664

Outer Membrane Protein and Lipopolysaccharide Heterogeneity among Eikenella corrodens Isolates C-K Casey Chen and Mark E. Wilson

Deparonent of Oral Biology, School of Dental Medicine, From the Department of New York at Buffalo State University of

Eikenella corrodens is a facultatively anaerobic gramnegative bacillus that is a common inhabitant of the oral cavity [1, 2], upper respiratory tract, and mucosal surfaces of the intestinal and genital tracts [3, 4]. There is conflicting evidence regarding the periodontopathic potential of this organism [1, 5-10]. However, E. corrodens is recognized as an infrequent pathogen capable of causing serious extraoral infection in immunocompromised and immunocompetent hosts [11-13]. of cell-surface components, including capThe importattce ofcell-surface sular polysaccharide, lipopolysaccharide (LPS), and outer membrane proteins (OMP), as potential targets ofthe immune response and as potential virulence factors associated with gram-negative infection has been established. Information regarding virulence factors that may contribute to the pathogenicity of E. corrodens and immune responses that may be relevant to host defense against this organism is sparse. However, the existence of several serogroups of E. corrodens, some of which are found only in periodontally diseased subjects, ofthis suggests that members of this species may differ in their pathogenic potential [14]. Although DNA hybridization studies indicate that E. cor-

Received 22 December 1989; revised 22 March 1990. DE-08240, and DE-07034 (US Public Health Grant support: DE-04898, DE-Q8240, Service). Wilson, Department of Oral Reprints and correspondence: Dr. Mark E. Wuson, Biology, State University of New York at Buffalo, 3435 Main St., Foster Hall, Buffalo, NY 14214. The Journal of Infectious Diseases 1990;162:664-671 © 1990 by The University of Chicago. All rights reserved. © 0022-1899/90/6203-0015$01.00

rodens strains are homogeneous [15-17], other evidence indicates that this species exhibits phenotypic diversity. It has been observed that E. corrodens exhibits variable colony morphology, biochemical reactivity, and susceptibility to serum bactericidal activity [16-18]. To assess the importance of such interstrain variation in the pathogenesis and epidemiology of E. corrodens infection, it is essential to have a convenient system for classification of this group of bacteria. Analysis of OMP or LPS electrophoretic mobility on SDSpolyacrylamide gels is a convenient means by which to evaluate phenotypic heterogeneity among strains ofvarious species of gram-negative bacteria. While some species such as Pseudomonas aeruginosa exhibit relatively conserved OMP profiles injluenzae exhibit sub[19,20], species such as Haemophilus influenzae stantial interstrain variation in both OMP and LPS profiles [21-23]. Subtyping systems based on differences in OMP or LPS electrophoretic mobilities have been useful in epidemiologic studies of infection due to various gram-negative bacteTyping schemes based on OMP patterns have ria [23-25]. lYping also been used to identify strains that are more pathogenic infection [23, 26-28]. and are associated with certain types of ofinfection We examined the OMP and LPS electrophoretic patterns of 27 reference and clinical isolates of E. corrodens to evaluate interstrain OMP and LPS heterogeneity, determine the feasibility of subtyping these organisms on the basis of OMP or LPS phenotype, and ascertain whether unique OMP or LPS patterns are associated with strains isolated from distinct clinical settings. Materials and Methods Strains. 1\venty-seven Twenty-seven E. corrodens strains were used in this study (table 1). The laboratory strains have been maintained in vitro for

Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 25, 2015

Outer membrane protein (OMP) and lipopolysaccharide (LPS) phenotypic diversity among 27 oral and extraoral strains of Eikenella corrodens was assessed by SDS-PAGE. Each strain exhibited one to three major protein bands in the 35- to 41.5-kDa range and one or two protein bands of lesser density in the 24.5- to 28-kDa range. Eleven OMP patterns were distinguished among the strains. While oral strains obtained from periodontaUy periodontally healthy and diseased subjects exhibited diverse OMP patterns, five of six strains from extraoral sites of infection expressed an identical OMP pattern. Comparison of the electrophoretic mobilities of LPS from these same strains revealed that E. corrodens LPS consists primarily oflow apparent molecular mass forms. Sixteen different LPS phenotypes were differentiated among the strains, with no apparent correlation between LPS phenotype and clinical setting. Strains expressing the same OMP pattern frequently expressed variable LPS phenotypes and vice versa. Analysis of OMP or LPS pattern by SDS-PAGE may be useful in taxonomic and epidemiologic studies of E. corrodens. Additional studies assessing the potential influence of OMP composition on invasiveness of this organism appear warranted.

