1093
Bacterial Concentration Fluorescence Immunoassay (BCFIA) for the Detection of Periodontopathogens in
Plaque Larry F. Wolff, * LuAnn Anderson, * Gregory P. Sandberg/ Lowell Reither, * Christine A. Binsfeld/ Giuseppe Corinaldesi,* and Charles E. Shelburnef A BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAY (BCFIA) Was developed tO rapidly detect periodontopathic bacteria in human plaque samples. The BCFIA utilized fluorescent-tagged monoclonal antibodies (MAbs) directed against the lipopolysaccharide of selected Gram-negative bacteria. Microorganisms identified in plaque using the BCFIA
included Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, and Fusobacterium nucleatum. The immunoassay procedure involved combining a patient's plaque sample with a species-specific fluorescein isothiocyanate-labeled
MAb and then incubating the mixture in a specialized microtiter plate allowing the MAb to bind to its homologous bacteria. Bound and unbound fluorescent-tagged MAbs were separated by filtration and total bound bacterial fluorescence was determined with a fluorimeter. The relative number of a bacterial species in a given plaque sample was estimated by reference to a standard curve carried through the BCFIA. The BCFIA had a lower detection limit of near 104 specific bacterial cells in a mixed bacterial preparation or plaque sample. When compared to cultivable flora procedures in detecting the 4 periodontopathogens, the BCFIA had high levels of statistical sensitivity, 97% to 100%, while statistical specificity ranged between 57% and 92%. There was a 71% to 82% agreement between BCFIA and DNA probe methodology in detecting periodontopathogens in plaque. The BCFIA, when compared to cultivable flora, offers the advantage of evaluating both live and dead bacterial cells in plaque. This may in part, if not fully, explain the lower specificity values of the BCFIA when compared to cultivable flora. Screening plaque samples for periodontopathic bacteria is considerably faster and results in a greater frequency of detection with BCFIA than cultivable flora based methods. / Periodontol 1992; 63:1093-1101.
Key Words:
Dental
plaque/microbiology; immunoassay,
fluorescence.
Although the gingival crevice supports the growth of a wide variety of microbial species, only a few of these microorganisms have been closely associated with periodontal disease.1"4 The identification of these microorganisms in subgingival plaque using cultural methods, although precise and specific, is labor-intensive, time-consuming and requires the viability of the bacterial cells. This severely limits research surrounding the evaluation of plaque for periodontal pathogens, especially in large population groups. Additional techniques are needed to more rapidly assess plaque for specific bacteria and to a certain extent these technologies, DNA probes, enzyme-linked immunosorbent "University of Minnesota Clinical Research Center for Minneapolis, MN. tBiosciences Laboratory, 3M Company, St. Paul, MN.
eases,
Periodontal Dis-
(ELISA), and immunofluorescence, have been developed.5"11 Jolley and co-workers introduced a particle concentration fluorescence immunoassay (PCFIA) which utilized polystyrene beads coated with Immunoglobulin as the solid phase.12 These investigators reported the PCFIA was as or more sensitive and could more quickly evaluate samples than the ELISA. Other investigators have used the PCFIA to quantitate Immunoglobulins and interleukin-4 in human serum and also alpha-1 antitrypsin from Escherichia coli lysates.13"16 A modified PCFIA was developed using bacterial cells as the solid phase, instead of polystyrene beads.17 The bacterial solid phase, Escherichia coli, was used to analyze both polyclonal and monoclonal antibodies against surface antigens of this bacterium. These investigators inassay
J Periodontol 1094
BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAV
dicated the bacterial concentration fluorescence immunoassay (BCFIA) showed up to 50-fold greater sensitivity in bacterial cell detection and was more quickly performed than the ELISA. We have modified and utilized the BCFIA for the rapid identification of periodontopathic bacteria in human plaque specimens. Monoclonal antibodies have been used in the BCFIA by our laboratory while both monoclonal and polyclonal antibodies have been utilized with indirect immunofluorescence microscopy by other investigators.9'18 21 The differentiation between monoclonal and polyclonal antibody production has been thoroughly described by Milstein.22 In this report, emphasis will be placed on a series of monoclonal antibodies which have been developed to a selected group of periodontal disease-associated pathogens including Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, and Fusobacterium nucleatum.2324 These antibodies recognize specific sites on the lipopolysaccharide (LPS) molecules of these microorganisms and form the basis for the BCFIA. Bacteria are the solid phase to which the LPS-specific fluorescein isothiocyanate-tagged MAbs are directed. The bound and unbound fluorescent-tagged MAbs are separated by filtration in a modified microtiter plate and the total bacterialbound fluorescence is determined with a fluorimeter. The BCFIA described in this report offers the advantage of evaluating both live and dead bacterial cells in plaque and with respect to screening plaque samples for periodontopathic bacteria is considerably faster than cultivable flora-based methods. MATERIAL AND METHODS Bacterial Strains The strains of bacteria used in this investigation were obtained from the American Type Culture Collection (ATCC) (Table 1). Bacteria were grown in media as recommended by the ATCC or under conditions as previously de-
scribed.1-25"28
Monoclonal Antibody Production, Screening and Fluorescein Isothiocyanate Labeling Hybridizations between Balb/c NS-1 myeloma cells and spleen cells from immunized Balb/c mice were performed as described previously.29 Briefly, Balb/c mice were immunized with a bacterial strain, either A. actinomycetemcomitans 29522, P. intermedia 25611, F. nucleatum 25586, or P. gingivalis 33277, 3 to 5 times by intraperitoneal injection with a final injection of 10 million cells given intravenously 3 days prior to aseptic removal of the spleens. Splenic leukocytes were then combined with a NS/1 myeloma cell line in 35% polyethylene glycol. The resulting hybridoma culture supernatants containing mouse Immunoglobulin were screened for the production
of antibacterial antibody using either an ELISA30 and/or a BCFIA.31 Cultures represented by supernatants with sig-
December 1992
Table 1. Bacterial Strains Used in Cross-Reactivity Testing A. actinomycetemcomitans, P. gingivalis, P. F. nucleatum Monoclonal Antibodies. A.
