Chromosome 5q deletion and epigenetic suppression of the gene encoding alpha-catenin (CTNNA1) in myeloid cell transformation

Interstitial loss of all or part of the long arm of chromosome 5, or del(5q), is a frequent clonal chromosomal abnormality in human myelodysplastic syndrome (MDS, a preleukemic disorder) and acute myeloid leukemia (AML), and is thought to contribute to the pathogenesis of these diseases by deleting one or more tumor-suppressor genes. Although a major commonly deleted region (CDR) has been delineated on chromosome band 5q31.1 (refs. 3-7), attempts to identify tumor suppressors within this band have been unsuccessful. We focused our analysis of gene expression on RNA from primitive leukemia-initiating cells, which harbor 5q deletions, and analyzed 12 genes within the CDR that are expressed by normal hematopoietic stem cells. Here we show that the gene encoding alpha-catenin (CTNNA1) is expressed at a much lower level in leukemia-initiating stem cells from individuals with AML or MDS with a 5q deletion than in individuals with MDS or AML lacking a 5q deletion or in normal hematopoietic stem cells. Analysis of HL-60 cells, a myeloid leukemia line with deletion of the 5q31 region, showed that the CTNNA1 promoter of the retained allele is suppressed by both methylation and histone deacetylation. Restoration of CTNNA1 expression in HL-60 cells resulted in reduced proliferation and apoptotic cell death. Thus, loss of expression of the alpha-catenin tumor suppressor in hematopoietic stem cells may provide a growth advantage that contributes to human MDS or AML with del(5q).     

Genomic organization and amplification of the human desmosomal cadherin genes DSC1 and DSC3, encoding desmocollin types 1 and 3

The desmosomal cadherins comprise the desmocollins and desmogleins and are involved in epithelial cell-cell adhesion. There are three desmocollins (DSC 1-3) and three desmogleins (DSG 1-3) that are expressed in a tissue- and development-specific manner. Desmosomal proteins have been implicated in a number of disorders characterized by loss of cell-cell adhesion and trauma-induced skin fragility. Therefore, the desmocollins are potential candidates for genodermatoses involving epithelial tissues. In order to screen the entire DSC1 and DSC3 genes, we have characterized their intron-exon organization. The DSC1 gene comprises 17 exons spanning approximately 33 kb on 18q12.1, and the DSC3 gene comprises 17 exons spanning approximately 49 kb on 18q12.1. We have also developed a comprehensive PCR-based mutation detection strategy for desmocollins 1, 2, and 3 using primers placed on flanking introns followed by direct sequencing of the PCR products.     

Identification of a Drosophila homologue of alpha-catenin and its association with the armadillo protein

The cadherin cell adhesion system plays a central role in cell-cell adhesion in vertebrates, but its homologues are not identified in the invertebrate. alpha-Catenins are a group of proteins associated with cadherins, and this association is crucial for the cadherins' function. Here, we report the cloning of a Drosophila alpha-catenin gene by low stringent hybridization with a mouse alpha E-catenin probe. Isolated cDNAs encoded a 110-kD protein with 60% identity to mouse alpha E- catenin, and this protein was termed D alpha-catenin. The gene of this protein was located at the chromosome band 80B. Immunostaining analysis using a mAb to D alpha-catenin revealed that it was localized to cell- cell contact sites, expressed throughout development and present in a wide variety of tissues. When this protein was immunoprecipitated from detergent extracts of Drosophila embryos or cell lines, several proteins co-precipitated. These included the armadillo product which was known to be a Drosophila homologue of beta-catenin, another cadherin-associated protein in vertebrates, and a 150-kD glycoprotein. These results strongly suggest that Drosophila has a cell adhesion machinery homologous to the vertebrate cadherin-catenin system.     

Molecular cloning and characterization of the human topoisomerase IIalpha and IIbeta genes: evidence for isoform evolution through gene duplication

Human DNA topoisomerase II is essential for chromosome segregation and is the target for several clinically important anticancer agents. It is expressed as genetically distinct alpha and beta isoforms encoded by the TOP2alpha and TOP2beta genes that map to chromosomes 17q21-22 and 3p24, respectively. The genes display different patterns of cell cycle- and tissue-specific expression, with the alpha isoform markedly upregulated in proliferating cells. In addition to the fundamental role of TOP2alpha and TOP2beta genes in cell growth and development, altered expression and rearrangement of both genes are implicated in anticancer drug resistance. Here, we report the complete structure of the human topoisomerase IIalpha gene, which consists of 35 exons spanning 27.5 kb. Sequence data for the exon-intron boundaries were determined and examined in the context of topoisomerase IIalpha protein structure comprising three functional domains associated with energy transduction, DNA breakage-reunion activity and nuclear localization. The organization of the 3' half of human TOP2beta, including sequence specifying the C-terminal nuclear localization domain, was also elucidated. Of the 15 introns identified in this 20 kb region of TOP2beta, the first nine and the last intron align in identical positions and display the same phases as introns in TOP2alpha. Though their extreme 3' ends differ, the striking conservation suggests the two genes diverged recently in evolutionary terms consistent with a gene duplication event. Access to TOP2alpha and TOP2beta gene structures should aid studies of mutations and gene rearrangements associated with anticancer drug resistance.     

