Gonadotropin and prostaglandins binding sites in rough endoplasmic reticulum and Golgi fractions of bovine corpora lutea.

Highly purified rough endoplasmic reticulum and three subfractions of golgi were prepared from 105,000g pellet of the homogenate by centrifugation in floatation and sedimentation discontinuous sucrose gradients. Highly purified plasma membranes were also prepared from 9,000g pellet of the same homogenates for assessment under the same experimental conditions. Although 5′-nucleotidase, a marker for plasma membranes, was markedly enriched in plasma membranes, very little or none of this enzyme activity was found in other fractions. Very little or no NADH cytochrome c reductase activity, a marker for rough endoplasmic reticulum, was found in fractions other than rough endoplasmic reticulum. Galactosyl transferase, a marker for golgi, was found and enriched in all the fractions; however, enrichment in golgi fractions was higher than in other fractions. Very little or no lysosomal marker activity, i.e., acid phosphatase, was found in rough endoplasmic reticulum or golgi fractions as compared to lysosomes. These marker enzyme data suggested that rough endoplasmic reticulum and golgi fractions were relatively pure with little or no cross contamination with other organelles. The [¹²⁵I]human choriogonadotropin ([¹²⁵I]hCG), [³H]prostaglandin (PG)E₁, and [³H]PGF_2a specifically bound to rough endoplasmic reticulum and golgi fractions in addition to plasma membranes. The enrichments of binding in the former two fractions, in some cases, were as high as plasma membranes itself. The specific binding of some of the ligands was found to be partially latent in rough endoplasmic reticulum and golgi fractions but not in plasma membranes. Marker enzyme data, ratio between bindings and marker enzyme activities (an index of organelle contamination), and partial latency of binding suggest that rough endoplasmic reticulum and golgi fractions intrinsically contain gonadotropin and PGs binding sites.     

Receptors for gonadotropins and prostaglandins in lysosomes of bovine corpora lutea.

The total mitochondrial fraction of bovine corpus luteum specifically bound [³H]prostaglandin (PG) E₁, [³H] PGF_2α, and ¹²⁵I-labeled human lutropin (hLH) despite very little 5′-nucleotidase activity, a marker for plasma membranes. Since the total mitochondrial fraction isolated by conventional centrifugation techniques contains both mitochondria and lysosomes, it was subfractionated into mitochondria and lysosomes to ascertain the relative contribution of these fractions to the binding. Subfractionation resulted in an enrichment of cytochrome c oxidase (a marker for mitochondria) in mitochondria and of acid phosphatase (a marker for lysosomes) in lysosomes. The lysosomes exhibited little or no contamination with Golgi vesicles, rough endoplasmic reticulum, or peroxisomes as assessed by their appropriate marker enzymes. Subfractionation also re ulted in [³H] PGE₁, [³H] PGF_2α, and ¹²⁵I-labeled hLH binding enrichment with respect to homogenate in lysosomes but not in mitochondria. The lysosomal binding enrichment and recovery were, however, lower than in plasma membranes. The ratios of marker enzyme to binding, an index of organelle contamination, revealed that plasma membrane and lysosomal receptors were intrinsic to these organelles. Freezing and thawing had markedly increased lysosomal binding but had no effect on plasma membrane binding. Exposure to 0.05% Triton X-100 resulted in a greater loss of plasma membrane compared to lysosomal binding. In summary, the above results suggest that lysosomes, but not mitochondria, in addition to plasma membranes, intrinsically contain receptors for PGs and gonadotropins. Furthermore, lysosomes overall contain a greater number of PGs and gonadotropin receptors compared to plasma membranes and these receptors are associated with the membrane but not the contents of lysosomes.     

Interactions of liver encoplasmic reticulum membranes and polysomes in vitro.

The interactions of various preparations of endoplasmic reticulum membranes and polysomes have been studied by means of a sandwich sucrose gradient that clearly isolates free ribosomes, smooth endoplasmic reticulum (S.E.R.) and rough endoplasmic reticulum (R.E.R.) from the microsomal fraction of rat liver homogenates. Reconstructed rough membranes separate well from the native R.E.R. but occupy the same position along the gradient as the S.E.R. and the rough membranes, stripped of their ribosomes by means of LiCl. Native R.E.R. and S.E.R. do not bind any added labeled polysomes at 0 degree C; previous treatment with LiCl does not modify the behavior of S.E.R. The presence of cell sap during the binding reaction does not increase polysome fixation by stripped-rough membranes but protects in some way the polysomes and preserves all their original functional capacity of amino acid incorporation into protein.     

