share (Bashford et al, 1984). In the present investigation the action of haemolytic pore-formers at non-lytic concen- trations was studied using Lettre cells, ...
BIOCHEMICAL SOCIETY TRANSACTIONS
708 virus (Huang et al., 1981; Lenard & Miller, 1981) at acid pH; (2) toxins such as the a-toxin of Staphylococcus aureus (Arbuthnott, 1982; Bhakdi & Tranum-Jensen, 1984) and the bee venom protein melittin (Tosteson & Tosteson, 198 I); (3) activated complement (Mayer, 1972; Bhakdi & Tranum-Jensen, 1983). Although the details of pore-formation by each agent differ, there are a number of features of the pore-forming process which all the agents share (Bashford et al, 1984). In the present investigation the action of haemolytic pore-formers at non-lytic concentrations was studied using Lettre cells, a non-erythroid, mouse acites tumour line derived from a transplantable mammary carcinoma (Lettre et al., 1972), from which were recorded: plasma-membrane potential using oxonol-V as the indicator (Bashford & Pasternak, 1984; Bashford et al., 1 9 8 5 ~ ) the ; cellular cation and [3H]phosphorylcholine content after pelleting the cells through oil (Impraim et al., 1980; Bashford et a/., 1983). The effect of Sendai virus and complement on the cation composition and [3H]phosphorylcholine content of labelled cells suspended in hysiological saline are illustrated in Figs. I(a) and 1 6 ) (open symbols). The leakage of Na+ into, and K + and [3H]phosphorylcholine from, cells is presented as the percentage of that found from fully permeabilized cells. The onset of ion and phosphorylcholine leakage induced either by virus or by complement is preceded by a lag (Micklem et al., 1985; Bashford et al., 1985b), as is the onset of depolarization (data not shown; see also Bashford et al. 1984, 1985b, Pasternak et al., 1985a,b). Leakage of ions and ghosphorylchline continues after the cells are cooled to 4 C (Fig. 1, halffilled symbols), a temperature at which new pores are not induced/inserted. The leakage at low temperature is sensitive to the presence of Zn2+ (Fig. 1, filled symbols). Zn2+ blocks the leakage induced by toxins, virus and complement in a manner similar to that reported for Ca2+ (Pasternak et al., 1985a,b). Significantly, removal of ZnZ+ by chelation with EGTA at 4'C stimulates leakage from treated cells (Fig. I). These observations suggest that one of the actions of ZnZ+ is to obstruct the pores created by viruses, toxins or complement. Ca2+ and Mg2+ can also antagonize the effects of these agents on cell permeability but they are only effective at higher concentrations: 50% inhibition of leakage is found at lo-'M-Zn2+,at around M-CaZ+and at around lo-' M Mg*+: In addition, each agent has a different susceptibility to inhibition by divalent cations, Sendai virus being most sensitive and a-toxin the least (Pasternak et al., 1985a,b). It is of interest that CaZ+and Zn2+ antagonize pore-formation at concentrations similar to those found in body fluids (Pasternak et al., 1985b), and it is possible that such cations may have a role in the pathophysiology of diseases caused by the agents used here.
A simple scheme suggesting a common pathway for pore-formation, at least in terms of the consequences of pore-formation, is shown in Fig. l(c). In the case of each agent leakage follows events at the cell surface, such as binding, followed by the accumulation of a sufficient number of potentially leaky viral envelopes fused with the cell membrane (Micklem et al., 1985) or the insertion of pore-forming polypeptides into the membrane. The observation that larger molecules leak more slowly than, and after a longer lag than, smaller molecules suggests that the size of the pores increases with time. Under some conditions pores of sufficient size are induced which allow small proteins to leak from cells (Bashford el al.. 1985b); such leakage is substantially less than that of the smaller molecules and ions. Divalent cations are able to inhibit the leakage of all the molecules and ions we have measured, suggesting that they can block pores regardless of their state of development. We thank Mrs. E. Penfold and Mrs. B. Bashford for preparing the typescript,and the Cell Surface Research Fund for financial support. Arbuthnott. J . P. (1982) in Molecular Action o f Toxins and Viruses (Cohen, P. & van lleyningcn, S., eds.), pp. 107-129, Elsevier Biomedical Press, Amsterdam Bashford, C. L. & Pasternak, C. A. (1984) J. Memhr. Biol. 7 9 , 275-284 Bashford, C. L., Alder, G., Micklern, K . J . & Pasternak, C. A. (1983) Biosci. Rep. 3, 631-642 Bashford, C. L., Alder, G. M., Patel, K . & Pasternak, C. A. (1984) Biosci. Rep. 4,797-805 Bashford, C. L., Alder, G. M., Gray, M . A . , Micklem, K. J . , Taylor, ) Cell. Physiol. C. C., Turek, P. J . & Pasternak, C. A. ( 1 9 8 5 ~ J. 123, in thc press Bashford, C. L., Micklern, K . J . & Pasternak, C . A. (19856) Biochim. Biophys. Acta in the press Bhakdi, S. & Tranum-Jensen, J . (1983) Biochim. Biophys. Acta 737.343-372 Huang, R. T. C., Roth, R . & Klenk, H.