Oral Microbiol Immunol 1990: 5: 275-279

Benzoyi-arginine naphthylamide(BANA) hydroiysis by Treponemadenticoia and/or Bacteroidesgingivalis in periodontal piaques

W, A, Bretz, D. E. Lopatin,W. J, LoescheDepartment of Biologic and IWaterialsSciences, School of Dentistry, University offvliehigan, Ann Arbor, Michigan, USA

Bret: WA, Lopatin DE, Loesche WJ. Betizoyl-arginine naphthylamide (BANA)hydrolysis by Treponema denticoia andjor Bacteroides gingivalis in periodontalplaques.Oral Microbiol Immunol 1990: 5: 275-279.

Treponema denlieola and Bacteroides gingivalis are among the few recognizedspecies found in periodontal pockets that can hydrolyze the synthetic peptide N-benzoyl-DL-arginine-2-naphthylamide (BANA), We determined the presence ofthese periodontal pathogens in BANA-positive and -negative plaque samplesthrough the use of indirect immunofluorescence antibody techniques. Eighteen of27 diseased sites gave BANA-positive reactions, and 9 gave BANA-negativereactions, F. denticoia was present in 16 of 18 BANA-positive reactions, whereas*• gingivalis was detected in 9 of the 18 BANA-positive reactions, F. denticoia*as present in 1 and B. gingivalis in 2 of the 9 BANA-negative reactions. Neitherorganism was detected in the 19 healthy sites that were negative for BANA, Altmeasured differences between BANA-positive and BANA-negative ptaques ob-tained in the same individuats were statistically significant. The accuracy of theB A N A test, compared with clinical parameters such as bleeding upon probingand increased probing depth, was about 80%, The accuracy of the test in detectingthe presence of F. denticoia was 93%, for B. gingivalis. 76% and for F. denticoiaand/or B. gingivalis. 96%, This study indicated that BANA-positive plaques wereassociated with the presence of F. denticoia and/or B. gingivatis, that F. denticoia*as found at a greater frequency and levels in BANA-positive plaques than E.Singivalis, and that the presence of these organisms was associated with clinicaldisease.

Key words: BANA test; immunofluorescence;Treponema denticoia: Bacteroides gingivalis:periodontal plaque

Walter A. Bretz, DDS, f^PH, 3204 Sehool ofDentistry, University of Miefiigan, Ann Arbor,f^l 48109-1078, USA

Aecepted for pubiieation Oetober 20, 1989

Tfeponetna denticoia (a small spiro-chete) and Bacteroides gingivalis have°wn strongly associated with perio-dontal lesions (7, 8, 18, 19), Both organ-'sms produce a number of virulence fac-tors that may play a role in periodontaltissue destruction, which would includeCollagenase and lipopolysaccharides (3,^' 21), Given the itrtportance of these 2Organisms in periodontal disease, rapid^rid simple assays for monitoringSubgingival plaques for these perio-dontal pathogens would have great clin-'cal value,

T. denticoia and B. gingivalis areamong the few recognized species found"1 gingival crevices or periodontalPockets that can hydrolyze the syntheticPfiptide N-benzoyt-DL-arginine-2-"'iphthylamide (BANA) (4, 5), It was•" tbund that Bacteroides forsy-

thus also possesses BANA activity (20),BANA hydrolysis by plaque sampleshas been significantly associated withincreasing probing depth and attach-ment loss and with high levels and pro-portions of spirochetes (9), tt was notpossible to associate BANA hydrolysiswith F denticoia. as this organism can-not be quantitatively recovered fromplaque samples by cultural procedures(17), The plaque BANA assay was sub-sequently shown to be a sensitive andspecific screening tool for recognizinguntreated and treated sites in perio-dontal patients (15),

tn this investigation, we determinedthe presence of F. denticoia and B. gin-givalis in BANA-positive or -negativeplaque samples, through the use of in-direct immunotluorescence techniquesemploying polyclonal antibodies specif-

ic to each organism. Thus, this reportfocuses on comparisons of BANA hy-drolysis with clinical parameters andwith the presence and levels of theseperiodontopathogens in the plaque.

