INFECTION AND IMMUNITY, Aug. 1992, p. 3360-33680019-9567/92/083360-09$02.00/0Copyright © 1992, American Society for Microbiology

Treponema denticola Induces Actin Rearrangement andDetachment of Human Gingival Fibroblasts

PIERRE C. BAEHNI,l* MEJA SONG,2 CHRISTOPHER A. G. McCULLOCH,2AND RICHARD P. ELLEN2

School of Dental Medicine, University of Geneva, 19 Rue Barthellemy-Menn, 1211 Geneva 4, Switzerland,and Faculty of Dentistry, University of Toronto, Toronto, Ontario, Canada MSG 1G62

Received 26 December 1991/Accepted 7 May 1992

Spirochetes are associated with destructive periodontal diseases, and one cultivatable oral species,

Treponema denticola, binds to mammalian cells and perturbs metabolism. To evaluate the cytoskeletalresponses and attachment functions of human gingival fibroblasts (HGF) exposed to T. denticola, monolayersofHGF were incubated with T. denticola strains ATCC 35405, e, and e' in serum-free medium. HGF retractedpseudopods, rounded up, and ultimately detached from the substratum. Scanning electron microscopy showedspirochetes in close contact with HGF surfaces; occasionally, bacteria were partially submerged between foldsin the HGF membrane. Blebbing and numerous microvilli formed on the cell surface as the HGF retracted. Byconfocal microscopy, spirochetes were detected in contact with the HGF surface but were never found on theventral surface of fibroblasts between the substratum and cell. Morphological alterations were associated withand preceded by actin assembly, as measured by microscopic fluorimetry: there was a 263% increase in actinfluorescence over controls within 30 min. Detachment of fibroblasts from the substratum was related toincubation time and was dependent on the concentration of T. denticola. Detachment was observed for allstrains tested and was also dependent on the viability of T. denticola: UV light, heat, and metronidazoletreatment markedly reduced the HGF detachment response. Detachment was also significantly reduced by theprotease inhibitor phenylmethylsulfonyl fluoride. HGF viability was not significantly affected by coincubationwith spirochetes, as measured by lactate dehydrogenase release. Thus, T. denticola induces rapid cytoskeletalremodelling followed by cell detachment, which might be stimulated by a bacterially associated protease but isnot likely directly mediated by proteolytic degradation at the cell-substratum adhesive contact points.

Periodontal diseases are a group of inflammatory disorderswhich affect the tooth-supporting tissues and may lead totooth loss. There is strong evidence for a bacterial etiology(33): microbiological studies have shown quantitative andqualitative differences between the microbiota associatedwith healthy gingiva and those present in cases of periodon-titis (35). For example, spirochetes are sparse in healthysites, while in progressive periodontitis, they may compriseup to 30% of the flora (16). Furthermore, large-scale reduc-tions in the proportions and numbers of spirochetes occur

after successful treatment by mechanical debridement or byantibiotics (19, 21).Treponema denticola is a small spirochete which has been

cultivated from periodontal pockets. It binds to fibronectin(8), attaches to epithelial cells (25, 29) and fibroblasts (38), iscytotoxic for epithelial cells (2, 37), inhibits fibroblastic (4)and endothelial cell (34) proliferation, inhibits degranulation(3) and oxidative bursts of neutrophils, (17, 30) and sup-presses lymphocyte activation (32). In addition, T. denticolapossesses proteolytic (14, 24, 36) as well as fibrinolytic andcollagenolytic (20, 23) activities and produces a variety ofpotentially toxic metabolites (18). Taken together, these dataindicate that this spirochete is likely to be an importantperiodontal pathogen which has the potential to mediatetissue destruction by direct action against host cells andmatrix proteins.While much is already known at the molecular level about

perturbation of host cell function by overtly invasive bacte-ria causing enteric or urogenital infections, including their

* Corresponding author.

stimulation of cytoskeletal rearrangements (9, 10, 12), thereis very little basic information on cell surface interactions,and their biological consequences, when indigenous patho-gens contact host cells. The purpose of this investigation wasto determine whether contact between T. denticola andhuman gingival fibroblasts stimulates host cell responsesanalogous to those triggered by exogenous pathogens.

