JOURNAL OF BACTERIOLOGY, Dec. 2008, p. 8145–8154 0021-9193/08/$08.00⫹0 doi:10.1128/JB.00983-08 Copyright © 2008, American Society for Microbiology. All Rights Reserved.
Vol. 190, No. 24
Characterization of a Streptococcus sp.-Veillonella sp. Community Micromanipulated from Dental Plaque䌤 Natalia I. Chalmers,1,2 Robert J. Palmer, Jr.,2 John O. Cisar,2 and Paul E. Kolenbrander2* Department of Biomedical Sciences, University of Maryland Dental School, Baltimore, Maryland 21201,1 and National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, Maryland 208922 Received 16 July 2008/Accepted 8 September 2008
Streptococci and veillonellae occur in mixed-species colonies during formation of early dental plaque. One factor hypothesized to be important in assembly of these initial communities is coaggregation (cell-cell recognition by genetically distinct bacteria). Intrageneric coaggregation of streptococci occurs when a lectinlike adhesin on one streptococcal species recognizes a receptor polysaccharide (RPS) on the partner species. Veillonellae also coaggregate with streptococci. These genera interact metabolically; lactic acid produced by streptococci is a carbon source for veillonellae. To transpose these interactions from undisturbed dental plaque to an experimentally tractable in vitro biofilm model, a community consisting of RPS-bearing streptococci juxtaposed with veillonellae was targeted by quantum dot-based immunofluorescence and then micromanipulated off the enamel surface and cultured. Besides the expected antibody-reactive cell types, a non-antibodyreactive streptococcus invisible during micromanipulation was obtained. The streptococci were identified as Streptococcus oralis (RPS bearing) and Streptococcus gordonii (adhesin bearing). The veillonellae could not be cultivated; however, a veillonella 16S rRNA gene sequence was amplified from the original isolation mixture, and this sequence was identical to the sequence of the previously studied organism Veillonella sp. strain PK1910, an oral isolate in our culture collection. S. oralis coaggregated with S. gordonii by an RPS-dependent mechanism, and both streptococci coaggregated with PK1910, which was used as a surrogate during in vitro community reconstruction. The streptococci and strain PK1910 formed interdigitated three-species clusters when grown as a biofilm using saliva as the nutritional source. PK1910 grew only when streptococci were present. This study confirms that RPS-mediated intrageneric coaggregation occurs in the earliest stages of plaque formation by bringing bacteria together to create a functional community. surface molecule found on many strains of S. oralis and S. mitis (15). It mediates coaggregation by its role as the recognition molecule for lectinlike adhesins found on actinomyces, veillonellae, and other streptococci. Six RPS types have been identified in oral streptococci (9). Each type is composed of a distinct hexa- or heptasaccharide repeating unit which contains one of two host-like disaccharide recognition motifs, GalNAc1-3Gal (Gn type) or Gal1-3GalNAc (G type). The lectin-like adhesins on actinomyces (8) and on veillonellae (16) recognize the Gn and G types of RPS, whereas certain streptococci bear GalNAc-specific adhesins that recognize only the Gn types (9). Intergeneric coaggregation of RPS-bearing streptococci and actinomyces (9) or veillonellae (17) is prevalent and is thought to contribute to the formation of pioneer multispecies communities on enamel (30, 31). Importantly, widespread intrageneric coaggregation of streptococci has been postulated to be a major factor in initial multispecies community formation (19), and such coaggregation is consistent with the hypothesis that streptococci are the dominant initial colonizers (12, 29). Although the species diversity of initial plaque (12), as well as that of mature plaque (1), has been described using molecular phylogenetics, this information does not reveal spatial relationships between species within communities. A retrievable enamel chip model (32) has been used to examine spatial relationships in initial, undisturbed, human plaque communities. In a fluorescence in situ hybridization (FISH) study using this model, streptococci were shown to be part of small communities that also contained nonstreptococcal cells (12). Im-
Dental plaque is a multispecies biofilm whose development is initiated by adherence of pioneer species to the salivary proteins and glycoproteins adsorbed on tooth enamel. Although more than 700 phylotypes have been detected in the human oral cavity, fewer than 100 phylotypes are found in a typical individual (1). The biofilm is not formed by random simultaneous colonization by these species; selective, reproducible, sequential colonization occurs (12, 29). The initial colonizers are a specific subset of the oral microflora, and Actinomyces, Neisseria, Prevotella, Streptococcus, and Veillonella predominate (12, 29). Streptococci constitute 63% of the culturable bacteria after 4 h of plaque formation (29) and account for 66% of 16S rRNA gene sequences cloned from 4-h plaque samples (12). The vast majority of the streptococcal sequences belong to the Streptococcus oralis-Streptococcus mitis cluster (12). Secondary colonizers, such as fusobacteria and capnocytophagae, coaggregate with pioneer species (18) and add to the multispecies transitions in the repetitive developmental process. Coaggregation, defined as cell-cell recognition and binding between genetically distinct bacteria, is characteristic of oral bacteria and has been postulated to play a role in biofilm development (18, 20). Receptor polysaccharide (RPS) is a cell
* Corresponding author. Mailing address: National Institutes of Health/NIDCR, Building 30, Room 310, 30 Convent Drive, MSC 4350, Bethesda, MD 20892-4350. Phone: (301) 496-1497. Fax: (301) 4020396. E-mail:
[email protected]. 䌤 Published ahead of print on 19 September 2008. 8145
8146
CHALMERS ET AL.
