Correspondence: Annelie Brauner, Department of Microbiology, Tumor and Cell Biology (MTC), Clinical Microbiology, Karolinska Institutet, Karolinska University Hospital, S-171 76 Stockholm, Sweden. Tel.: +46 8 5177 3914; fax: +46 8 3080 99; e-mail: firstname.lastname@example.org
Urinary tract infections (UTI) are one of the most common infectious diseases worldwide. The majority is caused by uropathogenic Escherichia coli. Emerging resistances against conventional antimicrobial therapy requires novel treatment strategies. Beside its role in erythropoiesis, erythropoietin has been recognized to exert tissue-protective and immunomodulatory properties. Here, we investigated the nonerythropoietic erythropoietin analogue ARA290 for potential properties to modulate uroepithelial infection by E. coli in a cell culture model. Expression of the erythropoietin receptor was increased by bacterial stimuli and further enhanced by ARA290 in bladder epithelial cell lines and primary cells as well as in the monocytic cell line THP-1. Stimulation with ARA290 promoted an immune response, inducing a strong initial, but temporarily limited interleukin-8 induction. Moreover, the invasion of bladder epithelial cells by E. coli was significantly reduced in cells costimulated with ARA290. Our results indicate that the erythropoietin analogue ARA290 might be a candidate for the development of novel treatment strategies against UTI, by boosting an early immune response and reducing bacterial invasion as a putative source for recurrent infections.
Urinary tract infections (UTI) are one of the most common infectious diseases worldwide. Uropathogenic Escherichia coli (UPEC) are the causative agent in >80% of uncomplicated UTI. Mechanisms of the innate immune system are considered of prime importance in the defense of the urinary tract against invading organisms (Sivick & Mobley, 2010), although adaptive immunity has been described to contribute to the protection (Thumbikat et al., 2006; Song & Abraham, 2008). Immune response to UPEC is initiated by bacterial contact with the uroepithelium, which induces the production of proinflammatory cytokines, for example interleukin-8 (IL-8) and tumor necrosis factor (TNF)-α, recruitment of neutrophils and clearance of the infection (Song & Abraham, 2008; Sivick & Mobley, 2010). On the other hand, an excessive and prolonged inflammatory response may lead to complications due to tissue damage (Sivick & Mobley, 2010).
Autocrine and paracrine secretion of erythropoietin (Epo) has been discovered to participate in universal stress responses by limiting the self-amplifying proinflammatory cascade (Brines & Cerami, 2008). Expression of the Epo receptor (EpoR) is upregulated by proinflammatory cytokines, for example TNF-α (Taoufik et al., 2008), whereas Epo secretion is downregulated in a concentration-dependent manner by proinflammatory cytokines (Jelkmann, 1998). Therefore, Epo is produced primarily at the periphery of the lesion. This situation allows the usage of exogenous Epo to limit general inflammation and protect the viable tissue (Bernaudin et al., 1999; Li et al., 2006; Lifshitz et al., 2009).
The tissue-protective and immunomodulatory functions of Epo on the one hand and erythropoiesis on the other are mediated by different EpoR (Brines et al., 2004; Brines & Cerami, 2008). The hematopoietic receptor is a homodimer of EpoR subunits with a very high affinity to Epo, corresponding to picomolar concentrations of circulating Epo. The tissue-protective receptor, in contrast, is a heterodimer consisting of one EpoR subunit disulfide-linked to the β common receptor (CD131). Its affinity for Epo is lower and local concentrations of Epo therefore need to be higher. Efforts have been made to design Epo analogues with confined receptor specificity, allowing tissue-protective, but not erythropoietic activity (Brines et al., 2008).
The pyroglutamate helix B surface peptide (ARA290) is a short peptide of 11 amino acids, designed for specificity to the EpoR–CD131 heterocomplex and without erythropoietic function (Brines et al., 2008). The tissue-protective and lack of erythropoitetic activity have been reported for ARA290 with in vitro and animal studies.
Here, we sought to investigate the influence of ARA290 on two parameters crucial for UTI pathogenesis, early immune response and cellular infection by UPEC, using a cell culture model of E. coli UTI.
