WO2009035402A1

WO2009035402A1 - Diagnosis of cancer by determining polymorphisms in blm, top3a or rmil

Info

Publication number: WO2009035402A1
Application number: PCT/SE2008/051009
Authority: WO
Inventors: Håkan OLSSON; Mattias HÖGLUND; Karin Broberg
Original assignee: Forskarpatent I Syd Ab
Priority date: 2007-09-11
Filing date: 2008-09-10
Publication date: 2009-03-19

Abstract

The present invention relates to a method for monitoring the risk of obtaining a cancer disease, in particular colon cancer, lymphoma, prostate cancer, malignant skin non- melanoma, malignant melanoma, leukaemia, bladder cancer, breast cancer, in particular malignant melanoma, leukaemia, bladder cancer, and breast cancer, whereby the frequencies of genetic variants of Bloom syndrome protein, and topoisomerase IIIa (TOP3A) and RECQ-mediated genome instability 1 protein (RMI1) are determined, and when a significant increased presence of certain alleles and/or combinations of alleles ispresence, the risk is significant.

Description

Diagnosis of cancer by determining polymorphisms in BLM, TOP3A or RMIl.

DESCRIPTION Technical field

The present invention relates to a method for monitoring the risk of being caught by a cancer disease, in particular colon cancer, lymphoma, prostate cancer, non-melanoma skin cancer, malignant melanoma, leukaemia, bladder cancer, breast cancer.

Background of the invention

Bloom syndrome is a condition characterised by growth inhibition, light sensitivity and high incidence of cancer in early life. Although there appears to be a predominance of lymphocytic leukaemia and lymphoma, many cancer types are seen in this condition. Represented in Bloom's syndrome are cancers that commonly affect the general population, including colon cancer, non-melanoma skin cancer and other carcinomas (German et al., 1997). A defining feature of Bloom's syndrome is an elevated frequency of sister chromatide exchanges. These arise from crossing over of chromatide arms during homologous recombination, a ubiquitous process that exists to repair DNA double-stranded breaks and damaged replication forks. Whereas crossing over is required in meiosis, it can in mitotic cells be associated with a detrimental loss of heterozygosity, a common feature in neoplastic cells.

(German J. Bloom's syndrome. XX. The first 100 cancers. Cancer Genet Cytogenet. 1997 Jan;93(1 ):100-6.)

BLM (Bloom syndrome protein), the helicase mutated in Bloom syndrome, is found in protein complexes together with topoisomerase IMa (TOP3A) and a newly identified member, the RECQ-mediated genome instability 1 (RMH ) protein, that process double Holliday junction intermediates into non-crossover recombinants. This dissolution activity of the BLM-TOP3A- RMM complex is thought to be critical for the suppression of DNA crossover formation in mitotic cells and cancer avoidance in humans. The complex might process many other DNA structures as well, such as stalled replication forks and has been implicated in checkpoint signalling and checkpoint responses during DNA damage.

Since mutations that alter BLM function are associated with elevated cancer susceptibility, genetic variants of BLM and other proteins that form a complex with BLM might affect cancer risk for different cancer forms as well. In this study the polymorphisms in the RMH, TOP3A and BLM, and their association with cancer risk in leukaemias (acute myeloid leukaemia and myelodysplastic syndromes), and malignant melanoma (MM), bladder cancer, and breast cancer was analysed.

Summary of the present invention During this study it was shown possible to obtain a diagnostic method for monitoring the risk of obtaining certain cancer forms in a subject. In particular it was shown possible to foresee a risk in obtaining colon cancer, lymphoma, prostate cancer, non-melanoma skin cancer, malignant melanoma, leukaemia, bladder cancer, breast cancer, in particular malignant melanoma, leukaemia, bladder cancer, and breast cancer.

Detailed description of the present invention

In particular the invention relates to a method for monitoring the risk of obtaining a cancer disease, in particular colon cancer, lymphoma, prostate cancer, non-melanoma skin cancer, malignant melanoma, leukaemia, bladder cancer, breast cancer, in particular malignant melanoma, leukaemia, bladder cancer, and breast cancer, whereby the frequencies of genetic variants of Bloom syndrome protein (BLM), and topoisomerase IMa (TOP3A) and RECQ-mediated genome instability 1 protein (RMM ) are determined, and when a significant increased presence of certain alleles and/or combinations of alleles is presence the risk is significant.

In a preferred embodiment the presence of rs401549 of BLM is monitored, rs401549 having the sequence:

GATGCTGCAAAGTGCTTCACTCTTAC[AZG]AACAAAAGTTGTGAAAATTAAATAA (SEQ. ID. NO. 1 )

In a further preferred embodiment the presence of rs2532105 of BLM is monitored, rs2532105 having the sequence:

CTCTGTTGTGGCTTCTTCTTGCTACC[CZT]CTGCAAAGGCTGGGAGGTCGATTCT (SEQ. ID. NO. 2)

In an another preferred embodiment the presence of rs1563634 of TOP3A is monitored, rs1563634 having the sequence:

CGGCCAGCACAAGAAACCTGCTGCGA[AZG]ATCACCTGCACTCTCCGGAACCCAC (SEQ. ID. NO. 3)

In still another preferred embodiment the presence of rs12945597 of TOP3A is monitored, rs 12945597 having the sequence: CCTTCCAGTCAACGTTGACGGGGTAG[AZG]ATGAGCTGCATGCTGCTAACCTGAG (SEQ. ID. NO. 4)

In a further preferred embodiment the presence of rs296887 of RMH is monitored, rs296887 having the sequence:

CGGCAGCGGAGGGACGGCAGTCTCGC[AyG]CGGTGAGGAGCCGGGTTGGGGGAGC (SEQ. ID. NO. 5)

In another preferred embodiment of the invention the presence of rs2532105 and rs1563634 and rs12945597 are monitored.