Material and methods

Subgingival plaque samples {2-A sitesper patient) were collected by means ofa sterile curette from 17 untreated adultperiodontitis patients with no history ofantibiotic usage in the past 6 tnonths, Atotal of 46 samples were collected from27 diseased sites and 19 healthy sites.The supragingival plaque was removedand discarded, and subgingival speci-mens were inoculated into a vial con-taining 0,2 ml of Sorensen phosphatebuffer (pH 7,2) and dispersed for 20 s

276 Bretz et al.

Utilizing a vortex mixer (working sus-pension).

Dark-field microscopy

Ten microUters of the working suspen-sion was ptaced on a gtass slide, coveredwith a cover slip, sealed, and examinedunder a dark-field microscope. The totalnumber of bacteria per high-powermicroscopic field (HPF) and the totalnumber of spirochetes/HPF (whichwere further subdivided into small, in-termediate and large spirochetes) wereobtained by counting 200 organisms orthe number of organisms in 20 HPF,whichever event occurred first,

BANA assay

One hundred microliters of the syntheticsubstrate n-benzoyl-DL-arginine-2-naphthylamide (BANA), previously di-luted (to p\ of 44 mg BANA/mt inDMSO, in 1 ml Sorensen buffer), wasadded to 100 p\ of the working suspen-sion and incubated in aerobic con-ditions (37 °C) for an average of 18 h.

The BANA test resutts were read visu-ally using scoring procedures describedelsewhere (1), Separate studies usingpure cultures of B. gingivalis and F. den-tieola demonstrated that weak-positiveresults contained about 5 x 1 0 ' colony-forming units (CFU) of B. gingivalis andt X 10" CFU of F. denticoia (10), Forpurposes of statistical analysis, weak-positive and positive results weregrouped as positive resutts.

indirect immunofluorescence assay

The remaining plaque suspension (90/il) was added to an equal volume of1 % formaldehyde, yielding a final con-centration in the reaction mixture of0,5% formaldehyde, and stored at 4 Cfor further analysis.

Ten microliters of the formaldehyde-fixed subgingival plaque samples wasapplied to a 12-wett microscope stide,heat-fixed and Gram-stained, Thesecounts provided total bacterial countsper HPF. Subsequently, 10 p\ of theptaque samples was applied to a second12-well stide, heat-fixed and stored at

Table L Cross-reaction between selected antibaeterial antibodies

Polyelonal antibody (1/1000 titer)

Baeteria

Aclinomyces viseosus(human isolates)Fusobacterium nucleatum(human isolates)Streptoeoccu.s sanguis(ATCC 10556)Selenomonas .sputigena(human isolates)Aetinobaeiliu.i actinomycetemeomitans (Y4)Bacteroides gingivalis(ATCC 33277)Bacteroides gingivalis(human isolates)Bacteroides intermedius(human isolates)Bacteroides melaninogenicus(human isolates)Capnoeytophaga oehraeeus(human isolates)Treponema denticoia(ATCC 35405)Treponema vincentii(ATCC 35580)Treponema socranskii(ATCC 35536)UnspeeiatedTreponema

T. denlieola(ATCC 35405)

0"

0

0

00

0

ND

ND

ND

ND

-1-4

+2

+ 2

+ 2

B. gingivaiis(ATCC 33277)

0

0

0

00

+ 4

-1-4

0

0

0

0

0

0

0

' Charaeteristies of positive eells: -1-4: brilliant fluoreseenee with good deftnition of eell wall,dark eenter; + 3 : moderate fluorescence with good definition of eell wall, dark center; -1-2:faint fluoreseenee with poor defmition of eell wall; + 1: barely deteetable fluorescence, singleeells not distinguishable; 0: no detectabte ftuorescence, ND = not done.

room temperature. Two wetts were usedfor a positive controt (pure bacteriat cul-ture + antibody + fluorescein isothiocy-anide (FtTC)-conjugate) and a negativecontrot (pure bacterial culture-H buf-fer -h FITC-conjugate),

The plaque samples were stained for10 min with \Q p\ of antisera diluted toworking titers in PBS containing 0,05%Tween 20, washed in phosphate-buf-fered saline (PBS) and rinsed with dis-tilled water. The slides were then incu-bated with 10 p\ of goat anti-rabbit IgGconjugated with FtTC, Slides wereagain rinsed and washed, and subse-quently mounted with glycerol in PBS(2:1 v/v) containing paraphenylenedia-mine (pH 9,0), sealed with a cover slipand nail polish,