MATERIALS AND METHODS

Fibroblast cultures. Human gingival fibroblasts (HGF)were established from human tissue biopsy samples as

described previously (5). Cells were cultured in a minimalessential medium (a-MEM) containing 0.17% penicillin G,0.1% gentamicin, and 15% fetal bovine serum (Flow Labo-ratories, McLean, Va.) and maintained in a humidified, 5%CO2 atmosphere at 37°C. The medium was changed every 3to 4 days, and the cells were subcultured every week. Forexperiments, HGF were harvested by trypsinization andresuspended in a-MEM supplemented with serum. Cellswere plated at a concentration of 9 x 104 cells per well in24-well plates (Corning Glass Works, Corning, N.Y.) or 4.5x 105 cells per dish in tissue culture dishes (60 by 15 mm;Falcon, Becton Dickinson, Labware, Lincoln Park, N.J.)and incubated for 18 h prior to use. HGF were used betweenthe 8th and 27th passages.

Bacterial cultures. T. denticola ATCC 35405 was providedby W. J. Loesche, University of Michigan, and T. denticolastrains e and e' were provided by E. C. Chan, McGillUniversity. Stock cultures were maintained and grown in a

complex broth medium containing brain heart infusion,tryptic peptone, yeast extract, and volatile fatty acids and

3360

Vol. 60, No. 8

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF T. DENTICOLA ON GINGIVAL FIBROBLASTS 3361

supplemented with rabbit serum, as previously described(8). Cells were grown at 37°C for 3 to 4 days in an anaerobicchamber containing 80% nitrogen, 10% hydrogen, and 10%carbon dioxide (Coy, Ann Arbor, Mich.). Streptococcussanguis was used as a control species, as it is not considereda periodontal pathogen yet is common in dental plaque. S.sanguis 34 was originally obtained from R. J. Gibbons,Forsyth Dental Center, and maintained as lyophilized stocksin our laboratory. For experiments, S. sanguis was grownanaerobically in tryptic soy broth for 24 h. Cells wereharvested by centrifugation at 17,400 x g for 5 min andwashed two times in phosphate-buffered saline (PBS, pH7.4). Cells were resuspended in a-MEM without serum, andthe concentration was adjusted to an optical density at 550nm of 0.1 to 0.4 (approx. 1 x 109 to 4 x 109 bacteria per ml,standardized by microscopic count).For some experiments, freshly grown T. denticola organ-

isms were killed by heat treatment (60°C for 1 h) or UV lighttreatment (1 h in an ice bath). Killing was also performed byincubating 3-day T. denticola cultures with 1.0 mg of met-ronidazole (Novopharm Ltd., Scarborough, Canada) per mlfor 24 h in anaerobic conditions. Cells were then washed andprepared as described above. Bacterial death was deter-mined by propidium iodide staining (31).

Culture conditions and cell detachment assay. Monolayersof HGF in 24-well plates were washed two times in PBS, and0.5 ml of T. denticola suspension was added to each well.Controls consisted of HGF exposed to a-MEM withoutserum. Cultures were incubated for 1 to 17 h at 37°C. At theend of the incubation, the cultures were gently washed twicewith PBS to remove detached cells. The remaining adherentcells were harvested by trypsinization, and the number ofcells was determined with a Coulter counter (ZM model;Coulter Electronics Ltd., Luton, England). Counts of cellsin the medium of control cultures were obtained to computetotal cell number.To determine the effect of protease inhibitors, T. denticola

suspensions were exposed to phenylmethylsulfonyl fluoride(PMSF; Sigma Diagnostics, St. Louis, Mo.) at 170 ,ug/ml orN-a-p-tosyl-L-lysine chloromethyl ketone (TLCK; Sigma) at150 ,ug/ml overnight before use. HGF were then incubatedwith T. denticola in the presence of inhibitors at the sameconcentrations. Inhibition of proteolytic activities in T.denticola suspensions was monitored by using N-benzoyl-DL-arginine-2-naphthylamide (BANA; Perioscan ReagentCard; Oral-B, Redwood City, Calif.) and N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide (SAAPNA; Sigma) as substrates.PMSF inhibited chymotrypsinlike activity against SAAPNAbut not trypsinlike activity against BANA, whereas TLCKinhibited trypsinlike but not chymotrypsinlike activity. Todetermine the potential effect of cytoskeletal inhibitors,experiments were carried out in the presence of cytochalasinD (1 ,g/ml; Sigma) and phalloidin (10-' M; Sigma) in theincubation medium.

Viability of fibroblasts. Cell viability was determined bymeasuring lactate dehydrogenase (LDH) activity in theculture supematants with a commercially available enzymekit (Sigma). Results were expressed as a percentage of thetotal LDH released after the cultured cells were exposed to0.1% Triton X-100.

