Palmitate Induced IL-6 and MCP-1 Expression in Human Bladder Smooth Muscle Cells Provides a Link between Diabetes and Urinary Tract Infections

Background Urinary tract infections (UTI) are more frequent in type-2 diabetes mellitus patients than in subjects with normal glucose metabolism. The mechanisms underlying this higher prevalence of UTI are unknown. However, cytokine levels are altered in diabetic patients and may thus contribute to the development of UTI. Increased levels of free fatty acids (FFA), as observed in obese patients, can induce IL-6 production in various cell types. Therefore we studied the effects of the free fatty acid palmitate and bacterial lipopolysaccharide (LPS) on interleukin-6 (IL-6) and monocyte chemotactic protein-1 (MCP-1) expression and secretion in cultured human bladder smooth muscle cells (hBSMC). Methodology/Principal Findings Biopsies were taken from patients undergoing cystectomy due to bladder cancer. Palmitate or LPS stimulated hBSMC were analysed for the production and secretion of the IL-6, gp80, gp80soluble, gp130, MCP-1, pSTAT3, SOCS3, NF-κB and SHP2 by quantitative PCR, ELISA, Western blotting, and confocal immunofluorescence. In signal transduction inhibition experiments we evaluated the involvement of NF-κB and MEK1 in IL-6 and MCP-1 regulation. Palmitate upregulates IL-6 mRNA expression and secretion via NF-κB dependent pathways in a concentration- and time-dependent manner. MCP-1 was moderately upregulated by palmitate but was strongly upregulated by LPS involving NF-κB and MEK1 dependent pathways. Soluble IL-6 receptor (gp80soluble) was downregulated by palmitate and LPS, while membrane-bound gp80 was moderately upregulated. LPS increased SOCS3 and SHP2, whereas palmitate only induced SOCS3. Secondary finding: most of the IL-6 is secreted. Conclusions/Significance Bacterial infection (LPS) or metabolic alterations (palmitate) have distinct effects on IL-6 expression in hBSMC, (i) short term LPS induced autocrine JAK/STAT signaling and (ii) long-term endocrine regulation of IL-6 by palmitate. Induction of IL-6 in human bladder smooth muscle cells by fatty acids may represent a pathogenetic factor underlying the higher frequency and persistence of urinary tract infections in patients with metabolic diseases.


Introduction
Urinary tract infections (UTI) are more frequent in patients with diabetes mellitus than in subjects with normal glucose metabolism and take a more severe course [1,2]. Women with diabetes require longer and more aggressive antimicrobial treatment for UTI, and have more recurrences of their UTI than non-diabetic women. The hospitalization due to complications of the UTI occurs significantly more often in women with diabetes [3]. Urine cytokine levels and an increased adherence of the microorganisms to the uroepithelial cells seem to be the main predictors of increased prevalence of both asymptomatic and symptomatic bacteriuria in diabetic patients [4]. Whereas some authors suggest glucosuria as diabetes-specific variable potentially associated with symptomatic infection [5], no association between bacteriuria and indicators of glycemic control, such as the blood glucose level and the glycosylated hemoglobin value, have been found by others [6]. In addition to chronic hyperglycemia, altered fatty acid metabolism belongs to the metabolic alterations associated with type-2 diabetes [7]. However, it is not known whether alterations in circulating free fatty acids (FFA) may contribute to the increased UTI frequency in patients with diabetes.
Bacterial UTI lead to upregulation of cytokines and growth factors and recruitment of inflammatory cells by LPS [8,9]. IL-6 was shown to be the single most prominent cytokine detected in the urine patients with UTI [10]. Low-grade chronic inflammation is reflected by a 2-3-fold increase in the systemic levels of certain cytokines [11] as well as C-reactive protein (CRP), and an association has been confirmed between low-grade systemic inflammation and type-2 diabetes [12].
There is growing evidence that increased levels of FFA can induce IL-6 production in various cell types, and might therefore be involved in the pathophysiology of UTI [13][14][15]. Elevated FFAs in obese patients may provide a mechanistic link between increased fat mass and the development of insulin resistance, glucose intolerance, and beta-cell dysfunction that promote the onset of diabetes [7,16]. Palmitate induced accumulation of IL-6 is regulated dependently of c-Jun N-terminal kinase in 3T3-L1 adipocytes [17], but is regulated via NF-kB in myotubes [13], pointing to the existence of potential tissue/cell specific regulatory mechanisms in inflammation. In addition, palmitate could regulate monocyte chemotactic protein-1 (MCP-1) expression in adipose tissue [18], human vascular umbilical vein cells (HUVEC) and rat vascular smooth muscle cells [19]. . IL-6 protein content (A) was measured in cytosol (white bars, ng/mg) and supernatant (black bars, ng/ml), while MCP-1 protein was measured exclusively in supernatant (C, ng/ml). All bars indicate difference to medium control. For each measurement a medium treated control was used. Data are shown as mean and SEM. Significant differences are indicated by lines. mRNA (B, D) was normalized to natural logarithm LN. doi:10.1371/journal.pone.0010882.g001 The aim of the present study was to investigate IL-6 and MCP-1 regulation in hBSMC by acute bacterial infection (LPS stimulation) and metabolic alterations (palmitate) as potential risk factor for chronic bladder inflammation.

