Familial Cancer DOI 10.1007/s10689-014-9730-7
ORIGINAL ARTICLE
Rapid and cost effective screening of breast and ovarian cancer genes using novel sequence capture method in clinical samples ´ rvai • Pe´ter Horva´th • Bernadett Balla • Kristo´f A Anna M. T} oke´s • Ba´lint Tobia´s • Istva´n Taka´cs • Zsolt Nagy • Pe´ter Lakatos • Ja´nos P. Ko´sa
Ó Springer Science+Business Media Dordrecht 2014
Abstract BRCA1 and BRCA2 are two well-known genes in the background of hereditary breast and ovarian cancer. There is also evidence that several other genes play an important role in the pathogenesis of these two malignancies. Latest population-scaled studies showed that certain mutations in different genes could cause similar risk elevation like BRCA2 mutations. In this study we present a new method to analyse the risk assessment of women to breast and ovarian cancer. Using Haloplex, a novel sequence capture method combined with next-generation sequencing we were able to perform rapid and cost-effective screening of 16 genes that could be associated with an increased risk of breast and ovarian cancer. The rapid and cost effective analysis of this 16-gene cohort can reveal the genetic background of approximately 30 % of hereditary and familiar cases of breast and ovarian cancers. Thus, it opens up a new and high-throughput approach with fast turnaround time to the genetic diagnostics of these
Pe´ter Lakatos and Ja´nos P. Ko´sa have equally contributed to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s10689-014-9730-7) contains supplementary material, which is available to authorized users. ´ rvai P. Horva´th J. P. Ko´sa (&) K. A 1st Department of Internal Medicine, Semmelweis University, Budapest, Hungary e-mail: [email protected] ´ rvai B. Balla B. Tobia´s I. Taka´cs Z. Nagy K. A P. Lakatos J. P. Ko´sa PentaCore Laboratory, Budapest, Hungary A. M. T}oke´s MTA-SE Tumor Progression Research Group, SE-2nd Department of Pathology, Budapest, Hungary
disorders and may be helpful to investigate other familial genetic disorders as well. Keywords Breast cancer Next generation sequencing Ovarian cancer Tumor genetics
Introduction So far, two prime high-penetrance genes associated with susceptibility to breast and ovarian cancer—breast cancer susceptibility gene 1 (BRCA1) and breast cancer susceptibility gene 2 (BRCA2)—have been referred. The presence of a deleterious mutation in any of these genes means a lifetime risk to develop breast cancer with a chance between 60–85 % and a 15–40 % risk for ovarian cancer, respectively [1]. A vast array of evidence indicates that other high- or low-penetrance susceptibility mutations may increase the risk of breast or ovarian cancer. Probably several variants in low-penetrance susceptibility genes are more common than mutations in high-penetrance alleles, hence low-penetrance alleles could add more risk to develop breast or ovarian cancer [2, 3]. The aim of research efforts to find new genes to explain the missing heritability in BRCA negative cancer families targeted genes that could interact with the BRCA pathway or the BRCA proteins themselves. Additional high penetrance alleles were found for breast or ovarian cancer, such as TP53 [4] and STK11 [5]. The genetic elements of Fanconi Anemia pathway have also been found as moderate penetrance alleles in these cancers including PALB2, BRIP1, RAD51C [6–9]. Similar penetrance level has been identified in case of other genes which are also involved in the homologous recombination repair pathway, such as ATM and CHEK2. The CHEK2
123
´ rvai et al. K. A
truncating mutations seems to be double the breast cancer risk [10] and recent, population based discoveries suggest that ATM, CHEK2, BRIP1 and PALB2 mutations has similar effect like BRCA2 mutations [11]. Family history of diffuse gastric cancer also mean a higher risk for breast cancer [12] but a mutation in CDH1 gene could increase breast cancer risk without familial gastric cancer syndrome [13]. Moreover, recent evidence suggests that a pathogenic BRCA1/2 mutation could also increase the risk of pancreatic or prostate cancer [14–16]. However, about 70 % of breast or ovarian cancer inheritance is still missing. All the genes mentioned above have in general dozens of coding exons and a complete lack of certain hotspot regions, hence traditional Sanger-sequencing could be very time-consuming and labor intensive. Also, massively parallel sequencing is capable of performing complex genetic analysis in a short time which was not possible before. This technique changed the way of genomic research for sure [17], but clinical applications of next-generation sequencing is still could be difficult due to the need of more complex equipments and highly experienced personnel. However, the latest generation of the benchtop instruments is much more affordable and user-friendly, producing sequencing data (up to a few gigabases) in a day or two [18, 19]. Furthermore, the coverage of certain regions is often low or absent in whole genome or whole exome sequencing studies [20]. Different library preparation methods exist, thus, here we demonstrate a novel sequence capture method in a clinical setting. Targeting the coding sequences of the genes of interest provides high coverage in every captured region. All the deleterious or important results need to be validated yet using different methods, such as Sanger-sequencing which is still a gold standard to investigate germline variants.
Materials and methods Biological samples and DNA isolation Twenty patient with different familiar history were selected from our routine genetic counseling database and anonymized for this study. Genomic DNA was isolated from 200 ll of peripheral blood using Reliaprep Blood gDNA Miniprep System (Promega, Madison, WI). Briefly, the blood samples were digested with Proteinase K solution in the presence of Cell Lysis Buffer, and after 10 min of incubation at 56 °C, DNA was bound to ReliaPrepTM Binding Column. After three washes, DNA was eluted into 50 ll of nuclease-free water. The concentration of the isolated DNA was determined with Qubit dsDNA HS Assay Kit (Life Technologies, Carlsbad, CA). The study was approved by the Semmelweis University’s Regional
123
and Institutional Committee of Science and Research Ethics (TUKEB-7-1/2008), and all patients gave written informed consent. Capture design The target list was carefully prepared based on literature data and the information on the NIH Genetic Home Reference site (http://ghr.nlm.nih.gov/). SureDesign software (Agilent, Santa Clara, CA) was used to design the custom HaloPlex capture assay. The following genes RefSeq coding exons were targeted with an extra 10 bases upstream and downstream: ATM (99.09 %), BARD1 (100 %), BRCA1 (100 %), BRCA2 (99.40 %), BRIP1 (100 %), CDH1 (99.76 %), CHEK2 (96.07 %), DIRAS3 (100 %), ERBB2 (100 %), KRAS (100 %), NBN (100 %), PALB2 (100 %), RAD50 (99.77 %), RAD51C (100 %), STK11 (100 %), TP53 (100 %) with 99.59 % of total coverage. Total amplicon number was 6123, target size 62.80 kb and the size of the total design was 164.03 kb. Sequence capture and library preparation For sequence capture, HaloPlex Target Enrichment System Kit, ION (Agilent) was used, according to the manufacturer’s instructions [21]. Briefly, in the first step, gDNA samples are digested in eight different restriction reactions, each containing two