- Email: [email protected]
Effect of oral Saccharomyces boulardii treatment on the activity of Clostridium difficile toxins in mouse digestive tract
r~~, vot. 30. No. tz ~. tss?~tsa9, t~z. Prmua ~ G~au t~.
ao4t-0tot~t ss.ao + .ao ® tv~z Paa.moo t~ ua
EFFECT OF ORAL SACCHAROMYCES BOULARDII TREATMENT ON THE ACTIVITY OF CLOSTRIDIUM DIFFICILE TOXINS IN MOUSE DIGESTIVE TRACT G. CORTFITEIt, I F. Lues,t S. JowERT2 and F. C~xz 'Unité d'l;.cologie et de Phy~ologie du Système Digestif; 78352 Jouy-en-Josas, France; and =Unité de Recherche en Immunologie, Faculté de Pharmacie, Université de Montpellier I, 34060 Montpellier, France (Received 30 March 1992 ; accepted 11 July 1992)
G. CORTI~It, F. Luc.+s, S. JOUVmtT and F. C~+.srsx. Effect of oral Saccharomyces botdardü treatment on the activity of Clostridium docile toxins in mouse digestive tract. Toxicon 30, 1583-1589, 1992.-Human antibioticassociated diarrhoea and pseudomembranous colitis are partly due to toxin production by Clostridium dfff:cile. It is now well documented that Saccharomyces boulardü protects against C. dt~cile induced diseases . In an attempt to understand better the mechanism of this protective effect, the action of S. boulardü on a crude toxin preparation was studied in vitro and in vivo. The results showed that the yeast had no effect on the toxins in vitro but was able to protect mice inoculated with these toxins . Furthermore, the observation by scanning electron microscopy that the mucosa of S. boulardü protected mice was not damaged suggests that the yeast mainly acts on the intestinal mucosa.
INTRODUCTION
Clostridium docile producing toxin A and B is responsible for pseudomembranous colitis and some forms of antibiotio-associated diarrhoea in humans (BARTI.SIT et al., 1978; L>BBY et al., 1982; GEORG): et al., 1978; LnRSON et al., 1978; LY1ntLY et al., 1988 ; SMALL, 1968). These diseases can be experimentally reproduced in clindamycin-treated hamsters and in gnotobiotic rodents (BARTLiri'T et al., 1978; $MALL, 1968; CORTFITER et al., 1985; CzupRnvsxl et al., 1983) . On the other hand, several authors have developed models to study the action of isolated toxins of C. dt~cile (POTHOULAKL4 et al., 1986; TRIADAFILOPOULOB et al., 1987) . It has been largely demonstrated in rodent models that oral treatment with Saccharomyces boulardü, a medically important yeast used to treat human diarrhoea, prevents the animals from dying (CASTFJC et al., 1990; CORTI~It et al., 1986; ELM)3R and McFARLAxD, 1987; MASSOT et al., 1984; TooT> "Iwxlat and ELa~lt, 1984) . This therapy has been successfully used to treat similar diseases in man ($URAWICZ et al., 1989a,b). Although the exact mechanism of action of S. boulardü is unknown, the yeast reduces toxinogenesis and protects the intestinal mucosa (CAST137C et al., 1990; CORTIiIER et al., 1986; ELM)31t and McFARLAxD, 1987; TooTxAlont and ELMER, 1984). 1583
1584
G. CORTHIER et aJ.
The purpose of this study was to determine whether S. boulardü has an antitoxic effect in vitro and in S. boulardü treated mice challenged orally with C. di,~ftcile toxins . The
effects of the toxins on the intestinal mucosa were visualized by scanning electron microscopy (SEM).
