関 直彦

The Down syndrome [DS) region has been defined by analyses of partial trisomy 21. The 2.5-Mb region between D21S17 and ERG is reportedly responsible for the main features of DS. Within this 2.5-Mb region, we focused previously on a distal 1.6-Mb region from an analysis of Japanese DS patients with partial trisomy 21. Previously we also performed exon-trapping and direct cDNA library screening of a fetal brain cDNA library and identified a novel gene TPRD. Further screening of a fetal heart cDNA library was performed and a total of 44 possible exons and 97 cDNA clones were obtained and mapped on a BamHI map. By rescreening other cDNA libraries and a RACE. reaction, we isolated nearly full-length cDNAs of three additional genes [holocarboxylase synthetase (HCS), G protein-coupled inward rectifier potassium channel 2 (GIRK2), and a human homolog of Drosophila minibrain gene (MNB)] and a coding sequence of a novel inward rectifier potassium channel-like gene (IRKK). The gene distribution and direction of transcription were determined by mapping both ends of the cDNA sequences. We found that these genes, except IRKK, are expressed ubiquitously and are relatively large, extending from 100 kb to 300 kb on the genome. These nearly full-length cDNA sequences should facilitate understanding of the detailed genome structure of the DS region and help to elucidate their role in the etiology of DS.

Chromosomal locations of the Atm (ataxia-telangiectasia (AT)-mutated) and Acat1 (mitochondrial acetoacetyl-CoA thiolase) genes in mouse, rat, and Syrian hamster were determined by direct R-banding FISH. Both genes were colocalized to the C-D band of mouse chromosome 9, the proximal end of q24.1 of rat chromosome 8, and qa4-qa5 of Syrian hamster chromosome 12. The regions in the mouse and rat were homologous to human chromosome 11q. Fine genetic linkage mapping of the mouse AT region was performed using the interspecific backcross mice. Atm, Acat1, and Npat, which is a new gene isolated from the AT region, and 12 flanking microsatellite DNA markers were examined. No recombinations were found among the Atm, Npat, Acat1, and D9Mit6 loci, and these loci were mapped 2.0 cM distal to D9Mit99 and 1.3 cM proximal to D9Mit102. Comparison of the linkage map of mouse chromosome 9 (MMU9) and that of human chromosome 11 (HSA11) indicates that there is a chromosomal rearrangement due to an inversion between Ets1 and Atm-Npat-Acat1 and that the inversion of MMU9 originated from the chromosomal breakage at the boundary between Gria4 and Atm-Npat-Acat1 on HSA11. This type of inversion appeared to be conserved in the three rodent species, mouse, rat, and Syrian hamster, using additional comparative mapping data with the Rck gene. (C) 1996 Academic Press, Inc.

Ataxia telangiectasia (AT) is an autosomal recessive disease of unknown etiology associated with cerebellar ataxia, oculocutaneous telangiectasia, immunodeficiency, and hypersensitivity to ionizing radiation. Although AT has been divided into four complementation groups by its radioresistant-DNA synthesis phenotype, the ATM gene has been isolated as the candidate gene responsible for all AT groups. We identified a new gene, designated NPAT, from the major AT locus on human chromosome 11q22-q23. The gene encoded a 1421-amino-acid protein containing nuclear localization signals and phosphorylation target sites by cyclin-dependent protein kinases associated with E2F. The messenger RNA of NPAT was detected in all human tissues examined, and its genomic sequence was strongly conserved through eukaryotes, suggesting that the NPAT gene may be essential for cell maintenance and may be a member of the housekeeping genes. Analysis of the genomic region of NPAT surprisingly revealed that the gene existed only 0.5 kb apart from the 5' end of the ATM transcript with opposite transcriptional direction. It may be possible to propose the idea that the promoter region could be shared by both housekeeping genes and that each gene could influence the expression of the other.