Immunofluorescence was evaluatedwith a Leitz Dialux microscopeequipped with a Ploempak 2,3 fluor-escence ittuminator for epiftuorescence.The light source was a 200-W HBO mer-cury superpressure lamp. Fluorescencewas graded from 0 to 4-1-, with grades3+ and 4 + considered serologicallypositive reactions. Staining was con-sidered positive for the selected micro-organism(s) if it exhibited strong fluor-escent cells with well defined outlinesand dark or Ughtly shining centers, com-prising more than 1% of the total cellcount. Data were calculated as positivecells per HPF, The antibodies used onlystained celts with a spirochetal mor-phology in the case of the F. denticoiaantibody, or cells with a cocco-baciHimorphology in the case of the B. gin-givalis antibody, thereby indicating thespecies specificity of these antibodies.

Antibodies

American Type Culture Collection(ATCC) reference strains of B. gingivalis(ATCC 33277) and F. denticoia (ATCC35405) were grown as previously de-scribed (t2, t4). The cultures weregrown under anaerobic conditions(85% N2, 10% H2, and 5% CO2) at35°C until maximum turbidity wasreached (about 3 to 5 d). The bacteriawere harvested by centrifugation andwashed 3 times in 0.15 M NaCt, Thecetls were then resuspended in I % form-aldehyde in PBS and stored at 4 C ior4 d. The celts were then washed 3 time*in PBS to remove formaldehyde and ly'ophilized,

High-titer polyclonal rabbit anti-bodies were prepared by subcutaneousimmunization of female New Zealand

BANA hydrolysis by T. denticola and/or B, gingivalis 277

Table 2. Frequency distribution of clinical status, BANA hydrolysis and presence of T. denticolaand/or B. gingivatis in subgingival plaque satnples

Presence of speeies in plaque

Treponema denticolayesno

Baeteroides gingivalisyesno

Either presentBoth absent

Diseased (n = 27)BANA hydrolysis

positiven = I 8

162

99

180

negative

18

27

27

Healthy (n=19)BANA hydrolysis

positiven = 0

00

00

00

negative«=19

019

019

019

white rabbits with 2 mg of the lyophil-ized immunization antigen in completeFreund's adjuvant at 0 and 1 week, fol-lowed by booster immunization in in-complete Freund's adjuvant at 7 weeksfollowing first immunization, Pre-im-fnune and immune (post 7 weeks) bleed-ings were obtained and antibody titersdetermined. All sera were aliquotted3nd frozen.

Clinical parameters of periodontai sites

Diseased (n = ll) or healthy (/7=19)Periodontal sites were identified basedon bleeding upon probing (yes or no),probing depth measurements, and theclinical judgement of whether or notthat periodontal site should undergoPeriodontal treatment beyond that as-sociated with a routine dental cleaningby a hygienist,

Statisticai anaiysis

Frequency tables, 4-fold tables andPaired ^-tests were calculated using the^AS computer program (SAS Institute,Cary, NC).

Resuits

Optimal working titrations for both B.gingivalis and T. dentieola antibodies forindirect immunofluorescence werefound to be approximately 1/1000. Theanti-7: denticola antibody reactedstrongly with the homologous strain,exhibited low cross-reactivity with Tre-ponema vincentii, Treponema soeranskiiand an unspeciated Treponema, but didnot react with the other tested species(Table 1), A score of 2+ denotes a faintfluorescence with poor definition of cellwall. Thus, the cross-reaction observedwithin spirochetes did not prevent usfrom recognizing T dentieola specifi-cally in the plaque sample. The anti-bodies prepared against B. gingivalis re-acted only with the homologous strains,thereby demonstrating the species speci-ficity of this immunologic reagent(Table 1),

Ninety-five percent of the healthysites did not bleed upon probing, where-as 89% of the diseased sites bled uponprobing. Ninety-five percent of thehealthy sites had probing depths <3mm, whereas 100% of the diseased sites

Table 3. Mean (average + SD) differences between BANA-positive and BANA-negativeplaques

BANA-positive(1=11)

BANA-negative(paired r-test)

IniniunofluorescenceTreponema dentieolaBaeteroides gingivalis

IJark-field mieroscopyTotal bacteriaTotal spirochetesSmall spirochetesIntermediate spirochetesLarge spirochetes

ClinicalProbing depth

' Average ±SD.

