Supplemental Table 1. Detection and Validation of 63 BRCA1 alternative splicing events.
Stage 2 assays (fragment sizes are indicated)2
Stage 11
Full-length
(1B)
(1B)-6
1A-3
1A-6
1A-11q
3-8
452bp
205bp
391bp
911bp
427bp
199bp
385bp
186bp
372bp
98bp
286bp
7-11q
7-12
8-13
10-12
10-13
448bp 3793bp 3837bp 3530bp 3661bp
12-13
12-14
12-16
13-22
14-22
16-22
16-24
20-24
166bp
288bp
769bp
1099bp
923bp
418bp
556bp
269bp
452bp
353bp
299bp
216bp
D1Aq
D1Aq,2
D1Aq_3
D1Aq_3+▼4
231bp
348bp
153bp
D1Aq_10
▼1aA
▼1aA+D2_17
216bp
294bp
480bp
192bp
378bp
104bp
292bp
not validated in the present study (see Supplemental Table 3 for further details)
D2
D2,3
237bp
354bp
159bp
D2_10
D3
D3_5*
▼4
D5
D5q
D5q,6
D6,7
D8p
222bp
398bp
337bp
not validated in the present study (see Supplemental Table 3 for further details)
508bp
543bp
374bp
313bp
349bp
430bp
369bp
405bp
not validated in the present study (see Supplemental Table 3 for further details)
not validated in the present study (see Supplemental Table 3 for further details)
424bp
445bp
296bp
219bp
D9
D9,10
D9_11
401bp
D10
D10,11
371bp
327bp
244bp
288bp
201bp
290bp
337bp
247bp
244bp
D11
367bp
412bp
104bp
322bp
148bp
319bp
11D3110
11D3240
D11q
(IRIS)
D13p
484bp
535bp
235bp
145bp
420bp
551bp
287bp
421bp
221bp
355bp
BRCA1-IRIS was analyzed with a non-fluorescent dedicated assay
163bp
285bp
116bp
D13_19*
▼13A
4
597bp
113bp
not validated in the present study (see Supplemental Table 3 for further details)
354bp
351bp
D14p
285bp
642bp
451bp
D14_17
D14_18
382bp
304bp
263bp
578bp
D15_17
D15_19
D17_19*
D18_20
D19
509bp
333bp
390bp
214bp
330bp
468bp
340bp
478bp
334bp
472bp
363bp
501bp
not validated in the present study (see Supplemental Table 3 for further details)
not validated in the present study (see Supplemental Table 3 for further details)
not validated in the present study (see Supplemental Table 3 for further details)
D21
D22
1
214bp
366bp
79bp
427bp
140bp
482bp
195bp
420bp
133bp
495bp
209bp
Output3
Full-length
(1B)
(1B)+D2
(1B)+D2,3
(1B)+D2_5#
D1Aq
D1Aq,2p
D1Aq,2
D1Aq_3
D1Aq_3+▼4
D1Aq_5
D1Aq_10
▼1aA
?
D2p
D2
D2,3
D2,3+▼4
D2_5
D2_10
D3
?
▼4
D5
D5q
?
?
D8p
D8,9
D8_10
D9
D9,10
D9_11
D9_12
D10
D10,11
D10_12
D10_13p
D11
D11,12
D11_13p
11D3110
11D3240
D11q
(IRIS)
D13p
D13
D13,14p
?
▼13A
▼13A+14p
D14p
D14
D14_15
D14_17
D14_18
D14_19
D15
D15_17
D15_19
D17
?
D18
?
?
