The Pix pilus adhesin of the uropathogenic Escherichia coli strain X2194 (O2 : K- : H6) is related to Pap pili but exhibits a truncated regulatory region

Andreas Lügering, Inga Benz, Sabine Knochenhauer, Michael Ruffing and M. Alexander Schmidt

Institut für Infektiologie, Zentrum für Molekularbiologie der Entzündung (ZMBE), Von Esmarch Strasse 56, Westfälische Wilhelms-Universität Münster, D-48149 Münster, Germany

Correspondence
M. Alexander Schmidt
infekt{at}uni-muenster.de


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Adhesins provide a major advantage for uropathogenic Escherichia coli in establishing urinary tract infections (UTIs). A novel gene cluster responsible for the expression of a filamentous adhesin of the pyelonephritogenic E. coli strain X2194 has been identified, molecularly cloned, and characterized. The ‘pix operon’ contains eight open reading frames which exhibit significant sequence homology to corresponding genes in the pap operon encoding P pili, the prevalent E. coli adhesins in non-obstructive acute pyelonephritis in humans. Although a pixB gene corresponding to the PapB regulator was identified, a papI homologue could not be found in the pix operon. Instead, a fragment of the R6 gene of the highly uropathogenic E. coli strain CFT073 was identified upstream of pixB. The R6 gene is located in a pathogenicity island containing several pilus-encoding sequences and shows homology to a transposase of Chelatobacter heintzii. In a pixA–lacZ fusion system it was demonstrated that the expression of Pix pili is regulated at the transcriptional level by the R6 gene sequence. A significantly reduced transcription was observed by deleting this fragment and by lowering the growth temperature from 37 to 26 °C. In contrast to other filamentous adhesin systems, Pix pili are mainly expressed in the steady state growth phase and were not repressed by the addition of glucose.


Abbreviations: PAI, pathogenicity island; UTI, urinary tract infection

The EMBL accession number for the sequences of the genes identified in the pix operon is AJ307043.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Urinary tract infections (UTIs) are considered one of the most common infectious diseases in humans. Particularly in non-hospitalized patients, more than 80 % of non-obstructive UTIs are caused by uropathogenic Escherichia coli strains. Different virulence factors have been identified as contributing to their pathogenicity, such as the secreted haemolysin, and specific capsular as well as O-serotypes, for example (Foxman et al., 1995; Ikameimo et al., 1993; Korhonen et al., 1985). The expression of specific adhesins which enable bacteria to attach to and to successfully colonize host tissues in the urinary tract is of particular importance (Wullt et al., 2000). Especially well-studied bacterial adhesin systems are associated with the expression of filamentous surface appendages, called pili or fimbriae. The majority of clinical isolates in non-obstructive acute pyelonephritis express P pili (Pap) which are critical virulence factors (Roberts et al., 1994). P pili bind to glycolipids of the globoseries where they recognize the Gal–Gal disaccharide as the principal receptor moiety (Källenius et al., 1980). Adhesion via P pili already mediates the induction of gene expression in target cells (Zhang & Normark, 1996). Fewer UTI strains produce S pili (sfa), which, however, are more frequently associated with newborn meningitis (Parkkinen et al., 1988).

The genetic organization, the regulation of expression, and the assembly process leading to functional filaments of P pili have been addressed by numerous investigators and have been elucidated in great detail. P pili are encoded by the 11 genes of the pap operon (Hultgren et al., 1993). Expression is regulated by a region directly upstream of the gene of the major pilus subunit PapA encompassing the papI and papB genes. Transcription of the polycistronic mRNA is additionally influenced by global regulators such as the leucine-responsive protein (Lrp), the catabolite activator protein (CAP) and H-NS (Nou et al., 1995; Weyand et al., 2001; White-Ziegler et al., 1998). The different pilus systems show remarkable similarities in the coding sequences as well as in the protein structures of their different subunits (Smyth et al., 1996).

The transcriptional regulation of pili expression can be modulated by environmental factors, such as temperature, carbon source and growth rate, for example (Uhlin, 1994). In several pilus systems, the regulatory mechanisms involve changes in primary DNA sequence thereby inducing a phase variation due to an ‘on’ or ‘off’ state at the transcriptional level (Jonsson et al., 1992; Swansson & Koomey, 1989; Willems et al., 1990). In the pap operon, differential methylation patterns of GATC sites located in the promoter region are responsible for inhibiting and inducing pilus transcription. Two pap operon genes (papB and papI) are involved in this process, acting in concert with global regulators such as Lrp (Nou et al., 1995; Weyand et al., 2001).

Here we report on the molecular cloning and characterization of a novel filamentous adhesin derived from the clinical UTI isolate X2194 (O2 : K- : H6). This adhesin has been designated ‘Pix’ for pilus involved in E. coli X2194 adhesion. The Pix adhesin exhibits genetic and structural similarity to the well-characterized P pilus; however, it lacks a papI equivalent. Instead, a fragment of the R6 gene which had been identified recently in the pathogenicity island (PAI) of the UTI E. coli CFT073 (Guyer et al., 1998; Kao et al., 1997) was detected upstream of pixB. Initial analysis of transcriptional fusions of this region indicated that the expression of Pix pili is influenced by the R6 gene fragment. Binding and inhibition studies indicated that Pix pili apparently recognize surface antigens which are distinct from the known UTI E. coli adhesin receptors.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Plasmids and bacterial strains.
Plasmids used in this study are listed in Table 1. All strains were routinely stored at -80 °C in Standard I Medium (Merck) with 15 % glycerol. E. coli X2194 was recovered from a female UTI patient and was classified as an O2 : K? : H6 strain by the International Escherichia and Klebsiella Centre (Copenhagen). The strain exhibited mannose-resistant haemagglutination of human erythrocytes. E. coli laboratory strains HB101 (recA13 F- {lambda}- supE44 lacY1, courtesy of I. van Die, Amsterdam), DH5{alpha} [supE44 {Delta}lacU169 (flacZ{Delta}M15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1], CMK603 (thr- leu- supE recBC lacY-, courtesy of E. Jacob, Ludwigshafen) and JM109 [recA1 endA1 gyrA96 thi hsdR17 supE44 relA1 {lambda}- {Delta}(lac proAB) (F'traD36 proAB lacIqZ{Delta}M15)] were used.


