Papers in Press. Published November 5, 2009 as doi:10.1373/clinchem.2009.132274
The latest version is at http://www.clinchem.org/cgi/doi/10.1373/clinchem.2009.132274
Clinical Chemistry 56:1
000 – 000 (2010)
Molecular Diagnostics and Genetics
Alternative Splicing and Molecular Characterization of
Splice Site Variants:
BRCA1 c.591C>T as a Case Study
Vanesa Dosil,1 Alicia Tosar,1 Carmen Cañadas,1 Pedro Pérez-Segura,1,2 Eduardo Dı́az-Rubio,1,2
Trinidad Caldés,1,2 and Miguel de la Hoya1*
BACKGROUND: Deleterious mutations in BRCA1 (breast
cancer 1, early onset; MIM 113705) increase human
breast and ovarian cancer [HB(O)C] risk; however,
many variants cannot be readily classified as deleterious or neutral. Unclassified variants (UVs) pose serious problems in genetic counseling. RNA-splicing
analysis is essential for the assessment of many UVs.
alternative splicing isoforms need to be considered
when assessing the role of BRCA1 UVs on splicing.
© 2009 American Association for Clinical Chemistry
Our data support a nonpathogenic role
for the BRCA1 c.591C⬎T variant. Naturally occurring
Deleterious germ-line mutations in BRCA1 (breast
cancer 1, early onset; MIM 113705) confer a high risk of
human breast and ovarian cancer [HB(O)C].3 BRCA1
genetic testing has revealed a large number of sequence
variants, but only frameshifts and premature stop
codons are generally assumed to be pathogenic without
further analysis (1 ). Therefore, a growing number of
variants are being detected that cannot be readily distinguished as either disease-causing or neutral. In the
Breast Cancer Information Core (BIC) database, 691
such variants were listed as of October 2009 (2 ). These
so-called unclassified variants (UVs) pose serious
problems in genetic counseling, ranging from the impossibility of accurate risk assessment to psychological
aspects (3 ).
Several approaches have been proposed to reclassify BRCA1 UVs into either disease-causing or neutral
variants (1 ). Often, testing the effect of UVs on RNA
splicing becomes essential to assess their pathogenicity.
For instance, exon skipping or the activation of a cryptic splice site (leading to the deletion/insertion of exonic/intronic sequences) provides strong evidence that
the variant is pathogenic. On the contrary, typical RNA
processing suggests that the variant is neutral (4 –14 ).
Surprisingly, the existence of naturally occurring
BRCA1 splicing isoforms (15 ) is mostly overlooked
when the role of BRCA1 variants on splicing is tested.
We report the identification and characterization
of allele-specific splicing (ASS) linked to the nonpathogenic BRCA1 variant c.591C⬎T. Our study illustrates
the importance of taking into consideration naturally
1
3
METHODS: Denaturing gradient gel electrophoresis was
used to genotype the BRCA1 c.591C⬎T variant in 685
index cases of HB(O)C families, 326 sporadic breast
cancer cases, and 450 healthy controls from Spain. In
silico tools were used to predict the effect of the
c.591C⬎T variant on splicing. In vitro splicing analysis
was performed in 7 c.591C⬎T carriers and 10 noncarriers. cDNAs were PCR-amplified with primers designed to detect BRCA1 alternative splicing isoforms.
The products were analyzed by capillary electrophoresis. Peak areas were used to quantify the relative abundance of each isoform. Sequencing through exonic
single-nucleotide polymorphisms (SNPs) enabled us
to discriminate wild-type and variant transcripts.
RESULTS:
c.591C⬎T was detected in HB(O)C cases
(1.5%), breast cancer cases (0.3%), and controls
(0.9%). c.591C⬎T induced BRCA1 exon 9 skipping
and modified the relative expression of ⌬(9,10),
⌬(9,10,11B), ⌬11B, and full-length isoforms. The
mean ratio of ⌬(9,10) to the full-length isoform increased from 0.25 in noncarriers to 1.5 in carriers. The
mean ⌬(9,10,11B)/⌬11B ratio increased from 0.2 to 4.
Overall expression levels of c.591C⬎T and wild-type
alleles were similar.
CONCLUSIONS:
Laboratorio de Oncologı́a Molecular and 2 Servicio de Oncologı́a Médica, Hospital Clı́nico San Carlos, Madrid, Spain.
