BRCA1 IVS16+6T→C Is a Deleterious Mutation ThatCreates an Aberrant Transcript by Activating aCryptic Splice Donor Site

Thomas Scholl,1* Michael T. Pyne,1 Donna Russo,2 and Brian E. Ward1

1Myriad Genetic Laboratories, Salt Lake City, Utah2Cancer Genetics Program, Columbia-Presbyterian Medical Center, New York, New York

Results and conclusions are presented thatcharacterize BRCA1 IVS16+6T→C as a del-eterious mutation. BRCA1 transcripts fromperipheral blood mononuclear cells of abreast cancer patient with the transitionIVS16+6T→C show the loss of a heterozy-gous base within codon 871. Additionally, anaberrant RNA splicing product which incor-porates 69 bases of the 5* end of intron 16 atthe junction of exons 16 and 17 is producedsolely from the allele with IVS16+6T→C.This insertion contains two in-frame stopcodons and encodes a protein truncated atresidue 1662 (plus 13 residues encoded bythe intron). The aberrant transcript is spe-cifically associated with the intronic vari-ant since it was contained within the inser-tion. Furthermore, sequence analysis of theheterozygous base within codon 871 demon-strates that the two RNA products, produc-tive mRNA and aberrantly spliced RNA, arecontributed to exclusively by separate al-leles. Finally, the aberrant transcript is pro-duced by the activation of a cryptic splicesite which has greater homology with theprimate consensus splice sequence than themutated exon 16 donor site. Am. J. Med.Genet. 85:113–116, 1999. © 1999 Wiley-Liss, Inc.

KEY WORDS: BRCA1; splice; mutation;RNA

INTRODUCTIONThe tumor suppressor breast/ovarian cancer suscep-

tibility gene, BRCA1, is located in a 100-kilobase (kb)region of chromosome 17 [Miki et al., 1994]. Mutationsin BRCA1 may confer a 74% risk of breast cancer and28% risk of ovarian cancer during a woman’s lifetime

[Whittemore et al., 1997]. The 5.5-kb messenger RNAis encoded by 23 exons. More than 300 genetic variantsdispersed throughout BRCA1 have been identified viathe analysis of both patients and controls [Breast Can-cer Information Core, 1998]. Over 50 intronic geneticvariants have been described; only a few of these havebeen characterized as mutations through their strongcorrelation with disease [Friedman et al., 1994; Serovaet al., 1996].

A variety of genetic alterations can lead to changes inRNA splicing. Highly conserved nucleotide sequenceswhich are recognized by the cellular splicing machin-ery are required at the junctions of exons with introns[Shapiro and Senapathy, 1987]. Another less conservednucleotide sequence, the branch site, resides within in-trons and is important in the formation of intermediateRNA-splicing structures [Ruskin and Green, 1985].Also, changes in base composition can create these con-sensus sequences which can then compete with thenormal sites for recognition by splicing factors and pro-duce additional RNA products [Chen et al., 1998; DeKlein et al., 1998].

MATERIALS AND METHODSSubject

The genetic variant IVS16+6T→C was identifiedduring clinical full-sequence analysis of BRCA1. Theuncertain clinical significance of this variant was re-ported to the patient through genetic counseling. Thepatient elected to submit another blood sample forRNA analysis as part of a research protocol.

Nucleic Acid TechniquesComplementary DNA was synthesized from total

RNA isolated from peripheral blood mononuclear cellsusing commercially available kits (RNAeasy, Qiagen,Chatsworth, CA and Superscript Preamplification Kit,Gibco BRL, Gaithersburg, MD). Samples of total RNAisolated using this technique contain some residual ge-nomic DNA. Aliquots of these preparations weretreated with RNase A and this DNA was sequenced asan additional confirmation of the original genotypes. Inexperiments examining cDNA, the lengths of the in-trons prevent amplification from the genomic DNA.The primers used in these experiments contained the

*Correspondence to: Dr. Thomas Scholl, Director of ProductDevelopment, Myriad Genetic Laboratories, 320 Wakara Way,Salt Lake City, UT 84108. E-mail: tscholl@myriad.com

Received 22 September 1998; Accepted 2 February 1999

American Journal of Medical Genetics 85:113–116 (1999)

© 1999 Wiley-Liss, Inc.

