Screening BRCA1 and BRCA2 unclassified variants for splicing mutations using reverse transcription PCR on patient RNA and an ex vivo assay based on a splicing reporter minigene

Abstract

Many unclassified variants (UV) of BRCA1 or BRCA2 may have an effect on pre-mRNA splicing. Patient blood samples suitable for RNA extraction are not always available for testing UVs at the RNA level.
Analyses of RNA from patient peripheral blood were performed, using a one-step reverse transcriptase-PCR (RT-PCR) protocol, and were compared with an ex vivo splicing assay based on PCR-amplified patient DNA inserted into a splicing reporter minigene. Using both methods 20 variants found in 17 patients were examined.
Data from patient RNA and from the minigene assay were fully concordant, but the ex vivo splicing assay, which is monoallelic, clarified several ambiguities in the patient RNA data. Two intronic variants induced strong splicing defects: BRCA1 c.4987-5T-->A (IVS16-5T-->A) induced exon 17 skipping and BRCA2 c.316+5G-->C (IVS3+5G-->C) induced complete skipping of exon 3. Of the exonic variants, BRCA2 c.7805G-->C (p.Arg2602Thr), at the last base of exon 16, induced both exon skipping and activation of a cryptic exonic donor site, and BRCA2 c.8023A-->G (p.Ile2675Val) generated a strong donor site within exon 18. These four variants were thus classified as pathogenic, because of the total absence of a normal transcript from the corresponding allele. Variant BRCA2 c.9501+3A-->T (IVS25+3A-->T) induced incomplete skipping of exon 25, suggesting a mutation with incomplete penetrance, and BRCA2 c.8257_8259del (p.Leu2753del) modified the alternative splicing of exons 17 and 18.
We show that functional analysis using a splicing reporter minigene is sensitive and specific, and should be used for initial screening of potential splicing defects, especially when patient RNA is not readily available.

Full text

Screening BRCA1 and BRCA2 unclassified variants
for splicing mutations using reverse transcription
PCR on patient RNA and an ex vivo assay based on a
splicing reporter minigene
C Bonnet,
1
S Krieger,
2
M Vezain,
1
A Rousselin,
2
I Tournier,
1
A Martins,
1
P Berthet,
3
A Chevrier,
6
C Dugast,
4
V Layet,
5
A Rossi,
6
R Lidereau,
7
T Fre´bourg,
1,6
A Hardouin,
2
M Tosi
1
cAdditional tables are
published online only at http://
jmg.bmj.com/content/vol45/
issue7
1
Inserm U614, IFRMP, Faculty
of Medicine and Department of
Genetics, University Hospital,
Institute for Biomedical
Research, Rouen, France;
2
Laboratoire de Biologie Clinique
et Oncologique, Centre Franc¸ois
Baclesse, Caen, France;
3
Consultation d’Oncoge´ne´tique,
Centre Franc¸ois Baclesse, Caen,
France;
4
Consultation
d’Oncoge´ne´tique, Centre Euge`ne
Marquis et CHU, Rennes,
France;
5
Unite´deGe´ne´tique,
Hoˆpital Flaubert, Le Havre,
France;
6
Unite´deGe´ne´tique
Clinique, CHU de Rouen, France;
7
Centre Rene´ Huguenin,
FNCLCC, Inserm U735, St-Cloud,
France
Correspondence to:
Dr M Tosi, Inserm U614, IFRMP,
Faculty of Medicine, Institute for
Biomedical Research and French
Northwest Canceropole, 22
boulevard Gambetta, 76183
Rouen, France;
mario.tosi@univ-rouen.fr
CB and SK contributed equally to
this work.
Received 18 December 2007
Revised 12 March 2008
Accepted 19 March 2008
Published Online First
18 April 2008
ABSTRACT
Background: Many unclassified variants (UV) of BRCA1
or BRCA2 may have an effect on pre-mRNA splicing.
Patient blood samples suitable for RNA extraction are not
always available for testing UVs at the RNA level.
Methods: Analyses of RNA from patient peripheral blood
were performed, using a one-step reverse transcriptase-
PCR (RT-PCR) protocol, and were compared with an ex
vivo splicing assay based on PCR-amplified patient DNA
inserted into a splicing reporter minigene. Using both
methods 20 variants found in 17 patients were examined.
Results: Data from patient RNA and from the minigene
assay were fully concordant, but the ex vivo splicing
assay, which is monoallelic, clarified several ambiguities in
the patient RNA data. Two intronic variants induced
strong splicing defects: BRCA1 c.4987-5TRA (IVS16-
5TRA) induced exon 17 skipping and BRCA2
c.316+5GRC (IVS3+5GRC) induced complete skipping
of exon 3. Of the exonic variants, BRCA2 c.7805GRC
(p.Arg2602Thr), at the last base of exon 16, induced both
exon skipping and activation of a cryptic exonic donor site,
and BRCA2 c.8023ARG (p.Ile2675Val) generated a
strong donor site within exon 18. These four variants were
thus classified as pathogenic, because of the total
absence of a normal transcript from the corresponding
allele. Variant BRCA2 c.9501+3ART (IVS25+3ART)
induced incomplete skipping of exon 25, suggesting a
mutation with incomplete penetrance, and BRCA2
c.8257_8259del (p.Leu2753del) modified the alternative
splicing of exons 17 and 18.
Conclusions: We show that functional analysis using a
splicing reporter minigene is sensitive and specific, and
should be used for initial screening of potential splicing
defects, especially when patient RNA is not
readily available.
A major difficulty in the molecular diagnosis of
breast and ovarian cancer predisposition results
from the large number of variants of unknown
biological and clinical significance found in the
BRCA1 and the BRCA2 genes and denoted here as
UVs (unclassified variants). These include the
majority of missense mutations and non-coding
nucleotide changes, especially intronic variants that
affect degenerate positions of the splice sites—that
is, those distinct from the conserved dinucleotides
GT and AG at the 59and 39end of introns,
respectively. Our current inability to determine the
consequences of these variants precludes optimum
risk assessment for a large number of patients. For
missense substitutions, although assays focused on
particular protein domains are usually unable to
reflect the functional complexity of these proteins,
statistical analyses can be used to improve the
classification of variants.
