Multiple differences between the nucleic acid sequences of the IgG2aa and IgG2ab alleles of the mouse

DNA-Based Delivery of Checkpoint Inhibitors in Muscle and Tumor Enables Long-Term Responses with Distinct Exposure

Checkpoint-inhibiting antibodies elicit impressive clinical responses, but still face several issues. The current study evaluated whether DNA-based delivery can broaden the application of checkpoint inhibitors, specifically by pursuing cost-efficient in vivo production, facilitating combination therapies, and exploring administration routes that lower immune-related toxicity risks. We therefore optimized plasmid-encoded anti-CTLA-4 and anti-PD-1 antibodies, and studied their pharmacokinetics and pharmacodynamics when delivered alone and in combination via intramuscular or intratumoral electroporation in mice. Intramuscular electrotransfer of these DNA-based antibodies induced complete regressions in a subcutaneous MC38 tumor model, with plasma concentrations up to 4 and 14 μg/mL for anti-CTLA-4 and anti-PD-1 antibodies, respectively, and antibody detection for at least 6 months. Intratumoral antibody gene electrotransfer gave similar anti-tumor responses as the intramuscular approach. Antibody plasma levels, however, were up to 70-fold lower and substantially more transient, potentially improving biosafety of the expressed checkpoint inhibitors. Intratumoral delivery also generated a systemic anti-tumor response, illustrated by moderate abscopal effects and prolonged protection of cured mice against a tumor rechallenge. In conclusion, intramuscular and intratumoral DNA-based delivery of checkpoint inhibitors both enabled long-term anti-tumor responses despite distinct systemic antibody exposure, highlighting the potential of the tumor as delivery site for DNA-based therapeutics.

Expression of IgG Monoclonals with Engineered Immune Effector Functions

The therapeutic development of monoclonal antibodies requires robust and reliable methods for their recombinant expression and characterization. In this context, an increasingly important aspect in the antibody development process is to determine the contribution of Fc-mediated immune effector functions to therapeutic activity. Here we describe steps for the cloning and mammalian expression of mouse and human IgG monoclonals with reduced immune effector functions, based on mutation of Fc-gamma receptor and complement-binding sites. The resulting antibody preparations contain low levels of endotoxin and are suitable for testing in animal models of disease.

FURTHER SIMULATION STUDIES ON EVOLUTION BY GENE DUPLICATION

In order to understand the origin of multigene families, Monte Carlo simulations were performed to see how a genetic system evolves under unequal crossing-over, mutation, random genetic drift and natural selection, starting from a single gene copy. Both haploid and diploid models were examined. Beneficial, neutral, and detrimental mutations were incorporated, and "positive" selection favors those chromosomes (haploid) or individuals (diploid) with more beneficial mutations than others. The same model for haploids was previously investigated with special reference to the evolution of gene organization, and the ratio of the numbers of beneficial genes to pseudogenes was found to be a rough indicator of the relative strengths of positive and negative (against deleterious alleles) natural selection (Ohta, 1987b). In the present paper, the evolution of gene organization and of sequence divergence among genes in the multigene family is examined. It is shown that positive selection accelerates the accumulation of arrays containing different beneficial mutations, but that total divergence including both neutral and beneficial mutations is not very sensitive to positive selection, under this model. The proportion of beneficial mutations in the total mutations accumulated is a better indicator of positive selection than is the total divergence. It is pointed out that various observed examples in which amino-acid substitutions are accelerated, as compared with synonymous substitutions in duplicated genes (Li, 1985), may reflect the effect of selection similar to the present scheme. The diploid model is shown to be more efficient for accumulating beneficial mutations in duplicated genes than the haploid one, and the relevance of this finding to the advantage of sexual reproduction is discussed.

Murinization and H Chain Isotype Matching of the Anti-GITR Antibody DTA-1 Reduces Immunogenicity-Mediated Anaphylaxis in C57BL/6 Mice

Recent advances in immuno-oncology have shown that the immune system can be activated to induce long-term, durable antitumor responses. For immuno-oncology drug development, immune activation is often explored using rat Abs in immunocompetent mouse models. Although these models can be used to show efficacy, antidrug immune responses to experimental protein-based therapeutics can arise. Immunogenicity of surrogate Abs may therefore represent an important obstacle to the evaluation of the antitumor efficacy of immunomodulator Abs in syngeneic models. A recent publication has shown that anti-glucocorticoid-induced TNFR family-related protein agonistic Ab DTA-1 (rat or murinized IgG2a) can induce the development of anaphylaxis in C57BL/6 mice upon repeated i.p. dosing because of an anti-idiotypic anti-drug Ab immune response. This study was undertaken to address the impact of the immunogenicity derived from the Fc and variable domains. To this end, chimerized (rat V domains/mouse constant regions) and murinized (95% mouse sequence) DTA-1-based surrogate Abs with a murine IgG2c H chain isotype were created. Chimerization and murinization of DTA-1 did not affect receptor binding and glucocorticoid-induced TNFR family-related protein-induced T cell agonistic properties. Similar in vivo antitumor efficacy and intratumoral CD8⁺/regulatory T cells were also observed. Finally, treatment of C57BL/6 mice with the chimerized and murinized DTA-1 Abs on a C57BL/6-matched IgG2c isotype resulted in reduced development and severity of anaphylaxis as measured by decline of body temperature, behavioral effects, serum IL-4, IgE, and anti-drug Ab levels. These results suggest that careful murinization and selection of a strain-matched H chain isotype are critical to generate ideal surrogate Abs for testing immuno-oncology mechanisms in vivo.

Isotype-specific glycosylation analysis of mouse IgG by LC-MS

With mice being the top model organism in immunology and with Fc glycosylation being increasingly recognized as important modulator of antibody function, the time has come to take a look at the glycosylation of mouse IgG isotypes. Tryptic glycopeptides of mouse IgG1, IgG2, and IgG3 differ in mass and so these three isoforms can be easily discriminated by MS. Commercial IgG contained a rare IgG1 variant but no IgG3, which, however, was found in sera of C57BL/6 and BALB/c mice. These strains deviated with regard to IgG2a and IgG2b alleles. The Ig2a B allele was not observed in any of the four samples investigated. All a/c isotypes contain the same glycopeptide sequence, which deviates from that of IgG2b by containing Leu instead of Ile. The Leu/Ile glycopeptide variants were separated by RP chromatography and the order of elution was determined. The major glycoforms on all isotypes were fucosylated with no and one galactose (GnGnF and GnAF) followed by fully galactosylated AAF and smaller amounts of mono- and disialylated N-glycans. In the commercial serum pool, the relative ratios of glycans differed between isotypes. Sialic acid exclusively occurred as N-glycolylneuraminic acid. Fucosylation was essentially complete. No bisected and no α1,3-galactosylated glycans were found.