In another preferred embodiment of the invention the presence of rs2532105 and rs1563634 are monitored.

In another preferred embodiment of the invention the presence of rs2532105 and rs12945597 are monitored.

In another preferred embodiment of the invention the presence of one or more of rs2532105, rs12945597, rs401549 and rs296887 are monitored to monitor the risk of malignant melanoma.

In another preferred embodiment of the invention the presence of one or more of rs2532105, rs1563634, rs12945597, and rs401549 and are monitored to monitor the risk of breast cancer.

In another preferred embodiment of the invention the presence of one or more of rs2532105, rs12945597, and rs296887 are monitored to monitor the risk of leukaemia.

In another preferred embodiment of the invention the presence of one or more of rs2532105, rs12945597, and rs401549 are monitored to monitor the risk of bladder cancer.

The invention will be described in the following with reference to a study made to determine the dependency of certain alleles and the risk of cancer, as well as a method for determining the presence of such alleles and other examples. Materials and methods Study populations

The study has been approved by the Ethics Committee of Lund University. All cancer patients were voluntarily recruited from the Southern Health Care Region of Sweden. Study population characteristics for different cancer forms are shown in Table 1.

Table 1. Study characteristics of cases and controls for different cancer forms.

Leukaemia MM Bladder cancer Breast cancer

Cases Cases Controls Cases Controls Cases Controls

N 152 170 119 61 T56 193 189

Median age 72 66 68 69 69 59 63

(range) (22-95) (24-80) (22-96) (35-85) (21-90) (34-86) (23-85)

Women/Men 73/79 87/83 53/66 9/52 34/122 193/0 131/58

(%) (48/52) (51/49) (44.5/55.

Leukaemia and malignant melanoma

These study populations have been described in more detail by Broberg et al.. In short, the diagnoses were: 78 cases of acute myeloid leukaemia and, 56 cases of myelodysplastic syndromes, and 18 subjects of acute myeloid leukaemia preceeded by a myelodysplastic syndrome. All the participating melanoma patients (N=170) had a primary cutaneous melanoma diagnosed except for two patients, who presented with metastatic disease without known primary site of their melanoma. The leukaemia patients' samples were collected during the years 1998-2004 and the MM samples during 2001-2005. The control group was drawn randomly from the Regional Population Registry during 2001-2004, frequency matched with the cases series with respect to sex, year of birth, as well as country of living.

Bladder cancer

The study groups have been studied earlier and are described more in detail in Broberg et al. The bladder cancer patients' and the control individuals' samples (mouth washes) were collected during 1995-2000. The control group was drawn randomly from the Regional Population Registry during 1995-2000, frequency matched with the cases with respect to sex, year of birth, as well as country of living. Breast cancer

The breast cancer patients all came for treatment at the Oncology Clinic at Lund University Hospital between 1990 and 1998, where they accepted to participate in a study to be made and donated blood samples. The case group encompassed 193 patients, all of which were women. The control group consisted of 189 accompanying spouses to cancer patients, recruited from Southern Sweden. The control individuals filled in a questionnaire about their health status and donated a blood sample. Spouses without former cancer were selected for the study.

Polymorphism selection and initial screening

Polymorphisms were selected in Bloom syndrome related genes (RMH, TOP3A and BLM) from the HapMap data (www.hapmap.org). TagSNPs (tagSNPs, i.e., tagged single nucleotide polymorphisms) were identified from the CEU population data (CEPH, Utah residents with ancestry from northern and western Europe), and by running the data in Haploview (Barrett et al. 2005). The TagSNPs chosen showed a minor allelic frequency >10%. DNA extractions for all samples were made with QIAmp 96 DNA blood kit (Qiagen, Hilden, Germany) by SWEGENE resource centre for Profiling Polygenic Diseases in Malmo University hospital, Sweden, apart from the blood samples from the breast control group, which were extracted for DNA with Wizard Genomic DNA Purification Kit (Promega, Madison, Wl, USA). The polymorphisms were analyzed with matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry (Sequenom™, San Diego, CA) at the SWEGENE resource centre for Profiling Polygenic Diseases in Malmo University hospital. Out of 28 tagSNPs analysed, 2 failed in the analysis, which left us with 26 tagSNPs. The frequency of individuals with missing genotype data ranged between 1.3%- 6.8%/SNP assay for leukaemia and malignant melanoma, and 7.7%-12.7%/SNP assay for bladder cancer. Selection of tagSNPs for further analysis of the breast cancer study population was based on changes in effect estimates (computed as odds ratios, ORs) in similar directions in different cancer forms, as well as on similar allele frequencies in the cases and control groups.

Genotyping of polymorphisms in the breast cancer/control material The genotyping was performed on 193 cases of breast cancer and 189 controls. The selected polymorphisms were rs1563634 (O ^O assay number for Taqman® SNP Genotyping Assays from Applied Biosystems, Foster City, CA) and rs12945597 (O ^ n\-o;.$0 i < >) in TOP3A, rs401549 (O J:^>^ i ) and rs2532105 (O >o

10 corresponds to rs7725932 which is tightly linked to rs1563634. Taqman assays and allelic discrimination was run on an ABI PRISM 7000 (Applied Biosystems). The Taqman assay reaction volume was 25 μl containing: 1X Universal Taqman mix (Applied Biosystems), each primer at 0.45 μM mixed with each probe at 0.10 μM, and 5-12 ng of DNA template. The thermal cycle protocol was 95⁰C for 10 min, 95⁰C for 15 sec and 6O⁰C for 1 min (40 cycles). Plate reading for allelic discrimination was performed under 6O⁰C for 1 min. For all assays, at least 5% of the samples were reanalyzed and the concordance rate of these analyses was 100%.

