Marker enzymes,phospholipids and acyl group composition of a somal plasma membrane fraction isolated from rat cerebral cortex: a comparison with microsomes and synaptic plasma membranes

Subjecting brain homogenates to differential speed and sucrose density gradient centrifugation resulted in the isolation of a membrane fraction from the post-mitochondrial supernatant with properties and marker enzyme profiles typical of plasma membranes. This membrane fraction is compared with the microsomes and the synaptic plasma membranes isolated from synaptosomes. Like the synaptic plasma membranes, membranes obtained from the post-mitochondrial supernatant were enriched five-fold in 5′-nucleotidase activity. However, the latter membranes were lower in (Na⁺, K⁺)-ATPase activity and higher in NADPH-cytochrome C reductase activity as compared to the synaptic plasma membranes. The post-mitochondrial plasma membranes were also different from the microsomes in their respective marker enzyme activities. Electron microscopic examination indicated largely membranous vesicles for both plasma membrane fractions with little contamination by myelin, mitochondra and intact synaptosomes. The phospholipid and acyl group profiles of the two plasma membrane fractions were surprisingly similar, but they were different from the characteristic profiles of myelin and mitochondria. It is concluded that plasma membranes isolated from the post-mitochondrial supernatant fraction are derived largely from neuronal and glial soma and are thus designated the somal plasma membrane fraction.     

Isolation and characterization of outer and inner envelope membranes of cyanelles from a glaucocystophyte, Cyanophora paradoxa

Envelope membranes were isolated by sucrose density gradient floatation centrifugation from the homogenate of cyanelles prepared from Cyanophora paradoxa. Two yellow bands were separated after 40 h of centrifugation. The buoyant density of one of the two fractions (fraction Y2) coincided with that of inner envelope membranes of spinach or plasma membranes of cyanobacteria. The other yellow fraction (fraction Y1) migrated to top of sucrose-gradient even at 0% sucrose. Pigment analysis revealed that the heavy yellow fraction was rich in zeaxanthin while the light fraction was rich in β-carotene, and the both fractions contained practically no chlorophylls. Another yellow fraction (fraction Y3) was isolated from the phycobiliprotein fraction, which was the position where the sample was placed for gradient centrifugation. Its buoyant density and absorption spectra were similar to outer membranes of cyanobacteria. We have assigned fractions Y2 and Y3 as inner and outer envelope membrane fractions of cyanelles, respectively. Protein compositions were rather different between the two envelope membranes indicating little cross-contamination among the fractions. H. Koike and Y. Ikeda contributed equally.     

Localization of 2',3'-cyclic nucleotide-3'-phosphohydrolase of rabbit brain by sedimentation in a continuous sucrose gradient.

Abstract— The enzyme 2′,3′-cyclic nucleotide-3′-phosphohydrolase (CNP) has been assayed in fractions from a continuous sucrose density gradient zonal centrifugation of rabbit brain homogenates. Basic protein (BP) was also assayed by a radioimmunomethod. Fractions were examined by SDS-polyacrylamide gel electrophoresis and by electron microscopy. These studies show that the major membrane fractions in the gradient differ greatly in the content of CNP and BP, and of high molecular weight proteins (HMW). The lightest membrane fractions contained numerous multilamellae, the highest content of BP and the lowest content of CNP and HMW, while the heaviest membrane fractions contained single membrane fragments and vesicles of unknown origin, the lowest content of BP and the highest content of CNP and HMW. The fraction containing the largest amount of membrane measured by turbidity, protein content, and water-washed dry weight contained only half the CNP specific activity of a denser fraction in the gradient. CNP specific activity in the lightest fractions was insignificant compared to that of denser fractions. Thus, we conclude that this enzyme may be absent from the typical multilamellar myelin structures but present in the single-membrane structures associated with myelin, such as the glial membrane and the paranodal segments of myelin adjacent to the axon. BP appears to occupy the opposite positions, highest in the multilamellae and lowest in the single-membrane structures of myelin. These studies do not exclude the possibility that CNP may not be bound to myelin membranes, but rather to a membrane of different origin. Evidence that this enzyme is a myelin-marker enzyme is circumstantial. Our evidence indicates the enzyme could be present either in a unique portion of myelin membranes or in another membrane structure.     