A poorly neutralizing IgG2a/c response elicited by a DNA vaccine protects mice against Japanese encephalitis virus

We demonstrated previously that immunization with a DNA vaccine expressing the Japanese encephalitis virus (JEV) envelope (E) protein conferred a high level of protection through a poorly neutralizing antibody response. Here, we further investigated the role of the IgG subclass in this antibody-dependent protection using cytokine co-immunization and cytokine-deficient mice. A significant difference in IgG2a/c but not IgG1 was observed between mice that survived or died following a lethal challenge. Correspondingly, the IgG2a/c response and protection increased in IL-4-deficient mice but decreased in IFN-γ-deficient mice, highlighting the importance of IgG2a/c. In addition, the restoration of protection and E-specific IgG2a/c production in IFN-γ-deficient mice by a T helper (Th) type 1-biased intramuscular immunization suggested that IgG2a/c but not IFN-γ was the major component for protection. The failure of protection against a direct intracranial challenge indicated that IgG2a/c-mediated protection was restricted to outside the central nervous system. Consistent with this conclusion, passive transfer of E-specific antisera conferred protection only pre-exposure to JEV. Therefore, our data provided evidence that the IgG subclass plays an important role in protection against JEV, particular in poorly neutralizing E-specific antibodies, and Th1-biased IgG2a/c confers better protection than Th2-biased IgG1 against JEV.

Anti-CTLA-4 antibodies of IgG2a isotype enhance antitumor activity through reduction of intratumoral regulatory T cells

Antitumor activity of CTLA-4 antibody blockade is thought to be mediated by interfering with the negative regulation of T-effector cell (Teff) function resulting from CTLA-4 engagement by B7-ligands. In addition, a role for CTLA-4 on regulatory T cells (Treg), wherein CTLA-4 loss or inhibition results in reduced Treg function, may also contribute to antitumor responses by anti-CTLA-4 treatment. We have examined the role of the immunoglobulin constant region on the antitumor activity of anti-CTLA-4 to analyze in greater detail the mechanism of action of anti-CTLA-4 antibodies. Anti-CTLA-4 antibody containing the murine immunoglobulin G (IgG)2a constant region exhibits enhanced antitumor activity in subcutaneous established MC38 and CT26 colon adenocarcinoma tumor models compared with anti-CTLA-4 containing the IgG2b constant region. Interestingly, anti-CTLA-4 antibodies containing mouse IgG1 or a mutated mouse IgG1-D265A, which eliminates binding to all Fcγ receptors (FcγR), do not show antitumor activity in these models. Assessment of Teff and Treg populations at the tumor and in the periphery showed that anti-CTLA-4-IgG2a mediated a rapid and dramatic reduction of Tregs at the tumor site, whereas treatment with each of the isotypes expanded Tregs in the periphery. Expansion of CD8(+) Teffs is observed with both the IgG2a and IgG2b anti-CTLA-4 isotypes, resulting in a superior Teff to Treg ratio for the IgG2a isotype. These data suggest that anti-CTLA-4 promotes antitumor activity by a selective reduction of intratumoral Tregs along with concomitant activation of Teffs.

Immunoglobulin locus associates with serum IgG levels and albuminuria

The interaction between IgG and Fc-γ receptors in glomeruli contributes to the development of several types of proteinuric glomerular disease, but the involvement of immunological mechanisms in hypertensive renal injury is incompletely understood. Here, we investigated serum IgG levels in SHR-A3 rats, which develop hypertensive injury, and compared them with the injury-resistant SHR-B2 line. At 18 weeks old, SHR-A3 rats had serum total IgG levels nearly twice those of SHR-B2 rats, although subclass IgG2b was undetectable in SHR-A3 rats compared with mean levels (± SEM) of 80.7 ± 12.8 mg/dl (18 weeks) and 116.6 ± 19.0 mg/dl (30 weeks) in SHR-B2 rats. In addition, these two strains had significantly different serum levels of IgG1, IgG2a, and IgG2c; differences persisted at 30 weeks for all subclasses except IgG2a. Genetic mapping revealed that a locus on chromosome 6 linked to IgG subclass levels that affected IgG1, IgG2b, and IgG2c but not IgG2a. The mapped haplotype block contains IgH, suggesting regulation of three of four serum IgG subclass levels in cis. Resequencing revealed variation in the sequence of the Fc portion of the IgG heavy chain, which predicts important functional changes. To examine whether there is any relationship between this haplotype block and susceptibility to renal injury, we examined the effect of SHR-A3 and SHR-B2 alleles at this block on albumin excretion in an F2 intercross. Albuminuria doubled with inheritance of SHR-A3 alleles. In summary, allelic variation in IgH or nearby genes may modulate the susceptibility to hypertensive renal injury in SHR-A3 rats.

Protective and pathogenic roles for B cells during systemic autoimmunity in NZB/W F1 mice

Delineating the relative contributions of B lymphocytes during the course of autoimmune disease has been difficult. Therefore, the effects of depleting all mature B cells using a potent CD20 mAb, or of depleting circulating and marginal zone B cells using a ligand-blocking CD22 mAb, were compared in NZB/W F(1) mice, a model for human systemic lupus erythematosus. Single low-dose mAb treatments depleted B cells efficiently in both NZB/W F(1) and C57BL/6 mice. Prophylactic B cell depletion by repeated CD20 mAb treatments prolonged survival during pristane-accelerated lupus in NZB/W F(1) mice, whereas CD22 mAb had little effect. Despite effective B cell depletion, neither mAb treatment prevented autoantibody generation. In addition, CD20, CD22, and control mAb-treated NZB/W F(1) mice developed anti-mouse IgG autoantibodies in contrast to parental NZB and NZW strains, which may have reduced the effectiveness of B cell depletion. Despite this, low-dose CD20 mAb treatment initiated in 12-28-wk-old mice, and administered every 4 wk thereafter, significantly delayed spontaneous disease in NZB/W F(1) mice. By contrast, B cell depletion initiated in 4-wk-old mice hastened disease onset, which paralleled depletion of the IL-10-producing regulatory B cell subset called B10 cells. B10 cells were phenotypically similar in NZB/W F(1) and C57BL/6 mice, but were expanded significantly in young NZB/W F(1) mice. Thus, B cell depletion had significant effects on NZB/W F(1) mouse survival that were dependent on the timing of treatment initiation. Therefore, distinct B cell populations can have opposing protective and pathogenic roles during lupus progression.

