MICROBIAL ECOLOGY IN HEALTH A N D DISEASE VOL. 2: 47-59 (1989) Characterisation of Protective Antibodies in Hamsters Immunised Against Clostridium diflcile Toxins A and €3 PYEUNG-HYEUN KIM and RIAL D. ROLFE* Department of Microbiology, Texas Tech University Health Sciences Center, Lubbock, Texas 79430, USA. Received 9 August 1988; revised 17 October 1988 Toxigenic Clostridium dtficile is the major cause of antimicrobial agent-associated pseudomembranous colitis in humans and of ileocaecitisin golden Syrian hamsters. The pathogenicity of C. dt@ciZe is believed to be dependent on the production of two immunologically and biochemically distinct toxins: toxin A (enterotoxin) and toxin B (cytotoxin). In earlier investigations we demonstrated that hamsters immunised against toxins A and B are protected against clindamycin-induced C. dtficile-associated ileocaecitis and that this protection is transferred to infant hamsters through maternal milk. In the present investigation, the class-specific immunoglobulin response in adult and infant hamsters immunised against C. djicile toxins A and B was examined using enzyme linked immunosorbent assays. Parenteral immunisation of adult hamsters against toxins A and B induced IgG, IgA and IgM antibodies in hamster sera specific for these two toxins. IgG antibodies to toxins A and B predominated in the milk and intestines of adult hamsters immunised against toxins A and B and in the sera and intestine of infants from immunised adult hamsters. No significant IgA or IgM antibodies to toxins A and B were detected in maternal milk and intestines or in infant sera and intestines. IgG antibodies to toxins A and B were detected in infant caecal contents up until 10 d and in sera up until approximately 24 d of age. The results from this investigation suggest that serum derived IgG antibody to C. djicile toxins is primarily responsible for protection of adult and infant hamsters against C. dzficile-associated ileocaecitis when immunised parenterally against toxins A and B. WORDS-clostridium Cytotoxin. KEY dzficife; Toxin A; Toxin €3: Antibody; Immunisation; Hamster; Ileocaecitis; Enterotoxin; INTRODUCTION Clostridium dificile-induced intestinal disease in humans is a health problem of significant clinical importance. Toxigenic C. dificile is the sole cause of antibiotic-induced pseudomembranous colitis, a severe and life-threatening intestinal disease, and is the cause of approximately 20 per cent of cases of antimicrobial agent-associated non-specific colitis and diarrhoea without c o l i t i ~ .All ' ~ major classes of antimicrobial agents have been reported to induce C. dijicile-associated intestinal disease in humans. In addition, C. dijicile is the cause of certain intestinal diseases not related to antimicrobial therapy.13.14,24.37 The mechanisms by which C. dificile induces intestinal disease are poorly understood. However, C. dificile-associated intestinal disease appears to involve many parameters including both the virulence of the microorganism (e.g., toxin production) and its interaction with the physiological status of the host (e.g., antibody). *Authorto whom correspondence should be addressed. 089ILo6OX/89/01oO47-13$06.50 0 1989 by John Wiley & Sons, Ltd. The pathogenicity of C. dificile, at least in part, is related to the production of two biochemically and immunochemically distinct toxins, 1 , 2 1 , 3 1 referred to as toxin A, an enterotoxin, and toxin B, a cytotoxin. It is not clearly understood which of the two toxins is responsible for the pathological changes seen in C. dificile-induced intestinal disease, though some evidence implicates both toxins. For example, both toxins are consistently detected in faecal specimens from humans and experimental animals with C. dificile-associated intestinal disIn addition, the intestinal pathology and death caused by purified toxins A and B in hamsters mimic the pathology observed during the natural disease On the other hand, there is evidence to suggest that toxin A plays a more important role in the disease process than toxin B. Mitchell et aLZ3 have shown that the secretory and tissue damaging properties of C. dificile broth filtrates on rabbit ileum are due to toxin A and not toxin B. It has also been reported that purified 48 P.-H. KIM AND R. D. ROLFE toxin A, but not toxin B, causes severe epithelial trate from a non-toxigenic strain of C. dijicile were cell necrosis with destruction of villi and poly- not protected. Antibody against the immunising morphonuclear infiltration in rabbit ileal loops.36 toxoid could be demonstrated in the serum, milk Investigators have postulated that the ability of and caecal contents of hamsters. Protection against humans to produce toxin-neutralising antibodies C. dzficile-induced intestinal disease could be may not only influence the severity of C. dificile- passively transferred from immunised mothers to associated disease, but also give protection against their infants, and neutralising antibodies to toxins A further infection with this microorganism. Circu- and B could be demonstrated in infant serum and lating antibodies to C. dzficile toxins can be intestinal contents. Foster mothering experiments demonstrated in humans and an antibody response demonstrated that maternal protection of infants frequently occurs in patients with C. dzficile- against C. dificile-associated ileocaecitis was transassociated d i a r r h ~ e a . ~For . ~ ~example, 82 per ferred to infant hamsters primarily through breast cent of individuals over the age of two years have milk and not across the placenta. antibody to toxin A and/or toxin B, showing seroIn the present investigation, the class-specific conversion occurs early in infancy.38 Analyses of immunoglobulin response in adult and infant serologic responses among patients with C. dzficile- hamsters immunised against C. dificile toxins A and induced intestinal disease show that approximately B was examined. 