- Email: [email protected]
Nucleotide sequence of Clostridium difficile toxin A gene fragment and detection of toxigenic strains by polymerase chain reaction
FEMS MicrobiologyLetters 70 (1990) 1-6 Publishedby Elsevier
l
FEMSLE04014
Nucleotide sequence of Clostridium difficile toxin A gene fragment and detection of toxigenic strains by polymerase chain reaction Brendan W. Wren, Christopher L. Clayton and Soad Tabaqchali Department of Medical Microbiolo~,, St. Bartholomew'sHospital Medical College, West Smithfield. London, U.K. Received25 January1990 Revisionreceived9 February1990 Accepted10 February1990 Key words: CIostridium difficile; Toxin A; Nucleotide sequence; PCR
1. SUMMARY
2. INTRODUCTION
A 1947 base pair (bp) fragment of the toxin A gene of Clostridium difficile was sequenced. A continuous open reading frame was found, which contained 4 distinct groups of repeat nucleotide sequence with 88 to 100% identity within each group. The arrangement of the groups (A, 81 bp, B, C and D, 63 bp) was ABCCCDABCDDABCCCDABCCDABCDABC. Based on nucleotide sequence data from the C repeat group, a pair of oligonucleotide primers were synthesised and used in the polymerase chain reaction (PCR) to amplify fragments from the toxin A gene. Several products of multiples of 63 bp length were amplified for all 33 toxigenic C. difficile strains tested in contrast to the 12 non-toxigen-;c s:raJns tested which failed to amplify any product. This rapid technique is of potential use in the specific identification of toxigenie C. difficile strains in mixed culture and from clinical specimens.
Clostridium difficile is the aetiological agent of pseudomembraneous colitis and antibiotic-associated diarrhoea in humans. The pathogenicity of C. difficile is, in part, related to the production of at least two toxins, toxin A, an enterotoxin, and toxin B, a cytopathic toxin [1,2]. Currently, diagnosis depends on the isolation and identification of the organism, which takes up to 48 h, a n d / o r the demonstration of toxins in faecal specimens of patients using tissue culture or ELISA techniques. More rapid diagnosis is essential to enable prompt treatment. We have undertaken the cloning and sequencing of the toxin A and B genes to develop improved methods for the specific identification of toxigenic strains and to gain an insight into the molecular basis of C. difficile-related disease. Previously, we described the clone ~tA5 with a 14.3 kb insert which encodes the toxin A gene [3]. The protein product of at least 235 kDa expressed by ?~tA5 reacted with antisera to the purified toxin, agglutinated rabbit erythrocytes and had a cytopathic effect on Chinese hamster ovary cells, properties consistent with purified toxin A protein. A 4.5 kb Pstl subclone, designated pBWW47, was isolated, ¢~hich expressed a 140 kDa protein that
Correspondence to: BrendanW. Wren,Departmentof Medical Microbiology, St. Bartholomew'sHospital Medical College, West Smithfield,LondonEC1A7BE, U.K. Some of the work siren in this paper is the subjectof British patent applicationNo. 8929293.2
0378-1097/90/$03.50© 1990F~derationof EuropeanMicrobiologicalSocictics
retained haemagglutinating capacity but was no longer cytopathic. Sequencing of this fragment revealed 4 distinct groups of repeat nucleotide sequence. Recently, the polymerase chain reaction (PCR) has been shown to be useful for the detection of slow-growing organisms such as Mycobacterium tuberculosis [4,5] and Helicobacter pylori [6]. PCR requires nucleotide sequence data to design oligonucleotide primers for the polymerisation step. The aim of this study was to utilise nucleotide sequence data from tandemly arranged repeat groups to design suitable oligonucleotide primers for the PCR amplification of the toxin A gene, to aid in the development of a rapid, sensitive and specific method for the identification of toxigenic C. difficile strains.
3. MATERIALS AND METHODS
3.1. Isolation of DNA and restriction enzyme analysis DNA from clone ~tA5 containing the toxin A gene from C. difficile strain W1, was isolated using the plate lysate method [7]. DNA from the plasmid clone pBWW47 was isolated by alkaline lysis [8]. Both clones were mapped with a variety of restriction enzymes (Amersham International, Amersham, U.K.) by established procedures [7].
3.2. M13 cloning and DNA sequencing procedures DNA fragments within pBWW47 were subcloned in both orientations into M13 and both strands were sequenced by the dideoxy chain termination procedure [9]. Extended polyacrylamide gel runs were performed to confirm the arrangement of overlapping repeat groups of nucleotide sequence.
