The Collagen Binding Protein Cnm Contributes to Oral Colonizationand Cariogenicity of Streptococcus mutans OMZ175

James H. Miller,a Alejandro Avilés-Reyes,a,b Kathy Scott-Anne,a Stacy Gregoire,a Gene E. Watson,c,d Edith Sampson,e

Ann Progulske-Fox,e Hyun Koo,f William H. Bowen,a José A. Lemos,a,b Jacqueline Abranchesa,b

Center for Oral Biology,a Department of Microbiology and Immunology,b Department of Dentistry,c and Department of Environmental Medicine,d University of RochesterSchool of Medicine and Dentistry, Rochester, New York, USA; Department of Oral Biology and Center for Molecular Microbiology, University of Florida College of Dentistry,Gainesville, Florida, USAe; Biofilm Research Laboratories, Levy Center for Oral Health, Division of Pediatric Dentistry and Community Oral Health, School of DentalMedicine, University of Pennsylvania, Philadelphia, Pennsylvania, USAf

Streptococcus mutans is the etiological agent of dental caries and one of the many bacterial species implicated in infective endo-carditis. The expression of the collagen-binding protein Cnm by S. mutans has been associated with extraoral infections, but itsrelevance for dental caries has only been theorized to date. Due to the collagenous composition of dentinal and root tissues, wehypothesized that Cnm may facilitate the colonization of these surfaces, thereby enhancing the pathogenic potential of S. mu-tans in advancing carious lesions. As shown for extraoral endothelial cell lines, Cnm mediates the invasion of oral keratinocytesand fibroblasts by S. mutans. In this study, we show that in the Cnm� native strain, OMZ175, Cnm mediates stringent adhesionto dentinal and root tissues as well as collagen-coated surfaces and promotes both cariogenicity and carriage in vivo. In vitro, exvivo, and in vivo experiments revealed that while Cnm is not universally required for S. mutans cariogenicity, it contributes to(i) the invasion of the oral epithelium, (ii) enhanced binding on collagenous surfaces, (iii) implantation of oral biofilms, and (IV)the severity of caries due to a native Cnm� isolate. Taken together, our findings reveal that Cnm is a colonization factor that con-tributes to the pathogenicity of certain S. mutans strains in their native habitat, the oral cavity.

The oral cavity is colonized by hundreds, if not thousands, ofbacterial species that occupy specialized niches within the

mouth. One such species, Streptococcus mutans, is a hard-tissuecolonizer and is considered the major microbial etiological agentof dental caries. In addition, oral streptococci, including S. mu-tans, are implicated as causative agents of extraoral infections,such as infective endocarditis (1). The enhanced capacity of S.mutans to form biofilms in the presence of sucrose, its ability tometabolize a wide range of carbohydrates, and its high tolerance offluctuations in pH and nutrient availability are all considered cru-cial traits for its survival, persistence, and causation of dental car-ies (2–4). Current paradigms of S. mutans attachment focus on itsability to bind oral surfaces, particularly the enamel and proteinswithin the salivary pellicle, via sucrose-dependent and -indepen-dent mechanisms. In the extensively studied sucrose-dependentmechanism, secreted glucosyltransferases (Gtfs) serve the dualrole of priming the tooth enamel surface with glucans for adhesionby surface glucan binding proteins (Gbps) and developing an ex-tracellular polysaccharide (EPS) superstructure that anchors thebiofilm and supports matrix-delineated pH microenvironmentsin situ (4–9). The sucrose-independent mechanism involves directsubstrate recognition by bacterial surface adhesins that bind eitherto constituents of the salivary pellicle on the tooth surface, such asSpaP (also known as P1, PA, or antigen I/II) (10, 11), or to com-ponents of underlying tissues that become exposed due to demin-eralization of the enamel, e.g., collagen from dentin and roots(11–13).

S. mutans is among several lactic acid bacteria in the oral cavitythat produce organic acids as fermentative end products fromdietary carbohydrates; these organic acids, in turn, acidify the oralenvironment. Biofilm accumulation slows the diffusion of organicacids and restricts the access of pH-buffering saliva, subjecting theenamel to repeated and prolonged exposure to an acidic pH

(pH �5.5) and thereby initiating the dental caries process (3, 14).Left untreated, further demineralization of the enamel and exten-sion of the lesion expose the underlying dentin, presenting colla-gen and other, additional substrates for bacterial colonization (15,16). Other oral tissues, such as the cementum, root, and periodon-tal ligament fibers, are also rich in collagen (17) and, if exposed tothe oral environment, may be vulnerable to attachment and col-onization by microbes equipped with collagen-binding adhesins(13, 18–20). In S. mutans, SpaP, WapA, Cnm, and Cbm have beenidentified as collagen-binding adhesins, but thus far, only SpaPhas been examined for its role in attachment to dentin (10). How-ever, the collagen-binding affinity of SpaP has been found to bemarginal compared to those of the more recently discovered ad-hesins Cnm and Cbm (12, 21, 22).

The gene coding for the collagen- and laminin-binding surfaceprotein Cnm has been found in approximately 10 to 20% of S.

Received 3 December 2014 Returned for modification 6 January 2015Accepted 21 February 2015

Accepted manuscript posted online 2 March 2015

Citation Miller JH, Avilés-Reyes A, Scott-Anne K, Gregoire S, Watson GE, SampsonE, Progulske-Fox A, Koo H, Bowen WH, Lemos JA, Abranches J. 2015. The collagenbinding protein Cnm contributes to oral colonization and cariogenicity ofStreptococcus mutans OMZ175. Infect Immun 83:2001–2010.doi:10.1128/IAI.03022-14.

Editor: A. Camilli

Address correspondence to Jacqueline Abranches,jacqueline_abranches@urmc.rochester.edu.

Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.03022-14.

Copyright © 2015, American Society for Microbiology. All Rights Reserved.

doi:10.1128/IAI.03022-14

May 2015 Volume 83 Number 5 iai.asm.org 2001Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

mutans clinical isolates (23), and the closely related cbm gene wasprevalent in approximately 2% (24). Although serotype c strainspredominate in dental plaque (�70 to 80%), cnm and cbm arefound mostly in strains belonging to the less prevalent serotypes e,f, and k (23–26). Previously, we showed that Cnm mediates theinvasion of human coronary artery endothelial cells (HCAEC) invitro and that Cnm� strains exhibit increased virulence in theGalleria mellonella insect model (26). Moreover, recent investiga-tion of S. mutans collagen-binding proteins (CBPs), such as Cnmand Cbm, has linked these adhesins with extraoral infections, in-cluding infective endocarditis, hemorrhagic stroke, and athero-sclerotic plaque development (1, 27–29). To date, studies havefocused on the role of Cnm in the extraoral virulence of S. mutansbut have yet to elucidate whether Cnm confers any advantage onS. mutans in the colonization of oral tissues and the developmentof dental caries. Here we utilized in vitro, ex vivo, and in vivoapproaches to evaluate the role of Cnm in colonization, persis-tence, and virulence in the oral cavity—the natural habitat of S.mutans.

MATERIALS AND METHODSBacterial strains and growth conditions. The strains used in this study(Table 1) were routinely grown in brain heart infusion (BHI) at 37°Cunder a 5% CO2 atmosphere. The integrative vectors pBGK (30) andpBGE (31) were used to incorporate the nonpolar kanamycin (pBGK) anderythromycin (pBGE) resistance cassettes into the gtfA locus of the S.mutans chromosome for the purpose of strain differentiation duringcompetitive binding assays. The nisin-inducible Cnm-expressing con-struct pcnm (26) was introduced into OMZ175 to produce the Cnm-overproducing strain OMZ175pcnm. Where applicable, kanamycin sul-fate (1 mg ml�1), erythromycin (10 �g ml�1), or nisin (500 ng ml�1) wasadded to the growth medium.

