Type 1 secretion systems (T1SS) transport a wide range of substrates across both membranes of Gram-negative bacteria and are composed of an outer membrane protein, a membrane fusion protein and an ABC (ATP-binding cassette) transporter. The ABC transporter HlyB (haemolysin B) is part of a T1SS catalysing the export of the toxin HlyA in E. coli. HlyB consists of the canonical transmembrane and nucleotide-binding domains. Additionally, HlyB contains an N-terminal CLD (C39-peptidase-like domain) that interacts with the transport substrate, but its functional relevance is still not precisely defined. In the present paper, we describe the purification and biochemical characterization of detergent-solubilized HlyB in the presence of its transport substrate. Our results exhibit a positive co-operativity in ATP hydrolysis. We characterized further the influence of the CLD on kinetic parameters by using an HlyB variant lacking the CLD (HlyB∆CLD). The biochemical parameters of HlyB∆CLD revealed an increased basal maximum velocity but no change in substrate-binding affinity in comparison with full-length HlyB. We also assigned a distinct interaction of the CLD and a transport substrate (HlyA1), leading to an inhibition of HlyB hydrolytic activity at low HlyA1 concentrations. At higher HlyA1 concentrations, we observed a stimulation of the hydrolytic activities of both HlyB and HlyB∆CLD, which was completely independent of the interaction of HlyA1 with the CLD. Notably, all observed effects on ATPase activity, which were also analysed in detail by mass spectrometry, were independent of the HlyA1 secretion signal. These results assign an interdomain regulatory role for the CLD modulating the hydrolytic activity of HlyB.

INTRODUCTION

Type 1 secretion systems (T1SS) are membrane-embedded transport machineries involved in the secretion of a wide range of substrates in Gram-negative bacteria. Substrate transport occurs in one step from the cytosol directly into the exterior, passing both the inner and outer membrane [1]. Substrates of T1SS include lipases, proteases, toxins and S-layer proteins, as well as haem-binding proteins [2]. The paradigm of T1SS is the haemolysin transport system from Escherichia coli, which transports HlyA (haemolysin A), a 110 kDa toxin produced by several uropathogenic E. coli strains [3,4]. HlyA belongs to the large RTX (repeats in toxins) protein family [5]. RTX proteins contain nonapeptide repeats, so-called GG-repeats, with the consensus sequence GGxGxDxUx, where x can be any amino acid residue and U represents a large hydrophobic amino acid [6]. These repeats are able to bind Ca2+ ions with micromolar affinity, inducing protein folding to build up a so-called β-roll motif [6,7]. The number of GG-repeats varies within the transported substrates and correlates in general with the size of the protein [8,9].

In general, T1SS substrates contain a non-cleavable C-terminal secretion signal that initiates the assembly of the whole transport complex and are essential and sufficient for protein secretion [1012]. In the case of HlyA, the secretion signal corresponds to the last 50–60 C-terminal amino acids [1315].

The T1SS of HlyA in E. coli is composed of three components and spans the periplasm from the inner membrane to the outer membrane. Two proteins are located in the inner membrane, the ABC (ATP-binding cassette) transporter HlyB (haemolysin B) and the MFP (membrane-fusion protein) HlyD (haemolysin D) [10,11,16]. Upon substrate interaction, the MFP recruits the third component of the secretion machinery, the OMP (outer membrane protein) TolC [12]. The functional complex transports the substrate protein in an unfolded state in one step across both membranes [1,11].

The general ABC transporter blueprint includes two hydrophobic TMDs (transmembrane domains) and two hydrophilic NBDs (nucleotide-binding domains). In the case of HlyB, there is an additional 123-amino-acid large N-terminal domain (Figure 1), which possesses a striking sequence and structural homology with C39-peptidases. These C39-peptidases are part of ABC transporters that usually mediate the transport of class II microcins. The proteolytically active N-terminal domain cleaves the N-terminal signal sequence of the precursor peptide of the transport substrate and thereby evolves the mature and active bacteriocin [17]. In contrast, the N-terminal domain of HlyB contains a degenerated catalytic triad and shows no proteolytic activity. Therefore the N-terminal domain of HlyB was called a CLD (C39-peptidase-like domain). Despite the lack of proteolytic activity, the CLD interacts specifically with unfolded HlyA and is indispensable for protein secretion [18]. Furthermore, it was proposed that the CLD tethers the substrate in a secretion-competent state [18], but its precise molecular function during secretion remains elusive.

Schematic model of the spatial orientation of the HlyB domains

Figure 1
Schematic model of the spatial orientation of the HlyB domains

HlyB is schematically shown as a dimer with the suggested spatial orientation of the TMDs, the NBDs and the CLD.

Figure 1
Schematic model of the spatial orientation of the HlyB domains

HlyB is schematically shown as a dimer with the suggested spatial orientation of the TMDs, the NBDs and the CLD.

In a previous study, HlyA was replaced by a C-terminal fragment (amino acids 807–1024) called HlyA1 [19] (Figure 2). This truncated protein contains the C-terminal secretion signal and three GG-repeats. It was demonstrated that HlyA1 is secreted with an efficiency comparable with that of HlyA [20]. Additionally, such HlyA fragments also bind Ca2+ ions with the same affinity leading to folding into a stable conformation as observed for full-length HlyA [19,21].

In the present paper, we describe the purification of full-length HlyB and its co-operative ATPase activity and provide strong evidence that the N-terminal domain of HlyB, the CLD, exhibits a novel regulatory role by inhibiting the ATPase activity. Furthermore, we demonstrate stimulation of the ATPase activity of HlyB in the presence of its transport substrate, which we show to be independent of the CLD and its secretion signal. We also analysed different cross-linking patterns of HlyB and its substrate by mass spectrometry, which correlate to stimulation and inhibition of HlyB ATPase activity. On the basis of our results, we propose a negative intermolecular regulation of ATPase activity of HlyB by the CLD. These findings represent a further step to understand the molecular principles of the HlyA Type 1 secretion process.

EXPERIMENTAL

Cloning of HlyB, HlyB∆CLD, HlyB-H662A and HlyB-Y20A

For cloning of the HlyB expression plasmid primers P1 and P2 (Table 1) were used to amplify hlyB (2124 bp) from the template plasmid pLG814 [14,22]. The HlyB sequence was subcloned into vector pET401 by using compatible and cohesive ends (EcoRI, BamHI). An N-terminal decahistidine tag with a protease cleavage site (Factor Xa) was introduced by addition of an oligonucleotide pair at the 5′ end of hlyB (primers P3 and P4; Table 1). Subsequently, the tagged hlyB gene was cloned into the expression vector pBADHisB using the NcoI (5′) and the SacI (3′) restriction site. The latter was obtained from subcloning into vector pET401. The resulting expression vector was named pBADHisHlyB. For investigation of HlyB lacking the N-terminal CLD, we created an expression plasmid with hlyB lacking the N-terminal 123 amino acids and termed this construct HlyB∆CLD. The construct was obtained using the In-Fusion Advantage PCR Cloning kit (Clontech) as recommended by the manufacturer. The plasmid pBADHisHlyB was linearized with primers P5 and P6 (Table 1) and hlyB was amplified from pBADHisHlyB using primers P7 and P8 (Table 1). After successful cloning, we obtained the expression plasmid pBADHisHlyB∆CLD. All expression constructs were used to express and subsequently purify HlyB or HlyB∆CLD in E. coli. To evaluate our purification method, we used an HlyB-H662A mutant, which is deficient in ATP hydrolysis [23]. A H662A mutation was inserted into the plasmids pBADHisHlyB and pBADHisHlyB∆CLD using the QuikChange® site-directed mutagenesis kit (Stratagene) following the manufacturer's instructions. The Y20A mutation was also introduced by QuikChange® site-directed mutagenesis following the manufacturer's instructions [18]. The correctness of all constructs was verified by sequencing analysis.

Schematic overview of the HlyA variants

Figure 2
Schematic overview of the HlyA variants

(A) HlyA with the putative membrane insertion domain (M, blue), two lysine residues which are acylated by HlyC (haemolysin C), six RTX repeats of the consensus sequence GGxGxDxUx (x is any amino acid residue; U is a large hydrophobic amino acid residue) (green) and a C-terminal secretion signal (red). (B) HlyA1, the C-terminal fragment of HlyA with three RTX repeats. (C) HlyA2, the C-terminal fragment of HlyA without the last 57 amino acid residues harbouring the secretion signal.

Figure 2
Schematic overview of the HlyA variants

(A) HlyA with the putative membrane insertion domain (M, blue), two lysine residues which are acylated by HlyC (haemolysin C), six RTX repeats of the consensus sequence GGxGxDxUx (x is any amino acid residue; U is a large hydrophobic amino acid residue) (green) and a C-terminal secretion signal (red). (B) HlyA1, the C-terminal fragment of HlyA with three RTX repeats. (C) HlyA2, the C-terminal fragment of HlyA without the last 57 amino acid residues harbouring the secretion signal.

Table 1
Overview of the primers used and their sequences
Name Sequence (5′→3′) 
P1 GCTATCCATGGCGAATTCTGATTCTTGTCATAAAATTGATTATGGG 
P2 CGTGTCTAGATTACAGGATCCCGTCTGACTGTAACTGATATAAGTAACTG 
P3 AATTCTCATCACCATCACCATCACCATCACCATCATAGCATCGAAGGGCGC 
P4 AATTGCGCCCTTCGATGCTATGATGGTGATGGTGATGGTGATGGTGATGAG 
P5 TCTAGAGCGGCCGCCACCG 
P6 ATGGTGATGGTGATGGTGATGAGAATTCGCCATGGT 
P7 CATCACCATCACCATCACCATCATGAGAATCTTTATTTTCAGGGCGCTTCCCGTTCTTCTGTTGC 
P8 GGCGGCCGCTCTAGATTACAG 
P9 CGCCTTGATGATGCGTTTTCACGGAATG 
P10 CATTCCGTGAAAACGCATCATCAAGGCG 
P11 CGCCTTGATGATTGGTTTTCACGGAATG 
P12 CATTCCGTGAAAACCAATCATCAAGGCG 
Name Sequence (5′→3′) 
P1 GCTATCCATGGCGAATTCTGATTCTTGTCATAAAATTGATTATGGG 
P2 CGTGTCTAGATTACAGGATCCCGTCTGACTGTAACTGATATAAGTAACTG 
P3 AATTCTCATCACCATCACCATCACCATCACCATCATAGCATCGAAGGGCGC 
P4 AATTGCGCCCTTCGATGCTATGATGGTGATGGTGATGGTGATGGTGATGAG 
P5 TCTAGAGCGGCCGCCACCG 
P6 ATGGTGATGGTGATGGTGATGAGAATTCGCCATGGT 
P7 CATCACCATCACCATCACCATCATGAGAATCTTTATTTTCAGGGCGCTTCCCGTTCTTCTGTTGC 
P8 GGCGGCCGCTCTAGATTACAG 
P9 CGCCTTGATGATGCGTTTTCACGGAATG 
P10 CATTCCGTGAAAACGCATCATCAAGGCG 
P11 CGCCTTGATGATTGGTTTTCACGGAATG 
P12 CATTCCGTGAAAACCAATCATCAAGGCG 

