Purpose: The PALB2 gene has an essential role in BRCA2-mediated DNA double-strand break repair and intra–S phase DNA damage checkpoint control, and its mutations are moderately associated with breast cancer susceptibility. This study was designed to investigate the common variants of PALB2 and their association with breast cancer risk.

Experimental Design: Four single nucleotide polymorphisms (SNP; rs249954, rs249935, rs120963, and rs16940342) which tagged all 19 of the reported SNPs (minor allele frequency >0.05) covering PALB2 were selected and genotyped in 1,049 patients with breast cancer and 1,073 cancer-free controls in a female Chinese population.

Results: Based on the multiple hypothesis testing with the Benjamini-Hochberg method, tagging SNPs (tSNP) rs249954, rs120963, and rs16940342 were found to be associated with an increase of breast cancer risk (false discovery rate–adjusted P values of 0.004, 0.028, and 0.049, respectively) under the dominant model. tSNP rs249954 was associated with a 36% increase of breast cancer risk [adjusted odds ratio (OR), 1.36; 95% confidence intervals (CI), 1.13-1.64; P = 0.001; TT/TC versus CC genotypes]. The adjusted OR for rs120963 was 1.25 (95% CI, 1.04-1.49; P = 0.014; CC/CT versus TT genotypes). For rs16940342, the adjusted OR was 1.21 (95% CI, 1.02-1.45; P = 0.037; GG/GA versus AA genotypes). Based on an additive model, tSNPs rs249954 and rs120963 were associated with an increase of breast cancer risk (P = 0.005 and 0.019; respectively), with the false discovery rate–adjusted P values being 0.020 and 0.038, respectively.

Conclusions: Our data suggest that the variants of PALB2 confer low-penetrance breast cancer susceptibility in a Chinese population.

Translational Relevance

Although many genetic and environmental factors have been suggested to be involved in the etiology of breast cancer, the underlying molecular and genetic mechanisms are not fully understood. The present investigation is a continuation of our efforts to determine the genetic risk factors involved in the development of human breast cancer. In this study, we identified several common PALB2 polymorphisms that confer low-penetrance susceptibility to breast cancer. Our findings support the concept that genetic variants of the PALB2 gene confer breast cancer susceptibility and that this gene can be a potential target for the prevention and treatment of breast cancer. This discovery also offers an opportunity to use PALB2 as a biomarker in disease prediction and diagnosis of breast cancer. In addition, PALB2 variants, as personal genetic/genomic profiling factors, can be used to predict cancer predisposition, leading to personalized approaches to cancer prevention and/or clinical management. Thus, these results may be applied to future clinical and basic cancer research and practice for improvements in the diagnosis, prevention, and treatment of breast cancer.

Breast cancer is the most common cancer of women, with an estimated 178,000 new cases and 40,000 deaths in the United States in 2007 (1). Although the incidence of breast cancer in China is one-fifth of that in the United States, there has been a recent increase in the incidence of this disease (2, 3). The incidence of breast cancer is about twice as common in women with first-degree relatives affected by the disease compared with the general population, and the risk ratio increases with increasing numbers of affected first-degree relatives. This indicates that hereditary factors are involved in the development of breast cancer (4). In the 1990s, mutations of BRCA1, BRCA2, and TP53, which confer a high risk (5- to 10-fold) of breast cancer, were identified (57). However, these mutations account for less than 5% of all patients with breast cancer, and less than 25% of those with familial cancers (8). Moderate-penetrance mutations of CHEK2, BRIP1, and ATM, which are rare but interact with BRCA1 and/or TP53, confer a 2-fold risk of breast cancer, have also been identified (911).

Breast cancer susceptibility is thought to be associated with a “polygenic” model, in which a number of loci contribute to susceptibility, with each locus contributing only a small effect (12). Obviously, common variants of the abovementioned six genes are good candidates for breast cancer susceptibility. Their associations with breast cancer have been evaluated by a number of research groups, but studies have found conflicting results or no significant association (1320). To date, several genome-wide association studies have been accomplished in which several novel susceptibility loci have been identified (2124).

Fanconi anemia (FA) is an autosomal recessive disorder characterized by cellular hypersensitivity to DNA cross-linking agents and predisposition to cancer (25). Biallelic BRCA2 mutations cause the D1 subtype of FA, and the phenotype of this subtype is different from other types, with a high risk of childhood solid tumors (26). Two research groups recently identified a new subtype of FA, FA-N, which has a similar phenotype as FA-D1 but lacks BRCA2 mutations (26, 27).

PALB2 (partner and localizer of BRCA2), a nuclear partner of BRCA2, is required for the intranuclear localization and stability of BRCA2 to execute its functions in error-free DNA double-strand break repair by homologous recombination and checkpoint control in intra–S phase DNA damage processes (28). Reduction of the expression of PALB2 by small interfering RNA, similar to a BRCA2 small interfering RNA, sensitized HeLa cells to mitomycin C, and resulted in a phenotype typical of FA (28). Because the defects in PALB2 were related to this new FA subtype, and affected subjects had a tendency to develop childhood cancer, PALB2 is also known as FANCN (FA complementation group N). In a subsequent study in the United Kingdom, monoallelic truncating mutations of PALB2 were found to be associated with a 2.3-fold increased risk of breast cancer in individuals with familial breast cancer (29). In Finland, a new PALB2 mutation, found in patients with or without a family history of breast cancer, conferred a 4-fold increased risk (30). Other studies have also found that truncating mutations in PALB2 conferred a moderate risk (31, 32). These data indicate that PALB2 mutations are moderate-penetrance susceptibility factors for breast cancer. Still, these mutations are rare, they occur in less than 1% of unselected breast cancers, and in less than 3% of patients with familial breast cancers (2932).