12
However, exonic var-
iants (coding or silent) quite often induce splicing
alterations, because they alter splice site sequences
or affect internal exonic elements involved in
regulation of splicing, such as ESEs (exonic splicing
enhancers) or ESS (exonic splicing silencers). Both
are short (6–8 bp) degenerate sequences that are
recognised by splicing factors that either favour or
inhibit recognition of suboptimum splice sites.
34
Analyses at the mRNA level are essential to detect
UVs that induce splicing alterations, but such
analyses are often hampered by at least three major
difficulties. First, it is not easy to collect on a routine
basis blood samples suitable for RNA extraction
from large numbers of probands carrying UVs.
Second, the large BRCA1 or BRCA2 mRNAs undergo
several alternative splicing events. Third, the major-
ity of splicing alterations induce frameshifts and
activate the nonsense-mediated decay (NMD) sys-
tem for mRNA, which in most cases cannot be
readily prevented when collecting blood samples.
Functional analyses of splicing using splicing
reporter minigenes have the advantage of relying
only on genomic DNA, from which the relevant
region carrying the variant can be amplified by
PCR and tested in transient cell transfection
assays. Several minigene systems have been
described for the analysis of cis elements affecting
splicing.
45
In this paper, we illustrate the advan-
tage of using improved reverse transcriptase (RT)-
PCR conditions for the analysis of BRCA1 and
BRCA2 mRNA species in patient RNA in combina-
tion with an ex vivo assay based on patient DNA
inserted into a novel splicing reporter minigene,
and we report the sensitivity and specificity of this
splicing reporter minigene assay.
METHODS
All patients and controls gave signed informed
consent.
Patients
The 20 BRCA1 and BRCA2 variants tested in this
work were detected in 17 index cases, selected from
Original article
438 J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

families undergoing oncogenic consultations in the western
French Network. The criteria for diagnostic mutation screening
of BRCA genes were according to the current French recom-
mendations. All families were tested for presence of BRCA1/2
point mutations and/or small deletions/insertions using denatur-
ing high-performance liquid chromatography followed by direct
sequencing. Screening for large genomic rearrangements of
BRCA1 was performed for all probands with a combination of
the multiplex ligation-dependent probe amplification and quan-
titative multiplex PCR of short fluorescent fragments methods
67
and these analyses failed to reveal causative mutations. Nine of
the variants tested here were registered in the BIC breast cancer
database (http://research.nhgri.nih.gov/bic//).
Nomenclature
The DNA mutation numbering is based on the cDNA
sequence for BRCA1 (GenBank U14680) and BRCA2
(GenBank NM_000059). The nomenclature system for the
description of changes in DNA, RNA and protein follows the
recommendations of the Human Genome Variation Society
(HGVS). Intronic variants were also named according to the
BIC database.
Blood samples, nucleic acid extraction and reverse transcriptase
PCR analysis
Patient peripheral blood samples were collected in EDTA tubes.
DNA was isolated using either a manual (QIAamp DNA mini
kit) or automated (EZ1 DNA Blood kit on BioRobot EZ1
workstation) procedure (both Qiagen, Hilden, Germany). For
RNA analysis, whole blood was collected from patients and
voluntary control subjects (PAXgene Blood RNA tubes;
Qiagen). RNA extraction was made according to the manu-
facturer’s protocols including DNAse treatment. RT-PCR was
performed with a commercial kit (One Step RT-PCR kit;
Qiagen), using 500 ng to 1 mg of RNA and a gene-specific primer
for reverse transcription (table 1). All amplifications were
performed using a touch-down PCR program. The thermocycler
parameters for the RT-PCR were 50uC for 30 minutes for the
reverse transcription step, 95uC for 15 minutes for the initial
DNA polymerase activation, 26 cycles of 94uC for 40 seconds,
68uCor63uC (depending on primer as shown in table 1) for
1 minute with a decrease of 0.5uC/cycle, and 72uC for
45 seconds, followed by 26 cycles of: 94uC for 40 seconds,
55uCor50uC (depending on primer) for 20 seconds and 72uC for
45 seconds, and a final extension at 72uC for 10 minutes. Three
controls consisting of blood RNA from individuals not carrying
the UV were amplified in parallel. All experiments were
repeated at least once, each time with three controls, with
qualitatively identical results. Several RNAs from patients
carrying mutations affecting the conserved positions GT or
AG at the 59and 39end, respectively, of introns were used as
positive controls for splicing defects, and exon skipping was
observed in each case. RNA from a patient carrying the known
splicing mutation BRCA1 213-11TRG, which induces intron
inclusion,
8
was also used as a positive control for comparisons
with data obtained in the splicing reporter minigene assay (see
below). RT-PCR products were separated by electrophoresis on a
1.5% agarose gel containing ethidium bromide and visualised by
exposure to ultraviolet light. RT-PCR products were sequenced
directly on an automated sequencer (ABI 3130; Applied
Biosystems, Foster City, California) using a cycle sequencing
reaction kit (Big Dye Terminator kit; Applied Biosystems), or after
excision from the gel and purification with a commercial kit
(NucleoSpin; Machery-Nalgen, Du¨ren, Germany).
Splicing predictions
We compared our results with the outcome of four different
splice-site prediction algorithms: Splice Site Finder (http://violin.
genet.sickkids.on.ca/,ali/splicesitefinder.html), Gene Splicer (http://
www.tigr.org/tdb/GeneSplicer/gene_spl.html), Splice Site Predic-
tion by Neural Network (NNSplice; http://www.fruitfly.
org/seq_tools/splice.html) and MaxEntScan (http://genes.mit.
edu/burgelab/maxent/Xmaxentscan_scoreseq.html). We simulta-
neously investigated Splice Site Finder, MaxEntScan and
GeneSplicer using the software Alamut V.1.3 f (Interactive
Biosofware; www.interactive-biosoftware.com). This software
also allows interrogation of the ESEfinder and the ESErescue
algorithms for analysis of several potential splicing regulatory
elements. In silico predictions are shown in detail in supplemen-
tary tables S1 and S2 (available online).
