`
`' 1994 Nature Publishing Group http://www.nature.com/naturegenetics
`
`article
`
`Confirmation of BRCAI by
`analysis of germline mutations
`linked to breast and ovarian
`cancer in ten families
`
`Lori S. Friedman, Elizabeth A. Ostermeyer, Csilla I. Szabo, Patrick Dowd, Eric D. Lynch,
`Sarah E. Rowell & Mary-Claire King
`
`We provide genetic evidence supporting the identity of the candidate gene for BRCA1
`through the characterization of germline mutations in 63 breast cancer patients and 10
`ovarian cancer patients in ten families with cancer linked to chromosome 1 7q21. Nine
`different mutations were detected by screening BRCA1 DNA and ANA by single-strand
`conformation polymorphism analysis and direct sequencing. Seven mutations lead to
`protein truncations at sites throughout the gene. One missense mutation (which occurred
`independently in two families) leads to loss of a cystelne in the zinc binding domain. An intmnic
`single basepair substitution destroys an acceptor site and activates a cryptic splice site,
`leading to a 59 basepair insertion and chain termination. The four families with both breast
`and ovarian cancer had chain termination mutations in the N-terminal half of the protein.
`
`Department of
`Molecular d’ Cell
`Biology and School
`of Public Health,
`University of
`California,
`Berkeky, California
`94720, USA
`
`In October 1994, a candidate gene for BRCA1, which is
`responsible for inherited predisposition to breast and
`ovarian cancers in some families, was isolated bypositional
`cloning’. In this report, we confirm that BRCA1 is this
`predisposing gene by analysing germline mutations in the
`families that originally defined the linked phenotype’, as
`well as in other families with breast cancer, and often
`ovarian cancer, linked to chromosome 17q2 1.
`For this purpose, the BRCAI gene was screened using
`single strand conformation polymorphism (SSCP) analysis
`in 20 BRCAI-linlced families, using both genomic DNA
`and cDNA prepared from lysnphoblast RNA. Criteria for
`distinguishing cancer-predisposing mutations from
`polymorphisms in BRCAI were (i) cosegregation of the
`variant with breast cancer, and with ovarian cancer, if it
`appeared; (ii) absence of the variant in control
`chromosomes; and (iii) amino acid substitution in, or
`truncation of, the BRCAI protein encoded by the variant
`sequence.
`
`Detection of mutations in genomic and cDNA
`Primers used to screen genomic DNA and cDNA for
`variation in BRCAI by SSCP analysis are indicated in
`Table 1. All primer pairs were used to screen all 20 families
`by SSCP analysis. Primers flanking exons 2, 3,4 (an Mu
`sequence not found inmost cDNA clones), 5,8,9, 10 and
`12-24 are from the ftp file atmorgan.med.utah.edu’. New
`primers were designed to flank exon 1, which is 100
`basepairs (bp) 5’ of the start site and for which surrounding
`genomic sequence was reported in Genbank (Li 82093) as
`an STS on 17q21 (ref. 3). in order to screen splice junctions
`of exons 6 and 7, intron 6 was sequenced and primers
`defined to yield 200-250 bp amplified products flanking
`Nature Genetics volume 8 december 1994 (cid:9)
`
`exons 6 and 7 independently. Primers were also designed
`to amplify exon 11 in 250-300 bp fragments, using the
`BRCAI cDNA sequence published as Genbank accession
`number U14680 (Table 1). New SSCP primers were
`designed to amplify from cDNA, and are designated
`cl-cl! in Table!.
`Each variant detected by SSCP analysis was sequenced
`using several templates: the variant band from the SSCP
`gel, genomic DNA and/or cDNA from the proband, and
`genomic DNA from family members. The 18 mutations
`(M) or polymorphisms (P) detected in genomic and/or
`cDNA are also indicated in Table 1 next to the primers
`with which each was observed.
`
`Analysis of BRCAI mutations
`Two types of BRCAI variants fulfilling the criteria for
`cancer-predisposing mutations are illustrated by families
`1 and 84. In family 1, an SSCP pattern reflecting a small
`deletion cosegregates with breast and ovarian cancer and
`the BRCA1-linked haplotype (Fig. 1). The haplotype
`analysis predicted that persons #1 and #8 did not inherit the
`linked chromosome, and the mutation analysis confirmed
`their BRCAI sequence was wild type. Thus their breast
`cancers at ages 71 and 45 are sporadic. The variant sequence
`shown has a 2 bp deletion at nucleotides (nts) 2800-2801
`(codon 894), creating a stop at codon 901.
`In family 84, an SSCP pattern representing a single
`nucleotide substitution cosegregates with breast cancer
`and the BRCAI-linked hapiotype (Fig. 2). The variant
`genomic sequence has a thymine to guanine mutation at
`nt 300 of sequence U14680, resulting in a Cys6lGly
`substitution. This mutation removes the penultimate
`cysteine of the putative C3HC4 zinc-binding motif(cid:151)the
`399
`
`GeneDX 1004, pg. 1
`
`
`
`z (cid:9)
`
`co
`
`6 3
`
`’(cid:149) G Zt
`2. 0.
