ABSTRACT
Uropathogenic Escherichia coli (UPEC) is responsible for the majority of uncomplicated urinary tract infections (UTI) and represents the most common bacterial infection in adults. UPEC utilizes a wide range of virulence factors to colonize the host, including the novel repeat-in-toxin (RTX) protein TosA, which is specifically expressed in the host urinary tract and contributes significantly to the virulence and survival of UPEC. tosA, found in strains within the B2 phylogenetic subgroup of E. coli, serves as a marker for strains that also contain a large number of well-characterized UPEC virulence factors. The presence of tosA in an E. coli isolate predicts successful colonization of the murine model of ascending UTI, regardless of the source of the isolate. Here, a detailed analysis of the function of tosA revealed that this gene is transcriptionally linked to genes encoding a conserved type 1 secretion system similar to other RTX family members. TosA localized to the cell surface and was found to mediate (i) adherence to host cells derived from the upper urinary tract and (ii) survival in disseminated infections and (iii) to enhance lethality during sepsis (as assessed in two different animal models of infection). An experimental vaccine, using purified TosA, protected vaccinated animals against urosepsis. From this work, it was concluded that TosA belongs to a novel group of RTX proteins that mediate adherence and host damage during UTI and urosepsis and could be a novel target for the development of therapeutics to treat ascending UTIs.
INTRODUCTION
Repeat-in-toxin (RTX) proteins are widespread among Gram-negative bacteria, with more than 1,000 family members detected in a survey of genome sequences from 251 bacterial species (21). Two common features present in known RTX family members are a characteristic glycine- and aspartate-rich repeat near the C terminus of the protein and a conserved type 1 secretion system (T1SS) that exports the protein into the extracellular environment, bypassing the periplasmic space (21, 37). The model protein for the RTX family, alpha-hemolysin, inserts into host membranes and forms pores that allow an influx of Ca2+ into host cells, altering host physiology or leading to cell death (37). Additional RTX family members have displayed a wide array of functions in bacterial pathogens; secreted proteases (24, 29) and lipases (9, 40), cross-linkers of cellular actin that cause host cell rounding (28), and surface-associated coats of protein that form the bacterial S-layer (26, 30) represent a few examples of these diverse functions. However, most RTX family members remain uncharacterized. Given the widespread distribution and diverse roles that known family members contribute to bacterial pathogenesis, the identification and characterization of novel RTX family members remains an important area of research.
One of the best characterized RTX family members, alpha-hemolysin, enhances host damage in the urinary tract during an Escherichia coli infection (32). In addition, this protein contributes to disseminated infections (5, 12), such as those that result from an ascending urinary tract infection (1, 6, 11). Indeed, it is estimated that Gram-negative bacilli, including E. coli, are the cause of up to 24% of cases of bacteremia annually in the United States alone (10). Despite the importance of this virulence determinant, not all uropathogenic E. coli (UPEC) strains produce this toxin (4, 18). However, additional uncharacterized RTX family members have been detected in the genomic sequences of uropathogens (21), raising the possibility that these proteins serve as alternatives to alpha-hemolysin.
The human pyelonephritis/urosepsis isolate E. coli CFT073 contains two annotated RTX family members: hlyA, encoding alpha-hemolysin, and tosA (originally annotated as upxA) (38). In silico analysis has previously indicated that tosA shares little homology with other characterized RTX family members (27) and might represent a novel virulence factor of UPEC. Subsequent analysis of the region surrounding tosA revealed that the gene is carried on a pathogenicity-associated island (PAI) and appears to be linked to genes encoding a T1SS composed of a tolC homolog and homologs of hlyB and hlyD (27). Our previous work (34) revealed that tosA is only expressed in the host environment and contributes significantly to the success of this UPEC strain in colonizing an animal model of an ascending urinary tract infection (UTI). However, it remains unclear whether this novel RTX protein functions in a similar way as alpha-hemolysin, or whether it confers a different set of advantages to the bacterium.
To explore this question, we characterized the expression of the genes in the region of the tosA locus and identified niches in the host that induce tosA expression in an animal model of ascending UTI. We then explored the function of TosA using in vitro tissue culture systems and in vivo models of UTI, bacteremia, and sepsis to cover the full range of environments UPEC encounter during their natural course of infection. Finally, we explored the use of purified TosA protein in an experimental animal vaccine model to protect against urosepsis. These results indicate that while the tos locus shares many features with the hly locus, TosA appears to play a distinct role in UPEC pathogenesis.
MATERIALS AND METHODS
Bacterial strains, plasmids, and primers.Strains and plasmids are listed in Table 1, and primers are presented in Table 2. E. coli CFT073 was isolated from blood and urine cultures from a patient with acute pyelonephritis (36). An isogenic strain lacking the gene tosA was already constructed by our group for use in a previous study (22). A plasmid containing the gene for green fluorescent protein (GFPmut3.1) under the constitutive em7 promoter, pGENmut3.1 (M. C. Lane, unpublished data), was electroporated into wild-type CFT073 cells for use in the murine model of ascending UTI.
Strains and plasmids used in this study
Primers used in this study
The pBAD-tosA plasmid was constructed by PCR amplifying the tosA gene using the following conditions: 1 μg of CFT073 genomic DNA was mixed with 4 μl of 2.5 mM deoxynucleoside triphosphate mixture, 5 μl of 5× GC Phusion buffer (NEB), 2 μl of 10 μM primer mixture, 1 U of Phusion polymerase in a 0.25-μl volume (NEB), and 12.75 μl of distilled water. PCRs were cycled according to the following protocol: (i) 98°C for 30 s, (ii) 98°C for 30 s, (iii) 64°C for 30 s, (iv) 72°C for 4 min, (v) repeat steps ii to iv 30 times, and finally (vi) 72°C for 5 min. PCR products were ligated into TOPO Blunt PCR cloning vector (Invitrogen) according to the manufacturer's instructions. A clone containing the correct insert was selected, and the plasmid was extracted and purified using a Qiagen plasmid miniprep kit and cut with BglII and EcoRI restriction enzymes (NEB). The DNA fragment containing the tosA sequence was purified after agarose gel electrophoresis using a Qiagen gel extraction kit and ligated overnight with a BglII/EcoRI (NEB)-cut pBAD/Myc-HisA vector (Invitrogen) using T4 DNA ligase (NEB) according to the manufacturer's instructions. Ligation products were electroporated into E. coli TOP10 (Invitrogen), and the resulting colonies were screened for protein expression after culture in Luria-Bertani (LB) medium supplemented with 10 mM l-arabinose at 37°C for 3 h (referred to as inducing conditions).
