JOURNAL OF PATHOLOGY, VOL.

163: 3 15-32 1 (1 99 1)

A NEW MONOCLONAL ANTIBODY, PM-2K, SPECIFICALLY RECOGNIZES TISSUE MACROPHAGES BUT NOT BLOOD MONOCYTES MOTOHIRO TAKEYA, TATSUAKI TSUCHIYA, YASUHIRO SHIMOKAWA AND KIYOSHI TAKAHASHI

Second Department of Pathology, Kumamoto University Medical School, Kumamoto, Japan Received I 7 September I990 Accepted I2 November I990

SUMMARY A new monoclonal antibody, PM-2K, was raised against 24-h cultured human peritoneal macrophages. Immunohistochemically, PM-2K recognized most tissue macrophages in lymphoreticular organs such as the thymus, spleen, lymph node, and tonsil. Kupffercells of the liver. alveolar macrophages, and macrophages in the interstitial tissue of the kidney, pancreas, and many other organs were also positively labelled. On the other hand, PM-2K failed to recognize blood monocytes, freshly isolated peritoneal macrophages, microglial cells, osteoclasts, and dendritic cells such as Langerhans cells, interdigitating cells, and follicular dendritic cells. In various pathological conditions, PM-2K labelled a wide variety of proliferating macrophages. Reaction products of PM-2K were observed by immunoelectron microscopy on the cytoplasmic membrane of cultured peritoneal macrophages. The molecular weight of the antigen recognized by PM-2K was determined to be 150 kD by Western blotting. As no cells other than macrophages were reactive with PM-ZK, this antibody is considered to be very useful not onIy in the investigation of macrophage differentiation and maturation, but also in the diagnosis of various proliferative disorders of macrophages. KEY

wow-Monoclonal

antibody, macrophages, proliferative disorders of macrophages, immunohistochemistry.

INTRODUCTION Macrophages are widely distributed throughout the body to play a crucial role in defence mechanisms.' Although it is accepted that macrophages originate in bone marrow,2 they are heterogeneous in morphology, cytochemistry, functions, and surface determinants3 Dendritic cells such as interdigitating cells, veiled cells, indeterminate cells, and Langerhans cells are also thought to belong to the mononuclear phagocyte ~ y s t e m , although ~,~ some investigators consider them to be of a different Various morphological, cytochemical, and immunological techniques have been used to identify monocytes/macrophages or dendritic cells. Among these, immunohistochemical labelling with monoclonal antibodies is one of the most specific and reliable techniques. A number of monoclonal Addressee for correspondence: Motohiro Takeya, Second Department of Pathology, Kumamoto University Medical School, 2-2-1, Honjo, Kumamoto, 860, Japan.

0022-341 7/91/0403 1 5 4 7 $05.00 0 1991 by John Wiley& Sons, Ltd.

antibodies have been raised against human monocytes/macropha es; however, most of them also label other cells.i This paper describes an anti-macrophage monoclonal antibody, PM-2K, which is highly specific for tissue macrophages but not reactive with blood monocytes. The reactivities of PM-2K in various proliferative disorders of macrophages are also reported. MATERIALS AND METHODS Production of monoclonal antibody Human peritoneal cells were collected from seven patients with endometriosis at the time of surgery as described elsewhere.' The cell suspensions were cultured in RPMI 1640 with 10 per cent fetal calf serum (FCS) on plastic dishes. After 24 h incubation in a 5 per cent CO, incubator, the adherent cells were collected and used for immunization. BALB/c mice were immunized three times by intraperitoneal injection of about 5 x lo6of the adherent peritoneal

316

M. TAKEYA E T A L .

cells each. Fusion of spleen cells with NS- 1 myeloma cells, and subsequent selection of hybridomas were performed as described elsewhere.” A clone PM-2K which produces specific antibody to tissue macrophages was selected in this study. Ascites fluid was obtained by injecting 5 x lo6 PM-2K cells into pristane-primed BALB/c mice. The isotype of PM2K was determined by a mouse monoclonal antibody isotyping kit (Amersham, U.K.).

from Wistar rats, guinea pigs of Hartley strain, golden hamsters, New Zealand white rabbits, cats, beagles, goats, horses and cows.