Interactions of liver endoplasmic reticulum membranes and polysomesin vitro

Summary The interactions of various preparations of endoplasmic reticulum membranes and polysomes have been studied by means of a sandwich sucrose gradient that clearly isolates free ribosomes, smooth endoplasmic reticulum (S.E.R.) and rough endoplasmic reticulum (R.E.R.) from the microsomal fraction of rat liver homogenates. Reconstructed rough membranes separate well from the native R.E.R. but occupy the same position along the gradients as the S.E.R. and the rough membranes, stripped of their ribosomes by means of LiCl. Native R.E.R. and S.E.R. do not bind any added labeled polysomes at 0°C; previous treatment with LiCl does not modify the behavior of S.E>R. The presence of cell sap during the binding reaction does not increase polysome fixation by stripped-rough membranes but protects in some way the polysomes and preserves all their original functional capacity of amino acid incorporation into protein.     

Phosphomannosyl-enzyme receptors in rat liver. Subcellular distribution and role in intracellular transport of lysosomal enzymes

beta-Hexosaminidase B purified from human fibroblast secretions was used as a ligand to study phosphomannosyl-enzyme receptors in membranes from rat tissues. Enzyme binding to rat liver membranes was saturable, competitively inhibited by mannose 6-phosphate, not dependent on calcium, and destroyed by prior treatment of the hexosaminidase with either alkaline phosphatase or endoglycosidase H. Most (90%) of the phosphomannosyl-enzyme receptors were found in endoplasmic reticulum, Golgi apparatus, and lysosomes; 9.5% in the plasma membrane, and less than 1% in nuclei and mitochondria. Receptors were vesicle-enclosed in all fractions except plasma membrane. Receptors in the endoplasmic reticulum apparently were occupied by endogenous ligands, but most receptors in lysosomes and plasma membrane were unoccupied. Most of the endogenous beta-hexosaminidase was in lysosomes and was released from vesicles by detergent treatment. Displacement of the residual receptor-bound endogenous beta-hexosaminidase (mostly in endoplasmic reticulum and Golgi apparatus) from detergent-treated membranes by mannose 6-phosphate released high uptake enzyme with properties expected for phosphomannosyl-enzymes. Mannose 6-phosphate-inhibitable enzyme receptor activity was found in nine rat organs and correlated roughly with their lysosomal enzyme content. These data support a general model for lysosomal enzyme transport in which the phosphomannosyl-enzyme receptor acts as a vehicle for delivery of newly synthesized acid hydrolases from the endoplasmic reticulum to lysosomes.     

Characterization of gonadotropin binding sites in the intracellular organelles of bovine corpora lutea and comparison with plasma membrane sites

The specific binding of 125I-human choriogonadotropin (hCG) to plasma membranes, nuclear membranes, lysosomes, rough endoplasmic reticulum, heavy golgi, and medium and light golgi of bovine corpora lutea was dependent on the amount of protein, 125I-hCG concentration and incubation time. The bound hormone in all the organelles was able to rebind to fresh corresponding organelles. Scatchard analysis revealed a homogenous population of gonadotropin binding sites in plasma membrane, rough endoplasmic reticulum, heavy golgi, and medium and light golgi, whose binding affinities (Kd = 8.6-11.0 X 10(-11) M) were similar but whose number of available gonadotropin binding sites varied. Scatchard analyses of nuclear membranes and lysosome binding, on the other hand, were heterogenous (Nuclear membranes, 11 and 23 X 10(-11) M lysosomes, 3.4 and 130 X 10(-11) M). The rate constants for association (5.9 to 12.1 X 10(6) M-1 S-1) and dissociation (7.4 to 9.0 X 10(-4) S-1) were similar among different subcellular organelles except for nuclear membranes and lysosomes, where rate constants for association were significantly lower. The ligand binding specificity, lower effectiveness of human luteinizing hormone as compared to hCG in competition, the optimal pH, the lack of ionic requirements for binding, and the molecular size of 125I-hCG-gonadotropin binding site complexes solubilized from various intracellular organelles were similar to those observed for plasma membranes. Numerous differences were also observed between intracellular organelles and plasma membranes as well as among intracellular organelles themselves with respect to binding losses due to exposure to low and high pH values, di- and monovalent cations, increasing preincubation temperatures, and a variety of enzymes and protein reagents. The possible reasons for these similarities as well as differences observed are discussed. The differences are viewed as an additional indication that contamination cannot solely explain the presence of gonadotropin binding sites in various intracellular organelles.     

