BALB/c mice were repeatedly immunized with a galactosyl transferase-rich microsomal fraction of rat myeloma cells. Spleen cells were subsequently fused with Sp2/0 mouse myeloma cells, the resulting hybridomas were cloned, and their secreted Ig was screened for reactivity with antigens belonging to the Golgi complex. One such monoclonal antibody, 6F4C5, gave especially intense immunofluorescent staining of the Golgi area of myeloma cells and fibroblasts. It recognized two proteins bands on immunoblots of gel-fractionated cell lysates: a major one with an estimated Mr of 54,000 and a minor one at 86,000. Both proteins were concentrated in microsomal fractions isolated at low ionic strength. They were hydrophilic judging from partitioning of a Triton X-114 cell lysate. Both were cytoplasmically oriented as demonstrated by protease and high KCl treatments of postmitochondrial supernatants and microsomal fractions. Neither was retained by columns of insolubilized wheat germ agglutinin or concanavalin A, which suggests that they are not glycoproteins. Their more detailed location in the Golgi complex was studied by immunoelectron microscopy, using a saponin permeabilization procedure and peroxidase-conjugated reagents. The observed staining was restricted to two or three cisternae in the medial part of the stack. Nevertheless, differential centrifugation experiments indicated that the two antigens may be recovered in distinct subcellular fractions: this may be related to the unexpected observation that rather low salt concentrations strip the antigens from microsomal fraction.

We report here that interleukins have a dramatic effect on extracellular matrix production by cultured endothelial cells. Human umbilical vein endothelial cells incubated with growth media conditioned by lectin-activated human peripheral blood mononuclear leukocytes undergo marked changes in cell shape and elaborate a highly organized extracellular material that is not detectable in untreated cultures. This material has the following characteristics: (a) it is not recognizable by electron microscopy unless the cationic dye, Alcian blue, is added to the fixative; (b) it is visualized as a network of branching and anastomosing fibrils of various thickness that can be resolved into bundles of fine filaments; (c) it is associated with the cell surface, extends between contiguous cells, and coats the culture substrate; (d) it is removed by digestion with glycosaminoglycan-degrading enzymes, such as crude heparinase and chondroitinase ABC. These results demonstrate that soluble factors released by activated peripheral blood mononuclear leukocytes (interleukins) stimulate cultured human umbilical vein endothelial cells to produce a highly structured pericellular matrix containing glycosaminoglycans (probably chondroitin sulfate and/or hyaluronic acid) as a major constituent. We speculate that this phenomenon corresponds to an early step of angiogenesis as observed in vivo as a consequence of interleukin release.

We have studied the behavior of cloned capillary endothelial cells grown inside a three dimensional collagen matrix. Cell monolayers established on the surface of collagen gels were covered with a second layer of collagen. This induced the monolayers of endothelial cells to reorganize into a network of branching and anastomosing capillary-like tubes. As seen by electron microscopy, the tubes were formed by at least two cells (in transverse sections) delimiting a narrow lumen. In addition, distinct basal lamina material was present between the abluminal face of the endothelial cells and the collagen matrix. These results showed that capillary endothelial cells have the capacity to form vessel-like structures with well-oriented cell polarity in vitro. They also suggest that an appropriate topological relationship of endothelial cells with collagen matrices, similar to that occurring in vivo, has an inducive role on the expression of this potential. This culture system provides a simple in vitro model for studying the factors involved in the formation of new blood vessels (angiogenesis).

We investigated the subcellular sites of glycoprotein oligosaccharide maturation by using lectin conjugates to stain lightly-fixed, saponin-permeabilized myeloma cells. At the electron microscopic level, concanavalin A-peroxidase stains the cisternal space of the nuclear envelope, the rough endoplasmic reticulum, and cisternae along the proximal face of the Golgi stack. Conversely, wheat germ agglutinin-peroxidase stains cisternae along the distal face of the Golgi stack, associated vesicles, and the cell surface. These observations confirm the existence of two qualitatively distinct Golgi subcompartments, show that the lectin conjugates can be employed as relatively proximal or distal Golgi markers under conditions of excellent ultrastructural preservation, suggest that the asymmetric distribution of qualitatively distinct oligosaccharides is a property of underlying cellular components and not simply of the principal secretory product, and suggest that the oligosaccharide structure recognized by wheat germ agglutinin is attained during transport from the proximal toward the distal face of the Golgi stack.

