No area of science has advanced our understanding of human disorders as rapidly as molecular genetics and biology. In recent years, the application of many molecular and genetic research techniques to inherited diseases affecting the gastrointestinal and hepatobiliary tracts has had a substantial impact on our understanding of these disorders. Such advances are significant, because they not only provide us with insight about the pathogenesis of these inherited diseases, but they also establish a framework for better understanding of other more common noninherited disorders. For example, the identification and characterization of the gene responsible for familial adenomatous polyposis coli has given us better insight into the complex process of tumor formation and the association between the common noninherited form of colorectal polyps and carcinoma (1). Therefore, discoveries in the molecular biology and genetic makeup of inherited disorders can have an impact on understanding noninherited disorders and may suggest new prospects for therapy.
The objective of this two-part series is to provide a comprehensive review of progress made since 1995 in our understanding of the molecular basis of inherited disorders of the gastrointestinal (Part I), and pancreatic and hepatobiliary (Part II) tracts. The emphasis is on specific breakthroughs in the discovery of genes responsible for a wide variety of disorders. In addition, in those disorders for which the gene is not known, recent advancements in chromosomal localization will be discussed. Finally, developments in the biology of disorders for which the defective protein has been previously defined will be reviewed.
Comprehensive tables are included that organize disorders on the basis of our current understanding of their molecular pathogenesis. Disorders are divided into three categories: those for which a) the gene(Table 1) or, b) the chromosome (Table 2), that contains the allele has been identified; and, c) disorders for which neither gene nor chromosome has been identified (Table 3). For the sake of completeness, several disease entities will be discussed that do not have a significant gastrointestinal phenotype but are included because their pathogenesis is based on an abnormal gene expressed in the digestive tract. Finally, those disorders of the hepatobiliary system that result in serious complications in the liver (i.e., hepatitis and liver failure) will be reviewed in Part II.
A question may be raised about whether genetic testing is available for any of these disorders. Although this specific question will not be answered for each disorder, several general rules on genetic testing may be useful to remember. First, testing of common disorders that result from a limited number of mutations in a single gene is increasingly provided by commercial groups. Second, genetic testing of rare disorders is occasionally provided by research groups that have a specific interest in the disorder. Generally, these researchers can be identified by searching through recent publications containing reports regarding the disorder. Disorders that are less amenable to testing are those that are caused by mutations in more than one gene or by multiple mutations in the same gene. However, a patient with a more severe, a milder, or an unusual clinical phenotype, compared with that of the typical patient with the disorder, is frequently of interest in research. Therefore, genetic testing is of particular interest in these groups of patients.
Finally, rapid advancements made in genetic research make keeping abreast of the field very difficult. Although several standard textbook resources provide an outstanding background for a wide range of these clinical disorders(2-6), the brisk pace of genetic research means that textbook content is rarely up-to-date with the advancements made in the field. The World Wide Web, however, is a medium that is amenable to this rapid pace. Several databases on the Web that are sponsored by the National Institutes of Health and the National Library of Medicine (Bethesda, Maryland, U.S.A.) are particularly useful. PubMed is an on-line service that provides up-to-date access to periodicals and textbooks(7). More importantly, Online Mendelian Inheritance in Man (OMIM) is an invaluable and frequently updated resource that summarizes those publications that pertain to a particular genetic disorder(8). The frequent use of these and other sites available on the World Wide Web should simplify the seemingly daunting task of keeping up-to-date with changes in the field of genetic disorders.
Disorders of Digestion, Secretion, and Absorption
Hereditary hemochromatosis (OMIM 235200) is an autosomal recessive disorder, characterized by excessive iron absorption by the intestine and secondary multiorgan failure related to excessive tissue iron content. Hemochromatosis is probably the most common genetic disorder among people of European ancestry, with an estimated carrier frequency of approximately 1 in 8 (9). The high incidence may have resulted from a selective advantage of enhanced iron absorption conferred by the heterozygous state. Support for this hypothesis comes from results of studies that showed that heterozygote people have an elevated serum concentration of iron, transferrin saturation, and ferritin concentration when compared with those levels in control subjects (10). These data support the notion that the heterozygote state may be a selective advantage, especially to pregnant women in whom iron deficiency anemia is a common and harmful complication.
The hereditary hemochromatosis locus was mapped to the major histocompatibility region on chromosome 6p. Last year, the defective gene responsible for hemochromatosis was identified when a careful physical map of the region was developed that narrowed the locus to a 250-kb stretch of genomic DNA (11). Sequencing of this span of DNA identified at least 15 genes, including 12 histone genes and a major histocompatibility complex class I-like gene named HLA-H. A cysteine-to-tyrosine mutation of residue 282 (Cys282Tyr) was identified in the HLA-H gene in 83% of 179 people with clinical evidence of hemochromatosis. A second mutation was identified that alters residue 63 (His63Asn) from a histidine-to-asparagine and accounts for a minority of hemochromatosis patients. Thus, approximately 87% of patients with hemochromatosis were either homozygote for Cys282Tyr or were heterozygote for both the Cys282Tyr and His63Asn mutations. Results in additional studies have confirmed those in this initial report by identifying the homozygote Cys282Tyr mutation in 82% of patients with hemochromatosis and in none from a group of 193 normal subjects(12). Nearly 7% of hemochromatosis patients were heterozygote for the Cys282Tyr mutation, with each containing theHis63Asn mutation on the opposite allele.
Although the precise function of the HLA-H protein has yet to be determined, the common Cys282Tyr mutation is believed to alter significantly the protein's function, because the cysteine residue at position 282 forms a cysteine disulfide bond that is critical for maintaining the protein's structure. Immunohistochemical data suggest that the HLA-H protein is present in the intracellular compartment of crypt cells of the intestine (13). These data suggest that theHLA-H protein is unlikely to be the iron transporter located on the apical brush border membrane of the small intestinal enterocyte. In fact, the intestinal iron transporter was recently identified using well-established methods of expression cloning (14). The protein(divalent cation transporter; DCT1) transports various cations, including iron, copper, and zinc, and is expressed in a variety of tissues, including the intestine. It is interesting that genomic mapping of an established mouse model line (mk) with autosomal recessive microcytic anemia identified a mutation in the murine isoform ofDCT1(15). Nevertheless, whether theHLA-H protein interacts with the iron transporter, serves as an iron sensor or has another critical role in intracellular iron transport is unknown.
Included in the analysis of Beutler et al. were four infants with the poorly understood pediatric counterpart of the disease, called neonatal hemochromatosis(12). In contrast to hereditary hemochromatosis, the clinical finding of the neonatal form is fulminant liver failure during the first several weeks of life, in the presence of excessive tissue iron content. A normal HLA-H genotype was identified in two of the neonatal hemochromatosis patients, and a third was heterozygote for theHis63Asn mutation. Although other novel mutations in theHLA-H gene were not evaluated, these data support the notion that the neonatal form of the disease is nonallelic with the hemochromatosis gene and imply that genetic testing is not useful in the evaluation of these patients.
Further genetic evidence that distinguishes hemochromatosis from its neonatal counterpart was described by Verloes et al., who identified two families with several members who were half-siblings and had neonatal hemochromatosis (16). These data shed doubt on the proposed autosomal recessive pattern of inheritance of neonatal hemochromatosis, suggesting instead either an autosomal dominant, complex, mitochondrial, or a noninherited disorder.
The presence of only two easily tested mutations will simplify genomic testing of presymptomatic patients with hereditary hemochromatosis. Such testing will be particularly useful, because the complications associated with hemochromatosis are related to long-term exposure to excess iron, and the identification of the disorder in presymptomatic people at risk should allow for earlier and more effective prevention.
Genetic testing is also available for the X-linked disorder Menkes' disease (OMIM 309400), which results from mutations in a copper transport(mnk) present on intestinal mucosa and other tissue, including that of the central nervous system (17). Previous attempts to treat infants with Menkes' disease with intravenous copper sulfate (or ethylenediaminetetraacetic acid [EDTA]) resulted in very poor clinical outcomes. Two unrelated patients were treated with copper-histidine and experienced an unusually mild clinical course. Analyses of their mnk genes revealed two single-base-pair deletions, resulting in a severely truncated and defective protein. The unusually mild symptoms that occurred in these patients despite severe mutations, suggest that copper-histidine treatment may improve clinical outcomes.
A useful animal model for Menkes' disease is the mottled-brindled(MobrJ) mouse that has a naturally occurring mutation in themnk gene that results in death during the first several weeks of life. These mice were previously used to establish that the copper retained within the intestinal enterocytes (secondary to failure to transport across the basolateral membrane) is bound to the soluble protein metallothionein. Kelly and Palmiter studied offspring from mice with targeted disruption of the metallothionein gene bred to a MobrJ genetic background and found that the double-knockout resulted in embryonic death by day 11(18). These data infer that metallothionein proteins play a critical role in decreasing the toxicity of copper excess in placenta and other tissues that have impaired excretion of cellular copper. Further investigations will be needed to determine the potential therapeutic usefulness of metallothionein proteins in patients with Menkes' disease.
