Affiliation to mature B cell repertoire and positive selection can be separated in two distinct processes.

Using an 'oligoclonal' model, we have previously shown that mice transgenic for a mu chain (H3) and deficient for kappa chain expression display a mature B cell repertoire largely dominated by the H3/lambda1 pair, while the four H3/lambda available combinations can be observed in the immature B cell compartment. This led us to propose the existence of a positive selection process. To test this hypothesis, we have introduced the SJL lambda locus coding for a defective lambda1 chain (lambda1(s)) that creates a dysfunctional Ig receptor complex during B cell differentiation. Our results show that the lambda1(s) defect impairs the development of mature B cells when the H3-mu transgene insert is present in the hemizygous state. This suggests that the Gly --> Val substitution present in the C(lambda)1(s) chain at position 155 is sufficient to abrogate the selection of the H3/lambda1 pair. Unexpectedly, when the H3-mu transgene array is present in a homozygous state in lambda1(s) mice but not in 'wild-type' lambda1 mice (lambda1(+)), a significant number of mature B cells expressing all H3/lambda combinations can be developed. These results indicate that the overriding H3/lambda1 dominance observed in lambda1(+) mice is due to a positive selection process and not to a negative selection of other H3/lambda combinations. They also show that the export of B cells to the periphery can be controlled by the expression of the mu chain.

Tolerance to maternal immunoglobulins: resilience of the specific T cell repertoire in spite of long-lasting perturbations.

T cell tolerance is established and maintained through various mechanisms, the critical component being the persistence of the specific Ag. However, at the molecular level, the nature of the recovering TCR repertoire following breakdown of tolerance is unknown. We address this important question by following kappa light chain constant region (C kappa)-specific CD4+ T cells of kappa light chain knock-out (kappa-/-) mice born to kappa+/- mothers. These cells, which were in contact with maternal kappa+ Igs from early ontogeny until weaning, were strongly tolerized. Tolerance was reversible and waned with the disappearance of peptide C kappa 134-148 presentation in lymphoid organs, including the thymus. Whereas three specific V beta-J beta rearrangements emerged in the peptide C kappa 134-148-specific CD4+ T cell response of all regular kappa-/- mice, soon after breakdown of tolerance only one of these rearrangements was detected. The two others displayed a significant delay in reappearance and were still rare at 26 wk of age, while the control proliferative response had already recovered 3 mo earlier. At 52 wk of age, a complete recovery of the three canonical V beta-J beta rearrangements was observed. Thus, although profoundly perturbed for several months, the T cell repertoire returns to equilibrium, highlighting the resilient nature of this system.

Role of maternal Ig in the induction of C kappa-specific CD8+ T cell tolerance.

Although the influence of maternal Ig on the B cell repertoire and subsequent Ab response has been extensively studied, much less attention has been devoted to their effects on T cell responses of the offspring. To address this question, we have studied the influence of maternal kappa-positive Ig (Ig kappa) on the C kappa-specific CD8+ T cell response of kappa knock-out (kappa-/-) pups resulting from various crosses and foster nursings. These systems allowed control of physiologic transmission of Ig kappa at defined periods of ontogeny. Our data show that conventional transfer of maternal Ig via the placenta plus colostrum/milk or adoptive transfer via only the colostrum/milk were the most efficient at tolerizing C kappa-specific CD8+ responses. Surprisingly, tolerance was not detected in kappa-/- pups born to kappa+/- females obtained by cesarean delivery and suckled by kappa-/- mothers (transplacental supply only). Tolerance, which was strong until 5 wk of age, was reversible and waned with the decrease of Ig kappa serum concentration. Depletion of CD4+ T cells at the time of C kappa peptide immunization abolished the tolerance of C kappa-specific CD8+ T cells. These data suggest that an oral supply of Ig is very efficient at inducing and maintaining tolerance of C kappa-specific CD8+ T cells, at least for several weeks after birth, and that suppression rather than deletion is responsible for this tolerance. In addition, they strengthen the view that tolerance of CD8+ T cells to a soluble Ag is never permanently acquired even if it is present in large quantities during ontogeny.