Protein purification of HlyB, HlyB∆CLD, HlyB-H662A and HlyB-Y20A

E. coli BL21(DE3) cells were transformed with pBADHlyB and exposed on selective agar plates containing 100 μg/ml ampicillin. A selective overnight culture was inoculated with a single E. coli colony and incubated for 16 h at 37°C and vigorously shaking at 200 rev./min. A main culture with selective 2YT medium [1.6% (w/v) tryptone, 1% (w/v) yeast extract and 0.5% NaCl] was inoculated from the overnight culture and grown to a OD600 of 4.0 before expression was started by addition of 10 mM arabinose. After 2 h of incubation, cells were harvested by centrifugation and resuspended in buffer A (25 mM NaH2PO4, 100 mM KCl and 20% glycerol, pH 8). Bacteria were disrupted by multiple passes through a cell disrupter (Constant Systems) at 2.5 kbar (1 bar=100 kPa). The homogenate was centrifuged at 12000 g and the supernatant was centrifuged again at 120000 g. The pellet containing cell membranes was resuspended in buffer A, solubilized with 1% Fos-Choline-14 at 4°C for 1 h and centrifuged again at 120000 g for 30 min. The supernatant was loaded on an IMAC (immobilized metal-ion-affinity chromatography) column [5 ml HiTrap Chelating HP column (GE Healthcare) loaded with ZnSO4] and washed with buffer A including 0.02% LMNG (lauryl maltose neopentyl glycol). Non-specifically bound protein was removed by washing with buffer A including 0.02% LMNG and 40 mM histidine. HlyB was eluted with buffer A including 0.02% LMNG and 150 mM histidine. Fractions containing HlyB were pooled and concentrated by ultrafiltration using an Amicon Ultra centrifugal filter unit (100000 Da molecular-mass cut-off; Merck Millipore). To increase the protein purity and to change the buffer for long-term protein stability, SEC (size-exclusion chromatography) was performed using a Superdex 200 10/300 GL column (GE Healthcare) using buffer B (10 mM Caps and 20% glycerol, pH 10.4) including 0.02% LMNG [24]. HlyB-containing fractions were flash-frozen in liquid nitrogen and stored at −80°C until further use. HlyB∆CLD, HlyB-H662A and HlyB-Y20A were purified accordingly.

Protein purification of HlyA1 and HlyA2

The HlyA derivatives HlyA1 (amino acids 807–1024; 24.9 kDa) and HlyA2 (amino acids 807–966; 17.7 kDa) were purified as described previously [19] with the following modifications. For overexpression of HlyA1, E. coli BL21(DE3) cells were transformed with pSOI-HlyA1 and exposed on selective 2YT agar plates containing 100 μg/ml ampicillin. A selective overnight culture was inoculated with a single E. coli colony and incubated for 16 h at 37°C and vigorously shaken at 200 rev./min. A main culture with selective 2YT medium was inoculated from the overnight culture and grown to a OD600 of 1.0 and induced by the addition of 1 mM IPTG. After incubation for 3 h at 37°C, cells were removed by centrifugation at 5800 g for 10 min. HlyA1 was purified from the medium supernatant. Supernatant from 200 ml of medium was concentrated to a final volume of 2 ml using an Amicon Ultra centrifugal filter unit (10000 Da molecular-mass cut-off; Merck Millipore) and subjected to SEC using a HiLoad Superdex 75 16/600 prep grade column (GE Healthcare), equilibrated with 100 mM HEPES (pH 8) and 250 mM NaCl. Protein-containing fractions were pooled, concentrated to 4 mg/ml and stored at −20°C.

For the purification of HlyA2, E. coli BL21(DE3) cells were transformed with pSOI-HlyA2 and expressed as described above for HlyA1. The main culture was grown to a OD600 of 1.0 and induced with 4 mM arabinose. After incubation for 90 min at 37°C, cells were harvested by centrifugation at 5800 g for 10 min. The cell pellet was resuspended in buffer C (10 mM Tris/HCl, 100 mM KCl, 0.2 mM EGTA and 10% glycerol, pH 8) and homogenized by three passages through a cell disruptor at 2.5 kbar (Constant Systems). The cell lysate was centrifuged at 120000 g for 45 min at 4°C and the supernatant was loaded on an IMAC column [5 ml HiTrap Chelating HP column (GE Healthcare) loaded with NiSO4] equilibrated with buffer C containing 10 mM imidazole. Non-specifically bound protein was washed off the column with buffer C containing 25 mM imidazole and HlyA2 was eluted with buffer containing 500 mM imidazole. Fractions containing HlyA2 were pooled and diluted 15-fold in buffer D (10 mM Tris/HCl, 10 mM KCl, 0.2 mM EGTA and 10% glycerol, pH 8). Further purification of HlyA2 was performed via anion-exchange chromatography using a HiTrap Q HP column (GE Healthcare) equilibrated with buffer D. HlyA2 was eluted with a KCl gradient from 10 mM to 500 mM over a volume of 200 ml. HlyA2-containing fractions were pooled and the buffer was exchanged with 10 mM CAPS (pH 10.4) via a PD-10 desalting column (GE Healthcare). The HlyA2 His tag was removed by TEV (tobacco etch virus) protease. The purification of the TEV protease was performed as described in [25]. HlyA2 was digested for 16 h in 100 mM HEPES with 150 mM NaCl (pH 8). The residual His tag and TEV protease were removed by an IMAC step and untagged HlyA2 was collected, concentrated using an Amicon Ultra centrifugal filter unit (10000 Da molecular-mass cut-off; Merck Millipore) and stored at −80°C.

Electrophoresis and immunological technique

SDS/PAGE and subsequent CBB (Coomassie Brilliant Blue) staining were used to analyse proteins. The immunodetection was performed with anti-HlyB antibody rabbit polyclonal serum at 1:8000 dilution in TBS-T (TBS with Tween 20: 20 mM Tris base, 300 mM NaCl and 0.05% Tween 20, pH 8).

Solubilization screen via the dot-blot technique

Membranes were thawed on ice and the protein concentration was adjusted to 5 mg/ml. The solubilization was performed at 4°C for 1 h with gentle mixing. The detergents were used at a concentration of 1% (w/v) or higher, depending on their CMC (critical micellar concentration). After solubilization, the samples were centrifuged for 30 min at 100000 g and 4°C. The supernatant was supplemented with SDS sample buffer and heated to 65°C for 10 min. Subsequently 3 μl of the probe was spotted on a nitrocellulose membrane and dried for 12 h. The membrane surface was blocked by incubation for 3 h in TBS-T supplemented with 5% (w/v) non-fat dried skimmed milk powder. The target protein was detected by immune detection using a polyclonal anti-HlyB antibody.

ATPase assays

The Malachite Green assay was used for quantification of hydrolytic activity as described in [26] with the following modifications. Reactions were performed in 100 mM HEPES (pH 7) containing 20-fold CMC of LMNG detergent and ATP concentrations ranging from 0 mM to 5 mM. HlyB and HlyB∆CLD were used at 1 μM concentration and diluted into assay buffer. Reactions were started by the addition of 10 mM MgCl2, incubated at 25°C and stopped after 30 min by transferring 25 μl of the reaction volume into 175 μl of 20 mM H2SO4. Free inorganic phosphate (Pi) was stained by adding 50 μl of dye solution [0.096% Malachite Green, 1.48% (w/v) ammonium molybdate and 0.173% Tween 20 in 2.36 M H2SO4] to the quenched reaction volume. Quantification was performed spectroscopically 10 min after adding dye solution by measuring the absorbance at 595 nm. For data evaluation, all appropriate controls were subtracted. Data points were fitted using Prism software (GraphPad) to one of the following equations.

Eqn (1), the Hill equation, is:

 
formula
1

Here, v corresponds to the ATPase activity as a function of the substrate concentration [S], Vmax is the maximum enzyme activity, h is the Hill coefficient, and K0.5 is the substrate concentration at which 50% enzyme binding sites are occupied.

Eqn (2), which assumes two independent binding sites, is:

 
formula
2

Here, v corresponds to the ATPase activity as a function of the substrate concentration [S]. K1 represents the substrate concentration of half-maximum inhibition, K2 represents the substrate concentration of half-maximum activation of ATPase activity, v0 is the basal activity of HlyB in the absence of the substrate, v1 is the minimal enzyme activity and v2 is the maximum enzyme activity.

Eqn (3), the classic Michaelis–Menten equation, is:

 
formula
3

Here, v corresponds to the ATPase activity as a function of the substrate concentration [S], Vmax is the maximum enzyme activity, and Km is the Michaelis–Menten constant.

Substrate-modulated kinetics were performed as described above, but the reaction buffer contained additionally up to 20 μM HlyA1 or HlyA2. ATPase assays with folded substrate also included 4 mM CaCl2 in the assay buffer. Appropriate controls confirmed that HlyB hydrolytic behaviour is not influenced by the assay buffer containing 4 mM CaCl2.

In vitro cross-linking of HlyB and HlyA2

Reactions were performed in 100 mM HEPES (pH 7) containing 20-fold CMC of LMNG detergent and 2 mM ATP. HlyB was diluted in assay buffer to a final concentration of 1 μM. Additionally HlyA2 was added to a final concentration of 1 μM or 20 μM respectively. Reactions were started by the addition of 10 mM MgCl2. After incubation at 25°C for 30 min, proteins were cross-linked by employing BS(PEG)5 [bis-N-succinimidyl-penta(ethylene glycol) ester] (Thermo Fisher Scientific), which is a homo-bifunctional lysine-specific cross-linker with a maximal arm length of 21.7 Å (1 Å=0.1 nm). BS(PEG)5 was used at a final concentrations of 0.8 mM (1 μM HlyA2) or 3.6 mM (20 μM HlyA2) respectively. The reaction mixture was incubated further at 25°C for 1 h and finally quenched by adding 50 mM Tris/HCl (pH 7). Samples were analysed further by SDS/PAGE, LC and MS.

Liquid chromatography and mass spectrometry

Three independent preparations of cross-linked ‘inhibition’ (1 μM HlyA2) or ‘stimulation’ (20 μM HlyA2) samples were analysed by LC–MS.

Samples were separated by SDS/PAGE (10% gel) and, after staining with CBB, protein-containing bands were excised and processed as described in [27]. Briefly, bands were washed, reduced, alkylated with iodoacetamide and digested overnight at 37°C with 0.05 μg of trypsin (Serva) in 50 mM NH4HCO3. After peptide extraction with 1:1 (v/v) 0.1% TFA (trifluoroacetic acid)/acetonitrile and vacuum concentration, peptides were resuspended in 0.1% TFA.

