ER Cisternae Research Articles

Influence of triethyllead on growth and ultrastructure of tobacco pollen tubes

The toxic effects of triethyllead (TriEL) on semi-vivo cultured pollen tubes of Nicotiana sylvestris were investigated by light and electron microscopy. TriEL caused severe retardation of pollen tube growth and induced three obvious alterations in pollen tube ultrastructure: (a) thickening of tube wall and accumulation of clumped fibrous wall materials in the apex, (b) formation of electron dense inclusions in the dilated portions of the rough ER cisternae and in the periplasmic space between the plasma membrane and tube wall, and (c) disappearance of microtubules in the sperm cells. A 50% reduction in pollen tube bundle lenght (ED 50 value) was observed at a TriEL concentration of 0.683 mg/l (2.1 μM). The tube growth inhibition might result from an irregular orientation and clumping of fibrous non-callosic, non-cellulosic polysaccharide wall materials and their accumulation in the apex and outer lateral wall layer.

Immunocytochemical localization of BGP in human bones in various developmental stages and pathological conditions

Bone gamma-carboxyglutamic acid containing protein (BGP) was isolated from human bone and anti-BGP antibody was produced in rabbits. Localization of BGP was investigated immunocytochemically by light and electron microscopy in human bones in various developmental stages and pathological conditions. In the bones of a 12 week fetus, osteoblasts stained strongly in areas of bone formation. However in bones of late fetal stages and in newborns and adults, BGP was localized predominantly in the osteoid and bone matrix in the ossifying front. Osteoblasts and osteocytes also stained positive, but less dominantly than in the early fetus. Electron microscopy showed that BGP was localized in the ER and Golgi cisternae of osteoblasts and osteocytes, and the collagen fibers of the osteoid and bone matrix. The intensity and distribution of staining were not significantly different in osteoporosis and osteoarthritis. These observations indicate that BGP is synthesized by osteoblasts most actively in early fetal life and is then deposited on collagen fibers of the osteoid and bone matrix.

Immunocytochemistry of calciosomes in liver and pancreas.

Calciosomes are small cytoplasmic vacuoles identified in various nonmuscle cell types by their content of protein(s) similar to calsequestrin (CS), the Ca2+ storage protein of the muscle sarcoplasmic reticulum (SR). These entities have been interpreted as the "primitive" counterpart of the SR, and suggested to be the organelle target of inositol-1,4,5-triphosphate action (Volpe, P., K. H. Krause, S. Hashimoto, F. Zorzato, T. Pozzan, J. Meldolesi, and D. P. Lew. Proc. Natl. Acad. Sci. USA. 85:1091-1095). Immunoperoxidase and immunogold experiments carried out in both thick and ultrathin cryosections of rat hepatocytes and pancreatic acinar cells by using antimuscle CS antibodies revealed a specific labeling widely distributed in the entire cytoplasm, while nuclei were negative. Individual calciosomes appeared as small (105 nm) membrane-bound vacuoles intermingled with, and often apposed to ER cisternae and mitochondria. Other calciosomes were scattered in the Golgi area, in between zymogen granules and beneath the plasma membrane. The cumulative volume of the CS-positive organelles was measured to account for the 0.8 and 0.45% of the cytoplasm in liver and pancreas cells, respectively. The real total volume of the calciosome compartment is expected to be approximately twice as large. In hepatocytes, structures similar to CS-positive calciosomes were decorated by antibodies against the Ca2+ ATPase of muscle SR, while ER cisternae were not. By dual labeling, colocalization was revealed in 53.6% of the organelles, with 37.6% positive for the ATPase only. CS appeared preferentially confined to the content, and the Ca2+ ATPase to the contour of the organelle. The results suggested a partial segregation of the two antigens, reminiscent of their well-known segregation in muscle SR. Additional dual-label experiments demonstrated that hepatic calciosomes express neither two ER markers (cytochrome-P450 and NADH-cytochrome b5 reductase) nor the endolysosome marker, luminal acidity (revealed by 3-[2,4-dinitroanilino]-3'-amino-N-methyl dipropylamine). Calciosomes appear as unique cytological entities, ideally equipped to play a role in the rapid-scale control of the cytosolic-free Ca2+ in nonmuscle cells.

Cytochemical analysis of the reconstitution of endoplasmic reticulum after microinjection of rat liver microsomes into Xenopus oocytes.

