1. INTRODUCTION
The tellurite resistance of human pathogens is known nearly 100 years. More than 20 years are known the genes encoding for the highest resistance phenotype, so called ter genes, or ter operon. However, the mechanism of the resistance remains unclear.
At our Department of Molecular Biology, these genes have been described on a large conjugative plasmid pTE53 and their relation to the operon found in an emergent food-born pathogen, enterohaemorhagic Escherichia coli (E. coli) O157:H7 was shown.
Expression of these genes seems to be constitutive, regulated also on the translational level, however the inducibility of them in some isolates of E. coli O157:H7 as well as those located on the chromosome of urinary tract pathogen Proteus mirabilis has been reported.
In our work we report the constitutive character of the ter operon found in uropathogenic
E. coli, its organization on the level of transcription and we compare its expression with the operon originated from E. coli O157:H7.
To study the biological mechanism of ter gene products involved in the resistance phenotype, we have cloned all four essential genes, e.g. terB, terC, terD and terE into pET28 expression system and here we describe the preparation of the system to show the biological function of pET-expressed TerB protein. This protein is fused with His-tag and here we report also its purification using Nickel-based Promega system. It is the initial step to express and study the function of potential multiprotein complex involved in tellurite resistancemediated by the ter operon
2. LITERARY OVERVIEW
2.1. Tellurium (Te)
Tellurium (atomic number 52) is a semimetal that belongs to the same group (VIA) in Periodic Table as the elements sulfur and selenium. Estimated to rank 75th in abundance of all the elements in the earth's crust, tellurium is found in low concentrations (about 0,002 ppm) throughout the environment (Schroeder et al., 1967).
The oxyanions of tellurium, tellurite (K2TeO3) and tellurate, are highly toxic for most micro-organisms, causing direct oxidation of cellular thiols (Turner et al., 1999) or, following reduction to telluride, it is inappropriately incorporated in place of sulfur in amino acids (Garberg et al., 1999; Taylor, 1999). The human body contains fairly large amounts of tellurium (up to 600 mg) (Nason and Schroeder, 1967).
Tellurium compounds have a long history as antimicrobial and therapeutic agents. Prior to the development and marketing of today's antibiotics, they were used to treat conditions such as leprosy, tuberculosis, dermatitis, cystitis and eye infections. Alexander Fleming reported the antibacterial properties of penicillin in 1950 and potassium tellurite in 1932. Tellurite has been used for at least 80 years in selective media for the isolation of pathogens, including Corynebacterium diphtheriae, Staphylococcus aureus, Vibrio cholerae and Shigella spp. (Zadik et al., 1993). For example, selective medium for the “hamburger disease” bacterium (verocytotoxigenic Escherichia coli O175:H7) was recently developed by Biosynth™. In 1977, resistance to K2TeO3 was linked to the presence of plasmids in enteric bacteria and in Pseudomonas (Summers and Jacoby, 1977).
Tellurium is used as a catalyst in the film industry, in the manufacture of batteries, alloys, and rubbers, and as a coloring agent in glass (Browning, 1969). An immunomodulating compound containing tellurium has been tested in rats for possible use in treating AIDS and cancer patients (Sredni et al., 1987; Nyska et al., 1989).
2.2. Tellurite toxicity
Nitrate reductase is responsible for the basal level of tellurite resistance found in E. coli (Avazeri et al., 1997).Tellurite, which circumvents the initial line of defense, is acted on by glutathione and/or other reduced thiols in the cytoplasm to generate additional
TeO3-2→Te0 reduction. Superoxide dismutase then acts on the O2- resulting from this reduction. The elimination of components of this cascade leads to the oxidation of cellular thiols, resulting in the shutdown of DNA synthesis, protein synthesis and most reductases (Turner et al., 1995).
To survive in the presence of tellurite, in absence of efflux, bacteria must convert it to a form that is less toxic. Electron microscopy and imaging have shown that TeR bacteria, which form black colonies, contain intracellular crystals of black metallic tellurium that are often located just inside the inner membrane (Taylor, 1999).
