d 1.53 0.47d 0.58d 1.13

a Mutations in amino acid residue(s) in SecA are shown in one-letter code. pET29b is a plasmid vector lacking the secA gene, which is present in the wild type. b Indicates if the mutated residue lies within the signal peptide-binding site determined by NMR or FRET, by both methods, or by neither. NA, not applicable. c The alkaline phosphatase ratio was determined by plotting the alkaline phosphatase activity versus time, as described in Materials and Methods, for the wild type (WT), vector only (pET29b), or indicated secA mutant and then dividing the slope of each line by the comparable slope of the isogenic strain carrying the wild-type secA plasmid. d secA mutants that have 75% or less of the wild-type alkaline phosphatase level.

310

jb.asm.org

order to acquire its enzymatic activity (26), thus allowing for a simple colorimetric assay for screening; and (iii) the PhoA signal peptide was used in the previous FRET-based mapping study (2). For this purpose we constructed the pCDF-BAD-phoA plasmid containing the phoA gene under the control of the tightly regulated araBAD promoter that has a replication origin derived from CloDF13 along with a streptomycin/spectinomycin resistance gene, thereby allowing for plasmid compatibility and antibiotic selection with our pT7secA-his vector containing our secA mutations (see Materials and Methods) (18). This system allowed us to induce alkaline phosphatase expression only after sufficient depletion of chromosomal secA expression had been achieved by growth at the restrictive temperature and therefore to assess the function of the plasmid-encoded secA gene. We found that 100 min of growth at 39°C was sufficient to reduce subsequent alkaline phosphatase expression and export in BL21.20(pET29b, pCDFBAD-phoA) lacking any plasmid-borne secA gene to less than 10% of the level present in BL21.20(pT7secA-his, pCDF-BAD-phoA) containing the wild-type secA gene (Table 1). Furthermore, in an initial pilot study to correlate signal peptide-binding affinity with the alkaline phosphatase secretion rate, we found that purified SecA protein with the I225R mutation [SecA(I225R)] displayed a 4-fold decrease in signal peptide-binding affinity compared to wild-type SecA (Table 2), and its alkaline phosphatase secretion was 73% of the wild-type level. Using this information, we narrowed our focus down to 24 mutants that possessed 75% or less of wild-type alkaline phosphatase secretion levels (Table 1). While the choice of this value was based on only a single mutant, it allowed us to focus on a tractable number of secA mutants for which additional studies, particularly of a detailed biochemical and physiological nature, were feasible. We also noted that the 24 mutations selected here provided good coverage of the three regions of interest, the signal peptide-binding sites determined by NMR and FRET approaches as well as the two-helix finger region of SecA (Fig. 1C). It is certainly conceivable that, within the subset of secA mutants eliminated by this approach, some could be involved in SecA signal peptide recognition. Biochemical studies. SecA protein was purified from the 24 mutants selected for further analysis along with a few additional mutants that were created to separate out the contribution of double or triple substitutions in adjacent residues to the observed phenotype. SecA signal peptide-binding assays were conducted by fluorescence anisotropy using IANBD-labeled PhoA signal peptide. Wild-type SecA protein bound PhoA signal peptide with an affinity of 2.4 ⫾ 0.8 ␮M, in agreement with previously published values (Table 2; see also Fig. S1 in the supplemental material) (2, 3, 31). Most of our SecA mutant collection had signal peptidebinding affinity constants that were within 3-fold of wild-type SecA, indicating that the relevant substitution had a modest effect or no effect on SecA signal peptide binding. However, SecA proteins with substitutions at I224A I225A S226A, I225R, S226A, V310R, E806R, and F808R displayed greater binding defects in the 3- to 5-fold range (Table 2; see Fig. S1A to C). These mutants do not display a saturated binding curve, so we carried out the titration to the highest SecA concentration possible (40 ␮M) before aggregation occurs and found that there was little change in the binding constants (data not shown). Since these substitutions resulted in the greatest defects in SecA signal peptide binding within our collection, we focused on them for the remainder of this study. Because we were also interested in the potential role of the two-

Journal of Bacteriology

SecA Signal Peptide Interaction

TABLE 2 Signal peptide-binding constants of SecA mutant proteins SecA protein

Kd (␮M)a

WT D189A N190A N190A I224A I225A S226A I225R I224A I225A S226A V239R I304A L306A V310R T371R V651R A658R D668R E708R E757A M758R M767R L785R I789R D798A E802A E802R E806R F808R L815R E821R V822R