3.3. Bacterial strains Nine standard C. difficile strains from different typing groups [10,11] designated A to E and W to Z, were used for initial PCR experiments, Strains B, D, E, W, X and Z were previously shown to be toxigenic by a direct sandwich ELISA using antitoxin A [12] in contrast to the non-toxigenic strains
A, C and Y. Thirty six other C. difficile strains isolated from patients attending St. Bartholomew's Hospital including 9 non-toxigenic strains were tested in PCR experiments. Strains were grown anaerobically at 37 °C for 36 h on selective media (cycloserine, cefoxitin and fructose agar, Oxoid Ltd., Basingstoke, U.K.) and their identity confirmed by colonial morphology, smell, Gram stain and gas-liquid chromatographic analysis of volatile fatty acids [13]. Sixteen other clostridial strains [14] from 12 other species including 3 C. sordellii strains wece also studied by PCR analysis.
3.4. Synthetic oligonucleotides All oligonucleotides were synthesised on an Applied Biosystems ~ynthesiser with the automated phosphoramidite coupling method. Oligonucleotides used as primers were designed based on nucleotide sequence data from repeat group C (BW 69 5"-GAAGCAGCTACTGGATGGCA and BW 70 5'-AGCAGTGTTAGTATTAAAGT).
3.5. Preparation of samples for PCR and PCR methodology A single colony was scraped with a 9 inoculating loop into a 1.5 ml polypropylene tube containing 200 /~1 of sterile water. After boiling for 10 min the samples were spun at 14000 × g in a bencbtop microfuge lot 5 min. Two microlitres of supernatant liquid were added to a 100/~! reaction volume with 1.5 mM magnesium chloride, 10 mM tris-HC! (pH 8.3), 0.01% gelatin, 200/~M each of deoxyribonucleotides, 100 pM oligonucleotide primers and 2.5 U Taq polymerase (Perkin Elmer Cetus, California U.S.A.). Reaction mixtures were overlaid with paraffin oil (100 ~al), placed in a thermal cycler (Hybaid Ltd, Twickenham, Middlesex, U.K.) and amplified for 30 cycles. A rapid two step cycle of 94°C and 4 6 ° C for 30 s each was chosen with the first 94°C step extended to 3 min to ensure denaturation of the initial sample. Twenty microlitres of the amplified products were electrophoresed in a horizontal 4% Nusieve GTG (ICN Biomedicais Ltd, High Wycombe, U.K.) agarose gel containing 0.5 /~g ethidium bromide per ml and the bands were visualised by excitation under ultraviolet light.
To determine the sensitivity of the PCR in detecting toxigenic C. d i f f i c i l e strains in the presence of bacteria from a faecal specimen, the number of organisms present in a PCR sample was quantified by serial dilutions and a viable count on Columbia horse blood agar plates (Oxoid Ltd.).
4. RESULTS
DNA isolated from the clones ),tA5 and pBWW47 was mapped with a variety of restriction enzymes shown in Fig. 1. Sequencing of a 1947 bp internal fragment of the toxin A gene revealed the presence of 4 distinct groups of repeat nucleotide sequence (A, 81 bp, B, C and D, 63 bp) with 88 to 100% identity within groups. The arrangement of the groups which accounts for 1935 of the 1947 bp sequenced is showa in Fig. 2, and the other 12 bp were 2 1: king areas of 6 bp between the underlined A and D groups. Also shown in Fig. 2 is the consensus nucleotide sequence and translated peptide sequence for groups A, B, C and D. The 4 groups showed areas of similarity to each other, in particular, the central region of the B and C
I ' I ' I ~
iT;Tr
Okb
7 %_-_
I ' I ' 1 4 I~
Fig. 1. Partialrestrictioncleavagemap of ktA5and pBWW47. B= BamHl, E = EcoR!, Ev = E c o R V , H = HindIll, P = Pstl, S = Sau3A, X = Xbal and Xh = Xhol.
