Serum antibody responses of cats to soluble whole cell antigens of feline Porphyromonas gingivalis

Veterinary Microbiology 73 (2000) 37±49

Serum antibody responses of cats to soluble whole cell antigens of feline Porphyromonas gingivalis Jacqueline M. Norris, Daria N. Love* Department of Veterinary Anatomy and Pathology, University of Sydney, Sydney, NSW 2006, Australia Received 26 July 1999; received in revised form 23 November 1999; accepted 16 December 1999

Abstract The whole cell soluble antigens of two strains (VPB 3457 and VPB 3492) of feline Porphyromonas gingivalis were analysed by Western blotting using serum taken from 40 domestic cats with various grades of periodontal disease. Five strongly immunogenic protein bands (70, 34, 27, 24 and 19 kDa) from VPB 3457 and seven from VPB 3492 (58, 44, 34, 27, 25, 24 and 21 kDa) were selected for further study. A signi®cant positive correlation was found between the serum antibody response to the 70, 34, 27, 24 and 19 kDa bands of VPB 3457 and the 58, 44, 25, 24 and 21 kDa bands of VPB 3492 and the overall periodontal grade. A signi®cant positive correlation was also found between the serum antibody response to the 24 kDa band of VPB 3457 and the total colony forming units of P. gingivalis. N-terminal sequencing of the 44 kDa band of VPB 3492 showed 75% identity with the translated amino acids from the hag A (haemagglutinin) gene of a human strain of P. gingivalis and N-terminal amino acid sequence of the 27 kDa band of VPB 3457 showed 88% identity with the amino acid sequences translated from DNA of purported genes coding for variously named proteinases of human strains of P. gingivalis. # 2000 Elsevier Science B.V. All rights reserved. Keywords: Porphyromonas gingivalis; Cat; Immune response Ð cat; Periodontitis; Proteinases

*

Corresponding author. Tel.: ‡61-2-9351-2454; fax: ‡61-2-9351-7348. E-mail address: [email protected] (D.N. Love) 0378-1135/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 1 3 5 ( 0 0 ) 0 0 1 5 3 - X

38

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

1. Introduction Knowledge of the primary bacterial agents involved, their interactions with each other and the periodontal tissue, and the modifying role of the host are essential to our understanding of the pathogenesis of periodontal disease in humans and animals. The precise means by which organisms such as Porphyromonas gingivalis (PgH) colonize and then initiate and sustain tissue destruction in periodontal disease in humans is still uncertain. Bacterial surface antigens of PgH and other suspected periodontal pathogens are major targets of the host response (Gmur et al., 1988) and reports have focused on determining the major immunogenic cell surface antigens of PgH as a means of gaining a greater understanding of the interaction between host and parasite and to identify new virulence factors or to place greater credence on those identi®ed previously. Using Western blotting, Curtis et al. (1991) determined that the 50, 47 and 40 kDa proteins were important surface antigens of PgH and that they were highly reactive with serum IgG from adult periodontal patients. The 50 kDa band has been characterized subsequently as a proteinase antigen prpR1, that is synthesized as part of a large polyprotein precursor to a number of different forms of the extracellular product (AduseOpoku et al., 1995). Boutsi et al. (1996) analyzed whole cell extracts and ®mbriae from PgH 381 using sera from patients with adult periodontitis, rapidly progressive periodontitis or juvenile periodontitis and healthy patients. Using Western analysis, reactivity to two major protein bands of 43 and 41 kDa (the latter of which corresponded to the ®mbrillin protein) was prominent in patients with periodontitis. Chen et al. (1995) found the 75, 55 and 43 kDa protein antigens of PgH were the components most frequently and strongly recognized from patients with rapidly progressive periodontitis. Laosrisin et al. (1990) studied antibody responses to PgH 381 whole cell sonicates and demonstrated that the 82, 57 and 44 kDa bands were more frequently reactive in patients with adult periodontitis compared with healthy subjects. Ebersole and Steffen (1995) found immunological responses in patients to a similar molecular weight range of antigen bands when they compared four serotypes of PgH. While there were differences in the serum responses of patients with periodontitis to the various antigens of the different serotypes, they concluded that PgH had a fairly consistent antigenic composition that would allow strategies to be devised to interfere immunologically with disease caused by PgH. Similar studies to determine the important periodontopathogens of cats have not been reported. Norris and Love (1999) have shown a signi®cant association between grade of periodontal disease and numbers of feline strain Porphyromonas gingivalis (PgF), thus establishing their importance in this disease process. In preliminary investigations, Norris and Love (1995) found that cats with periodontal disease showed an immune response to certain protein antigens of PgF and other feline Porphyromonas species. This report details the investigation of two strains of PgF which share 68±76% DNA homology with P. gingivialis ATCC 33277T. These ®ndings show a signi®cant association between the grade of serum antibody response to particular immunoreactive antigens of PgF and the overall periodontal grade. Two of these antigens were identi®ed by N-terminal sequencing as having signi®cant similarity to gene sequences of important purported proteins of PgH.