Ex vivo splicing assay using a splicing reporter minigene
BRCA1 or BRCA2 exons relevant to this study were amplified by
PCR, with 150 bp of 59and 39flanking intronic sequences, using
Table 1 Design of reverse transcription PCRs for BRCA1 and BRCA2 transcripts
Extent of RT-PCR*
Primers 59R3Size
(bp)
Initial
AT (uC)Forward Reverse
BRCA1
Exons 2–8 ATTGGAACAGAAAGAAATGGAT CCAATTCAATGTAGACAGACGT 563 63
Exons 6–11 TACGAGATTTAGTCAACTTG GTATTTGTTACATCCGTCTC 480 63
Exons 15–21 CATTAGATGATAGGTGGTA AAGGGTGAATGATGAAAGC 899 63
Exons 21–24 AGAAATCTGTTGCTATGGGC CAGTAGTGGCTGTGGGGGAT 300 68
BRCA2
59UTR–exon 5 TACTCCGGCCAAAAAAGAAC CTGACTTATCTCTTTGTGG 550 63
Exons 2–5 ACTTATTTACCAAGCATTG CTGACTTATCTCTTTGTGG 514 63
Exons 5–8 CATGTAACACCACAAAGAGA TTCAGATGCTTCTTCATTT 207 63
Exons 7–10 CACCACCCACCCTTAGTTCT ACATTTGGCATTGACTTTCC 271 68
Exons 15–17 TCTGCGTGTTCTCATAAAC TAGCTGCCAGTTTCCATAT 304 63
Exons 17–20 CTAATAGATGCCTAAGCCCAGA TCTTCCTCTCTTTCATTGCGAA 616 68
Exons 16–19 GAGTCTTTTCAGTTTCACAC CAACATTTCCTCCATCACTG 775 63
Exons 21–27 TATGAAGCAGTGAAGAATGC TCTTCCTCTCTTTCATTGCGAA 1340 68
Exons 23–27 ACATACAGTTAGCAGCGACA GCCATACAAAGTGATAAAG 634 63
AT, annealing temperature; UTR, untranslated reading frame.
*Exon 11 of BRCA1 and exon 11 of BRCA2 are not covered.
Original article
J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895 439
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

forward and reverse primers carrying 59tails that contained
restriction sites for BamHI and MluI, respectively. Primer
sequences are available on request. When the region to be
amplified contained one of these restriction sites, we included
into the 59tails of the primers the restriction sites BglII or AscI,
which are compatible with the BamHI and MluI sites,
respectively, present in the splicing reporter minigene pCAS. A
detailed description of this splicing reporter minigene has been
published elsewhere.
9
Minigenes carrying the wild type and the
variant allele were then identified by sequencing and were
Table 2 Effect of variants on RNA splicing
Family Gene Location Nucleotide change{
Presence in
French
database"
Effect predicted
in silico (by
fraction of
algorithms)"
Effect observed in
patient RNA
Effect observed In the ex
vivo splicing assay Interpretation**
1* BRCA1 Intron 2 c.81249ARC* 2 times No effect (4/4) None None Unlikely to be
pathogenic
2BRCA1 Intron 2 c.81265GRC 2 times No effect (4/4) None None Unlikely to be
pathogenic
3BRCA1 Intron 8 c.548217GRT 4 times No effect (4/4) None None Unlikely to be
pathogenic
4BRCA1 Intron 10 c.670+8CRT 2 times New Dss at
c.670+6 (2/4)
Reduced skipping of
exons 9 and 10
Complete inclusion of exon
10
Unlikely to be
pathogenic
5 BRCA1 Intron 16 c.498725TRA 1 time Weaker Ass (4/4)
Cryptic Ass at
IVS16+1 (3/4)
r.4987_5074del
Skipping of exon 17
Skipping of exon 17 but also
replacement of exon 17 with
153 bp of intron 17. Absence
of wild type product
Pathogenic
6BRCA1 Intron 20 c.5277+48_59dup12 39 times (3)1No effect (4/4) None None Unlikely to be
pathogenic
7BRCA1 Intron 22 c.5406+8TRC 5 times (1)1No effect (4/4) None None Unlikely to be
pathogenic
8BRCA2 Intron 3 c.316+5GRC 5 times Weaker Dss (3/4) r.68_316del Skipping
of exon 3 (full
exclusion)
Skipping of exon 3. Absence
of wild type product
Pathogenic
9BRCA2 Intron 6 c.516+14CRT 4 times No effect (4/4) None None Unlikely to be
pathogenic
1* BRCA2 Intron 8 c.681+7ART* Once No effect (4/4) None None Unlikely to be
pathogenic
10 BRCA2 Exon 16 c.7805GRC, last
base (p.Arg2602Thr)
3 times Weaker Dss and
cryptic Dss
activated (3/4)
r.7618_7805del Exon
16 skipping
Exon 16 skipping, but also
partial inclusion of 88 bp of
exon 16. Absence of wild
type product
Pathogenic
11{BRCA2 Exon 18 c.8023ARG{
(p.Ile2675Val)
3 times New Dss (3/4) r.8023_8331del
Generation of new
donor site: inclusion
of only 46
nucleotides of exon
18
Generation of new donor site:
inclusion of only 46
nucleotides of exon 18.
Absence of wild type product
Pathogenic
Intron 17 c.7976+35CRA{5 times c.7976+35CRA not tested
independently
c.7976+35CRA
Unlikely to be
pathogenic
12 BRCA2 Exon 18 c.8162TRG
(p.Leu2721Arg)
3 times No effect (4/4) None None Still unclassified
13 BRCA2 Exon 18 c.8257_8259delCTT
p.Leu2753del
3 times No effect (4/4) Modification of the
alternative splicing of
exons 17 and 18
Partial skipping of exon 18 Still unclassified
14{BRCA2 Intron 18 c.8331+22GRC Once No significant
effect (4/4)
None None Unlikely to be
pathogenic
c.8332254TRG Once No significant
effect (4/4)
No effect of c.8332254TRG
tested independently
Unlikely to be
pathogenic
15 BRCA2 Intron 24 c.92572113TRG 6 times No significant
effect (4/4)
None None Unlikely to be
pathogenic
16{BRCA2 Intron 25 c.9501+3ART{3 times Weaker Dss (4/4) r.9257_9501del Exon 25 skipping. Presence
of wild type product
Biologically
significant change,
suggesting
mutation with
reduced penetrance
Intron 24 c.9257216TRC{47 times (7)1Exon 25 skipping No effect of c.9257216TRC
tested independently
17 BRCA2 Intron 26 c.9648+106delT Once No effect (4/4) None None Unlikely to be
pathogenic
*Proband carries a UV in BRCA1 and a UV in BRCA2.{Two UVs found in the proband in the BRCA2 gene. We did not obtain separate functional information for the BRCA2
c.7976+35CRA variant (fig 3). {Cases (n) with the particular variant deposited in the French BRCA1/2 database (February 2008), including those analysed here. 1Cases (n)
associated with clearly deleterious mutations, but the phase is not known. "In silico predictions were obtained using four different algorithms (see Methods), and detailed
predictions are shown in supplementary tables S1 and S2 online. **Pathogenicity or non-pathogenicity are evaluated here only with regard to splicing; pathogenic effects at a level
different from splicing cannot be excluded.
