Experimental modulation of occluding junctions in a cultured transporting epithelium.

The experimental opening and resealing of occluding junctions in monolayers of cultured MDCK cells (epithelioid of renal origin) was explored by measuring changes in the electrical resistance across the monolayer and by freeze-fracture electron microscopy. As in natural epithelia, the function of occluding junctions as permeability barriers specifically depends on extracellular Ca++ concentration and fails if this ion is replaced by Mg++ or Ba++. The removal of Ca++ and the addition of EGTA to the bathing medium opened the junctions and reduced the transepithelial resistance. Resealing was achieved within 10-15 min by restoring Ca++. Quantitative freeze-fracture electron microscopy showed that junctional opening, caused by lack of Ca++, was accompanied by simplification of the pattern of the membrane strands of the occluding junction without disassembly or displacement of the junctional components. Resealing of the cellular contacts involved the gradual return to a normal junctional pattern estimated as the average number of strands constituting the junction. The occluding junctions were also opened by the addition of the ionophore A23187, suggesting that the sealing of the contacts requires high Ca++ on the extracellular side and low Ca++ concentration of the cytoplasmic compartment. The opening process could be blocked by low temperature (7.5 degrees C). Resealing did not depend on serum factors and did not require protein synthesis; therefore, it seems to be caused by reassembly of preexisting membrane junctional components. The restoration of the junctions occurred simultaneously with the establishment of ion-selective channels; the Na+/Cl- and the cation/cation selectivity were recovered with the same time-course as the electrical resistance. The role of the cytoskeleton in the process of junctional reassembly is reported in the companion article.

Magnesium and octylguanidinium inhibition of monovalent cation translocation in mitochondria.

We have studied the inhibitory effects of several cations (OG+, Mg2+, Ca2+, Sr2+, Ba2+, Mn2+, La3+) on the uniport and exchange pathways for Na+ and K+ of rat liver mitochondria. Swelling of mitochondria suspended in sodium or potassium acetates indicates that: 1. Sodium passive influx to inhibited mitochondria is not affected by OG+ (in the microM range), and mM concentrations of polyvalent cations only induce a poor inhibition, the sequence being: La3+ greater than Mn2+ greater than Ca2+ greater than Mg2+ greater than Sr2+ = 0. 2. Sodium active influx is 50% inhibited by 60 microM Mg2+ or 90 microM OG+. La3+, Mn2+, Ca2+ and Sr2+ also inhibit Na+ influx (mM range). 10 mM Mg2+ or 35 microM OG+ are required to inhibit 50% K+ active influx. 3. Alkali cation efflux from partially swollen inhibited mitochondria is 50% blocked by 2 mM Mg2+ or 105 microM OG+ when sodium is the major permeable cation in the bathing solution. 3 mM Mg2+ or 3.8 microM OG+ are required for 50% inhibition when mitochondria are suspended in potassium acetate. 4. Alkali cation efflux from partially swollen respiring mitochondria suspended in sodium acetate is promoted by concentrations above 0.1-0.2 mM Mg2+ or 50-100 microM OG+. These data fit a mechanism including an energy-dependent Mg2+ and OG+ sensitive inward sodium translocator and a Mg2+ and OG+ insensitive cation/H+ exchanger working in dynamic balance.

Mechanism of ammonium translocation in rat liver mitochondria. Finger-printing of the translocator.

We have studied whether ammonia crosses the cristae membrane of rat liver mitochondria as a charged (NH4+) or an uncharged (NH3) species. Passive swelling of mitochondria suspended in ammonium and/or sodium acetates showed that: swelling depends upon ammonium concentration in the same way it depends on sodium (at constant acetate concentration). The curves reach a plateau at 115 mM ammonium and 100 mM sodium. A two units change of pH induces a two orders of magnitude change in the concentration of NH3, while NH4+ remains almost constant. However, the extent and initial rates of swelling are not significantly modified. The results are discussed as ammonia being translocated as a charged species via a transport system. The transport system was characterized studying the swelling reaction of inhibited rat liver mitochondria suspended in acetate salts of alkali, ammonium and nitrogenous cations: The selectivity pattern of the translocator is: NH4+ greater than Na+ greater than Li+ greater than K+ greater than Rb+ greater than Cs+ (Eisenman's sequence X). For homologous nitrogenous cations permeability decreases as molecular weight increases or free solution mobility decreases. Ions with Ladd Radius shorter than 3.7 to 3.8 A and cross sections below 15 A2 are permeable, those with longer radius or larger areas are not. As the chain of monosubstituted nitrogenous cations is lengthened, permeability decreases, passes through a minimum around NC = 4 and then increases. Nitrogenous cations which do not make hydrogen bonds do not enter mitochondria, those with one, two or three donor protons have similar permeabilities, and those with one, two or three oxygen acceptors increase their permeability almost linearly. Formamidine, acetamidine, TEA, guanidine derivatives, N,N-dicyclohexylcarbodiimide and N-ethyl maleimide failed to inhibit sodium-induced mitochondrial passive swelling.