`m a
`. C
`0 (cid:9) (cid:149) 3
`
`.0 OUR
`. ma
`
`(cid:151)
`
`7V 0
`
`.
`
`; (cid:9)
`
`Z
`
`a
`9 (cid:9)
`
`v
`
`-.
`o
`a (cid:9)
`
`00
`3
`
`c.
`8)
`
`o Zp8 (cid:151)
`a> R Lo.
`(9 r..
`
`o 690
`
`Sc
`8) a a
`E 5
`03 gr
`’<:t. Z
`
`0
`2
`C
`3P)3
`
`0
`
`CL -*
`
`Spa
`S
`a
`0)
`
`U,
`
`69
`c c
`- - (cid:151)
`o a:t,
`’9
`-"-lb
`
`r
`
`4.
`
`C a 4.
`
`0
`0-
`1) a
`
`"0
`’0
`
`Primer oair
`exonl
`C
`exon2
`exon3
`c2
`exan4
`exonS
`exon6
`exon7
`c3
`exon8
`c4
`exon9
`axon 10
`exonllAi
`exonilA
`exonllB
`exon1lB2
`exonllC
`exonllD
`exonllD2
`exonhiE
`exon1lE2
`exonlil
`exonhiG
`exonhlH
`e,conlhH2
`exonill
`exonl 112
`exonhli
`exonlll(
`exonllK2
`axon llL
`exonllM
`exonliN
`exonllO
`exonllP
`c5-wonhlPi
`CS
`excn12
`exon13
`c6
`c7
`exon14
`axon 15
`c8
`exon16
`C9
`CIO
`exonl7
`exonh8
`exonl9
`exon20
`cli
`exon2i
`exon22
`exon23
`exon24
`
`Tm (cid:9) Size (bpi 06
`U 14680 bo
`genomic 1-100
`315
`55
`cDNA
`28-280
`55
`253
`101-199
`250
`genomic
`60
`genomic 200-253
`60
`300
`212-530
`55
`319
`cONA
`genomic Alu
`60
`200
`genomic 254-331
`200
`60
`genomic 332-420
`200
`55
`genomic 421-560
`55
`250
`258
`cONA
`390-647
`55
`220
`genomic 561-665
`60
`cDNA
`581-878
`55
`298
`genofnic 666-712
`200
`60
`220
`genomic 713-788
`60
`genomic 789-1090
`55
`301
`858-1166
`309
`both
`55
`1068-1363
`300
`both
`55
`both
`1209-1417
`55
`209
`both
`1299-1593
`55
`295
`1505-1758
`55
`254
`both
`1584-1855
`both
`55
`272
`1716-1985
`269
`both
`55
`1718-2127
`410
`both
`55
`1947-2219
`273
`both
`55
`2142-2460
`55
`319
`both
`2198-2509
`both
`55
`312
`2335-2620
`286
`both
`55
`280
`both
`2485-2764
`55
`2598-2930
`333
`both
`55
`both
`2716-3003
`288
`55
`2897-3201
`55
`305
`both
`both
`2998-3293
`55
`296
`3138-3438
`270
`both
`55
`both
`330-3662
`55
`270
`3552-3804
`both
`55
`253
`3683-3791
`both
`55
`289
`3857-4170
`314
`both
`55
`genomic 4105-4215
`200
`55
`4105-4416
`cDNA
`312
`55
`genomic 4216-4302
`220
`60
`ges’iomic 4303-4476
`280
`60
`287
`cDNA
`4323-4609
`55
`CDNA
`4445-4734
`55
`290
`genotnic 4477-4603
`60
`250
`genomic 4604-4794
`250
`60
`4645-4910
`cDNA
`55
`266
`genomic 4795-5105
`60
`375
`4837-5143
`307
`cDNA
`55
`290
`eDNA
`5092-5381
`55
`genomic 5106-5193
`350
`60
`genomic 5194-5273
`350
`60
`genonuc 5274-5310
`60
`220
`genomic 5311-5396
`220
`60
`cDNA
`5331-5648
`55
`318
`genomic 5397-5451
`60
`275
`genomic 5452-5526
`60
`275
`250
`genomic 5527-5586
`60
`genomic 5587-5711
`60
`275
`
`Forward orhner
`5’- TM CCC ITO OTT TCC GIG -3’
`5 1 - MC MG CTG TOG GOT TIC -3’
`51- GM OTT GTC AlT TTA TM ACC ITT -3’
`5- ICC TGA CAC MC MA CAT hA -3’
`5’- CM GGA ACC TOT CTC CM AM GTG -3’
`5’- GTC MA GM AlA GM TOT GAO C -3’
`5- CTC TTA AGO GCA GIl GIG AG -3’
`5’- CTT ATI hA GTG 1CC VIA MA GO -3’