2.4 ⫾ 0.8 6.5 ⫾ 2.4 1.4 ⫾ 0.7 12.2 ⫾ 0.9b 11.0 ⫾ 2.7b 1.9 ⫾ 0.7 4.5 ⫾ 0.5 10.0 ⫾ 1.3b 5.7 ⫾ 1.1 3.4 ⫾ 1.6 12.4 ⫾ 3.0b 1.3 ⫾ 0.5 1.1 ⫾ 0.3 2.8 ⫾ 0.9 2.8 ⫾ 0.5 3.5 ⫾ 0.5 4.6 ⫾ 1.6 4.5 ⫾ 1.4 4.4 ⫾ 0.4 3.9 ⫾ 1.0 5.1 ⫾ 0.5c 5.2 ⫾ 1.5 3.5 ⫾ 0.9 1.8 ⫾ 0.7 8.9 ⫾ 1.3b 7.4 ⫾ 1.9b 4.1 ⫾ 0.8 6.1 ⫾ 1.1 4.3 ⫾ 0.9

a The equilibrium binding constant of the wild-type (WT) or indicated SecA mutant protein for alkaline phosphatase signal peptide was determined by fluorescence anisotropy, as described in Materials and Methods. The values given are an average of at least three experiments. b SecA mutant protein with an ⬃3- to 5-fold decrease in alkaline phosphatase signal peptide binding. c I789R was selected for further analysis, given that it is located in the two-helix finger of SecA, which was previously implicated in a signal peptide-binding defect (47).

helix finger in signal peptide binding, we included the I789R mutant in our study, given that it was previously reported to have a 2-fold defect in binding the 3K7L signal peptide, as assessed by surface plasmon resonance (47). It was notable that for the SecA(I224A I225A S226A) triple mutant protein, most of the binding defect was due to the S226A substitution although the less conservative substitution I225R also resulted in a significant binding defect. We obtained relatively normal signal peptide-binding affinity for SecA(I304A L306A) (Kd of 3.4 ⫾ 1.6 ␮M), which was specifically created because it was previously reported by isothermal calorimetry to possess an 8-fold reduction in its affinity for the KRR-LamB signal peptide (17). This discrepancy may be due to the different signal peptides or methodologies employed in these measurements (e.g., calorimetry was done at much higher SecA protein and signal peptide concentrations than were utilized here). To rule out the possibility that these mutations were affecting SecA signal peptide interaction indirectly (e.g., due to conformational changes in SecA), we performed trypsinolysis studies on the mutant SecA proteins. All of the SecA proteins analyzed here with the exception of SecA(V310) gave trypsin digestion patterns similar to the pattern of wild-type SecA (see Fig. S2 in the supplemen-

January 2012 Volume 194 Number 2

tal material). Thus, the signal peptide-binding defect observed for SecA(V310R) may be of a more indirect nature (i.e., a poor conformation for signal peptide binding). We also measured the endogenous and membrane ATPase activities of our mutant SecA proteins since the former activity provides information about the correct conformation of the SecA DEAD motor, while the latter activity is dependent on the ability of SecA to productively bind SecYEG and accelerate its nucleotide exchange rate (33, 41). We found that most of the mutant SecA proteins had ATPase activities similar to wild-type levels with the exception of SecA(I789R) and SecA(E806R), whose endogenous ATPase activities were modestly increased (less than 2-fold) (see Fig. S3 in supplemental material). The latter result is consistent with the location of these two substitutions within the intramolecular regulator of ATPase1 (IRA1) region of the two-helix finger, which has been shown to be important for allosteric communication between the HSD domain and DEAD motor regions of SecA (47). We also utilized a second independent methodology to directly assess the signal peptide-binding affinity of our mutant SecA proteins, namely, the ability of functional signal peptides to specifically stimulate SecA ATPase activity in the presence of liposomes (28). We found that SecA proteins with substitutions at S226A, I789R, E806R, and F808R displayed significant defects in signal peptide-stimulated SecA ATPase activity (Fig. 2A and B). Of interest, while the SecA(V310R) protein was also clearly defective in signal peptide interaction at low signal peptide concentrations, it displayed a cooperative increase in its signal peptide-stimulated ATPase activity, suggesting that the presence of lipids helped to partially suppress this defect. This result may be due to the altered conformation of this protein, as detected in our trypsinolysis study above (see Fig. S2 in the supplemental material), and its suppression by lipid binding. The latter suggestion is consistent with the normal membrane ATPase activity noted above for SecA(V310R) (see Fig. S3). We also found that the SecA(I304A L306A) mutant protein had a modest decrease in SecA signal peptide interaction by this assay, at least qualitatively supporting the signal peptide-binding defect noted previously (17). The residues we have identified as important for SecA signal peptide interaction are located in two very different regions of SecA. Ser226 lies in the ascending ␤-strand connecting the PPXD domain with NBD-I, while Val310 resides within the third ␣-helix of the bulbous portion of PPXD (see Fig. 5A below). In contrast, Ile789 is located at the tip of the two-helix finger subdomain, while Glu806 and Phe808 lie toward the middle of the second ␣-helix of this subdomain. Collectively, these residues fall into two different groups in reference to the prior studies: S226A and V310R lie within the SecA signal peptide-binding sites determined by both NMR and FRET approaches, while I789R, E806R, and F808R are consistent only with the latter site. Physiological studies. Given the discrepancy between the NMR- and FRET-determined signal peptide-binding sites as well as in the behavior of the SecA(I304A L306A) mutant in the present versus previous study (17), we wondered whether SecA might recognize the PhoA and KRR-LamB signal peptides somewhat differently. In order to test this idea, we constructed a variant of our pCDF-BAD-phoA plasmid that contained the KRR-LamB signal sequence substituted for the PhoA signal sequence, and we compared alkaline phosphatase secretion directed by these two signal peptides in our secA mutants that were defective in PhoA signal peptide binding. We found that the levels of alkaline phosphatase

