Type III Pilus of Corynebacteria: Pilus Length Is Determined by the Level of Its Major Pilin Subunit

Abstract

Multiple pilus gene clusters have been identified in several gram-positive bacterial genomes sequenced to date, including the Actinomycetales, clostridia, streptococci, and corynebacteria. The genome of Corynebacterium diphtheriae contains three pilus gene clusters, two of which have been previously characterized. Here, we report the characterization of the third pilus encoded by the spaHIG cluster. By using electron microscopy and biochemical analysis, we demonstrate that SpaH forms the pilus shaft, while SpaI decorates the structure and SpaG is largely located at the pilus tip. The assembly of the SpaHIG pilus requires a specific sortase located within the spaHIG pilus gene cluster. Deletion of genes specific for the synthesis and polymerization of the other two pilus types does not affect the SpaHIG pilus. Moreover, SpaH but not SpaI or SpaG is essential for the formation of the filament. When expressed under the control of an inducible promoter, the amount of the SpaH pilin regulates pilus length; no pili are assembled from an SpaH precursor that has an alanine in place of the conserved lysine of the SpaH pilin motif. Thus, the spaHIG pilus gene cluster encodes a pilus structure that is independently assembled and antigenically distinct from other pili of C. diphtheriae. We incorporate these findings in a model of sortase-mediated pilus assembly that may be applicable to many gram-positive pathogens.

Over 35 years ago, using electron microscopy of corynebacterial species, Yanagawa and colleagues observed proteinaceous filaments, namely, pili or fimbriae, on the surface of many corynebacterial species (30, 31). Since then, pili have been identified in several gram-positive bacteria (2, 4, 10-13, 19, 21, 29). However, the mechanisms of pilus assembly have remained obscure. It has been proposed that gram-positive bacterial pili are covalently linked to the cell wall peptidoglycan and require sortase (14, 20, 25, 32), a transpeptidase that is found in all gram-positive organisms (5, 17, 23). The available genome sequences have revealed multiple pilus gene clusters with a putative sortase(s) in each of several gram-positive bacterial species, including the Actinomycetales, clostridia, corynebacteria, and streptococci (9, 26). In Corynebacterium diphtheriae, with genes encoding pilin subunits and putative sortases organized into three clusters, it has been shown that the organism assembles two different pilus structures encoded by spaABC and spaDEF gene clusters, designated SpaA-type and SpaD-type pili, respectively (9, 26). It is likely that the third cluster, spaHIG, also encodes a distinct pilus organelle.

Previous work has shown that the SpaA-type pilus of corynebacteria is composed of three pilin subunits with SpaA polymers forming the pilus shaft, SpaB observed at regular intervals, and SpaC positioned at the tip (26). Similarly, the SpaD-type pilus is composed of SpaD, SpaE, and SpaF, which are counterparts of SpaA, SpaB, and SpaC, respectively (9). Although C. diphtheriae has six sortase genes, srtA, srtB, srtC, srtD, srtE, and srtF (26), the assembly of each pilus type requires a specific one (9, 26). SrtA, encoded by srtA, which is the only sortase found in the spaABC pilus gene cluster, is required for the assembly of SpaABC pili (24, 26). Likewise, sortases SrtB and SrtC are essential for the formation of SpaDEF pili; SrtB is specifically required for the incorporation of SpaE into SpaDF pili, whose assembly requires either SrtB or SrtC, while other remaining sortases are dispensable (9). In addition to sortase as a central catalyst for pilus assembly, a conserved pilin motif is necessary for the polymerization of pilin subunits into a high-molecular-weight (HMW) structure (9, 26). It is believed that sortase cleaves the sorting signal of the pilin precursors to generate covalent linkages between pilin subunits at the cleaved polypeptides and the side chain amino groups of pilin motif sequences (25) (see Fig. 7). The precise mechanism of this reaction still remains to be determined. Both sortase and pilin motifs are features that are conserved in the genome of many gram-positive pathogens (25), suggesting the existence of a common pathway in the assembly of gram-positive bacterial pili.

FIG. 7.