Results
Treatment of hBSMC with FFA palmitate (up to 48 hrs) resulted in significantly increased IL-6 protein concentrations in the supernatant, but not in the cytosol (Fig. 1A) A similar pattern was observed for palmitate stimulated IL-6 mRNA expression (Fig. 1B). Increased IL-6 secretion into the medium was dependent on palmitate concentration and time.
In contrast, stimulation with LPS led to IL-6 protein upregulation in the supernatant while levels decreased with time in the cytosol ( Fig. 2A). There was no concentration-dependent regulation of IL-6 mRNA (Fig. 2B). In addition total IL-6 expression was only about 25% of IL-6 protein produced upon palmitate stimulation ( Fig. 2A).
While MCP-1 showed similar regulation of the mRNA (Fig. 1D), this concentration-and time-dependent upregulation was not . IL-6 protein content (A) was measured in cytosol (white bars, ng/mg) and supernatant (black bars, ng/ml), while MCP-1 protein was measured exclusively in supernatant (C, ng/ml). All bars indicate difference to medium control. For each measurement a medium treated control was used. Data are shown as mean and SEM. Significant differences are indicated by lines. Post-hoc Bonferroni test was used after ANOVA. Significance level was p,0.05. mRNA (B, D) was normalized to natural logarithm LN. doi:10.1371/journal.pone.0010882.g002 reflected by MCP-1 protein content. Protein expression was constant over 48 hrs and was highest at low palmitate concentration (Fig. 1C). Upon LPS stimulation the MCP-1 protein in the supernatant and in the cytosol decreased time-and concentrationdependent (Fig. 2C). MCP-1 mRNA showed a comparable downregulation with time but only moderate decrease with increasing concentrations of LPS (Fig. 2D).
We found a transient upregulation of the mRNA of IL-6 receptor protein gp80 and its soluble form gp80 soluble (Fig. 3A). After 16 hrs of incubation the gp80 mRNA was still high while the mRNA of gp80 soluble was significantly downregulated. This opposite regulatory effect on mRNA levels was most prominent with low palmitate concentration (0.125 mM) but was still present at higher concentrations (Fig. 3A). Gp80 protein was significantly downregulated as early as after 8 hrs incubation lasting up to 24 hrs. However, after 48 hrs incubation with 0.25 mM palmitate the effect was reversed. This high level expression of gp80 was concentration-independent (Fig. 3B).
The effects of LPS stimulation on gp80 and gp80 soluble mRNA are comparable to palmitate induced regulation (Fig. 4A, C), except for a significant initial upregulation of gp80 and gp80 soluble after 8 hrs. After 48 hrs there was an opposite regulation of cytosolic gp80 protein, which was increased by palmitate (0.25 mM) but decreased by LPS (1 mg/ml). Prominent downregulation of pSTAT3 protein was evident in LPS stimulated cells on a short time scale up to 6 hrs ( Fig. 4B), whereas there was an upregulation of pSTAT3 protein after 8 hrs of palmitate stimulation (Fig. 3B).
The membrane IL-6 receptor protein gp130 showed only moderate mRNA upregulation (Fig. 3C). Both, STAT3 mRNA and protein were significantly downregulated after 24 hrs incubation with 0.25 mM palmitate, which was the most effective concentration at 48 hrs, too (