jb.asm.org 311

Grady et al.

FIG 3 Comparison of alkaline phosphatase secretion driven by the PhoA or KRR-LamB signal peptides in different secA mutants. The percentage of alkaline phosphatase activity was determined by plotting the alkaline phosphatase activity versus time, as described in Materials and Methods, for the wild-type (WT) or indicated secA mutant carrying either the pCDF-BAD-phoA plasmid with the native PhoA signal peptide or the pCDF-BAD-KRR-lamB-phoA plasmid with the KRR-LamB signal peptide and then dividing the slope of each line by the comparable slope of the isogenic wild-type secA strain and multiplying by 100.

FIG 2 SecA binding affinity for the alkaline phosphatase signal peptide as measured by enhancement of SecA lipid ATPase activity. Signal peptidedependent stimulation of SecA lipid ATPase activity was measured at the indicated alkaline phosphatase (PhoA) signal peptide concentration for the wild-type (WT) or the indicated mutant SecA protein as described in Materials and Methods. Data from triplicate assays and are divided for purposes of clarity.

secretion directed by either the PhoA or KRR-LamB signal peptides were similar for each of our secA mutants, irrespective of the severity of the different secA alleles on in vitro signal peptide binding or in vivo alkaline phosphatase secretion (Fig. 3). This result indicates that, within the limits of our set of signal peptide recognition-defective alleles to provide readout on SecA signal peptide binding, SecA recognizes both of these signal peptides quite similarly. We next addressed the concern that some of the secA mutants utilized in this experiment may also be defective in a step downstream of signal peptide recognition. In order to distinguish the signal peptide recognition function of SecA from its later role in driving the stepwise protein translocation at SecYEG, we made use of the DsbA signal peptide that has been shown to redirect SecA-dependent preproteins to the translocon via the signal recognition particle (SRP)-dependent

312

jb.asm.org

pathway (39). Preproteins that contain more hydrophobic signal peptides like DsbA are targeted to the translocon in a cotranslational manner utilizing the SRP system for signal peptide recognition, after which they are transferred to SecA for translocation across SecYEG into the periplasm. Thus, comparison of the SecA secretion levels of alkaline phosphatase containing either the native PhoA or DsbA signal sequence should allow us to distinguish between these two different steps in SecA-dependent protein transport. Accordingly, we constructed a variant of our pCDFBAD-phoA plasmid that contained the DsbA signal sequence substituted for the PhoA signal sequence, pCDF-BAD-dsbA-phoA, and we compared alkaline phosphatase secretion levels from these two plasmids in our secA mutants of interest. We initially noticed that alkaline phosphatase secretion from BL21.20(pET29b, pCDF-BAD-dsbA-phoA) was significantly higher than that from BL21.20(pET29b, pCDF-BAD-phoA) (Fig. 4A). This result indicates that as SecA is depleted by our growth regimen at 39°C, its signal peptide-binding function is more drastically affected than its translocation motor function at SecYEG. Stated another way, at rate-limiting concentrations of wild-type SecA, the signal peptide recognition step appears to be the rate-limiting step in the protein translocation pathway. We found that secA mutants S226A, I304A L306A, V310R, and F808R had DsbA-directed alkaline phosphatase secretion levels similar to or greater than the wild-type strain, but their PhoA-directed alkaline phosphatase secretion levels were reduced by comparison (Fig. 4A). This result indicates that these mutants are defective in the SecA signal peptide recognition step that is bypassed using SRP targeting, whereas they appear to be relatively functional in the subsequent SecA-dependent translocation steps at SecYEG. The reason for the higher-than-wild-type level of DsbA-directed alkaline phosphatase noted for the S226A mutant is unclear presently, but it may relate to an effect of this secA allele on kinetic proofreading (see next paragraph). In con-