21.0 + 28.63.0± 3.0

18.8+13,86.0± 4.72.8+ 1.81.8+ 1.51,3± 1.5

6.5 ± 1.9

0,0 ±0,00.3 +0,6

5,3 +4.80.1 ±0.30.02 + 0,030.06 + 0.10.05±0.1

2.6 ±0.9

<0.034<0.010

<0.0I7<0.001<0.001<0.002<0.018

<0.001

had probing depths >4 mm. Table 2shows the frequency of diseased andhealthy sites, BANA hydrolysis and im-munologic detection of T denticola and/or B. gingivalis in the plaque samples.Eighteen of 27 plaques from diseasedsites gave BANA-positive plaques, and9 plaques from diseased sites gaveBANA-negative plaques. Indirect im-munofluorescence analysis revealed T.denticola in 16 of 18 BANA-positiveplaques and B. gingivalis in 9 of 18BANA-positive plaques. T. denticolaand B. gingivalis were present in 1 and2 (respectively) of the 9 BANA-negativeplaques. All 19 healthy sites were nega-tive for BANA and T. dentieola and/orB. gingivalis were not detected.

All measured differences betweenBANA-positive and BANA-negativeplaques obtained in the same individ-uals were statistically significant (Table3). BANA-positive plaques harbored T.dentieola at levels of 21 cells/HPF,whereas B. gingivalis was detected atlevels of 3 cells/HPF. Similarly, totalnumber of bacteria/HPF as well as thelevels of spirochetes as counted by dark-field microscopy were significantly dif-ferent between BANA-positive andBANA-negative plaques. BANA-posi-tive reactions were associated withplaques taken from pockets 6.5 mmdeep on average, whereas BANA-nega-tive results were associated with plaquestaken from shallow pocketes of 2,6 mmon average.

The sensitivity and specificity of theBANA test reaction measured againstthe clinical parameters is shown in Table4, One would be interested in determin-ing the overall ability of the test, todistinguish between diseased andhealthy sites. This can be derived by thesum of true positives plus true negativesdivided by the total number of compari-sons (accuracy). When the BANA testwas measured against the subjectivelydetermined clinical parameters, the ac-curacy of the BANA test was about80% (Table 4). The accuracy of theBANA test in terms of detecting thepresence of species in plaques (objec-tively measured parameters) was 93%for T. denticola, 76% for B. gingivalisand 96% for T. dentieola and/or B. gin-givalis (Table 5),

Discussion

Several studies have associated B. gin-givalis with periodontal disease status(7, 18, 19, 22) and a few have associated

278 Bretz et al.

Table 4. Relationship between BANA test and clinical parameters

Clinicaliudgement"

Bleeding uponprobing''

Probingdepth'

SensitivitySpecificityFalse ( - )False ( + )Aecuracy

66.7100,033,30.0

80.0

64.090.536.0

9.576.1

66.7100.033.30.0

80.0

(+ + ) = BANA test positive; clinical parameter po.sitive. ( ) = BANA test negative; clinicalparameter negative. (—h) = BANA test negative; clinical parameter positive. (H—) = BANApositive; clinical parameter negative. T = total number of observations. ' The judgement of ahealthy versus diseased site based on probing depth and bleeding on probing. '' Bleeding uponprobing; yes = positive; no = negative.' Probing depth; <3 mm = negative; > 4 mm = positive.

Table 5. Relationship between BANA test and T. dentieola and/or B. gingivulis

T dentieola B. gitigivalis

T. denticolaand/or

B. gingivalis

SensitivitySpecificityFalse ( - )False ( + )Accuracy

94.193.1

5.96.9

93.4

81,974.218.125.876.0

90,0100,0

10,00.0

95.7

( + + ) = BANA test positive; T. dentieola and/or B. gingivalis positive. ( ) = BANA testnegative; T. denticola and/or B. gingivalis negative, (—I-) = BANA test negative; T. denticotaand/or B. gingivalis positive. ( + - ) = BANA test positive; T. denticola and/or B. gingivalisnegative. T = total number of observations.

