Single nucleotide polymorphisms in Wnt signaling ...

Carcinogenesis vol.31 no.8 pp.1381–1386, 2010 doi:10.1093/carcin/bgq082 Advance Access publication April 19, 2010

Single nucleotide polymorphisms in Wnt signaling and cell death pathway genes and susceptibility to colorectal cancer Bernd Frank, Michael Hoffmeister, Norman Klopp1, Thomas Illig1, Jenny Chang-Claude2 and Hermann Brenner Division of Clinical Epidemiology and Aging Research, German Cancer Research Center, 69115 Heidelberg, Germany, 1Helmholtz Zentrum Mu¨nchen, German Research Center for Environmental Health, Institute of Epidemiology, 85764 Neuherberg, Germany and 2Division of Cancer Epidemiology, German Cancer Research Center, 69120 Heidelberg, Germany

It is well known that 90% of colorectal cancer (CRC) cases originate from the constitutive activation of the canonical Wnt signaling pathway. There is increasing evidence that genetic variation both in Wnt and apoptotic pathway genes affects CRC susceptibility and progression. This population-based case–control study, including 1795 CRC cases and 1805 controls, investigates the association between common, putative functional polymorphisms in DNFA5, HIF1A, NDRG1, PYGO1, SFRP2, SFRP4, WISP1 and WISP3 genes and CRC risk. We found no evidence for an association between the selected allelic variants and risk of CRC. Subsite analyses, however, revealed a significant association of HIF1A c. 191T>C with rectal cancer risk [odds ratio (OR) 5 1.25, 95% confidence interval (CI), 1.03–1.51, P 5 0.03] comparing minor allele carriers with major allele homozygotes. In addition, homozygosity for the minor allele of SFRP4 P320T was significantly associated with rectal cancer risk (OR 5 1.37, 95% CI, 1.06–1.79, P 5 0.02) and early-stage CRC (OR 5 1.33, 95% CI, 1.05–1.69, P 5 0.02). This study does not support the hypothesis that Wnt signaling- and apoptosis-related polymorphisms contribute to CRC risk. However, our results provide evidence that CRC subsets may be affected. If confirmed, this knowledge may be used to assess individual susceptibility and to target potential measures of cancer prevention.

Introduction Colorectal cancer (CRC) is the third most common malignant neoplasm and the fourth leading cause of cancer deaths worldwide (1–4), accounting for 1 million new cases in 2002 (9.4% of all new cases of cancer) (1). Considerable variation in the risk of different cancers does exist, depending on geographic area (1). However, most of the international variation is due to the exposure to known or suspected risk factors related to lifestyle, environment and genetics (4–6). Constitutive activation of the canonical Wnt signaling pathway is regarded as the initiating event in 90% of CRCs (7,8), originating from mutations in APC (adenomatosis polyposis coli), AXIN1 (axis inhibition protein 1) or b-catenin/CTNNB1 (7–10). In the absence of Wnt signaling, cytoplasmic b-catenin interacts with APC and AXIN1 and serves as a substrate for the kinases CK1 (casein kinase 1, alpha 1-like) and GSK3b (glycogen synthase kinase 3 beta), resulting in b-catenin phosphorylation, ubiquitination and degradation by the proteasome (7,11). Wnt signaling is initiated through Wnt ligand binding to a Frizzled (Fz) family receptor and to co-receptors LRP5 or LRP6 (low density lipoprotein receptor-related proteins 5 and 6). After formation Abbreviations: APC, adenomatosis polyposis coli; AXIN1, axis inhibition protein 1; CRC, colorectal cancer; CI, confidence interval; DACHS, Darmkrebs: Chancen der Verhu¨tung durch Screening; Fz, Frizzled; HIF, hypoxiainducible factor; OR, odds ratio; SFRP, secreted frizzled-related protein; SNP, single nucleotide polymorphism.

Material and methods Study population CRC cases and controls were drawn from the German DACHS (Darmkrebs: Chancen der Verhu¨tung durch Screening) study, a population-based case– control study carried out in the Rhine-Neckar-Odenwald and Heilbronn regions in the southwest of Germany (32–34). The analyses comprised 1795 unrelated male and female case patients (33–94 years, median 5 69 years) with a first histologically confirmed diagnosis of primary invasive CRC (ICD10 codes C18-C20). Cases and controls were recruited between January 2003 and December 2007. Cases were mostly recruited during first hospitalization due to cancer treatment or shortly afterward at their homes. Controls consisted of 1805 subjects (34–98 years, median 5 70 years) who were randomly selected from lists of residents supplied by population registries and frequencymatched to cases by 5-year age groups, sex and county of residence. Case patients and control individuals were 99% Caucasian, being eligible if they were 30 years of age, lived in the study region, were German speaking and mentally and physically able to participate in a personal interview of 1 h. All of them gave written informed consent. The study was approved by the ethics committees of the University of Heidelberg and the State Medical Boards of Baden-Wuerttemberg and Rhineland-Palatinate, Germany. Data collection Case patients were informed about the study by their caring physicians, preferentially during their hospital stay some days after surgery. They were notified to the study center upon receipt of informed consent. Based on statistics of CRC patients treated in the hospitals, the recruited patients constituted 50% of the expected total number of eligible patients in the study region. The response rate among eligible control individuals (n 5 4770) was 50%. All study subjects were asked to participate in an inperson interview and to donate blood. A minority of controls that was not willing to give a full-length interview participated by completing a shorter questionnaire. These subjects were not asked for a blood sample and hence not included in this analysis. From 98% of the interviewed participants we

