Intranuclear and cytoplasmic hemoglobin in human erythroblasts during maturation

Summary Changes in the hemoglobin level in human bone marrow erythroblasts associated with cell maturation were studied by the electron microscopic immunocytochemical technique using protein A-gold.Intense reaction of gold to hemoglobin was observed diffusely in the cytoplasm, but the reaction was weak in the Golgi zone. No reaction was observed in mitochondria or granules. Cytoplasmic hemoglobin was noted in basophilic erythroblasts and increased with maturation. Hemoglobin was also noted in the nucleus, especially in the euchromatin, though in smaller amounts than in the cytoplasm. Since intranuclear hemoglobin tended to increase in the euchromatin but to decrease in the heterochromatin with erythroblast maturation, the ratio of the amount of hemoglobin in the euchromatin to that in the heterochromatin increased with maturation.     

Chlorpromazine induced aggregation of normal human hemoglobin. Electron microscopic evidence.

Normal human hemoglobin exceeding a certain minimum concentration (called critical aggregation concentration) undergoes aggregation in presence of the psychotherapeutic drug chlorpromazine (CPZ). The critical aggregation concentration decreases with the increase of CPZ concentration. Electron micrographs of CPZ-treated hemoglobin clearly indicate that the aggregates of hemoglobin are in filamentous form of average width 75 +/- 8 A. A possible mechanism for such aggregation has been discussed.     

Electron microscopic examination of common red cell ghosts: cytoplasmic contamination.

Red blood cell ghost preparations are often cited as providing unequivocal or convincing evidence for the active transport of solutes from a solution of low concentration across a membrane to a solution of higher concentration. Electron microscopic examination of the more widely used ghost preparations show that a considerable quantity of cytoplasmic macromolecules (including hemoglobin) remain within the treated red blood cells. That is, many of the ghost preparations are not hollow membrane perparations. It is concluded that the problem of active solute transport in red blood cell ghost preparations should be reexamined. Furthermore, experiments with ghost preparations purporting to demonstrate active transport should include electron photomicrographs of the preparation utilized.     

Cytoskeletal distribution and function during the maturation and enucleation of mammalian erythroblasts

We have used murine splenic erythrolasts infected with the anemia- inducing strain of Friend virus (FVA cells), as an in vitro model to study cytoskeletal elements during erythroid maturation and enucleation. FVA cells are capable of enucleating in suspension culture in vitro, indicating that associations with an extracellular matrix or accessory cells are not required for enucleation to occur. The morphology of FVA cells undergoing enucleation is nearly identical to erythroblasts enucleating in vivo. The nucleus is segregated to one side of the cell and then appears to be pinched off resulting in an extruded nucleus and reticulocyte. The extruded nucleus is surrounded by an intact plasma membrane and has little cytoplasm associated with it. Newly formed reticulocytes have an irregular shape, are vacuolated and contain all cytoplasmic organelles. The spatial distribution of several cytoskeletal proteins was examined during the maturation process. Spectrin was found associated with the plasma membrane of FVA cells at all stages of maturation but was segregated entirely to the incipient reticulocyte during enucleation. Microtubules formed cages around nuclei in immature FVA cells and were found primarily in the incipient reticulocyte in cells undergoing enucleation. Reticulocytes occasionally contained microtubules, but a generalized diffuse distribution of tubulin was more common. Vimentin could not be detected at any time in FVA cell maturation. Filamentous actin (F-actin) had a patchy distribution at the cell surface in the most immature erythroblasts, but F-actin bundles could be detected as the cells matured. F-actin was found concentrated between the extruding nucleus and incipient reticulocyte in enucleating erythroblasts. Newly formed reticulocytes exhibited punctate actin fluorescence whereas extruded nuclei lacked F-actin. Addition of colchicine, vinblastine, or taxol to cultures of FVA cells did not affect enucleation. In contrast, cytochalasin D caused a complete inhibition of enucleation that could be reversed by washing out the cytochalasin D. These results demonstrate that F-actin plays a role in enucleation while the complete absence of microtubules or excessive numbers of polymerized microtubules do not affect enucleation.     

Electron microscopic localization of cytoplasmic myosin with ferritin- labeled antibodies