munofluorescence was used to reveal veillonellae juxtaposed with RPS-bearing streptococci (30). A study using immunofluorescence and nucleic acid stains (31) identified RPS-bearing streptococci juxtaposed with streptococci that lacked RPS and also revealed type-2-fimbria-bearing actinomyces juxtaposed with RPS-bearing streptococci. The latter juxtaposed pair, in which a cell bearing a specific coaggregation-mediating adhesin was juxtaposed with a cell bearing the complementary receptor molecule, provided strong evidence for the hypothesis that intergeneric coaggregation has a function in the assembly of biofilms in nature (31). However, while there is much evidence demonstrating that coaggregation has a role in plaque development, definitive proof requires isolation and subsequent culture of juxtaposed cells and reassembly of the cultured cells into physically and metabolically integrated communities in vitro. Veillonellae and streptococci have been postulated to be linked metabolically through streptococcal fermentation of sugars to lactic acid, which is a carbon source for the nonsaccharolytic veillonellae. In vivo studies using gnotobiotic rats demonstrated that veillonellae were unable to establish monoinfections, yet when a strain of Veillonella was inoculated into rats already monoinfected with a strain of Streptococcus mutans that coaggregates with that Veillonella strain, the number of veillonellae on the teeth of the coinfected animals was 1,000-fold higher than the number when a noncoaggregating Veillonella strain was used (25). Also in gnotobiotic rats, lower caries and plaque scores were obtained for two-species biofilms than for monospecies colonization by streptococci (41), and veillonellae have been shown to reduce caries activity and demineralization of the enamel surface by streptococci (26, 27). More recently, spatial relationships between these species have been reported to influence gene regulation in vitro; diffusible-signal exchange between the coaggregating partners Veillonella sp. strain PK1910 and Streptococcus gordonii V288 resulted in upregulation of an amylase gene (amyB) promoter in the streptococcus strain (13). Further, it has been shown that veillonellae are close to RPS-bearing streptococci in initial communities in vivo and that a rapid succession of veillonella phylotypes occurs in the communities (30). Because initial dental plaque communities are often composed of just a few cells of different species, a community containing RPS-bearing streptococci juxtaposed with veillonellae might consist of only coaggregating species. Furthermore, the cells might be able to form mixed-species biofilms in an in vitro model using saliva as the sole carbon source. Verification of these hypotheses would conclusively demonstrate that coaggregation has a role in establishment of initial dental plaque communities. MATERIALS AND METHODS Micromanipulation of an initial in vivo community. An 8-h-old plaque sample was obtained by using the retrievable enamel chip model (31, 32). Briefly, small chips of human enamel were carried in a mandibular stent in a volunteer’s mouth for 8 h, after which they were removed and stained in a disinfected (70% ethanol) chamber. All staining solutions were filter sterilized. Quantum dot (QD)-labeled primary antibodies were used at a concentration of 30 nM (6) to select communities for manipulation. Anti-RPS, which reacts with a subset of RPS-bearing streptococci that includes representatives of the G and Gn structural types (31), was conjugated to QD 655 (Invitrogen, Carlsbad, CA) for micromanipulation. Anti-R1 (30) reacted with almost all culturable veillonellae from the volunteer’s
J. BACTERIOL. mouth. This antibody was conjugated to QD 525 for micromanipulation. For some samples not destined for micromanipulation, 4⬘,6-diamidino-2-phenylindole dihydrochloride (DAPI) (Invitrogen) was applied at a concentration of 5 g/ml for detection of non-antibody-reactive cells. After staining, a chip was attached to a microscope slide using dental wax, sterile water was applied as an immersion fluid, and the biofilm was examined with a 63x 0.9 NA water-immersible lens that was wiped with 70% ethanol and was mounted on a DM LB2 upright microscope (Leica, Bannockburn, IL). A community that contained both of the antibody-reactive cell types (i.e., at least one anti-RPS-reactive cell together with at least one anti-R1-reactive cell) was identified, and dual TransferMan NK2 micromanipulators (Eppendorf, Westbury, NY) equipped with ethanol flame-sterilized microneedles or microspades (tip diameter, 25 m; Minitool, Los Gatos, CA) were used to transfer the community to anaerobic modified Schaedler’s medium (MSM) (4) in which lactic acid (21 ml of 60% lactic acid syrup/liter) was substituted for glucose. Identification of community members. After 48 h of anaerobic growth in MSM at 37°C with an H2-CO2-N2 (5:5:90) atmosphere (Bactron glovebox; Sheldon Manufacturing, Cornelius, OR), slight turbidity was observed. This enrichment culture was concentrated fivefold, made 20% with respect to glycerol, and then frozen at ⫺70°C in aliquots, which were regrown in fresh MSM broth for further work. Members of the enrichment culture were then isolated by serial dilution onto MSM agar plates, some of which contained vancomycin (7.5 g/ml). No growth was obtained on the veillonella-selective vancomycin-containing plates. However, anti-R1-reactive cells were always present in the enrichment culture, and PCR using forward primer A(C/T)CAACCTGCCCTTCAGA) and reverse primer CGTCCCGATTAACAGAGCTT targeting the 16S rRNA gene of veillonellae (34) also verified the presence of Veillonella sp. cells in the enrichment. Colonies were picked from MSM plates without vancomycin and were screened for RPS by dot immunoblotting (43). Membranes were spotted by hand with 0.7 l of an overnight bacterial culture and incubated with a primary antibody mixture that identified all RPS-bearing bacteria, and the RPS-bearing strains revealed by using horseradish peroxidase-conjugated secondary antibody. The RPS-positive isolates were then reblotted and screened using single antibodies to characterize the specific structural type of RPS on each isolate (43). All isolates (RPS positive and RPS negative) were subjected to repetitive extragenic palindromic PCR (REP-PCR) analysis to examine clonality (2). DNA was extracted from 5 l of overnight culture with GeneReleaser (BioVentures, Inc., Murfreesboro, TN), amplification was performed using the JumpStart ReadyMix REDTaq PCR mixture (Sigma, St. Louis, MO), and the initial denaturation at 95°C was for 2 min. The primer sequences were as follows: REP1RDt, IIINCGNCGNCATCNGCC; and REP2-Dt, NCGNCTTATCNGGCCTAC. The PCR products were separated by agarose gel electrophoresis. The phylogenetic relationships of the clones with other streptococci were examined using superoxide dismutase (sodA) gene sequences (14). The primer sequences were as follows: forward primer, TRCAYCATGAYAARCACCAT; and reverse