groups. Translation of the entire 1947 bp sequence (EMBL data bank accession No X17194) revelaed a continuous open reading frame (or0 confirming that the DNA encodes an internal region of a large polypeptide. From nucleotide sequence data oligonucleotide primers BW 69 and 70 were synthesised from the tandemly repeated C group. Using these primers in the PCR with toxigenic strains B, D, E, W, X and Z, revealed DNA bands of 63, 126, 189, 252 and 333 bp length which were absent from nontoxigenic strains, A, C and Y (Fig. 3). A total of 45 C . d i f f i c i l e strains were tested in the PCR and all 33 toxigenic strains were positive in contrast to the 12 non-toxigenic strains. Of the other 16
(i) ABCCCDABCDDABCCCDABCCDABCDABC
(ii) A = 5••RT•••••T•TTT•R••••CCT••••••TTT•••T•TTTTGC•••T•CT••T•CTT•T•RT••T••••T••••••TC••••T I G V F K G P K G F E Y F A P R H T Y N H H I E G Q A B =
5'-RTAGTTTATCAARGTRARTTCTTAACTTTrRATGGTAAARAATATTACTTTGATRRTRACTCA 3' I V Y Q S K F L T L N G K K Y Y F D H H S
C = 5'-GAAGCAGCTACTGGRTGGCAAACTATTGATGGTAAAAARTATTACTTTAATRCTAACACTGCT 3' E A A T G W Q T I D G K K Y Y F N T H T A D = 5'-ATAGCTTCAACTGGTTATACAATTATTAATGTTARACATTTTTATTTTAATACTGRTGGTATT 3' I R S T G Y T I I N G K H F Y F N T D G I
~&2~i)~t~f~u~tid¢~at~psA~B~C~dD~(ii)~n~susnuc~tid~and~tid~u~n~r~at~ps A,B,C~dD.
3'
4
M
A
B
C
D
E
W
X
Y
Z
1
2
3
Fig. 3. PCR of toxiganicand non-toxigenieC. difficile strains using ofigonucleotideprimers BW 69 and 70: Lanes B, D, E, W, X, and Z, toxigenic C. difficile strains from pure colonies,lanes A, C and Y non-toxigenicC. dlffici/estrains from pure colonies, lanes 1, 2 and 3 serialdilution of C. difficile strain W with 300, 30 and 3 organismsrespectivelymixed with l0 Tother anaerobic enteric bacteria. Lane M: DIqA size marker of visiblebands, 67, 80,110,147,190, 242 and 267 bp (Bshl and Mspl digests of pHC314). clostridial strains from 12 different species tested, 3 C. sordellii strains were positive (Table 1). To determine the number of bacteria necessary to give visible bands on agarose gels after PCR, samples containing known numbers of bacteria from viable counts were serially diluted and subjected to amplification. Visible bands were seen Table 1 Summary of PCR experimentswith various clustridial strains
Toxigenic C difficile Non-toxigenie C. ¢hfficile Other Clostridial species
No. tested
PCR positive
PCR negative
33
33
0
12
0
12
16
3
13
with as few as 300 0ane 1, Fig. 3) and 30 (lane 2, Fig. 3) toxigenic bacteria even in the presence of 107 anaerobic enteric bacteria.
5. D I S C U S S I O N This study has revealed within the toxin A gene the presence of several repeat groups of nucleotide sequence which translate into groups of peptide sequence similar to those reported by Johnson et al. [15] for C. difficile strain VPI 10463. The expressed product of clone pBWW47 which includes the area of D N A sequenced retained the ability to agglutinate rabbit erythrocytes (unpublished results), thus the peptide portion of the toxin A gene encoded by the repeat sequence may play a role in the specificity of the toxin protein to bind to receptor molecules on the brush border
membrane of the human intestine. Interestingly, a haemagglutinin from Myxococcus xanthus has been shown to contain 4 highly conserved repeat groups of 67 amino acid residues [16]. The presence of at least 11 C repeat groups of nucleotide sequence may have heightened the sensitivity of the PCR which could specifically detect the presence of as few as 30 toxin-producing bacteria in the presence of 107 other bacteria. This sensitivity is approximately 3 times greater than that reported by Karch and Meyer [17] for the detection of Shiga-like toxin-producing E. coli strains using PCR methodology. The PCR products of 63, 126, and 189 bp seen in Fig. 3 can be explained by the amplification of D N A between the 3 adjacent C groups. However, amplification" of the fourth 252 bp band, which was clearly observed for all toxigenic strains tested, is difficult to explain from the arrangement of the repeat groups as 4 adjacent C groups or 2 C groups separated by 2 other 63 bp groups are not apparent. It is possible that such an arrangement of repeat groups exists outside the area of sequence reported in this study. The presence of the fifth band of approximately 333 bp can be explained by the 2 C groups separated by D, A and B groups. The profiles from the PCR products appeared identical for all 35 toxigenic strains tested which were from a variety of typing groups suggesting that the area of D N A to which the primers annealed are highly conserved between different toxigenic strains. The variation in PCR conditions and *,he use of new pairs and various combinations of primers may reveal further banding patterns which apart from their diagnostic potential should prove useful in typing different C. difficile strains. The failure to detect D N A amplification products from all 12 non-toxigenic strains tested can be explained by the absence of at least part of the toxin A gene. These observations have been confirmed by hybridisation studies using the 4.5 kb Pstl fragment as a toxin A gene-specific probe [18]. The positive PCR results obtained with the C. sordellii strains were not surprising as the organism has a haemorrhagic toxin which crossreacts immunologically [19] and shows D N A homology with the toxin A gene [18,20].
