Restricted Diffusion of cAMP is Mediated by Mitochondrial Localized PKA Buffering in Living Cells

Abstract

Different G-protein coupled receptors can elicit unique functional responses by selectively utilizing specific subsets of spatially restricted cAMP signaling domains. Previous studies have concluded that this can only be achieved if the rate of cAMP movement within cells is much slower than the rate of free diffusion. However, this prediction has not been demonstrated experimentally. In this study, we directly test the hypothesis that the cytoplasmic diffusion of cAMP is indeed slow and then address the long standing question of what regulates its movement. We used Raster Image Correlation Spectroscopy (RICS) to directly measure the diffusion coefficients of various molecules in intact HEK 293 cells and adult ventricular cardiac myocytes to evaluate the effect of particle size and cell morphology on cAMP movement. Fluorescein and pharos-450 (φ450), fluorescent molecules similar in size to cAMP, exhibited diffusion coefficients 3-4 fold slower than their expected rates of free diffusion. The diffusion coefficient of φ450-labeled cAMP (φ450-cAMP) was >6 fold slower than fluorescein or φ450 free dye alone. This extremely slow rate of diffusion could not be explained by its size. In cardiac myocytes, φ450-cAMP co-localized with the type II regulatory (RII) subunit of protein kinase A (PKA). In addition, disruption of RII subunit anchoring to A kinase anchoring proteins (AKAPs) significantly increased the diffusion coefficients of both RII subunits and φ450-cAMP. Furthermore, RII subunits and φ450-cAMP co-localized with mitochondria. These results demonstrate that the cytosolic movement of cAMP is indeed much slower than the predicted rate of free diffusion and that this is due to interactions with PKA regulatory subunits associated specifically with mitochondria. These findings have important implications with respect to cAMP signaling in all cell types.