Statistical analysis The Hardy-Weinberg equilibrium (HWE) test was undertaken using the chi-square test in the control groups. Effect estimates were computed as odds ratios (ORs) (defined as the ratio of the odds of an event occurring in one group to the odds of it occurring in another group) with 95% confidence intervals (95% CIs) (an interval estimate of a population parameter, CIs are used to indicate the reliability of an estimate) by logistic regression using SPSS 14.0 (version 14; SPSS, Chicago, IL, USA). The analyses were performed both without any adjustments, as well as with adjustments for sex and age. When stratifying for age, subjects were categorized into two groups, based on the median age among the cancer cases and control group combined as a cut-off (<65 and ≥65 years). The stratified analysis was adjusted for gender.

To assess false positives, the False Positive Report Probability (FPRP) (Wacholder et al. 2004) was calculated, based on observed association data, according to the formula; α ^*(1-ττ)/( α ^*(1-ττ) + (1-β)^*ττ) wherein α denotes the p-values from the logistic regression analyses, 1 -β denotes the statistical power for the tests and π denotes the prior probability of a true association of the tested genetic variant and outcome. The likely OR for a noteworthy finding was set to 1.5, which is a plausible value for important biologic effects (Engel et al. 2002, Marcus et al 2000). The prior probability employed was set to 0.01 (low probability) and 0.1 (high probability) for all SNPs. The FPRP value for noteworthiness was set to 0.5, thus statistically significant SNPs with a FPRP above 0.5 were in this study not considered reliable to classify as true positives. These values, as well as the values for ORs and 95% CIs were entered into the online spreadsheet included in Wacholder et al. (2004) for FPRP calculations.

Bioinformatics

SNPs with a statistically significant association were further evaluated in silico for potential function. PupaSNP (www.pupasnp.org) was employed in order to detect SNPs potentially affecting transcription factor binding sites, exonic splicing enhancers/silencers, triplexes, splice sites and microRNA target sites. Emboss CpGPIot (http://www.ebi.ac.uk/emboss/cpgplot/) was employed to detect CpG-rich areas.

Results

All SNPs were in Hardy Weinberg equilibrium (HWE: the Hardy-Weinberg principle states that the genotype frequencies in a population remain constant or are in equilibrium from generation to generation unless specific disturbing influences are introduced. Genetic equilibrium is a basic principle of population genetics) in the control groups for leukaemia, malignant melanoma, and bladder cancer. The four selected SNPs were in HWE in the breast cancer control group as well (data not shown). The results from the initial screening of all SNPs and their association with leukaemia, malignant melanoma, and bladder cancer, are shown in Tables 2-4. Few statistically significant findings were found for each cancer form. The SNP rs296887 in RMH was significantly associated with increased cancer risks for leukaemia, as well as malignant melanoma, but not with bladder cancer. In TOP3A, rs12945597 was associated with increased risk in leukaemia and malignant melanoma, and a non-significantly increased risk was observed in bladder cancer as well. In BLM, rs401549 was associated with significantly increased risk in bladder cancer and a non-significantly increased risk in malignant melanoma. Rs2532105 showed increased risk in malignant melanoma and bladder cancer, and a non-significant risk increment in leukaemia. In leukaemia, rs393974 and rs6496724 were also significantly associated with cancer risk.

Table 2. Influence of polymorphisms in RMH, TOP3A, and BLM on risk for leukaemia. Statistically significant associations (p ≤ 0.05) are denoted in bold ^a

Gene SNPs Genotype Cases Controls OR 95% Cl

(Allelic % variant allele cases/controls)