An analytical approach to the preparation and characterization of subcellular membranes from canine mesenteric arteries

Subcellular membrane fractions were isolated from dog mesenteric arteries by differential and isopynic sucrose density gradient centrifugations. Isolated membrane fractions were characterized by marker enzyme activities, morphological features and sodium dodecyl sulfate-polyacrylamide gel electrophoretic patterns. Our results show that the microsomal fraction isolated by conventional differential centrifugation was highly heterogenous and contained substantial amount of plasma membranes which could be further enriched as a light density membrane fraction on a discontinuous sucrose density gradient. The microsomal fraction and its subfractions were vesicular in appearance under electron microscope and were capable of binding and actively transporting Ca2+. The binding of Ca2+ and ATP-supported Ca2+-transport in the presence or absence of oxalate paralleled the distribution of plasma membrane marker enzyme activities suggesting that plasma membranes in vascular smooth muscle may play a major role in handling Ca2+ and thus the control of contractile function.     

Studies on isolated subcellular components of cat pancreas. I. Isolation and enzymatic characterization.

Pancreas of the cat was fractionated into its subcellular components by centrifugation through an exponential ficoll-sucrose density gradient in a zonal rotor. This enables a preparation of four fractions enriched in plasma membranes, endoplasmic reticulum, mitochondria and zymogen granules, respectively. The first fraction, enriched by 9- to 15-fold in the plasma membrane marker enzymes, hormone-stimulated adenylate cyclase, (Na+K+)-ATPase, and 5'-nucleotidase, is contaminated by membranes derived from endoplasmic reticulum but is virtually free from mitochondrial and zymogen-granule contamination. The second fraction from the zonal gradient shows only moderate enrichment of the above marker enzymes but contains a considerable quantity of plasma membrane marker enzymes and represents mostly rough endoplasmic reticulum. The third fraction contains the bulk of mitochondria and the fourth mainly zymogen granules as assessed by electron microscopy and marker enzymes for both mitochondria and zymogen granules, namely succinic dehydrogenase, trypsin and amylase. Further purification of the plasma membrane fractions by differential and sucrose step-gradient centrifugation yields plasma membranes enriched 40-fold in basal and hormone-stimulated adenylate cyclase and (Na+K+)-ATPase.     

Isolation and partial characterization of chick brain synaptic plasma membranes.

An isolation procedure for synaptic plasma membranes from whole chick brain is reported that uses the combined flotation-sedimentation density gradient centrifugation procedure described by Jones and Matus (Jones, D. H. and Matus, A. I. (1974) Biochim. Biophys. Acta 356, 276-287) for rat brain. The particulate of the osmotically shocked and sonicated crude mitochondrial fraction was used for a flotation-sedimentation gradient step. Four fractions were recovered from the gradient after 30 min centrifugation. The fractions were identified and characterized by electron microscopy and by several markers for plasma membrane and other subcellular organelles. Fraction 2 was recovered from the 28.5-34% (w/v) sucrose interphase and contained the major part of the activities of the neuronal plasma membrane marker enzymes. The specific activities of the (Na+ +K+)-activated ATPase (EC 3.6.1.3), acetylcholinesterase (EC 3.1.1.7) and 5'-nucleotidase (EC 3.1.3.5) were, respectively, 4.5, 2.0 and 1.2 times higher than in the homogenate. However, Fraction 2 also contained considerable amounts of activities of putative lysosomal and microsomal markers in addition to lower amounts of mitochondrial and myelin markers. Although no prepurification of synaptosomes from the crude mitochondrial fraction was performed, the synaptic plasma membranes obtained showed many properties analogous to similar preparations from rat brain described in recent years.     

Isolation and partial characterization of chick brain synaptic plasma membranes

An isolation procedure for synaptic plasma membranes from whole chick brain is reported that uses the combined flotation-sedimentation density gradient centrifugation procedure described by Jones and Matus (Jones. D. H. and Matus. A. I. (1974) Biochim. Biophys. Acta 356, 276–287) for rat brain. The particulate of the osmotically shocked and sonicated crude mitochondrial fraction was used for a flotation-sedimentation gradient step. Four fractions were recovered from the gradient after 30 min centrifugation. The fractions were identified and characterized by electron microscopy and by several markers for plasma membrane and other subcellular organcelles. Fraction 2 was recovered from the 28.5–34% (w/v) sucrose interphase and contained the major part of the activities of the neuronal plasma membrane marker enzymes. The specific activities of the (Na⁺+K⁺)-activated ATPase (EC 3.6.1.3), acetylcholinesterase (EC 3.1.1.7) and 5′-nucleotidase (EC 3.1.3.5) were, respectively, 4.5. 2.0 and 1.2 times higher than in the homogenate. However, Fraction 2 also contained considerable amounts of activities of putative lysosomal and microsomal markers in addition to lower amounts of mitochondrial and myelin markers. Although no prepurification of synaptosomes from the crude mitochondrial fraction was performed, the synaptic plasma membranes obtained showed many properties analogous to similar preparations from rat brain described in recent years.     

