INFECTION AND IMMUNITY, Nov. 1983, p. 659-6670019-9567183/110659-09$02.00/0Copyright 0 1983, American Society for Microbiology

Vol. 42, No. 2

Relationship Between Oxygen Tension and SubgingivalBacterial Flora in Untreated Human Periodontal Pockets

W. J. LOESCHE,1* F. GUSBERTI,2 G. METTRAUX,2 T. HIGGINS,2 AND S. SYED2University ofMichigan School of Dentistry2 and Department of Microbiology, University of Michigan School

of Medicine,1 Ann Arbor, Michigan 48109

Received 18 May 1983/Accepted 16 August 1983

The predominance of anaerobic bacteria in subgingival plaque samples suggeststhat the pocket environment is anaerobic. In the present investigation, a smalloxygen tension (PO2) electrode was inserted into the base of the pocket and theP02 was recorded. In addition, the plaque in these pockets was examinedculturally and microscopically. The oxygen tension at the bottom of 36 pockets (5to 10 mm in depth) ranged from S to 27 mmHg (1 mmHg = 133.3 Pa) with a meanvalue of 13.3 mmHg. Moderate pockets (5 and 6 mm) exhibited a mean PO2 of 15.7mmHg, which was significantly higher than the 12.0 mmHg found in the deeperpockets. The deep pockets had higher percentages of spirochetes and Bacteroidesintermedius, whereas the moderate pockets had elevated proportions of Actino-myces naeslundii and Streptococcus mutans. The sites with oxygen tensionsequal to or less than 15 mmHg had significantly higher percentages of spirochetes,whereas the microaerophilic Capnocytophaga species were found in pockets witha P02 greater than 15 mmHg. The presence of bleeding in the pocket wasassociated with higher proportions of B. intermedius, Capnocytophaga sp., andA. naeslundii. These P02 readings of periodontal pockets indicated that there is aspectrum Of PO2 values which seem to define, in a general way, the microbiologi-cal composition of the pocket.

Bacteria which colonize the root surfaces ofthe teeth constitute the primary etiological com-ponent for destructive periodontitis (15, 19).There is evidence that the flora associated withactive periodontitis is characterized by a distinc-tive flora which if not uniquely involved with theadjacent tissue inflammation is at least diagnos-tic of that inflammation (16, 20, 29, 32, 33). Lightand electron microscopic examinations of sub-gingival plaque taken from relatively healthysites or from diseased sites routinely show highproportions of cocci in the healthy sites and highproportions of spirochetes in the diseased sites(15, 16).

Culture studies indicate that the majority ofthese subgingival organisms are anaerobes (1,6). These anaerobic bacteria, however, differwith regard to their oxygen sensitivity (18, 27).The small cultivable oral spirochete Treponemadenticola was classified as a strict anaerobebecause it only initiated growth when there wasless than 0.5% oxygen in the cultural atmo-sphere, whereas the pigmented and nonpigment-ed Bacteroides species were classified as moder-ate anaerobes because they initiated growth inthe presence of oxygen levels as high as 8% (18).The predominance of spirochetes, including

T. denticola, in subgingival plaque samples (22)suggested that the pocket environment is anaer-obic. This supposition was supported by theobservations that oxygen comprised about 1% ofthe gas atmosphere in the vicinity of the dento-gingival surfaces (4, 5) and that the redox potenwtial (Eh) of supragingival plaque dropped fromabout +200 mV to as low as -140 mV as theplaque aged (7).There is no information on oxygen tension

(PO2) in the periodontal pockets, or on how thisrelates to the pocket microbiota. A preliminarystudy (26) described a method to measure theP02 in human periodontal pockets by using aP02 electrode that was small enough to beinserted to the base of the pocket. The purposeof the present study was to relate the PO2 inuntreated periodonal pockets to the composi-tion of the subgingival bacterial flora.

MATERIALS AND METHODSTen medically healthy patients who showed clinical

and radiological evidence of advanced periodontitisparticipated in this study. No periodontal treatmenthad been performed in any of these patients in the last5 years, nor did they give a history of antibiotic usagein the preceding 6-month period. Thirty-six pockets

659

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

660 LOESCHE ET AL.

(three or four per patient) ranging from 5 to 10 mm indepth were selected for the PO2 measurements and thebacterial studies.P02 measurement. A flexible intravascular com-

bined P02 electrode (Roche, Bioelectronics Kontron,ZH, Switzerland), which was modified so as to havean active tip length of 2.5 mm and a cross section of 0.6by 1.3 mm (26), was inserted into the base of eachpocket. Because of these probe dimensions, it was notpossible to measure the PO2 in pockets <4 mm.The oxygen electrode contains a thin polyethylene

sensor tube which acts as a hydrophobic gas-perme-able membrane. This tube is filled with an electrolytesolution and contains two silver wires which representthe cathode and the anode. The cathode is polarizedwith -800 mV relative to the anode (2, 3). Any gasescapable of diffusing through the hydrophobic mem-brane and of being reduced at about 800 mV willgenerate an electric current. As oxygen is the mostlikely gas with these characteristics found in biologicalsystems, the electrolysis current is directly propor-tional to the rate of oxygen diffusion through themembrane and serves as a continuous indicator of thepartial pressure of oxygen (PO2) in the sample. Thereis no detectable interference with this measurement byCO2 and pH in physiological ranges (3).The electrode was designed to measure PO2 in a

large homogeneous liquid volume such as the blood-stream, where the oxygen consumption by the redoxprocess is negligible. However, in a small heteroge-neous system such as the periodontal pocket, wherethe plaque is bathed by gingival crevicular fluid orblood or both, the oxygen consumption by the elec-trode could result in a decrease of the actual PO2 if theoxygen supply is not maintained. The measured PO2values reached their lowest levels within 5 to 10 minand then remained stable. This suggests that oxygenconsumption by the electrode was not of any conse-quence, as sufficient oxygen diffused from the tissuesso as to maintain a steady PO2 within the pocket.The sterile, disposable electrodes were calibrated

before every measurement to the actual PO2 in theatmosphere, i.e., 21% of the atmospheric pressureminus the partial pressure of water vapor (47 mmHg at37°C; 1 mmHg = 133.3 Pa). Since the calibrationtemperature differed from the temperature in the peri-odontal pockets, which is about 34 to 36°C (28), anadjustment on the monitor to a temperature of 35°Cwas performed.The oxygen sensors were positioned carefully in the

pockets until resistance from the attachment tissueswas felt. Millimeter marks on the electrode indicatedthe actual depth of the sensor in the pocket. Thepatients were asked to close their lips during themeasurements so as to keep the intraoral temperatureand oxygen tension constant. The PO2 readings de-clined in the first few minutes in situ, reaching aplateau within 5 to 10 min after electrode insertion.The PO2 values on this plateau were then recorded.

