Reverse signaling through the costimulatory ligand CD137L in epithelial cells is essential for natural killer cell-mediated acute tissue inflammation

Abstract

Renal ischemia-reperfusion injury (IRI) after kidney transplantation is a major cause of delayed graft function. Even though IRI is recognized as a highly coordinated and specific process, the pathways and mechanisms through which the innate response is activated are poorly understood. In this study, we used a mouse model of acute kidney IRI to examine whether the interactions of costimulatory receptor CD137 and its ligand (CD137L) are involved in the early phase of acute kidney inflammation caused by IRI. We report here that the specific expressions of CD137 on natural killer cells and of CD137L on tubular epithelial cells (TECs) are required for acute kidney IRI. Reverse signaling through CD137L in TECs results in their production of the chemokine (C-X-C motif) receptor 2 ligands CXCL1 and CXCL2 and the subsequent induction of neutrophil recruitment, resulting in a cascade of proinflammatory events during kidney IRI. Our findings identify an innate pathogenic pathway for renal IRI involving the natural killer cell-TEC-neutrophil axis, whereby CD137-CD137L interactions provide the causal contribution of epithelial cell dysregulation to renal IRI. The CD137L reverse signaling pathway in epithelial cells therefore may represent a good target for blocking the initial stage of inflammatory diseases, including renal IRI.

Fig. 1.

CD137L signals are required for production of CXCL1 and CXCL2 in TECs. (A) BALB/c mice were subjected to 35-min ischemia, followed by 4-h reperfusion. Kidney cells were isolated from naive and IRI mice and stained with anti–Ep-CAM and anti-CD137L mAbs. Expression of CD137L was analyzed on gated Ep-CAM⁺ cells. (Left) Thin and thick lines in FACS histograms indicate isotype control and anti-CD137 mAbs, respectively. (Right) Frozen sections were prepared from naive and IRI mice and stained with anti-CD137L plus anti-cytokeratin mAbs. (Magnification: 40×.) (B) Frozen sections were prepared from kidneys of naive and 4-h IRI WT, CD137^−/−, and CD137L^−/− C57BL/6 mice and stained with anti-CXCL1 or anti-CXCL2 plus anti-cytokeratin mAbs. (Magnification: 40×.) (C) BALB/c TECs (1 × 10⁵ per well) were cultured in six-well plates precoated with human IgG1 or CD137-Fc (1 μg/mL) for 4 h. The quantities of CXCL1 and CXCL2 contained in the culture medium were measured using ELISA. A chemotaxis assay for purified neutrophils (5 × 10⁵ per well) was conducted using conditioned medium from TEC cultures. To block CXCR2, anti-CXCR2 mAb (50 μg/mL) was added. (D) BALB/c TECs (1 × 10⁵ per well) were cultured in six-well plates precoated with human IgG1 or CD137-Fc (1 μg/mL). (Upper) TECs were harvested at the indicated time points. Cell lysates (5 μg per sample) were used for immunoblotting. (Lower) TECs (5 × 10⁴ per well) were cultured for 5 h in wells precoated with human IgG1 or CD137-Fc (1 μg/mL) for 5 h in the presence or absence of SB203580 (5 mM and 10 mM) or SP600125 (5 mM and 10 mM), respectively. The concentrations of CXCL1 and CXCL2 contained in the culture medium were measured using ELISA. Data are the mean ± SEM (n = 3 per group) and are representative of two to four independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 2.

CD137L engagement with CD137-Fc exacerbates renal IRI. CD137-Fc or control human IgG1 (50 μg per mouse) was administered into CD137^−/− (A–D) and WT (E–H) BALB/c mice, respectively, 1 h before ischemia. Mice were subjected to 35-min ischemia followed by 24-h reperfusion. (A and E) Serum creatinine and BUN levels. (B and F) Percentages of neutrophils in kidneys. (C and G) Photomicrographs of representative sections from the respective group. (Magnification: 400×.). (D and H) Tubular injury scores. Data are the mean ± SEM from six to eight mice per group in two independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 3.

Implantation of WT TECs under the kidney capsule significantly recovers renal IRI in CD137L^−/− mice. TECs (2 × 10⁵) from WT or CD137L^−/− C57BL/6 mice were implanted under the kidney capsules of WT and CD137L^−/− C57BL/6 mice. Immediately thereafter, kidney IRI was induced. (A) Serum creatinine and BUN levels. (B) Percentages of neutrophils in kidneys. (C) Photomicrographs of representative sections from the three groups at 400× (Upper) or 200× (Lower) magnification. Outer medullary areas are shown in upper panels, and cortex, and subcapsular areas containing implanted TECs are shown in lower panels. (D) Tubular injury scores. Data are the mean ± SEM (n = 6–7 per group) and are representative of two independent experiments. *P < 0.05; **P < 0.01.

