Identification and molecular characterization of the gene encoding coli ...

FEMS Microbiology Letters 239 (2004) 131–138 www.fems-microbiology.org

Identification and molecular characterization of the gene encoding coli surface antigen 20 of enterotoxigenic Escherichia coli Ha˚vard Valvatne a, Hans Steinsland a,b, Harleen M.S. Grewal a, Ka˚re Mølbak c, Jens Vuust d, Halvor Sommerfelt b,* a

The Gade Institute, Section for Microbiology and Immunology, University of Bergen and Haukeland University Hospital, Norway b Centre for International Health, University of Bergen, Norway c Danish Epidemiology Science Center, Statens Serum Institut, Copenhagen, Denmark d Department of Clinical Biochemistry, Statens, Serum Institut, Copenhagen, Denmark Received 21 June 2004; received in revised form 17 August 2004; accepted 20 August 2004 First published online 11 September 2004 Edited by M. Schembri

Abstract Enterotoxigenic Escherichia coli (ETEC) is a major cause of diarrhea among children living in developing countries and of travelersÕ diarrhea. Current ETEC vaccine designs aim to induce an anti-colonizing immunity by including the ETEC surface colonization factor antigens. We isolated and characterized the structural gene of the coli surface antigen 20 (CS20). CS20 has an N-terminal amino acid sequence similar to that of CS18. We therefore used a DNA fragment carrying the CS18 fotA gene as a probe in a hybridization assay to detect the corresponding gene in a CS20-positive strain isolated from an Indian child. Cross hybridizing DNA was isolated and found to contain an open reading frame encoding a polypeptide of 195 amino acids, including a 22 amino acid signal peptide. The gene, which we named csnA, shows a high degree of identity to the major fimbrial subunits of CS12, CS18 and F6 (also referred to as 987P), a CS of porcine ETEC. The coding region of csnA was inserted into an expression system to generate a polypeptide confirmed to be CS20 by Western blot. A CS20 colony hybridization assay using a DNA probe derived from csnA was developed. 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. Keywords: CS20; csnA; Colonization factor; Enterotoxigenic; ETEC

1. Introduction Enterotoxigenic Escherichia coli (ETEC) is a major cause of diarrhea among children living in developing countries and of travelersÕ diarrhea [1]. ETEC produces at least one of the three enterotoxins heat-labile (LT), porcine heat-stable (STp) and the human heat-stable (STh) enterotoxins [1,2]. ETEC adheres to and colonizes the intestinal epithelium using colonization factors *

Corresponding author. Tel.: +47 55 97 4987; fax: +47 55 97 4979. E-mail address: [email protected] (H. Sommerfelt).

(CFs), most of which are fimbrial structures on the bacterial surface. Current vaccine development efforts are focused on inducing an anti-colonizing immunity by including CF antigens (CFAs) [3]. Best characterized are the CFAs I, II and IV. CFA/I is a uniform fimbrial structure, while CFA/II and CFA/IV are composed of distinct coli surface antigens (CSs) in various permutations. Thus, strains bearing CFA/II express CS3 alone or in combination with CS1 or CS2 fimbriae. Likewise, CFA/IV strains express CS6 alone or together with CS4 or CS5 fimbriae. The other factors described are CS7, CS8 (originally CFA/III), CS10 (antigen 2230), CS11

0378-1097/$22.00 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.femsle.2004.08.028

132

H. Valvatne et al. / FEMS Microbiology Letters 239 (2004) 131–138

(PCFO148), CS12 (PCFO159), CS13 (PCFO9), CS14 (PCFO166), CS15 (antigen 8786), CS17, CS18 (PCFO20), CS19, CS20, CS21 (longus) and CS22 [2,4]. Epidemiological studies reveal that many ETEC strains do not possess known CFs. These are referred to as 0-ETEC [5]. Since colonization is probably a prerequisite for virulence, it is assumed that most clinical ETEC isolates produce 1 or more adhesins [2]. The identification of new adhesins is important for extending the knowledge of ETEC and considered to be critical for developing vaccines against ETEC diarrhea [3]. We have previously described the major subunit of CS20 using immunoblot, slide agglutination and immunoelectron microscopy [6]. Purified CS20 are composed of protein subunits with an apparent molecular weight of 21 kDa. The N-terminal amino acid sequence of the major fimbrial subunit of CS20 shows 65% and 60% identity to CS18 and F6 (also referred to as 987P) fimbriae of human and porcine ETEC, respectively [6], and 30% identity to CS12 [7]. Here, we report the molecular cloning and the complete nucleotide sequence of the