Does positive selection determine the B cell repertoire?

To know whether each newly formed B cell has an equal chance of survival in the organism, we analyzed the composition of the B cell repertoire of extremely limited diversity by generating mu-transgenic kappa-knockout mice. Surprisingly, in both types of mice studied, the B cell repertoire is mainly composed of cells expressing the mu-transgene-encoded chain associated with only one out four available lambda types depending on the mu transgene. Moreover, B cell differentiation cultures in vitro show that newly formed B cells can express the various lambda types regardless of the presence or absence of the mu transgenes. These results show a drastic impact of the heavy chain on the lambda light chain repertoire expressed in the periphery. The overexpression of a unique heavy/light chain pairing therefore results from selective processes. The immature B cells may be positively selected to provide the immunocompetent B cells in the periphery.

Emergence in C kappa knockout mice of a diverse cytotoxic T lymphocyte repertoire that recognizes a single peptide from the immunoglobulin constant kappa light chain region.

Allotype- or idiotype-specific CD4+ T cells have been reported to recognize immunoglobulin (Ig) peptides presented by class II molecules. In contrast, few data are available concerning the generation of Ig peptide-specific CD8+ T cells. We have therefore investigated whether T-depleted spleen cells from Ig kappa light chain-expressing 129/Sv mice (129 kappa +/+) could induce, in C kappa knockout mice (129 kappa -/-), the generation of Ig constant kappa light chain region (C kappa)-specific cytotoxic T lymphocytes (CTL). The determination of TCR beta chain expressed by nine CTL clones, together with the use of a library of overlapping peptides spanning the whole C kappa sequence, show that the B cells from kappa +/+ mice are able to elicit in C kappa knockout mice, the emergence of a diverse CTL repertoire that recognizes one single C kappa peptide presented by the H-2Kb class I molecule. In addition, these data support the notion that B cells are able to process and present on their class I molecules, peptides generated from their own kappa light chains.

Various V-J rearrangement efficiencies shape the mouse lambda B cell repertoire.

The diversity of the B cell repertoire of C kappa knockout mice is limited by the expression of four lambda light chain types. Among the spleen B cells, lambda 1 is expressed by the majority (58%) of cells, and lambda 3 by the minority (8%), while lambda 2 (V2) and lambda 2 (Vx) are expressed in intermediate quantities (18% and 16%, respectively). To assess the influence of mechanistic pressures on the lambda subtype distribution, the proportions of the different lambda rearrangements were determined in various B cell subpopulations divided on the basis of the lambda subtype expressed, and the V lambda J lambda junction sequences were studied at different steps of B cell differentiation (pre-B, immature and mature B cells). The data show that (1) the ratio of productive/non-productive VJ junctions is determined by the nature of the lambda segments that are rearranged as can be observed in the pre-B cells, (2) V1-J1 non-productive rearrangements are often found in the lambda 1-negative B cells in the periphery, and (3) V1J3 junctions are often non-productive regardless of the nature of the cells analyzed. Our results, therefore, suggest that a strong probability of initiating a V1-J1 rearrangement and a weak probability of giving a productive V1J3 junction are responsible for the lambda 1 dominance and the lambda 3 under-expression, respectively. The intermediate proportion of lambda 2(V2) subtype is most likely due to a probability of obtaining a productive joint that is better than that for V1J3 and a probability of initiating a rearrangement that is lower than that for V1J1. However, the lambda 2(Vx) cell proportion cannot be determined only by these parameters.

Conserved distribution of lambda subtypes from rearranged gene segments to immunoglobulin synthesis in the mouse B cell repertoire.