Considering the fact that PALB2 is important for FA and DNA double-strand break repair, common variants of PALB2 are candidates for genetic susceptibility to breast cancer. However, no previous study has investigated common single nucleotide polymorphisms (SNP) of PALB2 for association with breast cancer susceptibility. In the present study, we analyzed common SNPs of PALB2 and their associations with breast cancer risk in a case-control study using a SNP-tagging method.

Study participants. In this study, 1,058 patients with clinically confirmed breast cancer were recruited from the Departments of Breast Surgery of the Cancer Hospital of Jiangsu Province (Nanjing), the First Affiliated Hospital of Nanjing Medical University, and the Nanjing Gulou Hospital, Jiangsu Province, China, between January 2004 and May 2007. A vast majority of participating subjects were included in a previously published study (33). All subjects were genetically unrelated ethnic Han Chinese women from hospitals in Nanjing City and surrounding regions served by these hospitals. Patients with a previous history of cancer, metastatic cancer, or previous radiotherapy or chemotherapy were excluded. The 1,078 cancer-free controls were ethnic Han Chinese women frequency-matched to the cases based on age (±5 years) and residential area (urban versus rural). These women were randomly selected from a pool of 20,000 individuals participating in a community-based screening program for noninfectious diseases conducted in Jiangsu Province from 2004 to 2005. All subjects provided a 5-mL venous blood sample. After informed consent was obtained, each participant was personally interviewed by trained interviewers using a pretested questionnaire to obtain information about demographic data, menstrual data, reproductive history, and the cancer history of first-degree relatives (parents, siblings, and children). If first-degree relatives had a history of any kind of cancer, it was recorded as “yes”, whereas patients without any first-degree relatives with a cancer history were marked as “no.” The study was approved by the Institutional Review Board of Nanjing Medical University.

Selection of tag SNPs for PALB2. The PALB2 gene, which is ∼38 kb long and is located on chromosome 16p12.1, contains 13 exons and 12 introns and encodes 1,186 amino acids [∼130 kDa; National Center for Biotechnology Information (NCBI) database]. The 5′-end of PALB2 is ∼130 bp from the 5′-end of DCTN5 (dynactin 4), and the 3′-end of PALB2 is ∼7 kb from the NADH dehydrogenase 1 (NOUFAB1). There have been 183 SNPs reported in the dbSNP database,3

none of which are the common SNPs [i.e., minor allele frequency (MAF) > 0.05] in the coding regions or in the promoter region for Asian populations. Based on the HapMap CHB database (HapMap data, Rel 22/phase II Apr07, on NCBI B36 assembly, dbSNP b126; population: Han Chinese in Beijing, China; MAF are >0.05) and the NCBI database (MAF >0.05 in the Chinese population), we identified 18 common SNPs in the PALB2 gene (including the region ∼300 bp upstream from the 5′-end, the whole PALB2 gene sequence and ∼6.5 kb downstream the 3′-end; ∼45 kb total). The HapMap database region was chr16:23515500.23560521 for tagging SNPs (tSNP) selection. In addition, a new common variant, rs13330119, was recently identified in this region based on the recently updated HapMap database (HapMap data, Rel 23a/phase II Mar08, on NCBI B36 assembly, dbSNP b126). We applied the criterion that the r2 threshold must be >0.8 in a “pairwise tagging only” mode with the Haploview v.3.32 software program4 to select tSNPs. Four SNPs (rs120963, rs249935, rs249954, and rs16940342) were selected as tSNPs for the association analysis with a mean r2 being 0.958 according to the updated database. The linkage disequilibrium (LD) between the 19 common variants is shown in Fig. 1.

Fig. 1.

LD of the 19 common variants (MAF > 0.05) based on the HapMap CHB database. The number in each square represents the correlation (r2) between each pair of SNPs; the squares with no number represent r2 = 1. Four SNPs (rs120963, rs249935, rs249954, and rs16940342) were chosen to tag the 19 common variants.

Fig. 1.

LD of the 19 common variants (MAF > 0.05) based on the HapMap CHB database. The number in each square represents the correlation (r2) between each pair of SNPs; the squares with no number represent r2 = 1. Four SNPs (rs120963, rs249935, rs249954, and rs16940342) were chosen to tag the 19 common variants.

Close modal

DNA isolation and genotyping. Genomic DNA was extracted from leukocytes by proteinase K digestion followed by phenol-chloroform extraction and ethanol precipitation. The tSNPs of PALB2 were genotyped using the GenomeLab SNPstream 12-plex Genotyping System (Beckman Coulter, Inc.) following the manufacturer's instructions. This platform uses a single base pair extension reaction to incorporate two-color fluorescence terminal nucleotides which are detected by a specialized imager (34). This high-throughput method has been previously validated (35, 36). Briefly, a 100 to 150-bp DNA sequence including each tSNP site was amplified from the genomic DNA using HotStar Taq DNA Polymerase (Qiagen, Germany) in a multiplex PCR. Primers for each tSNP are listed in Table 1. The PCR products were purified with exonuclease I (U.S. Biochemical Corporation) and shrimp alkaline phosphatase (U.S. Biochemical Corporation). After the purification step, an extension primer (Table 1) for each tSNP was used, together with the extension mix (Beckman Coulter) to make a specific extension product, which was then hybridized with the SNPware Plate (Beckman Coulter). The plate was screened by the specialized imager of this SNPstream system, and the genotype data were analyzed by GenomeLab SNPstream version 2.2 software (Beckman Coulter). To ensure genotyping quality, we added 24 (6.25%) randomly selected replications and 4 blanks for each 384-well plate and found a concordance rate of >99%. For each loci assayed, nine cases and five controls were excluded in further analyses for fail genotyping, the final analyses included 1,049 cases and 1,073 controls.

Table 1.