The fraction of algorithms supporting the global prediction shown in this column is indicated in parentheses.
Original article
440 J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

transfected separately into HeLa cells during the same
experiment. The splicing patterns corresponding to wild type
and to the mutant allele were then compared by RT-PCR
analysis of the RNA from transfected cells, using universal
primers in the exons of the pCAS minigene and by sequencing
all RT-PCR products.
RESULTS
Comparisons of the reverse transcriptase PCR analyses of in vivo
processed RNA with data from the ex vivo splicing assay
We compared various RT-PCR protocols to detect BRCA1 and
BRCA2 transcripts and found that using gene-specific primers in
the RT step in a one-tube RT-PCR produced stronger signals
than using oligo(dT) or random hexamers as primers (data not
shown). We also optimised primer selection in the regions
carrying the 20 UVs (table 1). RNA from 17 index cases carrying
different variants was tested using these optimised RT-PCR
conditions (table 2). DNA from the same 17 index cases was
then tested using the splicing reporter minigene assay. For the
three individuals carrying .1 variant in BRCA2, separate
information could be obtained from each variant in two cases.
The effects induced by each variant in this assay are listed in
table 2 and are compared with the data obtained from the
analysis of patient RNA.
The detailed RNA analysis of each of the variants that
induced alterations of the splicing pattern is illustrated (figs 1–3,
left panels). The previously described
8
splicing mutation BRCA1
213-11TRG was included as a positive control, especially for
the ex vivo splicing assay, because it activates an upstream
intronic cryptic splice site. RT-PCR analyses of patient RNA
(fig 1A) revealed, as expected, the insertion into BRCA1 mRNA
of 59 nucleotides corresponding to the 39end of intron 5, due to
weakening of the normal acceptor site of exon 6 and activation
of a cryptic acceptor site, which is present 59 bp upstream. The
corresponding minigene assay was carried out on a genomic
fragment encompassing exons 6 and 7 and confirmed the
inclusion of the same segment of intron 5 (fig 1B). Moreover,
this monoallelic minigene assay indicated a marked effect of the
variant, because the wild-type band, corresponding to the
normal inclusion of exons 6 and 7, was absent.
Four of the variants tested are located near exon boundaries
and affect degenerate positions of the splice site sequences. For
variant BRCA1 c.4987-5TRA, strong skipping of exon 17 was
observed in both assays (fig 1C,D). The minigene assay not only
confirmed the strong alteration of splicing, but also revealed
another defect, ie, replacement of the entire exon by 153
nucleotides derived from the 59end of intron 17 (fig 1D). The
most likely explanation for the production of this second
aberrant RNA species is the presence, at position c.5074+154 of
intron 17, of a cryptic donor splice site, together with a rather
uncommon characteristic of the 39boundary of exon 17, which
reveals a cryptic acceptor site with a rather high score, in
addition to its natural role of donor site (sequences denoted
AGgt in the schematic). This combination of cryptic splice sites
may be used preferentially in the context of the splicing reporter
minigene (panel D), because we were unable to show the
presence of the corresponding aberrant species in the RNA from
the patient (fig 1C). However, this aberrant transcript is
predicted to undergo degradation by the NMD system, because
of the presence of a premature termination codon within the
stretch of 153 nucleotides shown to replace exon 17 in the
minigene assay.
Variants affecting donor splice-site sequences induced splicing
alterations with variable levels of intensity (fig 2). Complete
BRCA2 exon 3 skipping was found with the variant
c.316+5GRC, as shown in panel A by the monoallelic RT-
PCR performed with a forward primer positioned on the SNP
c.226GRA in exon 2. Complete skipping of exon 3 was also
found in the minigene assay (fig 2B). The BRCA2 c.7805GRC
variant at the last base of exon 16 also induced strong exon
skipping (panel C). The ex vivo splicing analysis confirmed the
almost complete exclusion of this exon, but also revealed an
abundant species in which the first 88 nucleotides of exon 16
were present. Sequence analysis of a very weak species of about
200 bp present in the RT-PCR from patient RNA (fig 2C)
showed that this stretch of 88 bp of exon 16 is also included to
some extent in vivo. The most likely explanation for these
findings is the presence of a cryptic donor site at the
corresponding position of exon 16. In fact, a strong cryptic
donor site is predicted at position c.7706 (table 2), using several
bioinformatics programs.
In contrast to the acceptor-site mutation and the two donor-
site mutations illustrated above, the BRCA2 c.9501+3ART
variant in intron 25 had a moderate effect (fig 2E). The minigene
assay, which is monoallelic, also indicated an incomplete
splicing alteration, as an important fraction of the transcripts
observed with the splicing reporter minigene contained the
normal exon 25 sequence (panel F).
Two variants induced splicing alterations, despite being
located at internal exon positions. The BRCA2 variant
c.8023ARG in exon 18 (fig 3A), found in family 11 together
with a second UV in the same region of BRCA2 (c.7976+35 in
intron 17), was associated with loss from the mRNA of a stretch
of 309 nucleotides at the 39terminal of exon 18. A particularly
abundant RT-PCR product corresponding to this aberrant
species was observed, because the RNA deletion is in frame
and thus does not activate the NMD system. The ex vivo assay
was carried out either in the presence of exon 18 alone, or in the
presence of exons 17 and 18. Inclusion of only a short 59
terminal segment of exon 18 (46 nucleotides) was observed in
both cases and no species with normal inclusion of exon 18 was
found, indicating a major splicing defect. The deletion from the
mRNA of 309 nucleotides of exon 18 can be explained by the
generation, by the ARG variant, of a novel donor splice site
within exon 18, as shown in the schematic.
Finally, the one-codon deletion BRCA2 c.8257_8259del
induced incomplete inclusion of exon 18 and modified a known
alternative splicing pattern
10
involving exons 17 and 18,
suggesting the presence, at this position, of a splicing regulatory
element (fig 3C). This effect was reproduced in the minigene
assay, by using either a genomic DNA fragment covering exon
18 alone or a fragment encompassing both exon 17 and exon 18.
