Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Collins, R. F. Articles by Derrick, J. P. Search for Related Content PubMed PubMed Citation Articles by Collins, R. F. Articles by Derrick, J. P.

 Previous Article  |  Next Article 

Journal of Bacteriology, September 2007, p. 6389-6396, Vol. 189, No. 17
0021-9193/07/$08.00+0     doi:10.1128/JB.00648-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Purification and Three-Dimensional Electron Microscopy Structure of the Neisseria meningitidis Type IV Pilus Biogenesis Protein PilG

Richard F. Collins, Muhammad Saleem, and Jeremy P. Derrick^*

Faculties of Life Sciences and Engineering/Physical Sciences, Manchester Interdisciplinary Biocentre, University of Manchester, 131 Princess Street, Manchester, United Kingdom

Received 25 April 2007/ Accepted 26 June 2007

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Type IV pili are surface-exposed retractable fibers which play a key role in the pathogenesis of Neisseria meningitidis and other gram-negative pathogens. PilG is an integral inner membrane protein and a component of the type IV pilus biogenesis system. It is related by sequence to the extensive GspF family of secretory proteins, which are involved in type II secretion processes. PilG was overexpressed and purified from Escherichia coli membranes by detergent extraction and metal ion affinity chromatography. Analysis of the purified protein by perfluoro-octanoic acid polyacrylamide gel electrophoresis showed that PilG formed dimers and tetramers. A three-dimensional (3-D) electron microscopy structure of the PilG multimer was determined using single-particle averaging applied to samples visualized by negative staining. Symmetry analysis of the unsymmetrized 3-D volume provided further evidence that the PilG multimer is a tetramer. The reconstruction also revealed an asymmetric bilobed structure approximately 125 Å in length and 80 Å in width. The larger lobe within the structure was identified as the N terminus by location of Ni-nitrilotriacetic acid nanogold particles to the N-terminal polyhistidine tag. We propose that the smaller lobe corresponds to the periplasmic domain of the protein, with the narrower "waist" region being the transmembrane section. This constitutes the first report of a 3-D structure of a member of the GspF family and suggests a physical basis for the role of the protein in linking cytoplasmic and periplasmic protein components of the type II secretion and type IV pilus biogenesis systems.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The gram-negative pathogen Neisseria meningitidis is the causative agent of meningococcal meningitis and septicemia and a serious public health problem. Type IV pili are long, thin, and mechanically robust fibers which extend from the surface of the bacterium and bind to host cell surface receptors (37). They have been shown to play a pivotal role in colonization and infection by N. meningitidis (27). Type IV pilus tip adhesins mediate binding to epithelial cell receptors (37) and so play a role in colonization (27). Type IV pili are also known to be involved in autoagglutination (38) and natural competence for DNA uptake (18). The ability of type IV pili to retract has been linked to twitching motility, a form of bacterial propulsion along solid surfaces (3). Despite the importance of type IV pilus biogenesis, the present understanding of the process at a molecular level is rudimentary. A survey has identified 15 different proteins involved in the biogenesis of type IV pili in N. meningitidis (5). Of these, only six (PilD, -F, -M, -N, -O, and -P) are essential for assembly of the major pilus protein, PilE, into pili (4). Other proteins appear to be involved in the functional maturation of the pilus (PilC, -I, -J, -K, and -W), its retraction (PilT), or its transit across the outer membrane (PilQ). Interestingly, although pilG mutants in pathogenic Neisseria are devoid of pili (40), Carbonnelle et al. found that a pilG pilT double mutant was piliated, with adhesive and aggregative properties which were close to those of the wild-type strain (4). These authors concluded that PilG plays a role in the counterretraction of the pilus fiber rather than a primary role in pilus assembly.

Several lines of evidence have pointed to similarities between type IV pilus biogenesis and the type II secretion system (23, 28). For example, the GspE family (GSP is for general type II secretory pathway) of hexameric ATPases, which are involved in type II secretion (23, 43), have orthologs which appear to carry out similar functions in type IV pilus biogenesis (4, 14, 20). The PulD and PilQ secretins, which mediate the transport of pullulanase and type IV pili, respectively, across the outer membrane, show some homology in sequence within their C-terminal regions and have broad structural similarities (6, 12). PilG is a member of the large GspF family of integral inner membrane proteins, which includes proteins involved in type IV pilus biogenesis and type II secretion. Evidence from studies on both systems suggests that GspF orthologs function as a complex with other proteins involved in the secretion process. For example, Py et al. showed that in the type II secretion system in Erwinia chrysanthemi, GspF formed part of a complex with GspE, -L, and -M (30). Crowther et al., working on the mechanism of bundle-forming pilus formation in enteropathogenic Escherichia coli, showed that the GspF ortholog BfpE is involved in recruiting the ATPase BfpD to the cytoplasmic membrane (13). They also provided evidence that BfpE interacted with a separate ATPase, BfpF, to promote pilus retraction. Further work has shown that peptides derived from another pilus biogenesis component protein, BfpC, and BfpE stimulated the ATPase activity of BfpD in an allosteric manner (14). These observations support the view that the BfpE/PilG pilus biogenesis components play a crucial role in the formation of a multiprotein complex which could traverse the inner membrane.

