GENE DISCOVERY ACTIVITIES ON FINISHED CHROMOSOME 1 SEQUENCES.

Susan Rhodes¹, Richard M. Bruskiewich², Andrew King², Laurens Wilming², David Bentley¹, Tim Hubbard²

¹Human Genetics and ²Human Analysis/Informatics, Sanger Centre, Wellcome Trust Genome Campus, Cambridge, United Kingdom

We are undertaking systematic gene discovery by detailed human manual annotation and experimental analysis on 'finished' sequenced clones originating from the Sanger Centre human high throughput sequencing program on chromosome 1. Our human manual annotation activities consist of expert assessment of gene feature evidence generated by semi-automatic analysis of finished genomic sequences by a large suite of computer software based analytical tools, the results of which are presented within the graphical ACEDB database environment. These analytical tools use available evidence (e.g. similarity matches to entries in public sequence databases) to generate putative gene structures for assessment by the human expert. Our experimental approach consists primarily of cDNA rescue from one of 17 human cDNA libraries using a polymerase chain reaction (PCR) based screen with primer systems designed from preliminary expert human analysis of homology, EST or related data or, in some instances, by 'Ab initio' predicted gene structures. Each cDNA library consists of approximately 0.5 million clones for screening by PCR. PCR experiments are undertaken for the following situations (using inferred exons for primer design) in the following rough order of priority: Genes annotated by the Sanger Centre Human Analysis Group (Andrew King/Laurens Wilming) or exact matches to expressed sequence tags (ESTs) where there is some sort of conflict between the published data and the genomic sequence, usually splice site or problematic UTR's: ESTs which exhibit some partial exon/intron structure: Human EST clusters where ESTs are at 90%+ aligned and similar to the genomic sequence over most of their length, do not contain large repeats and are not adjacent to a run of A's in the genomic sequence: 'Ab initio' predicted genes of significant size (several exons long) and exhibiting appreciable exact exon match agreement between predictions from two or more prediction programs (at least, highly compatible Genscan and Fgenes predictions). Our poster will outline the global results of our analyses and specific results, where applicable, of our gene discoveries on the chromosome (e.g. identification of known, mapped but previously unsequenced genes).

AN EFFORT TO SEARCH FOR THE RP18 RESPONSIBLE GENE IN 1Q21 CHROMOSOMAL REGION

Loris Bernard¹, Sara Volorio¹, Giovanni Porta⁵, Mita Mancini⁴, Alessandro Bulfone¹, Claudio Gattuso¹, Andreas Gal³, Shomy Bhattacharya², Andrea Ballabio¹, Massimo Zollo¹

¹Telethon Institute of Genetics and Medicine (TIGEM), Milan, Italy
²Ophthalmology Department of Molecular Genetics, University College London, London, UK.
³Institut fur Humangenetik UKE, Hamburg, Germany
⁴HSR, YAC Screening Centre, Milano, Italy
⁵Dipartimento di patologia Umana ed Ereditaria, II Facolta' di Medicina e Chirurgia, Universita' di Pavia, Pavia, Italy

Our laboratory has been recently engaged in the characterization of the human homologue of the Drosophila prune gene located in 1q21 chromosomal region, through dbEST searches (1). We have excluded the gene as responsible for the Autosomal Dominant Retinitis Pigmentosa (adRP18: OMIM 601414) by mutation analysis performed in the Danish and English family previously reported (2), (3). In order to identify the RP18 responsible gene we have started a pilot mapping project in the region between the microsatellites markers (D1S1664 and D1S442) corresponding to approximately 3 cM physical distance and, in particular, around D1S498 marker. Three main contigs constructed with PACs clones derived from the Human genomic PAC library -RPCI-5 distributed at IGeR facility (http://www.spr.it/iger/home.html ) were used to cover approximately 800 kb of genomic DNA.
Furthermore, we have isolated the positioned EST clusters in the chromosomal 1q21 region available at "http://www.ncbi.nlm.nih.gov/Schuler/UniGene/" and ready mapped described at "http://www.ncbi.nlm.nih.gov/genemap98/" (4). Selected ESTs were then analyzed by mRNA "in situ" in mouse embryos at E14.5 to verify expression in developing eyes. Within this strategy, we have started to determine, only for few candidate genes, the exon-intron sequences boundaries in order to perform mutation analysis in the above DNA family collection. Finally, in order to narrow down the critical region, we are searching for new (CA)n microsatellites and positioning them in the map. A summary results status and analysis will be presented.

1. A. Bulfone, et al., Hum Mol. Genet 7, 1997 (1998).
2. C. F. Inglehearn, et al., J Med. Genet 35, 788 (1998).
3. S. Y. Xu, T. Rosenberg, A. Gal, Hum-Genet. 102, 493 (1998).
4. P. Deloukas, et al., Science 282, 744 (1998).

MAPPING DIRECT SELECTION TRANSCRIPTS IN THE CHARCOT-MARIE-TOOTH TYPE 2 (CMT2A) REGION OF 1p36.

A. Lindstrom, J.M. Stajich, B.Seo, D. Shand, J.M. Vance

Duke University Medical Center, Center for Human Genetics, Durham, NC, USA

CMT2A is an axonal neuropathy previous mapped to 1p35-36. To identify genes expressed in the axonal nerve we have constructed a partial PAC contig from the region, and performed a direct selection using peripheral nerve and spinal cord from approximately 40 PACs in the region. These have been sequenced, and screened versus existing EST, Sanger chromosome 1 ACE and gene data bases. STS have been constructed for unique clones and mapped back against the PAC contigs and radiation hybrids in the region. Approximately 300 initial clones were identified. Maps and contigs will be presented.

A GENE FOR INHERITED CUTANEOUS VENOUS ANOMALIES ("GLOMANGIOMAS") LOCALIZES TO CHROMOSOME 1p21-22.

