assemblymechanismoffctregiontype1piliinserotype … · plasmid was integrated into the chromosome...

Assembly Mechanism of FCT Region Type 1 Pili in SerotypeM6 Streptococcus pyogenes*□S

Received for publication, March 20, 2011, and in revised form, August 22, 2011 Published, JBC Papers in Press, August 31, 2011, DOI 10.1074/jbc.M111.239780

Masanobu Nakata‡, Keiji Richard Kimura‡, Tomoko Sumitomo‡, Satoshi Wada‡, Akinari Sugauchi‡, Eiji Oiki§,Miharu Higashino‡, Bernd Kreikemeyer¶1, Andreas Podbielski¶1, Nobuo Okahashi�, Shigeyuki Hamada**,Ryutaro Isoda‡, Yutaka Terao‡, and Shigetada Kawabata‡2

From the ‡Department of Oral and Molecular Microbiology, Osaka University Graduate School of Dentistry, 1-8 Yamadaoka, Suita-Osaka, 565-0871, Japan, §Center for Medical Research and Education, Graduate School of Medicine, Osaka University, 2-2Yamadaoka, Suita-Osaka, 565-0871, Japan, ¶Institute of Medical Microbiology, Virology and Hygiene, University of Rostock, D-18057, Rostock, Germany, �Department of Oral Frontier Biology, Osaka University Graduate School of Dentistry, 1-8 Yamadaoka,Suita-Osaka, 565-0871, Japan, and **Department of Medical Sciences, Research Collaboration Center on Emerging andReemerging Infections (RCC-ERI) 6F, Ministry of Public Health, Tiwanon Road, Muang Nonthaburi 11000, Thailand

Background: Group A streptococcal pili are diverse.Results: Mutational analyses identified a specific transpeptidase and lysine residue of the shaft protein required for pilusassembly in a serotype M6 strain.Conclusion: Assembly and cell wall anchoring of the pili are accomplished by concerted actions of transpeptidases.Significance:Our findings lead to predictions of homologous pilus assembly mechanisms in other species.

The human pathogen Streptococcus pyogenes producesdiverse pili depending on the serotype. We investigated theassembly mechanism of FCT type 1 pili in a serotype M6 strain.The pili were found to be assembled from two precursor pro-teins, the backbone protein T6 and ancillary protein FctX, andanchored to the cell wall in amanner that requires both a house-keeping sortase enzyme (SrtA) and pilus-associated sortaseenzyme (SrtB). SrtB is primarily required for efficient formationof the T6 and FctX complex and subsequent polymerization ofT6,whereas proper anchoring of the pili to the cell wall ismainlymediated by SrtA. Because motifs essential for polymerizationof pilus backbone proteins in other Gram-positive bacteria arenot present in T6, we sought to identify the functional residuesinvolved in this process. Our results showed that T6 encom-passes the novelVAKSpilinmotif conserved in streptococcal T6homologues and that the lysine residue (Lys-175) within themotif and cell wall sorting signal of T6 are prerequisites for iso-peptide linkage of T6 molecules. Because Lys-175 and the cellwall sorting signal of FctX are indispensable for substantialincorporation of FctX into the T6 pilus shaft, FctX is suggested

to be located at the pilus tip, which was also implied by immu-nogold electron microscopy findings. Thus, the elaborateassembly of FCT type 1 pili is potentially organized by sortase-mediated cross-linking between sorting signals and the aminogroup of Lys-175 positioned in the VAKS motif of T6, therebydisplaying T6 and FctX in a temporospatial manner.

Pili are long filamentous structures that extend from the bac-terial cell surface. After their discovery in a variety of patho-genic Gram-positive bacteria, multiple roles of pili in the infec-tion process and their impact related to vaccine candidates havebeen reported (1, 2). In contrast to pili of Gram-negative bac-teria, pilus subunits of Gram-positive bacteria are covalentlyattached to each other and anchored to cell wall peptidoglycans(3, 4). Typically, Gram-positive pili consist of one major andone or two minor subunits. Polymerization of the major sub-unit constitutes the pilus backbone shaft, whereas the minorsubunits are covalently incorporated into the backbone asancillary pilus proteins. These pilus subunits contain cell wallsorting signals (CWSSs),3 including a canonical LPXTG orLPXTG-like motif (5). Pilus subunits are assembled via thismotif, i.e. polymerization of major subunits and formation ofmajor andminor subunit complexes, and finally cross-linked tothe peptidoglycan layer. These processes are mediated by amembrane-bound transpeptidase termed sortase, including ahousekeeping sortase enzyme (SrtA) and pilus-associated sor-tase enzyme (SrtB in this study) (6–8).The mechanism of cell wall anchoring of surface proteins

governed by SrtA has been extensively investigated in Staphy-lococcus aureus (7). A large variety of surface proteins harbor-ing the LPXTG motif in CWSS are recognized by SrtA. SrtA

* This work was supported in part by Grant-in-aid for Scientific Research (B)20390465 from the Japan Society for the Promotion of Science and Grant-in-aid for Scientific Research on Priority Areas 18073011; Grant-in-aid forScientific Research for Young Scientists (A) 2189048; and Grants-in-aid forScientific Research for Young Scientists (B) 19791343, 20791336, and21791786 from the Ministry of Education, Culture, Sports, Science andTechnology.

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Figs. S1–S7 and Tables S1–S3.

The nucleotide sequence(s) reported in this paper has been submitted to the Gen-BankTM/EBI Data Bank with accession number(s) AB669020.

1 Supported by Deutsche Forschungsgemeinschaft (German Research Foun-dation) Grants Po391/12-1 and Kr1765/2-1 and the Bundesministerium fürBildung und Forschung (Federal Ministry of Education and Research) in theframework of ERANet PathoGenoMics I and II and SysMO Systems Biologyof Microorganisms I and II.

2 To whom correspondence should be addressed. Tel.: 81-6-6879-2896; Fax:81-6-6879-2180; E-mail: [email protected].

3 The abbreviations used are: CWSS, cell wall sorting signal; HLB, high molec-ular weight ladder band.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 286, NO. 43, pp. 37566 –37577, October 28, 2011© 2011 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

37566 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 286 • NUMBER 43 • OCTOBER 28, 2011

by guest on June 30, 2019http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/cgi/content/full/M111.239780/DC1

http://www.jbc.org/

cleaves LPXTG between threonine (Thr) and glycine (Gly) res-idues and then makes an acyl enzyme intermediate by linkingthe active cysteine residue to the carboxyl group of the threo-nine in the motif. Subsequently, the intermediate is relieved bynucleophilic attack followed by linkage of the carboxyl group ofthreonine to a free amino group in the growing cell wall.Regarding the pilus assembly, polymerization of major sub-

units and incorporation of minor subunits also require sortaseenzymatic activity (4, 9). In many cases, the pilus-associatedsortase catalyzes the formation of a peptide bond between thre-onine in the CWSSmotif of one subunit and an �-amino groupof a lysine (Lys) residue conserved in a pilin motif of the othersubunit (11, 12). Namely, the side chain of the lysine residue inthe pilus subunit functions as a nucleophile to polymerize themajor subunits and forms a complex of major and minor sub-units. Thus, these intermolecular isopeptide bonds formedbetween pilus subunits are fundamental for pilus assembly(13–15).Streptococcus pyogenes is a human pathogen that colonizes

the mucous membranes of the upper respiratory tract and epi-dermal layer of the skin. Consequently, S. pyogenes is responsi-ble for a wide variety of human diseases, ranging from self-limiting purulent infections to severe necrotizing fasciitis andautoimmune diseases, such as acute rheumatic fever and glo-merulonephritis (16). To successfully colonize anatomical sites,S. pyogenes achieves firm attachment to host cell surfaces.Among the numerous adhesins required for tissue-specificadherence, pili have been suggested to play a central role (1, 17,18).The pilus genes of S. pyogenes are encoded in the FCT