The araBP-tosC construct of CFT073 was engineered using the lambda red recombinase system (7) (Fig. 1A). Two PCR products were created; one included a cassette conferring kanamycin resistance amplified from the pKD4 plasmid, and the other from the pKD46 plasmid included the arabinose-inducible promoter. These two products were created such that the 5′ end of the pKD4 fragment and the 3′ end of the pKD46 product contained regions of homology to the genomic region upstream of tosC. The 3′ end of the pKD4 product contained a linking primer that was the reverse complement of a linking primer carried on the 5′ end of the pKD46 product. Each of the two PCR products was created (see above) in separate PCR products according to the PCR protocol outlined for pBAD-tosA vector creation, except that the extension step was modified to 60°C for 5 min. The resulting PCR products were gel purified as described above and combined in the same PCR to create a chimeric PCR product containing the two products linked end-to-end in a PCR as described above with the one modification. Ten thermal cycles (steps ii to iv above) were conducted without any primers to combine the two products. After 10 cycles, the cycling was paused, PCR primers corresponding to the 5′ end of pKD4 and the 3′ end of PKD46 fragments were added, and cycling was continued for an additional 30 cycles to amplify the combined product. Chimeric PCR products were purified using a DNA Clean & Concentrator kit (Zymo Research). Purified PCR products were electroporated into CFT073(pKD46), and standard protocols were followed to complete the Lambda Red recombination procedure (7). Clones were checked by Western blotting with TosA antiserum for protein production after growth in the arabinose-inducing conditions described above.
In vitro TosA expression constructs and putative tos operon structure. (A) Schematic illustrating construction of CFT073 araBP-tosC. A chimeric PCR construct consisting of two plasmid fragments generated from pKD4 and pKD46, containing a Kanr-conferring gene and an arabinose-inducible promoter, respectively, were amplified with primers that contain homology to the region between tosR and tosC (HR). A second set of primers, one the reverse complement of the other, were used as linking primers to combine the two PCR products together in a third PCR (LP). A schematic shows each ORF in the 15-kb region of tosA. The area denoted by brackets below genes, marked P1 to P4, illustrates amplicons that span adjacent genes. (B) Western blot of TosA induction. Cultures of CFT073 and CFT073 araBP-tosC were induced with 0.1 mM arabinose. At the indicated times, glucose was added to the cultures to give a final concentration of ca. 0.2%. Cultures were harvested at 3 h and processed for Western blot analysis. Proteins were reacted with TosA antiserum. Overexpression of TosA at 30, 60, and 90 min obscures the upper portion of these lanes. (C) RT-PCR of amplicons spanning the region between the two genes denote below each section. gDNA, genomic CFT073 DNA; (−) no RT, no-reverse-transcriptase control; cDNA, cDNA prepared from RNA from arabinose-induced bacteria. (D) qPCR results using RNA isolated from bacteria collected either from five mice transurethrally infected with wild-type CFT073 or from LB culture of wild-type CFT073 incubated at 37°C with aeration. Three replicates were normalized to the expression of gapA. In vivo, tosA expression was comparable to gapA expression (that is, well expressed), whereas in vitro expression of tos genes represented only a fraction of gapA expression. (E) TosA purified from the cytoplasmic fraction of arabinose-induced CFT073(pBAD-tosA) by gel filtration chromatography. A 0.8-μg portion of protein was loaded. M, protein standard markers; P, purified TosA. (F) Western blot with polyclonal TosA antiserum of wild-type (CFT073) and CFT073 araBP-tosC cells cultured under arabinose-inducing conditions. A total of 25 μl of late-exponential culture was loaded per lane; TosA marks the position just above the 250-kDa size standard. (G) Predicted molecular weights and amino acid identity to homologs for tos operon-encoded proteins. Homologs are as follows: PapB of Edwardsiella tarda (45% over 73 amino acids [aa]), TolC of Proteus mirabilis (44% over 392 amino acids [aa]), LssB family member of Neisseria sicca (62% over 707 aa), HlyD-like P. mirabilis (77% over 405 aa), outer membrane adhesin-like Shewanella woodyi protein (28% over 1,795 aa), LuxR/Sigma 70 family member of Citrobacter koseri (32% over 153 aa), and LuxR/UhpA family member of Vibrio campbelli (22% over 73 aa).
Reverse transcription-PCR (RT-PCR) and quantitative PCR (qPCR).RNA was isolated from bacteria collected from urine expelled from infected mice or from bacteria cultured in LB medium with aeration as described by Vigil et al. (34). Briefly, three groups of five CBA/J female mice each were transurethrally inoculated with 108 CFU of wild-type CFT073. Starting at 5 h postinfection, urine samples were collected from infected mice three times daily for 3 days. The samples were centrifuged at 13,000 rpm in a benchtop centrifuge for 1 min to pellet the bacteria, the supernatant was removed, and 20 μl of RNA Protect (Qiagen) was added to each pellet. Pellets were frozen at −80°C until later use. RNA was extracted using an RNeasy minikit (Qiagen), and cDNA was created with a SuperScript II reverse transcriptase kit (Invitrogen), according to the manufacturer's instructions.
RT-PCR was performed on the same cDNA preparation or on a no-reverse transcriptase control using primers that spanned the 3′ end of one gene to the 5′ end of the adjacent gene according to the above protocol. The PCR products were visualized on a 1.2% agarose gel following electrophoresis, stained with ethidium bromide, and visualized on a ChemiDoc imaging system (Bio-Rad).
Purification of TosA, generation of antiserum, and cell fractionation.For the purification of TosA, E. coli TOP10(pBAD-tosA) was cultured in LB medium with ampicillin (100 μg/ml) at 37°C with aeration (200 rpm) until reaching the early exponential phase. At this point, l-arabinose was added to a final concentration of 10 mM, and the cultures were incubated for an additional 3 h. Cells were pelleted by centrifugation (8,000 × g, 15 min, 4°C) and frozen at −80°C until purification was conducted. Pellets were thawed in 3 ml of phosphate-buffered saline (PBS) and passed three times through a French pressure cell press at 20,000 lb/in2. Cell lysates were centrifuged (112,000 × g, 30 min, 4°C) to pellet the unlysed cells and cell membrane material. The clarified supernatant was filtered through a 0.2-μm-pore-size syringe filter (Millipore). The lysate was immediately loaded onto a HiPrep Sephacryl-300 26/60 gel filtration column (GE Healthcare), and the column was run with PBS containing 0.5 M NaCl and 0.25 M urea. Collected fractions were analyzed by SDS-PAGE for the presence of TosA. Pooled fractions were concentrated using Centricon Plus-70 filter units with a 100-kDa cutoff and washed twice with 70 ml of Dulbecco PBS (pH 7.4; DPBS).