Primary information of the genotyping procedure for tSNPs of the PALB2 gene

tSNP*Amplification PCR primerPCR product (bp)Extension primer (45 bp)
rs120963 (T>C) 5′-CACATTTATTAGTCTTAGGTATTGAAATTG-3′ (sense)5′-TCCGTCTCTATTACAAAAAAGATTAAA-3′ (antisense) 155 5′-GACCTGGGTGTCGATACCTATAAATGTTAATACTTGTGGAAGAAT-3′ 
rs249935 (A>G) 5′-CAGAAAATTTACCAAGCAATCAC-3′ (sense)5′-TCTCATTAAGTGAGAGCAGTAAGTCA-3′ (antisense) 151 5′-AGATAGAGTCGATGCCAGCTATTGCTATCCAATTTTGAAAAAAGA-3′ 
rs249954 (C>T) 5′-AAGAATATTCTCTTCCAGTGTCTCG-3′ (sense)5′-AAATCTGTTTTTCTGAATCTGTTTACC-3′ (antisense) 133 5′-GCGGTAGGTTCCCGACATATTCTTACATATGAAAACTTTAATGAA-3′ 
rs16940342 (A>G) 5′-CAAGTTCCTAAAACAAAAAAACAAC-3′ (sense)5′-TAATTAGTCACAGCTTATAGACCAAGTC-3′ (antisense) 100 5′-ACGCACGTCCACGGTGATTTTGCAAAAATAGGTAGCAGGCTGCTC-3′ 
tSNP*Amplification PCR primerPCR product (bp)Extension primer (45 bp)
rs120963 (T>C) 5′-CACATTTATTAGTCTTAGGTATTGAAATTG-3′ (sense)5′-TCCGTCTCTATTACAAAAAAGATTAAA-3′ (antisense) 155 5′-GACCTGGGTGTCGATACCTATAAATGTTAATACTTGTGGAAGAAT-3′ 
rs249935 (A>G) 5′-CAGAAAATTTACCAAGCAATCAC-3′ (sense)5′-TCTCATTAAGTGAGAGCAGTAAGTCA-3′ (antisense) 151 5′-AGATAGAGTCGATGCCAGCTATTGCTATCCAATTTTGAAAAAAGA-3′ 
rs249954 (C>T) 5′-AAGAATATTCTCTTCCAGTGTCTCG-3′ (sense)5′-AAATCTGTTTTTCTGAATCTGTTTACC-3′ (antisense) 133 5′-GCGGTAGGTTCCCGACATATTCTTACATATGAAAACTTTAATGAA-3′ 
rs16940342 (A>G) 5′-CAAGTTCCTAAAACAAAAAAACAAC-3′ (sense)5′-TAATTAGTCACAGCTTATAGACCAAGTC-3′ (antisense) 100 5′-ACGCACGTCCACGGTGATTTTGCAAAAATAGGTAGCAGGCTGCTC-3′ 
*

rs249935 tagged rs414000, rs447529, rs463804, rs484226, rs463399, rs249932, rs811892, rs513313, rs249942, rs425629, and rs465017; rs120963 tagged rs12162020 and rs3096145; rs249954 tagged rs420259; and rs16940342 tagged rs13330119, according to the HapMap data (Rel 23a/phase II Mar08), on NCBI B36 assembly, dbSNP b126 CHB database.

Statistical methods. Differences in demographic characteristics and selected variables were evaluated with the χ2 test (for categorical variables) and Student's t test (for continuous variables). The χ2 test (1 df) was used to assess the Hardy-Weinberg equilibrium for each tSNP of the PALB2 gene. Unconditional logistic regression was used to compare the genotype frequencies of each tSNP between cases and controls. The common homozygote was used as the reference to calculate genotype-specific odds ratio (OR) and its 95% confidence intervals (CI) with adjustment for age, menopausal status, age at menarche, and first-degree relatives family history under the three assumed (additive, dominant, and recessive) models. The trend test was also carried out in the analysis. We carried out multiple hypothesis testing using the Benjamini-Hochberg method to control the false discovery rate (FDR) in the unconditional logistic regression analysis (37). An FDR of 0.05 was used as a critical value to assess whether the obtained P values were significant. For the three tSNPs with FDR-adjusted P < 0.05, their association with breast cancer risk was analyzed with stratification by age, menopausal status (premenopausal or postmenopausal), age at menarche (below or above 16), and first-degree relative cancer history (yes or no) under the dominant model. All statistical analyses were done with Statistical Analysis System software (v.9.1.3e; SAS Institute).

In general, we found a significant difference between cancer patients and controls with regard to the age at menarche (P < 0.0001), age at their first live birth (P < 0.0001), and the existence of a first-degree relative with a history of cancer (P < 0.0001; Table 2). The call rate of each tSNP was >90%, and the Hardy-Weinberg equilibrium test for each tSNP passed at the P = 0.05 level (Table 3).

Table 2.

Characteristics of breast cancer patients and cancer-free controls and the distributions of select variables

VariableCases, n (%)Ntotal = 1,049Controls, n (%) Ntotal = 1,073P
Age, mean years (±SD) 51.59 ± 11.18 51.77 ± 11.10 0.712 
Age at menarche, mean years (±SD) 15.29 ± 1.94 16.03 ± 1.92 <0.0001 
Age at menopause, mean years (±SD)* 48.99 ± 4.42 49.26 ± 4.23 0.307 
Age at first live birth, mean years (±SD) 25.58 ± 3.42 24.94 ± 3.20 <0.0001 
Menopause status   0.33 
    Premenopause 509 (48.52%) 498 (46.41%)  
    Postmenopause 540 (51.48%) 575 (53.59%)  
First-degree relative family history of cancer   <0.0001 
    Yes 305 (29.08%) 231 (21.53%)  
    No 744 (70.92%) 842 (78.47%)  
VariableCases, n (%)Ntotal = 1,049Controls, n (%) Ntotal = 1,073P
Age, mean years (±SD) 51.59 ± 11.18 51.77 ± 11.10 0.712 
Age at menarche, mean years (±SD) 15.29 ± 1.94 16.03 ± 1.92 <0.0001 
Age at menopause, mean years (±SD)* 48.99 ± 4.42 49.26 ± 4.23 0.307 
Age at first live birth, mean years (±SD) 25.58 ± 3.42 24.94 ± 3.20 <0.0001 
Menopause status   0.33 
    Premenopause 509 (48.52%) 498 (46.41%)  
    Postmenopause 540 (51.48%) 575 (53.59%)  
First-degree relative family history of cancer   <0.0001 
    Yes 305 (29.08%) 231 (21.53%)  
    No 744 (70.92%) 842 (78.47%)  
*

