Transcriptomics profiling study of breast cancer from Kingdom of Saudi Arabia revealed altered expression of Adiponectin and Fatty Acid Binding Protein4: Is lipid metabolism associated with breast cancer?

Breast cancer (BC) is the most common cancer that affects females worldwide and is the second most frequent cause of cancer-related deaths among women in the United States [1]. According to the National Cancer Registry in the Kingdom of Saudi Arabia (KSA), BC was ranked as the most prevalent form of cancer among females, accounting for 25.1% of all newly diagnosed female cancers (5,205) in the year 2009 [2]. The ASR was 22.7/100,000 for the female population. While the median age at presentation is around 63 years in the United States and Western Europe, the median age at presentation in the KSA is 48 years [2]. The genetic variability between patients and tumors drive this clinically heterogeneous disease [3]. Much of the knowledge on the molecular genetics on BC and other cancers has been resulting from examining European and US patients. However, growing scientific knowledge suggests the likelihood of variability in molecular signature between cancers from patients of different ethnic groups. Clear differences in the patterns of p53 mutations in BC were found between Midwest US Caucasian, African- American, Austrian, and Japanese women [4, 5]. Besides, varying patterns and the occurrence of germline mutations in BRCA1 and BRCA2 between ethnic groups were also reported [6]. Recently, National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program show that the age-adjusted BC frequency among minority groups are significantly lower than those among white women [7] Increase in BC mortality was also found in African American women when compared to white women [8]. All these differences continue to be mostly unexplained [9]. Moreover, changes in occurrence and clinical characteristics linked with ethnicity or race has got only limited consideration [10]. The current lack of measures to aid in stratifying BC treatment indicates the necessity for a new methodology to formulate better prognostication and therapy prediction. Although the parameters such as histological grade, stage, and tumor size are accepted as prognostic markers for BC [11], about 50% of the patients with ER-positive cancer fail on tamoxifen treatment [12, 13].

Transcriptomics approaches that allow screening of thousands of genes in an experiment, has had a major impact on BC research over the past 10 years [14] Several groups have carried out gene expression profiling of BC and classified clinically distinct subclasses of tumors [14], hereditary BC [15], and treatment prediction [16]. Many microarray studies so far reported have led to the discovery of several genes associated with BC. However, most of the gene expression profiling studies has been carried out mainly in the Caucasian population, and inclusion of the non-Caucasian population has been minimal [17, 18]. Significant risk factors of BC in the western countries such as nulliparity, low parity, first time pregnancy at late age, having no history of breast feeding, etc., are usually not practiced in the Saudi society, yet BC incidence is still high among women from the KSA [19]. In addition, a study of Taiwanese population reported that women who had more than three full-term pregnancies; first full-term pregnancy at the age of below 30; and three or more years of breast feeding showed significantly reduced risk of BC [20].

These studies suggest that there is an important need to better understand the molecular mechanisms underlying BC from different populations. In view of the existence of genetic differences across regional or geographical locations, more data specific to indigenous populations are needed for comparative analysis of molecular changes in breast tumors among Saudi population. More comprehensive molecular and genomic analyses of breast cancers are important to understand the complexity and severity of the disease and to find ‘druggable’ molecular targets for the development effective treatments. In order to obtain relevant target genes associated with BC patients in the KSA, we aim to start comprehensive microarray-based gene expression profiling study of BC in the Kingdom. Transcriptomic analysis coupled with functional and pathway analysis could lead to new insight into biomarkers and signatures associated with the disease. Deregulated signaling pathways are thought to drive functional processes such as cell growth, cellular proliferation, and invasion of cancer cells. Thus, identifying such underlying driving changes will be vital for studies of tumor progression, for the identification of novel therapeutic targets.