The immunoglobulin lambda light chain system displays a limited diversity in inbred mice. Indeed, the lambda locus is organized in two recombination units: V lambda 2-V lambda x-J lambda 2-C lambda 2-psi J lambda 4-psi C lambda 4, which can produce either lambda 2(V2) or lambda 2(Vx) chains; and V lambda 1-J lambda 3-C lambda 3-J lambda 1-C lambda 1, which can produce either lambda 1 or lambda 3 chains. Each of these units is associated with an enhancer, E lambda 2-4 or E lambda 1-3, at the 3' side. The expression of each lambda chain is, therefore, controlled by distinct promoter and/or enhancer regions. To clarify the basis of these controls, we measured, by quantitative polymerase chain reaction, the proportions of each lambda subtype in BALB/c spleen mRNA and among genomic rearrangements. It appears that these distributions are similar to and consistent with the relative cellular frequencies in the spleen, as evaluated by flow cytometry. These results suggest that, in resting cells, the transcription rates are identical, regardless of the lambda subtype. After lipopolysaccharide (LPS) stimulation, the transcription rates per cell remain similar for all lambda subtypes despite different regulatory sequences. To detect eventual post-transcriptional regulations, we estimated the lambda light chain distribution in IgM secreted by LPS-stimulated B cells and in serum IgG. These distributions are still similar to those of lambda-expressing cells, lambda mRNA or genomic rearrangements. We conclude that the lambda subtype distribution is conserved from productive V-J rearranged genes to secreted lambda immunoglobulins, despite different regulatory sequences.

Compartmentalization of lambda subtype expression in the B cell repertoire of mice with a disrupted or normal C kappa gene segment.

The establishment of the B cell repertoire depends on two major parameters. The first is determined by mechanistic processes that give rise to a great diversity of B cell receptors from a combination of multiple gene segments. The second is dominated by selective processes that recruit B cell clones via their immunoglobulin receptors. To assess the impact of these parameters on the composition of B cell repertoire, we constructed a mouse model displaying a B cell repertoire limited in its diversity. To this end, we disrupted the C kappa segment by gene targeting. B cells from such mutant mice do not express the kappa light chain. Their light chain repertoire is therefore limited by the expression of only four main lambda light chains: lambda 1, lambda 2(V2), lambda 2(Vx) and lambda 3. In this study we described the proportions of each lambda subtype in various lymphoid compartments. Our results show that the lambda 1 subtype is dominant in the spleen and the bone marrow. Moreover, lambda 1 prevalence is independent of the wild or mutant C kappa genotype. These results suggest that the mechanistic processes are mainly responsible for the bias in lambda subtype expression. On the other hand, the lambda 2(V2) and/or lambda 3 subtypes are expressed at higher levels in the peritoneal cavity. Their prevalence is again observed regardless of the C kappa genotype and seems to be due to B1 cells. These results suggest that different mechanistic processes could control lambda subtype expression in B1 and B2 cell lineages.

Mechanistic and selective constraints act on the establishment of V lambda J lambda junctions in the B cell repertoire.

Only four different subtypes of lambda Ig chains have been described in the mouse: lambda 1, lambda 2(V2), lambda 2(Vx), and lambda 3. These chains are encoded by gene segments all sequenced and well localized in chromosome 16. Although the lambda Ig system is both simple and well characterized, no exhaustive analysis has been done concerning V lambda J lambda junctions in nonintentionally stimulated B cells. To get an insight into the lambda B cell repertoire, we analyzed a large number of V lambda J lambda rearrangements isolated from spleen mRNA or genomic DNA of unimmunized adult BALB/c mice. By PCR amplification, more than 160 clones were obtained covering all V lambda J lambda recombinations. Simple recombinations of trimmed gene segments explain most sequences. Certain junctions have been found to be prevalent in each subtype, and an analysis of V lambda J lambda recombination sites shows that the splenic lambda repertoire can result from both differential efficiencies of rearrangement and selective processes.

Available lambda B cell repertoire in the mouse: evidence of positive selection by environmental factors.