Induced Expression of Murine γ2a by CD40 Ligation Independently of IFN-γ

IgG2a, with gamma2a H chains, is important for protection against viruses and other intracellular pathogens. Although a large portion of IgG2a expression is dependent upon IFN-gamma, some germline transcription and switch recombination to the murine gamma2a H chain gene expression are independent of IFN-gamma. We found that agonistic anti-CD40 Abs injected into IFN-gamma-deficient mice induce a > 200-fold increase in the amount of serum Ig2a, while other Ig isotypes are increased by 16-fold or less. In vitro, ligation of CD40 on B cells, without the addition of other B cell activators or cytokines, results in germline transcription and switch recombination that are largely restricted to the gamma2a gene. These results suggest that some immune responses to infectious agents can result in large amounts of IgG2a expression through ligation of CD40, without the expression of IFN-gamma by Th1 or other cells.

CD28 costimulation is required for the expression of T-cell-dependent cell-mediated immunity against blood-stage Plasmodium chabaudi malaria parasites

Mice suppress the parasitemia of acute blood-stage Plasmodium chabaudi malaria by an antibody- or T-cell-dependent cell-mediated mechanism of immunity (AMI and CMI, respectively) or by both mechanisms. To determine whether CD28 costimulation is required for expression of these polar immune responses, we first compared the time courses of P. chabaudi malaria in CD28-deficient (CD28(-/-)) and CD28-intact (CD28(+/+)) mice. Acute infections in both knockout (KO) and control mice followed similar time courses, with the period of descending parasitemia being prolonged approximately 2 weeks in KO mice followed by intermittent low-grade chronic parasitemia. Infected CD28(-/-) mice produced primarily the immunoglobulin M antibody, which upon passive transfer provided partial protection against P. chabaudi challenge, suggesting that the elimination of blood-stage parasites by CD28(-/-) mice was achieved by AMI. To determine whether CD28(-/-) costimulation is required for the expression of CMI against the parasite, we compared the time courses of parasitemia in B-cell-deficient double-KO (J(H)(-/-) x CD28(-/-)) mice and control (J(H)(-/-) x CD28(+/+)) mice. Whereas control mice suppressed parasitemia to subpatent levels within approximately 2 weeks postinoculation, double-KO mice developed high levels of parasitemia of long-lasting duration. Although not required for the suppression of acute P. chabaudi parasitemia by AMI, CD28 costimulation is essential for the elimination of blood-stage parasites by CMI.

Immunity to blood-stage murine malarial parasites is MHC class II dependent

To determine whether MHC class II antigen presentation is essential for the induction of protective immunity against blood-stage malarial parasites, we used gene-targeted knockout (KO) mice to follow the time-course of nonlethal Plasmodium yoelii and Plasmodium chabaudi infections in two models of MHC class II deficiency. Infection of MHC class II KO (A(-/-)) mice with either parasite species resulted in an unremitting hyperparasitemia, whereas MHC-intact control mice resolved their parasitemia. In contrast, invariant chain KO (Ii(-/-)) mice, which present antigen via recycled but not nascent MHC class II molecules, eventually cured their infections when infected with P. yoelii. P. chabaudi parasitemia declined to subpatent levels in most Ii(-/-) mice but then recrudesced. Immunity to blood-stage malaria may be achieved by cell-mediated and antibody-mediated mechanisms of immunity, as such, the findings in A(-/-) mice indicate an essential role for MHC class II presentation of malarial antigens. Moreover, they suggest that protective immune responses to malarial antigens capable of eliminating blood-stage parasites are T cell dependent and can be induced with antigens processed in early and late endosomes.

Plasmodium chabaudi adami: interferon-gamma but not IL-2 is essential for the expression of cell-mediated immunity against blood-stage parasites in mice

Cell-mediated immunity (CMI) may be important in immunity against blood-stage malaria. Accordingly, we examined the role of type 1 cytokines in the resolution of Plasmodium chabaudi adami malaria in mice genetically modified to have type 1 cytokine gene defects. Parasitemia was prolonged in double knockout (IL-2(-/-), IFNgamma(-/-)) mice compared to control mice. Despite deficiencies in gammadelta T cell and B cell subsets, these mice produced anti-malarial antibodies and eventually cured their infections, possibly by antibody-mediated immunity. However, because acute P. c. adami parasitemia may also be suppressed by CMI, the requirements for IL-2 and IFNgamma were evaluated in mice lacking B cells and functional IL-2 or IFNgamma genes. Acute malaria in J(H)(-/-), IL-2(-/-) mice was prolonged, but eventually cured. In contrast, J(H)(-/-), IFNgamma(-/-) mice developed unremitting parasitemia. These data strongly suggest that IFNgamma, but not IL-2, plays an essential role in the expression of CMI against P. c. adami infections. This finding may prove useful in developing malarial vaccines aimed at inducing CMI.

Characterization of mouse switch variant antibodies by matrix-assisted laser desorption ionization mass spectrometry and electrospray ionization mass spectrometry

The amino acid sequences of mouse monoclonal antibodies have been characterized completely by mass spectrometry. Antibodies used in the present study were derived from mouse switch variant cell lines that produce four kinds of immunoglobulin Gs (IgGs). The amino acid sequences of these antibodies had not been estimated from the corresponding DNA sequence, so the sequences of IgGs derived from other strains were used as references in this study. Intra- and interchain disulfide bonds of the IgGs were reduced and carboxymethylated and the products were subjected to proteolytic digestion. The existence of N-linked oligosaccharides also was taken into account. The capabilities and limitations of matrix-assisted laser desorption ionization-time-of-flight mass spectrometry and capillary liquid chromatography-electrospray ionization mass spectrometry are discussed in the structural characterization of the antibodies. Based on our results, allotypes of the antibodies examined are discussed. This study shows that amino acid sequences of proteins, such as IgG, can be investigated without information about the corresponding DNA sequence if appropriate reference sequences derived from other strains can be used.

