INTRODUCTION

Breast cancer is the most common malignancy and the leading cause of cancer-related death among women worldwide¹. In Mexico, the mortality rate during 2012 was 14.76 deaths per 100,000 women². Genetic and other risk factors play a fundamental role in the onset of this disease. Diverse studies have suggested that obesity increases the risk of breast cancer in postmenopausal women and raises the incidence of metastatic breast tumors, resulting in higher rates of recurrence and increased mortality³. In Mexico, the latest National Health and Nutrition Survey demonstrated that the overall prevalence of obesity has risen from 21.5% in 1993 to 33.3% in 2016 being more pronounced in women (38.6%)⁴.

Mutations in mitochondrial DNA (mtDNA) have been found to play a role in the genesis of numerous cancers including breast cancer⁵-⁸. The mitochondrial genome is a 16.5-kilobase double-stranded circular DNA that contains no introns and includes 13 genes that encode proteins involved in aerobic respiration, 22 transfer RNAs, and 2 ribosomal RNAs. mtDNA is more vulnerable to mutations, due to high levels of reactive oxygen species (ROS) formed during oxidative phosphorylation, to a limited repair DNA system, and to the absence of protective histones⁹. Therefore, ROS are thought to be involved in the initiation, progression, and maintenance of tumorigenesis¹⁰.

Despite the current novel strategies for high-throughput screening of the mitochondrial genome such as the MitoChip used for investigating mtDNA mutations in various cancers¹¹, studies aimed at detecting mutations in breast cancer often only analyze specific regions of mtDNA and not the entire mitochondrial genome¹²,¹³. Because mitochondrial defects and oxidative stress have been related to obesity¹⁴, and breast cancer is more frequent and more aggressive in postmenopausal women with obesity, we investigated whether Mexican-Mestizo postmenopausal women with breast cancer and obesity presented different somatic mutations in the mtDNA when compared to women with normal body mass index (BMI), using the second generation of MitoChip for sequencing all the mitochondrial genome.

SUBJECTS AND METHODS

RESULTS

Three women had a normal BMI and three presented obesity; all patients were postmenopausal. One of the women with a normal BMI was the oldest in the group (82 years) and presented a locally advanced breast cancer (Stage IIB). In the group of women with obesity, one woman had a higher proliferation index (Ki67) of 75%. Regarding the immunohistochemical subtypes, hormonal receptor (+)/HER2 (−) was found in all cases. General characteristics are presented in table 1. After sequencing the complete mitochondrial genome and comparing it with that reported in the Revised Cambridge Reference Sequence (rCRS) of the Human mtDNA (www.mitomap.org/foswiki/bin/view/MITOMAP/MutationsSomatic) and mtDB - Human Mitochondrial Genome Database (www.mtdb.igp.uu.se/), we found 64 genetic variants, 34 (53.1%) were in the tumor tissue and were not present in the DNA of blood leukocytes (somatic mutations), and 30 (46.9%) were polymorphisms (Tables 2-4). We observed only two heteroplasmic mutations, both in the non-coding region (Table 2). Of all somatic variants, 15 (44.1%) were in the non-coding region, 9 (60.0%) in patients with normal BMI, and 6 (40.0%) in patients with obesity (Table 2). On the other hand, we found 19 (55.9%) mutations in the tumor tissue that was not present in blood leukocytes DNA, in genes encoding mitochondrial proteins. Patients with normal BMI presented 14 (73.7%) mutations, 5 were synonymous, and the rest changed the amino acids (Table 3). Regarding the patients with obesity, we found only 5 (26.3%) mutations, being all non-synonymous changes (Table 3).

Polymorphisms were more frequent in patients with obesity, finding 21 variants in the coding region (66.7% of them were polymorphisms synonymous) and 5 in the D-loop region (Table 4). The four polymorphisms detected in patients with normal BMI were two in D-loop region and the other two in genes that encode mitochondrial proteins (Table 4). We also found 25 genetic variations that were present in both, tumor tissue and DNA of blood leukocytes, 60% of which had not been reported previously. 40% of these genetic variants were polymorphisms, and 80% were found in the non-coding region (data not shown).

DISCUSSION

Obesity is a major risk factor for breast cancer, and it has been suggested that adiposity increases the risk of recurrence and mortality in affected women, worsening quality of life³. Mitochondrial dysfunction has been implicated in the pathophysiology of both, obesity and breast cancer¹⁴,²⁰. Several somatic mutations have been reported in mtDNA in women with breast cancer. However, these studies have been conducted in women of Caucasian and Asian ethnicity⁶,⁷,²¹,²²; for that reason, we considered it important to study somatic mitochondrial mutations in patients with breast cancer and different BMI who belonged to a distinct ethnic group. The latter is relevant since different lineages of mtDNA are associated with diverse global population groups²³. This raises the possibility that mtDNA variations could contribute to differences observed in disease prevalence reported among different ethnic groups. Likewise, it is important to consider that Mexican-Mestizo population is constituted by a mixture of Europeans and Africans with native Indian individuals²⁴. The proportion of genes in this ethnic group has been reported as 56% Amerindian, 40% Caucasian, and 4% African genes²⁵.

