Search

The germinal center reaction is essential for efficient humoral immunity, but it can also give rise to B cell lymphomas. Cre/loxP-mediated conditional gene knock-out or knock-in can be used for the genetic manipulation of germinal center B cells in vivo. Here we present a novel allele, Cγ1-CreERT2, that allows for timed activation of Cre recombinase in a small fraction of germinal center B cells. This allele will be useful to study normal and malignant germinal center B cell development in vivo.

Medullary thymic epithelial cell (mTEC)-restricted expression of autoimmune regulator (Aire) is essential for establishment of immune tolerance. Recently, Aire was also shown to be expressed in cells of hematopietic and reproductive lineages. Thus, the generation of Airefl/fl mouse strain enables the investigation of the cell-specific function of Aire.

Germinal centers (GCs) are generally considered to be the sole site of memory B-cell generation. However, recent studies demonstrate that memory B cells can also develop in response to a T-cell dependent (TD) antigen before the onset, and independently of, the GC reaction. These two classes of memory cells persist equally over long periods of time and attain functional maturation through distinct but related transcriptional programs. Although the development of both memory B-cell types requires classical T-cell help, the generation of GC-dependent memory B cells requires TFH -cell help, while the generation of GC-independent memory cells does not. These findings led to the conclusion that B-cell memory is generated along two fundamentally distinct cellular differentiation pathways. In this review, we focus on the GC-independent pathway of memory B-cell development, and discuss how the unique features of memory B cells are maintained in the GC-independent pathway.

When I started my work at the Institute for Genetics in Cologne in 1964, I was the only immunologist among ambitious young molecular biologists. They used mutagenesis in bacteria and bacteriophages to understand biological processes, while I studied antibody responses and immunological tolerance to hapten-carrier conjugates in rabbits. Although this led to a first model of T cell-B cell collaboration, it was frustratingly clear that ultimately immunological problems would have to be resolved by methods of molecular genetics. Twenty five years later immunology had become one of our Institute's major research areas in molecular biology. Developing conditional gene targeting in mice to study immune regulation along the lines of the work of the young molecular biologists at the time, a dream had come true. In this essay, I present this story which represents the slightly modified text of a five-minute speech given in acceptance of the Novartis Prize for basic immunology in August 2007 in Rio de Janeiro.

The expression of the preB cell receptor (preBCR), composed of the mu chain, surrogate light chains and the Igalpha /Igbeta signal transduction unit, permits further differentiation of Bcell precursors. C57BL/6 mice homozygous for an inactivating mutation of the membrane exon of the mu chain gene (C57BL/6muMT/muMT)) cannot form a preBCR and are, consequently, devoid of mature B lymphocytes. Here we present evidence that the block of B cell development by the muMT mutation is incomplete in BALB/c mice. Unlike C57BL/6muMT/muMT), BALB/cmuMT/muMT) mice generate small numbers of mature B cells, accumulate plasma cells and produce high levels of all immunoglobulin isotypes, except IgM. The observed phenomenon seems to be controlled by a single genetic locus that is not linked to IgH.

The interaction of CD95 with its ligand CD95L is important for negative selection of B cells during the germinal center (GC) reaction. Recently, mutations confering restistance to CD95-induced apoptosis have been described for human GC B cells. Hence, as has been demonstrated for CD95-deficient mice, also GC-derived autoreactive B cells carrying somatic CD95 gene mutations may potentionally survice negative selection and participate in the development of autoimmune diseases. Here, single plasmablasts (PB) which are implicated in the production of autoantibodies in systemic lupus erythematosus (SLE) patients as well as ten human B cell lines producing autoantibodies were analyzed for destructive somatic CD95 gene mutations. However, inactivating CD95 gene mutations were very rare in PB and not detected in the cell lines. Sequence analysis of V gene rearrangements amplified from single PB confirmed that the cells are (post) GC B cells and additionally demonstrated massive clonal expansion of these cells in two of four SLE patients. We conclude that CD95 gene mutations play little if any role in the generation of the pool of PB in SLE patients and that mutations in the CD95 gene are rare among autoantibody-producing B cells in SLE and rheumatoid arthritis.

Somatic hypermutation (SH) in B cells undergoing T cell-dependent immune responses generates high-affinity antibodies that provide protective immunity. Most current models of SH postulate the introduction of a nick into the DNA and subsequent replication-independent, error-prone short-patch synthesis by one or more DNA polymerases. The PolK (DinB1) gene encodes a specialized mammalian DNA polymerase called DNA polymerase κ (polκ), a member of the recently discovered Y family of DNA polymerases. The mouse PolK gene is expressed at high levels in the seminiferous tubules of the testis and in the adrenal cortex, and at lower levels in most other cells of the body including B lymphocytes. In vitro studies showed that polκ can act as an error-prone polymerase, although they failed to ascribe a clear function to this enzyme. The ability of polκ to generate mutations when extending primers on undamaged DNA templates identifies this enzyme as a potential candidate for the introduction of nucleotide changes in the immunoglobulin (Ig) genes during the process of SH. Here we show that polκ-deficient mice are viable, fertile and able to mount a normal immune response to the antigen (4-hydroxy-3-nitrophenyl)acetyl-chicken γ-globulin (NP-GC). They also mutate their Ig genes normally. However, polκ-deficientembryonic fibroblasts are abnormally sensitive to killing following exposure to ultraviolet (UV) radiation, suggesting a role of polκ in translesion DNA synthesis.

In order to determine whether V gene replacement accompanies somatic hypermutation in the germinal center (GC) reaction in the human, we analyzed V(kappa)J(kappa) and V(lambda)J(lambda) joints and the kappa-deleting element in single lambda(+) naive and post GC B cells for rearrangements at the kappa and lambda loci. Among 265 lambda(+) post GC B cells, not a single unequivocal and only two potential examples of a cell that switched to lambda light chain expression after accumulation of (unfavorable) mutations in its productive V(kappa) rearrangement were observed. Taking the PCR efficiency into account, the frequency of such cells is likely below 3 %. In addition, heavy and light chain gene rearrangements were amplified and sequenced from the oligoclonal population of IgD-only peripheral blood post GC B cells which display extensive intraclonal sequence diversity. Among 61 IgD-only B cells belonging to 15 clones with intraclonal diversity, no combination of V gene rearrangements indicating receptor revision during clonal expansion was observed. Moreover, among 124 and 49 V(H) genes amplified from IgD-only and class-switched B cells, respectively, not a single example of V(H) revision through V(H) hybrid generation was detected. These results suggest that in the human GC reaction V gene replacement either does not usually accompany somatic hypermutation or is mostly counterselected.

