Fibronectin in saliva inhibits Porphyromonas gingivalis¢mbria-induced expression of in£ammatory cytokine gene

in mouse macrophages

Yukio Murakami a, Shigemasa Hanazawa b;*, Shoji Tanaka a, Hiroyoshi Iwahashi b,Shigeo Kitano b, Seiichiro Fujisawa a

a Department of Oral Diagnosis, Meikai University School of Dentistry, 1-1 Keyakidai, Sakado city, Saitama 350-0283, Japanb Department of Oral Microbiology, Meikai University School of Dentistry, 1-1 Keyakidai, Sakado city, Saitama 350-0283, Japan

Received 9 June 1998; received in revised form 4 August 1998; accepted 5 August 1998

Abstract

The aim of this study was to examine whether fibronectin in saliva plays a regulatory role in Porphyromonas gingivalisfimbria-mediated pathogenesis in adult periodontal disease. Our previous study demonstrated that fibronectin is one of thebinding proteins of P. gingivalis fimbrillin. In the present study, we observed that fibronectin in saliva binds specifically to thefimbrillin. The fibronectin content in saliva of adult periodontal patients was significantly lower than that of healthy subject. Inaddition, the inhibitory action of salivary fibronectin from adult periodontal patients toward P. gingivalis fimbria-inducedexpression of the neutrophil chemoattractant KC gene in mouse macrophages was clearly weak compared with that of thefibronectin from healthy subjects. These results suggest that fibronectin in saliva plays an important role as a regulator in thepathogenesis of P. gingivalis fimbriae in adult periodontal disease. z 1998 Federation of European Microbiological Soci-eties. Published by Elsevier Science B.V. All rights reserved.

Keywords: Porphyromonas gingivalis ; Fimbria; Salivary ¢bronectin; Adult periodontal disease; Macrophage; In£ammatory cytokine gene

1. Introduction

Porphyromonas gingivalis is the predominantpathogen of adult periodontal disease. P. gingivalis¢mbria is an important cell structure that is involvedin adherence of the bacterium to host cells. Our pre-vious studies [1^7] have demonstrated that the ¢m-briae induce expression of several in£ammatory cy-tokines in mouse macrophages and human gingival

¢broblasts and also that the ¢mbriae stimulate boneresorption via endogenous interleukin-1L in vitro.The ¢mbriae may function as pathogenic factorsthrough various biological actions of in£ammatorycytokines produced by the host cells stimulated bytheir adherence.

On the other hand, we [4,7] recently showed thatP. gingivalis ¢mbrillin, a major component of its¢mbriae, is one of the ¢bronectin-binding proteinsand that ¢bronectin is able to inhibit the ¢mbria-induced expression of in£ammatory cytokine genesin mouse macrophages as well as ¢mbria-stimulated

0928-8244 / 98 / $19.00 ß 1998 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.PII: S 0 9 2 8 - 8 2 4 4 ( 9 8 ) 0 0 0 9 8 - 4

FEMSIM 936 17-11-98

* Corresponding author. Tel. : +81 (492) 85-5511;Fax: +81 (492) 87-6657.

FEMS Immunology and Medical Microbiology 22 (1998) 257^262

bone resorption in vitro. These ¢ndings suggested tous that ¢bronectin plays a regulatory role in thepathogenic mechanism of the ¢mbriae in periodontaldisease. Although Amano et al. [8] indicated thatstathelin, a salivary component, binds to ¢mbriae,it is still not demonstrated whether the ¢bronectincontent in the saliva of healthy subjects and adultperiodontal patient di¡ers. It is also unknown if sali-vary ¢bronectin binds to the ¢mbria and is able toregulate the ¢mbria-induced expression of in£amma-tory cytokine genes in macrophages, therefore, weinvestigated these points in the present study.

2. Materials and methods

2.1. Subjects

Healthy subjects and subjects having adult perio-dontal disease (age range, 20 to 70 years) at MeikaiUniversity Hospital were classi¢ed into two groupsby the periodontal index (PI) described by Russell[9] : healthy (n = 48, PI = 1.0) and patient (n = 55,PI = 4.0 to 8.0). None of the subjects had receiveddental treatment or antibiotic therapy for at least 6months before this study.