restriction enzymes, to create a library of gDNA restriction fragments. In the second step, the collection of gDNA restriction fragments is hybridized to the HaloPlex probe capture library. HaloPlex probes are designed to hybridize selectively to fragments originating from target regions of the genome and to direct circularization of the targeted DNA fragments. During the hybridization process, Ion Torrent sequencing motifs, including IonXpress barcode sequences, are incorporated into the targeted fragments. In the third, capture step, the circularized target DNA-HaloPlex probe hybrids, containing biotin, are captured on streptavidin beads. In the next step, DNA ligase is added to the capture reaction to close nicks in the circularized HaloPlex probe-target DNA hybrids. The following step is to elute the captured DNA libraries from the magnetic beads. The final step of the library preparation is the PCR amplification of the captured target libraries. The amplified target DNA was purified twice, using AMPure XP beads (Beckman Coulter). The concentration of the captured libraries was determined with Qubit dsDNA HS Assay Kit (Life Technologies). Ion Torrent sequencing The HaloPlex libraries were diluted to 26 pM concentration, than 20 ll of diluted library was added into the emulsion-
Rapid and cost effective screening of breast and ovarian cancer genes
PCR with Ion Sphere Particles (ISPs) using automated template preparation on Ion One Touch (Life Technologies) instrument with Ion One Touch v2 DL kit (Life Technologies). As a result of this reaction, amplicons were clonally amplified and bound to the surface of the ISPs. Non-templated beads were removed from the solution in an automated enrichment step using Ion One Touch ES instrument (Life Technologies). ISPs were loaded into Ion 316 chips and the sequencing runs were performed using Ion PGM 200 Sequencing kit (Life Technologies) with 500 flows. Validation Sanger sequencing The PCR primers were designed using Primer3Plus (http:// primer3plus.com/) software. Roche FastStart TaqMan Probe Master (Roche) kit was used to amplify the target regions and the PCR program was as follows: 95 °C 10 min, 40 cycles of 95 °C 30 s, 60 °C 30 s, 72 °C 45 s and the final step was 72 °C 5 min. PCR products were enzymatically cleaned using ExoSAP IT (Affymetrix, Santa Clara, CA) according to the manufacturer’s instructions. Sanger sequencing was performed using BigDye Terminator v3.1 Cycle Sequencing Kit (Life Technologies) using an ABI 3130 instrument (Life Technologies). Data analysis Data from the Ion Torrent runs were analyzed using the platform-specific pipeline software Torrent Suite v3.2.1 for base-calling, trim adapter and primer sequences, filter out poor quality reads, and de-multiplex the reads according to the barcode sequences. Briefly, TMAP (https://github.com/ iontorrent/TMAP) algorithm was used to align the reads to the hg19 human reference genom, then the variantCaller plugin was selected to run to search for germline variants in the targeted regions. The analysis pipeline has not any PCR-duplicates removing step, because of the enzymatic DNA fragmentation and the post-capture PCR amplification makes identifying PCR duplicates almost not even possible. Integrative Genomics Viewer was used for visualization of the mapped reads [22]. Variants were reviewed and annotated using dbSNP (http://www.ncbi.nlm.nih.gov/ projects/SNP/) and BIC (http://research.nhgri.nih.gov/pro jects/bic/) database. For variant interpretation, Ingenuity Variant Analysis Pipeline (Ingenuity, Rewood City, CA) was also used. Pathogenic status of the variant was stated if it was a missense variant with\1 % minor allele frequency and/or the variant was listed in the literature or in the databases as a pathogenic alteration. Called and deleterious variants were confirmed by Sanger-sequencing. The Sanger sequence data were investigated using ABI Sequence Scanner 1.0 (Life Technologies) and BioEdit (http://www. mbio.ncsu.edu/bioedit/bioedit.html).
Results We achieved 75–87 % chip loading density and the sequencing runs generated read numbers between 2.7 million to 3.5 million per chip. The average read number was 494787 (92647-890855) reads per sample, generating 40 Mb (9–74 Mb) of sequencing data per sample. The average base coverage depth was 716 (120–1,335) and the average 19 target coverage was about 99 % with a mean raw accuracy over 99 %. The uniformity of coverage was between 68.8 and 96.1 % (Supplementary table 1). The variability of uniformity of coverage may come from the unequal design of HaloPlex amplicons, regions covered by more amplicons achieved higher number of reads compared to target regions covered by less amplicons (Fig. 1). We identified 76 different variants in our samples (Supplementary table 2). We have identified two exons (ATM exon 59 and RAD51C exon 6 (Supplementary figure 1 and 2) across all samples which are tend to get less than 10 reads even if the whole target area are covered couple of hundred-fold. Our goal was the achive 209 coverage at least 80 % of the target area and our results shows that, we achieved this depth of coverage over 90 % of the targeted regions (the average 209 coverage was 96.45 %). Despite of the variety of the number of the on-target reads and the depth of coverage (Supplementary figure 3) the average 19 and 209 coverage and the uniformity of coverage shows only minor differences (Supplementary figure 4). The most polymorphic gene was BRCA2 with 17 different variants and the following genes contained no called variants at all: CDH1, RAD51C. Deleterious mutations and variants with 100 % frequency were validated using Sanger-sequencing, and all of the called variants were confirmed. We have found two different pathogenic BRCA1 mutations: c.181T [ G and c.2587delG. We identified a novel deletion in PALB2 (c.658delA) (Fig. 2), two deleterious variants in TP53 (c.637C[T and c.737T[C), one pathogenic mutation in ATM (c.6154G[A) and one pathogenic CHEK2 allele (c.470T[C) (Table 1). We chose two samples for re-capture and re-sequence to investigate the reproducibility of the custom capture assay, the called variants list was identical compared to the first capture and sequencing reaction. We have summarized the basic sequencing run metrics of each individual sample in Table 2.
Discussion Here, we demonstrated a novel sequence capture method to analyze known susceptibility genes in breast and ovarian cancer in clinical settings utilizing the capability of benchtop next-generation sequencing technique. Our goal
123
´ rvai et al. K. A
Fig. 1 Unequal HaloPlex amplicon design can lead higher variability in target coverage. The achieved highest coverage (upper gray bars) was 3,219-fold and the lowest coverage was 353-fold in this region
Fig. 2 Identifying and confirming novel PALB2 c.658delA mutation using the described assay. Visualizing the alignment of the sequencing reads covering the novel PALB2 deletion. Total coverage was 278-fold: G 3-fold (1 %, 1?, 2-), T 275-fold (99 %, 225?, 50-), del
152-fold (a). Validating our finding with Sanger-sequencing, red arrow indicates the position of the deletion. The deletion is present in both directions (b)
was to investigate the feasibility of the application of such a novel method for routine screening of patients, who probably have higher risk to develop a malignant disease.