Ó The Author 2010. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

1381

Downloaded from http://carcin.oxfordjournals.org/ at Bibliothek Medizinische Hochschule Hannover on November 11, 2014

To whom correspondence should be addressed. Tel: þ49 6221 548143; Fax: þ49 6221 548142; Email: [email protected]

of the Wnt/LRP5/6 complex, Dishevelled and AXIN1 proteins are recruited to the cell membrane, the kinase activity of the APC/AXIN1/ CK1/GSK3b ‘destruction complex’ is blocked. Thus, b-catenin remains unphosphorylated, accumulates in the cytoplasm and is subsequently translocated to the nucleus. This enables b-catenin to complex with Tcell factor/lymphoid enhancer-binding factor family transcription factors and to activate Wnt downstream target genes (7,8,11). Mutations that mimic Wnt stimulation—inactivating APC or activating b-catenin/ CTNNB1 mutations—lead to nuclear accumulation of free b-catenin that associates with T-cell factor/lymphoid enhancer-binding factor, stimulating gene transcription independent of upstream Wnt signals (7,8,11). Besides aberrant downstream target activation, numerous upstream Wnt signaling components and Wnt antagonists—such as secreted frizzled-related proteins (SFRPs)—have been found to be deregulated in CRC as well (12–16). Notably, the activation of the Wnt signaling pathway has been shown to both positively and negatively modulate apoptosis (17– 19). There is increasing evidence that defects in apoptosis regulation may encourage development of CRC, and its poor response to chemotherapy and radiation (19,20). Recent studies have identified a series of candidate genes that may affect CRC development and progression due to their involvement in Wnt signaling and/or cell death pathways. In this study, we tested the hypothesis that common sequence variants in apoptosis and Wnt signaling pathway gene polymorphisms affect CRC susceptibility, focusing on eight genes, contributing to both of the pathways: DNFA5 (21), HIF1A (22,23), NDRG1 (24), PYGO1 (25,26), SFRP2 (14,15), SFRP4 (14,15), WISP1 (27,28) and WISP3 (28,29) (Table I). To our knowledge, this is the first study to investigate the involvement of these variants in cancer predisposition.

B.Frank et al.

Table I. Candidate apoptosis and Wnt signaling pathway polymorphisms Pathway

Gene name

Gene

Chromosome

SNP ID

SNP/location

Predicted functiona,b

Apoptosis Apoptosis Apoptosis Wnt signaling Wnt signaling Wnt signaling Wnt signaling Wnt signaling

Deafness, autosomal dominant 5 Hypoxia-inducible factor 1 alpha N-myc downstream regulated 1 Pygopus 1 Secreted frizzled-related protein 2 Secreted frizzled-related protein 4 WNT1 inducible signaling pathway protein 1 WNT1 inducible signaling pathway protein 3

DFNA5 HIF1A NDRG1 PYGO1 SFRP2 SFRP4 WISP1 WISP3

7p15 14q23 8q24 15q21 4q31 7p14 8q24 6q21

rs12540919 rs2057482 rs1049697 rs11858624 rs4643790 rs1802073 rs2929970 rs1230345

V207M 3’ region 3’ region P299H A45V P320T 3’ region Q56H

Benigna; damagingb n/aa; n/ab n/aa; n/ab Benigna; toleratedb Benigna; toleratedb Possibly damaginga; toleratedb n/aa; n/ab Benigna; damagingb

obtained a blood sample for DNA extraction. In the rare cases in which blood samples were not available (3%), a mouthwash was used. Information on demographic factors, anthropometric measures, medical history and lifestyle factors was collected by trained interviewers, utilizing a standardized questionnaire. More details of the data collection procedures for the DACHS study are reported elsewhere (32,33). Selection of single nucleotide polymorphisms Candidate single nucleotide polymorphisms (SNPs) were selected by means of well-defined methods and criteria: public literature resources and databases—NCBI PubMed, dbSNP and GeneCards—were extensively searched for epigenetically modulated (15,35) and/or CRC-related candidate genes (13,29,36–38) and previous epidemiologic findings (39,40), indicating associations with cancer susceptibility, being major criteria for SNP selection. In addition, SNPs were tested for evolutionary conservation at least among human, mouse and rat (WU-BLAST2) (41) and putative microRNA targets (PicTar) (42) and putative functional effects of the non-synonymous SNPs were predicted by PolyPhen and SIFT (30,31) as ancillary information and/ or affirmation of selection. SNPs with a minor allele frequency 0.05 in the HapMap CEU population (Utah residents with northern and western European ancestries from the Centre d’Etude du Polymorphisme Humain collection) were included in the study. Further basic selection criterion was an r2 value 0.8, excluding strong linkage disequilibrium between adjacent variants (43). As it was not possible to select all apoptosis- and Wnt signaling-related gene variants matching these criteria, we focused on eight most promising candidate SNPs for analysis on CRC risk: DFNA5 V207M (rs12540919), HIF1A c. 191T.C (rs2057482), NDRG1 c. 956T.C (rs1049697), PYGO1 P299H (rs11858624), SFRP2 A45V (rs4643790), SFRP4 P320T (rs1802073), WISP1 c. 1184G.A (rs2929970) and WISP3 Q56H (rs1230345). Table I summarizes the main characteristics of genes and SNPs, including chromosomal location, relevant pathways and predicted functional effects of the SNPs. DNA preparation and genotyping Genetic analysis was performed as described (33,44). Genomic DNA was isolated from blood and mouthwash samples with FlexiGene DNA and QIAamp DNA Mini Kits (QIAGEN GmbH, Hilden, Germany), respectively. Sequenom’s MassARRAYÒ system (Sequenom, San Diego, CA) was applied for genotyping, performing iPLEXÒ single base primer extension and matrixassisted laser desorption ionization time-of-flight mass spectrometry as described elsewhere (44). Genotyping calls were made in real time with the MassARRAYÒ RT software. A random selection of .5% of all samples was genotyped twice for quality control. Successfully genotyped duplicate samples displayed an average concordance rate of 99.7% (range: 99.1–100%) for the eight SNPs. Statistics Each SNP was tested for deviation from Hardy–Weinberg equilibrium in controls by comparing observed and expected genotype frequencies, using Pearson’s v2-tests with one degree of freedom. Unconditional logistic regression was applied to estimate odds ratios (ORs) and corresponding 95% confidence intervals (CIs) adjusted for age and sex, using dominant and co-dominant models, respectively. Additional adjustment for body mass index and smoking was also carried out, but as results remained essentially unchanged, they are not shown. Tests for linear trend were additionally employed. All tests were two-sided and considered statistically significant with P , 0.05. Accessory subsite analyses were conducted for colon and rectal cancers, as well as for the CRC stages I þ II and III þ IV.

