Journal of Dental Research

Journal of Dental Research http://jdr.sagepub.com/

Serum Antibody Response to Antigens of Oral Gram-negative Bacteria by Cats with Plasma Cell Gingivitis-Pharyngitis T.J. Sims, B.J. Moncla and R.C. Page J DENT RES 1990 69: 877 DOI: 10.1177/00220345900690031001 The online version of this article can be found at: http://jdr.sagepub.com/content/69/3/877

Published by: http://www.sagepublications.com

On behalf of: International and American Associations for Dental Research

Additional services and information for Journal of Dental Research can be found at: Email Alerts: http://jdr.sagepub.com/cgi/alerts Subscriptions: http://jdr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jdr.sagepub.com/content/69/3/877.refs.html

Downloaded from jdr.sagepub.com by guest on July 11, 2011 For personal use only. No other uses without permission.

Serum Antibody Response to Antigens of Oral Gram-negative Bacteria by Cats with Plasma Cell Gingivitis-Pharyngitis T.J. SIMS4, B.J. MONCLA" 2, and R.C. PAGE" 3'4'5

'Departments of Periodontics and 2Oral Biology, School of Dentistry; 3Department of Pathology, School of Medicine; and the 'Research Center in Oral Biology SM-42, Health Sciences Center, University of Washington, Seattle, Washington 98195 The etiology of a form ofperiodontal disease in domestic cats known as plasma cell gingivitis-pharyngitis is not understood. Actinobacillus

actinomycetemcomitans and Bacteroides species have been strongly implicated as the cause of periodontitis in humans and other mammalian species, and most affected patients manifest serum antibodies reactive with the infecting bacteria. We and others have isolated Bacteroides species from the oral flora of cats. Using enzyme-linked immunosorbent assay (ELISA) and immunoblot procedures, we measured serum antibodies in affected and control cats reactive with human isolates of A. actinomycetemcomitans, B. gingivalis, and B. intermedius, and purified lipopolysaccharide (LPS) from these and other species, and Bacteroides of cat origin. Affected cats had serum antibody titers reactive with these Gram-negative anaerobic bacteria that were significantly elevated relative to those of normal control cats. The quantitatively major antigens recognized by cat serum antibodies are proteins; this contrasts sharply with serum antibodies from humans with juvenile periodontitis, where LPS is the quantitatively major antigen fraction. Our data support the idea that plasma cell gingivitis-pharyngitis in cats may have a bacterial etiology, and that Gram-negative anaerobes similar to those that cause periodontitis in humans and other mammals may be involved. J Dent Res 69(3):877-882, March, 1990

1984) have been implicated (Frost and Williams, 1986), but it is clear that neither virus is found in all affected cats. The bacterial etiology of gingivitis and periodontitis in humans and in all other animals in which these diseases have been studied is firmly established, and specific species of predominantly Gram-negative anaerobes including Actinobacillus actinomycetemcomitans and various species of Bacteroides have been implicated (Page and Schroeder, 1982; Moore, 1987; Haffajee et al., 1988). A similar etiology seems likely in cats, although evidence is lacking. Our studies were aimed at seeking evidence for participation of Gram-negative bacteria similar to those associated with human periodontitis, and to learn whether cats recognize the same classes of antigens as human patients. Recently, B. gingivalis and other Bacteroides species have been isolated from the oral flora of cats (Love et aL, 1981, 1984, 1987). Most, if not all, humans respond to periodontal infection by production of serum antibodies reactive with the antigens of the infecting bacteria (Williams et al., 1985; Gunsolley et al., 1987). We have assessed the serum of normal and affected cats for antibodies reactive with antigens of known periodontal pathogens. We suggest that plasma cell gingivitispharyngitis in domestic cats may have a microbial etiology.

Introduction. Inflammatory periodontal and mucosal diseases are reported to rather high frequency in domestic cats (Von Schlup, 1982; Reichart et al., 1984). These consist of several different diseases that have not yet been sorted out into individual welldefined entities. A form of special interest has been referred to as plasma cell gingivitis-pharyngitis (Frost and Williams, 1986). In this form, the gingivae become hyperplastic and acutely inflamed, and the inflammatory process may extend onto the glossopalatine arches, pharynx, hard palate, and tongue. Ulcerations may or may not be present. As the name implies, the lesions are characterized by a dense infiltrate of plasma cells and lymphocytes, with variable numbers of granulocytes. These features, however, are no different from those seen in the more commonly occurring forms of periodontitis in man and other animals (Page and Schroeder, 1982). Affected cats manifest dysphagia, ptyalism, halitosis, loss of appetite, and severe pain when eating. The etiology and pathogenesis of plasma cell gingivitispharyngitis in cats remain obscure. Based on a study of nine affected animals, Johnessee and Hurvitz (1984) reported a polyclonal elevation in serum gammaglobulin, although they did not present data. Immune suppression resulting from infection with feline leukemia virus (Barrett et al., 1975; Cotter et al., 1975) and infection by calicivirus (Thompson et al., occur with a

