I $ !i ' 1996 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`letters
`
`’Myriad Genetics
`Inc., Salt Lake City
`Utah USA
`2Laboratory of
`Molecular
`Endocrinology,
`CHUL Research
`Center and Lava!
`University, Quebec
`City, Canada
`’Department of
`Genetics, Hospital
`for Sick Children,
`University of
`Toronto, Ontario,
`Canada
`4Deparment of
`Hematology-
`Oncology,
`University of
`Pennsylvania,
`Philadelphia, USA
`5Genetic
`Epidemiology
`Group, Department
`ofMedical
`Informatics and
`Departrnenrof
`Internal Medicine,
`University of Utah
`School of Medicine,
`Salt Lake City,
`Utah, USA
`6Departmentof
`Internal Medicine,
`University of
`Michigan Medical
`School, Ann Arbor,
`Michigan, USA
`’Molecular and Cell
`Biology Research
`Laboratory,
`Icelandic Cancer
`Society, Rekjavik,
`Iceland
`5Departmentof
`Human Genetics,
`Memorial Sloan-
`Kettering Cancer
`Center, New York,
`New York, USA
`9Unite de Genetique
`Oncologique,
`Institute Curie,
`Paris, France
`’(cid:176)Division of
`Biology, California
`Institute of
`Technology,
`Pasadena,
`California. USA
`"Unit of Genetic
`Epidemiology,
`International
`Agency for Research
`on Cancer, 150
`cours Albert
`Thomas, 63972,
`Lyon, Cedex 08,
`France
`
`Correspondence
`should be addressed
`to D.E.G.’1
`
`The complete BRCA2 gene
`and mutations in
`chromosome 13q-linked
`kindreds
`
`S.V. Tavtigia&, J. Simard2, J. Rommens 3, F. Couch4,
`D. Shattuck-Eidens’, S. Neuhausen 5, S. Merajver6,
`S. Thorlacius 7, K. Oflitt, D. Stoppa-Lyonnet 9,
`C. Belanger2, R. Bell’, S. Berry’, R. Bogden’,
`Q. Chen’, T. Davis’, M. Dumont2 , C. Frye’,
`T. Hattier’, S. Jammulapati’, T. Janecki’, P. Jiang’,
`R. Kehrer’, J.-F. Leblanc2, J.T. Mitchell1 ,
`J. McArthur-Morrison 3, K. Nguyen5, Y. Peng4 ,
`C. Samson 2, M. Schroeder’, S.C. Snyder’,
`L. Steele 5, M. Stringfellow’, C. Stroup’,
`B. Swedlund’, J. Swensen5, D. Teng’, A. Thomas’,
`T. Tran’, T. Tran5, M. Tranchant 2 ,
`J. Weaver-Feldhaus’, A.K.C. Wong’, H. Shizuya’(cid:176),
`J.E. Eyfjord7, L. Cannon-Albright 5, F. Labrie2 ,
`M.H. Skolnick"’, B. Weber’, A. Karnb’
`& D.E. Goldgar5"
`
`Breast carcinoma is the most common malignancy
`among women in developed countries. Because
`family history remains the strongest single predictor
`of breast cancer risk, attention has focused on the
`role of highly penetrant, dominantly inherited genes
`in cancer-prone kindreds 1 . BRCA 1 was localized to
`chromosome 17 through analysis of a set of high-
`risk kindreds 2, and then identified four years later by
`a positional cloning strategy 3. BRCA2 was mapped
`to chromosomal 13q at about the same time 4. Just
`fifteen months later, Wooster et al. 5 reported a partial
`BRCA2 sequence and six mutations predicted to
`cause truncation of the BRCA2 protein. While these
`findings provide strong evidence that the identified
`gene corresponds to BRCA2, only two thirds of the
`coding sequence and 8 out of 27 exons were isolat-
`ed and screened; consequently, several questions
`remained unanswered regarding the nature of
`BRCA2 and the frequency of mutations in 13q-linked
`families. We have now determined the complete
`coding sequence and exonic structure of BRCA2
`(GenBank accession #U43746), and examined its
`pattern of expression. Here, we provide sequences
`for a set of PCR primers sufficient to screen the
`entire coding sequence of BRCA2 using genomic
`DNA. We also report a mutational analysis of BRCA2
`in families selected on the basis of linkage analysis
`and/or the presence of one or more cases of male
`breast cancer. Together with the specific mutations
`described previously, our data provide preliminary
`insight into the BRCA2 mutation profile.
`BRCA2 lies neat the centre of a 1.4-megabase (Mb)
`interval flanked by markers D13S1444 and D13S310 (F.C.
`et al., unpublished), completely within a 0.3-Mb
`homozygous deletion identified in a pancreatic carcino-
`ma xenograft6 The full-length sequence of the BRCA2
`transcript was assembled by combination of several
`smaller sequences obtained from hybrid selection, exon
`
`trapping, cDNA library screening, genomic sequencing
`and inter-clone PCR experiments using cDNA as tem-
`plate for amplification (’island hopping’; Fig. 1 a). The
`extreme 5’ end of the mRNA, including the predicted
`translational start site, was identified by a modified 5’
`RACE protocol’. The first nucleotide in the sequence (nt
`1) is a non-template G, an indication that the mRNA cap
`is contained in the sequence. A portion of exon 11,
`which is nearly 5 kilobases (kb) in length, was identified
`by analysis of roughly 900 kb of genomic sequence in the
`public domain (ftp://genome.wustl.edu/pub/gscl/brca) .
