ARTICLE IN PRESS
Anaerobe 9 (2003) 161–163
Clostridium difficile and enterotoxigenic Bacteroides fragilis strains isolated from patients with antibiotic associated diarrhoea$ ! H. Pituch, P. Obuch-Woszczatynski, M. Łuczak, F. Meisel-Miko"ajczyk* Department of Medical Microbiology, Medical University of Warsaw, Warsaw, Poland Received 28 August 2002; received in revised form 9 July 2003; accepted 15 July 2003
Abstract From the fecal samples of 332 patients with a clinical diagnosis of antibiotic associated diarrhoea (AAD), 131 Clostridium difficile strains were isolated. For detection of toxin A in the isolated strains the enzymatic immunoassay was used. The cytopathic effect was determined on McCoy cell line. PCR was used for the detection of non-repeating and repeating sequences of toxin A gene and nonrepeating sequences of toxin B gene. One hundred and six isolated C. difficile strains were TcdA+TcdB+, 10 strains TcdA TcB+ and 15 were non-toxigenic TcdA TcdB . Out of the same fecal samples 50 Bacteroides fragilis strains were isolated. All B. fragilis strains were tested in PCR reaction for fragilysine gene detection (bft). In 9 strains (18%) this gene was detected and the strains could be assumed as enterotoxigenic Bacteroides fragilis (ETBF). In 4 fecal samples toxigenic C. difficile (TcdA+TcdB+) was found simultaneously with ETBF. One sample contained C. difficile (TcdA TcdB+) and ETBF. Out of 4 fecal samples only ETBF was isolated. The cytotoxicity of ETBF strains was tested on HT29/C1 human colon carcinoma cell line. The cytotoxicity titer in the range of 20 and 80 was observed. r 2003 Elsevier Ltd. All rights reserved. Keywords: Clostridium difficile; AAD; PCR; Bacteroides fragilis (ETBF)
1. Introduction It is well known that the gut ﬂora containing different bacteria (where anaerobes are outnumbering other organisms by a ratio of 1000:1) protect the human organism against the pathogenic agents (colonisation resistance factor). The worldwide increase in the application of different antimicrobials leads to the suppression of the normal gut ﬂora of the treated patients  what may lead to antibiotic associated diarrhoea (AAD), and even to psedomembranous colitis (PMC), where Clostridium difficile is recognised as the main etiological agent . Normal gut ﬂora contains different Gram-positive and Gram-negative anaerobes. Gram-negative rod Bacteroides fragilis is a member of the normal gut ﬂora. $ Paper from Anaerobe Olympiad 2002. The 6th Biennial Congress of the Anaerobe Society of the Americas, Park City, Utah, 29 June– 2 July 2002. *Corresponding author. Instytut Biostruktury, Zaklad Mikrobiologii Lekarskiej, Akademia Medyczna w Warszawie ul Chalubinskiego 5, 02-004 Warszawa, Poland. Tel./fax: +48-22-628-2739. E-mail address: [email protected]
1075-9964/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S1075-9964(03)00119-7
This micro-organism, producing many different virulence factors can also act as an etiological agent of different pyogenic infections . Since 1984  it is known that several B. fragilis strains are producing enterotoxin as a new virulence factor. Those strains since 1987 are observed as etiological agents of human diarrhoea . In 1999  we have reported the isolation of enterotoxin (fragilysine) producing B. fragilis strains together with C. difficile from fecal samples of patients with a clinical diagnosis of AAD. Up to now time from the 50 fecal samples we have isolated 26 C. difficile strains. Eighteen of them were toxin producing (TcdA+TcdB+). From the same fecal samples 17 B. fragilis strains were isolated, 4 of which produced fragilysin (enterotoxigenic Bacteroides fragilis—ETBF). In three samples toxin-positive C. difficile was cultured simultaneously with ETBF. Since the number of patients with a clinical diagnosis of AAD increased worldwide recently and the number of fecal samples submitted for testing of C. difficile also increased, we decided to search for the reasons. The association between C. difficile and enterotoxin producing B. fragilis in AAD remained unclear and needed
ARTICLE IN PRESS 162
H. Pituch et al. / Anaerobe 9 (2003) 161–163
further investigations. As our ﬁrst observations were done on a relatively small number of strains, in the present study we decided to analyse a greater number of fecal samples and isolated strains. The aim of this study was the isolation of the C. difficile and B. fragilis (ETBF) strains from another set of fecal samples of patients with a clinical diagnosis of AAD in order to compare the results with our earlier observations.