primer, ARRTARTAMGCRTGYTCCCARACRTC. MEGA version 4 (38) was used to construct a ClustalW alignment of the sodA sequences, and a tree was constructed using a neighbor-joining algorithm (37). Spatial relationship of RPS-bearing streptococci, other streptococci, and veillonellae in vivo. A protocol for simultaneous use of FISH and immunofluorescence was developed and used for biofilms on chips. Samples were labeled with Alexa Fluor 546-conjugated anti-RPS at a concentration of 5 g/ml for 20 min, washed with 1% phosphate-buffered saline (PBS)-bovine serum albumin (BSA), and then fixed at 4°C for 3 h with 4% paraformaldehyde in PBS. FISH was then carried out as previously described (12) by using the genus-level veillonella FISH probe VEI488 (CCGTGGCTTTCTATTCCG) designed with ARB software (21) or by using the genus-level streptococcal FISH probe STR405 (39). FISH probes were synthesized and labeled by Operon Biotechnologies, Inc. (Huntsville, AL). The specificity of VEI488 was tested, and this probe was shown to hybridize to Veillonella clinical isolates R1 and R2 (30), Veillonella sp. strain PK1910, Veillonella parvula ATCC 10790, and Veillonella atypica ATCC 17744. The negative controls used for VEI488 were S. gordonii DL1, S. oralis 34, S. mitis ATCC 49456, S. mutans ATCC 700610, S. oralis ATCC 10557, Streptococcus sanguinis ATCC 10556, S. gordonii ATCC 49818, Streptococcus salivarius ATCC 259750, Actinomyces naeslundii T14V, Fusobacterium nucleatum ATCC 10953, Prevotella intermedia ATCC 15032, and Porphyromonas gingivalis ATCC 53978. Probe VEI488 was tested with all negative controls and was shown not to hybridize to any of them. Reconstruction of the community in vitro. (i) Growth of biofilms on polystyrene pegs. Biofilms were grown in 25% human saliva on transferable solid-phase polystyrene pegs (Nunc 445497; Nunc-Immuno TSP) (5, 24) mounted in U96
VOL. 190, 2008
COAGGREGATION IN MICROMANIPULATED COMMUNITY
MicroWell plates (Nunc 163320) (24). Overnight cultures of the two Streptococcus isolates were grown in brain heart infusion broth (Difco, Detroit, MI). Veillonella sp. strain PK1910 was chosen as a surrogate for the uncultivated Veillonella sp. in the community because its 16S rRNA gene sequence is identical to that retrieved from the community. Overnight cultures of Veillonella sp. strain PK1910 were grown in MSM broth. Microtiter plate wells were filled with 200 l of 25% human saliva, and the pegs were then inserted and incubated for 30 min at room temperature to obtain a conditioning film. Twenty microliters of an overnight culture was added to each of the wells to obtain an optical density at 600 nm of approximately 0.1. The plates were placed in a humidity chamber and incubated anaerobically at 37°C for 24 or 48 h. The pegs that were incubated for 48 h were transferred to fresh reduced 25% saliva after 24 h. In preliminary experiments, total biomass was quantified by crystal violet staining; the transferable solid-phase unit was removed, air dried for 30 min at room temperature, stained with 200 l of 0.2% (wt/vol) crystal violet (Sigma), washed twice with deionized water, and then dried. The stain was eluted in 70% ethanol–5% acetic acid, and the absorbance at 540 nm of the elution wash solution was determined using a Victor3 plate reader (PerkinElmer, Inc., Waltham, MA). (ii) Real-time Q-PCR quantification of species in biofilms. DNA was extracted from biofilms by a modified alkaline lysis protocol (14). Biofilm-covered pegs were immersed in 40 l of sterile ultrapure water plus 160 l of 0.05 M sodium hydroxide and incubated at 60°C for 45 to 60 min, after which 18.4 l of 1 M Tris-HCl (pH 7.0) was added to neutralize the pH. The resulting extract was used as the template DNA for the quantitative PCR (Q-PCR) analyses (14). Bacterial genomic DNA used to obtain standard curves was extracted from overnight cultures of the clinical isolate of S. gordonii and Veillonella sp. strain PK1910 with a DNA extraction kit (Qiagen) used according to the manufacturer’s instructions. Genomic DNA was stored at ⫺20°C. Species-specific primers used for quantification were designed with AlleleID6 (PREMIER Biosoft International, Palo Alto, CA). The primers specific for streptococci were forward primer CGACGATACATAGCCGACCTGAG and reverse primer TCCATTGCCGAAGATTCCCTACTG, and the annealing temperature was 60°C. The primers specific for veillonellae were forward primer CCGTGATGGGATGGAAACTGC and reverse primer CCTTCGCCACTGG TGTTCTTC, and the annealing temperature was 60°C. Streptococci and veillonellae in the biofilms were quantified by performing real-time Q-PCR with the SYBR green dye to detect the 16S rRNA gene amplicons. Each reaction mixture (final volume, 20 l) contained 3 l template, 10 l FAST Power SYBR green PCR Master Mix (Applied Biosystems, Foster City, CA), 375 nM forward primer, and 375 nM reverse primer. The Q-PCR was performed with an MX3005P thermocycler (Stratagene, La Jolla, CA) using the thermocycling conditions recommended for FAST Power SYBR green PCR Master Mix (95°C for 20 s and 40 cycles of 3 s at 95°C and 30 s at 60°C). Dissociation curves were generated by incubating reaction products at 95°C for 1 min and at 56°C for 30 s and then incrementally increasing the temperature to 95°C. Fluorescence data were collected at the end of the 60°C primer annealing step for 40 amplification cycles and throughout the dissociation curve analysis. Analysis of the melting curves with both primer sets revealed a single sharp peak. DNA concentrations (ng/ml) were calculated based on standard curves obtained by using 10-fold serial dilutions of bacterial DNA isolated with a DNA extraction kit (Qiagen) and quantified using the PicoGreen fluorescence assay (Invitrogen). To convert nanograms of DNA to numbers of cells, the following weights and genome sizes were used: 2.05 fg/genome and 2 Mb for streptococci (42) and 3.08 fg/genome and 3 Mb for veillonellae (23). The data presented below were obtained for three independent biofilms. (iii) Labeling of peg biofilms and microscopy. Anti-RPS conjugated to Alexa Fluor 546 was used to identify S. oralis, anti-DL1 (31) conjugated to Alexa Fluor 488 was used to identify S. gordonii (which lacks RPS), and anti-1910 (30) conjugated to Alexa Fluor 633 was used to identify Veillonella sp. strain PK1910. Antibodies (5 g/ml) were applied for 20 min to peg biofilms immersed in PBS-BSA. The biofilms were then washed twice with 1% PBS-BSA after transfer to new microtiter plates. The pegs were then cut out and attached with dental wax to a microscope slide. Confocal microscopy was performed with a TCS SP2 confocal microscope (Leica Microsystems, Exton, Pa.) using a 63x 0.9NA LWD water-immersible lens. Nucleotide sequence accession numbers. The sodA sequences of the two Streptococcus isolates have been deposited in the GenBank database under accession numbers EU488871 and EU488872. The 343-bp sequence obtained with Veillonella-specific 16S rRNA gene primers has been deposited in the GenBank database under accession number EU488873.
