Our laboratory has recently shown that locally produced endothelin (ET) is involved in the atrial natriuretic peptide (ANP) response to a physical stimulus, stretch. The aim of this study was to determine if factors locally produced in the atria were involved in the ANP response to a chemical stimulus, anoxia. Reduced oxygen tension is a potent stimulus of ANP release, and our results show that, when isolated perfused atria were exposed to anoxic conditions, the ANP secretion rate increased by a maximum of 129 +/- 8% of the baseline. Exposure to anoxia caused neither an elevation in perfusate creatinine phosphokinase, a change in atrial morphology detectable by electron microscopy, nor interfered with the return toward the baseline ANP secretion rate with reoxygenation, suggesting that this response was not due to myocyte damage. When the atria were pretreated with either 3 microM BQ-123, an endothelin receptor inhibitor, or 10 microM indomethacin, a cyclooxygenase inhibitor, the ANP response to anoxia was nearly abolished. To clarify the association between ET and prostaglandins, we showed that the ANP response to 50 nM ET-1 was totally blocked at both high and low pressure by 10 microM indomethacin, but the increased contractility response to ET was unaffected. Therefore, we have concluded that the anoxia-induced ANP response is mediated by locally produced ET, which, in turn, stimulates the production of prostaglandins. Prostaglandins appear to be responsible for the increased ANP secretion rate.