(3) A PORIN WITH PROPERTIES UNDISTINGUISHABLE FROM THOSE OF MITOCHONDRIAL VDAC IS PRESENT IN CAVEOLAE AND CAVEOLAE-RELATED DOMAINS

G. Bathori (1) I. Parolini (2), F. Tombola (4), I. Szabú (4), A. Messina (3), M. Oliva (3), V. De Pinto (3), M. Sargiacomo (2) and M. Zoratti (4)
(1) Department of Physiology, Semmelweis Medical University, Budapest, Hungary
(2) Istituto Superiore di Sanità, Laboratorio di Ematologia, Roma, Italia
(3) Istituto di Scienze biochimiche e Farmacologiche, Catania, Italia
(4) Centro studi sulle Biomembrane del C.N.R., Padova, Italia

Porin, or VDAC (voltage-dependent anion channel), is a small family of channel-forming homologous proteins which confer permeability to the outer mitochondrial membrane. Since 1989 several reports have suggested that it may be present in non-mitochondrial locations, including the plasma membrane and endosomes, but the finding has remained controversial, and the scientific community is clearly divided on the issue. One of the arguments against the presence of porin in caveolae has been the possibility that VDAC may become solubilized during the preparation procedure, associating itself with caveolae because of its known affinity for cholesterol. The main aim of this work thus was to establish if the reported porin localization in the plasmamembrane was just an artifact or if it is a physiological feature of the cell. This question is of the highest biological importance, since the main recognized task of porin is to create a large channel in membranes, something apparently incompatible with plasma membrane function.

To solve this problem, we first ascertained the presence of VDAC in preparations of caveolae and lipid rafts from various sources, obtained with and without the use of detergents. VDAC was found to co-localize in density gradients with caveolin and Src-family tyrosine kinases (caveolar markers), and to be enriched in the caveolae or caveolae-like domains in comparison with total cell homogenates. The lack in these preparations of an abundant mitochondrial membrane protein used as a marker was also established by Western blots, indicating that they were free of mitochondrial contamination. To test the functionality of these preparations, brain caveolae were reconstituted into planar bilayers, where they exhibited channel activity undistinguishable from that arising from purified mitochondrial porin (conductance, voltage dependence, selectivity and kinetic behavior were studied).

To definitely establish the biochemical and biophysical features of the porin present in the plasma membrane, CEM cells (a lymphocytic cell line) were biotinylated with an impermeable reagent. Immunoprecipitation, followed by streptavidin staining, of plasmamembrane marker proteins resident in caveolae-related domains and known to be exposed either to the outside (CD4) or only to the cytoplasm (PKCq), as well as dye exclusion assays, confirmed that only surface-exposed proteins were labeled. Lipid rafts were prepared from the biotinylated cells, and biotinylated porin was isolated by a procedure including affinity chromatography on streptavidin-agarose. This porin preparation showed biochemical properties undistinguishable from those of the porin purified from mitochondria (the same electrophoretic mobilities and immunochemical reactivities with poly- and monoclonal Abs). Furthermore, the protein also exhibited electrophysiological behaviours undistinguishable from those of the other preparations we used and from those of mitochondrial VDAC. Finally, the presence in the plasma membrane of CEM cells of a regulated channel having properties similar to those of isolated porin was established by patch-clamp experiments. In our opinion, the present work provides indisputable evidence that porin is present both in the mitochondrial outer membrane and in the plasmamembrane.

Acknowledgements: the authors acknowledge the support from Telethon, Italy. VDP acknowledge also funding from CNR and MURST-COFIN.