Defects in the common cytokine receptor gamma chain (gammac) in man result in X-linked severe combined immunodeficiency disease (SCIDX1) characterized by an absence of alphabeta T cells, gammadelta T cells and NK cells, with the presence of circulating B cells. Mice made deficient for gammac lack gammadelta T cells and NK cells, but in contrast to SCIDX1 patients have appreciable numbers of alphabeta T cells, while B cells are reduced about tenfold in numbers and disappear with age. Here we show that when gammac- mice are rendered T cell deficient, B cell numbers are still reduced but the age-dependent loss of B cells does not occur. The peripheral B cells which persisted in gammac-/ nude and gammac-/TCRbeta-/- mice were able to respond to mitogen stimulation in vitro and to mount antigen-specific T-independent Ig responses in vivo. These results demonstrate that gammac- B cells are functionally competent and suggest that residual alphabeta T cells are implicated in the B cell loss in gammac mice. The gammac-/nude and gammac-/TCRbeta-/- mice provide new models to dissect the role of gammac-dependent receptors during murine B cell differentiation.

Mesenteric lymph nodes (MLN) drain the gut where nutritive antigens and pathogens are encountered by lymphocytes of the gut-associated lymphoid tissue. We sought to determine how lymphocytes enter the MLN by studying mice double deficient for beta7 integrins and L-selectin. beta7/L-selectin double-deficient lymphocytes did not migrate into MLN. Most importantly, MLN formation was drastically impaired in beta7/L-selectin double-deficient mice. Lymphocyte numbers in MLN from beta7/L-selectin double-deficient mice were tenfold reduced compared to control mice. A high percentage of the few lymphocytes still detected in MLN from beta7/L-selectin double-deficient mice were CD44hi CD18hi, suggesting alternate migration pathways independent of L-selectin and beta7 integrin for these cells. We conclude that the combination of both molecules, L-selectin and beta7 integrin, is indispensable for MLN formation and that these molecules may mediate lymphocyte migration to MLN in a sequential and synergistical manner.

Despite accounting for only a minor fraction of all cells in Hodgkin's lymphoma tissue, the Hodgkin and Reed-Sternberg (HRS) cells represent the malignant tumor cell clone in Hodgkin's disease (HD). By far the most abundant cell type in the tumor tissue are CD4+ T cells. Some of them intimately associate with HRS cells forming rosettes around them. This study addresses the question whether the rosetting phenomenon reflects a specific interaction between T and HRS cells by asking whether the rosettes are composed of T cells expressing a restricted TCR repertoire. Single rosetting T cells were micromanipulated from frozen sections of tumor tissue in two cases of nodular sclerosing HD and one case of lymphocyte predominant HD. TCR Vbeta gene rearrangements were amplified from these single cells by PCR. Of 83 potentially functional Vbeta gene rearrangements obtained altogether from the three cases, 81 were found to be clonally unrelated. Furthermore, they did not show signs of selection of the receptor chains for recognition of common epitopes: The usage of Vbeta and Jbeta gene segments as well as the distribution of complementarity-determining region (CDR) 3 lengths was similar to what was seen in a collection of 60 Vbeta gene rearrangements from blood of healthy donors and no recurrent CDR3 amino acid motifs were found. These data suggest that the HRS cells attract CD4+ T cells nonspecifically.

In the Revised European American Lymphoma (REAL) classification, several subtypes of high-grade lymphomas were combined in the entity diffuse large cell lymphoma (DLL). In the present study, a total of 19 cases of DLL (10 cases of centroblastic lymphoma, 5 cases of mediastinal B cell lymphoma, 2 cases of immunoblastic lymphoma, 1 case of T cell-rich B lymphoma and one case of large cell anaplastic lymphoma) were analyzed for somatically mutated immunoglobulin V region genes. Somatic mutations are acquired in the course of the germinal center (GC) reaction and are thus found in GC B cells and their descendants, i.e. memory B cells. The V gene sequences revealed that the tumor cells of all five subtypes of DLL harbored mutated V region genes and are thus derived from antigen-experienced (post) GC B cells. This indicates that from the point of view of the stage of development of the tumor precursor, the combination of those five subtypes to one entity, i.e. DLL, seems reasonable. In some cases, an unusually high frequency of somatic mutations was detected. This may indicate that DLL are derived from GC B cells, which, due to transforming events, stayed in the GC for prolonged periods of time, thereby accumulating a high load of somatic mutation. An analysis of the mutation pattern suggests that the tumor clone or its precursor were selected for antibody expression while acquiring somatic mutations. The latter observation discriminates DLL from classical Hodgkin's disease, where we recently also observed a high load of somatic mutation within rearranged V region genes, but a frequent occurrence of crippling mutations.

We have generated a mutant mouse in which the most D-proximal V(H) gene (V(H)81X) has been disrupted by introducing a neomycin-resistance gene into the V(H)81X exon by means of gene targeting in embryonic stem cells. The mutant mice generated are unable to express the V(H)81X gene but appear to display a normal pattern of B cell differentiation as well as normal numbers of bone marrow and peripheral B cells from fetal life all through ontogeny. They mount normal immune responses to several different antigens tested. In contrast, the distribution of V(H) gene rearrangements in the V(H)7183 family is altered in homozygous mutant mice. Thus, the antibody repertoire of the targeted mice is modified, at least as far as the expression of V(H)7183 genes is concerned.

Five cases of monocytoid B cell lymphoma (MBCL) were analyzed for somatic mutations in the rearranged V region genes. Somatic mutations were found in four of the five cases, whereas one unusual CD5+ lymphoma harbored unmutated V region genes. Since somatic mutations are introduced into V region genes of antigen-activated B cells in the course of T cell-dependent immune responses, these results suggest a derivation of the tumor B cells in MBCL from antigen-experienced mature B cells. An analysis of the kappa-deleting element in two of the cases in which mutated VH but unmutated and nonfunctional V kappa gene rearrangements were found suggests that somatic hypermutation does not take place in human rearranged V kappa region genes when the C kappa gene and the kappa enhancers have been deleted in cis by rearrangement of the kappa-deleting element.

We investigated gene rearrangements in the mutant IgH locus of a mouse strain generated by insertion of a rearranged heavy chain variable region gene (VT15) into the DQ52-JH region through gene targeting. In more than half of the B cells of heterozygous mutant mice, the mutant IgH locus was silenced by the rearrangement of an endogenous DH or DH and VH gene to the inserted VT15 gene. In these cases, a functional VHDHJH gene was present on the wild-type allele. The silencing rearrangement appeared to be mediated by recombination signal sequence (RSS)-like elements present in the "recipient" VT15 gene. Among the many such elements on the inserted VT15 gene, which apparently met the requirement for an RSS with respect to nucleotide sequence, only two were observed in the actual rearrangements. This indicates that targeting of the recombination machinery involves sequences in addition to the RSS motifs as they have been characterized so far. In homozygous mutant mice, most B cells appeared to carry the intact VT15 gene on both mutant IgH alleles, although single-cell polymerase chain reaction revealed that silencing rearrangements occurred frequently in B cell progenitors in the bone marrow. This observation indicates that once silencing rearrangements are initiated in a cell, they involve both VT15 genes in most cases, reminiscent of normal DH-JH rearrangement. B cells which did not initiate such rearrangements develop to populate the peripheral B cell compartment.