P. Brouillard¹, L. M Boon^1,2, A. Irrthum¹, L. Karttunen³, M. L. Warman⁴, R. Rudolph⁵, J. B Mulliken⁶, B. R Olsen³, M. Vikkula¹

¹Laboratory of Human Molecular Genetics, Christian de Duve Institute
²Center for Vascular Anomalies, Division of Plastic Surgery, UniversitE9 catholique de Louvain, Brussels, Belgium
³Department of Cell Biology, Harvard School of Dental Medicine, Boston, MA, USA
⁴Department of Genetics, Case Western Reserve University School of Medicine, Cleveland, OH, USA
⁵Division of Plastic and Reconstructive Surgery, University of California, La Jolla, CA, USA
⁶Division of Plastic Surgery, Children's Hospital, Boston, MA

Venous malformations (VMs) are congenital abnormalities of veins. Although VMs usually occur sporadically, we have been able to collect families with autosomal dominant inheritance. Using linkage analysis we have established that some families with inherited VMs link to chromosome 9p21 (Boon et al., Hum. Mol. Gen. 1994); the mutation causes ligand-independent activation of an endothelial cell specific receptor tyrosine kinase, TIE-2 (Vikkula et al., Cell, 1996). Now we show that VMs with glomus cells (known as "glomangiomas"), inherited as an autosomal dominant trait in five families, are not linked to 9p21, but instead link to a new locus on 1p21-p22, called VMGLOM (LOD score 12.70 with q 3D 0,00), bordered by AFMa205XD5 and D1S2775. Penetrance is close to 100%, but with variable expressivity for the number and size of lesions, time of occurrence of lesions, and painfulness of lesions. In contrast to the individuals with a TIE-2 mutation, members of the families showing linkage to VMGLOM did not exhibit any mucosal lesions. The linked locus is 4-6 cM (or 4-6 Mbp) based on current linkage and physical maps. We exclude three known positional candidate genes, DR1 (depressor of transcription 1), TGFBR3 (TGFb receptor 3) and TFA (tissue factor). We hypothesize that cutaneous venous anomalies ("glomangiomas") are caused by mutations in a novel gene that may act to regulate angiogenesis, in concert with the TIE-2 signaling pathway.

COMPREHENSIVE INTEGRATION AND CATALOGUING OF CHROMOSOME 1 USING COMPVIEW.

P.S. White¹, E.P. Sulman¹, T.C. Matise²

¹Division of Oncology, Children's Hospital of Philadelphia, Philadelphia PA, USA
²Lab of Statistical Genetics, The Rockefeller University, New York, NY, USA

Comprehensive representations of human chromosomes combining diverse genomic datasets, localizing expressed sequences, and reflecting physical distance are essential for disease gene identification and sequencing efforts. We have developed a method (CompView) for integrating and cataloguing genomic information derived from available cytogenetic, genetic linkage, radiation hybrid, physical, and transcript-based mapping approaches. CompView generates chromosome representations with substantially higher resolution, coverage, and integration than current maps of the human genome. The CompView process was used to build a representation of human chromosome 1, yielding a map with over 13,000 unique elements, 3500 transcripts, 1000 polymorphisms, an effective resolution of 900 kilobases, and a marker density of 50 kilobases. An associated Website offering a variety of text-based, tabular, and interactive graphical query and query return options will be launched at the workshop (http://genome.chop.edu). The chromosome 1 dataset was used to determine correlations between gene density and cytogenetic bands on a chromosome-wide basis. CompView creates comprehensive and fully integrated depictions of a chromosome's clinical, biological, and structural information.

FINE MAPPING OF A NEUROBLASTOMA TUMOR SUPPRESSOR CANDIDATE GENE REGION IN 1p36.2-3, AND MUTATION ANALYSIS OF THE CORTISTATIN GENE LOCALIZED TO THE REGION

T. Martinsson¹, F. Abel¹, K. Ejeskar¹, R.M. Sjoberg¹, P. Kogner²

¹Dept Clin Genetics, Sahlgrenska Univ Hosp, Gothenburg
²Childhood Cancer Research Unit, Karolinska Hosp, Stockholm, Sweden

Neuroblastoma is a childhood tumor originating from the neural crest. A common genetic feature of neuroblastomas, also being an important prognostic factor, are deletions of chromosome region 1p. The deletion of 1p often involves a deletion of varying size, with a consensus region within the most distal bands 1p36.2-3. The neuroblastoma SRO (shortest region of overlap of deletions) presented earlier by our group is defined distally by the cluster of loci D1S80/D1Z2/CDC2L1 and proximally by loci D1S244 i.e. app.25 cM. The 1p deletions are, however, not restricted to neuroblastoma tumors. In fact a large spectrum of tumor types display deletions to varying degree of distal 1p. We have exploited the possibility of using deletions of other tumor types, preferentially that of germ cell tumors, and combining those with that of the neuroblastoma SRO. Also in germ cell tumors distal 1p-deletion have been shown to have prognostic significance. We found in our germ cell tumors a SRO ranging from D1S508 to D1S200. Interestingly, this region only partially overlapped (app. 5cM) with our neuroblastoma SRO in region D1S508 to D1S244. We have thus focused on analyzing this smaller region in search for genes involved in the genesis of different cancers. We have performed radiation hybrid mapping of a large number of markers, STSs, ESTs and others known to reside in 1p. We have also initiated the development of a BAC contig of the region. FISH, and fiber-FISH mapping of BACs were also performed. Cortistatin (CORT) is highly homologous to somatostatin, and the presence of somatostatin in neuroblastoma tumors has been shown to have a favorable prognostic value. Here we have localized CORT to the neuroblastoma consensus region, defined the genomic organization of the gene and performed mutation analyses and gene expression studies of neuroblastoma primary tumors. No evidence has yet however been found to confirm CORT to be a neuroblastoma tumor suppressor gene.

CONTINUATION OF THE GENOME DATABASE PROJECT

Christopher J Porter, C Conover Talbot Jr., A Jamie Cuticchia

The Johns Hopkins University School of Medicine, Baltimore MD, USA The Hospital for Sick Children, Toronto ON, Canada

A little over a year ago, at the fourth Chromosome 1 workshop, we reported on the termination of funding for the Genome Database and the attempts being made to ensure that GDB's data would remain available. Following the announced termination of GDB we received a large number of messages from GDB users supporting continuation of the database.
Towards the end of 1998, staff at the Bioinformatics Supercomputing Centre at the Hospital for Sick Children (HSC) in Toronto located a source of private funds for GDB continuation. The primary editable copy of database has now been moved to Toronto and replication to the international nodes has resumed. GDB operations will continue both at HSC and at a data curation center at The Johns Hopkins University. Access to the database via the World Wide Web is unchanged at http://www.gdb.org/ or at the international nodes, which should now once again be receiving regular updates.

