Human immunoglobulin heavy chain genes: evolutionary comparisons of C mu, C delta and C gamma genes and associated switch sequences

Evidence for somatic selection of natural autoantibodies

Natural autoantibodies are primarily immunoglobulin M (IgM) antibodies that bind to a variety of self-antigens, including self-IgG. Accounting for a large proportion of the early B cell repertoire, such polyspecific autoantibodies are speculated to contribute to the homeostasis and/or competence of the primary humoral immune system. Recent studies indicate that the leukemia cells from most patients with chronic lymphocytic leukemia (CLL) also express such IgM autoantibodies. Similarly, the leukemia cells from many CLL patients react with murine monoclonal antibodies (mAbs) specific for crossreactive idiotypes (CRIs) associated with human IgM autoantibodies. In particular, leukemic cells frequently react with G6, a mAb specific for an Ig heavy chain (H chain)-associated CRI, and/or with 17.109, a mAb that defines a kappa light chain (L chain)-associated CRI. Generated against IgM rheumatoid factor (RF) paraproteins, G6 and 17.109 each recognize a major CRI that is present in many IgM RF paraproteins. Furthermore, over 90% of the IgM paraproteins found to bear both H and L chain-associated CRIs also are found to have RF activity. Molecular characterization of these CRIs demonstrates that each is a serologic marker for expression of a highly conserved Ig V gene. As such, the frequent production of IgM polyspecific autoantibodies in CLL simply may reflect the frequent use of such highly conserved autoantibody-encoding Ig V genes with little or no somatic mutation. To test this hypothesis, we generated murine transfectomas to pair the 17.109-reactive kappa L chain of SMI, a 17.109/G6-reactive CLL population, with the Ig H chain of SMI or other G6-reactive leukemia cells or tonsillar lymphocytes. Cotransfection of vectors encoding the Ig H and L chains of SMI generated transfectomas that produce IgM kappa RF autoantibodies reactive with human IgG1 and IgG4. In contrast to G6/17.109-reactive IgM kappa RF Waldenstrom's paraproteins, the SMI IgM kappa also reacts with several other self-antigens, including myoglobin, actin, and ssDNA. However, cotransfection of the SMI L chain with a vector encoding any one of 10 different G6-reactive Ig H chains generated transfectomas that produce IgM kappa antibodies without detectable polyspecific autoantibody activity. These results indicate that polyspecific antiself-reactivity of G6/17.019-reactive Ig is dependent on the somatically generated Ig third complementarity determining region. Collectively, these studies imply that selection may be responsible for the frequent expression of polyspecific autoantibodies in CLL and early B cell ontogeny.

Role of the transmembrane and cytoplasmic domains of surface IgM in endocytosis and signal transduction

The cross-linking of membrane IgM (mIgM) triggers the activation and differentiation of B lymphocytes. One very rapid result of the cross-linking is the activation of phospholipase C, the subsequent mobilization of free calcium from internal stores and the activation of protein kinase C. This is followed by a redistribution of the receptor-ligand complexes to a small cap on the B cell surface, the first step in endocytosis and antigen processing. Cross-linking of major histocompatibility complex (MHC) class I neither stimulates the release of intracellular calcium nor does it induce capping and endocytosis of the cell surface receptors. In this study, we sought to determine the role of the two carboxyterminal domains of the mu heavy chain in signal transduction, capping and endocytosis of mIgM. We took advantage of the clear differences between MHC class I molecules and mIgM, replacing the transmembrane and cytoplasmic domains of mu by their MHC class I equivalents. Our results show that the hybrid heavy chain could still associate with light chains and assemble into a tetramer on the cell surface. However, cross-linking of the hybrid cell receptor produced neither release of calcium from internal stores, nor capping and endocytosis. These observations demonstrate that the two carboxy-terminal domains of mu are critical to both signal transduction and modulation of the mIgM-ligand complexes from the surface of B lymphocytes.

Rapid cloning of rearranged immunoglobulin heavy chain genes from human B-cell lines using anchored polymerase chain reaction

A general method to obtain the variable region DNA sequence of the immunoglobulin heavy chain using anchored polymerase chain reaction is described. Based on this DNA sequence, clone-specific oligonucleotides were designed to anneal to the complementarity determining regions. These were used to identify the original B-cell clone in serial dilutions of polyclonal lymph node DNA with high specificity and sensitivity. This method should be useful for studying minimal residual disease in B-cell neoplasia.

Monokine gene expression in normal human liver: selective involvement of the portal compartment

Monokines play a major role in the regulation of hepatocyte functions. To document a possible in situ production of these mediators under physiological conditions, expression of IL-1 beta and of IL-6 genes was analyzed by in situ hybridization in four histologically normal human livers. We detected cells containing IL-1 beta mRNA or IL-6 mRNA in all cases. Cells expressing either the IL-1 beta gene or the IL-6 gene were found with equal frequency and were similarly distributed. Although present in all liver compartments, they were selectively enriched in portal areas, in which they were detected both in endothelial positions and in perivascular connective tissues. Few positive cells were observed in hepatic lobules, most of them being located in the walls of centrolobular veins, in an endothelial position. Subcapsular cells were also shown to express monokine genes. The location of positive cells and their pattern of labelling suggested that macrophages, fibroblasts and endothelial cells were the main cell populations expressing monokine genes. In contrast, Kupffer cells, biliary epithelial cells and hepatocytes did not express monokine genes. No marker of immune activation other than monokine gene expression was detected in these histologically normal livers. The expression of the IL-1 beta gene and of the IL-6 gene may be induced by gut-derived LPS, and could play a role in the modulation of hepatocyte function in normal liver.

B cell lymphoma-associated chromosomal translocation involves candidate oncogene lyt-10, homologous to NF-kappa B p50

A B cell lymphoma-associated chromosomal translocation, t(10;14)(q24;q32), juxtaposes the immunoglobulin C alpha 1 locus to a novel gene, lyt-10. The normal lyt-10 cDNA codes for a 98 kd protein which displays amino-terminal homology with the rel (DNA-binding) domain of the NF-kappa B-rel family of transcription factors and carboxy-terminal homology with the NF-kappa B p50 precursor protein, including the putative proteolytic cleavage domain (poly-G) and the ankyrin-like repeat domains. The lyt-10 protein can bind to kappa B sequences in vitro, although with different specificity from NF-kappa B p50, and in vitro DNA-binding is activated by removal of the ankyrin domain. Chromosomal translocation generates an lyt-10-C alpha 1 fusion gene coding for a protein that retains the rel effector domain, lacks the ankyrin regulatory domain, and binds kappa B sequences in vitro, suggesting its constitutive activation in vivo. Analogous rearrangements of the lyt-10 gene have been found in an additional three cases of lymphoid neoplasia. The lyt-10 gene defines a new subfamily (rel/poly-G/ankyrin) of NF-kappa B-rel transcription factors with potential for oncogenic activation in human cancer.

