Dimerization of the Pseudomonas aeruginosa Translocator Chaperone PcrH Is Required for Stability, Not Function

Abstract

Type III secretion systems rely on hydrophobic translocator proteins that form a pore in the host cell membrane to deliver effector proteins into targeted host cells. These translocator proteins are stabilized in the cytoplasm and targeted for export with the help of specific chaperone proteins. In Pseudomonas aeruginosa, the chaperone of the pore-forming translocator proteins is PcrH. Although all translocator chaperones dimerize, the location of the dimerization interface is in dispute. Moreover, it has been reported that interfering with dimerization interferes with chaperone function. However, binding of P. aeruginosa chaperone PcrH to its cognate secretion substrate, PopD, results in dissociation of the PcrH dimer in vitro, arguing that dimerization of PcrH is likely not important for substrate binding or targeting translocators for export. We demonstrate that PcrH dimerization occurs in vivo in P. aeruginosa and used a genetic screen to identify a dimerization mutant of PcrH. The mutant protein is fully functional in that it can both stabilize PopB and PopD in the cytoplasm and promote their export via the type III secretion system. The location of the mutation suggests that the dimerization interface of PcrH mirrors that of the Yersinia homolog SycD and not the dimerization interface that had previously been reported for PcrH based on crystallographic evidence. Finally, we present data that the dimerization mutant of PcrH is less stable than the wild-type protein in P. aeruginosa, suggesting that the function of dimerization is stabilization of PcrH in the absence of its cognate cargo.

INTRODUCTION

Type III secretion systems (T3SSs) are nanomachines used by a wide variety of Gram-negative pathogens to directly inject effector proteins into targeted host cells (1, 2). Effector secretion is triggered by cell contact, and the delivery of effector proteins into the host cell relies on a set of translocator proteins. In animal pathogens, two translocator proteins multimerize to form a pore in the host cell plasma membrane to which the needle is likely docked via a specialized structure at the tip of the secretion needle (3).

In the case of Pseudomonas aeruginosa, the pore-forming translocator proteins are PopB and PopD. Both proteins bind to the chaperone protein PcrH in the cytoplasm of P. aeruginosa. Binding to PcrH stabilizes both translocator proteins (4–6). PcrH is also required for the efficient export of PopB and PopD. Targeting PopB and PopD for export can be genetically separated from the stabilization function of PcrH and involves a region of the protein that includes residues 49 and 50 (7).

PcrH is a class II chaperone, which is a conserved class of chaperones involved in stabilizing pore-forming translocator proteins (8). This class of chaperones is characterized by a structurally conserved motif consisting of four tetratricopeptide repeats (TPRs), which assume the shape of hand that binds the translocator protein in its palm (9–11). Members of this family have been found to dimerize in vitro, although, based on crystallographic evidence, the orientation of the individual proteins within the dimer seemed to differ. The biological dimer of PcrH was reported to involve an interaction between the outside surfaces of the TPR “palms” (11). SycD and LcrH of Yersinia enterocolitica and Y. pestis, respectively, were reported to dimerize in a head-to-head fashion (9, 12). With IpgC, the Shigella flexneri translocator chaperone, dimer formation involves binding of the amino-terminal domain of one monomer to the convex outer surface of the palm formed by the TPR repeats of the second monomer (10, 13). An alternative dimer in which IpgC monomers interact in a head-to-head fashion, akin to the dimer reported for SycD, has also been reported (14).

Binding of PopD by PcrH in vitro dissolves the PcrH dimer, arguing that PcrH is monomeric when stabilizing its cognate translocator protein and targeting it for export (11). The same has also been observed in the case of Shigella IpgC binding to the translocator protein IpaC (15). However, mutating two residues of the Y. enterocolitica homolog SycD involved in the head-to-head dimerization interface of this protein, alanine 61 and leucine 65, to glutamic acid, while disrupting the dimer in vitro, also abolished SycD function in vivo, arguing that the dimer is functionally important (9). Mutating two residues involved in the asymmetric IpgC dimer, alanine 94 and valine 95, similarly abolished IpgC dimerization in vitro and function in vivo (10). Since the loss of function observed with the SycD and IpgC dimerization mutants mirrors the null mutant phenotype, an alternate explanation is that these mutants, which were designed based on crystallographic data, are misfolded in vivo and therefore not functional.

Here, we used a genetic technique to identify a pcrH mutation that abolishes PcrH dimerization, while not interfering with binding to PopD. The mutant protein was analyzed for dimerization using a bacterial two-hybrid system in Escherichia coli, as well as chemical cross-linking in vitro and in P. aeruginosa. We also assayed the ability of the dimerization mutant to stabilize and promote the export of PopB and PopD in vivo, as well as the impact of dimerization on PcrH stability in P. aeruginosa. In contrast to the crystallographic evidence, our data suggest that PcrH dimerizes in a head-to-head manner akin to SycD and LcrH of Yersinia sp. Moreover, unlike SycD, dimerization is not required for PcrH function. Instead, dimerization promotes PcrH stability.

MATERIALS AND METHODS

Media and culture conditions.