genomic region (loci encoding fibronectin-binding proteins,collagen-binding proteins, and T antigens) (19). Based on thevariability of gene composition and sequences, this region isclassified into at least nine subtypes, designated FCT types 1–9(20, 21). Among these, pilus components of FCT type 1 aremutually distinctive from those of the other types as shown byphylogenetic analyses of both the FCT region nucleotidesequences and the amino acid sequences of backbone and ancil-lary pilus proteins (20, 21). In serotype M6 strains belonging toFCT type 1, the pilus-related genomic region contains genesencoding Lancefield T6 antigen, FctX, and SrtB (16, 19, 22).Mora et al. (23) showed thatT6 is themajor subunit of FCT type1 pili, whereas FctX (also termed AP-1) serves as an ancillarypilus protein. The composition of FCT type 1 pili is uniquebecause FCT types 2–8 are composed of three pilus proteins.No remarkable homology on the protein level was observedamongT6, FctX, and FCT type 2/3 pilus proteins, i.e.Cpa, FctA,and FctB (21). Although the assembly mechanism and expres-sion mode of FCT type 2/3 pili have been investigated, those ofother types of pili are poorly understood (23–25).In the present study, we demonstrate for the first time the

assembly mechanism of FCT type 1 pili in a serotype M6 strainin which SrtA and SrtB were shown to collaborate to assemblepilus subunits and anchor pili to the peptidoglycan layer. Wealso found a lysine residue within a novel VAKS pilin motif ofT6 required for T6 polymerization and incorporation of FctX atthe tip of the pilus shaft. These unique characteristics of the

assembly mechanism of FCT type 1 pili give insights into thecomplexity of the biological nature of streptococcal pili.

EXPERIMENTAL PROCEDURES

Bacterial Strains and Culture Conditions—The S. pyogenesserotype M6 strain TW3558 was isolated from a patient withtonsillitis (26). The Escherichia coli strain TOP10 (Invitrogen)was used as a host for derivatives of plasmids pFW5-luc,pSET4s, and pAT18 (27–29). E. coli XL10-gold (Stratagene)served as a host for plasmid pQE30 derivatives (Qiagen). AllE. coli strains were cultured in Luria-Bertani (LB) medium at37 °C with agitation. The S. pyogenes wild type and isogenicmutant strainswere cultured inTodd-Hewitt broth (BDBiosci-ences) supplemented with 0.2% yeast extract (BD Biosciences)(THY medium) at 37 °C in an ambient atmosphere. For selec-tion andmaintenance ofmutants, antibiotics were added to themedia at the following concentrations: ampicillin (Wako), 100�g/ml for E. coli; kanamycin (Sigma-Aldrich), 50 �g/ml forE. coli; chloramphenicol (Sigma-Aldrich), 10 �g/ml for E. coli;spectinomycin (Wako), 100 �g/ml for E. coli and 60 �g/ml forS. pyogenes; and erythromycin (Sigma-Aldrich), 150 �g/ml forE. coli and 1 �g/ml for S. pyogenes.DNA and Computational Techniques—Purification of chro-

mosomal DNA from S. pyogeneswas done using a DNA extrac-tion kit (Takara). Plasmid DNA was purified with a plasmidpurification kit (Macherey-Nagel). Transformation of E. coliand S. pyogeneswas performed as described previously (30, 31).The DNA sequence of the FCT region of strain TW3558 wasdetermined using the reported genome sequence ofMGAS10394 (GenBank accession number NC_006086) as areference (32). The homologue of T6 was searched using thePSI-Blast algorithm. Selected amino acid sequences of thehomologues (e value, ��20; coverage, �90%; identity, �80%)were aligned using ClustalW software and visualized by CLCMain workbench software (CLC bio). The signal sequence ofT6 and FctX was deduced using the SignalP 3.0 server. Adomain search was conducted using the SMART server.Construction of Recombinant Vectors and Mutant Strains of

S. pyogenes—The construction of in-frame deletion mutantswas conducted using a temperature-sensitive shuttle vector,pSET4s, as reported previously (33, 34). The summarized pro-cedure is depicted in supplemental Fig. S1. During the course ofconstruction, a merodiploid strain was created after the firstallelic replacement and then resolved to possess either mutantor wild type alleles after the second allelic replacement. To ruleout the effects of secondary mutations that may have arisenduring mutagenesis, a clone possessing the wild type allele wasalso used as a revertant strain in this study. Both an in-framedeletion mutant and a revertant strain arose from the samemerodiploid ancestor. The correct in-frame deletion of geneswas confirmed by site-specific PCR and sequence analyses ofthe chromosomal locus encompassing the recombination sites.All primers used are listed in supplemental Table S1.For the construction of the fctX operon-luc reporter strain,

the 3�-end of the operon was amplified with PCR using primersFctXOpe-LucF and FctXOpe-LucR (supplemental Table S1)and cloned into multiple cloning site I of a luciferase reporterplasmid, pFW5-luc, via the NheI/BamHI sites. The resulting

Assembly Mechanism of Streptococcal FCT Type 1 Pili

OCTOBER 28, 2011 • VOLUME 286 • NUMBER 43 JOURNAL OF BIOLOGICAL CHEMISTRY 37567


ww

.jbc.org/D

ownloaded from




http://www.jbc.org/

plasmid was integrated into the chromosome of strainTW3558. Correct chromosomal integration of the luciferasegene was confirmed by site-specific PCR.To generate mutated tee6, an overlapping PCR strategy was

used with the primers listed in supplemental Table S2. Themutated fragment was cloned into the streptococcal shuttlevector pAT18-PgyrA (24). Introduction of the mutation wasconfirmed by DNA sequencing. Wild type tee6 was also clonedinto pAT18-PgyrA. The resultant plasmids were transformedinto the tee6 in-frame deletionmutant (�tee6) and grown in thepresence of erythromycin.To create an fctX expression vector, the fctX gene, including

the putative promoter, was amplified using primers 3558OpFand 3558FctXR (supplemental Table S3) and Phusion hot startDNA polymerase (Finnzymes). The PCR product was clonedinto pDONR221 (Invitrogen) using a BP recombination reac-tion. Then, utilizing an LR recombination reaction, the fctXgene fragment was cloned into pAT18-ccdB in which a blunt-ended cassette containing attR sites flanking the ccdB gene andthe chloramphenicol resistance gene (Invitrogen) was insertedinto the SmaI site of pAT18 according to the manufacturer’sinstructions. Themutant fctX fragment was created by overlap-ping PCR using primers 3558OpF and 3558FctXR as well as theprimers with mutations listed in supplemental Table S3. Thefragment was also cloned into pAT18-ccdB. Correct introduc-tion of themutation was confirmed by DNA sequencing. Theseconstructs were transformed into the in-frame fctX deletionmutant (�fctX). The procedures are briefly depicted in supple-mental Fig. S2.Preparation of Antisera against T6 and FctX—Recombinant

T6 and FctX proteins were prepared as described before (24).Briefly, recombinant proteins were hyperexpressed in E. colicells and purified from E. coli lysates using a QIAexpress pro-tein purification system (Qiagen) according to themanufactur-er’s instructions. Eluted proteins were dialyzed against phos-phate-buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 5.4 mM