8147
FIG. 1. Confocal micrographs of 8-h dental plaque. (A) QD-based primary immunofluorescence revealing RPS-bearing streptococci reactive with QD655-conjugated anti-RPS (red) juxtaposed with veillonellae reactive with QD525-conjugated anti-R1 (green). A community representative of the cells selected for micromanipulation is circled. (B) Same field of view as that in panel A but with DAPIstained cells (blue) also shown. The general nucleic acid stain DAPI revealed non-antibody-reactive cells, one of which was located in the representative community. DAPI was not used with micromanipulated samples. Bar, 10 m.
RESULTS Micromanipulation of an initial community from undisturbed plaque. The initial communities used for micromanipulation were selected based on the presence of cells reactive with anti-RPS juxtaposed with cells reactive with anti-R1. Figure 1A shows what was seen during targeting and manipulation of the community (small numbers of juxtaposed cells with at least one cell reactive with each antibody). DAPI staining (Fig. 1B) showed that non-antibody-reactive cells can also occur in such communities; however, nucleic acid-binding stains were not used during manipulation to minimize photodamage. Four communities from three independent biofilms were micromanipulated. Three micromanipulated communities were transferred to reduced MSM broth and incubated anaerobically. After 48 h the medium became slightly turbid. For each outgrowth the presence of anti-RPS-reactive and anti-R1-reactive cells was determined by using primary immunofluorescence. One outgrowth containing both cell types was studied further. Serial dilutions of the outgrowth were plated onto MSM with and without vancomycin and incubated anaerobically for 48 h. Growth occurred only on MSM without vancomycin, and 160 single colonies were retrieved for further study. The fourth micromanipulated community was plated directly onto MSM agar, and no growth was visible after 72 h of anaerobic incubation. Characterization of streptococci. The 160 isolates were screened by using a cocktail of anti-RPS antibodies that recognizes all types of RPS (43), after which the isolates were separated into two groups: an RPS-positive group (41 isolates)
8148
CHALMERS ET AL.
FIG. 2. REP-PCR patterns of four randomly selected RPS-bearing Streptococcus isolates (lanes 2 to 5) and four randomly selected RPSnegative Streptococcus isolates (lanes 6 to 9). Lanes 1 and 10 contained 1-kb molecular size markers.
and an RPS-negative group (119 isolates). Subsequently, the 41 RPS-positive isolates were tested with individual antibodies against each of the four recognized serotypes (9, 43) All 41 isolates reacted with antibody specific for RPS serotype 1. These isolates also coaggregated with S. gordonii DL1, indicating the presence of Gn-type RPS. Based on morphology and antibody reactivity, all 41 RPS-positive isolates were presumed to be streptococci. 16S rRNA gene sequence analysis of 10 randomly selected RPS-bearing isolates, as well as 10 randomly selected RPS-negative isolates, showed that these isolates were streptococci. However, 16S rRNA gene sequences do not distinguish oral streptococci at the species level; therefore, other methods described below were used. REP-PCR was used to assess the genotypic heterogeneity in all 160 isolates. REP-PCR provides a highly reproducible multiband PCR product fingerprint for each genotype (2). All 41 RPS-bearing streptococcal isolates produced identical REP-PCR fingerprints, and all 119 RPS-negative streptococcal isolates produced a single fingerprint distinct from that of the RPS-bearing isolates (Fig. 2). These data indicate that the micromanipulated community consisted of only two streptococcal genotypes. Phylogenetic identification of streptococci at the species level was accomplished by comparing the sequences of the superoxide dismutase (sodA) genes (14) of the isolates with the sequences of the superoxide dismutase genes of other nonhemolytic streptococci (Fig. 3). Based on sequencing results, the RPS-bearing Streptococcus spp. clustered with S. oralis, and the RPS-negative Streptococcus spp. clustered with S. gordonii (14). Oral streptococci participate in numerous types of intergeneric coaggregation, but they also exhibit extensive intrageneric coaggregation (19). The micromanipulated S. gordonii was compared with the reference strain S. gordonii DL1 to study its ability to coaggregate in vitro with a reference set of streptococcal strains bearing RPS type 1Gn, 2Gn, 4Gn, 2G, or
J. BACTERIOL.
FIG. 3. Phylogenetic tree based on streptococcal sodA sequences. The neighbor-joining method was used to construct the tree. Filled diamonds indicate the two clinical isolates. Scale bar ⫽ 5% difference in nucleotide sequence. The type strains S. oralis ATCC 35037 and S. gordonii ATCC 10558 are included for reference.
3G. S. gordonii DL1 has GalNAc-specific adhesins on its surface (9). The coaggregation of S. gordonii DL1 with the reference set of RPS-bearing streptococci was indistinguishable from the coaggregation of the micromanipulated S. gordonii strain with these RPS-bearing streptococci (data not shown), indicating that the coaggregation was Gn specific. An RPSnegative mutant of the reference strain S. oralis 34 (bearing 1Gn RPS) did not coaggregate with either S. gordonii strain, further supporting the hypothesis that Gn-specific adhesins were present on the micromanipulated S. gordonii strain. Collectively, these data documented that the micromanipulated S. oralis strain bears a 1Gn-type RPS, that the micromanipulated S. gordonii strain bears a Gn-specific adhesin, and that intrageneric coaggregation is fundamental within initial communities on enamel. Characterization of uncultured Veillonella sp. Veillonellae are typically isolated from clinical samples using selective agar based on vancomycin resistance (35, 36). After growth appeared in the original outgrowth inoculated with the micromanipulated community, serial dilutions were plated onto MSM agar with vancomycin (7.5 g/ml), but no colonies were evident after 48 to 72 h of anaerobic incubation. When vancomycin was omitted, the colonies were predominantly streptococcal colonies, but there were some mixed colonies in which anti-R1reactive cells were observed. Attempts to culture veillonellae from these colonies were unsuccessful. The procedures used to enrich for Veillonella cells included (i) growth on media containing preferred carbon sources other than lactate (e.g., pyruvate), (ii) plating on agar prepared with spent medium from the streptococcal clinical isolates grown in MSM, and (iii) magnetic capture using anti-R1-conjugated Dynabeads (Invitrogen). No colonies were recovered when these procedures were used. However, anti-R1-reactive cells were always detected by primary immunofluorescence in the original outgrowth of the micromanipulation-inoculated mixed culture (Fig. 4). The anti-R1-reactive cells were occasionally quite numerous and
VOL. 190, 2008
COAGGREGATION IN MICROMANIPULATED COMMUNITY
FIG. 4. Transmitted light micrograph (inset) of a wet mount of micromanipulated cells after outgrowth in an MSM broth culture and immunofluorescence microscopy (large image) of the same field of view showing cells labeled with anti-R1 antibody. The arrows indicate anti-R1-reactive cells in the two images. Note the non-antibody-reactive cells (presumed to be streptococci) in the transmitted light micrograph. Bars, 20 m.
