B29 and mb-1 expression during human B cell differentiation

Eur. J. Immunol. 1992. 22: 2753-2756

2753

Short paper David Y. Mason, Care1 J. M. van NoeseP, Jacqueline L. Cordell, W. Marieke Comans-Bittero., Kingsley Micklem, Albert G. D. Tse+", Ren6 A . W. van LierA and Jacques J. M. van Dongen. Department of Haematology, John Radcliffe Hospital, Headington, Central Laboratory of the Netherlands Red Cross Blood Transfusion Senice., Amsterdam, Department of Paediatricso, Sophia Children's HospitalRJniversity Hospital, Rotterdam, Department of Immunology., Erasmus University, Rotterdam and MRC Cellular Immunology Unit+, Sir William Dunn School of Pathology, Oxford

The B29 and mb-1 polypeptides are differentially expressed during human B cell differentiation* Surface immunoglobulin on mouse B cells is associated with a heterodimer comprising the products of the mb-1 and B29 genes. Here we report that antibodies raised against a peptide sequence from the intracytoplasmic C terminus of the B29 murine gene product detect the 37-kDa component of the human heterodimer, indicating that this component in man is also encoded by the B29 gene. The immunocytochemical reactivity of these anti-B29 antibodies was compared with those of antibodies to the mb-1 protein. Of 25 cases of acute lymphoblastic leukaemia (ALL), 24 were positive for mb-1 whereas B29 was expressed in only 13 cases. Most of these B29-positive ALL expressed immunoglobulin p heavy chain in their cytoplasm (pre-B ALL). In lymphoid tissue sections, anti-B29-labeled B cell follicles in a similar fashion to anti-mb-1, with the striking exception that plasma cells were unreactive for B?9, but positive for mb-1. These results suggest that the synthesis of B29 begins later in precursor B cells than that of mb-1, and ceases before the terminal plasmacyte phase.

1 Introduction In both man and mouse immunoglobulin (Ig) on the surface of B lymphocytes is associated with a heterodimer of invariant polypeptide chains [1-61. In man the 47-kDa Ig-associated chain is known to be encoded by the homologue of the murine mh-1 gene [5, 61. Here we describe the production of antibody to a synthetic peptide corresponding to the B29 gene, which shows that the 37-kDa partner to the 47-kDa chain is encoded (as in the mouse) by the B29 gene [7-101. Immunocytochemical studies revealed both human precursor B cells and plasma cells which express mb-1 protein chains but are negative for B29 chains. This suggests that synthesis of B29 chains during human B cell differentiation may start later than that of mb-1 chains and cease earlier. a phenomenon which may be related to the regulation of surface membrane Ig expression.

2 Materials and methods 2.1 Cells and tissue samples

Acetone-fixed cryostat sections were prepared as described elsewhere [ 111 and paraffin-embedded tissue sections were prepared from human tonsil exposed to Carnoy's fixative overnight. Bone marrow aspirates were obtained at diag-

nosis from patients with multiple myeloma and mononuclear cells were separated by gradient centrifugation. Bone marrow and blood samples from cases of acute lymphoblastic leukaemia (ALL) were obtained from one of the authors' (JJMvD) institutions as were precursor B cell lines. These comprised four which express cytoplasmic p chain (MIC-ALL, 697, NALM-6, SMS-SB) and six (RS4;11, TOM1, BV173, NALM-1, REH and 380) which are negative for this component. The Thiel myeloma cell line was obtained from Dr. W. P. Z. Zeijlemaker. 2.2 Antibodies The anti-B29 antibodies were produced by immunization of a rabbit with a synthetic peptide, C G E V K W S V G E H P G Q E, (representing the C terminus, from amino acid 215, of murine B29 polypeptide with a terminal cysteine added for protein coupling [7]), conjugated to thyroglobulin, and purified by elution from bovine serum albumin : peptide conjugate coupled to Sepharose. Anti-B29 was coupled to protein A-Sepharose CL-4B beads by incubating 10 pl of packed beads for 2 h with 30 1.11 of anti-B29 antiserum in 300 p1 of buffer. Production of the monoclonal anti-mb-1 antibody is described elsewhere [12]. 2.3 Immunoprecipitation

[I 103511 rhi\ work was supported by the Leukaemia Research Fund. Present address: Departmcnt ot Immunobiology, Dana-Farber Cancer Institute. 44 Binney Street, Boston, MA 02115, USA. Correspondence: David Y. Mason, Department of Haematology, John Radcliffe Hospital, Headington, Oxford, OX3 9DU, GB

