Aug 17, 1987 - Campylobacter jejuni is one of the commonest pathogens of acute bacterial ... colonization by C. jejuni might be attained by the anti-flagellar ...
Microbiol. Immunol. Vol. 31 (12), 1161-1171, 1987
Protection against Campylobacterjejuni Infection in Suckling Mice by Anti-Flagellar Antibody Yuji
UEKI,*
Masao Department
Akiko
UMEDA,
MITSUYAMA,•‚
of Bacteriology,
(Accepted
Faculty
and
of Medicine,
for
publication,
Shull
FUJIMOTO
Kazunobu Kyushu
AMAKO University,
August
17,
Fukuoka,
Fukuoka
812
1987)
Abstract We obtained two monoclonal antibodies of IgM class and IgA class of immunoglobulin prepared from mouse spleen cells immunized with crude flagellar preparation, and a polyclonal antibody raised against purified flagellin monomer of Campylobacter jejuni in a rabbit. The specificity of the reaction of these antibodies for flagellar filament was confirmed by Western blotting and by immunoelectron microscopy. These antibodies caused agglutination of the bacteria and inhibited the motility of the bacteria. When a strain of C. jejuni was treated with IgM class monoclonal antibody before being inoculated into suckling mice , it reduced colonization of the intestinal tract by this bacteria. Inhibition of the colonization by IgA class monoclonal antibody was less effective than that of IgM class, and the polyclonal antibody consisting mostly of IgG class immunoglobulin was without effect.
Campylobacterjejuni is one of the commonest pathogens of acute bacterial gastroenteritis (3). There are several reports suggesting that LPS (13, 19, 20), enterotoxin (9, 22), invasiveness (6), and flagella (13, 15, 18) may be virulence factors of the bacteria. We reported that the flagellum or the movement of the bacterium was important for colonization of the intestinal tract by C. jejuni (15). In infections with some other bacteria, prevention can be attained by specific antibodies against the colonization factors (14, 16). Therefore, it can be assumed that inhibition of the intestinal colonization by C. jejuni might be attained by the anti-flagellar antibody. Various attempts have been made to study the effects of active and passive immunization by flagellar proteins against C. jejuni (5, 17). The effectiveness of anti-flagellar antibody has not been clearly shown. On the other hand, the surface antigens of C. jejuni were analyzed by gel electrophoresis and four major antigens were detected (12). Among these proteins , the 63,000 dalton protein proved to be an antigen of the flagellar filament. Human sera obtained after convalescence from C. jejuni infection contained various classes of specific antibodies against these proteins (1, 2), but little is known of the protective roles of these antibodies in vivo. ≠ Present
Niigata,
Niigata
address:
Department
of
Bacteriology,
951. 1161
Faculty
of
Medicine
, Niigata
University,
1162
Y. UEKI
ET AI,
Based on this background, two monoclonal antibodies and a polyclonal antiserum against flagellar protein of C. jejuni were produced and their effects on intestinal colonization were examined. MATERIALS
AND
METHODS
Bacteria. The Campylobacter jejuni FUM158432 used in this study was obtained from Fukuoka University Hospital. The origin and details of the culture conditions of this strain were as described (15). Basically, Brucella agar plates (Difco Laboratories, Detroit, Mich., U.S.A.) supplemented with 10% (v/v) defibrinated sheep blood (Nippon Bio-Supp. Center) (Brucella Blood Agar: BBA) was used in a catalystfree anaerobic jar of the GasPak anaerobic system (BBL). Samples from the intestinal tract of mice were cultured at 42 C on BBA made selective for C.jejuni by the addition of vancomycin 10 mg/liter, polymyxin B 2,500 IU/liter and trimethoprim 5 mg/liter. Purificationof flagella. C. jejuni FUM158432 grown on BBA plates were collected by scraping, suspended in 50 mm Tris-HC1 buffer (pH 7.5), and then treated with sonic vibration for a few minutes in a Branson Sonifier Cell Disruptor, model 200 (Branson Sonic Power Co., Danbury, Conn., U.S.A.), at room temperature. Bacteria and detached flagella were separated by three cycles of differential centrifugation at 5,000 xg for 30 min and 100,000 x g for 60 min. The final pellet was suspended in a Tris-HC1 buffer. Flagellar protein was purified from this crude flagellar preparation by ion exchange column chromatography, gel filtration, and acid dissociation at pH 2.0. Productionof anti-flagellar antibodies. A modification of the method of Koehler and Milstein (10) was used to produce monoclonal antibody. BALB/c mice (6 weeks old, male) were immunized subcutaneously with 160 ,ug of crude flagella in complete Freund's adjuvant. Two weeks later, a second immunization was given with 80 