Morphological changes. The time course of morphologicalalterations of HGF cultures exposed to T. denticola wasmonitored by phase contrast microscopy with an invertedmicroscope (Diaphot-TMD; Nikon, Tokyo, Japan) equippedwith a constant-temperature incubator and a videotape re-corder. Videotapes were made to record the dynamics of

early events following exposure of HGF to T. denticola. Forscanning electron microscopy (SEM), experiments werecarried out with cells grown on glass coverslips placed in24-well plates. After incubation with T. denticola, cultureswere washed, fixed in 2.5% glutaraldehyde in PBS, washedin PBS, and dehydrated in a graded ethanol series. Sampleswere critical-point dried and coated with a 30-nm gold layer.Specimens were examined with a Hitachi S-2500 scanningelectron microscope.To determine whether spirochetes reached the space

between the substratum and fibroblast, cultures were pro-cessed for confocal microscopy. Following incubation withT. denticola, HGF monolayers were washed, fixed in 2.5%glutaraldehyde for 60 min, and stained by indirect immuno-fluorescence with anti-whole T. denticola antiserum (undi-luted) followed by fluorescein isothiocyanate (FITC)-conju-gated mouse anti-rabbit immunoglobulin antibody (1:50dilution) (Biocan, Mississauga, Canada). Propidium iodidewas used as a counterstain (31). Cultures were then imagedby confocal microscopy (Bio-Rad MRC-6000 series; OntarioLaser and Lightwave Research Center, University of Tor-onto), and 0.5-,um optical sections were obtained sequen-tially from the dorsal cell surface to the substratum.

F-actin labeling. HGF were grown in tissue culture dishesas described above. Cultures were exposed to T. denticolafor 2.5 to 60 min. At the end of incubation, cell cultures werewashed with warm PBS, fixed in 3.7% paraformaldehyde inPBS, and processed for filamentous (F-) actin labeling by theprocedure of Wymann et al. (40). Briefly, cultures werestained with FITC-phalloidin (Sigma) at 4 x 10-6 M in PBScontaining 0.1% Nonidet P-40 for 15 min. Cultures werewashed in PBS, mounted with a coverslip, and examinedwith a microscope spectrofluorimeter (MVP-SP; Leitz, Wet-zlar, Germany). F-actin labeling was measured by quantita-tive spectrofluorimetry and image analysis of FITC-phalloi-din-stained cells as described previously (26). The results ofF-actin labeling in cultures exposed to T. denticola wereexpressed as the percent increase in fluorescence above thatof control cultures not exposed to bacteria.

RESULTS

After exposure to T. denticola, HGF changed progres-sively from a flattened, stellate shape to a fusiform shapeover the course of 30 to 60 min. As seen under lightcinemicroscopy, pseudopods retracted, and the cellsrounded up and ultimately detached from the culture dishafter 1 to 2 h. Morphological examination by SEM showedthat HGF initially exhibited a smooth cell surface (Fig. 1A).Many HGF with bound T. denticola exhibited rapid cellsurface changes, including blebbing (Fig. 1B). As HGF cellsstarted to retract, long cytoplasmic processes remainedattached to the dish. Fragments of cell cytoplasm were oftenobserved in the vicinity of the trailing edge as remnants ofpseudopods (Fig. 1C). All rounded cells exhibited densemicrovilli, or ruffles, on their surfaces. The ruffles were oftenseen in association with spirochetes which were partlysubmerged in invaginations of the HGF membrane (Fig. 1D).By confocal microscopy, spirochetes were detected in con-tact with the dorsal HGF surface and in deeper opticalsections, suggesting their partial penetration of the cells.However, they were never found on the ventral surface offibroblasts, near focal contacts, between the substratum andcell (Fig. 2).

Control HGF cultures stained with FITC-phalloidin exhib-ited typical arrays of actin filaments, organized into stress

VOL. 60, 1992

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

3362 BAEHNI ET AL.

FIG. 1. SEM of HGF exposed to T. denticola. (A) Incubation time, 20 min. Spirochetes are in close contact with HGF (arrows).Magnification, x 5,000. (B and C) Initial morphological changes, characterized by alterations of the cell surface with blebbing and cytoplasmicprocesses as cells started to retract. Magnifications: B, x4,900; C, x2,500. (D) High magnification showing spirochete partially trapped bythe cytoplasmic membrane. Magnification, x 10,000. (E and F) Incubation time, 1 h. HGF exhibited numerous microvilli as cells rounded up.Magnifications: E, x5,000; F, x4,000.