Rapid formation of inositol 1,4,5-trisphosphate in rat pancreatic acini stimulated by carbamylcholine

Cholinergic stimulation of inositol phosphate formation was studied in isolated rat pancreatic acini, prelabelled with myo-[2-3H]inositol. Carbamylcholine increased incorporation of radioactivity into Ins(1,4,5)P3 and InsP4 within 5 s. Increases in [3H]Ins(1,3,4)P3 were delayed with marked stimulation occurring between 10 s and 1 min. Inositol polyphosphate formation was less sensitive to carbamylcholine concentration than was stimulation of amylase release. At a low (0.3 microM) carbamylcholine concentration, no increase in inositol polyphosphate formation was detected, whereas stimulation of amylase release, which was not dependent on extracellular calcium, was observed. Ins(1,4,5)P3 was shown to release actively accumulated 45Ca2+ from isolated rough endoplasmic reticulum membranes to a similar extent as that released from rough endoplasmic reticulum following cholinergic stimulation of pancreatic acini (Richardson, A.E. et al. (1984) Biochem. Soc. Trans. 12, 1066-1067). The data is consistent with Ins(1,4,5)P3 being produced rapidly enough to release sufficient calcium from the rough endoplasmic reticulum to cause an observed increases in cytoplasmic free Ca2+.     

Induction of high-density-lipoprotein receptors in rat corpus luteum by human choriogonadotropin. Evidence of protein synthesis de novo.

The present studies investigated the specific binding of 125I-labelled high-density lipoprotein (125I-HDL) to plasma membranes. Golgi, rough endoplasmic reticulum and mitochondria/lysosomes, prepared from ovaries of rats injected with human choriogonadotropin (hCG) or 0.9% NaCl. Treatment in vivo with hCG resulted in 2-3-fold induction of 125I-HDL binding activity in all the subcellular organelles. The specific binding of HDL to various subcellular organelles was dependent on the amount of protein, lipoprotein concentration and incubation time. Equilibrium-binding studies revealed comparable Kd values (13-22 micrograms of HDL protein/ml) for HDL binding in all the subcellular organelles tested. Treatment with cycloheximide (2.0 mg/kg body wt.) before hCG administration abolished the induction of HDL receptors, suggesting the involvement of a protein-synthesis-dependent process in receptor induction. Analysis of equilibrium dissociation constants (Kd) for 125I-HDL binding in membranes from hCG-, cycloheximide-and saline-treated animals suggests that the increase in binding was due to an increase in the number of binding sites rather than a change in the affinity. Additionally, pretreatment with tunicamycin, an inhibitor of N-linked glycosylation, had no effect on hCG-mediated receptor induction, suggesting that glycosylation of the receptor may not be necessary for the interaction of HDL with its receptors.     

Effect of ribosome stripping procedures on antigenicity and conformation of endoplasmic reticulum membrane.

The effects of 3 different procedures for stripping ribosomes from membranes on theantigeniticity and conformation of isolated rough and smooth endoplasmic reticulum from rat liver were examined by microcomplement fixation and circular dichroism. Some of the blocked antigenic binding sites in rough endoplasmic reticulum became available after stripping of ribosomes. None of the 3 methods used is capable of stripping ribosomes completely from rough endoplasmic reticulum without the concomitant removal of protein from the membrane. Such loss of membrane protein by the stripping treatments is probably involved in the observed changes in rough endoplasmic reticulum, since a marked reduction in complement fixing capacity and in ellipticity of circular dichroism is observed also in smooth endoplasmic reticulum after similar treatments.     

Association of poliovirus proteins with the endoplasmic reticulum.

Poliovirus proteins, except P3-7c, are associated with the endoplasmic reticulum after extraction of the cytoplasm and centrifugation of membranes to equilibrium in sucrose gradients. Proteins P3-2, P2-X, and P3-9 are found preferentially among the rough endoplasmic reticulum, whereas P3-7c is located in smooth endoplasmic reticulum fractions. P3-7c is probably not membrane associated, since it can be separated from membranes after centrifugation in buffer. However, P3-4a, P2-5b, P2-X, and P3-9 are avidly bound to membranes and cannot be dislodged with high-ionic-strength buffer containing EDTA or 4 M urea. These proteins are digested by trypsin, indicating peripheral rather than internal localization.     

