Immunology Letters, 33 (1992) 1-8
0165 - 2478 / 92 / $ 5.00 © 1992 ElsevierSciencePublishers B.V. All rights reserved IMLET 01797
Dll, a novel monoclonal antibody specific for human mature macrophages and peripheral blood monocytes T . D . R u d i n s k a y a a, V.S. P o l t o r a n i n a a, V , N . B a r a n o v a, N . N . P e t r o v i c h e v b, B.O. V o y t e n k o v c a n d L . O . T r e t y a k o v a aLaboratory of Immunochemistry and bLaboratory of Clinical Cytology, Cancer Research Centre, Russian Academy of Sciences, Moscow, Russia, and ~Laboratory of Oncoimmunology, Institute of Oncology, St.-Petersburg, Russia
(Received 4 March 1992; accepted 18 March 1992)
1. Summary
2.
A new monoclonal antibody designated Mab D ~ is described, which shows a restricted reactivity to cells of the monocyte/macrophage system. When tested by light and electron microscopic immunoperoxidase methods, Mab D~l specifically reacts with blood monocytes and stains resident macrophages in a wide variety of human tissues; it does not mark the macrophages of other species, i.e., rat, swine and mouse. Antigen-presenting cells, e.g., Langerhans cells, are Mab Dll negative. Mab Oil reveals the antigen on cryostat and paraffin tissue sections. Ultrastructurally the antigen recognized by Mab Dll in all macrophage types studied is located on the plasma membrane and within cytoplasmic structures including lysosomes. On immunoblotting, Mab D~I detects the 125-kDa antigen in human liver and the 135-kDa protein in tumours of histioeytic origin. The similarity of Mab Dl~ to known "panmacrophage" monoclonal antibodies is discussed.
The recent advances in the study of cells of the monocyte/macrophage system are mainly due to obtaining a large panel of monoclonal antibodies that recognize not only universal antigenic markers of phagocytic cells but also identify the specific determinants of mature and mediated cell forms during the differentiation of phagocyte cell lineage [1 3]. Most of the known monoclonal antibodies specific to mononuclear phagocytes are reactive with tissue macrophages as well as with monocytes, thus revealing that the antigens have been expressed in the earliest stages of phagocyte differentiation [4-7]. Of special interest among these monoclonal specificities are antibodies that react strongly with the majority of macrophages in a number of organs, and which for this reason are classified as "pan-macrophage" antibodies. According to results presented at the Fourth International Conference on Human Leucocyte Differentiation Antigens, "pan-macrophage" antibodies have the following specificities (CD 68 reagents): E B M l l [8], y 2/131 [9], y 1/82A [10], Ki-M6 [11], KP 1 [12], obtained in different laboratories. In the present report we describe a new antimacrophage monoclonal antibody, Dll (Mab D1~) that fulfils the criteria for "pan-macrophage" reagent. Mab D ~ labels all types of human tissue macrophages to a high degree and also reacts with freshly isolated peripheral blood
Key words: Monoclonal antibody; Human macrophage;
Monocyte; Light and electron microscopic immunolocalization Correspondence to." Dr. T.D. Rudinskaya, Laboratory of Immunochemistry, Cancer Research Centre, 115478, Moscow, Kashirskoye shosse 24, Russia.
Introduction
monocytes. Mab D11 is found to stain macrophages in the same way when tested on routinely fixed paraffin embedded and frozen tissues. In this paper some data on the ultrastructural localization and biochemical characteristics of antigens detected by Mab Dll are also given. 3.
Materials and Methods
3.1. Monoclonal antibody production The "ghosts" of human hepatocytes (a ghost is an intact complex of cell inner and external membranes) were used to immunize mice. The cell ghosts were obtained from a cell suspension of human liver by successive extractions with increasing concentrations of sodium chloride, according to the method of Bruemmer and Thomas [13]. BALB/c O' mice were injected intraperitoneally with 107 ghosts in 1.0 ml of phosphate-buffered saline. A month later the same dosage of ghosts was injected intraperitoneally again. On the third day cell fusion was performed according to Davidson and Gerald [14]. Hybridoma screening was carried out on frozen and paraffin-embedded liver sections using the indirect immunoperoxidase method described below.
3.2. Light microscopy." immunoperoxidase method Biopsies of normal human tissues (surgically removed specimens) were snap-frozen in liquid nitrogen. Five-/~m cryostat sections were cut, airdried and fixed in cold acetone (+4°C; 10 min). The paraffin sections were prepared from human tissues fixed by 4% para-formaldehyde (Sigma, St.Louis, MO, USA), followed by embedding in Histoplast as described previously [15]. Peripheral blood mononuclear cells from'healthy volunteers were separated by density centrifugation over Ficoll-Verographin (p= 1.077), and washed twice in cold PBS. Unfractionated mononuclear cells were resuspended in DMEM medium to 5 x 105/ml, then placed on the glass slide in a volume of 20 pl, air dried and fixed in acetone (4°C, 10 min) or methanol (-20°C, 10 rain). Monocytes were separated from unfractionated cells by glass adherence. In brief, 20 pl of mononuclear cells resuspended in DMEM (5 × 106/
ml) were placed on the glass slide, incubated in humid chamber (37°C, 4% CO2) for 2 h, washed thoroughly, air-dried and fixed as above. Tissue sections and blood cells were stained with monoclonal antibodies by the indirect immunoperoxidase method. Before treatment with antibodies the endogenous peroxidase was inhibited by 0.01 M Na-periodate and 0.1 M NaBH4 [16] or methanol with 0.3% H202 [17]. Then preparations were incubated successively with Mab Dll (culture supernatant, diluted 1:10) and peroxidase-conjugated rabbit anti-mouse IgG (diluted 1:100 in PBS, Dakopatts, Denmark), and neutralized with donor human serum. Peroxidase activity was visualized with DAB (3-Y-diaminobenzidine, Sigma, St.Louis, MO, USA), followed by contrasting with 0.4% OsO4. As a positive control, Mab BMA-0210 (Behringwerke AG, Marburg, Germany), specific to monocytes, was included.
