INFECTION AND IMMUNITY, Dec. 2008, p. 5624–5631 Vol. 76, No. 120019-9567/08/$08.00�0 doi:10.1128/IAI.00534-08Copyright © 2008, American Society for Microbiology. All Rights Reserved.

A Conserved C-Terminal 13-Amino-Acid Motif of Gap1 Is Requiredfor Gap1 Function and Necessary for the Biogenesis of a

Serine-Rich Glycoprotein of Streptococcus parasanguinis�

Meixian Zhou,1 Zhixiang Peng,1 Paula Fives-Taylor,2 and Hui Wu1*Departments of Pediatric Dentistry and Microbiology, University of Alabama at Birmingham Schools of Medicine and Dentistry,

Birmingham, Alabama 35244,1 and Department of Microbiology and Molecular Genetics, College of Medicine andCollege of Life Science and Agriculture, Burlington, Vermont 054052

Received 30 April 2008/Returned for modification 17 May 2008/Accepted 29 September 2008

Adhesion of Streptococcus parasanguinis to saliva-coated hydroxyapatite (SHA), an in vitro tooth model, ismediated by long peritrichous fimbriae. Fap1, a fimbria-associated serine-rich glycoprotein, is required forfimbrial assembly. Biogenesis of Fap1 is controlled by an 11-gene cluster that contains gly, nss, galT1 and -2,secY2, gap1 to -3, secA2, and gtf1 and -2. We had previously isolated a collection of nine nonadherent mutantsusing random chemical mutagenesis approaches. These mutants fail to adhere to the in vitro tooth model andto form fimbriae. In this report, we further characterized these randomly selected nonadherent mutants andclassified them into three distinct groups. Two groups of genes were previously implicated in Fap1 biogenesis.One group has a mutation in a glycosyltransferase gene, gtf1, that is essential for the first step of Fap1glycosylation, whereas the other group has defects in the fap1 structural gene. The third group mutantproduces an incompletely glycosylated Fap1 and exhibits a mutant phenotype similar to that of a glycosylation-associated protein 1 (Gap1) mutant. Analysis of this new mutant revealed that a conserved C-terminal13-amino-acid motif was missing in Gap1. Site-directed mutagenesis of a highly conserved amino acid tryp-tophan within this motif recapitulated the deletion phenotype, demonstrating the importance of the Gap1C-terminal motif for Fap1 biogenesis. Furthermore, the C-terminal mutation does not affect Gap1-Gap3protein-protein interaction, which has been shown to mediate Fap1 glycosylation, suggesting the C-terminalmotif has a distinct function related to Fap1 biogenesis.

Streptococcus parasanguinis is an early colonizer of the den-tal plaque and has been associated with the pathogenesis ofinfective endocarditis(5, 14, 17). Bacterial adhesion to hosts isimportant in these physiological and pathophysiological pro-cesses (16, 18). Therefore, investigation of molecular mecha-nisms of bacterial adhesion will facilitate our understandingthe biology of streptococcal fitness and pathogenicity.

S. parasanguinis possesses long peritrichous fimbriae. Ma-ture Fap1, a 200-kDa glycoprotein, is a major fimbrial subunitand is responsible for bacterial adhesion in an in vitro toothmodel, saliva-coated hydroxyapatite (SHA) (29). Deficiency inan accessory Sec protein, SecA2, blocks Fap1 export and alsoinhibits secretion of FimA (7), a key virulence factor of S.parasanguinis-induced infective endocarditis (4). Fap1 ho-mologs have been isolated as important adhesins in othergram-positive bacteria, including Streptococcus gordonii (2),Streptococcus sanguinis (20), Streptococcus pneumoniae (26),Streptococcus agalactiae (22), Staphylococcus aureus (23), andStaphylococcus epidermidis (31). Fap1-like proteins in S. gor-donii and Staphylococcus aureus have been implicated in thepathogenesis of infective endocarditis (23, 25). Fap1, a glyco-sylated cell-wall-anchored adhesin, contains a long signal se-quence at the amino terminus which directs Fap1 export in

conjunction with its C-terminal classic cell wall sorting signal.Fap1 also possesses two extensive serine-rich repeat regionsthat appear to be modified by O-linked glycan moieties (28,29). Biogenesis of Fap1 is mediated by a cluster of 11 genes(gly, nss, galT1 and -2, secY2, gap1 to -3, secA2, gtf1, and gtf2).The functions of many of these genes have been determined(27, 30). For instance, inactivation of the Gtfs renders themutants unable to glycosylate Fap1 and leads to the productionof an unglycosylated high-molecular-mass Fap1 precursor A(3, 27). SecY2 and Gap3 mutants fail to express mature Fap1and instead produce an incompletely glycosylated and distincthigh-molecular-mass Fap1 precursor B (19, 27).

In the process of identifying adhesins of S. parasanguinis thatmediate bacterial interactions with salivary components, wehad previously isolated a series of chemically mutagenizedstrains which do not attach to SHA and cannot form fimbriae(11, 13). These nonadherent mutants were isolated based ontheir inability to interact with SHA (13). Among these mu-tants, VT321 has been partially characterized and has alreadybeen linked to the Fap1 biogenesis defect (12), albeit thenature of mutation(s) has not been studied. However, it isunknown whether other nonadherent mutants also have de-fects in Fap1 and Fap1 biogenesis. Further characterization ofthese mutants will facilitate our understanding of Fap1 biogen-esis and Fap1-dependent and -independent bacterial adhesionmechanisms.

In this report, we characterized these mutants using a for-ward genetic complementation strategy, determined that a newmutant group VT324 was defective in glycosylation-associated

* Corresponding author. Mailing address: Department of PediatricDentistry, University of Alabama at Birmingham School of Dentistry,Birmingham, AL 35244. Phone: (205) 996-2392. Fax: (205) 975-6251.E-mail: hwu@uab.edu.

� Published ahead of print on 13 October 2008.

5624

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

protein Gap1, and show for the first time that a conservedC-terminal 13-amino-acid motif of Gap1 is required for Gap1function and is necessary for Fap1 maturation.