Immunohistochemistry Tissue sections, cell smears, epidermal sheet, and spread omentum were stained with hybridoma supernatant or ascites diluted 1 : l O 000 using the indirect immunoperoxidase method.” Control slides were incubated with non-immunized mouse serum or phosphate-buffered saline (PBS), then Tissue samples processed as described above. To assess the reacThe following fresh human tissues were obtained tivity of PM-2K with interdigitating cells, double from the Kumamoto University Hospital at staining with polyclonal rabbit anti-S- 100 protein autopsy or at the time of surgical resection: thymus, (Dakopatts, Glostrup, Denmark) and PM-2K were tonsil, spleen, lymph node, heart, lung, liver, kidney, performed. The sections of thymus, spleen, and pancreas, small intestine, thyroid, adrenal, salivary lymph nodes were processed to block endogenous glands, brain, skeletal muscles, bone marrow, peroxidase activity’, and incubated with anti-S- 100 omentum, and skin, Tissues were obtained from protein followed by treatment with species-specific proliferative disorders of monocytes/macrophages anti-rabbit Ig [F(ab’),] conjugated with peroxidase including the liver and spleen of a case of Gaucher’s (Amersham). After the visualization of peroxidase disease, the synovial tissue of a case of multicentric activity with 3,3’-diaminobenzidine (DAB)-H,O,, reticulohistiocytosis, the spleen and lymph nodes of the same sections were stained with PM-2K using two cases ofmalignant histiocytosis, and the tumour the alkaline phosphatase anti-alkaline phosphatase tissue of four cases of malignant fibrous histiocy(APAAP) m e t h ~ d . ’Alkaline ~ phosphatase was toma (MFH) and three cases of giant cell tumour visualized using naphthol AS-MX phosphate and (GCT) of the bone. The tissues were fixed with 2 per fast blue BB salt. For immunoelectron microscopy, cent periodate-lysine-paraformaldehyde fixative peritoneal cells obtained as described above were for 4 h and processed to make cryostat sections.” incubated on Petri dishes with RPMII640 suppleEpidermal sheets of the skin were prepared as mented with 10 per cent FCS at 37°C for 24 h. After described in ref. 11. The omentum was spread removal of the non-adherent cells, the cells were on albumin-coated slides without sectioning. fixed with 0.05 per cent glutaraldehyde in 0,165 M Peripheral blood cells obtained from healthy volun- cacodylate buffer, pH 7.4, for 10 min. Then the cells teers were centrifuged on a Lymphoprep (Nycomed were washed with PBS containing normal sheep AS, Oslo, Norway) to collect mononuclear cells. serum and incubated with culture supernatant of Some of the blood mononuclear cells were spread PM-2K overnight, followed by incubation with onto slides using Cytospin (Shandon, Cheshire, peroxidase-conjugated anti-mouse Ig [F(ab’),] U.K.), and the rest were cultured on four chamber- (Amersham) for 3 h. To observe the endogenous slides(Miles, Elkhart, U.S.A.) in RPMI1640 with 10 enzyme activity simultaneously, blocking of per cent FCS for I , 3, or 5 days in an atmosphere of 5 endogenous peroxidase was not performed. Peroxiper cent CO, and 95 per cent air. Peritoneal macrodase activity was visualized by incubating the cells in phages were obtained as described above. Alveolar a solution of 0.05 per cent DAB and 0.01 per cent macrophages were collected from the bronchoalveoH,O, in 0.05 M Tris-HC1 buffer (pH 7.6) for 30 min. lar lavage fluid of healthy volunteers. Freshly Then thecells werepost-fixed with 1 percent osmium obtained peritoneal or alveolar macrophages were tetroxide for 1 h, dehydrated, and detached from the used to make Cytospin preparations, and some of dishes by adding propylene oxide. The cell pellets them were cultured for 24 h in RPMI1640 supple- were embedded in resin and processed as usual for mented with 10 per cent FCS. Peripheral blood and electron microscopy. bone marrow samples obtained from patients without haematological malignancies were fixed with cold acetone for 10 min and subjected to immuno- Estimation of molecular weight Since many infiltrated macrophages in M F H histochemistry. To evaluate interspecies reactivities of PM-2K, lymph nodes and spleens were obtained were positive for PM-2K, the tumour tissue of M F H