50 per cent of patients with C. dzficile-associated diarrhoea develop a serum antibody response to one MATERIALS AND METHODS or both of C. dzficile toxins during the d i s e a ~ e . ~ . ~ Furthermore, high antibody titres to C. dificile Maintenance of Hamsters toxins correlate to a less severe clinical course of Golden Syrian hamsters (Sasco, Inc., Omaha, i n f e ~ t i o n . ~Other ,~ than the indirect evidence NE) were used throughout the course of this investidescribed above, there is no definitive data to show gation. Adult ( 23 mths of age) and infant hamsters that circulating antibody is protective against C. were housed in conventional rooms and maintained dzficile-associated intestinal disease in adults. ad libitum on Purina Laboratory Rodent Chow The Syrian Golden hamster has been used exten- 5001 (Ralston Purina Co., Richmond, IN) and sively as an animal model in the study of C. dzficile- water. Hamsters were randomly assigned to treatassociated disease for several reasons.6 First, treat- ment or control groups. Female hamsters were bred ment with antimicrobial agents induces an intestinal by housing one female and one male hamster disease in hamsters similar to pseudomembranous together for 7 d after which the females were caged colitis in humans. Second, human-derived isolates individually. All infants in a litter received the same of C. dificile cause disease in hamsters. Third, C. treatment. dzficile toxins appear to play a major role in the pathogenesis of the disease in both humans and Purzjication and Characterisation of Clostridium hamsters. For example, in 1979, A110 et at.’ demonstrated that passive immunisation with Clostridium difficile Toxins A and B sordellii antitoxin, which cross-reacts with C. dzfiClostridium dificile toxins A and B were purified cile toxins, protected hamsters from clindamycin- using previously established techniques. 17,21,3 induced ileocaecitis and death. Much of what is The purification steps included preparation of a presently known about the pathogenesis, epidemi- cell-free dialysate, ultrafiltration to collect material ology, treatment and prevention of C. dzficile- > 100,000 molecular weight (Mr), Sepharose CL-6B associated intestinal disease was originally disco- chromatography to collect Mr size fractions of approximately 500,000, two different steps of vered in hamsters and later confirmed in humans. In earlier investigations, we examined the role of fractionation on DEAE Sepharose CL-6B and antibodies against C . dificile toxins A and B in finally an acetate precipitation step on the toxin protecting adult and infant hamsters against C. A preparation. Protein concentrations were deterdzficile-associated ileocaecitis.” Adult female mined using the protein-dye binding assay of hamsters immunised parenterally with toxoid A or Bradford6 available from Bio-Rad Laboratories a mixture containing both toxoids A and B were (Richmond, CA). Bovine gamma globulin was used protected against clindamycin-induced C. dificile- as the standard. associated fatal ileocaecitis. On the other hand, The homogeneity of purified preparations of C. hamsters immunised with toxoid B or a broth fil- dificile toxins A and B were characterised by ANTIBODIES TO C. DIFFICILE TOXINS A AND B 49 polyacrylamide gel electrophoresis (PAGE) in 7.5 Collection of Hamster Specimens per cent native gels” and Western blot analy~es,~’ Two weeks after completion of immunisation, crossed-immunoelectrophoresis,40 cytotoxin neu- female hamsters were mated with non-immunised tralisation, and enzyme linked immunosorbent male hamsters. The infant hamsters were removed assays with affinity purified rabbit antibodies to from their mothers 6 d post-partum, sacrified, and toxins A and B. At the final stages of the purification intestinal homogenates and serum prepared from process, the purified toxin preparations contain each infant. Maternal milk was collected from each only one (toxin A) or a few (toxin B) major compofemale hamster using a negative pressure suction nents. This indicated that the toxin preparations apparatus 24 h later.’ The female hamsters were were substantially enriched and free of major levels bled by cardiac puncture and serum prepared, and of contaminants. More importantly, the purified then sacrified and their intestinal tracts removed. toxin A preparation did not contain detectable Intestinal homogenates were prepared by homogetoxin B and vice versa. nising the intestinal tracts (tissue and contents) in equal volumes (w/v) of PBS. The homogenates were centrifuged and the supernatants filter sterilised. Immunisation of Hamsters Against Toxins A and B Isolation of Immunoglobulin A (IgA) from Hamster Female adult hamsters were immunised with Milk toxoid A, toxoid B, a mixture containing toxoids A Hamster IgA was isolated from the collected milk and B (toxoid AB) or a broth filtrate prepared from by a modified method of Bienenstock.’ Twenty ml a non-toxigenic strain of C . dzjicile (NTBF). of milk was mixed with 10ml normal saline and Toxoids were prepared as described by Libby et centrifuged at 100,000 x g for I h at 4°C. After al.” Briefly, toxoids A or B were prepared by the centrifugation, the clear aqueous layer between the addition of 1 ml(O.5 mg) of purified toxins A or B to upper fatty layer and the pellet was collected. The 9ml (6mg of protein) of NTBF. Toxoid AB was solution was brought to p H 4 with I M HCI and prepared by the addition of 0.5 ml of purified toxin centrifuged at 30,OOOxg for 30min at 4°C. The A (0.5 mg) and 0.5 ml of purified toxin B (0.5 mg) to pellet (mostly casein) was discarded and the super9ml (6mg of protein) of NTBF. The NTBF was natant neutralised with 2 M Tris and centrifuged prepared from a 72 h dialysis bag broth culture of again at 30,000 x g for 30 min at 4°C. The superC. dijicile VPI 2037 as described previo~sly.’