MATERIALS AND METHODS

Bacterial strains, plasmids, and growth conditions. The bacterial strains andplasmids used in this study are listed in Table 1. S. parasanguinis strains weregrown statically in 5% CO2 in Todd-Hewitt (TH) medium or on TH agar platesat 37°C. Erythromycin (10 �g/ml) and kanamycin (125 �g/ml) were used to select

the designed streptococcal transformants. Escherichia coli cells were cultured inLuria-Bertani (LB) medium or on LB agar plates at 37°C. Erythromycin (300�g/ml) and kanamycin (25 �g/ml) were used for E. coli transformants.

DNA manipulation. The genomic DNA of S. parasanguinis was extracted usingthe Puregene DNA isolation kit (Gentra System). Plasmid DNA was isolatedusing the QIAgen spin miniprep kit (Qiagen, Inc., Santa Clarita, CA). Theprimers used in this study are listed in Table 2. PCR amplifications were per-formed using Taq DNA polymerase (Promega) or KOD hot start DNA poly-merase (Novagen) according to the suppliers’ instructions. DNA restrictionenzyme digestion, ligation, and transformation were performed using standardmethods (21).

Expression of Fap1 in nonadherent mutants. To examine the Fap1 expressionin nonadherent mutants, 1 ml of mutant cells grown to exponential phase (opticaldensity at 470 nm of �0.6) were harvested and lysed in 40 �l lysis buffer (1:10dilution of LambdaSa2 lysin [GenBank no. AE014275] in phosphate-bufferedsaline buffer [0.1 M NaCl, 0.01 M NaH2PO4, pH 7.4]) for 10 min at roomtemperature before adding 5� sodium dodecyl sulfate (SDS) loading buffer (250mM Tris-HCl [pH 6.8], 10% SDS, 0.5% bromophenol blue, 50% glycerol; 500mM �-mercaptoethanol). The extracted cell lysates were subjected to electro-phoresis on 4 to 12% precast gradient gels (Cambrex) and Western blottinganalysis. Culture supernatants of nonadherent mutants were prepared as de-scribed previously (27). In brief, proteins precipitated from 0.5 ml of culturemedia of each nonadherent mutant were dissolved in 30 �l 1� SDS loadingbuffers. Fifteen microliters of each of the prepared samples was subjected toWestern blotting analysis using Fap1-specific antibodies. Expression of Fap1 inwild-type strain FW213 and expression in the fap1 mutant were used as a positiveand negative control, respectively. The defined secY2, gap1, gap3, secA2, gtf1, andgtf2 mutants were used as reference strains to determine relevant gene mutationsin nonadherent mutants.

Genetic complementation of nonadherent mutants analyzed by BactELISAand Western blotting. BactELISA and Western blotting analyses were per-formed to assay the genetic complementation of nonadherent mutants. Thefull-length DNA sequences coding for SecY2, Gap1, Gap3, Gtf1, and Gtf2 wereamplified from the genomic DNA of S. parasanguinis FW213 using the gene-specific primers listed in Table 2. The amplified PCR products for SecY2, Gap1,Gap3, and Gtf2 were digested with the restriction enzymes SalI and KpnI andinserted into the same restriction sites of the E. coli-Streptococcus shuttle vectorpVT1666 (8) to create corresponding recombinant plasmids. The Gtf1 PCRproduct was digested with the restriction enzyme BglII and cloned into theBamHI site of the E. coli-Streptococcus shuttle vector pVT1666 to yield pVPT-Gtf1. The resulting pVPT derivatives were transformed into related mutants.Plasmid pVA838::VT1175 (6, 28) with the full-length fap1 gene was transformedinto the fap1-defective nonadherent mutants. The chosen nonadherent mutantsand their complemented strains were analyzed by BactELISA as previouslydescribed (10). Fap1-specific antibody mAbF51 was used as a primary antibodyto detect Fap1 expression. The assays were performed in triplicate in threeindependent experiments.

Western blot analyses were used to further confirm the genetic complemen-tation of nonadherent mutants with three Fap1-specific antibodies, mAbE42,mAbD10, and mAbF51.

Adhesion of complemented strains to SHA. Adhesion of S. parasanguinisFW213 to SHA was performed as described previously (29). FW213, the fap1mutant, nonadherent mutants, and their complemented strains were labeled with[3H]thymidine and grown to an optical density at 470 nm of 0.8. One milliliter ofradiolabeled bacteria was incubated with pretreated SHA beads in scintillationvials at 37°C for 2 h. The unbound bacteria remaining in supernatants weretransferred to new scintillation vials, the beads bound with bacteria were washedthree times with 0.05 M phosphate buffer (pH 6.0), and the correspondingradioactivities were counted and used to calculate adhesion levels of the testedstrains. Fifty microliters of radiolabeled bacteria was used to calculate the label-ing efficiency of the chosen strains.

Sequence analysis. PCR products of the gtf1 gene were amplified from thegenomic DNA of strains VT343 and VT508 using the primer pair Gtf1-BglII-Fand Gtf1-BglII-R. The PCR product of the gap1 gene was amplified from thegenomic DNA of VT324 using the primer pair Gap1-SalI-F and Gap1-KpnI-R.The PCR product corresponding to the N-terminal region of Fap1 was amplifiedfrom the genomic DNA of VT325 using the primer pair Fap1-N-F and Fapl-N-R.PCR products encoding the cell wall anchor (CWA) domain of Fap1 wereamplified from the genomic DNA samples of VT321, VT361, VT377, VT379,and VT380 using the primer pair Fap1-C-F1/F2 and Fapl-C-R. The amplifiedPCR products were sequenced using same gene-specific primer pairs and ana-lyzed with ClustalW, DNAstar, and NCBI tools.