`This genomic sequence was condensed with our own
`genomic sequence into a set of 160 sequence contigs.
`When the condensed sequence was scanned for open
`reading frames (ORFs), a contiguous stretch of nearly 5
`kb was identified which was spanned by long ORFs. This
`sequence was linked together by island hopping experi-
`ments with two previously identified candidate gene
`fragments (EC. et al., unpublished)’. Our composite
`BRCA2 cDNA sequence consists of 11,385 bp but does
`not include the polyadenylation signal or poly(A) tail.
`Conceptual translation of the cDNA reveals an ORF
`beginning at nt 229 and encoding a protein of 3,418
`amino acids. The peptide bears no similarity to other
`proteins apart from sequence composition. There is no
`signal sequence at the N terminus, and no obvious
`membrane-spanning regions. Like BRCA1, the BRCA2
`protein is highly charged; roughly one quarter of the
`residues are acidic or basic. However, there are few clues
`as to the biochemical function of BRCA2.
`The BRCA2 gene structure was determined by com-
`paring cDNA and genomic sequences. BRCA2 is com-
`posed of 27 exons distributed over roughly 70 kb of
`genomic DNA (Fig. 1 b). A CpG-rich region at the 5’ end
`of BRCA.2 extending upstream suggests the presence of
`regulatory signals often associated with CpG ’islands.’
`Unlike most human genes, the coding sequence is AT-
`rich (> 60%). Based on Southern blot experiments,
`BRCA2 appears to be unique, with no close homologue
`in the human genome (data not shown).
`Hybridization of labelled cDNA to human multiple
`tissue northern filters revealed an 11-12 kb transcript
`detectable in thymus and testis (Fig. 2a), suggesting that
`little of the BRCA2 mRNA sequence is missing from our
`composite cDNA. Because the northerns did not
`include mammary gland RNA, we performed RT-PCR
`experiments using a BRCA2 eDNA amplicon that spans
`the last splice junction on a set of human tissue RNAs
`(Fig. 2b). All of the samples produced positive signals.
`The highest levels of expression were observed in breast
`and thymus, with slightly lower levels in lung, ovary and
`spleen. This pattern of expression is similar to that pro-
`duced by BRCAI amplicons 3 .
`Individuals from 18 putative BRCA2 kindreds were
`screened for BRCA2 germline mutations by DNA
`sequence analysis’. Twelve kindreds have at least one
`case of male breast cancer; four have two or more cascs
`and four include at least one individual affected with
`ovarian cancer who shares the linked BRCA2 haplotype.
`Each of the 18 kindreds has a posterior probability of
`harboring a BRCA2 mutation of at least 69% and 9 kin-
`dreds have probabilities greater than 90%. Based on
`these combined probabilities, 16 of the 18 kindreds are
`expected to segregate BRCA2 mutations. The entire cod-
`ing sequence and associated splice junctions were
`
`nature genetics volume 12 march 1996 (cid:9)
`
`333
`
`GeneDX 1026, pg. 1
`
`

`

`letters (cid:9)
`
`' 1996 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`a (cid:9)
`
`________
`1000 bp
`
`-ock,a (cid:9)
`
`0, OGO0 000 (cid:9)
`
`o,00 0
`
`,,020-,000 (cid:9)
`
`- (cid:9)
`
`,*oO-cOI (cid:9)
`
`&020.1001
`
`-
`_________
`
`
`
`
`
`-
`
`Fig. 1 a, Cloning of (cid:9)
`BRCA2. Sequence-
`space relationships
`between the cDNA
`hybrid
`clones, (cid:9)
`selected (cid:9)
`clones,
`cDNA PCR prod-
`ucts, and genomic (cid:9)
`sequences used to
`assemble (cid:9)
`the
`transcript
`8RCA2
`sequence.
`2-Br-C:RACE is a (cid:9)
`biotin-capture (cid:9)
`product
`RACE (cid:9)
`obtained from both
`human breast and
`human (cid:9)
`thymus (cid:9)
`cDNA. The cDNA
`clone X sC713.1 (cid:9)
`was identified by (cid:9)
`screening a pool of (cid:9)
`human testis and (cid:9)
`HepG2 (cid:9)
`cDNA (cid:9)
`libraries with hybrid (cid:9)
`
`clone clone (cid:9)
`selected (cid:9)
`GT (cid:9)
`713. (cid:9)
`The (cid:9)
`5 e q U e n c e (cid:9)
`1-6r:CG026(cid:150)s5 kb (cid:9)
`was (cid:9)
`
`generated generated
`from a PCR prod- (cid:9)
`uct beginning at
`the axon 7/8 junc-
`tion (within ? sC71 3.1) and terminating within an hybrid selected clone that is part of axon 11. The sequence of axon 11 was corrected by comparison to
`hybrid selected clones, genomic sequence in the public domain, and radioactive DNA sequencing gels. Hybrid selected clones located within that axon
`(clone names beginning with mH or GT) are placed below it. The cDNA clones ? wCPF188.1, X wCPF1A5.1, ? wCPF1A5.12, ? wCBF1B6.2, and
`? wCBF1 B6.3 were identified by screening a pool of human mammary gland, placenta, testis, and HepG2 cDNA libraries with the axon trapped clones
`wXBF1 88, wXPF1A5, and wXBF1 B6. The clone ? wCBF1 86.3 is chimaeric (indicated by the dashed line), but its 5 end contained an important overlap
`b, Genomic organization of BRCA2. The exons (boxes
`with wCPF1A5.1. I’ denotes the translation initiator. I denotes the translation terminator.