PM-ZK, AN ANTI-MACROPHAGE ANTIBODY

was used to estimate the molecular weight of the antigen recognized by PM-2K. Proteins were extracted from the tumour tissue with 1 per cent Nonidet P-40 (Sigma, St Louis, MO, U.S.A.) in PBS containing 1 mM phenylmethylsulphonyl fluoride (Sigma) for 1 h at 4°C. After centrifugation at 20 OOOg for 30 min, the supernatant was mixed with the sample buffer, run on an 8 per cent SDSpolyacrylamide gel, and blotted on a Durapore GVHP membrane (Millipore, Bedford, MA, U.S.A.) as described in ref. 14. After the blotting, the membranes were treated with 10 per cent dried milk, incubated with the culture supernatant of PM2K for 1 h at room temperature, and incubated again with peroxidase-labelled anti-mouse Ig [F(ab’),] (Amersham) for 1 h. Peroxidase activity was visualized using DAB as substrate. RESULTS

Reactions of PM-2K on normal tissues PM-2K was found to react with a wide variety of tissue macrophages in different organs and tissues. In the lymphoid organs, PM-2K recognized most tissue macrophages such as thymic cortical and medullary macrophages (Fig. 1A), splenic red pulp macrophages, sinus macrophages, and interfollicular macrophages in the lymph nodes (Figs 1B and 1C). Macrophages in most non-lymphoid organs were also positive for PM-2K. Kupffer cells of the liver (Fig. 1D), alveolar macrophages, and omentum macrophages were strongly labelled. Macrophages in the interstitium of the kidney (Fig. IE), heart, thyroid, adrenal, salivary glands, and pancreas were also positively stained. However, tingible body macrophages in the germinal centres, microglial cells, and osteoclasts were not reactive. By double staining with anti-S-100 protein, PM-2K was found not to be reactive with interdigitating cells in the thymic medulla, splenic white pulp, or paracortical area of the lymph nodes (Fig. 1C). Langerhans cells in the epidermal sheets or skin sections were also negative for PM-2K.Cells other than macrophages including all epithelial cells were invariably negative for PM-2K. Reactions of PM-2K with peritoneal macrophages, alveolar macrophages, and blood cells N o positive cells were found in freshly obtained peritoneal cells. After 24 h culture, however, about 10 per cent of the adherent cells became positive for

317

PM-2K (Fig. 2A), whereas most alveolar macrophages (more than 90 per cent) were labelled by PM2K without cultivation (Fig. 2B). This positivity was not changed after 24 h of culture. No peripheral blood cells or bone marrow cells including monocytes, myelomonocytic precursor cells, and megakaryocytes were labelled with PM-2K. Even after 5 days in culture, the epitope against PM-2K did not appear on blood monocytes.

Reactivities of PM-2K in proliferative disorders of macrophages In Gaucher’s disease, multicentric reticulohistiocytosis, and malignant histiocytosis, proliferating macrophages were clearly labelled by PM-2K. The antibody was strongly reactive with infiltrating macrophages in MFH, while the tumour cells were not stained (Fig. 3). In GCT of the bone, PM-2K recognized some stromal cells, whereas it did not label osteoclast-like multinucleated giant cells (Fig. 4). The reactivity of PM-2K was the same for all cases of each disease. Reactivities against macrophages ofother species PM-2K was reactive with macrophages of cats and beagles. No PM-2K-positive cells were found in other species including rats, guinea pigs, hamsters, rabbits, goats, horses, and cows. Immunoelectron microscopic localization of PM-2K reaction Reaction products of PM-2K were clearly demonstrated by immunoelectron microscopy on the cytoplasmic membrane of peritoneal macrophages cultured for 24 h (Fig. 5). These PM-2K-positive macrophages showed no endogenous peroxidase activity. Molecular weight ofantigen and isotype of PM-2K Under non-reducing conditions, PM-2K immunostained a single protein band corresponding to 150 kD when the tissue extract of malignant fibrous histiocytoma was used as the antigen (Fig. 6). After addition of mercaptoethanol, no positive bands against PM-2K were observed. The isotype of PM-2K was IgGI, R. DISCUSSION In this study, a monoclonal antibody, PM-2K, with high specificity for tissue macrophages was

