Target genes for the identification and detection of ...

OECD Workshop Molecular Methods for Safe Drinking Water

Interlaken ‘98

Target genes for the identification and detection of potentially hazardous bacteria Joachim Frey, André Burnens, Katja Stuber, Jacques Nicolet and Peter Kuhnert Institute for Veterinary Bacteriology, University of Berne, Länggasstr. 122, CH-3012 Bern, Switzerland Keywords: Toxin genes, virulence genes, hybridisation, detection, broad range gene probes.

Summary The discovery during the last two decades of a large number of bacterial toxins and other virulence factors such as adhesins and invasins has led to a significantly better understanding of mechanisms of bacterial pathogenicity and powerful molecular methods for the accurate detection and identification of pathogenic bacteria. Hence, the different medically relevant pathogenic types of bacterial species such as Escherichia coli, Clostridium perfringens and several Pasteurella spp. can be rapidly and accurately identified and clearly distinguished from avirulent strains belonging to the same species by means of gene probes or PCR. However, for many pathogenic bacterial species no virulence factors or genes are known yet. For them, the differentiation of pathogenic types from less- or non-pathogenic types must rely on secondary markers or remains still impossible. We are currently developing broad range gene probes for the detection of potential virulence genes in possibly most or all bacterial species. For this purpose, all known toxin genes were grouped to toxin gene families, according to their genetic relationships. From this knowledge, broad range gene probes were derived by cloning and amplifying the most characteristic and best conserved parts of the different toxin gene families. Using low to medium stringent hybridisation conditions, these sets of gene probes detect known toxin genes which were not directly included in the gene probes and also indicate the presence of virulence genes in pathogenic bacterial species isolated from clinical samples as well as in type- and referencestrains of species from which no virulence genes were known yet. Using this approach, we have identified several new toxin genes from different pathogenic bacteria. Ultimately we are assessing broad range gene probes for type III secretion genes as general indicators for virulent bacteria. We prospect the broad range probes to be useful for the identification of waterborne pathogens.

Introduction The main goal in microbial quality control of drinking water is to ensure that it is free of microbes potentially hazardous to humans and animals. Therefore current quality standards prescribe that certain bacterial species generally known to be pathogenic must not be present or detectable in a given volume of drinking water, while other species which are known to be generally of no clinical significance might be tolerated up to a certain amount. This approach is widely and also successfully used in order to insure the control of safe drinking water. However, in the view of rapidly emerging new pathogens it is unclear whether this approach is still advisable. It is almost certain that some bacterial isolates of clinical and public health significance have not yet been recognised as pathogens. The most intriguing discovery in this context is the fact that in some bacterial species pathogenicity may vary considerably among isolates. Such species can represent strains ranging from highly dangerous isolates to entirely non-pathogenic strains that are biologically safe and can be used in fermentation processes or for the J. Frey et al.

1


Interlaken ‘98

manufacturing of medications. An approach, which takes advantage of the knowledge of molecular mechanisms of pathogenicity, is to search for virulence factors or their respective genes. To produce disease, pathogenic bacteria possess special structures called virulence factors which enable them to attach to, penetrate and multiply in the host cells or to produce toxic substances which impair the essential metabolism or structure of the host cell (Falkow, 1988). The combination of such virulence factors in a given bacterium and also the efficiency with which these genes are expressed under given circumstances during infection are the main determinants of the degree of virulence of a pathogen (Finlay and Falkow, 1989; Finlay and Falkow, 1997; Strauss and Falkow, 1997). Hence the highly pathogenic enterohemorrhagic Escherichia coli O157:H7 which caused several severe epidemics during the last few years, apparently emerged by the acquisition of toxin converting phages by a strain already possessing virulence factors, which in turn may have been acquired through horizontal spread (Whittam et al., 1993). It is therefore most evident to use virulence genes for the assessment of bacterial pathogenicity. The number of such virulence genes and of their varieties among the different bacterial species however is huge and it is assumed that most of them are still undiscovered (Strauss and Falkow, 1997). The assessment of bacterial virulence genes including yet unknown ones, respectively detection of bacteria harbouring such genes requests methodologies which target more generally the common genetic motives involved in virulence rather than specific sites of virulence genes. We are currently using three main approaches for this purpose. One method implying reverse dot blot hybridisation allows us to get the complete picture of virulence genes of any E. coli isolate. A second strategy involves the design of broad range gene probes for the detection of toxin gene families in Gram-negative and Gram-positive bacteria, which allows the detection and isolation of new toxin genes. Finally we are evaluating the use of common gene structures of type III secretion genes as general indicators for virulent bacteria.