To date, there have been no reports of three-dimensional (3-D) structures of members of the GspF family. Blank and Donnenberg used a fusion protein approach to study the topology of BfpE and identified four transmembrane segments, three within the center of the gene and one at the C terminus (1). These observations led to a proposed topology for BfpE, with a large cytoplasmic N-terminal domain, three transmembrane -helices, a smaller periplasmic domain, and a fourth C-terminal transmembrane -helix. Single-particle averaging (SPA) is a method which can be applied to determine the 3-D structure of a large protein (>200 kDa) by a process of averaging projections of many thousands of individual protein particles (17). We have previously described the application of SPA methods to determine the 3-D structure of another membrane-associated component from the Neisseria type IV pilus biogenesis system, the outer membrane secretin PilQ (10, 12). The results established that PilQ is a dodecamer and adopts a cage-like structure with a large internal cavity which is filled when PilQ associates with type IV pili (9, 11). These observations illustrate the potential of SPA methods to provide structural information which can suggest clues to the function of the protein. Here we describe the 3-D structure of PilG from Neisseria meningitidis by SPA. We show that PilG forms a tetrameric bilobed structure, which agrees well with the Blank and Donnenberg topology model for the ortholog BfpE. The significance of the structure, in terms of its implications for the function of this class of pilus biogenesis proteins, is discussed.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Materials. BD Co-TALON resin was obtained from BD Biosciences. n-Dodecyl-D-maltoside (DDM) was from Anatrace Ltd., United Kingdom. Oligonucleotide primers were purchased from Sigma-Aldrich, and Pfu polymerase was purchased from Epicenter Biotechnologies. The E. coli strain DH5 was obtained from Invitrogen Ltd., United Kingdom. E. coli strains BL21(DE3)* and BL21(DE3)pLysS were from Novagen (Merck Biosciences Ltd., United Kingdom). Perfluoro-octanoic acid (PFO) was supplied by Fluorochem (Old Glossop, Derbyshire, United Kingdom). Ni-nitrilotriacetic acid (NTA) nanogold was from Nanoprobes (Yaphank, NY).

Cloning and overexpression of recombinant PilG. DNA sequence coding for the integral membrane protein PilG was amplified from genomic DNA isolated from Neisseria meningitidis MC58 (39) by use of high-fidelity Pfu polymerase and the following two primers: 5'-CGGGGCTAGCATATGGCTAAAAACG and 5'-GGATCTGTGCCTCGAGTCAGGCGACCACG. An NdeI site was introduced into the 5' end of the PCR product before the start codon, and an XhoI site was positioned after the stop codon at the 3' end. The PCR product was ligated directionally into the corresponding sites of pET-28a (Novagen) to yield the expression vector pNm-pilG, which introduces a hexahistidine tag at the N terminus of the protein. For expression, pNm-pilG was transformed into E. coli BL21(DE3)* (Invitrogen), and transformants were grown in 2YT medium (34) to an optical density at 600 nm of 0.8 to 1.0 at 37°C. At this point, expression was induced by the addition of isopropyl-ß-D-thiogalactoside (IPTG) to 0.1 mM, and the cells were transferred to a temperature of 25°C. Growth was allowed to continue for a further 16 h after induction at 25°C, before the cells were harvested by centrifugation at 6,000 x g_av for 20 min.

Purification of recombinant His-tagged PilG. The bacterial cell pellet was resuspended in lysis buffer (50 mM Tris [pH 8.0], 200 mM NaCl, 1 mM EDTA, and 0.5 mM dithiothreitol); 10 ml of lysis buffer was used per liter of culture. Lysozyme and DNase were also added to aid cell lysis. To minimize proteolysis, a "cocktail" of protease inhibitors (complete, EDTA free, catalog no. 11873580001; Roche) was added at the standard concentration recommended by the manufacturer. Cells were broken using a probe sonicator, and unbroken cells were removed by low-speed centrifugation at 10,000 x g_av for 20 min. The supernatant was filtered through a 0.45-µm syringe filter and retained. A membrane-rich preparation was subsequently isolated by centrifugation at 100,000 x g_av for 1 h at 4°C. The supernatant was discarded and the pellet washed with 50 mM Tris (pH 8.0) and 200 mM NaCl, to which protease inhibitors had been added. PilG was solubilized from this membrane-rich preparation by resuspension in 50 mM Tris (pH 8.0), 200 mM NaCl, 1% (wt/vol) DDM, also including proteinase inhibitors, at 20°C for 1 h. Insoluble material was subsequently removed by centrifugation at 100,000 x g_av for 30 min at 22°C. Co-TALON resin, which had been preequilibrated in 50 mM Tris (pH 8.0), 200 mM NaCl, 0.08% (wt/vol) DDM, was added to the supernatant, and the suspension was mixed gently for 16 h at 4°C. The Co-TALON resin was packed into a benchtop column and washed with 10 column volumes of Tris-NaCl-DDM buffer. PilG was eluted by application of 50 mM Tris-HCl, 200 mM NaCl, 0.08% (wt/vol) DDM, and 200 mM imidazole (pH 8.0). The imidazole was removed by dialysis against 50 mM Tris-HCl, 200 mM NaCl, 0.08% (wt/vol) DDM (pH 8.0), and the purified PilG was concentrated by ultrafiltration. The identity of the eluted protein was confirmed by matrix-assisted laser desorption ionization-time of flight mass spectrometry.