Materials and methods
All cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and maintained in an appropriate medium (Gibco, Carlsbad, CA) at 37 °C in a 5% CO2 and humidified atmosphere. The human bladder cell lines T24 (HTB-4) and 5637 (HTB-9) were cultured in McCoy's medium and RPMI-1640 medium containing l-glutamine, respectively, supplemented with 10% fetal bovine serum. Primary human bladder epithelium progenitor cells were purchased from CELLnTEC (Bern, Switzerland). Cells were maintained in CnT-58 medium supplemented with antibiotics to final concentrations of 100 U mL−1 penicillin, 100 μg mL−1 streptomycin and 250 ng mL−1 amphotericin B (CELLnTEC) in a 5% CO2 and humidified atmosphere at 35 °C following the instructions of the supplier. For all the experiments, cells reaching confluence were used. The monocytic cell line THP-1 (TIB-202) was maintained in RPMI-1640 medium containing l-glutamine and supplemented with 10% fetal bovine serum, 1 mM HEPES and 0.05 mM 2-mercaptoethanol. In all the experiments, 106 THP-1 cells mL−1 were used.
The E. coli cystitis strain NU14 was used for cell stimulation. Bacteria were grown in a static Luria–Bertani broth to enhance the expression of type 1 fimbriae and collected by centrifugation at 3500 g for 10 min. Bacteria were inactivated by the addition of gentamicin to the cell culture medium (40 μg mL−1) to allow longer stimulation without perturbing the viability of epithelial cells. Alternatively, bacteria were heat-inactivated when cells were used for subsequent infection assays. For this purpose, E. coli NU14 was washed three times and diluted in phosphate-buffered saline (PBS) to a final concentration of 107 CFU mL−1 as determined spectrophotometrically. This suspension was then incubated at 70 °C for 60 min. Inactivation efficiency was checked after an overnight incubation of aliquots plated on blood agar plates. For cell infection assays, the E. coli pyelonephritis strain CFT073 was used. Bacteria were grown on blood agar plates and prepared in PBS as described above and then added to cells at a final concentration of 106 CFU mL−1.
The nonerythropoietic Epo analogue ARA290 was synthesized as described previously. Stock solutions (1–100 μM) were prepared in PBS, filter sterilized (0.2 μm) and kept at 4 °C for up to 4 weeks.
Experiments were performed in 24-well cell culture plates (Costar, Corning, NY). Inactivated bacteria were added to the medium at a final inoculum equivalent to 104, 106 and 108 CFU mL−1 for the initial dose–response experiments. Following this, an inoculum of 106 CFU mL−1 was used. Bacteria were used either alone or together with ARA290 at indicated concentrations (10–1000 nM). As a control, an equal volume of PBS was added to the medium without ARA290. Cells were stimulated for 1–24 h at 37 °C in a 5% CO2 and humidified atmosphere.
Isolation of total RNA, reverse transcription and real-time PCR
Cells were stimulated with gentamicin-inactivated E. coli NU14 as described above. Cells were collected before stimulation and after 1, 3, 6, 12 and 24 h. Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hamburg, Germany) according to the manufacturer's recommendations. RNA was stored at −80 °C until further use. An aliquot of <1 μg was transcribed to cDNA using the DyNAmo cDNA Synthesis kit (Finnzymes, Espoo, Finland). The expression of IL-8, EpoR, LL-37 and β1-integrin was analyzed using gene-specific TaqMan Gene Expression Assays (Applied Biosystems, Carlsbad, CA) according to the manufacturer's instructions. The location of the probes in all assays excluded the detection of genomic DNA. The relative expression of the genes was determined using the ΔΔCT method with GAPDH as an endogenous control (Applied Biosystems).
Determination of IL-8 production
Supernatants from cells stimulated as described for RNA isolation were collected, centrifuged at 300 g for 10 min at 4 °C to remove detached cells and stored at −20 °C until analysis. Aliquots in appropriate dilutions were analyzed for IL-8 protein levels by enzyme-linked immunosorbent assay (ELISA) using the DuoSet ELISA Development System as described by the manufacturer (R&D Systems, Abingdon, UK).