Highly purified plasma membranes from rat hepatocytes following rate-isopycnic zonal centrifugation

Plasma membranes from liver parenchymal cells were isolated by rate-isopycnic zonal centrifugation. A method is described for the Beckman size 15 zonal rotor. It involved preparation from a perfused liver of a parenchymal cell-enriched homogenate in isoosmotic sucrose. The nuclear fraction containing membranes was recovered by centrifugation. The resuspended pellet was applied on the gradient of the zonal rotor. The isolated membranes had the same isopycnic banding density as 37% sucrose (w/w). The specific activity of 5′-nucleotidase, a widely used plasma membrane marker, was 105 μmoles·(mg protein)^?1·h^?1 being enriched by a factor of 50 as compared with parenchymal cell homogenate. The plasma membrane fraction was free of the mitochondrial and lysosomal enzymes, succinate dehydrogenase and acid phosphatase. No DNA and 10 μg RNA per mg plasma membrane protein were found. The purity of the membranes and their morphological appearance were controlled by electron microscopy. The preparation consisting of large membrane sheets showed a considerable purification away from other cellular components. A comparison with similar methods indicates that plasma membranes of a higher degree of purity can be obtained from parenchymal cells.     

Basal-lateral membranes from rabbit renal cortex prepared on a large scale in a zonal rotor

Basal-lateral membranes from the renal cortex of the rabbit were isolated by sucrose gradient centrifugation in a zonal rotor which allows for a large-scale preparation of these membranes. A heterogeneous population of membranes (P4) which contained 29% of the (Na+ + K+)-ATPase found in the homogenate of renal cortex was prepared by differential centrifugation. When pellet P4 was subjected to centrifugation in a sucrose gradient the activity of (Na+ + K+)-ATPase, a marker for basal-lateral membranes, could be separated from enzymatic markers of other organelles. The specific activity of (Na+ + K+)-ATPase was enriched 12-fold at a density of 1.141 g/cm3. Membranes (P alpha) contained in the (Na+ + K+)-ATPase-rich fractions consisted primarily of closed vesicles which exhibited probenecid inhibitable transport of rho-aminohippurate. These membranes did not exhibit Na+-dependent, phlorizin-inhibitable D-glucose transport. Sodium dodecyl sulfate polyacrylamide gel electrophoresis of proteins from P alpha revealed at least six major protein bands with molecular weights of 91000, 81000, 73000, 65000, 47000 and 38000. A small fraction of total alkaline phosphatase found in the homogenate was found in pellet P4. Membranes containing this alkaline phosphatase activity were distributed widely over the gradient, with peak activity found at a density of 1.141 g/cm3. In contrast, when brush borders were subjected to gradient centrifugation under the same conditions as P4, alkaline phosphatase was found in a narrow distribution, with peak activity at a density of 1.158 g/cm3. The principle subcellular localization of the alkaline phosphatase found in P4 could not be determined unambiguously from the data, but the activity did not seem to be primarily associated with classical brush borders.     

Isolation of plasma membranes from corn roots by sucrose density gradient centrifugation: an anomalous effect of ficoll

An investigation was conducted into the isolation of plasma membrane vesicles from primary roots of corn (Zea mays L., WF9 × M14) by sucrose density gradient centrifugation. Identification of plasma membranes in cell fractions was by specific staining with the periodic-chromic-phosphotungstic acid procedure. Plasma membrane vesicles were rich in K⁺-stimulated ATPase activity at pH 6.5, and equilibrated in linear gradients of sucrose at a peak density of about 1.165 g/cc. It was necessary to remove mitochondria (equilibrium density of 1.18 g/cc) from the homogenate before density gradient centrifugation to minimize mitochondrial contamination of the plasma membrane fraction. Endoplasmic reticulum (NADH-cytochrome c reductase) and Golgi apparatus (latent IDPase) had equilibrium densities in sucrose of about 1.10 g/cc and 1.12 to 1.15 g/cc, respectively. A correlation (r = 0.975) was observed between K⁺-stimulated ATPase activity at pH 6.5 and the content of plasma membranes in various cell fractions. ATPase activity at pH 9 and cytochrome c oxidase activity were also correlated.     