~ natant was passed through a 0.45pm filter. To each preparation, 0.1 ml of 40 per cent (v/v) Twenty-six ml(35 mg of protein) of the filtrate were formaldehyde was added and the preparations were applied to a Sephadex G-200 column (2.5 x 120 cm) incubated for 36 h at 37°C. Formaldehyde-treated which had previously been equilibrated with 0.1 M NTBF was made by the addition of 1 ml of 50 mM sodium phosphate buffer (pH 6.8). The filtrate was phosphate buffered saline, pH 7.2 (PBS) to 9 ml eluted with the same buffer at 4°C. The effluent was (6 mg of protein) of NTBF and then treated with monitored for optical absorbance at 280 nm, and formaldehyde as described above for toxoids A and 12ml fractions collected. The first major peak, B. which eluted at the void volume, was pooled and Hamsters, weighing 70 to 80 g, were immunised concentrated by ultrafiltration. The retentate subcutaneously once per week for 4 wks with 0.2 ml contained primarily IgA and was characterised by of a 1:l mixture of Freund’s complete adjuvant ELISA. (GIBCO Laboratories, Chagrin Falls, Ohio) and toxoid A, toxoid B, toxoid AB or the NTBF.23 The hamsters were next immunised subcutaneously once Isolation of Immunoglobulin M (IgM) from per week for six weeks with 0.2 ml of a 1:1 mixture of Hamster Serum Freund’s incomplete adjuvant and the toxoid prepAdult hamsters were bled by cardiac puncture. arations. Animals were bled by cardiac puncture Hamster IgM was isolated from blood by a modi. ~ 30 ml of blood 7 d after the final immunisation. Serum anti-toxin fied method of C O ~Approximately titres were determined by using a cytotoxicity neu- was allowed to clot for 1 h at room temperature and tralisation assay performed in HeLa tissue culture then overnight at 4°C. Fifteen ml of serum was colcells and enzyme-linked immunosorbent assays lected and dialysed against 2 mM sodium phosphate buffer (pH 6.0) for 3 d at 4°C. The dialysate was (ELISAs). 50 centrifuged at 2,OOOxg for 10min at 4°C. The supernatant was discarded and the precipitate was re-suspended in cold phosphate buffer and re-centrifuged. The washing was repeated once. The final euglobulin precipitate was dissolved in 8ml of 0.1 M PBS (pH 8.0) and dialysed against 1 litre of the same buffer overnight at 4°C. Eight ml (35mg of protein) of dialysate was applied to a Sephadex G-200 column (2.5 x 120 cm) previously equilibrated with 0.1 M PBS (pH 8.0). The sample was eluted with the same buffer at 4°C. The effluent (7.5 ml/fraction) was monitored for optical absorbanceat 280 nm. Those fractions which eluted at the void volume were pooled and concentrated by ultrafiltration. Six ml(9 mg of protein) of the concentrated retentate were re-applied to a Sephadex G-200 column and eluted with 0.1 M PBS (pH 8.0). Two ml fractions were collected and each fraction which eluted at the void volume was characterised by ELISA. Characterisation of Mouse and Hamster ClassSpecijic Immunoglobulins The specific reactivities of goat anti-mouse IgG, IgA and IgM against mouse and hamster class-specific immunoglobulins were examined by enzyme linked immunosorbent assays. One hundred microlitres of serial two-fold dilutions of mouse or hamster class specific immunoglobulins were added to each well of a 96-well microtitre plate (Immunolon 11; Dynatech Laboratories, Inc., Alexandria, Va.). The mouse immunoglobulins were purchased from commercial sources; IgG and IgM were obtained from Calbiochem (La Jolla, CA) and IgA was obtained from Sigma Chemical Co. (St Louis, MO). Hamster IgG was purchased from PelFreez Biologicals (Rogers, AK). Hamster IgM and IgA were prepared as described above. After incubation overnight at 4"C, the wells were washed five times with 200 pl of PBS containing 0.05 per cent Tween (PBST). After washing, 0.1 ml of goat antimouse IgG (y chain specific), IgA ( a chain specific), or IgM (p chain specific) was added to the wells. All antisera were purchased from Sigma Chemical Co. The plates were incubated for 1 h at 37°C. The wells were then washed and rabbit anti-goat IgG alkaline phosphatase conjugate (Sigma Chemical Co.) was added to each well and incubated for 1 h at 37°C. After washing five times, 0.1 ml of phosphatase substrate (1 mg of p-nitrophenyl phosphate, disodium, per ml dissolved in diethanolamine buffer, pH9.8) was added to each well, and plates were P.-H. KIM AND R.D. ROLFE incubated for 30 min at room temperature. The reaction was terminated by the addition of 20 pl of 5 N NaOH to each well. The optical density at A,,, was measured in a microplate colorimeter. ELISA for Detection of Hamster Antibodies to Toxins A and B Purified toxins A and B were used to prepare ELISA plates for detecting antibodies to toxins A and B in hamster sera, intestinal homogenates, and maternal milk. C. dijicile toxins A and B were each diluted to a final concentration of 0.5 pg/ml in PBS. One hundred microlitres of the diluted toxin were added to each well of a 96-well microtitre plate. After incubation overnight at 4"C, the wells were washed five times with 200 p1 of PBST. Serial twoor five-fold dilutions of the materials to be tested for anibody were prepared in PBST containing 0.05 per cent (wt/vol) gelatin (PBSTG) and 0.1 ml of each dilution was added to a microtitre plate well. After incubation for 1 h at 37"C, wells were washed five times with PBST, and 0-1ml of goat anti-mouse IgG, IgA or IgM was added to the wells. Following incubation for 1 h at 37"C, the wells were washed and rabbit anti-goat IgG alkaline phosphatase conjugate was added to each well and incubated at 37°C. The wells were then washed as described above and the enzymatic activity in each well determined with p-nitrophenyl phosphate disodium, as described above. As controls, firstly some of the microtitre wells were coated with bovine serum albumin instead of toxin, secondly PBSTG was added to some wells in place of antibody samples, and thirdly samples of known antitoxin titre were diluted and added to several of the wells. Net absorbance was calculated by subtraction of the mean absorbance in wells coated with bovine serum albumin from the mean absorbance in wells coated with toxin. Cytotoxicity Neutralisation Assay Biological specimens prepared from hamsters (i.e., sera, maternal milk, and intestinal homogenates) were examined for their ability to neutralise the cytotoxic activity of purified preparations of toxins A and B in a HeLa tissue culture cell assay. Serial dilutions of the biological specimens were prepared in PBS, and each dilution mixed with an equal volume of toxins A or B. The toxins were diluted to concentrations 8-fold greater than the highest dilution of each toxin that caused complete rounding of 