SOCS3, SHP2
Since SOCS3 and SHP2 are important regulators of IL-6 signaling pathways, we further examined their regulation by palmitate and LPS. SOCS3 mRNA and SOCS3 protein increased after 48 hrs of palmitate treatment (Fig. 6A, B). SHP2 mRNA and SHP2 protein levels decreased (Fig. 6C, D). Interestingly, LPS caused upregulation of SOCS3 mRNA in the first 6 hrs of treatment, but levels returned to baseline level after 48 hrs, while SOCS protein content was highest after 48 hrs (Fig. 6E, F). SHP2 mRNA and protein levels were unaltered after LPS treatment (Fig. 6G, H). These findings were confirmed by confocal immunofluorescence. We found highest immunoreactivity for pSTAT3, SOCS3, and SHP2 after palmitate treatment, while the changes were only moderate in LPS treated cultures (Fig. 5).

NF-kB p65 regulation and activation
NF-kB p65 is the main transcription factor regulating IL-6 gene expression [20]. We found concentration-and time-dependent downregulation of the NF-kB p65 mRNA (Fig. S1A, B) and pNF-kB p65 protein (Fig. 7A, B). This was in striking contrast to the effect of LPS, which led to concentration-dependent upregulation of both, NF-kB p65 mRNA (Fig. S1B) and pNF-kB p65 protein (Fig. 7B).
In double confocal immunofluorecence experiments we found time-dependent translocation of NF-kB p65 into the nuclei as evidence of NF-kB activation after 48 hrs of LPS (1 mg/ml) and palmitate (0.25 mM) stimulation. We found more NF-kB p65 in nuclei of palmitate stimulated cells than after LPS stimulation. To quantify this translocation we performed dot blot analysis using a specific antibody against phosphorylated (Ser536) activated NF-kB p65. Basic ratio was 2.760.66 (mean6SD) in medium control and moderate higher in LPS stimulated cells (3.560.92), while the ratio was significant higher after 48 h of palmitate stimulation (5.761.1).

