Subpopulations of B cells in germinal centers. III. HJ6, a monoclonal antibody, binds globoside and a subpopulation of germinal center B cells.

To identify surface Ag uniquely expressed on human germinal center B cells, we produced a mouse mAb, HJ6. When tonsillar lymphocytes were examined, HJ6 did not label T cells and labeled only about half of PNA+ B cells that were HK23-. HJ6 did not label mononuclear cells from peripheral blood, splenocytes, and any of 29 cell lines including 23 B cell lines. This binding pattern of HJ6 was very similar to that of a mAb named 5B5. It was shown previously that 5B5 bound a glycolipid named CTH (CD77) and its Ag was expressed on HK23- PNA+ tonsillar lymphocytes and Burkitt's lymphoma cell lines. Despite the similarity, HJ6 differed from 5B5: HJ6 did not stain Burkitt's lymphoma cell lines and stained PNA+ tonsillar lymphocytes in the presence of a large concentration of galactose. When its binding to isolated glycolipids was studied, HJ6 was found to bind globoside and Forssman Ag and not to other glycolipids including CTH. When its binding to neutral glycolipids extracted from tonsillar lymphocytes was studied, HJ6 bound only globoside; Forssman Ag was not detected in tonsillar lymphocytes. Taken together, we conclude that globoside is a B cell Ag expressed on a subpopulation of germinal center B cells.     

The effects of IL-4 and IL-5 on the IgA response by murine Peyer's patch B cell subpopulations

IL-5 has been shown to specifically enhance IgA secretion in LPS-stimulated splenic B cell cultures. Maximum enhancement of IgA in such cultures, however, requires IL-4 in addition to IL-5. Because the Peyer's patches (PP), compared with spleen and lymph nodes, are enriched for precursors of IgA-secreting cells, we tested whether IL-4 and IL-5 would have a more profound effect on IgA secretion by polyclonally stimulated PP cells than spleen cells. The combination of IL-4 and IL-5 causes a comparable enhancement of IgA secretion in both LPS-stimulated PP and splenic B cell cultures. The majority of IgA secreted in LPS-stimulated PP cell cultures is derived from the sIgA- population. Furthermore, the binding high level of peanut agglutinin, germinal center subpopulation of PP cells is essentially nonresponsive to LPS, even in the presence of lymphokines; the majority of secreted IgA in these cultures is derived from the binding low level of peanut agglutinin population. In contrast to LPS-stimulated cultures, PP B cells secrete considerably more IgA than splenic B cells when polyclonally stimulated by a clone of autoreactive T cells in the presence of IL-4 and IL-5. The majority of IgA made by T cell-stimulated PP cell cultures is derived from the sIgA+ population. In these cultures, sIgA- PP cells and spleen cells secrete comparable levels of IgA and other non-IgM isotypes suggesting that sIgA- PP B cells are similar to splenic B cells in their potential to switch to IgA. In T cell-stimulated cultures the majority of IgA as well as of all other isotypes is also derived from the nongerminal center, binding low level of peanut agglutinin population.     

Bm1-Bm5 classification of peripheral blood B cells reveals circulating germinal center founder cells in healthy individuals and disturbance in the B cell subpopulations in patients with primary Sjögren's syndrome

Analyses of B cells in the bone marrow and secondary lymphoid tissues have revealed a broad range of cell surface markers defining B cell subpopulations, but only a few of these have been used to analyze B cell subpopulations in peripheral blood (PB). We report here the delineation of circulating PB B cell subpopulations by staining for CD19, CD38, and IgD in combination with CD10, CD44, CD77, CD95, CD23, IgM, and the B cell memory marker CD27. The utility of this approach is shown by the demonstration of disturbances of circulating B cell subpopulations in patients with autoimmune disease. Five mature B cell (Bm) subpopulations were identified in normal PB that were comparable with the tonsillar Bm1, Bm2, early Bm5, Bm5 subpopulations and, surprisingly, to the germinal center (GC) founder cell subpopulation (Bm2' and Bm3delta-4delta), suggesting that some GC founder cells are circulating. No PB B cells resembled the Bm3 and Bm4 GC cells. Remarkably, some cells with the CD38-IgD+ phenotype, previously known as naive Bm1 cells, expressed CD27. The CD38-IgD+ subpopulation therefore includes both naive Bm1 cells and IgD+ memory B cells. This new classification of B cell developmental stages reveals disturbances in the proportions of B cell subpopulations in primary Sj?gren's syndrome (pSS) patients compared with healthy donors and rheumatoid arthritis patients. Patients with pSS contained a significantly higher percentage of B cells in two activated stages, which might reflect a disturbance in B cell trafficking and/or alteration in B cell differentiation. These findings could be of diagnostic significance for pSS.     