Detection of Escherichia coli virulence genes Table 1: Virulence factors in E. coli Primary virulence factors

Secondary virulence factors

Shiga toxins (Stx1, Stx2)

Adherence factors/Fimbriae (P-, S-, F1Cfimbriae, Bfp, AAF)

Heat labile toxins (LTI, LTII)

Colonization factors (CFA/CS)

Heat stable toxins (STa, STb, EAST)

Invasion factors (Ipa)

Hemolysins (Hly, Ehx/Ely)

Iron transport systems (aerobactin/Iuc)

Cytotoxic CNF2)

necrotizing

factors

(CNF1, Capsule

Type III secretion system(s) (Sep) Intimin (Eae)

The widespread species Escherichia coli includes a broad variety of different types, ranging from highly pathogenic strains causing worldwide outbreaks of severe disease (Tauxe, 1997; Griffin and Tauxe, 1991; Bell et al., 1994) to avirulent isolates which belong to the normal intestinal flora or which are well known and safe laboratory strains (Muhldorfer et al., 1996). The pathogenicity of a given strain is mainly determined by specific virulence factors which include adhesins, invasins, toxins and capsule (Table 1). They are often organized in large genetic blocks, called pathogenicity islands located either on the chromosome or on large plasmids and which are

J. Frey et al.

2


Interlaken ‘98

often transmitted by bacteriophage or other mobile elements (Blum et al., 1994; Mcdaniel et al., 1995). Table 2: Probes used for detection of E. coli virulence- and target genes Probe

Name

Target gene

Reference

1a

Ferrichrom-iron receptor (Fhu)

fhuA

(Coulton et al., 1986)

2a

Type-1 fimbriae (Fim)

fimA

(Klemm, 1984)

3

P-fimbriae (Pap)

papA

(Marklund et al., 1992)

4

S-fimbriae (Sfa)

sfaA

(Schmoll et al., 1987)

sfaS

(Schmoll et al., 1989)

5 6

F1C fimbriae

foc

(van Die et al., 1984)

7

Bundel forming pilus (Bfp)

bfpA

(Donnenberg et al., 1992)

8

Colonization fimbriae (CFA/I)

cfa/I

(Karjalainen et al., 1989)

9

Colonization fimbriae (CS3)

cfa/II

(Boylan et al., 1988)

10

Aggregative adherence fimbriae (AAF/I)

aaf/I

(Schmidt et al., 1995b)

11

Intimin (Eae)

eae

(Jerse et al., 1990)

12

Invasion-plasmid antigen (Ipa)

ipaH

(Hartman et al., 1991)

13

Aerobactin

iucC

(Martinez et al., 1994)

14

K1 capsule antigen

neuA+ neuC

(Zapata et al., 1989; Zapata et al., 1992)

15

K5 capsule antigen

kfiB

unpublished

16

Heat-stable toxins (ST)

stIa/stIb

(Abe et al., 1990)

17

Shiga Toxin I (Stx1)

stxI

(Paton et al., 1993)

18

Shiga Toxin II (Stx2)

stxII

(Weinstein et al., 1988)

19

Heat-labile toxin I (LT-I)

eltI

(Spicer and Noble, 1982)

20

Heat-labile toxin II (LT-II)

eltIIa

(Pickett et al., 1987)

21

Alpha-hemolysin (HlyA)

hlyA

(Hess et al., 1986)

22

Enterohemolysin (ElyA)

elyA

(Schmidt et al., 1995a)

23

Cytotoxic necrotizing factor (CNF1)

cnf-1

(Falbo et al., 1993)

24

Low MW heat-stable toxin (EAST1)

astA

(Yamamoto and Echeverria, 1996)

a

The ferrichrom-iron receptor and the Type-1 fimbriae are not considered virulence factors. These gene probes are used as positive controls.

In order to efficiently test E. coli strains for the presence of known virulence genes we have developed a reverse dot-blot procedure. Specific segments of all currently known virulence genes of E. coli were designed to have similar hybridization parameters and were then subcloned on J. Frey et al.

3


Interlaken ‘98

plasmids and subsequently amplified by PCR as unlabelled probes in amounts sufficient to be bound to nylon membranes. Known pathogenic E. coli isolates and well known non-pathogenic E. coli laboratory strains were analyzed by labeling the genomic DNA of these strains with digoxigenin followed by hybridization to the prepared nylon membranes. These hybridization results demonstrated, that the ferrichrom-iron receptor gene (fhuA) seems to be ubiquitous in the species E. coli and that the Type-1 fimbria gene (fimA) is present in both pathogenic as well as in non-pathogenic E. coli. These genes therefore cannot be considered as targets for the detection of virulence but they are used as technical controls for the method (Kuhnert et al., 1997a). However, none of the other virulence genes (Table 2) reacted with the non-pathogenic E. coli strains, while clinical isolates of pathogenic E. coli control strains showed hybridization patterns indicating the typical virulence genes in these pathogens as expected from the clinical symptoms they caused. The described probes and their easy application on a single filter were shown to provide a useful tool for the assessment of the pathogenicity of E. coli strains in general and for safety assessment of strains to be used as hosts in biotechnological processes or for the identification and characterization of clinically significant E. coli isolates from human and animal species in particular (Kuhnert et al., 1997a).