The aim of this study was not to search for new predisposing genes or other elements, but perform a more complex risk assessment using our present knowledge, than the
123
Rapid and cost effective screening of breast and ovarian cancer genes Table 1 List of pathogenic variants in our samples
Gene ID
dbSNP
Coding variant
Protein effect
Variant effect
MAF*
ATM
–
c.6154G[A
p.Glu2052Lys
Deleterious
–
BRCA1
rs28897672
300T[G (c.181T[G)
p.Cys61Gly
Deleterious
–
BRCA1
–
2706delG (c.2587delG)
p.Val863Phefs*30
Deleterious
–
CHEK2
rs17879961
c.470T[C
p.Ile157Thr
Deleterious
–
PALB2
–
c.658delA
p.Ser220 fs*
Deleterious
–
*The minor allele frequency is the second most frequent allele value
TP53
–
c.737T[C
p.Met246Arg
Deleterious
–
TP53
–
c.637C[T
p.Arg213 = *
Deleterious
–
Table 2 Sequencing run metrics of each individual sample
Sample #
Number of ontarget reads
Average base coverage depth
19 coverage (%)
209 coverage (%)
Uniformity of base coverage (%)
Number of variants
1
890,855
1,335
99.10
97.10
86.80
25
2
660,111
923
98.46
94.21
73.01
19
3
391,200
577
98.70
94.90
83.74
27
4
813,106
713
98.26
94.16
76.24
29
5
406,871
603
98.47
94.90
80.13
26
6
362,491
538
98.22
94.53
82.45
25
7
733,356
1,066
98.84
96.57
89.92
23
8
267,504
403
99.70
99.11
96.11
19
9
264,623
399
99.44
97.34
90.86
26
10 11
515,458 795,440
778 965
99.24 97.43
97.99 90.72
90.61 68.82
29 23
12
92,647
120
98.39
92.06
90.46
23
13
168,181
223
99.10
94.99
89.47
23
14
288,580
416
99.62
96.91
90.35
36
15
565,061
879
99.23
97.26
86.66
27
16
591,377
930
99.70
99.52
95.49
34
17
442,211
752
99.69
99.49
94.53
31
18
646,496
1,125
99.70
99.51
94.51
26
19
791,078
1,262
99.70
99.52
94.29
21
20
209,109
332
99.61
98.33
94.68
24
ordinary BRCA1 and BRCA2 genetic test could do. Considering the recently published more conventional PCR approach to amplify the coding regions of BRCA1/2 genes [23], we chose a restriction digestion and probe hybridization-based sequence capture instead, to bypass the large number of primers required for amplification of total 289 coding exons. Multiplex PCR approaches often drop a few regions during the primer design, and that is not acceptable with a clinical purpose in mind and time-consuming optimization of the PCR amplicons that is often inevitable. Our custom sequence capture design targets almost every nucleotide (99.59 %) of coding exons of the genes of interest. However, one of the limitations of this technique may come from the unequal probe design of the assay: regions which are covered by more overlapping amplicons will
achieve higher coverage compared to regions covered by less amplicons, and that may lead to lower uniformity of coverage and makes it more complicated to calculate the proper number of necessary sequence reads to reach the level of clinical sensitivity. The second limitation could be the need of knowledge of variant interpretation compared to conventional Sanger-sequencing. Both false negatives and positives occur very rarely using Sanger-sequencing on blood-derived DNA samples but using next-generation sequencing both false cases could arise due the higher error-rate of this technique [24]. False positive variant calls tend to be found in the same location across samples and often originate from PCR artifacts. It is possible to filter them out after ample experience gained regarding the applied methods. Reducing the use of PCR may reduce the probability of a false positive variant calls. False negatives
123
´ rvai et al. K. A
could be decreased by alleviating the parameters in the variant calling algorithm. As we were investigating germline variants we does not looked for variants with low allelic frequency. The applied sequencing technology has some difficulties to determine to correct length of certain homopolimer regions, if the analysis pipeline called a variant in homopolimer area we designed a Sanger assay to validate such variants. Overall, we have identified 76 different variants in the 20 samples, 4 variants were intronic, 22 variants were synonymous, 7 variants were pathogenic and 43 were missense polymorphisms. We have found known deleterious mutations in BRCA1 (c.181T[G and c.2587delG), in TP53 (c.637C[T and c.737T[C) in ATM (c.6154G[A) and in CHEK2 (c.470T[C). The BRCA1 c.181T[G mutation is one of the most common pathogenic mutation in the Hungarian population and the BRCA1 c.2587delG also reported previously in Hungary [25]. We identified a novel frameshift mutation caused by deletion in PALB2 (c.658delA) gene. The patient with this novel mutation had breast, ovarian, and lung cancer and this novel alteration may have contributed to the development of these diseases. We have found 5 variants with 100 % frequency across our samples and this result is corresponding to the known allele frequencies in the CEU region. All of the identified variants have been confirmed by Sanger sequencing. No falsepositive Ion Torrent variant has been seen after filtering. The rapid and cost effective analysis of this 16-gene cohort can reveal the genetic background of approximately 30 % of hereditary and familiar cases of breast and ovarian cancers. Thus, it opens up a new and high-throughput approach with fast turnaround time to the genetic diagnostics of these disorders and may be helpful to investigate other familial genetic disorders as well. The interpretation of incidental findings (which is a less important issue regarding targeted sequencing) and variants with uncertain significance still remains a challenge, but our conclusion is similar to others [26, 27]: next-generation sequencing is a more viable and accurate option for hereditary cancer predisposition testing with a similar or lower cost and faster turnaround time compared to traditional methods. Conflict of interest
All authors claim no conflict of interest.
References 1. Wooster R, Weber BL (2003) Breast and ovarian cancer. New Engl J Med 348(23):2339–2347. doi:10.1056/NEJMra012284 2. Levy-Lahad E, Friedman E (2007) Cancer risks among BRCA1 and BRCA2 mutation carriers. Br J Cancer 96(1):11–15. doi:10. 1038/sj.bjc.6603535
123
3. Frank TS, Deffenbaugh AM, Reid JE, Hulick M, Ward BE, Lingenfelter B, Gumpper KL, Scholl T, Tavtigian SV, Pruss DR, Critchfield GC (2002) Clinical characteristics of individuals with germline mutations in BRCA1 and BRCA2: analysis of 10,000 individuals. J Clin Oncol 20(6):1480–1490 4. Carraro DM, Koike Folgueira MA, Garcia Lisboa BC, Ribeiro Olivieri EH, Vitorino Krepischi AC, de Carvalho AF, de Carvalho Mota LD, Puga RD, do Socorro Maciel M, Michelli RA, de de Lyra EC, Grosso SH, Soares FA, de Souza Waddington Achatz MI, Brentani H, Moreira-Filho CA, Brentani MM (2013) Comprehensive analysis of BRCA1, BRCA2 and TP53 germline mutation and tumor characterization: a portrait of early-onset breast cancer in Brazil. PLoS One 8(3):e57581. doi:10.1371/ journal.pone.0057581 5. Guenard F, Pedneault CS, Ouellette G, Labrie Y, Simard J, Durocher F (2010) Evaluation of the contribution of the three breast cancer susceptibility genes CHEK2, STK11, and PALB2 in non-BRCA1/2 French Canadian families with high risk of breast cancer. Genet Test Mol Biomark 14(4):515–526. doi:10. 