Clin. exp. Immunol. (1988) 72, 351-356

An analysis of myeloma phenotype using defined at the IIrd international workshop on human leucocyte differentiation

N. JACKSON, N.R. LING,* JENNIFER BALL, ELAINE BROMIDGE, P.D. NATHAN* & I.M. FRANKLIN Departments of Haematology and *, The Medical School, Birmingham

(Acceptedfor publication 12 January 1988)

SUMMARY Fresh bone marrow from 43 cases of myeloma and three cases of plasma cell leukaemia has been phenotyped both by indirect immune-rosetting and, on fixed cytospin preparations, by indirect immunofluorescence. Both clustered and unclustered associated antibodies from the IIrd International Workshop on Human Leucocyte Differentiation Antigens were used. The results confirm the lack of many pan-B antigens on the surface of myeloma plasma cells, i.e. CD19-23, 37, 39, w40. Strong surface reactivity is seen with CD38 antibodies and with one CD24 (HB8). Weak reactions are sometimes obtained with CD9, 10 and 45R. On cytospin preparations CD37, 39 and w40 are sometimes weakly positive, and anti-rough endoplasmic reticulum antibodies are always strongly positive. Specific and surface-reacting antiplasma cell antibodies are still lacking. Keywords myeloma phenotype antibody cluster

INTRODUCTION antibodies from CD24 (HB8 and VIB-E3). Before and after the IIrd workshop, we carried out a phenotypic analysis of fresh Multiple myeloma (MM) is a neoplastic disorder of the B cell MM bone marrow cells using both clustered and unclustered lineage. It is characterized by an excess ofneoplastic plasma cells monoclonal antibodies (MoAb) supplied for the workshop, in the bone marrow, although the nature of the clonogenic cell including one from each ofthe new and the previously defined B remains uncertain. On terminal differentiation to plasma cells, cell associated clusters. Some of these results were presented at cells of the B lineage cease to express many characteristic B cell the Workshop (Nathan et al., 1987). antigens. These include antigens detected by antibodies of clusters CD19-23 as defined at the 1Ind International Work- shop on Human Leucocyte Differentiation Antigens (Nadler, METHODS AND PATIENTS 1986) and surface immunoglobulin (sIg) (Katagiri et al., 1984; Foon, Schroff & Gale, 1982). The majority of plasma cells also Patients lack major histocompatibility complex (MHC) class II antigens Forty-six patients who fulfilled the Medical Research Council (Wu et al., 1976; Kuritani & Cooper 1982; Katagiri et al., 1984), criteria for entry into the IVth Myelomatosis trial (Medical and Fc and C3 receptors (Stashenko et al., 1980). Recently, Research Council, 1985) were studied, including three with some cases of myeloma have been found to express CDIO plasma cell leukaemia (i.e. peripheral blood plasma cell count (CALLA; common acute lymphoblastic leukaemia ) on .2 x 109/1). Their paraprotein type and treatment status are some of the cells in blood and bone marrow (Ruiz-Arguelles et shown in Table 1. al., 1984; Durie & Grogan, 1985; Caligaris-Cappio et al., 1985). The I1Ird International Workshop (McMichael et al., 1987) Surface antigen detection: indirect immune-rosetting defined five new B cell associated antibody clusters (CD 37-w40 Indirect immune-rosetting was used for surface antigen detec- and 45R), and also included studies on other antibodies which tion, a technique well established in this laboratory (Ling, were not clustered but which were found to react with plasma Bishop & Jefferis, 1977; Kalland, 1977). The MoAb, as listed in cells. Plasma cell reactivity was tested mainly against plasmacy- Table 2, were diluted (1 Ml in 50 pi of phosphate-buffered saline toid cell lines (JVM-3, U-266, HFB-1, RPMI-8226). Cells from (PBS) containing 10% fetal calf serum (FCS) and 0-1 % sodium only two cases of MM were investigated and they showed azide. An anti-idiotype MoAb generated against an unrelated moderate reactivity with antibodies of CD38, CD39 and two myeloma paraprotein was used as a control. Then, 2 5 x I05 cells to be analysed were added (in 5-20 p1 of culture medium) and Correspondence: Dr N. Jackson, Department of Haematology, incubated at 4°C for 30 min. After washing twice in Hepes- Royal Free Hospital, Pond Street, London NW3 2QG, UK. buffered RPMI 1640 medium with 2% FCS, the cells were mixed 351 352 N. Jackson et al. with approximately 10 x 106 sheep erythrocytes coated with a Table 1. Patient characteristics with their paraprotein sheep anti-mouse immunoglobulin in 150 p1 of the same types medium. This mixture was left to stand at room temperature for 10 min, centrifuged at 100 g for 30 s, and left to stand for a Myeloma Plasma cell leukaemia further 10 min. One drop of Acridine Orange (18 75 pg/ml) was added and after gently resuspending the cells, one drop was Paraprotein Untreated Treated Untreated spread under a coverslip on a multi-spot Polytetrafluoroethy- lene-coated slide. The percentage of rosetted non-myeloid cells IgG 15 11 1 was enumerated by counting the number of white cells with at IgA 4 7 2 least four SRBC attached out of a total of 100 or 200 cells. A Light chain 1 5 fluorescence microscope (Leitz dialux, Leitz Instruments, Bed- only ford UK) was used, set for Fluorescein excitation by epi- illumination to visualize the fluorescing nuclei, with slight background illumination with ordinary light to visualize the Cytoplasmic antigen detection: indirect immunofluorescence erythrocytes. As Acridine Orange makes myeloid cell cytoplasm Approximately 1-2 x 105 cells were cytospun onto slides, and fluoresce red under these circumstances, these cells were elimi- fixed in acetone for 10 min. The slide was first incubated with 1 nated from the counting in order to concentrate on the plasma p1 ofMoAb in 50 pi PBS with 10% FCS at room temperature for cells, which were also recognized by their characteristic morpho- 30 min, and then washed for 30 min in PBS. The slide was then logy. However, this led to some other non-myeloid cells being incubated with a I-in-100 dilution of fluorescein-conjugated counted, e.g. lymphocytes. Hence, small percentage values sheep anti-mouse Ig (prepared in this department by Dr G. D. (< 10%) were regarded as negative; counts between 10-30% as Johnson) for 30 min, and washed in PBS for 30 min. The slides weak positivity of plasma cells with that antibody; counts were then dipped in propidium iodide (1 pg/ml) for 5 min for > 30% as definite positivity with that antibody. nuclear counterstaining, and mounted using polyvinyl alcohol

