Membrane Ig-cytoskeletal interactions. I. Flow cytofluorometric and biochemical analysis of membrane IgM-cytoskeletal interactions

Membrane IgM (mIgM) and mIgD are the receptors for Ag on the surface of B lymphocytes, mIg is soluble in detergent; however, when mIg is cross-linked with anti-Ig, the mIg becomes associated with the cytoskeletal matrix and is rendered detergent-insoluble. By a novel flow cytofluorometric assay and by biochemical analysis, it has been shown that anti-isotype-specific antibodies induce mIgM and mIgD to associate with the cytoskeleton of B lymphocytes in an isotype-specific fashion. The detergent solubility of other prominent B lymphocyte surface proteins, such as class I and class II MHC proteins were unaffected by cross-linking of mIg. A panel of mu-specific mAb was analyzed for their ability to induce mIgM-cytoskeletal association. Although all mAb bound mIgM, only three out of seven rendered mIgM cytoskeletally associated. Further analysis revealed a strict correlation in the capacity of mu-specific mAb to induce capping and to induce the association of mIgM with the cytoskeleton.     

Evolutionary origin of the T-lymphocyte receptor. II. Production and partial characterization of antisera to the immunoglobulin-like cell membrane protein from thymocytes of the carp

An immunoglobulin-like protein from Triton X-100 solubilized cell surface proteins from thymocytes can be isolated by specific precipitation using anti-IgM antibodies. Immunizing rabbits with those immune complexes we succeeded to elicit anti-thymocyte mIg sera reacting with both thymocyte mIg and IgM. SDS-PAGE analysis revealed that anti-IgM and anti-thymocyte mIg sera recognize the same thymocytic surface protein. The cross-reactivity between thymocyte mIg and IgM is caused by protein determinants very similar to each other. Anti-thymocyte mIg sera absorbed with IgM lose their capacity to bind thymocyte mIg. However, following absorption with thymocytes anti-IgM sera still possess amounts of antibodies reacting with humoral IgM.     

Structure and biologic functions of human IgD. XI. Identification and ontogeny of a rat lymphocyte immunoglobulin having antigenic cross-reactivity with human IgD.

Chicken anti-human IgD antiserum (anti-delta) has demonstrated an antigenically cross-reactive homologue on rat lymphocytes. IgD and IgM are the only cell surface immunoglobulins detectable by the lactoperoxidase radiolabeling technique employed. The results indicate that, although rat surface IgD is antigenically distinct from rat IgM, the respective H chains co-electrophorese in 10% polyacrylamide-SDS gels. Rat delta-chain has an apparent m.w. of 73,000 daltons and exhibits a minor 65,000 dalton component which probably represents a partially degraded delta-chain. The ontogenic emergence of rat IgD occurs approximately 3.5 weeks after birth whereas IgM, in contrast, is apparent by 6 days of age. Thus, as in the human, IgM develops before IgD. IgD receptors are undetectable in the thymus but are present in increasing levels in spleen, blood, lymph nodes, and Peyer's patches.     

Identification of three major target molecules of IgM antilymphocyte autoantibodies in systemic lupus erythematosus

Three cell lymphocyte antigens of m.w. 55,000, 70,000, and 105,000 to 110,000 were identified by Western blotting as targets of IgM autoantibodies in serum from a group of 49 patients with systemic lupus erythematosus. The 55- and 70-kDa antigens were well expressed on unstimulated peripheral T cells, whereas the 105- to 110-kDa target was demonstrable only on mitogen-activated T cells and lymphoblastoid T cell lines. Localization of these molecules to the plasma membrane was established by cytoabsorption experiments in which IgM antibody staining of blotted antigens was specifically absorbed from systemic lupus erythematosus serum during 4 degrees C incubations with intact lymphocytes, and by their detection in purified lymphocyte plasma membranes. While the identity of these target antigens vis a vis known surface determinants was not defined, their expression on peripheral T cells from multiple donors and on cell lines of both undifferentiated (HSB-2) and phenotypically mature (Jurkat; HUT 78) types excluded alloantigens, major histocompatibility complex-encoded determinants, and most T cell differentiation antigens as candidates in this regard. Expression of the IgM autoantibody targets on HSB-2 cells argues against discrete T subset specificities as well. IgM reactivity with the 55-, 70-, and 105- to 110-kDa antigens by blotting was highly correlated with antilymphocyte antibody activity in complement-dependent cytotoxicity assays (Fisher's p less than 0.001), and paralleled flow microfluorimetric and microcytotoxicity quantitation of IgM antibody activity in serial observations of individual patients studied during different phases of disease activity. Taken together, these data suggest that IgM lymphocytotoxic antibodies in systemic lupus erythematosus are directed predominantly against a limited number of non-T cell subset-specific antigens.     

