Interleukin-1 receptor-mediated inflammation impairs the heat shock response of human mesothelial cells

Abstract

Bioincompatibility of peritoneal dialysis fluids (PDF) limits their use in renal replacement therapy. PDF exposure harms mesothelial cells but induces heat shock proteins (HSP), which are essential for repair and cytoprotection. We searched for cellular pathways that impair the heat shock response in mesothelial cells after PDF-exposure. In a dose-response experiment, increasing PDF-exposure times resulted in rapidly increasing mesothelial cell damage but decreasing HSP expression, confirming impaired heat shock response. Using proteomics and bioinformatics, simultaneously activated apoptosis-related and inflammation-related pathways were identified as candidate mechanisms. Testing the role of sterile inflammation, addition of necrotic cell material to mesothelial cells increased, whereas addition of the interleukin-1 receptor (IL-1R) antagonist anakinra to PDF decreased release of inflammatory cytokines. Addition of anakinra during PDF exposure resulted in cytoprotection and increased chaperone expression. Thus, activation of the IL-1R plays a pivotal role in impairment of the heat shock response of mesothelial cells to PDF. Danger signals from injured cells lead to an elevated level of cytokine release associated with sterile inflammation, which reduces expression of HSP and other cytoprotective chaperones and exacerbates PDF damage. Blocking the IL-1R pathway might be useful in limiting damage during peritoneal dialysis.

Figure 1

PDF induces insufficient levels of HSP. Expression of Hsp72 as assessed by Western blot analysis and LDH release as a parameter for cellular damage over various exposure times. Hsp72 is shown in values normalized to the expression of β-tubulin to compensate for a potential decrease in cell number with increasing PDF exposure time and relative to control. Control samples underwent sham treatment with normal growth medium. The recovery time was 16 hours for all samples. Mean values are presented as percent-fraction of the observed maximum and were computed from three independent biological experiments with three biological replicates each (n = 9 per data point). LDH release of control samples was set to 0%. The error bars represent the standard error. Expression of Hsp72 dropped below the detection range (n.d.) in case of 4-hour exposure to PDF.

Figure 2

Modality of cell death in experimental PD. Results from flow cytometry experiments after exposure of MCs to control medium (A) or PDF (B, C). The modality of cell death was assessed using standard propidium iodide (y axis) and AnnexinV-fluorescein isothiocyanate (x axis) labeling. Thresholds were set according to an unstained control (not shown) and applied to all samples as shown in the representative panels. The mean values and standard deviations of quadrant populations presented in D were computed from three independent biological experiments. PI, propidium iodide; FITC, fluorescein isothiocyanate.

Figure 3

Proteomics and bioinformatics approach. A: A representative Coomassie brilliant blue stained two-dimensional electrophoresis gel of the proteome of MCs. Proteins were separated by isoelectric point (pI) in the first dimension and molecular weight (MW) in the second dimension. Spots with significantly altered abundance (P < 0.05), where proteins could be identified by MS, are marked with arrows and labeled by Swiss-Prot identifiers according to Table 1 (“human” omitted). B: Representation of the PIN resulting from the expanded primary candidate list on the basis of OPHID interaction data. Proteomic (primary) candidates are depicted using black circles and next neighbors (secondary candidates) are depicted using light gray triangles, the lines denote the interaction edge.

Figure 4

Necrotic cell material (NCM) induces sterile inflammation. Secretion of the pro-inflammatory cytokines IL-6 (A) and IL-8 (B) into the medium supernatant after pre-incubation of MCs with NCM from homogenized cells as assessed by enzyme-linked immunosorbent assay. Mean values are shown relative to control without NCM pretreatment (white bars) and are representative for three independent biological experiments with two biological replicates each (n = 6 per data point). The error bars represent the standard error. NCM pre-treatment is shown as gray bars. Hatched bars show the effects of the addition of the IL-1Ra anakinra (single hatched) and the anti-TNF-mab infliximab (cross-hatched). Significant differences in group-wise comparisons from analysis of variance tests are indicated with brackets (* P < 0.05, ** P < 0.01).

Figure 5

PDF induces a pattern of pro-inflammatory interleukins similar to NCM. Secretion of the pro-inflammatory cytokines IL-6 (A) and IL-8 (B) into the medium supernatant after exposure of MCs to PDF as assessed by enzyme-linked immunosorbent assay. Mean values are shown relative to control without PDF exposure (white bars) and are representative for three independent biological experiments with two biological replicates each (n = 6 per data point). The error bars represent the standard error. PDF exposure is shown as gray bars. Hatched bars show the effects of the addition of the IL-1Ra anakinra (single hatched) and the anti-TNF-mab infliximab (cross-hatched). Significant differences in group-wise comparisons from analysis of variance tests are indicated with brackets (* P < 0.05, ** P < 0.01).

Figure 6

Addition of anakinra enables higher expression of Hsp72. The graph shows the results of a time-course experiment with PDF (black bars) or PDF supplemented with the IL-1Ra anakinra (shaded bars). Control samples underwent sham treatment with normal growth medium. Data are representative for three independent biological experiments with two biological replicates each (n = 6). Bars represent mean values relative to control; error bars are shown as standard error. A: LDH release was assessed after various exposure times. Enzymatic measurements were carried out in the supernatant after a recovery of 16 hours. B: Hsp72 was assessed by Western blot from total protein lysates after a recovery of 16 hours. Values are normalized to the expression of β-tubulin to compensate for a potential decrease in cell number with increasing PDF exposure time. Asterisks indicate time-points with significant differences in pair-wise comparisons between PDF and PDF supplemented with anakinra (t-test; P < 0.05).

Figure 7

Proteomics evaluation of anakinra effects on MCs after PDF exposure. Expression profiles of proteins with known association to PDF stress after addition of the IL-1Ra anakinra to PDF. The abundance as measured from two-dimensional gel staining is given relative to control conditions of the initial experiment. All proteins presented in this graph show significant differences in spot abundance as well between control and PDF treatment as between PDF and PDF supplemented with anakinra (P < 0.05). Expression profiles are labeled with Swiss-Prot entry names as given in Table 1. The two profiles for Hsp27 (HSPB1_HUMAN) refer to two separate spots on the two-dimensional gels. Data are representative for three independent biological experiments with two biological replicates each (n = 6). Error bars represent the SE.

Figure 8

New concept of peritoneal dialysis fluid (PDF) stress. Scheme of mesothelial cell monolayer exposed to peritoneal dialysis PDF. The current model of cytotoxic damage is extended with the model of sterile inflammation, which is revealed by pathway analysis and confirmed in experiments simulating sterile inflammation. Danger signals, released from injured cells, lead to sterile inflammation in adjacent cells via IL-1 receptor (IL-R)-mediated pathways. This sterile inflammation affects cytoprotective responses and exacerbates cell death on PD fluid exposure.