Table 2. Antibodies used, their cluster designation (CD) where applicable, and the sources from which they were obtained

Antibody CD Source

control An anti-idiotype MoAb produced in the Immunology lab, Birmingham University BU 16 9 Immunology lab, Birmingham University Anti-CALLA 10 Becton-Dickinson* BU12 19 Immunology lab, Birmingham University BCI 20 Gallart, Barcelona BA5 21 Le Bien, Minnesota 29-110 22 Kraft, Melbourne MHM6 23 McMichael, Oxford ALB9 24 Boucheix, Villejuif, France HB8 24 Cooper, Alabama HB9 24 Cooper, Alabama VIB C5 24 Knapp, Vienna VIB E3 24 Knapp, Vienna ALIa 24 Ravoet, Brussels LC66 24 Ravoet, Brussels CLB/granBly I 24 Tetteroo, Amsterdam Anti-IL2 receptor 25 Becton-Dickinson* WR-17 37 Moore, Southampton OKTIO 38 American Type Culture Collection HB7 38 Tedder, Boston T16 38 Bourel, Rennes, France T168 38 Bourel, Rennes, France AC2 39 Rowe, Birmingham G28.5 w40 Ledbetter, Seattle 4KB5 45R Pulford, Oxford PCA-l Nadler, Boston BU25 Anti-MHC class II Immunology lab, Birmingham University KB61 Pulford, Oxford 7F7 Schulz, Innsbruck BU 11,18 Anti-RER Immunology lab, Birmingham University RFD6 Anti-RER Janossy, London

* Purified antibody; others were ascitic fluid. Myeloma phenotype 353

WeaNs | W*-90r-.ve positive 1_ Definitely positiVe CD 10 20 30 40 50 60 70 80 90 100 _~~~,.~~ I 9 V 10 ,i Dea so l 19 20 I

21 I t 22 la I 23 I 24 (HB8) 24 (except HB8) s

25 o3 ,

37 38

39

w40 it 45R

Fig. 1. Surface reactivity ofmyeloma plasma cells with clustered antibodies: percentage ofpositive cells by indirect immune-rosetting. (0) Untreated; (0) treated; CD, cluster designation.