MATERIALS AND METHODS Clostridium di~cile toxins Strain VPI 10463 was kindly provided by N. M. Sullivan (Virginia Polytechnic Institute, Blackesburg, VA, U.S.A.). Crude toxin preparation was obtained by culturing C. diffcile within a dialysis bag (Sttu.rv~uv et al., 1982) in flasks containing autoclaved Brain Heart Infusion (Difw Laboratories, Detroit, MI, U.S .A .) . The medium did not contain added glucose to avoid acidiScation ; flasks were incubated 4 days at 37°C in an anaerobic chamber. Toxin A was quantified by enzyme immunoassay (M~ et al., 1987) and toxin H was determined by tissue culture as previously described (Coxr~rt et al., 1985, 1989). Toxin A and H concentrations were expressed as logic n8 1~ ml . Saccharomyces boulardü suspensions Hatches of S. boulardü suspensions, containing approximately 10'° viable yeasts per ml, were provided by Laboratories Biocodex (Montrouge, France). Effect of S. boulardü on toxins in vitro Three different dilutions of the crude toxin in 7.5% bicarbonate solution pH 7.7 were incubated with the suspensions of S. boulardü for 2 hr at 37°C . The concentrations of the toxins before and after incubation were evaluated by previously described methods (COR7iIIHt et al., 1985; M~ et al., 1987 ; Corthier et al., 1989). The toxin activity remaining after contact with S. boulmdü was determined after sedimenting the yeast cells and administering 0.5 ml of the supernatants to three groups (ten mice per group) of wnventional Ba1bC mice (4 montés old) by the orogastric route. Control mice received the diluted toxin preparations that had not beta exposod to the yeast. During a 3-day period following inoculation, mortality was evaluated. Effect of S. boulardü on toxins in vivo Ba1bC mice (13 per group) received S. Iwulardü in their drinking water (5 .109 viable yeasts per ml) during a 9-day experimental period. A freshly prepared suspension of yeast was given daily. On day 6, the mice rectivod by the orogastric route, 0.5 ml of a dilution of the crude toxin preparation mixed with a washed S. boulardü suspension to obtain the same final concentration as in drinking water. A preliminary study using dye showed that the ingestion of a 0.5 ml sample induced flushing of the intestinal contents within 5 min from the first half of the small intestine. For this reason, we added S. boulardü to the toxin preparations to compensate for the clearance of the yeast from the digestive tract. Eighteen hours post-inoculation, three mice were killed and portions of their duodenum, jejunum and ileum were immediately prepared for SEM (C~st~r et al., 1990). The ten remaining mice served to estimate mortality. Control mice receiving only the toxins, or only S. Loulardil or no treatment at all were also investigated. RESULTS
The action of S. boulardü on C. docile toxins was first investigated in vitro. Saccharomyces boulardü did not cause any reduction in the concentrations of toxins at any
of the three dilutions tested (Table 1). This suggests that yeast had no direct action on the toxins . This was confirmed by the absence of reduction in mortality when mice were given the supernatants of the toxin dilutions mixed with these S. boulardü suspensions compared with control mice that had only received the toxin dilutions (Table 1). Next, we tested the ability of S. boulardü included in the drinking water to protect mice challenged orally with toxin dilutions . The results in Table 1 show that all the mice pretreated with the yeast survived the administration of toxin whatever the dilution tested .
Mortality after oral administration (%) and toxin treatments
3.6 t0.2 2.8 f0.4 2.3 t 0.2
Dilution 1 Dilution 2 Dilution 3
C. difficüe supernatant were administered to mice pretreated with S. tïoulardü and biological activity was leafed
3.7 t0.3 2.9 t0.4 2.5 t0.3
4.1 t0.3 3.7 f0.2 3.1 t0.4
3.5 f0.2 2.7 f0.4 2.2 f0.4 2.2 f0.3 1.7 f0.4 1.5 t 0.4
3.0 f0.4 2.5 t 0.3 2.0 t0.4 2.7 f0.4 1 .5 t0.3 1 .1 f0.4
0 0 0
100 l00 20
" Three experiments were performed each with a different dilution of crude extract toxin preparation. t Amounts of toxins (log,o ng/ml) fstandard error of the mean.
4.1 f0.4 3.7 f0.4 3.1 f03
Dilution 1 Dilution 2 Dilution 3 70 20 0
70 90 20
0 0 0
100 100 50
100 100 60 0 0 0
90 90 SO 80 40 0
90 40 0
100 100 60
Day 3 Toxin B Day 1 Day 2 Crude toxin Toxin A preparation" S. Iwulardü Control S. Loultndii Control S. boulardü Control S. liotdardü Control S. boulardü Control
C. dlffecile supernatant was incubated in vitro with S, Gmdardü then toxin concentrations and biological activity were tested
Protocols
Toxin eonoentrationst and toxin treatments
TeetB 1. l~tcacv o~ S. boWmdli ox C. diffreik sur~erv~rexr corrr~o mxnv A ~uvn B. At~ in vitro rxrwre~r rr~ roxnvs w~a nvocat.~r~ nvro sac$ oR rs~ sace rASrRe~T® wrrtt S. boulmdü xacmv~ rr~ C. djJrcüe roxtNS Hv ow+t. Aouze
158 6
G. CORTHIER eJ al.
FIG .
1.
SCANNING
ELECTRON
MICROSCOPY OF THE JEJUNUM (HAR, 100 ~JIII~.