actinomycetemcomitans 29522 (b)* 29523 29524
A. A. B. B. C. C. C. E. E. E. F. H. H. H. . .
israelii 12102
(a) (b)
naeslundii 12104
forsythus 43037 fragilis 25285
concisas 33237 ochracea 33569 sputigena 33612 alactotyticum 23263 corrodens 23834 limosa 8486 nucleatum 25586* 23726 10953 aphrophilus 13252 influenzae 8142 9795 segnis 7901 sicca 9913 subflava 14799
Experiments for intermedia, and
P. buccae 33573 33574 P. denticela 33185 P. intermedia 25611* 33563 P. loescheii 15930 P. melaninogenicus 15032 15033 P. gingivalis 33277* P. micros 33270 S. mutans 10449 27607 S. sanguis 10566 S. sobrinus 33478 S. sputigena 33150 T. denticela 33405 33520 33521 T. vincentii 35580 V. caviae 33540 W. curva 335244 W. recta 33238
"Indicates those bacterial strains to which monoclonal
veloped.23"24
(Supplement)
antibody was
de-
nificantly elevated antibody binding over negative controls, that is the mean relative fluorescence units of the negative controls plus 3 standard deviations, were expanded for further examination. Positive controls using diluted serum from the immunized mice were included along with hyperimmune rabbit serum to each bacteria. Supernatants exhibiting reactivity were then cloned by limiting dilution in 96-well plates. Stable cloned hybrids were expanded by growth in culture and approximately 107 cells were injected intraperitoneally into Balb/c mice to induce ascites. The ascitic fluid was collected, antibodies purified using mixed bed ion-exchange HPLC* and dialyzed in 0.05 M sodium carbonate buffer, pH 9.5 at 2 mg/ml. For the BCFIA (described below), MAb was labeled by dialyzing 3 ml of purified antibody against fluorescein isothiocyanate (FITC), 3 mg FITC/100 ml carbonate buffer for 18 hrs, and then performing chromatography over G-25 sepharose5 to remove free FITC. The fluorescein to protein ratio (F/P) was determined by comparing the molar amounts of FITC and protein at 495 and 280 nm, respectively. Acceptable preparations for use in the BCFIA had a F/P ratio of between 3 and 12.
Monoclonal Antibody Binding to Purified LPS LPS was prepared from 500 mg of lyophilized late log phase cultures of Treponema denticola, Eikenella corrodens, Haemophilius aphrophilus, A. actinomycetemcomitans, P. intermedia, F. nucleatum, and P. gingivalis using the method of Darveau and Hancock.32 Escherichia coli, Salmonella minnesota, and Vibrio cholera LPS were com*Bakerbond ABX, JT Baker Research §Pharmacia LKB, Piscataway, NJ.
Products, Phillipsburg, NJ.
Volume 63 Number 12
WOLFF, ANDERSON, SANDBERG, REITHER, BINSFELD, CORINALDESI,
mercially obtained.11 Reactivity of each LPS preparation with the 4 MAbs was evaluated using an ELISA. Briefly, purified LPS was resuspended in PBS at 100 vg/ml and 50 µ amounts of each placed in individual wells of 96-well microtiter plates. After incubation at room temperature for 18
hours the unbound LPS was removed and 300 µ of PBS with 0.25% bovine serum albumin was added for 4 hours. The plates were then washed 3 times with PBS-Tween 20. Fifty µ of MAb was added to each of the wells and incubated for 1 hour followed by 3 washes with PBS-Tween. A peroxidase-labeled goat anti-mouse IgG was then added (50 µ /well) and incubated for an additional hour. After 3 additional washes in PBS-Tween 20, 50 µ of peroxidase substrate (1 mg/ml o-phenylenediamine, 0.03% hydrogen peroxide, 0.05 M sodium phosphate, pH 4.5) was added and the reaction stopped with 50 µ of 2 M H2S04 after 30 minutes. The optical density was determined at 490 nm using a microplate reader.' Monoclonal Antibody Specificity Forty-four ATCC strains of bacteria were washed 3 times 107 bacteria/ml. in PBS and resuspended in PBS at 5 to individual of each of the strains added were µ Twenty wells in a fluoricon assay plate.* Twenty µ of either FITClabeled anti-bacterial MAb, 10 µg/ml, or a similarly labeled MAb with a non-bacterial specificity, which served as a negative control, was then added to the wells and the plates incubated 20 minutes at room temperature in the dark. Plates were washed twice with PBS-NP40, and the bound fluorescence determined. The percent cross-reactivity was calculated by dividing the difference of the binding of heterologous anti-bacterial MAb and the non-bacterial MAB by the total bacterial MAB bound in a homologous reaction, followed by multiplication by 100. Bacterial Concentration Fluorescence
Immunoassay
(BCFIA) The general principle of the BCFIA has been described in an earlier report.31 The modified BCFIA in relation to evaluating plaque samples or mixed bacterial cultures for specific bacteria or identifying bacteria in a single bacterial suspension is outlined in Figure 1. The BCFIA was performed in a specially designed 96-well microtiter plate.1 Whole bacterial cells served as the solid phase in the BCFIA procedure. Prior to carrying each plaque sample or bacterial preparation through the immunoassay, cells were washed twice in distilled water-0.05% Tween and then resuspended in 500 µ PBS-0.05% Tween plus 1% goat serum. Twenty µ from 500 µ of the bacterial or plaque cell suspension was placed in each of 2 wells of the microtiter plate. To 1 well was added 20 µ of a bacterial species-specific FITC-
"Sigma, St. Louis, MO. 'Dynatech, Chantilly, VA. *IDEXX Corporation, Portland,
^ •000000000000 cooooooooooor
»oooooooooodL 000000000000
c000000000000
®
Bacteria
1095
MAB Bound to
(Solid Phase)
Specific Bacteria
^OOOOOOOOOOOO F! (jo resesi
Isothiocyanate
Labeled MAB (20µ )
w
gingivalis.
intermedia,
azi neTiyc01C~!CCTii'^S. F. nucleatum
Washed
Plaque Suspension
© Unbound MABj Pass Through
Filter. 0.2µ Pore Size
.
--
*
Figure 1. Principles of the BCFIA procedure for evaluating plaque samples, mixed bacterial cultures, or a specific bacterial suspension. labeled** MAb and incubated in the dark at 22°C for 20 minutes and then washed twice with PBS plus 0.0125% NP-40 and filtered under 25 mm vacuum. The other well containing the bacterial or plaque suspension served as a control and was treated with a non-bacterial binding antibody, FITC labeled MOPC, a myeloma protein at a concentration of 5 µg/ml and a F/P of 10, and assayed in parallel to the test mixture. Additional controls included carrying buffer as well as an FITC-labeled MAb suspension through the immunoassay without plaque or bacteria. Species specific MAbs utilized in the immunoassay included those against P. gingivalis, P. intermedia, A. actinomycetemcomitans, and F. nucleatum. The pore size of the membrane filter was 0.2 µ , allowing bacteria with or without tagged MAb to be trapped on the membrane surface while the unbound MAb passed through the filter with the buffer wash. Total bacterial bound fluorescence (485 excitation nm) was determined in a fluorimetertt at a wavelength of 535 nm. Total bound bacterial-specific fluorescence was obtained by subtracting non-specific fluorescent binding in the control well from the total fluorescent antibody binding in the test well. The relative bacterial cell equivalents were determined by comparison to a standard curve (see below). BCFIA Evaluation with Single Cultures of Periodontopathic Bacteria The sensitivity or bacterial detection limit of the immunoassay for screening single cultures of P. gingivalis, P. intermedia, F. nucleatum, or A. actinomycetemcomitans was determined using the BCFIA as described above. Twenty µ of a washed log-phase bacterial cell preparation were added to wells of the immunoassay plate followed by 1 of the FITC-labeled species-specific MAbs. The final total number of bacteria carried through the BCFIA ranged be"Isomer I, Research Organics, Cleveland, OH. fluorescence Concentration Analyzer, IDEXX
ME.
SHELBURNE
ME.