First, peptides were separated by reverse-phase LC on an UltiMate 3000 RSLCnano system (Thermo Scientific). Here, peptides were initially pre-concentrated on a trap column (Acclaim PepMap100, 3 μm C18 particle size, 100 Å pore size, 75 μm inner diameter, 2 cm length, Thermo Scientific) for 10 min at a flow rate of 6 μl/min and subsequently separated at a flow rate of 300 nl/min on an analytical column (Acclaim PepMapRSLC, 2 μm C18 particle size, 100 Å pore size, 75 μm inner diameter, 25 cm length, Thermo Scientific) using 0.1% TFA as the mobile phase and increasing concentrations of acetonitrile for 1 h.

In a second step, separated peptides were analysed with a Q Exactive hybrid quadrupole-Orbitrap mass spectrometer coupled to an LC system via a nanoelectrospray ionization source equipped with distal coated SilicaTip emitters (New Objective).

The mass spectrometer was operated in positive data-dependent mode, a spray voltage of 1400 V was applied and the capillary temperature set to 250°C. First, full scans were recorded with a resolution of 70000 over a scan rage from 200 to 2000 m/z in the Orbitrap analyser in profile mode with a maximum ion time of 50 ms and the target value for the automatic gain control set to 3000000. Subsequently, up to 20 precursors (charge states 2–5) were isolated within a 2 m/z isolation window. Isolated precursors were individually fragmented by HCD (higher-energy collisional dissociation) in the HCD cell of the instrument and MS/MS spectra were recorded within the Orbitrap analyser with a maximal ion time of 50 ms and the target value for the automatic gain control set to 100000. MS/MS spectra were acquired over an available scan range of 200 to 2000 m/z at a resolution of 17500. Already fragmented precursors were excluded from further isolation for the next 10 s.

To initially identify HlyB and HlyA2 peptide sequences, Proteome Discoverer (version 1.1.1.14, Thermo Scientific) was used to trigger searches by MS Amanda in a database consisting of 787 E. coli K12 entries downloaded on 14 October 2015 and additional sequences of HlyB and HlyA2. Searches were conducted with tryptic cleavage specificity with a maximal number of two missed cleavages, a mass tolerance of 5 p.p.m. for precursor and 20 p.p.m. for fragment spectra. Carbamidomethyl at cysteine residues was set as fixed modification and methionine oxidation, acetylation at protein N-termini and semihydrolysed BS(PEG)5 linker attached to lysine residues (+320.1471 Da) were considered as dynamic modifications.

For the detection of cross-linked peptides, StavroX version 3.5.1 [28] was applied for mapping spectra against HlyB and HlyA2 sequences. Here, tryptic cleavage specificity was considered with a maximum of one missed cleavage after arginine residues and two missed cleavages after lysine residues. Carbamidomethyl at cysteine residues was set as fixed and oxidation of methionine as variable modification. A reaction of the cross-linker was considered between lysine residues as well as N-termini between peptides, within peptides and as dead-end linker after reaction with water. The precursor precision was set to 10 p.p.m., the fragment precision to 20 p.p.m. and a slow precise scoring was applied. For individual searches, the cut-off scores were 41–55 for detecting decoy hits at a 5% false discovery rate and 52–67 at a 1% false discovery rate.

Interesting candidate peptides and dipeptides were evaluated further using the Thermo Xcalibur 3.0.63 Qual Browser (Thermo Scientific) by manual inspection of spectra and quantification of peptide intensities by generating extracted ion chromatograms from precursor spectra within a mass window of 10 p.p.m.

Secretion analysis of HlyA in the presence of HlyB mutants

Plasmid pK184-HlyBD was mutated by site-directed mutagenesis to insert single point mutations in hlyB. Appropriate oligonucleotides (primers P9–P12; Table 1) were used to replace Lys322 of HlyB with alanine or tryptophan respectively. The correctness of all constructs was verified by sequencing analysis.

Chemically competent E. coli BL21(DE3) cells were transformed with pK184-HlyBD and pSU-HlyA and grown on selective 2YT agar plates containing 100 μg/ml ampicillin and 30 μg/ml kanamycin. A selective overnight culture was inoculated with a single E. coli colony and incubated for 16 h at 37°C and vigorously shaken at 200 rev./min. A main culture with selective 2YT medium was inoculated from the overnight culture and grown to a OD600 of 1.0. The expression of HlyA, HlyB and HlyD was induced with 1 mM IPTG, and CaCl2 was added to the medium at a final concentration of 5 mM. Cells were grown for 2 h at 200 rev./min at 37°C. Subsequently a 1 ml aliquot was taken and centrifuged at 14000 g for 5 min. Cells in the aliquot were adjusted with water to OD600 equivalents of 0.1. The expression level of HlyB was determined via Western blotting using a polyclonal anti-HlyB antibody. To determine the amount of secreted HlyA, aliquot supernatants were adjusted correlating to a OD600 of 3 and 15 μl were analysed via CBB-stained SDS/PAGE gels.

RESULTS

Cloning, expression and purification of the ABC transporter HlyB

We cloned hlyB (2124 bp) from the template plasmid pLG814 into the expression vector pBADHisB [14,22]. The construct encoded HlyB with a theoretical total mass of 82.2 kDa including an N-terminal decahistidine tag. For investigation of HlyB lacking the N-terminal CLD, we also created an expression plasmid encoding HlyB lacking the N-terminal 123 amino acids and termed this construct HlyB∆CLD (67.9 kDa). We also used the HlyB-H662A and HlyB-Y20A mutants, which display a deficiency in ATP hydrolysis [23] or strongly reduced substrate binding [18,23] respectively.

All HlyB derivatives described above were expressed and purified using the same protocol. To identify suitable detergents for the solubilization of HlyB, a dot-blot detergent screen [29] was used highlighting that Fos-Choline-14 was one of the few detergents appropriate for solubilizing HlyB from the membranes (Supplementary Figure S1 and Supplementary Table S1). On the basis of the results obtained from the dot-blot analysis, isolated E. coli membranes were adjusted to 50 mg/ml and solubilized by addition of 1% (w/v) Fos-Choline-14 with gentle agitation at 4°C for 1 h. To separate solubilized His-tagged HlyB from solubilized contaminants and to exchange the detergent, we used a IMAC column (loaded with ZnSO4) equilibrated with buffer A containing 0.02% LMNG. HlyB solubilized in Fos-Choline-14 showed no ATPase activity, therefore we exchanged the detergent to improve HlyB homogeneity as observed by SEC and to retain its ability to hydrolyse ATP. Fractions containing HlyB as judged from a CBB-stained SDS/PAGE gel were pooled, concentrated and subjected to SEC using buffer B containing 0.02% LMNG (Figure 3). After this purification step, approximately 1.5 mg of homogeneous HlyB per litre of bacterial culture was obtained.

SEC profile, SDS/PAGE and Western blot analysis of purified HlyB and HlyB∆CLD

Figure 3
SEC profile, SDS/PAGE and Western blot analysis of purified HlyB and HlyB∆CLD

(A) HlyB protein was concentrated using an Amicon Ultra centrifugal filter unit and 500 μl (3 mg/ml) was loaded on to a Superdex 200 10/300 GL column. The buffer contained 10 mM Caps, 20% glycerol and 0.02% LMNG (pH 10.4). The arrow indicates the void volume of the column. mAU, milli-absorbance units. (B) Purified HlyB or (C) HlyB∆CLD was analysed by (I) SDS/PAGE and stained with CBB or detected by (II) Western blot analysis using polyclonal antisera. Including the decahistidine tag, HlyB has a theoretical mass of 82.2 kDa and HlyB∆CLD has a theoretical mass of 67.9 kDa.

Figure 3
SEC profile, SDS/PAGE and Western blot analysis of purified HlyB and HlyB∆CLD

(A) HlyB protein was concentrated using an Amicon Ultra centrifugal filter unit and 500 μl (3 mg/ml) was loaded on to a Superdex 200 10/300 GL column. The buffer contained 10 mM Caps, 20% glycerol and 0.02% LMNG (pH 10.4). The arrow indicates the void volume of the column. mAU, milli-absorbance units. (B) Purified HlyB or (C) HlyB∆CLD was analysed by (I) SDS/PAGE and stained with CBB or detected by (II) Western blot analysis using polyclonal antisera. Including the decahistidine tag, HlyB has a theoretical mass of 82.2 kDa and HlyB∆CLD has a theoretical mass of 67.9 kDa.

HlyB ATPase modulation by its transport substrate

ATPase activity of HlyB was monitored by the quantification of released Pi using a Malachite Green assay. ATP was applied at concentrations ranging from 0 to 5 mM while the HlyB concentration was kept constant at 1 μM. Results of the ATPase activity measurements are summarized in Figure 4(A) and followed a non-linear dependence on ATP concentration. The Hill equation (eqn 1) was used to analyse the results. According to this equation, the maximum velocity (Vmax) of the reaction was calculated to be 8.1±0.7 nmol of Pi/min per mg, the Hill coefficient h was 1.5±0.4 and the kinetic constant K0.5 was 0.3±0.1 mM ATP (Table 2). The value for the Hill coefficient indicates positive co-operativity of ATP hydrolysis as observed previously for the isolated HlyB NBD [30]. An HlyB mutant defective in ATP hydrolysis (HlyB H662A) showed no hydrolytic activity under the experimental conditions demonstrating that the observed activity was derived from HlyB [23]. The co-operative nature of ATPase activity was also supported by data analysis using a double logarithmic evaluation {‘Hill plot’, log[v/(Vmaxv)] against logATP (Figure 4B)}. Importantly, the value of the Hill coefficient was identical within experimental error for both evaluation procedures.

ATPase activity of purified HlyB and HlyB∆CLD

Figure 4
ATPase activity of purified HlyB and HlyB∆CLD

(A) ATPase activity of HlyB (circles) and HlyB∆CLD (triangle) in dependency of ATP concentration. Both datasets were analysed using the Hill equation (eqn 1). The plots are additionally labelled with schematic representations of HlyB and HlyB∆CLD. Hill plots of the data are shown for HlyB (B) and HlyB∆CLD (C). Data were also fitted according to the Michaelis–Menten equation (eqn 3), broken lines, or the Hill equation (eqn 1), unbroken lines. Results are means±S.D. from at least two independent experiments.

Figure 4
ATPase activity of purified HlyB and HlyB∆CLD

(A) ATPase activity of HlyB (circles) and HlyB∆CLD (triangle) in dependency of ATP concentration. Both datasets were analysed using the Hill equation (eqn 1). The plots are additionally labelled with schematic representations of HlyB and HlyB∆CLD. Hill plots of the data are shown for HlyB (B) and HlyB∆CLD (C). Data were also fitted according to the Michaelis–Menten equation (eqn 3), broken lines, or the Hill equation (eqn 1), unbroken lines. Results are means±S.D. from at least two independent experiments.

