.4rihs urn1 Bio[. i-01. 37. So. 5. pp. 597401. Pnn~cd:n Great Bnr~rn..+.I1nghrs ressned

1991 Copyright

0003-9969 92 S5.00 + 0.00 C 1991 Pergamon Press Lrd

SUSCEPTIBILITY OF PORPHYROMONAS GIZVGWKJS AND P. ASACCHAROLYTICA TO THE NON-OXIDATIVE KILLING MECHANISMS OF HUMAN NEUTROPHILS E. W. ODELL and P. J. WV Department of Oral Medicine and Pathology, United Medical and Dental Schools. Guy’s Hospital, London, U.K. (Accepted

8 April 1992)

Summav-Xeutrophils are essential for host defence against bacterial dental plaque and the pathogenic bacterial species within it. but in anaerobic environments such as the gingival crevice neutrophils can kill bacteria only with non-oxidative microbicidal compounds stored in their granules. Porphyromonas gingicalis W83, a pathogenic plaque species, and the avirulent non-oral type-strain P. osaccharol~fica were incubated anaerobically with intact neutrophils and with compounds extracted from normal human neutrophil granules. The killing of bacteria and the inactivation of lysozyme, cathepsin G. elastase, bacterialpermeability increasing factor and defensins by culture supematants were assayed. P. asaccharolytica but not P. gingicalis was killed under anaerobic conditions by intact neutrophils. P. gingkalis was also resistant to neutrophil granule compounds, its viability being reduced from a mean of 3.3 x IO6to 6.1 x 10pc.f.u/ml in 60 min by 400 pg,‘ml neutrophil granule extract, as compared to a reduction from 4.4 x IO6to 2.3 x IO’ c.f.u,ml for P. asaccharoi~rica. P. gingicalis culture supernatant inactivated cathepsin G, elastase, bacterial-permeability increasing factor and defensins. Resistance to neutrophil non-oxidative killing mechanisms may be an important virulence factor for P. gingicalis. Key words: Porphyronlonas

gingirnlis, neutrophils, bactericidal effect, non-oxidative killing, gingivain.

Several families of antimicrobial compounds are recognized: the defensins (Ganz er al., 1985), cathepsin G (Shafer er al., 199l)/cationic antimicrobial protein 37iazurocidin (Shafer, Martin and Spitznagel, 1984) and elastase (Janoff and Blondin, 1973), and the bactericidal permeability-increasing protein/cationic antimicrobial protein 57 family (Weiss et al., 1978). All these agents kill bacteria independently of any enzymic or bacteriolytic activity and each has a different range of activity (Elsbach and Weiss, 1988; Spitznagel, 1990). Many of these agents are highly target specific and by virtue of their protein nature might be susceptible to inactivation by proteolytic enzymes such as gingivain, a highly potent cysteine protease produced by P. gingivalis (Shah et al., 1991). Our aim now was therefore to determine the susceptibility of P. gingiualis to neutrophil non-oxidative killing mechanisms and compare its sensitivity to that of P. asaccharoLyrica, the type strain for the genus Porphyromonas, which is non-oral (Mayrand et al., 1984) and avirulent in periodontal disease.

INTRODUCTIOS Porph_vromonas gingitalis is a black-pigmented obligate anaerobe strongly associated with the progression of adult periodontal disease and juvenile periodontitis (Slots and Genco, 1984). Several putative virulence factors may be of importance in perio-

dontitis: these include polysaccharide capsule, lipopolysaccharide. fimbriae, epithelial and fibroblast cytotoxins, and several enzymes (Mayrand and Holt, 1988). many of which are secreted as structural components of membrane vesicles (Smalley and Birss, 1987; Shah et al.. 1991). Xeutrophils are the major host defence mechanism acting against dental plaque bacteria (Van Dyke, Le\-ine and Genco, 1982) and P. gingicalis is known to interfere with neutrophit function in several ways including inhibition of chemotaxis (Van Dyke er al., 1982a), resistance to (Schenkein, 1988) and degradation of (Grenier. Mayrand and McBride, 1989) opsonins, resistance to phagocytosis (Sundqvist er al., 1982) and by secretion of soluble toxins (Rotstein er al.. 1985). Less attention has been paid to the mechanisms by which neutrophils kill P. gingicalis. Under anaerobic or reducing conditions. such as occur in periodontal pockets and plaque, neutrophils kill bacteria by nonoxidative means, mediated by the release of an array of preformed compounds stored in primary granules.

