Conformational change in the mitochondrial channel, VDAC, detected by electron cryo-microscopy.

Crystalline arrays of the voltage-dependent channel, VDAC, can be produced by treatment of Neurospora mitochondrial outer membranes with phospholipase A2. The membrane crystals undergo a lateral phase transition (lattice contraction) that can be induced by an amphipathic polyanion, which also reduces the channel's gating potential. Electron cryo-microscopy of frozen-hydrated crystals indicates that the mean projected diameters of the channels do not decrease with lattice contraction. Instead, contraction is associated with the disappearance of lateral protein "arms" that normally extend between the channels. A model is presented that explains the changes in channel packing and gating potential in terms of a conformational change involving the movement of a protein "arm" between the bilayer and the channel.     

Interaction of bilirubin with human erythrocyte membranes. Bilirubin binding to neuraminidase- and phospholipase-treated membranes.

Saturable bilirubin binding to human erythrocyte membranes was measured before and after digestion with neuraminidase and phospholipases. Neuraminidase-treated erythrocyte membranes did not show any change in their binding properties, indicating that gangliosides could be excluded as candidates for saturable bilirubin-binding sites on erythrocyte membranes. Although bilirubin-binding properties of the membranes did not change after phospholipase D digestion, either, phospholipase C treatment greatly enhanced bilirubin binding. Thus it is suggested that a negatively charged phosphoric acid moiety of phospholipids on the membrane surface may play a role to prevent a large amount of bilirubin from binding to the membranes. Further saturable bilirubin binding to inside-out sealed erythrocyte membrane vesicles showed values comparable with those of the right-side-out sealed membranes, suggesting that the bilirubin-binding sites may be distributed on both outer and inner surfaces of the membranes, or may exist in the membranes where bilirubin may be accessible from either side.     

Phospholipase C digestion of rat liver plasma membrane stimulates adenylate cyclase

Brief treatment of rat liver plasma membranes with phospholipase C of Clostridium welchii increased both the ratio of saturated to unsaturated fatty acids and the ratio of cholesterol to phospholipids. Using 5-doxylstearic acid spin probes two breaks at 29 and 19.6 °C could be observed in the order parameter, S_A, vs temperature curve for untreated membranes. Upon phospholipase C digestion the lower phase transition temperature was shifted to 23 °C, while the higher phase transition temperature could not be detected up to 40 °C. The order parameter, S_A, was consistently higher at all temperatures in the phospholipase C-treated membranes. As phospholipase C is known to attack the outer lamella, these results can be interpreted as indicating an increase in ordering (i.e., decrease in fluidity) of the outer membrane lamella. On the other hand, an increase in basal activity of adenylate cyclase of the treated membranes was observed with an apparent reduction of the activation energies both below and above the break (at 20 °C) in the Arrhenius plot of enzyme activity. Phospholipase C treatment did not affect the temperature of the break in Arrhenius kinetics of the enzyme. The results are discussed in terms of the role of the ordering state of membrane lipids in adenylate cyclase activity.     

Evidence that the crystalline arrays in the outer membrane of Neurospora mitochondria are composed of the voltage-dependent channel protein

Antibodies were raised in rabbits against the outer membrane of Neurospora mitochondria. Antibodies were obtained that were specific for this membrane's major polypeptide ($\text{M, 31 000}$) and its slower-migrating derivatives on SDS-polyacrylamide gels. These antibodies inhibited the insertion into phospholipid bilayers of voltage-dependent ion channels from detergent extracts of the mitochondrial outer membranes. The same antibodies bound preferentially to membranes containing crystalline surface arrays in outer mitochondrial membrane fractions. These results indicate that the 31 kDa polypeptide is a component both of the ion channels and of the membrane arrays, suggesting identity between the functional and structural entities.     

