Exome sequencing and case-control analyses identify RCC1 as a candidate breast cancer susceptibility gene.

1 Exome sequencing and case-control analyses identify RCC1 as a candidate breast cancer susceptibility gene Aouatef Riahi (1,2), Hoda Radmanesh (1,3), Peter Schürmann (1), Natalia Bogdanova (1,4), Robert Geffers (5), Rym Meddeb (2,6), Maher Kharrat (2), Thilo Dörk (1) (1) Gynaecology Research Unit, Hannover Medical School, Hannover, Germany; (2) Laboratoire Génétique Humaine, Faculté de Médecine de Tunis, University Tunis El Manar, Tunis, Tunisia; (3) Medical Genetic Research Center (MGRC), School of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran; (4) Radiation Oncology Research Unit, Hannover Medical School, Hannover, Germany; (5) Genome Analytics Unit, Helmholtz Centre for Infection Research, Braunschweig, Germany; (6) Department of Hereditary and Congenital Disorders, Charles Nicolle Hospital, Tunis, Tunisia. Conflict of interest statement: The authors declare no potential conflicts of interest. Keywords: Breast carcinoma, replication, mitosis, founder mutation, RCC1 Contact: Aouatef Riahi, PhD/ Thilo Dörk, PhD Hannover Medical School Gynaecology Research Unit (OE 6411) Carl-Neuberg-Str. 1 D-30625 Hannover, Germany Phone: +44 511 532 6075 Fax: +44 511 532 6081 E-mail: [email protected] 2 Abstract: Breast cancer is a genetic disease but the known genes explain a minority of cases. To elucidate the molecular basis of breast cancer in the Tunisian population, we performed exome sequencing on six BRCA1/BRCA2 mutation-negative patients with familial breast cancer and identified a novel frameshift mutation in RCC1, encoding the Regulator of Chromosome Condensation 1. Subsequent genotyping detected the 19-bp deletion in additional 5 out of 153 (3%) breast cancer patients but in none of 400 female controls (p=0.0015). The deletion was enriched in patients with a positive family history (5%, p=0.0009) and co-segregated with breast cancer in the initial pedigree. The mutant allele was lost in 4/6 breast tumours from mutation carriers which may be consistent with the hypothesis that RCC1 dysfunction provides a selective disadvantage at the stage of tumour progression. In summary, we propose RCC1 as a likely breast cancer susceptibility gene in the Tunisian population. Novelty & Impact Statement: A truncating mutation in RCC1 was exclusively found in breast cancer patients from a founder population. RCC1 is important for replication control and proper mitotic spindle formation but has not previously been associated with hereditary cancer. Our results suggest RCC1 as a novel breast cancer susceptibility gene. 3 Introduction Breast cancer is a genetically heterogeneous disease with several susceptibility loci contributing to the disease.[1-4] Genomic biomarkers for breast cancer are comprised of rare highly penetrant mutations of genes such as BRCA1 or BRCA2, moderately penetrant mutations of genes such as ATM or CHEK2, as well as more common genomic variants, including single nucleotide polymorphisms, associated with modest effect sizes.[4] Founder populations have been very helpful to map and identify familial breast cancer susceptibility genes, including BRCA1, CHEK2, or PALB2, and whole exome sequencing continues to propose new candidates.[5,6] The products of these genes frequently cooperate in monitoring the fidelity of DNA replication, DNA damage repair and cell division. Because these pathways are complex and involve networks of hundreds of proteins, several more susceptibility genes are likely to exist. The presently known variants explain less than half of all breast cancers cases.[1,7] In Tunisia, the incidence of breast cancer has increased over the past years and now accounts for some 30% of all female cancers.[8] Founder mutations in BRCA1 and BRCA2 have been detected in about 25% of familial breast cancer patients but the majority of familial cases have remained unexplained.