Interactions between the Lipoprotein PilP and the Secretin PilQ in Neisseria meningitidis

Bacterial strains and growth conditions.Meningococcal strains were grown overnight in 5% CO₂ at 37°C on GC agar plates supplemented with IsoVitaleX (Becton Dickinson, United States). When required, kanamycin at a final concentration of 100 μg/ml was added. Escherichia coli strain ER2566 (New England Biolabs, United Kingdom) was used for plasmid propagation and was grown in LB medium or on LB plates containing kanamycin (50 μg/ml) or ampicillin (100 μg/ml) at 37°C. The bacterial strains and plasmids employed in this study are listed in Table 1.

Search for signature motifs.A search for functional domains and signatures in the meningococcal MC58 pilP deduced amino acid sequence (NMB1811) was carried out using DOLOP (31) and PROSITE (28). The MC58 PilP sequence was assessed for disorder prediction on DISPROD2 (58) and VSL1 (56).

Cloning and overexpression of PilP.The pilP gene was PCR amplified from MC58 genomic DNA using primers as listed in Tables S1 and S2 in the supplemental material. The gene fragments encoding full-length, N-terminal, C-terminal, and internal deletions of the pilP gene were cloned into the vector pET28-b(+) (Novagen, United States), with the C-terminal six-His tag in frame, or into the vector pCR7 TOPO (Invitrogen, United States), with an N-terminal His tag. For the generation of the internal deletion mutants, splicing by overlap extension (SOEing) PCR was employed (27). All recombinant proteins were overexpressed in E. coli ER2566 (New England Biolabs, MA) or E. coli BL21(DE3)* (Invitrogen, United States).

Construction of pilP mutants.The pilP knockout mutant was made by deleting 81 bp beginning from the start codon and inserting a kanamycin resistance cassette (Kan^r) into a SalI site created with the primers SVB25 and SVB26. The plasmid obtained, pUP6-pilP_Δ1-27::kan, was transformed into meningococcus strain M1080 by natural transformation using kanamycin resistance selection (55). A frameshift mutation was constructed by SOEing PCR using the primers SVB24, SVB77, SVB27, and SVB78 (see Tables S1 and S2 in the supplemental material). A single base at the start codon was deleted and a restriction site (EcoRI) was created to verify the frameshift mutation. The PCR product was digested with BamHI and ligated into the BamHI site of pUP6 to create the plasmid pUP6::pilP_fs. Plasmid pUP6::pilP_fs was transformed into N. meningitidis M1080, and transformants were detected by PCR after nonselective transformation. For the complementation studies, cloning of pilP into vector p2/16/1 (59) was performed using the primers SVB71-SVB74 (see Tables S1 and S2 in the supplemental material) and the resulting product was digested with SacI and cloned into a unique SacI site in the vector.

Construction of a meningococcal in-frame PilP-His fusion. (i) Genetic construct.A meningococcal in-frame pilP-His fusion was designed to encode a six-histidine tag in the predicted loop region at amino acid 74. For the generation of the construct encoding the internal PilP-His fusion, PCR SOEing was employed. The 5′ and 3′ fragments were produced with the primer combinations SVB31/SVB19 and SVB20/SVB32 (see Table S1 in the supplemental material), respectively. The final PCR product was amplified with the external primers SVB31 and SVB32, digested with BamHI, and ligated into the BamH1-digested vector pUP6 (47). The final vector, pUP6-pilP _74-6xHis, was used for the transformation of wild-type meningococcus M1080.

(ii) Nonselective transformation.Agglutinating colonies were densely streaked on blood agar plates and grown for 4 to 6 h at 37°C in 5% CO₂. Cells were harvested and resuspended in prewarmed GC broth supplemented with IsoVitaleX (Becton Dickinson, United States) and 7 mM MgCl₂. Plasmid DNA was added to the cell suspension. The sample was briefly mixed and incubated without agitation for 30 min at 34°C. Prewarmed supplemented medium was added, and the suspension was spread on blood agar plates and incubated overnight.

(iii) PilP-His fusion screening.Crude DNA samples for PCR testing from clones were made by freezing and thawing in distilled water and removal of debris by centrifugation. Specific primers SVB22 and SVB23 were designed to produce small PCR products covering the points of mutation (see Table S1 in the supplemental material).