Fragments of rough and smooth endoplasmic reticulum purified from rat liver were injected into Xenopus oocyte cytoplasm. Light and electron microscopy, cytochemistry, immunocytochemistry, and enzyme assay were employed to determine the fate of heterologous membranes in the host cytoplasm. The in vivo-incubated microsomes disappeared in a time-dependent manner. Within 3 hr, rough microsomes were replaced by flattened ER cisternae and smooth microsomes were replaced by a network of anastomosing tubules. Polyclonal antibodies against rat liver microsomes and protein A-gold complexes were applied to glycol methacrylate sections of microinjected oocytes. Specific labeling was observed over discrete rough and smooth ER cisternae 3 hr after microinjection. Endogenous ER was not labeled by this technique, and label was not observed when sections were treated with pre-immune antibodies. Diaminobenzidene cytochemistry of microinjected rat lacrimal gland microsomes revealed enzyme activity in heterologous microsomes after 3 hr of in vivo incubation. Control injected microsomes (inactivated by heat denaturation) became associated with autophagic vacuoles, coincident with changes in lysosomal activity. Freshly isolated un-denatured microsomes did not provoke changes in lysosomal activity, and glucose-6-phosphatase activity associated with microinjected membranes could be detected 21 hr after in vivo incubation. Since rat liver microsomes reconstitute after in vivo incubation into cytoplasmic structures resembling those from which they were derived, we conclude that the microinjected membrane fragments act as templates for their own three-dimensional organization.

Pentachlorophenol affects mitochondria and induces formation of golgi apparatus-endoplasmic reticulum hybrids in tobacco pollen tubes

The toxic effect of pentachlorophenol (PCP) on the growth and ultrastructure of tobacco pollen tubes was tested using a semivivo technique of tube culture. In this technique the pollen tubes were allowed to grow in the pistilin situ for 24 hr before they protruded from the cut end of the style and came into contact with the medium containing PCP. The inhibitory effect of different PCP concentrations was determined by measuring the length of tube bundles. The intracellular action of PCP was analysed by electron microscopy. This biocide caused four obvious alterations in the pollen tube ultrastructure: (1) swelling of the mitochondrial saccules; (2) enlargement of the dictyosomes by the increase of the cisternal diameter and the number of cisternae per stack; (3) formation of cup-shaped Golgi apparatus-endoplasmic reticulum hybrid structures (GER hybrids) showing continuities of ER and Golgi cisternae; (4) formation of stacked and/or concentric arrangements of rough ER cisternae. It is suggested that swelling of saccules was directly due to the uncoupling of oxidative phosphorylation whereas the changes of the endomembrane system were caused by energy depletion due to the inhibition of ATP synthesis. These changes are consistant with dynamic concepts of dictyosome and ER function when membrane formation exceeds membrane use in the production of secretory and transition vesicles. Thus, the enlargement of the dictyosomes and the formation of GER hybrids are thought to result from inhibition of budding of vesicles from the Golgi apparatus or from both the ER and the Golgi apparatus, respectively.

Changes in carbohydrates and ultrastructure in xylem ray cells ofPopulus in response to chilling

The ultrastructure of xylem ray cells inPopulus was studied in conjunction with their content of individual sugars and of starch. They differ considerably in structure and in carbohydrate content at the three chosen stages,i.e., of starch deposition (August), of starch maximum (November), and of starch dissolution (January). The transition from the summer to winter stage was also induced experimentally by storage of tissue at 0°C. Both in nature and after cold-storage, sucrose and its galactosides raffinose and stachyose were accumulated to a great extent, contributing up to 69.7 and 57.3% of total sugar content, respectively. They originated parallel to the breakdown of starch and to the appearance of abundant vesicular and dilated ER cisternae. Results indicating that they are the specific sites of sucrose accumulation, and/or its galactosides, are discussed. The occurrence of phytoferritin-like crystalloids in amyloplasts and of vacuolar flocculent material, which condenses into electron-dense bodies of suspectedly proteinaceous nature, is described.