* Address correspondence to this author at: Laboratorio de Oncologı́a Molecular,
Planta baja sur, Hospital Clı́nico San Carlos, c/Martı́n Lagos s/n, Madrid 28040,
Spain. Fax 0034913303544; e-mail [email protected].
Received June 16, 2009; accepted October 8, 2009.
Previously published online at DOI: 10.1373/clinchem.2009.132274
Nonstandard abbreviations: HB(O)C, human breast and ovarian cancer; BIC,
Breast Cancer Information Core; UV, unclassified variant; ASS, allele-specific
splicing; SNP, single-nucleotide polymorphism; BDGP, Berkeley Drosophila Genome Project; NG2, NetGene2; HSF, Human Splicing Finder; ESE, exonic splicing
enhancer; RESCUE-ESE, RESCUE ESE hexamer; PESE, putative 8-mer ESE; PESS,
putative 8-mer exonic splicing silencer; EIE, exon-identity element; IIE, intronidentity element; FAM, carboxyfluorescein.
1
Copyright (C) 2009 by The American Association for Clinical Chemistry
occurring BRCA1 splicing isoforms when assessing the
pathogenicity of BRCA1 UVs affecting splicing.
Materials and Methods
PARTICIPANT RECRUITMENT
Index cases of HB(O)C from 685 families were identified by the Familial Cancer Clinic of our hospital during the period 1998 –2009. HB(O)C families fulfilled at
least one of the following inclusion criteria: 3 breast
and/or ovarian cancer cases diagnosed in 2 generations
of the same parental branch, 2 breast and/or ovarian
cancer cases diagnosed at ⬍50 years of age in the same
parental branch, or 1 breast cancer case diagnosed at
⬍35 years of age.
We recruited 450 healthy female blood donors visiting the blood bank of our hospital during the period
2002–2009. None of the blood donors reported any
history of cancer in first-degree relatives.
Sporadic breast cancer cases (n ⫽ 326) were identified by the Oncology Department of our hospital during the period 2002–2008. All female breast cancer
cases classified as sporadic were eligible for the present
study. We classified breast cancer cases as sporadic if no
first-degree relatives with breast cancer had been
reported.
All participants provided written informed consent. The institutional ethics committee approved the
study.
BRCA1 GENOTYPING
Genomic DNA was extracted from peripheral blood
lymphocytes in a fully automated nucleic acid–
purification workstation (MagNA Pure Compact System; Roche Diagnostics). MagNA Pure Compact DNA
Isolation Kits were used according to the manufacturer’s protocol. Index cases from HB(O)C families were
scanned for the presence of BRCA1 germ-line mutations by denaturing gradient gel electrophoresis and
sequencing, as previously described (16 ). We used denaturing gradient gel electrophoresis to genotype
c.591C⬎T (p.197⫽), c.5123C⬎A (p.A1708E), and 14
BRCA1 tagging single-nucleotide polymorphisms
(SNPs) in sporadic breast cancer cases and in healthy
controls. BRCA1 tagging SNPs have been described
previously (17 ). Human Genome Variation Society
guidelines (http://www.hgvs.org) were used for BRCA1
mutation nomenclature.
IN SILICO SPLICING ANALYSIS
The following splice-site analysis tools were used: (a)
Berkeley Drosophila Genome Project (BDGP) (http://
www.fruitfly.org/seq_tools/splice.html); (b) NetGene2
(NG2) (http://www.cbs.dtu.dk/services/NetGene2);
2
Clinical Chemistry 56:1 (2010)
and (c) Human Splicing Finder (HSF) (http://www.
umd.be/HSF/).
Human default parameter settings were used in all
analyses. BDGP, NG2, and HSF matrices were used to
analyze the effect of the c.591C⬎T variant on donor
site efficiency. In addition, the following HSF matrices
were used to analyze the effect of the c.591C⬎T variant
on the following putative splicing regulatory sequences: HSF integrated matrices for serine/arginine-rich
proteins (SRp40, SC35, SF2/ASF, SF2/ASF IgM/
BRCA1, and SRp55), exonic splicing enhancer (ESE)
motifs from HSF (ESE-HSFs), RESCUE ESE hexamers
(RESCUE-ESE), putative 8-mer ESEs (PESEs) and putative 8-mer exonic splicing silencers (PESSs), exonidentity and intron-identity elements (EIEs and IIEs),
heterogeneous nuclear ribonucleoprotein– binding
motifs, and Fas exonic splicing silencers. A comprehensive description of these splicing regulatory elements can be found elsewhere (18 ).
