Adenomatous Polyposis Families That Screen APC Mutation每Negative by Conventional Methods Are Genetically Heterogeneous
http://www.100md.com
▲還散笫雖悝◎
the Department of Medical Genetics, Institute of Dentistry, and Institute of Biotechnology, University of Helsinki
Laboratory of Molecular Genetics, Department of Oral and Maxillofacial Diseases, and Department of Surgery, Helsinki University Central Hospital, Helsinki, Finland
ABSTRACT
PURPOSE: One third of families with classical adenomatous polyposis (FAP), and a majority of those with attenuated FAP (AFAP), remain APC mutation每negative by conventional methods. Our purpose was to clarify the genetic basis of polyposis and genotype-phenotype correlations in such families.
PATIENTS AND METHODS: We studied a cohort of 29 adenomatous polyposis families that had screened APC mutation每negative by the protein truncation test, heteroduplex analysis, and exon-specific sequencing. The APC gene was investigated for large genomic rearrangements by multiplex ligation-dependent probe amplification (MLPA), and for allelic mRNA expression by single nucleotide primer extension (SNuPE). The AXIN2 gene was screened for mutations by sequencing.
RESULTS: Four families (14%) showed a constitutional deletion of the entire APC gene (three families) or a single exon (one family). Seven families (24%) revealed reduced or extinct mRNA expression from one APC allele in blood, accompanied by loss of heterozygosity in the APC region in six (75%) of eight tumors. In 15 families (52%), possible APC involvement could be neither confirmed nor excluded. Finally, as detailed elsewhere, three families (10%) had germline mutations in genes other than APC, AXIN2 in one family, and MYH in two families.
CONCLUSION: "APC mutation每negative" FAP is genetically heterogeneous, and a combination of MLPA and SNuPE is able to link a considerable proportion (38%) to APC. Significant differences were observed in clinical manifestations between subgroups, emphasizing the importance of accurate genetic and clinical characterization for the proper management of such families.
INTRODUCTION
Familial adenomatous polyposis (FAP) is an autosomal dominant syndrome characterized by the development by age 20 years of multiple colorectal polyps; one or several of these polyps progresses to cancer by approximately 40 years of age.1 In classical FAP, at least 100 (often hundreds to more than a thousand) adenomas are present, whereas polyposis with fewer than 100 adenomas per patient is referred to as attenuated polyposis (AFAP or AAPC). In addition to colorectal polyps, affected individuals may develop various extracolonic manifestations, such as desmoid tumors, duodenal adenomas, mandibular osteomas, and hypertrophic pigmentary lesions of the retina. In AFAP, the age at onset of adenomas and cancer is higher, and the lifetime risk of cancer is lower compared with classical FAP.2 While the true incidence and frequency of AFAP is unknown,3 it may account for up to 10% of adenomatous polyposis families.4
Germline mutations of the adenomatous polyposis coli (APC) tumor suppressor gene located on 5q21-22 cause both FAP and AFAP.5-7 To date, more than 800 pathogenic mutations in APC have been described.8 These are typically frameshift or nonsense changes that lead to truncated protein products,9 forming the basis for their detection by the protein truncation test (PTT).10 In classical FAP, mutations may be scattered all over the gene, with some hot spots in the somatic mutation cluster region between codons 1250 and 1450 in exon 15.11 The position of APC mutation correlates to some extent with clinical features12-14; for example, mutations between codons 1444 to 1578 are associated with classical adenomatous polyposis and multiple extracolonic manifestations.15,16 AFAP phenotype is associated with mutations in the 5' end (before codon 157 in exon 4), in the alternatively spliced part of exon 9, and in the distal 3' end of APC (beyond codon 1595).17
Conventional methods for APC mutation detection include the PTT test and various exon-specific screening methods such as single-strand conformation polymorphism assay, heteroduplex analysis, and DNA sequencing. These methods are sensitive to point mutations, but they fail to detect large genomic rearrangements as well as possible transcription defects. Conventional techniques leave approximately 30% of families with classical FAP and approximately 90% of AFAP families APC mutation每negative.12,18-22 This apparent mutation negativity may suggest either that APC has alterations that escape detection by routine techniques, or alternatively, that other (known or unknown) genes are involved in FAP predisposition. Among known genes, recent evidence implicates at least the Mut Y homolog MYH encoding a DNA glycosylase.23 Recessive inheritance and variable polyp count (lower or greater than 100) are features of MYH-associated polyposis.24,25 MYH mutations may explain up to 25% of "APC mutation每negative" FAP, leaving the majority of FAP families without identifiable APC mutations molecularly unexplained.
We recently conducted a study on 65 classical FAP families from Finland, which resulted in the identification of APC germline mutations in 72%.22 The present report addresses the genetic basis of polyposis in the third (approximate) of FAP families that shows no APC mutation by conventional methods.
PATIENTS AND METHODS
Subjects
We investigated 29 families (38 individuals) representing the mutation-negative fraction (30%) among a total of 98 adenomatous polyposis families from the Hereditary Colorectal Cancer Registry of Finland that had been screened for APC germline mutations by common diagnostic methods (PTT for the entire coding region and heteroduplex analysis of exons 1 to 14, or sequencing of exons 1 to 15).22 Table 1 presents clinical features of the families. Eighteen FAP families with truncating mutations in APC (18 different mutations, 26 individuals) were studied for comparison. As controls, two healthy individuals (unaffected with any cancer), as well as 20 colon cancer patients from DNA mismatch repair gene mutation每negative families with hereditary nonpolyposis colorectal carcinoma (HNPCC), were tested. This study was approved by the institutional review board of the Helsinki University Central Hospital (Helsinki, Finland).
DNA and RNA Extraction
DNA was extracted from cultured lymphocytes or lymphoblasts.26 Total cellular RNA was extracted according to the guanidium thiocyanate method.27 Normal and tumor DNA from paraffin-embedded tissue sections was extracted according to Isola et al.28
MLPA for Large Genomic Rearrangements
Multiplex ligation-dependent probe amplification (MLPA)29 was used to screen DNA samples for large genomic rearrangements of APC according to the manufacturer's instructions (Salsa P043 kit; MRC-Holland, Amsterdam, the Netherlands). The MLPA method is based on sequence-specific probe hybridization to genomic DNA, followed by polymerase chain reaction (PCR) amplification of the hybridized probe (with one amplification primer fluorescently labeled), and semiquantitative analysis of the PCR products. Target-specific products are identified according to their differential length. The Salsa P043 kit contains probes for each coding exon of APC, three probes for the APC promoter region, and 11 probes for other human genes located on different chromosomes. FAM-labeled PCR products of MLPA reactions were analyzed on an ABI 3730 sequencer using Genemapper software, version 3.0 (Applied Biosystems, Foster City, CA). Deletion was suspected when the peak area was reduced more than 40% to 50% compared with normal controls. The diagnosis was based on at least two repeated MLPA experiments and was confirmed by the observation of the involvement of contiguous exons (whole-gene deletions) or the demonstration of a deletion-specific fragment in cDNA (exon 4 deletion).
SNuPE for Allelic mRNA Expression
The basic principles of the Single Nucleotide Primer Extension (SNuPE) method were described in our previous publication.30 For the present analyses, the coding polymorphism C/T at nucleotide 1458 of APC exon 11 was selected because of its high heterozygosity values and location in a short exon flanked by introns, making it possible to design gDNA- and cDNA-specific PCR reactions appropriate for this purpose. SNuPE analyses were performed on reverse transcription (RT) -PCR products of blood RNA encompassing the polymorphic site (amplified with forward primer from exon 11, 5'-GACTACAGGCCATTGCAGAA-3', and reverse primer from exon 12, 5'-GCCTTGTTGGCTACATCTCC-3'). The results were compared with amplification products obtained from genomic DNA of the same individuals (using the same forward primer as above, combined with reverse primer 5'-AAAGCCATTCCAGCATATCG-3' from exon 11). With these amplification products as templates, extension reaction was carried out from reverse direction using a fluorescently labeled primer (5'-GTGTAATACTGTAGTGGTCATTAGTAAG-3') and dATP, dCTP, and dTTP for chain elongation, and ddGTP for chain termination. Primer extension products, 34 bp for G (C) allele and 44 bp for A (T) allele, were run on an automated sequencer and analyzed using Genemapper. Ratios of allelic peak areas in cDNA relative to gDNA were calculated, and values below 0.6 or above 1.7 (indicating that the transcript of one allele had decreased 40% or more) were considered abnormal based on validation experiments performed in a large number of individuals.
