Several proteins associate with surface IgM to form the antigen receptor. We show that just two, the alpha and beta associated chains, are sufficient to reconstitute an IgM surface receptor in fibroblasts. Contrary to expectation, a common alpha chain associates with all five immunoglobulin classes. We propose that B-cell antigen receptors consist of a common alpha/beta heterodimer associated with each immunoglobulin class. But the classes differ both in the glycosylation of their associated alpha chain and in their dependence on alpha/beta for surface transport.

The mouse CTLA-4 gene has been shown to code for an activated lymphocyte-associated sequence belonging to the Ig gene superfamily. We now report on the molecular cloning and study of the human corresponding gene isolated from a genomic library and designated Hu-CTLA-4. The Hu-CTLA-4 gene exists as a single copy per human haploid genome and maps to band q33 of chromosome 2. It comprises 3 exons notwithstanding the leader sequence. The first exon encodes a V-like domain of 116 amino acids, the second one a hydrophobic putative transmembrane region of 37 amino acids and the third one a 34 amino acid putative cytoplasmic domain. Whereas the overall homology between the human and murine CTLA-4 proteins is 76%, there is, remarkably, a complete identity of their cytoplasmic domains. This complete interspecies conservation comes in support of an important role for this domain in CTLA-4 function.

A lymphoid-specific transcription enhancer element has recently been identified at the far 3' end of the rat immunoglobulin heavy chain (IgH) locus. Sequence analysis presented here reveals that this enhancer is flanked by a 350-bp invert repeat, giving a structure reminiscent of a transposable element. We therefore screened for the equivalent enhancer in the mouse to determine whether its presence was conserved during evolution. A mouse homologue was indeed identified and is located 16 kb downstream of the C alpha 1 exon. It is also flanked by invert repeats and these are not repeated throughout the genome. The mouse and rat enhancers retain high sequence homology. As regard activity, the IgH 3'-enhancer is lymphoid specific. However, this activity was detected in two plasmacytoma lines tested but not in two B cell lymphomas nor in HeLa cells suggesting that the enhancer may only play a stage-specific role during lymphocyte differentiation. As regards function within the IgH locus, we found that inclusion of the mouse IgH 3'-enhancer (in addition to the intron-enhancer) on mu gene expression plasmids effected a small increase in mu mRNA levels in stable plasmacytoma transfectants.

The locus of the human T cell rearranging gamma (TRG) or T cell receptor gamma chain genes comprises at least 14 variable genes (TRGV) belonging to four subgroups, five joining segments (TRGJ) and two constant region genes (TRGC). Nine V gamma genes belong to subgroup I, whereas subgroups II, III and IV each consists of a single gene respectively designated V9, V10 and V11. T cells expressing the gamma chain (TRG+) and recognized by the anti‐Ti gamma A monoclonal antibody have been shown to rearrange the V9 gene. In order to assess the N diversity at the V‐J junction in the TRG+ cells, the germline sequences of the segments involved in the V‐J rearrangements must be known. In this paper, we report the sequences of the germline V9 and V10 genes. Comparison of the V‐J junction and N region from transcripts or rearranged TRG genes belonging to the different subgroups shows no evidence of D segments in the human TRG locus. Sequences of the rearranged V11 gene from the JM cell line and those of the VA and VB pseudogenes, located upstream of V9 and V11 respectively, are given. Our results bring the number of human V gamma genes whose sequence is known to 13 and reveal unexpected homology with the mouse V gamma genes.

In the human T cell receptor gamma (TRG) locus, fourteen variable (TRGV) genes belonging to four subgroups have been identified upstream of two constant region (TRGC) genes. Three joining segments, JP1, JP and J1, have been localized upstream of TRGC1, and two others, JP2 and J2, upstream of TRGC2. In this report, we demonstrate that a unique Xho I fragment of 120 kilobases (kb) contains the fourteen TRGV genes and that the hybridization of that fragment in pulsed-field gel electrophoresis (PFGE) allows linkage of the variable region to the constant region locus. We also show that the variable and the constant regions are remarkably close to each other since the distance between V11, the most 3' V gamma gene, and JP1, the most 5' J gamma segment, is only 16 kb. With its 14 V gamma genes, spanning 100 kb, the two C gamma genes and 5 joining segments covering less than 40 kb and only 16 kb separating the most 3' V gene from the most 5' J segment, the human TRG locus spans 160 kb of genomic DNA and represents a particularly condensed locus compared to the other rearranging gene loci.