View this table:
[in this window]
[in a new window]
 
Table 1. Plasmids used in this study

 
Media and growth conditions.
Bacteria were grown in Luria–Bertani (LB) medium at 37 °C. For maintenance of plasmids, 100 µg ampicillin ml-1 (Sigma) was added. Test cultures for lacZ fusion assays were performed in M9CA medium containing 0·1 % glucose or glycerol (Sambrook et al., 1989). Cultures were grown at 37 °C (except indicated otherwise). HeLa cells (ATCC CCL 2; human cervical epitheloid carcinoma) were routinely grown at 37 °C in a 10 % CO2 atmosphere in Dulbecco's minimal essential medium (DMEM) supplemented with 10 % (v/v) fetal calf serum (FCS), containing 1 mM glutamine, penicillin (100 U ml-1) and streptomycin (100 µg ml-1).

Haemagglutination assay and inhibition by carbohydrates.
A single colony was grown in 1 ml LB medium, and the overnight culture was pelleted and resuspended in 0·1 % BSA/PBS containing 1 % mannose to inhibit haemagglutination mediated by type 1 pili expression. Fifty microlitres of the bacterial suspension was incubated with 50 µl of a 1 % suspension of human erythrocytes (HRBCs in PBS) for 1 h at 20 °C. HRBCs of blood group A and O were used. For inhibition the following compounds were added (16 mM final concentration): D-arabinose, D-fructose, D-galactose, D-glucose, lactose, D-mannose, D-melibiose, L-rhamnose, D-ribose, D-xylose, N-acetylgalactosamine, N-acetylglucosamine, N-acetylneuraminosyl-D-lactose, sucrose, asialofetuin, fetuin, transferrin, Tamm–Horsfall protein, glycophorin AM and L-serine. Agglutination of Gal-{beta}(1,4)Gal-coated P-latex beads (Chembiomed) was tested in PBS.

Adhesion assay.
Bacterial adherence to HeLa cells was monitored essentially as described by Cravioto et al. (1979) with modifications. For each assay, 108 bacteria grown overnight at 37 °C with aeration in LB medium were incubated for 5 min in 1 ml PBS containing 0·5 % D-mannose. The bacterial suspension was added to HeLa cell monolayers, which had been cultured overnight on cover-slips, just before reaching confluency. After 1 h incubation at 37 °C, the cells were washed extensively with PBS to remove non-adherent bacteria. The cells were fixed in 70 % methanol, stained with Giemsa (10 % solution in water) for better contrast, and evaluated by light microscopy.

Recombinant DNA techniques.
All DNA manipulations were performed by standard genetic and molecular biology techniques (Sambrook et al., 1989). Plasmid DNA was purified using a Qiagen kit. Restriction and DNA-modifying enzymes were obtained from New England Biolabs and used according to the manufacturer's instructions. Oligonucleotides were synthesized in a Beckman oligonucleotide synthesizer. The sequences of oligonucleotide primers are summarized in Table 2. E. coli strains were transformed by the CaCl2 method as described by Hanahan (1983).


View this table:
[in this window]
[in a new window]
 
Table 2. Sequences of oligonucleotide primers used in this study

 
Molecular cloning of the X2194 pilus adhesin genes (pix operon).
For cloning of the adhesion-mediating genes, 100 µg chromosomal DNA of E. coli X2194 was partially digested with Sau3A and the fragments were separated by gel electrophoresis (0·8 % agarose). Fragments ranging from 6 to 12 kb were eluted, ligated into the BamHI-restricted pBR322 vector, and transformed into competent HB101 cells. Adhesin-expressing clones were identified by haemagglutination of human erythrocytes and by adhesion to HeLa cells as described above.

Cloning of the 5'-regulatory region.
Cloning of sequences upstream of the major pilin gene pixA was performed by a modified chromosome walking method as described by Dominguez & Lopez-Larrea (1994). In brief, a pixA-specific primer (SP1) was used in combination with a non-specific primer (USP1) for PCR (20 pmol of each primer, 2 U Taq polymerase, 2·5 mM MgCl2) in standard PCR buffer (Roche Biochemicals) in a Biometra Trio PCR block. After initial denaturation at 95 °C for 3 min, five cycles were performed with the following settings: 95 °C for 60 s, 25 °C for 120 s, 72 °C for 120 s with a temperature ramping of 0·5 °C s-1 from 25 to 72 °C. An additional 30 cycles were carried out at 95 °C for 60 s, 55 °C for 90 s, and 72 °C for 90 s. Amplified fragments were evaluated by gel electrophoresis (1·0 % agarose). A 1·3 kb fragment could be identified by Southern blotting using a DIG-labelled pixA-specific oligonucleotide probe (Southern, 1975). This fragment was purified from a 1·0 % agarose gel by using the Qiaex gel extraction kit (Qiagen) according to the manufacturer's instructions, reamplified by PCR and cloned into the pGEM-T vector, resulting in pAL1. A second chromosomal walking step was carried out as described above using a different pair of primers (SP2 and USP2). A positive fragment (950 bp) was identified by digestion with BglII (removal of 252 bp from the 3'-end) and cloned into the pGEM-T vector (pAL4). Subsequently, the complete fragment was reamplified by PCR using the 5'-primer pixIBA and the 3'-primer pixABI (Table 2). Double-stranded plasmid DNA sequencing of the pix operon was carried out by the dideoxy-chain termination procedure with a T7 sequencing kit (Pharmacia). The software package HUSAR (Heidelberg Unix Sequence Analysis Resources, version 4.0) of the DKFZ Heidelberg was used for sequence analysis.

The sequences of the genes identified in the pix operon have been deposited in the EMBL database (accession no. AJ307043).

Construction of lacZ fusions for the evaluation of transcriptional activity.
For the evaluation and quantification of the putative transcriptional activity of the 5'-flanking sequence of pixB, two fragments were cloned into the promoter test vector pCB192, which harbours a promoterless lacZ gene (Schneider & Beck, 1986).

