Membrane blebbing is associated with Ca2+-activated hyperpolarizations induced by serum and alpha 2-macroglobulin

We have reported previously that serum and alpha 2-macroglobulin (alpha 2M) induce Ca2+-activated hyperpolarizations in the membrane potential of a clonal rat osteosarcoma cell line (ROS 17/2) (Dixon and Aubin, J. Cell, Physiol., 132:215-225, 1987). In this report, we describe morphological changes that accompany these hyperpolarizations. Both cell surface blebbing (zeiosis) and transient hyperpolarizations were induced by application of 10% fetal bovine serum (FBS) or alpha 2M; neither was induced by serum-free medium, a suspension of latex beads, or purified bovine serum albumin. Following a brief application of FBS or alpha 2M at time 0, electrical activity typically occurred between 7-40 s and was always followed by blebbing activity that began at 30 s and persisted for 3-5 min. In contrast, continuous exposure to FBS resulted in the persistence of both blebbing activity and transient hyperpolarizations for periods of at least several hours. Scanning electron microscopy (SEM) revealed that the blebs appeared concomitantly with the disappearance of microvilli and the appearance of surface pits that measured 100-300 nm in diameter. Coated pits and vesicles, similar in size to the pits observed by SEM, were observed using transmission electron microscopy (TEM). By TEM, blebs were found to contain few organelles other than centrally located free ribosomes. Fluorescence microscopy of nitrobenzooxadizole-phallacidin-labeled cells indicated that blebs contained filamentous actin and that microfilament bundles remained primarily on the substratum side of blebbed cells. We propose that blebbing results from a dynamic local reorganization of microfilaments initiated by ligand-induced transient increases in intracellular Ca2+.     

Quantitative ultrastructure of isolated Kupffer cells of rat liver

The average volume of isolated Kupffer cells of rat liver is 821 +/- 64 microns 3, the average surface being 423 +/- 24 microns 2 (599 microns 2, with cell processes included). The surface structure (pseudopodia, lamellipodia, filopodia, microvilli) of isolated cells is much less developed than that of Kupffer cells in situ. By morphometric characterization volume densities are 0.1264 +/- 0.0077 (SE) for mitochondria and 0.3591 +/- 0.0169 for lysosomal structures. The volume of mitochondria amount to 0.79 +/- 0.04 microns 3.     

A scanning electron microscope study of mouse peritoneal mesothelium.

As seen in the scanning electron microscope, peritoneal mesothelial cells of the mouse diaphragm, anterior abdominal wall and intestinal serosa carry numerous microvilli. These microvilli are absent over certain areas of the cell surface and are sometimes, interlocked in meshwork patterns or coronal formation. The apical cell membranes of the mesothelium at the base of the microvilli, are invaginated by many plasmalemmal vesicles and vacuoles and carry a number of protruding spherical structures. Deep circular craters, giving the impression of stomata, are also visible.     

Anchorage-independent growth of normal human mesothelial cells: A sensitive bioassay for EGF which discloses the absence of this factor in fetal calf serum

Summary This laboratory recently reported that normal human mesothelial cells require epidermal growth factor (EGF) and hydrocortisone (HC), in addition to fetal calf serum and a complex defined medium component, in order to grow optimally in surface culture (9). We report here that this normal cell type also forms large colonies at high efficiency in semi-solid medium, but exhihits more stringent serum and EGF requirements for anchorage-independent than for surface growth. Mesothelial cells are unable to divide at all in semi-solid medium with added EGF or with less than 2% serum, whereas they grow slowly but progressively in surface culture under such conditions. In semi-solid medium containing 20% serum and HC, mesothelial cells are stimulated to divide by the addition of as little as 30 pg/ml purified EGF. Human urine or male mouse plasma could substitute for purified EGF, yielding growth commensurate with the levels of EGF in these biological fluids previously measured by others using radioreceptor and radioimmune assays. Thus growth of mesothelial cells in semi-solid medium can serve as a highly sensitive assay of EGF biological activity which is unaffected by the presence of serum proteins. In addition, our results demonstrate that fetal calf serum does not provide mitogenic levels of EGF to cultured cells, raising the question of the identity of plasma and serum mitogens. This work was supported by NIH grants RO1 AG02048 and RO1 CA26656 to James G. Rheinwald and by NIH postdoctoral fellowship F32 AG05303 to Paul J. La Rocca.     