The Yaa gene-mediated acceleration of murine lupus: Yaa- T cells from non-autoimmune mice collaborate with Yaa+ B cells to produce lupus autoantibodies in vivo

The BXSB Y chromosome-linked mutant gene, Yaa, promotes autoimmune responses in mice predisposed to a lupus-like autoimmune disease. We have previously shown that a cognate interaction of T cells with B cells expressing the Yaa gene appears to be responsible for the accelerated production of autoantibodies. To investigate whether T cells that provide help for autoantibody production by Yaa+ B cells need to express the Yaa gene, we have made radiation bone marrow chimeras containing two sets of T and B cells from mice with or without the Yaa gene and differing by the Thy-1 and Igh allotypes. We then determined autoantibody production following the selective elimination of T cells of Yaa+ origin by treating mice with allele-specific anti-Thy-1 monoclonal antibody. Our results demonstrated that the selective production of autoantibodies by Yaa+ B cells in Yaa(+)-Yaa- double bone marrow chimeras can be mediated as efficiently by T cells from non-autoimmune mice lacking the Yaa gene as by T cells from autoimmune mice bearing the Yaa gene. This indicates that T cells from non-autoimmune Yaa- mice are capable of providing help for autoimmune responses by collaborating with Yaa+ B cells. These data thus strongly suggest that the Yaa gene defect is not functionally expressed in T cells, but only in B cells, and contrast with parallel experiments in the lpr model, in which defects of the Fas antigen in both T and B cells are crucial for the lpr gene-mediated promotion of autoantibody production.

Collagen-induced arthritis in the BB rat. Prevention of disease by treatment with CTLA-4-Ig

Antigen-specific T cell activation requires two independent signalling events, one mediated through T cell receptor engagement by the antigen-presenting cell-expressed peptide/class II major histocompatibility complex, and the second through the cognate interactions of costimulatory molecules expressed on the T cell and antigen-presenting cell. There is evidence from in vitro and in vivo experimental systems suggesting that the CD28/B7 costimulatory pathway is crucial for induction of maximal T cell proliferation and T helper-B cell collaboration for IgG production. This pathway can be blocked by CTLA-4-Ig, a soluble form of CTLA-4 which binds with high avidity to the CD28 ligands, B7-1 and B7-2. Here, we show that CTLA-4-Ig treatment prevents clinical and histological manifestations of disease in a collagen-induced arthritis model of rheumatoid arthritis in the diabetes resistant BB/Wor rat, when therapy is initiated before immunization with bovine type II collagen (BIIC). Anti-BIIC antibody titers are reduced in CTLA-4-Ig-treated rats compared to diseased control animals. Histologically, joints from CTLA-4-Ig-treated animals show no histological abnormalities, in contrast to control antibody-treated animals, which show complete erosion of the articular cartilage and bone. Despite the efficacy of CTLA-4-Ig in preventing clinical and histological signs of arthritis and reducing antibody responses to BIIC, delayed type hypersensitivity responses to collagen 18 d or more after CTLA-4-Ig treatment ends are similar in CTLA-4-Ig-treated and untreated rats, suggesting that the prolonged disease suppression observed does not result from induction of T cell anergy.

Long-term inhibition of murine experimental autoimmune encephalomyelitis using CTLA-4-Fc supports a key role for CD28 costimulation

T cell activation involves not only recognition of antigen presented by the MHC, but also nonspecific interactions termed "costimulation." The costimulatory molecules B7-1 and B7-2 are ligands on antigen-presenting cells for the CD28 and CTLA-4 receptors on T cells. Previously, a fusion protein consisting of human CTLA-4 linked to human Fc was shown to bind B7-1 and B7-2 with high avidity and to prevent specific T cell activation. Here we investigated the effects of a recombinant fusion protein consisting of the extracellular domain of human CTLA-4 bound to mouse IgG2a Fc (CTLA-4-Fc) upon experimental autoimmune encephalomyelitis, a T cell-mediated disease that serves as a model for multiple sclerosis. CTLA-4-Fc prevented experimental autoimmune encephalomyelitis in 26 of 28 CTLA-4-Fc-treated mice (median maximum score 0), whereas 28 of 30 mice treated with control mouse IgG2a developed disease (median maximum score 2.75). Less inflammation and virtually no demyelination or axonal loss occurred in CTLA-4-Fc-treated compared with control-treated mice. Activated splenocytes from CTLA-4-Fc-treated mice were able to transfer disease adoptively to naive recipients. These results indicate a key role for the B7/CD28 system in the development of actively induced murine experimental autoimmune encephalomyelitis, suggesting an area of investigation with therapeutic potential for multiple sclerosis.

Activation of effector functions by immune complexes of mouse IgG2a with isotype-specific autoantibodies

Analysis of five monoclonal autoantibodies, rheumatoid factors produced by hybridomas generated from spleen cells of BALB/c mice repeatedly infected with A/PR/8/34 human influenza A virus, revealed that they recognized distinct but spatially related epitopes. The differing isoallotypic specificity of the IgM and IgA monoclonal antibodies correlated with the presence of Ile258 and Ala305, respectively. Although these data suggest that the epitopes recognized are within the CH2 domain, all antibodies failed to inhibit IgG antigen reactivity with Staphylococcus aureus protein A (SpA), C1q, mouse C3, human Fc gamma RI or mouse Fc gamma RII, activities known to be predominantly determined by CH2 domain structures. Reactivity of the IgA antibody, Z34, with IgG2b allowed further specificity studies using a panel of 26 mutant IgG2b proteins, each having single amino acid replacements over the surface of the CH2 domain. The only substitution that affected Z34 reactivity was Asn/Ala297, which destroyed the glycosylation sequon, resulting in secretion of an aglycosylated IgG molecule. The epitope recognized by Z34 therefore seems to be located outside of the Fc gamma R and C1q binding sites, but to be dependent on the presence of carbohydrate for expression. In contrast to the binding studies, complement activation by aggregated IgG2a, through classical or alternative pathways, was inhibited by the presence of autoantibodies. The functional significance of isotype-specific autoantibody in immune regulation is discussed.

Th1 clones that suppress IgG2ab specifically recognize an allopeptide determinant comprising residues 435-451 of gamma 2ab

We have previously reported that gamma 2ab/I-A(d)-specific Th1 clones from BALB/c mice (gamma 2aa, H-2d) mediated a long-lasting, selective suppression of serum IgG2ab levels when transferred to newborn (BALB/c x B10.D2)F1 (gamma 2a/b, H-2d) mice (Bartnes, K. and Hannestad, K. Eur. J. Immunol. 1991. 21: 2365). We here analyze the peptide specificity of hybridomas derived from two suppressive T cell clones. The shortest synthetic peptide with optimal antigenicity comprises gamma 2ab residues 435-451 (Kabat numbering). The determinant core encompasses the gamma 2ab 440-446 (KLRVQKS) sequence which contains an I-A(d) allele-specific motif. Challenge with single amino acid-substituted gamma 2ab 435-447 analogs revealed that residues K440, R442 and K445 which are shared by the autologous and allogeneic gamma 2a, as well as residues Q444 and S446 which represent allogeneic differences, are critical for recognition. We obtained evidence that K440, R442 and Q444 are epitope residues, while K445 and S446 contribute to anchoring of the peptide to I-A(d). Amino acids located outside of the core also influence antigenicity, the most striking effect being a 340-870-fold augmentation of potency when gamma 2ab 437-451 is extended by F436. IgG2ab required processing in order to stimulate the hybridomas. The data support the contention that the Th1 clones specific for Fc of gamma 2ab mediated IgG2ab suppression by cognate interaction with sIgG2ab+ B cells that presented a C gamma 2ab peptide(s) derived from their endogenous Ig on major histocompatibility complex class II. The T cells cross-reacted weakly with peptide 435-451 of the autologous gamma 2aa allotype. This opens the possibility that self-peptides from Ig C regions can target B cells for regulatory interactions with autologous Th cells.

