Idiotype vaccination leads to the emergence of a stable surface Ig-negative variant of the mouse lymphoma BCL1, with different growth characteristics

Vaccination of mice with tumor-derived idiotypic IgM from the B cell lymphoma, BCL1, induces an anti-idiotypic immune response which suppresses tumor development. One of the mechanisms by which tumor cells can escape attack is by failing to express significant levels of idiotypic immunoglobulin at the cell surface, and a stable variant of this phenotype has been isolated. The variant, termed SNAG 1, continues to synthesize idiotypic IgM, which can be detected in the cytoplasm, but it neither secretes nor expresses IgM on the cell surface (less than 10% of the levels of the original BCL tumor), even though the H and L chains show no gross structural changes. The SNAG 1 cells resemble the parent BCL cells in morphology, in expression of MHC class I and II Ag and in bearing FcR. A significant difference between the BCL lymphoma cells and the variant cells is that the latter fail to respond to LPS by either DNA synthesis or secretion of IgM, suggesting that surface Ig might be required for such a response. The variant has a slower rate of division than the parent tumor both in vitro and in vivo, and a rather different organ distribution. Study of such variants might allow analysis of the mechanisms involved in surface Ig expression and its possible role in tumor cell growth and migration.     

Differential expression of HLA class II antigens in fetal human spleen: relationship of HLA-DP, DQ, and DR to immunoglobulin expression

Frozen sections of human fetal spleen from 12 to 20 wk gestation were examined by using polyclonal antibodies to Ig isotypes, monoclonal antibodies to HLA class II subregion locus products, B and T cells, and follicular dendritic cells. Scattered lymphoid cells in spleen sections from fetuses of 12 to 13 wk gestational age expressed IgM but not IgD. The appearance of lymphoid cells expressing IgD occurred at 14 to 15 wk before the formation of loose clusters of B cells at 16 wk. IgD expression was associated mainly with cells in these clusters, which by 17 wk had become definite follicles. Follicular dendritic cells were not detectable until 20 wk. OKT3-positive T cells were not detected until 17 wk, and at 20 wk constituted 5% of the nucleated cell population. HLA-DR- and DP-positive lymphocytes and macrophages were detectable in fetal spleen from 12 wk onward; DR was expressed on more cells than DP, and the numbers of cells stained by HLA-DR-specific monoclonal antibodies exceeded the number of Ig-positive cells in all spleens examined. HLA-DQ was expressed by consistently fewer cells than HLA-DR and -DP in all spleens tested. The small number of DQ-positive cells in spleens from 12- to 13-wk fetuses had the morphology of macrophages; HLA-DQ expression by lymphoid cells followed a similar pattern to IgD expression and was associated mainly with follicular lymphocytes. It could be demonstrated by double-labeling experiments that all follicular IgM-positive cells in 17- to 20-wk spleens expressed HLA-DP, DQ, and DR antigens: IgM-positive cells in 12- to 16-wk spleens and interfollicular IgM-positive cells in 17- to 20-wk spleens all expressed HLA-DR, but only 59% and 43% expressed DP and DQ, respectively. Ninety-one to 100% of IgD-positive cells in all spleens examined expressed HLA-DQ in addition to DR and DP. In these experiments IgD-negative, DQ-positive cells had the morphologic appearance characteristic of macrophages. These data suggest that class II antigens are differentially expressed on developing lymphoid cells; DR and DP expression occurring in the earliest spleens examined, with expression of DP on a subpopulation of DR-positive cells; IgD and DQ expression appears to be coincident on maturing B cells as they begin to form follicles. An immunoregulatory role for HLA-DQ in B cell development is implicated and remains to be fully investigated.     