1382

Table II. Polymorphisms in apoptosis and Wnt signaling genes and their associations with CRC Gene CRC variant/genotype cases N (%)

Controls

ORa (95% CI)

P

N (%)

DFNA5 V207M (rs12540919) GG 1457 (82.1) 1494 (83.4) 1 GA 308 (17.4) 284 (15.8) 1.11 (0.93–1.33) 0.23 AA 9 (0.5) 14 (0.8) 0.67 (0.29–1.56) 0.35 AA þ GA 317 (17.9) 298 (16.6) 1.09 (0.92–1.30) 0.31 Ptrend HIF1A 3’ region (rs2057482) TT 1259 (71.2) 1319 (73.5) 1 TC 477 (27.0) 441 (24.6) 1.13 (0.98–1.32) 0.10 CC 32 (1.8) 34 (1.9) 1.00 (0.61–1.63) 0.99 CC þ TC 509 (32.9) 475 (26.5) 1.12 (0.97–1.30) 0.12 Ptrend NDRG1 3’ region (rs1049697) TT 1177 (67.1) 1177 (65.9) 1 TC 521 (29.7) 548 (30.7) 0.95 (0.82–1.10) 0.49 CC 57 (3.2) 61 (3.4) 0.93 (0.64–1.35) 0.70 CC þ TC 578 (32.9) 609 (34.1) 0.95 (0.83–1.09) 0.46 Ptrend PYGO1 P299H (rs11858624) CC 1522 (86.9) 1509 (86.8) 1 CA 224 (12.8) 223 (12.8) 1.00 (0.82–1.21) 0.96 AA 5 (0.3) 7 (0.4) 0.70 (0.22–2.22) 0.54 AA þ CA 229 (13.1) 230 (13.2) 0.99 (0.81–1.20) 0.89 Ptrend SFRP2 A45V (rs4643790) CC 1175 (68.4) 1179 (67.2) 1 CT 488 (28.4) 499 (28.4) 0.99 (0.85–1.15) 0.86 TT 56 (3.3) 77 (4.4) 0.73 (0.51–1.04) 0.08 TT þ CT 544 (31.6) 576 (32.8) 0.95 (0.83–1.10) 0.50 Ptrend SFRP4 P320T (rs1802073) CC 686 (39.1) 708 (39.9) 1 CA 788 (44.9) 824 (46.4) 0.99 (0.86–1.14) 0.86 AA 281 (16.0) 243 (13.7) 1.19 (0.97–1.46) 0.09 AA þ CA 1069 (60.9) 1067 (60.1) 1.03 (0.90–1.18) 0.63 Ptrend WISP1 3’ region (rs2929970) GG 450 (25.4) 467 (26.1) 1 GA 876 (49.5) 884 (49.4) 1.03 (0.88–1.21) 0.73 AA 443 (25.0) 438 (24.5) 1.05 (0.87–1.26) 0.61 AA þ GA 1319 (74.6) 1322 (73.9) 1.04 (0.89–1.20) 0.65 Ptrend WISP3 Q56H (rs1230345) GG 958 (54.2) 917 (51.3) 1 GT 665 (37.6) 723 (40.5) 0.88 (0.77–1.01) 0.07 TT 145 (8.2) 147 (8.2) 0.95 (0.74–1.22) 0.68 TT þ GT 810 (45.8) 870 (48.7) 0.89 (0.78–1.02) 0.09 Ptrend a

ORs were adjusted for age and sex.

5 0.45

5 0.17

5 0.46

5 0.82

5 0.24

5 0.19

5 0.61

5 0.19


n/a, not applicable; SNP, single nucleotide polymorphism; SNP ID, single nucleotide polymorphism identification. a Fuctional predictions were based on the in silico tool PolyPhen, referring to the variant allele (30,31). b Fuctional predictions were based on the in silico tool SIFT, referring to the variant allele (30,31).