Further characterisation of this potential exonic regulatory
element is under way.
Comparisons of experimental results with in silico predictions
Four different algorithms were used to test if the concordant
results obtained by RT-PCR analysis of patient RNA and by
using the splicing reporter minigene assay could have been
predicted (see the supplementary tables S1 and S2 online). A
summary of these in silico predictions is shown in the sixth
column of table 2, in which the fraction of algorithms
contributing to each prediction is indicated. Only changes of
predicted scores that amounted to at least 10% were considered
significant, whereas a recent work
11
used more conservative
estimates and considered only variations of at least 20% to be
significant. For five of the six UVs, for which we experimentally
showed an effect on splicing, an effect was predicted by at least
Original article
J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895 441
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

three of the four algorithms. In this group, only the modifica-
tion of alternative splicing of BRCA2 exons 17 and 18, induced
by the variant BRCA2 c.8257_8259delCTT could not be
predicted. For the 14 variants tested with no experimental
evidence of a splicing defect, only minor variations of the
predicted scores were observed (supplementary table S2), except
for BRCA1 c.670+8CRT in intron 10 (table 2), for which two
algorithms predicted the generation of a new strong donor splice
site. As the predicted strength of this site is comparable with
that of the natural one, which is only six nucleotides upstream,
this new site, if used, would produce a mRNA species including
six nucleotides of intron sequences, which should not be
affected by the NMD mechanism in vivo and should also be
seen in the minigene-based assay. However, in spite of having
sequenced all RT-PCR products, we did not find any evidence
for such an mRNA species in our two assays, indicating that the
natural donor splice site is the major and possibly the only
donor site used. It is intriguing that the only effect found
associated with this variant, both in patient RNA and in the
minigene assay, was increased inclusion of BRCA1 exon 10.
DISCUSSION
By using an ex vivo splicing assay combined with RT-PCR
analyses on patient RNA, we examined the effect on splicing of
20 UVs: 7 in BRCA1 and 13 in BRCA2. Of these, 16 were
intronic, and all affected positions distinct from the absolutely
conserved GT or AG dinucleotides at intron boundaries. Several
BRCA1 or BRCA2 UVs have previously been found to alter
splicing,
12–15
but the variants found to alter RNA splicing in our
study have not previously been examined at the RNA level.
Unclassified variants of BRCA1 characterised as splicing defects
Of the seven BRCA1 variants tested in this study, we did not
find any effect on splicing for five. Variant c.5277+48_59dup12
in intron 20 has often been reported as a variant of controversial
clinical significance, more often as neutral
16–18
but also as a
potential regulatory sequence change.
19
Our data show that it
does not affect splicing.
The BRCA1 variant c.4987-5TRA (IVS16-5TRA) showed
clearly deleterious effects in patient RNA, because of disruption
of the open reading frame on a large fraction of and possibly on
all transcripts derived from the affected allele (fig 1C). The
minigene splicing assay, which is monoallelic, confirmed the
very strong effect of this variant, as no normal inclusion of the
corresponding exon was seen using this assay (fig 1D). Another
variant of BRCA1 tested, c.670+8CRT (IVS10+8CRT), induced
complete inclusion of the alternatively spliced exons 9 and 10
and this effect was confirmed using the minigene splicing assay
(table 2). Although this observation indicates that the donor
splice site of BRCA1 exon 10 is strengthened by the CRT
transition at position +8, it is presently difficult to envisage a
pathogenic role for this variant, because its effects seem to be
limited to reduction of an alternative splicing present in normal
individuals. The augmented inclusion of exon 10 induced by
this variant is apparently not due to the presence of an
additional donor splice site at position c.670+6, a scenario
predicted by two algorithms (table 2), because no mRNA species
containing the six initial bases of intron 10 was detected in
patient RNA or ex vivo using the minigene assay (data not
shown). Additional studies with a cell line carrying this BRCA1
variant are needed to understand its effect.
Figure 1 In vivo and ex vivo analysis of
BRCA1 variants that alter the strength of
acceptor-splice sites. Left panels, RT-PCR
products obtained from patient RNA. The
position of the reverse transcriptase (RT)-
PCR primers is shown above the panel.
All RT-PCR products shown have been
sequenced and their sizes are marked
alongside the left-hand margin. Lanes C1
and C2, control subjects; lane P, patient.
Right panels, corresponding ex vivo
assays. The genomic DNA segments
tested are shown schematically above
the panel. Black boxes (marked V)
represent exons of the splicing reporter
minigene. For simplicity, product sizes are
not indicated for the ex vivo analysis, but
the splicing patterns, determined by
sequencing each product, are shown
schematically alongside the right-hand
margin of the mutant lane. M, mutant; wt,
wild type. (A) Previously described
8
splicing mutation BRCA1 c.213-11TRG.
(B) Splicing analysis of this variant; the
schematic below the panel shows a
cryptic acceptor site (denoted ag) in
intron 5, at position 259. (C) BRCA1
c.4987-5TRA variant. (D) splicing
analysis of this variant; the schematic
shows a cryptic donor site (gt) at position
c.5074+154 of intron 17 and cryptic
acceptor site (AG) present at the end of
exon 17, immediately before the normal
donor site.
Original article
442 J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

Unclassified variants of BRCA2 characterised as splicing defects
Of the 13 BRCA2 variants tested separately in the minigene
assay (table 2), 8 failed to reveal any effect on splicing. Of these,
the exonic variant BRCA2 c.8162TRC in exon 18 affects a
position only three nucleotides upstream of the mutation
BRCA2 c.8165CRG (predicting p.T2722R), which is known to
cause exon skipping by disrupting several ESE sites.
20
Two variants, BRCA2 c.316+5GRC (IVS3+5GRC) and
c.7805GRC, at the last base of exon 16, induced strong effects
on splicing (figs 2A,B and 2C,D, respectively). Several other
defects resulting in exon 3 skipping have been described since
the initial reports by Nordling et al
21
and Santarosa et al.
22
Although exon 3 skipping is in frame, 83 amino acids are
removed from the BRCA2 protein, in a region that displays a
rather high degree of evolutionary conservation. Important
functions have been attributed to this BRCA2 region, especially
a transcriptional activation function and interaction with the
product of the EMSY gene, which is amplified in sporadic breast
cancer and in higher-grade ovarian cancer.