Association of androgen binding sites with the endoplasmic reticulum of rat ventral prostate

Microsomes from ventral prostate of 24-h castrated rats contain a single set of tissue-specific high-affinity, low-capacity androgen binding sites. These sites are indigenous to the endoplasmic reticulum, as shown by purification procedures associated with marker enzymes and electron microscopic analyses. When prostatic microsomal membranes are separated from plasma membranes using the nuclear or the mitochondrial pellets as the source of fractionation in sucrose gradients, the androgen binding activity is selectively associated with fractions rich in rough endoplasmic reticulum and ribosomes. Eighty-four percent of the total content of Na+/K+ adenosine triphosphatase (ATPase) and only 27% of the total binding capacity were concentrated in fractions rich in smooth-surfaced vesicular membranes, when nuclear suspensions constituted the membrane source. In contrast, the region of the same gradient when enriched in rough endoplasmic reticulum and deficient in plasma membrane content contained 73% of the androgen-binding capacity and only 14% of the ATPase. For fractions collected using mitochondrial suspensions as starting material, the ratio (total glucose-6-phosphatase/total binding capacity) was closer to 1.0 than similar ratios of ATPase/binding capacity, indicating co-sedimentation of binding sites with microsomal membranes and not with plasma membranes. Na+/K+ ATPase, but not 5' nucleotidase, is a valid plasma membrane marker for ventral prostate. Microsomal androgen receptors may constitute a new level of regulation of androgen action in target cells.     

Distribution of insulin receptors among mouse liver endomembranes

Specific binding of insulin to highly purified preparations of rough endoplasmic reticulum, Golgi apparatus, and plasma membrane of mouse liver was determined. ¹²⁵I-labeled insulin bound maximally to the plasma membrane in radio-receptor assays. Golgi apparatus fractions exhibited binding 10–20% that of plasma membrane and rough endoplasmic reticulum exhibited only 1–2% of plasma membrane binding. Binding was proportional to membrane concentration and dose vs. response curves were very similar for the different fractions. Scatchard analysis of the insulin binding data for the plasma membrane and Golgi apparatus fractions showed curvilinear plots yielding similar apparent binding affinities (0.9 and 3.0 · 10⁸ M^?1, respectively). Purity of the isolated endomembranes was analyzed by morphometry and (Na⁺ + K⁺ + Mg²⁺)-ATPase and these preparations displayed less than 1% contamination by plasma membrane. These findings provide important confirmation of the presence of insulin receptors in Golgi apparatus membranes comparable to those located on the plasma membrane. Finally, the present study did not allow us to verify the existence of insulin receptors in the endoplasmic reticulum.     

Polyribosomal attachment to rat liver and hepatoma endoplasmic reticulum in vitro. A method for its study

A system for study and measurement of the attachment in vitro of exogenous polyribosomes to membranes has been presented. Its main features are use of low temperature, post-microsomal supernatant, pyrophosphate and citric acid to remove ribosomes from the surface of rough endoplasmic reticulum, and a method for quantitative separation of unattached from membrane-associated polyribosomes. The following were found. (1) Rough endoplasmic reticulum, from which ribosomes had been removed by treatment with pyrophosphate and citrate, bound over 50% of added polyribosomes, whereas the untreated (or control) rough and smooth endoplasmic reticulum and the smooth endoplasmic reticulum treated with pyrophosphate-citrate did not bind polyribosomes. (2) The polyribosome-binding capacity of rough endoplasmic reticulum stripped of its ribosomes decayed upon storage of the membranes at 0-4 degrees C. The half-life of this decay was about 6 days whereas that of the polyribosome-binding capacity of hepatoma stripped rough endoplasmic reticulum was about 1.5 days. (3) Preparations of stripped rough endoplasmic reticulum after reassociation with polyribosomes in vitro were quite similar to preparations of native rough endoplasmic reticulum as viewed with the electron microscope. Evidence is presented to support the contention that association of polyribosomes with membranes was the result of polyribosomal reattachment to the membranes rather than trapping of the polyribosomes between vesicles of the membranes.     