PFO-PAGE of oligomeric PilG. PFO polyacrylamide gel electrophoresis (PFO-PAGE) was performed using a variation on previously reported methods (31). Ten microliters of the sample, at a protein concentration of 600 µg/ml, was mixed 1:1 with PFO sample buffer (100 mM Tris [pH 7.5], 8% [wt/vol] Na-PFO, and 20% [vol/vol] glycerol) and incubated at 4°C for 10 min. Samples were then loaded onto 7.5% slab gels containing 375 mM Tris-HCl (pH 7.5) and run in a buffer precooled to 4°C and containing 25 mM Tris-HCl (pH 7.5), 200 mM glycerol, and 8% (wt/vol) Na-PFO for 2 h at 4°C.

CD. Circular dichroism (CD) spectra of PilG were obtained using a JASCO J-10 spectrometer and a quartz cell with a 1-mm path length. PilG was at a concentration of 0.130 mg/ml in 50 mM Tris-HCl (pH 7.5), 200 mM NaCl, and 0.08% (wt/vol) DDM. The spectrum was recorded at 20°C, and data from a blank, containing the same buffer but with no protein, were subtracted. Measurements were made only down to the wavelength where the instrument dynode voltage indicated the detector was still in its linear range.

Electron microscopy of PilG. Samples of PilG were diluted to 25 µg/ml in 50 mM Tris-HCl, 200 mM NaCl, and 0.08% (wt/vol) DDM and negatively stained with uranyl acetate as previously described (11). Transmission electron microscopy (TEM) images were recorded with a Tecnai 10 operating at 100 KeV onto Kodak SO-163 film. Data were scanned using a UMAX 2000-Powerlook scanner at 1,600 dots per inch, providing a specimen level increment of 3.66 Å/pixel.

Determination of PilG 3-D structure. A total of 9,440 particles were interactively selected using BOXER (25) and contrast normalized. The contrast transfer function (CTF) for each particle in the data set was determined using the program CTFIT (25), and corrections for amplitudes and phases were applied using a GroEL structure factor data set (provided by S. Ludtke). The data set was initially analyzed by calculating 55 reference-free projection averages that revealed the complex orientated in multiple particle positions. By following methods previously described (12), a preliminary 3-D model was determined from SPA classes that represented different views of the oligomer. In outline, this process involved the generation of reference-free class averages followed by the imposition of a low-pass resolution filter to 20 Å. Poorly sampled or noisy images were deleted manually. A variety of classes that were representative of different sample orientations, i.e., exhibiting rotational and bilateral symmetry as well as a variety of intermediate orientations, were subsequently selected. Classes which were apparently duplicated were deleted. From these classes, an unsymmetrized start model was generated and then refined, with the appropriate symmetry imposed. As part of the symmetry analysis process, a range of symmetries was applied in separate structural calculations: C1 (i.e., no rotational symmetry imposed), D2, C2, and C4 symmetries were imposed during independent computations. The relative orientations of the characteristic views were determined using a Fourier common-lines routine, and the resulting averages were combined to generate the preliminary 3-D model. The 3-D structure was subsequently refined using eight rounds of iterative projection matching. Each refinement was assessed by examining convergence with the Fourier shell correlation (FSC) of the 3-D models generated from each iteration. The resolution was determined to be 22 Å by FSC analysis as described previously (12). FSC effectively splits the particles into two independent datasets: calculation of a 3-D volume from each data set generated two structures which had essentially the same features as the final structure. Furthermore, the principal features of the final model were not dependent on the starting model: a repeat of the procedure using an amorphous "blob" as a start model yielded a similar reconstruction. The labeling of PilG with Ni-NTA nanogold particles was carried out as described for the secretin protein PilQ (19). A total of 1,100 gold-labeled particles were selected interactively, and a 3-D structure was calculated using the refined PilG structure as a start model for the gold-labeled data set.

Symmetry analysis. Density map correlation coefficients were calculated by rotation of the unsymmetrized (C1) 3-D map about the long axis of the PilG complex by use of MAPROT (36), and the density map correlation coefficient was calculated using OVERLAPMAP (2); both programs are from the CCP4 suite (8).

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Expression and purification of PilG. As part of a program to investigate the structures and functions of the proteins involved in the biogenesis of type IV pili in Neisseria, we sought to express and purify the integral membrane protein PilG in quantities suitable for structural analysis. An expression vector for PilG was constructed by insertion of an NdeI-XhoI fragment incorporating the entire pilG open reading frame into pET-28a to create pNm-pilG. Following the transformation of pNm-pilG into E. coli BL21(DE3)*, expression was monitored by Western blotting under a variety of different growth conditions. Optimal expression of PilG was obtained by transferring the cells to grow at 25°C after induction with IPTG (data not shown). Following cell breakage by sonication, total membranes were isolated by ultracentrifugation before solubilization with the detergent DDM. PilG was then purified by metal affinity chromatography using a Co-TALON affinity resin. Examination of the purified protein by sodium dodecyl sulfate (SDS)-PAGE gave a band with an estimated subunit molecular mass of about 43 kDa, close to the predicted value of 48.2 kDa expected from the sequence (Fig. 1A). The procedure produced a yield of about 1 mg purified protein from 6 liters of bacterial culture.