Figure 1: Reverse dot-blot hybridization of environmental water isolates and control E. coli strains. The probes listed in table 2 were bound on nylon membranes as unlabeled PCR fragments. Numbering of the probes is the same as in table 2. Digoxigenin labeled total genomic DNA of E. coli K-12 (C600), the water isolates JF1585 and JF1586 and the uropathogenic E. coli (UPEC) RZ475 were hybridized to the nylon filters. After washing the filters under medium stringent

J. Frey et al.

4


Interlaken ‘98

conditions, signal was detected by chemiluminescence on X-ray films using CDP Star. Panel A: C600 (K-12), panel B: JF1585, panel C: JF1586, panel D: (RZ475) (UPEC) Using this method, we have compared randomly chosen E. coli strains isolated from surface water, with the well known non-pathogenic E. coli K-12 derivative which is approved as a biological safety strain and with an uropathogenic E. coli (UPEC) RZ475 isolated from a severely diseased patient. As documented in figure 1, E. coli K-12 (panel A) only shows signal with the probes for fhuA (probe # 1)and fimA (# 2) control genes and not with virulence gene probes, while the UPEC strain (panel D) in addition shows signals with the specific fimbrial genes pap (# 3), sfa (# 4, # 5) and foc (# 6) found in UPEC as well as the toxin genes hlyA (# 21), cnf-1 (# 23), the capsule K5 gene (# 15 and the gene probe for aerobactin iucC (# 13). The E. coli strains isolated from water differ in their hybridization pattern. Whereas strain JF1585 only gave a signal with the positive control fhuA and can be considered as apathogenic, strain JF1586 hybridized with various gene probes including the two positive controls, the fimbrial genes (pap, sfa, foc), and the toxins hlyA and cnf-1. These results confirm previous observations which indicated in the environment the presence of E. coli isolates that are devoid of virulence genes as well as isolates with toxin genes (Muhldorfer et al., 1996). Based on this report virulence genes specific for extraintestinal E. coli are actually often present in environmental water samples (approximately 40% of strains contain one or more of these genes). Nevertheless, the virulence gene combination/pattern of strain JF1586 is striking similar to uropathogenic representatives of extraintestinal E. coli and differs from virulence gene patterns found normally in environmental E. coli (Muhldorfer et al., 1996). These results show the capability of the detection method, since it does not only give information about the type of E. coli found but in addition it gives an assessment of its virulence potential.

Broad host range probes for the detection of unknown virulence genes. In order to enlarge the spectrum of applications of gene probes to the detection of potential, not yet known virulence genes in a broad range of bacterial species, we have developed broadrange virulence gene probes and special low to medium stringent hybridization conditions. This approach requested to establish families of genetically related toxins and major virulence factors of Gram-negative and Gram-positive bacterial species from DNA sequences as well as from biochemical data of toxins (Table 3). For each toxin family, phylogenetic trees were established based on the DNA sequences. These data enabled us to estimate the genetic relationship of these toxins and to select toxin genes for the construction of gene probes. The probes were then developed by PCR amplification of a set of fragments from the most common and characteristic parts of the selected toxin genes. Using this set of probes, we were able to detect a broad range of toxin genes of a given toxin family using controlled hybridization conditions. Typically, such broad range toxin gene probes were established for the family of RTX toxins (Kuhnert et al., 1997b) and for ADP ribosylating toxins. Gene probes for UDP glycosylating toxins and other toxin families are currently under investigation. Table 3: Families of major toxins and virulence factors RTX toxins ADP-ribosylating toxins UDP-glycosylating toxins Type III secretion systems Shiga toxins Cytotoxic necrotizing factors J. Frey et al.