Analysis of APC Promoter Region for Mutations
The APC promoter region (U02509; http://www.ncbi.nlm.nih.gov) was amplified with primers 5-GCTAGCATAGCTTTTCTGGTAAC-3' and 5'-CAGTGACACCCTGGCGGGCTG-3' and sequenced for alterations.
APC Haplotype Analysis
For haplotype analysis, the following microsatellite markers or single nucleotide polymorphisms were studied (genetic or physical distance between markers in parentheses): centromere, D5S409, (9.6 cM), D5S1965, (200 kb) [APC promoter a/g polymorphism, rs 201972031, APC intron 7 a/t polymorphism, rs 1914,31 APC exon 11 C/T polymorphism (nt. 1458), APC exon 15 A/G polymorphism (nt. 5037)],31 (< 100 kb), D5S346, (100 kb), MCC,32 telomere. Primer sequences for microsatellite markers can be found at http://www.gdb.org.
Loss of Heterozygosity in Tumor Tissue
APC exon 11 polymorphism was analyzed by SNuPE as described above, except that paired tumor and normal DNA samples were studied. Additionally, microsatellite markers closely linked to APC, D5S1965 (200 kb upstream of APC), and D5S346 (< 100 kb downstream of APC), were investigated. The marker-specific fluorescent amplification products were run on an automated sequencer and analyzed by Genemapper. The ratios of allelic peak areas in tumor DNA relative to normal DNA were interpreted to suggest loss of heterozygosity (LOH) if below 0.60 or above 1.67 (indicating that one allele had decreased 40% or more), and putative LOH if between 0.60 and 0.80, or between 1.25 and 1.67 (indicating the decrease of 21% to 39% for one allele).
AXIN2 Mutation Screen
The coding region of AXIN2 was sequenced for mutations exon by exon using primers and conditions as previously described (Lammi et al33 and http://helsinki.fi/science/dentgen).
Statistical Methods
The t test for independent samples or Fisher's exact test (two-tailed) was used to evaluate the statistical significance of differences between groups.
RESULTS
Large Genomic Rearrangements of APC
By MLPA, large genomic deletions were found to underlie 4/29 (14%) of families that tested APC mutation每negative by conventional techniques (group I, Table 2). Three families (FAP5, FAP91, and FAP108) harbored a deletion of the whole APC gene in one chromosomal homolog as suggested by the consistent reduction of APC-specific peaks relative to peaks originating from control sequences (Fig 1). Moreover, samples from two polyposis patients were available from FAP5, and both showed the deletion. The results from MLPA analysis were in agreement with lack of heterozygosity for markers located within APC (listed under Haplotype analysis in Patients and Methods). Heterozygosity for flanking markers D5S1965 and D5S346 provided borders for the APC deletion in FAP5, whereas lack of heterozygosity on the centromeric side was compatible with larger deletions in the other two families. In FAP115, MLPA indicated a deletion of APC exon 4 (Fig 1). This finding was confirmed by amplifying a segment from APC exon 3 to exon 5 from cDNA, which yielded two products: a full-length fragment and a shorter fragment that was shown to lack exon 4 by sequencing (data not shown).
Unbalanced Allelic mRNA Expression of APC
Our primer extension (SNuPE) experiments utilizing the C/T (G/A) polymorphism at nucleotide 1458 in exon 11 identified seven families with unbalanced allelic mRNA expression as the sole constitutional abnormality of APC (group II, seven of 29, 24%; Table 2; Fig 2). The degree of expression reduction was variable and ranged from the level of 68% relative to that of the normal allele (FAP103 with a G-A ratio of 1.46 in cDNA relative to gDNA), to complete silencing of one allele (FAP93). In contrast, families with balanced expression from the present series (indicated as "Normal" in Table 2) showed G-A ratios very close to 1 (ranging from 0.90 to 1.01). For comparison, families with truncating mutations in APC were investigated, and all nine informative patients, who carried eight different mutations, showed balanced mRNA expression in blood, with G-A ratios ranging from 0.97 to 1.07 (Table 3). The same was true for patients from mismatch repair gene mutation每negative HNPCC families with microsatellite-stable tumors (11 informative individuals with G-A ratios between 0.80 and 1.11), as well as for healthy controls as shown in the present study and elsewhere.34 These data suggested that unbalanced mRNA expression of APC was specific to FAP families from the present group II. As further proof of pathogenic significance of unbalanced expression, constitutional reduction of mRNA expression from one APC allele was accompanied by loss of heterozygosity in the APC region in six (75%) of eight informative tumors, and the lost allele was the one not involved in the constitutional abnormality in all cases in which allelic origin could be determined (Table 2). The APC promoter region was unaltered by sequencing, therefore providing no explanation for unbalanced mRNA expression.
Families With No Detectable Germline Changes in APC
Besides showing no changes in the initial screen by PTT, heteroduplex analysis, and exon-specific sequencing, 18 of 29 polyposis families could not be linked to APC even in our scrutinized screen by MLPA and SNuPE. While our investigation into APC was ongoing, a mutation in the Wnt signaling regulator gene AXIN2 was identified in one of these families (FAP113) as a result of a separate effort aiming to identify the gene defect underlying severe tooth agenesis in this family.33 We subsequently screened AXIN2 for germline mutations in the remaining families, but could not find any obvious disease-associated changes. In parallel, samples from our families were also submitted to a MYH screen, and two families (FAP83 and FAP119) revealed biallelic MYH mutations.35 Families with mutations in genes other than APC form group IV in Table 2 leaving 15 (52%) of 29 families without a defined molecular etiology (group III in Table 2).
In group III, APC germline involvement could not be excluded since, for example, seven families were uninformative for SNuPE, and the fact that most families were in fact solitary cases prevented genetic linkage analysis. In two families (FAP57 and FAP74), samples were available from multiple patients, and both displayed a shared disease haplotype within the family (data not shown). Strict or putative LOH in the APC region was present in 11 of 20 informative tumors from group III (55%; Table 2), and, according to Knudson's two-hit hypothesis, could serve to expose either a somatic or a hereditary "first hit." The LOH rate was higher than that seen in APC mutation-positive families (five of 16; 31%; Table 3), and it might be reasonable to expect that the first hit occurred in the germline in at least some cases (eg, FAP57) with additional linkage support. On the other hand, group III comprised some families (such as FAP92, FAP104, and FAP73), in which no APC alterations were found in constitutional or tumor tissues, which might suggest the involvement of (an) other, as yet unknown susceptibility gene(s).
Genetic Versus Clinical Features
Table 1 presents detailed clinical characteristics of the present cohort of FAP families, divided into four groups according to their genetic etiology. Based on polyp counts higher than 100 in at least one member, all families except for FAP78, FAP100, FAP111, FAP113, and FAP119 could be classified as having classical polyposis. Additionally, the consistently low polyp number in several individuals from FAP83 was compatible with the diagnosis of attenuated or atypical polyposis in this family.