Received for publication August 11, 1989 Accepted for publication November 23, 1989 5To whom correspondence and requests for reprints should be addressed This investigation was supported in part by USPHS Research Grants DE08555 and DE08229 from the National Institute of Dental Research, National Institutes of Health, Bethesda, Maryland 20892.

Materials and methods. Bacterial culture and preparation.-Type strains obtained from the American Type Culture Collection and maintained in our laboratories included Actinobacillus actinomycetemcomitans serotype b ATCC 43718, Bacteroides intermedius ATCC 25611, and Bacteroides gingivalis ATCC 33277. Cultures were maintained on pre-reduced blood agar plates incubated at 350C in Oxoid anaerobe jars, in an anaerobic atmosphere generated using Oxoid anaerobe gas packs (Oxoid, Ltd., London, England). Batch cultures of black-pigmented Bacteroides were grown on enriched trypticase soy agar plates (Syed, 1980), as described above, except that cells were collected after three or four days of growth. Cells were scraped from the plate surface with a sterile dacron-tip swab and placed in 10 mL phosphatebuffered saline, and washed 2 x in the same buffer and 1 x in sterile distilled water. Cell pellets were lyophilized and stored at -20TC until needed. The feline Bacteroides species used in this study were isolated from a sample of exudate from an oral palatal infection cultured on pre-reduced blood agar plates, as described above, for seven to ten days. Individual colonies were subcultured until pure, and identified as described by Holdeman et al. (1977, 1984) and Love et al. (1981, 1984, 1987). Bacteroides species isolated included a catalase-positive B. gingivalis and an asaccharolytic black-pigmented species corresponding to Group 2, and another closely resembling Group E, as described by Love et al. (1987). Antigen preparations. -Lyophilized bacteria were suspended in distilled water at 10 mg/mL, heated in a boiling water bath for 30 min for inactivation of heat-labile proteases, disrupted with a Cole-Parmer 4710 ultrasonic homogenizer (5 min, 50% pulse mode, 40% power), and diluted in buffer 877


878

J Dent Res March 1990

SIMS et al.