5


Interlaken ‘98

Phospholipases Iron-transport systems (Siderophores) Adhesion/Invasion poteins

Detection of new genes encoding RTX toxins and ADP ribosylating toxin using broad range gene probes. The family of RTX toxins (repeats in the structural toxin) are pore forming, cytolytic or cytotoxic protein toxins which are widely spread in Gram-negative bacterial species where they are playing an important role in virulence (Welch, 1991; Welch et al., 1995). By comparing the genetic relationship of the RTX toxin genes a set of 10 gene probes was established to be used for screening of RTX toxin genes, including yet unknown ones in bacterial species. The probes include parts of apxIA, apxIIA and apxIIIA from Actinobacillus pleuropneumoniae, cyaA from Bordetella pertussis, frpA from Neisseria meningitidis, prtC from Erwinia chrysanthemi, hlyA and elyA of Escherichia coli, aaltA from Actinobacillus actinomycetemcomitans and lktA of Pasteurella haemolytica. A panel of pathogenic and non-pathogenic Gram-negative bacteria was investigated for the presence of RTX toxin genes. The probes detected all known genes for RTX toxins. Moreover, it showed positive hybridization signals for potential RTX toxin genes in several pathogenic bacterial species for which no such toxins are known yet. With this method we discovered and investigated a new RTX operon, named paxCABD in Pasteurella aerogenes. The entire operon was cloned and sequenced. It could be detected in various clinical isolates of P. aerogenes which were all associated with abortion cases in pigs. Further studies revealed that the Pax protein is not produced by P. aerogenes under culture conditions, but seems to be expressed during infection in pigs (P. Kuhnert and J. Frey, unpublished results). Two other RTX genes were found with this method and further confirmed by analysis in Campylobacter rectus a Gram-negative, anaerobic motile bacterium often associated with periodontal disease in humans. In addition, signals of further potential RTX toxin genes were also detected in pathogenic strains of Yersinia enterocolitica and in Pseudomonas spp. (P. Kuhnert and J. Frey, unpublished results). Similar to the RTX gene probes, we developed broad range gene probes for ADP ribosylating toxins. This family includes very potent toxins, which are produced by several highly pathogenic Gram-positive and Gram-negative bacteria. Bacteria containing ADP ribosylating toxins like Vibrio cholerae, Corynebacterium diphtheriae and Bordetella pertussis generally cause important diseases such as cholera, diphtheria and whopping cough, respectively. From the 22 currently known genes of ADP ribosylating toxins characteristic conserved domains from the genes of 11 selected toxins have been used in order to construct a set of toxin probes. This set of probes detects all genes of the known ADP ribosylating toxins and indicates the presence of new, yet unknown ADP ribosylating toxins in certain bacterial isolates. These results show the capability of our approach to detect new toxin genes in bacterial species where no toxins were yet described and demonstrate the importance of toxin gene detection in bacteriological quality and safety assessment procedures.

Type III secretion genes as general indicators for virulent bacteria Pathogenicity islands and type III secretion systems have been shown to be two major traits that are present in a broad range of bacterial pathogens (Hacker et al., 1997). Pathogenicity islands are large unstable pieces of DNA carrying clusters of genes essential for pathogenicity. All pathogenicity islands discovered so far show evidence for acquisition by horizontal spread from foreign micro-organisms, like G+C content differing from the bulk of the genome, their instability, the presence of repeated or insertion elements at their ends, and their insertion at similar sites in J. Frey et al.

6


Interlaken ‘98

the genome (e.g. tRNA’s coding for rare amino acids). Pathogenicity islands are missing in nonvirulent isolates of the same species or bacterial species closely related to the respective pathogen. They constitute a potent means for evolving virulence by a single genetic event, i.e. the acquisition of the pathogenicity island by a bacterial host. Thus, E. coli K-12 was shown to become fully virulent in an animal model upon transfer of the pathogenicity island of enteropathogenic E. coli (the LEE locus of EPEC) (Mcdaniel and Kaper, 1997). The majority of pathogenicity islands discovered so far were shown to contain a type III secretion system. Type III secretion systems provide a mode for bacteria to pilot virulence factors directly to host cells. These secretion systems therefore differ from other secretion systems by the fact that they are directly involved in mechanisms of bacterial virulence, in contrast to the general secretory (or type II) system which has more general (housekeeping) tasks in bacteria and to the type I system which appears to be specialised for the secretion of certain proteins like RTX proteins. The recently discovered type III secretion systems are triggered when a pathogen comes in close contact with host cells and targets virulence factors directly at host cells. Type III secretion systems are complex systems involving several different determinants with different functions (Table 4, see appendix). The virulence factors then modify host cell functions, e.g. phosphorylation of signal transduction molecules, to the pathogens benefit.

Figure 2: Phylogenetic relation of the lcrD homologues of the type III secretion system. The 5’terminal 800 bp of the genes, which constitute the hybridisation probes, were aligned with the PILEUP programme and phylogenetic relationships were established with the Mega 1.02 programme. Corrected distances were calculated with the Jukes-Cantor algorithm, and a tree was constructed by using the neighbour-joining method. Bootstrap values, calculated from 1000 trees are indicated at the branchings. The scale bar represents 0.1% sequence divergence. Genes that were used for the constructions of the broad-spectrum probes are represented in bold. Screening bacterial isolates for the presence of type III secretion systems could be used for the general detection of potential pathogenicity of bacteria and to indicate the presence of pathogenicity islands. Due to the complex structure of type III secretion systems, a simple rationale for the universal screening for type III secretion systems needed to be developed. Because it seems plausible that the type III secretion systems evolved from the apparatus for flagellar assembly (based on the structural homology of the proteins involved) we concentrated on one integral component of the secretion apparatus which appeared to be a relatively well conserved protein, rather than the secreted virulence factors. The advantage of this approach would thus be the ability to detect organisms carrying type III secretion systems with a limited set of probes. After studying the amino acid homologies of the proteins respectively the nucleotide J. Frey et al.

