Vol. 60, No. 3

INFECrION AND IMMUNITY, Mar. 1992, p. 804-809

0019-9567/92/030804-06$02.00/0 Copyright C) 1992, American Society for Microbiology

Characterization of an Antigenically Conserved Heat-Modifiable Major Outer Membrane Protein of Branhamella catarrhalis JAWAD SARWAR,"12 ANTHONY A. CAMPAGNARI,' CHARMAINE KIRKHAM,"12 AND TIMOTHY F. MURPHY' 2*

Division of Infectious Diseases, Department of Medicine, and Department of Microbiology, School of Medicine and Biomedical Sciences, State University of New York at Buffalo, 1 and Buffalo VAMC, Medical Research 151, 3495 Bailey Avenue,2* Buffalo, New York 14215 Received 19 September 1991/Accepted 2 December 1991

Branhamella catarrhalis is a common cause of otitis media in children and of respiratory infections in adults with chronic bronchitis. Little is known about the antigenic structure of the outer membrane proteins (OMPs). In this study, two murine monoclonal antibodies, 7D6 and 5E8, were developed and used to characterize the major heat-modifiable OMP (OMP C/D) of B. catarrhalis. Immunoblot assays indicated that OMP C/D is heat modifiable, having a molecular mass of 55 kDa at room temperature and a mass of 60 kDa when heated under reducing conditions. Expression of the epitopes is independent of growth phase and growth media. Both epitopes are present in 51 of 51 strains of B. catarrhalis tested and are highly specific for Branhamella strains, being absent from a variety of other gram-negative species. Antibody 5E8 recognizes an epitope which is expressed on the surface of the intact bacterium. We conclude that OMP C/D is a major, heat-modifiable OMP antigen that expresses at least one stable, conserved epitope on the surface of B. catarrhalis. Future studies should focus on the role of OMP C/D in pathogenesis and on its potential role as a vaccine antigen.

species specificity of two epitopes recognized by monoclonal antibodies. (This work was presented in part at the Buffalo Conference on Microbial Pathogenesis, April 1990, Buffalo, N.Y., and at the Fifth International Symposium on Recent Advances in Otitis Media, May 1991, Fort Lauderdale, Fla.)

Branhamella catarrhalis has emerged as an important human pathogen in the past decade (5). The principal site of colonization of B. catarrhalis is the upper respiratory tract, and the organism is predominantly a mucosal pathogen. It is the third most common cause of otitis media in infants and children after nontypeable Haemophilus influenzae and Streptococcus pneumoniae (5, 19). Based on recent prospective studies utilizing cultures of middle ear fluid obtained by tympanocentesis, B. catarrhalis causes approximately 20% of cases of otitis media (3, 15, 27, 31). The incidence of otitis media caused by B. catarrhalis is increasing. As strategies for preventing otitis media caused by pneumococcus and nontypeable H. influenzae are developed, the relative importance of B. catarrhalis as a cause of otitis media will increase in the next decade. B. catarrhalis is also an important cause of lower respiratory tract infections in adults, particularly in the setting of chronic bronchitis and emphysema (6, 12, 17, 24, 25, 29, 34). The bacterium causes pneumonia and purulent exacerbations of chronic bronchitis. Finally, B. catarrhalis causes sinusitis in children and adults (4, 9, 30, 32, 33) and occasionally causes invasive disease (7, 8, 10, 11, 13, 26). The emergence of B. catarrhalis as an important human pathogen has stimulated interest in studies of the outer membrane proteins (OMPs) of the organism. Studies of OMPs are important to identify virulence factors, to elucidate structure-function relationships in the outer membrane, to characterize surface antigens, and to understand the immune response to infection. Recent studies have begun to shed light on the basic characteristics of the OMPs of B. catarrhalis (2, 20, 23, 28). However, little is known about the antigenic structure of OMPs of this bacterium. In this study, we describe features of a major OMP, OMP C/D, its heat modifiability, and the conservation, surface exposure, and