23
The criteria for the
interpretation of BRCA2 exon 3 skipping have recently been
debated, following the description of an Alu insertion into exon
3 (c.156–157 insAlu), which is a founder mutation of Portuguese
origin.
24 25
Our data show that exon 3 skipping alters essentially
100% of transcripts from the affected allele, both in patient
blood cells (fig 2A), by using allele-specific primers, and in the
minigene assay (fig 2B). Therefore, despite suggestions that
BRCA2 mRNA species lacking exon 3 sequences may be rather
ubiquitous and may have some function,
25
the effect of the
c.316+5GRC variant described here cannot be considered as an
augmentation of a natural phenomenon, but should be
Figure 2 In vivo and ex vivo analysis of
BRCA2 variants that alter the strength of
donor-splice sites. Left panels, RT-PCR
products obtained from patient RNA. The
position of the reverse transcriptase (RT)-
PCR primers is shown above the panel.
All RT-PCR products shown have been
sequenced and their sizes are marked
alongside the left-hand margin. Lanes C1
and C2, control subjects; lane P, patient.
Right panels, corresponding ex vivo
assays. The genomic DNA segments
tested are shown schematically above
the panel. Black boxes (marked V)
represent exons of the splicing reporter
minigene. For simplicity, product sizes are
not indicated for the ex vivo analysis, but
the splicing patterns, determined by
sequencing each product, are shown
schematically alongside the right-hand
margin of the mutant lane. M, mutant; wt,
wild type. (A) RNA from patient carrying
the BRCA2 c.316+5GRC variant was
tested in two distinct RT-PCRs together
with RNA from control individuals: using a
forward primer in the 59UTR sequences or
within exon 2, at the polymorphic
c.226GRA position. (B) Ex vivo splicing
analysis of this variant. (C) BRCA2
c.7805GRC ; an arrow points to the
inclusion of only 88 nucleotides of exon
16, as shown by sequencing (data not
shown). (D) Ex vivo analysis of this
variant, confirming exon skipping and the
inclusion of only 88 nucleotides of exon
16, because of a cryptic donor site at
position c.7706 of this exon, marked by
gt. (E) BRCA2 c.9501+3ART. (F) Ex vivo
splicing assay of the same variant.
Original article
J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895 443
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

interpreted as complete loss of functions specified by this
BRCA2 region.
We also consider BRCA2 c.7805GRC, which affects the last
position of exon 16, to be a pathogenic mutation because both
defective mRNA forms it induces (exon 16 skipping and
inclusion of only 88 nucleotides of exon 16) result in frameshifts
and because of the total absence of normal-size transcript in the
monoallelic minigene-based assay. The same arguments can be
applied to the variant BRCA2 c.8023ARG in exon 18, which
generates a stronger internal donor splice site within exon 18
than the natural one (fig 3A). The allele tested in this work also
carries a CRA transition 35 bp within intron 17, but a
contribution of this second variant to the splicing alteration
seems unlikely, although it cannot be formally excluded, as we
have not tested this second variant alone.
Unclassified variants that diminish, but do not abolish exon
inclusion
Other BRCA2 variants had partial effects on splicing. BRCA2
c.9501+3ART induced exon 25 skipping in about 50% of the
transcripts (fig 2E,F). The evidence for pathogenicity of this
variant is not compelling, because the splicing alteration affects
only a fraction of transcripts from the variant allele, at least in
patient blood cells and in the minigene assay.
The one-codon deletion c.8257_8259del in exon 18, which has
not been reported in previous studies, induced a modification of
the alternative splicing involving exons 17 and 18 (fig 3C) and
shifts this alternative splicing towards partial exclusion of exon
18 alone. The amount of transcript carrying both exon 17 and
exon 18 appear to be similar in patients and in control
individuals (panel C). In the ex vivo splicing assay, low levels
of exon 18 exclusion are also observed when this exon is tested
alone, and the assay with both exons recapitulates the shift
towards preferential exclusion of exon 18 (fig 3D). These data
suggest that an exonic splicing regulatory element is altered by
the one-codon deletion BRCA2 c.8257_8259del, but no ESE is
predicted at this position by the web-based program ESEfinder
(http://rulai.cshl.edu/tools/ese/). This position is distinct from
the ESE region highlighted in exon 18 by the aforementioned
mutation c.8165CRG
20
and also from the position of the
c.8536GRA variant, reported in the kConFab database and
predicted to reduce exon 18 inclusion.
26
Figure 3 In vivo and ex vivo analysis of
exonic variants. Left panels, RT-PCR
products obtained from patient RNA. The
position of the reverse transcriptase (RT)-
PCR primers is shown above the panel.
All RT-PCR products shown have been
sequenced and their sizes are marked
alongside the left-hand margin. Lanes C1
and C2, control subjects; lane P, patient.
Right panels, corresponding ex vivo
assays. The genomic DNA segments
tested are shown schematically above
the panel. Black boxes (marked V)
represent exons of the splicing reporter
minigene. For simplicity, product sizes are
not indicated for the ex vivo analysis, but
the splicing patterns, determined by
sequencing each product, are shown
schematically alongside the right-hand
margin of the mutant lane. M, mutant; wt,
wild type. (A) BRCA2 c.8023ARG variant
(note in lanes C1 and C2 the weak
product from alternative splicing of exon
18 and, in lane P, the strong 307 bp
product corresponding to the in-frame
deletion of 309 nucleotides from the
mRNA, because of a new donor splice
site generated within the exon, as shown
schematically). (C) BRCA2
c.8257_8259del variant (B) Ex vivo assay
of the same variant, performed on the
exon 18 region either alone, or together
with exon 17. (C) BRCA2 c.8257_8259del
variant. Note the modification of the
alternatively spliced forms involving exon
17 and exon 18. (D) Ex vivo splicing
assay of the same variant confirming the
patient-specific modification of alternative
splicing using either the exon 18 region or
both exons 17 and 18.
Original article
444 J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

Contribution of splicing reporter minigene assays to the
classification of unclassified variants
A recent genetic assessment of a large number of UVs in BRCA1
or BRCA2
27
has shown that statistical analysis of data from
personal and family history of cancer, from cosegregation of
UVs with disease in pedigrees and from co-occurrence in trans
with deleterious mutations defines a fraction of UVs that is very
probably deleterious. Those authors found that all 11 changes
they examined affecting conserved GT or AG dinucleotides at
intron boundaries fell within a group of 43 variants with high
odds in favour of causality, but they also found that 5 additional
UVs affecting degenerate splice donor or splice acceptor
positions fell into the same group. Therefore, variants affecting
splicing are notable in that study as a major group of UVs that
are likely to be causative for cancer predisposition.