3.2. Electron microscopy." immunoperoxidase method Specimens (2 x 2 mm) of the normal human liver, kidney and colon were fixed for 4 h at 4°C in 4% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.2), containing 0.05% saponin. The freezing of tissues, preparation of sections, treatment by antibodies (culture supernatant) and embedding process were performed as described elsewhere [18]. Mab Dll and peroxidase-conjugated rabbit anti-mouse IgG were in the concentrations used for the light microscopy test.
3.3. SDS-PAGE and immunoblotting For the Mr determination of D11 antigen, normal liver and two malignant tumours (histiocytoma and osteosarcoma), enriched in Dll antigen according to the immunohistochemical data, were used. One gram of each tissue was homogenized in 3 ml of PBS or PBS with 1% sodium dodecyl sulphate (SDS), each solution containing Trasylol (100 TIU/ml) and 2 mM phenylmethylsulphonylfluoride (PMSF; Sigma, St.Louis, MO, USA). The homogenates were extracted for 30 min at 4°C, then centrifuged at 8000 rev./min for 40 min. Cleared supernatants were electrophoresed
and transferred according to Laemmli [19] and Towbin [20], respectively. Blots were developed as reported earlier [21]. For incubation with blots, a 10-fold dilution of Mab Dl~ (culture supernatant) and a 500-fold dilution of peroxidaselabelled antibodies were used. 4.
Results
The screening of hybridomas produced after immunization with hepatocyte "ghosts" displayed various antibody specificities. The majority of hybridomas produced antibodies to liver cell lysosomes. Only one hybridoma turned out to react with Kupffer cells. 4.1.
Specificity of Mab Dr1 & tissue sections." light and electron microscopy
Mab D1~ was shown to react strongly with Kupffer cells. The intensity of staining was the same in cryostat and routinely fixed paraffin sections. D~1 antigen detected by Mab DI1 w a s evidently species-specific, because livers of mouse, rat and swine were Mab Dl~ negative. Since Mab D~I marked Kupffer cells, which are specialized tissue-fixed macrophages of liver, the antibody was investigated for reactivity on various tissue macrophages. The tissue specificity results for Mab D~ are summarized in Table 1. In the liver, Mab Dll reacted with the majority TABLE I Reactivity of Mab D1 j with macrophages in human tissues Samples
Stained cells
Liver
Kupffer cells; macrophages of portal connective tissue Macrophages of lamina propria, submucosal layer, muscularis mucosae, lymphoid follicles in lamina propria Macrophages of intertubular space, macrophages in connective tissue of blood vessels Macrophages alveolar, macrophages in connective tissue of blood vessels Macrophages of dermis, basal membrane. Langerhans cells are not stained Macrophages of red pulp, marginal macrophages of periarteriolar zones
Colon
Kidney Lung Skin Spleen
of Kupffer cells throughout whole epithelial layer (Fig. 1). In addition to Kupffer cells, the macrophages of the portal connective tissue were strongly stained by Mab D I I . A more detailed picture of the labeled Kupffer cell is seen at the ultrastructural level (Fig. 2). The sinusoidal part of the plasma membrane and the adjoining supramembrane layer usually displayetl a very high positive reaction. The remaining parts of the plasma membrane were also positive. In the cytoplasm various numbers of positively stained vesicles, tubular structures and lysosomes containing electron-dense material were present. In the colon, a strong positive reaction with macrophages of the lamina propria was found by light microscopy (Fig. 3). More weakly stained macrophages were scattered throughout the submucosal layer and muscularis mucosa. In the kidney, positively stained cells were detected in the intertubular spaces and also in the connective tissue that accompanies blood vessels of the glomeruli and medullary rays. At the ultrastructural level in the colon and kidney, Mab D~ showed a weaker reaction with macrophages than in the liver (Fig. 4). The reaction product was observed on the plasma membrane and inside lysosome-like structures. In the skin, Mab Dll detected dermal macrophages and stained basal membrane, but did not react with epidermal Langerhans cells. In the spleen, macrophages of red pulp and marginal macrophages of the periarteriolar zones displayed a strong positive reaction (Fig. 5). The stained macrophages of the white pulp are widely scattered. We did not find any cross-reaction of Mab DI~ with lymphocytes, fibroblasts, muscle cells, endothelial cells, dendritic cells of skin and spleen or any kind of epithelial cells in any of the tissues investigated. 4.2.