Working model of pilus assembly. SpaH pilins are synthesized in the cytoplasm as precursors with the C-terminal sorting signal consisting of the conserved LPXTG motif (white box), a hydrophobic domain (gray box), and a positively charged tail (black box). After being translocated by the Sec machinery and inserting into the membrane, the SpaH precursor forms an acyl-enzyme intermediate with sortase. Pilus polymerization involves a nucleophilic attack between two acyl-enzyme intermediates via the pilin motif (oval) and the LPXTG motif (product 1 of reaction I). Pili grow from the base as more pilins are polymerized (reaction II). Polymerization is terminated when the polymer is incorporated into the cell wall via a lipid II precursor by a transglycosylation reaction. The order of assembly is numbered, and different forms of sortases are indicated (see the text for details).

In the aforementioned pilus types of C. diphtheriae, SpaABC and SpaDEF pili, a major subunit, SpaA or SpaD, is essential for pilus assembly, while the minor components SpaB and SpaC or SpaE and SpaF, respectively, are not required (9, 26). Complementation of the ΔspaA mutant with wild-type spaA on a multicopy plasmid not only restores the SpaA-positive pilus but also leads to the formation of extended fibers (26). A similar phenotype is also observed when the SpaD pilin is overproduced (9). Thus, it seems likely that the amount of the major pilins regulates pilus length, a phenomenon that has previously been observed for the assembly of gram-negative bacterial pili (7, 27).

In this report, we have characterized the third pilus gene cluster of C. diphtheria, containing spaG-spaH-srtD-srtE-spaI. Using specific antisera against the pilus subunits, we show that spaH encodes a major pilin protein, SpaH, forming the pilus shaft, while spaG and spaI encode two minor proteins, SpaG and SpaI, respectively, which are dispensable for the formation of the pilus structure. We also show that the SpaHIG pilus is antigenically distinct from the other two corynebacterial pili and that the assembly of SpaHIG requires specific sortases located within the pilus gene cluster. Finally, using an inducible system, we demonstrate that the length of the filament is driven by the abundance of the major pilin SpaH.

MATERIALS AND METHODS

Bacterial strains, plasmids, and media.

C. diphtheriae strains (Table 1) were grown on heart infusion broth, heart infusion agar (HIA), or trypticase soy agar supplemented with 5% sheep blood (TSASB). Escherichia coli strains were grown in Luria broth (26). Kanamycin was added at 50 μg/ml for C. diphtheriae and E. coli strains as needed. Reagents were purchased from Sigma unless otherwise indicated.

TABLE 1.

Corynebacterium diphtheriae strains used in this study

Strain	Genotype	Reference or source
NCTC13129	Wild type	26
HT1	ΔsrtF	26
HT2	ΔsrtA	26
HT4	ΔsrtD	26
HT6	ΔsrtB	26
HT8	ΔsrtC	26
HT11	ΔspaA	26
HT12	ΔsrtE	26
HT14	ΔspaD	9
HT19	ΔsrtDE	This study
HT21	ΔsrtA-F	26
HT22	ΔspaH	This study
HT23	ΔspaG	This study
HT24	ΔspaI	This study

Generation of rabbit-raised polyclonal antibodies.

Appropriate oligonucleotide primers (Table 2) and C. diphtheriae chromosomal DNA were used for the PCR amplification of coding sequences for SpaG, SpaH, and SpaI, omitting signal peptide and sorting signal sequences. DNA fragments were cloned between appropriate sites of the expression vector pQE30 (QIAGEN), thereby generating expression signals and coding sequences for six-histidine-tagged recombinant proteins (N-terminal tag). The recombinant plasmids were transformed into E. coli XL1-Blue. DNA sequencing was used to verify the recombinant plasmids. The recombinant proteins were purified and injected into rabbits to raise antibodies as described previously (9).

TABLE 2.