Culture of an epithelial cell line (MDCK) with a serum substitute.

This paper reports the suitability of culturing a line of dog kidney epithelial cells, MDCK, in the presence of a serum substitute, Ultroser G. Serial subcultivation with this product was possible for at least 10 passages without any change in cell shape and size, saturation density, dome-forming ability, transepithelial resistance, and growth curve. Adhesion of newly plated cells to plastic was somewhat lower than in fetal calf serum but the trypsin-harvesting kinetics were essentially the same. However, the membrane ion transport system was altered: cell sodium influx was greatly diminished, suggesting a deep change in the amiloride-sensitive Na+ channels; sodium efflux was highly enhanced (both active and passive).

Changes in paracellular and cellular ionic permeabilities of monolayers of MDCK cells infected with influenza or vesicular stomatitis viruses.

MDCK cells (epithelioid line derived from the kidney of a normal dog) form monolayers which retain the properties of transporting epithelia. In these cells viruses bud asymmetrically: influenza from the apical, and vesicular stomatitis (VSV from the basolateral membrane (E. Rodríguez-Boulán and D.D. Sabatini, Proc. Natl. Acad. Sci. USA 75: 5071-5075, 1978; E. Rodríguez-Boulán and M. Pendergast, Cell 20: 45-54, 1980). In the present study, we analyzed whether these viruses affect specific ion-translocating mechanisms located in the plasma membrane. We studied the effect of infection on membrane and transepithelial conductance, passive and active unidirectional fluxes of Na+ and K+, intracellular potentials, cellular content of Na+ and K+, and formation of blisters which, in these preparations, are due to the vectorial transport of fluid. Two main observations are derived from these studies. First, infection with VSV caused an increase in transepithelial electrical conductance, due to the opening of tight junctions, 5 to 6 hr after the start of infection, coincident with the accumulation of envelope protein in the cell surface and with the rise in the curve of virus budding. Infection with influenza, on the other hand, increased the transepithelial conductance only late in the infection (12 to 14 hr) when virus production has already stopped. Second, viruses did affect membrane permeability. Yet, the changes observed may not be ascribed to a perturbation of the specific translocating mechanisms for Na+ and K+ which operate in the same region of the plasma membrane that the viruses use to penetrate and leave MDCK cells. The methods used in the present study are not suitable to decide whether the nonspecific changes in permeability elicited by the viruses occur over the whole cell membrane or are restricted to a given region.

Repolarization of Na+-K+ pumps during establishment of epithelial monolayers.

Madin-Darby canine kidney (MDCK) cells plated at confluence and incubated for 20 h in low (5 microM) Ca2+ have no tight junctions (TJs), and their Na+-K+-ATPase is randomly distributed over the surface. On transfer to normal Ca2+ levels (1.8 mM) ("Ca2+ switch"), TJs and transepithelial resistance develop quickly, trapping a considerable fraction (35%) of the surface Na+-K+-ATPase on the apical (incorrect) side. This misplaced enzyme is subsequently removed from this region or inactivated, demonstrating that polarization proceeds despite TJs. Simultaneously, the amount of Na+-K+-ATPase on the basolateral side increases in a higher proportion (125%), than could be accounted for by relocation of the misplaced apical enzyme. This incorporation is prevented by cycloheximide, ammonium chloride, primaquine, or chloroquine, suggesting that Na+-K+-ATPase originates in an intracellular pool and that its surface insertion requires synthesis of new enzyme or of a protein factor, since it is carried to the surface membrane through a mechanism of exocytosis. In summary, asymmetric distribution of ion pumps depends 1) on polarized insertion of Na+-K+-ATPase as well as 2) on removal or inactivation of misplaced enzyme.