RMH rs296887 " GG 68 70 1.0 " -

(32/25) AG 63 38 1.7 1.0-2.9

AA 15 10 1.5 0.65-3.7 rs296891 CC 40 32 1.0 -

(47/47) CT 78 62 1.0 0.57-1.8

TT 31 24 1.0 0.51-2.1 r_s754144 GG 68 55 1.0

(33/31 ) AG 64 52 1.0 0.60-1.7

AA 17 11 1.3 0.54-2.9 rs3737134 AA 87 59 1.0 -

(26/34) AC 51 50 0.7 0.42-1.2

CC 6 8 0.5 0.17-1.5 rs6559754 CC 42 35 1.0 -

(45/45) CG 79 59 1.1 0.64-2.0

GG 28 24 0.97 0.48-2.0 rs12001293 TT 108 85 1.0

(15/16) AT 37 31 0.94 0.54-1.6

AA 3 3 0.79 0.16-4.0

TOP3A rs1563634^d GG 84 52 1.0 -

(26/34) AG 54 52 0.64 0.38-1.1

AA 11 14 0.49 0.21-1.2 rs2294913 GG 74 67 1.0 -

(27/26) AG 61 41 1.3 0.80-2.3

AA 9 10 0.82 0.31-2.1 rs2294914 GG 121 88 1.0

(10/14) AG 24 28 0.62 0.34-1.1

AA 2 2 0.73 0.10-5.3 rs4925159 GG 46 45 1.0

(46/39) AG 64 55 1.1 0.66-2.0

AA 35 18 1.9 0.94-3.8 rs12945597 GG 61 64 1.0 -

(37/25) AG 63 49 1.3 0.81-2.3

AA 22 5 4.6 1.6-13

BLM rs387833 CC 88 73 1.0

(22/22) CT 54 39 1.1 0.69-1.9

TT 5 7 0.59 0.18-1.9 rs389480 CC 54 39 1.0

(39/42) CT 73 60 0.88 0.52-1.5

TT 21 20 0.76 0.36-1.6 rs393974 TT 38 45 1.0 -

(43/39) GT 82 44 2.2 1.3-3.9

GG 19 20 1.13 0.53-2.4 rs401549 AA 65 57 1.0

(31/30) AG 75 51 1.29 0.78-2.1

GG 9 10 0.79 0.30-2.1 rs2270132 TT 58 45 1.0

(36/38) GT 76 57 1.0 0.62-1.7

GG 15 16 0.73 0.33-1.6 rs2518967 AA 84 72 1.0 -

(25/22) AG 57 39 1.3 0.75-2.1

GG 8 7 0.98 0.34-2.8 rs2518968 CC 43 42 1.0

(45/41 ) CG 79 55 1.40 0.81-2.4

GG 27 21 1.26 0.62-2.6 rs2532105 CC 103 90 1.0

(16/12) CT 43 25 1.50 0.85-2.7

TT 3 2 1.31 0.21-8.0 rs6496724 AA 86 58 1.0 _ (23/30) AC 55 47 0.79 0.47-1.3

CC 6 12 0.34 0.12-0.95 rs7165790 AA 59 40 1.0

(38/43) AG 65 55 0.80 0.47-1.4

GG 24 23 0.71 0.35-1.4 rs7184015 GG 73 61 1.0 -

(30/31) GT 64 40 1.34 0.79-2.3

TT 12 17 0.59 0.26-1.3 rs8031341 AA 97 67 1.0

(19/24) AG 48 45 0.74 0.44-1.2

GG 4 6 0.46 0.13-1.7 rs8037430 CC 65 46 1.0

(35/.39) CT 65 53 0.87 0.51-1.5

TT 19 19 0.71 0.34-1.5 rs16944863 AA 117 92 1.0 -

(11/11) AG 30 26 0.91 0.50-1.6

GG 1 0 - - rs16944894 AA 95 69 1.0

(20/22) AG 49 45 0.79 0.48-1.3

GG 5 4 0.91 0.24-3.5 ^a Logistic regression, unadjusted. b Accession number for polymorphisms in the SNP database of National Centre of

Biotechnology Information, http://www.ncbi. nlm.nih.gov/sites/entrez?db=snp c Reference category. d SNPs labelled with yellow were chosen for further analysis in the breast cancer/control material.

Table 3. Influence of polymorphisms in RMH, TOP3A, and BLM on risk for malignant melanoma. Statistically significant associations (p ≤ 0.05) are denoted in bold ^a

Gene SNPs Genotype Cases Controls OR 95% Cl

(Allelic % variant allele cases/controls)