Apoptosis and Wnt pathway SNPs in CRC

Table III. Polymorphisms in apoptosis and Wnt signaling genes and their associations with colon and rectal cancers Gene variant/genotype

Controls

Colon cancer

N (%)

N (%)

ORa (95% CI)

P

Rectal cancer

ORa (95% CI)

P

N (%)

905 (83.3) 177 (16.3) 5 (0.5) 182 (16.7)

1 1.03 (0.84–1.27) 0.60 (0.22–1.67) 1.01 (0.83–1.24)

HIF1A 3’ region (rs2057482) TT 1319 (73.5) TC 441 (24.6) CC 34 (1.9) CC þ TC 475 (26.5)

785 (72.4) 279 (25.7) 20 (1.9) 299 (27.6)

1 1.05 (0.88–1.25) 1.01 (0.57–1.76) 1.05 (0.89–1.24)

NDRG1 3’ region (rs1049697) TT 1177 (65.9) TC 548 (30.7) CC 61 (3.4) CC þ TC 609 (34.1)

730 (68.0) 311 (29.0) 33 (3.1) 344 (32.0)

1 0.91 (0.77–1.08) 0.87 (0.56–1.34) 0.91 (0.77–1.07)

PYGO1 P299H (rs11858624) CC 1509 (86.8) CA 223 (12.8) AA 7 (0.4) AA þ CA 230 (13.2)

938 (87.6) 131 (12.2) 2 (0.2) 133 (12.4)

1 0.94 (0.74–1.18) 0.44 (0.09–2.13) 0.92 (0.73–1.16)

SFRP2 A45V (rs4643790) CC 1179 (67.2) CT 499 (28.4) TT 77 (4.4) TT þ CT 576 (32.8)

719 (68.1) 304 (28.8) 33 (3.1) 337 (31.9)

1 1.00 (0.84–1.18) 0.72 (0.47–1.09) 0.96 (0.82–1.13)

SFRP4 P320T (rs1802073) CC 708 (39.9) CA 824 (46.4) AA 243 (13.7) AA þ GA 1067 (60.1)

434 (40.4) 479 (44.6) 161 (15.0) 640 (59.6)

1 0.94 (0.80–1.11) 1.08 (0.86–1.37) 0.98 (0.84–1.14)

WISP1 3’ region (rs2929970) GG 467 (26.1) GA 884 (49.4) AA 438 (24.5) AA þ GA 1322 (73.9)

264 (24.4) 551 (50.9) 268 (24.7) 819 (75.6)

1 1.10 (0.92–1.33) 1.08 (0.87–1.34) 1.09 (0.92–1.30)

WISP3 Q56H (rs1230345) GG 917 (51.3) GT 723 (40.5) TT 147 (8.2) TT þ GT 870 (48.7)

586 (54.1) 407 (37.5) 91 (8.4) 498 (45.9)

1 0.88 (0.75–1.04) 0.97 (0.73–1.28) 0.90 (0.77–1.04)

0.76 0.33 0.90 Ptrend 5 0.93

0.57 0.99 0.58 Ptrend 5 0.62

0.27 0.51 0.23 Ptrend 5 0.23

0.58 0.31 0.49 Ptrend 5 0.40

0.99 0.12 0.63 Ptrend 5 0.34

0.49 0.50 0.75 Ptrend 5 0.78

0.30 0.49 0.31 Ptrend 5 0.49

0.12 0.82 0.16 Ptrend 5 0.30

552 (80.3) 131 (19.1) 4 (0.6) 135 (19.7)

1 1.24 (0.98–1.56) 0.77 (0.25–2.37) 1.21 (0.97–1.52)

474 (69.3) 198 (28.9) 12 (1.8) 210 (30.7)

1 1.27 (1.04–1.55) 0.97 (0.50–1.90) 1.25 (1.03–1.51)

447 (65.6) 210 (30.8) 24 (3.5) 234 (34.4)

1 1.01 (0.83–1.23) 1.06 (0.65–1.73) 1.02 (0.84–1.23)

584 (85.9) 93 (13.7) 3 (0.4) 96 (14.1)

1 1.11 (0.86–1.45) 1.16 (0.30–4.54) 1.11 (0.86–1.44)

456 (68.8) 184 (27.8) 23 (3.5) 207 (31.2)

1 0.98 (0.80–1.20) 0.76 (0.47–1.22) 0.95 (0.78–1.15)

252 (37.0) 309 (45.4) 120 (17.6) 429 (63.0)

1 1.04 (0.86–1.27) 1.37 (1.06–1.79) 1.12 (0.93–1.34)

186 (27.1) 325 (47.4) 175 (25.5) 500 (72.9)

1 0.94 (0.76–1.16) 1.06 (0.80–1.31) 0.97 (0.79–1.18)

372 (54.4) 258 (37.7) 54 (7.9) 312 (45.6)

1 0.88 (0.73–1.07) 0.90 (0.64–1.26) 0.89 (0.74–1.06)

0.07 0.65 0.09 Ptrend 5 0.14

0.02 0.94 0.03 Ptrend 5 0.06

0.91 0.80 0.87 Ptrend 5 0.82

0.43 0.83 0.42 Ptrend 5 0.42

0.83 0.25 0.59 Ptrend 5 0.39

0.66 0.02 0.23 Ptrend 5 0.04

0.55 0.86 0.73 Ptrend 5 0.87

0.19 0.54 0.18 Ptrend 5 0.24

a

ORs were adjusted for age and sex.