INFECT. IMMUN.

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF T. DENTICOL4 ON GINGIVAL FIBROBLASTS 3363

FIG. 2. Laser scanning confocal micrographs showing the first (A), fourth (B), seventh (C), and tenth (D) optical transverse sections(nominally 0.5-p.m thick) of HGF exposed to T. denticola for 30 min, stained by indirect immunofluorescence and counterstained withpropidium iodide. Note the spirochetes contacting HGF (arrows). At the tenth optical section through the cell-substrate junctions, nospirochetes are observed. Magnification, x 1,000.

fibers (Fig. 3A). Only faint staining of cortical actin wasobserved in the rest of the cell. After addition of T. denticolato the cultures, there was a rapid rearrangement of actinwithin the first few minutes (Fig. 3B through F). Stress fibersdisappeared, cortical actin staining increased, and there wasbright, perinuclear staining in all cells undergoing rounding.Quantitation of F-actin by microscopic fluorimetry in timedexperiments confirmed the morphological observations: af-ter 2.5 min of incubation, there was a 166% increase inF-actin labeling compared with control cells (Fig. 4). Maxi-mum levels of F-actin were reached after 10 min, well beforesignificant HGF detachment, and tended to remain constantbetween 20 and 60 min.

All strains of T. denticola that were tested caused HGFdetachment (Fig. 5A). In some experiments, strains ATCC35405 and e' appeared to be slightly more potent than e inproducing this effect. Cell detachment was related to thetime of incubation and was dependent on the concentrationof spirochetes in suspension (Fig. 5B). Under similar exper-imental conditions, addition of S. sanguis to HGF culturescaused only 30% cell detachment, which was similar to thatcaused by a-MEM alone (data not shown). Bacterial culturesupernatants of T. denticola ATCC 35405 (200 ,ul of spiro-chete supernatant plus 300 ,ul of ot-MEM) did not affect HGFmorphology and did not produce more cell detachment than

the a-MEM control. SAAPNA- but no BANA-degradingactivity was detected in these bacteria-free culture superna-tants.HGF detachment caused by T. denticola was not directly

related to cell death, as indicated by LDH release. Measure-ments of LDH activity in the culture supernatant exposed tothe various strains showed from 4.5 to 6.1% of total possiblerelease by Triton X-100 after 6 h and 13.8 to 14.7% after 17h of incubation. In addition, when detached HGF werewashed, resuspended in treponeme-free a-MEM with se-rum, and plated, cells reattached to the culture dish andexhibited normal morphology.The effect of T. denticola on HGF was almost completely

abolished by killing the spirochetes with heat or UV light(Fig. 6). Metronidazole treatment of spirochetes resulted in amarked reduction but not complete inhibition of HGF de-tachment. Assessment of bacterial viability by propidiumiodide staining indicated that heat treatment killed 90 to100% of the spirochetes, whereas UV or metronidazoletreatment killed only 60 to 90%. Inclusion of PMSF mark-edly inhibited the effect, whereas TLCK had no inhibitoryeffect on the detachment of HGF (Fig. 6). HGF incubatedwith both 10' M phalloidin and T. denticola exhibited 90%detachment after 1 h, while incubation with spirochetesalone induced 60% detachment. Control cultures of HGF

VOL. 60, 1992

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

3364 BAEHNI ET AL.

"ItI Pl

FIG. 3. HGF exposed to T. denticola ATCC 35405 and stained with FITC-phalloidin. (A) Control cells in a-MEM. (B through F) Afterincubation for (B) 5 min, (C) 15 min, (D and E) 30 min, and (F) 60 min. Original magnification, x800.

INFECT. IMMUN.

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF T. DENTICOLA ON GINGIVAL FIBROBLASTS 3365

300

rz

rz

200

100

0

TIME (min)

FIG. 4. Increase in F-actin labeling in HGF cultures exposed toT. denticola ATCC 35405. Fluorescence was measured by quanti-tative spectrofluorimetry after FITC-phalloidin staining. Values are

the mean ± standard deviation of three separate experiments.

incubated with phalloidin demonstrated 45% detachmentafter 1 h. Inclusion of cytochalasin D at 1 p,g/ml in themedium did not inhibit cell detachment.