We have recently shown that, from two BALB/c mice treated with rabbit anti-C lambda 2/C lambda 3 antibodies coupled to lipopolysaccharide, variable heavy chain (VH) family repertoires associated with lambda 2 or lambda 3 light chains can differ from one lambda subtype to another and from one individual mouse to another. Indeed, 4 out of 6 lambda 2 (VxJ2) hybridomas from one mouse preferentially expressed the VH10 family while 3 out of 8 lambda 2 (V2J2) and 5 out of 8 lambda 2 (VxJ2) hybridomas from a second mouse preferentially expressed the S107 and VGAM3.8 VH families, respectively. In this report, we describe the structural basis of such preferential pairings by sequence analysis of the 12 lambda 2 hybridomas. The sequence comparison of their VH regions show that each preferential association of a VH family to one V lambda region is restricted to the use of a single member or very closely related members inside a VH family and that a great variability of CDR3 of heavy chain is observed. We, therefore, suggest that environmental factors can modify the available lambda B cell repertoire through a positive selection of particular VH/V lambda pairings. Moreover, our data support that this selection does not require clonal expansion and punctual somatic mutation.

Preferential VH/V lambda pairings occur in the available B cell repertoire of adult BALB/c mice.

To gain insights into the composition of the B cell repertoire, we have investigated VH gene family expression associated with individual light chains. For this purpose, we have examined the use of 12 VH gene families in a large collection of hybridomas expressing one of the four lambda light chains [lambda 1 (V1J1), lambda 2 (V2J2 and V x J2) and lambda 3 (V1J3)]. Our results show that the distribution of the VH families is very different from one lambda subtype to another. This suggests that a few substitutions between VL regions are sufficient to generate very different associated repertoires by strong selection mechanisms. Moreover, we assume that the global VH expression pattern is not random but rather composed of many preferential VH/VL associations.

Ontogeny of the response of mouse thymocytes to interleukin 1 and interleukin 2.

In the present report we demonstrate that the in vitro proliferative response of the newborn thymocytes to interleukin (IL) 1 and IL 2, which is remarkably stronger than the adult thymocyte response, is associated with a considerable increase of CD4-CD8- cells expressing a gamma/delta T cell receptor (TcR). By polymerase chain reaction analysis we show that the V gamma gene segment usage in the adult and newborn responding cells reflects the developmentally regulated expression of the V gamma gene segments, suggesting that the increase in TcR gamma/delta+ cells results from the polyclonal expansion of pre-existing clones. Surprisingly, although the fetal thymocyte populations contain higher numbers of TcR gamma/delta+ cells than the adult and newborn ones, the highest proliferative response to IL 1 and IL 2 is obtained with the newborn thymocytes. Non mutually exclusive hypotheses are discussed to explain these results.

Cellular basis of the resistance of newborn mice to the pathogenic effects of anti-CD3 treatment.

We have analysed the mechanisms underlying the differences in the susceptibilities of adult and newborn mice to the pathogenic effects of anti-CD3 mAbs. Our data show that the thymus cell number in adults is reduced by 93% 48 h after one single injection of 5 mg/kg of Ab whereas the same dose in newborns induces only a 30% decrease. In the adult, this effect is associated with a marked depletion of CD4+ CD8+ double positive (DP) cells and with the appearance of important areas of cell necrosis in the thymic cortex. In newborns, the DP cells are less affected and the thymic cortex does not present any cell necrosis even after an injection of 45 mg/kg of mAbs. Pre-treatment of adults with anti-CD4 and anti-CD8 Abs, while completely abolishing the toxic side-effects induced by anti-CD3 mAbs, does not protect the thymus from the depletion of DP cells. In vitro, anti-CD3 mAbs induce the proliferation of thymocytes and spleen cells from adults but not from newborns. Tumour necrosis factor-alpha (TNF alpha) is found in the serum of adults 90 min after injection of anti-CD3 but is never detected in the serum of anti-CD3 treated newborns. Taken together our data support the view that anti-CD3 mAbs act by two different mechanisms. The first one results from the binding of anti-CD3 on the CD3+ thymocytes which induces a direct toxicity for only the CD4+ CD8+ cells.(ABSTRACT TRUNCATED AT 250 WORDS)

Junctional diversity of H and L chains allows the coexpression of two mutually exclusive idiotopes (IdI104 and IdI558).

