Squirrel Monkeys. serum and was associated with a significant reduction in

Transcription

1 INFCTION AND IMMUNITY, June 1991, p /91/ $2./ Copyright 1991, Amerian Soiety for Mirobiology Vol. 59, No. 6 Immune Modulation of Prevotella intermedia Colonization in Squirrel Monkeys W. B. CLARK, 12* I. MAGNUSSON,"2 J.. BM,"2 J. M. JUNG 1,2 R. G. MARKS 1,3 AND W. P. MARTHUR"'2 Periodontal Disease Researh Center' and Departments of Oral Biology2 and Statistis,3 University of Florida, Gainesville, Florida 3261 Reeived 15 Otober 199/Aepted 15 Marh 1991 Colonization of the gingival revie by blak-pigmented Porphyromonas or Prevotella spp. (BP/P), inluding Porphyromonas gingivalis (formerly Bateroides gingivalis) and Prevotella intermnedia (formerly Bateroides intermedius), is thought to be an important eologial event whih may result in the destrution of onnetive tissues supporting the teeth. Theoretially, periodontal diseases ould be prevented if these or other periodontal pathogeni miroorganisms did not olonize the subgingival area. The humoral immune response is one mehanism whih may modulate baterial olonization in the gingival revie. In the present study, we tested the effet of systemi humoral immunity on subgingival olonization by indigenous P. intermedia in squirrel monkeys (Saimiri siureus). Animals rendered essentially free of detetable BP/P by a single saling, 1 days of tetrayline therapy, and toothbrushing three times per week were immunized with P. intermedia 1447 or were sham immunized with phosphate-buffered saline. Subsequently, all oral hygiene proedures were disontinued and five teeth in one quadrant were ligated with baterium-soaked suture material to failitate BP/P olonization. Immunization resulted in a signifiant inrease in the level of immunoglobulin G anti-p. intermedia antibody in serum. Two weeks after ligation was initiated, P. intermedia ould be deteted in five of six sham-immunized and three of six immunized animals. Immunization was assoiated with a redution in the emergene of indigenous P. intermedia in the gingival revie. Blak-pigmented gram-negative anaerobi rods suh as Porphyromonas gingivalis (formerly Bateroides gingivalis) and Prevotella intermedia (formerly Bateroides intermedius), referred to as BP/P, are among the organisms that have been deteted in subgingival plaque (3) and have been impliated as prinipal etiologi agents of periodontitis (26). Substantial mirobiologi and immunologi data impliate P. gingivalis and P. intermedia in destrutive forms of human adult periodontitis (16, 21, 26). limination or prevention of olonization by speifi periodontal pathogens suh as P. gingivalis or P. intermedia is thought to ontrol the progression or onset of periodontal disease (22). Moreover, other baterial diseases an be immunologially modulated by vaination with whole bateria or purified baterial antigens (7). The presene of antibodies speifi for periodontal miroorganisms (25), as well as omplement and polymorphonulear leukoytes in the gingival revie (2), suggests that omponents of the immune system required to immunologially modulate baterial olonization an potentially ome in ontat with periodontal pathogens (18). Preliminary experiments suggest that it is possible to modulate olonization of periodontal pathogens in the gingival revie. For example, Okuda and o-workers (17) have shown that the mean reovery of P. gingivalis (an organism not indigenous to rodents) in hamsters with ligated teeth an be signifiantly redued by subutaneous immunization with whole ells or by passive loal immunization with rabbit anti-p. gingivalis or anti-baterial hemagglutinin serum. Reently, using a nonhuman primate (Saimiri siureus, the squirrel monkey) model, we demonstrated that immunization with P. gingivalis resulted in an inreased level of immunoglobulin G (IgG) anti-p. gingivalis antibodies in * Corresponding author serum and was assoiated with a signifiant redution in BP/P olonization in the gingival revie (13). We further observed that squirrel monkeys immunized with P. gingivalis harbored signifiantly fewer P. gingivalis organisms in the gingival revie when ompared with sham-immunized animals, but no differene in the levels of P. intermedia in the revie was deteted (13). The purpose of the present study was to assess the potential for vaination with P. intermedia whole ells to modulate the reolonization or emergene