The role of T cells in immunoglobulin class switching of specific antibody production system in vitro in humans

Only antibodies of the IgM class were produced in vitro by peripheral blood mononuclear cells stimulated with streptococcal carbohydrate. B cells of the peripheral blood mononuclear cells, however, synthesized both IgM and IgG class antibodies when combined with tonsillar T cells, suggesting that T cells inducing immunoglobulin class switching are present in the tonsils. Peripheral blood T cells also became capable of inducing B cells to produce IgG class antibodies when the T cells were incubated with antigen-pulsed macrophages. Surface IgM-positive, IgG-negative high-density B cells produced IgG antibodies for streptococcal carbohydrate in the presence of these T cells or tonsillar T cells. The culture supernatant solutions from these T cells or tonsillar T cells, however, failed to cause the B cells to produce IgG, indicating that class switching is not mediated by factors released from T cells. Lymphokines such as interleukin-2, human B cell growth factor, helper T cell factor, or interferon-gamma were also incapable of inducing IgG production. These results suggest that the cognate interaction between T cells and B cells is necessary for the immunoglobulin class switching.     

Idiotype variants emerging after anti-idiotype monoclonal antibody therapy of a murine B cell lymphoma

One of the difficulties encountered with the treatment of human B cell malignancies with anti-Id antibodies is the emergence of Id variants. The current study was designed to investigate this phenomenon further by using the murine B cell lymphoma model 38C13. Tumors were harvested that developed despite treatment with the anti-Id antibody S1C5 in mice inoculated with 38C13 cells and evaluated by immunofluorescence. Various phenotypes were found among escaping tumor cells. Some cells continued to react with S1C5 whereas others lost S1C5 reactivity. Among these latter cells, some continued to express surface IgM kappa, whereas others no longer expressed surface mu or kappa. After Id variant cell lines were established, immunofluorescence and ELISA of cell lysates from the surface IgM kappa- lines revealed persistent intracellular mu H chain but no detectable kappa. Surface IgM kappa+ lines were fused with myeloma cells and the Ig proteins secreted by the resultant hybridomas analyzed. The apparent m.w. of the mu-chains of these rescued Ig was the same as wild-type 38C13, whereas the kappa-chains were either the same or different in m.w. from the wild type. The IgM kappa of the variant line, T3C, weakly reacted with S1C5 and did not react with other anti-Id antibodies. The IgM kappa of the other variants were nonreactive with all the antibodies. Immunofluorescence of these surface Ig+ variants confirmed this finding. Some of the surface Ig+ and Ig- variant lines grew identically to wild-type tumor in vivo, but only the weakly S1C5-reactive variant T3C was inhibited in its growth by S1C5. Moreover, T3C was the only one of these lines capable of being lysed in vitro with S1C5 by antibody-dependent cellular cytotoxicity. Further studies revealed that surface Ig+ and Ig- variants emerge in escaping tumors with similar frequency and that these variants represent a major mode of tumor escape from anti-Id treatment in this model.     

VH replacement rescues progenitor B cells with two nonproductive VDJ alleles

Inaccurate VDJ rearrangements generate a large number of progenitor (pro)-B cells with two nonproductive IgH alleles. Such cells lack essential survival signals mediated by surface IgM heavy chain (muH chain) expression and are normally eliminated. However, secondary rearrangements of upstream VH gene segments into assembled VDJ exons have been described in mice transgenic for productive muH chains, a process known as VH replacement. If VH replacement was independent of muH chain signals, it could also modify nonproductive VDJ exons and thus rescue pro-B cells with unsuccessful rearrangements on both alleles. To test this hypothesis, we homologously replaced the JH cluster of a mouse with a nonproductive VDJ exon. Surprisingly, B cell development in IgHVDJ-/VDJ- mice was only slightly impaired and significant numbers of IgM-positive B cells were produced. DNA sequencing confirmed that all VDJ sequences from muH chain-positive B lymphoid cells were generated by VH replacement in a RAG-dependent manner. Another unique feature of our transgenic mice was the presence of IgH chains with unusually long CDR3-H regions. Such IgH chains were functional and only modestly counter-selected, arguing against a strict length constraint for CDR3-H regions. In conclusion, VH replacement can occur in the absence of a muH chain signal and provides a potential rescue mechanism for pro-B cells with two nonproductive IgH alleles.     

