Portal de Periódicos da CAPES

MLH1 Germline Epimutations as a Factor in Hereditary Nonpolyposis Colorectal Cancer

2005; Elsevier BV; Volume: 129; Issue: 5 Linguagem: Inglês

10.1053/j.gastro.2005.09.003

1528-0012

Megan P. Hitchins, Rachel Williams, Kayfong Cheong, Nimita Halani, Vita Lin, Deborah Packham, Sue Ku, Andrew Buckle, Nicholas J. Hawkins, John Burn, Steven Gallinger, Jack Goldblatt, Judy Kirk, Ian Tomlinson, Rodney J. Scott, Allan D. Spigelman, Catherine M. Suter, David I. K. Martin, Graeme Suthers, Robyn L. Ward,

Colorectal Cancer Screening and Detection

Background & Aims: Hereditary nonpolyposis colorectal cancer (HNPCC) is caused by heterozygous germline sequence mutations of DNA mismatch repair genes, most frequently MLH1 or MSH2. A novel molecular mechanism for HNPCC has recently been suggested by the finding of individuals with soma-wide monoallelic hypermethylation of the MLH1 gene promoter. In this study, we determined the frequency and role of germline epimutations of MLH1 in HNPCC. Methods: A cohort of 160 probands from HNPCC families who did not harbor germline sequence mutations in the mismatch repair genes were screened for methylation of the MLH1 and EPM2AIP1 promoters by combined bisulfite and restriction analyses. Allelic expression and family transmission of MLH1 were determined using polymorphisms in intron 4 and the 3′ untranslated region. Results: One of 160 individuals had monoallelic MLH1 hypermethylation in peripheral blood, hair follicles, and buccal mucosa, indicative of a soma-wide alteration. Monoallelic transcription of the paternal MLH1 allele was shown using a heterozygous expressed polymorphism within the 3′ untranslated region. The hypermethylated allele was maternally transmitted, however, the mother and siblings who inherited the same maternal homologue were unmethylated at MLH1, suggesting the epimutation arose as a de novo event. Conclusions: Germline MLH1 epimutations are functionally equivalent to an inactivating mutation and produce a clinical phenotype that resembles HNPCC. Inheritance of epimutations is weak, so family history is not a useful guide for screening. Germline epimutations should be suspected in younger individuals without a family history who present with a microsatellite unstable tumor showing loss of MLH1 expression. Background & Aims: Hereditary nonpolyposis colorectal cancer (HNPCC) is caused by heterozygous germline sequence mutations of DNA mismatch repair genes, most frequently MLH1 or MSH2. A novel molecular mechanism for HNPCC has recently been suggested by the finding of individuals with soma-wide monoallelic hypermethylation of the MLH1 gene promoter. In this study, we determined the frequency and role of germline epimutations of MLH1 in HNPCC. Methods: A cohort of 160 probands from HNPCC families who did not harbor germline sequence mutations in the mismatch repair genes were screened for methylation of the MLH1 and EPM2AIP1 promoters by combined bisulfite and restriction analyses. Allelic expression and family transmission of MLH1 were determined using polymorphisms in intron 4 and the 3′ untranslated region. Results: One of 160 individuals had monoallelic MLH1 hypermethylation in peripheral blood, hair follicles, and buccal mucosa, indicative of a soma-wide alteration. Monoallelic transcription of the paternal MLH1 allele was shown using a heterozygous expressed polymorphism within the 3′ untranslated region. The hypermethylated allele was maternally transmitted, however, the mother and siblings who inherited the same maternal homologue were unmethylated at MLH1, suggesting the epimutation arose as a de novo event. Conclusions: Germline MLH1 epimutations are functionally equivalent to an inactivating mutation and produce a clinical phenotype that resembles HNPCC. Inheritance of epimutations is weak, so family history is not a useful guide for screening. Germline epimutations should be suspected in younger individuals without a family history who present with a microsatellite unstable tumor showing loss of MLH1 expression. The majority of colorectal cancers characterized by microsatellite instability (MSI) occur sporadically in older women and are caused by impaired mismatch repair activity.1Malkhosyan S.R. Yamamoto H. Piao Z. Perucho M. Late onset and high incidence of colon cancer of the mutator phenotype with hypermethylated hMLH1 gene in women.Gastroenterology. 