Coordinate recruitment of E-cadherin and ALCAM to cell-cell contacts by alpha-catenin

Here we report on the role of alpha-catenin in the cellular localization of activated leukocyte cell adhesion molecule, ALCAM, and cadherin-mediated cell adhesion in human prostate cancer cells. Cell lines that have a functional E-cadherin-mediated cell adhesion (DU-145 and LNCaP) show ALCAM staining at cell-cell contacts. In contrast, in cell lines that lack alpha-catenin expression (ALVA-31, PC-3, and PPC-1), E-cadherin-mediated adhesion is disturbed and ALCAM staining is cytoplasmic. A role of alpha-catenin in the recruitment of E-cadherin and ALCAM to cell-cell contacts was established by transfection of an alpha-N-catenin construct into cell lines ALVA-31 and PC-3. This resulted not only in the correct assembly of E-cadherin/alpha-catenin complexes at the cell membrane but also in localization of ALCAM to cell-cell contacts, indicating that indeed alpha-catenin affects ALCAM localization.     

Human acetylcholinesterase and butyrylcholinesterase are encoded by two distinct genes

1. Various hybridization approaches were employed to investigate structural and chromosomal interrelationships between the human cholinesterase genes CHE and ACHE encoding the polymorphic, closely related, and coordinately regulated enzymes having butyrylcholinesterase (BuChE) and acetylcholinesterase (AChE) activities. 2. Homologous cosmid recombination with a 190-base pair 5' fragment from BuChEcDNA resulted in the isolation of four overlapping cosmid clones, apparently derived from a single gene with several introns. The Cosmid CHEDNA included a 700-base pair fragment known to be expressed at the 3' end of BuChEcDNA from nervous system tumors and which has been mapped by in situ hybridization to the unique 3q26-ter position. In contrast, cosmid CHEDNA did not hybridize with full-length AChEcDNA, proving that the complete CHE gene does not include AChE-encoding sequences either in exons or in its introns. 3. The chromosomal origin of BuChE-encoding sequences was further examined by two unrelated gene mapping approaches. Filter hybridization with DNA from human/hamster hybrid cell lines revealed BuChEcDNA-hybridizing sequences only in cell lines including human chromosome 3. However, three BuChEcDNA-homologous sequences were observed at chromosomal positions 3q21, 3q26-ter, and 16q21 by a highly stringent in situ hybridization protocol, including washes at high temperature and low salt. 4. These findings stress the selectivity of cosmid recombination and chromosome blots, raise the possibility of individual differences in BuChEcDNA-hybridizing sequences, and present an example for a family highly similar proteins encoded by distinct, nonhomologous genes.     

Alpha-catenin is a molecular switch that binds E-cadherin-beta-catenin and regulates actin-filament assembly

Epithelial cell-cell junctions, organized by adhesion proteins and the underlying actin cytoskeleton, are considered to be stable structures maintaining the structural integrity of tissues. Contrary to the idea that alpha-catenin links the adhesion protein E-cadherin through beta-catenin to the actin cytoskeleton, in the accompanying paper we report that alpha-catenin does not bind simultaneously to both E-cadherin-beta-catenin and actin filaments. Here we demonstrate that alpha-catenin exists as a monomer or a homodimer with different binding properties. Monomeric alpha-catenin binds more strongly to E-cadherin-beta-catenin, whereas the dimer preferentially binds actin filaments. Different molecular conformations are associated with these different binding states, indicating that alpha-catenin is an allosteric protein. Significantly, alpha-catenin directly regulates actin-filament organization by suppressing Arp2/3-mediated actin polymerization, likely by competing with the Arp2/3 complex for binding to actin filaments. These results indicate a new role for alpha-catenin in local regulation of actin assembly and organization at sites of cadherin-mediated cell-cell adhesion.     