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

39

2. Materials and methods 2.1. Bacterial strains Two strains of feline Porphyromonas gingivalis (PgF) were used in the study. Veterinary Pathology and Bacteriology (VPB) 3457 was isolated from a subcutaneous abscess and VPB 3492 was isolated from a case of pyothorax. Both were identi®ed by phenotypic characteristics and DNA-DNA hybridization and share 68 and 76% DNA homology, respectively, with P. gingivalis ATCC 33277T (Love et al., 1987). 2.2. Preparation of soluble whole cell antigens Single colonies from ®ve-day anaerobic growth of these organisms were subcultured onto individual supplemented 5% de®brinated sheep blood agar plates and grown anaerobically for three days. Bent glass Pasteur pipettes were used to harvest the bacteria (200 mg of cells per strain) separately into chilled microcentrifuge tubes containing 400 ml of sterile phosphate buffered saline (PBS). Cells were washed three times by centrifugation and re-suspension in PBS. The tubes were gently vortexed for 2 min to ensure an even cell suspension before cell disruption in a cup horn of a sonicator cell disruptor (Model W-375, Heat Systems Ultrasonics, Plainview, NY) at 60% duty cycle, output 6 for ®ve bursts of 3 min each. Samples were kept chilled with iced water circulating in the cup horn during sonication and were kept on ice between bursts. The cell debris and unbroken cells were removed by centrifugation using a Beckman microcentrifuge Model 12 at 12 000g for 15 min at 48C. The supernatant ¯uid comprised the soluble whole cell antigen preparation which was diluted with sterile PBS to a protein concentration of 1 mg/ml estimated using Bio-Rad Protein Assay (Bio-Rad Laboratories, CA). Each diluted sample was mixed with an equal volume of 2SDS sample buffer (0.5 M Tris HCl, pH 6.8, 2% SDS, 200 mM DTT, 20% glycerol, 0.05% bromophenol blue) to give a ®nal concentration of 0.5 mg/ml soluble whole cell proteins. Before use, each cell sample was heated for 3 min at 958C in a water bath, cooled on ice and then centrifuged at 12 000g for 3.min at 48C. 2.3. Selection of cats, periodontal grading and microbial sampling These methods were as described previously (Norris and Love, 1999). Brie¯y, 40 domestic cats ranging in age from 6 months to 17 years, both sexes and various breeds with different grades of periodontal disease were included in the study. Blood samples were taken from each cat and serum stored at ÿ208C. The right upper canine and premolar periodontia were assessed and graded according to the method of West-Hyde and Floyd (1995) and an overall periodontal grade (OPG) determined by comparing grades at each site. Where the grade varied between sites, the higher grade was used as the OPG for that mouth. The colony forming units isolated at the periodontium of the canine and premolar sites were added together to form the total colony-forming units (TCFUs). TCFUs of PgF were enumerated from colony lifts taken from replicate plates

40

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

made from samples taken from each site and probed with a digoxigenin-labeled whole chromosome DNA probe of PgF VPB 3457 (Love et al., 1992; Norris and Love, 1999). The identity of each cat, its overall grade of periodontal disease and the TCFU of PgF are presented in Table 2. 2.4. SDS-Polyacrylamide gel electrophoresis, detection and selection of immunogenic proteins and assessment of serum responses of cats Cell extracts were run on 12% SDS-PAGE using the method of Laemmli (1970). Western blotting was performed as described previously (Love et al., 1993). Whole cell protein extracts (0.5 mg/ml protein) were transferred to nitrocellulose using a semi-dry transfer method (Kyhse-Andersen, 1984) and probed with a standard dilution of serum from each of the 40 cats described in Section 2.3. Serum from each cat was run a minimum of three times against the antigen preparation. Three blots were then chosen from each cat and the responses to each band used to determine the most reactive bands recognized by each cat. From this analysis, the bands recognized most consistently and strongly in each strain across the 40 cats was determined and these were selected for further study. The serum antibody response of each cat to each of the selected protein bands was graded as strong (‡‡‡), moderate (‡‡), weak (‡), or absent (ÿ) by averaging the three responses of each cat to each band. 2.5. Statistical analysis Data were collected regarding the grade of serum antibody response from each cat to each soluble whole cell antigen, the OPG and the TCFU, and analyzed statistically from programs in Microsoft Excel Version 7.01. 2.6. N-terminal sequencing Cell extracts from VPB 3457 and VPB 3492 were each mixed with an equal volume of freshly prepared SDS sample buffer and incubated at 378C for 10 min. Samples were electrophoresed using a running buffer of 0.384 M glycine, 0.1% SDS in 0.05 M Tris buffer, pH 8.3 in a 12% SDS-polyacrylamide gel that had been left for 24 h to polymerize at room temperature. Electrophoresed samples were then transferred to polyvinylidene di¯uoride (PVDF) membrane (BioRad Transblot) by the semi-dry method (KyhseAndersen, 1984) using a buffer containing 10 mM 3-[cyclohexylamino]-1-propanesulfonic acid (CAPS), 10% methanol, pH 11.0. A reference strip was cut from the PVDF membrane and immunostained using serum from a cat with a ‡‡‡ reaction to the protein of interest to enable location of the required protein. The remaining part of the membrane was stained with 0.0125% Coomassie blue in 50% methanol for 3 min before being destained in 50% methanol for 3 min. The required protein was identi®ed by alignment with the immunostained portion and was then cut out and N-terminal sequencing performed by Edmon degradation on a Porton Instruments PI2090 equipped with a Beckman Systems Gold micro PTH (phenylthiohydantoin) amino acid analyzer.