Corporation, Portland,
J Periodontol 1096
BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAY
tween
102 and 106. The MAb concentration and F/P for
species-specific MAb preparation used in these experiments were: P. gingivalis 2.6 µ^ !, 4.6; F. nucleatum 3.7 µ& 1, 4.2; A. actinomycetemcomitans 4.3 µ / , 4.5; P. intermedia 7.1 µ& 1, 11.1; respectively. each
BCFIA Evaluation with Periodontopathic Bacteria in a Mixed Culture In an effort to determine if a mixed culture of bacteria would have an effect on the ability of the BCFIA to detect
specific periodontopathic bacteria, competition experiperformed. A description of these experiments is given in Table 2. Dilution buffers A, B, C, and D were prepared by adding log phase cultures of the Gram-negative bacteria P. gingivalis, P. intermedia, F. nucleatum, A. actinomycetemcomitans, E. corrodens, and/or the Gram-positive bacterium Streptococcus mutans to phosphate-buffered saline, pH 7.4, at the concentrations shown. A portion of buffers A, B, C, and D were then spiked with 2.5 107/ ml P. gingivalis, P. intermedia, F. nucleatum, or A. actinomycetemcomitans, respectively. The final total concentration of bacteria in each buffer with the spiked bacteria was 108/ml. This suspension of bacteria was then diluted in buffers A, B, C, or D (without the spiked bacteria) and titrated against species-specific MAbs at a concentration of 1 µ& 1 using the BCFIA. With this dilution scheme, the concentration of the spiked bacteria decreased while the a
ments were
concentration of the other bacteria remained the same. The F/P for the MAb preparations used in these experiments were: P. gingivalis, 5.6; F. nucleatum, 6.2; A. actinomycetemcomitans, 4.2; and P. intermedia 7.1.
Preparation of the Standard Curve for the BCFIA The standard curve was established by examining the relationship between fluorescence and relative bacterial num-
December 1992
(Supplement)
whole bacterial cells. The LPS-coated beads were assayed with their homologous MAbs, using the same procedure as described above in the BCFIA. Simultaneously, suspensions of each of the 5 bacteria from log phase cultures were adjusted, using a Petroff-Hausser chamber, at concentrations ranging from 103 to 106 bacteria/well and then immunoassayed. Each preparation of antigen-coated beads was then diluted with bovine serum albumin plus 1% goat serumcoated beads to give a standard corresponding to a relative number of bacteria in the BCFIA. Specific bead preparations (P. gingivalis, P. intermedia, A. actinomycetemcomitans, or F. nucleatum) ranged in concentration from 0.250% to 0.016% (mg of beads/volume of buffer) and were used to prepare the standard curve. Calibrated bead suspensions were stored at 4°C until used in the BCFIA. Each microtiter plate which contained plaque or other bacterial samples also contained primary standards prepared in duplicate. The average of the 2 primary standards for 5 or more different bead concentrations was used to prepare the standard curve and detemiine the relative number of bacteria in test samples. Bacterial Concentration Fluorescence Immunoassay Compared with Cultivable Flora and DNA Probe in
Detecting Periodontopathogens
Plaque sampling procedures and microbial evaluation of plaque have been described in detail in earlier reports.1-28'33 Eighty-one sites in individuals with a periodontal status ranging from health/gingivitis (n 30) to mild, moderate, or severe Periodontitis ( 51) were included in the phase of the study comparing the BCFIA with cultivable flora. After removal of supragingival plaque, a subgingival plaque sample was obtained with a sterile curet and placed into 250 µ of pre-reduced anaerobically-sterilized diluent and immediately transferred to an anaerobic chamber. The plaque sample was further diluted with 2.0 ml of the pre-reduced diluent, sonicated for 20 seconds,88 serially diluted and plated on supplemented blood agar media for detection of total =
=
bers in the BCFIA. The least squares method was used to determine the best fitting mathematical model. Preparation of the standard curve utilized LPS-enriched coated polystyrene particles as the solid phase in the BCFIA. For each of the bacteria, P. gingivalis, P. intermedia, A. actinomycetemcomitans, and F. nucleatum, 200 ml log phase cultures were washed 3 times in Dulbeco's PBS buffer and resuspended in 20 ml of the same buffer, pH 7.4. The cells were sonicated** for 3 1-minute bursts at an output of 10% and centrifuged for 15 minutes at 12,000 xg to remove unbroken cells. The supernatant from each of the bacterial preparations was combined with 2.0 ml of 0.85 µ fluoricon polystyrene assay particles1 and rotated at 0.5 rpm for 12 to 18 hours at 22°C. The beads were then washed in PBS plus 0.0125% NP-40 and resuspended in 40 ml of the same buffer with 0.5% sodium azide. The LPS-coated beads were then standardized by immunoassay to suspensions of
anaerobic cultivable flora, Fusobacterium spp. and colonies showing dark pigmentation. The dark-pigmenting colonies were screened for the identification of P. intermedia and P. gingivalis using the AN-IDENT system1111 and the APIZYM system as described by Laughon and co-workers.34 Selective agar media was used for the detection of A. actinomycetemcomitans.35 A portion of each plaque sample 20°C until it was carried through the BCFIA. was stored at The phase of the study comparing the BCFIA with DNA probe for detecting periodontopathogens included 28 plaque samples collected from sites with greater than 5 mm probing depth (mean ± SD 8.0 mm ± 1.5 mm) and greater than 3 mm loss of attachment (mean ± SD 7.9 mm ± 1.9 mm). These plaque samples were from different individuals than those used in the cultivable flora studies. Two
**Branson Sonifier, Model 350, Branson Sonic Power
§§Kontes Cell
bury,
CT.
Company,
Dan-
-
=
=
Disruptor, Kontes, Vineland, NJ. "Analylab Products, Sherwood Medical, Plainview, NY.
Volume 63 Number 12
WOLFF, ANDERSON, SANDBERG, REITHER, BINSFELD, CORINALDESI,
Table 2. Protocol
Periodontopathic
Describing Bacterial
Bacteria in
a
Concentration Fluorescence Mixed Culture
1097
SHELBURNE
Immunoassay Evaluation of
Dilution Buffers*
(bacteria/ml)
Bacteria P. P. F. A. E. S.
gingivalis
0
107 107 107 107
intermedia nucleatum
actinomycetemcomitans corrodens
Spike
P. gingivalis P. intermedia F. nucleatum A. actinomycetemcomitans
2.5
Total bacteria/ml
*Bacteria
were
3.5
107
3.5
mutans
107 0 107 107 IO7 IO7
C
D
IO7 IO7 0 IO7 IO7
IO7 IO7 IO7 0 IO7 IO7
3.5
3.5
IO7
2.5
IO7
107
x
2.5
x
IO7 2.5
IO8
10s
IO7
x
IO8
IO8
suspended in phosphate buffered saline, pH adjusted to 7.4. Specificity of Monoclonal Antibody Binding to Lipopolysaccharides Based on EUSA23"24'40
Table 3. Porphyromonas gingivalis (ATCC 33277)
Actinobacillus
actinomycetemcomitans
(ATCC 29522)
Lipopolysaccharide from: A. Prevotella intermedia
Fusobacterium nucleatum
Relative Bacterial Numbers
Relative Bacterial Numbers
(ATCC 25611)
(ATCC 25585)
Figure 2. Species-specific monoclonal antibodies titrated against their homologous bacteria using the BCFIA. sterile paper points were simultaneously inserted to the base of a pocket. After 10 seconds, the paper points were removed. One was placed into 250 µ of pre-reduced diluent and carried through the immunoassay as described above. The other paper point was placed in a sterile vial and forwarded to a commercial laboratory11 for DNA probe analysis. RESULTS Monoclonal Antibodies Reacting With Homologous Bacteria The MAbs were titrated against their homologous bacteria using the BCFIA (Fig. 2). The immunoassay had a lower detection limit for pure cultures of P. gingivalis, P. intermedia, F. nucleatum, and A. actinomycetemcomitans of between 1000 and 10,000 bacteria. It may also be observed that an approximately linear response was seen between relative fluorescent units and bacteria over the range of 104 to
106 bacteria.