The aforementioned analyses were carried out using Statistical Analysis Software version 9.1 (SAS Institute, Cary, NC). Haploview was used to examine measures of linkage disequilibrium (D# and r2) between adjacent SNPs and to define haplotype block structures based on the definition by Gabriel et al. (43). Power calculations were employed with the power and sample size software PS version 3.0.7 (45). With the present sample size, we had a power of 80% at a significance level of 0.05 to detect ORs 1.28 (rs12540919), 1.24 (rs2057482), 1.22 (rs1049697), 1.31 (rs11858624), 1.23 (rs4643790), 1.22 (rs1802073), 1.25 (rs2929970) and 1.21 (rs1230345) (45). For some subsite analyses, however, the statistical power was limited. Given the hypothesis-driven candidate SNP approach of the study, we purposely did not perform multiple testing corrections, in order to not eliminate potentially important results from the investigation. There-

fore, it will be of prime importance to replicate the findings in independent data sets.

Results The majority of DACHS cases and controls were males and between 60 and 79 years of age (median ages: 69 and 70 years, respectively). A family history of CRC was slightly more common among cases than among controls (14 versus 11%). Two-thirds of cancers were located in the sigmoid colon and rectum, and slightly more than half were diagnosed stage I or II (54 versus 46%), subdivided into 56% of colon and 52% of rectal cancers. Main characteristics of the DACHS study population have been described previously (34).

1383


DFNA5 V207M (rs12540919) GG 1494 (83.4) GA 284 (15.8) AA 14 (0.8) AA þ GA 298 (16.6)

B.Frank et al.

Table IV. Polymorphisms in apoptosis and Wnt signaling genes and their associations with CRC UICC stages Gene variant/genotype

Controls

CRC stages I þ II

N (%)

N (%)

ORa (95%CI)

P

CRC stages III þ IV

ORa (95% CI)

P

N (%)

794 (82.3) 166 (17.2) 5 (0.5) 171 (17.7)

1 1.10 (0.89–1.36) 0.68 (0.24–1.88) 1.08 (0.88–1.33)

HIF1A 3’ region (rs2057482) TT 1319 (73.5) TC 441 (24.6) CC 34 (1.9) CC þ TC 475 (26.5)

686 (71.2) 263 (27.3) 14 (1.5) 277 (28.8)

1 1.15 (0.96–1.37) 0.79 (0.42–1.49) 1.12 (0.94–1.34)

NDRG1 3’ region (rs1049697) TT 1177 (65.9) TC 548 (30.7) CC 61 (3.4) CC þ TC 609 (34.1)

625 (65.4) 299 (31.3) 32 (3.3) 331 (34.6)

1 1.03 (0.87–1.22) 0.99 (0.64–1.53) 1.02 (0.87–1.21)

PYGO1 P299H (rs11858624) CC 1509 (86.8) CA 223 (12.8) AA 7 (0.4) AA þ CA 230 (13.2)

835 (87.1) 121 (12.6) 3 (0.3) 124 (12.9)

1 0.98 (0.77–1.24) 0.77 (0.20–2.98) 0.97 (0.77–1.23)

SFRP2 A45V (rs4643790) CC 1179 (67.2) CT 499 (28.4) TT 77 (4.4) TT þ CT 576 (32.8)

640 (68.1) 269 (28.6) 31 (3.3) 300 (31.9)

1 0.99 (0.83–1.19) 0.74 (0.48–1.14) 0.96 (0.81–1.14)

SFRP4 P320T (rs1802073) CC 708 (39.9) CA 824 (46.4) AA 243 (13.7) AA þ GA 1067 (60.1)

352 (36.8) 444 (46.4) 161 (16.8) 605 (63.2)

1 1.08 (0.91–1.29) 1.33 (1.05–1.69) 1.14 (0.97–1.34)

WISP1 3’ region (rs2929970) GG 467 (26.1) GA 884 (49.4) AA 438 (24.5) AA þ GA 1322 (73.9)

234 (24.3) 491 (51.1) 236 (24.6) 727 (75.7)

1 1.11 (0.92–1.34) 1.08 (0.86–1.34) 1.10 (0.92–1.32)

WISP3 Q56H (rs1230345) GG 917 (51.3) GT 723 (40.5) TT 147 (8.2) TT þ GT 870 (48.7)

521 (54.1) 360 (37.4) 82 (8.5) 442 (45.9)

1 0.88 (0.74–1.04) 0.98 (0.73–1.31) 0.89 (0.76–1.05)

0.37 0.45 0.46 Ptrend 5 0.60 0.13 0.47 0.20 Ptrend 5 0.34 0.76 0.95 0.79 Ptrend 5 0.84 0.86 0.70 0.82 Ptrend 5 0.78 0.94 0.17 0.64 Ptrend 5 0.38 0.36 0.02 0.11 Ptrend 5 0.02 0.29 0.53 0.32 Ptrend 5 0.52 0.12 0.90 0.16 Ptrend 5 0.33

659 (81.9) 142 (17.6) 4 (0.5) 146 (18.1)

1 1.14 (0.91–1.42) 0.66 (0.22–2.01) 1.12 (0.90–1.39)

569 (71.0) 214 (26.7) 18 (2.2) 232 (29.0)

1 1.13 (0.93–1.37) 1.26 (0.70–2.24) 1.14 (0.94–1.37)

551 (69.3) 219 (27.5) 25 (3.1) 244 (30.7)

1 0.86 (0.71–1.03) 0.87 (0.54–1.41) 0.86 (0.72–1.03)

684 (86.8) 102 (12.9) 2 (0.3) 104 (13.2)

1 1.01 (0.78–1.30) 0.63 (0.13–3.06) 1.00 (0.78–1.28)

531 (68.5) 219 (28.3) 25 (3.2) 244 (31.5)