E. coli strains were routinely grown at 37°C in Luria-Bertani (LB) medium containing 10 g of NaCl/liter. P. aeruginosa was grown at 37°C in a modified LB medium (LB-MC) containing 200 mM NaCl, 0.5 mM CaCl₂, and 10 mM MgCl₂. The strains and plasmids used in the present study are listed in Table 1.

Table 1.

Strains and plasmids

Strain or plasmid	Genotype or relevant features	Source or reference
Strains
KDZif1ΔZ	E. coli two-hybrid analysis strain lacking rpoZ (encoding ω) and harboring a test promoter-lacZ fusion to detect Zif-dependent two-hybrid interactions	20
RP5795	PAO1F ΔfleQ pcrH(Δ46-71)	7
RP3014	PAO1F ΔfleQ Δ(pcrH popBD)	This study
Plasmids
pBRω-GP	Two-hybrid plasmid to create fusions to the C terminus of the ω subunit of E. coli RNAP, carb R, ColE1 oriR, lacUV5 promoter	23
pACtr-VSVG-ZifAP	Two-hybrid plasmid to create fusions to the C terminus of Zif (zinc-finger DNA-binding domain of murine Zif268 protein), VSV-G epitope tag fused to the N terminus of Zif, tetR, p15A oriR, lacUV5 promoter	23
pACtr-VSVG-ZifAP-popD	Two-hybrid plasmid encoding PopD fused to the C terminus of the RNA polymerase omega subunit	7
pBRω-GP-pcrH	Two-hybrid plasmid encoding PcrH fused to the C terminus of the RNA polymerase omega subunit	7
pBRω-GP-pcrH(A61S)	Two-hybrid plasmid encoding PcrH(A61S) fused to the C terminus of the RNA polymerase omega subunit	This study
pBRω-GP-pcrH(L65M)	Two-hybrid plasmid encoding PcrH(L65M) fused to the C terminus of the RNA polymerase omega subunit	This study
pBRω-GP-pcrH(A61S/L65M)	Two-hybrid plasmid encoding PcrH(A61S/L65M) fused to the C terminus of the RNA polymerase omega subunit	This study
pACtr-VSVG-ZifAP-pcrH	Two-hybrid plasmid encoding PcrH fused to the C terminus of Zif	This study
pACtr-VSVG-ZifAP-pcrH(A61S)	Two-hybrid plasmid encoding PcrH(A61S) fused to the C terminus of Zif	This study
pACtr-VSVG-ZifAP-pcrH(L65M)	Two-hybrid plasmid encoding PcrH(L65M) fused to the C terminus of Zif	This study
pACtr-VSVG-ZifAP-pcrH(A61S/L65M)	Two-hybrid plasmid encoding PcrH(A61S/L65M) fused to the C terminus of Zif	This study
pCLARA	Low-copy-number plasmid, spectinomycin resistance, arabinose-inducible promoter	This study
pCLARA-popD	popD cloned into pCLARA under the control of the pBAD promoter	This study
pPSV37	colE1 oriR, gentR, PA origin, oriT, lacUV5 promoter, lacIq, stops in every reading frame preceding the MCS and T7 terminator following the MCS relative to the lacUV5 promoter	24
pET28bTEV-pcrH	pET28b-derived vector with TEV site replacing the thrombin cleavage site separating the His tag and pcrH	This study
pET28bTEV-pcrH(A61S/L65M)	pET28b-derived vector with TEV site replacing the thrombin cleavage site separating the His tag and pcrH	This study
pET28bTEV-pcrH-Myc	pET28b-derived vector with TEV site replacing the thrombin cleavage site separating the His tag and pcrH	This study
pET28bTEV-pcrH(A61S/L65M)-Myc	pET28b-derived vector with TEV site replacing the thrombin cleavage site separating the His tag and pcrH	This study
pP37-pcrH-Myc	pcrH fused to a 2×Myc tag at the C terminus under the control of the lacUV5 promoter in pPSV37	7
pP37-pcrH-4VG-Myc	pcrH fused to four copies of the VSV-G epitope tag followed by a 2×Myc tag at the C terminus under the control of the lacUV5 promoter in pPSV37	This study
pP37-pcrH(A61S)-Myc	pcrH(A61S) fused to a 2×Myc tag at the C terminus under the control of the lacUV5 promoter in pPSV37	This study
pP37-pcrH(A61S/L65M)-Myc	pcrH(A61S/L65M) fused to a 2×Myc tag at the C terminus under the control of the lacUV5 promoter in pPSV37	This study

Plasmid construction.

E. coli strain DH5α was used for cloning, and E. coli strain Sm10λpir was used to mate plasmids into P. aeruginosa. pCLARA was constructed by amplifying araC and pBAD from pBAD30 using the primers Ara-Bgl and Ara-XhoI and combining it with the vector backbone of the low-copy-number plasmid pCL1920 (16), which was amplified using primers pCL-Xho and pCL-Bgl. The primers used here are listed in Table 2. popD was cloned into pCLARA as an EcoRI/HindIII fragment. PCR products were cloned into pBRω-GP and pACtr-VSVG-ZifAP as NotI/HindIII fragments and into pPSV37 as EcoRI/HindIII fragments.

Table 2.