Isolation and characterization of plasma membrane from lactating bovine mammary gland

Plasma membranes were isolated from lactating bovine mammary gland. Two crude membrane fractions; medium/d 1.033 (light membrane) and 1.033/1.053 interfaces (heavy membrane), were obtained by Ficoll density gradient centrifugation of osmotically washed microsomal fraction. Two crude membranes were further purified separately by sucrose density gradient centrifugation. Both light and heavy membranes banded at a sucrose density of 1.14. The purified membranes appeared as heterogeneous smooth membrane vesicles on electron microscopy. The contaminating suborganelles were not detected. The yield of the purified membranes relative to the homogenate was 1.2%. The degree of purity of the membranes was shown by a great increase in the specific activity of 5′-nucleotidase over the homogenate of 20-fold for light membrane and of 16-fold for heavy membrane. The relative activities of Mg²⁺-ATPase, (Na⁺ + K⁺)-ATPase, γ-glutamyl transpeptidase, phosphodiesterase I, akaline phosphatase and xanthine oxidase were also high (12–18-times) and nearly 20% of these enzymes was recovered. The activity of marker enzyme for mitochondria, endoplasmic reticulum and Golgi apparatus was very low, while that of acid phosphatase for lysosome was relatively high (5-times). DNA and RNA contents were very low. The major polypeptides rich in other suborganelles were not detected profoundly in the membrane fraction and the polypeptide compositions in both light and heavy membranes were similar upon SDS-polyacrylamide gel electrophoresis.     

Localization of Mg2+-ATPase in the membranes of the skeletal muscles and myocardium of the frog Rana temporaria

Using differential centrifugation in sucrose density gradient, from muscles of the frog fractions were obtained which contain fragments of sarcolemma, as well as membranes of T-system tubules and sarcoplasmic reticulum. In isolated membrane fractions, studies were made on the activity of cation-stimulated ATPases (Na+, K+-, Ca2+, Mg2+- and Mg2+-ATPases). Enzymic and electrophoretic analyses showed that the highest content of Mg2+-ATPases is typical of the fractions which are located on the surface of 35% sucrose. The data obtained indicate that Mg2+-ATPase is the enzyme which is specific for the membranes of T-system tubules in skeletal muscles of not only birds but amphibians as well. From cardiac muscle of the frog, membrane fraction was isolated which is similar (with respect to its predominant content of Mg2+-ATPase) to the membranes of T-system tubules. It is suggested that the presence of Mg2+-ATPase in these membranes is a common property of phasic striated muscle fibers in all mature vertebrate animals.     

Isolation and characterization of Sertoli cell plasma membranes and associated plasminogen activator activity

Plasma membranes were isolated from the cultured Sertoli cells of 20-day-old rat testes by differential centrifugation and sucrose density fractionation. The distribution and purity of subcellular components was determined by marker enzyme analysis of gradient fractions. The plasma membrane fraction showed an enrichment in two plasma membrane marker enzymes, 5'-nucleotidase and ouabain-sensitive Na+/K+-ATPase-specific activities, of 9- and 23-fold, respectively. Forty-two percent and 52% of the total cellular 5'-nucleotidase and ouabain-sensitive Na+/K+-ATPase activities, respectively, were found in the membrane fraction. The protein yield of plasma membrane was approximately 6% of the total cellular protein. Two-dimensional polyacrylamide gel electrophoresis was used to compare [35S] methionine- and [3H] glucosamine-labeled membrane proteins. The incorporation of [35S] methionine and [3H] glucosamine was increased in several proteins when the cultured Sertoli cells were treated with follicle-stimulating hormone, insulin, retinol, and testosterone. Isolated Sertoli cell membranes contained a membrane-associated form of plasminogen activator. Analysis of this plasminogen activator demonstrated that the membrane-associated enzyme existed primarily as a single 38,000-40,000-Mr form.     

Isolation of the Ochromanas danica plasma membrane and identification of several membrane enzymes

Ochromonas danica cell homogenate can be fractionated by differential centrifugation into chloroplast, mitochondrial, ribosome, lysosomal, plasma membrane and soluble fractions. The plasma membrane fraction was further purified by discontinuous sucrose density gradient centrifugation and was found to be enriched 4–16-fold in the following enzymes: β-galactosidase, acid phosphatase, alkaline phosphatase, 5′-nucleotidase, and (Na⁺, K⁺)-ATPase. The role of plasma membrane phosphatase in the phosphate metabolism of plants is discussed.     