1089/gtmb.2010.0027 6. Catucci I, Milgrom R, Kushnir A, Laitman Y, Paluch-Shimon S, Volorio S, Ficarazzi F, Bernard L, Radice P, Friedman E, Peterlongo P (2012) Germline mutations in BRIP1 and PALB2 in Jewish high cancer risk families. Fam Cancer 11(3):483–491. doi:10.1007/s10689-012-9540-8 7. Coulet F, Fajac A, Colas C, Eyries M, Dion-Miniere A, Rouzier R, Uzan S, Lefranc JP, Carbonnel M, Cornelis F, Cortez A, Soubrier F (2013) Germline RAD51C mutations in ovarian cancer susceptibility. Clin Genet 83(4):332–336. doi:10.1111/j.13990004.2012.01917.x 8. Tischkowitz M, Sabbaghian N, Hamel N, Pouchet C, Foulkes WD, Mes-Masson AM, Provencher DM, Tonin PN (2013) Contribution of the PALB2 c.2323C[T [p. Q775X] founder mutation in well-defined breast and/or ovarian cancer families and unselected ovarian cancer cases of French Canadian descent. BMC Med Genet 14:5. doi:10.1186/1471-2350-14-5 9. Gracia-Aznarez FJ, Fernandez V, Pita G, Peterlongo P, Dominguez O, de la Hoya M, Duran M, Osorio A, Moreno L, GonzalezNeira A, Rosa-Rosa JM, Sinilnikova O, Mazoyer S, Hopper J, Lazaro C, Southey M, Odefrey F, Manoukian S, Catucci I, Caldes T, Lynch HT, Hilbers FS, van Asperen CJ, Vasen HF, Goldgar D, Radice P, Devilee P, Benitez J (2013) Whole exome sequencing suggests much of non-BRCA1/BRCA2 familial breast cancer is due to moderate and low penetrance susceptibility alleles. PLoS One 8(2):e55681. doi:10.1371/journal.pone.0055681 10. Weischer M, Nordestgaard BG, Pharoah P, Bolla MK, Nevanlinna H, Van’t Veer LJ, Garcia-Closas M, Hopper JL, Hall P, Andrulis IL, Devilee P, Fasching PA, Anton-Culver H, Lambrechts D, Hooning M, Cox A, Giles GG, Burwinkel B, Lindblom A, Couch FJ, Mannermaa A, Grenaker Alnaes G, John EM, Dork T, Flyger H, Dunning AM, Wang Q, Muranen TA, van Hien R, Figueroa J, Southey MC, Czene K, Knight JA, Tollenaar RA, Beckmann MW, Ziogas A, Christiaens MR, Collee JM, Reed MW, Severi G, Marme F, Margolin S, Olson JE, Kosma VM, Kristensen VN, Miron A, Bogdanova N, Shah M, Blomqvist C, Broeks A, Sherman M, Phillips KA, Li J, Liu J, Glendon G, Seynaeve C, Ekici AB, Leunen K, Kriege M, Cross SS, Baglietto L, Sohn C, Wang X, Kataja V, Borresen-Dale AL, Meyer A, Easton DF, Schmidt MK, Bojesen SE (2012) CHEK2*1100delC heterozygosity in women with breast cancer associated with early death, breast cancer-specific death, and increased risk of a second breast cancer. J Clin Oncol 30(35):4308–4316. doi:10.1200/jco. 2012.42.7336 11. Southey MC, Teo ZL, Winship I (2013) PALB2 and breast cancer: ready for clinical translation! Appl Clin Genet 6:43–52. doi:10.2147/tacg.s34116
Rapid and cost effective screening of breast and ovarian cancer genes 12. Benusiglio PR, Malka D, Rouleau E, De Pauw A, Buecher B, Nogues C, Fourme E, Colas C, Coulet F, Warcoin M, Grandjouan S, Sezeur A, Laurent-Puig P, Moliere D, Tlemsani C, Di Maria M, Byrde V, Delaloge S, Blayau M, Caron O (2013) CDH1 germline mutations and the hereditary diffuse gastric and lobular breast cancer syndrome: a multicentre study. J Med Genet 50(7):486–489. doi:10.1136/jmedgenet-2012-101472 13. Kluijt I, Siemerink EJ, Ausems MG, van Os TA, de Jong D, Simoes-Correia J, van Krieken JH, Ligtenberg MJ, Figueiredo J, van Riel E, Sijmons RH, Plukker JT, van Hillegersberg R, Dekker E, Oliveira C, Cats A, Hoogerbrugge N (2012) CDH1-related hereditary diffuse gastric cancer syndrome: clinical variations and implications for counseling. Int J Cancer 131(2):367–376. doi:10.1002/ijc.26398 14. Iqbal J, Ragone A, Lubinski J, Lynch HT, Moller P, Ghadirian P, Foulkes WD, Armel S, Eisen A, Neuhausen SL, Senter L, Singer CF, Ainsworth P, Kim-Sing C, Tung N, Friedman E, Llacuachaqui M, Ping S, Narod SA (2012) The incidence of pancreatic cancer in BRCA1 and BRCA2 mutation carriers. Br J Cancer 107(12):2005–2009. doi:10.1038/bjc.2012.483 15. Mocci E, Milne RL, Mendez-Villamil EY, Hopper JL, John EM, Andrulis IL, Chung WK, Daly M, Buys SS, Malats N, Goldgar DE (2013) Risk of pancreatic cancer in breast cancer families from the breast cancer family registry. Cancer Epidemiol Biomark Prevent 22(5):803–811. doi:10.1158/1055-9965.EPI-120195 16. Castro E, Goh C, Olmos D, Saunders E, Leongamornlert D, Tymrakiewicz M, Mahmud N, Dadaev T, Govindasami K, Guy M, Sawyer E, Wilkinson R, Ardern-Jones A, Ellis S, Frost D, Peock S, Evans DG, Tischkowitz M, Cole T, Davidson R, Eccles D, Brewer C, Douglas F, Porteous ME, Donaldson A, Dorkins H, Izatt L, Cook J, Hodgson S, Kennedy MJ, Side LE, Eason J, Murray A, Antoniou AC, Easton DF, Kote-Jarai Z, Eeles R (2013) Germline BRCA mutations are associated with higher risk of nodal involvement, distant metastasis, and poor survival outcomes in prostate cancer. J Clin Oncol 31(14):1748–1757. doi:10. 1200/jco.2012.43.1882 17. Need AC, Shashi V, Hitomi Y, Schoch K, Shianna KV, McDonald MT, Meisler MH, Goldstein DB (2012) Clinical application of exome sequencing in undiagnosed genetic conditions. J Med Genet 49(6):353–361. doi:10.1136/jmedgenet-2012100819 18. Rothberg JM, Hinz W, Rearick TM, Schultz J, Mileski W, Davey M, Leamon JH, Johnson K, Milgrew MJ, Edwards M, Hoon J,
19.
20.
21. 22.
23.
24.
25.
26.
27.
Simons JF, Marran D, Myers JW, Davidson JF, Branting A, Nobile JR, Puc BP, Light D, Clark TA, Huber M, Branciforte JT, Stoner IB, Cawley SE, Lyons M, Fu Y, Homer N, Sedova M, Miao X, Reed B, Sabina J, Feierstein E, Schorn M, Alanjary M, Dimalanta E, Dressman D, Kasinskas R, Sokolsky T, Fidanza JA, Namsaraev E, McKernan KJ, Williams A, Roth GT, Bustillo J (2011) An integrated semiconductor device enabling non-optical genome sequencing. Nature 475(7356):348–352. doi:10.1038/ nature10242 Merriman B, Rothberg JM (2012) Progress in ion torrent semiconductor chip based sequencing. Electrophoresis 33(23):3397– 3417. doi:10.1002/elps.201200424 Walsh T, Lee MK, Casadei S, Thornton AM, Stray SM, Pennil C, Nord AS, Mandell JB, Swisher EM, King MC (2010) Detection of inherited mutations for breast and ovarian cancer using genomic capture and massively parallel sequencing. Proc Natl Acad Sci USA 107(28):12629–12633. doi:10.1073/pnas. 1007983107 Agilent (2014) Haloplex target enrichment protocol Thorvaldsdottir H, Robinson JT, Mesirov JP (2013) Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinf 14(2):178–192. doi:10.1093/bib/bbs017 Costa JL, Sousa S, Justino A, Kay T, Fernandes S, Cirnes L, Schmitt F, Machado JC (2013) Nonoptical massive parallel DNA sequencing of BRCA1 and BRCA2 genes in a diagnostic setting. Hum Mutat 34(4):629–635. doi:10.1002/humu.22272 Quail MA, Smith M, Coupland P, Otto TD, Harris SR, Connor TR, Bertoni A, Swerdlow HP, Gu Y (2012) A tale of three next generation sequencing platforms: comparison of Ion Torrent, Pacific Biosciences and Illumina MiSeq sequencers. BMC Genom 13:341. doi:10.1186/1471-2164-13-341 Van Der Looij M, Szabo C, Besznyak I, Liszka G, Csokay B, Pulay T, Toth J, Devilee P, King MC, Olah E (2000) Prevalence of founder BRCA1 and BRCA2 mutations among breast and ovarian cancer patients in Hungary. Int J Cancer 86(5):737–740 Ulahannan D, Kovac MB, Mulholland PJ, Cazier JB, Tomlinson I (2013) Technical and implementation issues in using next-generation sequencing of cancers in clinical practice. Br J Cancer 109(4):827–835. doi:10.1038/bjc.2013.416 Ku CS, Cooper DN, Iacopetta B, Roukos DH (2013) Integrating next-generation sequencing into the diagnostic testing of inherited cancer predisposition. Clin Genet 83(1):2–6. doi:10.1111/ cge.12028
123
ORIGINAL ARTICLE