Purification of recombinant PilP and PilQ proteins and pilus fibers.The full-length and partial PilP and PilQ recombinant proteins were overexpressed in E. coli and purified to homogeneity. All recombinant PilP proteins were affinity purified using Ni-nitrilotriacetic acid (NTA) agarose (QIAGEN, Germany) or, for PilP_Δ1-19 only, Talon cobalt metal affinity resin (Clontech, United States) under native conditions according to the manufacturer's protocol. The eluted proteins, except PilP_Δ1-19, were immediately dialyzed against 50 mM sodium phosphate buffer (pH 7.0) containing 300 mM NaCl and 10% glycerol. PilP_Δ1-19 was subjected to a second purification step by anion exchange using a MonoQ HR 10/10 column (GE Healthcare, United Kingdom). Purification of recombinant PilQ proteins and meningococcus type IV pilus fibers was performed as previously described (13).

Native purification of PilQ complex and His-tagged PilP.The PilQ complex was purified from meningococcus outer membranes by detergent solubilization and size exclusion as previously described (14). For native PilP purification, meningococcus M1080-PilP_74-6xHis cells were grown on blood agar plates in 5% CO₂ at 34°C overnight, harvested in cold phosphate-buffered saline (PBS) buffer and collected by centrifugation at 4,000 × g for 15 min. The cell pellet was resuspended in cold Tris-buffered saline (pH 7.5) containing 1× EDTA-free complete protease inhibitor cocktail (Roche Applied Science, Germany) and passed twice through a French press at 25,000 lb/in². Cell debris was removed by two succeeding centrifugations at 4,000 × g for 10 min. The supernatant was collected, and the membrane fraction pelleted by centrifugation at 35,000 rpm for 35 min in an SW40Ti rotor (Beckman, Ireland). The membrane pellet was dissolved in lysis buffer (10 mM imidazole [Sigma, Germany], 300 mM NaCl, 50 mM NaH₂PO₄, 0.5% Triton X-100, 10 mM β-mercaptoethanol [Sigma, Germany], 3 mM MgCl₂, pH 8.0) containing 1× EDTA-free complete protease inhibitor cocktail and incubated with rotation at 4°C overnight. The dissolved membrane fraction was centrifuged at 4,000 × g for 15 min, and the clear supernatant was loaded onto a Ni-NTA agarose column (QIAGEN, Germany). Bound proteins were eluted with elution buffer (200 mM imidazole, 300 mM NaCl, 50 mM NaH₂PO₄, pH 8.0) and immediately dialyzed against 50 mM sodium phosphate buffer (pH 7.0) containing 300 mM NaCl and 10% glycerol.

Antibody production.Rabbit polyclonal antibodies were raised against the purified full-length recombinant PilP proteins to generate antiserum K824 as previously described (54, 55). Antisera against PilQ were produced as previously described (22, 54).

Reverse transcription-PCR (RT-PCR).Total RNA was isolated from MC58 cells using TRIzol (GIBCO Life Technologies, United States). After treatment with DNase I, the RNA concentration was spectrophotometrically quantitated, and the quality was assessed by agarose gel electrophoresis. RT was performed in a final volume of 30 μl containing 1× RT buffer, 7 units AMV reverse transcriptase (Promega, United States), 28 units RNasin RNase inhibitor (Promega, United States), 15 mM deoxynucleoside triphosphates (GE Healthcare, Sweden), 10 μM reverse primer pilQ-6485, and 25 ng RNA template. The primer and RNA were denatured at 94°C for 1 min and immediately cooled on ice for 5 min before being added to the reaction mixture. The reaction was incubated at 42°C for 1 h. One microliter of 10 μg/ml RNase A was added to the controls prior to incubation, and samples were incubated at room temperature for 5 to 20 min. RT products were incubated at 98°C for 3 min to separate the strands. Five microliters of the cDNA-containing product was amplified in a 50-μl PCR mixture containing 1× PCR buffer and 1 unit Taq polymerase (BioLine, CA), as well as 10 μM of forward primer pilQ-5675 or pilP-5075 and reverse primer pilQ-6485. After denaturation for 5 min at 95°C, PCR was performed for 30 cycles (30 s at 95°C, 30 s at 59°C, and 60 s at 72°C) with a 7-min final extension of the products at 72°C. Controls were included to verify the absence of DNA in the RNA preparation, as well as nonspecific RT in the absence of the reverse primer. The PCR products were analyzed by agarose gel electrophoresis.