Na2HPO4, 1.8mMKH2PO4, pH 7.4), and the concentrationwasdetermined using a BCA protein assay kit (Thermo Scientific).Mouse antisera against FctX and T6 were raised by immu-

nizing BALB/c mice with the purified recombinant proteins asdescribed before (24). Rabbit antiserum against FctXwas raisedby immunizing male New Zealand White rabbits with purifiedrecombinant FctX proteins. Briefly, the rabbits were vaccinatedintradermallywith amixture of 400�g of recombinant proteinsand complete Freund adjuvant (Difco) followed by four boostswith 200 �g of recombinant proteins with incomplete Freundadjuvant (Difco) at 2-week intervals. The antiserum was puri-fied for the immunogold labeling of FctX as described previ-ously (35).Extraction of Cell wall and Supernatant Fractions and Prep-

aration for Immunoblotting—S. pyogenes strains were grown tothe late exponential phase or overnight in THY medium andwashed twice with PBS. The cells were suspended in protoplas-ting buffer (0.1 M KPO4, pH 6.2, 40% sucrose, 10 mM MgCl2)containing Complete EDTA-free protease inhibitors (RocheApplied Science) and 250 �g/ml N-acetylmuramidase (Seika-gaku Biobusiness), and then incubation was performed for 3 hat 37 °C. The resulting protoplasts were sedimented by centrif-

ugation at 16,000� g for 10min, and the resultant supernatantswere subjected to immunoblot analyses as cell wall fractions.Total proteins in the culture supernatants were precipitated

with 10% trichloroacetic acid on ice for 1 h followed by centrif-ugation at 16,000� g for 10min. The pellets were washed twicewith ice-cold ethanol and resuspended in 1 M Tris buffer, pH8.0.The proteins of each fraction were separated by sodium

dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 3–10% acrylamide gels (Wako) and blotted ontonitrocellulose membranes. The membranes were blocked withcasein-based blocking reagent (Snow Brand), incubated for 1 hwith mouse antisera diluted 1:2,000 in PBS containing 0.2%Tween 20 (PBST), washed three times with PBST, and incu-bated with horseradish peroxidase-conjugated anti-mouse IgGantibody (Cell signaling) diluted 1:2,000 in PBST for 1 h. Fol-lowing washing steps, the membranes were developed withPierce Western blotting substrate (Thermo Scientific).Immunoprecipitation—The cell wall fraction of the wild type

strain was prepared as noted above. The fraction was then dia-lyzed against PBS and incubated with rabbit anti-FctX antise-rum or non-immune rabbit serum at 4 °C overnight. Subse-quently, protein G-coated magnetic beads (Invitrogen) wereadded to the mixture (10%, v/v) and incubated at 4 °C for 5 h.The beads were extensively washed with PBST five times andresuspended in SDS-containing sample buffer (50 mM Tris, pH6.8, 1% SDS, 0.1% bromphenol blue, 10% glycerol) containing0.1 M dithiothreitol followed by boiling for 5 min and brief cen-trifugation. The resultant supernatant samples were subjectedto immunoblot analysis using mouse anti-T6 antiserum andnon-immune mouse serum.RT-PCR—A total RNA sample was prepared from the wild

type strain grown to the late exponential phase (36). Synthesisof cDNA from total RNA was conducted with a Transcripterhigh fidelity cDNA synthesis kit (Roche Applied Science)according to themanufacturer’s instructions. The possibility ofDNA contamination was excluded by PCR analysis withnon-RT samples. All primers used are listed in supplementalTable S1. The RT-PCR amplifications were performed using ExTaq (Takara).Quantitative Assays for Luciferase Activity—An fctX operon-

luc reporter strain was cultured in THY medium at 37 °C. Tomeasure luminescence, aliquots from bacterial cultures werewithdrawn at 30-min intervals and processed as described pre-viously (37), using D-luciferin (Wako) and a GloMax 20/20nluminometer (Promega).Immunogold Electron Microscopy—Immunogold labeling of

whole bacteria was conducted as described previously with aminormodification (38).Wild type and isogenicmutant strainswere grown to the midexponential phase (A600�0.5), thenwashed twice with PBS, and resuspended in PBS. A drop ofbacterial suspension was placed onto carbon-coated Formvar-covered nickel grids (Pyser-SGI) and allowed to settle for 15min. Bacteria on the grids were fixed with 1% paraformalde-hyde (Wako) for 5min at room temperature and incubatedwithblocking buffer (PBS containing 2%bovine serumalbumin (Sig-ma-Aldrich) and 100 �g/ml human Fc fragment (JacksonImmunoResearch Laboratories) for 1 h at room temperature.




ww

.jbc.org/D

ownloaded from








http://www.jbc.org/

Pili were labeled with polyclonal mouse or rabbit antiseradiluted 1:10 with PBS containing 10% (v/v) blocking buffer for1 h followed by washing five times with PBS. Subsequently, thegrids were incubated with 15- or 20-nm gold-conjugated sec-ondary antibodies (BB International) diluted 1:20 in PBS con-taining 10% (v/v) blocking buffer. After washing steps with PBSand distilled water, the grids were air-dried. Gold-labeled pilusproteins were observed with an H-7650 transmission electronmicroscope (Hitachi) at an accelerating voltage of 80 kV.

RESULTS

FCTType 1 Pilus Genes Are Transcribed as Operon—To gaininsight into the genetic features of the FCT type 1 pilus region,we first performed transcriptional analysis of the region usingserotype M6 strain TW3558, hereafter referred to as WT. Thepilus region of the strain contains the tee6, fctX, and srtB genesencoding T6, FctX, and SrtB, respectively. These genes arelocated in tandem between the M6 spy0158 and M6 spy0162genes (Fig. 1A). To analyze whether these contiguous genes aretranscribed in a polycistronic manner, RT-PCR analysis wasperformed using four sets of primers designed to amplify theboundaries of each gene (Fig. 1A, four pairs of arrowheads).Amplicons between fctX and tee6 as well as tee6 and srtB weredetected in the PCRs using cDNAofWTas a template, whereasthe regions betweenM6 spy0158 and fctX and between srtB andM6 spy0162were not amplified (Fig. 1B). In addition, no ampli-con was detected in the non-RT control samples. Because theputative transcriptional terminator located downstream of

both the M6 spy0158 and srtB genes was present (data notshown), it is most likely that the fctX, tee6, and srtB genes aretranscribed in a polycistronic manner as an operon (referred toas fctX operon).To analyze the temporal transcription profile of the fctX

operon, a luciferase gene reporter fusion system was used todetermine temporal activities (Fig. 1C). After 5 h in a batchculture reaching the late exponential phase, the fctX operon-lucconstruct demonstrated a sharp peak in luciferase activity. Ourdata indicated that FCT type 1 pilus genes are transcribed at themaximum amount in the late exponential phase.Roles of T6, FctX, and SrtB in Assembly of FCT Type 1 Pili—

As a first step toward understanding the mechanism by whichFCT type 1 pili are assembled, we sought to clarify the roles ofT6, FctX, and SrtB in that process. Because the correspondinggenes were shown to be transcribed in a polycistronic manner,markerless in-frame deletion mutants and revertant strainswere constructed to avoid polar effects on the expression ofneighboring genes. The growth rates of all constructedmutantswere similar to that of WT (data not shown).Cell wall fractions ofWT and themutant strains were immu-

noblotted with anti-T6 and anti-FctX antisera (Fig. 2). In thecell wall fraction ofWTand all revertant strains, highmolecularweight ladder bands (HLBs) reflecting polymerization of T6and the incorporation of FctX into T6 were detected with bothanti-T6 and FctX antisera on immunoblots (Fig. 2, A and B).Furthermore, formation of the T6 and FctX complex was sub-