27
Despite our initial concern that a fraction of splicing
alterations might escape detection by examining UV-containing
regions in a heterologous context, the present study showed a
high level of concordance between patient RNA data and
predictions from the splicing reporter minigene assay (table 2,
figs 1–3). Besides the major difficulty of obtaining blood samples
or cell lines for mRNA analyses for diagnosis, biallelic RT-PCR
analyses on patient blood often cannot be interpreted quantita-
tively, especially if mRNA degradation by NMD cannot be
prevented. The value of minigene assays is confirmed by
additional studies that we have carried out using the same
splicing reporter minigene on UVs found in the mismatch repair
genes involved in Lynch syndrome.
9
The interpretation of variants that reduce, but do not abolish
exon inclusion, as illustrated here, is difficult at present.
Whether such ex vivo assays or the analysis of RNA from
patient blood cells reflect the amplitude of a given splicing
alteration in cells that are more relevant to tumour formation
remains an open question, which might eventually be answered,
at least in some cases, if genetic data concerning the penetrance
of these mutations become available. For mutations that affect
splicing regulatory elements and sometimes induce relatively
weak effects on splicing in patient blood cells or in minigene
assays, it will be important to establish experimentally if they
are able to cause strong effects in tumour cells, because of
changes in the balance of trans-acting regulatory factors.
Complete absence of the full-size mRNA expressed from the
variant allele, combined with evidence that the aberrant mRNA
forms that are produced either induce frameshifts or determine
loss of a functionally important segment of the protein, can be
taken as biological indicators of pathogenicity. In this regard,
the consequences of the four variants considered in table 2 as
pathogenic should not differ substantially from the conse-
quences of mutations affecting the canonical AG/GT splice
sites. We consider UVs that did not show any effect in patient
RNA, nor in the minigene-based assay, to be most likely non-
pathogenic, but do not exclude the possibility of pathogenic
effects at a level other than splicing. Formal proof of non-
pathogenicity requires analyses of family segregation and of co-
occurrence with clearly deleterious changes. Although such
deleterious changes have been found associated, in several
French patients, with some of the UVs that failed to reveal
splicing defects in this study (table 2), further studies are needed
as we do not know at this stage if they are in cis or in trans with
the pathogenic mutation.
In conclusion, this work documents the power of combining
patient RNA analysis and ex vivo analysis using a splicing
reporter minigene. The sensitivity of the ex vivo approach
justifies its use for the initial characterisation of the effects of
UVs on splicing, especially when blood samples are not readily
available for patient RNA analysis.
Acknowledgements: We thank Dr O Sinilnikova for expert advice in the initial phases
of this work. This project was supported by the French North West Canceropole and
by the Fondation de France. CB was supported by a fellowship from the University
Hospital of Nancy, France; IB was supported by a fellowship from ARC (Association
pour la Recherche sur le Cancer) and MV is supported by a fellowship from the Ligue
Contre le Cancer, Comite´ de l’Eure.
Competing interests: None.
REFERENCES
1. Phelan CM, Dapic V, Tice B, Favis R, Kwan E, Barany F, Manoukian S, Radice P, van
der Luijt RB, van Nesselrooij BP, Chenevix-Trench G, kConFab, Caldes T, de la Hoya
M, Lindquist S, Tavtigian SV, Goldgar D, Borg A, Narod SA, Monteiro AN.
Classification of BRCA1 missense variants of unknown clinical significance. J Med
Genet 2005;42:138–46.
2. Tavtigian SV, Deffenbaugh AM, Yin L, Judkins T, Scholl T, Samollow PB, de Silva D,
Zharkikh A, Thomas A. Comprehensive statistical study of 452 BRCA1 missense
substitutions with classification of eight recurrent substitutions as neutral. J Med
Genet 2006;43:295–305.
3. Faustino NA, Cooper TA. Pre-mRNA splicing and human disease. Genes Dev
2003;17:419–37.
4. Baralle D, Baralle M. Splicing in action: assessing the disease causing sequence
changes. J Med Genet 2005;42:737–48.
5. Cooper TA. Use of minigene systems to dissect alternative splicing elements.
Methods 2005;37:331–40.
6. Hogervorst FB, Nederlof PM, Gille JJ, McElgunn CJ, Grippeling M, Pruntel R,
Regnerus R, van Welsem T, van Spaendonk R, Menko FH, Kluijt I, Dommering C,
Verhoef S, Schouten JP, van’t Veer LJ, Pals G. Large genomic deletions and
duplications in the BRCA1 gene identified by a novel quantitative method. Cancer Res
2003;63:1449–53.
7. Casilli F, Di Rocco ZC, Gad S, Tournier I, Stoppa-Lyonnet D, Frebourg T, Tosi M.
Rapid detection of novel BRCA1 rearrangements in high-risk breast-ovarian cancer
families using Multiplex PCR of short fluorescent fragments. Human Mutat
2002;20:218–26.
8. Friedman LS, Ostermeyer EA, Szabo CI, Dowd P, Lynch ED, Rowell SE, King MC.
Confirmation of BRCA1 by analysis of germline mutations linked to breast and ovarian
cancer in ten families. Nat Genet 1994;8:399–404.
9. Tournier I, Vezain M, Martins A, Charbonnier F, Baert-Desurmont S, Olschwang S,
Wang Q, Buisine MP, Soret J, Tazi J, Fre´bourg T, Tosi M. A large proportion of
unclassified variants of the mismatch repair genes MLH1 and MSH2 are associated
with splicing defects. Human Mutat 2008, in press.
10. Tesoriero AA, Wong EM, Jenkins MA, Hopper JL, Brown MA, Chenevix-Trench G,
Spurdle AB, Southey MC. 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.
11. Houdayer C, Dehainault C, Mattler C, Michaux D, Caux-Moncoutier V, Page`s-
Berhouet S, Dubois d’Enghien C, Lauge´ A, Castera L, Gauthier-Villars M, Stoppa-
Lyonnet D. Evaluation of in silico splice tools for decision-making in molecular
diagnosis. Hum Mutat Child. Published Online First: 30 April 2008; doi: 10.1002/
humu.20765.