Clinical parameters. The depths of the selectedpockets were subsequently measured to the nearestmillimeter with a graduated periodontal probe. Thedifference between the pocket depth and the sensordepth averaged about 0.6 mm. The gingival inflamma-tion about these teeth was estimated on a scale of 0 to3 according to the gingival index (17).

Bacteriological samples. The supragingival plaque

was removed with a curette, and then the tooth surfacewas cleaned with a sterile gauze before the insertion ofthe probe. The probe was left in place for the PO2readings and then removed. The adherent plaque onthe tip of the instrument was immediately transferredto a vial containing 0.5 ml of reduced transport fluidwithout EDTA (35). If more plaque was required, itwas taken by means of a curette. The plaque sampleswere placed immediately into an anaerobic chamber(1), and after dispersing for 20 s with a Vortex mixer, aSO-j.l portion of each sample was removed for micro-scopic examination. This degree of dispersal separatedmost of the plaque aggregates without lysing thespirochetes. The remaining sample was diluted to 4 ml,sonically dispersed for 20 s with a Kontes sonifier(Kontes Glass Company, Vineland, N.J.), seriallydiluted in reduced transport fluid, and plated automati-cally with a spiral plater (Spiral System, Inc., Cincin-nati, Ohio) on a variety of media (23).The total anaerobic count, the counts of black-

pigmented Bacteroides species, Capnocytophaga spe-cies, Fusobacterium nucleatum, and the red-brownpigmented colonies of Actinomyces odontolyticuswere obtained from growth on enriched tryptic soyagar (37). Either all or representative colonies ofblack-pigmented Bacteroides species were subcul-tured and differentiated on the basis of their ability touse glucose, to hydrolyze esculin, and to produceindole (24).Actinomyces viscosus and Actinomyces naeslundii

were differentiated and enumerated on a cadmiumsulfate-metronidazole selective medium (11). Strepto-coccus sanguis and Streptococcus mutans were identi-fied by their colony morphology on MM10 sucroseagar (23). Veillonella colonies were enumerated on amedium containing the MM10 base minus the sucroseand blood but supplemented with lactate, vancomycin(30), and a 0.004% bromocresol purple indicator. Thetotal count offacultative organisms was obtained fromgrowth on enriched tryptic soy agar containing 20 ,ug ofmetronidazole per ml, which was incubated anaerobi-cally (24).Ten microliters of the portion removed for the

microscopic examination was applied to a microscopicslide and covered with a cover slip. The slide wasexamined by dark-field microscopy at a magnificationof x 1,200. Generally, 200 bacteria or 20 fields selectedat random, whichever occurred first, were counted.The bacteria were classified into the following mor-phological categories: spirochetes (large, intermedi-ate, small), fusiforms, motile organisms, rods, andcocci.

Statistical analysis. The bacteria profiles of plaquesat different PO2 values or at different pocket depthswere statistically analyzed by using the Student t-testas well as the nonparametric Mann-Whitney U andmedian tests (31). An analysis of variance with theScheffe test was used for comparisons betweengroups.

RESULTS

The PO2 at the bottom of the untreated pock-ets ranged from 5 to 27 mmHg, with a meanvalue of 13.3 mmHg. The 95% confidence limitfor the mean P02 was between 12.3 and 14.2

INFECT. IMMUN.

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

OXYGEN TENSION AND SUBGINGIVAL BACTERIAL FLORA 661

TABLE 1. Proportions of bacteria in moderate and deep periodontal pockets"Proportion (mean % ± SE) of bacteria in:b

Organism Moderate Deep (7-10 mm) Significance'(5-6 mm)

pces( 0pockets (n = 16) pockets (n = 20)

CultivableBlack-pigmented Bacteroides 7.1 ± 2.1 9.5 ± 1.5B. intermedius 1.4 ± 0.6 4.3 ± 1.6B. gingivalis 3.6 ± 1.8 3.5 ± 1.5B. melaninogenicus 2.1 ± 1.5 1.7 ± 1.5

Veillonella sp. 2.3 ± 1.2 1.3 ± 0.3A. viscosus 2.3 ± 0.9 3.2 ± 1.2A. odontolyticus 1.0 ± 0.5 0.9 ± 0.3A. naeslundii 2.6 ± 1.4 0.1 ± 0.1 P= 0.005F. nucleatum 7.1 ± 1.9 7.5 ± 2.1Selenomonas sp. 3.2 ± 1.5 3.0 ± 0.8Capnocytophaga sp. 7.0 ± 1.8 6.7 ± 1.5Facultative organisms 29.7 ± 5.3 27.6 ± 3.6S. mutans 3.3 ± 1.3 0.8 ± 0.5 P = 0.03S. sanguis 5.0 ± 3.1 1.3 ± 0.5

Microscopic countsSpirochetes (total) 29.6 ± 4.4 40.7 ± 3.7 P = 0.07

Large 4.3 ± 1.5 7.6 ± 1.5Intermediate 5.1 ± 1.2 8.1 ± 1.6Small 20.2 ± 3.4 25.3 ± 2.0