On the ultrastructure of granulosa lutein cells in porcine corpus luteum

Summary In young corpora lutea the endoplasmic reticulum membranes are sparse. A marked increase of smooth membranes then follows up to the peak of dioestrus. Continuities between smooth and rough endoplasmic reticulum are obvious during the same period. These observations suggest that the agranular membranes develop from the granular ones.During the most intense development of the endoplasmic reticulum the membranes show a tendency to be arranged in whorls. Since these are numerous only during the period of high progesterone secretion, a multitude of whorls constitutes a useful morphologic sign of high functional activity in the porcine granulosa lutein cells.During the first half of the oestrous cycle the increase in endoplasmic reticulum in general also parallels the increase in progesterone secretion. However, this secretion as well as ⁵-3-hydroxysteroid dehydrogenase activity declines earlier and more rapidly than the endoplasmic reticulum regresses. Steroid hormone synthesis may therefore be lacking although the agranular membranes appear morphologically normal.The mechanisms of induction of the endoplasmic reticulum membranes and enzymes active in steroid synthesis are discussed and it is suggested that luteinizing hormone (LH) may act as a trigger by increasing transport across membranes.Read at the Meeting of the Swedish Society for Pathology in Umeå, September 25, 1965 (Bjersing, 1966).This investigation was supported by grants from the Swedish Medical Research Council (Projects No. 13 X-78-01, 12 X-78-02, and 12 X-78-03).     

Characterization of gastric mucosal prostaglandin E2 receptor

1. The binding characteristics of gastric mucosal prostaglandin (PG) E2 (PGE2) receptor were investigated using mucosal cell membranes from rat stomach. The binding was found to be dependent upon PGE2 and membrane protein concentration, the time of incubation and the pH of the mixture, being highest at pH 3.0. 2. Scatchard analysis of the binding data revealed a curvilinear plot with high affinity binding (Kd = 2 nM; Bmax = 0.106 pmol/mg protein) and low affinity binding (Kd = 319 nM; Bmax = 2.262 pmol/mg protein) sites. 3. Competitive displacement study indicated that the receptor was specific for PGs of the E series, as PGF2 alpha and 6-keto-PGF1 alpha failed to displace the PGE2. 4. The study is the first report to provide biochemical parameters of specific PGE receptors in rat gastric mucosa.     

Subcellular distribution of the mannan-binding protein and its endogenous inhibitors in rat liver

The subcellular distribution of the mannan-binding protein from rat liver, a lectin specific for mannose and N-acetylglucosamine, was studied. Approximately 75% of the binding activity of the homogenate was recovered in microsomes, approximately 76% of which was accounted for by rough microsomes. Rough microsomes had the highest specific activity of binding, followed by the Golgi apparatus and smooth microsomes, whereas plasma membranes, lysosomes, mitochondria, and the soluble fraction had little or no binding activity. A topographical survey indicated that the binding protein was localized exclusively on the cisternal surface of microsomal vesicles. Thus, the binding protein of microsomal vesicles was protected from protease digestion and was released from the vesicles by mild detergent treatment. Competitive inhibitors, which presumably represent endogenous ligands of the binding protein, were found among subcellular fractions. More than 50% of the inhibitory activity of the homogenate was recovered in rough microsomes, while the highest specific activity of inhibition was found in lysosomes. The K_i values estimated for rough microsomes and lysosomes were 25.9 and 8.67 μg/ml, respectively. The distribution profiles of inhibitors were correlated roughly with those of the binding protein, resulting in masking of the binding activity in organelles up to the level of 86%. On the basis of the known localization and topology of the binding protein and endogenous inhibitors (ligands), possible physiological functions of the binding protein relevant to the transport of biosynthetic intermediates of glycoproteins from the rough endoplasmic reticulum to the Golgi apparatus and from the Golgi apparatus to lysosomes were discussed.     

Intracellular distribution of Sindbis virus membrane proteins in BHK-21 cells infected with wild-type virus and maturation-defective mutants.

The association of Sindbis virus proteins with cellular membranes during virus maturation was examined by utilizing a technique for fractionating the membranes of BHK-21 cells into three subcellular classes, which were enriched for rough endoplasmic reticulum, smooth endoplasmic reticulum, and plasma membrane. Pulse-chase experiments with wild-type (strain SVHR) virus-infected cells showed that virus envelope proteins were incorporated initially into membranes of the rough endoplasmic reticulum and subsequently migrated to the smooth and plasma membrane fractions. Large amounts of capsid protein were associated with the plasma membrane fraction even at the earliest times postpulse, and relatively little was found associated with the other membranes, suggesting a rapid and preferential association of nucleocapsids with the plasma membrane. We also examined the intracellular processing of the proteins of two temperature-sensitive Sindbis virus mutants in pulse-chase experiments at the nonpermissive temperature. Labeled virus proteins of mutant ts-20 (complementation group E) first appeared in the rough endoplasmic reticulum and were then transported to the smooth and plasma membrane fractions, as in wild-type (strain SVHR) virus-infected cells. In cells infected with ts-23 (complementation group D), the pulse-labeled virus proteins appeared initially in the rough membrane fraction and were transported to the smooth membrane fraction, but only limited amounts reached the plasma membrane. Thus, in ts-23-infected cells, the transport of the virus-encoded proteins from the smooth membranes seemed to be defective. In both ts-20- and ts-23-infected cells the envelope precursor polypeptide PE2 was not processed to E2, and no label was incorporated into free virus at the nonpermissive temperature.     

