Association of Mitochondrial DNA D-loop (CA)_n Dinucleotide Repeat Polymorphism with Breast Cancer Risk and Survival among Chinese Women

Abstract

Mitochondrial genome alternations may be involved in carcinogenesis. The non-coding region of the mitochondrial DNA (mtDNA) displacement loop (D-loop) has emerged as a mutational hotspot. Using data from a population-based case-control study conducted among Chinese women in Shanghai, we evaluated associations of breast cancer risk and survival with the mtDNA D-loop (CA)_n dinucleotide repeat polymorphism. Included in the study were 1,058 cases and 1,129 age-frequency-matched community controls who participated in the Shanghai Breast Cancer Study (SBCS) between 1996 and 1998. Breast cancer patients were followed to determine intervals of overall survival (OS) and disease-free survival (DFS). Overall, there was no association between the mtDNAD-loop (CA)_n repeat polymorphism and breast cancer risk. Patients with multiple alleles of the mtDNAD-loop (CA)_n polymorphism (heteroplasmy) had significantly poorer DFS than those with one allele of the mtDNAD-loop (CA)_n polymorphism (hazard ratio1.62, 95% CI: 1.16–2.26). These results suggest that the mtDNAD-loop (CA)_n repeat polymorphism may be associated with breast cancer survival. Additional studies with a larger sample size are warranted.

Introduction

The development of cancer involves the accumulation of various genetic alterations, which are present in both the mitochondrial and nuclear genomes. Mitochondria play a critical role in energy production and oxidative phosphorylation (1). Defects in mitochondrial function are suspected to contribute to the development and progression of cancer (2–4). Human mitochondrial DNA(mtDNA) is a circular molecule consisting of 16,571 bp encoding 2 rRNAs, 22 tRNAs, and 13 polypeptides. The displacement loop (D-loop) is a non-coding region of 1,124 bp (spanning nucleotide positions 16,024 to 516), which acts as a promoter for both the heavy and light strands of mtDNA, and contains essential transcription and replication elements (4). Reactive oxygen species (ROS) have long been thought to damage both nuclear DNA and mtDNAand to play a key role in carcinogenesis (2). Increased ROS generation may alter signal transduction pathways, resulting in activation of oncogenes or inactivation of tumor suppressor genes (5). Since the D-loop region contains the leading strand for the origin of replication and a number of major promoters for transcription (6), it is conceivable that genetic variability in the D-loop region may affect the function of the respiration chain that is responsible for high ROS levels and could contribute to nuclear genome damage and cancer initiation and promotion (7, 8). Moreover, respiratory chain alteration may cause a dysfunction in mitochondrion-induced apoptosis (9).

MtDNA is highly polymorphic (Mitomap, www.mitomap.org). One of the most informative mtDNA variations is the (CA)_n dinucleotide repeat polymorphism between nucleotide position 514 and 523 in the third hypervariable region (HV3) (10, 11), which has been well recognized as a marker of mitochondrial genome instability and is a valuable tool in forensic identity testing and the analysis of crime scene stains (3, 12, 13). In a small study (n=40), Richard et al. reported that 42.5% of breast cancer tissue showed (CA)_n allele changes compared to adjacent normal tissue (3). However, the relationship of the germline (CA)_n dinucleotide repeat polymorphism with risk and survival of breast cancer has not been evaluated.

We used data from the Shanghai Breast Cancer Study (SBCS), a large-scale, population-based case-control study conducted in urban Shanghai from 1996 to 1998, to evaluate the association of the mtDNAD-loop (CA)_n repeat polymorphism with breast cancer risk and survival.

Materials and Methods

Study population

The SBCS is a population-based case-control study conducted among Chinese women in Shanghai. Detailed study methods have been published elsewhere (14). Cancer cases were identified through a rapid case ascertainment system, supplemented by the population-based Shanghai Cancer Registry, which has a virtually complete ascertainment of all incident cancer cases diagnosed among residents in urban Shanghai. Between 1996 and 1998, a total of 1,602 eligible breast cancer cases were identified, of which 1,459 (91.1%) cases completed in-person interviews. Cancer diagnoses for all patients were reviewed and confirmed by two senior pathologists. Controls were randomly selected from the general population of Shanghai using the Shanghai Resident Registry, a population registry containing demographic information for all adult residents of urban Shanghai. The inclusion criteria for controls were identical to those for cases with the exception of a cancer diagnosis. Of the 1,724 eligible women, 1,556 (90.3%) completed in-person interviews. The study protocol was approved by the relevant committees on the use of human subjects in research of all institutes involved in the study, and written, informed consent was obtained from all participants prior to interview.