Quantitative transformation assay.Natural competence for DNA transformation was quantified by counting the number of CFU obtained after the transformation with plasmid pSY6 harboring a nalidixic acid-resistance-conferring gene (49). Transformation was performed as described above for nonselective transformation, except that 1 μg/ml nalidixic acid was used for selection. The ratio of nalidixic acid-resistant CFU to the total number of CFU obtained without antibiotic selection yielded the transformation rate.

SDS-PAGE, immunoblotting, and solid-phase overlay assays.Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), immunoblotting, and solid-phase overlay assays in the form of far-Western analysis were carried out as previously described (13). For phenol treatment, equal volumes of the sample and phenol were mixed by vortexing for 2 min, and samples were further vortexed after adding 2 volumes of ice-cold acetone. The proteins were pelleted by centrifugation at 12, 000 × g for 15 min after overnight incubation at −20°C. The pellet was washed twice with 70% ethanol before it was resuspended in SDS-PAGE sample buffer.

Isolation of outer membrane vesicles (OMV).Bacteria were grown on blood agar plates at 37°C for 15 h, harvested in serotype antigen buffer (0.2 M LiCl, 0.1 M NaAc, pH 5.8) (21) and inactivated at 60°C for 30 min. A bacterial suspension with an optical density at 600 ng of 20 was prepared and shaken with 2-mm glass beads at 230 rpm for 15 min at room temperature. Cell debris was removed by two rounds of centrifugation at 4,000 × g for 20 min and 18,000 × g for 15 min, respectively. The supernatant was ultracentrifuged at 140,000 × g for 90 min and the pellet resuspended in H₂O and ultracentrifuged again at 140,000 × g for 90 min. The resulting pellet was resuspended in H₂O and analyzed by immunoblotting.

Separation of outer and inner membranes by sucrose gradient centrifugation.Meningococcal cells were grown on blood agar plates at 37°C in 5% CO₂, harvested in PBS buffer, and washed three times with PBS. The cells were resuspended in 50 mM Tris buffer, pH 8.0, containing 50 μg RNase and DNase (Sigma Aldrich, United Kingdom) and processed twice in a French press at 15,000 lb/in². The cell debris was removed by two centrifugations at 10,000 × g for 10 min. Sucrose gradient analysis was carried out as described by Masson and Holbein (33), with slight modifications. The supernatant was transferred onto two layers of sucrose consisting of 3 ml of 55% (wt/wt) sucrose and 4 ml 15% sucrose in H₂O with 3 mM EDTA, pH 8.0, and centrifuged at 217,000 × g for 5 h at 4°C in an SW40Ti rotor (Beckman, Ireland). The membrane fraction positioned at the interface of the two sucrose layers was collected and diluted in H₂O with 3 mM EDTA, pH 8.0, to a sucrose concentration of 30%. The membrane fraction was then applied to a discontinuous sucrose gradient, consisting of 3 ml of 50, 45, 40, and 35% sucrose in 3 mM EDTA, pH 8.0. This gradient was centrifuged in an SW40Ti rotor for 35 h at 180,000 × g at 4°C and fractionated from the top into 1-ml fractions. The density of each fraction was measured by using a refractometer from Precision Instruments (Bellingham and Stanley, United Kingdom). An aliquot of each fraction was diluted in 3 mM EDTA and ultracentrifuged at 35,000 rpm in an SW40Ti rotor for 3 h. The resulting pellet was resuspended in 50 μl 3 mM EDTA, and the level of d-lactate dehydrogenase (LDH) activity was measured as described by Osborn et al. (39).

PilP-PilQ interaction image analysis.Sample preparation, cryonegative staining, TEM, and image analysis were carried out as described previously (13). Natively purified recombinant PilPΔ_1-19 (100 μg/ml) was incubated with PilQ complex purified from outer membranes, with or without Ni-NTA-nanogold (Nanoprobes, United States), with agitation for 24 h at 4°C. Samples were then centrifuged at 13,000 rpm in a benchtop centrifuge for 5 min before being prepared for cryonegative staining as previously described (9). Grids were placed in an Oxford system cryostage, and data were recorded at ∼100 K in a FEG CM200 used in conjunction with a Gatan 4K charge-coupled-device (CCD) camera. Individual gold-labeled particles were interactively selected in 64-pixel boxes (Å/pixel = 3.1) by using BOXER. Following contrast transfer function correction, data were contrast normalized, centered, and low-pass filtered to a 25-Å resolution. After a 190-Å circular mask was applied to the data, six rounds of iterative refinement were performed in C4 symmetry, using our previously calculated 3D structure of PilQ (14) filtered to a 25-Å resolution as a start model.