FIGURE 1. Transcriptional analysis of FCT type 1 pilus region of serotype M6 strain. A, shown is a genetic map from serotype M6 strain TW3558. Genedesignations are presented above each arrow. The M6 spy numbers were assigned according to the corresponding genes in the M6 genome sequence (strainMGAS10394, GenBank accession number NC_006086). Black arrow, transcriptional regulator rofA; dark gray arrow, fibronectin-binding protein prtF1; light grayarrow, pilus proteins tee6 and fctX; light gray hatched arrow, pilus region-associated sortase srtB; white arrow, heat shock protein, putative transposase, andhypothetical protein. B, RT-PCR analysis of the pilus region. cDNA was synthesized from total RNA recovered from a late exponential phase culture of a wild typestrain. Total RNA subjected to the same reaction without reverse transcriptase (Non-RT) served as a negative control of PCRs. Four sets of primers (shown asarrowheads in A) were used to amplify fragments 1– 4 (shown as bars in A). The PCR products were subjected to electrophoresis on 1.5% Tris borate-EDTAagarose gels and visualized by staining with ethidium bromide. The numbers for each lane correspond to the expected PCR fragments 1– 4. The DNA markersize is indicated on the left. M, 100-bp ladder DNA marker. C, temporal transcriptional activity of fctX operon. Luciferase activities (black circles) and culturedensities (white circles) of the fctX operon-luc strain were determined. Luciferase activity is represented as relative light units (RLU). Three experiments wereperformed with data from a representative experiment shown. Values are presented as the mean � S.D. of triplicate samples. Error bars represent S.D.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

stantiated by results of immunoprecipitation assays (supple-mental Fig. S3). The bands corresponding to the monomericforms of T6 and FctXwith amolecularmass of50 and 70–80kDa, respectively, were also detected (Fig. 2, A and B, arrow-heads). The deduced molecular mass of FctX without a signalsequence and cell wall sorting signal is 108 kDa, thus, the70–80-kDa band detected in the blot with anti-FctX antiseramay correspond to either the breakdownproduct or a protease-cleaved product of FctX (Fig. 2B). In contrast, in the cell wallfraction of�tee6,HLBs completely disappeared, and themono-meric form of FctX was clearly detected, indicating that T6 isthe shaft protein and that FctX is the ancillary protein of pili(Fig. 2B, lane 2). On the other hand, the intensity of the HLBswas decreased, and the migration pattern changed with dele-tion of fctX (Fig. 2A, lane 4), which suggests that FctX isrequired for efficient polymerization of T6. The bands detectedbelow those representing the (T6)n homopolymer might havebeen the proteolytic products of (T6)n. The intensity of theHLBs also decreased, and the pattern changed in the cell wallfraction of the srtB deletionmutant (�srtB; Fig. 2,A and B, lane6). However, the pattern of HLBs of�srtB did not coincide withthat of �fctX, indicating that all HLBs of �srtB reflect the T6and FctX complex, (T6)n-(FctX)1, based on molecular masssizes. Of note, pilus assembly was not stopped even in theabsence of SrtB.Both SrtA and SrtB Are Required for Efficient Assembly and

Cell Wall Anchoring of FCT Type 1 Pili—Because the polymer-ization of T6 and incorporation of FctX into T6 were not abol-ished by srtB deletion, we investigated the possible involvementof the housekeeping sortase SrtA in pilus assembly. Mutantstrains with srtA (�srtA) deletion and srtA/srtB double deletion(�srtA/srtB) were constructed. For double knock-out of thesrtA and srtB genes, a deletion of srtB was introduced into

�srtA. Therefore, the genotype of the revertant strain (desig-nated WrsrtB/�srtA in Fig. 3) was the same as that of �srtA.Cell wall and culture supernatant fractions were prepared

from these mutants, including �srtB and its revertant strain,and subjected to Western blot analysis using anti-T6 and anti-FctX antisera (Fig. 3, A and B). HLBs that reacted with bothantisera were still present in the cell wall fraction of �srtA;however, the signal intensitywas remarkably decreased as com-pared with that ofWT (Fig. 3, A and B, lane 1 versus lane 4). Ofnote, HLBs were readily detected in the culture supernatant of�srtA andWrsrtB/�srtA (Fig. 3, C andD, lanes 4 and 7), whichstrongly indicates that SrtA mainly functions in the finalanchoring of pili to the cell wall. Because a lower level of T6monomer was detected in both fractions of �srtA, it can bespeculated that T6 polymerization in �srtA is more efficientthan inWT.Whenboth srtA and srtBwere deleted, noHLBwasdetected in the cell wall fraction (Fig. 3, A and B, lane 6). Takentogether, these findings indicate that SrtB is primarily requiredfor T6 polymerization and the complex formation of T6 andFctX. However, in �srtB, SrtA can compensate to a certainextent for the role of SrtB in formation of that complex and T6polymerization. In contrast, SrtA apparently plays a central rolein cell wall anchoring of assembled pili. Hence, the functions ofboth sortases are distinctive during the course of pilusassembly.LPSTG Motif of T6 Is Not Required for Formation of T6 and

FctX Complex—As shown in Fig. 3, SrtB is primarily requiredfor T6 polymerization and formation of the T6 and FctX com-plex. To verify whether the LPSTG motif in the CWSS of T6,which might be processed by sortase enzymes, is necessary forT6 polymerization and complex formation, we introduced amutation into this motif and investigated the effect on pilusassembly. Mutated tee6 in which the canonical LPSTG motif

FIGURE 2. Immunoblot analyses of T6 and FctX in cell wall extracts from wild type and mutants of FCT type1 pilus region genes. Total proteins in the cellwall fraction of the wild type and its derivatives were subjected to immunoblotting with mouse antisera against T6 (A) and FctX (B). Lane 1, wild type; lane 2, tee6deletion mutant (�tee6); lane 3, tee6 revertant (Wrtee6); lane 4, fctX deletion mutant (�fctX); lane 5, fctX revertant (WrfctX); lane 6, srtB deletion mutant (�srtB);lane 7, srtB revertant (WrsrtB). The position of monomeric forms of T6 and FctX is indicated by an arrowhead. The deduced positions of the complex forms of T6and FctX are indicated on the right. The protein marker sizes are indicated on the left. M, protein size marker.




ww

.jbc.org/D

ownloaded from



http://www.jbc.org/

was replaced with LG was expressed in �tee6 using a shuttlevector (�PST), whereas the vector harboring wild type tee6 andan empty vector were also transformed and used as controls(wild type andmock, respectively). In the cell wall extract of thewild type, HLBs detected with anti-T6 and anti-FctX antiserawere visible in immunoblots (Fig. 4, A and B, lane 2). In con-trast, HLBs were not detected in the fractions of mock and�PST (Fig. 4,A andB, lanes 1 and 3). In addition, noHLBs weredetected in the concentrated culture supernatant of the �PSTmutant (supplemental Fig. S4, lane 11), indicating that theLPSTG motif is essential for T6 polymerization.

Despite the lack of HLBs in the cell wall extract of the�PST mutant, a band with a molecular mass of 130 kDawas visible (Fig. 4, A and B, lane 3, arrowheads). Because this130-kDa band, which reacted with both anti-T6 and anti-FctX antisera, was not visible in the extracts of the fctX dele-tion mutant transformed with an empty vector (�fctX) ormock (Fig. 4,A and B, lanes 1 and 5) and a band with the samemolecular mass was also detected in the immunoprecipita-tion assay (supplemental Fig. S3), it was identified as a het-erodimer band, which reflects the complex of T6 and FctX,(T6)1-(FctX)1. Therefore, we concluded that the LPSTG

FIGURE 3. Both SrtA and SrtB are required for assembly and cell wall anchoring of FCT type1 pili. Total protein in cell wall fractions (A and B) and culturesupernatants (C and D) of S. pyogenes strain TW3558 and its sortase mutants was subjected to immunoblotting with mouse antisera against T6 (A and C) andFctX (B and D). Lane 1, wild type; lane 2, srtB deletion mutant (�srtB); lane 3, srtB revertant (WrsrtB); lane 4, srtA deletion mutant (�srtA); lane 5, srtA revertant strain(WrsrtA); lane 6, srtA/srtB deletion mutant (�srtA/srtB); lane 7, srtB revertant strain in the background of srtA/srtB deletion mutant (WrsrtB/�srtA). The positionsof monomeric forms of T6 and FctX are indicated by arrowheads. The deduced positions of the complex forms of T6 and FctX are indicated on the right. The sizesof the molecular mass standards are indicated on the left. M, protein size marker.