Our understanding of the biology of intestinal ion transport is limited when compared with our knowledge of the complex process of bile acid transport. The enterohepatic circulation of bile acids involves a multistep process of hepatocyte and enterocyte transport of a diverse group of bile acid species. The first step to understanding the molecular genetics of the intestinal portion of this process is now possible because of the cloning of the human ileal Na+/bile acid cotransporter (SLC10A2) gene by Dawson (19).
Primary bile acid malabsorption (OMIM 601295) is a rare autosomal recessive disorder caused by a defect in the ileal transport of bile acids. Patients with this disorder have diarrhea, steatorrhea, failure to thrive, and low plasma levels of low-density lipoprotein cholesterol. Oelkers et al. screened SLC10A2 for mutations in a patient with bile acid malabsorption, using the single-strand conformational polymorphism method(20). Four mutations were identified including three missense mutations (A171S, L243P, T262M), and a splice-site defect in exon 3. In vitro analysis in COS cells revealed that of the missense mutations, none alters the proper targeting of the protein to the cell membrane. However, bile acid (taurocholate) transport by SLC10A2 was impaired if the protein contained either the L243P or theT262M mutation. The common A171S mutation (present in 28% of the population) appears to represent a benign polymorphism and has no consequence in bile acid absorption, at least when tested in vitro. In the bile acid transport gene of a patient with Crohn's disease, Wong et al. surprisingly identified a heterozygote mutation (P290S) that abolishes taurocholate transport (21). The defective protein is not believed to play a significant role in the pathogenesis of Crohn disease. Nevertheless, whether homozygous or heterozygote carriers of the A171S or P290S mutation have normal cholesterol homeostasis and bile acid metabolism awaits further investigation.
A membrane protein distantly related to the bile acid transporter is the Na+/glucose cotransporter gene (SGLT1). Thirty-three novel mutations in SGLT1 have been identified in patients with the autosomal recessive disorder glucose-galactose malabsorption (OMIM 182380) (22,23). Severe, life-threatening diarrhea develops in patients shortly after birth while they are fed a glucose-, galactose-, or lactose-based diet. Most are missense mutations, and functional assays of those with the Xenopus heterologous expression system identified either impaired sugar transport by improper trafficking of the transporter to the plasma membrane or disruption of transport activity, or both. These naturally occurring mutant transporters will provide a tool to investigate further the biology of the malabsorption syndrome and of the biosynthesis, trafficking, structure, and function of the SGLT1 cotransporter. Although prenatal diagnosis was successfully performed on a family at risk for glucose-galactose malabsorption, the apparent absence of a prevalent mutation makes such screening technically difficult and time consuming (24). Because of the complexity of genomic screening, careful small bowel biopsy and histologic evaluation, breath hydrogen test, and dietary manipulations may suffice in establishing the diagnosis.
Occasionally, studying the genetics of a rare disorder may provide important information on the biology of more common clinical problems. Such is the case with glucose-galactose malabsorption and water transport in the intestine. Oral rehydration solutions composed of Na+ and glucose remain the mainstay of treatment for diarrheal disorders throughout the world. The mechanism by which these substrates aid in water reabsorption in the small intestine has been poorly understood. Recently, Loo et al. provided direct evidence that the SGLT1 protein also serves as a type of water channel and may account for the movement of water across the intestinal membrane barriers (25). Using the well-establishedXenopus oocyte expression system, they identified substrate-(Na+-glucose-) induced water movement through the SGLT1 protein that was blocked by the competitive inhibitor phlorizin. Further studies identifying the precise mechanism of water transport will be aided by using the naturally occurring glucose-galactose malabsorption mutations shown to alter Na+ and glucose transport(22,23).
Monosaccharide transport on the apical membrane of the enterocyte occurs either through the Na+/glucose-galactose cotransporter(SGLT1), or the high-affinity fructose transporter(GLUT5). Fructose malabsorption has been implicated in the common noninherited toddler's diarrhea and in the rare autosomal recessive disorder isolated fructose malabsorption(26,27). Glucose transporter-5 (GLUT5) represents the fructose transporter and is the obvious candidate gene for isolated fructose malabsorption (28). Rand analyzed eight patients, including the original kindred with presumed isolated fructose malabsorption, for mutations in the GLUT5 gene(29). Analysis of its coding sequence and intron-exon boundaries failed to identify a specific mutation suggesting that eitherGLUT5 is not defective or that the alterations identified in the yet to be characterized 5′-upstream promoter region accounts for the disorder. Linkage analysis examining the inheritance pattern of this specific locus has not been performed and may shed further light on the validity of the candidate gene and on the actual inheritance pattern of the disorder.
The autosomal recessive disorder congenital chloride diarrhea(OMIM 214700) is associated with severe neonatal secretory diarrhea that frequently requires longterm parenteral fluid replacement. An earlier linkage analysis of several families identified the chloride diarrhea locus to a 450-kb region of chromosome 7q31 located in the vicinity of the cystic fibrosis gene CFTR(30). The gene down-regulated in adenomas (DRA) represented a viable candidate for chloride diarrhea because of its chromosome localization and its expression in colonocytes. Furthermore, the gene encodes a protein with physical characteristics of a membrane protein as well as transport function(31).
Höglund et al. identified several mutations in the DRA gene in patients with chloride diarrhea. An identical homozygoteΔVal317 mutation was discovered in most Finnish people, implicating the mutation as a founder effect in Finland, where chloride diarrhea is common (32). In addition, missense(His124Leu) and frameshift (344delT) mutations where identified in other chloride diarrhea patients of eastern European origin. Therefore, genetic testing will be easier in a patient of Finnish inheritance and more cumbersome in others. Furthermore, analysis of the precise transport function of this protein and the effects that the missense mutation has on its transport characteristics have not been described.
Monosaccharides are the product of complex carbohydrate digestion by salivary and pancreatic amylases and by membrane-bound sucrase-isomaltase and lactase-phlorizin enzymes of the small intestine. Congenitalsucrase-isomaltase deficiency (OMIM 222900) is a recessively transmitted disorder that appears in early infancy with symptoms of osmotic diarrhea caused by intolerance of a sucrose- and maltose-based diet. Although the sucrase-isomaltase gene has been cloned for several years, a mutation attributable to sucrase-isomaltase deficiency was only recently identified(33). A missense mutation (Gln1098Pro) in the gene encoding the membrane-bound enzyme sucrase-isomaltase results in improper folding of the enzyme complex. Analysis of in vitro cells and of patients' intestinal biopsy specimens provided direct evidence that inappropriate folding of the protein interferes with its processing from the endoplasmic reticulum to the Golgi apparatus and eventually results in protein degradation by a poorly understood quality control system.
The cellular degradation of a mutant protein is not unique to sucrase-isomaltase deficiency and has been described in many disorders. For instance, in cystic fibrosis (OMIM 219700) the commonΔF508 mutation of the CFTR protein does not alter the protein's transport function but instead results from the mutant protein's failure to reach its normal destination in the plasma membrane(34,35). Such improper targeting is caused by misfolded proteins that move from the endoplasmic reticulum by proteasomes for degradation (35). Immunohistochemical analysis of biopsy specimens from patients with a mutation that causes a trafficking defect may be a useful diagnostic tool. For instance, if most mutations of a particular gene result in improper targeting of the protein to the plasma membrane, cytochemical staining showing an absence of the protein on the membrane or a diffuse pattern of cytoplasmic staining would be consistent with the disorder.
Mucosal Gastrointestinal Disorders
Our understanding of the genetic basis of celiac disease (OMIM 212750) and inflammatory bowel disease (IBD) has not progressed rapidly in comparison with understanding of other disorders, in part because the complex nature of the gene's inheritance makes results of standard linkage analysis less reliable. However, new approaches to genetic mapping of complex disorders have begun to provide insight into the molecular and genetic basis of these relatively common disorders.
The diagnosis of celiac disease can be established by identifying villous atrophy and the presence of elevated antiendomysial immunoglobulin (Ig) A antibodies. Gliadin is the component of gluten that activates an immunemediated destruction of the intestinal epithelium. The precise mechanism of this destruction and the role of antiendomysial antibodies is not well understood. Anti-endomysial IgA antibodies were recently shown to recognize the endogenous protein tissue transglutaminase (36). Transglutaminase enzymes activate cross-linking reactions between various proteins (acceptors) that contain a specific type of lysine residue and a finite number of proteins (donor) with a certain glutamine group. Gliadin served as an effective donor protein and is cross-linked by the transglutaminase enzyme. The investigators (Dieterich et al.) raised the possibility that the formation of gliadin-transglutaminase complexes may form new antigenic targets that cause an immune-based destruction of the intestinal epithelial layer.
