Acute cystitis strains elicit an IL-1β response in human bladder epithelial cells

To address how infection creates a hyper-inflammatory state in patients with acute cystitis, we first infected the human bladder epithelial cell line HTB-9 in vitro and quantified inflammatory mediators in cell supernatants. We detected an increase in IL-1β secretion, four hours after infection with acute cystitis (CY) strains CY-17, CY-92, CY-132 or the uropathogenic Escherichia coli strain CFT073 (P < 0.001, compared to uninfected cells, two-tailed unpaired t-test). In contrast, the IL-1β response was low in cells infected with the ABU strain E. coli 83972, indicating a virulence-association (Fig 1A). IL-1β secretion was not detected in kidney epithelial cell supernatants after infection with the same strains, suggesting specificity for the bladder epithelium (Fig 1B). Western blot analysis confirmed that mature IL-1β was present in the supernatants of the infected HTB-9 cells (4 hours), as well as unprocessed pro-IL-1β and N-terminal fragment (Fig 1C). A rapid increase in IL-1β staining intensity was observed by confocal microscopy, in cells infected for one hour with 10⁵ CFU/ml of CY-17, CY-92 and CFT073 compared to uninfected cells or cells infected with the ABU strain (Fig 1D and 1E, two-tailed unpaired t-test). This increase in cellular IL-1β levels was confirmed by Western blot analysis of whole cell extracts (Fig 1F and 1G). At this time (1 hour), low levels of pro-IL-1β were detected. There was no significant reduction in cell viability after one (≥ 95% viable) or four hours (≥ 90% viable), as quantified by PrestoBlue staining and no evidence of pyroptosis after one hour, when the increase in cellular IL-1β levels was detected (S1 Fig). The presence of unprocessed pro-IL-1β in the 4 hour supernatant might be due to secretion of unprocessed IL-1β via exosomes [30].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Acute cystitis strains activate an IL-1β response in human bladder epithelial cells.

(A) IL-1β response in human bladder epithelial carcinoma cells (HTB-9) infected with acute cystitis strains CY-92, CY-17, CY-132 and CY-49 (4 hours). CFT073 and E. coli 83972 (ABU) were used as reference strains. CY-17, CY-92 and CY-132 triggered high IL-1β responses, as did CFT073. IL-1β was quantified, in cell supernatants, by ELISA (means ± SEMs of three independent experiments, *** P < 0.001 compared to PBS, two-tailed unpaired t-test). (B) Lack of IL-1β secretion by infected human kidney epithelial carcinoma cells (A498), infected as in Fig 1A. (C) Increased pro-IL-1β and mature IL-1β levels in HTB-9 cells infected with CY-92, CY-17 and CY-132 (Western blot analysis of cell supernatants, 4 hours, one representative experiment of several repeats). (D) IL-1β staining of HTB-9 cells infected with CY-17, CY-92, CY-132, CFT073 or ABU compared to the background in uninfected cells (PBS). Scale bars = 20 μm (upper panel) and 10 μm (lower panel). One representative experiment is shown. (E) Quantification of the cellular IL-1β response to infection. Increase in total fluorescence intensity (open pin-hole) after subtraction of the background staining in uninfected cells (PBS) (means ± SEMs of 50 cells per sample, ** P < 0.01 and *** P < 0.001 compared to PBS, two-tailed unpaired t-test). One of three experiments is shown. (F, G) IL-1β response to infection (1 hour) quantified by Western blot of whole cell extracts. Quantification of integrated density relative to GAPDH normalized against the background of uninfected cells. One representative experiment of several repeats. (H) IL-1β response to an epidemiologically defined collection of pediatric acute cystitis strains (n = 67) compared to ABU strains (n = 62), obtained from children in the same geographic area. IL-1β was quantified in infected cell supernatants, by ELISA. Pie chart depicting the frequency of bacterial strains activating IL-1β responses: high (orange), intermediate (blue), low (purple) or negative (green). Histogram (inset) of the mean IL-1β response to CY versus ABU strains (means ± SEMs, *** P < 0.001, two-tailed Mann Whitney test). (I) IL-1β activation plotted against hemolytic activity in the collection of CY and ABU strains. No significant association was detected (n = 18–21, Hly+ versus Hly-, two-tailed Mann Whitney test).

https://doi.org/10.1371/journal.ppat.1005848.g001

To address if IL-1β activation is a characteristic of acute cystitis strains, we infected human bladder epithelial cells with an epidemiologically defined collection of pediatric acute cystitis isolates (n = 67, [31, 32]). The majority of these strains (85%) triggered an IL-1β response > 5 pg/ml and 64% of those triggered a high response (40–1,000 pg/ml, Fig 1H). We also examined a collection of pediatric ABU strains (n = 62, [31, 33]), which was obtained by screening infants and children in the same geographic area for bacteriuria in the absence of urinary tract infection (UTI) symptoms. In contrast to the CY strains, most of the ABU strains did not trigger a strong IL-1β response (61% < 5 pg/ml), resulting in significantly higher mean IL-1β concentrations in supernatants of bladder cells infected with the CY strains than the ABU strains (121.8 and 32.4 pg/ml respectively, P < 0.001, Fig 1H). To address if the secretion of IL-1β was influenced by bacterial hemolysin [19, 20], IL-1β concentrations were examined as a function of hemolytic activity in 40 CY and 38 ABU strains (Fig 1I). There was no significant difference in IL-1β response between hemolysin positive and negative strains (P = 0.07, Mann Whitney unpaired test). In the cystitis subset, a significant difference between hemolysin positive and negative strains was observed, however (P = 0.01, Mann Whitney unpaired test), suggesting that hemolytic cell lysis may contribute to the IL-1β activating virulent phenotype of the acute cystitis strains, for example by assisting the release of IL-1β from cells infected with hemolysin-producing strains.

The results suggest that the majority of acute cystitis strains activate an IL-1β response in human bladder epithelial cells.

Genetic control of acute cystitis in the murine UTI model

As IL-1β is processed by the inflammasome, we subsequently examined if mice with intact or defective inflammasome function develop acute cystitis. We infected mice with genetic defects affecting the NLRP-3 inflammasome: NLRP-3 deficient mice (Nlrp3^-/- [34]) or ASC deficient mice (Asc^-/- [35]). In addition, Il1b^-/- [36] and Casp1^-/- [37] mice were used and C57BL/6 WT mice were included as controls (Fig 2). For sample sizes and number of experiments, please see each figure legend and an overview in S1 Table. The mice were infected by intravesical inoculation with E. coli strains that triggered high IL-1β responses in human bladder epithelial cells, in vitro (CFT073, CY-17 or CY-92). Infected bladders were evaluated macroscopically, at sacrifice after 7 days and assigned a gross pathology score, defined by size, edema and hyperemia. Tissue pathology was further evaluated by hematoxylin and eosin (H&E) staining and immunohistochemistry of frozen tissue sections, and a histo-pathology score was assigned to each mouse. Histology was scored, independently, by two experienced researchers. The analysis was not blinded. Infection kinetics was followed in urine samples obtained after 6 and 24 hours, 3 and 7 days.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Genetic control of inflammation in acute cystitis.