We localized myosin in vertebrate nonmuscle cells by electron microscopy using purified antibodies coupled with ferritin. Native and formaldehyde-fixed filaments of purified platelet myosin filaments each consisting of approximately 30 myosin molecules bound an equivalent number of ferritin-antimyosin conjugates. In preparations of crude platelet actomyosin, the ferritin-antimyosin bound exclusively to similar short, 10-15 nm wide filaments. In both cases, binding of the ferritin-antimyosin to the myosin filaments was blocked by preincubation with unlabeled antimyosin. With indirect fluorescent antibody staining at the light microscope level, we found that the ferritin-antimyosin and unlabeled antimyosin stained HeLa cells identically, with the antibodies concentrated in 0.5-microns spots along stress fibers. By electron microscopy, we found that the concentration of ferritin-antimyosin in the dense regions of stress fibers was five to six times that in the intervening less dense regions, 20 times that in the cytoplasmic matrix, and 100 times that in the nucleus. These concentration differences may account for the light microscope antibody staining pattern of spread interphase cells. Some, but certainly not all, of the ferritin-antimyosin was associated with 10-15-nm filaments. In mouse intestinal epithelial cells, ferritin- antimyosin was located almost exclusively in the terminal web. In isolated brush borders exposed to 5 mM MgCl2, ferritin-antimyosin was also concentrated in the terminal web associated with 10-15-nm filaments.     

Electron microscopic studies of the intracellular polymerization of sickle hemoglobin

Transmission electron microscopy has been used to study intracellular sickle hemoglobin polymer in unfractionated cells from the arterial and venous blood of patients and after external deoxygenation. We detect polymerized hemoglobin in up to 10% of the cells in the venous circulation, especially in cells that are "cigar-shaped" and appear to be irreversibly sickled. We could not see well-defined polymer in mixed arterial samples; nevertheless, we found electron opaque spots, which could be ferritin granules, hemosiderin, or small aggregates of hemoglobin S. However, upon sequential chemical deoxygenation using 1.0% sodium metabisulphite, polymer formation was seen at oxygen saturation values of 75%-85%. Cells that were physically deoxygenated using gas mixtures containing nitrogen-carbon dioxide-oxygen mixtures were found to contain distinct polymers of deoxyhemoglobin S at oxyhemoglobin saturation values of 50%-75%. As deoxygenation increases, we detect short, randomly arranged polymer in a loose network, with occasional long polymers. Upon further deoxygenation, the length and number of polymer forms increased. Between 0% and 50% saturation, most erythrocytes were full of long, parallel, closely packed polymers that tend to align and run parallel to the cell membrane. In both chemical and physically deoxygenated blood samples, cells were seen at 50%-75% oxyhemoglobin saturation that retained their normal biconcave disc shape, although they contained significant amounts of polymer. The structural changes in sickle erythrocytes seen in vitro due to physical or chemical deoxygenation of cells, may reflect in vivo intracellular changes in the sickle cell patient.     

Intranuclear maturation pathways of rat liver ribosomal ribonucleic acids.

The maturation of pre-rRNA (precursor to rRNA)in liver nuclei is studied by agar/ureagel electrophoresis, kinetics of labelling in vivo with [14C] orotate and electron-microscopic observation of secondary structure of RNA molecules. (1) Processing starts from primary pre-rRNA molecules with average mol. wt. 4.6X10(6)(45S) containing the segments of both 28S and 18S rRNA. These molecules form a heterogeneous peak on electrophoresis. The 28S rRNA segment is homogeneous in its secondary structure. However, the large transcribed spacer segment (presumably at the 5'-end) is heterogeneous in size and secondary structure. A minor early labelled RNA component with mol.wt. about 5.8X10(6) is reproducibly found, but its role as a pre-rRNA species remains to be determined. (2) The following intermediate pre-rRNA species are identified: 3.25X10(6) mol.wt.(41S), a precursor common to both mature rRNA species ; 2.60X10(6)(36S) and 2.15X10(6)(32S) precursors to 28S rRNA; 1.05X10(6) (21S) precursor to 18S rRNA. The pre-rRNA molecules in rat liver are identical in size and secondary structure with those observed in other mammalian cells. These results suggest that the endonuclease-cleavage sites along the pre-rRNA chain are identical in all mammalian cells. (3) Labelling kinetics and the simultaneous existence of both 36S and 21S pre-rRNA reveal that processing of primary pre-rRNA in adult rat liver occurs simultaneously by at least two major pathways: (i) 45S leads to 41S leads to 32S+21S leads to 28S+18S rRNA and (ii) 45S leads to 41S leads to 36S+18S leads to 32S leads to 28S rRNA. The two pathways differ by the temporal sequence of endonuclease attack along the 41 S pre-rRNA chain. A minor fraction (mol.wt.2.9X10(6), 39S) is identified as most likely originating by a direct split of 28S rRNA from 45S pre-rRNA. These results show that in liver considerable flexibility exists in the order of cleavage of pre-rRNA molecules during processing.     

Electron microscopic immunocytochemistry. Silver enhancement of colloidal gold marker allows double labeling with the same primary antibody

Electron microscopic sections, immunocytochemically labeled with colloidal gold, can be prepared for double labeling by applying the "EM-silver enhancement" procedure. This method, a photographic, so-called physical, development, increases the size of the gold marker to a predeterminable value and thereby inactivates the anti-species antibody present on the gold grain, thus allowing the labeling of a second antigen with antibody raised in the same species.