Rapid and cost effective screening of breast and ovarian cancer genes using novel sequence capture method in clinical samples ´ rvai • Pe´ter Horva´th • Bernadett Balla • Kristo´f A Anna M. T} oke´s • Ba´lint Tobia´s • Istva´n Taka´cs • Zsolt Nagy • Pe´ter Lakatos • Ja´nos P. Ko´sa
Ó Springer Science+Business Media Dordrecht 2014
Abstract BRCA1 and BRCA2 are two well-known genes in the background of hereditary breast and ovarian cancer. There is also evidence that several other genes play an important role in the pathogenesis of these two malignancies. Latest population-scaled studies showed that certain mutations in different genes could cause similar risk elevation like BRCA2 mutations. In this study we present a new method to analyse the risk assessment of women to breast and ovarian cancer. Using Haloplex, a novel sequence capture method combined with next-generation sequencing we were able to perform rapid and cost-effective screening of 16 genes that could be associated with an increased risk of breast and ovarian cancer. The rapid and cost effective analysis of this 16-gene cohort can reveal the genetic background of approximately 30 % of hereditary and familiar cases of breast and ovarian cancers. Thus, it opens up a new and high-throughput approach with fast turnaround time to the genetic diagnostics of these
Pe´ter Lakatos and Ja´nos P. Ko´sa have equally contributed to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s10689-014-9730-7) contains supplementary material, which is available to authorized users. ´ rvai P. Horva´th J. P. Ko´sa (&) K. A 1st Department of Internal Medicine, Semmelweis University, Budapest, Hungary e-mail: [email protected] ´ rvai B. Balla B. Tobia´s I. Taka´cs Z. Nagy K. A P. Lakatos J. P. Ko´sa PentaCore Laboratory, Budapest, Hungary A. M. T}oke´s MTA-SE Tumor Progression Research Group, SE-2nd Department of Pathology, Budapest, Hungary
disorders and may be helpful to investigate other familial genetic disorders as well. Keywords Breast cancer Next generation sequencing Ovarian cancer Tumor genetics
Introduction So far, two prime high-penetrance genes associated with susceptibility to breast and ovarian cancer—breast cancer susceptibility gene 1 (BRCA1) and breast cancer susceptibility gene 2 (BRCA2)—have been referred. The presence of a deleterious mutation in any of these genes means a lifetime risk to develop breast cancer with a chance between 60–85 % and a 15–40 % risk for ovarian cancer, respectively [1]. A vast array of evidence indicates that other high- or low-penetrance susceptibility mutations may increase the risk of breast or ovarian cancer. Probably several variants in low-penetrance susceptibility genes are more common than mutations in high-penetrance alleles, hence low-penetrance alleles could add more risk to develop breast or ovarian cancer [2, 3]. The aim of research efforts to find new genes to explain the missing heritability in BRCA negative cancer families targeted genes that could interact with the BRCA pathway or the BRCA proteins themselves. Additional high penetrance alleles were found for breast or ovarian cancer, such as TP53 [4] and STK11 [5]. The genetic elements of Fanconi Anemia pathway have also been found as moderate penetrance alleles in these cancers including PALB2, BRIP1, RAD51C [6–9]. Similar penetrance level has been identified in case of other genes which are also involved in the homologous recombination repair pathway, such as ATM and CHEK2. The CHEK2
123
´ rvai et al. K. A
truncating mutations seems to be double the breast cancer risk [10] and recent, population based discoveries suggest that ATM, CHEK2, BRIP1 and PALB2 mutations has similar effect like BRCA2 mutations [11]. Family history of diffuse gastric cancer also mean a higher risk for breast cancer [12] but a mutation in CDH1 gene could increase breast cancer risk without familial gastric cancer syndrome [13]. Moreover, recent evidence suggests that a pathogenic BRCA1/2 mutation could also increase the risk of pancreatic or prostate cancer [14–16]. However, about 70 % of breast or ovarian cancer inheritance is still missing. All the genes mentioned above have in general dozens of coding exons and a complete lack of certain hotspot regions, hence traditional Sanger-sequencing could be very time-consuming and labor intensive. Also, massively parallel sequencing is capable of performing complex genetic analysis in a short time which was not possible before. This technique changed the way of genomic research for sure [17], but clinical applications of next-generation sequencing is still could be difficult due to the need of more complex equipments and highly experienced personnel. However, the latest generation of the benchtop instruments is much more affordable and user-friendly, producing sequencing data (up to a few gigabases) in a day or two [18, 19]. Furthermore, the coverage of certain regions is often low or absent in whole genome or whole exome sequencing studies [20]. Different library preparation methods exist, thus, here we demonstrate a novel sequence capture method in a clinical setting. Targeting the coding sequences of the genes of interest provides high coverage in every captured region. All the deleterious or important results need to be validated yet using different methods, such as Sanger-sequencing which is still a gold standard to investigate germline variants.
Materials and methods Biological samples and DNA isolation Twenty patient with different familiar history were selected from our routine genetic counseling database and anonymized for this study. Genomic DNA was isolated from 200 ll of peripheral blood using Reliaprep Blood gDNA Miniprep System (Promega, Madison, WI). Briefly, the blood samples were digested with Proteinase K solution in the presence of Cell Lysis Buffer, and after 10 min of incubation at 56 °C, DNA was bound to ReliaPrepTM Binding Column. After three washes, DNA was eluted into 50 ll of nuclease-free water. The concentration of the isolated DNA was determined with Qubit dsDNA HS Assay Kit (Life Technologies, Carlsbad, CA). The study was approved by the Semmelweis University’s Regional
123
and Institutional Committee of Science and Research Ethics (TUKEB-7-1/2008), and all patients gave written informed consent. Capture design The target list was carefully prepared based on literature data and the information on the NIH Genetic Home Reference site (http://ghr.nlm.nih.gov/). SureDesign software (Agilent, Santa Clara, CA) was used to design the custom HaloPlex capture assay. The following genes RefSeq coding exons were targeted with an extra 10 bases upstream and downstream: ATM (99.09 %), BARD1 (100 %), BRCA1 (100 %), BRCA2 (99.40 %), BRIP1 (100 %), CDH1 (99.76 %), CHEK2 (96.07 %), DIRAS3 (100 %), ERBB2 (100 %), KRAS (100 %), NBN (100 %), PALB2 (100 %), RAD50 (99.77 %), RAD51C (100 %), STK11 (100 %), TP53 (100 %) with 99.59 % of total coverage. Total amplicon number was 6123, target size 62.80 kb and the size of the total design was 164.03 kb. Sequence capture and library preparation For sequence capture, HaloPlex Target Enrichment System Kit, ION (Agilent) was used, according to the manufacturer’s instructions [21]. Briefly, in the first step, gDNA samples are digested in eight different restriction reactions, each containing two restriction enzymes, to create a library of gDNA restriction