Primers

Primer	Sequence (5′–3′)a	Description
pCL-Xho	AAAAActcgagCCCCGACACCCGCCAACACC	Amplify pCL1920 backbone, XhoI site
pCL-Bgl	AAAAAagatctAAGCCTGTTGATGATACCGC	Amplify pCL1920 backbone, BglII site
Ara-Bgl	AAAAAagatctTGTCAAATGGACGAAGCAGGG	Amplify araC and pBAD of pBAD30, BglII site
Ara-Xho	AAAAActcgagACAAAGAGTTTGTAGAAACGC	Amplify araC and pBAD of pBAD30, XhoI site
pcrH5R-opt	AAAAAgaattcGAAACATCAGGAGAAGGCAACCATCATGAACCAGCCGACCCCTTCCGAC	pcrH 5′ primer with exoS translation start site
CMyc2-3H	AAAAAaagcttTCACAGGTCCTCCTCCGAGAT	3′ primer to amplify genes tagged at the C terminus with a 2×Myc tag
pcrH-5Not	AAAAAgcggccgcAAACCAGCCGACCCCTTCCGAC	Primer to fuse pcrH to 3′ end of Zif or omega
popD-5Not	AAAAAgcggccgcAATCGACACGCAATATT	popD 5′ primer with a NotI site
pcrH3H	AAAAAaagcttTCAAGCGTTATCGGATTCATAT	pcrH 3′ primer with an HindIII site
pcrHA61S-3-1	CAGAAGATCTTCCAGtcgCTGTGCATGCTCGACCAC	5′ primer to amplify pcrH(A61S)
pcrHA61S-5-2	GTGGTCGAGCATGCACAGCGACTGGAAGATCTTCTG	3′ primer to amplify pcrH(A61S)
pcrHA61SL65M-3-1	CAGAAGATCTTCCAGTCGCTGTGCATGATGGACCACTACGACGCCCGCTAC	5′ primer to amplify pcrH(A61S/L65M)
pcrHA61SL65M-5-2	GTAGCGGGCGTCGTAGTGGTCCATCATGCACAGCGACTGGAAGATCTTCTG	3′ primer to amplify pcrH(A61S/L65M)
popD5R	AAAAAgaattcTTAGGAGGCGCCCCCATGATCGACACGCAATATTCCCT	5′ primer for popD ORF with an EcoRI site
popDEX3	AAAAAaagcttCGCGCGGAGACGGCTCAGACCACT	popD 3′ primer with an HindIII site
pcrHmut16-5Not	AAAAAGCGGCCGCAATGAACCAGCCGACCCCTTCCGACACCGACCAGCAACAGGCGCTGNNNGCCTTCCTGCGCGACGGCGGC	5′ primer to mutate codon 16 of pcrH
pcrHL65M-3-1	CTTCCAGGCACTGTGCATGATGGACCACTACGACGCCCGCTAC	Introduce L65M mutation into pcrH
pcrHL65M-5-2	GTAGCGGGCGTCGTAGTGGTCCATCATGCACAGTGCCTGGAAG
H-4VG-2M-5-2	CGATATCGGTGTACTTGCCCAGCCGGTTCATCTCGATGTCGGTGTATTTTCCTAATCTATTCATTTCAATATCTGTATAAGCGTTATCGGATTCATATGC	Insert 4×VSV-G tag between the 3′ end of PcrH and the 2×Myc tag
H-4VG-2M-3-1	CGGCTGGGCAAGTACACCGATATCGAAATGAACAGGCTCGGGAAGTACACCGACATTGAAATGAATCGCCTAGGAAAGGAACAAAAGTTGATTTCTGAAG

^aLowercase letters indicate restriction sites.

RECC assay.

An assay to Re-Establish Calcium Control (RECC), a modified secretion assay, was performed as previously described (17). Bacteria were grown in LB-MC with the addition of EGTA (5 mM final concentration) to induce effector secretion and upregulate production of the T3SS and its secretion substrates. Cultures were removed at an optical density at 600 nm (OD₆₀₀) of 0.4 to 0.6, bacteria were pelleted and either resuspended in prewarmed LB-MC (+calcium condition) or LB-MC with 5 mM EGTA (−calcium condition). The cultures were then incubated for an additional 25 min and chilled on ice for 5 min before pelleting the bacteria from 1 ml of culture and removing 0.5 ml of the supernatant into a separate tube. The remaining supernatant was discarded and the bacteria pellet was resuspended in 1× sodium dodecyl sulfate (SDS) sample buffer, normalized to an OD₆₀₀ of 10. The supernatant proteins were precipitated with trichloroacetic acid (10% final concentration), pelleted by centrifugation, washed once with acetone, dried, and resuspended in 1× SDS sample buffer to match the concentration of the cognate cell-pellet sample. Complementation plasmids were induced with 500 μM IPTG (isopropyl-β-d-thiogalactopyranoside).

Western blot.