Changes in the membrane components of nondividing cambial cells

Horse-chestnut cambial cells are characterized by the formation of numerous plasmalemma invaginations and the accumulation of membrane whorls in the vacuoles during the transition from activity to rest. This suggests an active membrane trafficking which was investigated with conventional electron microscopy methods combined with selective cytochemical staining. After the cessation of meristematic activity, cell wall thickening is accompanied by increased dictyosome activity. The incorporation of dictyosome vesicles into the plasma membrane produces an increase in plasmalemma surface area in these nongrowing cells. This increase is compensated for by endocytosis accomplished by the formation of saclike plasmalemma invaginations into the peripheral cytoplasm. These invaginations often contain vesicles and tubules. When these invaginations come in contact with a vacuole, they appear to push the tonoplast into the vacuole and form double-membrane protrusions which may eventually separate from the plasmalemma. ER cisternae situated in the intermembrane zone also appear to be transported into the vacuole. Other cisternae may be directly sequestered into the vacuole or take part in the formation of the myelinlike structures which were observed in the cytoplasm. Thus, the vacuoles appear to fill progressively with complex membranous structures of various origins (plasmalemma, tonoplast, ER). It is suggested that their subsequent disappearance during the winter is a consequence of the hydrolytic action of vacuolar contents.

Endocytic apparatus and transcytosis in epithelial cells of the vas deferens in the rat.

The apex of the principal epithelial cells lining the vas deferens of the rat contains coated pits in continuity with the apical plasma membrane and large subsurface-coated vesicles (100-125 nm). In the apical cytoplasm, large, pale, uncoated vesicles (150-300 nm), small coated and uncoated vesicles (50-60 nm), uncoated vesicles about 75-90 nm, and membranous apical tubules are present, in addition to large, vacuolar, pale, multivesicular bodies, dense multivesicular bodies, and secondary lysosomes seen deeper in the cytoplasm amongst numerous ER cisternae, saccules of the Golgi apparatus, and mitochondria. The endocytic activity of these cells was investigated by using cationic ferritin (CF) as a marker of adsorptive endocytosis and native ferritin (NF) for demonstrating fluid-phase endocytosis. These tracers were injected separately into the lumen of the vas deferens, and the animals were killed at various time intervals thereafter from 2 to 90 minutes. At 2 minutes CF was seen bound predominantly to microvilli and to areas of the apical plasma membrane delimiting coated pits as well as in large, coated vesicles. At 5 and 15 minutes the tracers were seen in apical tubules and pale multivesicular bodies; at 30 minutes moderately dense multivesicular bodies were labeled. At 1 hour and longer time intervals dense multivesicular bodies and secondary lysosomes were labeled. NF followed the same pathway as CF; however, no binding to microvilli or areas delimiting coated pits was observed. The numerous other vesicular structures, i.e., the large uncoated vesicles (150-300 nm) and the small coated and uncoated vesicles (50-60 nm), never became labeled with the tracers and therefore were not involved in the endocytic process. There was, however, an exception in the case of several small (75-90 nm) uncoated vesicles seen deeper in the apical cytoplasm of these cells which were labeled exclusively with CF. With time such vesicles appeared along the lateral and basal surfaces of these cells and discharged their content of CF into the lateral intercellular space or the connective tissue space at the base of these cells. Thus the principal epithelial cells in addition to sequestering the endocytosed tracers within secondary lysosomes where they are presumably degraded also appear to be involved in the transcytosis of material from the lumen of the vas deferens to the underlying lamina propria.

Ca2+-binding sites and phosphatase activities in sieve element reticulum and P-protein of chick-pea phloem. A cytochemical and X-ray microanalysis survey

In stem ofCicer arietinum, the loss of ribosomes attached to the rough ER cisternae during sieve element ontogeny results in the formation of “sieve element reticulum” (SER). By enhancing contrast of the SER, the OsFeCN postfixation/staining of material prefixed in glutaraldehyde in presence of calcium enables a good visualization of this membrane system. The pattern of staining in the SER is slightly lower when Mg2+ is substituted for Ca2+. These results support the view that the OsFeCN staining requires divalent cations and that the SER can accumulate Ca2+. The detection of Ca2+ by means of the pyroantimonate method in conjunction with X-ray microanalysis and the cytochemical localization of Ca2+ -ATPase in the SER cisternae provides evidence for Ca2+ sequestration by the SER. On the other hand, Ca2+-binding sites and ATPase activity are localized in P-protein. The ability to bind Ca2+ seems to enable the SER to function as an effective Ca2+ sink which may participate—with the sieve tube plasma membrane and mitochondria—in the maintenance of low Ca2+ concentration in phloem sap. In addition, the close association between P-protein and SER membranes exhibiting Ca2+-binding capabilities suggests that a Ca2+-mediated functional relationship may exist between the two structures. It is postulated that the SER may play a role in putative Ca2+ control of P-protein organization.