Reactivity of Mab DH on blood cells
M a b D I I gave a moderate reaction with cells on smears of unfractioned mononuclear cells from the peripheral blood. All stained cells displayed distinct cytoplasmic labelling, with a contrasting membrane in some cells. The amount of positive cells varied between 3-20%. On parallel smears, monocytes were identified using Mab
Fig. 1. Light microscopy. Section of human liver incubated with Mab Dll. Clear staining of Kupffer cells is visible, x 400. Fig. 2. lmmunoelectron microscopy. Kupffer cell stained by Mab Dlb The plasma membrane (short arrow), lysosomes (L) and cytoplasmic vesicles (long arrow) contain considerable amounts of the reaction product, x 5300. Fig. 3. Light microscopy. Section of human colon incubated with Mab Dll. Stained macrophages (arrow) are located in lamina propria. The cryptal cells are devoid of the reaction product, x 400. Fig. 4. lmmunoelectron microscopy. The colon macrophage stained by Mab D~I. The electron-dense granules are present in lysosome (L). More faint reaction is visible in cytoplasm and on the plasma membrane (arrow). x 5300. Fig. 5. Light microscopy. Section of human spleen incubated with Mab D~t. Macrophages of red pulp are strongly labelled. White pulp contains widely scattered stained macrophages (arrow). x 250. Fig. 6. Monocytes adhering to substrate after 2 h of incubation and stained by Mab D~. Note the intense labelling of cytoplasm and cell surface. Unstained monocytes are also present (arrow). x 640.
extracts were analysed (Fig. 7, lanes 5,6). Dll antigens detected in two different mglignant tumours, histiocytoma and osteosarcoma, exhibited identical sizes, 135 k D a (Fig. 7, lanes 1-4). Besides the major band revealed by M a b D~I on blots, the antibody reacted non-specifically with two or more high-molecular-weight components in all antigenic preparations. D11 antigen was well solubilized in PBS. The Mrs of the DI~ antigen solubilized by PBS and in the presence of ionic detergent are similar (125 kDa) (Fig. 7, lanes 5,6). 5.
Fig. 7. Immunoblotting Ag Di1 isolated from (a) histiocytoma/SDS lysate; original (lane 1) and diluted 1/2 (lane 2); (b) osteosarcoma/SDS lysate; original (lane 3) and diluted 1/2 (lane 4); (c) human liver - SDS lysate (lane 5), PBS extract (lane 6). Each sample (40 #1) given for electrophoresis contains 100-200 #g of protein. All samples were heated for 5 min at 90°C in SDS-buffer. The blots were quenched with 2% egg albumin (EA) for 1 h, incubated with a 20-fold dilution of Mab D~I (culture supernatant) in 1% EA for 16 h at 4°C, and then a 500-fold dilution of peroxidase-labelledrabbit anti-mouse IgG in 1% EA for 2 h at room temperature. Substrate: 0.01% H202 and chloronaphthol (0.5 mg/ml in PBS pH 7.4). The sizes of Mr standards are given in kDa. BMA-0210, known to recognize monocytes. Monocytes were found to be more frequent (1020%) than cells detected by M a b D~l. DII antigen was expressed on most monocytes attached to the substrate after 2 h of incubation (Fig. 6). In all cases about 10% unstained monocytes were observed. Some attached lymphocytes, which always contaminated monocyte preparations, did not react with M a b Dll.
4.3.
Molecular weight determination
The epitope revealed by M a b DI~ was destroyed by mercaptoethanol treatment. Therefore, native non-reduced antigenic preparations were used for immunoblotting. The size of the Dla antigen was measured as 125 k D a when liver
Discussion
The anti-macrophage specificity of M a b D ~ was originally demonstrated on Kupffer cells. M a b D~I recognized the corresponding antigen on cryostat and paraffin tissue sections. This property of the antibody makes it suitable for clinical use, where embedding of tissues in Histoplast is routinely applied. M a b Dll reacted strongly with the majority of free and fixed macrophages in a large number of normal tissues. No cross-reaction was found with epithelial or connective tissue cells in the sections. In view of the immunohistochemical data, we were unable to conclude that M a b D~I distinguished some definite subpopulation of macrophages in human organs. With respect to subpopulations of macrophages, some were immunohistochemically defined in rat lymph nodes and spleen [22] and human lymphoid organs [23]. M a b Dll revealed freshly isolated and substrate-attached monocytes of the peripheral blood. Not all, but some subset of monocytes were shown to be labelled by the antibody. The moderate staining of blood cells pointed to weak expression of the antigen. Attachment of cells to the substrate resulted in more intense labelling. The increase in antigen concentration in tissue macrophages seems to reflect the maturation of monocytes to macrophages. According to our electron microscopic data, the D~l antigen is located on the plasma membrane, in the supramembranous layer, and within cytoplasmic structures in Kupffer cells, including lysosomes. In both kidney and colon macrophages, DII antigen was found to be specifically
bound to plasmalemma and lysosome-like structures. Dll antigen was easily extracted from cells with PBS. Nevertheless, the membrane origin of the antigen is not excluded, because the treatment of tissues with ionic detergent SDS increased the antigen solubility (data on dot-blot reaction). The size of the D ~ antigen of normal liver was shown to be 125 kDa. In tumours (histiocytoma and osteosarcoma), this was increased to 135 kDa, probably due to the greater extent of glycosylation observed for the proteins of embryonic and malignant tissues [24]. DI~ antigen with " t u m o u r " Mr (135 kDa) might be an additional diagnostic trait for neoplasia of histiocytic origin, since the "normal" Mr of Dll antigen of any non-malignant macrophage is always less - 125 kDa. Mab D ~ shows a reactivity closely restricted to circulating blood monocytes and resident macrophages, i.e., cells that are prominently phagocytic. As has been reported, Mab DI1 did not reveal Langerhans cells, which together with other dendritic cells are believed to exert a specific function during antigen presentation to the lymphocyte [25]. Although the reactivity of Mab D ~ with other types of dendritic cells was not completely investigated, the preliminary data on immunohistochemical localization of Mab D11 supported the notion that we are dealing with an antibody which is not specific to "accessory cells" of the T and B immune responses. Mab D ~ is very similar to other known "panmacrophage" antibodies, EBM 11 [18], Ki-M6 [11] and KP 1 [12]. Recent studies using cells transfected with c D N A clone encoding KP 1 antigen and immunoabsorbent columns had demonstrated that all the above-mentioned specificities recognized the same l l0-kDa glycoprotein. Furthermore, these monoclonal antibodies reacted with different antigen epitopes of this protein [9]. The greatest similarity appears to exist between Mab Dll and KP 1. Indeed, both antibodies displayed very close molecular weight values for corresponding antigens (125 and 110 kDa), and similar patterns of tissue reaction and resistance of epitopes to paraffin embedding. According to the results obtained, Mab Dll and KP 1 may recognize epitopes present on the same antigen and distinguished by their stability to reducing agents:
KP 1, in contrast to Mab D11, detects the stable determinant. KP 1 was generated against the lysosome fraction and demonstrated the reaction with lysosomes on cells. Mab Dll also defines the lysosome-like structures in all macrophage types studied according to electron microscopic data. In any case, only direct comparison of Mab D~x with other pan-macrophage antibodies in the same experiment could help to elicit their complete identity.