The bfp, a member of the RING finger family, has been shown to be predominantly expressed in brain and up-regulated in neural differentiation of P19 embryonic carcinoma cells, Chromosome mapping of the bfp gene by fluorescence in situ hybridization reveals that human BFP (ZNF179) is located at 1.7p1.1.2, mouse Bfp at 11B1.3, and rat BFP at 1.0q22, These results provide additional evidence that the mouse 11B region displays conserved linkage homology with the 17p11.2 region of the human genome and the 10q22 region of the rat genome. (C) 1996 Academic Press, Inc.

Ataxia telangiectasia (AT) is an autosomal recessive disease and AT cells exhibit chromosomal instability, hypersensitivity to the killing effect of ionizing radiation, and defective G2/M cell cycle checkpoint. Very recently, the candidate gene for AT, ATM, has been identified by Savitsky et al. (Science 268s1749-1753, 1995). To analyze AT locus further, we constructed a long-range physical map of YAC clones spanning the AT region and identified two new house keeping genes by screening cDNA libraries with the YACs. Analysis of genomic structure showed that one of the new genes links closely to the ATM gene by head to head array. In order to clarify the role of the genes mapped in AT locus, we are attempting to analyze function of these genes as well as to search mutations of the genes in AT patients.

Eps8 is a recently identified SH3-containing substrate for tyrosine kinase receptors. To understand the role of eps8 in receptor-mediated signaling, we cloned cDNAs encoding proteins that bind to its SH3 domain. One of these cDNAs predicts the synthesis of an 828 amino acid protein with homology to the N-terminal region of the tre oncogene. We designated this protein RN-tre for Related to the N-terminus of tre. RN-tre is ubiquitously expressed and maps to 10p13, a region known to be involved in translocations in various leukemias. In addition, a 10p13 monosomy syndrome, characterized by developmental alterations, has been reported. The regional homology between RN-tre and tre, which is limited to their N-terminal portion, prompted us to investigate the origin of the tre oncogene transcriptional unit. We were able to show that tre is the fusion product of a 5' genetic element, homologous to RN-tre and a 3' element, encoding a de-ubiquinating enzyme. Moreover, we identified, within the N-terminus of RN-tre and tre, a domain (named TrH, for Tre Homology), which is conserved within several proteins from yeast to mammals and has protein-binding properties in vitro.

As part of our continuing efforts to accumulate information on the coding region of unidentified human genes, we newly determined the sequences of 40 cDNA clones of human cell line KG-1 which correspond to relatively long and nearly full-length transcripts, and predicted the coding sequences of the corresponding genes, named KIAA0161 to 0200. The average size of the cDNA clones analyzed was approximately 5.0 kb. A computer search of the sequences in public databases indicated that the sequences of 20 genes were unrelated to any reported genes, while the remaining 20 genes carried sequences which show some similarities to known genes. Among the genes in the latter category, KIAA0167 contained a Zn-finger motif with significant structural similarity to that of the yeast transcription factor GCS1, and KIAA0189 was classified into the RhoGAP gene family. Stretches of typical CAG (Gin) repeats, which were often correlated with genetic disorders, were found in KIAA0181 and KIAA0192. Another novel repeat composed of alternating Arg and Glu was identified in KIAA0182. Northern hybridization analysis demonstrated that 10 genes are expressed in a cell- or tissue-specific manner.

Ataxia telangiectasia (AT) is an autosomal recessive disease of unknown etiology associated with cerebellar ataxia, oculocutaneous telangiectasia, immunodeficiency, and hypersensitivity to ionizing radiation. Although AT has been divided into four complementation groups by its radioresistant-DNA synthesis phenotype, the ATM gene has been isolated as the candidate gene responsible for all AT groups. We identified a new gene, designated NPAT, from the major AT locus on human chromosome 11q22-q23. The gene encoded a 1421-amino-acid protein containing nuclear localization signals and phosphorylation target sites by cyclin-dependent protein kinases associated with E2F. The messenger RNA of NPAT was detected in all human tissues examined, and its genomic sequence was strongly conserved through eukaryotes, suggesting that the NPAT gene may be essential for cell maintenance and may be a member of the housekeeping genes. Analysis of the genomic region of NPAT surprisingly revealed that the gene existed only 0.5 kb apart from the 5' end of the ATM transcript with opposite transcriptional direction. it may be possible to propose the idea that the promoter region could be shared by both housekeeping genes and that each gene could influence the expression of the other.