2.3. Bacterial Resistance to Tellurium compounds
Some gram-positive bacteria, including Corynebacterium diphtheriae (Conradi and Troch, 1912), Streptococcus faecalis (now called Enterococcus faecalis) (Skadhauge, 1950; Appleman and Heinmiller, 1961), and most strains of Staphylococcus aureus (Hoeprich et al., 1960), are naturally resistant to potassium tellurite. For this reason, tellurite medium has long been used as a means of identification of C.diphtheriae. In addition, the ability of mycobacteria to produce black colonies when grown on medium containing tellurite has been used to test for the viability of the tubercle bacilli (Corper, 1915; Kilburn et al., 1969). The resistance of Wautersia (Alcaligenes) to tellurite is used to differentiate A. faecalis and A. denitrificans from Bordetella bronchiseptica (Jonson and Sneath, 1973). Some coliform bacteria, which are extremely sensitive to tellurite, can be acclimatized to higher concentrations by passage in tellurite broth (Fleming and Young, 1940).
Gram-negative tellurite-resistant bacteria are frequently isolated from hospital and urban sewages and film-reprocessing sludge (Taylor and Summers, 1979). In the family Enterobacteriaceae, TeR is often mediated by plasmids.
At least five TeR determinants have been identified by genetic studies and DNA sequencing, all apparently unrelated to one another neither on the DNA or protein level (Taylor, 1999).
2. 3. 1. Plasmid-mediated TeR
Plasmids of the incompatibility group (Inc)HI2 in Enterobacteriaceae and the IncP2 group in Pseudomonas sp. are often associated with TeR (Summers and Jacoby, 1977; Taylor and Summers, 1979). IncHII plasmids found in Enterobacteriaceae also mediated TeR (Bradley et al., 1982).
2. 3. 1. 1. The TeR Determinants of the IncHIPlasmid, pMER610
The IncHI and IncHII plasmids are large (greater than 150 kb), conjugative plasmids that encode multiple drug resistances (Anderson and Smith, 1972; Anderson, 1975). The plasmid pMER610 belongs to the IncHI2 subgroup and was originally isolated from a Wautersia (Alcaligenes) species of bacteria. It is a large (>250 kb) plasmid that encodes resistance to mercury and tellurium compounds (Jobling and Ritchie, 1987).
The TeR determinant on pMER610 is contained within a 3,55-kb DNA fragment. Nucleotide sequence analysis indicated that the pMER610 system consisted of five genes, designed terA, -B, -C, -D, and -E, with predicted poroducts of 37, 14, 38, 20, and 20 kDa, respectively . Four polypeptides were detected using “maxicell” analysis. The genes have been sequenced and five open reading frames (ORFs) were detected. ORFs 1 and 2 encode polypeptides of 37 and 14 kDa in size, respectively. Insertion of the transposon Tn1000 into these genes resulted in only partial loss of resistance to tellurite, suggesting that ORFs1 and 2 may be regulatory rather than structural genes. Insertion of Tn1000 into ORFs 3 and 4 resulted in complete loss of resistance to tellurite, suggesting that these are the structural genes determining TeR. ORF3 encodes a protein of 38 kDa , which is highly hydrophobic and has nine potential membrane-spanning domains.
ORFs 4 and 5, both of which encode 20-kDa polypeptides, have 66% nucleotide homology and may have arisen probably from a gene duplication event.
Expression of TeR by Escherichia coli carrying pMER610 and its subclones appeared to be inducible by prior exposure to subtoxic levels of potassium tellurite. However, studies using transcriptional and translational fusions to β-galactosidase suggest that the TeR genes themselves are constitutively expressed (Jobling and Ritchie, 1988).
2. 3. 1. 2. The TeR Determinant of the IncHII Plasmid, pHH1508a
The 208 kb IncHII plasmid pHH1508a was originally isolated from a strain of Klebsiella aerogenes obtained from a patient with typhoid fever (Datta et al., 1981). Expression of TeR by pHH1508a and other IncHII and IncHI2 plasmids has frequently been linked to other unusual properties such as inhibition of coliphage development and colicin B resistance; however, no functional relationship has yet been shown (Taylor and Summers, 1979; Maher et al., 1989).
A 1,8 kb region responsible for expression of TeR by pHH1508a has been cloned and sequenced (Walter et al., 1991). This region, which shows no homology to the TeR determinant on pMER610, encodes two genes that have been named tehA and tehB. The tehB gene encodes a 23 kDa polypeptide that is relatively hydrophilic and is probably located in the cytoplasm. The tehA gene encodes a 36 kDa polypeptide, which migrates with an apparent molecular mass of 28 kDa in SDS-polyacrylamide gel electrophoresis (SDS-PAGE) due to its high hydrophobicity. It appears to have between 5 and 10 membrane-spanning domains and is likely to be located in the inner membrane of the bacterium. Three cysteine residues appear to be located within 3 central putative membrane-spanning domains and could potentially have a role in resistance by binding ligands.