appropriate for polyacrylamide gel electrophoresis (PAGE), enzyme-linked immunosorbent assay (ELISA), or immunoblot assay, as described below. For preparation of purified protein and lipopolysaccharide (LPS) fractions, the heated cell homogenates were again sonicated in detergent (1% sodium dodecylsulfate [SDS] and 1% 2-mercaptoethanol), heat-treated again, and centrifuged at 100,000 g for one h for removal of the peptidoglycan and insoluble debris. The supernatant was diluted 1:2 with cold 20% trichloroacetic acid (TCA) for precipitation of the proteins, and centrifuged at 12,000 g at 40C for 30 min. The TCA precipitate was suspended in water and dissolved by titration to neutral pH with 0.01 mol/L NaOH, dialyzed against PBS followed by water, and lyophilized. The TCA-soluble fraction was dialyzed in the same way and sedimented at 100,000 g for two h so that aggregated LPS could be harvested. In addition, LPS was prepared from A. actinomycetemcomitans, B. gingivalis, and B. intennedius by the method of Westphal and Jann (1965), and LPS from E. coli, Salmonella minnesota, and Pseudomonas aeruginosa prepared by the same method were purchased (Sigma, St. Louis, MO). The total protein content of each bacterial homogenate and fraction was measured (Lowry et al., 1951). Total sugar and heptulose were measured by the phenol-sulfuric acid method (DuBois et al., 1956) and by cysteine-sulfuric acid procedures (Wright and Rebers, 1972), respectively. The ratio of nucleic acid and protein in the fractions was determined by UV absorption at 260 and 280 nm. LPS fatty acid composition and contamination of other fractions by LPS were assessed by gasliquid chromatography by a Hewlett-Packard 5830A system equipped with a Supelco SP-2100 column, as follows. Fatty acid methyl ester derivatives were obtained for the fraction or whole cells being heated at 10 mg/mL in 3% methanolic HCl (0.1 mg/mL pentadecanoic acid internal standard) in sealed vials at 100'C for 18 h. Derivative peaks were identified by comparison of their retention times with those of a mixture of bacterial methyl ester standards (Supelco No. 4-5436). The amount of each fatty acid found in a given fraction was calculated based on the area of a given peak relative to that of the internal standard. In addition, fractions of cell sonicates were analyzed by PAGE (Laemmli, 1970). Protein bands were stained with Coomassie blue, and LPS ladder patterns were made visible with ammoniated silver nitrate (Hitchcock and Brown, 1983). Nitrocellulose immunoblots. -Protein or LPS fractions separated by PAGE were electrophoretically immobilized on sheets of nitrocellulose (NC), as described by Towbin et al. (1979). Fraction samples were suspended in buffer (100 mmol/L TRIS, 10% sucrose, 1% SDS, 1% 2-mercaptoethanol, pH 6.8) at 1 mg/mL, heated in a boiling water bath, and separated in 8% polyacrylamide gels. The gels Were placed in contact with NC sheets, immersed in blot buffer (25 mmol/L TRIS, 100 mmol/ L glycine, 20% methanol, pH 8.8) between two wire grid electrodes, and subjected to a seven volt/cm electrical field for two h. The NC sheets were washed with blocking buffer (10 mmol/L TES, 1% BSA, 0.1% TWEEN-20, pH 7.5) and incubated for two h with cat or human serum diluted 1:300 on a rocking platform, washed, incubated two h in goat anti-cat IgG (-y- and X-specific) alkaline phosphatase conjugate (PelFreeze, Rodgers, AK), or goat anti-human IgG (y) alkaline phosphatase (Sigma, St. Louis, MO) diluted 1:300, washed again, and incubated in 5-bromo-4-chloro-3-indolyl-phosphate/nitroblue tetrazolium phosphatase substrate solution (Kirkegaard and Perry, Gaithersburg, MD) until antibody-bearing bands were visible. For identification of periodate-sensitive antigens on electrophoretic immunoblots, two NC sheets bearing the same band pattern were processed as described above, except that just

prior to the serum incubation step, one was incubated for three h in 50 mmol/L ammonium acetate buffer (pH 4.2) with 100 mmol/L periodic acid, and the other in buffer alone. Both blots

were washed extensively with blocking buffer crose (1%), and were processed, as described

containing suabove, so that residual antibody-binding bands could be made visible. For the dot-blot assay, bacterial sonicates and LPS or purified protein fractions were suspended at a concentration of 100 p.g/mL in 100 mmol/L sodium carbonate buffer containing 20 mmol/L MgCl2 and diluted serially 1:1 in microtest plate wells. A 3.5-j±L sample from each well was then spotted onto a sheet of NC with use of a 12-channel micropipetter. Sheets bearing rows of spots representing dilutions of different antigen preparations were then incubated with blocking buffer, washed, incubated for two h with diluted feline or human serum, and processed to reveal bound IgG, as described for electrophoretic immunoblots above. However, to stain bound cat IgM, we had to use a three-stage procedure in which incubation with the cat serum was followed by sheep anti-cat IgM (Nordic Immunological), which in turn was followed by donkey anti-sheep alkaline phosphatase conjugate (Sigma, St. Louis, MO). Enzyme-linked immunosorbent assay (ELISA). -LPS preparations were coated onto the surfaces of polystyrene microtest plate wells (Flow Labs) by the method of Schenck (1985), and bacterial proteins and sonicates by the method of Ebersole et al. (1980), so that optimal binding of antigen could be ensured in each case. Coating solutions were washed from the wells, and unbound sites on the plastic surface were blocked with 0.1% TWEEN-20 in 10 mmol/L TES-buffer saline by use of a Titertek automatic plate washer. Serum (1:100) was diluted serially across rows of 12 wells, and the plate was incubated on a rotating platform (140 rpm) for two h. The plate was washed lOx, goat anti-cat or anti-human IgG (1:500) alkaline phosphatase (as above) was added to the wells, and rotation continued for two h. Plates were again washed 10 x, then rinsed with 50 mmol/L carbonate-20 mmol/L MgCl2 buffer (pH 9.6), and 200 pug/mL p-nitrophenyl phosphate in the same buffer were added to each well. UV absorbance at 405 nm was then measured for each well, after an incubation time of 30 min, by use of a Titertek Multiskan MC photometer. Dilutions of a standard cat serum pool were included on each plate to monitor interplate assay variation. Antibody titers for individual sera from comparable plates were calculated as the reciprocal of the highest dilution corresponding to an absorbance of 0.25 (roughly 20% of the standard curve maximum absorbance), as determined by linear interpolation between dilution points (Caulfield and Shaffer, 1984). Titer values were log-transformed so that equality of variance in affected and control cat data and normality of distribution would be ensured, and differences between the mean titers of the two groups were assessed by a one-way analysis of variance. Serum samples. -Serum samples were obtained from 11 systemically and periodontally normal cats residing in the Vivarium at the University of Washington. These cats were obtained from Fab Laboratories (New York) and were random source. Samples from affected cats were obtained from smallanimal veterinary practices from cats diagnosed by veterinarians as having plasma cell gingivitis-pharyngitis. Samples were stored frozen until tested.