Subpopulations of mature murine thymocytes: properties of CD4-CD8+ and CD4+CD8- thymocytes lacking the heat-stable antigen

The heat-stable antigen (HSA), recognized by the monoclonal antibodies M1/69, B2A2, and J11d, is low or absent on the surface of most murine peripheral T cells but present on all but 3% of thymocytes. The CD4-CD8+ and CD4+CD8- or "single positive" thymic populations may be divided into further subgroups based on surface HSA expression. One group, CD4-CD8+ and expressing very high levels of HSA (HSA++), is an immature, T cell antigen receptor (TcR) negative, outer cortical blast cell. However, a further subdivision of CD4-CD8+ and CD4+CD8- single positives may be made, into those negative to low for HSA (HSA-) and those expressing moderate amounts of HSA (HSA+). The proportion of HSA- single positives is low in the thymus of young mice, whereas the proportion of HSA+ single positives is similar to that of the adult. Both the HSA- and the HSA+ subsets of single positive thymocytes from adult mice are CD3+ and express the normal peripheral T cell incidence of V beta 8 determinants on the TcR. On stimulation with concanavalin A in limit-dilution culture both HSA- and HSA+ subsets of single positive thymocytes give a high frequency of proliferating clones, and the clones from both HSA- and HSA+ subsets of CD4-CD8+ thymocytes are cytotoxic. Thus both HSA- and HSA+ single positive thymocytes are functionally mature. The HSA- subsets of single positive thymocytes differ from the HSA+ subsets in being slightly larger in size, in expressing higher levels of MEL-14, in binding more peanut agglutinin, and in including a proportion of cells expressing high levels of the Pgp-1 glycoprotein. It is suggested that HSA- CD4-CD8+ and HSA- CD4+CD8- thymocytes are more mature than their HSA+ counterparts, and might represent a previously activated or "memory" thymic subpopulation.     

Induction of a germinal center phenotype in B cells in vitro by a Th2 cell line

We have investigated the contribution of various stimuli for generating in vitro the changes in surface phenotype characteristic of B cells responding to a T-dependent antigen in a germinal center (GC). We show that, unlike many other stimuli such as B cell mitogens, cytokines, and surrogate antigen, alone or in combination, an alloreactive Th2 clonal line induces splenic B cells to become cell surface peanut agglutinin (PNA)(hi), Ig(lo), CD62L(lo), and CD44(hi) to produce mRNA for M17 and to express a GC-specific transgene even without B cell receptor ligation. Neither proliferation nor prior activation of responding B cells is needed, but B cells from CD45-null mice show reduced efficiency of this induction. These findings open up possibilities for separation and dissection of the various components of the GC response.     

Memory B cells express a phenotype consistent with migratory competence after secondary but not short-term primary immunization

