Identification of the 45-base-long primordial building block of the entire class I major histocompatibility complex antigen gene.

Within the 885-base-long incomplete coding sequence for the mouse class I major histocompatibility complex (MHC) antigen H-2Kb heavy chain, we have identified 32 recurring base oligomers (1 decamer, 4 octamers, 9 heptamers, and 18 hexamers), the decamer being T-C-A-C-C-C-T-G-A-G. The compilation of these recurring base oligomers led to the identification of the 45-base-long primordial building block for this class of genes which is: C-C-T-G-C-G-G-A-C-A-A-T-G-T-C-A-C-C-C-T-G-A-G-C-T-G-C-T-G-G-G-C-T-C-T-G-G-G-C-G - T-T-G-A-C. Inasmuch as these base oligomers that recurred within the mouse H-2Kb coding sequence proper also recurred within the adjacent, transcribed noncoding sequence, we conclude that the entire approximately equal to 4,000-base-long germ-line DNA segment containing eight or nine separate exons of the coding sequence for class I MHC antigen heavy chains can be construed as having evolved from tandem repeats of this one 45-base-long primordial building block.     

The 48-base-long primordial building block of immunoglobulin light-chain variable regions is complementary to the primordial building block of heavy-chain variable regions.

The ancestral gene for immunoglobulin light-chain variable regions (Ig VLs) of the kappa as well as the lambda class apparently arose from about 12 tandem repeats of the 48-base-long primordial building block sequence TCT-TGC-GCA-GTA-AGT-CCA-CTC-CAG-GTC-ATA-TCC-AGT-CAG-GCT-GCT-GAA. Even today, amino acid residues 67 to 82 of each Ig V kappa L are still specified by a direct descendant in toto of the above-noted primordial building block, whereas amino acid residues 14 to 25 are invariably specified by its truncated copy. The Ig VL primordial building block presently identified is 100% complementary to the Ig VH (heavy-chain variable region) primordial building block previously identified. In the recognition of specific antigenic determinants by antibodies, Ig VL and Ig VH of light-chain--heavy-chain dimers have to complement each other. It is perhaps fitting that the primordial building blocks of the two are represented by the complementary strands of the same 48-base-pair-long DNA sequence.     

Concordant evolution of coding and noncoding regions of DNA made possible by the universal rule of TA/CG deficiency-TG/CT excess.

The universal rule of TA/CG deficiency-TG/CT excess previously proposed as the construction principle of coding sequences applies to noncoding regions of the gene as well. Analysis of a 1989-base-long gene sequence for mouse immunoglobulin gamma 2a heavy-chain constant region as well as the 19,002-base-long gene sequence for human serum albumin revealed deficiency and overabundance of very similar sets of base trimers and tetramers in the coding and noncoding regions of the same gene, in spite of the fact that noncoding regions were considerably richer in A + T. Inasmuch as this universal rule does not discriminate one strand of DNA double helix from another, two complementary DNA strands of the entire gene maintained nearly perfect symmetry. That is to say, the degrees of excesses, deficiencies of the 64-base trimers remained nearly identical between two complementary strands, and this symmetry was only slightly disturbed in the coding region. It would thus appear that the universal rule as an intrinsic force has been exerting far greater influence than natural selection in the evolution of genes.     

Mouse immunoglobulin coding sequences for the heavy-chain variable region arose as repeats of the two short building blocks.

The coding sequence for the 97-amino-acid-residue-long immunoglobulin heavy-chain variable (VH) regions of the mouse apparently arose as repeats of the two short building blocks. Three of the recognizable copies of the one 21-base-long prototype sequence A-C-T-G-G-A-T-A-T-G-A-C-C-T-G-G-A-G-T-G-G are invariably found to occupy the fixed positions within the 5' half of each VH coding sequence. Interestingly, the first and third copies specify the relatively invariant regions represented by the 7th to 13th and 41st to 47th amino acid residues (the first and second framework regions), whereas the second copy specifies the first hypervariable region (31st to 35th amino acid residues). These copies maintain at least 57.2% (12 out of 21) base sequence homology to the above-noted prototype building block. Base sequences of the other 14- to 15-base-long prototype building block differ from each other by as much as 60% between individual VHS. Yet one of its copies invariably occupies the terminal region of each VH coding sequence, thus specifying the very invariant third framework region. Other copies occupy unfixed positions in the VH and its attendant hydrophobic leader coding sequence as well as in adjacent noncoding sequences. The homology thus revealed between the VH coding sequence and its adjacent noncoding sequences suggests their concordant evolution.     

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.     