Information for age at menopause was available for 522 (49.76%) cases with breast cancer and for 518 (48.27%) controls.

Age at first live birth information was available for 1,006 (95.90%) cases with breast cancer and for 1,043 (97.20%) controls.

Table 3.

Summary data for correlation, call rate, and Hardy-Weinberg equilibrium test for each tSNP of the PALB2 gene

tSNPGenotypeCases n (%)
Controls n (%)
OR (95% CI)OR (95% CI)*PP-trend*PCall rate
HWE P
n = 1,010n = 1,03396.3%
rs249954 CC 331 (32.77) 412 (39.88) 1.00 (Reference) 1.00 (Reference)      
 TC 524 (51.88) 477 (46.18) 1.36 (1.12-1.65) 1.37 (1.13-1.66)      
 TT 155 (15.35) 144 (13.94) 1.32 (1.01-1.72) 1.33 (1.01-1.75) 0.005 0.006 0.020  0.701 
 TT/TC vs. CC (Dom T)   1.35 (1.13-1.62) 1.36 (1.13-1.64) 0.001  0.004   
 TT vs. TC/CC (Rec T)   1.10 (0.86-1.41) 1.11 (0.87-1.43) 0.414     
           

 

 
n = 997
 
n = 1,008
 

 

 

 

 

 
94.5%
 

 
rs120963 TT 428 (42.93) 488 (48.41) 1.00 (Reference) 1.00 (Reference)      
 CT 459 (46.04) 436 (43.25) 1.20 (1.00-1.45) 1.20 (0.99-1.45)      
 CC 110 (11.03) 84 (8.33) 1.50 (1.10-2.04) 1.49 (1.08-2.05) 0.019 0.007 0.038  0.331 
 CC/CT vs. TT (Dom T)   1.25 (1.05-1.49) 1.25 (1.04-1.49) 0.014  0.028   
 CC vs. CT/TT (Rec C)   1.36 (1.01-1.84) 1.36 (1.01-1.85) 0.041  0.164   
           

 

 
n = 1,031
 
n = 1,052
 

 

 

 

 

 
98.2%
 

 
rs16940342 AA 561 (54.41) 620 (58.94) 1.00 (Reference) 1.00 (Reference)      
 GA 423 (41.03) 377 (35.84) 1.24 (1.04-1.48) 1.25 (1.04-1.50)      
 GG 47 (4.56) 55 (5.23) 0.95 (0.63-1.42) 0.97 (0.64-1.46) 0.050 0.115 0.067  0.813 
 GG/GA vs. AA (Dom G)   1.20 (1.01-1.43) 1.21 (1.02-1.45) 0.037  0.049   
 GG vs. GA/AA (Rec G)   0.87 (0.58-1.29) 0.88 (0.59-1.33) 0.479     
           

 

 
n = 1,035
 
n = 1,058
 

 

 

 

 

 
98.6%
 

 
rs249935 AA 720 (69.57) 762 (72.02) 1.00 (Reference) 1.00 (Reference)      
 GA 290 (28.02) 264 (24.95) 1.16 (0.96-1.41) 1.14 (0.93-1.39)      
 GG 25 (2.42) 32 (3.02) 0.83 (0.49-1.41) 0.88 (0.51-1.52) 0.221 0.453   0.123 
 GG/GA vs. AA (Dom G)   1.13 (0.93-1.36) 1.11 (0.92-1.35) 0.216     
 GG vs. GA/AA (Rec G)   0.79 (0.47-1.35) 0.85 (0.50-1.46) 0.392     
tSNPGenotypeCases n (%)
Controls n (%)
OR (95% CI)OR (95% CI)*PP-trend*PCall rate
HWE P
n = 1,010n = 1,03396.3%
rs249954 CC 331 (32.77) 412 (39.88) 1.00 (Reference) 1.00 (Reference)      
 TC 524 (51.88) 477 (46.18) 1.36 (1.12-1.65) 1.37 (1.13-1.66)      
 TT 155 (15.35) 144 (13.94) 1.32 (1.01-1.72) 1.33 (1.01-1.75) 0.005 0.006 0.020  0.701 
 TT/TC vs. CC (Dom T)   1.35 (1.13-1.62) 1.36 (1.13-1.64) 0.001  0.004   
 TT vs. TC/CC (Rec T)   1.10 (0.86-1.41) 1.11 (0.87-1.43) 0.414     
           

 

 
n = 997
 
n = 1,008
 

 

 

 

 

 
94.5%
 

 
rs120963 TT 428 (42.93) 488 (48.41) 1.00 (Reference) 1.00 (Reference)      
 CT 459 (46.04) 436 (43.25) 1.20 (1.00-1.45) 1.20 (0.99-1.45)      
 CC 110 (11.03) 84 (8.33) 1.50 (1.10-2.04) 1.49 (1.08-2.05) 0.019 0.007 0.038  0.331 
 CC/CT vs. TT (Dom T)   1.25 (1.05-1.49) 1.25 (1.04-1.49) 0.014  0.028   
 CC vs. CT/TT (Rec C)   1.36 (1.01-1.84) 1.36 (1.01-1.85) 0.041  0.164   
           