Burkitt lymphoma cell lines are prone to recombination in the switch region of the Ig mu heavy chain locus

Different recombinations have been found at the Ig heavy chain gene loci in a number of sublines of the Burkitt lymphoma (BL) cell line Namalwa, following prolonged in vitro culture. The Namalwa sublines examined are DNA fingerprint-identical and derived from a monoclonal source. Recombinant DNA clones were used to map the Ig heavy chain gene mutations to a region between the VDJ and C mu segment of the locus. This region is associated with Ig heavy chain class switching in normal B cells. Of 24 clones established from one subline, three were found to have additional VDJ-C mu region mutations, indicating a high frequency of mutation at this locus.

An epitope on membrane-bound but not secreted IgE: implications in isotype-specific regulation

Immunoglobulins (Igs) on the surface of B lymphocytes are isotype-specific immunological markers of the B-cell subsets expressing them. Since these membrane-bound Igs (mIgs) are antigen receptors, their interaction with antibodies could be explored for modulating the activity of specific B-cell subsets. Targeting mIgs by antibodies in vivo, however, has not been feasible because of the presence of Igs in the circulation and the frequent association of Igs with various cell types via Fc receptors. To circumvent these problems, we proposed that the extracellular portions of the membrane-anchoring segments of the heavy chains of mIgs, referred to as "mIg isotype-specific" or "migis" peptides, may provide the antigenic sites for the isotype-specific targeting of B cells in vivo. Here we describe the exemplary development of monoclonal antibodies (mAbs) recognizing this unique epitope of mIgE.

mRNA transcripts initiating within the human immunoglobulin mu heavy chain enhancer region contain a non-translatable exon and are extremely heterogeneous at the 5' end

Transcription events are thought to precede gene rearrangement in the immunoglobulin (Ig) loci and may be the mechanism by which the various gene regions are made accessible for recombination. If this is the case, identification and characterization of transcripts from the Ig loci should permit a better understanding of the gene rearrangement process. We have isolated a 2.3 kb cDNA clone from the human pre-B cell line Nalm-1 that contains enhancer-specific sequences from the Ig heavy (H) chain gene locus. The 2.3 kb transcript initiated within the enhancer region and showed extreme 5' heterogeneity, with more than 50 initiation sites mapping near the Ig-specific octamer ATTTGCGT. Sequencing of the cDNA clone demonstrated that 644 nucleotides from the Ig enhancer region were incorporated as a leader exon spliced to the mu constant (Cmu) region. This leader exon contained many translation termination codons and may function to inhibit the translation of sterile Cmu polypeptides. Using an enhancer-derived probe, we detected two low-abundancy mRNA transcripts with sizes of 2.3 and 12 kb. Northern blot analysis suggested that the 12 kb transcript was the unspliced precursor mRNA of a VDJ rearrangement. The potential role of these enhancer-containing transcripts in the opening of the IgH chain gene for rearrangement and for class switching is discussed.

Sequencing of selected regions of the human immunoglobulin heavy-chain gene locus that completes the sequence from JH through the delta constant region

Much of the nucleotide sequence between the start of the joining region and the end of the immunoglobulin heavy chain delta gene has already been determined. However, two gaps existed in potentially functionally important regions in this sequence: the region between the 3' end of the joining region and the heavy chain enhancer region and that between the enhancer and the mu constant region. We have determined the nucleotide sequences of these regions. The 734 bp between the joining and enhancer regions contained no additional joining regions. The 4525 bp region between the heavy chain enhancer and the mu constant region contains the mu switch region, which consists of pentameric repeats. Approximately 60% of these repeats are GGGCT and GAGCT. With the determination of these sequences, the entire region of the heavy chain locus starting upstream of the joining region to downstream of the last exon of the delta constant region (a total of more than 29 kb) has now been sequenced.

Human chronic lymphocytic leukemia cells regularly express mRNAs of the protooncogenes lck and c-fgr

Protooncogenes, the cellular homologues of transforming retroviral oncogenes, have been suggested to participate in malignant transformation of various cell types. The role activated protooncogenes might play in chronic lymphocytic leukemia (CLL) is not yet clear. We investigated the expression of the protooncogenes lck and c-fgr encoding two protein tyrosine kinases by Northern-blot analysis in 21 patients with B-CLL and transcripts of both genes were detected in all cases. In contrast to CLL B cells, normal B cells from peripheral blood of a healthy donor separated by magnetic sorting did not express c-fgr and lck. The results suggest a possible role for these two protooncogenes of the src family in the activation and/or transformation of human B lymphocytes.

B cells of chronic lymphatic leukemia express V genes in unmutated form

In order to investigate whether the leukemic B cells in B-CLL express immunoglobulin genes in mutated or unmutated form, independent cDNA clones of one patient with B-CLL expressing heavy and light chain V region genes were sequenced. The sequences of both the VH and V kappa clones were identical. The V genes could be assigned to known germline V genes. No somatic mutations were found. At the V kappa-J kappa border there is an insertion of N-sequences which are only rarely found in immunoglobulin light chain genes. Our study confirms other published data on V gene expression in B-CLL in that the surface immunoglobulins in these tumors are unmutated.

Identification of two distinct phosphoproteins as components of the human B cell antigen receptor complex

In human B cells, the molecules that, upon receptor occupancy, couple membrane immunoglobulin to intracellular signal transduction pathways have never been identified. We here describe two phosphoproteins as integral parts of the B cell antigen receptor complex. Membrane IgM is non-covalently associated with a disulfide-linked heterodimer of glycoproteins. These molecules can be demonstrated on B cell lines and freshly isolated polyclonal B cell populations and are subject to phosphorylation at serine residues. Identification of these constituents of the B cell receptor complex opens up the opportunity to study coupling of the B cell antigen receptor to the intracellular signal transduction machinery at the molecular level.