DISCUSSION

Binding. In this study, we have demonstrated and quanti-fied the cellular responses ofHGF exposed to washed wholecells of T. denticola. SEM demonstrated adhesion of T.denticola to HGF and partial penetration of some

treponemes into the cells. Previous morphological studies on

epithelial cell or fibroblast interactions with T. denticolahave yielded similar results (25, 29, 38). Confocal micros-copy showed that spirochetes were bound to the HGFmembrane on the dorsal surface but were absent from theventral surface of the cells, indicating preferential surface-dependent binding. The mechanisms of binding of spiro-chetes to mammalian cells are not well understood, but likesome other mucosal pathogens with Arg-Gly-Asp (RGD)peptide-binding functions, T. denticola strains with such a

phenotype might interact with integrin receptors on the HGFmembranes either directly or through bridging via fibronec-tin or other RGD-containing extracellular matrix proteins.Indeed, T. denticola has been shown to adhere specifically toimmobilized fibronectin and to RGD peptides and to exhibitaugmented adhesion to immobilized fibronectin in the pres-

ence of soluble fibronectin (8).Actin. Incubation of T. denticola cells with fibroblasts was

associated with rapid collapse of stress fibers and the reor-

ganization of filamentous actin into a brightly staining,perinuclear array. The fluorescence probe, FITC-phalloidin,does not quantitate the proportions of fluorescent species inseparate filamentous actin pools (6) and does not distinguishbetween short and long actin filaments. It is likely, however,that the observed rapid increases in fluorescence due toFITC-phalloidin staining reflect the reorganization of stress

fibers into much shorter filaments that move preferentially toperinuclear sites. Trypsin treatment of cultured eukaryoticcells also induces loss of stress fibers (28), although themechanisms underlying cell shape changes provoked byproteases are not well understood (1). It is conceivable thatthe rapid movement of actin to perinuclear sites mightprovide specific protection against nuclease activity elabo-rated by bacteria. Previous work by Lazarides and Lindberg(15) has identified actin as a specific inhibitor of DNase, andin this context it is noteworthy that T. denticola killed verylow proportions of fibroblasts, as detected by LDH releaseor propidium iodide exclusion. Thus, after bacterial chal-lenge, actin rearrangement of fibroblasts might represent ahost cell mechanism to protect the integrity of genomicDNA.The role of actin rearrangement in fibroblast detachment

cannot be inferred directly from the experimental designused here. However, several lines of evidence indicate thatactin rearrangements and cell detachment may be dissoci-ated events. Indeed, the fluorescence intensity of FITC-phalloidin binding to filamentous actin peaked at 10 min,almost an hour before significant numbers of cells began todetach. In addition, cytochalasin blockage of barbed fila-ment ends (7) to prevent growth of actin filaments had nodetectable effect on detachment. Lastly, stabilization ofexisting filamentous actin by addition of phalloidin wasassociated with increased detachment of fibroblasts, indicat-ing that the retention of a small proportion of actin filamentsdid not block the detachment process. Together, these datado not directly implicate actin depolymerization in themediation of cell detachment after exposure to T. denticolaand its proteases.Detachment. Contact of T. denticola with HGF ultimately

resulted in cell detachment from the substratum. Cell de-tachment may be caused, or in someway induced, by pro-teolytic cell-associated trypsinlike and chymotrypsinlike ac-tivities (14, 24, 36). The latter activity, localized on thebacterial surface (11), has the potential to bind and degradeextracellular fibronectin and perhaps other matrix proteins(11). Chymotrypsinlike activity may affect cell attachmentby disrupting cell-cell or cell-matrix adhesion (37). Extractsof another highly proteolytic periodontal pathogen, Porphy-romonas gingivalis, can bind and degrade the basementmembrane matrix (39) and affect the attachment of fibro-blasts in vitro (22, 27). Our finding that the protease inhibitorPMSF but not TLCK arrested the detachment of HGF mightbe interpreted to suggest that the morphological changeswere simply manifestations of direct chymotrypsinlike pro-teolysis at sites of cellular adhesion to the matrix. Inaddition, we found that bacterial culture supernatants whichwere trypsin negative and had little chymotrypsin activitydid not induce cell detachment. However, several otherobservations suggest that alternative mechanisms may beinvolved in this process.