with 1, 4-diazobicyclo-(2,2,2)-octane (Johnson et al., 1982). The Table 3. Cytoplasmic reactivity of myeloma slides were examined using a fluorescence microscope set for plasma cells with clustered antibodies fluorescein excitation (fluorescein, green; propidium iodide, red), and results graded as negative, weak, moderate, or strong. Cytoplasmic reactivity This method will also stain surface antigens that are not destroyed by acetone fixation. CD No. Negative Weak Moderate Strong

9 1 8 I RESULTS 10 8 2 19 6 4 20 8 Reactivity ofmyeloma plasma cells with clustered antibodies 21 6 Bone marrow mononuclear cells from 46 patients have been 22 9 1 analysed with various combinations of antibodies. The 16 most 23 4 2 recent samples have been tested against the full spread of B cell 24 (HB8) 6 6 2 24 (other) 6 antibody clusters, as defined by the IlIrd International Work- 25 12 shop. The results for antibodies from these clusters are shown in 37 7 1 Fig. 1 (surface typing), and Table 3 (cytoplasmic typing). In the 38 6 11 13 majority of cases the plasma cells were HB8+, but most other 39 5 3 antibodies from CD24 gave negative results. Most cases (26/36) 40 3 3 I were CD38 (OKT10)+, but some were negative by indirect resetting. In eight out of the 10 CD38- cases, the pattern of CD, cluster designation. positivity ofthe immunofluorescence on the cytospins suggested that the cells were in fact surface CD38+. One case was CDl0+ CDl9+, CD20+, and CD23+, and occasional weak surface Reactivity ofmyeloma plasma cells with unclustered antibodies reactivity was also obtained with CD9 (three cases), CD25 (two The results using unclustered antibodies are shown in Fig. 2 and cases), CDw4O (two cases), CD45R (five cases). Reactivity on Table 4. Antibody KB61 showed definite surface reactivity in cytospin preparations was found with CD38 and HB8 (strong), six/ 18 cases, and weak surface reactivity in a further five cases. and CD9, 23, 37, 39 and w40 (weak to moderate). There was very weak MHC Class II expression in a few cases. 354 N. Jackson et al.

Negative | Week positive II Definitely positive 1

Antibody 0 10 20 30 40 50 60 70 80 90 100 T-~ I I I I I I I I J05 00 00 1 I 0 (MHC Class II) 00 @00 O00 0 I

00 00 0 7F7 00 o I 00 *0

0 1 . *0 KB 61 0 0* I 00 00 I BUll ..I 0f 000*90 1 00

00 00 00 a BU 18 00 000 *: *0

00 00 0o 00 RFD 6 08 0S0 00 oo PCA.I o f

Fig. 2. Surface reactivity ofmyeloma plasma cells with unclustered antibodies: percentage ofpositive cells by indirect immune-rosetting. Jo5 is an anti-MHC class II MoAb; BUI 1, BU18 and RFD6 are directed against determinants on rough endoplasmic reticulum. (0) Untreated; (-) treated.