Primers used in this study

Primer	Sequence
SpaG-Ab-5′	AAGGATCCAAAGCTGCCAAGAAAGCTGTTT
SpaG-Ab-3′	AAGGATCCTTATTTACCAAGTTCGTTCTTAAA
SpaH-Ab-5′	AAGGATCCCAAACCGAGCCGTCCGCT
SpaH-Ab-3′	AAGGATCCTTAGCGCTTGACGTTCTCGA
SpaI-Ab-5′	AAGGATCCCGCACAATCACCGGCGCT
SpaI-Ab-3′	AAGGATCCCTATTCCTTACGGAACTTCTC
SrtD-5′	AAACCCGGGCTGCCGCTGACCGGTGG
SrtD-3′	AAACCCGGGTTCCCCGCCGTTTCAGAGTT
SrtE-A-5′	AAAAAGCTTTGAGTTCGATTGGCTTTTTTTC
SrtE-B-3′	AAAGAATTCCCGGTGCCTCATCTAGCG
SrtE-C-5′	AAAGAATTCATGGCGATGAACTCTGAAACGGCGG
SrtE-D-3′	AAGCTTTTACTGGGACTCCTCGGCCGCT
SpaH-A-5′	AAAAAGCTTTGAGTTCGATTGGCTTTTTTTC
SpaH-B-3′	AAAGAATTCCCGGTGCCTCATCTAGCG
SpaH-C-5′	AAAGAATTCCTGAAAGGAAAACACCCTATG
SpaH-D-3′	AAAAAGCTTCACGCGACAAATCAATTACTC
SpaH-rbs-5′	CGCGGTACCATAAGTTGAGGGATTGTCCTGAAAGGAAAACACCCT
SpaH-Hind-3′	CGCGAATTCCAATTACTCGTTGATACGGC
SpaH-K202A-5′	GACGTCAACGTCTTCCCGGCCAACGGTAAGACTAAGCTC
SpaH-K202A-3′	GAGCTTAGTCTTACCGTTGGCCGGGAAGACGTTGACGTC

Generation of C. diphtheriae deletion mutants.

C. diphtheriae deletion mutants were generated by allelic exchange and verified by PCR and Western and Southern blotting techniques as previously described (9). Briefly, gene deletion cassettes were constructed by crossover PCR (15) and cloned between appropriate restriction sites of pK18mobsacB, and recombinant plasmids were transformed into E. coli S17-1 (6). Cultures of E. coli S17-1 and C. diphtheriae that were grown overnight were mixed in equal volumes and spread onto agar plates at 30°C for 16 h. Corynebacterial cointegrates were next isolated by plating conjugal bacteria onto HIA plates supplemented with 35 μg ml⁻¹ nalidixic acid and 50 μg ml⁻¹ kanamycin. Finally, C. diphtheriae deletion mutants were selected by plating cointegrates onto HIA plates supplemented with 10% sucrose and 35 μg ml⁻¹ nalidixic acid.

Plasmid construction. (i) Plasmid pSrtD.

Primers SrtD-5′ and SrtD-3′ were used to amplify the 5′ promoter sequence and untranslated region (UTR) of srtD and the coding sequence of srtD from C. diphtheriae NCTC13129 chromosomal DNA, while providing the amplified corynebacterial DNA segment with flanking SmaI sites for cloning purposes (Table 2). The PCR-amplified fragments were digested with appropriate enzymes, purified, and ligated into the cleaved SmaI sites of the E. coli/Corynebacterium shuttle vector pCGL0243 (26). The recombinant plasmid was electroporated into C. diphtheria cells.

(ii) Plasmids pSrtE and pSpaH.

Primers SrtE-A-5′ and SrtE-B-3′ or primers SrtE-C-5′ and SrtE-D-3′ were used to amplify, while appending EcoRI and HindIII sites to the fragments, the 5′ promoter sequence and UTR of the spaA or the srtE coding sequence, respectively, from C. diphtheriae NCTC13129 chromosomal DNA. The PCR-amplified fragments were digested with EcoRI and HindIII and ligated with the cleaved HindIII sites of the E. coli/Corynebacterium shuttle vector pCGL0243 (26) to generated pSrtE. The same strategy was employed with primers SpaH-A-5′, SpaH-B-3′, SpaH-C-5′, and SpaH-D-3′ to clone the 5′ promoter sequence and UTR of the spaA or spaH coding sequence into vector pCGL0243 to generate pSpaH. The recombinant plasmids were electroporated into C. diphtheria cells.

(iii) Plasmid pPlac-SpaH.

Primers SpaH-rbs-5′ and SpaH-Hind-3′ were used to PCR amplify the coding sequence of spaH while appending the ribosomal binding site of spaA from C. diphtheriae NCTC13129 chromosomal DNA. This fragment was digested with EcoRI/KpnI and ligated into the EcoRI/KpnI-cleaved Corynebacterium/E. coli shuttle vector pEKEx2, which contains an IPTG (isopropyl-β-d-thiogalactopyranoside)-inducible promoter (8). The recombinant plasmid was electroporated into C. diphtheria cells.

Site-directed mutagenesis of spaH.