For generation of TosA antiserum, TosA was prepared by separating cell lysate from the pBAD-tosA, prepared as described above, using SDS-PAGE. Bands of ∼250 kDa, corresponding to TosA, were excised from multiple gels, pooled, and sent for commercial production of polyclonal TosA antiserum in rabbits (Rockland Immunochemical). Affinity purification of the resulting TosA antiserum was performed by cross-linking purified TosA to Ultralink Biosupport resin (Thermo Scientific) according to the protocols of the manufacturer.
Cell fractionation was performed on E. coli TOP10(pBAD-tosA) and the CFT073 araBP-tosC construct cultured in LB medium with ampicillin (100 μg/ml) and 1 mM calcium chloride or with kanamycin (25 μg/ml) and 1 mM calcium chloride, respectively. At the early exponential phase, each culture was treated with 0.1 mM l-arabinose for 40 min; after this induction period, glucose was added to a final concentration of 0.2%, and the cultures were incubated for 4 h postinduction. Bacteria were harvested by centrifugation (8,000 × g, 15 min, 4°C) and frozen at −20°C. The cells were thawed in 8 ml of PBS, lysed by passage through a French pressure cell press at 20,000 lb/in2, and centrifuged briefly (5,200 × g [CFT073 araBP-tosC] to 18,840 × g [pBAD-tosA], 10 min, 4°C) to remove unlysed cells, and the supernatant was centrifuged (112,000 × g, 45 min, 4°C). The supernatant, representing soluble cytoplasmic and periplasmic proteins, was carefully removed and stored; pellets, representing membrane fractions, were washed two to three times with PBS and subsequently treated with 1% Triton X-100 for 30 min at room temperature. The membranes were spun again as described above; the supernatant, enriched for inner membrane proteins, was collected and stored. The remaining pellet, enriched for outer membrane proteins, was washed two to three times, resuspended with PBS, and stored at −20°C. Filtered culture supernatant was collected and treated with ammonium sulfate to a final concentration of 3.40 M at 4°C. Precipitated protein was centrifuged (16,900 × g), and the protein pellet was resuspended in PBS. The resuspended material was dialyzed against two changes of PBS, at 4°C. Millipore centrifugal filter units with a 10,000-molecular-weight pore size were used to concentrate the dialyzed material. Concentrated culture supernatant, cytoplasmic/periplasmic, and membrane fractions, equivalent to 20 μg of total protein, were analyzed by Western blotting for the presence of TosA.
Limited proteolysis of the surface protein.The proteinase K assay was performed on the pBAD-tosA and araBP-tosC constructs cultured and induced under the same conditions as in the membrane fractionation assay, with the exception that each was cultured for 3 h postinduction. However, after these cells were pelleted and resuspended in 5 ml of PBS, this material was divided into four 1-ml treatments containing either 10, 50, or 500 μg of proteinase K or no proteinase K. Each treatment was digested for 1 h at 37°C; after this period, the reactions were stopped by the addition of phenylmethylsulfonyl fluoride to a final concentration of 500 μM. These cells were pelleted and washed three times with PBS before being stored at −20°C. From each of the whole-cell treatments, 20 μg of total protein (CFT073 araBP-tosC construct) or 8 μg of total protein (pBAD-tosA construct) was analyzed by Western blotting for the presence of TosA.
tosA expression.The tosA expression assay was conducted using CFT073 or the CFT073 araBP-tosC construct cultured in LB medium or in LB medium plus kanamycin (25 μg/ml), respectively. At the early exponential phase, each culture was treated with 0.1 mM l-arabinose and subsequently treated with glucose to a final concentration of 0.2%, at 15, 30, 60, or 90 min postinduction. All cultures were incubated for 3 h postinduction and, after this period, the cells were pelleted and frozen at −20°C. The equivalent of 50 μl from the most concentrated culture, as determined by the optical density at 600 nm (OD600), was analyzed by Western blotting for the presence of TosA in each treatment.
Animal models of infection.All mouse studies were approved by the University of Michigan Committee on the Use and Care of Animals. A previously described murine model of ascending UTI (14) was utilized as described below for immunocytochemistry experiments. A recently developed murine model of bacteremia (31) was used for nonlethal competition studies.
A zebrafish model of ExPEC pathogenesis as described by Wiles et al. (39) was used as a lethal sepsis model. Briefly, bacteria were prepared as described above for the murine model, and 1 nl of a suspension containing 1,000 CFU was microinjected into either the pericardial cavity or into the blood via the circulation valley of 48 h postfertilization zebrafish embryos. Cochallenge was conducted using equal mixtures of both strains and injecting 1,000 CFU total into each body site. To induce TosA synthesis, wild-type CFT073 and CFT073 araBP-tosC were cultured in minimal medium overnight, washed with sterile PBS, and suspended in PBS containing 50 mM l-arabinose 0.5 h prior to injection. Infection proceeded as described above, but zebrafish were maintained in water supplemented with 10 mM l-arabinose during the course of the experiment. Lethality was assessed at regular intervals for 2 days following infection. CFU counts were determined by homogenizing infected fish embryos in 500 μl of sterile PBS containing 0.5% Triton X-100 and plating serial dilutions on LB agar.
Cell adherence and intact bladder adherence assays.Adherence to cell lines cultured in vitro was measured using the following cell lines: Hs 769.T, UM-UC-3, MM55.k, Vero, and HEK293 (American Type Culture Collection) cultured in Dulbecco modified Eagle medium containing 10% fetal bovine serum (FBS), and SV Huc1 cultured in F12K containing 10% FBS and RT4 grown in McCoy's 5a medium containing 10% FBS (Invitrogen). All cell lines were cultured at 37°C and 5% CO2. Cells were cultured in sterile six-well plates until 90 to 95% confluence was reached and then washed with 1 ml of sterile DPBS. Wild-type and CFT073 araBP-tosC strains were cultured in LB medium supplemented with 10 mM l-arabinose for 3 h prior to infection, an OD600 reading taken, and the samples were diluted in sterile DPBS and used to inoculate washed mammalian cells at an average multiplicity of infection (MOI) of 0.6 (a low MOI was used to maximize sensitivity of the adherence assay and avoid saturation of receptors). The plates were centrifuged (500 × g, 5 min) and incubated at 37°C and 5% CO2 for 10 min. The cells were washed twice with 1 ml of sterile PBS and then lifted off in 1 ml of 0.9 mM EDTA–0.25% trypsin solution (Invitrogen). Serial dilutions of cell suspensions were spread onto LB agar plates, and CFU counts were obtained after overnight growth at 37°C. The input dilution of bacteria was also plated to determine the CFU count for each inoculum.