Gene segments encoding membrane domains of the human immunoglobulin gamma 3 and alpha chains

The carboxyterminal region of the heavy chains, according to its hydrophilic or hydrophobic properties, determines whether the immunoglobulin will be secreted or membrane-bound. We have determined the nucleotide sequences of the human IGHG3, IGHA1, and IGHA2 membrane exons isolated from genomic DNA libraries. The IGHG3 M1 and M2 exons are separated by a long intron of 2.1 kilobases (kb) containing an highly repeated motif of 34 base pairs (bp). The IGHA1 and IGHA2 genes, like the mouse Igh-A gene, have a single exon encoding the extracellular, transmembrane, and cytoplasmic regions. For each class of immunoglobulins, the sequences of membrane exons are highly conserved between human and mouse, but no alignment is possible for the flanking regions. In contrast, for a same species, the sequences of the heavy chain membrane exons differ from one class to another. While the hydrophobic profile of the membrane core is well conserved, the cytoplasmic region differs in length and in composition. None of the intracellular domains presents the sequence implied in signal transduction, implying that membrane immunoglobulins need other proteins, which probably interact with the constant or membrane domain, to transmit signals leading to B-cell activation.

Preferential utilization of conserved immunoglobulin heavy chain variable gene segments during human fetal life

The ability to respond to specific antigens develops in a programmed fashion. Although the antibody repertoire in adults is presumably generated by stochastic combinatorial joining of rearranged heavy variable, diversity, and joining (VH-DH-JH) and light (VL-JL) chains, experimental evidence in the mouse has shown nonrandom utilization of variable gene segments during ontogeny and in response to specific antigens. In this study, we have performed sequence analysis of 104-day human fetal liver-derived, randomly isolated constant region C+ mu transcripts and demonstrate a consistent preference during fetal life for a small subset of three highly conserved VH3 family gene segments. In addition, the data show that this preferential gene segment utilization extends to the DHQ52 and the JH3 and JH4 loci. Sequence analysis of two "sterile" DH-JH transcripts suggests that transcriptional activation of the JH-proximal DHQ52 element may precede initiation of DH-JH rearrangement and influence fetal DH utilization. Sequence comparisons reveal striking nucleotide polymorphism in allelic gene segments which is poorly reflected in the peptide sequence, implying considerable evolutionary selection pressure. Although vertebrate species utilize a variety of strategies to generate their antibody repertoire, preferential utilization of VH3 elements is consistently found during early development. These data support the hypothesis that VH3 gene segments play an essential role in the development of the immune response.

Molecular cloning and characterization of the human double-stranded RNA-activated protein kinase induced by interferon

The double-stranded (ds) RNA-activated protein kinase from human cells is a 68 kd protein (p68 kinase) induced by interferon. On activation by dsRNA in the presence of ATP, the kinase becomes autophosphorylated and can catalyze the phosphorylation of the alpha subunit of eIF2, which leads to an inhibition of the initiation of protein synthesis. Here we report the molecular cloning and characterization of several related cDNAs from which can be deduced the full-length p68 kinase sequence. All of the cDNAs identify a 2.5 kb RNA that is strongly induced by interferon. The deduced amino acid sequence of the p68 kinase predicts a protein of 550 amino acids containing all of the conserved domains specific for members of the protein kinase family, including the catalytic domain characteristic of serine/threonine kinases. In vitro translation of a reconstructed full-length p68 kinase cDNA yields a protein of 68 kd that binds dsRNA, is recognized by a monoclonal antibody raised against the native p68 kinase, and is autophosphorylated.

Immunoglobulin D switching can occur through homologous recombination in human B cells

Prototypical class switching in mouse and human immunoglobulin heavy chains occurs through recombination of tandem blocks of short repeats located 5' to each heavy chain constant region (CH) except C delta. Deletion of C mu in immunoglobulin D (IgD)-secreting murine plasmacytomas occurs illegitimately. We demonstrate here that in human IgD-secreting myeloma cells freshly isolated from patient bone marrow and in normal peripheral blood B lymphocytes, an IgD switch can occur through homologous recombination of a direct repeat consisting of a 442-bp sequence 1.5 kbp 3' of the JH complex and a 443-bp sequence that is duplicated almost perfectly (96% similarity) 1.7 kbp 5' of the C delta gene (442/443-base-pair [bp] repeat). This homologous recombination mechanism is not exclusive for IgD switching, since C mu deletion endpoints in two established IgD-secreting myeloma cell lines fall outside the 442/443-bp repeat. The 442/443-bp mediated recombination shows cell type specificity, and we propose that it represents a unique mode for increased levels of IgD secretion in humans.

Report of an unusual small lymphocytic B-cell lymphoma selectively involving the B-zone of lymph node

Selective involvement of the B-cell compartment of lymph node by B-cell malignant lymphomas is an occasional finding related to early phases of lymph node infiltration. The authors have observed a unique case of diffuse small lymphocytic lymphoma that consisted of immunohistologically and genotypically proven B-clonal population exhibiting a repetitive pattern of infiltration in three lymph node samples obtained from the patient during a 9-year period. This pattern consisted of a selective and complete replacement of the B-areas with disappearance of follicles and widening of the medullary cords, an expanded T-zone showing features consistent with dermatopathic lymphadenitis and well-preserved sinuses. Clinically, multiple involved sites at presentation (lymph nodes, spleen, skin, bone marrow, and peripheral blood) and during the 9-year follow-up (testis) were detected, and the disease was associated with a relative indolent course like other low-grade lymphomas. The phenotypic profile of lymphoma cells studied by immunoperoxidase method, and by single-labeling and double-labeling flow cytometric analyses (SIg+, K+, LN2+, MB1+, MB2+, HLA-DR+, CD 9+, CD19+, CD 20+, CD 21+, CD 22+, CD 24+, Leu 8+, CD 5-, CD 10-, CD 11b-, CD 11c-, CD 25-, CD 38-, PCA-1-, FMC-7-, CD 23-) was consistent with a B-cell proliferation at an intermediate stage of differentiation but distinct from other well-defined B-cell neoplasms. Whether such unique B-zone pattern was due to an intrinsic property of this lymphoma or it is to be related to the coexisting reactive T-zone expansion remains controversial.