Localization of phosphatidylcholine in outer envelope membrane of spinach chloroplasts

We have examined the effects of phospholipase C from Bacillus cereus on the extent of phospholipid hydrolysis in envelope membrane vesicles and in intact chloroplasts. When isolated envelope vesicles were incubated in presence of phospholipase C, phosphatidylcholine and phosphatidylglycerol, but not phosphatidylinositol, were totally converted into diacylglycerol if they were available to the enzyme (i.e., when the vesicles were sonicated in presence of phospholipase C). These experiments demonstrate that phospholipase C can be used to probe the availability of phosphatidylcholine and phosphatidylglycerol in the cytosolic leaflet of the outer envelope membrane from spinach chloroplasts. When isolated, purified, intact chloroplasts were incubated with low amounts of phospholipase C (0.3 U/mg chlorophyll) under very mild conditions (12 degrees C for 1 min), greater than 80% of phosphatidylcholine molecules and almost none of phosphatidylglycerol molecules were hydrolyzed. Since we have also demonstrated, by using several different methods (phase-contrast and electron microscopy, immunochemical and electrophoretic analyses) that isolated spinach chloroplasts, and especially their outer envelope membrane, remained intact after mild treatment with phospholipase C, we can conclude that there is a marked asymmetric distribution of phospholipids across the outer envelope membrane of spinach chloroplasts. Phosphatidylcholine, the major polar lipid of the outer envelope membrane, is almost entirely accessible from the cytosolic side of the membrane and therefore is probably localized in the outer leaflet of the outer envelope bilayer. On the contrary, phosphatidylglycerol, the major polar lipid in the inner envelope membrane and the thylakoids, is probably not accessible to phospholipase C from the cytosol and therefore is probably localized mostly in the inner leaflet of the outer envelope membrane and in the other chloroplast membranes.     

Phospholipid composition modifications influence phospholipase A2 activity in rat liver plasma membranes

The influence of the phospholipid composition and the physico-chemical properties of rat liver plasma membranes on the activity of membrane-bound phospholipase A2 has been investigated. The plasma membrane composition was modified by the aid of exogenous phospholipases A2, C and D, and by butanol treatment. The partially delipidated membranes thus obtained were enriched with different phospholipids. The steady-state fluorescent anisotropy of 1,6-diphenyl-1,3,5-hexatriene and the lipid order parameter-SDPH in the modified membranes were calculated. It was established that the activity of the membrane-bound phospholipase A2 was higher in rigid membranes and was decreased when the membrane lipid bilayer was fluidized.     

Asymmetry of the site of choline incorporation into phosphatidylcholine of rat liver microsomes.

[14C]Choline was incorporated into microsomal membranes in vivo, and from CDP-[14C]choline in vitro, and the site of incorporation determined by hydrolysis of the outer leaflet of the membrane bilayer using phospholipase C from Clostridium welchii. Labelled phosphatidylcholine was found to be concentrated in the outer leaflet of the membrane bilayer with a specific activity approximately three times that of the inner leaflet. During incorporation of CDP-choline and treatment with phospholipase C the vesicles retained labelled-protein contents indicating that they remained intact. When the microsomes were opened with taurocholate after incorporation of [14C]choline in vivo, the labelled phosphatidylcholine behaved as a single pool. Selective hydrolysis of labelled phosphatidylcholine in intact vesicles is not, therefore, a consequence of specificity of phospholipase C. These results indicate that the phosphatidylcholine of the outer leaflet of the microsomal membrane bilayer is preferentially labelled by the choline-phosphotransferase pathway and that this pool of phospholipid does not equilibrate with that of the inner leaflet.     

Relation between various phospholipase actions on human red cell membranes and the interfacial phospholipid pressure in monolayers.

The action of purified phospholipases on monomolecular films of various interfacial pressures is compared with the action on erythrocyte membranes. The phospholipases which cannot hyorolyse phospholipids of the intact erythrocyte membrane, phospholipase C from Bacillus cereus, phospholipase A2 from pig pancreas and Crotalus adamanteus and phospholipase D from cabbage, can hydrolyse phospholipid monolayers at pressure below 31 dynes/cm only. The phospholipases which can hydrolyse phospholipids of the intact erythrocyte membrane, phospholipase C from Clostridium welchii phospholipase A2 from Naja naja and bee venom and sphingomyelinase from Staphylococcus aureus, can hydrolyse phospholipid monolayers at pressure above 31 dynes/cm. It is concluded that the lipid packing in the outer monolayer of the erythrocyte membrane is comparable with a lateral surface pressure between 31 and 34.8 dynes/cm.     

