Veterinary Microbiology, 30 ( 1 9 9 2 ) 2 5 7 - 2 6 6 Elsevier Science Publishers B.V., A m s t e r d a m
257
Profiles of toxin production by thermophilic Campylobacter of animal origin B.A. McFarland and S.D. Neill Bacteriology Department, Veterinary Research Laboratories, Stormont, Belfast, BT4 3SD, UK (Accepted 20 June 1991 )
ABSTRACT McFarland, B.A. and Neill, S.D., 1992. Profiles of toxin production by thermophilic Campylobacter of animal origin. Vet. MicrobioL, 30: 257-266. Seventy-five strains of Campylobacterjejuni and C. coli, which were isolated from a variety of animal species, primarily poultry, were examined for production of toxin. Polymyxin extracts were tested in in vitro assays using CHO-KI, FCL (foetal calf lung), Vero, HeLa and CEF (chicken embryo fibroblast ) cells. The toxic effects observed were cell rounding and death. Extracts from almost all C. jejuni and C. coli strains were toxic to both CHO-KI and FCL cells and 69.0% of C. jejuni isolates and 75% of C, coli isolates were also toxic to CEF cells. 50.7% of C. jejuni extracts were toxic to Vero cells and 46.5% toxic to HeLa cells. None of the C. coli isolates were toxic to either of these cell lines. None of the strains tested produced cytotonic enterotoxin. No differences in toxigenicity patterns were evident between Campylobacter isolated from different sources.
INTRODUCTION
Campylobacter are recognised as aetiological agents of disease in both humans and animals. In particular, Campylobacterjejuni has been reported to be one of the more common causes of human diarrhoeal disease (Bruce et al., 1977; Butzler and Skirrow, 1979; Blaser and Reller, 1981 ). From clinical presentations, invasion of host tissues (Duffy et al., 1980; Manninen et al., 1982 ) and toxin production (Ruiz-Palacios et al., 1985; Honda et al., 1986; Walker et al., 1986 ) have been described as possible mechanisms of pathogenesis. Thermophilic Campylobacter can readily be recovered from poultry (Neill et al., 1984; Smitherman et al., 1984) and consumption of poultry has been widely reported as an important source of C. jejuni involved in human enteritis (Blaser et al., 1980; Grant et al., 1980; Park et al., 1981 ). Although their association with avian hepatitis (Peckman, 1958) and with wet litter conditions often found in broiler houses (Neill et al., 1984) has been noted, the isolation ofthermophilic Campylobacterfrom poultry has usually been in the absence of overt clinical disease. 0 3 7 8 - 1 1 3 5 / 9 2 / $ 0 5 . 0 0 © 1992 Elsevier Science Publishers B.V. All rights reserved.
258
B.A. MCFARLAND AND S.D. NEILL
The aim of the present study was to examine a selection of thermophilic
Campylobacter isolates, primarily C. jejuni recovered from commercial poultry, for production of toxin (s) which could potentially be involved in avian a n d / o r h u m a n diseases. MATERIALS AND METHODS
Strains examined The Campylobacter isolates were obtained from a variety of specimens submitted for bacteriological examination to the laboratory. Isolates were presumptively identified as Campylobacter by microscopic examination and subsequently speciated using standard biotyping tests (Lander and Gill, 1985 ). The C. jejuni strains consisted of 48 avian, 8 bovine, 4 canine, 2 ovine and 1 porcine strain. The four isolates of C. coli examined consisted of two avian and two porcine strains. In addition six h u m a n isolates of C. jejuni were obtained from a local hospital. C. jejuni strains 3969 and 4182 which had previously been reported to produce cytotoxin and cytotonic toxin respectively (Johnson and Lior, 1986 ) were included in the study. In total, 75 strains were examined. All strains were stored at - 7 0 °C in trypticase soy broth containing 0.05% Campylobacter FBP supplement (Oxoid) and 0.075% agar. Prior to testing all strains were grown on blood agar base no. 2 (Difco) containing 5% lysed horse blood. Escherichia coli strains NCTC 11601 (a heat-labile toxin (LT) producing strain) and NCTC 11603 (a heat-stable toxin (ST) producing strain) were included as controls for enterotoxin.