The analysis was performed with a 400-bp sequence
centered on the c.591C⬎T variant. In accordance with
other studies (19 ), a scoring decrease ⬎10% was considered predictive of splicing alterations.
cDNA SYNTHESIS
We analyzed 7 c.591C⬎T carriers (3 index cases, 2 relatives, and 2 healthy controls) and 10 noncarriers (7
healthy controls, 2 index cases carrying BRCA1 diseasecausing mutation c.421-1G⬎A, and 1 index case carrying BRCA1 disease-causing mutation p.L974X). Total
RNA was extracted from peripheral blood lymphocytes
in a MagNA Pure Compact workstation. MagNA Pure
Compact RNA Isolation Kits were used according to
the manufacturer’s protocol. First-strand cDNA was
synthesized with random hexamers (Promega) and a
mixture of reverse transcriptases from avian myeloblastosis virus and Moloney murine leukemia virus
(Promega).
PCR AMPLIFICATION OF BRCA1 SPLICING ISOFORMS
To amplify different BRCA1 splicing isoforms, we designed primers with the following sequences (5⬘ to 3⬘):
CATCCAAAGTATGGGCTACAGA (exon7fw), TGTC
CAACTCTCTAACCTTG (exon8fw), ACATTGAATT
GGCTGCTTGTG (exon8-11Afw), GTCTACATTG
AATTGGTGTGGGAG (exon8-9fw), CCAGGGATG
AAATCAGTTTGGA (exon10fw), GAATCCAAACT
GATTTCATCC (exon10rev), TGGCTCCACATGCAA
GTTTGT (exon11Brev), CGTCTCTGAAGACTGC
TCA (exon12fw), CTGAGAGGATAGCCCTGAGCA
(exon12rev), and TGGAAGGGTAGCTGTTAGAAGG
(exon12fw). All primers were designed with the open
source software, Primer3 (20 ). Reverse primers were
labeled with 6-carboxyfluorescein (FAM). PCRs were
performed with FastStart Taq DNA polymerase (Roche
Characterization of Splice Site Variants
Table 1. Splice-site prediction analysis of the BRCA1 variant c.591C>T.a
a
b
Wild type
Score
Variant
Score
Variation, %
CAGgtgagt
98.84
TAGgtgagt
96.86
⫺2
CAGgtgagt
10.67
TAGgtgagt
8.83
⫺17.2
CAGgtgagt
1.00
TAGgtgagt
0.99
CAGgtgagt
0.81
TAGgtgagt
0.67
Method
HSF matrices
⫺0.01
⫺17.2
MESb
BDGP
NG2
The 3⬘ end of exon 9 is displayed in capital letters. The polymorphic position is displayed in boldface.
MES, MaxEnt scoring.
Diagnostics). PCR was conducted under the following
cycling conditions, irrespective of the primers used: denaturation at 94 °C for 4 min; 25 cycles of 94 °C for 30 s,
55 °C for 30 s, and 72 °C for 1 min; and extension at
72 °C for 10 min. The number of PCR cycles was kept
low to approach linearity of the relative probe signal
with the target sequence. The reaction volume was 25
␮L in all cases.
CAPILLARY ELECTROPHORESIS
Two microliters of the amplification product was diluted to 15␮L with formamide (Applied Biosystems),
and 0.2 ␮L of 500 LIZ fluorescent size standard (Applied Biosystems) was added to each sample. The mixture was denatured at 95 °C for 2 min and cooled to
4 °C. Amplified PCR products were separated with an
ABI 3130 Genetic Analyzer (Applied Biosystems) with
polymer POP-7 (Applied Biosystems). Size calling was
performed with Genotyper software v4.0 (Applied Biosystems). The same software was used to quantify peak
areas. The BRCA1 ⌬9 normalized signal ratio was calculated as the (⌬9 peak area)/(⌬9 peak area ⫹ fulllength isoform peak area), as measured after PCR amplification with primers exon8fw and exon10rev.
Similarly, the BRCA1 ⌬(9,10) normalized signal ratio
was calculated as [⌬(9,10) peak area]/[⌬(9,10) peak
area ⫹ full-length isoform peak area], as measured after PCR amplification with primers exon7fw and
exon11Brev. Finally, the BRCA1 ⌬(9,10,11B) normalized signal ratio was calculated as [⌬(9,10,11B) peak
area]/[⌬(9,10,11B) peak area ⫹ ⌬11B peak area], as
measured after PCR amplification with primers
exon8fw and exon12rev. Normalized signal ratios were
assessed in triplicate in all samples. Mean normalized
signal ratios were compared with the Student t-test.