Allelic polymorphism of mouse Igh-J locus, which encodes immunoglobulin heavy chain joining (JH) segments

The mouse genome contains four functional JH genes, which encode immunoglobulin heavy chain joining segments. The JH gene cluster is located a few kilobases 5' from the constant region genes (C genes) on chromosome 12. The polymerase chain reaction (PCR)-technique was used to amplify DNA stretches from mouse genome of approximately 1,340 nucleotides in length containing all four JH genes (Igh-J locus). PCR products were directly used as templates in Sanger's dideoxy-sequencing, and sequences were determined. Twelve inbred mouse strains belonging to ten different Igh-C haplotypes were studied. The strains were: BALB/c, C58/J, RIII, DBA/2, CE, RF, CBA, NZB/J, AKR, C57BL/10, SJL, and A/J. Five allelic forms of the Igh-J locus were found among these strains. The A/J mouse has an allele (e) which differs from the BALB/c allele (a) by 15 nucleotides. C57BL and SJL have the allele (b) with eight differences from BALB/c. The CBA allele (j) has two differences, and the CE allele (f) has a single nucleotide difference compared with the BALB/c sequence. Based on the JH, variable (V) and constant (C) region sequences we conclude that independent reshuffling of VH, JH, and CH gene clusters occurred during the evolution of Mus musculus.

Presentation of IgG2a antigens to class II-restricted T cells by stably transfected B lymphoma cells

Here we describe a panel of BALB/c T cells specific for IgG2a of the b allotype in association with I-Ad. We used DNA-mediated gene transfer techniques to localize antigenic determinants recognized by responding T cells. Initially a truncated IgG2aa gene comprising a variable domain and the CH3 domain (not including the membrane exons) from the BALB/c IgG2aa heavy chain was introduced into myeloma cells. The V-CH3 protein was expressed at high levels under control of the Ig heavy chain enhancer. Secretion of the V-CH3 protein did not require assembly of H-H dimers or an association with light chains. To generate stably transfected B cell lines that would stimulate our class II-restricted T cells, we replaced most of the BALB/c IgG2aa CH3 exon with CH3 coding sequences from a C57BL/6 IgG2ab cDNA clone and introduced these constructs into Ia+ B lymphoma cells. The IgG2ab CH3-transfected B cells were recognized by BALB/c Igh-1b-specific T cell hybrids in the absence of exogenous antigen. Experiments using glutaraldehyde-fixed cells as stimulators indicate that presentation of the secreted form of V-IgG2ab CH3 requires processing. We found that a significant fraction of the endogenously synthesized V-IgG2ab CH3 protein was, however, present as already processed antigen.

Genetic variability between two breeds based on restriction fragment length polymorphisms (RFLPs) of major histocompatibility complex class I genes in the pig

Restriction fragment length polymorphism analyses of SLA class I genes were performed on 55 Duroc and 24 Hampshire boars from the 1986-87 national performance tests of each breed. Few boars were inbred. Southern blotting and hybridization procedures were performed on genomic DNA isolated from white blood cells by using Pvu II, Bam HI, and Eco RI endonucleases and a swine MHC class I probe. Genetic variability within and between the two breeds was estimated in terms of nucleotide diversity, by using a mathematical analysis based on the different RFLP patterns. The nucleotide diversity calculated within each breed was less than that between the two breeds. The results from the nucleotide diversity analysis suggested that genetic variability was greater in the Duroc breed than in the Hampshire breed. A relatively high level of genetic variability was shown in the class I major histocompatibility complex genes in the pig.

The mouse Igh-1a and Igh-1b H chain constant regions are derived from two distinct isotypic genes

Genetic and structural analyses of the mouse genes encoding constant region of immunoglobulin subclass (Igh-C) have shown that recombination is rare within this cluster which is inherited as a set designated the Igh haplotype. Recent molecular analyses have demonstrated that either DNA exchanges or gene duplications have probably occurred during the evolution of this set of genes. In order to assess the generality of the duplication processes, the presence and expression of two allelic forms of the Igh-1 (gamma 2a) gene (Igh-1a and Igh-1b) were examined in a large panel of wild mice belonging to Mus musculus domesticus and Mus musculus musculus species. Our data indicate that certain M. m. domesticus animals and most animals in the M. m. musculus group coexpress the two allelic forms of Igh-1. Moreover, genetic studies show that these two immunoglobulin types are encoded by tandemly arranged genes. We propose that wild mice, from which laboratory mice are derived, carry three isotypic gamma 2 genes (Igh-1a, Igh-1b, Igh-3), and these have given rise to the two isotypes seen in laboratory strains by a deletion/insertion mechanism.

Evolution of protein inhibitors of serine proteinases: positive Darwinian selection or compositional effects?

In at least two instances involving serine proteinase inhibitors it has been shown that functionally important sites evolve faster and exhibit more interspecific variability than functionally neutral sites. Because these phenomena are difficult to reconcile with the neutral theory of molecular evolution, it has been suggested that the accelerated rate of amino acid substitution at the reactive sites is brought about by positive Darwinian selection. We show that differences in the amino acid composition in the different regions of proteinase inhibitors can account for the differences in the rates of amino acid substitution. By using an index of protein mutability [D. Graur (1985) J Mol Evol 22:53-62], we show that the amino acid composition of the reactive center in the ovomucoids and Spi-2 gene products is such that, regardless of function, they are expected to evolve more rapidly than any other polypeptide for which the rate of substitution is known. In addition, the reactive region in the Spi-2 proteins is shown to be free of compositional constraint. Positive Darwinian selection need not be invoked at the present time in these cases.