Table 2
Kinetic parameters of HlyB, HlyBΔCLD and HlyB-Y20A

(HlyA1), kinetic parameters were determined in dependence of HlyA1 concentrations. (HlyA2), kinetic parameters were determined in dependence of HlyA2 concentrations. (HlyA1, folded), kinetic parameters were determined in the presence of HlyA1 and Ca2+ ions (see the Materials and methods section).

(a) 
 K0.5 (mM) Vmax (nmol/min per mg) h Slope (‘Hill plot’)§ kcat (min−1
HlyB* 0.3±0.1 8.1±0.7 1.5±0.4 1.3±0.1 0.7±0.1 
HlyBΔCLD* 0.40±0.02 124.3±3.1 1.5±0.1 1.4±0.1 8.4±0.2 
(b) 
 Km/K0.5 (μM) Vmax (nmol/min per mg) h Slope (‘Hill plot’)§ 
HlyBΔCLD (HlyA1)† 1.2±0.3 151.6±2.3 (+33%) – – 
HlyBΔCLD (HlyA2)† 2.0±0.3 154.4±1.3 (+29%) – – 
HlyB-Y20A (HlyA1)* 0.9±0.2 9.1±0.2 (+20%) 1.5±0.4 1.4±0.4 
HlyB (HlyA1, folded)* 2.0±0.3 9.5±0.2 (+33%) 1.8±0.4 1.6±0.1 
(c) 
 K1 (μM) K2 (μM) v0 (nmol/min per mg) v1 (nmol/min per mg) v2 (nmol/min per mg) 
HlyB (HlyA1)‡ 1.0±0.5 19.4±3.1 13.9±0.4 8.8±0.9 (–37%) 28.4±0.1 (+104%) 
HlyB (HlyA2)‡ 0.4±0.1 4.7±0.3 17.9±0.2 16.6±0.2 (–7%) 21.7±0.2 (+21%) 
(a) 
 K0.5 (mM) Vmax (nmol/min per mg) h Slope (‘Hill plot’)§ kcat (min−1
HlyB* 0.3±0.1 8.1±0.7 1.5±0.4 1.3±0.1 0.7±0.1 
HlyBΔCLD* 0.40±0.02 124.3±3.1 1.5±0.1 1.4±0.1 8.4±0.2 
(b) 
 Km/K0.5 (μM) Vmax (nmol/min per mg) h Slope (‘Hill plot’)§ 
HlyBΔCLD (HlyA1)† 1.2±0.3 151.6±2.3 (+33%) – – 
HlyBΔCLD (HlyA2)† 2.0±0.3 154.4±1.3 (+29%) – – 
HlyB-Y20A (HlyA1)* 0.9±0.2 9.1±0.2 (+20%) 1.5±0.4 1.4±0.4 
HlyB (HlyA1, folded)* 2.0±0.3 9.5±0.2 (+33%) 1.8±0.4 1.6±0.1 
(c) 
 K1 (μM) K2 (μM) v0 (nmol/min per mg) v1 (nmol/min per mg) v2 (nmol/min per mg) 
HlyB (HlyA1)‡ 1.0±0.5 19.4±3.1 13.9±0.4 8.8±0.9 (–37%) 28.4±0.1 (+104%) 
HlyB (HlyA2)‡ 0.4±0.1 4.7±0.3 17.9±0.2 16.6±0.2 (–7%) 21.7±0.2 (+21%) 

*Parameters were determined according to eqn (1), the Hill equation.

†Parameters were determined according to eqn (3), the Michaelis–Menten equation.

‡Parameters were determined according to eqn (2) which assumes two independent binding sites.

§Parameters were determined according to a double logarithmic evaluation (‘Hill plot’, log[v/(Vmaxv)] against logATP.

Next, we examined whether the transport substrate influences the hydrolytic activity of HlyB. Instead of HlyA we employed HlyA1 because of its higher stability and ease of purification [19] (Figure 2). ATPase activity was measured at a constant HlyB concentration of 1 μM and a constant ATP concentration of 2 mM, while the HlyA1 concentration varied from 0 μM to 20 μM. The applied concentration of ATP was approximately 6-fold above the determined K0.5 value and ensured that HlyB was always saturated under these conditions. We observed both an inhibitory and a stimulatory effect of unfolded HlyA1 on HlyB ATPase activity (Figure 5). The two opposite effects depended on substrate concentration suggesting different modes of interaction between HlyA1 and HlyB, since no hydrolytic activity of HlyA1 was observed. The data were analysed according to eqn (2) assuming two independent binding sites [31]. The first binding site inhibits ATPase activity by 37% and the second binding site stimulates hydrolytic activity of HlyB by 104% (Table 2). However, higher concentrations of HlyA1 were not possible to measure due to aggregation of the protein. Therefore the K2 value is only an estimation and represents a lower limit of the interaction (Table 2). We visualized the inhibitory as well as the stimulatory effect independently of each other by simulations (Figure 5). These simulations revealed a maximum inhibition of 28% and a maximum stimulation of 52%. A simple addition of both curves overlapped well with our experimental results and indicates that our assumption of inhibition at low substrate concentration, which is followed by stimulation at higher substrate concentrations, is valid.

Modulation of ATPase activity of HlyB by unfolded HlyA1 and simulations of the inhibitory and stimulatory parts of ATPase activity

Figure 5
Modulation of ATPase activity of HlyB by unfolded HlyA1 and simulations of the inhibitory and stimulatory parts of ATPase activity

Experimental results of HlyB ATPase activity in the presence of unfolded substrate HlyA1 are shown in black (circles). Datasets were analysed according to eqn (2). Results are means±S.D. from at least two independent experiments. A simulation of the inhibitory influence of HlyA1 on HlyB ATPase activity is shown in green, whereas a simulation of the stimulatory part is shown in blue. The red curve represents the summation of both effects.

Figure 5
Modulation of ATPase activity of HlyB by unfolded HlyA1 and simulations of the inhibitory and stimulatory parts of ATPase activity

Experimental results of HlyB ATPase activity in the presence of unfolded substrate HlyA1 are shown in black (circles). Datasets were analysed according to eqn (2). Results are means±S.D. from at least two independent experiments. A simulation of the inhibitory influence of HlyA1 on HlyB ATPase activity is shown in green, whereas a simulation of the stimulatory part is shown in blue. The red curve represents the summation of both effects.

The N-terminal CLD of HlyB inhibits ATPase activity

Next we investigated the role of the N-terminal CLD on the ATPase activity of HlyB. We used HlyB∆CLD lacking the N-terminal domain and measured its ATPase activity by quantifying released Pi by a Malachite Green assay as described above. The results of the ATPase activity measurements are summarized in Figure 4 and followed a non-linear dependence of activity on ATP concentration similar to HlyB. The Hill equation (eqn 1) was used to fit the data. The maximum velocity (Vmax) of the reaction was calculated to be 124.3±3.1 nmol of Pi/min per mg, the Hill coefficient h was 1.5±0.1 and kinetic constant K0.5 was 0.40±0.02 mM (Table 2). Again, data analysis using a double logarithmic evaluation {‘Hill plot’, log[v/(Vmaxv)] against logATP (Figure 4C)} supported the co-operative nature of ATP hydrolysis with the value of the Hill coefficient being identical within experimental error for both evaluation procedures. In comparison with HlyB, the lack of the CLD (HlyB∆CLD) resulted in a 15-fold increased Vmax value, whereas the K0.5 value and the Hill coefficient remained identical within experimental error for both. These results indicate an intermolecular regulatory role of the CLD, which reduces the NBDs’ basal hydrolytic activity.

Furthermore, we also examined for HlyB∆CLD whether the transport substrate HlyA1 influences hydrolytic activity. The ATPase activity profile followed Michaelis–Menten kinetics and was fitted according to eqn (3) (Figure 6A, black circles). Remarkably, the co-operative nature of ATPase activity of HlyB∆CLD was abolished in the presence of unfolded HlyA1. The absence of co-operativity was supported by data analysis using a double-logarithmic evaluation {‘Hill plot’, log[v/(Vmaxv)] against logATP (Figure 6B)}. Analysis of the data revealed an increase of 33% in reaction velocity in comparison with HlyB∆CLD in the absence of the transport substrate (Table 2). For HlyB∆CLD, we observed only stimulation of ATPase activity, but an inhibition as in the case of HlyB (Figure 5) was not detected. Both sets of experiments were performed under identical conditions and differed only in the presence (HlyB) or absence (HlyB∆CLD) of the CLD. Therefore differences in the observed activity should originate from an interaction of the unfolded transport substrate HlyA1 with the CLD. These results emphasize further the regulatory role of the CLD and demonstrate that the observed stimulatory effects are independent of the CLD.

ATPase activity of HlyB, HlyB∆CLD and HlyB-Y20A in the presence of unfolded or folded HlyA1 respectively

Figure 6
ATPase activity of HlyB, HlyB∆CLD and HlyB-Y20A in the presence of unfolded or folded HlyA1 respectively

(A) Datasets were analysed according to the Michaelis–Menten equation (eqn 3) for results of HlyB∆CLD in the presence of unfolded HlyA1 (black circles). The Hill equation (eqn 1) was used to analyse the results of HlyB-Y20A in the presence of unfolded HlyA1 (red triangles) and HlyB in the presence of folded HlyA1 (blue squares, folding induced by adding 4 mM CaCl2). Hill plots of the data are shown for HlyB∆CLD (B), HlyB-Y20A (C) and HlyB (D) and were fitted according to the Michaelis–Menten equation (eqn 3), broken lines, or Hill equation (eqn 1), unbroken lines. Results are means±S.D. from at least two independent experiments.

Figure 6
ATPase activity of HlyB, HlyB∆CLD and HlyB-Y20A in the presence of unfolded or folded HlyA1 respectively

(A) Datasets were analysed according to the Michaelis–Menten equation (eqn 3) for results of HlyB∆CLD in the presence of unfolded HlyA1 (black circles). The Hill equation (eqn 1) was used to analyse the results of HlyB-Y20A in the presence of unfolded HlyA1 (red triangles) and HlyB in the presence of folded HlyA1 (blue squares, folding induced by adding 4 mM CaCl2). Hill plots of the data are shown for HlyB∆CLD (B), HlyB-Y20A (C) and HlyB (D) and were fitted according to the Michaelis–Menten equation (eqn 3), broken lines, or Hill equation (eqn 1), unbroken lines. Results are means±S.D. from at least two independent experiments.