Fig. I-Light microscopic immunohistochemical reactivities of PM-2K in normal tissues (immunoperoxidase, counterstained with haematoxylin except for (C)). (A) Cortical (c) and medullary (m) macrophages in the thymus are positively stained. (B) In the lymph node, macrophages in the sinus (s) and interfollicular area are labelled. Tingible body macrophages in the follicles (f)are not stained. (C) Double staining of the lymph node with PM-2K and anti-S-100. Anti-S-100 (brown) and PM-2K (blue) label different cell populations in the paracortical area (immunoperoxidase and APAAP methods, no counterstaining). (D) Positive staining of Kupffer cells in the liver. (E) Macrophages between renal tubules and a few mesangial cells are positive Fig. 2-Reactivity of PM-2K with peritoneal and alveolar macrophages (immunoperoxidase, counterstained with haematoxylin). (A) A few peritoneal macrophages became positive after 24 h culture. (B) Most freshly isolated alveolar macrophages are positive Fig. 3-In MFH, only infiltrated macrophages are labelled with PM-2K (immunoperoxidase, counterstained with haematoxylin) Fig. 4-Stromal cells of GCT of the bone are stained by PM-2K. Multinucleated giant cells are not labelled (immunoperoxidase, counterstained with haematoxylin)

PM-ZK, AN ANTI-MACROPHAGE ANTIBODY

319

Fig. 5-Immunoelectron microscopic demonstration of PM-2K reaction on an adherent peritoneal macrophage. Reaction products of PM-2K are observed on the cytoplasmic membrane. No staining is observed in the intracytoplasmic space, organelles, or nucleus (immunoperoxidase staining, counterstained with uranyl acetate, bar = 1 pm)

A

200 ,o

-

B

4 1 50 ,O

116,3*

97,4* 66.2

*

42,6* Fig. &Western blotting of the PM-2K antigen in MFH. (A) Molecular weight standard. (B) PM-2K stains a single band of 150kD

produced using adherent human peritoneal macrophages as the immunogen. Immunohistochemically, PM-2K recognized a wide variety of tissue macrophages in various organs. However, tingible body macrophages in the germinal centre, microglial cells, and osteoclasts were not labelled by PM-2K. Interestingly, the epitope of PM-2K did

not appear on monocytes even after 5 days in culture. The intracellular distribution of endogenous peroxidase activity has been used to distinguish the differentiation stages of peritoneal macroImmunoelectron microscopic study showed that PM-2K-positive macrophages possess no endogenous peroxidase activity. Exudate macrophages characterized by the presence of peroxidasepositive lysosomal granules were not labelled by PM-2K. These findings indicate that the antigen recognized by PM-2K appears at a late differentiation stage of the macrophages. Among the antimacrophage monoclonal antibodies reported, 24" and Ki-MS" are strongly reactive with macrophages but only weakly with monocytes. In contrast to PM-2K, however, the molecules recognized on monocytes by antibody 24 and Ki-M8 increased when they were cultured'' or stimulated.'' The reactivities of Ki-M8 and PM-2K with tissue macrophages are very similar; however, the former recognizes an intracytoplasmic antigen'' and the latter a membrane antigen. An anti-macrophage monoclonal antibody, 25F9, failed to react with freshly isolated monocytes, but recognized monocytes after 3 days in culture." Accordingly, PM-2K was thought to be a unique antibody not previously reported.