Results. Fig. 1 shows the results of assessment by ELISA of IgG antibody titers of serum samples for each normal and affected cat, with the following being used as plate antigens: homogenates of human oral isolates of A. actinomycetemcomitans,


ANTIBODIES IN CATS WITH PERIODONTAL DISEASE

Vol. 69 No. 3

C

B

A

A

E

D

879

B

D

C

E

0

8-

0

9

0

.

S

5-

I

0

S

4-

S 0

0

,x

A

8 0

11V

to

0

x~

3

so

0

I

0

00 0 00

.

o0

0 0

A

N

N

A

N

A

N

B. intermedius, and B. gingivalis, and cat isolates of B. gingivalis and Bacteroides Group E. Although titers for individual cats in the normal and affected groups overlap, the mean titers for affected cats were significantly elevated relative to control cats for all antigen preparations tested, with probability values ranging from 0.001 to 0.026. Fig. 2 shows ELISA titers performed with LPS obtained from human isolates of B. intermedius and B. gingivalis, as well as E. coli, S. Minnesota, and P. aeruginosa as plate antigens. Mean titers (log-transformed) for affected cats were significantly elevated relative to those of normal cats against LPS from B. intennedius and B gingivalis, as well as E. coli (p = 0.023, 0.002, 0.008, respectively). Anti-LPS antibody titers against P. aeruginosa and S. Minnesota were elevated in affected relative to normal cat serum, but the differences were not statistically significant (p = 0.102, 0.085, respectively). Although the ELISA data suggested that LPS accounted for at least part of the antibody-binding activity of unfractionated cells, they did not indicate whether LPS was a major or minor contributor to the total activity relative to other antigen classes such as proteins. For this question to be addressed, protein fractions were isolated from A. actinomycetemcomitans and B. gingivalis, and their binding activities were compared with that of LPS by a quantitative dot-blot assay. Different dilutions of fractions or whole-cell sonicates were made, and dots of each were placed on a sheet of nitrocellulose, which was subsequently incubated in diluted pooled serum from affected cats or human subjects with periodontitis. The lowest concentration of a given fraction positive for antibody binding was then determined visually. For the homogenate of A. actinomycetemcomitans (feline, lane C), the dots were essentially negative after the fourth antigen dilution; in contrast, the dots for A. actinomycetemcomitans purified protein fraction (lane A) on the same sheet were still positive at the eighth antigen dilution, but positive for LPS (lane B) only to the second antigen dilution. Thus, the proteins of A. actinomycetemcomitans appeared to be the quantitatively important antigens, with little or no reactivity

0 0

0o 0 0

8 0 000

IE

A

0

0

2

01a)

A Fig. 1-Comparison of normal (o) and clinically affected (e) cats with regard to serum IgG titers (ELISA) against antigens of ultrasonically disrupted bacteria (A, A. actinomycetemcomitans ATCC 43718; B, B. intermedius ATCC 25611; C, B. gingivalis ATCC 33277; D, cat isolate of Bacteroides Group E; E, cat isolate of B. gingivalis). Mean titer values for normal (A) and affected (A) populations are shown with standard errors. Data were analyzed by a one-way analysis of variance. N

o

01

I

8{

1

C? 0

x3

0

0 0

0

4 0

0 0

2-

0

0

AS0

0

3-

.

5

0

S 0 0 S 0

11

°T

8

080

0

1

8

0 0

0 0 0

i

0 I

0

0

0

8

0

0(00 I

0 0 0o 0

.

