FEMS Microbiology Letters 78 (1991) 75-80 ~ 1991 Federation of European Microbiological Societies 0378-1097/91/$03.50

75

A D O N I S 037810979100105S FEMSLE 04302

Molecular screening of Clostridium difficile toxins A and B genetic determinants and identification of mutant strains D a v i d E. McMillin, L y c u r g u s L. M u l d r o w , S h w a n d a J. Leggette, Y a h y a A b d u l a h i a n d U d o u d o M. E k a n e m e s a n g D.E.M. Science Research Institute, Atlanta, CA, and Clark Atlanta Unioersity, Atlanta, GA, U.S.A.

Received 3 October 1990 Accepted 12 October 1990

Key words: Clostridium difficile; Polymerase chain reaction; Toxin A; Toxin B; Pseudomembranous colitis

1. S U M M A R Y Three separate sets of polymerase chain reaction primers were designed to specifically detect the presence of a toxin A gene fragment, a toxin B gene fragment, and the entire toxin B gene. In addition toxin gene fragments that were amplified from well characterized toxic strains were tagged fluorescently and used as hybridization probes to screen C. difficile strains. A survey of 37 toxic strains and 10 non-toxic strains demonstrated that toxic strains normally contain the genetic composition for toxin A and toxin B simultaneously; whereas, non-toxic strains typically did not contain detectable toxin determinants. The only exception found was strain 39, which had the genetic composition for toxins A and B, but was not cytotoxic under the conditions tested.

Correspondence to: L.L. Muldrow, D.E.M. Science Research Institute, 440 Westview Drive, S.W., Atlanta, GA 30310, U.S.A.

2. I N T R O D U C T I O N

Clostridium difficile, which is the causative agent of antibiotic associated pseudomembranous colitis (PMC) in humans, produces two toxins, A and B. The production of toxin A and B may vary considerably in different strains ranging from very high levels of the toxins, to the apparent absence of the toxin proteins or activity [1,2]. Whereas non-toxic strains do not have detectable toxins, toxic strains of C. difficile appear to possess both the toxin A protein and toxin B activity simultaneously [1,2]. It also has been suggested that the hemorrhagic enterotoxin A and the potent cytotoxin B act synergistically to induce colitis and PMC [3]. At the genetic level it has been reported that toxic strains of C. difficile contain the toxin A gene, while non-toxic strains do not carry the genetic determinant for this protein. DNA hybridization studies by Price et al. [4] with a 2.1 kb PstI fragment of the toxin A gene demonstrated that six toxigenic strains of C. difficile contained

76 the toxin A gene, while four non-toxic strains did not. In two recent publications by Wren et al. [5,6], a 4.5 kb PstI fragment (which contained the above mentioned 2.1 kb PstI fragment), hybridized to 58 toxic strains, but did not hybridize to 17 non-toxic strains; whereas a 63 bp repeated region within this 4.5 kb fragment polymerase chain reaction (PCR) amplified in 33 toxic strains, but could not be amplified from 12 non-toxic strains. A correlation of toxicity with the presence of toxin A DNA has been established; however, studies on the genetic composition of toxic and non-toxic strains of C. difficile for toxin B have not been reported, nor has documentation been presented to determine if these two toxin genes occur simultaneously in different strains of C. difficile. Therefore, the aim of this study was to evaluate C. difficile strains for the presence of the toxin A and toxin B genes.

3. MATERIALS A N D M E T H O D S 3.1. Bacterial strains All but two strains of C. difficile were obtained from the Center for Disease Control, Atlanta, GA. Strain 9689 was obtained from the American Type Culture Collection (Rockville, MD), and strain 10463 was kindly supplied by Tracy D. Wilkins, Virginia Polytechnic Institute, Blacksburg, VA. Purity of the strains was maintained using the cycloserine, cefoxitin and fructose agar selective medium [7]. 3.2. High molecular mass chromosomal DNA isolation High molecular mass chromosomal DNA was isolated by growing the C. difficile cells for 24 to 48 h in brain heart infusion media supplemented with 1.6% glycine. The cells were spun down at 3000 × g and suspended in T E buffer. A 1 : 1 mixture of phenol and chloroform was added and then the microcentrifuge tubes were heated for 2 min at 6 0 ° C prior to centrifugation. Two chloroform extractions were performed followed by ethanol precipitation. The high molecular mass D N A was suspended in TE buffer.