`5’- CAC MC AM GAG CAT MA TAG GO -3’
`5’- TOT OCT ITT CM CIT GAG ACA 06-3’
`5’- TOT TAG CIG ACT OAT OAT 001 -3’
`5’- CCA ACT CTC TM CCI TOO AM TOT G -3’
`5’- CCA CAG TAO ATG CTC AOl AM IA -3’
`5’- TOG TCA OCT TIC TOT MT CO -3’
`5’- GGA ATT AM IGA 4MG ACT ATG MC -3’
`5’- MC MC ACT GAG MG COT GCA 6-3’
`5’- CM CAT MC MA TOO OCT OGA AC -3’
`5’- CCI MA OAT ACT GM OAT OTT CCI TGG -3’
`5’- GOT TCT OAT GAC TCA CAT OAT 606 -3’
`5’- GM MC CIA TOG GM GM OGC AM -3’
`5’- GAG CCACAOATAATA CAA GAG CGTC-3’
`5 1 - ATC AGO GM CIA ACC AM CGG AG -3’
`5’- TCA 600 MC TM CCA AAC OGA 0-3’
`5’- AGO CTO AGO AGO MO TCT TCT ACC -3’
`5’- OCA ACT GGA 0CC AMMO ACT MC -3’
`5’- CM COA TM ITT CCC MA OCT 0-3’
`5’-AMTGICIAATAATOCTGAM ACC CC-3’
`5’- GCA CTC TAG GGA AGO CM MA CAG -3’
`5 1 - GM GOC ITT MG TAT CCA fIG 06-3’
`5’- 0CC MT CAT ITO CTC COT ITT C -3’
`5’- bC AGO CIT ICC TOT 60110-3’
`5’- OCA AGO AM CTG GM ICA TEA CIC -3’
`5- TCA ATG TCACCT GM MA GM AIG 0-3’
`5’- ITO MT OCT ATG CII MA TEA 6060 -3’
`5’- OTT TOT TCT GAG ACA OCT OAT GACC -3’
`5 1 - GAG TCC TAG CCC TTT CAC CCA TAO -3’
`51 - COT TGO TAC CGA GTG TCT GIC TM 0-3’
`5’- AM 6CC AGO GAO TTG GTC TGA 6-3’
`5’- AM 0CC MG GAG ITO GTC TGA 0-3’
`5’- GTC CTG CCA ATG AGA MA AA -3’
`5-MT OGA AAG CIT CTC AAA GTA ’3’
`5’- CAT MC CTG AlA AAG CTC CAG CM 0-3’
`5’- OCO AM ICC MA ACA MO CAC ATC -3’
`5’- CIA MC TGA ATT ATC ACT ATC A -3’
`5’-TGGCTGCCCACOAAC TAT G-3’
`5’- MA TOO ACA GIT OCT CTG GGA 0-3’
`5’-MTTCT TMCACACACCAGMC-3’
`5’- AIG MC CTG MT CTO ATC CIT CTO -3’
`5’- CCC CM AM MI ITA TOO ICG TO -3’
`5’- GIG TM MC GIG CM OAT 10-3’
`5’- GOC TCT TTA OCT ICI TAG GAG -3’
`5’- CTG TCA TIC TIC CTG TGC IC -3’
`5’- ATA TOA COT GTC TGC TCC AC -3’
`5’- GGA OAT GIG GTC MT GGA MA AAC -3’
`5’- MO CTC TIC CIT ITT GM MT C -3’
`5 1 - ICC CAT IGA GAG GTC ITO CI -3’
`5-CM MC AM MC CTO TCT C -3’
`5’- ATG MT IGA CM TAA TCT CTO C -3’
`
`Revww orimer
`5’-TCA CAA OGCCTTACGCCTC-3’
`5’- TOG ITO MA ACT TIC MC ATG -3’
`5’- TOT CTT TIC TIC CCI AOl ATO 1 -3’
`5’- 116 OAT ITT CGT TCT CM TTA -3’
`5’- AM TCT ITT GOC MG OTT ICT 0-3’
`5’- CCC GTC TCT ACA GM MC AC -3’
`5’- TIC CIA CTO TOO ITO CIT CC -3’
`5’- ITT CAT GGA CAG CAC TIC MT G ’3’
`S’- ICGGCTTCACTCIGT AGAAG-3’
`5’- CGT CIT ITO AGO 116 TAT CCG CTG -3’
`5’- ATC CAG CM TTA TEA hA MT AG -3’
`5’- TOC AGO CU dC ACT GOT OTT C -3’
`5’- TM GM MT MC MC ITC AlA GA -3’
`5’- GTA TCT MC CAC TCT CIT CII CAG -3’
`5’- CTT CCA 0CC CAT CIG ’ETA IGI 10 -3’
`5’-CTC ACA CACOOGATCAGCATTC-3’
`5’- MG ICC MT MA TCA OCT ACT TIG 0-3’
`5’- 0CC ACT MO TCT All TTC TCT GM GM CC ’3’
`5’- TCT GIG OCT CAC TM CM ATG CTC -3’
`5- ICA ICA CII GAG CAT ICT OCT CC -3’
`5’- OCA OAT ICT Ill TOG ACT OAT TCT ATI 660-3’
`5’- CGC ATG MT AIG CCI GOT MA AG
`5’- CCA bA GIl GTA GOT TIC hOC TG -3’