open reading frame (ORF) encoding the CS20 major fimbrial subunit. To confirm that the ORF codes for CS20 and to test whether it was possible to produce recombinant CS20, expression studies of the cloned CS20 structural gene were performed. Furthermore, we developed a DNA–DNA colony hybridization assay based on a probe from the CS20 major fimbrial subunit gene, and evaluated its sensitivity and specificity on a battery of ETEC reference strains and ETEC strains isolated from children in North India.

2. Materials and methods 2.1. Bacterial strains, plasmids and primers The characteristics of the reference strains used in the study are presented in Table 1. The plasmids and primers used in the study are shown in Table 2. All ETEC strains were grown overnight on CFA agar plates [6] at 37 C before use.

Table 1 Characteristics of reference strains used in the study No. of strains

1 1 1 1 1 4

5

1 1 2 3

1 1 1 1 1 1 1 1 1 1 1 1 1

Name

H721A H721Ab(pIVB3-100) H10407 247425-1a 248750-1a 309900-2a 259325-1b 278556-1a 253488-1a VX67356a VX68011 VM66257 VM66252 E11881A E17018A VM75688 E3135A E519/66A 31-10A E19475A Z-26-5-6M E29101A 334A 2230 350C1 E7476A 8786 E20738A ARG2 H595C DS168-1 E9034 ARG3 9504225

Enterotoxin gene profile

Colonization factor designation

Ref.

ST

LT

+ + + + + +

+ + + + + +

CS20 CS20 CFA/I CS1CS3 CS2CS3 CS3

[6] [6] [15] [16] [16] [16]

+

+

CS4CS6

[17] [18]

+ + +

+

CS5CS6 CS5CS6 CS6

[18] [17] [18]

+

CS6CS8

+ + + + + + + + + +

+ + + + + + + +

CS7 CS7 CS10 CS12 CS14 CS15 CS17 CS18 CS19 CS19 CS3CS21 CS22 F6 (987P)

[18] [19] [20] [21] [21] [22] [23] [19] [24] [25] [11] [26] [27] [28] [4] [29]


133

Table 2 Characteristics of primers and plasmids used in the study Plasmid pBluescript II pCR2.1 pET101/D-TOPO pBS-CS20-probe-1 pET-CS20 pIVB-CS18

Description Cloning vector; 2.9 kb. Apr Cloning vector; 3.9 kb. Apr. Kanr In vitro expression vector, 5.75 kb, Apr Clone 1, containing the 409 bp at 3 0 end of csnA from PCR with csnA1 probe and csnA2 probe Expression plasmid containing the 588 bp insert of csnA from PCR with csnA1 expr and csnA2 expr Containing the 662 bp at 3 0 end of fotA gene

Primer csnA1 csnA2 csnA1probe csnA2probe csnA1expr csnA2expr M13Forward M13Reverse T7 T3

Description ATG CCC TTT TAA CTA TGG TGT CC GCA TAA CCT CTC CCG GAA TG CTG TGA CAG CCT GAC TGA CG GCA TAA CCT CTC CCG GAA TG CACC ATG AAA AAA AAT GAT TAT GCC TTA CGG AGT TGC ATC CGC C CAC GAC GTT GTA AAA CGA C CAG GAA ACA GCT ATG AC TAA TAC GAC TCA CTA TAG GG ATT AAC CCT CAC TAA AGG GA

Accession numbers AF438155 AF438156 AF438157

CS20 sequence obtained from strain H721A CS20 sequence obtained from strain H683A CS20 sequence obtained from strain H49A

