|Home | About | Journals | Submit | Contact Us | Français|
In an effort to uncover the role of the high affinity Zn(II) uptake system in uropathogenic Escherichia coli CFT073, we deleted the znuB gene, which encodes for the transmembrane component of the ZnuABC transporter system. The null mutant for znuB did not grow on minimal medium unless supplemented with excess Zn(II) (50 µM ZnCl2). In contrast, the E. coli K-12 ΔznuB cell line grew well on minimal medium that was not supplemented with Zn(II). The ΔznuB mutant was significantly deficient in formation of biofilm under static conditions and also showed a substantially-reduced migration front of swarm cells. Because motility and biofilm formation are important for E. coli CFT073 pathogenicity, we propose that the high affinity Zn(II) uptake system may contribute to the virulence of this pathogen in the urinary tract.
Zinc is an essential element that plays an important catalytic and structural role in a number of proteins. Efficient zinc acquisition systems are used to scavenge zinc from the environment under zinc-restricted conditions (Blencowe & Morby, 2003, Hantke, 2001, Hantke, 2005). Available information suggests that Zn(II) import in E. coli is accomplished by ZupT (broad metal ion transport system) (Grass, et al., 2005, Grass, et al., 2002) and ZnuABC (Patzer & Hantke, 1998) (high affinity Zn(II) uptake system). The gene znuA encodes for the periplasmic, zinc-binding component of the transporter, znuB encodes for the transmembrane component, and znuC encodes for the ATPase subunit. In E. coli, the znuA, znuC, and znuB genes are divergently-oriented, and their transcriptional starting points are separated by a few base pairs. Zn(II)-loaded Zur, which belongs to the Fur family of metalloregulatory proteins, binds to a nearly perfect palindrome found in the bidirectional promoter region of znuA and znuCB, blocking both znuA and znuCB transcription (Patzer & Hantke, 2000). However, under Zn(II)-deficient conditions, Zur de-represses and induces the expression of this ABC transporter. In our previous study, we found that znuA in E. coli K-12 was up-regulated over 42 times in Zn(II) deficient conditions(Sigdel, et al., 2006). ZupT appears to be a housekeeping divalent metal uptake transporter, expressed constitutively at low levels (Grass, et al., 2002). In addition to the ZnuABC and ZupT systems, the phosphate uptake system (PitA) may also be involved in Zn(II) uptake (Beard, et al., 2000). Although these import pathways exist in E. coli K12 and other pathogenic strains, Zn(II)-inducible ZnuABC plays the primary role in acquiring Zn(II) when Zn(II) levels are low in the environment.
Although E. coli strains such as K-12 live as commensals in the large intestines of humans and other animals, other E. coli strains are pathogenic and can lead to neonatal meningitis and urinary tract infections (Kaper, et al., 2004). The latter groups of pathogens are responsible for 70–90% of infections causing acute cystitis and cases of pyelonephritis (Welch, et al., 2002). Both pathogenic and non pathogenic E. coli have evolved through a complex process so that only about 39% of genes have a common ancestor (Welch, et al., 2002). Most of the other genes are newly-acquired via horizontal gene-transfer events (Kao, et al., 1997). Uropathogenic E. coli (UPEC) are uniquely endowed with various virulence traits, enabling them to survive and grow in urine and other extraintestinal environments. UPEC strains are more likely to carry multiple determinants for particular adhesins (pyelonephritis-associated pili (pap), P and S fimbriae) and toxins (α-hemolysin (hly) and cytotoxic necrotizing factor) than strains found in feces (Kao, et al., 1997). The importance of iron in microbial infections has been demonstrated, and additional pathways to acquire iron in the urinary tract have already been discussed (Welch, et al., 2002). E. coli has evolved several mechanisms to acquire iron, including the production of siderophores, such as enterobactin and aerobactin (Torres, et al., 2001). Aerobactin, an iron-chelating siderophore, is more common in E. coli isolated from urine than from feces. As for E. coli CFT073, it has an additional TonB-mediated iron acquisition system, and this additional iron uptake system is important for the virulence of uropathogenic E. coli strains (Torres, et al., 2001).