The PCR experiments in this study were designed for the rapid aniplification of target DNA. The 7 2 ° C elongation step normally incorporated in the PCR was omitted as the relatively small length of D N A required to be synthesised could be completed by the Tat/polymerase as the 46 ° C to 9 4 ° C cycle transiently passes the optimum temperature for polymerisation. The 10 rain boiling of the samples released enough bacterial D N A for the PCR to yield positive results. This rapid and convenient method requires no D N A extraction procedure in contrast to that reported for the PCR on mycobacteria [5]. The PCR-based method described in this study, which does not require hybridisation technology o2 the use of radioactivity, should be useful for the rapid identification of toxigenic C. difficile strains from pure and mixed culture and, with suitable adaptations, could be applied directly to clinical specimens.
ACKNOWLEDGEMENTS We thank John Keyte for the synthesis of oligonucleotides. This work was supported by the Wellcome Trust, U.K.
REFERENCES [1] Bartlett,J.G., Onderdonk, A.B., Cisners, R.L. and Kasper, D.L. (1977) J. Infect. Dis. 136, 701-705. [2] Taylor, N.S., Thome, G.N. and Bartlett, J.G. (1981) Infect. lmmun. 34,1036-1043. [3] Wren, B.W., Clayton. C.L., Mullany, P. and Tabaqchali, S. (1987) FEBS Lett. 225, 82-86. [4] Hance,AJ., Grandchamp, B. and Levy-Frebault,V. (1989) Molec. MicrobioL3, 843-849. [5] Brisson-Noel, A., Gicquei, B., Lecossier, D., LevyFrebault, V., Nassif, X. and Hance, AJ. (1989) Lancet it, 1069-1071. [6] Clayton, C.L., Kleanthous, H., Wren, B.W. and Tabaqchali, S. (1990) J. Med. Microbiol.(in press). [7] Maniatis, T., Fritsch, E. and Sambrook. J. (1982) Molecular cloning: a laboratory approach. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. [8] Bimboim,H.C. and Doly, J. (1979) Nucleic Acids Res. 7, 1514-1523. [9] Sanger, F., Nicklen, S. and Coulson, A.R. (1977) P~oc. Natl. Acad. Sci. U.S.A.74, 5463-5467.
[10] Tabaqchali, S., O'FarreU, S., Holland, D. and Silman, R. (1984) Lancet i, 935-938. [11] Wren, B.W. and Tabaqchali, S. (1987) J. Clin. Microbiol. 25, 2402-2404. [12] Wren, B.W., Heard, S. and Tabaqchali, S. (1987) 3. Clia. Pathol. 40, 1397-1401. [13] Holderman, L.V., Cato, E.P. and Moore, W.E.C. (1977) Anaerobe laboratory manual, 4th edition. Blacksburg Virginia Polytechnic and State University. [141 Wren, B.W., Mullany, P.M., Clayton, C.L. and Tabaqchali, S. (1988) ~EMS Microbiol. Immunol. 47,163-168. [15] John: ~±n, J.L., Dove, C.H., Price, S.B., Sickles, T.W., Pbelp'i, C.J. and Wilkins, T.D. (1988) In Anaerobes today
(Hardie, J.M. and Borriello, S.P., eds.), pp. 115-123, Wiley, Chichester, U.K. [16] Romeo, J.M., Esmon, B. and Zusman, D.R. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 6332-6336. [17] Karch, H. and Meyer, T. (1989) J. Clin. Microbiol. 27, 2751-2757. [18] Wren, B.W., Castledine, N.B., Clayton, C.L. and Tabaqchali, S. (1990) J. Med. Microbiol. (in press). [19] Martinez, R.D. and Wilkins, T.C. (1988) Infect. Immun 56,1215-1221. [20] Price, S.B., Phelps, C.J., Wilkins, T.D. and Johnson, J.J. (1987) Cuff. Microbiol. 16, 82-86.