of indigenous P. intermedia in immunized squirrel monkeys. MATRIALS AND MTHODS Animals. Twelve adult female squirrel monkeys (S. siureus), approximately 6 to 7 years of age, were used in these experiments. Animals were housed and fed as previously desribed (5) in the University of Florida's Health Center Animal Resoures Department, in aordane with institutional guidelines. This faility is aredited by the Amerian Assoiation for Areditation of Laboratory Animal Care. These animals had naturally ourring plaque and mild gingivitis and low levels of serum anti-p. intermedia IgG, IgM, and IgA (<2 U of eah per ml). Before the immunization protool was initiated, BP/P were deteted in 1 of 12 animals or 3 of 48 quadrants. In an effort to redue or eliminate detetable BP/P, animals were treated with dental prophylaxis and tetrayline therapy (2 mg/kg three times a day for 1 days). One week after the prophylaxis and antibioti therapy had been ompleted, the animals were resampled to assess the effetiveness of therapy. A very low level (<1% of total ultivable mirobiota) of BP/P was deteted in only one quadrant of one animal (1 of 48 total quadrants). To maintain low or undetetable levels of olonization during the immunization regimen, the animals' teeth Downloaded from on January 1, 219 by guest

2 1928 CLARK T AL. were brushed with.2% hlorohexidine gluonate three times per week. Immunization. Monkeys (n = 6) were immunized as previously desribed for P. gingivalis 381 (13) but with 19 CFU of formalin-fixed (.6% formalin) P. intermedia This strain had been derived earlier from the subgingival plaque of a squirrel monkey. The bateria were emulsified in inomplete Freund adjuvant at a ratio of.5 ml of adjuvant and.5 ml of baterial suspension. The teeth were ligated on day ; therefore, the immunization protool was begun 18 days before the teeth were ligated. We designated the day of the primary immunization as -18 days. One milliliter of P. intermedia 1447 (19)-inomplete Freund adjuvant was injeted subutaneously at four sites on the hest and abdomen of eah animal. Sham-immunized animals reeived injetions of 1 ml of emulsified inomplete Freund adjuvant without bateria. Immunized animals reeived three booster injetions of 19 formalin-fixed P. intermedia at -145, -11, and -14 days before ligatures were plaed in quadrant 4. Thus, immunization was ompleted before the ligatures were plaed. These injetions onsisted of four subutaneous injetions on the hest and abdomen of 19 formalin-fixed P. intermedia ells suspended in phosphate-buffered saline solution (ph 7.2). Speifi antibody detetion. Systemi IgG antibody responses to P. intermedia 1447 were monitored by a modified enzyme-linked immunosorbent assay (LISA) (5, 8, 13, 19) in whih the relative levels of speifi antibody are expressed as units from a standard urve. Sera were olleted at speifi times and assayed for IgG anti-p. intermedia antibody by using P. intermedia 1447 soni extrat as the antigen on the plates. The data were reported as relative units of IgG anti-p. intermedia based on a standard urve of known amounts of monkey IgG bound to the test wells (4, 12). Speifi antibody levels in saliva were not measured in this study. Mirobiology. Mirobiologi sampling of the gingival reviular area and ulturing methods have been desribed previously (5, 13). Total ultivable mirobiota and BP/P were enumerated on enrihed trypti soy agar (24). Immunoenzymati differentiation was onfirmed by biohemial tests and gas-liquid hromatography as previously desribed (5). Identifiation and differentiation of P. gingivalis and P. intermedia olonies were performed with immunoenzymati detetion by using speies-speifi monolonal antibodies as desribed for Atinobaillus atinomyetemomitans (14). Ligation and inoulation. After the seond booster injetion (-11 days), toothbrushing proedures were stopped. During the 96-day interval between the seond and third booster injetions, the animals were sampled five times and remained essentially free of detetable BP/P; only 3 of 24 possible quadrants were minimally olonized (<1% BP/P). To failitate reolonization or emergene of BP/P, the subgingival area around five teeth in the mandibular right quadrant (quadrant 4) of eah animal was ligated with sterile 3- silk sutures 14 days after administration of the third booster injetion. The day the ligatures were plaed was designated