Functional analysis of antigen 43 in uropathogenic Escherichia coli reveals a role in long-term persistence in the urinary tract - PubMed (original) (raw)
Functional analysis of antigen 43 in uropathogenic Escherichia coli reveals a role in long-term persistence in the urinary tract
Glen C Ulett et al. Infect Immun. 2007 Jul.
Abstract
Escherichia coli is the primary cause of urinary tract infection (UTI) in the developed world. The major factors associated with the virulence of uropathogenic E. coli (UPEC) are fimbrial adhesins, which mediate specific attachment to host receptors and trigger innate host responses. Another group of adhesins is represented by the autotransporter subgroup of proteins. The best characterized of these proteins, antigen 43 (Ag43), is a self-recognizing adhesin that is associated with cell aggregation and biofilm formation in E. coli K-12. The sequenced genome of prototype UPEC strain CFT073 contains two variant Ag43-encoding genes located on pathogenicity islands. The biological significance of both of these genes and their role in UPEC pathogenesis have not been investigated previously. Here we performed a detailed molecular characterization analysis of Ag43a (c3655) and Ag43b (c1273) from UPEC CFT073. Expression of Ag43a and Ag43b in a K-12 background revealed that they possess different functional properties. Ag43a produced a strong aggregation phenotype and promoted significant biofilm growth. Deletion mutants and strains constitutively expressing Ag43a and Ag43b were also constructed using CFT073. When these mutants were analyzed in a mouse model of UTI, Ag43a (but not Ag43b) promoted long-term persistence in the urinary bladder. Our findings demonstrate that Ag43a contributes to UPEC disease pathogenesis and reveal that there are pathogenicity-adapted variants of Ag43 with distinct virulence-related functions.
Figures
FIG. 1.
(A) Coomassie brilliant blue-stained SDS-PAGE gel demonstrating Ag43 expression in E. coli MG1655 Δ_flu_ strains, including MS1231 (pCO3, Ag43b), MS1232 (pCO4, Ag43a), and MS1230 (pCO2, Ag43 K-12). Production of each Ag43 variant was induced by 0.2% arabinose. Protein preparations were obtained by subjecting cells to a brief heat treatment, centrifugation, and purification of heat-released proteins by TCA precipitation. Lane M contained markers. (B) Western blot analysis using a serum directed against the alpha-domain of Ag43 from E. coli K-12. Expression of Ag43b (MS1231), Ag43a (MS1232), and Ag43 from K-12 (MS1230) was detected only following induction with 0.2% arabinose. The serum reacted most strongly against the K-12 Ag43 alpha-domain subunit; the Ag43a and Ag43b variants reacted more weakly, but the levels for these variants were comparable. The asterisk indicates specific degradation products of the different Ag43 variants.
FIG. 2.
(A) Autoaggregation assay demonstrating the settling profiles for liquid suspensions of E. coli OS56 (MG1655Δ_flu_, Gfp+) containing the following plasmids: (i) pBADMycHisA (MS1199), (ii) pCO2 (K-12 Ag43, MS1230), (iii) pCO3 (Ag43b, MS1231), and (iv) pCO4 (Ag43a, MS1232). All strains were grown in the presence of 0.2% arabinose to induce Ag43 expression. Expression of Ag43 led to significant autoaggregation of MS1230, MS1231, and MS1232 compared to the autoaggregation of control strain MS1199 (at 60 to 120 min, P < 0.05 for each strain, as determined by the independent sample t test). Furthermore, MS1232 cells expressing Ag43a exhibited significantly stronger autoaggregation than their MS1231 Ag43b-expressing counterparts (at 60 to 120 min, P < 0.002, as determined by the independent sample t test). (B) Same assay in the presence of type 1 fimbria expression. MS1199, MS1230, MS1231, and MS1232 were transformed with pHHA13 (which encodes the ability to express type 1 fimbriae) and grown in the presence of 0.2% arabinose as described above to induce Ag43 expression. The presence of type 1 fimbriae on the cell surface resulted in prevention of Ag43-mediated aggregation. (C) Western blot of Ag43 alpha-domain subunit proteins liberated from E. coli MS1231 (Ag43b+), MS1232 (Ag43a+), MS1231(pHHA13) (Ag43b+ Fim+), and MS1232(pHHA13) (Ag43a+ Fim+). Overexpression of type 1 fimbriae did not affect the amount of Ag43 present at the cell surface of each strain.