RMH rs296887^u GG 78 70 1.0 " -

(32/25) AG 71 38 1.7 1.0-2.8

AA 19 10 1.7 0.74-3.9 rs296891 CC 54 32 1.0 -

(44/47) CT 78 62 0.75 0.43-1.3

TT 35 24 0.86 0.44-1.7 rs754144 GG 95 55 1.0

(26/31 ) AG 58 52 0.65 0.39-1.1

AA 15 11 0.79 0.34-1.8 rs3737134 AA 101 59 1.0 -

(22/28) AC 55 50 0.64 0.39-1.1

CC 8 8 0.58 0.21-1.6 rs6559754 CC 57 35 1.0 -

(43/45) CG 79 59 0.82 0.48-1.4

GG 32 24 0.82 0.42-1.6 rs12001293 TT 110 85 1.0

(18/16) AT 55 31 1.4 0.81-2.3

AA 2 3 0.52 0.084-3.2

TOP3A rs1563634 ^d GG 96 52 1.0 -

(25/34) AG 60 52 0.63 0.38-1.0

AA 12 14 0.46 0.20-1.1 rs2294913 GG 80 67 1.0 -

(32/26) AG 67 41 1.4 0.83-2.3

AA 20 10 1.7 0.73-3.8 rs2294914 GG 137 88 1.0

(11/14) AG 24 28 0.55 0.30-1.0

AA 6 2 1.9 0.38-9.8 rs4925159 GG 59 45 1.0

(42/39) AG 76 55 1.1 0.63-1.8

AA 33 18 1.4 0.70-2.8 rs12945597 GG 78 64 1.0 -

(32/25) AG 72 49 1.2 0.74-2.0

AA 18 5 3.0 1.0-8.4 rs387833 CC 88 73 1.0

BLM (26/22) CT 71 39 1.5 0.92-2.5

TT 7 7 0.83 0.28-2.5 rs389480 CC 52 39 1.0

(44/42) CT 82 60 1.0 0.60-1.7

TT 32 20 1.2 0.60-2.4 rs393974 TT 71 45 1.0 -

(37/39) GT 64 44 0.92 0.54-1.6

GG 28 20 0.89 0.45-1.8 rs401549 AA 62 57 1.0

(36/30) AG 90 51 1.6 0.99-2.7

GG 16 10 1.5 0.62-3.5 rs2270132 TT 56 45 1.0

(42/38) GT 84 57 1.2 0.71-2.0

GG 28 16 1.4 0.68-3.0 rs2518967 AA 87 72 1.0 -

(26/22) AG 73 39 1.5 0.94-2.6

GG 8 7 0.95 0.33-2.7 rs2518968 CC 46 42 1.0

(47/41 ) CG 87 55 1.4 0.84-2.5

GG 35 21 1.5 0.77-3.0 rs2532105 CC 109 90 1.0

(18/12) CT 56 25 1.9 1.1-3.2

TT 3 2 1.2 0.20-7.6 rs6496724 AA 94 58 1.0 _ (25/30) AC 61 47 0.80 0.49-1.3

CC 11 12 0.57 0.23-1.4 rs7165790 AA 58 40 1.0

(38/43) AG 90 55 1.1 0.67-1.9

GG 19 23 0.57 0.28-1.2 rs7184015 GG 85 61 1.0 -

(28/31) GT 72 40 1.3 0.78-2.1

TT 11 17 0.46 0.20-1.1 rs8031341 AA 112 67 1.0

(19/24) AG 48 45 0.64 0.38-1.1

GG 7 6 0.70 0.23-2.2 rs8037430 CC 65 46 1.0

(35/39) CT 89 53 1.2 0.72-2.0

TT 14 19 0.52 0.24-1.1 rs16944863 AA 126 92 1.0 -

(13/11) AG 38 26 1.1 0.61-1.9

GG 3 0 - rs16944894 AA 109 69 1.0

(20/22) AG 50 45 0.70 0.43-1.2

GG 9 4 1.4 0.42-4.8 ^a Logistic regression, unadjusted. b Accession number for polymorphisms in the SNP database of National Centre of Biotechnology Information, http://www.ncbi. nlm.nih.gov/sites/entrez?db=snp ^c Reference category. d SNPs labelled with yellow were chosen for further analysis in the breast cancer/control mat

Table 4. Influence of polymorphisms in RMH, TOP3A, and BLM on risk for malignant melanoma. Statistically significant associations (p < 0.05) are denoted in bold ^a

Gene SNPs Genotype Cases Controls OR 95% Cl

(Allelic % variant allele cases/controls)

RMH rs296887^u GG 32 63 1.0 " -

(26/30) AG 23 68 0.67 0.35-1.3

AA 4 8 0.98 0.28-3.5 rs296891 CC 15 47 1.0 -

(51/43) CT 29 65 1.4 0.68-2.9

TT 16 28 1.8 0.77-4.2 r_s754144 GG 23 67 1.0

(35/30) AG 32 61 1.5 0.81-2.9

AA 5 12 1.2 0.39-3.8 rs3737134 AA 35 76 1.0 -

(23/26) AC 22 47 1.0 0.53-1.9

CC 3 12 0.54 0.14-2.0 rs6559754 CC 15 50 1.0 -

(50/41 ) CG 30 65 1.5 0.75-3.2

GG 15 25 2.0 0.85-4.7 rs12001293 TT 45 105 1.0

(16/14) AT 13 31 0.98 0.47-2.0

AA 3 4 1.8 0.38-8.1

TOP3A rs1563634 ^d GG 25 66 1.0 -

(32/33) AG 32 54 1.6 0.83-3.0

AA 3 19 0.42 0.11-1.5 rs2294913 GG 35 72 1.0 -

(25/29) AG 21 54 0.80 0.42-1.5

AA 5 13 0.79 0.26-2.4 rs2294914 GG 54 115 1.0

(6/9) AG 7 24 0.62 0.25-1.5

AA 0 0 rs4925159 GG 22 55 1.0

(42/38) AG 24 62 0.97 0.49-1.9

AA 13 22 1.5 0.63-3.4 rs12945597 GG 26 73 1.0 -

(36/30) AG 25 49 1.4 0.74-2.8

AA 9 17 1.5 0.59-3.7

BLM rs387833 CC 35 86 1.0

(23/20) CT 19 50 0.93 0.48-1.8

TT 4 2 4.9 0.86-28 rs389480 CC 18 41 1.0

(43/44) CT 33 73 1.0 0.52-2.1

TT 10 24 0.95 0.38-2.4 rs393974 TT 20 53 1.0 -

(44/39) GT 23 57 1.1 0.53-2.2

GG 13 23 1.5 0.64-3.5 rs401549 AA 26 64 1.0

(37/30) AG 24 67 0.88 0.46-1.7

GG 10 8 3.1 1.1-8.7 rs2270132 TT 23 48 1.0

(40/41 ) GT 26 69 0.79 0.40-1.5

GG 11 23 1.0 0.42-2.4 rs2518967 AA 32 84 1.0 -

(26/22) AG 25 50 1.3 0.70-2.5

GG 3 6 1.3 0.31-5.6 rs2518968 CC 18 40 1.0

(46/43) CG 29 79 0.82 0.41-1.6

GG 13 21 1.4 0.57-3.3 rs2532105 CC 40 111 1.0

(20/10) CT 16 27 1.6 0.80-3.4

TT 4 1 11 1.2-102 rs6496724 AA 37 68 1.0 _ (23/30) AC 19 58 0.60 0.31-1.2

CC 4 13 0.57 0.17-1.9 rs7165790 AA 26 48 1.0

(34/40) AG 27 73 0.68 0.36-1.3

GG 7 19 0.68 0.25-1.8 rs7184015 GG 33 76 1.0 -

(26/25) GT 21 56 0.86 0.45-1.6

TT 5 7 1.6 0.49-5.6 rs8031341 AA 41 85 1.0

(18/22) AG 16 48 0.69 0.35-1.4

GG 3 6 1.0 0.25-4.4 rs8037430 CC 30 53 1.0

(28/36) CT 26 73 0.63 0.33-1.2

TT 4 14 0.50 0.15-1.7 rs16944863 AA 47 106 1.0 -

(11/13) AG 13 31 0.95 0.45-2.0

GG 0 2 - - rs16944894 AA 40 89 1.0

(18/20) AG 18 45 0.89 0.46-1.7

GG 2 6 0.74 0.14-3.8 a Logistic regression, unadjusted. bAccession number for polymorphisms in the SNP database of National Centre of Biotechnology Information, http://www.ncbi. nlm.nih.gov/sites/entrez?db=snp c Reference category. d SNPs labelled with yellow were chosen for further analysis in the breast cancer/control material.