iug of the same antigen in incomplete Freund's adjuvant. At ten days after the second injection, the mice were given intravenously 90 ,ug of the soluble antigen, as a booster. Three days later, the spleens were removed and lymphocytes fused with P3U1 myeloma cells in the presence of 35% (w/v) polyethylene glycol 1000. The fused cells were cultured in modified Dulbecco's minimal essential medium (GIBCO Laboratories, Life Technologies, Inc., Grand Island, N.Y., U.S.A.). The presence of flagellum-specific antibody in the culture supernatant of the hybridomas was tested by ELISA with the crude flagella as a solid phase. Positive hybridomas were cloned by the limiting dilution method. Culture supernatants of the hybridomas were used in the following experiments and P3U1 culture supernatant was used as a control. Polyclonal anti-flagellar antiserum was raised by immunizing New Zealand White rabbits (3 kg, female) with purified flagellar protein. Rabbits were immunized subcutaneously with 2 mg of purified flagellar protein in complete Freund's adjuvant. Seven days later, a second immunization was carried out with the same dose of the antigen in incomplete Freund's adjuvant. After two further intravenous booster injections with 1 mg of soluble flagellar protein, the rabbits were
PROTECTION
AGAINST
C. JEJUNI
INFECTION
1163
bled. Separated sera were inactivated at 56 C for 30 min, dialyzed against phosphate-buffered saline (PBS), pH 7.8, and stored at -70 C. Normal rabbit serum dialyzed against PBS was used as a control in the following experiments. Sodium dodecylsulfate-polyacrylamide gel electrophoresis(SDS-PAGE). SDS-PAGE was done according to the method of Laemmli (11), with some modification. The concentration of acrylamide used to stack and separate was 4.0% (w/v) and 12.5% (w/v), respectively. Samples to be analyzed were boiled for 5 min in a sample buffer (62.5 mm Tris-HC1 pH 6.8, 10% (w/v) glycerol, 5% (v/v) 2-mercaptoethanol, 2.3% (w/v) SDS, 0.5% (w/v) bromophenol blue). Electrophoresis was done on a slab gel at 10 mA per gel constant current until the stacking dye reached the separating gel, at which time the constant current was increased to 20 mA per gel. After electrophoresis, the gel was stained with 0.1 % (w/v) Coomassie brilliant blue dye. Immunoblotting. Electrophoresed flagellar proteins were transferred electrophoretically from the polyacrylamide gel to a nitrocellulose sheet (membrane filters, 0.45 earn, Schleicher & Schuell, Dassel, West Germany) by the method of Towbin et al (23). After transfer, the sheet was blocked by soaking in 0.5% (w/v) bovine serum albumin-containing PBS-Tween (phosphate-buffered saline, pH 7.8, supplemented with 0.05% (v/v) Tween 20) and then washed three times with PBS-Tween. The sheet was incubated with anti-flagellar antibody at room temperature for 90 min, and then washed. The reactivity of antibody probe to the flagellar proteins was detected by incubating the sheet with peroxidase-conjugated goat anti-mouse or rabbit immunoglobulins (Zymed Laboratories, Inc., San Francisco, Calif., U.S.A.) at room temperature for 90 min. After three washings with PBS-Tween, the sheet was soaked in a solution of 4-chloro-l-naphtol (40 mg substrate in 100 ml 50 mm Tris-HC1, pH 7.5, to which 35 1il of 30% (v/v) hydrogen peroxide had been added). Light microscopy. To determine the effects of these antibodies on the movement of C. jejuni, the motility of the bacteria preincubated with anti-flagellar antibodies was observed by a wet mount slide. The strain of C.jejuni cultured on a BBA plate was suspended in one ml of Brucella broth. One drop of the suspension was placed on a clean slide glass and the bacterial movement in the drop was observed by a wet mount slide. To observe the effects of antibodies on the bacterial movement, an equal volume of antibody solution and the bacterial suspension were mixed, kept for 10 min at room temperature, and the bacterial movement was observed by a wet mount slide. Electronmicroscopy. To confirm the specificity of anti-flagellar polyclonal antibodies, the immuno-gold method was used. Protein A-colloidal gold was prepared by the method of Frens (7) and Roth et al (21). The specificity of monoclonal antibodies was confirmed by using colloidal gold-labelled antimouse IgM or IgA (Janssen Life Sciences Products, Beerse, Belgium). The specimens were suspended in a small volume of 2% (w/v) ammonium acetate solution and examined by a negative staining method. Enzyme-linkedimmunosorbentassay (ELISA). Wells of polyvinylchloride microtitration plates were coated with 50 ,u1of flagellar antigen in 0.015 Msodium carbonate buffer (coating buffer pH 9.6) (2 ,ug/well) at 4 C for 16 hr, and blocked with 0.5%
1164
Y.
