Using phase relations to identify potential mechanisms for metabolic oscillations in isolated β-cell mitochondria

There is a great deal of evidence for the existence of metabolic oscillations in pancreatic β-cells. Mechanisms that have been proposed for these oscillations include glycolytic oscillations; oscillations due to the feedback of Ca2+ onto the mitochondrial inner membrane and on dehydrogenases; and oscillations intrinsic to the tricarboxylic (TCA) cycle or the downstream reactions of oxidative phosphorylation. MacDonald and co-workers (J. Biol. Chem., 278:51894-51900, 2003) showed examples of oscillations in TCA intermediates in isolated mitochondria from liver cells and pancreatic β-cells. These oscillations were clearly not due to oscillations in glycolysis or Ca2+ feedback. In this article we consider several potential mechanisms for these TCA oscillations, using mathematical modeling to determine the phase relations that would result between the citrate and NAD+ concentrations in each case. We demonstrate that negative feedback at only one feedback point, isocitrate dehydrogenase, produces the correct phase relation if oscillations are intrinsic to the TCA cycle. Alternatively, the correct phase relation results if oscillations are due to oscillations in oxidative phosphorylation feeding back onto the TCA cycle. This analysis shows that the observed phase relation between citrate and NAD(P) places strict limits on the potential mechanism for the metabolic oscillations in isolated mitochondria that were observed by MacDonald and co-workers.