OF
TOXIN-FREE
CONTROL
MICE
The SEM examinations revealed that the aspect of the small intestinal mucosa of the untreated mice and the control mice only treated with S. boulardü was the same (Fig. 1). The small intestinal mucosa of the mice only treated with the toxins had an abnormal aspect . The villi appeared flabby and joined at the regions of contacts (Fig. 2) . The top of the villi had wrinkles (Fig. 3). These phenomena were mainly observed for the duodenum and the jejunum and to a lesser extent for the ileum of mice challenged with the highest concentration of toxins A and B. For the two other concentrations, the phenomena were occasionally observed . In contrast, the small intestinal mucosa of mice pretreated with S. boulardü and challenged with the toxin had a normal aspect with yeast cells usually grouped on the surface (Fig. 4). In some fields, the villi were free from yeasts (Fig. 5) but the mucosa was still quite similar to the toxin-free controls .
FIG . Z . GENERAL ASPECT OF TJJE JEJUNUM OF MICE INOCULATED WITH THE MOST CONCENTRATED CRUDE TOXIN PREPARATION (HAR, ZOO FUII~.
Oral S. boulardü Treatment
158 7
FIG. 3 . WRINKLED ASPECT' OF THE VILW OF TFIE JFJIJNUM OF MICE INOCULATED WITH THE MOST' CONCENTRATED CRUDE TOXIN PREPARATION (BAR, TOO
ulll).
DISCUSSION
The results corroborate the previously described ability of S. boulardü to protect mice against the pathological affects caused by administering whole bacteria (CASTex et al ., 1990 ; CORTFIIER et al., 1986) . As in the last reported study (CASTER et al ., 1990), here again the SEM observations showed that the small intestinal mucosa remained intact when the mice were pretreated with S. boulardü . Since we also demonstrated in the present report that the yeast did not need to act directly on C. diffJCile toxins to exert its protective effect, the other hypothesis that can be put forth to explain this phenomenon is that the yeast acts on the intestinal mucosa. When only toxins were administered to mice, death occurred rapidly ; and since the gut of the moribund mice showed no marked injury but only flabby villi that appeared to fuse,
FiG. 4 . SCANNING ELECTRON MICR08COPY OF TIC JEJUNUM OF S. bouialdii PROTECTED MICE SHOWING A NORMAL MUCOSA WITH GROUPS OF YEAST CELLS (HAR, TOO um).
158 8
G.
FIG .
S.
CORTHIER et
al.
SCANNING ELECTRON MICROSCOPY OF THE JFJUNUM OF S. boulardü PROTECTED MICE SHOWING NORMAL V1LLI, FREE OF YEASTS (BAR, IOOltm) .
it is possible that the toxins crossed the mucosa and caused death by systemic action. In protected mice, S. boulardü may also exert a direct action rendering the enterocytes impermeable to toxins . It has been shown that the yeast prevents C. docile toxins from damaging intestinal cells in culture (CZERUCKA et al., 1991). Other mechanisms that have been proposed to explain the protection of S. boulardü include stimulation of the immune system (Burs et al., 1990) and modification of the toxin A brush border receptor by activation of intestinal enzymes (Burs et al., 1986). Furthermore, S. boulardü might also cause a reduction in toxin A receptors as has been reported for the digestive microflora (Lucas et al., 1989). In conclusion, the present data indicate that even under the drastic conditions of our model where the mice received the crude toxin preparation directly in the stomach, S. boulardü proved to be a highly effective protective agent. Acknowledgement-The
authors thank Dr
S.
L. SALHI for critical reading
of the
manuscript.
REFERENCES
J . G ., MooN, N ., CHANG, T . W., TAYLOR, N . and ONDERDONK, A. B . (1978) Role of Clostridium docile in antibiotic-associated pseudomembranous colitis . Gastroenterology 75, 778-782 . Burs, J . P., BERNASCONI, P., VAN CRAYNFST, M . P., MALDAGUE, P. and DE MEYER, R . (1986) Response of human and rat small intestinal mucosa to oral administration of S . boulardü. Pediatr . Res . 