fragments. In the second step, the collection of gDNA restriction fragments is hybridized to the HaloPlex probe capture library. HaloPlex probes are designed to hybridize selectively to fragments originating from target regions of the genome and to direct circularization of the targeted DNA fragments. During the hybridization process, Ion Torrent sequencing motifs, including IonXpress barcode sequences, are incorporated into the targeted fragments. In the third, capture step, the circularized target DNA-HaloPlex probe hybrids, containing biotin, are captured on streptavidin beads. In the next step, DNA ligase is added to the capture reaction to close nicks in the circularized HaloPlex probe-target DNA hybrids. The following step is to elute the captured DNA libraries from the magnetic beads. The final step of the library preparation is the PCR amplification of the captured target libraries. The amplified target DNA was purified twice, using AMPure XP beads (Beckman Coulter). The concentration of the captured libraries was determined with Qubit dsDNA HS Assay Kit (Life Technologies). Ion Torrent sequencing The HaloPlex libraries were diluted to 26 pM concentration, than 20 ll of diluted library was added into the emulsion-
Rapid and cost effective screening of breast and ovarian cancer genes
PCR with Ion Sphere Particles (ISPs) using automated template preparation on Ion One Touch (Life Technologies) instrument with Ion One Touch v2 DL kit (Life Technologies). As a result of this reaction, amplicons were clonally amplified and bound to the surface of the ISPs. Non-templated beads were removed from the solution in an automated enrichment step using Ion One Touch ES instrument (Life Technologies). ISPs were loaded into Ion 316 chips and the sequencing runs were performed using Ion PGM 200 Sequencing kit (Life Technologies) with 500 flows. Validation Sanger sequencing The PCR primers were designed using Primer3Plus (http:// primer3plus.com/) software. Roche FastStart TaqMan Probe Master (Roche) kit was used to amplify the target regions and the PCR program was as follows: 95 °C 10 min, 40 cycles of 95 °C 30 s, 60 °C 30 s, 72 °C 45 s and the final step was 72 °C 5 min. PCR products were enzymatically cleaned using ExoSAP IT (Affymetrix, Santa Clara, CA) according to the manufacturer’s instructions. Sanger sequencing was performed using BigDye Terminator v3.1 Cycle Sequencing Kit (Life Technologies) using an ABI 3130 instrument (Life Technologies). Data analysis Data from the Ion Torrent runs were analyzed using the platform-specific pipeline software Torrent Suite v3.2.1 for base-calling, trim adapter and primer sequences, filter out poor quality reads, and de-multiplex the reads according to the barcode sequences. Briefly, TMAP (https://github.com/ iontorrent/TMAP) algorithm was used to align the reads to the hg19 human reference genom, then the variantCaller plugin was selected to run to search for germline variants in the targeted regions. The analysis pipeline has not any PCR-duplicates removing step, because of the enzymatic DNA fragmentation and the post-capture PCR amplification makes identifying PCR duplicates almost not even possible. Integrative Genomics Viewer was used for visualization of the mapped reads [22]. Variants were reviewed and annotated using dbSNP (http://www.ncbi.nlm.nih.gov/ projects/SNP/) and BIC (http://research.nhgri.nih.gov/pro jects/bic/) database. For variant interpretation, Ingenuity Variant Analysis Pipeline (Ingenuity, Rewood City, CA) was also used. Pathogenic status of the variant was stated if it was a missense variant with\1 % minor allele frequency and/or the variant was listed in the literature or in the databases as a pathogenic alteration. Called and deleterious variants were confirmed by Sanger-sequencing. The Sanger sequence data were investigated using ABI Sequence Scanner 1.0 (Life Technologies) and BioEdit (http://www. mbio.ncsu.edu/bioedit/bioedit.html).
Results We achieved 75–87 % chip loading density and the sequencing runs generated read numbers between 2.7 million to 3.5 million per chip. The average read number was 494787 (92647-890855) reads per sample, generating 40 Mb (9–74 Mb) of sequencing data per sample. The average base coverage depth was 716 (120–1,335) and the average 19 target coverage was about 99 % with a mean raw accuracy over 99 %. The uniformity of coverage was between 68.8 and 96.1 % (Supplementary table 1). The variability of uniformity of coverage may come from the unequal design of HaloPlex amplicons, regions covered by more amplicons achieved higher number of reads compared to target regions covered by less amplicons (Fig. 1). We identified 76 different variants in our samples (Supplementary table 2). We have identified two exons (ATM exon 59 and RAD51C exon 6 (Supplementary figure 1 and 2) across all samples which are tend to get less than 10 reads even if the whole target area are covered couple of hundred-fold. Our goal was the achive 209 coverage at least 80 % of the target area and our results shows that, we achieved this depth of coverage over 90 % of the targeted regions (the average 209 coverage was 96.45 %). Despite of the variety of the number of the on-target reads and the depth of coverage (Supplementary figure 3) the average 19 and 209 coverage and the uniformity of coverage shows only minor differences (Supplementary figure 4). The most polymorphic gene was BRCA2 with 17 different variants and the following genes contained no called variants at all: CDH1, RAD51C. Deleterious mutations and variants with 100 % frequency were validated using Sanger-sequencing, and all of the called variants were confirmed. We have found two different pathogenic BRCA1 mutations: c.181T [ G and c.2587delG. We identified a novel deletion in PALB2 (c.658delA) (Fig. 2), two deleterious variants in TP53 (c.637C[T and c.737T[C), one pathogenic mutation in ATM (c.6154G[A) and one pathogenic CHEK2 allele (c.470T[C) (Table 1). We chose two samples for re-capture and re-sequence to investigate the reproducibility of the custom capture assay, the called variants list was identical compared to the first capture and sequencing reaction. We have summarized the basic sequencing run metrics of each individual sample in Table 2.
Discussion Here, we demonstrated a novel sequence capture method to analyze known susceptibility genes in breast and ovarian cancer in clinical settings utilizing the capability of benchtop next-generation sequencing technique. Our goal
123
´ rvai et al. K. A
Fig. 1 Unequal HaloPlex amplicon design can lead higher variability in target coverage. The achieved highest coverage (upper gray bars) was 3,219-fold and the lowest coverage was 353-fold in this region
Fig. 2 Identifying and confirming novel PALB2 c.658delA mutation using the described assay. Visualizing the alignment of the sequencing reads covering the novel PALB2 deletion. Total coverage was 278-fold: G 3-fold (1 %, 1?, 2-), T 275-fold (99 %, 225?, 50-), del
152-fold (a). Validating our finding with Sanger-sequencing, red arrow indicates the position of the deletion. The deletion is present in both directions (b)
was to investigate the feasibility of the application of such a novel method for routine screening of patients, who probably have higher risk to develop a malignant disease.