`and/or vertical lines) are parsed across the publicly available genomic sequences (horizontal lines) such that their sizes and spacing are proportional. The
`name of each genomic sequence is given on the left. The sequences 92M18.00541 and 92M18.01289 actually overlap. Distances between the other
`genomic sequences are not known. No databases contained genomic sequences overlapping with axon 21. The extent of the peptide sequence published
`by Wooster et al. 5 is indicated below the parsed exons. Exons 1, 11 and 21 are numbered. * denotes two adjacent axons spaced closely enough that they
`are not resolved at this scale.
`
`
`
`5kb
`
`1
`2140(23.00685 II (cid:9)
`2140(23.00424 (cid:9)
`2140(2301357 (cid:9)
`2140(23.00276
`
`I
`
`4(cid:151)tI
`
`I
`
`2140(2301355
`2140(23.51225
`noseqence (cid:9)
`2140(23.00522 (cid:9)
`92M15.00541
`
`2M 8.01292
`
`I
`
`II
`
`I (cid:9)
`
`Inlor1aI of BR52 reported in ref s (cid:9)
`
`I
`
`screened for mutations in multiple individuals from 9
`kindreds using either cDNA or genomic DNA (Table
`in). Individuals from the remaining 9 kindreds were
`screened for mutations using only genomic DNA (Table
`I b). These latter screening experiments encompassed
`99% of the coding sequence (all exons excluding exon
`15) and all but two of the splice junctions.
`We identified potentially deleterious sequence alter-
`ations in 9 of the 18 kindreds (Table I). All except one (cid:151)
`a deletion of three nucleotides (kindred 10 19) - involved
`nucleotide deletions that altered the reading frame, lead-
`ing to truncation of the BRCA2 protein. The 3-nt deletion
`was not observed in 36 unrelated breast cancer cases,
`hence we have included it in our mutation tally although
`its effect on BRCA2 function must be proved. All 9 muta-
`tions are distinct. In most cases, segregation studies show
`that the mutations are present in multiple haplotype car-
`riers and absent in noncarriers (data not shown). In addi-
`tion to these mutations, three silent and three missense
`substitutions were detected. Based on their frequencies in
`a set of control chromosomes, we have classified these
`lc).
`variants as neutral polymorphisms (Table
`Nine of the 18 kindreds were tested for transcript loss.
`Specific polymorphic sites known to be heterozygous in
`genomic DNA (Table Ic) were examined in cDNA from
`kindred individuals. The appearance of hemizygosity
`was interpreted as evidence for a mutation leading to
`reduction in mRNA levels. Two of the 9 kindreds dis-
`played signs of reduced transcript levels, However, one
`of these kindreds (1018) contained a previously identi-
`
`fied frameshift mutation, while the second (2367) con-
`tained an aberrantly spliced BRCA2 mRNA that lacked
`exon 2 (data not shown). The abundance of this mutant
`transcript was estimated at roughly 20% of wild-type.
`This implies that some mutations in the BRCA2 coding
`sequence may destabilize the transcript in addition to
`disrupting the protein sequence, similar to BRCAJ (ref.
`9). In no case was a purely regulatory mutation inferred.
`In summary, 56% of the kindreds (10/18) contained an
`altered BRCA2 gene. Half of our kindreds contained
`microdeletion mutations, mostly frameshifts; none con-
`tained missense or nonsense mutations.
`BRCA2 is remarkably similar to BRCA1. Both genes
`encode exceptionally large, highly charged proteins; both
`have many exons; both have a large exon 11(3,426 bp for
`BRCA1 and 4,932 bp for BRCA2); both have translational
`start sites in exon 2; both have coding sequences that are
`AT-rich; both span approximately 70 kb of genomic DNA;
`and, both are expressed at high levels in testis. Whether or
`not BRCAJ and BRCA2 participate in the same pathway
`of tumour suppression in breast epithelium is not known.
`The different phenotypes of the two mutant genes, partic-
`ularly the role of BRCA2 in male breast cancer, suggest
`that they may not function in the same genetic pathway.
`Mutational analysis of BRCA2 reveals other features
`in common with BRCAJ, The distribution of mutations
`in BRCA2 appears to be uniform based on our data and
`data from Wooster et al. 5 . Mutations have been identi-
`fied in 6 of the 26 coding exons of BRCA2 (exons 2, 9,
`10, 11, 18 and 23). Nine mutations have been detected
`
`334 (cid:9)
`
`nature genetics volume 12 march 1996
`
`GeneDX 1026, pg. 2
`
`(cid:9)
`(cid:9)
`

`

`' 1996 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`letters
`
`a
`
`coW (cid:9)
`
`0
`C
`
`(om
`-> (cid:9)
`’- 0 > (cid:149)- (cid:9) 0
`(I)FQ..l-OO..J
`
`kb
`
`t) (cid:9)
`
`W (cid:9)
`> (cid:9)
`s, (cid:9)
`(cid:176) (cid:9) E E
`:3 (cid:9)
`0 ( >’-
`’n
` U
`4-’
`> (cid:9)
`C
`2 0.
