1. Isolated livers from fed male rats were perfused for 2 h with T4 (L-thyroxine), T3 (L-3,3′,5-tri-iodothyronine) or rT3 (L-3,3′,5′-tri-iodothyronine) at different pH values (7.1-7.6) in a fully synthetic medium, whereby normal metabolic functions were maintained without addition of rat blood constituents or albumin. 2. T3 output into the medium and net T3 production reached a maximum at a pH of the medium of 7.2 and significantly decreased with alteration of the pH when livers were perfused with T4 as a substrate. 3. However, the net T4 and T3 uptake by the liver, as well as the hepatic T4 and T3 content after perfusion, were not dependent on the pH of the perfusion when livers were offered T4 or T3 as substrates respectively. 4. Determination of intracellular pH by the analysis of the distribution of the weak acid dimethyloxazolidinedione allows the conclusion that the pH optimum of iodothyronine 5′-deiodinase in the intact perfused liver corresponds to the maximum determined in vitro for the membrane-bound enzyme localized in the endoplasmic reticulum. 5. The rapid 5′-deiodination of rT3 to 3,3′-T2 (L-3,3′-di-iodothyronine), the fast disappearance of 3,3′-T2, and the fact that no net rT3 production from T4 could be detected, supports the hypothesis that in rat liver iodothyronine 5′-deiodinase activity seems to predominate over iodothyronine 5-deiodinase activity. 6. Thus the rat liver can be considered in normal physiological situations as an organ forming T3 from T4 and deiodinating rT3 originating from extrahepatic tissues, whereby the cellular iodothyronine 5′-deiodination rate is controlled by the intracellular pH.