20, 192-196 . Burs, J. P ., BERNASCONI, P ., VAERMAN, J . P . and DtvE, C . (1990) Stimulation of secretory IgA and secretory component of immunoglobulins in small intestine of rats treated with S. boulardü. Dig. Dis. Sci. 35, 231-236. CASTEX, F ., CORTHIER, G ., JOUVERT, S ., ELMER, G. W ., LUCAS, F . and BAS'rIDE, M . (1990) Prevention of Clostridium di,~Jicile-induced experimental pseudomembranous colitis by Saccharomyces boulardü: a scanning electron microscopic and miaobiological study . !. Gen . Mùrobiol. I36, 1085-1089. CORTHIER, G ., DUBOS, F. and RAIBAUD, P . (1983) Modulation of cytotoxin production by C. docile in the intestinal tracts of gnotobiotic mice inoculated with various human intestinal bacteria. Appl . Envir. Micrabiol. 49, 250-232 . CORTt-igR, G ., Dueos, F . and DUCLUZEAU, R . (1986) Prevention of Clostridium difficile induced mortality in gnotobiotic mice by Saccharomyces boulardü. Can . J. Microbiol. 32, 894-896. Coxrl-nER, G ., MULLER, M . C., ELMER, G . W . and LucAS, F . (1989) Interrelationships between digestive proteolytic activities and production and quantitation of toxin in pseudomembranous colitis induced by Clostridium difficile in gnotobiotic mice . Inject Immw. 57, 3922-3927. BARTLEIT,
Oral S. boulardil Treatment
1589
Czsauctce, D., Nexo, J. L., Bi:ariescoxr, P. and Rer~er., P. (1991) Protective effects of Saccharomyces boulardü against the action of Clostrldhon difficile toxine. Gattroenterol. Clin . Biol. 15, 22-27. Czueatraxr, C. J., Jora~sox, W. J., Ber.rsEt, E. and Wn .xnvs, T. D. (1983) Pseudomembraaous colitis in Closrridhon difficile monoaseoeiated rata . I>!Ject Immun. 39, 1368-1376. Et .t~x, G. W. and McFeauxn, L. V. (1987) Supprea~on by Saccharwnyces boulardü of toxigenic Clostridhon diJfictk overgrowth after vancomycin treatment in hamsters . Antimicrob . Agents. Chemother. 31, 129-131 . GHOAüE, W. L., Snnoxns, J. M., Dusoctt, F., Bitowx, J. D., Aaear, Y., Srm~ieGewe, N., IC~a~v, M. R. B., Atexexnne-Wrr.r.uVS, J. and Buaoox, D. W. (1978) Identification of Clostridhaw difficile as a cause of pseudomembranoue colitis. Br Mcd. 1, 695. Lensox, H. E., Hoxous, P., Parce, A. B. and BoRA>Er.rA, S. P. (1978) Clostridium diffcile and the aetiology of pseudomembranous colitis. Lmicet i, 1063-1066. l:.mev, J. M., DoxTe, S. T. and Wn rigs, T. D. (1982) Effects of two toxins of C. di,~cile in antibiotio-aseociated acids in hamsters . Inject . Immun. 36, 822 29 . Lues, F., Eras, G. W., Bxor-Leaoct~, E. and Coarrt~a, G. (1989) Fixation of Clostridlion diffcile toxin A and cholera toxin to intestinal brush border membranes from axenic and conventional mice. Inject Immun. 57, 1680-1683. Lvear.v, D. M., Knmex, H. C. and Wn .s:nvs, T. D. (1988) Clostridüon difficile: its disease and toxins. Clin. Microbial. Rev. 2, 1-118. Mew, S., Coxrr~e, G. and Drmos, F. (1987) Effect of various diets on toxin production by two strains of Clostridium difficile in gnotobiotic mice . Infect . Immun. 55, 1801-1805. Messar, J., Sexc~z, O., Coucrrv, R., AsTOrx, J. and PeAOnr, A. L. (1984) Bacterio-pharmacological activity of Saccharomyces boulardü in clindamycin-induced colitis in the hamster. ArzneLnittelforscMoig 34, 794-797 . PorxouLAKL4, C., Twenetrsn.orouwa, G., Ctea¢, M., Frtexzateu, C. and Lessoxr, J. T. (1986) Clostridiion difficile cytotoxin inhibits protein synthesis in fibroblasts and intestinal mornes . Gastroenterology 91, 1147-1153. S~"~?, J. D. (1968) Fatal enterocolitis in hamsters given lincomycin hydrochloride. Lab. Mi. Care 18, 4l 1-420. Stn.uvex, N. M., PetrEr-r', S. and Wtr.xnvs, T. D. (1982) Purification and characterization of toxins A and B of Clostridium difficile . Infect . Immun. 35, 1032-1040. Suaewicz, C. M., Et.~e, G. W., Sr~r.itex, P., McFeauxn, L. V., Cxrx, J. and Vex-Ber.r~, G. (1989x) Prevention of antibiotio-assaciated diarrhoea by Saccharomyces boulardii: a prospective study. Gastroenterology 96, 981=988. Suxewrcz, C. M., McFeatexn, L. V., Er.t~t, G. W. and Crnxx, J. (1989b) Treatment of recurrent Clostridhon doffcüe colitis with vancomycin and Saccharomyces boulardü. Am. I. Gastrocnt. 84, 1285-1287. Toortter®t, R. D. and Er.e~rt, G. W. (1984) Prevention of clinaamycin-induced mortality in hamsters by Saccharomyces boulardii. Mtimicrob. Agents Chemother. 26, 552-556. Tarenerrn.orour.os, G., Porxour.ears, C., O'Bnw, M. J. and LewoxT, J. T. (1987) Differential effects of Clostridiwn diff tilt toxins A and B on rabbit ileum. Gastroenterology 93, 273-279.
