Heavy alcohol drinking downregulates ALDH2 gene expression but heavy smoking up-regulates SOD2 gene expression in head and neck squamous cell carcinoma

Head and neck squamous cell carcinoma (HNSCC), the fifth most common type of cancer, accounts for 600,000 newly diagnosed cancer cases with 300,000 deaths worldwide every year [1]. Despite novel surgical and pharmaceutical treatment methods developed for HNSCC, its survival rate has not been improved over the last two decades [2]. It is widely accepted that HNSCC is strongly associated with alcohol drinking and tobacco smoking [3]. It has been estimated that 75% of HNSCC cases in the USA are caused by alcohol drinking and tobacco smoking [4].

In general, the metabolism of ethanol, the major component of alcoholic beverages, consists of two steps [5]. First, alcohol dehydrogenase (ADH) enzyme converts ethanol to acetaldehyde. Second, aldehyde dehydrogenase (ALDH) converts acetaldehyde to acetate. Acetaldehyde is a toxic material that has been established as a strong carcinogen [6]. The second oxidation step is largely dependent on ALDH2 enzyme encoded by ALDH2 gene.

As for smoking, tobacco smoke produces a variety of free radicals, reactive oxygen species (ROS), and reactive nitrogen species (RNS). They will produce superoxide, hydrogen peroxide, and nitric oxide which can cause oxidative damage and oxidative stress to cells [7]. Superoxide dismutases (SODs) are enzymes that can catalyze the dismutation of superoxide into oxygen or hydrogen peroxide [8]. There are three distinctive SODs: copper/zinc SOD (Cu/ZnSOD or SOD1), manganese SOD (MnSOD or SOD2), and extracellular SOD (ecSOD or SOD3, and Cu/ZnSOD) [9]. Among them, SOD2 is localized at the mitochondria. It acts as an endogenous antioxidant enzyme that limits the oxidative burden [10]. SOD2 can convert superoxide into hydrogen peroxide and oxygen in the mitochondria. It plays a major role in protecting cells from oxidative damage [11].

A number of studies have demonstrated the correlation between ALDH2 gene or SOD2 gene polymorphisms and susceptibility to HNSCC [12‐17]. However, the expression pattern of these two genes in HNSCC patients in terms of the amount of alcohol drinking and tobacco smoking has not been reported.

Therefore, the objectives of this study were to analyze the expression pattern of ALDH2 and SOD2 genes using available microarray data (GSE65858 and GSE39368) [18] and determine the relationship between the expression level of these two genes and clinical parameters such as alcohol drinking, tobacco smoking, primary site of HNSCC, and human papilloma virus (HPV) state.

In this study, we evaluated the association between expression levels of alcohol metabolism genes or ROS metabolism genes and the amount of alcohol drinking or tobacco smoking. We also determined the association of gene expression patterns with clinical variables such as primary site, HPV state, stage, and survival result. Among alcohol metabolism genes, ALDH2 gene expression pattern divided the microarray data into two groups: ALDH2 high-expression group (N = 152) and ALDH2 low-expression group (N = 118) (Fig. 1a). Interestingly, heavy alcohol consumption group showed low expression levels of ALDH2 with poor overall survival results (Table 1 and Fig. 3a). Microarray data (N = 270) were also divided into two groups by SOD2 expression pattern (Fig. 1b). However, in contrast to the relationship between ALDH2 expression and alcohol consumption amount, heavy smoking group showed high SOD2 gene expression (Table 1). Interestingly, there was no correlation between SOD2 expression level and survival results (Fig. 3b).