*

MATERIALS AND METHODS

Bacteria. B. catarrhalis 4223 and 5191 are clinical isolates recovered by tympanocentesis from the middle ears of two different children with otitis media in Buffalo, N.Y. B. catarrhalis 56 was recovered from the sputum of an adult with a lower respiratory tract infection in Johnson City, Tenn. Other bacteria were used to test the strain and species specificity of monoclonal antibodies 7D6 and 5E8. These are listed in Table 1. The identity of B. catarrhalis strains was confirmed by established criteria, including Gram stain; colonial morphology; DNase, catalase, and cytochrome oxidase production; and the inability to utilize maltose, glucose, lactose, and sucrose as carbohydrate sources (2). Strains were grown on chocolate agar plates at 37°C in 5% CO2 or in Mueller-Hinton broth. Bacteria were stored at -70°C in Mueller-Hinton broth plus 10% glycerol. Development of hybridomas. Antibody 7D6 was developed by immunizing 4-week-old BALB/c mice intraperitoneally with 109 whole cells of B. catarrhalis 4223 on days 0 and 28. Antibody 5E8 was developed by immunizing mice with Zwittergent-extracted outer membrane of B. catarrhalis 56 with the following schedule: day 0, 10 ,ug with complete Freund's adjuvant, subcutaneously; days 7 and 14, -25 ,ug with incomplete Freund's adjuvant, subcutaneously; days 21 and 28, -45 ,ug with no adjuvant, intraperitoneally. For both the 7D6 and 5E8 fusions, the mice were sacrificed on day 32 after the initial injection, and the spleens were removed and perfused to collect splenocytes. A modification of the procedure of Kennett (14) and our previously described method

Corresponding author. 804

VOL. 60, 1992

OUTER MEMBRANE PROTEIN C/D OF B. CATARRHALIS

TABLE 1. Species specificity of the epitopes recognized by antibodies 7D6 and 5E8 Strain

Branhamella catarrhalis Branhamella speciesa

Moraxella

species'

Kingella speciesc Haemophilus influenzaed Haemophilus speciese Neisseria gonorrhoeae Neisseria meningitidis Actinobacillus species Other gram-negativesf

No. of strains tested

51 3 3 3 10 12 10 10 3 11

No. of strains reactive with:

7D6

5E8

51 3 1 0 0 0 0 0 0 0

51 2 1 0 0 0 0 0 0 0

a One strain each of B. ovis, B. caviae, and B. cuniculi. One strain each of M. osloensis, M. phenylpyruvica, and M. atlantae. c One strain of K kingae and two strains of K denitnificans. d Three serotype b strains and seven nontypeable strains. ' Three strains of H. parainfluenzae, two strains each of H. ducreyi and H. aphrophilus, and one strain each of H. parahaemolyticus, H. paraphrophilus, H. paraphrohaemolyticus, H. haemolyticus, and H. segnis. f Five strains of E. coli and one strain each of Klebsiella pneunoniae, Proteus mirabilis, Proteus vulgaris, Pseudomonas aeruginosa, Serratia marcescens, and Enterobacter aerogenes. b

(1) was used for the fusion. Briefly, splenocytes and Sp210Agl4 plasmacytoma cells were centrifuged together in a 5:1 ratio at 170 x g for 10 min. The resulting pellet was loosened gently, and 1 ml of 50% polyethylene glycol in 0.075 M HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid) (Boehringer Mannhein Biochemicals, Indianapolis, Ind.) was slowly added at room temperature and incubated for 1 min. The concentration of polyethylene glycol was then diluted by the sequential addition of Dulbecco's modified Eagle medium over 7 min. Finally, 8 ml of Dulbecco's modified Eagle medium plus 20% heat-inactivated fetal bovine serum was added, and the suspension was centrifuged at 200 x g for 10 min at 27°C. The supernatant was completely removed, and the cells were suspended in Dulbecco's modified Eagle medium plus 20% fetal bovine serum with supplements at 3 x 105 plasmacytoma cells per ml. Aliquots of 0.05 ml were placed into each well of 96-well microtiter plates. On the day following the fusion, 0.05 ml of HAT medium containing hypoxanthine (13.6 ,g/ml), aminopterin (0.36 ,g/ml), and thymidine (3.9 ,ug/ml) was added. Microtiter plates were incubated under 5% CO2 and 95% humidity at 35°C. Wells were examined for the presence of clones after 2 weeks. Screening of hybridomas. Whole-cell lysate, Zwittergentextracted outer membrane, and Triton-insoluble fractions were prepared from the homologous strains as previously described (21, 23). Tissue culture supernatants from hybridomas were screened initially by immunodot assays with these three preparations of the homologous strain. The antigen preparations were dotted onto nitrocellulose, air dried and incubated in 5% nonfat dry milk in buffer A (0.01 M Tris, 0.15 M NaCl, pH 7.4) for 1 h. The nitrocellulose was washed with buffer A and overlaid with tissue culture supernatant from hybridomas and incubated overnight at room temperature. The nitrocellulose was washed and incubated in a 1:1,000 dilution of goat anti-mouse immunoglobulin G (IgG)-IgM-IgA peroxidase-labeled conjugate (Kirkegaard & Perry Laboratories, Gaithersburg, Md.) for 1 h at room temperature. The immunodot assays were developed