Peripheral blood B cells of a 67-year-old person were separated into IgM+IgD+, IgM+IgD-, and IgM-IgD- subsets, and nucleotide sequences of expressed immunoglobulin light chain variable (V) regions encoded by V kappa 3 and V kappa 4 gene family members were determined from amplified cDNA. V region sequences from IgM+IgD+ cells (the major B cell population in the blood) showed no or little somatic mutation (0.3%), in contrast to V kappa sequences from IgM+IgD- and IgM-IgD- B cells (2.0% and 3.9%, respectively). This suggests that in the human like in the mouse, and independently of age, somatically mutated memory B cells accumulate in the compartment of IgM-IgD- cells, whereas the IgM+IgD+ subpopulation consists of cells whose antibody repertoire is mainly determined by V region gene rearrangements and N-region insertion, at the molecular level. The somatically mutated IgM+IgD- cells may represent early descendants of IgM+IgD+ cells recruited into the memory cell compartment.

Antibody diversity is generated by a random gene recombination process with the inherent risk of the production of autoreactive specificities. The current view suggests that B cells expressing such specificities are negatively selected at an early developmental stage. Using the knock-in model system of the 3–83 autoreactive B-cell antigen receptor (BCR) in combination with precursor-BCR (pre-BCR) deficiency, we show here that the 3–83 BCR mediates efficient generation of B cells in the presence, but not the absence, of a strongly recognized auto-antigen. Experiments with mixed bone marrow chimeras showed that combining the 3–83 BCR with the corresponding auto-antigen resulted in efficient reconstitution of B-cell development in immune-deficient mice. These results suggest that B cells are positively selected by recognition of self-antigens during developmental stages that precede receptor editing. Moreover, the data indicate that the pre- BCR functions as a specialized autoreactive BCR to initiate positive selection at a stage where the cells express immunoglobulin heavy but not light chains.

The accumulation of somatic mutants in splenic B lymphocytes early after primary immunization with the hapten (4-hydroxy-3-nitro-phenyl)acetyl (NP) coupled to chicken gamma globulin (CG) was determined. Rearranged V186.2 heavy chain genes were amplified by the polymerase chain reaction from genomic DNA and subjected to nucleotide sequence analysis. Somatic antibody mutants become detectable on day 6 after immunization, and most of the somatic mutations accumulating in the memory compartment are introduced until day 14. At this time strong selection for mutants expressing high binding affinity for NP is apparent. Extrapolation from the mutation frequency increases between day 6 and day 14 to the previously determined mutation frequency at week 6 (Weiss. U. and Rajewsky, K., J. Exp. Med. 1990, 172: 1681) leads to the prediction that the process of mutant generation ceases to operate around day 22 after primary immunization.

The murine interleukin 4 (IL4)-encoding cDNA expressed under the control of the immunoglobulin heavy chain enhancer/promoter was inserted into the mouse germ line. All B cells present in the IL4 transgenic mice show a marked increase in expression of class II histocompatibility (Ia) antigens compared to B cells from control mice. In vivo 5-bromo-2'-deoxyuridine incorporation and cell cycle analysis revealed no significant increase in B cell proliferation in spleens of IL4 transgenic mice compared to littermate controls. Likewise, serum immunoglobulin concentrations in IL4 transgenic mice were not increased. From these experiments we conclude that Ia hyperexpression on B cells per se does not induce activation of autoreactive T cells leading to B cell proliferation or hypergammaglobulinemia.

Somatic hypermutation (SH) in B cells undergoing T cell-dependent immune responses generates high-affinity antibodies that provide protective immunity. Most current models of SH postulate the introduction of a nick into the DNA and subsequent replication-independent, error-prone short-patch synthesis by one or more DNA polymerases. The Pol kappa (DinB1) gene encodes a specialized mammalian DNA polymerase called DNA polymerase kappa (pol kappa), a member of the recently discovered Y family of DNA polymerases. The mouse PolK gene is expressed at high levels in the seminiferous tubules of the testis and in the adrenal cortex, and at lower levels in most other cells of the body including B lymphocytes. In vitro studies showed that pol kappa can act as an error-prone polymerase, although they failed to ascribe a clear function to this enzyme. The ability of pol kappa to generate mutations when extending primers on undamaged DNA templates identifies this enzyme as a potential candidate for the introduction of nucleotide changes in the immunoglobulin (Ig) genes during the process of SH. Here we show that pol kappa-deficient mice are viable, fertile and able to mount a normal immune response to the antigen (4-hydroxy-3-nitrophenyl)acetyl-chicken gamma-globulin (NP-GC). They also mutate their Ig genes normally. However, pol kappa-deficient embryonic fibroblasts are abnormally sensitive to killing following exposure to ultraviolet (UV) radiation, suggesting a role of pol kappa in translesion DNA synthesis.

The interaction of CD95 with its ligand CD95L is important for negative selection of B cells during the germinal center (GC) reaction. Recently, mutations conferring resistance to CD95-induced apoptosis have been described for human GC B cells. Hence, as has been demonstrated for CD95-deficient mice, also GC-derived autoreactive B cells carrying somatic CD95 gene mutations may potentially service negative selection and participate in the development of autoimmune diseases. Here, single plasmablasts (PB) which are implicated in the production of autoantibodies in systemic lupus erythematosus (SLE) patients as well as ten human B cell lines producing autoantibodies were analyzed for destructive somatic CD95 gene mutations. However, inactivating CD95 gene mutations were very rare in PB and not detected in the cell lines. Sequence analysis of V gene rearrangements amplified from single PB confirmed that the cells are (post) GC B cells and additionally demonstrated massive clonal expansion of these cells in two of four SLE patients. We conclude that CD95 gene mutations play little if any role in the generation of the pool of PB in SLE patients and that mutations in the CD95 gene are rare among autoantibody-producing B cells in SLE and rheumatoid arthritis.

The expression of the preB cell receptor (preBCR), composed of the mu chain, surrogate light chains and the Igalpha /Igbeta signal transduction unit, permits further differentiation of Bcell precursors. C57BL/6 mice homozygous for an inactivating mutation of the membrane exon of the mu chain gene (C57BL/6muMT/muMT)) cannot form a preBCR and are, consequently, devoid of mature B lymphocytes. Here we present evidence that the block of B cell development by the muMT mutation is incomplete in BALB/c mice. Unlike C57BL/6muMT/muMT), BALB/cmuMT/muMT) mice generate small numbers of mature B cells, accumulate plasma cells and produce high levels of all immunoglobulin isotypes, except IgM. The observed phenomenon seems to be controlled by a single genetic locus that is not linked to IgH.