in the cytotoxin assays and speciﬁc PCR tests for the detection of non-repeating toxin gene sequences. One C. difficile strain GAI 95 601 (TcdA TcdB+) isolated in Institute of Anaerobic Bacteriology, Gifu University School of Medicine, Japan, was used as a control for the detection of repeating sequences in the tcdA gene. In addition, two reference B. fragilis strains were included: one enterotoxigenic ATCC 43858 (ETBF) and one nonenterotoxigenic IPL E 323 (NTBF). 2.4. Bacterial toxins detection
2. Material and methods 2.1. Clinical samples During a 2-year-period (2000–2001) 332 fecal samples from patients suspected of AAD were submitted to our anaerobe unit with a request from clinicians to examine for C. difficile. The clinical criteria of an AAD were: loose or liquid stools, prolonged hospitalisation, antibiotic therapy (immunosuppression), severe underlying disease and a lack of another enteric pathogen. Two hundred and seventy-two stool samples from adults hospitalised in different wards (transplantology=86, general surgery=52, internal medicine=60, orthopedic surgery=40) of the University hospital and 60 fecal samples from children hospitalised in the gastroenterology and the haematology wards of a paediatric hospital localised in a different area were tested for the presence of C. difficile. 2.2. Isolation of bacterial strains Stool samples for detection of C. difficile were inoculated on a selective medium (CCCA–cycloserine– cefoxitine–amphotericin B) (bio-Merieux, Marcyl’Etoile, France). Plates were incubated anaerobically at 37 C for at least 96 h. Isolates were identiﬁed as C. difficile by a characteristic morphology of the colonies, speciﬁc horse odour, green-yellow ﬂuorescence under UV light, Gram staining and biochemical test Api-20 A (bio-Merieux, Marcy-l’Etoile, France). Simultaneously fecal samples were cultured on a selective BBE (Brucella Bile Esculine) medium. Plates were incubated anaerobically at 37 C. Isolates were identiﬁed as B. fragilis on the basis black colour of the colonies, Gram staining and biochemical test Api-20 A (bio-Merieux, Marcy-l’Etoile, France). 2.3. Bacterial strains Three reference C. difficile strains were used for comparative reasons. The reference set consisted of a single toxigenic C. difficile strain VPI 10 463 (TcdA+TcdB+), a non-toxigenic C. difficile strain NIH BRIGGS 8050 (TcdA TcdB ) as control strains
The presence of C. difficile toxin A directly in the fecal sample and in the isolated strain was determined using EIA test the ‘‘Culturette Brand Toxin CD’’ for detection of toxin A (Becton-Dickinson, Meyland, France). Toxin B (cytotoxin) of C. difficile was tested on McCoy cell line using supernatants of 48 h cultures on BHI medium. Brieﬂy, one colony of a C. difficile strain was used to inoculate BHI medium and incubated anaerobically. Supernatants were collected and tenfold serial dilutions of the culture ﬁltrate were added in duplicate to McCoy cells and incubated for 24 h at 37 C in a 5% CO2 atmosphere. CPE was observed by inverse microscopy . B. fragilis enterotoxin (fragilysin) was detected in BHI culture supernatants on human carcinoma cell line HT29/C1 (cytotoxicity test). Brieﬂy, serial double dilutions of the supernatants were added to the HT29/ C1 cell line and incubated. The results were read after 4 h of incubation. Last dilution causing a characteristic rounding and detachment of more than 50% of cells was considered as a titer of the enterotoxin. 2.5. Study of the DNA isolated from tested C. difficile and B. fragilis strains For PCR analysis of the tested strains DNA isolation was performed using Genomic DNA PREP PLUS (A&A Biotechnology, Poland). DNA ampliﬁcation was performed in a Techne thermocycler with selected pair of primers. Determination of non-repeating sequences of tcdA and tcdB genes of C. difficile. For tcdA gene detection primers YT28 (GCA TGA TAA GGC AAC TTC AGT GG) and YT29 (GAG TAA GTT CCT CCT GCT CCA TCAA) were used. For tcdB gene detection primers YT17 (GGT GGA GCT TCA ATT GGA GAG) and YT18 (GTG TAA CCT ACT TTC ATA ACA CCAG) were used. Throughout these PCR studied the reaction mixtures were heated at 94 C for 2 min, plus 35 cycles of 94 C (45 s), 55 C (30 s) and 70 C (45 s) [8,9]. Determination of repeating sequences of gene A (tcdA) of C. difficile was done using primers NK9 (CCA CCA GCT GCA GCC ATA) and NKV 011
ARTICLE IN PRESS H. Pituch et al. / Anaerobe 9 (2003) 161–163
(TTT TGA TCC TATAGA ATC TAA CTT AGT AAC), according to Kato et al. . Detection of the B. fragilis enterotoxin (fragilysin) gene was performed using primers 404 (GAG CCG AAG ACG GTG TAT GTG ATT TGT) and 407 (TGC TCA GCG CCC AGT ATA TGA CCT AGT). In this PCR program the reaction mixtures were heated: 94 C for 4 min, 40 cycles of 94 C (1 min), 52 C (1 min) and 74 C (1 min) .