Ahbreriarions:

L-BAPNA. benzoyl-L-arginine-p-nitroanilide; 3-PDS, 22’dipyridyl disulphide.

MATERIALS AND

METHODS

Bacterial strains and culture supernatants P. gingivalis strain W83 (formerly Bacteroides gingivaIis) and P. asaccharolytica provided by Dr

H. Shah (The London Hospital Medical College, London, U.K.), were maintained anaerobically on blood agar plates (Gaspak system, BBL Microbiology systems, Cockeysville, MD, U.S.A.). Acinetobacter calcoacericus ATCC 14987 was obtained from the

597

E. W. ODELLand P. J. Wu

598

American Type Culture Collection, Rockville, MD, U.S.A. and Escherichia coli strain C600 from the Department of Medical Microbiology. UMDS, Guy’s Hospital. London, U.K., and strains were confirmed by routine microbiological methods. Porph_womonas spp. maintained on 5% blood agar plates were inoculated into BM medium supplemented with haemin (10 pg,iml) and menadione (1 pgim!) (Shah er al., 1976) and cultured for 5 days at 37’C anaerobically (Gaspak system). Supernatants were centrifuged at 20,000 g for 30 min to remove bacteria, concentrated 20 times by ultrafiltration through a 10,000 mol. wt cut-off membrane (Amicon TM!0 membrane, Amicon Corp., Danver, MA, U.S.A.) and then maintained under anaerobic conditions at -70’C. Supernatants were free of bacteria as determined by culture but retained membrane vesicles. Gingivain activity was assayed kinetically by the hydrolysis of L-BAPNA (Shah et al., 1991) in a Perkin Elmer E.2 spectrophotometer. Gingivain was inactivated using specific inhibitor 2-PDS (Aldrich) (Shah er al., 1991), kindly provided by Dr H. Shah. Preparation extract

of

neutrophils

and neutrophil

granule

Neutrophils and the extract of their primary granules were prepared as before (Ode11 and Segal, 1991). In brief, normal human neutrophils were separated by dextran sedimentation and Ficoll-Hypaque centrifugation, disrupted by nitrogen cavitation and fractionated by sucrose density-gradient centrifugation. Microbicida! compounds were extracted from primary granules, dialysed against I54 mM NaC!, concentrated to 1 mgjm! protein (Pierce BCA protein assay system, Pierce, Rockford, IL, U.S.A.) retaining low molecular-weight components, and stored at -2O’C. Bacterial

killing by intact neutrophils

Bacterial killing by neutrophils was determined in a solid-phase assay. Five alternate layers of IO9c.f.u. of P. gingicalis or P. asaccharolytica, preopsonized in 50% pooled normal human serum for 15 min at 37’C, and IO8 neutrophils were centrifuged to the tip of a microfuge tube and incubated for 60 min at 37’C in Hanks balanced salt solution. The pellet was then lysed in water. pH 11 (Gargan, Brumfitt and HamiltonMiller, 1989), and bacterial viability assessed. The effect of anaerobiosis was determined in an anaerobic cabinet (oxygen concentration of five parts/lo6 or less). Bacterial