Current GDB activities are focussed on recommendations received from HUGO's Human Gene Mapping Committee. We are coordinating our data with other databases, and investigating methods for integrating GDB data with the rapidly emerging genomic sequence. New search and editing tools are under development, and we welcome suggestions and requests from GDB users.

A MINIMAL DELETED REGION IN HEPATOCELLULAR CARCINOMA MAPS WITHIN HUMAN CHROMOSOME 1p36.1-36.2

Wei Fang¹, Zhe Piao¹, Manuel Perucho¹, Daniela Simon², Jin-Chuan Sheu³, Shi Huang¹

¹The Burnham Institute, La Jolla, CA, USA
²Department of Pathology, MCP-Hahnemann School of Medicine, Philadelphia, PA, USA

Loss of heterozygosity on 1p36 is commonly found in human hepatocelluar carcinoma (HCC). The proximal border of deletion has been previously mapped to marker D1S96 which is proximal to the center of the minimal deleted region in neuroblastoma (NB). To define a minimal deleted region in HCC, we studied 97 cases of primary HCC for LOH of genes and STS markers located within a 18 cM region immediately proximal to the NB consensus deletion center, including (telemere-D1S96)-D1S508-D1S160-D1S244-D1S489-RIZ-D1S228-D1S507- D1S436-D1S199-(centromere). 37 (36%) cases showed LOH for at least two markers. 25 of 67 (37%) informative cases showed LOH of RIZ specific markers. The minimal deleted region was mapped between D1S160 and D1S436 which retained heterozygosity in at least two cases. The results confirm previous finding that the HCC locus is proximal to the NB locus. Our data also mapped RIZ within the minimal deleted region which, in turn, suggest a role for RIZ as a candidate HCC suppressor. Consistently, decreased expression of the RIZ1 product of RIZ locus was found to be most common in HCC cell lines (80%) compared to other cancer cell lines. More importantly, forced RIZ1 expression suppressed HCC tumorigenicity in nude mice (Jiang, et al., Int. J. Cancer. in press). That loss of RIZ1 expression is causally linked to tumor formation was proven by generating RIZ1-deficient mice which were tumor prone (Steele-Perkins, et al., manuscript submitted). However, single strand conformation polymorphism (SSCP) analysis did not show mutations in the PR domain region of RIZ1 in 49 cases of HCC examined. Our data suggest that the more common way of RIZ1 inactivation may not involve intragenic mutations.

CONSTRUCTION OF A PHYSICAL MAP AT 1q21-22 TO AID IN IDENTIFICATION OF THE GENE DEFECTIVE IN PARTIAL LIPODYSTROPHY

D.J.Lloyd¹, S.Shackleton¹, S.N.J.Jackson¹, R.Gwilliams², S.G.Gregory², R.C.Trembath¹

¹Division of Medical Genetics, Department of Genetics, Univeristy of Leicester, Leicester, UK
²The Sanger Centre, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK

Partial lipodystrophy (PLD) is characterised by the combination of a regional absence of subcutaneous fat, insulin resistance and hyperlipidaemia. This autosomal dominant trait is variable in extent and may overlap with more common diseases associated with glucose intolerance, including type II diabetes mellitus. We and others have previously mapped the PLD locus to a 5.3 cM region at chromosome 1q21-22.

Recombination-based analysis using a panel of microsatellites has identified a refined disease interval and a common disease haplotype between PLD families. We have generated partial BAC contigs covering this interval, completion of this physical map is being achieved by BAC-end clone walking. ESTs and candidate genes have been placed on the contigs and the analysis of their expression patterns, with particular reference to adipocytes, is underway.

The evidence of allelic association has prompted identification of SNPs within the PLD gene interval. These have now been placed on the physical map and the genotyping of these SNPs will provide useful information to support the existence of linkage disequilibrium. These resources will significantly aid in the refinement of the critical interval and, ultimately, the identification of the PLD disease gene.

COMPARATIVE MAPPING OF THE MOUSE AND HUMAN HOMOLOGOUS CHROMOSOME 1 REGIONS CONTAINING THE MOUSE NTD MUTANT Lp LOCUS

K. Doudney¹, J. Eddleston¹, M. Tham¹, J. Murdoch², C. Paternotte², S.G. Gregory³, A. Copp², P. Stanier¹

¹Division of Paediatrics, Obstetrics and Gynaecology, Queen Charlotte's and Chelsea Hospital, London, UK
²Neural Development Unit, Institute of Child Health, University College London, London, UK
³The Sanger Centre, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK

The distal portion of mouse chromosome 1 is homologous to human chromosome 1q21-q25. The homology has been established primarily at the level of gene content although there is also evidence for the conservation of gene order. We have exploited this information to enrich the transcript map of the mouse region to aid in the end stage of a positional cloning project designed to identify the gene for the severe NTD mouse mutant loop-tail (Lp). Human EST clusters that approximately map to the homologous region to Lp were used to identify the orthologous mouse genes by cross-screening EST databases for each species. STSs to these genes were then used in the construction of a ~3 Mb YAC contig spanning the Lp region. As a result, 31 genes were mapped and ordered in this interval. Within this region, a 700 kb PAC and P1 contig was constructed which contained the flanking markers which precisely defined the Lp critical interval. EST database screening significantly aided the identification of candidate genes for Lp with at least 11 genes now known to map within the critical region. To further study the level of conservation between the two species in this interval, a human BAC contig was constructed between SHGC-13385 and KCNJ9. The human ESTs initially used to identify the mouse genes were used to screen a BAC library. A combination of fingerprint analysis and common STS content demonstrated complete overlap across the region with the 55 clones identified. A minimum tiling path could be achieved with 6 BAC clones spanning an estimated 610 kb. From the comparative location of each of the mouse and human genes studied, it is clear that the physical size of the interval, gene content and order are very highly conserved in this region. The human and mouse contigs provide a sequence ready resource which will allow more detailed comparative analysis and aid in the identification of the Lp gene.

