Critical test of a sister chromatid exchange model for the immunoglobulin heavy-chain class switch

B lymphocytes may switch from producing an immunoglobulin heavy chain of the mu class to that of the gamma, epsilon or alpha class. To maintain the specificity, the new heavy chain must keep the original variable (V) region; this is achieved by deleting DNA sequences so that the V (consisting of joined VH, diversity (DH) and joining (JH) gene segments) and C (constant) gene segments coding for the new heavy chain are brought into close proximity (reviewed in ref. 5; we do not consider here the mu-delta situation). There are, in principle, three types of chromosomal rearrangements that yield a deletion: rearrangement within a chromatid; unequal sister chromatid exchange (as suggested by Obata et al.); and unequal recombination between chromosomal homologues. We have analysed the arrangement of C mu DNA in clones of the pre-B-cell line 18-81 that switches in vitro from mu to gamma 2b. The clones examined produce either mu, gamma 2b or no immunoglobulin chain. We report here that all the gamma 2b clones had lost at least one copy of C mu and no clones contained three copies of C mu. These findings formally exclude both unequal sister chromatid exchange and recombination between homologues as mechanisms for creating a gene encoding the gamma 2b chain.     

Immunoglobulin Cmu RNA in T lymphoma cells is not translated

It is widely believed that immunoglobulin genes might encode at least part of the receptor for antigen on the T lymphocyte. Evidence supporting this comes from the effects of anti-immunoglobulin idiotype antibodies on cellular immune networks and from the presence of idiotypes on immunologically active factors from T cells. Detailed molecular characterization of the receptors, however, has been seriously hampered by the lack of a suitable cellular source from which it might be isolated. The recent demonstration of Kemp et al. that thymocytes and certain cultured lines of mouse T lymphoma cells contain polyadenylated RNA molecules encoded by the immunoglobulin Cmu gene (Cmu RNA) prompted us to identify the corresponding protein molecules in those cells. As the haploid mouse genome contains a single Cmu gene, any polypeptide encoded by this gene should react with at least some of the antibodies present in rabbit anti-mouse IgM antiserum. In this letter we report that a number of T lymphoma lines, regardless of whether they contain Cmu RNA, synthesize no detectable mu polypeptides.     

The immunoglobulin mu constant region gene is expressed in mouse thymocytes

It has been a matter of controversy whether the functional capacity of T cells to discriminate between antigens is mediated via immunoglobulin, an immunoglobulin-like molecule, or by the product(s) of unrelated genes. The progenitors of immunoglobulin-secreting cells, B cells, express membrane-bound immunoglobulin as the antigen-specific receptor on their surface. For T cells, although products of immunoglobulin heavy chain variable region genes are implicated as receptor components, there has been no compelling immunochemical evidence for participation of either immunoglobulin light chains or heavy chain constant regions (see refs 2-6 for the disparate views). Recently, using cloned immunoglobulin DNA sequences as hybridization probes, we have demonstrated that the immunoglobulin Cmu gene, but not the Cmu gene, is expressed as polyadenylated RNA in some T cell tumour (T lymphoma) cell lines. Individual T lymphoma lines yielded up to three discrete mu RNA species of different sizes (1.9, 2.2 and 3.0 kilobases), each species being different in size from the major mu RNA species present in B lymphoma cells (2.4 and 2.7 kilobases). We show here that cells from the normal mouse thymus contain mu RNA species, indistinguishable in size from those in T lymphoma cells, but contain little if any kappa RNA.     

Unusual sequences in the murine immunoglobulin mu-delta heavy-chain region

The delta heavy (H) chain of mouse immunoglobulin D (IgD) is unusual both in its structure and in its differential expression relative to immunoglobulin M (IgM; reviewed in ref. 1). The region of DNA between IgM and IgD H-chain constant-region genes is probably implicated in this control. So far only fragments of the area have been sequenced. Now, however, we present the complete sequence as well as the sequence of the introns of the C delta gene. We have found several interesting features (Fig. 1), including an open reading frame (ORF) between Cmu and C delta which encodes 146 amino acids that might represent a previously unsuspected domain-like protein; three blocks of simple repetitive sequences; a 162-base pair (bp) unique-sequence inverted repeat; and a domain-like pseudogene in the large intron of C delta. We have not found, however, any sequence 5' of C delta resembling the switch (S) recombination sequences associated with class switching in other heavy chains. Moreover, we have determined the 3' deletion end point of an IgD-producing myeloma and find no sequences reminiscent of switch sites nearby.     