After digestion of pAL1 with SphI and SalI, the insert was cloned into the vector pBS(+) digested with the same enzymes to generate pAL2. Subsequently, the fragment now flanked by HindIII and PspAI sites was directionally cloned as a 1·3 kb insert into pCB192 (pAL3), generating a pixAlacZ fusion for the investigation of the putative promoter activity residing in the pixBA region. Positive clones were identified by a low activity of lacZ and were additionally characterized by restriction analysis.

For construction of pAL6, chromosomal DNA of E. coli X2194 was amplified using the two primers pixABI and pixIBA (Table 2) in a standard PCR reaction (30 cycles of 95 °C for 60 s, 60 °C for 90 s, 72 °C for 60 s). These primers are flanking the 1923 bp fragment of the pix operon, including the R6 gene fragment, and additionally contain a BamHI restriction site at their 5'-end to allow cloning into pCB192. The orientation of the insert was checked by digestion with EcoRI/SphI. Positive clones showed a high activity of lacZ as detected on X-Gal-containing LB-agar plates. {beta}-Galactosidase activity in bacteria was determined as described by Miller (1972).

Influence of growth conditions on pix operon expression (growth phase, temperature, carbon source).
The influence of different culture conditions was investigated by growing the bacteria in M9CA medium containing glucose as carbon source. Cultures were started with the addition of a 1 : 100 dilution of an overnight culture. To investigate the effect of growth phase on Pix pilus expression, samples were taken at different time points and the OD600 and lacZ activity were measured. To analyse the effect of temperature on pixA expression, cultures were grown at 37 and 26 °C, and lacZ activity was monitored at different time points. To assess the influence of different carbon sources, bacteria were grown in M9CA medium containing 0·1 % glucose or 0·1 % glycerol.

Electron microscopy.
Expression of pili in E. coli X2194 and in recombinant clones was evaluated by transmission electron microscopy using uranyl acetate for contrast. Bacteria suspended in PBS were placed on a Formvar- and carbon-coated copper grid. In some experiments, the pili were stabilized by the addition of 1 % BSA/PBS. After applying a 1 % uranyl acetate solution for negative staining, the sample was inspected using a Philips 400 electron microscope set for transmission electron microscopy or in the scanning mode.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Haemagglutination and adherence of E. coli X2194 to HeLa cells
Bacteria causing UTIs express adhesins to resist cleansing mechanisms of the host and to establish themselves on host cells and tissues. These adhesins usually recognize carbohydrate structures linked to glycolipids or glycoproteins. To investigate the adherence properties of the uropathogenic clinical isolate E. coli X2194 (O2 : K- : H6), we perfomed in vitro adherence assays using epitheloid HeLa cells for assays in tissue culture and erythrocytes in haemagglutination assays. E. coli X2194 adhered to human epithelial cells and haemagglutinated human, mouse, duck, dog, cat and rat erythrocytes. Interestingly, strain X2194 did not agglutinate red blood cells obtained from sheep or goat. Inhibition experiments with various carbohydrates (D-arabinose, D-fructose, D-galactose, D-glucose, lactose, D-mannose, D-melibiose, L-rhamnose, D-ribose, D-xylose, N-acetylgalactosamine, N-acetylglucosamine, N-acetylneuraminosyl-D-lactose and sucrose) as well as glycoproteins (asialofetuin, fetuin, transferrin, Tamm–Horsfall protein and glycophorin AM) and L-serine were not successful and indicated that the X2194 adhesin did not belong to the known group of adhesins exhibiting G-, M-, S- or mannose-specificity. As E. coli X2194 also did not agglutinate p-latex beads, the presence of Pap pili could be excluded. This showed that adherence of X2194 was not mediated by the common type-1 or by Pap pili adhesins representing the most frequently encountered filamentous adhesins in UTI E. coli.

Molecular cloning of the E. coli X2194 pilus adhesin
To further characterize the E. coli X2194 adhesin by molecular cloning, chromosomal DNA was partially digested with Sau3A and fragments between 6·0 and 12·0 kb were ligated into a BamHI-restricted pBR322. After transformation of CMK603 mannose-resistant haemagglutination of human erythrocytes was used to screen for the expression of the functional adhesin. Adherence-positive bacteria were found to harbour a recombinant plasmid named pAD1 which carried an 8·0 kb chromosomal DNA insert. Directional cloning of the 8·0 kb insert from pAD1 into the pBSIISK(-) vector carrying a multiple cloning site was achieved by excising the insert with the flanking HindIII and SalI sites. Ligation of the HindIII–SalI fragment (8·6 kb) encompassing the original 8·0 kb insert with some flanking pBR322 vector sequences into the pBSIISK(-) vector generated plasmid pSK12. The 8·0 kb insert also mediated adhesion of E. coli CMK603(pAD1) and of E. coli JM109(pSK12) to HeLa cells. The 8·0 kb insert harboured by pAD1 (also present in pSK12) contained all the genes necessary for the functional expression of the specific pilus adhesin of E. coli strain X2194 (Pix pili).

Pix pilus characterization by electron microscopy
The wild-type E. coli strain X2194 as well as the recombinant strain harbouring plasmid pAD1 were investigated by electron microscopy using negative staining. As demonstrated in Fig. 1, the original isolate (Fig. 1a, b) as well as the recombinant type 1 pili-negative HB101 strain harbouring the pAD1 plasmid (Fig. 1f) expressed numerous filamentous appendages. Inspection of the images revealed the pili to have a diameter of about 10 nm and a maximal length of approximately 3·5 µm. Interestingly, these pili appeared to be not as rigid as pili expressed by other UTI strains, like the Pap pili, for example.



View larger version (136K):
[in this window]
[in a new window]
 
Fig. 1. Visualization of Pix pili by transmission and scanning electron microscopy. (a, b) Pix pili of E. coli strain X2194 (O2 : K- : H6) (a: bar, 0·9 µm; b: bar, 0·06 µm). (c, d) Scanning electron micrographs of JM109(pSK12) expressing recombinant Pix pili adhering to Hep-2 cells (c: bar, 3·8 µm; d: bar, 1 µm ); the arrows in (d) indicate the presence of Pix pili. (e) HB101 without plasmid – no pili can be detected (bar, 0·45 µm). (f) Recombinant HB101 harbouring pAD1 expresses Pix pili (bar, 0·45 µm).