2.2. Collection of saliva

Para¤n-stimulated whole saliva was collected andcentrifuged at 15 000Ug and then the supernatant asthe test sample was collected and stored at 320³Cuntil used.

2.3. Preparation of P. gingivalis ¢mbriae

P. gingivalis ATCC 33277 ¢mbriae were preparedand puri¢ed from cell washings by the method ofYoshimura et al. as described previously [10]. Theprotein content of the ¢mbriae was measured bythe method of Bradford [11].

2.4. Preparation of mouse peritoneal macrophages

Thioglycollate-stimulated peritoneal exudate cellsfrom 6 to 8 week old BALB/c mice were harvested,and the peritoneal macrophages were prepared asdescribed earlier [12].

2.5. Western blotting assay

The puri¢ed ¢mbriae (10 Wg protein per ml) weresubjected to sodium dodecyl sulfate polyacrylamidegel electrophoresis (SDS-PAGE) in a 12.5% poly-acrylamide gel. After SDS-PAGE, the separatedsamples were transferred to nitrocellulose paper bythe method described by Towbin et al. [13]. Aftertransfer, the nitrocellulose paper was treated for 2 hat 37³C with saliva from a healthy subject, washed,and then further incubated for 2 h with rabbit serum(control) or rabbit polyclonal antiserum (SigmaChem. Co., St. Louis, MO, USA) against human¢bronectin. The salivary ¢bronectin bound to the¢mbriae was visualized by use of horseradish perox-idase (HRP)-conjugated goat anti-rabbit IgG (Bio-Rad, Richmond, CA, USA), and HRP color devel-opment reagent (Bio-Rad).

2.6. Quanti¢cation of ¢bronectin in saliva

Fibronectin in saliva was measured by an enzymelinked immunosorbent assay (ELISA). In brief, sali-va samples (50 Wl) suspended in 500 Wl of sodiumcarbonate bu¡er were inoculated into a 96-welltype assay plate (NUNK, Kamstrup, Denmark).After incubation overnight at 4³C, the plate wastreated for 30 min with 3% (w/v) gelatin dissolvedin phosphate-bu¡ered saline (PBS) and washed withPBS containing 0.001% saponin (PBS-saponin).After washing, the plate was treated for 60 minwith anti-human ¢bronectin polyclonal antibodyand then washed three times with PBS-saponin.The treated plate was next incubated for 60 minalkaline phosphatase (ALP)-conjugated goat anti-rabbit IgG (ICN Pharmaceutical Inc., Aurora, OH,USA), washed, and then visualized by use of p-nitro-phenylphosphate (1 mg ml31) dissolved in 10% di-ethanol amine. Fibronectin contents were calculatedby use of puri¢ed human ¢bronectin (Sigma) as astandard.

2.7. cDNA hybridization probe

A plasmid containing mouse neutrophil chemoat-tractant KC cDNA sequence was provided by C.D.Stiles. Also, a plasmid with L-actin cDNA was ob-tained from the Japanese Cancer Research Bank

FEMSIM 936 17-11-98

Y. Murakami et al. / FEMS Immunology and Medical Microbiology 22 (1998) 257^262258

(JCRB, Tokyo, Japan). The methods used for plas-mid preparation were described previously [14].

2.8. Preparation of RNA and Northern blottinganalysis

The macrophage monolayers prepared frommouse peritoneal exudate cells (107 cells) weretreated or not with the ¢mbriae that has been pre-treated or not for 2 h at room temperature withsaliva sample. Thereafter, preparation of total cellu-lar RNA and Northern blot analysis was performedas described previously [15]. L-actin was used as aninternal standard for the quanti¢cation of totalmRNA on each lane of the gel.

2.9. Reagents

Human ¢bronectin and its polyclonal antibodiesraised in rabbits were purchased from SigmaChem. Co.; and ALP-conjugated goat anti-rabbitIgG, from ICN Pharmaceuticals Inc. p-Nitrophenylphosphate was obtained from Sigma Chem. Co.HRP-conjugated goat anti-rabbit IgG and HRP col-or development reagent were from Bio-Rad.

2.10. Data analysis

All experiments were repeated three times. Thesigni¢cance of di¡erences between the di¡erentgroups were determined by Student's t-test.