100 per cent of the HeLa cells in a 51 ANTIBODIES TO C. DIFFICILE TOXINS A AND B Table 1 . Summary of the specificity and sensitivity of anti-mouse immunoglobulins in ELISA P Valuef Anti-mouse5 IgG (Pool) Mouse IgG 16 Anti-mouse IgA Mouse IgA 32 Anti-mouse IgM MouseIgM 32 Anti-mouse IgG (Pool) Hamster IgG Anti-mouse IgA Hamster IgA Anti-mouse IgM Hamster IgM 250 63 63 IgG/IgG: IgA/ IgG : IgM/IgG: IgG/IgA: IgA/IgA: IgM/IgA: IgG/IgM: IgA/IgM: IgM/IgM: IgG/IgG: IgA/IgC: IgM/IgC: IgG/IgA: IgA/IgA: IgM/IgA: IgG/IgM: IgA/IgM: IgM/IgM: lOO0/0 8 Yo 8O h 4 yo loooh 13% 8% 6% 100% 100% 19% 18% 6 Yo 100% 15% 10% 22 Yo < 0.001 <0.002 < 0.001 < 0.001 <0.002 < 0.001 < 0.001 < 0.002 < 0.001 < 0.002 <0.001 < 0.00 1 100% *Sensitivity was determined by examining serial two-fold dilutions of each immunoglobulin by ELISA. The lowest concentration of immunoglobulins which gave an OD,,o significantly greater than background was defined as the sensitivity of the ELISA. ?Specificity is defined as the A,,, value of test immunoglobulin/A,,, value of reference immunoglobulin (noted in bold print) x 100. $Student’s t-test. $Goat anti-mouse immunoglobulins were prepared as described in Results. microtitre well (TCD,,,). The TCD,,, of the their infants. Since anti-hamster class-specific partially purified toxin A and B preparations were immunoglobulins are not commercially available 190 ng of protein (38.7 ng/ml) and 4.5 pg of protein we decided to substitute anti-mouse class-specific (90.4 pg/ml), respectively. After incubation at 37°C immunoglobulins for anti-hamster class-specific for 1 h, 50 p1 of each mixture were added to a micro- immunoglobulins in the ELISAs. titre well containing a monolayer of HeLa cells In the initial investigations, mouse immunogloand 200 p1 of cell culture medium. The HeLa cells bulins were characterised by ELISA. Optimum concentrations of goat anti-mouse classwere prepared as described p r e v i o ~ s l y .The ~ ~ ~ working ~~ antitoxin titre was defined as the reciprocal of the specific immunoglobulins for use in ELISA were greatest dilution which completely inhibited round- determined by checkerboard titration. Pooled ing of the HeLa cells after 24 h incubation at 37°C. anti-mouse IgG was prepared by mixing equal The differences between experimental groups were volumes of anti-mouse IgGl (diluted 1:3200 in PBSTG), IgG2a (1 :1600),IgG2b (1 :1600),and IgG3 analysed by Student’s t test (P-values). (1:3200). Goat anti-mouse IgA and IgM were 1:12 800 in PBSTG. These dilutions of goat diluted RESULTS anti-mouse immunoglobulins reacted specifically Characterisation of Mouse and Hamster Immunoglo- against 50 ng of the respective mouse immunoglobulins bulins in ELISA (Table 1). Hamster immunoglobulins were next examined ELISAs were developed to detect hamster classspecific antibodies to toxins A and B in various for reactivity with goat anti-mouse immunoglobubiological specimens obtained from hamsters lins (Table 1). Goat anti-mouse IgG was prepared immunised parenterally with toxoids A and B and by mixing equal volumes of anti-mouse IgGl 52 P.-H. KIM AND R. D. ROLFE 1.5 -. 1.5 h - h 5 1.0 - 1.0- * 0 0 w z z 0 . P U a c1 Lli . C 0.5 - $ CI 0.5 - a < -2 0.0 0 I I 1 2 , , , , 3 4 5 6 7 Serum Dilution (log2) I , , , 8 9 10 0.0 0 1 1 I 1 2 3 I I I I I 5 6 7 Serum Dilution (log2) 8 9 10 4 1 I Figure 1. Specificity of ELISAs for IgG antibody to Clostridium dijicile toxins A and B. Sera samples were obtained from hamsters imrnunised against toxin A, toxin B or NTBF. Each serum sample was diluted 1 :625 in PBSTG. Sera from hamsters immunised against toxin A ( 0 ) toxin ; B (A);NTBF (0) (diluted 1300 in PBSTG), IgG2a (1 :400), IgG2b (1 : 400), and IgG3 (1:800). Goat anti-mouse IgA and IgM were diluted 1 :3200 and 1:6400, respectively. Similar to the observations made with mouse immunoglobulins, the above dilutions of goat anti-mouse immunoglobulins reacted specifically against 50 ng of the respective hamster immunoglobulins in ELISA. However, overall sensitivity of goat antimouse immunoglobulins in detecting hamster immunoglobulins by ELISA was at least two times lower than for mouse immunoglobulins. Specijicity of ELISAs f o r Characterisation of Antibody to Toxins A and B Serum samples were obtained from hamsters immunised against toxins A, B, or the NTBF. Sera collected from hamsters immunised against toxins A or B neutralised the cytotoxic activity of each respective toxin; no cross-neutralisation was demonstrated. Sera collected from hamsters immunised against the NTBF did not neutralise the cytotoxic activity of either toxins A or B. These sera were tested in the two ELISAs developed for measuring class-specificantibodies to toxins A or B. No significant cross-reaction occurred between the two ELISAs for detection of IgG antibody to toxins A and B (Figure I). Sera collected from hamsters immunised against toxin A reacted in the ELISA for detection of IgG antibody to toxin A but not in the ELISA for detection of IgG antibody to toxin B. Similarly, sera collected from hamsters immunised against toxin B reacted in the ELISA for detection of IgG antibody to toxin B but not in the ELISA for detection of IgG antibody to toxin A. Similar results were obtained in the detection of IgA and IgM antibodies to these toxins. Characterisation of Class-SpeciJicImmunoglobulins in Adult Hamsters Immunised Parenterally against Toxins A and B Specimens collected from adult hamsters immunised parenterally against toxoid AB or the NTBF were characterised for class-specific antibodies to toxins A and B. The sera from hamsters immunised against toxins A and B possessed cytotoxicity neutralisation titres of 25 against toxin A and 625 against toxin B. The sera from hamsters immunised against the NTBF did not neutralise the cytotoxicity of either toxins A or B. Figure 2 shows that levels of IgG and IgM antibodies to both toxins predominated in sera although low but significant (p<O.OOl) levels of IgA antibody could also be demonstrated in the sera. The ileum and caecum from immunised adult hamsters were also examined for class-specificantibodies to toxins A and B. Significant levels of specific IgG antibodies