Table 4. Cytoplasmic reactivity of myeloma DISCUSSION plasma cells with unclustered antibodies These results confirm that most cells from the majority of cases of myeloma are CD38+, HB8(CD24)+, and CDlO,19-23- (cf Cytoplasmic reactivity Nadler, 1986; Katagiri et al., 1984). Antigens of the newly MoAb Negative Weak Moderate Strong designated clusters CD37, 39, w40, and 45R are also not usually present. However there is some heterogeneity of phenotype, Jo5 11 with a few cases showing weak surface reactivity with CD9, 7F7 6 2 CD1O, CD19, CD20, CD23, CD25 and CD45R. The unclus- KB61 4 2 2 3 tered antibody KB61 (Pulford et al., 1987), which is B cell BU 11,18 2 6 3 5 associated, binds to the neoplastic cells ofabout halfthe cases of RFD6 3 6 3 myeloma. PCA-I 6 CD9 Jo5 is an anti-MHC class MoAb, BUllI, This antigen is expressed on a number of different haemopoietic BU18 and RFD6 are directed against determi- cells including platelets, neutrophil and eosinophil polymorpho- nants on rough endoplasmic reticulum. nuclear leucocytes, B cells (weak), common acute lymphoblastic and non-lymphoblastic leukaemia blasts and a range of other Antibodies reactive with determinants on rough endoplasmic tissue cells (Horton & Hogg, 1987). Its previously reported reticulum (BU 1, BU 18, RFD6) were strongly reactive with the presence on monocytes is now disputed (Ohto et al., 1987). San cytoplasm but negative with the cell surface. Antibody 7F7, Miguel et al. (1986) found surface expression of FMC8 (CD9) in which reacted with plasmacytoid lines in the Workshop was 32 out of 37 cases of myeloma. In contrast, we found only 3/19 negative on myeloma cell surface and cytoplasm. Antibody cases with a weak positive surface reaction. Katagiri et al. (1984) PCA-I was found unreactive with myeloma cells, but there was also demonstrated the reactivity of myeloma cells with BA2 known to be a problem with this antibody as supplied to the (CD9) in three patients, but Caligaris-Cappio et al. (1985) failed Workshop, many centres finding it unreactive in most tests (e.g. to demonstrate any BA2+ myeloma cells out of 25 cases, even Johnson et al., 1987). Three cases of true plasma cell leukaemia though 1 1 of these cases expressed CD1O. The technique we use are included in this study: their phenotype was not different for demonstrating cytoplasmic antigens will also identify sur- from the other cases, i.e. CD38+, HB8+, CD0-, KB61+/-. face antigens if they are not destroyed by acetone fixation, and Myeloma phenotype 355 weak CD9 reactivity was found on immunofluorescence of CD38 cytospin preparations. It seems, therefore, that myeloma cells The CD38 antigen occurs on both T and B cells. It was first express CD9 weakly. The function of the CD9 molecule, being described as present on thymocytes and activated T cells found on such a wide variety of cell types, is unknown: CD9 (Reinherz et al., 1980; Reinherz & Schlossman, 1981), but it is antibodies will aggregate platelets (Boucheix et al., 1987), and also present on B cell progenitors (e.g. on common acute do not inhibit polymorph chemotaxis (Ohto et al., 1987). It is lymphoblastic leukaemia and B progenitor lines) and germinal one of a number of antigens shared between plasma cells and centre B cells (Ling et al., 1987, p. 321). OKT1O is the prototype myeloid and/or monocytic cells. antibody, but all the antibodies reacted with the majority of myeloma cells. However, it was again evident that these CDJO antibodies did not always perform well in the indirect rosetting In this series, 1/16 cases contained significant numbers of cells assay. Eight out of 10 cases negative by indirect immune- expressing CD10: this case also weakly expressed CD 19, CD20 rosetting were found to be CD38+ by indirect immunofluores- cence. Little is known and CD23, and was in a terminal relapse after a very prolonged ofthe function ofthe CD38 molecule, but clinical course with multiple treatments. Another nine patients there are other antigens shared between T cells and plasma cells, had between one and 10% non-myeloid CDl0+ cells detectable: e.g. Tp44 (CD28) (Kozbor et al., 1987). CD28 antibodies can it is unlikely that these are