Familial Cancer (2011) 10:515–520 DOI 10.1007/s10689-011-9436-z
A putative Lynch syndrome family carrying MSH2 and MSH6 variants of uncertain significance—functional analysis reveals the pathogenic one Jukka Kantelinen • Thomas v. O. Hansen • Minttu Kansikas • Lotte Nylandsted Krogh • Mari K. Korhonen • Saara Ollila • Minna Nystro¨m • Anne-Marie Gerdes • Reetta Kariola
Published online: 24 March 2011 Ó Springer Science+Business Media B.V. 2011
Abstract Inherited pathogenic mutations in the mismatch repair (MMR) genes, MSH2, MLH1, MSH6, and PMS2 predispose to Lynch syndrome (LS). However, the finding of a variant or variants of uncertain significance (VUS) in affected family members complicates the risk assessment. Here, we describe a putative LS family carrying VUS in both MSH2 (c.2768T[A, p.Val923Glu) and MSH6 (c.3563G[A, p.Ser1188Asn). Two colorectal cancer (CRC) patients were studied for mutations and identified as carriers of both variants. In spite of a relatively high mean age of cancer onset (59.5 years) in the family, many CRC patients and the tumor pathological data suggested that the missense variation in MSH2, the more common susceptibility gene in LS, would be the predisposing alteration. However, MSH2 VUS was surprisingly found to be MMR proficient in an in vitro MMR assay and a tolerant alteration in silico. By supplying evidence that instead of MSH2 p.Val923Glu the MSH6 p.Ser1188Asn variant is completely MMR-deficient, the present study confirms the J. Kantelinen M. Kansikas M. K. Korhonen S. Ollila M. Nystro¨m R. Kariola (&) Department of Biosciences, Genetics, University of Helsinki, P.O. Box 56 (Viikinkaari 5), 00014 Helsinki, Finland e-mail:
[email protected] T. v. O. Hansen Department of Clinical Biochemistry, Section of Genomic Medicine, Rigshospitalet and Copenhagen University, Copenhagen, Denmark A.-M. Gerdes Department of Clinical Genetics, Rigshospital and Copenhagen University, Copenhagen, Denmark L. N. Krogh Department of Clinical Genetics, Odense University Hospital, Odense, Denmark
previous findings, and suggests that MSH6 (c.3563G[A, p.Ser1188Asn) is the pathogenic mutation in the family. Moreover, our results strongly support the strategy to functionally assess all identified VUS before predictive gene testing and genetic counseling are offered to a family. Keywords Functional analysis Lynch syndrome MSH2 MSH6 Variants of uncertain significance (VUS) Abbreviations AC Amsterdam criteria CRC Colorectal cancer EPCAM Epithelial cell adhesion molecule HNPCC Hereditary non-polyposis colorectal cancer IHC Immunohistochemical LS Lynch syndrome MMR Mismatch repair MSI Microsatellite instability NE Nuclear extract TE Total extract VUS Variants of uncertain significance WT Wild type
Introduction Lynch syndrome (LS; hereditary non-polyposis colorectal cancer syndrome, HNPCC; MIM # 120435) is associated with the malfunction of a highly conserved postreplicative DNA mismatch repair (MMR) mechanism. Approximately 97% of all reported LS germline mutations are found in three different MMR genes, MLH1 (NM_000249.2), MSH2 (NM_000251.1) and MSH6 (NM_000179.2) (http://www. insight-group.org/mutations/; http://www.lovd.nl) [1].
