Analysis of BRCA and HNPCC mutations

Analysis of HNPCC and BRCA mutations:screening of unclassified variants (UVs)for splicing defectsMario TosiInserm U614, Faculté de Médecine et de Pharmacie de RouenInstitut Hospitalo-Universitaire de Recherche Biomédicale de RouenLondon, 3rd March 2008Eurasnet workshop : RNA splicing and genetic diagnosisWorshop organizers : Diana Baralle, Nick Cross

HNPCC (Hereditary Non Polyposis Colorectal Cancer)OR LYNCH SYNDROME2 - 5 % of colorectal cancerscrccrccrc < 50 years- Colon rectum +++- Endometrium +++- Small bowel- Urinary tract- Stomach- Ovary- Biliary tractAutosomal dominant genetic predispositionMLH1 (3p21)MSH2 (2p22-p21)MSH6 (2p16)PMS2 (7p22)MMR (mismatch repair) genesMLH1MSH2MSH6PMS2

MMR GENE ALTERATIONS IN THE LYNCH SYNDROMEData for families recruited with stringent criteria (Amsterdam)-Pathogenic point mutationsMLH1 25 %MSH2 25 %MSH6, PMS2 5 %- MSH2 gene rearrangements 10 %- MLH1 rearrangements < 4 %QMPSFMLPA< 70 %+ Many variants of unknown significance (UVs)Missense variants, silent (at the level of translation), intronic variantsDi Fiore et al., J Med Genet 2004, 41:18-20

SPLICING REGULATORY ELEMENTSISE Intronic Splicing EnhancerISS Intronic Splicing SilencerESE Exonic Splicing EnhancerESS Exonic Splicing SilencerISEISSESE ESSGU A YRYYRY AG+ -GUAG

LEVELS OF ANALYSISPATIENT RNAin most cases not availableIn silico predictionsnot sufficient for diagnosisnot always accurateFunctional splicing assaystime-consuming but very usefulSplicing reporter minigene assays, reviewed in:Baralle and Baralle J. Med. Genet. 2005;42;737-748Cooper TA. Methods 2005; 37:331-340

SPLICING ASSAYPatient blood samples (PaxGene) for RNAoften not availablePatient DNABamHIExon + mutationMluI150 bp150 bpCMVEx2Ex3PolyAdenylationpCDNA3.1(-) / SERPING1/C1NH

SPLICING ASSAYWild-type exon1) CloningMutant exon2) Transient transfectioninto HeLa cellsPuromycin10ug/mL 5h303) RT-PCR after 24hEx2Ex XEx3Ex2Ex34) Agarose gel – SYSTEMATIC sequencing

EXAMPLES OF VARIANTS THAT MODIFYOR GENERATE EXON BOUNDARIESExon + mutationBamHI MluI150 bp150 bpCMVP1 P2PolyApCAS vector(pcDNA3.1(-) / SERPING1/C1NH)pCAS vectorEx 10 wt MLH1c.884+4A>GEx 15 wt MLH1c.1731G>Ac.1731+5G>AEx 11 wt MSH2c.1759G>CEx 13 wt MSH2c.2006-6T>CEx 15 wt MSH2c.2634+5G>TMLH1 MSH2All RT-PCR products are sequencedTournier et al. (2008). Human Mutation, in press

MLH1, MSH2 VARIANTS ANALYSED (FEB. 2008)74 Missense mutations 16 Silent mutations 34 Intronic mutations5 single codon deletionsother than conserved AG/GT1 del/ins130 MLH1 (73), MSH2 (57) variants analysed33 / 130 (25%) affect splicing27 Alteration of splice sites, orGeneration of new splice sites, orActivation of crytpic splice site6 Mutations of putative splicingregulatory elements

MUTATIONS OF PUTATIVE SPLICING REGULATORY ELEMENTSMLH1Exon 10, 3rth exon position4th exon positionIn silico predictionsVariantESEfinderRescueESEPESXASSAc.793C>T----c.794G>ASC35, SRp40 abolished--SC35pCAS splicing assayc.793C>T : partial exon 10 skipping +++c.794G>A : partial exon 10 skipping ++VectorEx 10 WT MLH1c.793C>Tc.794G>AIdentification of a new ESE element ?

MUTATIONS OF PUTATIVE SPLICING REGULATORY ELEMENTSEffect of four MLH1 or MSH2 exonic variantsand one MLH1 intronic UV in 3 different cell typesEx 9 WT MLH1c.790+10A>GEx 10 WT MLH1c.793C>Tc.794G>AEx 11 WT MLH1c.988_990delEx 5 WT MSH2c.815C>TVectorHeLaDLD1COS7Same variants in Hela cells usig pSPL3 : similar results

SC35-3xc.840C>TVALIDATION OF THE ESE-DEPENDENT SPLICING ASSAYPutative ESE sequence (~30 bp)EcoRIBamHICMVEx1Ex2PolyApcDNA-DupMiddle exon…cacccttagGCTGAATTC ( ) GGATCCGGCAGgttggtatc…SMN1Ex 7MCADEx 5SF2/ASF-3xEmptyBR2 Int 11wtwtc.351A>Cc.351A>Cc.362C>Tc.362C>TInteraction ESS-ESENielsen et al. (Andresen BS) Am J Hum Genet 80:416 (2007)