The cell surface phenotype of dinitrophenol (DNP)-specific memory B cells, defined by their capacity to transfer IgG responses into syngeneic irradiated recipients, was assessed using two markers of relevance to lymphocyte migratory properties: (i) peanut agglutinin, which binds to terminal galactosyl residues expressed at high levels by several nonmigrating lymphocyte subsets and, among lymph node B cells, is highly specific for germinal center cells; and (ii) MEL-14, a monoclonal antibody specific for lymphocyte surface receptors required for migration from the blood into peripheral lymph nodes. At various times after primary or secondary immunization with DNP-keyhole limpet hemocyananin (KLH), lymph node B cells were separated by fluorescence-activated cell sorting on the basis of staining with PNA and/or MEL-14, and the presence of B-memory cells in each fraction was assessed by adoptive transfer with antigen (DNP-KLH) and helper T cells. One week after immunization, most of the memory sorted in the PNAhi population, confirming a previous report by R. F. Coico, B. S. Bhogal, and G. J. Thorbecke (J. Immunol. 131, 2254, 1983) that early memory B cells or their precursors are contained within the germinal center cell population, a population which is known to be MEL-14- and migratory-incompetent. Six weeks after primary stimulation, however, the bulk of memory cells, unlike germinal center cells, were MEL-14hi. After secondary immunization, memory was still predominantly MEL-14+ and PNAlo, although in some experiments adoptive responses were transferred by all sorted fractions. The results are consistent with the hypothesis that antigen-specific B cells initially undergo local (sessile) differentiation and proliferation in germinal centers, where they develop the capacity for adoptive transfer of antigen-specific secondary responses, but that with continued development their long-lived memory-containing progeny express a phenotype permitting their reentry into the recirculating lymphocyte pool.     

The role of ICOS in the CXCR5+ follicular B helper T cell maintenance in vivo

ICOS is a new member of the CD28 family of costimulatory molecules that is expressed on activated T cells. Its ligand B7RP-1 is constitutively expressed on B cells. Although the blockade of ICOS/B7RP-1 interaction inhibits T cell-dependent Ab production and germinal center formation, the mechanism remains unclear. We examined the contribution of ICOS/B7RP-1 to the generation of CXCR5+ follicular B helper T (T(FH)) cells in vivo, which preferentially migrate to the B cell zone where they provide cognate help to B cells. In the spleen, anti-B7RP-1 mAb-treated or ICOS-deficient mice showed substantially impaired development of CXCR5+ T(FH) cells and peanut agglutinin+ germinal center B cells in response to primary or secondary immunization with SRBC. Expression of CXCR5 on CD4+ T cells was associated with ICOS expression. Adoptive transfer experiments showed that the development of CXCR5+ T(FH) cells was enhanced by interaction with B cells, which was abrogated by anti-B7RP-1 mAb treatment. The development of CXCR5+ T(FH) cells in the lymph nodes was also inhibited by the anti-B7RP-1 mAb treatment. These results indicated that the ICOS/B7RP-1 interaction plays an essential role in the development of CXCR5+ T(FH) cells in vivo.     

Germinal center cells are a major IL-5-responsive B cell population in peripheral lymph nodes engaged in the immune response

Germinal center formation and the development of B cell memory in lymphoid tissue is a T cell-dependent process. The specific B cell-T cell interactions, and/or cytokines, resulting in germinal center cell growth have not yet been identified. Germinal center B cells were separated from other lymph node (LN) B cells by panning on peanut agglutinin (PNA)-coated dishes. Resulting fractions enriched for PNA+ (germinal center) B cells, and the PNA- (other) LN B cells from immune SJL mice were assayed for proliferation in the presence of cytokines. PNA+ and PNA- B cells responded equally to IL-4 in the anti-mu co-stimulator assay. In contrast, PNA+ B cells responded to murine (r)IL-5 or human B cell growth factor in the dextran sulfate (DxS) co-stimulator assay, to a much greater degree than did PNA- B cells. The same results were obtained with PNA+ and PNA- cells from LAF1 mice. Unfractionated LN B cells from nonimmunized SJL or BALB/c mice did not respond to IL-5 with or without DxS. B cell populations from BALB/c mice such as from spleen and peritoneal cavity, which are known to be high in Ly-1+B cells, responded to IL-5 alone, and more dramatically, to IL-5 as a co-stimulator with DxS. Such populations of cells from SJL mice, which are known to contain low numbers of Ly-1+B cells, responded markedly less. These results are consistent with those of others which show that in nonimmunized mice, Ly-1+B cells are a major IL-5 responsive subpopulation. IL-1 enhanced the proliferation of PNA+ cells in response to rIL-5 and had no effect on PNA- cells. IL-4 and IL-5 did not enhance each other's effects as co-stimulators of proliferation. In contrast to PNA+ B cells from immune LN, B cells activated by Escherichia coli endotoxin exhibited no responses to rIL-5. The present results indicate that in immune LN, PNA+, germinal center B cells constitute a prominent IL-5-responsive population.     