ao4t-0tot~t ss.ao + .ao ® tv~z Paa.moo t~ ua
EFFECT OF ORAL SACCHAROMYCES BOULARDII TREATMENT ON THE ACTIVITY OF CLOSTRIDIUM DIFFICILE TOXINS IN MOUSE DIGESTIVE TRACT G. CORTFITEIt, I F. Lues,t S. JowERT2 and F. C~xz 'Unité d'l;.cologie et de Phy~ologie du Système Digestif; 78352 Jouy-en-Josas, France; and =Unité de Recherche en Immunologie, Faculté de Pharmacie, Université de Montpellier I, 34060 Montpellier, France (Received 30 March 1992 ; accepted 11 July 1992)
G. CORTI~It, F. Luc.+s, S. JOUVmtT and F. C~+.srsx. Effect of oral Saccharomyces botdardü treatment on the activity of Clostridium docile toxins in mouse digestive tract. Toxicon 30, 1583-1589, 1992.-Human antibioticassociated diarrhoea and pseudomembranous colitis are partly due to toxin production by Clostridium dfff:cile. It is now well documented that Saccharomyces boulardü protects against C. dt~cile induced diseases . In an attempt to understand better the mechanism of this protective effect, the action of S. boulardü on a crude toxin preparation was studied in vitro and in vivo. The results showed that the yeast had no effect on the toxins in vitro but was able to protect mice inoculated with these toxins . Furthermore, the observation by scanning electron microscopy that the mucosa of S. boulardü protected mice was not damaged suggests that the yeast mainly acts on the intestinal mucosa.
INTRODUCTION
Clostridium docile producing toxin A and B is responsible for pseudomembranous colitis and some forms of antibiotio-associated diarrhoea in humans (BARTI.SIT et al., 1978; L>BBY et al., 1982; GEORG): et al., 1978; LnRSON et al., 1978; LY1ntLY et al., 1988 ; SMALL, 1968). These diseases can be experimentally reproduced in clindamycin-treated hamsters and in gnotobiotic rodents (BARTLiri'T et al., 1978; $MALL, 1968; CORTFITER et al., 1985; CzupRnvsxl et al., 1983) . On the other hand, several authors have developed models to study the action of isolated toxins of C. dt~cile (POTHOULAKL4 et al., 1986; TRIADAFILOPOULOB et al., 1987) . It has been largely demonstrated in rodent models that oral treatment with Saccharomyces boulardü, a medically important yeast used to treat human diarrhoea, prevents the animals from dying (CASTFJC et al., 1990; CORTI~It et al., 1986; ELM)3R and McFARLAxD, 1987; MASSOT et al., 1984; TooT> "Iwxlat and ELa~lt, 1984) . This therapy has been successfully used to treat similar diseases in man ($URAWICZ et al., 1989a,b). Although the exact mechanism of action of S. boulardü is unknown, the yeast reduces toxinogenesis and protects the intestinal mucosa (CAST137C et al., 1990; CORTIiIER et al., 1986; ELM)31t and McFARLAxD, 1987; TooTxAlont and ELMER, 1984). 1583
1584
G. CORTHIER et aJ.
The purpose of this study was to determine whether S. boulardü has an antitoxic effect in vitro and in S. boulardü treated mice challenged orally with C. di,~ftcile toxins . The
effects of the toxins on the intestinal mucosa were visualized by scanning electron microscopy (SEM).