Surface configuration of mesothelial cells in effusions. A comparative light microscopic and scanning electron microscopic study.

Surface configuration of mesothelial cells identified by light microscopy (LM) has been studied by scanning electron microscopy (SEM). It has been shown that mesothelial cells may have a variable SEM appearance. The surfaces of a small proportion of mesothelial cells are covered by regular microvilli (MV) and show openings of the pinocytotic vesicles. The surfaces of the majority of these cells are covered by vesicles or blebs. An intermediate population of mesothelial cells, i.e., cells displaying side-by-side blebs and MV, has also been observed. The latter cells no longer display pinocytotic vesicles. Occasional mesothelial cells have smooth surfaces. It has been shown by LM and transmission electron microscopy that cells with blebs are viable and capable of mitotic activity. It is concluded that mesothelial cells, detached from their epithelial setting, lose microvilli and pinocytotic vesicles and acquire surface blebs. The possible relationship between mesothelial cells and macrophages based on surface features has been discussed.     

Ultrastructural analysis of the organization and distribution of insulin receptors on the surface of 3T3-L1 adipocytes: rapid microaggregation and migration of occupied receptors

Monomeric ferritin-insulin and high-resolution electron microscopic analysis were used to study the organization, distribution, and movement of insulin receptors on differentiated 3T3-L1 adipocytes. Analysis of the binding to prefixed cells showed that insulin initially occupied single and paired receptors preferentially located on microvilli. The majority of receptors (60%) were found as single molecules and 30% were pairs. In 1 min at 37% C, 50% of the receptors on nonfixed cells were found on the intervillous plasma membrane and more than 70% of the total receptors had microaggregated. By 30 min only 7% of the receptors were single or paired molecules on microvilli. The majority were on the intervillous membrane, with 95% of those receptors in groups. The receptor groups on the intervillous plasma membrane could be found in both noncoated invaginations and coated pits. The concentration of occupied receptors in the noncoated invaginations and the coated pits was similar; however, ten times more noncoated invaginations than coated pits contained occupied insulin receptors. The observations in this study contrast with those reported on rat adipocytes using identical techniques (Jarett and Smith, 1977). Insulin receptors on adipocytes were initially grouped and randomly distributed over the entire cell surface and did not microaggregate into larger groups. Insulin receptors on rat adipocytes were found in noncoated invaginations but were excluded from the coated pits. The differences in the organization and behavior of the insulin receptor between rat and 3T3-L1 adipocytes suggest that the mechanisms regulating the initial organization of insulin receptors and the aggregation of occupied receptors may be controlled by tissue-specific processes. Since both of these cell types are equally insulin sensitive, the differences in the initial organization and distribution of the insulin receptors on the cell surface may not be related to the sensitivity or biological responsiveness of these cells to insulin but may affect other processes such as receptor regulation and internalization. On the other hand, the microaggregates of occupied receptors on both cell types may relate to biological responsiveness.     

Effect of cytochalasins on the surface topography of neoplastic cells in suspension

The effect of cytochalasins B or D on the surface topography of mouse neoplastic fibroblasts of L line (detached from glass by trypsin-EDTA solution) or of its LS subline (adapted to the growth in suspension in vitro), as well as on that of Ehrlich's ascites tumor cells was investigated by screening electron microscopy. Incubation of suspended cells with cytochalasin B (2 micrograms/ml) or cytochalasin D (0.2 microgram/ml) for 30-180 min led to the following changes: (I) progressive decrease of the proportion of the cells with a microvillous surface relief and simultaneous increase in the percentage of the cells with a blebbed microrelief; (2) shortening of the microvilli and decrease of their density on the cell surface; (3) appearance of surface areas with a rough folded relief; (4) formation of very large blebs on the LS or L cell surfaces; (5) unusual "polar" distribution of blebs on Ehrlich's tumor and L cells: the blebs were concentrated in one locus on the cell surface. The data show that normal organization of the actin microfilament system in the cell cortex is necessary for formation of the microvilli but not for the blebs.     