Oral cell line attachment and invasion assays. The capacity of S. mu-tans for attachment to, and invasion of, oral cell lines was assessed bymethods described elsewhere, with minor modifications (32). Immortal-ized human gingival fibroblasts (HGF-1 [CRL-2014]; ATCC, Manassas,VA) were maintained in a 37°C incubator under 5% CO2 in Dulbecco’smodified Eagle medium (DMEM) with 4.5 g liter�1 glucose, sodium py-ruvate, and L-glutamine (Cellgro, Corning, NY) supplemented with 10%fetal bovine serum (HyClone, Logan, UT), passaged as many as 8 times byusing Accutase solution (Innovative Cell Technologies, San Diego, CA)according to the manufacturer’s instructions, and seeded in 24-well platesfor assay purposes. Briefly, stationary-phase cultures were harvested,washed, resuspended in tissue culture medium, and kept on ice. Mean-while, HGF-1 cells were preincubated on ice blocks for 30 min, followedby coculturing with S. mutans at a multiplicity of infection (MOI) of 100.Cocultures were centrifuged for 10 min at 800 rpm and 4°C to synchro-

nize attachment and invasion. For attachment assays, centrifuged cocul-tures were maintained on ice for 20 min (total of 30 min of incubation forattachment), washed three times with ice-cold Hanks’ buffered saline so-lution (HBSS), and lysed by incubation at 4°C with ice-cold sterile water.The inoculum and lysates were serially diluted and were plated on BHIagar for CFU enumeration.

Bacterial invasion was assessed by using an antibiotic protection assaywith the preparations described above. Centrifuged cocultures werewarmed rapidly for 3 min in a 37°C water bath, incubated at 37°C under a5% CO2 atmosphere for 2 h, and washed twice with HBSS. Cocultureswere then incubated for 3 h with fresh culture medium containing peni-cillin G (10 �g ml�1) and gentamicin (0.3 mg ml�1) to kill extracellularbacteria, followed by two washes with warm phosphate-buffered saline(PBS). Lysates were recovered by the procedures described above.

Human oral keratinocytes (HOK) (ScienCell, Carlsbad, CA) were cul-tured in oral keratinocyte medium (OKM) with a penicillin-streptomycinsolution (ScienCell). For assay purposes, cells were seeded at 2 � 105/wellin 24-well plates in medium without antibiotics. Invasion and attachmentassays were performed using the same procedures as those described forthe HGF-1 cell assays. In order to compare results to those for the controlstrain (OMZ175), attachment and invasion efficiencies from each inde-pendent experiment were normalized to the experimental averages forOMZ175 infections and were compared statistically by the Kruskal-Wallistest with Dunn’s post hoc multiple-comparison test. Statistical analyseswere performed using GraphPad Prism, version 6.0e for Macintosh(GraphPad Software, San Diego, CA).

TEM of oral cell line invasion by OMZ175. Transmission electronmicroscopy (TEM) was performed on HGF-1 cells and HOK coincubatedwith OMZ175 in order to observe host invasion and the localization ofinvading cells as described elsewhere (33). Sections were examined using aHitachi H7000 TEM (Hitachi High Technologies America Inc., Schaum-burg, IL), and images were acquired with a Veleta 2k � 2k camera withiTEM software (Olympus Soft Imaging Solutions Corp., Lakewood, CO).Samples were fixed, embedded, and imaged by the Interdisciplinary Cen-ter for Biotechnology Research Electron Microscopy and BioimagingCore facility at the University of Florida.

Adherence to saliva- and collagen-coated HA beads. The potential ofOMZ175, OMZ175�cnm, and UA159 for binding to hydroxyapatite(HA) beads, a tooth enamel surrogate, coated with either collagen type I,clarified whole saliva, or both was assessed using methods described else-where, with minor modifications (20, 34, 35). Whole stimulated saliva wasobtained from healthy adult volunteers under a protocol approved by theResearch Subjects Review Board of the University of Rochester (RSRB no.00030432). Briefly, cultures were grown to mid-exponential phase in low-molecular-weight broth (LMWG) (0.25% [wt/vol] Bacto tryptone, 0.15%[wt/vol] Bacto yeast extract, 25 mM KH2PO4, 4 mM MgSO4, [pH 7.35],filtered to a 10-kDa cutoff and supplemented with 1% glucose) and triti-ated thymidine (10 �Ci ml�1). The cultures were washed with absorptionbuffer (AB) (1 mM pH 6.5 potassium phosphate buffer plus 50 mM KCl,

TABLE 1 Streptococcus mutans strains used in this study

Strain Expt(s)a Relevant characteristic(s) Source or reference

OMZ175 a, b, c, d, e, f Cnm� B. GuggenheimOMZ175�cnm a, b, c, d, e, f Cnm� Kanr 26OMZ175�cnm/pcnm c, e, f Complementary Cnm expression; Kanr Ermr 26OMZ175pcnm c, e, f Adjunct Cnm expression; Ermr This study; 26OMZ175-pBGE c Cnm� Ermr This study; 31OMZ175-pBGK c Cnm� Kanr This study; 30UA159 a, b, d, e, f, g Cnm� University of AlabamaUA159-cnm c, d, e, f, g Cnm� Ermr 21UA159-pBGE c, d, e, f Ermr 21, 31UA159-pBGK c Kanr This study; 30a a, antibiotic protection assays; b, in vitro binding to coated hydroxyapatite beads; c, ex vivo tooth substrate binding competition assay; d, in vivo colonization and caries formationin Sprague-Dawley rats; e, Western blotting for Cnm; f, in vitro binding to collagen type I; g, immunofluorescent detection of Cnm.

Miller et al.

2002 iai.asm.org May 2015 Volume 83 Number 5Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

1 mM CaCl2, and 0.1 mM MgCl2) and were concentrated to a 2� volume,followed by disruption of cell chains by four brief sonication cycles (15 s at7 W with 30-s intervals). The culture density was adjusted with AB to anoptical density at 600 nm (OD600) of 1.0, and the emitted beta radiationwas determined by scintillation counts. Ceramic hydroxyapatite type Ibeads (80 �m; Bio-Rad, Hercules, CA) were prepared in one of three ways:(i) saliva treatment (immersion in clarified saliva on a rocker for 1 h at37°C, followed by 3 washes with buffer AB2 [AB with 0.1 M phenylmeth-ylsulfonyl fluoride plus 0.02% sodium azide] and incubation in buffer ABfor 1 h at 4°C), (ii) collagen treatment (immersion for 1 h at 4°C inacid-solubilized rat tail collagen type I [Sigma, St. Louis, MO] diluted inbuffer AB [0.5 mg ml�1], followed by washes with buffer AB2 and incu-bation again in buffer AB for 1 h at 4°C), or (iii) collagen/saliva treatment(the same protocol as for the collagen treatment, followed by incubationin saliva for 1 h at 4°C after washes with AB2). Nonspecific binding wasminimized by blocking all bead preparations with 0.5 mg ml�1 bovineserum albumin (BSA) in AB for 1 h at 37°C. Prepared beads were incu-bated with culture aliquots on a rocker at 37°C for 1 h, followed by washesin AB and emission measurement by liquid scintillation. Bacterial attach-ment efficiencies were determined by the average emission per CFU andwere normalized to the rates of attachment observed for strain OMZ175,and analysis of variance (ANOVA) was performed on the log10-trans-formed data with Bonferroni post hoc comparisons.

