Smoking can cause a variety of tumors and non-tumor diseases. It is estimated that 40% of human cancers are smoking-related which are responsible for more than 5 million preventable deaths worldwide per year (Ambrose and Barua, 2004). Many chemicals in tobacco are toxic. It has been reported that there are more than 69 harmful components, including 10 kinds of polycyclic aromatic hydrocarbons, 10 kinds of nitrosamines, eight kinds of heterocyclic amines, and six kinds of heterocyclic hydrocarbons generated when smoking (Benowitz, 2003). Smoking can impair natural killer cell function in the innate immune system, thereby weakening the function of tumor surveillance and inhibiting atypical cell growth (Yasuma et al., 2021). Like most environmental exposures, it does not directly cause certain diseases. On the contrary, it can increase the susceptibility to chronic diseases, such as coronary atherosclerosis and metabolic disorders. The pathogenesis caused by cigarette smoking includes inflammation and immune suppression, reactive oxidative species generation, and apoptosis (Yoon et al., 2021). However, the exact mechanism leading to smoking-related diseases and deaths remains largely unknown. Recently, epigenetic regulation has become the focus of research, which can effectively explain many genomic effects that cannot be explained by protein-coding genes (Nakamura et al., 2021; J. L.; Zhang et al., 2019).

MicroRNA is a single-stranded RNA with no protein-coding function, which regulates gene expression after transcription in numerous pathogenic signal transduction cascades with very little variation among species (Xu et al., 2017). Schembri et al. recently identified 28 miRNAs in smokers. Compared with non-smokers, these miRNAs are differentially expressed in the tracheal epithelium of smokers. Most of these miRNAs (82%) are down-regulated (Zhang et al., 2019). The changes in miRNA expression caused by smoking are thought to regulate the inflammation, generation of reactive oxidative species, xenobiotic metabolism, and carcinogenesis (Schembri et al., 2009). A recent study further showed that smoking caused 126 miRNAs in lung tissue to be down-regulated by at least 2 times, and 24 miRNAs were down-regulated by more than 3 times. These miRNA expression changes are related to the up-regulation of 107 genes, which regulate the expression of protein-coding genes related to stress, cell cycle regulation, and vascular formation (Willinger et al., 2017; Xu et al., 2018).

Recently, it was found that alterations of DNA methylation could be possible pathogenesis that mediate CS-induced diseases (Buck et al., 2021). It has been reported that, for current smokers, differential CpG island methylation occurred in almost one-third of human transcriptome (Bakulski et al., 2019). In addition, maternal smoking exposure has a great impact on the DNA methylation of infants. Suter et al. compared 36 matched placental samples with gene expression arrays and methylation arrays. 438 deferentially expressed genes showed a significant correlation with aberrant methylation. Several of them belong to characteristic pathways of oxidative phosphorylation, which indicated smokers have increased oxidative damage markers. These CpGs sites can be used as potential biomarkers to predict smoking-related disorders (Andersen et al., 2021). DNA methylation is mainly catalyzed by DNA methyltransferase (DNMT). Cigarette smoking can up-regulate DNMT1 expression, thus promoting the methylation process. The researchers revealed that compared with non-smokers, the expression of DNMT1 in the lung tissue of smokers was significantly higher (Lai et al., 2017).

So far, innumerable studies have reported abnormal DNA methylation and aberrant miRNA expression between smokers and non-smoking controls (Nozoe et al., 2002; Stroud et al., 2016; Sun et al., 2015). However, a comprehensive regulatory network on smoking-related microRNA and DNA methylation changes has not been established yet. Therefore, a thorough analysis of the regulatory network is needed to depict the correlation between genetic and epigenetic changes. In this study, the mRNA expression profile microarray data, miRNA expression microarray, and DNA methylation microarray were systematically analyzed to determine the core genes and pathways that lead to smoking-related damage through epigenetic regulation.

Promoter sequence methylation and changes in miRNA expression play an important role in smoking-induced diseases by altering gene transcription activities (Lamadrid-Romero et al., 2018). As a result, DNA methylation patterns and particular miRNAs can be regarded as instrumental biomarkers to evaluate smoking-related pathological conditions (L. Wang et al., 2018). In the current study, the data of miRNA microarray, DNA methylation microarray, and mRNA microarray were systematically analyzed, as shown in Figures 1, 2, and the differential analysis between the smoking group and the non-smoking group samples was compared. Pivotal genes and pathways responsible for key events in epigenetic changes regulated by microRNA binding and DNA methylation have been identified in the present study.