TABLE 1. Strains and plasmids used in this study

Strain or plasmid Characteristic(s) Source orreference

StrainsE. coli Top10 Host for recombinant

plasmidsInvitrogen

S. parasanguinisFW213

Wild type 9

fap1 mutant fap1::aphA3 Kanr 29secY2 mutant secY2::aphA3 Kanr 27gap1 mutant gap1::aphA3 Kanr Wu et al.,

unpublishedgap3 mutant gap3::aphA3 Kanr 19secA2 mutant secA2::Tn5 7gtf1 mutant gtf1::aphA3 Kanr 27gtf2 mutant gtf2::aphA3 Kanr 3VT321 Nonadherent chemically

mutagenized strain13

VT324 Nonadherent chemicallymutagenized strain

13

VT325 Nonadherent chemicallymutagenized strain

13

VT343 Nonadherent chemicallymutagenized strain

13

VT361 Nonadherent chemicallymutagenized strain

13

VT377 Nonadherent chemicallymutagenized strain

13

VT379 Nonadherent chemicallymutagenized strain

13

VT380 Nonadherent chemicallymutagenized strain

13

VT508 Spontaneous F51-negativenonadherent mutant

11

PlasmidspBD-Gap3 gap3 cloned in pGBKT7;

KanrWu et al.,

unpublishedpVT1666 E. coli-Streptococcus

shuttle vector; Ermr8

pVPT-SecY2 secY2 cloned in pVT1666;Ermr

This study

pVPT-Gap1 gap1 cloned in pVT1666;Ermr

This study

pVPT-Gap1(W518A) gap1(W508A) cloned inpVT1666; Ermr

This study

pVPT-Gap1(W518Y) gap1(W508Y) cloned inpVT1666; Ermr

This study

pVPT-Gap3 gap3 cloned in pVT1666;Ermr

This study

pVPT-SecA2 secA2 cloned in pVT1666;Ermr

8

pVPT-Gtf1 gtf1 cloned in pVPT1666;Ermr

This study

pVPT-Gtf2 gtf2 cloned in pVT1666;Ermr

This study

pVA838::VT1175 fap1 cloned in pVA838;Ermr

6

VOL. 76, 2008 GENES INVOLVED IN S. PARASANGUINIS Fap1 BIOGENESIS 5625

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

Gap1 C-terminal site-directed mutagenesis. Plasmid pVPT-Gap1 (Table 1)containing the full-length gap1 gene was used as a template for Gap1 C-terminalsite-directed mutagenesis to mutate a conserved amino acid, tryptophan (W),into a disfavored amino acid, alanine (A), and favored amino acid, tyrosine (Y).Site-directed mutagenesis was carried out by PCR utilizing a QuickChange XLmutagenesis kit (Stratagene, La Jolla, CA). Two primer pairs, Gap1 (W518A)-Fand Gap1 (W518A)-R and Gap1 (W518Y)-F and Gap1 (W518Y)-R (Table 2),were used to construct site-directed mutations. The resulting plasmids, pVPT-Gap1(W518A) and pVPT-Gap1(W518Y) (Table 1), were confirmed by DNAsequencing and then transformed into a Gap1-null mutant, respectively, to detectthe Fap1 expression profile. The expression of FimA in the wild-type strain, fap1mutant, VT324, Gap1 W518A mutant, and Gap1 W518Y mutant was used as aloading control.

In vitro GST pull-down assays. To examine the in vitro interaction betweenGap1, Gap1(�513–525), and Gap3, the DNA sequences of the full-length Gap1and the truncated Gap1 were amplified from the genomic DNA samples of S.parasanguinis FW213 and VT324, respectively, using primer pairs Gap1-EcoRI-Fand Gap1-BamHI-R or Gap1(�513–525)-BamHI-R (Table 2). The amplifiedPCR products were then digested with the restriction enzymes EcoRI andBamHI and inserted into the same restriction sites of pGEX-5X-1 to producefusion plasmids pGEX-5X-1::Gap1 and pGEX-5X-1::GST-Gap1(�513–525), re-spectively. The glutathione S-transferase (GST), GST-Gap1, and GST-Gap1(�513–525) fusion proteins were purified using glutathione Sepharose 4Bbeads (Amersham) according to the manufacturer’s instructions. The purifiedproteins were subjected to SDS-polyacrylamide gel electrophoresis analysis byCoommassie brilliant blue staining. Plasmid pBD-Gap3 (Table 1) was used as atemplate to in vitro translate the c-Myc-tagged Gap3 using the TnT quick-coupled transcription/translation system (Promega) following the manufacturer’sinstructions. The expression of c-Myc–Gap3 protein was detected by Westernblot analysis using anti-c-Myc antibody (1:500 dilution). For the in vitro GSTpull-down assay, 5 �g of GST, GST-Gap1, or GST-Gap1(�513–525) fusionproteins in NETN binding buffer (20 mM Tris-HCl [pH 7.2], 50 mM NaCl, 1 mMEDTA, 0.2% NP-40) was mixed with 5 �l of in vitro-translated c-Myc–Gap3separately in a final volume of 200 �l NETN binding buffer and incubated on arotary shaker at 4°C overnight. The beads were washed three times with 600 �lNETN binding buffer, and the bound proteins were eluted by boiling in the SDSsample loading buffer and subjected to Western blotting analysis with anti-c-Mycantibody (1:500 dilution).

RESULTS AND DISCUSSION

Using chemical mutagenesis schemes, we had previously iso-lated nine nonadherent mutants of S. parasanguinis (11, 13).They have been classified into different phenotypic groupsbased on their twitching motility, relative hydrophobicity, saliva-induced aggregation, and coaggregation (13). A close associa-tion between the loss of adherence ability and a decrease in cellfimbriation exists, albeit the nature of mutations is unknown.In this report we further characterized these mutants.