805

with horseradish peroxidase color developer as previously described (22). A positive control for each assay included polyclonal mouse antiserum recovered from the mice whose splenocytes were used in the fusions. Three negative controls for each assay included (i) Sp2/0-Agl4 tissue culture supernatant in place of hybridoma supernatant, (ii) buffer A in place of tissue culture supernatant, and (iii) buffer A in place of the peroxidase conjugate. After antibodies 7D6 and 5E8 were identified as described in the Results, the hybridomas were cloned by limiting dilution and the isotypes were determined by using the Mono Ab ID EIA kit (Zymed Laboratories, Inc., San Francisco, Calif.). Ascites fluid was prepared by injecting hybridoma cells into the peritoneums of pristane-primed BALB/c mice, and the fluid was collected 2 to 4 weeks later. SDS-PAGE and immunoblot assays. Samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 11% separating gels with 3% stacking gels and were either stained with Coomassie blue or subjected to immunoblot assays by previously described methods (22). Once the antigens were transferred to nitrocellulose, the immunoblot assays were performed as described above for the immunodot assays. Immunoadsorption experiments. To determine whether the antibodies recognize epitopes that are expressed on the surface of the intact bacterium, we used the immunoadsorption method of Loeb (16). Tissue culture supematants were precipitated with ammonium sulfate by standard methods before these experiments were performed. The ammonium sulfate-precipitated antibodies were tested at various dilutions in an immunodot assay with whole-cell lysates to determine the dilution at which to perform the adsorptions. Bacteria were grown in broth to the mid-logarithmic phase and harvested by centrifugation at 5,000 x g for 15 min. Cells were suspended in phosphate-buffered saline with 0.5 mM MgCl2 plus 0.15 mM CaCJ2 added (PCM). The equivalent of 50 ml of cells was gently suspended in 0.8 ml of PCM. To this suspension was added 0.8 ml of diluted monoclonal antibodies which were then gently mixed. The suspensions were incubated at room temperature for 60 min with occasional mixing. Cells were centrifuged, and the adsorbed antibody preparations were tested in immunoblot assays. To control for the possibility that antibodies might adhere nonspecifically to bacteria, aliquots of the antibodies were also adsorbed with a strain of Kingella denitrificans and a strain of Escherichia coli. These bacteria do not express the epitopes recognized by 7D6 and 5E8. Immunofluorescence. To further assess whether the antibodies recognized epitopes that are expressed on the surface of the intact bacterium, we performed immunofluorescence assays. Bacterial cells were harvested from an agar plate, suspended in 0.9% sodium chloride, placed on printed slides, and allowed to air dry. A volume of 0.05 ml of tissue culture supernatant was placed on each section of the slide and incubated for 1 h at room temperature. After the slides were thoroughly washed with 0.9% NaCl plus 0.05% Tween 20, goat anti-mouse Ig conjugated to fluorescein was placed on the slide and incubated for 30 min at room temperature. The slides were washed with saline and observed under a fluorescent microscope for fluorescence of individual bacterial cells. Three negative controls were included with each experiment. (i) Sp2/0-Agl4 tissue culture supernatant was used in place of antibody. (ii) Saline was used in place of antibody. (iii) K denitrificans F8841, which lacks the epitopes recognized by antibodies 7D6 and 5E8, was used in place of B. catarrhalis strains. In all experiments, the

806

SARWAR ET AL.

INFECT. IMMUN.