To this end, this study aimed to survey the genes that are differentially regulated in forty five freshly frozen BC tissue samples compared to eight normal controls. The ethnic group of the cohort in this study provides unique dimension to BC research. We examined genome-wide gene expression in order to elucidate the molecular mechanisms underlying malignancy in breast tissues. Based on the differential expression signature, we found biological functions and pathways that are significantly altered in BC. Pathway-based clustering was apparent across the tissue samples. Pathway analysis revealed canonical pathways that have not been implicated in BC, previously. We also demonstrated additional utility of our BC signature by comparative analysis between the USA and the KSA samples reveals marked biological differences in signature. Furthermore, Connectivity Map (cmap) analysis offered hypotheses regarding potential treatments for differential regulation observed in the BC [21]. Thus, our finding might open up new avenues of BC research and assist discover possible new targets for BC treatment where ethnicity plays major role.

Breast cancer therapeutic strategies today are generally based on histopathological characterization, tumor size, grading, and axillary lymph node and receptor status [1]. Even so, patients when diagnosed with similar conditions and when treated with similar drug can go through extensive differences in the development and relapse of BC. Studies show that women, who had a full-term pregnancy at an age below 30; and who went through three full-term pregnancies, and three or more years of breast feeding, were significantly protected against BC [20]. However, BC incidence is high among women from the KSA. BC risk factors including nulliparity or low parity, first full term pregnancy at very late age, no breast feeding etc., are not common in the Saudi society [19]. These factors combined with the early onset of BC among women, in this ethnicity, prompted us to study the molecular mechanism underlying the malignancies.

In the present study, we identified transcriptomic signature in BC from the KSA that is associated with clinical and histological parameters. The ethnicity of this cohort renders the power to prospectively examine, the roles of lifestyle and genetic susceptibility in the onset of BC. The genome-wide expression analysis contains several overrepresented functional gene classes and has substantial overlaps with transcriptomic signatures of metastatic human BC. The commonalities found between different populations in terms of increased cell cycle regulation and DNA integrity checkpoints were observed. A similar transcriptomic signature for many of these genes has been reported for human BC before [22–26].

Genes associated with lipid metabolism and small molecule biochemistry was significantly downregulated in the BC tissues. The role of these genes on the tumorigenesis of BC has been reported previously [27–29]. Disruption of lipid metabolism has been implicated in tumorigenesis in several studies, especially in cancer of breast [30, 31]. The differential regulation of lipid metabolism between normal and cancer subjects might reveal transformation in the metabolism of the cancer patient in this ethnicity [31]. Decreased ADIPOQ levels interrupt cellular signaling networks that are linked to cell survival, angiogenesis, proliferation, and cell-cycle regulation [32]. Adipose tissues secrete ADIPOQ and LEP-ADIPOQ axis has been well implicated in breast cancer tumorigenesis [32]. Levels of ADIPOQ have been inversely correlated with obesity. Interestingly, studies show that women with increased ADIPOQ concentrations possess 65% reduced risk for breast cancer [27]. In another study, it has been shown that patients with low ADIPOQ levels had a significantly increased likelihood of cancer recurrence.

Unsupervised clustering analysis also revealed the gene expression signatures to be associated with histologic grade, age and triple negative status. This signature could also be used for improving stratification of patients with BC in this population. Further pathway analysis of differentially regulated genes provides novel hypotheses underlying metastatic progression of BC. In addition, cmap analysis can formulate novel therapeutics for BC, starting from a cellular approach to investigating the effects of the compound determined by cmap. It may be possible to develop stratified therapy by combining both standard chemotherapy and pathway-specific therapies.

ADIPOQ, an adipocyte secreted protein, was found to one of the most downregulated gene in our analysis and it’s lower expression levels have been shown to be associated with obesity, insulin resistance, and type 2 diabetes mellitus [33]. It has also been shown to be reduced in pre-menopausal women with endometrial cancer, which is tightly linked with obesity and insulin resistance. A case control study found significantly inverse relationship of serum ADIPOQ with BC [34].

