Gene structure of the human MET proto-oncogene Fuh-Mei Duh 1 , Stephen W Scherer 3 , Lap-Chee Tsui 3 , Michael I Lerman 2 , Berton Zbar 2 and Laura Schmidt 1 1 Intramural Research Support Program, SAIC Frederick, and 2 Laboratory of Immunobiology, National Cancer Institute-Frederick Cancer Research and Development Center, Frederick, Maryland 21702, USA; 3 Department of Genetics, the Hospital for Sick Children, Toronto, Ontario, M5G 1X8, Canada By direct sequencing of cosmids using primers designed from the known cDNA sequence, we identified 19 exons in the human MET proto-oncogene, and sequenced the corresponding 5’ and 3’ exon-intron junctions. By homology search in the database of the Washington University Genome Sequence Center (GSC), we identi- fied one additional exon. These 20 exons, together with a previously reported exon, bring the total exon number of MET to 21. Oligonucleotide primers were designed to amplify each exon and adjacent intronic sequences to permit examination of each exon for mutations. By restriction mapping, we assembled a 110 kb genomic contig that covered almost the entire MET proto- oncogene. This information is relevant for the screening of recently reported mutations of the MET gene which cause hereditary papillary renal carcinomas and for the search for additional mutations of the same gene which may play a role in the pathogenesis of common human carcinomas including carcinomas of the breast, ovary and pancreas. Keywords: MET proto-oncogene; HGF/SF receptor; renal carcinoma Introduction The MET receptor tyrosine kinase transduces motility, proliferation and morphogenic signals of the ligand hepatocyte growth factor/scatter factor (HGF/SF) in epithelial cells (Weidner et al., 1993). During embry- ogenesis, the MET receptor – HGF/SF pathway is required for normal muscle and liver development (Bladt et al., 1995). The MET proto-oncogene has been implicated in human neoplasia. Increased MET protein expression has been found in a number of common human carcinomas (Di Renzo et al., 1992, 1994). The human MET cDNA encodes a protein of 1408 amino acids (aa) consisting of a large extracel- lular domain containing 950 residues, a transmembrane domain (aa 951 – 973), and an intracellular tyrosine kinase domain (aa 1102 – 1351) (Cooper et al., 1984; Park et al., 1987). Alternative splicing at the 5’ end of the gene has been reported; this leads to a protein of 1390 aa (Ponzetto et al., 1991). The processed form of the MET protein consists of an alpha and a beta chain; the alpha chain consists of aa 1 – 307; the beta chain consists of aa 308 – 1408 or aa 308 – 1390 in the alternatively spliced form. Hereditary papillary renal carcinoma (HPRC) is a recently described, inherited form of human cancer characterized by a predisposition to develop bilateral papillary renal tumors (Zbar et al., 1994, 1995). Recent studies demonstrated missense mutations in the tyrosine kinase domain of MET in patients with HPRC (Schmidt et al., 1997). The cosegregation of these mutations with the disease clearly indicates that they predispose to the development of papillary renal carcinomas. Of particular interest, three of the mutations were located in codons that are homo- logous to codons in c-KIT and the RET proto- oncogene that are targets of naturally-occurring mutations (Hofstra et al., 1994; Nagata et al., 1995). The availability of a full length cDNA sequence of the MET gene, as well as genomic cosmids of the MET gene, made it possible for us to use a direct sequencing approach for the identification of exon-intron bound- aries. Primers were designed every 200 – 300 base pairs based on the known MET cDNA sequence. These primers were used to sequence genomic cosmids that contained portions of the MET gene. Discrepancies between the genomic and cDNA sequences pointed to exon-intron boundaries. The strategies used in the present work led to the definition of 20 exon-intron boundaries defining 20 coding exons. These exons are in addition to a 5’ noncoding exon identified by Gambarotta et al. (1994). In addition, we report 5’ and 3’ flanking intronic sequences at each of the 20 newly identified boundaries, and a panel of primer pairs suitable for amplifying and scanning the exons for mutations. Results and discussion We defined the exon-intron boundaries, and flanking intronic sequences of the MET proto-oncogene by directly sequencing three overlapping cosmids that cover a 70 kb genomic region. We identified 19 exons in the genomic region that corresponds to the MET cDNA sequence from nucleotide 1395 to 4626 bp. We were unable to obtain any cosmids that covered the region that corresponded to MET cDNA nucleotides 181 – 1395. However, by performing homology searches, we found the genomic sequence that corresponded to MET cDNA sequence 8 – 2458 in the GSC database (Genome Sequencing Center, personal communication). We identified an additional exon at position 181 – 1394 and confirmed the published exon at position 8 – 180 (Gambarotta et al., 1994). The 3’ and 5’ intronic sequences adjacent to each exon were defined (Table 1). Presented in Table 1 are the 20 nucleotides adjacent to each of the exon-intron Correspondence: L Schmidt Received 10 May 1997; revised 11 June 1997; accepted 11 June 1997 Oncogene (1997) 15, 1583 – 1586  1997 Stockton Press All rights reserved 0950 – 9232/97 $12.00