Analysis of PilP structure.In the pathogenic members of the Neisseriae, the pilP gene is located downstream of pilO and upstream of pilQ (Fig. 1A), and these genes all encode proteins required for pilus biogenesis (19, 32). In order to identify functional domains or signatures of the PilP protein, computational predictions were made based on the MC58 pilP deduced amino acid sequence. According to the DOLOP database (31), the PilP signal sequence was predicted to be 16 amino acids long, with a C-terminal lipobox (signal peptide cleavage site), LSAC, at positions 13 to 16 (Fig. 1B). In addition, a prokaryotic membrane attachment site (PROSITE entry PS00013) was identified, suggesting that PilP might be anchored in the membrane. A proline-rich region was identified from amino acid 40 to 60 (Fig. 1B). Proline-rich regions tend to be very flexible and are often implicated in protein-protein interaction domains (2). Previous NMR studies of a PilP fragment spanning residues 20 to 181 and a 3D structure determination of the C-terminal region covering residues 69 to 181 (24) showed a single folded domain from residue 85 to 158; the remaining N-terminal (amino acids 25 to 84) and C-terminal (amino acids 159 to 181) regions are disordered in solution. The folded domain is a novel β-sandwich structure, made up of a single 3₁₀-helix and seven β-strands. Based on these results, a scheme for the generation of pilP subclones was devised for the production of full-length and truncated forms of recombinant PilP (Fig. 1C).

• Open in new tab
• Download powerpoint

FIG. 1.

Gene organization and pilP constructs. (A) Chromosomal organization of the pilP locus of N. meningitidis. (B) Structural features of PilP. The lipobox and proline-rich regions are shaded in gray. Diagonal and dotted boxes indicate disordered and folded regions, respectively. (C) Schematic representation of the PilP clones used in this study. Internal deletions of the PilP protein are shaded in gray. Positions of the six-His tag are marked with H's, and the kanamycin cassette insertion in N. meningitidis M1080 pilP is also indicated. Relative molecular masses in kDa are given. na, not applicable; *, the PilP_Δ1-19 product contains an N-terminal His tag with 35 amino acids derived from the vector.

PilP is expressed in all meningococcal strains examined.Using rabbit antibodies raised against full-length recombinant PilP, PilP expression was examined in 16 clinical meningococcal isolates (data not shown). PilP was present in all of these strains, indicating that it is an important cellular component. When recombinant PilP proteins were reacted with rabbit antiserum raised against the full-length recombinant PilP, all partial proteins displayed strong reactivity, except the C-terminal PilP_Δ1-77 (Fig. 2A). An in-frame PilP polyhistidine insertion construct was made and introduced into the meningococcal genome in order to purify native PilP from meningococcal membranes. When the cell lysate from this meningococcus M1080-PilP_74-6xHis was tested by immunoblotting, the migration of PilP_74-6xHis (Fig. 2B, lane 2) was retarded relative to the migration of the wild type (Fig. 2B, lane 1), consistent with its increased molecular mass. The presence of the polyhistidine tag was demonstrated with anti-His antibody (Fig. 2B, lane 4). Based on the results of the immunoblot analysis of truncated PilP proteins, the N-terminal region seemed to be immunodominant with respect to denatured epitopes compared to the C-terminal region. The C terminus was recognized only after seminative gel electrophoresis (data not shown), suggesting that this region contains immunogenic epitopes in the tertiary structure. This observation agrees with the results of our NMR studies, which established that the C terminus of the PilP protein consisted of a folded domain but that the N terminus is disordered (24).

• Open in new tab
• Download powerpoint

FIG. 2.

PilP immunodetection. (A) Purified recombinant PilP proteins were immunodetected. Lanes: 1, full-length PilP; 2, PilP_Δ1-12; 3, PilP_Δ1-15; 4, PilP_Δ1-16; 5, PilP_Δ1-77; 6, PilP_Δ61-80; 7, PilP_Δ114-181. (B) Detection of M1080-PilP_74-6xHis with anti-PilP antiserum K824 (lanes 1 and 2) and anti-His (lanes 3 and 4). Lanes 1 and 3, wild-type M1080; lanes 2 and 4, M1080-PilP_74-6xHis. PilP protein is denoted with arrowheads I and II. Positions of the molecular mass standard proteins are indicated on the left in kDa.