PCR-based site-directed mutagenesis of double-stranded DNA was employed in this study (28). Plasmid DNA was used as a template for PCR amplification with Pfu DNA polymerase using primer sets (5′ and 3′) flanking six codons on both sides of the K202 mutation. Following PCR, the amplified plasmids were digested overnight at 37°C with DpnI to select against parental DNA molecules. The digested plasmids were transformed into E. coli cells. Mutant plasmids were verified by DNA sequencing and transformed into C. diphtheriae cells by electroporation according to a previously published protocol (26).

Extraction of C. diphtheria pili.

Extractions of C. diphtheriae pili were carried out as previously described (26). Briefly, C. diphtheriae strains were scraped from TSASB plates after growth overnight and washed in SMM buffer (0.5 M sucrose, 10 mM MgCl₂, and 10 mM maleate, pH 6.8). Cell pellets were suspended in the same buffer and treated with mutanolysin (300 units ml⁻¹) at 37°C for 6 h or overnight. Solubilized pili were isolated from the supernatant after centrifugation at 16,000 × g, followed by trichloroacetic acid precipitation and acetone washing and drying protein samples under a vacuum. Pilus preparations were boiled in sodium dodecyl sulfate (SDS) sample buffer, separated on an SDS-polyacrylamide gel electrophoresis (PAGE) gel, subjected to immunoblotting with rabbit antisera (anti-SpaH at a 1:5,000 dilution, anti-SpaA at a 1:20,000 dilution, and anti-SpaD at a 1:20,000 dilution) and anti-rabbit horseradish peroxidase-linked immunoglobulin G (IgG) antibody, and detected by chemiluminescence.

Electron microscopy and immunogold labeling.

Electron microscopic experiments were carried out as previously described (9). Briefly, bacterial strains were grown overnight on agar plates (TSASB), washed in 0.1 M NaCl, and stained with 1% uranyl acetate. For immunogold labeling, a drop of bacterial suspension was placed onto carbon grids, washed three times with phosphate-buffered saline (PBS) containing 2% bovine serum albumin (BSA), and blocked for 1 h in PBS with 0.1% gelatin. Pili were stained with a primary antibody diluted 1:100 in PBS with 2% BSA for 1 h, followed by washing and blocking. Pili were stained with gold goat anti-rabbit IgG (Jackson ImmunoResearch, PA) diluted 1:20 in PBS with 2% BSA for 1 h, followed by washing in PBS with 2% BSA. For double-labeling experiments, the same procedure was applied using another primary antibody and goat anti-rabbit IgG conjugated with gold particles of different sizes. The grids were washed five times with water before staining with 1% uranyl acetate was done. Samples were viewed by using a Jeol 100CX electron microscope.

Induced expression of SpaH.

Cultures of C. diphtheriae grown overnight were diluted 1:50 in heart infusion broth supplemented with kanamycin and allowed to grow to an optical density at 600 nm of 0.4 at 37°C. IPTG was then added to a final concentration of 1 mM. Duplicate aliquots of bacterial cultures were taken at various time intervals, and cell pellets were harvested by centrifugation. An equal amount of cells for each sample was adjusted by optical density. One set of samples was treated with mutanolysin (300 U/ml) to solubilize the pili. Solubilized pili were analyzed by immunoblotting with specific antibodies as described above. The other set of samples was subjected to immunoelectron microscopy (IEM) as described above.

RESULTS

Components of the third pilus of Corynebacterium diphtheriae.

The third pilus gene cluster of C. diphtheria contains spaG-spaH-srtD-srtE-spaI, with srtD and srtE encoding the putative sortases SrtD and SrtE, respectively (see Fig. 4A). To determine whether SpaH, SpaG, and SpaI are pilus components, we examined corynebacteria by IEM using a specific rabbit antibody raised against a purified pilin as well as gold particles conjugated with goat anti-rabbit IgG. When the wild-type cells were reacted with anti-SpaH and gold-labeled IgG, we observed immunogold labeling of stubby fibers external to the cell body (Fig. 1A). No SpaH-labeled structures were observed in the isogenic strain of C. diphtheriae HT22 (ΔspaH), which lacks the spaH gene (Fig. 1B). Moreover, control rabbit sera did not result in the labeling of pilus fibers (data not shown). Complementation of ΔspaH mutant HT22 with wild-type spaH on a multicopy plasmid not only restored the SpaH-positive pilus but led to the formation of extended pili (Fig. 1C), similar to the extended pilus phenotype previously observed with overproduced SpaA and SpaD (9, 26). We conclude that not only is SpaH necessary to form pilus structures, it is also sufficient to drive the assembly of long polymers.