Southern blot analyses of immunoglobulin light chain gene rearrangements in human leukemias and myelomas indicated that lambda loci in kappa-producing cells are largely unrearranged while kappa loci in lambda producers are often rearranged and inactivated by rearrangements of the kappa-deleting element (KDE). For a systematic analysis of the regulation of light chain rearrangements during early B cell development in normal human B cells also considering functionality of the rearrangements, we used FACS-sorted single naive kappa- and lambda-expressing B cells from peripheral blood of healthy humans. V(kappa)J(kappa) and V(lambda)J(lambda) joints and rearrangements involving the KDE were amplified simultaneously from single cells and sequenced. Whereas only 2 - 3 % of kappa-expressing cells carry V(lambda)J(lambda) joints, nearly all lambda-expressing cells have rearranged kappa loci and indeed carry V(kappa)J(kappa) joints. The V(kappa)J(kappa) joints in lambda-expressing cells exhibit preferential J(kappa)4 and J(kappa)5 over J(kappa)1 and J(kappa)2 usage compared to kappa-expressing cells. Thirty percent of the V(kappa)J(kappa) joints in lambda producers are rearranged in-frame. These data indicate extensive sequential V(kappa)-J(kappa) rearrangements and inactivation of functional V(kappa)J(kappa) joints in lambda-expressing cells, presumably before V(lambda)J(lambda) joining.

B cell chronic lymphocytic leukemia (B-CLL) is a disease of the elderly and is characterized by a malignant clone of CD5+ B cells. In old mice, clonal expansions of CD5+ B cells are a common feature, and these animals frequently develop B-CLL. To investigate whether clonal expansion of CD5+ B cells also occurs in elderly humans, predisposing for the development of B-CLL, we analyzed VH gene rearrangements of CD5+ B cells from blood samples of four healthy, 65-82-years-old volunteers as markers of clonality. CD5+ and CD5-B cells were obtained by cell sorting, CDRIII of rearranged VH genes were amplified by polymerase chain reaction, and fragment length analysis was performed. All samples demonstrated a polyclonal pattern of VH gene length distribution. In addition, VH gene rearrangements were amplified and sequenced from sorted single cells of two of the donors. No clonally related CD5+ or CD5- B cells were observed. Thus, development of dominant clones of CD5+ peripheral blood B cells is unlikely to be a common trait of elderly individuals.

The CD5 antigen is expressed at a high level on T cells from early on in thymocyte ontogeny and continues to be expressed on the surface of all mature T cells. In addition, it marks a population of B lymphocytes (B-1a) with distinct physiological properties. To study the in vivo function of CD5, the murine gene was inactivated using the technique of homologous recombination in embryonic stem cells. In homozygous mutant mice the CD5 antigen is not expressed on the surface of either T or B lineage cells, indicating that in both cell populations this antigen is encoded by the same gene. CD5-deficient (CD5T) mice are healthy and populations of T and B lymphocytes in these mice look unchanged when compared to control mice. The mutant mice are able to mount effective immune responses to T cell-dependent and -independent antigens.

Interleukin (IL)-2 and IL-4 are considered as important regulators of growth and differentiation of lymphocytes. We report that in mice made deficient for both IL-2 and IL-4 by gene targeting all major T cell subsets and B cells were normal, indicating that IL-2 and IL-4 are not essential for development of the immune system. Paradoxically, proliferation of T cells was increased in both IL-2 and IL-4-deficient homozygous mice.

In the present study we have addressed the question of whether the composition of the peripheral B cell pool is similar in mice raised under conventional, specific pathogen-free or germ-free breeding conditions. For this purpose, the representation of various B lineage subsets was quantitated by flow cytometry and the lifespan of B cells analyzed by labeling dividing cells with the thymidine analogue 5-bromo-2'-deoxyuridine in vivo. The results indicate that except for an increased fraction of CD45R (B220)dull B lineage cells in the bone marrow and a marginally faster turnover of splenic B cells in germ-free compared to conventional mice, there is no obvious difference in the representation and turnover of B cells in the different kinds of mice. The data indicate that a stable pool of peripheral B cells is generated even in the absence of microbial stimulation.

Labeling of dividing cells with the thymidine analogue 5-bromo-2′-deoxyuridine (BrdUrd) has been frequently used to determine the life-span of cells in various lymphocyte subsets. In this study we address the question of whether administration of BrdUrd in vivo allows labeling of proliferating lymphocytes at early as well as later stages of maturation with high efficiency. For this purpose, the proportion of BrdUrd-labeled cells among all cells synthesizing DNA was determined in the bone marrow and spleen of mice that had received a single injection of BrdUrd or, alternatively, had been fed with BrdUrd for 1 week. On the average, 87% of bone marrow and 83% of spleen cells in the S phase of the cell cycle, as determined by DNA content, were labeled with BrdUrd 1.5 h following injection of the drug. We conclude from these data that administration of BrdUrd in vivo allows labeling of the vast majority of the dividing cells in both primary and secondary lymphoid organs and thus represents an efficient method to determine the life-span of lymphoid cells in vivo.

The concentration of IgG2a in the serum of C57BL/6 mice suppressed for the production of IgG2a was measured over a period of 3 months after suppression. The elimination of a minimum of 96% of the endogenous IgG2a molecules from the serum follows an exponential decay with a half-life of 6-8 days. A monoclonal IgG2a antibody injected into suppressed or normal mice is eliminated from the serum in the same fashion. We conclude that there are few, if any, "long-lived" IgG2a molecules present in the serum of C57BL/6 mice under physiological conditions.

We have partially sequenced the light chain variable regions expressed in three IgM-producing hybridomas generated from newborn mice or from manipulated animals suppressed for IgM production. In these lines a new V gene (V-lambda-X), exhibiting less than 60% homology to any known lambda or kappa V gene, is rearranged to J-lambda-2. The light chains produced by these cells contain the lambda-2 constant domain, but are not recognized by goat antisera raised against conventional mouse lambda light chains.

We determined the half-lives of several sets of murine monoclonal antibodies spanning all immunoglobulin isotypes in the serum. The antibodies in each set possess the same V region. With this approach, the differences in half-life observed between the different isotypes are independent of the V region carried by the monoclonal antibodies and therefore must relate to each other in the same way as the half-lives of each class of serum immunoglobulins. The half-life of a monoclonal antibody of the gamma 2a isotype is identical to the average half-life of serum IgG2a as previously determined (6-8 days; P. Vieira and K. Rajewsky, Eur. J. Immunol. 1986. 16:871). Therefore, the half-lives determined with monoclonal antibodies possessing the same V region represent the half-life of the serum immunoglobulins. In this way we calculated the half-life of IgM as 2 days, IgG3 and IgG1 as 6-8 days, IgG2b has a half-life of 4-6 days. IgE has a half-life of 12 h. A polymeric form of IgA was found to be eliminated from the serum with a half-life of 17-22 h.