`c (cid:9)
`a (cid:9)
`>
`- (cid:9)
`..J 0 0. 0. (1) V) H
`
`.0
`
`U,
`M
`’-
`cc
`
`’- (cid:9)
`cc
`
`c C)
`C
`
`
`
`up (cid:9)
`
`40
`
`Fig. 2 a, Analysis of BRCA2 expression.
`Upper panel: Multiple tissue northern
`(MTN) filters (Clontech) probed with the
`1.55-kb 1-Br.00026--45-kb PCR product.
`The 11-12 kb BPCA2 transcript is detect-
`ad in testis. Lower panel: the same filter
`probed with glyceraldehyde-3-phosphate
`dehydrogenase. b, RT-PCR analysis of
`BRCA2 expression. A BRCA2 eDNA ampli-
`con spanning the last splice )unction was
`amplified from 0.4 ng or random-primed A
`cDNA from the indicated tissues.
`
`9.5-
`7.5 -
`4.4-
`
`2.4-
`
`1.35- (cid:9)
`
`GAPDH
`
`in exon 11, and six elsewhere (very
`close to the expected spread). in
`addition, of the 9 kindreds for which
`t5 r- 1JJJ,J1J cDNA was available one revealed a
`splicing defect. No confirmed regu-
`latory mutations (0/4) could be
`inferred. Thus, the rate of such
`mutations in BRCA2 appears to be
`comparable (or lower) to the rate
`observed in BRCA1 (10_20%) 10.
`The mutation profile of BRCA2, however, may differ
`from BRCA1. Of the 15 sequence alterations described so
`far in BRCA2 (9 here, 6 from ref. 5), all involve deletions
`of 1-6 nucleotides. In contrast, microinsertions plus
`point mutations in BRCAI are about as common as
`
`inicrodeletions. Furthermore, no point mutations have
`been detected so far in BRCA2, whereas they constitute a
`significant fraction of BRCA1 variants. Thus, BRCA2
`may be especially vulnerable to deletion mutations. If
`confirmed, this trait has clear implications for mutation
`detection. The lack of multiple observations of specific
`mutations indicates that the number of independent
`BRCA2 mutations in the population may be even greater
`than the number of BRCAI mutations.
`Approximately two thirds as many mutations were
`detected in BRCA2 as expected among our 18 families.
`This may imply the presence of mutations in regions
`that are difficult to detect, or in regions that were not
`screened. For example, only half of the kindreds were
`screened for regulatory mutations due to lack of cDNA
`
`Table 1 BRCA2 mutations and polymorphisms
`
`a, Families screened for complete coding sequence (informative eDNA sample)
`Prior
`BRCA2
`Family
`Number of cancer cases
`Lod
`score probability mutation
`FBC (cid:9)
`FBC (cid:9) Ov MBC
`<Soyrs
`
`exon codon
`
`effect
`
`UT-107
`UT-1018
`UT-2044
`UT-2367
`UT-2327
`UT-2388
`UT-2328
`UT-4328
`Ml-1016
`
`20 (cid:9)
`ii (cid:9)
`8 (cid:9)
`9 (cid:9)
`13 (cid:9)
`3 (cid:9)
`10 (cid:9)
`4 (cid:9)
`4 (cid:9)
`
`18 (cid:9)
`9 (cid:9)
`6 (cid:9)
`5 (cid:9)
`6 (cid:9)
`3 (cid:9)
`4 (cid:9)
`3 (cid:9)
`2 (cid:9)
`
`2 (cid:9)
`0 (cid:9)
`4 (cid:9)
`1 (cid:9)
`0 (cid:9)
`1 (cid:9)
`0 (cid:9)
`0 (cid:9)
`0 (cid:9)
`
`3
`1
`1
`0
`0
`0
`1
`0
`1
`
`5.06
`2.47
`2.13
`2.09
`1.92
`0.92
`0.21
`0.18
`0.04
`
`1.00
`1.00
`1.00
`0.99
`0.99
`0.92
`0.87
`0.69
`0.81
`
`277 delAC
`982del4
`4706 del4
`SP
`ND
`ND
`ND
`ND
`ND
`
`b, Families screened for all exam except 15 (no eDNA sample available)
`
`2
`9
`11
`
`17
`252
`1493
`
`termination codon at 29
`termination codon at 275
`termination codon at 1502
`deletion of exon 2 in mRNA