7


Interlaken ‘98

homologies of their genes involved in type III secretion systems known so far (Table 4), the most obvious candidates were the members of the lcrD-invA-flhA superfamily constituting the genes of the membrane spanning protein of the type III secretion apparatus. This superfamily includes the low calcium response (lcrD) of Yersinia pestis, the invasin (invA) of Salmonella typhimurium and the flagellar hook protein (flhA) of Escherichia coli (Table 5). The phylogenetic relationship of the different members of this superfamily is shown in Figure 2. Since the flhA gene is not originating from a type III secretion system but rather from a putative predecessor, the flagellar hook of E. coli which is also present in the apathogenic K-12 derivatives, flhA was taken as an outgroup in our calculations. By comparing the sequences encoding the family of membrane spanning proteins of known type III secretion systems with the flhA sequence of E. coli K-12 as an outgroup, a region of ca. 800 base pairs of sequence homology comprised in all members of the superfamily was delineated. The overall homology found was around 70 %. The corresponding fragments of the lcrD (Yersinia enterocolitica), mxiA (Shigella sonnei), sepA (enterohemorrhagic E. coli, EHEC), invA (Salmonella typhimurium), hrcV (phytopathogenic Erwinia amylovora), and flhA (E. coli K-12) genes (Table 5) were amplified by PCR, and the fragments were cloned on plasmids. Probes were made from the cloned fragments and labelled with DIG-dUTP. Hybridisation experiments showed that there was a certain degree of cross-hybridisation among the probes as expected. However, the reaction obtained with the non-virulence related flhA outgroup probe, could always be separated clearly. The probe based on flhA therefore was used as a control rather than an indicator for the presence of type III secretion system genes. A dot-blot hybridisation assay was used to validate the detection of the homologous virulence gene in a representative set of isolates of the species where the corresponding gene of the type III secretion system is expected. For this validation, 146 clinical isolates were analysed. They included 62 strains of E. coli belonging to established pathogenicity groups (enterovirulent: ETEC, EPEC, EIEC, EHEC; uropathogenic: UPEC), 41 strains of Salmonella sp., 31 shigellae belonging to the four known serogroups, and 15 strains belonging to the genus Yersinia. The results of the hybridisation experiments with the probes derived from the lcrD-invA-flhA superfamily (Table 3) were in excellent agreement with clinical conditions from which the strains were isolated. Positive reactions with one or several of the probes (due to the expected cross hybridisations) were found for 134 of the 149 strains tested. From the 15 strains that were negative on our hybridisation assay, 4 strains were E. coli UPEC strains. Analysis of these strains with the probes for the E. coli - specific virulence genes revealed that they contained the genes which are characteristic for UPEC (Fig. 1). The negative hybridisation results of these four strains with the probes for the lcrD homologues could indicate that they do not contain type III secretion genes or else that the UPEC type III secretion system, if present, shows insufficient homology with the probes used. So far, no type III secretion systems have been reported for strains of UPEC (Hueck, 1998). The other 11 strains showed unexpected negative reactions. Further investigations using the virulence gene probes showed that these strains had lost their virulence plasmid during propagation on culture medium. The virulence plasmid is known to be lost very rapidly in vitro in strains of Shigella spp. and Y. enterocolitica under certain conditions of culture. In these bacteria the ‘pathogenicity island’ is carried on the virulence plasmid (Burnens et al., 1996). Loss of pathogenicity islands under in vitro culture conditions is known to result in loss of virulence in all established infection models. Thus the negative reactions with our probes is predictive of the observed lack of pathogenicity showing that the probes based on the genes of the lcrD-invA-flhA superfamily gave reliable results on the potential pathogenicity of a given isolated strain. This consideration may also be important in the view of water hygiene where strains that originated from diseased persons of animals and initially were pathogenic may under certain circumstances loose their virulence factors and become no longer hazardous. J. Frey et al.

8


Interlaken ‘98

Table 5: Genes used for the construction of type III specific gene probes Gene

Protein name

Species

Reference

lcrD

low calcium response protein Yersinia enterocolitica

(Plano et al., 1991)

invA

invasin

Salmonella typhimurium

(Galan et al., 1992)

MxiA

membrane associated expression of invasion plasmid antigens

Shigella sonnei

(Andrews and Maurelli, 1992)

SepA

secreted enteropathogenic protein

Escherichia coli

(Jarvis et al., 1995)

HrcV

hypersensitive response and pathogenicity (hrp)

Erwinia amylovora

(Wei and Beer, 1993)

flhA

flagellar hook protein

Escherichia coli K-12

(Itoh et al., 1996)

The validation of the probes was further extended to strains isolated from the environment and from healthy individuals as well as to properly defined non-pathogenic laboratory strains, mainly belonging to the family of Enterobacteriaceae which were chosen as control group for non-pathogenic organisms. As expected, 48 of 51 isolates, including all strains of E. coli used for biotechnology (10 strains) were devoid of sequences hybridising with probes for type III secretion systems. However, three E. coli strains gave a signal with the sepA probe for the type III system of pathogenic E. coli. Further analysis of these three strains using the above described virulence gene probes revealed that they belonged to the enteropathogenic group of E. coli, EPEC and hence must be considered as pathogenic E. coli. The results from our investigation indicate that the gene probes derived from the genes of the lcrD-invA-flhA superfamily of the type III secretion system is a valuable system to trace pathogenic bacteria belonging to the family of Enterobacteriaceae. We are currently evaluating the system further for its ability to be used as a more general system for the tracing of pathogenic bacteria.