-D. (1981) Virology 110, 243-241 Impraim, C . C., Foster, I95%) of band 4.5 polypeptides, and detergent-free material was shown to be capable o f reversibly binding nitrobenzylthioinosine (NBMPR). BALB/c mice were injected intraperitoneally with the immunogen (30pg) and boosted three times 13, 24 and 27 days after the initial injection. One day after the last injection, mice were killed and spleen cells were fused with myeloma P3/NSI/l-Ag4-1 cells (Kohler & Milstein, 1976; GalfrC & Milstein, 1981). Culture supernatants from hybridoma clones were initially screened for antibodies against protein-depleted human erythrocyte membranes using an enzyme-linked immunosorbent assay system and those that were positive were then screened for antibodies that immunoprecipitated nucleoside or glucose 1 1 I L transporter proteins (Fig. 1) or that inhibited nucleoside 0 60 1 20 transport by human erythrocytes. Two immunoglobulin G (I&)-producing hybridomas (64C7 and 65D4) were selected for further study and were subcloned twice. Antibodies used in this report were isolated from ascites Fig. 1. I m m u n o p r e c i p i t a t i o n of the nucleoside a n d glucose fluid or from cell-culture supernatants and were further transporters purified by ammonium sulphate precipitation and protein ‘Protein-depleted’ human erythrocyte membranes were A-Sepharose chromatography. equilibrated with either 60 nM-[ 3 H ] NBMPR, supplemented A direct interaction between the antibodies and transporter proteins was demonstrated by immunoprecipi- with dithiothreitol (final concentration 50 mM) or 2pM[ 3H] cytochalasin B, and exposed t o high-intensity U . V . tation of solubilized human erythrocyte membranes photolabelled with either [3H]NBMPR or [3H]cytochalasin B, light from a 450 W mercury arc lamp (Wu e t a l . , 1983) for specific inhibitors of nucleoside and hexose transport, 45 s. Non-covalently bound radioactivity was removed by washing with 50mM-Tris/HC1 (pH 7.5) and the washed respectively (Fig. 1). Band 4.5 proteins are selectively labelled by both ligands (Carter-Su er a f . , 1982; Wu er af., membrane pellets were solubilized with 1% Triton X-100. Non-solubilized membrane was removed by centrifugation 1983). Antibody 65D4 immunoprecipitated both [3H](lOOOOOg, 1 h ) and t h e solubilized membranes were NBMPR- and [3H]cytochalasin B-labelled membranes incubated with purified 6584 antibody ( 0 , 0) or non(see Fig. 1). Similar results were obtained with antibody immune mouse IgG (A, A) in a medium containing 1%64C7 (data not shown). In contrast, polyclonal mouse IgG prepared from non-immunized animals did not immuno- Triton X-100, 100 mM-NaC1 and 50 mM-Tris/HCl (pH 7.5). After 18 h, 1 0 0 ~ of 1 Pansorbin was added for 1 h at 4OC precipitate radioactivity. Electroblotting of human erythrocyte membrane proteins to nitrocellulose confirmed and the immunprecipitates were washed twice by centifugation at 10 OOOg with 50 mM-Tns/HCl (pH 7.5) that both antibodies reacted with band 4.5, although containing 0.1% Triton X-100. Radioactivity associated antibody 64C7 also reacted with proteins with an apparent MI of 70000-90000. Control experiments with non- with the immunoprecipitates was determined by liquid scintillation counting 0 , A, [ 3H] NBMPR-labelled immune mouse IgG did not label any protein bands. 0,A [3H]cytochalasin B-labelled membranes. membranes; Additional elect rob lott ing studies with t rypsin-t reat ed ‘ghosts’ demonstrated that antibody 65D4 reacted with a broad band (apparent M , 40 000-26 000) whereas 64C7 reacted with a sharp 50000 MI band. Endoglycosidase F This work was funded by grants from the Medical Kesearch treatment of partially purified band 4.5 proteins resulted Council o f Canada. S. M. 1. is a Scholar of the Alberta Heritage in conversion of the diffuse band t o a sharp band at 46 000. Foundation for Medical Research. C. E. C. and A. R. P. P are Subsequent electroblotting experiments demonstrated that Research Associates of the National Cancer Institute of Canada. antibodies 64C7 and 65D4 labelled a sharp peak corresponding t o MI 46 000 after endoglycosidase F treatment, suggesting that the antibodies were not reacting Carter-Su, C., Pessin, J. E., Mora, K., Gitorner, W. & Czech, M. P. (1982) J. Biol. Chem. 257,5419-5425 with carbohydrate residues. In conclusion, antibodies 64C7 and 65D4 appear to GalfrB, G. & Milstein, C. (1981) Methods Enzymol 73, 3-46 react with both the nucleoside and glucose transporters Kohler, G. & Milstein, C. (1976) Eur. J. Immunol 6 , 511-519 although the site of recognition on the protein is probably Jarvis, S. M. & Young, J . D. (1981) Biochem J. 194, 331-339 A. R. P., Kolassa, N. & Cass, C. t. (1981) Pharmacol different for the two antibodies. Preliminary experiments Paterson, Ther. 12,515-536 suggest that treatment of human erythrocytes with anti- Tse, C.-M., Belt, J . A., Jarvis, S. M., Paterson, A. R. P., Wu, J-S. R. bodies 64C7 and 65D4 did not inhibit (a) high-affinity &Young, J . D. (1985)J. Biol. Chem. 260,3506-3511 binding of ’H NBMPR or ( b ) zero-trans influx of [3H]- Widdas, W. F. (1980) Curr. Top. Membr. Tramp. 14, 165-223 uridine or [ 5Hladenosine. 1 ’ Further studies are in progress t o Wu, J.-S. R . , Kwong, €.. Y. P., Jarvis, S. M. & Young, J . D. (1983) attempt to raise a monoclonal antibody specific for the J. Biol. Chem. 258,13745-13151 Young, J . D. & Jarvis, S. M. (1983) Biosci. Rep. 3, 309-322 nucleoside transporter.
P
I
Vol. 13