Acknowledgements We thank Dr. A.K. Yasova for her kind help with hybridoma technique, Mrs. O.A. Salnikova for skilled technical assistance and Mrs. O.V. Chistyakova for donor sera. We also thank the surgeons of Moscow Hospital N 7 for tissue biopsies.
References [1] Leenen, P.I.M., Melis, M., Slicker, W.A.T. and Ewijk, W.V. (1990) Eur. J. Immunol. 20, 27. [2] Zwadlo, G., Br6cker, E.-B., von Bassewitz, D.-B., Feige, U. and Sorg, C. (1985) J. lmmunol. 134, 1487. [3] Maruyama, S., Naito, T., Kolkita, H., Kishimoto, S., Yamamura, Y. and Kishimoto, T. (1983) J. Clin. Immunol. 3, 57. [4] Dimitriu-Bona, A., Burmester, G.R., Waters, S.I. and Winchester, R.I. (1983)J. Immunol. 130, 145. [5] Radzun, H.J. and Parwaresch, M.R. (1983) Cell lmmunol. 82, 174. [6] Todd, R.F.III and Schlossman, S.F. (1982) Blood 59, 775. [7] Franklin, W.A., Mason, D.Y., Pulford, K., Falini, B., Bliss, E., Gatter, K.C., Stein, H., Clarke, L.C. and McGee, J.O'D. (1986) Lab. Invest. 54, 322. [8] Kelly, P.M.A., Bliss, E., Morton, J.A., Burns, J. and McGee, J.O'D. (1988) J. Clin. Pathol. 41, 510. [9] Micklem, K., Cordelt, J., Rigney, E., Simmons, D., Pulford, K., Stross, P. and Mason, D. (1989) Leucocyte Typing, IV. White Cell Differentiation Antigens (W. Knapp, Ed.) pp. 843 846, Oxford University Press, Oxford. [10] Davey, T.R., Cordell, J.L., Erber, W.N., Pulford, K.A.F., Gatter, K.C. and Mason, D.Y. (1988) J. Clin. Pathol. 41, 753. [11] Parwaresch, M.R., Radzun, H.J., Kreipe, H., Hansmann, M.-L. and Barth, J. (1986) Am. J. Pathol. 124, 141. [12] Pulford, K.A.F., Rigney, E.M., Micklem, K.I., Jones, M., Stross, P.W., Gatter, K.C. and Mason, D.Y. (1988) J. Clin. Pathol. 42, 414. [13] Bruemmer, N.C. and Thomas, L.E. (1957) Exp. Cell. Res. 13, 103.
[14] Davidson, R.L. and Gerald, P.S. (1977) Methods Cell Biol. 15, 325. [15] Engelhardt, N.V., Gleiberman, A.S., Lazareva, M.N. and Poltoranina, V.S. (1979) in: Carcinoembryonic Antigens. Chemistry, Biology, Clinical Application (F.G. Lehman, Ed.), pp. 111-119. Elsevier-North-Holland, Amsterdam. [16] Taylor, C.R. (1978) Arch. Pathol. Med. 102, 113. [17] Baranov, V.N., Poltoranina, V.S. and Goussev, A.I. (1986) Bull. Exp. Biol. Med. 2, 199 (in Russian). [18] Kuprina, N.I., Baranov, V.N., Yasova, A.K., Rudinskaya, T.D., Escribano, M., Cordier, J., Gleiberman, A.S. and Goussev, A.I. (1990) Histochemistry 94, 179. [19] Laemmli, U.K. (1970) Nature 227, 680.
[20] Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350. [21] Rudinskaya, T.D., Kuprina, N.I., Yasova, A.K., Escribano, M.X., Cordier, J., Goussev, A.I. and Galperina, T.E. (1987) Biol. Membr. 4, 194 (in Russian). [22] Chao, D. and Macpherson, G.G. (1989) Eur. J. Immunol. 19, 1237. [23] Baroni, C.D., Vitolo, D., Remrotti, D., Biondi, A., Pezzella, F., Ruco, L.P. and Uccini, S. (1987) Histopathology 11, 1029. [24] Rothbard, J.B., Brackenbury, R., Cunningham, B.A. and Edelman, G.M. (1982)J. Biol. Chem. 257, 11064. [25] Steinman, R.M. and Zweig, M.C. (1980) lmmunol. Rev. 53, 127.