Comparison of the Fc receptors for IgE on human lymphocytes and monocytes

Fc receptors for IgE (Fc epsilon R) on human peripheral blood lymphocytes and monocytes and cultured lymphoblastoid and macrophage-like cell lines were compared with respect to: 1) binding affinity for radiolabeled IgE, 2) inhibition of IgE-specific rosette formation and inhibition of binding of radiolabeled IgE by an antiserum raised against Fc epsilon R isolated from a lymphoblastoid cell line, and 3) m.w. of radiolabeled cell surface proteins precipitated with the anti-Fc epsilon R serum. Scatchard analysis of 125I-IgE binding to lymphocytes, monocytes, and their corresponding cell lines showed biphasic binding curves with all cell types, from which 2 binding affinities were calculated to be KA = 6.2 +/- 1.1 and 2.0 +/- 0.5 x 10(7) M-1. The anti-Fc epsilon R serum inhibited both IgE rosette formation and binding of radiolabeled IgE by lymphocytes and monocytes but did not inhibit IgE rosettes formed by basophils. The inhibitory activity of the anti-Fc epsilon R serum could be absorbed with Fc epsilon R(+) but not with Fc epsilon R(-) cell lines. The anti-Fc epsilon R serum precipitated 2 peptides having m.w. of approximately 47,000 and 23,000 daltons from lysates of both cell surface-labeled lymphocyte and macrophage cell lines. These data indicate that Fc epsilon R on normal lymphocytes and monocytes, as well as on cultured lymphoblastoid and macrophage-like cells, are related structurally, since they share antigenic determinants, bind IgE with a similar affinity, and have similar m.w. However, they differ in all 3 parameters from Fc epsilon R on basophilic granulocytes.     

Snapper (Pagrus auratus) leucocyte proliferation is synergistically enhanced by simultaneous stimulation with LPS and PHA

Important channels of communication between mammalian leucocytes have long been recognised. Here, data are reported that suggest similar integrations may occur between snapper leucocytes upon mitogen stimulation. Cell surface immunoglobulin (IgM) expression was used in conjunction with intracellular fluorescence staining and flow cytometry to differentiate proliferating peripheral blood leucocyte subsets (PBLs). Independent activation using phytohaemagglutinin (PHA) or lipopolysaccharide (LPS) drove both mIg(-)and mIg(+)cells into cycle. It is not known if the proliferation of mIg(+)cells was mediated by a mutually exclusive effect of the mitogen on each cell population, cognate cellular interaction or a soluble growth factor. Simultaneous activation of PBLs with PHA and LPS consistently induced significantly more cells to proliferate than the sum of proliferating cells stimulated solely with PHA or LPS. Together, the results suggest that different leucocyte subsets have the capability to influence their respective responses to mitogenic stimulation. Therefore, like in the mammalian immune system, communication may occur between snapper leucocyte subsets.     

The MB-1/B29 heterodimer couples the B cell antigen receptor to multiple src family protein tyrosine kinases.

The B cell Ag receptor complex is comprised of membrane (m)IgM or mIgD noncovalently associated with one or more heterodimers, each containing one subunit of MB-1 (IgM alpha or IgD alpha) and one of B29 (Ig beta or Ig gamma). It is known that cross-linking of the B cell Ag receptor results in protein tyrosine kinase activation. Recent reports from other laboratories have demonstrated that mIg coprecipitates with multiple src family protein tyrosine kinases, including blk, lyn, and fyn. However, the mechanism by which these kinases are physically coupled to the Ag receptor has not been confirmed. It has been hypothesized that the mIg-associated proteins MB-1 and B29 provide a physical link between the Ag receptor (mIg) and one or more protein tyrosine kinases. In this study, we confirm previous findings demonstrating that the B cell Ag receptor coprecipitates with the MB-1/B29 heterodimer as well as the protein tyrosine kinases blk, lyn, and fyn under mild detergent conditions (1% digitonin). Additionally, we demonstrate that in detergent conditions (1% Nonidet P-40 (NP-40)) which disrupt the association between mIg and the MB-1/B29 heterodimer, no protein tyrosine kinase activity can be detected in association with mIg. These findings indicated that NP-40 effectively dissociates the B cell Ag receptor from ancillary signal transducing proteins. MB-1 and B29 were however, found to coprecipitate with blk, lyn, and fyn isolated from B cell lysates containing 1% NP-40. No significant difference was observed in the stoichiometry of association between the kinases and the MB-1/B29 heterodimer in the presence of 1% NP-40 when compared to 1% digitonin. It was further determined that in resting B cells, only a small fraction (approximately 1-3%) of the MB-1/B29 heterodimers appear to be complexed with protein tyrosine kinases. Finally, based on preclearing experiments, it appears that individual heterodimers may associate with a single species of protein tyrosine kinase. These data support the hypothesis that the MB-1/B29 heterodimer couples the antigen receptor to protein tyrosine kinases, thereby providing a physical link that facilitates Ag receptor-mediated regulation of kinase activity.     