Motile organisms 4.9 ± 1.1 4.7 ± 0.9Fusiforms 7.6 ± 1.2 5.5 ± 1.1Rods 30.6 ± 3.2 24.4 ± 3.2 P = 0.04Cocci 20.1 ± 4.3 15.8 ± 2.0a Total cultivable bacteria counts (CFU x 105 per plaque ± standard error): moderate pockets, 59.7 ± 13.2;

deep pockets, 133.8 ± 2.0. Significantly different at P = 0.02.b PO2 values: moderate pockets, 15.7 mmHg; deep pockets, 12.0 mmHg.c Determined with the Mann-Whitney U and median tests.

mmHg. The pockets were separated into moder-ate (5 and 6 mm) and deep (7 to 10 mm) pockets,and their P02 values were compared. The mod-erate pockets exhibited a mean P02 of 15.7mmHg, which was significantly higher than the12.0 mmHg found in the deeper pockets. Thisgrouping of moderate and deep pockets wasused to analyze the bacteriological data, primari-ly because of this significant difference in P02and also because of the clinical relevance ofpocket depth in assessing the effect of periodon-tal therapy (10).The proportions of the monitored bacterial

species and morphotypes found in 16 moderatepockets and 20 deep pockets are compared inTable 1. The deep pockets, due to their largervolume, had about twice the number of viableorganisms as did the moderate pockets (Table1). In absolute numbers there were significantlymore bacteria, spirochetes, Bacteroides inter-medius, F. nucleatum, and Capnocytophaga sp.in the deeper pockets (data not shown).The counts were normalized according to the

percentage of viable bacteria to determinewhether any of the monitored organisms exhibit-ed a preference for the deeper pockets. The

percentages of total spirochetes and B. interme-dius were higher in the deep pockets, but notsignificantly so (Table 1). A. naeslundii, S. mu-tans, and the proportions of rods observed mi-croscopically were significantly elevated in themoderate pockets (Table 1). These findings sug-gest that the deeper pockets, which had anaverage P02 of 12.0 mmHg, tended to select forcertain anaerobic species, whereas the moder-ately deep pockets, which had a PO2 of 15.7mmHg, selected for certain microaerophilic spe-cies.The data were then analyzed according to the

P02 found in the pockets. The viable countswere comparable in both oxygen groups (Table2). The sites with P02 equal to or less than 15mmHg had significantly higher percentages ofspirochetes than were found in the sites withP02 greater than 15 mmHg (Table 2). Althoughall sizes of spirochetes exhibited this distributionpattern, only the intermediate-size spirocheteswere significantly elevated in the more anaero-bic pockets. The microaerophilic Capnocyto-phaga species were significantly elevated in thepockets with the higher P02 (Table 2).The distribution of the monitored organisms

VC)L. 42, 1983

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

TABLE 2. Proportions of bacteria in periodontal pockets with P02 higher or lower than 15 mmHga

Organism Proportion (mean ± SE) of bacteria in pockets with pO,: Significance'-15 mmHg (n = 25) >15 mmHg (n = 11)

CultivableBlack-pigmented Bacteroides 9.2 ± 2.2 6.4 ± 3.1

B. intermedius 3.9 ± 1.3 0.8 ± 0.4B. gingivalis 3.8 ± 1.4 3.1 ± 1.8B. melaninogenicus 1.5 ± 1.3 2.6 ± 2.2

Veillonella sp. 2.3 ± 0.7 0.5 ± 0.2A. viscosus 3.5 ± 1.1 1.3 ± 0.5A. odontolyticus 1.1 ± 0.4 0.7 ± 0.5A. naeslundii 0.6 ± 0.4 2.8 ± 2.2F. nucleatum 5.9 ± 1.2 10.6 ± 3.9Selenomonas sp. 3.9 ± 1.1 1.1 ± 0.5Capnocytophaga sp. 5.5 ± 0.9 9.9 ± 3.1 P = 0.05Facultative organisms 30.7 ± 3.9 22.8 ± 5.2S. mutans 2.6 ± 0.9 0.1 ± 0.02S. sanguis 1.9 ± 0.7 4.9 ± 4.2

Microscopic countscSpirochetes (total) 38.5 ± 3.0 29.7 ± 6.5 P = 0.01Large 6.2 ± 1.3 4.7 ± 2.2Intermediate 7.9 ± 1.2 4.1 ± 1.8 P = 0.01Small 23.9 ± 1.9 20.9 ± 4.7

Motile organisms 4.8 ± 0.9 4.7 ± 1.3Fusiforms 6.1 ± 0.8 7.1 ± 1.8Rods 27.7 ± 2.8 26.0 ± 3.1Cocci 15.4 ± 1.9 23.1 ± 5.6a Total cultivable bacteria counts (CFU x 105 per plaque ± standard error): P02 < 15 mmHg, 103.7 ± 20.0;

PO2 > 15 mmHg, 90.3 ± 35.1. Not significantly different.b Determined with the Mann-Whitney U and median tests.C Total microscopic count per 20 high-power fields: P02 < 15 mmHg, 200.9 ± 13.2; P02 > 15 mmHg, 158.0 ±

27.1.

as a function of both pocket depth and P02 isshown in Table 3. The proportions of strictlyanaerobic spirochete species tended to be ele-vated in deep pockets, with a predilection shownfor sites with low P02. In the moderate pockets,the spirochetes also preferred the low P02. Theintermediate size spirochetes, in particular,were significantly reduced in the moderate pock-ets when the P02 was above 15 mmHg (Table 3).Among the moderate anaerobes, B. interme-

dius was significantly elevated in those deeppockets that had low P02 (Table 3). Bacteroidesgingivalis was found in both moderate and deeppockets but tended to be reduced in those deeppockets with the higher P02. F. nucleatum ap-peared to tolerate higher PO2, as their propor-tions were significantly elevated in those deeppockets with a P02 greater than 15 mmHg.Among the microaerophilic species, A. naes-

lundii was significantly elevated in the moderatepockets that exhibited higher PO2. The propor-tions of Capnocytophaga sp. in both moderateand deep pockets were elevated in those siteswith higher P02 (Table 3). The other microaero-philic organism monitored, A. viscosus, behavedin an anomalous manner in that its proportions

tended to be higher in the low P02 sites, regard-less of pocket depth (Table 3).