2000; 119: 598Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar, 2Miyakura Y. Sugano K. Konishi F. Ichikawa A. Maekawa M. Shitoh K. Igarashi S. Kotake K. Koyama Y. Nagai H. Extensive methylation of hMLH1 promoter region predominates in proximal colon cancer with microsatellite instability.Gastroenterology. 2001; 121: 1300-1309Abstract Full Text Full Text PDF PubMed Scopus (169) Google Scholar, 3Hawkins N. Norrie M. Cheong K. Mokany E. Ku S.L. Meagher A. O'Connor T. Ward R. CpG island methylation in sporadic colorectal cancers and its relationship to microsatellite instability.Gastroenterology. 2002; 122: 1376-1387Abstract Full Text Full Text PDF PubMed Scopus (324) Google Scholar They usually arise following somatic hypermethylation of both alleles of the MLH1 gene promoter, which causes transcriptional silencing.4Cunningham J.M. Christensen E.R. Tester D.J. Kim C.Y. Roche P.C. Burgart L.J. Thibodeau S.N. Hypermethylation of the hMLH1 promoter in colon cancer with microsatellite instability.Cancer Res. 1998; 58: 3455-3460PubMed Google Scholar, 5Herman J.G. Umar A. Polyak K. Graff J.R. Ahuja N. Issa J.P. Markowitz S. Willson J.K. Hamilton S.R. Kinzler K.W. Kane M.F. Kolodner R.D. Vogelstein B. Kunkel T.A. Baylin S.B. Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma.Proc Natl Acad Sci U S A. 1998; 95: 6870-6875Crossref PubMed Scopus (1728) Google Scholar Hereditary nonpolyposis colorectal cancer (HNPCC) is an autosomal dominant cancer predisposition syndrome characterized by early-onset MSI colorectal cancers and an increased risk of additional MSI tumors, most notably of the endometrium, small bowel, and urinary tract.6Green S.E. Bradburn D.M. Varma J.S. Burn J. Hereditary non-polyposis colorectal cancer.Int J Colorectal Dis. 1998; 13: 3-12Crossref PubMed Scopus (21) Google Scholar HNPCC is caused by heterozygous germline mutations of the mismatch repair genes, most commonly MSH2 and MLH1, each accounting for approximately one third of index cases.6Green S.E. Bradburn D.M. Varma J.S. Burn J. Hereditary non-polyposis colorectal cancer.Int J Colorectal Dis. 1998; 13: 3-12Crossref PubMed Scopus (21) Google Scholar Mutations of MSH6,7Plaschke J. Engel C. Kruger S. Holinski-Feder E. Pagenstecher C. Mangold E. Moeslein G. Schulmann K. Gebert J. von Knebel Doeberitz M. Ruschoff J. Loeffler M. Schackert H.K. Lower incidence of colorectal cancer and later age of disease onset in 27 families with pathogenic MSH6 germline mutations compared with families with MLH1 or MSH2 mutations the German Hereditary Nonpolyposis Colorectal Cancer Consortium.J Clin Oncol. 2004; 22: 4486-4494Crossref PubMed Scopus (203) Google Scholar PMS2,8Nakagawa H. Lockman J.C. Frankel W.L. Hampel H. Steenblock K. Burgart L.J. Thibodeau S.N. de la Chapelle A. Mismatch repair gene PMS2 disease-causing germline mutations are frequent in patients whose tumors stain negative for PMS2 protein, but paralogous genes obscure mutation detection and interpretation.Cancer Res. 2004; 64: 4721-4727Crossref PubMed Scopus (136) Google Scholar and MLH39Wu Y. Berends M.J. Sijmons R.H. Mensink R.G. Verlind E. Kooi K.A. van der Sluis T. Kempinga C. van der Zee A.G. Hollema H. Buys C.H. Kleibeuker J.H. Hofstra R.M. A role for MLH3 in hereditary nonpolyposis colorectal cancer.Nat Genet. 2001; 29: 137-138Crossref PubMed Scopus (147) Google Scholar have collectively been associated with a small percentage of HNPCC cases. The onset of tumorigenesis is most frequently due to somatic loss of the wild-type allele, in accordance with Knudson's "two-hit" hypothesis.10Hemminki A. Peltomaki P. Mecklin J.P. Jarvinen H. Salovaara R. Nystrom-Lahti M. de la Chapelle A. Aaltonen L.A. Loss of the wild type MLH1 gene is a feature of hereditary nonpolyposis colorectal cancer.Nat Genet. 