occurred together with other spherical cells presumed to be streptococci. Molecular techniques were also used to confirm the presence of Veillonella cells in the original mixed culture. Veillonella-specific 16S rRNA gene primers (34) amplified a 343-bp sequence that is identical to the sequences of other uncultured veillonellae, including sequences from the same volunteer (12). The sequence also clustered with the sequences of other anti-R1-reactive cells that are most closely related to V. parvula (30). The same study revealed that Veillonella sp. strain PK1910, a strain in our culture collection which was identified using nonmolecular approaches and which clusters together with V. parvula strains, is also very closely related to the uncultured Veillonella sp. from the captured community. Therefore, PK1910 was selected as a surrogate veillonella strain for in vitro studies with the micromanipulated S. oralis and S. gordonii isolates. Spatial relationship between phylotypes of the micromanipulated community members in vivo. The micromanipulated community consisted of three members: an uncultured Veil-
8149
lonella sp. and two Streptococcus spp. (RPS-bearing S. oralis and RPS-negative S. gordonii). To study the spatial arrangement of these organisms in the community in vivo, immunofluorescence was combined with FISH. The FISH probe VEI488 (which recognizes all Veillonella cells) was used simultaneously with anti-RPS immunofluorescence. VEI488-reactive cells were distributed rather evenly over the enamel surface (Fig. 5A), as were the anti-RPS-reactive clusters of cells (Fig. 5B). The VEI488-reactive cells were almost exclusively close to anti-RPS-labeled cells, indicating that the distribution of veillonellae and streptococci in vivo is not random (Fig. 5C). These findings support those of a previous study in which these communities were identified by fluorescent antibody labeling of veillonellae and RPS-bearing streptococci (30) and extend the observations to include all veillonellae reactive with the VEI488 FISH probe, regardless of their antigenic reactivity. The FISH probe STR405 recognizes all Streptococcus cells (39) and was used together with anti-RPS immunofluorescence to investigate the prevalence of streptococcal communities composed of RPS-bearing cells (immunoreactive and STR405 reactive) and RPS-negative cells (only STR405 reactive) in an undisturbed dental plaque biofilm stained with acridine orange (Fig. 6A). A subset of the Streptococcus cells identified by FISH (Fig. 6B) was also anti-RPS reactive (Fig. 6C). Many anti-RPS-reactive cells were close to RPS-negative streptococci (Fig. 6C) and nonstreptococcal cells (Fig. 6C). Area analysis of multiple images of biofilms labeled with anti-RPS and STR405 using the DAIME software (11) revealed that the RPS-bearing streptococci accounted for 38% ⫾ 8% of the total streptococcal population. Reconstruction of three-species biofilms in vitro. The juxtaposition in vivo of the RPS-positive organism S. oralis, the adhesin-bearing organism S. gordonii, and the uncultured Veillonella sp. indicates that there was coaggregation-mediated colonization. Veillonella sp. strain PK1910 was chosen as a surrogate for the uncultured Veillonella sp. in the micromanipulated community because it coaggregated with the micromanipulated S. oralis and S. gordonii isolates, as well as with the reference strain S. oralis 34, but not with the RPS-negative mutant of the latter strain. Therefore, reconstruction of a three-species biofilm community in vitro was attempted. Bio-
FIG. 5. Confocal micrographs of immunofluorescence- and FISH-treated 8-h plaque on enamel showing (A) Veillonella cells reactive with the VEI488 FISH probe for veillonellae 16S rRNA (green), (B) RPS-bearing streptococci reactive with anti-S. oralis 34 RPS (red), and (C) an overlay of panels A and B showing juxtaposition of veillonellae and RPS-bearing streptococci. All images are maximum projection images. Bar, 40 m.
8150
CHALMERS ET AL.
J. BACTERIOL.
film formation by the streptococcal clinical isolates and PK1910 using saliva as the sole nutritional source was studied with a static model in which bacteria adhered to a salivaconditioned polystyrene peg suspended in a microtiter well. Veillonellae can use the metabolic products of streptococci; therefore, coculture of veillonellae with streptococci could enhance the growth of veillonellae in a multispecies biofilm. The initial results obtained using crystal violet staining of biofilms formed on the polystyrene pegs indicated that each streptococcal isolate could form a monospecies biofilm, but PK1910 could not do this (data not shown). However, culture of PK1910 together with either streptococcal isolate resulted in accumulation of a large amount of biomass. Three-species cultures showed biomass accumulation similar to that seen with two-species streptococcus cultures. Although these results supported the hypothesis that metabolic interaction existed in the multispecies biofilms, they did not quantify species biomass or reveal the spatial relationship between the organisms. To quantify veillonellae and streptococci in the biofilms, Q-PCR with species-specific primers was used to amplify part of the 16S rRNA gene. No cross amplification with Streptococcus- and Veillonella-specific primers occurred. In monospecies biofilms, S. oralis and S. gordonii formed biofilms by 24 h, and the number of cells was greater at 48 h (Fig. 7). In two-species streptococcal biofilms, the cumulative number of cells of the two streptococci was not higher than the numbers of cells when the organisms were grown as monocultures. In monoculture, PK1910 formed a minimal biofilm. However, in two-species biofilms with each streptococcus, the number of veillonella cells at the initial 24-h time point was higher than that in monoculture, and the number of cells increased significantly over the following 24 h. The same was true for the threespecies biofilms. These data show that each streptococcal community member can grow on its own using saliva as the sole nutrient source, whereas PK1910 cannot grow unless a streptococcal partner is present. Further, these data indicate that all members of the three-species community can grow together on saliva. Architecture of three-species biofilms in vitro. Each species was labeled using primary immunofluorescence, and the biofilms were examined using laser scanning confocal microscopy (Fig. 8). At the initial 24-h time point, all three cell types were found to be members of multispecies coaggregates, and they were not randomly distributed as single-species colonies over the peg surface, an impressive finding given that the system was
FIG. 6. Confocal micrographs of immunofluorescence- and FISHtreated 8-h plaque on enamel showing the distribution of RPS-bearing streptococci among other streptococci and nonstreptococcal bacteria. (A) All cells stained with the general nucleic acid stain acridine orange (green). (B) Streptococcus cells reactive with the 16S rRNA for streptococci appear blue-green through combination of acridine orange (green; shows all cells) with the streptococcal 16S rRNA probe (blue). (C) Streptococcus cells reactive with Alexa Fluor 546-conjugated antiRPS. RPS-bearing Streptococcus cells are red with a white center. The bright white pixels in the center of large colonies result from colocalization of red, green, and blue. All images are maximum projection images. The arrowheads indicate RPS-negative streptococci (bluegreen) that are close to RPS-bearing streptococci (red with white centers). Bar, 8 m.