Facultative species such as S. mutans and S.sanguis were present in higher proportions inthe moderate pockets, with S. mutans beingsignificantly elevated in those moderate pocketswhich had the lowest P02 (Table 3).

Several of the indigenous subgingival plaqueorganisms are known to derive nutritional bene-fit from host products (21). As these nutrientsmight be elevated in the presence of gingivalbleeding, an effort was made to determine whateffect bleeding, combined with PO2, could haveupon the proportional distribution of these or-ganisms in the periodontal pockets.The presence or absence of gingival bleeding

did not affect the proportional distribution of thetotal spirochetes (Table 4). However, in the highP02 sites, the proportions of large spirochetes inthe two pockets that exhibited no bleeding weresignificantly elevated relative to the bleedingsites (Table 4). The intermediate-size spiro-chetes were elevated in the nonbleeding sitescompared to bleeding sites at both high and lowoxygen tensions.

B. intermedius was significantly elevated in

INFECT. IMMUN.662 LOESCHE ET AL.

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

OXYGEN TENSION AND SUBGINGIVAL BACTERIAL FLORA 663

TABLE 3. Proportions of bacteriaa in moderate versus deep pockets as a function of pO2aProportion (mean % ± SE) found in:

Moderate pockets Deep pocketsIn vitro 02 Species PO2 ' 15 pO, > 15 pO -

15 pO, > 15sensitivity mmHg mmHg mmHg mmHg

(n =9) (n 7) (n =16) (n =4)

Strict Spirochetes, total 32.3 ± 4.7 26.1 ± 7.9 41.9 ± 3.7b 35.9 ± 12.3anaerobes Spirochetes, large 4.1 ± 1.0 4.6 ± 3.1 7.5 ± 1.8 5.0 + 2.6

Spirochetes, 7.6 + 1.8 1.8 ± 0.8' 8.2 ± 1.7 7.8 ± 4.5intermediateSpirochetes, small 20.6 ± 3.2 19.7 + 6.9 25.8 ± 2.3 23.1 ± 5.5

Moderate Black-pigmented 5.8 ± 2.9 8.9 ± 4.7 11.4 ± 3.4 2.2 ± 1.1anaerobes Bacteroides

B. intermedius 2.1 ± 0.9 0.5 ± 0.3 5.1 ± 1.9 1.2 ± 1.0B. gingivalis 3.2 ± 2.6 4.2 ± 2.9 4.2 ± 1.8 0.9 ± 0.9B. melaninogenicus 0.5 ± 0.2 4.1 ± 3.5 2.1 ± 1.8 0.02 + 0.01

F. nucleatum 6.4 ± 2.4 8.2 ± 3.5 5.7 ± 1.3 15.0 + 9.5A. odontolyticus 1.0 ± 0.5 1.0 ± 0.9 1.1 ± 0.4 0.3 ± 0.3Veillonella sp. 3.3 ± 1.8 0.7 ± 0.4 1.6 ± 0.5 0.07 ± 0.03

Microaero- Capnocytophaga sp. 5.6 + 2.0 9.0 ± 3.5 5.5 ± 1.1 11.5 ± 6.4philic A. naeslundii 1.3 ± 0.7 4.4 ± 3.3 0.1 ± 0.1 0.06 + 0.04

A. viscosus 2.9 + 1.5 1.5 + 0.7 3.8 ± 1.4 0.8 ± 0.4Facultative S. mutans 5.8 ± 1.7 0.07 ± 0.03 1.0 ± 0.6 0.02 ± 0.01organisms S. sanguis 2.8 ± 0.7 8.0 ± 6.5 1.6 ± 0.7 0.3 ± 0.2a Total viable counts (CFU x 106 per plaque sample ± standard error): moderate, PO2 < 15 mmHg, 5.9 ± 1.8;

moderate, PO2 > 15 mmHg, 6.0 ± 2.1; deep, PO2 < 15 mmHg, 13.1 ± 2.3; deep, P02 > 15 mmHg, 14.3 ± 9.2.b Values that are connected by arrows are significantly different; Scheffe test, P values indicated over arrows.c Values in boldface are significantly different from other values in row. Scheffe test, P < 0.05.

those low PO2 sites where bleeding occurred(Table 4). This suggested that some nutrient(s)was derived from the bleeding, but that thisbenefit could best be realized at low P02. F.nucleatum was more prominent in the absenceof bleeding, especially in the high P02 pockets.The Capnocytophaga species were signifi-

cantly elevated only in those bleeding sites thatexhibited the higher P02 (Table 4). This suggest-ed that higher P02 were more important thanbleeding for their ascendency in a site. The samewas true for A. naeslundii, as the proportions ofthis organism were significantly higher in thosebleeding sites associated with the higher P02. S-mutans was significantly elevated in bleedingsites with low PO2, whereas S. sanguis tended tobe higher in those bleeding sites with the higherP02 (Table 4).The interactions of pocket depth and bleeding

upon the bacterial flora were also examined.Only two pockets were both deep and withoutbleeding, and these pockets had significantlyelevated proportions of intermediate size spiro-chetes and F. nucleatum relative to the otherpocket configurations (Table 5). Pocket depthappeared to be a more important determinantthan bleeding in regard to proportions of spiro-chetes, as more spirochetes were present in thedeep pockets than in the moderate pockets (Ta-ble 5). Bleeding in the moderate pockets was

associated with significantly lower proportionsof total spirochetes compared to bleeding sites inthe deep pockets. Bleeding also seemed to dis-criminate against the intermediate-size spiro-chetes and F. nucleatum. Bleeding, however,was associated with increased proportions of B.intermedius in the deep pockets and significantincreases of A. naeslundii, S. mutans, and S.sanguis in the moderate pockets (Table 5).