(A) Macroscopic evidence of acute cystitis in CFT073 infected mice, 7 days after infection. Slight increase in edema and hyperemia in C57BL/6 WT mice compared to uninfected controls, massively enlarged, hyper-inflamed bladders in Asc^-/- and Nlrp3^-/- mice and intact morphology in bladders from Il1b^-/- mice. The macroscopic pathology score was based on edema, hyperemia and size on a scale of 0–10, where 10 the most edematous, hyperemic and largest bladder). Individual pathology scores are indicated. Scale bar = 1 mm. Two representative bladders are shown for each genotype. (For uninfected Asc^-/-, Nlrp3^-/- and Il1b^-/- mice see S2A Fig). (B) Bladder tissue structure in H&E-stained tissue sections. Bladders from Asc^-/- and Nlrp3^-/- mice showed extensive edema, a loss of tissue structure and epithelial hypertrophy, compared to C57BL/6 WT and Il1b^-/- mice or uninfected control mice. The histology score was assessed using H&E stained bladder sections based on neutrophil infiltration, tissue architecture and epithelial thickness on a scale of 0–10, where 10 is highest neutrophil infiltration, least preserved tissue architecture and maximum epithelial thickness). Individual histology scores are indicated. Scale bar = 200 μm. (C) Gross bladder pathology score in infected mice (7 days) and uninfected C57BL/6 WT controls (means ± SEMs of two experiments, *** P < 0.001, two-tailed unpaired t-test compared to WT mice). (D) Elevated neutrophil counts and (E) bacterial counts in urine and (F) bladder tissues from Asc^-/- and Nlrp3^-/- mice, compared to WT and Il1b^-/- mice. Means ± SEMs of two experiments, *** P < 0.001, ** P < 0.01, * P < 0.05, two-tailed unpaired t-test compared to WT mice. (G) Detection of neutrophils and bacteria in the mucosa of infected and control mice, by immunohistochemistry. Increased staining for mucosal neutrophils (red) and bacteria (green) in Asc^-/- and Nlrp3^-/- mice, compared to C57BL/6 WT, Il1b^-/- mice and uninfected controls. Tissues obtained 7 days after infection. Scale bar = 50 μm. (H) Mucosal IL-1β staining in bladder tissue sections, obtained 24 hours after infection. Infection increased epithelial IL-1β staining in C57BL/6 WT, Asc^-/- and Nlrp3^-/- mice. Scale bar = 50 μm. Experiments included 14 Asc^-/-, 11 Nlrp3^-/-, 11 C57BL/6 WT and 10 Il1b^-/- mice.

https://doi.org/10.1371/journal.ppat.1005848.g002

Major, genotype-specific differences in bladder inflammation and pathology were detected after seven days (Fig 2). Two disease end points were distinguished. 1. Severe, progressive cystitis in mice lacking ASC or NLRP-3, resembling chronic human disease. 2. A moderate, self-limiting form of acute cystitis in C57BL/6WT mice with intact inflammasome function, resembling sporadic human cystitis.

Infected Asc^-/- and Nlrp3^-/- mice developed severe, acute cystitis with enlarged bladders, edema and hyperemia compared to uninfected bladders (Fig 2A and S2A Fig). By histology, most bladders from Asc^-/- mice showed extensive loss of tissue structure, defined by a pronounced mucosal and submucosal edema, with disappearance of the tissue folds that characterizes the healthy mucosa. In addition, a massive inflammatory cell infiltrate was present, along the mucosal border and in the submucosa (10/14 mice, 71%). Similar tissue destruction was observed in bladders from Nlrp3^-/- mice (7/11 mice, 64%), (Fig 2B). Mean gross bladder pathology score of infected Asc^-/- and Nlrp3^-/- mice were 7.9 and 7.2, respectively (Fig 2C).

Bladder pathology was accompanied by high bacterial counts in urine and bladder tissue (Fig 2D–2F). In Asc^-/- mice, the neutrophil influx accelerated until day 7, indicating a loss of homeostatic control and progression to chronic inflammation. Infiltrating bacteria and neutrophil aggregates or micro-abscesses were detected in the mucosa of Asc^-/- and Nlrp3^-/- mice, with extensive sloughing of epithelial cells into the lumen. Bacteria were mainly localized along the mucosal surface, with no evidence of bacterial invasion (Fig 2G). This hyper-inflammatory phenotype was also observed after infection of Asc^-/- mice with the acute cystitis strains CY-92 and CY-17, which triggered high IL-1β responses in vitro (S3A–S3D Fig). In contrast, there was no disease phenotype in Asc^-/- mice infected with the ABU strain E. coli 83972 or in C57BL/6 WT mice after 24 hours or 7 days (S3E Fig). There was no evidence of kidney involvement or pathology in mice infected with CY-92 and CY-17, despite positive bacterial cultures from renal tissues.

In contrast, C57BL/6 WT mice infected with CFT073 showed moderate macroscopic evidence of acute cystitis including a small increase in size, edema and hyperemia compared to uninfected controls (Fig 2A and 2C, mean pathology score 1.5). The low level of edema was confirmed by histology, with no evidence of tissue damage (Fig 2B). Infection was accompanied by an increase in urine neutrophil numbers (Fig 2D) and bacterial numbers reached a peak after 24 hours and then declined (Fig 2E and 2F). By immunohistochemistry, bacterial staining was weak and very few neutrophils were detected in the bladder mucosa (Fig 2G).

Il1b^-/- mice showed an even more attenuated phenotype after infection with CFT073, consistent with a key role of IL-1β for bladder inflammation and pathology compared to uninfected bladders (Fig 2A and S2A Fig). There was no macroscopic evidence of acute cystitis (Fig 2A and 2C, mean pathology score 0.7) and bladder tissue morphology remained intact (Fig 2B, mean histo-pathology score 0.9, P = 0.003 compared to C57BL/6 WT mice). Il1b^-/- mice had fewer infiltrating neutrophils and lower bacterial counts than the C57BL/6 WT mice on day seven (P < 0.001 and P < 0.05), (Fig 2D–2F). Furthermore, Casp1^-/- mice, which have a functional IL-1β deficiency due to defective IL-1β processing and secretion [37], did not develop acute cystitis (S4A–S4D Fig). The bladders were enlarged and hyperemic, but there was no evidence of inflammatory changes or tissue damage. Casp1^-/- mice showed reduced IL-1β secretion and tissue retention of IL-1β (S4E–S4G Fig). As a result, IL-1β dependent gene expression was low and Casp1^-/- mice showed a lack of inflammation in bladder tissues. Neutrophils and bacteria were present in urine but did not accumulate in the tissues and the mucosal morphology was intact (S4 Fig).

