CYTOKINES STIMULATE THE CRH BUT NOT THE VASOPRESSIN NEURONAL SYSTEM - EVIDENCE FOR A MEDIAN-EMINENCE SITE OF INTERLEUKIN-6 ACTION

Antigen-activated immune cells acutely release cytokines which, besides their effects on the immune system, increase hypothalamopituitary-adrenocortical (HPA) function to counteract the inflammatory process. The present study was designed to test, using in vitro paradigms, whether there exists a hypothalamic and/or a median eminence site of action, whereby different substances derived from the immune system could stimulate the CRH and/or the arginine-vasopressin (AVP) neuronal pathway. For this purpose, whole medial basal hypothalamus (containing the median eminence) were dissected from female rats and incubated in vitro with several concentrations of interleukin-1 (IL-1)beta, interleukin-6 (IL-6), tumor necrosis factor (TNF)-alpha, thymosin fraction 5 (TF5) or bacterial lipopolysaccharide (LPS). After a 40-min incubation period, the amounts of CRH and AVP released into the incubation medium were measured by specific radioimmunoassays (RIAs). Additional experiments were carried out by superfusing isolated rat median eminence fragments with the different test substances; CRH and AVP released into the medium were also measured by RIAs. The results indicated that IL-1-beta (10(-11) to 10(-7) M), IL-6 (0.06 x 10(-10) to 0.4 x 10(-10) M), TNF-alpha (6 x 10(-9) to 6 x 10(-7) M) and TF5 (5-500-mu-g/ml) but not LPS (1-100 ng/ml) significantly enhanced hypothalamic CRH secretion above baseline in a concentration-related fashion. Additionally, superfusion experiments demonstrated that, among all test substances, only IL-6 possesses a direct and dose-dependent CRH-releasing activity at the median eminence level. Conversely, no preparation enhanced basal AVP release in either in vitro design. For the pure cytokines, hypothalamic incubation studies further suggested a rank order of CRH releasing activity, as follows: IL-1-beta > TNF > IL-6; the partial purifed thymic preparation, TF5, was approximately as potent as IL-1-beta whereas LPS had no effect. Our results provide strong evidence for a hypothalamic site of action of several cytokines on CRH secretion. Additionally they suggest that IL-6 also contributes to stimulate HPA function by inducing CRH output from neuron terminals present at the median eminence level, where no effective blood-brain barrier is present. These data further suggest that the CRH neuronal system is one of the most important interfaces between the immune and neuroendocrine axes and that vasopressin is not involved in such an effect.