'OmniGene, Inc., Cambridge, MA.
actinomycetemcomitans
(Aa ATCC 29522) P. intermedia (Pi ATCC 25611) F. nucleatum (Fn ATCC 25586) P. gingivalis (Pg ATCC 33277) T. denticola (ATCC 33521) E. corrodens (ATCC 23834) H. aphrophilus (ATCC 13252) E. coli 055:B5 E. coli 0111:B4 S. minnesota Re595 V. cholerar569B
Purified
Optical Density 490 nm Antibody to: Fn Pg
Monoclonal Aa Pi 1.43 .000 .031 .002 .000 .000 .000 .000 .020 .000 .013
.000 .862 .000 .000 .071 .016 .003 .024 .000 .000 .000
.031 .000 .771 .002 .021 .019 .000 .021 .004 .000 .000
.002 .077 .002 1.33 .000 .021 .011 .000 .000 .000 .017
Specificity of MAbs Used in the Bacterial Concentration Fluorescence Immunoassay The specificity of the MAbs were evaluated against a battery of 44 ATCC bacterial type strains (Table 1). The heterologous cross-reactions with both Gram-positive and Gramnegative bacterial species were at least one-tenth and in most cases one-twentieth the intensity of the homologous reaction for each of the 4 bacterial-specific MAbs. The specificity of the MAbs for LPS was determined using purified LPS from a number of different bacteria as targets in an ELISA format (Table 3). The average optical density of triplicate wells demonstrated heterologous reactions were less than 5% of the homologous reactions. Detection of Periodontopathic Bacteria in a Mixed Culture Using the BCFIA The effect of a mixed culture of bacteria on the sensitivity of the BCFIA for detecting specific bacteria may be seen in Figure 3. A linear response was observed between fluorescent units and relative number of bacterial cells. In
J Periodontol 1098
O
BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAY
December 1992
(Supplement)
20 m
F. nucleatum
O
(25586)
c
o
actinomycetemcomitans
3
(29522) ß> >
ce
mixed
14 Relative Number
addition, even in the presence of high levels of other bacteria, using each of the 4 species-specific MAbs, the BCFIA had
a
lower detection limit value of
near
104 bacterial cell
105
10 6
Porphyromonas gingivalis Cells
Figure 4. Method used to convert fluorescence for a plaque sample determined by the BCFIA into Porphyromonas gingivalis cell numbers. Heavy solid line/circles represent standard curve.
equivalents. BCFIA Determination of Bacterial Equivalents in a Plaque Sample Based on the Standard Curve After testing a variety of models of least squares curvefitting, a linear model of the logarithm of dependent and independent variables provided the best fit. The standard curve has the form log y a + b log x, where y indicates fluorescence and indicates relative number of bacterial cells. The fluorescence of a plaque sample corrected for background was determined in the BCFIA and compared to the standard curve. Based on the standard curve, plaque sample fluorescent units were converted into relative numbers of bacteria. An example of converting fluorescent values obtained for a plaque sample using the BCFIA into specific P. gingivalis bacterial counts is illustrated in Figure 4. The relationship —3.88 + (1.36 log x) was derived for the log y standard curve with both logarithms corresponding to base 10. A plaque sample with a fluorescence of 18,962 and corresponding background fluorescence of 2,132 units was determined in the BCFIA to have 9.13 x 105 P. gingivalis cells; see "A" in Figure 4. Solving the above equation for "x" we get: =
=
log
=
·
(log (18,962
-
2,132)
+
3.88)/1.36
5.96 9.13 x 105 P.
=
gingivalis cell equivalents. Twenty µ of the original 500 µ of the plaque sample was used in the immunoassay. Therefore, the number of P. gingivalis cells in the plaque sample collected was determined by multiplying 9.13 x 105 by 25 (500 µ1/20 µ ) to give 2.28 x 107 P. gingivalis cell equivalents in the total plaque sample collected. Another example of using the =
standard curve to determine the relative number of moderate
levels of P. gingivalis cells in a plaque sample is described below and is illustrated in "B" of Figure 4:
log
=
(log (5,591 1,103) + 3.88/1.36 (log (4,488) + 3.88)/1.36 -
=
=
=
3.45
x
7.53/1.36
105 P.
=
5.54
gingivalis cell equivalents.
Comparison of BCFIA with Cultivable Flora and DNA Probe in Detecting Plaque Periodontopathogens For these experiments, we wanted to determine how the BCFIA compared with cultivable flora in detecting 4 bacterial pathogens in a large number of plaque samples. Sites included had a periodontal status ranging from healthy to severe Periodontitis. Statistical comparison of the BCFIA with cultivable flora in detecting periodontopathogens in plaque may be seen in Table 4. Sensitivity was 100% for A. actinomycetemcomitans, P. gingivalis, and P. intermedia, while the sensitivity for F. nucleatum was 97% when the BCFIA was compared with cultivable flora. The specificity of the BCFIA when compared with cultivable flora methodology was 68% for A actinomycetemcomitans, 57% for P. gingivalis, 73% for P. intermedia, and 92% for F. nucleatum. In addition to cultivable flora, the BCFIA was compared to the DNA probe in detecting periodontopathogens in plaque. In evaluating plaque samples from patients with severe Periodontitis, the BCFIA agreed with the DNA probe results 75, 71, and 82% of the time for A. actinomycetemcomitans, P. gingivalis, and P. intermedia, re-
spectively (Fig. 5).
DISCUSSION The BCFIA is an effective immunoassay system for the screening of bacterial mixtures for specific pathogenic bac-
Volume 63 Number 12
WOLFF, ANDERSON, SANDBERG, REITHER, BINSFELD, CORINALDESI,
Table 4.
Comparison of Bacterial Concentration Fluorescence Immunoassay (BCFIA) With Cultivable Flora in Detecting Plaque Periodontopathogens A.
P.
actinomycetemcomitans
cultivable
flora
flora
+
BCFIA
4 0
+
+
25 52
-
BCFIA
-
+
+
5 0
33 43
-
-
sensitivity: 100% specificity: 68%
sensitivity: 100% specificity: 57%
P. intermedia
F. nucleatum
cultivable
cultivable
flora
flora
+
BCFIA
gingivalis
cultivable
30 0
+
14 37 -
BCFIA
-
+
67 2
1 11 -
-
sensitivity: 100% specificity: 73%
sensitivity: 97% specificity: 92%
100% m o *.
