Molecular and Cellular Probes (1990) 4, 193-203

A comparison of alkaline phosphatase and radiolabelled gene probes with bioassays for enterotoxigenic Escherichia coil

Cheryl A. Bopp, 1 Velda L . Threatt, 2 Steve L . Moseley, 3 Joy G . Wells' and 1 . Kaye Wachsmuth l 'Enteric Bacteriology Section, Enteric Diseases Branch, Division of Bacterial Diseases, Center for Infectious Diseases, Centers for Disease Control, Atlanta, Georgia 30333, USA, 2School of Veterinary Medicine, Tuskegee University, Tuskegee, Alabama-36088, USA and 3 Department of Microbiology, School of Medicine, University of Washington, Seattle, Washington 98195, USA (Received 19 November 1989, Accepted 4 December 1989)

Alkaline phosphatase-conjugated oligonucleotide probes (APO), 32 P-labelled oligonucleotide (RO) and cloned polynucleotide (RP) probes were evaluated for their ability to detect enterotoxigenic Escherichia colt (ETEC) as defined by bioassay . These three sets of probes were applied to 301 E. coli strains that had previously been defined by the Y1 adrenal cell assay for heat-labile enterotoxin (LT) and the infant mouse assay for heat-stable enterotoxin (ST) . The correlation of the APO probe for LT with the bioassay was 98% with five discrepancies and a positive predictive value (PPV) of 100% . For the APO/ST probe the correlation with the bioassay was 98% with seven discrepancies and a PPV of 96% . The correlation, of the RO probe for LT was 99% with four discrepancies and a PPV of 100%, while the overall correlation for the two RO/ST probes was 97% with eight discrepancies and a PPV of 97% . For the RP probes, the correlation for LT was 99% with four discrepancies and a PPV of 100% and for ST-was 98% with seven discrepancies and a PPV of 98% . These findings suggest that tisé'ÂPO probes were as sensitive as the RO and RP probes in detecting ETEC by colony hybridization and could be a practical alternative to bioassays and radiolabelled probes for ETEC since they do not require expensive equipment or extensive technical training .

KEYWORDS : Escherichia coli, enterotoxigenic, gene probes, oligonucleotide, alkaline phosphatase .

"Author to whom correspondence should be addressed . This work performed at CDC and the University of Washington .

0890-8508/90/030193 + 11 $03 .00/0 193



194

C . A . Bopp et al.

INTRODUCTION Enterotoxigenic Escherichia coli (ETEC) is an important cause of diarrhoea in developing countries, where it is one of the most common enteric pathogens of young children ; it is also an important and frequent cause of traveller's diarrhoea .` Although infrequently isolated from sporadic illnesses in the United States, ETEC has caused several outbreaks ranging from infant diarrhoeas in a hospital nursery to large-scale community outbreaks ." ETEC may produce a heat-labile enterotoxin (LT), a heat-stable enterotoxin (ST) or both LT and ST . Escherichia coli strains of human and porcine origin produce two antigenic variants of LT, LTh and LTp . Recently, another heat-labile enterotoxin has been described in an E. coli strain isolated from an ill water buffalo and is now referred to as LT-II . LT-II, like LT-I, induces rounding of Y1 adrenal cells but is not neutralized by antisera to either LTh-l or LTp-I . 10 Two classes of ST are also known, designated STa (ST-I) and STb (ST-II), and are differentiated by their activity in animal models and solubility in methanol ." Unlike STa, STb is not active in the infant mouse model and is infrequently isolated from human S .12,13 Furthermore, two STa toxins have been identified that are structurally distinct but show a high degree of ," homology ." STa-1 (STp) was originally purified from a porcine strain, while STa-2 (STh) was isolated from a human ETEC s train . S T and LT have been purified, synthesized and sequenced . LT-1 is a high molecular weight toxin of two subunits, while STa-1 and STa-2 are small peptides of 18 or 19 amino acids . Historically the rabbit ilea) loop assay was used to detect intestinal fluid accumulation mediated by both LT and ST activity, but more recently the infant mouse assay for STa and the Y1 adrenal cell assay for LT are considered the standard activity assays for E. coli enterotoxins . 16' 17 These two assays require animal and tissue culture facilities that are not available in most small hospitals or clinical laboratories . Alternatively, several enzyme-linked immunosorbent assays (ELISAs) and other immunological assays have been developed using one or more specific antisera to each of the toxins ; however, these antisera are not generally available commercially . 18 An additional problem encountered in all toxin assays is that precise media and culture conditions are usually required for the production and elaboration of toxin in detectable quantities . Although some of the ELISAs can be read visually, most require quantitative spectrophotometry. It is possible to identify ETEC by direct detection of plasmid-encoded LT and ST genes using DNA hybridization . DNA probes have been constructed from both cloned genes and synthesized oligonucleotides . 19-21 The preparation of those cloned polynucleotide probes required purification of the LT and ST genes from vector plasmids carried in bacterial cells, digestion of the purified plasmid DNA with restriction endonucleases and separation of the resulting fragments by preparative gel electrophoresis . The appropriate DNA fragments were then extracted from the agarose or acrylamide gel, purified and labelled with 32 P . Subsequently, chemically synthesized pieces of DNA (oligonucleotides) that correspond to specific sequences within the LT and ST genes were produced, labelled with 32 P and used as probes . 21 More recently, oligonucleotides have been conjugated directly to alkaline phosphatase, avoiding the problems associated with radioisotopes and auto radiography . 22--26 In this study, alkaline phosphatase-conjugated oligonucleotide probes (Molecular