Subcellular localization of a membrane-associated transglutaminase activity in rat liver

Fractionation of rat liver by homogenization and differential centrifugation revealed that only about 83% of the transglutaminase activity in the tissue is in a soluble form, and that the remainder is associated with the particulate fraction. This latter activity remained with the membranes even after they were extensively washed to remove 99% of such soluble enzymes as lactate dehydrogenase and aldolase. Subsequent fractionation of the membranes by isopycnic density gradient centrifugation in sucrose resulted in a single band of transglutaminase activity at a density of 1.194 g/cm3. This activity was coincident with the major band of plasma membranes, which was identified by its content of 5'-nucleotidase, alkaline phosphodiesterase I, alkaline phosphatase and leucine aminopeptidase activities. After treatment with digitonin and fractionation on sucrose gradients, the transglutaminase activity and the plasma membrane marker enzyme activities were found at a new density of 1.210 g/cm3, while the enzyme markers for the other membrane fractions remained unchanged. From these data, we conclude that approximately 17% of the transglutaminase activity in rat liver is specifically associated with the plasma membranes.     

Purification and characterization of two plasma membrane domains from ejaculated bull spermatozoa

Plasma membranes were detached from ejaculated bull spermatozoa by a brief sonication in a moderately hypotonic medium, and the released plasma membranes were partially purified by differential centrifugation. The resulting fraction was enriched 8- and 15-fold in alkaline phosphatase and 5' nucleotidase activities, respectively, compared with the starting sonicated spermatozoa. This total plasma membrane fraction was separated into two distinct fractions by equilibrium density centrifugation on a continuous linear sucrose gradient. Two peaks of light scattering material were formed at densities of 1.117 and 1.148 g/ml. The denser peak contained most of the protein of the plasma membrane fraction, whereas nearly all the concanavalin A binding activity was found in the lighter peak. The two bands had distinctly different polypeptide compositions when analyzed by SDS PAGE. Polyclonal antibodies were raised in rabbits against a major integral membrane glycoprotein of each fraction (Mr of 92,000 in the light peak and 98,000 in the dense peak). The two antigens were detected on the surface of intact spermatozoa by indirect immunofluorescence microscopy. The 92-kD protein (present in the lighter band) was detected only on the plasma membrane of the acrosomal and anterior postacrosomal regions of the head. The 98-kD antigen, present in the heavier band, was localized to the surface of the postacrosomal region of the head, to the principal piece of the tail, and to the connecting piece between the head and tail. The exclusive localization of the 92-kD polypeptide to the surface of the anterior portion of the head was confirmed by immunoelectron microscopy. These data show that the two fractions isolated on the sucrose gradient originate from different regions of the sperm cell plasma membrane.     

Purification and partial characterization of plasma membranes from bovine spermatozoa

Bovine epididymal spermatozoa were subjected to nitrogen cavitation (600 psi for 10 min) to remove plasma membrane. Examination of the cavitated cells by electron microscopy revealed that the plasma membrane was preferentially removed from the periacrosomal and flagellar regions. Nuclear, mitochondrial and acrosomal membranes remained intact and attached to the spermatozoa, but the cytoplasmic droplets were frequently disrupted and their internal membrane-bound vesicles were released. Lower pressures (less than 200 psi) were relatively ineffective in removing the periacrosomal plasma membrane, while an intermediate pressure (400 psi) removed this membrane from about 70% of the spermatozoa. No apparent selectivity for removal of the periacrosomal and flagellar plasma membrane was observed as a function of cavitation pressure. The cavitated cells were separated from the plasma membranes by differential followed by linear sucrose density gradient centrifugation. Two distinct membrane populations were resolved on sucrose gradients and were designated Band I and Band II. Band I contained only spherical vesicles which arose from the plasma membrane. Surface labeling of intact cells confirmed the plasma membrane as the origin of Band I. The membranes of higher density comprising Band II were heterogeneous consisting of both spherical and flattened vesicles. When purified cytoplasmic droplets were cavitated and centrifuged on the sucrose gradient only Band II was obtained. These studies indicate that nitrogen cavitation of bovine epididymal spermatozoa can result in significant contamination of plasma membrane fractions by cytoplasmic droplet membranes unless appropriate differential centrifugation is used to separate the membrane fractions.     