Retroviral gag and DNA endonuclease coding sequences in IgE-binding factor gene

Immunoglobulin-binding factors are known to regulate the synthesis of B-cell-derived immunoglobulin heavy-chain isotypes. Cloning and nucleotide sequence determination of complementary DNA encoding rodent IgE-binding factors (IgE-BF) revealed that messenger RNA encodes a glycoprotein of 557 amino acids which is expressed as a precursor of relative molecular mass (Mr) 60,000 (60K) in COS7 monkey cells. We report here that the 3' two-thirds of the IgE-BF coding sequence shows a surprising homology (72%) at the DNA level with coding sequences of the gag and pol (DNA endonuclease) genes of the Syrian hamster intracisternal A particle (IAP H18), an endogenous retrovirus. This marked homology demonstrates that the rodent gene encoding IgE-BF is a hybrid gene which evolved very recently by integrating genes of viral origin, and that the encoded polypeptide comprises three separate domains: an IgE-BF domain and retrovirus-derived gag and DNA endonuclease-like domains. This may represent the first report of a cellular gene containing a virus-derived coding sequence.     

Tissue-specific expression of three distinct types of rabbit protein kinase C

We examined the structure of protein kinase C in an attempt to understand the molecular events connecting protein kinase C activation with the cellular response. Rabbit complementary DNA clones coding for three distinct types of protein kinase C, named alpha, beta and gamma, have been identified and sequenced. The deduced amino acid sequence for alpha, beta and gamma (673, 671 and 672 amino acids, respectively) are closely related. Kinases alpha and beta share an identical N-terminal sequence of 621 amino acid residues and their messenger RNAs arise from a single gene. The C-terminal halves of alpha, beta and gamma are protein kinase domains and are highly homologous to other protein kinases. The mRNAs for alpha, beta and gamma are expressed in various tissues with strikingly different tissue specificities. The one for gamma is found ubiquitously among various tissues, while those for alpha and beta predominate in the brain.     

IgH class switching and translocations use a robust non-classical end-joining pathway

Immunoglobulin variable region exons are assembled in developing B cells by V(D)J recombination. Once mature, these cells undergo class-switch recombination (CSR) when activated by antigen. CSR changes the heavy chain constant region exons (Ch) expressed with a given variable region exon from Cmu to a downstream Ch (for example, Cgamma, Cepsilon or Calpha), thereby switching expression from IgM to IgG, IgE or IgA. Both V(D)J recombination and CSR involve the introduction of DNA double-strand breaks and their repair by means of end joining. For CSR, double-strand breaks are introduced into switch regions that flank Cmu and a downstream Ch, followed by fusion of the broken switch regions. In mammalian cells, the 'classical' non-homologous end joining (C-NHEJ) pathway repairs both general DNA double-strand breaks and programmed double-strand breaks generated by V(D)J recombination. C-NHEJ, as observed during V(D)J recombination, joins ends that lack homology to form 'direct' joins, and also joins ends with several base-pair homologies to form microhomology joins. CSR joins also display direct and microhomology joins, and CSR has been suggested to use C-NHEJ. Xrcc4 and DNA ligase IV (Lig4), which cooperatively catalyse the ligation step of C-NHEJ, are the most specific C-NHEJ factors; they are absolutely required for V(D)J recombination and have no known functions other than C-NHEJ. Here we assess whether C-NHEJ is also critical for CSR by assaying CSR in Xrcc4- or Lig4-deficient mouse B cells. C-NHEJ indeed catalyses CSR joins, because C-NHEJ-deficient B cells had decreased CSR and substantial levels of IgH locus (immunoglobulin heavy chain, encoded by Igh) chromosomal breaks. However, an alternative end-joining pathway, which is markedly biased towards microhomology joins, supports CSR at unexpectedly robust levels in C-NHEJ-deficient B cells. In the absence of C-NHEJ, this alternative end-joining pathway also frequently joins Igh locus breaks to other chromosomes to generate translocations.     

Domains and the hinge region of an immunoglobulin heavy chain are encoded in separate DNA segments.

A 6.8-kilobase DNA fragment containing the sequence coding for the constant region of the mouse immunoglobulin gamma1 heavy chain was cloned from total cellular DNA. Electron microscopic and nucleotide sequencing studies showed that the three protein domains and the hinge region are encoded in separate DNA segments.     