Eikenella corrodens OMP & LPS Phenotypes

JID 1990;162 (September)

us

corrodens Table 1. Sources and phenotypic variations of Eikenella co"odens isolates. Strain (origin)

Type strain and laboratory strains SVNYaB 1 (NY) SVNYaB 4 (NY) SVNYaB5 (NY) SVNYaB 9 (NY) FDC 373 (MA) FDC 1073 (MA) ATCC 23834 (ATCC) (ATCe) Isolates from periodontitis patients VB 27 (NY) VB 30 (NY) VB 56 (NY) VB 80 (NY) VB 140 (NY) VB 163 (NY) VB 190 (NY) VB 344 (NY) VB 367 (NY) Isolates from extraoral infections D 3846 (KS) E 86 (DE) E 5178 (VA) E 9024 (MI) E 960E (AL) F 3624 (CO) Isolates from healthy subjects VB 67 (NY) VB 96 (NY) VB 105 (NY) VB 243 (NY) VB 281 (NY)

Source

Colony type

UP/dental plaque AP/dental plaque UP/dental plaque AP/dental plaque AP/dental plaque* AP/dental plaque Sputum

1 1 3 2 2 2 2

plaquett AP/dental plaque AP/dental plaque AP/dental plaque AP/dental plaque AP/dental plaque plaquett AP/dental plaque AP/dental plaque plaquett UP/dental plaque AP/dental plaque

1 2 1 2 1 2 1 1 2

Blood Blood Blood Blood Bite wound Neck abscess

3 2 3 2 2 3

PH/dental PH/dental PH/dental PH/dental PH/dental

3 2 3 3 2

plaque plaque plaque plaque plaque

NOTE. All NY isolates were from Buffalo; all MA isolates were from Boston. AP. localized juvenile periodontitis; PH, periodontally healthy. AP, adult periodontitis; UP, UP,loca1izedjuvenile All isolates were positive for ornithine-decarboxylase activity. * Isolate negative for lysine-decarboxylase activity; all others were positive. t Isolates positive for catalase activity; all others were negative.

*

Laemmli Laemrnli gel system [32]. Outer membrane fractions were incubated with an equal volume of2x of 2 x treatment buffer (0.125 MTris-CI M Tris-CI (pH 6.8), 4% SOS, SDS, 20% glycerol, 10% 2-mercaptoethanol, 0.006% bromophenol blue) at 100°C for 15 min. Electrophoresis was carried out on a vertical slab gel apparatus (Thll Mighty Small Vertical Slab Unit; Hoefer, San Francisco). Protein (1'\.11 (""1 pog) JLg) was loaded on each lane in a gel consisting ofa 3.8 % stacking gel and a 12%resolv12 % resolving gel containing 0.1% SOS. SDS. In some instances, 14% resolving gels containing 2.8 M M urea were used. Electrophoresis was carried out at 6 W W of constant power until the tracking dye was about at the bottom of the gel. Protein bands were visualized by staining with Coomassie brilliant blue R250 (Sigma). The LPS electrophoretic patterns were assessed with polyacryl% stacking gel and a 16.5 % resolving amide gels consisting of a 3.8 3.8% 16.5% gel containing 3.2 M M urea. Bromophenol blue was added to the LPS samples to 0.003 %, and the samples were electrophoresed at 6 W of constant power until tracking dye was at the bottom of the gel. LPS bands were visualized by silver staining as described by Hitchcock and Brown [31]. Phenol-extracted LPS from E. coli 011l:B4 Oll1:B4 (Sigma) was used as reference.

Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 25, 2015

>30 passages. The fresh clinical isolates were passed no more than five times in vitro with no overt changes in colony morphology. All isolates were examined for colony morphology and biochemical activity. E. corrodens co"odens is a gram-negative rod that is oxidase-, lysine-, and ornithine-decarboxylase positive; is catalase, urease, esculin hydrolysis, and indole negative; does not produce acid from glucose, xylose, mannitol, lactose, sucrose, or maltose; and reduces nitrate A few catalase-positive or lysine decarboxylase-negato nitrite [29]. A tive strains, identified as E. corrodens on the basis of ONA DNA homology shared with type strain ATCC 23834 [16], were also included in this study. Three of the fresh clinical isolates from periodontitis patients were obtained from a single patient. One of these (VB (UB 30) (UB exhibited different colony morphology from the other isolates (VB UB 56). VB UB 27 and VB UB 56 differed in catalase activity. Sources 27 and VB and characteristics of the E. corrodens strains are summarized in table 1. To examine the colony morphology, freshly thawed organisms were streaked on sheep blood agar and cultured at 37°C in humidified 55% % C02 for 3 days. A single colony was picked and dispersed in sterile distilled water and plated onto sheep blood agar with a bent glass rod. After incubation for 3 days, the colonies on the plate were photographed with a Zeiss stereomicroscope under reflected light. Growth conditions. Bacteria were maintained on sheep blood 0.001% pog/ml equine hemin agar supplemented with 0.001 % menadione, 5 JLglrnl % yeast extract and incubated at 3rC 37°C in a humidified ill, and 0.1 % To obtain log-phase bactechamber supplemented with 55% % C02. 1b co"odens was suspended in 10 rnl ml of ria, a heavy inoculum of E. corrodens Todd-Hewitt broth (supplemented with 2 mglrnl mg/ml KN03, 5 JLglrnl pog/ml hepog/ml L-cysteine, pH 7.2) and incubated overnight at min, and 50 JLglrnl 37°C in a 55% % CO2 chamber. Subsequently, the bacterial suspension was diluted 1:25 with fresh broth and incubated for an additional12-16 h. Stationary-phase organisms were obtained under the same conditions except the cells were incubated for 48 h. fractions. Outer membrane fracPreparation ofOMP and LPSfractions. tions were isolated by the procedure of Filip et al. [30]. Briefly, brothgrown bacteria were harvested by centrifugation at 14,500 g for 20 min. The bacterial pellet was suspended in 10 mM HEPES (pH 7.4) and disrupted with a sonicator (Sonicator, Heat System-Ultrasonics, of2Q-sec intervals at 50 50% % maximum Farmingdale, NY) for five bursts of20-sec output. Intact cells and insoluble debris were removed by centrifugation at 2500 g for 20 min, and the total membrane fraction in the supernatant was recovered by centrifugation at 100,000 g for 1 h at 4°C. The total membrane fraction was subsequently treated with 1% sodium lauroyl sarcosinate (Sigma Chemical, St. Louis) for 30 1% min at room temperature followed by centrifugation at 100,000 g for 11h. h. The pellet (sarkosyl-insoluble outer membrane fraction) was then suspended in distilled water. Protein concentration in the outer membrane fractions was determined by a dye-binding assay (BioRad Laboratories, Richmond, CA) using BSA (bovine serum albumin) as a standard. The LPS fraction was prepared by a modified procedure of Hitchcock and Brown [31]. The sarkosyl-insoluble outer membrane fractions were dispersed in 100 pol JL1 of lysing buffer (4% SOS, 10% glycerol, and 1.0 MTris, pH 6.8) 2-mercaptoethanol, 2% SDS, and solubilized by incubation at 100°C for 10 min. The material pol of proteinase was cooled to room temperature, after which 10 JLl mg/ml freshly prepared in lysing buffer) K (at a concentration of 10 mglrnl was added. The solution was incubated at 60°C 4-6 h to allow complete protein digestion. SDS-PAGE. OMP profiles were assessed using a modified

665

666

Chen & Wilson

JID 1990;162 (September)

