Differentially Expressed Genes Analysis in the Human Small Airway Epithelium of Healthy Smokers Shows Potential Risks of Disease Caused by Oxidative Stress and Inflammation and the Potentiality of Astaxanthin as an Anti-Inflammatory Agent

Cigarette smoke (CS) was known for its effect of increasing oxidative stress that could trigger tissue injury and endothelial dysfunction mediated by free radicals and reactive oxygen species (ROS). ROS itself is a key signaling molecule that plays a role in the development of inflammatory disorders. Nuclear factor erythroid2 related factor2 (Nrf2) is the main regulator of antioxidant cellular response to cell and tissue-destroying components caused by CS. Nrf2 protein that is significantly activated in the smokers' small airway epithelium is followed by a series of gene expression changes in the same cells. This study aims to observe differentially expressed genes (DEGs) in the human small airway epithelium of smokers compared to genes whose expression changes due to astaxanthin (AST) treatment, an antioxidant compound that can modulate Nrf2. Gene expression data that was stored in the GEO browser (GSE 11952) was analyzed using GEO2R to search for DEG among smokers and nonsmokers subject. DEG was further compared to those genes whose expression changes due to astaxanthin treatment (AST) that were obtained from the Comparative Toxicogenomics Database (CTD; https://ctdbase.org/). DEG (p < 0.05) analysis result shows that there are 23 genes whose expression regulation is reversed compared to gene expression due to AST treatment. The gene function annotations of the 23 DEGs showed the involvement of some of these genes in chemical and oxidative stress, reactive oxygen species (ROS), and apoptotic signaling pathways. All of the genes were involved/associated with chronic bronchitis, adenocarcinoma of the lung, non-small-cell lung carcinoma, carcinoma, small cell lung carcinoma, type 2 diabetes mellitus, emphysema, ischemic stroke, lung diseases, and inflammation. Thus, AST treatment for smokers could potentially decrease the development of ROS and oxidative stress that leads to inflammation and health risks associated with smoking.


Introduction
Oxidative stress occurs due to an imbalance between the increased production of free radicals and decreased antioxidant capacity [1]. Under physiological conditions, oxidative stress will trigger an increase in the expression of endogenous antioxidant genes and cytoprotective proteins to prevent or limit tissue damage. Tis process is mediated by nuclear factor erythroid2 related factor2 (Nrf2) activity which then activates transcription way for antioxidant gene and enzyme detoxifcation [1,2]. Tus, impaired activation of Nrf2 will cause a decrease in antioxidant capacity.
Cigarette smoke (CS) component that is dissolved in water is known to directly increase oxidative stress that could trigger tissue injury. Smoking tobacco has also been associated with vascular endothelium dysfunction through causative methods depending on the dose. Tis is mainly related to tobacco content of reactive oxygen species (ROS), nicotine, and infammation driven by oxidative stress [3]. In particular, chronic CS exposure to respiratory tract tissue causes an increase in radical concentration, volatile compound (particularly oxygen species and reactive nitrogen), and CS condensate deposition, which will trigger a pleiotropic adaptive response, aimed at restoring tissue homeostasis [4]. Chronic exposure to CS generally encounters a cellular defense system characterized by activation of Nrf2. Nrf2 as the main regulator of antioxidant cellular response is proven to regulate the frst line of defense against CSinduced cell and tissue-damaging components. Tis is indicated by the higher expression of Nrf2 in PBMC in moderate smokers compared to nonsmokers (p < 0.01). An increase in Nrf2 was not found in heavy smokers who possess a high level of nuclear transcription factor (NF-kB) and C-reactive protein (CRP) (p > 0.01) [5]. Tis indicates disruption of the Nrf2 role in heavy smokers with an infammatory problem. Nrf2 genetic efect also afects smokers' health status. Tis is indicated by the signifcant interaction between genotype rs6726395 with accompanied by the decrease of forced expiratory volume in one second (FEV1) (p � 0.011) [6]. Te Haplotype rs2001350T/ rs6726395A/rs1962142A/rs2364722A/rs6721961T is also associated with a lower annual decline in FEV1 (p � 0.004) [6].
Astaxanthin (AST) is a food xanthophyll that is often found in sea organisms, and because of its unique molecular feature, it possesses good antioxidant activity. More evidence has suggested AST's protective role to counter several diseases where oxidative stress and infammation occur continuously. AST is known to modulate Nrf2 binding to antioxidant response elements (AREs) in the promoter region of most cytoprotective or detoxifying enzymes [7]. Recent studies have also shown that AST modulates the NF-B signaling network by increasing infammation and oxidative stress in various experimental models [8,9]. Several studies have shown that the anti-aging efect, as well as attenuation of oxidative stress and infammation of AST, is carried out through Nrf2 activation and NF-kB inhibition [10][11][12].
Diferentially Expressed Gene (DEG) is important to understand the biological diference between a healthy and ill condition. Identifcation of genes involved in disease is an important tool for revealing the molecular mechanisms of disease development. In pharmaceutical and clinical studies, DEG also plays an important role to choose biomarker candidates, therapeutic targets, and genetic signatures for diagnosis [13]. In this study, an analysis of changes in gene expression patterns due to smoking was carried out in the Gene Expression Omnibus (GEO) database [14,15] which was compared with changes in profle expression genes due to AST treatment obtained from Comparative Toxicogenomics Database (CTD; https://ctdbase.org/) [16]. By comparing these two research data, it is intriguing to know what genes have the potential expression to be afected by AST, so it is hoped that it could further explain the potential of AST as a candidate for antioxidant supplements in terms of its mechanism of action in reducing the health efects that can appear on smokers.