VOL. 190, 2008
COAGGREGATION IN MICROMANIPULATED COMMUNITY
8151
FIG. 7. Q-PCR quantification of S. oralis, S. gordonii, and Veillonella sp. strain PK1910 in one-, two- and three-species biofilms at 24 and 48 h. For the two- and three-species biofilms, the number of streptococcal cells (S. oralis RPS-bearing isolate and S. gordonii RPS-negative isolate) is indicated by light gray bars, and the number of PK1910 cells is indicated by dark gray bars. The number of streptococcal cells increased between 24 and 48 h in the one-, two- and three-species biofilms. PK1910 did not form a single-species biofilm, but its biomass increased significantly in two- and three-species biofilms.
not a flowing model system. After an additional 24 h of growth, the sizes of the cell clusters had increased, and the majority of the clusters contained three species. The images (Fig. 8) revealed the intimate three-species interdigitation and emphasized the importance of coaggregation to the development of these communities. DISCUSSION A multispecies oral biofilm community was obtained from a retrievable human enamel surface by using a novel approach that preserves interspecies interactions. The community consisted of two streptococci (S. oralis and S. gordonii) and an uncultivated Veillonella sp. The streptococci exhibited coaggregation; the S. oralis strain had a GalNAc-containing RPS and bound to the S. gordonii strain, which had a GalNAc-specific adhesin. These findings define an important role for intrageneric coaggregation in the development of plaque communities in vivo. The streptococci also coaggregated in vitro with Veillonella sp. strain PK1910, a strain indistinguishable on the basis of the 16S rRNA gene sequence from the uncultivated Veillonella sp. of the captured community. Reconstruction of the community in vitro, using saliva as the sole carbon source and PK1910 as a surrogate for the micromanipulated Veillonella sp., demonstrated that the three organisms interacted through coaggregation to form biofilms composed of discrete interdigitated multispecies colonies whose structure was similar to that of the original community captured from the tooth surface. Furthermore, PK1910 could not grow without interaction with
at least one of the streptococci, thereby demonstrating the metabolic dependence of veillonellae on other bacteria for growth in saliva. These results support the concept that this three-species community was a fundamental building block of the initial oral biofilms. One methodological advance required for micromanipulation of the community from the enamel surface was identification of target bacteria based on criteria other than cell shape. Previously, micromanipulation of oral bacteria was based on an unusual and easily identifiable morphology: the “corn cob” consortium (28). This consortium was shown to consist of a long rod, eventually named Corynebacterium matruchotii (10), surrounded by a tufted streptococcus that was eventually classified as Streptococcus cristatus (40). The occasional isolation of a Veillonella-like bacterium was noteworthy. Subsequent to isolation, antisera against these bacteria were produced, and the juxtaposition of antibody-reactive bacteria within corncobs was confirmed by secondary immunofluorescence. However, the Veillonella-like bacterium was rarely seen and appeared only at the tip of the otherwise densely populated corncobs. The Veillonella-like bacterium was likely dependent on the other bacteria for growth. The corncobs were obtained from disrupted plaque samples that had none of the original biofilm architecture. In these micromanipulations, the sample was spread across a thin agar coating on a microscopy coverglass, which formed the upper part of a glass chamber with the agar surface facing downward. Phase-contrast light microscopy with an upright microscope was used to view the sample through the
8152
CHALMERS ET AL.
J. BACTERIOL.
FIG. 8. Representative confocal micrographs of 24-h (A and C) and 48-h (B and D) in vitro biofilms showing the intimate interaction between the RPS-bearing streptococcal isolate (S. oralis, labeled red by Alexa Fluor 546-conjugated anti-RPS), the RPS-negative streptococcal isolate (S. gordonii, labeled green by Alexa Fluor 488-conjugated anti-DL1), and the surrogate organism Veillonella sp. strain PK1910 (labeled blue by Alexa Fluor 633-conjugated anti-1910). (A and B) Distribution and juxtaposition of the three species on a peg surface after 24 h (A) and 48 h (B) of biofilm growth on saliva as the sole nutritional source. Significant growth of all species occurred at 48 h. (C and D) Three-dimensional volume renderings of the communities indicated by the squares in panels A and B, showing the interdigitation and spatial relationships of the three species. The arrowheads indicate interdigitation of the three species. Bars, 40 m.