`5’- CAO CTC TOG GM AOl AIC OCT 0-3’
`5’- CCT GAO TGC CAT MT CAG TAG CAG 0-3’
`5’- TCT OTT ITT 6CC TIC CCI MA GIG -3’
`5’- CCC MT OGA TAG TTA AM CCT TCT G -3’
`5’- CAT ICC TCT TCT OCA ITT CCT GO -3’
`5’- CII ATC ITT CTG MC AM CAC MG -3’
`5’- CGT TGC CTC TGA ACT GAO ATO ATAG -3’
`5’- 6CC TM ITO TGC ICA CTG TAG ITO 0 -3’
`5’- MI ACT GGA 6CC CM TIC AlT ACT AC -3’
`5’- CM OAT OCT TAG MT TM TIC CM 0-3’
`5’- GAd OCT TTT OCT AM MC MC AC -3’
`5’- AOl OfF GGA MC AGO GM OCT C 4
`5’-GTOATGTTCCTGAGATGCCTTTG-3’
`5’- MC CCG lid CTC ITT CIT CAT C -3
`5’- GTG CIC CCA MA GCA TM A -3’
`5’- TGA TOG AM GOT MC IGI TAG AM 0-3’
`5’- TOT CM CM ACC TM GM TOT -3’
`5’- ATO ITO GAG CIA GOT CCI TAG -3’
`5’- OAT GAG CII ICC ACT CCT GOT IC -3’
`5’- OCT GTE OCT CCT CCA CAT CM C -3’
`5’- GIG TAT MA TGC dIG TAT GCA .3’
`5’- MC CAO MT ATC ITT ATO TAG GA -3’
`5’- GOT TGA MA TOG TAT OTT 0CC MC -3’
`5- AM ACT CIT ICC MA ATO ITO 1 -3’
`5’- GIG ATO TGG TGT ITT CTO OCA MC 4
`5’- ITC TCT IOC TCG CII TOG MC -3’
`5’-TCGCCT CAT GIGGTIITA-3’
`5’- GM MC AU ITC CCA GCA IC -3’
`5’- CAT TOT TM GOA AM TOO TGC -3’
`5’- 600 MT CCA AAITM ACA GC-3’
`5" IGC TAG ACT GIC CM CAC CCA CTC -3’
`5’- GTAOM AM TAG MT MC CTC T -3’
`5’- GAG MG ACT ICT GAG OCT AC -3’
`5’- ACT OTOCTACTCAAG CAC CA-3’
`5’- GTA 0CC MC MA GIA GM OOA -3’
`
`Variant (by)
`
`14(300). 14(332)
`
`14(300)
`14(332)
`
`P(intron)
`
`PC 1186)
`
`P(2201); P(2430); M(241 5)
`P(2430); 14(2415)
`P(2430); M(241 5)
`P(2430)
`14(2800); 14(2863)
`14(2800); 14(2863)
`
`P(3232)
`P(3232); 14(3238)
`
`P(3667); 14(3726)
`
`14(4184)
`14(4184)
`P(intron)
`
`P(4956)
`P(4956)
`
`P(intron)
`
`14(5677)
`
`S
`
`'
`
`(0
`to
`
`z
`0)
`C
`CD
`-D
`C
`C’
`(4)
`
`(0
`C)
`0
`C
`’C
`
`GeneDX 1004, pg. 2
`
`(cid:9)
`(cid:9)
`
`
`lee YM6 ' 1994 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`Part of Family I
`
`6,71 6,31 7977 6.45 0v48 6r36 (cid:9)
`
`53 (cid:9)
`
`55
`
`’TI 6(cid:243)
`
`Pr 79 IJu1
`T FftiE UEU i I
`
`article
`
`Fig. 1 Framestiift mutation in BRCAI
`linked to breast and ovarian cancer in
`family 1. Symbols under each relative
`Indicate chromosome 17q21
`haplotypes linked (filled rectangles) or
`not linked (open rectangles) to
`BRCAI. Cancers are indicated as
`follows: Br, breast cancer; Ov, ovarian
`cancer; Pr, prostate cancer. Ages at
`diagnosis are also indicated. V
`indicates variant sequence and wt
`indicates wildtype sequence for
`BPCAI, based on SSCP analysis of
`15 relatives, shown below the
`pedigree. The mutant sequence,
`shown on the right, has a 2 bp
`deletion (AA) at nts 2800-2801
`(arrow), leading to a stop at codon
`901. The wild type sequence is
`TCCTrAMGMCAAAGTCCAAAGTC-
`AC1TTTGAATGTGAACA, with the
`mutated sequence underlined.
`
`GATC (cid:9)
`
`appeared twice, in families 84 and 4.
