Oral Microbiol Immunol 2001: 16: 163–169 Copyright C Munksgaard 2001 Printed in Denmark . All rights reserved ISSN 0902-0055 Z. Metzger1,2, L. G. Featherstone2, Kinetics of coaggregation of W. W. Ambrose2,M.Trope1, R. R. Arnold2,3 1Department of Endodontics, 2Dental Research Center and 3Departments of Porphyromonas gingivalis with Periodontics and Diagnostic Sciences, School of Dentistry, University of North Carolina at Fusobacterium nucleatum using Chapel Hill, USA an automated microtiter plate assay Metzger Z, Featherstone LG, Ambrose WW, Trope M, Arnold RR. Kinetics of coaggregation of Porphyromonas gingivalis with Fusobacterium nucleatum using an automated microtiter plate assay. Oral Microbiol Immunol 2001: 16: 163–169. C Munksgaard, 2001. Coaggregation between Porphyromonas gingivalis and Fusobacterium nucleatum strains was previously studied using either a semi-quantitative macroscopic assay or radioactive tracer assays. A new automated microtiter plate assay is introduced, in which the plate reader (Vmax) was adapted to allow quantitative evaluation of the kinetics of coaggregation. F nucleatum PK 1594 coaggregated with P. gingivalis HG 405 with a maximal coaggregation rate of 1.05 mOD/ min, which occurred at a P. gingivalis to F. nucleatum cell ratio of 1 to 2. F. nucleatum PK 1594 failed to do so with P. gingivalis strains A 7436 or ATCC Key words: coaggregation; Fusobacterium 33277. Galactose inhibition of this coaggregation could be quantitatively nucleatum; Porphyromonas gingivalis measured over a wide range of concentrations to demonstrate its dose- dependent manner. P. gingivalis HG 405 failed to coaggregate with F. nucleatum Zvi Metzger, The Goldschleger School of strains ATCC 25586 and ATCC 49256. The assay used in the present study is Dental Medicine, Tel Aviv University, Ramat Aviv, Tel Aviv 69978, Israel a sensitive and efficient quantitative automated tool to study coaggregation and may replace tedious radioactive tracer assays. Accepted for publication October 8, 2000 Coaggregation between oral bacterial tion. Many of them were directed at coaggregation (2). However, since this strains is dependent on a plethora of high-aggregating pairs of strains, which method allows only semi-quantitative bacterial adhesins that mediate this coaggregate and sediment fast. evaluation, methods based on radioac- phenomenon. The coaggregation be- Interactions between P. gingivalis tive labeling of the bacteria have also tween Porphyromonas gingivalis and Fu- and F. nucleatum may not be limited been used to provide quantitative analy- sobacterium nucleatum has been exten- only to dental plaque formation. Syner- sis of coaggregation and of its inhi- sively studied by Kolenbrander et al. gistic pathogenicity of P. gingivalis and bition by specific sugars (5, 11, 16). (11, 12), by Kinder & Holt (9, 10) and F. nucleatum has also been reported in In the present study we introduce a recently by Shaniztki et al. (14). It has subcutaneous abscess models (1, 5, 15). new approach that allows quantitative been attributed to proteinaceous adhes- Therefore, the present study focused on evaluation, as well as determination of ins on F. nucleatum that recognize P. gingivalis strains that were recently the kinetics of the coaggregation of the galactose-containing saccharides on the studied for their pathogenicity in the bacteria studied. The coaggregation porphyromonas. Most of these coag- murine subcutaneous chamber model properties of three P. gingivalis strains gregation studies were aimed at elucid- (6, 7). with three potential F. nucleatum part- ating the role of these adhesins in the Most previous coaggregation studies ners were evaluated using a new spec- complex events of dental plaque forma- used a macroscopic evaluation scale for trophotometric method. 164 Metzger et al. We report here on an adaptation of (Maxisorp-immunoplates, Nunc) were Macroscopic coaggregation assay the kinetic automated microtiter plate pretreated by filling the wells with a reader Vmax for an effective and quan- The coaggregation assay originally de- solution of 0.05% Tween-20 in PBS, ad- titative measurement of the kinetics of scribed by Cisar et al. (2) and modified justed to pH 6.8, for 30 min. The buffer coaggregation between bacterial strains by Kolenbrander was used (11, 12). was then discarded and the plates were and the kinetics and other character- Briefly, bacteria of an overnight culture, allowed to dry. Bacterial suspensions of istics of coaggregation of F. nucleatum which reached a late exponential or each of the test strains in aggregation PK 1594 with P. gingivalis HG 405 and early stationary phase, were washed buffer (as described above) were ad- its absence with P. gingivalis A 7436 or three times in an aggregation buffer justed to A660½0.5 and added to the P. gingivalis ATCC 33277. consisting of 0.1 mM CaCl2, 0.1 mM wells, fusobacteria first, followed by the MgCl2 0.15 M NaCl and 0.02% NaN3, porphyromonads. The total volume in dissolved in 1.0 mM Tris and adjusted each well was 