In vitro collagen binding assay. The potential of study strains to bindto immobilized collagen type I on microtiter plates was assessed usingmethods described elsewhere (21), with minor modifications. Cells weregrown and treated as described under “Cnm detection” below, followedby washing and resuspension in PBS to achieve 4-fold concentrations.Viable-cell densities of cell preparations were checked by serial dilutionand plating on BHI agar to ensure similar loading. The A575 for cellsbinding to no-collagen control wells (background) was subtracted fromreadings for collagen-bound cells from the same inoculum, and the result-ing values were normalized to that for OMZ175. Statistical analyses wereperformed on the log10-transformed data by using ANOVA with Bonfer-roni post hoc comparisons.

Ex vivo binding competition assays on tooth substrates. The avidityand specificity of the binding of S. mutans to tooth substrates were eval-uated by competition assays using strain pairs mixed 1:1 and coincubatedwith sectioned teeth. Surgically extracted nonimpacted third molars werecleaned of all soft tissue detritus and were cut horizontally using a circularsaw blade at the cementum-enamel junction to separate the crown fromthe roots (see Fig. 3A). Tooth sections were sonicated in sterile water toremove any bound microorganisms and were screened for microbial con-taminants by plating sonicants on tryptic soy agar (TSA). Crown sectionsintended for binding experiments involving only the enamel surface(identified as “enamel” in this study) were masked with a melted wax capon the cut face to block binding to the exposed dentin. This cap wasremoved using a sterile scalpel before the bacterial release step. Toothparticles labeled “dentin” had the apical portion of the crown removedand no wax applied. All tooth sections were bathed in clarified sterilesaliva for 30 min at 37°C prior to coinfection. Briefly, cultures grown inLMWG medium to mid-exponential phase were mixed 1:1 and were in-cubated with the tooth substrates for 5 min at 37°C under a 5% CO2

atmosphere. The tooth sections then underwent three washes (1, 2, and 7min) in 14 ml sterile PBS (pH 7.2) on a rotator (60 rpm) to remove looselybound cells. Stringently bound cells were released by three sonicationcycles of 10 s each, with 30-s rests on ice between cycles. Sonicants werethen serially diluted in 0.1 M glycine (pH 7.2), followed by plating on TSAsupplemented with erythromycin or kanamycin. CFU were enumeratedafter 48 h of incubation at 37°C under a 5% CO2 atmosphere. For com-petition assays in the presence of sucrose, cells were grown as describedabove in LMWG medium, pelleted, resuspended in LMWG medium plus1% sucrose, incubated for 15 min at 37°C under a 5% CO2 atmosphere,and then coincubated with tooth sections for 5 min. Statistical analyseswere performed on the log10-transformed competitive indices using

ANOVA with Bonferroni post hoc comparisons to compare substrate pref-erences among strain pairs.

Rat caries model. The cariogenic potentials of OMZ175, OMZ175�cnm,UA159, UA159-cnm, and UA159-pBGE were assessed in Sprague-Dawleyrats by methods described elsewhere (36, 37). Approval for this study wasobtained from the University of Rochester Committee on Animal Re-sources (UCAR no. 100583), and test subject sample sizes were chosen onthe basis of previous studies using similar model parameters so as to en-sure sufficient power to detect distribution differences (38, 39). FemaleSprague-Dawley pups aged 19 days were purchased with their dams (Har-lan Laboratories, Indianapolis, IN). Animals received Diet 2000 (HarlanLaboratories, Indianapolis, IN) and sterile 5% sucrose water ad libitum forthe duration of the study. S. mutans cultures were grown to late-exponen-tial phase in LMWG and were swabbed onto all interior surfaces of themouth anterior to the oropharynx for 4 consecutive days, starting at 19days of age. Dams were similarly infected on the same days with the samecultures as those for their litters and were housed with their pups for 1week to facilitate strain transmission and implantation. At the age of 26days, each pup was paired with a similarly infected pup from a differentlitter, and the pairs were housed in wire-bottom cages, in order to mini-mize litter-based biases (40). Pups were screened for strain implantationat the ages of 25, 27, and 29 days by plating oral swabs on MSB (mitissalivarius agar plus 200 U bacitracin liter�1), and the identities of thestrains were confirmed by PCR to ensure that there was no cross-contam-ination. Primer pair 1876F/1877R (23) produced a 1.4-kbp product fromUA159 and 7.0- and 7.8-kbp products from OMZ175 and OMZ175�cnm,respectively.

After 30 days of infection, the pups were killed by CO2 asphyxiation,and the lower left jaws were aseptically removed and placed in sterile salinesolution. The jaws were sonicated using the same parameters as thosedescribed for the ex vivo experiments, serially diluted, and plated for totalbacterial counts on TSA II (BD Bioscience) and for S. mutans counts onMSB. Statistical differences were ascertained either by a Kruskal-Wallistest with Dunn’s post hoc comparisons or by a Kolmogorov-Smirnov test.Carious lesions were evaluated by a single calibrated examiner blinded tothe study groups. The extension and severity of lesions on smooth andsulcal surfaces of the teeth were analyzed using Larson’s modification ofKeyes’ method (41, 42), and statistical significance was determined byeither ANOVA with Bonferroni’s comparisons or Student’s 2-tailed t test,using the arcsine normalization transform of enumerated lesions (40).Exceptions to analysis techniques were made for the comparison of exten-sive sulcal lesions in Fig. 4E due to the absence of such lesions in allOMZ175�cnm-infected rats (Kruskal-Wallis test with Dunn’s post hoctest).

pH drop and acid killing. The glycolytic capacities of UA159,OMZ175, and their derivative strains were compared by pH drop exper-iments using standard protocols (34, 43). For acid killing experiments,cultures grown to mid-exponential phase were washed once with 0.1 Mglycine (pH 7.0) and were resuspended in 0.1 M glycine (pH 2.8). Sampleswere stirred continuously at room temperature; aliquots of cells wereremoved at predetermined intervals; and viability was determined by CFUenumeration on BHI agar. The resulting distributions were analyzed byANOVA to evaluate survivability distributions for each time point.

Biofilm analyses. The biofilm composition (intracellular water-solu-ble and extracellular insoluble polysaccharides) and dry weight of S. mu-tans strains grown in the presence of sucrose were assessed using methodsdescribed previously (34, 44). The results within each assessment werecompared using the Kruskal-Wallis test with Dunn’s post hoc compari-sons.

Cnm detection. The production and localization of Cnm by studystrains were confirmed by Western blotting and immunofluorescencetechniques described previously (45). Briefly, cells were grown in LMWGto an OD600 of 0.2, at which point cultures were treated with nisin, whereapplicable, and were grown for an additional 2 h. Whole-cell lysates wereobtained and were resolved by 10% SDS-PAGE, followed by transfer to a

Cnm Facilitates S. mutans Oral Colonization

May 2015 Volume 83 Number 5 iai.asm.org 2003Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

polyvinylidene difluoride (PVDF) membrane. The blots were probed forCnm using a rabbit anti-Cnm antibody (21) with a horseradish peroxi-dase-coupled secondary antibody. Immunodetection was performed us-ing the ECL chemiluminescent system (GE Life Sciences, Piscataway, NJ),and blots were exposed to X-Omat LS film (Carestream, Rochester, NY).The export of Cnm to the cell surface was validated using immunofluo-rescent labeling of surface-bound Cnm as described elsewhere (45).