Genotype markers and proto-oncogene analysis in the CD30-positive "malignant histiocytosis" DEL cell line with t(5;6)(q35;p21)

The DEL cell line isolated from a patient who died of malignant histiocytosis exhibits a reciprocal chromosomal translocation t(5;6)(5q35;6p21). The cells were analyzed for Ig(Jh), TCR beta-gene rearrangements and proto-oncogene expression pattern, using a panel of molecularly cloned probes that included c-fms, c-myc, c-myb, c-pim, c-fos, N-myc, c-sis, c-fgr as well as the virally derived probes v-ki-ras and v-src. Consistent levels of expression of c-fms, c-myc, c-myb, c-ki-ras and c-fgr were identified in cells from several in vitro passages as well as from the heterotransplanted tumors in nude mice. Transcripts homologous to the c-fos, c-src and c-sis were not observed. Southern blot study of DNA showed that the banding pattern of the screened proto-oncogenes was not altered. Furthermore, Southern blot analysis demonstrated monoallelic immunoglobulin heavy chain (IgJh) rearrangement but a normal germ-line configuration of the kappa light chain and TCR beta-genes. These results appear to imply that a T- or B-cell origin can be eliminated and that several activated proto-oncogenes, usually expressed in immature MPS cells (c-fms) and myeloblastic cells (c-fgr), may be implicated in the proliferative activity of the DEL cell line, the stem of which may be a primitive, ancestral myelomonocytic cell.

Immunoglobulin D switching can occur through homologous recombination in human B cells

Prototypical class switching in mouse and human immunoglobulin heavy chains occurs through recombination of tandem blocks of short repeats located 5' to each heavy chain constant region (CH) except C delta. Deletion of C mu in immunoglobulin D (IgD)-secreting murine plasmacytomas occurs illegitimately. We demonstrate here that in human IgD-secreting myeloma cells freshly isolated from patient bone marrow and in normal peripheral blood B lymphocytes, an IgD switch can occur through homologous recombination of a direct repeat consisting of a 442-bp sequence 1.5 kbp 3' of the JH complex and a 443-bp sequence that is duplicated almost perfectly (96% similarity) 1.7 kbp 5' of the C delta gene (442/443-base-pair [bp] repeat). This homologous recombination mechanism is not exclusive for IgD switching, since C mu deletion endpoints in two established IgD-secreting myeloma cell lines fall outside the 442/443-bp repeat. The 442/443-bp mediated recombination shows cell type specificity, and we propose that it represents a unique mode for increased levels of IgD secretion in humans.

Presence of immunoglobulin (Ig) M and IgG double isotype-bearing cells and defect of switch recombination in hyper IgM immunodeficiency

We established a transformed B cell line expressing both IgM and IgG on the cell surface from a patient with hyper IgM immunodeficiency using Epstein-Barr viruses. DNA and RNA of the cells were analyzed. DNA rearrangements of Ig JH gene loci were observed on both chromosomes. Cloning and DNA sequence analyses showed that one has a VHDHJH structure while the other has a DHJH structure. Southern hybridization with 5'-S mu and S gamma region-containing probes indicated germline configuration in the switch regions of mu and gamma genes on both chromosomes. To test expression of mu and gamma chains in the transformed cells at the mRNA-level, we used the polymerase chain reaction with three kinds of synthetic oligonucleotides as primers, one of which was part of the VH gene, while the other two were complementary to parts of C mu and C gamma genes. Sequence analysis of the amplified products showed that the same VHDHJH sequence is directly connected with either the C mu or the C gamma sequence in the mRNAs. This is direct evidence showing that in double isotype-bearing cells one VHDHJH exon in the transcript is alternatively spliced to C mu or C gamma without DNA rearrangement. The defect in this disease could be at the S-S recombination stage.

Three B-cell surface molecules associating with membrane immunoglobulin

Three cell-surface molecules associated with surface Ig on murine B cells have been identified. Surface-iodinated resting B cells were lysed in digitonin or Nonidet P-40. The lysates were then examined by immunocoprecipitation with anti-Ig reagents and SDS-PAGE. By this procedure, a 73,000 MW heterodimer, composed of 35,000 and 38,000 MW disulphide-linked subunits, was found to be associated with both mIgM and mIgD. This association was detectable in the detergent digitonin, but not Nonidet P-40. As an alternative approach, resting B cell mIg was capped and internalized by anti-Ig induced mIg cross-linking. The resulting 'modulated' cells were then screened for surface reactivity with a panel of monoclonal antibodies. Most antibodies tested (anti-class I, -class II, -LFA, -FcR gamma, -FcR epsilon) reacted equally well with control and modulated B cells. With two, however, there was a significant reduction in surface representation on modulated B cells. One determinant corresponds to a previously undescribed 90,000 MW murine surface marker, and the other was a polydisperse glycoprotein recognized by the mAb J11D.

Recurrent mutation of immunoglobulin and c-myc genes and differential expression of cell surface antigens occur in variant cell lines derived from a Burkitt lymphoma

The phenotypic and molecular genetic characteristics of 4 variant sublines of the Burkitt lymphoma cell line Namalwa have been examined. The sublines are DNA-fingerprint-identical and derived from a monoclonal tumour, as shown by a rearrangement of the T-cell-receptor beta-chain gene common to the 4 sublines. There is non-co-ordinate expression of MHC class-I MHC class-II, surface immunoglobulin and a number of antigens recognized by CD MAbs on the different sublines. These different phenotypes of the cells are reminiscent of B cells arrested in varying states of cellular maturity. On Southern blots there are different patterns of restriction fragments hybridizing with Ig heavy- and light-chain gene probes among the sublines, indicating that multiple rearrangements or mutations of Ig genes have occurred in the cells. Different patterns of hybridizing fragments among the sublines were also found by using c-myc probes, implying the existence of different mutations of the c-myc locus. The c-myc mutation found in one of the sublines mapped to the 5' flanking sequence and in another 3' to the c-myc locus. Using the J17BS8 probe, which detects a restriction fragment length polymorphism in the 3' flanking region of the c-myc gene, a 4-fold variation in the gene copy number among the subline was found and one of the sublines was shown to be hemizygous for c-myc. Examination of DNA from early cultures of Namalwa cells showed that the alternations in Ig and c-myc structure had occurred on prolonged culture of the cells.

De novo Ph negative T-cell lymphoblastic leukaemia associated with bcr gene rearrangement

One case of de novo Ph-negative T-cell acute lymphoblastic leukaemia has been found to have a classical breakpoint cluster region (bcr) rearrangement of the type seen in chronic granulocytic leukaemia. There were no haematological features to suggest a previous chronic phase. This case represents the first report of this rearrangement in Ph negative acute T-lymphoblastic leukaemia at presentation. The implications for various therapeutic options in such patients are discussed.