View larger version (72K): [in this window] [in a new window]

FIG. 1. SDS-PAGE and PFO-PAGE of purified PilG. (A) SDS-PAGE of purified PilG on a 4 to 20% continuous-gradient gel. Molecular mass markers are indicated on the right of the gel. (B) PFO-PAGE of PilG on a 4 to 12% gradient gel. Lanes: 1, ferritin (440 kDa); 2, immunoglobulin G (150 kDa); 3, bovine serum albumin (66 kDa); 4, PilG. Both gels were stained with Coomassie blue.

Biophysical properties of purified PilG. Previous work using a yeast two-hybrid system had suggested that the PilG ortholog BfpE is multimeric, but the precise quaternary structure of any member of the GspF family has yet to be established (13). PFO-PAGE is a technique that allows the quaternary structure of an oligomeric complex to be quantitatively assessed because the detergent PFO maintains subunit interactions in many protein oligomers (31). Separation of purified PilG by PFO-PAGE identified two main bands which migrated at estimated molecular masses of 40 kDa and 80 kDa (Fig. 1B) and were identified as monomeric and dimeric species, respectively. A higher-molecular-mass band, estimated at 180 kDa, was also detectable and could correspond to a trimer or tetramer. The quaternary structure of PilG was also examined by ultracentrifugation: sedimentation equilibrium analysis produced an overall molecular mass estimate of 124 ± 4 kDa. Sedimentation velocity analysis showed the presence of three species: around two-thirds of the sample was dimeric, with the remaining fraction consisting mostly of tetramers and, to a lesser extent, higher-order species (data not shown). The ultracentrifugation was carried out using PilG solubilized in DDM: this detergent appeared to preserve the quaternary structure of the protein better than PFO. Nevertheless, the results provided clear evidence for the formation of dimers and tetramers by PilG.

Secondary structure bioinformatic predictions (33) indicated a high content of -helix in PilG, at 73%. In order to establish that the purified PilG sample was correctly folded, a CD spectrum was obtained (Fig. 2). The results show a double minimum in the CD signal at 208 and 220 nm, characteristic of a protein with high -helical content. Noise in the signal in the 190- to 200-nm region complicated an estimate of the proportion of -helix in the protein.

View larger version (14K): [in this window] [in a new window]

FIG. 2. CD spectrum of purified PilG. degM, •••.

TEM of purified PilG. Figure 3A shows a sample area of a TEM of purified PilG visualized by negative staining. The population of PilG oligomers was homogeneous, generally nonaggregated, and uniform in size (80 to 130 Å). Determination of a 3-D structure by SPA using the EMAN software (25) requires that a preliminary 3-D model be generated and then subjected to an iterative refinement procedure to generate the final structure. To generate this preliminary model, a set of reference-free class averages needs to be generated from the particle data set: a representative selection of those for PilG is shown in Fig. 3B. Each class average represents an averaged projection of a particular orientation of the PilG particle. The results show a variety of different projection types, demonstrating that the PilG oligomer was positioned on the carbon support film in different orientations. Although many projection classes were apparent, two types of high-contrast projections could be broadly distinguished from the majority of the data. One type of projection (Fig. 3B, top left) was 80 to 90 Å in width, had a squared four-lobed appearance, and was well stained in the center. The other type of high-contrast projection (Fig. 3B, bottom right) was more elongated. The other class averages represented various intermediate orientations between these two projections (Fig. 3B, middle rows) and so were suitable for calculating a 3-D structure using Euler angle determination and cross-common lines projection matching (25). The readily apparent visual correlation between each class average (Fig. 3B, right column) and its equivalent back-projection from the refined 3-D structure (Fig. 3B, left column) demonstrates an excellent correspondence between the data and the final structure.

View larger version (116K): [in this window] [in a new window]

FIG. 3. Analysis of PilG by TEM. (A) Sample TEM data of PilG negatively stained with 2% (wt/vol) uranyl acetate. PilG particles are highlighted in square boxes. The image has been CTF corrected and contrast enhanced for presentation. Scale bar = 500 Å. (B) Selected class averages generated from SPA as applied to the PilG data set. The 14 average pairs shown are representative samples from the total classes and show a back-projection from the C4-symmetrized 3-D volume paired with the corresponding unsymmetrized class average used in the calculation. Excellent visual correlation between the pairs is apparent. Box size is 234 Å by 234 Å.

3-D structure of solubilized PilG. A total of 9,440 unique PilG particles were processed to produce a 3-D reconstruction at an estimated resolution of 22 Å. Structure calculation carried out with no symmetry imposed (i.e., C1), generated a volume which was analyzed by map density correlation analysis (Fig. 4). The results showed a strong signal for twofold symmetry, with slightly weaker peaks at 90° and 270°. A top view of the unsymmetrized volume showed an apparent fourfold symmetrical arrangement (Fig. 4, inset). Given the results of the biochemical analysis of the quaternary structure detailed above, the most plausible symmetry arrangement would be D2, C2, or C4. Repeating the structure calculation, but with D2 symmetry imposed, produced a rectangle-shaped volume, which was inconsistent with the unsymmetrized class averages shown in Fig. 3 (data not shown). The possibility of D2 symmetry was therefore discounted. 3-D structural processing was repeated, imposing C2 or C4 rotational symmetry: the resulting reconstructions (Fig. 5A) were essentially similar in all major features. Volumes from both structures, when contoured at 2.6 above the mean density, would accommodate an estimated molecular mass of about 240 kDa. This is consistent with a PilG tetramer, assuming a contribution of 192 kDa from the protein and 48 kDa from the bound DDM micelle. It is not possible, however, to distinguish between a dimer-of-dimers quaternary arrangement (i.e., C2) and a tetramer related by fourfold rotational symmetry (C4) on the basis of these data. As the C2 and C4 models show essentially the same features, further analysis will use the C4 structure, although the conclusions drawn also pertain to the C2 reconstruction. It should be noted that the sample would undoubtedly have contained PilG dimers, but their size (120 kDa) is too small to be visualized as a particle well stained for SPA. Consequently, the particles selected and analyzed by SPA corresponded only to the tetrameric species, and the dimeric species were effectively omitted.