Statistical analysis was carried out with the aid of MedCalc software (MedCalc Software).
SEQUENCING
Samples were sequenced with the BigDye Terminator
Cycle Sequencing Kit (Applied Biosystems) according
to the manufacturer’s protocol.
Results
To date, we have performed BRCA1 mutation scanning
in a cohort of 685 index cases of HB(O)C families from
Spain. The BRCA1 c.591C⬎T variant was identified in
10 index cases (1.5%). In addition, we have detected
the variant in 4 of 450 healthy controls (0.9%) and in 1
of 326 (0.3%) cases of sporadic breast cancer.
BRCA1 alleles are described almost completely by
10 canonical haplotypes defined by 14 tagging SNPs
(17 ); however, haplotypes 1 and 2 are the commonest
BRCA1 haplotypes (at least in Western populations),
accounting for roughly 78% of all chromosomes (17 ).
In our cohort, the c.591C⬎T variant was associated
only with haplotype 1 (data not shown); however, most
haplotype 1 alleles do not contain this variant. Interestingly, the disease-causing mutation c.5123C⬎A was
detected only in index cases carrying the c.591C⬎T variant. Segregation analysis confirmed that c.5123C⬎A was
linked to c.591C⬎T, at least in our population.
The c.591C⬎T variant is located at position ⫺3 of
the intron 9 donor splice site. We used 4 splice site–
analysis tools to predict the effect on splicing, with conflicting results (Table 1). The donor splice site scoring
was either decreased (MaxEnt scoring and NG2) or
preserved (HSF matrices and BDGP). In addition, the
c.591C⬎T variant was predicted to affect potential
splicing regulatory sequences located at the 3⬘ end of
exon 9. Interestingly, HSF predicted both the disruption of splicing enhancers and the creation of splicing
silencers (Table 2). RESCUE-ESE, PESE, EIE, and ESEHSF matrices found no differences between variant
and wild-type sequences (data not shown).
To investigate a possible effect of the c.591C⬎T
variant on splicing, we analyzed cDNA samples derived
from 7 carriers and 10 noncarriers. After cDNA amplification with primers exon8fw and exon10rev, we observed the expected 201-bp product corresponding to
the full-length BRCA1 transcript in all tested samples.
An additional product of 155 bp was observed only in
c.591C⬎T carriers (Fig. 1A). Because BRCA1 exon 9
spans 46 nucleotides, the data were compatible with
Clinical Chemistry 56:1 (2010)
3
Table 2. Predicted effect of BRCA1 variant c.591C>T on splicing regulatory motifs.a
Role
a
b
Motif
Wild type
Score
Variant
Score
Variation, %
SE
SRp40
TTGCAGg
82.51
TTGTAGg
0
⫺100 (site broken)
SE
SRp55
TGCAGg
75.65
TGTAGg
0
⫺100 (site broken)
SE
SF2/ASF
TGCAGgt
72.77
TGTAGgt
0
⫺100 (site broken)
SE
SF2/ASF
TGCAGgt
74.61
TGTAGgt
0
⫺100 (site broken)
SS
Fas-ESS
Absent
NA
TGTAGg
NA
New site
SS
Fas-ESS
Absent
NA
GTAGgt
NA
New site
SS
PESS
Absent
NA
TTATTGTA
NA
New site
SS
PESS
Absent
NA
TATTGTAG
NA
New site
SS
IIEs
Absent
NA
ATTGTA
NA
New site
SS
IIEs
Absent
NA
TGTAGg
NA
New site
SS
IIEs
Absent
NA
GTAGgt
NA
New site
SS
hnRNP
CAGgtg
70.24
TAGgtg
85.48
⫹21.69
b
The 3⬘ end of exon 9 is displayed in capital letters. The polymorphic position is displayed in boldface.
SE, splicing enhancer; SS, splicing silencer; NA, not applicable; FAS-ESS, Fas exonic splicing silencer; hnRNP, heterogeneous nuclear ribonucleoprotein. SRp40,
SRp55, and SF2/ASF are serine/arginine-rich proteins.
exon 9 skipping. The BRCA1 ⌬9 signal was consistently
lower than the full-length signal in all carriers (Fig. 1A).
The mean (SD) normalized signal ratio was 0.22 (0.05).