Primary anti-NP [(4-hydroxy-3-nitrophenyl)acetyl] antibodies and NP-specific receptors isolated from T cells of mice carrying the Ig-1b allotype express the NPb idio-type and a fine specificity marker (heteroclicity) which are both controlled by V genes in the heavy chain linkage group. While in antibodies both markers are invariably expressed in association with Ig light chains of the Λ type, antigen-binding T cell-derived receptors appear not to follow this rule in that they do not significantly bind to anti-Λ antisera. Mice of the SJL strain also carry the Ig-lb allotype but express a defect in Λ chain synthesis, and their primary anti-NP response is thus almost completely devoid of Λ-bearing heteroclitic molecules of the NPb idiotype. In contrast, both NPb idiotype and heteroclicity are expressed in the majority of T cell-derived receptors. This result permits a positive discrimination between T cell-derived receptors and humoral antibodies in the SJL strain and shows that V gene expression follows different rules in the T and the B cell compartment. The expression of Ig V regions in isolated T cell-derived receptors was also analyzed with antisera binding to presumed “framework” determinants of the VH and the VΛdomains of Ig. A reaction of anti-VH antiserum with isolated T cell receptors could clearly be demonstrated. The anti-VΛ antiserum may bind to a minor fraction of the NP-specific molecules in our T cell receptor preparations.

Antigen-independent cooperation between T and B lymphocytes is demonstrated in vitro in two different experimental protocols: (a) B cells from A/J mice immunized in vivo either with Group A streptococcal vaccine (Strep.A) or with the IgG1 fraction of guinea pig anti-idiotypic antibody to the A5A idiotype, mature into plaque-forming cells (PFC) with specificity for Group A streptococcal carbohydrate (A-CHO) during a 4-day culture together with T cells from A/J mice immunized in vivo with A5A idiotypic antibody. (b) B cells from A/J mice immunized in vivo with Strep.A generate PFC specific for A-CHO when cultured in the presence of small concentrations of anti-A5A idiotypic antibody and of T cells primed with Strep.A. In both cases, antigenindependent cooperation is idiotypically selective, such that only those B cells respond that secrete antibody with the A5A idiotype. The data are interpreted to suggest that, in addition to antigen-specific helper cells, idiotype-specific may participate in antibody responses, and that the latter type of help may be responsible for the idiotypic selectivity in T-B cooperation observed previously. Furthermore, idiotype-specific cooperation may be a means to generate and maintain B cell diversity during the evolution and ontogeny of the immune system.

Eight monoclonal antibodies from the primary response of C57BL/6 mice against the hapten (4-hydroxy-3-nitrophenyl)acetyl (NP) were isolated. The antibodies carry lambda 1 light chains and have similar affinities for the immunizing hapten. Sequence analysis at the level of mRNA reveals that all antibodies express the VH gene 186.2 and all but one the DFl 16.1 gene segment. The J segment of the heavy chain is JH2 in six cases and JH4 in two. Somatic point mutations are scarcely detectable in the antibodies, but there is extensive sequence variability at the boundaries of the D gene segment, mainly at its 5' end. However, seven of eight antibodies express tyrosine in position 99 of the heavy chain, encoded either by the 5' codon of DFl 16.1 or by presumed N sequences. In the former case, the tyrosine is the first of a stretch of three (positions 99-101). In the latter, a similar stretch (positions 99, 101, 102) is interrupted by aspartic acid, asparagine or cysteine in position 100. These variations profoundly affect idiotypic specificity. Six of the eight monoclonal antibodies came from mice neonatally suppressed by an anti-idiotope antibody whose target idiotope is regularly expressed in primary anti-NP responses and depends upon a non-germ-line-encoded aspartic acid in position 100 of the heavy chain. The sequence data show that the mice circumvent suppression by expressing antibodies which lack this aspartic acid but are otherwise structurally very similar to anti-NP antibodies from normal animals. Since suppression in the animals is partly controlled by regulatory T cells, we conclude that these T cells are highly restricted in their specificity in that they preferentially see a determinant which also depends upon the aspartic acid in position 100. The data suggest that the VH to D boundary serves as a target of idiotypic selection.

Monoclonal antibodies detecting idiotopes on the germ line-encoded anti-(4-hydroxy-3-nitrophenyl)acetyl (NP) antibody B1-8 were injected at various doses into newborn mice and the expression of B1-8 idiotopes was measured in anti-NP responses in later life. Suppression was long lasting, and a 100-fold increase in the dose of anti-idiotope delayed recovery from suppression by 5-6 weeks. Upon injection of a single anti-idiotope, suppression was observed for all B1-8 idiotopes to various degrees. Certain idiotopically defined antibody phenotypes were much more efficiently suppressed, and later recovered from suppression, than others. This specificity pattern was observed at the level of both B and T cells from the manipulated animals, as demonstrated in cell transfer experiments in which such cells were mixed with normal T and B cells. In these experiments, there was evidence for suppression mediated by regulatory T (and possibly also B) cells. Whereas the B cells from the manipulated animals were idiotypically unresponsive in a T cell-dependent adoptive primary response, the frequency of lipopolysaccharide-reactive B cells expressing the target idiotype was only slightly reduced in these animals as compared to control mice. Together with data on the elimination of anti-idiotope antibody from the neonatally injected animals these results are interpreted in the following way: idiotype suppression is induced through the reaction of anti-idiotope with idiotopes expressed on the surface of newly generated B cells, at microgram concentrations of anti-idiotope. When the concentration of anti-idiotope falls below that level, recovery from suppression sets in. Two types of suppression are induced. The first, namely, direct blockade of B cell maturation, is short-lived. The second involves the induction of regulatory cells, perhaps through idiotope-bearing antibody V regions complexed by anti-idiotope. This type of suppression is long-lived and its specificity depends upon the distribution of the target idiotope in the antibody repertoire and/or peculiarities of the T cell receptor repertoire. It impinges on the selection of the B cell repertoire in the animal as expressed in T cell-dependent (and possibly other) responses and is thus hardly seen at the level of lipopolysaccharide-reactive (immature) cells. Idiotype suppression by regulatory cells may be perpetuated by antigen interacting with idiotypic antibodies on the B cell surface and may therefore play a role in establishing tolerance not only for the expressed antibody repertoire, but for self antigens in general.