2. 3. 1. 3. The Cryptic TeR Determinant of the IncPα Plasmids, RK2
Plasmids of the P incompatibility group have the ability to transfer between, and replicate and be stably maintained in a wide variety of gram-negative bacteria (reviewed by Thomas and Smith, 1987). RK2 and plasmids named RP4, RP1, R18 and R68 were originally isolated from gram-negative bacteria obtained from a hospital in Birmingham (UK) in 1969 (Holloway and Richmond, 1973; Ingram et al., 1973). Through the use of Tn7 insertion mutagenesis, the RK2 TeR determinant was located between the kilA and korA genes, which are involved in plasmid replication control (Taylor and Bradley, 1987). The 3 kb region contains an operon specifying three genes, kilA, telA, and telB , all of which are necessary for expression of high-level resistance to tellurite (Walter, 1990; Goncharoff et al., 1991). No DNA or amino acid sequence homology to the IncHI or IncHII TeR determinants could be detected. The first gene in the operon is the kilA gene, which encodes a 28-kDa hydrophilic polypeptide that is probably located in the cytoplasm (Walter et al., 1991). This gene was previously identified by Figurski et al. (1982) and named for the lethal effect that was observed, when it was cloned separately from the korA (kill-override), which negatively regulates transcription from the kilA promoter (Young et al., 1985, 1987). Further work by this group suggested that each of the three genes (kilA, telA, and telB) is able to express a host-lethal phenotype (Goncharoff et al., 1991).
The second gene, telA, encodes a 42 kDa hydrophilic polypeptide, which is likely to be located in the cytoplasm (Walter et al., 1991). The last gene in the operon, telB, encodes a 32 kDa hydrophobic polypeptide.
Analysis of the amino acid sequence suggests that the TelB polypeptide has four membrane-spanning domains. A proposed model for the topology of TelB in the inner membrane has been based on hydrophobicity plot analysis and TnphoA insertion mutagenesis (Walter, 1990). Given this model, it appears that the two cysteine residues (Cys-125 and Cys-132) in TelB are located on the cytoplasmic face of the protein, where they could potentially be involved in binding to ligands. The importance of Cys-125 in the expression of TeR was demonstrated by comparison of the DNA sequences of TeR and TeS isolates of RK2. The DNA sequence of the kilA-telAB operon on a TeS of RK2 is identical to the TeR sequence except for a single nucleotide change in amino acid position 125 in the telB gene resulting in a Ser (tellurite sensitive) instead of Cys (tellurite resistance) (Walter et al.,1991; Goncharoff et al., 1991).
2. 3. 1. 4. TeR Determinant from plasmid R478
pR478 is a 272-kb self-transmissible plasmid of the incompatibility subgroup IncHI2, which mediates resistance to kanamycin, chloramphenicol, tetracycline, mercuric chloride, potassium tellurite (TeR), and arsenic compounds, as well as resistance to some bacteriophages (Phi) and to pore-forming colicins (PacB) (Maher et al., 1989; Rodriguez-Lemoine, 1992; Summers and Jacoby, 1977; Taylor and Levine, 1980; Taylor and Summers, 1979; Whelan and Colleran, 1992; Whelan et al., 1995).
The TeR, Phi and PacB phenotypes were linked by insertion mutagenesis to a 6,2 kb DNA fragment that contained seven genes, terZ, -A, -B, -C, -D, -E, and -F ( Whelan et al., 1995). Insertion mutagenesis within terZ, terC, or terD reduced or abolished resistance to phage T5, colicins A, B, and K, and potassium tellurite. The amino acid sequences deduced from terD, terE, and terZ were found to be related. Identity was also observed between the amino acid sequences of terA and terF; in addition, highly related amino acid domains were noted among various subsets of the five latter putative polypeptides (Whelan et al., 1995). The mutagenesis system indicated that terD is essential for expression of the three phenotypes, terZ is necessary for colicin and tellurite resistance, and terC is required for tellurite resistance. Components of the gene cluster encoding TeR, Phi, and PacB cloned from plasmid R478 are lethal to E. coli cells and bring about filamentous morphology. A polypeptide of 17 kDa, specified by the terW gene, was mapped to a 1,95 kb BamHI fragment, which directly protected cells from lethaly. Complementation experiments between pKFW4A and pDT2665 suggested that terW (carried on pDT2665) could not reverse filamentation. Plasmid pKFW4A does not mediate the PacB phenotype, and the upstream region of terZ is absent. It may be that TerW interacts with this upstream region. Sequence analysis did not reveal any coding region within the upstream region (0,7 kb SalI-BamHI region), suggesting that protection is more likely to be related to transcription of this cluster and that the upstream region may constitute a TerW binding site (Whelan et. al, 1996).