Infection was accompanied by strong mucosal IL-1β staining in bladder tissue sections in C57BL/6 WT mice, Asc^-/- and Nlrp3^-/- mice after 24 hours (Fig 2H). Staining was mainly epithelial and was not seen in Il1b^-/- mice or uninfected C57BL/6 WT mice. In parallel with the epithelial staining IL-1β was detected by ELISA in the urine of infected Asc^-/- and Nlrp3^-/- mice, with lower levels in C57BL/6 WT mice. By Western blot analysis, bands of approximately 36 and 18 kDa were detected (S2B Fig).

These studies identify genetic determinants of host susceptibility to acute cystitis. Asc and Nlrp3 were defined as key resistance determinants and IL-1β activation as a crucial step in the pathogenesis of acute cystitis.

Asc and Nlrp3 control gene expression in infected bladders

To define the mechanism of bladder pathology, we extracted total bladder RNA from infected Asc^-/- and Nlrp3^-/- mice with the highest pathology score after seven days and from C57BL/6 WT and Il1b^-/- mice, with low pathology scores (Experiments 1 and 2 in S1 Table) and from uninfected bladders. The RNA was amplified, hybridized onto Mouse Genome array strips, washed, stained and scanned using the GeneAtlas system. Significantly altered genes were identified, by comparing infected- to uninfected mice of the same genetic background (P-values < 0.05 and absolute fold change > 1.41) and sorted by relative expression using 2-way ANOVA [38]. Heat-maps were constructed by Gitools 2.1.1 software and differentially expressed genes and regulated pathways were analyzed by Ingenuity Pathway Analysis software (see Materials and Methods).

We identified a set of strongly upregulated genes in Asc^-/- and Nlrp3^-/- mice with the highest bladder pathology score, but not in C57BL/6 WT mice or Il1b^-/- mice (2,228 specifically regulated genes). The heat map in Fig 3A illustrates the similarities in gene expression between 5/7 Asc^-/- and 2/5 Nlrp3^-/- mice analyzed by this technology. Those mice also had high and comparable histology scores, defined by evaluation of the H&E-stained bladder tissue sections from the corresponding mice. To further understand the disease process, we identified the most strongly upregulated genes in these mice. Genes with a FC > 100 included metalloproteinase Mmp7, the neutrophil and monocyte chemoattractants Cxcl6 and Cxcl3, the genes encoding calprotectin S100a8 and a9 and the stefin gene Stfa1 (Fig 3B and S2 Table). By analysis of top-scoring canonical pathways, these genes were shown to control granulocyte and leucocyte diapedesis and signaling, acute phase responses including IL-6 and IL-1β signaling, IL-1R expression and NF-κB-signaling and dendritic cell maturation (S5A Fig). These genes and pathways were not significantly regulated in C57BL/6 WT or Il1b^-/- mice, supporting a disease association.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Hyper-activation of IL-1β dependent gene expression and bladder pathology in Asc^-/- and Nlrp3^-/- mice.

Transcriptomic analysis of whole bladder RNA from infected mice (CFT073, 7 days), compared to uninfected controls of each genotype (cut off FC 1.41, P < 0.05). (A) Heatmap of regulated genes in Asc^-/- and Nlrp3^-/- mice with the highest bladder pathology score, defined by neutrophil infiltration, loss of tissue structure and epithelial thickness in H&E stained bladder tissue sections. Scale FC -4 to 4, red = upregulated, blue = downregulated. A distinct gene set distinguished the Asc^-/- and Nlrp3^-/- mice with a high histopathology score from C57BL/6 WT mice or Il1b^-/- mice without pathology. (B) Top up-regulated genes in the pathology-associated gene set, compared to uninfected controls of each genotype. Means ± SEMs of 5 mice for Asc^-/- mice and 2 Nlrp3^-/- mice. (C) Analysis of IL-1β, inflammasome activators and effectors in Asc^-/- and Nlrp3^-/- mice, detecting massive over-expression compared to Il1b^-/- and WT mice. Red = upregulated, blue = suppressed. The data set included gene expression profiles from 7 Asc^-/- and 5 Nlrp3^-/- mice, and two each of the C57BL/6 WT and Il1b^-/- controls. Uninfected control RNA of each genotype were used to define significantly regulated genes (≥ 2 mice per genotype). Histopathology scores and group numbers for individual mice (see also Experiments 1, 2 and 3 in S1 Table).

https://doi.org/10.1371/journal.ppat.1005848.g003

The activation of inflammasome-genes and IL-1β-dependent gene network was analyzed, using Qiagen’s list of 84 key inflammasome genes (Fig 3C). In the Asc^-/- and Nlrp3^-/- mice with a high histo-pathology score, Il1b expression was activated (FC 10–41), as were IL-1β-dependent and inflammasome-related genes including Cxcl1, Cxcl3 and Il33 (FC 5–200), (Fig 3C). These genes and pathways were not significantly regulated in C57BL/6 WT or Il1b^-/- mice (Fig 3C). Il1a expression was activated in Asc^-/- and Nlrp3^-/- mice with bladder pathology, but not among the 40 top-regulated genes (FC 1.6–8.7). Il18, Casp11 and inflammasome-related NLRP genes were not transcriptionally regulated (S2 Table).

The results identify IL-1β driven pro-inflammatory genes that are activated, exclusively in Asc^-/- and Nlrp3^-/- mice with severe bladder pathology. This response was not detected in the kidneys of infected C57BL/6 WT mice (S5B Fig).