Protein was separated by SDS-PAGE on a 12% polyacrylamide gel (Bio-Rad), transferred to polyvinylidene difluoride (PVDF) membrane, blocked with 5% milk, and probed with specific antisera. Primary antibodies were detected using horseradish peroxidase (HRP)-conjugated secondary antibodies and a chemiluminescent detection reagent (WesternBright Quantum [Advansta]). Antibodies to PopB, PopD, and ExoT were generated in rabbits using a commercial service (Covance). The antisera were further purified using an affinity purification protocol. Antibodies directed against RpoA (Neoclone) and Myc tag (Thermo) were obtained commercially. Blots were imaged using a GE ImageQuant LAS 4000 digital imaging system. Scanned images were processed for brightness and contrast using only the levels function of Adobe Photoshop applied to the entire image before cropping.

pcrH mutant library screen.

A pcrH open reading frame was randomly mutagenized by PCR (Taq with a mutagenic deoxynucleoside triphosphate mix in which G and C are present at a 10-fold-lower concentration than are A and T) and cloned into pACtr-VSVG-ZifAP as a NotI/HindIII fragment. This pACtr-VSVG-ZifAP-pcrH* library consisted of ∼300,000 clones with 0.6% religation. The pcrH* library was cotransformed into reporter strain E. coli KDZif1ΔZ with pBRωGP-pcrH and plated on LB medium with 20 μg of tetracycline/ml, 60 μg of carbenicillin/ml, 25 μM IPTG, 300 μM 2-phenylethyl β-d-thiogalactoside (TPEG), and 40 μg of X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside)/ml. Pale blue colonies were picked after growing overnight at 37°C. 118 light blue colonies were restruck to verify pale blue color and picked into 96-well plates with 200 μl of LB medium. The isolates were grown overnight, and 150 μl of each column on the 96-well plate (8 isolates) was pooled and miniprepped. DNA from each pool was retransformed with plasmid pBRωGP-popD and screened for interaction with PopD to rule out unstable/misfolded or prematurely terminated PcrH mutant proteins. Plasmid DNA was isolated from 12 blue colonies, diluted, and retransformed into DH5α to reisolate the individual pACtr-VSVG-ZifAP-pcrH* mutant plasmids. The nature of the mutation was determined by DNA sequencing. Two mutations were identified, A61S and E103G. More careful characterization of these mutants by two-hybrid analysis, both with regard to PcrH-dimerization and PopD binding, demonstrated that the E103G mutant was also somewhat defective in binding to PopD (data not shown), suggesting that the protein is destabilized by the introduction of the glycine residue and was therefore discarded.

β-Galactosidase assay.

E. coli KDZif1ΔZ was cotransformed with plasmids containing Zif and ω fusions and grown overnight in LB medium with 20 μg of tetracycline/ml and 60 μg of carbenicillin/ml. Where indicated, popD was coexpressed from pCLARA maintained with 50 μg of spectinomycin/ml. The overnight cultures were diluted 1:300 into fresh LB medium with or without IPTG induction (25 μM) and 0.4% arabinose to induce expression of popD where indicated. Strains were grown to the mid-logarithmic phase. After cell permeabilization with chloroform and SDS, the β-galactosidase activity was assayed as previously described (18). All assays were performed on three consecutive days. The data are presented as means ± the standard deviations.

Protein purification and in vitro cross-linking.

PcrH, PcrH(A61S/L65M), PcrH-Myc, and PcrH(A61S/L65M)-Myc were produced as His-tagged proteins in E. coli BL21 DE3/CodonPlus-RP (Stratagene) using a T7 expression system. His-tagged proteins were purified from E. coli lysates under native conditions using nickel-affinity purification (Qiagen). The His tags were removed by cleavage with a His-tagged version of TEV protease (Invitrogen), and the cleaved proteins were recovered by passing the reaction mixes over a nickel column and recovering the flowthrough. Proteins were then dialyzed against a buffer containing 50 mM sodium phosphate (pH 8) and 100 mM NaCl and concentrated using Amicon Ultra centrifuge filters (3-kDa cutoff). The protein concentration was determined by using a Bradford protein assay.

For cross-linking experiments, all proteins were adjusted to a concentration of 0.5 mg/ml. Portions (10 μl) of the proteins solution was mixed with 1.1 μl of 20 mM ethylene glycol bis[succinimidylsuccinate] (EGS) in dimethyl sulfoxide (DMSO) or 1.1 μl of DMSO (no cross-linker control) and incubated at room temperature in the dark for 20 min. The reaction was stopped by the addition of 1.2 μl of a 500 μM glycine solution, followed by incubation for 5 min at room temperature, at which point 4 μl of 4× SDS sample buffer was added to each sample. All samples were boiled for 10 min before loading 5 μl on a 12% TGS-precast gel (Bio-Rad). Proteins were stained using SimplyBlue SafeStain (Invitrogen).

In vivo EGS cross-linking.