Using the cDNA clone from mouse testis which carries the conserved sequences among meiosis-specific recA-like genes, a highly homologous cDNA clone was isolated from a cDNA library of human testis. The clone had the coding capacity of a protein consisting of 340 amino acid residues, which coincides with the average size of putative eukaryotic recA-like proteins. When expression of the corresponding gene, named HsLIM15, in various tissues was examined by reverse transcription-PCR, products of two different sizes were detected in testis: While the longer was predominantly seen in the testis, the shorter was commonly to all the tissues including the testis. Analysis of the sequences indicated that the longer product corresponded to the above cDNA clone, and the shorter one was its deletion product missing an internal 165 bp portion. The result strongly suggests that the mRNA species coding for the putative meiosis-specific RecA-like protein in human is predominantly expressed in testis possibly as an alternative splicing product of a ubiquitously expressed gene.

In this series of projects regarding the accumulation of sequence information of unidentified human genes, we newly deduced the sequences of 40 full-length cDNA clones of human cell line KG-1, and predicted the coding sequences of the corresponding genes, named KIAA0121 to 0160. The results of a computer search of public databases indicated that the sequences of 13 genes were unrelated to any reported genes, while the remaining 27 genes carried sequences which showed some similarities to known genes. Obvious unique sequences noted were as follows. A stretch of triplet repeats was contained in each of three genes: These were GAG(Glu) in KIAA0122 and KIAA0147, and TCC(Ser) in KIAA0150. A stretch of 10 amino acid-residues was repeated 21 times in KIAA0139, and a homologous sequence of 76-78 nucleotides was found repeated 6 times in the untranslated region of KIAA0125. northern hybridization analysis demonstrated that 13 genes were expressed in a cell- or tissue-specific manner. Although a vast number of expressed sequence tags (ESTs) have been registered for comprehensive analysis of cDNA clones, our sequence data indicated that their distribution is very unbalanced: e.g. while no EST hit 7 genes, 85 ESTs fell in a single gene.

We isolated full-length cDNA clones from size-fractionated cDNA libraries of human immature myeloid cell line KG-1, and the coding sequences of 40 genes were newly predicted. A computer search of the GenBank/EMBL databases indicated that the sequences of 14 genes were unrelated to any reported genes, while the remaining 26 genes carried some sequences with similarities to known genes. Significant transmembrane domains were identified in 17 genes, and protein motifs that matched those in the PROSITE motif database were identified in 11 genes. Northern hybridization analysis with 18 different cells and tissues demonstrated that 10 genes were apparently expressed in a cell-specific or tissue-specific manner. Among the genes predicted, half were isolated from the medium-sized cDNA library and the other half from the small-sized cDNA library, and their average sizes were 4 kb and 1.4 kb, respectively. As judged by Northern hybridization profiles, small-sized cDNAs appeared to be expressed more ubiquitously and abundantly in various tissues, compared with that of medium-sized cDNAs.

Ataxia telangiectasia (AT) is an autosomal recessive disease of unknown etiology associated with cerebellar ataxia, telangiectasia, immune dysfunction, higher cancer risk, genomic instability and hypersensitivity to ionizing radiation. The major AT loci, AT-A and AT-C, are shown to be closely linked at chromosome 11q22-q23. The most recent genetic linkage mapping and linkage disequilibrium analysis have localized the major AT loci to a sequence of approximately 850 kb between the markers D11S1819 and D11S1818. The isolation of yeast artificial chromosomes spanning the AT region is an essential step to identify the gene or genes responsible for the mutation(s). We isolated a total of 20 YAC clones from three independent YAC libraries, using sequence tagged sites mapped in the AT region as primers for PCR-based YAC screening. The PCR assay for the presence or absence of 16 different DNA markers allowed us to construct and to order four YAC contigs at the AT region. One of the contigs which consists of the 10 YAC clones, covers about 2 Mb of DNA at the boundary between Giemsa-positive band 11q22.3 and Giemsa-negative band 11q23.1 and includes the entire region of the major AT locus between D11S1819 and D11S1818. Thus, the YAC contigs will facilitate the positional cloning approach for searching transcribed sequences from the defined genomic region.