D20
D21
D21,22
D21_23
D22
D22,23
D23
Splicing events displayed in blue boxes were supported by experimental evidence of contributing laboratories. By contrast, splicing events displayed in white
boxes were supported solely by public domain scientific data 2 Predicted size-fragments according to Ensemble reference transcript ENST00000357654. The
actual size-fragment observed in CE might be slightly different (±2bp) 3 Splicing events displayed in blue boxes have been validated in the present study. * Stage
2 amplicons were designed prior to the publication of these findings. 4 Exon1B terminal events have been analyzed by one contributing laboratory only. Since
(E1B)+Δ2_5 has not been sequenced, this event do not qualify for a validated event according to our own criteria (see methods). However, we have validated this
imputed splicing event based on the following data. First, BRCA1 Δ2, and Δ2,3 events have been validated in both E1B and E1A containing transcripts,
suggesting that alternative splicing events at the 5' end of the BRCA1 gene is similar, regardless of the transcription star site used. Second, Δ2_5 has been
validated in E1A transcripts.
1
Supplemental Table 2. Stage 2 and 3 Data
detected (assessed)3
PBMCs PBLs LCLs BREAST
Designation
QA1
Coverage 2
Detection Rate 2
(1B)
(1B)+D2
(1B)+D2,3
(1B)+D2_5
D1Aq
D1Aq,2p
D1Aq,2
D1Aq_3
D1Aq_3+▼4
D1Aq_5
D1Aq_10
▼1aA
D2p
D2
D2,3
D2,3+▼4
D2_5
D2_10
D3
▼4
D5
D5q
D8p
D8,9
D8_10
D9
D9,10
D9_11
D9_12
D10
D10,11
D10_12
D10_13p
D11
D11,12
D11_13p
11D3110
11D3240
D11q
(IRIS)
D13p
D13
D13,14p
▼13A
▼13A+14p
D14p
D14
D14_15
D14_17
D14_18
D14_19
D15
D15_17
D15_19
D17
D18
D20
D21
D21,22
D21_23
D22
D22,23
D23
Total
Average
not assesable
35
100%
10(10)
6(6)
7(7)
10(10)
2(2)
(1B)-6
not assessable
35
51%
7(10)
0(6)
4(7)
5(10)
2(2)
(1B)-6
not assessable
35
26%
0(10)
0(6)
2(7)
6(10)
1(2)
(1B)-6
not assessable
35
17%
4(10)
0(6)
1(7)
1(10)
0(2)
(1B)-6
predominant
61
100%
8(8)
10(10)
31(31)
10(10)
2(2)
1A-6
minor
61
38%
4(8)
0(10)
9(31)
8(10)
2(2)
1A-6
minor
61
65%
6(8)
5(10)
17(31)
10(10)
2(2)
1A-6
minor
61
28%
1(8)
0(10)
7(31)
8(10)
2(2)
1A-6
minor
61
21%
2(8)
1(10)
3(31)
7(10)
0(2)
1A-6
minor
61
21%
2(8)
2(10)
3(31)
1(10)
2(2)
1A-6
minor
76
46%
5(9)
5(8)
17(40)
7(17)
1(2)
1A-11q
minor
61
43%
4(8)
7(10)
10(31)
3(10)
2(2)
1A-6
minor
61
38%
4(8)
0(10)
9(31)
8(10)
2(2)
1A-6
minor
61
65%
6(8)
5(10)
17(31)
10(10)
2(2)
1A-6
minor
61
28%
1(8)
0(10)
7(31)
8(10)
2(2)
1A-6
minor
61
21%
2(8)
1(10)
3(31)
7(10)
0(2)
1A-6
minor
61
21%
2(8)
2(10)
3(31)
1(10)
2(2)
1A-6
minor
76
17%
2(9)
0(8)
9(40)
2(17)
0(2)
1A-11q
minor
61
47%