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

41

Fig. 1. Western blot of soluble whole cell antigens of feline Porphyromonas gingivalis (a) VPB 3457 and (b) VPB 3492 using sera from cats with various grades of periodontal disease. M, molecular weight of the selected proteins. Antigen reacted against serum from cats: (a) 1, J82; 2, J81; and 3, J80. Grades of periodontal disease and grading of serum responses of these cats are given in Table 2. (b) 1, J.34; 2, J50, 3, J36. Grades of periodontal disease of these cats is given in Table 2.

3. Results 3.1. Serum antibody responses to feline Porphyromonas gingivalis VPB 3457 3.1.1. Percentage of cats with an immunological response to selected soluble whole cell antigens Five immunogenic bands were chosen as shown in Fig. 1a. The estimated molecular weight of these protein bands was 70, 34, 27, 24, and 19 kDa. Table 1 lists the percentage

Table 1 The percentage of cats in each overall periodontal grade with a serum antibody response to each of the selected immunogenic protein bands of feline Porphyromonas gingivalis strains VPB 3457 and VPB 3492 Overall periodontal grade (n)a

Molecular weight (kDa) of selected protein bands VPBb 3457 70

0 1 2 3 4 5 6

(3) (4) (6) (19) (3) (3) (2) a

c

0 25 100 74 67 100 100

VPB 3492

34

27

24

19

58

44

34

27

25

24

21

67 75 100 95 100 100 100

100 50 67 95 100 100 100

0 0 50 68 100 100 0

0 0 50 63 100 100 0

100 75 100 100 100 100 100

67 50 100 100 100 100 100

33 25 67 95 100 100 50

33 50 100 95 100 100 100

33 50 100 100 100 100 100

33 50 100 100 100 100 100

0 0 67 95 100 100 50

Number of cats in each grade. Veterinary Pathology and Bacteriology culture collection number c Percentage of cats with serum antibody responses to each immunogen. b

42

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

of cats in each OPG with a serum antibody response to each of the selected protein bands of PgF VPB 3457. 3.1.2. Relationship between the overall periodontal grade and grade of serum antibody response to the selected immunogenic protein bands Table 2 shows the grade of serum antibody response to the selected immunogenic protein bands of VPB 3457, the OPG, and the TCFU of PgF isolated from each cat and arranged in ascending order of periodontal grade. When a regression analysis was performed, a positive correlation was found between the OPG and the grade of serum antibody response to the 70 kDa band (pˆ0.0254), the 34 kDa (pˆ0.0105), the 27 kDa band (pˆ<0.001), the 24 kDa (pˆ0.0147) and 19 kDa band (pˆ0.0258). 3.1.3. Relationship between the total colony forming units and the grade of serum antibody response to the selected immunogenic protein bands Regression analysis of the TCFU of PgF isolated against the grade of serum antibody response to each of the selected protein bands of VPB 3457 showed that a signi®cant correlation existed between the 24 kDa band and the TCFU of PgF (pˆ0.0344). Examination of the serum antibody responses of cats to the 27 kDa band showed that three cats (J41, J75, J74) with a large number of PgF isolated (between 4.84105 and 1.36106) and OPGs of three, six and six, respectively, had weak serum antibody responses to the 27 kDa band, while one cat (J71) with a very small number of PgF isolated and an OPG of two, had a strong serum antibody response to this band. It would appear therefore that, while positive correlation can be seen between some of the protein bands and the total CFU of PgF, the relationship between these variables was not direct or simple. 3.1.4. N-terminal sequence for 27 kDa band of PgF VPB 3457 The N-terminal amino acid analysis yielded the sequence ANEAKVVLVADNVQGDN which has been deposited in Swiss-Prot as Accession No. P81784. This sequence shares 88% identity with amino acid sequences translated from DNA of PgH deposited in the GenBank database as L27483, L26341, U75366, U68468, D83258, U15282, X82680, U42210, AF017059, U54691 and another reported 27 kDa band (Table 3). These gene sequences have been described as prtH, prpR1, and prtP and encode for variously named proteinases, such as arginine-speci®c thiol protease, lysine-speci®c cysteine proteinase, cysteine protease, haemagglutinin, Lys-gingipain, Arg-gingipain-1 proteinase, and porphypain. 3.2. Serum antibody responses to feline Porphyromonas gingivalis VPB 3492 3.2.1. Percentage of cats with an immunological response to selected soluble whole cell antigens Seven immunogenic protein bands with estimated molecular weights of 58, 44, 34, 27, 25, 24, and 21 kDa were selected as seen in Fig. 1b. Table 1 lists the percentage of cats in each OPG with a serum antibody response to each of the selected protein bands of PgF VPB 3492.