 

 
n = 1,031
 
n = 1,052
 

 

 

 

 

 
98.2%
 

 
rs16940342 AA 561 (54.41) 620 (58.94) 1.00 (Reference) 1.00 (Reference)      
 GA 423 (41.03) 377 (35.84) 1.24 (1.04-1.48) 1.25 (1.04-1.50)      
 GG 47 (4.56) 55 (5.23) 0.95 (0.63-1.42) 0.97 (0.64-1.46) 0.050 0.115 0.067  0.813 
 GG/GA vs. AA (Dom G)   1.20 (1.01-1.43) 1.21 (1.02-1.45) 0.037  0.049   
 GG vs. GA/AA (Rec G)   0.87 (0.58-1.29) 0.88 (0.59-1.33) 0.479     
           

 

 
n = 1,035
 
n = 1,058
 

 

 

 

 

 
98.6%
 

 
rs249935 AA 720 (69.57) 762 (72.02) 1.00 (Reference) 1.00 (Reference)      
 GA 290 (28.02) 264 (24.95) 1.16 (0.96-1.41) 1.14 (0.93-1.39)      
 GG 25 (2.42) 32 (3.02) 0.83 (0.49-1.41) 0.88 (0.51-1.52) 0.221 0.453   0.123 
 GG/GA vs. AA (Dom G)   1.13 (0.93-1.36) 1.11 (0.92-1.35) 0.216     
 GG vs. GA/AA (Rec G)   0.79 (0.47-1.35) 0.85 (0.50-1.46) 0.392     

Abbreviations: Rec, recessive; Dom, dominant; HWE, Hardy-Weinberg equilibrium.

*

Adjusted by age, age at menarche, menopause status, and first-degree relative family history of cancer.

Two-sided χ2 test.

FDR-adjusted P value for multiple hypothesis testing using the Benjamini-Hochberg method.

Three tSNPs were significantly associated with breast cancer risk under the additive model (Table 3). The association for tSNP rs249935 was not significant (P = 0.221), whereas the association between tSNP rs249954 and breast cancer risk was the strongest (adjusted OR, 1.33; 95% CI, 1.01-1.75 for the TT genotype and adjusted OR, 1.37; 95% CI, 1.13-1.66 for the TC genotype; Table 3), compared with the common homozygote CC (P = 0.005 and P-trend = 0.006). The tSNP rs120963 was also associated with an increased risk of breast cancer (adjusted OR, 1.49; 95% CI, 1.08-2.05 for CC and adjusted OR, 1.20; 95% CI, 0.99-1.45 for CT compared with the TT genotype; P = 0.019; Table 3) in a dose-dependent manner (P-trend = 0.007). The tSNP rs16940342 showed a marginal association with breast cancer risk (P = 0.050) without an obvious trend (P-trend = 0.115). However, after being determined by multiple hypothesis testing, only rs249954 and rs120963 were found to be associated with breast cancer risk, with the FDR-adjusted P values being 0.020 and 0.038, respectively.

Under the assumption of a dominant model, tSNPs rs249954 and rs129063 were associated with increased breast cancer risk (P = 0.001 and 0.014, respectively; Table 3) with FDR-adjusted P < 0.05 (adjusted P = 0.004 and P = 0.028 for rs249954 and rs120963, respectively). tSNP rs16940342 also showed a significant risk association (adjusted OR, 1.21; 95% CI, 1.02-1.45; P = 0.037; Table 3), and had a FDR-adjusted P < 0.05 (adjusted P = 0.0493), indicating that G is another allele that associated with breast cancer risk. Under the assumption of a recessive model, only tSNP rs120963 showed a marginal association (P = 0.041) with an adjusted OR of 1.37 (95% CI, 1.01-1.85). However, based on the multiple hypothesis testing, this association was not significant (FDR-adjusted P = 0.164). No significant association between rs249935 and breast cancer risk was detected under either of the assumptions (P = 0.216 and 0.392, respectively; Table 3).

We stratified the three positively associated tSNPs by selected factors: age, age at menarche (below or above 16), menopausal status (premenopausal or postmenopausal), and whether the subject had a first-degree relative with a history of cancer (yes or no) for further exploration of the association of PALB2 variants with breast cancer risk based on a dominant model (Table 4). We found that the association of rs249954 with breast cancer risk was more evident in the subjects with an age at menarche of >16 (adjusted OR, 1.54; 95% CI, 1.19-2.00) and those without a first-degree relative with a history of cancer (adjusted OR, 1.37; 95% CI, 1.10-1.69). The associations between rs120963 or rs16940342 and breast cancer risk were also stronger in participants with an age at menarche of >16 (adjusted OR, 1.29; 95% CI, 1.00-1.66 for rs120963 and adjusted OR, 1.32; 95% CI, 1.03-1.69 for rs16940342) and those without a first-degree relative with a history of cancer (adjusted OR, 1.26; 95% CI, 1.02-1.56 for rs120963 and adjusted OR, 1.23; 95% CI, 1.00-1.51 for rs16940342). On the other hand, only rs120963 was strongly associated with premenopausal women (adjusted OR, 1.32; 95% CI, 1.02-1.72).

Table 4.