[9-11] In the present study, we performed exome sequencing of germline DNA from six BRCA1 and BRCA2 mutation-negative Tunisian patients with familial breast cancer to uncover additional susceptibility genes. A novel c.1067_1086del19 mutation in RCC1 was further genotyped in a breast cancer case-control association study. Methods Study Population The Tunisian patients included in this study (n=159, including 92 familial and 67 sporadic cases) had been referred to the Salah Azaiz Cancer Institute, Tunis, between 2008-2010 (48 familial and 67 sporadic cases) and between 2016-2017 (44 familial cases). The mean age at diagnosis was 46 years (range 33–81 years) for sporadic and 41 years (range 23-82 years) for familial cases, with 10 (15%) of the sporadic cases and 41 (45%) of the familial cases diagnosed below age 40. Among the sporadic cases, 28% of the tumours were ER-ve and 30% were PR-ve. Among 4 the familial cases, 35% of the tumours were ER-ve and 34% were PR-ve. Healthy female controls (n = 400) matched for ethnicity were selected from the Department of Hereditary and Congenital Disorders, Charles Nicolle Hospital, Tunis biobank. The mean age of the healthy women was 37 years at the time of blood draw. Informed consent was obtained from all participants and approved by the Ethical Committee of the institute in Tunis. Six high-risk breast cancer families, negative for any pathological BRCA1 and BRCA2 mutation, with at least three cases of multi-generational breast cancer were selected for whole exome sequencing. Selected exome variants were then genotyped in the remaining Tunisian breast cancer patients (n=153) and population controls (n=400). To evaluate population specificity, the RCC1 deletion was genotyped in additional breast cancer patients from Saudi Arabia (n=101) and Iran (n=116). These series have been described previously.[12,13] Whole-Exome Sequencing Exome sequencing was performed on six genomic DNA samples (3 µg) from Tunisian patients with familial breast cancer at 100x coverage. For this purpose, exonic sequences were enriched using the SureSelect XT Huma All Exon V6 library (Agilent Technologies, Santa Clara, CA, USA) and were sequenced on a HiSeq2500 platform (Illumina Inc., San Diego, CA, USA). Raw exome sequencing data were called, de-multiplexed and aligned according to the GATK pipeline, and variants were annotated using the SnpEff tool (http://snpeff.sourceforge.net/). Mutations were filtered according to their absence or low prevalence (MAF< 0.001) in the NCBI SNP and/or 1000Genomes databases and according to their predicted effects. We prioritized truncating mutations in genes with a known role in DNA replication or DNA repair, because these pathways are particularly important in breast cancer etiology. We identified four such genes with rare or novel truncating mutations in single exomes: MCM7, POLE, POLN, and RCC1 (Supplementary Table S1). Of these, the mutation in POLE was not confirmed by Sanger sequencing and may have constituted a false positive. The mutations in MCM7 (p.Arg527Ter) and POLN (c.467_470del4) were confirmed by Sanger sequencing, but no additional carriers were observed in a case-only screening of our Tunisian breast cancer series. A list of variants in any gene of interest is available upon request. 5 PCR Analysis and Sanger Sequencing Genomic DNA samples were amplified by PCR using the primer pair 5´- GCCACCCATTTTGCCTGTAG– 3´ and 5´-CTCACCATCCTTGGTCACAG– 3´, and PCR products were separated on a 3% agarose gel. PCR product sizes were 220 bp for the wildtype and 201 bp for the mutant allele. For semi-quantitative fragment analysis, a FAM-labelled forward primer was used. PCR products were then mixed with GeeSa™500 LIZ Size Standard and loaded onto a 3130 Genetic Analyzer (Applied Biosystems). Loss of heterozygosity (LOH) was determined as relative peak height ratio of tumour versus matched germline DNA sample. Using peak area ratios gave similar results. For direct sequencing, PCR products from mutation carriers were purified and directly sequenced using BigDye terminator v3.1 chemistry on a 3130 Genetic Analyzer (Applied Biosystems). The following GenBank reference sequence was used for variant annotation: RCC1, NM_001048194. Statistical Analyses Case-control analyses were performed with 2x2 tables using two-sided Fisher´s exact tests (because at least one cell had an expected frequency less than 5), and a p-value < 0.05 was considered significant. RCC1 was the first gene to be screened in this study so that we have not applied correction for multiple testing. We omitted the 6 hypothesis-generating families with fully sequenced exomes, including the initial mutation carrier and her