Mechanism of atypical IL-1β processing in infected bladders

The Mmp7 gene, which encodes the matrix metalloproteinase (MMP)-7 [39] was strongly upregulated in Asc^-/- and Nlrp3^-/- mice with a high histo-pathology score (Fig 3B). MMP-7 expression was therefore examined as a function of the histo-pathology score (Fig 4A). In Asc^-/- and Nlrp3^-/- mice, Mmp7 expression showed a clear association to the overall bladder tissue pathology score and was not regulated in the Il1b^-/- or C57BL/6 WT mice (Fig 4A). High MMP-7 protein expression was confirmed, by immunohistochemistry, in bladder tissue sections from Asc^-/- and Nlrp3^-/- mice (Fig 4B). Importantly, staining was exclusively epithelial, with shedding of MMP-7 positive cells into the bladder lumen. Epithelial MMP-7 activation was detected as early as 24 hours after infection and importantly, MMP-7 showed no detectable co-localization with neutrophils in the mucosa or sub-mucosa (S6 Fig) To further evaluate the involvement of MMP-7 in acute cystitis, we infected Mmp7^-/- mice [40] with CFT073 and used Asc^-/- mice as disease controls. Consistent with their intact inflammasome function, Mmp7^-/- mice developed transient cystitis similar to C57BL/6 WT mice (Fig 4C). A moderate mucosal IL-1β response was observed by immunohistochemistry (Fig 4C). IL-1β levels in urine (Fig 4C) and IL-1β-dependent gene expression was comparable to that in WT mice, with expression of Ccl5, Nlrc5, Irf1, Ctsb, Birc3, and MyD88. Thus, Mmp7 did not drive pathology in mice with intact ASC or NLRP-3 function.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. IL-1β processing by MMP-7.

(A) Gene expression profiling identified Mmp7 as one of the top up-regulated gene in Asc^-/- and Nlrp3^-/- mice with bladder pathology (CFT073 infected mice, 7 days). Log₂ fold change of Mmp7 expression levels in individual mice are shown relative to the H&E pathology score. Mmp7 was not regulated in C57BL/6 WT mice or in Il1b^-/- or Casp1^-/- mice. (B) Strong epithelial MMP-7 staining in Asc^-/- and Nlrp3^-/- mice with bladder pathology. MMP-7 staining was very low in C57BL/6 WT, Il1b^-/- and Casp1^-/- mice. Scale bars = 50 μm. (C) Phenotype of Mmp7^-/- mice, 7 days after infection with CFT073. Intact mucosal tissue structure with inflammatory cell infiltration. Low bacterial and neutrophil counts in urine compared to Asc^-/- mice (n = 5 mice per group, means ± SEMs, ** P < 0.01, *** P < 0.001, two-tailed unpaired t-test). Scale bar = 1 mm. IL-1β levels were elevated in the urine of Asc^-/- mice but not in Mmp7^-/- mice, as detected by ELISA. (D) Proteolytic cleavage of pro-IL-1β by MMP-7 in vitro, using purified enzyme and GST-tagged pro-IL-1β. The IL-1β fragments generated by proteolysis were 18 and 16 kDa, defined by Western blot using an antibody specific for the mature form of IL-1β. Recombinant mature IL-1β and GST-tagged pro-IL-1β were used as controls, as well as recombinant MMP-7. One representative experiment out of three, see also S7A Fig. (E) Bioassay for IL-1β activity, measuring the PGE₂ response of human bladder epithelial cells to the IL-1β fragments generated by MMP-7 proteolysis of GST-tagged pro-IL-1β. The cleaved products activated PGE₂ but MMP-7 and pro-IL-1β had no effect (means ± SEMs of two experiments, ** P < 0.01, two-tailed Mann Whitney test).

https://doi.org/10.1371/journal.ppat.1005848.g004

To address if MMP-7 cleaves IL-1β, we exposed recombinant GST-tagged pro-IL-1β to recombinant active MMP-7 in vitro and detected proteolytic fragments by Western blot, using IL-1β specific antibodies (Fig 4D and S7A Fig). Kinetic analysis detected a time-dependent cleavage of IL-1β with a reduction in full-length protein from 10 to 60 minutes (Fig 4D). Using antibodies with higher affinity for the mature IL-1β, a band of 18 kDa was detected corresponding in size to the recombinant, active and mature IL-1β control. With increasing time, a band of 16 kDa was also observed (Fig 4D). Using the same experimental set up, we observed that ASC was degraded by MMP-7 over time (S7B Fig) while recombinant NLRP-3 was not cleaved by MMP-7 and therefore served as a negative control for unspecific effects of the enzyme (S7C Fig).

To address if the cleaved IL-1β fragments were biologically active, reaction mixtures containing pro-IL-1β and MMP-7 were collected after 30 minutes, when the mature product was detected by Western blot (Fig 4D). Human bladder epithelial cells were stimulated with the reaction mixture for one hour, and IL-1β activity was quantified, by measuring the prostaglandin E2 (PGE₂) response [41]. The 30 minutes reaction mixture activated a dose-dependent PGE₂ response but recombinant MMP-7 and pro-IL-1β (280 and 840 ng/ml) alone had no effect (Fig 4E).

The results identify a new, MMP-7-dependent mechanism of pro-IL-1β processing in Asc^-/- and Nlrp3^-/- mice.

NLRP-3 and ASC act as negative regulators of MMP7 expression

To understand the mechanism of increased MMP-7 expression in infected Asc^-/- and Nlrp3^-/- mice, we examined if ASC and/or NLRP-3 may act as negative regulators of MMP7 expression. After infection of human bladder epithelial cells, with CY-17 and CY-92, we detected a significant increase in MMP-7 staining (confocal microscopy, Fig 5A and 5B). In contrast, ASC staining was reduced after infection with the virulent strains (P < 0.001) and NLRP-3 showed a weaker staining (P < 0.01). The MMP-7 response to infection and the decrease in ASC and NLRP-3 levels were confirmed by Western blot analysis (Fig 5C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Regulation of MMP7 expression by ASC and NLRP-3.

(A) MMP-7, ASC and NLRP-3 responses to infection were visualized by confocal microscopy. An increase in MMP-7 and decrease in ASC staining were detected after infection of HTB-9 cells with CY-17 and CY-92 for 1 hour, compared to uninfected control cells. NLRP-3 staining was weakly affected. (B) Quantification of total fluorescence intensity (open pin-hole) after subtraction of the background staining in uninfected cells (PBS). Medians ± SEMs of 50 cells, * P < 0.05, ** P < 0.01, *** P < 0.001, compared to PBS control, two-tailed unpaired t-test (MMP-7 and NLRP-3) or two-tailed Mann Whitney test (ASC). One of three experiments is shown. (C) Western blot confirming the change in cellular content of MMP-7, ASC and NLRP-3, 1 hour after infection with the indicated strains. Fold change compared to PBS of normalized values (against GAPDH). One experiment out of 2 is shown. (D) Increase in MMP-7 expression in HTB-9 cells transfected with siRNAs specific for ASC or NLRP3 and infected with CY-17 (4 hours, scale bars = 20 μm). (E) Western blot confirming the knock-down of ASC or NLRP-3 with siRNAs. A further reduction in ASC expression was detected after CY-17 infection (4 hours, quantified in S8B Fig, one experiment out of 2 is shown.). (F) PCR amplification of a 259 bp fragment in the MMP7 promoter (P1, -18/-276 relative to the transcription start site). (G) EMSA of the amplified fragment and nuclear extract from CY-17 infected HTB-9 cells (4 hours). Binding of ASC and NLRP-3 to P1 was identified as a band shift (arrow indicating protein-DNA complex). The band shift was inhibited by ASC- or NLRP-3-specific antibody. Free DNA formed a single low molecular weight band (arrow indicating free probe). The band shift was not affected by the IgG isotype control. One of three similar experiments is shown. (H) EMSA of the 259 bp MMP7 promoter fragment P1 and recombinant ASC. Dose-dependent formation of an ASC-P1 complex is shown as a band shift (arrow indicating ASC-DNA complex), which was inhibited by 0.5 and 1.0 μg of anti-ASC antibodies. The band shift was not affected by negative control murine IgG control. One of three similar experiments is shown.