Phenotypes of gtf1 and fap1 nonadherent mutants. SinceFap1 is required for bacterial adhesion, we examined Fap1expression profile of the isolated mutants by Western blottinganalysis using three available Fap1-specific antibodies. mAbE42 isa peptide-specific antibody, whereas mAbD10 and mAbF51are glycan-specific antibodies recognizing different epitopes.Only mature Fap1 reacts with mAbF51 (27). Wild-type bacte-ria expressed the mature 200-kDa Fap1 that reacts with allthree monoclonal antibodies (Fig. 1A, lane 1). The first groupof mutants, VT343 and VT508, expressed a 360-kDa polypep-tide when probed with mAbE42 (Fig. 1A, top panel, lanes 3and 4). This high-molecular-mass protein did not react withglycan-specific antibodies (Fig. 1A, middle and bottom panels,lanes 3 and 4), suggesting it is not glycosylated. This Fap1expression profile is similar to that observed with the gtf1 orgtf2 mutant (Fig. 1A, lanes 5 and 6). Therefore, we trans-formed VT343 and VT508 with pVPT-Gtf1 and pVPT-Gtf2,respectively, to carry out genetic complementation experi-ments. Only pVPT-Gtf1 restored the production of matureFap1 (Fig. 1B), suggesting both mutants are defective in thegtf1 gene. The complementation was further confirmed byWestern blot analysis. Like the wild-type bacteria, the comple-

TABLE 2. Primers used in this study

Primer Characteristic(s) Sequencea

SecY2-KpnI-F Amplifies secY2 gene GATCAGGTACCTTGAGTATGTTAAGAAAACTTCSecY2-KpnI-R Amplifies secY2 gene GATCAGGTACCGTACAAGTTTTTGTATTTTTTCCAACGap1-SalI-F Amplifies gap1 gene CGTCAGTCGACATGTTTTATTTTGTACCTTCGap1-KpnI-R Amplifies gap1 gene CCGCGCGGTACCTTTCTTTTTTAGCATACCTTTCGap1-EcoRI-F Amplifies gap1 gene CGGAATTCATGTTTTATTTTGTACCTTCGap1-BamHI-R Amplifies gap1 gene CGGGATCCTTTCTTTTTTAGCATACCTTTCCGap1(�513–525)-

BamHI-RAmplifiese gap1 (�513–525) gene GATCAGGATCCGCCACTTGTGTAATCAGAC

Gap1(W518A)-F Amplifies gap1 site-directed mutation allele GTGGCAAAATTTTGAAGCAAGCGAAAGGTATGCTAAAAAAGGap1(W518A)-R Amplifies gap1 site-directed mutation allele CTTTTTTAGCATACCTTTCGCTTGCTTCAAAATTTTGCCACGap1(W518Y)-F Amplifies gap1 site-directed mutation allele CAAGTGGCAAAATTTTGAAGCAATACAAAGGTATGCTAAA

AAAGGap1(W518Y)-R Amplifies gap1 site-directed mutation allele CTTTTTTAGCATACCTTTGTATTGCTTCAAAATTTTGCCA

CTTGGap3-SalI-F Amplifies gap3 gene CGGCCGTCGACATGACTAAACAGTTAATTTCGap3-KpnI-R Amplifies gap3 gene CGCCGCGGTACCAATATATTCTATTAAATTTTTCGtf1-BglII-F Amplifies gtf1 gene GATCAAGATCTATGACAATCTATAATATTAATTTAGGtf1-BglII-R Amplifies gtf1 gene GATCAAGATCTATCATTTAACATCTCCTCGtf2-SalI-F Amplifies gtf2 gene GCAGCGTCGACATGATTAGGTTGTTTGAATGGtf2-KpnI-R Amplifies gtf2 gene CGGCCCGGTACCATCTACATTACTAACCAATACFap1-N-F Amplifies N-terminal region of Fap1 TTGGAATACGGTGCTAAAATGFap1-N-R Amplifies N-terminal region of Fap1 CTGGTTGTATCTGAAGGTATFap1-C-F1 Amplifies for CWA domain of Fap1 ATCCGTATCAGAATCAGTGFap1-C-F2 Amplifies for C-terminal region of Fap1 GTCAGTATCAGAGTCAGTAAGFap1-C-R Amplifies for CWA domain of Fap1 CATCCATGTAAATCTCTG

a Restriction enzyme sites are underlined.

5626 ZHOU ET AL. INFECT. IMMUN.

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

mented strains all produced mature 200-kDa Fap1 (Fig. 1C,lanes 1, 4, and 6). These results demonstrate that VT343 andVT508 are gtf1 mutants.

Sequence analysis of the mutated gtf1 alleles revealed aG-to-T substitution at codon 67 and 203 in VT343 and VT508,respectively (Fig. 1D). Such changes led to premature termi-nations of Gtf1 at the amino acid residues G67 in VT343 andY203 in VT508.

Since Fap1 is responsible for bacterial adhesion, we per-formed adhesion assays to characterize the complemented mu-tants. VT343 and VT508 can be partially complemented by thefull-length gtf1 gene; the adherence levels of the comple-mented VT343 and VT508 reached 70% and 65% of the wild-type level (Fig. 1E). VT508 was selected by bacterial inabilityto agglutinate Fap1-specific mAbF51; therefore, it is not sur-prising to find that the mutant has a defect in Gtf1 as Gtf1 isessential to catalyze the first step of the Fap1 glycosylation (3).However, it is interesting to note that VT343 selected by itsadhesion deficiency had the same mutation, indicating the im-portance of Gtf1 in the Fap1 biogenesis. Failure to fully com-plement the adhesion defect in spite of complete restoration ofmature Fap1 expression may indicate that the overexpressionof glycosyltransferase Gtf1 deregulates some Fap1-dependentor -independent adhesion process.