2.2. Isolation of the CS20 major fimbrial gene

2.3. DNA sequencing

Using the QIAprep Spin Miniprep Kit (Qiagen, Germantown, MD), wild type plasmid DNA was extracted from ETEC strains H721A (CS20+), ARG2 (CS18+), 9504225 (F6+) and H10407 (CFA/I+). The plasmids were digested with PstI (New England BioLabs, Beverly, MA) for 4 h at 37 C. The resulting fragments were separated on a 1.5% agarose gel and transferred to nylon membranes (Hybond-N+, Amersham International, UK) by the Southern blot capillary transfer method. The membranes were hybridized with a DIG-labeled CS18 probe (Table 2) overnight at 45 C and washed under low stringency conditions [8]. The fragment corresponding to the DNA to which the CS18 probe hybridized was excised from the gel, purified using the QIAquick Gel Extraction Kit (Qiagen) and ligated into pBluescript (Stratagene, La Jolla, CA). The construct was heatshock transformed into TOP10FÕ cells as described by the manufacturer (Invitrogen, Carlsbad, CA). The CS20 major fimbrial subunit gene was isolated from this construct with PCR using 10 mM Tris–Cl, pH 8.3, 50 mM KCl, 1.5 mM MgCl2, 0.1 mM dNTP, 2.5 U Taq DNA polymerase, 0.5lM each of sense and antisense primers (Table 2), and 0.5 lg template DNA. Amplification reactions consisted of 10 min at 94 C, 35 cycles of 94 C for 30 s, 58 C for 30 s, and 72 C for 1 min, and a final extension of 10 min at 72 C. The resulting PCR fragments were separated with agarose gel electrophoresis, and the bands were purified using QIAquick Gel Extraction Kit and cloned into pCR2.1 and transformed into TOP10 using the manufacturerÕs protocol for TA cloning (Invitrogen).

DNA constructs were sequenced using an ABI Prism 377 DNA sequencer. TA clones were used as template DNA and oligonucleotides corresponding to the M13 forward and reverse promotor regions flanking the cloned DNA were used as primers. Nucleotide and protein database searches were performed using Blast Network Service, and sequence alignments were performed with CLUSTAL X. Putative post-translational modifications were analyzed at the SignalP 3.0 World Wide Web server at Centre for Biological Sequence Analysis, Denmark (CBS) (http://www.cbs. dtu.dk/services/SignalP). 2.4. Expression of recombinant CS20 Amplification of the gene assumed to encode the CS20 major subunit was carried out using a forward primer designed to encode a CACC overhang, the start codon and the subsequent 18 nucleotides of the gene (Table 2). The reverse primer included the 19 last nucleotides of the gene, including a TAA stop codon (Table 2). PCRs were performed as described above but with an annealing temperature of 48 C. The ligation and transformation procedures are described in the protocol for TOPO cloning (Invitrogen). The colonies from which plasmids were purified were grown in 10 ml LB containing 100 lg ampicillin. A construct which we named pET-CS20 with correctly inserted CS20 genes were identified with PCR using the T7 forward and the csnA2expr primers (Table 2). BL21(DE3) One Shot E. coli was heatshock transformed by 10 ng pET-CS20