Results

Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 25, 2015

Three colony morphotypes were differentiated among these strains (figure 1). Type 2 colonies are circular with an irregular margin and typically ~3 mm rom in diameter. The colony contains a rough, nonrefractile opaque perimeter and a moist umbonate center and appears to pit into the agar surface. Type 1 colonies are usually larger and thicker than type 2, contain a large glistening umbonate center with a relatively narrow nonrefractile perimeter, and have a yellowish and shiny appearance under reflective light. Type 3 colonies exhibit a small, round, smooth surface without the granular perimeter and do not appear to "corrode" the agar surface. Some E. corrodens strains were found to convert from type 2 to type 3 colony morphotype or vice versa. However, we noted that these strains adopt a preferred colony morphotype that is usually stable even after numerous in vitro passages (unpublished data). The type 1 colony morphotype is stable and can be easily differentiated from type 2 or 3. In no instance have we observed conversion between type 1 and type 2 or 3 colonies after in vitro passage. In preliminary studies we determined that OMP and LPS electrophoretic profiles of selected E. corrodens strains were not influenced by the phase ofgrowth in which the organisms were harvested (data not shown). Moreover, fresh isolates maintained in vitro for >15 passages exhibited a stable OMP and LPS phenotype (data not shown). Nevertheless, to facilitate interstrain comparisons, OMP and LPS electrophoretic profiles were assessed using bacteria grown under identical conditions and harvested in logarithmic phase, and when applicable, only fresh isolates (fewer than five passages in vitro) were used. Th gain insight into the extent of OMP or LPS phenotypic heterogeneity among E. corrodens, outer membranes of 27 strains representing type and laboratory strains, fresh oral strains from periodontally healthy and diseased subjects, and strains isolated from extraoral sites of infection were analyzed by SOS-PAGE (table 1). The OMP profiles of oral and extraoral strains of E. corrodens are depicted in figure 2. Each of the strains tested exhibited one to three densely staining bands in the 35- to 41.5-kDa range. OMP in this group often exhibited only minor differences in apparent molecular mass. Mixing experiments were performed in which samples containing OMPs of different mobility were coelectrophoresed in the same lane of an SOS-polyacrylamide gel. The results of such experiments confirmed that the minor differences in apparent molecular mass of the major OMP represent distinct OMP phenotypes (data not shown). 24.S to A second band of lesser density appeared in the 24.5 28-kDa range. There appeared to be less variability in this OMP group than was evident for the high-molecular-mass major OMP. However, some strains exhibited t\m two closely spaced bands of comparable staining intensity, and three strains expressed an additional OMP of 31 kDa (figure 20, lanes 1, 3, and 7).

EikenelIa corFigure 1. Colony morphology of representative Eikenel/a rodens isolates. Isolates were plated onto sheep blood agar and incubated at 37°C in a humidified chamber supplemented with 55% % CO2 for 3 days. A, E. corrodens type strain AlCC ATCC 23834, type 2 colony; D, B, UB 56, type 1 colony; C, SUNYaB 5, type 3 colony. Bar = 1 mm. rom.

110 1ID 1990;162 (September)

Eiunel/a corrodens OMP & LPS Phenotypes Eikene/fa

667

66-

66-

66-

45-

45-

45-

123456

C

14.2 -

---

12345670

123456789

Figure 2. Outer membrane protein patterns of Eikene/la corrodens assessed by SDS-PAGE. A, E. corrodens isolates from dental plaque of periodontally healthy subjects; lanes 1-5: VB UB 67, VB UB 96, VB UB 105, VB UB 243, and VB UB 281. B, D, E. corrodens isolates from extraoral infection; lanes 1-6: D 3846 (blood isolate), E 86 (blood isolate), E 5178 (blood isolate), E 9024 (blood isolate), E 960E (bite wound), and F 3624 (neck abscess). C, E. corrrxkns corrodens type strain and laboratory strains recovered from dental plaque; lanes 1-7, ATCC 23834, FDC 373, FDC 1073, SUNYaB 1, SUNYaB 4, SUNYaB 5, and SUNYaB 9. D, E. corrodens isolates from dental plaque of periodontitis patients; lanes 1-9: VB UB 27, VB UB 30, VB UB 56, VB UB 80, VB UB 140, UB VB 163, VB UB 190, VB UB 344, and VB UB 367. Isolates in lanes 1-3 were from one periodontitis patient. Molecular mass standards (MW-SDS-70L Kit, Sigma) are bovine albumin (66 IDa), kDa), egg albumin (45 kDa), g1yceraldehyde-3phosphate dehydrogenase (36 kDa), carbonic anhydrase (29 kDa), trypsinogen (24 kDa), soybean trypsin inhibitor (20.1 kDa), and a-lactalbumin (14.2 kDa).