Illegitimate recombination generates a class switch from C mu to C delta in an IgD-secreting plasmacytoma.

We present here the sequence characterization of a C mu----C delta immunoglobulin (Ig) heavy chain class switch. In the murine IgD-secreting plasmacytoma TEPC 1017, deletion of most of the mu switch recombination region (S mu) and the entire C mu gene occurred in the absence of switch region sequences 5' to C delta. This unique rearrangement resulted from an illegitimate recombination of sequences with only patchy homology to each other. The infrequent and variable nature of illegitimate recombination may explain the low frequency of IgD-secreting plasma cells in normal mouse tissues.     

Organization and sequences of the diversity, joining, and constant region genes of the human T-cell receptor beta chain.

The organization and sequences of the human beta-chain T-cell receptor diversity, joining, and constant region segments are described. The beta chain of the human T-cell receptor, analogous to the mouse counterpart, consists of two distinct constant region genes approximately equal to 10 kilobases apart. The two constant region genes, C beta 1 and C beta 2, are very similar not only in sequence but also in genomic organization. The coding sequences of each of these C beta constant region genes are divided into four exons. The first two exons encode most of the extracellular constant domain. The third exon encodes a major part of the presumed transmembrane portion, and the last exon contains the cytoplasmic coding sequence as well as 3' untranslated sequences. Except for a stretch of approximately equal to 95 highly conserved nucleotides extending 3' of the first exon of the C region genes, little homology can be found between the intron sequences of C beta 1 and C beta 2. A small cluster of joining region (J beta) gene segments is located approximately equal to 5 kilobases upstream of each of these two constant regions. The first cluster, J beta 1, contains six functional J gene segments while the second, J beta 2, contains seven functional J gene segments. In addition, diversity region (D beta) gene segments are located approximately equal to 600 base pairs upstream of each J beta. Recombinational signals containing highly conserved heptamer and nonamer sequences separated by 12 or 23 bases are found adjacent to all of these D beta and J beta gene segments. These signal sequences are thought to be involved in the somatic recombination processes. These results indicate that what appears to be a gene duplication event giving rise to these two distinct regions must have arisen a long time ago in the evolution of this gene locus.     

Intervening sequences divide the gene for the constant region of mouse immunoglobulin mu chains into segments, each encoding a domain.

To elucidate the structure of the gene for the constant region of immunoglobulin mu chains, we have cloned a 9.9-kilobase-pair fragment of mouse DNA bearing a gene for the constant region of the mu chain (C mu gene) from an IgM-secreting mouse plasmacytoma. The sequence around this gene has apparently undergone somatic rearrangement; the gene occurs in an EcoRI restriction endonuclease fragment of a different size from that in embryo or liver DNA and no C mu-bearing fragment of embryo size remains in the plasmacytoma. The cloned sequence lacks a variable region gene; hence, if this C mu gene is active, its position within the clone indicates that the gene for the variable region of a heavy chain (VH gene) must be more than 3.7 kilobase pairs away. The C mu gene is divided by three intervening sequences into four coding segments, each of which encodes one of the domains (homology units) of the polypeptide. The nucleotide sequence coding for amino acids near the V-C junction is not present within the C mu clone or clones bearing homologous embryonic VH genes. This suggests that an immunoglobulin heavy chain, in common with light chains, is encoded not only by a V and C gene, but also by an independent joining region (JH) gene.     

Cloning and partial nucleotide sequence of human immunoglobulin mu chain cDNA from B cells and mouse-human hybridomas.

Purified mRNAs coding for mu and kappa human immunoglobulin polypeptides were translated in vitro and their products were characterized. The mu-specific mRNAs, derived from both human lymphoblastoid cells (GM607) and from a mouse-human somatic cell hybrid secreting human mu chains (alpha D5-H11-BC11), were copied into cDNAs and inserted into the plasmid pBR322. Several recombinant cDNAs that were obtained were identified by a combination of colony hybridization with labeled probes, in vitro translation of plasmid-selected mu mRNAs, and DNA nucleotide sequence determination. One recombinant DNA, for which the sequence has been partially determined, contains the codons for part of the C3 constant region domain through the carboxy-terminal piece (155 amino acids total) as well as the entire 3' noncoding sequence up to the poly(A) site of the human mu mRNA. The sequence A-A-U-A-A occurs 12 nucleotides prior to the poly(A) addition site in the human mu mRNA. Considerable sequence homology is observed in the mouse and human mu mRNA 3' coding and noncoding sequences.     

Separation of sequences defining basal expression from those conferring alpha gene recognition within the regulatory domains of herpes simplex virus 1 alpha genes.