134


and inoculated into 10 ml LB supplemented with 200 lg/ ml carbencillin and 1% glucose. IPTG was added to a final concentration of 1 mM to induce expression of recombinant CS20. Crude protein extracts were run through a SDS–PAGE gel and stained with Coomassie blue or electroblotted onto nitrocellulose membranes (Bio-Rad Laboratories, Richmond, CA). The CS20 specific bands were detected as described earlier using the / 78 rabbit antiserum raised against CS20 [6]. 2.5. DNA–DNA colony hybridization and CS20 gene probe assay evaluation The 3 0 end of the gene assumed to encode the CS20 major subunit was TA-cloned (Invitrogen) from the wild type plasmid DNA from strain H721A using the primers csnA1probe and csnA2probe (Table 2). The insert was transferred from pCR2.1 to pBluescript using EcoRI (New England BioLabs). The DIG-labeled CS20 gene probe was produced using the PCR DIG probe Synthesis Kit (F. Hoffmann-La Roche Ltd, Basel, Switzerland) with T7 and T3 as primers as described elsewhere [8]. Bacterial isolates were inoculated onto Hybond N+ membranes overlaying CFA agar plates. Hybridization was performed using 1 ll of unpurified PCR-produced DIG probe solution per ml of hybridization solution [8]. To optimize the performance of the CS20 gene probe assay, several parallel Hybond-N+ membranes with colony lysates from six ETEC strains with known CS profiles [H721A (CS20+) and one strain each of types CS18+, F6+, CS12+, CS5+/6+ and CFA/I+] were used in a stringency calibration [8]. After washing at 65 C in 3· SSC, 0.5% SDS, the colony lysates of the CS18+, F6+, CS12+, CS5+/6+ and CFA/I+ strains yielded signals which gradually disappeared when the stringency was increased. These signals disappeared completely after the probe was washed in 0.6· SSC, 0.1% SDS at 65 C, while the H721A lysate maintained its intensity. Thus, in the final assay, after hybridization for 6 h, the stringent washes were performed in 0.6· SSC, 0.1% SDS at 65 C. The assay was then evaluated using a panel of phenotypically well characterized ETEC strains, including 90 pediatric isolates from studies in India with known CFs [5–7] and a panel of 34 reference strains (Table 1). In addition, 21 pediatric isolates from North India with unknown CS profiles were included [5–7].

3. Results 3.1. Southern blot hybridization, cloning and nucleotide sequence of the CS20 fimbrial gene Plasmids from ETEC strains H721A, ARG2, H10407 and 9504255 were digested with PstI, run on an agarose gel and transferred to a nylon membrane. On hybridiza-

tion with the CS18 gene probe at low stringency, a positive fragment was detected with DNA from the CS18+ ARG2 strain. A DNA fragment from the CS20+ H721A strain hybridized, while there was no cross-hybridization with DNA fragments from the CFA/I+ H10407 or the F6+ 9504255 strains. The cross-hybridizing DNA fragment from strain H721A was isolated and cloned into the pBluescript II vector and sequenced. The nucleotide sequence of the complete 588 bp ORF was determined (GenBank Accession No. AF438155). The gene that we named csnA,1 encoded a polypeptide of 195 amino acids with a theoretical molecular mass of 19.8 kDa. After cleavage of the signal peptide, the mature protein is predicted to have a mass of 17.5 kDa. Identical ORFs were also sequenced from the strains H683A and H49A. 3.2. Genetic relationship between CS12, CS18, CS20 and F6 Database searches and alignments showed that the csnA sequence had a high identity score with the human ETEC fimbrial subunit genes encoding CS12 and CS18 and the porcine ETEC fimbrial subunit gene encoding F6 (Fig. 1). Alignment of the nucleotide sequences reveals 376 bases of a 618 nucleotide overlap (61% identity) between csnA and fotA (encoding CS18). Comparing csnA with cswA (encoding CS12) shows an overlap of 317 of the 592 nucleotides (54% identity). The csnA gene was also closely related to the fasA (encoding F6) with 336 of 588 nucleotides that overlap (57% identity). Comparison of the deduced amino acid sequence of CS20 with those of the fimbrial subunit of CS12, CS18 and F6 showed an overall identity of 46%, 58% and 53%, respectively. 3.3. Expression of csnA To confirm that csnA encodes the structural subunit of CS20 and to assess whether CS20 could be produced recombinantly, we amplified the coding sequence and cloned it into the TOPO cloning vector pET. The resulting plasmid, pET-CS20 (Table 2), was transformed into the E. coli strain BL21(DE3) and subsequently, the whole cell lysates were examined by immunoblotting using antiserum /78. As shown in Fig. 2, a CS20-specific band at approximately 24 kDa was observed 1 h after 1

Based on an unpublished agreement (Wim Gaastra, personal communication), it is proposed that genes encoding CFs of ETEC isolated from humans should start with cs (coli surface), followed by the first available letter in the word denoting the CF number. Thus, for CS20, the first available letter is n (t, w and e in ‘‘twenty’’ have already been used to name the genes of other CFs). Furthermore, each gene in the CF operon is given a last letter that denotes the function of the corresponding protein; the genes that encode the major structural protein have the final letter A.