UEKI
ET AL
(w/v) bovine serum albumin in coating buffer. After washing with PBS-Tween, 2-fold serial dilutions of anti-flagellar antibodies were made with PBS-Tween and 50 ,u1 was added to each well. The plates were incubated at room temperature for 90 min. The reactivity of antibody with the antigen was detected with peroxidase-conjugated goat anti-mouse or rabbit immunoglobulins (Zymed Laboratories). After 90 min, binding of the second antibody was detected with 50 ,u1per well of 40 mg orthophenylenediamine in 100 ml phosphate citrate buffer to which was added 35 ,u1 of 30% (v/v) hydrogen peroxide. After the addition of 2N H2SO4, the absorbance of the reaction product was measured at a wavelength of 490 nm by an Elisaminireader (Dynatech Laboratories Inc., Alexandria, Va., U.S.A.). Wells that showed a higher absorbance than that of control wells were considered positive and antibody titers were expressed as reciprocal of the highest dilution. P3U1 culture supernatant or normal rabbit serum were used as the control. Agglutinationtest. Agglutination titers of anti-flagellar antibodies were assayed by the microtiter method. Fifty ,u1 of bacterial suspension of C. jejuni FUM158432 in PBS (approximately 5 x 108 CFU/ml) was added to each of 96-wells V-shaped microtiter plates, after which 50 ,u1 of 2-fold serial diluted antibodies was added. After being incubated at 37 C for 2 hr, the plates were kept overnight at 4 C and the agglutination was read. Oral challenge. The method of oral inoculation of C. jejuni was essentially the same as described (15). One- to four-day-old suckling mice of the ddY strain were used. The bacterial inoculum treated with anti-flagellar antibodies was prepared as follows. C. jejuni FUM 158432 stored at -70 C was spread on BBA plates and then incubated at 37 C for 24-48 hr. After an additional subculture on BBA, the bacteria were collected and suspended in Brucella broth. The bacterial suspension and the same volume of anti-flagellar antibodies whose agglutination titer were adjusted to 512 were mixed and the preparation was then incubated at 37 C. The viable count in the suspension was determined by diluting and plating on BBA plates. The mixture was inoculated in a volume of 50 duiinto the stomach of suckling mice through a fine polyethylene tube attached to the tip of a 23-gauge needle fitted to a 1 ml syringe. At various times after inoculation, mice were killed and the colon was removed. The number of bacteria colonized in the colon of these mice was counted by plating the homogenate of colon prepared by a Potter-type glass homogenizer, on selective BBA plates. RESULTS
Propertiesof Anti-Flagellar Antibodies From hundreds of hybridomas, two clones, 1AC12 and 9EE1, with a high antibody titer for flagellar antigen in ELISA were obtained. I AC 12 produced IgM class and 9EE1 produced IgA class of immunoglobulin. Antibody titers of these two monoclonal antibodies and that of polyclonal antiserum were in the range of 4098 to 16384 in ELISA, and 512 to 2048 in the agglutination method, as listed in Table 1. Agglutination titer of these anti-flagellar antibodies was adjusted by
PROTECTION Table
1.
AGAINST Properties
C. JEJUNI of anti-flagellar
INFECTION
116.5
antibodies
Fig. 1. SDS-PAGE analysis of flagellar protein of C. jejuni FUM158432. Lane 1, molecular weight marker; lane 2, crude flagella after differential centrifugation; lane 3, purified flagellar protein.
Fig. 2. Immunoblotting of purified flagellar protein with anti-flagellar antibodies. Lane 1, purified protein stained with amide black; lane 2, blotted with polyclonal antiserum diluted 1: 200 with PBS-Tween; lane 3, blotted with IAC12 monoclonal antibody; lane 4, blotted with 9EE I monoclonal antibody. Peroxidase-conjugated second antibody was diluted 1: 800 with PBS-Tween.