It has been recently shown that ZnuABC, the high affinity Zn(II) transporter, plays a critical role in infection of enteric pathogen Salmonella enterica Serovar Typhimurium (Ammendola, et al., 2007). Compared to the human gastrointestinal (GI) tract, the urinary tract has low Zn(II) levels, containing < 5 µM Zn(II) (Cano Pavon, et al., 1986, Horng & Lin, 1997, Kimura, et al., 2005). We hypothesize that the high affinity Zn(II) uptake system plays an important role in Zn(II) uptake of bacteria in the urinary tract. In this study we have characterized the role of ZnuABC, the high affinity Zn(II) uptake system in the pyelonephritic E. coli CFT073 strain and compared the results with those using the E. coli K-12 strain. These data clearly show that ZnuABC affects virulence-associated phenotypes of uropathogenic Escherichia coli (CFT073).
Deletion of znuB in E. coli CFT073 was accomplished by using the lambda Red recombination method described by Datsenko and Wanner (Datsenko & Wanner, 2000). A linear DNA fragment encoding the CmR gene was generated by PCR using oligonucleotide primers with a 40-base-pair region of homology to regions containing or flanking the znuB gene. Fragments were introduced into the E. coli CFT073 strain by electroporation, and CmR recombinants were isolated. Deletions were confirmed by PCR analysis using locus-specific primers. The deletion mutant of E. coli K12 was constructed using the same primers and procedure.
Luria-Bertani (LB) and M9 minimal media were prepared and used as described (Paliy & Gunasekera, 2007). E. coli cultures were first grown at 37 °C in M9 medium supplemented with 0.2% glucose or LB medium. Cells were harvested and washed twice with phosphate buffered saline (PBS) and resuspended to the desired starting optical density in the appropriate fresh medium. Inoculated cultures were grown in Erlenmeyer flasks at 37 °C until they reached late log or stationary phase. Growth was monitored spectrophotometrically by measuring each culture’s optical density (OD600nm) periodically. The growth curves of at least three independent cultures were carried out for each strain in M9 minimal medium with 50 µM ZnCl2 or without zinc.
Plasmid pTSG110 was constructed by cloning the coding regions of the compete znuABC operon into the pACYC-177 vector using HindIII and XhoI (see Table 1 for primers used to clone the genes). The resulting plasmid was transformed into E. coli CFT073 ΔznuB by electroporation, and complementation was determined by conducting growth curves and metal analyses.
Bacterial cultures were grown in M9 minimal medium, and 2.5 × 109 cells (equivalent to 1 ml of cells having an OD600nm of 1) were collected by centrifugation at 10,000 × g for 5 min. Cytoplasmic fractions were separated from periplasmic fractions by using the PeriPreps™ periplasting kit (Epicentre, Madison, WI). The metal content of the cytoplamsic fractions was measured using an inductively coupled plasma instrument with mass spectrometry detection (ICP-MS). A calibration curve with 4 standards and a correlation coefficient of greater than 0.999 was generated using Zn(II) reference solutions from Fisher Scientific. Cytoplasmic fractions were diluted with 1% nitric acid and analyzed for zinc. At least three independent sets of cultures/metal analyses were conducted. In our calculations of Zn(II) content in the cytoplasm, we used the volumes of these spaces found at http://gchelpdesk.ualberta.ca/CCDB/cgi-bin/STAT_NEW.cgi.
Cell motility assays of E. coli CFT073 were performed in M9 minimal medium supplemented with 0.2% glycerol and 0.2% agar and in LB and ½ LB (LB diluted 1:1 with Nanopure water). Bacteria were grown in LB medium, and cells were washed twice with PBS and diluted to OD600nm = 0.2. Five microliter aliquots were subsequently spotted in the middle of the agar plates. Plates were incubated at 37 °C and then inspected for cell migration after 8, 24, and 48 hrs. Migration distances were measured from the center of the spot to the edge of the growth circle, and the reported migration distances were averages of four independent experiments.
Bacteria were grown in LB medium, and recovered cells were washed twice with PBS and diluted to OD600nm = 0.02 with 5 ml of fresh LB or M9 minimal medium. Duplicate cultures for each sample were incubated for 16h stagnant at 37 °C. One tube was sonicated immediately for 5 s with a thin probe, and the OD600nm of the culture was determined to estimate total cell biomass. The supernatant of the other tube was aspirated and rinsed thoroughly with distilled water. The cells attached to the tube walls were visualized and quantified by staining with crystal violet and solubilization with ethanol–acetone as described by Tomaras (Tomaras, et al., 2003). The OD580/OD600 ratio (OD580 is used to measure the stain while OD600 is used to measure cell density) was used to normalize the amount of biofilm formed to the total cell content of each sample tested to avoid variations due to differences in bacterial growth under different experimental conditions. All assays were done at least twice using fresh samples each time. In addition, every test was done in duplicate each time.