`
`CU-20
`CU-159
`UT-2043
`IC-2204
`MS-075
`UT-1019
`UT-2027
`MS-036
`IJT-2171
`
`4 (cid:9)
`8 (cid:9)
`2 (cid:9)
`3 (cid:9)
`4 (cid:9)
`5 (cid:9)
`4 (cid:9)
`3 (cid:9)
`5 (cid:9)
`
`3 (cid:9)
`4 (cid:9)
`2 (cid:9)
`1 (cid:9)
`1 (cid:9)
`1 (cid:9)
`4 (cid:9)
`2 (cid:9)
`4 (cid:9)
`
`2 (cid:9)
`0 (cid:9)
`1 (cid:9)
`0 (cid:9)
`0 (cid:9)
`0 (cid:9)
`0 (cid:9)
`0 (cid:9)
`2 (cid:9)
`
`2
`0
`1
`4
`1
`2
`0
`1
`0
`
`1.09
`0.99
`0.86
`0.51
`0.50
`nd
`0.39
`nd
`ad
`
`1
`0.94
`0.97
`0.98
`0.93
`0.95
`0.79
`0.90
`nd
`
`18
`8525 deiC
`23
`9254 del5
`4075 delGT 11
`999 del5
`9
`6174 delI
`ii
`4132 c1913
`11
`ND
`ND
`ND
`
`2766
`3009
`1283
`257
`1982
`1302
`
`termination codon at 2776
`termination codorr at 3015
`termination codon at 1285
`termination codon at 273
`termination codon at 2003
`deletion ofthri302
`
`c, Common polymorphisms in BRCA2
`
`Polymorphism
`
`Description
`
`Effect
`
`5UTR-203
`PM-1342
`PM-2457
`PM-3199
`PM-3668
`PM-4035
`PM-7470
`3UTR-10,854
`3UTR-11,316
`
`TACCM(G/A)CATTG
`GTA GCA (C/A)AT CAG
`GTA CAA CA(T/C)TCA
`TACATG(A/G)ACMI
`CCI GM A(A/G)C CAG
`GAT TCT GT(T/C) G’TT
`ACT ,AAA TC(NG) CAT
`AAAAGM(G/A)CAITTCTA
`ATTIAlllllllll(t)CMC
`
`His-Asn
`His-His
`ASn-ASP
`Asn-+Ser
`Val-+VaI
`Ser-Ser
`
`Ig vs, T10
`
`A
`
`9
`24
`2
`37
`34
`0
`34
`17
`19:(7)
`
`Number of chromosomes
`G (cid:9)
`
`C
`
`T
`
`Total
`
`0
`14
`0
`0
`0
`4
`0
`0
`
`25 (cid:9)
`0 (cid:9)
`0 (cid:9)
`3 (cid:9)
`6 (cid:9)
`0 (cid:9)
`6 (cid:9)
`15 (cid:9)
`T10:(29)
`
`0
`0
`38
`0
`0
`36
`0
`0
`
`34
`38
`40
`40
`40
`40
`40
`32
`36
`
`Mutations and polymorphisms are given by nucleotide position. SP-inferred splice mutation. ND-none detected, nd-not determined. FBC-female breast
`cancer. Ov-ovaian cancer. M8C-male breast cancer.
`
`nature genetics volume 12 march 1996 (cid:9)
`
`335
`
`GeneDX 1026, pg. 3
`
`(cid:9)
`(cid:9)
`

`

`letters (cid:9)
`
`I S ! ' 1996 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`Table 2 Primers used to amplify and mutation screen
`BRCA2 from genomic DNA
`Reverse Primer
`GTACTGGGTITTTAGCMGCA
`ATTTGCCCAGCATGACACA
`G7AGGAAATG,7rcA,TrM
`GGGOOTMA nflAGGGGM
`AATTGCCTGTATGAGGCAGA
`ATOGTCAG1TACTMCACAC
`cAGGmAGAGAcmcTc
`GTCMGA(cid:176)aQGTANGGTM
`C07AGTC11GCTAG7TC1T
`GAG1TTGATACCCTGMATO
`CATGTATACAGATGATGCCTMO
`ATACATCTTGATTG1T1TCCAV
`TTAQATTTGTGTITTGGTTGAP.
`ccTAcrG77AT0-rrCAGAdAG
`C1TGCTGCTGTCTACCTI3
`CCPAAA(cid:176)GTTSAATCT3ACA
`CGTCTGGAGAEG1TTCCTGAC
`AGTACC7TGCTCTTT1TCATC
`TICGr3AGAGATGA3TITFGTC
`1T7ITGAITATATCTCG1TG
`GACGTAGI3TGMTAGTGAAGA
`TGAGA01TrGGTFCCTAATAC-
`CCCCCWCTGACTAGACM
`UGGAGAGGCAGGTGI5AT
`ATWAc000aNGTG1TSa0r
`CACCACGAAEGGGGG4AA
`ATGOO4ATAOANITACACTCTGTC
`TAGTTCGAGAGACAG1TA6G’
`AECCITMGCCATACTGCC
`GAPATTQAGCATCCUAGTM
`TfACACACACCM.Ex,AAGTCA
`C1TGTTGCT.ATTCTTTGTCTA
`GCGAGAGAGTCTAAMCAG
`ATT1TGTTAGTASGGTCA1TTTT
`CCGTGGCTGGT5ATCTG
`ACCGGTACA,AJkGC1TTCATTG
`
`Forward Primer
`Exon
`TGTTCCCATCCTCACAGTMG
`Exon 2
`GGTTAA.5ACTAAGOTGGGA
`Noon 3
`1TtCCCAGTA7AGAGGAGA
`E. 4
`ATCWAGTAGTA1TCCAACA
`Enorr 5
`GAATAAGTCAI3(3TATtSA7r
`Eons 6
`CIGCMTICAGTMACGUM
`0-ron 7
`OTOTcATGTMTCriAATAGr
`Noon N
`OGACCTAGGTTOATTGCA
`NoonS
`CTATGAGAkEGGTTGTGAG
`Coon iO-r
`A5CAG71GTAGATACCTCTGAA
`Noon 10-2
`CAGCATCTTGMTCTCATACAG
`Noon 10-3