Exon 9 skipping was confirmed by direct sequencing
(Fig. 1B). To assess any contribution of the variant allele to the population of full-length transcripts, we sequenced PCR products obtained with a forward primer
designed to anneal across the exon 8 – exon 9 junction
(exon8-9fw). The chromatogram showed that fulllength transcripts arise predominantly, but not exclusively, from the wild-type allele (Fig. 1C). The data suggest that the c.591C⬎T variant decreases the efficiency
of, but does not abolish, the intron 9 donor splice site.
BRCA1 splicing variants ⌬(9,10) and ⌬(9,10,11B)
do not require a functional intron 9 donor site for
proper splicing, at least in theory; however, we were
interested in investigating any effect of the c.591C⬎T
variant on these isoforms and so conducted additional
experiments.
After cDNA amplification with the primers
exon7fw and exon11Brev, we observed double peaks of
446/449 bp and 323/326 bp that corresponded to fulllength and ⌬(9,10) isoforms, respectively, in all investigated samples (Fig. 2). In c.591C⬎T carriers, we also
observed an extra double peak of 400/403 bp (compatible with exon 9 skipping) (Fig. 2). Sequencing (data
not shown) confirmed that the double peaks corresponded to wobble splicing occurring at the 5⬘ end of
BRCA1 exon 8 (15 ).
cDNA amplification with primers exon8fw and
exon12rev generated 2 PCR products of 396 bp and 273
bp, corresponding to the ⌬11B and ⌬(9,10,11B) iso4
Clinical Chemistry 56:1 (2010)
forms, respectively (Fig. 2). The products corresponding to the full-length and ⌬(9,10) isoforms were too
large to be amplified under the same conditions. An
extra product of 350 bp was generated only in
c.591C⬎T carriers, indicating that exon 9 skipping also
occurs in the ⌬11B isoform (Fig. 2).
Interestingly, the mean normalized signal ratios of
the ⌬(9,10) and ⌬(9,10,11B) isoforms were higher in
c.591C⬎T carriers. These findings were statistically
significant (Fig. 2). The data suggested that c.591C⬎T
modifies the relative expression level of ⌬(9,10),
⌬(9,10,11B), ⌬11B, and full-length isoforms.
To demonstrate ASS associated with the variant
allele, we selected 4 samples (2 c.591C⬎T carriers and 2
noncarrier healthy controls) heterozygous for BRCA1
haplotypes 1 and 2 and amplified cDNAs with 3 different forward primers (exon12fw; exon8-11Afw, a
primer designed to anneal across the exon 8 – exon 11A
junction; and exon10fw) in combination with 1 reverse
primer (exon13rev). Use of primers exon8-11Afw and
exon10fw individually in combination with primer
exon13rev amplified products specific for ⌬(9,10,11B)
and ⌬(11B), respectively. All PCR products were sequenced with primer exon12fw. We used the BRCA1
tagging SNP rs1060915 located in exon 13
(c.4308T⬎C) to distinguish the expression of T (haplotype 1) and C (haplotype 2) alleles. Sequencing of the
PCR fragment encompassing exons 12 and 13 indicated similar expression of the 2 alleles, irrespective of
c.591C⬎T status (Fig. 3). With the exception of
BRCA1-IRIS (21 ), both exon 12 and exon 13 are
present in all BRCA1 isoforms that have been described
Characterization of Splice Site Variants
Fig. 1. Detection of BRCA1 exon 9 skipping in c.591C⬎T carriers.
(A), Representative examples of cDNA amplification with BRCA1 primers exon8fw and exon10rev. (B), Exon 9 skipping
demonstrated by direct sequencing. (C), Sequencing a PCR product generated with a forward primer annealing across the exon
8 – exon 9 junction demonstrated that the variant allele produces full-length transcripts.
Clinical Chemistry 56:1 (2010)
5
Fig. 2. BRCA1 ⌬(9,10) and ⌬(9,10,11B) normalized signal ratios increase in c.591C>T carriers.