320

M. TAKEYA E T A L .

Peritoneal macrophages and alveolar macrophages are very different in function, activities of various enzymes, and cell surface properties.’0321As indicated by the fact that the alveolar macrophages exhibit a higher activity for most lysosomal enzyme^,'^.^^ alveolar macrophages are considered to be in a more activated state than peritoneal macrophages because of their direct exposure to exogenous materials. It is worth noting that the epitope of PM-2K was not observed on the freshly isolated peritoneal macrophages, but was already present on freshly obtained alveolar macrophages. Further experiments need to be done to elucidate the functional properties of the PM-2K antigen. The reactivity of previously reported antimonocyte/macrophage monoclonal antibodies with dendritic cells is variable. For example, CD14 and CD35 are reactive with follicular dendritic cells,24 and antibodies 63D3, 61D3,25 Ki-M1,26 and 2417 with interdigitating cells. On the other hand, KiM7,27Ki-M8,’* PHM2, and PHM328 are not reactive with any dendritic cells. The reactivities of CD68 antibodies with dendritic cells are not uniform. EBM 1 1 is reactive with Langerhans cells and interdigitating cells,29 whereas Ki-M6,” Y1/82A,” and KP1” do not label dendritic cells. In the present study, PM-2K did not recognize any dendritic cells in epidermal sheets. The absence of reactivity of PM-2K with interdigitating cells in the thymus-dependent area of lymphoid organs was confirmed by double staining with S- 100 protein, one of the most reliable markers for interdigitating ~ ~ 1 1 sAccordingly, .’~ PM-2K is considered to be one of the markers which can discriminate between macrophages and dendritic cells. In studies of pathological tissue, PM-2K stained proliferating macrophages of both reactive and malignant disorders such as Gaucher’s disease, multicentric reticulohistiocytosis, and malignant histiocytosis. In the tumour tissue of MFH, infiltrating macrophages were clearly labelled with PM-2K, but tumour cells were not stained. This indicates a non-histiocytic origin of MFH as indicated by us in the experimental rat MFH.34Interestingly, normal osteoclasts in normal bone marrow and osteoclast-like giant cells in GCT were unreactive with PM-2K. As some other monoclonal anti-macrophage antibodies also fail to recognize osteoclasts and osteoclast-like giant the association of osteoclasts and cells of the mononuclear phagocyte system needs to be re-evaluated. It is not surprising that PM-2K labelled some of the mononuclear stromal cells, because some stromal

cells express Ia- and monocyte/macrophage antigens,36.37 although it is still controversial whether these cells are reactive or neoplastic. Many monoclonal antibodies which recognize human monocytes/macrophages have been reported to date; however, most of them are cross-reactive with other cells. For example, CD13, CD16, CD31, CD35,24.38 and MAC38739are reactive with granulocytes. Antibodies CD31, CD32, and CD36 label platelet^,^^^^^ and CD13 and CD35 recognize some epithelial cells.24 The CD68 antibodies are considered to be one of the most reliable markers for monocyte/macrophages. However, they are also cross-reactive with cells other than macrophages, such as m e g a k a r y o ~ y t e s , ~the ~ , ~epithelium ’ of renal proximal tubule^,^^,^' hepat~cytes,~’ and some granu~ocytes.’~~~~ Through the immunohistochemical observation of various human tissues, PM-2K was shown to be specific against tissue macrophages. Cells other than macrophages, such as epithelial cells, endothelial cells, granulocytes, platelets, and megakaryocytes, were not recognized by PM-2K. Accordingly, PM2K is considered to be one of the most reliable markers for tissue macrophages. It should be very useful not only in investigating macrophage differentiation and maturation, but also in identifying macrophages in inflammatory, reactive, and neoplastic conditions. REFERENCES I . Carr I. Daems WT (eds). The macrophage: a bird’s-eye view. In: The Rettculoendothehal System, Vol. 1 . New York and London: Plenum, 1980: 1-17. 2. van Furth R, Cohn ZA. The origin and kinetics of mononuclear phagocytes. J Exp Med 1968; 128: 415435. 3. Forster 0, Landy M. Heterogeneity of Mononuclear Phagocytes. London: Academic Press, 1981 4. Hashimoto K, Tarnowski WM. Some new aspects of the Langerhans cells. Arch Dermatol 1968; 97: 4 5 W 6 4 . 5 Veerman AJP. On the interdigitating cells in the thymus-dependent area of the rat spleen: a relation between the mononuclear phagocyte system and T lymphocytes. Cell Tissue Res 1974; 148 247-257. 6. Steinman RM, Cohn ZA. Identification of a novel cell type in peripheral lymphoid organs of mice. 1. Morphology, quantitation, tissue distribution. J E.rp Med 1973; 137: 1142-1 162. 7. Steinman RM. Dendritic cells. 7‘ran.rpkantation 1981; 31: 151-155. 8. Hogg N. Human mononuclear phagocyte molecules and the use of mononuclear antibodies in their detection. Clin E.xp lmmunol 1987: 69 687-694. 9. Miyamura S, Okamura H, Naito M , Takahashi K . Ultracytochemical localization of peroxidase activity in human peritoneal macrophages. J Clin Electron Microsc 1987; 2 0 5 4 . 10. Takeya M, Hsiao L, Takahashi K . A new monoclonal antibody, TRPM-3, binds specifically to certain rat macrophage populations: immunohistochemical and immunoelectron microscopic analysis. J Leuk Biol1987;41: 187-195. I I . Hsiao L, Takeya M, Arao T, Takahashi K. An immunohistochemical and immunoelectron microscopic study of the ontogeny of rat