(i) HGF detachment depended on the viability of T.denticola. Although killing by heat would be expected todenature and/or inactivate bacterial proteases, cell-associ-ated protease activity should not have been diminished byeither UV light or metronidazole treatment. Both of thesetreatments significantly reduced the ability of T. denticola toinduce HGF detachment, despite the fact that proteaseactivity was still detectable. These treatments would also beexpected to reduce T. denticola motility.

(ii) Fibroblast viability also appeared to be essential forcell detachment, as HGF killed by 70% ethanol treatment didnot detach after contact with T. denticola (data not shown).

VOL. 60, 1992

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

3366 BAEHNI ET AL.

1 2 3TIME (hours)

4 5 6

1007 B

80

q

a..C:

60

40

20- r/

r w W W ~ ~ ~ ~~~I* ****I0.01 0.1 1

O.D. 550

FIG. 5. (A) Cell detachment in response to various strains of T. denticola, expressed as the percentage of HGF detached with time.Strains: 0, ATCC 35405; *, e; A, e'; El, control cultures in a-MEM. (B) HGF detachment in response to various concentrations of T.denticola ATCC 35405. Cells were exposed to spirochetes for 2 h (O) or 4 h (A). Values are the means + standard deviations of quadruplicatecultures.

(iii) Confocal microscopy demonstrated that spirocheteswere absent at the interface between the cells and thesubstratum but were found on the dorsal surface of the HGF,suggesting that the cell-associated proteolytic activity of T.denticola (11) occurred at a site remote from the substratuminterface. In addition, bacterial culture supernatant had no

effect on HGF detachment above that found for controlcultures.

Pathogenic mechanisms. We have found that the effects ofT. denticola on HGF structure and adhesion are apparentlymore profound than its direct cytotoxic effects. Indeed, thecell detachment induced by T. denticola was not related toHGF death, as indicated by LDH release, propidium iodidestaining, and the observation that detached HGF, whenwashed and replated, did reattach to the culture dish. In thecontext of cell-pathogenic mechanisms, the hypothesis of a

possible protective role of actin against endogenous or

bacterial DNase is attractive and warrants further investiga-tion.

If such cytopathic effects were expressed in vivo, howmight cytoskeletal rearrangements or detachment fromextracellular matrix proteins disrupt the homeostasis ofperiodontal tissues? One way in which fibroblasts contributeto natural tissue remodelling is by collagen turnover throughphagocytosis (13). Considering the magnitude of the cyto-skeletal changes induced by spirochetes, HGF might be-come refractory to the signals required for efficient phago-cytosis. Since cell motility also depends on contact withextracellular matrix proteins and actin assembly, the repairor regenerative ability of infected tissues might be compro-mised. These effects would not necessarily depend on sig-nificant cell killing. Whatever its implications for pathogen-

A

p

100 -

80 -

60 -

a

40 -

20 -

0

0

INFECT. IMMUN.

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

EFFECTS OF T. DENTICOLA ON GINGIVAL FIBROBLASTS 3367

o0

a-I-

n

al C

O- I

TP..j* .ff.f|.e......

o

k~~~~~~~~~1

U cn

i.. ". S aX i > yffi~~~~:z

FIG. 6. HGF detachment in response to T. denticola ATCC35405. Incubation was for 1 h, and the spirochete concentration was2 x 109 cells per ml. The control was HGF in a-MEM. Untreat.,HGF exposed to fresh T. denticola; TCLK, 150 ,ug/ml in theincubation medium; PMSF, 170 ±g/ml; heat, 60°C for 1 h; UV light,1 h in an ice bath; Metron., metronidazole (1.0 mg/ml) for 24 h.Values are the means standard deviations of quadruplicatecultures.

esis, it is apparent that the adhesion of T. denticola to HGFis far from inconsequential at the cellular level.

ACKNOWLEDGMENTS

We are grateful to J. Beamish for helpful discussions. We thankGisele Knowles for technical assistance and Jacqueline Liberiouxand Margot Saunders for help in the preparation of the manuscript.

This work was supported by grants MT-5619 and MA-8903 fromMRC Canada and a grant from Vipont Pharmaceuticals, USA.

REFERENCES1. Aggeler, J. 1990. Cytoskeletal dynamics in rabbit synovial

fibroblasts. II. Reformation of stress fibers in cells rounded bytreatment with collagenase-inducing agents. Cell Motil. Cy-toskeleton 16:121-132.