Genomic alterations in a case of alpha heavy chain disease leading to the generation of composite exons from the JH region

Human alpha heavy chain disease (HCD) is characterized by the presence in patient's serum of a short Ig alpha chain devoid of light chains. We analyzed the serum protein, the alpha HCD mRNA and the productive rearranged H chain gene from the leukemic cells of a new case (YAO) of alpha HCD. The abnormal YAO alpha 1 Ig was devoid of VH and CH1 domains and started at the beginning of the hinge region. The alpha HCD mRNA was shorter than normal alpha mRNA and the cDNA prepared from YAO mRNA encoded a leader sequence, an insert of 70 nucleotides and the CH2 and CH3 exons. The origin of the inserted sequence was assessed by cloning and sequence analysis of the alpha 1 productive gene. It started with a leader exon, a leader-VH intron and the first 11 bp of a VH exon. Then the VH region was deleted and replaced by a 19-nucleotide sequence that turned out to correspond to the 3' part of a modified JH5 exon. It was followed by a 221-bp sequence homologous to the JH5-psi JH3 intron and by an inserted sequence of unknown origin. The 3' part of this insertion and the remnant of a JH6 exon delineated a third exon that was followed by a relatively conserved JH6-C alpha intron. These two composite exons were flanked by splicing sites and accounted for the 70-nucleotide insert of the cDNA. The genomic nucleotide sequence also revealed a large deletion in the switch CH1 region which eliminated normal splicing sites and resulted in splicing of the third exon directly to the CH2 exon.

A human immunoglobulin IGHG3 allele (Gmb0,b1,c3,c5,u) with an IGHG4 converted region and three hinge exons

The five human IGHG genes consist of three constant domain exons plus one of or four hinge exon(s), the quadruplicated hinge region being characteristic of the IGHG3 gene. Besides this structural difference, the IGHG genes are polymorphic, as demonstrated by the restriction fragment length polymorphism and, at the protein level, by the Gm allotypic antigenic determinants. In this paper, we report the sequence of the G3m(b0, b1, c3, c5,u) IGHG3 allele, typical of the Black African populations and of populations with Negroid admixture, found in a homozygous Tunisian designated as LAT. We demonstrate that this G3 allele contains only three hinge exons instead of four (the probable result of an unequal crossing over) and that IGHG3 genes with triplicated hinge exons (and therefore encoding shorter gamma 3 chains) are present in healthy individuals from different populations. Moreover, we show that the LAT G3m (b0, b1, c3, c5, u) coding sequence results from the conversion, in the CH3 exon, of the G3m (b0, b1, b3, b4, b5, u, v) allele, the most frequent IGHG3 gene in the Negroid populations, by the homologous region of a IGHG4 gene. The structural features of the LAT IGHG3 allele, which are the lack of one hinge exon and its conversion by the IGHG4 gene, demonstrate that both crossing-over and gene conversion events occur in the evolution of the human IGHG genes.

Interleukin 6 induces secretion of IgG1 by coordinated transcriptional activation and differential mRNA accumulation

The molecular mechanism by which interleukin 6 (IL-6) induces terminal differentiation of B cells was investigated in a subpopulation of the clonal human B-lymphoblastoid cell line CESS selected for high density of cell surface IgG1. Induction of CESS cells with IL-6 resulted in a 15-fold preferential accumulation of secreted-specific gamma 1 (gamma 1s) mRNA but not of the alternatively processed membrane-specific gamma 1 (gamma 1m) mRNA. Similarly, microseconds mRNA but not the microns mRNA of the nonproductively rearranged mu heavy-chain allele was also increased. Accompanying the differential accumulation of gamma 1s mRNA was a 4.5-fold increase in lambda light-chain mRNA, leading to secretion of IgG1. Analyses of transcription in isolated nuclei demonstrated that transcriptional activation was the primary mechanism for quantitative increase of immunoglobulin mRNAs (5.5-fold for gamma 1 and mu and at least 2-fold for lambda). Since polymerase loading is diminished by 75% before reaching the downstream gamma 1m polyadenylylation site in CESS cells, irrespective of IL-6 induction, transcriptional pausing/termination appears intrinsic and contributes to the selection of gamma 1s and gamma 1m polyadenylylation sites in activated B cells. Furthermore, differential mRNA stabilization is likely to contribute to the alteration of the gamma 1s/gamma 1m mRNA ratio at IL-6 induction.

Transcriptional regulation of immunoglobulin expression in a chronic lymphocytic leukemia cell line

A cell line derived from a B-type chronic lymphocytic leukemia, Corinna II, was found to down-regulate its steady-state level of IgM mRNA after treatment with phorbol 12-myristate 13-acetate while the proliferative capacity of the cell line was unaffected. No changes in the splicing pattern of the IgM transcript could be observed after treatment. The down-regulatory effect on IgM RNA expression was found to be inhibited by cycloheximide, suggesting that functional protein synthesis was needed for the effect. Transfection experiments showed that the down-regulatory effect of phorbol myristate acetate was exerted at the level of transcriptional initiation. The DNA element mediating the down-regulatory effect was found to be present within 140 bp 5' of the mRNA cap site in an immunoglobulin promoter.

Characterization of immunoglobulin mRNA expression in Burkitt lymphoma cell lines

Immunoglobulin heavy- and light-chain mRNA of 11 Burkitt lymphoma (BL) cell lines (9 African and 2 American) were analyzed for various structural characteristics. In agreement with previous results at the protein level, all the BL cell lines express heavy-chain mRNA transcripts of the mu class. Surprisingly, a high mu s/mu m mRNA ratio was found in 2 IgM-producing BL cell lines (Raji and CCL85), that do not secrete immunoglobulin. Variable region gene use was also assessed in the cell lines: while 4 out of 7 endemic BL cell lines use VH genes that belong to the VH3 gene family, no clear bias in the expression of particular VH or VL gene families among this sampling of BL lines was found. Northern blot analysis of immunoglobulin transcripts in endemic BL cell lines did show that 2 such lines (AG876 and HTB62) expressed truncated heavy-chain transcripts; RNA sequence analysis of the VH region demonstrated different abnormal 5'-localized RNA splicing events for the 2 shortened mu transcripts. The light-chain mRNA in these 2 cell lines also showed structural abnormalities and, in the case of HTB62, 3 different kappa light-chain transcripts are produced (of elongated, native and truncated sizes). In vitro translation of mRNA from HTB62 showed mu and kappa chain proteins corresponding with the relative size for each message.

Pulsed-field gel analysis of human immunoglobulin heavy-chain constant region gene deletions reveals the extent of unmapped regions within the locus

The human immunoglobulin heavy-chain constant region gene locus is organized in three main gene groups, the physical distances of which are unknown. Different types of gene deletions, originated by unequal crossingover, have been found to encompass one or more genes in the locus. We have analyzed some of these deletions by means of pulsed-field gel electrophoresis, which allows resolution of large DNA fragments. By identifying a fragment containing two of the main gene groups and by observing the size reduction of this fragment in subjects with deletions, we were able to estimate the distance between the two groups and better locate the pseudogene in-between.