Transferrin receptor and its recycling in HeLa cells.

The transferrin receptor is a 180 000-dalton protein which can be dissociated to two 90 000-dalton polypeptides under reducing conditions. It can be labelled by lactoperoxidase-catalysed iodination on the cell surface at 0 degree C. Trypsin digestion of labelled cells at 0 degree C can be used to degrade those receptors on the cell surface; they release a 70 000-dalton soluble fragment which binds to transferrin. When cells are labelled at 0 degree C, then warmed to 37 degrees C, the labelled receptors enter the cells and become trypsin resistant. These receptors enter the cells, probably via coated pits, with a half-life of approximately 5 min. Since there is about three times as much receptor inside cells as on the surface, this means that transit through the cell to the cell surface takes approximately 21 min, if all receptors are on the same cycling pathway.     

Comparison of rapid changes in surface morphology and coated pit formation of PC12 cells in response to nerve growth factor, epidermal growth factor, and dibutyryl cyclic AMP

Scanning and transmission electron microscopic studies were carried out on the rapid cell surface response of PC12 pheochromocytoma cells to treatment with nerve growth factor (NGF), epidermal growth factor (EGF), and dibutyryl cyclic AMP. EGF induced a rapidly initiated series of surface changes identical to those previously observed with NGF. Ruffles appear over the dorsal surface of the cells by 30 s, are prominent at 3 min, and are absent by 7 min. Microvilli disappear as dorsal ruffles become prominent. Peripheral ruffles are seen by 3 min, are prominent on most of the cells by 7 min, and are virtually absent by 15 min. Large blebs are present on 50% of the cells by 2 h and are markedly decreased by 4 h. Within 30 s after NGF or EGF addition, an increase in the density of 60-130-nm coated pits per unit membrane is detectable. This reaches a maximum of two- to threefold in from 1 to 3 min and gradually decreases. Combined treatment with NGF and EGF increases surface ruffling and, after an early peak in coated pits which at 3 min is similar in magnitude to that observed for the separately administered factors, maintains a greater number of pits per unit area than either treatment alone. 3-d pretreatment with NGF greatly reduces the response of the cells to EGF both with respect to surface ruffling and coated pit formation while 4-h NGF pretreatment has no effect on the EGF response. Dibutyryl cyclic AMP induced none of the rapidly onsetting changes caused by NGF or EGF, and therefore it seems unlikely that cyclic AMP mediates these surface changes. Changes in cell surface architecture induced by NGF and EGF on PC12 cells and by NGF in normal sympathetic neurons (as previously described) indicates that such responses may be a widespread phenomenon associated with the interaction of at least some peptide growth factors/hormones with their receptors. These responses may represent or reflect primary events in the mechanism by which these factors act.     

Ultrastructure of the full-term shark yolk sac placenta. II. The smooth, proximal segment

The smooth, proximal portion of the yolk sac placenta of the sandbar shark, Carcharhinus plumbeus is comprised of: (1) An outermost epithelial ectoderm; (2) an intervening collagenous stroma; and (3) an inner mesothelium. The surface epithelium may be one to three cell layers thick. The surface epithelium comprises two cell types. A cuboidal cell that has a dome-like apical surface covered with microvilli and an ovoid nucleus predominate. These cells contain lipid inclusions, many cytoplasmic filaments, and are joined by desmosomes. The second cell type has a convoluted nucleus and a flattened cell apex with microvilli, cilia, and paddle cilia. Golgi complexes and elements of the endoplasmic reticulum are relatively uncommon in the cytoplasm of both cell types. Microplicae also occur on the surface of some cells. The smooth, proximal portion of the placenta is sparsely vascularized. The innermost cellular elements of the surface epithelium rest on a prominent basal lamina. A collagenous zone separates the epithelial basal lamina from the basal lamina of the mesothelium. The mesothelial cells are squamous with a fusiform nucleus, many pinocytotic pits and vesicles, and a large number of cytoplasmic filaments. The endoplasmic reticulum, except for occasional patches of the rough type, and the Golgi complex are poorly developed. Ultrastructural tracer studies show that this portion of the placenta does not absorb horseradish peroxidase (HRP) and trypan blue.     