The complete gene for human thrombopoietin (TPO) has been cloned by screening a human genomic library using human TPO cDNA as a probe. This gene is 6.2 kb in length and contains six exons and five introns. It is shown that the human genome contains a single copy of the human TPO gene according to Southern blotting analysis. The transcription initiation site was determined by S1 nuclease mapping. The human TPO gene expressed TPO activity when transfected into COS-1 cells. The human TPO gene has been mapped to chromosome 3q27 by in situ hybridization using a biotin-labeled probe. © 1994.

Fragile X syndrome is the most common familial form of mental retardation and known to be associated with the fragile site at Xq27.3 (FRAXA). The syndrome has recently been characterized by a unique genetic mechanism which involves dynamic mutation due to a heritable unstable DNA sequence, p(CCG)n repeat, in the FRAXA locus. We were asked to make a genetic diagnosis on the case of a normal male who has two brothers and a maternal uncle with mental retardation. We performed the pedigree analysis of the fragile X syndrome using both cytogenetic and molecular techniques. The affected two brothers and the uncle showed cytogenetic expression of the fra (X)(q27.3) and carried hypermethylated full mutation in the FRAXA locus. The phenotypically normal mother also exhibited fragile X expression and was found to be a carrier of premutation. Via female transmission, the premutation converted to full mutation and exhibited somatic heterogeneity and hypermethylation. However, both cytogenetic and molecular data did not show any evidence of fragile X mutation in the normal male client and, thus, excluded the possibility of his being a carrier.<br>

A human chromosomal region 11q23-specific DNA library has been constructed by means of microdissection-microcloning method with polymerase chain reaction (PCR) technique (Seki et al., Genomics 16: 1993). DNA sequences were determined for 25 microclones that contained approximately 300-500 bp insert and gave a unique (single copy) signal in Southern blot analysis. The sequence tagged site (STS) was designed and appropriate condition for PCR was determined for each unique microclone. Twelve STSs were established and used for PCR-screening of human genomic libraries constructed with yeast artificial chromosome (YAC). Thirteen YAC clones have been isolated from eight STSs. These chromosomal region-specific STSs and YAC clones will be useful in the positional cloning of disease-related genes localized to the q23 region of chromosome 11. © 1994 The Japan Society of Human Genetics.

By applying the protocol previously established, we isolated and sequenced full-length cDNA clones longer than 2 kb from a cDNA library of human immature myeloid cell line KG-1, and the coding sequences of 40 new genes were predicted. A computer search of the sequences indicated that 29 genes contained sequences with similarities to reported genes in the GenBank/EMBL databases. Significant transmembrane domains were identified in 9 genes, 5 of which harbored multiple hydrophobic regions. Protein motifs that matched those in the PROSITE motif database were identified in 13 genes. In terms of sequence similarities and protein motifs, 5 genes were related to transcriptional factors. Repetitive sequences were found in the 3′-untranslated region of 8 genes. Northern hybridization demonstrated that the expression of 9 genes was tissue-specific, while the remaining 31 genes were expressed ubiquitously. It was also noted that 17 genes yielded different sizes of bands possibly due to either alternative splicing or alternative initiation. The chromosomal location of these genes has been determined.