PilP copurifies with the meningococcal inner membrane.We demonstrated that PilP was not present in OMV (Fig. 3). In order to determine the subcellular localization of PilP, we isolated the inner and outer membranes by using sucrose density gradient centrifugation. The cellular fractions were assessed for LDH activity as a marker for the inner membrane fraction and PilQ and OpcA as markers for the outer membrane fraction (Fig. 4). The LDH activity was detected at a lower density, corresponding to fractions 4 to 8, whereas PilQ and OpcA were mainly localized at a higher density, corresponding to fractions 10 to 13 (Fig. 4). The analysis of the fractions by immunoblotting, using the PilP antiserum, demonstrated that PilP was located in the inner membrane fractions. The presence of PilP in the inner membrane gradient fractions, in combination with the absence of PilP in OMV, strongly suggests that PilP is associated with the inner membrane. Immunoblot analysis of membrane fractions from the pilP frameshift mutant M1080pilPfs demonstrated that PilQ was still localized in the outer membrane (data not shown), indicating that PilP does not have an effect on PilQ membrane localization.

• Open in new tab
• Download powerpoint

FIG. 3.

PilP is not present in OMV. Immunoblotting of the whole-cell lysates (lane 1) and purified OMV (lane 2) from N. meningitidis MC58 detected with anti-PilP antiserum (K824), anti-PilQ antiserum (K010), and anti-OpcA antibody. α, anti.

• Open in new tab
• Download powerpoint

FIG. 4.

PilP copurifies with the inner membrane. The subcellular localization of N. meningitidis PilP was assessed by sucrose gradient centrifuge analysis of membrane fractions. Fractions enriched in the inner membranes were identified by monitoring LDH activity (graph, fractions 4 to 8). The presence of PilP protein was detected by immunoblotting using anti-PilP antiserum K824 (upper gel) and coincided with the LDH activity (fractions 4 to 8). Outer membrane fractions (10 to 13) enriched in the outer membrane proteins PilQ and OpcA (middle and lower gels) were detected by immunoblotting with anti-PilQ and anti-OpcA antibodies, respectively. OD/min, optical density per minute; •, LDH activity; ▪, sucrose density; α, anti.

The pilP and pilQ genes are cotranscribed.Since the genes encoding PilP and PilQ are located in the same locus, RT-PCR was employed to analyze the transcription patterns of the pilP and pilQ genes (Fig. 5). As expected, 1,410-bp and 810-bp products were obtained when primer pairs pilP-5075/pilQ-6485 and pilQ-5675/pilQ-6485 (see Table S1 in the supplemental material), respectively, were used, indicating that pilP and pilQ are cotranscribed.

• Open in new tab
• Download powerpoint

FIG. 5.

The genes encoding PilP and PilQ are cotranscribed. RT-PCR analysis of pilP and pilQ from N. meningitidis M1080. Ten microliters of each reaction mixture was subjected to electrophoresis on a 2% agarose gel. Lanes: 1, DNA size marker; 2 and 3, RT-PCRs on RNA treated with RNase for 5 and 20 min, respectively; 4, RT-PCR on RNA template using pilP- and pilQ-specific primers pilP-5075 and pilQ-6485; 5, PCR on RNA with pilQ-specific primers pilQ-5675 and pilQ-6485 using Taq polymerase; 6, RT-PCR on RNA without pilQ reverse primer; 7, RT-PCR on RNA using pilQ-specific primers pilQ-5675 and pilQ-6485. The products in lane 6, obscuring the interpretation of results, indicate a fold-back artifact of RNA, which can provide a 3′ terminus appropriate for reverse transcriptase to act on even in the absence of reverse primer. Molecular size markers are indicated on the left in kilobases.