2. Baehni, P. 1986. Interactions between plaque microorganismsand human oral epithelial cells, p. 143-153. In T. Lehner and G.Cimasoni (ed.), Borderland between caries and periodontaldisease, 3rd European symposium. Editions Medecine et Hy-giene, Geneva.

3. Boehringer, H. R., P. H. Berthold, and N. S. Taichman. 1986.Studies on the interaction of human neutrophils with plaquespirochetes. J. Periodontal Res. 21:195-209.

4. Boehringer, H. R., N. S. Taichman, and B. J. Shenker. 1984.Suppression of fibroblast proliferation by oral spirochetes. In-fect. Immun. 45:155-159.

5. Brunette, D. M., A. H. Melcher, and H. K. Moe. 1976. Cultureand origin of epithelium-like and fibroblast-like cells from por-cine periodontal ligament explants and cell suspensions. Arch.Oral Biol. 21:393-400.

6. Cassimeris, L., H. McNeill, and S. H. Zigmond. 1990. Chemoat-tractant-stimulated polymorphonuclear leukocytes contain twopopulations of actin filaments that differ in their spatial distri-butions and relative stabilities. J. Cell Biol. 110:1067-1075.

7. Cooper, J. A. 1987. Effects of cytochalasin and phalloidin onactin. J. Cell Biol. 105:1473-1478.

8. Dawson, J. R., and R. P. Ellen. 1990. Tip-oriented adherence ofTreponema denticola to fibronectin. Infect. Immun. 58:3924-3928.

9. Finlay, B. B. 1990. Cell adhesion and invasion mechanisms inmicrobial pathogenesis. Curr. Opin. Cell Biol. 2:815-820.

10. Finlay, B. B., S. Ruschkowski, and S. Dedhar. 1991. Cytoskel-etal rearrangements accompanying Salmonella entry into epi-thelial cells. J. Cell Sci. 99:283-296.

11. Grenier, D., V.-J. Uitto, and B. C. McBride. 1990. Cellularlocation of Treponema denticola chymotrypsinlike protease andimportance of the protease in migration through the basementmembrane. Infect. Immun. 58:347-351.

12. Isberg, R. R. 1991. Discrimination between intracellular uptakeand surface adhesion of bacterial pathogens. Science 252:934-938.

13. Knowles, G. C., M. McKeown, J. Sodek, and C. A. G. McCul-loch. 1991. Mechanism of collagen phagocytosis by humangingival fibroblasts: importance of collagen structure in cellrecognition and internalization. J. Cell Sci. 98:551-558.

14. Laughon, B. E., S. A. Syed, and W. J. Loesche. 1982. API ZYMsystem for identification of Bacteroides spp., Capnocytophagaspp., and spirochetes of oral origin. J. Clin. Microbiol. 15:97-102.

15. Lazarides, E., and U. Lindberg. 1974. Actin is the naturallyoccurring inhibitor of deoxyribonuclease I. Proc. Natl. Acad.Sci. USA 71:4742-4746.

16. Listgarten, M., and L. Heliden. 1978. Relative distribution ofbacteria at clinically healthy and periodontally diseased sites inhumans. J. Clin. Periodontol. 5:115-132.

17. Loesche, W. J. 1988. The role of spirochetes in periodontaldiseases. Adv. Dent. Res. 2:275-283.

18. Loesche, W. J., and B. E. Laughon. 1982. Role of spirochetes inperiodontal disease, p. 62-75. In R. J. Genco and S. E.Mergenhagen (ed.), Host-parasite interactions in periodontaldiseases. American Society for Microbiology, Washington,D.C.

19. Loesche, W. J., S. A. Syed, E. C. Morrison, G. A. Kerry, T.Higgins, and J. Stoll. 1984. Metronidazole in periodontitis. I.Clinical and bacteriological results after 15 to 30 weeks. J.Periodontol. 55:325-335.

20. Miakinen, K. K., S. A. Syed, P.-L. Makinen, and W. J. Loesche.1986. Benzoylarginine peptidase and iminopeptidase profiles ofTreponema denticola strains isolated from the human periodon-tal pocket. Curr. Microbiol. 14:85-89.

21. Mousques, T., M. A. Listgarten, and R. W. Phillips. 1980. Effectof scaling and root planing on the composition of the humansubgingival microbial flora. J. Periodontal Res. 15:144-151.

22. Nakamura, R., D. Hinode, H. Terai, and M. Morioka. 1991.Extracellular enzymes of Porphyromonas (Bacteroides) gingi-valis in relation to periodontal destruction, p. 129-141. In S.Hamada, S. C. Holt, and J. R. McGhee (ed.), Periodontaldisease: pathogens and host immune responses. QuintessencePublishing Co. Ltd., Tokyo.