gel

7D6

5E8

7D6 -94 -67

-94

B

-67 - 43

-43

-30 -30

G

b

a

c

a

0%t

H

-20.1 a

b

a

b

a

negative controls did not produce fluorescence of bacterial cells. RESULTS Antigenic specificity of antibodies 7D6 and 5E8. To analyze antigenic determinants on the major OMP of B. catarrhalis, we developed monoclonal antibodies. Mice were immunized as described above, and hybridoma supernatants were tested by immunodot assays for the presence of antibodies to whole-organism lysates and purified outer membrane preparations of the homologous strains. Initial screening identified hybridomas, labeled 7D6 and 5E8, that produced positive immunodot assays with whole-organism lysates, Zwittergent-extracted outer membranes, and Triton-insoluble preparations. Antibody 7D6 was identified as an IgG2a, and antibody 5E8 was an IgM isotype. Immunoblot assays were performed to further characterize these antibodies. Figure 1 shows a gel and immunoblot assay of Zwittergentextracted outer membrane of B. catarrhalis 4223 in which two lanes were stained with Coomassie blue and four lanes were transferred onto nitrocellulose and probed with antibodies 7D6 and 5E8. In the samples which were solubilized at room temperature in the absence of 3-mercaptoethanol, a broad band of approximately 55 kDa is apparent. In the samples solubilized at 100°C in the presence of ,B-mercaptoethanol, a double band of approximately 60 kDa is detected. The molecular mass and heat modifiability of the bands art characteristic of OMP C/D. Furthermore, since the Zwittergent extract is free of cellular components other than outer membrane (21), and since bands other than C/D are not apparent in the Zwittergent extract in this molecular mass range, we conclude that antibodies 7D6 and 5E8 are directed against OMP C/D. The molecular mass of OMP C/D was determined by measurement of the Rf and calculation from standard curves with molecular mass standards. This was performed on four separate gels. Based on these measurements, the heatmodified form of OMP C/D is approximately 60 kDa and the non-heat-modified form is approximately 55 kDa. These sizes for OMP C/D are somewhat lower than previous measurements which were based on standard curves from a previous series of gels (2). This difference appears to be a result of variability in mobility of proteins in SDS-PAGE

C

a

b C -20.1 10%l

5E8

b

FIG. 1. Coomassie blue-stained SDS gel (left panel) and immunoblot assays with antibody 7D6 (center panel) and antibody 5E8 (left panel). All three panels are from the same gel, and lanes contain preparations of strain 4223. Lanes: a, Zwittergent-extracted outer membrane solubilized at room temperature; b, Zwittergent-extracted outer membrane solubilized at 100°C. Major OMPs are noted on the left, and molecular mass standards are noted (in thousands of daltons) on the right.

b 5

-94 -67 ~~~~~-43

I

30

a

b

c

a

b

c

a

b

c

20.1

10% 0%t 5%r FIG. 2. Immunoblot assays of whole-organism lysates of strain 4223. The concentration of 13-mercaptoethanol in the sample buffer is noted at the bottom. Samples were incubated as follows: lanes a, room temperature, 5 min; lanes b, 100°C, 5 min; lanes c, 100°C, 60 min. Panel 7D6 was incubated with antibody 7D6 and then with protein A-peroxidase. Panel 5E8 was incubated with antibody 5E8 and then with peroxidase-conjugated anti-mouse IgM. Molecular mass standards are noted (in thousands of daltons) on the right.

performed under different conditions with reagents from different suppliers. Heat modifiability of OMP C/D. The changes in apparent molecular mass of OMP C/D with various solubilization temperatures and reducing agents were studied with antibodies 7D6 and 5E8. Whole-cell lysates were solubilized in SDS sample buffers containing 0, 5, and 10% 13-mercaptoethanol. Individual samples were incubated at room temperature for 5 min, at 100°C for 5 min, and at 100°C for 1 h. Figure 2 shows the results of immunoblot assays of these preparations. All lanes contain an identical amount of protein. When samples were solubilized at room temperature, a band of approximately 55 kDa was seen in all three preparations (lanes a). When samples were heated, a progressive shift to a doublet of approximately 60 kDa was observed (lanes b and c). The shift occurred independent of ,-mercaptoethanol. This doublet appears to represent the fully heatmodified form of the protein. Increasing amounts of 13mercaptoethanol, substitution of 3-mercaptoethanol with dithiothreitol, or the addition of iodoacetamide resulted in no additional modification. Antibody 5E8 is directed against both bands of the doublet in all samples subjected to heat. Antibody 7D6 produces a less prominent signal on the upper band in several samples but strongly detects the lower band in all samples subjected to heat. Therefore, the 5E8 epitope is stably expressed on both bands of the doublet, whereas the 7D6 epitope appears somewhat labile on the upper band but is stably expressed on the lower band. Effect of growth phase. To assess the effect of the bacterial growth phase on the expression of the 7D6 and 5E8 epitopes, we harvested cells of a growing broth culture of strain 4223 by centrifugation at different phases of growth over time. Cells were solubilized in sample buffer and subjected to SDS-PAGE and immunoblot assay. The 7D6 and 5E8

OUTER MEMBRANE PROTEIN C/D OF B. CATARRHIALIS

VOL. 60, 1992

807

7D6

a

b

c

d

e;