123
516
J. Kantelinen et al.
Patients carrying MLH1 or MSH2 mutations often display typical clinical and tumor pathological features of the syndrome, such as the frequent occurrence of colorectal cancer (CRC), early age of onset and high microsatellite instability (MSI) in tumor cells. Contrary to such classical LS families, mutations in the MSH6 gene are often associated with later age of cancer onset and low or no MSI in tumors [2–5]. Regardless of the gene in question, truncating MMR gene mutations are generally considered to be disease causing, whereas amino acid substitutions that does not necessarily destroy protein expression complicate their interpretation. Such alterations are referred to as variants of uncertain significance (VUS). Today’s focus in international collaborative research for gastrointestinal hereditary tumors (http://www.insight-group.org/) is to functionally characterize all VUS found in suspected LS families and thus distinguish non-pathogenic variants from pathogenic ones facilitating gene testing and genetic counseling in these families [6]. This study is related to our previous functional assessment of MSH2, c.2768T[A, p.Val923Glu [7, 8], which was found in a family resembling the LS phenotype. The family members were found to carry VUS in both MSH2 and MSH6 (c.3563G[A, p.Ser1188Asn), but high prevalence of CRCs and the tumor pathological data in the family suggested that the MSH2 VUS would be the predisposing variation. However, our results from an in vitro MMR analysis showed that the MSH2 p.Val923Glu variant corrected mismatches as efficiently as the wild type MSH2 protein and in silico analysis supported the nonpathogenic nature of this variant [8]. Here, we continue the functional assessment of the two VUS found in the family by analyzing the effects of the variants both separately and combined. The present study
underscores the importance of doing a complete mutation analysis of all LS predisposing genes and to functionally assess all VUS found in a putative Lynch syndrome family.
Fig. 1 Pedigree of the putative Lynch syndrome family. Plus (?), the family member verified as carrying two VUS (MSH2 c.2768T[A, p.Val923Glu and MSH6 c.3563G[A, p.Ser1188Asn). Minus (-), the family member tested negative for both VUS. Number inside a
diamond, number of children of unspecified sex. Tumor types (BCC, basocellular carcinoma; CRA, colorectal adenoma; CRC, colorectal cancer; EC, endometrial cancer; LU, lung cancer) and ages at onset are marked
123
Materials and methods Genetic and clinical data of the putative Lynch syndrome family The pedigree of the studied family is shown in Fig. 1. Of seven affected family members, two CRC patients (III:2 and III:7) were screened for MLH1, MSH2 and MSH6 mutations and both were found to be carriers of two germline variations, MSH2 c.2768T[A, p.Val923Glu and MSH6 c.3563G[A, p.Ser1188Asn. Furthermore, all available family members in generation IV were screened for these two VUS. Since the carriers (III:2 and III:7) represent different branches of the family tree, their parents (II:I and II:4) must be carriers of both VUS as well (Fig. 1). The MSH2 VUS has been previously described [9] and functionally tested [7, 8], whereas the MSH6 VUS has not been analyzed before. Mutation, MSI and immunohistochemical analyses In mutation analysis, MLH1, MSH2 and MSH6 sequences were amplified using intronic primer pairs flanking each exon. PCR products were pre-screened by denaturing high performance liquid chromatography (dHPLC) using the Wawe system (Transgenomic, Glasgow, UK) and sequenced using an ABI3730 DNA analyzer (Applied Biosystems, Foster City, CA, USA). Sequence variations were verified from new blood samples. Moreover, large genomic
Functional testing of MSH2 and MSH6 VUS
517
rearrangements including EPCAM (epithelial cell adhesion molecule) deletions were excluded by multiplex ligationdependent probe amplification (MLPA) as recommended by the manufacturer (MRC-Holland, Amsterdam, the Netherlands). The Bethesda panel (BAT-25, BAT-26, D2S123, D5S346 and D17S250) was applied for the MSI analysis [7]. The immunohistochemical (IHC) stainings of MLH1, MSH2, MSH6 and PMS2 proteins was performed as described previously [7].
results in the proteolytic degradation of its counterparts MSH3 and MSH6 [16, 17]. Nuclear proteins were extracted as previously described [10]. The MMR capability of the protein variants were analyzed by using two different heteroduplexes, a G-T mismatch (50 GT) and a single nucleotide insertion/deletion loop (50 IDL1) as substrates [10].