0165 - 2478 / 92 / $ 5.00 © 1992 ElsevierSciencePublishers B.V. All rights reserved IMLET 01797
Dll, a novel monoclonal antibody specific for human mature macrophages and peripheral blood monocytes T . D . R u d i n s k a y a a, V.S. P o l t o r a n i n a a, V , N . B a r a n o v a, N . N . P e t r o v i c h e v b, B.O. V o y t e n k o v c a n d L . O . T r e t y a k o v a aLaboratory of Immunochemistry and bLaboratory of Clinical Cytology, Cancer Research Centre, Russian Academy of Sciences, Moscow, Russia, and ~Laboratory of Oncoimmunology, Institute of Oncology, St.-Petersburg, Russia
(Received 4 March 1992; accepted 18 March 1992)
1. Summary
2.
A new monoclonal antibody designated Mab D ~ is described, which shows a restricted reactivity to cells of the monocyte/macrophage system. When tested by light and electron microscopic immunoperoxidase methods, Mab D~l specifically reacts with blood monocytes and stains resident macrophages in a wide variety of human tissues; it does not mark the macrophages of other species, i.e., rat, swine and mouse. Antigen-presenting cells, e.g., Langerhans cells, are Mab Dll negative. Mab Oil reveals the antigen on cryostat and paraffin tissue sections. Ultrastructurally the antigen recognized by Mab Dll in all macrophage types studied is located on the plasma membrane and within cytoplasmic structures including lysosomes. On immunoblotting, Mab D~I detects the 125-kDa antigen in human liver and the 135-kDa protein in tumours of histioeytic origin. The similarity of Mab Dl~ to known "panmacrophage" monoclonal antibodies is discussed.
The recent advances in the study of cells of the monocyte/macrophage system are mainly due to obtaining a large panel of monoclonal antibodies that recognize not only universal antigenic markers of phagocytic cells but also identify the specific determinants of mature and mediated cell forms during the differentiation of phagocyte cell lineage [1 3]. Most of the known monoclonal antibodies specific to mononuclear phagocytes are reactive with tissue macrophages as well as with monocytes, thus revealing that the antigens have been expressed in the earliest stages of phagocyte differentiation [4-7]. Of special interest among these monoclonal specificities are antibodies that react strongly with the majority of macrophages in a number of organs, and which for this reason are classified as "pan-macrophage" antibodies. According to results presented at the Fourth International Conference on Human Leucocyte Differentiation Antigens, "pan-macrophage" antibodies have the following specificities (CD 68 reagents): E B M l l [8], y 2/131 [9], y 1/82A [10], Ki-M6 [11], KP 1 [12], obtained in different laboratories. In the present report we describe a new antimacrophage monoclonal antibody, Dll (Mab D1~) that fulfils the criteria for "pan-macrophage" reagent. Mab D ~ labels all types of human tissue macrophages to a high degree and also reacts with freshly isolated peripheral blood
Key words: Monoclonal antibody; Human macrophage;
Monocyte; Light and electron microscopic immunolocalization Correspondence to." Dr. T.D. Rudinskaya, Laboratory of Immunochemistry, Cancer Research Centre, 115478, Moscow, Kashirskoye shosse 24, Russia.
Introduction
monocytes. Mab D11 is found to stain macrophages in the same way when tested on routinely fixed paraffin embedded and frozen tissues. In this paper some data on the ultrastructural localization and biochemical characteristics of antigens detected by Mab Dll are also given. 3.
Materials and Methods
3.1. Monoclonal antibody production The "ghosts" of human hepatocytes (a ghost is an intact complex of cell inner and external membranes) were used to immunize mice. The cell ghosts were obtained from a cell suspension of human liver by successive extractions with increasing concentrations of sodium chloride, according to the method of Bruemmer and Thomas [13]. BALB/c O' mice were injected intraperitoneally with 107 ghosts in 1.0 ml of phosphate-buffered saline. A month later the same dosage of ghosts was injected intraperitoneally again. On the third day cell fusion was performed according to Davidson and Gerald [14]. Hybridoma screening was carried out on frozen and paraffin-embedded liver sections using the indirect immunoperoxidase method described below.
3.2. Light microscopy." immunoperoxidase method Biopsies of normal human tissues (surgically removed specimens) were snap-frozen in liquid nitrogen. Five-/~m cryostat sections were cut, airdried and fixed in cold acetone (+4°C; 10 min). The paraffin sections were prepared from human tissues fixed by 4% para-formaldehyde (Sigma, St.Louis, MO, USA), followed by embedding in Histoplast as described previously [15]. Peripheral blood mononuclear cells from'healthy volunteers were separated by density centrifugation over Ficoll-Verographin (p= 1.077), and washed twice in cold PBS. Unfractionated mononuclear cells were resuspended in DMEM medium to 5 x 105/ml, then placed on the glass slide in a volume of 20 pl, air dried and fixed in acetone (4°C, 10 min) or methanol (-20°C, 10 rain). Monocytes were separated from unfractionated cells by glass adherence. In brief, 20 pl of mononuclear cells resuspended in DMEM (5 × 106/
ml) were placed on the glass slide, incubated in humid chamber (37°C, 4% CO2) for 2 h, washed thoroughly, air-dried and fixed as above. Tissue sections and blood cells were stained with monoclonal antibodies by the indirect immunoperoxidase method. Before treatment with antibodies the endogenous peroxidase was inhibited by 0.01 M Na-periodate and 0.1 M NaBH4 [16] or methanol with 0.3% H202 [17]. Then preparations were incubated successively with Mab Dll (culture supernatant, diluted 1:10) and peroxidase-conjugated rabbit anti-mouse IgG (diluted 1:100 in PBS, Dakopatts, Denmark), and neutralized with donor human serum. Peroxidase activity was visualized with DAB (3-Y-diaminobenzidine, Sigma, St.Louis, MO, USA), followed by contrasting with 0.4% OsO4. As a positive control, Mab BMA-0210 (Behringwerke AG, Marburg, Germany), specific to monocytes, was included.