Six nonallelic immunoglobulin A constant region genes have been previously characterized on a 40-kilobase stretch of DNA. The nucleotide sequences of the three upstream genes of this cluster (Cj1, Cx2, Ck3) have been determined by other workers and shown to encode, respectively, the isotypic Mcg, Kern-Oz-, and Kern-Oz+ constant region of the A chains. In this paper, we report the sequence of the three downstream genes of this cluster and show that two of them (CA4 and Cx5) are pseudogenes. However, Ck6 encodes a Kern+Oz-chain and corresponds to the fourth isotype described among the A proteins sequenced so far. A potentially active J4 (oining) segment, with the canonical heptamer and nonamer sequences for rearrangement, is located 1.5 kilobases upstream of Cx6. The amino acid sequence encoded by the Cx6 gene is compared with the constant region sequences of various monoclonal Bence Jones A proteins. Allotypic and isotypic differences confirm the polymorphism and complexity of the human Ck locus.In humans, the constant (C) region of the immunoglobulin X light chains consists of at least four nonallelic or isotypic forms that differ by limited amino acid substitutions to produce the serological markers Kern (Ke) (1, 2), Oz (3-5), and Mcg (6, 7). Several additional substitutions have been described (8-16), but it is unknown whether these represent allelic variants or distinct isotypes. The human immunoglobulin X light chain genes have been mapped to chromosome 22 (17) at band q1l (18,19), and six nonallelic X C region genes (CxJ to Cx6) have been characterized on a 40-kilobase (kb) stretch of DNA (20). The number of CA genes varies between six and nine per haploid genome (21). These variations were detected by restriction fragment length polymorphism (21) and seem to have arisen from unequal meiotic crossing-over with a duplication of the CA2 and CA3 genes. Moreover, three additional CA-like genes have been recently identified, which map on different stretches of DNA and are nonallelic (22). One of these is a pseudogene, whereas the two others encode a putative A chain C region whose sequence differs from that of the X chains described so far.Only three QC genes (CA1, Cx2, and CA3) belonging to the cluster described by Hieter have been sequenced (20), and they have been shown to encode, respectively, the Mcg, Ke-Oz-, and Ke-Oz+ C region of the A chains. In this paper, we report the sequences* of the three genes located downstream in this cluster and show that two of them (CA4 and CA5) are pseudogenes, whereas CA6 encodes a Ke+Oz-chain, the fourth isotype described among the proteins sequenced so far. This CK6 gene has a potentially active JX6 joining region, with the canonical heptamer and nonamer sequences for rearrangement, 1.5 kb upstream of the coding C region. MATERIALS AND METHODSConstruction of a Phage Library from LY67 DNA. DNA prepared from LY67 cells (a X-producing Burkitt's lymphoma) (23)

No abstract

Intracellular expression of Ab fragments has been efficiently used to inactivate therapeutic targets, oncogene products, and to induce viral resistance in plants. Ab fragments expressed in the appropriate cell compartment may also help to elucidate the functions of a protein of interest. We report in this study the successful targeting of the protein tyrosine kinase Syk in the RBL-2H3 rat basophilic leukemia cell line. We isolated from a phage display library human single-chain variable fragments (scFv) directed against the portion of Syk containing the Src homology 2 domains and the linker region that separates them. Among them, two scFv named G4G11 and G4E4 exhibited the best binding to Syk in vivo in a yeast two-hybrid selection system. Stable transfectants of RBL-2H3 cells expressing cytosolic G4G11 and G4E4 were established. Immunoprecipitation experiments showed that intracellular G4G11 and G4E4 bind to Syk, but do not inhibit the activation of Syk following FcεRI aggregation, suggesting that the scFv do not affect the recruitment of Syk to the receptor. Nevertheless, FcεRI-mediated calcium mobilization and the release of inflammatory mediators are inhibited, and are consistent with a defect in Bruton’s tyrosine kinase and phospholipase C-γ2 tyrosine phosphorylation and activation. Interestingly, FcεRI-induced mitogen-activated protein kinase phosphorylation is not altered, suggesting that intracellular G4G11 and G4E4 do not prevent the coupling of Syk to the Ras pathway, but they selectively inhibit the pathway involving phospholipase C-γ2 activation.