3. Results

3.1. Binding of ¢bronectin in saliva to P. gingivalis¢mbriae

Since our previous study [7] demonstrated that¢mbriae are able to bind to ¢bronectin with higha¤nity, we ¢rst examined by Western blot assaythe saliva of healthy subjects for the presence of¢bronectin. Fig. 1 shows that ¢bronectin, whichbound to the 43-kDa ¢mbrillin molecule, was presentin all of the saliva samples tested. These observationssuggested a possibility that ¢bronectin may play afunctional role in the regulation of ¢mbria-mediatedpathogenesis.

3.2. Fibronectin content in saliva is decreased in adultperiodontal patients

We next explored whether ¢bronectin content insaliva di¡ers between healthy subjects (n = 48) andperiodontal patients (n = 55). As shown in Fig. 2,the ELISA test showed that the ¢bronectin contentsin saliva of adult periodontal patients was signi¢-cantly lower than that of healthy (normal) subjects(P6 0.01).

FEMSIM 936 17-11-98

Fig. 1. Binding of ¢bronectin in saliva to P. gingivalis ¢mbrillin.

Fig. 2. Fibronectin content in saliva is decreased in adult perio-dontal patients.

Y. Murakami et al. / FEMS Immunology and Medical Microbiology 22 (1998) 257^262 259

3.3. Fibronectin in saliva inhibits the ¢mbria-inducedexpression of the KC gene in mouse macrophages

As we had previously demonstrated that the ¢m-briae strongly induce expression of the KC gene ofmouse peritoneal macrophages [2,6], we examined by

the Northern blot assay whether ¢bronectin in thesaliva is able to inhibit the ¢mbria-induced expres-sion of this gene in these cells. The cells were incu-bated for 1 h with ¢mbriae (1 Wg of protein ml31)that had been treated or not with healthy subject-derived saliva pretreated or not with anti-human ¢-bronectin antibody. Then total RNA was isolatedfrom the cells and KC gene expression in the cellswas analyzed by Northern blot assay. As shown inFig. 3, the ¢mbria-induced expression of the KCgene in the macrophages was markedly inhibited bypretreatment with saliva (30 Wg protein ml31 at ¢nalconcentration) from healthy subjects. This markedinhibitory action was eliminated by pretreatmentingthe saliva with anti-human ¢bronectin antibody at adilution of 1:400. These results strongly suggest aninhibitory action of ¢bronectin in saliva toward the¢mbria-induced expression of the KC gene in thecells.

In addition, we explored whether the inhibitory

FEMSIM 936 17-11-98

Fig. 3. Fibronectin in saliva inhibits P. gingivalis ¢mbria-inducedexpression of the KC gene in mouse peritoneal macrophages.

Fig. 4. Inhibitory e¡ect of saliva from healthy and adult periodontal subjects on P. gingivalis ¢mbria-induced expression of the KC genein mouse peritoneal macrophages.

Y. Murakami et al. / FEMS Immunology and Medical Microbiology 22 (1998) 257^262260

action of saliva from adult periodontal patients to-ward the ¢mbria-induced expression of the KC genewas lower than that from healthy subjects. The cellswere incubated for 1 h with ¢mbriae (1 Wg of proteinml31) that had been treated or not with saliva fromhealthy or adult periodontal subjects. Then totalRNA was isolated from the cells and KC gene ex-pression in the cells was analyzed by Northern blotassay. Fig. 4B and D were quanti¢cation of the KCgene expression in panel `A' and `C'. They respec-tively were done by densitometry, and are expressedas a % of the maximum.

Fig. 4A, B, C, D show that the inhibition withsaliva samples from several healthy subjects wasclearly stronger than that of the samples from perio-dontal patients. Taken together with the data inFigs. 3 and 4, these ¢ndings suggested that ¢bronec-tin in saliva may function as a potent regulator in thepathogenesis of P. gingivalis ¢mbriae in adult perio-dontal disease.