FISHING cDNAs FROM A FREQUENTLY AMPLIFIED REGION OF CHROMOSOME 1 IN SARCOMAS

Meza-Zepeda L.A.¹, Forus A.¹, Bagstevold A.¹, Godager L.H.¹, Marenholz I.², Mischke D.², South A.³, Nizetic D.³, Lioumi M.⁴, Ragoussis I.⁴, Serra M.⁵, Myklebost O¹

¹Department of Tumour Biology, The Norwegian Radium Hospital, Oslo, Norway
²Institut fFCrExperimentelle Onkologie un Transplantationsmedizin, Berlin, Germany
³Centre for Applied Molecular Biology, School of Pharmacy, University of London, London, UK
⁴Genomics Laboratory, Division of Medical Genetics UMDS, Guy's Hospital, London, UK
⁵ Laboratorio di Ricerca Oncologica, Instituto Ortopedico Rizzoli, Bologna, Italy

Recurrent amplification of the 1q21-q22 region has frequently been observed in human sarcomas and other malignancies. We have studied the amplification status of yeast artificial chromosomes (YAC) from a 6Mb contig partially covering the region by fluorescent in situ hybridisation (FISH). The YACs containing the marker D1S3620 were predominantly amplified in the samples tested. We have chosen the most frequently and highly amplified YAC (789f2) to capture cDNAs by direct cDNA selection. A linked cDNA library was constructed from an osteosarcoma cell line (Saos-2) with the relevant amplification and hybridised to biotinylated YAC sequences. cDNA complementary to YAC sequences were allowed to anneal under stringent conditions. The genomic/cDNA hybrids were captured using streptavidine coated magnetic beads and magnetic selection. Hybrids were washed under stringent conditions and the cDNA was eluted from the genomic sequences. Complementary DNA's were amplified and cloned. Characterisation of the captured cDNAs by sequence information, chromosomal localisation, expression and DNA copy number has been done. Sequence analysis revealed three novel genes, two belonging to gene families (novel members) and one novel gene. DNA copy number and expression studies indicate the over-representation and expression of these sequences in a subset of sarcoma. The isolated cDNAs may represent possible candidate genes for amplification in the region.

AN 8kb DELETION/INSERTION POLYMORPHISM IN THE VAN DER WOUDE SYNDROME CRITICAL REGION AT 1q32-q41.

Y. Watanabe¹, B.C. Bjork¹, P.-W. Chiang², S.G. Gregory³, D.M. Kurnit², J.C. Murray¹, B.C. Schutte¹

¹Dept. of Pediatrics, Univ. of Iowa, Iowa City, IA, USA
²Dept. of Pediatrics, Univ. of Michigan, Ann Arbor, MI, USA
³The Sanger Centre, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK

VWS is an autosomal dominant disorder whose cardinal features are cleft lip and/or palate and lip pits. The VWS gene was localized to a 1.6cM region of 1q32-q41 flanked by markers D1S491 and D1S205. A contig of BAC and PAC clones was constructed across the VWS critical region (B. Bjork et al., submitted). Eleven overlapping BAC and PAC clones, spanning approximately 1.1 Mb, were selected for high-throughput genomic (HTG) sequence analysis by the Sanger Centre (UK). From sequence analysis, the BAC clone 321i20 and the PAC clone 782d21 overlap, however the clone 321i20 was missing a 7922 bp region. Imbedded in this 7922 bp region was a short tandem repeat (STR). To verify that this nearly 8kb deletion in the clone bk321i20 was not a cloning artifact, we genotyped unaffected control samples using two sets of PCR primers, 186/7 that flank the STR and 583/4 that flank the putative 8 kb deleted sequence. The genotype data with 186/7 showed that the STR was polymorphic. The genotype data with 583/4 showed that the 8 kb deletion mutation was present in 54% of the chromosomes tested. The presence of the 8 kb deletion in the control samples was confirmed by QPCR. These results demonstrate the presence of an STRP within a common 8 kb deletion/insertion at the distal end of the VWS critical region. In addition, the deletion breakpoint sequences from 11 unrelated individuals were identical, suggesting that the 8 kb deletion/insertion may be an old mutation. Since this polymorphism is present in control samples and appears to be an old mutation, it is not a mutation that is etiologic for VWS. However, it is located near the distal breakpoints for two known microdeletions that are etiologic for VWS, providing additional evidence that this may be a region of genetic instability.

A PRIMARY TRANSCRIPT MAP FOR THE VAN DER WOUDE SYNDROME (VWS) CRITICAL REGION DERIVED FROM 900 kb OF GENOMIC SEQUENCE AT 1q32-q41.

B. C. Bjork¹, B. C. Schutte², M. Malik², K. Coppage², S. G. Gregory⁵, D. J. Scott⁵, L.Brentzell⁴, Y. Watanabe², M. J. Dixon⁴ and J. C. Murray^1,2,3

University of Iowa, Iowa City, ¹Genetics Ph.D. Program, Departments of ²Pediatrics and ³Biology
⁴Departments of Dental Medicine and Surgery, University of Manchester, Manchester, UK
⁵The Sanger Centre, Welcome Trust Genome Campus, Hinxton, Cambridge, UK

VWS is an autosomal dominant form of clefting with bilateral lower lip pits. The VWS gene was localized to a 1.6 cM region of 1q32 between D1S491 and D1S205. We constructed a sequence-ready BAC contig across the VWS critical region (VWSCR) using gene-based and anonymous STSs from our previous YAC contig. All STSs and BAC clones were shared with the Sanger Centre (U.K.) which developed a contig of 97 PAC clones over the same region. 11 overlapping clones from both contigs, spanning approximately 1.1 Mb, were selected for high-throughput genomic (HTG) sequence analysis. The complete sequence has been generated for 9 of the clones. This 1.1 Mb region is comprised of two contiguous sequences of 721 kb (contains the 350 kb VWSCR) and 129 kb that are separated by a gap of about 300 kb of which 150 kb is partially sequenced. These sequence contigs were analyzed and revealed novel polymorphisms, including an 8 kb deletion/insertion (Watanabe et al., submitted) and 11 STRPs. To identify genes in these sequences, we first used the computer program RepeatMasker2 to "mask" repetitive sequences, then performed BLAST homology searches against the non-redundant and dbEST databases in Genbank. Various gene prediction programs, utilized singly and in combination, with Genotator, were used to identify putative exons in the genomic sequence. These analyses revealed 6 genes previously mapped to the region, 9 novel genes, 8 putative gene fragments and 3 pseudogenes. As they are determined, exon sequences are screened for the presence of etiologic mutations in a panel of 97 affected individuals by SSCV analysis and direct DNA sequencing. To date, 31 normal sequence variants have been identified from this analysis of 92 exons, but no disease-causing mutations were observed. We continue to determine the expression patterns, transcript sizes and cDNA sequence of gene fragments to prioritize genes for mutation analysis.