Effect of phospholipase C, trypsin and neuraminidase on binding of bilirubin to mammalian erythrocyte membranes.

Binding of bilirubin to erythrocyte membranes of human, buffalo, sheep and goat was studied after phospholipase C, trypsin and neuraminidase treatment. Phospholipase C and trypsin treatment of membranes greatly enhanced the bilirubin binding in all mammalian species, whereas, neuraminidase treatment resulted into a small increase in the membrane-bound bilirubin. Human erythrocyte membranes bound the highest amount of bilirubin, whereas buffalo, sheep and goat erythrocyte membranes showed different mode of bilirubin binding. The order of bilirubin binding to unmodified as well as neuraminidase-treated erythrocyte membranes was: human>sheep>buffalo>goat; the order was: human>buffalo>sheep>goat; in phospholipase C- and trypsin-treated erythrocyte membranes. These binding results indicate that membrane phospholipids are directly involved in the interaction of bilirubin with the membranes as the differences observed in the membrane-bound bilirubin among mammalian species were directly correlated with the sum of choline phospholipids, especially phosphatidylcholine and sphingomyelin content of the erythrocyte membranes. The negatively charged phosphate moiety of phospholipids of the membranes appears to inhibit a large amount of bilirubin binding to the membrane as its removal by phospholipase C greatly enhanced the binding. Furthermore, membrane proteins and carbohydrate also seem to play a significant regulatory function on the binding as their degradation and/or removal in the form of glycopeptides by trypsin expose a large number of bilirubin binding sites.     

Modification of phospholipids in erythrocyte membranes by phospholipase D. A fluorescence and ESR spectroscopic study

The conversion of more than 65% of the phospholipids in human erythrocyte membranes to phosphatidyl-methanol and phosphatidic acid by incubation with phospholipase D and methanol increased the dissociation constant of the fluorescence probe ANS compared to untreated membranes, but did not affect the number of binding sites and the limiting fluorescence enhancement at maximal binding (Imax). On the contrary, the cationic fluorescence probe dansylcadaverin showed additional binding sites without a change in Kd and an increase of Imax upon incubation with phospholipase D treated erythrocyte membranes compared to incubations of membranes with the original phospholipid pattern. The characteristic temperature-dependence of the quenching of the membrane protein fluorescence by a membrane-bound nitroxide-labeled stearic acid was not influenced by the modification of the phospholipids. A slight reduction of the order parameter, S, determined by ESR-spectroscopy with the same nitroxide spin-labeled fatty acid incorporated into modified membranes compared to controls was found at 40 degrees C, but not at 25 degrees C. The results were interpreted as an indication of membrane domains that retained their physical properties and lipid composition during the incubation with phospholipase D.     

Gonadotropin receptors in plasma membranes of bovine corpus luteum. I. Effect of phospholipases on the binding of 125I-choriogonadotropin by membrane-associated and solubilized receptors.

The ability of bovine corpus luteum plasma membranes to bind 125I-choriogonadotropin has been examined after prior treatment of the membranes with phospholipases A, C, and D. Treatment of the purified membranes with low concentrations of phospholipases A and C resulted in the inhibition of the binding of 125I-choriogonadotropin to its receptors, whereas phospholipase D had no effect. Receptor activity was decreased by low concentrations of phospholipase A from either bee venom, Vipera russelli or Crotalus terrificus terrificus. Similarly, low concentrations of phospholipase C from Clostridium perfringens and Clostridium welchii also inhibited the binding activity while comparatively higher concentrations of phospholipase C from Bacillus cereus were required to achieve comparable inhibition. The time required to produce 50% inhibition of in vitro binding by phospholipases A and C was found to be 6 and 23 min, respectively. Upon either removal or chelation of calcium ions by ethylene glycol bis(beta-aminoethyl ether)-N,N'-tetraacetic acid (EGTA) both enzymes were completely inhibited as evidenced by the complete retention of the membrane binding activity. The decrease in the specific binding of choriogonadotropin to membranes after phospholipase digestion resulted in a decrease in the number of binding sites and was not accompanied by a change in the affinity of the hormone-receptor complex. The rates of association and dissociation of the 125I-choriogonadotropin-receptor complex and the equilibrium dissociation constant (Kd) were nearly identical in untreated and phospholipase-treated membranes. Phospholipases did not have any effect on the preformed hormone-receptor complex or on solubilized receptor. Filtration through Sepharose 6B of solubilized 125I-choriogonadotropin-receptor complex from untreated membranes or membranes which had been pretreated with phospholipase C prior to carrying out hormone binding did not alter the profile (Kav 0.38). Gel filtration of membranes treated with phospholipase A showed two peaks of bound radioactivity with distribution coefficients (Kav) of 0.08 and 0.35, respectively.     