FIG. 4.

Sortases SrtD and SrtE are required for the assembly of SpaH-type pili. (A) Graphic representation of the third gene cluster containing two sortase genes (srtD and srtE) along with three sortase-mediated pilus assembly genes (spaH, spaI, and spaG). The conserved LPXTG motif of Spa proteins is indicated. Arrows indicate the positions of predicted promoters as well as the direction of transcription. The locations of designated genes in the genome are shown by numbers. (B) C. diphtheriae strain NCTC12139 (wild type [WT]) or its isogenic derivatives carrying an individual or combined deletion of sortase genes were treated with muramidase (M) prior to extraction with hot SDS sample buffer. Proteins were separated on 4 to 12% SDS-PAGE gels and detected by immunoblotting with the specific antiserum, anti-SpaH (α-SpaH).

FIG. 1.

Localization of SpaH, SpaI, and SpaG in pili on the surface of Corynebacterium diphtheriae. Wild-type (WT) C. diphtheriae strain NCTC12139 (A) or its isogenic derivatives carrying a deletion of the spaH gene (ΔspaH) (B) transformed with plasmid pSpaH encoding wild-type spaH (C to F) were immobilized on carbon grids and stained with specific antiserum, anti-SpaH (α-SpaH) and IgG-labeled 6-nm gold particles (open arrow). Double-labeling experiments (D and F) were followed by staining of the cells with anti-SpaI (D) or anti-SpaG (F) and 12-nm gold-labeled IgG (filled arrows). An enlarged area of D is shown in E. Samples were viewed by transmission electron microscopy. Bars indicate a distance of 0.5 μm.

To determine if SpaI and SpaG are associated with the SpaH pilus structure, we carried out a double-labeling experiment (26). Cells were first reacted with anti-SpaH followed by 6-nm gold-labeled IgG. These washed cells were then treated with anti-SpaI followed by 12-nm gold-labeled IgG. SpaI staining (Fig. 1D and E, filled arrow) occurred along the SpaH pilus structures (Fig. 1D and E, open arrow), indicating that both SpaH and SpaI assemble onto the same pilus shaft (Fig. 1D and E). A similar experiment was performed with anti-SpaH and anti-SpaG, with 12-nm gold-labeled SpaG appearing to locate at the tip of an SpaH structure observed in a majority of the cell population (>95%) (Fig. 1F, filled arrow). Together, the data indicate that SpaH forms the pilus shaft, SpaI decorates the pilus structure, and SpaG may be positioned at the pilus tip, as is the case with SpaC or SpaF (9, 26).

The SpaHIG pilus is expressed and assembled independently from other corynebacterial pili.

Previous work showed that the expression and assembly of the SpaD-type pilus is independent of other pili. To determine if this was the case for the SpaHIG pilus, we examined the polymerization of pili with a previously developed assay (26). Corynebacterial pili were isolated by treatment of cells with a murein hydrolase, muramidase, and the hydrolyzed material was then boiled in SDS sample buffer with a reducing agent followed by electrophoresis in polyacrylamide gels and immunoblotting with rabbit antisera (9). Muramidase treatment solubilized pili of wild-type strain NCTC13129, and the pilin species were identified by immunoblotting with anti-SpaH, anti-SpaA, or anti-SpaD (Fig. 2). Immunoreactive HMW SpaH was detected within the 4% SDS-PAGE stack with a mass greater than 200 kDa, whereas SpaH monomers (SpaH_M) migrated just above a 55-kDa marker (SpaH precursor predicted mass, 59 kDa) (Fig. 2A). Deletion of spaH (ΔspaH) eliminated the synthesis and polymerization of SpaH monomers as expected (Fig. 2A). The introduction of wild-type SpaH into strain HT22 (ΔspaH) via an expression plasmid restored the synthesis and assembly of SpaH (Fig. 2A). In contrast, the deletion of spaG, spaI, spaA, or spaD did not affect the polymerization of SpaH precursor proteins (Fig. 2A). Similarly, the formation of SpaA and SpaD pili was independent of SpaH pilus synthesis and assembly, as the deletion of spaH, spaG, spaI, or sortases srtDE did not abolish SpaA or SpaD pili (Fig. 2B and C and data not shown).

FIG. 2.