4. Discussion

We recently demonstrated that ¢bronectin func-tions as an inhibitor in the pathogenesis of P. gingi-valis infection via binding to the bacterial ¢mbriae[7]. Since ¢bronectin inhibited ¢mbria-stimulatedbone resorption in vitro [4], it was of interest todetermine if the ¢bronectin in saliva of adult perio-dontal patients in vivo plays a regulatory role in thepathogenesis of the ¢mbriae. We showed here thatthe ¢mbria-induced expression of the KC gene inmouse macrophages was markedly inhibited bytreatment with saliva of healthy subjects and thatthe inhibitory action in saliva from adult periodontalpatients was clearly lower. The inhibition was attrib-uted to salivary ¢bronectin, for antibody against thiscomponent blocked the inhibitory e¡ect. We alsohave observed in human monocytic THP-1 cells theinhibitory action of ¢bronectin in saliva on expres-sion of in£ammatory cytokine genes such as IL-1Land IL-6 (unpublished data). Therefore, the presentstudy suggests that ¢bronectin in saliva plays an im-portant protective role in preventing the initiationstage of pathogenesis of P. gingivalis, which involvesbinding of bacteria to host cells via the ¢mbriae.

Pierschbacher et al. [16] demonstrated that ¢bro-

nectin is fragmented into three domains, cell attach-ment- and gelatin- and heparin-binding domains, bychymotrypsin. Several studies have shown the pres-ence of fragmented domains of ¢bronectin in salivaand crevicular £uid [17,18], and we detected byWestern blot analysis the presence of two domain-bearing fragments in saliva of healthy subjects andadult periodontal patients (unpublished data). Asour previous studies showed that ¢bronectin bindsto P. gingivalis ¢mbriae via its cell attachment- andheparin-binding domains [2,4,6,7], these two do-mains in saliva may be involved in inhibition ofthe ¢mbria-induced expression of the KC gene inmacrophages observed in this study.

Salivary proline-rich protein (PRP) and statherinalso bind to P. gingivalis ¢mbriae [8,19,20], therefore,in further experiments, it will be important for us toexplore whether these proteins also are involved inregulation of the ¢mbria-induced expression of theKC gene.

In conclusion, we have demonstrated here that¢bronectin in saliva functions as a potent regulatorin one stage of the pathogenesis of P. gingivalis inadult periodontal disease by binding to the ¢mbriaeof the bacterium.

Acknowledgments

We thank C.D. Stiles for providing mouse KCcDNA probe. This work was supported by a grant-in-aid for scienti¢c research (No. 09771859) from theMinistry of Education, Science, and Culture of Ja-pan.

References

[1] Hanazawa, S., Murakami, Y., Hirose, K., Amano, S., Oh-mori, Y., Higuchi, H. and Kitano, S. (1991) Bacteroides (Por-phyromonas) gingivalis ¢mbriae activating mouse peritonealmacrophages and gene expression and production of interleu-kin-1. Infect. Immun. 59, 1972^1977.

[2] Hanazawa, S., Murakami, Y., Takeshita, A., Kitami, H.,Ohta, K., Amano, S. and Kitano, S. (1992) Porphyromonasgingivalis ¢mbriae induce expression of the neutrophil chemo-tactic factor KC gene of mouse peritoneal macrophages: Roleof protein kinase C. Infect. Immun. 60, 1544^1549.

[3] Kawata, Y., Hanazawa, S., Amano, S., Murakami, Y., Mat-sumoto, T., Nishida, K. and Kitano, S. (1994) Porphyromonas

FEMSIM 936 17-11-98

Y. Murakami et al. / FEMS Immunology and Medical Microbiology 22 (1998) 257^262 261

gingivalis ¢mbriae stimulate bone resorption in vitro. Infect.Immun. 62, 3012^3016.

[4] Kawata, Y., Iwasaka, H., Kitano, S. and Hanazawa, S. (1997)Porphyromonas gingivalis ¢mbria-stimulated bone resorptionis inhibited through binding of the ¢mbriae to ¢bronectin.Infect. Immun. 65, 815^817.

[5] Murakami, Y., Hanazawa, S., Nishida, K., Iwasaka, H. andKitano, S. (1993) N-acetyl-D-galactosamine inhibits TNF-Kgene expression induced in mouse peritonal macrophages by¢mbriae of Porphyromonas (Bacteroides) gingivalis, an oralanaerobe. Biochem. Biophys. Res. Commun. 192, 826^832.