View larger version (22K): [in this window] [in a new window]

FIG. 4. Rotational symmetry analysis of SPA PilG volume. Real-space self-correlation analysis of the unsymmetrized PilG volume is shown as a function of the angle of rotation about the main axis of the complex. Correlation coefficients were calculated from a 3-D volume produced with no symmetry averaging applied (C1). The volume was calculated using procedures described previously (12). The inset shows a top-view back-projection of the unsymmetrized volume.

View larger version (42K): [in this window] [in a new window]

FIG. 5. 3-D reconstruction of PilG. (A) Surface-rendered symmetrized volumes of the PilG complex: the structure in the left column has C2 symmetry imposed, and the structure on the right has C4 symmetry. Both volumes are displayed at a threshold of 2.6 above the mean density. The volume was low-pass filtered to a 22-Å resolution. The main features and dimensions of the complex are indicated. (B) Slabs taken through the C4-symmetrized 3-D volume from panel A, highlighting structural features at 3 (red), 3.5 (orange), 4 (yellow), 4.5 (pale blue), and 5 (dark blue) above the mean density.

The PilG tetramer is 125 Å in length and has a bilobed, asymmetric, "missile-like" appearance when viewed from the side. A narrowed waist divides the molecule into two unequal halves, which can be characterized as "fin-like" and "cone-like" features (Fig. 5A). Viewed from the top, the particle has a distinctive square appearance, maximally 80 Å wide (Fig. 5). In order to identify which end of the molecule corresponded to the N-terminal domain, we employed selective labeling with Ni-NTA nanogold, which binds to the N-terminal polyhistidine tag. In a separate reconstruction, a data set of 1,100 particles of the Ni-NTA nanogold-PilG complex was collected: some examples of the images used are shown in Fig. 6A. Using the unliganded PilG structure as a starting model, a volume for the Ni-NTA nanogold-PilG complex was determined (Fig. 6B). A major strength of the nanogold labeling methodology is that the nanogold particles have scattering characteristics different from those of the PilG volume in negative staining: the positive density in Fig. 6B, in green, is attributable to PilG, and the negative density, in yellow, corresponds to the nanogold particles. The location of the nanogold label can therefore be readily distinguished from background noise. The PilG portion of the structure has the characteristics of the PilG reconstruction alone but is inevitably noisier, due to the lower number of particles used in the reconstruction. The main features of the PilG structure—the "fins" at the top, the narrower "waist" region, and the "cone" at the bottom of the structure—are readily apparent. It is clear that the majority of the nanogold labels associate with the bottom half of the structure. It should be noted that the yellow nanogold density in Fig. 6B and C is somewhat diffuse, possibly because the N-terminal polyhistidine tag is surface exposed and consequently mobile. Nevertheless, the results provide clear evidence that the N terminus is associated with the "cone" portion of the PilG structure. A composite image of the (unliganded) PilG structure in green with the nanogold density peaks superimposed in yellow is shown on the left side of Fig. 6C.

View larger version (66K): [in this window] [in a new window]