to toxins A and B could be detected in both caecal tissue and contents (Figure 3). However, none of the ileal or caecal specimens contained antibody which neutralised the cytotoxic activity of toxins A or B. We have previously demonstrated that ELISAs are more sensitive than cytotoxicity neutralisation in detecting antibodies to toxins A and B.17 Specific IgA and IgM anti- 53 ANTIBODIES TO C. DIFFICLE TOXINS A AND B 2.5 1 2.0 h 5 s s 1.5 ai :: 2 1.0 0 2 0.5 0.0 IgG IgM IgA Figure 2. Characterisation of class-specificimmunoglobulins in serum of hamsters immunised against toxins A and B or NTBF. Serum was diluted 1:25 in PBSTG. Antibody to toxin A (D); Antibody to toxin B (0);NTBF sera reactivity in ELISAs (El). Data are the means+SEM (vertical bars) of four samples Caecal Tissue 1.2 Caecal Contents 1 l.01 1.2 T e 1 .o 0.8 3 0 s a, 0.6 E a 0.4 a d 0.2 0.0 IgG IgA IgM IgC IgM Figure 3. Class-specificantibodies to toxins A and Bin the caecum of adult hamsters immunised against toxins A and B or NTBF. For the preparation of caecal contents, the mucous lining of the caecum was added to the luminal contents. The samples were then diluted five times (w/v) in PBSTG, centrifuged and the supernatant filter sterilised. The intestinal tissue samples were diluted five times (w/v) in PBSTG, homogenised, centrifuged and the supernatant filter sterilised. Antibody to toxin A (El); Antibody to toxin B (0);NTBF sera reactivity in ELISAs (a).Data are the meansf SEM (vertical bars) of four samples bodies in caecal tissue and contents of hamsters immunised against toxins A and B were not significantly greater than in hamsters immunised against the NTBF. Similar results were obtained with ileal tissue and contents obtained from immunised adult hamsters (data not shown). IgG was also the predominate class-specific immunoglobulin to toxins A and B in the milk from hamsters immunised with toxoids A and B; no significant levels of IgA or IgM were present in the milk of these hamsters (Figure 4). The milk from hamsters immunised against toxins A and B possessed cytotoxicity neutralisation titres of 1:5 against toxin A and 1:125 against toxin B. Milk from hamsters immunised against the NTBF did not neutralise the cytotoxic activity of either toxins A or B. Milk samples from hamsters immunised with toxoid A and non-immunised hamsters were characterised for class-specificimmunoglobulins (Figure 5). Comparable levels of non-specific IgG and IgA were 54 P.-H. KIM AND R. D. ROLFE present in the milk from hamsters immunised against toxoid A, while IgA was the predominant immunoglobulin in non-immune milk. However, none of these differences in immunoglobulin concentrations in immune and non-immune milk were statistically significant ( p > 0.05) although this could be related to the small population size (i.e., duplicate samples). 2'01 T -- Characterisation of Class-Specific Immunoglobulins to Toxins A and B in Infant Hamsters The intestines and sera of 7 d-old infant hamsters from mothers immunised parenterally against toxins A and B when examined for class-specific antibodies to these toxins yielded results similar to those seen in maternal milk (Figure 6 ) with only significantlevels of IgG antibodies to toxins A and B detected. Levels of class-specificantibodies to toxins A and B were also measured in the caeca and sera of infant hamsters of different ages from mothers immunised parenterally against toxins A and B (Figure 7). Significant levels of IgG antibodies to these toxins were present in the caeca of infants up to 10 d old, whereas significant levels of IgG antibodies to toxins A and B could be demonstrated in the sera up to approximately 24 d of age. Significant levels of IgA and IgM could not be demonstrated in any of the hamsters (data not shown). DISCUSSION The ability of animals to produce toxin-neutralising Ik!G IgA I%M antibodies may not only influence the severity of C. Figure 4. Characterisation of class-specific immunoglobulins in dificile-associated disease, but also give protection milk from hamsters immunised against toxins A and B or NTBF. against further infection with C. dificile. The presMilk samples were collected 7 d post-partum, centrifuged, the supernatant filter sterilised, and the filtrate diluted 1:25 in ence of toxin-neutralising antibodies could also PBSTG. Antibody to toxin A (a); Antibody to toxin B (0); account for asymptomatic colonisation of adults NTBF sera reactivity in ELISAs (@). Data are the means+ SEM and infants by toxigenic C. dificile as well as the (vertical bars) of four samples failure of C. dificile to cause disease throughout the colon. Studies have demonstrated that antibodies to toxins A and B are present in the majority of older infants and adults and, furthermore, that patients with C. dificile-associated intestinal disease develop serologic responses to one or both In this investigation, the hamster model of clindamycin-induced C. dificile-associated ileocaecitis was used to study the role of antibodies to toxins A and B in protection against disease. ELISAs were -I developed to permit detection of hamster classspecific immunoglobulins directed against C. dzficile toxins A and B. Initial attempts were made to purify heavy chains of hamster class-specific immunoglobulins which were to be used in rabbit immunisation for obtaining anti-hamster class-specific immunoglobulins for use in ELISAs. However, it was not possible to obtain sufficient quantities of 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Absorbance (410 nm) purified hamster immunoglobulin heavy chains for Figure 5. Comparison of non-specific immunoglobulin levels in rabbit immunisation. Consequently, anti-mouse non-immune and immune milk. Milk samples were collected 7 d immunoglobulins were adopted for use in characpost-partum, centrifuged and the supernatants filter sterilised. terising hamster antibodies to toxins A and B. Samples were diluted 1 :15625 before being used to coat the wells Hamster immunoglobulins have not been as well of microtitre plates. Immune milk (8); Non-immune milk (FA). characterised as mouse and human immunoglobuData are the means of duplicate samples 55 ANTIBODIES TO C. DIFFICLE TOXINS A AND B Infant Intestine 2.5 -E 2 -? 