000

N A N A N A N A Fig. 2-Comparison of normal (0) and clinically affected (e) cats with regard to serum IgG titers (ELISA) against phenol-extracted LPS from various bacteria (A, B. intennedius ATCC 25611; B, B. gingivalis ATCC 33277; C, E. coli 0127:B8; D, S. Minnesota; E, P. aeruginosa). Mean (± standard errors) are shown for normal (A) and affected (A) populations. N

A

above background levels with purified LPS. Binding to B. gingivalis LPS (lane D) was significantly less than binding to B. gingivalis whole homogenate (lane E). Binding to B. intermedius whole homogenate (lane G) was slightly less than that to B. gingivalis, but binding to B. intennedius LPS (lane F) was barely detectable. Binding to E. coli LPS was also low (lane H) and comparable with that for A. actinomycetemcomitans LPS (lane B). Dot blots were also performed as above, but with use of a detection system for cat IgM rather than IgG (data not shown). The results did not differ substantially from those observed for IgG, except that binding was less intense. Binding of antibodies present in human juvenile periodontitis serum to the same antigens differed significantly from that observed for cats. Binding to whole homogenate of A. actinomycetemcomitans (human, lane C) and to purified LPS (lane B) was much greater for human than for cat serum, while binding to purified protein (lane A) was about the same. More binding of human than cat antibodies was also observed for LPS fractions from B. gingivalis (lane B), B. intermedius (lane F), and E. coli (lane H). Thus, human serum appeared to contain either many more IgG antibodies specific for LPS than cat serum, or antibody of much greater avidity. The observations with use of dot blots were confirmed and extended by electrophoretic immunoblots showing the relative IgG binding activity of serum from human patients and affected cats to PAGE-separated antigen preparations (Fig. 4). In these experiments, IgG binding to antigen preparations previously subjected to periodate oxidation, a procedure that destroys carbohydrate antigens, was compared with untreated preparations. Periodate oxidation appeared to have little effect on affected cat serum IgG binding to components of B. intermedius, B. gingivalis, or A. actinomycetemcomitans (lanes A,a; B,b; C,c), respectively, or to the purified protein fraction from A. actinomycetemcomitans (lane Ee). Furthermore, antibody failed to bind in detectable amounts to LPS from A. actinomycetemcomitans, oxidized or not (lanes Dd). In marked contrast, intense binding of antibodies in human serum was observed to whole homogenates (lane g) and purified LPS (lane f) from



SIMS et al.

880 2

34

5

6

78

1

2

3

FeI ine

4

56

8

7

Human

Fig. 3-Dot immunoblots showing relative IgG binding activity of various bacterial sonicates and fractions prepared from ATCC strains and spotted onto nitrocellulose and incubated with pooled sera from affected cats (n = 8) or from humans with juvenile periodontitis (n = 8). Each strip (rows A-H) represents dots made from eight serial 1:1 dilutions of a given antigen preparation, starting at 100 pg/mL (column 1) and ending at 0.8 pg/mL (column 8). Two identical strips made from TCA-purified protein fraction of A. actinomycetemcomitans (A), A. actinomycetemcomitans LPS (B), A. actinomycetemcomitans sonicate (C), B. gingivalis LPS (D), B. gingivalis sonicate (E), B. intennedius LPS (F), B. intemnedius sonicate (G), or E. coli LPS (0127:B8) (H) are shown side-by-side so that direct comparison of feline and human serum activities may be made.

A

b

cDdEe

:f Gg

Fig. 4-Western immunoblots showing the relative feline and human IgG binding activity of PAGE-separated whole homogenate constituents of B. intennedius (A,a), B. gingivalis (B,b), A. actinomycetemcomitans (C,c,G,g), LPS from A. actinomycetemcomitans (D,d,F,f), and TCA-purified protein fraction fromA. actinomycetemcomitans (E,e), with oxidation of the blot with 100 mmol/L periodic acid (upper case) or without oxidation (lower case). Blot pairs A,a through E,e were incubated with pooled sera from eight cats with plasma cell gingivitis-pharyngitis, and pairs F,f and G,g were incubated with pooled serum from patients with juvenile periodontitis (n = 8), each at a dilution of 1:300.

