Lipocalin-2 expressed in innate immune cells is an endogenous inhibitor of inflammation in murine nephrotoxic serum nephritis

Abstract

Lipocalin-2 (Lcn-2) is involved in divergent processes such as acute kidney injury or bacterial host defence. Our study was designed to evaluate the functional role of Lcn-2 in nephrotoxic serum nephritis (NTS). Since Lcn-2 is expressed in tubular epithelial cells as well as in cells of innate immunity such as macrophages and polymorphonuclear neutrophils (PMN), we induced NTS in wild-type (WT), Lcn-2 knock-out (KO) mice and WT/Lcn-2 KO chimeras. Mice lacking Lcn-2 exhibited more glomerular damage with increased proteinuria and interstitial leukocyte accumulation compared to WT mice. Chimeras able to express Lcn-2 in macrophages and PMN but not in epithelial cells were found to develop NTS comparable to wild-type controls. In contrast, chimeras expressing Lcn-2 in tubular epithelial cells with no expression in innate immune cells developed increased NTS due to decreased concerted apoptosis but increased necrosis and formation of damage-associated molecular patterns (DAMPs) such as high-mobility group box 1 (HMGB-1) in the kidney. In vivo blockade of HMGB-1, a toll-like receptor (TLR)-2 agonist, significantly reduced inflammation and NTS in Lcn-2 knock-out mice. In parallel, TLR-2 signalling was found to drive Lcn-2 transcription in vitro. Taken together, Lcn-2 expressed in innate immune cells is protective in NTS by inducing concerted apoptosis and inhibiting the formation of HMGB-1 thereby limiting cytokine production via TLR-2 signalling. In parallel, TLR-2 dependent transcription of Lcn-2 is an endogenous inhibitor of inflammation in NTS.

Figure 1. Lcn-2 is expressed in tubular epithelial and innate immune cells NTS.

C57BL/6 mice were subjected to NTS and were followed for 7 days. Kidneys were stained for Lcn-2. (A) The isotype control is shown. (B+C) Representative pictures are presented. Black arrows indicate infiltrating immune cells stained positively for Lcn-2.

Figure 2. Lcn-2 KO mice are increasingly susceptible to NTS.

WT (black bar; n = 13) and Lcn-2 KO mice (white bar; n = 12) were subjected to NTS. (A) Albuminuria was evaluated on day 0 and day 7 after NTS induction. (B) On day 7 after NTS induction kidney sections were evaluated for PAS positive deposits. (C) The percentage of crescentic glomeruli was quantified in WT and Lcn-2 KO mice. Data are given as means ± SEM. *p<0.05. (D) Representative PAS-stained sections of WT and Lcn-2 KO mice are shown. PAS positive glomeruli are marked by open arrows. A small crescent formation is marked by an asterix. Tubular cells with PAS positive material inserted because of heavy proteinuria are marked by closed arrows. Magnification x400.

Figure 3. Lcn-2 KO mice display increased inflammatory cell infiltration of kidneys after NTS.

WT (black bar; n = 13) and Lcn-2 KO mice (white bar; n = 12) were subjected to NTS. (A) Seven days after induction of NTS kidney sections were evaluated for the infiltration of CAE⁺ cells. The number of positive cells in 6 high power fields is given. (B) cDNA isolated from kidney samples was analysed for the KC expression. The relative abundance of KC to the house-keeping gene HPRT is provided. (C) Representative pictures of kidneys stained for Gr-1⁺ (left column) and CAE⁺ (right column) cells are given. The CAE⁺ cells are stained in dark blue, the Gr-1⁺ cells in brown. Kidney sections were analyzed for Gr-1 and CD68 (D), CD4 (E), and CD8 (F). (G) cDNA isolated from kidney samples was analyzed for the expression of the respective genes. The fold increase compared to healthy controls is provided. All data are given as mean ± SEM. Af = analysis field. Hpf = high power field. *p<0.05, **p<0.01, and ***p<0.001.

Figure 4. Lcn-2 KO show decreased apoptosis but increased proliferation in NTS.