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

43

Table 2 Association between the overall periodontal grade and the grade of serum antibody response to the selected immunogenic protein bands of Porphyromonas gingivalis VPB 3457 in each cat Cat no.

J31 J33 J45 J46 J30 J32 J69 J28 J78 J65 J60 J76 J71 J67 J68 J53 J36 J41 J35 J40 J34 J39 J73 J27 J64 J77 J59 J54 J63 J51 J79 J58 J50 J61 J80 J56 J82 J81 J75 J74 a b

Overall periodontal grade

TCFUa (103)

Grade of serum antibody response to respective immunogenb 70

34

27

24

19

0 0 0 1 1 1 1 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 5 5 5 6 6

27 7 80 93 12 68 114 33 8 62 40 93 10 118 200 248 110 484 259 536 75 260 33 90 308 590 168 358 172 210 228 396 750 800 164 300 832 544 560 1360

ÿ ÿ ÿ ÿ ÿ ÿ ‡‡ ‡ ‡‡ ‡‡‡ ‡ ‡‡‡ ‡ ‡ ÿ ÿ ‡ ‡ ‡‡ ‡ ‡‡ ‡ ‡‡ ‡ ÿ ‡‡ ‡‡ ‡‡ ‡‡ ÿ ÿ ‡‡ ‡‡ ‡ ÿ ‡‡ ‡‡ ‡‡ ‡ ‡‡

ÿ ‡ ‡‡ ÿ ‡ ‡ ‡‡ ‡ ‡ ‡‡ ‡‡ ‡‡ ‡‡‡ ÿ ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡‡ ‡‡ ‡‡ ‡‡ ‡‡ ‡‡‡ ‡‡‡ ‡‡‡ ‡‡‡ ‡‡‡ ‡‡‡ ‡ ‡‡ ‡‡‡ ‡ ‡‡‡ ‡‡‡ ‡‡ ‡‡‡

‡ ‡ ‡ ÿ ÿ ‡ ‡ ÿ ‡ ÿ ‡‡ ‡‡ ‡‡‡ ‡ ÿ ‡ ‡ ‡ ‡‡ ‡‡ ‡‡‡ ‡ ‡‡ ‡‡ ‡‡ ‡‡ ‡ ‡ ‡‡ ‡‡ ‡‡ ‡‡ ‡‡‡ ‡‡ ‡‡‡ ‡‡‡ ‡‡ ‡‡‡ ‡ ‡

ÿ ÿ ÿ ÿ ÿ ÿ ÿ ÿ ‡ ÿ ‡ ‡ ÿ ÿ ÿ ‡ ‡ ‡ ‡ ‡ ‡‡ ÿ ÿ ‡ ÿ ‡‡ ÿ ‡ ‡ ‡ ‡ ‡ ‡‡ ‡ ‡ ‡ ‡‡ ‡ ÿ ÿ

ÿ ÿ ÿ ÿ ÿ ÿ ÿ ‡ ÿ ÿ ‡ ‡ ÿ ÿ ‡ ‡‡ ‡ ÿ ‡ ‡‡ ‡ ÿ ÿ ‡‡ ÿ ‡ ÿ ‡ ‡‡ ‡ ÿ ‡ ‡ ‡ ‡‡ ‡ ‡‡ ‡ ÿ ÿ

Total colony-forming units. ‡‡‡, Strong; ‡‡, moderate; ‡, weak; ÿ, absent response to each band.

44

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

Table 3 N-terminal sequence of the 27 kDa band of feline Porphyromonas gingivalis VPB 3457, compared with amino acid sequences translated from DNA of human strains of Porphyromonas gingivalis (the amino acids in bold type represent mismatches) Database Accession No.

