Infection and Immunity, February 2002, p. 1027-1031, Vol. 70, No. 2
0019-9567/01/$04.00+0 DOI: 70.2.1027-1031.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Department of Microbiology and Immunology, University of Maryland School of Medicine, Baltimore, Maryland 21201,1 Department of Medical MicrobiologyImmunology,2 Laboratory of Genetics, University of Wisconsin Madison, Wisconsin 53706;,3 Institute of Biochemistry, Biological Research Center, Hungarian Academy of Science, Szeged, Hungary 67014
Received 29 June 2001/ Returned for modification 8 October 2001/ Accepted 12 November 2001
|
|
|---|
background. However, when the gene fragment is transformed back into the native EDL933 background, the enzymatic activity of the cloned determinants is undetectable. We speculate that an unidentified trans-acting factor in enterohemorrhagic E. coli (EHEC) is responsible for this regulation of ure expression. In addition, Fur-like recognition sites are present in three independent O157:H7 isolates upstream of ureD and ureA. Enzymatic assays confirmed a difference in urease expression of cloned EHEC ure clusters in E. coli MC3100
fur. Likewise, interruption of fur in O157:H7 isolate IN1 significantly diminished urease activity. We propose that, similar to the function of Fur in regulating the acid response of Salmonella enterica serovar Typhimurium, it modulates urease expression in EHEC, perhaps contributing to the acid tolerance of the organism. |
|
|---|
As part of a comparative study of E. coli genomes, the chromosome of EHEC strain EDL933 (38) was recently sequenced at the E. coli genome center (http://www.genome.wisc.edu) (35). Amidst the 1.4 Mb of DNA that was not homologous with E. coli K-12 sequence, two urease gene clusters were identified in EDL933 at nonadjacent loci (35). Both gene clusters are predicted to encode the structural proteins and accessory polypeptides necessary for the assembly of urease. This complex nickel metalloenzyme catalyzes the hydrolysis of urea into ammonia and carbon dioxide (32). It has been demonstrated previously that ammonium ions accumulate in the proximal surroundings of urease-positive bacteria, mediating an increase in pH (32). It is postulated elsewhere that some acid-tolerant organisms like Helicobacter pylori, Streptococcus salivarius, and Yersinia enterocolitica exploit this enzyme to buffer themselves in acidic niches within a host (6, 13, 32, 41).
Similar to these pathogens, E. coli O157:H7 is known to have a high tolerance to acid (1, 29). This phenotype is partially inhibited by chloramphenicol, suggesting that de novo protein synthesis is probably required (2, 11). Interestingly, urease production in H. pylori and S. salivarius has been demonstrated elsewhere to increase under acidic culture conditions (6, 42). We speculate that urease production could be part of a metabolic adaptation employed by EHEC to transiently neutralize acidity experienced during an oral route of infection (2, 11) or with exposure to organic acids present in the human intestinal tract (12).
In this report, we demonstrate that EDL933 carries functional urease genes, which can be expressed in an E. coli K-12 background. EHEC ureases are preceded by Fur-like boxes, which appear to regulate ure expression as detected by enzymatic assays. This regulation is illustrated in both a recombinant system and an isogenic fur mutant generated in this study. The significance of Fur regulating acid tolerance is well documented elsewhere for Salmonella enterica serovar Typhimurium (17, 21, 36). Variations in ure regulation noted among O157:H7 strains used in this study suggest that additional mechanisms of regulation probably exist.
|
|
|---|
The structural proteins UreA, UreB, and UreC are predicted to share 80 to 96% similarity with the K. aerogenes polypeptides, whereas the translated accessory gene products of ureD, ureE, ureF, and ureG are predicted to share 80, 77, 76, and 93% similarity with K. aerogenes homologues, respectively. These urease accessory proteins are speculated to assist in procuring and coordinating the nickel ions in the active site of the apourease (10, 23, 27, 43).
|
|
|---|
However, a 5,000-bp DNA fragment encoding ureDABCEFG (pEDL ure) derived from EDL933 conferred urease activity on E. coli DH5
(39) upon transformation (550 nmol of NH3 generated/min/mg of protein) (Fig. 1). This gene cluster was PCR amplified from genomic DNA preparations of EDL933 using the Expand Long Template PCR system (Boehringer Mannheim) with primer pairs 5"-TCGGAGCTCTCTGCCTGATTCACTGGATAA-3" (upstream primer) and 5"-AACGCCAACTTGGATCCTTCCTTCTGATAA-3" (downstream primer) (3, 39). The gel-purified PCR product, which extends from 44 bp upstream of the first possible ureD translational start site to 53 bp downstream of the ureG stop codon (Fig. 2), was cloned into the EcoRV site of pBS SK+ for expression (3, 39). Because the two ure gene clusters of EDL933 are identical, it is not clear which of the two loci was cloned.
