Genomic structure and chromosomal localization of the human hepatocyte growth factor activator inhibitor type 1 and 2 genes.

Hepatocyte growth factor activator inhibitor type 1 (HAI-1) and type 2 (HAI-2) are recently discovered Kunitz-type serine protease inhibitors which can be purified and cloned from human stomach cancer cell line MKN45 as specific inhibitors against hepatocyte growth factor activator (HGFA). HAI-2 was identical with the protein originally reported as placental bikunin. Both proteins contain two Kunitz inhibitor domains (KDs), of which the first domain (KD1) is mainly responsible for the inhibitory activity against HGFA, and are expressed ubiquitously in various tissues. In this study, we cloned the genes coding for these two structurally similar proteins by screening of human genomic bacterial artificial chromosome (BAC) library and their genomic structures were compared. HAI-1 and -2 genes consist of 11 and 8 exons spanning 12 kbp and 12.5 kbp, respectively. Three exons were inserted between KD1 and KD2 of each gene, of which the middle one was the low-density lipoprotein (LDL) receptor-like domain (HAI-1) and the testis specific exon (HAI-2). Apparently homologous regions between HAI-1 and -2 were not found in 5'-flanking region and neither TATA nor CAAT box was present. The genes were mapped to chromosome 15q15 (HAI-1) and 19q13.11 (HAI-2). These results suggested that although HAI-1 and -2 genes might be derived from same ancestor gene, they acquired distinctive in vivo roles during their evolution.     

Genomic organization and identification of a novel alternative splicing variant of mouse mitochondrial thioredoxin reductase (TrxR2) gene

Eukaryotic mitochondria are equipped with a complete thioredoxin system, composed of thioredoxin and thioredoxin reductase, which has been implicated in the protection against the reactive oxygen intermdiates generated during the respiratory process in this organelle. Like its cytosolic counterpart, mammalian mitochondrial thioredoxin reductase is a homodimeric selenoprotein. We report here the genomic organization of the mouse mitochondrial thioredoxin gene (TrxR2) that spans 53 kb and consists of 18 exons ranging from 20 to 210 bp. All splicing sites conformed to the GT/AG rule with the exon-intron boundaries located exactly at the same position as the human TrxR2 gene, the only mammalian mitochondrial thioredoxin reductase gene whose genomic structure has been elucidated to date. In addition, we have identified a novel mRNA splicing variant lacking intron 14 resulting in a protein subunit with a shorter interface domain. This new splicing variant provides a framework for further analysis of this important enzyme as its predicted homodimeric conformation can now be expanded to a putative heterodimeric structure as well as a small subunit homodimer with the obvious implications at the regulatory level.     

Genomic sequence and structural organization of mouse slow skeletal muscle troponin T gene

Three muscle type-specific troponin T (TnT) genes are present in vertebrate to encode a number of protein isoforms via alternative mRNA splicing. While the genomic structures of cardiac and fast skeletal muscle TnT genes have been documented, this study cloned and characterized the slow skeletal muscle TnT (sTnT) gene. Complete nucleotide sequence and genomic organization revealed that the mouse sTnT gene spans 11.1kb and contains 14 exons, which is smaller and simpler than the fast skeletal muscle and cardiac TnT genes. Potentially representing a prototype of the TnT gene family, the 5'-region of the sTnT gene contains fewer unsplit large exons, among which two alternatively spliced exons are responsible for the NH2-terminal variation of three sTnT isoforms. The sTnT gene structure shows that the alternatively spliced central segment found in human sTnT cDNAs may be a result from splicing using an alternative acceptor site at the intron 11-exon 12 boundary. Together with the well-conserved protein structure, the highly specific expression of sTnT in slow skeletal muscles indicates a differentiated function of this member of the TnT gene family. The determination of genomic structure and alternative splicing pathways of sTnT gene lays a foundation to further understand the TnT structure-function evolution as well as contractile characteristics of different types of muscle fiber.     

Mapping and characterization of the mouse and human SS18 genes, two human SS18-like genes and a mouse Ss18 pseudogene.

We have previously isolated and characterized a mouse cDNA orthologous to the human synovial sarcoma associated SS18 (formerly named SSXT and SYT) cDNA. Here, we report the characterization of the genomic structure of the mouse Ss18 gene. Through in silico methods with sequence information contained in the public databases, we did the same for the human SS18 gene and two human SS18 homologous genes, SS18L1 and SS18L2. In addition, we identified a mouse Ss18 processed pseudogene and mapped it to chromosome 1, band A2-3. The mouse Ss18 gene, which is subject to extensive alternative splicing, is made up of 11 exons, spread out over approximately 45 kb of genomic sequence. The human SS18 gene is also composed of 11 exons with similar intron-exon boundaries, spreading out over about 70 kb of genomic sequence. One alternatively spliced exon, which is not included in the published SS18 cDNA, corresponds to a stretch of sequence which we previously identified in the mouse Ss18 cDNA. The human SS18L1 gene, which is also made up of 11 exons with similar intron-exon boundaries, was mapped to chromosome 20 band q13.3. The smaller SS18L2 gene, which is composed of three exons with similar boundaries as the first three exons of the other three genes, was mapped to chromosome 3 band p21. Through sequence and mutation analyses this gene could be excluded as a candidate gene for 3p21-associated renal cell cancer. In addition, we created a detailed BAC map around the human SS18 gene, placing it unequivocally between the CA-repeat marker AFMc014wf9 and the dihydrofolate reductase pseudogene DHFRP1. The next gene in this map, located distal to SS18, was found to be the TBP associated factor TAFII-105 (TAF2C2). Further analogies between the mouse Ss18 gene, the human SS18 gene and its two homologous genes were found in the putative promoter fragments. All four promoters resemble the promoters of housekeeping genes in that they are TATA-less and embedded in canonical CpG islands, thus explaining the high and widespread expression of the SS18 genes.     