DISCUSSION

The interpretation of the value of a particularredox couple, i.e., 1/2 02-H20, in a complexsystem such as plaque bathed by gingival crevi-cular fluid or blood or both and approximated bythe gingival equilibrium would normally be ten-tative. In this study the characteristics of theelectrode were such that only gases capable ofdiffusing through the hydrophobic polyethylenemembrane would reach the electrode and thenonly gases electrolytically reduced at 800 mVcould generate a current (3). This electrodedesign would favor the detection of oxygen evenin a complex environment such as the periodon-tal pocket. The fact that steady-state readingscould be obtained is indirect evidence that a

single redox couple was being monitored.The actual value of the PO2 is pH and tem-

perature dependent. We were unable to monitor

VOL. 42, 1983

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

TABLE 4. Proportions of bacteria in pockets with P02 higher than 15 mmHg versus lower than 15mmHg asa function of the gingival indexa

Proportion (mean % ± SE) found with:

In vitro 02 Speciesb P02 ' 15 mmHg PO2 > 15 mmHgsenstivity No bleeding Bleeding No bleeding Bleeding

(n =6) (n =19) (n =2) (n =9)

Strict anaerobes Spirochetes, total 36.8 ± 6.3 39.0 ± 3.5 32.2 ± 2.0 29.1 ± 8.0Spirochetes, large 3.8 ± 1.2 7.1 ± 1.6 13.5 + 8.9*c 2.8 ± 1.3*Spirochetes, inter- 10.2 ± 3.1* 7.2 ± 1.2 5.3 ± 2.3 3.7 ± 2.2*

mediate 22.7 ± 4.2 24.3 ± 2.2 13.6 ± 4.6 22.6 + 5.5Spirochetes, small

Moderate anaer- Black-pigmented Bac- 7.1 ± 3.8 10.0 ± 2.7 0.6 ± 0.4 7.7 ± 3.6obes teroides 0.9 ± 0.4 5.0 ± 1.7* 0.5 ± 0.5 0.9 ± 0.5*

B. intermedius 5.9 ± 3.9 3.1 ± 1.4 0.03 ± 0.02 3.7 + 2.2B. gingivalis 0.4 ± 0.2 1.9 ± 1.6 0.03 ± 0.02 3.2 ± 2.7B. melaninogenicus 7.9 ± 3.6 5.3 ± 1.1 33.5 ± 8.3 5.6 ± 2.1

F. nucleatum 1.7 ± 1.1 2.6 ± 1.1 0.03 ± 0.01 0.6 ± 0.4Veillonella sp. 0.3 ± 0.1 1.4 ± 0.4 0.6 ± 0.5 0.8 + 0.7A. odontolyticus

Microaerophilic Capnocytophaga sp. 7.7 ± 2.6 4.7 ± 1.0* 8.5 ± 2.8 10.3 ± 3.7*A. naeslundii 1.0 ± 0.6 0.4 ± 0.3 0.03 ± 0.01 3.4 ± 2.6A. viscosus 3.6 ± 2.1 3.4 ± 1.2 3.1 + 2.0 0.9 ± 0.4

Facultative or- S. mutans 1.7 ± 1.3 2.8 ± 1.0 0.1 ± 0.1 0.03 ± 0.02ganisms S. sanguis 1.4 ± 0.4 2.1 ± 0.9 0.03 ± 0.01 6.1 ± 5.2a Gingival index: 0 and 1, no bleeding; 2 and 3, bleeding.b Total viable counts (CFU x 106 per plaque sample ± standard error): P02 s 15 mmHg, no bleeding, 6.5 ±

2.5; P02 < 15 mmHg, bleeding, 11.8 ± 4.2; P02 > 15 mmHg, no bleeding, 2.2 ± 0.6; P02 > 15 mmHg, bleeding,10.6 ± 4.2.

c, Values significantly different from each other in the same row so indicated. Scheffe test, P < 0.05.d Values in boldface are significantly different by the Scheffe test from all other values in the same row; P <

0.05.

the actual pH and temperature in the monitoredpockets and based our calculations on a pH of 7and a temperature of 35°C in the pockets. ThepH of resting subgingival plaque, such as thatpresent in this investigation, is about 7.0 (9).However, some subgingival plaque organismsare saccharolytic, i.e., Capnocytophaga spe-cies, Bacteroides melaninogenicus, S. mutans,and S. sanguis, and it is possible that if sufficientglucose was present in the gingival crevicularfluid the pH in the plaques in which theseorganisms dominated would be lower. In thiscase, the observed P02 values would be higherthan those reported here. Recent reports indi-cate that the temperatures of pockets '5 mmrange from 34 to 360C (28). For purposes of thepresent study, the temperature of the pocketwas assumed to be 35°C. With these reserva-tions in mind, the results were interpreted asfollows.The P02 at the base of untreated periodontal

pockets is about 13.3 mmHg or 1.8% oxygen.This is low compared with atmospheric P02 (155mmHg), arterial blood P02 (95 mmHg), andvenous blood P02 (20 to 40 mmHg) (8). Howev-er, the P02 in the buccal vestibule, which mightreflect the PO2 at the orifice of the pocket, islower, i.e., about 5 mmHg (4, 5). Thus, theslightly higher P02 found in these untreatedpockets might reflect oxygen which diffuses into

the pocket from the adjacent epithelial surfaces.The tissue P02 ranged from 15 to 40 mmHg (2),which could maintain such a diffusion gradient.Also, bleeding into the pocket could elevate thepocket P02 levels. Some support for this was theobservation that sites which exhibited spontane-ous bleeding (gingival index = 3) had significant-ly higher PO2 readings than did sites with lowergingivitis scores (G. Mettraux, F. A. Gusberti,and H. Graf, manuscript in preparation).The pockets studied in this investigation were

divided into those with relatively high P02 andthose with relatively low PO2, using 15 mmHg asthe partition value. This value seemed to definean important environmental parameter with re-gard to the distribution pattern of several of themonitored species. Thus, the spirochetes, whichwere presumed to contain strict anaerobesamong their numbers, were significantly elevat-ed in the pockets with low PO2, and Capnocyto-phaga species, which are known microaerophils(14), were significantly elevated in the high PO2pockets (Table 2). The intermediate-size spiro-chetes, in particular, were significantly elevatedin the low P02 sites.Pocket depth is the clinical parameter used to

assess periodontal morbidity and treatment effi-cacy. Deep pockets were found to have both alower P02 relative to moderate pockets and tocontain higher proportions of spirochetes (Table

