The isolation of microsomal subfractions of mouse plasmacytoma cells: The effect of salt concentration during nitrogen cavitation

Following disruption of MPC-11 cells by nitrogen cavitation the microsomes have been fractionated by centrifugation on discontinuous sucrose gradients. When the homogenization buffer contained 25 mm KCl three fractions were observed: smooth microsomes, light rough microsomes, and heavy rough microsomes. When it contained 100 mm KCl, however, only smooth and light microsomal fractions were found. Under the latter conditions the heavy rough microsomal vesicles were apparently not released as separate organelles but instead sedimented together with the endoplasmic reticulum which remains attached to the nuclei.     

Newly made phosphatidylserine and phosphatidylethanolamine are preferentially translocated between rat liver mitochondria and endoplasmic reticulum

The translocation of: (i) phosphatidylserine (PtdSer) from its site of synthesis on microsomal membranes to its site decarboxylation in mitochondrial membranes and (ii) phosphatidylethanolamine (PtdEtn) from the mitochondria to its site of methylation to phosphatidylcholine on microsomal membranes has been reconstituted in cell-free systems consisting of rat liver mitochondria and microsomes. Two types of systems have been reconstituted. In one, the translocation of newly made PtdSer or PtdEtn was examined by incubation of microsomes and mitochondria with [3-3H]serine. In the other, membranes were prelabeled with radioactive PtdSer or PtdEtn, and the transfer of these two lipids between mitochondria and microsomes was monitored. For the transfer of both PtdSer from microsomes to mitochondria and PtdEtn from mitochondria to microsomes, newly made phospholipids were translocated much more readily than pre-existing phospholipids. The data suggest that with respect to their translocation between these two organelles, the pools of newly synthesized PtdSer and PtdEtn were distinct from the pools of "older" phospholipids pre-existing in the membranes. Transfer of neither phospholipid in vitro depended on the presence of cytosolic proteins (i.e. soluble phospholipid transfer proteins) or on the hydrolysis of ATP, although there was some stimulation of PtdSer transfer by ATP and several other nucleoside mono-, di-, and triphosphates. The data are consistent with a collision-based mechanism in which the endoplasmic reticulum and mitochondria come into contact with one another, thereby effecting the transfer of phospholipids. The proposal that there is contact between the endoplasmic reticulum and mitochondria is supported by the recent isolation of a membrane fraction having many, but not all, of the properties of the endoplasmic reticulum, but which was isolated in association with mitochondria (Vance, J. E. (1990) J. Biol. Chem. 265, 7248-7256).     

A comparative study of the molecular structures of the plasma membranes and the smooth and the rough endoplasmic-reticulum membranes from rat liver

Plasma-membrane as well as smooth-, rough- and degranulated-endoplasmic-reticulum-membrane fractions were isolated from the microsomal pellet of rat liver. The purity of these fractions, as determined by marker-enzyme activities, electron microscopy, cholesterol content and RNA content, was found to be adequate for a comparative structural study. Major differences in lipid and protein composition were found to exist between the plasma membrane and the endoplasmic reticulum, but not between the smooth and the rough fractions of the endoplasmic reticulum. Differences in the location of membrane protein thiol groups and the mobility of the membrane phospholipids were observed between the plasma membranes and the endoplasmic reticulum, and these could be explained by differences in protein and lipid composition. However, by employing fluorescence and spin-labelling techniques structural changes were also observed between the smooth and the rough endoplasmic-reticulum fractions. These results suggest that the structural heterogeneity existing between the two latter membrane fractions occurs near or on their membrane surfaces and is not due to the greater number of ribosomes bound to the rough endoplasmic-reticulum fraction.     

Biogenesis of very-low-density lipoproteins in rat liver. Intracellular distribution of apolipoprotein B