ww

.jbc.org/D

ownloaded from



http://www.jbc.org/

motif of T6 is not necessary for formation of isopeptidebonds between T6 and FctX proteins.The 120-kDa band detected with anti-FctX antiserum, but

not with anti-T6 antiserum, in supernatant fractions may rep-resent released intact FctX monomers (supplemental Figs. S4,lanes 9–16, and S5, lanes 10–12 and 14–16). We consider thatthe prominent180-kDa band detected with both anti-T6 and-FctX antisera in the supernatant fraction of the �PST mutantis most likely the heterodimer of T6 and intact FctX (supple-mental Fig. S4, lane 11).When the canonical LPSTG motif of T6 was replaced with

the analogous LPSSG motif present in FctX, the intensity ofbands representing (T6)n was increased and that of those rep-resenting (T6)n-(FctX)1 was decreased in the cell wall fractionas compared with that of the wild type (Fig. 4, A and B, lane 2versus lane 4). In contrast, HLBs were detected more clearly inthe culture supernatant of the mutant (supplemental Fig. S4, Aand B, lane 10 versus lane 12). These findings indicated that theintroduced LPSSG sequence in T6 drives T6 homopolymeriza-tion and is not suitable for cell wall anchoring of pili.Whenwildtype tee6 or mutated tee6 genes were expressed in �tee6�srtB,HLBs and the heterodimer band completely disappeared inboth the cell wall and culture supernatant fractions (supple-mental Fig. S4), indicating a principal role of SrtB in complexformation and subsequent T6 polymerization.LPSSGMotif of FctX Is Required for Formation of T6 and FctX

Complex—The sorting signal of FctX includes an LPXTG-likemotif, i.e. LPSSG, which is likely recognized by sortases. Todetermine whether the LPSSG motif in CWSS of FctX func-

tions in formation of the complex, either wild type fctX ormutated fctX in which the LPSSG motif was replaced with LGwas expressed in �fctX (wild type and �PSS, respectively), andthe cell wall fractions of these mutants were immunoblottedwith anti-T6 and anti-FctX antisera (Fig. 5). In the cell wallextract of �fctX transformed with an empty vector (mock), theintensity of HLBs as detected with anti-T6 antiserum wasdecreased and themigration pattern changed as comparedwiththose of the wild type (Fig. 5A, lane 1 versus lane 2). The samewas observed with the extract of �PSS (Fig. 5A, lane 3). Theheterodimer band was virtually undetected in the cell wallextracts as well as the concentrated culture supernatants of�PSSmutant,mock, and�tee6 transformedwith an empty vec-tor (Fig. 5, lanes 1, 3, and 5 and supplemental Figs. S5, lanes 9and 11, and S4, lane 9). Thus, the LPSSG motif of FctX is sug-gested to be essential for isopeptide bonds between FctX andT6 proteins. On the basis of data obtained thus far, we alsoassumed that FctX is located at the tip of the T6 polymer.When the canonical LPSSG motif of FctX was substituted

with the LPSTG motif, HLBs detected in the cell wall fractionrepresenting (T6)n-(FctX)1 were shifted to a slightly higherposition (Fig. 5, lane 2 versus lane 4). In the culture supernatantfractions, HLBs detected in the LPSTGmutant also had highermolecular weights (supplemental Fig. S5, lanes 10 and 12), indi-cating that T6 polymerization was more efficient in the FctXLPSTG mutant. It is most likely that the complex of T6 andFctX was formed more efficiently in the LPSTGmutant, possi-bly accelerating the polymerization of T6. When wild type ormutated fctX genes were expressed in �fctX�srtB, HLBs and

FIGURE 4. Introduction of mutation into LPSTG motif of T6 abrogated T6 polymerization but had no effect on heterodimer formation of T6 and FctX.Cell wall fractions were extracted from a tee6 deletion mutant (�tee6) transformed with an empty shuttle vector (mock; lane 1) or a shuttle vector harboring wildtype tee6 (lane 2) or mutated tee6. In mutated tee6, the LPSTG motif of T6 was replaced with either LG (�PST; lane 3) or LPSSG (lane 4). These extracts weresubjected to immunoblotting with mouse antisera against T6 (A) and FctX (B). The fraction of the fctX deletion mutant transformed with an empty shuttle vector(�fctX; lane 5) served as a control. The position of the heterodimer composed of T6 and FctX is denoted by an arrowhead. The deduced positions of themonomeric and complex forms of T6 and FctX are indicated on the right. The molecular mass sizes of the protein marker (M) are indicated on the left.




ww

.jbc.org/D

ownloaded from












http://www.jbc.org/

the heterodimer band completely disappeared in both the cellwall and culture supernatant fractions (supplemental Fig. S5),which again suggests a principal role of SrtB in pilus assembly.Lysine Residue 175 of T6 Is Essential for T6 Polymerization—

To identify the lysine residues of T6 that play a critical role inpilus assembly, the amino acid sequences of T6 and its homo-logues were aligned (supplemental Fig. S6). Selected homo-logues included the recently identified Streptococcus suis pilussubunit (39), T6 counterparts in serotypeM23 andM24 S. pyo-genes strains, and homologous proteins present in Clostridiumperfringens, Ruminococcus obeum, and Eubacterium ventrio-sum. Twenty-two Lys residues, which are well conservedamong homologues, especially in streptococci, and locatedbetween the putative signal peptidase cleavage site and LPSTGmotif, were selected for point mutation analysis, i.e. Lys-35,Lys-42, Lys-56, Lys-103, Lys-175, Lys-183, Lys-194, Lys-195,Lys-216, Lys-221, Lys-268, Lys-321, Lys-331, Lys-341, Lys-345,Lys-356, Lys-404, Lys-407, Lys-410, Lys-476, Lys-494, and Lys-512. We replaced each lysine residue with alanine andexpressed either wild type tee6 or a series of mutated tee6 on abackground of �tee6 (wild type, and mutants referred to asK35A for example).Cell wall extracts of each mutant strain were subjected to

Western blot analysis. HLBs detected with anti-T6 antiserumwere visible for all point mutated strains except for K175A (Fig.6A, lane 3, and supplemental Fig. S7, left panel, lane 7). Becausethe monomeric form of K175A with a molecular mass of 52kDa was readily detected, the mutant protein was preciselyexpressed, and disappearance of HLBs was not due to protein

FIGURE 6. Conserved lysine residue of T6 required for T6 polymerizationand heterodimer formation of T6 and FctX. Cell wall fractions wereextracted from the tee6 deletion mutant (�tee6) transformed with an emptyshuttle vector (mock; lane 1) or with a shuttle vector harboring wild type tee6(lane 2) or mutated tee6 (K175A, K183A, and K356A; lanes 3, 4, and 5, respec-tively) in which Lys residues Lys-175, Lys-183, and Lys-356 were pointmutated with alanine (Ala) residue. The fraction of an fctX deletion mutanttransformed with an empty shuttle vector (�fctX; lane 6) served as a control.These extracts were subjected to immunoblotting with mouse antiseraagainst T6 (A) and FctX (B). The deduced positions of the monomer and com-plex forms of T6 and FctX are indicated on the right. The sizes of the molecularmass standards are indicated on the left. M, protein size marker.

FIGURE 5. Introduction of mutation into LPSSG motif of FctX abrogated heterodimer formation of T6 and FctX. Cell wall fractions were extracted from anfctX deletion mutant (�fctX) transformed with an empty shuttle vector (mock; lane 1) or a shuttle vector harboring wild type fctX (lane 2) or mutated fctX. Inmutated fctX, the LPSSG motif of FctX was replaced with either LG (�PSS; lane 3) or LPSTG (lane 4). These extracts were subjected to immunoblotting withmouse antisera against T6 (A) and FctX (B). The fraction of the tee6 deletion mutant transformed with an empty shuttle vector (�tee6; lane 5) served as a control.The position of the heterodimer composed of T6 and FctX is denoted by an arrowhead. The deduced positions of the monomer and complex forms of T6 andFctX are indicated on the right. The sizes of the molecular mass standards are indicated on the left. M, protein size marker.