RESULTSCnm mediates the invasion of oral epithelial cell lines. Previ-ously, we and others showed that Cnm facilitates enhanced inva-sion and attachment of two nonoral cell lines, HCAEC and humanumbilical vein endothelial cells (HUVEC) (22, 26). Here we de-termined the contribution of Cnm toward S. mutans attachmentto, and invasion of, two distinct cell lines of the oral epithelium:human gingival fibroblasts (HGF-1) and human oral keratino-cytes (HOK). The median intracellular viable counts of UA159recovered from both HGF-1 cells and HOK were zero. Coincuba-tion of HGF-1 cells with OMZ175 (Cnm�), OMZ175�cnm(Cnm�), or UA159 (Cnm�) showed similar levels of attachmentto the host cell surface (Fig. 1A) (P 0.686). A previous study

assessing attachment to HCAEC revealed that Cnm contributed tohigher attachment rates by Cnm� S. mutans strains (26). Thiscould be attributed both to differences in host cell lines and todifferences in methodology that included centrifugation of thecoculture to facilitate cell-cell contact. Nevertheless, the recoveryof intracellular S. mutans bacteria from HGF-1 cells showed sig-nificant differences between the Cnm� strain OMZ175 and theCnm� strains (P � 0.0089). In addition, coincubation of thesestrains with HOK revealed significant differences between theCnm� and Cnm� strains in both attachment (P � 0.0042) andinvasion (P � 0.0017) (Fig. 1B). TEM analysis of HGF-1 cellsinfected with OMZ175 for 2 h validated the antibiotic protectionassay results, revealing cell membrane extrusions surrounding S.mutans cells and intracellular bacteria contained within mem-brane-bound vacuoles (Fig. 1C and D). Thus, our findings suggestthat Cnm may confer an advantage on S. mutans in colonizing oralepithelial tissues, thereby expanding its possible niches for persis-tence in the oral cavity.

Cnm enhances binding to collagen-coated hydroxyapatitesurfaces. Whereas tooth enamel is a noncollagenous mineralized

FIG 1 Attachment to, and invasion of, human gingival fibroblasts (HGF-1) by OMZ175 (Cnm�), OMZ175�cnm (cnm knockout), or UA159 (Cnm�). (A andB) Antibiotic protection assay to assess S. mutans attachment to, and invasion of, human gingival fibroblasts (A) and human oral keratinocytes (B). Medians andinterquartile ranges for eight replicates were normalized to the average results for OMZ175 infections. Asterisks indicate significant differences (P � 0.05). (C)Transmission electron microscopy of HGF-1 cells coincubated with OMZ175 for 2 h revealed S. mutans bacteria attached to HGF-1 cells, enveloped bylamellipodium-like extensions (black arrows), and intracellular S. mutans bacteria within membrane-bound vacuoles (white arrows). (D) Detail of transmissionelectron micrograph. Projections from the surfaces of OMZ175 cells were visible (black arrows) and appeared to be in contact with the HGF-1 vesicle wall (whitearrows).

Miller et al.

2004 iai.asm.org May 2015 Volume 83 Number 5Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

tissue, the underlying dentin and adjacent roots are much lessmineralized and have an organic matrix composed mainly of typeI collagen (46). In order to determine whether S. mutans bindscollagen type I in a context similar to that found in the oral cavity,hydroxyapatite (HA) beads were coated with either collagen typeI, saliva, or both. The coated beads were incubated with OMZ175,OMZ175�cnm, or UA159 cells, and bound cells were quantified(Fig. 2). On HA beads coated with saliva alone, UA159 cells boundin higher numbers than OMZ175 or OMZ175�cnm cells (P �0.0006). However, on beads coated either with collagen type Ialone or with a combination of collagen and saliva, the Cnm�

strain OMZ175 bound significantly better than either Cnm�

strain (P � 0.0001).Cnm facilitates enhanced binding to dentinal and root sub-

strates. Based on the observation that OMZ175 displayed agreater capacity to bind to collagen-coated HA beads than Cnm�

strains, we evaluated the contribution of Cnm toward binding tobiologically pertinent substrates through competitive binding as-says. For this purpose, we developed an ex vivo approach to eval-uate the avidity of binding of our study strains to (i) enamel, (ii)exposed dentin, and (iii) root surfaces (Fig. 3A). Competitivebinding indices (CI) were obtained for the following groups:

OMZ175 versus OMZ175�cnm, OMZ175 versus UA159, andOMZ175�cnm versus UA159 (Fig. 3B). In all cases, similar bind-ing to enamel surfaces was observed for the three groups com-pared. However, the Cnm� strain OMZ175 outcompeted theCnm� strains on both dentin and root sections by approximately1 log unit (P � 0.0153). Inactivation of cnm in OMZ175 impairedits ability to bind dentin and root sections, resulting in affinitiesfor these substrates similar to those of the Cnm� native strainUA159. The advantage in binding to collagenous tooth substratesconferred on OMZ175 by Cnm was reduced to modest levels uponexposure to sucrose (see Fig. S6 in the supplemental material).While OMZ175 was still able to outcompete the Cnm� strains ondentin and root sections in sucrose, the differences between theindices obtained for collagenous and noncollagenous substrateswere not statistically significant in most cases.

Cnm complementation, overexpression in OMZ175, and ex-pression in UA159. To further confirm that collagen-bindingdominance ex vivo was conferred by Cnm in OMZ175, Cnm wasexpressed in trans (pcnm) using the nisin-inducible plasmidpMSP3535 in the �cnm strain (OMZ175�cnm/pcnm) (26). In ad-dition, the effects of Cnm overproduction by pcnm in the nativeCnm� background (OMZ175pcnm) and of Cnm expression inthe Cnm� strain (UA159-cnm) were evaluated in vitro and ex vivo.In vitro binding to collagen type I by the complemented strainOMZ175�cnm/pcnm showed nearly complete reversal of the col-lagen-binding defect observed in OMZ175�cnm (see Fig. S1A inthe supplemental material). This partial restoration may be attrib-uted to the fact that Cnm is not expressed in its native, fully gly-cosylated form (see Fig. S1B) (45), since the plasmid-encodedCnm migrates at a slightly smaller size than Cnm in OMZ175. Inthe ex vivo competition binding assays, OMZ175�cnm/pcnm out-competed OMZ175�cnm (see Fig. S2 in the supplemental mate-rial), although when coincubated with wild-type OMZ175,OMZ175�cnm/pcnm was outcompeted on all substrates.

Overexpression of Cnm in OMZ175 (OMZ175pcnm) did notenhance its capacity to bind collagen type I in vitro (see Fig. S1Aaand g in the supplemental material). Unexpectedly, the overex-pressing strain outcompeted OMZ175 in the ex vivo studies on allthree surfaces regardless of the apparent collagen content of thesubstrate. Both the complemented and overexpressing strains dis-

FIG 2 Binding of S. mutans strains to hydroxyapatite beads coated with eithersaliva, collagen, or saliva over collagen. Data are averages � standard devia-tions for three replicates normalized to results for OMZ175. Asterisks indicatesignificant differences (P � 0.05).

FIG 3 Ex vivo tooth substrate binding competition assay model. (A) Schematic of tooth sectioning to simulate three distinct substrates for competitive bindingassays. (B) Tooth sections were coated with saliva and were coincubated for 5 min with equal amounts of two strains containing different antibiotic resistancemarkers. Competition indices were then derived from the ratio of the number of viable cells recovered from the sections to the number of viable cells recoveredfrom the inocula. The dashed line indicates equal binding by competing strains. Averages � standard deviations from at least three independent experiments arepresented. Asterisks indicate significant differences (P � 0.05).

Cnm Facilitates S. mutans Oral Colonization

May 2015 Volume 83 Number 5 iai.asm.org 2005Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

played inconsistent phenotypes with regard to selective binding tocollagenous hard surfaces and were thus not considered useful forstudy in vivo. Next, we evaluated the contribution of Cnm to oralcolonization by expressing it in the Cnm� strain UA159. As ex-pected, we observed a gain of function for strain UA159-cnm,which bound to collagen at levels similar to those of OMZ175 invitro (see Fig. S1A). Cnm expression and proper localization to thecell surface for UA159-cnm were confirmed by Western blottingand immunofluorescence, respectively (see Fig. S1B and S3 in thesupplemental material). In addition, glycosylation by the nativemachinery in UA159 appears to properly decorate Cnm, as shownby a migration pattern similar to that of OMZ175 by Westernblotting (see Fig. S1B), and to confer a stability phenotype uponexposure to proteinase K (45). Ex vivo binding assays with UA159and UA159-cnm showed nearly equal binding to enamel surfaces(see Fig. S2) but advantageous adhesion to dentin and root sub-strates (P, 0.063 and 0.005, respectively) for UA159-cnm. Thus,heterologous expression of Cnm confers a clear advantage in ad-herence to dentin and root surfaces.

