Phagocytosis of concanavalin a in normal and enucleated cultures of mammalian cells

Enucleated mammalian cells in culture are similar to their normal counterparts in that they can attach to a substratum (Prescott, Myerson & Wallace, 1972; Goldman, Pollack & Hopkins, 1973), respond in the same way as enucleated cells to agents which cause shape changes (Schröder & Hsie, 1973), and maintain their surface concanavalin A-binding sites (CABS) for at least 24 h after enucleation (Wise & Larsen, 1976). With respect to the presence of the CABS on the enucleated cells, questions arise as to whether or not the sites can be internalized as in the case of lymphocytes (e.g. Barat & Avrameas, 1973; Unanue, Perkins & Karnovsky, 1972) or polymorphonuclear leukocytes (Oliver, Ukena & Berlin, 1974). Moreover, if they are phagocytozed in the enucleated cell one can ask how this compares to the manner and rate of phagocytosis in normal, nucleated cells.

To answer these questions, normal and enucleated cells were incubated in Con A for 30 min (pulse) and then placed in fresh medium for varying amounts of time before being fixed for electron microscopy. Detection of the fate of the bound Con A by incubating the fixed cells in horseradish peroxidase suggests that the Con A and presumably the sites to which it is bound are phagocytozed in a similar manner in both normal and enucleated mouse L cells. Thus, the immediate presence of a nucleus is not necessary for the phagocytosis of surface CABS and possibly is not needed for the turnover of the CABS.

Cultures of mouse L cells were grown to near confluency on round glass coverslips at 37 °C in a mixture of Ham’s F-12 medium and 10% calf serum. For enucleation experiments, the coverslips were inverted, placed into centrifuge tubes containing 10 μg cytochalasin B/ml of medium and centrifuged for 20 min at 1 o 000 rev /min in an RC-2B Sorvall centrifuge prewarmed to 37 °C. This technique devised by Prescott et al. (1972) results in the nuclei and their surrounding thin rim of cytoplasm being pulled to the bottom of the tube whereas the resulting enucleated cells remain attached to the coverslips. The coverslips were then removed from the tubes, rinsed in Earle’s Basic Salt Solution (EBSS) and placed in fresh medium devoid of cytochalasin B. Within 30 min to 1 h the enucleated cells had regained their normal fibroblast morphology.

After enucleates had been in fresh medium for 1 h following enucleation, they were rinsed in EBSS and then placed in a solution of 10 μg Con A/ml EBSS for 30 min. The cells were then rinsed in EBSS and some of them were fixed for electron microscopy (pulse) whereas the remainder were placed in fresh nutrient medium for either 2 or 20 h before being fixed for electron microscopy (chase). Normal, nucleated cells were concurrently processed in the same manner.

All cells were fixed for 45 min in 2 % glutaraldehyde in 01 M cacodylate buffer (pH 7·4). Following fixation they were rinsed 3 times in cacodylate buffer and placed in a solution of 1 mg horseradish peroxidase/ml of 0·1 M cacodylate buffer for 45 min. They were then rinsed 3 times in buffer and incubated for 20 min in a solution of 10 mg of 3,3’-diaminobenzidene (DAB) in 20 ml of Tris-HCl buffer (pH 7·6) and 0·02 % H₂O₂ (Graham & Karnovsky, 1966).

Following incubation in DAB, the cells were rinsed 3 times in cacodylate buffer and postfixed 1 h in 1·0 % OsO₄. They were then dehydrated in a graded series of ethanols, transferred to propylene oxide and embedded in Araldite 6005. Sections were cut on a Porter-Blum MT2B ultramicrotome, stained with uranyl acetate and lead citrate and examined with an AEI-801 electron microscope.

The Con A binds to the glucose or mannose residues on the cell surface but because it has 4 binding sites, other sites on the molecule can bind to the glycoprotein horseradish peroxidase. Rendering the peroxidase electron-opaque by DAB treatment then enables one to visualize by electron microscopy where the Con A-binding sites are on the cell.

To ensure that the peroxidase is binding to the Con A, the cells (normal and enucleates) were incubated only in peroxidase and DAB with no prior treatment of Con A. Other controls utilized adding a sugar, D-mannose, which is a competitive inhibitor of Con A binding (Hirano et al. 1972). D-mannose was added to both the incubation mixtures of Con A and peroxidase.

As reported in a previous paper (Wise & Prescott, 1973) the enucleated cell initially appears to have a normal complement of cytoplasmic organelles after enucleation. For example, a 2-h enucleate contains a morphologically normal Golgi apparatus, mitochondria and rough endoplasmic reticulum (Fig. 1) and the cytoplasm essentially is indistinguishable from that of a normal, nucleated L cell. In the 20-h enucleates the morphology is altered such that the Golgi apparatus is either absent or disrupted and the cell usually has numerous vacuoles (Wise & Prescott, 1973).

Following a 30-min pulse with Con A, the enucleated or normal L cells appear to have bound the Con A uniformly over their entire surface (Fig. 2) as indicated by the presence of the oxidized DAB reaction product. The staining of the surface is light, primarily because of the low concentration of Con A used in the experiments. Occasionally after the 30-min pulse, some cellular invaginations (phagocytic channels) are observed with an increased amount of reaction product at the base of the channel (Fig. 3).