Concluding remarks Bacterial pathogenicity is a complex multi-factorial mechanism involving a large number of virulence factors. They include adhesins, attachment functions, host cell surface modifying factors (effacement factors), invasins, and many different toxins as well as secretion systems, which export toxins and other virulence factors and pilot them to the target host cells. During evolution, virulence factors and the genes coding for them adapted in their particular hosts to specialised tasks during infection leading to the creation of an immense amount of different virulence genes of which probably only a small part has been discovered yet. This variability is almost certain to increase dramatically, if pathogenicity islands are indeed able of horizontal spread between genera of pathogens as suggested by recent data (Schubert et al., 1998). Since virulence factors are directly involved in the mechanisms of bacterial pathogenicity, their genes represent ideal targets for the development of molecular methods for quality control of water resources destined as drinking water either directly or indirectly after processing. The immense variations of bacterial virulence genes, and their ability of horizontal spread, calls for the development of rational approaches for this task. Our studies show that the common structures of the most prominent bacterial toxins and other central virulence determinants such as type III secretion systems were J. Frey et al.

9


Interlaken ‘98

able to clearly distinguish between non-pathogenic and pathogenic bacterial isolates. They can therefore serve as main targets for the molecular genetic detection of pathogenic or toxigenic bacterial contaminants. The use of these target sequences in combination with the rapidly developing multi-array gene sensor systems (biochips) will constitute a powerful tool for the molecular detection of pathogenic micro-organisms in future quality control for safe drinking water. REFERENCES Abe, A., Komase, K., Bangtrakulnonth, A., Ratchtrachenchat, O.A., Kawahara, K., and Danbara, H. (1990) Trivalent heat-labile- and heat-stable-enterotoxin probe conjugated with horseradish peroxidase for detection of enterotoxigenic Escherichia coli by hybridization. J.Clin.Microbiol. 28, 2616-2620. Andrews, G.P. and Maurelli, A.T. (1992) mxiA of Shigella flexneri 2a, which facilitates export of invasion plasmid antigens, encodes a homolog of the low-calcium-response protein, LcrD, of Yersinia pestis. Infect.Immun. 60, 3287-3295. Bell, B.P., Goldoft, M., Griffin, P.M., Davis, M.A., Gordon, D.C., Tarr, P.I., Bartleson, C.A., Lewis, J.H., Barrett, T.J., Wells, J.G., Baron, R., and Kobayashi, J. (1994) A multistate outbreak of Escherichia coli O157:H7- associated bloody diarrhea and hemolytic uremic syndrome from hamburgers - The Washington experience. JAMA 272, 1349-1353. Blum, G., Ott, M., Lischewski, A., Ritter, A., Imrich, H., Tschape, H., and Hacker, J. (1994) Excision of large DNA regions termed pathogenicity islands from tRNA-specific loci in the chromosome of an Escherichia coli wild- type pathogen. Infect.Immun. 62, 606-614. Boylan, M., Smyth, C.J., and Scott, J.R. (1988) Nucleotide sequence of the gene encoding the major subunit of CS3 fimbriae of enterotoxigenic Escherichia coli. Infect.Immun. 56, 3297-3300. Burnens, A.P., Frey, A., and Nicolet, J. (1996) Association between clinical presentation, biogroups and virulence attributes of Yersinia enterocolitica strains in human diarrhoeal disease. Epidemiol.Infect. 116, 27-34. Coulton, J.W., Mason, P., Cameron, D.R., Carmel, G., Jean, R., and Rode, H.N. (1986) Protein fusions of beta-galactosidase to ferrichrome-iron receptor of Escherichia coli K-12. J.Am.Vet.Med.Assoc. 165, 181-192. Donnenberg, M.S., Giron, J.A., Nataro, J.P., and Kaper, J.B. (1992) A plasmid-encoded type IV fimbrial gene of enteropathogenic Escherichia coli associated with localized adherence. Mol.Microbiol. 6, 3427-3437. Falbo, V., Pace, T., Picci, L., Pizzi, E., and Caprioli, A. (1993) Isolation and nucleotide sequence of the gene encoding cytotoxic necrotizing factor-1 of Escherichia-coli. Infect.Immun. 61, 4909-4914. Falkow, S. (1988) Molecular Koch's postulates applied to microbial pathogenicity. Rev.Infect.Dis. 10 Suppl 2, S274-S276 Finlay, B.B. and Falkow, S. (1989) Common themes in microbial pathogenicity. Microbiol.Rev. 53, 210-230. Finlay, B.B. and Falkow, S. (1997) Common themes in microbial pathogenicity revisited. Microbiol.Mol.Biol.Rev. 61, 136-169.