Discussion
Here we studied the effects of palmitate and LPS on human bladder smooth muscle cells with the aim to elucidate potential novel mechanisms underlying the observation that urinary tract infections are more frequent in patients with diabetes mellitus and take a more severe course [1,2]. Defects in the local urinary cytokine secretions and an increased adherence of pathogenic microorganisms to the uroepithelial cells have been discussed as mechanisms for increased prevalence of UTI in patients with diabetes [4]. In a proteom study Yohannes et al. [21] found significant alterations in rat detrusor muscle after streptozotocin-induced diabetes, suggesting that diabetes depending bladder dysfunction may involve dysregulation of structural cell integrity, cell adhesion, proliferation and inflammation [21].
It is well known that during acute infection LPS triggers intracellular signaling cascades via Toll-like receptor TLR4 that rapidly induce inflammatory cytokines that initiate a variety of overlapping immune responses [22].
However, fatty acids and their metabolites act also directly and indirectly to regulate metabolism and immune function through their interactions with specific enzymes [23]. Exaggerated and prolonged postprandial hyperlipidemia is an important characteristic of diabetic dyslipidemia [7,24].
In the present study we report that palmitate treated human bladder smooth muscle cells (hBSMC) are a source of MCP-1 and IL-6. We used specific inhibitors to interfere with the signaling cascade. We found regulation of IL-6 via NF-kB, which is in line with the regulation of IL-6 by palmitate stimulation in human skeletal muscle cells as reported earlier [13].
Unexpectly, pNF-kB p65 protein content was significantly lower in palmitate stimulated cells as compared to medium control (Fig. 7A). This finding needs further investigation, since on the other hand we found short-term upregulation of pNF-kB p65 by LPS (Fig. 7B). One might suggest that the palmitate stimulus is much stronger than LPS and might therefore lead to increased protein turn-over.
MCP-1 also showed NF-kB dependent regulation as recently reported for IL1b stimulated human aorta smooth muscle cells  [25]. However, MCP-1 was downregulated by MEK1 inhibitor (PD98059) in hBSMC, indicating regulation via MEK1 pathway in contrast to IL-6. It is well known that MCP-1 promotes migration of monocytes [26] and this mechanism has also been proposed for bladder inflammation [27]. MCP-1 is also related to insulin resistance and the risk of metabolic syndrome [28].
In hBSMC we found higher secretion of MCP-1 after LPS than after palmitate stimulation. In contrast IL-6 secretion was much higher in palmitate stimulated cells than after LPS treatment (Fig. 1, Fig. 2). To the best knowledge of the authors this is the first study to show that the vast amount of palmitate stimulated IL-6 is secreted. Only basic cytosolic amounts are detectable (Fig. 1). This supports the view that the human bladder smooth muscle cells are involved in immunological processes and may have endocrine function. To examine possible regulatory function of IL-6 onto bladder smooth muscle cells we studied the IL-6 receptor pathway in hBSMC.
The IL-6 signaling cascade depends on the common cytokine receptor family gp130 subunit and the IL-6 receptor specific gp80 subunit forming a hexameric receptor-IL-6 complex. Gp80 exists in a membrane-bound and soluble form, produced as alternative splice variant [29]. For long-term autocrine stimulation by IL-6 the whole receptor complex is needed. We could show, that gp80 mRNA and protein were time-dependent down-regulated by both, palmitate and LSP (Fig. 3, 4). This implies preferential exocrine cytokine signaling by BSMC.
We found short term upregulation of pSTAT3 in LPS and palmitate stimulated cells, but long term downregulation of pSTAT3 (Fig. 3, 4) in conjunction with upregulation of SOCS3 (Fig. 5), which implies negative feedback to IL-6 signaling as reported by Croker et al. [30]. Hence it is evident that both, LSP and palmitate also induced short-term autocrine regulation of IL-6 receptor pathway. However, conditional knockout studies have shown that SOCS3 depletion can induce a number of inflammatory and metabolic disorders [30]. Therefore, upregulation of SOCS3 could be a protection mechanism for the hBSMC to cope with chronically elevated IL-6. At present it is not known which genes are regulated by pSTAT3 in hBSMC.
In conclusion, there is a short-term autocrine IL-6 signaling of BSMC, but a long-term exocrine IL-6 signaling providing cytokine stimulation to local or invading cells equipped with the IL-6 receptor complex.
Our study provides evidence for the urinary bladder smooth muscle cells being an immunologically active tissue, which can be triggered by inflammatory conditions, but which is especially sensitive to metabolic alterations as provoked by the saturated FFA palmitate. In the light of our findings, chronically elevated FFA could well be a risk factor for urinary infections and idiopathic bladder inflammation in diabetes.

Cell Culture
Human bladder smooth muscle cell (hBSMC) cultures were established from human detrusor muscle after radical cystectomy of tumour patients. Written informed consent was obtained from all patients. The study was approved by the ethics committee of the University of Leipzig (Reg.No. 773). Cells were grown in SMC Growth Medium 2 (PromoCell, Heidelberg, Germany) and subcultured up to the forth passage (P4). At 80% confluence cells were treated with palmitate or with LPS. Palmitate was used in combination with 2% bovine serum albumin (BSA) as carrier of free fatty acids. Medium treated cells were used as control. For pathway analyses, cells were preincubated for 1 h with 20 mM PD98059 (Map Kinase Kinase/Erk Kinase 1 inhibitor) or 40 mM MG123 (proteasome inhibitor; Calbiochem, San Diego, USA). If not indicated otherwise all chemicals were from Sigma-Aldrich, Steinheim, Germany.

RNA Extraction and Real Time PCR
Total RNA was isolated with RNeasy Mini kit (Qiagen, Hilden, Germany). Quantitative PCR was performed with the ABI 7500 Real Time PCR Instrument (Applied Biosystems, Foster, USA) using the Brilliant SYBR Green QPCR Core Reagent kit (Stratagene, La Jolla, USA) and custom primers (MWG-Biotech, Ebersberg, Germany; Table 1). Human 36B4 (acidic ribosomal phosphoprotein P0) served as internal standard.

Western Blotting
Cells were washed with PBS and scrapped off in 16 lyses buffer (Cell Signaling, Danvers, USA) supplemented with 1 mM PMSF and 1/100 volume protease inhibitor cocktail (Sigma-Aldrich). After sonication, the cell extracts were cleared by centrifugation. 75 mg of total protein were separated by sodium dodecyl sulfate polyacrylamide (10%) gel electrophoresis and transferred to polyvinylidene fluoride (PVDF) membrane by tank blotting (Mini-PROTEAN II; BioRad, Munich, Germany). Blots were incubated with primary antibodies for 2 h at room temperature or over night at 4uC. Detection and visulization was done with horseradish peroxidase-conjugated anti-rabbit or anti-mouse IgG (1:1000) for 1 h at room temperature and DAB (FAST 3,39-diaminobenzidine tablets) as a substrate.