FIG. 3.
Biofilm formation by E. coli MG1655Δ_flu_ cells harboring plasmids expressing Ag43a and Ag43b. The effect of each Ag43 variant on biofilm formation was assessed in E. coli OS56 (MG1655Δ_flu_, Gfp+) cells containing the following plasmids: (i) pBADMycHisA (MS1199), (ii) pCO3 (Ag43b, MS1231), and (iii) pCO4 (Ag43a, MS1232). All strains were grown in the presence of 0.2% arabinose to induce Ag43 expression. Three different assays were employed. (A) Static biofilm formation in polystyrene microtiter plates. Expression of Ag43 led to significant biofilm formation by MS1231 and MS1232 compared to the biofilm formation by control strain MS1199 (P < 0.05 for each strain, as determined by the independent sample t test). (B) Dynamic biofilm formation in a microfermentor system. Expression of Ag43a led to significant biofilm formation by MS1232 compared to the biofilm formation by control strain MS1199 (P < 0.05, as determined by the independent sample t test). MS1231 cells expressing Ag43b exhibited a small increase in biofilm formation compared to the biofilm formation by control strain MS1199, although the difference was not statistically significant. (C) Dynamic biofilm formation using a flow chamber model. Biofilm development was monitored by confocal scanning laser microscopy. The images are representative horizontal sections collected for each biofilm and vertical sections (to the right of and below each larger panel, representing the yz plane and the xz plane, respectively) at the positions indicated by the white lines.
FIG. 4.
RT-PCR analysis of fluA and fluB expression by CFT073. Total RNA was extracted from CFT073 during exponential growth in LB. Lane 1, _fluA_-specific PCR product (523 bp) obtained using cDNA as the template; lane 2, negative control using RNA prior to cDNA synthesis; lane 3, _fluB_-specific PCR product (518 bp) obtained using cDNA as the template; lane 4, negative control using RNA prior to cDNA synthesis; lane M, 1-kb Plus DNA ladder (Invitrogen).
FIG. 5.
(A) Schematic representation of the OxyR binding sites and adjacent nucleotide sequences of the flu gene from E. coli K-12 and the fluA and fluB genes from CFT073. The putative OxyR binding sites are underlined, the GATC methylation sites are indicated by bold type, and the nucleotide sequence differences are indicated by shading. (B) Ag43 expression in CFT073 flu and oxyR mutants. Protein preparations were obtained by subjecting cells to a brief heat treatment, centrifugation, and subsequent SDS-PAGE and immunodetection analysis using a serum directed against the α-domain of Ag43 from E. coli K-12. Whereas no Ag43 variants could be detected in wild-type CFT073 in this image, longer exposure of the same membrane revealed the presence of Ag43a but not Ag43b in CFT073 (data not shown). Deletion of oxyR resulted in significant expression of both Ag43a and Ag43b, demonstrating that OxyR represses both the fluA and fluB genes in wild-type strain CFT073. (C) Autoaggregation assay demonstrating the settling profiles for liquid suspensions of CFT073, CFT073_fluA_, CFT073_fluB_, CFT_fluA fluB_, CFT073_oxyR_, CFT073_oxyR fluB_, CFT073_oxyR fluA_, and CFT073_oxyR fluA fluB_. Deletion of oxyR did not produce an aggregation phenotype. (D) Biofilm formation by CFT073, CFT073_fluA_, CFT073_fluB_, CFT_fluA fluB_, CFT073_oxyR_, CFT073_oxyR fluB_, CFT073_oxyR fluA_, and CFT073_oxyR fluA fluB_. Deletion of oxyR gave rise to a biofilm phenotype similar to that of the CFT073_fluA_+, CFT073_fluB_+, and CFT073_fluA_+ fluB+ strains.