Our group I, with large genomic deletions, had clinical features similar to those previously described for FAP families with truncating APC mutations from the same population,22 including profuse polyposis, early age at onset, and frequent extracolonic manifestations. The mean age at polyposis diagnosis was remarkably low (27.3 years) in probands from families with deletion of the entire APC gene (FAP5, FAP91, FAP108). Compared with these, FAP115 with exon 4 deletion showed a milder phenotype, consistent with the location of the change in the "AFAP region" (see Introduction). Families with unbalanced mRNA expression of APC alleles (group II) all had classical FAP, but interestingly, the mean age at onset was higher, and less than half displayed extracolonic manifestations. Compared with APC mutation每positive FAP (35.6 years),22 the mean ages at onset in probands from groups II (47.3 years) and III (45.2 years) were significantly higher (with P values of .04 and .02, respectively, by two-tailed t test). The six families with attenuated/atypical polyposis in our series belonged to groups III (with no direct evidence of APC involvement) and IV (with mutations in AXIN2 or MYH). In particular, none of the patients with MYH-associated polyposis had more than 150 polyps; similarly, the highest polyp count in patients from FAP113 with the AXIN2 R656X mutation33 was 70. Consistent with previous reports,24,25 the present two families with MYH-associated polyposis were among those without clear dominant transmission.
DISCUSSION
Unlike most previous studies that have typically addressed one abnormality at a time, we applied a comprehensive approach to clarify the molecular etiology of "APC mutation每negative" FAP. The families fell into four categories: (I) families with large genomic deletions of APC (n = 4; 14%); (II) families with unbalanced allelic mRNA expression of APC (n = 7; 24%); (III) families with no evidence of genomic deletions or unbalanced mRNA but with no convincing evidence against APC, either (n = 15; 52%); and (IV) families with germline mutations in genes other than APC (AXIN2 in one family, 4%, and MYH in two families, 7%).
Several techniques are available for detecting deletions or insertions of a few base pairs (common types of mutations described for APC9) or very large, cytogenetically visible deletions,36 but difficulties arise in the diagnosis of deletions or duplications of a few kilobases, such as a deletion of a single exon or a deletion of the entire predisposing gene. For this reason, the true frequency of genomic rearrangements in APC mutation每negative FAP is unknown. Based on real-time quantitative PCR or RT-PCR experiments, APC deletion frequencies of 12% to 33% were reported, and genomic deletions were almost exclusively restricted to classical FAP (as opposed to AFAP).31,37 Recently, two new methods, multiplex amplification and probe hybridization and MLPA were designed to screen up to 40 target sequences for deletions or duplications in a single reaction tube.38 Available experience suggests that MLPA provides an easy and reliable approach for the diagnosis of dosage alterations in several disease-associated genes, including HNPCC-associated DNA mismatch repair genes39-41 and the breast and ovarian cancer每associated BRCA1.42 The present data show that MLPA is also valuable for the detection of APC deletions in FAP families. The overall deletion frequency (14% among APC mutation每negative families) and type of deletions (whole gene deletion in three fourths, single exon deletion in one fourth), and association with classical FAP were compatible with results obtained from other populations using earlier methodology (see above). While our manuscript was in preparation, another MLPA study of FAP families was published reporting APC deletions in six (25%) of 24 among APC mutation每negative FAP families from Wessex, United Kingdom; the clinical features of the families were not specified.43
Alterations in one allele may occasionally be masked by the normal sequence present in the other allele, making allele-specific approaches necessary. Concurrent with the development of the PTT test for truncating APC mutations, a radioactive oligoligation-based assay was devised to detect reduced mRNA expression from one APC allele in cases not showing truncations.10 This method was able to link three (27%) of 11 FAP families without truncations to APC. Later, physical isolation of the maternal and paternal APC alleles by the newly developed Conversion technology, combined with digital single nucleotide polymorphism for mRNA quantification, confirmed these initial observations and showed that even modest decreases in transcripts (approximately 50% of that of the normal allele) can lead to FAP predisposition.44 We developed a simple semiquantitative fluorescent primer extension assay to monitor allelic mRNA expression in our families, and seven (24%) of 29 of these, all with classical FAP, exhibited a relative expression reduction varying from the level of 68% of that of the normal allele to complete silencing of one allele. Constitutionally decreased expression from one allele was accompanied by the selective loss of the wild-type allele in tumor tissue, suggesting that reduced dosage of the APC gene product contributed to tumor development in such families.
The mechanism of reduced mRNA expression remains to be identified. It is not associated with apparent DNA sequence changes in the coding region or in the 5' or 3' flanking areas.44 A previously undescribed mechanism may be involved since in our investigation, unbalanced mRNA expression was specific to the subgroup of seven families, being absent in all FAP patients with truncating APC mutations (P = .022 by two-tailed Fisher's exact test) and in patients with microsatellite-stable HNPCC (P = .021). The lack of unbalanced mRNA expression of APC in mutation-positive FAP is in marked contrast to truncating HNPCC-associated DNA mismatch repair gene mutations, which we found to regularly lead to under-representation of the mutation-containing transcripts, presumably due to nonsense-mediated RNA decay.30,45
Our results suggest that MLPA and SNuPE (or direct assays for genomic changes underlying unbalanced expression if available in the future) are valuable additions to standard APC mutation diagnostics. However, as shown by the present group III, even intensive research efforts may not result in the identification of germline changes in all cases with apparent FAP phenotype. In a clinical setting, it is important to be aware that because of genetic heterogeneity, a negative test result does not rule out inherited susceptibility to polyposis and colorectal cancer. In such cases, endoscopic screening should be offered for the siblings and descendants of the affected individuals, as was done in the pretesting era.46
Previous investigations focusing on families with APC mutation每negative FAP have observed a difference in their clinical features compared with those of APC mutation每positive FAP. Two studies found a milder phenotype,20,22 whereas one study47 reported a more severe phenotype for the mutation-negative families. The recognition of genetic heterogeneity was an important advantage of the present investigation making it possible to assess genotype-phenotype correlations separately for the individual subgroups. Our group I with large genomic deletions of APC had clinical features indistinguishable from those described for FAP families with truncating APC mutations. Families with unbalanced mRNA expression of APC alleles (group II) had a generally "milder" phenotype, supporting a distinct mechanism for predisposition. The same broadly applied to group III (families with association to APC neither confirmed nor excluded). Finally, families with germline mutations in AXIN2 or MYH (group IV) often showed attenuated or atypical FAP. Our ability to stratify the original single cohort into at least four molecularly and clinically distinct subgroups significantly defines the genetic basis of FAP and facilitates the design of more tailored strategies for the management of such families.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank Daniel Fix for advice in MLPA analyses; Saila Saarinen, Marjatta Kivek?s, and Jonna Jalanka for expert technical assistance; Jinmin Miao for cell cultures; and Tuula Lehtinen and Katja Kuosa for assistance with sample collection.