107 up-regulated genes have been screened out via the strategy of overlapping DEGs and DEM target genes, as plotted in Figure 2E. The molecular function enriched was mainly relevant to ubiquitination, indicating that the regulation of ubiquitin may play an important role in the pathogenesis caused by smoking. KEGG analysis revealed enriched pathways including amyotrophic lateral sclerosis, TGF-beta signaling pathway, autophagy, mTOR signaling pathway, and EGFR tyrosine kinase inhibitor resistance pathway, as shown in Figure 3A. Transforming growth factor-β (transforming growth factor-β, TGF-β) is a multi-functional cytokine that plays a dual role of anti-inflammatory and pro-inflammatory, and can act on the body’s immune response, inflammation, tissue repair, and embryonic development (Grunwald et al., 2021; Kim and Christian, 2021; Qiu et al., 2021; Wang et al., 2021; Xu et al., 2021). Studies have found that TGF-β is involved in the development of periodontal inflammation. Smoking promotes the development of chronic periodontitis by increasing the expression levels of IL-17 and TGF-β (C. Wang et al., 2021; Woods et al., 2021; Xu et al., 2021). Mechanistic Target of Rapamycin (mTOR) is an atypical serine/threonine-protein kinase, belonging to the phosphoinositide 3-kinase (PI3K)- related kinase family. The mTOR signal pathway can be activated by various external factors such as growth factors, stress, oxygen, and amino acid, and is involved in many physiological processes, such as liposome synthesis, energy metabolism, and autophagy. The inducible knockout of mTOR in mouse airway epithelial cells and alveolar type II epithelial cells will exacerbate smoking-induced inflammation, autophagy, necrosis, and apoptosis, which will lead to airway inflammation and emphysema (Liu et al., 2021; Wu et al., 2021; Ye et al., 2021). MircoRNA-mRNA regulatory network in Figure 3C demonstrated that KPNA6 and TOR1AIP2 were targeted by four miRNAs. SMAD4, BCL2L11 were targeted by three miRNA. Smad4 is a key molecule in the TGFbeta pathway. The phosphorylated Smad2 and Smad3 binding with Smad4 can form a molecular compound to achieve TGF-β1 signal transduction into the cell nucleus, thus initiating the biological effect (Dai et al., 2021; Zhang and Qiu, 2021). However, the functions of the other three genes in smoking-related pathological changes are poorly understood.

A total of 65 lower-expressed genes were filtered by taking the intersection of DEGs and DEM target genes, as illustrated in Figure 2F. It was found via enrichment that proteins encoded were mostly expressed around nerve synapses and mainly involved in biological processes related to catabolism and apoptosis. P53 signaling pathway, prostate cancer, and viral carcinogenesis pathway enriched by KEGG database suggested that these genes may play an important part in the pathogenesis of smoking-related cancers, as shown in Figure 3B. Figure 3D demonstrated that more than eight genes were targeted by hsa-miR-16-5p, hsa-miR-6127, hsa-miR-6124, and hsa-miR-25-3p, respectively. Sachin Kumar et al. found that hsa-miR-25-3p was significantly down-regulated in the serum of patients with non-small cell lung cancer, and the expression was significantly correlated with the stage of cancer (Kumar et al., 2020). He et al. found that hsa-miR-16 can inhibit the proliferation of nasopharyngeal carcinoma cells by inhibiting the activation of PI3K/AKT and MAPK signaling pathways (He and Huang, 2017). Therefore, the down-regulation of miR-16 in smokers may be related to carcinogenesis caused by smoking. However, the role of hsa-miR-6127 and hsa-miR-6124 play in cigarette smoking-related pathology has not been reported.