A modification of the murine model of ascending UTI described above was developed to test adherence of UPEC on intact bladder epithelium. Female C57BL/6 mice were transurethrally inoculated with 2 × 106 CFU of either wild-type CFT073 or CFT073 araBP-tosC cultured under the arabinose-inducing conditions described above in a total volume of 25 μl. At 30 min after inoculation, these mice were euthanized, and the bladders were removed and cut in half with a sterile scalpel blade. The bladders were washed in 1 ml of sterile DPBS for 5 min on a rotating microcentrifuge mixer under low speed, transferred to new tubes, and washed a second time to remove nonadherent bacteria. Washed bladders were transferred to 3 ml of sterile PBS, homogenized, and plated on LB agar plates for the determination of CFU. Recovered bacteria were compared to CFU counts of the inoculum.
Immunogold-transmission electron microscopy.Immunogold labeling of intact bacterium was carried out by culturing wild-type CFT073, E. coli TOP10 pBAD-tosA, and CFT073 araBP-tosC cells in LB medium supplemented with 10 mM arabinose for 3 h until late exponential phase of growth. Bacterial suspension (10 μl) was spotted on nickel coated Formvar/carbon film nickel-coated TEM grids (EMS). After 15 min, the liquid was wicked away with filter paper and blocked with 10 μl of DPBS containing 5% goat normal serum and 5% bovine serum albumin (BSA) for 15 min. Blocking solution was exchanged with 10 μl of TosA antiserum diluted 1:250 in DPBS with 5% goat serum, 0.1% cold water fish skin gelatin, and 0.1% BSA-c (EMS). After 15 min, excess fluid was wicked away with filter paper and exchanged for 10 μl of incubation solution for 5 min. The wash was repeated and then exchanged with 10 μl of goat anti-rabbit IgG conjugated with 10-nm gold particles (EMS) diluted 1:250 in incubation solution. After 15 min, grids were washed twice with incubation solution and twice with distilled water. Grids were air dried and imaged on a Philips CM-100 transmission electron microscope. Grids for negative staining were incubated with 10 μl of 1% uranyl acetate for 3 min.
Immunofluorescent imaging.Female C57BL/6 mice were transurethrally inoculated with 108 CFU of CFT073(pGENmut3.1) as described above. At 48 h postinoculation, the mice were euthanized, and the bladder, kidneys, and spleen were removed from each mouse and fixed in 10% buffered formalin for 4 h. Organs were processed through a sucrose gradient and frozen in OCT (Andwin Scientific). Frozen sections (10 μm thick) were cut from each organ.
Tissue sections were blocked in DPBS containing 5% goat normal serum and 5% donkey normal serum for 15 min. TosA antiserum and chicken antiserum to GFP (Abcam) were each diluted 1:500 in blocking solution and incubated for 30 min, followed by two washes in DPBS for 10 min. Anti-rabbit IgG Dylight 549 and anti-chicken IgY Dylight 488 (Jackson Immunoresearch) were diluted 1:1,000 and 1:500, respectively, in blocking solution and 1 U of Alexa Fluor 647-phalloidin conjugate was added. Tissue sections were incubated in this solution for 30 min and then washed two times in DPBS. Sections were mounted with ProLong Gold antifade solution (Invitrogen) and imaged on a Zeiss LSM-510 META confocal laser scanning microscope.
Vaccination model.Purified TosA was diluted with DPBS containing 10% Imject Alum (Thermo Scientific) and administered via subcutaneous injection to deliver 100 μg of protein in a 100-μl volume. Control mice were given a solution of DPBS containing 10% alum. On the same day, retro-orbital eye bleeds were collected from each animal, and the serum was separated and stored at −20°C until further analysis. At 1 week and 2 weeks after primary vaccination, the animals were boosted with 25 μg of protein in 100 μl of DPBS containing 5% alum. Control mice were vaccinated with DPBS containing 5% alum. At 1 week after the second boost, 3 weeks after the primary vaccination, retro-orbital eye bleeds were collected, and the animals were challenged with wild-type CFT073. UTI challenge was performed according to the transurethral model, and bacteremia challenge was performed according to the protocols outlined above.
Data analysis.Statistical tests were carried out in the Prism statistical software package (GraphPad Software). Cell adherence data were analyzed by using a Student t test. Cochallenge data were analyzed by calculating the log10 transformation of competitive index (mutant CFU/wild-type CFU) and performing a Wilcoxon rank-sum test with a hypothetical median of 0. Survival curves were analyzed by using the Mantel-Cox test.
RESULTS
The tos locus of E. coli CFT073 includes the genes for a RTX family member, a T1SS, and putative regulators.The gene content and organization of the 15-kb region of the CFT073 genome containing tosA (Fig. 1A) is contained within the PAICFT073-aspV pathogenicity-associated island (22, 23, 38). This genomic island had previously been implicated in enhancing the fitness of UPEC in a murine model of an ascending UTI (22) by contributing significantly to the virulence of UPEC strain CFT073 in both bladder and kidney tissues of infected mice (34). The putative type 1 secretion system (T1SS) genes, tosCBD, were identified in a previous in silico analysis (27) as components of a conserved T1SS apparatus predicted to mediate the export of an RTX family member (Fig. 1A). tosC is predicted to encode a homolog of the outer membrane protein TolC; TosB and TosD are predicted to form the inner-membrane ABC transporter and the membrane fusion protein of the T1SS, respectively (Fig. 1G). Consistent with this predicted mode of export, TosA does not carry an identifiable cleavable leader peptide. Two additional open reading frames (ORF), designated c0364 and c0365, follow the putative RTX family member tosA in the same orientation. In addition, our analysis uncovered a previously unannotated ORF, designated tosR, directly upstream of the start codon of tosC (52 nucleotides separate the two genes) (Fig. 1A); the ORF has homology to papB, a regulator of the P fimbrial operon (2) (Fig. 1G). The tos locus is flanked by fragments of genes predicted to encode remnants of a transposon system, possibly indicating that this region was acquired by horizontal gene transfer.