PM-2K, A N ANTI-MACROPHAGE ANTIBODY

12.

13.

14. 15. 16.

17.

18.

19.

20. 21.

22. 23. 24.

25. 26

Langerhans cell lineage with anti-macrophage and anti-Ia monoclonal antibodies. J Invesr Dermntul 1989; 93: 78&786. lsobe S, Nakane PK, Brown WR. Studies on translocation of immunoglobulins across intestinal epithelium. 1. Improvements in the peroxidase-labeled antibody method for application to study of human intestinal mucosa. Acra Histochem C p c h e m 1977; 10: 167-1 71. Cordell JL, Falini 8 , Erber WN, er a / . lmmunoenzymatic labeling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP complexes). J Hisrochem Cjrochem 1984; 32: 219-229. Shimokawa Y, Takeya M, Miyauchi Y, Takahashi K. A monoclonal antibody, RbM2, specific for a lysosoiiial membrane antigen ofrahbit monocyteirnacrophages. Immunologj 1990; 7 0 5 13-5 19. Beelen RHJ, Fluitsma D M , van der Meer JWM, Hoefsmit ECM. Development of different peroxidatic activity in peritoneal macrophages in vivo and in vitro. J Reticuloendothel Soc 1979; 25: 5 13-523. Daems WT. Peritoneal macrophages. In: I Carr, WT Daems, eds. The Reticuloendothelial System, Vol. I . New York and London: Plenum, 19x0; 57-127. Hogg N, Selvendran Y. An anti-human monocyte/macrophage antibody, reacting most strongly with macrophages in lymphoid tissue. Crlllmmunol1985; 92: 247-253. Radzun HJ, Kreipe H, Bodewadt S, Hansmann M-L, Barth J, Parwaresch MR. Ki-M8 monoclonal antibody reactive with an intracytoplasmic antigen of nionocyteimacrophage lineage. B h o d 1987; 6 9 1320-1327. Zwadlo G, Brocker E-B, von Bassewitz D-B, Feige U, Sorg C. A monoclonal antibody to a differentiation antigen present on mature human macrophages and absent from monocytes. J lmmuuol 1985; 134 1487-1492. Hocking WG. Golde DW. The pulmonary-alveolar macrophage (first of two parts). N Engl J Med 1979; 301: 58G587. Bowden HB, Adamson YR. Pulmonary macrophages. In: I Carr, WT Daems, eds. The Reticuloendothelial System, Vol. 1. New York and London: Plenum, 1980; 607-633. Cohn ZA, Weiner E. The particulate hydrolases of macrophages. 1. Comparative enzymology, isolation, and properties. J Exp Med 1963; 118: 991-1008. Cohn ZA, Weiner E. The particulate hydrolases of macrophages. II. Biochemical and morphological response to particle ingestion. J Exp Met1 1963; 118: 1009-1020. Hogg, N, Horton MA. Myeloid antigens: new and previously defined clusters. In: AJ McMichael, ed. LeucocyteTyping 111. Oxford: Oxford University Press, 1987; 57C602. Ugolini V, Nunez G, Smith RG, Stastny P, Capra JD. Initial characterization of monoclonal antibodies against human monocytes. Proc Nut/ AradSci U S A 1980; 77: 67646768. Radrun HJ, Parwaresch MR, Feller AC, Hansmann M-L. Monocytei macrophage-specific monoclonal antibody Ki-M 1 recognizes interdigitating reticulum cells. Am J Pathol1984;117: 441450.