Bacterial Concentration Fluorescence Immunoassay (BCFIA) for the Detection of Periodontopathogens in
Plaque Larry F. Wolff, * LuAnn Anderson, * Gregory P. Sandberg/ Lowell Reither, * Christine A. Binsfeld/ Giuseppe Corinaldesi,* and Charles E. Shelburnef A BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAY (BCFIA) Was developed tO rapidly detect periodontopathic bacteria in human plaque samples. The BCFIA utilized fluorescent-tagged monoclonal antibodies (MAbs) directed against the lipopolysaccharide of selected Gram-negative bacteria. Microorganisms identified in plaque using the BCFIA
included Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, and Fusobacterium nucleatum. The immunoassay procedure involved combining a patient's plaque sample with a species-specific fluorescein isothiocyanate-labeled
MAb and then incubating the mixture in a specialized microtiter plate allowing the MAb to bind to its homologous bacteria. Bound and unbound fluorescent-tagged MAbs were separated by filtration and total bound bacterial fluorescence was determined with a fluorimeter. The relative number of a bacterial species in a given plaque sample was estimated by reference to a standard curve carried through the BCFIA. The BCFIA had a lower detection limit of near 104 specific bacterial cells in a mixed bacterial preparation or plaque sample. When compared to cultivable flora procedures in detecting the 4 periodontopathogens, the BCFIA had high levels of statistical sensitivity, 97% to 100%, while statistical specificity ranged between 57% and 92%. There was a 71% to 82% agreement between BCFIA and DNA probe methodology in detecting periodontopathogens in plaque. The BCFIA, when compared to cultivable flora, offers the advantage of evaluating both live and dead bacterial cells in plaque. This may in part, if not fully, explain the lower specificity values of the BCFIA when compared to cultivable flora. Screening plaque samples for periodontopathic bacteria is considerably faster and results in a greater frequency of detection with BCFIA than cultivable flora based methods. / Periodontol 1992; 63:1093-1101.
Key Words:
Dental
plaque/microbiology; immunoassay,
fluorescence.
Although the gingival crevice supports the growth of a wide variety of microbial species, only a few of these microorganisms have been closely associated with periodontal disease.1"4 The identification of these microorganisms in subgingival plaque using cultural methods, although precise and specific, is labor-intensive, time-consuming and requires the viability of the bacterial cells. This severely limits research surrounding the evaluation of plaque for periodontal pathogens, especially in large population groups. Additional techniques are needed to more rapidly assess plaque for specific bacteria and to a certain extent these technologies, DNA probes, enzyme-linked immunosorbent "University of Minnesota Clinical Research Center for Minneapolis, MN. tBiosciences Laboratory, 3M Company, St. Paul, MN.
eases,
Periodontal Dis-
(ELISA), and immunofluorescence, have been developed.5"11 Jolley and co-workers introduced a particle concentration fluorescence immunoassay (PCFIA) which utilized polystyrene beads coated with Immunoglobulin as the solid phase.12 These investigators reported the PCFIA was as or more sensitive and could more quickly evaluate samples than the ELISA. Other investigators have used the PCFIA to quantitate Immunoglobulins and interleukin-4 in human serum and also alpha-1 antitrypsin from Escherichia coli lysates.13"16 A modified PCFIA was developed using bacterial cells as the solid phase, instead of polystyrene beads.17 The bacterial solid phase, Escherichia coli, was used to analyze both polyclonal and monoclonal antibodies against surface antigens of this bacterium. These investigators inassay
J Periodontol 1094
BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAV
dicated the bacterial concentration fluorescence immunoassay (BCFIA) showed up to 50-fold greater sensitivity in bacterial cell detection and was more quickly performed than the ELISA. We have modified and utilized the BCFIA for the rapid identification of periodontopathic bacteria in human plaque specimens. Monoclonal antibodies have been used in the BCFIA by our laboratory while both monoclonal and polyclonal antibodies have been utilized with indirect immunofluorescence microscopy by other investigators.9'18 21 The differentiation between monoclonal and polyclonal antibody production has been thoroughly described by Milstein.22 In this report, emphasis will be placed on a series of monoclonal antibodies which have been developed to a selected group of periodontal disease-associated pathogens including Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis, Prevotella intermedia, and Fusobacterium nucleatum.2324 These antibodies recognize specific sites on the lipopolysaccharide (LPS) molecules of these microorganisms and form the basis for the BCFIA. Bacteria are the solid phase to which the LPS-specific fluorescein isothiocyanate-tagged MAbs are directed. The bound and unbound fluorescent-tagged MAbs are separated by filtration in a modified microtiter plate and the total bacterialbound fluorescence is determined with a fluorimeter. The BCFIA described in this report offers the advantage of evaluating both live and dead bacterial cells in plaque and with respect to screening plaque samples for periodontopathic bacteria is considerably faster than cultivable flora-based methods. MATERIAL AND METHODS Bacterial Strains The strains of bacteria used in this investigation were obtained from the American Type Culture Collection (ATCC) (Table 1). Bacteria were grown in media as recommended by the ATCC or under conditions as previously de-
scribed.1-25"28
Monoclonal Antibody Production, Screening and Fluorescein Isothiocyanate Labeling Hybridizations between Balb/c NS-1 myeloma cells and spleen cells from immunized Balb/c mice were performed as described previously.29 Briefly, Balb/c mice were immunized with a bacterial strain, either A. actinomycetemcomitans 29522, P. intermedia 25611, F. nucleatum 25586, or P. gingivalis 33277, 3 to 5 times by intraperitoneal injection with a final injection of 10 million cells given intravenously 3 days prior to aseptic removal of the spleens. Splenic leukocytes were then combined with a NS/1 myeloma cell line in 35% polyethylene glycol. The resulting hybridoma culture supernatants containing mouse Immunoglobulin were screened for the production
of antibacterial antibody using either an ELISA30 and/or a BCFIA.31 Cultures represented by supernatants with sig-
December 1992
Table 1. Bacterial Strains Used in Cross-Reactivity Testing A. actinomycetemcomitans, P. gingivalis, P. F. nucleatum Monoclonal Antibodies. A.