High anti-TNP plaque-forming cell potential of residual mIg+ cells in a T cell population

In the course of experiments designed to study the immune response of purified populations of B lymphocytes to thymus-independent (TI) antigens, a variety of cell purification procedures were followed. In using anti-immunoglobulin-coated dishes to separate lymphocytes bearing membrane immunoglobulin (mIg) from mIg- lymphocytes, it was found that the nonadherent fraction, which was predominantly mIg-, complement receptor negative, and nonresponsive to the B cell mitogen lipopolysaccharide, gave very substantial anti-TNP plaque-forming cell responses to 2 TI antigens. These responses could be inhibited by incubation of such cells in the presence of anti-mu and thus appeared to be attributable to mIg+ cells. The evidence suggests the existence of a population of B lymphocytes that constitute a minor component of mIg+ cells having a high potential to make in vitro antibody responses. Users of techniques that utilize anti-Ig as a tool for separating B and T lymphocytes should carefully assess the extent to which residual B lymphocytes in the mIg- population contribute to antibody responses being studied.     

Tyrosine phosphorylation of phospholipase C-gamma 2 upon cross-linking of membrane Ig on murine B lymphocytes.

When membrane Ig (mIg) on the surface of B lymphocytes is cross-linked using anti-Ig antibodies, the enzyme phospholipase C (PLC) is activated to cleave inositol phospholipids. Tyrosine kinase inhibitors have been reported to inhibit this event. Therefore, we investigated the effect of cross-linking of mIg on the state of tyrosine phosphorylation of PLC activity in two murine B cell lines and in normal resting mouse B cells. Proteins from lysates of stimulated or unstimulated cells were immunoprecipitated with an antiphosphotyrosine antibody and subsequently assayed for PLC activity. Treatment of the B cell line WEHI-231 with anti-IgM led within 15 to 30 s to a 10- to 20-fold increase in tyrosine-phosphorylated PLC activity. Inositol trisphosphate generation by WEHI-231 cells stimulated under the same conditions demonstrated similar kinetics. Normal resting B cells treated with anti-IgM or anti-IgD demonstrated 2.5- and 4-fold increases, respectively, of tyrosine-phosphorylated PLC activity. To identify the isozyme of PLC that was phosphorylated, we immunoprecipitated PLC-gamma 1 or PLC-gamma 2 with specific antibodies and assessed the amount of tyrosine phosphorylation of these proteins by antiphosphotyrosine immunoblotting. Treatment of WEHI-231 or Bal17 cells with anti-IgM induced an increase in PLC-gamma 2 tyrosine phosphorylation over background levels. There was no detectable tyrosine phosphorylation of PLC-gamma 1 in treated or untreated WEHI-231 cells, whereas anti-IgM-treated Bal17 cells did exhibit low but detectable levels of tyrosine phosphorylation of PLC-gamma 1. In normal resting mouse B cells, there was no detectable PLC-gamma 1, but PLC-gamma 2 was abundant. These observations suggest that PLC-gamma 2 is a significant substrate for the mIg-activated protein tyrosine kinase and may be responsible for mediating mIg stimulation of inositol phospholipid hydrolysis in murine B cells.     