SpaH pili are synthesized independently of SpaA or SpaD pili. Wild-type (WT) Corynebacterium diphtheriae strain NCTC12139, the isogenic derivatives carrying an individual deletion of spa genes, and the spaH deletion mutant (ΔspaH) harboring a plasmid containing spaH were treated with muramidase prior to extraction with hot SDS sample buffer. Proteins were separated on SDS-PAGE gels (4% stack) and detected by immunoblotting with the specific antisera, anti-SpaH (α-SpaH) (A), anti-SpaA (B), and anti-SpaD (C). The monomer (e.g., Spa_M) and high-molecular-weight mature products (e.g., HMW) of pilus assembly and the positions of molecular weight markers (in thousands) are indicated.

To examine whether the assembly of SpaH pili was independent of SpaA and SpaD pilus organelles, we introduced the wild-type spaH gene on a multicopy plasmid into mutant ΔspaA or ΔspaD and examined pilus structures by using IEM. The overexpression of major subunit SpaH allowed us to easily observe pili that are longer than those expressed from the chromosome, as described in previous studies (9). Deletion of spaA or spaD did not abrogate the expression and assembly of the SpaH pilus structures (Fig. 2 and 3A and B). To differentiate the SpaH-type pili from the other two types, plasmid pSpaH was introduced into wild-type corynebacterial strain NCTC13129 (wild type). We next confirmed that the SpaH structures are different from SpaA pili by double-labeling experiments (9). We observed two separate pilus types, extended SpaH pili labeled with 6-nm gold particles (Fig. 3C, open arrow) and shorter SpaA pili labeled with 12-nm gold particles (Fig. 3C, filled arrow). Likewise, when SpaH was overexpressed with the pSpaH plasmid, two different pili were detected using IEM with anti-SpaH and anti-SpaD on the surface of the wild-type strain, long pili labeled by anti-SpaH (open arrow) and short pili labeled by anti-SpaD (filled arrow) (Fig. 3D). In addition, no labeled SpaB, SpaE, or SpaF was detected along the SpaH fibers when the corresponding gold-labeled antibodies were used (data not shown). Together, these results demonstrate that SpaH, SpaG, and SpaI form a pilus structure distinct from those of other pilus types of C. diphtheriae.

FIG. 3.

Distinct structure of SpaH-type pili on the surface of Corynebacterium diphtheriae. Wild-type (WT) C. diphtheriae strain NCTC12139 (C and D) or its isogenic derivatives carrying a deletion of the spaA gene (ΔspaA) (A) or the spaD gene (ΔspaD) (B) transformed with plasmid pSpaH encoding wild-type spaH were immobilized on carbon grids and stained with specific antiserum, anti-SpaH (α-SpaH) and IgG-labeled 12-nm gold particles (A and B). In double-labeling experiments (C and D), SpaH pili were stained with anti-SpaH and 6-nm gold-labeled IgG followed by anti-SpaA (C) or anti-SpaD (D) and 12-nm gold-labeled IgG. Samples were viewed by transmission electron microscopy.

SrtD and SrtE sortases are required for the polymerization of SpaHIG pili.

Six sortase genes are found in the genome of C. diphtheriae NCTC13129, five of which are associated with the three pilus gene clusters (26). In the spaHIG gene cluster, sortases srtD and srtE are located after the spaH gene (Fig. 4A). In the first gene cluster, which contains only sortase srtA, it was previously shown that SrtA is essential for the assembly of SpaABC pili (26). Likewise, the assembly of SpaDEF pili requires only sortase genes srtB and srtC, located within the spaD pilus gene cluster (9). To investigate whether the same is true for the SpaHIG pilus, we analyzed strain HT21 (ΔsrtA-F), which contains a deletion of all six sortases (srtA, srtB, srtC, srtD, srtE, and srtF), as well as strains lacking individual sortase genes (26). The polymerization of pili was examined by SDS-PAGE as described above. As expected, an individual deletion of srtA, srtB, srtC, or srtF did not affect SpaH pilus assembly and the production of high-molecular-weight SpaH species (SpaH_HMW) (Fig. 4B). However, the deletion of all six sortase genes in strain HT21 abolished the polymerization of SpaH precursors into high-molecular-weight species, and the SpaH precursor accumulated, migrating as a 59-kDa protein (Fig. 4B). This effect of sortase deletion on SpaH is similar to that on SpaA or SpaD in strain HT21 (9, 26). The same effect on SpaH was also observed in the strain that lacks only sortase genes srtD and srtE, indicating that sortases SrtD and SrtE are specific for the assembly of SpaH pili (Fig. 4B). In the strain lacking either srtD or srtE, the polymerization of SpaH monomers into high-molecular-weight polymers is significantly slower than the that in the wild-type strain, thus leading to the accumulation of SpaH monomers in high abundance. This defect is more profound in the srtD deletion mutant. This is expected if SpaH polymerization is dependent upon both SrtD and SrtE. A high-molecular-mass species (200 kDa) accumulated in the srtDE double mutant complemented with an srtE plasmid. The nature of this species and significance, if any, are not known.