Circulating monoclonal B lymphocytes in multiple myeloma

Peripheral blood mononuclear cells from 28 patients with multiple myeloma (MM) and nine patients with monoclonal gammopathy of unknown significance (MGUS) were studied by immunoglobulin gene analysis. Clonal immunoglobulin gene rearrangements in peripheral blood mononuclear cells (PBIGRA) were demonstrated in 10 of the 28 MM patients (36%). Bone marrow and peripheral blood mononuclear cells were studied simultaneously in five of these 10 patients, and identical gene rearrangements were demonstrated in both. The incidence of such gene arrangements was higher in patients with active disease (cases at presentation or relapsed = 10/19 [47%]) compared to remission status (0/9) and higher in untreated (47%) compared to treated patients (11%) (P less than 0.05). Patients with this phenomenon had higher serum calcium levels (P less than 0.001), and higher bone marrow plasma cell counts (P less than 0.05). Serum creatinine and beta 2-microglobulin were also higher but did not reach statistical significance. None of the patients with monoclonal gammopathy of uncertain significance had gene arrangements. Our findings confirm that circulating B lymphocytes are part of the malignant clone in MM and their presence correlates with high tumour volume.

Nucleotide sequence of Suncus murinus immunoglobulin mu gene and comparison with mouse and human mu genes

Both primates and rodents apparently originated from insectivores and then evolved separately. We isolated the immunoglobulin mu gene from DNA of the insectivore Suncus murinus and determined its nucleotide sequence. The gene organization was CH1 exon (318 bp)-intron (89 bp)-CH2 exon (345 bp)-intron (80 bp)-CH3 exon (318 bp)-intron (85 bp)-CH4 exon (392 bp)-coding sequence and 3'-untranslated region. Comparison of nucleotide sequences of mu genes between mouse, human and Suncus murinus indicated that the evolutionary distance between human and mouse is equal to that between Suncus murinus and human, and that mouse is deviated further from Suncus murinus than the two other combinations. This conclusion was further supported by sequence comparison of non-coding regions.

The 5' splice site: phylogenetic evolution and variable geometry of association with U1RNA

The 5' splice site sequences of 3294 introns from various organisms (1-672) were analyzed in order to determine the rules governing evolution of this sequence, which may shed light on the mechanism of cleavage at the exon-intron junction. The data indicate that, currently, in all organisms, a common sequence 1GUAAG6U and its derivatives are used as well as an additional sequence and its derivatives, which differ in metazoa (G/1GUgAG6U), lower eucaryotes (1GUAxG6U) and higher plants (AG/1GU3A). They all partly resemble the prototype sequence AG/1GUAAG6U whose 8 contigous nucleotides are complementary to the nucleotides 4-11 of U1RNA, which are perfectly conserved in the course of phylogenetic evolution. Detailed examination of the data shows that U1RNA can recognize different parts of 5' splice sites. As a rule, either prototype nucleotides at position -2 and -1 or at positions 4, 5 or 6 or at positions 3-4 are dispensable provided that the stability of the U1RNA-5' splice site hybrid is conserved. On the basis of frequency of sequences, the optimal size of the hybridizable region is 5-7 nucleotides. Thus, the cleavage at the exon-intron junction seems to imply, first, that the 5' splice site is recognized by U1RNA according to a "variable geometry" program; second, that the precise cleavage site is determined by the conserved sequence of U1RNA since it occurs exactly opposite to the junction between nucleotides C9 and C10 of U1RNA. The variable geometry of the U1RNA-5' splice site association provides flexibility to the system and allows diversification in the course of phylogenetic evolution.

Predominant immunoglobulin gene rearrangements in two patients with immunodeficiency: restricted use of V gene segments and DNA hypermethylation

The majority of common variable immunodeficiencies (CVID) is caused by intrinsic B cell defects which impede distinct stages of B cell differentiation. B cell differentiation is accompanied by the rearrangement of immunoglobulin (Ig) genes. The first step in the rearrangement process is the assembly of IgH genes, and subsequently, IgL genes are rearranged. During B cell maturation, Ig genes are demethylated in a stepwise, locus-specific manner. Here, we examined the Ig gene rearrangements of four patients with classical CVID and of one child suffering from an unusual immunodeficiency associated with CD5+ B cell lymphocytosis. In one of the four adult patients with CVID, we observed a predominant type of VHDJH-gene rearrangement. In the child, different polyclonal VHDJH-gene rearrangements were found together with a predominant type of kappa light chain gene rearrangement. The rearranged kappa chain genes were methylated (as in the pre-B cell stage). These findings together with the cell phenotype analysis and the clinical course of the disease in the child suggests that in some patients with primary immunodeficiency a maturation arrest may occur in B cells leading to a predominant Ig V gene rearrangement.

New types of multiple and single gene deletions in the human IgCH locus

The locus for human immunoglobulin heavy chain constant region genes (IgCH) is characterized by a significant frequency of deleted or duplicated haplotypes, due to unequal crossing-over events. Four types of deletions and one duplication have been reported so far. We describe here a molecular study of four cases of IgCH deletions. Two of the three types of deletions are reported here for the first time. Analysis of genetic markers associated with the deleted haplotypes pointed to the independent origin of similar deletions and the involvement of intergenic sequences in the mispairing-recombination process. The reduced or absent transcription of the C gamma 4 gene in two C gamma 2-deleted haplotypes offers an insight into the requirements for the isotype switch mechanism.

T cell receptor delta gene rearrangements in acute lymphoblastic leukemia

Using a newly isolated cDNA clone encoding the TCR-delta gene and genomic probes, we have analyzed T cell receptor (TCR) delta gene rearrangement in 19 patients with T cell acute lymphoblastic leukemia (T-ALL) and 29 patients with B-precursor ALL. Five out of seven CD3- T-ALL and 4 of 12 CD3+ T-ALL showed bi-allelic rearrangements of the TCR-delta gene. In three CD3+ patients, a single allelic TCR-delta gene rearrangement was observed with rearrangement of the TCR-alpha gene on the other allele. In five CD3+ patients with bi-allelic rearrangements of the TCR-alpha gene, the TCR-delta gene locus was deleted. Transcription of the TCR-delta gene was also analyzed in six T-ALL. Five patients expressed TCR-delta transcripts. Only one T-ALL, presumably derived from the most immature T lineage cells, did not have TCR-delta transcripts, but expressed TCR-gamma and 1.0-kb truncated TCR-beta transcripts. In B-precursor ALL, 20 patients (69%) showed rearrangements of the TCR-delta gene. The frequency of TCR-delta gene rearrangement was higher than TCR-alpha (59%), gamma (52%), or beta (31%) genes. These findings suggest that TCR-alpha gene rearrangements may take place after rearrangements of the TCR-delta gene with concomitant deletion of rearranged TCR-delta genes in T cell differentiation. Among leukemic cells of B lineage, the TCR-delta gene is the earliest rearranging TCR gene, followed by TCR-gamma and beta gene rearrangements.