32 1

27. Kreipe H, Radzun HJ, Parwaresch MR, Haislip A, Hansmann M-L. Ki-M7 monoclonal antibody specific for myelomonocytic cell lineage and macrophages in human. J Hisrochem Cjrorhem 1987; 3 5 I 1 17-1 126. 28. Becker GJ, Hancock WW, Kraft N, Lanyon HC, Atkins RC. Monoclonal antibodies to human macrophage and leukocyte common antigens. Pathologj 1981; 13: 669-680. 29. Kelly PMA, Bliss E, Morton JA, McGee JO'D. Monoclonal antibody EBM/I I : high cellular specificity for human macrophages. J Clin Puthol1988; 41: 51G515. 30. Parwaresch MR, Radzun HJ, Kreipe H , Hansmann M-L, Barth J. Monocyte/macrophage-reactive monoclonal antibody Ki-M6 recognizesan intracytoplasmicantigen. A m J Pathol1986; 124 141-151. 31. Davey FR, Cordell JL, Erber WN, Pulford KAF, Gatter KC, Mason DY. Monoclonal antibody (Y l/82A) with specificity towards peripheral blood monocytes and tissue macrophages J Clin Purhol 1988; 41: 753-758. 32. Pulford KAF, Rigney EM, Micklem KJ, el ul. KPI: a new monoclonal antibody that detects a monocyte/macrophage associated antigen in routinely processed tissue sections. J C/in Putholl989;42: 4 1 U 2 1 . 33. Wood GS, Turner RR, Shiurba RA, Eng L, Warnke RA. Human dendritic cells and macrophages. A m J Parhol1985; 1 1 9 73-82. 34. Kato T, Takeya M, Takagi K, Takahashi K . Chemically induced transplantable malignant fibrous histiocytoma of the rat: analysis with rmmunobistochemistry, immunoelectron microscopy and ['Hlthymidine autoradiography. Lab lnvrsr 1990; 62: 635-645. 35. Athanasou NA, Bliss E, Gatter KC, Heryet A, Woods CG, McGee JO'D. An immunohistological study of giant-cell tumour of bone: evidence for an osteoclast origin of the giant cells. J Pathol 1985; 147: 153-158. 36. Ling L, Klein MJ, Sissons HA, Steiner GC, Winchester RJ. Expression of la and monocyte-macrophage lineage antigens in giant cell tumor ofboneand related lesions. Arch Pathol Lab Med 1988; 112 65-69. 37. Tze-Chun L, Zhong-Min J, Liang-Tang W. Giant cell turnours of bone. An immunohistochemical study. Purhol Res Pract 1989; 185 448453 38. Stain C, Jager U, Majdic 0, er al. The phenotyping of human basophils with myeloid workshop panel. In: AJ McMichael, ed. Leucocyte Typing 111. Oxford: Oxford University Press, 1987; 72&722. 39. Flavell DJ, Jones DB. Wright D H . Identification of tissue histiocytes on paraffin sections by a new monoclonal antibody. J Histvchem Cytochem 1987;45: 1217-1226. 40. Horton MA, Hogg N. Platelet antigens: new and previously defined cluaters. In: AJ McMichael, ed. Leucocyte Typing 111. Oxford: Oxford University Press, 1987; 733-745. 41. Pulford KAF, Sipos A, Cordell JL, Stross WP, Mason DY. Distribution of the CD68 macrophage/myeloid associated antigen. In/ Immuno/1990; 2 973-980.