Shown are representative examples of cDNA amplification with BRCA1 primers exon7fw and exon11Brev, and with BRCA1
primers exon8fw and exon12rev. The mean normalized signal ratios (⫾2 SD) are plotted.
so far (15 ). Therefore, the data indicated that the expression levels of wild-type and variant alleles were
roughly equivalent. As expected, sequencing of
⌬(9,10,11B) and ⌬11B isoforms showed similar contributions from the 2 alleles in healthy control samples
(Fig. 3). In c.591C⬎T carriers, sequencing of
⌬(9,10,11B) products revealed a predominant contribution of the variant (haplotype 1) allele (Fig. 3). On
the contrary, sequencing of ⌬11B products revealed a
predominant contribution of the wild-type (haplotype
2) allele (Fig. 3). Taken together, the data indicated that
the population of messengers arising from the variant
allele was characterized by a substantial reduction of
full-length and ⌬11B isoforms that was compensated
by an increase in the population of ⌬(9,10) and
⌬(9,10,11B) isoforms, demonstrating ASS.
6
Clinical Chemistry 56:1 (2010)
Discussion
In this study, we have shown that the BRCA1
c.591C⬎T variant is associated with ASS. Several
groups have considered that this variant is most likely a
neutral polymorphism (22–24 ). The variant is recorded 31 times in the BIC database (Cys197Cys according to the BIC traditional nomenclature) as a silent
change of no clinical importance (2 ). Our study further supports a nonpathogenic role for this change.
First, the variant is present in 0.9% of the healthy
control population. Second, the expression levels of
variant and wild-type alleles are similar. Lastly,
c.5123C⬎A, a known BRCA1 disease-causing mutation (25 ), is observed in cis with c.591C⬎T in HB(O)C
families.
Characterization of Splice Site Variants
Fig. 3. BRCA1 exon 13 sequencing in cDNAs derived from c.591C⬎T carriers and controls (noncarriers).
The arrow indicates the location of tagging SNP rs1060915. The c.591C⬎T variant is linked to the rs1060915 T allele.
To our knowledge, this report is the first of nonpathogenic ASS occurring in the BRCA1 gene. The ASS
is characterized by a reduced population of full-length
and ⌬11B transcripts, which is counterbalanced by an
increased population of ⌬(9,10) and ⌬(9,10,11B) transcripts such that the overall expression level is similar
to that of the wild-type allele. In addition, ASS is characterized by the presence of transcripts lacking exon 9.
A schematic representation of c.591C⬎T–associated
ASS is shown in Fig. 4.
RNA-splicing analysis is essential for the assessment of many BRCA1 UVs (8, 12, 19, 26, 27 ). In most
Fig. 4. Proposed model for BRCA1c.591C⬎T allele-specific splicing.
We have represented the proposed splicing events occurring in wild-type (black) and variant (gray) alleles. Solid lines indicate
major events; dashed lines indicate minor events.
Clinical Chemistry 56:1 (2010)
7
cases, analysis is performed by amplifying cDNA with
primers flanking the exon likely to be affected by the
variant. This approach is designed to detect singlecassette exon events. Interestingly, it will produce misleading results if it is used to test the effect of the BRCA1
c.591C⬎T variant on splicing. This is the case because
cDNA amplification with primers located in exons
8 –10 (flanking exons) restricts the analysis to the fulllength isoform, thus detecting exon 9 skipping (a
frameshift alteration) but missing an increase in the
skipping of exons 9 –10 (a naturally occurring BRCA1
isoform preserving the open reading frame). The latter
probably explains why the BRCA1 c.591C⬎T variant
does not have a pathogenic effect [to be more precise, it
does not cause an autosomal dominant HB(O)C syndrome], although a role as an allele with a lowpenetrance risk cannot be fully discarded. Moreover, it
is important to point out that ASS has been characterized in blood cells and does not necessarily reflect the
situation in other tissues, such as the mammary and
ovary epithelia.
Our study was not intended to investigate the molecular mechanisms responsible for the association between c.591C⬎T and ASS, so we can only speculate. In
our opinion, the mere localization of the variant points
to a reduced efficiency of the intron 9 splicing donor
site as a likely causative mechanism, but other mechanisms are possible. For instance, the variant may target
some splicing regulatory sequences. Both hypotheses
find some support from in silico analysis (Tables 1 and
2). Regardless of the true mechanisms involved, we
propose that inclusion/skipping of exons 9 –10 are
competing splicing reactions, so that hampering one of
the reactions shifts the process to the alternative reaction. In this scenario, a sequence change slightly disrupting any splicing site (or splicing regulatory) sequence is more likely to affect the outcome of the
splicing reaction.