following designations and gene-specific sequences (allnumeric base pair designations conform to GenBanksubmission U14680 [Miki et al., 1994]: 11F, 58-AGTGTGCAGCATTTGAAAACCC (bases 2530→2551);11R, 58-TCCAGTTTCGTTGCCTCTGAACT (bases3111→2989); 14F, 58-GTGTCTGCAGATAGTTCTAC-CAGTA (bases 4554→4578); 15R, 58-CAACTGTGCAT-GTACCACCTATCATCTA (bases 4657→4630); 16 and17R, 58-TCTGGCAAACTTGTACACGAGCATAA(bases 5208→5184); 21R, 58-TGAAGGGCCCATAG-CAACAGA (bases 5516→5496). All primers carriedM13 tails with either the sequence 58-GTTTTC-CCAGTCACGACG or 58-AGGAAACAGCTATGACCATdepending on whether they directed polymerization inthe forward or the reverse directions, respectively. Thenucleotide sequence of polymerase chain reaction(PCR) products was determined on both strands byfluorescent thermocycle chemistry detected on auto-mated instruments (Model 377, Applied Biosystems,Foster City, CA). While conclusions were drawn fromthe sequence data from both strands, all figures showdata from the reverse strand which consistently pro-duced data with slightly better quality.

RESULTSDetection of Genetic Variants Within

BRCA1 Introns

The patient described here was diagnosed withatypical medullary carcinoma of the left breast at age31. The patient derives from a family with a stronghistory of breast and ovarian cancer, including the fol-lowing diagnoses: her mother with breast cancer atages 41 and 46, one maternal aunt with breast cancerat age 42, another maternal aunt with ovarian cancerat age 46 and breast cancer at ages 52 and 53, hermaternal cousin with breast cancer at age 25. All diag-

noses were confirmed by pathology reports. The pater-nal family history is unremarkable. The patient sub-mitted a sample as part of a research protocol forBRCA1 genomic sequence analysis (Myriad GeneticLaboratories, Salt Lake City, UT). The testing processidentified the BRCA1 genetic variants IVS16+6T→Cand the heterozygous missense codon P871L (CCG forproline and CTG for leucine). This research was ap-proved by the Columbia Presbyterian Medical CenterIRB. The control for this study was a 37-year-old malewith no personal or family history of cancer.

Analysis of RNA TranscriptsThe procedure to evaluate BRCA1 transcripts from

the patient is depicted schematically in Figure 1. Aregion of approximately 3 kb within the cDNA includ-ing the polymorphic codon 871 and the junction at ex-ons 16 and 17 was amplified using the primer pair 11Fand 21R. The control specimen for these experimentsalso contained P871L. The fragments were isolated fol-lowing size separation by agarose gel electrophoresis.The fragment lengths were indistinguishable betweenthe patient and control reactions. The isolated productswere used as a target in a second set of semi-nestedamplifications using primer pairs 11F with 11R and14F with 21R. The products from these secondary re-actions were sequenced.

The nucleotide sequence of the transcripts isolatedfrom the patient displayed a loss of the heterozygousbase in codon 871. Only the sequence CCG, encodingproline, was detected indicating that the PCR productwas derived from only one allele (Fig. 2). Furthermore,the junction at exons 16 and 17 was interrupted by a 69base insertion derived from the 58 end of intron 16 (Fig.3). This product is derived solely from the allele con-taining IVS16+6T→C which is present within the in-sertion and not heterozygous. This insertion termi-nates the open reading frame with the stop codon TAA

Fig. 1. Schematic depiction of regions of BRCA1 and the PCR products analyzed to characterize IVS16+6T→C. The 38 region of BRCA1 including aportion of exon 11 through part of exon 22 is depicted. PCR products amplified for this analysis are designated by the primers used (above) and their basepositions within the cDNA (below). The position of the polymorphic codon 871 is designated with an arrow.

114 Scholl et al.

at the fourteenth position in the insertion (Fig. 3).Identical analysis was performed on a control RNAsample that is also heterozygous at the polymorphiccodon P871L. The results from the secondary amplifi-cation and sequencing showed that the polymorphiccodon (CCG/CTG) was present in the RNA (Fig. 2) andthat the exon 16–17 junction was normal. The sequencederived from the genomic DNA of both the patient andthe control contain the polymorphism within codon 871(Fig. 2).

These data show that the genetic variantIVS16+6T→C results in an altered transcript which isnot produced by the wild-type allele. From these re-

sults, it is a reasonable conclusion that this transcriptrepresents an aberrant rather than an alternativeRNA-splicing product. However, these data do not ex-plain the absence of the wild-type transcript in the am-plified products. Also, it is unclear if the allele contain-ing IVS16+6T→C produces any RNA with the wild-type sequence. Alternatively spliced RNA products candiffer dramatically in stability in vivo [Beelman andParker, 1995]. These differences in stability can resultin altered representation of the various RNA specieswithin the cell. Possibly, the aberrant transcript de-tected in the PCR products from the patient mask de-tection of the wild-type product by this mechanism, al-though usually it is the mutant transcript which ispreferentially degraded and underrepresented in theRNA [Pohlenz et al., 1998].