B cells in patients with X-linked agammaglobulinemia

X-linked agammaglobulinemia (XLA) has been described as a disorder in which pre-B cells fail to differentiate into B cells. However, a small number of B cells have been seen occasionally in patients with this disorder. Because the phenotype of these cells might be helpful in defining the site of the defect in XLA, immunofluorescent staining techniques were used to characterize the B cells that can be found in patients with XLA. Surface IgM-positive B cells could be detected in the peripheral circulation of all seven patients studied. These B cells constituted a very small percentage of the total lymphocytes (0.01 to 0.3% compared with 3.2 to 13.7% in controls) and differed in phenotype from control B cells. They were much more brightly stained for surface IgM (p less than 0.001) and less brightly stained for Ia (p less than 0.01). This phenotype is similar to that described for immature B cells in the mouse. Over 80% of the patients' B cells expressed surface IgD, and all expressed the B cell marker B1, but only 35% expressed the B cell marker B2. This B cell marker, which is the C3d receptor and the Epstein-Barr virus receptor, is expressed later in ontogeny than B1 and can be detected on over 80% of control B cells. All B cells expressed either kappa or lambda light chain. These findings indicate that the defect in differentiation of pre-B cells into B cells is not absolute in patients with XLA. The immature phenotype of the B cells additionally suggests that there may be a block in the maturation of B cells at more than one stage of differentiation in this disorder.     

Functional characterization of two stromal cell lines that support B lymphopoiesis.

Bone marrow stromal cells have well documented effects on the production of B lymphocytes, but whether or not stromal cell signals are involved in the pre-B to B cell transition is unclear. The potential of two stromal cell lines, S10 and S17, in this process was examined. Initial experiments, using a short term liquid culture, indicated that S10 and S17 stroma efficiently supported the generation of clonable B cells (B lymphocyte CFU) from their immediate precursors in fresh bone marrow. The contribution of macrophages and other accessory cells in those experiments was minimized through use of a colony assay system that permits the direct effects of stromal cell signals on single B cell progenitors to be evaluated. The results indicated that soluble mediators from the S10 and S17 lines could support colony formation from fresh or cultured surface Ig- bone marrow cells. Colonies supported by S17 stroma appeared on day 15 and contained cells that expressed the B220 Ag; surface IgM expression was never observed. S10 supported colonies appeared on day 7 and routinely included surface IgM+ cells. Individual colonies were capable of undergoing additional growth when picked and replated directly onto the different stroma. Those colonies replated onto S10 stroma generated surface IgM expressing cells in up to 60% of experiments, but colonies transferred onto the S17 cell line included B cells only 10% of the time. These data demonstrate that stromal cells alone can provide the signals necessary for generating a surface IgM+ B cell from precursors but that not all stromal cell lines are equally efficient at doing so.     

Expression, distribution and specificity of Fc receptors for IgM on murine B cells