Journal of Bacteriology

SecA Signal Peptide Interaction

trast, since alkaline phosphatase secretion remained at background levels (similar to or lower than levels of the pCDF-BADdsbA-phoA- or pCDF-BAD-phoA-containing strains with the pET29b vector lacking any secA gene) for secA mutants I789R and E806R, these two SecA proteins appear to be relatively inactive in both the proximal and distal SecA-dependent steps. Additional physiological experiments where chromosomally derived SecA was more extensively depleted suggested that the E806R mutant protein was slightly active, while the I789R mutant protein was inactive (Fig. 4B). It occurred to us that improper recognition of signal peptides by our mutant SecA proteins might reduce alkaline phosphatase secretion due to the proofreading mechanism that normally rejects preproteins with defective signal peptides. Such proofreading appears to involve signaling between SecA and SecYEG when a signal peptide is properly bound to this complex. Mutations in these components have been isolated that bypass the proofreading step, of which prlA4-2 (an allele of secY) is one of the strongest since it allows translocation of preproteins devoid of signal sequences by stabilization of SecA-SecYEG association and preactivation of the channel complex (13, 25, 35, 45). To determine whether the SecA signal peptide recognition-defective mutants affected the proofreading step due to improper binding of the alkaline phosphatase signal peptide, we measured alkaline phosphatase secretion in isogenic strains containing prlA⫹ and prlA4-2 alleles. We found similar alkaline phosphatase secretion levels in both strain backgrounds for most of our secA mutants with the exception of S226A, where the presence of the prlA4-2 allele did increase alkaline phosphatase secretion by approximately 2-fold (Fig. 4C). The latter result suggests that this SecA protein may normally be somewhat defective in signaling the presence of a bound signal peptide to SecYEG and that proofreading may account in part for the reduced alkaline phosphatase secretion level present in the prlA⫹ background. The latter result may be due to improper signal peptide binding, an altered conformation of SecA(S226A), or involvement of Ser226 in normally promoting signaling of the signal peptide-binding event to the SecYEG complex in order to induce channel activation. DISCUSSION

The present study was undertaken in order to further elucidate how SecA recognizes signal peptides and, in particular, to investigate the differences between the proposed signal peptide-binding sites determined by NMR and FRET approaches (2, 17). To accomplish this task, we engineered 59 mutations into the relevant region of SecA that should allow us to investigate the importance of the two proposed signal peptide-binding sites as well as to assess the importance of the two-helix finger subdomain in constituting

FIG 4 Comparison of alkaline phosphatase secretion driven by the PhoA or DsbA signal peptides for different secA mutants. The percent alkaline phosphatase activity was determined by plotting the alkaline phosphatase activity versus time, as described in Materials and Methods, for the wild-type (WT), vector only (pET29b), or indicated secA mutant strain carrying either the pCDF-BAD-phoA plasmid with the native PhoA signal peptide or the pCDFBAD-dsbA-phoA plasmid with the DsbA signal peptide and then dividing the

January 2012 Volume 194 Number 2

slope of each line by the comparable slope of the isogenic wild-type secA strain and multiplying by 100. (A) Alkaline phosphatase activity was determined in cultures depleted of chromosomal SecA for 100 min at the nonpermissive temperature, prior to alkaline phosphatase induction with arabinose. D209N was included as an example of a secA null mutant. (B) Alkaline phosphatase activity was determined in the indicated strains after depletion of chromosomal SecA for 200 min at the nonpermissive temperature, prior to alkaline phosphatase induction with arabinose. (C) Alkaline phosphatase activity was determined for the indicated secA mutant in the isogenic BL21.20(pCDFBAD-phoA) prlA⫹ (black bars) or BL21.22(pCDF-BAD-phoA) prlA4-2 (gray bars) host after 100 min at the nonpermissive temperature prior to alkaline phosphatase induction with arabinose.

jb.asm.org 313

Grady et al.