3.3. Cytotoxic assays The standard cytotoxicity and antibody neutralization assays were performed to test toxin activity using chinese hamster ovary (CHO) cells as described previously [8,9]. Toxic activity was expressed as the greatest dilution that caused 50% cell death. Toxin antiserum was generated from purified toxin as described previously [10]. 3.4. DNA hybridization and electrophoresis DNA dot blot analysis and Southern blot hybridization were performed using chemiluminescence labeled probes. Approximately 600 ng of PCR amplified DNA was labeled using the procedure outlined in the ECLT M chemiluminescence kit (Amersham, Arlington Heights, IL, U.S.A.). The DNA samples were blotted onto a nitrocellulose membrane using a dot blot apparatus or from a 1% agarose gel. Hybridization was conducted under high stringency conditions ( 4 2 ° C and 0.5 M NaC1) overnight as described in the ECL protocol. Non-denaturing and denaturing polyacrylamide gel electrophoresis (PAGE) was performed in 5-15% gradient slab gels at pH 8.3 [11]. Electrophoretic transfer (Western blotting) of proteins from polyacrylamide gels to nitrocellulose paper was performed as described by Towbin et al. [12]. 3.5. Oligonucleotide primers The primers used for PCR were toxA-P1 ( G G A A A T T T A G C T G C A G C A T C T G A C ) , toxAP2 ( T C T A G C A A A T T C G C T T G T G T T G A A ) , toxB-P4 ( A G G T G A A C T A C T G T G C A T T C ) , the d e g e n e r a t i v e p r i m e r toxB-P3 ( G A A A A A ATGGC(T)or(A)AATGT), toxB-P1 ( G G T G A T A T G G A G G C A T C A C C A C T A G ) and toxB-P2 (TCCAGGATAAGTCTCCTCTACGTTG). All oligonucleotides were synthesized on a Milligen model 7500 DNA synthesizer. 3.6. Amplification of the toxin A and toxin B genes Amplification of the C. difficile toxin A gene fragment, the toxin B gene fragment, and the entire toxin B gene was carried out in a 50 /tl reaction volume containing 50 ng of C. difficile DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KC1, 1.5 mM MgC12, 2.5 units of Taq polymerase (Perkin Elmer, Norwalk, CT, USA), 200 ~M d N T P rrfix,

77

and oligonucleotide primers. The primer concentration of toxA-P1 and toxA-P2 used to amplify the toxin A gene fragment was 0.2 uM. A concentration of 0.15 # M was used for the toxB-P1 and toxB-P2 primers, while a primer concentration of 1.84 ttM was used for toxB-P3 and toxB-P4. To ensure that all nuclease activity was destroyed, the reaction tube was heated for 5 min at 9 4 ° C prior to adding the Taq polymerase. Amplification was conducted for 37 cycles for all toxin genes using a Perkin Elmer Cetus thermocyler (Norwalk, CT, U.S.A.). The cycle used to amplify the 1.2-kb toxin A gene fragment consisted of denaturation at 9 4 ° C for 1 min, annealing at 55 ° C for 1 min, and extension at 65 ° C for 1 min. The 1-kb toxin B gene fragment was amplified using a cycle that consisted of a 1-min, denaturation step at 94°C, annealing at 4 1 ° C for 1 min, and extension at 7 2 ° C for 1 min. The entire toxin B gene was amplified using a cycle that consisted of denaturation for 1.25 min at 9 4 ° C , annealing for 2.25 min at 41°C, and extension at 6 2 ° C for 10 rain. At the conclusion of the PCR cycles, all tubes were incubated for 7 min at 7 2 ° C to allow the amplification process to go to completion.

1

2

3

4

1

i

66kb-4.4kb --

9.~b 6.6~ 4,4~

2

3

3. 7. Isolation of toxin A using rabbit erythrocyte

ghosts Rabbit erythrocyte ghosts were prepared using the procedure of Krivan et al. [13]. Toxin A from strains 63 and 10463 was isolated using rabbit erythrocyte membranes as described by Wilkins and Krivan [14], and generally outlined as follows. Erythrocyte ghosts were mixed with concentrated C. difficile supernatant containing toxin A for 3 h at 4 ° C , washed 3 times, and incubated at 3 7 ° C for 1 h to dissociate the bound toxin A from the erythrocyte membranes.

