The potential role of HMGB1 release in peritoneal dialysis-related peritonitis

Abstract

High mobility group box 1 (HMGB1), a DNA-binding nuclear protein, has been implicated as an endogenous danger signal in the pathogenesis of infection diseases. However, the potential role and source of HMGB1 in the peritoneal dialysis (PD) effluence of patients with peritonitis are unknown. First, to evaluate HMDB1 levels in peritoneal dialysis effluence (PDE), a total of 61 PD patients were enrolled in this study, including 42 patients with peritonitis and 19 without peritonitis. Demographic characteristics, symptoms, physical examination findings and laboratory parameters were recorded. HMGB1 levels in PDE were determined by Western blot and ELISA. The concentrations of TNF-α and IL-6 in PDE were quantified by ELISA. By animal model, inhibition of HMGB1 with glycyrrhizin was performed to determine the effects of HMGB1 in LPS-induced mice peritonitis. In vitro, a human peritoneal mesothelial cell line (HMrSV5) was stimulated with lipopolysaccharide (LPS), HMGB1 extracellular content in the culture media and intracellular distribution in various cellular fractions were analyzed by Western blot or immunofluorescence. The results showed that the levels of HMGB1 in PDE were higher in patients with peritonitis than those in controls, and gradually declined during the period of effective antibiotic treatments. Furthermore, the levels of HMGB1 in PDE were positively correlated with white blood cells (WBCs) count, TNF-α and IL-6 levels. However, pretreatment with glycyrrhizin attenuated LPS-induced acute peritoneal inflammation and dysfunction in mice. In cultured HMrSV5 cells, LPS actively induced HMGB1 nuclear-cytoplasmic translocation and release in a time and dose-dependent fashion. Moreover, cytosolic HMGB1 was located in lysosomes and secreted via a lysosome-mediated secretory pathway following LPS stimulation. Our study demonstrates that elevated HMGB1 levels in PDE during PD-related peritonitis, at least partially, from peritoneal mesothelial cells, which may be involved in the process of PD-related peritonitis and play a critical role in acute peritoneal dysfunction.

Figure 1. HMGB1 levels in peritoneal dialysis effluents (PDE).

(A) Levels of HMGB1 in PDE of patients with or without peritonitis were detected by western blotting. PD patients without peritonitis served as controls (Con). (B) Densitometry of HMGB1 in immunoblots. Data are means ± SE (n = 3), *P<0.05 versus control subjects. (C) Representative immunoblot for HMGB1 in PDE among patient subgroups, including patients without peritonitis, with Gram-positive (G⁺) and Gram-negative (G⁻) peritonitis. (D) Quantitative determination of the relative abundance of HMGB1 protein among different groups. Data are means ± SE (n = 3), *P<0.05 versus control subjects. (E) Levels of HMGB1 in PDE of patients with or without peritonitis were quantified by ELISA. (F) Levels of HMGB1 in PDE among patient subgroups were assayed by ELISA. The box plot in E and F represents (from the top) values of the maximum, the third quartile, the median, the first quartile and the minimum, respectively (n = 4). *P<0.05 versus no peritonitis, ^# P<0.05 versus Gram-positive peritonitis.

Figure 2. Serial changes in HMGB1 levels in PDE during peritonitis.

(A) Representative HMGB1 immunoblot on PDE samples after antibiotic treatment. (B) Quantitative determination of relative HMGB1 levels in PDE after treatment. Data are expressed as mean ± SE from 3 independent experiments, *P<0.05 versus HMGB1 levels before treatment.

Figure 3. Levels of TNF-α and IL-6 in PDE during peritonitis.

TNF-α levels (A) and IL-6 levels (B) in PDE were measured by ELISA from PD patients with or without peritonitis. The concentrations of TNF-α (C) and IL-6 (D) in PDE among subgroups of patients were determined by ELISA. The values represented the maximum, the third quartile, the median, the first quartile and the minimum, respectively (n = 4). *P<0.05 versus no peritonitis, ^# P<0.05 versus Gram-positive peritonitis.

Figure 4. Correlation between PDE levels of HMGB1 and WBCs as well as cytokines during peritonitis.

(A) Correlation between levels of HMGB1 and WBC counts in PDE (r = 0.86, P<0.01). (B) Correlation between levels of HMGB1 and TNF-α in PDE (r = 0.75, P<0.01). (C) Correlation between levels of HMGB1 and IL-6 in PDE (r = 0.81, P<0.01).

Figure 5. Effects of HMGB1 inhibitor on peritoneal inflammation and function.

(A) Representative HE staining showed morphological changes and inflammatory infiltrate in both parietal and visceral peritoneum in each condition. Original magnification×200. (B) The total number of WBCs and the percentage of leukocytes in PDE among different group. (C–E) PD transport parameters to ultrafiltration, urea and glucose among different groups. Values are expressed as net ultrafiltration, D/P urea or D/D₀ glucose. Data in C, D and E are mean ± SE (n = 6), *P<0.05 versus control, # P<0.05 versus LPS-treated without glycyrrhizin (GL) administration.

Figure 6. Effects of LPS on HMGB1 release in HMrSV5 cells.

(A) Cells were treated with LPS at various concentrations for 48 hr. Cell culture media were collected and analyzed by immunoblotting with HMGB1 antibody. (B) Densitometry of HMGB1 proteins in immunoblots. (C) Cell viability was evaluated by MTT assay after treatment with LPS at the indicated concentrations for 48 hr. (D) Immunoblot analysis of HMGB1 in cell culture supernatants following 2 µg/ml LPS stimulation for the indicated time. (E) Quantitative determination of the relative abundance of HMGB1 proteins among different groups. (F) Cell viability was assayed by MTT at different time following LPS incubation. Data in B, C, E and F are expressed as mean ± SE (n = 6). *P<0.05 versus control group.

Figure 7. HMGB1 nuclear-cytoplasmic translocation in LPS-induced HMrSV5 cells.

Cells were treated with 2 µg/ml LPS for the indicated time period. (A) Representative confocal microscopic images showed the cellular localization of HMGB1 (red) and nuclear staining (blue) by indirect immunofluorescence staining in cells. Original magnification×400. (B) HMGB1 content in cytoplasm and nuclear fractions after LPS stimulation were assessed by Western blot analysis. (C) Quantitative determination of the relative abundance of HMGB1 in the cytoplasm and the nucleus among different groups. Data are expressed as mean±SE of three experiments. *P<0.01 versus negative control in the cytoplasm; #P<0.01 versus negative control in the nucleus.

Figure 8. HMGB1 was secreted via lysosome-mediated secretory pathway in response to LPS stimulation.

(A). HMGB1 was present in vesicles cofractionating with lysosomes after LPS administration. Lysosome fractions were untreated (lanes 1 and 2, Try) or solubilized (lanes 3 and 4, TyrTx) with Triton X-100 (TX) before trypsin (Try) digestion and subjected to Western blot analysis with anti-Cathepsin D and anti-HMGB1 antibodies. (B) Densitometry of HMGB1 content in different groups. Data are expressed as mean ± SE, n = 3 per treatment, *P<0.05 versus Try treated only group. (C) Cells were treated with LPS (2 ug/ml) for 24 hr. Representative immunofluorescence analysis of cellular localization of HMGB1 (green) and LAMP2a (red), a maker of lysosome in cells. Original magnification×400. (D) HMGB1 protein contents in lysosome fractions following LPS treatment were determined by Western blotting. (E) Quantitative determination of the relative abundance of HMGB1 among different groups. Data are expressed as mean ± SE, n = 3 per treatment, *P<0.05 versus control group, # P<0.05 versus LPS treated for 24 hr.