PilP is not required for PilQ multimer stability.Previous reports have yielded contradictory results on PilQ stability in pilP-knockout mutants (8, 19). Drake et al. (19) found that an N. gonorrhoeae pilP null mutant expressed reduced amounts of PilQ complex. In contrast, Carbonnelle and coworkers (8) reported that in an N. meningitidis pilP null mutant, the amounts of PilQ monomer and complex formation were not affected. We therefore tested the PilQ multimer stability in our pilP mutants. Immunoblotting of whole cell lysates revealed that, in the M1080PilP_74-6xHis and the M1080pilPfs mutants, PilQ multimers were detected in the stacking gel as for the wild type, but PilQ complex was found in greatly reduced amounts in the M1080ΔpilP mutant (Fig. 6). The reduction of PilQ multimers in M1080ΔpilP was due to reduced PilQ monomer production, since the overall amount of this protein after phenol treatment of the extract also was reduced (Fig. 6B). Complementation of the pilPfs and ΔpilP mutants with PilP in trans did not exert any effect on PilQ expression or complex levels (Fig. 6A). The analysis of PilP expression in the pilQ mutant exhibited partial degradation of PilP (Fig. 6C). These experiments clearly show that PilP is not required for PilQ stability and therefore cannot be classified as a pilotin.

• Open in new tab
• Download powerpoint

FIG. 6.

PilP is not required for PilQ multimer stability. Meningococcal PilQ and PilP expression in whole-cell lysates monitored by immunoblotting. (A) Samples detected with anti-PilQ antiserum (K010); (B) samples detected with anti-PilQ antiserum after phenol extraction; (C) samples detected with anti-PilP antiserum K824. Lanes: 1, wild-type M1080; 2, M1080pilP _74-6xHis; 3, M1080ΔpilP; 4, M1080ΔpilP iga::pilP; 5, M1080pilPfs; 6, M1080pilPfs iga::pilP; 7, M1080ΔpilQ. Arrowheads indicate the positions of proteins. Equal amounts of starting materials were used to make cell lysates. Positions of the molecular mass standard proteins are indicated on the left in kDa.

Pilus expression and transformation are abolished in pilP mutants.Phenotypic analyses of the pilP mutants were carried out by testing piliation, agglutination, and competence for transformation with the M1080ΔpilQ mutant as a negative control (Table 2). The results showed wild-type levels of pili in M1080-PilP_74-6xHis, whereas the expression of pili was abolished in the M1080pilPfs and M1080ΔpilP mutants (Table 2). Complementation of pilP deletion and frameshift mutants restored the pilus expression to the wild-type level (Table 2). Compared to the competence for transformation in wild-type M1080 and M1080-PilP_74-6xHis, competence for transformation was abolished in the M1080ΔpilP and M1080pilPfs mutants (Table 2). These observations clearly showed that PilP is important for pilus biogenesis and competence for transformation.

The N-terminal and C-terminal segments of PilP interact with the central portion of PilQ.Solid-phase overlay assay was employed to assess the interaction between the PilP and PilQ proteins, including the full-length, N-terminal, central, and C-terminal portions of the proteins (Fig. 7A) (13, 22). The interactions of another neisserial outer membrane protein, hemoglobin binding receptor (HmbR), as well as bovine serum albumin, with PilP were also included as controls. Full-length PilP exhibited an interaction with full-length PilQ and the central portion of PilQ and a weak interaction with N-terminal PilQ, but no interaction with the C-terminal PilQ section (Fig. 7B, panel I). PilP proteins with the deletion of amino acids 1 to 12, 1 to 15, 40 to 60, 61 to 80, and polyhistidine-tagged protein (M1080-PilP_74-6xHis) exhibited a binding pattern similar to that of full-length PilP protein (Table 3). Deletion of residues 1 to 77 and only residues 1 to 16 abolished binding completely (Table 3). C-terminal PilP deletions which spanned residues 81 to 102, 107 to 116, and 114 to 181 of the protein abolished binding to all three truncated recombinant PilQ fragments, but not to the full-length PilQ. The results suggest that the N-terminal region of PilP, around the lipidated residue Cys₁₆, mediates substantial interaction with PilQ and that the region around the folded domain at the C terminus also contributes to PilQ recognition. For PilQ, the middle section of the polypeptide chain, between residues 213 and 349, seems to be the major site for PilP binding.

• Open in new tab
• Download powerpoint

FIG. 7.

