Alpha-ketoisocaproate is not a true substrate for ATP production by pancreatic beta-cell mitochondria.

The ability of alpha-ketoisocaproate (KIC) to induce ATP production in isolated mitochondria from pancreatic beta-cells was examined with a bioluminometric method. There was no ATP production from KIC when tested alone or in combination with malate (1 mmol/l), nor did DL-beta-hydroxybutyrate induce mitochondrial ATP production, whereas palmitoyl-carnitine and pyruvate were efficient stimulators of mitochondrial ATP production in the presence of an equimolar concentration of malate. However, KIC stimulated the mitochondrial ATP production when tested in combination with glutamate (10 mmol/l). The concentration necessary to obtain half-maximal stimulation was approximately 50 micromol/l KIC, and maximal activity, comparable to that obtained with fatty acids, was reached at 1 mmol/l KIC. Higher KIC concentrations inhibited the mitochondrial ATP production, whereas a plateau was attained at 1 mmol/l KIC in the presence of glutamine. Ca2+ stimulated the maximal mitochondrial ATP production induced by KIC. Maximal stimulation was obtained with 300 nmol/l Ca2+ in the presence of 0.3 mmol/l KIC. Ca2+ reduced the concentration of KIC necessary for half-maximal stimulation to <30 micromol/l. Leucine stimulated the mitochondrial ATP production in the presence of glutamate to the same extent as KIC. Half-maximal stimulation was observed with 2 mmol/l leucine. There were no additive effects on mitochondrial ATP production when KIC and leucine were tested in combination. The results demonstrate that KIC by itself is not a mitochondrial substrate for ATP production. KIC must transaminate with glutamate or glutamine to yield alpha-ketoglutarate and leucine. Since leucine allosterically activates glutamate dehydrogenase, which also produces alpha-ketoglutarate, the insulinogenic effect of KIC may in part be due to the intramitochondrial generation of alpha-ketoglutarate. Since KIC-induced ATP production reaches a plateau already at micromolar concentrations (i.e., far below the concentrations at which KIC induces insulin release), it is proposed here that the catabolism of KIC may induce additional signals related to insulin release.