All ORFs in the tos locus are expressed at significant levels in vivo (similar to gapA), whereas their expression is very low (compared to gapA) during in vitro culture (Fig. 1D). qPCR conducted on mRNA purified from bacteria voided in the urine of UPEC-infected mice verified that significant tosA expression was limited to the in vivo environment (34).
To extend this analysis, a third set of probes was used to assess the transcriptional organization of the tos locus (Fig. 1A). These data suggest a possible operon structure for these genes. RT-PCR on extracted RNA using primers that span the region from the 3′ end of one gene to the 5′ end of the adjacent gene (Fig. 1C) indicated that the genes tosR through tosA are transcribed as part of a continuous transcript. Although qPCR was sufficiently sensitive to detect expression of c0364 and c0365 in vivo (Fig. 1D), RT-PCR was not able to detect these genes as a part of the tos transcript (data not shown). Thus, there is insufficient evidence at present to include c0364 and c0365 in the operon. This result is consistent with an operon structure consisting of tosRCBDA, but additional experiments are required to define the full extent of the tos operon.
tosA expression is limited to the host in vivo environment.E. coli progresses through the host urinary tract in an ascending manner (1, 6, 11), encountering a variety of environments in the process. To determine whether the tosA expression observed in Fig. 1D was limited to a particular site in the urinary tract, we utilized a murine model of an ascending UTI. At 48 h after transurethral inoculation, the bladders, kidneys, and spleens were removed and processed for immunofluorescence microscopy. A plasmid that constitutively expresses GFP was utilized to mark wild-type CFT073 bacteria and TosA antiserum was used to stain bacteria expressing TosA. Tissue sections of the kidney (Fig. 2A) and spleen (Fig. 2B) demonstrated that UPEC expresses TosA protein in each organ at 48 h postinoculation. In addition, in separate experiments, our murine bacteremia model reveals that TosA is expressed at 24 h postinoculation in spleen and liver tissue (Fig. 2C and D, respectively), two sites where UPEC was previously determined to display enhanced fitness and survival over nonpathogenic E. coli (31). Although the results of these two experiments suggest that TosA expression is important for UPEC during both ascending UTI and disseminated infections, it will be important to examine tissue from mice infected with the ΔtosA mutant.
TosA is expressed in infected bladder, kidney, spleen, and liver. Female C57/Bl6 mice inoculated transurethrally (A and B) or via tail vein injection (C and D) with CFT073(pGENmut3.1). At 48 h postinoculation, the kidneys (A) and spleens (B) and at 24 h postinfection the spleens (C) and livers (D) were removed and processed for immunofluorescence microscopy. GFP antiserum staining is shown in green, TosA antiserum in red, phalloidin staining in white, and DAPI in blue. Measurements of individual bacteria are denoted by scale bars.
Genetic manipulation of strain CFT073 allows in vitro expression of TosA.Because significant tosA expression is primarily limited to the in vivo environment (34), two constructs allowing inducible expression of tosA were engineered to allow additional experimentation. A plasmid-based system, pBAD-tosA, that carries the cloned tosA gene under an arabinose-inducible promoter, but which lacks the T1SS genes and accessory gene content illustrated in Fig. 1A, allows in vitro TosA production and purification (Fig. 1E). However, the TosA produced remains primarily confined to the soluble cytoplasmic fraction of bacteria (see below).
To construct a strain that expresses tosA in vitro and localizes the protein to its wild-type location, we used the Lambda Red recombinase system and a hybrid PCR product containing an arabinose-inducible promoter to recombine into the chromosome of CFT073, placing the araB promoter sequence upstream of the start codon of tosC (Fig. 1A). Upon arabinose induction, this araBP-tosC construct produces TosA protein that is detectable using polyclonal TosA-antiserum (Fig. 1B and F). In contrast, wild-type CFT073 cultured under identical conditions does not produce detectable levels of TosA (Fig. 1F).
TosA protein localizes to the outer surface of E. coli.Immunogold labeling of arabinose-induced CFT073araBP-tosC bacteria expressing TosA, imaged by transmission electron microscopy, localizes TosA to the outer surface of the bacterium. When wild-type CFT073 (Fig. 3A) and CFT073araBP-tosC (Fig. 3B) were cultured in vitro under arabinose-inducing conditions, only CFT073araBP-tosC bacteria expressing TosA were marked with immunogold particles (>35 beads visible), while wild-type CFT073 remained unlabeled (no beads visible). Negative staining of these bacteria revealed that the immunogold particles marking TosA localize to the surface of the bacterium (data not shown). Finally, pBAD-tosA bacteria cultured under inducing conditions showed no immunogold staining (no beads visible) (Fig. 3C), confirming that TosA requires a specific transport mechanism, which is not found in the E. coli TOP10 strain that carries the pBAD-tosA plasmid, to be exported out of the cytoplasm. While the specific transport system is believed to be encoded by tosCBD, additional experiments will be required to confirm this.
TosA localizes to the outer membrane. (A to C) Immunogold-TEM micrographs of TosA antiserum-stained bacteria. Bacteria were cultured under arabinose-inducing conditions. (A) CFT073 ΔtosA show no immunogold particles. (B) CFT073 araBP-tosC marked with numerous immunogold particles. (C) TOP10(pBAD-tosA) cells show no immunogold staining. All images were acquired at ×64,000 magnification; scale bars denote 100 nm.
When arabinose-induced E. coli CFT073 araBP-tosC, which expresses the entire tos operon, was subjected to cell fractionation into cytosol, inner-membrane, outer-membrane, and supernatant fractions, TosA localized to the outer-membrane fraction (Fig. 4A). TosA was not observed in uninduced bacterial fractions. When TosA was arabinose-induced from pBAD-tosA, the soluble fraction contained the majority of the detectable TosA. An equivalent volume of concentrated, filtered culture supernatant from both cultures indicated that, in contrast to the majority of known RTX family members (21), TosA is not secreted extracellularly in significant amounts from the bacterium (Fig. 4A). Surface-expressed TosA from arabinose-induced E. coli CFT073 araBP-tosC was susceptible to proteinase K digestion, whereas cytoplasmic TosA from arabinose-induced E. coli TOP10(pBAD-tosA) was not degraded by the protease treatment (Fig. 4B).