0 VCH Verlagsgesellschaft mbH. D-6940 Weinheim, 1992

For cell surface radioiodination, 40 x lo6 viable Daudi or EB4B cells were suspended in PBS and labeled with 1mCi Na1251(Amersham Co., Amersham, GB) using lactoperoxidase as a catalyst.The cells were then washed three times in serum-free medium and lysed in immunoprecipitation buffer (IPB) consisting of 0.01 M triethanolamine-HCI, pH 0014-298019211010-2753$3.50+ ,2510

2154

D.Y. Mason, C. J. M. van Noesel, J. L. Cordell et al. A

Mr

200

97 68

43

B

C

D

E

F G H

-

Eur. J. Immunol. 1992. 22: 2753-2756

precipitation was performed with the mb-1 and B29 specific antibodies coupled to protein A-Sepharose CL-4B beads. All samples were analyzed by SDS-PAGE according to a modification of the Laemmli procedure on a 5%-15% gradient gel, and proteins were visualized by autoradiography at - 70 “C, using Kodak XAR-5 film in combination with intensifier screens (Cromex, Dupont Chemical Co., Newtown, CT).

2.4 Immunocytochemical staining 2918-

Figure I . The heterodimeric molecules that are associated with surface membrane IgM and surface membrane IgG are reactive with anti-B29 and anti-mb-1 anti-peptide antibodies. Surface membrane IgM (lane A) and surface membrane IgG (lane E) were isolated from digitonin lysates of surface radiolabeled Daudi or EB4B cells by use of anti-IgM and anti-IgG antibodies, respectively. The surface membrane IgM associated heterodimer was eluted using 10% NP40 whereas the membrane IgG-associated heterodimer was dissociated using 0.1% SDS. Lanes B and F show the proteins which remained on the beads after these elution procedures. The eluates were, after reduction and alkylation, reprecipitated using antibodies specific for B29 (lanes C and G) or for mb-1 (lanes D and H). The B29 precipitate in lane G is barely visible in this photograph but was clearly seen on the original autoradiograph.

Tissue sections and cytocentrifuge preparations were fixed in acetone for 10 min prior to incubation with monoclonal anti-mb-1 (antibody HM57) or polyclonal anti-B29 (diluted 1/20-1/130). Tissue sections and the myeloma cytocentrifuge preparations were then stained by the APAAP immuno-alkaline phosphatase technique while the acute leukemia samples were incubated with FITC-conjugated goat anti-mouse-Ig or FITC-conjugated goat anti-rabbit-Ig , respectively and then evaluated by conventional epiillumination fluorescence microscopy (Zeiss, Oberkochen, FRG). Samples from two cases of myeloma were analyzed by flow cytometry using anti-B29 and anti-mb-1 antibodies following permeabilization of cells, as described elsewhere