l
FEMSLE04014
Nucleotide sequence of Clostridium difficile toxin A gene fragment and detection of toxigenic strains by polymerase chain reaction Brendan W. Wren, Christopher L. Clayton and Soad Tabaqchali Department of Medical Microbiolo~,, St. Bartholomew'sHospital Medical College, West Smithfield. London, U.K. Received25 January1990 Revisionreceived9 February1990 Accepted10 February1990 Key words: CIostridium difficile; Toxin A; Nucleotide sequence; PCR
1. SUMMARY
2. INTRODUCTION
A 1947 base pair (bp) fragment of the toxin A gene of Clostridium difficile was sequenced. A continuous open reading frame was found, which contained 4 distinct groups of repeat nucleotide sequence with 88 to 100% identity within each group. The arrangement of the groups (A, 81 bp, B, C and D, 63 bp) was ABCCCDABCDDABCCCDABCCDABCDABC. Based on nucleotide sequence data from the C repeat group, a pair of oligonucleotide primers were synthesised and used in the polymerase chain reaction (PCR) to amplify fragments from the toxin A gene. Several products of multiples of 63 bp length were amplified for all 33 toxigenic C. difficile strains tested in contrast to the 12 non-toxigen-;c s:raJns tested which failed to amplify any product. This rapid technique is of potential use in the specific identification of toxigenie C. difficile strains in mixed culture and from clinical specimens.
Clostridium difficile is the aetiological agent of pseudomembraneous colitis and antibiotic-associated diarrhoea in humans. The pathogenicity of C. difficile is, in part, related to the production of at least two toxins, toxin A, an enterotoxin, and toxin B, a cytopathic toxin [1,2]. Currently, diagnosis depends on the isolation and identification of the organism, which takes up to 48 h, a n d / o r the demonstration of toxins in faecal specimens of patients using tissue culture or ELISA techniques. More rapid diagnosis is essential to enable prompt treatment. We have undertaken the cloning and sequencing of the toxin A and B genes to develop improved methods for the specific identification of toxigenic strains and to gain an insight into the molecular basis of C. difficile-related disease. Previously, we described the clone ~tA5 with a 14.3 kb insert which encodes the toxin A gene [3]. The protein product of at least 235 kDa expressed by ?~tA5 reacted with antisera to the purified toxin, agglutinated rabbit erythrocytes and had a cytopathic effect on Chinese hamster ovary cells, properties consistent with purified toxin A protein. A 4.5 kb Pstl subclone, designated pBWW47, was isolated, ¢~hich expressed a 140 kDa protein that
Correspondence to: BrendanW. Wren,Departmentof Medical Microbiology, St. Bartholomew'sHospital Medical College, West Smithfield,LondonEC1A7BE, U.K. Some of the work siren in this paper is the subjectof British patent applicationNo. 8929293.2
0378-1097/90/$03.50© 1990F~derationof EuropeanMicrobiologicalSocictics
retained haemagglutinating capacity but was no longer cytopathic. Sequencing of this fragment revealed 4 distinct groups of repeat nucleotide sequence. Recently, the polymerase chain reaction (PCR) has been shown to be useful for the detection of slow-growing organisms such as Mycobacterium tuberculosis [4,5] and Helicobacter pylori [6]. PCR requires nucleotide sequence data to design oligonucleotide primers for the polymerisation step. The aim of this study was to utilise nucleotide sequence data from tandemly arranged repeat groups to design suitable oligonucleotide primers for the PCR amplification of the toxin A gene, to aid in the development of a rapid, sensitive and specific method for the identification of toxigenic C. difficile strains.
3. MATERIALS AND METHODS
3.1. Isolation of DNA and restriction enzyme analysis DNA from clone ~tA5 containing the toxin A gene from C. difficile strain W1, was isolated using the plate lysate method [7]. DNA from the plasmid clone pBWW47 was isolated by alkaline lysis [8]. Both clones were mapped with a variety of restriction enzymes (Amersham International, Amersham, U.K.) by established procedures [7].
3.2. M13 cloning and DNA sequencing procedures DNA fragments within pBWW47 were subcloned in both orientations into M13 and both strands were sequenced by the dideoxy chain termination procedure [9]. Extended polyacrylamide gel runs were performed to confirm the arrangement of overlapping repeat groups of nucleotide sequence.