Gene probes for enterotoxigenic E . coli

195

Biosystems, Inc ., San Diego, CA), 32 P-labelled oligonucleotide and cloned polynucleotide probes were evaluated for their relative ability to detect by colony hybridization strains of ETEC isolated in the United States and other countries . This is the first study to evaluate all three sets of probes on a large, diverse group of ETEC strains .

MATERIALS AND METHODS Bacterial strains The 301 E . coli strains included in this study were part of the Centers for Disease Control (CDC) collection received from 1960 to 1987 . Thirty strains were isolated from asymptomatic persons, eight strains were animal isolates, and the remainder were isolated from patients with diarrhoea . When originally isolated, 271 strains were determined to be ETEC (i .e . positive in either the Y1 adrenal cell assay or the infant mouse assay or both) and 30 were non-toxigenic . Of those for which serotyping data was available, 27 strains were serogroup 027, 20 strains were serogroup 06, 18 strains were 0128, 17 strains were serogroup 015, 17 strains were group 078, and 11 strains were serogroup 0148 . All other serogroups were represented by 10 or fewer strains . The United States (including Puerto Rico) was the source for 146 strains, while 132 were received from seven foreign countries (Central African Republic, El Salvador, Truk, Burma, Bangladesh, Pakistan and Mexico), 15 from cruise ship outbreaks, and eight were animal isolates . Strains were stored lyophilized or on stock slants . After rehydration of lyophilized strains, they were stocked in paraffin-corked blood agar base slants at room temperature . Strains were streaked to fresh heart infusion agar slants before being assayed for enterotoxin by gene probes or bioassays .

Bioassays Strains that had divergent results for the bioassays and gene probes were repeated simultaneously in both assays . Strains to be assayed for LT and STa production were inoculated into Evans' CAYE (2% casamino acids, 0 . 15% yeast extract, 0 . 25% NaCl, 0-871% K 2 HPO4 , 0 . 25% glucose, 0 . 1 % trace salts solution (5% MgSO 4, 0E5% MnC1 2 , 0 . 5% FeCl 3 . 6H 2O1, pH 8. 5) . The broth cultures were incubated for 24 h at 37°C on a roller drum rotating at 20 rpm . Broth cultures were centrifuged at 3500 x g for 20 min . The supernatants were removed and assayed in Y1 adrenal cells for LT and 16,17 infant mice for ST as previously described . Toxin neutralization was determined with polyclonal and monoclonal antibodies to LT and CT . Supernatants of the assayed strains were mixed with an equal volume of serially diluted antiserum and incubated for 1 h at 37°C, and 0 . 05 ml samples were applied to Y1 adrenal cell monolayers in microtitre plates . After overnight incubation, the monolayers were examined microscopically for the cytotonic (cell rounding) effect typical of LT .