Despite the identification of transglutaminase as a potential autoantigen, only those people who are genetically predisposed to celiac disease will mount an immune response that destroys the epithelial cells. This genetic predisposition was linked earlier to the HLA region of chromosome 6. However, several other non-HLA regions were recently linked to celiac disease, including chromosome 6p (lod score, 4.66), 11p (lod score, 3.92), 7q (lod score, 2.99), 22 (lod score, 2.69), and 15q (lod score, 2.12)(37). In general, lod scores of greater than three indicate linkage of chromosomal markers to the inheritance of a disorder. In celiac disease, the strongest association is with chromosome 6p; however, the region identified is actually distinct from the HLA loci. This non-HLA region of chromosome 6p does not contain an obvious candidate gene, including transglutaminase, suggesting that a novel gene in this region increases susceptibility to the disorder. Weak linkage was also established to a region of chromosome 15q, a previously identified locus for insulin-dependent diabetes mellitus, a disorder associated with celiac disease.
The inheritance pattern of IBD is even less well understood than that of celiac disease. Complex inheritance indicates the role of at least two genes(or loci) in the development of the disorder. The first successful genomic linkage analysis of Crohn's disease (OMIM 266600) took into consideration its complex pattern of inheritance (38). Initial attempts using 270 locus markers in 25 families performed with linkage parameters set for an autosomal recessive disorder failed to establish significant linkage to any site. When the loci markers were examined for identity-by-descent in affected sibling pairs, 4 of 270 markers linked to three sites on chromosome 16 and to a single site on chromosome 11. Analysis of 53 additional families demonstrated linkage only to chromosome 16, where 7 out of 8 additional markers within this region also linked with the disease. This region is called IBD1, and is located in the pericentromer region of chromosome 16q12-q13. This area of the chromosome contains several good candidate genes, including α-integrins, CD19 antigen, and the interleukin-4 receptor.
Ohmen et al. independently confirmed the location of a moderate susceptibility locus at chromosome 16 and performed linkage with a more extensive list of markers in this specific region of the chromosome(39). The results more precisely define the locus and suggest further that it is particularly useful in non-Jews and not among Ashkenazi Jews with Crohn's disease. In results of both studies,ulcerative colitis (OMIM 191390) failed to map to this particular region.
The analysis of Ohmen et al. was limited to chromosome 16, but a second more complete analysis was performed by Satsangi et al. who screened 260 microsatellite markers of 89 pairs of affected siblings with IBD and their relatives (40). Overall, they demonstrated linkage to chromosomes 7, 3p21.2, and 12p13.2-q24.1. The region of chromosome 16 previously identified by Hugot (38) and subsequently confirmed by Ohmen et al. failed to demonstrate significant linkage in the patients studied. This apparent discrepancy may have occurred because the microsatellite markers in chromosome 16q12-13 used by Satsangi et al. were not in the immediate vicinity of merkers used by the other groups.
Nevertheless, the identification of chromosome 3, 7, and 12 as susceptibility loci for IBD raises the possibility that several candidate genes located in one or more chromosomes may influence the development of IBD. The highest lod score was obtained for chromosome 12p and was designated theIBD2 locus. Several interesting candidate genes lie within the vicinity including CD9 and CD27 receptors, guanine nucleotide binding protein-3, and tumor necrosis factor receptor(TNF-1). Similarly, chromosome 3p also contained several potential genes of interest including inhibitory guanine binding protein-α,transforming growth factor-β-receptor,β1-catenin, and the DNA-mismatch repair enzyme,MutL (Escherichia coli) homologue-1 (hMLH-1). Several of these genes are particularly important, especiallyinhibitory guanine binding protein-α, which was shown in results of knockout mouse experiments to result in inflammation of the intestine that resembles ulcerative colitis (41).β1-Catenin and MLH1 are also of particular interest, because people with long-standing ulcerative colitis are predisposed to colonic cancers, and the role of these genes in the development of spontaneous colorectal cancer and hereditary nonpolyposis colorectal cancer (OMIM 120435) has been well defined (discussed later). Finally, the region on chromosome 7 that demonstrated linkage to IBD is located near the intestinal gene mucin-3, which was previously implicated in IBD(42).
The previous association between Crohn's disease and the HLA locus suggested a potential role of a gene(s) in this region to the pathogenesis of the disorder. This association was recently confirmed by Plevy et al. who identified an allele within the TNF locus that was associated with the Crohn's disease phenotype (43). The alleleTNFa2blc2d4el represents a microsatellite region of theTNF-α gene. The possible effect that this particular allele has on Crohn's disease was not demonstrated, but the investigators suggested that mutations in the non-coding region of the TNF-α gene may influence the expression of the cytokine.
Genetic analysis with more extensive chromosomal markers and sequencing of candidate genes is required before the causative gene within each locus can be identified. However, a thorough understanding of the genetics of IBD will require not only the identification of each of these genes, but also knowledge of the biologic interaction of these genes with one another.
Hermansky-Pudlak syndrome (OMIM 203300) is a well-documented, autosomal recessive disorder with the clinical features of IBD, albinism, interstitial pulmonary fibrosis, and cardiomyopathy. The cellular organelle isolated from these people have distinct features, including defective lysosomes and melanosomes. Hermansky-Pudlak syndrome is especially common in Puerto Rico but has been reported in other countries as well. Earlier linkage analysis identified chromosome 10q23 as the Hermansky-Pudlak syndrome locus(44). A region of genomic DNA of approximately 180 kb was believed to contain the candidate gene; and through exon trapping, DNA sequencing, and complementary DNA (cDNA) selection, a gene named HPS was identified (45). Several nonsense mutations(A441X, P324X, and a 16-bp duplication in exon 15) were identified in the different Hermansky-Pudlak syndrome populations. Furthermore, the naturally occurring mouse model pale ear (ep) was determined to contain nonsense mutation in HPS(46). This gene is expressed in various tissues and appears to encode a novel transmembrane protein of unknown function. The mechanism underlying IBD in people with Hermansky-Pudlak syndrome and the role of the HPS protein in sporadic IBD remain unclear. However, the failure of IBD linkage studies to identify chromosome 10q23 suggests that defects in HPS are specific for Hermansky-Pudlak syndrome but do not contribute to the common nonsyndromic form of IBD.
In summary, significant strides have been made in elucidating the genetic factors that contribute to the development of IBD. Their complex inheritance pattern implies that multiple genes working by themselves or with others contribute to the development of IBD. This complexity has also been confirmed by the various natural and knockout mouse models in which IBD develops. More specifically, deletion of interleukin-2, interleukin10, Gαi2, N-cadherin, α or β T-cell receptor, and SCID mice reconstituted withCD+CD45RBhigh results in various forms of intestinal inflammation that resemble Crohn's disease or ulcerative colitis(47,48). Therefore, the multigenic nature of Crohn's disease and ulcerative colitis and patient heterogeneity may be attributable to the multiple susceptibility sites identified in results of linkage and knockout studies.
Intestinal Dysmotility Disorders
Intestinal dysmotility disorders are a relatively large group of poorly understood inherited diseases that affect either neuronal or muscle cell migration, differentiation, or function. Clearly, Hirschsprung's disease (HSCR) (OMIM 142623) is the most common and extensively studied of this group of disorders. Soon after birth, HSCR patients usually have obstructive symptoms that include constipation, and analysis of rectal biopsy specimens shows an absence of enteric ganglia. Hirschsprung's disease is a relatively common disorder with a complex pattern of inheritance that is caused by abnormalities in several proteins, includingrearranged during transfection (Ret), glial cell line-derived neurotrophic factor (GDNF), endothelin-B receptor (Ednrb), and endothelin-3 (Edn-3)(49).
Linkage analysis of a large North American Mennonite kindred identified a missense mutation of residue 276 (Cys276Thr) in the endothelin-B receptor (Ednrb) (50). A similar phenotype was identified in a mouse model with selective disruption of the Ednrb gene and in the naturally occurring piebald-lethal mouse model that contains a mutation in the Ednrb gene (51). However, analysis of the Ednrb gene in a non-Mennonite population identified only eight heterozygote mutations out of 223 patients (4%) withHSCR, suggesting that these individuals may have an additional modifying mutation in a yet unidentified gene(52-54).
Endothelin-3 (Edn-3) is the soluble ligand for the endothelin-B receptor and previous reports with gene disruption studies, and a naturally occurring mouse model (lethal spotting) implicated this protein as a potential cause of HSCR(55). Last year, Hofstra et al. identified for the first time a homozygote mutation (Phe159Cys) in the Edn-3 gene in a single kindred with the combination of aWaardenburg's syndrome type 2 and HSCR phenotype (OMIM 277580) (56). This mutation presumably alters a residue that is critical for proper posttranslational processing of the precursor(prepro) form of endothelin.
Congenital central hypoventilation syndrome (OMIM 209880) is a rare disorder that is frequently associated with HSCR (16%). Because the hypoventilation syndrome presumably results from abnormalities of cells in the neural crest, Bolk et al. suggested that it may be allelic withHSCR. The attempt at identifying a mutation in Ret in 14 probands with hypoventilation syndrome was unsuccessful(57). However, a single heterozygote frameshift mutation in Edn-3 was identified in one patient (58). Other investigators who analyzed 17 sporadic cases of classic short-segmentHSCR did not identify further mutations in Edn-3, suggesting that defects in the gene are important in the few patients who have hypoventilation syndrome or Waardenburg's syndrome type 2.