In an effort to uncover the role of the high affinity Zn(II) uptake system in E. coli CFT073, we deleted the znuB gene, which encodes for the transmembrane component of the transporter system. We chose to delete znuB, rather than znuA or znuABC, because ZnuA has a pronounced effect on the equilibrium concentrations of Zn(II) in the periplasm by pulling Zn(II) from other Zn(II) metalloproteins, such as CuZn superoxide dismutase (Berducci, et al., 2004). The alteration of the periplasmic Zn(II) equilibrium would affect the amount of Zn(II) available to enter the cytoplasm by other Zn(II) and nonspecific divalent metal importers and likely complicate the subsequent interpretation of the results. The growth of E. coli CFT073 ΔznuB mutant line was compared to that of its parent strain and the znuB-deleted E. coli K12 (BW25113) strain. Although the znuB deleted strain of E. coli CFT073 did initiate some growth, total biomass density did not exceed an OD600nm of 0.4 when the cells were grown in M9 minimal medium after 24 hrs (Figure 1A). However, this phenotype was fully rescued by adding Zn(II) to the M9 minimal medium (Figure 1C), suggesting the growth deficiency of the mutant strain is due to the loss of the high-affinity Zn(II) import system. Although the znuB mutant cell grew well on LB medium, growth of the znuB mutant was inhibited when Zn(II) was chelated from the growth medium by TPEN (Figure 1B and Figure 2). However, the wild-type E. coli CFT073 strain grew well and formed visible colonies on TPEN-treated LB-agar plates (Figure 2). We complemented the ΔznuB mutant by inserting znuB into the pACYC-177 vector (pTSG121) and expressing ZnuB. The expression of ZnuB rescued cell growth when cells were cultured in LB medium containing 20 µM TPEN (Figure 1B). These results confirm that ZnuABC plays an important role in Zn(II) uptake from the environment in the CFT073 strain.
Metal analyses were used to determine the amount of Zn(II) in the cytoplasm of the cells when cultured in different media. Our LB medium contained 10 ± 3 µM Zn(II), and the minimal medium contained 100-fold less Zn(II) than the LB. The wild-type CFT073 strain contained 0.56 ± 0.18 mM and 0.31 ± 0.04 mM Zn(II) in the cytoplasm of the cells when cultured in LB or M9 media, respectively. Similar concentrations of Zn(II) in E. coli have been previously reported by other groups (Graham, et al., 2009, Outten & O'Halloran, 2001). The ΔznuB mutant of the CFT073 strain contained 0.20 ± 0.04 mM and 0.17 ± 0.03 mM Zn(II) in the cytoplasm of the cells when cultured in LB or M9 media, respectively. This drop in Zn(II) is attributed to the loss of ZnuB, because the same number of cells were analyzed for metal content. The complemented cells (CFT073 ΔznuB + pTSG110) contained 0.41 ± 0.14 mM Zn(II) in the cytoplasm of the cells when cultured in LB medium, which is consistent with the rescued growth phenotype of this strain (Figure 1B). These data suggest that ZnuABC in the CFT073 strain must play a vital role in the sequestration of this metal in the urinary tract.
Under similar conditions the znuB-deleted E. coli K-12 strain did not show significant growth reduction. Although growth rates were slower than its counterpart parent strain, K12 ΔznuB achieved final OD600 of 2.5 when grown in M9 medium, which is significantly higher than the growth of E. coli CFT073 ΔznuB strain (Figure 1A). In this study we found that the high affinity ZnuABC Zn(II) transporter plays a far more important role in the acquisition of zinc from the environment for uropathogenic E. coli CFT073 than for the K-12 strain, and this transporter system is critical for the CFT073 pathogen. The differences likely reflect genetic differences, the lifestyles of these two E. coli strains, and the bioavailability of zinc in their natural habitats. Furthermore, the Zn(II) demand for CFT073 is most likely higher as this organism has a larger genome size compared to K-12 strain. The two strains share only 39% of their genes, and the different genes/proteins in CFT073 may require different amounts of Zn(II) (Welch, et al., 2002).