Translocation of protein kinase C during membrane immunoglobulin-mediated transmembrane signaling in B lymphocytes

Previous studies have implicated a role for protein kinase C (PKC) in transmembrane signal transduction by B cell surface immunoglobulin (Ig). Specifically, the pharmacologic PKC activator phorbol myristate acetate mimics the biologic effects of mIg cross-linking ligands, and cross-linking of membrane Ig (mIg) induces polyphosphoinositide hydrolysis generating diacylglycerol, a potent activator of PKC. Studies described here additionally implicate PKC in mIg-mediated signaling by demonstrating rapid translocation of activatable PKC (PKCa) from cytosol to Triton-soluble membrane fractions after cross-linking of cell surface IgM or IgD. This response, which is also induced by phorbol myristate acetate and lipolysaccharide, is detectable within 1 min of mIg cross-linking and is followed within 4 min by additional translocation of PKCa to a Triton-insoluble particulate compartment. The ability of dbcAMP plus theophylline to inhibit polyphosphoinositide hydrolysis, PKCa translocation, and the B cell's subsequent biological response suggests that these events may be causally related.     

Impaired lymphocyte functions in heterozygous rabbits recovering from neonatal allotype suppression

Lymphocytes from heterozygous rabbits suppressed for an allotypic determinant on kappa light chains by exposure to maternally derived antibodies specific for the paternal gene product were analyzed for their capacity to express membrane-bound and secreted immunoglobulin (Ig). Individual cells displaying allotypic membrane Ig (mIg) were enumerated by a rosette test, while Ig-secreting cells were assessed by means of a hemolytic plaque assay. In a group of suppressed rabbits varying in age from 3 to 19 months, the proportion of cells with mIg of the paternal type was markedly higher than that of cells secreting that type of Ig. The same high proportion of lymphocytes displaying mIg of the suppressed type was observed whether lymphocytes from blood, spleen, or lymph nodes of suppressed rabbits were examined. In contrast, similar analyses performed with cells of normal heterozygous rabbits showed no discrepancy between mIg expression and secretion of either allotype. Lymphocytes synthesizing Ig of the paternal type were also defective in responses to lipopolysaccharide (LPS), which stimulates differentiation to Ig secretion in normal B lymphocytes. These results support the idea that B lymphocytes capable of synthesizing the suppressed type of Ig have functional impairments affecting secretion and responses to environmental stimuli.     

In vitro effects of steroid hormones on IgM-secreting cells and IgM secretion in common carp (Cyprinus carpio)

The effects of steroid hormones on in vitro IgM-secreting cells (IgMSC) and IgM secretion by lymphocytes of the lymphoid organs in common carp, Cyprinus carpio were examined by ELISPOT and ELISA assay, respectively. Cells isolated from peripheral blood (PB), spleen and head kidney were cultured for 12, 24 and 48 h either in the absence or in the presence of steroid hormones, i.e. cortisol, testosterone, 11-ketotestosterone (11-KT), and estradiol-17beta (E(2)) at doses of 1, 10 and 100 ng/ml. Cortisol reduced the IgMSC numbers and IgM secretion by cells from all organs. In addition, cortisol induced apoptosis in lymphocytes from all organs. High dose of testosterone showed tissue-specific functions; it reduced the number of IgMSC and amount of IgM secretion by cells from spleen and head kidney, but not in PB, though IgM secretion was suppressed. However, no effects of sex steroids were observed in this study. The results show that sex-specific steroid hormones may have no immunosuppressive effects in common carp.     

Functional and morphologic characterization of human T cells continuously grown in vitro.

Long-term growth (now over 13 months) of thymus-derived lymphocytes from numerous normal human bone marrow and peripheral blood cell samples was accomplished by using a factor present in media obtained from mitogen-stimulated human peripheral blood lymphocytes. This long-term growth could neither be initiated nor maintained by mitogens alone. All cell cultures were greater than 90% E rosette-positive, whereas the tests for B cell markers, surface IgG and IgM, and EAC rosette were routinely negative. There was no evidence for the presence of granulocytes, monocytes, and their precursors in these cultures. The E rosette-positive cells were then tested to see if they had T cell functions. PHA, Con A, and pokeweed mitogens stimulated lymphproliferative responses in these cultures comparable to those of fresh peripheral blood cells. These proliferating cells were also able to release cell mediators, such as interferon and colony-stimulating activity. Further evidence for the T lymphocyte nature of these cultured cells was obtained from one-way mixed leukocyte cultures in which these cells responded to but were unable to stimulate allogeneic cells. The functional and morphologic characteristics of these cultured cells show that these cells are T cells that grow continuously in vitro.     