Methods
Tis study gathers data from Gene Expression Omnibus (GEO) database, a study conducted by Hübner et al. [17]. Te inclusion criteria of a healthy nonsmoker and smoker referred to those study. Healthy nonsmoker was people with normal physical examination, lung function, and chest Xray, with smoking-related blood and urine within the nonsmoker range. Te criteria for a healthy smoker were current smoking history, followed by normal physical examination, lung function, chest X-ray, and smoking-related blood and urine parameters consistent with current smokers. In the study, the age of the subjects was not distinguished [17]. Human small airway epithelium samples was obtained using fber-optic bronchoscopy of 38 healthy nonsmokers and 45 healthy smokers, and Nrf2-associated gene expression was assessed using the Afymetrix HG-U133 Plus 2.0 microarray. Compared to healthy nonsmokers, it was found that the Nrf2 protein was signifcantly activated in the human small airway epithelium of healthy smokers and localized in the nucleus (p < 0.05). Te research gene expression data stored in the GEO browser (GSE 11952) was then analyzed using GEO2R to look for DEG between smokers and nonsmokers subjects. Furthermore, DEGs of smoker's vs nonsmokers were compared with genes that changed expression due to AST treatment obtained from the CTD, with the target of fnding genes that were opposite in expression between the 2 datasets.
Genes with opposite expressions were then made into protein networks and clustered using STRING (stringdb.org) [18]. In these genes, gene function annotations were made using the GO biological process [19,20]. Relationships between genes and smoking-related diseases obtained from CTD. At this stage, it was expected to know the role of AST on changes in biological processes that occur in smokers and the diseases that can accompany them based on gene expression profles.