We established a protocol for the prediction of the coding sequences of unidentified human genes based on the double selection and sequence analysis of cDNA clones with inserts carrying unreported 5′-terminal sequences and with insert sizes corresponding to nearly full-length transcripts. By applying the protocol, cDNA clones with inserts longer than 2 kb were isolated from a cDNA library of human immature myeloid cell line KG-1, and the coding sequences of 40 new genes were predicted. A computer search of the sequences indicated that 20 genes contained sequences similar to known genes in the GenBank/EMBL databases. The sequences of the remaining 20 genes were entirely new, and characteristic protein motifs or domains were identified in 32 genes. Other sequence features noted were that the coding sequences of 23 genes were followed by relatively long stretches of 3′-untranslated sequences and that 5 genes contained repetitive sequences in their 3′-untranslated regions. The chromosomal location of these genes has been determined. By increasing the scale of the above analysis, the coding sequences of many unidentified genes can be predicted.

Fragile X syndrome, associated with the fragile site at Xq27.3 (FRAXA), is the most common form of familial mental retardation. The fragile X mutation has recently been characterized as a heritable unstable DNA sequence, p(CCG)n/p(CGG)n, in the FRAXA locus. In the present study, a correlation between fragile X‐genotypes in the FRAXA locus and hypermethylation of an adjacent CpG island was examined in four Japanese families with fragile X syndrome. We show here that the heritable unstable DNA sequences in the fragile X chromosome usually increase in size when transmitted by female carriers, and that the degree of methylation in the CpG island correlated with the increased sizes of the unstable DNA sequences. When a hypermethylated full mutation was transmitted by a male to his daughters, both the size of the unstable DNA sequence and the degree of the methylation reduced to the premutation range. Our observations suggest that female meiosis has a greater potential for amplifying unstable DNA sequences and that amplified DNA sequences can be transmitted through germ cells, while male germ cells seem not to be able to tolerate highly amplified unstable DNA sequences. 1993 Blackwell Munksgaard

Fragile X syndrome, one of the most common human genetic diseases, is characterised by a unique genetic mechanism which involves dynamic mutation because of a heritable unstable DNA sequence and abnormal DNA methylation. Direct detection of the dynamic mutation and its methylation status at the DNA level would facilitate reliable tests for prenatal and postnatal diagnosis of the disease and for carrier detection. However, it has been suggested that DNA methylation can not be used as the basis for prenatal diagnosis as the CpG island is not always methylated in chorionic villus DNA. We report here a male fetus exhibiting both extensive somatic heterogeneity and abnormal hypermethylation of the full fragile X mutation in chorionic villus DNA as well as in fetal tissue DNA. Our results indicate that both somatic heterogeneity and hypermethylation of the full fragile X mutation are events that are clearly detectable in the 11th to 12th week of pregnancy.

Yamauchi M, Nagata S, Seki N, Toyama Y, Harada N, Niikawa N, Masuno I, Kajii T, Hori T. Prenatal diagnosis of fragile X syndrome by direct detection of the dynamic mutation due to an unstable DNA sequence. Clin Genet 1993: 44: 169–172. © Munksgaard, 1993 The fragile X syndrome is the most common familial form of mental retardation. The mutation causing the syndrome is dynamic mutation due to an unstable DNA (CCG)n repeat localized at Xq27.3. We have previously reported a PCR procedure to prepare a diagnostic probe, pPCRfx1, which can be used to determine the genotype of fragile X mutation individuals by Southern blot analysis. In the present study, pPCRfx1 was applied to the prenatal diagnosis, using chorionic villus cells, of a fetus which was at risk of having fragile X syndrome. In the PstI assay, the Southern blot showed the typical pattern of a female carrier with the full mutation. Analysis of the DNA methylation patterns by EcoRI + EagI assay showed that the EagI restriction site was not methylated on the mutated X chromosome of chorionic villi, but the sites were totally methylated in the brain and other tissues of the fetus. Thus the fetus was diagnosed to be a heterozygous female carrier of the dynamic mutation involved in the fragile X syndrome. Copyright © 1993, Wiley Blackwell. All rights reserved