As we all know, lncRNA and cirRNA, as epigenetic regulatory molecules, are widely found in the regulation of gene expression (Gu et al., 2021; Li et al., 2021). miRNA induces gene silencing and down-regulates gene expression by binding to mRNA. However, the upstream molecules circRNA and lncRNA can influence the binding of miRNA to downstream target gene mRNA by competing for binding with miRNA response elements, thereby up-regulating downstream target gene expression. This interaction is called the ceRNA network (Li et al., 2021). It was reported that CSE inhibits insulin production by upregulating TXNIP via MALAT1-mediated downregulation of miR-17 causing reduced β-cells function (Sun et al., 2018). Ma et al. found that cigarette smoke extract (CSE) caused elevated epithelial-mesenchymal transition (EMT) and the increases of circ0061052 in HBE cells. By binding miR-515-5p competitively to regulate the expression of FoxC1/Snail, circ0061052 was involved in the epithelial-mesenchymal transition of epithelial cells during cigarette smoke-induced airway remodeling (Ma et al., 2020). Luo et al. (2020) reported the role of long noncoding RNA MEG3/miR-378/GRB2 axis in neuronal autophagy and neurological functional impairment in ischemic stroke. GRB2 is one of the hub genes in our study and its binding to miR-378a-3p has been predicted, as illustrated in Figure 6D. As a result, ceRNA regulatory network of GRB2 may also play an important part in CSE induced pathogensis. In addition, the ceRNA networks of hub genes were predicted and constructed providing research clues for further exploration, as plotted in Figures 3E, 6D‒E.

It was showed that 324 up-regulated genes with DNA hypomethylation were mainly enriched in neurodegeneration, amyotrophic lateral sclerosis, Alzheimer’s disease, and Huntington disease in terms of the KEGG pathway, as shown in Figure 4A. Therefore, these genes may play an important role in the neurological pathogenesis caused by cigarette smoking. It has been frequently reported that smoking may increase the risk of various neuropsychiatric diseases, such as cognitive disorder, psychopathy, depression, and anxiety. Inflammation and oxidative stress may play a crucial role in the pathogenesis of smoking-induced nervous injury (Liu et al., 2020; Na et al., 2014). After the application of the MOCDE plugin, 10 hub genes have been screened out, among which UBE2N is the gene with the highest degree of connectivity. UBE2N gene which encodes the E2 ubiquitin-conjugating enzyme responsible for the repair of DNA after replication plays an important role in ubiquitin-mediated protein degradation. Modification of proteins with ubiquitin is an important cellular mechanism for targeting abnormal or short-lived proteins for degradation (Dikshit et al., 2018; Zhu Q. et al., 2020). The up-regulation of UBE2N may be regarded as a defense mechanism against the smoking-induced abnormality of protein.

There were 204 low expressed genes with hypermethylation. GO and KEGG enrichment analysis suggested that they were mainly enriched in cardiovascular system functions, including adrenergic signaling in cardiomyocytes, platelet activation, and regulation of actin filament-based process, as shown in Figure 4C. Platelet activation is commonly seen in the blood of smokers. Platelets isolated from smokers exhibit spontaneous aggregation (Nardin et al., 2020). After exposure to smoker serum, platelets isolated from non-smokers exhibit a high degree of aggregation (Pamukcu et al., 2011). Smoking may reduce the availability of platelet-derived NO and reduce the sensitivity of platelets to exogenous NO, leading to increased activation and adhesion (Korniluk et al., 2019). Smoking can also lead to impaired vasodilation. Several studies have shown that both active and passive smoking are related to decreased vasodilation. For humans, exposure to cigarette smoke impairs endothelial dependence vasodilation (EDV) in the large vascular bed (such as coronary and brachial arteries) and the microvascular bed (Tousoulis et al., 2012). The enrichment results indicated that the Adrenergic signaling pathway may play a part in smoking-induced vascular spasm.

It is worth noting that miRNA and DNA methylation may synergistically regulate the abnormal expression of certain genes in smokers. Fourty three genes such as GRB2, UBE2N, TRMT2A, AMBRA1, CAMKK1, and ZNF516 showed enhanced expression accompanied by DNA hypomethylation and reduced miRNA expression. Similarly, 14 genes including FBXL20, CLN8, PDZD8, and NAV1 were down-regulated with the same dual regulation pattern. KEGG enrichment analysis showed that these genes were involved in the EGFR signaling pathway, insulin pathway, focal adhesion kinase-mediated signaling events, mTOR signaling pathway, and internalization of ErbB1, as plotted in Figure 5A. EEF2 and CNBP were highly expressed in the hematopoietic system and vasculature, as shown in Figure 5C. Among the down-regulated genes plotted in Figure 5D, GFRA1 showed low expression in the hematopoietic system and vasculature. It is suggested that these three genes may play an important role in the changes of blood vessels and the hematopoietic system (including monocyte and osteoclastogenesis) caused by cigarette smoking.