A cobblestone cell isolated from the human omentum: the mesothelial cell; isolation,identification, and growth characteristics

Summary Normal human mesothelial cells (NHMC) were isolated from pieces of human omentum. The cell yield was approximately one million cells per square centimeter omentum. The mesothelial cells were identified by their positive staining with monoclonal antibodies against cytokeratins 6 and 18. Transmission electronmicroscopy of cultured NHMC revealed many microvilli on the apical surface and many mitochondria and pinocytotic vesicles in the cytoplasm, indicating active transmembrane transport. Growth of NHMC was directly related to the concentration of human serum or of fetal bovine serum in the growth medium. Addition of epidermal growth factor with or without hydrocortisone resulted in a significant increase of NHMC growth; when endothelial cell growth factor, insulin, or hydrocortisone were added no such increase was observed. Seeding NHMC at densities less than 3000/cm² did not result in monolayer formation. The mesothelial cells were serially passed in growth medium M199 with added 10% fetal bovine serum up to 7 passages. However, after Passage 4 the cells changed into giant cells with an irregular pattern, and a lack of intracellular cytokeratin expression was observed for most of the cells.     

Cholinergic stimulation of chromaffin cells induces rapid coating of the plasma membrane

Stimulation of primary bovine adrenal chromaffin cells by carbachol produced a 6-fold increase in cell surface coated pits within 30 s. This coat appeared not to be recruited from a preformed pool at the plasma membrane, but from some pool transparent to electron microscopy. The number of coated pits appeared to decrease rapidly after 1 to 2 min stimulation, but processing for electron microscopy using tannic acid to enhance contrast indicated that both coated pits and closed coated vesicles were increased relative to unstimulated cells for up to 30 min. Analyses of purified adrenal medulla coated vesicles showed a lipid composition close to that expected for cell surface membrane, but there were only trace levels of plasma membrane marker enzymes. Coated vesicles contained significant amounts of both membrane and content proteins characteristic of the chromaffin granule, suggesting that medulla coated vesicles preferentially carry secretion granule proteins. The kinetics of stimulus-dependent formation of coated membrane in the cortical zone of chromaffin cells is closely similar to that observed for secretion granule membrane retrieval.     

Gravitational and magnetic activation methods applied to cultured cells (LK 35.2)

Monoclonal antibody 10.2-16 is directed toward the mouse class II major histocompatibility complex gene product 1-Ak expressed on the cell line LK35.2. Instead of activating cells by fluorophor we used (acrylamide-coated) heavy and magnetic microspheres of 0.6 micron in radius. These microspheres are chemically coupled (carbodiimide method) with the antibody toward the surface antigen. The cells are observed through a microscope with horizontal alignment, as they sediment in a (temperature controlled) tube with square cross-section. Stokes Law allows the determination of the density of cells (first alone) using viscosity and density of Dulbecco's modified Eagle's Medium together with the observed mean sedimentation velocity (66 microns/min) and a mean diameter of 10 microns. We found a density of 1.0558 +/- 0.0028 g/cm3 at 10 degrees C. Independently, thinly coated, heavy (and magnetizable) microspheres with the cited antibody are attached to cells and observed likewise. The increased sedimentation velocity permits us to show that the cells were fully covered with microspheres (290 per cell). A magnetic field gradient opposing gravity moved these cells against gravity with two different mean velocities, 340 microns/min and 850 microns/min. The higher velocity resulted in 290 particles per cell, the lower one in 130 particles per cell. The limits for the expansion of this method to smaller particle sizes (down to 10 nm) are evaluated.     