Moore et al. (59), Pichel et al.(60), and Sanchez et al. (61) independently performed gene disruption analysis of GDNF, and determined that homozygote mice had an absence of parasympathetic enteric neurons in the entire small and large intestine. Homozygote mice also had renal and ureteral agenesis, whereas a single group identified a range of subtle renal abnormalities (60). The phenotypic similarities between the GDNF knockout mice and those in a previously described Ret knockout model raised the possibility thatGDNF represented the elusive Ret ligand(62). Treanor and Goodman confirmed that GDNF binds the protein GDNF-α, which in turn forms a complex with the tyrosine kinase receptor Ret(63). Glial cell line-derived neurotrophic factor-α is a membrane protein with an extracellular domain that is believed to bind GDNF and to increase its interaction with Ret. Salomon et al. identified three mutations in the GDNF gene from a group of 173 HSCR patients (1.7%)(64). These mutations were either not essential or could not solely account for the HSCR phenotype, suggesting that mutations in the GDNF gene may contribute or may modify yet another unidentified gene locus. Such a locus may be represented by the GDNF binding protein (GDNF-α) RETL2, or the currently undefined gene Dom that was disrupted in a third mouse model ofHSCR(64,65).
Iwashita et al. identified five novel extracellular missense mutations of the Ret gene in HSCR patients which significantly reducedRet transport and its transforming capabilities(66). A single mutation that results in a mild reduction in Ret transforming activity was associated with the milder phenotype of short-segment HSCR, whereas the more severely abnormal mutations were associated with the long-segment HSCR phenotype. These data suggest that genotype-phenotype comparisons are possible at least in some patients with Ret mutations. Such comparisons will clearly be difficult for other alleles like Ednrb, for which even among siblings, the identical mutations have different phenotypes and suggest that other yet unidentified modifying loci contribute to the penetrance of the disease and its eventual phenotype (50).
Taken together, these data indicate that HSCR is a very complex disorder that may result from defects in several genes. This complexity is caused by the process of neural crest cell migration and differentiation and the roles that various proteins play in this process. Overall, unraveling the mechanism of how these various receptors and ligands mediate the intricate process of neural crest cell migration and differentiation will undoubtedly be the forthcoming challenge in this interesting area of research.
The complicated genetics of HSCR may be a good indication of what lies ahead in understanding the genetics of the heterogenous disorders of intestinal pseudoobstruction. One such disorder, neuronal intestinal dysplasia (X-linked) (OMIM 300048), was recently mapped to chromosome Xq28(67). Patients with this rare disorder have short bowel syndrome, malrotation, pyloric stenosis, and abnormalities of the argyrophil neurons of the myenteric plexus. It is interesting that in a recent report of a patient with hydrocephalus and HSCR, a mutation was identified in the L1 cell adhesion molecule (L1CAM) (68). Mutations in this neuronal cell adhesion molecule have been identified in patients with X-linked hydrocephalus (OMIM 307000). Based on chromosomal localization and clinical phenotype, the L1CAM protein represents the prime candidate gene for X-linked neuronal intestinal dysplasia. Moreover, the L1CAM gene may be an important gene modifier that accounts for the higher incidence of HSCR in males. Further genetic testing is required in a larger population of patients before this can be concluded.
Barone drew parallels between potential defects in the migration and maturation of neuroblasts of the neural crest as a potential cause for bothHSCR and a non-X-linked form of neuronal intestinal dysplasia type B (OMIM 601223) (69). However, their analysis of two kindreds failed to show linkage to the Ret locus frequently mutated in HSCR. Other potential loci that were not investigated include GDNF, GDNF-α, Ednrb and Edn-3, all of which have been implicated in HSCR(70). Linkage analysis at these specific locus should be useful for screening these candidate genes.
The role of the homeobox gene Enx (Hox11L1), which is expressed in myenteric and submucosal neurons of the colon and ileum, was investigated by Shirasawa et al. in a knockout mouse model(70). Mice homozygote for a targeted disruption of theEnx gene had a high mortality rate after weaning that was associated with toxic megacolon. These Enx (-/-) mice have pathologic findings that strongly resemble neuronal intestinal dysplasia in humans. InEnx (-/-) mice, the number of neurons per ganglion was increased in the colon and reduced in the ileum when compared with the number in wild-typeEnx (+/+) mice. In addition, the Enx (-/-) mice had hypertrophy of the myenteric ganglia. As a transcriptional factor,Enx would be expected to alter the transcriptional expression of other genes in the myenteric neurons. Further studies in humans should determine whether neuronal intestinal dysplasia is predominately caused by a defect in the Enx gene or of its unknown downstream targets of transcription.
Several years ago, mice with a targeted disruption of the neuronal nitric oxide synthase (NOS1) gene exhibited a clinical phenotype that resembled infantile pyloric stenosis (OMIM 179010)(71). This result led several groups to investigate the possible connection between NOS1 and pyloric stenosis(72). Because NOS1 represents the best candidate gene, Chung performed linkage analysis of 27 families with pyloric stenosis and confirmed the nNOS locus as a possible susceptibility locus (72). As expected, locus heterogeneity was also identified and is consistent with the pyloric stenosis complex pattern of inheritance. Despite good evidence implicating the nNOS locus in the pathogenesis of pyloric stenosis, an actual mutation in the NOS gene has not been identified in any patient. Moreover, scanning of the entire genome by linkage analysis looking for other pyloric stenosis loci has not been reported; therefore, it is possible that other potential susceptibility markers have yet to be identified.
Analysis in eight kindreds with the recessively inherited triple-A syndrome (OMIM 231550), which consists of adrenal insufficiency, achalasia, and alacrima, identified linkage chromosome 12q13 (73). In this investigation, other previously proposed candidate genes were ruled out, including the vasoactive intestinal peptide (VIP) and adrenocorticotropin (ACTH) receptors, because they are located on different chromosomes. Chromosome 12q13 contains several keratin genes (3 and 7), including the cytokeratin protein perpherin (PRPH), which may be a reasonable candidate because of the hyperkeratosis that is associated with triple-A syndrome.
Gastrointestinal Disorders of Polyposis, Polyps, and Neoplasia
Our understanding of the pathogenesis and genetics of various intestinal polyposis and cancer syndromes has progressed at a rapid pace in recent years. The recent identification of the gene defective in multiple endocrine neoplasia-type 1 (MEN1) (OMIM 131100) is such an example. Multiple endocrine neoplasia is an autosomal dominant disorder with the clinical phenotype of Zollinger-Ellison syndrome (hypergastrinemia) and various other endocrine tumors. The MEN1 locus was mapped to chromosome 11q13 nearly 10 years ago. Chandrasekharappa et al. developed a careful physical map of the proposed region and identified 33 candidate genes within the vicinity of chromosome 11q13 (74,75). The researchers used endocrine tumors that demonstrate evidence of loss of heterozygosity (LOH) to narrow the chromosomal region, and approximately 100 kb of DNA was sequenced. Loss of heterozygosity is characteristic of tumors that have lost the normal function of a tumor-suppressor gene(76). This section of genomic DNA contained approximately 8 genes that were previously isolated independently and were expressed in various tissue. Menin, a single gene from this region, is expressed in a wide assortment of tissue, including those of endocrine origin. Forty-eight mutations were identified in menin in patients with familial and sporadic forms of MEN1(74,77). Although the precise function of themenin protein remains a mystery, it appears to be a cytoplasmically located soluble protein that is not homologous with other previously identified proteins and most likely represents a tumor-suppressor gene. Further investigations of other kindreds will be useful in determining the potential role of molecular diagnostic testing, especially in presymptomatic members of MEN1 families.
Mutations in menin were also identified in patients with sporadic primary hyperparathyroidism (78). Thirty percent of patients with primary hyperparathyroidism have LOH in chromosome 11q13, the approximate location of MEN1. Somatic mutations in menin were recently identified in all patients analyzed with LOH of this region. This raises the distinct possibility that mutations in menin may be responsible for sporadic cases of Zollinger-Ellison syndrome. Although LOH has not been reported in tumors isolated from sporadic cases, analysis of the role that both the menin gene and its homologues have in the pathogenesis of Zollinger-Ellison syndrome and other endocrine tumors is likely to be forthcoming.