actinomycetemcomitans 29522 (b)* 29523 29524
A. A. B. B. C. C. C. E. E. E. F. H. H. H. . .
israelii 12102
(a) (b)
naeslundii 12104
forsythus 43037 fragilis 25285
concisas 33237 ochracea 33569 sputigena 33612 alactotyticum 23263 corrodens 23834 limosa 8486 nucleatum 25586* 23726 10953 aphrophilus 13252 influenzae 8142 9795 segnis 7901 sicca 9913 subflava 14799
Experiments for intermedia, and
P. buccae 33573 33574 P. denticela 33185 P. intermedia 25611* 33563 P. loescheii 15930 P. melaninogenicus 15032 15033 P. gingivalis 33277* P. micros 33270 S. mutans 10449 27607 S. sanguis 10566 S. sobrinus 33478 S. sputigena 33150 T. denticela 33405 33520 33521 T. vincentii 35580 V. caviae 33540 W. curva 335244 W. recta 33238
"Indicates those bacterial strains to which monoclonal
veloped.23"24
(Supplement)
antibody was
de-
nificantly elevated antibody binding over negative controls, that is the mean relative fluorescence units of the negative controls plus 3 standard deviations, were expanded for further examination. Positive controls using diluted serum from the immunized mice were included along with hyperimmune rabbit serum to each bacteria. Supernatants exhibiting reactivity were then cloned by limiting dilution in 96-well plates. Stable cloned hybrids were expanded by growth in culture and approximately 107 cells were injected intraperitoneally into Balb/c mice to induce ascites. The ascitic fluid was collected, antibodies purified using mixed bed ion-exchange HPLC* and dialyzed in 0.05 M sodium carbonate buffer, pH 9.5 at 2 mg/ml. For the BCFIA (described below), MAb was labeled by dialyzing 3 ml of purified antibody against fluorescein isothiocyanate (FITC), 3 mg FITC/100 ml carbonate buffer for 18 hrs, and then performing chromatography over G-25 sepharose5 to remove free FITC. The fluorescein to protein ratio (F/P) was determined by comparing the molar amounts of FITC and protein at 495 and 280 nm, respectively. Acceptable preparations for use in the BCFIA had a F/P ratio of between 3 and 12.
Monoclonal Antibody Binding to Purified LPS LPS was prepared from 500 mg of lyophilized late log phase cultures of Treponema denticola, Eikenella corrodens, Haemophilius aphrophilus, A. actinomycetemcomitans, P. intermedia, F. nucleatum, and P. gingivalis using the method of Darveau and Hancock.32 Escherichia coli, Salmonella minnesota, and Vibrio cholera LPS were com*Bakerbond ABX, JT Baker Research §Pharmacia LKB, Piscataway, NJ.
Products, Phillipsburg, NJ.
Volume 63 Number 12
WOLFF, ANDERSON, SANDBERG, REITHER, BINSFELD, CORINALDESI,
mercially obtained.11 Reactivity of each LPS preparation with the 4 MAbs was evaluated using an ELISA. Briefly, purified LPS was resuspended in PBS at 100 vg/ml and 50 µ amounts of each placed in individual wells of 96-well microtiter plates. After incubation at room temperature for 18
hours the unbound LPS was removed and 300 µ of PBS with 0.25% bovine serum albumin was added for 4 hours. The plates were then washed 3 times with PBS-Tween 20. Fifty µ of MAb was added to each of the wells and incubated for 1 hour followed by 3 washes with PBS-Tween. A peroxidase-labeled goat anti-mouse IgG was then added (50 µ /well) and incubated for an additional hour. After 3 additional washes in PBS-Tween 20, 50 µ of peroxidase substrate (1 mg/ml o-phenylenediamine, 0.03% hydrogen peroxide, 0.05 M sodium phosphate, pH 4.5) was added and the reaction stopped with 50 µ of 2 M H2S04 after 30 minutes. The optical density was determined at 490 nm using a microplate reader.' Monoclonal Antibody Specificity Forty-four ATCC strains of bacteria were washed 3 times 107 bacteria/ml. in PBS and resuspended in PBS at 5 to individual of each of the strains added were µ Twenty wells in a fluoricon assay plate.* Twenty µ of either FITClabeled anti-bacterial MAb, 10 µg/ml, or a similarly labeled MAb with a non-bacterial specificity, which served as a negative control, was then added to the wells and the plates incubated 20 minutes at room temperature in the dark. Plates were washed twice with PBS-NP40, and the bound fluorescence determined. The percent cross-reactivity was calculated by dividing the difference of the binding of heterologous anti-bacterial MAb and the non-bacterial MAB by the total bacterial MAB bound in a homologous reaction, followed by multiplication by 100. Bacterial Concentration Fluorescence
Immunoassay
(BCFIA) The general principle of the BCFIA has been described in an earlier report.31 The modified BCFIA in relation to evaluating plaque samples or mixed bacterial cultures for specific bacteria or identifying bacteria in a single bacterial suspension is outlined in Figure 1. The BCFIA was performed in a specially designed 96-well microtiter plate.1 Whole bacterial cells served as the solid phase in the BCFIA procedure. Prior to carrying each plaque sample or bacterial preparation through the immunoassay, cells were washed twice in distilled water-0.05% Tween and then resuspended in 500 µ PBS-0.05% Tween plus 1% goat serum. Twenty µ from 500 µ of the bacterial or plaque cell suspension was placed in each of 2 wells of the microtiter plate. To 1 well was added 20 µ of a bacterial species-specific FITC-
"Sigma, St. Louis, MO. 'Dynatech, Chantilly, VA. *IDEXX Corporation, Portland,
^ •000000000000 cooooooooooor
»oooooooooodL 000000000000
c000000000000
®
Bacteria
1095
MAB Bound to
(Solid Phase)
Specific Bacteria
^OOOOOOOOOOOO F! (jo resesi
Isothiocyanate
Labeled MAB (20µ )
w
gingivalis.
intermedia,
azi neTiyc01C~!CCTii'^S. F. nucleatum
Washed
Plaque Suspension
© Unbound MABj Pass Through
Filter. 0.2µ Pore Size
.
--
*
Figure 1. Principles of the BCFIA procedure for evaluating plaque samples, mixed bacterial cultures, or a specific bacterial suspension. labeled** MAb and incubated in the dark at 22°C for 20 minutes and then washed twice with PBS plus 0.0125% NP-40 and filtered under 25 mm vacuum. The other well containing the bacterial or plaque suspension served as a control and was treated with a non-bacterial binding antibody, FITC labeled MOPC, a myeloma protein at a concentration of 5 µg/ml and a F/P of 10, and assayed in parallel to the test mixture. Additional controls included carrying buffer as well as an FITC-labeled MAb suspension through the immunoassay without plaque or bacteria. Species specific MAbs utilized in the immunoassay included those against P. gingivalis, P. intermedia, A. actinomycetemcomitans, and F. nucleatum. The pore size of the membrane filter was 0.2 µ , allowing bacteria with or without tagged MAb to be trapped on the membrane surface while the unbound MAb passed through the filter with the buffer wash. Total bacterial bound fluorescence (485 excitation nm) was determined in a fluorimetertt at a wavelength of 535 nm. Total bound bacterial-specific fluorescence was obtained by subtracting non-specific fluorescent binding in the control well from the total fluorescent antibody binding in the test well. The relative bacterial cell equivalents were determined by comparison to a standard curve (see below). BCFIA Evaluation with Single Cultures of Periodontopathic Bacteria The sensitivity or bacterial detection limit of the immunoassay for screening single cultures of P. gingivalis, P. intermedia, F. nucleatum, or A. actinomycetemcomitans was determined using the BCFIA as described above. Twenty µ of a washed log-phase bacterial cell preparation were added to wells of the immunoassay plate followed by 1 of the FITC-labeled species-specific MAbs. The final total number of bacteria carried through the BCFIA ranged be"Isomer I, Research Organics, Cleveland, OH. fluorescence Concentration Analyzer, IDEXX
ME.
SHELBURNE
ME.