 
Sequence analysis of the putative pix operon and homologies to the pap operon
To further analyse the genes encoding the Pix pili, we determined the nucleotide sequence of the 8·0 kb insert using the pSK12 plasmid. By sequence analysis we identified a gene cluster of seven complete open reading frames (ORFs) which appeared to be organized in an operon structure (Fig. 2a). These ORFs showed significant homology to genes in the pap operon (Hultgren et al., 1993; Källenius et al., 1981) encoding the expression of Pap pili representing the major pilus type found among uropathogenic E. coli strains. The ORFs in the putative pix operon were therefore named according to their corresponding homologues in the Pap pilus system. A summary of the identified ORFs and the degree of their various homologies to their corresponding Pap proteins and to the very recently identified sfp system (Brunder et al., 2001) is given in Table 3. The sequences of the putative pix operon have been deposited in the EMBL database (accession no. AJ307043). Interestingly, no homologues of the Pap pilus PapE and PapK subunits could be found and, furthermore, only a truncated version of the pixB gene was identified at the 5'-end of the insert harboured by the pAD1 plasmid. By analogy to the papI and papB genes located in the 5'-regulatory region of the pap operon, we therefore assumed that part of the corresponding sequence potentially harbouring regulatory genes in the pix operon was still missing in the originally cloned chromosomal 8·0 kb fragment.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 2. (a) Genetic organization of the pix operon and of the upstream region harbouring the AT-rich sequences and the R6 fragment. The pix genes have been named according to the corresponding genes in the pap operon. (b) Construction of pixA–lacZ transcriptional fusions. To quantitatively monitor 5'-region-dependent pixA expression, pixA–lacZ transcriptional fusions were constructed containing either a truncated R6 fragment ({Delta}1–573 bp of R6 fragment) leading to pAL3 or the complete 5'-region (pAL6).

 

View this table:
[in this window]
[in a new window]
 
Table 3. Similarities between the predicted pix ORFs and the deduced amino acid sequences of Sfp and Pap proteins

 
Major pilus subunit PixA
The predicted mature PixA pilin subunit has 69·8 % amino acid similarity to the major pilin subunit of the F13 P-pili (PapAF13). The structural homology of the major pilin subunit PixA to the corresponding PapA was further emphasized by the recognition of the 18 kDa PixA protein in Western blotting experiments by antisera directed against synthetic peptides aa 5–15 and aa 8–22 of the PapA N-terminal amino acid sequence (data not shown). Pix pili were isolated from the bacterial culture supernatant and subjected to SDS-PAGE. N-terminal sequencing of the major pilin subunit (PixA) identified the first 25 aa and supported the significant homology with the PapA amino acid sequence. A comparison of PapA-related N-terminal pilin sequences derived from uropathogenic E. coli isolates is given in Table 4.


View this table:
[in this window]
[in a new window]
 
Table 4. Comparison of N-terminal amino acid sequences of UTI E. coli pilins

Conserved residues are shaded and residues representing the Ic subfamily sequence signature motif of the S1 segment (Girardeau et al., 2000) are indicated in bold.

 
According to the classification of Girardeau et al. (2000), PixA is a member of the Ic subfamily together with E. coli PapA, Serratia marcescens SmfA and Proteus mirabilis MrpA and PmpA. The sequence signature motif in segment S1 characteristic for the Ic subfamily (GxG[KT]V[TS]FxG[TS]V[VI]DAP) (strongly conserved residues in the S1 motif are indicated in bold) is conserved, including the phenylalanine at position 7 and the proline at position 15 (Table 4). The sequence of the S1 segment of PixA is therefore represented as GQG[V]V[N]FKG[T]V[I]DAP (additional substitutions in the S1 segment motif are underlined).

Nucleotide sequence of the region upstream of pixA
The cloned fragment of the pix operon contained seven complete ORFs which exhibited significant homology to the genes of the well-characterized pilin subunits of the pap operon. Additionally, an incomplete coding sequence for a papB homologue, denoted pixB, was identified upstream of pixA. This suggested that an incomplete operon had probably been cloned, which might have been truncated in the 5'-regulatory sequences. To complete the pixB gene and to further characterize the sequences upstream of pixA, a modified PCR technique was employed for ‘chromosome walking’ steps for completing the pix operon sequence. After two walking steps we could additionally clone 1518 bp upstream of pixB. Sequence alignment revealed, besides a now completed ORF for pixB, a potential promoter site for a pixB–pixA transcript followed by a 370 bp long AT-rich region. This AT-rich sequence contained several direct as well as inverted repeats suggesting a potential effector site for transcriptional regulators (data not shown). In contrast to the corresponding region in the pap operon, the sequence contained no apparent sites for DNA methylation (GATC). Interestingly, a putative coding sequence for an ORF homologous to papI could also not be detected.

Furthermore, the fragment upstream of the AT-rich region showed significant sequence identities to the first part of the R6 gene which had been identified in the PAI of the highly virulent uropathogenic E. coli strain CFT073. The R6 gene of E. coli CFT073 has recently been reported to express significant homology to a transposase gene of Chelatobacter heintzii (Xu et al., 1997). It was interesting to see whether the R6 gene fragment represented just the remnant of a transposable element or if this sequence might be involved in the regulation of Pix pilus expression. Expression of a truncated R6 protein is not possible because the insertion of an additional nucleotide leads to a frameshift mutation. Homologies to ORFs B–D detected in the upstream sequence could not be identified.