PNA-binding glycans are expressed at high levels on horse mature and immature T lymphocytes and a subpopulation of B lymphocytes

In mammals, the binding of peanut agglutinin (PNA) on the plasma membrane defines subpopulations among lymphocytes from peripheral blood and lymphoid organs. PNA binds Gal 1,3GalNAc residues provided that they are not sialylated. Here, we studied the expression of PNA-binding glycans on healthy horse peripheral blood, thymus, lymph node and spleen lymphocytes. We first demonstrated the binding specificity of PNA for galactose residues by competition experiments and the inhibitory role of sialic acids in PNA binding by sialidase digestion. Unlike human and murine lymphocytes, all equine lymphocytes were found positive by flow cytometry analysis. Double-staining analyses showed that lymphocytes expressing high levels of PNA-binding glycans (PNA^high lymphocytes) were made up of the great majority of CD5⁺, CD4⁺ and CD8⁺ cells, and of 30 and 50% of sIg-bearing lymphocytes in peripheral blood and in lymph nodes or spleen, respectively. Lectin histochemistry suggested that lymph node germinal centres contained PNA^high B cells. Contrary to what is found in humans and mice, PNA staining intensity on CD5⁺, CD4⁺ and CD8⁺ cells did not differentiate immature from mature T lymphocytes in the equine thymus. The functional consequences of these differences are discussed. Published in 2005.     

Relationship of germinal centers in lymphoid tissue to immunologic memory. VI. Transfer of B cell memory with lymph node cells fractionated according to their receptors for peanut agglutinin

We isolated germinal center B cells by exploiting their high affinity for peanut agglutinin (PNA). The PNA+ and PNA- B cells, fractionated by panning on PNA-coated petri dishes, were examined for their ability to transfer memory responses to irradiated recipients at various times after priming. With such fractionated B cells from lymph nodes taken at the peak of germinal center formation, the largest response was obtained in recipients of the PNA+ B cell population. At 4 to 5 wk after priming, and 10 days after challenge with an unrelated antigen, memory responses were approximately equal in recipients of PNA+ or PNA- B cells. At 14 wk after priming, memory responses were found only in recipients of the PNA- B cell population. Memory B cells from the spleen, taken from mice primed in the footpad 8 wk earlier, were also PNA-. Finally, we show that boosting with a TNP-conjugate in the footpad, 6 mo after priming in the same footpad, induced the reappearance of marked memory responsiveness in the PNA+ B cell fraction of the draining node.     

The human lymph node germinal center cell: characterization and isolation by using two-color flow cytometry

The germinal center of lymphoid tissues is a critical microenvironmental site of B cell activation and differentiation in response to antigenic stimuli. However, characterization of germinal center cells (GCC) in tissue sections has proved technically difficult. Therefore, we have employed two-color flow cytometric analysis of suspended human tonsillar lymphocytes in order to define more precisely the immunologic features of GCC. These cells were identified in suspension by virtue of their specific surface binding of the lectin peanut agglutinin (PNA), confirmed by tissue immunoperoxidase studies. Phycoerythrin-labeled lectin was used in combination with a variety of fluorescein-labeled antibodies in order to identify subpopulations of tonsillar lymphocytes. The majority of PNA+ cells were B cells, and both PNA+ and PNA- B cells stained for surface immunoglobulin light chains. PNA+ cells lacked surface IgD, but included cells with surface IgG and IgM. Both PNA+ and PNA- cells stained for B1, B2, BA-1, Leu-12, Leu-14, CR-I, and HLA-DR antigens, whereas CALLA was present only on PNA+ cells. There were differences between PNA+ and PNA- cells in the relative expression of B1 and B2 antigens, possibly reflecting differences in B cell activation or maturation. A small proportion of T cells were PNA+, including both helper/inducer and suppressor/cytotoxic phenotypes. PNA+ cells included both small and large lymphoid cells, and almost all DNA synthetic activity was associated with the large PNA+ cells. PNA+ B cells isolated by cell sorting had morphologic features characteristic of GCC. Therefore, PNA+ cells in suspension appeared to represent GCC, and features of these cells that cannot be convincingly shown in tissue section studies were demonstrated by flow cytometry.     