Results
Te results of DEG analysis of research data from Hübner et al. [17] stored in the GEO browser (GSE 11952), showed that there were 4912 signifcantly diferentially expressed genes (DEGs) in the human small airway epithelium of smokers compared to nonsmokers (p < 0.05). If the 4192 DEGs was compared with genes or proteins whose expression changed because of AST administration obtained from CTD, then the results are as shown in Table 1. Table 1 shows 23 Nrf2-related genes/proteins that expression regulation was opposite between smokers (against nonsmokers) with the efect of AST treatment. Te efect of AST in infuencing the gene expression/protein could happen directly or indirectly. In the case of it happening indirectly, the AST slows down the reaction that infuences a particular gene expression/protein. For instance, slowing down LEPR mutant reaction. LEP is known to be associated with leptin receptor (LEPR) and took part in activating several intracellular signaling channels [21]. Te increase of LEP in the lungs and serum is associated with potentially worsening or hastening the development of lung diseases, including acute lung injury (ALI), acute respiratory distress syndrome, chronic obstructive pulmonary disease (COPD), airway remodeling associated with asthma, and lung cancer [21]. In addition, the presence of polymorphism LEPR is known to show a statistically signifcant diference between lung cancer patients and controls (p � 0.007) [22]. LEPR mutant is also known to cause kidney [23] and bone marrow fbrosis [24].
Te relationship between these 23 genes and compounds found in environmental tobacco smoke (ETS) can be seen in Table 2. ETS is smoke that originated from burning tobacco products and smoke exhaled by smokers [25]. ETS consist of 40 biologically and toxicologically active compounds according to Hofmann's list [26]. Tree compounds from Hofmann List are produced in milligrams per cigarette (tar, nicotine, and CO), while the remaining are in nanograms or micrograms level per cigarette [26,27]. In Table 2, it could be seen that 6 out of 23 genes that undergo changes in profle expression in human small airway epithelium on smoker subjects are associated with the compounds from ETS according to the Hofmann List. In addition to explaining how smoking can afect the expression profle of these genes, this can also clarify the potential benefts of giving AST to smokers.
From the 23 mentioned genes, the protein networks were made using STRING (string-db.org) as could be seen in Figure 1. Tere are 4 clusters with the red node as the central cluster (C1). Te central cluster consists of 7 genes/proteins: SOD1, IDH1, TKT, PRDX1, GPX3 SKAP2, and BECN1 with SOD1 as central nodes. If Nrf2 (NFE2L2) was administrated into the networks, it could be seen that these genes were in the central cluster (C1). In Table 3, the annotations of the 7 (seven) genes/proteins which contain 10 (ten) groups of gene annotations (gene ontology and GO biological process) can be seen based on the smallest adjusted p value [28]. In Table 3, the genes/proteins can be seen to be involved in chemical and oxidative stress, reactive oxygen species (ROS), and apoptotic signaling pathway. Smoking is known to induce oxidative stress, as well as activate infammatory response pathways, which trigger a cascade of events in which ROS production is an early but indispensable step [29]. CS is also known to induce in vivo epithelial cell apoptosis, however, fbrotic changes occur only after a viral exacerbation [29,30].
Gene function annotation (GO biological process) of other clusters (C2, C3, and C4) can also be seen in Table 3. C2 cluster is related to the regulation of extrinsic apoptotic signaling pathways and the regulation of immune response. C3 cluster is related to transcription process regulation, programmed cell death, signaling receptor activity, and deacetylation reaction. While the C4 cluster is related to aerobic and cellular respiration, as well as the electron transport process in mitochondria. Based on the relationship between the central cluster and other clusters, it can be seen how chemical oxidative stress caused by ROS due to smoking activity could afect the transcription process regulation, signaling process related to apoptosis and receptor activity, as well as electron transfer process and other cellular processes.
If the 23 genes that change expression due to smoking were associated with diseases caused by smoking, it was known that all of genes were involved/associated with chronic bronchitis, adenocarcinoma of the lung, nonsmall-cell lung carcinoma, carcinoma, small cell lung carcinoma, type 2 diabetes mellitus, emphysema, ischemic stroke, lung diseases, and infammation (Table 4). While the ones related to pulmonary heart disease are known to be as many as 19 genes. Te interesting thing is that the 23 genes are also related to the infammation disease category. It appears that this evidence suggests that the association between ROS and oxidative stress induced by smoking and smoking-related disease may be mediated by the infammatory process. On the other hand, the administration of AST, thus has the potential to reduce the risk of the development of these diseases in smokers.