Cell infection assays
Confluent cells in 24-well plates were stimulated with heat-inactivated E. coli NU14 with or without ARA290 in different concentrations. Each condition was analyzed in triplicate. After 6 h of stimulation, E. coli CFT073 was added to each well at a final concentration of 106 CFU mL−1. Plates were centrifuged at 300 g for 5 min to expedite bacterial contact with host cells and then incubated for 30 min at 37 °C. To determine the total number of cell-associated bacteria, cells were washed three times with warm PBS to remove loosely attached bacteria, lysed with 1% Triton X-100 in PBS and plated in different dilutions on blood agar plates. To assess the number of intracellular bacteria, plates were washed and then incubated for another 60 min in a fresh medium. Then, extracellular bacteria were killed by incubation with a medium containing gentamicin (100 μg mL−1) for 30 min. After washes with warm PBS, the cells were lysed and lysates were plated as above. Bacterial recovery was determined after an overnight incubation. The invasion rate was determined as the relation of intracellular bacteria to the total count from the same experiment.
Cell proliferation assay
To determine the possible influence of ARA290 on cell proliferation and viability, the XTT assay was used (Sigma-Aldrich, St. Louis, MO). Cells were grown in 96-well plates (Costar) until reaching confluence and stimulated for 24 h as described above. Cells incubated in medium alone served as controls. Triplicates were analyzed for each condition. After 24 h, cells were washed three times in PBS and incubated for 4 h with 250 μL freshly prepared XTT–menadione solution (1 mg mL−1 and 12.5 μM, respectively) at 37 °C. The formazan concentration was then measured at 490 nm.
Immunoprecipitation and Western blot analysis
For immunoprecipitation, cells were seeded in six-well plates (Costar). After reaching confluence, the cells were stimulated and infected as described for cell infection assays. After centrifugation at 300 g for 5 min, cells were incubated for further 5, 15 or 25 min at 37 °C or collected directly. Cells were washed with ice-cold PBS, lysed with lysis buffer [137 nM NaCl, 1% IGEPAL CA-630, 20 mM Tris (pH 8.0), 200 μM phenylmethylsulfonyl fluoride, 10% glycerol, complete protease inhibitor (1 : 100, Sigma-Aldrich), phosphatase inhibitor cocktail (1 : 100, Sigma-Aldrich)] and cleared by centrifugation for 20 min at 10 000 g and 4 °C. The protein concentration in the lysates was measured using BCA Protein Assay reagent (Pierce, Thermo Scientific, Rockford, IL) and samples were adjusted to equal protein concentrations. Lysates were then incubated for 1 h at room temperature with Protein G-coated beads (Dynabeads Protein G; Dynal, Oslo, Norway) to remove unspecifically bound proteins. Cleared lysate was incubated with goat anti-focal adhesion kinase (anti-FAK) antibody A-17 (Santa Cruz Biotechnology, Santa Cruz, CA) overnight at 4 °C. The FAK–antibody complex was then precipitated with Protein G-coated beads for 1 h at room temperature. After three washes with PBS, collected proteins were eluted from the beads by heating the samples in sodium dodecyl sulfate (SDS) sample buffer (Bio-Rad Laboratories, Hercules, CA) supplemented with 0.5%β-mercaptoethanol at 95 °C for 5 min. Proteins were subjected to SDS-polyacrylamide gel electrophoresis on a 10% polyacrylamide gel (Tris-HCl Ready Gel Precast Gel, Bio-Rad Laboratories) and transferred to a polyvinylidene fluoride membrane (Invitrogen, Carlsbad, CA). The membrane was blocked with 5% milk in 0.05% Tween 20 in Tris-buffered saline (TBST) for 1 h at room temperature and incubated with rabbit anti-pFAK antibody against pTyr397 (1 : 1000, Sigma-Aldrich) in blocking buffer overnight at 4 °C. The membrane was then washed three times in TBST, incubated with an anti-rabbit horseradish peroxidase-conjugated secondary antibody (1 : 3000, Bio-Rad Laboratories) for 1 h at room temperature and washed again. Proteins were visualized using the SuperSignal West Pico Chemiluminescent Substrate (Pierce). To ensure equal amounts of FAK in all samples, the membrane was stripped and reprobed with rabbit anti-FAK antibody C-309 (1 : 200 in blocking buffer, Santa Cruz Biotechnologies). Digital images of the membrane were analyzed using imagej software (NIH, Bethesda, MD, http://rsb.info.nih.gov/ij/).
Data were analyzed by one-way anova with the Newman–Keuls multiple comparison post test using the graphpad prism version 5.02 (GraphPad Software, San Diego, CA). Differences with P-values <0.05 were considered statistically significant.