Expression levels of SpaH pilins determine the pilus length.

In the experiments described above, the overexpression of SpaH led to the formation of extended fibers compared to the wild-type pili (Fig. 1C and 3), suggesting that the amount of SpaH precursors may dictate the pilus length. To address this notion experimentally, we generated a construct containing the spaH coding sequence under the control of an IPTG-inducible promoter in the Corynebacterium/E. coli shuttle vector pEKEx2 (8) (Fig. 5A). Mid-log-phase corynebacteria of strain HT22 (ΔspaH) harboring the generated plasmid were induced with 1 mM IPTG in liquid culture at 37°C. At various times, aliquots of cells in duplicate were removed from the induced cultures and harvested by centrifugation. One set of samples of washed cells was subjected to IEM as described above; for the other set, samples with equivalent amounts of cells were treated with muramidase and analyzed by immunoblotting with anti-SpaH (Fig. 5B). No SpaH_HMW was detected at time zero or in uninduced cells at 60 min. When induced by IPTG, the first SpaH polymers could be detected after 4 min, while SpaH monomers became more visible. As time increased, higher-molecular-weight species of SpaH were readily observed (Fig. 5B), indicating that the length of the pilus structure increased. Among high-molecular-weight species, a band migrating around 110 kDa appeared to be prominent, consistent with a dimer's molecular mass that might be a result of overproducing monomers. A specific antibody against SrtD, a sortase required for the assembly of SpaHIG pili, was used as a lysis and loading control, which showed no apparent differences by quantification (Fig. 5B and data not shown). In a parallel experiment, no SpaH_HMW was detected at various time intervals in a strain harboring the pEKEx2 vector that induced the expression of SpaH with an alanine substitution at the conserved lysine 202 of the SpaH pilin motif (Fig. 5C). Thus, long pilus structures are still dependent upon the covalent cross-linking of subunits involving the conserved lysine residue.

FIG. 5.

Inducible expression of SpaH pilins. (A) Plasmid pPlac-spaH was generated by cloning the coding sequence of spaH under the control of an IPTG-inducible promoter of a C. diphtheriae/E. coli shuttle vector, pEKEx2. Conserved lysine 202 of the SpaH pilin motif was substituted by an alanine in a control plasmid used in C. (B) Corynebacteria harboring pPlac-spaH were induced by the addition of 1 mM IPTG, and samples with an equal number of cells were taken at various time intervals for treatment with muramidase. Solubilized pilins were boiled in SDS sample buffer, separated on 4 to 12% SDS-PAGE gels, and detected by immunoblotting with the specific antiserum, anti-SpaH (α-SpaH). (C) A parallel experiment as described in B was performed with corynebacteria harboring a control plasmid that expressed mutant SpaH pilins. SrtD was used as a lysis and loading control.

We extended these experiments by immunoelectron microscopy. When labeled with ant-SpaH and gold-conjugated IgG, at time zero, no labeled SpaH pili were observed on the cell surface (Fig. 6A). When SpaH was induced, SpaH labeling became evident starting at 4 min (Fig. 6B). As the induction period increased, SpaH-labeled pili increased in length, and SpaH labeling was distributed uniformly along the pilus shaft (Fig. 6C to E). No labeled SpaH was detected in uninduced cells at 1 h (Fig. 6F). These phenotypes were observed in a majority of the cell population (>90%). Altogether, we conclude that the spaHIG pilus gene cluster encodes the type III pilus, which is distinct from the other two types of C. diphtheriae pili. Moreover, the pilus length is governed by the abundance of the major subunit SpaH.

FIG. 6.