`This mutation probably arose
`independently in the two families,
`because chromosome 17q2 1 marker
`alleles defining the BRCA1 haplotypes
`of the two families differ. The ancestry
`of family 8415 Polish and of family 4
`German; insofar as the families are
`aware, they are not related. The only
`other missense mutation in our
`families leads to the substitution of
`asparagine for serine atcodon 1040 in
`all patients with breast cancer linked
`to BRCA1 in family 14. The functional significance of this
`mutation is currently unknown and might represent either
`a disease-predisposing allele or a rare polymorphism.
`
`Wt (cid:9)
`4 (cid:9)
`
`V
`57
`
`_a (cid:9)
`
`13
`
`t_IJW
` 6r29 49 6r45
`We
`
`T F [ T T (cid:9)
`
`14 4j (cid:9)
`
`6r25 (cid:9)
`
`I T
`
`15 16
`
`6.27 11
`
`I 2 3 4 (cid:9)
`
`5 6 7 $ 9 10 11 (cid:9)
`
`12 13 14 (cid:9)
`
`15
`
`-. -
`
`only recognizable domain so far in the BRCA1 protein.
`Thisvariantdoes notappearin l20 control chromosomes,
`suggesting that it is not simply a rare polymorphism.
`In all, we identified mutations
`cosegregating with breast and ovarian
`cancer and linked to BRCAJ in ten (cid:9)
`families (Table 2). None of these
`mutations has been previously (cid:9)
`observed (D. Goldgar, personal
`communication)’. The ten families
`have nine different mutations: the
`same thymineto guanine substitution
`at nt 300 (the Cys6lGly mutation)
`
`Family84
`
`1
`
`Wt
`
`37 (cid:9)
`
`6r34 (cid:9)
`
`33 (cid:9)
`
`Or 29
`
`EDlEflhl
`
`Wt (cid:9)
`
`V (cid:9)
`
`Wt (cid:9)
`
`V (cid:9)
`
`Fig. 2 Missense mutation in the zinc-binding domain
`of BACA1 linked to breast cancer in Family 84.
`Symbols are as described in Fig. 1; individual #6 had a
`prophylactic mastectomy at age 28, ten years prior to
`entrance into the study. The mutant sequence, shown
`on the right, illustrates the Ito G substitution at nt 300
`(arrow), which leads to replacement of the penultimate
`Cys in the zinc binding domain with a Gly residue. This
`Cys6l Gly mutation also occurred, probably
`Independently, in Family 4. The wild type sequence Is
`GCTGAAACTrCTCAACCAGAAGW-
`GGGccTrcAcAGTGIGTCCmATGTAAGAATGATATA-
`ACCMAAGGTAT, with the mutated sequence
`underlined.
`
`Or SS
`
`Or 55
`
`ILU
`1 (cid:9)
`
`Or 32,37 pm 28
`
`V (cid:9)
`
`Wt
`
`G AIC
`
`a
`
`- a
`-a--
`
`12 345678
`
`.-.
`
`Nature Genetics volume 8 december 1994 (cid:9)
`
`401
`
`GeneDX 1004, pg. 3
`
`(cid:9)
`(cid:9)
`(cid:9)
`
`
`article
`
`U
`
`' 1994 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`Table 2. Gesmllno mutations cosegregating with breast and ovarian cancer In BRCAI-linked families
`
`Family ARCA1 nt (cid:9)
`4
`300
`84
`300
`332
`82
`2415
`3
`1
`2800
`102
`2863
`14
`3238
`74
`3726
`2
`4184
`77
`5677
`
`Exon
`5
`5
`Intron 5
`11
`11
`11
`11
`11
`11
`24
`
`Mutation
`Ito G substitution (rGT-i.GGT)
`T to G substitution (TGT-*GGT)
`Ito 0 substitution-+59 bp insertion
`AG deletion
`AA deletion
`TO deletion
`GtoAsubstitution (AGC-+AAC)
`C to T substitution (CGA.-+TGA)
`ICAA deletion
`A insertion
`
`Amino acid change (cid:9)
`Cys61 Gly
`Cys6l Gly
`75Stop
`Ser766Stop
`901Stop
`Ser915Stop
`SerlO4OAsn
`Argl 2O3Stop
`1 364Stop
`Iyrl853Stop
`
`Predicted effect
`lose zinc-binding motif
`lose zinc-binding motif
`protein truncation
`protein truncation
`protein truncation
`protein truncation
`missense
`protein truncation
`protein truncation
`protein truncation
`
`Table 2b Sites, ages at diagnosis and lateral lty of cancers associated with BRCAI mutations
`
`Family
`
`4
`84
`82
`3
`1
`102
`14
`74
`2
`77
`
`Breast cancer (cid:9)
`llateral) Avg. age at dx (cid:9)
`Cases (cid:9)
`10(1)
`39.7
`5(1)
`41.6
`40.3 (cid:9)
`6(2)
`3 (0)
`37.3 (cid:9)
`16(0)
`37.8 (cid:9)
`45.2 (cid:9)
`5(0)
`51.1
`6(2)
`6 (1)
`45.5
`38.2
`5(1)
`7(1)
`32.9
`
`Ovarian cancer
`Avg. age at dx
`Cases (cid:9)
`
`Other cancers with with BRCAI mutations
`
`2 (cid:9)
`4 (cid:9)
`2 (cid:9)
`3 (cid:9)
`
`52.1
`46.3
`54.5
`51.7
`
`Small Intestine (dx 46)
`
`3 Prostate (dx 57, 77, 79)
`
`KIdney (dx 55); Colon (dx. 40)
`
`However, we have not found the mutation in 120 control
`chromosomes.