day. At approximately 14-day intervals postligation, the animals were sampled for the presene of P. intermedia and P. gingivalis. In the nonligated quadrants (1 = maxillary right, 2 = maxillary left, 3 = mandibular left), sterile endodonti paper points were used to sample the gingival revie around eah tooth (three premolars and first and seond molars). Samples from teeth in the same quadrant were pooled. In the ligated quadrant, ligatures around TABL 1. Anti-P. intermedia 1447 IgG antibody levels Bleeding U of IgG/ml of seruma (mean [range]) time (days postligation) Immunized Sham immunized 5,813.2 (3,43-8,867) 3.7 (1-66) 35 4,43.7 (2,167-6,983) 14.8 (19-197) 7 3,181. (1,974-4,75) (14-362) 91 2,479.5 (1,598-3,62) (23-298) a IgG anti-p. intermedia 1447 levels were determined with an LISA, using P. intermedia 1447 soni extrat as the antigen. eah tooth were removed and pooled together as the sample for that quadrant. After samples were taken, the ligatures were replaed in quadrant 4. The sampling and ligation regimen was repeated seven times over the next 1-day period. Statistial methods. Although the P. intermedia and P. gingivalis responses are numerial in nature, they were lustered in three groups: zero, very low ounts, or high ounts. For analysis, we treated the data on a nominal sale as infeted (low and high ounts) or noninfeted (zero ounts). To evaluate differenes in the frequeny of ourrene of P. intermedia or P. gingivalis, we developed a repeated-measures regression model to ompare the two groups, using all eight sampling times between day and day 1. This method measures differenes between monkeys by nesting within eah group (13, 15). The analysis on this nominal response variable is appropriate sine it produes an adequate sample size to ompare sham-immunized with immunized monkeys aross all time periods. All analyses were performed by the Statistial Analysis System (2). RSULTS INFCT. IMMUN. Influene of immunization on levels of anti-p. intermedia 1447 in serum. Completion of the immunization protool during the period -18 to -14 days prior to ligation resulted in levels of IgG anti-p. intermedia antibody in serum at, 35, 7, and 95 days after ligation whih were learly greater than respetive levels in sham-immunized animals (Table 1). No additional injetions were administered after the fourth injetion at -14 days. The mean serum IgG anti-p. intermedia antibody response observed at ligation ( days) was approximately 19-fold higher than that observed in the sham-immunized animals. Over the nearly 1 days during whih teeth in quadrant 4 were ligated, the mean IgG anti-p. intermedia response delined by more than one-half the peak response in the immunized animals, while the mean level of IgG anti-p. intermedia antibody inreased by more than threefold over the baseline levels in the sham-immunized animals (Table 1). However, the mean IgG anti-p. intermedia 1447 antibody levels in the immunized animals were always approximately 22-fold or greater than the mean IgG anti-p. intermedia 1447 antibody levels in the sham-immunized animals. Influene of immunization on P. intermedia levels. Figure 1 reports the perentage of monkeys harboring P. intermedia at eah time period by quadrant; as shown, P. intermedia was more frequently deteted in sham-immunized than in immunized animals. Only at the first sampling after the ligatures were plaed (day 16) were more immunized animals infeted with P. intermedia (in one quadrant) than were sham-immunized animals. At that sampling, one immunized animal was infeted with P. intermedia in quadrant 4 (ligated Downloaded from on January 1, 219 by guest

3 VOL. 59, 1991 IMMUN MODULATION OF P. INTRMDIA COLONIZATION 1929 Quadrant Quadrant 1 a- C) U 'a m Immunized * Sham-immun. Ula - n a- U n 'U o 8 I I.. I.._ I._ Quadrant 3 I oo la a) 4- C.) a) C C I-O Quadrant Immunized Sham-immun. Downloaded from FIG. 1. Mean perentages of sham-immunized squirrel monkeys (n = 6) and squirrel monkeys immunized with P. intermedia 1447 (n = 6) harboring P. intermedia in subgingival plaque from quadrants 1, 2, 3, and 4 (ligated quadrant) at eah sampling time. quadrant) and no sham-immunized animals were infeted in quadrant 4. At 56 and 7 days, the sham-immunized group exhibited a muh greater frequeny of P. intermedia olonization than did the immunized group in quadrants 1, 2, and 4. It is interesting that in the immunized