135

Fig. 1. Alignment of the nucleotide sequences for the genes encoding the major subunits of CS12 (GenBank Accession No. AY009096), CS18 (GenBank Accession No. U31413), CS20 (GenBank Accession No. AF438155) and F6 (GenBank Accession No. U50547) Nucleotide sequences identical to those of CS20 are indicated by black boxes. Grey boxes indicate regions where some of the sequences are identical. Dashes indicate gaps introduced to maximize the alignment.

IPTG induction (Fig. 2, lane 3). In addition, a weak band of 21 kDa was also observed in the same extract. Two hours after the IPTG induction, the 24 and the

21 kDa protein bands showed equal strengths, which were maintained during the whole expression experiment (Fig. 2, lanes 4–6). A third, somewhat weaker,

Fig. 2. Immunoblot analysis. Lanes:1, heat extract of BL21(pET-CS20) before IPTG induction; 2, heat extract of BL21(DE) 4 h growth in the presence of IPTG; 3, heat extract of BL21(pET-CS20) 1 h incubation with IPTG; 4, heat extract of BL21(pET-CS20) 2 h incubation with IPTG; 5, heat extract of BL21(pET-CS20) 3 h incubation with IPTG; 6, heat extract of BL21(pET-CS20) 4 h incubation with IPTG; 7, heat extract of the cfaR-transformant H721A(pIVB3-100) grown in CFA agar containing ampicillin. A molecular weight standard is indicated to the right.

136


band at approximately 30 kDa appeared 3 h after induction (Fig. 2, lane 5). The cfaR-transformed ETEC strain H721Ab(pIVB3-100) yielded a strong CS20 specific band at approximately 21 kDa (Fig. 2, lane 7). 3.4. Detection of CS20-positive ETEC by colony hybridization Seven of the Indian ETEC strains tested positive with the csnA probe. One of these (H721A) was the reference CS20+ strain [6] from which the probe was derived. Furthermore, five isolates, H683A, F459A, F67B, 387II and H49A, were positive in the hybridization assay. Two of these strains have previously been shown to produce polypeptides with an N-terminal aminoacid sequence identical to that of CS20 after cfaR-transformation. All five strains were also genetically related to H721A and therefore already assumed to contain the CS20 structural gene [7]. In addition, the STLT+ 0-ETEC strain H596A was positive in the csnA probe assay. The remaining pediatric Indian and reference ETEC strains were all negative in the hybridization assay.

4. Discussion The fimbrial structural gene, csnA, was isolated following cross-hybridization on a Southern blot of plasmid DNA from the CS20+ ETEC strain H721A using a CS18 polynucleotide probe at low stringency. The sequence of the DNA fragment revealed an ORF of 588 bp. CsnA has three potential in-frame ATG initiation codons, but the presence of two lysine residues following the first methionine residue suggests that the first start codon could represent the initiation of translation. The resulting product, a polypeptide of 195 amino acids, would then contain a typical prokaryotic signal peptide [9]. An amino acid sequence identical to the N-terminal amino acid sequence of the CS20 protein [6] was found 23 residues after the start codon. This is in agreement with a putative cleavage site of the pre-fimbrial protein, which is proposed to be between residue 22 and 23 using databases for prediction of post-translational modification [10]. Removal of the first 22 amino acids would yield a mature fimbrial protein subunit of 17.5 kDa. A previous estimate based on SDS-PAGE was considerably higher, at approximately 21 kDa [6]. The basis for this discrepancy in size is unknown, but unexpected migration in SDS-PAGE appears to be a common feature for other fimbrial subunits such as the 17.5 kDa subunit of F6, which migrates at 21 kDa, and the 18.1 kDa subunit of CS18, which migrates at 25 kDa [11]. The predicted amino acid sequence of the fimbrial subunit of strain H721A shows a high degree of identity with the fimbrial subunits CS12 and CS18 (46% and 58%, respectively) of ETEC isolated in humans, and

