Microbiol. Immunol. Vol. 36 (8), 791-801, 1992
Production of a Monoclonal Antibody That Recognizes the Lipopolysaccharide of a Campylobacter-Like Organism Steven Kok Chee KUAN,* Peter John COLOE, and Malcolm Raymond ALDERTON Departmentof AppliedBiologyand Biotechnology,Royal MelbourneInstitute of G.P.O.
(Accepted for publication,
May 6, 1992)
Abstract A monoclonal antibody was produced to a Campylobacter-likeorganism (RMIT 32A) which was isolated from the terminal ileum of a pig with proliferative enteritis. Isotyping of the antibody revealed that it was an IgG2a with kappa light chains. Immunoblots using the antibody against proteinase-K-treated whole cell lysates of RMIT 32A, a selection of Campylobacterspecies and other enteric bacteria showed that the antibody was specific for RMIT 32A and was directed against the lipopolysaccharide. This antibody can be used for the specific detection of RMIT 32A.
Porcine proliferative enteritis (PPE) is a disease of pigs characterized by adenomatosis of mucosal epithelial cells in the ileum (18, 23, 43, 45). The disease appears to present in several pathological forms such as necrotic enteritis, proliferative hemorrhagic enteropathy, regional ileitis, intestinal adenomatosis and terminal enteritis. However, these diseases are commonly reported as separate entities (18, 20, 30, 42, 44, 45). The causative agent of the disease has not been fully identified but several Campylobacter species have been reported to be associated with the disease (6, 8, 9, 18, 23, 29, 35, 36, 43). Histological examination of affected ilea has revealed the presence of Campylobacterand Campylobacter-likeorganisms that are not membrane bound but present within the apical cytoplasm of intestinal cells (24-26). Several Campylobacter species have been isolated from intestinal lesions and feces of affected pigs. C. sputorumsubspecies mucosalisoriginally described by Lawson and co-workers (18, 19), has been suggested as a causative agent of PPE. Other researchers have since reported the isolation of C. sputorumsubsp mucosalisfrom intestinal lesions of pigs with PPE (23, 26, 35, 36, 42). Two other Campylobacterspecies, C. hyointestinalisand C. coli, have also been isolated from diseased pigs (8, 9,28-30, 36, 48). Although C. sputorumsubsp mucosalis,C. hyointestinalisand C. coli have been suggested as possible causative agents of PPE, efforts to reproduce the disease by oral inoculation of pigs with pure cultures of these organisms have been mostly unsuccessful. Oral inoculation of pigs with homogenized ilea from pigs with PPE has been the only way to reproduce the disease (24, 25, 29, 30, 42). 791
McOrist and co-workers (32, 33) reported the isolation of a non-culturable intracellular Campylobacter-likeorganism from infected pig ileal sections. Protein banding profiles in sodium dodecyl sulfate-polyacrylamide gels and immunoblot analysis
may be a new Campylobacter species.