daughter, from the subsequent breast cancer case-control association study. Results Exome sequencing on germ-line DNA samples from six BRCA1/BRCA2 mutation-negative Tunisian patients with familial breast cancer revealed a heterozygous carrier of a novel frameshift mutation, c.1067_1086del19, in RCC1, that is the gene for Regulator of Chromosome Condensation 1 (Suppl. Fig. S1). The frameshift deletion of 19 nucleotides was subsequently confirmed in her affected daughter, who had breast cancer at age 44 (IV.1), and in one of two daughters who were healthy at the time of analysis (IV.2; age 39); the mutation was also transmitted to the healthy young granddaughter in the affected lineage (V.1) (Fig. 1). The 6 RCC1*c.1067_1086del19 mutation has not been listed in the NCBI SNP database nor has it been reported by the Exome Aggregation Consortium. The potential association between the RCC1*c.1067_1086del19 mutation and breast cancer was then investigated in a set of 86 familial and 67 sporadic breast cancer patients from Tunisia. Unexpectedly, the mutation was detected in further five unrelated cases, including four familial cases (F75, F79, F83, F104) and one sporadic case (S27) (Fig. 2, Suppl. Fig. S2), comprising a 4.7% carrier frequency among patients with family history of breast cancer and 1.5% among sporadic cases. By contrast, no single mutation carrier was found among 400 healthy female controls from the general Tunisian population. The difference in carrier frequencies was nominally significant for all breast cancer cases and controls (p=0.0015, 2df) and for familial breast cancer vs. all controls (p= 0.0009, 2df). The ExAc database records loss of function mutations in RCC1 in only 4/48035 individuals which again is significantly different from the 5/153 patients with a RCC1 mutation in our breast cancer series (p= 3.8x10-11, 2 df). Clinical-pathological characteristics of breast cancer patients with the RCC1 mutation are summarized in Table 1. Mutation carriers had a median age at diagnosis of 46.5 years, and the mutation was associated with estrogen receptor-positive tumours in 5 of 6 cases. When we genotyped RCC1*c.1067_1086del19 in matched tumor DNA from mutation carriers, the mutation was observed in the heterozygous state in only two breast tumors, whereas loss of heterozygosity (LOH) was observed in four tumours (Table 1, Fig. 2). Interestingly, all four tumours with LOH showed loss of the mutant allele, raising the possibility of a selective disadvantage at the stage of tumour progression. Relative peak height ratios in these four tumour samples were 6%, 7%, 15% and 42%, respectively. There was no obvious correlation of LOH with other tumour variables but numbers were small. To further evaluate the population specificity of the mutation, the RCC1*c.1067_1086del19 mutation was also genotyped in breast cancer patients from Saudi Arabia (n=101) and Iran (n=116). No carrier was identified in these series, suggesting a specific occurrence of RCC1*c.1067_1086del19 in the Tunisian population. 7 Discussion This study provides first evidence for a mutation in RCC1 as a breast cancer susceptibility allele. RCC1 is a chromatin-bound guanine-nucleotide releasing factor that promotes the exchange of GDP by GTP in the Ras-related nuclear protein, Ran.[14] It has initially been identified as a regulator for the onset of chromosome condensation in the S phase.[15] The RCC1-Ran complex (together with other proteins) acts as a component of a signal transmission pathway that detects unreplicated DNA and prevents it to enter mitosis.[16] RCC1 creates the anisotropy of the distribution of RanGTP (the RanGTP gradient) underlying mitotic spindle assembly and nuclear pore and nuclear envelope formation.[17] RCC1-dependent activation of Ran accelerates cell cycle and DNA repair and inhibits DNA damage-induced cell senescence.[18] Hence, RCC1 is a well-known key protein in assuring the fidelity of DNA replication and subsequent entry of faithfully replicated chromosomes into cell division. In the present study, we identify through exome sequencing and association by genotyping a novel mutation in RCC1 that has exclusively been found in Tunisian breast cancer patients. Consistent with a genetic predisposition, the mutation was mainly detected in familial cases where it accounted for 4% of the hitherto unexplained cases. RCC1 has not previously been implicated in cancer risk, although a homolog of RCC1, RCCD1, is encoded near one of the breast cancer susceptibility loci identified through genome-wide association studies.