To confirm a distinct interaction of the CLD with HlyA1 as the source of the observed negative regulation of the ATPase activity, we used the HlyB-Y20A mutant. It was already shown that the CLD interacts with unfolded HlyA1 and CSP (chemical shift perturbation) experiments suggested that Tyr20 is part of the HlyA-binding region [18]. In vivo secretion experiments with the HlyB-Y20A mutant confirmed this hypothesis as the mutation resulted in a significantly reduced secretion level when compared with HlyB. ATPase assays were performed with 1 μM HlyB-Y20A and 2 mM ATP, while the HlyA1 concentration varied from 0 μM to 20 μM. Results of the ATPase activity measurements are summarized in Figure 6(A) (red triangles) and the best fit of the data was obtained for the Hill equation (eqn 1). The value of the Hill coefficient was supported by data analysis using a double-logarithmic evaluation {‘Hill plot’, log[v/(Vmaxv)] against logATP (Figure 6C)}. HlyB-Y20A showed only ATPase stimulation by 20% in comparison with HlyB-Y20A in the absence of the unfolded substrate (Table 2). This general behaviour correlates with the results obtained from HlyB∆CLD. No ATPase inhibition was observed in the Y20A background, suggesting that the interaction of the mutated CLD and the substrate does not take place. This observation is in line with our previous results [18] and confirmed a CLD–HlyA1 interaction resulting in an inhibition of ATPase activity of HlyB.

Substrate folding prohibits HlyB ATPase inhibition

GG-repeats are characteristic features of RTX domains and folding experiments demonstrated that these GG-repeats trigger folding of the substrate by binding of Ca2+ [8]. Since HlyA is transported in an unfolded state, we assumed that a folded substrate is incompatible with an interaction with HlyB [1,32], as already demonstrated by pull-down experiments with the purified CLD [18]. In the present study, we examined whether folding of the substrate also induces an inhibition or stimulation of the ATPase activity of HlyB. Therefore substrate was folded by pre-incubating HlyA1 with 4 mM Ca2+ ions and ATPase assays of HlyB in the presence of folded HlyA1 were performed as described above. Data were analysed using the Hill equation (eqn 1) and the results are summarized in Figure 6(A) (blue squares). Again, data analysis using a double-logarithmic evaluation {‘Hill plot’, log[v/(Vmaxv)] against logATP (Figure 6D)} supported the value of the Hill coefficient. We determined a stimulation of 33% in comparison with the activity without the folded transported substrate (Table 2). In comparison with the ATPase assay in the presence of unfolded HlyA1 (Figure 5), no inhibition was observed (Figure 6A). Thus the CLD–HlyA1 interaction, the inhibition, does not take place if the transport substrate adopts a tertiary structure. However, the stimulation is independent of the folding state of HlyA1.

HlyB ATPase modulation occurs independently of the HlyA secretion signal

The observed stimulation of ATPase activity was independent of the CLD. We therefore examined whether the secretion signal of HlyA1 represents the interaction platform that modulates hydrolytic activity of HlyB in an inhibitory or stimulatory way. We used HlyA2 (Figure 2), which corresponds to HlyA1 but lacks the C-terminal 57 amino acid residues and therefore the secretion sequence. We abandoned measuring the influence of folded HlyA2 because of the missing extreme C-terminus of HlyA2. It has a major influence on protein folding [33] and it is suggested to stabilize the whole protein and especially the RTX domain [21]. The resulting data were analysed using eqns (2) and (3) for the data obtained for HlyB and HlyB∆CLD respectively (Figure 7). For HlyB∆CLD only, stimulation by 29% was observed following Michaelis–Menten kinetics (Table 2). For HlyB in the presence of HlyA2, the velocity was inhibited by 7% with 1.5 μM HlyA2 and stimulated by 21% with 20 μM HlyA2 (Table 2). Taken together, the secretion signal of the transport substrate has no influence on the modulation of ATPase inhibition and stimulation in principal, either in the case of HlyB or in the case of HlyB∆CLD. However, we observed a variation of HlyB kinetic parameters depending on the presence of HlyA1 or HlyA2 (Table 2). In general, the substrate affinities in the presence of HlyA2 are higher (K1=0.4±0.1 μM, K2=4.7±0.3 μM) and the impact on ATPase inhibition (v1) or ATPase stimulation (v2) is lower in comparison with HlyA1.

ATPase activity modulation of HlyB and HlyB∆CLD by unfolded HlyA2 (substrate without the secretion signal)

Figure 7
ATPase activity modulation of HlyB and HlyB∆CLD by unfolded HlyA2 (substrate without the secretion signal)

Datasets were analysed using eqn (2) in the case of HlyB (black circles) and the Michaelis–Menten equation (eqn 3) in the case of HlyB∆CLD (blue triangles). Results are means±S.D. from at least two independent experiments. The plots are additionally labelled with schematic representations of HlyB and HlyB∆CLD.

Figure 7
ATPase activity modulation of HlyB and HlyB∆CLD by unfolded HlyA2 (substrate without the secretion signal)

Datasets were analysed using eqn (2) in the case of HlyB (black circles) and the Michaelis–Menten equation (eqn 3) in the case of HlyB∆CLD (blue triangles). Results are means±S.D. from at least two independent experiments. The plots are additionally labelled with schematic representations of HlyB and HlyB∆CLD.

Overall, both substrates showed the same general influence concerning inhibition and stimulation, whereas the absence of the CLD abolished ATPase inhibition. Therefore we conclude that the modulation of ATPase activity occurs independently of the secretion signal.

HlyB and its substrate interact concentration-dependently

Cross-linking studies were performed to analyse the interaction sites of HlyB and HlyA2 in the presence of ATP. We chose HlyA2 as the substrate for this experiment since the modulation of HlyB ATPase velocity occurs independently of the secretion signal. We compared the cross-linking pattern of both proteins at two distinct molar ratios according to the maximum inhibition and maximum stimulation of ATPase velocity. Therefore we used a 1:1 molar ratio of HlyB and HlyA2 (sample called ‘inhibition’) and a 1:20 molar ratio of HlyB and HlyA2 (sample called ‘stimulation’). The concentration of ATP was kept constant at 2 mM. The new protein species obtained by cross-linking are shown in Figure 8(A). These species with higher molecular masses could only be detected by a CBB-stained SDS/PAGE gel or by Western blot analysis with an anti-HlyB antibody, but not with an anti-HlyA antibody. For further analysis of the cross-linked samples, we used LC–MS analysis and detected an intra-HlyB cross-link in the ‘inhibition’ sample linking the peptides SSVAGK and LAK (Figure 8B). The normalized signal intensity of this cross-linked dipeptide was at least 5-fold higher in comparison with the ‘stimulation’ sample. The cross-linked amino acids at positions 134 and 137 are located between the CLD and the TMD of HlyB.

Analysis of cross-linked HlyB/HlyA2 products and influence of HlyB-K322 on HlyA secretion

Figure 8
Analysis of cross-linked HlyB/HlyA2 products and influence of HlyB-K322 on HlyA secretion

(A) Analysis of cross-linked HlyB and HlyA2 products by Western blotting with a polyclonal anti-HlyB antibody. HlyB and HlyA2 were cross-linked by BS(PEG)5 in the presence of ATP and MgCl2. According to the maximum inhibition (1:1) and maximum stimulation (1:20) of ATPase velocity, we used two different molar ratios. We observed only in the presence of the cross-linking reagent a shift of the anti-HlyB antibody signal. (B) Representative fragment spectrum of the 2+ charged precursor mass m/z 590.8380 matching to an intra-HlyB cross-link linking the peptides SSVAGK and LAK (StavroX score 80, theoretical precursor mass 1180.667 Da, precursor mass deviation 1.39 p.p.m.). Fragment spectra have been identified from ‘inhibition’ samples of three independent replicates and two ‘stimulation’ samples. The signal intensity from extracted ion chromatograms (mass 590.8380, retention time 26 min) normalized with two non-cross-linked peptides of HlyB [YLIFDLEQR (residues 96–104), HLLALPISYFESR (residues 234–246)] is at least four times higher in ‘inhibition’ compared with ‘stimulation’ samples in three independent replicates. (C) Representative fragment spectrum of the 4+ charged precursor mass m/z 961.2247 which corresponds to the mass of the cross-linked peptides KAFEYQQSNNKVSYVYGHDA from HlyA2 (residues 944–963) and RRLDDKFSR from HlyB (residues 317–325) (StavroX score 65). The theoretical mass of the dipeptide is 3841.878 Da resulting in mass deviation of the measured precursor of −0.24 p.p.m. Thirteen fragment spectra generated from 3+ and 4+ charged precursor ions from three independent replicates of ‘stimulation’ samples could be mapped to this dipeptide (StavroX score between 52 and 284), whereas the dipeptide was not found in ‘inhibition’ samples. The extracted ion chromatogram signal from the corresponding peptides (retention time 56 min) is at least 69-fold higher in the ‘stimulation’ compared with the ‘inhibition’ sample in three independent replicates. (D) Schematic overview of a HlyB dimer and the cross-link results. The intra-HlyB cross-link between the CLD and the TMD of the ‘inhibition’ sample is marked with two red asterisks (*) connected by a straight line. The amino acid of HlyB, which cross-linked to the N-terminal part of HlyA2 in the ‘stimulation’ sample, is marked with a blue cross (+). (E) Secretion analysis of HlyA by SDS/PAGE in presence of wild-type (Wt) HlyB, HlyB-K322A and HlyB-K322W. Results are representative of three independent experiments. (F) Analysis of the expression level of wild-type (Wt) HlyB, HlyB-K322A and HlyB-K322W by Western blotting with a polyclonal anti-HlyB antibody. Results are representative of three independent experiments.

Figure 8
Analysis of cross-linked HlyB/HlyA2 products and influence of HlyB-K322 on HlyA secretion

(A) Analysis of cross-linked HlyB and HlyA2 products by Western blotting with a polyclonal anti-HlyB antibody. HlyB and HlyA2 were cross-linked by BS(PEG)5 in the presence of ATP and MgCl2. According to the maximum inhibition (1:1) and maximum stimulation (1:20) of ATPase velocity, we used two different molar ratios. We observed only in the presence of the cross-linking reagent a shift of the anti-HlyB antibody signal. (B) Representative fragment spectrum of the 2+ charged precursor mass m/z 590.8380 matching to an intra-HlyB cross-link linking the peptides SSVAGK and LAK (StavroX score 80, theoretical precursor mass 1180.667 Da, precursor mass deviation 1.39 p.p.m.). Fragment spectra have been identified from ‘inhibition’ samples of three independent replicates and two ‘stimulation’ samples. The signal intensity from extracted ion chromatograms (mass 590.8380, retention time 26 min) normalized with two non-cross-linked peptides of HlyB [YLIFDLEQR (residues 96–104), HLLALPISYFESR (residues 234–246)] is at least four times higher in ‘inhibition’ compared with ‘stimulation’ samples in three independent replicates. (C) Representative fragment spectrum of the 4+ charged precursor mass m/z 961.2247 which corresponds to the mass of the cross-linked peptides KAFEYQQSNNKVSYVYGHDA from HlyA2 (residues 944–963) and RRLDDKFSR from HlyB (residues 317–325) (StavroX score 65). The theoretical mass of the dipeptide is 3841.878 Da resulting in mass deviation of the measured precursor of −0.24 p.p.m. Thirteen fragment spectra generated from 3+ and 4+ charged precursor ions from three independent replicates of ‘stimulation’ samples could be mapped to this dipeptide (StavroX score between 52 and 284), whereas the dipeptide was not found in ‘inhibition’ samples. The extracted ion chromatogram signal from the corresponding peptides (retention time 56 min) is at least 69-fold higher in the ‘stimulation’ compared with the ‘inhibition’ sample in three independent replicates. (D) Schematic overview of a HlyB dimer and the cross-link results. The intra-HlyB cross-link between the CLD and the TMD of the ‘inhibition’ sample is marked with two red asterisks (*) connected by a straight line. The amino acid of HlyB, which cross-linked to the N-terminal part of HlyA2 in the ‘stimulation’ sample, is marked with a blue cross (+). (E) Secretion analysis of HlyA by SDS/PAGE in presence of wild-type (Wt) HlyB, HlyB-K322A and HlyB-K322W. Results are representative of three independent experiments. (F) Analysis of the expression level of wild-type (Wt) HlyB, HlyB-K322A and HlyB-K322W by Western blotting with a polyclonal anti-HlyB antibody. Results are representative of three independent experiments.