group, P. intermedia tended to disappear by 56 and 7 days but then started to reappear, whereas in the sham-immunized animals, P. intermedia remained in high numbers throughout the study. Data summarizing the number of quadrants olonized by P. intermedia and P. gingivalis for all dates in immunized and sham-immunized squirrel monkeys are shown in Table 2. The P. intermedia mean perentage of total mirobiota in subgingival plaque for all dates is also shown for eah quadrant (in parentheses, Table 2). The results of the repeated-measures analysis for eah quadrant, in whih differenes among monkeys are measured by nesting within eah group (13, 15), are presented in Table 2. Sham-immunized animals were more frequently olonized by P. intermedia than were immunized animals. As shown in Table 2, these differenes were at or near statistial signifiane in all four quadrants. In ontrast, there was essentially no differene between immunized and sham-immunized animals in the frequeny of olonization by P. gingivalis. DISCUSSION These data demonstrate that immunization of squirrel monkeys with P. intermedia leads to an inreased level of IgG anti-p. intermedia antibodies in serum and is assoiated with a redution in the frequeny of reolonization in the gingival revie by indigenous P. intermedia. Before the immunization protool was begun, animals were treated with mehanial and antimirobial therapy to eliminate or redue indigenous BP/P below the level of detetion. After immunization was ompleted, the mandibular right quadrant (quadrant 4) was ligated with sterile silk suture material to TABL 2. Quadrantsa olonized by P. intermedia in immunized and sham-immunized squirrel monkeys No. olonized No. olonized with P. intermedial with P. gingivalisl Quadrant no. of monkeys no. of monkeys Immunized Sham pb Immunized Sham 1 11/48 (.7%) 25/48 (1.6%) <.9 2/48 4/48 2 8/48 (.8%) 22/48 (1.6%) <.7 2/48 1/48 3 4/48 (.1%) 18/48 (.3%) <.3 1/48 2/48 4 9/48 (1.5%) 25/48 (5.2%) <.8 3/48 2/48 Total 32/192 19/192 <.6 8/192 9/192 Data by quadrant are pooled for all weeks. b Signifiane levels are from a repeated-measures regression model omparing groups after adjusting for monkey within-group effet. Number in parentheses is P. intermedia mean perentage of total mirobiota in subgingival plaque pooled for all weeks by quadrant. on January 1, 219 by guest

4 193 CLARK T AL. failitate olonization or emergene of BP/P. P. intermedia ould be deteted in the animals within 16 days. The umulative results show differenes in the number of infeted sham-immunized and infeted immunized animals that were at or near statistial signifiane in all quadrants (Table 2). Note that the number of animals (six immunized and six sham immunized) does not provide suffiient data for a statistial analysis for eah time period. Also, of most interest is a omparison of the two groups over time: the six monkeys in eah group over eight time periods produed 96 observations, enough for an adequate omparison of groups by quadrant. In addition, for eah quadrant, the P. intermedia mean perentage of total mirobiota in subgingival plaque was lower in immunized than in sham-immunized animals. Although differenes between the immunized and sham-immunized groups in the number of animals infeted seem similar in quadrants 1, 2, 3, and 4, the proportions of P. intermedia olonizing quadrant 4 are greater than in the other quadrants in both immunized and sham-immunized animals. P. gingivalis olonization ourred infrequently in both groups, and there was no statistially signifiant differene between the groups. Sine P. gingivalis infetion in these animals was undetetable prior to initiation of this experiment (data not shown), it is not surprising that P. gingivalis infetion remained low. In addition, P. intermedia and P. gingivalis do not ross-reat, so we did not expet to detet a signifiant differene between immunized and sham-immunized squirrel monkeys in emergene or reolonization of P. gingivalis. However, these results are partiularly interesting when onsidered with an earlier study whih showed that olonization by P. gingivalis was signifiantly redued in squirrel monkeys immunized and hallenged with a monkeyderived isolate of P. gingivalis but that olonization of P. intermedia was not redued (6, 13). With the exeption of quadrant 1, by day 28 fewer immunized than sham-immunized animals were infeted with