MATERIALS AND METHODS Clostridium di~cile toxins Strain VPI 10463 was kindly provided by N. M. Sullivan (Virginia Polytechnic Institute, Blackesburg, VA, U.S.A.). Crude toxin preparation was obtained by culturing C. diffcile within a dialysis bag (Sttu.rv~uv et al., 1982) in flasks containing autoclaved Brain Heart Infusion (Difw Laboratories, Detroit, MI, U.S .A .) . The medium did not contain added glucose to avoid acidiScation ; flasks were incubated 4 days at 37°C in an anaerobic chamber. Toxin A was quantified by enzyme immunoassay (M~ et al., 1987) and toxin H was determined by tissue culture as previously described (Coxr~rt et al., 1985, 1989). Toxin A and H concentrations were expressed as logic n8 1~ ml . Saccharomyces boulardü suspensions Hatches of S. boulardü suspensions, containing approximately 10'° viable yeasts per ml, were provided by Laboratories Biocodex (Montrouge, France). Effect of S. boulardü on toxins in vitro Three different dilutions of the crude toxin in 7.5% bicarbonate solution pH 7.7 were incubated with the suspensions of S. boulardü for 2 hr at 37°C . The concentrations of the toxins before and after incubation were evaluated by previously described methods (COR7iIIHt et al., 1985; M~ et al., 1987 ; Corthier et al., 1989). The toxin activity remaining after contact with S. boulmdü was determined after sedimenting the yeast cells and administering 0.5 ml of the supernatants to three groups (ten mice per group) of wnventional Ba1bC mice (4 montés old) by the orogastric route. Control mice received the diluted toxin preparations that had not beta exposod to the yeast. During a 3-day period following inoculation, mortality was evaluated. Effect of S. boulardü on toxins in vivo Ba1bC mice (13 per group) received S. Iwulardü in their drinking water (5 .109 viable yeasts per ml) during a 9-day experimental period. A freshly prepared suspension of yeast was given daily. On day 6, the mice rectivod by the orogastric route, 0.5 ml of a dilution of the crude toxin preparation mixed with a washed S. boulardü suspension to obtain the same final concentration as in drinking water. A preliminary study using dye showed that the ingestion of a 0.5 ml sample induced flushing of the intestinal contents within 5 min from the first half of the small intestine. For this reason, we added S. boulardü to the toxin preparations to compensate for the clearance of the yeast from the digestive tract. Eighteen hours post-inoculation, three mice were killed and portions of their duodenum, jejunum and ileum were immediately prepared for SEM (C~st~r et al., 1990). The ten remaining mice served to estimate mortality. Control mice receiving only the toxins, or only S. Loulardil or no treatment at all were also investigated. RESULTS
The action of S. boulardü on C. docile toxins was first investigated in vitro. Saccharomyces boulardü did not cause any reduction in the concentrations of toxins at any
of the three dilutions tested (Table 1). This suggests that yeast had no direct action on the toxins . This was confirmed by the absence of reduction in mortality when mice were given the supernatants of the toxin dilutions mixed with these S. boulardü suspensions compared with control mice that had only received the toxin dilutions (Table 1). Next, we tested the ability of S. boulardü included in the drinking water to protect mice challenged orally with toxin dilutions . The results in Table 1 show that all the mice pretreated with the yeast survived the administration of toxin whatever the dilution tested .
Mortality after oral administration (%) and toxin treatments
3.6 t0.2 2.8 f0.4 2.3 t 0.2
Dilution 1 Dilution 2 Dilution 3
C. difficüe supernatant were administered to mice pretreated with S. tïoulardü and biological activity was leafed
3.7 t0.3 2.9 t0.4 2.5 t0.3
4.1 t0.3 3.7 f0.2 3.1 t0.4
3.5 f0.2 2.7 f0.4 2.2 f0.4 2.2 f0.3 1.7 f0.4 1.5 t 0.4
3.0 f0.4 2.5 t 0.3 2.0 t0.4 2.7 f0.4 1 .5 t0.3 1 .1 f0.4
0 0 0
100 l00 20
" Three experiments were performed each with a different dilution of crude extract toxin preparation. t Amounts of toxins (log,o ng/ml) fstandard error of the mean.
4.1 f0.4 3.7 f0.4 3.1 f03
Dilution 1 Dilution 2 Dilution 3 70 20 0
70 90 20
0 0 0
100 100 50
100 100 60 0 0 0
90 90 SO 80 40 0
90 40 0
100 100 60
Day 3 Toxin B Day 1 Day 2 Crude toxin Toxin A preparation" S. Iwulardü Control S. Loultndii Control S. boulardü Control S. liotdardü Control S. boulardü Control
C. dlffecile supernatant was incubated in vitro with S, Gmdardü then toxin concentrations and biological activity were tested
Protocols
Toxin eonoentrationst and toxin treatments
TeetB 1. l~tcacv o~ S. boWmdli ox C. diffreik sur~erv~rexr corrr~o mxnv A ~uvn B. At~ in vitro rxrwre~r rr~ roxnvs w~a nvocat.~r~ nvro sac$ oR rs~ sace rASrRe~T® wrrtt S. boulmdü xacmv~ rr~ C. djJrcüe roxtNS Hv ow+t. Aouze
158 6
G. CORTHIER eJ al.
FIG .
1.
SCANNING
ELECTRON
MICROSCOPY OF THE JEJUNUM (HAR, 100 ~JIII~.
OF
TOXIN-FREE
CONTROL
MICE
The SEM examinations revealed that the aspect of the small intestinal mucosa of the untreated mice and the control mice only treated with S. boulardü was the same (Fig. 1). The small intestinal mucosa of the mice only treated with the toxins had an abnormal aspect . The villi appeared flabby and joined at the regions of contacts (Fig. 2) . The top of the villi had wrinkles (Fig. 3). These phenomena were mainly observed for the duodenum and the jejunum and to a lesser extent for the ileum of mice challenged with the highest concentration of toxins A and B. For the two other concentrations, the phenomena were occasionally observed . In contrast, the small intestinal mucosa of mice pretreated with S. boulardü and challenged with the toxin had a normal aspect with yeast cells usually grouped on the surface (Fig. 4). In some fields, the villi were free from yeasts (Fig. 5) but the mucosa was still quite similar to the toxin-free controls .