During the in-person interview, a structured questionnaire was used to elicit detailed information on demographic factors, menstrual and reproductive history, hormone use, dietary habits, previous disease history, physical activity, tobacco and alcohol use, weight history, and family history of cancer. All participants were measured for their current weight, waist and hip circumferences, and sitting and standing height by trained study interviewers using a standard protocol (15). 1,193 (82%) cases and 1,310 (84%) controls donated a blood sample. All of the specimens were collected in the morning before any meals. These samples were processed on the same day, typically within 6 hours of sample collection, and stored at −70°C until relevant bioassays were carried out.

Usual dietary habits over the past 5 years were assessed by in-person interview, using a validated quantitative food frequency questionnaire. The food frequency questionnaire included 76 food items or groups, 30 fresh vegetables and nine fruits, covering over 85% of foods commonly consumed in Shanghai.

The methodology for the follow-up of cancer cases in the SBCS has been described previously (16, 17). Of the 1,455 eligible patients, 1,378 responded to at least one of the active follow-up surveys. During the follow-up period, 266 deaths were identified, 237 from breast cancer, 26 from other diseases, and three from unclear causes. Survival status for the 77 participants who did not respond to the follow-up surveys was established by linkage to the death certificate registry, and 47 deaths were identified with all but one due to breast cancer. The remaining 30 patients had no match in the death certificate registry and were assumed to be still living on December 2004, 6 months before our last search of the registry to allow for a possible delay of entry of death certificates into the registry. Breast cancer relapse information was unavailable for the 30 patients who had no match in the death certificate registry and for one patient who died from other causes before being contacted. Including three women for whom we lacked detailed information on cause of death through active follow-up, 34 women in total were excluded from the disease-free survival analysis.

Genotyping Method

Genomic DNA was extracted from blood samples. Genotyping for the (CA)_n polymorphism was performed by detection of fluorescent amplimers on an ABI PRISM 3700 automated DNA analyzer. Primers were designed using a tailing strategy to promote full non-templated nucleotide addition by AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA), providing unambiguous detection of alleles separated by 1 bp (18). The primers were: forward 5'-CCAGCCTAACCAGATTTC and reverse 5'-gtgtCTTTGAGGAGGTAAGCTA-3'. The forward primer was labeled with 6-carboxyfluorescein. Each 2.2 µl of PCR mixture included 0.1 unit of AmpliTaq Gold DNA polymerase, 1x Buffer II, 2.5 mM MgCl2, 0.25 mM dNTPs, 335 nM concentrations of each primer, and 1 ng of DNA. Thermal cycling conditions were as follows: 95°C for 10 min followed by 10 cycles of 94°C for 15 s, 55°C for 15 s, and 72°C for 30 s; 20 cycles of 89°C for 15 s, 55°C for 15 s, and 72°C for 30 s with a final extension step of 72°C for 10 min.

Allele fragment size estimation was accomplished using the internal size standard Genescan 400HD ROX and the Local Southern algorithm of GENESCAN software. Editing of alleles was performed in GENOTYPER. Allele binning and adjustment of run mobility according to control alleles of CEPH 1347-02 were accomplished by custom software. Each 96-well plate of genomic DNA contained multiple controls, including one water blank, two samples of CEPH 1347-02, two public study control duplicates, and two blinded study control duplicates. Duplicates were distributed across separate 96-well plates. The ABI3700 DNA analyzer has a single laser and an approximate 3-fold attenuation of signal across the capillary array, translating as a weaker signal in wells to the left in a 96-well plate. Consequently, the genotyping assay failure rate could be higher among those samples. To preclude this as a potential source of bias, samples were arrayed such that equal numbers of cases and controls were present in any given plate column. Genotyping data were obtained from 1,058 (88.7%) cases and 1,129 (86.2%) controls who donated blood samples. The major reasons for incomplete genotyping were insufficient DNA and unsuccessful PCR amplification.