To determine whether this observed intra-HlyB cross-link depends on the substrate, we repeated the experiment in the absence of the substrate. Here the cross-link fragments found previously were not observed, which is no final proof but suggests that the observed intra-HlyB cross-link depends on the substrate.

In comparison, the ‘stimulation’ sample contained a nearly exclusive cross-link between the HlyB peptide RRLDDKFSR (residues 317–325) and the HlyA2 peptide KAFEYQQSNNKVSYVYGHDA (corresponding to residues 944–963) (Figure 8C). The HlyA2 peptide identified is located at the C-terminal end of the protein, whereas the HlyB peptide is located in the TMD at position 322. The cross-linked peptides of HlyB detected are visualized schematically in Figure 8(D).

To verify in vivo the relevance of the identified HlyB–HlyA2 cross-link, we mutated Lys322 of HlyB to alanine and tryptophan respectively and examined the effect on HlyA secretion. Secretion supernatants were analysed by SDS/PAGE (Figure 8E). The secretion level of HlyB-K322A is similar to the secretion with wild-type HlyB. In contrast, the secretion level of HlyB-K322W is significantly reduced in comparison with wild-type HlyB. This difference is not due to a different expression level of HlyB as demonstrated by Western blot analysis (Figure 8F).

DISCUSSION

In the present paper, we report for the first time the purification and characterization of HlyB, the full-length ABC transporter of the HlyA T1SS of E. coli. Despite the large and diverse field of ABC transporters, proteins of this superfamily share a common organization consisting of at least two TMDs and two NBDs that can be arranged in any possible combination [34]. Many of these transporters possess positive co-operativity as has been observed in systems such as P-glycoprotein, BmrA, maltose or histidine importers [3538]. Also the isolated soluble NBDs of HlyB displayed positive co-operativity, suggesting ATP-induced protein dimerization [30,39]. In the present study, solubilized full-length HlyB also exhibited positive co-operativity in ATP hydrolysis, which suggests that HlyB acts as a functional dimer. The range of our HlyB kinetic parameters is in line with those of, e.g., MacB, the ABC transporter of the macrolide-exporter system of E. coli [40].

The ABC transporter HlyB contains a CLD localized at its N-terminus [18]. To investigate the role of the CLD in the context of its ABC transporter, we cloned, expressed and purified a HlyB derivative lacking this domain (HlyB∆CLD). Kinetic studies of HlyB∆CLD revealed a K0.5 and a Hill coefficient almost identical with that of HlyB, but the maximal reaction velocity was increased approximately 15-fold. This comparison suggests a significant influence of the CLD on the hydrolytic velocity of the NBDs, whereas no influence on the substrate-binding affinity or co-operativity was detected.

For further investigation, we compared the activity of HlyB and HlyB∆CLD in the presence of a transport substrate. When increasing amounts of unfolded HlyA1 were added, HlyB ATPase activity showed a biphasic behaviour (Figure 5). By using the HlyB-Y20A mutation, we clearly assigned the distinct interaction of the CLD and HlyA1 being the reason for an inhibitory effect on HlyB ATPase activity. Lecher et al. [18] found that Tyr20 plays an important role, since the Y20A mutation resulted in a drastic reduction of substrate secretion, whereas the deletion of the CLD (HlyB∆CLD) abolished substrate secretion completely. Additionally, our results show that HlyB∆CLD possesses no co-operativity in the presence of HlyA1 or HlyA2, whereas an ‘inactivated’ CLD (e.g. Y20A mutant) shows co-operative interaction (Table 2). This points to a communication between the CLD and the NBD modulating hydrolytic activity (Figure 1). Overall, our results clearly confirm a distinct interaction of the substrate with the CLD leading to a reduction in HlyB hydrolytic velocity. This points to a regulatory role for the CLD as an autoinhibitory domain.

A regulatory role for the CLD by a hypothetical spatial reorientation within HlyB has also been supported by our cross-linking results (Figure 8). We suggest at least two different spatial conformations of HlyB with respect to the orientation of the CLD. Our findings imply a spatial reorientation of the CLD relative to the TMD as we observed a change in the cross-link pattern in the presence of low (‘inhibiting’) and high (‘stimulating’) substrate concentrations. As hydrolytic activity at the ‘inhibiting’ protein ratio is observed, we assume that the CLD position is also fixed during ATP binding and hydrolysis. This is in contrast with the structure of the PCAT (C39-peptidase-containing ABC transporter), in which the peptidase domain was not visible in the electron density of the ATP-bound state [41]. When we employed ‘stimulating’ HlyA2 concentrations, we observed a new cross-link product between the C-terminus of HlyA2 (residue 954) and the HlyB TMD (residue 322). Finally, the reduction of the HlyA secretion upon a K322W mutation supports the relevance of this position.

Terminal regulatory domains have already been described before for other ATPases and different systems such as the γ subunit of SpoIIIE, SecA, TrwK (type IV secretion system) and EccC (type VII protein secretion) [4245]. For example, plant plasma membrane H+- and Ca2+-ATPases contain autoinhibitory domains, which are located C-terminally or N-terminally respectively [46,47]. In the case of P2B Ca2+ pumps, calmodulin activates the pump upon binding [48]. In good agreement with the increased hydrolytic activity of HlyB∆CLD, it was also demonstrated that the removal of the autoinhibitory domain of the plant plasma H+-ATPase results in an activated pump [49]. In principle, this regulatory mode is in line with our results. As it is generally accepted that exporters regulate ATPase activity upon substrate binding, we suggest that the CLD acts as an autoinhibitory and interdomain-regulating subunit. In this model, the CLD locks HlyB in an ATP-bound state with strongly reduced hydrolytic capability in the absence of the substrate. In the presence of the substrate, the lock (CLD) is removed, and the transporter operates with increased hydrolytic activity. This seems plausible as the intracellular ATP concentration is approximately 10-fold higher than the Km of most ABC transporters, suggesting that NBDs are always present in the ATP-bound state in vivo [50]. A similar model has already been proposed for an ABC uptake system on the basis of the structures of MBP–MalFGK2 and ModB2C2A with the substrate-binding protein acting as a ‘lock-release’ [51]. Moreover, in the case of the ABC importer ModBC from Methanosarcina acetivorans, a molybdate ion inhibits the transporter by binding to a regulatory cytosolic domain, a mechanism termed ‘transinhibition’ [52]. Substrate-induced inhibition of the ATPase activity of HlyB variants or parts of HlyB has been described, probably caused by an interaction between HlyB-NBD and HlyA1 [19]. ATPase activities of GST–HlyB-NBD fusion protein as well as His-tagged HlyB-NBD were partially inhibited by HlyA or a C-terminal fragment of HlyA respectively [19,53]. Besides HlyB, the ABC transporter PrtD was also inhibited by its cognate C-terminal secretion signal of PrtG or PrtB metalloproteases [54]. However, the exact mechanistic details and amino acid residues involved in the interaction of the CLD and the NBD or TMD cannot be defined on the basis of the data from the present study.

Additionally, we investigated the role of the HlyA secretion signal in the context of substrate–CLD interaction. We observed that the absence of the secretion signal had only a minor impact on the kinetic parameters (Table 2). Variations of HlyB substrate affinities (K1, K2) and velocities (v1, v2) may be caused by different physical characteristics of HlyA1 and HlyA2. We found no impact on the general behaviour concerning ATPase stimulation or inhibition. Thus the interaction occurs independently of the secretion sequence. The N-terminal part of HlyA1 harbours the RTX domains that include the GG-repeats. The crystal structure of two members of the RTX toxin superfamily revealed that GG-repeats bind Ca2+ ions, and studies also showed that this binding induces substrate folding at the exterior of the cell [6,5557]. The results of Lecher et al. [18] support our data and suggest an exclusive interaction of the unfolded substrate/RTX domain with the CLD. In turn, Ca2+-induced folding of the substrate might be the reason that HlyA1 does not inhibit the HlyB ATPase activity in the presence of Ca2+. Therefore participation of the GG-repeats in CLD interaction seems likely and results in a decrease in reaction velocity. The secretion of HlyA is dependent on the presence of its C-terminal secretion signal [20,58,59]. Thus our observed secretion-signal-independent interaction could only represent one step within a well-regulated and complex secretion mechanism. Although we have to take into account that solubilized HlyB shows a low uncoupled basal ATPase activity, which is not likely to be present in vivo, we demonstrated a distinct modulation of HlyB ATPase activity in the presence of its substrate.

Finally, we suggest that at least two different binding sites of HlyA exist within HlyB. We clearly identified a CLD–substrate interaction leading to a reduction in hydrolytic velocity. This underlines an interdomain and regulatory role of the CLD modulating the NBDs catalytic activity depending on the concentration of the substrate. Furthermore, we also observed stimulatory effects on the hydrolytic activity. Stimulatory effects were determined for HlyB as well as for HlyB∆CLD and HlyB-Y20A. This strongly suggests that the stimulation occurred independently of the CLD itself and the HlyA-binding ability to the CLD (Y20A mutant). The NBD of HlyB presumably can also be excluded to modulate stimulation, since at least the isolated domain interacts exclusively with the secretion signal of HlyA. In contrast, we observed a secretion-signal-independent stimulation of HlyB ATPase activity [19]. This subsequently leads to the assumption that a second binding site could be present within the TMD of HlyB stimulating the ATPase activity, which is conceptually similar to the proposal for MsbA [60,61] and supported by our cross-linking data (Figure 8).

In summary, our data assign an interdomain regulatory role for the CLD, the N-terminal appendix of HlyB. Such an arrangement has not been identified for ABC export systems previously and suggests that the secretion of the substrate is a highly ordered and regulated process, in which the CLD possesses a key role.