110 ml, and the total A660 to pH 8.0. The microorganisms were was kept at 0.5, unless otherwise speci- Material and methods then resuspended in this buffer at an fied. Bacterial strains and growth conditions absorbance at 660 nm (A660)½1.0 and Different proportions between the Three P. gingivalis strains, HG 405, A kept at 4æC until used in the experi- coaggregating partners have been pre- 7436 and ATCC 33277, were used in ments. viously reported to dramatically influ- this study as well as Prevotella interme- To determine coaggregation, 150 ml ence the outcome, even to the extent dia CB21. All of the above were tested of an A660½1.0 suspension of each that the same pair of bacteria seemed for coaggregation with three F. nucleat- bacterium of the pairs of strains tested to lack coaggregating capacity when um strains: PK 1594, ATCC 25586 and were mixed at room temperature. Co- mixed at other than optimal ratios (11). ATCC 49256. aggregation was then followed visually, Therefore, each pair of bacteria was P. gingivalis A 7436 has been shown using the scale described by Cisar et tested with a range of P. gingivalis: F. to be highly pathogenic and disseminat- al. (2). In this scale, 4¹ corresponded nucleatum ratios, including 10:1, 3:1, ing in the mouse model (6, 7). Both P. to formation of large, fast settling ag- 2:1, 1:1, 1:2, 1:3 and 1:10, and controls gingivalis strains HG 405 and ATCC gregates, with the supernatant remain- were included that consisted of each of 33277 were capable of establishing a ing water-clear; 3¹ was defined by the the strains alone. The various bacterial locally contained infection in the formation of large settling aggregates, proportions were achieved by dispens- chamber model, which progressed to lo- but with slightly turbid supernatant; ing the proper volumes of each of the cal suppurative lesions, but were in- 2¹ represented the presence of definite suspensions to maintain an A660 of 0.5. capable of dissemination even at high aggregates which did not settle im- The assays were carried out in quad- infectious dose (6, 7). mediately; 1¹ was characterized by ruplicate. F. nucleatum PK 1594 has been finely dispersed aggregates in a turbid Immediately after the suspension of studied previously for its coaggregation background, and 0 had no visible ag- the second partner was added, the with several other P. gingivalis strains gregates with no reduction in turbidity. plates were inserted into the Vmax plate and has been defined as having a galac- In experiments in which bacterial ra- reader and the kinetics of the coaggre- tose-inhibitable adhesin that mediates tios other than 1:1 were tested, the gation followed for 30 min by monitor- its coaggregation with P. gingivalis PK total A660 of the mixture was also kept ing the reduction in optical density. The 1924 (11, 14), as well as its attachment at 1.0. reader was set to read each well every to mammalian cells (16). F. nucleatum 14 s with repeated shaking and the limit strains ATCC 25586 (subsp. F. nucleat- set to optical density »0.05. The optical Microscopic examination um) and ATCC 49256 (subsp. Prevotella density diminished gradually with the vincentii) are two of the type strains for Each mixed bacterial suspension that aggregation, and the maximal slope of sub-species groups of F. nucleatum pro- was macroscopically tested for coaggre- the resulting curve was expressed as posed by Dzink et al. (3). P. intermedia gation was also examined microscopi- Vmax value (expressed as change in (D) CB 21 has been studied by Baumgart- cally (phase contrast at ¿1000 magni- mOD/min), which was used to estimate ner et al. for its synergistic pathoge- fication). The observation was done on different rates of aggregation. The cor- nicity with F. nucleatum VPI 10197 (1) a ‘‘wet mount’’ as described by Kolen- relation coefficient of the resulting and was used in the present study as a brander & Andersen (11). curves was usually Ͼ93%, unless other- non–P. gingivalis positive control that wise noted. For each group consisting presented high coaggregation with F. of four wells, a mean Vmax value was Vmax automated coaggregation assay nucleatum PK 1594. calculated with a standard error of the All of these strains were grown an- To quantitatively evaluate the coaggre- mean usually within 5%. aerobically in Wilkins-Chalgren anaer- gation rate of weak coaggregating pairs Fusobacteria were always added first, obic broth (Oxoid), in an atmosphere of and to quantitatively evaluate the in- while the second partner was added 5% CO2, 10% H2 and 85% N2 at 37æC fluence of competitors with this coag- after a 2-min pause. This proved to be (Coy anaerobic chamber). Bacterial gregation, the Vmax kinetic microtiter important, as it eliminated from the as- strains were kept as frozen stocks and plate spectrophotometric reader (Mol- say the slight background spontaneous were started and passed three times in ecular Devices Corp., Sunnyvale, CA) sedimentation encountered with some this growth medium before being used was utilized.