NOTES
Supported by grants from the Sigrid Juselius Foundation, the Academy of Finland, the Finnish Cancer Foundation, K. Albin Johansson Foundation, Helsinki University Hospital Research and Development Funds (EVO), and Helsinki Biomedical Graduate School.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Talbot IC, Burt R, J?rvinen H, et al: Familial adenomatous polyposis, in Hamilton SR, Aaltonen LA (eds): World Health Organization Classification of Tumours: Pathology and Genetics of Tumours of the Digestive System. Lyon, France, IARC Press, 2000, pp 120-125
Friedl W, Meuschel S, Caspari R, et al: Attenuated familial adenomatous polyposis due to a mutation in the 3' part of the APC gene: A clue for understanding the function of the APC protein. Hum Genet 97:579-584, 1996
Lynch HT, Watson P: AFAP: Variety is the spice of life. Gut 43:451-452, 1998
Vasen HF: Clinical diagnosis and management of hereditary colorectal cancer syndromes. J Clin Oncol 18:81S-92S, 2000
Kinzler KW, Nilbert MC, Su LK, et al: Identification of FAP locus genes from chromosome 5q21. Science 253:661-665, 1991
Groden J, Thliveris A, Samowitz W, et al: Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66:589-600, 1991
Joslyn G, Carlson M, Thliveris A, et al: Identification of deletion mutations and three new genes at the familial polyposis locus. Cell 66:601-613, 1991
Sieber OM, Tomlinson IP, Lamlum H: The adenomatous polyposis coli (APC) tumour suppressor: Genetics, function and disease. Mol Med Today 6:462-469, 2000
Laurent-Puig P, Beroud C, Soussi T: APC gene: Database of germline and somatic mutations in human tumors and cell lines. Nucleic Acids Res 26:269-270, 1998
Powell SM, Petersen GM, Krush AJ, et al: Molecular diagnosis of familial adenomatous polyposis. N Engl J Med 329:1982-1987, 1993
Nagase H, Nakamura Y: Mutations of the APC (adenomatous polyposis coli) gene. Hum Mut 2:425-434, 1993
Spirio L, Olschwang S, Groden J, et al: Alleles of the APC gene: An attenuated form of familial polyposis. Cell 75:951-957, 1993
Varesco L, Gismondi V, Presciuttini S, et al: Mutation in a splice-donor site of the APC gene in a family with polyposis and late age of colonic cancer death. Hum Genet 93:281-286, 1994
van der Luijt RB, Vasen HF, Tops CM, et al: APC mutation in the alternatively spliced region of exon 9 associated with late onset familial adenomatous polyposis. Hum Genet 96:705-710, 1995
Caspari R, Olschwang S, Friedl W, et al: Familial adenomatous polyposis: Desmoid tumours and lack of ophthalmic lesions (CHRPE) associated with APC mutations beyond codon 1444. Hum Mol Genet 4:337-340, 1995
Miyoshi Y, Ando H, Nagase H, et al: Germ-line mutations of the APC gene in 53 familial adenomatous polyposis patients. Proc Natl Acad Sci U S A 89:4452-4456, 1992
Knudsen AL, Bisgaard ML, Bulow S: Attenuated familial adenomatous polyposis (AFAP): A review of the literature. Fam Cancer 2:43-55, 2003
van der Luijt RB, Meera Khan P, Vasen HF, et al: Germline mutations in the 3' part of APC exon 15 do not result in truncated proteins and are associated with attenuated adenomatous polyposis coli. Hum Genet 98:727-734, 1996
Lamlum H, Al Tassan N, Jaeger E, et al: Germline APC variants in patients with multiple colorectal adenomas, with evidence for the particular importance of E1317Q. Hum Mol Genet 9:2215-2221, 2000
Heinimann K, Thompson A, Locher A, et al: Nontruncating APC germ-line mutations and mismatch repair deficiency play a minor role in APC mutation-negative polyposis. Cancer Res 61:7616-7622, 2001
Friedl W, Caspari R, Sengteller M, et al: Can APC mutation analysis contribute to therapeutic decisions in familial adenomatous polyposis? Experience from 680 FAP families. Gut 48:515-521, 2001
Moisio AL, Jarvinen H, Peltomaki P: Genetic and clinical characterisation of familial adenomatous polyposis: A population-based study. Gut 50:845-850, 2002
Al-Tassan N, Chmiel NH, Maynard JF, et al: Inherited variants of MYH associated with somatic G:CT:A mutations in colorectal tumors. Nat Genet 30:227-232, 2002
Sampson JR, Dolwani S, Jones S, et al: Autosomal recessive colorectal adenomatous polyposis due to inherited mutations of MYH. Lancet 362:39-41, 2003
Sieber OM, Lipton L, Crabtree M, et al: Multiple colorectal adenomas, classic adenomatous polyposis, and germ-line mutations in MYH. N Engl J Med 348:791-799, 2003
Lahiri DK, Nurnberger JI: A rapid non-enzymatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res 19:5444, 1991
Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159, 1987
Isola J, DeVries S, Chu L, et al: Analysis of changes in DNA sequence copy number by comparative genomic hybridization in archival paraffin-embedded tumor samples. Am J Pathol 145:1301-1308, 1994
Schouten JP, McElgunn CJ, Waaijer R, et al: Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res 30:e57, 2002
Renkonen E, Zhang Y, Lohi H, et al: Altered expression of MLH1, MSH2, and MSH6 in predisposition to hereditary nonpolyposis colorectal cancer. J Clin Oncol 21:3629-3637, 2003
Sieber OM, Lamlum H, Crabtree MD, et al: Whole-gene APC deletions cause classical familial adenomatous polyposis, but not attenuated polyposis or "multiple" colorectal adenomas. Proc Natl Acad Sci U S A 99:2954-2958, 2002
Peltomaki P, Sistonen P, Mecklin JP, et al: Evidence that the MCC-APC gene region in 5q21 is not the site for susceptibility to hereditary nonpolyposis colorectal carcinoma. Cancer Res 52:4530-4533, 1992
Lammi L, Arte S, Somer M, et al: Mutations in AXIN2 cause familial tooth agenesis and predispose to colorectal cancer. Am J Hum Genet 74:1043-1050, 2004
Yan H, Yuan W, Velculescu VE, et al: Allelic variation in human gene expression. Science 297:1143, 2002
Alhopuro P, Parker AR, Lehtonen R, et al: A novel functionally deficient MYH variant in individuals with colorectal adenomatous polyposis. Hum Mut (2005, in press)
Herrera L, Kakati S, Gibas L, et al: Gardner syndrome in a man with an interstitial deletion of 5q. Am J Med Genet 25:473-476, 1986
Venesio T, Balsamo A, Rondo-Spaudo M, et al: APC haploinsufficiency, but not CTNNB1 or CDH1 gene mutations, accounts for a fraction of familial adenomatous polyposis patients without APC truncating mutations. Lab Invest 83:1859-1866, 2003
Sellner LN, Taylor GR: MLPA and MAPH: New techniques for detection of gene deletions. Hum Mut 23:413-419, 2004
Gille JJ, Hogervorst FB, Pals G, et al: Genomic deletions of MSH2 and MLH1 in colorectal cancer families detected by a novel mutation detection approach. Br J Cancer 87:892-897, 2002
Taylor CF, Charlton RS, Burn J, et al: Genomic deletions in MSH2 or MLH1 are a frequent cause of hereditary non-polyposis colorectal cancer: Identification of novel and recurrent deletions by MLPA. Hum Mut 22:428-433, 2003
Nakagawa H, Hampel H, de la Chapelle A: Identification and characterization of genomic rearrangements of MSH2 and MLH1 in Lynch syndrome (HNPCC) by novel techniques. Hum Mut 22:258, 2003
Hogervorst FB, Nederlof PM, Gille JJ, et al: Large genomic deletions and duplications in the BRCA1 gene identified by a novel quantitative method. Cancer Res 63:1449-1453, 2003
Bunyan DJ, Eccles DM, Sillibourne J, et al: Dosage analysis of cancer predisposition genes by multiplex ligation-dependent probe amplification. Br J Cancer 91:1155-1159, 2004
Yan H, Dobbie Z, Gruber SB, et al: Small changes in expression affect predisposition to tumorigenesis. Nat Genet 30:25-26, 2002
Renkonen E, Lohi H, Jarvinen HJ, et al: Novel splicing associations of hereditary colon cancer related DNA mismatch repair gene mutations. J Med Genet 41:e95, 2004
Church J, Lowry A, Simmang C: Practice parameters for the identification and testing of patients at risk for dominantly inherited colorectal cancer: Supporting documentation. Dis Colon Rectum 44:1404-1412, 2001
Bisgaard ML, Ripa R, Knudsen AL, et al: Familial adenomatous polyposis patients without an identified APC germline mutation have a severe phenotype. Gut 53:266-270, 2004(Elise T. Renkonen, Pekka )
Laboratory of Molecular Genetics, Department of Oral and Maxillofacial Diseases, and Department of Surgery, Helsinki University Central Hospital, Helsinki, Finland
ABSTRACT
PURPOSE: One third of families with classical adenomatous polyposis (FAP), and a majority of those with attenuated FAP (AFAP), remain APC mutation每negative by conventional methods. Our purpose was to clarify the genetic basis of polyposis and genotype-phenotype correlations in such families.