Results and discussion
EpoR is expressed by bladder epithelial cells
The expression of EpoR in nonerythropoietic tissue is debated (Ghezzi et al., 2010; Sinclair et al., 2010; Swift et al., 2010; Xiong et al., 2010). A prediction for our hypothesis was, however, that EpoR, as part of the heteromer with CD131, is expressed in the bladder epithelium. We therefore tested the bladder epithelial cell lines and primary bladder epithelial cells used in our cell infection model for EpoR expression. We could detect low constitutive levels of EpoR-specific mRNA in all three bladder cell types investigated in this study (Fig. 1) as well as in the monocytic cell line THP-1. Discrepancies among the findings reported by others might result from the different sensitivities of methods or interpretation criteria (Ghezzi et al., 2010).
Bacterial stimuli lead to upregulation of EpoR expression, which is further enhanced by ARA290
Contact between E. coli and bladder epithelial cells induces a general inflammatory response. In other nonerythropoietic tissues, TNF-α-dependent upregulation of EpoR has been described to mediate the tissue-protective action of Epo (Brines & Cerami, 2008). To investigate whether this also applies for bladder epithelial cells, we exposed cells to bacterial stimuli, E. coli NU14, and determined the mRNA expression of EpoR at different time points after stimulation. The expression of EpoR was induced in a bimodal manner, with a first peak at three (5637 cells) or 6 h (primary cells) and a second upregulation after 24 h of stimulation (Fig. 2a). This first peak was very low in T24 cells stimulated with bacteria alone. When, however, these cells were costimulated with ARA290, EpoR expression was upregulated 3 h after costimulation (P<0.05; Fig. 2b). Enhanced and earlier EpoR upregulation in the presence of ARA290 was also observed for 5637 and primary bladder epithelial cells, although the effect was less pronounced (data not shown). In the monocytic cell line THP-1, a similar pattern was observed, but expression peaked earlier, after 1 and 12 h of stimulation, respectively (Fig. 2a). Additional stimulation with ARA290 showed no obvious additive effect. Synergistic positive effects of stress stimuli and Epo on EpoR expression have been reported previously in endothelial cells (Beleslin-Cokic et al., 2004). Our observation suggests that this effect becomes more evident when the basal levels of EpoR expression are low and ARA290 is applied in nanomolar concentrations. Based on these initial results, we chose an incubation time of 6 h and 10 nM or 100 nM ARA290 as an appropriate condition to prestimulate cells in further experiments.
ARA290 does not enhance proliferation of urothelial cells
Epo and its analogues have been described to enhance proliferation in healthy tissue, tumors and cell lines (Kumar et al., 2005; Hardee et al., 2007). Such activity would clearly constitute a strong adverse effect for the usage of ARA290 in the urinary tract. In addition to its clinical relevance, pronounced differences in cell growth would also skew the results from in vitro assays. Therefore, we investigated the cell proliferation and viability of cells cultured in the presence of ARA290 for 24 h and performed an XTT assay. On applying the assay, we could not detect any significant difference in cell proliferation and viability between treated and control cells in concentrations used for further experiments (T24: 102.7±5.8% for 100 nM ARA290; 5637: 97.1±3.2% for 100 nM ARA290) nor at higher concentrations (for 1 μM ARA290, T24: 90.33±7.6%; 5637: 98.3±0.7%). No changes were observed when cells were costimulated with inactivated bacteria (data not shown).