Infectious entry pathway of influenza virus in a canine kidney cell line

The entry of fowl plague virus, and avian influenza A virus, into Madin- Darby canine kidney (MDCK) cells was examined both biochemically and morphologically. At low multiplicity and 0 degrees C, viruses bound to the cell surface but were not internalized. Binding was not greatly dependent on the pH of the medium and reached an equilibrium level in 60-90 min. Over 90% of the bound viruses were removed by neuraminidase but not by proteases. When cells with prebound virus were warmed to 37 degrees C, part of the virus became resistant to removal b neuraminidase, with a half-time of 10-15 min. After a brief lag period, degraded viral material was released into the medium. The neuraminidase- resistant virus was capable of infecting the cells and probably did so by an intracellular route, since ammonium chloride, a lysosomotropic agent, blocked both the infection and the degradation of viral protein. When the entry process was observed by electron microscopy, viruses were seen bound primarily to microvilli on the cell surface at 0 degrees C and, after warming at 37 degrees C, were endocytosed in coated pits, coated vesicles, and large smooth-surfaced vacuoles. Viruses were also present in smooth-surfaced invaginations and small smooth-surfaced vesicles at both temperatures. At physiological pH, no fusion of the virus with the plasma membrane was observed. When prebound virus was incubated at a pH of 5.5 or below for 1 min at 37 degrees C, fusion was, however, detected by ferritin immunolabeling. t low multiplicity, 90% of the prebound virus became neuraminidase- resistant and was presumably fused after only 30 s at low pH. These experiments suggest that fowl plague virus enters MDCK cells by endocytosis in coated pits and coated vesicles and is transported to the lysosome where the low pH initiates a fusion reaction ultimately resulting in the transfer of the genome into the cytoplasm. The entry pathway of fowl plague virus thus resembles tht earlier described for Semliki Forest virus.     

Cell surface glycoproteins of CHO cells. II. Surface distribution and pathway of internalization

The surface distribution and pathway for internalization of the major cell surface proteins of Chinese hamster ovary (CHO) cells have been investigated after reacting cells at 4 degrees C with the membrane-impermeant reagent trinitrobenzenesulfonate. Molecules, haptenized with trinitrophenol groups, the majority of which are in a group of high molecular weight acidic glycoproteins (HMWAG), were labelled at 4 degrees C with anti-dinitrophenol immunoglobulins coupled to fluorescein isothiocyanate (FITC), horseradish peroxidase, or colloidal gold and either immediately fixed for mapping their distribution or followed intracellularly after warming to allow endocytosis to proceed. The distribution of label on the CHO cell surface was non-random with a large proportion arranged in clusters from 100 to 300 nm in diameter. Antibody label was concentrated heavily on microvilli, and about 10% of the molecules were always associated with clathrin-coated pits. Upon warming the cells to 37 degrees C, HMWAG were internalized immediately into smooth-membraned tubules (less than 80 nm luminal diameter) that appeared to connect with vesicles (less than 300 nm luminal diameter) located in the cortical cytoplasm. By 60 min, labelled antibody was located within larger vesicles (greater than 300 nm luminal diameter) that had a morphology characteristic of multivesicular bodies and not lysosomes. There was no evidence for entry of labelled molecules into either electron-dense, secondary lysosomes or into the Golgi cisternae, suggesting that neither compartment is involved in the major pathway of cell surface endocytosis. Our results are consistent with the view that the majority of plasma membrane protein are internalized as small discrete domains by a pathway very similar to that described by others for adsorptive endocytosis.     