There are few studies, in breast cancer patients, where the whole mtDNA is sequenced in search of somatic mutations. Furthermore, these studies do not consider the presence of obesity as a risk factor. In several studies, it has been reported that the D-loop region of mtDNA is the most frequent site of mutations in breast cancer⁵-⁷,²¹,²². Parrella et al.⁵ examined by direct sequence analysis, the mtDNA of 16 women with primary breast cancer. Despite having analyzed only 84% of the mtDNA, they identified 12 somatic mutations and several polymorphisms; the highest frequencies of mutations were detected in the D-loop region. Tan et al.⁶, studying the complete sequence of the mtDNA of breast cancer tissue, found that 74% of the analyzed tumors had somatic mutations, and most of these mutations were also found in the hypervariable D-loop region. These authors mentioned that some of these somatic mutations had been described in other cancers (i.e., ovarian carcinomas, esophageal cancer, colorectal, and gastric tumors). Likewise, Zhu et al.⁷ sequenced the entire mtDNA of 15 breast cancer tissue samples detecting several polymorphisms and 45 somatic mutations in 93% of the tumors. Most of these mutations were homoplasmic and the majority of them, as well as the polymorphisms, were also located within the D-loop region. However, unlike the studies cited above, we found more somatic mutations in those genes that encode proteins involved in aerobic respiration, then somatic mutations located in the non-coding region. Moreover, we did not find the mononucleotide repeat, between nucleotides 303 and 315 (D310), which is a conserved sequence of mtDNA and has been considered a hotspot region⁵,⁷.

Differences between our results and those previously reported might be due to the screening approaches used in the studies mentioned above. In most of the previous studies, only a portion of the mitochondrial genome was screened. Likewise, some of the screening techniques used, such as temperature gradient gel electrophoresis, may miss some mutations. On the other hand, we used an Affymetrix MitoChip, and as has been proposed by Xie et al.²⁶, this tool has a high sensitivity and accuracy needed for reliable, whole mitochondrial genome sequencing. Moreover, we observed that women with breast cancer and a normal BMI had more somatic mutations in genes that encode mitochondrial proteins than women with obesity. These mutations were present in NADH-dehydrogenase (ND) subunit 1 (ND1), subunit 2 (ND2), subunit 3 (ND3), subunit 4 (ND4), subunit 5 (ND5), ATP synthase 6 (ATP6), and cytochrome oxidase subunit I genes; most of these changes were non-synonymous. Interestingly, patient 3 showed 9 somatic mutations, where 8 of them changed amino acids. Probably in this patient, advanced age (82 years) and the presence of a more advanced cancer (IIB stage) contributed to the existence of a higher number of mutations.

In relation to those somatic mutations observed in the non-coding region in women with obesity, only the mutation at position 664 of the 12S ribosomal RNA (present in all three patients with obesity) had been reported previously in human ovarian carcinoma²⁷. In addition, of the five non-synonymous somatic mutations present in genes that encode mitochondrial proteins, found in our women with obesity, four of them were present in patient 6 and almost all were in ND2. It has been hypothesized that ND2 is involved in proton translocation across the inner mitochondrial membrane, contributing to pH cell regulation. It has been demonstrated that the hydrophobic subunit of complex I is highly conserved throughout evolution, from bacteria to mice and to humans, indicating a central physiological role²⁸.

Noteworthy, the patients with breast cancer and obesity presented a significant number of somatic polymorphisms. Most of these polymorphisms were in genes that encode mitochondrial proteins; some of them have been described in thyroid, prostate, pancreatic, and breast cancer (www.mitomap.org/foswiki/bin/view/MITOMAP/MutationsSomatic). This indicates that mtDNA is highly susceptible to mutation, probably by oxidative stress²⁹. It has been proposed that these mtDNA variants, together with somatic mtDNA mutations, could create an environment with a slight increase in levels of ROS inside the mitochondria³⁰; from the consequent signal transduction, cell proliferation may be augmented.

The main limitation of this study was that we only examined six patients with breast cancer; despite this limitation, the strengths of the study are that we analyzed the whole sequence of the mtDNA, and we compared women with normal BMI versus women with obesity.

To the best of our knowledge, this is the first report where the complete sequence of mtDNA in Mexican-Mestizo postmenopausal women with breast cancer and different BMI is performed, using human microarray data from the MitoChip. However, we did not find a higher frequency of mitochondrial somatic mutations in postmenopausal women with breast cancer and obesity. The latter may be due to the small number of women with obesity or normal BMI studied, so we could consider this a pilot study that needs to be confirmed in future studies with larger sample sizes.

Furthermore, due to the limited number of women with obesity and breast cancer studied, we could not demonstrate whether mitochondrial mutations in obesity are a risk factor for developing this type of cancer. We found that women with obesity had a greater number of polymorphisms which could be relevant for the presence of this cancer since the synonymous or synonymous codons usage could affect other aspects of protein biogenesis or regulate the expression level, which could be linked to human diseases³¹. Besides, our results diverge from those reported in Caucasian and Asian populations, possibly due to ethnic differences. We believe that further studies are needed, where a larger number of patients with breast cancer are included. Furthermore, extensive biochemical and molecular studies will be necessary to determine the effects on energy metabolism in malignant cells.