Comparison of BC data from USA with KSA further revealed distinct expression levels for LEP and other genes involved in lipid metabolism [35]. In our analysis, glycerolipid metabolism was found to be a highly significantly altered canonical pathway. However, comparative analysis of these findings is different from the recent transcriptomic profiles of BC from the USA data, as glycerolipid metabolism was ranked 41^st in the USA data [35]. In the KSA samples, glycerolipid metabolism was one the most overrepresented pathways, whereas, PPARα / RXRα activation pathway is the most disrupted one in the USA data.

To comprehend population differences in the severity of BC, studies linking both genetic and environmental factors and examining cases from different populations are important. Researchers have been investigating issues surrounding lifestyle, circadian rhythm, obesity, and adverse exposures in an attempt to identify susceptibility factors in initiating BC. Since diet and obesity contribute to changes in circulating hormone levels, they both play a key role in the development of BC. Understanding the underlying genetics and complex interactions of lifestyle, diet, and other known risk factors will remain a key area of research.

The prevalence of obesity among Saudi patients adults was 55.56% in comparison to 33.8% in USA adults, however combined obese and overweight percentage were almost same with 68.0% in western and 73 % in Saudi BC patients [36]. Obesity is primarily characterized by excess fat storage, adipocyte mass, and increase in certain types of lipids. Normally healthy women have 14–28% of fat mass in the body, but it may increase up to 60–70% in morbidly obese individuals [37]. Earlier white adipose tissue function was assigned for purely an energy storage tissue, however, recent finding has uncovered the endocrine and metabolic properties that has led to several mechanisms implicated in how obesity drives cancer prevalence and cancer deaths. Many possible mechanisms have been proposed to explain the increased risk of breast cancer associated with obesity such as increased lipids and lipid signaling, inflammatory responses, insulin resistance, adipokines, altered immune responses, and oxidative stress. Study had shown that cancer cells can access lipids from neighboring adipocyte stores and can directly use these transferred lipids as an energy source, which in turn promotes tumor growth [38]. Thus, it is critical to understand the physiological impact of obesity on breast cancer development and progression. In recent years, obesity has been identified as a significant modifiable increased risk factor for breast cancers among postmenopausal women but is unrelated or inversely related to risk among premenopausal women not in pre-menopausal women [39–43]. A study examined obesity and mortality from cancer in 495,477 US women, and found that obese women had double the death rate (relative risk, 2.12) than women with normal BMI (25 or less) [44].

Fat tissue of obese women are most important source of estrogen after ovary, secreting higher level of estrogen potentially leading to more rapid growth of estrogen-responsive breast tumors [45]. Fat cells may also promote inflammation linked cancer through tumor growth regulators and tumor necrotic factors. Obesity often cause insulin resistance, which may promote the development of breast cancer [46, 47]. Adipose tissue secrete TNF-α leading to inflamation promoting cancer. Fat cells produce hormones, called adipokines, that may stimulate or inhibit cell growth [48]. For example, adiponectin, a down regulated gene in present study, has been reported to be inhibitor for proliferetion and negative regulator of angiogenesis [37, 49, 50]. Further a significant inverse correlation between serum adiponectin levels and poor-prognosis breast cancer were confirmed. Thus, low adiponectin in obese individuals could increase risk of developing tumors with aggressive angiogenesis [34, 51].

This study described, to our knowledge, the first genome-wide profiling of BC from Saudi ethnic females. Our analysis reveals appropriate biological relevance and a number of molecular pathways that may serve as targets for novel therapeutics. Our finding of 73.3% obese vs 26.7% normal weight clearly suggest that obesity increases the risk of breast cancer occurrence. Further studies are needed to determine the relationship between lipid metabolism dysregulation and the mechanisms underlying BC. This study may open new avenues of experimental strategies for further examination in larger cohorts of all subtypes, correlating ethnicity to molecular signature and life-style.

The authors declare that they have no potential conflicts of interest.