Inactivation of cnm decreases the colonization and cario-genic potentials of OMZ175. Our in vitro and ex vivo findingssuggested that Cnm might facilitate the colonization of dentin androots, leading us to assess the cariogenic potential of Cnm� strainsin the rat caries model. Since OMZ175 has never been tested for itsability to cause caries in rats, we included the widely characterizedUA159 strain in the study as a reference. Rat pups were infectedwith either UA159 (Cnm�), OMZ175 (Cnm�), or the �cnm mu-tant (OMZ175�cnm). The average total flora counts recoveredfrom rats infected with OMZ175�cnm were significantly lower

than those from rats infected with either UA159 or OMZ175 (P �0.001) (Fig. 4A). Similarly, S. mutans carriage was significantlylower in OMZ175�cnm infections than in either UA159 orOMZ175 infections (P � 0.001) (Fig. 4A). Interestingly, propor-tions of S. mutans to total flora tended to be higher in OMZ175infections (60.0% 19.8%) than in infections by either UA159 orOMZ175�cnm (43.6% 9.6% and 44.4% 8.9%, respectively),although these differences were not statistically significant (P 0.074).

Scoring of carious lesions on smooth and sulcal surfaces wasdivided into parameters evaluating either lesion extension(enamel [E]) or lesion severity (slight [Ds], moderate [Dm], orextensive [Dx]) as defined by Larson (41). All three strains pro-duced similar levels of average lesion extension (E) on bothsmooth and sulcal surfaces (Fig. 4B). However, comparisons oflesion severity (Fig. 4C to E) revealed significant differences result-ing from infections with different strains. In most cases, rats infectedwith OMZ175�cnm had significantly lower (P � 0.05) severity scoreson sulcal and smooth surfaces than rats infected with either OMZ175or UA159. These results reveal that Cnm plays an important role inoral colonization by OMZ175 and in its cariogenicity.

Despite an obvious role for Cnm in caries severity due toOMZ175 infections, animals infected with UA159 had lesions thattended to be even more severe than those caused by OMZ175. Inan attempt to determine the basis for these differences, UA159,OMZ175, and OMZ175�cnm were assayed for their abilities towithstand (aciduricity) and produce (acidogenicity) acids, as wellas to synthesize extracellular polysaccharides (EPS) and intracel-lular polysaccharides (IPS) in the presence of sucrose. Collec-

FIG 4 An in vivo caries study on Sprague-Dawley rats infected with S. mutans strains was carried out for 30 days, with 12 animals per infecting strain. (A) MedianCFU (error bars, interquartile ranges) of total flora and S. mutans bacteria recovered from the left mandible 30 days postinfection. Asterisks indicate significantdifferences (P � 0.05). (B through E) Caries scores defined by Larson (41). Results are averages � standard deviations. Asterisks indicate significant differences(P � 0.05). (B) Lesion extension levels on smooth and sulcal tooth surfaces. (C to E) Severity of lesions on smooth and sulcal tooth surfaces.

Miller et al.

2006 iai.asm.org May 2015 Volume 83 Number 5Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

tively, these traits encompass some of the key aspects of S. mutanscariogenic potential (3). Inactivation of cnm in OMZ175 did notsignificantly alter any of these characteristics from those of theparental strain (see Fig. S4 in the supplemental material). UA159and OMZ175 did not differ significantly in either acidogenicity oraciduricity (see Fig. S4A and B). However, analysis of biofilmsproduced by strains grown in the presence of sucrose (see Fig.S4C) revealed that UA159 produced moderately higher biofilmdry masses than OMZ175 (P � 0.08) with significantly higher IPScontents (P � 0.003). No significant differences in the amounts ofsoluble and insoluble EPS were observed between the UA159 andOMZ175 background strains.

Expression of Cnm in UA159 does not increase cariogenicity.Due to the enhanced binding to collagenous substrates attributedto Cnm expression by UA159 in the ex vivo assays, the cariogenicpotential of UA159-cnm was evaluated in the rat caries model (seeFig. S5 in the supplemental material). Infections were performedusing UA159-cnm or a UA159 variant transformed with the emptypBGE vector, the same vector used to integrate cnm into the chro-mosome to produce UA159-cnm (21). Although the differencewas not significant, the carriage of total flora and S. mutans wasslightly lower in UA159-cnm-infected rats than in rats infectedwith UA159-pBGE (P 0.67). The colonization profiles resultingfrom infection by UA159-cnm or UA159-pBGE were nearly iden-tical (52.4% 13.4% and 54.9% 16.5%, respectively). In addi-tion, average lesion extension (E) and severity (Ds, Dm, and Dx)were similar for the two strains. Thus, it appears that UA159, anaturally occurring Cnm� isolate, employs strategies for toothadherence, colonization, and persistence that outweigh any en-hanced function conferred by Cnm expression in vivo.

DISCUSSION

Bacterial surface proteins that mediate cellular binding to extra-cellular matrix (ECM) components, known as MSCRAMM (mi-crobial surface components recognizing adhesive matrix mole-cules), constitute a class of virulence factors that facilitate bacterialinfections through adhesion to, and colonization of, host tissues(47, 48). Avid binding of collagenous tissues is an importantmechanism utilized by a subset of S. mutans strains expressingCBPs to successfully colonize oral and extraoral tissues (1, 22, 24,26, 27, 49). This study highlights two important characteristicsthat can be attributed to Cnm, (i) enhanced invasion of epithelialcell lines relevant to soft tissues of the oral cavity and (ii) stringentadherence to collagenous oral hard tissues, which enhance thecolonization and cariogenic potentials of certain S. mutans strains.

Earlier studies have demonstrated that streptococci dominatethe intracellular flora of buccal cells in vivo (50) and have alsolinked Cnm expression in S. mutans with invasion of HCAEC (26)and HUVEC (27) and adherence to hepatocytes (49). These ob-servations prompted our investigation into Cnm-mediated at-tachment to, and invasion of, oral epithelial cell lines. As observedwith HCAEC and HUVEC, we found that Cnm expression is di-rectly linked to enhanced invasion of HGF-1 cells and HOK, sug-gesting that intracellular invasion may provide a protective reser-voir for Cnm� S. mutans strains. This would facilitate persistencein the oral cavity through mitigation of environmental challengessuch as host defenses, administered antibiotics, competition withbacterial commensals, and mechanical/chemical removal (51).Moreover, attachment to, and invasion of, epithelial cells, partic-ularly those directly involved in the healing process (e.g., gingival

fibroblasts), may further facilitate the access of S. mutans to thebloodstream through injury sites.

The presence of collagen on HA surfaces, regardless of coatingwith saliva, provided the substrate needed for stringent binding byCnm-expressing strains (OMZ175). Similarly, ex vivo tooth-bind-ing assays demonstrated that Cnm� strains outcompeted theirCnm� variants in adherence to root and dentin sections, but notto enamel, emphasizing the specificity of Cnm for collagen-richsubstrates. The dentin organic matrix is composed of approxi-mately 90% collagen type I, and surrounding tissues, such as thecementum, periodontal ligament fibers, and roots, are also rich incollagen (17). These tissues are ordinarily shielded from the oralflora either by protective hard surfaces (enamel) or by soft tissues(gingiva). However, physical trauma, dentin-level carious lesions,or gingival recession can expose these various collagenous tissuesto the oral flora. Thus, adhesion to, and persistence on, collage-nous tissues, mediated by Cnm, may enhance the ability of S.mutans to colonize tooth surfaces, especially when sucrose avail-ability is low.

