Previous investigations demonstrated that repeated stresses before an ethanol exposure sensitize ethanol withdrawal-induced anxiety-like behavior ('anxiety'). In addition to activating the hypothalamic-pituitary-adrenal axis, acute stress also elevates cytokines in brain. Initially, to test possible cytokine involvement in this stress/withdrawal protocol, cytokines were increased in brain with 2 weekly repeated lipopolysaccharide (LPS) administrations (1000 microg/kg) [corrected] (LPS/withdrawal protocol) or with twice weekly intracerebroventricular (i.c.v.) administrations of the cytokines IL-1 beta, CCL2 (MCP-1) or TNFalpha (cytokine/withdrawal protocol) before exposure and withdrawal from a 5-day cycle of chronic ethanol diet. Both protocols sensitized withdrawal-induced anxiety and confirm cytokine involvement in the sensitized anxiety response. Testing of various doses of LPS (16-1000 microg/kg) and TNFalpha (3-100 ng, i.c.v.) demonstrated the dose-related nature of these protocols to sensitize withdrawal-induced anxiety. The sensitized anxiety was not produced by a single 5-day ethanol diet cycle or by repeated LPS or cytokine treatments alone. Likewise, sensitized anxiety in these protocols could not be attributed to differences in ethanol ingestion. When challenged with a subsequent re-exposure to a 5-day ethanol diet cycle 16 days after completion of the LPS/withdrawal or cytokine/withdrawal protocols, an increase in withdrawal-induced anxiety was observed-an indication of induction of an underlying persistent adaptive change. Finally, just as found previously with the stress/withdrawal protocol, administration of the benzodiazepine receptor antagonist flumazenil before the LPS or TNF treatments prevented anxiety sensitization. Together, these findings indicate that increased cytokine activity induces adaptive change that supports sensitization of ethanol withdrawal-induced anxiety that may be linked to GABA(A)-receptor function.