Spleen cells from C57BL/6 mice sensitized to the hapten (4-hydroxy-3-nitrophenyl)acetyl (NP) were hybridized with myeloma cells, and a variety of hybrid cell lines was isolated which secreted homogeneous anti-NP antibodies. The antibodies were purified by affinity chromatography and their chain composition, affinity and fine specificity were determined. All antibodies recovered from the primary immune response carried λ light and μ or γ1 heavy chains. Their variable portions were nonidentical but similar in terms of hapten-binding specificity with a higher affinity for the cross-reacting haptens (4-hydroxy-3,5-dinitro-phenyl)acetyl (NNP) and (4-hydroxy-5-iodo-3-nitro-phenyl)acetyl (NIP) than for the homologous hapten NP. This heteroclitic property as well as the presence of λ, μ and γ1 chains are characteristic for primary anti-NP sera of C57BL/6 mice. In contrast, four families of anti-NP antibodies, each with a characteristic fine specificity pattern, were found among the hybrid cell antibodies derived from the hyperimmune anti-NP response. The antibodies of one of these families were related to the antibodies recovered from the primary immune response in that they were heteroclitic and carried λ light chains. All members of the other groups expressed k chains and were nonheteroclitic. The finding of well-defined antibody families in this system and the isolation of their members enable us to approach the problem of V gene expression and diversification in a new way.

A self-I-A reactive T cell clone has been isolated that proliferates in the presence of irradiated syngeneic spleen cells in mouse serum. The cells directly stimulate approximately every hundredth B cell, which includes small B cells, into both proliferation and antibody formation in the absence of added antigen. There is evidence to suggest that cells of this type are generally present in murine T cell populations. The cells may recognize self-I-A as such. We failed to obtain evidence, but cannot exclude that they recognize an antibody idiotype in addition. The clone also augments antigen-specific antibody responses in vivo and in vitro in the presence of antigen. In the in vitro response to a T-independent TI-2 antigen, the T cells in co-culture with B cells induced specifically the production of IgG1 antibodies, thus mimicking the function of major histocompatibility complex-restricted, antigen-specific helper T cells. However, IgG1 production was not observed when splenic T cells were added to this system. Thus, self-I-A reactive T cells, generally inactive in the system, may locally help B cell responses to antigens which the T cells do not themselves recognize. In such instances one might attribute to the T cells antigen or idiotype-binding specificities which the cells in reality do not display.

H-2-linked immune response (Ir) genes specifically and distinctly control immunological responsiveness of mice to the non-cross-reacting enzyme antigens LDHA and LDHB. In a previous study (Melchers, I., Rajewsky, K. and Shreffler, D. C., Eur. J. Immunol. 1973. 3: 754) two levels of responsiveness to LDHB were found and mice could be classified into high and low responders (HR or LR). We now detect a third, intermediate level of responsiveness. In addition, in certain crosses, “superhigh” responses as well as responses in between the low and intermediate level are found. An apparently highly polymorphic system of Ir genes thus determines the level of responsiveness in a surprisingly discrete and precise way. In general, high or intermediate responsiveness was dominant over low responsiveness in F1 animals. In one out of four cases the cross between an intermediate responder (IMR) and a LR, which were congenic for H-2, resulted in a level of responsiveness in between that of the two parental strains. In crosses between IMR strains, the F1 animals were either again IMR or showed functional complementation to high responsiveness in two out of four combinations. The results can be explained in terms of two polymorphic loci which interact in the control of responsiveness to LDHB. Surprisingly, the B10.HTT strain, which carries the recombinant H-2t3 chromosome, is an IMR, although the H-2t3 chromosome carries the HR H-2s allele in the Ir-1B region, to which Ir-LDHB has been previously mapped. Complementation to high responsiveness is achieved when the H-2t3 chromosome is combined in an F1 cross with the H-2 chromosome from LR strains carrying “silent” HR alleles in the I-C and/or S regions (H-2a and H-2h2). This demonstrates that the response to LDHB is controlled by at least two separate, interacting loci within the H-2 complex. One of these loci maps, at least in the case of H-2s and H-2d, to the left of the I-C region, whereas the other is located in I-C or S, as again shown for the H-2d and H-2s haplotypes. One may speculate that these loci also mediate complementation between our IMR strains (see above) and exert distinct functions in the Ir gene control of the immune response.

It was previously shown that in C57BL/6 mice chronic suppression of an idiotypically defined subset of lambda 1 chain-bearing anti-(4-hydroxy-3-nitrophenyl)acetyl (NP) antibodies is achieved by neonatal administration of allogeneic monoclonal anti-idiotope antibodies reactive with this subset, or by NP coupled to mouse immunoglobulin. The present data show that isologous monoclonal anti-idiotope antibodies have the same effect. In contrast, antibodies against constant region determinants of lambda 1, mu or delta immunoglobulin chains failed to induce chronic suppression of the same antibody subset. Furthermore, the effect of the anti-idiotope antibodies was neutralized by idiotypic antibodies of the IgG1 class, injected before or together with the anti-idiotype. These results suggest that the mere complexing of idiotypic molecules on the B cell surface or in the circulation is insufficient for the induction of chronic idiotype suppression. In the present system, induction appears to require the binding of a ligand to idiotype-bearing receptor V regions, expressed on the surface of B (or T?) cells.

In the present study we use two monoclonal anti-idiotope antibodies (Ac38 and Ac146; Ab1) against the germ line-encoded, lambda 1 chain-bearing and (4-hydroxy-3-nitrophenyl)acetyl (NP)-binding antibody B1-8 (Ab1) for the induction of complementary antibodies (Ab3) and ask the question to what extent antibodies bearing B1-8-like idiotopes (Ab3 beta) are represented in the Ab3 population. In this experimental system, Ab3 beta is distinguished from the remaining Ab3 population (Ab3 alpha) by three properties which Ab3 beta may share with B1-8: (a) the binding of NP, (b) the binding to Ac146 (if induced by Ac38) or the binding to Ac38 (if induced by 146) and (c) that they carry lambda 1 chains. Antibodies with all three properties are induced in low amounts by both anti-idiotopes. Also, Ac146 induces only Ab3 alpha (bearing kappa chains and not binding NP and Ac38). In contrast, Ac38 triggers almost exclusively a lambda 1 chain-bearing response, i.e. Ab3 beta. The response has an unusually large size, reaches its maximum after a week and is long-lasting. An analysis at the level of lipopolysaccharide-reactive precursor B cells demonstrates that, in this case, cells expressing Ab3 beta occur at exceedingly high frequency (approximately equal to 10(-3] and are at least 10 times more frequent than cells expressing Ab3 alpha. The high frequency of Ab3 beta-expressing cells correlates with the contribution of several VH, D and J genes to the expression of this particular idiotope. In the case of the Ac146 anti-idiotope antibody, the response is dominated by Ab3 alpha. Ab3 beta represents less than 10% of the total response, reaches maximal levels 2 weeks after immunization and declines rapidly. This correlates with a low frequency of precursor B cells expressing Ab3 beta (approximately equal to 10(-5] and a restriction of the corresponding idiotope to rare VH-D combinations, caused presumably by a stringent contribution of the H chain to this idiotope which covers the NP-binding site. Our data suggest that anti-idiotypic antibodies (Ab2) against Ab1 will preferentially induce antibodies idiotypically related to Ab1 if the corresponding idiotopes are expressed in high frequency in the B cell compartment. This is expected in cases where Ab2 recognizes an idiotope that can be formed by many germ line-encoded VH-D-VL-combinations.