196

C . A. Bopp et al.

Preparation of filters for hybridization with radiolabelled probes To prepare colony blots for hybridization, strains were inoculated onto MacConkey agar plates and incubated at 37°C overnight . Colonies were blotted onto Whatman 541* cellulose filters (Whatman Ltd ., Maidstone, England), and the DNA was extracted and denatured as previously described . 27

Preparation of 32 P-labelled cloned polynucleotide probes (RP) and hybridization conditions Restriction fragments to be used as hybridization probes for LT, STa-1 and STa-2 were isolated from plasmids pWD299, pDAS101 and pDAS100, respectively, as previously described . 28 Restriction fragments were labelled with a- 32 P nucleoside triphosphates (New England Nuclear Corp ., Boston, MA) by the random primer method . 29 Preparation of colonies and conditions of hybridization have been previously described . 3o

Preparation of 32 P-labelled oligonucleotide probes (RO) and hybridization conditions The oligonucleotide probes for LT, STa-1 and STa-2 were provided by Walter Hill (eltAll, 5' GCG AGA GGA ACA CAA ACC GG 3' ; STa-1, 5' GCT GTG AAC TTT GTT GTA ATC C 3' ; STa-2, 5' GCT GTG AAT TGT GTT GTA ATC C 3') and were 5'-end labelled with 32 P with T4 polynucleotide kinase as described by Hill et al. 2 ' Hybridization of filters was performed at 40°C overnight in a solution containing 6 X SSC (0 .9 M NaCl plus 0 .09 M sodium citrate), 2% salmon sperm DNA, and 5 X Denhardt's solution (0 . 1 % Ficoll, 0 . 1 % polyvinyl pyrrolidone and 0. 1 % BSA) . The filters were rinsed in a 6 X SSC solution for 10 s, followed by two washes at 50°C in 6 X SSC for 1 h . The filters were air dried, exposed to X-ray film at -70°C for 18 h and developed according to manufacturer's instructions .

DNA hybridization with the alkaline phosphatase-labelled probes Two test kits were provided by Molecular Biosystems, Inc. : one for detecting LT genes, and another for detecting STa genes containing a mixture of STa-1 and STa-2 oligonucleotides . These 26-base non-radioactive oligonucleotide probes were constructed by linking alkaline phosphatase to the C-5 position of a thymidine base through a 12-atom spacer arm . 22 Probes were supplied in buffer containing 30 mm Tris, 3 M NaCl, 1 MM MgCl2 , 0 . 1 mM ZnCl2 , pH 7 . 6 with 0 . 05% sodium azide . All reagents, nylon membrane test strips and controls were provided with the kits . Test colonies, the positive control (nucleic acids containing partial LT or ST sequences, minimum concentration 400 lag ml - ' in Tris buffer with 0 . 02% sodium azide), and the negative control (E . coli DNA, minimum concentration 400 Ag ml - ' in Tris buffer * Use of trade names is for identification only and does not imply endorsement by the Public Health Service or by the US Department of Health and Human Services .



Gene probes for enterotoxigenic E . coli

197

with 0 .02% sodium azide) were spotted onto the membrane test strips provided (the manufacturer recommends up to six *ples per strip) . The DNA was extracted by placing the membranes colony side up onto filter paper pads soaked with 2 ml of extraction reagent (1% sodium lauroyl sarcosine, 0-5m sodium hydroxide, and antimicrobial agents 5-chloro-2-methyl .-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one) . Membranes were incubated for 15 min in a 50°C water bath followed by a 30 s wash in distilled water . Membranes were neutralized for 5 min at room temperature by placing them on filter paper pads soaked with 2 ml of 0 . 2 N HCI . After a 30 s wash in distilled water, the membranes were prehybridized in 2 ml hybridization buffer (0 .75 M sodium citrate, 1 .0% N-lauroyl sarcosine, 05% bovine serum albumin, and antimicrobial agents 5-chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one) for 15 min at 50 ° C . Membranes) were then placed in fresh hybridization buffer containing 20 µl of LT or ST probe solution and incubated at 50 ° C for 15 min . Membranes were placed in membrane wash solution #1 (5 X concentrate : 0 . 75 M sodium chloride, 0 . 075 M sodium citrate, 1 . 25% sodium lauroyl sulphate, 1 . 25% N-lauroyl sarcosine, pH 7 . 0, and antimicrobial agents 5chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one) for 20 min at 50°C for the ST probe or 37°C for the LT probe . A second wash 'followed in membrane wash solution #2 (5 X concentrate : 0 . 75 M sodium chloride, 0075 M sodium citrate, 2 . 5% Triton X-100, and antimicrobial agents 5-chloro-2-methyl-4isothiazolin-3-one and 2-methyl-4-isothiazolin-3-one) for 10 min at room temperature with agitation . Membranes were developed in substrate buffer (01 M Tris buffer, 0. 1 M NaCl, 1 . 0% MgCl 2, 0. 136% zinc chloride, pH 8 . 5 and 002% sodium azide) with nitro blue tetrazolium (7 . 50/o in 70% dimethylformamide) and 5-bromo-4chloro-3-indolyl phosphate (5 . 0% in dimethylformamide) and incubated for 30 min for the ST test and 1 h for the LT test . A positive was indicated by a dark purple spot similar in intensity to the positive control .