3.3. Bacterial strains Nine standard C. difficile strains from different typing groups [10,11] designated A to E and W to Z, were used for initial PCR experiments, Strains B, D, E, W, X and Z were previously shown to be toxigenic by a direct sandwich ELISA using antitoxin A [12] in contrast to the non-toxigenic strains
A, C and Y. Thirty six other C. difficile strains isolated from patients attending St. Bartholomew's Hospital including 9 non-toxigenic strains were tested in PCR experiments. Strains were grown anaerobically at 37 °C for 36 h on selective media (cycloserine, cefoxitin and fructose agar, Oxoid Ltd., Basingstoke, U.K.) and their identity confirmed by colonial morphology, smell, Gram stain and gas-liquid chromatographic analysis of volatile fatty acids [13]. Sixteen other clostridial strains [14] from 12 other species including 3 C. sordellii strains wece also studied by PCR analysis.
3.4. Synthetic oligonucleotides All oligonucleotides were synthesised on an Applied Biosystems ~ynthesiser with the automated phosphoramidite coupling method. Oligonucleotides used as primers were designed based on nucleotide sequence data from repeat group C (BW 69 5"-GAAGCAGCTACTGGATGGCA and BW 70 5'-AGCAGTGTTAGTATTAAAGT).
3.5. Preparation of samples for PCR and PCR methodology A single colony was scraped with a 9 inoculating loop into a 1.5 ml polypropylene tube containing 200 /~1 of sterile water. After boiling for 10 min the samples were spun at 14000 × g in a bencbtop microfuge lot 5 min. Two microlitres of supernatant liquid were added to a 100/~! reaction volume with 1.5 mM magnesium chloride, 10 mM tris-HC! (pH 8.3), 0.01% gelatin, 200/~M each of deoxyribonucleotides, 100 pM oligonucleotide primers and 2.5 U Taq polymerase (Perkin Elmer Cetus, California U.S.A.). Reaction mixtures were overlaid with paraffin oil (100 ~al), placed in a thermal cycler (Hybaid Ltd, Twickenham, Middlesex, U.K.) and amplified for 30 cycles. A rapid two step cycle of 94°C and 4 6 ° C for 30 s each was chosen with the first 94°C step extended to 3 min to ensure denaturation of the initial sample. Twenty microlitres of the amplified products were electrophoresed in a horizontal 4% Nusieve GTG (ICN Biomedicais Ltd, High Wycombe, U.K.) agarose gel containing 0.5 /~g ethidium bromide per ml and the bands were visualised by excitation under ultraviolet light.
To determine the sensitivity of the PCR in detecting toxigenic C. d i f f i c i l e strains in the presence of bacteria from a faecal specimen, the number of organisms present in a PCR sample was quantified by serial dilutions and a viable count on Columbia horse blood agar plates (Oxoid Ltd.).
4. RESULTS
DNA isolated from the clones ),tA5 and pBWW47 was mapped with a variety of restriction enzymes shown in Fig. 1. Sequencing of a 1947 bp internal fragment of the toxin A gene revealed the presence of 4 distinct groups of repeat nucleotide sequence (A, 81 bp, B, C and D, 63 bp) with 88 to 100% identity within groups. The arrangement of the groups which accounts for 1935 of the 1947 bp sequenced is showa in Fig. 2, and the other 12 bp were 2 1: king areas of 6 bp between the underlined A and D groups. Also shown in Fig. 2 is the consensus nucleotide sequence and translated peptide sequence for groups A, B, C and D. The 4 groups showed areas of similarity to each other, in particular, the central region of the B and C
I ' I ' I ~
iT;Tr
Okb
7 %_-_
I ' I ' 1 4 I~
Fig. 1. Partialrestrictioncleavagemap of ktA5and pBWW47. B= BamHl, E = EcoR!, Ev = E c o R V , H = HindIll, P = Pstl, S = Sau3A, X = Xbal and Xh = Xhol.