1994; 8: 405-410Crossref PubMed Scopus (284) Google Scholar, 11Tannergard P. Liu T. Weger A. Nordenskjold M. Lindblom A. Tumorigenesis in colorectal tumors from patients with hereditary non-polyposis colorectal cancer.Hum Genet. 1997; 101: 51-55Crossref PubMed Scopus (69) Google Scholar Last year, our group reported the finding of "germline epimutations" of MLH1, which manifested as soma-wide monoallelic hypermethylation of the gene promoter, in association with a phenotype consistent with HNPCC.12Suter C.M. Martin D.I. Ward R.L. Germline epimutation of MLH1 in individuals with multiple cancers.Nat Genet. 2004; 36: 497-501Crossref PubMed Scopus (396) Google Scholar In that study, we described 2 unrelated individuals who had multiple primary MSI tumors, including colorectal cancers, younger than 50 years of age and endometrial cancer in the female proband. In both individuals, the molecular defect was restricted to a single allele and was present in somatic tissues derived from all 3 embryonic germ cell layers, suggesting the error had most likely been inherited or arisen de novo in the germline. Five additional individuals with early-onset colorectal cancer (plus additional tumors in a proportion of these cases) have been described with hemiallelic hypermethylation of the MLH1 promoter in peripheral blood.13Gazzoli I. Loda M. Garber J. Syngal S. Kolodner R.D. A hereditary nonpolyposis colorectal carcinoma case associated with hypermethylation of the MLH1 gene in normal tissue and loss of heterozygosity of the unmethylated allele in the resulting microsatellite instability-high tumor.Cancer Res. 2002; 62: 3925-3928PubMed Google Scholar, 14Miyakura Y. Sugano K. Akasu T. Yoshida T. Maekawa M. Saitoh S. Sasaki H. Nomizu T. Konishi F. Fujita S. Moriya Y. Nagai H. Extensive but hemiallelic methylation of the hMLH1 promoter region in early-onset sporadic colon cancers with microsatellite instability.Clin Gastroenterol Hepatol. 2004; 2: 147-156Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar While these reports provide evidence for the existence of further cases with germline MLH1 epimutations, methylation analyses were limited to peripheral blood, and so the somatic distribution and probable origin of the molecular defect in these individuals remain undefined. Nevertheless, the identification of a small number of individuals with a clinical classification of HNPCC in association with a confirmed or probable germline MLH1 epimutation suggests this molecular defect may represent a novel etiological mechanism for HNPCC. To study the frequency and role of MLH1 germline epimutations in HNPCC, we screened a cohort of 160 probands who had undergone germline testing for HNPCC and in whom no sequence mutations of the mismatch repair genes had been identified using standard mutation detection methods. From this cohort, we identified one further male individual with early-onset colorectal cancer displaying MSI and no family history who had soma-wide hemiallelic promoter hypermethylation and monoallelic expression of the MLH1 gene. The results suggested that germline MLH1 epimutations were rare in individuals with colorectal cancer with a family history suggestive of HNPCC. A total of 160 suspected HNPCC families in whom previous testing had not identified germline sequence mutations in MSH2 or MLH1 were identified by analyzing the medical records of 8 family cancer clinics in Australia, Canada, and the United Kingdom. With approval of the Human Research and Ethics Committee of St Vincent's Hospital, constitutive DNA from 97 female and 63 male probands (mean age, 51.7 ± 13.1 years; range, 26–80 years) was submitted for inclusion in this study. Eleven of these individuals carried sequence alterations of unknown functional significance in one of the mismatch repair genes. Of the 160 families, 42 (26.3%) fulfilled the Amsterdam I criteria,15Vasen H.F. Mecklin J.P. Khan P.M. Lynch H.T. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC)Dis Colon Rectum. 1991; 34: 424-425Crossref PubMed Scopus (1778) Google Scholar 28 (17.6%) met Amsterdam II criteria,16Vasen H.F. Watson P. Mecklin J.P. Lynch H.T. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative Group on HNPCC.Gastroenterology. 1999; 116: 1453-1456Abstract Full Text Full Text PDF PubMed Scopus (2146) Google Scholar and 70 (43.8%) met one of the revised Bethesda criteria.17Umar A. Boland C.R. Terdiman J.P. Syngal S. de la Chapelle A. Ruschoff J. Fishel R. Lindor N.M. Burgart L.J. Hamelin R. Hamilton S.R. Hiatt R.A. Jass J. Lindblom A. Lynch H.T. Peltomaki P. Ramsey S.D. Rodriguez-Bigas M.A. Vasen H.F. Hawk E.T. Barrett J.C. Freedman A.N. Srivastava S. Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite instability.J Natl Cancer Inst. 2004; 96: 261-268Crossref PubMed Scopus (2609) Google Scholar An additional 20 individuals (12.5%) did not meet clinical criteria for HNPCC, but clinical suspicion had warranted germline testing (7 individuals with an MSI colorectal cancer tumor, 5 individuals with an HNPCC non–colorectal-related tumor but no family history, and 8 individuals with colorectal cancer aged 60–65 years but reason for germline testing not stated). Most individuals presented with colorectal cancer (130 cases; 80%); however, other primary lesions included endometrial cancer (n = 6), urothelial tumors (n = 4), ovarian cancer (n = 4), upper gastrointestinal cancer (n = 3), breast cancer (n = 1), liposarcoma (n = 1), and renal cell carcinoma (n = 2). Although the latter 4 individuals did not present with an HNPCC-related tumor, they subsequently developed multiple tumors, including colorectal cancer. The histological type of the primary tumor could not be verified in 10 individuals. A total of 36 individuals in this study had more than 1 primary tumor, and 10 patients had 3 or more. The results of tumor microsatellite testing and immunostaining were only available on a small subset of cancers because such testing was not routine in most centers. Of the 56 tumors tested, 49 (87%) showed MSI, including 37 colorectal adenocarcinomas, 1 endometrial cancer, 1 renal cell cancer, and 1 gastric tumor. The 7 microsatellite stable cancers were colorectal tumors with normal expression of the mismatch repair proteins. Forty-seven of the 49 MSI tumors were examined immunohistochemically for expression of mismatch repair enzymes. Unequivocal loss of MSH2 protein was found in 5 tumors, MLH1 expression was lost in 33 tumors, and 9 tumors demonstrated normal expression of both enzymes. A population of 300 Australian Red Cross blood donors (139 females and 161 males; mean age, 46.5 ± 13.5 years; range, 17–71 years) was also enrolled in this study. DNA from peripheral blood lymphocytes (PBLs) was extracted using the standard phenol-chloroform method, and DNA from buccal mucosa and hair follicles was extracted using the BuccalAmp DNA Extraction Kit (Epicentre Biotechnologies, Madison, WI). The DNA from individuals with HNPCC was screened using combined bisulfite and restriction analyses (COBRA) for the A and C regions of the MLH1 promoter and the overlapping EPM2AIP1 promoter on the opposite strand (Figure 1A). The normal control population was screened for methylation by COBRA of the MLH1-C locus only. Genomic DNA (1 μg) was treated with sodium bisulfite as previously described.18Arnaud P. Monk D. Hitchins M. Gordon E. Dean W. Beechey C.V. Peters J. Craigen W. Preece M. Stanier P. Moore G.E. Kelsey G. Conserved methylation imprints in the human and mouse GRB10 genes with divergent allelic expression suggests differential reading of the same mark.Hum Mol Genet. 2003; 12: 1005-1019Crossref PubMed Scopus (128) Google Scholar Polymerase chain reaction (PCR) amplification was performed using primers designed to coamplify methylated and unmethylated templates equally from sodium bisulfite–converted DNA. Primer sequences (listed as a forward and a reverse, respectively, here and elsewhere), PCR product sizes, and annealing temperatures were as follows: MLH1-A region: GTTTTGAYGTAGAYGTTTTATTAGGGT and TTAACCCTACTCTTATAACCTCCC, 281 base pairs, 52°C; MLH1-C region: TATTTTAGTAGAGGTATATAAGTTYGG and CCTTCAACCAATCACCTCAATACC, 322 base pairs, 51°C; EPM2AIP: AAATTTTTTAATTTTGTGGGTTG and ACTTCCATCTTACTTCTTTTAAAC, 342 base pairs, 47°C. DNA samples that were methylated by