AUTHOR CONTRIBUTION

Sven Reimann, Sander Smits, Lutz Schmitt, Gereon Poschmann and Kai Stühler conceived and designed the experiments and analysed the data. Sven Reimann and Gereon Poschmann performed the experiments. Kerstin Kanonenberg contributed HlyA1 protein. Sven Reimann, Sander Smits, Lutz Schmitt and Gereon Poschmann wrote the paper.

We thank Christian Schwarz and Michael Lenders for valuable discussions. We thank Diana Kleinschrodt and Iris Fey, Protein Production Facility of Heinrich-Heine-University, for their support. We gratefully acknowledge support (and training) from the International NRW Research School BioStruct, granted by the Ministry of Innovation, Science and Research of the State North Rhine-Westphalia, the Heinrich-Heine-University Düsseldorf and the Entrepreneur Foundation at the Heinrich-Heine-University of Düsseldorf.

FUNDING

This project was funded in part by the Deutsche Forschungsgemeinschaft (DFG) [CRC 1208, projects A01 to L.S. and Z01 to K.S.]

Abbreviations

     
  • ABC

    ATP-binding cassette

  •  
  • BS(PEG)5

    bis-N-succinimidyl-penta(ethylene glycol) ester

  •  
  • CBB

    Coomassie Brilliant Blue

  •  
  • CLD

    C39-peptidase-like domain

  •  
  • CMC

    critical micellar concentration

  •  
  • HCD

    higher-energy collisional dissociation

  •  
  • HlyA

    haemolysin A

  •  
  • HlyB

    haemolysin B

  •  
  • HlyD

    haemolysin D

  •  
  • IMAC

    immobilized metal-ion-affinity chromatography

  •  
  • LMNG

    lauryl maltose neopentyl glycol

  •  
  • MFP

    membrane-fusion protein

  •  
  • NBD

    nucleotide-binding domain

  •  
  • RTX

    repeats in toxins

  •  
  • SEC

    size-exclusion chromatography

  •  
  • T1SS

    Type 1 secretion system

  •  
  • TBS-T

    TBS with Tween 20

  •  
  • TEV

    tobacco etch virus

  •  
  • TFA

    trifluoroacetic acid

  •  
  • TMD

    transmembrane domain

References

References
1
Bakkes
P.J.
Jenewein
S.
Smits
S.H.
Holland
I.B.
Schmitt
L.
The rate of folding dictates substrate secretion by the Escherichia coli hemolysin type 1 secretion system
J. Biol. Chem.
2010
, vol. 
285
 (pg. 
40573
-
40580
)
[PubMed]
2
Holland
I.B.
Schmitt
L.
Young
J.
Type 1 protein secretion in bacteria, the ABC-transporter dependent pathway
Mol. Membr. Biol.
2005
, vol. 
22
 (pg. 
29
-
39
)
[PubMed]
3
Ludwig
A.
Schmid
A.
Benz
R.
Goebel
W.
Mutations affecting pore formation by haemolysin from Escherichia coli
Mol. Gen. Genet.
1991
, vol. 
226
 (pg. 
198
-
208
)
[PubMed]
4
Welch
R.A.
Forestier
C.
Lobo
A.
Pellett
S.
Thomas
W.
Jr
Rowe
G.
The synthesis and function of the Escherichia coli hemolysin and related RTX exotoxins
FEMS Microbiol. Immunol.
1992
, vol. 
5
 (pg. 
29
-
36
)
[PubMed]
5
Thomas
S.
Holland
I.B.
Schmitt
L.
The Type 1 secretion pathway: the hemolysin system and beyond
Biochim. Biophys. Acta
2014
, vol. 
1843
 (pg. 
1629
-
1641
)
[PubMed]
6
Baumann
U.
Wu
S.
Flaherty
K.M.
McKay
D.B.
Three-dimensional structure of the alkaline protease of Pseudomonas aeruginosa: a two-domain protein with a calcium binding parallel β roll motif
EMBO J.
1993
, vol. 
12
 (pg. 
3357
-
3364
)
[PubMed]
7
Ostolaza
H.
Soloaga
A.
Goñi
F.M.
The binding of divalent cations to Escherichia coli α-haemolysin
Eur. J. Biochem.
1995
, vol. 
228
 (pg. 
39
-
44
)
[PubMed]
8
Linhartová
I.
Bumba
L.
Mašín
J.
Basler
M.
Osička
R.
Kamanová
J.
Procházková
K.
Adkins
I.
Hejnová-Holubová
J.
Sadílková
L.
, et al. 
RTX proteins: a highly diverse family secreted by a common mechanism
FEMS Microbiol Rev.
2010
, vol. 
34
 (pg. 
1076
-
1112
)
[PubMed]
9
Delepelaire
P.
Type I secretion in Gram-negative bacteria
Biochim. Biophys. Acta
2004
, vol. 
1694
 (pg. 
149
-
161
)
[PubMed]
10
Letoffe
S.
Delepelaire
P.
Wandersman
C.
Protein secretion in Gram-negative bacteria: assembly of the three components of ABC protein-mediated exporters is ordered and promoted by substrate binding
EMBO J.
1996
, vol. 
15
 (pg. 
5804
-
5811
)
[PubMed]
11
Thanabalu
T.
Koronakis
E.
Hughes
C.
Koronakis
V.
Substrate-induced assembly of a contiguous channel for protein export from E.coli: reversible bridging of an inner-membrane translocase to an outer membrane exit pore
EMBO J.
1998
, vol. 
17
 (pg. 
6487
-
6496
)
[PubMed]
12
Balakrishnan
L.
Hughes
C.
Koronakis
V.
Substrate-triggered recruitment of the TolC channel-tunnel during type I export of hemolysin by Escherichia coli
J. Mol. Biol.
2001
, vol. 
313
 (pg. 
501
-
510
)
[PubMed]
13
Ghigo
J.M.
Wandersman
C.
A carboxyl-terminal four-amino acid motif is required for secretion of the metalloprotease PrtG through the Erwinia chrysanthemi protease secretion pathway
J. Biol. Chem.
1994
, vol. 
269
 (pg. 
8979
-
8985
)
[PubMed]
14
Kenny
B.
Taylor
S.
Holland
I.B.
Identification of individual amino acids required for secretion within the haemolysin (HlyA) C-terminal targeting region
Mol. Microbiol.
1992
, vol. 
6
 (pg. 
1477
-
1489
)
[PubMed]
15
Kenny
B.
Chervaux
C.
Holland
I.B.
Evidence that residues −15 to −46 of the haemolysin secretion signal are involved in early steps in secretion, leading to recognition of the translocator
Mol. Microbiol.
1994
, vol. 
11
 (pg. 
99
-
109
)
[PubMed]
16
Koronakis
V.
Sharff
A.
Koronakis
E.
Luisi
B.
Hughes
C.
Crystal structure of the bacterial membrane protein TolC central to multidrug efflux and protein export
Nature
2000
, vol. 
405
 (pg. 
914
-
919
)
[PubMed]
17
Havarstein
L.S.
Diep
D.B.
Nes
I.F.
A family of bacteriocin ABC transporters carry out proteolytic processing of their substrates concomitant with export
Mol. Microbiol.
1995
, vol. 
16
 (pg. 
229
-
240
)
[PubMed]
18
Lecher
J.
Schwarz
C.K.
Stoldt
M.
Smits
S.H.
Willbold
D.
Schmitt
L.
An RTX transporter tethers its unfolded substrate during secretion via a unique N-terminal domain
Structure
2012
, vol. 
20
 (pg. 
1778
-
1787
)
[PubMed]
19
Benabdelhak
H.
Kiontke
S.
Horn
C.
Ernst
R.
Blight
M.A.
Holland
I.B.
Schmitt
L.
A specific interaction between the NBD of the ABC-transporter HlyB and a C-terminal fragment of its transport substrate haemolysin A
J. Mol. Biol.
2003
, vol. 
327
 (pg. 
1169
-
1179
)
[PubMed]
20
Nicaud
J.M.
Mackman
N.
Gray
L.
Holland
I.B.
The C-terminal, 23 kDa peptide of E. coli haemolysin 2001 contains all the information necessary for its secretion by the haemolysin (Hly) export machinery
FEBS Lett.
1986
, vol. 
204
 (pg. 
331
-
335
)
[PubMed]
21
Thomas
S.
Bakkes
P.J.
Smits
S.H.
Schmitt
L.
Equilibrium folding of pro-HlyA from Escherichia coli reveals a stable calcium ion dependent folding intermediate
Biochim. Biophys. Acta
2014
, vol. 
1844
 (pg. 
1500
-
1510
)
[PubMed]
22
Mackman
N.
Holland
I.B.
Functional characterization of a cloned haemolysin determinant from E. coli of human origin, encoding information for the secretion of a 107K polypeptide
Mol. Gen. Genet.
1984
, vol. 
196
 (pg. 
129
-
134
)
[PubMed]
23
Zaitseva
J.
Jenewein
S.
Jumpertz
T.
Holland
I.B.
Schmitt
L.
H662 is the linchpin of ATP hydrolysis in the nucleotide-binding domain of the ABC transporter HlyB
EMBO J.
2005
, vol. 
24
 (pg. 
1901
-
1910
)
[PubMed]
24
Zaitseva
J.
Holland
I.B.
Schmitt
L.
The role of CAPS buffer in expanding the crystallization space of the nucleotide-binding domain of the ABC transporter haemolysin B from Escherichia coli
Acta Crystallogr. D Biol. Crystallogr.
2004
, vol. 
60
 (pg. 
1076
-
1084
)
[PubMed]
25
Blommel
P.G.
Fox
B.G.
A combined approach to improving large-scale production of tobacco etch virus protease
Protein Expr. Purif.
2007
, vol. 
55
 (pg. 
53
-
68
)
[PubMed]
26
Baykov
A.A.
Evtushenko
O.A.
Avaeva
S.M.
A malachite green procedure for orthophosphate determination and its use in alkaline phosphatase-based enzyme immunoassay
Anal. Biochem.
1988
, vol. 
171
 (pg. 
266
-
270
)
[PubMed]
27
Poschmann
G.
Seyfarth
K.
Besong Agbo
D.
Klafki
H.W.
Rozman
J.
Wurst
W.
Wiltfang
J.
Meyer
H.E.
Klingenspor
M.
Stühler
K.
High-fat diet induced isoform changes of the Parkinson's disease protein DJ-1
J. Proteome Res.
2014
, vol. 
13
 (pg. 
2339
-
2351
)
[PubMed]
28
Gotze
M.
Pettelkau
J.
Schaks
S.
Bosse
K.
Ihling
C.H.
Krauth
F.
Fritzsche
R.
Kühn
U.
Sinz
A.
StavroX: a software for analyzing crosslinked products in protein interaction studies
J. Am. Soc. Mass Spectrom.
2012
, vol. 
23
 (pg. 
76
-
87
)
[PubMed]
29
Ellinger
P.
Kluth
M.
Stindt
J.
Smits
S.H.
Schmitt
L.
Detergent screening and purification of the human liver ABC transporters BSEP (ABCB11) and MDR3 (ABCB4) expressed in the yeast Pichia pastoris
PLoS One
2013
, vol. 
8
 pg. 
e60620
 
