Leptin is an endogenous protective protein against the toxicity exerted by tumor necrosis factor

Abstract

Tumor necrosis factor (TNF) is a central mediator of a number of important pathologies such as the systemic inflammatory response syndrome. Administration of high TNF doses induces acute anorexia, metabolic derangement, inflammation, and eventually shock and death. The in vivo effects of TNF are largely mediated by a complex network of TNF-induced cytokines and hormones acting together or antagonistically. Since TNF also induces leptin, a hormone secreted by adipocytes that modulates food intake and metabolism, we questioned the role of leptin in TNF-induced pathology. To address this question, we tested mouse strains that were defective either in leptin gene (ob/ob) or in functional leptin receptor gene (db/db), and made use of a receptor antagonist of leptin. Ob/ob and db/db mice, as well as normal mice treated with antagonist, exhibited increased sensitivity to the lethal effect of TNF. Exogenous leptin afforded protection to TNF in ob/ob mice, but failed to enhance the protective effect of endogenous leptin in normal mice. We conclude that leptin is involved in the protective mechanisms that allow an organism to cope with the potentially autoaggressive effects of its immune system.

Figure 1

Sensitivity of ob/ob and db/db mice to mTNF-induced lethality. Mice (wt, ob/ob and db/db) were challenged with 500 μg/kg mTNF intravenously. The percentage of survival was plotted as a function of time (hours after challenge). n = 10 (wt) and 5 (ob/ob and db/db). ○, wt; □, ob/ob; ▵, db/db. ****P = 0.0001 (wt versus ob/ob); ***P = 0.0006 (wt versus db/db). The results shown are representative of three independent experiments.

Figure 2

Effect of leptin pretreatment in ob/ob mice receiving a lethal challenge with mTNF. Mice (wt and ob/ob) were pretreated as described in Materials and Methods. Doses of 2A5 antibody and leptin were both 100 μg/mouse. Mice were challenged with 500 μg/kg mTNF. The percentage of survival was plotted as a function of time (hours after challenge). For each group, n = 5. ○, 2A5 in wt mice; □, 2A5 in ob/ob mice; ▵, leptin + 2A5 in ob/ob mice. **P = 0.0025 (wt versus ob/ob); *P = 0.0079 (ob/ob-2A5 versus ob/ob-leptin). The results shown are representative of three independent experiments.

Figure 3

Effect of leptin and leptin antagonist on mTNF-induced lethality. C57BL/6 mice were pretreated with leptin (100 μg/mouse) or R128Q (100 μg/mouse) in combination with 2A5 antibody (100 μg/ mouse) as described in Materials and Methods. Mice were challenged with a low dose (A, 375 μg/kg) or a high dose (B, 750 μg/kg) of mTNF. The percentage of survival was plotted as a function of time (hours after challenge). Each group consisted of 13 mice. The result is a cumulative sum of two independent experiments. (A) ○, 2A5; □, leptin + 2A5; ▵, R128Q + 2A5; ***P = 0.0008 (2A5 versus R128Q + 2A5); (B) •, 2A5; ▪, leptin + 2A5; ▴, R128Q + 2A5. **P = 0.0063 (2A5 versus R128Q + 2A5).

Figure 3

Effect of leptin and leptin antagonist on mTNF-induced lethality. C57BL/6 mice were pretreated with leptin (100 μg/mouse) or R128Q (100 μg/mouse) in combination with 2A5 antibody (100 μg/ mouse) as described in Materials and Methods. Mice were challenged with a low dose (A, 375 μg/kg) or a high dose (B, 750 μg/kg) of mTNF. The percentage of survival was plotted as a function of time (hours after challenge). Each group consisted of 13 mice. The result is a cumulative sum of two independent experiments. (A) ○, 2A5; □, leptin + 2A5; ▵, R128Q + 2A5; ***P = 0.0008 (2A5 versus R128Q + 2A5); (B) •, 2A5; ▪, leptin + 2A5; ▴, R128Q + 2A5. **P = 0.0063 (2A5 versus R128Q + 2A5).

Figure 4

Immune neuroendocrine network involving TNF and leptin. TNF induces a proinflammatory cascade on the one hand and an antiinflammatory protection mechanism on the other hand. → , induction; ¢, inhibition; ⇒ , binding to a receptor. GCH, glucocorticoid hormone; OB-R, leptin receptor long form; MC2-R, melanocortin-2 receptor; MCx-R, melanocortin-1 receptor or possibly other subtypes.