The second mutant group contains six mutants. Five ofthem, including VT321 (Fig. 2A, lane 3), VT361 (lane 5),VT377 (lane 6), VT379 (lane 7), and VT380 (lane 8), secreteda large amount of mature Fap1 into their culture media. Thissecreted protein reacted with all three Fap1-specific antibodies

(Fig. 2A, lanes 3 and 5 to 8). Notably, these mutants did notretain detectable mature Fap1 in the cell lysates (Fig. 2B, lanes3 and 5 to 8), suggesting a defect in the CWA domain of Fap1.Another mutant, VT325, did not react with any Fap1-specificantibody (Fig. 2A and B, lane 4), implying a defect in the fap1structural gene. We transformed the full-length fap1 gene intoeach mutant and then analyzed the mature Fap1 production inthe complemented strains. All six mutants were complementedby the full-length fap1 as determined by enzyme-linked immu-nosorbent assay (ELISA) analysis (data not shown), demon-strating that they have defects in fap1. Western blot analysiswith Fap1-specific antibodies confirmed that fap1-comple-mented strains expressed mature 200-kDa Fap1 in both culturemedia (Fig. 2C, lanes 4, 6, 8, 10, 12, and 14) and cell lysates(Fig. 2D, lanes 4, 6, 8, 10, 12, and 14), further supporting theconclusion that these mutants belong to fap1 mutant groups.

As the fap1 gene is large (8,000 bp) and contains extensiverepetitive sequence (28), it is difficult to reliably clone andsequence the entire gene, especially the large repetitive region;therefore, we analyzed different fap1 regions based on theFap1 expression profile. As five mutants may have defects intheir ability to anchor, we analyzed the DNA sequence fromthe anchor domain of these mutants. Examination of the am-plified anchor region of VT321 revealed two mutation sites,leucine (L) 2545 to histidine (H) and valine (V) 2548 to as-partic acid (D) (Fig. 2E). The L and V residues are located inthe CWA domain of Fap1. The CWA is highly conserved inmany gram-positive cell surface proteins and contains threemotifs, the LPxTG motif, followed by a hydrophobic region

FIG. 1. Characterization of nonadherent gtf1 mutants. (A) Fap1 expression profile of the first group of mutants. The Fap1 expression profilesof cell lysates of the wild type (WT) (lane 1), fap1 mutant (lane 2), VT343 (lane 3), VT508 (lane 4), gtf1 mutant (lane 5), and gtf2 mutant (lane6) were determined by Western blotting analysis with three Fap1-specific antibodies, mAbE42 (top panel), mAbD10 (middle panel), and mAbF51(bottom panel). (B) Genetic complementation of VT343 and VT508 with genes coding for glycosyltransferases Gtf1 and Gtf2 and analyzed byBactELISA using mAbF51. (C) Western blotting analyses of the wild type (lane 1), fap1 mutant (lane 2), VT343 (lane 3), and VT508 (lane 5) andtheir complemented strains (lanes 4 and 6) using Fap1-specific antibodies. (D) Diagrammatic depiction of gtf1 alleles from VT343 and VT508.(E) Adhesion of S. parasanguinis FW213 and its derivatives. Wild type FW213, the fap1 mutant, VT343, and VT508 and their complemented strainswere assayed for their ability to bind to SHA.

VOL. 76, 2008 GENES INVOLVED IN S. PARASANGUINIS Fap1 BIOGENESIS 5627

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

and a small charged tail (24). Fap1 possesses a typical cellanchor domain (28). Mutagenesis of two nonpolar amino ac-ids, L and V, into the basic amino acid H and the acidic aminoacid D altered the Fap1 secretion profile (Fig. 2C and D, lanes1 and 3), suggesting the hydrophobic region in the anchordomain is very important for Fap1 secretion. The mutations inthe hydrophobic region can alter the polarity of the anchordomain of Fap1, thereby inhibiting the function of the Fap1anchor domain and consequently leading to the export of Fap1into culture media. This new information can help to designsurface proteins that are readily secreted into culture media tofacilitate protein purification in gram-positive bacteria.

There was no mutation identified in the CWA domain ofFap1 from the VT361, VT377, VT379, and VT380 mutants,despite the fact that the mutant phenotypes are very similar tothat observed for VT321, suggesting that the mutation in thesefapl genes is located in the RII region that is adjacent to theCWA domain of Fap1. We attempted to amplify this regionbut failed to obtain satisfactory PCR products that are ame-nable to sequencing. This is likely due to the presence ofextensive repeated DNA sequences (28). In fact, a definedmutant constructed by inserting an erythromycin resistancecassette at the junction regions between RII and CWA dis-played a very similar phenotype (28).

FIG. 2. Characterization of nonadherent fap1 mutants. (A and B) Fap1 expression profiles of culture supernatants (A) and cell lysates (B) ofthe wild type (WT) (lane 1), fap1 mutant (lane 2), VT321 (lane 3), VT325 (lane 4), VT361 (lane 5), VT377 (lane6), VT379 (lane 7), and VT380(lane 8), as determined by Western blotting analysis with three Fap1-specific antibodies. Arrows point to the positions corresponding to the Fap1protein. (C and D) Western blotting analyses of culture media (C) and cell lysates (D) of the wild type (lane 1), fap1 mutant (lane 2), and fap1chemically mutagenized strains and their complemented counterparts VT321 (lane 3), VT321/fap1� (lane 4), VT325 (lane 5), VT325/fap1� (lane6), VT361 (lane 7), VT361/fap1� (lane 8), VT377 (lane 9), VT377/fap1� (lane 10), VT379 (lane 11), VT379/fap1� (lane 12), VT380 (lane 13), andVT380/fap1� (lane 14) with Fap1-specific peptide antibodies. (E) Diagrammatic depiction of fap1 alleles from VT321, VT325, and FW213.(F) Adhesion of S. parasanguinis FW213 and its fap1 mutant derivatives. Wild-type strain FW213, the fap1 mutant, and VT321, VT325, VT361,VT377, VT379, and VT380 and their complemented strains were evaluated for their ability to bind to SHA.

5628 ZHOU ET AL. INFECT. IMMUN.

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

As VT325 failed to produce any Fap1-reactive band, wehypothesize that it might have a mutation at the very beginningof the amino terminus. Sequencing analysis of the 5� fap1fragment revealed the presence of a mutation in the NRIIregion of fap1. A C-to-T substitution at codon 224 led topretermination of Fapl translation (Fig. 2E). The truncatedFap1 does not react with Fap1-peptide specific antibodymAbE42, indicating the E42 epitope is located outside of thisregion.