At least 33 different BRCA1 variants recorded in
the BIC database (2 ) are located in the genomic region
encompassing exons 9 and 10. Of these variants, only
4 are considered disease-causing mutations: 3 exonic
variants that have been reported only once
(c.625_626ins20, c.668_669insA, c.668-669delA) and 1
intronic variant reported 11 times (c.594-2A⬎C). The
pathogenic nature of the latter is supported by experimental data demonstrating exon 10 skipping (12 );
however, we suggest that the relevant outcome of intron 9 acceptor site deficiency could be (as is the case
for intron 9 donor site deficiency) an increased population of ⌬(9,10) transcripts, a possibility that has not
yet been investigated.
Our study highlights the fact that splice variants
other than the full-length isoform should be considered when analyzing the effect of certain BRCA1 variants (or variants in any other tumor suppressor gene)
on splicing. This consideration is important because
restricting the in vitro analysis to single-cassette events
(namely, select primers for splicing analysis in flanking
exons) may introduce bias that affects the interpretation of biological relevance. In addition, our study suggests that the false-negative rate of splicing-prediction
programs may increase if alternative splicing variants
(naturally occurring alternative splicing pathways) do
exist.
Author Contributions: All authors confirmed they have contributed to
the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design,
acquisition of data, or analysis and interpretation of data; (b) drafting
or revising the article for intellectual content; and (c) final approval of
the published article.
Authors’ Disclosures of Potential Conflicts of Interest: Upon
manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: None declared.
Consultant or Advisory Role: None declared.
Stock Ownership: None declared.
Honoraria: None declared.
Research Funding: Research grant FIS 05/0864 ISCIII, Spanish Ministry of Science and Innovation; research grant RD 06/0020/0021
from RTICC, ISCIII, Spanish Ministry of Science and Innovation;
research grants FMMA-06 and FMMA-08, Fundación Mutua
Madrileña.
Expert Testimony: None declared.
Role of Sponsor: The funding organizations played no role in the
design of study, choice of enrolled patients, review and interpretation
of data, or preparation or approval of manuscript.
Acknowledgments: We thank the patients and their families for their
cooperation.
References
1. Goldgar DE, Easton DF, Deffenbaugh AM, Monteiro AN, Tavtigian SV, Couch FJ, Breast Cancer
Information Core (BIC) Steering Committee. Integrated evaluation of DNA sequence variants of
unknown clinical significance: application to
BRCA1 and BRCA2. Am J Hum Genet 2004;75:
535– 44.
2. The Breast Cancer Information Core (BIC) Database. National Human Genome Research Insti-
8
Clinical Chemistry 56:1 (2010)
tute. National Institutes of Health. http://
research.nhgri.nih.gov/bic/ (Accessed October
2009).
3. Vink GR, van Asperen CJ, Devilee P, Breuning
MH, Bakker E. Unclassified variants in diseasecausing genes: nonuniformity of genetic testing
and counselling, a proposal for guidelines. Eur J
Hum Genet 2005;13:525–7.
4. Petrij-Bosch A, Peelen T, van Vliet M, van Eijk R,
Olmer R, Drüsedau M, et al. BRCA1 genomic
deletions are major founder mutations in Dutch
breast cancer patients. Nat Genet 1997;17:
341–5.
5. Scholl T, Pyne MT, Russo D, Ward BE. BRCA1
IVS16⫹6T⬎C is a deleterious mutation that creates an aberrant transcript by activating a cryptic
splice donor site. Am J Med Genet 1999;85:
113– 6.
Characterization of Splice Site Variants
6. Laskie Ostrow K, DiCioccio RA, McGuire V, Whittemore AS. A BRCA1 variant, IVS23⫹1G3 A,
causes abnormal RNA splicing by deleting exon
23. Cancer Genet Cytogenet 2001;127:188 –90.
7. Claes K, Poppe B, Machackova E, Coene I, Foretova L, De Paepe A, Messiaen L. Differentiating
pathogenic mutations from polymorphic alterations in the splice sites of BRCA1 and BRCA2.
Genes Chromosomes Cancer 2003;37:314 –20.
8. Campos B, Dı́ez O, Domènech M, Baena M, Balmaña J, Sanz J, et al. RNA analysis of eight
BRCA1 and BRCA2 unclassified variants identified
in breast/ovarian cancer families from Spain. Hum
Mutat 2003;22:337.
9. Keaton JC, Nielsen DR, Hendrickson BC, Pyne MT,
Scheuer L, Ward BE, et al. A biochemical analysis
demonstrates that the BRCA1 intronic variant
IVS10 –2A3 C is a mutation. J Hum Genet 2003;
48:399 – 403.