Modulation of membrane composition of swine vascular smooth muscle cells by homologous lipoproteins in culture

Swine vascular smooth muscle cells were exposed to homologous low-density or high-density lipoprotein fractions for 24 h. Total cell membranes were isolated from the post-nuclear supernatant of the cell homogenates, fractionated by sucrose density gradient centrifugation and characterized by enzyme assays. The membrane fraction with the lowest density was enriched in plasma membrane marker enzymes. Cholesterol analysis showed that cells exposed to low-density lipoprotein had higher cholesterol-to-protein ratios in total cells, total cell membranes and individual membrane fractions than had the cells exposed to high-density lipoproteins. Cholesterol-to-phospholipid ratios of the plasma membrane-enriched fraction from cells exposed to low-density lipoprotein were higher than the same membrane fraction of cells exposed to high-density lipoprotein. Studies with iodinated lipoproteins showed that these compositional changes could not be due to lipoprotein contamination. Membrane microviscosity was determined by fluorescence depolarization with diphenylhexatriene and the microviscosity of the plasma membrane-enriched fraction was different in the cells exposed to the two different lipoprotein fractions. This difference in membrane microviscosity was significant only when the medium cholesterol content was 40 μg per ml or greater; cells exposed to low-density lipoprotein gave membranes with higher microviscosity.These results demonstrate that the properties of vascular smooth muscle cell membranes are influenced by exposure of the cells to homologous lipoprotein fractions.     

Extraction of plasma membranes from smooth muscle cells of the rabbit small intestine

A method is offered for isolation of subcellular fractions from small intestinal smooth muscle cells enriched by plasma membranes (PM). The method is based on differential centrifugation over sucrose density gradient. According to the localization of marker enzymes, the membrane fraction obtained with the use of 30% sucrose is considered to be optimal. The PM fraction is superior to the homogenate 10-fold on the average in the magnitude of Na, K-ATPase, 17-fold in Mg2+-ATPase, and 15-fold in that of 5'-nucleotidase activity. ATPase of PM is activated by Ca2+ in micro- and millimolar concentrations. It is suggested that Mg2+-dependent Ca-activated ATPase of PM is related to the Ca2+ content control in the cell.     

Isolation and characterization of plasma and smooth membranes of the marine diatom Nitzschia alba

A procedure is described for isolating plasma, smooth and other cellular membranes from hypotonically lysed protoplasts of the marine diatom, Nitzschia alba. From starting material of approximately 10 g wet weight (10¹⁰ cells), about 168 mg (organic weight) of a membrane-enriched fraction, exclusive of mitochondria, is obtained by differential centrifugation. From this, six membrane fractions are separated on a discontinuous sucrose gradient by isopycnic centrifugation.The plasma membranes, from the density region 1.23-1.29 g/cc, consist of small vesicles and sheets. They are purified approximately 20-fold, based on the increase in specific activity of a (Na⁺-K⁺-Mg²⁺)-ATPase, an enzyme found predominantly in these membranes. They also contain the highest specific and total activity of a (Mg²⁺)-ATPase and, in addition, are distinguished chemically by their high sterol specific content and high molar ratio of sterol/phospholipid (0.792-0.854). The carbohydrate/ protein ratio (0.070-0.072) is appreciably lower than that of the smooth membranes.The smooth membranes separate into two distinct fractions, a light and heavy component, which occur at the top of the sucrose gradient in densities of 1.13 and 1.18 g/cc, respectively. Both fractions are composed of relatively large membrane vesicles and membrane sheets and are distinguished from other membrane fractions by an exceptionally high carbohydrate/protein ratio (0.194-0.294).The light component shows the highest specific content of lipid, phospholipid, neutral lipid, carbohydrate, sialic acid, and RNA, and the highest specific activity of NADPH cytochrome c reductase, 5′-nucleotidase and phosphodiesterase compared to the other five fractions. It shows the lowest Na⁺ plus K⁺ stimulation of the (Mg²⁺)-ATPase. This fraction is probably enriched in endoplasmic reticulum.The heavy component contains some Golgi-like vesicles, sacs and tubules. It is characterized by the highest total content of chemical constituents analyzed, with the exception of RNA, and by the highest specific activity of thiamine pyrophosphatase, uridine diphosphatase, acid and alkaline phosphatase, and glucose-6-phosphatase, suggesting that this component is enriched in Golgi membranes approximately 13-fold.A most striking feature of these diatom membranes is the presence in all fractions of (Mg²⁺)-ATPase activity which is stimulated 5- to 10-fold by the presence of equimolar Na²⁺ plus K⁺. The data clearly differentiate these membrane fractions from each other as well as from membranes prepared from animal cells.