`AAcTrAGTGPAAAATAmAGTGA
`Soon (cid:9)
`1-1
`AGAE0CMC1TTGTCC7TM
`Enon 11-2
`ATGGAAASGAATCMbATGTAr
`Noon 11-3
`GTGT4A5GOAGCATATAA(cid:176)ANT
`Noon 11-4
`GcATAAmMcACOTAGccA --
`Eons 11-5
`OGcyITrATrcTGcTCATGGC
`AACOt3ACTTGCTATITACTGA
`Coon 11-6
`cAaCTAGcoGGA8N,AaGTrA
`Exor, 11-7
`GCC1TAGC1TrTrACACAA
`Noon Il-N
`CCA1T5uw,TFGTCCATATCTA
`Noon 11-9
`Eons 11-10 GMGATAGTACCANQ.4AGTC
`Noon 11_11 GTC7TCACTA1TCACCTACG
`ACTCrnGAAhOATTAGGTCA
`Enon 12
`ITTATGCTGA1TTCTGTTGTAT
`Eons 13
`GUTAGAlaNACAGITACCAGA
`ExOn 14
`AT1TCAAT11TATrnTGC7’
`Noon 15
`ATGTTTITGTAGTGAAGATTCT
`Eon 16
`CAGAGMTAGTFGTAGTTGTT
`Noon 17
`None 18
`T1TTA1TCTCAGTrAITOAGTG
`ATA1TITTSAGGCAG1TCTAGA
`Coon 19
`TGMTG17ATATATGTGAG1T1T
`Eons 20
`C11TrAGCAGTTATATAGTTTC
`Noon 21
`TIT31TGTATIT3TC0TGTTTA
`Eons 22
`ATCA01TC1TCCATTGCA7C
`Eons 23
`CTGGTAOCTCCAOCTAATC
`Noon 24
`cTATrrr(sAmGcTmATrATr GCTAITT0CTTGATACTG(3AC
`Noon 25 (cid:9)
`ACUACAGGAGCCACATANC
`TTGGMsACATS0MTATGTGGG (cid:9)
`Noon 26 (cid:9)
`CTAGATAATTATGATAGGCTNCG GTACTANTGTGTGGT1TGMA
`Eons 27 (cid:9)
`TCAsrGcAfleucucCTcAGc
`
`Nested Primer
`
`
`
`mAGTGANFGTGATTGATGGr
`TAGCTCTITr000ACMTIC
`GCTACCTCCMAACTGTGA
`AGTGGTCTTMGATAGTcAr
`
`TrA1TCT0OrrGTTTrccrrA
`TCAkATrCIDTCTAACACTCC
`AQTSONCGAACA1TCAGACCAG
`AGCATACCMGTCTACTGMr
`CrATAGAGGGAGSACAGAr
`cTGTGAGYTAmGGTGcAr
`AAATGA000TCTGCANCPAN
`TACACTCTGTCATAAAAGCC
`GAGTT17GGmOTIthTAATrG
`TTCAGTATCATCCTATGTI3G
`MTICTAGAGTCACACYICC
`TGMANCTC71ATGATATGTGr
`CCCTAGATACTAA 5AMTANAG
`C1TrOGGTGTflTATGCTrd
`G1TcTGA1-rGcrrrrTATTCC
`
`Sephadex column and reamplifled using the primer RAG. The
`products were then digested with EcoRl, size selected on agarose
`gels, and ligated into pBluescript (Stratagene) that had been
`digested with EcoRI and treated with calf alkaline phosphataae
`(Boehringer Mannheim). Ligation products were transformed
`into competent DH5cz E.coli cells (Life Technologies, Inc.).
`Characterization of retrieved cDNAs. 200 to 300 individual
`colonies from each ligation (from each 250 kb of genomic DNA)
`were picked and gridded into niicrotitre plates for ordering and
`storage. Cultures were replica transferred onto Hybond N mem-
`branes (Amersham) supported by LB agar with ampiciin.
`Colonies were allowed to propagate and were subsequently lysed
`with standard procedures. Initial analysis of the eDNA clones
`involved a prescreen for ribosomal sequences and subsequent
`cross screenings for detection of overlap and redundancy.
`Approximately 10-25% of the clones were eliminated as they
`hybridized strongly with radiolabelled cDNA obtained from
`total RNA. Plasmids from 25 to 50 clones from each selection
`experiment that did not hybridize in the prescreening were iso-
`lated for further analysis. The retrieved eDNA fragments were
`verified to originate from individual starting genomic clones by
`hybridization to restriction digests of DNAa of the starting
`clones, of a hamster hybrid cell line (GM10898A) that contains
`chromosome 13 as its only human material and to human
`genotnic DNA. The clones were tentatively assigned into groups
`based on the overlapping or nonoverlapping intervals of the
`genomic clones. Of clones tested, approximately 85% mapped
`appropriately to the starting clones.