COBRA were cloned into the pGEMTeasy vector (Promega Corp, Madison, WI), and individual clones were fluorescently sequenced on an ABI 3100 (Applied Biosystems, Foster City, CA) using M13 vector primers. Any sequences indicating partial nonconversion during bisulfite treatment were excluded from analysis. The methylation status at the promoter regions of the genes APC, BLM1, BRCA1, BRCA2, CDH1, CDH13, CDKN2A, HIC1, RASSF1A, and TMEFF2, which are frequently associated with somatic hypermethylation in cancer, was also analyzed by COBRA. Primers specific to sodium bisulfite–converted DNA for these loci are available on request. To screen for exonic deletions or duplications of the MLH1 and MSH2 genes, the multiplex ligation-dependent probe assay (MLPA) was performed using the SALSA MLPA Kit P003 MLH1/MSH2 as per the manufacturer's instructions (MRC-Holland, Amsterdam, The Netherlands). Briefly, 50 ng genomic DNA was denatured at 95°C for 5 minutes and then hybridized with the P003 probe mix in MLPA buffer at 60°C for 16 hours. Probe ligation was performed at 54°C for 15 minutes, and the ligated products were then amplified using 6-carboxyfluorescein–labeled universal primers. The products were separated by capillary electrophoresis on the ABI 3700 system and analyzed using Genescan and Genotyper software (Applied Biosystems). A peak representing each exon of MLH1 and MSH2 was detected, and peak heights were normalized by division with the mean of the control peak heights. The criterion for detection of an exonic deletion or duplication was a peak height deviation >35% from that of normal controls. To screen for sequence mutations within the MLH1 promoter, genomic DNA was PCR amplified using primers GAAAACTAGAGCCTCGTCGACTT and TAGATGCTCAACGGAAGTGCCTT. PCR amplification was performed using FastStart GC-RICH buffer and FastStart Taq polymerase (Roche, New South Wales, Australia) with annealing at 55°C. Purified PCR products were fluorescently sequenced using the original primers. To screen for heterozygous polymorphisms within the 3′ untranslated region of MLH1, the final exon of MLH1 was PCR amplified and sequenced from PBL DNA using intronic primers AGGCTTATGACATCTAATGTGT and GCATCTGAACTGACACAATATA. Total RNA was extracted from PBLs following red cell lysis, using the standard guanidinium isothiocyanate method.19Chomczynski P. Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction.Anal Biochem. 1987; 162: 156-159Crossref PubMed Scopus (66797) Google Scholar First-strand complementary DNAs were generated by standard reverse transcription (RT) using poly dT oligonucleotides. RT-PCR was performed using forward primer TGGGACGAAGAAAAGGAATG from the penultimate exon of MLH1 and reverse primer GTCTTAAGTGCTACCAACAC from the final exon. The RT-PCR product was purified and sequenced in both directions, across the polymorphic site, using nested primers GTTTTGAAAGCCTCAGTAAAG and GCTACCAACACTTATGTTGG to determine allelic expression of MLH1. An informative G/A single nucleotide polymorphism (SNP) within intron 4 (National Center for Biotechnology Information SNP database reference no. rs4647224) of MLH1 was used in family transmission studies and to screen for somatic loss of the normal allele in tumor DNA. Alleles were distinguished by TaiI restriction digestion of the amplicons generated from a PCR reaction using primers GATGGCATTAAGGAGAAATTATG and CTGGGTAGAAAGATATCCAACAGG. Amplicons containing the G were sensitive and those with the A were resistant to restriction digestion. For loss of heterozygosity analysis,20Cawkwell L. Bell S.M. Lewis F.A. Dixon M.F. Taylor G.R. Quirke P. Rapid detection of allele loss in colorectal tumours using microsatellites and fluorescent DNA technology.Br J Cancer. 1993; 67: 1262-1267Crossref PubMed Scopus (191) Google Scholar microdissected DNA from paraffin-embedded tumor and normal colonic mucosa was amplified as previously described, except that the reverse primer was carboxyfluorescein labeled and the products were electrophoresed on an ABI 377 PRISM DNA Sequencer (Applied Biosystems). Loss of heterozygosity was scored when there was a reduction (ratio of alleles 2.0) or total loss of one allele in tumor DNA relative to normal DNA.20Cawkwell L. Bell S.M. Lewis F.A. Dixon M.F. Taylor G.R. Quirke P. Rapid detection of allele loss in colorectal tumours using microsatellites and fluorescent DNA technology.Br J Cancer. 