IDENTIFYING ESE ELEMENTS OF MLH1 AND MSH2Sequences to be tested (~30bp) inserted into a central exon,which is strictly dependent on presence of a ESE~ 30 pbVecteurExon 10 MLH1 WTc.793C>Tc.794G>Ac.988_990delc.815C>TExon 11 MLH1 WTExon 5 MSH2 WTpSPDUPHeLaWildtype sequence : exon inclusion3 new ESE elements found in this workDemonstration that 11 previously published mutations of MLH1 ou MSH2are within ESE elementsTournier et al. (2008). Human Mutation, in press

COMPARISONS WITH PATIENT RNA DATA130 MLH1, MSH2 variants analysed25 % (31 / 130) variants affect splicing37 / 130 variants tested on patient RNA35 Concordant results 2 Discordant results(but NMD not prevented)

APPLICATIONS TO BRCA1 AND BRCA2French Network Breast cancer geneticsWe compared patient RNA analyses and the minigene assayRouen + Caen (Dr. A. Hardouin)20 variants BRCA1, BRCA2 (17 families)Analysed in parallel onPatient RNAEx vivo minigene assay6 Variants affect splicing. Perfect agreementbetween in vivo and ex vivo data

EXAMPLE 1BRCA2 c.316+5G>C, IVS3Patient RNAMinigene assay5’UTR-ex5550 bp301 bp ∆ ex 3VE3+5G>CVex2-ex5, monoallelic514 bp265 bpC1 C2 P∆ ex 3wtMC1 C2 PNo normal mRNA expressed from mutant allele, pathogenicC. Bonnet, S. Krieger et al., revised manuscript submitted

EXAMPLE 2BRCA2 c.9501+3A>T (IVS25+3A>T)Patient RNART ex 23-ex 27Minigene assay+3A>TE25634 bp389 bp∆ ex 25C1 C2 PwtMMinigene assay : ~ 50% normal mRNA from mutant allelePartial splicing defect, still unclassified

DISCREPANCIES BETWEEN IN SILICO, IN VIVO AND EX VIVOBRCA1 c.670+8C>T (IVS10)Several in silico predictions :New Dss at +6, of samestrength as natural oneRT-PCR on patient RNA :Only natural Dss used+ Automated Splice Site AnalysesUsing sofware « Alamut » from Interactive BiosoftwarepCAS Minigene splicing assay :Only natural Dss usedUnlikely to be pathogenicBut this variant promotesinclusion, both in vivo and inthe minigene assay. Does itaffect an ISS?

SUMMARYLarge numbers of UVs screened with minigene assay130 in MLH1, MSH2 and 20 in BRCA1/2Good concordance with patient RNA data35/37 for MLH1, MSH2 and 20/20 for BRCA1/225% of UVs classified as splicing defectsMainly splice site alterations, activations of cryptic sitesor generation of new sitesFive mutations found in new exonic regulatory elementsDefinition of 12 ESE-containing short regions8 in MLH1, 4 in MSH2

ADVANTAGES OF MINIGENE ASSAYVery useful for screening UVs, because in most casespatient RNA is not availableWhen patient RNA is available, minigene assay facilitatesthe interpretation, because monoallelicAlthough exons are tested in heterologous context,good concordance with patient RNA dataAllows evaluation of in silico predictionsAlgorithms predicting 5’ and 3’ splice sites : rather accurateAlgorithms predicting ESE, ESS : neither sensitive nor specific

INTERPRETATION OFTEN REQUIRESAdding genetic evidence:segregation, co-occurrence of déléterious changesAdding information obtained from tumours(e.g. for HNPCC: IHC, microsatellite instability…)Interrogating databasesFrequent meetings of network members

Inserm U614Faculté de Médecine, RouenService de Génétique, CHU RouenIsabelle TournierMyriam VezainCéline BonnetAudrey KillianAlexandra MartinsStéphanie Baert-DesurmontMario TosiThierry FrébourgConsultations d’OncogénétiqueDominique Stoppa-LyonnetInstitut Curie, ParisCatherine DugastCentre Eugène Marquis, RennesRosetteLidereau,Centre René Huguenin, FNCLCC,St-CloudCancéropôle Nord-OuestAgnès Hardouin, Sophie KriegerCentre François Baclesse, CaenRéseau laboratoires HNPCCSylviane OlschwangInstitut Paoli-Calmettes, MarseilleMarie-Pierre BuisineCHU, LilleQing WangCentre Léon Bérard, LyonJean-Marc ReyCHU Hôpital Arnaud de Villeneuve,MontpellierOlivier CaronCHU Hôpital Civil, StasbourgChristophe PhilippeCHU, NancyStéphane BézieauCHU, NantesIGMM, UMR 5535, MontpellierJohann SoretJamal Tazi