The hepatic subcellular distribution of apolipoprotein B (apo B) was studied quantitatively by using an enzyme immunoassay developed for apo B and by immunoadsorption-precipitation of [3H]leucine-labelled apo B. Over 50% (of 0.59 microgram/mg protein) of the apo B was located in the microsomal fraction. Further subfractionation of the microsomes revealed that 47% of the microsomal apo B was in the Golgi apparatus, while another 43% was associated with the rough endoplasmic reticulum. The smooth endoplasmic reticulum accounted for only 4% of the total. When rat livers were labelled with [3H]leucine for 10 min, the rough endoplasmic reticulum accounted for 80% of the total immunoadsorbed precipitable apo B radioactivity while the smooth accounted for 20%, with no contribution from the Golgi. However, only 8.7% of the total radioactive immunoadsorbed precipitable apo B was lipoprotein-associated, the remainder being membrane-bound. Lipoprotein-associated apo B radioactivity in the smooth endoplasmic reticulum accounted for 40%, with the rough contribution attributed at 50% and the Golgi at 9%. We concluded that (a) there are two major pools of apo B in rat liver microsomes; (b) although the apo B mass may be negligible in the smooth endoplasmic reticulum, the latter does play a role in lipoprotein biogenesis. The possible function of apo B associated with membranes of the microsomes is also discussed.     

Intracellular transport of mouse kidney β-glucuronidase induced by gonadotrophin

1. The response of renal beta-glucuronidase with time to the injection of gonadotrophin was investigated in each submicrosomal fraction of rough and smooth microsomal fractions of mouse kidney homogenate. 2. The increase in beta-glucuronidase activity appeared initially in membranes of the rough microsomal fraction, 24h after injection. 3. Afterwards the newly synthesized enzyme appeared in the contents of the rough microsomal fraction and was subsequently found in the smooth microsomal fraction, reaching a maximum concentration in this fraction at 72h. 4. At this juncture, a decrease in the enzyme activity was observed in rough microsomal contents whereas the lysosomal fraction had reached its maximum value. 5. The time-course of the appearance of beta-glucuronidase in the submicrosomal fractions after the gonadotrophin stimulation suggests that the newly synthesized enzyme at the site of membrane-bound ribosomes is transferred across the membrane into cisternae of the rough endoplasmic reticulum, and then is transported into lysosomes via the smooth endoplasmic reticulum. 6. The properties of microsomal and lysosomal beta-glucuronidases were compared.     

Biogenesis of endoplasmic reticulum membranes. I. Structural and chemical differentiation in developing rat hepatocyte

The development of the endoplasmic reticulum of rat hepatocytes was studied during a period of rapid cell differentiation, i.e., from 3 days before to 8 days after birth. Before birth, the ER increases in volume, remaining predominantly rough surfaced; after birth, the increase continues but affects mainly the smooth-surfaced part of the system. These changes are reflected in variations of the RNA/protein and PLP/protein ratios of microsomal fractions: the first decreases, while the second increases, with age. The analysis of microsomal membranes and of microsomal lipids indicates that the PLP/protein ratio, the distribution of phospholipids, and the rate of P³² incorporation into these phospholipids show little variation over the period examined and are comparable to values found in adult liver. Fatty acid composition of total phosphatides undergoes, however, drastic changes after birth. During the period of rapid ER development in vivo incorporation of leucine-C¹⁴ and glycerol-C¹⁴ into the proteins and lipids of microsomal membranes is higher in the rough-than in the smooth-surfaced microsomes, for the first hours after the injection of the label; later on (~10 hr) the situation is reversed. These results strongly suggest that new membrane is synthesized in the rough ER and subsequently transferred to the smooth ER.     

The phospholipid composition of embryonic chick liver microsomes

1. The phospholipid composition of hepatic microsomal fractions from different developmental stages of embryonic chick was established. The major components were phosphatidylcholine (approx. 66%), phosphatidylethanolamine plus phosphatidylserine (approx. 21%) and sphingomyelin (approx. 9%). 2. There were no significant changes in the phospholipid composition during embryonic development from 9 to 20 days. 3. When microsomal subfractions were prepared it was found that the smooth-microsomal fractions (Ia and Ib) had a significantly greater sphingomyelin content than the rough-microsomal fraction (II). This was compensated by a lower phosphatidylcholine content in fractions Ia and Ib and an increase of phosphatidylcholine in fraction II. 4. The significance of the differences in the phospholipid composition of smooth and rough microsomes is discussed with particular reference to the origin and interrelation of smooth and rough endoplasmic reticulum.     

Can mitochondria and synaptosomes of guinea-pig brain synthesize phospholipids?