The adherence levels of VT321, VT361, VT377, VT379, andVT380 were well restored by the fap1 complementation (Fig.2F). These data suggest the identified mutations are responsi-ble for the observed adhesion phenotypes, supporting the con-cept that Fap1 is a major bacterial adhesin.

Interestingly another fap1-related mutant, VT325, behaveddifferently. The complemented strain could produce matureFap1 (Fig. 2C and D, lane 6); however, it failed to restorebacterial adhesion, suggesting this strain may have an addi-tional Fap1-independent mutation. To further characterize thestrain, we determined the labeling efficiency of each mutant.VT325 and its fap1-complemented derivative had remarkablylow efficiency to incorporate [3H]thymidine (data not shown),suggesting a global defect in bacterial DNA metabolism. Thismay explain why even the complemented VT325 cannot ad-here well. The global defects may affect the Fap1-independentbacterial adhesion mechanism as diverse adhesion molecules

are present in oral streptococci and can interact with host cellsin a concerted or independent manner.

Gap1 is required for complete glycosylation of Fap1. Thethird mutant group, VT324 (Fig. 3A, lane 3), displayed a dis-tinct Fap1 expression profile which is similar to the phenotypeobserved in secY2, gap1, gap3, and secA2 mutants (Fig. 3A,lanes 4 to 7). They all expressed a 470-kDa polypeptide thatreacted with peptide-specific antibody mAbE42 (Fig. 3A, toppanel, lanes 3 to 7) and one glycan-specific antibody, mAbD10(Fig. 3A, middle panel, lanes 3 to 7), but not with the otherglycan-specific antibody, mAbF51 (Fig. 3A, bottom panel,lanes 3 to 7). These data suggest that the Fap1 precursor lacksF51-specific glycan epitope(s) and is not completely glycosy-lated in VT324.

Since gap1, secY2, gap3, and secA2 mutants exhibited thesame phenotype as VT324, we transformed these full-lengthgenes into VT324, respectively, and determined Fap1 expres-sion. Only pVPT-Gap1 restored the mature Fap1 productionby ELISA (data not shown), suggesting that VT324 is a gap1mutant. Western blotting analysis confirmed this result. VT324can be complemented only by a full-length gap1 gene (Fig. 3B,lane 4) but not by other genes (Fig. 3B, lanes 5 to 7). Like thewild type strain, the complemented strain produced matureFap1 (Fig. 3B, lanes 1 and 4), demonstrating VT324 is a gap1mutant. The adhesion level of VT324 can be fully comple-

FIG. 3. Characterization of a nonadherent gap1 mutant. (A) The Fap1 expression profiles of cell lysates of the wild type (lane 1), fap1 mutant(lane 2), VT324 (lane 3), secY2 mutant (lane 4), gap1 mutant (lane 5), gap3 mutant (lane 6), and secA2 mutant (lane 7) were determined by Westernblotting analysis with three Fap1-specific antibodies. (B) Western blotting analyses of the wild type (lane 1), fap1 mutant (lane 2), VT324 (lane3), and VT324 complemented with the full-length gap1 (lane 4), secY2 (lane 5), gap3 (lane 6), and secA2 (lane 7) genes. (C) Adhesion of S.parasanguinis FW213 and gap1 mutant derivatives to SHA.

VOL. 76, 2008 GENES INVOLVED IN S. PARASANGUINIS Fap1 BIOGENESIS 5629

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

mented (Fig. 3C), concurrent with the production of the ma-ture Fap1 protein.

The C-terminal 13-amino-acid motif is required for Gap1function. The mutation in the gap1 gene of VT324 was deter-mined to be an A-to-T substitution at codon 513 (Fig. 4A).This mutation led to premature termination of Gap1. Loss ofthe C-terminal 13 amino acids inhibited mature Fap1 produc-tion, suggesting the C terminus is very important for the main-tenance of the Gap1 function. The 13 amino acids are con-served in other Gap1 homologs from S. agalactiae, S. gordonii,S. sanguinis, S. pneumoniae, Staphylococcus haemolyticus, orLactobacillus johnsonii (Fig. 4B).

Based on the aligned sequence in this region (Fig. 4B), wedetermined that the W518 residue is conserved across manyspecies, suggesting it is important for Gap1 function. To sup-port this, we performed site-directed mutagenesis and changedthe amino acid W into disfavored amino acid A and favoredamino acid Y and then transformed the mutant alleles into theGap1-null mutant, respectively. Western blot analysis revealedthat the W518A mutation significantly increased the expres-sion of the 470-kDa partially glycosylated protein (Fig.4CI,lane 4) that is characteristic of the Gap1 mutant (Fig.4CI, lane3), while the production of mature Fap1 in the mutant wasminimal (Fig. 4C, lane 4). However, the W518Y mutation didnot promote the production of the 470-kDa protein: the mu-tant retained the ability to produce mature Fap1 (Fig. 4C, lane5). These data further support the concept that this C-terminalregion is important for Gap1 function and necessary for Fap1glycosylation. In fact, this region is structurally aligned andshows homology to a functional domain of several known gly-cosyltransferases, including PimA of mycobacteria (15), as de-termined by the Protein Homology/analogY Recognition En-gine (PHYRE) program (1). Among these functional domains,W is predicted to be the key amino acid responsible for theglycosylation function.