4(8)
4(10)
9(31)
10(10)
2(2)
1A-6
minor
61
49%
4(8)
4(10)
11(31)
10(10)
1(2)
1A-6
predominant
124
14(14)
18(18)
37(69)
19(20)
2(3)
1A-6,3-8
predominant
124
73%
68%
14(14)
10(18)
37(69)
20(20)
3(3)
1A-6,3-8
predominant
116
100%
10(14)
16(16)
63(63)
20(20)
3(3)
3-8,7-11B
minor
55
13%
0(10)
0(8)
4(25)
3(10)
0(2)
7-11B
minor
55
38%
2(10)
1(8)
9(25)
8(10)
1(2)
7-11B
predominant
55
62%
3(10)
4(8)
15(25)
10(10)
2(2)
7-11B
predominant
55
100%
10(10)
8(8)
25(25)
10(10)
2(2)
7-11B
predominant
77
71%
15(15)
14(14)
8(29)
16(16)
2(3)
7-12,8-13
minor
25
28%
0(6)
3(6)
0(5)
3(6)
1(2)
8-13
minor
55
33%
6(10)
1(8)
10(25)
1(10)
0(2)
7-11B
minor
77
28%
7(15)
10(14)
1(29)
3(16)
1(3)
7-12,8-13
minor
22
43%
2(6)
3(6)
0(5)
3(3)
1(2)
8-13
minor
19%
0(6)
0(6)
0(5)
3(3)
1(2)
8-13
minor
22
115
93%
19(20)
21(22)
43(49)
20(20)
4(4)
8-13,10-12,10-13
minor
65
81%
7(12)
12(14)
22(27)
10(10)
2(2)
8-13,10-13
minor
65
66%
4(12)
10(14)
17(27)
10(10)
2(2)
8-13,10-13
minor
98
11%
3(14)
1(16)
1(44)
6(20)
0(4)
10-12,10-13
LEUs
Assays4
minor
98
59%
6(14)
13(16)
19(44)
20(20)
0(4)
10-12,10-13
predominant
115
99%
20(20)
22(22)
48(49)
20(20)
4(4)
8-13,10-12,10-13
not assessable
18
66%
6(6)
3(3)
0(6)
3(3)
not tested
dedicated assay
predominant
56
87%
4(6)
8(8)
26(31)
10(10)
1(1)
12-13
minor
56
54%
0(6)
8(8)
13(31)
10(10)
0(1)
12-14
minor
56
54%
0(6)
8(8)
13(31)
10(10)
0(1)
12-14
minor
56
56%
2(6)
8(8)
19(31)
2(10)
1(1)
12-14
minor
56
56%
2(6)
8(8)
19(31)
2(10)
1(1)
12-14
predominant
57
100%
6(6)
8(8)
31(31)
10(10)
2(2)
12-14
minor
18
17%
1(4)
0(8)
0(1)
2(4)
0(1)
12-16
minor
18
17%
1(4)
0(8)
0(1)
2(4)
0(1)
12-16
minor
65
9%
0(6)
0(8)
3(40)
3(10)
0(1)
13-22
minor
65
3%
0(6)
0(8)
0(40)
2(10)
0(1)
13-22
minor
65
3%
0(6)
0(8)
2(40)
0(10)
0(1)
13-22
minor
18
22%
3(4)
0(8)
0(1)
1(4)
0(1)
12-16
minor
110
25%
4(12)
5(16)
11(60)
6(20)
1(2)
14-22,13-22
minor
110
44%
6(12)
8(16)
24(60)
10(20)
1(2)
14-22,13-22
minor
103
27%
1(10)
3(16)
12(55)
12(20)
0(2)
16-22,16-24
minor
103
40%
0(10)
8(16)
14(55)
19(20)
0(2)
16-22,16-24
minor
51
19%
0(6)
1(8)
4(26)
4(10)
1(1)
16-24
minor
163
51%
8(21)
13(24)
28(80)
29(30)
5(8)
16-24,20-24,16-22
minor
115
43%
6(17)
8(16)
20(55)
12(20)
3(7)
16-24,20-24
minor
64
17%
0(11)
0(8)
10(29)
0(10)
1(6)
20-24
minor
115
70%
12(17)
13(16)
30(55)
20(20)
6(7)
16-24,20-24
minor
64
58%
4(11)
8(8)
13(29)
10(10)
2(6)
20-24
minor
115
49%
5 (17)
10(16)
20(55)
20(20)
1(7)
16-24,20-24
4281
618
681
2055
773
154
68
10
11
33
12
2
1
Qualitative abundance (QA) based on visual inspection of splicing assays (see methods for further details) 2 Coverage refers to data-points per splicing event.