A novel type of aberrant recombination in immunoglobulin genes and its implications for V-J joining mechanism

Functional kappa light chain genes are formed during B-lymphocyte differentiation by the joining of initially separate V and J gene segments. It has been suggested that the intervening DNA is deleted, however the recent reports of what appear to be the reciprocal products of V and J recombination (back-to-back conserved V and J flanking sequences, called f-fragments) in DNA from mature lymphocytes make a simple deletion model unlikely. An alternative scheme involving unequal sister chromatid exchange has been proposed, supported by the evidence that the f-fragments seem to have segregated from the chromosome carrying the reciprocal complete kappa light chain gene (this and other schemes are briefly reviewed in ref. 8). We report here the analysis of a mouse myeloma (MOPC 41), in which a productive (kappa+) and a non-productive (kappa-) rearrangement has occured, which may help to clarify the mechanism of V-J joining. The aberrant rearrangement has led to the joining of a J1 gene segment to a sequence unrelated to any V gene (L10), and which in the germ line is flanked by a sequence resembling a V region recombination signal sequence. In this case no segregation of the reciprocal recombination products (kappa-41 and f41), which is a required step in sister chromatid exchange models, has taken place. An inversion model provides the simplest explanation of this J rearrangement.     

Circular DNA is a product of the immunoglobulin class switch rearrangement

The class of immunoglobulin is defined by the constant region of its heavy chain. When a B lymphocyte switches the class of heavy chain it produces, the constant region of mu-type heavy chain is replaced; this occurs through a DNA rearrangement that brings the gene segment encoding the new constant region close to the VDJ segment encoding the variable region. The pre-B-cell line 18-81, which switches from heavy chain mu to gamma 2b production in culture, occasionally abnormally rearranges the heavy chain locus so that DNA sequences between the switch regions of mu and gamma 2b are inverted. Because looping-out is an intermediate step in generating an inversion, the switch rearrangement could occur by looping-out and deletion. Provided that recombination is reciprocal, this would produce a circle of DNA. Indeed, circular DNA molecules have been isolated as products of rearrangement among gene segments encoding the variable regions of the T-cell receptor and of the immunoglobulin heavy chain and light chain. But whereas the breakpoints for the variable region rearrangement are precisely defined, the breakpoints for any given heavy chain class switch are scattered over a length of greater than 6 kilobases, including both switch regions. We have now isolated circular DNA containing the sequences deleted by class-switching, thereby showing that the immunoglobulin heavy chain class switch occurs through looping-out and deletion.     

Recombination between an expressed immunoglobulin heavy-chain gene and a germline variable gene segment in a Ly 1+ B-cell lymphoma

The early stages of murine B-cell differentiation are characterized by a series of immunoglobulin gene rearrangements which are required for the assembly of heavy(H) and light(L)-chain variable regions from germline gene segments. Rearrangement at the heavy-chain locus is initiated first and consists of the joining of a diversity (DH) gene segment to a joining (JH) gene segment. This forms a DJH intermediate to which a variable (VH) gene segment is subsequently added. Light-chain gene rearrangement follows and consists of the joining of a VL gene segment to a JL gene segment: once a productive light-chain gene has been formed the cell initiates synthesis of surface immunoglobulin M (sIgM) receptors (reviewed in ref. 1). These receptors are clonally distributed and may undergo further diversification either by somatic mutation or possibly by continued recombinational events. Such recombinational events have been detected in the Ly 1+ B-cell lymphoma NFS-5, which has been shown to rearrange both lambda and H-chain genes subsequent to the formation of sIgM (mu kappa) molecules. Here we have analysed a rearrangement of the productive allele of NFS-5 and found that it is due to a novel recombination event between VH genes which results in the replacement of most or all of the coding sequence of the initial VHQ52 rearrangement by a germline VH7183 gene. Embedded in the VH coding sequence close to the site of the cross-over is the sequence 5' TACTGTG 3', which is identical to the signal heptamer found 5' of many DH gene segments. This embedded heptamer is conserved in over 70% of known VH genes. We suggest that this heptamer mediates VH gene replacement and may play an important part in the development of the antibody repertoire.     

Identification of reciprocal translocation sites within the c-myc oncogene and immunoglobulin mu locus in a Burkitt lymphoma

The association between certain human tumours and characteristic chromosomal abnormalities has led to the hypothesis that specific cellular oncogenes may be involved and consequently 'activated' in these genetic recombinations. This hypothesis has found strong support in the recent findings that some cellular homologues of retroviral onc genes are located in chromosomal segments which are affected by specific tumour-related abnormalities (see ref. 4 for review). In the case of human undifferentiated B-cell lymphoma (UBL) and mouse plasmacytomas, cytogenetic and chromosomal mapping data have identified characteristic chromosomal recombinations directly involving different immunoglobulin genes and the c-myc oncogene (for review see refs 5, 6). In UBLs carrying the t(8:14) translocation it has been shown that the human c-myc gene is located on the region of chromosome 8 (8q24) which is translocated to the immunoglobulin heavy-chain locus (IHC) on chromosome 14. Although it is known that the chromosomal breakpoints can be variably located within or outside the c-myc locus and within the IHC mu (refs 9, 11) or IHC gamma locus, the recombination sites have not been exactly identified and mapped in relation to the functional domains of these loci. We report here the identification and characterization of two reciprocal recombination sites between c-myc and IHC mu in a Burkitt lymphoma. Nucleotide sequencing of the cross-over point joining chromosomes 8 and 14 on chromosome 14q--shows that the onc gene is interrupted within its first intron and joined to the heavy-chain mu switch region. This recombination predicts that the translocated onc gene would code for a rearranged mRNA but a normal c-myc polypeptide.     