5E8

a

b

c

d

e

FIG. 3. Immunoblot assay. All lanes of the top panel contain Zwittergent-extracted outer membrane of B. catarralis 4223. The strips were developed with aliquots of antibody 7D6 which were adsorbed with the following: lane a, unadsorbed; lane b, B. catarrhalis 4223; lane c, B. catarrhalis 5191; lane d, K denitrificans 8841; lane e, E. coli Y1090. All lanes of the bottom panel contain Zwittergent-extracted outer membrane of strain 5191. The strips were developed with aliquots of antibody 5E8 which were adsorbed with the same strains as for antibody 7D6. The antibodies recognize both the heat-modified and non-heat-modified forms of OMP C/D.

epitopes are expressed in all phases of growth. The epitopes are expressed by cells grown on chocolate agar and in Mueller-Hinton broth. Surface exposure of the epitopes recognized by antibodies 7D6 and 5E8. To determine whether antibodies 7D6 and 5E8 recognize epitopes that are expressed on the surface of the intact bacterium, aliquots of antibodies were adsorbed with whole bacterial cells, and the resulting adsorbed antibody preparations were tested in immunoblot assays. In this assay, disappearance or decrease in the intensity of the band compared with unadsorbed antibody indicates that the antibody binds to an epitope that is present on the bacterial surface. Figure 3 shows that adsorption of antibody 7D6 with B. catarrhalis 4223 and 5191 causes a decrease in the intensity of the bands compared with unadsorbed antibody. Adsorption of antibody 5E8 causes a complete disappearance of the bands compared with unadsorbed antibody. The specificity of this observation is confirmed by the persistence of the bands in aliquots of the antibodies which were adsorbed with K denitnificans F8841 and E. coli Y1090, two strains which do not express the 7D6 and 5E8 epitopes (lanes d and e). We conclude that the epitope recognized by antibody 5E8 is expressed on the surface of the intact bacterium. The epitope recognized by antibody 7D6 may be partially exposed on the bacterial surface, but it is difficult to draw definitive conclusions regarding the surface exposure of this epitope based on these experiments. To further assess the question of surface exposure, we tested antibodies 7D6 and 5E8 with strains 4223 and 5191 in an indirect immunofluorescence assay. Antibody 5E8 produced prominent fluorescence of cells of both strains (Fig. 4). Antibody 7D6 yielded equivocal results. Three controls included (i) Sp2/0-Agl4 tissue culture supematant in place of antibody (Fig. 4), (ii) buffer in place of antibody, and (iii) K denitnficans F8841 as the organism on the slide. Antibody 7D6 produced inconsistent staining, and differentiation be-

FIG. 4. Immunofluorescence assay with B. catarrhalis 4223. Cells in the left panel were incubated with antibody 5E8 tissue culture supernatant, and cells in the right panel were incubated with SP2 tissue culture supernatant. Both panels were then incubated with fluorescein-conjugated anti-mouse IgM. Original magnification, xl1,000.

tween control and experimental wells was not possible. By contrast, all experiments involving 5E8 produced prominent staining, and the difference between experimental and control assays was obvious. Competitive immunodot assays. To determine whether antibodies 7D6 and 5E8 recognized different epitopes on the C/D protein, we dotted Zwittergent-extracted outer membranes of strain 4223 onto nitrocellulose. After blocking, the antigens were incubated with antibody 5E8 for 8 h. Following washing, the antigens were then incubated with antibody 7D6 for 16 h. Binding of antibody 7D6 to its epitope was then detected with protein A-peroxidase and horseradish peroxidase color developer. Figure 5 shows the results. Preincubation of outer membrane with antibody 5E8 does not inhibit the binding of antibody 7D6 to its epitope. Similarly, the

bu ffer

5E8

SES

7 1: 6

7D6

4 buffer

4 706

5E8

Prot A

Prot A

Prol A

Prot A

4

4

4

4

buffer

7D6

7D6

SE8

5E8

buffer

5E8

7D6

anti 1gM

4

44 IgM anti IgM

anti IgM anti

4

4

4

FIG. 5. Competitive immunodot assays. Purified outer membrane of B. catarrhalis 4223 was dotted onto nitrocellulose and incubated at room temperature with the reagents noted in the figure. The initial incubation of each dot (top of columns) was for 8 h, and the second incubation (second reagent in column) was for 16 h. Protein A-peroxidase (Prot A) and peroxidase-conjugated antimouse IgM (anti IgM) were incubated for 1 h. Immunodots were developed with horseradish peroxidase color developer. Antibody 7D6 is an IgG2a and binds protein A. Antibody 5E8 is an IgM and binds anti-mouse IgM.