Production of recombinant protein variants
The present study was done to continue the functional assessment of the two VUS, MSH2 (c.2768T[A, p.Val923Glu) and MSH6 (c.3563G[A, p.Ser1188Asn), found in a putative LS family. The MSH2-V923E variant was previously found to be MMR proficient in the in vitro MMR assay and tolerant based on a 0.89 SIFT score (an amino acid substitution is predicted to affect protein function when \0.05) [7]. Later, the MSH2-V923E variant was, however, suggested to have slightly reduced mismatch binding and release capacity compared to the wild type MSH2 protein [8], but these deficiencies were not statistically significant. Here, our results confirm the previous findings and suggest that MSH6 VUS is the pathogenic mutation in the family. Production of VUS with its wild type partner (MSH2V923E/MSH6-WT and MSH2-WT/MSH6-S1188N) and with the mutated partner (MSH2-V923E/MSH6-S1188N) was successful in Sf9 insect cells (Fig. 2a). Mismatch repair capacity of the variants both separately and combined were tested in the in vitro MMR assay. The MSH6S1188N variant was totally deficient, whereas MSH2V923E repaired mismatches as WT protein (Fig. 2b). Because MSH2-WT/MSH6-S1188N was not able to correct a G-T mismatch or a single nucleotide insertion/ deletion loop at all (Fig. 2b), we could not see a potential compound effect created by the presence of the two variations (MSH2-V923E/MSH6-S1188N). The C-terminal part of the MSH6 polypeptide consists of an evolutionarily highly conserved ABC-ATPase domain between amino acids 1,076–1,360. In this sequence, amino acids 1,180–1,186 form a disordered loop structure, which may play a crucial role in ATP binding [18]. Two cancer associated mutations have been reported in the vicinity of that region, the truncating mutation MSH6 c.3558_3565delTGAAAGTA, p.Gly1186fsX1190, which was detected in a CRC patient at the age of 27 [19], and the missense variation, MSH6 c.3577G[A, p.Glu1193Lys, which was identified in two endometrial cancer patients with late age of onset (59 and 60 years) and poor family history [11]. Remarkably, both the truncating and the missense variation turned out to be MMR deficient [11, 19], indicating that the region between amino acids
Preparation of the mutated MSH2 and MSH6 cDNAs, expression of the mutated and the wild type MSH2 and MSH6 proteins by using the Bac-to-Bac Baculovirus expression system (Invitrogen, Carlsbad, CA, USA) and analyses of the protein variants in the in vitro MMR assay followed the previously described protocols [7, 10, 11]. The missense variations in MSH2 (c.2768T[A, NM_000251.1) and MSH6 (c.3563G[A, NM_000179.2) were constructed with a PCR based site-directed mutagenesis kit according to manufacturer’s instructions (QuickChance Site-directed mutagenesis, Stratagene), substituting in the produced proteins valine to glutamine in codon 923 (p.Val923Glu) and serine to asparagine in codon 1188 (p.Ser1188Asn), respectively. The mutated MSH2 and MSH6 cDNAs were introduced into a pFastBac1 vector (Invitrogen) and sequenced by ABIPrism 3100 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The primer sequences and PCR parameters are available upon request. For protein production, Sf9 insect cells were co-infected with MSH2 and MSH6 recombinant baculoviruses in the following combinations: MSH2-WT (wild type) with MSH6-WT, MSH2-V923E (p.Val923Glu) with MSH6-WT, MSH2-WT with MSH6-S1188N (p.Ser1188Asn) and MSH2-V923E with MSH6-S1188N. Co-infection was used since both in vivo study in mice [12] and in vitro studies in human cells [13, 14] have shown that the MSH6 protein is unstable without its cognate partner MSH2. In vitro MMR assay The repair efficiency of the MSH6 and MSH2 protein variants were studied by complementing MSH6- and MSH2-deficient nuclear extract (NE) of LoVo cells (American Type Culture Collection, Manassas, VA, USA) with the total protein extract (TE) of Sf9 cells including over expressed human MSH6 and MSH2 proteins [15]. MMR proficient HeLa cell line (used as a positive control) and MMR deficient LoVo cell line were cultured according to providers’ instructions. In LoVo cells, the MSH2 gene is inactivated causing the lack of MSH2 protein, which also