To identify the legitimate transcript, primers weredesigned that annealed across the exon 16–17 junction.Amplification with these primers would exclude thecDNA products derived from the transcript containingthe insertion. The products from amplification withprimers 11F and 16 and 17R were gel purified and thisfragment was used a target for a second round of am-plification with primers 11F and 11R. The product fromthe secondary amplification was sequenced and the pa-tient sample contained the sequence CTG at codon 871encoding leucine. The control sample was again poly-morphic at codon 871. This result is most conclusiveregarding the clinical significance of IVS16+6T→C anddemonstrates the loss of the heterozygous base on thewild-type transcript. Therefore, the allele containingIVS16+6T→C produces only the aberrantly splicedproduct and does not contribute to BRCA1 expression(Fig. 2). As an additional control, cDNA was amplifiedusing the primer pair 11F and 15R; 15R anneals up-stream of the junctions affected by the splice mutation.When these products were reamplified to sequencecodon 871, the patient sample showed the loss of theheterozygous base and read GCC, while the control waspolymorphic at that position (Fig. 2). This control ex-cludes preferential amplification due to the insertionas the cause for the loss of the wild-type fragment inthe PCR reactions using primers 11F and 21R. Thisresult supports the conclusion that the aberrant tran-script is more abundant than the legitimate transcriptin the RNA preparations.

The mutant RNA splice product results from the ac-

Fig. 3. A 69 base insertion between exons 16 and 17 results from IVS16+6T→C. A portion of a tracing generated by fluorescent sequencing of a PCRproduct amplified using primers 14F and 21R. The target for that reaction was the gel-purified fragment from an amplification with the primer set 11Fand 21R. The reverse strand is shown. The intron/exon boundaries are indicated by bars (). IVS16+6T→C and the first stop codon are designated by anarrow and a bracket respectively.

Fig. 2. Sequence analysis of the polymorphic BRCA1 codon 871. Por-tions of fluorescent sequence data (extracted with ABI Prism sequencinganalysis software version 2.1.1) surrounding codon 871 were assembled forcomparison; reverse strands shown. In each case, the sequence was derivedfrom the product of amplification with the primer pair 11F and 11R. Thetarget for these reactions was genomic DNA [11F-11R (DNA)] or the gelpurified fragments from amplification with the primer set indicated abovethe data pairs on cDNA [11F-21R, 11F-15R, and 11F-16&17R]. While thebase was designated by the software in the control samples for 11F-15Rand 11F-16&17R, it is clearly heterozygous.

BRCA1 IVS16+6T→C Is a Deleterious Mutation 115

tivation of a downstream cryptic splice donor site dueto the T→C transition within the exon 16 donor site(Fig. 4). The exon 16 donor site matches the primateconsensus sequence completely at its 38 end, containedwithin the intron. IVS16+6T→C alters the most 38 baseof the donor site away from the consensus sequence. Incomparsion, the cryptic site which is activated due tothis mutation also deviates from the consensus withinthe intron. However, the cryptic site matches the pri-mate consensus at the 58 dinucleotide, AG, which oc-curs within the exon. The 58 dinucleotide of the wild-type site, TT, does not match the consensus.

CONCLUSION

The intronic mutation BRCA1 IVS16+6T→C is char-acterized conclusively. Foremost, there is a loss of aheterozygous base in the wild-type transcript whichdemonstrates that the productive messenger RNA isproduced from only one allele. Also, an aberrantlyspliced RNA product was detected which is producedsolely from the other allele which bears IVS16+6T→C;the mutant base is contained on this product. This mu-tant RNA contains a 69 base insertion from the 58 endof intron 16 and includes stop codons which terminatethe open reading frame. In all cases, the nucleotidesequencing indicated that each of these two RNA prod-ucts was produced exclusively by one allele.

Large and complex genes, such as BRCA1, whichcontain dispersed and unidentified mutations, warrantgenomic DNA sequence analysis for clinical diagnosisof predisposition to disease. Sequencing is the besttechnique to provide for the detection of novel geneticvariants, as well as an unmatched degree of accuracy.However, problems with interpretation arise sincevariants of uncertain clinical significance are detected.