The aim of this study was not to search for new predisposing genes or other elements, but perform a more complex risk assessment using our present knowledge, than the
123
Rapid and cost effective screening of breast and ovarian cancer genes Table 1 List of pathogenic variants in our samples
Gene ID
dbSNP
Coding variant
Protein effect
Variant effect
MAF*
ATM
–
c.6154G[A
p.Glu2052Lys
Deleterious
–
BRCA1
rs28897672
300T[G (c.181T[G)
p.Cys61Gly
Deleterious
–
BRCA1
–
2706delG (c.2587delG)
p.Val863Phefs*30
Deleterious
–
CHEK2
rs17879961
c.470T[C
p.Ile157Thr
Deleterious
–
PALB2
–
c.658delA
p.Ser220 fs*
Deleterious
–
*The minor allele frequency is the second most frequent allele value
TP53
–
c.737T[C
p.Met246Arg
Deleterious
–
TP53
–
c.637C[T
p.Arg213 = *
Deleterious
–
Table 2 Sequencing run metrics of each individual sample
Sample #
Number of ontarget reads
Average base coverage depth
19 coverage (%)
209 coverage (%)
Uniformity of base coverage (%)
Number of variants
1
890,855
1,335
99.10
97.10
86.80
25
2
660,111
923
98.46
94.21
73.01
19
3
391,200
577
98.70
94.90
83.74
27
4
813,106
713
98.26
94.16
76.24
29
5
406,871
603
98.47
94.90
80.13
26
6
362,491
538
98.22
94.53
82.45
25
7
733,356
1,066
98.84
96.57
89.92
23
8
267,504
403
99.70
99.11
96.11
19
9
264,623
399
99.44
97.34
90.86
26
10 11
515,458 795,440
778 965
99.24 97.43
97.99 90.72
90.61 68.82
29 23
12
92,647
120
98.39
92.06
90.46
23
13
168,181
223
99.10
94.99
89.47
23
14
288,580
416
99.62
96.91
90.35
36
15
565,061
879
99.23
97.26
86.66
27
16
591,377
930
99.70
99.52
95.49
34
17
442,211
752
99.69
99.49
94.53
31
18
646,496
1,125
99.70
99.51
94.51
26
19
791,078
1,262
99.70
99.52
94.29
21
20
209,109
332
99.61
98.33
94.68
24
ordinary BRCA1 and BRCA2 genetic test could do. Considering the recently published more conventional PCR approach to amplify the coding regions of BRCA1/2 genes [23], we chose a restriction digestion and probe hybridization-based sequence capture instead, to bypass the large number of primers required for amplification of total 289 coding exons. Multiplex PCR approaches often drop a few regions during the primer design, and that is not acceptable with a clinical purpose in mind and time-consuming optimization of the PCR amplicons that is often inevitable. Our custom sequence capture design targets almost every nucleotide (99.59 %) of coding exons of the genes of interest. However, one of the limitations of this technique may come from the unequal probe design of the assay: regions which are covered by more overlapping amplicons will
achieve higher coverage compared to regions covered by less amplicons, and that may lead to lower uniformity of coverage and makes it more complicated to calculate the proper number of necessary sequence reads to reach the level of clinical sensitivity. The second limitation could be the need of knowledge of variant interpretation compared to conventional Sanger-sequencing. Both false negatives and positives occur very rarely using Sanger-sequencing on blood-derived DNA samples but using next-generation sequencing both false cases could arise due the higher error-rate of this technique [24]. False positive variant calls tend to be found in the same location across samples and often originate from PCR artifacts. It is possible to filter them out after ample experience gained regarding the applied methods. Reducing the use of PCR may reduce the probability of a false positive variant calls. False negatives
123
´ rvai et al. K. A
could be decreased by alleviating the parameters in the variant calling algorithm. As we were investigating germline variants we does not looked for variants with low allelic frequency. The applied sequencing technology has some difficulties to determine to correct length of certain homopolimer regions, if the analysis pipeline called a variant in homopolimer area we designed a Sanger assay to validate such variants. Overall, we have identified 76 different variants in the 20 samples, 4 variants were intronic, 22 variants were synonymous, 7 variants were pathogenic and 43 were missense polymorphisms. We have found known deleterious mutations in BRCA1 (c.181T[G and c.2587delG), in TP53 (c.637C[T and c.737T[C) in ATM (c.6154G[A) and in CHEK2 (c.470T[C). The BRCA1 c.181T[G mutation is one of the most common pathogenic mutation in the Hungarian population and the BRCA1 c.2587delG also reported previously in Hungary [25]. We identified a novel frameshift mutation caused by deletion in PALB2 (c.658delA) gene. The patient with this novel mutation had breast, ovarian, and lung cancer and this novel alteration may have contributed to the development of these diseases. We have found 5 variants with 100 % frequency across our samples and this result is corresponding to the known allele frequencies in the CEU region. All of the identified variants have been confirmed by Sanger sequencing. No falsepositive Ion Torrent variant has been seen after filtering. The rapid and cost effective analysis of this 16-gene cohort can reveal the genetic background of approximately 30 % of hereditary and familiar cases of breast and ovarian cancers. Thus, it opens up a new and high-throughput approach with fast turnaround time to the genetic diagnostics of these disorders and may be helpful to investigate other familial genetic disorders as well. The interpretation of incidental findings (which is a less important issue regarding targeted sequencing) and variants with uncertain significance still remains a challenge, but our conclusion is similar to others [26, 27]: next-generation sequencing is a more viable and accurate option for hereditary cancer predisposition testing with a similar or lower cost and faster turnaround time compared to traditional methods. Conflict of interest
All authors claim no conflict of interest.
References 1. Wooster R, Weber BL (2003) Breast and ovarian cancer. New Engl J Med 348(23):2339–2347. doi:10.1056/NEJMra012284 2. Levy-Lahad E, Friedman E (2007) Cancer risks among BRCA1 and BRCA2 mutation carriers. Br J Cancer 96(1):11–15. doi:10. 1038/sj.bjc.6603535
123
3. Frank TS, Deffenbaugh AM, Reid JE, Hulick M, Ward BE, Lingenfelter B, Gumpper KL, Scholl T, Tavtigian SV, Pruss DR, Critchfield GC (2002) Clinical characteristics of individuals with germline mutations in BRCA1 and BRCA2: analysis of 10,000 individuals. J Clin Oncol 20(6):1480–1490 4. Carraro DM, Koike Folgueira MA, Garcia Lisboa BC, Ribeiro Olivieri EH, Vitorino Krepischi AC, de Carvalho AF, de Carvalho Mota LD, Puga RD, do Socorro Maciel M, Michelli RA, de de Lyra EC, Grosso SH, Soares FA, de Souza Waddington Achatz MI, Brentani H, Moreira-Filho CA, Brentani MM (2013) Comprehensive analysis of BRCA1, BRCA2 and TP53 germline mutation and tumor characterization: a portrait of early-onset breast cancer in Brazil. PLoS One 8(3):e57581. doi:10.1371/ journal.pone.0057581 5. Guenard F, Pedneault CS, Ouellette G, Labrie Y, Simard J, Durocher F (2010) Evaluation of the contribution of the three breast cancer susceptibility genes CHEK2, STK11, and PALB2 in non-BRCA1/2 French Canadian families with high risk of breast cancer. Genet Test Mol Biomark 14(4):515–526. doi:10. 