Results and discussion
123
518
J. Kantelinen et al.
Fig. 2 a Western blot analysis showing total protein extracts (TE) of Sf9 cells co-infected with baculovirus constructs expressing MSH6 wild type protein (MSH6-WT) with MSH2-WT and with MSH2V923E variant, MSH2-WT with MSH6-S1188N, and MSH2-V923E variant with MSH6-S1188N variant. The proteins were blotted onto nitrocellulose membranes (Hypond, PVDF, Amersham Pharmacia biotech, Uppsala, Sweden) and visualized with anti-MSH2 (Calbiochem, San Diego, CA, USA, MSH2- Ab1, NA-26, 0.2 lg/ml) and anti-MSH6 (BD Transduction Laboratories, clone 44, 0.02 lg/ml) monoclonal antibodies. b In vitro MMR efficiency. The MMR activity of the recombinant protein variants with their cognate partners were studied in LoVo (L) NE. The fragment lengths on the left indicate the migration of the unrepaired linearized plasmid DNA (3,193 bp) and of the two fragments (1,833 and 1,360 bp) produced following the correction of the 50 IDL1 (insertion/deletion loop) or
50 GT mispair, which makes the DNA susceptible to cleavage with the restriction endonuclease BglII [10]. The numbers below the panel represent fractions (R%) of repaired DNA. The repair percentages were analyzed with Gene Tools 3.08 (SynGene, Cambridge, England) and calculated as an average of 3 independent experiments. HeLa is MMR-proficient NE used as a positive control. Mock contains heteroduplex DNA with no added NE or recombinant protein. LoVo NE (MSH2-/-) is MMR-deficient. LoVo NE with recombinant MSH2 protein is also MMR deficient demonstrating degradation of native MSH6 protein in LoVo cells. Repair efficiencies and standard deviation were 50 IDL1: MSH2-WT/MSH6-WT (14% ± 2%), MSH2V923E/MSH6-WT (13% ± 2%), MSH2-WT/MSH6-S1188N (0%), MSH2-V923E/MSH6-S1188N (0%) and 50 GT: MSH2-WT/MSH6WT (22% ± 6%), MSH2-V923E/MSH6-WT (18% ± 2%), MSH2WT/MSH6-S1188N (0%), MSH2-V923E/MSH6-S1188N (0%)
1186–1193 is extremely important in repair function. By supplying evidence, that MSH6 (c.3563G[A, p.Ser1188Asn) caused complete loss of protein function in the MMR reaction, our results reinforce the impression that this region in MSH6 is particularly important. Results of mutation analysis as well as tumor analyses including IHC staining of MLH1, MSH2, MSH6, and PMS2 proteins, and MSI results are collected in Table 1. Although, the IHC analysis shows problems in MSH2 expression in two family members (III:2 and III:7) carrying MSH2 (c.2768T[A, p.Val923Glu) and MSH6 (c.3563G[A, p.Ser1188Asn) variations, our results demonstrate that the expression problems of both MSH2 and MSH6 as well as the high MSI in one of the mutation carriers (III:7) is rather associated with MSH6 than MSH2 deficiency. Since, immunohistochemical analysis of MMR protein expression cannot distinguish between EPCAM deletion carriers and MSH2 mutation carriers [20], MLPA analysis was performed in the carrier III:7 to exclude a germline EPCAM deletion. Previously, functionally studied MSH6 VUS have been associated with low cancer susceptibility [11]. Here, the clinical features such as the late mean age of cancer onset (59.5 years) (Fig. 1), not completely lost but reduced or heterogenous expression of MSH6 in the tumor of the
mutation carrier III:2 (Table 1) and loss of MSH6 expression in the carrier III:7 support the MSH6 predisposition. Although, the tumors of both mutation carriers showed expression deficiencies of MSH6 and MSH2 (Table 1), neither MSH6-S1188N nor MSH2-V923E showed expression problems in our in vitro Sf9 expression system (Fig. 2a). The explanation for this discrepancy is most probably in the in vitro expression system, in which the protein is abundantly expressed under strong virus promoter. Thus, the successful MMR protein production in Sf9 insect cells does not necessarily mean stability in the tumors and it is not possible totally exclude the MSH2 VUS contribution to the LS in these patients. However, its nonpathogenicity was supported by both the in vitro MMR and in silico analyses, recently found to be a reliable assay combination to verify pathogenicity/nonpathogenicity of a MMR VUS [21]. Finally, in the fourth generation, there are several family members who do not carry either VUS but still have colorectal adenomas (CRA) at young age. The fact that also other mutations including large genomic rearrangements in MLH1, MSH2, and MSH6 were excluded and that all the MMR proteins were expressed in the carrier0 s IV:4 CRA (Table 1) suggest that predisposition to CRA in these individuals is associated with something else than MMR deficiency.