We incubated mouse peritoneal macrophages for 3-8 min at 37 degrees C with antibody-coated sheep erythrocytes and examined regions of close interaction between the two cell types by electron microscopy. At sites of focal macrophage-erythrocyte contact we observed a distinctive specialization of the macrophage plasma membrane consisting of a prominent subplasmalemmal band of electron-dense material, approximately 25-35 nm in thickness. In many instances, this band showed a periodic substructure similar to that seen in clathrin coats. Moreover, many slender erythrocyte processes penetrated into invaginations of the macrophage surface which were bristle-coated at their blind extremity. As previously shown for clathrin-coated pits, the segments of the macrophage plasma membrane beneath which the defense material was found were selectively resistant to the membrane-perturbing effect of the antibiotic, filipin. This structural specialization of the macrophage plasma membrane at sites of ligand-receptor interaction during immune phagocytosis of antibody-coated erythrocytes may represent the morphological counterpart of the zipper mechanism of phagocytosis previously demonstrated by functional studies.

Immunoglobulin M (IgM)-secreting murine plasmablasts have been used to explore the cytologic site(s) of the successive modifications of the polypeptide H and L chains (steps of glycosylation, chain assembly, and polymerization) which occur during intracellular transport (ICT) and the interrelationships between these events. A combination of pulse-chase biosynthetic labeling protocols (using amino acids and sugars), subcellular fractionation, and electron microscope autoradiography was used in conjunction with inhibitors of glycosylation and agents (carboxyl cyanide m-chlorophenyl hydrazone [CCCP] and monensin) which block Ig exit from the rough endoplasmic reticulum (RER) or Golgi cisternae. The data are consistent with the following conclusions: (1) Sugar addition and modification occur in three main steps: (a) en bloc addition of core sugars to nascent H chains, (b) partial trimming of these oligosaccharide chains in the RER, (c) quasiconcerted addition of terminal sugars (galactose, fucose, and sialic acid) in a very distal compartment between monensin-sensitive Golgi cisternae and the cell surface. (2) H and L chain assembly occurs between nascent H chains and a pool of free light chains present in the RER, followed by interchain disulfide bonding and rapid assembly of monomers into J chain-containing pentamers in the RER. Small amounts of various apparently non-obligatory intermediates in polymerization are also formed. (3) Carbohydrate addition is not required for chain assembly, polymerization, and secretion since completely unglycosylated chains (synthesized in the presence of deoxyglucose or tunicamycin) undergo polymerization and are secreted (although at a reduced rate). (4) Surface 8s IgM molecules do not represent a step in the IgM secretory pathway.

The physiology of protein intracellular transport and secretion by cell types thought to be free from short-term control has been compared with that of the pancreatic acinar cell, using pulse-chase protocols to follow biosynthetically-labeled secretory products. Data previously obtained (Tartakoff, A.M., and P. Vassalli. J. Exp. Med. 146:1332-1345) has shown that plasma-cell immunoglobulin (Ig) secretion is inhibited by respiratory inhibitors, by partial Na/K equilibration effected by the carboxylic ionophore monensin, and by calcium withdrawal effected by the carboxylic ionophore A 23187 in the presence of ethylene glycol bis (beta-aminoethylether)-N,N,N',N'-tetraacetic acid (EGTA) and absence of calcium. We report here that both inhibition of respiration and treatment with monensin slow secretion by fibroblasts, and also macrophages and slow intracellular transport (though not discharge per se) by the exocrine pancreatic cells. Attempted calcium withdrawal is inhibitory for fibroblasts but not for macrophages. The elimination of extracellular calcium or addition of 50 mM KCl has no major effect on secretory rate of either fibroblasts or macrophages. Electron microscopic examination of all cell types shows that monensin causes a rapid and impressive dilation of Golgi elements. Combined cell fractionation and autoradiographic studies of the pancreas show that the effect of monensin is exerted at the point of the exit of secretory protein from the Golgi apparatus. Other steps in intracellular transport proceed at normal rates. These observations suggest a common effect of the cytoplasmic Na/K balance at the Golgi level and lead to a model of intracellular transport in which secretory product obligatorily passes through Golgi elements (cisternae?) that are sensitive to monensin. Thus, intracellular transport follows a similar course in both regulated and nonregulated secretory cells up to the level of distal Golgi elements.