Cellular localization and protease susceptibility of TosA. (A) The indicated bacterial constructs were induced or not with arabinose and fractionated as described in Materials and Methods. Samples were subjected to SDS-PAGE and Western blotting with TosA antiserum. Fractions: Cyt, cytosol; IM, inner membrane; OM, outer membrane; Sup, supernatant. (B) The indicated constructs were induced with arabinose as described in Materials and Methods. Whole bacterial cells in suspension were incubated for 1 h with the indicated amount of proteinase K (Protease). Bacteria were treated with SDS-gel sample buffer, electrophoresed, and subjected to Western blotting with TosA antiserum. Blank lanes in the lower right panel contained no sample.
tosA expression enhances adherence to the epithelial cells that line the urinary tract.Given the surface localization of TosA, we reasoned that this protein might mediate adherence to epithelial cells that line the host urinary tract. When wild-type CFT073 and araBP-tosC bacteria, both cultured in arabinose, were incubated with a murine kidney epithelial cell line MM55.K in vitro, 5.4% of wild-type CFT073 bacteria in the inoculum adhered to the epithelial cells after a wash with PBS. Bacteria expressing TosA, however, adhered under these conditions at nearly three times this level (15.4% of the inoculum) (Fig. 5A, white bars), suggesting that this novel RTX family member mediates adherence to epithelial cells that line the host urinary tract. No cytopathic effects were observed for these cultured cells or any other cell line in the present study exposed to TosA-expressing bacteria.
TosA mediates adherence to epithelial cells of the upper urinary tract. Wild-type CFT073 (WT) and CFT073 araBP-tosC (TosA), cultured under arabinose-inducing conditions, were used to test adherence to confluent monolayers of epithelial cells in vitro at an MOI averaging 0.6. After incubation and washing, the cells were lifted off and plated for CFU determination. The data are expressed as the percentage of the inoculum that was adherent to the monolayer. (A) MM55.K cell adherence assay conducted with wild type (WT) and TosA (white bars) and repeated after cells were pretreated for 5 min with TosA antiserum (black bars). (B) Adherence assay conducted with seven distinct cell lines derived from throughout the urinary tract and expressed as the percentage of inoculum that was adherent to the monolayer. Raw data are supplied in Table S1 in the supplemental material.*, P < 0.05 as calculated by a Student t test.
The adherence assay was repeated with wild-type and TosA-expressing bacteria that had been preincubated for 5 min with TosA antiserum. The results of the adherence assay showed an ablation of the TosA-mediated increase in adherence back to wild-type levels but had little impact on wild-type bacteria treated with the same antiserum (Fig. 5A, black bars). Given the lack of reactivity against wild-type bacteria cultured under identical conditions (Fig. 1F), we expected that only bacteria expressing TosA would be impacted by the TosA antiserum pretreatment. To confirm that bacteria treated with TosA antiserum were viable and that this result was not the result of complement-mediated lysis of TosA-producing bacteria, bacterial suspensions were treated with TosA antiserum and plated to assess viability. No significant difference was observed in viability between the number of wild-type bacteria (91% ± 3% of cells viable) and bacteria expressing TosA (89% ± 2% of cells viable). This result confirmed that the increase in adhesion was specific to the production of TosA protein.
The adherence assay described above was repeated with different cell lines that together represent the epithelial cell types that line the entire host urinary tract. UPEC adhered strongly to cell lines derived from a human transitional cell carcinoma of the urethra (Hs 769.T) and at lower levels in two cell lines derived from a human transitional cell carcinoma and a transitional cell papilloma of the bladder (UM-UC-3 and RT4, respectively). However, no differences were observed in adherence levels between wild-type CFT073 and CFT073 araBP-tosC cells expressing TosA (Fig. 5B). Thus, wild-type CFT073, a prototypical human pyelonephritis/urosepsis isolate (38), adhered to cells derived from the epithelial lining of the lower urinary tract efficiently without the need for TosA.
In contrast, bacteria expressing TosA adhered in significantly higher numbers than wild-type CFT073 to all tested cell lines derived from the upper urinary tract. A transformed transitional cell epithelial line derived from a normal adult human specimen (SV Huc1) allowed wild-type bacteria to adhere in high numbers, but a significant increase in adherence was observed when TosA was expressed. Epithelial cells derived from embryonic human kidney tissue (HEK293) showed a modest but significant increase of adhesion with TosA expression, and an even greater increase was observed for a transformed adult primate kidney epithelial cell line (Vero) (Fig. 5B).
To confirm that the result obtained in cell lines derived from the lower urinary tract under in vitro conditions (no contribution of TosA to adherence) was relevant during the course of infection, a similar adherence assay was developed by modifying an established murine model of ascending UTI (14). At 30 min after transurethral inoculation of either 108 CFU wild-type or araBP-tosC bacteria induced with arabinose, the mice were euthanized, and the bladders were removed, cut in half, and weighed. The bladder tissue was subjected to a PBS washing in a fashion similar to the in vitro adherence assays to remove nonadherent bacteria. The number of bacteria that remained adherent demonstrated no significant difference between adherence levels of wild-type (0.24% of the inoculum) and TosA-producing (0.17% of the inoculum) bacteria (P > 0.1). Taken together with the previous results, this indicates that TosA can mediate adherence to the epithelial cell lining of the upper urinary tract but does not contribute significantly to the adherence of UPEC in the urethra or bladders of infected hosts.
tosA enhances fitness in two animal models of bacteremia and sepsis.Due to the fact that tosA was discovered in an urosepsis isolate and is expressed in the liver and spleen during bacteremia (Fig. 2C and D), the possibility that the protein has a secondary role during bacteremia and sepsis was investigated using two different animal models of extraintestinal pathogenic E. coli (ExPEC) pathogenesis. First, a ΔtosA strain that had been previously demonstrated to have reduced fitness in the murine model of ascending UTI during cochallenge (22) and during independent challenge (34) was tested in our recently developed murine model of bacteremia (31). Our group developed this model to assess the impact of known and putative UPEC virulence factors in the success of bacteria that spread from the kidneys to distant organ sites. We found that many genes that confer an advantage in the urinary tract also confer a fitness advantage in this model. Using this model, CFT073 ΔtosA was recovered in significantly lower CFU in both liver (8 of 8 mice) and spleen (6 of 8 mice) tissue at 24 h postinfection with competitive indices of 0.17 (P = 0.008) and 0.28 (P = 0.023), respectively (Fig. 6), suggesting a secondary role for this gene during bacteremia.
TosA enhances fitness during bacteremia. A murine model of bacteremia was used to compete wild-type CFT073 against CFT073 ΔtosA. An equal mixture of both strains was diluted to deliver 106 CFU in 100 μl, which was injected in the tail vein of female C57BL/6 mice. At 24 h postinoculation, liver and spleen tissue was removed, and the CFU of each strain were determined. The log10 competitive index (mutant CFU/wild-type CFU) is shown, with lines representing median values. P values are indicated on the graph.