[6] Murakami, Y., Hanazawa, S., Watanabe, A., Naganuma, K.,Iwasaka, H., Kawakami, K. and Kitano, S. (1994) Porphy-romonas gingivalis ¢mbriae induce a 68-kilodalton phos-phorylated protein in macrophages. Infect. Immun. 62,5242^5246.

[7] Murakami, Y., Iwahashi, H., Yasuda, H., Umemoto, T., Na-mikawa, I., Kitano, S. and Hanazawa, S. (1996) Porphyromo-nas gingivalis ¢mbrillin is one of the ¢bronectin-binding pro-teins. Infect. Immun. 64, 2571^2676.

[8] Amano, A., Sojar, H.T., Lee, J.Y., Sharma, A., Levine, M.J.and Genco, R.J. (1994) Salivary receptors for recombinant¢mbrillin of Porphyromonas gingivalis. Infect. Immun. 62,3372^3380.

[9] Russell, A.L. (1956) A system of classi¢cation and scoring forprevalence surveys of periodontal disease. J. Dent. Res. 35,350^359.

[10] Yoshimura, F., Takahashi, K., Nodasaka, Y. and Suzuki, T.(1984) Puri¢cation and characterization of a novel type of¢mbriae from the oral anaerobe Bacteroides gingivalis. J. Bac-teriol. 160, 949^954.

[11] Bradford, M.M. (1976) A rapid and sensitive method for thequantities of protein utilizing the principle of protein-dyebinding. Anal. Biochem. 72, 248^254.

[12] Hanazawa, S., Nakada, K., Ohmori, Y., Miyoshi, T., Amano,S. and Kitano, S. (1985) Functional role of interleukin-1 in

periodontal disease: induction of interleukin-1 production byBacteroides gingivalis lipopolysaccharide in peritoneal macro-phages from C3H/HeN and C3H/HeJ mice. Infect. Immun.50, 262^270.

[13] Towbin, H., Staehelin, T. and Gordon, J. (1979) Electropho-retic transfer of proteins from polyacrylamide gels to nitro-cellulose sheets: procedure and some applications. Proc. Natl.Acad. Sci. USA 76, 4350^4354.

[14] Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) MolecularCloning: A Laboratory Manual, pp. 365^389. Cold SpringHarbor Laboratory Press, Cold Spring Harbor, NY.

[15] Hanazawa, S., Takeshita, A., Amano, S., Semba, T., Nirazu-ka, T., Katoh, H. and Kitano, S. (1993) Tumor necrosis fac-tor-alpha induces expression of monocyte chemoattractant JEvia fos and jun genes in clonal osteoblastic MC3T3-E1 cells. J.Biol. Chem. 268, 9526^9532.

[16] Pierschbacher, M.D., Hayman, E.G. and Ruoslahti, E. (1981)Location of the cell-attachment site in ¢bronectin with mono-clonal antibodies and proteolytic fragments of the molecule.Cell 26, 259^267.

[17] Kanehisa, J., Doi, S., Yamanaka, T. and Takeuchi, H. (1991)Salivary ¢bronectin in man: An immunoblotting, radioimmu-noassay and immunohistochemical study. Arch. Oral Biol. 36,265^272.

[18] Pellat, B., Planchenault, T., Keil, D.V. and Pellerin, C. (1988)Fibronectin degrading activity in human crevicular £uid, gin-gival explants culture medium and bacterial plaques. J. Biol.Buccale 16, 51^57.

[19] Gibbons, R.J., Hay, D.I. and Schlesinger, D.H. (1991) Delin-eation of a segment of adsorbed salivary acidic proline-richproteins which promotes adhesion of Streptococcus gordonii toapatitic surfaces. Infect. Immun. 59, 2948^2954.

[20] Kataoka, K., Amano, A., Kuboniwa, M., Horie, H., Nagata,H. and Shizukuishi, S. (1997) Active sites of salivary proline-rich protein for binding to Porphyromonas gingivalis ¢mbriae.Infect. Immun. 65, 3159^3164.

FEMSIM 936 17-11-98

Y. Murakami et al. / FEMS Immunology and Medical Microbiology 22 (1998) 257^262262