ww

.jbc.org/D

ownloaded from




http://www.jbc.org/

instability. Moreover, HLBs were not detected in the concen-trated culture supernatant of the K175A mutant (data notshown). Accordingly, Lys-175 in T6 is suggested to participatein the formation of intramolecular isopeptide bonds betweenT6 pilin subunits. Among the streptococcal homologues of T6,the VAKS amino acid sequence, including Lys-175, is con-served (supplemental Fig. S6); thus, the sequence VAKS wasdesignated as the FCT type 1 pilin motif.It is noteworthy that both the monomeric and polymeric

forms of K356A migrated slower than those of wild type T6(Fig. 6, A and B, lane 5, and supplemental Fig. S7, right panel,lane 6). Because lysine residues are also involved in intramolec-ular isopeptide bonds in many pilus proteins (15), it is quitepossible that T6 also contains intramolecular bonds and thatLys-356 participates in bond formation.Involvement of Lys-175 in Formation of T6 and FctX

Complex—To identify the lysine residue of T6 crucial for incor-poration of FctX, the formation of the T6 and FctX complexwas examined by immunoblotting using cell wall extracts of thewild type; a series of T6 mutant strains, including K175A,K183A, and K356A; and empty vector-transformed �tee6(mock). The (T6)1-(FctX)1 heterodimer band with a molecularmass of 130 kDa was detected in the cell wall extracts of thewild type and T6 mutants K183A and K356A (Fig. 6, lanes 2, 4,and 5). However, the heterodimer band was not detected in themock or K175A mutant (Fig. 6, lane 3). Thus, Lys-175 of T6 isalso suggested to be responsible for cross-linking between T6and FctX proteins.Localization of FctX in T6 Pilus Shaft—To visualize T6 pili

and the position of FctX in T6 pili, immunogold electronmicroscopy analysis was performed using whole bacteria. First,we labeled T6 inWTwith anti-T6mouse polyclonal antiserumand a secondary anti-mouse IgG conjugated with 15-nm gold

particles. �tee6 was used as a negative control. As expected,WTproduced abundant T6 pili, which protruded circumferen-tially from the cell surface (Fig. 7A). In contrast, no gold-labeledpili were observed in �tee6 (Fig. 7B).Next, double labeling of T6 and FctX in WT was conducted

to examine the location of FctX. T6 was labeled with 15-nmgold particles in the same manner described above, whereasFctX was detected using anti-FctX rabbit polyclonal antiserumand a secondary anti-rabbit IgG conjugated with 20-nm goldparticles. As a negative control of FctX labeling after T6 label-ing, non-immune rabbit serum and an anti-rabbit IgG conju-gated with 20-nm gold particles were utilized. With this nega-tive control, nonspecific 20-nm gold particles were sparselydetected only on the cell surface (Fig. 7C). In contrast, althoughthe majority of FctX labeled with larger gold particles wasdetected on the cell surface, labeled FctX was occasionallyobserved on the tips of T6 shafts labeled with smaller 15-nmgold particles (Fig. 7, D, E, and F), indicating the localization ofFctX at the tip of T6 shafts, which is implied by previous immu-noblot data. Based on our findings, we propose an assemblymodel of FCT type 1 pili (Fig. 8 and see “Discussion”).

DISCUSSION

Among the various adhesins thatmediate firm attachment tohost human tissues and tropism for specific infection sites by S.pyogenes, the pilus has become the center of attention in explo-rations of the initial step of infection (1, 40). The variability of S.pyogenes pili on both the gene and protein levels is outstandingamong streptococcal species, which is reflected by the fact thatthe antigenicity of pilus proteins is one of the determinants forT serotyping (20, 23, 24, 41). To reveal the involvement of S.pyogenes pili in the infection mechanism and tissue tropism,detailed analyses of functions, assembly mechanisms, and the

FIGURE 7. Immunogold labeling of T6 and FctX. A and B, a wild type strain (A) and tee6 deletion mutant (B) were incubated with anti-T6 antiserum followedby incubation with 15-nm gold-conjugated anti-mouse IgG antibody. Bar, 0.5 �m. C, T6 of the wild type strain was labeled with 15-nm gold particles as notedabove followed by incubation with non-immune rabbit serum and a 20-nm gold-conjugated anti-rabbit IgG. Bar, 0.5 �m. D, E, and F, T6 of the wild type strainwas labeled with 15-nm gold particles as noted above. Subsequently, FctX was labeled with rabbit anti-FctX antiserum and 20-nm gold-conjugated anti-rabbitIgG. FctX observed at the tip of T6 pilus shaft is shown by arrowheads. Bar, 0.5 �m.




ww

.jbc.org/D

ownloaded from




http://www.jbc.org/

expressionmode are crucial. Althoughmolecular characteriza-tions of FCT type 2 and 3 pili have been reported, characteriza-tions of other types of pili remain elusive (23–25). To beginmolecular characterization of thus far uncharacterized pili, wefocused on the assembly mechanism of FCT type 1 pili in aserotype M6 strain.For assembly of pili in Gram-positive bacteria, pilus subunits

are covalently linked to each other by the action of a transpep-tidase enzyme called sortase. In a serotype M6 strain, two sor-tase enzymes are encoded in the genome, i.e. the housekeepingsortase SrtA and pilus region-associated sortase SrtB (8). Thepresent findings demonstrated that SrtB is required for thisassembly, and SrtA can partially compensate in the absence ofSrtB. Therefore, it ismost likely that both recognize LPSTGandLPSSG motifs buried in the CWSSs of T6 and FctX, respec-tively. Based on ambiguous data obtained with swapping of themotif, theremay be a sequence preference for each sortase, andwe speculate that SrtA prefers the canonical LPSTG sequenceas compared with the LPSSG sequence. Because the assemblyof FCT type 1 pili was scarcely influenced by deletion of the srtAgene, it is reasonable to speculate that SrtB primarily mediatesheterodimer formation and subsequent T6 polymerization. Onthe other hand, deletion of the srtA gene caused a greaterrelease of pili into the culture supernatant; thus, SrtA mightfunction mainly for cell wall anchoring of pili as has beenreported for many different types of pili (42–44). Therefore,designation of the complex or monomer pilus proteins towardthe amino group of either cell wall or pilus proteins would bedetermined by the availability of SrtA and SrtB, which is likely

regulated by the expression mode of the srtA gene and fctXoperon.An unexpected finding in this study is that SrtAmay have an

ability to generate the complex and polymerize T6 in certainconditions. Immunoblot data for the cell wall fraction extractedfrom the srtB deletion mutant showed that all detected HLBsrepresented the (FctX)1-linked T6 polymer based onmolecularmass size, i.e. (T6)n-(FctX)1, which suggests that T6 polymer-izations without the linkage of FctX do not occur solely by theaction of SrtA. In other words, SrtA has the ability to polymer-ize T6 when FctX is covalently linked to T6.Moreover, becausenoHLBswere detectedwhen pilus geneswere expressed ectop-ically in the background of srtB deletion, the assembly by SrtAmight require temporospatial regulation of the expression ofpilus genes. Because the ability of the housekeeping sortase topolymerize pilus major subunits has not been reported, theunique interaction between SrtA and the (T6)1-(FctX)1 com-plex is expected to change the present understanding of its con-ventional functions.Another unexpected finding is that deletion of fctX caused a

decreased intensity of HBLs in immunoblotting, which indi-cates that FctX is required for proper polymerization of T6. Thesame finding was recently reported in a study of S. suis pili (39).Although there is no experimental data to explain the phenom-enon, it can be speculated that the heterodimer formation ofFctX and T6 accelerates polymerization of T6. In FCT type2/3pili, the signal peptidase I homologue SipA/LepA encoded inthe pilus region may function as a chaperone (24, 45). Such achaperone-like function of FctX may serve as a trigger factor

FIGURE 8. Proposed model of assembly of FCT type1 pili based on present and previous findings. Pilus-associated sortase SrtB cleaves the LPSSG motif ofFctX between serine and glycine residues (A); this is accompanied by formation of an acyl enzyme intermediate. SrtB also forms an intermediate with T6 (A).FctX is further transferred to an �-amino group of a lysine residue of T6 (Lys-175) by nucleophilic attack (B). The complex of FctX and T6 is then transferred toLys-175 of T6 (C). Consequently, T6 polymerizes, and FctX becomes localized at the tip of the T6 pilus shaft (D). For cell wall anchoring of pili, the housekeepingsortase SrtA cleaves the LPSTG motif of T6 between the threonine and glycine residues (E) and forms an acyl enzyme intermediate (F). Finally, the pili arecovalently linked to a free amino group of the peptidoglycan layer (G).