Preparation of Polymyxin Extracts Field strains were initially cultured on blood agar plates (Difco cat. no. 0696) containing 5% lysed horse blood and incubated at 37°C for 48 hours in a reduced oxygen atmosphere of 90% nitrogen, 5% carbon dioxide and 5% oxygen. Colonies were subsequently inoculated onto lysed horse blood agar (two plates per strain), which were incubated in a reduced oxygen atmosphere at 37°C for 48 hours. Growth was harvested into 2 mg ml -~ polymyxin B sulphate (Sigma) in ice-cold saline ( 1 ml per plate) and the suspension was incubated at 37°C for 5 min. The preparation was centrifuged at 3000 g for 25 min and the supernatant sterilised by passage through a 0.22 /zm disposable filter (Millipore) before testing for toxin. The E. coli strains were grown on lysed horse blood agar for 18 h. Single colonies were inoculated into casamino acid yeast extract broth containing 4 mg ml-1 lincomycin (Sigma) and incubated aerobically at 37 °C for 24 hours. Growth was aerated by constant shaking. The resulting suspension was centrifuged at 3000 g for 25 min and the supernatant filter sterilised before testing in the assays.
PROFILES OF CAMPYLOBACTER TOXIN
259
Tissue culture All media used contained nystatin ( 50 U m l - l ), penicillin ( 100 U m l - 1), streptomycin sulphate ( 1 0 0 / t g ml - I ) and glutamine (2.0 m M ) . Chinese hamster ovary (CHO-KI) cells were grown in Hams F 12 m e d i u m ( Flow Laboratories, cat. no. 12-422-54) supplemented with 10% foetal calf serum (FCS), foetal calf lung (FCL) cells were maintained in Glasgow modified Eagle's m e d i u m (Gibco, cat. no. 072-02100) supplemented with 10% heat inactivated (56°C, 45 min) horse serum and Vero cells were grown in Iscove's modification of Dulbecco's minimal essential m e d i u m (DMEME) (Gibco, cat. no. 074-02200 ) supplemented with 5% horse serum (5%). HeLa and chick embryo fibroblast (CEF) cells were maintained in Eagles minimal essential m e d i u m (Flow Laboratories, cat. no. 14-100) supplemented with 10% FCS. The cells were seeded in tissue culture flasks (50 ml; Nunc Products) and incubated in 10% CO2 at 37 °C until required. The tissue cultures were examined daily to ensure a healthy status. Tissue culture assay Toxin assays were carried out using the media described above, but with serum concentrations of 3% for CHO-KI and FCL cells, 5% for HeLa cells, 2% for CEF cells and 1% for Vero cells. To perform the assay, 50/d of appropriate assay m e d i u m was added to each well of a 96-well flat-bottomed microtitre plate (Nunc Products). The appropriate bacterial extract was added to the first well (50/A) and doubling dilutions were made to a final dilution of 3.3X 10 -7. AS controls, polymyxin solution (2.0 mg m1-1 in saline) and the appropriate tissue culture m e d i u m were similarly titrated. Preliminary testing of Campylobacter extracts indicated that the addition of tissue culture cells to the extract dilutions was more sensitive for detecting toxin than the addition of extract to the cells after incubation of the cells for 24 hours (McFarland, 1990). Freshly trypsinised cells were therefore added immediately to each well at a concentration of 2.0X 106 cells per ml, diluted in the appropriate medium. A lid was placed on the microtitre plate, which was incubated in 10% CO2 for up to 72 hours. Wells were observed at 24-h intervals under phase contrast microscopy using an inverted microscope. Cytotoxin titres were recorded as the inverse of the highest dilution of extract which caused rounding of more than 70% of the cells. Cytotonic toxin titres were recorded as the inverse of the highest dilutions which caused distinct elongation of more than 20% of the cells in each field. RESULTS
Although the extracts of all the Campylobacter strains exhibited a cytotoxic effect against the tissue cultures none caused a cytotonic effect. The cytotoxic effect could be observed as complete destruction of the cell monolayers due
260
B.A. MCFARLAND AND S.D. NEILL
to a rounding up of the cells. Figs. l a - c illustrate this effect on C H O cells. Similar monolayer destruction was observed using CEF, HeLa, F C L and Vero cell lines.Toxic activity was most clearly demonstrated after 48 hours, as incubation for longer periods resulted in less distinct end-points of titration relative to the control cells. Cytotoxic and cytotonic activities were clearly demonstrable against the C H O cells using the extracts prepared from E. coli strains N C T C 11603 (ST producing) and N C T C 11601 (LT producing). Neither polymyxin solution nor the tissue culture m e d i u m had any visible effect on any of the cell lines used in the assays.