3.2. Electron microscopy." immunoperoxidase method Specimens (2 x 2 mm) of the normal human liver, kidney and colon were fixed for 4 h at 4°C in 4% paraformaldehyde in 0.1 M cacodylate buffer (pH 7.2), containing 0.05% saponin. The freezing of tissues, preparation of sections, treatment by antibodies (culture supernatant) and embedding process were performed as described elsewhere [18]. Mab Dll and peroxidase-conjugated rabbit anti-mouse IgG were in the concentrations used for the light microscopy test.
3.3. SDS-PAGE and immunoblotting For the Mr determination of D11 antigen, normal liver and two malignant tumours (histiocytoma and osteosarcoma), enriched in Dll antigen according to the immunohistochemical data, were used. One gram of each tissue was homogenized in 3 ml of PBS or PBS with 1% sodium dodecyl sulphate (SDS), each solution containing Trasylol (100 TIU/ml) and 2 mM phenylmethylsulphonylfluoride (PMSF; Sigma, St.Louis, MO, USA). The homogenates were extracted for 30 min at 4°C, then centrifuged at 8000 rev./min for 40 min. Cleared supernatants were electrophoresed
and transferred according to Laemmli [19] and Towbin [20], respectively. Blots were developed as reported earlier [21]. For incubation with blots, a 10-fold dilution of Mab Dl~ (culture supernatant) and a 500-fold dilution of peroxidaselabelled antibodies were used. 4.
Results
The screening of hybridomas produced after immunization with hepatocyte "ghosts" displayed various antibody specificities. The majority of hybridomas produced antibodies to liver cell lysosomes. Only one hybridoma turned out to react with Kupffer cells. 4.1.
Specificity of Mab Dr1 & tissue sections." light and electron microscopy
Mab D1~ was shown to react strongly with Kupffer cells. The intensity of staining was the same in cryostat and routinely fixed paraffin sections. D~1 antigen detected by Mab DI1 w a s evidently species-specific, because livers of mouse, rat and swine were Mab Dl~ negative. Since Mab D~I marked Kupffer cells, which are specialized tissue-fixed macrophages of liver, the antibody was investigated for reactivity on various tissue macrophages. The tissue specificity results for Mab D~ are summarized in Table 1. In the liver, Mab Dll reacted with the majority TABLE I Reactivity of Mab D1 j with macrophages in human tissues Samples
Stained cells
Liver
Kupffer cells; macrophages of portal connective tissue Macrophages of lamina propria, submucosal layer, muscularis mucosae, lymphoid follicles in lamina propria Macrophages of intertubular space, macrophages in connective tissue of blood vessels Macrophages alveolar, macrophages in connective tissue of blood vessels Macrophages of dermis, basal membrane. Langerhans cells are not stained Macrophages of red pulp, marginal macrophages of periarteriolar zones
Colon
Kidney Lung Skin Spleen
of Kupffer cells throughout whole epithelial layer (Fig. 1). In addition to Kupffer cells, the macrophages of the portal connective tissue were strongly stained by Mab D I I . A more detailed picture of the labeled Kupffer cell is seen at the ultrastructural level (Fig. 2). The sinusoidal part of the plasma membrane and the adjoining supramembrane layer usually displayetl a very high positive reaction. The remaining parts of the plasma membrane were also positive. In the cytoplasm various numbers of positively stained vesicles, tubular structures and lysosomes containing electron-dense material were present. In the colon, a strong positive reaction with macrophages of the lamina propria was found by light microscopy (Fig. 3). More weakly stained macrophages were scattered throughout the submucosal layer and muscularis mucosa. In the kidney, positively stained cells were detected in the intertubular spaces and also in the connective tissue that accompanies blood vessels of the glomeruli and medullary rays. At the ultrastructural level in the colon and kidney, Mab D~ showed a weaker reaction with macrophages than in the liver (Fig. 4). The reaction product was observed on the plasma membrane and inside lysosome-like structures. In the skin, Mab Dll detected dermal macrophages and stained basal membrane, but did not react with epidermal Langerhans cells. In the spleen, macrophages of red pulp and marginal macrophages of the periarteriolar zones displayed a strong positive reaction (Fig. 5). The stained macrophages of the white pulp are widely scattered. We did not find any cross-reaction of Mab DI~ with lymphocytes, fibroblasts, muscle cells, endothelial cells, dendritic cells of skin and spleen or any kind of epithelial cells in any of the tissues investigated. 4.2.