[PubMed]
30
Zaitseva
J.
Jenewein
S.
Wiedenmann
A.
Benabdelhak
H.
Holland
I.B.
Schmitt
L.
Functional characterization and ATP-induced dimerization of the isolated ABC-domain of the haemolysin B transporter
Biochemistry
2005
, vol. 
44
 (pg. 
9680
-
9690
)
[PubMed]
31
Litman
T.
Nielsen
D.
Skovsgaard
T.
Zeuthen
T.
Stein
W.D.
ATPase activity of P-glycoprotein related to emergence of drug resistance in Ehrlich ascites tumor cell lines
Biochim. Biophys. Acta
1997
, vol. 
1361
 (pg. 
147
-
158
)
[PubMed]
32
Jones
H.E.
Holland
I.B.
Baker
H.L.
Campbell
A.K.
Slow changes in cytosolic free Ca2+ in Escherichia coli highlight two putative influx mechanisms in response to changes in extracellular calcium
Cell Calcium
1999
, vol. 
25
 (pg. 
265
-
274
)
[PubMed]
33
Jumpertz
T.
Chervaux
C.
Racher
K.
Zouhair
M.
Blight
M.A.
Holland
I.B.
Schmitt
L.
Mutations affecting the extreme C terminus of Escherichia coli haemolysin A reduce haemolytic activity by altering the folding of the toxin
Microbiology
2010
, vol. 
156
 (pg. 
2495
-
2505
)
[PubMed]
34
Biemans-Oldehinkel
E.
Doeven
M.K.
Poolman
B.
ABC transporter architecture and regulatory roles of accessory domains
FEBS Lett.
2006
, vol. 
580
 (pg. 
1023
-
1035
)
[PubMed]
35
Senior
A.E.
Bhagat
S.
P-glycoprotein shows strong catalytic cooperativity between the two nucleotide sites
Biochemistry
1998
, vol. 
37
 (pg. 
831
-
836
)
[PubMed]
36
Liu
C.E.
Liu
P.Q.
Ames
G.F.
Characterization of the adenosine triphosphatase activity of the periplasmic histidine permease, a traffic ATPase (ABC transporter)
J. Biol. Chem.
1997
, vol. 
272
 (pg. 
21883
-
21891
)
[PubMed]
37
Davidson
A.L.
Laghaeian
S.S.
Mannering
D.E.
The maltose transport system of Escherichia coli displays positive cooperativity in ATP hydrolysis
J. Biol. Chem.
1996
, vol. 
271
 (pg. 
4858
-
4863
)
[PubMed]
38
Steinfels
E.
Orelle
C.
Fantino
J.R.
Dalmas
O.
Rigaud
J.L.
Denizot
F.
Di Pietro
A.
Jault
J.M.
Characterization of YvcC (BmrA), a multidrug ABC transporter constitutively expressed in Bacillus subtilis
Biochemistry
2004
, vol. 
43
 (pg. 
7491
-
7502
)
[PubMed]
39
Benabdelhak
H.
Schmitt
L.
Horn
C.
Jumel
K.
Blight
M.A.
Holland
I.B.
Positive co-operative activity and dimerization of the isolated ABC ATPase domain of HlyB from Escherichia coli
Biochem. J.
2005
, vol. 
386
 (pg. 
489
-
495
)
[PubMed]
40
Lin
H.T.
Bavro
V.N.
Barrera
N.P.
Frankish
H.M.
Velamakanni
S.
van Veen
H.W.
Robinson
C.V.
Borges-Walmsley
M.I.
Walmsley
A.R.
MacB ABC transporter is a dimer whose ATPase activity and macrolide-binding capacity are regulated by the membrane fusion protein MacA
J. Biol. Chem.
2009
, vol. 
284
 (pg. 
1145
-
1154
)
[PubMed]
41
Lin
D.Y.
Huang
S.
Chen
J.
Crystal structures of a polypeptide processing and secretion transporter
Nature
2015
, vol. 
523
 (pg. 
425
-
430
)
[PubMed]
42
Besprozvannaya
M.
Pivorunas
V.L.
Feldman
Z.
Burton
B.M.
SpoIIIE protein achieves directional DNA translocation through allosteric regulation of ATPase activity by an accessory domain
J. Biol. Chem.
2013
, vol. 
288
 (pg. 
28962
-
28974
)
[PubMed]
43
Karamanou
S.
Vrontou
E.
Sianidis
G.
Baud
C.
Roos
T.
Kuhn
A.
Politou
A.S.
Economou
A.
A molecular switch in SecA protein couples ATP hydrolysis to protein translocation
Mol. Microbiol.
1999
, vol. 
34
 (pg. 
1133
-
1145
)
[PubMed]
44
Peña
A.
Ripoll-Rozada
J.
Zunzunegui
S.
Cabezón
E.
de la Cruz
F.
Arechaga
I.
Autoinhibitory regulation of TrwK, an essential VirB4 ATPase in type IV secretion systems
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
17376
-
17382
)
[PubMed]
45
Rosenberg
O.S.
Dovala
D.
Li
X.
Connolly
L.
Bendebury
A.
Finer-Moore
J.
Holton
J.
Cheng
Y.
Stroud
R.M.
Cox
J.S.
Substrates control multimerization and activation of the multi-domain ATPase motor of type VII secretion
Cell
2015
, vol. 
161
 (pg. 
501
-
512
)
[PubMed]
46
Axelsen
K.B.
Palmgren
M.G.
Inventory of the superfamily of P-type ion pumps in Arabidopsis
Plant Physiol.
2001
, vol. 
126
 (pg. 
696
-
706
)
[PubMed]
47
Baekgaard
L.
Fuglsang
A.T.
Palmgren
M.G.
Regulation of plant plasma membrane H+- and Ca2+-ATPases by terminal domains
J. Bioenerg. Biomembr.
2005
, vol. 
37
 (pg. 
369
-
374
)
[PubMed]
48
Harper
J.F.
Hong
B.
Hwang
I.
Guo
H.Q.
Stoddard
R.
Huang
J.F.
Palmgren
M.G.
Sze
H.
A novel calmodulin-regulated Ca2+-ATPase (ACA2) from Arabidopsis with an N-terminal autoinhibitory domain
J. Biol. Chem.
1998
, vol. 
273
 (pg. 
1099
-
1106
)
[PubMed]
49
Palmgren
M.G.
Sommarin
M.
Serrano
R.
Larsson
C.
Identification of an autoinhibitory domain in the C-terminal region of the plant plasma membrane H+-ATPase
J. Biol. Chem.
1991
, vol. 
266
 (pg. 
20470
-
20475
)
[PubMed]
50
Neidhardt
F.C.
, et al. 
Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology
1987
American Society for Microbiology Publishers
 
Vol. 1
51
Oldham
M.L.
Khare
D.
Quiocho
F.A.
Davidson
A.L.
Chen
J.
Crystal structure of a catalytic intermediate of the maltose transporter
Nature
2007
, vol. 
450
 (pg. 
515
-
521
)
[PubMed]
52
Gerber
S.
Comellas-Bigler
M.
Goetz
B.A.
Locher
K.P.
Structural basis of trans-inhibition in a molybdate/tungstate ABC transporter
Science
2008
, vol. 
321
 (pg. 
246
-
250
)
[PubMed]
53
Koronakis
V.
Hughes
C.
Koronakis
E.
ATPase activity and ATP/ADP-induced conformational change in the soluble domain of the bacterial protein translocator HlyB
Mol. Microbiol.
1993
, vol. 
8
 (pg. 
1163
-
1175
)
[PubMed]
54
Delepelaire
P.
PrtD, the integral membrane ATP-binding cassette component of the Erwinia chrysanthemi metalloprotease secretion system, exhibits a secretion signal-regulated ATPase activity
J. Biol. Chem.
1994
, vol. 
269
 (pg. 
27952
-
27957
)
[PubMed]
55
Hamada
K.
Hata
Y.
Katsuya
Y.
Hiramatsu
H.
Fujiwara
T.
Katsube
Y.
Crystal structure of Serratia protease, a zinc-dependent proteinase from Serratia sp. E-15, containing a β-sheet coil motif at 2.0 Å resolution
J. Biochem.
1996
, vol. 
119
 (pg. 
844
-
851
)
[PubMed]
56
Sánchez-Magraner
L.
Viguera
A.R.
García-Pacios
M.
Garcillán
M.P.
Arrondo
J.L.
de la Cruz
F.
Goñi
F.M.
Ostolaza
H.
The calcium-binding C-terminal domain of Escherichia coli α-hemolysin is a major determinant in the surface-active properties of the protein
J. Biol. Chem.
2007
, vol. 
282
 (pg. 
11827
-
11835
)
[PubMed]
57
Sánchez-Magraner
L.
Cortajarena
A.L.
García-Pacios
M.
Arrondo
J.L.
Agirre
J.
Guérin
D.M.
Goñi
F.M.
Ostolaza
H.
Interdomain Ca2+ effects in Escherichia coli α-haemolysin: Ca2+ binding to the C-terminal domain stabilizes both C- and N-terminal domains
Biochim. Biophys. Acta
2010
, vol. 
1798
 (pg. 
1225
-
1233
)
[PubMed]
58
Gray
L.
Mackman
N.
Nicaud
J.M.
Holland
I.B.
The carboxy-terminal region of haemolysin 2001 is required for secretion of the toxin from Escherichia coli
Mol. Gen. Genet.
1986
, vol. 
205
 (pg. 
127
-
133
)
[PubMed]
59
Felmlee
T.
Pellett
S.
Lee
E.Y.
Welch
R.A.
Escherichia coli hemolysin is released extracellularly without cleavage of a signal peptide
J. Bacteriol.
1985
, vol. 
163
 (pg. 
88
-
93
)
[PubMed]
60
Aller
S.G.
Yu
J.
Ward
A.
Weng
Y.
Chittaboina
S.
Zhuo
R.
Harrell
P.M.
Trinh
Y.T.
Zhang
Q.
Urbatsch
I.L.
Chang
G.
Structure of P-glycoprotein reveals a molecular basis for poly-specific drug binding
Science
2009
, vol. 
323
 (pg. 
1718
-
1722
)
[PubMed]
61
Smriti, Zou
P.
McHaourab
H.S.
Mapping daunorubicin-binding Sites in the ATP-binding cassette transporter MsbA using site-specific quenching by spin labels
J. Biol. Chem.
2009
, vol. 
284
 (pg. 
13904
-
13913
)
[PubMed]

Supplementary data