Rapid, sequential changes in surface morphology of PC12 pheochromocytoma cells in response to nerve growth factor

The effect of nerve growth factor (NGF), a substance that promotes the differentiation and maintenance of certain neurons, was studied via scanning electron microscopy utilizing the PC12 clonal NGF-responsive pheochromocytoma cell line. After 2-4 d of exposure to NGF, these cells acquire many of the properties of normal sympathic neurons. However, by phase microscopy, no changes are discernible within the first 12-18 h. Since the primary NGF receptor appears to be a membrane receptor, it seemed likely that some of the initial responses to the factor may be surface related. PC12 cells maintained without NGF are round to ovoid and have numerous microvilli and small blebs. After the addition of NGF, there is a rapidly initiated sequential change in the cell surface. Ruffles appear over the dorsal surface of the cells with 1 min, become prominent by 3 min, and almost disappear by 7 min. Microvilli, conversely, disappear as the dorsal ruffles become prominent. Ruffles are seen at the the periphery of cell at 3 min, are prominent on most of the cells by 7 min and are gone by 15 min. The surface remains smooth from 15 min until 45 min when large blebs appear. The large blebs are present on most cells at 2 h and are gone by 4 h. The surface remains relatively smooth until 6-7 h of NGF treatment, when microvilli reappear as small knobs. These microvilli increase in both number and length to cover the cell surface by 10 h. These changes were not observed with other basic proteins, with α-bungarotoxin (which binds specifically to PC12 membranes), and were not affected by an RNA synthesis inhibitor that blocks initiation of neurite outgrowth. Changes in the cell surface architecture appear to be among the earlist NGF responses yet detected and may represent or reflect primary events in the mechanism of the factor’s action.     

Mesothelial cells activate the plasma kallikrein-kinin system during pleural inflammation

Abstract Pleural inflammation underlies the formation of most exudative pleural effusions and the plasma kallikrein-kinin system (KKS) is known to contribute. Mesothelial cells are the predominant cell type in the pleural cavity, but their potential role in plasma KKS activation and BK production has not been studied. Bradykinin concentrations were higher in pleural fluids than the corresponding serum samples in patients with a variety of diseases. Bradykinin concentrations did not correlate with disease diagnosis, but were elevated in exudative effusions. It was demonstrated, using a range of primary and transformed mesothelial and mesothelioma cell lines, that cells assembled high molecular weight kininogen and plasma prekallikrein to liberate bradykinin, a process inhibited by novobiocin, a heat shock protein 90 (HSP90) inhibitor, cysteine, bradykinin and protamine sulphate. Of the common plasma prekallikrein activators, mesothelial cells expressed HSP90, but not prolylcarboxypeptidase or Factor XII. Calcium mobilisation was induced in some mesothelium-derived cell lines by bradykinin. Des-Arg(9)-bradykinin was inactive, indicating that mesothelial cells are responsive to bradykinin, mediated via the bradykinin receptor subtype 2. In summary, pleural mesothelial cells support the assembly and activation of the plasma KKS by a mechanism dependent on HSP90, and may contribute to KKS-mediated inflammation in pleural disease.     

Changes in cell surface primary cilia and microvilli concurrent with measurements of fluid flow across the rabbit corneal endothelium ex vivo

Primary cilia and microvilli have been reported on the mammalian rabbit corneal endothelium but their relationship to cell function is undefined. Six corneas from healthy 2 kg female albino rabbits were glutaraldehyde-fixed post mortem (15:00 h) or twelve corneal stroma-endothelial preparations incubated at 37 degrees C under an applied hydrostatic pressure of 20 cm H2O for 4 h prior to fixation. The corneal endothelium was assessed by quantitative scanning electron microscopy. Cells fixed immediately post mortem were decorated with small stubby microvilli (average 21 +/- 13/100 micron 2), and only 25% of the cells were decorated with primary cilia having an average length of 2.44 +/- 1.56 microns. Following 4 h ex vivo incubation with a phosphate-buffered Ringer solution, conspicuous microvilli developed to an average density of 40 +/- 19/100 micron 2 and primary cilia were found on 12% of the cells, having on average length of 2.27 +/- 1.38 microns. Following 4 h incubation in a bicarbonate-buffered Ringer solution, small stubby microvilli developed to a density of 49 +/- 18/100 micron 2, and 40% of the cells showed primary cilia with an average length of 4.31 +/- 1.93 microns; the net trans-endothelial fluid flow in the latter set was 60% greater. These studies indicate that the primary cilia on corneal endothelial cells might be responsive to fluid flow, but that mild mechanical and/or chemical stress could also be the cause of the change since the elaboration of primary cilia can be accompanied by microvilli as well.     