APPLICATION OF SAGE TO GENERATE AN INTEGRAL 1p35-36 TRANSCRIPTION MAP IN NEUROBLASTOMA.

Rogier Versteeg, Huib Caron, Peter van Sluis, Ilja Roobeek, Kathy Boon

Dept. of Human Genetics, Academic Medical Center, University of Amsterdam, The Netherlands

Deletions of chromosome 1p35-36 are frequent in many tumor types. However, searches for a commonly deleted region have yielded conflicting results. In the best studied tumor, neuroblastoma, evidence exists that several different 1p35-36 subregions play a role. The 1p35-36 region encodes for an estimated one thousand genes, with many evident candidate tumor suppressor genes among them. RH mapping has identified many of these genes, and sequencing of the entire region is well under way. As a next step in understanding the role of this region, we set out to construct an integral expression map of all 1p35-36 genes in neuroblastoma.

We applied the SAGE (Serial Analysis of Gene Expression) technique to obtain an integral expression profile of neuroblastoma. Five SAGE libraries were constructed, two from neuroblastoma tumors and three from cell lines. We have presently sequenced 65.000 SAGE tags, each corresponding to and identifying a single transcript. The libraries are successfully used to identify downstream targets of the N-myc oncogene and other transcripts with a potential role in neuroblastoma.

We are presently analyzing the SAGE tags that are derived from transcripts of chromosome 1p35-36 genes. Starting from Human Gene Map 99, we use the CGAP SAGE database to identify reliable SAGE tags corresponding to genes and ESTs from this region. These tags are used to screen our 65.000 TAG neuroblastoma database. Pilot studies of specific subregions show that about 40% of the genes is expressed. This percentage will probably increase upon continued sequencing of the libraries. These maps can be compared with expression profiles of other tumors and tissues, which enables discrimination between 'household' genes and neuroblastoma specific genes. Further sequencing of the different libraries will eventually yield expression maps of different neuroblastoma subtypes and will identify 1p35-36 genes with a potential role in neuroblastoma pathogenesis.

SEQUENCE DETERMINATION OF HUMAN CHROMOSOME 1q22-q23 WHICH IS PARALOGOUS TO THE HLA REGION ON CHROMOSOME 6p21.3.

T.Shiina¹, N.Takishima¹, A.Shigenari¹, A.Ando¹, E.Kikkawa¹, K.Iwata¹, Y.Kuwano¹, Y.Kitamura¹, K.Watanabe², Y.Fukuzumi², M.Yamazaki², H.Tashiro², G.Tamiya¹, M.Kimura¹, E.Soeda³, T.Ikemura⁴, H.Inoko¹

¹Dept. Molecular Life Science, Tokai University School of Medicine
²Bioscience Research Lab., Fujiya Co., Ltd.
³RIKEN Gene Bank, The Institute of Physical and Chemical Research
⁴Genome Research, National Institute Radiological Science

The HLA region has paralogous genes or sequences in the 1q21-q25, 9q33-q34, and 19p13.1-p13.3 indicating segmental chromosome duplication during the course of evolution. In the last meeting, we have revealed that 1.4 Mb of the 1q22-q23 region around the CD1 genes contained a cluster of genes which had immunological importance such as SPTA1, IFI-16, FY, FCERIA and CD3Z so far. In order to clarify the molecular structure and gene organization of the 1q22-q23 region considered to be evolutionally related to the HLA region and also to investigate how large duplication occurred in genome, we have determined the 246 kb of nucleotide sequence of two PAC clones, RPCI 893N23 and 622B6 including the CD1B - CD1E, and FY1 - FCER1A genes, respectively. Nucleotide sequence determination was carried out by the shotgun method. Homology search and gene prediction of the sequence data with the aid of computer analyses revealed the existence of 5 expressed genes in total, namely 4 previously mapped genes (CD1B and CD1E, FY1 and FCER1A), 1 newly mapped gene (BL1A : anonymous cell adhesion molecule in 622B6) in the segment. The GC-content is 39% on average, which is justified by gene and LINE sequence densities observed in this region.

HIGH RESOLUTION MAPPING OF CHROMOSOME 1P DELETIONS IN HEREDITARY AND SPORADIC CRC CANCERS BY COMBINED HIGH DENSITY SINGLE NUCLEOTIDE POLYMORPHISM AND MICROSATELLITE ANALYSIS.

RB Chadwick¹, B Yuan¹, GA Bennington¹, CK Johnson¹, MW Stevens¹, LA Aaltonen², PT Peltomaki², A de la Chapelle²

¹Division of Human Cancer Genetics, The Ohio State University, Comprehensive Cancer Center, Columbus, Ohio, USA
²Department of Medical Genetics, Haartman Institute, University of Helsinki, Finland

The short arm of chromosome 1 is one of the most frequently deleted regions in the human genome in many different types of cancer. Genomic deletions in tumours are indicative of regions that potentially contain tumour suppressor genes. Traditionally, genomic deletions are mapped by loss of heterozygosity (LOH) microsatellite analysis of normal/tumour DNA pairs. However in hereditary nonpolyposis colorectal cancers (HNPCC) it is difficult to determine genomic deletions by microsatellite LOH analysis due to defects in DNA mismatch repair which cause microsatellite instability (MSI) in repetitive regions of the genome. Thus this study was undertaken to map a potential tumour suppressor gene on chromosome 1p in HNPCC and sporadic colorectal cancers by a combination of high-density single nucleotide polymorphism (SNP) and microsatellite analysis. Nineteen fluorescent microsatellite markers and 12 SNPs were amplified and analyzed for 24 HNPCC normal/tumor pairs and 8 sporadic CRC normal/tumor pairs. Loss of heterozygosity determination of SNPs was done by direct fluorescent sequencing of SNP PCR products and measurement and comparison of the respective heterozygote fluorescent intensity. Several new SNPs were discovered in the process. Microsatellite LOH determination was done by standard fluorescent ratio comparison of microsatellite normal/tumour alleles. Some sporadic tumors showed LOH for the entire telomeric part of chromosome 1p with a return to heterozygosity at approximately 30 to 35 cM from 1p tel. Some HNPCC tumors showed localized LOH in this region. More markers are being analyzed to more closely define this critical region. Integration of the critical region of LOH in the SNP and microsatellite data and determination of tumour suppressor candidate genes are facilitated by genome sequencing data from the chromosome 1 database.