Statistic Analysis

χ² statistics were used to evaluate case-control differences in the distribution of genotypes. Logistic regression models were used to estimate odds ratios (ORs) for the association of mtDNAD-loop (CA)_n polymorphisms with breast cancer risk. Analyses stratified by menopausal status, age, body mass index (BMI), waist-to-hip ratio (WHR), years of menstruation, and intake of fruits, vegetables, vitamin supplements, selenium and antioxidant vitamins were conducted to evaluate the homogeneity of the association. Adjusting for age, education, and major breast cancer risk factors had no appreciable effect on OR estimations. Thus, only crude ORs were presented. A composite dietary antioxidant index was derived to incorporate information on intake of four antioxidant nutrients (i.e. selenium and vitamins A, C, and E)(19). The Cox proportional hazard models were applied to evaluate hazard ratios (HRs) for the association of mtDNAD-loop (CA)_n polymorphisms with the overall survival (OS) and disease-free survival (DFS) with adjustment for age, menopausal status, TNM stage, and estrogen receptor/ progesterone receptor (ER/PR) status. Survival time was calculated as the time from cancer diagnosis to the endpoints of the study, censoring at the date of last contact or non-cancer death. For subjects who had died of breast cancer without information on the date of recurrence or metastasis, total survival time was substituted for DFS time (16). The 5-year survival rate was estimated using the Kaplan-Meier method. The log-rank test was applied to test the differences of survival rates across comparison groups.

Results

Selected demographic characteristics and major risk factors for cases and controls are presented in Table 1. Breast cancer cases and controls were comparable in age and education level. With the exception of a family history of breast cancer, statistically significant associations were observed for virtually all major risk factors of breast cancer. There was no appreciable difference between cases included in the genotyping study and the whole study. The same was true for the controls included in the genotyping study and the whole study (data not shown).

Table 1.

Comparison of cases and controls by selected descriptive characteristics, Shanghai Breast Cancer Study, 1996–1998.

Subject Characteristic	Cases (n=1058)	Controls (n=1129)	P-valuea
Age (years), mean ± SD	47.5 ± 7.9	47.1 ± 8.8	0.324
Education, %
Elementary school or below	12.2	14.8
Middle school	44.4	42.4
> Middle school	43.4	42.8	0.195
Breast cancer in first-degree relative, %	3.2	2.4	0.243
Ever had breast fibroadenoma, %	9.6	5.1	<0.001
Age at menarche (years), mean ± SD	14.5 ± 1.7	14.7 ± 1.7	0.004
Ever had a live birth, %	95.2	95.8	0.523
Number of live births, mean ± SD	1.48 ± 0.81	1.53 ± 0.86	0.166
Age at first live birth (years), mean ± SD	26.8 ± 4.1	26.2 ± 3.8	<0.001
Oral contraceptive use, %	20.9	22.0	0.539
Hormone replacement therapy use, %	2.6	2.4	0.804
Post-menopausal, %	32.6	36.1	0.084
Age at menopause (years), mean ± SD	48.1 ± 4.6	47.5 ± 5.1	0.075
Physically active past 10 years, %	19.1	26.3	<0.001
BMI (kg/m2), mean ± SD	23.6 ± 3.5	23.2 ± 3.4	0.041
Waist-to-hip ratio, mean ± SD	0.81± 0.06	0.80± 0.06	0.008

^aFor χ² test (categorical variables) or t test (continuous variables).

A total of 8 (CA)_n repeat alleles were observed in our study population, ranging from 4 repeats [denoted as (CA)₄ ] to 11 repeats [denoted as (CA)₁₁] (Table 2). Alleles (CA)₅ (52.2%) and (CA)₄ (41.9%) were the two most common alleles in this Chinese population. Allele frequencies of 8 to 11 repeats were low (1.1 % in cases and 0.7% in controls), and these alleles were combined into one group [denoted as (CA)_8–11] in the analyses. Table 2 shows the association between mtDNA D-loop (CA)_n repeat polymorphisms and breast cancer risk. Overall, there were no associations of breast cancer risk with the mtDNAD-loop (CA)_n repeat polymorphism. Using the most common alleles [(CA)₅] as the reference group in the OR estimations, allele (CA)₇ was found to be statistically associated with decreased breast cancer risk (OR = 0.50; 95% CI, 0.27–0.93). However, the sample size in the (CA)₇ group is small.