1993; 67: 1262-1267Crossref PubMed Scopus (191) Google Scholar Screening by COBRA identified one individual (ST) from the cohort of 160 cases (0.6%) with a restriction pattern suggestive of monoallelic methylation at both the A and C regions of the MLH1 promoter and the overlapping EPM2AIP1 promoter on the antisense strand (Figure 1B, proband ST). No evidence of methylation was detected in the subject at 10 additional cancer-related loci, including APC, BLM1, BRCA1, BRCA2, CDH1, CDH13, CDKN2A, HIC1, RASSF1A, and TMEFF2, indicating the methylation defect was specific to the MLH1 locus. Individual ST did not have a family history of cancer (Figure 1B) but fulfilled the Bethesda criteria for HNPCC because he developed colorectal cancer at the age of 39 years. Furthermore, this tumor was right sided, microsatellite unstable (marked alteration in repeat length at 5 loci: BAT 25, BAT 26, BAT 40, D2S123, and D17S250 with stability at only D5S346) and failed to express MLH1 (Figure 2). COBRA and sodium bisulfite sequence analyses of the MLH1 A and C promoter regions and EPM2AIP1 in PBLs, buccal mucosa, and hair follicles from the subject revealed approximately equal proportions of hypermethylated and unmethylated alleles in each tissue (Figure 3), consistent with soma-wide monoallelic methylation. Some mosaicism in the methylation patterns was observed between alleles, with some alleles showing partial methylation. This was particularly notable for all 3 regions in the hair follicles and the EPM2AIP1 promoter in all tissues (Figure 3), suggesting some mitotic instability in this molecular defect.Figure 3Bisulfite sequencing of the MLH1 and EPM2AIP1 promoters in somatic tissues from individual ST. HF, hair follicles; BM, buccal mucosa. Each horizontal row represents a single allele and each circle a CpG dinucleotide. Methylated CpG sites are black, and unmethylated CpGs are white. The subject showed approximately equal proportions of methylated and unmethylated alleles in each cell type derived from all 3 embryonic germ cell layers, suggestive of soma-wide monoallelic epimutation.View Large Image Figure ViewerDownload Hi-res image Download (PPT) COBRA analyses of immediate family members showed normal, unmethylated patterns in PBL DNA from the parents and siblings, including a sister and monozygotic twin brothers, suggesting de novo occurrence of the defect in patient ST (Figure 1B). COBRA of the C region of the MLH1 promoter in a control population of 300 normal individuals did not identify any aberrant methylation of this locus, indicating this molecular defect is disease associated and does not represent an epigenetic polymorphism. To determine whether hypermethylation was associated with allele-specific transcriptional inactivation of MLH1 in individual ST, we screened for a heterozygous polymorphism within the 3′ untranslated region in genomic DNA, which could be used to trace the allelic pattern of expression. Proband ST was heterozygous for a CTT deletion polymorphism at nucleotide position 2363–2365 with respect to the messenger RNA sequence (GenBank accession no. NM00249). RT-PCR analysis of PBLs across the transcribed polymorphic site showed monoallelic expression of the 2363ΔCTT allele (Figure 4). Transcriptional silencing of the wild-type allele provides evidence for promoter inactivation mediated by hypermethylation of this particular allele. Analysis of parental DNA showed that the father of subject ST was heterozygous for the 2363ΔCTT polymorphism, whereas his mother was homozygous for the wild-type allele. The functional 2363ΔCTT allele was thus paternally derived, indicating the affected allele was maternally inherited. All of the subject's siblings were homozygous for the wild-type allele, so this polymorphism was noninformative with respect to which maternal homologue they had inherited. The