Our understanding of the genetics and biology of many of the polyposis syndromes has continued to advance rapidly. The recent identification of the gene defective in Muir-Torre syndrome (OMIM 158320) is such an example. Muir-Torre syndrome is a rare autosomal dominant disorder with the clinical feature of sebaceous skin tumors and gastrointestinal tumors. The clinical similarities between Muir-Torre syndrome and hereditary nonpolyposis colorectal cancer (HNPCC) patients have led some to speculate that they may be clinical variants of one another (79). Furthermore, malignant tumors in Muir-Torre syndrome exhibit the same microsatellite instabilities that are characteristic of tumors from HNPCC patients. The gene locus for Muir-Torre syndrome is on chromosome 2p22-21, which corresponds to the approximate location of the DNA-mismatch repair enzyme mutS (E. coli) homologue-2 (hMSH2) gene, which is abnormal in many patients with HNPCC (80). However, a mutation in the hMSH2 gene has never been definitively identified. Bapat et al. reported a mutation in another DNA-mismatch repair gene hMLH-1, suggesting the genetic heterogeneity of this syndrome and its similarity to HNPCC (79).
Another polyposis syndrome is the autosomal dominant disorder Cowden disease (OMIM 158350). The clinical features of Cowden disease are hamartomatous intestinal polyps, breast cancer, megaloencephaly, and other symptoms of the central nervous system. Nelen et al. performed linkage analysis of 12 families with Cowden disease and established strong linkage to 10q22-23 (lod score, 8.92) (81). Several tumors were previously shown to have LOH in this region of chromosome 10, suggesting that it represents the location of a tumor-suppressor gene. Phosphatase and tensin homologue (PTEN) is a tumor-suppressor gene that is located in this region of chromosome 10. Mutations (LOH) in PTEN have been described in various sporadic tumors, especially in glioblastomas(82). Liaw et al. analyzed five families with Cowden disease for germline mutations in the PTEN gene(83). The probands in two families have an identical missense mutation of amino acid 129 (G129E), which resides inPTEN's critical tyrosine phosphatase domain and probably alters the function of the protein. More severe nonsense mutations that resulted in truncated protein were identified in two kindreds in whom the affected members had macrocephaly. These data suggest that the more deleterious mutations in the PTEN gene are associated with a more severe clinical outcome(Lhermitte-Duclos disease). The identification of the Cowden disease gene should provide the tools needed to perform presymptomatic screening of people at risk for the disease.
Loss of heterozygosity on chromosome 10q22 was recently identified in juvenile polyps of patients with sporadic solitary juvenile polyps orjuvenile polyposis coli (OMIM 174900) (84). Genomic heterozygosity was assessed in this region after an earlier report that a patient with rare juvenile polyposis coli and abnormalities of the head and extremities had a germline interstitial deletion of chromosome 10q22.3-q24.1 (85). Genetic analysis of thePTEN gene will determine whether juvenile polyposis coli is allelic with Cowden disease or results from a mutation in a yet unidentified tumor-suppressor gene located on chromosome 10q22.
Another autosomal dominant disorder is familial infiltrative fibromatosis (OMIM 135290), a rare disease associated with multiple desmoid tumors and nonpolyposis colorectal cancer. Scott et al. investigated four apparently unrelated kindreds with infiltrative fibromatosis for truncated mutations in the adenomatous polyposis coli (APC) gene(86). Three families were determined to have the identical 4-bp frameshift mutation in codon 1962, suggesting that3′ APC mutations may predispose to desmoid tumors. Eccles et al. independently studied a three-generation kindred with infiltrative fibromatosis that had no history of intestinal polyps and identified a single nonsense mutation in this region at position 1924 (87). These data are consistent with those in other studies in which results have demonstrated an association between the clinical phenotype and truncation mutations in certain locations of APC.
Peutz-Jeghers syndrome (OMIM 175200) is a distinctly autosomal dominant disorder with characteristic buccal mucosal pigmentation, intestinal polyps, and carcinoma of the pancreas and breast(88,89). Hemminki et al. hypothesized that the hamartomatous tumors from patients with Peutz-Jeghers would reveal evidence of LOH (76). Tumors from Peutz-Jeghers syndrome patients were cytogenetically analyzed for evidence of LOH, and a consistent region near the teleomere of chromosome 19 was identified. Comparative genomic hybridization was then used to define further this region of chromosome 19p13.3. Subsequently, standard linkage analysis revealed a very high lod score of 7, strongly suggesting the approximate location of the disease-causing gene. Potential candidate genes in this location include the oncogene VAV, and several transcriptional factors, including zinc finger protein 14 (KOX 6), and the transducin-like enhancer of split 1 and 2.
Similar linkage analysis has been recently reported in a single large kindred with hereditary mixed polyposis syndrome (OMIM 601228), an autosomal dominant disorder characterized by juvenile polyposis and colorectal carcinoma. The linkage analysis identified a particular region in chromosome 6q. This location is not known to contain a tumor-suppressor gene or other potential candidate genes (90). Moreover, LOH in chromosome 6q has not been identified in colorectal tumors; and thus, the defective gene will probably be a novel protein. Because of the apparent absence of a likely candidate gene, further linkage analysis that includes more informative families would allow investigators to narrow the candidate locus to a smaller, more manageable size.
Hereditary nonpolyposis colorectal cancer represents the most common inherited disorder associated with colorectal cancer and is inherited in an autosomal dominant pattern. Genomic DNA, isolated from tumors of patients with HNPCC, has features of genomic instability or microsatellite mutator phenotype(MMP) (1). The MMP from such tumors resembles abnormalities seen in bacteria that contained mutations in particular DNA-mismatch repair enzymes (1). Combining both extensive linkage data of HPNCC kindreds with the genetics and chemistry of invertebrate DNA-mismatch repair enzymes will hasten the identification of the human isoform of this family of DNA repair enzymes. Germline mutations in such DNA repair enzymes are seen in the majority of patients with HPNCC(1).
Recently, Liu et al. performed a comprehensive analysis of 74 families with HNPCC for mutations in DNA-mismatch repair genes (80). The results showed that 92% of HNPCC kindreds displayed evidence of microsatellite instability in colonic tumor cells, and 70% exhibited germline mutations. The germline mutations involved alterations in four DNA-mismatch repair enzymes hMSH2, hMLH1 (MutL [E. coli] homologue 1),hPMS1, and hPMS2 (postmeiotic segregation increased-2), whereas no mutations were identified in GTBP(G/T binding protein). Wijnen et al. studied 34 kindreds with HNPCC for mutations in the hMLH1 gene and identified 10 germline mutations, of which 50% were located in exons 15 and 16(91). Such mutation hot spots could be useful in simplifying attempts to perform genetic screening. Germline mutations in either MLH1 or MSH2 were identified in 86% of Finnish kindreds with HNPCC (92). Nearly 90% of these mutations occurred in the MLH1 locus. Other investigators have identified mutations in the hMHS2 and hMLH1 gene in patients with HNPCC, as well as in several people who do not meet the classic definition of this disorder (93-94). Vasen et al. examined the variability in clinical symptoms of the extended members of 17 kindreds with identified mutations in either hMSH2 or hMLH1(96). The relative risk of developing colon cancer in these people was 80% and did not vary regardless of which DNA repair enzyme was abnormal. When compared with those in non-HNPCC subjects, mutations in either gene significantly increased the risk of small intestinal carcinoma.
Taken together, these data may be interpreted to suggest that germline mutations in the various DNA-mismatch repair enzymes result in genomic instability in the somatic cell. However, the genes that are disrupted by such genomic instabilities, and the role that they may have in tumor formation is not well understood. Toward this end, microsatellite instability has been reported to result frequently in mutations in the insulin-like growth factor II receptor (IGFRII) gene (97). A second gene implicated in colon tumor pathogenesis is transforming growth factor-β(TGF-β). Eppert et al. recently delineatedTGF-β's downstream pathway that includes the phosphorylation of an MAD-related protein (MADR2)(98). MADR2 is located on chromosome 18q21, a region of the genome implicated in colon cancer because of reports of LOH in tumors from patients with HNPCC. Furthermore, several somatic mutations that inactivate MADR2 were identified and provided convincing evidence that TGF-β and other proteins in its signal transduction pathway have a tumor-suppressor role.
To identify in somatic cells the downstream consequences of mutations in DNA-mismatch repair enzymes, Rampino et al. examined tumors that exhibited evidence of genomic instability (MMP) for somatic mutations in the protein inBAX(99). The BAX protein forms either a homodimer or a heterodimer with the BCL2 protein, which represses cell death during the process of apoptosis. Approximately 50% of MMP(+) cells had evidence of somatic mutations within a specified hot spot of the BAX gene. These mutations resulted in the inactivation of theBAX protein and therefore dysregulation in the normal process of cell death. In contrast, no MMP(-) cells had a mutation in the BAX gene; and thus, there was no alteration in the normal apoptotic process of the colonic epithelia.
Germline mutation in the tumor-suppressor protein APC gene has a well-documented role in familial adenomatous polyposis (FAP). Because of the genomic instabilities that characterize the colonocytes of patients with HNPCC, somatic mutations in the APC gene may be possible. Huang et al. provided direct evidence of somatic mutations in the APC genes of patients with HNPCC and sporadic colorectal cancer(100). It is interesting that the incidence ofAPC mutations was unexpectedly just as common in MMP(+) as in MMP(-) cells. However, APC mutations in MMP(+) cells were more frequently severe nonsense mutations resulting in usually less deleterious truncated proteins, whereas MMP(-) cells had a higher incidence of the missense mutations.