Peanut agglutinin. I. A new tool for studying T lymphocyte subpopulations.

Fluorescein-coupled peanut agglutinin (PNA) has been used at the single-cell level to study mouse lymphocyte subpopulations. PNA not only binds to most thymocytes, as has already been shown by other authors, but also binds to a small fraction of peripheral lymphocytes that are all T cells (theta+Ig-) or null cells (theta-Ig-). Most PNA-positive thymocytes are sensitive to in vivo corticosteroids and irradiation (450 rads) treatments. Conversely, the positive spleen cells (5% of total spleen lymphocytes) are essentially resistant to corticosteroids and irradiation. Study of PNA binding during ontogenesis shows the occurrence of PNA-positive cells in the fetal liver before thymus constitution and in the very beginning of embryonic thymus and spleen development. These data indicate that PNA is a marker of early T cell subpopulations but that there are probably several distinct subsets of PNA-positive T cells.     

Human follicular dendritic cells promote the APC capability of B cells by enhancing CD86 expression levels

Follicular dendritic cells (FDCs) are an essential cellular component of the germinal center (GC) and are believed to exert regulatory effects on the various stages of GC reactions. According to our previous reports, human FDCs express prostacyclin synthase, and prostacyclin analogues augment adhesion and co-stimulatory molecules on the surface of activated B cells. These findings prompted us to investigate whether FDCs would contribute to the antigen-presenting capability of B cells by using the well-established FDC-like cells, HK cells, and tonsillar B cells. Our results show that HK cells significantly enhance the expression levels of CD54, CD80, and CD86 on the surface of activated B cells. The enhancing effect of HK cells on CD86 is impeded by indomethacin and an EP4 antagonist, implying that a certain prostaglandin is mediating the up-regulation. Prostacyclin indeed recapitulates the enhancing effect on CD86, which is inhibited by EP4 as well as IP antagonists. B cells co-cultured with HK cells exhibit an augmented APC activity, which is inhibited by CD86 neutralization. These results reveal another unrecognized function of human FDC.     

Induction of isotype switching and Ig production by CD5+ and CD10+ human fetal B cells.

In the present study the capacity of early fetal B cells to produce Ig was investigated. It is shown that B cells from fetal liver, spleen, and bone marrow (BM) can be induced to produce IgM, IgG, IgG4, and IgE, but not IgA, in response to IL-4 in the presence of anti-CD40 mAb or cloned CD4+ T cells. Even splenic B cells from a human fetus of only 12 wk of gestation produced these Ig isotypes. IFN-alpha, IFN-gamma, and transforming growth factor-beta inhibited IL-4-induced IgE production in fetal B cells, as described for mature B cells. The majority of B cells in fetal spleen expressed CD5 and CD10 and greater than 99% of B cells in fetal BM were CD10+. Highly purified CD10+, CD19+ immature B cells and CD5+, CD19+ B cells could be induced to produce Ig, including IgG4 and IgE, in similar amounts as unseparated CD19+ B cells. Virtually all CD19+ cells still expressed CD10 after 12 days of culture. However, the IgE-producing cells at the end of the culture period were found in the CD19-,CD10- cell population, suggesting differentiation of CD19+,CD10+ B cells into CD19-,CD10- plasma cells. Pre-B cells are characterized by their lack of expression of surface IgM (sIgM). Only 30 to 40% of BM B cells expressed sIgM. However, in contrast to sIgM+,CD10+,CD19+ immature B cells, sorted sIgM-,CD10+,CD19+ pre-B cells failed to differentiate into Ig-secreting cells under the present culture conditions. Addition of IL-6 to these cultures was ineffective. Taken together, these results indicate that fetal CD5+ and CD10+ B cells are mature in their capacity to be induced to Ig isotype switching in vitro as soon as they express sIgM.     

Resolution of three nonproliferative immature splenic B cell subsets reveals multiple selection points during peripheral B cell maturation.