Overnight cultures were diluted 1:300 into 3 ml of LB-MC with 15 μg of gentamicin/ml and 5 mM EGTA, protein production was induced with IPTG [pcrH-Myc, 50 μM in popBD⁺ and 100 μM in ΔpopBD strain backgrounds; pcrH(A61S/L65M)-Myc, 150 μM in popBD⁺ and 1 mM in ΔpopBD strain backgrounds; pcrH-4VG-Myc, 100 μM in popBD⁺ and 200 μM in ΔpopBD strain backgrounds; as well as pcrH-Myc and pcrH(A61S/L65M)-Myc with 500 μM IPTG to mimic the conditions in Fig. 3], and cultures were grown to mid-logarithmic phase. Cells were pelleted by centrifugation and then resuspended in PBS-MC (phosphate-buffered saline with 10 mM MgCl₂ and 0.5 mM CaCl₂) with 2 mM phenylmethylsulfonyl fluoride and 5 mM EGS or with an equivalent amount of DMSO in the no-cross-linker control sample. The bacteria were incubated for 20 min at room temperature in the dark, at which point the reaction was stopped through the addition of glycine (50 mM final concentration). The samples were incubated for an additional 5 min at room temperature. The bacteria were then pelleted by centrifugation, washed once with 1 ml of PBS-MC, pelleted again, and resuspended in 1× SDS sample buffer. Samples were normalized for OD₆₀₀, boiled, and separated by SDS-PAGE. PcrH-Myc and PopB were detected by Western blotting.

Fig 3.

Dimerization is not required for PcrH function. The ability of the PcrH dimerization mutant A61S/L65M to stabilize PopB and PopD in the cell, as well as promoter their export via the T3SS was measured by RECC assay (a modified secretion assay [17]). The nature of the PcrH examined (Myc-tagged and/or dimerization mutant A16S/L65M) is indicated. Calcium was removed from the medium where indicated by the addition of the chelator EGTA. ExoT represents an effector, which is only exported after triggering of effector secretion. RpoA is a loading control. A Myc-tagged version of PcrH was used to detect the chaperone. The production of PcrH was induced with 500 μM IPTG in all instances.

PcrH stability determination.

Overnight cultures of PAO1 lacking pcrH or pcrH popBD harboring the indicated pcrH expression vector encoding Myc-tagged versions of PcrH were diluted 1:300 into 3 ml of fresh LB-MC with 15 μg of gentamicin/ml and 5 mM EGTA. Protein production was induced by adding IPTG to the growth medium [pcrH-Myc, 50 μM in popBD⁺ and 100 μM in Δ(pcrH popBD) strain backgrounds; pcrH(A61S/L65M)-Myc, 150 μM in popBD⁺ and 1 mM in Δ(pcrH popBD) strain backgrounds]. At mid-log phase, protein synthesis was stopped through the addition of 500 μg of tetracycline/ml to the medium. Portions (500 μl) of culture aliquots were removed at the time of tetracycline addition, as well as after 7.5, 15, and 30 min of incubation in the presence of the antibiotic and placed on ice. Cells were pelleted by microcentrifugation, the supernatants were removed, and the cell pellets were resuspended in 50 μl of 1× SDS sample buffer. Samples were subsequently boiled, separated by SDS-PAGE (12% TGX gel; Bio-Rad), transferred to PVDF membrane, and probed with mouse monoclonal antibodies to Myc (Thermo) and RpoA (Neoclone). In some instances, the primary antibodies were detected using a fluorescent secondary antibody (Sigma, catalog no. CF647) rather than an HRP-conjugated secondary antibody. Blots were imaged using a GE ImageQuant LAS 4000 digital imaging system, quantitated using ImageJ software (National Institutes of Health), and plotted using the Prism6 software package (GraphPad). The data points in Fig. 4 represent the averages of three independent experiments ± the standard deviations.

Fig 4.

Dimerization stabilizes PcrH. Myc-tagged versions of PcrH were produced either in the presence or absence of PopB and PopD. Protein synthesis was stopped by adding tetracycline to the medium. Aliquots were removed at 0, 7.5, 15, and 30 min, cells were pelleted, lysed, and separated by SDS-PAGE. PcrH-Myc levels were determined by Western blotting and normalized to the amount present at the 0-min time point. The average of three independent experiments was plotted, and the slopes of the curves were determined by fitting the data to a decay curve using Prism6 (GraphPad). A representative Western blot is shown below the graph. Expression of the pcrH-Myc constructs was induced with the following concentrations of IPTG: pcrH-Myc, 50 μM in popBD⁺ and 100 μM in ΔpopBD strain backgrounds; and pcrH(A61S/L65M)-Myc, 150 μM in popBD⁺ and 1 mM in ΔpopBD strain backgrounds.

RESULTS

To test the role of dimerization in PcrH function, we initially attempted to mutate residues along the outer, convex surface of the TPR-repeat domain of PcrH based on the published dimerization interface (11). This dimer conformation corresponds to the organization of the PcrH monomers in the asymmetric unit of the published PcrH crystal structure. An interaction between the amino-terminal domains of PcrH reminiscent of the SycD dimer had also been observed between monomers of adjacent asymmetric units, but this interaction was disfavored by the authors due to the fact that it involved fewer residues than the “back-to-back” dimer. However, we were never able to generate a mutant that interfered with dimerization without also affecting binding to PopD, suggesting that the mutations had unforeseen structural consequences that prevented substrate binding as well (data not shown). We therefore used a bacterial two-hybrid system (19), a genetic technique, to identify a mutation that prevents PcrH dimerization while still allowing binding to PopD.