![]() View larger version (31K): [in a new window] |
FIG. 1. Urease activities of cloned EHEC ure clusters expressed in E. coli DH5 and native EDL933 background. Strains were cultured overnight in M9 medium modified with 1.0 µM NiCl2. French-pressed lysates were standardized for protein concentration and assayed for urease activity, reported as nanomoles of NH3 per minute per milligram of protein (30, 48). Each column represents triplicate measurements averaged over three independent experiments. Error bars indicate standard deviations. Cloned EDL933 ure (pEDLure) wasreisolated from EDL933 and transformed into E. coli DH5 (indicated by arrows). As a control, assay results with recombinant P. mirabilis ure (pMID1010) (24) transformants are included. E. coli DH5 (pBS SK+) routinely hydrolyzes <10 nmol of urea/min/mg of protein. The inset is an anti-UreC immunoblot of lysates of E. coli DH5 carrying cloned EHEC ure clusters (3, 22). Relative molecular masses are indicated on the left. The UreC homologue is predicted to migrate at 60 kDa (right arrow).
|
![]() View larger version (44K): [in a new window] |
FIG. 2. A 300-bp DNA sequence upstream of the EHEC ure cluster in strains EDL933, IN1, and MO28. Recombinant EHEC ure constructs used in this study include DNA sequence extending to the region labeled as ure upstream primer. This stretch of DNA encodes three putative translational start sites for the ureD homologue (marked as "1st SC," "2nd SC," and "3rd SC"). A 70-like promoter (boldface) was identified between the first and second putative ureD start codons (SC). Based on the consensus sequence reported in the work of Escolar et al. (16), two Fur-like boxes were found upstream and overlapping the 70-like promoter, together with another recognition sequence identified further downstream. These Fur-like boxes are highlighted within the sequence (boxes) with an indication of their relative direction (arrows).
|
cells transformed with the cloned ure clusters of EDL933, IN1, and MO28 (Fig. 1). Furthermore, cell lysates immunoblotted with antiserum to P. mirabilis UreC, the largest structural subunit, cross-reacted with appropriately sized polypeptides (Fig. 1, inset) (3, 22).
These observations indicate that EHEC strains EDL933, IN1, and MO28 carry functional urease genes, despite the fact that ure gene expression in the native EDL933 cannot be detected on Christensen's urea agar or stationary-phase L broth cultures. In contrast, O157:H7 strains IN1 and MO28 do produce an enzymatically active urease during stationary-phase growth in L broth cultures (200 ± 10 and 85 ± 20 nmol of NH3/min/mg of protein, respectively). Note that this activity is less than that expressed by the cloned determinants in E. coli DH5
(Fig. 1).
|
|
|---|
background, pEDLure was transformed into EDL933 and subsequently analyzed for catalytic activity. As with the native strain, EDL933 (pEDLure) transformants do not produce urease. Thus, the cloned ure genes were repressed in their native background (Fig. 1). However, derepression could be demonstrated upon isolation of pEDLure and retransformation into E. coli DH5
(700 ± 300 nmol of NH3/min/mg of protein) (Fig. 1). Based on this observation, we postulated that a trans-acting factor regulates EDL933 ure expression.
Furthermore, the cloned EDL933 ure determinants were not expressed in EDL933
ure, in which DNA sequences extending 760 bp upstream and 840 bp downstream of both ure loci have been deleted by a method described in the work of Pósfai et al. (37; data not shown). Thus, we postulated that the sequence encoding this trans-acting factor is not immediately adjacent to either of the ure loci.