123
Functional testing of MSH2 and MSH6 VUS
519
Table 1 Results of mutation, immunohistochemical (IHC) and MSI analyses from the family Family member
Mutation testinga
IHCb
MSH2
MSH6
MLH1
III:1
ND
ND
?
III:2
?
?
?
III:3
–
–
ND
III:5
ND
ND
III:7
?
?
III:8
ND
IV:1 IV:3
– ?
IV:4 IV:5
MSI status MSH6
PMS2
?
±
ND
-
±
ND
ND
ND
ND
ND
ND
±
?
±
?
Stable
?
±
–
ND
Highc
ND
ND
ND
ND
ND
ND
– ?
ND ND
ND ND
ND ND
ND ND
ND ND
–
–
?
?
?
?
ND
–
–
ND
ND
ND
ND
ND
IV:8
–
–
ND
ND
ND
ND
ND
IV:9
–
–
ND
ND
ND
ND
ND
IV:10
?
?
ND
ND
ND
ND
ND
IV:11
–
–
ND
ND
ND
ND
ND
a b c
MSH2
Stable
? mutation present (MSH2 (c.2768T[A, p.Val923Glu) and MSH6 (c.3563G[A, p.Ser1188Asn), - mutation not present ? normal expression, ± reduced or heterogenous expression, - loss of expression Bat25, Bat26 and D17S250 loci were unstable (D2S123 and D5S346 were stable)
ND, not determined
Our study clearly demonstrates that when LS is suspected, all MMR susceptibility genes should be included in the mutation analyses and all identified VUS should be functionally analyzed. As seen in this family, when multiple VUS are found in the genes MSH2 and MSH6, both located on the chromosome 2, VUS may be inherited in the same chromosome thus showing similar patterns of segregation. Especially, in these cases, functional analyses are required to ensure the LS predisposing variation. Acknowledgments We thank Mikko Frilander for providing the HeLa cell line and Merja Salmitie for the preparation of 50 IDL1. Friedrik Wikman at Molecular Diagnostics at Skejby Hospital and Henrik Okkels at Clinical Biochemistry at Aalborg Hospital are acknowledged for their contributions to the molecular genetics analyses. Claus Fenger at Pathologic Department at Odense University Hospital is acknowledged for performing the immunohistochemical analyses. This study was supported by grants: Sigrid Juselius Foundation; European Research Council (2008-AdG-232635); Finnish Cancer Organisations, The Biocentrum Helsinki Organisation; The Research Foundation of the University of Helsinki and Kuopio Naturalists’ Society.