2.0 2.0 h 2 s 1.5 E Y "e 2k W s: 3 Infant Serum 2.5 1.5 W x 2L 1.0 1.0 A . 2 0.5 0.5 0.0 0.0 IgG IgA IgM IgA IgM Figure 6. Characterisation of class-specificimmunoglobulins to toxins A and B in the intestine and serum of 7 d-old infant hamsters immunised against toxins A and B or NTBF. Intestinal samples (large and small intestine) were collected, diluted 5 times (w/v) in PBSTG, centrifuged, and the supernatant filter sterilised. Serum samples were diluted 1:25 in PBSTG. Antibody to toxin A (a); Antobody to toxin B (0);NTBF sera reactivity in ELISAs (El). Data are the means SEM (vertical bars) of four samples + Intestine 10 17 24 31 38 Age (days) Serum 48 Control 10 17 24 31 Age (days) 38 48 Control Figure 7. IgG antibody to toxins A and B in the caecum and serum of infant hamsters of different ages. Mother hamsters were immunised against toxins A and B. Caecal samples were diluted five times (w/v) in PBSTG, homogenised, centrifuged, and the supernatant filter sterilised. Each serum sample was diluted 1:25 in PBSTG. Control caecal homogenates and sera were obtained from 7 d-old infants from non-immunised mothers. Antibody to toxin A (El); Antibody to toxin B (El).Data are the means+ SEM (vertical bars) of three samples lins. Hamster immunoglobulins are composed of IgA, IgM and two subclasses of IgG (GI, G2) while mouse immunoglobulins are composed of IgA, IgM and three subclasses of IgG (Gl, G2a, G2b, G3).'s8 There is, however, evidence that hamster IgG2 is heterogeneous, indicating that at least two types of IgG2 exist. The exact evolutionary relationship between the hamster and mouse is unknown. Simpson3' places the divergence point somewhere in the late Miocene or early Pliocene period, approximately 20 million years ago. There is indirect evidence that mouse and hamster immunoglobulins are closely related. McGuire et a1.22compared hamster, mouse and human constant regions of IgM heavy chain (Cp) genes for phylogenetic conservation. Homologies with the hamster gene for the Cp 5' region are 8 1 per cent, 59 per cent, and 5 1 per cent for mouse, rabbit, and human, respectively, and the corresponding homologies for the Cp 3'region are 94 per cent, 83 per cent, and 81 per cent, respectively. Cross-reactivity of goat anti-mouse IgG (heavy chain specific) against hamster IgG was first examined in double immunodiffusion agar plates and a single precipitation line was observed (data not 56 shown). This promising finding prompted further analyses of the specific cross-reactions between anti-mouse immunoglobulins and hamster immunoglobulins by ELISA, which is a much more sensitive method than Ouchterlony double immunodiffusion. Overall sensitivity of anti-mouse immunoglobulins for detecting hamster immunoglobulins was lower than the sensitivity for detecting mouse immunoglobulins (Table 1). This could be due to a fewer number of antigenic sites on hamster immunoglobulins as compared to mouse immunoglobulins and/or a decreased affinity of the antibodies for the antigenic sites. However, overall specificity of the ELISAs for detection of hamster immunoglobulins was comparable to the specificity for mouse immunoglobulins. In addition, the ELISA for antibody to toxin A and the ELISA for antibody to toxin B were specific for each class of immunoglobulin (Figure I). Although the ELISAs using anti-mouse class-specific immunoglobulins permitted the detection of hamster class-specific antibodies, it would be advantageous to eventually use anti-hamster class-specific immunoglobulins because of possible increases in sensitivity and specificity. In earlier studies we demonstrated that adult hamsters immunised parenterally against only toxin A or both toxins A and B were protected against antibiotic-induced C. dzficile-associated ileocaecitis while hamsters immunised against only toxin B or a broth filtrate from a non-toxigenic strain of C. difJicile were not protected against disease. l 7 The protection observed in these investigations presumably reflected the inactivation of the biological activity of C. difJicile toxins before they could induce pathological damage in the intestinal tract. In the present investigation, parenteral immunisation of adult hamsters with toxoids A or B elicited toxin-specific IgG, IgA and IgM immunoglobulins in hamster sera. As shown in Figure 2, hamster class-specific antibody profiles to toxins A and B were almost identical, indicating that the immunisation protocols induced antibodies to both toxins with equal efficacy. Interestingly, significant levels of IgM antibody to toxins A and B were present in the sera even though antibody analyses were performed 14 weeks after the start of immunisation. Parenteral immunisation did not induce a toxinspecific mucosal IgA immune response since the ileal and caecal tissue contained only toxin-specific IgG (Figure 3). These results were unexpected since previous investigators have reported that parenteral immunisation with Vibrio cholerae and Shigella P.-H. KIM AND R. D. ROLFE jexneri elicits a specific IgA as well as IgG response in the mucosal secretions of rabbit^.'^.^^ Furthermore, it is well established that antibodies produced locally in the intestine, particularly sIgA, are important for specific immune protection against many enteric infections. Nonetheless, serum-derived IgG antibody protected adult hamsters against C. dificile-associated ileocaecitis. Since only very low levels of toxin-specificantibodies were detected in the luminal contents of the ileum and caecum it appears that toxin specific IgG acts at the level of the mucosa and/or submucosa. As described previously, maternal protection against C. dificile-associated ileocaecitis can be transferred to offspring through milk.” Therefore, it was of interest to identify the class-specific antibodies in milk which may account for the protection of infant hamsters against disease. Similar to the observations made with adult intestines, IgG was the predominate class-specific antibody to toxins A and B in the milk from hamsters immunised with toxoids A and B; once again, no significant levels of IgA or