ARA290 induces an early, but limited inflammatory response to bacterial stimulation
The neutrophil-attractant chemokine IL-8 serves a crucial function during UTI in mediating the elimination of bacteria (Hedges et al., 1994; Agace, 1996). The treatment with recombinant Epo has repeatedly been demonstrated to reduce lipopolysaccharide-induced cytokine induction in leukocytes (Schultz et al., 2008; Strunk et al., 2008; Yazihan et al., 2008). To test whether ARA290 modulated this immune response, we costimulated bladder epithelial cell lines with E. coli NU14 and ARA290 in different concentrations. During the period corresponding to basal levels of EpoR expression, the additional presence of ARA290 enhanced IL-8 mRNA expression. At 3 h, an increase in the IL-8 mRNA levels was observed in T24 cells after costimulation with 100 nM ARA290, compared with stimulation with bacteria alone (127% of 0 nM ARA290, P<0.05; Fig. 3a). This early proinflammatory effect was even stronger with 10 nM ARA290 (155% of 0 nM ARA290, P<0.05). Consequently, IL-8 protein levels were higher in cell culture supernatants 12 h after costimulation with 100 nM ARA290 (115% of 0 nM ARA290, Fig. 3b) or 10 nM ARA290 (125% of 0 nM ARA290, P<0.05). At later time points, when EpoR expression was upregulated, ARA290 costimulation did not further promote immune induction. In contrast, IL-8 levels were reduced on mRNA (61% of 0 nM ARA290, P<0.05; Fig. 3a) and protein levels (78% of 0 nM ARA290, P<0.05; Fig. 3b). This downregulation was also observed at 10 nM ARA290, even though not as pronounced (91% for mRNA and 81% of 0 nM ARA290 for protein, P<0.05). A similar pattern of IL-8 mRNA expression was observed for bladder epithelial cells 5637 (data not shown). In the monocytic cell line THP-1, where upregulation of EpoR expression occurred very early (Fig. 1), reduction of IL-8 mRNA was accordingly detected already 1 h after costimulation with ARA290.
ARA290 decreases E. coli invasion into prestimulated bladder cells
To establish infection, E. coli firmly adheres and eventually invades the epithelial cells in the urinary bladder (Wu et al., 1996; Martinez et al., 2000). Intracellular bacteria are able to multiply and persist in the bladder epithelium, likely constituting the reservoir for recurrent infection (Mysorekar & Hultgren, 2006). We therefore investigated whether ARA290 influenced these two crucial steps of bacterial infection. In 5637 bladder epithelial cells, the total number of E. coli did not differ after any treatment. In contrast, invasion was reduced when cells were costimulated with inactivated bacteria and 100 nM ARA290 (P<0.05; Fig. 4). A similar effect was obtained in the bladder epithelial cell line T24 by costimulation with 10 nM ARA290 (data not shown).
ARA290 reduces the activation of FAK during infection
To understand the mechanism underlying reduced bacterial invasion, we investigated the pathways known to be activated during E. coli invasion into bladder epithelial cells. Type 1 fimbriae expressed by virtually all UPEC bind to different cell surface markers on uroepithelial cells, including β1 integrins (Martinez et al., 2000; Eto et al., 2007). Activated β1 integrin signals to FAK, which becomes phosphorylated and further activates phosphoisonitol-3-kinase. Eventually, bacterial binding induces rearrangement of the cellular actin cytoskeleton and uptake into the cell (Martinez & Hultgren, 2002). We assessed the influence of ARA290 on the activation of this pathway by determining the content of phosphorylated FAK (pFAK) at 5, 15 and 25 min after infection with E. coli CFT073. As expected, infection with CFT073 induced increased levels of pFAK (Fig. 5). Interestingly, activation of FAK was diminished in cells costimulated with ARA290, indicated by lower levels of pFAK compared with cells exposed to bacterial stimuli only. The total FAK levels were not affected by this treatment as determined by reprobing the blot with anti-FAK antibody. It thus remains to be determined whether reduced FAK activation was due to the specific inhibition of FAK phosphorylation, or whether upstream signals, i.e. β1 integrin signaling was impaired. However, we did not observe changes in β1 integrin mRNA expression, nor could we detect changes on the protein level, either in the total or in the membrane protein fraction (data not shown).
With emerging resistance against conventional antimicrobial therapy, new treatment strategies are needed. In this study, we investigate whether the nonerythropoitetic Epo analogue ARA290 might be a candidate for such an approach. Using an in vitro model of E. coli UTI, we reveal two mechanisms by which ARA290 modulates E. coli infection. Firstly, ARA290 expedites an early immune response, but at the same time limits the peak of IL-8 production. This might support an early, efficient elimination of bacteria while reducing inflammation-associated tissue damage. Secondly, ARA290 directly reduces cellular infection due to interference with bacterial invasion. Because the intracellular niche is regarded as a relevant reservoir for E. coli, this may confer protection against recurrence of the infection. Taken together, the combination of these effects makes ARA290 a promising substance both to boost the immune response during acute UTI and to prevent recurrence of the infection.
This work was supported by grants from the Swedish Research Council (56X-20356) and ALF Project Funding and Karolinska Institutet.