In vivo caries experiments using rats infected with OMZ175 orits �cnm counterpart (OMZ175�cnm) strongly support the hy-pothesis that Cnm contributes to colonization by, and persistenceof, the Cnm� native S. mutans strain in the oral cavity. SinceOMZ175�cnm and the parent strain did not differ in acid toler-ance, acidogenicity, biofilm biomass, or IPS and EPS production,it can be inferred that the decreased caries severity index observedfor rats infected with OMZ175�cnm is directly associated with thereduced ability of this strain to bind to collagen. Furthermore,Cnm appears to modulate the carriage of both S. mutans and totalbiofilms, since rats infected with OMZ175�cnm had lower S. mu-tans and total flora counts than rats infected with the parentalstrain. Interestingly, animals from the same experiment infectedwith the Cnm� strain UA159 had colonization profiles similar tothose of rats infected by OMZ175 but overall displayed a greaternumber of severe lesions. Comparisons of biofilm characteristicsrevealed that UA159 had significantly higher average IPS accumu-lation than OMZ175. Heightened production of IPS is consideredimportant to the organism’s long-term persistence and has beenlinked to a lower bacterial fasting pH, caries-active status in hu-man patients, and caries development in animal models (52, 53).Additionally, in contrast to UA159, OMZ175 lacks the glucan-binding adhesin GbpA and expresses low levels of SpaP (54).GbpA has been implicated in sucrose-dependent biofilm forma-tion previously (5), and SpaP is a multifunctional adhesin impli-cated primarily in mediating adherence to the salivary pellicle (55,56). This observation is bolstered by the more-robust binding byUA159 than by OMZ175 to HA beads coated with saliva only,suggesting that UA159 is a more effective colonizer of intactenamel surfaces. A study by Nomura et al. (22) found that levels ofcollagen binding and HUVEC attachment/invasion were highestfor strains expressing CBPs only, intermediate for strains express-ing both CBPs and SpaP, and lowest for strains expressing SpaPonly. Taking these findings together, it appears that there is atrade-off between stringent binding to the salivary pellicle andstringent binding to collagenous surfaces, whereby Cnm� isolatesare able to colonize through optimized adherence via glucan andsalivary pellicle binding, and Cnm� isolates colonize by using col-lagen as a preferential binding substrate.

Expression of Cnm by UA159 has been shown previously toincrease the binding affinity of this strain for collagen and laminin

Cnm Facilitates S. mutans Oral Colonization

May 2015 Volume 83 Number 5 iai.asm.org 2007Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

substrates (21). Here we show that Cnm confers on UA159 anadvantage in binding to dentin and root surfaces ex vivo. However,these traits increased neither the colonization nor the cariogenic-ity of UA159 in the rat caries model. One reason might be that thein vivo model employed here utilizes a high-sucrose diet to induceand accelerate caries development. While effective at evaluatingthe contribution of glucan-mediated colonization to S. mutanspathogenicity, the high-sucrose diet may mask the relevance ofsucrose-independent adherence mechanisms, such as collagenbinding by CBPs. Despite this possibility, it should be noted thatthe production of glucans from sucrose is a major virulence factorof S. mutans, which may help explain why Cnm� strains do notpredominate in the oral cavity. On the other hand, Cnm� strainsmay take advantage of previously damaged enamel or exposedroots to further the disease process.

The success of S. mutans as an etiological agent of dental cariesrelies on a variety of mechanisms that the organism employs toadhere to hard surfaces in the oral cavity. Recently, mountingevidence has suggested that CBPs expressed by a subset of S. mu-tans strains enhance the potential of this pathogen as a causativeagent of infective endocarditis. We propose a mechanism bywhich CBPs may enable S. mutans to exploit predamaged tissuesso as to facilitate effective colonization and subsequent pathogen-esis, both inside and outside the oral cavity (Fig. 5). The expres-sion of CBPs by S. mutans confers targeted adhesion to collage-nous tissues and may facilitate invasion of the oral epithelium, aniche in which there is likely less competition with commensalsand other S. mutans strains than in supragingival plaque. Whereassucrose-mediated adherence capitalizes on the development ofbiofilms on nonshedding hard surfaces, stringent adhesion to col-lagen by CBPs anchors microbes to exposed dentin and root sur-faces. Dental procedures, trauma to the mouth, or poor oral hy-giene may facilitate the introduction of CBP-expressing strains

into the bloodstream, through which they are disseminatedthroughout the body, creating states of transient bacteremia. Fur-thermore, preexisting damage to the vascular endothelium andother host risk predispositions, combined with sufficient bacterialloads, increase the likelihood for extraoral infections and disease.

In summary, this study reveals that Cnm is required for thevirulence of OMZ175 with regard to dental caries, whereas UA159employs different mechanisms for colonization and pathogenesisin the oral cavity, further bolstering the hypothesis that S. mutansstrains have enough genetic variability to allow for different colo-nization, survival, and persistence strategies (57). Therefore, ourstudy reinforces the importance of collagen binding as an appara-tus for oral establishment utilized by a subset of S. mutans strains(10). Contemporary studies aimed at the identification, character-ization, and clinical surveillance of S. mutans strains expressingcollagen-binding adhesins have focused on the role of these pro-teins in S. mutans colonization and virulence outside the oral cav-ity. However, our study reveals that Cnm may also contribute tothe success of certain S. mutans strains in colonization, and thuspathogenesis, within its conventional habitat, the oral cavity.

ACKNOWLEDGMENTS

This work was supported by American Heart Association grant10GRNT4210049 (to J.A.), NIDCR grant DE022559 (to J.A. and J.A.L.),NSF grant EFRI-1137186 (to H.K.), and NIDCR grant DE022690 (to A.P.-F.). A.A.-R. was supported by an NIDCR training grant in oral sciences(T90 DE021985) and by NHLBI grant F31HL124952.

We thank J. Fantuzzo, J. Roger, and B. Bezerra for kindly providing theextracted teeth for ex vivo studies.

We declare no competing financial interests.

REFERENCES1. Nakano K, Nomura R, Matsumoto M, Ooshima T. 2010. Roles of oral

bacteria in cardiovascular diseases—from molecular mechanisms to clin-ical cases: cell-surface structures of novel serotype k Streptococcus mutansstrains and their correlation to virulence. J Pharmacol Sci 113:120 –125.http://dx.doi.org/10.1254/jphs.09R24FM.

2. Burne RA, Ahn SJ, Wen ZT, Zeng L, Lemos JA, Abranches J, Nasci-mento M. 2009. Opportunities for disrupting cariogenic biofilms. AdvDent Res 21:17–20. http://dx.doi.org/10.1177/0895937409335593.

3. Lemos JA, Burne RA. 2008. A model of efficiency: stress tolerance byStreptococcus mutans. Microbiology 154:3247–3255. http://dx.doi.org/10.1099/mic.0.2008/023770-0.

4. Bowen WH, Koo H. 2011. Biology of Streptococcus mutans-derived glu-cosyltransferases: role in extracellular matrix formation of cariogenic bio-films. Caries Res 45:69 – 86. http://dx.doi.org/10.1159/000324598.

5. Banas JA, Vickerman MM. 2003. Glucan-binding proteins of the oralstreptococci. Crit Rev Oral Biol Med 14:89 –99. http://dx.doi.org/10.1177/154411130301400203.

6. Banas JA. 2004. Virulence properties of Streptococcus mutans. Front Bio-sci 9:1267–1277. http://dx.doi.org/10.2741/1305.

7. Koo H, Falsetta ML, Klein MI. 2013. The exopolysaccharide matrix: avirulence determinant of cariogenic biofilm. J Dent Res 92:1065–1073.http://dx.doi.org/10.1177/0022034513504218.