groups. Translation of the entire 1947 bp sequence (EMBL data bank accession No X17194) revelaed a continuous open reading frame (or0 confirming that the DNA encodes an internal region of a large polypeptide. From nucleotide sequence data oligonucleotide primers BW 69 and 70 were synthesised from the tandemly repeated C group. Using these primers in the PCR with toxigenic strains B, D, E, W, X and Z, revealed DNA bands of 63, 126, 189, 252 and 333 bp length which were absent from nontoxigenic strains, A, C and Y (Fig. 3). A total of 45 C . d i f f i c i l e strains were tested in the PCR and all 33 toxigenic strains were positive in contrast to the 12 non-toxigenic strains. Of the other 16
(i) ABCCCDABCDDABCCCDABCCDABCDABC
(ii) A = 5••RT•••••T•TTT•R••••CCT••••••TTT•••T•TTTTGC•••T•CT••T•CTT•T•RT••T••••T••••••TC••••T I G V F K G P K G F E Y F A P R H T Y N H H I E G Q A B =
5'-RTAGTTTATCAARGTRARTTCTTAACTTTrRATGGTAAARAATATTACTTTGATRRTRACTCA 3' I V Y Q S K F L T L N G K K Y Y F D H H S
C = 5'-GAAGCAGCTACTGGRTGGCAAACTATTGATGGTAAAAARTATTACTTTAATRCTAACACTGCT 3' E A A T G W Q T I D G K K Y Y F N T H T A D = 5'-ATAGCTTCAACTGGTTATACAATTATTAATGTTARACATTTTTATTTTAATACTGRTGGTATT 3' I R S T G Y T I I N G K H F Y F N T D G I
~&2~i)~t~f~u~tid¢~at~psA~B~C~dD~(ii)~n~susnuc~tid~and~tid~u~n~r~at~ps A,B,C~dD.
3'
4
M
A
B
C
D
E
W
X
Y
Z
1
2
3
Fig. 3. PCR of toxiganicand non-toxigenieC. difficile strains using ofigonucleotideprimers BW 69 and 70: Lanes B, D, E, W, X, and Z, toxigenic C. difficile strains from pure colonies,lanes A, C and Y non-toxigenicC. dlffici/estrains from pure colonies, lanes 1, 2 and 3 serialdilution of C. difficile strain W with 300, 30 and 3 organismsrespectivelymixed with l0 Tother anaerobic enteric bacteria. Lane M: DIqA size marker of visiblebands, 67, 80,110,147,190, 242 and 267 bp (Bshl and Mspl digests of pHC314). clostridial strains from 12 different species tested, 3 C. sordellii strains were positive (Table 1). To determine the number of bacteria necessary to give visible bands on agarose gels after PCR, samples containing known numbers of bacteria from viable counts were serially diluted and subjected to amplification. Visible bands were seen Table 1 Summary of PCR experimentswith various clustridial strains
Toxigenic C difficile Non-toxigenie C. ¢hfficile Other Clostridial species
No. tested
PCR positive
PCR negative
33
33
0
12
0
12
16
3
13
with as few as 300 0ane 1, Fig. 3) and 30 (lane 2, Fig. 3) toxigenic bacteria even in the presence of 107 anaerobic enteric bacteria.
5. D I S C U S S I O N This study has revealed within the toxin A gene the presence of several repeat groups of nucleotide sequence which translate into groups of peptide sequence similar to those reported by Johnson et al. [15] for C. difficile strain VPI 10463. The expressed product of clone pBWW47 which includes the area of D N A sequenced retained the ability to agglutinate rabbit erythrocytes (unpublished results), thus the peptide portion of the toxin A gene encoded by the repeat sequence may play a role in the specificity of the toxin protein to bind to receptor molecules on the brush border
membrane of the human intestine. Interestingly, a haemagglutinin from Myxococcus xanthus has been shown to contain 4 highly conserved repeat groups of 67 amino acid residues [16]. The presence of at least 11 C repeat groups of nucleotide sequence may have heightened the sensitivity of the PCR which could specifically detect the presence of as few as 30 toxin-producing bacteria in the presence of 107 other bacteria. This sensitivity is approximately 3 times greater than that reported by Karch and Meyer [17] for the detection of Shiga-like toxin-producing E. coli strains using PCR methodology. The PCR products of 63, 126, and 189 bp seen in Fig. 3 can be explained by the amplification of D N A between the 3 adjacent C groups. However, amplification" of the fourth 252 bp band, which was clearly observed for all toxigenic strains tested, is difficult to explain from the arrangement of the repeat groups as 4 adjacent C groups or 2 C groups separated by 2 other 63 bp groups are not apparent. It is possible that such an arrangement of repeat groups exists outside the area of sequence reported in this study. The presence of the fifth band of approximately 333 bp can be explained by the 2 C groups separated by D, A and B groups. The profiles from the PCR products appeared identical for all 35 toxigenic strains tested which were from a variety of typing groups suggesting that the area of D N A to which the primers annealed are highly conserved between different toxigenic strains. The variation in PCR conditions and *,he use of new pairs and various combinations of primers may reveal further banding patterns which apart from their diagnostic potential should prove useful in typing different C. difficile strains. The failure to detect D N A amplification products from all 12 non-toxigenic strains tested can be explained by the absence of at least part of the toxin A gene. These observations have been confirmed by hybridisation studies using the 4.5 kb Pstl fragment as a toxin A gene-specific probe [18]. The positive PCR results obtained with the C. sordellii strains were not surprising as the organism has a haemorrhagic toxin which crossreacts immunologically [19] and shows D N A homology with the toxin A gene [18,20].