Fig. 4.

Stimulation of TEC CD137L with NK cell CD137 results in CXC chemokine production. Kidneys were harvested from WT and CD137^−/− C57BL/6 mice 4 h after IRI induction. (A) Expression of CD137 was analyzed on gated CD45⁺CD3⁻NK1.1⁺ NK cells. Thin and thick lines in FACS histograms indicate isotype control and anti-CD137 mAb, respectively. (B) Immunohistochemical analysis for the basement layer structure and NK cell infiltration. Frozen sections were stained with anti-laminin or anti-NK1.1 and anti-cytokeratin mAbs. (Magnification: 40×.) (C–F) NK cells were purified from spleens of WT, CD137L^−/−, and CD137^−/− C57BL/6 mice. TECs were harvested from kidneys of WT and CD137L^−/− C57BL/6 mice. (C) NK cells were activated in the presence of IL-2 for 12 h, and expression of CD137 was confirmed on WT and CD137L^−/− cells, but not CD137^−/− NK cells, after activation. (D) Activated NK cells (5 x10⁵) were cocultured with TECs at the ratio of 5:1 for 4 h, and culture medium was harvested. CXCL1 and CXCL2 levels in the culture medium were quantified using ELISA. (E) Cocultured cells were harvested and stained with anti- CXCL1 and anti-cytokeratin mAbs. (Magnification: 20×.) (F) Conditioned medium of NK cell/TEC cocultures was used for chemotaxis assay. Data are the mean ± SEM (n = 3 per group) and are representative of two independent experiments. *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 5.

Adoptive transfer of WT NK cells restores renal IRI in CD137^−/− mice. (A) Kidney cells were harvested 4 h after IRI induction and stained with mAbs to CD45, CD3, and DX5. Data shown are the percentage of CD45⁺CD3⁻DX5⁺ cells in WT and CD137^−/− BALB/c mice (n = 7–8 per group). (B–E) CD45⁺CD3⁻DX5⁺CD122⁺ NK cells were purified from spleens of WT or CD137^−/− BALB/c mice using the MoFlow cell sorter and were adoptively transferred into CD137^−/− mice 1 h before induction of IRI. Mice that received WT or CD137^−/− NK cells were subjected to 35-min ischemia, followed by 24-h reperfusion. (B) Serum creatinine and BUN levels. (C) Percentages of neutrophils in kidneys. (D) Photomicrographs of representative sections from the respective groups. (Magnification: 400×.) (E) Tubular injury scores. Data are the mean ± SEM (n = 4–5 per group). *P < 0.05.

Fig. 6.

Depletion of NK cells inhibits neutrophil recruitment and kidney IRI. C57BL/6 mice were administered anti-NK1.1 mAb (200 μg per mouse). Twenty-four hours later mice were subjected to 35-min ischemia, followed by 24-h reperfusion. (A) Serum creatinine and BUN levels. (B) Percentages of neutrophils in kidneys. (C) Photomicrographs of representative sections from the respective group. (Magnification: 400×.) (D) Tubular injury scores. Data are the mean ± SEM from five mice per group. NK cell depletion was confirmed by analyzing spleen and liver NK cells.*P < 0.05; **P < 0.01; ***P < 0.001.

Fig. 7.

Depletion of neutrophils does not result in kidney IRI. BALB/c mice were administered anti–Gr-1 mAb (250 μg per mouse). Twenty-four hours later, mice were subjected to 35-min ischemia, followed by 24-h reperfusion. (A) Serum creatinine and BUN levels. (B) Percentages of neutrophils in kidneys. (C) Photomicrographs of representative sections from the respective group. (Magnification: 400×.) (D) Tubular injury scores. Data are the mean ± SEM from five mice per group. *P < 0.05; **P < 0.01; ***P < 0.001.

Fig. P1.

The role of CD137–CD137L interactions in the NK cell–TEC–neutrophil axis. After injury characterized by ischemia-reperfusion (reduced blood flow, followed by recovered blood flow), infiltrated NK cells (step 1) use their cell-surface molecule CD137 to stimulate CD137L on the surface of TECs (step 2). Signaling through CD137L results in the production of additional signaling molecules, CXCL1 and CXCL2, in TECs (step 3). Once infiltrated (step 4), neutrophils participate in active tissue destruction (step 5).