A second animal model using zebrafish to study ExPEC pathogenesis during sepsis (39) was also used to study the same ΔtosA strain. In the zebrafish model, wild-type bacteria were recovered in numbers equal to those for ΔtosA bacteria during cochallenge in the bloodstream and when injected into the pericardial cavity (Fig. 7A). During independent challenge in the bloodstream, identical survival curves were obtained for the two strains (Fig. 7B), indicating that in this model system tosA does not confer an advantage to bacterial survival.
TosA enhances lethality in a zebrafish model of ExPEC pathogenesis. At 48 h postfertilization zebrafish embryos were microinjected with bacteria into either the pericardial cavity around the developing heart or into the circulation valley to initiate bacteremia and sepsis. (A) Wild-type (WT) CFT073 and CFT073 ΔtosA were mixed in equal proportions and inoculated into the pericardial cavity (PC) or the circulation valley (blood) with 1,000 CFU in 1 nl. At 16 h postinoculation, embryos were homogenized, and the CFU load of each strain was determined. The data are expressed as competitive indexes (mutant CFU/wild-type CFU), which were not significantly different. (B) Survival curve of zebrafish embryos injected in the circulation valley with 1,000 CFU of either wild-type CFT073 or CFT073 ΔtosA. (C) Immunofluorescence microscopy of the pericardial cavity (PC) or tails (blood) of embryos injected with wild-type CFT073 or araBP-tosC bacteria cultured under arabinose-inducing conditions. Fish were maintained in water supplemented with arabinose and, 16 h postinoculation, fixed and processed for imaging. Anti-E. coli antibody staining marks bacteria red and TosA antiserum staining is displayed in green. The merged files in the right panels appear green (and not yellow) due to the far more intense anti-TosA staining than the anti-E. coli staining. Scale bars, 100 μm. (D) Survival curves of fish infected with CFT073 araBP-tosC, incubated with or without arabinose induction. Mock-infected embryos were incubated in the presence of arabinose. n = 40 embryos for each survival curve.
The discrepancy between the two animal models was resolved by staining zebrafish infected with wild-type CFT073 for TosA and imaging by immunofluorescence microscopy. In mice, this strain expresses TosA in the spleen following transurethral inoculation and in the spleen and liver following intravenous injection (Fig. 2), but no detectable TosA protein was observed on bacteria either in the pericardial space of the fish (Fig. 7C, upper left panel) or on bacteria that had disseminated through the circulatory system (Fig. 7C, lower left panel).
In fish that had been infected with the araBP-tosC strain, the bacteria produced TosA protein if the fish were maintained in water supplemented with arabinose (Fig. 7C, right panels) and allowed this model system to be utilized with this construct of CFT073. When arabinose-induced araBP-tosC bacteria were inoculated directly into the pericardial cavity (Fig. 7C, PC) or circulation valley (Fig. 7C, blood) of zebrafish that were maintained in water supplemented with arabinose, TosA-expressing bacteria were uniformly observed (Fig. 7C, right panels) and nearly 80% of infected fish died by 18 h postinoculation (Fig. 7D). Fish infected with the wild-type strain showed E. coli bacteria but no TosA expression (Fig. 7C, left panels) and took nearly 36 h to reach this level of mortality (P < 0.01) (Fig. 7D). When zebrafish were maintained in arabinose-supplemented water and mock infected with PBS supplemented with arabinose, >80% of the fish survived to 48 h (Fig. 7D), indicating that the arabinose treatment per se did not contribute to the enhanced killing kinetics observed with TosA expression.
Vaccination against TosA confers protection against urosepsis caused by a human pyelonephritis UPEC isolate.In the final set of experiments, we determined that TosA protein contributes to the ability of UPEC to spread from the kidneys to distant organ sites. TosA was purified from the pBAD-tosA system, mixed with alum as adjuvant and administered to mice by subcutaneous injections following the vaccination schedule outlined in Fig. 8A. Vaccinated mice were transurethrally inoculated with wild-type CFT073, and the CFU in the bladder, kidney, and spleen were quantified and compared to the CFU of mock-vaccinated mice that only received PBS-alum injections. Vaccination had no impact on bacterial survival in the urinary tract, but a significant decrease in CFU/g of spleen tissue was observed at 48 h postinfection (P = 0.004) (Fig. 8B). The vaccination protocol was repeated, and mice were challenged by intravenous injection; the CFU levels in the spleen and liver were determined at 24 h postinoculation (Fig. 8C). In this model, no protection was observed among vaccinated mice (data not shown), suggesting that the protection observed in the previous vaccination was specific to bacteria that ascended the urinary tract and later disseminated into the bloodstream.
TosA vaccination protects against urosepsis. (A) Purified TosA was utilized in an experimental vaccine model according to the depicted schedule: female C57/Bl6 mice were vaccinated with 100 μg of protein and then boosted one and 2 weeks later with 25 μg. Three weeks after primary vaccination the mice were challenged with wild-type CFT073 via transurethral infection (B) or tail vein injection (C). The CFU/g of tissue was determined for each organ site at 48 h postinfection (B) or 24 h postinfection (C) and compared to mice mock-vaccinated with PBS. *, P = 0.004.
DISCUSSION
The gene sequence of tosA had previously been annotated as a putative RTX family exoprotein based on its association with the conserved T1SS (27) and the presence of characteristic glycine- and aspartate-rich repeats in the C terminus of the protein (21). Our group has independently identified nine repeats (Fig. 9) that match closely to the GGXGXD consensus RTX sequence (21) in the C terminus of the protein. In addition, a novel repeat was discovered in tosA consisting of five large tandem repeats ∼1 kb in length (Fig. 9).