The R6 fragment participates in pixA transcription
To evaluate the potential influence of the upstream sequence of pixB including the R6 fragment on expression, two different pixA–lacZ fusions were constructed (Fig. 2b) and transformed into E. coli DH5{alpha} to assess the specific influence of the R6 homology region. In addition to the potential regulator pixB and the AT-rich region, the construct pAL3 contained the sequence downstream of nucleotide 706 and consequently only a truncated R6 fragment. The construct pAL6 harboured the complete sequence for the R6 fragment of E. coli strain X2194. As an indicator for the specific relative influence of the 5'-regions on expression, the {beta}-galactosidase activities were compared in strains DH5{alpha}(pAL3) and DH5{alpha}(pAL6) grown in M9CA medium containing glucose as sole carbon source. The {beta}-galactosidase activities show that in the absence of the R6 fragment in pAL3, pixA transcription is reduced about 40-fold in comparison to pAL6 containing the R6 fragment (data not shown), suggesting an essential role of the R6 sequence for pix transcription.

Monitoring growth-phase-specific and temperature-dependent PixA expression
To assess whether the effect of the 5'-sequence on the expression of Pix pili is influenced by environmental parameters, the pAL3 and pAL6 plasmids encoding a PixA–LacZ fusion protein with either a truncated (pAL3) or a complete (pAL6) R6-homology region (see Fig. 2B) were transformed into DH5{alpha}. Stationary LB cultures of DH5{alpha}(pAL3) or DH5{alpha}(pAL6) were diluted 1 : 100 in M9CA medium containing 0·2 % glucose as carbon source. With DH5{alpha}(pCB192) serving as vector control, {beta}-galactosidase activity was determined as an indicator of PixA pilin synthesis and bacterial growth was monitored by OD600 (Fig. 3). During the exponential growth phase of DH5{alpha}(pCB192), the rate of {beta}-Gal production appeared to be limited as the enzyme activity increased only twofold. By the late exponential growth phase, transcription of the gene encoding the major pilus subunit PixA was markedly induced in DH5{alpha}(pAL6) as indicated by a sevenfold increase in activity of LacZ compared to the basal level. In contrast, no induction of pixA transcription could be detected in DH5{alpha}(pAL3).



View larger version (15K):
[in this window]
[in a new window]
 
Fig. 3. Expression of pixA is dependent on the growth phase of the recombinant strains DH5{alpha}(pAL3) ({circ}) and DH5{alpha}(pAL6) ({square}) as indicated by the {beta}-galactosidase activity of the pixA–lacZ transcriptional fusions. DH5{alpha} harbouring the vector pCB192 ({triangleup}) was used as control.

 
To monitor whether pixA expression might additionally also be influenced by temperature, DH5{alpha}(pAL6) was grown at 26 and 37 °C and {beta}-galactosidase activity was monitored. Lowering the temperature from 37 to 26 °C almost completely abolished pixA transcription, reducing the activity to a similar level as observed in DH5{alpha}(pAL3) (Fig. 4).



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 4. Expression of pixA in DH5{alpha}(pAL6) is dependent on temperature as indicated by the {beta}-galactosidase activity of the pixA–lacZ transcriptional fusions. {blacksquare}, 37 °C; {blacklozenge}, 26 °C.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this study, we identified and characterized the adhesin of the uropathogenic E. coli strain X2194 (O2 : K- : H6) which had been isolated from a patient with severe pyelonephritis. E. coli X2194 adhered to human epithelial cells and haemagglutinated erythrocytes derived from different species, including man. By using electron microscopy with negative staining we could demonstrate the presence of pili (Fig. 1). A chromosomal DNA fragment was identified by shotgun cloning which transferred the adhesive properties of the wild-type strain onto laboratory E. coli strains, such as CMK603, HB101 or JM109, for example. Sequencing of the cloned insert revealed a novel gene cluster containing seven complete ORFs which exhibited significant sequence similarity to genes of the pap operon. Pap pili are expressed in over 90 % of the clinical isolates found to be associated with severe upper urinary tract infections in non-hospitalized patients. In contrast, only 7 % of faecal isolates express Pap pili in healthy subjects (Källenius et al., 1981). In comparison with the pap operon, a gene corresponding to papB could be identified in front of the major pilin gene pixA; however, a homologue to papI could not be detected. Interestingly, no genes corresponding to the major component of the tip fibrillum, PapE, or the adaptor protein encoded by papK were found (Jacob-Dubuisson et al., 1993). In fact, the Pix pilus seems to be more related to the recently described Sfp pili from strain O147 : H7 (Brunder et al., 2001) as this adhesin is also lacking papE and papK homologues. Furthermore, PixG shows higher sequence homology to SfpG than to PapG. In addition, in the sfp operon, homologues of papI and papB could not be detected (Brunder et al., 2001). As the pix operon apparently does not contain genes encoding a tip-associated fibrillum, these data might indicate that the actual adhesin might not be located at the tip but instead could be periodically intercalated in the pilus rod (Ponniah et al., 1991).

The expression of fimbrial adhesins is usually regulated at the transcriptional level, whereas DNA rearrangement occurs in type 1 pili (Abraham et al., 1985). Methylation-dependent transcriptional regulation has been described for the pap operon, where the gene encoding the transcriptional activator PapB is co-transcribed with the gene of the major pilin subunit PapA and genes encoding additional pilus subunits from a promoter upstream of papB. The amount of the major pilin subunit PapA is a critical parameter for pilus synthesis (Nilsson & Uhlin, 1991). The second specific regulator for the synthesis of the Pap pilus is PapI. The gene encoding this is located upstream of papB and is transcribed in the opposite direction. Regulation of papBA transcription is dependent on the co-operative binding of Lrp to defined DNA sequences containing GATC motifs, which are differentially methylated by deoxyadenosine methylase (Braaten et al., 1994). The papB gene product increases pap operon expression and acts as a transcriptional regulator of papI as well. Therefore, PapB has an autoregulatory function (Forsman et al., 1989). The gene products of fanA and fanB serving as regulatory proteins in K99-pili production share significant homology to PapB but have been shown to work as transcriptional anti-terminators (Roosendaal et al., 1989). PapI, the second pilus-specific regulator, increases the affinity of Lrp for definite pap DNA sites and thereby induces a translocation of Lrp leading to an activation of pilus transcription. Similar mechanisms were described for the plasmid-encoded fimbriae (Pef) of Salmonella typhimurium (Nicholson & Low, 2000), but also some modifications were observed, e.g. the function of Pef. Other global regulators of Pap pilus transcription are the catabolite activator protein and H-NS (White-Ziegler et al., 1998).