In summary, these data are consistent with the premise that dominantly inherited mutations in a member of the DNA-mismatch repair enzyme family of proteins result in poor genomic DNA repair and in instability. These subsequent somatic mutations in such genes as IGFRII, MADR2, andBAX, to name a few, will interrupt the normal function of the protein, which could have a deleterious effect on cell growth and programmed cell death. This may result in poorly controlled cell growth and in tumor formation. Identifying the other targets of somatic mutations will be particularly important to understanding the common noninherited forms of colorectal polyps and carcinoma (1).
The autosomal dominant disorder familial adenomatous polyposis(FAP), (OMIM 175100) results in severe intestinal polyposis and colorectal cancer before the fourth decade of life. Until recently, the precise action ofAPC and its role in the complex process of cell growth and tumor formation was poorly understood. The large APC protein interacts through various distinct binding sites with several other proteins, includingβ1-catenin and GSK-3β(1,101-103).β1-Catenin has been the particular focus of current investigations because of its central role in tumor formation(1,101).
β1-Catenin forms a heterodimer with the transcriptional factor T-cell factor (hTcf-4) and enhances the transcriptional activation of various unidentified target genes(101). In the presence of a functional APC protein, free β1-catenin concentration is low, resulting in a decline in β1-catenin/hTcf-4 heterodimers and a subsequent decline in the gene expression of various downstream targets. Morin et al. have shown that truncated mutations of the APC gene (e.g., those described in FAP patients) that eliminate as few as a singleβ1-catenin-binding segment elevate the concentration ofβ1-catenin and increaseβ1-catenin/hTcf-4 transactivation(103). In results reported in other studies, Morin et al. determined that tumor cell lines containing two wild-type copies of theAPC gene had a unique mutation in the regulatory phosphorylation site (serine 45) of β1-catenin. Normally, phosphorylation of wild-type β1-catenin reduces its function (104). Therefore, either an APC mutation that disrupts a β1-catenin binding site or a mutation in the β1-catenin gene that causes its constitutive high levels of activation results inβ1-catenin/hTcf-4-induced transactivation of downstream genes and in subsequent polyposis and colorectal cancer(1).
Taken together, these data were interpreted to suggest that either germline(FAP) or somatic (HNPCC) mutations in the APC gene result in unabated β1-catenin/hTcf-4-induced upregulation of various genes that are presumably important for controlling normal cell growth and death. This was confirmed by Morin et al., who used the colorectal cell line HT-29 that expresses two native nonfunctional APC transcripts to test the consequence of expressing an exogenous functional APC protein (105). The introduction of a functionalAPC protein results in a dramatic decline in cell growth caused by increased apoptosis of these cells.
To define the potential target of β1-catenin/hTcf-4 transactivation, Zhang et al. identified 83 transcripts that are either increased or repressed 10-fold in colon tumors compared with those found in the normal colon (106). Included among the transcripts whose levels were altered were insulin-like growth factor II, fibronectin, and various ribosomal proteins, to name a few. Assessingβ1-catenin/hTcf-4's direct or indirect role in altering the transcription of these specific genes and determining the role that some of these proteins play in the downstream process of tumor formation will be the future direction of many projects in colorectal cancer. The identification and elucidation of these various proteins and their roles in tumor cells will eventually provide new targets to combat tumor formation in patients with FAP, HNPCC, and the common noninherited forms of colorectal cancer.
It was recently reported that germline mutations in the APC gene occur in 6% of Ashkenazi Jews (107). The mutation was identified using an in vitro synthesized protein assay that detected a truncated APC protein from a single patient. It is notable that although the expected nonsense mutation was not identified in the germline, a missense mutation (I1307K) was detected on one allele. This was particularly surprising, because missense mutations of the APC gene are generally nonpathogenic; and more importantly, the I1307K mutation would not be expected to result in a truncated protein. The peculiar mutation changes the wild-type sequence from GAAATAAAA to GAAAAAAAA, at codons 1306 to 1308. Tumors from patients with the germline I1307K mutation had a variety of somatic nonsense mutations located in the vicinity of the 1307 codon. These data were interpreted to suggest that the I1307K mutation results in (A)8 tract that is incorrectly read by the polymerases and invariably results in somatic mutations with deleterious effects. The I1307K mutation segregates in families with probands with colorectal cancers and is particularly common in Ashkenazi Jews with a history of colorectal cancer, especially before the age of 66. The identification of this hypermutable region will have significant implications that reach beyond the field of colorectal cancer and must be considered in other genetic diseases. More importantly, this mutation represents the most common cancer-causing defect identified in any particular population, and genetic testing should be readily available to test people at greatest risk.
Basal cell nevus syndrome (BCNS) (OMIM 109400) is a complex disorder associated with basal cell carcinoma, CNS tumors, gastric hamartomatous polyps, and various skeletal malformations. Johnson identified that BCNS is due to a defect in the human homolog of a gene namedpatched (PTC) (108).Patch is a segment polarity gene originally identified and characterized in Drosophila and is the receptor for the sonic hedgehog protein (109). The human homolog ofpatched was isolated and mapped to chromosome 9q22.3, a region previously determined by linkage studies to be the location of theBCNS gene. Johnson tested the potential role of PTC in the pathogenesis of BCNS by cloning the gene and screening for mutations. Several nonsense mutations were identified. PTC is a 1447 amino acid membrane protein known to repress transforming growth factor-β (TGF-β) expression, and thus counters the action of it's ligand, sonic hedgehog.
Significant strides have also been made in identifying the new treatments for people with FAP. Results of several studies have demonstrated a role for nonsteriodal antiinflammatory drugs (NSAID) in decreasing the incidence of polyps in humans and mice with FAP (110). The activity of the cyclooxygenase-1 (COX-1) and -2 enzymes decreases with NSAID use and results in a decline in the prostaglandin G2 level. What remains unclear is the precise mechanism by which NSAIDs alter tumorigenesis and colorectal cancer. Oshima et al. crossed APC knockout mice(APC1716) with mice that had undergone COX-2 disruption and showed that the extent of polyp formation was significantly lower in COX-2 knockout mice (111). More importantly, they demonstrated that a COX-2-selective inhibitor (MF tricyclic) was significantly more effective at decreasing polyposis than were general inhibitors of COX-1 and -2. Similarly, Boolbol et al. showed that in the Min mouse model of polyposis, the COX-2 inhibitor sulindac dramatically decreased the incidence of intestinal tumors 10-fold (112). Cyclooxygenase-2 levels are elevated in the Min mouse, consistent with the idea that COX-2 activity is involved in polyp formation(113). These data infer a role for COX-2-specific inhibitors in controlling polyp formation; however, clinical trials comparing the efficacy of COX-1 and -2 in humans with FAP have not been reported.
Acknowledgment: The author thanks Dr. Richard Aranda for his suggestions and careful review of the manuscript and Dr. Ernest M. Wright for his continuing support and mentoring.
1. Kinzler K, Vogelstein B. Lessons from hereditary colorectal cancer. Cell
2. Scriver CR, Beaudet AL, Sly WS, Valle D. The metabolic and molecular bases of inherited disease
. Vol 1-3. New York: McGraw-Hill, Health Professions Division, 1995.
3. Yamada T, Alpers DH, Owyang C, Powell DW, Silverstein FE.Textbook of gastroenterology
. Philadelphia: Lippincott, 1995.
4. Mowat AP. Liver disorders in childhood
. 3rd ed. Oxford and Boston: Butterworth-Heinemann, 1994.
5. Walker WA. Pediatric gastrointestinal disease: Pathophysiology, diagnosis, management
. 2nd ed. St. Louis: Mosby, 1996.
6. Suchy FJ. Liver disease in children
. 1st ed. St. Louis: Mosby, 1994.