conditions. In some experiments gingivain in culture supernatants was inhibited by 2-PDS using the minimum amount required to abolish activity. titrated using the BAPNA assay. Samples were then assayed for cathepsin G, elastase, lysozyme, bacterial-permeability increasing factor and defensins. Lysozyme was assayed spectrophotometrically by lysis of Micrococcus Iysodeikticus cell walls according to Klass et al. (1977). Cathepsin G was assayed spectrophotometrically by the hydrolysis of .I--succinyl-Ala-AlaPro-Phe p-nitroanilide (Sigma, St Louis. MO, U.S.A.) accordine to De! Mar et al. (1979) and elastase by hydrolyss of h’-r-Boc Ala-Pro-NVA p-chlorothiobenzyl ester (S&ma) according,to Harper et al. (1984). Bacterial-permeability increasmg factor and defensin activities were quantitated by bioassay using the specific targets E. coli C600 (Weiss et al.. 1978) and A. calcoaceticus (Greenwald and Ganz, 1987). Bacterial

riability

Bacteria! viability was determined by passing samples through a 25-gauge needle to break up clumps, serial dilution in anaerobic 154 mM saline, followed by colonv counting on blood agar plates incubated anaerobically for 6 days. All counts were made in triplicate and results were calculated as the mean and SEM from at least three experiments. Statistical comparisons were made with the Student’s t-test. RESULTS In the presence of oxygen, 5 x 10” intact human neutrophils reduced the viability of both P. gingiralis (from mean 8.8 x 1Oa to 4.36 x I0”c.f.u.. p ~0.05) and P. asnccharoly~tica (from mean 4.39 x IO8 to 1.6 x IO8c.f.u., p < 0.05) but under anaerobic conditions, although P. asaccharolytica was killed (mean 7.2 x 10’ to 2.38 x 10sc.f.u., p

F

40

2

t 20

LYSOZYME

t

-1:o1:l EXTRACT

1:L : SUPERNATANT

1:lO

RATIO

Fig. 3. Inactivation of cathepsin G, elastase and lysozyme in neutrophil primary granule extract by P. gingiwlis W83 (a) and P. usnccharolytica 9337 (A) culture supernatants. The effect of inhibiting gingivain in W83 culture supernatants with 2-PDS is shown (A). Differences between species are significant at all ratios of supernatant to neutrophi1 primary granule extract (p < 0.05, n = 24, mean value from enzyme assays made in triplicate).

the chief host defence against periodontal disease (Van Dyke, Levine and Genco, 1982b). In oico, however, neutrophils must interact with P. gingidis at low oxygen concentration and in a heavily infected, reducing environment. Under such conditions the neutrophil’s oxidative killing ability is reduced (Loesche et al., 1988) because hypochlorous acid and hydrogen peroxide are not generated by the respiratory burst (Weiss et al., 1982), phagosomal pH control is disturbed (Segal et al., 1981), and oxidants are scavenged (Thomas et al., 1988). Some bacterial species are, however, killed equally well under such conditions as under aerobic conditions (Mandell, 1974; Vel er a!., 1984) because they are susceptible to neutrophils’ antimicrobial granule contents, which act independently of oxygen (Elsbach and

600

E. W.

BACTERICIDAL

J 0

PERMEA31_‘;”

100

INCREASING

Granule

extract

FACTOR

i

300

230

ODELL and

400

,ug/ml

DEFENS!NS

. lo3 c

\

-r,

extract

9337I .