10. Brose MS, Volpe P, Paul K, Stopfer JE, Colligon
TA, Calzone KA, Weber BL. Characterization of
two novel BRCA1 germ-line mutations involving
splice donor sites. Genet Test 2004;8:133– 8.
11. Sharp A, Pichert G, Lucassen A, Eccles D. RNA
analysis reveals splicing mutations and loss of
expression defects in MLH1 and BRCA1. Hum
Mutat 2004;24:272.
12. Tesoriero AA, Wong EM, Jenkins MA, Hopper JL,
Brown MA, Chenevix-Trench G, et al. Molecular
characterization and cancer risk associated with
BRCA1 and BRCA2 splice site variants identified
in multiple-case breast cancer families. Hum Mutat 2005;26:495.
13. Chen X, Truong TT, Weaver J, Bove BA, Cattie K,
Armstrong BA, et al. Intronic alterations in
BRCA1 and BRCA2: effect on mRNA splicing fidelity and expression. Hum Mutat 2006;27:427–
35.
14. Bonatti F, Pepe C, Tancredi M, Lombardi G,
Aretini P, Sensi E, et al. RNA-based analysis of
BRCA1 and BRCA2 gene alterations. Cancer
Genet Cytogenet 2006;170:93–101.
15. Orban TI, Olah E. Emerging roles of BRCA1 alternative splicing. Mol Pathol 2003;56:191–7.
16. de la Hoya M, Pérez-Segura P, Van Orsouw N,
Dı́az-Rubio E, Caldés T. Spanish family study on
hereditary breast and/or ovarian cancer: analysis of
the BRCA1 gene. Int J Cancer 2001;91:137– 40.
17. Judkins T, Hendrickson BC, Deffenbaugh AM,
Eliason K, Leclair B, Norton MJ, et al. Application
of embryonic lethal or other obvious phenotypes
to characterize the clinical significance of genetic
variants found in trans with known deleterious
mutations. Cancer Res 2005;65:10096 –103.
18. Desmet FO, Hamroun D, Lalande M, CollodBéroud G, Claustres M, Béroud C. Human Splicing
Finder: an online bioinformatics tool to predict
splicing signals. Nucleic Acids Res 2009;37:e67.
19. Vreeswijk MP, Kraan JN, van der Klift HM, Vink
GR, Cornelisse CJ, Wijnen JT, et al. Intronic variants in BRCA1 and BRCA2 that affect RNA splicing can be reliably selected by splice-site prediction programs. Hum Mutat 2009;30:107–14.
20. Rozen S, Skaletsky H. Primer3 on the WWW for
general users and for biologist programmers.
Methods Mol Biol 2000;132:365– 86.
21. ElShamy WM, Livingston DM. Identification of
BRCA1-IRIS, a BRCA1 locus product. Nat Cell Biol
2004;6:954 – 67.
22. Takahashi H, Behbakht K, McGovern PE, Chiu HC,
Couch FJ, Weber BL, et al. Mutation analysis of
the BRCA1 gene in ovarian cancers. Cancer Res
1995;55:2998 –3002.
23. Durocher F, Tonin P, Shattuck-Eidens D, Skolnick
M, Narod SA, Simard J. Mutation analysis of the
BRCA1 gene in 23 families with cases of cancer of
the breast, ovary, and multiple other sites. J Med
Genet 1996;33:814 –9.
24. Dı́ez O, Osorio A, Durán M, Martinez-Ferrandis JI,
de la Hoya M, Salazar R, et al. Analysis of BRCA1
and BRCA2 genes in Spanish breast/ovarian cancer patients: a high proportion of mutations
unique to Spain and evidence of founder effects.
Hum Mutat 2003;22:301–12.
25. Osorio A, Milne RL, Honrado E, Barroso A, Diez O,
Salazar R, et al. Classification of missense variants of unknown significance in BRCA1 based on
clinical and tumor information. Hum Mutat 2007;
28:477– 85.
26. Vega A, Campos B, Bressac-De-Paillerets B, Bond
PM, Janin N, Douglas FS, et al. The R71G BRCA1
is a founder Spanish mutation and leads to aberrant splicing of the transcript. Hum Mutat 2001;
17:520 –1.
27. Anczuków O, Buisson M, Salles MJ, Triboulet S,
Longy M, Lidereau R, et al. Unclassified variants
identified in BRCA1 exon 11: consequences on
splicing. Genes Chromosomes Cancer 2008;47:
418 –26.
Clinical Chemistry 56:1 (2010)
9