PATIENTS AND METHODS: We studied a cohort of 29 adenomatous polyposis families that had screened APC mutation每negative by the protein truncation test, heteroduplex analysis, and exon-specific sequencing. The APC gene was investigated for large genomic rearrangements by multiplex ligation-dependent probe amplification (MLPA), and for allelic mRNA expression by single nucleotide primer extension (SNuPE). The AXIN2 gene was screened for mutations by sequencing.
RESULTS: Four families (14%) showed a constitutional deletion of the entire APC gene (three families) or a single exon (one family). Seven families (24%) revealed reduced or extinct mRNA expression from one APC allele in blood, accompanied by loss of heterozygosity in the APC region in six (75%) of eight tumors. In 15 families (52%), possible APC involvement could be neither confirmed nor excluded. Finally, as detailed elsewhere, three families (10%) had germline mutations in genes other than APC, AXIN2 in one family, and MYH in two families.
CONCLUSION: "APC mutation每negative" FAP is genetically heterogeneous, and a combination of MLPA and SNuPE is able to link a considerable proportion (38%) to APC. Significant differences were observed in clinical manifestations between subgroups, emphasizing the importance of accurate genetic and clinical characterization for the proper management of such families.
INTRODUCTION
Familial adenomatous polyposis (FAP) is an autosomal dominant syndrome characterized by the development by age 20 years of multiple colorectal polyps; one or several of these polyps progresses to cancer by approximately 40 years of age.1 In classical FAP, at least 100 (often hundreds to more than a thousand) adenomas are present, whereas polyposis with fewer than 100 adenomas per patient is referred to as attenuated polyposis (AFAP or AAPC). In addition to colorectal polyps, affected individuals may develop various extracolonic manifestations, such as desmoid tumors, duodenal adenomas, mandibular osteomas, and hypertrophic pigmentary lesions of the retina. In AFAP, the age at onset of adenomas and cancer is higher, and the lifetime risk of cancer is lower compared with classical FAP.2 While the true incidence and frequency of AFAP is unknown,3 it may account for up to 10% of adenomatous polyposis families.4
Germline mutations of the adenomatous polyposis coli (APC) tumor suppressor gene located on 5q21-22 cause both FAP and AFAP.5-7 To date, more than 800 pathogenic mutations in APC have been described.8 These are typically frameshift or nonsense changes that lead to truncated protein products,9 forming the basis for their detection by the protein truncation test (PTT).10 In classical FAP, mutations may be scattered all over the gene, with some hot spots in the somatic mutation cluster region between codons 1250 and 1450 in exon 15.11 The position of APC mutation correlates to some extent with clinical features12-14; for example, mutations between codons 1444 to 1578 are associated with classical adenomatous polyposis and multiple extracolonic manifestations.15,16 AFAP phenotype is associated with mutations in the 5' end (before codon 157 in exon 4), in the alternatively spliced part of exon 9, and in the distal 3' end of APC (beyond codon 1595).17
Conventional methods for APC mutation detection include the PTT test and various exon-specific screening methods such as single-strand conformation polymorphism assay, heteroduplex analysis, and DNA sequencing. These methods are sensitive to point mutations, but they fail to detect large genomic rearrangements as well as possible transcription defects. Conventional techniques leave approximately 30% of families with classical FAP and approximately 90% of AFAP families APC mutation每negative.12,18-22 This apparent mutation negativity may suggest either that APC has alterations that escape detection by routine techniques, or alternatively, that other (known or unknown) genes are involved in FAP predisposition. Among known genes, recent evidence implicates at least the Mut Y homolog MYH encoding a DNA glycosylase.23 Recessive inheritance and variable polyp count (lower or greater than 100) are features of MYH-associated polyposis.24,25 MYH mutations may explain up to 25% of "APC mutation每negative" FAP, leaving the majority of FAP families without identifiable APC mutations molecularly unexplained.
We recently conducted a study on 65 classical FAP families from Finland, which resulted in the identification of APC germline mutations in 72%.22 The present report addresses the genetic basis of polyposis in the third (approximate) of FAP families that shows no APC mutation by conventional methods.
PATIENTS AND METHODS
Subjects
We investigated 29 families (38 individuals) representing the mutation-negative fraction (30%) among a total of 98 adenomatous polyposis families from the Hereditary Colorectal Cancer Registry of Finland that had been screened for APC germline mutations by common diagnostic methods (PTT for the entire coding region and heteroduplex analysis of exons 1 to 14, or sequencing of exons 1 to 15).22 Table 1 presents clinical features of the families. Eighteen FAP families with truncating mutations in APC (18 different mutations, 26 individuals) were studied for comparison. As controls, two healthy individuals (unaffected with any cancer), as well as 20 colon cancer patients from DNA mismatch repair gene mutation每negative families with hereditary nonpolyposis colorectal carcinoma (HNPCC), were tested. This study was approved by the institutional review board of the Helsinki University Central Hospital (Helsinki, Finland).
DNA and RNA Extraction
DNA was extracted from cultured lymphocytes or lymphoblasts.26 Total cellular RNA was extracted according to the guanidium thiocyanate method.27 Normal and tumor DNA from paraffin-embedded tissue sections was extracted according to Isola et al.28
MLPA for Large Genomic Rearrangements
Multiplex ligation-dependent probe amplification (MLPA)29 was used to screen DNA samples for large genomic rearrangements of APC according to the manufacturer's instructions (Salsa P043 kit; MRC-Holland, Amsterdam, the Netherlands). The MLPA method is based on sequence-specific probe hybridization to genomic DNA, followed by polymerase chain reaction (PCR) amplification of the hybridized probe (with one amplification primer fluorescently labeled), and semiquantitative analysis of the PCR products. Target-specific products are identified according to their differential length. The Salsa P043 kit contains probes for each coding exon of APC, three probes for the APC promoter region, and 11 probes for other human genes located on different chromosomes. FAM-labeled PCR products of MLPA reactions were analyzed on an ABI 3730 sequencer using Genemapper software, version 3.0 (Applied Biosystems, Foster City, CA). Deletion was suspected when the peak area was reduced more than 40% to 50% compared with normal controls. The diagnosis was based on at least two repeated MLPA experiments and was confirmed by the observation of the involvement of contiguous exons (whole-gene deletions) or the demonstration of a deletion-specific fragment in cDNA (exon 4 deletion).
SNuPE for Allelic mRNA Expression
The basic principles of the Single Nucleotide Primer Extension (SNuPE) method were described in our previous publication.30 For the present analyses, the coding polymorphism C/T at nucleotide 1458 of APC exon 11 was selected because of its high heterozygosity values and location in a short exon flanked by introns, making it possible to design gDNA- and cDNA-specific PCR reactions appropriate for this purpose. SNuPE analyses were performed on reverse transcription (RT) -PCR products of blood RNA encompassing the polymorphic site (amplified with forward primer from exon 11, 5'-GACTACAGGCCATTGCAGAA-3', and reverse primer from exon 12, 5'-GCCTTGTTGGCTACATCTCC-3'). The results were compared with amplification products obtained from genomic DNA of the same individuals (using the same forward primer as above, combined with reverse primer 5'-AAAGCCATTCCAGCATATCG-3' from exon 11). With these amplification products as templates, extension reaction was carried out from reverse direction using a fluorescently labeled primer (5'-GTGTAATACTGTAGTGGTCATTAGTAAG-3') and dATP, dCTP, and dTTP for chain elongation, and ddGTP for chain termination. Primer extension products, 34 bp for G (C) allele and 44 bp for A (T) allele, were run on an automated sequencer and analyzed using Genemapper. Ratios of allelic peak areas in cDNA relative to gDNA were calculated, and values below 0.6 or above 1.7 (indicating that the transcript of one allele had decreased 40% or more) were considered abnormal based on validation experiments performed in a large number of individuals.