8. Online Mendelian Inheritance in Man, OMIM
. Baltimore, MD: Center for Medical Genetics, Johns Hopkins University. Bethesda, MD: National Center for Biotechnology Information, National Library of Medicine. 1996. World Wide Web URL:http://ww.ncbi.nlm.nih.gov/Omim/
9. Smith BN, Kantrowitz W, Grace ND, et al. Prevalence of hereditary hemochromatosis in a Massachusetts corporation: Is Celtic origin a risk factor. Hepatology
10. Bulaj Z, Griffen LM, Jorde LB, Edwards CQ, Kushner JP. Clinical and biochemical abnormalities in people heterozygous for hemochromatosis. N Engl J Med
11. Feder JN, Gnirke A, Thomas W, et al. A novel MHC class I-like gene is mutated in patients with hereditary haemochromatosis.Nat Genet
12. Beutler E, Gelbart T, West C, et al. Mutation analysis in hereditary hemochromatosis. Blood Cell Mol Dis
13. Parkkila S, Waheed A, Britton RS, et al. Immunohistochemistry of HLA-H, the protein defective in patients with hereditary hemochromatosis, reveals unique pattern of expression in gastrointestinal tract. Proc Natl Acad Sci USA
14. Gunshin H, Mackenzie B, Berger UV, et al. Cloning and characterization of a mammalian proton-coupled metal-ion transporter.Nature
15. Fleming MD, Trenor CC, Su MA, et al. Microcytic anaemia mice have a mutation in Nramp2, a candidate iron transporter gene. Nat Genet
16. Verloes A, Temple IK, Hubert AF, et al. Recurrence of neonatal haemochromatosis in half sibs born of unaffected mothers. J Med Genet
17. Tumer Z, Horn N, Tonnesen T, Christodoulou J, Clarke J, Sarkar B. Early Copper-histidine treatment for Menkes ´Disease.Nat Genet
18. Kelly EJ, Palmiter RD. A murine model of Menkes disease reveals a physiological function of metallothionein. Nat Genet
19. Wong MH, Oelkers P, Craddock AL, Dawson PA. Expression cloning and characterization of the hamster ileal sodium-dependent bile acid transporter. J Biol Chem
20. Oelkers P, Kirby LC, Heubi JE, Dawson PA. Primary bile acid malabsorption caused by mutations in the ileal sodium-dependent bile acid transporter gene (SLC10A2). J Clin Invest
21. Wong MH, Oelkers P, Dawson PA. Identification of a mutation in the ileal sodium-dependent bile acid transporter gene that abolishes transport activity. J Biol Chem
22. Martin MG, Turk E, Lostao MP, Kerner C, Wright EM. Defects in Na+ glucose cotransporter (SGLT1) trafficking and function cause glucose-galactose malabsorption. Nat Genet
23. Martin MG, Lostao MP, Turk E, Lam E, Kreman M, Wright EM. Compound missense mutations in the sodium/D-glucose cotransporter result in trafficking defects. Gastroenterology
24. Martin MG, Turk E, Kerner C, Zabel B, Wirth S, Wright EM. Prenatal identification of a heterozygous status in two fetuses at risk for glucose galactose malabsorption. Prenat Diagn
25. Loo DDF, Zeuthen T, Chandy G, Wright EM. Cotransport of water by the Na+
/glucose cotransporter. Proc Natl Acad Sci USA
26. Nobigrot T, Chasalow FI, Lifshitz F. Carbohydrate absorption from one serving of fruit juice in young children: Age and carbohydrate composition effects. J Am Coll Nutr
27. Tajiri H, Hamamoto T, Harada T, Okada S. Fructose malabsorption in a patient with chronic diarrhea. J Pediatr Gastroenterol Nutr
28. Rand EB, Depaoli AM, Davidson NO, Bell GI, Burant CF. Sequence, tissue distribution, and functional characterization of the rat fructose transporter GLUT5. Am J Physiol
29. Wasserman D, Hoekstra JH, Tolia V, et al. Molecular analysis of the fructose transporter gene (GLUT5
) in isolated fructose malabsorption. J Clin Invest
30. Höglund P, Haila S, Scherer SW, et al. Positional candidate genes for congenital chloride diarrhea suggested by high-resolution physical mapping in chromosome region 7q31. Genome Res
31. Silberg DG, Wang W, Moseley RH, Traber PG. TheDown Regulated in Adenoma
) gene encodes an intestine-specific membrane sulfate transport protein. J Biol Chem
32. Höglund P, Haila S, Socha J, et al. Mutations of the Down-regulated in adenoma (DRA
) gene cause congenital chloride diarrhoea. Nat Genet
33. Ouwendijk J, Moolenaar CEC, Peters WJ, et al. Congenital sucrase-isomaltose deficiency-Identification of a glutamine to proline substitution that leads to a transport block of sucrase-isomaltose in a pre-Golgi compartment. J Clin Invest
34. Ward CL, Omura S, Kopito RR. Degradation of CFTR by the ubiquitin-proteasome pathway. Cell
35. Kopito RK. ER Quality Control: The cytoplasmic connection. Cell
36. Dieterich W, Ehnis T, Bauer M, Donner P, Volta U, Riecken EO, Schuppan D. Identification of tissue transglutaminase as the autoantigen of celiac disease Nat Med
37. Zhong F, McCombs CC, Olson JM, et al. An autosomal screen for genes that predispose to celiac disease in the western countries of Ireland. Nat Genet
38. Hugot JP, Laurent-Puig P, Gower-Rousseau C, et al. Mapping of a susceptibility locus for Crohn's disease on chromosome 16.Nature
39. Ohmen JD, Yang HY, Yamamoto KK, et al. Susceptibility locus for inflammatory bowel disease on chromosome 16 has a role in Crohn's disease, but not in ulcerative colitis. Hum Mol Genet
40. Satsangi J, Parkes M, Louis E, et al. Two stage genome-wide search in inflammatory bowel disease provides evidence for susceptibility loci on chromosomes 3, 7 and 12. Nat Genet
41. Rudolph U, Finegold MJ, Rich SS, et al. Ulcerative colitis and adenocarcinoma of the colon in Gai2
-deficient mice.Nat Genet
42. Sperber K, Ogata S, Sylvester C, et al. A novel human macrophage-derived intestinal mucin secretagogue: Implications for the pathogenesis of inflammatory bowel disease. Gastroenterology
43. Plevy SE, Targan SR, Yang HY, Fernandez D, Rotter JI, Toyoda H. Tumor necrosis factor microsatellites define a Crohn's disease-associated haplotype on chromosome 6. Gastroenterology
44. Fukai K, Oh J, Frenk E, Almodovar C, Spritz RA. Linkage disequilibrium mapping of the gene for Hermansky-Pudlak syndrome to chromosome 10q23.1-q23.3. Hum Mol Genet
45. Oh J, Bailin T, Fukai K, et al. Positional cloning of a gene for Hermansky-Pudlak syndrome, a disorder of cytoplasmic organelles.Nat Genet
46. Feng GH, Bailin T, Oh J, Spritz RA. Mouse pale ear
) is homologous to human Hermansky-Pudlak syndrome and contains a rare `AT-AC' intron. Hum Mol Genet
47. Abreu-Martin MT, Targan SR. Regulation of immune responses of the intestinal mucosa. Crit Rev Immunol
48. Aranda R, Sydora BC, McAllister PL, et al. Analysis of intestinal lymphocytes in mouse colitis mediated by transfer of CD4+, CD45RBhigh
, T cells to SCID recipients. J Immunol
49. Chakravarti A. Endothelin receptor-mediated signaling in Hirschsprung disease. Hum Mol Genet
50. Puffenberger EG, Hosoda K, Washington SS, et al. A missense mutation of the endothelin-B receptor gene in multigenic Hirschsprung's disease. Cell
51. Hosoda K, Hammer RE, Richardson JA, et al. Targeted and natural (piebald-lethal) mutations of endothelin-B receptor gene produce megacolon associated with spotted coat color in mice. Cell
52. Kusafuka T, Wang YP, Puri P. Novel mutations of the endothelin-B receptor gene in isolated patients with Hirschsprung's disease.Hum Mol Genet
53. Amiel J, Atti T, Jan D, et al. Heterozygous endothelin receptor B (EDNRB
) mutations in isolated Hirschsprung disease.Hum Mol Genet
54. Auricchio A, Casari G, Staiano A, Ballabio A. Endothelin-B receptor mutations in patients with isolated Hirschsprung disease from a non-inbred population. Hum Mol Genet
55. Baynash AG, Hosoda K, Giaid A, et al. Interaction of endothelin-3 with endothelin-B receptor is essential for development of epidermal melanocytes and enteric neurons. Cell
56. Hofstra RMW, Osinga J, Tan-Sindhunata G, et al. A homozygous mutation in the endothelin-3 gene associated with a combined Waardenburg type 2 and Hirschsprung phenotype (Shah-Waardenburg syndrome). 1996;12:445-7.
57. Bolk S, Angrist M, Schwartz S, et al. Congenital central hypoventilation syndrome: Mutation analysis of the receptor tyrosine kinase ret. Am J Med Genet
58. Bolk S, Angrist M, Xie J, et al. Endothelin-3 frameshift mutation in congenital central hypoventilation syndrome. 1996;13:395-6.