In the response to NP-lipopolysaccharide or NP-Ficoll predominantly anti-NP antibodies of the IgM class are produced in mice with lower amounts of IgG3 and IgG2b but little or no IgG1 and IgG2a. In contrast, in the primary T-dependent response to NP-keyhole limpet hemocyanin (KLH) or NP-chicken gamma globulin high amounts of all IgG isotypes are induced. To investigate whether isotype-specific T cells are responsible for these differences we carried out cell transfer experiments using carrier-specific T cell lines. Two such lines were established and one of the two could be cloned. Upon activation by antigen the T cell lines induced unprimed syngeneic splenic B cells to proliferate and differentiate into antibody-secreting cells in vitro in an antigen-nonspecific way. Antigen-specific activation of unprimed B cells in a cell transfer system in vivo showed that high concentrations of hapten-specific antibodies of all IgG isotypes are induced through both carrier-specific T helper lines. The isotypic pattern of these antibodies is similar to that produced via heterogeneous splenic T cells in the cell transfer system, or in normal animals on immunization with the same antigen. These results suggest that isotype-specific T cells are not required for the production of IgG isotypes in a primary anti-NP response and thus not responsible for the differences seen in isotypic patterns between T-dependent and T-independent responses.

The relationship of serologically defined determinants to determinants recognized by cytotoxic T cells on molecules encoded by the Kk gene of the murine major histocompatibility complex (H-2) has been analyzed. For this purpose we used three somatic variants of a Kk-expressing lymphoma line lacking individual determinants of the Kk molecule, as defined by monoclonal antibodies (mAb), as target cells for Kk-specific alloreactive and Kk-restricted cytotoxic T lymphocytes (CTL) cloned by limiting dilution. Neither alloreactive nor fluorescein isothiocyanate, influenza- or Newcastle disease virus-specific Kk-restricted CTL clones were found to distinguish between variants and wild type cells, indicating that the serologically defined determinants lost by the variants were not essential for antigen recognition of CTL with these specificities. On the other hand, two of the variants lacking either one of a pair of serological determinants were discriminated from Kk wild type cells by about 40% of Kk-restricted, trinitrophenol (TNP)-specific CTL clones. The third variant, lacking both of the determinants, however, was lysed by all CTL clones to the same extent as wild type cells. From these results we conclude that the determinants restricting the TNP-specific CTL were also not identical with those defined by mAb. In experiments performed to optimize the conditions for the limiting dilution analysis we found that the specificity of the CTL stimulation was strongly dependent on the concentration of T cell growth factor (interleukin 2) in the cultures during CTL stimulation. High concentrations of IL2 resulted in a drastic increase in the frequency of CTL clones. Part of these clones, however, were found not to be specific for antigens present on the stimulator cells.

In the present analysis we dissect the idiotype repertoire, independently of hapten-binding specificity, by immunizing different strains of mice with cross-linked monoclonal anti-idiotope antibodies against antibody B1-8. B1-8 is a monoclonal antibody with specificity for the hapten (4-hydroxy-3-nitro-phenyl)acetyl (NP) and carries a germ line gene-encoded variable region. The results demonstrate that the expression of B1-8 idiotopes and their association with each other and with NP-binding specificity are strain-specific. Certain idiotopes are expressed on antibodies differing in antigen-binding specificity, whereas one of the idiotopes appears strictly associated with NP-binding antibodies. The genetic analysis provides strong evidence that the strain specificity of the idiotope repertoire is a result of V region polymorphism in the mouse.

Splenic T and B lymphocytes from sensitized mice were adsorbed to antigen-coated nylon discs and realesed from the discs by temperature shift as described by Kiefer, H., (Eur. J. Immunol. 1973. 3: 181 and 1975. 5:624). After cell release, lymphocyte-derived antigen-binding material could be recovered from the discs. A major fraction of the activity (70 -80%) binds to anti-immunoglobulin immunosorbents. A minor fraction does not detectably cross-react with gamma, mu, alpha and kappa immunoglobulin poly-peptide chains. When filtrated through Sephadex G-200, the bulk of activity of both fractions elutes in the region of 7 S serum antibody. However, as demonstrated for (4-hydroxy-3-nitro-phenyl)acetyl (NP)-binding material from C57BL/6 lymphocytes, the average affinity for antigen of the two fractions is drastically lower than that of humoral antibody, and the average affinity of the minor fraction is lower than that of the major fraction. The fraction of lymphocyte-derived antigen-binding material that does not adsorb to insolubilized anti-immunoglobulin serum was found to be proportional to the fraction of T lymphocytes in the input cell population and may therefore represent antigen-binding T lymphocyte surface receptors.

Strain C57BL/6 mice produce a highly restricted primary response to the (4-hydroxy-3-nitrophenyl)acetyl (NP) group. This response is composed of molecules having mu or gamma1 heavy chains and light chains of the lambda type. A guinea pig anti-idiotype was raised to the purified C57BL/6 primary anti-NP immunoglobulin. After suitable absorption, this anti-idiotype was shown to detect markers present on the primary anti-NP immunoglobulin only of those strains which express the Ig-1b allotype. Breeding experiments demonstrated that the marker segregated with the heavy chain linkage group.

Transfer of peritoneal cells from adult donors into newborn, allotype-congenic mice led to colonization of the recipient mice by donor-derived B lymphocytes expressing the Ly-1 surface marker (Ly-1 B cells). These cells not only persisted in the recipient mice for at least 5 months, but also increased in number. In contrast, bone marrow-derived stem cells did not or scarcely give rise to B cells after intraperitoneal injection into congenic newborn recipients under the same experimental conditions. Approximately half of the IgM in the serum of peritoneal cell-recipients was produced by donor-derived Ly-1 B cells, suggesting that high levels of serum IgM in a normal mouse are produced by this B cell subpopulation. The transferred Ly-1 B cells were able to respond in a normal fashion to alpha (1----3)dextran, but they did not participate in thymus-dependent and -independent (TI II) immune responses to the hapten (4-hydroxy-3-nitrophenyl)acetyl (NP). In neither the immune response to alpha (1----3)dextran nor that to NP were we able to detect an influence of the transferred Ly-1 B cells on the selection of the idiotypic repertoire in the recipient mice.

