Uromodulin triggers IL-1β-dependent innate immunity via the NLRP3 inflammasome

Abstract

Uromodulin/Tamm-Horsfall protein is not immunostimulatory in the tubular lumen, but through unknown mechanisms it can activate dendritic cells and promote inflammation in the renal interstitium. Here, we noted that uromodulin isolated from human urine aggregates to large, irregular clumps with a crystal-like ultrastructure. These uromodulin nanoparticles activated isolated human monocytes to express costimulatory molecules and to secrete the mature proinflammatory cytokines, including IL-1β. Full release of IL-1β in response to uromodulin depended on priming of pro-IL-1β expression by Toll-like receptors, TNF-α, or IL-1α. In addition, uromodulin-induced secretion of mature IL-1β depended on the NLRP3 inflammasome, its linker molecule ASC, and pro-IL-1β cleavage by caspase-1. Activation of NLRP3 required phagocytosis of uromodulin particles into lysosomes, cathepsin leakage, oxidative stress, and potassium efflux from the cell. Taken together, these data suggest that uromodulin is a NLRP3 agonist handled by antigen-presenting cells as an immunostimulatory nanoparticle. Thus, in the presence of tubular damage that exposes the renal interstitium, uromodulin becomes an endogenous danger signal. The inability of renal parenchymal cells to secrete IL-1β may explain why uromodulin remains immunologically inert inside the luminal compartment of the urinary tract.

Figure 1.

Uromodulin ultrastructure. (A and B) Uromodulin aggregation in aqueous solution as seen by phase-contrast microscopy. (C and D) TEM revealed a filamentous network ultrastructure of uromodulin polymers (C). The diameter of the filaments is approximately 5 nm. (D) Storage at −20°C breaks the filament network into single filaments of even small oligomers with a diameter of approximately 5 nm. (E and F) SEM illustrates uromodulin aggregates (E) composed of crystal-like particles with a diameter of about 1 μm (F). Original magnification, ×200 in A and B; ×74,000 in C and D.

Figure 2.

Uromodulin activates human PBMCs to secrete IL-1β. (A) LPS-primed (1 µg/ml) PBMCs were treated with increasing concentrations of uromodulin or ATP (5 mM) and IL-1β was measured in the supernatants by ELISA after 6 hours. (B) Human PBMCs were treated with LPS or uromodulin and the amount of IL-1β (p17 and ³¹P) and caspase-1 (p10 and p22) in cell culture supernatants were visualized by Western blot. (C) LPS-primed (1 µg/ml) or unprimed PBMCs were incubated either with uromodulin (50 µg/ml) or proteinase K–digested uromodulin for 6 hours and IL-1β release was assessed in the supernatants ELISA. Data in A and C are means ± SDs from three independent experiments all performed in triplicate. (D) Flow cytometry of PBMCs after incubation with or without uromodulin (50 µg/ml) for 48 hours. Shown are representative histograms and median fluorescence intensities of CD80 and CD40 of gated CD14⁺ monocytes. (E) SEM images of PBMCs treated with or without uromodulin. Capacity of uromodulin to activate PBMCs is also illustrated by enlarged and fluffier appearing cells due to increased cell membrane foldings. ***P<0.001 versus medium control. UMOD, uromodulin; casp1, caspase-1.

Figure 3.

Uromodulin-induced IL-1β secretion requires NLRP3, ASC, and caspase-1. (A and C) Freshly isolated PBMCs were transfected with 200 pmol of each control siRNA or siRNA specific for NLRP3, ASC, and caspase1 genes. The efficiency of knockdown was determined by quantitative PCR for indicated genes, which was further confirmed by agarose gel electrophoresis. (B) After transfection, cells were primed with LPS (1 µg/ml) and subsequently incubated with uromodulin (50 µg/ml) for 6 hours and IL-1β secretion was measured by ELISA in the cell culture supernatants. Data are means ± SDs from three independent experiments all performed in triplicate. ^#P<0.05, ^##P<0.01, ^###P<0.001 versus control siRNA. siRNA, small interfering RNA; casp1, caspase-1; UMOD, uromodulin; Cntrl, control.

Figure 4.

Uromodulin-induced NLRP3 activation involves particle phagocytosis, oxidative stress, and potassium efflux. (A) PBMCs were primed with LPS (1 µg/ml) and incubated with immunogold-labeled uromodulin to determine localization. Gold particles are mostly localized within intracellular lysosomes, a few on the cell surface, as well as, in the cytosol (black arrow, ×200 nm). (B) The same cells were incubated with uromodulin (50 µg/ml) in the presence or absence of the phagocytosis inhibitor cytochalasin-D (5 µM) and supernatants were collected 6 hours later for IL-1β ELISA. (C–E) LPS-primed PBMCs were also treated with CA-07-Me (10 µM and 25 µM for monosodium urate), a cathepsin inhibitor; N-acetyl cysteine (10 mM), an antioxidant; and quinidine (250 µM), a potassium channel blocker. In B–E, monosodium urate (100 µg/ml) was used as a positive control without LPS prestimulation. (F) As another tool to block potassium efflux from the cell, we applied either 75 mM KCl or NaCl (as a control for hyperosmolarity) in a serum-free medium for 30 minutes, followed by uromodulin stimulation for 6 hours. IL-1β secretion was measured in supernatants by ELISA. ATP was used as a positive control. Data in A and C–F are means ± SDs from three independent experiments all performed in triplicate. **P<0.01 versus medium control. UMOD, uromodulin; NS, not significant.