The C-terminal 13-amino-acid motif is not required for theinteraction between Gap1 and Gap3. Recently we have deter-mined that Gap1 interacts with Gap3 and an N-terminal motifin Gap1 is important for the interaction and also required forbiogenesis of Fap1. The involvement of the C-terminal motif inthe Fap1 glycosylation suggests that Gap1 has a distinct func-tional domain. As the C-terminal motif is structurally alignedto a functional glycosylation domain, it may not be involved inthe interaction between Gap1 and Gap3. To test this, we con-structed and purified recombinant GST, GST-Gap1, and GST-Gap1(�513–525) proteins (Fig. 5A) and performed in vitroGST pull-down assays to determine the effect of the C-termi-nal deletion on the Gap1-Gap3 interaction. The interactionwas not affected by the deletion (Fig. 5B, lanes 2 and 3),suggesting the C-terminal Gap1 is not required for the protein-

FIG. 4. The Gap1 C-terminal motif is important for Gap1 function and Fap1 biogenesis. (A) Diagrammatic depiction of a gap1 allele fromVT324. (B) Homology comparison of the C-terminal domain of Gap1 of S. parasanguinis and its homologs from S. agalactiae (AAZ95528), S.gordonii (AAK16998), S. sanguinis (ABN44261), S. pneumoniae (AAK75837), Staphylococcus haemolyticus (BAE03637), and Lactobacillus johnso-nii (AAS08375). The conserved amino acid residues are highlighted. (C) A key amino acid residue in the C-terminal motif is important for Gap1function. Cell lysates prepared from the wild type (lane 1), fap1 mutant (lane 2), and VT324 (lane 3) as well as Gap1 site-directed mutants W518A(lane 4) and W518Y (lane 5) were analyzed by Western blotting with three Fap1-specific antibodies, mAbE42 (I), mAbD10 (II), and mAbF51 (III),or anti-FimA antibody (IV). Arrows point to the positions corresponding to Fap1 or FimA.

FIG. 5. The C-terminal deletion does not affect interaction be-tween Gap1 and Gap3. Gap1(�513–525) and Gap1 interactions withGap3 were determined by in vitro GST pull-down assays. (A) SDS-polyacrylamide gel electrophoresis analyses of GST (lane 1), GST-Gap1 (lane 2), and GST-Gap1(�513–525) (lane 3) fusion proteinspurified using glutathione Sepharose beads from respective bacterialstrains. The gel was stained with Coomassie blue. M, protein marker.(B) In vitro GST pull-down assays. The purified GST (lane 1), GST-Gap1 (lane 2), or Gap1(�513–525) (lane 3) glutathione Sepharosebeads were incubated with in vitro-translated c-Myc–Gap3. The cap-tured protein complexes were subjected to Western blot analyses withc-Myc antibody. “Input” represents in vitro-translated c-Myc–Gap3.

5630 ZHOU ET AL. INFECT. IMMUN.

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from

protein interaction. This result is consistent with the notionthat the C terminus has a glycosylation functional domain andrepresents a distinct functional domain for Gap1.

In summary, we have shown that all nonadherent mutantswe isolated via their ability to interact with the in vitro toothmodel are defective in Fap1 and Fap1 biogenesis. The forwardgenetic approach demonstrated the importance of Fap1 inbacterial adhesion and revealed the important function ofGap1 and its last 13 amino acid residues in Fap1 biogenesis. AsFap1-like proteins are highly conserved in many streptococciand staphylococci as adhesins, such approaches are invaluablein determining genes related to the biogenesis of Fap1-depen-dent and -independent adhesion molecules.

ACKNOWLEDGMENTS

This study was supported by K22 DE014726, R01DE017954, andR01DE11000 (H. Wu) from the National Institute of Dental andCraniofacial Research.

REFERENCES

1. Bennett-Lovsey, R. M., A. D. Herbert, M. J. Sternberg, and L. A. Kelley.2008. Exploring the extremes of sequence/structure space with ensemble foldrecognition in the program Phyre. Proteins 70:611–625.

2. Bensing, B. A., and P. M. Sullam. 2002. An accessory sec locus of Strepto-coccus gordonii is required for export of the surface protein GspB and fornormal levels of binding to human platelets. Mol. Microbiol. 44:1081–1094.

3. Bu, S., Y. Li, M. Zhou, P. Azadin, M. Zeng, P. Fives-Taylor, and H. Wu. 2008.Interaction between two putative glycosyltransferases is required for glyco-sylation of a serine-rich streptococcal adhesin. J. Bacteriol. 190:1256–1266.

4. Burnette-Curley, D., V. Wells, H. Viscount, C. L. Munro, J. C. Fenno, P.Fives-Taylor, and F. L. Macrina. 1995. FimA, a major virulence factorassociated with Streptococcus parasanguis endocarditis. Infect. Immun. 63:4669–4674.

5. Carlsson, J., H. Grahnen, G. Jonsson, and S. Wikner. 1970. Establishment ofStreptococcus sanguis in the mouths of infants. Arch. Oral Biol. 15:1143–1148.

6. Chen, Q., B. Sun, H. Wu, Z. Peng, and P. M. Fives-Taylor. 2007. Differentialroles of individual domains in selection of secretion route of a Streptococcusparasanguinis serine-rich adhesin, Fap1. J. Bacteriol. 189:7610–7617.

7. Chen, Q., H. Wu, and P. M. Fives-Taylor. 2004. Investigating the role ofsecA2 in secretion and glycosylation of a fimbrial adhesin in Streptococcusparasanguis FW213. Mol. Microbiol. 53:843–856.

8. Chen, Q., H. Wu, R. Kumar, Z. Peng, and P. M. Fives-Taylor. 2006. SecA2is distinct from SecA in immunogenic specificity, subcellular distribution andrequirement for membrane anchoring in Streptococcus parasanguis. FEMSMicrobiol. Lett. 264:174–181.

9. Cole, R. M., G. B. Calandra, E. Huff, and K. M. Nugent. 1976. Attributes ofpotential utility in differentiating among “group H” streptococci or Strepto-coccus sanguis. J. Dent. Res. 55:A142–A153.

10. Elder, B. L., D. K. Boraker, and P. M. Fives-Taylor. 1982. Whole-bacterialcell enzyme-linked immunosorbent assay for Streptococcus sanguis fimbrialantigens. J. Clin. Microbiol. 16:141–144.

11. Elder, B. L., and P. Fives-Taylor. 1986. Characterization of monoclonalantibodies specific for adhesion: isolation of an adhesin of Streptococcussanguis FW213. Infect. Immun. 54:421–427.

12. Fachon-Kalweit, S., B. L. Elder, and P. Fives-Taylor. 1985. Antibodies thatbind to fimbriae block adhesion of Streptococcus sanguis to saliva-coatedhydroxyapatite. Infect. Immun. 48:617–624.