The PCR experiments in this study were designed for the rapid aniplification of target DNA. The 7 2 ° C elongation step normally incorporated in the PCR was omitted as the relatively small length of D N A required to be synthesised could be completed by the Tat/polymerase as the 46 ° C to 9 4 ° C cycle transiently passes the optimum temperature for polymerisation. The 10 rain boiling of the samples released enough bacterial D N A for the PCR to yield positive results. This rapid and convenient method requires no D N A extraction procedure in contrast to that reported for the PCR on mycobacteria [5]. The PCR-based method described in this study, which does not require hybridisation technology o2 the use of radioactivity, should be useful for the rapid identification of toxigenic C. difficile strains from pure and mixed culture and, with suitable adaptations, could be applied directly to clinical specimens.
ACKNOWLEDGEMENTS We thank John Keyte for the synthesis of oligonucleotides. This work was supported by the Wellcome Trust, U.K.
REFERENCES [1] Bartlett,J.G., Onderdonk, A.B., Cisners, R.L. and Kasper, D.L. (1977) J. Infect. Dis. 136, 701-705. [2] Taylor, N.S., Thome, G.N. and Bartlett, J.G. (1981) Infect. lmmun. 34,1036-1043. [3] Wren, B.W., Clayton. C.L., Mullany, P. and Tabaqchali, S. (1987) FEBS Lett. 225, 82-86. [4] Hance,AJ., Grandchamp, B. and Levy-Frebault,V. (1989) Molec. MicrobioL3, 843-849. [5] Brisson-Noel, A., Gicquei, B., Lecossier, D., LevyFrebault, V., Nassif, X. and Hance, AJ. (1989) Lancet it, 1069-1071. [6] Clayton, C.L., Kleanthous, H., Wren, B.W. and Tabaqchali, S. (1990) J. Med. Microbiol.(in press). [7] Maniatis, T., Fritsch, E. and Sambrook. J. (1982) Molecular cloning: a laboratory approach. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. [8] Bimboim,H.C. and Doly, J. (1979) Nucleic Acids Res. 7, 1514-1523. [9] Sanger, F., Nicklen, S. and Coulson, A.R. (1977) P~oc. Natl. Acad. Sci. U.S.A.74, 5463-5467.
[10] Tabaqchali, S., O'FarreU, S., Holland, D. and Silman, R. (1984) Lancet i, 935-938. [11] Wren, B.W. and Tabaqchali, S. (1987) J. Clin. Microbiol. 25, 2402-2404. [12] Wren, B.W., Heard, S. and Tabaqchali, S. (1987) 3. Clia. Pathol. 40, 1397-1401. [13] Holderman, L.V., Cato, E.P. and Moore, W.E.C. (1977) Anaerobe laboratory manual, 4th edition. Blacksburg Virginia Polytechnic and State University. [141 Wren, B.W., Mullany, P.M., Clayton, C.L. and Tabaqchali, S. (1988) ~EMS Microbiol. Immunol. 47,163-168. [15] John: ~±n, J.L., Dove, C.H., Price, S.B., Sickles, T.W., Pbelp'i, C.J. and Wilkins, T.D. (1988) In Anaerobes today
(Hardie, J.M. and Borriello, S.P., eds.), pp. 115-123, Wiley, Chichester, U.K. [16] Romeo, J.M., Esmon, B. and Zusman, D.R. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 6332-6336. [17] Karch, H. and Meyer, T. (1989) J. Clin. Microbiol. 27, 2751-2757. [18] Wren, B.W., Castledine, N.B., Clayton, C.L. and Tabaqchali, S. (1990) J. Med. Microbiol. (in press). [19] Martinez, R.D. and Wilkins, T.C. (1988) Infect. Immun 56,1215-1221. [20] Price, S.B., Phelps, C.J., Wilkins, T.D. and Johnson, J.J. (1987) Cuff. Microbiol. 16, 82-86.