Based on our results and from previously-published information (Yang, et al., 2006), ZnuABC plays an important role in the acquisition of Zn(II) from the environment. In addition, ZnuABC has an important role in the pathogencity Salmonella enterica serovar Typhimurium (Ammendola, et al., 2007). In this study, the authors deleted znuA and infected mice with the mutant strain, and a dramatic reduction in the pathogenicity was observed in mice, independent of the route of infection (intraperitoneal or oral) or of the genetic background of the mice (BALB/c and DBA-2) (Ammendola, et al., 2007). In addition, znuA mutants in B. abortus, H. ducreyi, and P. multocida were found to be significantly less virulent than wild-type strains when tested in animal models. Yang et al. (Yang, et al., 2006) assessed the potential of using the ΔznuA mutant as a live vaccine against infections of Brucella melitensis. These results demonstrate that the ZnuABC transporter is absolutely required for full bacterial virulence.
Although we have not tested pathogenicity of the uropathgenic E. coli ΔznuA mutant in an animal model, we have found that the znuB mutant exhibits two reduced virulence phenotypes (less biofilm formation and less swimming motility) than the corresponding parent strain. Cell motility assays in M9, ½ LB, or LB growth media demonstrated 40–80% reduction in migration, depending on the medium used (Figure 3 and Figure 4). In addition, we evaluated the ability of the wild-type and ΔznuB CFT073 strains to form biofilm using the assay of Tomaras (Tomaras, et al., 2003). In M9 medium, wild-type CFT073 exhibited an OD580/600 of 0.78 ± 0.07, while the ΔznuB line exhibited an OD580/600 of 0.46 ± 0.09. In LB medium, wild-type CFT073 exhibited an OD580/600 of 0.065 ± 0.012, while the ΔznuB line exhibited an OD580/600 of 0.016 ± 0.007. Motility can have a profound impact on the colonization of surfaces by formation of adherent microbial assemblies (biofilms) and subsequent accumulation and lateral expansion of adherent biomass. In a microarray study, we found that flagellar biosynthesis was down-regulated in the E. coli K-12 strain under Zn(II)-limited conditions, (Gunasekera and Crowder, unpublished results). Consistent with this previous observation, we confirm that the ΔznuB mutant line is less motile. Taken together, these observations suggest that the expression of flagellar biosynthetic genes is Zn(II)-dependent.
However, free Zn(II) concentrations in mammalian hosts are very low (Gaither & Eide, 2001), and survival and proliferation of bacteria within animal hosts are highly-dependent on the uptake and sequestration of Zn(II) from Zn(II)-depleted environments such as the urinary tract. In our previous studies on E. coli K-12 (Sigdel, et al., 2006), we found that znuABC is significantly up-regulated in Zn(II)-deficient conditions. Although E. coli CFT073 has unexpectedly high laterally-transferred genes and additional heme uptake systems, zinc uptake systems in addition to ZupT (constitutively-expressed broad range metal iron uptake system) have not been reported. The ZupT transporter is constitutively-expressed and plays an housekeeping role to uptake Zn(II) from the environment. The data presented herein suggest that ZupT, nor any other Zn(II) importer, is not constitutively-expressed in CFT073; therefore, it is not clear how this uropathogenic strain, which contains a larger genome and exists in a low Zn(II) environment, can survive. It is possible that ZupT is expressed in CFT073; however, it cannot import Zn(II) at the low levels found in the minimal medium. It is also possible that the ZnuABC system in CFT073 may sequester Zn(II) better than the comparable system in K-12, thereby eliminating the need for redundant, constitutively-expressed Zn(II) importers. Future DNA microarray or quantitative PCR along with protein studies are needed to address this interesting hypothesis. The work described herein demonstrates that ZnuABC from E. coli CFT073 is a potential antibiotic target and the understanding of Zn(II) homeostasis in bacteria may identify other potential targets for the generation of a novel class of antibiotics that target metal ion transport.
The authors would like to thank the National Institutes of Health (GM079411 to M.W.C.), Volwiler Professorship (to M.W.C.), and Miami University (URG to A.H.H.) for funding this work.