T. dentieola (13, 16). This study, how-ever, is the first to use imtnunofluor-escence analysis to associate the fre-quency and the levels of both T dentieo-la and B. gingivalis in subgingivalplaque samples with periodontal healthor disease. Both organisms were presentin BANA-positive plaques of perio-dontally diseased sites, were detected ata low frequency in BANA-negativeplaques removed from diseased sites,and were absent from healthy sites thatwere unifonnly BANA-negative (Table2). The frequency of occurrence of T.denticola in BANA-positive plaques wasapproximately 2-foid higher than thatof B. gitigivalis (Table 2) and the levelsof T denticola in BANA-positiveplaques were greater than the levels ofB. gingivalis (Table 3).

The accuracy of the BANA test v.$clinical parameters never exceeded 80%(Table 4), perhaps indicating the subjec-tive nature of the clinical parameters,which are affected by tneasurement er-ror and examiner variability (6). In thisrespect, the BANA test may be a moreobjective measure of identifying perio-dontal sites potentially at risk fordeveloping periodontal destructionthrough the detection of periodontalpathogens, such as T. denticola, B. gin-givalis and Bacteroides forsythus. Alter-

nately, the failure to obtain 100% accu-racy with the clinical parameters couldindicate that, in some plaques, bacteriaother than T. denticola and/or B. gin-givalis were associated with clinical dis-ease.

The accuracy of the BANA test indetecting T denticota and/or B. gingival-is was about 96% (Table 5), whichmeans that the BANA test gives infor-mation that is as reliable as the infor-mation obtained with the immunologicreagents. Given the fact that T denticotacannot be quantified by culturalmethods (17), the BANA test could thenbe considered as a substitute for bothcultural and immunologic methods fordetecting T. denticola and possibly B.gingivalis in plaque samples.

The BANA test can detect T. dentico-la and B. gingivalis when 10' to 10* CFUare present in plaque samples (10).These detection limits make the BANAtest less .sensitive than other methods,such as DNA probes and cultural pro-cedures, which can detect 10' to 10'CFU. If clinical disease is preceded byor associated with an overgrowth ofthese periodontal pathogens in plaquesamples, then a less sensitive test maybe more indicative of an infection thatputs periodontal sites at risk for de-veloping periodontal disease.

Several methods, such as cultural,microscopic examination, immunologicreagents, enzyme activity, and DNAprobes, could help the clinician monitorplaques for the presence of periodontalpathogens in conjunction with clinicalparameters. The BANA test offers thepotential to identify plaques that mayharbor T. denticola, B. gingivalis and athird organism, B. forsythus, which hasrecently been associated with the pro-gressing periodontal lesion (2), TheBANA test would then be a simplemethod of detecting periodontal anaer-obic infections.

In summary, this study has indicatedthat: BANA-positive plaques are associ-ated with the presence of T. denticolaand/or B. gingivalis; T. denticola wasfound at greater frequency and levelsin BANA-positive plaques than was B-gingivalis; and the presence of the or-ganisms is associated with clinical iO'flammation in deep periodontalpockets, A BANA test in a solid statethat can be performed at chairside hasbeen recently evaluated and shows anaccuracy of 84% with B. gitigivalis andT. denticola (11),

Acknowledgemertts

This research was supported by grantNo, DE02731 from the National Insti-tute of Dental Research and by a giftfrom Oral-B Laboratories, RedwoodCity, CA,

References

1. Bretz WA, Loesche WJ. Characteristicof trypsin-like reaction in subgingi^^plaque samples, i Dent Res 1987: 66:1668-1672.

2. Dzink JL, Socransky SS, Haffajee AD-The predominant cultivable microbio'^of active and inactive lesions of destruc-tive periodontal diseases, i Clin Pen"'dontol 1988: 15: 316^323,

3. Larjava H, Uitto VJ, Eerola E, Haapasa-lo M. Inhibition of gingival fibrobli'^growth by Bacteroides gingivalis. lni'̂ '-Immun 1987: 55: 201-205.