In control experiments, P. mirabilis ure genes (pMID1010) (24) were shown to confer the same level of enzymatic activity upon EDL933 and DH5
(Fig. 1). Thus, the regulatory effect appears specific for the EDL933 ure genes.
|
|
|---|
70-like promoter, situated halfway between the two possible start sites (Fig. 2). The second and third Fur boxes overlap and lie downstream of the putative promoter, respectively (Fig. 2). For comparison with other O157:H7 isolates, a 600-bp region immediately adjacent to the ure subclones of strains IN1 and MO28 was PCR amplified with primer pair 5"-TCATCGTACCCCCATTAGTTAGAGC-3" (upstream primer) and 5"-GCTGTGATTATGCCTTCTCGCTCAG-3" (downstream primer) and sequenced at the University of Maryland, Baltimore Biopolymer Core Facility (3, 40). This region was highly conserved between EDL933 and these two urease-positive strains (data not shown).
To examine whether these Fur-like boxes had any effect on O157:H7 ure transcription, cloned EHEC ure determinants were transformed into E. coli MC3100 and an isogenic fur mutant (44). Cloned EDL933 ure expression, measured by enzymatic activity (30, 48), increased a modest fivefold in the fur mutant background (P = 0.042) (Fig. 3B). In contrast, the urease activities associated with the cloned IN1 and MO28 gene clusters decreased dramatically in a
fur background. E. coli MC3100(pINure) produced 99.5 nmol of NH3/min/mg of protein, a 50-fold excess over that of the isogenic
fur strain (2.1 nmol of NH3/min/mg of protein) (P < 0.001) (Fig. 3B). Likewise, E. coli MC3100(pMOure) urease activity exceeds that of the
fur mutant by 25-fold (Fig. 3B) (P < 0.001).
![]() View larger version (16K): [in a new window] |
FIG. 3. Urease activity of cloned ure gene cluster of EHEC strains EDL933, MO28, and IN1 in the E. coli MC3100 fur strain. Cloned EHEC ure clusters were transformed into E. coli MC3100 and an isogenic fur strain. French-pressed lyates were prepared from overnight cultures grown in M9 medium modified with 1.0 µM NiCl2. Lysates were standardized for protein concentration and assayed for urease activity, reported as nanomoles of NH3 per minute per milligram of protein (30, 48) (B). Each bar represents the average of triplicate measurements of three separate experiments. Student t test results are reported for paired strains. As a positive control, a recombinant EDL933 chuA-phoA fusion (45) was introduced in trans into MC3100 and MC3100 fur. Subsequent assays of alkaline phosphatase activity are reported as arbitrary units based on the method of Brickman and Beckwith (5) (A).
|
fur strain (45). As an indicator of chuA transcription, alkaline phosphatase activity was measured in mid-exponential-phase L broth cultures according to the method of Brickman and Beckwith (5). As described in previous studies, chuA expression was fivefold greater in the
fur strain than in the wild type (P = 0.014) (45) (Fig. 3A).
These data suggest that Fur can regulate recombinant ure genes of multiple O157:H7 isolates (Fig. 3B). It is unclear why the phenotypes of the IN1 and MO28 cloned ure determinants were altered dramatically in a
fur background while EDL933 ure showed only a small change in activity despite the high degree of homology upstream of the ure clusters, including the Fur-like boxes.
To confirm that Fur influences EHEC ure expression, nonpolar insertional mutations were constructed within the fur loci of IN1 and EDL933. This was accomplished by PCR amplifying a 2,400-bp region encoding fur and flanking sequences from genomic DNA with the primer pair 5"-GACTGCCTGTTCTGCTATGATTG-3" (upstream primer) and 5"-TACTTCCTGCAACGTATGACTCC-3" (downstream primer) (3). PCR products were cloned into the EcoRV site of pBS SK+ and subsequently interrupted at an internal XmnI site with a 1,200-bp nonpolar aphA cassette originating from pUC4K (47). After confirmation of the construction of fur::aphA by DNA sequence analysis (40), this insertional mutation was transferred into the SmaI site of the
pir-dependent suicide vector pCVD442 (15). Mutations in fur were generated in nalidixic acid-resistant isolates of IN1 and EDL933 by conjugation with E. coli S17
pir transformed with the suicide construct (20). Transconjugants having undergone replacement recombination were confirmed by PCR amplification of genomic DNA with oligomers (described above) flanking fur, generating a single 3,500-bp product (3; data not shown). Control PCR amplifications were performed with genomic DNA from IN1 and EDL933 or plasmid DNA (pfur::aphA442) isolated from E. coli S17
pir, which generated 2,400- and 3,500 bp DNA products, respectively (data not shown).