On the basis of variation in electrophoretic mobilities of the high- and low-molecular-mass major OMP of these 27 E. corrodens strains, 11 different OMP patterns were identified:- A schematic representation of the OMP patterns, along with the number of strains and source(s) of each phenotype, is depicted in figure 3. Although additional strains must be evaluated to define associations between OMP pattern and anatomic location or clinical setting, a number of observations can be made. First, five of six extraoral strains of E. corrodens (obtained from Centers for Disease Control, Atlanta) of diverse geographic origin expressed a single OMP phenotype (figures 2B and 3, lane 3) and shared a common low-molecular-mass band. Similarly, three of five strains from periodontally healthy subjects shared a common OMP pattern and low-molecular-mass band (figures 2A and 3, lane 8). In contrast, E. corrodens strains from patients with adult periodontitis exhibited diverse OMP profiles spanning 8 of the 11 recognizable OMP phenotypes (figures 20 2D and 3). Hence, there was no indication of an association between OMP pattern and adult periodontitis. In addition, three oral strains from subjects with localized juvenile periodontitis exhibited three 2D and 3), again suggesting distinct OMP patterns (figures 20 the absence of an association between OMP phenotype and the presence of E. corrodens in periodontally diseased sites. Interstrain phenotypic variation among E. corrodens strains was also demonstrated by analysis ofLPS electrophoretic mobility on silver-stained SOS-polyacrylamide SDS-polyacrylamide gels. E. corrodens

strains displayed only limited LPS microheterogeneity, in con&chtrast to the extensive microheterogeneity observed with Escherichia coli 0111:B4 LPS (figure 4). Most strains displayed two or three major bands oflow apparent molecular mass consistent with a lipooligosaccharide structure. Sixteen distinct LPS electrophoretic profiles were identified among the 27 E. corrodens strains examined. There was no apparent relation between LPS pattern and the source of the strains or their clinical setting (figure 3). In some instances, strains sharing a common OMP profile also exhibited the same or similar LPS electrophoretic profiles (e.g., compare figure 2A, lanes 1 and 2, with figure 4A, lanes 2 and 3; figure 2C, lanes 1-3, with figure 4C, lanes 2-4). In contrast, other strains with identical OMP patterns (figure 2B, lanes 4-6) expressed distinct LPS profiles (figure 4B, lanes 5-7). Conversely, strains with indistinguishable LPS profiles exhibited different OMP patterns. Three distinct colony morphotypes were distinguished among the E. corrodens strains examined (figure 1). Strains ofthe same colony morphotype were found to exhibit different OMP and LPS patterns, suggesting the absence of a relationship between LPS or OMP phenotype and colony morphotype. Strains VB UB 27, VB UB 30, and UB VB 56 were from one periodontitis patient and exhibited differences in colony morphology and catalase activity (table 1). The OMP and LPS elecUB 27 and VB UB 56 were identical but trophoretic patterns of VB 2D, lanes 1-3; figure 40, 4D, lanes differed from VB UB 30 (figure 20,

Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 25, 2015

------

8 12 23 34 45 58

2924-

14.2-

14.2 -

14.2-

36-

2924-

2924-

2924-

45-

36-

36-

36-

A

66-

668

JIO 1990;162 (September) 110

WLison Chen & Wtlson

OMP Patterns of E. corrodens Isolates 41.5 41 40

39

38

1

---- --

37 36 35

U8140 UB140 (AP)

5

4

6

7

8

-

-- - - - -

--

-

..........-:: .........-:=

~ -;:::ft ::::0 rr

3

0030 UB30 (AP)

UB163 UBl63 CAP) (API

Strain designation (source of isolation)

UBBO UB80 (AP)

--

-

--

UB281 (PH)

UB27 (AP)

SUNYaB 1 (UP)

UB56 (AP)

SUNYaB 4 (AP) {AP}

UB190 UB 190 (AP)

9

-

-

UB67 (PH)