N-terminal amino acid sequence or translated amino acid sequence

Nucleotide position

Reading frame of DNA

Swiss-Prot P81784 GenBank L27483 GenBank L26341 GenBank U75366 GenBank U68468 GenBank D83258 GenBank U15282 GenBank X82680 GenBank U42210 GenBank AF017059 GenBank U54691 27 kDac

ANEAKVVLVADNVQGDNa ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN ANEAKVVLAADNVWGDN

n.ab 1540±1590 4378±4428 2907±2957 2906±2956 2931±2981 5236±5286 4921±4971 2907±2957 2949±2999 2863±2913 n.a

n.a ‡1 ‡1 ‡3 ‡2 ‡3 ‡1 ‡1 ‡3 ‡3 ‡1 n.a

a

Amino acids in bold represent mismatches from the feline P. gingivalis sequence. Not applicable. c Pike et al. (1994). b

3.2.2. Relationship between the overall periodontal grade and the grade of serum antibody response to the selected immunogenic protein bands Regression analysis showed there was a signi®cant positive correlation between the OPG and the grade of serum antibody response to the 58 kDa band (pˆ0.0352) and the 44 kDa band (pˆ0.0421). However, no statistically signi®cant associations were found between the OPG and the grade of serum antibody response to the 34 and 27 kDa bands. Closer examination of the data showed that four of 27 cats (J75, J74, J77, J61) with an OPG of 3±6 had high numbers of PgF (>5.60105), but weak or absent serum antibody responses to these bands. There was a positive correlation between the OPG and the grade of serum antibody response to the 25 kDa band (p<0.001), the 24 kDa band (pˆ0.0412) and the 21 kDa band (pˆ0.0238). 3.2.3. Relationship between the total colony forming units and the grade of serum antibody response to the selected immunogenic protein bands Regression analysis of the TCFU of PgF isolated from each cat against the grade of serum antibody response to each of the selected immunogenic protein bands of VPB 3492 showed there was no signi®cant correlation between these variables. Examination of serum antibody responses to the 58, 44, and 34 kDa bands showed that four of 27 cats (J61, J74, J75, J77) with an OPG of between three and six, had high numbers of PgF (>5.60105) but had weak or absent serum antibody responses to PgF VPB 3492. The serum antibody response of cats to the 27 and 21 kDa bands was variable and did not correlate directly with the TCFU of PgF. It is noteworthy that strong serum antibody responses were only seen in cats with <3.96105 of PgF.

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

45

3.2.4. N-Terminal sequence for 44 kDa band of PgF VPB 3492 The N-terminal amino acid analysis of the 44 kDa band of PgF VPB 3492 yielded the sequence APYQKRNI which has been deposited in Swiss-Prot as Accession No. P81886. It shares 75% identity with the amino acid sequence (APYQERTI) translated from nucleotides 1464±1487 of GenBank Accession No. U41807 from PgH 381. 3.3. Comparison between PgF VPB 3457 and PgF VPB 3492 Polyacrylamide Gel Electrophoresis (PAGE) showed different protein patterns for each strain (data not shown). These differences were also re¯ected in the serum antibody response of cats as assessed by Western blots. Five dominant immunogenic protein bands were selected from VPB 3457, while seven were selected for VPB 3492. Both strains had immunogenic bands with approximate molecular weights of 34, 27 and 24 kDa, although identity across strains of these bands was not established. The grade of serum antibody response in each cat to the 34 kDa band of VPB 3457 and VPB 3492 was the same between strains in nine of 40 cats (22.5%). For the 27 kDa band of VPB 3457 and VPB 3492, the grade of serum antibody response in each cat was the same between strains in 13 of 40 cats (32.5%). For the 24 kDa band of VPB 3457 and VPB 3492, the grade of serum antibody response in each cat was the same between strains in eight of 40 cats (20%). A positive correlation was found between the OPG and the grade of serum antibody response to 24 kDa band in both strains. 4. Discussion The current study represents the ®rst report in which the serum antibody responses of cats to the whole cell antigens of PgF have been analyzed in conjunction with data relating to the periodontal grade and the number of each Porphyromonas species isolated from each cat. In so doing, these studies have ful®lled one of the many important criteria required to establish an organism as a periodontal pathogen (Haffajee and Socransky, 1994), i.e. by determining that serum antibody responses of cats were directed against these species. For PgF VPB 3457, the cats showed strong serum antibody responses to the 27 kDa band and a positive correlation was found between the grade of serum antibody response to this band and the OPG. Watanabe et al. (1989) analyzed sonicated whole cell extracts from PgH W50 by Western analysis and found that amongst the 12 major immunogenic bands, the 27 kDa band reacted more frequently with sera from adult patients with severe periodontitis than sera from controls and those with mild periodontitis. The 86% identity of the N-terminal amino acid sequence of the 27 kDa protein from PgF VPB 3457 with the variously named proteinase of PgH, most frequently called Arg-gingipain, may, by implication, give greater importance to PgF in periodontal disease. Arg-gingipain has been implicated in many of the destructive and evasive properties of PgH, including the disruption of neutrophil function, degradation of immunoglobulin, complement factors and collagen, haemagglutination, and the activation of the coagulation and in¯ammatory pathways (Shah et al., 1992; Kadowaki et al., 1994; Imamura et al., 1995, 1997; Jagels et al., 1996; Tokuda et al., 1998) and its role in the