POSITIONAL CLONING OF THE TRMA GENE.

Valentina Labay^1*, Dana Baron^1*, Tal Raz^1*, Hanna Mandel², Hawys Williams³, Timothy Barret⁴, Raymonde Szargel¹, Louise McDonald³, Adel Shalata¹, Kazuto Nosaka⁵, Simon Gregory³, Nadine Cohen¹

^*These authors equally contributed to the work

¹Department of Genetics. Technion-Israel institute of Technology, Bruce Rappaport Faculty of Medicine
²Department of Pediatrics, Rambam Medical Center
³The Sanger Center, Wellcome Trust Genome Campus Cambridge, England
⁴University of Birmingham, UK. 5 Kyoto Prefectural University of Medicine, Japan

Thiamine responsive megaloblastic anemia (TRMA), also known as Rogers syndrome, is an early-onset, autosomal, recessive disorder defined by the occurrence of megaloblastic anemia, diabetes mellitus and sensorineural deafness, responding in varying degree to thiamine treatment (OMIM #249270). The TRMA locus was previously narrowed from 16-cM to 4-cM (D1S2658 - D1S2851) on chromosomal region 1q23.3 (ref 1,2). In the present study, a physical map of the region was assembled. Two high resolution P1-artificial chromosome (PAC) contigs (196 and 13) including sixteen PAC clones, represented the minimum tiling path between D1S2658 (centromeric end) and D1S2851 (telomeric end). This clarified the order of known markers across the TRMA locus. Using the available sequence of the PAC clones nine new CA-repeat markers were identified and found to be polymorphic in the population tested. Genetic analysis was performed using these markers and full haplotypes were constructed for all informative individuals in 6 families previously described by us 1,2. Based on loss of homozygosity by descent and linkage disequilibrium, a new TRMA candidate region was obtained between the new markers TMG738C and TMG780. Recently, the TRMA locus has been further refined to a 1.4-cM region 3. This enabled us to narrow the TRMA locus to an approximately 400kb region, based on the size of the PAC clones spanning the new interval. A search for candidate genes in the critical region, using computer analysis, identified four known genes and two novel genes. One of these genes was found to be mutated in TRMA affected individuals. Results will be discussed at the workshop.

1 Neufeld et al. Localization of the gene for Thiamine-Responsive Megaloblastic Anemia syndrome on the long arm of chromosome 1 by homozygosity mapping. Am. J. Hum. Genet. 61, 1335-1341 (1997).
2 Raz et al. Refine mapping of the gene for Thiamine-responsive Megaloblastic Anemia syndrome in a 4cM region and evidence for genetic homogeneity. Hum. Genet. 103, 455-461 (1998).
3 Banikazemi et al. Localization of the Thiamine-Responsive Megaloblastic Anemia syndrome locus to a 1.4-cM region of 1q23. Mol. Genet. Metab, 66, 193-198 (1999).

TRANSCRIPTION MAP OF CHROMOSOME 1q42-43: CANDIDATE GENES FOR THE SYNDROME OF CONGENITAL HYPOPARATHYROIDISM, GROWTH, MENTAL RETARDATION AND DYSMORPHISM (HRD).

Ruti Parvari¹, Sharon Shatzki¹, Eli Hershkovitz² Rivka Carmi¹

¹Genetics Institute
²Pediatric Department, Soroka Medical Center and the Faculty of Health Sciences Ben Gurion University of the Negev, Beer Sheva, Israel

The syndrome of hypoparathyroidism associated with growth retardation and developmental delay and dysmorphism (HRD) is a newly described autosomal recessive congenital disorder with severe, often fatal consequences. Since the syndrome is very rare with parents of affected individuals being consanguineous, the disorder is presumed to be caused by homozygous inheritance of a single recessive mutation from a common ancestor. To localize the HRD gene, a genome wide screen using DNA pooling and homozygosity mapping was performed using apparently unlinked kindreds. We have previously mapped the HRD on chromosome 1q42-43 between markers D1S1540 and D1S2678 . Additional meioses identified during the past year helped in further narrowing this interval on the telomeric side to D1S235.

To identify the mutated gene causing the HRD we aimed at creating the transcription map in the HRD interval. 61 ESTs and genes that have been broadly mapped between D1S459 (251.2) and D1S304 (272) were analysed for their presence on the contigs created by the Sanger Center between D1S1540 and D1S2678. 21 ESTs and 5 genes were placed on the contigs by PCR amplifications using the PACs' DNA. Many of these ESTs were placed on the same PACs in several regions, either representing different parts of the same gene or clustered genes. The sum of this analysis suggests hotspots for transcription in 1q42-43 and provides candidates for HRD. Mutational analysis was started on one of the genes placed on the contigs.

MAPPING OF NOVEL AMPLICON 1q21-q25 IN HUMAN HEPATOCELLULAR CARCINOMA BY FISH INTERPHASE CYTOGENETICS.

Nathalie Wong¹, Paul Lai², Sui-Wah Lee¹, Joseph Lau², Philip Johnson¹

¹Department of Clinical Oncology at the Sir Y.K. Pao Centre for Cancer Centre
²Department of Surgery, The Chinese University of Hong Kong, SAR Hong Kong, China

Hepatocellular carcinoma (HCC) is a highly malignant tumour that is prevalent in Southeast Asia and sub-Saharan Africa. Risk factors associated include chronic viral hepatitis (types B & C), alcoholic cirrhosis and exposure to dietary aflatoxin. We have previously investigated the pattern of genomic imbalances in HCC by comparative genomic hybridization (CGH) and found an astoundingly high incidence of chromosome 1q copy number gain (72%; 48/67HCC samples). A minimal overlapping amplicon 1q21-q25 was also identified. Fine mapping of this novel amplicon to obtain a higher resolution on the amplified region was subsequently carried out by FISH interphase cytogenetics. Dual-labeled FISH analysis on CGH positive cases was performed using YACs on 1q21-q25 and 1p31. Relative signals scored between the q- and p-arm probes indicated considerable clonal variation in the amplification pattern. Although signal gain was often detected in about 50% of the nuclei examined, the most frequently amplified YAC was found be 955E11. Signal clustering of this probe was also observed in one case. Our present FISH analysis confirmed the presence of an amplicon 1q21-q25 in the genetic aberration of HCC. Delineation of this amplicon further pinpointed the affected region to the YAC 955E11, which maps to 1q21.1.