RESULTS Identification of ETEC phenotype by gene probes Of the 62 strains that produced both LT and STa (LT+ST+) in the bioassays, all were correctly identified as that phenotype by the APO probe, and 61 were correctly identified by the RP and RO probes (Table 1) . Of 52 LT + ST- strains, 47 were correctly identified by the RP probes and 46 by the RO and 43 by the APO probes . Of 70 LT - ST + strains, 68 were correctly identified by the APO probes and 66 by the RP and RO probes . The strains that gave discrepant results in the bioassays and the gene probes are listed in Table 2 . For seven strains the probe results were in agreement for all three probes but disagreed with the bioassay : One strain (EDL 1264) was negative in the bioassays but was identified as LT-ST+ by all three sets of gene probes . Two LT+ST- strains (138 and T923) were identified by all three sets of probes as LT+ST+ . Strain EDL 1273 was positive in the infant mouse assay but negative in the three sets of STa gene probes . Three strains (SSU 6491, SSU 6492 and SSU 6494) were positive in the Y1 adrenal cell assay but negative by all the LT probes . The toxin produced by these three strains was neutralized by both anti-LT and anti-CT in Y1



C. A. Bopp et al.

198

Table 1 .

Accuracy of identification of ETEC by colony hybridization with APO, RO and RP probes Number of ETEC correctly identified

ETEC Phenotype LT/STa LT STa Total Overall sensitivity

Bioassays'

APO'

RO 3

RP'

62 52 70 184

62 43 68 173

61 46 66 173

61 47 66 174

96%

96%

96%

1 Bioassays : Y1 adrenal cell assay for LT, infant mouse assay for ST .' APO : alkaline phosphatase-labelled probes .' RO : radiolabelled oligonucleotide probes . "RP : radiolabelled cloned polynucleotide probes .

adrenal cells . For the remaining 11 strains, the APO, RO and RP probes differed in their identification of LT and ST genes .

Comparison of Y1 adrenal cell assay and colony hybridization with LT probes Of 301 strains included in the study, 114 were positive for LT in the Y1 adrenal cell assay, while 110 were positive by the RP probe, 109 by the APO probe, and 110 by the RO probe . Four false-negatives were detected by the RP probe, five by the APO probe, and four by the RO probe. No false-positives were detected by any of the LT probes . The correlation between the Y1 assay and the RP, RO, and APO probes for LT was 99%, 99% and 98%, respectively . Table 3 lists the sensitivity, specificity and positive and negative predictive values for each probe .

Comparison of infant mouse assay and colony hybridization with the STa probes STa was detected by the infant mouse assay in 132 of 301 strains tested . The APO probe for STa identified 130 strains as STa-positive ; the RP and RO probes for STa-1 and STa-2 each identified 128 STa-positives . Two false-negatives were detected by the APO probe, four false-negatives by the RP probes, and four-false negatives by the RO probes . Three false STa-positives were detected by the RP probes, four by the RO probes, and five by the APO probes . The correlation between the infant mouse assay and the RP, APO and RO probes for STa were 98%, 98% and 97%, respectively (Table 3) .