46

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

pathogenicity of PgH has attracted recent interest. Imamura et al. (1997) found that Arggingipain readily induced blood coagulation in vitro and suggested that bacterial proteinases such as this may induce disseminated intravascular coagulation during sepsis. Therefore, the purported functions attributed to this cysteine proteinase may provide PgH with the ability to obtain its speci®c nutritional requirements, assist in binding to the dental plaque, evade the immune response, and assist in the destruction of host tissue. The present study has established the presence of a similar proteinase in PgF. The 75% amino acid identity of the N-terminal sequence of the 44 kDa protein band of VPB 3492, with a region in the DNA sequence of PgH strain 381 reported by Han et al. (1996) to be a haemagglutinin, could be of interest to those workers suggesting its importance as a virulence factor in PgH. Haemagglutinins are generally expressed on the cell surface of bacteria in association with either ®lamentous structures, such as ®mbriae, or non®mbrial surface components. The haemagglutinating activity of PgH and PgF has been reported previously (Okuda et al., 1981; Love et al., 1984, 1987). PgH can utilize hemin as an iron source (Brahmanti and Holt, 1991) and, therefore, the ability to attach to and lyse red blood cells has been considered an important survival mechanism. In addition, there is mounting evidence that the PgH haemagglutinins are closely associated with proteinases produced by this organism. Pike et al. (1994) established that the 50 kDa arg-gingipain molecule of PgH H66 complexed with the 44 kDa haemagglutinins and postulated that this was of great bene®t to the organism in the binding and degradation of the red blood cell surface. Yoneda and Kuramitsu (1996) found that an isogenic mutant of PgH, de®cient in Arg-gingipain cysteine proteinase activity, also had signi®cantly reduced haemagglutination activity compared with wild-type strains. They suggested that the arg-gingipain and haemagglutinating activities of PgH were expressed from a single gene. PgF VPB 3457 and VPB 3492 showed distinctly different protein patterns in the SDSPAGE system indicating that PgF is heterogeneous. This was also shown during Western blotting in which ®ve immunogenic proteins of PgF VPB 3457 were strongly recognized by the cats, whereas seven proteins of PgF VPB 3492 were highly immunoreactive, with only three proteins sharing a common molecular weight between the two strains. Differences between strains of PgH with respect to antigenicity and pathogenicity have been realized for some time (Evans et al., 1992). Ebersole and Steffen (1995) used SDSPAGE and Western blotting to assess the antigenic composition of the outer envelopes of three serotypes of PgH: ATCC 33277T; W50; and A7A1-28. They found distinctly different protein patterns from each serotype and this was re¯ected in the Western blots. There was, however, molecular weight similarity between some of the major proteins from each serotype. Chen et al. (1995) found PgH ATCC 33277T and 381 were very similar, if not identical, antigenically as assessed by Western analysis, but that these two strains differed greatly from PgH A7A1-28 and W50. Drucker et al. (1998) used phospholipid analog pro®ling by fast-atom bombardment mass spectrometry to compare strains of P. gingivalis of human and animal origin. They found differences between strains within the same animal species and concluded that P. gingivalis was heterogeneous. The distinct differences in protein and antigenic pattern in the two strains of PgF used in this study provide further support for the heterogeneity of PgF. Throughout the study, there were numerous examples of cats that showed strong antibody responses to certain proteins, yet had a lower periodontal grade and small