For further analysis of breast cancer, rs12945597 in TOP3A, rs401549, and rs2532105 in BLM were selected, as well as rs1563634 in TOP3A, which showed similar non-significant changes in all three cancer forms for the variant homozygous carriers. The rs12945597 in TOP3A (Table 5A) and rs2532105 in BLM (Table 5B) was significantly associated with increased breast cancer risk. Dual-polymorphisms analyses (in two different genes) were performed for all cancer forms for the three SNPs showing main genetic effects (Table 6). An allele-dosage effect was found for the combination rs12945597 and rs2532105 for leukaemia, bladder and breast cancer, where carriers with two variant alleles displayed the highest risks (Table 6A).

Table 5. Associations of polymorphisms in A) TOP3A and B) BLM and cancer risk in different cancer forms. ^a

5A TOP3A rs1563634 " Allelic % AA AG GG risk allele cases/controls

Leukaemia 1.0 ^c 1.3 2.0

75/66 0.53-3.1 ⁶ 0.82-4.7

Malignant 1.0 1.4 2.2 melanoma 75/66 0.58-3.2 0.93-5.1

Bladder 1.0 3.9 2.5 cancer 68/67 1.1-14 0.67-9.1

Breast 1.0 1.5 1.2 cancer 68/68 0.69-3.1 0.55-2.4

TOP3A, rs12945597 Allelic % GG AG AA risk allele cases/controls

1.0 ' 1.3 4.9

37/25 0.76-2.2 1.7-14 1.0 1.2 3.0 32/25 0.74-2.0 1.0-8.4 1.0 1.4 1.5 36/30 0.75-2.8 0.58-3.7 1.0 1.6 1.1 31/26 1.0-2.7 0.53-2.3

5B BLM, rs401549

Allelic % AA AG GG risk allele cases/controls

Leukaemia 1.0 1.3 0.77

31/30 - 0.77-2.1 0.29-2.0

Malignant 1.0 1.6 1.5 melanoma 36/30 - 0.98-2.7 0.61-3.5

Bladder 1.0 0.90 3.1 cancer 37/30 - 0.47-1.7 1.1-8.7

Breast 1.0 1.0 0.99 cancer 34/33 - 0.66-1.7 0.44-2.2

BLM, rs2532105

Allelic % CC CT+TT risk allele cases/controls

1.0 1.4

16/12 - 0.83-2.5

1.0 1.8

18/12 - 1.1-3.2

1.0 2.1

20/10 - 1.0-4.2

1.0 2.0

17/13 - 1.2-3.3

^a Logistic regression, adjusted for age and gender. Statistically significant associations (p ≤ 0.05) are denoted in bold. ^b Accession number for polymorphisms in the SNP database of National Centre of

Biotechnology Information, http://www.ncbi. nlm.nih.gov/sites/entrez?db=snp c Reference category. d For rs1563634, the reference category was changed to the variant allele in order to display cancer risk increment. e 95% confidence interval

Table 6. Cancer risk estimates for gene-gene combinations of A) rs12945597 (TOP3A) and rs2532105 (BLM) and B) rs1563634 (TOP3A) and rs2532105 (BLM) in different cancer forms.³

6A. rs12945597+rs2532105

Genotype OR 95% Cl

Leukaemia GG+CC 1.0 -

GG+T- 1.4 0.61-3.0

A-+CC 1.6 0.89-2.8

A-+T- 2.4 1.1-5.4

Malignant melanoma GG+CC 1.0 -

A-+CC 1.7 0.98-3.0

GG+T- 2.5 1.2-5.2

A-+T- 2.4 1.1-5.4

Bladder cancer GG+CC 1.0 -

A-+CC 1.4 0.68-2.9

GG+T- 2.1 0.74-6.1

A-+T- 3.0 1.1-7.7

Breast cancer GG+CC 1.0 -

A-+CC 1.5 0.88-2.5

GG+T- 2.0 0.97-3.9

A-+T- 3.0 1.4-6.5

6B. rs1563634^b+rs2532105

Genotype OR 95% Cl

Leukaemia AA+CC 1.0 -

AA+T- 0.48 0.087-2.7

G-+CC 1.1 0.39-3.1

G-+T- 1.9 0.62-5.8

Malignant melanoma AA+CC 1.0 -

AA+T- 0.48 0.089-2.6

G-+CC 1.1 0.41-3.1

G-+T- 2.5 0.84-7.4

Bladder cancer AA+CC 1.0

AA+T- 1.7 0.12-23

G-+CC 3.0 0.64-14

G-+T- 6.5 1.3-33

Breast cancer AA+CC 1.0 -

G-+CC 2.7 1.0-7.0

AA+T- 12 2.1-70

G-+T- 4.5 1.6-13 a Statistically significant associations (p ≤ 0.05) are denoted in bold. The analyses were adjusted for age and gender. b For rs1563634, the reference category was changed to the variant allele in order to display cancer risk increment. Based on these results, we combined all cases (N=576) and controls (N=464), respectively and performed association analyses. The cases and controls showed a similar distribution in age (cases median age=65 years, controls= 64 years) but differed for sex (63% women among cases and 47% women among controls). For rs1563634, rs12945597, and rs2532105, a significantly increased risk for developing cancer was found, and for the two latter SNPs an allele dosage effect was detected (Table 7). The highest risk was found for the combination of rs1563634 of TOP3A and rs2532105 in BLM (OR=3.2, 95% Cl 1.8-5.7; Table 8). We also performed a three-polymorphism combination analysis for rs1563634, rs12945597, and rs2532105 and found an OR=3.5, 95% Cl 1.9-6.5, for carriers of three variant alleles.