Meningococcal PilP and PilQ monomers directly interact. Interaction of recombinant meningococcal PilP with recombinant PilQ protein detected by a solid-phase overlay assay (far-Western analysis). (A) Schematic diagram showing the pilQ gene constructs encoding the four recombinant proteins used in this study. H represents the position of the polyhistidine tag. (B) Far-Western analysis; the three panels, I to III, show gels with the same samples in each lane, but overlaid with different recombinant PilP proteins before immunodetection with anti-PilP antiserum K824. Panels: I, full-length recombinant PilP; II, recombinant PilP_Δ1-12; III, recombinant PilP_Δ61-80. Lanes: 1, full-length PilQ; 2, PilQ_25-354; 3, PilQ_217-478; 4, PilQ_350-761; 5, Hemoglobin-binding outer membrane receptor protein (HmbR); 6, bovine serum albumin. Positions of the molecular mass standard proteins are indicated on the left in kDa.

3D reconstruction of the PilP-PilQ oligomer complex.Having established the location of the PilP binding site within the primary structure of the PilQ oligomer, we extended the investigation to examine the interaction within the context of the 3D structure of the PilQ oligomer. PilP is a much smaller protein than PilQ, and it is therefore difficult to identify the density attributable to PilP from a low-resolution PilQ-PilP complex reconstruction determined by TEM. Consequently, two separate structures were determined. The first structural analysis involved the incubation of a recombinant version of PilP, with a polyhistidine tag at the N terminus (Fig. 1C), with the PilQ oligomer. The second involved the incubation of PilPΔ_1-19 with PilQ and a Ni-NTA-nanogold probe, which would specifically adhere to the polyhistidine tag within PilPΔ_1-19. The recombinant fragment PilPΔ_1-19 was selected for these studies because we have reported its structural characterization by NMR elsewhere (24).

A 3D reconstruction of the PilPΔ_1-19-PilQ complex was determined from approximately 8,000 particles, using a method similar to that described for the determination of the 3D structure of the PilQ oligomer alone (14) (Fig. 8A and B). The resulting structure was determined at a resolution of 16 Å: comparison with the original structure of the unliganded PilQ oligomer revealed changes within the “arm” and “cap” regions, as defined previously (11). As was the case for the structure of the PilQ oligomer alone, the PilPΔ_1-19-PilQ complex exhibited strong C4 symmetry (compare the unsymmetrized class averages in Fig. 8B with the back projections determined assuming C4 symmetry). For the second independent experiment, PilQ and PilPΔ_1-19 were incubated with Ni-NTA-nanogold: around one thousand particles were analyzed and processed to produce a 3D volume from which the gold particles could be readily identified based on their differential scattering characteristics. The gold particles, shown as yellow density peaks in Fig. 8C, were located around the cap feature of the complex, confirming that this is the location of the binding site for PilPΔ_1-19. The gold particles identify the location of the polyhistidine tag in PilPΔ_1-19: we have shown that the N terminus of this recombinant fragment of PilP is highly flexible in solution and does not adopt any regular secondary structure (24). Such an observation would explain why the locations of the gold labels in Fig. 8C appear to be adjacent to regions of density, rather than within them. Nevertheless, it is clear that the labeling is concentrated around the cap end of the PilQ complex.

• Open in new tab
• Download powerpoint

FIG. 8.

3D reconstruction of the complex between PilP_Δ1-19 and the PilQ oligomer. (A) Sample micrograph of PilQ particles, with PilP_Δ1-19 bound, in cryonegative stain (14). The positions of individual PilQ-PilP particles are circled. Scale bar = 1,000 Å. (B) PilQ-PilP particle classification. 2D class averages for each of the particle classes following reference-free alignment are shown with the corresponding 2D back projections used for the C4 3D reconstruction. Averages were calculated by using the software package EMAN. Box size = 250 Ų. (C) Comparison of 3D reconstructions of PilQ and PilQ-PilP oligomers. Isosurface renderings of volumes are displayed at thresholds appropriate to accommodate 900 to 1,000 kDa of mass. The top structure shows the structure of the PilQ oligomer alone (14). The middle structure shows the PilQ oligomer bound to PilP_Δ1-19, determined from 7,700 particles. The structure converged correctly using a symmetrical start model, a class-selected C1 start model, or an elliptical “blob”; model-dependent bias was therefore eliminated. The bottom structure shows the structure of the PilQ oligomer bound to PilP_Δ1-19 and Ni-NTA-nanogold (n = 1,010). The structure was calculated in the same way from a cryonegatively stained sample containing Ni-NTA-nanogold. Scale bar = 100 Å.