ANALYSIS OF 5 Mb REGION AT 1p36.1 HARBOURING A TUMOUR SUPPRESSOR LOCUS INVOLVED IN MYCN AMPLIFIED NEUROBLASTOMA.

Nicole Spieker, Mabel Beitsma, Peter van Sluis, Huib Caron and Rogier Versteeg.

Academic Medical Centre, Meibergdreef 15, 1105 AZ, Amsterdam, The Netherlands

Common genetic aberrations of neuroblastoma are deletions of the short arm of chromosome 1 (1p36) and MYCN amplification. Our deletion analysis of 25 tumor cell lines and 171 tumours strongly suggests that 1p harbours several tumour suppressor loci. Distinct loci are involved in MYCN single copy versus MYCN amplified neuroblastoma. Deletions in MYCN single copy tumours have a shortest region of overlap (SRO) of 20 cM at 1p36.3. MYCN amplified tumours have large deletions with an SRO of about 60 cM, from 1p36.1 till the telomere. This SRO is defined by the HMG17 gene (1p36.1), which was the most distal locus retained. Therefore, a suppressor gene associated with MYCN amplified tumours probably maps within a few megabases distal of HMG17. We mapped the breakpoint of the MYCN amplified neuroblastoma with the smallest 1p deletion between 56.6 and 57.2 cM from the top of chromosome 1. PFGE and Radiation Hybrid mapping was used to construct a 5 Mb physical map of this region. The map includes the region from 82.73 till 92.89 cR from the top of chromosome 1p. About half of it was isolated in P1 and PAC clones. The region turned out to harbour the genes FGR, SLC9A1, HMG17, EXTL1, AML2, RH, OP18 and 4 ESTs.
In order to identify a tumour suppressor gene involved in MYCN amplified neuroblastoma, we used two strategies. We searched for new genes within the 5Mb region using a series of PACs from the syntenic mouse region. We analysed the corresponding human and mouse PACs for conserved sequences. These sequences potentially represent new genes. In addition, a series of PACs and contigs from this region are being sequenced at the Sanger Centre. These data reveal a large number of new ESTs within the critical region.

The 5 Mb region was also screened for mutations in MYCN amplified neuroblastomas. The candidate tumour suppressor gene AML2 was analysed by SSCP, but no mutations were found in both MYCN single copy and MYCN amplified neuroblastomas. Furthermore, Southern blots of 200 patients and cell lines were hybridised with over 50 single copy probes isolated from the 5 Mb region, revealing a tumour specific rearrangement in patient N14. Currently, this rearrangement is being cloned and analysed.

Integration of gene function together with expression and mutation analysis is used for the selection of putative neuroblastoma tumour suppressor genes.

A PHYSICAL AND TRANSCRIPTION MAP OF 1q24.3-q31.1 AS A RESOURCE FOR IDENTIFYING HEREDITARY DISEASE GENES INCLUDING HPC1.

John Carpten¹, Christiane Robbins¹, Raman Sood¹, Tom Bonner⁴, Izabella Makalowska¹, Dietrich Stephan¹, Jeff Smith¹, Nyasha Scott¹, Mezbah Faruque¹, Heather Pinkett¹, Chris Graham¹, Tim Connors², Sharon Morgenbesser², Kui Su², Kathy Klingler², Greg Landes², Simon Gregory⁵, Hawys Williams⁵, Louise M. McDonald⁵, William Isaacs³ and Jeffrey Trent¹

¹Cancer Genetics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD, USA
²Genzyme Genetics Corporation, Framingham, MA, USA
³Johns Hopkins Hospital Brady Urological Clinic, Baltimore, MD, USA
⁴National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA
⁵The Sanger Centre, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK

Several hereditary disease loci have been genetically mapped to the chromosome 1q24.3-q31.1 interval including the HPC1 locus. A 20Mb YAC contig of the 1q24.3-q31.1 has been confirmed by STS content mapping and spans the interval flanked by D1S212 and AFMB330XE9. Over 300 BAC and PAC clones have been identified by PCR screening of commercially available library DNA pools with public and novel markers placed on the YAC contig. We have generated a complete 6Mb sequence ready BAC/PAC contig of the 1q25.2 interval by STS content mapping and chromosome walking. Approximately, 130 new STS's have been generated from this contig and shown to map to the 1q25.2 interval, as well as 18 novel simple sequence repeat polymorphisms which are being used for genetic refinement of multiple disease loci. BAC/PAC contig integrity has been confirmed by restriction fingerprinting and clones from this contig are being used as templates for human chromosome 1 genome sequencing. A transcription mapping effort has resulted in the precise localization of 18 known genes and 31 EST's by database searching, exon trapping, direct cDNA hybridization, as well as sample sequencing of BACs from the 1q25.2 region. An additional 11 known genes and ESTs have been placed within the larger 1q24.3-q31.1 interval. These transcription units represent candidate genes for multiple hereditary diseases including the HPC1 gene.

LINKAGE DISEQUILIBRIUM MAPPING OF A TYPE 2 DIABETES-LINKED REGION ON 1q21-q23

JK Wolford, RL Hanson, C Bogardus, M Prochazka

NIH/NIDDK, Phoenix, Arizona

Type 2 diabetes mellitus is a complex heritable disease. A genetic basis for the disease is especially evident in the Pima Indians of Arizona, who have the highest reported prevalence of this disease. We have previously demonstrated evidence for linkage to diabetes in the Pimas in a ~30 cM region on 1q21-q23 with the maximum multipoint LOD score (2.5) at D1S1677. This region has also been linked to type 2 diabetes in a Caucasian population (Diabetes 48:1175-1182, 1999). To narrow the area of linkage and identify potential candidate genes, we are evaluating densely spaced SNP markers by association analysis in 117 Pima sib-pairs discordant for diabetes and in 100 unrelated Pimas, including 50 affected and 50 unaffected subjects. We are utilizing a contig of 65 overlapping YACs from the CEPH Mega-YAC library that encompasses the region of linkage and spans approximately 34 cM between markers D1S442 and D1S452. Using a combination of automated sequencing, denaturing-HPLC, PCR-RFLP, and automated allelic discrimination technology, we have analyzed 47 SNPs. These SNPs include 13 previously reported markers and 34 novel, gene-specific markers, which were identified as polymorphic in Pimas. Our strongest association was found in four SNPs in the GIRK3 gene. GIRK3, located approximately 3 cM centromeric from the peak of linkage at D1S1677, encodes a G-protein-coupled inwardly rectifying potassium channel protein. The GIRK3 protein is known to regulate ion flow and hormone secretion in some tissues and thus represents a candidate gene for diabetes susceptibility. However, the GIRK3 SNPs only accounted for less than half of the linkage on 1q and for this reason we do not believe that it is the disease allele. We are investigating this gene further and constructing a BAC contig through the GIRK3-D1S1677 region to concomitantly identify additional candidate genes and locate novel SNPs for use in haplotype analysis.