664 LOESCHE ET AL. INFECT. IMMUN.

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

OXYGEN TENSION AND SUBGINGIVAL BACTERIAL FLORA 665

TABLE 5. Proportions of bacteria in moderate versus deep pockets as a function of the gingival index

Proportion (mean % ± SE) found in:'In vitro 02 Speciesb Moderate pockets Deep pocketssensitivity No bleeding Bleeding No bleeding Bleeding

(n =7) (n 10) (n =2) (n 18)

Strict anaer- Spirochetes, total 33.9 ± 5.6 27.0 ± 6.0* 40.8 ± 10.5 40.7 ± 4.0*obes Spirochetes, large 6.1 ± 3.3 3.2 ± 1.4 6.6 ± 7.0 7.1 ± 1.6

Spirochetes, inter- 6.6 ± 2.3 4.1 ± 1.6 - 16.1 ± 8.5 # 7.2 ± 1.5mediate

Spirochetes, small 21.2 ± 4.7 19.6 ± 4.8 18.1 ± 0.1 26.1 ± 2.2Moderate an- Black-pigmented Bac- 4.7 ± 0.4 8.7 ± 3.2 9.0 ± 8.9 9.6 ± 3.0

aerobes teroidesB. intermedius 0.8 ± 0.4 1.9 ± 1.0 0.9 ± 0.9 4.7 ± 1.7B. gingivalis 3.6 + 3.6 3.6 ± 2.1 7.8 ± 7.8 3.1 ± 1.4B. melaninogenicus 0.3 + 0.2 3.2 ± 2.4 0.3 ± 0.2 1.9 ± 0.6

F. nucleatum 9.8 ± 4.2 5.2 ± 1.6 - 26.8 ± 14.8 *-- 5.4 ± 1.2Veillonella sp. 1.3 ± 1.1 3.2 ± 1.7 1.4 ± 1.4 1.3 ± 0.4A. odontolyticus 0.3 ± 0.2 1.5 ± 0.7 0.7 ± 0.4 1.0 ± 0.3

Microaerophilic Capnocytophaga sp. 7.7 ± 7.6 6.5 ± 2.7 8.8 ± 2.5 6.5 ± 1.7A. viscosus 4.1 ± 7.1 1.1 ± 0.4 1.3 ± 0.2 3.4 ± 1.3A. naeslundii 1.0 ± 0.6 3.7 ± 7.4* 0.02 ± 0.02 0.1 ± 0.1*

Facultative or- S. mutans 1.8 ± 1.3 4.2 ± 1.9* 0.03 ± 0.01 0.8 ± 0.5*ganisms S. sanguis 0.8 ± 0.4 7.3 ± 4.6* 1.3 ± 1.2 1.3 ± 0.6*

a Gingival index: 0 and 1, no bleeding; 2 and 3, bleeding.b Total viable counts (CFU x 106 per plaque sample ± standard error): moderate pockets, no bleeding, 4.9 ±

2.3; moderate pockets, bleeding, 6.7 ± 1.5; deep pockets, no bleeding, 7.4 ± 5.9; deep pockets, bleeding, 14.4 ±3.1.

c *, Values significantly different from each other in the same row so indicated; Scheffe test, P < 0.05. Valuesconnected by arrows are significantly different from each other Scheffe test; P < 0.05.

1). The microaerophilic A. naeslundii and S.mutans were significantly elevated in the moder-ate pockets compared to the deep pockets (Ta-ble 1). Thus, the oxygen sensitivities of theseplaque organisms coincided with their propor-tional distribution in pockets of various P02 anddepths.Although spirochetes tended to be more nu-

merous in pockets with the lower PO2. they werenevertheless found in plaques over the P02range of 5 to 27 mmHg. This suggested thatsome or all of these plaque spirochetes weremore oxygen tolerant than previously suspect-ed. Previously, the small spirochete T. denticolawas found to be unable to initiate growth on agarsurfaces when the atmospheric PO2 was greaterthan 4 mmHg (18). Either the spirochetes ob-served at the higher PO2 were not as oxygensensitive as T. denticola, or in vivo, low P02 arenot the sole determinants of the ability of a strictanaerobe to survive in the periodontal pocket.Presumably some protective factor(s) exists

which prevents toxic accumulations of oxidessuch as hydrogen peroxide or the formation ofsuperoxide anion. Both facultative bacteria andaerobic host cells contain catalases, peroxi-dases, and superoxide dismutases, which couldinactivate these oxides in the pocket and permitgrowth of strict anaerobes. Thus, it is to beexpected that strict anaerobes are less sensitive

to oxygen in vivo than in vitro; as in the lattercase, the antioxidative contributions of host andfacultative bacterial cells are missing. Thiswould explain the observation that presumablystrict anaerobes such as the spirochetes could beisolated from pockets with oxygen levels as highas 3.6%.