FIG. 6.
(A) Persistence of wild-type strain CFT073 and CFT073 fluA_- and fluB_-deficient mutants in the bladders of C57BL/6 mice following intraurethral challenge. At day 1 following infection, equivalent numbers of both flu deletion mutants and wild-type strain CFT073 were recovered. At day 5 following infection, significantly lower numbers of the fluA deletion mutant than of wild-type strain CFT073 were recovered (P = 0.01). The data are the mean total numbers of CFU per 0.1 g of bladder tissue ± standard errors of the means for 10 mice. (B) Persistence of E. coli CFT073 and CFT_fluA fluB, CFT_fluA+ fluB, CFT_fluA fluB_+, and CFT_fluA_+ fluB+ mutants in the bladders of C57BL/6 mice following intraurethral challenge. At day 1 following infection wild-type strain CFT073 and the CFT_fluA fluB_ and CFT_fluA_+ fluB strains colonized the bladder epithelium significantly better than strains expressing Ag43b (P ≤ 0.028 for comparisons of all three of the strains with CFT_fluA fluB_+ and CFT_fluA_+ fluB+). At day 5 following infection constitutive expression of Ag43a in E. coli CFT073 promoted longer-term colonization in the bladder irrespective of the presence of Ag43b (P ≤ 0.044 for comparisons of CFT_fluA fluB_ and CFT_fluA fluB_+ with CFT073, CFT_fluA_+ fluB, or CFT_fluA_+ fluB). The data are the mean total numbers of CFU per 0.1 g of bladder tissue ± standard errors of the means for at least two independent experiments. The following numbers of mice were used: on day 1, 29 mice for CFT073, 15 mice for CFT_fluA fluB_, 14 mice for CFT_fluA fluB_+, 15 mice for CFT_fluA_+ fluB, and 15 mice for CFT_fluA_+ fluB+; on day 5, 18 mice for CFT073, 15 mice for CFT_fluA fluB_, 14 mice for CFT_fluA fluB_+, 16 mice for CFT_fluA_+ fluB, and 15 mice for CFT_fluA_+ fluB+. An asterisk indicates that mean bacterial titers were compared using the independent sample t test after log transformation of data to achieve normal distributions according to P-P plots and histogram analysis (SPSS v9.0.2; SPSS, Chicago, IL).
FIG. 7.
(A) Immunodetection analysis of Ag43a and Ag43b proteins produced in CFT073 as a result of constitutive expression from PcL promoter insertions. Protein preparations were obtained by subjecting cells to a brief heat treatment, centrifugation, and subsequent SDS-PAGE and immunodetection analysis using a serum directed against the α-domain of Ag43 from E. coli K-12. Each strain produced the appropriate Ag43 variant(s), as demonstrated by the different sizes of the alpha-domain and the degradation products (indicated by an asterisk). (B) Biofilm formation in microtiter plates by CFT073 expressing Ag43 variants and capsule and type 1 fimbria mutant derivatives. WT, wild type.
References
- Anderson, G. G., J. J. Palermo, J. D. Schilling, R. Roth, J. Heuser, and S. J. Hultgren. 2003. Intracellular bacterial biofilm-like pods in urinary tract infections. Science 301**:**105-107. -PubMed
- Bergsten, G., M. Samuelsson, B. Wullt, I. Leijonhufvud, H. Fischer, and C. Svanborg. 2004. PapG-dependent adherence breaks mucosal inertia and triggers the innate host response. J. Infect. Dis. 189**:**1734-1742. -PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Medical