Subpopulations of normal adult murine splenic B cells and a panel of murine B cell tumors were examined for their ability to bind murine IgM specifically. By using two-color flow cytometric analyses, we have demonstrated that 90 to 95% of surface (s)IgD+ B cells express surface membrane receptors for IgM (Fc mu R). The binding of pentameric murine IgM to splenocyte Fc mu R was IgM-specific since it was totally inhibited by other polymeric IgM proteins, but not by Ig of other H chain classes or by mAb specific for the murine IgG or IgE FcR. Binding of IgM to splenic cells was saturable. Fc mu R were co-expressed with the Fc gamma R as well as the Fc epsilon R on the majority of splenic B cells. Minor populations of splenic mononuclear cells expressed only an Fc mu R, Fc gamma R or Fc epsilon R. In a survey of B tumor cell lines representing different stages of B cell development, we observed that the Fc mu R was expressed on pre-B cell lines and that Fc mu R detection was maximal on immature B cell lines that expressed sIgM and low amounts of sIgD and Ia. Fc mu R were not detected on cell lines that had switched from sIgM to the expression of another sIg, or on plasmacytomas and hybridomas. The studies with normal splenocytes establish that the majority of sIgD+ B lymphocytes in adult BALB/c mice express surface membrane receptors that specifically bind IgM. The studies with B lineage tumor cells suggest that the expression of Fc mu R on B cells is developmentally regulated and that the pattern of expression exhibited by Fc mu R during B cell ontogeny differs from the patterns that have been previously found for IgG and IgE FcR. These observations raise the possibility that Fc mu R might have a functional significance in some aspect of B cell maturation and activation. By using a family of IgM H chain constant region domain deletional mutants, we have further demonstrated that, like the T cell Fc mu R, the B cell Fc mu R also requires a C mu 3 domain for binding to occur, raising the possibility that the T and B cell Fc mu R in mice may be structurally related molecules.     

Molecular components of the B cell antigen receptor complex of class IgD differ partly from those of IgM

Two classes of immunoglobulin, IgM and IgD, are present as antigen receptors on the surface of mature B lymphocytes. We show here that IgD molecules are noncovalently associated in the B cell membrane with a heterodimer consisting of two proteins of 35 kd (IgD-alpha) and 39 kd (Ig-beta), respectively. The two novel proteins are not found in the IgD-expressing myeloma J558L delta m, which fails to bring IgD antigen receptor onto the cell surface. In a surface IgD positive variant line of this myeloma, however, membrane-bound IgD molecules are associated with the heterodimer, suggesting that the formation of an antigen receptor complex is required for surface IgD expression. We further demonstrate that the IgD-associated heterodimer differs partly from that of the IgM antigen receptor and that its binding to the heavy chain only requires the presence of the last constant domain and the transmembrane part of the delta m chain.     

An in vitro line of the B cell tumor BCL1 can be activated by LPS to secrete IgM1

An in vitro line of the B cell tumor BCL1 was developed. The cell line carried u-, S-, and A-chains on the cell surface as judged by analysis of surface iodinated proteins but did not secret Ig. ASfter stimulation with LPS, limpid A, or bacterial lipoprotein, 20 to 40% of the tumor cells matured to IgM secretors when detected in a plaque assay. Two other polyclonal B cell activators, namely dextransulphate and PPD, had at most a marginal stimulatory effect. The ability of the cells to become activated to IgM secretion as well as the expression of cell surface IgM and IgD makes the BCL1 unique among murine B cell tumors.     

Anti-idiotypic antibodies recognizing stable epitopes limit the emergence of idiotype variants in a murine B cell lymphoma.

The emergence of Id variants is a major escape mechanism from anti-Id therapy of human B cell malignancies and of the murine B cell lymphoma 38C13. To determine what impact the epitope specificity of anti-Id antibodies has on the prevention of emergence of such Id variants in the 38C13 lymphoma, anti-Id mAb of varying epitope specificity for the Id of 38C13 tumor cells were produced and studied. Some antibodies, produced by immunizing mice with both the wild-type 38C13 IgM and variant IgM, cross-reacted with wild-type 38C13 IgM and with all four members of a panel of variant IgM. These anti-Id did not react with separated 38C13 IgM H or L chains by Western blot, but did react with the cytoplasmic H chain of the surface Ig- variant cell line T2D that expresses the same H chain as wild-type 38C13 in its cytoplasm but does not express any associated L chain. In contrast, anti-Id of narrower specificity did not react with this H chain. This indicated that the broadly cross-reactive antibodies recognized a stable epitope on 38C13 H chain. When a broadly cross-reactive antibody MS11G6 was compared to S1C5, an antibody of narrower specificity, MS11G6, was superior at preventing tumor growth in mice inoculated with 38C13 cells. Moreover, no surface Ig+ variants emerged in escaping tumors in the MS11G6-treated group, whereas such variants were common in the S1C5 treated group. Both anti-Id were of equal efficacy in eliminating wild-type 38C13 cells by using 38C13 cells in tumor inoculums that had just been cloned in vitro, but MS11G6 was also capable of preventing the growth of several surface Ig+ variant cell lines in vivo. We conclude that anti-Id recognizing more stable Id determinants can limit the emergence of Id variants and therefore be more effective therapeutic agents. This finding is of additional importance as additional in vivo and immunophenotypic studies demonstrated that the generation of Id variants was an ongoing process both in cloned parental 38C13 cells and its variants.     