Alcohol drinking and tobacco smoking are significant risk factors for head and neck cancer [3]. In general, alcohol metabolism involves two steps. First, alcohol is converted into acetaldehyde by ADH. Second, acetaldehyde is oxidized into nontoxic acetate by ALDH [5]. Among all classes of ADH and ALDH isoenzymes, ADH1B, ADH1C, and ALDH2 are the main ethanol-metabolizing enzymes [21, 22]. Therefore, we used these three genes to analyze the expression pattern of alcohol metabolism genes. As for ROS metabolism genes, SODs are enzymes that catalyze the removal of superoxide free radicals [8]. SODs are present in all aerobic living cells. This is probably because O₂ ⁻ is a common product of oxygen metabolic reactions. In mammals, there are three distinctive SODs: SOD1 (Cu/ZnSOD), SOD2 (MnSOD), and SOD3 (extracellular SOD). SOD1 is the major intracellular form of SODs. It is primarily cytosolic [23]. In contrast, SOD2 is exclusively localized in the mitochondrial matrix (MM) [23], while SOD3 is the secreted form mainly associated with the extracellular matrix of different tissues [24]. In this study, we used these three superoxide dismutases genes to analyze the expression patterns of ROS metabolism genes. Among these three alcohol-related genes, only ALDH2 gene showed distinguished expression pattern. Similarly, among the three smoking-related genes, only SOD2 gene showed distinguished expression subtype.

Several studies have demonstrated the correlation between ALDH2 gene or SOD2 gene polymorphisms and susceptibility to cancer development. Zhao et al. have reported that ALDH2 Glu504Lys SNP is a potential candidate genetic risk factor for a variety of chronic diseases such as cardiovascular disease, cancer, and late-onset Alzheimer’s disease [12]. Chung et al. have suggested that minor alleles of ADH1B (rs1229984) and ALDH2 (rs671) not only are associated with the risk of upper aerodigestive tract cancer, but also can potentiate the carcinogenic effects of alcohol [13]. Hidaka et al. have reported that ALDH2 AA allele carriers who drink > 150 g/week have an increased risk of cancer development compared to GG genotype carriers who drink 0 to < 150 g/week [15]. As for SOD2, Han et al. have reported that polymorphism of SOD2 T5482C might be closely associated with increased susceptibility to the development and differentiation of gastric cancer in Korean population [16]. All these SNP studies have emphasized the important roles of ALDH2 and SOD2 in alcohol drinking and smoking-induced cancer development. However, up to date, no study has reported the expression levels of these genes, especially using microarray mRNA-Seq data. In this study, we used mRNA-seq data and evaluated the relationship between expression levels of ALDH2 and SOD2 genes and the amount of alcohol drinking or smoking consumption.

In statistical analysis, the amount of alcohol consumption, primary site, and HPV state showed statistically significant difference between ALDH2 high-expression group and ALDH2 low-expression group (Table 1). Patients with heavy alcohol consumption showed low ALDH2 expression levels, while patients with no or light alcohol consumption showed high ALDH2 expression levels (Table 1). There are two possible hypotheses to explain such results. First, cumulative acetaldehyde itself might have downregulated the expression of ALDH2 gene directly or indirectly using other upstream or downstream molecules. Second, the liver functions of heavy alcoholics usually are not so good. Their anabolic functions such as mRNA synthesis and protein synthesis are usually decreased. The finding that patients in the heavy smoking group showed high SOD2 expression in this study supports the second hypothesis.

In this study, we used Ingenuity Pathway Analysis (IPA) to investigate up-regulated and downregulated pathways in ALDH2 high-expression group and SOD2 high-expression group. Intrinsic prothrombin activation pathway is up-regulated in ALDH2 low-expression group (z-score 2.236, Table 2). Garcia-Banuelos et al. have investigated the relationship between ALDH2 SNP and liver cirrhosis [25]. Heavy alcohol drinking can cause liver damage and activate intrinsic prothrombin activation pathway, especially in the ALDH2 low-expression group. Chao et al. have reported the interplay between superoxide dismutase genes and PPAR signaling [26]. In our study, the SOD2 high-expression group showed decreased PPAR signaling pathway (Table 2).

This study has some limitations. First, this study is a retrospective study. Second, no data about liver function state of patients enrolled in this study were available. Comparing liver functions between heavy alcoholics and those with no or light alcohol consumption is needed as we mentioned above. A prospective and functional study is needed to provide more precise information about the role and prognostic aspect of ALDH2 and SOD2 genes in head and neck cancer.