1166
Y. LIEKI
ET Al
Fig. 3. Electron micrographs of C. jejuni FUM158432. a, Treated with 1AC12 monoclonal antibody and colloidal gold-labeled anti-mouse IgM. b, Treated with anti-flagellar antiserum and protein A-gold.
dilution with PBS to 512 and used in the following experiments. Specificityof Anti-Flagellar Antibodies Specificity of antibodies for flagellar protein was tested by immunoblotting and by immunoelectron microscopy. An analysis of purified flagellar protein of C. jejuni by SDS-PAGE revealed a single polypeptide with an apparent molecular mass of 63 kd (Fig. 1, lane 3). In the immunoblotting method, both monoclonal antibodies and the polyclonal antibody reacted specifically with this peptide (Fig. 2), but not with others. The specificity of antibodies was also confirmed by immunoelectron microscopy, using an immuno-gold technique. Colloidal gold particles were found only on the flagellar filaments and not on the surface materials of the cells tested with 1AC12 monoclonal antibody (Fig. 3a) and the polyclonal antibody
PROTECTION
Fig.
4. bodies. mal
Percent
viability
Symbols: •œ, rabbit
serum
of 1AC12
dialyzed
AGAINST
C. jejuni
FUM
monoclonal against
C. JEJUNI
158432
after
antibody; PBS;
A,
0, Brucella
INFECTION
incubation P3U1
1167
with culture
anti-flagellar supernatant; •¬,
antinor-
broth.
(Fig. 3b). Similar findings were observed using the 9EE1 monoclonal antibody (data not shown). Ejects of Anti-Flagellar Antibodieson Bacterial Movement C. jejuni showed active and random movement in the wet mount slide. This typical movement was greatly inhibited after incubation with anti-flagellar antibodies (1 x 108 CFU/mixture). Some of the bacteria formed aggregates consisting of dozens of cells and other single cells showed no active movement. The formation of such aggregation was less frequent when the bacterial concentration in the suspension was lower than 1 x 105 CFU/mixture. The inhibitory effect on motility of the 1AC12 monoclonal antibody was greater than that of the 9EE1 monoclonal antibody and of the polyclonal antibody. Effectsof Anti-FlagellarAntibodieson Viabilityof C. jejuni Bacteria suspended in Brucella broth (1 x 108 CFU/ml) were mixed at 37 C with either monoclonal antibodies, polyclonal antiserum, P3U 1 culture supernatant, normal rabbit serum, or Brucella broth. The number of viable bacteria in the mixture was counted at various times during the incubation. The number of viable cells incubated with 1AC12 monoclonal antibody increased slightly in the initial 30 min and then decreased gradually. The same reducing patterns were seen in the P3U1 culture supernatant, normal rabbit serum and Brucella broth (Fig. 4). Similar results were obtained with the 9EE1 monoclonal antibody and the polyclonal antibody. Thus, the time for preincubation of bacteria with anti-flagellar antibodies for oral challenge was limited to 30 min. Protectionof Colonizationin the Colonof Suckling Mice by C. jejuni by Anti-Flagellar Antibodies To determine optimal conditions for the maximum colonization in control mice,
1168
Y. UEKI
Fig.
5. P3U1 ▲,
Colonization culture
normal
in
supernatant rabbit
Table 2.
colon
of
suckling
or normal
rabbit
ET Al.
mice serum.
by
C. jejuni Symbols: •œ
FUM
158432 , P3U1
preincubated culture
supernatant
with :
serum.