I

200

c Granule

,

pg/ml

Fig. 4. Bioassay for bacterial-permeability increasing factor and defensins. Killing of indicator control species, respectively. A. cnlcoaceticus ATCC 14987 and E. cob C600, is proportional to neutrophil primary granule extract dose (0). The effect of pretreatment of extract with culture supernatant from P. gingivalis W83 (0) and P. asaccharolytica (A), and of W83 supernatant pretreated with 2-PDS (A) is significantly different from control viability at granule extract concentrations marked (*) (p < 0.05, n = 24, colony counts made in triplicate. mean i SD). Weiss, 1988; Spitznagel, 1990). Resistance to neutrophi1 antimicrobial proteins has been shown to correlate with surface structure, virulence and resistance to neutrophil killing of several non-oral microbial species (Rest, Cooney and Spitznagel 1977; Stinavage, Martin and Spitznagel, 1989; Ode11 and Segal, 1991). Oral bacteria, including some Gram-negative anaerobes, have been shown to be killed by non-oxidative mechanisms including lysozyme (Iacono et al., 1983), lactoferrin (Kalmar and Arnold, 1988) and defensins (Miyasaki ef al., 1990). We found that P. gingicalis resisted killing by whole human neutrophils in an assay which mimics the proposed role of neutrophils in the gingival crevice (Wilton, 1982) and which assesses killing independently of phagocytosis. Killing was more efficient, though still not marked, under aerobic conditions, possibly because in this type of assay the bacteria tend to reduce the redox potential of the pellet and inhibit oxidative mechanisms even though oxygen is

P. J. ‘&‘u

present in the supernatant. and also perhaps due to inefficient opsonization (Schenkein. 1988). P. asaccharo&ica. in contrast. is killed almost equally well under aerobic and anaerobic conditions. These findings demonstrate that under anaerobic conditions P. gingi&s is resistant to killing by neutrophils. This resistance may be explained by the relative immunity of P. gingilalis to killing by the neutrophils’ nonoxidative microbicidal components. as demonstrated by incubation in primary granule extract. P. nsaccharolytica, in contrast. is significantly more susceptible to killing by neutrophil gfanule lysate. although killing of both species requires considerably higher doses than other opportunistic pathogens such as E. coli, Cundida spp. or Strrpl~~lococc~ts aweus (Odell and Segal. 1991). The difference in susceptibility between P. ginghwlis and P. asucchnrol~tica to neutrophil granule extract may be explained in part by the inactivation of its constituent antimicrobial compounds because culture supernatants from either species reduced the activity of cathepsin G. elastase. bacterial-permeability increasing factor and defensins. although the reduction in lysozyme activity by P. ginghalti culture supematant was small. In particular, supematants from both species abolished bacterial-permeability increasing factor activity in granule extract while cathepsin G and defensins were markedly more susceptible to P. gingiralis supernatants. Both defensins and cathepsin G have recently been shown to kill Actinobacillus actit~onl~~ceremcor~lirans and Eikenella corrodens under condiiions similar to those in the gingival crevice and to those in our study (Miyasaki et al., 1990), and bacterial-permeability increasing factor is active against a wide range of Gram-negative bacteria (Weiss ez al., 1978). The demonstration that granule extract can kill P. gingiculis under anaerobic conditions need not suggest that defensins are less important than cathepsin G and bacterial-permeability increasing factor because, exceptionally, defensins can also kill another plaque species A. uctinornycetenzconzitans under anaerobic conditions (Miyasaki et al., 1990). Inhibition of neutrophil granule extract by P. gingirulis supernatant does not, however, appear to involve gingivain as the effects were not abrogated by 2-PDS. conditions under which gingivain was inactive, but further studies using purified enzyme lvill be required to investigate its role. It is concluded that both P. ginghulis and P. macchurolyticu are resistant to killing by neutrophils under anaerobic conditions, and that this resistance is reflected in their insusceptibility to antimicrobial compounds extracted from human neutrophil primary granules. Although culture supernatants from both species are able to inactivate individual neutrophil antimicrobial proteins, P. gingirulis supernatant was more effective. P. gingimlis demonstrates resistance to neutrophil-mediated killing under anaerobic conditions and this feature is likely to be an important virulence determinant in the gingival crevice. REFERESCES

Del Mar E. G., Largman C., Brodrick 1. W. ar:d Geokas M. C. (1979) A sensitive new substrate for chymotrypsin. Analyr.

Blochem.

99, 3 16-320.