this site. By screening for defects in alkaline phosphatase secretion, we were able to narrow down our focus to a tractable number of 24 mutants for which detailed physiological and biochemical assays were performed. Utilizing two different signal peptide-binding assays, we found that Ser226, Val310, Ile789, Glu806, and Phe808 are important in the SecA signal peptide recognition step although some of these residues have additional roles in SecA function (described below), given the linkage between SecA signal peptide binding and activation of the downstream translocation steps. A number of general conclusions arise from our studies of SecA signal peptide recognition. First, given that the strongest signal peptide-binding defects that we isolated had reduced binding affinities of only 5-fold (Table 2) and that such mutants could be fully viable with only modest protein secretion defects (Table 1; see also S226A in Table S1 in the supplemental material), the signal peptide-binding site of SecA protein appears to be relatively accommodating of single amino acid substitutions. This observation is consistent with the size and complexity of the relevant binding surface that ordinarily recognizes both the aminoterminal basic and hydrophobic core regions of signal peptides (1, 2, 7, 17, 48). In that regard, the residues identified here that are important for SecA signal peptide binding represent only a limited subset of residues that constitute this domain. Exhaustive mutagenesis of every SecA residue within the two proposed regions and perhaps beyond would be required to pinpoint all of the residues that are important in this function. Since signal peptide binding to soluble SecA is relatively weak, based on our observed binding data as well as the moderately fast exchange rate observed by NMR (7), and since both the NMR- and FRET-determined sites are located at domain-domain interfaces, it seems likely that additional signal sequence binding capacity as well as diversity could be accommodated through dynamic movements of SecA that subtly change the relevant binding surfaces. Indeed, prlD alleles of secA modestly suppress signal peptide defects, and they are distributed along several of the relevant domain-domain interfaces, where they destabilize the SecA ground state (19). Such interdomain flexing may contribute to signal peptide trapping when SecA associates with other binding partners such as SecB or anionic phospholipids or, alternatively, when SecA binds SecYEG and the complex commits to translocation of its bound substrate. Second, since SecA appears to recognize both the PhoA and KRRLamB signal peptides similarly (Fig. 3), it seems likely that there are certain critical residues within the SecA signal peptide-binding site that are utilized in common to recognize most, if not all, SecA-dependent signal peptides. Third, a number of the residues that we identified here as critical for signal peptide recognition appear to serve additional downstream roles in SecA function. This observation was anticipated since SecA signal peptide recognition needs to facilitate downstream activation of the translocase complex. Presumably certain SecA residues within the signal peptide-binding site are utilized as sensors to transduce the presence of the correctly bound signal peptide to the SecA-SecYEG complex in order to facilitate channel activation and entry of the first segment of the preprotein into the protein-conducting channel. Indeed, we obtained evidence that Ser226 may participate in the signal peptide-proofreading step (Fig. 4C), while Ile789 and Glu806 are important not only for the initial signal peptide recognition step but also in facilitating the more distal SecA-dependent steps at SecYEG (Fig. 4A and B). Previous characterization of the I789R mutant showed that this protein displayed defects in nucle-

314

jb.asm.org

otide, signal peptide, and SecYEG binding as well as in allosteric communication between the SecA DEAD motor and its C domain that consists of HSD, HWD, and CTL (47). While such “wiring” of the SecA signal peptide-binding site to downstream biochemical steps complicates our study, it lies at the heart of understanding the interconnectivity and regulation of the various SecA functions in the protein translocation cycle, as we describe in the following paragraph. The five residues that were the focus of our study can be divided into two groups: Ser226 and Val310 lie within both the NMR- and FRET-determined signal peptide-binding sites, while Ile789, Glu806, and Phe808 are consistent with only the FRETdetermined site. Ser226 lies in the ascending ␤-strand that forms one of the “stems” connecting the PPXD domain with NBD-I, while Val310 resides within the third ␣-helix of the bulbous portion of PPXD (Fig. 5A). Both of these regions have been implicated in SecA signal peptide binding previously (3, 31, 37). It is likely that the V310R mutation perturbs the structure of at least this portion of PPXD, given the increased trypsin sensitivity of this mutant protein (see Fig. S2 in the supplemental material). Based on its high B-factor value, the PPXD domain is weakly structured, and binding of functional signal peptides is known to significantly enhance SecA proteolysis (19, 48). Ile789 is located at the tip of the two-helix finger subdomain, while Glu806 and Phe808 lie toward the middle of the second ␣-helix of this subdomain. Interpretation of the effect of these three residues on signal peptide binding is complicated by the fact that they reside within the IRA1 regulatory region, which has been shown to affect a number of SecA properties, including nucleotide, signal peptide, and SecYEG binding as well as cross talk between the SecA DEAD motor and C domain (47). Thus, it remains possible that these mutations could affect SecA signal peptide interaction indirectly. We do not favor this hypothesis, however, since our study did contain other equally strong mutations within IRA1 that are known to disrupt its function (e.g., L785R, E802A, E802R [47]), and yet these mutations had little or no effect on SecA signal peptide-binding affinity (Table 2). In conjunction with our earlier FRET mapping study (2), we have now obtained additional evidence to indicate that the signal peptide may bind to SecA through interactions with the two-helix finger as well as the stem and bulbous portions of PPXD, thereby aligning itself parallel rather than perpendicular to the former region (modeled in Fig. 5B). This suggestion is, in fact, consistent with the first SecA X-ray structure, which showed occupancy of this peptide-binding groove by 24 residues of SecA’s weakly structured CTL domain (19). A regulatory interaction of this type, indeed, appears to be physiologically relevant since the SecA carboxyl-terminus has been found to competitively inhibit its signal peptide-binding activity (17). Furthermore, our proposed SecA signal peptide-binding site is consistent with the positioning of the two-helix finger subdomain at the entrance to the proteinconducting channel in the SecA-SecYEG X-ray structure (Fig. 5C) as well as with the ability to cross-link an arrested translocation intermediate to the loop of this subdomain (50). In this orientation the two-helix finger ratchet could readily insert the bound signal peptide into the protein-conducting channel, where it could be transferred to the appropriate binding site adjacent to the lateral gate of SecYEG (44). Finally, as assessed by the SecASecYEG structure, the proposed location would provide a continuous path between the SecA-bound signal peptide and the early