FDC1073 (AP)

UB96 (PH)

FDC373 (AP)

UB243 (PH)

SUNYa85 SUNYaB5 (WP) (UP)

E9024 E!1024 (Blood) (BIDod)

UB367 (AP)

SUNYa89 SUNYaB9 (AP) lAP)

E960E

E5178 ES178 (Blood)

ATCC23834 (Sputum)

(Blewound) (Blewourd)

11

- -

E86 (Blood) (BIDod)

03846 03&46 (Blood) (BIDod)

10

- U8344 UB344 (UP)

00105 (PH)

F3624 (Neck (Ned!

abscess)

isoIaIes Total no. of isoIal

1

2

6

1

2

2

1

5

5

1

1

01 LPS patterns No. of

1

2

4

1

1

2

1

3

3

1

1

Figure 3. Schematic representation of 11 different outer membrane protein (OMP) patterns among 27 Eikenella corrodens isolates. Strain designation and sources of isolation are indicated below each OMP pattern. Numbers of different lipopolysaccharide (LPS) patterns expressed by strains with identical OMP patterns are also indicated. Sources of isolation: AP, adult periodontitis; PH, periodontally healthy; UP, localized juvenile periodontitis. Additional information is provided in table 1.

2-4). This suggests that this patient was colonized by at least two different E. corrodens strains.

Discussion During the past two decades, E. corrodens infections have been increasingly recognized. This organism is isolated most frequently from infections ofthe blood, head, neck, and central nervous system. In addition, E. corrodens appears to be a common pathogen in human bite wounds and closed-fist injuries complicated by osteomyelitis [12]. The organism may be isolated either as the sole infecting pathogen or as part of a mixed infection, often involving streptococci. Postanginal sepsis and a number ofother infections caused by E. corrodens have been associated with dental infection due to this organism [12, 33]. >80 % of subjects (both perioRecent studies indicate that >80% dontally healthy and diseased) harbor E. corrodens in the oral cavity, the primary ecologic niche being dental plaque [2]. Although the prevalence of E. corrodens in dental plaque of

periodontitis subjects is increased, the percentage of the total cultivable microftora represented by this species did not differ between periodontally healthy and diseased subjects. Such findings cast doubt on the role of E. corrodens in the etiology and pathogenesis of periodontal disease. It is intriguing that E. corrodens remains an uncommon pathogen despite its frequent occurrence in the oral cavity of periodontally healthy and periodontally diseased subjects. While the basis for this finding is unclear, there are numerous precedents in the literature demonstrating that subtypes within a given species of gram-negative bacteria can exhibit variable pathogenicity and can be associated with different types of infection [28, 34]. In this context, Badger and Tanner [14] identified seven distinct serogroups among E. corrodens strains recovered from periodontally healthy and diseased subjects, noting that certain serotypes were not represented in plaque samples from periodontally healthy sites. This raised the question as to whether both virulent and avirulent strains of E. corrodens may exist in the oral cavity, with only the former being involved in the pathogenesis of perioTo address this issue, it was dontal or extraoral infection. 1b

Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 25, 2015

28 26 25 24.5

~

2

Eikenella corrodens OMP & LPS Phenotypes

110 JID 1990;162 (September)

669

66-

66-

66-

45-

45-

45-

45-

3629-

36-

3629-

3629-

24-

24-

24-

14.2-

14.2-

14.2-

29-

24-

14.2 -

A

123456

B

1234567C

23456780

, 1

2 3 4 5 6 7 8 9 10

Figure 4. Silver-stained lipopolysaccharide (LPS) patterns of Eikenella corrodens isolates assessed in gels consisting of 3.8% 3.8 % stacking gel and 16.5 % resolving gel containing 3.2 M urea. Molecular mass markers are indicated. Lane 1I of each gel contained LPS from Escherichia coli strain 0111:84. Olll:B4. A, E. corrodens isolates recovered from periodontally healthy subjects; lanes 2-6 correspond to lanes 1-5 in figure 2A. B, E. corrodens isolates from extraoral infection; lanes 2-7 correspond to lanes 1-6 in figure 28. 2B. C, E. corrodens laboratory isolates; lanes 2-8 correspond to lanes 1-7 in figure 2C. D, E. corrodens clinical isolates from periodontitis patients; lanes 2-10 correspond to lanes 1-9 in figure 2D.