`The remaining seven mutations all lead to truncated
`proteins, mostly due to small insertions or deletions of 1
`to 4 bp (Table 2). Four mutations are immediate stops; the
`others are frameshifts leading to stops after 4, 7 and 9
`codons, respectively. In family 3, one AG doublet is
`deleted from an (AG) 3 repeat; the other small length
`changes appeared to be random throughout the sequence.
`In family 82, a thymine to guanine substitution occurs
`s(cid:231) of the exon 6 splice
`within the consensus sequence
`acceptor site (Fig. 3). This substitution leads to the
`preferential activation in intron 5 of a cryptic splice
`acceptor’. This generates a mutant transcript containing
`
`a 59 bp insertion of intronic sequences’ of exon 6, causing
`a stop at codon 75. Both the wild type and mutant
`transcripts are detected in lymphocyte cDNA with primer
`pair c2.
`One variant observed only at the cDNA level could not
`yet be classified as apredisposing mutation. In two patients
`from family 5 with BRCA1-linked breast and ovarian
`cancer, cDNA prepared from lymphoblast RNA lacked
`exon 3. (No RNA from breast tumour is available for this
`family.) Exon 3 splice junction sequences are normal in
`genomic DNA from these patients. The absence of exon 3
`has not been observed in lymphoblast cDNA from controls
`or from other breast cancer families. This case differs from
`the family 82 variant in that in family 5, an entire exon is
`
`Family 82
`
`6 (cid:9)
`
`84
`
`8,47
`
`8,43 (cid:9)
`
`16
`
`0v54
`
`Wt
`
`Wt
`
`Wt
`
`v
`
`Wt
`
`8,42,47 OvSI 41 (cid:9)
`
`45
`
`47 (cid:9)
`
`44 ntiS
`
`47
`
`8r29,33
`
`v (cid:9)
`
`v
`
`Wt (cid:9) Wt
`
`Wt
`
`M
`
`Wt
`
`Y
`
`Exon 5
`
`Exon 6
`
`...NNNNNcTFATTVFAGTG .........ACAAGrrAA1TrCAGGA. (cid:9)
`A
`ACMTrFAXVVFCAGGA (cid:9)
`
`Family 82 mutation
`
`Wild type acceptor
`
`402 (cid:9)
`
`Fig. 3 Aberrant splicing leading to a
`truncated 6RCA1 protein in Family 82.
`a, BRCAI-Iinked patients and
`unaffected relatives of Family 82. Mutant
`) or wildtype (wi) sequence was
`confirmed In living relatives. (Symbols
`are as given inFig. 1). b, Aberrant
`splicing of exons 5 and 6 In Family 82. A
`cryptic splice acceptor site In Intron 5
`(open box) leads to insertion of 59 bp
`(hatched box) into the mutant transcript
`Sequences of exons Sand 6 (filled
`boxes) are normal. The cryptic acceptor
`is activated by a T to substitution in
`the usual acceptor site. The concensus
`splice acceptor sequence is
`(Py)32NCAG/N.
`
`Nature Genetics volume 8 december 1994
`
`GeneDX 1004, pg. 4
`
`(cid:9)
`(cid:9)
`
`
`1;T8 IN ' 1994 Nature Publishing Group http://www.nature.com/naturegenetics
`
`article
`
`BRCA 1 nt
`intron 8
`1186
`2201
`2430
`3232
`3667
`intron 12
`4956
`intron 16 or 17
`
`Nucleotide variant
`T deleted
`A/G
`CIT
`TIC
`A’G
`A/G
`CA deleted
`A/G
`not sequenced
`
`Amino acid variant
`non-coding
`Gln356Arg
`silent
`silent
`Glu1038Gly
`Lysll83Arg
`non-coding
`Sen 61 3Gly
`non-coding
`
`.671.33 (cid:9)
`.671.33 (cid:9)
`.67433 (cid:9)
`.9505 (cid:9)
`
`69/31 (cid:9)
`
`associated with veryearly-onset breast
`Table 3 Neutral polymorphisms in BRCAI
`cancer in family 77. Bilateral breast
`Allele frequencies cancer was associated with mutations
`throughout the sequence, although
`67/33 (cid:9)
`ascertainment of bilaterality is biased
`:82/13 (cid:9)
`in veryhigh-risk families in that some
`women choose bilateral mastectomy
`following their first diagnosis ofbreast
`cancer. In this series, there is no
`association between age at onset and
`locale or type of mutation, but these
`families were selected for early age at
`breast cancer diagnosis and probably
`do not represent BRCA1 patients
`generally. Better evaluation of outcomes associated with
`BRCAJ mutations of different types and at different
`locations in the sequence will be based on identifying
`germline BR CAI mutations in population-based series of
`cases not selected for family history.
