Role of cytokines in liver inflammation

摘要

Because of its anatomical location and role in drug and xenobiotic detoxification, the healthy liver is confronted almost continuously with factors that induce injury-related, pro-inflammatory cytokines. For example, normal portal blood contains bacterial products, including lipopolysaccharide (LPS) endo-toxin, which is a potent inducer of tumour necrosis factor alpha (TNF-alpha) production by macrophages and other mononuclear cells. TNF induction is known to be one of the earliest events in hepatic inflammation, triggering a cascade of other cytokines that cooperate to kill hepatocytes, recruit inflammatory cells, and initiate a wound-healing response that includes fibrogenesis. However, although the liver contains one of the largest populations of resident macrophages of any organ in the body, and these cells are routinely exposed to portal blood LPS, cytokine gene expression is barely detectable in the healthy liver. Table 1 lists all of the cytokine mRNAs that either cannot be detected at all, or are demonstrated only inconsistently or at very low levels, by RNAase protection analysis of total RNA from healthy mouse livers. It is evident that LPS is not a very effective inducer of either pro- or anti-inflammatory cyto-kines in the healthy liver. Furthermore, acute treatment with exogenous pro-inflammatory cytokines, e.g. TNF-alpha, is generally well tolerated by the liver. For example, several investigators have reported that TNF-alpha treatment induces hepa-tocytes in healthy rats to proliferate, rather than die. The same response (i.e. hepatocyte proliferation) occurs when the healthy liver is confronted acutely by other insults, e.g. partial hepatectomy (PH), that acutely elicit hepatic production of pro-inflammatory cytokines. Indeed, antibody neutralization studies and experiments with type 1 TNF receptor-deficient mice indicate that TNF-alpha must activate its receptor that contains a 'death domain', in order for the liver to regenerate after PH.