Induction of differentiation in human leukemic B cells by interleukin 2 alone: differential effect on the expression of mu and J chain genes

The effects of interleukin 2 (IL2) on the proliferation and differentiation of B cells were analyzed separately using cells from two patients suffering from B-type chronic lymphocytic leukemia. The monoclonal B cells from these patients exhibited an opposite pattern of responsiveness upon in vitro culture with IL2 in the absence of other stimuli. In the first patient, IL2 alone was able to induce DNA synthesis and no Ig production. In the second patient, although no DNA synthesis was detected, B lymphocytes synthesized IgM upon stimulation with IL2 alone. Analysis of mRNA levels was performed on the cells of this latter patient after culture without or with IL2. In the presence of IL2 we observed a strong enhancement of C mu gene expression associated with an increase of the ratio between the secreted form and the membrane-bound form of mu mRNA. In contrast IL2 induced only a marginal enhancement of J chain mRNA. Thus, terminal B cell differentiation of selected monoclonal B cells can be obtained in the absence of DNA synthesis and IL2 alone can mediate this process. Moreover, IL2 can act at selective steps of the molecular events associated with IgM production. These results document the multiple effects of a given IL on the events leading to antibody production and strongly suggest that they can be conditioned by the maturation stage of a given responding cell.

An immunocytochemical method for the detection of fluorochrome-labelled DNA probes hybridized in situ with cellular RNA

Various detection systems for in situ hybridization of nucleic acids are currently used. We report here an immunocytochemical detection system which is based on the detection of FITC-labelled DNA/mRNA hybrids and takes advantage of FITC molecules attached covalently to the DNA probes prior to hybridization. In situ hybridization on cytocentrifuge spots is followed by the application of an anti-fluorescein antibody thus permitting detection of mRNA/DNA-FITC hybrids. The anti-FITC antibody reaction is demonstrated by an indirect immunocytochemical peroxidase-staining method. The T lymphoblast cell line Jurkat and the cDNA for the TCR-beta chain were chosen to establish the technique.

IgD/lambda plasmocytoma with immunoglobulin kappa light-chain genes in the germ-line configuration

Human immunoglobulin (Ig) genes are rearranged in an ordered sequence of events during B-cell differentiation: starting at the IgH locus, a productive VHDJH rearrangement leads to the expression of mu chains. Light-chain gene rearrangements have been found in pre-B cells which express mu chains. In these cells rearrangements of Ig kappa light-chain genes precede that of lambda genes. In an IgD/lambda-producing plasmocytoma, however, we found an apparent exception to this rule: the kappa genes were not rearranged. Together with the observation that roughly 90% of human IgD plasmocytomas produce lambda light-chain proteins, the finding reported here leads us to suggest that lambda light-chain genes are rearranged preferentially in IgD-producing plasma cells. Ig gene rearrangement, isotype switch, and the phenomenon of isotypic and allelic exclusion are discussed with special reference to our findings.

IgD in nasopharyngeal secretions and tonsils from otitis-prone children

Quantification of IgD was performed by ELISA in 180 plasma samples and 83 nasopharyngeal secretions (NPS) from children aged 2-162 months with varying degrees of recurrent otitis media. Furthermore, in 24 of the children the density of the IgD-immunocytes (IgD-cells) was calculated in immunoenzyme-stained cross-sections of their nasopharyngeal tonsils (NPT). Owing to a considerable variation of the IgD cell density throughout the NPT, a semi-quantitative counting system was applied. An irregular distribution of plasma IgD was observed during childhood and maximum levels were found between 48 and 72 months of age. However, an even distribution of plasma IgD was found among the four groups of children investigated. Based on calculations of the transudation of albumin from plasma to the NPS the amount of locally produced IgD in NPS (NPS-IgD (local] was estimated to 88% (range 41-99%). Significantly higher levels of NPS-IgD (local) were found in otitis-prone children than in the other groups. Moreover, a positive correlation was calculated between levels of NPS-IgD (local) and NPT-IgD cell density, indicating that NPT, being the local lymphoepithelial tissue, also functions as an important source of NPS-IgD. NPS-IgD was not found to be associated with secretory component, indicating a passive transfer of IgD through the mucosal membranes. Our results support the hypothesis of an association between the occurrence of IgD in the mucosa and secretions of the upper respiratory tract with localized inflammatory events.

Amino acid sequence of heavy chain from Xenopus laevis IgM deduced from cDNA sequence: implications for evolution of immunoglobulin domains

Present understanding of the evolution of immunoglobulins is derived almost entirely from studies of a few mammalian species. To obtain information about immunoglobulin genes in Xenopus laevis, a cDNA library was prepared in the expression vector lambda gt11 from mitogen-stimulated splenocytes of this species. Of approximately equal to 50,000 clones screened, 18 were found to express IgM epitopes. One of these, lambda XIg14, hybridized with RNA of RNA of approximately equal to 2 kilobases from splenocytes. The insert of this clone appears to encode a variable region and part of a mu constant region; that of another clone, lambda XIg8, appears to encode a variable region and a complete mu constant region. Both inserts contain sequence corresponding to the three gene segments (VH, DH, and JH) that encode heavy-chain variable regions. The heavy-chain constant region (CH) encoded by lambda XIg8 has the characteristic features of C mu, including a four-domain structure and a carboxyl-terminal tail. The amino acid sequences of two mu-chain peptides agree with the cDNA sequence. The identity in amino acid sequence between the corresponding Xenopus and mouse C mu domains ranges from 31 to 47%. The C mu domains vary in the extent to which their sequences resemble the sequences of other immunoglobulins, consistent with previous suggestions that the immunoglobulin domains have an independent evolutionary history.