IgM antibodies were present (Figure 4). Theodore et ai.34 also demonstrated that only specific IgG antibody is present in the colostrum and milk of rabbits immunised intravenously with respiratory syncytial virus. The analyses of total non-specific immunoglobulins in the milk of non-immunised hamsters showed that IgA was the predominant immunoglobulin (Figure 5). Similar observations have been made in the colostrum of non-immunised hamsters. On the other hand, comparable levels of non-specific IgA and IgG were present in milk from hamsters immunised with toxoid A. These results suggest that toxin-specific IgG antibody and/or antibody-producing cells are transferred from serum to milk. At least two factors may contribute to the high levels of IgG in maternal milk of immunised hamsters. First, the half-life of IgG is generally longer than the half-life of IgA and IgM. The catabolic half-life of IgG, IgM, and IgA are 23, 5, and 6 days in humans and 4, 0.5, and 1.3 days in mice, respectively. Second, toxin-specific IgM and IgA may be degraded during lactation. Ogra and Ogra2’ have reported that the levels of IgM and IgA in human colostrum and milk fall rapidly after the first three to four days, whereas IgG remains at a constant low level. This may explain why no toxin-specific IgM is detected in hamster maternal milk although a significant level of toxinspecific IgM antibody is present in the sera. The class-specificantibodies to toxins A and B in infant intestine and sera were similar to the anti- 57 ANTIBODIES TO C. DIFFICILE TOXINS A AND B bodies present in adult intestine and milk (Figure 6). This indicates that toxin specific IgG present in maternal milk is absorbed across the intestinal tract of infant hamsters into the systemic circulation. Other investigators have also reported the intestinal absorption of IgG in infant The levels of IgG in infant sera were higher and lasted longer than infant caeca, indicating that IgG antibody in serum is more stable. We have previously shown that infant hamsters from mothers immunised parenterally against toxin A are partially protected against C. dijicile-associated ileocaecitis up to 17 days of age." This suggests that toxin-specific IgG in infant serum piays an important protective role against C. dificile-associated ileocaecitis since toxin-specific IgG in infant caeca was not present at significant levels after ten days of age while infant sera possessed toxin-specific IgG up to approximately 24 days of age (Figure 7). The mechanisms by which serum IgG antibody to toxins A and B protects adult and infant hamsters against C. dijicile-associated intestinal disease are not known and merit further investigation. However, one possible mechanism is inhibition of an inflammatory response in the intestine. Triadafilopoulos et al.36 reported that toxin A causes severe epithelial cell necrosis with destruction of villi and polymorphonuclear infiltration. They speculated that inflammatory mediators released by neutrophils, mast cells, and/or macrophages are likely to contribute to the intestinal damage seen in C. dificile-associated pseudomembranous colitis and diarrhoea. Recently, Daubener et aL9 demonstrated that C. dificile toxins inhibit T cell proliferation by influencing monocytes in a dose-dependent fashion. They proposed that the influence of C. dificile toxins on monocytes may be an additional factor in the development of pseudomembranous colitis. One can postulate from the data in our investigation that circulating antibody to C. dijicile toxins gains access to intestinal tissue which in turn neutralises the toxins before they can bind to specific receptors and induce harmful inflammation in the intestine. In conclusion, the induction of C. dificileassociated intestinal disease clearly involves many parameters including the physiological and immunological status of the host. The hamster is an effective model to study the pathogenesis of C. dificile-associated intestinal disease since enteric colonisation with C. dijicile is possible, a vigorous immune response occurs, and resistance to subsequent re-infection has been demonstrated. It would be hazardous to conclude that the patho- genesis of C. dificile and the possible role of immunological factors in protection are the same in humans as in hamsters. However, the similarities between C. dificile-associated disease in hamsters and humans suggests that the results obtained with the hamster model will directly aid in our understanding of the pathogenesis of C. dijicileassociated disease in humans. ACKNOWLEDGEMENTS This work was supported, in part, by Public Health Service grant ROI-A1121489 from the National Institute of Allergy and Infectious Disease. We thank Woosun Song for excellent technical assistance during the course of these investigations. REFERENCES 1. Allo M, Silva J, Fekety R, Rifkin GD, Waskin H. 2. 3. 4. 5. 6. 7. 8. 9. 10. (1979). Prevention of clindamycin-induced colitis in hamsters by Clostridium sordellii antitoxin. Gastroenterology 76,351-355. Aronsson B, Granstrom M. (1988). Immunological response to Clostridium dijicile infection. In: Rolfe RD, Finegold SM (ed) Clostridium difficile: Its Role in Intestinal Disease. Academic Press, New York, pp 99-1 12. Aronsson B, Granstrom M, Molby R, Nord CE. (1983). Enzyme-linked immunosorbent assay (ELISA)for antibodiesto Clostridium dzficile toxins in patients with pseudomembranous colitis and antibiotic-associated diarrhea. Journal of Immunological Methods 60,341-350. Aronsson B, Molby R, Nord CE. (1984). Diagnosis and epidemiology of Clostridium dificile enterocolitis in Sweden. Journal of Antimicrobial Chemotherapy 14 (Suppl. D), 85-95. Bienenstock J. (1970). Immunoglobulins of the hamster. 11. Characterization of the yA and other immunoglobulins in serum and secretions. Journal of Immunology 104,1228-1235. Bradford MA. (1976). A rapid and sensitive method for quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Analytical Biochemistry 72,248-254. Coe JE. (1968). The immune response in the hamsters. I. Definition of the two 7s globulin classes. Journal of Immunology 100,507-5 15. Coe JE, Bell JF. (1977). Antibody response to rabies virus in Syrian hamsters. Infection and Immunity 16, 9 15-9 19. Daubener W, Leiser E, Eichel-StreiberC, Hadding U. (1988). Clostridium dificile toxins A and B inhibit human immune response in vitro. Infection and Immunity 56, 1107-1 112. Davis BJ. (1965). Disc electrophoresis. 