1. The use of ;marker' enzymes for investigating the contamination by endoplasmic reticulum of mitochondrial and synaptosomal (nerve-ending) fractions isolated from guinea-pig brain was examined. NADPH-cytochrome c reductase appeared to be satisfactory. With the synaptosomal preparation there was a non-occluded enzymic activity believed to arise from contaminating microsomes and an occluded form released by detergent, which probably was derived from some type of intraterminal smooth endoplasmic reticulum. 2. Isolated brain mitochondria, both intact and osmotically shocked, could not synthesize more labelled phosphatidylcholine from CDP-[Me-(14)C]choline or phosphoryl[Me-(14)C]choline than could be accounted for by microsomal contamination. They could synthesize only phosphatidic acid and diphosphatidylglycerol from a [(32)P]P(i) precursor and not nitrogen-containing phosphoglycerides or phosphatidylinositol. 3. The synaptosomal outer membrane and the intraterminal mitochondria could not synthesize phosphatidylcholine from CDP-[Me-(14)C]choline but the synaptic vesicles and probably the intraterminal ;endoplasmic reticulum' appeared to be capable of catalysing the incorporation of label from this substrate into their phospholipids. 4. Microsomal fractions and synaptosomes from guinea-pig brain could incorporate [Me-(14)C]choline into their phospholipids by a non-energy-requiring exchange process, which was catalysed by Ca(2+). Fractionation of the synaptosomes after such an exchange had taken place revealed that the label was predominantly in the intraterminal mitochondria and not associated with membranes containing NADPH-cytochrome c reductase. 5. On the intraperitoneal injection of [(32)P]P(i) into guinea pigs, incorporation of radioactivity into phosphatidylinositol and phosphatidic acid was much faster than into the nitrogen-containing phosphoglycerides. Mitochondria and microsomal fractions showed a roughly equivalent incorporation into individual phospholipids, and that into synaptosomes was appreciably less, whereas the phospholipids of myelin showed little (32)P incorporation up to 10h.     

Co-ordination between membrane phospholipid synthesis and accelerated biosynthesis of cytoplasmic ribonucleic acid and protein

1. The rate of synthesis of membrane phospholipid was studied in rat liver and seminal vesicles by following the incorporation of [(32)P]orthophosphate, [(14)C]choline and [(14)C]glycerol. Particular emphasis was laid on the endoplasmic reticulum, which was fractionated into smooth microsomal membranes, heavy rough membranes, light rough membranes and free polyribosomes. 2. Phospholipid labelling patterns suggested a heterogeneity in the synthesis and turnover of the different lipid moieties of smooth and rough endoplasmic membranes. The major phospholipids, phosphatidylcholine and phosphatidylethanolamine, were labelled relatively rapidly with (32)P over a short period of time whereas incorporation of radioisotope into the minor phospholipids, sphingomyelin, lysolecithin and phosphatidylinositol proceeded slowly but over a longer period of time. 3. The incorporation of orotic acid into RNA and labelled amino acids into protein of the four submicrosomal fractions was also studied. 4. Rapid growth of the liver was induced by the administration of growth hormone and tri-iodothyronine to hypophysectomized and thyroidectomized rats and by partial hepatectomy. Growth of seminal vesicles of castrated rats was stimulated with testosterone propionate. 5. The rate of labelling of membrane phospholipids was enhanced in all major subcellular particulate fractions (nuclear, mitochondrial and microsomal) during induced growth. However, it was in the rough endoplasmic reticulum that the accumulation of phospholipids, RNA and protein was most marked. The effect of hormone administration was also to accelerate preferentially the labelling with (32)P of sphingomyelin relative to that of phosphatidylcholine or phosphatidylethanolamine. 6. Time-course analyses showed that, in all four growth systems studied, the enhancement of the rate of membrane phospholipid synthesis coincided with the rather abrupt increase in the synthesis of RNA and protein of the rough endoplasmic reticulum. Growth hormone and tri-iodothyronine administered to hypophysectomized rats had additive effects in all the biosynthetic processes. The latent period of action of each hormone was maintained so that two waves of proliferation of endoplasmic reticulum occurred if the hormones were administered simultaneously. 7. It is concluded that there is some mechanism in the cell that tightly co-ordinates the formation of membranes, especially those of the endoplasmic reticulum, when an increased demand is made for protein synthesis.     

Properties of membrane-bound bilirubin UDP-glucuronyltransferase in rough and smooth endoplasmic reticulum and in the nuclear envelope from rat liver.