the F6 fimbrial subunit (53%) of porcine ETEC. Fimbriae of human ETEC are divided into different groups based on fimbrial gene sequence homology [2]. The first group is the CFA/I family where all members have fimbriae that are genetically related to CFA/I. The second group is the CS5 family and comprises the two helically shaped CS5 and CS7 and the more distantly related CS13. The third group contains the CS8 and CS21 that belongs to the Type IV like family, also found in other bacteria like Vibrio cholerae and Neisseria gonorrhoeae [12]. A small family, the CS15 family, comprises the newly identified CS15 and CS22 [4]. The new sequence information for CS12 and CS20 reveals that they are homologous to each other and to the F6 and CS18 major subunits, and not closely related to the other described families. This group, which we name the CS12 family, may also be a subgroup of Class I fimbriae produced by G- bacteria since both F6 and CS18 have been shown to contain structural relatedness to these [13]. The last family of CFs comprising CS3, CS6, CS10 and CS11 have no homology to each other or with any other known fimbria. Protein expression studies using a recombinant plasmid harboring the csnA gene were carried out in the BL21(DE3) strain. The IPTG-induced proteins were identified as CS20 using specific antibodies [6]. A band corresponding to 21 kDa, probably representing the mature CS20 polypeptide, was observed 1 h after IPTG induction. The expression studies also revealed an additional, about 3 kDa larger band, which probably represents the unprocessed pre-fimbrial protein; pre-CS20. When the transformed BL21(DE) strain was incubated for more than 3 h with IPTG, a third polypeptide of approximately 30 kDa appeared. This polypeptide probably represents translational overruns of the CS20 gene stop codon, resulting in a product combining the CS20 protein with the V5 epitope and the His tag in the expression vector. The tendency of ETEC strains to spontaneously lose genes encoding positive regulators of fimbrial expression indicates that DNA based methods may be preferable to phenotypic tests for the identification of ETEC CFs on stored and subcultured ETEC strains [5]. DNA hybridization assays have previously been shown to be suitable to identify unexpressed fimbrial genes [14]. In the present study, a polynucleotide probe specific for the csnA gene was developed and the resulting assay was highly specific and sensitive, with no observable cross-hybridization to the genes encoding the closely related CS12, CS18, and F6 fimbriae. Seven CS20+ strains (6%) were detected in the panel of the 111 Indian ETEC strains. Thus, the CS20 gene was accordingly identified in 7 (26%) of the 27 originally described 0-ETEC Indian strains [5]. As stated, ETEC is an important cause of diarrhea in children of developing countries and in visitors to these


countries. Since immunity follows exposure to ETEC, and CFs are believed to play a role in the development of this immunity, there is an interest in characterizing ETEC with regard to the CFs in order to develop an ETEC vaccine. The proportion of CS20-positive strains among the 0-ETEC in the Indian study [5] suggests that infections with these organisms may be prevalent in developing countries, a finding which is of relevance for developing a vaccine against childhood and travelersÕ diarrhea.

Acknowledgements We thank Nils Axelsen (Department of Clinical Biochemistry, Statens Serum Institute, Copenhagen), Gunnar Kva˚le (Centre for International Health, University of Bergen) as well as Karl Henning Kalland (Section for Microbiology and Immunology, The Gade Institute, University of Bergen), for providing excellent working environments. The constructive advice of Wim Gaastra (University of Utrecht), Michael Theisen, Birthe Storai, Mette Paulli Andersen and Ilona Rosenstadt (Statens Serum Institute) and Thor G. Theander (University of Copenhagen) is gratefully appreciated. We are particularly grateful to Maharaj K. Bhan (All India Institute of Medical Sciences, New Delhi), Ann-Mari Svennerholm and Gloria Viboud (University of Gothenburg), Carol O. Tacket (University of Maryland) and Moyra M McConnell (Central Public Health Laboratory, London) for useful discussions and for providing ETEC strains. The study was supported financially by The University of Bergen and The Graduate School of Research in International Health, Institute for Medical Microbiology and Immunology, University of Copenhagen.