Corporation, Portland,
J Periodontol 1096
BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAY
tween
102 and 106. The MAb concentration and F/P for
species-specific MAb preparation used in these experiments were: P. gingivalis 2.6 µ^ !, 4.6; F. nucleatum 3.7 µ& 1, 4.2; A. actinomycetemcomitans 4.3 µ / , 4.5; P. intermedia 7.1 µ& 1, 11.1; respectively. each
BCFIA Evaluation with Periodontopathic Bacteria in a Mixed Culture In an effort to determine if a mixed culture of bacteria would have an effect on the ability of the BCFIA to detect
specific periodontopathic bacteria, competition experiperformed. A description of these experiments is given in Table 2. Dilution buffers A, B, C, and D were prepared by adding log phase cultures of the Gram-negative bacteria P. gingivalis, P. intermedia, F. nucleatum, A. actinomycetemcomitans, E. corrodens, and/or the Gram-positive bacterium Streptococcus mutans to phosphate-buffered saline, pH 7.4, at the concentrations shown. A portion of buffers A, B, C, and D were then spiked with 2.5 107/ ml P. gingivalis, P. intermedia, F. nucleatum, or A. actinomycetemcomitans, respectively. The final total concentration of bacteria in each buffer with the spiked bacteria was 108/ml. This suspension of bacteria was then diluted in buffers A, B, C, or D (without the spiked bacteria) and titrated against species-specific MAbs at a concentration of 1 µ& 1 using the BCFIA. With this dilution scheme, the concentration of the spiked bacteria decreased while the a
ments were
concentration of the other bacteria remained the same. The F/P for the MAb preparations used in these experiments were: P. gingivalis, 5.6; F. nucleatum, 6.2; A. actinomycetemcomitans, 4.2; and P. intermedia 7.1.
Preparation of the Standard Curve for the BCFIA The standard curve was established by examining the relationship between fluorescence and relative bacterial num-
December 1992
(Supplement)
whole bacterial cells. The LPS-coated beads were assayed with their homologous MAbs, using the same procedure as described above in the BCFIA. Simultaneously, suspensions of each of the 5 bacteria from log phase cultures were adjusted, using a Petroff-Hausser chamber, at concentrations ranging from 103 to 106 bacteria/well and then immunoassayed. Each preparation of antigen-coated beads was then diluted with bovine serum albumin plus 1% goat serumcoated beads to give a standard corresponding to a relative number of bacteria in the BCFIA. Specific bead preparations (P. gingivalis, P. intermedia, A. actinomycetemcomitans, or F. nucleatum) ranged in concentration from 0.250% to 0.016% (mg of beads/volume of buffer) and were used to prepare the standard curve. Calibrated bead suspensions were stored at 4°C until used in the BCFIA. Each microtiter plate which contained plaque or other bacterial samples also contained primary standards prepared in duplicate. The average of the 2 primary standards for 5 or more different bead concentrations was used to prepare the standard curve and detemiine the relative number of bacteria in test samples. Bacterial Concentration Fluorescence Immunoassay Compared with Cultivable Flora and DNA Probe in
Detecting Periodontopathogens
Plaque sampling procedures and microbial evaluation of plaque have been described in detail in earlier reports.1-28'33 Eighty-one sites in individuals with a periodontal status ranging from health/gingivitis (n 30) to mild, moderate, or severe Periodontitis ( 51) were included in the phase of the study comparing the BCFIA with cultivable flora. After removal of supragingival plaque, a subgingival plaque sample was obtained with a sterile curet and placed into 250 µ of pre-reduced anaerobically-sterilized diluent and immediately transferred to an anaerobic chamber. The plaque sample was further diluted with 2.0 ml of the pre-reduced diluent, sonicated for 20 seconds,88 serially diluted and plated on supplemented blood agar media for detection of total =
=
bers in the BCFIA. The least squares method was used to determine the best fitting mathematical model. Preparation of the standard curve utilized LPS-enriched coated polystyrene particles as the solid phase in the BCFIA. For each of the bacteria, P. gingivalis, P. intermedia, A. actinomycetemcomitans, and F. nucleatum, 200 ml log phase cultures were washed 3 times in Dulbeco's PBS buffer and resuspended in 20 ml of the same buffer, pH 7.4. The cells were sonicated** for 3 1-minute bursts at an output of 10% and centrifuged for 15 minutes at 12,000 xg to remove unbroken cells. The supernatant from each of the bacterial preparations was combined with 2.0 ml of 0.85 µ fluoricon polystyrene assay particles1 and rotated at 0.5 rpm for 12 to 18 hours at 22°C. The beads were then washed in PBS plus 0.0125% NP-40 and resuspended in 40 ml of the same buffer with 0.5% sodium azide. The LPS-coated beads were then standardized by immunoassay to suspensions of
anaerobic cultivable flora, Fusobacterium spp. and colonies showing dark pigmentation. The dark-pigmenting colonies were screened for the identification of P. intermedia and P. gingivalis using the AN-IDENT system1111 and the APIZYM system as described by Laughon and co-workers.34 Selective agar media was used for the detection of A. actinomycetemcomitans.35 A portion of each plaque sample 20°C until it was carried through the BCFIA. was stored at The phase of the study comparing the BCFIA with DNA probe for detecting periodontopathogens included 28 plaque samples collected from sites with greater than 5 mm probing depth (mean ± SD 8.0 mm ± 1.5 mm) and greater than 3 mm loss of attachment (mean ± SD 7.9 mm ± 1.9 mm). These plaque samples were from different individuals than those used in the cultivable flora studies. Two
**Branson Sonifier, Model 350, Branson Sonic Power
§§Kontes Cell
bury,
CT.
Company,
Dan-
-
=
=
Disruptor, Kontes, Vineland, NJ. "Analylab Products, Sherwood Medical, Plainview, NY.
Volume 63 Number 12
WOLFF, ANDERSON, SANDBERG, REITHER, BINSFELD, CORINALDESI,
Table 2. Protocol
Periodontopathic
Describing Bacterial
Bacteria in
a
Concentration Fluorescence Mixed Culture
1097
SHELBURNE
Immunoassay Evaluation of
Dilution Buffers*
(bacteria/ml)
Bacteria P. P. F. A. E. S.
gingivalis
0
107 107 107 107
intermedia nucleatum
actinomycetemcomitans corrodens
Spike
P. gingivalis P. intermedia F. nucleatum A. actinomycetemcomitans
2.5
Total bacteria/ml
*Bacteria
were
3.5
107
3.5
mutans
107 0 107 107 IO7 IO7
C
D
IO7 IO7 0 IO7 IO7
IO7 IO7 IO7 0 IO7 IO7
3.5
3.5
IO7
2.5
IO7
107
x
2.5
x
IO7 2.5
IO8
10s
IO7
x
IO8
IO8
suspended in phosphate buffered saline, pH adjusted to 7.4. Specificity of Monoclonal Antibody Binding to Lipopolysaccharides Based on EUSA23"24'40
Table 3. Porphyromonas gingivalis (ATCC 33277)
Actinobacillus
actinomycetemcomitans
(ATCC 29522)
Lipopolysaccharide from: A. Prevotella intermedia
Fusobacterium nucleatum
Relative Bacterial Numbers
Relative Bacterial Numbers
(ATCC 25611)
(ATCC 25585)
Figure 2. Species-specific monoclonal antibodies titrated against their homologous bacteria using the BCFIA. sterile paper points were simultaneously inserted to the base of a pocket. After 10 seconds, the paper points were removed. One was placed into 250 µ of pre-reduced diluent and carried through the immunoassay as described above. The other paper point was placed in a sterile vial and forwarded to a commercial laboratory11 for DNA probe analysis. RESULTS Monoclonal Antibodies Reacting With Homologous Bacteria The MAbs were titrated against their homologous bacteria using the BCFIA (Fig. 2). The immunoassay had a lower detection limit for pure cultures of P. gingivalis, P. intermedia, F. nucleatum, and A. actinomycetemcomitans of between 1000 and 10,000 bacteria. It may also be observed that an approximately linear response was seen between relative fluorescent units and bacteria over the range of 104 to
106 bacteria.