Fig. 1. The effect of toxic extracts on CHO cells. (a) Cytotonic effect on monolayer (48 h) after addition of extract from E. coli strain NCTC 11601 (final dilution l /8). Mag.× 306. (b) Untreated cell monolayer (48 h) Mag.× 306. (c) Cytotoxic effect on monolayer (48 h ) after addition of extract from C. jejuni strain B 133A (final dilution 1/8). Mag.× 306.
261
PROFILES OF CAMPYLOBACTER TOXIN
With one exception extracts from all of the Campylobacter isolates showed toxic activity against the CHO-KI and FCL cells. This toxic activity was not demonstrated for all of the strains using other cell lines. The range of toxic titres obtained for the strains using the different cell types are shown in Table 1. The CHO-KI cell line appeared most sensitive. Of avian strains having titres in CHO assay in excess of 103, four had titres in excess of 10v. Four of the non-avian C. jejuni and one C. coli also gave titres in excess of 107. Titres demonstrated in all other cell lines were low and the toxic effect was usually not observed beyond one well dilution (titre 4). The characteristic cytotonic effect on CHO cells as seen with E. coli LT (Fig. 1 ) was not exhibited by any of the Campylobacter extracts. Strain 4182 (reported by Johnson and Lior, 1986 to produce cytotonic toxin) gave a titre in excess of 1 0 7 in CHO assay, titres of 4 in FCL and Vero cell assays, a titre of 8 in HeLa cell assay and was not toxic to CEF cells. Strain 3969 (reported by Johnson and Lior, 1986 to produce cytotoxin) gave a titre of 102 in CHO assay, and although giving titres of 8 and 4 in FCL and HeLa cell assays respectively, was not toxic to Vero or CEF cells. The toxigenic profiles associated with C. jejuni isolates from different sources are summarised in Table 2. TABLE 1 Effect o f Campylobacter p o l y m y x i n extracts on tissue culture cells Cell line
Titre
Source a o f C. jejuni Avian
C. coli
Non-avian
CHO
t > 103 1 0 2 < t < 103 0 < t < 102 t=0
20 20 8 0
8 8 8 0
4 0 0 0
FCL
t > 10 0 < t < 10 t=0
8 39 1
3 18 0
2 2 0
Vero
t > 10 0 10 0
257
Profiles of toxin production by thermophilic Campylobacter of animal origin B.A. McFarland and S.D. Neill Bacteriology Department, Veterinary Research Laboratories, Stormont, Belfast, BT4 3SD, UK (Accepted 20 June 1991 )
ABSTRACT McFarland, B.A. and Neill, S.D., 1992. Profiles of toxin production by thermophilic Campylobacter of animal origin. Vet. MicrobioL, 30: 257-266. Seventy-five strains of Campylobacterjejuni and C. coli, which were isolated from a variety of animal species, primarily poultry, were examined for production of toxin. Polymyxin extracts were tested in in vitro assays using CHO-KI, FCL (foetal calf lung), Vero, HeLa and CEF (chicken embryo fibroblast ) cells. The toxic effects observed were cell rounding and death. Extracts from almost all C. jejuni and C. coli strains were toxic to both CHO-KI and FCL cells and 69.0% of C. jejuni isolates and 75% of C, coli isolates were also toxic to CEF cells. 50.7% of C. jejuni extracts were toxic to Vero cells and 46.5% toxic to HeLa cells. None of the C. coli isolates were toxic to either of these cell lines. None of the strains tested produced cytotonic enterotoxin. No differences in toxigenicity patterns were evident between Campylobacter isolated from different sources.
INTRODUCTION
Campylobacter are recognised as aetiological agents of disease in both humans and animals. In particular, Campylobacterjejuni has been reported to be one of the more common causes of human diarrhoeal disease (Bruce et al., 1977; Butzler and Skirrow, 1979; Blaser and Reller, 1981 ). From clinical presentations, invasion of host tissues (Duffy et al., 1980; Manninen et al., 1982 ) and toxin production (Ruiz-Palacios et al., 1985; Honda et al., 1986; Walker et al., 1986 ) have been described as possible mechanisms of pathogenesis. Thermophilic Campylobacter can readily be recovered from poultry (Neill et al., 1984; Smitherman et al., 1984) and consumption of poultry has been widely reported as an important source of C. jejuni involved in human enteritis (Blaser et al., 1980; Grant et al., 1980; Park et al., 1981 ). Although their association with avian hepatitis (Peckman, 1958) and with wet litter conditions often found in broiler houses (Neill et al., 1984) has been noted, the isolation ofthermophilic Campylobacterfrom poultry has usually been in the absence of overt clinical disease. 0 3 7 8 - 1 1 3 5 / 9 2 / $ 0 5 . 0 0 © 1992 Elsevier Science Publishers B.V. All rights reserved.