https://doi.org/10.1371/journal.ppat.1005848.g005

ASC or NLRP-3 expression was subsequently inhibited by transfection of human bladder epithelial cells with ASC- or NLRP3- specific siRNAs and the effects on MMP-7 expression were examined by confocal imaging (Fig 5D and S8A Fig). MMP-7 expression increased drastically in transfected and infected cells, where the expression of ASC or NLRP3 had been inhibited, but not in cells transfected with negative control siRNA (Fig 5D). Inhibition efficiency of ASC and NLRP-3 expression by specific siRNAs was confirmed by Western blot analysis. Infection of the cells with CY-17 caused a further decrease in ASC and NLRP-3 staining (Fig 5E, quantified in S8B Fig). Two protein bands where detected, one of 24 kDa, corresponding to the common ASC variant and one of 20 kDa (ASC-b), corresponding to an ASC variant that enhances IL-1β secretion in human promyelocytic leukemia cells (HL60) [42] (Fig 5E).

To address if infection with cystitis strains modifies the interaction of ASC and NLRP-3 in cells, co-immunoprecipitation was performed. ASC was shown to pull down NLRP-3 in nuclear extracts of uninfected cells but after infection, a reduction in ASC/NLRP-3 interaction was detected suggesting that a loss of ASC/NLRP-3 interaction in the nuclear compartment accompanies MMP7 activation (S8C Fig).

To determine if ASC and NLRP-3 interact with the MMP7 promoter, DNA fragments spanning the entire promoter were used as probes in electrophoretic mobility shift assays (EMSA) (S9A Fig). A DNA fragment of 259 bp, adjacent to the transcription start site (P1, position -18/-276) was shown to interact with a nuclear protein extract from infected bladder cells, resulting in a significant band shift (Fig 5F and 5G). Specificity for ASC and NLRP-3 was confirmed by competition with specific antibodies (Fig 5G). In the absence of nuclear extract, the probe formed a single low molecular weight band, serving as a negative control. To confirm that ASC binds directly to the MMP7 promoter, recombinant ASC protein was incubated with the 259 bp DNA sequence and examined by EMSA. Strong dose-dependent binding of ASC to MMP7 promoter DNA was detected as a band shift, which was competitively inhibited by specific antibodies but not by the IgG isotype control (Fig 5H). Other MMP7 promoter sequences did not interact with ASC or NLRP-3 in this assay (S9A and S9B Fig).

The results suggesting that NLRP-3 and ASC act as negative regulators of MMP7 expression and identify an ASC binding site in MMP7 promoter DNA, adjacent to the transcription start site.

Therapeutic attenuation of the IL-1β response

To address if IL-1β serves as target for immunomodulatory therapy, we selected the most susceptible genotype (Asc^-/- mice) for treatment with the IL-1 receptor antagonist (IL-1RA) Anakinra. A dose of 1 mg per mouse in 100 μl of PBS was given intra-peritoneally, 30 minutes before infection and daily after infection with E. coli CFT073 (Fig 6A). This dose was selected based on previous studies in murine models [43].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Acute cystitis immunotherapy, using an IL-1 receptor antagonist (IL-1RA) or an MMP inhibitor.

(A) Overview of therapeutic regimen used to inhibit bladder pathology. Asc^-/- mice were pre-treated with Anakinra (IL-1RA), 30 min before infection and daily after infection with E. coli CFT073 (1 mg in 100 μl of PBS i.p. per mouse) and sacrificed 7 days after infection. Alternatively, Asc^-/- mice were pre-treated with the matrix metalloproteinase inhibitor (MMPI) Batimastat, 30 min before infection and daily after infection with E. coli CFT073 (0.5 mg in 100 μl of PBS i.p. per mouse, except day 3) (n = 7 per treatment group, total of two experiments). (B) Difference in gross bladder pathology between untreated controls and IL-1RA or MMPI treated mice. Two representative mice per group are shown. The IL-1RA therapy reduced macroscopic bladder pathology in Asc^-/- mice. Scale bar = 1 mm. The MMPI therapy showed a similar but less pronounced effect. (C) Edema, hyperemia and size of the bladders were used as scoring parameters for the pathology score. Pathology scores from individual mice are shown. The gross pathology score was reduced by the inhibitors (P < 0.001, for IL-1RA compared to untreated Asc^-/- mice and P = 0.002, for MMPI, compared to untreated Asc^-/- mice, means ± SEMs of two experiments, two-tailed Mann Whitney test). (D) Protection from bladder tissue pathology shown in H&E stained sections from treated versus control mice. Arrows indicate mucosal sloughing, edema and subepithelial abscesses in untreated mice. Inhibition of mucosal neutrophil aggregate formation in bladder sections from treated mice compared to untreated and infected mice. Scale bar = 200 μm (H&E) and 50 μm (immunofluorescence). (E) Kinetics of neutrophil recruitment and bacterial clearance in the urine of IL-1RA or MMPI treated Asc^-/- mice, compared to untreated mice. (n = 7 mice per group, means ± SEMs, * P < 0.05, ** P < 0.01, *** P < 0.001, two-tailed unpaired t-test).

https://doi.org/10.1371/journal.ppat.1005848.g006

A dramatic therapeutic effect was observed compared to infected Asc^-/- control mice. By macroscopic evaluation, the extent of edema, hyperemia and enlargement was reduced, resulting in a significantly lower pathology score (P < 0.001), (Fig 6B and 6C). By histology, a reduced inflammatory response was seen in the bladders of treated mice and mucosal pathology was inhibited compared to untreated controls that developed extensive bladder pathology (Fig 6D). Mucosal neutrophil infiltration, which accompanies pathology, was prevented and urine neutrophil numbers were low (Fig 6E). As a control for unspecific effects of Anakinra on the bacteria, CFT073 was grown in Luria-Bertani with or without 500 ng/ml of IL-1RA for 10 hours. No difference in bacterial growth rate was detected (S10 Fig).