8. Xiao J, Klein MI, Falsetta ML, Lu B, Delahunty CM, Yates JR, III,Heydorn A, Koo H. 2012. The exopolysaccharide matrix modulates theinteraction between 3D architecture and virulence of a mixed-species oralbiofilm. PLoS Pathog 8:e1002623. http://dx.doi.org/10.1371/journal.ppat.1002623.

9. Klein MI, Xiao J, Heydorn A, Koo H. 2011. An analytical tool-box forcomprehensive biochemical, structural and transcriptome evaluation oforal biofilms mediated by mutans streptococci. J Vis Exp 2011:2512. http://dx.doi.org/10.3791/2512.

10. Love RM, McMillan MD, Jenkinson HF. 1997. Invasion of dentinaltubules by oral streptococci is associated with collagen recognition medi-ated by the antigen I/II family of polypeptides. Infect Immun 65:5157–5164.

FIG 5 Proposed mechanism for enhancement of S. mutans infection by col-lagen-binding adhesins. (A and B) Dentin-level lesions (A) or gingival reces-sion (B) exposes collagen-rich tissues to stringent adhesion by S. mutansstrains expressing CBPs (gray). The ability to invade the oral epithelium pro-vides an additional colonization reservoir. (C) In situ colonization and acidi-fication deepen existing lesions and/or form subgingival lesions (root caries),exacerbating existing tooth damage and causing injury at new sites. (D) Dentalprocedures, trauma, and poor oral hygiene provide S. mutans access to thebloodstream, where CBPs act as potential virulence factors for extraoral infec-tions and pathogenesis.

Miller et al.

2008 iai.asm.org May 2015 Volume 83 Number 5Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

11. Nobbs AH, Lamont RJ, Jenkinson HF. 2009. Streptococcus adherence andcolonization. Microbiol Mol Biol Rev 73:407– 450. http://dx.doi.org/10.1128/MMBR.00014-09.

12. Han TK, Zhang C, Dao ML. 2006. Identification and characterization ofcollagen-binding activity in Streptococcus mutans wall-associated protein: apossible implication in dental root caries and endocarditis. Biochem BiophysRes Commun 343:787–792. http://dx.doi.org/10.1016/j.bbrc.2006.03.025.

13. Switalski LM, Butcher WG, Caufield PC, Lantz MS. 1993. Collagenmediates adhesion of Streptococcus mutans to human dentin. Infect Im-mun 61:4119 – 4125.

14. Bowen WH. 2013. The Stephan Curve revisited. Odontology 101:2– 8.http://dx.doi.org/10.1007/s10266-012-0092-z.

15. Dung TZ, Liu AH. 1999. Molecular pathogenesis of root dentin caries.Oral Dis 5:92–99.

16. Hashimoto K, Sato T, Shimauchi H, Takahashi N. 2011. Profiling ofdental plaque microflora on root caries lesions and the protein-denaturingactivity of these bacteria. Am J Dent 24:295–299.

17. Ho SP, Marshall SJ, Ryder MI, Marshall GW. 2007. The tooth attach-ment mechanism defined by structure, chemical composition and me-chanical properties of collagen fibers in the periodontium. Biomaterials28:5238 –5245. http://dx.doi.org/10.1016/j.biomaterials.2007.08.031.

18. Love RM, Jenkinson HF. 2002. Invasion of dentinal tubules by oralbacteria. Crit Rev Oral Biol Med 13:171–183. http://dx.doi.org/10.1177/154411130201300207.

19. Liu T, Gibbons RJ, Hay DI. 1990. Streptococcus cricetus and Streptococcusrattus bind to different segments of collagen molecules. Oral MicrobiolImmunol 5:143–148. http://dx.doi.org/10.1111/j.1399-302X.1990.tb00412.x.

20. Liu T, Gibbons RJ. 1990. Binding of streptococci of the “mutans” groupto type 1 collagen associated with apatitic surfaces. Oral Microbiol Immu-nol 5:131–136. http://dx.doi.org/10.1111/j.1399-302X.1990.tb00410.x.

21. Avilés-Reyes A, Miller JH, Simpson-Haidaris PJ, Lemos JA, AbranchesJ. 2014. Cnm is a major virulence factor of invasive Streptococcus mutansand part of a conserved three-gene locus. Mol Oral Microbiol 29:11–23.http://dx.doi.org/10.1111/mom.12041.

22. Nomura R, Naka S, Nemoto H, Otsugu M, Nakamura S, Ooshima T,Nakano K. 2013. Potential high virulence for infective endocarditis inStreptococcus mutans strains with collagen-binding proteins but lackingPA expression. Arch Oral Biol 58:1627–1634. http://dx.doi.org/10.1016/j.archoralbio.2013.06.008.

23. Nakano K, Nomura R, Taniguchi N, Lapirattanakul J, Kojima A, NakaS, Senawongse P, Srisatjaluk R, Gronroos L, Alaluusua S, MatsumotoM, Ooshima T. 2010. Molecular characterization of Streptococcus mutansstrains containing the cnm gene encoding a collagen-binding adhesin.Arch Oral Biol 55:34 –39. http://dx.doi.org/10.1016/j.archoralbio.2009.11.008.

24. Nomura R, Nakano K, Naka S, Nemoto H, Masuda K, LapirattanakulJ, Alaluusua S, Matsumoto M, Kawabata S, Ooshima T. 2012. Identifi-cation and characterization of a collagen-binding protein, Cbm, in Strep-tococcus mutans. Mol Oral Microbiol 27:308 –323. http://dx.doi.org/10.1111/j.2041-1014.2012.00649.x.

25. Nomura R, Nakano K, Taniguchi N, Lapirattanakul J, Nemoto H,Gronroos L, Alaluusua S, Ooshima T. 2009. Molecular and clinicalanalyses of the gene encoding the collagen-binding adhesin of Streptococ-cus mutans. J Med Microbiol 58:469 – 475. http://dx.doi.org/10.1099/jmm.0.007559-0.

26. Abranches J, Miller JH, Martinez AR, Simpson-Haidaris PJ, Burne RA,Lemos JA. 2011. The collagen-binding protein Cnm is required for Strep-tococcus mutans adherence to and intracellular invasion of human coro-nary artery endothelial cells. Infect Immun 79:2277–2284. http://dx.doi.org/10.1128/IAI.00767-10.

27. Nomura R, Naka S, Nemoto H, Inagaki S, Taniguchi K, Ooshima T,Nakano K. 2013. Potential involvement of collagen-binding proteins ofStreptococcus mutans in infective endocarditis. Oral Dis 19:387–393. http://dx.doi.org/10.1111/odi.12016.

28. Kesavalu L, Lucas AR, Verma RK, Liu L, Dai E, Sampson E, Progulske-Fox A. 2012. Increased atherogenesis during Streptococcus mutans infec-tion in ApoE-null mice. J Dent Res 91:255–260. http://dx.doi.org/10.1177/0022034511435101.

29. Nomura R, Otsugu M, Naka S, Teramoto N, Kojima A, Muranaka Y,Matsumoto-Nakano M, Ooshima T, Nakano K. 2014. Contribution ofthe interaction of Streptococcus mutans serotype k strains with fibrinogento the pathogenicity of infective endocarditis. Infect Immun 82:5223–5234. http://dx.doi.org/10.1128/IAI.02164-14.

30. Wen ZT, Burne RA. 2001. Construction of a new integration vector foruse in Streptococcus mutans. Plasmid 45:31–36. http://dx.doi.org/10.1006/plas.2000.1498.

31. Zeng L, Burne RA. 2009. Transcriptional regulation of the cellobioseoperon of Streptococcus mutans. J Bacteriol 191:2153–2162. http://dx.doi.org/10.1128/JB.01641-08.

