Abstract
Background The peroxisome proliferator–activated receptor alpha (PPARA) and apolipoprotein E (APOE) proteins are reported to be correlated with lipid metabolism, cardiovascular disease, and breast cancer.
Methods We screened APOE and PPARA (S24F and V227A) polymorphisms in 306 breast cancer patients and 300 noncancer controls and determined the relationship between their genetic polymorphisms and breast cancer risk. Interactions with clinical characteristics were also examined.
Results We found that the risk of breast cancer was associated with APOE genotypes (P = 0.014) but not with PPARA S24F or V227A genotypes. The combined effects of F24/APOE genotypes (P = 0.003) on breast cancer risk were more significant than the individual effect of APOE genotypes (P = 0.014). F24/[Latin Small Letter Open E]4 carriers had a higher tendency to develop breast cancer than F24/[Latin Small Letter Open E]3 carriers (P = 0.013), and this effect is stronger than with individual [Latin Small Letter Open E]4 carriers (P = 0.029). In addition, both F24/[Latin Small Letter Open E]4 and V227/[Latin Small Letter Open E]4 carriers were significantly enriched in the human epidermal growth factor receptor 2/neu negative status.
Conclusions These findings suggest that the APOE [Latin Small Letter Open E]4 genotype plays a major role in the prediction of breast cancer, but the PPARA F24 mutation enhances this outcome. The combined effects of F24/[Latin Small Letter Open E]4 genotypes are positively associated with risk of breast cancer.
Epidemiological studies have shown an association between high-fat diets and high incidence of breast cancer.1Previous reports suggest that some lipid-related genes, including estrogen receptor α (ERα), apolipoprotein E (APOE), and peroxisome proliferator–activated receptor alpha (PPARA) are associated with breast cancer risk.2,3Apolipoprotein E has 3 functionally distinct isoforms of the protein (E2, E3, and E4) encoded by the corresponding alleles [Latin Small Letter Open E]2, [Latin Small Letter Open E]3, and [Latin Small Letter Open E]4 in humans.4Our previous study showed that the [Latin Small Letter Open E]4 allele was associated with the breast cancer risk and human epidermal growth factor receptor 2 (HER2)/neu negative status in patients with breast cancer.2However, different studies show that the presence of an [Latin Small Letter Open E]4 allele had neither association with breast cancer risk nor influence on tumor cell proliferation.5,6Therefore, the presence of additional genetic variants is suggested to concomitantly influence the risk of breast cancer. Recently, the PPARA polymorphism (rs4253760) in intron 6 was suggested to be a candidate for increasing breast cancer risk.7However, except for the L162V mutation in exon 5, the correlation between PPARA variants in other exons and breast cancer risk is still unclear.7
The PPARA gene is located on chromosome 22 and consists of 8 exons that encode the PPARα protein. PPARα is a ligand-activated transcription factor that regulates lipid metabolism by controlling the gene expression of β-oxidation enzymes, apolipoproteins, and fatty acid transport proteins.8In rodent breast cancer models, activation of PPARα by long-chain fatty acids and synthetic ligands has been reported to reduce tumor incidence and progression.9In breast cancer cell models, activation of PPARα by arachidonic acid has been reported to stimulate8or inhibit10breast cancer cell proliferation. Genetic variants of PPARA were found to be correlated with lipoprotein levels, cardiovascular disease, obesity, type 2 diabetes, and cancers.11In particular, the L162V and V227A variants of PPARA were associated with dyslipidemia.12,13
We preliminarily screened the S24F, L162V, and V227A polymorphisms in the PPARA gene in 306 patients with breast cancer and 300 cancer-free subjects from previous data collection before 2004.2No significant correlation between the S24F, L162V, and V227A variants of PPARA and breast cancer risk was found (unpublished data). However, a positive association between PPARα and breast tumorigenesis was reported in isolated rat mammary gland epithelial cells14and breast cancer cell lines.8,15Therefore, we suggest that the PPARA gene may play a synergistic role in the development of breast cancer. In this study, we ask whether there is a possible combined effect of APOE and PPARA polymorphisms on the risk of breast cancer. To test this hypothesis, the combined effects of APOE and PPARA polymorphisms (S24F and V227A) on breast cancer risk were studied. Interactions with clinical characteristics were also examined.
MATERIALS AND METHODS
Study Population
The study population was composed of female Taiwanese ranging in age from 25 to 73 years. A total of 606 subjects were recruited between 2001 and 2004; details of the retrospective study population and data collection methods have been published previously.2Patients with breast cancer patients (n = 306; mean [SD] age, 48.5 [11.3]) were recruited from the China Medical University Hospital and Fong Yuan Hospital. Noncancer controls (n = 300; mean [SD] age, 41.3 [10.5]) were recruited from the Taichung Blood Center and China Medical University Hospital.2This study was approved by the Institutional Review Board of the China Medical University Hospital.