A morphologic analysis of neurons and neuropil in the dorsal lateral geniculate nucleus of aged rats

Light and electron microscopy were used to investigate the morphology of neuropil and neuronal cell bodies of the dorsal lateral geniculate nucleus (LGNd) of aged rats. Light microscopic examination reveals that, despite the optic tract showing signs of degeneration, the LGNd is scarcely affected. Thus, a slight but significant reduction in the diameters of both soma and nuclei is observed in aged neurons of the LGNd. Ultrastructural analysis demonstrates a few degenerating profiles of the neuropil. Neurons resembling relay cells exhibit typical features of aged neurons. Cells showing a very infolded nucleus, most of the ER cisternae connected with the nuclear envelope, abundant free polyribosomes and subsurface cisterns associated with mitochondria are similar to interneurons of adult rats. Therefore, aging and partial loss of visual input appear to induce small changes in the morphology of most of LGNd neurons.

Development and cell surface of a non-syncytial invasive tapetum inCanna: Ultrastructural, freeze-substitution, cytochemical and immunofluorescence study

The anther ofCanna indica L. ×C. sp. hybrid contains a hitherto uncharacterized non-syncytial, invasive category of tapetum. With the onset of prophase I the tapetal walls are dissolved and the released protoplasts migrate into the loculus, where they stay discrete. Concomitant with the dissolution of walls the tapetal protoplasts develop a 17 nm thick extracellular granulo-fibrillar cell coat. This feature develops in the synchronous phase of tapetal development. The cell coat reacts positively with ruthenium red, potassium ferrocyanide, ConA-FITC and in the Thiery reaction. Immunofluorescence microscopy using anti-tubulin revealed that even after the migration of tapetal cells into the loculus, the microtubules retain a predominant orientation in the cell cortex, probably derived from that in the original tapetal walled cells. This order is lost during late post-meiotic stages when the cells distort and can produce amoeboid processes. The microtubule orientation is correlated with that of the cell coat fibrils. Tapetal cells vary in ultrastructure and the density of cell coat fibrils after their migration into the loculus, but the cell coat persists until the cells degenerate. It is surmised that development of the cell coat relates to the lack of cell fusion and that the cortical microtubules help to sustain cell form. During post-meiotic stages the free tapetal cells develop massive peripheral arrays of interconnected ER cisternae, probably as part of a secretory apparatus which matures when the spores are producing their ornamented walls. Buds grown in colchicine solution showed accumulation of sporopolleninlike granules in all extracellular spaces of the anther cavity.

Ultrastructure of the early stages of carposporophyte development in the red algaChondria tenuissima (Rhodomelaceae, Ceramiales)

The ultrastructure of the early stages of carposporophyte development in the marine red algaChondria tenuissima has been studied. The diploid carposporophyte grows on the gametophyte. Apical gonimoblast cells develop into diploid carpospores. The basal gonimoblast cells cease to divide and undergo considerable cytoplasmic changes before they become incorporated into the expanding fusion cell. Nucleus and plastids degenerate gradually, while mitochondria remain intact. The smooth endoplasmic reticulum becomes prominent, it seems to produce small vesicles with electron dense contents. Simultaneously, numerous mucilage sacs are formed, presumably from dilating ER cisternae. The contents of the mucilage sacs are secreted by exocytosis. The pit connections between gonimoblast cells flare out. They remain as isolated bodies without connection to a wall after fusion. Secondary pit connections occur between vegetative gametophyte cells and sterile carposporophyte cells. There are three different morphological types of pit connections.

Ontogenesis of Tannin Coenocytes in Sambucus racemosa L. II. Mother Tannin Cells

Tannin coenocytes develop from mononucleate tannin mother cells. The process occurs within the whole of the first (youngest) internode and its development can be divided into three stages. In stage I the MTC is isodiametric and similar to the surrounding cells of the flank meristem, being present in the ninth cell layer from the apex surface. The nucleus becomes lanceolate and elongates, and large cytoplasmic vacuoles appear. A twofold elongation of both the cell and nucleus continues in the second stage, the cell-nucleus ratio indicating that it is due to the enlarged vacuole, which pushes a thin layer of cytoplasm close to the cell wall. In this layer of cytoplasm dilated ER cisterna occur together with small and large vacuoles, a fusion of the vacuoles increasing their volume. Such cells are diploid in spite of larger nuclear volume and rough structure of its chromatin.