32. Dorn BR, Dunn WA, Jr, Progulske-Fox A. 1999. Invasion of humancoronary artery cells by periodontal pathogens. Infect Immun 67:5792–5798.

33. Abranches J, Zeng L, Belanger M, Rodrigues PH, Simpson-Haidaris PJ,Akin D, Dunn WA, Jr, Progulske-Fox A, Burne RA. 2009. Invasion ofhuman coronary artery endothelial cells by Streptococcus mutansOMZ175. Oral Microbiol Immunol 24:141–145. http://dx.doi.org/10.1111/j.1399-302X.2008.00487.x.

34. Lemos JA, Abranches J, Koo H, Marquis RE, Burne RA. 2010. Protocolsto study the physiology of oral biofilms. Methods Mol Biol 666:87–102.http://dx.doi.org/10.1007/978-1-60761-820-1_7.

35. Gregoire S, Xiao J, Silva BB, Gonzalez I, Agidi PS, Klein MI, Ambati-pudi KS, Rosalen PL, Bauserman R, Waugh RE, Koo H. 2011. Role ofglucosyltransferase B in interactions of Candida albicans with Streptococ-cus mutans and with an experimental pellicle on hydroxyapatite surfaces.Appl Environ Microbiol 77:6357– 6367. http://dx.doi.org/10.1128/AEM.05203-11.

36. Koo H, Schobel B, Scott-Anne K, Watson G, Bowen WH, Cury JA,Rosalen PL, Park YK. 2005. Apigenin and tt-farnesol with fluoride effectson S. mutans biofilms and dental caries. J Dent Res 84:1016 –1020. http://dx.doi.org/10.1177/154405910508401109.

37. Bowen WH, Madison KM, Pearson SK. 1988. Influence of desalivation inrats on incidence of caries in intact cagemates. J Dent Res 67:1316 –1318.http://dx.doi.org/10.1177/00220345880670101401.

38. Yamashita Y, Bowen WH, Burne RA, Kuramitsu HK. 1993. Role of theStreptococcus mutans gtf genes in caries induction in the specific-pathogen-free rat model. Infect Immun 61:3811–3817.

39. Fozo EM, Scott-Anne K, Koo H, Quivey RG, Jr. 2007. Role of unsatu-rated fatty acid biosynthesis in virulence of Streptococcus mutans. InfectImmun 75:1537–1539. http://dx.doi.org/10.1128/IAI.01938-06.

40. Raubertas RF, Davis BA, Bowen WH, Pearson SK, Watson GE. 1999.Litter effects on caries in rats and implications for experimental design.Caries Res 33:164 –169. http://dx.doi.org/10.1159/000016511.

41. Larson RH. 1981. Merits and modifications of scoring rat dental caries byKeyes’ method, p 195–203. In Tanzer JM (ed), Animal models in cariol-ogy. Information Retrieval, Inc, Washington, DC.

42. Keyes PH. 1958. Dental caries in the molar teeth of rats. II. A method fordiagnosing and scoring several types of lesions simultaneously. J Dent Res37:1088 –1099.

43. Belli WA, Marquis RE. 1991. Adaptation of Streptococcus mutans andEnterococcus hirae to acid stress in continuous culture. Appl Environ Mi-crobiol 57:1134 –1138.

44. Koo H, Hayacibara MF, Schobel BD, Cury JA, Rosalen PL, Park YK,Vacca-Smith AM, Bowen WH. 2003. Inhibition of Streptococcus mutansbiofilm accumulation and polysaccharide production by apigenin and tt-farnesol. J Antimicrob Chemother 52:782–789. http://dx.doi.org/10.1093/jac/dkg449.

45. Avilés-Reyes A, Miller JH, Simpson-Haidaris PJ, Hagen FK, AbranchesJ, Lemos JA. 2014. Modification of Streptococcus mutans Cnm by PgfScontributes to adhesion, endothelial cell invasion, and virulence. J Bacte-riol 196:2789 –2797. http://dx.doi.org/10.1128/JB.01783-14.

46. Linde A. 1989. Dentin matrix proteins: composition and possible func-tions in calcification. Anat Rec 224:154 –166. http://dx.doi.org/10.1002/ar.1092240206.

47. Patti JM, Allen BL, McGavin MJ, Hook M. 1994. MSCRAMM-mediatedadherence of microorganisms to host tissues. Annu Rev Microbiol 48:585– 617. http://dx.doi.org/10.1146/annurev.mi.48.100194.003101.

48. Chagnot C, Listrat A, Astruc T, Desvaux M. 2012. Bacterial adhesion toanimal tissues: protein determinants for recognition of extracellular ma-trix components. Cell Microbiol 14:1687–1696. http://dx.doi.org/10.1111/cmi.12002.

49. Kojima A, Nakano K, Wada K, Takahashi H, Katayama K, Yoneda M,Higurashi T, Nomura R, Hokamura K, Muranaka Y, Matsuhashi N,Umemura K, Kamisaki Y, Nakajima A, Ooshima T. 2012. Infection ofspecific strains of Streptococcus mutans, oral bacteria, confers a risk ofulcerative colitis. Sci Rep 2:332. http://dx.doi.org/10.1038/srep00332.

50. Rudney JD, Chen R, Zhang G. 2005. Streptococci dominate the diverse

Cnm Facilitates S. mutans Oral Colonization

May 2015 Volume 83 Number 5 iai.asm.org 2009Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from

flora within buccal cells. J Dent Res 84:1165–1171. http://dx.doi.org/10.1177/154405910508401214.

51. Berlutti F, Catizone A, Ricci G, Frioni A, Natalizi T, Valenti P, PolimeniA. 2010. Streptococcus mutans and Streptococcus sobrinus are able to adhereand invade human gingival fibroblast cell line. Int J Immunopathol Phar-macol 23:1253–1260.

52. Spatafora G, Rohrer K, Barnard D, Michalek S. 1995. A Streptococcusmutans mutant that synthesizes elevated levels of intracellular polysaccha-ride is hypercariogenic in vivo. Infect Immun 63:2556 –2563.

53. Khoo G, Zhan L, Hoover C, Featherstone JD. 2005. Cariogenic virulencecharacteristics of mutans streptococci isolated from caries-active and car-ies-free adults. J Calif Dent Assoc 33:973–980.

54. Palmer SR, Miller JH, Abranches J, Zeng L, Lefebure T, Richards VP,Lemos JA, Stanhope MJ, Burne RA. 2013. Phenotypic heterogeneity of

genomically-diverse isolates of Streptococcus mutans. PLoS One 8:e61358.http://dx.doi.org/10.1371/journal.pone.0061358.

55. Lee SF, Boran TL. 2003. Roles of sortase in surface expression of the majorprotein adhesin P1, saliva-induced aggregation and adherence, and cario-genicity of Streptococcus mutans. Infect Immun 71:676 – 681. http://dx.doi.org/10.1128/IAI.71.2.676-681.2003.

56. Crowley PJ, Brady LJ, Piacentini DA, Bleiweis AS. 1993. Identification ofa salivary agglutinin-binding domain within cell surface adhesin P1 ofStreptococcus mutans. Infect Immun 61:1547–1552.

57. Cornejo OE, Lefebure T, Bitar PD, Lang P, Richards VP, Eilertson K,Do T, Beighton D, Zeng L, Ahn SJ, Burne RA, Siepel A, BustamanteCD, Stanhope MJ. 2013. Evolutionary and population genomics of thecavity causing bacteria Streptococcus mutans. Mol Biol Evol 30:881– 893.http://dx.doi.org/10.1093/molbev/mss278.

Miller et al.

2010 iai.asm.org May 2015 Volume 83 Number 5Infection and Immunity

on February 3, 2018 by guest

http://iai.asm.org/

Dow

nloaded from