After a 2-h chase in medium devoid of Con A, the surfaces of either enucleated or nucleated ceUs display a marked reduction of surface staining (Fig. 4). Reaction product is still observed on the surface but it is now discretely localized in clusters or patches on the membrane (Fig. 5). Moreover, following a 2-h chase and especially after a 20-h chase, phagocytic vacuoles are observed which contain reaction product (Fig. 6).

In about 2 % of the observed normal or enucleated cells some reaction product is present as a large patch on the cell surface (Fig. 7) even after a 20-h chase. Whether this represents an inability of the cell to phagocytoze the material or simply a failure to wash out some unbound peroxidase trapped in the cellular processes is not known. Perhaps it is even material that has been released back to the surface by exocytosis.

In control cells incubated with D-mannose, the staining reaction was greatly reduced (Fig. 8). In cells treated with peroxidase but not incubated in Con A, the cell surface was not stained although there was some variable staining of ribosomes and RER, as has been previously reported (Wise & Larsen, 1976).

The results of this study show that enucleated cells are capable of internalizing bound Con A in a manner similar to their normal, nucleated counterparts. Following exposure to the Con A for 30 min, the cells have phagocytozed much of the bound Con A by 2 h after being placed in fresh medium devoid of Con A.

Although the inherent distribution of the sites in normal and enucleated L cells is initially random (Wise & Larsen, 1976), incubating live cells in Con A followed by a chase period resulted in the lectin becoming concentrated into patches or clusters on the cell surface during the chase. The lectin and its sites then were phagocytozed. In contrast, lymphocytes or polymorphonuclear lymphocytes aggregate Con A into a cap before ingesting it (e.g. Unanue et al. 1972; Oliver et al. 1974).

The fate of bound Con A after it was localized in the phagocytic vacuoles could not be traced from them; i.e. even after a 20-h chase some vacuoles were still stained. In 3T3 cells in which the ingestion of another lectin, ricin, was followed (Nicolson, 629 1974), the lectin was also incorporated into phagocytic vesicles. However, breaks were then observed in the vacuole membranes and it was postulated that the escape of the ricin through these gaps resulted in the toxic effect exerted by ricin on these cells. In the L cells in this study, however, the phagocytic vacuole membranes were usually intact.

Because phagosomes usually become secondary lysosomes one would assume that Con A is degraded in the vacuole if primary lysosomes fuse with the vacuole. This view is supported by Goldman & Raz (1975) who reported that Con A induces vacuole formation in mouse peritoneal macrophages and that there is a significant increase in acid phosphatase activity in the cells as early as 1 h after exposure to Con A. However, Edelson & Cohn (1974a) reported no acid phosphatase activity in these vacuoles in macrophages. Moreover, the half fife of internalized proteins such as horseradish peroxidase was prolonged (Edelson & Cohn, 1974i). Since phagocytic vacuoles were still observed in the L cells in this study after a 20-h chase, the possibility exists that they have not become secondary lysosomes and will not degrade the Con A.

The answer about the fate of the vacuoles containing Con A simply may be that the phagocytic vacuoles release their product back to the cell exterior by exocytosis. Thus, the reaction product occasionally seen in a patch on the 20-h chased cells may have originated by fusion of a labelled vacuole with the cell membrane. Support for this possibility comes from the work on lymphocytes by Barat & Avrameas (1973) in which they suggested that the ingested Con A is released from the cell by exocytosis after incubation in Con A for 3 h. Without a subsequent chase period, however, it is difficult to determine the direction of flow of the labelled vacuoles within the lymphocytes.

Con A at a concentration of 50 μg/ml or more is toxic to some cells (Shoham, Inbar & Sachs, 1970)· In this study, however, a low concentration of Con A (10 μg/ml) was used to avoid this problem. The fact that the cells incubated in Con A remained attached to the substrate and maintained the same morphology (surface and internal) as did cells not exposed to Con A suggested that the concentration of Con A utilized was not toxic.

Although it has not yet been determined whether new surface CABS are added following internalization of existing surface CABS, previous studies have shown that enucleated cells can be fixed any time up to 24 h after enucleation and the surface CABS are present (Wise & Larsen, 1976). Thus, if there is normal phagocytosis of the surface CABS regardless of the presence or absence of Con A, the sites are being turned over because they can always be detected at the surface. Turnover would then be happening in the absence of a nucleus and the absence of a Golgi apparatus because the Golgi is either greatly morphologically distorted or absent by 24 h after enucleation (Wise & Prescott, 1973).

This study thus adds one more to the list of functions an enucleated cell is able to accomplish. This nuclear independence of the cell surface in terms of its maintenance and phagocytosis of surface CABS is surprising in light of the role the sites may play in signalling the nucleus to undergo mitosis (Burger & Noonan, 1970). One might suppose that removal of the nucleus would have an immediate effect upon the sites but apparently it does not. Consequently, these experiments are perhaps a complementary reciprocal to the studies which indicate that exposure of more sites by protease treatment does not stimulate cell growth (Glynn, Thrash & Cunningham, 1973) and to the experiments which indicate that covering of the sitesbysuccinylatedConA does not inhibit cell growth (Trowbridge & Hilborn, 1974).

The author is indebted to the Department of Biological Structure for its support of this study and to Joy Harrison for her skilled technical assistance.