Since there are no explicit drugs to improve or reverse the pathophysiological conditions caused by smoking, online databases were used to help predict some drugs. At present, CMap is a useful tool for exploring new drugs and reusing existing drugs. Its effectiveness has been confirmed by many studies (Gao et al., 2019). From the CMap database, 10 chemical substances were identified, including Gly-His-Lys, wortmannin, tetracaine, diethylstilbestrol, and calcium pantothenate, as listed in Table 1. They may have significant potential therapeutic effects for smoking-caused pathological conditions. Many studies have proved the important role of peptidergic mechanisms in regulating various body functions. Gly-His-Lys (Bobyntsev et al., 2015; Sever’yanova Lcapital and Dolgintsev, 2017) reveals the versatility of this effect. It has a regulatory effect on cell growth and differentiation, promotes wound healing by stimulating fibroblasts to synthesize collagen, promotes liver repair and liver cell regeneration, stimulates hair growth, and has an anti-inflammatory effect on damaged tissues. It can also act on the nervous system and function as an anti-anxiety drug (Lassila et al., 1989; Kukowska and Dzierzbicka, 2014; Kukowska et al., 2015).

After topological structure analysis of protein interaction of all overlapping dual-regulated genes, the top 4 pivotal genes turned out to be HSPA4, GRB2, PRKCA, and BCL2L1. The BCL2L1 encoded protein belongs to the BCL-2 protein family and acts as an anti- or pro-apoptotic regulator involved in a variety of cellular activities (Dai et al., 2019). It is located in the outer mitochondrial membrane and has been shown to regulate the opening of the outer mitochondrial membrane channel (VDAC) which plays an important role in apoptosis. The protein encoded by GRB2 binds epidermal growth factor receptor and contains one SH2 domain and two SH3 domains. Its two SH3 domains take part in the formation of complexes with the proline-rich regions of other proteins, and its SH2 domain binds to tyrosine phosphorylation sequences and participates in signal transduction pathways (Tari and Lopez-Berestein, 2001). The protein kinase C (PKC) encoded by the PRKCA gene is a family of serine and threonine-specific protein kinases that can be activated by calcium and second messenger diglycerides (Huang et al., 2020). Members of the PKC family phosphorylate multiple protein targets and are known to participate in multiple cell signaling pathways. It has been reported that this kinase plays a role in many different cellular processes, such as cell adhesion, cell transformation, cell cycle checkpoints, and cell volume control. Knockout studies in mice indicate that this kinase may be the basic regulator of myocardial contractility and Ca(2+) influx in cardiomyocytes (Li et al., 2017). Heat shock proteins promote cell survival by supporting protein folding, stabilization and the unfolding of denatured proteins. Heat shock protein 70 (HSP70) is a 70kd stress-inducing protein whose intracellular expression can increase the tolerance of cardiomyocytes to ischemia/reperfusion injury (Fernandez-Fernandez and Valpuesta, 2018). It also has been reported that extracellular HSP70 is a modulator of inflammatory response. The enhanced expression of HSPA4 encoding HSP70 was demonstrated in the pressure zone of experimental tooth movement, indicating that the intracellular defense system was activated to ensure cell survival (Fontaine et al., 2016; Zhang et al., 2018).

The study revealed a series of deferentially expressed genes that are associated with epigenetic variations in miRNA expression and DNA methylation in smokers. By overlapping DEGs and DEM-targeted genes, 107 low miRNA-targeted up-regulated genes and 65 high-miRNA down-regulated genes were obtained, which were enriched in autophagy signaling pathways and protein ubiquitination pathways, respectively. In addition, 324 low-methylation-high-expression genes and 204 high-methylation-low-expression genes are respectively related to the degeneration of the nervous system and the function of the cardiovascular system. Interestingly, 43 genes were up-regulated under the dual regulation of reduced miRNA and hypomethylation, while 14 genes were down-regulated under the dual regulation of increased miRNA and hypermethylation. Ten chemicals have been identified as putative therapeutic agents for pathological conditions caused by smoking. In addition, HSPA4, GRB2, PRKCA, and BCL2L1 were regarded as hub genes and could play a fundamental role in pathological conditions caused by smoking and may be used as the biomarkers for precise diagnosis and targets for future therapies of smoking-related diseases.