FIG. 6. Nanogold labeling of PilG and proposed orientation in the inner membrane. (A) Montage of raw multiple gold-labeled PilG complexes. Data were CTF corrected for presentation. Box size is 234 Å by 234 Å. (B) 3-D reconstruction of the PilG-nanogold complex, filtered to a 30-Å resolution. The PilG volume is shown by green netting contoured at 3 above the mean density. Peaks associated with nanogold particles are shown in yellow at –4.1 below the mean density. C4 symmetry has been imposed. The volume was calculated using the PilG reconstruction determined previously (Fig. 5) as a start model for the nanogold-labeled data set (n = 1,100) and processed as described previously (12). The volume is displayed following low-pass filtering to 25 Å. Scale bar = 50 Å. (C) The left panel shows a composite figure, with the volume from the reconstruction of PilG alone displayed at 2.6 (light green) and 5 (dark green) above the mean density in green surface rendering with 50% opacity. Superimposed on this density are the peaks associated with the bound nanogold particles, shown at –5 below the mean density in yellow surface rendering. Both volumes had C4 symmetry imposed. The cartoon on the right shows the PilG topology, with transmembrane helices predicted using PHDhtm (32), matched approximately to the distribution of mass in the PilG 3-D structure on the left.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The members of the GspF family are not well understood at the structural level, despite their established importance in type II secretion and type IV pilus biogenesis. The data presented here have provided the first indication of the tertiary structure of a member of the GspF family. In order to correlate the 3-D reconstruction of PilG with its predicted topology in the membrane, we analyzed the PilG sequence for transmembrane helices with a variety of prediction methods. Most methods—HMMTOP (41), DAS (15), PHDhtm (32), and TMPred (22)—predicted four helices. However, TMHMM (35) and SOSUI (21) found only three, missing the third transmembrane helix, which was detected only weakly by most methods, probably due to the relatively high number of charged residues within this portion of the sequence. The prediction of the number and location of transmembrane helices in membrane proteins is not completely reliable, although methods based purely on hydrophobicity scales are generally less accurate (7). In view of the preponderance of predictions in favor of four helices in PilG, rather than three, and considering the similarities between the roles of PilG and BfpE in pilus biogenesis, we propose that PilG is most likely to adopt a four-helix topology. This topological model would effectively divide PilG into two unequal halves: an N-terminal cytoplasmic domain of around 170 residues and a C-terminal periplasmic domain of 100 residues, separated by three transmembrane helices (Fig. 6C, right). Such an arrangement would have an obvious correlation with the 3-D reconstruction, with the N-terminal domain aligning with the bottom part of the structure, the fin-shaped top of the structure corresponding to the C-terminal domain, and the narrow "waist" region comprising the central three transmembrane helices (Fig. 6C). This conclusion would suggest that PilG not only spans the bacterial inner membrane but has substantial segments of polypeptide chain exposed at both the cytoplasmic and periplasmic sides. We would expect this conclusion to extend to other orthologs of PilG including, for example, PulF from Klebsiella pneumoniae (29) and OutF from Erwinia chrysanthemi (24).

The overall size of the PilG particles is at the limits of the potential of single-particle 3-D reconstructions (17). Nevertheless, the resulting 3-D reconstruction provided clear evidence for a tetrameric quaternary arrangement. This observation agrees well with PFO-PAGE and ultracentrifugation analysis, which established that although the dimeric form was the main species present in solution, PilG was capable of forming tetramers. The quaternary state of PilG when reconstituted into the membrane is harder to extrapolate from these data. It should be noted that the oligomeric state of functionally active membrane protein complexes can vary depending on the detergent used in the purification. For example, photosystem I may be purified as a stable monomer or trimer, but the relative ratio of monomeric to trimeric species varies depending on the detergent of choice (16).

Carbonnelle et al. have suggested that PilG plays a role in the retraction/counterretraction of the type IV pilus fiber rather than in the assembly of the PilE pilin protein into pili (4). The ATPase PilT is known to be specifically involved in retraction of the pilus fiber: pilT mutants can produce pili but not retract them (26, 42). A pilG pilT double mutant produces pilus fibers that are identical to those from wild-type organisms, and the pilus-mediated adhesion of the double mutant to human cells is only marginally impaired (4). The precise function of PilG is currently unknown, although the structure presented here suggests that it could potentially play a role in binding to other pilus biogenesis proteins in the cytoplasm and periplasm and perhaps even provide a link between these two compartments. We estimate that the cytoplasmic and periplasmic domains protrude about 50 and 30 Å, respectively, from the inner membrane and would present plausible surfaces for interaction with other biogenesis component proteins. The association of the four periplasmic "fins" forms a shallow, cone-shaped invagination at the top of the PilG complex. This is around 20 Å deep in the center (Fig. 5B) and could potentially provide a binding surface for other protein partners. An intriguing aspect of type IV pilus retraction is the way in which ATP hydrolysis in the cytoplasm is apparently coupled to the disassembly of the pilus fiber in the periplasm. The 3-D structure of PilG suggests a physical explanation for how this protein could play a role in physically linking these two processes.

 
We thank Andy Baron at the University of Leeds for providing ultracentrifugation and sedimentation data.

Funding associated with ultracentrifugation measurements was provided by the Wellcome Trust under the Joint Infrastructure Fund (JIF) initiative.

 
* Corresponding author. Mailing address: Faculties of Life Sciences and Engineering/Physical Sciences, Manchester Interdisciplinary Biocentre, University of Manchester, 131 Princess Street, Manchester, United Kingdom. Phone: 001-44-0612004207. Fax: 001-44-0612360409. E-mail: Jeremy.Derrick{at}manchester.ac.uk

Published ahead of print on 6 July 2007.