1 0.98 (0.81–1.19) 0.72 (0.45–1.15) 0.95 (0.79–1.14)

334 (42.0) 341 (42.9) 120 (15.1) 461 (58.0)

1 0.88 (0.73–1.05) 1.03 (0.80–1.33) 0.91 (0.77–1.08)

215 (26.7) 382 (47.5) 207 (25.7) 589 (73.3)

1 0.94 (0.77–1.15) 1.03 (0.82–1.30) 0.97 (0.80–1.17)

435 (54.3) 303 (37.8) 63 (7.9) 366 (45.7)

1 0.88 (0.74–1.06) 0.91 (0.67–1.25) 0.89 (0.75–1.05)

0.25 0.46 0.33 Ptrend 5 0.45 0.22 0.44 0.18 Ptrend 5 0.16 0.10 0.57 0.09 Ptrend 5 0.12 0.95 0.57 0.98 Ptrend 5 0.90 0.87 0.17 0.57 Ptrend 5 0.34 0.15 0.80 0.28 Ptrend 5 0.74 0.53 0.82 0.73 Ptrend 5 0.83 0.17 0.57 0.17 Ptrend 5 0.24

UICC, International Union Against Cancer. a ORs were adjusted for age and sex.

The average call rate for the SNPs analyzed was 98.2% (range: 96.5–99.1%), and it did not differ between cases and controls for any individual assay. Using Pearson’s goodness of fit v2-test with one degree of freedom, allele frequencies among controls were consistent with Hardy–Weinberg equilibrium for all SNPs (P 0.01). Table II summarizes the results of the analyses for the eight SNPs. Assuming a dominant model (comparing carriers of the minor allele with wild type homozygous subjects), there was no evidence of a significant association with overall CRC risk for any of the polymorphisms (Table II). When considering associations among subgroups, including colon cancer (Table III) and late-stage CRCs (Table IV), none of the analyzed SNPs was associated with disease risk. However, we revealed a significant association of HIF1A c. 191T.C with rectal cancer risk (OR 5 1.25, 95% CI, 1.03– 1.51, P 5 0.03; Table III) and of homozygosity for the minor allele

1384

of SFRP4 P320T with rectal cancer risk (OR 5 1.37, 95% CI, 1.06– 1.79, P 5 0.02; Table III) and with early-stage CRC (OR 5 1.33, 95% CI, 1.05–1.69, P 5 0.02; Table IV). Discussion In the present case–control study, we assessed the association of a series of apoptosis- and Wnt signaling pathway-related SNPs with CRC risk. We observed associations of HIF1A c. 191T.C and SFRP4 P320T with an increased risk of rectal cancer, and of SFRP4 P320T with early-stage CRC. Hypoxia-inducible factors (HIFs) regulate oxygen delivery and adaptation to oxygen deprivation, by regulating the expression of genes that are involved in cellular processes, such as glucose uptake and metabolism, angiogenesis, cell proliferation and apoptosis


DFNA5 V207M (rs12540919) GG 1494 (83.4) GA 284 (15.8) AA 14 (0.8) AA þ GA 298 (16.6)

Apoptosis and Wnt pathway SNPs in CRC

Funding German Research Council (Deutsche Forschungsgemeinschaft) (BR 1704/6-1, BR 1704/6-3, BR 1704/6-4 to H.B., CH 117/1-1 to J.C.-C.); German Federal Ministry for Education and Research (Bundesminis-

terium fu¨r Bildung und Forschung) (01 KH 0404 to H.B., 01 GS 08181 to H.B. and J.C.-C.). Acknowledgements We wish to thank the clinicians who supported the study, all patients and control individuals who participated in the study, and Ute Handte-Daub and Belinda-Su Kaspereit for excellent technical assistance. Conflict of Interest Statement: None declared.