A. actinomycetemcomitans, and this binding was abrogated by

periodate oxidation (lanes G,F, respectively). Discussion. Inflammatory periodontal diseases occur with a high frequency in domestic cats. In one study of 200 cats, 57.7% manifested mild to severe disease (Von Schlup, 1982), and in another, radiographic evidence of alveolar bone loss was observed around 77.3% of all premolars and molars in 15 cats averaging 6.8 years of age (Reichart et al., 1984). Both vertical and horizontal bone lesions occur, and these

can

be

seen

in all breeds at all ages. Periodontal diseases seem, however, to be more prevalent in older pure-bred animals. Plasma cell gingivitis-pharyngitis in domestic cats is important, since it appears to have a rather high prevalence, the etiology and pathogenesis are not understood, and no successful form of therapy exists (Frost and Williams, 1986). Furthermore, findings may help elucidate the nature of human periodontal diseases. Although a viral etiology has been suggested (Barrett et al., 1975; Cotter et al., 1975; Thompson et al., 1984), a definitive association between the disease and any known viral infection has not been forthcoming. It is notable that except for HIV infection in humans, viral infection has not been implicated in gingivitis or periodontitis in humans or in any other mammals. On the other hand, a bacterial etiology in humans and some animal species has been well-documented, and A. actinomycetemcomitans and various species of Bacteroides, especially B. gingivalis, are among the implicated pathogens (Page and Schroeder, 1982; Moore, 1987; Haffajee et al., 1988). The same may also be true for cats. Love et al. (1981, 1984, 1987) isolated several species of Bacteroides from the normal oral flora and from oral and other soft-tissue lesions of cats, and we have isolated strains of B. gingivalis, Bacteroides Group E, and Bacteroides species from an oral abscess in one cat. Our data show that a group of cats diagnosed as having plasma cell gingivitis-pharyngitis had significantly elevated mean titers of serum IgG reactive with antigens of human isolates of A. actinomycetemcomitans, B. gingivalis, and B. intennedius, and cat isolates of B. gingivalis and Bacteroides Group E. These observations are consistent with, but do not prove, infection by these or cross-reacting species. Recent evidence indicates that the class of antigens most recognized by high-titer human serum is carbohydrate (Califano et al., 1989; Sims et al, 1989). We performed additional tests to determine the serodominant antigen in affected cats. Titers against purified LPS from human isolates of B. gingivalis and B. intennedius, as well as with LPS from E. coli, were elevated when assessed by ELISA, indicating that LPS from these or cross-reacting species may have stimulated an immune response in affected cats. It is notable that E. coli has been isolated from the oral flora of cats (Johnessee and Hurvitz, 1984). Since the ELISA assays did not determine the relative extent of antibody binding to antigen classes, dot blots and Western blots were prepared. LPS is a quantitatively minor antigen when assessed with affected cat serum. The degree of binding of IgG to purified LPS was small relative to binding to whole bacterial homogenates, while binding to the purified protein fraction from A. actinomycetemcomitans was greatly enhanced, compared with that of homogenates (Fig. 3). Similar results were obtained for binding of IgM. Immunoblot analysis revealed that periodate oxidation, a process that destroys the carbohydrate moiety of LPS, did not significantly alter IgG binding to homogenates of A. actinomycetemcomitans, B. intennedius, B. gingivalis, or to purified protein from A. actinomycetemcomitans (Fig. 4). In marked contrast to cats, IgG in human periodontitis patient serum bound intensively to antigens in the LPS preparation (Figs. 3 and 4), and binding was abrogated by periodate oxidation of the LPS fraction (Fig. 4). While our data are consistent with antibody binding to an integral carbohydrate antigen of LPS, we cannot rule out binding to a contaminating polysaccharide. Contamination by capsular polysaccharide seems unlikely, however, since Zambon et al. (1984) reported that these molecules cannot enter the gels. The observation made when human patient serum was used serves as a positive control for the immunoblots developed with use of affected cat serum, which showed absence of anti-LPS an-


