Mutations in the SMAD4/DPC4 Gene in Juvenile Polyposis. Howe JR, Roth S, Ringold JC, Summers RW, Järvinen HJ, Sistonen P, Tomlinson IPM, Houlston RS, Bevan S, Mitros FA, Stone EM, Aaltonen LA.Science 1998;280:1086-1088.
Summary: Investigators at the University of Iowa recently identified a gene locus (18q21.1) linked to autosomal dominant juvenile polyposis coli (JPC) (Am J Med Genet 1998;62:1129-36). The 18q21.1 locus includes two candidate tumor-suppressor genes, DCC and SMAD4. In the report in Science, the investigators sequenced genomic polymerase chain reaction fragments from DCC and SMAD4 using DNA from an affected person belonging to a large Iowa kindred with JPC. A 4-bp deletion was detected in exon 9 of SMAD4. Sequence analysis predicted that the 4-bp deletion causes a frame shift and a premature stop codon resulting in a truncated SMAD4 protein. Critical regions needed for normal SMAD4 function are lost. This same truncating mutation was found in all 13 affected people in the JPC kindred and in 4 of 26 people considered "at risk." None of seven spouses and none of 242 unrelated people had the SMAD4 mutation, thereby greatly reducing the likelihood of a polymorphism in this region.
Unrelated patients with JPC were also analyzed for SMAD4 mutations. Seven patients from two kindreds, one in Mississippi and one in Finland, had an identical 4-bp deletion. In two sporadic cases, additional deleterious mutations were noted in SMAD4: a 2-bp deletion in exon 8 and a 1-bp insertion in exon 5 of SMAD4. Both of these mutations were also predicted to create a premature stop codon and a truncated protein. SMAD4 mutations were not detected in four other affected patients, suggesting heterogeneity in the genetics of JPC.
Comment: SMAD4 is a key cytoplasmic protein responsible for transforming growth factor (TGF)-β ligand/receptor signal transduction from the cell surface to the nucleus (Nature 1997;390:465-71). TGF-β is a well-known autocrine growth inhibitor for a wide variety of cell types, including normal gastrointestinal epithelial cells. Thus, SMAD4 and other proteins in the TGF-β signaling pathway have been considered putative tumor-suppressor genes. Indeed, somatic mutations in SMAD4 have been described in up to 50% of pancreatic tumors (Cancer Res 1996;56:490-4) and 15% of colorectal tumors (Gastroenterology 1996;111:1369-72). Loss of other components of the TGF-β pathway have also been implicated in carcinogenesis (Cytokine Growth Factor Rev 1996;7:93-102) emphasizing the importance of this growth factor in intestinal biology.
Juvenile polyps are among the most common colonic lesions encountered in clinical practice of pediatric gastroenterology. Most are single polyps and are not considered preneoplastic (Gastroenterology 1993;105:689-700). Notwithstanding the assurance of this general observation, on rare occasion, carcinoma has been reported in solitary juvenile polyps (Dis Colon Rectum 1990;33:980-4). Adenomatous change can be found in 3% to 5% of juvenile polyps (J Pediatr 1989;114:593-6). A more frequent and vexing situation occurs in the child with multiple juvenile polyps, which is not an uncommon clinical occurrence. Thirty-nine percent of children with juvenile polyps have two or more lesions (J Pediatr 1989;114:593-6). The problem can be stated as such: If there is essentially no risk of adenoma formation and tumor progression in children with solitary juvenile polyps and there is a clearly defined incidence of at least a 20% risk in patients with well-defined JPC, what about the substantial number of children with multiple juvenile polyps? Using clinical data obtained retrospectively, Jass et al. defined juvenile polyposis (presumably carrying a risk of colorectal neoplasia) in children with more than five colorectal juvenile polyps, with juvenile polyps throughout the gastrointestinal tract, or with any number of polyps and a family history of JPC (Histopathology 1988;13:619-30). Using an analogous approach, Giardiello et al. suggested that patients who have three or more juvenile polyps or a family history of juvenile polyps should undergo surveillance for colorectal neoplasia (Arch Dis Child 1991;66:971-5). If the Jass criteria are rigidly applied to our own experience at Vanderbilt Children's Hospital, 17% of 63 children undergoing pancolonoscopy for polyp removal would merit surveillance because five or more polyps are present. Obviously, even more patients require surveillance if the criteria proposed by Giardiello et al. are used.
The report by Howe et al. in Science offers a hopeful prospect for the development of more precise criteria based on genetic information for identification of children with multiple juvenile polyps at risk for neoplasia. Patients with genetic defects in tumor-suppressor gene products such as the SMAD signaling proteins, other components of the TGF-β ligand/receptor system, or other tumor-suppressor genes can be appropriately screened while the intensity of invasive, expensive monitoring can be reduced in patients without defects. This approach is conceptually similar to that currently used in families with familial adenomatous polyposis coli; however, much additional molecular and clinical work must be undertaken to reach this meritorious goal in the diagnosis and treatment of JPC.
Department of Pediatrics; Division of Gastroenterology and Nutrition; Vanderbilt University School of Medicine; Nashville, Tennessee