The genes of herpes simplex virus 1 form three major groups--alpha, beta, and gamma--whose expression is coordinately regulated and sequentially ordered in a cascade fashion. To determine how the infected cell differentiates between these gene groups, alpha-regulated chimeric genes were constructed in earlier studies by fusing the structural sequences of the thymidine kinase (TK) gene, a beta gene, to the 5' noncoding sequences of alpha genes. These studies showed that (i) one or more structural components of the virion act in trans to increase alpha gene expression and (ii) the 5' noncoding sequences of alpha genes contain cis-acting domains that promote gene expression and confer alpha-gene regulation. These two domains could be moved independently, but the regulatory domain required a promoter for its function. We report here the properties of three sequences containing features common to the regulatory regions of all alpha genes. Sequence 1, containing (G + C)-rich inverted repeats, increased the basal level of TK expression when fused 5' to either the alpha gene 4 promoter or the truncated beta TK promoter. The effect was to some extent orientation dependent. Moreover, sequence 1 restored beta regulation to the truncated beta TK promoter but did not confer alpha-specific regulation on any of the chimeric genes tested. Sequences 2 (49 base pairs) and 3 (29 base pairs), containing an (A + T)-rich homolog from alpha gene 27 and alpha gene 0, respectively, restored alpha-specific regulation to the alpha promoter gene but only sequence 2 conferred alpha regulation on the truncated beta promoter gene. Our results indicate that (i) in natural beta TK the promoter and regulatory domains overlap, (ii) sequence 1 determines basal level of expression and substitutes for a promoter component that is essential for beta but not alpha regulation, and (iii) conversion of a gene with a promoter into an alpha gene requires two elements. Sequence 2 may contain both whereas sequence 3 contains only one.     

Use of synthetic oligonucleotides as hybridization probes: isolation of cloned cDNA sequences for human beta 2-microglobulin.

We have synthesized two sets of 15-base-long oligodeoxyribonucleotides corresponding to all possible coding sequences for a small portion of human beta 2-microglobulin. Labeled oligonucleotides were used as hybridization probes to screen bacterial clones containing cDNA sequences primed with oligo(dT) and inserted into the plasmid vector pBR322. One beta 2-microglobulin cDNA clone was detected in the 535 bacterial plasmid clones that were screened. The clone has been characterized by blotting and nucleotide sequence analysis. The cloned beta 2-microglobulin sequence contains 217 base pairs of the 3' untranslated region of the mRNA and 328 base pairs (97%) of the coding region.     

Identification of IgF, a hinge-region-containing Ig class, and IgD in Xenopus tropicalis

Only three Ig isotypes, IgM, IgX, and IgY, were previously known in amphibians. Here, we describe a heavy-chain isotype in Xenopus tropicalis, IgF (encoded by C(phi)), with only two constant region domains. IgF is similar to amphibian IgY in sequence, but the gene contains a hinge exon, making it the earliest example, in evolution, of an Ig isotype with a separately encoded genetic hinge. We also characterized a gene for the heavy chain of IgD, located immediately 3' of C(mu), that shares features with the C(delta) gene in fish and mammals. The latter gene contains eight constant-region-encoding exons and, unlike the chimeric splicing of muC(H)1 onto the IgD heavy chain in teleost fish, it is expressed as a unique IgD heavy chain. The IgH locus of X. tropicalis shows a 5' V(H)-D(H)-J(H)-C(mu)-C(delta)-C(chi)-C(upsilon)-C(phi) 3' organization, suggesting that the mammalian and amphibian Ig heavy-chain loci share a common ancestor.     

Somatic rearrangements forming active immunoglobulin mu genes in B and T lymphoid cell lines.

We have cloned an active gene for an immunoglobulin mu heavy (H) chain, bearing the variable (VH), joining (JH), and constant (C mu) sequences expressed in the IgM-secreting mouse plasmacytoma HPC-76. The mu gene was formed by somatic recombination between a VH gene and one of several JH genes, which are located about 7.7 kilobase pairs from the C mu gene in embryo DNA. The JH-C mu intervening sequence has suffered a deletion of about 2.7 kilobase pairs in HPC-76. Because the delection encompasses sequences required to switch an expressed VH-JH gene from C mu to another CH gene, it may represent a mechanism for "freezing" a lymphocyte clone at the stage of IgM expression. For the second (inactive) C mu allele in HPC-76, the entire joining and switch regions have been deleted; functional inactivation of one allele may thus represent one mechanism by which a lymphocyte clone restricts expression to a single allele (allelic exclusion). Probes generated from the cloned mu gene allowed examination of the JH locus in B, Abelson "pre-B," and T lymphoma cell lines and a myeloid line, all of which cotain RNA species bearing C mu sequences. The B and pre-B lines exhibited recombination within both alleles of the JH locus, suggesting that both alleles may be expressed in some cells. The absence of the JH gene 5' to the recombination sites favors a deletion mechanism for VH-JH joining. Recombination within the JH locus was also detected in two out of four T lymphoma lines, but not in the myeloid line. This indicates that the mechanism by which B cells generate immunoglobulin diversity is operational in some T cells. Lines that synthesize mu RNA without JH rearrangement may have activated the C mu gene directly or have undergone recombination at a more distant locus.     