Detection Rate refers to the % of positive data-points. (See methods for further details). 3 Gray boxes indicate a minimum of one positive data-point. 4Splicing
assays interrogating the corresponding event. Some events are detected by two or more different assays (see Supp Table 1 for further details).
2
Supplemental Table 3. BRCA1 alternative splicing events not validated by our experimental approach.
This study
1
Previous studies
Designation
HGVS
Biotype
Functional
Annotation
Coverage1
GENCODE
Blood
Others
▼1aA+D2_17
r.-20_-19ins-20+1_-20+89 + r.-19_5074del
Splice donor shift + Multi-cassette
No FS2
72*
-
-
BCCLs, BCs(22)
D3_5
r.81_217del
Multi-cassette
PTC-NMD
61
-
PBMCs (17)
-
D5q,6
r.191_301del
Splice donor shift + cassette
PTC-NMD
124
-
PBMC (48)
-
D6,7
r.213_441del
Multi-cassette
PTC-NMD
124
-
PBMC (48)
-
D13_19
r.4186_5193del
Multi-cassette
No FS
20*
-
PBMCs (17)
-
D17_19
r.4987_5193del
Multi-cassette
No FS
223
-
PBMCs (17)
-
D18_20
r.5075_5277del
Multi-cassette
PTC-NMD
103
-
-
-
D19
r.5153_5193del
Cassette
PTC-NMD
103
-
-
-
Coverage was calculated pooling together splicing assays performed in LEUs, PBMCs, PBLs, LCLs, and BREAST samples (independent control samples plus technical replicas). 2 Exon 1Aa is a splice donor
shift event incorporating 89 nucleotides into exon 1A. The inserted nucleotides include an in-frame ATG. * Since none of the Stage 2 splicing assays (see Supp Table 1) are able to detected ▼1aA+D2_17 and
D13_19 events, we later performed dedicated screenings with splicing assays 1A-22 and 12-22.
Since ▼1aA+D2_17 has been detected in Breast Cancer Cell Lines (BCCLs) and Breast Cancers (BCs) but not in control samples, it does not necessarily qualify for a naturally occurring splicing event (it
might be, for instance, a by-product of the carcinogenic process). Yet, ▼1aA+D2_17 combines two biotypes, one of them a donor shift (▼1aA) detected in the present study as an individual event both in blood
and BREAST (Table 3). By contrast, BRCA1 D(5q,6), D(6,7), D3_5, D13_19, and D17_19 have been reported in blood related control samples, thus apparently qualifying for naturally occurring events. Yet,
the authors do not inform on relevant analytical parameters such as coverage, detection rates, and/or relative expression levels (or whether the analyses were performed in the presence of NMD inhibitors). It is
therefore possible that some of these events represent rather minor events. In this regard it is interesting to note that we have not detected D13_19 and D17_19, but we have detected other (multi)-cassette events
in the same BRCA1 region, most of them among the Stage 2 events with lower detection rates (see Supp. Table 2). We have identified in our study up to 23 PTC-NMD splicing events, suggesting that lack of
NMD inhibition (at least in capillary EP based analysis) is not a major concern (37). Yet, we cannot discard that certain PTC-NMD events, including D(5q,6), D(6,7), and D3_5, are particularly prone to NMD
degradation. BRCA1 D18_20 and D19 have been detected (imputed) by contributors of the present study (one contributor each). However, we do not have sequence evidence nor validation detection by a
second laboratory.
3
Supplemental Figure 1 Annotation
of
BRCA1
events
alternative
Panel
A
splicing
shows
a
representative example of a 7-11q
CE
assay.
Note
that
±3bp
(NAGNAG located at the 5’ end of
exon 8), ±46bp (exon 9), and ±77bp
(exon 10) spacing between peaks are
essential for imputation and proper
peak annotation. Interestingly, the
assay predicts the existence of one
splicing event (SE) involving ±16bp
(annotated as X in the figure).