Pilus length induced by controllable expression of SpaH pilins. Corynebacteria harboring pPlac-spaH (A to E) were induced by the addition of 1 mM IPTG, and samples with an equal number of cells were taken at various time intervals for IEM with the specific antiserum anti-SpaH (α-SpaH). No IPTG (−) was added in F as a control. Samples were viewed by transmission electron microscopy. Bars indicate a distance of 0.5 μm.

DISCUSSION

Studies reported in this paper add to the growing evidence that many gram-positive bacteria harbor multiple pilus-encoding loci, which might have been acquired by a horizontal gene transfer mechanism (2, 9). We have characterized the third pilus in C. diphtheriae encoded by the spaHIG gene cluster. The SpaHIG pilus is composed of SpaH, SpaI, and SpaG, with SpaH and SpaI forming the pilus shaft and SpaG localized at the tip. Although C. diphtheriae has six sortase genes, the assembly of the SpaHIG pilus requires a specific sortase located within the pilus gene cluster. Structurally homologous components and sortase specificity are common features among corynebacterial pilus gene clusters. In a related mechanism, i.e., surface protein anchoring in gram-positive bacteria, sortase specificity is often determined by its substrates. In the case of Staphylococcus aureus SrtA and SrtB, SrtA recognizes and cleaves only LPXTG-bearing surface proteins, whereas SrtB is required only for NPQTN-bearing surface proteins (18). A similar phenomenon has been observed for Listeria monocytogenes sortases SrtA and SrtB (3). Likewise, in Streptococcus pyogenes, SrtA is required for the cell wall anchoring of the LPXTG-containing proteins M6, F, ScpA, and GRAB, whereas SrtB is essential only for the LPXTG-containing T6 protein (1), which was later shown to be a pilus component (19). It has been speculated that these sortases may be differently regulated to allow the modulation of the surface display of factors required for bacterial colonization (1). It is increasingly clear that the requirement for sortase is specific, but how sortase has evolved to recognize multiple conserved motifs (pilin motif and sorting signal) for pilus assembly as well as cell wall anchoring of surface proteins is equally puzzling (25).

According to current models of pilus assembly in gram-positive bacteria (16, 22, 25), the process begins in the cytoplasm with the synthesis of pilin precursors containing a hydrophobic domain and a positively charged tail, which are then transported across the membrane by the Sec machinery. The hydrophobic domain and the positively charged tail allow the precursors to be retained within the secretory pathway for subsequent recognition and cleavage by sortase. A pilus-specific sortase cleaves the precursor between the threonine and glycine of the conserved LPXTG motif and forms an acyl-enzyme intermediate. Pilus polymerization occurs when a sortase-acyl intermediate with the first subunit is attacked by the free amino group of a conserved lysine residue (K) present in the pilin motif of the membrane-bound second subunit. Its sorting signal would in turn be cleaved by sortase and linked to the lysine of a third pilin subunit. The polymerization is terminated when polymers are incorporated into the cell wall via a lipid II precursor.

Based on available data, we favor a new, more economical model of pilus assembly whereby the nucleophilic attack occurs between two acyl-enzyme intermediates, each containing a thioester linkage that is more labile to serve as the leaving group in the polymerization/cross-linking reaction (Fig. 7). As the concentration of the pilin precursors increases (via regulation), more pilins will be added (reactions I and II), resulting in an increase in pilus length. Our model further suggests that the pilus structure grows from the base rather than from the top. This “base growth” mechanism does not involve an enormous amount of energy that would otherwise be needed for the alternative “tip growth” mechanism. In that case, as the pilus grows, the pilus polymer would have to bend for a nucleophilic attack from the amino group of the pilin subunit at the tip to an adjacent acyl-enzyme intermediate located on the membrane.

The evidence that pilus length is controllable by merely adjusting the level of the major pilin suggests that pilus genes are regulated. Such a regulation has previously been described for the pneumococcal pilus gene cluster, which involves a positive regulator, rlrA (rofA-like regulator), and a transcriptional repressor, mgrA (2). The deletion of mgrA in a pneumococcal strain in which pili are normally undetected leads to abundant pili on the cell surface (2). A similar regulatory system has been found in group B streptococci (2). Surprisingly, no rofA or rlrA homolog is apparent in the C. diphtheriae genome (unpublished data). How corynebacterial pili are regulated remains to be investigated. We speculate that factors that are involved in toxigenicity or pathogenesis may contribute to the regulated expression and assembly of pili by corynebacteria.