Killing of P. gingicalis by neutrophils Elsbach P. and Weiss J. (1988) InJ?ammation: Basic Principles and Clinical Correlares (Eds Gallin J., Goldstein I. and Snyderman R.). Chap. 24, p. 445. Raven Press. New York. Ganz T., Se/stead M., Szlarek D., Harwig S., Daher K.. Bainton D. and Lehrer T. (1985) Defensins. Natural peptide antibiotics of human neutrophils. J. c/in. Inrest. 76, 1427-1435 Gargan R. A., Brumfitt W. and Hamilton-Miller J. M. T. (1989) Failure of water to lyse polymorphonuclear neutrophils completely. J. immunol. Merh. 124, 289-29 I. Greenwald G. 1. and Ganz T. (1987) Defensins mediate the microbicidal activity of human neutrophil granule extract against Acinetobacrer calcoacelicus. Infect. Immun. 55, 1365-1368.

Grenier D., Mayrand D. and McBride B. C. (1989) Further studies of the degradation of immunoglobulins by black-pigmented Bacteroides. Oral microbial. Immunol. 4, 12-18.

Harper J. W., Cook R. R. Roberts R., McLaughlin 8. J. and Powers J. C. (1984) Active site mapping of the serine proteases human leukocyte elastase, cathepsin G, porcine pancreatic elastase, rat mast cell proteases I and II, bovine chymotrypsin Ax and Staphylococcus aureus protease VS using tripeptide thiobenzyl ester substrates. Biochemisrq 23, 2995-3002.

Iacono V. J., Boldt P. R., MacKay B. J., Cho M. I. and Pollock J. J. (1983) Lytic susceptibility of Acrinobacillus actinomyceremcomitans Y4 to lysozyme. Infecr. Immun. 40, 773-784.

Janoff A. and Blondin J. (1973) The effect of human granulocyte elastase on bacterial suspensions. Lab. Incesl. 29, 454-457.

Kalmar J. R. and Arnold R. R. (1988) Killing of Actinobacillus actinomyceremcomitans by human lactoferrin. Infect. Immun. 56, 2552-2557.

Klass H. J., Hopkins J., Neale G. and Peters T. J. (1977) The estimation of serum lysozyme: a comparison of four assay methods. Biochem. Med. 18, 52-57. Loesche W., Robinson J., Flynn M., Hudson J. and Duclue R. (1988) Reduced oxidative function in gingival crevicular neutrophils in periodontal disease. Infect. Immun. 56, 156-160.

Mandell G. L. (1974) Bactericidal activity of aerobic and anaerobic polymorphonuclear neutrophils. Infect. Immun. 9, 337-341. Mayrand D. and Halt S. (1988) Biology of asaccharolytic black pigmented Bacreroides species. Mirrobio/. Rm. 52, 134-152. Mayrand D., Bourgeau G., Grenier D. and Lacroix J. M. (1984) Properties of oral black-pigmented asaccharolytic Bacteroides. Can. J. Microbial. 30, 1133-1136. Miyasaki K. T., Bodeau A. L., Ganz T., Selstead M. E. and Lehrer R. I. (1990) In cirro sensitivity of oral, Gram negative, facultative bacteria to the bactericidal activity of human neutrophil defensins. Infect. Immun. 58, 3934-3940.

Ode11 E. W. and Segal A. W. (1991) Killing of pathogens associated with chronic granulomatous disease by the nonoxidative microbicidal mechanisms of human neutrophils. J. med. Microbial. 34, 129-135.

Rest R. F., Cooney M. H. and Spitznagel J. K. (1977) Susceptibility of lipopolysaccharide mutants to the bactetitidal action of human neutrophil lysosomal fractions. Infecf. Immun. 16, 145-151.

Rotstein 0. D., Pruett T. L., Fiegel V. D., Nelson R. D. and Simmons R. L. (1985) Succinic acid, a metabolic by-product of bacteroides species, inhibits polymorphonuclear leukocyte function. Infect. Immun. 48, 402-408.