FIG . Z . GENERAL ASPECT OF TJJE JEJUNUM OF MICE INOCULATED WITH THE MOST CONCENTRATED CRUDE TOXIN PREPARATION (HAR, ZOO FUII~.
Oral S. boulardü Treatment
158 7
FIG. 3 . WRINKLED ASPECT' OF THE VILW OF TFIE JFJIJNUM OF MICE INOCULATED WITH THE MOST' CONCENTRATED CRUDE TOXIN PREPARATION (BAR, TOO
ulll).
DISCUSSION
The results corroborate the previously described ability of S. boulardü to protect mice against the pathological affects caused by administering whole bacteria (CASTex et al ., 1990 ; CORTFIIER et al., 1986) . As in the last reported study (CASTER et al ., 1990), here again the SEM observations showed that the small intestinal mucosa remained intact when the mice were pretreated with S. boulardü . Since we also demonstrated in the present report that the yeast did not need to act directly on C. diffJCile toxins to exert its protective effect, the other hypothesis that can be put forth to explain this phenomenon is that the yeast acts on the intestinal mucosa. When only toxins were administered to mice, death occurred rapidly ; and since the gut of the moribund mice showed no marked injury but only flabby villi that appeared to fuse,
FiG. 4 . SCANNING ELECTRON MICR08COPY OF TIC JEJUNUM OF S. bouialdii PROTECTED MICE SHOWING A NORMAL MUCOSA WITH GROUPS OF YEAST CELLS (HAR, TOO um).
158 8
G.
FIG .
S.
CORTHIER et
al.
SCANNING ELECTRON MICROSCOPY OF THE JFJUNUM OF S. boulardü PROTECTED MICE SHOWING NORMAL V1LLI, FREE OF YEASTS (BAR, IOOltm) .
it is possible that the toxins crossed the mucosa and caused death by systemic action. In protected mice, S. boulardü may also exert a direct action rendering the enterocytes impermeable to toxins . It has been shown that the yeast prevents C. docile toxins from damaging intestinal cells in culture (CZERUCKA et al., 1991). Other mechanisms that have been proposed to explain the protection of S. boulardü include stimulation of the immune system (Burs et al., 1990) and modification of the toxin A brush border receptor by activation of intestinal enzymes (Burs et al., 1986). Furthermore, S. boulardü might also cause a reduction in toxin A receptors as has been reported for the digestive microflora (Lucas et al., 1989). In conclusion, the present data indicate that even under the drastic conditions of our model where the mice received the crude toxin preparation directly in the stomach, S. boulardü proved to be a highly effective protective agent. Acknowledgement-The
authors thank Dr
S.
L. SALHI for critical reading
of the
manuscript.
REFERENCES
J . G ., MooN, N ., CHANG, T . W., TAYLOR, N . and ONDERDONK, A. B . (1978) Role of Clostridium docile in antibiotic-associated pseudomembranous colitis . Gastroenterology 75, 778-782 . Burs, J . P., BERNASCONI, P., VAN CRAYNFST, M . P., MALDAGUE, P. and DE MEYER, R . (1986) Response of human and rat small intestinal mucosa to oral administration of S . boulardü. Pediatr . Res . 20, 192-196 . Burs, J. P ., BERNASCONI, P ., VAERMAN, J . P . and DtvE, C . (1990) Stimulation of secretory IgA and secretory component of immunoglobulins in small intestine of rats treated with S. boulardü. Dig. Dis. Sci. 35, 231-236. CASTEX, F ., CORTHIER, G ., JOUVERT, S ., ELMER, G. W ., LUCAS, F . and BAS'rIDE, M . (1990) Prevention of Clostridium di,~Jicile-induced experimental pseudomembranous colitis by Saccharomyces boulardü: a scanning electron microscopic and miaobiological study . !. Gen . Mùrobiol. I36, 1085-1089. CORTHIER, G ., DUBOS, F. and RAIBAUD, P . (1983) Modulation of cytotoxin production by C. docile in the intestinal tracts of gnotobiotic mice inoculated with various human intestinal bacteria. Appl . Envir. Micrabiol. 49, 250-232 . CORTt-igR, G ., Dueos, F . and DUCLUZEAU, R . (1986) Prevention of Clostridium difficile induced mortality in gnotobiotic mice by Saccharomyces boulardü. Can . J. Microbiol. 32, 894-896. Coxrl-nER, G ., MULLER, M . C., ELMER, G . W . and LucAS, F . (1989) Interrelationships between digestive proteolytic activities and production and quantitation of toxin in pseudomembranous colitis induced by Clostridium difficile in gnotobiotic mice . Inject Immw. 57, 3922-3927. BARTLEIT,