13. Fives-Taylor, P. M., and D. W. Thompson. 1985. Surface properties ofStreptococcus sanguis FW213 mutants nonadherent to saliva-coated hydroxy-apatite. Infect. Immun. 47:752–759.

14. Froeliger, E. H., and P. Fives-Taylor. 2001. Streptococcus parasanguis fim-bria-associated adhesin Fap1 is required for biofilm formation. Infect. Im-mun. 69:2512–2519.

15. Guerin, M. E., J. Kordulakova, F. Schaeffer, Z. Svetlikova, A. Buschiazzo, D.Giganti, B. Gicquel, K. Mikusova, M. Jackson, and P. M. Alzari. 2007.Molecular recognition and interfacial catalysis by the essential phosphatidyl-inositol mannosyltransferase PimA from mycobacteria. J. Biol. Chem. 282:20705–20714.

16. Jenkinson, H. F., and R. J. Lamont. 2005. Oral microbial communities insickness and in health. Trends Microbiol. 13:589–595.

17. Jenkinson, H. F., and R. J. Lamont. 1997. Streptococcal adhesion and col-onization. Crit. Rev. Oral Biol. Med. 8:175–200.

18. Klemm, P., V. Hancock, M. Kvist, and M. A. Schembri. 2007. Candidatetargets for new antivirulence drugs: selected cases of bacterial adhesion andbiofilm formation. Future Microbiol. 2:643–653.

19. Peng, Z., H. Wu, T. Ruiz, Q. Chen, M. Zhou, B. Sun, and P. Fives-Taylor.2008. Role of gap3 in Fap1 glycosylation, stability, in vitro adhesion, andfimbrial and biofilm formation of Streptococcus parasanguinis. Oral Micro-biol. Immunol. 23:70–78.

20. Plummer, C., H. Wu, S. W. Kerrigan, G. Meade, D. Cox, and C. W. IanDouglas. 2005. A serine-rich glycoprotein of Streptococcus sanguis mediatesadhesion to platelets via GPIb. Br. J. Haematol 129:101–109.

21. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold SpringHarbor, NY.

22. Seifert, K. N., E. E. Adderson, A. A. Whiting, J. F. Bohnsack, P. J. Crowley,and L. J. Brady. 2006. A unique serine-rich repeat protein (Srr-2) and novelsurface antigen (epsilon) associated with a virulent lineage of serotype IIIStreptococcus agalactiae. Microbiology 152:1029–1040.

23. Siboo, I. R., H. F. Chambers, and P. M. Sullam. 2005. Role of SraP, aserine-rich surface protein of Staphylococcus aureus, in binding to humanplatelets. Infect. Immun. 73:2273–2280.

24. Siezen, R. J. 1999. Multi-domain, cell-envelope proteinases of lactic acidbacteria. Antonie van Leeuwenhoek 76:139–155.

25. Takahashi, Y., E. Takashima, K. Shimazu, H. Yagishita, T. Aoba, and K.Konishi. 2006. Contribution of sialic acid-binding adhesin to pathogenesis ofexperimental endocarditis caused by Streptococcus gordonii DL1. Infect. Im-mun. 74:740–743.

26. Tettelin, H., K. E. Nelson, I. T. Paulsen, J. A. Eisen, T. D. Read, S. Peterson,J. Heidelberg, R. T. DeBoy, D. H. Haft, R. J. Dodson, A. S. Durkin, M.Gwinn, J. F. Kolonay, W. C. Nelson, J. D. Peterson, L. A. Umayam, O. White,S. L. Salzberg, M. R. Lewis, D. Radune, E. Holtzapple, H. Khouri, A. M.Wolf, T. R. Utterback, C. L. Hansen, L. A. McDonald, T. V. Feldblyum, S.Angiuoli, T. Dickinson, E. K. Hickey, I. E. Holt, B. J. Loftus, F. Yang, H. O.Smith, J. C. Venter, B. A. Dougherty, D. A. Morrison, S. K. Hollingshead,and C. M. Fraser. 2001. Complete genome sequence of a virulent isolate ofStreptococcus pneumoniae. Science 293:498–506.

27. Wu, H., S. Bu, P. Newell, Q. Chen, and P. Fives-Taylor. 2007. Two genedeterminants are differentially involved in the biogenesis of Fap1 precursorsin Streptococcus parasanguis. J. Bacteriol. 189:1390–1398.

28. Wu, H., and P. M. Fives-Taylor. 1999. Identification of dipeptide repeats anda cell wall sorting signal in the fimbriae-associated adhesin, Fap1, of Strep-tococcus parasanguis. Mol. Microbiol. 34:1070–1081.

29. Wu, H., K. P. Mintz, M. Ladha, and P. M. Fives-Taylor. 1998. Isolation andcharacterization of Fap1, a fimbriae-associated adhesin of Streptococcusparasanguis FW213. Mol. Microbiol. 28:487–500.

30. Wu, H., M. Zeng, and P. Fives-Taylor. 2007. The glycan moieties and theN-terminal polypeptide backbone of a fimbria-associated adhesin, Fap1, playdistinct roles in the biofilm development of Streptococcus parasanguinis.Infect. Immun. 75:2181–2188.

31. Zhang, Y. Q., S. X. Ren, H. L. Li, Y. X. Wang, G. Fu, J. Yang, Z. Q. Qin, Y. G.Miao, W. Y. Wang, R. S. Chen, Y. Shen, Z. Chen, Z. H. Yuan, G. P. Zhao, D.Qu, A. Danchin, and Y. M. Wen. 2003. Genome-based analysis of virulencegenes in a non-biofilm-forming Staphylococcus epidermidis strain (ATCC12228). Mol. Microbiol. 49:1577–1593.

Editor: A. Camilli

VOL. 76, 2008 GENES INVOLVED IN S. PARASANGUINIS Fap1 BIOGENESIS 5631

on February 18, 2021 by guest

http://iai.asm.org/

Dow

nloaded from