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

activity in serotypeM6 strains that encode no SipA/LepAcoun-terparts in the FCT type 1 pilus region.Our findings revealed that a lysine residue (Lys-175) and the

LPSTGmotif of T6 are simultaneously necessary to polymerizeT6. Lysine residues inmajor subunits of pili from other species,including SpaA of Corynebacterium diphtheriae, FctA of FCTtype 2 pili, RrgB of Streptococcus pneumoniae, and BcpA ofBacillus cereus, are known to function as nucleophiles topolymerizemajor subunits (11, 13, 14, 46). Because a conservedpilinmotif, YPKN, reported inC. diphtheriae andB. cereus (44),was not present in T6, we searched the T6-specific pilin motif,including the lysine residue, using point mutation analysis andsequence alignment of T6 and its homologues identified by aBlast search. Surprisingly, theVAKSmotif designated FCT type1 pilinmotif waswell conserved amongT6 and its streptococcalhomologues, including confirmed or inferredmajor subunits ofS. suis and Streptococcus equi subspecies zooepidemicus (datanot shown). Thus, the lysine residue corresponding to Lys-175of T6 present in the VAKS motif could be involved in thepolymerization of theirmajor subunits. Because homologues ofT6 andFctX are found in zoonotic streptococcal pathogens, it isquite possible that the FCT type 1 pilus region was horizontallytransferred among streptococcal species, which is also sup-ported by the fact that the gene upstream of fctX encodes atransposon remnant (19).We also revealed that Lys-175 of T6 participates not only in

T6 polymerization but also in the linkage of T6 and FctX. Intro-duction of the mutation into the LPSSG motif in the CWSS ofFctX hampered the linkage; thus, it is most likely that sortasescleave the motif between serine and glycine residues, and thenthe carboxyl group of the serine residue is covalently linked toLys-175 of T6, thereby forming the complex of T6 and FctX. Asimilar case has been reported regarding the interactionbetween Cpa and FctA from FCT type 3 pili in a serotype M3strain (25). It is possible that the mechanism in which the samelysine residue linksmajor subunits andmajor/minor subunits iscommon in Streptococcus species.

Intramolecular isopeptide bonds formed between lysine andasparagine residues have also been revealed by recent crystal-lographic andmass spectral analyses of several major pilus sub-units, including FctA in the FCT type 2 pili, BcpA in B. cereus,SpaA in C. diphtheriae, and RrgB in S. pneumoniae (10, 13, 14,46). The position of intramolecular isopeptide bonds differsamong those pilus proteins. In our point mutation study,immunoblotting results showed that the migration pattern ofK356A was considerably altered. Such a change was alsoreported in another study in which a mutation was introducedinto the lysine residue responsible for the intramolecular bondsof FctA (25). Therefore, Lys-356 of T6 could be involved inpotential intramolecular isopeptide bonds. Although themuta-tion did not affect polymerization, it is possible that thededuced isopeptide bond formed via Lys-356 contributes tothermostability and resistance against unfolding stress and pro-teolysis as reported for FctA (15).According to a previous report by Barnett and Scott (8), cell

wall anchoring of T6 is exclusively mediated by SrtB, whereasknock-out of SrtA has no effect on that anchoring in a serotypeM6 laboratory strain. However, the present study revealed that

the anchoring of T6 is mainly mediated by the action of SrtA.This discrepancy may be attributable to differences in experi-mental design, including mutation strategy (insertionalmutagenesis versus in-frame deletion), the antisera utilizedagainst T6 (typing serum versus T6- and FctX-specific), anddetection methods (whole cell dot blot assay versus immuno-blot analysis of cell wall fractions).Based on the present findings and the currently accepted

model of sortase function, we propose a three-step model ofsurface display of FCT type 1 pili (Fig. 8). First, FctX and T6form a complex. Next, T6 monomers polymerize to constitutethe pilus shaft and allow FctX to be localized at the tip of theshaft. Finally, the assembled pilus is covalently anchored to thecell wall. Synthesized T6 and FctX in the cytoplasm are guidedto the cell membrane via a Sec-dependent secretion pathway.Then, the LPSSGmotif in CWSS of FctX and the LPSTGmotifin CWSS of T6 are recognized and processed by SrtB, and acylenzyme intermediates are formed. Subsequently, FctX is trans-ferred to an �-amino group of Lys-175 of T6 by nucleophilicattack, thereby forming the complex of monomeric T6 andFctX. The T6-FctX heterodimer is then transferred to an �-amino group of the Lys-175 of neighboring T6. Repeatednucleophilic attack between the T6-FctX complex and the sidechain of monomeric T6 results in polymerization of T6 andpositioning of FctX at the tip of the T6 shaft (Fig. 8).Because deletion of srtB did not abolishT6 polymerization or

the complex formation of T6 and FctX, these processes mayalso occur at a low frequency by individual actions of SrtA thatare independent of SrtB. Of note, our immunoblot data showedthat T6 also polymerized without incorporation of FctX. Forcell wall anchoring of FCT type 1 pili, SrtA cleaves the LPSTGmotif of T6 at the base of the pilus structure and then forms anacyl enzyme intermediate. Finally, the assembled pilus is trans-ferred to a free amino group of the cell wall (Fig. 8). For theanchoring step, SrtA rather than SrtB plays a central role. Thus,FCT type 1 pili are elaborately assembled and anchored to thecell wall by the concerted actions of SrtA and SrtB.In summary, we revealed differential roles of SrtA and SrtB in

the assembly and cell wall anchoring of FCT type 1 pili.We alsofound that the lysine residue in the VAKS pilin motif of T6 isinvolved in intramolecular isopeptide bonds of T6 and forma-tion of the T6 and FctX complex. In addition, FctX was shownto be located at the tip of the T6 pilus shaft. Although the bio-logical importance of FCT type 1 pili remains elusive, pilus pro-teins positioned away from the cell surface and capsule mightfunction for the first interaction with the human host. Becausehomologous pili are potentially produced by zoonotic strepto-coccal pathogens, future investigation of the structure andfunctions of FCT type 1 pili will provide insights into the uni-versal interactions between streptococcal pili and the host.

Acknowledgments—We gratefully acknowledge T. Sekizaki (Univer-sity of Tokyo) and D. Takamatsu (National Institute of AnimalHealth) for providing the pSET4s plasmid. pAT18 was kindly pro-vided by P. Trieu-Cuot (Institut Pasteur). We also thank J. Yamauchifor helpful technical assistance.




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

REFERENCES1. Kreikemeyer, B., Gámez, G., Margarit, I., Giard, J. C., Hammerschmidt, S.,

Hartke, A., and Podbielski, A. (2011) Int. J. Med. Microbiol. 301, 240–2512. Telford, J. L., Barocchi, M. A., Margarit, I., Rappuoli, R., and Grandi, G.

(2006) Nat. Rev. Microbiol. 4, 509–5193. Proft, T., and Baker, E. N. (2009) Cell. Mol. Life. Sci. 66, 613–6354. Scott, J. R., and Zähner, D. (2006)Mol. Microbiol. 62, 320–3305. Schneewind, O., Model, P., and Fischetti, V. A. (1992) Cell 70, 267–2816. Mandlik, A., Swierczynski, A., Das, A., and Ton-That, H. (2008) Trends

Microbiol. 16, 33–407. Marraffini, L. A., Dedent, A. C., and Schneewind, O. (2006) Microbiol.

Mol. Biol. Rev. 70, 192–2218. Barnett, T. C., and Scott, J. R. (2002) J. Bacteriol. 184, 2181–21919. Hendrickx, A. P., Budzik, J. M., Oh, S. Y., and Schneewind, O. (2011)Nat.