Enucleation of phagocytic cells with adenine, guanine, and their nucleosides in combination with centrifugation

Several purine compounds, such as adenine, guanine, adenosine, guanosine, and their related compounds, exhibited enucleation activity on adherent mouse peritoneal exudate cells (macrophages) during centrifugation at 25,000 and 35,000 g for 60 min at 34 degrees-36 degrees C in medium containing one of these compounds. Enucleation activity, however, did not occur in cells treated with adenine nucleotides, inosine, xanthine, or any of the tested pyrimidines. The purine compounds also had enucleation activity on mouse macrophage-like cell lines (P388D1 and RAW 264) and mouse polymorphonuclear leukocytes, but not on other typical cell lines such as a human epithelial cell line (HeLa S-3) or a mouse fibroblast cell line (L929). Cytochalasin B (CB) treatment, however, resulted in the enucleation of all cell types tested, even at a centrifugal force as low as 5,000 g. The process of macrophage enucleation was observed by both light microscopy and scanning electron microscopy. In enucleated macrophages that had been treated with purine compounds, but not with CB, a newly formed cytoplasmic crater-like structure (about 3-9 microns in diameter) was observed at the original site of the nucleus. Surface structures, such as microvilli and membrane ruffles, remained relatively intact in macrophages that had been enucleated by treatment with purine compounds. By contrast, these surface structures were markedly changed in CB-treated macrophages. Purine compounds may affect cytoskeletal elements in ways similar to the well characterized effects of CB, and thus result in the enucleation of phagocytes. However, the characteristic differences in the enucleation activity exhibited by purine compounds and CB may indicate that purines have a mechanism of action different from that of CB.     

Conditionally immortal ovarian cell lines for investigating the influence of ovarian stroma on the estrogen sensitivity and tumorigenicity of ovarian surface epithelial cells

The tendency of the ovarian surface epithelium (OSE) to undergo metaplastic and morphogenetic changes during the life cycle, at variance with the adjacent peritoneal mesothelial cells, suggests that its biology may be regulated by underlying ovarian stromal cues. However, little is known about the role that the ovarian stroma plays in the pathobiology of the OSE, largely because of the lack of a suitable in vitro model. Here, we describe the establishment and characterization of conditionally immortalized ovarian stromal and surface epithelial cell lines from H-2K(b)-tsA58 transgenic mice that carry the thermolabile mutant of SV-40 large T antigen under the control of an interferon-gamma (IFN-gamma)-inducible promoter. These cells express functional T antigens, grow continuously under permissive conditions at 33 degrees C in the presence of IFN-gamma, and stop dividing when the activity and expression of the tumor antigen is suppressed by restrictive conditions without IFN-gamma at 39 degrees C. Morphological, immunohistochemical, and ultrastructural analyses show that conditionally immortal OSE cells form cobblestone-like monolayers, express cytokeratin and vimentin, contain several microvilli, and develop tight junctions, whereas stromal cells are spindle-like, express vimentin but not cytokeratin, and contain rare microvilli, thus exhibiting epithelial and stromal phenotypes, respectively. At variance with the reported behavior of rat epithelial cells, conditionally immortal mouse epithelial cells are not spontaneously transformed after continuous culture in vitro. More importantly, conditioned media from stromal cells cultured under permissive conditions increase the specific activity of the endogenous estrogen receptor in BG-1 human ovarian epithelial cancer cells and promote these cells' anchorage-independent growth, suggesting the paracrine influence of a stromal factor. In addition, stromal cells cultured under restrictive conditions retain this growth-stimulatory activity, which, therefore, appears to be independent of T antigen expression. These established cell lines should provide a useful in vitro model system for studying the role of cellular interactions in OSE cell growth and tumorigenesis.