3 Results and discussion

B cell antigen receptor complexes were isolated from 7.8, 0.15 M NaC1, 5 mM EDTA, 1 mM PMSF, 0.02 mg/ml digitonin lysates of the IgM-positive Daudi cell line or the ovomucoid trypsin inhibitor, 1 mM TLCK, 0.02 mglml IgG-positive EB4B cell line using anti-IgM or anti-IgG leupeptin, supplemented with 1% digitonin as detergent. antibodies (Fig. 1, lanes A and E, respectively). The Nuclear debris was removed by centrifugation for 15 rnin at 47137-kDa and 49135-kDa heterodimers were then selec13 000 x g and for 30 min at 100 000 x g and precleared by tively eluted from the immunoprecipitates (lanes B and F, three incubations with 30 yl of a 10% vlv suspension of respectively), and irreversibly dissociated by reductive protein A-CL 4B Sepharose beads (Pharmacia, Uppsala, alkylation. In agreement with our previous findings [13], Sweden) coated with normal mouse Ig. Two consecutive antibodies raised against a peptide sequence from the immunoprecipitations were carried out for 2 h each, using a murine B29 gene product selectively precipitated the 37monoclonal antibody specific for human 1.1chain or human and 35-kDa chains from these eluates (lanes C and G), y chain (antibodies CLB-MH15 and CLB-MH16, respec- while the 47- and 49-kDa chains were recognized, as tively, both from the Central Laboratory of the Blood demonstrated previously, by antibodies to the mb-1 gene Transfusion Service, Amsterdam) coupled covalently to product (lanes D and H) [6]. This result was confirmed by CNBr-activated Sepharose 4B beads. Anti-IgM was used Western blotting experiments, which showed that the for Daudi cells, and anti-IgG for EB4B cells, which express anti-B29 antibodies reacted with the smaller of the two surface membrane IgG but not surface membrane IgM. The polypeptide chains in affinity-purified Ig-associated heteroimmunoprecipitates were washed five times in IPB contain- dimer isolated from human hairy cell leukemia cells [12] ing digitonin, and aliquots were resuspended in sample and also in membrane extracts of human B cells. buffer. For reprecipitation, the disulfide-linked dimers were dissociated from the surface membrane Ig chains by The cellular distribution of the human B29 gene product, incubation of the immunoprecipitates for 1 h on ice in evaluated in normal human lymphoid tissue by immuno200 yl IPB containing 1% NP40 (for IgM immunoprecipi- histological staining, was very similar to that of the mb-1 tates) or the same volume of IPB containing 0.1% SDS protein [12] in that not only were secondary B cell follicles followed by 2% NP40 (for IgG immunoprecipitates). selectively labeled (Fig. 2a), but within these structures Proteins in the supernatant were then reduced and alky- follicular mantle zones were more strongly labeled than lated by adding 0.2% SDS, incubating for 5 min at 68 “C, germinal centers. Marginal zone B cells in human spleen adding 2 mM dithiothreitol, incubating for 30 min at 45 “C, were positive (although more weakly than mantle zone and then finally adding 20 mM iodoacetamide and incubat- cells) as were “monocytoid B cells”, a cell type which is ing for 30 min at room temperature. The eluates were thought to be related to marginal zone B cells [14]. A diluted with 5 volumes of IPB containing 1.5% NP40. striking difference between the two antibodies, however, Myoglobin (100 yg) was added as carrier protein, eluates was that many plasma cells in lymphoid tissue sections were were precleared two times with protein A-Sepharose negative for B29 but strongly labeled with anti-mb-1 CL-4B beads coated with normal mouse Ig, and immuno- (Fig. 2a-c). This was particularly easily seen in human

Eur. .I. lmmunol. 1992. 22: 2753-2756

B29 and mb-1 expression during human B cell differentiation

tonsil, in which plasma cells are present in large, often confluent clusters, in close association with the squamous epithelium of the tonsillar crypts. This difference in the reactivity of plasma cells for the two polypeptides was a very consistent observation and was seen when the primary antibodies were used at a range of different dilutions, so that the results do not appear to reflect differences in antibody activity. Cell samples were analyzed from five cases of multiple myeloma, and four were found to be positive for mb-1. Two of these four samples were also clearly positive for B29, one showed occasional positive cells, and the fourth case was negative (Fig. 2d). The single mb-1-negative multiple myeloma was also negative for B29. A sample from the human myeloma cell line Thiel [l5] contained many mb-1-positive cells and smaller numbers of cells expressing B29. Northern blot analysis of this cell line confirmed that the mb-I gene is expressed (unpublished data). Staining of 25 human ALL samples of precursor B cell origin was performed to investigate expression of B29 and

2755

mb-1 genes during early B cell differentiation. Whereas mb-1 staining was found in virtually all cases of precursorB-ALL (24 out of 25), B29 expression was found in only 13 cases. Most of these B29-positive samples (10 out of 13) were from cases of pre-B-ALL, a subtype of ALL which is characterized by the cytoplasmic expression of 1-1 Ig heavy chain. Staining was also performed of precursor B cell lines, both to confirm the data on leukemic cells and to obtain material for biochemical studies. However, two were negative for both mb-1 and B29 (RS4; 11and TOMI) while the other eight (see Sect. 2.1) expressed both of these polypeptides.

Our previous studies revealed that antibodies to a peptide sequence from human mb-1 react with many mammalian species [12]. Tissue sections from a number of animal species were therefore tested with the anti-B29 reagent. Labeling of B cell areas was seen in all mammalian species tested [calf, cat, guinea pig, horse, marsupial (Monodelphus domesticus), monkey and pig; Fig. 2e]. In addition, the anti-B29 antibodies labeled B cell areas in avian bursa of Fabricius. In one species (the cat) the mb-I-positive, B29-negative pattern characteristic of human plasma cells was also seen (in scattered cells within the germinal centers; Fig. 2e). These results show that surface membrane Ig on human B cells is associated with a heterodimer of transmembrane glycoproteins, which are encoded, as in the mouse, by two distinct genes, B29 and mb-1. The fact that antibodies against synthetic peptides from the intracytoplasmic portion of both the mb-1 and B29 gene products react with B cells from all mammalian species tested [12] suggests a high degree of evolutionary conservation of amino acid sequences in the cytoplasmic domains. This conservation resembles that of the CD3 and i; components of the TcR complex and points to a pivotal role of these molecules in