808

SARWAR ET AL.

lower portion of Fig. 5 shows that preincubation of outer membrane with antibody 7D6 does not inhibit the binding of antibody 5E8 to its epitope. Several controls are also shown in Fig. 5. These experiments confirm that antibodies 7D6 and 5E8 recognize different epitopes on the C/D protein. Strain and species specificity. To assess the degree of antigenic conservation of the epitopes recognized by antibodies 7D6 and 5E8 among strains of B. catarrhalis, 51 strains from diverse geographic and clinical sources were tested in an immunodot assay. Both antibodies recognized epitopes on all strains tested (Table 1). Immunoblot assays on selected strains showed that the molecular mass of the bands recognized by the antibodies was identical in all strains. A variety of gram-negative species was tested with antibodies 7D6 and 5E8 in immunodot assays, and the results are shown in Table 1. Antibody 7D6 recognized three of three Branhamella strains and one strain of Moraxella atlantae. Antibody 5E8 recognized one strain each of B. caviae and B. cuniculi and one strain of M. osloensis. Immunoblot assays of lysates of these strains revealed that the antibodies recognize bands that are similar in molecular mass but not identical to those in B. catarrhalis strains. In addition, although the lysates contained similar amounts of antigen, the intensity of the signal is substantially less in the other species compared with that in B. catarrhalis. DISCUSSION Recent studies involving the purification and electrophoretic characterization of the outer membrane of B. catarrhalis have identified eight major OMPs, designated A through H (2, 23). The molecular masses of the OMPs of B. catarrhalis demonstrate striking similarities among strains by SDSPAGE (2). Adsorption studies with polyclonal antisera have established that OMP E (-50,000 Da) and OMP G (-28,000 Da) express antigenic determinants on the surface of the intact bacterium (21). Little other information regarding the antigenic structure of individual OMPs of B. catarrhalis is available. In the present study, we developed and used monoclonal antibodies to characterize a major, heat-modifiable OMP. The observation that antibodies 7D6 and 5E8 bind to a somewhat broad band when OMP C/D is in its lowermolecular-mass conformation and to a double band when in its higher-molecular-mass conformation confirms previous studies with two-dimensional gel electrophoresis in which this characteristic of OMP C/D was described (23). Several lines of evidence indicate that antibodies 7D6 and 5E8 recognize different epitopes on the protein. First, in immunoblot assays of heated samples, antibody 7D6 binds consistently to the lower band but inconsistently to the upper band. By contrast, the 5E8 epitope is expressed on both bands in virtually all conditions (Fig. 2). Second, the species specificity of antibodies 7D6 and 5E8 is different (Table 1). Third, the surface exposure of the two antibodies is different (Fig. 3 and 4). Finally, preincubation of immunodot assays with one antibody does not block the binding of the other antibody (Fig. 5). The observations that the epitopes recognized by antibodies 7D6 and 5E8 are expressed on two bands in the heatmodified form of OMP C/D are consistent with two hypotheses. First, OMP C/D is a single protein that has two different stable conformations under the conditions used in these experiments. The second possibility is that OMP C/D actually represents two proteins encoded by different genes

INFEC-F. IMMUN.

and that the antibodies recognize epitopes on both proteins. This would be similar to the observation in E. coli that the three porin molecules, OmpF, OmpC, and PhoE, each encoded by its own gene, share extensive nucleotide homology and demonstrate antigenic cross-reactivity (18). Studies are under way to clone the gene(s) that encodes OMP C/D. This approach will resolve the question of whether OMPs C and D are the product of one or more genes. The specificity of these two monoclonal antibodies for strains of the species B. catarrhalis has potential implications regarding identification of the bacterium in clinical samples. B. catarrhalis is indistinguishable from Neisseria species by colonial morphology. A variety of Neisseria species are present as part of the normal flora of the human upper respiratory tract. Therefore, to identify B. catarrhalis in a clinical sample from the upper respiratory tract, it is necessary to test individual colonies to distinguish between commensal Neisseria species and B. catarrhalis. Many bacteriology laboratories in hospitals do not attempt specifically to identify B. catarrhalis in respiratory tract secretions. Antibodies 7D6 and 5E8 recognize all B. catarrhalis strains, and the antibody is highly specific for B. catarrhalis. Furthermore, the epitope is stably expressed. Therefore, these antibodies, or DNA probes corresponding to their epitopes, could represent the basis of an assay to identify B. catarrhalis in clinical samples. Antibody 5E8 recognizes an epitope that is expressed on the surface of the intact bacterium and that is conserved among all strains of B. catarrhalis tested. Indeed, the strikingly prominent signal in immunofluorescence assays (Fig. 4) and the complete adsorption of antibody 5E8 by intact bacterial cells (Fig. 3) suggest that the epitope is abundantly expressed on the bacterial surface. This observation has important implications with regard to identifying potential vaccine antigens. The presence of a conserved, surface-exposed epitope suggests that OMP C/D generates protective antibody. Future studies will evaluate the role of the C/D protein in the human immune response to infection and assess the potential role of antibody to OMP C/D in protection from infection. In summary, the present study establishes that OMP C/D is present in all strains of B. catarrhalis and has a molecular mass that is identical in all strains tested. Based on SDSPAGE of purified outer membrane, OMP C/D appears to be the most abundant protein in the outer membrane (Fig. 1). This study further shows that the protein expresses at least two epitopes that are antigenically conserved among strains of the species. The epitopes are highly specific for B. catarrhalis, being absent from a wide variety of other gram-negative bacteria. Expression of both epitopes is independent of growth phase and growth media. Finally, one of these epitopes is expressed on the surface of the intact bacterium. Molecular studies of OMP C/D will focus on determination of the function of the protein for the bacterium and further elucidation of the antigenic structure of this major outer membrane antigen of B. catarrhalis. This will lead to an understanding of the role of this OMP in pathogenesis and its potential role as a vaccine antigen. ACKNOWLEDGMENTS This work was supported by Public Health Service grant AI28304 (T.F.M.) from the National Institute of Allergy and Infectious Diseases. Our thanks to Maureen Wirth and Richard Karalus for expert technical assistance; to Jan Brentjens for assistance with the immunofluorescence; to Karen Chiavetta for performing the photography;