The molecular basis for the unexpected coexpression of the individual Id (IdI)558 and IdI104 Id by anti-alpha(1-3) DEX antibody (Ab) (126.33 and 414.2) derived from the MPW wild mouse strain has been investigated by the comparison of the structures of their VH and V lambda 1 chain regions with those of two other MPW-derived Ab (262.9 and 16.3) expressing either IdI558 or IdI104 Id. Our data show that 262.9 and 16.3 Ab display identical V lambda 1 and very similar VH regions when compared with BALB/c anti-alpha (1-3) dextran Ab expressing IdI104 or IdI558, respectively. The two Ab (414.2 and 126.33) that express both IdI104 and IdI558 Id display two main features. First, their VH CDR3 are different from those found in IdI104 or IdI558 expressing anti-alpha(1-3) dextran Ab. Second, their V lambda 1 are identical to those from BALB/c origin except for the presence of an additional residue, a phenylalanine at position 95A of CDR3. This additional residue is encoded by the V lambda 1 gene segment and results from a hitherto undescribed V lambda 1-J lambda 1 junction. The alteration of the length of the V lambda 1 CDR3 loop, in conjunction with particular residues within VH CDR3, allows the coexpression of two Id that were found to be mutually exclusive in laboratory mice.

Effects of neonatal injections of anti-CD3 epsilon monoclonal antibodies on the T cell functions of adult mice.

The effect of neonatal injections of anti-CD3 mAb on the subsequent immune responsiveness of adult mice has been investigated. The data indicate that neonatal treatment with three injections of 50 microliters of anti-CD3 epsilon ascitic fluid induce a profound depletion of T cells in the peripheral lymphoid organs but do not modify the absolute number of thymocytes. This treatment completely abolishes the T cell functions of mice at least 3 wk after the last injection of mAb. In the thymus, this suppression is associated with a decrease in the number of TcR/CD3 molecules in the dull CD3+ cells and with a drastic reduction the bright CD3+ population. However, the amount and the size of the TcR alpha,beta and CD3 epsilon mRNA transcripts are not modified, suggesting that the down-regulation of the TcR/CD3 complex induced by anti-CD3 mAb does not exert a feedback inhibition on the transcription of TcR genes. The suppression induced by neonatal injections of anti-CD3 epsilon mAb is reversible and the induction of immune responses requires the reappearance of a minimal number of bright CD3+ cells. However, this suppression can be maintained without side effects for several months provided that anti-CD3 mAb were administered at 7-day intervals from birth. This injection schedule should allow the study of the effect of anti-CD3 antibodies during the T cell ontogeny on the establishment of the B and T cell repertoires.

Mapping of Igk-V genes using backcrossed laboratory and wild mice.

In the mouse, the genes coding for the Ly-2 antigen, the beta chain of the T-cell receptor, and the immunoglobulin kappa light chain have been located on chromosome 6. Although a tentative order has been proposed for these genes, very few data have been reported concerning their genetic distance. To address this question, we have produced backcross mice between SJL and MAI (a wild-derived strain belonging to the Mus musculus), since these mice segregate for the Ly-2 and Igk-C proteins and for the Igk-V24, Igk-V21, Igk-V10, Igk-V8, and Igk-V4 genes. Twelve recombinants were obtained from 163 backcross mice studied. Two mice showed a recombination between the (Igk-V24, Igk-V10, Igk-V8, Igk-V4) and the (Ly-2, Igk-C, Igk-V21) groups, and ten mice displayed a recombination between the (Igk-V24, Igk-V10, Igk-V8, Igk-V4) group and the Tcrb-C loci. These data imply the following gene order: Tcrb-C .... (Igk-V24, Igk-V10, Igk-V8, Igk-V4) .... (Igk-V21, Igk-C, Ly-2). They indicate a distance of 6.1 cM between Tcrb-C and (Igk-V24, Igk-V10, Igk-V8, Igk-V4) and 1.2 cM between Igk-V24, Igk-V10, Igk-V8, Igk-V4 and the (Igk-V21, Igk-C, Ly-2) groups.