References [1] Levine, M.M. (1987) Escherichia coli that cause diarrhea: enterotoxigenic, enteropathogenic, enteroinvasive, enterohemorrhagic, and enteroadherent. J. Infect Dis. 155, 377–389. [2] Gaastra, W. and Svennerholm, A.M. (1996) Colonization factors of human enterotoxigenic Escherichia coli (ETEC). Trends Microbiol. 4, 444–452. [3] Svennerholm, A.M. and Steele, D. (2004) Microbial-gut interactions in health and disease. Progress in enteric vaccine development. Best Pract. Res. Clin. Gastroenterol. 18, 421–445. [4] Pichel, M., Binsztein, N. and Viboud, G. (2000) CS22, a novel human enterotoxigenic Escherichia coli adhesin, is related to CS15. Infect Immun. 68, 3280–3285. [5] Sommerfelt, H., Steinsland, H., Grewal, H.M., Viboud, G.I., Bhandari, N., Gaastra, W., Svennerholm, A.M. and Bhan, M.K. (1996) Colonization factors of enterotoxigenic Escherichia coli isolated from children in north India. J. Infect. Dis. 174, 768–776. [6] Valvatne, H., Sommerfelt, H., Gaastra, W., Bhan, M.K. and Grewal, H.M. (1996) Identification and characterization of CS20, a new putative colonization factor of enterotoxigenic Escherichia coli. Infect. Immun. 64, 2635–2642.

137

[7] Valvatne, H., Steinsland, H. and Sommerfelt, H. (2002) Clonal clustering and colonization factors among thermolabile and porcine thermostable enterotoxin-producing Escherichia coli. APMIS 110, 665–672. [8] Steinsland, H., Valentiner-Branth, P., Grewal, H.M., Gaastra, W., Molbak, K.K. and Sommerfelt, H. (2003) Development and evaluation of genotypic assays for the detection and characterization of enterotoxigenic Escherichia coli. Diagn. Microbiol. Infect. Dis. 45, 97–105. [9] Watson, M.E. (1984) Compilation of published signal sequences. Nucleic Acids Res. 12, 5145–5164. [10] Nielsen, H., Engelbrecht, J., Brunak, S. and von Heijne, G. (1997) Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Eng. 10, 1–6. [11] Viboud, G.I., Jonson, G., Dean-Nystrom, E. and Svennerholm, A.M. (1996) The structural gene encoding human enterotoxigenic Escherichia coli PCFO20 is homologous to that for porcine 987P. Infect. Immun. 64, 1233–1239. [12] Giroń, J.A., Gomez-Duarte, O.G., Jarvis, K.G. and Kaper, J.B. (1997) Longus pilus of enterotoxigenic Escherichia coli and its relatedness to other type-4 pili-a minireview. Gene 192, 39–43. [13] Girardeau, J.P., Bertin, Y. and Callebaut, I. (2000) Conserved structural features in class I major fimbrial subunits (Pilin). in gram-negative bacteria. Molecular basis of classification in seven subfamilies and identification of intrasubfamily sequence signature motifs which might be implicated in quaternary structure. J. Mol. Evol. 50, 424–442. [14] Grewal, H.M., Sommerfelt, H., Gaastra, W., Svennerholm, A.M., Bhan, M.K., Hamers, A.M., Kumar, R., Wiklund, G. and Bjorvatn, B. (1990) Detection of colonization factor antigen Ipositive enterotoxigenic Escherichia coli with a cloned polynucleotide probe. J. Clin. Microbiol. 28, 2264–2268. [15] Evans, D.G., Silver, R.P., Evans Jr., D.J., Chase, D.G. and Gorbach, S.L. (1975) Plasmid-controlled colonization factor associated with virulence in Esherichia coli enterotoxigenic for humans. Infect. Immun. 12, 656–667. ˚ hreń, C., Stoll, B., Barua, D.K., Ørskov, F., [16] Gothefors, L., A Salek, M.A. and Svennerholm, A.M. (1985) Presence of colonization factor antigens on fresh isolates of fecal Escherichia coli: a prospective study. J. Infect. Dis. 152, 1128–1133. [17] Clemens, J.D., Sack, D.A., Harris, J.R., Chakraborty, J., Neogy, P.K., Stanton, B., Huda, N., Khan, M.U., Kay, B.A. and Khan, M.R., et al. (1988) Cross-protection by B subunit-whole cell cholera vaccine against diarrhea associated with heat-labile toxinproducing enterotoxigenic Escherichia coli: results of a large-scale field trial. J. Infect. Dis. 158, 372–377. [18] McConnell, M.M., Thomas, L.V., Willshaw, G.A., Smith, H.R. and Rowe, B. (1988) Genetic control and properties of coli surface antigens of colonization factor antigen IV (PCF8775) of enterotoxigenic Escherichia coli. Infect. Immun. 56, 1974–1980. [19] McConnell, M.M. and Rowe, B. (1989) Prevalence of the putative colonization factors CFA/III and PCFO159:H4 in enterotoxigenic Escherichia coli. J. Infect. Dis. 159, 582–586. [20] Binsztein, N., Jouve, M.J., Viboud, G.I., Moral, L., Rivas, M., ˚ hreń, C. and Svennerholm, A.M. (1991) Lo´pez Ørskov, I., A Colonization factors of enterotoxigenic Escherichia coli isolated from children with diarrhea in Argentina. J. Clin. Microbiol. 29, 1893–1898. [21] Hibberd, M.L., McConnell, M.M., Field, A.M. and Rowe, B. (1990) The fimbriae of human enterotoxigenic Escherichia coli strain 334 are related to CS5 fimbriae. J. Gen. Microbiol. 136 (Pt 12), 2449–2456. [22] Darfeuille-Michaud, A., Forestier, C., Joly, B. and Cluzel, R. (1986) Identification of a nonfimbrial adhesive factor of an enterotoxigenic Escherichia coli strain. Infect. Immun. 52, 468– 475.