Dot Blot Analysis
Nuclear and cytosolic extracts were prepared according to Yerneni et al. 1999 [31]. 2 mg total protein of nuclear or cytosolic extracts were transferred in duplicates on nitrocellulose membrane by Dot Blotting (Dot Blot 96 System, Biometra, Goettingen, Germany). After blocking with Odyssee blocking buffer (Licor Biosciences, Bad Homburg, Germany) for 1 h the membranes were incubated with anti-Phospho-NF-kB p65 (Ser536) rabbit IgG (1:1000; Cell Signaling, Danvers, USA) over night at 4uC. Detection was done with anti-rabbit IRDye 680 (1:5000; Licor Biosciences) for 2 h. Membranes were scanned with Odyssey Infrared Imager and evaluated by Odyssey Infrared Imaging Software 3.0 (Licor Biosciences). Total protein was visualized by SYPRO Ruby blot stain (BioRad).

Enzyme-linked Immunosorbent Assay (ELISA)
Intracellular and secreted IL-6 was quantified by Quantikine human IL-6 immunoassay (R&D Systems, Minneapolis, MN). Cell extracts were diluted 1:10 with calibrator diluent. IL-6 in cell culture supernatants were measured after being centrifuged in 1:100 diluted samples. IL-6 receptor (gp80) and activated transcription factors (NF-kB, pSTAT3) were determined in undiluted cell extracts by Quantikine human IL-6 sR immunoassay (R&D Systems) and by PathScan Phospho-NF-kB p65 (pNF-kB; Ser536) and PathScan Phospho-STAT3 (pSTAT3; Tyr705) Sandwich ELISA (Cell Signaling). MCP-1 was measured in cell culture supernatants and cell extracts by Quantikine human CCL2/MCP-1 immunoassay (R&D Systems). For quantification Figure 6. Palmitate and LPS effects on SOCS3 and SHP2. Time-and concentration-dependent palmitate effects on SOCS3 mRNA (A), SHP2 mRNA (C), SOCS3 protein expression (B), and SHP2 protein expression (D). LPS effects on SOCS3 mRNA (E), SHP2 mRNA (F), SOCS3 protein expression (G), and SHP2 protein expression (H). Protein content was analysed in the cytosol using Western blotting analysis. All bars indicate difference to medium control. For each measurement a medium treated control was used. Data are shown as mean and SEM. Significant differences are indicated by lines. mRNA was normalized to natural logarithm LN. doi:10.1371/journal.pone.0010882.g006 Figure 7. Gene regulation and pathway analysis. Time-and concentration-dependent palmitate and LPS effects on pNF-kB protein expression (A, B). Protein content of pNF-kB was measured in cytosol (A, black bars) and is indicated as difference to medium control (AU in pg/mg) by ELISA. All bars indicate difference to medium control (A, B). For each measurement a medium treated control was used. Confocal double immunofluorescence images demonstrate nuclear translocation of NF-kB (green) and IL-6 (red). Nuclei are depicted by DAPI (blue). Dot blot analysis of phosphorylated NF-kB are shown as ratio of nuclear pNF-kB protein content to cytosolic pNF-kB protein content related to total protein content. Cells were stimulated for 48 h either with 0.25 mM palmitate or with 1 mg/ml LPS (C). Pathway analysis of palmitate induced IL-6 (D) and MCP-1 (E) regulation. Protein content was measured by ELISA in supernatants (ng/ml). Cells were stimulated with palmitate (0.5 mM) and inhibitor of proteasomal degradation of NF-kB (MG132, 40 mM) or MEK1 (PD98059, 20 mM) for 48 hrs. Medium control is indicated as black bar (D, E). Data are shown as mean and SEM.

Statistical Analysis
Complete data analysis was performed using Prism 5.0 (GraphPad Software, La Jolla, USA) statistical software. The data are presented as the mean +/2 SEM from at least three independent experiments. Statistical differences were analyzed by ANOVA. A p-value ,0.05 was considered statistically significant.