Journal of Bacteriology

SecA Signal Peptide Interaction

FIG 5 (A) Location of signal peptide binding-defective mutations on the SecA-bound KRR-LamB NMR structure. SecA and signal peptide are colored as described in the legend of Fig. 1B. Red spheres indicate the positions of the signal peptide-binding mutants. (B) E. coli NMR SecA structure with the KRR-LamB signal peptide modeled parallel to the two-helix finger. SecA is represented as a gray ribbon with the two-helix finger shown in yellow and the signal peptide shown in red. The region of SecA proposed to act as a preprotein-binding clamp is indicated by the blue arc (4). (C) Model of the position of the KRR-LamB signal peptide at the entrance of the SecY channel in the Thermotoga maritima SecA-SecYEG cocrystal. The signal peptide and SecA are colored as described for panel B, and SecYEG are shown in violet, light blue, and light green, respectively.

mature region bound to the SecA clamp (shown in Fig. 5B). This portion of SecA has been mapped recently to the large groove located between the PPXD and NBD-II domains both by X-ray crystallography of a SecA-bound hydrophilic peptide and through disulfide cross-linking analysis of arrested translocation intermediates (4, 49). In the latter studies the authors point out that a continuous path between these two distinct polypeptide-binding sites of SecA is not possible for the NMR-determined signal peptide-binding site since relevant portions of PPXD form critical contacts with cytosolic loops of SecY that stabilize the SecASecYEG complex (49). A similar problem occurs with this model during the transfer of the SecB-bound preprotein to the SecASecYEG complex, given the recently determined binding sites of preprotein and SecA on SecB (8, 24, 43). Preprotein transfer is known to occur efficiently from the SecB-bound preprotein complex to SecYEG-bound SecA (12). In contrast, both of these problems are avoided in our proposed orientation of the SecA-bound signal peptide, which is now adjacent to the clamp and accessible to solvent. Insertion of the signal peptide-bound two-helix finger ratchet into the mouth of the protein-conducting channel would allow for formation of a loop comprised of the signal peptide and early mature region of the preprotein that is necessary for the

January 2012 Volume 194 Number 2

initiation of protein translocation. It certainly remains possible, however, that SecA-signal peptide association is dynamic in nature, where both the prior and present studies capture legitimate intermediate states in the binding interaction. In conclusion, a combination of genetics, biochemistry, and physiology was used to identify SecA residues important for signal peptide recognition. The location of these residues within the signal peptide-binding domain of SecA suggests a novel orientation for the bound signal peptide and a clear mechanism for the initiation of protein translocation. Our studies also provide a window into the dual function of such residues in signal peptide recognition as well as signal transduction in order to facilitate downstream events at SecA-bound SecYEG. Our findings should promote future studies directed at further elucidation of important early steps in the SecA-dependent protein translocation pathway. ACKNOWLEDGMENTS We thank Ishita Mukerji for training in fluorescence spectroscopy and data analysis, Sarah Auclair, Aaron Olsen, Elena Georgieva, and Alison Hardy for technical support, and Sarah Auclair and Sanchaita Das for their intellectual input. This work was supported by grant GM42033 from NIGMS to D.O.

jb.asm.org 315

Grady et al.