Inhibition of the colonization of the colon of suckling mice by C. jejuni by anti-flagellar antibodies
C.jejuni cells suspended in Brucella broth were mixed with the same volume of P3U 1 culture supernatant or normal rabbit serum at 37 C for 30 min, and 50 ,u1 of the mixture was inoculated into the stomach of suckling mice (Fig. 5). On days 2, 4, and 7 after the challenge, the mice were killed and the number of bacteria colonized
PROTECTION
AGAINST
C. JEJUNI
INFECTION
1169
in the colon was counted. The maximum level of colonized bacteria was recovered on day 4. From this observation the colonization assay was usually done on samples taken 4 days after the challenge. When the bacterial suspension at the concentration lower than 1.9 x 104 CFU/mouse were treated with 1AC 12, an IgM isotype monoclonal antibody, no colonization was observed in the 20 mice inoculated. Intestinal colonization occurred at a bacterial concentration higher than 1.9 x 105 CFU/mouse. At the concentration of 1.9 x 106 CFU/mouse, the colonization ocurred in. 100% of the mice. Inhibition of the colonization was less effective in mice inoculated with the bacterial suspension treated with the other monoclonal antibody 9EE1, an IgA class antibody (Table 2, Exp. 1). At the bacterial concentration higher than 2.0 x 105 CFU/mouse, colonization ocurred in 100% of the mice. On the other hand, the polyclonal antibody did not inhibit the colonization (Table 2, Exp. 2). Even at the bacterial concentration of 3.5 x 101 CFU/mouse, C. jejuni colonized in about half of the mice inoculated, and at the concentration of 6.0 x 102 CFU/mouse no protective effect was observed. DISCUSSION
Motility is thought to be a virulence factor for some of the flagellated bacteria such as Vibriocholeraeand Pseudomonasaeruginosa,and protection of animals from the infection by these bacteria is afforded by immunizing the animals with the flagellar antigen preparation (4, 8, 24). In Campylobacterinfection, flagellum is reported to play an important role in the pathogenesis of acute gastroenteritis (15, 18). In this study, we prepared two monoclonal antibodies and a polyclonal antibody raised against the highly purified flagellar filament of C.jejuni and examined their protective effects on the intestinal infection of suckling mice by C. jejuni. These antibodies could agglutinate the bacteria and inhibit their motility. The inhibitory effects on the colonization of intestinal tract by C. jejuni differed, depending on the immunoglobulin class of these antibodies. Monoclonal antibody of IgM class was more effective than that of IgA class, and polyclonal antibody which consisted mostly of IgG class immunoglobulin had little effect on the colonization. The mechanisms of the protection afforded by anti-flagellar antibody are not well understood. Since the flagellated but non-motile strain of C. jejuni could not colonize in the intestinal tract, the motility is thought to be important for the colonization (15). Regarding the mechanisms of the inhibitory effect of the antibody, the following two possibilities can be considered. First, inhibition of the motility by the attachment of the antibody to the flagellar filaments may result in inhibition of the colonization. Newell reported that anti-flagellar monoclonal antibodies which could neither agglutinate the bacteria nor inhibit motility could not inhibit the colonization of C.jejuni in the intestinal tract of suckling mice (17). In this study we showed that all the antibodies used inhibited the bacterial movement and that the effects differed to some extent, depending on the immunoglobulin class of the antibody. Concerning the inhibition of colonization, however, the antibody of IgM class was more effective than that of IgA or IgG. Here the mechanical effects
1170
Y. UEKI
ET AL
on the movement caused by anti-flagellar antibodies have to be given attention. The IgM class antibody is of a larger molecular size than other classes of antibodies. The inhibitory effect on the colonization by the antibodies obtained in this study was proportional to the molecular size. Second, it was reported that the flagellar filament of C. jejuni was essential not only for motility but also for adherence to the intestinal mucosa (5). If our monoclonal antibody of IgM class recognizes a specific receptor-binding site, this antibody might inhibit the colonization more effectively than the other monoclonal antibody of IgA class. The polyclonal antibody obtained in this study, which presumably contained the antibody directed to the receptor-binding sites together with the other flagellar antigens was all but ineffective in inhibiting colonization. We previously reported that flagellar mutant M14, which had a morphologically similar flagellum but lacked motility, could not colonize in suckling mice (15). This mutant showed almost the same agglutinability with anti-flagellar antibodies used in this study. This evidence suggested that motility of this bacterium was essential for colonization, but the possibility that the flagella acted as adhesin was low. At present we should like to emphasize that the first possibility is the mechanism of the protection by transfer of specific antibody. We are intending to make other monoclonal antibodies of IgG class. By comparing their effects on motility and colonization, the mechanism of this protection afforded by anti-flagellar antibody may be elucidated. We thank M. Ohara for reviewing this manuscript and Seiya Jingushi for his technical assistance. REFERENCES
1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11)
Blaser, M.J., and Dunkan, D. J. 1984. Human serum antibody response to (ampylobacterjejuni infection as measured in an enzyme-linked immunosorbent assay. Infect. Immun. 44: 292-298. Blaser, M. J., Hopkins, J.A., and Vasil, M.L. 1984. Campylobacter jejuni outer membrane proteins are antigenic for humans. Infect. Immun. 43: 986-993. Blaser, M.J., and Reller, L.B. 1981. Campylobacterenteritis. N. Engl. J. Med. 305: 1444-1452. Craven, R.C., and Montie, T.C. 1981. Motility and chemotaxis of three strains of P. aeruginosa used for virulence studies. Can. J. Microbiol. 27: 458-460. Dolby, J.M., and Newell, D.C. 1986. The protection of infant mice from colonization with (ampylobacter jejuni by vaccination of the dams. J. Hyg. 96: 143-151. Duffy, M.C., Benson, J.B., and Rubin, S. J. 1980. Mucosal invasion in (ampylobacier enteritis. Am. J. Clin. Pathol. 73: 706-708. Frens, G. 1973. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nature Phys. Sci. 241: 20-22. Holder, LA., Wheeler, R., and Montie, T.C. 1982. Flagellar preparations from Pseudomonas aeruginosa: animal protection studies. Infect. Immun. 35: 276-280. Klipstein, F.A., and Engert, R.F. 1985. Immunological relationship of B subunit of (ampylobacter jejuni and Escherichia coli heat-labile enterotoxin. Infect. Immun. 48: 629-633. Koehler, G., and Milstein, C. 1975. Continuous cultures of fused cells secreting of predicted specificity. Nature 256: 493-497. Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685.
