Using an established biochemical approach, five subcellular fractions of human B lymphocytes were prepared by differential centrifugation. Crude membranes were passed over a lentil lectin column to enrich carbohydrate-coated cell surface microsomes. The lectin-bound fraction contained a high amount of plasma membrane-derived microsomes as indicated by cell surface markers. All subcellular fractions in Western blots proved to contain distinct but variable amounts of porin. There was a strong increase in porin content from crude membranes to plasma membrane-derived vesicles. The porin content of this fraction appeared to be higher than that of mitochondria. In the final step the plasma membrane-derived microsome fraction proved to be devoid of contamination by outer mitochondrial membranes, as revealed by antibodies against the established markers MAO B and Tom20 applied in Western blots. These data prove the extramitochondrial expression of human type-1 porin/ type-1 VDAC.

There is indirect evidence that the plasmalemma-integrated eukaryotic porin (the voltage-dependent anion-selective channel, VDAC) functions as the outwardly rectifying chloride channel (ORCC). The channel, which is believed to play a role in cell volume regulation, appears to be relevant for cystic fibrosis (CF) therapy, in that it may function as an alternative Cl(-) channel. In the present study we showed first that Gd(3+) altered the voltage dependence of human type-1 porin incorporated into artificial planar lipid bilayers. Next, using a light-scattering approach on transformed normal or CF human B-lymphocytes in hypotonic Ringer solution, we found slightly differing regulatory volume decrease (RVD) curves for the cell lines under study. Addition of 15-60 microM GdCl3 in hypotonic Ringer increased light scattering, pointing to cell swelling beyond normal values. RVD was not observed in those experiments. A corresponding effect was seen in isotonic Ringer containing GdCl3. In either osmotic situation Gd(3+)-induced cell swelling was abolished by monoclonal mouse anti-human type-1 porin antibodies. Agonist and antibody effects were dose dependent. Finally, videocamera-monitored control experiments with adherent HeLa cells verified the direct effect of the agonist on cell swelling in hypo- or isotonic situations and its prevention by the antibodies. We conclude that GdCl3 opens plasmalemma-integrated porin channels, allowing ions to following their gradients, resulting in cell swelling. Since respiratory epithelium expresses porin channels in the apical membrane, the use of gadolinium to activate ORCC may represent a new therapeutic approach in CF.

In mammalian cells porin channels are localised in both mitochondrial outer membranes and extra-mitochondrial membranes. We isolated mitochondriaderived porin of a human lymphoblastoid B cell line, determined its amino acid sequence and characterised its channel properties. Interestingly, the amino acid sequence of this porin preparation and, correspondingly, its electrophysiological characteristics in a reconstituted system were identical to those of 'Porin 31HL', the human type-1 porin purified from a crude membrane preparation of the same cell line using a different purification protocol. The results raise questions about targeting, insertion and orientation of human type-1 porin in different membranes.