CLONING AND CHARACTERIZATION OF GENES AT RECIPROCAL t(1;9)(p32.3;p21.2) TRANSLOCATION BREAKPOINT IN A PATIENT WITH NEUROBLASTOMA.

A Horii, HO Shiwaku, M Hoshi, H Tsuchiya, I Kamino and Y Kaneko

Department of Molecular Pathology, Tohoku University School of Medicine
Department of Pediatrics, Kumamoto University School of Medicine, Kumamoto
Special Reference Laboratory, Hachioji
Department of Cancer Chemotherapy, Saitama Cancer Center, Saitama, Japan

Neuroblastoma is a well-known malignant disease in infants, but its molecular mechanisms have not yet been elucidated. The most frequently observed a chromosomal aberration is loss of chromosome arm 1p in neuroblastomas, and this chromosome arm may harbor two or more tumor suppressor genes in this disease. We have established a cell line that was derived from lymphoblast of a patient with neuroblastoma. This patient has a constitutional reciprocal translocation t(1;9)(p32.3;p21.2). A contig spanning an approximately 800-kb consisting of ten BAC clones of minimal overlap was constructed and three overlapping clones that harbor the breakpoint were cloned. A cosmid library was also constructed with the DNA of the patient's cell line and the translocation breakpoint was further characterized in detail. We also performed cDNA screening and detected two transcripts from this region. Results of the characterization of the breakpoint as well as cDNA clones will be described in detail.

THE HUMAN CHROMOSOME 1 WEB SITE.

Tara C. Matise¹ and Peter S. White²

¹Rutgers University, New-Brunswick, NJ, USA
²Children's Hospital of Philadelphia, PA, USA

The Human Chromosome 1 Web Site (C1WS; http://compgen.rutgers.edu/chr1, mirrored at http://corba.ebi.ac.uk/Mapping/Chr1/) was launched in August, 1995. The initial purpose was to provide information on the Second Chromosome 1 Workshop, and the original site consisted of 5 web pages. Since then, the site has expanded substantially to provide resources for chromosome 1 via a data server and customized web links; facilitating interaction among researchers via a discussion forum and mailing list; providing information about and full on-line reports from the Chromosome 1 Workshops, and providing information for the general public.

The current site consists of 515 web pages and received an average of 3177 hits per month (2693 US, 484 UK) in the past year (7/98 to 6/99). A total of 7643 unique Internet sites (6077 US, 1566 UK) have accessed over 470 different documents during 16126 total visits (13451 US, 5809 UK) to the C1WS. The top 5 sites accessing the US C1WS most often are: fw1-jjnet.jnj.com, montespan.pasteur.fr, relay.lsl.co.uk, nat45.latino.web.co, and osmium.hgmp.mrc.ac.uk. Over 70 Internet sites have links to the C1WS.

The Data section of the C1WS contains 3 project-specific datasets: linkage map of chromosome 1, integrated map of 1p36, and ESTs mapped to 1q21. We are soliciting additional data for this page.

The discussion forum received 21 postings during 1998. Many of these were from lay people, most of whom have a family member or friend with a chromosome 1 deletion or translocation. These postings came from several countries, including: France, Italy, Norway, Singapore, and England. We personally responded to most of these postings, and in some cases we received positive feedback for our responses. We consider our "Information for the General Public" page to be a unique and valuable section of the C1WS and strongly encourage members of the chromosome 1 community to respond to any relevant postings.

PROGRESS OF CHROMOSOME 1 MAPPING AND SEQUENCING AT THE SANGER CENTRE.

Louise McDonald¹, Christine Bird² on behalf of the Sanger Centre Mapping and Sequencing Groups

Human Chromosome One ¹Mapping and ²Sequencing groups, the Sanger Centre, Wellcome Trust Genome Campus, , Hinxton, Cambridge, (UK)

The Sanger Centre has been funded to generate a comprehensive sequence map of human chromosome 1 as part of the international program to determine the sequence of the human genome. In close collaboration with the chromosome 1 community we are constructing a sequence map which will include all genes and other biologically important sequences.

In accordance with our landmark based strategy we have generated a marker density of >15 STS/Mb and used 5021 STSs (83% RH mapped) to identify 28897 large insert bacterial clones. Restriction digest fingerprinting of 22369 clones has produced an estimated 145.5 Mb of sequence contig coverage. Gaps in the bacterial clone map are being closed using a combination of de novo STS generation from contigs ends, utilizing unfinished and finished sequence homologies to TIGR GSS's, and by incorporating contigs from the whole genome fingerprinting of the GSC, St Louis.

Sequence production is based upon shotgun sequencing of 1.4-2.2 Kb subcloned M13 and pUCs. The generation of draft sequence utilizes only pUC clones and produces an estimated 3X coverage, 6X coverage established for full shotgun. Clones are sequenced using Big dye chemistries and in 384 well format. Significant improvements in throughput has been facilitated by using high density ABI 377's slab gels and by the introduction of ABI 3700 and MEGABASE capillary sequencing machines. As of 28th of June 1999 a total of 16.4Mb of finished sequence has been submitted to public databases. Abstract 1, this workshop, details analysis and annotation carried out on finished sequence prior to submission.

Release of chromosome 1 mapping and sequence data is in line with the Sanger Centre policy (http://www.sanger.ac.uk/Projects/release-policy.shtml). The current status of chromosome 1 mapping and sequencing, and unfinished and finished sequence, is available via the Sanger Centre chromosome 1 homepage at http://www.sanger.ac.uk/HGP/Chr1.