Discussion
Smoking activity is a major factor in various diseases, including immune-mediated infammation disease. Te concept of chronic or prolonged ROS production is central to the development of infammatory diseases [31]. On tobacco, ROS production is mainly contributed by nicotine, the main component in tobacco. A low concentration of nicotine (0.1 μM) could induce ROS to about 35%, however, a signifcant increase in the amount of ROS could be observed at 1 and 10 μM nicotine concentrations of 54% and 80%, respectively [32]. Aside from nicotine, ROS development is also stimulated by various agents such as pollutants in ETS such as heavy metals (lead, nickel, mercury, arsenic, cadmium, chromium, and cobalt), or other organic compounds such as hydroquinone, acrylonitrile, acrolein, formaldehyde, acetaldehyde, benzene, dan benzo(a)pyrene. Reactive oxygen species (ROS) is a key signaling molecule that plays an important role in the development of infammatory disorders. Te increase in ROS generation by neutrophils polymorphonuclear (PMN) at the sites of infammation can for example lead to endothelial dysfunction and tissue injury [31]. However, nicotine-induced neutrophil activation by nicotine is also known to be ROS-independent [33].
International Journal of Infammation Table  1: DEGs in Nrf2-associated human small airway epithelium of smoker vs nonsmokers compared with changes in gene/protein expression due to AST administration taken from CTD [16]. International Journal of Infammation Some of the associations between genes that changed expression in smokers with infammation, endothelial dysfunction, and tissue injury can be explained as follows: in Table 2, it is known that an increase in histone deacetylase2 (HDAC2) expression, which is also caused by nicotine, hydroquinone, and benzo(a)pyrene compound, happens in smokers. Te increase in expression also happens on secreted phosphoprotein1 (SPP1), transketolase (TKT), cytochrome b-245 beta chain (CYBB), and peroxiredoxin 1 (PRDX1) in smoker subjects compared to nonsmokers (Table 1). In atherosclerosis, overexpression of HDAC2 in endothelial cells under proatherogenic conditions and     International Journal of Infammation oxidative injury suppresses the expression of Arginase2 (ARG2), which further reduces the expression of endothelial nitric oxide synthase (eNOS) [34]. Endothelial dysfunction is known to be caused by a decrease in eNOS expression. In chronic diabetic foot ulcer (DFU) an increase of HDAC2 expression also happens where dysfunctional endothelial progenitor cells (EPCs) plays a major role in inhibiting vascular complication in DFU patient [35]. Inhibition of HDAC2 is known to prevent infammatory disorders and ROS production in EPCs with high glucose levels [35]. SPP1 is highly expressed after stimulation of oxidized low-density lipoprotein (oxLDL) and plays a role in causing infammation of human coronary artery endothelial cells (HCAECs) [36]. High TKT expression is also associated with advanced tumor stage and TKT inhibitors promote apoptosis of lung adenocarcinoma cells and cell cycle blockade [37]. CYBB, also known as NADPH-oxidase (NOX2) is known to be involved in angiotensin II-induced hypertension and endothelial dysfunction, as well as abundantly expressed in the endothelium [38]. PRDX1 is also signifcantly higher in stroke patients compared to control. PRDX1 level is also higher on blood samples taken 3 and 6 hours after the stroke attack compared to the control [39]. Te 23 DEGs generated from the analysis of gene expression data in the GEO browser (GSE 11952) were the genes expressed in the human small airway epithelium of smokers vs nonsmokers, where the Nrf2 protein is also signifcantly activated and localized in the nucleus of the same cell [17]. Tis can also indicate how these genes are related to Nrf2. Furthermore, if we look at the protein network in Figure 1 where Nrf2 (NFE2L2) was in the central cluster (C1), there is a strong indication that the 23 DEGs produced are related to Nrf2 (NFE2L2). In mammals, Nrf2 has long been known to function as an evolutionarily conserved intracellular defense mechanism against oxidative stress. Nrf2 has been shown to contribute to the regulation of the heme oxygenase1 (HO-1) axis, which is a strong antiinfammatory target, and has shown a relationship with the expression of infammatory mediators in the NF-kB pathway and macrophage metabolism through the Nrf2/antioxidant response element (ARE) system [40]. Lungs are highly vulnerable to oxidative stress-inducing factors such as infection, allergen, and pollutant such as ETS. Oxidative stress that triggers Nrd2 activation has been shown in several human respiratory diseases such as asthma and chronic obstructive pulmonary disease (COPD), or pulmonary parenchyma-related diseases such as acute respiratory distress syndrome (ARDS) and lung fbrosis [41]. In this study, it has been shown (in Table 4) the association of the 23 DEGs with these diseases and other smoking-related diseases such as pulmonary heart disease, ischemic stroke, and type 2 diabetes mellitus (T2D).

International Journal of Infammation
Tis study shows that AST could also act as a very good candidate to improve diseases related to infammation [42]. AST is also known to increase Nrf2 and HO-1 expression in the lung, and suppress emphysema due to cigarette smoke in rats [43]. From the various previous explanations, it can be concluded that AST treatment in smokers has the potential to reduce the formation of ROS and the occurrence of oxidative stress that triggers infammation, as well as the accompanying diseases. Te potential for AST can then be confrmed through the next stage of research (e.g., clinical trials) including through observation of changes in gene expression biomarkers of the 23 DEGs.

Conclusion
From this study, we found that the 23 DEGs (smokers vs nonsmokers) in the human small airway epithelium were found to be inversely regulated by genes that changed expression due to AST treatment. Based on the GO biological process, some of these genes are known to be related to oxidative stress and ROS. AST has been confrmed to be efcacious in relieving chronic and acute infammation in a variety of diseases, including neurodegenerative disorders, diabetes, gastrointestinal diseases, kidney infammation, and skin and eye diseases.

Data Availability
Te data used to support the fndings of the study are available on the database mentioned in the manuscript, as it used secondary data from the Gene Omnibus Ontology (GEO) database and Comparative Toxicogenomics Database (CTD; https://ctdbase.org/).