Two other newly identified genes, terX and terY, were located downstream of terW and were not required for maintenance of the resistance cluster genes. The function of TerX and TerY are DNA-binding proteins from their predicted amino acid sequences. The degrees of sequence similarity among TerX and three polypeptides (TerZ, TerD, and TerE) encoded by the TeR, Phi, and PacB resistance cluster suggest that they have a relationship, either at the functional or evolutionary level. The highly conserved motif of 13 residues specified by the amino acid GDN(R/L)TG(E/A)GDGDDE is present in the most of these genes (Whelan et al., 1996; Vavrova et al., 2006).
2. 3. 1. 5. Virulence plasmid pLVPK of Klebsiella pneumoniae CG43
pLVPK is a 219-kb virulence plasmid harbored in a bacteremic isolate of Klebsiella pneumoniae (Chen et al., 2004). A gene cluster encoding E. coli terZABCDE homolog was also identified. The terZABCDE has been shown previously to be a part of a PAI, which also contains integrase, prophage, and urease genes in E. coli EDL933. This gene cluster also provides the resistance to bacteriophage infection as well as resistance to pore-forming colicins. Although terBCDE are sufficient for the tellurite resistance property, the functions of each of these genes are unknown. The 14,7 kb region containing terZABCDE genes and 12 putative ORFs of pLVPK are comparable to the ter genes-containing region in the E. coli O157 genome. The homology is interrupted downstream of the terZABCDE region by an
E. coli pTE53 tellurite resistance terF homolog and IS903 gene. A recent study suggests that the TeR-containing pathogenecity island in enterohemorrhagic E. coli isolates was acquired from plasmid. With considerable degree of sequence homology (75-98 % amino acid sequence similarity respectively with that of the E. coli O157 terZABCDE), the ter genes of the pLVPK are likely horizontally acquired. It has been speculated that the ter system most likely plays other functional roles such as protection against host defenses so as to be stably maintained in the bacterium (Taylor et al., 2002).
2. 3. 1. 6. Clinical strain of E. coli KL53
The tellurite-resistant strain KL53 was found during the testing of a group of clinical isolates for antibiotic and heavy metal ion resistance (Burian et al., 1990). Strain KL53 harbors three large plasmids, two of which are conjugable. The resistance determinants are evidently located separately, on two distinct plasmids. Conjugal transfer of the third and smallest plasmid was not found; therefore this plasmid remains cryptic. DNA hybridization results suggested that the tellurite resistance determinant of pTE53 is a novel TeR determinant. The strains formed typical black colonies on solid LB medium with tellurite. The strain KL53 has MIC to tellurite ions on solid media 150μg/ml. A gradual increase in tellurite concentration resulted in increased resistance with MIC value of 1500µg/ml. Recombinant TeR clones were obtained only as a result of in vitro cloning into medium-copy-number vector pACYC184 (Burian et al., 1998) and by in vivo cloning (Tu et al., 2001).
The in vitro clone of pTE53 [Acc.N. AJ238043.1] named pLK18 contained the minimal part of the operon, genes terB, C, D, E, and terF, expressing tellurite resistance and genes terB, C, D, E were shown essential by Tn1737Km mediated gene disruption for the resistance (Kormutakova et al., 2000). For in vivo cloning a low copy mini-Mu derivative pPR46 was used resulting in pNT3B plasmid [Acc. N. AJ888883].
The in vitro cloning of the large conjugative plasmid pTE53 from E. coli strain KL53 resulted in obtaining of 5 kb fragment with fully functional tellurite resistance. The 5 ORFs encoding for 5 genes, which have been named terB to terF (Kormutakova et al., 2000).
The BLAST search of 5250 bp fragment from pTE53 against nucleotide database showed a significant homology with 3 known operons encoding for tellurite resistance genes. The first is carried by the plasmid pR478 from Serratia marcescens (EMBL ID P47TERZ, Whelan, 1997). The next is from Alcaligenes sp. plasmid pMJ606 (EMBL ID PLMTEAD, Jobling and Ritchie, 1988). The third is a chromosomal tellurite resistance operon from Proteus mirabilis (EMBL ID AF168355, Toptchieva et al., 1999). Only four of the seven genes are present in all the four operons. They are terB, terC, terD and terE, their role in tellurite resistance mechanism is probably essential but not known yet.