Polymorphism Analysis
Genomic DNA was extracted from whole blood by using the Viogene isolation kit (Viogene, Taiwan). APOE genotyping was performed according to the previous report.2The polymorphism V227A (rs1800234), which is located on exon 6 of PPARA, was analyzed using polymerase chain reaction (PCR) restriction fragment length polymorphism analysis. The amplicon was generated with the following PCR primers: forward primer (5′-TCCATAGTGGAAAGCCGA-3′) and reverse primer (5′-TTTCCATCTTCGCGGTCCTT-3′). The PCR product was digested by the Sau96I (New England Biolabs, Ipswich, MA) restriction enzyme. Restriction fragments were separated on a 3% agarose gel. The S24F polymorphism (TCT/serine to TTT/phenylalanine) is caused by a genetic mutation at exon 3 and does not alter any restriction site; a mismatch PCR restriction fragment length polymorphism method was therefore used to genotype all the individuals. The amplicon was generated with the following PCR primers: forward primer (5′-GATCTAGAGAGCCCGTGAT-3′) and reverse primer (5′-GATGGAGACCATCCTGGCTA-3′). The mismatched site, with G substituted for T, is located on the nucleotide 17 of forward primer. Accordingly, the PCR products generated from the mutants cannot be digested by the DpnII (New England Biolabs) restriction enzyme. Restriction fragments were separated on a 3% agarose gel.
Statistical Analysis
All data were analyzed using SPSS version 15.0 for Windows. The genotypes of patients with breast cancer and noncancer controls were compared using the χ2 test and the Fisher exact test. Effects of the PPARA and APOE interaction on clinical characteristics in patients with breast cancer were analyzed using the χ2 test and the Fisher exact test. Logistic regression analysis was performed to test the association of gene polymorphisms with both breast cancer risk and measures of clinical characteristics after adjustment for age. P < 0.05 was considered statistically significant.
RESULTS
Correlation Between Genetic Polymorphisms of APOE and PPARA and Breast Cancer Risk
The genotypes of PPARA and APOE are shown in Table 1. The S24F polymorphism of PPARA was classified as an S24 carrier (C/C) or an F24 carrier (C/T and T/T). The V227A polymorphism was classified as a V227 carrier (T/T) or an A227 carrier (T/C and C/C). The APOE polymorphism was classified as an [Latin Small Letter Open E]2 carrier (2/2, 2/3, or 2/4), an [Latin Small Letter Open E]3 carrier (3/3), or an [Latin Small Letter Open E]4 carrier (3/4 or 4/4). First, we evaluated whether only one gene variant was correlated with breast cancer risk (Table 1). We found that neither the S24F (P = 0.507) nor the V227A (P = 0.703) polymorphism was associated with breast cancer risk. A significant difference in APOE genotype levels between the patients with breast cancer and the noncancer controls was observed (P = 0.014). We then used logistic regression to explore the correlation between the APOE polymorphisms and the breast cancer risk after adjustment for age (Table 1). Because the APOE mutation types are classified as either wild type ([Latin Small Letter Open E]3) or mutation type ([Latin Small Letter Open E]2 and [Latin Small Letter Open E]4), all logistic regression analyses compare the mutation type with the wild type. The results showed that [Latin Small Letter Open E]4 carriers had a higher risk of breast cancer in comparison to [Latin Small Letter Open E]3 carriers (P = 0.029).
Combined Genetic Effects of S24F, V227A, and APOE Polymorphisms on Breast Cancer Risk
The combined genetic effects of 2 gene variants (PPARA/APOE) on the risk for breast cancer were measured (Table 2). The combined effect of S24/APOE carriers showed no significant association with breast cancer (P = 0.493), whereas that of F24/APOE carriers had a significant effect on breast cancer (P = 0.003). This suggested that there was a statistically significant interaction effect of S24F and APOE on breast cancer. A similar result was also found in the interaction between V227A and APOE. A significant correlation between V227/APOE carriers and breast cancer was found (P = 0.012), whereas no significant correlation between A227/APOE and breast cancer was observed (P = 0.772). These findings suggested that both F24 and V227 carriers enhanced the correlation between APOE genotypes and breast cancer. The F24/APOE carriers in particular showed an increased risk of breast cancer with a P value changing from 0.014 to 0.003. Logistic regression analysis showed that F24/[Latin Small Letter Open E]4 genotypes were associated with an increased risk of breast cancer (P = 0.013) in comparison to F24/[Latin Small Letter Open E]3 genotypes. Additionally, F24/[Latin Small Letter Open E]2 carriers showed an increased risk of breast cancer (P = 0.039). Patients with V227/[Latin Small Letter Open E]4 genotypes also had a significant influence on the risk of breast cancer (P = 0.018, V227/[Latin Small Letter Open E]4 vs V227/[Latin Small Letter Open E]3). In addition, when the F24/V227 genotypes and APOE polymorphisms were combined, an additive effect on breast cancer risk was observed in F24/V227/[Latin Small Letter Open E]4 genotypes (P = 0.015, F24/V227/[Latin Small Letter Open E]4 vs F24/V227/[Latin Small Letter Open E]3).