4. R E S U L T S

4.1. Toxins A and B screening Three separate sets of primers were designed to specifically detect the presence of a toxin A gene fragment, and the entire toxin B gene, as well as a toxin B gene fragment. The first set of primers, toxA-P1 and toxA-P2, was derived from the published toxin A gene sequence [15,16]. These primers amplified a 1216 bp fragment from toxigenic strains, but did not amplify D N A from non-toxigenic strains as seen in Fig. 1A. This pattern of

4

1

2

3

4

I

4.~b--

--I

7 lkb--

2

3

4

--71kb

-- 7 1 k b 23kb--

2.0kb --

2.~b 20 ~

- -

-

1.3k,b - I .Ikb --

13kb--

0.9kb --

1.1~--

i

1.1kb--

--

1Dkb

1.2~ 0.9~b

--

0.~b--

A

B

C

D

Fig. 1. PCR amplification of C. difficile toxin genes. Results of PCR amplification of: (A) the 1.2-kb toxin A fragment; (B) the entire toxin B open reading frame; (C) the 1.0-kb toxin B fragment; and (D) the 7.1-kb toxin B open reading frame and Southern blot hybridization of the 7.1-kb toxin B gene with the 1.0-kb toxin B D N A fragment. Lanes 1, 2, 3 and 4 from A, B, and C represent the D N A molecular mass marker, toxic strata 74, toxic strain 9689, and non-toxic strain 85 (A, and C) or strain 84 (B) respectively. Lanes l and 2 in D represent non-toxic strain 85 and toxic strain 9689, respectively; while lanes 3 and 4 represent the results of Southern hybridization of lanes 1 and 2, respectively.

78

Table

1

Characterization Strain

of

Clostridium difficile

Cytotoxic

Toxin

activity

PCR

A

Toxin

28

- 7

n e

+

a

dot

strains

A

Toxin

blot b

B

Toxin

dot blot

n

+

10463

- 6

+

+

+

+

77

- 5

n

+

n

+

47

-5

+

+

+

+

70

- 5

+

+

+

+

74

- 5

+

+

+

+

9689

- 5

+

+

+

+

62

- 4

+

+

n

+

59

- 4

+

+

n

+

57

- 4

+

+

n

+

55

- 4

+

+

n

+

89

-4

+

+

n

+

42

- 4

+

+

+

+

66

- 4

+

+

+

+

81

-4

+

+

+

+

25

- 3

n

+

n

+

27

- 3

n

+

n

+

12

- 3

n

+

n

+

40

- 3

n

+

+

+

32

- 3

+

+

n

+

31

- 3

+

+

n

+

36

- 3

+

+

n

+

53

-3

+

+

n

+

58

-3

+

+

n

+

4848

- 3

+

+

n

+

56

- 3

+

+

+

+

49

-3

+

+

+

+

63

- 2

n

+

n

+

41

- 2

n

+

n

+ +

37

- 2

n

+

n

23

- 2

+

+

n

+

48

- 2

+

+

n

+ +

10

-

+

+

n

35

-2

+

+

+

+

45

- 2

+

n

+

n

22

-2

+

+

+

+

34

- 2

+

+

+

+

0

n

-

n

-

4847

0

n

-

n

n

29

0

n

-

n

-

54

0

n

-

n

n

+

+

-

n

A914

2

39

0

+

09

0

.

30

0

-

71

0

.

85

0

.

.

.

.

84

0

.

.

.

.

Primers b Toxin c Primers

toxA-Pl A 1.2-kb toxB-P3

Toxin

B 1.0-kb

n, not

analyzed.

and

+ .

.

.