The clone pLK18 was subjected to the transposition with Tn1737 Km to disrupt determinant of the tellurite resistance. Disruption of terB, terC, terD and terE genes resulted in abolished of tellurite resistance, and insertion in terF gene showed that this gene is not essential for the conservation of tellurite resistance. The product of terF gene is not part of the tellurite resistance pathway (Kormutakova et al., 2000).
The in vivo cloning system based on mini-Mu derivatives was used for cloning of tellurite resistance genes. The Mu-phage and its derivatives combine the properties of both the temperate phage and transposable element, which are wide range utilized in in vivo cloning (Tu et al., 2001).
For in vivo cloning a low copy mini-Mu derivative pPR46 was used resulting in pNT3B plasmid, from which a significant part of ter operon has been sequenced and was also analyzed together with previous sequence of the operon, found on pTE53 E. coli plasmid (Vavrova, 2006).
Recently obtained sequence of in vivo cloning pNT3B that is a subclone of pTE53 plasmid revealed that the large conjugative plasmid possess the full operon similar as was described in Serratia marcescens as well as a part of chromosomal ter operon found in E. coli O157:H7.
2. 3. 2. Chromosomal TeRdeterminants
2. 3. 2. 1. Determinant tmp from Pseudomonas syringae pathovar pisi
Several different chromosomal TeR systems have been characterized until now. A TeR determinant from the pea blight pathogen Pseudomonas syringae pathovar pisi was isolated by a shotgun strategy involving TeR screening in E. coli. A single protein of 24, 5 kDa, encoded by the tmp gene, resembles a thiapurine methyltransferase (Cournoyer, 1998).
Cournoyer et al. have recently proposed that Tmp could be involved in the volatilization of tellurite into dimethyltelluride.
2. 3. 2. 2. Determinant tehA and tehB
The E. coli K12 genome contains two TeR genes (tehA and tehB) within an operon at 32,3 min, which encodes a minimal inhibitory concentration (MIC) of 128 µg/ml to K2TeO3 , when expressed on a multicopy plasmid or behind a strong promoter (Taylor et al., 1994). Sequence analysis shows that the TehB protein of E. coli K12 contains all three characteristic motifs identified in SAM-dependent N-methyltrasferases, including the SAM-binding motif (Fu et al. ,1996). E. coli mutations that could interfere with synthesis of SAM or with related biochemical pathways reduce, but do not eliminate TeR. TehA has ten membrane-spanning regions and there is evidence for efflux transporter activity (Turner et al., 1997), but apparently not for tellurite (Turner et al., 1995). Mutations in the cysteine, glutathione or thioredoxin biosynthetic pathways decrease TeR in the E. coli K12 tehAB system. Lower concentrations of cysteine are likely to decrease important thiol metabolites, such as glutathione and coenzyme A, and could consequently reduce TeR.
2. 3. 2. 3. Determinant from Rhodobacter sphaeroides trgAB
In Rhodobacter sphaeroides, a bacterium that can perform photosynthesis under anaerobic conditions as well as to fix nitrogen and carbone dioxide, two sets of unrelated determinants are involved. The trgAB (TeR genes) encode membrane associated proteins that confer TeR on a related bacterium Paracoccus denitrificans. Located downstream from trgAB is cysK, which encodes cysteine synthase and is responsible for the terminal step in cysteine biosynthesis in R. sphaeroides (O'Gara et al., 1997). Disruption of cysK results in decreased TeR in R. sphaeroides. The second TeR locus (telA) in R. sphaeroides encodes a deduced protein (TelA) of 396 amino acids that shows 65% similarity to TelA (KlaB), encoded on the IncPα plasmid.
2. 3. 2. 4. Escherichia coliO157:H7
Escherichia coli O157:H7 is of major interest in clinical practice, food safety and evolutionary biology. E. coli O157:H7 and other Shiga toxin (Stx)-producing E. coli (STEC)strains cause diarrhea, hemorrhagic colitis, and the hemolytic uremic syndrome (Tarr and Bilge, 1998).
Using an allele-specific probe for the gene (uidA) encoding β-glucuronidase and a multilocus enzyme electrophoresis technique, a stepwise model for evolution of E. coli O157:H7 was proposed by Feng (1995) and Whittman (1998) that the emergence of E. coli O157:H7 with phenotypes (SOR-, GUD-) or (SOR-, GUD+) is based on the discrete evolutionary events from an EPEC-like ancestor resembling most present-day commensal E. coli in terms of the ability to express β-glucuronidase (GUD+) and ferment sorbitol (SOR+) (Park et al., 1999).