The Zif-Omega two-hybrid system allows the detection of homomeric dimers by fusing one interaction partner to the monomeric DNA-binding domain Zif and the other to the monomeric RNA-polymerase subunit omega (20). If the two fusion proteins interact, the interaction results in recruitment of RNA polymerase to a test promoter, activating lacZ expression. Consistent with the in vitro evidence for dimerization of PcrH, coexpression of Zif-pcrH and omega-pcrH resulted in activation of the test promoter. Being able to detect the interaction by two-hybrid analysis allowed us to mutagenize one of the interaction partners (Zif-PcrH) and screen for mutants that abrogate the interaction (pale blue on X-Gal plates). Pale blue colonies were picked onto a second plate pooled, and plasmids were isolated from the pool of pale colonies. The plasmids were diluted and cotransformed with a construct specifying the production of a PopD-omega fusion protein and again screened for activation of β-galactosidase activity. Mutations that interfere with PcrH dimerization but do not otherwise adversely affect folding of mutant PcrH should retain the ability to bind to PopD. Indeed, we isolated one mutant, A61S, which interfered with PcrH-PcrH interaction but did not adversely affect the PcrH-PopD interaction (Fig. 1). Interestingly, A61, which is conserved in SycD, is one of the residues implicated in the SycD “head-to-head” dimer (9). We therefore mutated the second residue implied in the interaction, leucine 65, to determine whether we could increase the defect in dimerization observed with the A61S mutation. Since changing A61 and L65 to glutamic acid in SycD resulted in a protein that failed to stabilize YopD in Y. enterocolitica and therefore likely prevented binding of YopD by SycD, we opted to replace L65 of PcrH with methionine, simply increasing the bulk of the side chain, but not changing the charge of the residue. By itself, the L65M mutation only had a minor effect on dimerization (Fig. 1A and B). The A61S/L65M double mutant retained the ability to bind to PopD (Fig. 1C), while severely interfering with the PcrH-PcrH interaction, as assayed by two-hybrid analysis (Fig. 1A and B). Taken together, these data suggest that PcrH can dimerize in vivo and that it dimerizes via its amino terminus, akin to the published SycD dimer.

Fig 1.

PcrH dimerization mutants. (A) Two-hybrid analysis to detect the interaction between PcrH monomers. Wild-type PcrH was fused to the omega subunit of RNA polymerase. Wild-type PcrH or the indicated mutant protein were fused to Zif. An interaction was detected by coproducing both fusion proteins in an E. coli strain harboring a reporter gene which depends on the interaction of the two fusion proteins for recruiting RNA polymerase to the test promoter. Fusion protein synthesis was induced with 25 μM IPTG. (B) Two-hybrid analysis was performed as in panel A, except that both omega-PcrH and Zif-PcrH were mutated. (C) The ability of PcrH and its respective dimerization mutants to bind to PopD was tested by fusing the mutant proteins to Zif and coproducing omega fused to PopD. Expression of the indicated Zif-PcrH fusion was assayed by Western blotting using an α-VSV-G antibody, which detects the VSV-G epitope tag joining the Zif and PcrH fusion partners. RpoA was included as a loading control. The respective blots are shown below each β-galactosidase activity graph.

Although PcrH forms an unstable dimer on its own in vitro, it is stabilized in the monomeric form when bound to PopD. We therefore tested whether the dimer formed in our two-hybrid system is sensitive to coexpression of popD. Indeed, the PcrH-PcrH interaction was prevented when PopD was supplied from a third plasmid (Fig. 2A). Coexpression of popD did not interfere with the unrelated PcrG-PcrV interaction (Fig. 2A). PopD levels differed when popD was coexpressed while assaying the PcrH-PcrH interaction as opposed to the PcrG-PcrV interaction, most likely because PopD requires PcrH for stability (4).

Fig 2.

PcrH dimerization. (A) PcrH dimerization was assayed by two-hybrid analysis. popD was coexpressed from a third plasmid, pCLARA, which allowed for arabinose-inducible expression of the translocator gene. As a control, expression of PopD was also assayed for the PcrG-PcrV interaction. Expression of the Zif-fusion protein (α-VSV-G) and PopD were monitored by Western blotting. RpoA was included as a loading control. Production of the Zif and omega fusion proteins was induced with 25 μM IPTG. (B) PcrH dimerization was assayed using purified protein in vitro. PcrH, PcrH(A61S/L65M), PcrH-Myc and PcrH(A61S/L65M)-Myc were suspended at a concentration of 0.5 mg/ml and cross-linked using ethylene glycol bis[succinimidylsuccinate] (EGS). Cross-linked proteins were separated by SDS-PAGE and stained for total protein. (C) PcrH dimerization in P. aeruginosa was assayed by producing the indicated chaperone either in the presence or absence of PopB and PopD and cross-linking total protein with EGS. The position of monomer and cross-linked dimer as well as PcrH-PopB species is indicated by arrowheads. In the right panel, PopB was detected to identify cross-linked PcrH-PopB species. The IPTG concentrations used to express the indicated fusions were as follows: pcrH-Myc, 50 μM in popBD⁺ and 100 μM in ΔpopBD strain backgrounds; pcrH(A61S/L65M)-Myc, 150 μM in popBD⁺ and 1 mM in ΔpopBD strain backgrounds; and pcrH-4VG-Myc, 100 μM in popBD⁺ and 200 μM in ΔpopBD strain backgrounds; as well as pcrH-Myc and pcrH(A61S/L65M)-Myc with 500 μM IPTG in a popBD⁺ strain to mimic the conditions in Fig. 3.