PROTECTION
12)
Logan,
S.M.,
and
Campylobacter 13)
McSweegan, jejuni
14)
E.,
adhesins.
Morgan,
R.L.,
purified
Campylobacter
Walker,•@
or
Immun. A., J.
R.I.
Gen.
K99
pili:
22:
771-777.
and
Amako,
Microbiol.
INFECTION
identification
of
1171
surface
protein
antigens
of
675-682.
1986.
Moon,
C. JEIUNI
Molecular
42:
enterotoxigenic
987
Umeda, jejuni.
1983.
53:
R.E.,
against
Infect. T.,
Immun.
Isaacson,
dams
Morooka,
T. J. Immun.
and
pigs
with
Trust, Infect.
Infect.
of suckling
challenge. 15)
jejuni.
AGAINST
Identification
and
characterization
of
two
Campylobacter
141-148. H.W.,
Brinton,
Escherichia protection
K. 131:
C.C.,
correlates
1985.
and
coli-induced
Motility
To, diarrheal
with
as
an
pilus
intestinal
C.-C.
1978. disease
homology
colonization
Immunization by of
vaccinating vaccine
factor
and
for
1973-1980.
16) Nagy, B., Moon, H.W., Isaacson, R.E., To, C.-C., and Brinton, C.C. 1978. Immunization of suckling pigs against enteric enterotoxigenic Escherichiacoli infection by vaccinating darns with purified pili. Infect. Immun. 21: 269-274. 17) Newell, D.G. 1986. Monoclonal antibodies directed against the flagella of Campylobacter,jejuni: production, characterization and lack of effect on the colonization of infant mice. J. Hyg. 96: 131-141. 18) Newell, D.G., McBride, H., and Dolby, J. 1983. The significance of flagella in the pathogenesis of Campylobacterjejuni, p. 109. In Pearson, A.D., Skirrow, M.B., Rowe, B., Davies, J.R., and Jones, D.M. (eds), Campylobacter II, Public Health Laboratory Service, London. 19) Perez, G.P., and Blaser, M.J. 1985. Lipopolysaccharide characteristics of pathogenic Campylobacters. Infect. Immun. 47: 353-359. 20) Perez, G.P., Hopkins, J.A., and Blaser, M.J. 1985. Antigenic heterogeneity of lipopolysaccharides from Campylobacterjejuni and Campylobacterfetus. Infect. Immun. 48: 528-533. 21) Roth, J., Bendayan, M., and Orci, L. 1978. Ultrastructural localization of intracelullar antigens by use of protein A-gold complex. J. Histochem. Cytochem. 26: 1074-1081. 22) Ruiz-Palacios, G., Torres, N.1., Escamilla, E., Ruiz-Palacios, B.R., and Tamayo, J. 1983. Cholera-like enterotoxin produced by Campylobactercoli and Campylobacterjejuni strains. Lancet ii: 250-252. 23) Towbin, H., Staelin, T., and Gordon, T. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl. Acad. Sci. U.S.A. 76: 4350-4354. 24) Yancey, R. J., Willis, D.L., and Berry, L. J. 1979. Flagella-induced immunity against experimental cholera in adult rabbits. Infect. Immun. 25: 220-228. (Received for publication,
May 15, 1987)