Association of F24/APOE and V227/APOE Carriers With Clinical Characteristics
The associations of F24/APOE and V227/APOE genotypes with clinical characteristics are shown in Supplementary Table 1 (Supplemental Digital Content 1, http://links.lww.com/JIM/A11). No statistically significant relationship was found between F24/APOE genotypes and clinical characteristics including menopause, tumor grade, TNM classification, estrogen receptor, progesterone receptor, associated ductal carcinoma in situ, and lymphatic invasion (P > 0.05). However, the presence or absence of HER2/neu was associated with F24/APOE (P = 0.006) and V227/APOE (P = 0.005) polymorphisms in patients with breast cancer. Logistic regression analysis showed that patients with F24/[Latin Small Letter Open E]4 genotypes were also significantly enriched in HER2/neu negative status (P = 0.007, F24/[Latin Small Letter Open E]4 vs F24/[Latin Small Letter Open E]3; Table 3). Similar results were also found in patients with V227/[Latin Small Letter Open E]4 genotypes (P = 0.005, V227/[Latin Small Letter Open E]4 vs V227/[Latin Small Letter Open E]3). However, 2 gene polymorphisms (F24/APOE, P = 0.005; V227/APOE, P = 0.007) had no additive effect on HER2/neu status over one gene polymorphism (APOE, P = 0.005).
DISCUSSION
In this study, we found that the F24/[Latin Small Letter Open E]4 and V227/[Latin Small Letter Open E]4 genotypes were associated with an increased risk of breast cancer. According to χ2 test, the combined effects of the F24/APOE polymorphisms on breast cancer risk showed more statistical significance (P = 0.003; Table 2) than the single effect of APOE polymorphisms (P = 0.014; Table 1). Logistic regression analysis showed that subjects with F24/[Latin Small Letter Open E]4 genotypes (P = 0.013; Table 2) had a higher tendency to develop breast cancer than subjects with only an [Latin Small Letter Open E]4 genotype (P = 0.029; Table 1). A higher risk tendency was also found in F24/[Latin Small Letter Open E]2 carriers (P = 0.039; Table 2) than in only [Latin Small Letter Open E]2 carriers (P = 0.073; Table 1). Notably, no significant relationship between S24F polymorphisms and breast cancer risk was found with any single mutation of PPARA codon 24. This indicated that S24F polymorphisms did not directly link to breast cancer. These findings suggested that the APOE mutation plays a major role in the prediction of breast cancer and the PPARA F24 mutation enhances this outcome.
We are the first team to study the S24F polymorphism (TCT/serine to TTT/phenylalanine) in the PPARA gene. The T-allele frequency occurred in 79.1% of the breast cancer cases and 77.5% of the noncancer controls. The genotype frequency of C/T and T/T (F carrier) occurred in 92.5% of the breast cancer cases and 91.0% of the noncancer controls (Table 1). The S24F mutation is located on the A/B region of the PPARα protein, which has a ligand-independent activating function (AF-1).16Deletion of the first 25 amino acid residues from the N-terminal side was reported to completely remove the AF-1 activity.16Therefore, we hypothesized that the F24/[Latin Small Letter Open E]4 mutation correlated with an increased risk of breast cancer might link to change AF-1 activity. It is possible that Ser24 mutating to Phe24 might cause a structure change in the α-helix in the A/B region or a change in the AF-1 activity. However, we need more evidence to prove this hypothesis in subsequent studies.
Furthermore, V227/[Latin Small Letter Open E]4 genotypes were also associated with an increased risk of breast cancer (Table 2). The V227A polymorphism is located on the hinge region of PPARA, which can enhance recruitment of the nuclear receptor corepressor and attenuate transactivation functions.13Most studies reported that the V227A polymorphism was associated with dyslipidemia,12,13and few reports focused on the prediction of breast cancer. In this study, we found that the V227A polymorphism was not correlated with the risk of breast cancer (Table 1). In addition, the statistical results between the V227/APOE carriers (P = 0.012) and the APOE carriers (P = 0.014) showed only a small change in P value. Therefore, we suggested that V227A polymorphism might not be a direct cause of breast cancer.
No statistically significant relationship was found between F24/APOE carriers and any clinical characteristic except the absence of HER2/neu, which showed a positive correlation (P = 0.006; Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/JIM/A11). A similar result was also found in V227/APOE carriers (P = 0.006; Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/JIM/A11). HER2/neu negative status in patients with breast cancer may favorably influence response to adjuvant tamoxifen therapy.16 F24/[Latin Small Letter Open E]4 carriers enriched in the HER2/neu negative status might provide useful information in predicting the response to hormone therapy or HER2/neu–targeted therapy in patients with breast cancer.
CONCLUSIONS
This study indicates that the combined effects of the F24/APOE polymorphisms on breast cancer risk show more statistical significance than the single effect of APOE polymorphisms. Subjects with F24/[Latin Small Letter Open E]4 genotypes present an especially high tendency for breast cancer risk.