We had previously used a Myc-tagged version of PcrH to examine its function in vivo. In order to determine whether the Myc tag interferes with dimerization, we purified PcrH, as well as a C-terminally Myc-tagged version of PcrH and the corresponding dimerization mutants from E. coli. We trapped PcrH dimers using the primary amine-specific cross-linking reagent ethylene glycol bis[succinimidylsuccinate] (EGS) and resolved the cross-linked proteins by SDS-PAGE. The addition of the C-terminal Myc tag had no effect on dimer formation. Although dimer formation was incomplete, the result mirrors the efficiency with which the IpgC dimer could be trapped using BS₃ as a cross-linker in vitro (14). As with the two-hybrid analysis, the dimer was almost completely abolished by the introduction of the A61S/L65M double mutation (Fig. 2B).

We next determined whether PcrH also dimerizes in P. aeruginosa and whether this dimerization involves residues A61 and L65. We produced Myc-tagged versions of either wild-type PcrH or the A61S/L65M mutant protein in P. aeruginosa. Since binding of PcrH to PopD dissolved the PcrH-dimer in vitro, we produced these proteins either in the presence of absence of PopB and PopD. Expression of PopB and PopD was upregulated by removing calcium, which triggers effector secretion and upregulates production of the T3SS and its secretion substrates (21). Total protein was cross-linked using 5 mM EGS to trap protein-protein interactions and separated by SDS-PAGE, followed by Western blotting to detect the Myc-tagged PcrH. We detected a weak cross-linked species corresponding to the approximate predicted molecular mass of the PcrH dimer in the case of wild-type PcrH (Fig. 2C). In order to confirm that the cross-linked protein represents a dimer of PcrH, we examined dimer formation using a size-tagged version of PcrH. We produced a version of PcrH in which four copies of the 11-amino-acid VSV-G epitope tag had been inserted between PcrH and the C-terminal Myc tag (PcrH-4VG-Myc). The four VSV-G tags increased the apparent molecular mass of PcrH-Myc by 10 kDa. The dimer of PcrH-4VG-Myc displayed an apparent molecular mass of 60 kDa, which represents a 20-kDa increase compared to the 40 kDa of the putative PcrH-Myc dimer, indicating that the cross-linked species we detected indeed represents a cross-linked dimer of PcrH (Fig. 2C). Coexpression of PopB and PopD prevented dimer formation, indicating that substrate binding dissolves the dimer in P. aeruginosa as well. The dimer was completely abolished when the A61S/L65M mutant protein was assayed. The dimer species only represents a very small portion of the PcrH population. Since the dimer formation by PcrH in vitro requires a high protein concentration (11), it may simply be that we did not express enough PcrH to allow efficient dimer formation. It should be noted that the protein was produced somewhat below the level that is needed for full complementation in these experiments. This allowed us to normalize the amount of PcrH produced among the different conditions due to the instability of the dimerization mutant in the absence of PopB and PopD (see below). We also tested dimerization when PcrH was produced at a level that fully complements the null mutant. Here production of PcrH-Myc and PcrH(A61S/L65M)-Myc was induced with 500 μM IPTG. Although we were able to detect a PcrH dimer in the case of wild-type PcrH under these conditions, we could not detect a dimer of the A61S/L65M mutant protein (Fig. 2C). Presumably production of PcrH induced with 500 μM IPTG outstrips production of PopB and PopD, thereby allowing the dimer to form. Notably, we could also detect heterodimers of PopB and PcrH in these experiments. While we could not detect a PopD-PcrH heterodimer, this could be due to the fact that PopD outcompetes PopB for export (e.g., as seen in Fig. 3), thereby making the PopD-PcrH dimer more transient in nature, or that PopD does not have an appropriately positioned residue with a primary amine to allow for cross-linking to PcrH. Taken together, these data demonstrate that PcrH dimerizes in P. aeruginosa and that the dimer is dissolved by binding of PcrH to its cognate cargo, akin to the dissolution of the dimer observed in vitro (11). Moreover, dimerization is abrogated by mutation of residues A61 and L65, suggesting that the PcrH monomers form a “head-to-head” dimer in vivo, akin to that observed in the SycD crystal structure.

In order to test whether dimerization of PcrH is required for function, we complemented a pcrH-null mutant by producing Myc-tagged versions of wild-type PcrH or the A61S/L65M mutant and then assayed stabilization of PopB and PopD in the cytoplasm of P. aeruginosa, as well as their export via the T3SS (Fig. 3). As noted previously, the absence of PcrH destabilizes PopB and PopD but does not affect the stability or export of the effector protein ExoT. The Myc-tagged PcrH was fully functional, both stabilizing intracellular PopB and PopD and promoting their export into the culture supernatant. The dimerization mutant, PcrH(A61S/L65M), was similarly fully active, indicating that the inability of PcrH(A61S/L65M) to dimerize did not interfere with its chaperone function.