Stratified analysis of the three significantly associated loci of PALB2

Stratification factorsrs249954
rs120963
rs16940342
GenotypeCases, n = 1,010Controls, n = 1,033OR (95% CI)*GenotypeCases, n = 997Controls, n = 1,008OR (95% CI)*GenotypeCases, n = 1,031Controls, n = 1,052OR (95% CI)*
Age             
    <50 CC 172 205 1.00 TT 228 243 1.00 AA 282 311 1.00 
 TC + TT 342 302 1.35 (1.04-1.76) CT + CC 280 249 1.24 (0.96-1.60) GA + GG 240 209 1.26 (0.98-1.62) 
    ≥50 CC 159 205 1.00 TT 200 245 1.00 AA 279 309 1.00 
 TC + TT 337 321 1.38 (1.06-1.79) CT + CC 289 271 1.28 (0.99-1.65) GA + GG 230 223 1.18 (0.92-1.52) 
Age at menarche             
    <16 CC 186 177 1.00 TT 253 204 1.00 AA 332 250 1.00 
 TC + TT 393 242 1.54 (1.19-2.00) CT + CC 323 203 1.27 (0.98-1.64) GA + GG 263 175 1.14 (0.88-1.47) 
    ≥16 CC 145 233 1.00 TT 175 284 1.00 AA 229 370 1.00 
 TC + TT 286 381 1.23 (0.95-1.60) CT + CC 246 317 1.29 (1.00-1.66) GA + GG 207 257 1.32 (1.03-1.69) 
Menopause status             
    Premenopause CC 162 196 1.00 TT 216 237 1.00 AA 271 293 1.00 
 TC + TT 327 282 1.39 (1.07-1.82) CT + CC 270 226 1.32 (1.02-1.72) GA + GG 228 202 1.23 (0.95-1.58) 
    Postmenopause CC 169 214 1.00 TT 212 251 1.00 AA 290 327 1.00 
 TC + TT 352 341 1.33 (1.03-1.72) CT + CC 299 294 1.19 (0.92-1.52) GA + GG 242 230 1.21 (0.95-1.55) 
First-degree relative, family history of cancer             
    No CC 231 317 1.00 TT 300 374 1.00 AA 391 483 1.00 
 TC + TT 487 494 1.37 (1.10-1.69) CT + CC 405 406 1.26 (1.02-1.56) GA + GG 342 346 1.23 (1.00-1.51) 
    Yes CC 100 93 1.00 TT 128 114 1.00 AA 170 137 1.00 
 TC + TT 192 129 1.35 (0.93-1.94) CT + CC 164 114 1.20 (0.84-1.71) GA + GG 128 86 1.17 (0.81-1.67) 
Stratification factorsrs249954
rs120963
rs16940342
GenotypeCases, n = 1,010Controls, n = 1,033OR (95% CI)*GenotypeCases, n = 997Controls, n = 1,008OR (95% CI)*GenotypeCases, n = 1,031Controls, n = 1,052OR (95% CI)*
Age             
    <50 CC 172 205 1.00 TT 228 243 1.00 AA 282 311 1.00 
 TC + TT 342 302 1.35 (1.04-1.76) CT + CC 280 249 1.24 (0.96-1.60) GA + GG 240 209 1.26 (0.98-1.62) 
    ≥50 CC 159 205 1.00 TT 200 245 1.00 AA 279 309 1.00 
 TC + TT 337 321 1.38 (1.06-1.79) CT + CC 289 271 1.28 (0.99-1.65) GA + GG 230 223 1.18 (0.92-1.52) 
Age at menarche             
    <16 CC 186 177 1.00 TT 253 204 1.00 AA 332 250 1.00 
 TC + TT 393 242 1.54 (1.19-2.00) CT + CC 323 203 1.27 (0.98-1.64) GA + GG 263 175 1.14 (0.88-1.47) 
    ≥16 CC 145 233 1.00 TT 175 284 1.00 AA 229 370 1.00 
 TC + TT 286 381 1.23 (0.95-1.60) CT + CC 246 317 1.29 (1.00-1.66) GA + GG 207 257 1.32 (1.03-1.69) 
Menopause status             
    Premenopause CC 162 196 1.00 TT 216 237 1.00 AA 271 293 1.00 
 TC + TT 327 282 1.39 (1.07-1.82) CT + CC 270 226 1.32 (1.02-1.72) GA + GG 228 202 1.23 (0.95-1.58) 
    Postmenopause CC 169 214 1.00 TT 212 251 1.00 AA 290 327 1.00 
 TC + TT 352 341 1.33 (1.03-1.72) CT + CC 299 294 1.19 (0.92-1.52) GA + GG 242 230 1.21 (0.95-1.55) 
First-degree relative, family history of cancer             
    No CC 231 317 1.00 TT 300 374 1.00 AA 391 483 1.00 
 TC + TT 487 494 1.37 (1.10-1.69) CT + CC 405 406 1.26 (1.02-1.56) GA + GG 342 346 1.23 (1.00-1.51) 
    Yes CC 100 93 1.00 TT 128 114 1.00 AA 170 137 1.00 
 TC + TT 192 129 1.35 (0.93-1.94) CT + CC 164 114 1.20 (0.84-1.71) GA + GG 128 86 1.17 (0.81-1.67) 
*

Adjusted by age, age at menarche, menopausal status, and the first-degree relative family history of cancer (excluding the stratified factors in each stratum).

We used a tagging approach to evaluate the associations between the common variants of PALB2 and breast cancer risk. Based on the HapMap CHB database, a region of about 45 kb covering the whole PALB2 gene was analyzed; the SNPs identified in the PALB2 region were consistent with those reported in the NCBI database.3

For the present study, four SNPs with a mean correlation of 0.959 were selected from 19 tagged common SNPs. Our data showed that tSNPs rs249954 and rs120963 were associated with breast cancer risk when an additive model was used. Based on a dominant model, three of the four tSNPs (rs249954, rs120963, and rs16940342) were significantly associated with breast cancer risk in the single-locus analysis. Even after adjustment for multiple testing, these three tSNPs were still significantly associated with breast cancer risk, with a FDR-adjusted P < 0.05.