VOL. 60, 1992

OUTER MEMBRANE PROTEIN C/D OF B. CATARRHALIS

to Adeline Thurston for preparing the manuscript; and to Howard Faden, Diane Dryja, Joel Bernstein, Steven Berk, Carl Marrs, and Daniel Musher for providing isolates.

REFERENCES 1. Apicella, M. A., K. C. Dudas, A. Campagnari, P. Rice, J. M. Mylotte, and T. F. Murphy. 1985. Antigenic heterogeneity of lipid A of Haemophilus influenzae. Infect. Immun. 50:9-14. 2. Bartos, L. C., and T. F. Murphy. 1988. Comparison of the outer membrane proteins of 50 strains of Branhamella catarrhalis. J. Infect. Dis. 158:761-765. 3. Bluestone, C. D. 1986. Otitis media and sinusitis in children: role of Branhamella catarrhalis. Drugs 31(Suppl. 3):S132-S141. 4. Brorson, J.-E., A. Axeisson, and S. E. Holm. 1976. Studies on Branhamella catarrhalis (Neisseria catarrhalis) with special reference to maxillary sinusitis. Scand. J. Infect. Dis. 8:151155. 5. Catlin, B. W. 1990. Branhamella catarrhalis: an organism gaining respect as a pathogen. Clin. Microbiol. Rev. 3:293-320. 6. Chapman, A. J., D. M. Musher, S. Jonsson, J. E. Clarridge, and R. J. Wallace. 1985. Development of bactericidal antibody during Branhamella catarrhalis infection. J. Infect. Dis. 151: 878-882. 7. Christensen, J. J., and B. Bruun. 1985. Bacteremia caused by a beta-lactamase producing strain of Branhamella catarrhalis. Acta Pathol. Microbiol. Immunol. Scand. Sect. B 93:273-275. 8. Craig, D. B., and P. A. Wehrle. 1983. Branhamella catarrhalis septic arthritis. J. Rheumatol. 10:985-986. 9. Evans, F. O., Jr., J. B. Sydnor, W. E. C. Moore, G. R. Moore, J. L. Manwaring, A. H. Brill, R. T. Jackson, S. Hanna, J. S. Skaar, L. V. Holdeman, G. S. Fitz-Hugh, M. A. Sande, and J. M. Gwaltney, Jr. 1975. Sinusitis of the maxillary antrum. N. Engl. J. Med. 293:735-739. 10. Gray, L. D., R. E. Van Scoy, J. P. Anhalt, and P. K. W. Yu. 1989. Wound infection caused by Branhamella catarrhalis. J. Clin. Microbiol. 27:818-820. 11. Guthrie, R., K. Bakenhaster, R. Nelson, and R. Woskobnick. 1988. Branhamella catarrhalis sepsis: a case report and review of the literature. J. Infect. Dis. 158:907-908. 12. Hager, H., A. Verghese, S. Alvarez, and S. L. Berk. 1987. Branhamella catarrhalis respiratory infections. Rev. Infect. Dis. 9:1140-1149. 13. Hiroshi, S., E. J. Anaissie, N. Khardori, and G. P. Bodey. 1988. Branhamella catarrhalis septicemia in patients with leukemia. Cancer 61:2315-2317. 14. Kennett, R. H. 1979. Cell fusion. Methods Enzymol. 58:345359. 15. Kovatch, A. L., E. R. Wald, and R. H. Michael. 1983. Betalactamase-producing Branhamella catarrhalis causing otitis media in children. J. Pediatr. 102:261-264. 16. Loeb, M. R. 1984. Immunoblot method for identifying surface components, determining their cross-reactivity, and investigating cell topology: results with Haemophilus influenzae type b. Anal. Biochem. 143:196-204. 17. McLeod, D. T., F. Ahmad, M. J. Croughan, and M. A. Calder. 1986. Bronchopulmonary infection due to B. catarrhalis. Clini-