Interferon-gamma gene expression in human B-cell lines: induction by interleukin-2, protein kinase C activators, and possible effect of hypomethylation on gene regulation.

Human interferon-gamma (IFN-gamma) is an important immunomodulatory protein produced predominantly by T cells and large granular lymphocytes (LGLs). Whereas large amounts of data have been accumulated regarding IFN gamma gene expression in these two cell types, little information about IFN gamma expression in other cell types exists. In this study, we have analyzed the production of IFN gamma by the Epstein-Barr virus (EBV)-positive B-cell line, JLP(c), derived from a patient with Burkitt's lymphoma, and another human B-cell line, PA682BM-1, which was derived from an acquired immunodeficiency syndrome patient. Southern blot analysis indicates the presence of an Ig heavy chain gene rearrangement, but no rearrangement of the T-cell receptor beta chain gene or IFN gamma gene in these B-cell lines. Both cell lines were found to express surface IgD and other B-cell surface markers, thus confirming their B-cell lineage. Analysis for surface Ig, cytoplasmic Ig, and secreted Ig indicates that the two cell lines are in relatively early stages of the B-cell differentiation pathway. We now report that PA682BM-1 can be triggered by the protein kinase C (PKC) activators, phorbol 12-myristate 13-acetate (PMA) and (-)Indolactam-v, to secrete IFN gamma, whereas JLP(c) cells spontaneously produce low levels of IFN gamma that can be enhanced by PKC activators and interleukin-2 (IL-2). After activation of the cell lines with IL-2, (-)Indolactam-v, and PMA, increases in cytoplasmic messenger RNAs (mRNAs) of IFN gamma and the IL-2 receptor chains were also observed. The induction of IFN gamma mRNA and protein by IL-2 was completely blocked by a monoclonal antibody to IL-2 receptor p75 (beta chain), but not by the monoclonal antibody to p55 (alpha chain). Analysis of IFN gamma genomic DNA indicates that the gene is not amplified, but that hypomethylation in the 5' noncoding region of the IFN gamma gene has occurred in the B-cell line from the Burkitt's lymphoma patient that spontaneously produces IFN gamma. This finding suggests that the methylation state of the promoter region may play an important role in the control of IFN gamma gene expression in B cells.     

Identification of the 48-base-long primordial building block sequence of mouse immunoglobulin variable region genes.

Mouse immunoglobulin heavy-chain variable region (Ig VH) genes apparently arose from the approximately 600-base-pair-long (approximately 12 tandem repeats of the 48-base-pair-long primordial building block sequence TTC-AGC-AGC-CTG-ACT-GGA-TAT-GAC-CTG-GAG-TGG-ACT-TAC-TGC-GCA-AGA) that in the original reading frame specified the amino acid sequence Phe-Ser-Ser-Leu-Thr-Gly-Tyr-Asp-Leu-Glu-Trp-Thr-Tyr-Cys-Ala-Arg. The previously identified, shorter prototype building blocks merely represented particular portions of the above primordial sequence. Even today, the direct descendant in toto of this primordial sequence specifies the last one-sixth of each VH coding sequence: the 83rd to 98th amino acid residues. Furthermore, its four truncated derivatives specify the 4th to 14th, 17th to 23rd, 29th to 37th, and 38th to 48th amino acid residues. Accordingly, all three relatively invariant--therefore, conserved--framework regions (FW-1, FW-2, and FW-3) of VHs are specified by recognizable--therefore, conserved--descendants of the primordial sequence.     

Identification of genes for the constant region of rabbit T-cell receptor beta chains.