Requirement of interaction of nectin-1alpha/HveC with afadin for efficient cell-cell spread of herpes simplex virus type 1

We recently found a novel cell-cell adhesion system at cadherin-based adherens junctions (AJs), consisting at least of nectin, a Ca(2+)-independent homophilic immunoglobulin-like adhesion molecule, and afadin, an actin filament-binding protein that connects nectin to the actin cytoskeleton. Nectin is associated with cadherin through afadin and alpha-catenin. The cadherin-catenin system increases the concentration of nectin at AJs in an afadin-dependent manner. Nectin constitutes a family consisting of three members: nectin-1, -2, and -3. Nectin-1 serves as an entry and cell-cell spread mediator of herpes simplex virus type 1 (HSV-1). We studied here a role of the interaction of nectin-1alpha with afadin in entry and/or cell-cell spread of HSV-1. By the use of cadherin-deficient L cells overexpressing the full length of nectin-1alpha capable of interacting with afadin and L cells overexpressing a truncated form of nectin-1alpha incapable of interacting with afadin, we found that the interaction of nectin-1alpha with afadin increased the efficiency of cell-cell spread, but not entry, of HSV-1. This interaction did not affect the binding to nectin-1alpha of glycoprotein D, a viral component mediating entry of HSV-1 into host cells. Furthermore, the cadherin-catenin system increased the efficiency of cell-cell spread of HSV-1, although it also increased the efficiency of entry of HSV-1. It is likely that efficient cell-cell spread of HSV-1 is caused by afadin-dependent concentrated localization of nectin-1alpha at cadherin-based AJs.     

Structure and expression of the gene coding for the human serpin hLS2

We have analyzed genomic clones encoding human leuserpin 2 (hLS2). The gene covers about 14.5 kilobases and consists of 5 exons and 4 introns. The genes coding for hLS2, alpha 1-antitrypsin, alpha 1-antichymotrypsin, and rat angiotensinogen share an equivalent exon-intron structure and therefore constitute a distinct subgroup within the serpin gene family, which otherwise displays a highly variable exon-intron pattern. With the exception of a segment in the second exon, the sequence similarity of the genes coding for hLS2 and alpha 1-antitrypsin extends to all exons including one encoding the 5'-untranslated sequences. The implications of these findings with respect to the genesis of the amino-terminal heterogeneity in the serpin family are discussed.     

Chemokine PARC gene (SCYA18) generated by fusion of two MIP-1alpha/LD78alpha-like genes

Two loci in the human genome, chromosomes 4q12-q21 and 17q11.2, contain clusters of CXC and CC chemokine subfamily genes, respectively. Since mice appear to contain fewer chemokine genes than humans, numerous gene duplications might have occurred in each locus of the human genome. Here we describe the genomic organization of the human pulmonary and activation-regulated CC chemokine (PARC), also known as DC-CK1 and AMAC-1. Despite high sequence similarity to a CC chemokine macrophage inflammatory protein-1alpha (MIP-1alpha)/LD78alpha, PARC is chemotactic for lymphocytes and not for monocytes and does not share its receptor with MIP-1alpha. Analyses of the BAC clones containing the human PARC gene indicated that the gene is located most closely to MIP-1alpha (HGMW-approved symbol SCYA3) and MIP-1beta (HGMW-approved symbol SCYA4) on chromosome 17q11.2. Dot-plot comparison suggested that the PARC gene had been generated by fusion of two MIP-1alpha-like genes with deletion and selective usage of exons. Base changes accumulated before and after the fusion might have adapted the gene to a new function. Since there are variably duplicated copies of the MIP-1alpha gene called LD78beta (HGMW-approved symbol SCYA3L) in the vicinity of the MIP-1alpha gene, the locus surrounding the MIP-1alpha gene seems to be a "hot spring" that continuously produces new family genes. This evidence provides a new model, duplication and fusion, of the molecular basis for diversity within a gene family.     

Structure and chromosomal localization of human and mouse genes for hematopoietic prostaglandin D synthase. Conservation of the ancestral genomic structure of sigma-class glutathione S-transferase.

Hematopoietic prostaglandin D synthase (H-PGDS) is the key enzyme for the production of the D and J series of prostanoids, and the first recognized vertebrate homolog of sigma-class glutathione S-transferase (GST). We isolated the genes and cDNAs for human and mouse H-PGDSs. The human and mouse cDNAs contained a coding region corresponding to 199 amino-acid residues with calculated molecular masses of 23 343 and 23 226, respectively. Both H-PGDS proteins recombinantly expressed in Escherichia coli showed bifunctional activities for PGDS and GST, and had almost the same catalytic properties as the rat enzyme. Northern analyses demonstrated that the H-PGDS genes were expressed in a highly species-specific manner. Whereas the human gene was widely distributed, in contrast, the mouse gene was detected only in samples from oviduct and skin. By fluorescence in situ hybridization, the chromosomal localization of the human and mouse H-PGDS genes were mapped to 4q21-22 and 3D-E, respectively. The human and mouse H-PGDS genes spanned approximately 41 and 28 kb, respectively, and consisted of six exons divided by five introns. The exon/intron boundaries of both genes were completely identical to those of the sigma-class GST subfamily, although the amino-acid sequences of the latter were only 17.0-21.5% identical to those of either H-PGDS. These findings suggest that the H-PGDS genes evolved from the same ancestral gene as the members of the sigma-class GST family.     