We examined regulatory properties of bilirubin UDP-glucuronyltransferase in sealed RER (rough endoplasmic reticulum)- and SER (smooth endoplasmic reticulum)-enriched microsomes (microsomal fractions), as well as in nuclear envelope from rat liver. Purity of membrane fractions was verified by electron microscopy and marker studies. Intactness of RER and SER vesicles was ascertained by a high degree of latency of the lumenal marker mannose-6-phosphatase. No major differences in the stimulation of UDP-glucuronyltransferase by detergent or by the presumed physiological activator, UDPGlcNAc, were observed between total microsomes and RER- or SER-enriched microsomes. Isolated nuclear envelopes were present as a partially disrupted membrane system, with approx. 50% loss of mannose-6-phosphatase latency. The nuclear transferase had lost its latency to a similar extent, and the enzyme failed to respond to UDPGlcNAc. Our results underscore the necessity to include data on the integrity of the membrane permeability barrier when reporting regulatory properties of UDP-glucuronyltransferase in different membrane preparations.     

Glycosyltransferase activities in Golgi complex and endoplasmic reticulum fractions isolated from African trypanosomes

Highly enriched Golgi complex and endoplasmic reticulum fractions were isolated from total microsomes obtained from Trypanosoma brucei, Trypanosoma congolense, and Trypanosoma vivax, and tested for glycosyltransferase activity. Purity of the fractions was assessed by electron microscopy as well as by biochemical analysis. The relative distribution of all the glycosyltransferases was remarkably similar for the three species of African trypanosomes studied. The Golgi complex fraction contained most of the galactosyltransferase activity followed by the smooth and rough endoplasmic reticulum fractions. The dolichol- dependent mannosyltransferase activities were highest for the rough endoplasmic reticulum, lower for the smooth endoplasmic reticulum, and lowest for the Golgi complex. Although the dolichol-independent form of N-acetylglucosaminyltransferase was essentially similar in all the fractions, the dolichol-dependent form of this enzyme was much higher in the endoplasmic reticulum fractions than in the Golgi complex fraction. Inhibition of this latter activity in the smooth endoplasmic reticulum fraction by tunicamycin A1 suggests that core glycosylation of the variable surface glycoprotein may occur in this organelle and not in the rough endoplasmic reticulum as previously assumed.     

A rapid method for the fractionation of isolated hepatocytes

The fractionation of rat liver hepatocytes using a mechanical disruption technique followed by centrifugation is reported; the whole procedure requires approximately 10 min. Marker enzyme distribution data are in good agreement with distribution data from standard techniques connected with the production of three subcellular fractions—cytoplasmic, mitochondrial, and microsomal. Electrophoretic analysis of the mitochondrial and microsomal fractions show total band correspondence between the fractions produced by the method and traditional techniques. Examination of the fractions by electron microscopy supports the view that the mitochondrial fraction is comprised of both intact mitochondria and mitochondria from which the outer membrane has been removed. The microsomal fraction contains discrete vesicles derived from both rough and smooth endoplasmic reticulum.     

Localization of proteoglycan core protein in subcellular fractions isolated from rat chondrosarcoma chondrocytes

Chondrocytes from the Swarm rat chondrosarcoma were pulse-labeled with [3H]serine for 30 min and chased, in the presence of cycloheximide, for times up to 300 min. The movement of newly synthesized core protein precursor of the proteoglycan through elements of the endoplasmic reticulum and Golgi complex was examined. Rough and smooth microsome fractions were obtained by centrifuging postmitochondrial supernatants from cell homogenates on discontinuous sucrose gradients. The core protein precursor was identified in subcellular fractions by (a) immunoprecipitation with an antiserum directed against the hyaluronate binding region of the core protein and the link protein and (b) its size on polyacrylamide gels. Labeled core protein precursor decreased from the microsomes with a t1/2 of 60 +/- 8 min, nearly the same as for the appearance of label in completed proteoglycan monomer (t1/2 = 58 +/- 13 min), consistent with a precursor-product relationship. After correcting for incomplete recovery of the core protein precursor in the microsomal fractions and for cross-contamination of the smooth microsomes by elements of rough endoplasmic reticulum, the redistribution of core protein precursor and completed proteoglycan in the intracellular compartments and of labeled extracellular proteoglycan were fit to a three-compartment model. A t1/2 of 98 +/- 7 min for the loss of core protein precursor from the rough microsomes and a t1/2 = 10 +/- 4 min for the completed proteoglycan in the intracellular compartment (Golgi and secretory vesicles) was obtained. The data indicate that at least 70% of the intracellular transit time for the core protein precursor is spent in the rough endoplasmic reticulum. The addition of glycosaminoglycan chains followed by secretion from the cell occurs relatively rapidly, occupying less than 30% of the total intracellular dwell time.     