Interestingly, although the pix operon exhibits significant sequence homology to the pap operon, Pix pilus synthesis appears to be regulated differently. Sequence analysis of the cloned fragment encoding the whole pix operon identified an AT-rich region containing several direct and inverted repeats upstream of pixB. As AT-rich domains have been suggested to serve as a target for transcriptional regulators, this sequence might harbour a putative regulatory site. AT-rich regions seem to have special properties, e.g. the A tracts of the oligonucleotide 5'-GGAAATTTCC-3' show a gradual compression of the minor groove as determined by NMR (Katahira et al., 1990). Additionally, A or T tracts longer than three nucleotides result in curved DNA when positioned on the same side of the helix (Hagermann, 1990). Recent studies by Xia et al. (1998) could demonstrate that the PapB protein recognizes a motif including a 9 bp repeat sequence containing T/A triplets at conserved positions. Inhibitory experiments with distamycin, a minor groove DNA binding drug (Coll et al., 1987), could demonstrate a competition effect between PapB and distamycin, which further suggested that the DNA architecture might be a factor in the transcriptional control of adhesin expression.

In contrast to the homologous pap or pef operon, no Lrp-binding sites and GATC motifs could be observed within the regulatory region of the pix operon. Lrp is a 19 kDa DNA-binding protein that activates some genes and represses others in the E. coli chromosome (Calvo & Matthews, 1994). The pap-regulatory region contains six binding sites each sharing the consensus sequence GN2–3TTT. Lrp acts as a repressor when bound to pap DNA sites [1,2,3] and as an activator when bound to DNA sites [5,6]; each of these two binding regions contains a GATC motif which is methylated by Dam. Differential methylation and binding of Lrp enables a transition from ‘phase on’ to ‘phase off’ pilus transcription (van der Woude et al., 1996; Weyand et al., 2001). In the corresponding regulatory region of the pix operon we did not detect any similar features by sequence alignment suggesting that other regulatory mechanisms might be involved. PapI is the second pilus-specific regulator in the pap operon shown to be essential for the translocation of PapB from Lrp-binding site [1,2,3] to Lrp-binding site [5,6]. We were not able to detect Lrp-binding sites in the pix operon and we could also not identify any ORF corresponding to a PapI homologue. Employing database searches and sequence alignment we identified a 952 bp fragment of the R6 gene which had been found in the PAI of the highly uropathogenic E. coli CFT073 recently reported by Guyer et al. (1998). This PAI spans a 58 kb region, and among the 44 ORFs identified it contains several virulence genes, including a complete pap gene cluster. Additionally, the authors described the R6 gene as showing a highly significant homology to a gene inside a gene cluster encoding nitrilotriacetate monooxygenase and NADH : flavin mononucleotide oxidoreductase in C. heintzii ATCC 29600 (Xu et al., 1997). By sequence alignment an ORF of 1467 bp encoding a protein of 488 aa was suggested. The deduced amino acid sequence had a weak but still significant identity (~25 %) to several transposases, including those from E. coli IS21 and Bacillus thuringiensis IS232. It was therefore suggested that this gene might be part of an insertion element, but no further evaluation of its function has been reported. Interestingly, an IS element has also been identified at the 3'-end of the related sfp operon (Brunder et al., 2001), which might indicate the presence of a PAI. This might even suggest a distribution of these related genes by horizontal gene transfer as had been reported for pap and prs operons (Marklund et al., 1992).

To further investigate the relevance of the R6 fragment for Pix pilus transcription in the absence of a second pilin-specific regulator homologue, we carried out initial studies using the vector pCB192 which contains a promoterless lacZ gene. Two pixA–lacZ fusions were constructed: the first contained the complete R6 region as well as the ORF encoding PixB; in the second construct the R6 region was deleted. These studies clearly demonstrated the effect of the R6 fragment on Pix pilus expression. pixA transcription was markedly reduced after deletion of this regulatory region and no regulation of expression could be observed. As was found also with the Pap pilus system, PixB proved not to be sufficient for transcriptional activation (Göransson et al., 1988). We conclude that expression of the pix operon possibly involves the interaction of the PapB homologue PixB with the R6 fragment. Dissection of the mechanism of regulation of Pix pili expression will require additional studies that have to elucidate the function of this region in more detail and to address possible interactions with global regulators (Morschhäuser et al., 1994). Additionally, the pix operon may be part of a PAI on the chromosome of the clinical isolate E. coli X2194.

The expression of fimbrial operons is regulated by environmental signals such as temperature, growth rate, aliphatic amino acids, oxygen level, iron, osmolarity and carbon source (Uhlin, 1994). In this study, we demonstrated that temperature-dependent expression of Pix pili occurs predominantly during the stationary growth phase as has been described for other virulence factors. The effect of growth temperature is mediated in many fimbrial systems by histone-like protein H-NS. Whether temperature regulation of the pix operon is achieved in a similar manner has to be established. However, the observation that in contrast to other pilus systems Pix pili are expressed during the stationary growth phase reflects the growth-phase-dependent regulation of virulence gene expression as an adaptive response to different microenvironments (Puente et al., 1996).