Glycoproteins of 210,000 and 130,000 m.w. on activated T cells: cell distribution and antigenic relation to components on resting cells and T cell lines

A glycoprotein complex of 210,000 and 130,000 m.w., found on mitogen or alloantigen-stimulated human T cells and not on other hematopoietic cells, has been defined by a monoclonal antibody (Mab). The components of this complex are a subset of a larger family of proteins (210,000, 165,000 and 130,000 m.w.) defined by a second Mab. In a panel of hematopoietic cell lines and cell types, only activated T cells (including the cell line HUT-102) express the 210,000/130,000 complex and these cells also express the IL 2 receptor, a characteristic marker for activated T cells. The 210,000/130,000 m.w. complex (reactive with the Mab TS2/7) is present on all long-term activated T cells, including both the OKT4 and OKT8 subsets. The 210,000 m.w. subunit is expressed only on activated T cells. Other lymphoid cells express either the 130,000 m.w. subunit alone (unactivated lymphocytes, thymocytes, HUT-78) or the 130,000 subunit together with a 165,000 subunit (MOLT-4, HSB, and other leukemic T cell lines). The 210,000/130,000 m.w., 165,000/130,000 m.w. and 130,000 m.w. complexes are antigenically related in that all share reactivity with the Mab A- 1A5 . Among non-lymphoid hematopoietic cells and cell lines, none express the 210,000 m.w. chain; adherent cells (monocytes) and myeloid cell lines each express single proteins of 130,000 to 155,000 m.w. Granulocytes and red blood cells are negative and platelets express multiple bands (165,000 and 140,000 m.w.). Immunoperoxidase staining of tissue sections showed that a broad range of tissues and cell types had material cross-reactive with the lymphoid 130,000 m.w. protein. However, only a discrete subset of those tissues and cells including blood vessel walls, connective tissue, smooth muscle, kidney mesangial cells, and some non-cellular matrix tissue, had material cross-reactive with the 210,000 m.w. protein on activated T lymphocytes.     

IL-4 promotes Stat6-dependent survival of autoreactive B cells in vivo without inducing autoantibody production

Persistent cross-linking of hen egg lysozyme (HEL)-specific B cell membrane Ig (mIg) in double transgenic mice that express soluble HEL as a self Ag (HEL-Ig mice) decreases B cell mIgM expression, responsiveness, and life span. Because in vitro treatment with IL-4 inhibits T cell apoptosis through a Stat6-independent mechanism, increases mIg expression, and suppresses activation-induced B cell death, we studied IL-4 effects on B cell mIg expression, survival, and Ab secretion in Stat6-sufficient and deficient HEL-Ig mice. IL-4 treatment nearly normalized B cell number and greatly increased the percentage of mature B cells in HEL-Ig mice, but failed to normalize mIgM expression or spontaneous LPS-induced IgM secretion. IL-4 effects on B cell survival and maturation were CD4(+) T cell independent, but Stat6 dependent, and did not involve receptor editing. IL-4 had to be present while B cells were generated to have a detectable effect on autoreactive B cell survival; however, the survival of B cells generated in the presence of IL-4 was substantially increased even after IL-4 was withdrawn. These observations suggest that: 1) activation-induced B cell death and anergy are independent processes; 2) B cells that survive to maturity develop increased resistance to Ag-induced deletion; and 3) IL-4 promotes B and T cell survival through different mechanisms.     

Separation of human lymphocyte subpopulations by density gradient electrophoresis. I. Different mobilities of T (T mu, T alpha) and B lymphocytes from human tonsils

T and B lymphocytes from human tonsils were separated by density gradient electrophoresis on the basis of their surface charge. The high-mobility cell fractions were found to be highly enriched in T lymphocytes with only very small proportions of B cells. In contrast, the low-mobility fractions were predominantly B lymphocytes, and had only 10 to 30% contamination of T cells. The intermediate-mobility fractions contained both T and B lymphocytes in approximately equal proportions. IgM-bearing lymphocytes, as well as cells with receptors for mouse erythrocytes, the Fc portion of IgG, and complement were found in the intermediate- and low-mobility fractions. T lymphocytes, prepared by E rosetting, were also electrophoresed by this method and found to be of higher mobility as compared with peripheral blood T lymphocytes. T cells with Fc receptors for IgM (Tμ) or IgA (Tα) were found to be considerably heterodisperse with regard to surface charge and were present in all fractions. The separated cell fractions were treated in vitro with various concentrations of concanavalin A and thereafter examined for Tμ, Tγ, and Tα phenotypes. Low concentrations of Con A (2.5 μg/ml) had no effect on cell surface phenotypes. However, higher concentrations of Con A (20μg/ml) significantly reduced the numbers of T cells having IgM receptors (Tμ), but failed to alter the expression of the Tγ phenotype. The latter finding contrasts to that observed with T cells from the peripheral blood where high concentrations of Con A increase the proportions of the Tγ cells. This study demonstrates that density gradient electrophoresis can be used for the separation and study of lymphocyte subpopulations from human tonsils.     