11. Methods 58 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. P.-H. KIM AND R. D. ROLFE and application to human serum proteins. Annals of the New York Academy of Sciences 121,404427. Donta ST. (1988). Mechanism of action of Clostridium dificile toxins. In: Rolfe RD, Finegold SM (eds) Clostridium difficile: Its Role in Intestinal Disease. Academic Press, New York, pp 169-181. Escribano MJ, Haddada H, DeVaux Saint Cyr C (1982). Isolation of two immunoglobulin G subclasses, IgG2 and IgGl, from hamster serum protein-A-Sepharose. Journal of Immunological Methods 52,63-72. Fainstein V, Bodey GP, Fekety R. (I98 1). Relapsing pseudomembranous colitis associated with cancer chemotherapy. Journal of Infectious Diseases 143, 863-865. George WL. (1988). Antimicrobial agent-associated diarrhoea in adult humans. In: Rolfe RD, Finegold SM (eds) Clostridium difficle: Its Role in Intestinal Disease. Academic Press, New York, pp 3 1 4 . Hasegawa H, Nakamura A, Watanabe K, Brown WR, Nagura H. (1987). Intestinal uptake of IgG in suckling rats: Distinction between jejunal and ileal epithelial cells demonstrated by simultaneous ultrastructural localization of IgG and acid phosphatase. Gastroenterology 92,186191. Keren VF, McDonald RA, Carey JL. (1983). Effect of parenteral immunization on the local immunoglobulin A response of the intestine to ShigellaJlexneri antigens. Infection and Immunity 42,202-207. Kim P-H, Iaconis JP, Rolfe RD. (1987). Immunization of adult hamsters against Clostridium dificileassociated ileocecitis and transfer of protection to infant hamsters. Infection and Immunity 55, 29842992. Kim P-H, Rolfe RD. (1988). Characteristics of protective antibodies to Clostridium dij?cile toxins A and B. Abstracts of the Annual Meeting of the American Society for Microbiology, Abstract E24. Libby JM, Jortner BS, Wilkins TD. (1982). Effects of the two toxins of Clostridium dijicile in antibioticassociated cecitis in hamsters. Infection and Immunity 36,822-829. Lyerly DM, Sullivan NM, Wilkins TD. (1983). Enzyme-linked immunosorbent assay for Clostridium dij?cile. Journal of Clinical Microbiology 17, 72-78. Lyerly DM, Wilkins TD. (1988). Purification and properties of toxins A and B of Clostridium dij?cile. In: Rolfe RD, Finegold SM (eds) Clostridium difficile: Its Role in Intestinal Disease. Academic Press, New York, pp 145-167. McGuire KL, Duncan WR, Tucker PW. (1985). Phylogenetic conservation of immunoglobulin heavy chains: direct comparison of hamster and mouse Cp genes. Nucleic Acids Research 13, 561 15638. Mitchell TJ, Ketley JM, Halan SC, Stephen J, Burdon DW, Candy CC, Daniel R. (1986). Effects of 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 31. toxins A and B of Clostridium dificile on rabbit ileum and colon. Gut 27,78-85. Moskovitz M, Bartlett JG. (1981). Recurrent pseudomembranous colitis unassociated with prior antibiotic therapy. Archives of Internal Medicine 141,663464. Ogra SS, Ogra PL. (1978). Immunologic aspects of human colostrum and milk. I. Distribution characteristics and concentrations of immunoglobulins at different times after the onset of lactation. Journalof Pediatrics 92,546549. Onderdonk AB. (1988). Role of the hamster model of antibiotic-associated colitis in defining the etiology of the disease. In: Rolfe RD, Finegold SM (eds) Clostridium difficile: Its Role in Intestinal Disease. Academic Press, New York, pp 115-127. Peri BA, Rothberg RM. (1986). Transmission of maternal antibody prenatally and from milk into serum of neonatal rabbits. Immunology 57,49-53. Rolfe RD, Finegold SM. (1979). Purification and characterization of Clostridium dificile toxin. Infection and Immunity 25,191-201. Rolfe RD, Iaconis JP. (1983). Intestinal colonization of infant hamsters with Clostridium dificile. Infection and Immunity 42,480-486. Simpson GG. (1959). Cold Spring Harbor, Symposium on Quantitative Biology 24,255-271. Sullivan NM, Pellet S, Wilkins TD. (1982). Purification and characterization of toxins A and B of Clostridium di8cile. Infection and Immunity 35, 1032-1040. Svennerholm A, Holmgren J. (1 977). Immunoglobulin and specific-antibody synthesis in vitro by enteral and nonenteral lymphoid tissues after subcutaneous cholera immunization. Infection and Immunity 15,36&369. Taylor NS, Thorne GM, Bartlett JG. (1981). Comparison of two toxins produced by Clostridium dificile. Infection and Immunity 34,10361043. Theodore CM, Losonsky G, Peri B, Fishaut M, Rothberg RM, Ogra PL. (1982). Immunologic aspects ofcolostrum and milk: development of antibody response to respiratory syncytial virus and bovine serum albumin and rabbit mammary glands. In: Handon LA, Sell KW. (eds) Recent Advances in Mucosal Immunity. Raven Press, New York, pp 393403. Towbin H, Staehelin T, Gordon J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Proceeding of the National Academy of Sciences USA 76, 4650-4654. Triadafilopoulos G, Pothoulakis C, O’Brien MJ, LaMont JT. (1987). Differential effects of Clostridium dijicile toxins A and B on rabbit ileum. Gastroenterology 93,273-279. Trnka VM, LaMont JT. (1981). Association of Clostridium di8cile toxin with symptomatic relapse ANTIBODIES TO C. DIFFICILE TOXINS A AND B of chronic inflammatory bowel disease. Gastroenterology 80,693-696. 38. Visicidi R, Laughon BE, Yolken R, Bo-Linn P, Moench T, Ryder PW, Bartlett JG. (1983). Serum antibody response to toxins A and B of Clostridium diflciie. Journal of Infectious Diseases 148,93-100. 39. Wexler H. (1988). Other potential diagnostic techniques. In: Rolfe RD, Finegold SM (eds) 59 Clostridium difficile: Its Role in Intestinal Disease. Academic Press, New York, pp 309-338. 40. Wunderly CH. (1960). The technique of immunoelectrophoresis in agar gel. In: Alexander P, Block AJ (eds) A Laboratory Munuaf of Analytical Methods of Protein Chemistry, Vol. 2. Pergamon Press, New York, pp 143-167.
© Copyright 2026 Paperzz