subject was also heterozygous for a G/A SNP within intron 4 (National Center for Biotechnology Information SNP database reference no. rs4647224), and this SNP proved informative with regard to familial inheritance patterns (Figure 5A). The paternal DNA was homozygous G/G, the maternal DNA heterozygous G/A, and each of the siblings heterozygous G/A. Given obligate maternal inheritance of the A allele, these data indicate that the subject and each of his unaffected siblings inherited the same maternal MLH1 homologue. Analysis of tumor DNA from individual ST provided further evidence that the epimutant allele was maternally derived. Comparative fluorescent PCR analysis of the intron 4 SNP in tumor DNA showed somatic loss of the G allele in the tumor (Figure 5B). Because the G allele was paternally transmitted (Figure 5A), this indicated somatic loss of the functional paternal allele in the tumor, consistent with maternal inheritance of the epigenetically inactivated allele. Screens for germline mutations in the coding region of the MLH1 and MSH2 in subject ST had previously been negative. MLPA for intragenic deletions or duplications of MLH1 and MSH2 were also negative. To determine whether the subject harbored a de novo sequence mutation within the promoter, the region from +29 to −466 (with +1 representing the transcription start site) encompassing both the A and C regions of the promoter was screened by direct sequence analysis, but no sequence changes were identified. The absence of any detectable mutation in the subject rules against the likelihood that a sequence mutation underlies the epimutation in this individual. This study has identified a single individual with symmetric methylation of both sense and antisense strands of the MLH1 locus in cells derived from all 3 embryonic germ cell layers. These findings suggest this individual carried a germline MLH1 epimutation and that this event accounts for very few cases of HNPCC. Despite the rarity of epimutations in this population, our finding of monoallelic expression of the paternal MLH1 allele shows that germline epimutations can act as functional hemizygous mutations in affected individuals, mimicking the predisposition to tumor development conferred by heterozygous sequence mutations in HNPCC. The range of tumors documented in cases with confirmed or probable MLH1 epimutations is the same as that in HNPCC, with predominantly early-onset MSI colorectal and endometrial cancer.12Suter C.M. Martin D.I. Ward R.L. Germline epimutation of MLH1 in individuals with multiple cancers.Nat Genet. 2004; 36: 497-501Crossref PubMed Scopus (396) Google Scholar, 13Gazzoli I. Loda M. Garber J. Syngal S. Kolodner R.D. A hereditary nonpolyposis colorectal carcinoma case associated with hypermethylation of the MLH1 gene in normal tissue and loss of heterozygosity of the unmethylated allele in the resulting microsatellite instability-high tumor.Cancer Res. 2002; 62: 3925-3928PubMed Google Scholar, 14Miyakura Y. Sugano K. Akasu T. Yoshida T. Maekawa M. Saitoh S. Sasaki H. Nomizu T. Konishi F. Fujita S. Moriya Y. Nagai H. Extensive but hemiallelic methylation of the hMLH1 promoter region in early-onset sporadic colon cancers with microsatellite instability.Clin Gastroenterol Hepatol. 2004; 2: 147-156Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar Given these findings, there is no reason why the surveillance and treatment of individuals with germline epimutations should differ from those with sequence mutations in the mismatch repair genes. In clear distinction to the stable transmission of sequence mutations in the mismatch repair genes, the evidence to date suggests that germline epimutations in MLH1 are not inherited. Like patient ST, the 5 patients reported by Gazzoli et al and Miyakura et al had either no family history of cancer or no affected first-degree relatives.13Gazzoli I. Loda M. Garber J. Syngal S. Kolodner R.D. A hereditary nonpolyposis colorectal carcinoma case associated with hypermethylation of the MLH1 gene in normal tissue and loss of heterozygosity of the unmethylated allele in the resulting microsatellite instability-high tumor.Cancer Res. 2002; 62: 3925-3928PubMed Google Scholar, 14Miyakura Y. Sugano K. Akasu T. Yoshida T. Maekawa M. Saitoh S. Sasaki H. Nomizu T. Konishi F. Fujita S. Moriya Y. Nagai H. Extensive but hemiallelic methylation of the hMLH1 promoter region in early-onset sporadic colon cancers with microsatellite instability.Clin Gastroenterol Hepatol. 