Both authors contributed equally to the work.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Blank, T. E., and M. S. Donnenberg. 2001. Novel topology of BfpE, a cytoplasmic membrane protein required for type IV fimbrial biogenesis in enteropathogenic Escherichia coli. J. Bacteriol. 183:4435-4450.[Abstract/Free Full Text]
2
2. Branden, C. I., and T. A. Jones. 1990. Between objectivity and subjectivity. Nature 343:687-689.[CrossRef]
3
3. Burrows, L. L. 2005. Weapons of mass retraction. Mol. Microbiol. 57:878-888.[CrossRef][Medline]
4
4. Carbonnelle, E., S. Helaine, X. Nassif, and V. Pelicic. 2006. A systematic genetic analysis in Neisseria meningitidis defines the Pil proteins required for assembly, functionality, stabilization and export of type IV pili. Mol. Microbiol. 61:1510-1522.[CrossRef][Medline]
5
5. Carbonnelle, E., S. Helaine, L. Prouvensier, X. Nassif, and V. Pelicic. 2005. Type IV pilus biogenesis in Neisseria meningitidis: PilW is involved in a step occurring after pilus assembly, essential for fibre stability and function. Mol. Microbiol. 55:54-64.[CrossRef][Medline]
6
6. Chami, M., I. Guilvout, M. Gregorini, H. W. Remigy, S. A. Muller, M. Valerio, A. Engel, A. P. Pugsley, and N. Bayan. 2005. Structural insights into the secretin PulD and its trypsin-resistant core. J. Biol. Chem. 280:37732-37741.[Abstract/Free Full Text]
7
7. Chen, C. P., A. Kernytsky, and B. Rost. 2002. Transmembrane helix predictions revisited. Protein Sci. 11:2774-2791.[CrossRef][Medline]
8
8. Collaborative Computational Project, Number 4. 1994. The CCP4 suite: programs for protein crystallography. Acta Crystallogr. D 50:760-763.[CrossRef][Medline]
9
9. Collins, R. F., and J. P. Derrick. 2007. Wza: a new structural paradigm for outer membrane secretory proteins? Trends Microbiol. 15:96-100.[CrossRef][Medline]
10
10. Collins, R. F., R. C. Ford, A. Kitmitto, R. Olsen, T. Tønjum, and J. P. Derrick. 2003. Three-dimensional structure of the Neisseria meningitidis secretin PilQ determined from negative-stain transmission electron microscopy. J. Bacteriol. 185:2611-2617.[Abstract/Free Full Text]
11
11. Collins, R. F., S. Frye, S. Balasingham, R. C. Ford, T. Tønjum, and J. P. Derrick. 2005. Interaction with Type IV pili induces structural changes in the bacterial outer membrane secretin PilQ. J. Biol. Chem. 280:18923-18930.[Abstract/Free Full Text]
12
12. Collins, R. F., S. A. Frye, A. Kitmitto, R. C. Ford, T. Tønjum, and J. P. Derrick. 2004. Structure of the Neisseria meningitidis outer membrane PilQ secretin complex at 12 Å resolution. J. Biol. Chem. 279:39750-39756.[Abstract/Free Full Text]
13
13. Crowther, L. J., R. P. Anantha, and M. S. Donnenberg. 2004. The inner membrane subassembly of the enteropathogenic Escherichia coli bundle-forming pilus machine. Mol. Microbiol. 52:67-79.[CrossRef][Medline]
14
14. Crowther, L. J., A. Yamagata, L. Craig, J. A. Tainer, and M. S. Donnenberg. 2005. The ATPase activity of BfpD is greatly enhanced by zinc and allosteric interactions with other Bfp proteins. J. Biol. Chem. 280:24839-24848.[Abstract/Free Full Text]
15
15. Cserzo, M., E. Wallin, I. Simon, G. von Heijne, and A. Elofsson. 1997. Prediction of transmembrane alpha-helices in prokaryotic membrane proteins: the dense alignment surface method. Protein Eng. 10:673-676.[Abstract/Free Full Text]
16
16. Ford, R. C., A. Hefti, and A. Engel. 1990. Ordered arrays of the photosystem-I reaction center after reconstitution—projections and surface reliefs of the complex at 2nm resolution. EMBO J. 9:3067-3075.[Medline]
17
17. Frank, J. 2002. Single-particle imaging of macromolecules by cryo-electron microscopy. Annu. Rev. Biophys. Biomol. Struct. 31:303-319.[CrossRef][Medline]
18
18. Froholm, L., K. Jyssum, and K. Bovre. 1973. Electron microscopical and cultural features of Neisseria meningitidis competence variants. Acta Pathol. Microbiol. Scand. Sect. B 81:525-537.
19
19. Frye, S. A., R. Assalkhou, R. F. Collins, R. C. Ford, C. Petersson, J. P. Derrick, and T. Tønjum. Topology of the outer membrane secretin PilQ from Neisseria meningitidis. Microbiology, in press.
20
20. Herdendorf, T. J., D. R. McCaslin, and K. T. Forest. 2002. Aquifex aeolicus PilT, homologue of a surface motility protein, is a thermostable oligomeric NTPase. J. Bacteriol. 184:6465-6471.[Abstract/Free Full Text]
21
21. Hirokawa, T., S. Boon-Chieng, and S. Mitaku. 1998. SOSUI: classification and secondary structure prediction system for membrane proteins. Bioinformatics 14:378-379.[Abstract/Free Full Text]
22
22. Hofmann, K., and W. Stoffel. 1993. TMbase—a database of membrane spanning proteins segments. Biol. Chem. Hoppe-Seyler 374:166.
23