258
B.A. MCFARLAND AND S.D. NEILL
The aim of the present study was to examine a selection of thermophilic
Campylobacter isolates, primarily C. jejuni recovered from commercial poultry, for production of toxin (s) which could potentially be involved in avian a n d / o r h u m a n diseases. MATERIALS AND METHODS
Strains examined The Campylobacter isolates were obtained from a variety of specimens submitted for bacteriological examination to the laboratory. Isolates were presumptively identified as Campylobacter by microscopic examination and subsequently speciated using standard biotyping tests (Lander and Gill, 1985 ). The C. jejuni strains consisted of 48 avian, 8 bovine, 4 canine, 2 ovine and 1 porcine strain. The four isolates of C. coli examined consisted of two avian and two porcine strains. In addition six h u m a n isolates of C. jejuni were obtained from a local hospital. C. jejuni strains 3969 and 4182 which had previously been reported to produce cytotoxin and cytotonic toxin respectively (Johnson and Lior, 1986 ) were included in the study. In total, 75 strains were examined. All strains were stored at - 7 0 °C in trypticase soy broth containing 0.05% Campylobacter FBP supplement (Oxoid) and 0.075% agar. Prior to testing all strains were grown on blood agar base no. 2 (Difco) containing 5% lysed horse blood. Escherichia coli strains NCTC 11601 (a heat-labile toxin (LT) producing strain) and NCTC 11603 (a heat-stable toxin (ST) producing strain) were included as controls for enterotoxin.
Preparation of Polymyxin Extracts Field strains were initially cultured on blood agar plates (Difco cat. no. 0696) containing 5% lysed horse blood and incubated at 37°C for 48 hours in a reduced oxygen atmosphere of 90% nitrogen, 5% carbon dioxide and 5% oxygen. Colonies were subsequently inoculated onto lysed horse blood agar (two plates per strain), which were incubated in a reduced oxygen atmosphere at 37°C for 48 hours. Growth was harvested into 2 mg ml -~ polymyxin B sulphate (Sigma) in ice-cold saline ( 1 ml per plate) and the suspension was incubated at 37°C for 5 min. The preparation was centrifuged at 3000 g for 25 min and the supernatant sterilised by passage through a 0.22 /zm disposable filter (Millipore) before testing for toxin. The E. coli strains were grown on lysed horse blood agar for 18 h. Single colonies were inoculated into casamino acid yeast extract broth containing 4 mg ml-1 lincomycin (Sigma) and incubated aerobically at 37 °C for 24 hours. Growth was aerated by constant shaking. The resulting suspension was centrifuged at 3000 g for 25 min and the supernatant filter sterilised before testing in the assays.
PROFILES OF CAMPYLOBACTER TOXIN
259
Tissue culture All media used contained nystatin ( 50 U m l - l ), penicillin ( 100 U m l - 1), streptomycin sulphate ( 1 0 0 / t g ml - I ) and glutamine (2.0 m M ) . Chinese hamster ovary (CHO-KI) cells were grown in Hams F 12 m e d i u m ( Flow Laboratories, cat. no. 12-422-54) supplemented with 10% foetal calf serum (FCS), foetal calf lung (FCL) cells were maintained in Glasgow modified Eagle's m e d i u m (Gibco, cat. no. 072-02100) supplemented with 10% heat inactivated (56°C, 45 min) horse serum and Vero cells were grown in Iscove's modification of Dulbecco's minimal essential m e d i u m (DMEME) (Gibco, cat. no. 074-02200 ) supplemented with 5% horse serum (5%). HeLa and chick embryo fibroblast (CEF) cells were maintained in Eagles minimal essential m e d i u m (Flow Laboratories, cat. no. 14-100) supplemented with 10% FCS. The cells were seeded in tissue culture flasks (50 ml; Nunc Products) and incubated in 10% CO2 at 37 °C until required. The tissue cultures were examined daily to ensure a healthy status. Tissue culture assay Toxin assays were carried out using the media described above, but with serum concentrations of 3% for CHO-KI and FCL cells, 5% for HeLa cells, 2% for CEF cells and 1% for Vero cells. To perform the assay, 50/d of appropriate assay m e d i u m was added to each well of a 96-well flat-bottomed microtitre plate (Nunc Products). The appropriate bacterial extract