2004; 2: 147-156Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar Our previous report of 2 patients with epimutations met clinical criteria for HNPCC because 1 parent was affected by cancer. Unfortunately, because parental DNA was unavailable, no further studies on the potential inheritance of the defects were possible.12Suter C.M. Martin D.I. Ward R.L. Germline epimutation of MLH1 in individuals with multiple cancers.Nat Genet. 2004; 36: 497-501Crossref PubMed Scopus (396) Google Scholar The current study thus provided the first opportunity to screen for epimutations in the parents and siblings of an affected individual. COBRA and SNP analyses of DNA from the parents and siblings showed that none of the family members harbored the epimutation, even though the unaffected siblings shared the same maternal homologue with the proband. These findings suggested that the epimutation arose de novo in the germline (during oogenesis or in the zygote). We found no evidence for the existence of an underlying sequence alteration on the shared maternal allele, which might be responsible for the establishment of the epimutation, despite a thorough screen of the coding and promoter regions. While epimutations affect all somatic cells and are functionally equivalent to an inactivating mutation, the available evidence suggests they may only rarely be retained during gametogenesis and passed to the next generation. The low levels of methylated MLH1 alleles in the spermatozoa of our previously reported male proband were indicative of significant reversal of the epimutation during gametogenesis in this particular individual.12Suter C.M. Martin D.I. Ward R.L. Germline epimutation of MLH1 in individuals with multiple cancers.Nat Genet. 2004; 36: 497-501Crossref PubMed Scopus (396) Google Scholar Because the MLH1 epimutation is unstable in the germline, the affected alleles can be passed on without the epimutation. This implies an epimutation can be inherited, but in a pattern that in no way resembles the orderly transmission of genetic alleles. The potential risk of transmission of the defect in individual ST was not assessed because of a prior vasectomy. Regardless of the mechanism by which epimutations occur, the apparent low risk of transmission that seems to characterize this phenomenon has important clinical implications for the relatives of affected individuals. While risk assessment will need to be performed on a case-by-case basis, there appears to be little justification for the subject's siblings and parents to continue intensive surveillance. In clinical practice, genetic testing is usually restricted to individuals with a strong family history. While there is likely to be some variability in genetic testing criteria in this multi-institutional study, there is little doubt that the study population was enriched for those with a clear familial pattern of disease. Because inheritance of epimutations is typically weak, patients suspected of having a genetic disorder on the basis of a strong family history may paradoxically be less likely to carry an epimutation. Our work indicates that the individuals who are most likely to carry germline epimutations are those who manifest the clinical phenotype of a disease in the absence of a positive family history. More specifically, germline epimutations in MLH1 should be strongly suspected in younger individuals without a family history of cancer who develop an MSI tumor (especially colorectal or endometrial cancer) that shows loss of MLH1. The authors thank Drs Barbara Leggett, Jeremy Jass, and Joanne Young (Gastroenterology Laboratory, Queensland Institute of Medical Research, Herston, Queensland, Australia) for the provision of clinical samples and information.