2.3.2.4.1. Pathogenicity of E. coli O157:H7
The virulence of E. coli O157:H7 is known to be attributed to several factors. These factors are one or more Shiga toxins (Karmali et al., 1985), hemolysin (Beutin et al., 1989), the adhesion intimin (Tzipori et al., 1986, Donnenberg et al., 1993), Esp E (Calderwood et al., 1996), secreted proteins encoded in Type III secretion system (Kenny et al., 1996; Haigh et al., 1995; Lai et al., 1997), and O157 lipopolysaccharide O-side antigen (Bilge et al., 1996).
A. Shiga Toxins
The Shiga toxins (Stxs) are a family of bacterial cytotoxins produced by Shigella dysenteriae type 1 and STEC. These toxins were formerly called Shiga-like toxins (SLTs) or sometimes verocytotoxins (VTs). The role of Stx as a virulence factor is clearly demonstrated by several findings. First, HUS is caused by only Stx-producing bacteria, including STEC and only S. dysenteriae type 1. EPEC strains, which are most similar to EHEC in terms of virulence except for the production of Stxs, do not cause HUS. Second, streptomycin- treated mice that are fed certain STEC strains develop renal tubular necrosis, resulting in the death of the animal (Wadolkowski et al., 1990). Third, rabbits inoculated with a rabbit EPEC strain transduced with the Stx 1-converting bacteriophage develop more serious histological lesions, similar to the hemorrhagic colitis caused by EHEC (Sjogren et al., 1994).
A strain of Shiga toxins-producing E. coli may produce Stx1 or Stx2 (or a variant such as Stx2c, or Stx2d or Stx2e), or both of these toxins.
These were obtained HGT.
B. pO157
A 90-kb plasmid that is carried by almost all E. coli O157:H7 strains, was designated as pO157 by Toth et al. in 1990. EHEC-hemolysin is a highly active repeats-in-toxin (RTX) that belongs to the family of pore-forming proteins (Schmidt et al., 1995). Four open reading frames contained on pO157, which were highly related to both the genes of the
E. coli alpha-hemolysin operon and EHEC-hly operons, were termed EHEC-hlyC, EHEC-hlyA, EHEC-hlyB, andEHEC-hlyD (Brunder et al., 1997).
C. LEE
The bacterial genes involved in producing the A/E histopathology are located on a
35-kbp segment of chromosomal DNA termed the locus of enterocyte effacement (lee region) in EPEC, E. coli O157:H7, and many other serotypes of STEC (McDaniel et al., 1995, Perna et al., 1998, Jerse et al., 1990). Among the multiple virulence factors encoded in the LEE, the eae gene encoding intimin and the tir gene encoding a translocated receptor for intimin are located in the middle region. Downstream of eae there are esp genes, encoding secreted proteins responsible for inducing the epithelial cell signal transduction events leading to the A/E lesion. In the upstream region of eae and tir, there are esc and sep genes, encoding a type III secretion system involved in extracellular secretion of the protein encoded by esp genes (Perna et al., 1998, Jerse et al., 1990).
D. Intimin
Intimin, encoded by eae, is produced by A/E pathogens such as EPEC and E. coli O157:H7 (Rosenshine et al., 1996). Intimin, a 94 to 97 kDa outer membrane protein, mediates A/E lesions in epithelial cells.
The role of intimin is futher proven by an experiment, in which the intimate adherence to epithelial cells disappeared when EPEC strains mutated in the eae gene encoding intimin (Frankel et al., 1996). Although the role of intimin as an adhesion is clear, the interaction of intimin with eukaryotic cells for the formation of A/E lesion has yet to be fully elucidated. In EPEC, intimin expressed on the bacterial surface binds to Tir and perhaps to β1 integrins.
E. Tir
Tir (translocated intimin receptor) is produced in the bacterial cell as a 78 kDa unphosphorylated protein. Subsequently, the type III secretion system translocates the protein into the host cell, where the protein is then phosphorylated (at least in EPEC and STEC O26:H- strains), thereby increasing in size to 90 kDa (Calderwood et al., 1996).
After binding with intimin, Tir triggers additional host-signaling events and actin nucleation for the formation of the A/E lesion.