Characterization of the mouse hnRNP A2/B1/B0 gene and identification of processed pseudogenes

The mouse hnRNP A2/B1/B0 gene has been cloned using a PCR-based strategy and sequenced. Analysis of this sequence showed that the gene organization closely follows that of the human orthologue with 12 exons and 11 introns. The hnRNP A2/B1/B0 gene gives rise to four splice variants through alternative splicing of exons 2 and 9. RT-PCR assays indicated that all splice variants were expressed in mouse brain, skin, and stomach tissues of varying ages, although their ratios to one another varied with age and tissue type. We also identified a small subset of all polyadenylated splice variants that included intron 11, which shows 94% sequence identity between human and mouse. Several processed pseudogenes were identified in the mouse genome. A search of the mouse genome databases located five pseudogenes, four of which are presumed to be non-functional because of the presence of premature stop codons, large deletions or rearrangements within the coding region. The fifth, which possesses putative promoter elements and has a coding sequence identical to that of the hnRNP A2 mRNA variant, may be functional.     

Isolation, characterization and mapping of the mouse and human PRG4 (proteoglycan 4) genes

PRG4 (proteoglycan 4) has been identified as megakaryocyte stimulating factor and articular superficial zone protein. PRG4 has characteristic motifs including somatomedin B and hemopexin domains, a chondroitin sulfate-attachment site and mucin-like repeats. During a screen of genes implicated in ectopic ossification, we found a novel mouse gene highly homologous to human and bovine PRG4 genes. Here, we report isolation, characterization and mapping of the gene, Prg4 together with characterization of its human orthologue. Prg4 cDNA was 3,320 bp long, encoding a 1,054 amino-acid protein. Human and mouse PRG4 genes each consisting of 12 exons spanned 18 and 16 kb, respectively. Characteristic motifs were conserved across species; however, the mucin-like repeat regions were highly diverse in length between species with a tendency that larger animals had longer repeats. Expression of human and mouse PRG4 genes was similar and found not only in cartilage, but also in liver, heart, lung, and bone. Expression of the mouse gene increased with progression of ectopic ossification. Multiple tissue-specific splicing variants lacking some of the motifs were found in both human and mouse. Although a specific role in the articular joint has previously been reported, the presence of multi-functional motifs as well as unique expression and alternative splicing patterns suggest that PRG4 functions in several distinctive biological process including regulation of ossification.     

Membrane and secretory forms of mouse membrane cofactor protein (CD46) generated from a single gene through alternative splicing

 A cDNA encoding a new secretory form of mouse membrane cofactor protein (MCP, CD46) was identified additionally to the membrane form cDNA. The secretory MCP, predicted from the cDNA sequence, consisted of the conserved four short consensus repeats (SCRs) plus a four amino acid-stretch. Unlike human MCP which comprises many isoforms, mouse MCP cDNA predicted a single isoform of membrane MCP with cytoplasmic tail 1 (CYT1) and serine/threonine-rich domain C (ST^C). To clarify the genomic origin and monomorphic alteration of these cDNAs, we cloned and analyzed a mouse genomic DNA harboring the full coding sequence of MCP from a 129/SV mouse genomic library. The mouse Mcp was a single gene ∼50 kilobases long. Eleven of the 14 coding exons of the human MCP gene and intron-exon boundary sequences were found to be conserved in the mouse gene. The ST^C homologue but not the ST^A or ST^B homologue in the mouse exons was functional: the latter being due to deletions and lack of consensus sequences for splicing. The sequence equivalent to cytoplasmic tail 2 (CYT2) has not been identified in the Mcp genome. Thus, the three exons (ST^A, ST^B, and probably CYT2) responsible for the polymorphism of human MCP by differential splicing were missing in the mouse Mcp gene. Unlike the case in humans, no Mcp-related genes or pseudogenes were observed in the mouse genome. The single mouse Mcp gene was mapped to the R-positive H5 band of mouse Chromosome 1 by FISH. Strikingly, one alternative exon with 73 base pairs (encoding the four new amino acids and a TGA stop codon) was discovered between the SCRIV and the ST^C exons; alternative splicing causes the generation of the secretory form of mouse MCP. These results on mouse MCP, together with the information concerning other mouse SCR proteins, infer that the regulator of complement activation (RCA) gene cluster is genetically diverged between humans and mice. Received: 22 April 1999 / Revised: 21 June 1999