A B cell-deficient mouse by targeted disruption of the membrane exon of the immunoglobulin mu chain gene

Of the various classes of antibodies that B lymphocytes can produce, class M (IgM) is the first to be expressed on the membrane of the developing cells. Pre-B cells, the precursors of B-lymphocytes, produce the heavy chain of IgM (mu chain), but not light chains. Recent data suggest that pre-B cells express mu chains on the membrane together with the 'surrogate' light chains lambda 5 and V pre B (refs 2-7). This complex could control pre-B-cell differentiation, in particular the rearrangement of the light-chain genes. We have now assessed the importance of the membrane form of the mu chain in B-cell development by generating mice lacking this chain. We disrupted one of the membrane exons of the gene encoding the mu-chain constant region by gene targeting in mouse embryonic stem cells. From these cells we derived mice heterozygous or homozygous for the mutation. B-cell development in the heterozygous mice seemed to be normal, but in homozygous animals B cells were absent, their development already being arrested at the stage of pre-B-cell maturation.     

Molecular components of the B-cell antigen receptor complex of the IgM class

The antigen receptors on mature B lymphocytes are membrane-bound immunoglobulins of the IgM and IgD classes whose cross-linking by polyvalent antigens results in B-cell proliferation and differentiation. How these membrane-bound immunoglobulin chains, which lack a cytoplasmic tail, generate a cell activation signal is not at present known. We now show that the IgM molecule is non-covalently associated in the membrane of B cells with two proteins of relative molecular mass 34,000 (Mr 34 K; IgM-alpha) and 39 K (Ig-beta) which form a disulphide-linked heterodimer. Surface expression of IgM seems to require the formation of an appropriate complex between IgM and the heterodimer. A transfection experiment indicates that IgM-alpha is the product of mb-1, a B-cell specific gene encoding a transmembrane protein with sequence homology to proteins of the T-cell antigen receptor-CD3 complex.     

Characterization of 64-, 123- and 182-base-pair exons in the mouse alpha 2(IV) collagen gene

Genes encoding types I, II and III collagens (fibrillar collagens) contain many discrete-size exons, most of them 54 base pairs (bp) long, in addition to the 45-, 99-, 108- and 162-bp exons. It has been suggested that these collagen genes evolved from an ancestral coding unit of 54 bp. Type IV collagen is a specific component of basement membranes and contains two genetically distinct polypeptides, the alpha 1(IV) and alpha 2(IV) chains. It differs from the types I-III collagens in that it contains interruptions in the Gly-X-Y repeat sequence and does not form ordered fibrillar structures. We have isolated complementary DNA and genomic clones for the mouse alpha 2(IV) collagen chain and here characterize 64-, 123- and 182-bp exons in the Gly-X-Y coding domain of the gene. The data suggest that the alpha 2(IV) collagen gene may have evolved differently from those encoding the fibrillar collagens.     

Homology of 54K protein of signal-recognition particle, docking protein and two E. coli proteins with putative GTP-binding domains

Most proteins exported from mammalian cells contain a signal sequence which mediates targeting to and insertion into the membrane of the endoplasmic reticulum (ER). Involved in this process are the signal-recognition particle (SRP) and docking protein (DP), the receptor for SRP in the ER membrane. SRP interacts with the signal sequence on nascent polypeptide chains and retards their further elongation, which resumes only after interaction of the arrested ribosomal complex with the docking protein. SRP is a ribonucleoprotein particle comprising a 7S RNA and six polypeptides with relative molecular masses (Mr) of 9,000 (9K) 14K, 19K, 54K, 68K and 72K (ref. 1). The 9K and 14K proteins are essential for elongation arrest and the 68K-72K heterodimer is required for docking to the ER membrane. The 54K protein binds to the signal sequence when it emerges from the ribosome. Docking protein consists of two polypeptides, a 72K alpha-subunit (DP alpha) and a 30K beta-subunit (DP beta). No components structurally homologous to SRP and docking protein have yet been found in yeast or Escherichia coli. To understand the molecular nature of the interaction between the signal sequence and its receptor(s) we have characterized a complementary DNA coding for the 54K protein of SRP. Significant sequence homology was found to part of DP alpha and two E. coli proteins of unknown function. The homologous region includes a putative GTP-binding domain.