.

n .

toxA-P2

fragment and

.

was

toxB-P4

fragment

was

were used were used

used

for amplification.

as a probe. used

B

P'CR c

for amplification.

as a probe.

d

amplification held true for 28 toxigenic strains and five non-toxic strains with only one exception, the non-toxic strain 39 amplified the toxin A gene fragment (Table 1). The PCR data described above correlated with DNA hybridization patterns from 37 toxic and 9 non-toxic strains of C. difficile (Table 1). The fluorescent toxin A probe used for the dot blot hybridizations of HindlII digested chromosomal DNA, was generated by amplifying the 1.2-kb toxin A fragment with toxA-P1 and toxA-P2 from toxigenic strain 9689. All toxin-positive strains tested hybridized with this probe, and the only toxin-negative strain which gave a positive hybridization signal was strain 39 (Table 1). The second set of primers, toxB-P3 and toxBP4, demonstrated that the entire 7.1 kb toxin B open reading frame could be amplified from toxigenic strains of C. difficile (Fig. 1B). The toxB-P4 primer was chosen from a sequence down stream of the toxin B gene [15]; and the degenerate primer toxB-P3 was derived from the N-terminal amino acid sequence [17] of toxin B using the codon usage frequency determined from the published toxin A sequence [15] and CAT-D sequence [18] of C. difficile. ToxB-P3 was synthesized and tested prior to the recently published DNA sequence of toxin B [19] and contained either base A or T in position no. 12. (Position no. 12 was subsequently shown by Barroso et al. to be an A [19]). Despite this degeneracy, the entire toxin B gene amplified from the 15 toxic strains tested, and did not amplify from 5 non-toxic strains(Table 1). The non-toxic strain 39 again was the exception (Table 1). To demonstrate that the proposed 7.1 kb toxin B open reading frame, that was amplified with the primers toxB-P3 and toxB-P4, did n o t represent non, specific amplification, primers toxB-P1 and toxB-P2 were designed f r o m tIie recently published toxin B gene sequence [t,9], These two primers wet~e used to amplify a i ~ 9 bp probe from toxin slrain 10463, which was then Southern blot hybridized to the toxB-P3 : toxBTP4 amplified DNA (Fig. 1D). As illustrated in Fig. 1D, using the toxic strain 9689, the 1.0-kb toxin B probe only bound to the 7.1-kb amplified fragment. The 1.0-kb toxB-P1 :toxB-P2 probe was also used to

79 hybridize to the D N A of 35 cytotoxin-positive strains, and to verify that this D N A fragment was not present in 7 non-toxic strains, except strain 39 (Table 1).

4.2. Mutant strains U p o n screening for toxin A and B genetic determinants and the activity of toxin B, as well as the immunological presence of toxin A, two mutant toxin strains were identified. The first mutant strain 39 appears to contain the entire toxin B gene as demonstrated by PCR amplification of the 7.1-kb D N A fragment, but lacked cytotoxic activity under the conditions tested for toxin B (Table 1). This strain also contained the 1.2-kb toxin A D N A fragment (Table 1), but preliminary Western blot screening did not show evidence of the production of the toxin A protein. Additional characterization of this mutant is a major emphasis of our research laboratory. The second mutant that was identified was strain 63. This cytotoxic strain has the genetic determinants for toxin B and at least part of the toxin A gene (Table 1). However, Western blot data demonstrated that this strain contained a

Fig. 2. Western blot of mutant strain 63. Purified toxin A (lane A) and toxin B (lane B) and the culture supernatants from 5 toxin positive strains (10463, 63, 62, 55, and 46) and 2 non-toxic strains (4848 and 84) were run on a 5-15% non-denaturing gel, Western-blotted and developed with a 1:500 dilution of toxin A antiserum.

toxin A gene product that migrated at a faster rate in non-denaturing gels than the purified toxin A or toxin A from other strains of C. difficile (Fig. 2). Preliminary binding and cytotoxic data were obtained on this mutant toxin A protein by subjecting the supernatant of strain 63 (and a control strain 10463) to the erythrocyte ghost purification procedure. This crude preparation demonstrated cytotoxic activity for tissue culture cells which could be neutralized with toxin A antiserum; thus providing preliminary evidence that the mutant toxin contains carbohydrate binding and cytotoxic activities.