coverglass-agar sandwich, and the consortium was manipulated off using an angled needle inserted into the chamber from the side. In the present study, intact dental plaque biofilms on the opaque substratum of human tooth enamel were examined. An upright microscope with a water immersion objective was used to view a sample without a coverglass, and the manipulators approached through the water droplet between the lens and the sample. The candidate community, RPS-bearing streptococci juxtaposed with veillonellae, was composed entirely of coccoid organisms; therefore, primary immunofluorescence was required to distinguish the target organisms from the many
other coccoid bacteria in the biofilm. The use of primary immunofluorescence to target cells required antibodies conjugated with photostable QD fluorophors (6). Several minutes were needed to locate, select, and capture the targeted community. Therefore, as envisioned in a previous study (6), photostable QD luminescence was essential for this. QDs have narrow, symmetric emission spectra, as well as broad continuous excitation (3, 7). Thus, white light epifluorescence at a single wavelength was used to simultaneously excite QD655– anti-RPS conjugates together with QD525–anti-R1 conjugates for location and manipulation of the community. Micromanip-
VOL. 190, 2008
COAGGREGATION IN MICROMANIPULATED COMMUNITY
ulation of QD-labeled cells from opaque substrata could prove applicable to capture of communities from a broad spectrum of naturally occurring biofilms. Only veillonellae and RPS-bearing streptococci were visible during the manipulation. However, although not seen, other bacteria were likely to be part of the community because diversity within even very small biofilm communities (three to five cells) has been demonstrated (30, 31). One cell type not targeted in this study but which might be expected was actinomyces because it was shown to be associated with RPS-bearing streptococci in initial communities (31). Another expected cell type was adhesin-bearing streptococci; oral streptococci are known to coaggregate with one another (19), and a Gn-specific adhesin-bearing S. gordonii not visible during manipulation was indeed captured together with the immunofluorescencetargeted RPS-bearing organism S. oralis. That only two streptococcal genotypes were obtained in the absence of a variety of other species illustrates the robustness of targeting a small number of cell types when diverse yet small oral communities are isolated. Only about 50% of oral phylotypes are estimated to have been cultured (33). Veillonella spp. can be difficult to isolate from clinical specimens because other bacteria overgrow them unless the other bacteria are inhibited by an antibiotic or detergent (22, 36). However, in the presence of vancomycin, nothing grew from the micromanipulated sample known to contain cells reactive with veillonella-specific antibodies, as well as a veillonella 16S rRNA gene sequence. Furthermore, no veillonellae were obtained by other isolation methods, including using lactate or pyruvate as a nutritional source, immunobinding of cells to anti-R1-coated magnetic beads, or growth in spent streptococcal culture media. However, on MSM agar in the absence of antibiotics, anti-R1-reactive cells were found in colonies of streptococci. As demonstrated in the in vitro experiments, the surrogate strain PK1910 was dependent on association with at least one of the clinical streptococci for growth in saliva. Overall, the data suggest that streptococcal growth in saliva alone is sufficient to support the growth of veillonellae and that a metabolic product produced by streptococci during their growth on saliva may be essential for the survival of the uncultured Veillonella sp. in the micromanipulated community. This report demonstrates that metabolic dependence is facilitated by coaggregation of the participants; in the in vitro reconstruction, both streptococci coaggregated with PK1910, and the streptococci interacted by RPS-mediated coaggregation. The capture of a coaggregating pair of cells from a naturally occurring community containing a very small number of cells provides proof that coaggregation does occur in vivo and is the first step in establishment of a multispecies community. In particular, intrageneric coaggregation of streptococci and intergeneric coaggregation of streptococci and veillonellae are important factors in the initial formation of spatially distinct and metabolically cooperative communities during primary colonization of the tooth surface.
ACKNOWLEDGMENTS We thank Nicholas Jakubovics (Newcastle University) for his assistance with the Q-PCR.
8153
This research was supported in part by the Intramural Research Program of the National Institute of Dental and Craniofacial Research, National Institutes of Health. REFERENCES 1. Aas, J. A., B. J. Paster, L. N. Stokes, I. Olsen, and F. E. Dewhirst. 2005. Defining the normal bacterial flora of the oral cavity. J. Clin. Microbiol. 43:5721–5732. 2. Alam, S., S. R. Brailsford, R. A. Whiley, and D. Beighton. 1999. PCR-based methods for genotyping viridans group streptococci. J. Clin. Microbiol. 37: 2772–2776. 3. Bruchez, M., Jr., M. Moronne, P. Gin, S. Weiss, and A. P. Alivisatos. 1998. Semiconductor nanocrystals as fluorescent biological labels. Science 281: 2013–2016. 4. Celesk, R. A., and J. London. 1980. Attachment of oral Cytophaga species to hydroxyapatite-containing surfaces. Infect. Immun. 29:768–777. 5. Ceri, H., M. E. Olson, C. Stremick, R. R. Read, D. Morck, and A. Buret. 1999. The Calgary biofilm device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms. J. Clin. Microbiol. 37:1771– 1776. 6. Chalmers, N. I., R. J. Palmer, Jr., L. Du-Thumm, R. Sullivan, W. Shi, and P. E. Kolenbrander. 2007. Use of quantum dot luminescent probes to achieve single-cell resolution of human oral bacteria in biofilms. Appl. Environ. Microbiol. 73:630–636. 7. Chan, W. C., and S. Nie. 1998. Quantum dot bioconjugates for ultrasensitive nonisotopic detection. Science 281:2016–2018. 8. Cisar, J. O., A. L. Sandberg, C. Abeygunawardana, G. P. Reddy, and C. A. Bush. 1995. Lectin recognition of host-like saccharide motifs in streptococcal cell wall polysaccharides. Glycobiology 5:655–662. 9. Cisar, J. O., A. L. Sandberg, G. P. Reddy, C. Abeygunawardana, and C. A. Bush. 1997. Structural and antigenic types of cell wall polysaccharides from viridans group streptococci with receptors for oral actinomyces and streptococcal lectins. Infect. Immun. 65:5035–5041. 10. Collins, M. D. 1982. Reclassification of Bacterionema matruchotti (Gilmour, Howell and Bibby) in the genus Corynebacterium, as Corynebacterium matruchotii comb. nov. Zentbl. Bakteriol. Mikrobiol. Hyg. Abt. 1 Orig. Reihe C 3:364–367. 11. Daims, H., S. Lucker, and M. Wagner. 2006. daime, a novel image analysis program for microbial ecology and biofilm research. Environ. Microbiol. 8:200–213. 12. Diaz, P. I., N. I. Chalmers, A. H. Rickard, C. Kong, C. L. Milburn, R. J. Palmer, Jr., and P. E. Kolenbrander. 2006. Molecular characterization of subject-specific oral microflora during initial colonization of enamel. Appl. Environ. Microbiol. 72:2837–2848. 13. Egland, P. G., R. J. Palmer, Jr., and P. E. Kolenbrander. 2004. Interspecies communication in Streptococcus gordonii-Veillonella atypica biofilms: signaling in flow conditions requires juxtaposition. Proc. Natl. Acad. Sci. USA 101:16917–16922. 14. Hoshino, T., T. Fujiwara, and M. Kilian. 2005. Use of phylogenetic and phenotypic analyses to identify nonhemolytic streptococci isolated from bacteremic patients. J. Clin. Microbiol. 43:6073–6085. 15. Hsu, S. D., J. O. Cisar, A. L. Sandberg, and M. Kilian. 1994. Adhesive properties of viridans streptococcal species. Microb. Ecol. Health Dis. 7:125– 137. 16. Hughes, C. V., R. N. Andersen, and P. E. Kolenbrander. 1992. Characterization of Veillonella atypica PK1910 adhesin-mediated coaggregation with oral Streptococcus spp. Infect. Immun. 60:1178–1186. 17. Hughes, C. V., P. E. Kolenbrander, R. N. Andersen, and L. V. Moore. 1988. Coaggregation properties of human oral Veillonella spp.: relationship to colonization site and oral ecology. Appl. Environ. Microbiol. 54:1957–1963. 18. Kolenbrander, P. E., R. N. Andersen, D. S. Blehert, P. G. Egland, J. S. Foster, and R. J. Palmer, Jr. 2002. Communication among oral bacteria. Microbiol. Mol. Biol. Rev. 66:486–505. 19. Kolenbrander, P. E., R. N. Andersen, and L. V. H. Moore. 1990. Intrageneric coaggregation among strains of human oral bacteria: potential role in primary colonization of the tooth surface. Appl. Environ. Microbiol. 56:3890– 3894. 20. Kolenbrander, P. E., and J. London. 1993. Adhere today, here tomorrow: oral bacterial adherence. J. Bacteriol. 175:3247–3252. 21. Ludwig, W., O. Strunk, R. Westram, L. Richter, H. Meier, Yadhukumar, A. Buchner, T. Lai, S. Steppi, G. Jobb, W. Forster, I. Brettske, S. Gerber, A. W. Ginhart, O. Gross, S. Grumann, S. Hermann, R. Jost, A. Konig, T. Liss, R. Lussmann, M. May, B. Nonhoff, B. Reichel, R. Strehlow, A. Stamatakis, N. Stuckmann, A. Vilbig, M. Lenke, T. Ludwig, A. Bode, and K. H. Schleifer. 2004. ARB: a software environment for sequence data. Nucleic Acids Res. 32:1363–1371. 22. MacFarlane, T. W. 1977. A semiselective medium for the isolation of Veillonella species from the mouth. J. Clin. Pathol. 30:191–192. 23. Malinen, E., T. Rinttila, K. Kajander, J. Matto, A. Kassinen, L. Krogius, M. Saarela, R. Korpela, and A. Palva. 2005. Analysis of the fecal microbiota of irritable bowel syndrome patients and healthy controls with real-time PCR. Am. J. Gastroenterol. 100:373–382.