Figure 2. (a) Adjacent cryostat sections of human tonsil stained for (left) B29, and (right) mb-1, showing very similar patterns,with strong staining of mantle zone cells and much weaker labeling of germinal centers. A cluster of epithelium-associated plasma cells (arrowed) expresses mb-1 antigen but not B29. (b) Higher power view of the same sections (B29 reactivity on the left and mb-1 on the right). The mb-1-positive, B29-negative plasma cell cluster is arrowed and scattered cells with the same phenotype are seen in the germinal center. These correspond in distribution and number to germinal center B cells (“centrocytes”) containing cytoplasmic Ig. (c) Adjacent sections of paraffin-embedded (Carnoy’s fixed) human tonsil stained for mb-1 (left) and B29 (right) showing the typical plasmacytoid morphology and cytoplasmic labeling of mb-1-positive cells more clearly than in cryostat sections. In the section stained for B29, unlabeled plasma cells (arrowed) can be recognized by their nuclear morphology. (d) Bone marrow cells from a case of human myeloma stained with anti-mb-1 (left) and anti-B29 (right), showing the same phenotype as reactive plasma cells. (e) Labeling of B29 in animal tissue. On the left a cat lymph node (cryostat section) stained for mb-1 (above) and B29 (below) is shown, Scattered mb-1-positive, B29-negative cells (arrowed) in the germinal center resemble those seen in human lymphoid tissue (Fig. 2b). Arrowheads indicate follicular mantle zones. O n the right (M) a marsupial (Monodelphus domesticus) spleen (cryostat section) stained for B29 is seen, showing strong labeling of B cell areas. A germinal center (arrowed) is more weakly stained than the surroundingmantle and marginal zones. RP = Red pulp;T = T cell area.

2756

D. Y. Mason, C. J. M. van Noesel, J. L. Cordell et al.

coupling the B cell antigen receptor to the mechanism for intracellular signal transduction. Our immunohistological findings clearly demonstrate that mb-1 protein is present in normal polyclonal human plasma cells in the absence of detectable amounts of B29 protein. It is, therefore, tempting to speculate that it is cessation of R29 synthesis which accounts for the disappearance of surface membrane Ig at this terminal stage of B cell differentiation. An analogy may be made with the TcR/CD3 complex, which is only found on the surface of T cells when all of its polypeptide components are available [16, 171. It should be added that these observations on human plasma cells are in disagreement with studies of mRNA production in murine plasmacytoma cells, which suggested that production of the mb-1 protein ceases at an earlier point in terminal B cell differentiation than that of B29 [2, 71. However, the presence in a cell of mRNA does not necessarily indicate that the encoded protein is synthesized, nor does the absence of transcription indicate that the cell does not contain previously synthesized protein. Furthermore, it is possible that the plasmacytoma cell lines studied previously were not representative of normal plasma cells. In this context, it is to be noted that whereas one of the fresh human myeloma samples that we studied showed the same discordance between mb-1 and B29 staining as seen in normal plasma cells, other cases expressed both of these chains, suggesting that they may not correspond to the terminal B cell differentiation stage seen in human tissues. However, it is of interest that none of the myeloma cases expressed B29 in the absence of mb-1. We carefully examined the red pulp in sections of mouse spleen (where scattered plasma cells should be present) but we could not identify plasma cells with certainty. However, cellular staining in this region was clearly more extensive for mb-1 than for B29 which suggests that further investigation may reveal the same mb-1-positive, B29-negative phenotype in murine plasma cells as we have found in human plasma cells. Of further interest is the suggestion from our results that the expression of mb-1 and B29 protein may possibly be differentially regulated in early B cell ontogeny. In the series of 25 precursor-B ALL cases studied, expression of I329 was especially found in pre-B-ALL cases, identified by their expreosion of cytoplasmic Ig p chains. Surface membrane Ig is first detectable at the pre-B cell stage in the form of Ig p heavy chain bound to theVpre.B/hSsurrogate Ig light chain complex [18, 191 and it is conceivable that B29 may also play a role in governing the appearance of membrane Ig at this stage in B cell differentiation. However, it should be emphasized that our data are based on a relatively small number of precursor B cell leukemias, rather than on normal human precursor B cells, and also that we could not find discrepant B29/mb-1 expression in precursor B cell lines. Consequently, further work is needed to investigate this hypothesis more fully. It should be noted that the level of surface membrane Ig on pre-B cells is much lower than on mature B cells, so that there are possibly additional, as yet unidentified, elements which regulate the assembly and membrane expression of the B cell receptor complex.