However, we have not been able to
impute X.
Panel B shows a
representative example of a 12-14
assay.
In
this
assay,
±3bp
(NAGNAG located at the 5’ ends of
exon 13 and exon 14), ±172bp
(exon13), and ±66bp (exon13A)
spacing between peaks are essential
for imputation and proper peak
annotation. Panel C shows a detail of
an independent 12-14 assay (a
technical replica) in which one extra
peak in the 350bp region is observed. Note that the extra peak shown in Panel C is essential for ▼13A,14p annotation. Peaks representing a combination of two independent splicing
events are annotated as (SE1+SE2).Peaks compatible with different SE events are annotated as (SE1)/(SE2). *Annotations not supported by sequencing data (see Tables 1 to 4 for further
details). Note that, overall, the signal corresponding to transcripts containing two splicing events is consistently lower than that of the transcripts containing the corresponding individual
events (compare for instance D8p+D9,10 with D8p and D9,10), as expected from a random combination of independent elements.
4
Supplemental Figure 2. Stochastic
preferential amplification of minor
splicing events. CE analysis of 1624 assays allowed us to identify up
to 8 different alternative splicing
events.
Panel
A
shows
a
representative example. Compared
with the full-length, all them are
rather minor events (note that the
corresponding full-length peak is
saturated), so direct sequencing of
splice junctions is not feasible.
Panels B and C show examples of
stochastic preferential amplification
of minor alternative splicing events
(D23 and D18, respectively) which
allow direct sequencing of the
splicing junctions. To avoid D22
interference, D23 splice junction was
sequenced with a forward primer
located at exon 22. Panels D and E
show two assays performed in
parallel
with
the
same
sample
(technical replicas). Four alternative
splicing events are detected in both
assays (only the 420-520bp region is shown), but D17 is detected only once.
*Annotations
not supported by sequencing data (see Tables 1-4 for further details).
5
Supplemental
Figure
3.
BRCA1
alternative splicing in exon11 lacking
RNAs. Panel A shows a representative
example of a 7-12 CE assay. The fulllength fragment (3793bp) is out of range
for CE analysis. Dashes areas are shown
in greater detail in panel B and C (note
that panel B1 is a magnification of panel
A, but panels B2, C1 and C2 represent
independent assays. Several peaks have
been annotated, most of them representing
combinations
of
independent
events.
Other peaks have not been annotated.
Some
might
represent
unspecific
amplifications, but others (in particular
DB, DC, DD, and DX) are apparently
observed in combination with BRCA1
events, supporting that they are indeed
true BRCA1 splicing events that we have not been able to annotate. Interestingly, the DX event (±16bp) is probably the same DX event detected in 7-11q CE assays (see Figure 1). Peaks
representing a combination of independent splicing events (SE) are annotated as (SE 1+SE2+SEn). *Annotations not supported by sequencing data (see Tables 1-4 for further details). Note
that, overall, the signal corresponding to transcripts containing two or more splicing events is consistently lower than that of the transcripts containing the corresponding individual
events, as expected from a random combination of independent elements.
6
Supplemental Figure 4. NAGNAG sequence motifs at the 5’ end of BRCA1 exon 6, 8, 13 and 14 and local splice-site prediction. Screenshot from
Alamut software (Interactive Biosoftware) showing the local sequence at the 5’end of the exon 6, 8, 13 and 14, including the composition of the
NAGNAG motif. The 3’ predictions in green represent the outcomes for predictions of the strength of a splice acceptor site as predicted by the different
integrated software tools. The height of the vertical green bars is correlated with the height of the score (range of scores indicated in light gray next to
the different tools). The NAGNAG-motif is displayed next to the screenshot, including the experimentally (Exp) determined ratio between the use of the
proximal (P) or distal (D) acceptor site (NAGPNAGD). BayNAGNAG prediction scores (http://www.tassdb.info/baynagnag/form.html) are given (≥0.5
= significant result). PD = usage of both acceptors, P or D = constitutive splicing at either P or D acceptor site.