Identification of STa-1 and STa-2 producing strains A total of 132 strains were positive in the infant mouse assay, of which 72 were probe-positive for the STa-2 gene only, 50 probe-positive for the STa-1 gene only, and three positive for both STa-1 and STa-2 . As described above, three strains were positive in the bioassay but did not react with the RP and RO probes for STa-1 and



199

Gene probes for enterotoxigenic E . col!

N td F N

N N C O

C C

I

1

1

1

1

1

1

1

1

1

I+

1

I+

I

I+

r F NF

•

I I I+ 1 1+ 1 1 1+ 1 1 1 1 I

F

I

I

I

N N

I

I

I

ro

•

I I I I+ I++ I

I I+ I I I I

I

I

I I I++++++

1 1 1 1 1 1 1

I

• F N

•

I I I I++++ I

++ I I+++

+

F"

I

I

I

+

•

1++++++

1

1

1

1

1

1

1

I+ I

I

I

I

I+

I

N q) C

â O C Q)

o

oa)

Na) a 0 C 0

N N

I

I

I

I

I

+

O

â. 0 a Q

•

•0

v o C 1

N

I

I

I++++ I

I

I

I

I

o

• C

N d

O N z NN 0

1 I I I r r r ~t I I CD Ln O z=z 1 xz = =z _=== °E . rnr rM rM NL'a 1Co Con CCN 1n r Nvl ~ N N - IC n 000000000 0000000 0

I

L

v 0 C C C C C •E • •CC CC: 00 - •X d • v CA C 05 Q) > 2à, z-uçi ) >âô=m0m N trn ~-0 ro,at

•

I

I I I I I I I I I I + +++ + + + + F FFF-FF-FF V) FFFFFF--FFF tn N V1 N N N N V1 N Vl V) Vl V) V1 Ui V) F I + + + + + + + + + I I I I I I I + F- J FFFFFF--FFFJ J J J J J J J J FFF-FFFF JJJJ JJJ

.C

•

I

Û

rO Q J `p W Ol e r-4 V) 'A Lri 'n

O J) V) U) r- r r N F--- W

Co rNN a, -1 J J -1 N N r, 0 0 0 Co U •

Q U

n v

N F

Q ro C Û -c N V CT o av c Û0 av m C 1Y aa) Ca) Uo C av a) Ô b av o Co CC4 av a).0 N d ro ro to-flC Co_ vo vC ..o ro 4H O

200

Table 3 .

C . A . Bopp et al.

Comparison of results obtained with the APO, RO and RP probes Predictive value (%)

Probes

Sensitivity (%)

Specificity (%)

Positive

Negative

LT : APO'

96

ROz RP,

96 96

100 100

100 100

97 98

100

100

98

STa : APO RO RP

98 97 97

97 98 98

96 97 98

99 98

'Alkaline phosphatase-labelled oligonucleotide. 'Radiolabelled oligonucleotide. 3 Radiolabelled cloned polynucleotide .

STa-2 . The RO and RP probes differed in their identification of STa-1 and STa-2 genes for four mouse-positive strains (Table 2) . The APO probe for STa hybridized to all 72 bioassay-positive strains carrying the STa-2 gene (as differentiated by the RP and RO probes), to 49 of 50 strains that possessed the STa-1 gene only, and to all three strains that were STa-1 and STa-2 .

DISCUSSION Others have evaluated gene probes for use in detecting ETEC, but this is the first study to examine a large group of isolates from several geographical areas . This study included 301 isolates from the United States (15 states, the District of Columbia, and Puerto Rico) and seven foreign countries in Asia, Africa and Latin America, as well as isolates from cruise ship outbreaks and several animal isolates . Of the 271 strains identified as ETEC originally, only 184 were positive in one or both bioassays at the time of this study, indicating that many had lost toxin genes during storage . The overall correlation of the gene probes with the bioassays was 96% for all the probes . These results indicate that the three different probe formats used here are comparable in sensitivity and specificity . No difference was found between polynucleotide and oligonucleotide probes or between radiolabelled and alkaline phosphatase probes . Some investigators have demonstrated that when the colony hybridization technique is used, radiolabelled polynucleotide and oligonucleotide probes are similarly sensitive and specific, while others show polynucleotides to be more sensitive than oligonucleotides . 28,31 The relative advantages and disadvantages of polynucleotides and oligonucleotides as probes are outlined in Table 4. 32 No false-positives were detected by any of the LT probes . Culture supernatants of all six false-negative strains (four by RP and RO and five by APO) were neutralized by polyclonal anti-LTh-l and anti-cholera toxin but not by a monoclonal anti-LTh-1, indicating that the toxin represents a variant of LT-1 rather than LT-11 . 10 The strains include three serotypes isolated in geographically diverse areas and climates