Relative roles of mutation and selection in the maintenance of genetic variability

The extent and pattern of protein and DNA polymorphisms are discussed with emphasis on the mechanism of maintenance of the polymorphisms. Statistical studies suggest that a large proportion of genetic variability at the molecular level is maintained by a mutation-drift balance. At some loci, such as those for histocompatibility in mammals, however, a form of overdominant selection seems to be involved. In the presence of overdominant selection, polymorphic alleles may be maintained for tens of millions of years, so that the number of nucleotide differences between alleles is often very large, as in the case of self-incompatibility alleles in plants. There are also an increasing number of examples in which an adaptive change of a morphological or physiological character is caused by a single nucleotide substitution. Nevertheless, these mutations seem to be a small proportion of the total nucleotide changes that contribute to genetic variability and evolution. Although there are many examples of frequency-dependent selection, this form of selection is apparently unimportant for the maintenance of genetic variability except in some special cases. Observations on the evolutionary change of DNA suggest that the driving force of evolution is mutation rather than selection.

Mouse immunoglobulin allotypes: multiple differences between the nucleic acid sequences of the IgEa and IgEb alleles

To clarify the allotypic difference of the IgE antibody molecule, we determined the complete nucleotide sequence of the genes encoding the constant portion of mouse IgE of a (BALB/c) as well as b (B10.A) allotypes. A comparison of the sequences revealed that there were 12 single-base changes: 2 single-base changes in CH1 and CH2, 3 in CH3, and 7 in CH4. Five of them were silent changes, but seven resulted in amino acid substitutions. Although the silent changes are scattered through CH1 to CH4, the nonsilent substitutions were found only in CH3 (two substitutions) and CH4 (five). The allotypic determinant(s) that conventional antisera detect most likely reflects an amino acid difference(s) in CH3 and/or CH4.

Sequence and transcriptional start site of the Pseudomonas aeruginosa outer membrane porin protein F gene

Porin F is one of the major proteins of the outer membrane of Pseudomonas aeruginosa. It forms water-filled pores of variable size. Porin F is a candidate for a vaccine against P. aeruginosa because it antigenically cross-reacts in all serotype strains of the International Antigenic Typing Scheme. We have isolated the gene for porin F from a lambda EMBL3 bacteriophage library by using oligodeoxynucleotide hybridization probes and have determined its nucleotide sequence. Different peptide sequences obtained from isolated porin F confirmed the deduced protein sequence. The mature protein consists of 326 amino acid residues and has a molecular weight of 35,250. The precursor contains an N-terminal signal peptide of 24 amino acid residues. S1 protection and primer extension experiments, together with Northern (RNA) blots, indicate that the mRNA coding for porin F is monocistronic with short untranslated regions of about 58 bases at the 5' end and about 47 bases at the 3' end. The sequences in the -10 and -35 regions upstream of the transcriptional start site are closely related to the Escherichia coli promoter consensus sequences, which explains why the porin F gene is expressed in E. coli under the control of its own promoter. The amino acid sequence of porin F is not homologous to the different E. coli porins OmpF, OmpC, LamB, and PhoE. On the other hand, a highly homologous region of 30 amino acids between the OmpA proteins of different enteric bacteria and porin F of P. aeruginosa was detected. The core region of the homology to E. coli OmpA had 11 of 12 amino acid residues in common.

Isolation and complete amino acid sequence of human thymopoietin and splenin

Human thymopoietin and splenin were isolated from human thymus and spleen, respectively, by monitoring tissue fractionation with a bovine thymopoietin RIA cross-reactive with human thymopoietin and splenin. Bovine thymopoietin and splenin are 49-amino acid polypeptides that differ by only 2 amino acids at positions 34 and 43; the change at position 34 in the active-site region changes the receptor specificities and biological activities. The complete amino acid sequences of purified human thymopoietin and splenin were determined and shown to be 48-amino acid polypeptides differing at four positions. Ten amino acids, constant within each species for thymopoietin and splenin, differ between the human and bovine polypeptides. The pentapeptide active site of thymopoietin (residues 32-36) is constant between the human and bovine thymopoietins, but position 34 in the active site of splenin has changed from glutamic acid in bovine splenin to alanine in human splenin, accounting for the biological activity of the human but not the bovine splenin on the human T-cell line MOLT-4.

Nonrandom allele associations between unlinked protein loci: are the polymorphisms of the immunoglobulin constant regions adaptive?

Consistent linkage disequilibrium was observed between independently segregating protein loci. In natural populations of the European rabbit Oryctolagus cuniculus, highly significant, nonrandom associations between alleles of the constant regions of the immunoglobulin light and heavy chains were found, both within localities and between localities. We suggest that the population genetic data presented here are relevant to the adaptive significance of the genetic polymorphisms of the antibody constant regions.

Immunoglobulin heavy chain locus of the rat: striking homology to mouse antibody genes

DNA encoding the rat diversity segment (D), joining segment (JH), and constant (C) region mu, gamma 2a, gamma 1, gamma 2b, epsilon and alpha of the Ig heavy chain has been isolated from a cosmid library. Restriction mapping allowed us to identify two gene clusters: D-JH-C mu and C gamma 1-C gamma 2b-C epsilon-C alpha in addition to a single C gamma 2a gene. Analysis of genomic DNA by Southern blotting permitted identification of the C gamma 2c gene and led to the proposal of the following gene order for the rat Ig heavy chain locus: D-JH-C mu-C delta-(C gamma 2c, C gamma 2a)-C gamma 1-C gamma 2b-C epsilon-C alpha. There is striking homology between the rat and mouse Ig heavy chain loci as regards gene order and distance between CH genes. Partial DNA sequencing confirms this homology and shows that exon sequences are more conserved than are intron sequences. One of the most conserved intron regions between rat and mouse is that spanning the Ig heavy chain enhancer (91% homology). However, the relationship between the different C gamma subclasses in rat differs from that in mouse. Comparison of the C gamma CH3 domains shows that the rat C gamma 2b gene is most homologous to mouse C gamma 2a/b, whereas the rat C gamma 1 and C gamma 2a genes, both very similar to each other, are most homologous to the mouse C gamma 1 gene.

Allotypic differences in murine mu genes

We report the complete DNA sequence of a c-DNA clone of the heavy chain mu b allele of the C57BL/6 mouse. Comparisons have been made with the nucleotide sequences of the germ line BALB/c mu a and the plasmacytoma TEPC-183 mu a alleles reported elsewhere over the entire length of the coding and the 3' untranslated region. In contrast to the extensive differences between the gamma 2a a and b alleles we have reported earlier we see a very high degree of homology between the mu alleles. Only one of the nucleotide differences between C57BL/6 mu b and BALB/c mu a leads to an amino acid substitution. This single amino acid exchange must form the allotypic determinant of the mu b allele. A comparison of four different DNA sequences indicates that they are all distinct IgM alleles.