J. Frey et al.

10


Interlaken ‘98

Galan, J.E., Ginocchio, C., and Costeas, P. (1992) Molecular and functional characterization of the Salmonella invasion gene invA: homology of InvA to members of a new protein family. J.Bacteriol. 174, 4338-4349. Griffin, P.M. and Tauxe, R.V. (1991) The epidemiology of infections caused by Escherichia coli O157: H7, other enterohemorrhagic E. coli, and the associated hemolytic uremic syndrome. Epidemiol.Rev. 13, 60-98. Hacker, J., BlumOehler, G., Muhldorfer, I., and Tschape, H. (1997) Pathogenicity islands of virulent bacteria: Structure, function and impact on microbial evolution. Mol.Microbiol. 23, 1089-1097. Hartman, A.B., Venkatesan, M., Oaks, E.V., and Buysse, J.M. (1991) Sequence and molecular characterization of a multicopy invasion plasmid antigen gene, ipaH, of Shigella flexneri. J.Bacteriol. 172, 1905-1915. Hess, J., Wels, W., Vogel, M., and Goebel, W. (1986) Nucleotide sequence of a plasmid-encoded hemolysin determinant and its comparison with a corresponding chromosomal sequence. FEMS Microbiol.Letters 34, 1-11. Hueck, C.J. (1998) Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol.Mol.Biol.Rev. 62, 379-433. Itoh, T., Aiba, H., Baba, T., Hayashi, K., Inada, T., Isono, K., Kasai, H., Kimura, Kitakawa, M., Kitagawa, M., Makino, K., Miki, T., Mizobuchi, K., Mori, H., Mori, T., Motomura, K., Nakade, S., Nakamura, Y., Nashimoto, H., Nishio, Y., Oshima, T., Saito, N., Sampei, G., Seki, Y., and Horiuchi, T. (1996) A 460-kb DNA sequence of the Escherichia coli K-12 genome corresponding to the 40.1-50.0 min region on the linkage map. DNA Res. 3, 379392. Jarvis, K.G., Giron, J.A., Jerse, A.E., Mcdaniel, T.K., Donnenberg, M.S., and Kaper, J.B. (1995) Enteropathogenic Escherichia coli contains a putative type III secretion system necessary for the export of proteins involved in attaching and effacing lesion formation. Proc.Natl.Acad.Sci.U.S.A 92, 7996-8000. Jerse, A.E., Yu, J., Tall, B.D., and Kaper, J.B. (1990) A genetic locus of enteropathogenic Escherichia coli necessary for the production of attaching and effacing lesions on tissue culture cells. Proc.Natl.Acad.Sci.USA 87, 7839-7843. Karjalainen, T.M., Evans, D.G., So, M., and Lee, C.H. (1989) Molecular cloning and nucleotide sequence of the colonization factor antigen I gene of Escherichia coli. Infect.Immun. 57 , 1126-1130. Klemm, P. (1984) The fimA gene encoding the type-1 fimbrial subunit of Escherichia coli. Nucleotide sequence and primary structure of the protein. Eur.J.Biochem. 143, 395-399. Kuhnert, P., Hacker, J., Muhldorfer, I., Burnens, A.P., Nicolet, J., and Frey, J. (1997a) Detection system for Escherichia coli-specific virulence genes: Absence of virulence determinants in B and C strains. Appl.Environ.Microbiol. 63, 703-709. Kuhnert, P., HeybergerMeyer, B., Burnens, A.P., Nicolet, J., and Frey, J. (1997b) Detection of RTX toxin genes in gram-negative bacteria with a set of specific probes. Appl.Environ.Microbiol. 63, 2258-2265. Marklund, B.I., Tennent, J.M., Garcia, E., Hamers, A., Baga, M., Lindberg, F., Gaastra, W., and Normark, S. (1992) Horizontal gene transfer of the Escherichia coli pap and prs pili J. Frey et al.