Analysis of APC Promoter Region for Mutations
The APC promoter region (U02509; http://www.ncbi.nlm.nih.gov) was amplified with primers 5-GCTAGCATAGCTTTTCTGGTAAC-3' and 5'-CAGTGACACCCTGGCGGGCTG-3' and sequenced for alterations.
APC Haplotype Analysis
For haplotype analysis, the following microsatellite markers or single nucleotide polymorphisms were studied (genetic or physical distance between markers in parentheses): centromere, D5S409, (9.6 cM), D5S1965, (200 kb) [APC promoter a/g polymorphism, rs 201972031, APC intron 7 a/t polymorphism, rs 1914,31 APC exon 11 C/T polymorphism (nt. 1458), APC exon 15 A/G polymorphism (nt. 5037)],31 (< 100 kb), D5S346, (100 kb), MCC,32 telomere. Primer sequences for microsatellite markers can be found at http://www.gdb.org.
Loss of Heterozygosity in Tumor Tissue
APC exon 11 polymorphism was analyzed by SNuPE as described above, except that paired tumor and normal DNA samples were studied. Additionally, microsatellite markers closely linked to APC, D5S1965 (200 kb upstream of APC), and D5S346 (< 100 kb downstream of APC), were investigated. The marker-specific fluorescent amplification products were run on an automated sequencer and analyzed by Genemapper. The ratios of allelic peak areas in tumor DNA relative to normal DNA were interpreted to suggest loss of heterozygosity (LOH) if below 0.60 or above 1.67 (indicating that one allele had decreased 40% or more), and putative LOH if between 0.60 and 0.80, or between 1.25 and 1.67 (indicating the decrease of 21% to 39% for one allele).
AXIN2 Mutation Screen
The coding region of AXIN2 was sequenced for mutations exon by exon using primers and conditions as previously described (Lammi et al33 and http://helsinki.fi/science/dentgen).
Statistical Methods
The t test for independent samples or Fisher's exact test (two-tailed) was used to evaluate the statistical significance of differences between groups.
RESULTS
Large Genomic Rearrangements of APC
By MLPA, large genomic deletions were found to underlie 4/29 (14%) of families that tested APC mutation每negative by conventional techniques (group I, Table 2). Three families (FAP5, FAP91, and FAP108) harbored a deletion of the whole APC gene in one chromosomal homolog as suggested by the consistent reduction of APC-specific peaks relative to peaks originating from control sequences (Fig 1). Moreover, samples from two polyposis patients were available from FAP5, and both showed the deletion. The results from MLPA analysis were in agreement with lack of heterozygosity for markers located within APC (listed under Haplotype analysis in Patients and Methods). Heterozygosity for flanking markers D5S1965 and D5S346 provided borders for the APC deletion in FAP5, whereas lack of heterozygosity on the centromeric side was compatible with larger deletions in the other two families. In FAP115, MLPA indicated a deletion of APC exon 4 (Fig 1). This finding was confirmed by amplifying a segment from APC exon 3 to exon 5 from cDNA, which yielded two products: a full-length fragment and a shorter fragment that was shown to lack exon 4 by sequencing (data not shown).
Unbalanced Allelic mRNA Expression of APC
Our primer extension (SNuPE) experiments utilizing the C/T (G/A) polymorphism at nucleotide 1458 in exon 11 identified seven families with unbalanced allelic mRNA expression as the sole constitutional abnormality of APC (group II, seven of 29, 24%; Table 2; Fig 2). The degree of expression reduction was variable and ranged from the level of 68% relative to that of the normal allele (FAP103 with a G-A ratio of 1.46 in cDNA relative to gDNA), to complete silencing of one allele (FAP93). In contrast, families with balanced expression from the present series (indicated as "Normal" in Table 2) showed G-A ratios very close to 1 (ranging from 0.90 to 1.01). For comparison, families with truncating mutations in APC were investigated, and all nine informative patients, who carried eight different mutations, showed balanced mRNA expression in blood, with G-A ratios ranging from 0.97 to 1.07 (Table 3). The same was true for patients from mismatch repair gene mutation每negative HNPCC families with microsatellite-stable tumors (11 informative individuals with G-A ratios between 0.80 and 1.11), as well as for healthy controls as shown in the present study and elsewhere.34 These data suggested that unbalanced mRNA expression of APC was specific to FAP families from the present group II. As further proof of pathogenic significance of unbalanced expression, constitutional reduction of mRNA expression from one APC allele was accompanied by loss of heterozygosity in the APC region in six (75%) of eight informative tumors, and the lost allele was the one not involved in the constitutional abnormality in all cases in which allelic origin could be determined (Table 2). The APC promoter region was unaltered by sequencing, therefore providing no explanation for unbalanced mRNA expression.
Families With No Detectable Germline Changes in APC
Besides showing no changes in the initial screen by PTT, heteroduplex analysis, and exon-specific sequencing, 18 of 29 polyposis families could not be linked to APC even in our scrutinized screen by MLPA and SNuPE. While our investigation into APC was ongoing, a mutation in the Wnt signaling regulator gene AXIN2 was identified in one of these families (FAP113) as a result of a separate effort aiming to identify the gene defect underlying severe tooth agenesis in this family.33 We subsequently screened AXIN2 for germline mutations in the remaining families, but could not find any obvious disease-associated changes. In parallel, samples from our families were also submitted to a MYH screen, and two families (FAP83 and FAP119) revealed biallelic MYH mutations.35 Families with mutations in genes other than APC form group IV in Table 2 leaving 15 (52%) of 29 families without a defined molecular etiology (group III in Table 2).
In group III, APC germline involvement could not be excluded since, for example, seven families were uninformative for SNuPE, and the fact that most families were in fact solitary cases prevented genetic linkage analysis. In two families (FAP57 and FAP74), samples were available from multiple patients, and both displayed a shared disease haplotype within the family (data not shown). Strict or putative LOH in the APC region was present in 11 of 20 informative tumors from group III (55%; Table 2), and, according to Knudson's two-hit hypothesis, could serve to expose either a somatic or a hereditary "first hit." The LOH rate was higher than that seen in APC mutation-positive families (five of 16; 31%; Table 3), and it might be reasonable to expect that the first hit occurred in the germline in at least some cases (eg, FAP57) with additional linkage support. On the other hand, group III comprised some families (such as FAP92, FAP104, and FAP73), in which no APC alterations were found in constitutional or tumor tissues, which might suggest the involvement of (an) other, as yet unknown susceptibility gene(s).
Genetic Versus Clinical Features
Table 1 presents detailed clinical characteristics of the present cohort of FAP families, divided into four groups according to their genetic etiology. Based on polyp counts higher than 100 in at least one member, all families except for FAP78, FAP100, FAP111, FAP113, and FAP119 could be classified as having classical polyposis. Additionally, the consistently low polyp number in several individuals from FAP83 was compatible with the diagnosis of attenuated or atypical polyposis in this family.