8154
CHALMERS ET AL.
24. Mampel, J., T. Spirig, S. S. Weber, J. A. Haagensen, S. Molin, and H. Hilbi. 2006. Planktonic replication is essential for biofilm formation by Legionella pneumophila in a complex medium under static and dynamic flow conditions. Appl. Environ. Microbiol. 72:2885–2895. 25. McBride, B. C., and J. S. van der Hoeven. 1981. Role of interbacterial adherence in colonization of the oral cavities of gnotobiotic rats infected with Streptococcus mutans and Veillonella alcalescens. Infect. Immun. 33:467– 472. 26. Mikx, F. H., and J. S. van der Hoeven. 1975. Symbiosis of Streptococcus mutans and Veillonella alcalescens in mixed continuous cultures. Arch. Oral Biol. 20:407–410. 27. Mikx, F. H., J. S. van der Hoeven, K. G. Konig, A. J. Plasschaert, and B. Guggenheim. 1972. Establishment of defined microbial ecosystems in germfree rats. I. The effect of the interactions of Streptococcus mutans or Streptococcus sanguis with Veillonella alcalescens on plaque formation and caries activity. Caries Res. 6:211–223. 28. Mouton, C., H. Reynolds, and R. J. Genco. 1977. Combined micromanipulation, culture and immunofluorescent techniques for isolation of the coccal organisms comprising the “corn cob” configuration of human dental plaque. J. Biol. Buccale 5:321–332. 29. Nyvad, B., and M. Kilian. 1987. Microbiology of the early colonization of human enamel and root surfaces in vivo. Scand. J. Dent. Res. 95:369–380. 30. Palmer, R. J., Jr., P. I. Diaz, and P. E. Kolenbrander. 2006. Rapid succession within the Veillonella population of a developing human oral biofilm in situ. J. Bacteriol. 188:4117–4124. 31. Palmer, R. J., Jr., S. M. Gordon, J. O. Cisar, and P. E. Kolenbrander. 2003. Coaggregation-mediated interactions of streptococci and actinomyces detected in initial human dental plaque. J. Bacteriol. 185:3400–3409. 32. Palmer, R. J., Jr., R. Wu, S. Gordon, C. G. Bloomquist, W. F. Liljemark, M. Kilian, and P. E. Kolenbrander. 2001. Retrieval of biofilms from the oral cavity. Methods Enzymol. 337:393–403. 33. Paster, B. J., S. K. Boches, J. L. Galvin, R. E. Ericson, C. N. Lau, V. A.
J. BACTERIOL.
34.
35. 36.
37. 38.
39.
40.
41.
42.
43.
Levanos, A. Sahasrabudhe, and F. E. Dewhirst. 2001. Bacterial diversity in human subgingival plaque. J. Bacteriol. 183:3770–3783. Rinttila, T., A. Kassinen, E. Malinen, L. Krogius, and A. Palva. 2004. Development of an extensive set of 16S rDNA-targeted primers for quantification of pathogenic and indigenous bacteria in faecal samples by real-time PCR. J. Appl. Microbiol. 97:1166–1177. Rogosa, M. 1956. A selective medium for the isolation and enumeration of the veillonella from the oral cavity. J. Bacteriol. 72:533–536. Rogosa, M., R. J. Fitzgerald, M. E. Mackintosh, and A. J. Beaman. 1958. Improved medium for selective isolation of Veillonella. J. Bacteriol. 76:455– 456. Saitou, N., and M. Nei. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–425. Tamura, K., J. Dudley, M. Nei, and S. Kumar. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24:1596–1599. Thurnheer, T., R. Gmur, E. Giertsen, and B. Guggenheim. 2001. Automated fluorescent in situ hybridization for the specific detection and quantification of oral streptococci in dental plaque. J. Microbiol. Methods 44:39–47. Truper, H. G., and L. de’Clari. 1997. Taxonomic note: necessary correction of specific epithets formed as substantives (nouns) in “apposition.” Int. J. Syst. Bacteriol. 47:908. van der Hoeven, J. S., A. I. Toorop, and R. H. Mikx. 1978. Symbiotic relationship of Veillonella alcalescens and Streptococcus mutans in dental plaque in gnotobiotic rats. Caries Res. 12:142–147. Vickerman, M. M., S. Iobst, A. M. Jesionowski, and S. R. Gill. 2007. Genome-wide transcriptional changes in Streptococcus gordonii in response to competence signaling peptide. J. Bacteriol. 189:7799–7807. Yoshida, Y., S. Ganguly, C. A. Bush, and J. O. Cisar. 2005. Carbohydrate engineering of the recognition motifs in streptococcal coaggregation receptor polysaccharides. Mol. Microbiol. 58:244–256.