Table 2.

mtDNA D-loop (CA)_n repeat polymorphism, unadjusted and adjusted OR for breast cancer, Shanghai Breast Cancer Study, 1996–1998

D-loop (CA)n	Cases, N (%)	Controls, N (%)	OR (95% CI)a
(CA)5	555 (52.5)	587 (52.0)	1.00 (ref)
(CA)4	449 (42.4)	468 (41.5)	1.02 (0.85–1.21)
(CA)6	27 (2.6)	34 (3.0)	0.84 (0.50–141)
(CA)7	15 (1.4)	32 (2.8)	0.50 (0.27–0.93)
(CA)8–11	12 (1.1)	8 (0.7)	1.59 (0.64–3.91)
Total	1058 (100)	1129 (100)
D–loop (CA)n genotypes
Single alleleb	949 (89.7)	1013 (89.7)	1.00 (ref)
Multiple allelesc	109 (10.3)	116 (10.3)	1.00 (0.76–1.32)
Without any (CA)6–11 allele	57 (5.4)	51 (4.5)	1.19 (0.81–1.76)
With any (CA)6–11 allele	52 (4.9)	65 (5.8)	0.85 (0.59–1.24)
With all (CA)6–11 alleles	16 (1.5)	25 (2.2)	0.68 (0.36–1.29)
Analyses stratified by age
Age < 45
(CA)5	234 (54.0)	243 (51.8)	1.00 (ref)
(CA)4	175 (40.4)	193 (41.2)	0.94 (0.72–1.24)
(CA)6	12 (2.8)	14 (3.0)	0.89 (0.40–1.97)
(CA)7	6 (1.4)	13 (2.8)	0.48 (0.18–1.28)
(CA)8–11	6 (1.4)	6 (1.3)	1.04 (0.33–3.27)
Age : 45–49
(CA)5	110 (44.7)	121(52.2)	1.00 (ref)
(CA)4	124 (50.4)	96 (41.4)	1.40 (0.96–2.02)
(CA)6	6 (2.4)	9 (3.9)	0.72 (0.25–2.09)
(CA)7	4 (1.6)	6 (2.6)	0.72 (0.20–2.62)
(CA)8–11	2 (0.8)	0(0.0)	/
Age ≥ 45
(CA)5	211 (55.7)	223 (52.1)	1.00 (ref)
(CA)4	150 (39.6)	179 (41.8)	0.89 (0.67–1.18)
(CA)6	9 (2.4)	11 (2.6)	0.87 (0.35–2.13)
(CA)7	5 (1.3)	13 (3.0)	0.41 (0.14–1.16)
(CA)8–11	4(1.1)	2(0.5)	2.11(0.38–11.7)

^aCrude OR

^bOne mtDNA D–loop (CA)n repeat allele.

^cMore than one mtDNA D–loop(CA)_n allele.

We also evaluated the association between mtDNA D-loop (CA)_n repeat polymorphisms and breast cancer risk, stratified by BMI, WHR, total years of menstruation, and years of menstruation before first live birth, all of which are indicators of endogenous estrogen exposure. The association did not differ by age groups (<45 years vs 45–49 vs ≥45 years old at the time of diagnosis) (Table 2). There was no significant interaction of BMI, WHR, total years of menstruation, and years of menstruation before first live birth with mtDNAD-loop (CA)_n repeat polymorphisms in relation to breast cancer risk (data not shown). No evidence was found for an interaction between the mtDNAD-loop (CA)_n polymorphism with the intake of any dietary antioxidant (data not shown).

Two hundred and twenty-five women (10.3%) had more than one (CA)_n repeat allele [denoted as multiple alleles], suggesting heteroplasmy in these women. Associations of multiple alleles of the mtDNAD-loop (CA)_n repeat polymorphism and breast cancer risk are shown in Table 2. Overall, carrying multiple alleles of the mtDNAD-loop (CA)_n polymorphism was not associated with breast cancer risk (OR = 1.00; 95% CI, 0.76–1.32). Women with heteroplasmy and who carried any of the (CA)_6–11 alleles had a decreased risk of breast cancer. The ORs were 0.85 (95% CI: 0.59–1.24) for women with any (CA)_6–11 allele and 0.68 (95% CI: 0.36–1.29) for women who carried all (CA)_6–11 alleles (Table 2). The ORs, however, were not statistically significant.