The t(8;14) breakpoint of the EW 36 undifferentiated lymphoma cell line lies 5' of MYC in a region prone to involvement in endemic Burkitt's lymphomas

One of the best analyzed tumor-specific cytogenetic abnormalities is the t(8;14) chromosomal translocation observed in cases of Burkitt's and undifferentiated lymphomas (ULs), and acute lymphoblastic leukemias (ALLs). Here we analyze the cloned (8;14) chromosomal breakpoint of the UL cell line EW 36. We show that the region of chromosome 8 involved in the translocation is situated near a site previously demonstrated to harbor a cluster of endemic Burkitt's lymphoma breakpoints, approximately 50 kb 5' of MYC. In those cases, we demonstrated that malfunction of the V-D-J recombinase generated the translocations. However, in this case the isotype switch mechanism of translocation is implicated: at the breakpoint, S mu/S gamma and C gamma sequences are found on chromosome 14. Thus, the features of the EW 36 t(8;14) breakpoint are consonant with our model for B-cell lymphomagenesis which relates the precursor cell that gives rise to malignancy, the mechanism of translocation, and the phenotype of the tumor.

Quantitation of mu mRNA by in situ hybridization reveals a correlation between B-maturation associated antigens and IgM gene activation in acute lymphatic leukemias

Acute lymphatic leukemias, expressing the common acute leukemia antigen were investigated for expression of early T- and B-cell associated markers and the activation of the gene for the heavy chain of the immunoglobulin M. The gene activation as determined by quantification of the mu mRNA with a fluorochrome labeled gene probe in situ in individual cells showed a wide spectrum of positivity which was correlated to increasing expression of B-maturation markers. There was also a correlation between the amount of cellular immunoglobulin as determined by immunofluorimetry. Quantitation of mRNA in individual cells may allow to determine the degree of gene activation in antigenically defined individual cells and may thus contribute a new tool for classification of normal and malignant cells.

T-cell receptor beta-chain gene expression in B-lineage acute lymphoblastic leukemia

The T-cell receptor beta-chain (T beta) gene expression was examined in 16 children with B-lineage acute lymphoblastic leukemia (ALL), including eight patients with rearrangement of the T beta gene as well as immunoglobulin (Ig) heavy chain gene rearrangement. In contrast to the 1.3 kb full-length transcripts of the T beta observed in T-lineage leukemia and lymphoma cells, no transcript of the T beta gene was detected in 10 patients, including four with T beta gene rearrangement. Low levels of T beta transcripts were found in three patients with T beta gene rearrangement and two patients without T beta gene rearrangement, but those transcripts were truncated. In contrast to those findings, a single patient with T beta gene rearrangement showed abundant 1.3 kb T beta transcripts. These data indicate that T beta gene expression is not restricted to T-lineage cells and demonstrate the heterogeneity of B-lineage ALL at the expression level of the T beta gene. Our findings also suggest that T beta gene expression is not always accompanied with T beta gene rearrangement.

A new human immunoglobulin VH family preferentially rearranged in immature B-cell tumours

The prevalent forms of adult and childhood B-cell neoplasia are chronic lymphocytic (CLL) and acute lymphocytic (ALL) leukaemia, and are typified by a nearly monoclonal accumulation of cells expressing a single heavy (H) and light (L) chain variable (V) region. V gene selection could be random, or quite biased if the disease or the developmental status of the transformed cell somehow influenced DNA rearrangement. We have cloned and sequenced three germ-line VH gene segments that constitute a new human VH family (subgroup V) linked within 160 kilobase pairs of the DH-JH complex. One VH(V) member is rearranged in about 30% of patients with CLL and ALL, but not in IgM-expressing B-cell lines from peripheral blood. In some tumours, we detect a truncated (VH(V) RNA devoid of constant regions that originates from unrearranged VH(V) genes. In other tumours and in resting splenocytes, we detect large amounts of normally sized VH(V)-associated mRNA, although stimulation by mitogen of splenic B cells results in loss of VH(V)-hybridizing RNA. These features suggest that biased rearrangement of subgroup V may be under developmental selection.

T cell receptor alpha-chain gene rearrangements in B-precursor leukemia are in contrast to the findings in T cell acute lymphoblastic leukemia. Comparative study of T cell receptor gene rearrangement in childhood leukemia

We have analyzed T cell receptor alpha-chain gene configuration using three genomic joining (J) region probes in 64 children with acute lymphoblastic leukemia (ALL). 11 out of 18 T-ALLs were T3 positive; alpha-chain gene rearrangements were demonstrated in only two of 18, indicating that the majority of T-ALLs would have rearrangements involving J alpha segments located upstream of these probes. In contrast, 15 out of 46 B-precursor ALLs showed rearrangements of the alpha-chain gene and J alpha segments located approximately 20-30 kb upstream of the constant region were involved in 13 of these patients. Nine of 15 B-precursor ALLs with rearranged alpha-chain genes had rearrangements of both gamma- and beta-chain genes, whereas the remaining six had no rearrangements of gamma- and beta-chain genes. These findings indicated that alpha-chain gene rearrangement is not specific for T lineage cells and gamma- and/or beta-chain gene rearrangement does not appear essential for alpha-chain gene rearrangement, at least in B-precursor leukemic cells.

Isolation of genes encoding bovine IgM, IgG, IgA and IgE chains

Bovine C mu, C gamma, C alpha and C epsilon genes were cloned in an EMBL4 recombinant phage library using rabbit immunoglobulin switch mu (Su) and human C gamma as probes. Restriction mapping and Southern blot analyses of these clones identified one clone which hybridized with rabbit C mu and JH probes. The HG and C mu regions were separated by 6 kb of DNA. One C alpha and one C epsilon gene were found on overlapping clones and were separated by approximately 15 kb of DNA. Southern blot analysis of germline DNA with a bovine C alpha associated probe (S alpha) indicated that the germline contains a single C alpha gene. Similar analyses with a bovine C epsilon probe indicated that the germline contains either one C epsilon gene with allelic restriction polymorphism or two C epsilon genes. Three C gamma genes were cloned and did not overlap with one another. Southern blot analyses of germline DNA with a bovine C gamma probe indicated that the germline contains a total of four C gamma genes. The genes cloned correspond to three of the four genes identified by Southern blot analysis. The orientation of each CH gene was assigned by hybridization with S mu or S gamma probes. The S gamma probe hybridized to DNA immediately adjacent to all three C genes; the S probe hybridized to DNA immediately adjacent to the C mu, C alpha and C epsilon genes. Unexpectedly, the S mu probe also hybridized with a segment of DNA approximately 7 kb downstream of the C mu gene. This may represent a switch region for C gamma.