was added to the first well (50/A) and doubling dilutions were made to a final dilution of 3.3X 10 -7. AS controls, polymyxin solution (2.0 mg m1-1 in saline) and the appropriate tissue culture m e d i u m were similarly titrated. Preliminary testing of Campylobacter extracts indicated that the addition of tissue culture cells to the extract dilutions was more sensitive for detecting toxin than the addition of extract to the cells after incubation of the cells for 24 hours (McFarland, 1990). Freshly trypsinised cells were therefore added immediately to each well at a concentration of 2.0X 106 cells per ml, diluted in the appropriate medium. A lid was placed on the microtitre plate, which was incubated in 10% CO2 for up to 72 hours. Wells were observed at 24-h intervals under phase contrast microscopy using an inverted microscope. Cytotoxin titres were recorded as the inverse of the highest dilution of extract which caused rounding of more than 70% of the cells. Cytotonic toxin titres were recorded as the inverse of the highest dilutions which caused distinct elongation of more than 20% of the cells in each field. RESULTS
Although the extracts of all the Campylobacter strains exhibited a cytotoxic effect against the tissue cultures none caused a cytotonic effect. The cytotoxic effect could be observed as complete destruction of the cell monolayers due
260
B.A. MCFARLAND AND S.D. NEILL
to a rounding up of the cells. Figs. l a - c illustrate this effect on C H O cells. Similar monolayer destruction was observed using CEF, HeLa, F C L and Vero cell lines.Toxic activity was most clearly demonstrated after 48 hours, as incubation for longer periods resulted in less distinct end-points of titration relative to the control cells. Cytotoxic and cytotonic activities were clearly demonstrable against the C H O cells using the extracts prepared from E. coli strains N C T C 11603 (ST producing) and N C T C 11601 (LT producing). Neither polymyxin solution nor the tissue culture m e d i u m had any visible effect on any of the cell lines used in the assays.
Fig. 1. The effect of toxic extracts on CHO cells. (a) Cytotonic effect on monolayer (48 h) after addition of extract from E. coli strain NCTC 11601 (final dilution l /8). Mag.× 306. (b) Untreated cell monolayer (48 h) Mag.× 306. (c) Cytotoxic effect on monolayer (48 h ) after addition of extract from C. jejuni strain B 133A (final dilution 1/8). Mag.× 306.
261
PROFILES OF CAMPYLOBACTER TOXIN
With one exception extracts from all of the Campylobacter isolates showed toxic activity against the CHO-KI and FCL cells. This toxic activity was not demonstrated for all of the strains using other cell lines. The range of toxic titres obtained for the strains using the different cell types are shown in Table 1. The CHO-KI cell line appeared most sensitive. Of avian strains having titres in CHO assay in excess of 103, four had titres in excess of 10v. Four of the non-avian C. jejuni and one C. coli also gave titres in excess of 107. Titres demonstrated in all other cell lines were low and the toxic effect was usually not observed beyond one well dilution (titre 4). The characteristic cytotonic effect on CHO cells as seen with E. coli LT (Fig. 1 ) was not exhibited by any of the Campylobacter extracts. Strain 4182 (reported by Johnson and Lior, 1986 to produce cytotonic toxin) gave a titre in excess of 1 0 7 in CHO assay, titres of 4 in FCL and Vero cell assays, a titre of 8 in HeLa cell assay and was not toxic to CEF cells. Strain 3969 (reported by Johnson and Lior, 1986 to produce cytotoxin) gave a titre of 102 in CHO assay, and although giving titres of 8 and 4 in FCL and HeLa cell assays respectively, was not toxic to Vero or CEF cells. The toxigenic profiles associated with C. jejuni isolates from different sources are summarised in Table 2. TABLE 1 Effect o f Campylobacter p o l y m y x i n extracts on tissue culture cells Cell line
Titre
Source a o f C. jejuni Avian
C. coli
Non-avian
CHO
t > 103 1 0 2 < t < 103 0 < t < 102 t=0
20 20 8 0
8 8 8 0
4 0 0 0
FCL
t > 10 0 < t < 10 t=0
8 39 1
3 18 0
2 2 0
Vero
t > 10 0 10 0