F. Secreted Proteins by a Type III Secretion System
Secreted bacterial proteins have long been known to play central roles in bacterial-host interactions. Javis et al (1995) reported that E. coli O157:H7 secreted some immunoreactive proteins, including Esps via a Type III secretion system, which has been found in many Gram-negative bacteria, causing disease in animals and plants. Secreted proteins by Gram-negative bacteria must pass through two membranes, the inner membrane surrounding the cytoplasm and the outer envelope enclosing the periplasm. The type III secretion system is responsible for transporting the proteins directly from the cytoplasm to the cell surface, while the general secretory pathway transports proteins to the periplasm (Lee, 1997; Perry et al., 1998).
G. Lipopolysaccharide
E. coli O157:H7 has the typical lipopolysaccharide (LPS) surface structure of Gram-negative bacteria. LPS is known historically as the ‘O' or somatic antigen. The LPS is composed of lipid A and O polysaccharide. The O polysaccharide of E. coli O157:H7 appears to play a role in the adherence of the serotype to host epithelial cells. E. coli O157:H7 mutants deficient in O-antigen were more adherent to the host cells than was its E. coli O157:H7 wild type, suggesting that the O side chains of E. coli O157:H7 lipopolysaccharide interfere with the adherence of E. coli O157:H7 to host epithelial cells (Bilge et al., 1996; Cohen and Giannella, 1992).
H. TAI islands
A significant part of the ter operon has been found on the E. coli O157:H7 chromosome as a part of a large genomic island called Tellurite and Adhesin Island (TAI), bearing also other pathogenic determinants as adhesion, urease and phage protection proteins (Tarr et al., 2000). Horizontal gene transfer of such a pathogenicity island (PAI) represents an important tool of microbial evolution (Hacker and Carniel, 2001). E. coli O157:H7 has acquired by this way two important PAIs so called LEE locus (locus of enterocyte effacement, Park et al.,1999) and TAI (tellurite teristance- and adherence conferring island, Tarr et al., 2000). Significant genomic variability of this island bearing, the ter genes determinant within several different E. coli O157:H7 isolates originated worldwide, has been detected. Some strains were shown to bear two copies of TAI, some one and there are some without this locus, showing that these pathogenic E. coli strains have the same evolutionary origin
(E. coli encompassing O antigen), but the emergent pathogens have arisen many times during evolution (Taylor et al., 2002, Wick et al., 2005).
The genome sequences of two strains of enterohemorrhagic E. coli O157:H7 were recently completed (Hayashi et al., 2001; Perna et al., 2001). Comparisons between laboratory E. coli strain K-12 MG1655 (Blattner et al., 1997) and EDL933, an E. coli O157:H7 strain, demonstrated that they have a complex relationship since their divergence 4,5 million years ago. Homology between the two is interrupted by the presence of hundreds of islands of inserted DNA. K islands are DNA segments present in MG1655 (K-12) but absent from EDL933 (O157:H7), whereas O islands (OI) are present only in EDL933. In E. coli EDL933 two OI, designated OI 43 and OI 48, contain integrase, phage, tellurite resistance, and urease genes (Perna et al., 2001). One of these islands had been identified previously by cosmid cloning in E. coli O157:H7 strains and in more distantly related serotypes. Tarr and coworkers termed this region the tellurite resistance an adherence-conferring island (TAI). Since the island was found in E. coli O157:H7 but was absent from nontoxigenic E. coli O55:H7 and toxigenic E. coli O157:H- (flagellum-), pathogenic E. coli O157:H7 it is believed to have acquired the TAI only recently (Tarr et al., 2000).
2. 3. 2. 5. An inducible tellurite-resistance operon in Proteus mirabilis
The ter operon is also present in the chromosome of all recent clinical isolates of Proteus spp. as well as in some isolates of Morganella and Providencia.Futhermore, the ter operon was not located on large IncJ or IncT plasmids found in some strains of P. vulgaris and Providencia spp. (Boltner et al., 2002; Murata et al., 2002). A phylogenetic analysis of TeR proteins suggests a rather recent dissemination among the enteric bacteria, most likely accelerated by transmissible plasmids (Taylor et al., 2002).This conclusion is also supported by the high level of nucleotide identity (70-80%) among terBCDE genes from the IncHI plasmids and genomic sequences from E. coli O157:H7, Y. pestis and P. mirabilis.