Phosphoglycerides and phospholipase C in membrane fractions of Escherichia coli B.

The phospholipid composition and the phospholipase C activity of envelope fractions of Escherichia coli B were determined with special consideration of fractions containing sites at which an attachment of inner and outer membranes had been observed in the electron microscope (Int.M). Phosphoglycerides labeled with [14C]palmitic acid and [3H]serine were extracted from membrane fractions and identified by two-dimensional thin-layer chromatography. The amount of phosphatidylethanolamine was highest in the outer membrane, whereas the amounts of phosphatidylglycerol and cardiolipin were highest in the inner membrane. The Int.M fractions were observed to have concentrations of phospholipids intermediate to those of the inner and outer membranes. This result supports the assumption that a concentration gradient of inner membrane-outer membrane lipids might exist at the membrane contact sites. The highest phospholipase C activity was detected in the inner membrane and Int.M fractions. The presence of phospholipase C and other lipolytic enzymes in the Int.M fractions suggests a possible involvement of adhesion sites in lipid metabolism, adding a further set of activities to the function of these domains.     

Surface topography and molecular stoichiometry of the mitochondrial channel, VDAC, in crystalline arrays.

The mitochondrial outer membrane contains a protein, called VDAC, that forms large aqueous pores. In Neurospora crassa outer membranes, VDAC forms two-dimensional crystalline arrays whose size and frequency can be greatly augmented by lipase treatment of these membranes (C. Mannella, Science 224, 165, 1984). Fourier filtration and surface reconstruction of freeze-dried/shadowed (45 degrees) arrays produced detailed images of two populations of crystals, whose lattices are mirror images of each other. Most likely, this technique has revealed both surfaces of the same two-dimensional crystal with lattice parameters: a = 12.3 +/- 0.1 nm, b = 11.2 +/- 0.1 nm, and theta = 109 +/- 1 degree. Three-dimensional reconstructions of the surface reliefs on both sides of the crystal show them to be very similar. The majority of the protein forming the channel appears to be at or below the level of the membrane. To address the issue of the number of 30-kDa polypeptides that form a VDAC channel, measurements of mass per unit area were carried out by analyzing scanning transmission electron micrographs of unstained, freeze-dried arrays. The crystal form used for mass analysis contained the same motif of six stain-accumulating centers per unit cell, with p2 symmetry as in the oblique configuration, but it had a different orientation relative to the lattice lines. These data yielded a surface density of 1.9 +/- 0.2 kDa/nm2, indicating that there is a one-to-one ratio between VDAC polypeptides and the channels visualized in filtered electron micrographs, and that VDAC membrane crystals contain 68% protein and 32% lipid by mass.     