Table 3 presents the association of mtDNA(CA)_n repeat polymorphism and breast cancer survival after adjustment for potential confounding factors, including TNM stage, ER/PR status, radiotherapy, and age. Since the allele frequencies of (CA)₆, (CA)₇, (CA)_8–11 were low, these alleles were combined into one group [denoted as (CA)_6–11]. Overall, neither OS nor disease-free survival (DFS) was associated with the mtDNAD-loop (CA)₄ or (CA)_6–11 alleles when compared with the (CA)₅ allele (Table 3). However, women with multiple alleles (heteroplasmy) of the mtDNAD-loop (CA)_n repeat polymorphism had poorer survival rates. Compared with women with a single allele of the mtDNAD-loop (CA)_n repeat, women who carried multiple alleles had a reduced 5-year DFS rate (76.7% vs 65.6%), and the multivariate-adjusted hazard ratio (HR) was 1.62 (95% CI, 1.16–2.26). Subjects with multiple alleles of the D-loop (CA)_n repeat also had poorer OS rates, although the association was not statistically significant (Table 3). The HRs were 1.54 (95% CI, 0.89–2.65) and 1.63 (95% CI, 1.01–2.64) for overall survival and disease-free survival, respectively, for women with any (CA)₆ allele. We also analyzed the association of D-loop (CA)_n repeat polymorphisms and allele numbers with respect to prognostic factors such as breast cancer stage and ER/PR status. However, neither D-loop (CA)_n polymorphisms nor having multiple alleles (heteroplasmy) were associated with these prognostic factors (data not shown).

Table 3.

Association of mtDNA D-loop (CA)_n repeat polymorphism and breast cancer survival, Shanghai Breast Cancer Study,1996–1998.

	Overall survival				Disease-free survival
Variables	Cases	Events	5-year survival %	HRa (95% CI)	Events	5-year survival %	HR a(95% CI)
Polymorphisms of D–loop (CA)n
(CA)5	554	130	83.6	1.00 (Ref)	160	78.0	1.00 (Ref)
(CA)4	447	89	83.7	0.84(0.64–1.10)	115	73.9	0.86(0.68–1.09)
(CA)6–11	54	12	76.2	1.07(0.59–1.95)	16	69.1	1.13(0.67–1.90)
D-loop (CA)n genotypes
Single alleleb	946	202	83.9	1.00 (Ref)	250	76.7	1.00 (Ref)
Multiple allelesc	109	29	78.9	1.31(0.88–1.94)	41	65.6	1.62(1.16–2.26)
Without any (CA)6–11 allele	57	15	82.5	1.15(0.68–1.95)	23	65.4	1.61(1.05–2.49)
With any (CA)6–11 allele	52	14	71.2	1.54(0.89–2.65)	18	63.8	1.63(1.01–2.64)

HR, hazard ratio; CI, confidence interval; ER, estrogen receptor; PR, progesterone receptor.

^aAdjusted for age, education, TNM stage, radiotherapy, and ER/PR status.

^bOne mtDNA D–loop(CA)_n repeat allele.

^cMore than one mtDNA D-loop(CA)_n repeat allele.

Discussion

This study suggests that the mtDNAD-loop (CA)_n dinucleotide repeat polymorphism may not play a significant role in breast cancer etiology. Women with multiple alleles (heteroplasmy) of the mtDNA D-loop (CA)_n repeat, however, exhibited poorer survival compared with those carrying a single allele of the (CA)_n repeat, and the association with survival seemed to be independent of other clinical prognostic factors such as cancer stage or ER/PR status.