The P. mirabilis ter operon differed from both its plasmid and E. coli O157-borne counterparts, including (i) a larger terA product (44 aa larger) which may have resulted from a fusion of terD sequences with terA; (ii) the absence of terF, which is also absent from plasmids pMJ606 and pTE53 (Kormutakova et al., 2000; Taylor et al., 2002); and (iii) inducibility by tellurite and to a lesser extent by agents associated with oxidative stress. Alignment of the intervening DNA sequences between orf3 and terZ of P. mirabilis (~390 bp region containing putative regulatory motifs) with those from the published sequences of other ter operons revealed no nucleotide sequence identity, while that of the flanking genes was ~80%, suggesting that the regulatory region may be unique to the P. mirabilis ter operon. The ter operon of P. mirabilis was inducible by tellurite has since led to evaluation of TeR in E. coli O157:H7 strains. There is considerable variability in ter loci among clinical isolates with some strains containing duplications of ter genes while others show variability in the complement of ter genes. In some E. coli O157:H7 strains, prior exposure to tellurite led to increased resistance of strains when plated on K2TeO3 -containing medium and analysis of the ter genes by RT-PCR showed that mRNA of terBC and F genes was increased while that for terZD and E was not induced (Taylor et al., 2002). Interestingly, the study found that terA was not expressed, suggesting that a potential regulatory region might exist in this sequence.
2. 5. The mechanism of resistance to tellurite and other heavy metals
To survive in the presence of high levels of various toxic metals or antibiotics, bacteria have evolved ways of preventing the internal concentration of the noxious compound from rising to lethal levels. These mechanisms include (a) entry exclusion, for example, chromosomal resistance to cadmium in Bacillus subtilis (Laddaga et al., 1985); (b) extrusion, e.g., ATP-dependent efflux systems for arsenic and cadmium resistance (Mobley and Rosen, 1982; Tynecka et al., 1981); (c) detoxification, e.g., reduction of Hg 2+to volatile Hg0 in mercury resistance; (d) sequestration, e.g., cadmium resistance due to the production of cadmium-binding proteins (Higham et al., 1984; Perry and Silver, 1982); and (e) target modification, e.g., resistance to the macrolide-lincosamide-streptogramin B (MLS) family of antibiotics (Lai et al., 1973). This last mechanism has not yet been encountered in microbial resistance to metals, probably because of the more generalized nature of the targets of metal toxicity (Summers and Barkay, 1989).
There is preliminary experimental evidence that TeR can result from several of these possible mechanisms. Most isolates of E. coli are highly sensitive to tellurite, having a MIC of 0, 25 to µg/ml (Taylor et al., 1988). Spontaneous mutants of E. coli resistant to low levels of tellurite (~10µg/ml) well as to arsenate were obtained. These mutants were found to be defective in phosphate transport and were unable to grow on medium containing low levels of phosphate. Transport of phosphate was competitively inhibited by tellurite. Susceptibility to tellurite could be restored by a plasmid carrying the phoB region, which is involved in phosphate regulation. These results indicate that E. coli takes up tellurite by phosphate transport system and that reduced uptake results in a low level of resistance (Tomas and Kay, 1986). E. coli carrying the IncHII Te determinant (on plasmid pHH1508a or its derivative pDT1364) contain much more black metallic tellurium than those carrying the IncP TeR genes (on plasmids RP4 and pDT1558) (Taylor et al., 1988). Furthermore, the amount of black metallic tellurium was increased in cells carrying the TeR determinants on the high-copy-number plasmid vector pUC8. These observations suggested that the IncHII and IncP plasmids perhaps encoded different mechanisms of tellurite resistance and that reduction to metallic tellurium may be involved. IncHII TeR determinant on pDT1364 encodes a tellurite-detoxification system, whereas the TeR determinant on pDT1558 probably encodes a different mechanism of resistance such as efflux or decreased uptake.
The rate of reduction of tellurite to black metallic tellurium by TeR and TeS bacteria has been estimated using a Klett colorimeter (Walter, 1990). The preliminary findings suggested that the mechanism of tellurite resistance mediated by the IncHII plasmid was via the tellurite reduction. The sensitivity of E. coli bacterial cells to tellurite and selenite was enhanced by the presence of L-methionine (Scala and Williams, 1962, 1963). Since tellurite and selenite are chemically similar to sulfate, Scala and Williams proposed that tellurite and selenite could be reduced, and thus detoxified, by the sulfate reduction pathway. The presence of an exogenous reduced sulfur source such as methionine would repress this pathway, thus decreasing the rate of detoxification of selenite and tellurite, and therefore increasing sensitivity to these anions.