Therefore, in a microbial community as com-plex as subgingival plaque, a single environmen-tal determinant such as P02 will not exclusivelydefine the character of the flora. Other phenom-ena, which would include the host immunologi-cal response, the volume and composition of thegingival crevicular fluid, as well as bleeding intothe pocket, could play a role. Bleeding shouldboth increase the oxygen levels in the pocketdue to the dissolved oxygen in the blood, as wellas provide essential nutrients such as hemin,vitamin K, and steroids in the case of certainblack-pigmented Bacteroides species (13, 25),alpha-2-globulin in the case of certain smallspirochetes (34), and other products in the caseof A. naeslundii and Capnocytophaga (F. A.Gusberti, S. A. Syed, G. Bacon, N. Grossman,and W. J. Loesche, J. Periodontol., in press).Because of these nutritional considerations, cer-tain anaerobes could be selected for in pocketsites where bleeding occurred, despite the pres-ence of higher P02 levels.The interactions of PO2, bleeding, and pocket

VOL. 42, 1983

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

TABLE 6. Relative effect of environmental parameters on proportions of selected members of thesubgingival flora'

PO, Bleeding PocketOrganism

Low High Yes No Moderate Deep

Intermediate spirochete ++ (-) ++ +B. intermedius ++ ++ +F. nucleatum + + +Capnocytophaga sp. ++ +A. naeslundii ++ ++ ++S. mutans ++ + ++

a Low P02, <15 mmHg; high P02, >15 mmHg. Moderate pockets, 5 to 6 mm; deep pockets, .7 mm.+,positive tendency; ++, positive association; (-), negative association.

depth upon certain of the monitored organismsare summarized in Table 6. The intermediate-size spirochetes preferred low P02 and siteswithout bleeding. These conditions tended to bemet when the pockets were deep. This suggeststhat these spirochetes were not dependent uponblood-borne products to the extent that otherorganisms such as B. intermedius and A. naes-lundii were.

B. intermedius was significantly elevated inbleeding sites which had low P02 compared toeither nonbleeding sites with a low PO2 or bleed-ings sites with a higher P02 (Tables 4 and 6).This indicated that this moderate anaerobe de-rived a real benefit from blood products, butonly at low P02 tensions. Some strains of thisspecies require hemin, vitamin K (25), and ste-roids such as progesterone and estradiol (13) foroptimal growth in vitro. These or other yet-to-be-identified nutrients in the blood could ac-count for the ascendancy of B. intermedius inthese sites in a manner analogous to their selec-tion in plaques associated with pregnancy gingi-vitis (12).

F. nucleatum was particularly high in the twoplaque samples taken from nonbleeding, highP02 sites. Otherwise it tended to prefer non-bleeding sites. The Capnocytophaga speciespreferred the high P02 sites and seemed tobenefit from bleeding. A. naeslundii also exhibit-ed higher proportions in those sites with bleed-ing and a high P02. Thus, these microaerophilicspecies appeared to derive some selective ad-vantage from both high P02 levels and bleeding.In regard to the latter, both Capnocytophagaspecies and A. naeslundii increase in plaquesamples at the time of puberty and coincidentwith gingivitis, suggesting that some host-de-rived product was instrumental in their selection(Gusberti et al., in press).The facultative S. mutans and S. sanguis

species were confined to the moderate pocketsand tended to be elevated in the presence ofbleeding. S. mutans was elevated in sites withlow PO2, whereas S. sanguis preferred sites with

a higher P02 (Table 4). The presence of S.mutans in subgingival plaque is potentially omi-nous, as this organism is significantly associatedwith root surface caries (36).These P02 readings in periodontal pockets

indicated that there is a spectrum of P02 valueswhich seems to define, in a general way, themicrobiological composition of the pocket.Thus, in pockets where the P02 values are low,all sizes of spirochetes and B. intermedius in-crease proportionally.Spontaneous bleeding tends to raise the P02

values in pockets. However, rather than inhibit-ing the anaerobic species, there is a balancewhere the added nutrients provided by the bloodor the antioxidants contained therein promotethe growth of anaerobic species. This was ob-served in the cases of B. intermedius, the largespirochetes, and facultative S. mutans. Bleedingalso promotes the growth of microaerophilicspecies such as A. naeslundii and Capnocyto-phaga.The discussion has focused primarily on the

influence that P02 derived from the host canhave on shaping the microbiota of the pocket. Inreality the microbiota contribute significantly tothe environment by producing or excreting prod-ucts which affect the physical-chemical charac-teristics of the niche. For instance, microbeswould lower the P02 of the environment viaenzymes such as oxidases and superoxide dis-mutase and increase the P02 by producing prod-ucts such as antigens, endotoxins, organic acids,peptides, and sulfides, which act on the host toincrease bleeding. Thus the P02 detected by theoxygen electrode is the result of many complexinteractions, and the low values observed shouldbe interpreted as confirming only that the pocketenvironment is anaerobic, a fact that has beenwell demonstrated by the anaerobic culturingrequirements of the resident flora (6). However,the P02 of the pocket is not as low as would havebeen surmised, given the low P02 that is neces-sary for optimal bacteria recoveries in vitro.This would indicate that factors other than P02

INFECT. IMMUN.666 LOESCHE ET AL.

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

OXYGEN TENSION AND SUBGINGIVAL BACTERIAL FLORA 667

determine the ability of strict anaerobes to growin the periodontal pocket.

ACKNOWLEDGMENTS

This work was supported by Public Health Service grantsDE02731 and DE06030 from the National Institute of DentalResearch. F.G. was a recipient of a Swiss National Fellow-ship.The technical assistance of Cathy Haushalter and Janice

Stoll is gratefully acknowledged.

LITERATURE CITED1. Aranki, A., S. A. Syed, E. B. Kenney, and R. Freter. 1969.

Isolation of anaerobic bacteria from human gingiva andmouse cecum by means of a simplified glove box proce-dure. Appl. Microbiol. 17:568-576.

2. Clark, L. C., Jr. 1956. Monitor and control of blood andtissue oxygen tensions. Trans. Am. Soc. Artif. Intern.Organs 2:41-46.

3. Eberhard, P., W. Fehlmann, and W. Mindt. 1979. Anelectrochemical sensor for continuous intravascular oxy-gen monitoring. Biotelem. Patient Monit. 6:16-21.

4. Eskow, R. N., and W. J. Loesche. 1971. Oxygen tensionsin the human oral cavity. Arch. Oral Biol. 16:1127-1128.

5. Globerman, D. Y., and I. Kleinberg. 1979. Intra-oral pO2and its relation to bacterial accumulation on the oraltissues, p. 275-292. In 1. Kleinberg, S. A. Ellison, andI. D. Mandel (ed.), Proceedings, Saliva and Dental Car-ies, special supplement to Microbiol. Abstr.