The class of surface immunoglobulin on virgin and memory B lymphocytes.

The class of surface immunoglobulin receptors for antigen on B cell precursors of different classes of antibody-forming cells was determined by utilizing a technique of class-specific antigen suicide. Spleen cells are first treated with a class-specific antiserum under conditions that result in the stripping of that class from the cell surface. The cells are then permitted to bind a highly radioactive trinitrophenyl (TNP)-conjugated protein, which leads to lethal irradiation of all TNP-specific B cells except those whose TNP receptors had been removed by the class-specific stripping of surface immunoglobulin. In this way, the class of antibody-forming cells resulting from TNP stimulation of B cells with different classes of surface immunoglobulin can be examined. It was found that the virgin B cell precursors of IgM-producing cells are two types: cells bearing IgM receptors only and those bearing both IgM and IgD receptors. All virgin B cells that gave rise to IgG1 antibody-forming cells had both IgM and IgD on their surfaces, demonstrating that an antigen-dependent switch from IgM and IgD to IgG1 production is a common feature of B cell maturation. In contrast, memory B cell precursors of IgG1 antibody-forming cells had predominantly IgG1 as their surface antigen receptor. The implications of these findings on current models of B cell maturation are analyzed.     

Differential expression of the c-myb proto-oncogene marks the pre-B cell/B cell junction in murine B lymphoid tumors

A series of murine B lymphoid tumor cell lines which are representative of the pre-B cell, immature and mature B cell, and plasma cell stages of B cell development have been examined for expression of c-myb proto-oncogene mRNA. The pre-B cell lymphoma cell lines express equivalent high steady state levels of c-myb mRNA. In contrast, the B cell lymphoma and plasmacytoma cell lines express steady state c-myb mRNA at levels which are 0.005 to 0.1 times that of the pre-B cell lymphoma lines. These results correlate high levels of c-myb mRNA expression with the pre-B cell stage of development. Subclones of the 1881 pre-B cell lymphoma which express K light chain and are surface IgM-positive as well as two types of hybrid B lymphoid cell lines have been used to demonstrate that surface immunoglobulin expression is not sufficient to result in the down-regulation of c-myb mRNA levels or changes in the expression N-myc mRNA, lambda 5 mRNA, or the BP-1 surface antigen which are markers of the pre-B cell stage of development. Thus, changes in the expression of genes which are independent of immunoglobulin expression are associated with transition from the pre-B cell to the immature B cell stage of development.     

Functional properties of human B cell subpopulations defined by monoclonal antibodies HB4 and FMC7

Human B cell subpopulations can be distinguished by the expression of B cell-associated antigens. mAb directed against these structures allow the isolation and subsequent functional analysis of such subsets. Blood B cells from healthy individuals were separated on basis of the expression of the antigens recognized by the mAb HB4 and FMC7. The B cells present in the thus obtained populations were analyzed for their ability to secrete IgM and IgG after stimulation in vitro with polyclonal B cell activators (PWM, Staphylococcus aureus Cowan I), antigens (tetanus toxoid, type 4 pneumococcal polysaccharides), and soluble B cell differentiation factors. The results suggest that B cells capable of Ig and anti-tetanus toxoid or anti-type 4 pneumococcal polysaccharide antibody production after in vitro culture are localized in a relative small B cell subpopulation carrying the FMC7 determinant but lacking HB4. This holds true for both the IgM- and IgG-secreting B cells. These data further suggest that the majority of B cells found in the blood and which can be characterized as being surface IgM+/IgD+ HB4+ are immature cells unable to respond to differentiation-inducing signals.     