Idiotypic characteristics of lambda 1-bearing anti-alpha(1----3)dextran response in wild mice: high frequency of responder strains and co-expression of IdI 558 and IdI 104 on the same molecule.

In laboratory mouse strains, only mice with the IghCa phenotype are able to mount an immune response against the alpha(1----3) glycosidic linkage of the B1355 dextran. The response is highly restricted with respect to utilization of only V lambda 1 and few related members of the VH558 family, and characterized by the production of cross-reacting (IdX), and individual (IdI) 104 and 558 idiotopes. We report in this study data showing a large degree of polymorphism in the anti-alpha(1----3)Dex response among 35 wild-derived mouse strains belonging to 4 different species. In contrast to the anti-alpha(1----3)Dex response of common laboratory strains, neither this response nor expression of the IdX, and IdI 104 and 558 idiotopes are linked to the IghCa haplotype. We also demonstrate co-expression on the same molecule of IdI 104 and IdI 558 idiotopes in wild mouse-derived hybridomas. These data challenge the idea of the exclusive expression at the molecular level of IdI 104 or IdI 558.

Highly preferential VH-VL pairing in normal B cells results in antigen-independent selection of the available repertoire.

The aim of this work was to analyze the probabilities of combined expression of the various polymorphic forms of two well-defined VH and VK segments. The availability of two monoclonal antibodies specific, respectively, for the VHT15 and VK21 D-E gene products allowed us to study, both in the selected and in the preimmune repertoire, the expression of VHT15-VK21 D-E pairs in several mouse strains. Our data establish, first, that even before antigen encounters B cells, clones utilize a highly biased repertoire of VH-VL combinations and, secondly, that the level of productive VH-VL pairing depends primarily on the "allelic" form of the relevant VH and VK genes. Furthermore, clonal analysis revealed that this asymmetry in Ig gene expression affects a large proportion of newly arising B cells. Together, these studies demand reconsideration of the current estimates of the available antibody repertoire size, and provide new insights for our understanding of the phenomenon of clonal dominance.

Paucity of phosphorylcholine-specific clones in B cells expressing the VHT15 gene product.

The aim of this work was to study the cellular basis of the phenomenon of clonal dominance. To this end we analyzed two collections of BALB/c and C.B20 hybridomas that we selected on the basis of the expression of the VHT15 gene product independently from their antigen specificity. Our study demonstrates that none of the 28 BALB/c and only 2 of the 29 C.B20 hybridomas obtained have variable regions that bind PC. We conclude therefore that the domination of the immune response to PC by particular variable regions cannot be due to the establishment of clonal dominance prior to immunization.

Differential expression of sets of highly homologous variable region gene products in selected and preimmune repertoires of inbred mouse strains.

Using mAb that selectively recognize the various allelic forms of the VHT15 and Vk21D-E genes' products, we analyzed the influence of VH and Vk polymorphism on the probability of expression of these gene segments. Our data show that the frequency to which the VHT15 gene product becomes available in the preimmune repertoire is strongly influenced by the polymorphism of the relevant structural gene, suggesting therefore that VH genes cannot be randomly used in the various strains. Contrary to this, the frequency of Vk21D-E+ clones is similar in all mouse strains tested, and in all cases is higher than the frequency of VHT15 clones. This observation strongly suggests that Vk genes can be randomly expressed, and/or that their number is lower than that of their VH counterpart. Finally, analysis of the specificity associated to the expression of the VHT15 segment revealed that VH polymorphism strongly influences not only the probability of expression of each V gene, but also the specificity of the antibodies on which these VH genes are used.