necessary to assess genotypic and phenotypic diversity among strains of E. comxJens corrodens obtained from different clinical settings. The relation between surmee-exposed surfu~xposed outer membrane components, including LPS and OMP, and expression of virulence has been demonstrated. Moreover, classification schemes based on electrophoretic mobility of OMPs or LPS in SOSpolyacrylamide gels have been useful in identifying subtypes that exhibit more invasiveness in certain clinical settings than do other strains of the same species [28]. LPS and OMP electrophoretic analysis has also been shown to be useful in studying bacterial transmission [24, 27]. In this study, we used SOS-PAGE to assess OMP and LPS phenotypic diversity among oral and extraoral strains of E. corrodens. E. corrodens exhibited substantial electrophoretic heterogeneity with respect to both OMP and LPS structure. The OMP and LPS profiles appeared to be stable markers and were not influenced by growth phase or in vitro passage. Eleven OMP patterns and 16 LPS patterns were differentiated among the 27 strains examined. Strains sharing a common OMP pattern were found to exhibit distinct LPS patterns and vice versa. Three of five strains recovered from periodontally healthy subjects exhibited an identical OMP pattern. This suggests that certain subtypes of E. corrodens may constitute part of the oral flora of periodontally healthy subjects. On the other hand, nine different OMP patterns were detected among 15 strains derived from patients with adult or juvenile forms of periodontal disease. Hence, there was no evidence for an association between OMP subtype and periodontal disease.

However, the limited number of strains examined precludes definitive assessment of the number of possible OMP subtypes among E. corrodens and the distribution of subtypes among periodontally healthy and diseased subjects. Our study included an examination of the OMP profiles of six strains of E. corrodens obtained from various extraoral sites of infection (and of different geographic origins). Five of these strains exhibited an identical OMP pattern. Analysis of additional extraoral strains from diverse clinical settings is needed to determine whether this OMP subtype is characteristic of more invasive strains of E. corrodens. While the potential role of OMPs in defining virulence of E. corrodens has not been defined, Thfano et al. [35] noted that the 42-kDa major OMP of this species is a porin capable of activating ltg/mI, murine macrophages in vitro. At concentrations >10 ILglml, this molecule was toxic for macrophages. Maliszewski et al. [36] demonstrated that monoclonal antibody to the 42-kDa OMP of a single strain reacted with only 1 of 11 additional E. corrodens strains, indicating that this protein is both structurally and antigenically diverse. Analysis of E. corrodens LPS by SOS-PAGE revealed the presence of marked interstrain LPS electrophoretic heterogeneity, indicating that the chemical composition of LPS is of Kato et al. [37], different. This is consistent with the results ofKato who noted O-antigenic heterogeneity in LPS from seven strains of E. corrodens. There was no evidence indicating that certain LPg LPS phenotypes are associated with periodontal disease activity or extraoral infection. Multiple isolates obtained from a single periodontitis sub-

Downloaded from http://jid.oxfordjournals.org/ at University of Birmingham on August 25, 2015

66-

670

Chen & Wilson

E. corrodens. Acknowledgment We thank Paul M. Bronson for technical assistance, Paul Dressel for assistance in preparing illustrations, Mirdza Neiders for helpful discussions, and Tunothy F. Murphy for critical reading of this manuscript.

References

mc, Tarenzi LA, Agyare EO, Berger JR. Prevalence of Eikenella corrodens in dental plaque. J Clin Microbioll983;17:636-639 MicrobioI1983;17:636-639 JJ. Eikenella corrodens Zambon JI. Chen CKC, Dunford RG, Reynolds HS, l.ambon in the human oral cavity. J Periodontol 1989;60:611-616 Montclos H, Boude M, Den

Outer membrane protein and lipopolysaccharide heterogeneity among Eikenella corrodens isolates.

Outer membrane protein (OMP) and lipopolysaccharide (LPS) phenotypic diversity among 27 oral and extraoral strains of Eikenella corrodens was assessed...
900KB Sizes 0 Downloads 0 Views