   ACKNOWLEDGEMENTS
 
We like to thank L. Greune for excellent electron microscopy. The partial support of this work by grants of the Deutsche Forschungsgemeinschaft (SFB293/TPB5), the German Federal Ministry of Education and Research (Fö. 01KS9604/0) and the Interdisciplinary Clinical Research Center (IZKF) Münster (TP D4) is gratefully acknowledged.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Abraham, J. M., Freitag, C. S., Clements, J. R. & Eisenstein, B. I. (1985). An invertible element of DNA controls phase variation of type 1 pili of Escherichia coli. Proc Natl Acad Sci U S A 82, 5724–5727.[Abstract]

Båga, M., Normark, S., Hardy, J., O'Hanley, P., Lark, D., Olsson, O., Schoolnik, G. K. & Falkow, S. (1984). Nucleotide sequence of the papA gene encoding the Pap pilus subunit of human uropathogenic Escherichia coli. J Bacteriol 157, 330–333.[Medline]

Braaten, B. A., Nou, X., Kaltenbach, L. S. & Low, D. A. (1994). Methylation patterns in pap regulatory DNA control pyelonephritis-associated pili phase variation in E. coli. Cell 76, 577–588.[Medline]

Brunder, W., Salam Khan, A., Hacker, J. & Karch, H. (2001). Novel type of fimbriae encoded by the large plasmid of the sorbitol-fermenting enterohemorrhagic Escherichia coli O157 : H-. Infect Immun 69, 4447–4457.[Abstract/Free Full Text]

Calvo, J. M. & Matthews, R. G. (1994). The leucine-responsive regulatory protein, a global regulator of metabolism in Escherichia coli. Microbiol Rev 58, 466–490.[Medline]

Coll, M., Frederick, C., Wang, A. H. & Rich, A. (1987). A bifurcated hydrogen-bonded conformation in the d(AT) base pairs of the DNA dodecamer d(CGCAAATTTGCG) and its complex with distamycin. Proc Natl Acad Sci U S A 84, 8385–8389.[Abstract]

Cravioto, A., Gross, R. J., Scotland, S. M. & Rowe, B. (1979). An adhesive factor found in strains of Escherichia coli belonging to the traditional infantile enteropathogenic serotypes. Curr Microbiol 3, 95–99.

Denich, K., Blyn, L. B., Craiu, A., Braaten, B. A., Hardy, J., Low, D. A. & O'Hanley, P. D. (1991). DNA sequences of three papA genes from uropathogenic Escherichia coli strains: evidence of structural and serological conservation. Infect Immun 59, 3849–3858.[Medline]

Dominguez, O. & Lopez-Larrea, C. (1994). Gene walking by unpredictable primed PCR. Nucleic Acids Res 22, 3247–3248.[Medline]

Forsman, K., Göransson, M. & Uhlin, B. E. (1989). Autoregulation and multiple DNA interactions by a transcriptional regulatory protein in E. coli pili biogenesis. EMBO J 8, 1271–1277.[Abstract]

Foxman, B., Zhang, L., Palin, K., Tallman, P. & Marrs, C. F. (1995). Bacterial virulence characteristics of Escherichia coli isolates from first-time urinary tract infection. J Infect Dis 171, 1514–1521.[Medline]

Girardeau, J. P., Bertin, Y. & 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.[Medline]

Göransson, M., Forsman, K. & Uhlin, B. E. (1988). Functional and structural homology among regulatory cistrons of pili-adhesin determinants in Escherichia coli. Mol Gen Genet 212, 412–417.[Medline]

Guyer, M. D., Kao, J. S. & Mobley, H. L. T. (1998). Genomic analysis of a pathogenicity island in uropathogenic Escherichia coli CFT073: distribution of homologous sequences among isolates from patients with pyelonephritis, cystitis and catheter-associated bacteriuria and from fecal samples. Infect Immun 66, 4411–4417.[Abstract/Free Full Text]

Hagermann, P. J. (1990). Sequence-directed curvature of DNA. Annu Rev Biochem 59, 755–781.[CrossRef][Medline]

Hanahan, D. (1983). Studies on transformation of Escherichia coli with plasmids. J Biol Chem 166, 557–580.

Hultgren, S. J., Abraham, S., Caparon, M., Falk, P., St Geme, J. W. & Normark, S. (1993). Pilus and nonpilus bacterial adhesins: assembly and function in cell recognition. Cell 73, 887–901.[Medline]

Ikaheimo, R., Siitonen, A., Karkkainen, U., Kuosmanen, P. & Mäkelä, P. H. (1993). Characteristics of Escherichia coli in acute community-acquired cystitis of adult women. Scand J Infect Dis 25, 705–712.[Medline]

Jacob-Dubuisson, F., Heuser, J., Dodson, K., Normark, S. & Hultgren, S. (1993). Initiation of assembly and association of the structural elements of a bacterial pilus depend on two specialized tip proteins. EMBO J 12, 837–847.[Abstract]

Jonsson, A. B., Pfeifer, J. & Normark, S. (1992). Neisseria gonorrhoeae PilC expression provides a selective mechanism for structural diversity of pili. Proc Natl Acad Sci U S A 89, 3204–3208.[Abstract]

Källenius, G., Möllby, R., Svenson, S. B., Winberg, J., Lundblad, A. & Svensson, S. (1980). The pK antigen as receptor for pyelonephritic E. coli. FEMS Microbiol Lett 7, 297–300.[CrossRef]

Källenius, G., Möllby, R., Svenson, S. B., Helin, I., Hultberg, H., Cedergren, B. & Winberg, J. (1981). Occurrence of P-fimbriated Escherichia coli in urinary tract infections. Lancet 26, 1369–1372.

Kao, J. S., Stucker, D. M., Warren, J. W. & Mobley, H. L. T. (1997). Pathogenicity island sequences of pyelonephritogenic Escherichia coli are associated with virulent uropathogenic strains. Infect Immun 65, 2812–2820.[Abstract]

Katahira, M., Sugeta, H. & Kyogoku, Y. (1990). A new model for bending of DNAs containing the oligo(dA) tracts based on NMR observations. Nucleic Acids Res 18, 613–618.[Abstract]

Korhonen, T. K., Valtonen, M. V., Parkinnen, J., Vaisanen-Rhen, V., Finne, J., Ørskov, I., Ørskov, F., Svenson, S. B. & Mäkelä, P. H. (1985). Serotype, hemolysin production and receptor recognition of Escherichia coli strains associated to neonatal sepsis and meningitis. Infect Immun 48, 486–491.[Medline]

Marklund, B. I., Tennent, J. M., Garcia, E., Hamers, A., Baga, M., Lindberg, F., Gaastra, W. & Normark, S. (1992). Horizontal gene transfer of the Escherichia coli pap and prs-operons as a mechanism for the development of tissue specific adhesive properties. Mol Microbiol 6, 2225–2242.[Medline]

Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Morschhäuser, J., Vetter, V., Emödy, L. & Hacker, J. (1994). Adhesin regulatory genes within large, unstable DNA regions of pathogenic Escherichia coli: cross-talk between different gene clusters. Mol Microbiol 11, 555–566.[Medline]

Nicholson, B. & Low, D. (2000). DNA methylation-dependent regulation of pef expression in Salmonella typhimurium. Mol Microbiol 35, 728–742.[CrossRef][Medline]

Nilsson, P. & Uhlin, B. E. (1991). Differential decay of a polycistronic Escherichia coli transcript is initiated by RNaseE-dependent endonucleolytic processing. Mol Microbiol 5, 1791–1799.[Medline]

Nou, X., Braaten, B., Kaltenbach, L. & Low, D. A. (1995). Differential binding of Lrp to two sets of pap DNA binding sites mediated by PapI regulates Pap phase variation in Escherichia coli. EMBO J 14, 5785–5797.[Abstract]

Parkkinen, J., Korhonen, T. K., Pere, A., Hacker, J. & Soinila, S. (1988). Binding sites in the rat brain for Escherichia coli S pili associated with neonatal meningitis. J Clin Investig 81, 860–865.[Medline]

Ponniah, S., Endres, R. O., Hasty, D. L. & Abraham, S. N. (1991). Fragmentation of Escherichia coli type 1 pili exposes cryptic D-mannose-binding sites. J Bacteriol 173, 4195–4202.[Medline]

Puente, J. L., Bieber, D., Ramer, S. W., Murray, W. & Schoolnik, G. K. (1996). The bundle-forming pili of enteropathogenic Escherichia coli: transcriptional regulation by environmental signals. Mol Microbiol 20, 87–100.[Medline]

Roberts, J. A., Marklund, B., Ilver, D. & 7 other authors (1994). The Gal(1–4)Gal-specific Tip adhesin of Escherichia coli P-fimbriae is needed for pyelonephritis to occur in the normal urinary tract. Proc Natl Acad Sci U S A 91, 11889–11893.[Abstract/Free Full Text]

Roosendaal, B., Damoiseaux, J., Jordi, W. & de Graaf, F. K. (1989). Transcriptional organisation of the DNA region controlling expression of the K99 gene cluster. Mol Gen Genet 215, 250–256.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schneider, K. & Beck, C. F. (1986). Promoter-probe vectors for the analysis of divergently arranged promoters. Gene 42, 37–48.[CrossRef][Medline]

Smyth, C. J., Marron, M. B., Twohig, J. M. & Smith, S. G. (1996). Fimbrial adhesins: similarities and variations in structure and biogenesis. FEMS Immunol Med Microbiol 16, 127–139.[CrossRef][Medline]

Southern, E. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98, 503–517.[Medline]

Swansson, J. & Koomey, J. M. (1989). Mechanisms for variation of pili and outer membrane protein II in Neisseria gonorrhoea. In Mobile DNA, pp. 743–761. Edited by D. E. Berg & M. M. Howe. Washington, DC: American Society for Microbiology.

Uhlin, B. E. (1994). Regulation of E. coli fimbrial expression. In Pili: Adhesin, Genetics, Biogenesis and Vaccines, pp. 171–178. Edited by P. Klemm. Boca Raton, FL: CRC Press.

Van der Woude, M. W., Braaten, B. & Low, D. A. (1996). Epigenic phase variation of the pap-operon in Escherichia coli. Trends Microbiol 4, 5–9.[CrossRef][Medline]

van Die, I., Dijksterhuis, M., deCock, H., Hoekstra, W. & Bergmans, H. (1986). Structural variation of P-fimbriae from uropathogenic Escherichia coli. In Protein-Carbohydrate Interactions in Biological Systems, pp. 39–46. Edited by D. L. Lark. London: Academic Press.

Weyand, N. J., Braaten, B. A., van der Woude, M., Tucker, J. & Low, D. A. (2001). The essential role of the promotor-proximal subunit of CAP in pap phase variation: Lrp- and helical phase-dependent activation of papBA transcription by CAP from -215. Mol Microbiol 39, 1504–1522.[CrossRef][Medline]

White-Ziegler, C. A., Angus Hill, M. L., Braaten, B. A., van der Woude, M. W. & Low, D. A. (1998). Thermoregulation of Escherichia coli pap transcription: H-NS is a temperature-dependent DNA methylation blocking factor. Mol Microbiol 28, 1121–1137.[CrossRef][Medline]

Willems, R., Paul, A., van der Heide, H. G., ter Avest, A. R. & Mooi, F. R. (1990). Fimbrial phase variation in Bordetella pertussis: a novel mechanism for transcriptional regulation. EMBO J 9, 2803–2809.[Abstract]

Wullt, B., Bergsten, G., Conell, H., Röllano, P., Gebretsadik, N., Hull, R. & Svanborg, C. (2000). P fimbriae enhance the early establishment of Escherichia coli in the human urinary tract. Mol Microbiol 38, 456–464.[CrossRef][Medline]

Xia, Y., Forsman, K., Jass, J. & Uhlin, B. E. (1998). Oligomeric interaction of the PapB transcriptional regulator with upstream activating region of pili adhesin gene promotors in Escherichia coli. Mol Microbiol 30, 513–523.[CrossRef][Medline]

Xu, Y., Mortimer, M. W., Fisher, T. S., Kahn, M. L., Brockman, F. J. & Xun, L. (1997). Cloning, sequencing and analysis of a gene cluster from Chelatobacter heintzii ATCC29600 encoding nitrilotriacetate monooxygenase and NADH : flavin mononucleotide oxidoreductase. J Bacteriol 179, 1112–1116.[Abstract]

Zhang, J. P. & Normark, S. (1996). Induction of gene expression in Escherichia coli after pilus-mediated adherence. Science 273, 1234–1236.[Abstract]

Received 28 January 2003; accepted 3 March 2003.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Lügering, A.
Articles by Schmidt, M. A.
Articles citing this Article
PubMed
PubMed Citation
Articles by Lügering, A.
Articles by Schmidt, M. A.
Agricola
Articles by Lügering, A.
Articles by Schmidt, M. A.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2003 Society for General Microbiology.