Eur. J. Immunol. 1992. 22: 27.53-2756

4 Concluding remarks These results indicate that the two chains which comprise the Ig-associated heterodimer in human B cells can now be identified with antibody probes, and that their expression appears to be differentially regulated. The cross-reactivity of these antibodies will make it possible to extend these studies to other mammalian species. We are grateful to Sheena Gowing for preparation of the B29 peptide, to Margaret Jones for skilful immunocytochemical assistance, to G. M. W van Schijndel for expert technical assistance, and to Jannie Borst and Alan Williamsfor helpful discussion. The EB4B cell line was kindly provided by Dr. R. Jefferis (Dept. of Immunology, University of Birmingham, Birmingham, G. B.). Received February 12, 1992; in final revised form July 8, 1992.

5 References 1 Hombach, J., Leclercq, L., Radbruch, A , , Rajewsky, K. and Reth, M., EMBO J. 1988. 7: 34.51. 2 Sakaguchi, N., Kashiwamura, S., Kimoto, M., Thalmann, F! and Melchers, F., EMBO J. 1988. 7: 3457. 3 Campbell, K. S. and Cambier, J. C., EMBO J. 1990. 9: 441. 4 Hombach, J.,Tsubata,T., Leclercq, L., Stappert, H. and Reth, M., Nature 1990. 343: 760. 5 Van Noesel, C. J. M., Borst, J., de Vries, E . F. R. and van Lier, R. A. W., Eur. J. Immunol. 1990. 20: 2789. 6 Van Noesel, C. J. M., van Lier, R. A.W., Cordell, J. L. ,Tse, A. G. D., van Schijndel, G. M.W., deVries, E. F. R., Mason, D.Y. and Borst, J., J. Imrnunol. 1991. 146: 3881. 7 Hermanson, G. G., Eisenberg, D., Kincade, F!W. and Wall, R., Proc. Natl. Acad. Sci. USA 1988. 85: 6890. 8 Ishihara, K., Wood, W. J., Damore, M., Hermanson, G. G., Wall, R. and Kincade, PW., Proc. Natl. Acad. Sci. USA 1992. 89: 633. 9 Campbell, K. S., Hager, E. J., Freidrich, R. J. and Cambier, J. C., Proc. Natl. Acad. Sci. USA 1991. 88: 3982. 10 Reth, M., Hombach, J., Wienands, J. K., Campbell, K. S., Chien, N., Justement, L. B. and Cambier, J. C., Immunol. Today 1991. 12: 196. 11 Gatter, K. C., Falini, B. and Mason, D.Y., in Anthony, P. and MacSween, R. (Eds), Recent Advances in Histopathology, Vol. 12., Churchill Livingstone, Edinburgh 1984, p. 35. 12 Mason, D.Y., Cordell, J. L.,Tse, A. G. D., van Dongen, J. M., van Noesel, C. J. M., Micklem, K., Pulford, K. A. F. ,Valemi, F., Comans-Bitter, W. M., Borst, J. and Gatter, K. C., J. Immunol. 1991. 147: 2474. 13 Van Noesel, C. J. M., Brouns, G. S., van Schijndel, G. M. W., Bende, R. J., Mason, D.Y., Borst, J. and van Lier, R. A. W., J. Exp. Med. 1992. 175: 1511. 14 Sheibani, K., Kritz, R. M.,Winberg, C. D., Burke, J. S. and Rappaport, H., Amer. J. Clin. Pathol. 1984. 81: 453. 15 Weinreich, S. S. ,von dem Borne, A. E . G. Kr., van Lier, R. A. W., Feltkamp, C. A., Slater, R. M., Webster, R. and Zeijlemaker,W. l? Z., Br. J. Haematol. 1991. 79: 226. 16 Clevers, H., Alarcon, B. ,Wileman,T. and Terhorst, C., Annu. Rev. Immunol. 1988. 6: 629. 17 Van Dongen, J. J. M., Krissansen, G. W. ,Wolvers-Tettero, I. L. M., Comans-Bitter, W. M., Adriaansen, H. J., Hooijkaas, H., van Wering, E. R. and Terhorst, C., Blood 1988. 71: 603. 18 Nishimoto, N., Kubagawa, H., Ohno, T., Gartland, G. L., Stankovic, A. K. and Cooper, M. D., Proc. Natl. Acad. Sci. U S A 1991. 88: 6284. 19 Kerr, W. G., Cooper, M. D., Feng, L., Burrows, l? D. and Hendershot, L. M., Int. Immunol. 1989. 1: 355.