REFERENCES 1. Akita M, Shinkai A, Matsuyama S, Mizushima S. 1991. SecA, an essential component of the secretory machinery of Escherichia coli, exists as homodimer. Biochem. Biophys. Res. Commun. 174:211–216. 2. Auclair S, et al. 2010. Mapping of the signal peptide-binding domain of Escherichia coli SecA using Forster resonance energy transfer. Biochemistry 49:782–792. 3. Baud C, et al. 2002. Allosteric communication between signal peptides and the SecA protein DEAD motor ATPase domain. J. Biol. Chem. 277: 13724 –13731. 4. Bauer BW, Rapoport TA. 2009. Mapping polypeptide interactions of the SecA ATPase during translocation. Proc. Natl. Acad. Sci. U. S. A. 106: 20800 –20805. 5. Breukink E, et al. 1995. The C terminus of SecA is involved in both lipid binding and SecB binding. J. Biol. Chem. 270:7902–7907. 6. Brickman E, Beckwith J. 1975. Analysis of the regulation of Escherichia coli alkaline phosphatase synthesis using deletions and ␾80 transducing phages. J. Mol. Biol. 96:307–316. 7. Chou Y-T, Gierasch L. 2005. The conformation of a signal peptid bound by Escherichia coli preprotein translocase SecA. J. Biol. Chem. 280: 32753–32760. 8. Cooper D, et al. 2008. SecA, the motor of the secretion machine, binds diverse partners on one interactive surface. J. Mol. Biol. 382:74 – 87. 9. Crane J, et al. 2005. Mapping of the docking of SecA onto the chaperone SecB by site-directed spin labeling: insight into the mechanism of ligand transfer during protein export. J. Mol. Biol. 353:295–307. 10. Das S, Oliver D. 2011. Mapping of the SecA-SecY and SecA-SecG interfaces by site-directed in vivo photocrosslinking. J. Biol. Chem. 286: 12371–12380. 11. Das S, Stivison E, Folta-Stogniew E, Oliver D. 2008. Reexamination of the role of the amino terminus of SecA in promoting its dimerization and functional state. J. Bacteriol. 190:7302–7307. 12. den Blaauwen T, Terpetschnig E, Lakowicz J, Driessen A. 1997. Interaction of SecB with soluble SecA. FEBS Lett. 416:35–38. 13. Derman AI, Puziss JW, Bassford PJJ, Beckwith J. 1993. A signal sequence is not required for protein export in prlA mutants of Escherichia coli. EMBO J. 12:879 – 888. 14. du Plessis D, Nouwen N, Driessen A. 2011. The Sec translocase. Biochim. Biophys. Acta 1808:851– 865. 15. Erlandson KJ, et al. 2008. A role for the two-helix finger of the SecA ATPase in protein translocation. Nature 455:984 –988. 16. Fekkes P, van der Does C, Driessen A. 1997. The molecular chaperone SecB is released from the carboxy-terminus of SecA during initiation of precursor protein translocation. EMBO J. 16:6105– 6113. 17. Gelis I, et al. 2007. Structural basis for signal-sequence recognition by the translocase motor SecA as determined by NMR. Cell 131:756 –769. 18. Guzman L-M, Belin D, Carson MJ, Beckwith J. 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol. 177:4121– 4130. 19. Hunt JF, et al. 2002. Nucleotide control of interdomain interactions in the conformational reaction cycle of SecA. Science 297:2018 –2026. 20. Jilaveanu LB, Zito CR, Oliver D. 2005. Dimeric SecA is essential for protein translocation. Proc. Natl. Acad. Sci. U. S. A. 102:7511–7516. 21. Kumamoto C, Francetic O. 1993. Highly selective binding of nascent polypeptides by an Escherichia coli chaperone protein in vivo. J. Bacteriol. 175:2184 –2188. 22. Kusters I, et al. 2011. Quaternary structure of SecA in solution and bound to SecYEG probed at the single molecule level. Structure 19:430 – 439. 23. Lanzetta PA, Alvarez LJ, Reinach PS, Candia OA. 1979. An improved assay for nanomole amounts of inorganic phosphate. Anal. Biochem. 100: 95–97. 24. Lilly A, Crane J, Randall L. 2009. Export chaperone SecB used one surface of interaction for diverse unfolded polypeptide ligands. Protein Science 18:1860 –1868. 25. Maillard A, Lalani S, Silva F, Belin D, Duong F. 2007. Deregulation of the SecYEG translocation channel upon removal of the plug domain. J. Biol. Chem. 282:1281–1287. 26. Manoil C, Mekalanos JJ, Beckwith J. 1990. Alkaline phosphatase fusions: sensors of subcellular location. J. Bacteriol. 172:515–518. 27. Mao C, Hardy S, Randall L. 2009. Maximal efficiency of coupling be-