We subsequently treated susceptible Asc^-/- mice with the matrix metalloproteinase inhibitor (MMPI) Batimastat. The MMPI was given 30 minutes before infection and on days 0–2 and 4–6 after infection (0.5 mg in 100 μl of PBS i.p., Fig 6A). The MMPI had a significant protective effect (Fig 6B–6D), detected by macroscopic evaluation, resulting in a reduced pathology score (P = 0.002). By histology neutrophil infiltration was reduced (Fig 6D). As in the IL-1RA-treated mice, bacterial numbers remained elevated (Fig 6E). Batimastat (250 ng/ml) did not affect bacterial growth in vitro for up to 10 hours. No difference in bacterial growth rate was detected (S10 Fig). As Batimastat is a broad metalloproteinase inhibitor, unspecific effects on other proteases might occur. Proteases inhibited by Batimastat other than MMP-7, were not transcriptionally regulated in any of the mice with acute cystitis or controls. MMP-15, which is not susceptible to Batimastat, was weakly activated in mice with bladder pathology (FC 2.0). These findings suggest that the therapeutic effect of Batimastat reflects inhibition of MMP-7.

The results confirm the importance of IL-1β and MMP-7 for the pathogenesis of acute cystitis and identify these molecules as functional targets for immunomodulatory therapy. Bacterial counts remained elevated in the IL-1RA and the MMPI treated mice in the absence of inflammation, suggesting that the treated Asc^-/- mice might develop a condition more like asymptomatic bacteriuria than acute cystitis (Fig 6E).

IL-1β and MMP-7 responses in patients with acute cystitis

To examine the human relevance of the findings in the murine UTI model, we collected urine samples from patients with acute cystitis or ABU and quantified the IL-1β and MMP-7 levels, by ELISA (Fig 7). Samples from patients with sporadic episodes of acute cystitis were collected at the time of diagnosis, defined by a positive dipstick, dysuria, urgency and frequency of urination but no fever (n = 9). Samples were also obtained from patients with ABU (n = 161), who carried the prototype ABU strain E. coli 83972, following therapeutic inoculation [21]. The patients with ABU participated in a prospective study of E. coli 83972-inoculation with detailed monthly collection of symptom scores and urine samples. There were 20 patients with low symptom scores and 161 urine samples were obtained from this group.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Elevated concentrations of IL-1β and MMP-7 in the urine of patients with acute cystitis.

(A) IL-1β concentrations in urine samples from patients with acute cystitis (n = 9). (B) IL-1β concentrations in consecutive urine samples from patients with ABU, who were long-term asymptomatic carriers of E. coli 83972 [21] (means ± SEMs, 20 patients, 161 urine samples). Elevated levels of IL-1β in the cystitis patients compared to the ABU group. Pie chart (inset) depicts the distribution of IL-1β concentrations in each patient group. (C) Histogram compares IL-1β concentrations between the two patient groups (means ± SEMs, *** P < 0.001, two-tailed Mann Whitney test). (D) MMP-7 concentrations were higher in urine samples from the patients with acute cystitis than in patients with long-term ABU (means ± SEMs, *** P < 0.001, two-tailed unpaired t-test). Urine samples were obtained from patients with sporadic acute cystitis at the time of diagnosis. The patients with ABU participated in a prospective study of therapeutic inoculation with E. coli 83972 and were subjected to long-term follow up [21]. Multiple samples were obtained during asymptomatic carriage (3–14 samples per patient).

https://doi.org/10.1371/journal.ppat.1005848.g007

We found elevated concentrations of IL-1β in patients with acute cystitis compared to the asymptomatic patient group (Fig 7A–7C) resulting in a means of 264.5 pg/ml and 1.5 pg/ml, respectively (P < 0.001).

In addition, all the patients with acute cystitis had positive MMP-7 levels, above the detection limit of 0.15 ng/ml (Fig 7D). In a subset of 28 ABU urine samples, the mean MMP-7 concentration was low, resulting in mean concentrations of 15.4 ng/ml and 4.3 ng/ml, respectively (P < 0.001).

The results show that patients with acute cystitis have more elevated concentrations of IL-1β and MMP-7 in urine, than patients with ABU, identifying IL-1β and MMP-7 as potential biomarkers of acute cystitis.

Bacterial strains and cell infection procedure

Cystitis (CY) and asymptomatic bacteriuria (ABU) strains were prospectively isolated during a study of childhood UTI in Göteborg, Sweden, using standard microbiological techniques [32, 33]. The strain collection has been extensively studied and characterized with fimbrial genotype and phenotype, virulence factor expression, OKH antigen profiles and multilocus enzyme typing [31, 68]. The hemolytic activity was assessed with blood agar plates, where the hemolytic zone surrounding the central stab of bacteria is recorded. The phenotype has been compared to the hly genotype and found to be a very close fit. The UPEC strain, E. coli CFT073 (O6:K2:H1) [69] and the ABU strain E. coli 83972 (OR:K5:H-) [25] have been extensively characterized, including whole genome sequencing and were used as positive or negative controls. Bacteria were cultured on tryptic soy agar (TSA, 16 h, 37°C), harvested in phosphate-buffered saline (PBS, pH 7.2) and diluted as appropriate in RPMI without FCS. In the screen of IL-1β responses to acute cystitis or ABU strains, cells were exposed to 10⁸ CFU/ml of bacteria with Gentamicin for 4 hours. In remaining experiments, cells were exposed to 10⁵ CFU/ml for 1 hour or 4 hours without antibiotics. Overnight static cultures of E. coli CFT073, CY-17 or CY-92 or 83972 in Luria-Bertani (LB) broth were used for experimental infection.

Cell culture

Human bladder grade II carcinoma cells (5637, ATCC# HTB-9) were cultured, to 70–80% confluency on 8-well chamber Permanox slides (6x10⁴ cells/well), in 6-well plates (6x10⁵ cells/well) or 96-well plates (5x10⁴ cells/well), (all from Thermo Scientific) in RPMI-1640 supplemented with 1 mM sodium pyruvate, 1 mM non-essential amino acids and 10% heat-inactivated FBS (PAA) at 37°C with 5% CO₂. Gentamicin (50 μg/ml) was from GE Healthcare.

Cell viability assay

HTB-9 cells in 96-well plates were infected for 1h or 4h. PrestoBlue (Invitrogen, A13262) was added to each well to a final concentration of 10%. After 20 min of incubation at 37°C, total well fluorescence was measured using a microplate reader Infinite F200 (Tecan), with 585 nm excitation and 620 nm emission filters.