VoL 69 No. 3

ANTIBODIES IN CATS WITH PERIODONTAL DISEASE

tibodies, and demonstrates that periodate treatment effectively destroyed LPS carbohydrate antigens. Thus, while affected cat and human patient serum samples both contain antibodies reactive with antigens expressed by putative periodontal pathogens, cats appear to recognize a different class of antigens than do humans. The demonstration by ELISA of elevated anti-LPS antibodies in affected cat serum and the detection of little or no antiLPS antibody by dot-blot and immunoblot assay require explanation. Our ELISA data are endpoint titer values, and therefore reflect antibody binding at very high serum dilutions. Endpoint titers are more strongly influenced by antibody concentration than by avidity of binding (Caulfield and Shaffer, 1984). In contrast, the immunoblot assays are performed at much higher serum concentrations, and therefore are more dependent upon antibody avidity for the test antigens than on antibody concentration. The combined data are consistent with the idea that affected cats may in fact have elevated titers of anti-LPS antibody, but it may be of very low avidity and therefore lack the functional activity needed to confer protection against infection. A. actinomycetemcomitans and species of Bacteroides have been implicated in the etiology of periodontitis in humans, non-human primates, and dogs. One manifestation of infection in humans is the presence of elevated levels of serum antibodies reactive with the antigens of these bacteria. We and others have isolated Bacteroides species from the oral flora of domestic cats. These observations, combined with the known pathogenic potential of many of these species, and our observations of the elevated levels of serum antibodies reactive with antigens expressed by these bacteria, indicate their participation in the initiation and perpetuation of plasma cell gingivitispharyngitis in cats. If this is the case, additional studies in cats may shed new light on the role of the immune response in humans. Our observations do not bear on the more basic question of why, as in humans, some cats are susceptible to the disease while others are not. We speculate that susceptible cats may not have the ability to produce high-avidity antibodies or antibodies to carbohydrate antigens that might limit the infection. Alternatively, susceptible cats may have some other abnormality in host defense, such as in the phagocytic leukocytes, as is the case in humans with some forms of early-onset aggressive periodontitis (Page et al., 1983). REFERENCES BARRETT, R.E.; POST, J.E.; and SCHULTZ, R.D. (1975): Chronic Relapsing Stomatitis in a Cat Associated with Feline Leukemia Virus Infection, Feline Pract 5:34-38. CALIFANO, J.V.; SCHENKEIN, H.A.; and TEW, J.G. (1989): Immunodominant Antigen of Actinobacillus actinomycetemcomitans Y4 in High-Responder Patients, Infect Immun 57:1582-1589. CAULFIELD, M. and SHAFFER, D. (1984): A Computer Program for Evaluation of ELISA Data Obtained Using an Automated Microtiter Plate Absorbance Reader, J Immunol Methods 74:205215. COTTER, S.M.; HARDY, W.D.; and ESSEX, M. (1975): Association of Feline Leukemia Virus with Lymphosarcoma and Other Disorders in Cats, JAm Vet Med Assoc 166:449-454. DUBOIS, M.; GILLES, K.A.; HAMILTON, J.K.; REBERS, P.A.; and SMITH, F. (1956): Colorimetric Method for Determination of Sugars and Related Substances, Analyt Chem 28:350-355. EBERSOLE, J.L.; FREY, D.E.; TAUBMAN, M.A.; and SMITH, D.J. (1980): An ELISA for Measuring Serum Antibodies to Actinobacillus actinomycetemcomitans, J Periodont Res 15:621-632. FROST, P. and WILLIAMS, C.A. (1986): Feline Dental Diseases, Vet Clin NAm 16:851-873. GUNSOLLEY, J.C.; BURMEISTER, J.A.; TEW, J.G.; BEST, A.M.; and RANNEY, R.R. (1987): Relationship of Serum Antibody to