11


Interlaken ‘98

operons as a mechanism for the development of tissue-specific adhesive properties. Mol.Microbiol. 6, 2225-2242. Martinez, J.L., Herrero, M., and de Lorenzo, V. (1994) The organization of intercistronic regions of the aerobactin operon of pColV-K30 may account for the differential expression of the iucABCD iutA genes. J.Mol.Biol. 238, 288-293. Mcdaniel, T.K., Jarvis, K.G., Donnenberg, M.S., and Kaper, J.B. (1995) A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens. Proc.Natl.Acad.Sci.U.S.A. 92, 1664-1668. Mcdaniel, T.K. and Kaper, J.B. (1997) A cloned pathogenicity island from enteropathogenic Escherichia coli confers the attaching and effacing phenotype on E. coli K-12. Mol.Microbiol. 23, 399-407. Muhldorfer, I., Blum, G., Donohue-Rolfe, A., Heier, H., Olschlager, T., Tschape, H., Wallner, U., and Hacker, J. (1996) Characterization of Escherichia coli strains isolated from environmental water habitats and from stool samples of healthy volunteers. Res.Microbiol. 147, 625-635. Paton, A.W., Paton, J.C., Goldwater, P.N., Heuzenroeder, M.W., and Manning, P.A. (1993) Sequence of a variant Shiga-like toxin type-I operon of Escherichia coli O111-H-. Gene 129, 87-92. Pickett, C.L., Weinstein, D.L., and Holmes, R.K. (1987) Genetics of type IIa heat-labile enterotoxin of Escherichia coli: operon fusions, nucleotide sequence, and hybridization studies. J.Bacteriol. 169, 5180-5187. Plano, G.V., Barve, S.S., and Straley, S.C. (1991) LcrD, a membrane-bound regulator of the Yersinia pestis low-calcium response. J.Bacteriol. 173, 7293-7303. Schmidt, H., Beutin, L., and Karch, H. (1995a) Molecular analysis of the plasmid-encoded hemolysin of Escherichia coli 0157:H7 strain EDL 933. Infect.Immun. 63, 1055-1061. Schmidt, H., Knop, C., Franke, S., Aleksic, S., Heesemann, J., and Karch, H. (1995b) Development of PCR for screening of enteroaggregative Escherichia coli. J.Clin.Microbiol. 33, 701-705. Schmoll, T., Hacker, J., and Goebel, W. (1987) Nucleotide sequence of the sfaA gene coding for the S-fimbrial protein subunit of Escherichia coli. FEMS Microbiol.Letters 41, 229-235. Schmoll, T., Hoschutzky, H., Morschhauser, J., Lottspeich, F., Jann, K., and Hacker, J. (1989) Analysis of genes coding for the sialic acid-binding adhesin and two other minor fimbrial subunits of the S-fimbrial adhesin determinant of Escherichia coli. Mol.Microbiol. 3, 1735-1744. Schubert, S., Rakin, A., Karch, H., Carniel, E., and Heesemann, J. (1998) Prevalence of the ''highpathogenicity island'' of Yersinia species among Escherichia coli strains that are pathogenic to humans. Infect.Immun. 66, 480-485. Spicer, E.K. and Noble, J.A. (1982) Escherichia coli heat-labile enterotoxin. Nucleotide sequence of the A subunit gene. J.Biol.Chem. 257, 5716-5721. Strauss, E.J. and Falkow, S. (1997) Microbial pathogenesis: genomics and beyond. Science 276, 707-712. Tauxe, R.V. (1997) Emerging foodborne diseases: an evolving public health challenge. Emerg.Infect.Dis. 3, 425-434. J. Frey et al.

12


Interlaken ‘98

van Die, I., van Geffen, B., Hoekstra, W., and Bergmans, H. (1984) Type 1C fimbriae of a uropathogenic Escherichia coli strain: cloning and characterization of the genes involved in the expression of the 1C antigen and nucleotide sequence of the subunit gene. Gene 34, 187-196. Wei, Z.M. and Beer, S.V. (1993) HrpI of Erwinia amylovora functions in secretion of harpin and is a member of a new protein family. J.Bacteriol. 175, 7958-7967. Weinstein, D.L., Jackson, M.P., Samuel, J.E., Holmes, R.K., and O Brien, A.D. (1988) Cloning and sequencing of a Shiga-like toxin type II variant from Escherichia coli strain responsible for edema disease of swine. J.Bacteriol. 170, 4223-4230. Welch, R.A. (1991) Pore-forming cytolysins of gram-negative bacteria. Mol.Microbiol. 5, 521528. Welch, R.A., Bauer, M.E., Kent, A.D., Leeds, J.A., Moayeri, M., Regassa, L.B., and Swenson, D.L. (1995) Battling against host phagocytes: The wherefore of the RTX family of toxins? Infect.Agent.Dis. 4, 254-272. Whittam, T.S., Wolfe, M.L., Wachsmuth, I.K., Orskov, F., Orskov, I., and Wilson, R.A. (1993) Clonal relationships among Escherichia coli strains that cause hemorrhagic colitis and infantile diarrhea. Infect.Immun. 61, 1619-1629. Yamamoto, T. and Echeverria, P. (1996) Detection of the enteroaggregative Escherichia coli heat- stable enterotoxin 1 gene sequences in enterotoxigenic E. coli strains pathogenic for humans. Infect.Immun. 64, 1441-1445. Zapata, G., Crowley, J.M., and Vann, W.F. (1992) Sequence and expression of the Escherichia coli K1 neuC gene product. J.Bacteriol. 174, 315-319. Zapata, G., Vann, W.F., Aaronson, W., Lewis, M.S., and Moos, M. (1989) Sequence of the cloned Escherichia coli K1 CMP-N-acetylmuraminic acid synthetase gene. J.Biol.Chem. 264, 14769-14774.

J. Frey et al.

13


Interlaken ‘98

$SSHQGL[ 7DEOHPresumed currently known gene homologues of type III secretion proteins in pathogenic bacteria and of flagellar biosynthesis )XQFWLRQLQWKHW\SH,,,VHFUHWLRQV\VWHP %DFWHULDOVSHFLHV VHFUHWLRQV\VWHP

Outer Lipo- LQQHU ATPase inner inner inner inner Secreted membrane protein PHPEUDQH membrane membrane membrane membrane protein VSDQQLQJ

-

Surface Cytosolic Chaperon Secretion /secreted regulation protein