The localization of cholinephosphotransferase in the outer membrane of guinea-pig lung mitochondria

The activity of cholinephosphotransferase was measured in the subcellular fractions of guinea-pig lung. The specific activity of the enzyme was highest in a fraction, intermediate in density between mitochondria and microsomes. Similar subcellular distribution patterns were observed for both cholinephosphotransferase and rotenone-insensitive NADH-cytochrome c reductase, an enzyme associated with the outer membrane of mitochondria and endoplasmic reticulum, suggesting that cholinephosphotransferase may be localized in both of these organelles. The distribution of cholinephosphotransferase activity in the subfractions of mitochondria and the intermediate fractions recovered by linear density gradient paralleled that of the mitochondrial outer membrane marker enzyme, monoamine oxidase. RNA content of a subfraction enriched in cholinephosphotransferase and monoamine oxidase was not typical to that of either rough or smooth endoplasmic reticulum. The results of this study suggest that in guinea-pig lung, cholinephosphotransferase is localized in both the outer membrane of mitochondria, and the endoplasmic reticulum.     

Preparation of liver microsomes with high recovery of endoplasmic reticulum and a low grade of contamination

A modified procedure for preparing the microsomal fraction from rat liver was developed with the aim of increasing the recovery without increasing the degree of contamination. 87% of the membranes of the microsomal fraction isolated from the first mitochondrial (10 000 X g) supernatant originates from the endoplasmic reticulum, representing a 35% yield. By gentle resuspension of the 10 000 X g pellet followed by differential centrifugation a second crop of microsomes can be prepared which, together with the first crop, gives a 55% total recovery of microsomal markers. 87% of the protein in this second crop also originates from the endoplasmic reticulum and this fraction has properties similar to those of the first crop. Contaminating membranes include Golgi membranes (0.6% of the total protein), mitochondria (2.5%), lysosomes (5%) and plasma membranes (5%). Collecting further crops increases the contamination. Subfractionation studies revealed almost identical distributions of ribosome-rich, ribosome-poor and smooth membranes in the two crops of microsomal fractions. The results obtained after treatment of the animals with phenobarbital or methylcholantrene were similar to those obtained with control animals; but in the case of methylcholantrene treatment the second crop represents a larger portion of the total membranes of the endoplasmic reticulum.     

Studies on the latency of UDP-galactose-asialo-mucin galactosyltransferase activity in microsomal and Golgi subfractions from rat liver

The activity of UDPgalactose-asialo-mucin galactosyltransferase (EC 2.4.1.74) in microsomal and Golig subfractions was stimulated 2.4-fold after disruption of the membrane permeability barrier by hypotonic incubation. In the presence of Triton X-100, galactose transfer to asialo-mucin was increased 12-fold in rough microsomes and 5-fold in smooth microsomes both with and without hypotonic incubation; while in the Golgi subfractions no stimulation by detergent was observed. These experiments indicate differences in enzyme-lipid or enzyme-protein interactions in microsomes and Golgi membranes. Furthermore, these results strongly support the conclusion that the UDP-galactose-asialo-mucin galactosyltransferase activity in microsomal fractions is not due to contamination by Golgi vesicles but represents an enzyme activity endogenous to the endoplasmic reticulum.     

Biogenesis of endoplasmic reticulum membrane in rat liver cells. II. Discharge of the nascent peptides of NADPH-cytochrome c reductase and cytochrome b5 on the cytoplasmic side of the endoplasmic reticulum membrane.

The direction of discharge of the nascent peptides of NADPH-cytochrome c reductase and cytochrome b5 from bound polyribosomes of rough microsomes was investigated in order to elucidate the mechanism of separation of these membrane proteins from secretory proteins, which are also synthesized by the same class of ribosomes of rough endoplasmic reticulum. The nascent peptides of NADPH-cytochrome c reductase and cytochrome b5 in intact rough microsomes were accessible to externally added 125I-Fab's against these proteins, and were susceptible to trypsin digestion, whereas the nascent peptides of serum albumin were not. The nascent peptides of these two microsomal proteins were released into the cytoplasm by puromycin treatment of intact rough microsomes, while the nascent peptides of serum albumin were retained in the microsomal lumen. These observations suggest that the nascent peptides of microsomal proteins, which are present on the cytoplasmic surface of the endoplasmic reticulum membrane, are exposed on the surface of microsomal vesicles, while those of secretory proteins are enclosed inside the vesicles. Therefore, the topographical separation of microsomal membrane proteins from secretory proteins is accomplished at the step of their synthesis by the bound polyribosomes of rough endoplasmic reticulum.     