`Method 2: ( refs 14, 15): cDNA prepararion. Poly(A) enriched
`RNA from human mammary gland, brain, lymphocyte, and
`stomach were reverse transcribed using the tailed random
`primer XN 1 , l5’(cid:151)(NH 2 )-GTAGTGGAAGGGTCGAGAACN 12 ]
`and Superscript H reverse transcriptase (Gibco BRL). After 2nd
`strand synthesis and end polishing, the ds eDNA was purified on
`sepahrose CL-4B columns (Pharmacia). cONAs were ’anchored’
`by ligation of a double stranded oligo RP
`[5’(cid:151)(NH2)-TGAGTAGAATTGTAACGGGCGTCATTGTTC
`annealed to 5’(cid:151)GAACAATGAGGGCGGTTAGAATTCTAG-
`TCA-(NH 2)l to their 5’ ends (5’ relative to mRNA) using T4
`DNA ligase. Anchored ds cDNA was then repurified on sep-
`ahrose CL-4B columns. Selection. cDNAs from mammary gland,
`brain, lymphocyte, and stomach tissues were first amplified
`using a nested version of RP (RP.A: 5’(cid:151)TGAGTAGAATTC-
`TAACGGCCGTCAT) and XPCR [5’(cid:151)(104)-GTAGTGCAAG-
`GCTCGAGAAC] and purified by fractionation on Sepharose
`CL-4B. Selection probes were prepared from purified Pis, BACs,
`or PACs, by digestion with Hinfl and Exonuclease III. The single
`stranded probe was photo-labelled with photobiotin (Gibco
`BRL) according to the manufacturers recommendations. Probe,
`eDNA and Got-I DNA were hybridized in 2.4 M TEA-Cl, 10mM
`NaPO4, 1mM EDTA. Hybridized cDNAs were captured on
`streptavidin-paramagnetic particles (Dynal), eluted, reamplifled
`with a further nested version of RP RRB: 5’(cid:151)(PO4)-TGAGTA-
`GAATTCTAACGGCCGTCAYTG] and XPCR, and size selected
`on Sepharose GL-6B. The selected, amplified cDNA was
`hybridized with an additional aliquot of probe and Cot-1 DNA.
`Captured and eluted products were amplified again with RP.B
`and XPCR, size selected by gel electrophoresis and cloned into
`dephosphorylated Hincil-cut pUC18. Ligation products were
`transformed into XL2-Blue ultra-competent cells (Stratagene).
`Analysis. Approximately 192 colonies for each single-probe selec-
`tion experiment were amplified by colony PCR using vector
`primers and blotted in duplicate onto Zeta Probe nylon filters
`(Bio-Rad). The filters were hybridized using standard proce-
`dures with either random primed Cot-I DNA or probe DNA
`(P1, BAG, or PAC). Probe positive, Cot-1 negative clones were
`sequenced in both directions using vector primers on an ABI 377
`sequencer.
`
`Exon trapping. Exon amplification was performed using a mini-
`mally overlapping set of BAGs, Pis and PACs in order to isolate a
`number of gene sequences from the BRCA2 candidate region.
`
`Primers with a were used for sequencing
`Primers without a are replaced by toe internal rooted poorer in the third column for bus’ the second round of FOR and
`sequencing. For large enons requiring internal sequencing primers, signifies the primers used to amplify the eons
`
`samples. In addition, in only one of the two families in
`which transcript loss was detected was the actual
`sequence alteration identified. However, 7/9 kindreds
`with Lod scores over 0.39 revealed mutations. Thus, it is
`possible that the assumed prior probabilities for our
`kindreds were inflated. Several of the kindreds for which
`mutations have not been defined may not segregate
`BRCA2 mutations. Many have been screened without
`success for BRCAI mutations. Therefore, some of the
`families may represent sporadic clusters; others may be
`afflicted with breast cancer due to segregation of genes
`besides BRCAI and BRCA2.
`The characterization of two genes, BRCAI and
`BRCA2, that together may account for the vast majority
`of early-onset hereditary breast cancer, is a major step
`toward early detection of an important human disease.
`One of the significant goals ahead is the development of
`reliable diagnostic tests for BRCAI and BRCA2. In addi-
`tion, the definition of other genes that may contribute to
`breast cancer incidence is an important pursuit.
`
`Methods
`Hybrid selection. Two distinct methods of hybrid selection were
`used. Method 1: cDNA preparation and selection. P oly(A)h
`enriched RNA from human mammary gland, ovary, testis, fetal
`brain and placenta tissues and from total RNA of the cell line
`Caco-2 (ATCG HTB 37) were reverse transcribed using the tailed
`random primer RXGN 6 (5’(cid:151)CGGAATTCTGCAGATCTA’B’CN 6 )
`and M-MLV Reverse Transcriptase (Life Technologies, Inc.). First
`strand cDNA was poly(A) tailed, 2d strand synthesis was primed
`with the oligo RXGT 12 (5’(cid:151)CGGAATTCTGCAGATCT 12), and
`then the ds eDNA was expanded by amplification with the
`primer RAG (5’(cid:151)CGGAATTCTGGAGATCT). Hybrid selection
`was carried Out for two consecutive rounds to immobilized P1 or
`BAG DNA as described" , ". Groups of two to four overlapping
`Pt and/or BAG clones were used in individual selection experi-
`ments. Hybridizing cDNA was collected, passed over a G50 Fine
`
`336 (cid:9)
`
`nature genetics volume 12 march 1996
`
`GeneDX 1026, pg. 4
`
`

`

`I
`
`(cid:149) ! ' 1996 Nature Publishing Group http://www.nature.com/naturegenetics
`
`
`
`letters
`
`Pools of genomic clones were assembled, containing from
`100.-300 kb of DNA in the form of 1-3 overlapping genomic
`clones. Genomic clones were digested with Pstl or BansH 1 + Bgll 1
`and ligated into Pstl or BatnHl sites of the pSPL3 splicing vector.