59. Moore MW, Klein RD, Farias I, et al. Renal and neuronal abnormalities in mice lacking GDNF. Nature
60. Pichel JG, Shen LY, Sheng HZ, et al. Defects in enteric innervation and kidney development in mice lacking GDNF. Nature
61. Sanchez MP, Silos-Santiago I, Frisn J, He B, Lira SA, Barbacid M. Renal agenesis and the absence of enteric neurons in mice lacking GDNF. Nature
62. Pasini B, Ceccherini I, Romeo G. RET mutations in human disease. Trends Genet
63. Treanor JJS, Goodman L, De Sauvage F, et al. Characterization of a multicomponent receptor for GDNF. Nature
64. Salomon R, Atti T, Pelet A, et al. Germ-line mutations of the RET ligand GDNF are not sufficient to cause Hirschsprung disease.Nat Genet
65. Sanicola M, Hession C, Worley D, et al. Glial cell line-derived neurotrophic factor-dependent RET activation can be mediated by two different cell-surface accessory proteins. Proc Natl Acad Sci USA
66. Iwashita T, Murakami H, Asai N, Takashasi M. Mechanism of Ret
dysfunction by Hirschsprung mutations affecting its extracellular domain. Hum Mol Genet
67. Auricchio A, Brancolini V, Casari G, et al. The locus for a novel syndromic form of neuronal intestinal pseudoobstruction maps to Xq28. Am J Hum Genet
68. Okamoto N, Wada Y, Goto M. Hydrocephalus and Hirschsprung's disease in a patient with a mutation of L1CAM. J Med Genet
69. Barone V, Weber D, Luo Y, Brancolini V, Devoto M, Romero MF. Exclusion of linkage between Ret
and neuronal intestinal dysplasia type B. Am J Med Genet
70. Shirasawa S, Yunker AMR, Roth KA, Brown GA, Horning S, Korsmeyer SJ. Enx(Hox11Ll)-
deficient mice develop myenteric neuronal hyperplasia and megacolon. Nat Med
71. Huang PL, Dawson TM, Bredt DS, Snyder SH, Fishman MC. Targeted disruption of the neuromal nitric oxide synthase gene. Cell
72. Chung E, Curtis D, Chen G, et al. Genetic evidence for the neuronal nitric oxide synthase gene (NOS1
) as a susceptibility locus for infantile pyloric stenosis. Am J Hum Genet
73. Weber A, Wienker TF, Jung M, et al. Linkage of the gene for the triple A syndrome to chromosome 12q13 near the type II keratin gene cluster. Hum Mol Genet
74. Chandrasekharappa SC, Guru SC, Manickam P, et al. Positional cloning of the gene for Multiple Endocrine Neoplasia-Type 1.Science
75. Emmert-Buck MR, Lubensky IA, Dong Q, et al. Localization of the multiple endocrine neoplasia type I (MEN1) gene based on tumor loss of heterozygosity analysis. Cancer Res
76. Hemminki A, Tomlinson I, Markie D, et al. Localization of a susceptibility locus for Peutz-Jeghers syndrome to 19p using comparative genomic hybridization and targeted linkage analysis. Nat Genet
77. Agarwal SK, Kester MB, Debelenko LV, et al. Germline mutations of the MEN1 gene in familial multiple endocrine neoplasia type 1 and related states. Hum Mol Genet
78. Heppner C, Kester MB, Agarwal SK, et al. Somatic mutation of the MEN1 gene in parathyroid tumours. Nat Genet
79. Bapat B, Xia L, Madlensky L et al. The genetic basis of Muir-Torre Syndrome includes the hMLH1 locus. Am J Hum Genet,
80. Liu B, Parsons R, Papadopoulos N, et al. Analysis of mismatch repair genes in hereditary non-polyposis colorectal cancer patients.Nat Med
81. Nelen MR, Padberg GW, Peeters EAJ, et al. Localization of the gene for Cowden disease to chromosome 10q22-23. Nat Genet
82. Li J, Yen C, Liaw D, et al. PTEN, a putative protein tyrosine phosphatase gene mutated in human brain, breast, and prostate cancer.Science
83. Liaw D, Marsh DJ, Li J, et al. Germ-line mutations of the PTEN gene in Cowden disease, an inherited breast and thyroid cancer syndrome. Nat Genet
84. Jacoby RF, Schlack S, Cole CE, Sharbek M, Harris C, Meisner LF. A juvenile polyposis tumor suppressor locus at 10q22 is deleted from nonepithelial cells in the lamina propria. Gastroenterology
85. Jacoby RF, Schlack S, Sekhon G, Laxova R. Del(10)(q22.3q24.1) associated with juvenile polyposis. Am J Med Genet
86. Scott RJ, Froggatt NJ, Trembath RC, Evans DGR, Hodgson SV, Maher ER. Familial infiltrative fibromatosis (Desmoid tumors) (MIM135290) caused by a recurrent 3′ APC gene mutation. Hum Mol Genet
87. Eccles DM, Luijt R, Breukel C, et al. Hereditary Desmoid Disease due to a frameshift mutation at codon 1924 of the APC gene.Am J Hum Genet
88. Tomlinson IPM, Olschwang S, Abelovitch D, et al. Testing candidate loci on chromosomes 1 and 6 for genetic linkage to Peutz-Jeghers' disease. Ann Hum Genet
89. Markie D, Huson S, Maher E, Davies A, Tomlinson I, Bodmer WF. A pericentric inversion of chromosome six in a patients with Peutz-Jeghers' syndrome and the use of FISH to localise the breakpoints on a genetic map. Hum Genet
90. Thomas HJW, Whitelaw SC, Cottrell SE, et al. Genetic mapping of the hereditary mixed polyposis syndrome to chromosome 6q. Am J Hum Genet
91. Wijnen J, Khan PM, Vasen H, et al. Majority of hMLH1 mutations responsible for hereditary nonpolyposis colorectal cancer cluster at the exonic region 15-16. Am J Hum Genet
92. Nyström-Lahti M, Wu Y, Moisio AL, et al. DNA mismatch repair gene mutations in 55 kindreds with verified or putative hereditary non-polyposis colorectal cancer. Hum Mol Genet
93. Jeon HM, Lynch PM, Howard L, Ajani J, Levin B, Frazier ML. Mutation of the hMSH2 gene in two families with hereditary nonpolyposis colorectal cancer. Hum Mutat
94. Maliaka YK, Chudina AP, Belev NF, Alday P, Bochkov N, Buerstedde JM. CpG dinucleotides in the hMSH2 and hMLH1 genes are hotspots for HNPCC mutations. Hum Genet
95. Pensotti V, Radice P, Presciuttini S, et al. Mean age of tumor onset in hereditary nonpolyposis colorectal cancer (HNPCC) families correlates with the presence of mutations in DNA mismatch repair genes.Genes Chrom Cancer
96. Vasen HFA, Wijnen JT, Menko FH, et al. Cancer risk in families with hereditary nonpolyposis colorectal cancer diagnosed by mutation analysis. Gastroenterol
97. Souza RF, Appel R, Yin J, et al. Microsatellite instability in the insulin-like growth factor II receptor gene in gastrointestinal tumours. Nat Genet
98. Eppert K, Scherer SW, Ozcelik H, et al. MADR2 maps to 18q21 and encodes a TGFB-regulated MAD-related protein that is functionally mutated in colorectal carcinoma. Cell
99. Rampino N, Yamamoto H, Ionov Y, et al. Somatic frameshift mutations in the BAX gene in colon cancers of the microsatellite mutator phenotype. Science
100. Huang J, Papadopoulos N, McKinley AJ, et al. APC mutations in colorectal tumors with mismatch repair deficiency. Proc Natl Acad Sci USA
101. Korinek V, Barker N, Morin PJ, et al. Constitutive transcriptional activation by a B-Catenin-Tcf complex in APC-/-colon carcinoma. Science
102. Rubinfeld B, Albert I, Porfiri E, Fiol C, Munemitsu S, Polakis P. Binding of GSK3 B to the APC/B-Catenin complex and regulation of complex assembly. Science
103. Matsumine A, Ogai A, Senda T, et al. Binding of APC to the human homolog of the drosophila discs large tumor suppressor protein.Science
104. Morin PJ, Sparks AB, Korinek V, et al. Activation of B-Catenin-Tcf signaling in colon cancer by mutations in B-Catenin or APC.Science
105. Morin PJ, Vogelstein B, Kinzler KW. Apoptosis andAPC
in colorectal tumorigenesis. Proc Natl Acad Sci USA
106. Zhang L, Zhou W, Velculescu VE, et al. Gene expression profiles in normal and cancer cells. Science
107. Laken SJ, Petersen GM, Gruber SB, et al. Familial colorectal cancer in Ashkenazim due to a hypermutable tract in APC. 1997;17:79-83.
108. Johnson RL, Rothman AL, Xie JW, et al. Human homolog of patch: a candidate for the basal cell nevus syndrome. Science
109. Hooper JE, Scott MP. The drosophilia patch gene encodes a putative membrane protein required for segmental patterning.Cell
110. Tsujii M, Kawano S, DuBois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential.Proc Natl Acad Sci USA
111. Oshima M, Dinchuk JE, Kargman SL, et al. Suppression of intestinal polyposis in APCΔ716
knockout mice by inhibition of cyclooxygenase 2 (COX-2). Cell
112. Boolbol SK, Dannenberg AJ, Chadburn A, et al. Cycooxygenase-2 overexpression and tumor formation are blocked by sulindac in a murine model of familial adenomatous polyposis. Cancer Res
113. Williams CS, Luongo C, Radhika A, et al. Elevated cyclooxygenase-2 levels in Min