23. Johnson, T. L., J. Abendroth, W. G. J. Hol, and M. Sandkvist. 2006. Type II secretion: from structure to function. FEMS Microbiol. Lett. 255:175-186.[CrossRef][Medline]
24
24. Lindeberg, M., and A. Collmer. 1992. Analysis of eight out genes in a cluster required for pectic enzyme secretion by Erwinia chrysanthemi: sequence comparison with secretion genes from other gram-negative bacteria. J. Bacteriol. 174:7385-7397.[Abstract/Free Full Text]
25
25. Ludtke, S. J., P. R. Baldwin, and W. Chiu. 1999. EMAN: semiautomated software for high-resolution single-particle reconstructions. J. Struct. Biol. 128:82-97.[CrossRef][Medline]
26
26. Maier, B., L. Potter, M. So, H. S. Seifert, and M. P. Sheetz. 2002. Single pilus motor forces exceed 100 pN. Proc. Natl. Acad. Sci. USA 99:16012-16017.[Abstract/Free Full Text]
27
27. Nassif, X., C. Pujol, P. Morand, and E. Eugene. 1999. Interactions of pathogenic Neisseria with host cells. Is it possible to assemble the puzzle? Mol. Microbiol. 32:1124-1132.[CrossRef][Medline]
28
28. Nunn, D. 1999. Bacterial type II protein export and pilus biogenesis: more than just homologies? Trends Cell Biol. 9:402-408.[CrossRef][Medline]
29
29. Possot, O., C. Denfert, I. Reyss, and A. P. Pugsley. 1992. Pullulanase secretion in Escherichia coli K-12 requires a cytoplasmic protein and a putative polytopic cytoplasmic membrane-protein. Mol. Microbiol. 6:95-105.[CrossRef][Medline]
30
30. Py, B., L. Loiseau, and F. Barras. 2001. An inner membrane platform in the type II secretion machinery of gram-negative bacteria. EMBO Rep. 2:244-248.[CrossRef][Medline]
31
31. Ramjeesingh, M., L. J. Huan, E. Garami, and C. E. Bear. 1999. Novel method for evaluation of the oligomeric structure of membrane proteins. Biochem. J. 342:119-123.[CrossRef][Medline]
32
32. Rost, B., P. Fariselli, and R. Casadio. 1996. Topology prediction for helical transmembrane proteins at 86% accuracy. Protein Sci. 5:1704-1718.[Medline]
33
33. Rost, B., and C. Sander. 1994. Combining evolutionary information and neural networks to predict protein secondary structure. Proteins 19:55-72.[CrossRef][Medline]
34
34. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
35
35. Sonnhammer, E., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences, p. 175-182. In J. Glasgow, T. Littlejohn, F. Major, and R. Lathrop (ed.), Proceedings: Sixth International Conference on Intelligent Systems for Molecular Biology (ISMB98), Montreal, Canada.
36
36. Stein, P. E., A. Boodhoo, G. D. Armstrong, S. A. Cockle, M. H. Klein, and R. J. Read. 1994. The crystal structure of pertussis toxin. Structure 2:45-57.[Medline]
37
37. Strom, M. S., and S. Lory. 1993. Structure, function and biogenesis of the type IV pili. Annu. Rev. Microbiol. 47:565-596.[CrossRef][Medline]
38
38. Swanson, J., S. J. Kraus, and E. C. Gotschlich. 1971. Studies on gonococcus infection. I. Pili and zones of adhesion: their relation to gonococcal growth patterns. J. Exp. Med. 134:886-906.[Abstract]
39
39. Tettelin, H., N. J. Saunders, J. Heidelberg, A. C. Jeffries, K. E. Nelson, J. A. Eisen, K. A. Ketchum, D. W. Hood, J. F. Peden, R. J. Dodson, W. C. Nelson, M. L. Gwinn, R. DeBoy, J. D. Peterson, E. K. Hickey, D. H. Haft, S. L. Salzberg, O. White, R. D. Fleischmann, B. A. Dougherty, T. Mason, A. Ciecko, D. S. Parksey, E. Blair, H. Cittone, E. B. Clark, M. D. Cotton, T. R. Utterback, H. Khouri, H. Y. Qin, J. Vamathevan, J. Gill, V. Scarlato, V. Masignani, M. Pizza, G. Grandi, L. Sun, H. O. Smith, C. M. Fraser, E. R. Moxon, R. Rappuoli, and J. C. Venter. 2000. Complete genome sequence of Neisseria meningitidis serogroup B strain MC58. Science 287:1809-1815.[Abstract/Free Full Text]
40
40. Tønjum, T., N. E. Freitag, E. Namork, and M. Koomey. 1995. Identification and characterization of PilG, a highly conserved pilus-assembly gene in pathogenic Neisseria. Mol. Microbiol. 16:451-464.[Medline]
41
41. Tusnady, G., and I. Simon. 1998. Principles governing amino acid composition of integral membrane proteins: application to topology prediction. J. Mol. Biol. 283:489-506.[CrossRef][Medline]
42
42. Wolfgang, M., J. P. M. van Putten, S. F. Hayes, D. Dorward, and M. Koomey. 2000. Components and dynamics of fiber formation define a ubiquitous biogenesis pathway for bacterial pili. EMBO J. 19:6408-6418.[CrossRef][Medline]
43
43. Yamagata, A., and J. A. Tainer. 2007. Hexameric structures of the archaeal secretion ATPase GspE and implications for a universal secretion mechanism. EMBO J. 26:878-890.[CrossRef][Medline]

---

Journal of Bacteriology, September 2007, p. 6389-6396, Vol. 189, No. 17
0021-9193/07/$08.00+0     doi:10.1128/JB.00648-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Collins, R. F. Articles by Derrick, J. P. Search for Related Content PubMed PubMed Citation Articles by Collins, R. F. Articles by Derrick, J. P.