all residual normal pre-B cells but induce antigen-independent T cell activation, and although this they might be myeloma progenitors, as others have postulated. antigen is present on plasma cells, CD28 antibodies do not Ruiz-Arguelles et al. (1984) reported 2/14 myeloma cases having appear to induce proliferation of plasma cells (Kozbor et al., > 15% CDl0+ cells in the bone marrow, and these patients' 1987). myeloma cells also expressed other B cell antigens (HLA-Class II, slg, Bi (CD20), and B2 (CD21)). Durie & Unclustered antibodies Grogan (1985) Not found 4/21 cases of myeloma were CALLA(CD10)+ and surprisingly, antibodies reactive with rough-endoplasmic showed that these patients had a worse prognosis. Caligaris- reticulum (RER) (BUll, BU18 and RFD6) show strong with Cappio et al. (1985) found 9-27% CD10+ cells in bone marrows reactivity plasma cell cytoplasm. We were surprised to find from 11/25 cases of myeloma/plasma cell leukaemia using a complete absence of these antigens on the cell surface, as one indirect immunofluorescence on acetone-fixed cytocentrifuge might expect some carry-over ofthe RER antigens onto the cell specimens (the other cases contained < 1 % CDl0+ cells). These surface membrane during Ig secretion. These antibodies have CDI0+ cells were oflymphoid morphology, but did not express also been shown to react with other secretory cells; e.g. thyroid surface or cytoplasmic Ig, CD9, 22, 24, or TdT, but they were acinar epithelium and exocrine pancreatic cells (Johnson et al., HLA-DR+ and CD38+. These cells do not correspond to any 1987). We have confirmed very weak or absent MHC-Class II currently recognized stage ofnormal B cell differentiation. Their expression of myeloma cells, as reported by other workers phenotype is not typical of pre-B cells, but more closely (Halper et al., 1978; Katagiri et al., 1984; San Miguel et al., resembles that ofcentroblasts, which may be precursors ofbone 1986). PCA-1 (Anderson et al., 1983) as supplied to the marrow plasma cells (Maclennan & Gray, 1986). Centroblasts Workshop, appeared to have lost activity and was found are CD10+, CD38+ with low expression of other B cell markers unreactive in most tests undertaken, and no myeloma plasma such as CD23, 24, 39, sIg (Ling, Maclennan & Mason, 1987), cells reacted with PCA-I in our experience. Antibody 7F7 has Caligaris-Cappio et al. (1985) found that the CD10+, Ig- cells restricted B cell reactivity and was shown in the Workshop to could be induced to differentiate into plasma cells with tetrade- bind to three of the four plasmacytoid cell lines, but not to canoyl phorbol acetate, and they suggested that the clonogenic plasma cells in tissue sections (Ling et al., 1987, p. 318). In our precursor cell in myeloma could be CD10+. The terminal studies it reacted with neither the surface nor cytoplasm of aggressive phase ofmyeloma, sometimes associated with cells of myeloma cells. KB61 recognizes mature B cells but not germinal more immature morphology and reduced Ig secretion, could centre B cells, and certain macrophages (Pulford et al., 1987). It represent the outgrowth of this CD10+ cell. reacts with some plasma cells in tissue sections and in our studies it reacted with the surface of myeloma plasma cells in 50% of cases. An unclustered MoAb, 8A, has been reported to react CD24 with virtually all myeloma plasma cells (Tazzari et al., 1987). We Strong surface expression ofthe pan-B and granulocyte antigen have found 8A and a number of MoAb we have made ourselves CD24, as recognized by HB8, was shown by 24/27 samples. This to be strongly reactive with plasma cells, but they are broadly reactivity was not shown by other CD24 antibodies. There is reactive with a number of other cell types, especially on tissue known to be heterogeneity of CD24 antibodies, e.g. in their sections (G. D. Johnson and our unpublished observations). reactivity with vascular endothelium (Ling et al., 1987, p. 312), In conclusion, the majority of myeloma plasma cells have or with neuroblastoma cells, and in their cross-blocking activity been shown to lack expression of the B cell antigens defined by (Stockinger et al., 1987). It could be