881

Attachment Level Patterns in Young Adults with Juvenile Periodontitis or Generalized Severe Periodontitis, J Periodontol 58:314320. HAFFAJEE, A.D.; SOCRANSKY, S.S.; DZINK, J.L.; TAUBMAN, M.A.; and EBERSOLE, J.L. (1988): Clinical, Microbiological and Immunological Features of Subjects with Refractory Periodontal Diseases, J Clin Penodontol 15:390-398. HITCHCOCK, P.J. and BROWN, T.M. (1983): Morphological Heterogeneity Among Salmonella. Lipopolysaccharide Chemotypes in Silver-stained Polyacrylamide Gels, J Bacteriol 154:269-277. HOLDEMAN, L.V.; CATO, E.P.; and MOORE, W.E.C., Eds. (1977): Anaerobe Laboratory Manual, 4th ed., Blacksburg, VA: VPI Anaerobe Laboratory, Virginia Polytechnic Institute and State University, 152 pp. HOLDEMAN, L.V.; KELLY, R.W.; and MOORE, W.E.C. (1984): Family I. Bacteroidaceae Pribram 1983. In: Bergey's Manual of Systematic Bacteriology, Vol. 1, N.R. Kreig and J.G. Holt, Eds., Baltimore: Williams and Wilkins, pp. 602-662. JOHNESSEE, J.S. and HURVITZ, A.I. (1984): Feline Plasma Cell Gingivitis-Pharyngitis, JAm Animal Hosp Assoc 19:179-181. LAEMMLI, U.K. (1970): Cleavage of Structural Proteins During the Assembly of the Head of Bacteriophage T4, Nature (London) 227:680-685. LOVE, D.N.; JOHNSON, J.L.; JONES, R.F.; and CALVERLEY, A. (1987): Bacteroides salivosus sp. nov., an Asaccharolytic, BlackPigmented Species from Cats, Int J System Bacteriol 37:307-309. LOVE, D.N.; JONES, R.F.; and BAILEY, M. (1981): Characterization of Bacteroides Species Isolated from Soft Tissue Infections in Cats, JAppl Bacteriol 50:567-575. LOVE, D.N.; JONES, R.F.; and CALVERLEY, A. (1984): Asaccharolytic Black-Pigmented Bacteroides Strains from Soft Tissue Infections in Cats, Int J System Bacteriol 34:300-303. LOWRY, O.H.; ROSEBROUGH, N.J.; FARR, A.L.; and RANDALL, R.J. (1951): Protein Measurement with the Folin Phenol Reagent, J Biol Chem 193:265-275. MOORE, W.E.C. (1987): Microbiology of Periodontal Disease, J Periodont Res 22:335-341. PAGE, R.C.; BOWEN, T.; ALTMAN, L.; VANDESTEEN, E.; OCHS, H.; MacKENZIE, P.; OSTERBERG, S.; ENGEL, L.D.; and WILLIAMS, B.L. (1983): Prepubertal Periodontitis. I. Definition of a Clinical Disease Entity, J Periodontol 54:257-271. PAGE, R.C. and SCHROEDER, H.E. (1982): Periodontitis in Man and Other Animals. A Comparative Review, Basel: S. Karger, 330 pp. REICHART, P.A.; DURR, U.-M.; TRIADAN, H.; and VICKENDY, G. (1984): Periodontal Disease in the Domestic Cat. A Histopathologic Study, J Periodont Res 19:67-75. SCHENCK, K. (1985): IgG, IgA, and IgM Serum Antibodies Against Lipopolysaccharide from Bacteroides gingivalis in Periodontal Health and Disease, J Periodont Res 20:368-377. SIMS, T.; MONCLA, B.; DARVEAU, R.; and PAGE, R. (1989): Antigenic Reactivity of Sera from Periodontitis Patients Positive for Actinobacillus, J Dent Res 68:222, Abst. No. 325. SYED, S.A. (1980): Characteristics of Bacteroides asaccharolyticus from Dental Plaque of Beagle Dogs, J Clin Microbiol 11:522526. THOMPSON, R.R.; WILCOX, G.E.; CLARK, W.T.; and JANSEN, K.L. (1984): Association of Calicivirus Infection with Chronic Gingivitis and Pharyngitis in Cats, J Small Animal Pract 25:207210. TOWBIN, H.; STAEHELIN, T.; and GORDON, J. (1979): Electrophoretic Transfer of Proteins from Polyacrylamide Gels to Nitrocellulose Sheets: Procedure and Some Applications, Proc Natl Acad Sci USA 76:4350-4354. VON SCHLUP, D. (1982): Epidemiologische und morphologische Untersuchungen am Katzengebiss I. Mitteilung: Epidemiologische Untersuchungen, Kleinterpraxis 27:86-94. WESTPHAL, 0. and JANN, K. (1965): Bacterial Lipopolysaccharides: Extraction with Phenol-Water and Further Applications of the Procedure. In: Methods in Carbohydrate Chemistry, Vol. 5, R.L. Whistler, Ed., New York: Academic Press, pp. 83-91. WILLIAMS, B.L.; EBERSOLE, J.L.; SPEKTOR, M.D.; and PAGE, R.C. (1985): Assessment of Serum Antibody Patterns, and Analysis of Subgingival Microflora of Members of a Family with a High


882

SIMS et al.


Prevalence of Early-Onset Periodontitis, Infect Immun 49:742750. WRIGHT, B.G. and REBERS, P.A. (1972): Procedure for Determining Heptose and Hexose in Lipopolysaccharides, Analyt Biochem 49:307-313.

ZAMBON, J.J.; SLOTS, J.; MIYASAKI, K.; LINZER, R.; COHEN, R.; LEVINE, M.; and GENCO, R.J. (1984): Purification and Characterization of Serotype c Antigen from Actinobacillus actinomycetemcomitans, Infect Immun 44:22-27.


Journal of Dental Research

Journal of Dental Research

Suggest Documents

Dental Research Journal Dental Research Journal