Gene probes for enterotoxigenic E . coli

201

indicating that each variant evolved independently, however, strains SSU 6491, 6492 and 6494 may represent a cluster of illness caused by a single epidemic strain . Five false-positives were detected by one or more of the ST probes (five APO, four RO and three RP) . Of these five strains, two (EDL 1264 and T923) were infant mouse positive when originally isolated . Perhaps remnants or rearrangements of the toxin genes were detected by the probes or possibly altered regulatory genes prevented toxin expression . Seven false-negatives were detected by the ST probes, of which four were negative by the RO probes and RP probes, and two by the APO probes . In comparison to the bioassay, the APO probes were slightly more sensitive than the other two probes, indicating that the oligonucleotide mixture may detect a more conserved region of the STa gene than either of the separate ST probes . This increased sensitivity due to mixing of oligonucleotides was also noted in a previously published study ." Another possible explanation for false-negative reactions with any DNA probe could be the failure of a strain to lyse and release colony DNA onto the filters . 27 The APO probe lysing procedure and filter composition differed from those of the RO and RP probes . It has also been suggested that the failure of APO probes to detect some ETEC strains is due to difficulty in reaching target DNA because of the alkaline phosphatase linkage . 26 In that study the APO probes identified all ETEC strains when bacterial lysates were examined, but only 78% were identified by colony hybridization .

Table 4.

Features of cloned polynucleotides and synthetic oligonucleotides as gene probes 32

Type of probe

Advantages

Disadvantages

Polynucleotides (cloned probes)

Preparation does not require a DNA synthesizer

Difficult to obtain sufficient quantities of purified cloned gene probes

Not susceptible to a loss of sensitivity resulting from minor nucleotide mismatches

Vector (especially ColEl derivatives) contamination of the probe fragment may generate false-positive reactions

Can be prepared in large quantities overnight

Preparation requires a DNA synthesizer

Highly specific, can discriminate between sequences which differ by only 1 base and between variable and conservative regions

So specific that the sensitivity may be affected by minor nucleotide changes in the target sequences

More uniform than cloned polynucleotides

Hybridization and washing conditions critical

Oligonucleotide (synthetic probes)

Hybridization reaction faster Synthesized as a single strand and will not reanneal to itself No potential for vector contamination



2 02

C . A . Bopp et al.

The APO probes used in this study were oligonucleotides conjugated directly to alkaline phosphatase, and were as specific and sensitive as the radiolabelled probes, particularly for the LT genes . The STa-APO probe was a mixture of STa-1 and STa-2 oligonucleotides . Having one probe instead of two for STa simplifies the testing procedure for ETEC, but does not allow STa-1 and STa-2 producing strains to The distinguished . Other studies using alkaline phosphatase-labelled probes found them to be acceptably sensitive and specific for colony hybridization, although one study using a commercially available STa-2 probe found that STa-1 strains did not hybridize . 24--2' This strain collection had a larger percentage of STa-1 strains (40% of those identified by the radiolabelled STa-1 probes) than has generally been reported, and the STa-APO probe mixture identified all but one . 23-26 In this study, the RO, RP and APO probes were approximately equal in their correlation to bioassays, and any one of these probes would be suitable for use by clinical laboratories . By definition, the RP and RO probes suffer the disadvantages associated with radioisotopes . The short half-life of radioisotopes requires that probes be enzymatically labelled just before use . Isotopes also require special care and precautions during transport and handling in the laboratory . The APO probes in this study were synthesized and labelled at the point of manufacture, shipped at room temperature, could be used for 12 months and required no special precautions for handling . Although one of the recognized advantages of gene probes is their application in processing large numbers of specimens, the use of prelabelled, stable, non-isotopic probes means that small numbers of isolates can be assayed more rapidly and conveniently in a conventional laboratory setting .