6. Gordon, D. F., M. Stutman, and W. J. Loesche. 1971.Improved isolation of anaerobic bacteria from the gingivalcrevice area of man. AppI. Microbiol. 21:1016-1021.

7. Kenney, E., and M. M. Ash. 1969. Oxidation reductionpotential of developing plaque, periodontal pockets andgingival sulci. J. Periodontol. 40:630-633.

8. Kessler, M., D. F. Bruley, L. C. Clark, Jr., D. W. Lub-bers, I. A. Silver, and J. Strauss. 1973. Oxygen supply.University Park Press, Baltimore.

9. Kleinberg, I., and C. J. Hall. 1969. pH and depth ofgingival crevices in different areas of the mouths of fastinghumans. J. Periodontal Res. 4:109-117.

10. Knowles, J. W., F. G. Burgett, R. R. Nissle, R. A. Schick,E. C. Morrison, and S. P. Ramfjord. 1979. Results ofperiodontal treatment related to pocket depth and attach-ment level. Eight years. J. Periodontol. 50:225-233.

11. Kormman, K. S., and W. J. Loesche. 1978. New mediumfor the isolation of Actinomyces viscosus and Actinomy-ces naeslundii from dental plaque. J. Clin. Microbiol.7:514-518.

12. Kornman, K. S., and W. J. Loesche. 1980. The subgingivalmicrobial flora during pregnancy. J. Periodontal Res.15:111-122.

13. Kornman, K. S., and W. J. Loesche. 1982. Effect ofestradiol and progesterone on Bacteroides melaninogeni-cus and Bacteroides gingivalis. Infect. Immun. 35:256-263.

14. Leadbetter, E. R., S. C. Holt, and S. S. Socransky. 1979.Capnocytophaga: new genus of gram-negative glidingbacteria. 1. General characteristics, taxonomic consider-ations and significance. Arch. Microbiol. 122:9-16.

15. Listgarten, M. A. 1976. Structure of the microbial floraassociated with periodontal health and disease in man. J.

Periodontol. 47:1-18.16. Listgarten, M. A., and L. Hellden. 1978. Relative distribu-

tion of bacteria at clinically healthy and periodontallydiseased sites in humans. J. Clin. Periodontol. 5:115-132.

17. Loe, H., and J. Silness. 1963. Periodontal disease inpregnancy. I. Prevalence and severity. Acta Odontol.Scand. 21:533-551.

18. Loesche, W. J. 1969. Oxygen sensitivity of various anaer-obic bacteria. Appl. Microbiol. 18:723-727.

19. Loesche, W. J. 1976. Chemotherapy of dental plaqueinfections. Oral Sci. Rev. 9:63-105.

20. Loesche, W. J. 1980. The role of anaerobes in periodontaldisease, p. 61-72. In G. Gottschalk, N. Pennig, and H.Werner (ed.), Anaerobes and anaerobic infections. Gus-tav Fischer Verlag, Stuttgart.

21. Loesche, W. J. 1982. Dental caries: a treatable infection.Charles C Thomas, Publisher, Springfield, Ill.

22. Loesche, W. J., and B. Laughon. 1982. Role of spirochetesin periodontal disease, p. 62-75. In R. J. Genco and S. E.Mergenhangen (ed.), Host-parasite interactions in peri-odontal disease. American Society for Microbiology,Washington, D.C.

23. Loesche, W. J., and L. H. Straffon. 1979. Longitudinalinvestigation of the role of Streptococcus mutans inhuman fissure decay. Infect. Immun. 26:498-507.

24. Loesche, W. J., S. A. Syed, B. E. Laughon, and J. Stoll.1982. The bacteriology of acute necrotizing ulcerativegingivitis. J. Periodontol. 53:223-230.

25. Macdonald, J. B., R. J. Gibbons, and S. S. Socransky.1960. Bacterial mechanisms in periodontal disease. Ann.N.Y. Acad. Sci. 85:467-478.

26. Mettraux, G., and H. Graf. 1981. A new method tomeasure oxygen tension (pO2) in human periodontalpockets. Schweiz. Monatsschr. Zahnheilkd. 91:1087-1088.

27. Morris, J. G. 1977. Oxygen and the growth of the oralbacteria, p. 293-306. In I. Kleinberg, S. A. Ellison, andI. D. Mandel (ed.), Proceedings, Saliva and Dental Car-ies, a special supplement to Microbiol. Abstr.

28. Mukherjee, S. 1981. The temperature of the periodontalpockets. J. Clin. Periodontol. 8:17-20.

29. Newman, M. G., and S. S. Socransky. 1977. Predominantcultivable microbiota in periodontitis. J. Periodontal Res.12:120-128.

30. Rogosa, M. 1964. The genus Veillonella. J. Bacteriol.87:162-170.

31. Siegel, S. 1956. Nonparametric statistics for the behaviorsciences. McGraw-Hill Book Co., New York.

32. Slots, J. 1979. Subgingival microflora and periodontaldisease. (review). J. Clin. Periodontol. 6:351-386.

33. Socransky, S. S. 1977. Microbiology of periodontal dis-ease-present status and future considerations. J. Perio-dontol. 48:497-510.

34. Socransky, S. S., and C. Hubersak. 1967. Replacement ofascitic fluid or rabbit serum requirement of Treponemadentium by a-globulin. J. Bacteriol. 94:1795-1796.

35. Syed, S. A., and W. J. Loesche. 1972. Survival of humandental plaque flora in various transport media. Appl.Microbiol. 24:638-644.

36. Syed, S. A., W. J. Loesche, H. R. Pape, and E. Grenier.1975. Predominant cultivable flora isolated from humanroot surface plaque. Infect. Immun. 11:727-731.

37. Syed, S. A., M. Svanberg, and G. Svanberg. 1980. Predom-inant cultivable dental plaque flora of beagle dogs withgingivitis. J. Periodontal Res. 151:125-136.

VOL. 42, 1983

on May 27, 2020 by guest

http://iai.asm.org/

Dow

nloaded from