7
Supplemental Figure 5. Relative expression level of BRCA1 D1Aq in 4 blood related RNA sources
and one breast tissue. 33-cycle RT-PCR assays (E1A-E3) were analyzed by CE. The boxplot shows
the ratios between the peak area of D1Aq and the peak areas of the reference full-length transcript
(Low, Q1, Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small
circle. N indicates the number of individual data-points (different control samples plus technical
replicates). In BREAST, N indicates the number of technical replicas.
8
Supplemental Figure 6. Relative expression level of BRCA1 D5 in 4 blood related RNA sources and
one breast tissue. 33-cycle RT-PCR assays (E3-E8) were analyzed by CE. The boxplot shows the
ratios between the peak area of D5 and the peak areas of the reference full-length transcript (Low, Q1,
Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small circle. Extreme
outliers (>3 IQR) display a star sign. N indicates the number of individual data-points (different
control samples plus technical replicates). In BREAST, N indicates the number of technical replicas
performed.
9
Supplemental Figure 7. Relative expression level of BRCA1 D5q in 4 blood related RNA sources
and one breast tissue. 33-cycle RT-PCR assays (E3-E8) were analyzed by CE. The boxplot shows the
ratios between the peak area of D5q and the peak areas of the reference full-length transcript (Low,
Q1, Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small circle. N
indicates the number of individual data-points (different control samples plus technical replicates). In
BREAST, N indicates the number of technical replicas performed.
10
Supplemental Figure 8. Relative expression level of BRCA1 D8p in 4 blood related RNA sources
and one breast tissue. 33-cycle RT-PCR assays were analyzed by CE. The boxplot shows the ratios
between the peak area of D8p and the peak areas of the reference full-length transcript (Low, Q1,
Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small circle. Extreme
outliers (>3 IQR) display a star sign. N indicates the number of individual data-points (different
control samples plus technical replicates). In BREAST, N indicates the number of technical replicas
performed.
11
Supplemental Figure 9. Relative expression level of BRCA1 D9 in 4 blood related RNA sources and
one breast tissue. 33-cycle RT-PCR assays were analyzed by CE. The boxplot shows the ratios
between the peak area of D9 and the peak areas of the reference full-length transcript (Low, Q1,
Median, Q3, ad High values are indicated). N indicates the number of individual data-points (different
control samples plus technical replicates). In BREAST, N indicates the number of technical replicas
performed.
12
Supplemental Figure 10. Relative expression level of BRCA1 D9,10 in 4 blood related RNA sources
and one breast tissue. 33-cycle RT-PCR assays were analyzed by CE. The boxplot shows the ratios
between the peak area of D9,10 and the peak areas of the reference full-length transcript (Low, Q1,
Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small circle. Extreme
outliers (>3 IQR) display a star sign. N indicates the number of individual data-points (different
control samples plus technical replicates). In BREAST, N indicates the number of technical replicas
performed.
13
Supplemental Figure 11. Relative expression level of BRCA1 D13p in 4 blood related RNA sources
and one breast tissue. 33-cycle RT-PCR assays were analyzed by CE. The boxplot shows the ratios
between the peak area of D13p and the peak areas of the reference full-length transcript (Low, Q1,
Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small circle. N
indicates the number of individual data-points (different control samples plus technical replicates). In
BREAST, N indicates the number of technical replicas performed.
14
Supplemental Figure 12. Relative expression level of BRCA1 D14p in 4 blood related RNA sources
and one breast tissue. 33-cycle RT-PCR assays were analyzed by CE. The boxplot shows the ratios
between the peak area of D14p and the peak areas of the reference full-length transcript (Low, Q1,
Median, Q3, ad High values are indicated). Normal outliers (>1.5 IQR) display a small circle. N
indicates the number of individual data-points (different control samples plus technical replicates). In
BREAST, N indicates the number of technical replicas performed.
15
16