REFERENCES 1 . Aguero, M . E., Reyes, L ., Prado, V . et al. (1985) . Enterotoxigenic Escherichia coli in a population of infants with diarrhea in Chile . Journal of Clinical Microbiology 22, 576-81 . 2 . Echeverria, P ., Seriwatana, J ., Leksomboon, U . et al. (1984) . Identification by DNA hybridization of enterotoxigenic Escherichia coli in homes of children with diarrhea . Lancet i, 63-6 . 3 . Echeverria, P., Seriwatana, J ., Taylor, D . N . et al. (1985) . A comparative study of enterotoxigenic Escherichia coli, Shigella, Aeromonas, and Vibrio aetiologies of diarrhea in northeastern Thailand .

American Journal of Tropical Medicine and Hygiene 34, 547-54 . 4 . Shukry, S ., Zaki, A . M., DuPont, H . L . et al. (1986) . Detection of enteropathogens in fatal and potentially fatal diarrhea in Cairo, Egypt. Journal of Clinical Microbiology 24, 959-62 . 5 . Merson, M. H ., Morris, G . K ., Sack, D . A. et al. (1976) . Travelers' diarrhea in Mexico : A prospective study of physicians and family members attending a congress . New England Journal of Medicine 294,1299-305. 6 . Ryder, R. W ., Wachsmuth, I . K., Buxton, A . E . Wat: (1976) . Infantile diarrhea produced by heatstable enterotoxigenic Escherichia coli . New England Journal of Medicine 295, 849-53 . 7 . Rosenberg, M . L., Koplan, J . P., Wachsmuth, I . K . et al . (1977) . Epidemic diarrhea at Crater Lake from enterotoxigenic Escherichia coli . Annals of Internal Medicine 86, 714-18 . 8 . Taylor, W. R ., Schell, W . L., Wells, J . G . et al . (1982). A foodborne outbreak of enterotoxigenic Escherichia coli diarrhea . New England Journal of Medicine 306, 1093-5 . 9 . Wood, L . V., Wolfe, W . H ., Ruiz-Palacios, G . et al. (1983). An outbreak of gastroenteritis due to a heat-labile enterotoxin-producing strain of Escherichia coli. Infection and Immunity 41, 931-4 . 10 . Guth, B . E . C., Pickett, C . L ., Twiddy, E . M . et al. (1986). Production of type II heat-labile enterotoxin by Escherichia coli isolated from food and human feces . Infection and Immunity 59, 587-9 . 11 . Burgess, M. N ., Bywater, R . J ., Cowley, C . M . et al. (1978) . Biological evaluation of a methanolsoluble, heat-stable Escherichia coli enterotoxin in infant mice, pigs, rabbits, and calves . Infection and Immunity 21, 526-31 .