P. intermedia. Interestingly, after day 28, the number of infeted sham-immunized animals remained fairly onstant, whereas more immunized animals beame infeted after day 56. Sine there are insuffiient data for statistial analysis at eah time period, we do not know whether the observation is statistially signifiant. However, assuming these observations are real, it is tempting to speulate that the deline in the amount of or bioativity of serum anti-p. intermedia over time (Table 1) is responsible. Perhaps a booster injetion during this time might have extended the period of suppression of P. intermedia in the immunized animals. Alternatively, the organisms olonizing after day 56 may differ antigenially or genetially from the original organisms (i.e., antigeni variation). However, neither of these possibilities an be onfirmed in the present experiment. The experimental design in the work reported here was quite different from that used in a previous study (13) from the standpoint of the mehanism of infetion. The animals used in the present experiment harbored indigenous P. intermedia. Ligation was used to failitate olonization or emergene of P. intermedia from an indigenous soure. In the earlier experiments, infetion was aomplished by weekly ligations and inoulations of the total dentogingival margin with viable P. gingivalis. ssentially nothing is known about the mehanisms of infetion by the BP/P organisms in these animals. It ould be argued that emergene of or reinfetion by indigenous organisms may involve baterial strains that are better adapted for infetion (i.e., are possibly more virulent) than a seleted laboratory strain INFCT. IMMUN. isolated from a diseased site in a squirrel monkey, as was done in the previous study (13). Consequently, the modulation of an indigenous infetion in onjuntion with ligation may be more signifiant in light of the potential to limit natural infetion in humans. Ligature-indued periodontitis in the squirrel monkey (1, 9, 11), and other nonhuman primates (1, 23) has been well established. Although the mirobial etiology of ligatureindued periodontitis in the squirrel monkey (or any of the other ligature models) has not been preisely established, BP/P have been impliated, and their olonization is strongly favored by the presene of ligatures in the gingival revie (1, 23). We have also observed statistially signifiant inreases in P. gingivalis and P. intermedia levels in ligated quadrants aompanied by signifiant alveolar bone loss relative to nonligated sites with minimal alveolar bone loss (6). Periodontal diseases are thought to be multifatorial diseases involving multiple etiologies (1, 16, 21-25). In theory, prevention of infetion by one or more of the putative periodontal pathogeni organisms should prevent or redue disease. The observation that in squirrel monkeys with ligated teeth whole-ell P. intermedia and P. gingivalis vaines are assoiated with redutions in olonization or emergene of P. intermedia (in the present study) or P. gingivalis (in previous studies [13, 17]), demonstrates that it is possible to modulate infetion by two prominent periodontal pathogens in a periodontal disease model. Obviously, the vaination responses and influene on olonization need to be optimized. Theoretially, however, a multivalent vaine against P. intermedia and P. gingivalis might redue olonization by these organisms below a ritial threshold neessary to initiate or maintain disease. Assuming that P. intermedia and P. gingivalis are diretly involved in the initiation and/or progression of alveolar bone loss in this model, then immune modulation of the levels of these organisms in the subgingival revie of squirrel monkeys below the ritial threshold might also be expeted to influene alveolar bone loss. ACKNOWLDGMNTS The work presented here was supported by Publi Health Servie grant D7117 from the National Institute of Dental Researh. The manusript was written while W.B.C. was a visiting professor in the Department of Mirobiology, University of British Columbia, Vanouver, Canada. We thank D. F. Croft, R. Swift, and J. W. Maruniak for expert tehnial assistane, P. Padgett for assistane with statistial analysis, and J. van Mill for assistane in preparing the manusript. In addition, we thank Barry C. MBride for his assistane during the preparation of this manusript. RFRNCS 1. Adams, R. A., H. A. Zander, and A. M. Poison Interproximal and bual ell populations apial to the sulus before and during experimental periodontitis in squirrel monkeys. J. Periodontol. 