ELECTRON MICROSCOPIC RADIOAUTOGRAPHIC DETECTION OF SITES OF PROTEIN SYNTHESIS AND MIGRATION IN LIVER

The sites of synthesis of proteins and their subsequent migration in rat liver have been studied during a 75 min period after labeling of liver-slice proteins by exposure to leucine-H³ for 2 min. Incorporation of the label into protein began after 1 min and was maximal by 4 min. Electron microscopic radioautography showed that synthesis of proteins in hepatocytes occurs mainly on ribosomes, particularly those in rough endoplasmic reticulum and, to some extent, in nuclei and mitochondria. Most of the newly formed proteins leave the endoplasmic reticulum in the course of 40 min, and concurrently labeled proteins appear in Golgi bodies, smooth membranes, microbodies, and lysosomes. A likely pathway for the secretion of some or all plasma proteins is from typical rough endoplasmic reticulum to a zone of reticulum which is partially coated with ribosomes, to the Golgi apparatus, and thence to the cell periphery. The formation of protein by reticuloendothelial cells was measured and found to be about 5% of the total protein formed by the liver.     

Purification and functional characterization of membranes derived from the rough endoplasmic reticulum of Saccharomyces cerevisiae

Isolation and biochemical analysis of the components involved in protein translocation into the rough endoplasmic reticulum (ER) requires starting material highly enriched in membranes derived from this organelle. We have chosen to study the yeast Saccharomyces cerevisiae in order to profit from the ease of genetic manipulation. To date, however, no efficient scheme has been devised that allows the purification of functional rough ER-derived membranes from yeast, largely because proteins have yet to be identified that are rough ER-specific. In the experiments described here, we expressed the human rough ER marker ribophorin I to facilitate the analysis of subcellular fractionation. We found that the endoplasmic reticulum of yeast could be separated into two distinct domains by fractionation on continuous sucrose gradients. This procedure revealed a bimodal distribution of ER markers. The yeast homologue of the heavy chain-binding protein, BiP (encoded by the KAR2 gene), and the product of the SEC62 gene were present in two fractions having equilibrium densities of 1.146 and 1.192 g/ml, respectively. In contrast, our analysis showed that preprotein translocation activity and retention of the rough ER-specific protein ribophorin I were specific only to the membrane fraction with an equilibrium density of 1.192 g/ml. To prepare fractions highly enriched in translocation competent rough ER-derived membranes for analysis, we developed a density shift fractionation scheme that optimizes the purity of membranes containing human ribophorin I. Membranes obtained by this method were found to possess the majority of the appropriate functional markers, including ATP-independent preprotein binding, ribosome binding, and post-translational translocation. Mitochondria, the major contaminant of the 1.192 g/ml fraction, were significantly depleted in density-shifted membrane populations.     

EFFECTS OF ARGININE DEPRIVATION, ULTRAVIOLET RADIATION, AND X-RADIATION ON CULTURED KB CELLS : A Cytochemical and Ultrastructural Study

Cultured KB cells (derived from a human oral carcinoma) grown in monolayers were injured by one of three agents: starvation by arginine deprivation or treatment with high doses of either ultraviolet radiation or x-radiation. The different agents produced changes in nucleolar structure and varying accumulations of triglyceride and glycogen. All three agents produced an increase in number and size of lysosomes. These were studied in acid phosphatase preparations, viewed by both light and electron microscopy, and, occasionally, in vital dye, esterase, and aryl sulfatase preparations. Ultrastructurally, alterations in lysosomes suggested that "residual bodies" developed in a variety of ways, i.e., from the endoplasmic reticulum, multivesicular bodies, or autophagic vacuoles. Following all three agents the endoplasmic reticulum assumed the form of "rough" or "smooth" whorls, and, after two of the agents, arginine deprivation or ultraviolet radiation, it acquired cytochemically demonstrable acid phosphatase activity. Near connections between the endoplasmic reticulum and lysosomes raise the possibility that in KB cells, at least when injured, the endoplasmic reticulum is involved in the formation of lysosomes and the transport of acid phosphatase to them.