Cloning and chromosomal localization of human genes encoding the three chains of type VI collagen.

Type VI collagen is a heterotrimer composed of three polypeptide chains, alpha 1(VI), alpha 2(VI), and alpha 3(VI). By immunological screening of an expression cDNA library, human cDNAs specific for each chain were isolated and characterized. Major mRNA species encoding these chains have a size of 4.2 kb (alpha 1), 3.5 kb (alpha 2), and 8.5 kb (alpha 3). The cDNA clones were also used to map the genes on human chromosomes by somatic cell hybrid analysis and in situ hybridization. The alpha 1 (VI) and alpha 2(VI) collagen genes were both located on chromosome 21, in band q223. This represents a third example of a possible physical proximity of two collagen loci. The alpha 3(VI) collagen gene was localized to chromosome 2, in the region 2q37. The alpha 3(VI) collagen gene is the fifth extracellular matrix gene to be localized to 2q, as four other extracellular matrix genes--i.e., the alpha 1(III) and alpha 2(V) collagen genes, the elastin gene, and the fibronectin gene--have been previously mapped to the distal region of the long arm of chromosome 2.     

Genomic organization and chromosomal localization of three members of the UDP-N-acetylgalactosamine: polypeptide N- acetylgalactosaminyltransferase family

A homologous family of UDP- N -acetylgalactosamine: polypeptide N - acetylgalactosaminyltransferases (GalNAc-transferases) initiate O- glycosylation. These transferases share overall amino acid sequence similarities of approximately 45-50%, but segments with higher similarities of approximately 80% are found in the putative catalytic domain. Here we have characterized the genomic organization of the coding regions of three GalNAc-transferase genes and determined their chromosomal localization. The coding regions of GALNT1 , -T2 , and -T3 were found to span 11, 16, and 10 exons, respectively. Several intron/exon boundaries were conserved within the three genes. One conserved boundary was shared in a homologous C. elegans GalNAc- transferase gene. Fluorescence in situ hybridization showed that GALNT1 , -T2 , and -T3 are localized at chromosomes 18q12-q21, 1q41-q42, and 2q24-q31, respectively. These results suggest that the members of the polypeptide GalNAc-transferase family diverged early in evolution from a common ancestral gene through gene duplication.      

Regulation of cadherin-mediated cell-cell adhesion by the Rho family GTPases.

Reports in the past two years have shown that Cdc42, Rac1, and Rho - belonging to the Rho small GTPase family - participate in the regulation of cadherin-mediated cell-cell adhesion. IQGAP1, an effector of Cdc42 and Rac1, interacts with cadherin and beta-catenin and induces the dissociation of alpha-catenin from the cadherin-catenins complex leading to disruption of cell-cell adhesion: activated Cdc42 and Rac1 counteract the effect of IQGAP1. Thus, Cdc42 and Rac1 appear to regulate cadherin-mediated cell-cell adhesion acting through IQGAP1.     

Replication timing of extremely large genes on human chromosomes 11q and 21q

Many human genes have been mapped precisely in the genome. These genes vary from a few kb to more than 1 Mb in length. Previously, we measured replication timing along the entire lengths of human chromosomes 11q and 21q at the sequence level. In the present study, we used the newest information for human chromosomes 11q and 21q to analyze the replication timing of 30 extremely large genes (>250 kb) in two human cell lines (THP-1 and Jurkat). The timing of replication differed between the 5'- and 3'-ends of each of extremely large genes on 11q and 21q, and the time interval between their replication varied among genes of different lengths. The large genes analyzed here included several tissue-specific genes associated with neural diseases and genes encoding cell adhesion molecules: some of these genes had different patterns of replication timing between the two cell lines. The amyloid precursor protein gene (APP), which is associated with familial Alzheimer's disease (AD1), showed the largest difference in timing of replication between its 5'- and 3'-ends in relation to gene length of all the large genes studied on 11q and 21q. These extremely large genes were concentrated in and around genomic regions in which replication timing switches from early to late on both 11q and 21q. The differences of replication timing between the 5'- and 3'-terminal regions of large genes may be related to the molecular mechanisms that underlie tissue-specific expression.