organism which can be grown in vitro, has been isolated from a pig with PPE (1). This organism, designated as RMIT 32A, has different phenotypic characteristics from the other Campylobacterspecies commonly reported to be associated with PPE and oral inoculation of conventional cross-bred pigs with RMIT 32A initiated symptoms of PPE. In this report, we describe the production of monoclonal antibodies to RMIT 32A. MATERIALS
Bacterial strains. Campylobacterspecies (Table 1) and RMIT 32A were grown on 5% defibrinated horse-blood agar (HBA) with or without Skirrow antibiotic supplementation (Oxoid, England) at 37 C in sealed jars (BBL Microbiology Systems, U.S.A.) in a microaerophilic environment induced by gassing the jars for 2 to 3 min with 50% H2 and 50% CO2. Serpulina species (Table 2) were grown on HBA plates at 37 C in an anaerobic environment ("GasPak," BBL Microbiology System, U.S.A.). Other gramnegative bacteria listed in Table 2 were grown aerobically on Columbia base agar (Oxoid) plates at 37 C. Monoclonalantibodyproduction. RMIT 32A was grown for 4 days and harvested with sterile phosphate-buffered saline, pH 7.4 (PBS). After three washes with PBS, 106cells were injected intraperitoneally into an eight-week-old female BALB/c mouse. The mouse received three injections with the final booster injection given 3 days prior to removal of the spleen (11). Dulbecco's Modification of Eagle's Medium (Cytosystems, Australia), supplemented with 10% heat-inactivated newborn bovine serum (iron supplemented; HyClone, U.S.A.), was used as the culture medium. The non-secretory myeloma cell line of Sp2/0-Ag14 (Sp2; 46) and resultant hybridomas were cultured at 37 C in a 5% CO2 environment. Hybridomas were produced by fusion (7, 10) using polyethylene glycol 1500 (Boehringer Mannheim, Germany) and HAT selection procedures (21). Enzyme-linkedimmunosorbent assay (ELISA). Preliminary screening of hybridomas was performed by ELISA in microwell plates (Nunclon, Denmark). Culture supernatants were used to detect sonicated RMIT 32A. After the addition of substrate (2, 2'-Azino-bis[3-ethyl benzthiazoline sulfonate]), wells exhibiting a strong green color were scored as positives. Characterization of immunoglobulinsubclass. The isotyping of monoclonal antibodies were performed using "MISO" mouse monoclonal antibody isotyping kit (Commonwealth Serum Lab., Australia). Neat hybridomas culture supernatants were used.
washed by (0.1
Lab.) ; 20 for
M Tris-HCl blue)
(LPS) procedure (mPCP;
was (RIM; 38).
isolated LPS was treated with RNase prior to analysis. performed by colorimetry (16).
Quantification of LPS was
Preliminary screening using EL1SA identified 50 hybridoma clones which produced antibodies to RMIT 32A. These hybridomas were cloned twice by limiting dilution. A hybridoma clone, C3: 52c1, was found to be secreting IgG2a antibodies (kappa light chains) to RMIT 32A. The specificity of this monoclonal antibody to RMIT 32A was examined by immunoblotting. An immunoblot using whole cell lysate of RMIT 32A and C3: 52cl revealed bands at approximately 19 kDa and immediately behind the dye front (Fig. 1, lane 2). There was also weak recognition of bands at 35 to 80 kDa region. Immunoblots of whole cell lysates of the bacteria listed in Tables 1 and 2 of "MATERIALSAND METHODS," showed some weak cross-reactivity (data not shown). The monoclonal antibody was also used in immunoblotting analysis of proteinase-K-treated (60 C for 1 hr) whole cell lysate of RMIT 32A. No recognition of 35 to 80 kDa was observed but the monoclonal antibody reacted with a broad band
2 3 1 Fig. 1. Immunoblot analysis ofRMIT 32A antigens with the monoclonal antibody, C3: 52cl, after SDS-PAGE using 12.5% gels. Lanes: 1, molecular weight standards (kDa): 2, whole cell lysate of RMIT 32A; 3, proteinase-K-treated (6(1C for 1 hr) whole cell lysate of RMIT 32A.
Campylobacter species used in the study
located between the dye front and 20 kDa region (Fig. 1, lane 3). When the monoclonal antibody was used in immunoblotting analyses of proteinase-K-treated whole cell lysates of other tested bacteria, no recognition was observed. To study the effectiveness of proteinase-K treatment of whole cell lysate at 60 C for 1 hr, whole cell lysates of RMIT 32A and Campylobacter coli NCTC 11366 (Fig. 2, A and B) were also proteinase-K treated overnight at 37 C and then, 2 hr at 65 C. The products of both proteinase-K treatments were subjected to SDS-PAGE and the gel was Coomassie brilliant-blue stained (Fig. 2A). The 60 C treatment (Fig. 2A, lanes 4 and 5) appears to be as equally effective as the longer overnight treatment (Fig. 2A, lanes 6 and 7) because there are no visible signs of protein bands. However, when a matching gel was silver stained (Fig. 2B), low molecular weight bands could be seen in lanes that were loaded with the products of proteinaseK treatments (Fig. 2B, lanes 4 to 7). Whole cell lysates of RMIT 32A and C.coli NCTC 11366 that were treated with proteinase-K at 60 C for 1 hr have similar banding patterns to their respective counterparts that were treated overnight at 37 C and then, 21 hr at 65 C. The products of proteinase-K treatment of whole cell lysates of gram-negative bacteria have been shown to be of LPS (3, 13, 22, 27, 38-41). To further define the
Fig. 2. SDS-polyacrylamide gel (12.5%) after (A) Coomassie brillant-blue, and (B) silver staining. Lanes: 1, molecular weight standards (kDa); 2, whole cell lysate of RMIT 32A; 3, whole cell lysate of C. co/i NCTC 11366; 4 and 6, proteinase-K-treated whole cell lysate of RMIT 32A; 5 and 7, proteinase-K-treated whole cell lysate of C. co/i NCTC 11366. Lanes 4 and 5 were treated at 60 (1 for 1 hr. Lanes 6 and 7 were treated overnight at 37 C and then, 2 hr at 65 C.