Although immature/transitional peripheral B cells may remain susceptible to selection pressures before full maturation, the nature and timing of these selection events remain unclear. We show that correlated expression of surface (s) IgM (sIgM), CD23, and AA4 defines three nonproliferative subpopulations of immature/transitional peripheral B cells. We designate these populations transitional (T) 1 (AA4(+)CD23(-)sIgM(high)), T2 (AA4(+)CD23(+)sIgM(high)), and T3 (AA4(+)CD23(+)sIgM(low)). Cells within all three subsets are functionally immature as judged by their failure to proliferate following sIgM cross-linking in vitro, and their rapid rate of turnover in vivo as assessed by 5-bromo-2'-deoxyuridine labeling. These labeling studies also reveal measurable cell loss at both the T1-T2 and T2-T3 transitions, suggesting the existence of multiple selection points within the peripheral immature B cell pool. Furthermore, we find that Btk-deficient (xid) mice exhibit an incomplete developmental block at the T2-T3 transition within the immature B cell pool. This contrasts markedly with lyn(-/-) mice, which exhibit depressed numbers but normal ratios of each immature peripheral B cell subset and severely reduced numbers of mature B cells. Together, these data provide evidence for multiple selection points among immature peripheral B cells, suggesting that the B cell repertoire is shaped by multiple unique selection events that occur within the immature/transitional peripheral B cell pool.     

Antigen receptor engagement selectively induces macrophage inflammatory protein-1 alpha (MIP-1 alpha) and MIP-1 beta chemokine production in human B cells

We show herein that B cell Ag receptor (BCR) triggering, but not stimulation by CD40 mAb and/or IL-4, rapidly induced the coordinated expression of two closely related T cell chemoattractants, macrophage inflammatory protein-1 beta (MIP-1 beta) and MIP-1 alpha, by human B cells. Naive, memory, and germinal center B cells all produced MIP-1 alpha/beta in response to BCR triggering. In contrast to MIP-1 alpha/beta, IL-8, which is spontaneously produced by germinal center B cells but not by naive and memory B cells, was not regulated by BCR triggering. Culturing follicular dendritic cell-like HK cells with activated B cells did not regulate MIP-1 alpha/beta production, but it did induce production of IL-8 by HK cells. Microchemotaxis assays showed that CD4+CD45RO+ T cells of the effector/helper phenotype actively migrated along a chemotactic gradient formed by BCR-stimulated B cells. This effect was partially blocked by anti-MIP-1 beta and anti-CC chemokine receptor 5 Ab, but not by anti-MIP-1 alpha Ab suggesting that MIP-1 beta plays a major role in this chemoattraction. Since maturation of the B cell response to a peptide Ag is mostly dependent on the availability of T cell help, the ability of Ag-stimulated B cells to recruit T cells via MIP-1 alpha/beta, may represent one possible mechanism enabling cognate interactions between rare in vivo Ag-specific T and B cells.     

Immunophenotypic characterization of primary and secondary lymphoid follicles

The need for an immunophenotypical referential framework relative to lymphoid follicle has led us to apply a panel of monoclonal and polyclonal antibodies, by means of a sensitive immunostaining method. Lymphoid follicle is an immunophenotypically complex structure made up of three lymphoid populations (B, being its bulk, and a few T and NK cells), dendritic reticulum cells (DRCs) and Flemming's macrophages. Follicular B population is To 15 +, B1+, OKB 7 +, HLA-DR + and C3bR +. In secondary follicles there are differential characteristic reactivities for each topographic compartment: Mantle zone is positive for OKB 2 and surface IgM (sIgM) and IgD (sIgD); germinal center (GC) clear zone (with centrocytic predominance) for OKT 9, sIgM and weakly for OKB 2; and GC dark zone (with centroblastic predominance) only for OKT 9. In sections, OKT 10 allows one to see immunoblasts and plasma cells, the latter being with lymphoplasmacytoid cells the only intracytoplasmic immunoglobulin holders. 10% of GC lymphocytes are T cells, almost exclusively T-helper (Leu 3a +). Another 10% to 15% of lymphoid cells are Leu 7 (HNK-1) +. In histological sections, DRCs are specifically marked with R4/23 and Flemming's macrophages with anti-alpha1-antitrypsin and anti-alpha1-antichymotrypsin antibodies, both populations being negative to OKM 1 and OKM 5.