WT (black bar; n = 13) and Lcn-2 KO mice (white bar; n = 12) were subjected to NTS. Seven days after induction of NTS kidney sections were (A) TUNEL stained and stained for (B) active Caspase-3. Apoptotic cells are marked with arrows. (C) The number of TUNEL and active Caspase-3 positive cells was quantified. Furthermore, sections were (D) H&E stained and evaluated for tubular necrosis. (E+G) Additionally, kidneys of WT and Lcn-2 KO mice were evaluated for proliferating cells 7 days after NTS induction by performing PCNA stainings. (E) Quantification and (G) representative pictures are shown._Magnification x400. Open arrows show representative PCNA-positive tubular epithelial cells. Closed arrows reflect representative PCNA-positive infiltrating immune cells. (F) Protein isolated from kidneys of Lcn-2 KO and WT mice was analyzed for HMGB-1 by Western blotting. Beta-actin reblotting was performed as a loading control. One representative blot is shown. All data are given as means ± SEM. Hpf = high power field. Lpf = low power field. *p<0.05, **p<0.01 and ***p<0.001.

Figure 5. Effect of Lcn-2 and Lcn-2 siRNA on staurosporin-induced cell death of murine macrophages.

(A) Percentages of apoptotic macrophages were determined after visualization of fragmented chromatin using Hoechst dye. (B) Caspase-3/7 activity was quantitated spectrophotometrically in macrophages. Lcn-2 incubation significantly increased number of apoptotic cells, prestimulation with Lcn-2 si-RNA prior to induction of cell death reduced macrophage apoptosis. Statistically significant differences are depicted: *,p<0.05; **,p<0.01.

Figure 6. Lcn-2 expressed in circulating immune cells protects mice from NTS.

WT (black bar; n = 10), Lcn-2 KO (white bar; n = 8) and chimeras (WT→KO, dark grey bar, n = 6; KO→WT, light grey bar, n = 8) were subjected to NTS. (A) On day 14 after induction of NTS albuminuria was evaluated. (B) Representative PAS-stained kidney sections are shown. Magnfication x400. (C) Tubular necrosis was assessed in WT→KO and KO→WT chimeras by H&E staining. (D) WT mice and KO→WT chimeras were analyzed for their urinary Lcn-2 concentration 0, 7 and 14 days after NTS induction. (E) On day 14 the serum Lcn-2 concentration of the two groups was evaluated. Data are given as mean ± SEM. *p<0.05, **p<0.01.

Figure 7. TLR2 stimulation mediates Lcn-2 expression.

(A+B) Cultured distal convoluted tubular cells (DCT, white bar) and (C+D) a macrophage cell line (RAW, black bar) were pre-treated with TLR2 siRNA and evaluated for (A+C) Lcn-2 and (B+D) IL-6 mRNA production after stimulation with TLR2 ligands PGN and LTA. All data are provided as fold increase. *p<0.05, **p<0.01. At least three independent experiments were performed.

Figure 8. HMGB-1 blockade reduces NTS activity by limiting macrophage infiltration of the kidney of WT mice.

WT mice were treated with an isotype control antibody (black bar; n = 5) or with an anti-HMGB-1 antibody (shaded bar; n = 5). Seven days after disease induction (A) staining for Gr-1⁺ cells, CD68⁺ and CD8⁺ cells was performed. The number of positive pixels in 6 analysis fields is given. (B) Staining for CD4⁺ T cells was performed. The number of positive cells in 6 high power fields is given. (C) cDNA isolated from kidney samples was analysed for the expression of the respective genes. The fold change compared to WT with isotype control is shown. (D) Scoring for PAS-positive deposits was performed. All data are given as mean ± SEM. Hpf = high power field, Af = analysis field. *p<0.05 and **p<0.01.

Figure 9. HMGB-1 blockade reduces NTS activity by limiting macrophage infiltration of the kidney of Lcn-2 KO mice.

WT mice (black bar; n = 5), Lcn-2 KO mice treated with an isotype control antibody (white bar; n = 6) or with an anti-HMGB-1 antibody (grey bar; n = 5). Seven days after disease induction (A) staining for Gr-1⁺ cells, CD68⁺ and CD8⁺ cells was performed. The number of positive pixels in 6 analysis fields is given. (B) Kidney sections were stained for infiltration by CD4⁺ T cells. The number of positive cells in 6 high power fields is given. (C) cDNA isolated from kidney samples was analysed for the expression of the respective genes. The ratio of the respective gene to the housekeeping gene HPRT is provided. (D) Scoring for PAS-positive deposits was performed. All data are given as mean ± SEM. Af = analysis field; Hpf = high power field; *p<0.05 and **p<0.01 WT compared to Lcn-2KO mice+isotype. ^#p<0.05 and ^##p<0.01 Lcn-2KO mice+isotype compared to Lcn-2KO mice+a-HMGB-1 Ab.