Human B cell responsiveness to B cell growth factor after activation by phorbol ester and monoclonal anti-mu antibody

The effect of phorbol ester on human B cell activation was examined. Picomolar to nanomolar concentrations of phorbol ester induced a high level of proliferation in small IgM-positive B cells isolated from peripheral blood by fluorescence-activated cell sorting. The addition of optimal doses of anti-mu antibody resulted in enhanced proliferation of phorbol ester-activated B cells. The addition of B cell growth factor (BCGF) to phorbol ester-activated B cells also resulted in a dose-dependent synergistic effect and maximal enhancement on day 3. BCGF activity could be absorbed with either phorbol ester- or anti-mu-activated B cells, but not with resting B cells, thus confirming the induction of functional BCGF receptor expression. Cell proliferation was not necessary for the induction of functional BCGF receptors. Phorbol ester was a more efficient inducer of BCGF receptor expression than was anti-mu antibody; gamma-interferon treatment had no effect. BCGF enhanced transferrin receptor expression by phorbol ester-activated B cells. The results suggest that phorbol ester-activated small B cells can be used to monitor BCGF activity, and this synergistic combination may be useful in establishing BCGF-dependent B cell clones in culture.     

The CR2 receptor (CD21) shows increased expression in the more differentiated cells of an antigen-specific B cell line.

The complement receptor 2 (CR2; CD21), a 145,000 MW glycoprotein, has been useful as a marker of B lymphocyte maturation. It is expressed on the 1:13 monoclonal, EBV-transformed, B cell line which produces TNP-specific IgM-kappa and displays an in vitro capacity for differentiation. The line expresses the CD20+CD21+ phenotype. We studied whether CR2 receptor surface expression varied in relation to the cell cycle or state of differentiation in the 1:13 line. High CD21 and IgM expression occurred in the G1 phase of the cell cycle. In contrast to CD21, there were no distinctly brighter subpopulations of CD20 positive cells in the G1, S, or G2M compartments of the cell cycle. When sorted according to size, smaller cells were predominantly in G1, whereas a greater proportion of the larger cells were in the G2M phase of the cell cycle. The smaller 1:13 cells expressed more CD21 surface antigen and IgM than the larger cells. Cells which formed stable rosettes with TNP-SRBC expressed more surface IgM and CD21 antigen than nonrosetting cells. We have previously shown that the TNP-SRBC rosetting cells were more differentiated, resided in G1, and secreted more immunoglobulin than the nonrosetting cells. Thus increased CR2 expression occurred in the more differentiated cells of this human monoclonal B cell line.     

Comparison of membrane IgM expression in the murine B cell lymphoma 70Z/3 treated with LPS or supernatant containing T cell factors

The cultured murine B cell lymphoma, 70Z /3, can be induced to express membrane IgM ( mIgM ) after exposure to lipopolysaccharide (LPS) or T cell-derived factors. The kinetics and magnitude of the responses have been compared in wild type 70Z /3 cells and three variants by using flow cytometric analysis and immunoprecipitation. Wild type 70Z /3 cells respond to LPS more quickly and with twofold greater mIgM than to concanavalin A-induced spleen cell supernatant (CAS). Variants were selected for their abnormal mIgM expression in response to LPS, but individual variants also showed normal, aberrant, or no response to CAS. When cells were induced with suboptimal amounts of LPS and CAS, a synergistic effect on the magnitude of mIgM expression was seen in wild type and variant cells. This suggests that both inducing agents are utilizing some part of a common inductive mechanism. The different responses of the variant cell lines will allow further genetic dissection and comparison of the mIgM expression pathways used in response to LPS and CAS.