Phosphatidylglycerol in lung surfactant. II. Subcellular distribution and mechanism of biosynthesis in vitro.

Lamellar inclusion bodies, apparent precursors for alveolar surfactant lining, have remarkably similar phospholipid composition to surfactant from alveolar lavage, but distinctly different from other fractions studied: mitochondria, microsomal fraction containing endoplasmic reticulum membranes, plasma membranes and nuclei. Surfactant contained (as % of total phospholipid phosphate): 75.5-77.0% lecithin, 11.0-11.2% phosphatidylglycerol, 4.2-4.6% phosphatidylethanolamine, 3.0-3.2% phosphatidylinositol, 1.5-1.7% bis-(monoacylglycerol) phosphate, 1.2-1.9% phosphatidylserine, and 0.7-1.5% sphingomyelin. Fatty acids of phosphatidylglycerol from lamellar bodies were similar to those from microsomes but different from those in mitochondria. Lung homogenate in continuous sucrose density gradient displayed two major activity peaks of phosphatidylglycerol synthesis: the heavier from mitochondria; the lighter from endoplasmic reticulum. Studies on mechanism of phosphatidylglycerol synthesis in vitro revealed (in these two fractions) CDP-diglyceride and sn-glycerol phosphate precursors to phosphatidylglycerol phosphate, that hydrolysed to phosphatidylglycerol. In microsomes disaturated CDP-diglycerides were 1.6-1.9 times more active substrates than in mitochondria, whereas CDP-diglycerides from egg lecithin were almost equally active. In contrast to lung mitochondria no cardiolipin synthesis was detected in microsomes. The highest specific activities for phosphatidate cytidyltransferase, CDP-diglyceride-inositol phosphatidyltransferase, choline phosphotransferase, and phosphatidylethanolamine methyltransferase were all found in microsomes. The present in vitro studies and additional evidence (M. Hallman and L. Gluck, (1975) Fed. Proc. 34, 274) support the hypothesis that de novo synthesis of surfactant lecithin phosphatidylinositol and phosphatidylglycerol takes place in the endoplasmic reticulum of alveolar cells.     

Fatty acid synthesis in vivo. Transfer of lipids from microsomes to other subcellular fractions of rat liver

At various times after the intraperitoneal injection of Na acetate-1-C¹⁴ to male Wistar rats, the labelled fatty acids are nonuniformely distributed among the lipids of liver microsomes, mitochondria and cell sap. The changes observed in the specific radioactivity of the neutral and phospholipids support the hypothesis that a transfer of these lipids takes place from the site of synthesis (endoplasmic reticulum) to mitochondria and cell sap. This phenomenon is probably responsible for the decline of microsomal fatty acids in favour of the mitochondrial and soluble fractions. In this connection, the deacylation-reacylation process does not seem to be involved.     

Biogenesis of the mitochondrial matrix enzyme, glutamate dehydrogenase, in rat liver cells. II. Significance of binding of glutamate dehydrogenase to microsomal membrane

1. Glutamate dehydrogenase and malate dehydrogenase solubilized from liver microsomes were able to rebind to microsomal vesicles while the corresponding dehydrogenases extracted from mitochondria showed no affinity for microsomes. 2. Competition was noticed between microsomal glutamate dehydrogenase and microsomal malate dehydrogenase in the binding to microsomal membranes. Mitochondrial malate dehydrogenase or bovine serum albumin did not inhibit the binding of microsomal glutamate dehydrogenase to microsomes. 3. Binding of microsomal glutamate dehydrogenase to microsomal membranes decreased when microsomes was preincubated with trypsin. 4. Rough microsomal glutamate dehydrogenase was more efficiently bound to rough microsomes than smooth microsomes. Conversely, smooth microsomal glutamate dehydrogenase had higher affinity for smooth microsomes than for rough microsomes. 5. A difference was noticed among the glutamate dehydrogenase isolated from rough and smooth microsomes, and from mitochondria, which suggested the possibility of minor post-translational modification of enzyme molecules in the transport from the site of synthesis to mitochondria.