Cytochalasin B affects the structural polarity of statocytes from cress roots (Lepidium sativum L.).

The effect of cytochalasin B (CB; 25 micrograms ml-1 in 1% dimethylsulfoxide, DMSO) upon the structural polarity of statocytes in cress roots is demonstrated. If normal, vertically grown roots are incubated in CB, the structural polarity of the statocytes is altered according to the developmental stage of the root. Statocytes from young roots (13 or 17 hours, additionally 7 hours CB) are characterized by proximal ER cisternae and a sparsely developed distal ER-complex. Statocytes from older roots (24 hours, additionally 7 hours CB) still accumulate distal ER, as in control roots, but at the proximal cell pole in the vicinity of the nucleus additional ER is found. These effects are reversed by washing out the drug in DMSO. Growth of the roots under a continuous supply of CB yields statocytes with sedimented nuclei, proximal ER and almost no distal ER. Together with quantitative data from morphometric studies, a dynamic model of the expression of inherent cell polarity in structural polarity is proposed.

Spermatogenesis of the water strider, Gerris paludum (Heteroptera, Gerridae)

Fine structural observations during spermatogenesis in Gerris paludum are described, including changes in mitochondria, acrosome, and microtubules. Early spermatid perinuclear ER cisternae and later stage membrane-bound ER cisternae seem to form a membranous sleeve at mid spermatid. G. paludum has an extremely long and complex acrosome. The acrosomal vesicle of the early spermatid is nearly spherical and 6.7 μ m in diameter and the acrosome of the mature spermatozoon is 0.4 μ m in diameter at the base and 0.16 μ m at the tip. In mature bundled spermatozoa the acrosomes are arranged in a regular direction. The cytoplasm of the cyst cell contains a great number of modified microtubules with one to three long projections, all of which curve in the same direction. Microtubules along the interior rim of the cyst cell have one or no projections. This type of microtubule is evidently specific for the genus Gerris of the family Gerridae.

Light and electron microscopical changes in the liver of mice following treatment with aminotriazole

Light and electronmicroscopical alterations in the liver of mice were shown oral administration of aminotriazole. Intensity of damage depends on concentration. High dosage affects all inter- and intralobular portions. Presented light microscopical results show, next to an extreme hypertrophy of hepatocytes, well distinguishable nuclear deformations and increased pyknotic nucleoli. The cytoplasm contains numerous vacuoles and lipid droplets. Fine structural alterations after drug administration include defects on all cell organelles and membranes. Especially proliferations of smooth ER cisternae with simultaneous reductions of rough ER cisternae, increase of microbodies as well as numerous defects on mitochondria were obvious. Interpretations of the mode of action of aminotriazole is based on ultrastructural results and biochemical knowledge on this herbicide and compared with other toxicants.

Cellular compartments in mitotic cells: ultrahistochemical identification of Golgi elements in PtK-1 cells.

This investigation focuses on the identification, distribution, and transport of intracellular membrane systems during mitosis. The membranes of the Golgi apparatus can be identified cytochemically by staining for acid phosphatase (acPase) and thiamine pyrophosphatase (TPPase) activity. Using this approach we are able to study the disintegration of the Golgi apparatus during mitosis and to follow the dislocation as well as the organized reappearance of Golgi elements after the completion of mitosis. We are able to demonstrate that during mitosis the activity of both enzymes is strong enough to react with the substrate applied during the staining procedure. Furthermore, we observe a characteristic pattern of membrane distribution in mitotic cells. During interphase the TPPase reaction is characteristically limited to one or two cisternae of a dictyosomal stack. The acPase reaction stains the membranes of the total stack, of the GERL, of some vesicles and cisternae near the dictyosomes and lysosomes. After the mitotic breakdown of the dictyosomal stacks the forming vesicles still stain positively and are distributed over the entire cytoplasm. At late anaphase and early telophase the enzyme activity occurs not only in the reconstituting dictyosomes but also in the nuclear envelope and in some ER cisternae. The extended spectrum of membrane structures indicating Golgi enzyme activity becomes obvious. This phenomenon favors the idea that at least some functions of the Golgi apparatus persist during mitosis.