'OmniGene, Inc., Cambridge, MA.
actinomycetemcomitans
(Aa ATCC 29522) P. intermedia (Pi ATCC 25611) F. nucleatum (Fn ATCC 25586) P. gingivalis (Pg ATCC 33277) T. denticola (ATCC 33521) E. corrodens (ATCC 23834) H. aphrophilus (ATCC 13252) E. coli 055:B5 E. coli 0111:B4 S. minnesota Re595 V. cholerar569B
Purified
Optical Density 490 nm Antibody to: Fn Pg
Monoclonal Aa Pi 1.43 .000 .031 .002 .000 .000 .000 .000 .020 .000 .013
.000 .862 .000 .000 .071 .016 .003 .024 .000 .000 .000
.031 .000 .771 .002 .021 .019 .000 .021 .004 .000 .000
.002 .077 .002 1.33 .000 .021 .011 .000 .000 .000 .017
Specificity of MAbs Used in the Bacterial Concentration Fluorescence Immunoassay The specificity of the MAbs were evaluated against a battery of 44 ATCC bacterial type strains (Table 1). The heterologous cross-reactions with both Gram-positive and Gramnegative bacterial species were at least one-tenth and in most cases one-twentieth the intensity of the homologous reaction for each of the 4 bacterial-specific MAbs. The specificity of the MAbs for LPS was determined using purified LPS from a number of different bacteria as targets in an ELISA format (Table 3). The average optical density of triplicate wells demonstrated heterologous reactions were less than 5% of the homologous reactions. Detection of Periodontopathic Bacteria in a Mixed Culture Using the BCFIA The effect of a mixed culture of bacteria on the sensitivity of the BCFIA for detecting specific bacteria may be seen in Figure 3. A linear response was observed between fluorescent units and relative number of bacterial cells. In
J Periodontol 1098
O
BACTERIAL CONCENTRATION FLUORESCENCE IMMUNOASSAY
December 1992
(Supplement)
20 m
F. nucleatum
O
(25586)
c
o
actinomycetemcomitans
3
(29522) ß> >
ce
mixed
14 Relative Number
addition, even in the presence of high levels of other bacteria, using each of the 4 species-specific MAbs, the BCFIA had
a
lower detection limit value of
near
104 bacterial cell
105
10 6
Porphyromonas gingivalis Cells
Figure 4. Method used to convert fluorescence for a plaque sample determined by the BCFIA into Porphyromonas gingivalis cell numbers. Heavy solid line/circles represent standard curve.
equivalents. BCFIA Determination of Bacterial Equivalents in a Plaque Sample Based on the Standard Curve After testing a variety of models of least squares curvefitting, a linear model of the logarithm of dependent and independent variables provided the best fit. The standard curve has the form log y a + b log x, where y indicates fluorescence and indicates relative number of bacterial cells. The fluorescence of a plaque sample corrected for background was determined in the BCFIA and compared to the standard curve. Based on the standard curve, plaque sample fluorescent units were converted into relative numbers of bacteria. An example of converting fluorescent values obtained for a plaque sample using the BCFIA into specific P. gingivalis bacterial counts is illustrated in Figure 4. The relationship —3.88 + (1.36 log x) was derived for the log y standard curve with both logarithms corresponding to base 10. A plaque sample with a fluorescence of 18,962 and corresponding background fluorescence of 2,132 units was determined in the BCFIA to have 9.13 x 105 P. gingivalis cells; see "A" in Figure 4. Solving the above equation for "x" we get: =
=
log
=
·
(log (18,962
-
2,132)
+
3.88)/1.36
5.96 9.13 x 105 P.
=
gingivalis cell equivalents. Twenty µ of the original 500 µ of the plaque sample was used in the immunoassay. Therefore, the number of P. gingivalis cells in the plaque sample collected was determined by multiplying 9.13 x 105 by 25 (500 µ1/20 µ ) to give 2.28 x 107 P. gingivalis cell equivalents in the total plaque sample collected. Another example of using the =
standard curve to determine the relative number of moderate
levels of P. gingivalis cells in a plaque sample is described below and is illustrated in "B" of Figure 4:
log
=
(log (5,591 1,103) + 3.88/1.36 (log (4,488) + 3.88)/1.36 -
=
=
=
3.45
x
7.53/1.36
105 P.
=
5.54
gingivalis cell equivalents.
Comparison of BCFIA with Cultivable Flora and DNA Probe in Detecting Plaque Periodontopathogens For these experiments, we wanted to determine how the BCFIA compared with cultivable flora in detecting 4 bacterial pathogens in a large number of plaque samples. Sites included had a periodontal status ranging from healthy to severe Periodontitis. Statistical comparison of the BCFIA with cultivable flora in detecting periodontopathogens in plaque may be seen in Table 4. Sensitivity was 100% for A. actinomycetemcomitans, P. gingivalis, and P. intermedia, while the sensitivity for F. nucleatum was 97% when the BCFIA was compared with cultivable flora. The specificity of the BCFIA when compared with cultivable flora methodology was 68% for A actinomycetemcomitans, 57% for P. gingivalis, 73% for P. intermedia, and 92% for F. nucleatum. In addition to cultivable flora, the BCFIA was compared to the DNA probe in detecting periodontopathogens in plaque. In evaluating plaque samples from patients with severe Periodontitis, the BCFIA agreed with the DNA probe results 75, 71, and 82% of the time for A. actinomycetemcomitans, P. gingivalis, and P. intermedia, re-
spectively (Fig. 5).
DISCUSSION The BCFIA is an effective immunoassay system for the screening of bacterial mixtures for specific pathogenic bac-
Volume 63 Number 12
WOLFF, ANDERSON, SANDBERG, REITHER, BINSFELD, CORINALDESI,
Table 4.
Comparison of Bacterial Concentration Fluorescence Immunoassay (BCFIA) With Cultivable Flora in Detecting Plaque Periodontopathogens A.
P.
actinomycetemcomitans
cultivable
flora
flora
+
BCFIA
4 0
+
+
25 52
-
BCFIA
-
+
+
5 0
33 43
-
-
sensitivity: 100% specificity: 68%
sensitivity: 100% specificity: 57%
P. intermedia
F. nucleatum
cultivable
cultivable
flora
flora
+
BCFIA
gingivalis
cultivable
30 0
+
14 37 -
BCFIA
-
+
67 2
1 11 -
-
sensitivity: 100% specificity: 73%
sensitivity: 97% specificity: 92%
100% m o *.