Asymmetric distribution of phospholipids in membranes from Mycobacterium phlei

The distribution of phospholipids in the membranes of Mycobacterium phlei has been studied by the use of phospholipase C and trinitrobenzenesulfonic acid. In inverted membrane vesicles, whose external surface apparently corresponds topologically to the cytoplasmic surface of the membrane in intact cells, 80% of the phosphatidyl ethanolamine, 24% of diphosphatidyl glycerol, and 13% of phosphatidyl inositol are accessible to cleavage by phospholipase C. These results are in agreement with the finding that 70–75% of phosphatidyl ethanolamine in the membrane is accessible to chemical modification by trinitrobenzenesulfonic acid or dimethylsuberimidate at 4 °C. It can be inferred that in the inverted membrane the majority of phosphatidyl ethanolamine is present on the outer half of the lipid bilayer while inner half constitutes primarily other phospholipids namely phosphatidyl inositol and diphosphatidyl glycerol. Phospholipase C treatment of ETP membranes selectively impairs the active transport of Ca²⁺ without affecting the generation of a proton gradient, respiration, and coupled phosphorylation.     

Transbilayer distribution of phospholipids in photoreceptor membrane studied with trinitrobenzenesulfonate alone and in combination with phospholipase D

In a further study of the transbilayer distribution of phospholipids in rod disk membranes, the amino group reagent, trinitrobenzenesulfonate, and the phospholipid-hydrolyzing enzyme, phospholipase D, have been used alone and in combination.Under carefully defined conditions (1 mM trinitrobenzenesulfonate, pH 7.4, 20°C, darkness), trinitrobenzenesulfonate yields limited final levels of modification of phosphatidylethanolamine and phosphatidylserine, suggesting only minor reagent penetration and membrane disturbance under these conditions.Treatment of stacked disks with trinitrobenzenesulfonate under these conditions leads to a biphasic modification of the a aminophospholipids. Relatively fast (less than 1 h) modification of 50% phosphatidylethanolamine and 40% phosphatidylserine occurs, slowly rising (approx. 3 h) to 60 and 50%, respectively.Extensive treatment of stacked disks with phospholipase D leads to the hydrolysis of 55% phosphatidylcholine and 50% phosphatidylethanolamine, while phosphatidylserine is hardly attacked by this enzyme.Treatment of stacked disks with trinitrobenzenesulfonate after prior treatment with phospholipase D leads to no further modification than that maximally obtained with either reagent alone: about one-half of the three major phospholipid classes is accessible. Although both reagents differ greatly in molecular size, mode of action and other properties, they apparently see the same pool of phosphatidylethanolamine, their joint substrate. Considering that we start with the original right-side-out configuration, that all phospholipids can in principle be modified (no shielding) and that the membrane remains essentially intact, we conclude that the accessible lipid pool represents the outer face of the disk membranes.These results confirm our earlier conclusions from treatment with three phospholipases that the three major phospholipids are nearly symmetrically distributed over the two faces of the disk membrane.The divergence with the conclusions of other investigators is most likely explained by their use of disk membranes (disk vesicles) in which the original phospholipid distribution had not been maintained and/or of conditions under which trinitrobenzenesulfonate markedly penetrates the membrane.     

The role of lipid-protein interactions in NADH-cytochrome c reductase (rotenone-insensitive) of rat liver mitochondria

The phospholipid depletion of rat liver mitochondria, induced by acetone-extraction or by digestion with phospholipase A₂ or phospholipase C, greatly inhibited the activity of NADH-cytochrome c reductase (rotenone-insensitive). A great decrease of the reductase activity also occurred in isolated outer mitochondrial membranes after incubation with phospholipase A₂. The enzyme activity was almost completely restored by the addition of a mixture of mitochondrial phospholipids to either lipid-deficient mitochondria, or lipid-deficient outer membranes. The individual phospholipids present in the outer mitochondrial membrane induced little or no stimulation of the reductase activity. Egg phosphatidylcholine was the most active phospholipid, but dipalmitoyl phosphatidylcholine was almost ineffective. The lipid depletion of mitochondria resulted in the disappearance of the non-linear Arrhenius plot which characterized the native reductase activity. A non-linear plot almost identical to that of the native enzyme was shown by the enzyme reconstituted with mitochondrial phospholipids. Triton X-100, Tween 80 or sodium deoxycholate induced only a small activation of NADH-cytochrome c reductase (rotenone-insensitive) in lipiddeficient mitochondria. The addition of cholesterol to extracted mitochondrial phospholipids at a 1 : 1 molar ratio inhibited the reactivation of NADH-cytochrome c reductase (rotenone-insensitive) but not the binding of phospholipids to lipid-deficient mitochondria or lipid-deficient outer membranes.These results show that NADH-cytochrome c reductase (rotenone-insensitive) of the outer mitochondrial membrane requires phospholipids for its activity. A mixture of phospholipids accomplishes this requirement better than individual phospholipids or detergents. It also seems that the membrane fluidity may influence the reductase activity.     