that it is only the HB8 CDl9-23, 37, 39, and w40. They express strongly the CD38 epitope that is expressed on myeloma cells, perhaps being antigen and the HB8 epitope ofCD24. Weak expression ofCD9 incorporated into another molecule (Ling et al., 1987, p. 312). and CD1O is also occasionally seen. The existence of small and Alternatively, this pattern of reactivity could be due to HB8 sometimes large numbers of CD10+ cells, especially in aggres- having the highest affinity of the cluster. In immunofluorescent sive or late disease is compatible with the myeloma progenitor staining, HB8 always stained myeloid cells more brightly than cell being CD10+. However this does not mean that the plasma cells. PCA-l is also reported to react with both plasma progenitor cell is at the pre-B cell stage: it could be more closely cells and myeloid cells (Anderson et al., 1983), so this is further related to the centroblast, which is CD10+, CD38+. That the evidence of shared antigens between these two cell lineages. progenitor cell is at a late B cell stage is supported by studies on a 356 N. Jackson et al. plasma cell leukaemia cell line that we have recently developed 4128. (JJN-1). The phenotype ofJJN-1 is similar to that ofthe mature KURITANI, T. & COOPER, M.D. (1982) Human B cell differentiation. I. plasma cell, i.e. CD38+, HB8+, CD10-, CDl9-23- (manuscript Analysis of Immunoglobulin M, G, and A antibodies and pokeweed in preparation). We plan to use this cell line to try to make new mitogen induced plasma cell differentiation. J. exp. Med. 155, 839. LING, N.R., BISHOP, S. & JEFFERIS, R. (1977) Use ofantibody-coated red MoAb to supplement the few currently available plasma cell cells for the sensitive detection of antigen and in rosette tests for cells specific antibodies. bearing surface immunoglobulin. J. immunol. Methods 15, 279. LING, N.R., MACLENNAN, I.C.M. & MASON, D.Y. (1987) B cell and ACKNOWLEDGMENTS plasma cell antigens: new and previously defined clusters. In: We would like to thank all those who contributed antibodies to the Leucocyte Typing III (ed. by A.J. McMichael et al.) p. 302. Oxford Workshop, which were used in this study. The advice and encourage- University Press, Oxford. ment of Dr G. D. Johnson and Professor I.C.M. Maclennan, and the MACCLENNAN, I.C.M. & GRAY, D. (1986) Antigen-driven selection of skilled technical assistance of Miss Jenny Lowe and Mrs Brenda Brown virgin and memory B cells. Immunol. Rev. 91, 61, are gratefully acknowledged. Financial support was received from the MCMICHAEL, A.J. et al. (Eds) (1987) Leucocyte Typing III. White Cell Leukaemia Research Fund and the Central Birmingham Health Differentiation Antigens. Oxford University Press, Oxford. Authority Trust Funds. Dr N. Jackson is a Sheldon Clinical Research MEDICAL RESEARCH COUNCIL WORKING PARTY ON LEUKAEMIA IN Fellow of the West Midlands Regional Health Authority. ADULTS (1985) Objective evaluation of the role of vincristine in induction and maintenance therapy for myelomatosis. Br. J. Cancer 52, 153. REFERENCES NADLER, L.M. (1986) B cell/Leukaemia Panel Workshop: Summary and ANDERSON, K.C., PARK, E.K., BATES, M.P., LEONARD, R.C.F., HARDY, Comments. In: Leucocyte Typing II, Volume 2 Human BLymphocytes R., SCHLOSSMAN, S.F. & NADLER, L.M. (1983) Antigens on human (ed. by E.L. Reinherz, B.F. Haynes, L.M. Nadler & I.D. Bernstein) p. plasma cells identified by monoclonal antibodies. J. Immunol. 130, 3. Springer Verlag, New York. 1132. NATHAN, P.D., JACKSON, N., FRANKLIN, I.M., BALL, J. & JOHNSON, BoucHEIx, C., BENOIT, P., KRIEF, P., BILLARD, M., MISHAL, Z., G.D. (1987) Myeloma cell phenotyping with anti-plasma cell, T10, AZzARONE, B., RENDU, E., ESNOUF, J., HERMANT, L., BREDOux, R., CD24 and rough endoplasmic reticulum monoclonal antibodies. In: LEVY-TOLEDANO, S., SORIA, C., PERROT, J.Y., MIRSHA~i, M., GIAN- Leucocyte Typing III (ed. by A.J. McMichael et al.) p. 506. Oxford NONI, F., BERNADOU, A. & SORIA, J. (1987) Platelet aggregation University Press, Oxford. induced by CD9 mAbs. Mechanisms and comparison with platelet OHTO, H., HITONI, Y., KAMBAYASHI, H., SATO, T. & MAEDA, H. (1987) aggregating properties of mAbs directed against