5. D I S C U S S I O N It has been previously established that toxigenic strains of C. difficile contain the toxin A gene, while non-toxigenic strains do not [4-6]. To date, a survey for the toxin B genetic composition in toxic and non-toxic strains has not been reported. Also, the toxin B gene has been shown in strain 10463 to be located 1350 bp from the toxin A gene [15]; therefore, the two genes may be on the same operon. In this study a survey of 37 toxic strains and 10 non-toxic strains demonstrated that in the cytotoxic strains studied the toxin A and B genes were found; whereas non-cytotoxic strains typically did not contain detectable toxin determinants. The only exception found was strain 39, which had the genetic composition for toxin A and B, but did not exhibit cytotoxicity under the conditions tested. The results of this study support the hypothesis that the toxin genes could be part of an operon, since the toxin genes occurred simultaneously in the strains analyzed. For strain 39, it could be suggested that a mutation affecting transcription may exist in the leader sequence of the toxin B gene. Ongoing characterization of mutant strain 39 may provide future information which will aid in the understanding of toxin gene regulation in C. diffi'cile. The identification of additional mutants, such as strain 63, should provide a foundation for studies aimed at the elucidation of the molecular, biochemical and physiological mechanism of action of toxin-induced PMC.

80 ACKNOWLEDGEMENTS W e t h a n k J a y W i l s o n a n d T h o m a s B a r b e r for the s y n t h e s i s o f t h e o l i g o n u c l e o t i d e i n the R C M I , M B R L l a b o r a t o r y a n d B a r r y B a t e s for his t e c h n i cal assistance. T h i s w o r k was s u p p o r t e d b y P u b l i c Health Service Grants R29AI26813, S06RR08241, and 1RR03062.

REFERENCES [1] Lyerly, D.M., Sullivan, N.M. and Wilkins, T.D. (1983) J. Clin. Microbiol. 17, 72-78. [2] Wilkins, T.D., Krivan, H., Stiles, B., Carman, R. and Lyerly, D. (1985) in Microbial Toxins and Diarrhoeal Disease (Evered, D. and Whelan, J., eds.), pp. 230-241, Pitman, London. [3] Lyerly, D.M., Saum, K.E., MacDonald, D.K. and Wilkins, T.D. (1985) Infect. Immun. 47, 349-352. [4] Price, S.P., Phelps, C.J., Wilkins, T.D. and Johnson, J.L. (1987) Curr. Microbiol. 16, 55-60. [51 Wren, B.W., Clayton, C.L., Castledine, N.B. and Tabaqchali, S. (1990) J. Clin. Microbiol. 28, 1808-1812.

[6] Wren, B.W., Clayton, C.L. and Tabaqchali, S. (1990) FEMS Microbiol. Lett. 90, 1-6. 17] George, W.L., Sutter, V.L., Citron, D. and Finegold, S.M. (1979) J. Clin. Microbiol. 9, 214-219. [8] Ehrich, M., van Tassell, R.L., Libby, J.M. and Wilkins, T.D. (1980) Infect. Immun. 28, 1041-1043. [9] Muldrow, L.L., Archibold, E.R., Nunez-Montrel, O.L., and Sheehy, R.J. (1982) J. Clin. Microbiol. 16, 637-640. [10] Muldrow, L.L., Ibeanu, G.C., Lee, N.I., Bose, N.K. and Johnson, J. (1987) FEBS Lett. 213, 249-253. [111 Laemmli, U.K. (1970) Nature (Lond.) 227, 680-685. [12] Towbin, H., Stahelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 4350-4354. [13] Krivan, H.C., Clark, G.F., Smith, D.F. and Wilkins, T.D. (1986) Infect. Immun. 53, 573-581. [141 Krivan, H.C. and Wilkins, T.D. (1987) Infect. Immun. [15] Dove, C.H., Wang, S.Z., Price, S.B., Phelps, C.J., Lyerly, D.M., Wilkins, T.D. and Johnson, J.L. (1990) Infect. Immun. 58, 480-488. [161 Sauerborn, M. and von Eichel-Streiber, C. (1990) Nucleic Acids Res. 18, 1629. [17] Meador, J. and Tweten, R.K. (1988) Infect. lmmun. 56, 1708-1714. [18] Wren, B.W., Mullany, P., Clayton, C.L. and Tabaqchali, S. (1989) Nucleic Acids Res. 17, 4877. [19] Barroso, L.A., Wang, S.Z., Phelps, C.J., Johnson, J.L. and Wilkins, T.D. (1990) "Nucleic Acids Res. 18, 4004.

Molecular screening of Clostridium difficile toxins A and B genetic determinants and identification of mutant strains.

Three separate sets of polymerase chain reaction primers were designed to specifically detect the presence of a toxin A gene fragment, a toxin B gene ...
384KB Sizes 0 Downloads 0 Views