12. Campos B, Diez O, Domenech M, Baena M, Balmana J, Sanz J, Ramirez A, Alonso
C, Baiget M. RNA analysis of eight BRCA1 and BRCA2 unclassified variants identified
in breast/ovarian cancer families from Spain. Hum Mutat 2003;22:337.
13. 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.
14. Bonatti F, Pepe C, Tancredi M, Lombardi G, Aretini P, Sensi E, Falaschi E, Cipollini G,
Bevilacqua G, Caligo MA. RNA-based analysis of BRCA1 and BRCA2 gene alterations.
Cancer Genet Cytogenet 2006;170:93–101.
15. Chen X, Truong TT, Weaver J, Bove BA, Cattie K, Armstrong BA, Daly MB, Godwin
AK. Intronic alterations in BRCA1 and BRCA2: effect on mRNA splicing fidelity and
expression. Hum Mutat 2006;27:427–35.
16. Scholl T, Pyne MT, Ward B, Pruss D. Biochemical and genetic characterisation
shows that the BRCA1 IVS20 insertion is a polymorphism. J Med Genet
1999;36:571–2.
17. Dork T, Skawran B, Stuhrmann M, Bremer M, Karstens JH. BRCA1 expression is not
affected by the intronic 12 bp duplication. J Med Genet 1999;36:262–3.
18. Kozlowski P, Sobczak K, Jasinska A, Krzyzosiak WJ. Allelic imbalance of BRCA1
transcript in the IVS20 12-bp insertion carrier. Hum Mutat 2000;16:371.
19. Robledo M, Osorio A, Sentis C, Albertos J, Estevez L, Benitez J. The 12 base pair
duplication/insertion alteration could be a regulatory mutation. J Med Genet
1997;34:592–93.
20. Fackenthal JD, Cartegni L, Krainer AR, Olopade OI. BRCA2 T2722R is a deleterious
allele that causes exon skipping. Am J Hum Genet 2002;71:625–31.
21. Nordling M, Karlsson P, Wahlstrom J, Engwall Y, Wallgren A, Martinsson T. A large
deletion disrupts the exon 3 transcription activation domain of the BRCA2 gene in a
breast/ovarian cancer family. Cancer Res 1998;58:1372–5.
Original article
J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895 445
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

22. Santarosa M, Viel A, Boiocchi M. Splice variant lacking the transactivation domain
of the BRCA2 gene and mutations in the splice acceptor site of intron 2. Genes
Chromosomes and Cancer 1999;26:381–2.
23. Hughes-Davies L, Huntsman D, Ruas M, Fuks F, Bye J, Chin SF, Milner J, Brown LA,
Hsu F, Gilks B, Nielsen T, Schulzer M, Chia S, Ragaz J, Cahn A, Linger L, Ozdag H,
Cattaneo E, Jordanova ES, Schuuring E, Yu DS, Venkitaraman A, Ponder B, Doherty A,
Aparicio S, Bentley D, Theillet C, Ponting CP, Caldas C, Kouzarides T. EMSY links the
BRCA2 pathway to sporadic breast and ovarian cancer. Cell 2003;115:523–35.
24. Machado PM, Brandao RD, Cavaco BM, Eugenio J, Bento S, Nave M, Rodrigues P,
Fernandes A, Vaz F. Screening for a BRCA2 rearrangement in high-risk breast/ovarian
cancer families: evidence for a founder effect and analysis of the associated
phenotypes. J Clin Oncol 2007;25:2027–34.
25. Diez O, Gutierrez-Enriquez S, Cajal T, Alonso C, Balmana J, Llort G. Caution shouldbe
used when interpreting alterations affecting the exon 3 of the BRCA2 gene in breast/
ovarian cancer families. J Clin Oncol 2007;25:5035–6.
26. Pettigrew CA, Wayte N, Wronski A, Lovelock PK, Spurdle AB, Brown MA.
Colocalisation of predicted exonic splicing enhancers in BRCA2 with reported
sequence variants. Breast Cancer Res Treat Published Online First: 26 September
2007; doi: 10.1007/s10549-007-9714-5.
27. Easton DF, Deffenbaugh AM, Pruss D, Frye C, Wenstrup RJ, Allen-Brady K,
Tavtigian SV, Monteiro AN, Iversen ES, Couch FJ, Goldgar DE. A systematic genetic
assessment of 1,433 sequence variants of unknown clinical significance inthe
BRCA1 and BRCA2 breast cancer-predisposition genes. Am J Hum Genet
2007;81:873–83.
Submit an eLetter, and join the debate
eLetters are a fast and convenient way to register your opinion on topical and contentious medical
issues. You can find the ‘‘submit a response’’ link alongside the abstract, full text and PDF versions of
all our articles. We aim to publish swiftly, and your comments will be emailed directly to the author of
the original article to allow them to respond. eLetters are a great way of participating in important
clinical debates, so make sure your voice is heard.
Original article
446 J Med Genet 2008;45:438–446. doi:10.1136/jmg.2007.056895
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from

minigene
vivo assay based on a splicing reporter
transcription PCR on patient RNA and an ex
variants for splicing mutations using reverse
unclassifiedBRCA2 and BRCA1Screening
A Hardouin and M Tosi
Berthet, A Chevrier, C Dugast, V Layet, A Rossi, R Lidereau, T Frébourg,
C Bonnet, S Krieger, M Vezain, A Rousselin, I Tournier, A Martins, P
doi: 10.1136/jmg.2007.056895
2008 45: 438-446 originally published online April 18, 2008J Med Genet
http://jmg.bmj.com/content/45/7/438
Updated information and services can be found at:
These include:
Material
Supplementary http://jmg.bmj.com/content/suppl/2008/06/16/45.7.438.DC1.html
Supplementary material can be found at:
References #BIBLhttp://jmg.bmj.com/content/45/7/438
This article cites 24 articles, 11 of which you can access for free at:
service
Email alerting box at the top right corner of the online article.
Receive free email alerts when new articles cite this article. Sign up in the
Collections
Topic Articles on similar topics can be found in the following collections
(591)Epidemiology (1194)Molecular genetics
(221)Breast cancer
Notes
http://group.bmj.com/group/rights-licensing/permissions
To request permissions go to:
http://journals.bmj.com/cgi/reprintform
To order reprints go to:
http://group.bmj.com/subscribe/
To subscribe to BMJ go to:
group.bmj.com on April 18, 2015 - Published by http://jmg.bmj.com/Downloaded from