Determination of IgM and IgG on lymphocyte surfaces by means of registering rosette formation with anti-immunoglobulin-coated erythocytes

To identify IgM and IgG on the surface of normal bovine B-lymphocytes, the test of antiglobulin-rosette formation (anti-Ig rosettes) was used. IgM and IgG were found to exist on separate B-lymphocyte populations. The composition of receptors on the surface of lymphocytes, forming Ig rosettes, was investigated. The distribution of these lymphocytes in bovine blood and lymphoid organs was determined. IgM and IgG were detected on the surface of thymocytes, their origin remains open to question.     

Membrane immunoglobulin: anti-immunoglobulin interactions mediate the phosphorylation of actin associated proteins in the B-lymphocyte

The binding of membrane immunoglobulin (mIg) by anti-Ig antibodies is known to initiate a mitogenic signal in B lymphocytes. Because in many instances growth control appears to be correlated with phosphokinase activity, as well as with alterations in cytoskeletal architecture, we asked the question whether antibodies binding to mIg would also lead to the specific phosphorylation of lymphocyte actin-associated proteins. Utilizing a myosin affinity technique, we directly examined phosphoproteins that were associated with actin in the chicken B cell. We found that in a few instances the level of phosphorylation was indeed modulated by mIg:anti-Ig interactions. These actin-binding phosphoproteins may be important control elements in the lymphocyte cytoskeleton.     

Location of nucleotide pyrophosphatase and alkaline phosphodiesterase activities on the lymphocyte surface membrane.

1. Isolated mouse spleen lymphocytes hydrolysed UDP-galactose added to the medium. Nucleotide pyrophosphatase activity that accounted for this hydrolysis was enriched to a similar extent as alkaline phosphodiesterase and 5'-nucleotidase in a lymphocyte plasma-membrane fraction. 2. The cell surfaces of mouse spleen and thymus lymphocytes were iodinated with 125I by using the lactoperoxidase-catalysis method. Detergent extracts of the cells were mixed with a purified anti-(mouse liver plasma-membrane nucleotide pyrophosphatase) antiserum and the immunoprecipitates analysed by polyacrylamide-gel electrophoresis. Only one major radioactive component, similar in size (apparent mol.wt 110000-130000) to the liver enzyme, was observed. 3. Electrophoresis of an iodinated spleen plasma-membrane fraction indicated peaks of radioactivity, including one of apparent mol.wt 110000-130000. 4. When detergent extracts of spleen lymphocytes were passed through a Sepharose-bead column containing covalently attached anti-(nucleotide pyrophosphatase) antiserum, the nucleotide pyrophosphatase activity was retained by the beads, whereas protein and leucine naphthylamidase activity were eluted. 5. The results indicate that nucleotide pyrophosphatase and alkaline phosphodiesterase activities are due to the location of the same or similar enzymes at the outer aspect of the lymphocyte plasma membrane. Some possible functions of enzymes at this location are discussed.     

Memory B cell survival and function in the absence of secreted antibody and immune complexes on follicular dendritic cells

Ag, in the form of immune complexes retained on follicular dendritic cells, has been implicated in the development and maintenance of B cell memory. We addressed this question using a H chain transgenic (Tg) mouse model that lacks secreted Ig (mIg), and thus does not deposit Ag-containing immune complexes. We compared the ability of the mIg strain and a control Tg strain, which secretes IgM, to develop and maintain long-lived memory cells. After immunization, there was an increase of Ag-specific B cells in both strains that was maintained for at least 20 wk. We labeled the long-lived Ag-specific cells with BrdU and found that this population was similarly maintained. In addition, both Tgs were able to maintain a functional memory response as measured by secondary germinal center reactions. Our studies indicate that localization of Ag on follicular dendritic cells is not necessary for development and maintenance of B cell memory.