3. Results Out of 332 fecal samples of diarrheic patients with a suspicion of an antibiotic associated diarrhea submitted to our anaerobic unit 76 were positive for toxin A C. difficile. From these clinical materials 131 C. difficile strains were isolated. Among these C. difficile strains 106 were TcdA+TcdB+ as demonstrated by C. difficile Toxin A Test ‘‘Culturette Brand Toxin Test’’ and cytotoxicity test on McCoy cells line. Remaining 25 strains were toxin A negative (TcdA ). Among these strains 10 were TcdA TcdB+ because supernatant from ﬂuid culture of these tested strains demonstrated a cytopathogenic effect (CPE) on McCoy cell line. PCR ampliﬁcation with YT28/YT29 and YT17/YT18 generated products of 630 and 399 bp for the TcdA and TcdB genes, respectively. For 10 TcdA TcdB+ strains PCR with the NK9/NKV011 primer set generated a 700 bp product similar to that obtained from the Japanese GAI 95 601 strains. Remaining 15 isolated C. difficile strains were non-toxigenic (TcdA TcdB ). Out of the same 332 fecal samples 50 B. fragilis (15%) strains were isolated. Nine were enterotoxigenic, what was conﬁrmed by the determination on human carcinoma cell line (HT29/C1) and using PCR for detection of fragilysine gene (bft). Among those 9 ETBF strains 5 were isolated from children and 4 from adults. All those strains possessing the fragilysine gene were toxigenic for the HT29/C1 line. The toxigenicity titer observed ranged between 20 and 80. During our observations in 4 cases (1.2%) ETBF was isolated together with toxigenic C. difficile producing toxins A and B. From one patient ETBF was isolated together with C. difficile strain which produced toxin B only (TcdA TcdB+). In 4 cases of patients with a clinical diagnosis of AAD only ETBF was cultured, without the presence of C. difficile.
4. Conclusions In our anaerobic unit the fecal samples of patients with a clinical diagnosis of AAD are tested routinely. In comparison to our previous observations reported in
1999, in the present study the number of samples and the number of isolated strains were much larger. Why is the reason we decided to compare the strains isolated during years 2000–2001 as outlined in the ‘‘material and methods’’. From fecal samples of patients with a clinical diagnosis of AAD not only toxigenic (TcdA+TcdB+) C. difficile strains but also strains producing toxin B without toxin A could be isolated. Those strains could be isolated alone as etiological agents of AAD and simultaneously with ETBF. From fecal samples of patients with a suspicion of AAD, ETBF can be isolated as the only etiological agents from children and adults. The importance of C. difficile strains toxin A negative in AAD needs further investigation. The association and interaction between toxin producing C. difficile and B. fragilis ETBF remains unclear and needs further study.
References  Borriello SP, Barclay FE. Anaerobes in intestinal homeostasis. In: Duerden BI, Drasar BS. editors. Anaerobes in human disease. London, Melbourne, Auckland: Edward Arnold, A Division of Hodder & Stoughton; 1991. p. 343–50.  Borriello SP. Anaerobic bacteria in diarrhoea and colitis. In: Duerden BI, Drasar B.S. editors. Anaerobes in human disease. London, Melbourne, Auckland: Edward Arnold, A Division of Hodder & Stoughton; 1991. p. 351–61.  Meisel-Miko"ajczyk F. Anaerobes today. Nova Acta Leopoldina NF 80, No. 312, 1999. p. 207–17.  Myers LL, Firehammer BD, Shoop an Border MM. A possible cause of acute diarrheal disease in newborn lambs. Infect Immun 1984;44:241–4.  Myers LL, Shoop DS, Stackhouse LL, Newman FS, Flaherty RJ, Letson GW, Sack RB. Isolation of enterotoxigenic Bacteroides fragilis from humans with diarrhea. J Clin Microbiol 1987;25: 2330–3. !  Meisel-Miko"ajczyk F, Leszczynski P, van Belkum A, Pituch H, ! Obuch-Woszczatynski P, Rouyan GhS. Enterotoxin-producing Bacteroides fragilis (ETBF) strains in stool samples submitted for testing of Clostridium difficile and its toxins. Anaerobe 1999;5: 217–9. !  Martirosian G, Pituch H, Obuch-Woszczatynski P, Rouyan Gh, Meisel-Miko"ajczyk F. Evaluation of different methods for detection of C. difficile toxins in Poland. Acta Microbiol Pol 1999; 48:349–53.  Weikel CS, Grieco FD, Rueben J, Myers LL, Sacks RB. Human colonic epithelial cells, HT 29/C1 treated with crude Bacteroides fragilis enterotoxin dramatically alter their morphology. Infect Immun 1992;60:321–7.  Khul SJ, Tang YJ, Navarro L, Gumerlock PH, Silva Jr J. Diagnosis and monitoring of Clostridium difficile infections with the polymerase chain reaction. Clin Infect Dis 1993;16(Suppl.): 234–8.  Tang YJ, Gumerlock PH, Weiss JB, Silva Jr J. Speciﬁc detection of Clostridium difficile toxin A gene sequences in clinical isolates. Mol Cellular Probes 1994;8:463–7.  Kato H, Kato N, Katow S, Muegawa T, Nakamura S, Lyerly D. Deletion in the repeating sequences of the toxin A gene of toxin A-negative, toxin B-positive C. difficile strains. FEMS Microbiol Lett 1999;175:197–203.