316

jb.asm.org

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51.

tween ATP hydrolysis and translocation of polypeptides mediated by SecB requires two protomers of SecA. J. Bacteriol. 191:978 –984. Miller A, Wang L, Kendall D. 1998. Synthetic signal peptides specifically recognize SecA and stimulate ATPase activity in the absence of preprotein. J. Biol. Chem. 273:11409 –11412. Mitchell C, Oliver DB. 1993. Two distinct ATP-binding domains are needed to promote protein export by Escherichia coli SecA ATPase. Mol. Microbiol. 10:483– 497. Mori H, Ito K. 2006. Different modes of SecY-SecA interactions revealed by site-directed in vivo photo-crosslinking. Proc. Natl. Acad. Sci. U. S. A. 103:16159 –16164. Musial-Siwek M, Rusch S, Kendall D. 2005. Probing the affinity of SecA for signal peptides in different environments. Biochemistry 44: 13987–13996. Musial-Siwek M, Rusch S, Kendall D. 2007. Selective photoaffinity labeling identifies the signal peptide binding domain on SecA. J. Mol. Biol. 365:637– 648. Natale P, Swaving J, van der Does C, de Keyzer J, Driessen A. 2004. Binding of SecA to the SecYEG complex accelerates the rate of nucleotide exchange on SecA. J. Biol. Chem. 279:13769 –13777. Osborne AR, Rapoport TA. 2007. Protein translocation is mediated by oligomers of the SecY complex with one SecY copy forming the channel. Cell 129:97–110. Osborne RS, Silhavy TJ. 1993. PrlA suppressor mutations cluster in regions corresponding to three distinct topological domains. EMBO J. 12:3391–3398. Papanikolau Y, et al. 2007. Structure of dimeric SecA, the Escherichia coli preprotein translocase motor. J. Mol. Biol. 366:1545–1557. Papanikou E, et al. 2005. Identification of the preprotein binding domain of SecA. J. Biol. Chem. 280:43209 – 43217. Papanikou E, Karamanou S, Economou A. 2007. Bacterial protein secretion through the translocase nanomachine. Nat. Rev. Microbiol. 5:839 – 851. Schierle C, et al. 2003. The DsbA signal sequence directs efficient, cotranslational export of passenger proteins to the Escherichia coli periplasm via the signal recognition particle pathway. J. Bacteriol. 185: 5706 –5713. Sharma V, et al. 2003. Crystal structure of Mycobacterium tuberculosis SecA, a preprotein tranlsocating ATPase. Proc. Natl. Acad. Sci. U. S. A. 100:2243–2248. Sianidis G, et al. 2001. Cross-talk between catalytic and regulatory elements in a DEAD motor domain is essential for SecA function. EMBO J. 20:961–970. Studier WF, Rosenberg AH, Dunn JJ, Dubendorff JW. 1990. Use of T7 RNA polymerase to direct expression of cloned genes. Methods Enzymol. 185:60 – 89. Suo Y, Hardy S, Randall L. 2011. Orientation of SecA and SecB in complex, derived from disulfide cross-linking. J. Bacteriol. 193:190 –196. Van den Berg B, et al. 2004. X-ray structure of a protein-conducting channel. Nature 427:36 – 44. van der Wolk J, et al. 1998. PrlA4 prevents the rejection of signal sequence defective preproteins by stabilizing the SecA-SecY interaction during the initiation of translocation. EMBO J. 17:3631–3639. Vassylyev D, et al. 2006. Crystal structure of the translocation ATPase SecA from Thermus thermophilus reveals a parallel, head-to-head dimer. J. Mol. Biol. 364:248 –258. Vrontou E, Karamanou S, Baud C, Sianidis G, Economou A. 2004. Global co-ordination of protein translocation by the SecA IRA1 switch. J. Biol. Chem. 279:22490 –22497. Wang L, Miller A, Kendall DA. 2000. Signal peptide determinants of SecA binding and stimulation of ATPase activity. J. Biol. Chem. 275: 10154 –10159. Zimmer J, Rapoport TA. 2009. Conformational flexibility and peptide interaction of the translocation ATPase SecA. J. Mol. Biol. 394:606 – 612. Zimmer J, Nam Y, Rapoport TA. 2008. Structure of a complex of the ATPase SecA and the protein-translocation channel. Nature 455:936 –943. Zito C, Antony E, Hunt J, Oliver D, Hingorani M. 2005. Role of a conserved glutamate residue in the Escherichia coli SecA ATPase mechanism. J. Biol. Chem. 280:14611–14619.

Journal of Bacteriology