References 1. Woods MO, Williams P, Careen A et al (2007) A new variant database for mismatch repair genes associated with Lynch syndrome. Hum Mutat 28:669–673 2. Hendriks YM, Wagner A, Morreau H et al (2004) Cancer risk in hereditary nonpolyposis colorectal cancer due to MSH6 mutations: impact on counseling and surveillance. Gastroenterology 127:17–25
3. Berends MJ, Wu Y, Sijmons RH et al (2002) Molecular and clinical characteristics of MSH6 variants: an analysis of 25 index carriers of a germline variant. Am J Hum Genet 70:26–37 4. Wagner A, Hendriks Y, Meijers-Heijboer EJ et al (2001) Atypical HNPCC owing to MSH6 germline mutations: analysis of a large Dutch pedigree. J Med Genet 38:318–322 5. Wu Y, Berends MJ, Mensink RG et al (1999) Association of hereditary nonpolyposis colorectal cancer-related tumors displaying low microsatellite instability with MSH6 germline mutations. Am J Hum Genet 65:1291–1298 6. Couch FJ, Rasmussen LJ, Hofstra R et al (2008) Assessment of functional effects of unclassified genetic variants. Hum Mutat 29:1314–1326 7. Ollila S, Sarantaus L, Kariola R et al (2006) Pathogenicity of MSH2 missense mutations is typically associated with impaired repair capability of the mutated protein. Gastroenterology 131:1408–1417 8. Ollila S, Dermadi Bebek D, Jiricny J et al (2008) Mechanisms of pathogenicity in human MSH2 missense mutants. Hum Mutat 29:1355–1363 9. Bisgaard ML, Jager AC, Myrhoj T et al (2002) Hereditary nonpolyposis colorectal cancer (HNPCC): phenotype-genotype correlation between patients with and without identified mutation. Hum Mutat 20:20–27 10. Kantelinen J, Kansikas M, Korhonen MK et al (2010) MutSbeta exceeds MutSalpha in dinucleotide loop repair. Br J Cancer 102:1068–1073 11. Kariola R, Hampel H, Frankel WL et al (2004) MSH6 missense mutations are often associated with no or low cancer susceptibility. Br J Cancer 91:1287–1292 12. de Wind N, Dekker M, Claij N et al (1999) HNPCC-like cancer predisposition in mice through simultaneous loss of Msh3 and Msh6 mismatch-repair protein functions. Nat Genet 23:359–362 13. Chang DK, Ricciardiello L, Goel A et al (2000) Steady-state regulation of the human DNA mismatch repair system. J Biol Chem 275:18424–18431
123
520 14. Marra G, Iaccarino I, Lettieri T et al (1998) Mismatch repair deficiency associated with overexpression of the MSH3 gene. Proc Natl Acad Sci USA 95:8568–8573 15. Kariola R, Otway R, Lonnqvist KE et al (2003) Two mismatch repair gene mutations found in a colon cancer patient–which one is pathogenic? Hum Genet 112:105–109 16. Cannavo E, Marra G, Sabates-Bellver J et al (2005) Expression of the MutL homologue hMLH3 in human cells and its role in DNA mismatch repair. Cancer Res 65:10759–10766 17. Drummond JT, Genschel J, Wolf E et al (1997) DHFR/MSH3 amplification in methotrexate-resistant cells alters the hMutSalpha/hMutSbeta ratio and reduces the efficiency of base-base mismatch repair. Proc Natl Acad Sci USA 94:10144–10149
123
J. Kantelinen et al. 18. Warren JJ, Pohlhaus TJ, Changela A et al (2007) Structure of the human MutSalpha DNA lesion recognition complex. Mol Cell 26:579–592 19. Pinto C, Veiga I, Pinheiro M et al (2006) MSH6 germline mutations in early-onset colorectal cancer patients without family history of the disease. Br J Cancer 95:752–756 20. Kloor M, Voigt AY, Schackert HK et al (2011) Analysis of EPCAM protein expression in diagnostics of Lynch syndrome. J Clin Oncol 29:223–227 21. Kansikas M, Kariola R, Nystro¨m M (2011) Verification of the three-step model in assessing the pathogenicity of mismatch repair gene variants. Hum Mutat 32:107–115