601

Schenkein H. A. (1988) Failure of Bacreroides gingiralis WS3 to accumulate bound C3 following opsonisation with serum. J. periodonr. Res. 24, 20-27. Segal A.. Geisow M.. Garcia R.. Harper A. and Miller R. (1981) The respiratory burst of phagocytic cells is associated with a rise in vacuolar pH. Nafure 290, 406-409. Shafer W.. Martin L. and Spitznagel J. (1984) Cationic antimicrobial proteins isolated from human neurophil granulocytes in the presence of di-isopropylfluorophosphate. Infect. Immun. 45, 29-35. Shafer W., Pohl J.. Onunka V.. Bangalore N. and Travis J. (1991) Human lysosomal cathepsin G and granzyme B share a functionally conserved broad spectrum antibacterial peptide. J. biol. Chem. 266, I iZ-116. Shah H.. Williams R., Bowden G. and Hardie J. (1976) Comparison of the biochemical properties of Bacreroides melaninogenicus from the human dental plaque and other sites. J. appl. Liacf. 41, 473-192. Shah H.. Gharbia S., Kowlessur D., Wilkie E. and Brocklehurst K. (1991) Gingivain; a cysteine proteinase isolated from Porphyromonas gingiralis. Microb. ecol. health Dir. 4, 319-328.

Slots J. and Genco R. (1984) Microbial pathogenicity: black pigmented Bacteroides species, Cupnocvrophaga species and Acrinobacillus actinon~~cetemcomians in human periodontal disease, virulence factors in colonisation, survival and tissue destruction. J. denl. Res. 63, 412-421. Smalley J. and Birss A. (1987) Trypsin-like enzyme activity of the extracellular membrane vesicles of Bacreroides gingiraiis WSO. J. gen. Microbial. 133, 2883-2894. Spitznagel J. (1990) Antibiotic proteins of human neutrophils. J. c/in. Incesi. 86, 1381-1386. Stinavage P., Martin L. E. and Spitznagel J. K. (1989) 0 antigen and lipid A phosphoryl groups in resistance of Salmonella fyphimurium LT-2 to non-oxidative killing in human polymorphonuclear neutrophils. I&t. Immun. 57, 3894-3900.

Sundqvist G., Bloom G. D., Enberg K. and Johansson F. (1982) Phagocytosis of Bacleroides meluninogenicus and B. gingicalis in tirro by human neutrophils. J. periodont. Res. 17, 113-121.

Thomas E.. Learn D., Jefferson M. and Weatherred W. (1988) Superoxide-dependent oxidation of extracellular reducing agents by isolated neutrophils. J. biol. Gem. 263, 2178-2186. Van Dyke T. E., Bartholomew E., Genco R. J., Slots J. and Levine M. J. (1982a) Inhibition of neutrophil chemotaxis by soluble bacterial products. J. Periodont. 53, 502-508. Van Dyke T., Levine M. and Genco R. J. (1982b) Host Parasite Interactions in Periodontal Diseases (Eds Genco R. J. and Mergenhagen S.). p. 235. American Society of Microbiology, Washington, DC. Vel W. A., Namavar F., Verweij M. J., Pubben A. N. and McLaren D. M. (1984) Killing capacity of human polymorphonuclear leucocytes in aerobic and anaerobic conditions. J. med. Microbial. 18, 173-180. Weiss S., Elsbach P., Olsson I. and Odeberg H. (1978) Purification and characterisation of a potent and bactericidal and membrane-active protein from the granules of human polymorphonuclear leukocytes. J. biof. Chem. 253, 2664-2672.

Weiss S. J., Klein R., Slivka A. and Wei M. (1982) Chlorination of taurine by human neutrophils. Evidence for hypochlorous acid generation. J. c/in. Incesr. 70, 598-602.

Wilton J. (1982) The role of the polymorphonuclear leukocyte in the control of subgingival plaque formation. J. periodonf. Res. 17, 506-508.