We describe cDNA clones from thymus mRNA of a young rabbit that have sequences highly homologous to the human and murine T-cell receptor beta-chain constant region (C beta). In rabbit, man, and mouse there is a conserved extra cysteine in the constant region that could lead to a free thiol group or alternative disulfide bond formation depending on the locations and total numbers of cysteines in assembled receptor molecules. A cDNA clone (CL ANA 11) with 571 bases 5' of the C beta coding sequence has an open reading frame starting at a methionine codon that encodes 141 amino acids in frame with the C beta sequence. The encoded sequence has no resemblance to known immunoglobulin or beta-chain variable regions or other known proteins. An oligonucleotide probe from the 5' end of the encoded protein hybridizes to an approximately equal to 2-kilobase genomic DNA fragment that contains C beta gene sequences and to an approximately equal to 8-kilobase mRNA species in the thymus mRNA preparation from which the clone was derived. Within the 5' coding sequence there is a stretch of 211 bases containing strings of alternating purines and pyrimidines that may form Z-DNA. The sequence of the last 55 base pairs adjacent to C beta resembles the corresponding segment of mouse cDNA clone 86T3 that contains sequence 5' of the mouse C beta 1 gene. Although the function of a potential protein encoded by the 5' end of CL ANA 11 is unknown, it could play a role in regulation of thymocyte growth and differentiation.     

Induction of pre-B cell proliferation after de novo synthesis of the pre-B cell receptor

The assembly of a pre-B cell receptor (pre-BCR) composed of an Ig mu heavy chain (mu H-chain), the surrogate light (SL) chain, and the Ig alpha/beta dimer is critical for late pro-B cells to advance to the pre-B cell stage. By using a transgenic mouse model, in which mu H-chain synthesis is solely driven by a tetracycline-controlled transactivator, we show that de novo synthesis of mu H-chain in transgenic pro-B cells not only induces differentiation but also proliferation. This positive effect of mu H-chain synthesis on proliferation requires the presence of SL chain and costimulatory signals provided by stromal cells or IL-7. We conclude that pre-BCR signaling induces clonal expansion of early pre-B cells.     

Sequence and evolution of the human T-cell antigen receptor beta-chain genes.

We present the nucleotide sequences of the two genomic constant (C)-region gene segments, C beta 1 and C beta 2, encoding the beta chain of the human T-cell antigen receptor. The two C beta genes are organized identically to each other and to the corresponding mouse genes, both having four exons, whose boundaries were confirmed from the sequence of a C beta 2 cDNA clone from the T-cell line MOLT-4. The predicted amino acid sequences of human C beta 1 and C beta 2 differ at only five positions, which suggests that the proteins have very similar functions. This similarity is the result of strong nucleotide-sequence conservation in protein-coding regions, which extends to silent positions. A quantitative analysis of an alignment of the nucleotide sequences of the two human genes shows that whereas the 5' ends (including the first exon) are extremely homologous, the 3' ends are widely divergent, with other regions having intermediate levels of homology. Analysis of published data [Gascoigne, N.R.J., Chien, Y., Becker, D.M., Kavaler, J. & Davis, M.M. (1984) Nature (London) 310, 387-391] shows that the mouse C beta 1 and C beta 2 genes are also virtually identical in their first exons but more divergent in the remaining coding regions. Therefore, partial gene conversion events may have occurred during the evolution of both human and mouse C beta genes.     

The I region of the C57BL/10 mouse: characterization and physical linkage to H-2K of an SB beta-like class II pseudogene, psi A beta 3.

In the C57BL/10 mouse, 140 kilobases (kb) of the I region (I-Ab, I-Eb) were isolated as recombinant cosmids. The class II genes A beta 2, A beta 1, A alpha,E beta 1, E beta 2, and E alpha are located from centromere to telomere in a region of approximately equal to 110 kb, which shows that the I region in the b haplotype has a similar overall organization to those described for the d, k, and wr7 haplotypes. In addition to these genes, we have also isolated a class II gene, psi A beta 3, which is physically linked to the class I H-2K region, 75 kb telomeric to the H-2Kb gene. This orients the H-2K region on the genetic map with the H-2Kb gene being located toward the I region. The sequence of the beta 2 domain of psi A beta 3 is similar to the immunoglobulin-like domain of other class II genes. Interestingly, it shows 83% nucleotide homology to the human SB beta gene, the same homology that was seen previously between the immunoglobulin-like exons of A beta 1 and DC beta and between E beta 1 and DR beta, respectively. It is likely, therefore, that psi A beta 3 represents a member of a third SB-like class II gene family present in addition to I-A and I-E genes and that the divergence of the SB family predates the speciation of rodents and primates. Comparison of the DNA sequence of the exon encoding the beta 2 domain of psi A beta 3 in the b or k haplotypes with functional class II genes shows that a deletion of eight nucleotides has occurred, such that the psi A beta 3 sequence cannot be translated into a functional class II protein. This suggests that psi A beta 3 is a pseudogene.