Cytokine measurements

IL-1β concentrations in filtered supernatants (Syringe Filter w/0.2 μm PES, VWR) from cells infected with 10⁸ CFU/ml (with Gentamicin, 4 hours) were determined by Immulite 1000 (Siemens) and IL-1β concentrations in cell supernatants or urine by Human or Mouse IL-1β/IL-1F2 DuoSet ELISA kits (all from R&D Systems). Urine MMP-7 levels were quantified with Human total MMP-7 Immunoassay Quantikine ELISA (R&D Systems).

Confocal microscopy

Cells were infected, fixed (3.7% formaldehyde, 10 min), permeabilized (0.25% Triton X-100, 5% FBS, 15 min), blocked (5% FBS, 1h at RT), incubated with primary antibodies in 5% FBS overnight at 4°C (anti-IL-1 beta, 1:100, ab9722; anti-MMP7, 1:25, ab4044, all Abcam; anti-ASC, 1:50, sc-22514-R, Santa Cruz; anti-NLRP3/Cryo-2, 1:100, AG-20B-0014-C100, Adipogen) and appropriate secondary antibody (Alexa Fluor 488 goat anti-rabbit IgG, A-11034, or goat anti-mouse IgG, A-11001; Life Technologies), (1h at RT). After nuclear staining (DRAQ5, Abcam), slides were mounted (Fluoromount, Sigma-Aldrich), imaged by laser-scanning confocal microscopy (LSM510 META confocal microscope, Carl Zeiss) and quantified by ImageJ software 1.46r (NIH).

Western blotting

Cells were lysed with RIPA lysis buffer, supplemented with protease and phosphatase inhibitors (both from Roche Diagnostics) and fractionated using the NE-PER Nuclear and Cytoplasmic extraction reagents (Thermo Scientific). Supernatants were filtered and concentrated by trichloroacetic acid precipitation, followed by aceton desiccation. Proteins were run on SDS-PAGE (4–12% Bis-Tris gels, Invitrogen), blotted onto PVDF membranes (GE Healthcare) blocked with 5% bovine serum albumin (BSA) or non-fat dry milk (NFDM), incubated with primary antibody: rabbit anti-IL-1 beta (1:2,500 in 5% NFDM, ab9722, Abcam), rabbit anti-ASC (1:200 in 5% BSA, sc-22514-R, Santa Cruz), mouse anti-NLRP3/NALP3 (1:1,000 in 5% milk, Cryo-2, Adipogen) or rabbit anti-MMP7 (1:200 in 5% BSA, ab4044, Abcam), washed with PBS tween 0.1% and incubated with secondary antibodies in 5% NFDM (goat anti rabbit-HRP or goat anti-mouse-HRP, Cell Signaling). Bands were imaged using ECL Plus detection reagent (GE Health Care) and quantified using ImageJ. GAPDH (1:1,000, sc-25778, Santa Cruz) was used as loading control.

Co-immunoprecipitation

Nuclear fractions, extracted as described in Western blotting, were incubated with rabbit anti-ASC antibody (sc-22514-R, Santa Cruz, 1 μg/ml) overnight and complexes were collected with magnetic Dynabeads Protein G (Life technologies), analyzed by SDS-PAGE with rabbit anti-ASC and mouse anti-NLRP3 (Cryo-2, Adipogen) primary antibodies (1:200–1:1,000, 5% BSA), followed by secondary anti-rabbit (Cell Signaling) or anti-mouse (DAKO) antibodies (1:4,000, 5% NFDM).

Global gene expression

Total RNA was extracted from murine bladders or kidneys in RLT buffer with 1% β-Mercaptoethanol after disruption in a tissue homogenizer (TissueLyser LT, Qiagen) using Precellys Lysing kits (Bertin Technologies), with the RNeasy Mini Kit (Qiagen), 100 ng of RNA was amplified using GeneChip 3´IVT Express Kit, 6 μg of fragmented and labeled aRNA was hybridized onto Mouse Genome 430 PM array strips for 16 hours at 45°C, washed, stained and scanned using the Geneatlas system (all Affymetrix). All samples passed the internal quality controls included in the array strips (signal intensity by signal to noise ratio; hybridization and labeling controls; sample quality by GAPDH signal and 3’-5’ ratio < 3).

Transcriptomic data was normalized using Robust Multi Average implemented in the Partek Express Software (Partek) [70, 71]. Fold change was calculated by comparing infected (7 days) to uninfected mice of the same genetic background. Significantly altered genes were sorted by relative expression (2-way ANOVA model using Method of Moments, P-values < 0.05 and absolute fold change > 1.41) [38]. Heat-maps were constructed by Gitools 2.1.1 software. Differentially expressed genes and regulated pathways were analyzed by Ingenuity Pathway Analysis software (IPA, Ingenuity Systems, Qiagen). Qiagen’s list of 84 key inflammasome genes was selected for analysis.

The microarray data are available in the NCBI’s Gene Expression Omnibus repository (accession number GSE86096).

In vitro proteolysis

Recombinant human IL-1β, NLRP-3 or PYCARD (ASC) (280 ng, H00003553-P02, H00114548-P01 or H00029108-P01, Abnova) were incubated with recombinant active human MMP-7 (0.035U, #444270 Merck Millipore) in MMP reaction buffer (20 mM Tris, pH 7.6, 5 mM CaCl₂, 0.1 M NaCl) at 37°C until stopped with 100 mM DDT. Fragments were detected by Western blot, using rabbit anti-IL-1 beta (1:2 000, ab9722, Abcam), rabbit anti-ASC (1:1 000, p9522-75, US Biological) and rabbit anti-NLRP3 (1:500, sc-66846, Santa Cruz).

IL-1β activity assay

HTB-9 cells were treated with the products of the in vitro proteolysis of pro-IL-1β by MMP-7 at different concentrations or with pro-IL-1β or MMP-7 alone, serving as negative controls. Prostaglandin E2 (PGE₂) concentrations were measured in filtered supernatants (Syringe Filter w/0.2 μm PES, VWR) by ELISA (R&D systems).

siRNA transfection

HTB-9 cells were transfected with PYCARD/ASC and NLRP3 specific siRNAs (0.09 μM, FlexiTube GeneSolution, #GS29108 and #GS114548, Qiagen) or with AllStars Negative Control siRNA (#SI03650318, Qiagen) using the HiPerFect Transfection Reagent (#301705, Qiagen) for 17 hours, then infected. Transfection efficiency was assessed by Western blotting.

PCR analysis

MMP7 promoter and promoter flanks were amplified in 10 different fragments by PCR using 15 ng of total human genomic DNA. For forward and reverse primers (http://primer3.ut.ee/), see S3 Table. Thermal cycling conditions were as follows: 95°C for 2 min, 35 cycles (95°C for 30 s, 60°C for 30 s and 72°C for 40 s) and 72°C for 5 min.