Osmium impregnation of the endoplasmic reticulum correlates with the functional status of prostatic secretory cells.

The ultrastructure of prostatic secretory cells was studied with the osmium impregnation technique in order to determine if the ER reactivity, or its absence, and its three-dimensional organization correspond to specific functions possibly hormono-dependent. Thick sections (0.3 micron) of rat ventral prostate were made after a five-day impregnation with osmium tetroxide and examined by standard transmission electron microscopy at 80 kV. Studies were performed in normal adult rats, between the 3rd and 26th day following castration and in castrated rats treated with 5-alpha-dihydrotestosterone. In normal rats the impregnation technique delineated three secretory cell types (dark, greyish and clear), representing various degrees of reactivity in ER cisternae; however, despite this quantitative variation, they had similar morphological characteristics. In a longitudinal section, the ER network appeared to be made of saccules running parallel along the length of the cell and forming whorl-like patterns around the nucleus. Comparison of sections taken at various angles suggests that the ER network is made of concentric parallel saccules extending from the base to the apex of the cell and encircling the nucleus and the Golgi apparatus like a large multilayered cylinder. Whereas in dark cells the Golgi apparatus contained mostly clear vesicles, it was always heavily impregnated in clear cells. Noteworthy, osmium deposits were rarely observed on the nuclear envelope of secretory cells but were always present in basal cells. After castration, secretory cells became progressively cubic and the most conspicuous cytoplasmic change was observed in association with the ER. The Golgi apparatus decreased markedly in volume and became heavily stained with metallic osmium.(ABSTRACT TRUNCATED AT 250 WORDS)

Immunoelectron microscopic demonstration of prostatic acid phosphatase in human hyperplastic prostate

Immunoelectron microscopic studies were done on prostatic tissues obtained from patients with benign hyperplasia. Rabbit IgG-peroxidase conjugate against purified human prostatic acid phosphatase band 2 (HPAP-2) was used for studies. Under the light microscope, the columnar secretory epithelia of prostatic glands showed different intensity and distribution of immunostaining whereas the basal cells were unstained. Under the electron microscope, the secretory epithelial cells often showed electron-dense reaction product in the Golgi apparatus and secretory vesicles and vacuoles, and only sparingly in the cisternae of nuclear envelope and rough ER. Sometimes, fusion of secretory vacuolar membrane and plasma membrane and discharge of the vacuolar contents into the extracellular space were noted. The surfaces of microvilli at the apical portion of the columnar epithelia and the lumen of the glandular acini always showed reaction product. These findings suggest that HPAP-2 may be synthesized in the rough ER and transported to the Golgi apparatus where it is concentrated and transferred to the secretory vesicles and vacuoles. HPAP-2 is finally discharged into the extracellular spaces through exocytosis, a secretory mechanism similar to that of other secretory proteins.

Ultrastructure of granulopoiesis in tunica propria of the bursa of fabricius

Heterophilic and eosinophilic leukocyte series appear adjacent to blood vessels or surrounding the primordium of lymphoid follicles in the tunica propria at stages 38-45 (6) of chicken embryo bursa. Development of the heterophilic granulocytes promyelocytes are the largest of the heterophils (8~m diameter) and the first recognizable stage (fig. i). They have short, thick fingerlike cytoplasmic projections, an eccentric round nucleus with fine dense chromatin pattern, a small peripheral zone, central clumps of heterochromatin, and one or two nucleoli. Round and elongated mitochondria, free ribosomes and several cisternae of rough ER dilated to a variable degree can be seen. The juxtanuclear Golgi complex is surrounded by numerous medium dense vesicles. Spherical electron-dense granules (type I) of 0.1/~m diameter are present. Myelocytes are spherical cells of 7~m diameter (fig. 2) with an eccentric round and a slightly idented nucleus and more peripheral heterochromatin than in promyelocytes; one or two nucleoli were sometimes observed. Flattened cisternae of rough ER were arranged parallel to the cellular surface and free ribosomes were less numerous than in promyelocytes. Mitochondria were located in the cell periphery. A well developed Golgi complex surrounds the centrioles. Dense spherical and peripherally located (0.3-0.7~m of diameter) (type I), similar to those described in promyelocytes, and less dense spherical or oval granules (O.IXO.4~Im) of variable electron-density (type II) are distinguished.