Influence of phospholipid environment on the phosphatidylethanolamine: ceramide-phosphorylethanolamine transferase activity in rat liver plasma membranes.

1. Plasma membranes were treated with phospholipase A2, phospholipase C or phospholipase D. The phosphatidylethanolamine:ceramide-phosphorylethanolamine transferase was deactivated by phospholipase C treatment, whereas phospholipase A2 and phospholipase D did not affect the enzyme. 2. Incorporation of phosphatidylethanolamine and phosphatidylglycerol into partially delipidated plasma membranes resulted in significant stimulation of the transferase, whereas inclusion of sphingomyelin and phosphatidylserine suppressed the enzyme activity. Our results suggest that phosphatidylserine is a regulator of sphingomyelin level in membranes. 3. The activity of phosphatidylethanolamine:ceramide-phosphorylethanolamine transferase was not influenced by the fluidity of its lipid environment.     

Relation between various phospholipase actions on human red cell membranes and the interfacial phospholipid pressure in monolayers

The action of purified phospholipases on monomolecular films of various interfacial pressures is compared with the action on erythrocyte membranes. The phospholipases which cannot hydrolyse phospholipids of the intact erythrocyte membrane, phospholipase C from Bacillus cereus, phospholipase A₂ from pig pancreas and Crotalus adamanteus and phospholipase D from cabbage, can hydrolyse phospholipid monolayers at pressure below 31 dynes/cm only.The phospholipases which can hydrolyse phospholipids of the intact erythrocyte membrane, phospholipase C from Clostridium welchii phospholipase A₂ from Naja naja and bee venom and sphingomyelinase from Staphylococcus aureus, can hydrolyse phospholipid monolayers at pressure above 31 dynes/cm. It is concluded that the lipid packing in the outer monolayer of the erythrocyte membrane is comparable with a lateral surface pressure between 31 and 34.8 dynes/cm.     

The role of choline on the activity-temperature relationship of brush-border alkaline phosphatase

We have studied the effect of choline on the activity and temperature dependency of the brush-border alkaline phosphatase isoenzymes from rat intestine (tissue-specific type), and from kidney and placenta (tissue-nonspecific type). The removal of choline with phospholipase D resulted in the loss of enzyme activity in all the membranes, whereas in situ loss in the discontinuity of Arrhenius plots occurred in the kidney and the placental membranes, but not in the intestinal membranes. The lost activity was restored either by addition of free choline or phosphatidylcholine or by the removal of the enzyme from the membrane surface. Intestinal enzyme was removed by papain, while the tissue-nonspecific enzyme was released by subtilisin and by phosphatidylinositol-specific phospholipase C. The enzyme from kidney and placental membranes aggregated (rho = 1.13) upon removal of choline, and addition of choline resulted in disaggregation (rho = 1.03). Conversion of discontinuous to continuous linear plots of alkaline phosphatase in the kidney and placental membranes paralleled the increase in membrane phosphatidic acid content, and the decrease in total phosphatidylcholines. The intestinal enzyme produced plots with break points at all phosphatidic acid/phosphatidylcholine ratios. The change brought about by treatment with phospholipidase D was not due to changes in the half-saturation kinetics (Km) for the substrate. Based on these studies we conclude that the active site of the tissue-nonspecific phosphatase is approximated to exterior membrane cholines, as in the case of the intestinal isoenzyme; that despite similar effects on the membrane content of phospholipids, phospholipase D treatment caused much greater effects on the tissue-nonspecific enzyme, as assessed by Arrhenius plots and density centrifugation; that these effects are due to different protein structures rather than to a lipid milieu unique to each brush-border membrane.