`The exon amplification technique was performed’ 6 and the end
`products were cloned in the pAMPI plasmid from the Uradil DNA
`Glycosylase cloning system (BRL). Approximately 6,000 clones
`were picked, propagated in 96 well plates, stamped onto filters,
`and analysed for the presence of vector and repeat sequences by
`hybridization. Each clone insert was PCR amplified and tested for
`redundancy, localization, and human specificity by hybridization
`to grids of exons and dot blots of the parent genomic DNA.
`Unique candidate exons were sequenced, searched against the
`databases, and used for hybridization to cDNA libraries.
`
`mg/ml denatured salmon testis DNA and 2 ig/ml Poly-(A).
`Hybridizations were in the same solution with the addition
`dextran sulfate to 4% and probe. Stringency washes were in
`Dlx SSC/0.1% SDS at 50(cid:176)C.
`
`RT-PCR analysis. Poly-(A) RNA extracted from 11 human tis-
`sues was reverse transcribed using random primes and Super-
`script II reverse transcriptase (Gibco BIlL). Thereafter, 0.4 ng of
`each cDNA sample was amplified for 20 cycles using the BRCA2
`primers B2#F9833 (5’-CGTACACTGCTCAAATCATTC) and
`B2#R10061 (5’-GACTAACAGGTGGAGGTAAAG). Samples
`were diluted 10-fold, and then 2 pl aliquots reamplified for 18
`cycles using the primers B2#F9857 (5’-GTACAGGAAA-
`CAAGCTTCTGA) and 82#R10061.
`
`5’ RACE. The Send of BRCA2 was identified by a modified RACE
`protocol called biotin capture RACE. Poly(A) enriched RNA
`from human mammary gland and thymus was reverse transcribed
`using the tailed random primer XN 12 [5’-(NH2)-GTAGTGCAAG-
`GCTCGAGAACN 12] and Superscript II reverse transcriptase
`(Gibco BRL). The RNA strand was hydrolysed in NaOH and first
`strand cDNA purified by fractionation on Sepharose CL-4B
`(Pharmacia). First strand cDNAs were ’anchored’ by ligation of a
`double stranded oligo with a 7 bp random 5’ overhang [ds UCA
`5’CCUCACACGCGTATCGATFAGTCACN7-(NH2) annealed
`to 5’(PO4)-GTGACTAATCGATACGCGTGTGAAGGTGC] to
`their 3’ ends using T4 DNA ligase. After ligation, the anchored
`cDNA was repurified by fractionation on Sepharose CL-4B. The 5’
`end of BRCA2 was amplified using a biotinylated reverse primer
`[5’(li)-TrGAAGAACAACAGGAC7TTCACTA] and a nested ver-
`sion of UCA [UCP.A 5’-CACCTTCACACGCGTATCG]. PCR
`products were fractionated on an agarose gel, gel purified, and
`captured on streptavidin-paramagnetic particles (Dynal). Cap-
`tured cDNA was reamplified using a nested reverse primer
`j5’-GUCGTAATTGTTGTTTTTATGlTCAGJ and a further
`nested version of UCA [UCP.B: 5’-CCUCACACGCGTATC-
`GATTAGI. This PCR reaction gave a single sharp band on an
`agarose gel; the DNA was gel purified and sequenced in both
`directions on an ABI 377 sequencer.
`
`eDNA clones. Human cDNA libraries were screened with 32P
`labelled hybrid selected or exon trapped clones. Phage eluted
`from tertiary plaques were PCR amplified with vector specific
`primers and then sequenced on an ABI 377 sequencer.
`
`Northern blots, Multiple tissue northern (MTN) filters, which
`are loaded with 2 ig per lane of poly(V RNA derived from a
`number of human tissues, were purchased from Ciontech. 32P-
`random-primer labelled probes corresponding to the 1.55 kb
`PCR fragment 1-Br:CG026-45 kb (exon 7/8 junction into exon
`11), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH)
`were used to probe the filters. Prehybridizations were at 42(cid:176) C in
`50% formamide, 5x SSPE, 1% SOS, 5x Denhardt’s mixture, 0.2
`
`PCR amplification and mutation screening. All 26 coding exona
`of BRCA2 and their associated splice sites were amplified from
`genomic DNA as described". The DNA sequences of the
`primers, some of which lie in flanking intron sequence, used for
`amplification and sequencing appear in Table 2. Some of the
`exons (2 through 10, 11-5, 11-6, 11-7, and 23 through 27) were
`amplified by a simple one step method. The PCR conditions for
`those exons were: single denaturing step of 95 (cid:176)C 0 mm);