Table 7. Association of polymorphisms in TOP3A and BLM and risk of cancer (all cancer cases and all controls combined).⁸ SNP Hpnntypp DR Q^E;% r.i FPRP n 1^C FPRP n ni rs1563634" AA 1.0 - -

AG 1.6 1.0-2.5 0.47 0.90

GG 1.8 1.2-2.8 0.28 0.81 rs12945597 GG 1.0

AG 1.4 1.1-1.9 0.29 0.82

AA 1.7 1.1-2.6 0.31 0.835 rs401549 AA 1.0 - - -

AG 1.1 0.85-1.4

GG 1.2 0.74-1.9 - - rs2532105 CC 1.0 - - -

CT+TT 1.7 1.3-2.3 0.024 0.21

^a Statistically significant associations (p < 0.05) are denoted in bold. The analyses were adjusted for age and gender. b FPRP is only calculated for significant p-values. c For rs1563634, the reference category was changed to the variant allele in order to display cancer risk increment.

Table 8. Gene-gene associations of polymorphisms in TOP3A and BLM and risk of cancer (all cancer cases and all controls combined).⁸

Polymorphisms Genotype OR 95% Cl rs12945597+rs2532105 GG+CC 1.0 -

A-+CC 1.6 1.2-2.1

GG+T- 1.9 1.3-2.9

A-+T- 2.6 1.7-3.9 rs1563634+rs2532105 AA+CC 1.0

G-+CC 1.8 1.1-3.1

AA+T- 2.0 0.84-4.6

G-+T- 3.2 1.8-5.7 a Statistically significant associations (p ≤ 0.05) are denoted in bold. The analyses were adjusted for age and gender.

The impact of age on the associations between genetic markers of the BLM-complex and cancer risk was evaluated. For the two polymorphisms in TOP3A a stronger effect was found among elderly individuals (rs1563634 GG carriers ≥65 years OR=2.1 95% Cl 1.2-3.9, <65 years OR=1.4 95% Cl 0.74-2.7; rs12945597 AA carriers ≥65 years OR=2.4 95% Cl 1.3-4.5, <65 years OR=1.1 95%CI 0.61-2.1 ). No such effect could be detected for the two variants analysed in BLM (data not shown).

We evaluated the significance of the findings for all cancer cases and controls combined by using FPRP analysis (Table 7). At high prior probability (set as 0.1 ) all statistically significant findings demonstrated a low probability for false positive results. However, at a lower prior probability (0.01 ), only BLM rs2532105 displayed a FPRP value 0.5.

The bioinformatics analyses revealed that rs1563634 is situated in a CpG island downstream to the TOP3A gene, about 700 basepairs downstream to the start codon.

The present study shows that individuals carrying genetic variants of the BLM-TOP3A-RMI1 complex have an increased risk of leukaemia, malignant melanoma, bladder and breast cancer. The strongest genetic risk marker was found in BLM, i.e., the variant allele of rs2532105, which showed a statistically significantly increased risk for three out of four cancer forms analyzed. Taking two polymorphisms into consideration resulted in stronger associations, but there were no indications of a multiplicative interaction between genetic variants at different gene loci.

The number of study individuals for each cancer type was limited but the results for the four tumour types showed a consistent pattern. This suggests that some polymorphisms in the BLM complex are general cancer susceptibility markers and that the homologous recombination system may be involved in neoplastic transformation of several cell types. Our finding is in line with the observation that individuals with Blooms syndrome carrying mutations in BLM, essential for the homologous recombination complex, show elevated risk for various cancer types. There are increasing amount of evidence that homologous recombination generates loss of heterozygosity in various cancer types e.g. acute myeloid leukaemias, follicular lymphomas, breast cancer, bladder cancer, gastrointestinal stromal tumors and Barrett's oesophagus. A recent study have also shown the importance of the BLM-TOP3A-RMI1 for the maintenance of the genome stability by faithful chromosome segregation and prevention of anaphase bridges, a cytogenetic aberration found in several types of neoplastic cells. In our previous article about the Ser455Asn polymorphism in RMH, we reported an more pronounced effect of among elderly individuals. We found the same pattern in this study for the two polymorphisms in TOP3A, possibly reflecting the finding that mitotic recombination increases with age and, hypothetically, that protein variants involved in this process has a larger influence when the body burden of mitotic recombination increases.

The effect of the polymorphisms remains to be clarified. None of the three rs1563634, rs12945597, and rs2532105 result in non-synonymous exchanges. However, rs1563634 is positioned in a CpG island downstream to the TOP3A gene. The SNP does not in itself introduce or remove an extra CpG, but CpG regions tend to be associated with gene promoter regions and unmethylated CpGs allow gene expression.

For bladder cancer, the drop out in genotype data/assay was fairly high and these data should thus be cautiously interpreted. The cancer forms in this study are very different in etiology, and genetic heterogeneity is present in different tumour forms, but were combined based on similar effect estimates of the polymorphisms on cancer risk. Also the characteristics of the control groups differed; they were selected in different ways during different time frames, showed varying degree of participation, and varied in sex and age distribution. However, allelic frequencies in the different control groups were similar, indicating that the observed frequencies are good estimates of the true frequency in this region of Sweden. Moreover, when the cases from all cancer forms and the different control groups were combined in one analysis, the case and control groups were similar in age but differed somewhat in sex distribution.