The kinetics of appearance of newly made 60S and 40S ribosomal subunits in the free and membrane-bound ribosomal particles of P3K cells were explored by determining the specific radioactivities of their 18S and 28S RNA after various lengths of [3H]uridine pulse. Both 40S and 60S subunits enter free and membrane-bound polyribosomes at comparable rates from the cytoplasmic pool of newly made, free native subunits, the 40S subunits entering the native subunit pool and the polyribosomes slightly earlier than the 60S subunits. At all times, the specific radioactivity of the membrane-bound native 60S subunits was slightly lower than that of the polyribosomal 60S subunits. This indicates that the membrane-bound native 60S subunits are not precursors destined to enter membrane-bound polyribosomes and suggests that they result from the dissociation of ribosomes after chain termination. The results observed also suggest that the membrane-bound native 60S subunits are not reutilized before their release from the membranes, which probably takes place shortly after dissociation from their 40S subunits. The monoribosomes, both free and membrane-bound, had the lowest specific radioactivities in their subunits. Finally, a small amount of newly made native 40S subunits, containing 18S RNA of high specific radioactivity, and apparently also newly made messenger RNA were detected on the membranes. The high turnover of these membrane-bound native 40S subunits suggests that they may represent initiation complexes formed with mRNA which has just reached the membranes and which has not yet given rise to polyribosomes.

Mild ribonuclease treatment of the membrane fraction of P3K cells released three types of membrane-bound ribosomal particles: (a) all the newly made native 40S subunits detected after 2 h of [3H]uridine pulse. Since after a 3-min pulse with [35S]methionine these membrane native subunits appear to contain at least sevenfold more Met-tRNA per particle than the free native subunits, they may all be initiation complexes with mRNA molecules which have just become associated with the membranes; (b) about 50% of the ribosomes present in polyribosomes. Evidence is presented that the released ribosomes carry nascent chains about two and a half to three times shorter than those present on the ribosomes remaining bound to the membranes. It is proposed that in the membrane-bound polyribosomes of P3K cells, only the ribosomes closer to the 3' end of the mRNA molecules are directly bound, while the latest ribosomes to enter the polyribosomal structures are indirectly bound through the mRNA molecules; (c) a small number of 40S subunits of polyribosomal origin, presumably initiation complexes attached at the 5' end of mRNA molecules of polyribosomes. When the P3K cells were incubated with inhibitors acting at different steps of protein synthesis, it was found that puromycin and pactamycin decreased by about 40% the proportion of ribosomes in the membrane fraction, while cycloheximide and anisomycin had no such effect. The ribosomes remaining on the membrane fraction of puromycin-treated cells consisted of a few polyribosomes, and of an accumulation of 80S and 60S particles, which were almost entirely released by high salt treatment of the membranes. The membrane-bound ribosomes found after pactamycin treatment consisted of a few polyribosomes, with a striking accumulation of native 60S subunits and an increased number of native 40S subunits. On the basis of the observations made in this and the preceding papers, a model for the binding of ribosomes to membranes and for the ribosomal cycle on the membranes is proposed. It is suggested that ribosomal subunits exchange between free and membrane-bound polyribosomes through the cytoplasmic pool of free native subunits, and that their entry into membrane-bound ribosomes is mediated by mRNA molecules associated with membranes.

A cell fractionation procedure is described which allowed, by use of MOPC 21 (P3K) mouse plasmocytoma cells in culture, the separation of the cytoplasmic free and membrane-bound ribosomes in fractions devoid of mutual cross-contamination, and in which the polyribosomal structure was entirely preserved. This was achieved by sedimentation on a discontinuous sucrose density gradient in which the two ribosome populations migrate in opposite directions. A variety of controls (electron microscopy, labeling of membrane lipids, further repurification of the isolated fractions) provided no evidence of cross-contamination of these populations. However, when an excess of free 60S or 40S subunits, labeled with a different isotope, was added to the cytoplasmic extract before fractionation, the possibility of a small amount of trapping and/or adsorption of free ribosomal particles by the membrane fraction was detected, especially in the case of the 60S subunits; this could be entirely prevented by the use of sucrose gradients containing 0.15 M KC1. EDTA treatment of the membrane fraction detached almost all the 40S subunits, and about 70% of the 60S subunits. 0.5 M KC1 detached only 10% of the ribosomal particles, which consist of the native 60S subunits and the monoribosomes, i.e. the bound particles inactive in protein synthesis. Analysis in CsC1 buoyant density gradients of the free and membrane-bound polyribosomes and of their derived 60S and 40S ribosomal subunits showed that the free and membrane-bound ribosomal particles have similar densities.