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

47

numbers of Porphyromonas spp. (J71 serum antibody response to the 27 kDa protein of PgF 3457, for example). In these cases, the strong serum antibody response may re¯ect past exposure in which an effective antibody response has restricted the number of bacteria present. These may represent animals which are less likely to progress to advanced periodontitis due to control of bacterial numbers by the production of effective antibodies that do not contribute further to periodontal destruction. It is also feasible that these cats have been colonized by less virulent strains that are more easily restricted in number. There were also examples of cats that showed weak or absent serum antibody responses to certain proteins despite having large numbers of Porphyromonas present and a moderate-to-severe periodontal disease (e.g. serum antibody response of J74 to the 25 kDa and 17 kDa proteins of PgF 3457). These may represent cats in which a weak serum antibody response has allowed the numbers of Porphyromonas to increase and so allowed the progression to periodontitis. Alternatively, the strains of Porphyromonas present in these cats may not have expressed this protein or may have produced a protein with different epitopes exposed for antibody binding. Differences in the antigenicity of proteins produced by PgH have also been reported by others. Loos and Dyer (1992) found considerable heterogeneity in the gene encoding the ®mbrillin subunit (®mA) amongst strains of PgH and suggested that these differences may be important in the function and antigenicity of the ®mbriae. The lack of correlation between the grade of serum response to many of the proteins produced by PgF and the TCFU does not negate the potential importance of these organisms in periodontal disease. While the importance of PgH has clearly been established in periodontal disease in humans, Haffajee et al. (1995), who studied the serum antibody response (using ELISA) to several important periodontal pathogens including PgH and combined this with the number of CFU of each species as determine by a colony-lift method and DNA probes, found that the mean serum antibody response to a species was not directly related to the mean counts of that species in subgingival plaque samples. This study has successfully used the host serum antibody response to identify antigens of PgF using Western blots and has found that two of these proteins are similar to proteins identi®ed from PgH as possible virulence factors. The results of this study, in providing the basis of the antigenic composition of the feline oral Porphyromonas species, may allow the identi®cation of other potentially important proteins and thereby expand our understanding of the mechanisms involved in disease pathogenesis. The apparent heterogeneity of PgF also requires further investigation as it may be of importance not only to periodontal disease of cats, but may provide an explanation for the different clinical outcomes of cat bites in humans and cats. Acknowledgements This work was supported in part by the Australian Research Council. Valuable technical assistance was provided by Lana Patoka and Frank Taeker who prepared bacteriological media, and professional assistance was provided by Denise Wigney.

48

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

References Aduse-Opoku, J., Muir, J., Slaney, J.M., Rangarajan, M., Curtis, M.A., 1995. Characterization, genetic analysis, and expression of a protease antigen (PrpR1) of Porphyromonas gingivalis W50. Infect. Immun. 63, 4744± 4754. Boutsi, E.A., Koseki, T., Nishihara, T., Ishikawa, I., 1996. Characterization of the immunodominant antigens of Porphyromonas gingivalis 381 in high-responder patients. Oral Microbiol. Immunol. 11, 236±241. Brahmanti, T.E., Holt, S.C., 1991. Roles of porphyrins and host iron transport proteins in regulation of growth of Porphyromonas gingivalis W50. J. Bacteriol. 173, 7330±7339. Chen, H.A., Darveau, R.P., Engel, D., Page, R.C., 1995. Immunodominant antigens of Porphyromonas gingivalis in rapidly progressive periodontitis. Oral Microbiol. Immunol. 10, 193±201. Curtis, M.A., Slaney, J.M., Carman, R.J., Johnson, N.W., 1991. Identi®cation of the major surface protein antigens of Porphyromonas gingivalis using IgG antibody reactivity of periodontal case-control serum. Oral Microbiol. Immunol. 6, 321±326. Drucker, D.B., Tavana, A.M., Boote, V., Love, D.N., 1998. Is Porphyromonas gingivalis heterogenous? Rev. Med. Microbiol. 9, S11±S12. Ebersole, J.L., Steffen, M.J., 1995. Human antibody responses to outer envelope antigens of Porphyromonas gingivalis serotypes. J. Periodontol. Res. 30, 1±14. Evans, R.T., Klausen, B., Ramamurthy, N.S., Golub, L.M., S®ntescu, C., Genco, R.J., 1992. Periodontopathic potential of two strains of Porphyromonas gingivalis in gnotobiotic rats. Arch. Oral Biol. 37, 813±819. Gmur, R., Werner-Felmayer, G., Guggenheim, B., 1988. Production and characterization of monoclonal antibodies speci®c for Bacteroides gingivalis. Oral Microbiol. Immunol. 3, 181±186. Haffajee, A.D., Socransky, S.S., 1994. Microbial etiological agents of destructive periodontal diseases. Periodontol. 2000 5, 78±111. Haffajee, A.D., Dibart, S., Kent, R.L., Socransky, S., 1995. Factors associated with different responses to periodontal therapy. J. Clin. Periodontol. 22, 628±636. Han, N., Whitlock, J., Progulske-Fox, A., 1996. The haemagglutinin gene A (hagA) of Porphyromonas gingivalis 381 contains four large, contiguous, direct repeats. Infect. Immun. 64, 4000±4007. Imamura, T., Potempa, J., Pike, R.N., Travis, J., 1995. Dependence of vascular permeability enhancement on cysteine proteinases in vesicles of Porphyromonas gingivalis. Infect. Immun. 63, 1999±2003. Imamura, T., Potempa, J., Tanase, S., Travis, J., 1997. Activation of blood coagulation factor X by argininespeci®c cysteine proteinases (Gingipain-Rs) from Porphyromonas gingivalis. J. Biol. Chem. 272, 16062± 16067. Jagels, M.A., Travis, J., Potempa, J., Pike, R., Hugli, T.E., 1996. Proteolytic inactivation of the leukocyte C5a receptor by proteinases derived from Porphyromonas gingivalis. Infect. Immun. 64, 1984±1991. Kadowaki, T., Yoneda, M., Okamoto, K., Maeda, K., Yamamoto, K., 1994. Puri®cation and characterization of a novel arginine-speci®c cysteine proteinase (argingipain) involved in the pathogenesis of periodontal disease from the culture supernatant of Porphyromonas gingivalis. J. Biol. Chem. 269, 21371±21378. Kyhse-Andersen, J., 1984. Electroblotting of multiple gels: a simple apparatus without buffer tank for rapid transfer of proteins from polyacrylamide to nitrocellulose. J. Biochem. Biophys. Methods 10, 203± 209. Laosrisin, N., Nakashima, K., Ishikawa, I., 1990. Detection of Bacteroides gingivalis antigenic proteins by immunoblotting analysis. J. Periodontol. 61, 261±268. Laemmli, U.K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680±685. Loos, B.G., Dyer, D.W., 1992. Restriction fragment length polymorphism analysis of the ®mbrillin locus, ®mA, of Porphyromonas gingivalis. J. Dental Res. 71, 1173±1181. Love, D.N., Jones, R.F., Calverley, A., 1984. Asaccharolytic black-pigmented Bacteroides strains from soft tissue infections in cats. Int. J. System. Bacteriol. 34, 300±303. Love, D.N., Johnson, J.L., Jones, R.F., Calverley, A., 1987. Bacteroides salivosus sp. nov., an asaccharolytic, black-pigmented species from cats. Int. J. System. Bacteriol. 37, 307±309. Love, D.N., Bailey, G.D., Bastin, D., 1992. Chromosomal DNA probes for the identi®cation of asaccharolytic anaerobic pigmented bacterial rods from the oral cavity. Vet. Microbiol. 31, 287±295.