23. Nitzan, D., J. F. Sperry, and T. D. Wilkins. 1978. Fibrinolyticactivity of oral anaerobic bacteria. Arch. Oral Biol. 23:465-470.

24. Ohta, K., K. K. Mikinen, and W. J. Loesche. 1986. Purificationand characterization of an enzyme produced by Treponemadenticola capable of hydrolyzing synthetic trypsin substrates.Infect. Immun. 53:213-220.

25. Olsen, I. 1984. Attachment of Treponema denticola to culturedhuman epithelial cells. Scand. J. Dent. Res. 92:55-63.

26. Pender, N., and C. A. G. McCulloch. 1991. Quantitation of actinpolymerization in two human fibroblast sub-types responding tomechanical stretching. J. Cell Sci. 100:187-192.

27. Phillips, J. R., H. S. Nadim, and D. L. Layman. 1990. Alter-ations in cell morphology and cytoskeletal proteins in gingivalfibroblasts exposed to a Bacteroides gingivalis extract. J. Peri-odontal Res. 25:339-346.

28. Pollack, R., and D. Rifkin. 1975. Actin-containing cables withinanchorage-dependent rat embryo cells are dissociated by plas-min and trypsin. Cell 6:495-506.

29. Reijntjens, F. M. J., F. H. M. Mikx, J. M. L. Wolters-Lutger-horst, and J. C. Maltha. 1986. Adherence of oral treponemesand their effect on morphological damage and detachment ofepithelial cells in vitro. Infect. Immun. 51:642-647.

30. Sela, M. N., A. Weinberg, R. Borinsky, S. C. Holt, and T.Dishon. 1988. Inhibition of superoxide production in humanpolymorphonuclear leukocytes by oral treponemal factors. In-

VOL. 60, 1992

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

INFECT. IMMUN.

fect. Immun. 56:589-594.31. Shapiro, H. M. 1988. Practical flow cytometry, 2nd ed., p. 168.

Alan R. Liss, Inc., New York.32. Shenker, B. J., M. A. Listgarten, and N. S. Taichman. 1984.

Suppression of human lymphocyte responses by oral spiro-chetes: a monocyte-dependent phenomenon. J. Immunol. 132:2039-2045.

33. Socransky, S. S. 1977. Microbiology of periodontal disease.Present status and future considerations. J. Periodontol. 48:497-504.

34. Taichman, N. S., J. E. Klass, B. J. Shenker, E. J. Macarak, H.Boehringer, and C.-C. Tsai. 1984. Suspected periodontopathicorganisms alter in vitro proliferation of endothelial cells. Pro-ceedings of the 6th International Conference on PeriodontalResearch, 1984. J. Periodontal Res. 19:583-586.

35. Tanner, A. 1991. Microbial succession in the development ofperiodontal disease, p. 13-25. In S. Hamada, S. C. Holt, andJ. R. McGhee (ed.), Periodontal disease: pathogens and host

immune responses. Quintessence Publishing Co. Ltd., Tokyo.36. Uitto, V.-J., D. Grenier, E. C. S. Chan, and B. C. McBride.

1988. Isolation of a chymotrypsinlike enzyme from Treponemadenticola. Infect. Immun. 56:2717-2722.

37. Uitto, V.-J., D. Grenier, and B. C. McBride. 1989. Effects ofTreponema denticola on periodontal epithelial cells (abstr. no.223). J. Dent. Res. 68:894.

38. Weinberg, A., and S. C. Holt. 1990. Interaction of Treponemadenticola TD-4, GM-1, and MS25 with human gingival fibro-blasts. Infect. Immun. 58:1720-1729.

39. Winkler, J. R., V. Matarese, C. I. Hoover, R. H. Kramer, andP. A. Murray. 1988. An in vitro model to study bacterialinvasion of periodontal tissues. J. Periodontol. 59:40-45.

40. Wymann, M. P., P. Kernen, T. Bengtsson, T. Andersson, M.Baggiolini, and D. A. Deranleau. 1990. Corresponding oscilla-tions in neutrophil shape and filamentous actin content. J. Biol.Chem. 265:619-622.

3368 BAEHNI ET AL.

on October 8, 2020 by guest

http://iai.asm.org/

Dow

nloaded from