Electrophoretic mobility shift assay (EMSA)

Amplified DNA sequences from the MMP7 promoter were used as probes and labeled with GelGreen (Biotium). Each reaction contained 3–5 μg of DNA probe with, 5 μg of nuclear extract from infected HTB-9 cells, or 0.2–0.65 μg recombinant ASC (Abnova, H00029108-P01) or NLRP-3 (Abnova, H00114548-P01) in binding buffer (100 mM Tris, 500 mM NaCl and 10 mM DTT, pH 7). For the band shift competition assay, 0.5–1 μg of rabbit anti-ASC (Santa Cruz, sc-22514-R,) or 0.5 μg of rabbit anti-NLRP3 (Cryo-2, Adipogen) antibodies were used. Binding reactions were incubated at 15°C for 30 min, loaded onto a 6% non-denaturing, non-reducing polyacrylamide gel and ran in a 50 mM Tris (pH 7), 0.38 M glycine, and 2 mM EDTA buffer at 100 V for 2–3 hours. Mouse IgG2A isotype control (R&D Systems, MAB003) was used as negative control antibody. Gels were imaged using the Bio-RAD ChemiDoc system.

Experimental urinary tract infection

Mice were bred and housed in the specific pathogen-free MIG animal facilities (Lund, Sweden) with free access to food and water. Female C57BL/6 mice or Il1b^-/- [36], Nlrp3^-/- [34] Asc^-/- [35], Casp1^-/- [37], Mmp7^-/- [40] mice were used at 9–15 weeks of age. The Il1b^-/- mice have recently been shown to be functionally defective for IL-1α [72]. The Casp1^-/- mice were also deficient for Caspase-11 [73]. Nlrp3^-/- and Asc^-/- mice were from Jürg Tschopp's laboratory, Department of Biochemistry, University of Lausanne and Institute for Arthritis Research (aIAR). Mmp7^-/- and Casp1^-/- mice were purchased from The Jackson Laboratories, USA.

Mice were intravesically infected under Isofluorane anesthesia (10⁸ CFU in 0.1 ml), through a soft polyethylene catheter (outer diameter 0.61 mm; Clay Adams). Animals were sacrificed under anesthesia; bladders and kidneys were aseptically removed and macroscopic pathology was documented by photography. Tissues were fixed with 4% paraformaldehyde or frozen for sectioning and RNA extraction. Viable counts in homogenized tissues (Stomacher 80, Seward Medical) were determined on TSA (37°C, overnight). Urine samples were collected prior to and at regular times after infection and quantitatively cultured. Neutrophils in uncentrifuged urine were counted, using a hemocytometer.

Gross pathology was scored based on the macroscopic appearance of the bladders at sacrifice. The score was based on edema, hyperemia and size, on a scale of 0–10, where 0 is unchanged compared to the uninfected controls and 10 is most edematous, most hyperemic and largest size.

Histology and immunohistochemistry

Tissues were embedded in O.C.T. compound (VWR) and 5-μm-thick fresh cryosections on positively charged microscope slides (Superfrost/Plus; Thermo Scientific) were fixed with 4% paraformaldehyde or acetone-methanol (1:1 v/v). For H&E or immunohistochemistry, sections were blocked and permeabilized (0.2% Triton X-100, 5% goat normal serum (DAKO) or 1% BSA (Sigma), stained (anti-neutrophil antibody [NIMP-R14] (ab2557, Abcam), polyclonal E. coli antibody (1:100, NB200-579, Novus Biologicals), anti-IL-1 beta (1:50, ab9722, Abcam) or anti-MMP-7 (1:100, ab4044, Abcam), all rabbit antibodies). Alexa 488 anti-rat IgG or anti-rabbit IgG and Alexa 568 anti-rabbit IgG (A-21210, A-11001 and A-11011, Life Technologies) were secondary antibodies and nuclei were counterstained with DAPI (0.05 mM, Sigma-Aldrich). Imaging was by fluorescence microscopy (AX60, Olympus Optical). Richard-Allan Scientific Signature Series Hematoxylin 7211 and Eosin-Y 7111 (Thermo Scientific) were used to counterstain the tissue sections.

Histology was scored using H&E stained bladder sections. The score was based on neutrophil infiltration, tissue architecture and epithelial thickness on a scale of 0–10, where 0 is unchanged compared to uninfected controls and 10 the highest neutrophil infiltration, most destroyed tissue architecture and maximum epithelial thickness.

IL-1β and MMP-7 therapy

The IL-1 receptor antagonist, Anakinra (Kineret, SOBI) or the broad-spectrum MMP inhibitor, Batimastat (ab142087, Abcam) were injected intraperitoneally (i.p.) as described in Fig 6A.

Patients

Urine samples from patients with sporadic acute cystitis were obtained at two primary care clinics in Lund, Sweden. A diagnosis of acute cystitis was based on a urine dipstick analysis positive for bacteria and symptoms from the lower urinary tract, including frequency, dysuria and suprapubic pain. Midstream urine specimens were obtained at the time of diagnosis.

Patients with ABU were included in a placebo-controlled study of asymptomatic bacteriuria, following intravesical inoculation with E. coli 83972 [21]. Briefly, E. coli 83972 bacteriuria was established by intravesical inoculation (10⁵ CFU/ml in saline), daily for three days and the outcome was measured as the total number of UTIs during an optimal period of 12 months followed by a cross over to a similar period without E. coli 83972 bacteriuria. Urine samples were obtained for cytokine analysis during E. coli 83972 bacteriuria, with negative symptom scores [21].

Statistics

ELISA results, fluorescence intensity, Pathology, bacterial numbers and neutrophil responses were analyzed by unpaired two-tailed t-test or Mann-Whitney test after assessment of normality with the d’Agostino & Pearson omnibus normality test. Significance was accepted at P < 0.05 (*), P < 0.01 (**) or P < 0.001 (***). For animal numbers, see S1 Table. Data was examined using Prism (v. 6.02, GraphPad).

Ethics statement

Experimental infections were approved by the Malmö/Lund Animal Experimental Ethics Committee at the Lund District Court in Sweden (approval number M44-13). All animal care and protocols were governed by the European Parlement and Council Directive 2010/63/EU, the Swedish Animal Welfare Act (Djurskyddslag 1988:534), the Swedish Animal Welfare Ordinance (Djurskyddsförordning 1988:539) and Institutional Animal Care and Use Committee (IACUC) guidelines. Experiments were reported according to the ARRIVE guidelines. The clinical studies were approved by the Human Ethics Committee at Lund University (approval numbers LU106-02, LU236-99 and Clinical Trial Registration RTP-A2003, International Committee of Medical Journal Editors, www.clinicaltrials.gov). Patients gave their informed written consent.