References 1. Parkin,D.M. et al. (2005) Global cancer statistics, 2002. CA Cancer J. Clin., 55, 74–108. 2. Ferlay,J. et al. (2004) GLOBOCAN 2002. Cancer Incidence, Mortality and Prevalence Worldwide. IARC Cancer Base No. 5 Version 2.0. IARCPress, Lyon. 3. Pisani,P. et al. (2002) Estimates of the worldwide prevalence of cancer for 25 cancers in the adult population. Int. J. Cancer, 97, 72–81. 4. Potter,J.D. (1999) Colorectal cancer: molecules and populations. J. Natl Cancer Inst., 91, 916–932. 5. Lichtenstein,P. et al. (2000) Environmental and heritable factors in the causation of cancer—analyses of cohorts of twins from Sweden, Denmark, and Finland. N. Engl. J. Med., 343, 78–85. 6. de la Chapelle,A. (2004) Genetic predisposition to colorectal cancer. Nat. Rev. Cancer, 4, 769–780. 7. Klaus,A. et al. (2008) Wnt signalling and its impact on development and cancer. Nat. Rev. Cancer, 8, 387–398. 8. Fodde,R. et al. (2001) APC, signal transduction and genetic instability in colorectal cancer. Nat. Rev. Cancer, 1, 55–67. 9. Bienz,M. et al. (2000) Linking colorectal cancer to Wnt signaling. Cell, 103, 311–320. 10. Kinzler,K.W. et al. (1996) Lessons from hereditary colorectal cancer. Cell, 87, 159–170. 11. Huang,H. et al. (2008) Wnt/beta-catenin signaling: new (and old) players and new insights. Curr. Opin. Cell Biol., 20, 119–125. 12. Katoh,M. (2001) Frequent up-regulation of WNT2 in primary gastric cancer and colorectal cancer. Int. J. Oncol., 19, 1003–1007. 13. Huang,D. et al. (2010) SFRP4 was overexpressed in colorectal carcinoma. J. Cancer Res. Clin. Oncol., 136, 395–401. 14. Bovolenta,P. et al. (2008) Beyond Wnt inhibition: new functions of secreted Frizzled-related proteins in development and disease. J. Cell Sci., 121, 737– 746. 15. Suzuki,H. et al. (2004) Epigenetic inactivation of SFRP genes allows constitutive WNT signaling in colorectal cancer. Nat. Genet., 36, 417–422. 16. Suzuki,H. et al. (2003) A genomic screen for genes upregulated by demethylation and histone deacetylase inhibition in human colorectal cancer. Nat. Genet., 31, 141–149. 17. Shi,Y. et al. (2007) Inhibition of Wnt-2 and galectin-3 synergistically destabilizes beta-catenin and induces apoptosis in human colorectal cancer cells. Int. J. Cancer, 121, 1175–1181. 18. Olmeda,D. et al. (2003) Beta-catenin regulation during the cell cycle: implications in G2/M and apoptosis. Mol. Biol. Cell, 14, 2844–2860. 19. West,N.J. et al. (2009) Apoptosis in the colonic crypt, colorectal adenomata, and manipulation by chemoprevention. Cancer Epidemiol. Biomarkers Prev., 18, 1680–1687. 20. Watson,A.J. (1995) Review article: manipulation of cell death–the development of novel strategies for the treatment of gastrointestinal disease. Aliment. Pharmacol. Ther., 9, 215–226. 21. Masuda,Y. et al. (2006) The potential role of DFNA5, a hearing impairment gene, in p53-mediated cellular response to DNA damage. J. Hum. Genet., 51, 652–664. 22. Wincewicz,A. et al. (2007) Cumulative expression of HIF-1-alpha, Bax, Bcl-xL and P53 in human colorectal cancer. Pathology, 39, 334–338. 23. Greijer,A.E. et al. (2004) The role of hypoxia inducible factor 1 (HIF-1) in hypoxia induced apoptosis. J. Clin. Pathol., 57, 1009–1014. 24. Stein,S. et al. (2004) NDRG1 is necessary for p53-dependent apoptosis. J. Biol. Chem., 279, 48930–48940. 25. Andrews,P.G. et al. (2009) The transcriptional activity of Pygopus is enhanced by its interaction with cAMP-response-element-binding protein (CREB)-binding protein. Biochem. J., 422, 493–501. 26. Kramps,T. et al. (2002) Wnt/wingless signaling requires BCL9/leglessmediated recruitment of pygopus to the nuclear beta-catenin-TCF complex. Cell, 109, 47–60.

1385


(23,46). HIF1 is a basic helix-loop-helix (bHLH) PER–ARNT (arylcarbon receptor nuclear translocator)–SIM (PAS) transcription factor that binds DNA as heterodimer composed of the hypoxia-regulating HIF1A and the constitutively expressed non-hypoxia-related HIF1B subunits (36,46). HIF1A is considered to induce apoptosis by increasing the stability of the tumor suppressor TP53 that mediates cell death via regulation of proapoptotic proteins, such as BCL2-associated X protein or the release of cytochrome c from mitochrondria (22,23,47). HIF1A is overexpressed in multiple types of cancer, including lung, prostate, breast, colon and rectum carcinoma, and in regional or distant metastases, emphasizing its vital role in tumor progression (36,48). We found a significant association of the HIF1A c. 191T.C SNP in the 3# untranslated region of the gene with an increased risk of rectal cancer (Table III). A plausible mechanism, explaining the association of HIF1A c. 191T.C with rectal cancer risk, may be that this variant induces the overexpression of HIF1A under hypoxic conditions in rectal cancer cells, stimulating genes which are involved in apoptosis and cell proliferation in cancer cells. This coincides with the findings by Rasheed et al. (48), providing evidence for increased HIF1A expression in rectal cancer cases. In addition, HIF1A expression was shown to be a prognostic marker in rectal adenocarcinoma (49). Yet, it is questionable, if HIF1A c. 191T.C represents the causative variant itself. Interestingly, the associated SNP is in strong linkage disequilibrium with HIF1A P582S and A588T, which have been demonstrated to significantly enhance transcription activities under both normoxic and hypoxic conditions (50) and which are associated with ulcerative growth in CRC patients (51). On the other hand, HIF1A c. 191T.C is localized in immediate vicinity of a miR-199a and miR-199b-binding site in the 3# untranslated region, and it may influence messenger RNA stability, translational efficiency or gene expression. SFRPs are a family of secreted glycoproteins that directly bind Wnts, thereby altering their ability to attach the ‘destruction complex’ and inhibiting Wnt signaling (14,16). SFRPs interact with Wnt ligands through their characteristic cysteine-rich domain, sharing homology with the Fz cysteine-rich domain. They block canonical Wnt signaling by interacting with Wnt proteins to prevent them from binding to Fz proteins or by forming non-functional complexes with Fz (14). Functional loss of SFRPs contributes to the activation of Wnt signaling, leading to carcinogenesis through deregulation of cell proliferation and differentiation. SFRP1, SFRP2 and SFRP5 are frequent targets of epigenetic inactivation early in CRC progression (15,16), whereas SFRP4 was found to be overexpressed in CRC (13,14). This study found a significant association of SFRP4 P320T with risk of rectal cancer and early-stage CRC. This finding, along with the putative functional effect of P320T (Table I, PolyPhen), supports suggestions that this SNP may provoke elevated Wnt signaling, resulting in the attenuation of apoptotic processes, subsequent uncontrolled gene expression and tumorigenesis. Our study has both strengths and limitations. The strengths include its population-based design, the well-defined homogeneous study population, and the fairly large sample size. In addition, we included a number of well-founded SNPs for which an association with CRC is biologically plausible and/or has been previously reported for any forms of cancer. This study does not support the hypothesis that alterations in both cell death and Wnt signaling pathways contribute to CRC susceptibility or progression. However, our results suggest that a subset of CRCs, including rectal cancer and early-stage CRC, may be affected. Overall, our findings encourage the concept that genetic variants in these genes confer cancer risk, which may help to assess individual susceptibility and to target potential measures of disease prevention. Replication in further epidemiologic studies and functional analyses are warranted to confirm these findings.

