Background
Currently, cigarette smoking is an established risk factor for the development of clear cell renal cell carcinoma (ccRCC) [
1]. Indeed, authors of a meta-analysis involving 26 epidemiologic studies spanning 37 years concluded that the risk of ccRCC among ever smokers is approximately 40% higher compared to lifetime never smokers [
2]. From a population-based perspective, previous investigators have suggested that cigarette smoking alone accounts for approximately 20-25% of the ccRCCs diagnosed in the U.S. [
3,
4]. While smoking is an established risk factor for ccRCC, what remain unclear are the specific somatic molecular alterations that underlie this well-reported association. Identification of specific alterations at the cellular level that link smoking to ccRCC development has the potential to further solidify a causal association, advance our understanding of the etiology of this disease and possibly extend even further into more focused measures of early detection and prevention.
To address the need to better understand the molecular underpinnings of smoking-related ccRCC, we sought to identify candidate genes that are differentially expressed in ccRCC tumors that develop in smokers compared to non-smokers. Thus, we employed the Affymetrix U133 Plus 2.0 platform to compare somatic gene expression profiles between patient-matched ccRCC and normal kidney tissues from patients with and without a history of smoking, controlling for obesity status (the other primary risk factor for ccRCC [
4,
5]). Although other risk factors have been reported in the literature, smoking and obesity are the only epidemiological risk factors that have been consistently validated as increasing risk of ccRCC. Following our microarray-based discovery efforts, we then validated our top candidate genes by employing RT-PCR on an independent set of ccRCC and patient-matched normal kidney tissue samples from smokers and non-smokers. Using this multi-stage design, we report that ANKS1B is a smoking-related alteration in ccRCC.
Discussion
Based on the current literature, there is little question regarding the role of cigarette smoking in the etiology of ccRCC; however, what remains unclear is exactly how smoking acts within the body (specifically within the kidney itself) to increase a person’s risk of developing ccRCC. Related to this, tobacco smoke contains a vast number of chemicals, with about 50 of those chemicals being classified as human carcinogens [
9]. Inhaled chemical carcinogens from cigarette smoke, like any other chemical that enters the human body, are subject to extensive metabolism. The majority of this metabolism is directed toward deactivation of the particular chemical and eventual excretion. However, an important fraction of the metabolic process results in the conversion of the ingested compound to highly reactive metabolite(s) that possess the ability to bind to intercellular components (i.e. DNA) and induce changes in their structure; changes that may or may not lead to the transformation of normal cells to tumor cells. Given that the kidney is the main filtration organ of the blood and is known to locally produce enzymes involved in xenobiotic metabolism, it is theorized to be at high exposure to any smoking-related carcinogen. In fact, researchers have reported that the urine of smokers has increased mutagenic activity compared to non-smokers [
10]. While this primary theory of how smoking increases the risk of ccRCC does exist, little progress has been made towards illuminating the actual molecular target(s) that are altered by smoking carcinogens in the development of ccRCC.
ANKS1B, Ankyrin repeat and sterile alpha motif domain-containing protein 1B, is a tyrosine kinase signal transduction gene that is primarily expressed in the brain and testis. Here, we demonstrate for the first time that expression of the ANKS1B gene is associated with smoking-related ccRCC development. ANKS1B is involved in apoptosis and thus has the potential to play a key role in cancer development [
11]. From our observational data, we show that ANKS1B is up regulated in smokers relative to non-smokers in normal kidney tissue; however, it is down regulated in smokers relative to non-smokers in ccRCC tumor tissue. Thus, ANKS1B expression in smokers is down regulated in the tumor tissue in comparison to the patient-matched normal kidney tissue and this down regulation is potentially a key event that supports ccRCC development. Interestingly, Lin et al. [
12] recently evaluated the association of germline SNPs within apoptotic pathway genes with lung cancer risk – in which smoking is also a major risk factor – and identified 2 SNPs in ANKS1B (rs1549102 and rs11110099) that had statistically significant associations. What remains unclear is whether these SNPs are also found in lung cancer tissues and whether they are functionally associated with expression or activity of the ANKS1B protein. That notwithstanding, these results from another smoking-related cancer further suggest a possible role for ANKS1B to be a smoking-related molecular alteration in cancer and underscore the potential for these results to advance the knowledge of ccRCC etiology and prevention. Indeed, in addition to advancing our understanding of the pathways involved in smoking-related ccRCC, alterations in ANKS1B could also potentially be used for early detection and prevention in smokers. That being said, we acknowledge that our findings must first be validated at the protein level. Moreover, there is a need to link alterations in ANKS1B to smoking-related ccRCC in a more robust epidemiologic study design. Particularly, using a larger case–control study or a large prospective-cohort study where it would be feasible to adjust for additional reported risk factors, to study the dose–response relationship of smoking with ANKS1B and lastly, to study the association of smoking with molecularly-defined ccRCC subtypes.
We used a discovery-based approach to identify smoking-specific molecular alterations associated with ccRCC development that can be followed up in more focused investigations. Having said that, the key limitations of our approach include our focus on expression changes at the RNA level (compared to protein expression or alterations at DNA level) and our overall limited generalizability (tertiary referral center, >95% of patients are Caucasian). We acknowledge that our cohort has differences between the ccRCC tumors in the smokers and non-smokers that were studied. First, smokers were more likely to be male than non-smokers. Additionally, smokers in our study were more likely to have later stage disease compared to non-smokers in stage 1; however, the stage distributions were similar between smokers and non-smokers in stages 2-3. Since ANKS1B showed similar results in all 3 stages it is likely not simply a marker associated with later-stage disease. With those limitations in mind, the specific strengths of our design include the use of only clear cell RCC subtype (the most common histologic subtype), exclusion of late stage and high grade tumors in the discovery stages (to focus on events linked to early ccRCC development), use of packyears > 20 years to define smokers (those at theorized high exposure to smoking carcinogens) and access to data on obesity in order to account for the other primary risk factor for ccRCC.
Our study was designed specifically to identify smoking-related molecular alterations that are associated with ccRCC development. As a result, we evaluated patient-matched tumor and normal kidney samples from both smokers and non-smokers. Thus, our potential targets of interest were those that had a statistically significant smoking status-by-tissue type interaction. It is worth noting that if cancer is not of interest and future investigators are interested in simply identifying genes that are associated with only smoking, our publicly available data could be further explored to identify genes with a significant smoking main effect.
Conclusion
In summary, we demonstrated that ANKS1B expression is associated with smoking-related ccRCC. Interestingly, ANKS1B was recently shown to be associated with cancer by Lin et al. [
12], where they showed that 2 SNPs in ANKS1B are associated with risk of lung cancer. Here, we showed that ANKS1B is under expressed in ccRCC tumor tissue in comparison to patient-matched normal. Given the role of ANKS1B as an enhancer of apoptosis, down regulation of this gene could be involved in increasing the risk of ccRCC development.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JEP participated in the conception and design, assisted in the statistical analyses and drafted the manuscript. DJS performed the statistical analyses and helped draft the manuscript. BMB participated in the conception and design, performed the statistical analyses and revised the manuscript. RWJ participated in the interpretation and revised the manuscript. JCC reviewed the pathology of all subjects and revised the manuscript. ASP conceived the study, participated in the interpretation, drafting and final approval of the manuscript. All authors read and approved the final manuscript.