4. Laughon BE, Syed SA, Loesche WJ'Rapid identification of Bacteroides P"givalis. J Clin Microbiol 1982: •'•345 346.

5. Laughon BE, Syed SA, Loesche wJj,API-ZYM system for identification "̂Baeteroides sp., Capnoeytophaga sp. ' 'spirochetes of oral origin. J Clin Micfbiol 1982: 15: 97-102.

6. Listgarten MA. Periodontal probing-what does it mean? J Clin Pcriodont"1980: 7: 165-176.

BANA hydrolysis by T, denticola and/or B, gingivalis 279

7. Loesehe WJ, Syed SA, Schmidt E, Mor-rison EC. Bacterial profiles of subgingiv-al plaques in periodontitis. J Periodontol1985: 56: 447^56.

8. Loesche WJ. The role of spirochetes inperiodontal disease. Adv Dent Res 1988:2: 275-283.

9. Loesche WJ, Syed SA, Stoll J. Trypsin-like activity in subgingival plaque: a di-agnostic marker for spirochetes andperiodontal disease? J Periodontol 1987:58: 266-273.

10. Loesche WJ, Hujoel P. Microbiological-based diagnostic tests for periodontitis.Considerations in regard to sensitivity,speeificity and accuracy. The RoyalSociety of Medicine Symposium:Markers of Disease Susceptibility andActivity for Periodontal Diseases. Cam-bridge: Cambridge University Press, inpress.

U. Loesche WJ, Bretz WA, Lopatin DE etal. Multi-center clinical evaluation of achairside method for detecting certainperiodontopathic bacteria in periodontaldisease. J Periodontol 1990: 61: 189-196.

12. Mangan DF, Laughon BE, Bower B, Lo-patin DE, In vitro lymphocyte blasto-

genic responses and titers of humoralantibodies from periodontitis patients tooral spirochete isolates. Infect Immun1982:37:445-451.

13. Moore WEC, Holdeman LV, Cato EP etal. Comparative bacteriology of juvenileperiodontitis. Infect Immun 1985: 48:507-519.

14. Ohta K, Makinen KK, Loesche WJ.Purification and characterization of anenzyme from T. denticola capable of hy-drolyzing synthetic trypsin substrates.Infect Immun 1986: 53: 213-220.

15. Schmidt EF, Bretz WA, Hutchinson RA,Loesche WJ. Correlation of the hydroly-sis of benzoyl-arginine naphthylamide(BANA) by plaque with clinical par-ameters and subgingival levels of spiro-chetes in periodontal patients. J DentRes 1988: 67: 1505-1509.

16. Simonson LG, Goodman CH, Bial JJ,Morton HE. Quantitative relationship ofTreponema dentieola to severity of perio-dontal disease. Infect Immun 1988: 56:447^56.

17. Salvador SL, Syed SA, Loesche WJ.Comparison of three dispersion pro-cedures for quantitative recovery of cul-

tivable species of subgingival spirochetes,J Clin Microbiol 1987: 25: 2230-2232,

18. Slots J, Listgarten MA. Baeteroides gin-givalis, Baeteroides intermedius and Acti-nobacittus actinomyeetemeomitans in hu-man periodontal diseases, J Chn Perio-dontol 1988: 15: 85-93.

19. Tanner ACR, Socransky SS, GoodsonJM. Microbiota of periodontal pocketslosing crestal alveolar bone. J PeriodontRes 1984: 19: 279.

20. Tanner ACR, Ebersole JL, ListgartenMA, Strzempko MN. Baeteroides forsy-thus sp. nov., a slow-growing fusiformBaeteroides sp. from the human oral cav-ity. Int J Syst Bacteriol 1986: 36:213-221.

21. Uitto VJ, Larjava H, Heino J, Sorsa T,A protease of Baeteroides gingivalis de-grades eell surfaee and matrix glyeo-proteins of cultural gingival fibroblastsand induces secretion of collagenase andplasminogens aetivators. Infect Immun1989: 57: 213-218.

22. Zambon JJ, Reynolds HS, Slots J. Black-pigmented Baeteroides spp. in the humanoral cavity. Infect Immun 1981: 32:198-203.