SNP rs249954 tags SNP rs420259 (r2 = 1, D′ = 1), which was reported to be associated with the risk of bipolar disorder (P = 6.29 × 10−8) following a genome-wide association study of cases from the United Kingdom (38). This indicates that diseases may share some fraction of genetic factors. Because SNP rs249954 is located in intron 6 and rs420259 is in intron 9 of the PALB2 gene, it is uncertain which one of these two variants is the cause of the increase in breast cancer risk. In order to identify additional SNPs that may be in high LD with these two sites that could be associated with breast cancer risk, we screened all of the common variants (with MAF > 0.05) within an approximately 200-kb-long region around these two sites (∼100 kb upstream and ∼100 kb downstream of these loci) based on the CHB HapMap data resource (HapMap data, Rel 23a/phase II Mar08, on NCBI B36 assembly, dbSNP b126). We did not find any SNPs that correlated with these two loci under the “r2 > 0.80” level. This indicates that the association identified here may be unique.

tSNP rs120963 is ∼6.5 kb from the 3′-end of PALB2, but it is in complete LD with SNP rs3096145 (r2 = 1, D′ = 1) and in high LD with rs12162020 (r2 = 0.896, D′ = 0.947) of the 3′-end near the PALB2 gene; therefore, it is highly likely that the 3′-end near gene SNP rs3096145 of PALB2 may be the causal variant. Nevertheless, we cannot exclude the possibility that SNPs in other genes (such as rs152460 and rs152454 in gene UBPH, ∼30 kb from PALB2) found to be in high LD (r2 > 0.80) with rs120963 within a 200 kb region around this locus may be potential sources for the observed effect. In our study, tSNP rs16940342 (in intron 4) was only marginally associated with breast cancer risk (P = 0.05, P-trend = 0.115) in an additive model. However, under the assumption of a dominant model, it showed a stronger association with breast cancer risk (adjusted OR, 1.21; 95% CI, 1.02-1.45; P = 0.037) with a FDR-adjusted P < 0.05. The newly identified common variant rs13330119 is tagged by rs16940342 with a pairwise r2 of 0.941 and may also be the causal variant. SNP rs16940342 is located in intron 4, and rs13330119 is within intron 10 of the PALB2 gene; therefore, it is uncertain which of these two variants underlies the association with breast cancer risk. We also did not find any other SNPs in high LD with these two SNPs within a 200-kb-long region based on the HapMap data (Rel 23a/phase II Mar08), on the NCBI B36 assembly dbSNP b126 CHB database. Therefore, the association we found here may be caused by one single SNP or by the combined effects of these two loci. However, further analysis is needed to validate this finding.

Of note, some intricate effects were found with the variants in the stratification study. Although we attempted to control the confounding factors for the cases and controls used in the study, the definition of cancer family history used in the study may be too broad. Positive definition of family history—if first-degree relatives had a history of any kind of cancers, family history was recorded as “yes”—may be too broad and makes it difficult to interpret the subsequent observation of an increased breast cancer risk of “at-risk” genotypes (under the dominant model) in individuals with no family history of cancer. In future studies, a better-defined family history should be used.

In summary, there were low-penetrance effects of PALB2 variants on breast cancer risk in a Chinese population. Because the etiologic roles of these variants are unknown, their biological functions, and the functions of the variants in LD with the investigated variants, require further validation and mechanistic studies. To the best of our knowledge, this is the first study to evaluate the common variants in PALB2 and their association with breast cancer risk. To confirm our findings, larger epidemiologic studies with ethnically diverse populations are warranted.

No potential conflicts of interest were disclosed.

Grant support: Knowledge Innovation Program of the Chinese Academy of Sciences (KSCX-YW-R-73), Ministry of Science and Technology of China (2007CB947100), Science and Technology Commission of Shanghai Municipality (06DZ19021), Pujiang Talent Program (06PJ14107), Food Safety Research Center and Key Laboratory of Nutrition and Metabolism of the Chinese Academy of Sciences, Department of Education (Innovative Key Grant 705023), and Program for Changjiang Scholars and Innovative Research Team in University (IRT0631).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: P. Chen and J. Liang contributed equally to this work.

We thank our collaborators in the Department of General Surgery, Jiangsu Cancer Hospital (Jinhai Tang, Jianwei Qin) and in the Department of General Surgery (Shui Wang), The First Affiliated Hospital of Nanjing Medical University. We also thank Drs. Qingyi Wei, Donald L. Hill, and Elizabeth Rayburn for assistance in editing this manuscript.