A processed chicken pseudogene (CPS1) related to the ras oncogene superfamily

We describe the first polyA-containing processed pseudogene reported in the chicken. It includes a 0.52 kb open reading frame which could encode a 175 amino acid protein. The putative protein shows extensive homology to the ras oncogene superfamily, being most closely related to the yeast protein YP2. It is one of the two most divergent members of the ras superfamily yet described and is the most homologous of any ras-related protein to the G-protein alpha-transducin. The chicken genome contains at least one other gene highly homologous to CPS1; at least one member of the CPS1 family is active, but only early in chicken development. This pattern of expression, and the presence of mutations in regions known to activate human c-ras genes to oncogenicity, suggest that CPS1 may represent a new oncogene family.

Sequence of a human immunoglobulin gamma 3 heavy chain constant region gene: comparison with the other human C gamma genes

We report the first and complete nucleotide sequence of a human gamma 3 heavy chain constant region gene (C gamma 3). This gene displays the same organization than the others C gamma genes and exhibits normal RNA splice and polyadenylation sites. A comparison of its primary sequence with those of C gamma 1, C gamma 2 and C gamma 4 genes confirms the high degree of homology (95%) of the human family in both coding and non-coding regions, and the divergence of the hinge region. The C gamma 3 gene we sequenced codes for a Gm(b) gamma 3 chain (EZZ). Comparison with other known protein sequences reveals that only two specific aminoacids are involved in the Gm(b) and Gm(g) allotypes, which suggests an important part of the spatial configuration in the allotypic specificities.

Gene conversion in human immunoglobulin gamma locus shown by unusual location of IgG allotypes

The constant region of the gamma 1, gamma 2 and gamma 3 heavy chains of the human IgG1, IgG2 and IgG3 immunoglobulins carries antigenic determinants or G1m, G2m and G3m allotypes, which are genetic markers of these subclasses. The exceptional presence on gamma 1 and gamma 2 chains of Gm allotypes usually located on the CH3 domain of gamma 3 shows an unexpected clustering of base changes and subsequent identity of short DNA sequences in the CH3 exon of the non-allelic gamma 1, gamma 2 and gamma 3 genes. Such clusters of substitutions are not easily explained on the classical basis of point mutations. A gene conversion, which substituted a segment of the gamma 1 or gamma 2 gene with the homologous region of the non-allelic gamma 3 gene, is more likely. Other examples of possible conversion involving the gamma genes are described. The conservation or the restoration of short sequences produced by the conversion events might be related to the biological properties of the constant region of the heavy chains.

Monoclonal rheumatoid factors from B6-lpr/lpr mice

Monoclonal rheumatoid factors (MoRF) were prepared from autoimmune B6-lpr/lpr mice to investigate the influence of strain background on the specificity of these autoantibodies. Using screening assays for binding to heterologous rabbit IgG, four IgM MoRF were obtained. Three of these antibodies showed a broad pattern of reactivity with IgG antigen, binding BALB/c myeloma IgG1, IgG2a and IgG3 as well as heterologous IgG. One of the antibodies, however, had a distinct form of IgG interaction and was without reactivity with any of the BALB/c myelomas tested. None of the antibodies had significant reactivity with IgG2b. These results suggest common features of B6-lpr/lpr rheumatoid factor (RF) specificities, some of which may be shared by comparable products derived from MRL-lpr/lpr mice. Comparison of these antibodies with those in other reported series suggests that the background genome as well as the nature of the inducing mechanisms may affect the specificity of RF as well as their pathogenetic role.

The D segment defines the T15 idiotype: the immunoresponse of A/J mice to Pneumococcus pneumoniae

In the immune response of BALB/c mice (Igha) to Pneumococcus the majority of antibodies express the idiotype of the myeloma protein TEPC 15 (T15). In contrast mice of the A/J strain (Ighe) do not express this idiotype. Using (BALB/c X A/J)F1, F2 or backcross mice it could be shown that in allotype heterozygous animals (Igha/e) Pneumococcus pneumoniae preferentially stimulates B cells expressing a heavy chain (H) encoded by genes in the BALB/c H chain gene complex. Phosphorylcholine (PC)-specific hybridoma lines were established from BALB/c and A/J spleen cells and idiotypically analyzed using monoclonal antibodies (mAb) specific for the T15 idiotopes 32/65, 10/13, 16/13 or 21A5. Whereas the majority of the BALB/c PC-binding mAb express these idiotopes, only some of the A/J mAb are positive for one or the other of the idiotopes formed by the variable (V) regions of the H and the light chain of the myeloma protein T15. However, 80% of the A/J PC-binding hybridoma proteins were bound by the anti-idiotopic mAb 21A5. This mAb is specific for a determinant partially formed by the C alpha and partially by the V regions of the myeloma protein T15. The mRNA of one of these T15- A/J PC-binding hybridoma lines was sequenced. VH and V kappa were identical with sequences found for BALB/c T15-like antibodies. The sequence of the D segment was structurally very different. The importance of the D segment in the dominant expression of the T15 idiotype is discussed.

Structural analysis of the two tandemly repeated acid phosphatase genes in yeast

We have sequenced the genetically linked genes for repressible (PHO5) and and constitutive (PHO3) acid phosphatase from S. cerevisiae. Both genes are located on a 3.91 Kb BamHI and HpaI fragment, in the order (5') PHO5, PHO3 (3'). The mRNA transcripts have been analysed by S1-nuclease mapping. They show heterogenous initiation sites. Each of the PHO5 and PHO3 genes codes for 467 amino acids as deduced from the DNA sequence. The coding regions of the two genes show homology both at the nucleotide (82%) and the amino acid (87%) level. In the coding sequences, long stretches of homologous regions are flanked by small non-homologous regions. The nucleotide homology (65%) extends to some length into the 5' and 3' non-coding flanking sequences. Further upstream sequences are unrelated. The comparison of the NH2-terminal amino acid sequence deduced from the nucleotide sequence, with that of purified repressible acid phosphatase revealed the presence of a putative signal peptide.