[19-21] By contrast with low-penetrance variants at the RCCD1 locus, the RCC1 mutation described here is a truncating mutation. Due to the absence of this mutation in controls, the effect size could not be determined but it is likely to be high. Mitotic spindle assembly has as a crucial and vulnerable function for avoiding aneuploidy and is also targeted by hereditary mutations in other cancer syndromes.[22] The structure of human RCC1 has been solved to 1.7-A resolution by X-ray crystallography and consists of a seven-bladed propeller formed from internal repeats of 51-68 residues per blade.[23] The frameshift p.RLGLGEG356delfs caused by the Tunisian mutation would eliminate blade 7 and the carboxy-terminal half of blade 1. This disables RCC1 to form the molecular clasp that tightens the circular belt of the ring structure, a requirement for the binding of Ran.[23] It is presently unknown whether the p.RLGLGEG356delfs can give rise to any stable aberrant 8 protein, but the absence of RCC1 would similarly prevent its monitor function in DNA replication and mitotic entry. RCC1 mutations may become disadvantageous at later stages of tumour progression which depends on high replicative activity for fast proliferation. In fact, RCC1 blockage is being investigated as potential therapeutic means against aggressive breast tumours.[24] Selective advantage may explain why the RCC1 mutant allele is preferentially lost and why RCC1 rarely acquires somatic mutations in breast carcinomas. It is possible that the loss of the c.1067_1086del19 mutant allele is a spurious observation, given the small number of tumour samples at hand. Alternatively, it could serve to remove an intracellular barrier to unrestricted proliferation, such as a dominant negative p.RLGLGEG356delfs protein that might retain its ability of nucleosome binding [25] but blocks interaction with Ran.[23] If such a scenario is common, more breast cancer susceptibility genes and their driver mutations may have been missed in previous sequencing studies solely based on tumour material. In summary, we here describe a truncating mutation in RCC1 that was exclusively found in breast cancer patients from the Tunisian population. These data provide compelling genetic evidence for RCC1 as a novel breast cancer susceptibility gene and encourage further search for germline RCC1 mutations in cancer patients from other populations. 9 Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed by the authors. Authors' Contributions Conception and design: A. Riahi, M. Kharrat, T. Dörk Development of methodology: A. Riahi, N. Bogdanova, R. Geffers Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Riahi, H. Radmanesh, P. Schürmann, R. Geffers, R. Meddeb, T. Dörk Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Riahi, R. Geffers, T. Dörk Writing, review, and/or revision of the manuscript: A. Riahi, H. Radmanesh, P. Schürmann, N. Bogdanova, R. Geffers, R. Meddeb, M. Kharrat, T. Dörk Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Riahi, R. Geffers, T. Dörk Study supervision: A. Riahi, M. Kharrat, T. Dörk Grant Support This work was funded by a grant from the German Ministry of Education and Research and the Tunisian Ministry for Higher Education and Scientific Research (TUNGER-70). Acknowledgments We cordially thank the families who took part in this study. We gratefully acknowledge the support of clinicians at the Salah Azaiz Cancer Institute in Tunis, and Dr. Abdel-Hadi and Professor El-Harith for samples of breast cancer patients from the University Clinics of Dammam. 16 Figures Figure 1: Identification and segregation analysis of the RCC1*c.1067_1086del19 mutation. A: Pedigree of the family 89. +/-, heterozygous RCC1*c.1067_1086del19; +/+, wild-type; *, individual who underwent whole exome sequencing. B: Validation by direct sequencing of RCC1*c.1067_1086del19. (B.2, IV.1: heterozygous; B.1, IV.3: wildtype). C: Agarose gel electrophoresis for detection of heterozygous RCC1*c.1067_1086del19 carriers (IV.2, IV.1, V.1, III.1) by product size and heteroduplex formation; NC, negative control.