In C57BL/6 and BALB/c mice, 2–3% of immunoglobulin-positive cells express λ1 and 1–2% λ2 chains at the cell surface. These percentages are constant from the time around birth until adult age. The levels of λ1 chains in serum antibodies correspond closely to those in the B cell population in both mouse strains whereas in the case of λ2 chains, they are three times lower in serum antibodies than in the B cell population. In strain SJL, λ2-bearing B cells occur in similar frequencies as in BALB/c and C57BL/6, both in newborn and adult animals. However, the level of λ2-bearing immunoglobulins in the serum is 30 times below that expected on the basis of the cellular analysis and 10 times lower than in BALB/c and C57BL/6. B cells expressing λ1-like light chains were found in normal frequency in newborn SJL mice with a monoclonal rat anti-λ1 antibody. In adult animals such cells were present in a frequency 100 times below that characteristic for BALB/c and C57BL/6 mice. The concentration of immunoglobulin reactive with rat anti-λ1 antibody was 300 times lower in SJL than in BALB/c and C57BL/6 serum, in accord with previous data (W. Geckeler et al., J. Exp. Med. 1978. 148: 1122). The notion that in SJL mice a control mechanism specifically suppressess λ chain expression is further supported by an analysis of the early primary anti-(4-hydroxy-3-nitro-phenyl)acetyl (NP) response. We show that this response is initially dominated by antibody-forming cells secreting antibodies with λ1-like light chains. These cells rapidly disappear and appear to be selected against particulary in the IgG response. The monoclonal SJL anti-λ1 antibody Ls136 (M. Reth et al., Eur. J. Immunol. 1979. 9: 1004) does not stain any B cells in SJL mice, whereas it detects λ1-bearing cells in C57BL/6 and BALB/c animals in the expected frequency.

When the hapten (4-hydroxy-3-nitrophenyl)acetyl (NP) is presented on different carrier molecules, different anti-NP antibody responses are stimulated. On stimulation with NP-lipopolysaccharide (LPS) [T-independent type 1 (TI-1) antigen] kappa + antibodies are the major population, whereas on stimulation with NP-Ficoll [T-independent type 2 (TI-2) antigen], NP-keyhole limpet hemocyanin (KLH) or NP-chicken gamma globulin (CG) [T-dependent (TD) antigens], lambda 1+ antibodies dominate. The relative contribution of idiotopes Ac38 or Ac146 to the lambda 1+ anti-NP response was also different on comparison of TI-1 with TI-2 or TD anti-NP responses. We investigated whether light chain- or idiotype-specific T cells are responsible for these differences. Analysis of the anti-NP response of nude mice after immunization with NP-Ficoll showed lambda 1 dominance. Likewise primary adoptive transfer experiments using carrier-specific T cell lines to reconstitute the TD anti-NP response to NP-KLH or NP-CG, showed that help from carrier-specific T cells alone is capable of stimulating the characteristic lambda 1 dominant response. No significant difference could be found in the levels of Ac38 and Ac146 idiotope expression between mice reconstituted with splenic T cells and those reconstituted with T cell lines. These results suggest that light chain- or idiotype-specific T cells are required neither for the production of lambda 1 light chain dominance, nor for the appearance of idiotopes characteristic of the primary anti-NP response. The possible reasons for differences seen in both light chain and idiotope expression between primary anti-NP responses to the TI-1 antigen NP-LPS and those to TD or TI-2 antigens are discussed.

CBA and SJL mice were immunized against the hybrid cell antibody Bl-8 (μ,λ1) which we have previously isolated from the primary response of C57BL/6 mice against the hapten (4-hydroxy-3-nitro-phenyl)acetyl (NP) (Eur. J. Immunol. 1978. 8: 393). Cell lines secreting monoclonal antibodies against B1–8 idiotopes were isolated, and the anti-idiotope antibodies were used to characterize the variable portion of B1–8 and of other anti-NP antibodies all belonging to an antibody family which carries a major idiotypic marker, namely the NPb idiotype. Three groups of idiotopes were defined by the monoclonal anti-idiotope antibodies on the B1–8 protein; a first, in the NP binding site so that the corresponding anti- idiotopes compete with the NP hapten for binding to B1–8 (group 1); a second, presumably further apart from the binding site so that there is competition of the corresponding anti-idiotopes with a polyvalent NP-protein conjugate but not the free hapten for binding to B1–8 (group 2); and a third (group 3), specific for a determinant constructed by B1–8 μ chains and MOPC 21 χ chains which are still present in the B1–8 preparation as the product of the X63-Ag8 parent cell line. Group 1 idiotopes were expressed on 1–10% of χ1-bearing primary anti-NP antibodies of C57BL/6 origin and group 2 idiotopes on 10–100% of these antibodies. Group 3 idiotopes were not detected in anti-NP antisera. The analysis of a set of monoclonal hybrid cell anti-NP antibodies from C57BL/6 confirmed that group 2 idiotopes are more abundantly expressed in the primary anti-NP response than any given group 1 idiotope. Group 2 idiotopes can thus be expressed independently of group 1 idiotopes. In addition, group 1 idiotopes appear to exhibit a higher degree of diversity than idiotopes of group 2. The present analysis dissects the NPb idiotype into a large set of idiotopes which can be grouped in terms of localization on the antibody molecule and of variability and which are combined in various ways on a large number of χ chain-bearing antibody species. At least some of the idiotopes recognized by the monoclonal antibodies are under the control of the heavy chain linkage group, as shown by the idiotope analysis of serum antibodies from Igh-congenic mice. In the course of the fusion experiments, also cell lines secreting antibodies of SJL origin with specificity for the constant region of χ1 chains were isolated, and we also describe the preparation of an Fv fragment from B1–8, an IgM antibody.

Preinjection of C57BL/6 mice with nano-to microgram amounts of a monoclonal IgG1 antibody directed against a binding site-related idiotope of the anti-NP [(4-hydroxy-3-nitro-phenyl)acetyl] antibody B1-8 results in enhancement or suppression of the corresponding and of another B1-8 idiotope in a subsequent anti-NP response, depending on the dose of the injected anti-idiotope antibody. The enhancing and suppressive effects appear two weeks after anti-idiotope administration and are maximal after 6-8 weeks. They are predominantly expressed at the level of IgG, not IgM, antibodies. Enhancement of idiotype expression, i.e. idiotypic memory, can also be induced by the injection of idiotypic antibody of the IgM class, namely antibody B1-8. This effect might represent one of the general mechanisms by which immunological memory is established.