1
Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007.
CA Cancer J Clin
2007
;
57
:
43
–66.
2
Yang L, Parkin DM, Li L, Chen Y. Time trends in cancer mortality in China: 1987-1999.
Int J Cancer
2003
;
106
:
771
–83.
3
Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002.
CA Cancer J Clin
2005
;
55
:
74
–108.
4
Collaborative Group on Hormonal Factors in Breast Cancer. Familial breast cancer: collaborative reanalysis of individual data from 52 epidemiological studies including 58,209 women with breast cancer and 101,986 women without the disease.
Lancet
2001
;
358
:
1389
–99.
5
Garber JE, Goldstein AM, Kantor AF, Dreyfus MG, Fraumeni JF, Jr., Li FP. Follow-up study of twenty-four families with Li-Fraumeni syndrome.
Cancer Res
1991
;
51
:
6094
–7.
6
Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1.
Science
1994
;
266
:
66
–71.
7
Wooster R, Bignell G, Lancaster J, et al. Identification of the breast cancer susceptibility gene BRCA2.
Nature
1995
;
378
:
789
–92.
8
Wooster R, Weber BL. Breast and ovarian cancer.
N Engl J Med
2003
;
348
:
2339
–47.
9
Meijers-Heijboer H, van den Ouweland A, Klijn J, et al. Low-penetrance susceptibility to breast cancer due to CHEK2 (*) 1100delC in noncarriers of BRCA1 or BRCA2 mutations.
Nat Genet
2002
;
31
:
55
–9.
10
Renwick A, Thompson D, Seal S, et al. ATM mutations that cause ataxia-telangiectasia are breast cancer susceptibility alleles.
Nat Genet
2006
;
38
:
873
–5.
11
Seal S, Thompson D, Renwick A, et al. Truncating mutations in the Fanconi anemia J gene BRIP1 are low-penetrance breast cancer susceptibility alleles.
Nat Genet
2006
;
38
:
1239
–41.
12
Pharoah PD, Antoniou A, Bobrow M, Zimmern RL, Easton DF, Ponder BA. Polygenic susceptibility to breast cancer and implications for prevention.
Nat Genet
2002
;
31
:
33
–6.
13
Ishitobi M, Miyoshi Y, Ando A, et al. Association of BRCA2 polymorphism at codon 784 (Met/Val) with breast cancer risk and prognosis.
Clin Cancer Res
2003
;
9
:
1376
–80.
14
Tamimi RM, Hankinson SE, Spiegelman D, Kraft P, Colditz GA, Hunter DJ. Common ataxia telangiectasia mutated haplotypes and risk of breast cancer: a nested case-control study.
Breast Cancer Res
2004
;
6
:
R416
–22.
15
Cox DG, Hankinson SE, Hunter DJ. No association between BRCA2 N372H and breast cancer risk.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
1353
–4.
16
Cox DG, Kraft P, Hankinson SE, Hunter DJ. Haplotype analysis of common variants in the BRCA1 gene and risk of sporadic breast cancer.
Breast Cancer Res
2005
;
7
:
R171
–5.
17
Freedman ML, Penney KL, Stram DO, et al. A haplotype-based case-control study of BRCA1 and sporadic breast cancer risk.
Cancer Res
2005
;
65
:
7516
–22.
18
Einarsdottir K, Humphreys K, Bonnard C, et al. Linkage disequilibrium mapping of CHEK2: common variation and breast cancer risk.
PLoS Med
2006
;
3
:
e168
.
19
Baynes C, Healey CS, Pooley KA, et al. Common variants in the ATM, BRCA1, BRCA2, CHEK2 and TP53 cancer susceptibility genes are unlikely to increase breast cancer risk.
Breast Cancer Res
2007
;
9
:
R27
.
20
Song H, Ramus SJ, Kjaer SK, et al. Tagging single nucleotide polymorphisms in the BRIP1 gene and susceptibility to breast and ovarian cancer.
PLoS ONE
2007
;
2
:
e268
.
21
Easton DF, Pooley KA, Dunning AM, et al. Genome-wide association study identifies novel breast cancer susceptibility loci.
Nature
2007
;
447
:
1087
–93.
22
Hunter DJ, Kraft P, Jacobs KB, et al. A genome-wide association study identifies alleles in FGFR2 associated with risk of sporadic postmenopausal breast cancer.
Nat Genet
2007
;
39
:
870
–4.
23
Murabito JM, Rosenberg CL, Finger D, et al. A genome-wide association study of breast and prostate cancer in the NHLBI's Framingham Heart Study.
BMC Med Genet
2007
;
8
Suppl 1:
S6
.
24
Stacey SN, Manolescu A, Sulem P, et al. Common variants on chromosomes 2q35 and 16q12 confer susceptibility to estrogen receptor-positive breast cancer.
Nat Genet
2007
;
39
:
865
–9.
25
Kutler DI, Singh B, Satagopan J, et al. A 20-year perspective on the International Fanconi Anemia Registry (IFAR).
Blood
2003
;
101
:
1249
–56.
26
Reid S, Schindler D, Hanenberg H, et al. Biallelic mutations in PALB2 cause Fanconi anemia subtype FA-N and predispose to childhood cancer.
Nat Genet
2007
;
39
:
162
–4.
27
Xia B, Dorsman JC, Ameziane N, et al. Fanconi anemia is associated with a defect in the BRCA2 partner PALB2.
Nat Genet
2007
;
39
:
159
–61.
28
Xia B, Sheng Q, Nakanishi K, et al. Control of BRCA2 cellular and clinical functions by a nuclear partner, PALB2.
Mol Cell
2006
;
22
:
719
–29.
29
Rahman N, Seal S, Thompson D, et al. PALB2, which encodes a BRCA2-interacting protein, is a breast cancer susceptibility gene.
Nat Genet
2007
;
39
:
165
–7.
30
Erkko H, Xia B, Nikkila J, et al. A recurrent mutation in PALB2 in Finnish cancer families.
Nature
2007
;
446
:
316
–9.
31
Foulkes WD, Ghadirian P, Akbari MR, et al. Identification of a novel truncating PALB2 mutation and analysis of its contribution to early-onset breast cancer in French-Canadian women.
Breast Cancer Res
2007
;
9
:
R83
.
32
Tischkowitz M, Xia B, Sabbaghian N, et al. Analysis of PALB2/FANCN-associated breast cancer families.
Proc Natl Acad Sci U S A
2007
;
104
:
6788
–93.
33
Wang Y, Tian T, Hu Z, et al. EGF promoter SNPs, plasma EGF levels and risk of breast cancer in Chinese women. Breast Cancer Res Treat 2008;111:321–7.
34
Syvanen AC. Toward genome-wide SNP genotyping.
Nat Genet
2005
;
37
Suppl:
S5
–10.
35
Alcais A, Alter A, Antoni G, et al. Stepwise replication identifies a low-producing lymphotoxin-α allele as a major risk factor for early-onset leprosy.
Nat Genet
2007
;
39
:
517
–22.
36
Rogaeva E, Meng Y, Lee JH, et al. The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease.
Nat Genet
2007
;
39
:
168
–77.
37
Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing.
J R Stat Soc B
1995
;
57
:
289
–300.
38
The Wellcome Trust Case Control Consortium. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls.
Nature
2007
;
447
:
661
–78.