Gene probes for enterotoxigenic E . coli

203

12 . Echeverria, P ., Seriwatana, J ., Patamaroj, U . et al. (1984) . Prevalence of heat-stable II enterotoxigenic Escherichia coli in pigs, water, and people at farms in Thailand as determined by DNA hybridization. Journal of Clinical Microbiology 19, 489-91 . 13 . Weikel, C. S ., Tiemens, K. M ., Moseley, S . L . et al. (1986) . Species specificity and lack of production of STb enterotoxin by Escherichia coli strains isolated from humans with diarrheal illness . Infection and Immunity 52, 323-5 . 14 . Moseley, S . L ., Hardy, J . W ., Huq, M. I . et al . (1983). Isolation and nucleotide sequence determination of a gene encoding a heat-stable enterotoxin of Escherichia coli . Infection and Immunity 39, 1167-74 . 15 . So, M. & McCarthy, B . J . (1980) . Nucleotide sequence of bacterial transposon Tn1681 encoding a heat-stable (ST) toxin and its identification in enterotoxigenic Escherichia coli strains . Proceedings of the National Academy of Science, USA 77, 4011-15 . 16 . Dean, A . G ., Ching, Y .-C ., Williams, R. G . et al. (1972) . Test for Escherichia coli enterotoxin using infant mice: application in a study of diarrhea in children in Honolulu . Journal of Infectious Diseases 125, 407-11 . 17 . Sack, D . A . & Sack, R . B . (1975) . Test for enterotoxigenic Escherichia coli using Y1 adrenal cells in miniculture . Infection and Immunity 11, 334-6 . 18 . Germani, Y. (1986) . Identification and assay methods for Escherichia coli enterotoxins . Bulletin Institut Pasteur 84, 365-87 . 19 . Viscidi, R . P . & Yolken, R . G . (1987) . Molecular diagnosis of infectious diseases by nucleic acid hybridization . ,Molecular and Cellular Probes 1, 3-14 . 20 . Moseley, S . L ., Echeverria, P ., Seriwatana, J . et al. (1982) . Identification of enterotoxigenic Escherichia coli by colony hybridization using three enterotoxin gene probes . Journal of Infectious Diseases 145, 863-9 . 21 . Hill, W . E ., Wentz, B. A ., Payne, W . L. et al. (1986) . DNA colony hybridization method using synthetic oligonucleotides to detect enterotoxigenic Escherichia coli : collaborative study . Journal of the Association of Official Analytical Chemists 69, 531-6 . 22 . Jablonski, E., Moomaw, E . W ., Tullis, R . H. et al. (1986) . Preparation of oligodeoxynucleotide-alkaline phosphatase conjugates and their use as hybridization probes . Nucleic Acids Research 14, 611528 . 23 . Medon, P . P., Lanser, J . A ., Monckton, P . R. et al. (1988) . Identification of enterotoxigenic Escherichia coli isolates with enzyme-labeled synthetic oligonucleotide probes. Journal of Clinical Microbiology 26, 2173-6. 24. Nishibuchi, M ., Arita, M ., Honda, T . et al . (1988) . Evaluation of a nonisotopically labeled oligonucleotide probe to detect the heat-stable enterotoxin gene of Escherichia coli by the DNA colony hybridization test. Journal of Clinical Microbiology 26, 784-6 . 25 . Oprandy, J . J ., Thornton, S . A ., Gardiner, C . H . et al. (1988) . Alkaline phosphatase-conjugated oligonucleotide probes for enterotoxigenic Escherichia coli in travelers to South America and West Africa . Journal of Clinical Microbiology 26, 92-5 . 26 . Seriwatana, J ., Echeverria, P., Taylor, D . N . et al . (1987) . Identification of enterotoxigenic Escherichia coli with synthetic alkaline phosphatase-conjugated oligonucleotide DNA probes . Journal of Clinical Microbiology 25, 1438-41 . 27 . Maas, R . (1983) . An improved colony hybridization method with significantly increased sensitivity for detection of single genes . Plasmid 10, 296-8 . 28 . Sommerfelt, H ., Kalland, K . H ., Raj, P . et al . (1988). Cloned polynucleotide and synthetic oligonucleotide probes used in colony hybridization are equally efficient in the identification of enterotoxigenic Escherichia coli. Journal of Clinical Microbiology 26, 2275-8 . 29 . Feinberg, A. P. & Vogelstien, B . (1983) . A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity . Analytical Biochemistry 132, 6-13 . 30 . Moseley, S . L ., Dougan, G ., Schneider, R . A . et al . (1986) . Cloning of chromosomal DNA encoding the F41 adhesin of enterotoxigenic Escherichia coli and genetic homology between adhesins F41 and K88. Journal of Bacteriology 167, 799-804 . 31 . Echeverria, P ., Taylor, D . N., Seriwatana, J . et al . (1987) . Comparative study of synthetic oligonucleotide and cloned polynucleotide enterotoxin gene probes to identify enterotoxigenic Escherichia coli . journal of Clinical Microbiology 25, 106-9 . 32 . Wachsmuth, K., Strockbine, N . A ., Aye, T . et al . (1989) . DNA probes for enteric bacterial pathogens . In Rapid Methods and Automation in Microbiology and Immunology (Balows, A ., Tilton, R . C. & Turano, A ., eds), pp. 408-19. Brescia, Italy : Brixia Academic Press .