`The mutations found, andnot found, in these 20 families
`illustrate the complexities ofscreening for inherited BRCA1
`mutations in families, let alone in the general population 6’7.
`First, mutationswere identified inonly5o% ofthe families
`in this series, although both genomic DNA and RNA were
`screenedby SSCP analysis and all families have convincing
`evidence of BRCA1 linkage’,’. The families for which
`BRCA1 mutations have not yet been identified are as
`likely to carry mutations as those successfully screened.
`The posterior probability of linkage is ~!.95 for families
`with and without definable mutations. The two groups of
`families are similar in number and ages at diagnosis of
`breast and ovarian cancers linked to BRCA1; ovarian
`cancer occurred in four of ten families with mutations
`found and in five of ten families with mutations still
`undetected. Additional mutations may prove detectable
`by other screening techniques such as mismatch cleavage
`analysis, or by Southern or northern hybridization.
`However, it is likely that a large fraction of mutations will
`only be detectable when genomic sequence of the entire
`gene is available. The importance of genomic sequence is
`illustrated by the successful detection of two mutations in
`cDNA: the 59 bp intronic insertion in family 82 reported
`here and the loss of transcript in family 2035 previously
`described’. The intronic mutation leading to aberrant
`splicing in family82 could onlybe identified from genomic
`sequence. In general, mutations in introns and promoter
`regions can only be identified reliably when sequence of
`the entire 80 kb BRCAI gene is known and critical regions
`identified therein. The alternative would be to screen
`breast or ovarian RNA for mutations using multiple
`techniques, an approach not feasible on a wide scale.
`Second, most BRCAI variants are not predisposing
`mutations. Most individuals, including women at high
`risk of breast cancer, are heterozygous at one or more sites
`in the BRCA1 gene. Most of these variants are
`polymorphisms, some of which will be revealed, after
`sequencing, to be noncoding or silent. However, in this
`series, four missense mutations were revealed that altered
`peptide sequence. Because they occurred at similar
`frequencies in cases and controls, we concluded that they
`are polymorphisms. Even by screening cases and controls,
`some rare polymorphisms may be mistakenly identified
`as predisposing mutations. The mutation at nt 3238 in
`family 14 may prove to be of this type, and we would not
`predict risk for an asymptomatic woman based on that
`403
`
`missing and no intronic sequence has been introduced. It
`is possible that absence of exon 3 represents alternative
`splicing characteristic of lyrnphoblasts we immortalized
`for this family, rather than having any relationship to
`cancer predisposition. Alternatively, an inherited mutation
`in genomic DNA leading to consistent aberrant splicing
`would strongly suggest that the variant is disease-related.
`
`BRCAI polymorphisms
`Nine polymorphisms were detected in BRCAJ, six in the
`coding sequence and three in introns (Table 3). Four
`polymorphisms alter amino acid residues. Five
`polymorphisms are in complete disequilibrium with each
`other, and a sixth is in partial disequilibrium, based on
`genotypes of 42 chromosomes. Identification of these
`polymorphisms enabled us to screen for mutations causing
`loss of BRCAI transcript in the ten families for which no
`mutations were found bySSCP analysis of genomic DNA
`and cDNA.
`Genomic DNAs from probands of these families were
`genotyped for polymorphisms in the BRCA1 coding
`sequence. Probands of five families with undetected
`mutations were heterozygous for one or more
`polymorphisms. The cDNAofthese five probands were also
`heterozygous, indicating that mRNA was being expressed
`from both BR CAl alleles. For the remaining five families,
`patients heterozygous for coding sequence polymorphisms
`will be identified and the same test carried out
`
`Discussion
`Most of the BRCA1 mutations we have identified result in
`premature termination of protein translation. Such
`truncated proteins could conceivably compete with
`wildtype protein, thereby acting as dominant mutations.
`On the other hand, loss of the wildtype allele in the
`tumours of patients with BRCAI germline mutations
`would suggest the opposite role - that the termination
`mutation was recessive. Paraffin blocks with adequate
`malignant cells for loss-of-heterozygosity (LOH) studies
`were available for one tumour from family 1, five tumours
`from family 3 and two tumours from family 82. For all
`eight of these tumors, the wildtype BRCA1 allele was lost,
`an observation consistent with recessive mutations 5.
`The number of families with BRCAI mutations
`identified is still very small, so associations between
`genotype and phenotype are not at all certain. Cancer-
`related mutations occur throughout the BRCA1 sequence
`(Table 2). In our series, all four families with both ovarian
`and breast cancer had truncation mutations in the first
`half of the protein. However, termination of the BRCA1
`protein only 11 amino acids from the C-terminal end was
`Nature Genetics volume 8 december 1994 (cid:9)
`
`GeneDX 1004, pg. 5
`
`(cid:9)
`
`
`' 1994 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`article
`
`mutation without determining its biological function.