nature of the antigen recognized by C3: 52cl, LPS from RM IT 32A was isolated using two different procedures: the mPCP and RIM procedures. SDS-PAGE banding profiles of the isolated LPS are shown in Fig. 3. The immunoblot of the isolated LPS (using mPCP and RIM) and proteinase-K-treated whole cell lysate of RMIT 32A, probed with the monoclonal antibody, is shown in Fig. 4. There arc observable differences in the recognition of the antibody to mPCP-LPS, RIM-LPS and proteinase-K-treated whole cell lysate.
Fig. 3. Silver-stained SDS-polyacrylamide gel (10-20%) of LPS isolated from RMIT Lanes: 1, molecular weight standards (kl)a) 2, mPCP-LPS; 3, RIM-LPS.
Fig. 4. Immunoblot analysis of RMIT 32A LPS with the monoclonal antibody. C3: 52cl, after SDS-PAGE (l0 20% gel). Lanes: 1, molecular weight standards (kDa) ; 2, mPCP-LPS; 3, RI NI-1.PS; 4, proteinase-K-treated whole cell lysate of RMIT 32A.
In this report, we described the production of a monoclonal antibody, C3: 52cl, to a Campylobacter-likeorganism, RMIT 32A. Immunoblots of proteinase-K- and non-proteinase-K-treated whole cell lysates were used to study the specificity of the monoclonal antibody to RMIT 32A. The immunoblots using the monoclonal antibody and whole cell lysates of RMIT 32A, a selection of Campylobacterspecies and other gram-negative bacteria showed some weak cross-reactivities. After proteinase-K treatment of whole cell lysates at 60 C for 1 hr, the weak cross-reactivities have been completely eliminated, resulting in the specific recognition of an antigen of RMIT 32A by the antibody. The recognition (Fig. 1, lane 3) is located between the dye front and the 20 kDa region. Proteinase-K is a potent proteolytic enzyme and this analysis shows that both the 60 C treatment and the overnight treatment at 37 C and then 2 hr at 65 C, are equally effective in digesting proteins as there are no protein bands in the Coomassie brilliant-blue stained gel (Fig. 2A, lanes 4 to 7). However, bands were revealed when a matching gel was silver stained (Fig. 2B, lanes 4 to 7). The banding patterns, on the silver-stained gel of the 60 C treated whole cell lysates of RMIT 32A (Fig. 2B, lane 4) and C. coli NCTC 11366 (Fig. 2B, lane 5) were similar to their respective counterparts which were treated overnight at 37 C and then, 2 hr a t 65 C (Fig. 2B, lanes 6 and 7, respectively). Therefore, the 60 C treatment is the preferred treatment as it involved a shorter incubation duration. To further elucidate the nature of the antigen recognized by C3 : S2c1, LPS was isolated from RMIT 32A using two different procedures. The SDS-PAGE banding patterns of the isolated LPS (Fig. 3) is similar to Salmonellarough LPS chemotypes (13). The immunoblot of the monoclonal antibody against isolated LPS and proteinase-K-treated whole cell lysate of RMIT 32A (Fig. 4) showed recognition bands of low molecular weights (