18. 19.

20. 21. 22.

23. 24.

25. 26.

27.

28. 29.

30. 31.

32.

33. 34.

809

cal features and therapeutic response. Drugs 31(Suppl. 3):109112. Mizuno, T., M. Y. Chou, and M. Inouye. 1983. A comparative study on the genes for three porins of the Escherichia coli outer membrane. J. Biol. Chem. 258:6932-6940. Murphy, T. F. 1989. The surface of Branhamella catarrhalis: a systematic approach to the surface antigens of an emerging pathogen. Pediatr. Infect. Dis. J. 8:S75-S77. Murphy, T. F. 1990. Studies of the outer membrane proteins of Branhamella catarrhalis. Am. J. Med. 88:41S-45S. Murphy, T. F., and L. C. Bartos. 1989. Surface-exposed and antigenically conserved determinants of outer membrane proteins of Branhamella catarrhalis. Infect. Immun. 57:2938-2941. Murphy, T. F., L. C. Bartos, A. A. Campagnari, M. B. Nelson, and M. A. Apicella. 1986. Antigenic characterization of the P6 protein of nontypeable Haemophilus infiuenzae. Infect. Immun. 54:774-779. Murphy, T. F., and M. R. Loeb. 1989. Isolation of the outer membrane of Branhamella catarrhalis. Microb. Pathog. 6:159174. Nicotra, B., M. Rivera, J. I. Luman, and R. J. Wallace. 1986. Branhamella catarrhalis as a lower respiratory tract pathogen in patients with chronic lung disease. Arch. Intern. Med. 146:890893. Ninane, G., J. Joly, and M. Kraytman. 1978. Bronchopulmonary infection due to Branhamella catarrhalis: 11 cases assessed by transtracheal puncture. Br. Med. J. 1:276-278. O'Neill, J. H., and P. W. Matheison. 1987. Meningitis due to Branhamella catarrhalis. Aust. N.Z. J. Med. 17:241-242. Shurin, P. A., C. D. Marchant, C. H. Kim, G. F. Van Hare, C. E. Johnson, M. A. Tutihasi, and L. J. Knapp. 1983. Emergence of beta-lactamase-producing strains of Branhamella catarrhalis as important agents of acute otitis media. Pediatr. Infect. Dis. 2:34-38. Soto-Hernandez, J. L., S. Holtsclaw-Berk, L. M. Harvill, and S. L. Berk. 1989. Phenotypic characteristics of Branhamella catarrhalis strains. J. Clin. Microbiol. 27:903-908. Srinivasan, G., M. J. Raff, W. C. Templeton, S. J. Givens, R. C. Graves, and J. C. Melo. 1981. Branhamella catarrhalis pneumonia. Report of two cases and review of the literature. Am. Rev. Respir. Dis. 123:553-555. Tinkelman, D. G., and H. J. Silk. 1989. Clinical and bacteriologic features of chronic sinusitis in children. Am. J. Dis. Child. 143:938-942. Van Hare, G. F., P. A. Shurin, C. D. Marchant, N. A. Cartelli, C. E. Johnson, D. Fulton, S. Carlin, and C. H. Kim. 1987. Acute otitis media caused by Branhamella catarrhalis: biology and therapy. Rev. Infect. Dis. 9:16-27. Wald, E. R., C. Byers, N. Guerra, M. Casselbrant, and D. Beste. 1989. Subacute sinusitis in children. J. Pediatr. 115:28-32. Wald, E. R., G. J. Milmoe, A. Bowen, J. Ledesma-Medina, N. Salamon, and C. D. Bluestone. 1981. Acute maxillary sinusitis in children. N. Engl. J. Med. 304:749-754. West, M., S. L. Berk, and J. K. Smith. 1982. Branhamella catarrhalis pneumonia. South. Med. J. 75:1021-1023.