False-positive report probability (FPRP) analyses have been conducted in order to assess the risk of false positives. The prior probabilities were considered relatively high (0.1 ) due to the impact of mutations in BLM for cancer risk in Bloom syndrome patients and the functional relevance of the TOP3A for the BLM complex. However, since there was little evidence for functional effect for the SNPs analysed, based on bioinformatics analysis, the prior probability of 0.01 was also taken into consideration. With this cut-off, only the BLM rs2532105 remained as a significant finding. Thus, the results of this study needs to be taken under careful consideration.

There are very few other studies of polymorphisms of the BLM-complex and cancer risks known. As mentioned, we have previously studied the Ser455Asn polymorphism in RMH and found an association between increased risk of leukaemia and MM for variant carriers. In this study, we analysed the SNP rs296887, which is linked to RMH Ser455Asn and found a similar but weaker effect for leukaemia and MM but no effect in bladder cancer. Other studies have analyzed BLM Thr298Met and small cell lung carcinoma risk and BLM Pro868Leu and familial breast cancer, but none of these polymorphisms showed any association with cancer risk. However, there are several studies on other genes that are required for homologous recombination repair, i.e. RAD51, XRCC2 and XRCC3. The RAD51 -135 (G/C) variant allele has been associated with increased risk of AML and this relation was most pronounced for therapy-related AML. The results for the XRCC3 Thr241 Met variant have so far been not been conclusive: variant carriers was found to have increased risk of AML, bladder and breast cancer and malignant melanoma, whereas in a previous study of ours no association with this marker and bladder cancer was demonstrated. In one meta-analysis, there were no significant associations between XRCC3 Thr241 Met and cancer of the bladder, breast, lung or skin, but in another meta-analysis the Met/Met genotype showed a small cancer risk with the strongest effect found for breast cancer. Also for the XRCC2 gene the results have been contradictory; the variant allele of Arg188His has been associated with increased cancer risk for breast and pharyngeal cancer and smoking-related pancreatic cancer, whereas for epithelial ovarian cancer there was no effect or a reduced risk.

In conclusion, our data indicate that variants of the BLM-TOP3A-RMI1 recombination complex, has an impact on cancer risk. Since the SNPs are common among our subjects with leukaemia, MM, bladder and breast cancer, even the moderate increase in risk observed in this study is associated with a considerable impact.

Claims

1. Method for monitoring the risk of obtaining a cancer disease, in particular colon cancer, lymphoma, prostate cancer, malignant skin non-melanoma, malignant melanoma, leukaemia, bladder cancer, breast cancer, in particular malignant melanoma, leukaemia, bladder cancer, and breast cancer, whereby the frequencies of genetic variants of Bloom syndrome protein, and topoisomerase MIa (TOP3A) and RECQ-mediated genome instability 1 protein (RMH ) are determined, and when a significant increased presence of certain alleles and/or combinations of alleles is presence the risk is significant.

2. Method according to claim 1 , wherein the presence of rs401549 of BLM is monitored, rs401549 having the sequence:

GATGCTGCAAAGTGCTTCACTCTTAC[AZG]AACAAAAGTTGTGAAAATTAAATAA (SEQ. ID. NO. 1 )

3. Method according to claim 1 , wherein the presence of rs2532105 of BLM is monitored, rs2532105 having the sequence:

CTCTGTTGTGGCTTCTTCTTGCTACC[CZTJCTGCAAAGGCTGGGAGGTCGATTCT (SEQ. ID. NO. 2)

4. Method according to claim 1 , wherein the presence of rs1563634 of TOP3A is monitored, rs1563634 having the sequence:

CGGCCAGCACAAGAAACCTGCTGCGA[AZG]ATCACCTGCACTCTCCGGAACCCA C (SEQ. ID. NO. 3)

5. Method according to claim 1 , wherein the presence of rs12945597 of TOP3A is monitored, rs 12945597 having the sequence:

CCTTCCAGTCAACGTTGACGGGGTAG[AZG]ATGAGCTGCATGCTGCTAACCTGA G (SEQ. ID. NO. 4).

6. Method according to claim 1 , wherein the presence of rs296887 of RMM is monitored, rs296887 having the sequence:

CGGCAGCGGAGGGACGGCAGTCTCGC[AZG]CGGTGAGGAGCCGGGTTGGGGG AGC (SEQ. ID. NO. 5)

7. Method according to claim 1 , wherein the presence of rs2532105 and rs1563634 and rs12945597 are monitored.

8. Method according to claim 1 , wherein the presence of rs2532105 and rs1563634 are monitored.

9. Method according to claim 1 , wherein the presence of rs2532105 and rs12945597 are monitored.

10. Method according to claim 1 , wherein the presence of one or more of rs2532105, rs12945597, rs401549 and rs296887 are monitored to monitor the risk of malignant melanoma.

1 1. Method according to claim 1 , wherein the presence of one or more of rs2532105, rs1563634, rs12945597, and rs401549 and are monitored to monitor the risk of breast cancer.

12. Method according to claim 1 , wherein the presence of one or more of rs2532105, rs12945597, and rs296887 are monitored to monitor the risk of leukaemia.

13. Method according to claim 1 , wherein the presence of one or more of rs2532105, rs12945597, and rs401549 are monitored to monitor the risk of bladder cancer.