Reactivity of Mab DH on blood cells
M a b D I I gave a moderate reaction with cells on smears of unfractioned mononuclear cells from the peripheral blood. All stained cells displayed distinct cytoplasmic labelling, with a contrasting membrane in some cells. The amount of positive cells varied between 3-20%. On parallel smears, monocytes were identified using Mab
Fig. 1. Light microscopy. Section of human liver incubated with Mab Dll. Clear staining of Kupffer cells is visible, x 400. Fig. 2. lmmunoelectron microscopy. Kupffer cell stained by Mab Dlb The plasma membrane (short arrow), lysosomes (L) and cytoplasmic vesicles (long arrow) contain considerable amounts of the reaction product, x 5300. Fig. 3. Light microscopy. Section of human colon incubated with Mab Dll. Stained macrophages (arrow) are located in lamina propria. The cryptal cells are devoid of the reaction product, x 400. Fig. 4. lmmunoelectron microscopy. The colon macrophage stained by Mab D~I. The electron-dense granules are present in lysosome (L). More faint reaction is visible in cytoplasm and on the plasma membrane (arrow). x 5300. Fig. 5. Light microscopy. Section of human spleen incubated with Mab D~t. Macrophages of red pulp are strongly labelled. White pulp contains widely scattered stained macrophages (arrow). x 250. Fig. 6. Monocytes adhering to substrate after 2 h of incubation and stained by Mab D~. Note the intense labelling of cytoplasm and cell surface. Unstained monocytes are also present (arrow). x 640.
extracts were analysed (Fig. 7, lanes 5,6). Dll antigens detected in two different mglignant tumours, histiocytoma and osteosarcoma, exhibited identical sizes, 135 k D a (Fig. 7, lanes 1-4). Besides the major band revealed by M a b D~I on blots, the antibody reacted non-specifically with two or more high-molecular-weight components in all antigenic preparations. D11 antigen was well solubilized in PBS. The Mrs of the DI~ antigen solubilized by PBS and in the presence of ionic detergent are similar (125 kDa) (Fig. 7, lanes 5,6). 5.
Fig. 7. Immunoblotting Ag Di1 isolated from (a) histiocytoma/SDS lysate; original (lane 1) and diluted 1/2 (lane 2); (b) osteosarcoma/SDS lysate; original (lane 3) and diluted 1/2 (lane 4); (c) human liver - SDS lysate (lane 5), PBS extract (lane 6). Each sample (40 #1) given for electrophoresis contains 100-200 #g of protein. All samples were heated for 5 min at 90°C in SDS-buffer. The blots were quenched with 2% egg albumin (EA) for 1 h, incubated with a 20-fold dilution of Mab D~I (culture supernatant) in 1% EA for 16 h at 4°C, and then a 500-fold dilution of peroxidase-labelledrabbit anti-mouse IgG in 1% EA for 2 h at room temperature. Substrate: 0.01% H202 and chloronaphthol (0.5 mg/ml in PBS pH 7.4). The sizes of Mr standards are given in kDa. BMA-0210, known to recognize monocytes. Monocytes were found to be more frequent (1020%) than cells detected by M a b D~l. DII antigen was expressed on most monocytes attached to the substrate after 2 h of incubation (Fig. 6). In all cases about 10% unstained monocytes were observed. Some attached lymphocytes, which always contaminated monocyte preparations, did not react with M a b Dll.
4.3.
Molecular weight determination
The epitope revealed by M a b DI~ was destroyed by mercaptoethanol treatment. Therefore, native non-reduced antigenic preparations were used for immunoblotting. The size of the Dla antigen was measured as 125 k D a when liver
Discussion
The anti-macrophage specificity of M a b D ~ was originally demonstrated on Kupffer cells. M a b D~I recognized the corresponding antigen on cryostat and paraffin tissue sections. This property of the antibody makes it suitable for clinical use, where embedding of tissues in Histoplast is routinely applied. M a b Dll reacted strongly with the majority of free and fixed macrophages in a large number of normal tissues. No cross-reaction was found with epithelial or connective tissue cells in the sections. In view of the immunohistochemical data, we were unable to conclude that M a b D~I distinguished some definite subpopulation of macrophages in human organs. With respect to subpopulations of macrophages, some were immunohistochemically defined in rat lymph nodes and spleen [22] and human lymphoid organs [23]. M a b Dll revealed freshly isolated and substrate-attached monocytes of the peripheral blood. Not all, but some subset of monocytes were shown to be labelled by the antibody. The moderate staining of blood cells pointed to weak expression of the antigen. Attachment of cells to the substrate resulted in more intense labelling. The increase in antigen concentration in tissue macrophages seems to reflect the maturation of monocytes to macrophages. According to our electron microscopic data, the D~l antigen is located on the plasma membrane, in the supramembranous layer, and within cytoplasmic structures in Kupffer cells, including lysosomes. In both kidney and colon macrophages, DII antigen was found to be specifically
bound to plasmalemma and lysosome-like structures. Dll antigen was easily extracted from cells with PBS. Nevertheless, the membrane origin of the antigen is not excluded, because the treatment of tissues with ionic detergent SDS increased the antigen solubility (data on dot-blot reaction). The size of the D ~ antigen of normal liver was shown to be 125 kDa. In tumours (histiocytoma and osteosarcoma), this was increased to 135 kDa, probably due to the greater extent of glycosylation observed for the proteins of embryonic and malignant tissues [24]. DI~ antigen with " t u m o u r " Mr (135 kDa) might be an additional diagnostic trait for neoplasia of histiocytic origin, since the "normal" Mr of Dll antigen of any non-malignant macrophage is always less - 125 kDa. Mab D ~ shows a reactivity closely restricted to circulating blood monocytes and resident macrophages, i.e., cells that are prominently phagocytic. As has been reported, Mab DI1 did not reveal Langerhans cells, which together with other dendritic cells are believed to exert a specific function during antigen presentation to the lymphocyte [25]. Although the reactivity of Mab D ~ with other types of dendritic cells was not completely investigated, the preliminary data on immunohistochemical localization of Mab D11 supported the notion that we are dealing with an antibody which is not specific to "accessory cells" of the T and B immune responses. Mab D ~ is very similar to other known "panmacrophage" antibodies, EBM 11 [18], Ki-M6 [11] and KP 1 [12]. Recent studies using cells transfected with c D N A clone encoding KP 1 antigen and immunoabsorbent columns had demonstrated that all the above-mentioned specificities recognized the same l l0-kDa glycoprotein. Furthermore, these monoclonal antibodies reacted with different antigen epitopes of this protein [9]. The greatest similarity appears to exist between Mab Dll and KP 1. Indeed, both antibodies displayed very close molecular weight values for corresponding antigens (125 and 110 kDa), and similar patterns of tissue reaction and resistance of epitopes to paraffin embedding. According to the results obtained, Mab Dll and KP 1 may recognize epitopes present on the same antigen and distinguished by their stability to reducing agents:
KP 1, in contrast to Mab D11, detects the stable determinant. KP 1 was generated against the lysosome fraction and demonstrated the reaction with lysosomes on cells. Mab Dll also defines the lysosome-like structures in all macrophage types studied according to electron microscopic data. In any case, only direct comparison of Mab D~x with other pan-macrophage antibodies in the same experiment could help to elicit their complete identity.