1089/gtmb.2010.0027 6. Catucci I, Milgrom R, Kushnir A, Laitman Y, Paluch-Shimon S, Volorio S, Ficarazzi F, Bernard L, Radice P, Friedman E, Peterlongo P (2012) Germline mutations in BRIP1 and PALB2 in Jewish high cancer risk families. Fam Cancer 11(3):483–491. doi:10.1007/s10689-012-9540-8 7. Coulet F, Fajac A, Colas C, Eyries M, Dion-Miniere A, Rouzier R, Uzan S, Lefranc JP, Carbonnel M, Cornelis F, Cortez A, Soubrier F (2013) Germline RAD51C mutations in ovarian cancer susceptibility. Clin Genet 83(4):332–336. doi:10.1111/j.13990004.2012.01917.x 8. Tischkowitz M, Sabbaghian N, Hamel N, Pouchet C, Foulkes WD, Mes-Masson AM, Provencher DM, Tonin PN (2013) Contribution of the PALB2 c.2323C[T [p. Q775X] founder mutation in well-defined breast and/or ovarian cancer families and unselected ovarian cancer cases of French Canadian descent. BMC Med Genet 14:5. doi:10.1186/1471-2350-14-5 9. Gracia-Aznarez FJ, Fernandez V, Pita G, Peterlongo P, Dominguez O, de la Hoya M, Duran M, Osorio A, Moreno L, GonzalezNeira A, Rosa-Rosa JM, Sinilnikova O, Mazoyer S, Hopper J, Lazaro C, Southey M, Odefrey F, Manoukian S, Catucci I, Caldes T, Lynch HT, Hilbers FS, van Asperen CJ, Vasen HF, Goldgar D, Radice P, Devilee P, Benitez J (2013) Whole exome sequencing suggests much of non-BRCA1/BRCA2 familial breast cancer is due to moderate and low penetrance susceptibility alleles. PLoS One 8(2):e55681. doi:10.1371/journal.pone.0055681 10. Weischer M, Nordestgaard BG, Pharoah P, Bolla MK, Nevanlinna H, Van’t Veer LJ, Garcia-Closas M, Hopper JL, Hall P, Andrulis IL, Devilee P, Fasching PA, Anton-Culver H, Lambrechts D, Hooning M, Cox A, Giles GG, Burwinkel B, Lindblom A, Couch FJ, Mannermaa A, Grenaker Alnaes G, John EM, Dork T, Flyger H, Dunning AM, Wang Q, Muranen TA, van Hien R, Figueroa J, Southey MC, Czene K, Knight JA, Tollenaar RA, Beckmann MW, Ziogas A, Christiaens MR, Collee JM, Reed MW, Severi G, Marme F, Margolin S, Olson JE, Kosma VM, Kristensen VN, Miron A, Bogdanova N, Shah M, Blomqvist C, Broeks A, Sherman M, Phillips KA, Li J, Liu J, Glendon G, Seynaeve C, Ekici AB, Leunen K, Kriege M, Cross SS, Baglietto L, Sohn C, Wang X, Kataja V, Borresen-Dale AL, Meyer A, Easton DF, Schmidt MK, Bojesen SE (2012) CHEK2*1100delC heterozygosity in women with breast cancer associated with early death, breast cancer-specific death, and increased risk of a second breast cancer. J Clin Oncol 30(35):4308–4316. doi:10.1200/jco. 2012.42.7336 11. Southey MC, Teo ZL, Winship I (2013) PALB2 and breast cancer: ready for clinical translation! Appl Clin Genet 6:43–52. doi:10.2147/tacg.s34116
Rapid and cost effective screening of breast and ovarian cancer genes 12. Benusiglio PR, Malka D, Rouleau E, De Pauw A, Buecher B, Nogues C, Fourme E, Colas C, Coulet F, Warcoin M, Grandjouan S, Sezeur A, Laurent-Puig P, Moliere D, Tlemsani C, Di Maria M, Byrde V, Delaloge S, Blayau M, Caron O (2013) CDH1 germline mutations and the hereditary diffuse gastric and lobular breast cancer syndrome: a multicentre study. J Med Genet 50(7):486–489. doi:10.1136/jmedgenet-2012-101472 13. Kluijt I, Siemerink EJ, Ausems MG, van Os TA, de Jong D, Simoes-Correia J, van Krieken JH, Ligtenberg MJ, Figueiredo J, van Riel E, Sijmons RH, Plukker JT, van Hillegersberg R, Dekker E, Oliveira C, Cats A, Hoogerbrugge N (2012) CDH1-related hereditary diffuse gastric cancer syndrome: clinical variations and implications for counseling. Int J Cancer 131(2):367–376. doi:10.1002/ijc.26398 14. Iqbal J, Ragone A, Lubinski J, Lynch HT, Moller P, Ghadirian P, Foulkes WD, Armel S, Eisen A, Neuhausen SL, Senter L, Singer CF, Ainsworth P, Kim-Sing C, Tung N, Friedman E, Llacuachaqui M, Ping S, Narod SA (2012) The incidence of pancreatic cancer in BRCA1 and BRCA2 mutation carriers. Br J Cancer 107(12):2005–2009. doi:10.1038/bjc.2012.483 15. Mocci E, Milne RL, Mendez-Villamil EY, Hopper JL, John EM, Andrulis IL, Chung WK, Daly M, Buys SS, Malats N, Goldgar DE (2013) Risk of pancreatic cancer in breast cancer families from the breast cancer family registry. Cancer Epidemiol Biomark Prevent 22(5):803–811. doi:10.1158/1055-9965.EPI-120195 16. Castro E, Goh C, Olmos D, Saunders E, Leongamornlert D, Tymrakiewicz M, Mahmud N, Dadaev T, Govindasami K, Guy M, Sawyer E, Wilkinson R, Ardern-Jones A, Ellis S, Frost D, Peock S, Evans DG, Tischkowitz M, Cole T, Davidson R, Eccles D, Brewer C, Douglas F, Porteous ME, Donaldson A, Dorkins H, Izatt L, Cook J, Hodgson S, Kennedy MJ, Side LE, Eason J, Murray A, Antoniou AC, Easton DF, Kote-Jarai Z, Eeles R (2013) Germline BRCA mutations are associated with higher risk of nodal involvement, distant metastasis, and poor survival outcomes in prostate cancer. J Clin Oncol 31(14):1748–1757. doi:10. 1200/jco.2012.43.1882 17. Need AC, Shashi V, Hitomi Y, Schoch K, Shianna KV, McDonald MT, Meisler MH, Goldstein DB (2012) Clinical application of exome sequencing in undiagnosed genetic conditions. J Med Genet 49(6):353–361. doi:10.1136/jmedgenet-2012100819 18. Rothberg JM, Hinz W, Rearick TM, Schultz J, Mileski W, Davey M, Leamon JH, Johnson K, Milgrew MJ, Edwards M, Hoon J,
19.
20.
21. 22.
23.
24.
25.
26.
27.
Simons JF, Marran D, Myers JW, Davidson JF, Branting A, Nobile JR, Puc BP, Light D, Clark TA, Huber M, Branciforte JT, Stoner IB, Cawley SE, Lyons M, Fu Y, Homer N, Sedova M, Miao X, Reed B, Sabina J, Feierstein E, Schorn M, Alanjary M, Dimalanta E, Dressman D, Kasinskas R, Sokolsky T, Fidanza JA, Namsaraev E, McKernan KJ, Williams A, Roth GT, Bustillo J (2011) An integrated semiconductor device enabling non-optical genome sequencing. Nature 475(7356):348–352. doi:10.1038/ nature10242 Merriman B, Rothberg JM (2012) Progress in ion torrent semiconductor chip based sequencing. Electrophoresis 33(23):3397– 3417. doi:10.1002/elps.201200424 Walsh T, Lee MK, Casadei S, Thornton AM, Stray SM, Pennil C, Nord AS, Mandell JB, Swisher EM, King MC (2010) Detection of inherited mutations for breast and ovarian cancer using genomic capture and massively parallel sequencing. Proc Natl Acad Sci USA 107(28):12629–12633. doi:10.1073/pnas. 1007983107 Agilent (2014) Haloplex target enrichment protocol Thorvaldsdottir H, Robinson JT, Mesirov JP (2013) Integrative Genomics Viewer (IGV): high-performance genomics data visualization and exploration. Brief Bioinf 14(2):178–192. doi:10.1093/bib/bbs017 Costa JL, Sousa S, Justino A, Kay T, Fernandes S, Cirnes L, Schmitt F, Machado JC (2013) Nonoptical massive parallel DNA sequencing of BRCA1 and BRCA2 genes in a diagnostic setting. Hum Mutat 34(4):629–635. doi:10.1002/humu.22272 Quail MA, Smith M, Coupland P, Otto TD, Harris SR, Connor TR, Bertoni A, Swerdlow HP, Gu Y (2012) A tale of three next generation sequencing platforms: comparison of Ion Torrent, Pacific Biosciences and Illumina MiSeq sequencers. BMC Genom 13:341. doi:10.1186/1471-2164-13-341 Van Der Looij M, Szabo C, Besznyak I, Liszka G, Csokay B, Pulay T, Toth J, Devilee P, King MC, Olah E (2000) Prevalence of founder BRCA1 and BRCA2 mutations among breast and ovarian cancer patients in Hungary. Int J Cancer 86(5):737–740 Ulahannan D, Kovac MB, Mulholland PJ, Cazier JB, Tomlinson I (2013) Technical and implementation issues in using next-generation sequencing of cancers in clinical practice. Br J Cancer 109(4):827–835. doi:10.1038/bjc.2013.416 Ku CS, Cooper DN, Iacopetta B, Roukos DH (2013) Integrating next-generation sequencing into the diagnostic testing of inherited cancer predisposition. Clin Genet 83(1):2–6. doi:10.1111/ cge.12028
123