Similar to results in an E. coli MC3100
fur background (Fig. the IN1 fur::aphA mutant produced significantly less urease activity than did the isogenic wild-type strain. The urease specific activities for IN1 and IN1 fur::aphA were 199.2 ± 7.8 and 111.9 ± 8.7 nmol of NH3/min/mg of protein, respectively (P < 0.001 by the Student t test). These strains were also transformed with pchuA-phoA. The alkaline phosphatase activities for IN1 and IN1fur::aphA were 14.5 ± 1.6 and 79.1 ± 15.6 arbitrary units (5), respectively. (P was 0.001). EDLfur::aphA did not differ in ure expression from EDL933 because neither strain produced detectable urease activity in stationary-phase cultures (data not shown).
Altogether, these results suggest that Fur may enhance the basal level of ure expression in O157:H7 isolates IN1 and MO28 during stationary-phase growth. Preliminary characterizations of ureD transcriptional fusions suggest that both IN1 fur::aphA and EDL fur::aphA produce less transcript than do the isogenic wild-type strains (data not shown). Thus, Fur-dependent regulation is probably occurring in EDL933, similar to IN1, but secondary to another mechanism of regulation.
|
|
|---|
Bacterial cultures were grown overnight in 5.0 ml of M9 minimal medium containing 1 µM NiCl2 at 37°C with aeration. Several variations in culture conditions were assessed for effects on urease expression, including (i) M9 minimal medium in which NH4Cl was separately replaced with 0.2% (wt/vol) proline, 0.2% (wt/vol) histidine, and 0.2% (wt/vol) glutamine (4, 9); (ii) M9 minimal medium in which NH4Cl was replaced with 0, 100, 250, and 500 mM urea (34); and (iii) M9 minimal medium in which sodium phosphate was omitted and subsequently buffered with 2-(N-morpholino)ethanesulfonic acid (MES) at pH 5.0, 6.0, and 7.5 (6, 41).
Under all conditions tested, no urease activity was detected in O157:H7 strain EDL933. Likewise, polyclonal sera against UreC and UreD failed to react with proteins in whole-lysate preparations (data not shown). In contrast, IN1 had 1.9-fold-greater enzymatic activity in L broth at pH 5.0 than at pH 7.0 (148 ± 10.1 versus 77.4 ± 3.6 nmol of NH3/min/mg of protein) (P = 0.037), suggesting that enhanced urease expression may be correlated with exposure to acidic environments.
Introduction of a gene encoding a high-affinity Ni2+ transporter (nixA) of H. pylori also did not affect urease detection in EDL933 under any condition tested; thus, Ni2+ availability was not limiting for the metalloenzyme (18).
As the frequency and severity of E. coli O157:H7 outbreaks continue to increase, there is an accompanying need to understand the pathogenicity of this organism. Undoubtedly during the course of infection, EHEC encounters both gastric acidity and permeable, weak acids in the human intestinal tract (7). It is not surprising that EHEC has evolved a resilient mechanism to regulate acid resistance, offering protection from transient and prolonged exposure to various acids.
Overall, our findings suggest that O157:H7 strains encode functional urease genes, which may be regulated both by an unindentified trans-acting factor and, more conclusively, by Fur. Like the glutamate- and arginine-dependent systems described elsewhere for E. coli (26, 28, 29), urease expression could modify internal and/or surrounding anion concentrations, enabling EHEC to survive acidic conditions and perhaps contributing to its low infectious dose. To our knowledge, this report is the first description of a Fur-regulated urease.
|
|
|---|
|
|
|---|
fur (24) were generously supplied by P. F. Sparling at the University of North Carolina at Chapel Hill and may otherwise be known as W3100
lac and W3100 fur
lac (14). The chuA-phoA fusion was the gift of Alfredo Torres at the CVD, University of Maryland, Baltimore, and was constructed as pSHU251 as described in the work of Mills and Payne (31). This work was supported in part by Public Health Service grant AI23328 from the National Institutes of Health.
|
|
|---|
80 transducing phages. J. Mol. Biol. 96:307-316.[CrossRef][Medline]
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»