In the course of producing the Myc-tagged version of PcrH for the dimerization experiment, we noticed that synthesizing the A61S/L65M mutant protein at a level comparable to the corresponding wild-type protein required a higher level of the inducer IPTG. This observation suggested to us that the mutant protein may be less stable than wild-type PcrH. To test this hypothesis, we produced Myc-tagged version of wild-type PcrH or the A61S/L65M mutant protein either in the presence or in the absence of PopB and PopD and monitored the level of the protein over time after stopping protein synthesis through the addition of tetracycline to the growth medium (Fig. 4). We found that PcrH is stable both in the presence and in the absence of PopB and PopD (with approximate half-lives of 55 and 36 min, respectively). The A61S/L65M mutant protein, on the other hand, was stable in the presence of PopB and PopD (with an approximate half-life of 61 min) but decayed rapidly, with an approximate half-life of 18 min when expressed in a popBD-null mutant strain. These data suggest that the formation of the PcrH dimer stabilizes the protein when not bound to its cognate cargo.

DISCUSSION

Virulence-associated type III secretion systems rely on pore-forming translocator proteins to form a pore in the host cell membrane through which effector proteins are delivered into the cytoplasm of the targeted cell. These translocator proteins rely on specific signals to be exported before effectors and rely on a specific class of chaperone proteins for stability in the bacterial cytoplasm, as well as efficient export via the T3SS (6, 7, 22). These translocator chaperones contain a series of TPR repeats which resemble the cupped palm of a hand that cradles the chaperone-binding domain of the bound translocator protein (10, 11). The structure of these chaperones and their mode of interaction with their cognate translocator proteins are highly conserved. Although all translocator chaperones analyzed to date form dimers in solution, based on the crystallographic evidence the nature of the dimerization interface appears to differ between homologs. In the case of the P. aeruginosa translocator chaperone, PcrH, the dimer had been proposed to involve the convex portion of the PcrH monomer, opposite of the binding site for the translocators PopB and PopD (11). In the case of the closely related SycD protein of Y. enterocolitica and LcrH of Y. pestis, the dimerization is thought to instead be mediated by an interaction of the amino-terminal domain of two SycD/LcrH monomers (9, 12). Finally, in the case of S. flexneri IpgC, the dimer had been proposed to involve an asymmetric interaction in which the amino-terminal domain of one monomer interacts with the convex portion of the TPR “palm” of the second monomer (10). Mutating residues A61 and L65 of SycD to glutamic acid resulted in a protein that failed to dimerize in vitro and was unable to complement a sycD mutant in vivo (9). These data were interpreted to mean that dimerization is important for SycD function but left open the possibility that changing A61 and L65 to glutamic acid inadvertently also disrupted SycD structure to a degree where it interfered with translocator binding, thereby completely inactivating the protein.

To assess the role of dimerization in PcrH function, we performed a genetic screen in E. coli to identify mutants of PcrH that interfere with dimerization while not interfering with the binding of PopD. The screen resulted in the isolation of a mutant in which alanine 61 of PcrH had been changed to serine. Combining the A61S mutation with a L65M substitution resulted in a protein in which PcrH dimerization had been abolished without interfering with PopD binding. The location of the mutations suggests that PcrH dimerizes via its amino-terminal domain, similar to the dimerization observed in the case of SycD and LcrH. Although tyrosine 65 of IpgC, the residue corresponding to L65 of PcrH, is involved in the asymmetric dimer reported by Lunelli et al. (10), phenylalanine 61 was not identified as an important contributor to the IpgC asymmetric dimer. We favor the head-to-head dimer for PcrH because (i) it has been observed, at least as a minor species, with every chaperone crystallized to date (9, 11, 14) and (ii) the A61S mutation had a much greater effect on dimer formation by PcrH compared to the L65M change.

PcrH dimerizes in P. aeruginosa as well, and introducing the A61S/L65M double mutation resulted in the loss of dimerization. Production of translocator proteins prevented dimerization of PcrH in the E. coli two-hybrid system and in P. aeruginosa. These data suggest that the PcrH dimer is dissolved in P. aeruginosa by binding of PcrH to its cognate cargo, as had been observed previously in vitro.

The PcrH(A61S/L65M) mutant protein stabilized PopB and PopD in P. aeruginosa and promoted its export via the T3SS, demonstrating that dimerization is not required for PcrH function. Notably, in the absence of PopB and PopD, the mutant protein was less stable compared to wild-type PcrH, suggesting that the function of the PcrH dimer is to stabilize PcrH when not bound to its cognate secretion substrate.

It is interesting that PcrH transitions between dimer and monomer forms, depending on whether or not it is bound to its export substrate. Since the dimerization interface is in close proximity to the region of the protein that is required for targeting PopB and PopD for export, which involves residues A49 and G50 of PcrH (7), it is tempting to speculate that the PcrH dimer has to be dissolved in order to expose the targeting domain of PcrH. A mutation that locks PcrH in the dimer conformation when bound to PopB or PopD will be required to test this hypothesis directly. As it stands, our data suggest that PcrH dimerizes in P. aeruginosa in a head-to-head fashion involving residues A61 and L65. The dimer is not required for binding and stabilizing PopB and PopD, nor is it required for promoting their efficient export. Instead, dimerization helps stabilize free PcrH in the cell.