B.Frank et al.

1386

39. Akino,K. et al. (2007) Identification of DFNA5 as a target of epigenetic inactivation in gastric cancer. Cancer Sci., 98, 88–95. 40. Lee,J.Y. et al. (2008) Rare variant of hypoxia-inducible factor-1alpha (HIF1A) and breast cancer risk in Korean women. Clin. Chim. Acta, 389, 167– 170. 41. Lopez,R. et al. (2003) WU-Blast2 server at the European Bioinformatics Institute. Nucleic Acids Res., 31, 3795–3798. 42. Krek,A. et al. (2005) Combinatorial microRNA target predictions. Nat. Genet., 37, 495–500. 43. Gabriel,S.B. et al. (2002) The structure of haplotype blocks in the human genome. Science, 296, 2225–2229. 44. Kormann,M.S. et al. (2005) G-Protein-coupled receptor polymorphisms are associated with asthma in a large German population. Am. J. Respir. Crit. Care Med., 171, 1358–1362. 45. Dupont,W.D. et al. (1998) Power and sample size calculations for studies involving linear regression. Control. Clin. Trials, 19, 589–601. 46. Wang,G.L. et al. (1995) Purification and characterization of hypoxiainducible factor 1. J. Biol. Chem., 270, 1230–1237. 47. Hammond,E.M. et al. (2005) The role of p53 in hypoxia-induced apoptosis. Biochem. Biophys. Res. Commun., 331, 718–725. 48. Rasheed,S. et al. (2009) Hypoxia-inducible factor-1alpha and -2alpha are expressed in most rectal cancers but only hypoxia-inducible factor-1alpha is associated with prognosis. Br. J. Cancer, 100, 1666–1673. 49. Lu,X.G. et al. (2006) Clinical significance of immunohistochemical expression of hypoxia-inducible factor-1alpha as a prognostic marker in rectal adenocarcinoma. Clin. Colorectal Cancer, 5, 350–353. 50. Tanimoto,K. et al. (2003) Hypoxia-inducible factor-1alpha polymorphisms associated with enhanced transactivation capacity, implying clinical significance. Carcinogenesis, 24, 1779–1783. 51. Franse´n,K. et al. (2006) Association between ulcerative growth and hypoxia inducible factor-1alpha polymorphisms in colorectal cancer patients. Mol. Carcinog., 45, 833–840. Received February 17, 2010; revised April 8, 2010; accepted April 14, 2010


27. Xu,L. et al. (2000) WISP-1 is a Wnt-1- and beta-catenin-responsive oncogene. Genes Dev., 14, 585–595. 28. Pennica,D. et al. (1998) WISP genes are members of the connective tissue growth factor family that are up-regulated in wnt-1-transformed cells and aberrantly expressed in human colon tumors. Proc. Natl Acad. Sci. USA, 95, 14717–14722. 29. Thorstensen,L. et al. (2001) WNT1 inducible signaling pathway protein 3, WISP-3, a novel target gene in colorectal carcinomas with microsatellite instability. Gastroenterology, 121, 1275–1280. 30. Ramensky,V. et al. (2002) Human non-synonymous SNPs: server and survey. Nucleic Acids Res., 30, 3894–3900. 31. Kumar,P. et al. (2009) Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm. Nat. Protoc., 4, 1073–1081. 32. Brenner,H. et al. (2006) Does a negative screening colonoscopy ever need to be repeated? Gut, 55, 1145–1150. 33. Lilla,C. et al. (2006) Effect of NAT1 and NAT2 genetic polymorphisms on colorectal cancer risk associated with exposure to tobacco smoke and meat consumption. Cancer Epidemiol. Biomarkers Prev., 15, 99–107. 34. Frank,B. et al. (2010) Colorectal Cancer and Polymorphisms in DNA Repair Genes WRN, RMI1, and BLM. Carcinogenesis, 31, 442– 445. 35. Kim,M.S. et al. (2008) Aberrant promoter methylation and tumor suppressive activity of the DFNA5 gene in colorectal carcinoma. Oncogene, 27, 3624–3634. 36. Zhong,H. et al. (1999) Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. Cancer Res., 59, 5830– 5835. 37. Strzelczyk,B. et al. (2009) Identification of high-risk stage II colorectal tumors by combined analysis of the NDRG1 gene expression and the depth of tumor invasion. Ann. Surg. Oncol., 16, 1287–1294. 38. Khor,T.O. et al. (2006) A comparative study of the expression of Wnt-1, WISP-1, survivin and cyclin-D1 in colorectal carcinoma. Int. J. Colorectal Dis., 21, 291–300.