SUMMARYCapture ELISAs with biotinylated monospecific antibodies were developed to detect both C-reactive protein (CRP) and haptoglobin (Hp) in serum of adult periodontitis (AP) patients and normal subjects. Each acute-phase reactant was significantly increased in serum from AP patients with CRP at 9 : 12 Ϯ 1 : 61 mg/l versus 2 : 17 Ϯ 0 : 41 mg/l (P < 0 : 001) and Hp at 3 : 68 Ϯ 0 : 37 g/l versus 1 : 12 Ϯ 0 : 78 g/l (P < 0 : 001). Assessment of clinical characteristics of the patients' periodontal disease indicated that CRP and Hp levels were significantly increased in patients with the most frequent disease active episodes (P < 0 : 02 and P < 0 : 001, respectively). Longitudinal examination of the Hp levels showed a significant decrease following scaling and root planing (3 . 68 versus 2 . 38 g/l; P < 0 : 01). After a 2-year administration of 50 mg/b.i.d. Flurbiprofen (a non-steroidal anti-inflammatory drug), significantly decreased Hp levels were noted (P < 0 : 005). CRP levels declined by 35-40% after 1-2 years of treatment with the drug (P < 0 : 05). The findings indicated that localized infections resulting in increased inflammation and tissue loss in the periodontium elicit systemic host changes manifest by increases in two acute-phase reactants. The conclusions are that either these molecules are formed locally and distributed to the serum, or these presumably localized infections impact upon the systemic components of the host protective responses.

Selected cell envelope components of Porphyromonas gingivalis were tested in a BALB/c mouse model in an attempt to elucidate further the outer membrane components of this putative oral pathogen that might be considered as virulence factors in host tissue destruction. Lipopolysaccharide (LPS), outer membrane, and outer membrane vesicles of P. gingivalis W50, ATCC 53977, and ATCC 33277 were selected to examine an immunological approach for interference with progressing tissue destruction. Mice were actively immunized with heat-killed (H-K) or Formalin-killed (F-K) whole cells or with the outer membrane fraction, LPS, or outer membrane vesicles of the invasive strain P. gingivalis W50. The induction of invasive spreading lesions with tissue destruction and lethality were compared among different immunization groups in normal, dexamethasone-treated (dexamethasone alters neutrophil function at the inflammatory site), and galactosamine-sensitized (galactosamine sensitization increases endotoxin sensitivity) mice after challenge infection with the homologous strain (W50) and heterologous strains (ATCC 53977 and ATCC 33277). Enzyme-linked immunosorbent assay analyses revealed significantly elevated immunoglobulin G and M antibody responses after immunization with H-K or F-K cells or the outer membrane fraction compared with those of nonimmunized mice. The killed whole-cell vaccines provided significantly greater protection against challenge infection in normal mice (decreased lesion size and death) than did either the outer membrane fraction or LPS immunization. The lesion development observed in dexamethasone-pretreated mice was significantly enhanced compared with that of normal mice after challenge with P. gingivalis. Immunization with P. gingivalis W50 provided less protection against heterologous challenge infection with P. gingivalis ATCC 53977; however, some species-specific antigens were recognized and induced protective immunity. Only viable P. gingivalis induced a spreading lesion in normal, dexamethasone-treated, or galactosamine-sensitized mice; F-K or H-K bacteria did not induce lesions. The F-K and outer membrane vesicle immunization offered greater protection from lesion induction than did the H-K immunogen after challenge infection simultaneous with galactosamine sensitization. The H-K cell challenge with galactosamine sensitization produced 100% mortality without lesion induction, suggesting that LPS or LPS-associated outer membrane molecules were functioning like endotoxin. Likewise, P. gingivalis W50 LPS (1 ,ug per animal) administered intravenously produced 80%o mortality in galactosamine-sensitized mice. In contrast to the effects of immunization on lesion development, immunization with H-K or F-K cells or LPS provided no protection against intravenous challenge with LPS; 100% of the mice died from acute endotoxin toxicity. These findings suggest that the murine model will be useful in examining the tissuedestructive components of P. gingivalis.

OBJECTIVE(S): This report compares the virulence of selected strains of P. gingivalis, A. actinomycetemcomitans, C. rectus, F. nucleatum and T. denticola in a murine model as a measure of pathogenic potential of these oral microorganisms. The characteristics of the tissue destruction associated with these monoinfections were then related to a potential model for bacterial synergism in progressing periodontitis. DESIGN AND METHODS: All bacterial strains were grown to mid‐logarithmic to early stationary growth phase, harvested and used at various doses to challenge BALBlc normal and BALB/c dexamethasone (DEX) treated mice to mimic a neutrophil dysfunction. The characteristics of tissue destruction, and overt tissue destructive capacity of these species were examined as a function of challenge dose and time. OUTCOME MEASURES The mice were examined for an interval of approximately 15 days post‐challenge and the presencelabsence of lesions, localized or generalized nature of the lesion (including size in mm), and lethality of the infection were assessed. RESULTS: Comparison of the virulence of the various P. gingivalis strains related to lethality and lesion size associated with destruction of the connective tissue, indicated a virulence capacity of P. gingivalis strains 53977>W50 = T22>3079>33277>381. C. rectus elicited localized necrotic lesions which were limited to the epithelial layers of the skin. The size of the lesions also indicated a graded difference in virulence, such that C. rectus strains 234>576>>33238. A. actinomycetemcornitans caused the formation of classic localized abscesses with a PMN infiltrate and inflammatory exudates. Although, each of the A. actinornyceterncomitans strains exhibited a similar virulence pattern in this murine model, A. actinomycetemcomitans serotype b representative strains were potentially more pathogenic with a virulence capacity of 3113D‐N = 3975A>jP2>Y4>29523>33384. Both C. rectus and A. actinomycetemcomitans strains showed clear evidence that recent clinical isolates were more virulent than laboratory strains. Challenge with F. nucleatum resulted in tissue destructive responses which were different from those observed with the other strains used in this study. A rapid onset of dose‐dependent lesion development, related to the formation of either closed abscesses or open lesions, was observed with F. nucleaturn. Tissue involvement was also greater at lower F. nucleaturn doses when compared to the other bacteria. F. nucleaturn challenge of DEX‐treated mice resulted in a shift to open lesions. T. denticola appeared to be more tissue invasive than the other species examined in this study. Challenge of mice with T. denticola resulted in involvement of multiple tissues, including epithelial and connective tissues, as well as appearing to invade muscle layers and deeper tissues. In addition to invading deeper tissues, the resulting lesions took considerably longer to resolve. In the DEX‐treated mice (neutrophil depleted), P. gingivalis, C. rectus, and A. actinomycetemcomitans we...