Immunoglobulin kappa light-chain diversity in rabbit is based on the 3' length heterogeneity of germ-line variable genes

Antibody diversity is generated by the combinatorial association of multiple distinct genetic segments (variable (V), joining (J) and diversity (D) light (L) and heavy (H) chains--VL, JL and VH, D, JH) and amplified somatically by three or four different mechanisms. The kappa system in mouse and human consists of 50-100 V kappa segments associated with a cluster of four or five functional J kappa segments, located 2.5 kilobases (kb) 5' to a single C kappa gene. The third hypervariable region (CDR3), which is part of the antibody combining site, is usually nine amino acids long in human and mouse kappa chains. It is encoded by the last seven codons of the V kappa segment and the first two of the J kappa segment, one codon sometimes being added or deleted between V and J by junctional variation. In the rabbit, the C kappa 1 gene which encodes the major isotype, is associated with a cluster of five J kappa segments, only one of which seems to be functional, thus significantly decreasing the combinatorial potential. However, amino acid sequence comparison has revealed extensive heterogeneity in the length of rabbit CDR3 , suggesting the existence of a D segment analogous to that in the heavy-chain system. We show here that rabbit V kappa genes have several additional nucleotides at their 3' ends. Thus, even with a single functional J kappa segment, high CDR3 diversity can be generated based on the length heterogeneity of V kappa germ-line segments and their greater length, which might leave scope for an increased junctional deletion.

Is there a higher level genetic code that directs evolution?

Because the genetic code is redundant for most amino acids, different codons can be used in a given position without altering the structure of the protein for which the gene codes. This flexibility permits information encoding structural, and therefore functional, properties of RNA and DNA to be transmitted simultaneously by a protein-coding sequence of DNA. Among the other messages that might be transmitted, it is proposed, is one modulating the evolution of the DNA itself.

Recent gene conversion involving bovine vasopressin and oxytocin precursor genes suggested by nucleotide sequence

The nonapeptide hormones arginine vasopressin (AVP) and oxytocin (OT) are synthesized in the hypothalamus together with their carrier proteins, the neurophysins, as common polypeptide precursors. The organization of these precursors has been established by sequence determination of cloned bovine cDNAs encoding prepro-arginine vasopressin-neurophysin II (prepro-AVP-NPII) and prepro-oxytocin-neurophysin I (prepro-OT-NPI). When the mRNA sequences coding for the conserved middle part of the neurophysins were compared, we found that these sequences are not merely similar but identical. The primary structure of the chromosomal genes now determined shows that both genes, which appear to have arisen by a gene duplication, are split into three exons, each encoding a functional domain of the precursor polypeptide. Sequence comparison reveals that the stretch of sequence identity within the two mRNAs is probably the result of a gene conversion encompassing exon B, which encodes the conserved part of the neurophysins, and part of the preceding intron.

Evidence for gene conversion among immunoglobulin heavy chain variable region genes

We have previously reported that the VH region amino acid sequence of a phosphocholine (PC)-binding hybridoma antibody of CBA/J origin, HP101 6G6 (6G6), differs extensively from the VH regions of other PC-binding antibodies. The sequence of 6G6 VH appears to be derived from a gene homologous to the BALB/c V11 gene, a member of the PC VH (T15 VH) gene family not normally used to encode PC-binding antibodies. The 6G6 VH sequence differs from the translated sequence of V11 by six amino acids, four of which occur at the same position in other members of this gene family. This coincidence led to the proposal that the 6G6 VH gene was derived by gene conversion involving three genes of the PC VH gene family. We report here the nucleic acid sequence of the rearranged VH gene of hybridoma 6G6. This sequence supports our previous suggestion of gene conversion by confirming those differences, relative to the BALB/c V11 gene sequence, that are encoded by other members of this gene family, and extends this correlation to include three silent base pair substitutions as well. In addition, 5' noncoding region sequence and Southern blot analysis using probes derived from the coding and 5' noncoding regions confirm that the 6G6 VH gene is likely to be derived from the V11 homologue in CBA/J mice, and suggest that all three genes believed to be involved in the generation of the 6G6 VH gene are present in the CBA/J genome, a prerequisite for their involvement in gene conversion.

Ruminant globin gene structures suggest an evolutionary role for Alu-type repeats

Bovine fetal and adult globin genes were cloned and subjected to DNA sequence analysis. Both of these genes contained insertions of Alu-type repetitive DNA within their introns. Comparison of cow and goat beta-type globin genes indicates that intragenic DNA insertions played a role in their evolution. These data support the theory that Alu-type repeats maintain genetic diversity by inhibiting gene conversion.

Somatic mutation creates diversity in the major group of mouse immunoglobulin kappa light chains

Using a cloned cDNA of a mouse immunoglobulin kappa light chain synthesized in a myeloma MOPC321 (V kappa-21 subgroup C) as a probe we could detect 13 germ line V kappa gene segments. 11 of these were isolated. Using a set of overlapping cloned segments, we showed that nine of these germ line V kappa genes are arranged in two linkage clusters and that they all have the same transcriptional orientation (11, 12, 22). These two clusters occupy 90 and 30 kb of chromosomal DNA and contain six and three V kappa's, respectively. We determined the complete nucleotide sequences of five germ line V kappa's and showed that three of them encode the prototype sequence of V kappa-21 subgroups B, C, and E. None of these five germ line V kappa's encodes the variant amino acid sequences of known V kappa-21 subgroups. We thus conclude that, as in the lambda 1 light chains, the variant V regions are encoded by gene segments derived by a few somatic mutations from the corresponding germ line DNA. Such somatic mutations are not restricted to sequences encoding the hypervariable regions: they also occur in sequences encoding framework regions.

Idiotypic inheritance: allelic relationships between immunoglobin genes

An analysis of genes related to the expressed VH region of anti-NP antibodies in C57BL/6 mice showed that 24 distinct genes could be grouped into 3 subfamilies. Some of these genes showed striking homology with a family found in BALB/c mice. The extension of diversity through gene conversion seems to be very important in the VH family.

Structural relationships among mouse and human immunoglobulin VH genes in the subgroup III

The mouse VHIII subgroup is composed of four families which share sequence homology. We isolated a VH germ-line genomic clone, which cross hybridizes with a cDNA probe from one of these families, derived from a myeloma secreting an antigalactan antibody. We report here the nucleotide sequence of the cross hybridizing gene and show that very likely it has an anti-sheep red blood cell specificity. Comparison of its nucleotide sequence with those of the three other VHIII families shows that these genes share segmental homologies of variable lengths. This suggests that interchanges of sequence blocks between VH genes could be an important evolutionary mechanism for diversifying the germ-line repertoire. The strong homology (82%) with human VHIII genes suggests that efficient antibody sequences are strongly conserved. This conservation of homology is particularly striking when compared to the more limited homology (63%) between mouse and human C kappa genes.