138


[23] Tacket, C.O., Maneval, D.R. and Levine, M.M. (1987) Purification, morphology, and genetics of a new fimbrial putative colonization factor of enterotoxigenic Escherichia coli O159:H4. Infect. Immun. 55, 1063–1069. [24] Aubel, D., Darfeuille-Michaud, A. and Joly, B. (1991) New adhesive factor (antigen 8786) on a human enterotoxigenic Escherichia coli O117:H4 strain isolated in Africa. Infect. Immun. 59, 1290–1299. [25] McConnell, M.M., Hibberd, M., Field, A.M., Chart, H. and Rowe, B. (1990) Characterization of a new putative colonization factor (CS17). from a human enterotoxigenic Escherichia coli of serotype O114:H21 which produces only heat-labile enterotoxin. J. Infect. Dis. 161, 343–347. [26] Grewal, H.M., Valvatne, H., Bhan, M.K., van Dijk, L., Gaastra, W. and Sommerfelt, H. (1997) A new putative fimbrial coloniza-

tion factor, CS19, of human enterotoxigenic Escherichia coli. Infect Immun. 65, 507–513. [27] Wolf, M.K., Taylor, D.N., Boedeker, E.C., Hyams, K.C., Maneval, D.R., Levine, M.M., Tamura, K., Wilson, R.A. and Echeverria, P. (1993) Characterization of enterotoxigenic Escherichia coli isolated from US troops deployed to the Middle East. J. Clin. Microbiol. 31, 851–856. [28] Nishimura, L.S., Giron, J.A., Nunes, S.L. and Guth, B.E. (2002) Prevalence of enterotoxigenic Escherichia coli strains harboring the longus pilus gene in Brazil. J. Clin. Microbiol. 40, 2606–2608. [29] Nagy, B., Moon, H.W. and Isaacson, R.E. (1976) Colonization of porcine small intestine by Escherichia coli: ileal colonization and adhesion by pig enteropathogens that lack K88 antigen and by some acapsular mutants. Infect. Immun. 13, 1214–1220.