Our group I, with large genomic deletions, had clinical features similar to those previously described for FAP families with truncating APC mutations from the same population,22 including profuse polyposis, early age at onset, and frequent extracolonic manifestations. The mean age at polyposis diagnosis was remarkably low (27.3 years) in probands from families with deletion of the entire APC gene (FAP5, FAP91, FAP108). Compared with these, FAP115 with exon 4 deletion showed a milder phenotype, consistent with the location of the change in the "AFAP region" (see Introduction). Families with unbalanced mRNA expression of APC alleles (group II) all had classical FAP, but interestingly, the mean age at onset was higher, and less than half displayed extracolonic manifestations. Compared with APC mutation每positive FAP (35.6 years),22 the mean ages at onset in probands from groups II (47.3 years) and III (45.2 years) were significantly higher (with P values of .04 and .02, respectively, by two-tailed t test). The six families with attenuated/atypical polyposis in our series belonged to groups III (with no direct evidence of APC involvement) and IV (with mutations in AXIN2 or MYH). In particular, none of the patients with MYH-associated polyposis had more than 150 polyps; similarly, the highest polyp count in patients from FAP113 with the AXIN2 R656X mutation33 was 70. Consistent with previous reports,24,25 the present two families with MYH-associated polyposis were among those without clear dominant transmission.
DISCUSSION
Unlike most previous studies that have typically addressed one abnormality at a time, we applied a comprehensive approach to clarify the molecular etiology of "APC mutation每negative" FAP. The families fell into four categories: (I) families with large genomic deletions of APC (n = 4; 14%); (II) families with unbalanced allelic mRNA expression of APC (n = 7; 24%); (III) families with no evidence of genomic deletions or unbalanced mRNA but with no convincing evidence against APC, either (n = 15; 52%); and (IV) families with germline mutations in genes other than APC (AXIN2 in one family, 4%, and MYH in two families, 7%).
Several techniques are available for detecting deletions or insertions of a few base pairs (common types of mutations described for APC9) or very large, cytogenetically visible deletions,36 but difficulties arise in the diagnosis of deletions or duplications of a few kilobases, such as a deletion of a single exon or a deletion of the entire predisposing gene. For this reason, the true frequency of genomic rearrangements in APC mutation每negative FAP is unknown. Based on real-time quantitative PCR or RT-PCR experiments, APC deletion frequencies of 12% to 33% were reported, and genomic deletions were almost exclusively restricted to classical FAP (as opposed to AFAP).31,37 Recently, two new methods, multiplex amplification and probe hybridization and MLPA were designed to screen up to 40 target sequences for deletions or duplications in a single reaction tube.38 Available experience suggests that MLPA provides an easy and reliable approach for the diagnosis of dosage alterations in several disease-associated genes, including HNPCC-associated DNA mismatch repair genes39-41 and the breast and ovarian cancer每associated BRCA1.42 The present data show that MLPA is also valuable for the detection of APC deletions in FAP families. The overall deletion frequency (14% among APC mutation每negative families) and type of deletions (whole gene deletion in three fourths, single exon deletion in one fourth), and association with classical FAP were compatible with results obtained from other populations using earlier methodology (see above). While our manuscript was in preparation, another MLPA study of FAP families was published reporting APC deletions in six (25%) of 24 among APC mutation每negative FAP families from Wessex, United Kingdom; the clinical features of the families were not specified.43
Alterations in one allele may occasionally be masked by the normal sequence present in the other allele, making allele-specific approaches necessary. Concurrent with the development of the PTT test for truncating APC mutations, a radioactive oligoligation-based assay was devised to detect reduced mRNA expression from one APC allele in cases not showing truncations.10 This method was able to link three (27%) of 11 FAP families without truncations to APC. Later, physical isolation of the maternal and paternal APC alleles by the newly developed Conversion technology, combined with digital single nucleotide polymorphism for mRNA quantification, confirmed these initial observations and showed that even modest decreases in transcripts (approximately 50% of that of the normal allele) can lead to FAP predisposition.44 We developed a simple semiquantitative fluorescent primer extension assay to monitor allelic mRNA expression in our families, and seven (24%) of 29 of these, all with classical FAP, exhibited a relative expression reduction varying from the level of 68% of that of the normal allele to complete silencing of one allele. Constitutionally decreased expression from one allele was accompanied by the selective loss of the wild-type allele in tumor tissue, suggesting that reduced dosage of the APC gene product contributed to tumor development in such families.
The mechanism of reduced mRNA expression remains to be identified. It is not associated with apparent DNA sequence changes in the coding region or in the 5' or 3' flanking areas.44 A previously undescribed mechanism may be involved since in our investigation, unbalanced mRNA expression was specific to the subgroup of seven families, being absent in all FAP patients with truncating APC mutations (P = .022 by two-tailed Fisher's exact test) and in patients with microsatellite-stable HNPCC (P = .021). The lack of unbalanced mRNA expression of APC in mutation-positive FAP is in marked contrast to truncating HNPCC-associated DNA mismatch repair gene mutations, which we found to regularly lead to under-representation of the mutation-containing transcripts, presumably due to nonsense-mediated RNA decay.30,45
Our results suggest that MLPA and SNuPE (or direct assays for genomic changes underlying unbalanced expression if available in the future) are valuable additions to standard APC mutation diagnostics. However, as shown by the present group III, even intensive research efforts may not result in the identification of germline changes in all cases with apparent FAP phenotype. In a clinical setting, it is important to be aware that because of genetic heterogeneity, a negative test result does not rule out inherited susceptibility to polyposis and colorectal cancer. In such cases, endoscopic screening should be offered for the siblings and descendants of the affected individuals, as was done in the pretesting era.46
Previous investigations focusing on families with APC mutation每negative FAP have observed a difference in their clinical features compared with those of APC mutation每positive FAP. Two studies found a milder phenotype,20,22 whereas one study47 reported a more severe phenotype for the mutation-negative families. The recognition of genetic heterogeneity was an important advantage of the present investigation making it possible to assess genotype-phenotype correlations separately for the individual subgroups. Our group I with large genomic deletions of APC had clinical features indistinguishable from those described for FAP families with truncating APC mutations. Families with unbalanced mRNA expression of APC alleles (group II) had a generally "milder" phenotype, supporting a distinct mechanism for predisposition. The same broadly applied to group III (families with association to APC neither confirmed nor excluded). Finally, families with germline mutations in AXIN2 or MYH (group IV) often showed attenuated or atypical FAP. Our ability to stratify the original single cohort into at least four molecularly and clinically distinct subgroups significantly defines the genetic basis of FAP and facilitates the design of more tailored strategies for the management of such families.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank Daniel Fix for advice in MLPA analyses; Saila Saarinen, Marjatta Kivek?s, and Jonna Jalanka for expert technical assistance; Jinmin Miao for cell cultures; and Tuula Lehtinen and Katja Kuosa for assistance with sample collection.
