TNF-alpha mediates chemokine and cytokine expression and renal injury in cisplatin nephrotoxicity

Abstract

The purpose of these studies was to examine the role of cytokines in the pathogenesis of cisplatin nephrotoxicity. Injection of mice with cisplatin (20 mg/kg) led to severe renal failure. The expression of cytokines, chemokines, and ICAM-1 in kidney was measured by ribonuclease protection assays and RT-PCR. We found significant upregulation of TNF-alpha, TGF-beta, RANTES, MIP-2, MCP-1, TCA3, IL-1beta, and ICAM-1 in kidneys from cisplatin-treated animals. In addition, serum, kidney, and urine levels of TNF-alpha measured by ELISA were increased by cisplatin. Inhibitors of TNF-alpha production (GM6001, pentoxifylline) and TNF-alpha Ab's reduced serum and kidney TNF-alpha protein levels and also blunted the cisplatin-induced increases in TNF-alpha, TGF-beta, RANTES, MIP-2, MCP-1, and IL-1beta, but not ICAM-1, mRNA. In addition, the TNF-alpha inhibitors also ameliorated cisplatin-induced renal dysfunction and reduced cisplatin-induced structural damage. Likewise, TNF-alpha-deficient mice were resistant to cisplatin nephrotoxicity. These results indicate cisplatin nephrotoxicity is characterized by activation of proinflammatory cytokines and chemokines. TNF-alpha appears to play a central role in the activation of this cytokine response and also in the pathogenesis of cisplatin renal injury.

Figure 1

Time course of gene expression in cisplatin nephrotoxicity. Mice were injected with either saline (first lane) or 20 mg/kg cisplatin, sacrificed at the indicated times, and total RNA was isolated from the kidney. Levels of cytokine transcripts were measured by RPA as described in Methods.

Figure 2

Levels of TNF-α in serum (a), kidney (b), and urine (c). GM(–) is a structural analogue of GM6001 that does not inhibit TACE. All levels were measured 72 hours after injection. ND, not detectable. *P < 0.01 vs. saline, ⁺P < 0.01 vs. cisplatin. (a) n = 4–15; (b) n = 2–9; (c) n = 7–15. GM6001(–), GM(–); cis, cisplatin; pentoxi, pentoxifylline.

Figure 3

Effect of TNF-α inhibitors on renal cytokine expression. Mice were injected with either saline, cisplatin, or cisplatin with GM6001, or pentoxifylline and sacrificed at 72 hours. Cytokine transcripts were measured by RPA as in Figure 1. All lanes are from the same gel. Lanes 1 and 2 are shown in Figure 1 and are included here to facilitate comparison with the TNF-α inhibitor samples.

Figure 4

Effects of TNF-α inhibitors on cytokine gene expression. Cytokine gene expression was measured 72 hours after injection of cisplatin (black bars) or cisplatin and either GM6001 (white bars) or pentoxifylline (gray bars) by either RPA or real-time RT-PCR. In each experiment, the expression levels were normalized to the expression of either GAPDH or actin and are expressed relative to saline-treated control mice. n = 3–5 for each condition.

Figure 5

Effect of TNF-α inhibitors on cisplatin nephrotoxicity. Mice were injected with saline (squares), 20 mg/kg cisplatin (circles), cisplatin and GM6001 (triangles), or neutralizing TNF-α Ab (inverted triangles), as described in Methods. Blood urea nitrogen was measured at the indicated times. *P < 0.01 vs. cisplatin, n = 4–7.

Figure 6

Effects of cisplatin and TNF-α inhibitors on kidney morphology. Kidneys were removed 72 hours after injection with saline (a), cisplatin (b), or cisplatin and GM6001 (c). (a) Saline-treated kidneys show normal morphology with well-preserved brush border membranes and no loss of tubular epithelial cells. (b) Cisplatin-treated kidneys show extensive loss of tubular epithelial cells, tubular dilation, intratubular debris, and cast formation. (c) Kidneys from cisplatin plus GM6001 mice show relatively normal morphology. Sections were stained with PAS and photographed at ×100 magnification (inset, ×400).

Figure 7

Cisplatin nephrotoxicity in wild-type (circles) or TNF-α knockout (triangles, squares) mice. Mice were injected with saline (squares) or 20 mg/kg cisplatin (circles, triangles). Blood urea nitrogen was measured at the indicated times. *P < 0.01 vs. wild-type cisplatin-injected mice, n = 4–7.

Figure 8

Effect of cisplatin on kidney morphology in wild-type and TNF-α knockout mice. (a and b) Wild-type mice exhibit severe tubular necrosis, tubular dilation, and cast formation throughout the cortex. (c and d) TNF-α knockout mice have well-preserved renal morphology. Kidneys were removed 72 hours after injection with cisplatin. Sections were stained with PAS and photographed at ×100 magnification (a and c) or ×400 magnification (b and d).

Figure 9

TNF-α dependence of leukocyte infiltration in cisplatin nephrotoxicity. Sections of kidney harvested 72 hours after injection were stained for leukocytes as described in Methods. Thirty ×40 fields of kidney cortex were examined from each animal. The total number of leukocytes in those 30 fields is presented in the figure. *P < 0.02 vs. saline or cisplatin plus GM6001. n = 2–7. WT, wild type; KO, knockout.