TosA contains characteristic RTX repeats and a novel repeat structure. (A) DNA sequences of hlyA and tosA from the CFT073 genome were searched for the RTX repeat structure GGXGXD using PATTINPROT software program set to find regions with 75% identity to the consensus RTX sequence. The repeats were aligned and used to create sequence logos using WebLogo software (http://weblogo.berkeley.edu/logo.cgi). The height of the letters at each position indicates the level of conservation at that amino acid position (http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=npsa_pattinprot.html). HlyA, top panel; TosA, bottom panel. Twelve repeats were found for hlyA, and nine were found for tosA. (B) The CFT073 genome annotation originally described tosA (first annotated as upxA) as a 5,325-nucleotide ORF encoding a putative protein 1610 amino acids long (predicted molecular mass, 163,819 Da). Sequence analysis identified two tandem repeats in the nucleotide sequence, depicted in panel as gray bars underneath a schematic of tosA, which shows the repeats relative location proximal to the 5′ of the gene and N terminus of the predicted protein. (C) Cloning and expression of tosA indicate the gene is actually ∼8,000 nucleotides in length and encodes a protein slightly larger than 250,000 Da (see Fig. 1A). Restriction enzyme mapping of full tosA PCR products is consistent with the presence of five of these repeats in a tandem array (data not shown). (D) Amino acid sequence of the tandem repeats is 335 residues long.
The strict regulation of tosA expression (only in vivo induction) appears to be unique among RTX family members. Little is known about the regulation of RTX operons beyond the regulatory mechanisms defined for alpha-hemolysin (8, 15, 19). CFT073 displays hemolytic activity during in vitro growth (25) but does not express detectable levels of TosA (Fig. 1 and 4). This suggests that different regulation mechanisms operate on the two RTX systems present in this strain. Furthermore, the well-defined RTX operons, conferring diverse functions ranging from the pore-forming toxins to secreted proteases, do not appear to contain genes that potentially regulate gene expression (21). The presence of tosR, a gene with homology to the papB family of fimbrial gene expression regulators (17, 41), which is immediately upstream of a novel UPEC adherence factor gene, remains unexplained. Attempts to delete tosR met with no recombinants, however, and a definitive function for this gene has yet to be elucidated. One attractive hypothesis is that the UPEC, which is known to utilize papB family members to coordinate expression of the multitude of fimbriae encoded in the genomes of UPEC (16, 20, 42), adapted the same network of regulators to allow tosA expression to respond in concert with changes in expression of other adherence factors. Further work will be required to determine whether TosR is a PapB family member and determine its effects on the expression of other adherence factors. In addition, while RNA transcripts corresponding to ORFs c0364 and c0365 were detected, no function has yet been assigned to these sequences.
Surface localization of TosA also sets this RTX family member apart from other RTX proteins. Only the RTX members that comprise the S layer of certain bacteria, notably the motility-associated RTX of Cyanobacteria (3) and the RTX surface coat of Campylobacter spp., which protects against immune system attack (26), are known to remain associated with the bacterium. All other known RTX members are fully secreted into the external milieu. However, many RTX proteins can bind to host cell membranes, some with selectivity (33). The only adaptation that would be required to convert such a receptor-specific RTX into an adherence factor would be the incorporation of the protein into the outer membrane. Previous in silico analysis identified a putative transmembrane spanning region in the N terminus of TosA (27). However, additional study is required to determine whether this region allows TosA to embed in the outer membrane.
Three lines of additional evidence also support a possible secondary role for TosA, which is to mediate survival during bacteremia and urosepsis. First, a ΔtosA strain of CFT073 demonstrated reduced fitness when competed against wild-type bacteria in our murine model of bacteremia (Fig. 6). This model has determined that many UPEC virulence factors, including many nontoxins, enhance fitness in the host (31). Second, the protection observed in TosA-vaccinated mice from the final stage of an ascending UTI (Fig. 8B), urosepsis, may indicate that TosA plays a role or, at least, is expressed during the transition from the urinary tract to the bloodstream. Finally, the enhanced lethality of TosA-expressing bacteria in the zebrafish sepsis model (Fig. 7D) argues for an important role during disseminated infections. However, these data do not specifically address whether the protein has a direct role in host toxicity. Several attempts have been made to observe such a role, including incubating TosA-expressing bacteria or purified TosA with kidney epithelial cells or sheep erythrocytes in vitro, have failed to provide evidence for a direct role in cytotoxicity (data not shown). Given the results at hand, TosA may only be enhancing host damage and survival indirectly by allowing bacteria expressing the protein to bind to target cells, enhancing delivery of other secreted products including alpha-hemolysin and the autotransporter proteases.
A limitation of our virulence studies has been the lack of complementation. We speculate that difficulties arise because of the large size of tosA (8 kb) and the presence of multiple tandem repeat-in-toxin sequences within tosA, possibly leading to deletions during cloning. However, we have constructed independent deletion mutants that consistently demonstrate severe attenuation in the murine model.
The experimental vaccination results (Fig. 8) raise the possibility that a novel therapeutic agent could be developed to target TosA during urosepsis. This model utilizes an earlier endpoint than the bacteremia UTI model (24 versus 48 h). It is possible that the host immune system has not had sufficient time to mount an effective response in the short time frame. It is also possible that the difference in results between the two models was due to the manner in which bacteria entered the bloodstream. For the bacteremia model, bacteria were cultured in vitro, a condition where tosA is not expressed (34), and injected all at once. In contrast, it is expected that movement from the urinary tract into the bloodstream occurred throughout the course of the infection and, given the results of the present study, the bacteria in this model were already expressing TosA. This may have presented the immune system of vaccinated animals with two very different challenges (TosA-expressing and TosA-nonexpressing bacteria) and could explain the difference in the two vaccination experiments.
Regardless of the differences observed in the two vaccination models, two additional observations suggest that further development of anti-TosA therapeutics may prove beneficial. In a previous study, our group identified that bacteria flushed from the urinary tract in voided urine from women with UTI show detectable transcript levels of tosA (13), indicating that the restricted expression of tosA may include the human urinary tract. Finally, in a recent survey conducted by our group of urinary tract and fecal isolates of E. coli, we discovered that tosA is present in ∼25% of uncomplicated UTI isolates, but rare in fecal isolates. A total of 98% of strains that carry tosA were part of the B2 phylogenetic lineage of E. coli, and most belonged to a subgroup that were enriched for highly pathogenic E. coli (35). Thus, while not all UPEC express this protein, TosA may still be an attractive target for the development of multivalent vaccines to combat the development of dissemination to the bloodstream secondary to uncomplicated UTIs.
ACKNOWLEDGMENTS
This study supported in part by Public Health Service grants AI43363 and AI059722.
We thank Sara Smith for expert assistance with animal studies.
FOOTNOTES
- Received 2 August 2011.
- Returned for modification 15 September 2011.
- Accepted 7 November 2011.
- Accepted manuscript posted online 14 November 2011.
- Address correspondence to Harry L. T. Mobley, hmobley{at}umich.edu.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/IAI.05713-11.
REFERENCES
- Copyright © 2012, American Society for Microbiology. All Rights Reserved.