NOTES
Supported by grants from the Sigrid Juselius Foundation, the Academy of Finland, the Finnish Cancer Foundation, K. Albin Johansson Foundation, Helsinki University Hospital Research and Development Funds (EVO), and Helsinki Biomedical Graduate School.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Talbot IC, Burt R, J?rvinen H, et al: Familial adenomatous polyposis, in Hamilton SR, Aaltonen LA (eds): World Health Organization Classification of Tumours: Pathology and Genetics of Tumours of the Digestive System. Lyon, France, IARC Press, 2000, pp 120-125
Friedl W, Meuschel S, Caspari R, et al: Attenuated familial adenomatous polyposis due to a mutation in the 3' part of the APC gene: A clue for understanding the function of the APC protein. Hum Genet 97:579-584, 1996
Lynch HT, Watson P: AFAP: Variety is the spice of life. Gut 43:451-452, 1998
Vasen HF: Clinical diagnosis and management of hereditary colorectal cancer syndromes. J Clin Oncol 18:81S-92S, 2000
Kinzler KW, Nilbert MC, Su LK, et al: Identification of FAP locus genes from chromosome 5q21. Science 253:661-665, 1991
Groden J, Thliveris A, Samowitz W, et al: Identification and characterization of the familial adenomatous polyposis coli gene. Cell 66:589-600, 1991
Joslyn G, Carlson M, Thliveris A, et al: Identification of deletion mutations and three new genes at the familial polyposis locus. Cell 66:601-613, 1991
Sieber OM, Tomlinson IP, Lamlum H: The adenomatous polyposis coli (APC) tumour suppressor: Genetics, function and disease. Mol Med Today 6:462-469, 2000
Laurent-Puig P, Beroud C, Soussi T: APC gene: Database of germline and somatic mutations in human tumors and cell lines. Nucleic Acids Res 26:269-270, 1998
Powell SM, Petersen GM, Krush AJ, et al: Molecular diagnosis of familial adenomatous polyposis. N Engl J Med 329:1982-1987, 1993
Nagase H, Nakamura Y: Mutations of the APC (adenomatous polyposis coli) gene. Hum Mut 2:425-434, 1993
Spirio L, Olschwang S, Groden J, et al: Alleles of the APC gene: An attenuated form of familial polyposis. Cell 75:951-957, 1993
Varesco L, Gismondi V, Presciuttini S, et al: Mutation in a splice-donor site of the APC gene in a family with polyposis and late age of colonic cancer death. Hum Genet 93:281-286, 1994
van der Luijt RB, Vasen HF, Tops CM, et al: APC mutation in the alternatively spliced region of exon 9 associated with late onset familial adenomatous polyposis. Hum Genet 96:705-710, 1995
Caspari R, Olschwang S, Friedl W, et al: Familial adenomatous polyposis: Desmoid tumours and lack of ophthalmic lesions (CHRPE) associated with APC mutations beyond codon 1444. Hum Mol Genet 4:337-340, 1995
Miyoshi Y, Ando H, Nagase H, et al: Germ-line mutations of the APC gene in 53 familial adenomatous polyposis patients. Proc Natl Acad Sci U S A 89:4452-4456, 1992
Knudsen AL, Bisgaard ML, Bulow S: Attenuated familial adenomatous polyposis (AFAP): A review of the literature. Fam Cancer 2:43-55, 2003
van der Luijt RB, Meera Khan P, Vasen HF, et al: Germline mutations in the 3' part of APC exon 15 do not result in truncated proteins and are associated with attenuated adenomatous polyposis coli. Hum Genet 98:727-734, 1996
Lamlum H, Al Tassan N, Jaeger E, et al: Germline APC variants in patients with multiple colorectal adenomas, with evidence for the particular importance of E1317Q. Hum Mol Genet 9:2215-2221, 2000
Heinimann K, Thompson A, Locher A, et al: Nontruncating APC germ-line mutations and mismatch repair deficiency play a minor role in APC mutation-negative polyposis. Cancer Res 61:7616-7622, 2001
Friedl W, Caspari R, Sengteller M, et al: Can APC mutation analysis contribute to therapeutic decisions in familial adenomatous polyposis? Experience from 680 FAP families. Gut 48:515-521, 2001
Moisio AL, Jarvinen H, Peltomaki P: Genetic and clinical characterisation of familial adenomatous polyposis: A population-based study. Gut 50:845-850, 2002
Al-Tassan N, Chmiel NH, Maynard JF, et al: Inherited variants of MYH associated with somatic G:CT:A mutations in colorectal tumors. Nat Genet 30:227-232, 2002
Sampson JR, Dolwani S, Jones S, et al: Autosomal recessive colorectal adenomatous polyposis due to inherited mutations of MYH. Lancet 362:39-41, 2003
Sieber OM, Lipton L, Crabtree M, et al: Multiple colorectal adenomas, classic adenomatous polyposis, and germ-line mutations in MYH. N Engl J Med 348:791-799, 2003
Lahiri DK, Nurnberger JI: A rapid non-enzymatic method for the preparation of HMW DNA from blood for RFLP studies. Nucleic Acids Res 19:5444, 1991
Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159, 1987
Isola J, DeVries S, Chu L, et al: Analysis of changes in DNA sequence copy number by comparative genomic hybridization in archival paraffin-embedded tumor samples. Am J Pathol 145:1301-1308, 1994
Schouten JP, McElgunn CJ, Waaijer R, et al: Relative quantification of 40 nucleic acid sequences by multiplex ligation-dependent probe amplification. Nucleic Acids Res 30:e57, 2002
Renkonen E, Zhang Y, Lohi H, et al: Altered expression of MLH1, MSH2, and MSH6 in predisposition to hereditary nonpolyposis colorectal cancer. J Clin Oncol 21:3629-3637, 2003
Sieber OM, Lamlum H, Crabtree MD, et al: Whole-gene APC deletions cause classical familial adenomatous polyposis, but not attenuated polyposis or "multiple" colorectal adenomas. Proc Natl Acad Sci U S A 99:2954-2958, 2002
Peltomaki P, Sistonen P, Mecklin JP, et al: Evidence that the MCC-APC gene region in 5q21 is not the site for susceptibility to hereditary nonpolyposis colorectal carcinoma. Cancer Res 52:4530-4533, 1992
Lammi L, Arte S, Somer M, et al: Mutations in AXIN2 cause familial tooth agenesis and predispose to colorectal cancer. Am J Hum Genet 74:1043-1050, 2004
Yan H, Yuan W, Velculescu VE, et al: Allelic variation in human gene expression. Science 297:1143, 2002
Alhopuro P, Parker AR, Lehtonen R, et al: A novel functionally deficient MYH variant in individuals with colorectal adenomatous polyposis. Hum Mut (2005, in press)
Herrera L, Kakati S, Gibas L, et al: Gardner syndrome in a man with an interstitial deletion of 5q. Am J Med Genet 25:473-476, 1986
Venesio T, Balsamo A, Rondo-Spaudo M, et al: APC haploinsufficiency, but not CTNNB1 or CDH1 gene mutations, accounts for a fraction of familial adenomatous polyposis patients without APC truncating mutations. Lab Invest 83:1859-1866, 2003
Sellner LN, Taylor GR: MLPA and MAPH: New techniques for detection of gene deletions. Hum Mut 23:413-419, 2004
Gille JJ, Hogervorst FB, Pals G, et al: Genomic deletions of MSH2 and MLH1 in colorectal cancer families detected by a novel mutation detection approach. Br J Cancer 87:892-897, 2002
Taylor CF, Charlton RS, Burn J, et al: Genomic deletions in MSH2 or MLH1 are a frequent cause of hereditary non-polyposis colorectal cancer: Identification of novel and recurrent deletions by MLPA. Hum Mut 22:428-433, 2003
Nakagawa H, Hampel H, de la Chapelle A: Identification and characterization of genomic rearrangements of MSH2 and MLH1 in Lynch syndrome (HNPCC) by novel techniques. Hum Mut 22:258, 2003
Hogervorst FB, Nederlof PM, Gille JJ, et al: Large genomic deletions and duplications in the BRCA1 gene identified by a novel quantitative method. Cancer Res 63:1449-1453, 2003
Bunyan DJ, Eccles DM, Sillibourne J, et al: Dosage analysis of cancer predisposition genes by multiplex ligation-dependent probe amplification. Br J Cancer 91:1155-1159, 2004
Yan H, Dobbie Z, Gruber SB, et al: Small changes in expression affect predisposition to tumorigenesis. Nat Genet 30:25-26, 2002
Renkonen E, Lohi H, Jarvinen HJ, et al: Novel splicing associations of hereditary colon cancer related DNA mismatch repair gene mutations. J Med Genet 41:e95, 2004
Church J, Lowry A, Simmang C: Practice parameters for the identification and testing of patients at risk for dominantly inherited colorectal cancer: Supporting documentation. Dis Colon Rectum 44:1404-1412, 2001
Bisgaard ML, Ripa R, Knudsen AL, et al: Familial adenomatous polyposis patients without an identified APC germline mutation have a severe phenotype. Gut 53:266-270, 2004(Elise T. Renkonen, Pekka )