J.M. Norris, D.N. Love / Veterinary Microbiology 73 (2000) 37±49

49

Love, D.N., Bell, C.W., Pye, D., Edwards, S., Hayden, M., Lawrence, G.L., Boyle, D., Pye, T., Whalley, J.M., 1993. Expression of equine herpesvirus glycoprotein D by using a recombinant baculovirus. J. Virol. 67, 6820 ±6823. Norris, J.M., Love, D.N., 1995. Serum responses of cats with periodontal/gingival disease to feline members of the genus Porphyromonas. Clin. Infect. Dis. 20, S314±S316. Norris, J.M., Love, D.N., 1999. Associations amongst three feline Porphyromonas species from the gingival margin of cats during periodontal health and disease. Vet. Microbiol. 65, 195±207. Okuda, K., Slots, J., Genco, R..J., 1981. Bacteroides gingivalis, Bacteroides asaccharolyticus, and Bacteroides melaninogenicus subspecies: cell surface morphology and adherence to erythrocytes and human buccal epithelial cells. Curr. Microbiol. 6, 7±10. Pike, R., McGraw, W., Potempa, J., Travis, J., 1994. Lysine- and Arginine-speci®c proteinases from Porphyromonas gingivalis. J. Biol. Chem. 269, 406±411. Shah, H.N., Gharbia, S.E., Progulske-Fox, A., Brocklehurst, K., 1992. Evidence for independent molecular identity and functional interaction of the haemagglutinin and cysteine proteinase (gingivain) of Porphyromonas gingivalis. J. Med. Microbiol. 36, 239±244. Tokuda, M., Karunakaran, T., Duncan, M., Hamada, N., Kuramitsu, H., 1998. Role of arg-gingipain A in virulence of Porphyromonas gingivalis. Infect. Immun. 66, 1159±1166. Watanabe, H., Marsh, P.D., Ivanyi, L., 1989. Detection of immunodominant antigens of periodontopathic antigens of periodontopathic bacteria in human periodontal disease. Oral Microbiol. Immunol. 4, 159±164. West-Hyde, L.W., Floyd, M., 1995. Dentistry. In: Ettinger, S.J., Feldman, E.C. (Eds.), Textbook of veterinary internal medicine, 4th Edition. W.B Saunders Company, Philadelphia, pp. 1097±1123. Yoneda, M., Kuramitsu, H.K., 1996. Genetic evidence for the relationship of Porphyromonas gingivalis cysteine protease and hemagglutinin activities. Oral Microbiol. Immunol. 11, 129±134.