For variants occurring within introns, biochemicalanalysis is complicated when alternative RNA-splicingproducts are produced, as occurs in BRCA1 [Miki et al.,1994; Lu et al., 1996]. Since the types and the relativeratios of these different RNA species can vary betweentissues or depend on the techniques employed or the

condition of the samples at processing, it can be diffi-cult to differentiate between aberrant and alternativeRNA-splicing products. This is especially problematicwhen conclusions drawn from these data are used forclinical determinations. Herein, the contribution ofeach allele to particular transcripts was determinedthrough the use of a coding region polymorphism. Com-bined with the use of control RNA containing the samepolymorphism, IVS16+6T→C could be characterizedunequivocally as a mutation. Fortunately, in this case,the interpretation of the data was simplified since theRNA products from both alleles could be isolated andeach allele contributed solely to one of these RNA spe-cies. Hopefully, this approach will prove useful in char-acterization of other genetic variants of clinically im-portant genes.

REFERENCES

Beelman CA, Parker R. 1995. Degradation of mRNA in eukaryotes. Cell81:179–183.

Breast Cancer Information Core (BIC). 1998. http://www.nhgri.nih.gov/Intramural_research/Lab_transfer/Bic/index.html

Chen W, Kubota S, Teramoto T, Nishimura Y, Yonemoto K, Seyama Y.1998. Silent nucleotide substitution in the sterol 27-hydroxylase gene(CYP 27) leads to alternative pre-mRNA splicing by activating a cryptic58 splice site at the mutant codon in cerebrotendinous xanthomatosispatients. Biochemistry 37:4420–4428.

De Klein A, Riegman PH, Bijlsma EK, Heldoorn A, Muijtjens M, den Bak-ker MA, Avezaat CJ, Zwarthoff EC. 1998. A G→A transition creates abranch point sequence and activation of a cryptic exon, resulting in thehereditary disorder neurofibromatosis 2. Hum Mol Genet 7:393–398.

Friedman LS, Ostermeyer EA, Szabo CI, Dowd P, Lynch ED, Rowell SE,King MC. 1994. Confirmation of BRCA1 by analysis of germline mu-tations linked to breast and ovarian cancer in families. Nat Genet8:399–404.

Lu M, Conzen SD, Cole CN, Arrick BA. 1996. Characterization of func-tional messenger RNA splice variants of BRCA1 expressed in nonma-lignant and tumor-derived breast cells. Cancer Res 56:4578–4581.

Miki Y, Swensen J, Shattuck-Eidens D, Futreal A, Harshman K, TavtigianS, Liu Q, Cochran C, Bennett L, Ding, W, Bell R, Rosenthal J, HusseyC, Tran T, McClure M, Frye C, Hattier T, Phelps R, Haugen-Strano A,Katcher H, Yakumo K, Gholami Z, Shaffer D, Stone S, Bayer S, WrayC, Bogden R, Dayananth P, Ward J, Tonin P, Narod S, Bristow P,Norris F, Helvering L, Morrsion P, Rosteck P, Lai M, Barrett J, LewisC, Neuhausen S, Cannon-Albright L, Goldgar D, Wiseman R, Kamb A,Skolnick M. 1994. A strong candidate for the breast and ovarian cancersusceptibility gene BRCA1. Science 266:66–71.

Pohlenz J, Rosenthal IM, Weiss RE, Jhiang SM, Burant C, Refetoff S. 1998.Congenital hypothyroidism due to mutations in the sodium/iodide sym-porter. J Clin Invest 101:1028–1035.

Ruskin B, Green M. 1985. Role of the 38 splice site consensus in mamma-lian pre-mRNA splicing. Nature 317:732–734.

Serova O, Montagna M, Torchard D, Narod SA, Tonin P, Sylla B, LynchHT, Feunteun J, Lenoir GM. 1996. A high incidence of BRCA1 muta-tions in 20 breast-ovarian cancer families. Am J Hum Genet 58:42–51.

Shapiro M B, Senapathy P. 1987. RNA splice junctions of different classesof eukaryotes: Statistics and functional implications in gene expres-sion. Nucleic Acids Res 15:7155–7174.

Whittemore AS, Gong G, Itnyre J. 1997. Prevalence and contribution ofBRCA1 mutations in breast cancer and ovarian cancer: results fromthree U.S. population-based case-control studies of ovarian cancer. AmJ Hum Genet 60:496–504.

Fig. 4. Alignment of splice sites. The consensus primate donor splicesequence [Shapiro and Senapathy, 1987] is aligned with that for BRCA1exon 16 wild-type, BRCA1 IVS16+6T→C, and the cryptic site utilized inthe mutant transcript.

116 Scholl et al.