52: Attstrom, R., A. B. Laurel, U. Larsson, and A. Sjoholm Complement fators in gingival revie material from healthy and inflamed gingiva in humans. J. Periodontal Res. 1: Bailit, H. L., D. C. Baldwin, and.. Hunt, Jr Inreasing prevalene of gingival Bateroides melaninogenius with age in hildren. Arh. Oral Biol. 9: Butler, J.., T. L. Feldbush, P. L. Mgiven, and N. Stewart The enzyme-linked immunosorbent assay: a measurement of antibody onentration or affinity? Immunohemistry 15: Clark, W. B., I. Magnusson, C. Abee, B. Collins, J.. Beem, and W. P. MArthur Natural ourrene of blak-pigmented Downloaded from on January 1, 219 by guest

5 VOL. 59, 1991 IMMUN MODULATION OF P. INTRMDIA COLONIZATION 1931 Bateroides speies in the gingival revie of the squirrel monkey. Infet. Immun. 56: Clark, W. B., I. Magnusson, J.. Beem, J. M. Jung, R. G. Marks, and W. P. MArthur. Submitted for publiation. 7. Corbeil, L. B Strategies to overome baterial virulene mehanisms: an overview, p In J. A. Roth (ed.), Virulene mehanisms of baterial pathogens. Amerian Soiety for Mirobiology, Washington, D.C. 8. ngvall,., and P. Perlmann LISA III quantitation of speifi antibodies by enzyme linked immunoglobulin in antigen oated tubes. J. Immunol. 19: HeUl, L., B. R. Rifkin, and H. A. Zander Conversion of hroni gingivitis to periodontitis in squirrel monkeys. J. Periodontol. 47: Holt, S. C., J. bersole, J. Felton, M. Brunsvold, and K. S. Kornman Implantation of Bateroides gingivalis in nonhuman primates initiates progression of periodontitis. Siene 239: Kennedy, J.., and A. M. Polson xperimental marginal periodontitis in squirrel monkeys. J. Periodontol. 44: Makela, I., and F. Peterty Standard sera in solid-phase immunoassays. ur. J. Immunol. 13: MArthur, W. P., I. Magnusson, R. G. Marks, and W. B. Clark Modulation of olonization by blak-pigmented Bateroides speies in squirrel monkeys by immunization with Bateroides gingivalis. Infet. Immun. 57: MArthur, W. P., S. Stroup, and L. Wilson Detetion and serotyping of Atinobaillus atinomyetemomitans isolates on nitroellulose paper blots with monolonal antibodies. J. Clin. Periodontol. 13: Mendenhall, W Introdution to linear models and the design and analysis of experiments, p Wadsworth Publishing, Belmont, Calif. 16. Mouton, C., P. G. Hammond, J. Slots, and R. J. Geno Serum antibodies to oral Bateroides asaharolytius (Bateroides gingivalis): relationship to age and periodontal disease. Infet. Immun. 31: Okuda, K., T. Kato, Y. Naito, I. Takazoe, Y. Kikuhi, T. Nakamura, T. Kiyoshige, and S. Sasaki Protetive effiay of ative and passive immunizations against experimental infetion with Bateroides gingivalis in ligated hamsters. J. Dent. Res. 67: Page, R. C., and H.. Shroeder Periodontitis in man and animals. S. Karger, Basel. 19. Rodbard, D., and P. J. Munson Radioimmunoassay data proessing p In N. R. Rose and H. Friedman (ed.), Manual of linial immunology. Amerian Soiety for Mirobiology, Washington, D.C. 2. SAS Institute SAS User's Guide. SAS Institute, In., Cary, N.C. 21. Slots, J Subgingival miroflora and periodontal disease. J. Clin. Periodontol. 6: Slots, J., L. J. mrih, R. J. Geno, and B. G. Rosling Relationship between some subgingival bateria and periodontal poket depth and gain or loss of periodontal attahment after treatment of adult periodontitis. J. Clin. Periodontol. 12: Slots, J.,. Hausmann, C. Mouton, L. F. Ortman, P. G. Hammond, and R. J. Geno The relationship between periodontal miroflora and alveolar bone loss in Maaa artoides, p In D. W. Lambe, Jr., R. J. Geno, and K. J. Mayberry-Carson (ed.), Anaerobi bateria: seleted topis. Plenum Publishing Corp., New York. 24. Slots, J., and M. A. Listgarten Bateroides gingivalis, Bateroides intermedius and Atinobaillus atinomyetemomitans in human periodontal diseases. J. Clin. Periodontol. 15: Taubman, M. A., J. L. bersole, and D. Smith Assoiation between systemi and loal antibody and periodontal diseases, p In R. J. Geno and S. A. Mergenhagen (ed.), Host-parasite interations in periodontal diseases. Amerian Soiety for Mirobiology, Washington, D.C. 26. White, D., and D. Mayrand Assoiation of oral Bateroides with gingivitis and adult periodontitis. J. Periodontal Res. 16: Downloaded from on January 1, 219 by guest