other membrane CD9 defined antigen (p24) is expressed on eosinophils but not on proteins. In: Leucocyte Typing III (ed. by A.J. McMichael et al.) normal monocytes. In: Leucocyte Typing III (ed. by A.J. McMichael p.780. Oxford University Press, Oxford. et al.) p. 731. Oxford University Press, Oxford. CALIGARIS-CAPPIO, F., BERGUI, L., TEsio, L., PIzzoLo, G., MALAVASI, PULFORD, K.A.F., RALFKIAER, E., MACDONALD, S.M., ERBER, W.N., F., CHILOSI, M., CAMPANA, D., VAN CAMP, B. & JANOSSY, G. (1985) FALINI, B., GATTER, K.C. & MASON, D.Y. (1987) A unique monoclon- Identification ofmalignant plasma cell precursors in the bone marrow al antibody recognizing a 40 kDa antigen shared between a subpopu- of multiple myeloma. J. Clin. Invest. 76, 1243. lation of B-lymphocytes and macrophages. In: Leucocyte Typing III DURIE, B.G.M. & GROGAN, T.M. (1985) CALLA-positive multiple (ed. by A.J. McMichael et al.) p. 358. Oxford University Press, myeloma: an aggressive subtype with poor survival. Blood 66, 229. Oxford. FOON, K.A., SCHROFF R.W. & GALE, R.P. (1982) Surface markers on REINHERZ, E.L., KUNG, P.C., GOLDSTEIN, G., LEVEY, R.H. & SCHLOSS- leukaemia and cells: recent advances. Blood 60, 1. MAN, S.F. (1980) Discrete stages of human intrathymic differentia- HALPER, J., Fu, S.M., WINCHESTER, C.Y. & KUNKEL, H.G. (1978) tion: analysis ofnormal thymocytes and leukemic lymphoblasts ofT- Patterns of expression of human Ia-like antigens during the terminal cell lineage. Proc. natn Acad. Sci. USA. 77, 1588. stage of B cell development. J. Immunol. 120, 1480. REINHERZ, E.L. & SCHLOSSMAN S.F. (1981) The characterization and HORTON, M.A. & HOGG, N. (1987) Platelet antigens: new and previously function of human immunoregulatory T lymphocyte subsets. Immu- defined clusters. In: Leucocyte Typing III, (ed. by A.J. McMichael et nol. Today 2, 69. al.) p. 733. Oxford University Press, Oxford. RuIz-ARGUELLES, G.J., KATZMANN, J.A., GREIPP, P.R., GONCHOROFF, JOHNSON, G.D., DAVIDSON, R.S., MCNAMEE, K.C., RUSSELL, G., N.J., GARTON, J.P. & KYLE, R.A. (1984) Multiple myeloma: circulat- GOODWIN, D. & HOLBORROW, E.J. (1982) Fading of immunofluore- ing lymphocytes that express plasma cell antigens. Blood 64, 352. sence during microscopy: a study of the phenomenon and its remedy. SAN MIGUEL, J.F., CABALLERO, M.D., GONZALEZ, M., ZOLA, H. & J. Immunol. Meth. 55, 231. LOPEZ BORRASCA, A. (1986) Immunological phenotype of neoplasms JOHNSON, G.D., MACLENNAN, I.C.M., LING, N.R., NATHAN, P.D. & involving the B cell in the last step of differentiation. Br J. Haematol. HARDIE. D.L. (1987) Immunohistological analysis of the plasma cell 62, 75. Workshop antibodies on normal lymphoid tissue and secretory STASHENKO, P., NADLER, L.M., HARDY, R. & SCHLOSSMAN, S.F. (1980) epithalia. In: Leucocyte Typing III (ed. by A.J. McMichael et al.) p. Characterization of a human B lymphocyte-specific antigen. J. 507. Oxford University Press, Oxford. Immunol. 125, 1678. KALLAND, T. (1977) A simple and rapid method to distinguish between STOCKINGER, H., MAJDIC, O., KOLLER, U. & KNAPP W. (1987) rosettes and non-specific aggregation of erythrocytes in the rosette Subclustering of the CD24 cluster, In: Leucocyte Typing III (ed. by assay for human T lymphocytes. J. Immunol. Methods. 17, 279. A.J. McMichael et al.) p. 367. Oxford University Press, Oxford. KATAGIRI, S., YONEZAWA, T., TAMAKI, T., KANAYAMA, Y., KUYAMA, J., TAZZARI, P.L., GOBBI, M., DINOTA, A., BONTADINI, A., GRASSI, G., OHNISHI, M., TSUBAKIO, T., KURATA, Y. & TARUI, S. (1984) Surface CERATO, C., CAVO, M., PILERI, S., CALIGARIS-CAPPIO, F. & TURA, S. expression of human myeloma cells: an analysis using a panel of (1987) Normal and neoplastic plasma cell membrane phenotype. monoclonal antibodies. Acta Haematol. 72, 372. Studies with new monoclonal antibodies. Clin. exp. Immunol. 70, 192. KOZBOR, D., MORETTA, A., MESSNER, H.A., MORETTA, L. & CROCE, Wu, L.Y.F., BIANCO, A., COOPER, M.D. & LAWTON, A.R. (1976) C.M. (1987) Tp44 molecules involved in antigen-independent T cell Ontogeny of lymphocyte differentiation induced by pokeweed mito- activation are expressed on human plasma cells. J. Immunol. 138, gen. Clin. Immunol. Immunopathol. 5, 208.