Acknowledgements We thank Dr. A.K. Yasova for her kind help with hybridoma technique, Mrs. O.A. Salnikova for skilled technical assistance and Mrs. O.V. Chistyakova for donor sera. We also thank the surgeons of Moscow Hospital N 7 for tissue biopsies.
References [1] Leenen, P.I.M., Melis, M., Slicker, W.A.T. and Ewijk, W.V. (1990) Eur. J. Immunol. 20, 27. [2] Zwadlo, G., Br6cker, E.-B., von Bassewitz, D.-B., Feige, U. and Sorg, C. (1985) J. lmmunol. 134, 1487. [3] Maruyama, S., Naito, T., Kolkita, H., Kishimoto, S., Yamamura, Y. and Kishimoto, T. (1983) J. Clin. Immunol. 3, 57. [4] Dimitriu-Bona, A., Burmester, G.R., Waters, S.I. and Winchester, R.I. (1983)J. Immunol. 130, 145. [5] Radzun, H.J. and Parwaresch, M.R. (1983) Cell lmmunol. 82, 174. [6] Todd, R.F.III and Schlossman, S.F. (1982) Blood 59, 775. [7] Franklin, W.A., Mason, D.Y., Pulford, K., Falini, B., Bliss, E., Gatter, K.C., Stein, H., Clarke, L.C. and McGee, J.O'D. (1986) Lab. Invest. 54, 322. [8] Kelly, P.M.A., Bliss, E., Morton, J.A., Burns, J. and McGee, J.O'D. (1988) J. Clin. Pathol. 41, 510. [9] Micklem, K., Cordelt, J., Rigney, E., Simmons, D., Pulford, K., Stross, P. and Mason, D. (1989) Leucocyte Typing, IV. White Cell Differentiation Antigens (W. Knapp, Ed.) pp. 843 846, Oxford University Press, Oxford. [10] Davey, T.R., Cordell, J.L., Erber, W.N., Pulford, K.A.F., Gatter, K.C. and Mason, D.Y. (1988) J. Clin. Pathol. 41, 753. [11] Parwaresch, M.R., Radzun, H.J., Kreipe, H., Hansmann, M.-L. and Barth, J. (1986) Am. J. Pathol. 124, 141. [12] Pulford, K.A.F., Rigney, E.M., Micklem, K.I., Jones, M., Stross, P.W., Gatter, K.C. and Mason, D.Y. (1988) J. Clin. Pathol. 42, 414. [13] Bruemmer, N.C. and Thomas, L.E. (1957) Exp. Cell. Res. 13, 103.
[14] Davidson, R.L. and Gerald, P.S. (1977) Methods Cell Biol. 15, 325. [15] Engelhardt, N.V., Gleiberman, A.S., Lazareva, M.N. and Poltoranina, V.S. (1979) in: Carcinoembryonic Antigens. Chemistry, Biology, Clinical Application (F.G. Lehman, Ed.), pp. 111-119. Elsevier-North-Holland, Amsterdam. [16] Taylor, C.R. (1978) Arch. Pathol. Med. 102, 113. [17] Baranov, V.N., Poltoranina, V.S. and Goussev, A.I. (1986) Bull. Exp. Biol. Med. 2, 199 (in Russian). [18] Kuprina, N.I., Baranov, V.N., Yasova, A.K., Rudinskaya, T.D., Escribano, M., Cordier, J., Gleiberman, A.S. and Goussev, A.I. (1990) Histochemistry 94, 179. [19] Laemmli, U.K. (1970) Nature 227, 680.
[20] Towbin, H., Staehelin, T. and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350. [21] Rudinskaya, T.D., Kuprina, N.I., Yasova, A.K., Escribano, M.X., Cordier, J., Goussev, A.I. and Galperina, T.E. (1987) Biol. Membr. 4, 194 (in Russian). [22] Chao, D. and Macpherson, G.G. (1989) Eur. J. Immunol. 19, 1237. [23] Baroni, C.D., Vitolo, D., Remrotti, D., Biondi, A., Pezzella, F., Ruco, L.P. and Uccini, S. (1987) Histopathology 11, 1029. [24] Rothbard, J.B., Brackenbury, R., Cunningham, B.A. and Edelman, G.M. (1982)J. Biol. Chem. 257, 11064. [25] Steinman, R.M. and Zweig, M.C. (1980) lmmunol. Rev. 53, 127.