Rev. Microbiol. 9, 166–17610. Budzik, J. M., Poor, C. B., Faull, K. F., Whitelegge, J. P., He, C., and Sch-

neewind, O. (2009) Proc. Natl. Acad. Sci. U.S.A. 106, 19992–1999711. Budzik, J. M., Marraffini, L. A., Souda, P.,Whitelegge, J. P., Faull, K. F., and

Schneewind, O. (2008) Proc. Natl. Acad. Sci. U.S.A. 105, 10215–1022012. Ton-That, H., and Schneewind, O. (2003)Mol. Microbiol. 50, 1429–143813. Kang, H. J., Coulibaly, F., Clow, F., Proft, T., and Baker, E. N. (2007) Science

318, 1625–162814. Kang, H. J., Paterson, N. G., Gaspar, A. H., Ton-That, H., and Baker, E. N.

(2009) Proc. Natl. Acad. Sci. U.S.A. 106, 16967–1697115. Kang, H. J., Middleditch,M., Proft, T., and Baker, E. N. (2009)Biopolymers

91, 1126–113416. Cunningham, M. W. (2000) Clin. Microbiol. Rev. 13, 470–51117. Abbot, E. L., Smith W. D., Siou, G. P., Chiriboga, C., Smith, R. J., Wilson,

J. A., Hirst, B. H., and Kehoe, M. A. (2007) Cell. Microbiol. 9, 1822–183318. Manetti, A. G., Zingaretti, C., Falugi, F., Capo, S., Bombaci,M., Bagnoli, F.,

Gambellini, G., Bensi, G., Mora, M., Edwards, A. M., Musser, J. M., Gra-viss, E. A., Telford, J. L., Grandi, G., andMargarit, I. (2007)Mol.Microbiol.64, 968–983

19. Bessen, D. E., and Kalia, A. (2002) Infect. Immun. 70, 1159–116720. Falugi, F., Zingaretti, C., Pinto, V., Mariani, M., Amodeo, L., Manetti,

A.G., Capo, S.,Musser, J.M.,Orefici, G.,Margarit, I., Telford, J. L., Grandi,G., and Mora, M. (2008) J. Infect. Dis. 198, 1834–1841

21. Kratovac, Z., Manoharan, A., Luo, F., Lizano, S., and Bessen, D. E. (2007) J.Bacteriol. 189, 1299–1310

22. Schneewind, O., Jones, K. F., and Fischetti, V. A. (1990) J. Bacteriol. 172,3310–3317

23. Mora, M., Bensi, G., Capo, S., Falugi, F., Zingaretti, C., Manetti, A. G.,Maggi, T., Taddei, A. R., Grandi, G., and Telford, J. L. (2005) Proc. Natl.Acad. Sci. U.S.A. 102, 15641–15646

24. Nakata, M., Köller, T., Moritz, K., Ribardo, D., Jonas, L., McIver, K. S.,Sumitomo, T., Terao, Y., Kawabata, S., Podbielski, A., and Kreikemeyer, B.

(2009) Infect. Immun. 77, 32–4425. Quigley, B. R., Zähner, D., Hatkoff, M., Thanassi, D. G., and Scott, J. R.

(2009)Mol. Microbiol. 72, 1379–139426. Murakami, J., Kawabata, S., Terao, Y., Kikuchi, K., Totsuka, K., Tamaru,

A., Katsukawa, C., Moriya, K., Nakagawa, I., Morisaki, I., and Hamada, S.(2002) Epidemiol. Infect. 128, 397–404

27. Podbielski, A., Spellerberg, B., Woischnik, M., Pohl, B., and Lütticken, R.(1996) Gene 177, 137–147

28. Takamatsu, D., Osaki, M., and Sekizaki, T. (2001) Plasmid 46, 140–14829. Trieu-Cuot, P., Carlier, C., Poyart-Salmeron, C., and Courvalin, P. (1991)

Gene 102, 99–10430. Inoue, H., Nojima, H., and Okayama, H. (1990) Gene 96, 23–2831. Caparon, M. G., and Scott, J. R. (1991)Methods Enzymol. 204, 556–58632. Banks, D. J., Porcella, S. F., Barbian, K. D., Beres, S. B., Philips, L. E., Voyich,

J. M., DeLeo, F. R., Martin, J. M., Somerville, G. A., and Musser, J. M.(2004) J. Infect. Dis. 190, 727–738

33. Okahashi, N., Nakata, M., Sakurai, A., Terao, Y., Hoshino, T., Yamaguchi,M., Isoda, R., Sumitomo, T., Nakano, K., Kawabata, S., and Ooshima, T.(2010) Biochem. Biophys. Res. Commun. 391, 1192–1196

34. Sumitomo, T., Nakata, M., Higashino, M., Jin, Y., Terao, Y., Fujinaga, Y.,and Kawabata, S. (2011) J. Biol. Chem. 286, 2750–2761

35. Kreikemeyer, B., Nakata, M., Oehmcke, S., Gschwendtner, C., Normann,J., and Podbielski, A. (2005) J. Biol. Chem. 280, 33228–33239

36. Nakata,M., Podbielski, A., and Kreikemeyer, B. (2005)Mol.Microbiol. 57,786–803

37. Podbielski, A., Woischnik, M., Leonard, B. A., and Schmidt, K. H. (1999)Mol. Microbiol. 31, 1051–1064

38. Okahashi, N, Nakata, M., Terao, Y., Isoda, R., Sakurai, A., Sumitomo, T.,Yamaguchi, M., Kimura, R. K., Oiki, E., Kawabata, S., and Ooshima, T.(2011)Microb. Pathog. 50, 148–154

39. Okura,M.,Osaki,M., Fittipaldi, N., Gottschalk,M., Sekizaki, T., andTaka-matsu, D. (2011) J. Bacteriol. 193, 822–831

40. Bessen D. E., and Lizano, S. (2010) Future Microbiol. 5, 623–63841. Griffith, F. (1934) J. Hygiene 34, 542–58442. Budzik, J. M.,Marraffini, L. A., and Schneewind, O. (2007)Mol.Microbiol.

66, 495–51043. Nobbs, A. H., Rosini, R., Rinaudo, C. D., Maione, D., Grandi, G., and

Telford, J. L. (2008) Infect. Immun. 76, 3550–356044. Swaminathan, A., Mandlik, A., Swierczynski, A., Gaspar, A., Das, A., and

Ton-That, H. (2007)Mol. Microbiol. 66, 961–97445. Zähner, D., and Scott, J. R. (2008) J. Bacteriol. 190, 527–53546. Spraggon, G., Koesema, E., Scarselli, M.,Malito, E., Biagini,M., Norais, N.,

Emolo, C., Barocchi, M. A., Giusti, F., Hilleringmann, M., Rappuoli, R.,Lesley, S., Covacci, A., Masignani, V., and Ferlenghi, I. (2010) PLoS One 5,e10919




ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

Okahashi, Shigeyuki Hamada, Ryutaro Isoda, Yutaka Terao and Shigetada KawabataNobuoSugauchi, Eiji Oiki, Miharu Higashino, Bernd Kreikemeyer, Andreas Podbielski,

Masanobu Nakata, Keiji Richard Kimura, Tomoko Sumitomo, Satoshi Wada, Akinaripyogenes

StreptococcusAssembly Mechanism of FCT Region Type 1 Pili in Serotype M6

doi: 10.1074/jbc.M111.239780 originally published online August 31, 20112011, 286:37566-37577.J. Biol. Chem.

10.1074/jbc.M111.239780Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

Supplemental material:

http://www.jbc.org/content/suppl/2011/08/31/M111.239780.DC1

http://www.jbc.org/content/286/43/37566.full.html#ref-list-1

This article cites 46 references, 17 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M111.239780

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;286/43/37566&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/286/43/37566

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=286/43/37566&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/286/43/37566

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/suppl/2011/08/31/M111.239780.DC1

http://www.jbc.org/content/286/43/37566.full.html#ref-list-1

http://www.jbc.org/

assemblymechanismoffctregiontype1piliinserotype … · plasmid was integrated into the chromosome...

Documents