`Third, there are many predisposing mutations in
`BRCAJ, and the risks associated with most are notknown.
`The mutations identified so far have high penetrance,
`because they were identified in families ascertained for
`very high lifetime risk of breast cancer. Whether other
`BRCA1 mutations are associated with more moderate
`risk, possibly influenced by other genetic or
`environmental factors, remains to be seen.
`
`Methodology
`SSCP. PCR was carried out in 50 jil volumes containing 50 ng
`cDNA or genomic DNA, lx PCR buffer (Boehringer Mannheim),
`200 p dATP, dGTP, dITP (BoehringerMannheim), 1O PM dCTP
`(Boehringer Mannheim), 50 pmoles each primer from Table I
`(Operon), 05 pCi 32P-dCTP (NEN Dupont), 1.25 U Taq DNA
`polymerase (Boehringer Mannheim). PCR conditions for primer
`(Tm from
`pairs were 35 cycles of 94 (cid:176)C, 45 s annealing temperature
`Table 1) 305; 72 ’C; 30 s. PCR template was lymphocyte cDNA or
`genomic DNA from members of BRCA 1-linked families who carry
`the predisposing haplotype. Amplified samples were diluted 1:10 in
`formamide buffer (98% formamide, 10 mM EDTA pH 8,0.05%
`Bromophenol blue, 0.05% Xylene cyanol), held at 95(cid:176)C for 5 mm,
`then cooled rapidly to 4(cid:176)C. and held for 5 mm. For each sample, 5
`p1 was loaded onto an SSCP gel and run at W (constant power) for
`16 h in 0.6x TBE at room temperature. (An 80 ml gel solution
`contains: 0.5x MDE (AT Biochem), 0.6x TBE, 160 0 25%
`ammonium persulphate, 38 p1 TEMED.) Gels were dried on a
`
`Received 8 November; accepted 9 November 1994.
`
`vacuum gel dryer and exposed to film for 12-24h with an intensifying
`screen. Variant bands were cut out of the gel and rehydrated in 100
`p1 water.
`
`Sequencing. PCR was carried out in 50 0 volumes containing 50 n
`DNA or cDNA or I p1 of rehydrated SSCP gel fragment solution, lx
`PCR buffer (Boehringer Mannheim), 200 ;LM dATP, dCTP, dITP,
`dGTP (BoehringerMannheim), 1.25 U Taqpolymerase(Boehringer
`Mannheim), 50 pmoles each primer (Operon) that detected the
`SSCP variant Cycling conditions were 35 cycles of 94 ’C, 45 s,
`annealing temperature (Tm in Table 1) 30 s, 72 ’C, 30 s. Direct
`sequencingofthedouble-stranded DNA from the PCRamplification
`was performed with the USB PCR product sequencing kit, using (cid:176)S
`ATP as the isotope. Sequencing primers were the same as those used
`to amplify thetemplate. Sequence was run on 6% acrylamide/bis gels
`at 70 Wconstant power for 2 h, and exposed to X-ray film for 12-24
`h. Mutations were confirmed in multiple members of each familyby
`amplifying and directly sequencing from genomic DNA.
`Intron 6 sequence was obtained by amplifying genomic DNA
`using the primer pair axon 6&7 (ftp file at morgan.med.utah.edu ),
`with the reaction mixture described above. Direct sequencing was
`done on the 800 bp product as described above.
`
`Acknowldg.ments
`This work was supported in part by NIH grant ROl CA27632. L.S.F.
`is aKomen Foundation Fellow; C.I.S. is an NIH Fellow; M-C.K. is an
`American Cancer Society Professor.
`
`1. Mild, Y. stal. A strong candidate forthe breast and ovarian cancer susceptibility
`gene BRCA1. Science 266 66-71 (1994).
`2. Hat. J.M. eta). Unkage of early-onset familial breast cancer to chromosome
`17q21. Science 260, 16W1689(1990).
`3. Albertson, N.M. at at A physical map and candidate genes In the BRCAI
`region on chromosome 17q12-21. Nature Canal. 7, 472-479 (1994).
`4. Sharp, P.A. Split genes and ANA splicing. Cell 77, 805-815 (1994).
`5. Friedman, L.S. eta). The search for BRCA1 and the cloning of 22 genes from
`
`chromosome 17q21. Cancer Ass. Qn the press).
`8. American Society of Human Genetics. Statement on genetic testing for
`breast and ovarian cancer. Am. J. hum. Genet 56,860(1994).
`7. ROWalI, S., Newman, B.. Boyd, J. & King, M-C. Inherited predisposition to
`breast and ovarian cancer. Am. J. hum. Genel 56,881-885(1994).
`8. Arena, J.F. etaL Inherited breast cancer: Confirmation of lInkage to BRCA2,
`possible linkage to the estrogen receptor, and the clinical Implications of
`genetic complexity. J. Am. med. Assoc. (In the press).
`
`404 (cid:9)
`
`Nature Genetics volume 8 december 1994
`
`GeneDX 1004, pg. 6