Oral S. boulardil Treatment
1589
Czsauctce, D., Nexo, J. L., Bi:ariescoxr, P. and Rer~er., P. (1991) Protective effects of Saccharomyces boulardü against the action of Clostrldhon difficile toxine. Gattroenterol. Clin . Biol. 15, 22-27. Czueatraxr, C. J., Jora~sox, W. J., Ber.rsEt, E. and Wn .xnvs, T. D. (1983) Pseudomembraaous colitis in Closrridhon difficile monoaseoeiated rata . I>!Ject Immun. 39, 1368-1376. Et .t~x, G. W. and McFeauxn, L. V. (1987) Supprea~on by Saccharwnyces boulardü of toxigenic Clostridhon diJfictk overgrowth after vancomycin treatment in hamsters . Antimicrob . Agents. Chemother. 31, 129-131 . GHOAüE, W. L., Snnoxns, J. M., Dusoctt, F., Bitowx, J. D., Aaear, Y., Srm~ieGewe, N., IC~a~v, M. R. B., Atexexnne-Wrr.r.uVS, J. and Buaoox, D. W. (1978) Identification of Clostridhaw difficile as a cause of pseudomembranoue colitis. Br Mcd. 1, 695. Lensox, H. E., Hoxous, P., Parce, A. B. and BoRA>Er.rA, S. P. (1978) Clostridium diffcile and the aetiology of pseudomembranous colitis. Lmicet i, 1063-1066. l:.mev, J. M., DoxTe, S. T. and Wn rigs, T. D. (1982) Effects of two toxins of C. di,~cile in antibiotio-aseociated acids in hamsters . Inject . Immun. 36, 822 29 . Lues, F., Eras, G. W., Bxor-Leaoct~, E. and Coarrt~a, G. (1989) Fixation of Clostridlion diffcile toxin A and cholera toxin to intestinal brush border membranes from axenic and conventional mice. Inject Immun. 57, 1680-1683. Lvear.v, D. M., Knmex, H. C. and Wn .s:nvs, T. D. (1988) Clostridüon difficile: its disease and toxins. Clin. Microbial. Rev. 2, 1-118. Mew, S., Coxrr~e, G. and Drmos, F. (1987) Effect of various diets on toxin production by two strains of Clostridium difficile in gnotobiotic mice . Infect . Immun. 55, 1801-1805. Messar, J., Sexc~z, O., Coucrrv, R., AsTOrx, J. and PeAOnr, A. L. (1984) Bacterio-pharmacological activity of Saccharomyces boulardü in clindamycin-induced colitis in the hamster. ArzneLnittelforscMoig 34, 794-797 . PorxouLAKL4, C., Twenetrsn.orouwa, G., Ctea¢, M., Frtexzateu, C. and Lessoxr, J. T. (1986) Clostridiion difficile cytotoxin inhibits protein synthesis in fibroblasts and intestinal mornes . Gastroenterology 91, 1147-1153. S~"~?, J. D. (1968) Fatal enterocolitis in hamsters given lincomycin hydrochloride. Lab. Mi. Care 18, 4l 1-420. Stn.uvex, N. M., PetrEr-r', S. and Wtr.xnvs, T. D. (1982) Purification and characterization of toxins A and B of Clostridium difficile . Infect . Immun. 35, 1032-1040. Suaewicz, C. M., Et.~e, G. W., Sr~r.itex, P., McFeauxn, L. V., Cxrx, J. and Vex-Ber.r~, G. (1989x) Prevention of antibiotio-assaciated diarrhoea by Saccharomyces boulardii: a prospective study. Gastroenterology 96, 981=988. Suxewrcz, C. M., McFeatexn, L. V., Er.t~t, G. W. and Crnxx, J. (1989b) Treatment of recurrent Clostridhon doffcüe colitis with vancomycin and Saccharomyces boulardü. Am. I. Gastrocnt. 84, 1285-1287. Toortter®t, R. D. and Er.e~rt, G. W. (1984) Prevention of clinaamycin-induced mortality in hamsters by Saccharomyces boulardii. Mtimicrob. Agents Chemother. 26, 552-556. Tarenerrn.orour.os, G., Porxour.ears, C., O'Bnw, M. J. and LewoxT, J. T. (1987) Differential effects of Clostridiwn diff tilt toxins A and B on rabbit ileum. Gastroenterology 93, 273-279.