Endogenous estrogen plays a critical role in the pathogenesis of breast cancer (20, 21), and mitochondria are an important early target of estrogen action. Studies of other genes have suggested that the number of CA repeats in the promoter region is inversely correlated with transcription activity, with an up to 5-fold decrease in activity depending on the number of repeats (22). Variations in the length of the CA dinucleotide repeat may affect the transcription of mtDNAcoding genes, since the mtDNAD-loop acts as a promoter for both the heavy and light strands of mtDNA. Given that mitochondrial transcription is enhanced by estrogen treatment (20, 23), estrogen-induced mitochondrial transcription is likely to participate in breast carcinogenesis. A longer (CA)_n repeat in the mtDNA D-loop may antagonize this effect by decreasing mitochondrial transcription activity, so women carrying alleles with longer (CA)_n repeat lengths would theoretically be less susceptible to oxidatively-generated DNA damage. Our findings, however, do not support this hypothesis.

The presence of multiple alleles (heteroplasmy) of the (CA)_n repeat may be an indicator of mitochondrial genome instability and mtDNA malfunction and thus may be associated with poorer breast cancer prognosis. This hypotheses is supported, in part, by the findings of a recent in vitro study suggesting that mitochondrial respiration deficiency leads to activation of the Akt survival pathway through nicotinamide adenine dinucleotide (NADH)-mediated inactivation of phosphatase and tensin homologue (PTEN), which contributes to increased survival and drug resistance of cancer cells (24, 25). Intriguingly, we found that women who carried multiple alleles of mtDNAD-loop (CA)_n had lower DFS rates compared with those carrying one mtDNA D-loop (CA)_n repeat allele. More studies are needed to better understand the association of this polymorphism with breast cancer prognosis and the biological mechanisms underlying its effects.

Previous studies of mtDNAD-loop variation have focused on SNPs or point mutations. Only a few studies have evaluated the association of germline mtDNA variation in the D-loop region with cancer. Recently, Bai et al reported that the T16519C polymorphism in the D-loop was associated with increased breast cancer risk (OR = 1.98; 95% CI, 1.25–3.12) in a small case-control study, although this finding was not replicated in another study (26). Several studies have investigated the association of somatic D-loop mutations with breast cancer. In a study conducted using samples from 19 breast cancer patients, 14 of 19 tumors (74%) displayed at least one somatic mtDNA mutation; 22 of the somatic mutations were in the D-loop region (27). In a study conducted in 15 breast cancer patients using cancer tissue samples and matched nipple aspirate fluid, it was found that the frequency of mtDNA mutation was higher in the D-loop region than in non D-loop (i.e., coding) regions (28). More recently, in a study of 60 Taiwanese breast cancer patients, 30% of breast cancers displayed somatic mutations in the mtDNAD-loop region (29). These findings suggest that instability of the mtDNAD-loop region may be involved in breast carcinogenesis. Studies that analyze both germ line and somatic mutations in the mtDNAD-loop region may provide additional insight as to the role of mtDNA variations in breast cancer risk and survival.

Strengths of this study include the population-based study design and high response rate, which minimized potential selection bias. The detailed exposure information collected in the study enabled an evaluation of gene-environment interactions. Information on cancer characteristics and treatment was obtained from the vast majority of patients, allowing an evaluation of the possible modifying effects of these factors. Additionally, Chinese women living in Shanghai are relatively homogeneous in ethnic background; over 98% are classified into a single ethnic group (Han Chinese). Thus, potential confounding by ethnicity is not a major concern for our study. There are a few limitations in this study. The frequencies of the (CA)_6–11 alleles were relatively low (5.85%) and only 10.3% of women had multiple alleles of the (CA)_n repeat in our study population, which may have limited the statistical power for stratified analyses. Given the sample size, power = 0.80, and α= 0.05, the smallest detectable ORs for this study would be 1.65, 1.46, and 1.34 for risk genotype with frequencies of 5%, 10%, and 20%, respectively. Additional studies are needed to confirm these findings.

In summary, our study suggests that the mtDNAD-loop (CA)_n repeat polymorphism may not be associated with breast cancer risk. Carrying multiple alleles (heteroplasmy) of this polymorphism, however, may be associated with poorer breast cancer survival. This is the first study to evaluate the association of the mtDNAD-loop (CA)_n repeat polymorphism with breast cancer risk and survival. The results need to be confirmed in other large-scale studies.

Abbreviations

CI

confidence interval

DFS

disease-free survival

D-loop

displacement loop

HR

hazard ratio

mtDNA

mitochondrial DNA

OR

odds ratio

OS

overall survival

ROS

reactive oxygen species

PTEN

phosphatase and tensin homologue