The biologic effects of cigarette smoke on cancer cells


  • Samantha L. Sobus BS,

    1. Department of Molecular Pharmacology and Cancer Therapeutics, Roswell Park Cancer Institute, Buffalo, New York
    2. Department of Radiation Oncology, Medical University of South Carolina, Charleston, South Carolina
    Search for more papers by this author
  • Graham W. Warren MD, PhD

    Corresponding author
    1. Department of Radiation Oncology, Medical University of South Carolina, Charleston, South Carolina
    2. Department of Cell and Molecular Pharmacology and Experimental Therapeutics, Medical University of South Carolina, Charleston, South Carolina
    • Corresponding author: Graham Warren MD, PhD, Department of Radiation Oncology, Medical University of South Carolina, 169 Ashley Avenue MSC 318, Charleston, SC 29425; Fax: (843) 876-2297;

    Search for more papers by this author


Smoking is one of the largest preventable risk factors for developing cancer, and continued smoking by cancer patients is associated with increased toxicity, recurrence, risk of second primary cancer, and mortality. Cigarette smoke (CS) contains thousands of chemicals, including many known carcinogens. The carcinogenic effects of CS are well established, but relatively little work has been done to evaluate the effects of CS on cancer cells. In this review of the literature, the authors demonstrate that CS induces a more malignant tumor phenotype by increasing proliferation, migration, invasion, and angiogenesis and by activating prosurvival cellular pathways. Significant work is needed to understand the biologic effect of CS on cancer biology, including the development of model systems and the identification of critical biologic mediators of CS-induced changes in cancer cell physiology. Cancer 2014;120:3617–3626. © 2014 American Cancer Society.


Cigarette smoking and other forms of tobacco use generally are considered the largest preventable causes of cancer as well as heart disease, pulmonary disease, and many other diseases.[1] There are 7000 compounds in cigarette smoke (CS), and the primary addictive substance nicotine leads most individuals to start smoking before age 18 years.[2] Overwhelming evidence demonstrates that smoking causes a wide spectrum of cancers, and significant work has been done to understand the fundamental biologic processes associated with tobacco-induced carcinogenesis.[3] However, there are proportionately far fewer data on the clinical effects of smoking in cancer patients. The 2014 Surgeon General's Report presents a convincing case associating continued smoking by cancer patients with increased cancer recurrence, treatment toxicity, risk of second primary cancer, and mortality.[1] The systemic effects of smoking on cancer treatment suggest that common biologic processes may be involved, but there are relatively few data on the effects of smoking on cancer biology. The objectives of this review were to briefly introduce the different types of CS and review the known effects of CS on cancer cells and to discuss the effects on cell proliferation, angiogenesis, migration, immune modulation, and cell death.

The Composition of Cigarette Smoke

CS is a complex mixture of aerosolized chemicals and tobacco toxicants, which can be classified according to the location of origin and composition of the tobacco leaf. The burning of cigarettes generates three primary types of smoke: 1) mainstream CS (MCS), which is produced from the cigarette butt and drawn directly into the smokers lungs (active smoking); 2) sidestream CS (SCS), which is produced from the continued cigarette smoldering between puffs (passive smoking); and 3) environmental tobacco smoke (ETS), which is a combination of MCS and SCS generated within the vicinity of the smoker.[4] The way in which a cigarette is smoked effects the ratio of MCS and SCS generated. There are standardized smoke conditions, as defined by the Federal Trade Commission (FTC) and the International Organization for Standardization (ISO), defined as a 35-mL puff volume with 2-second duration at a 1-minute puff frequency, but these conditions may not accurately represent current cigarette use behavior.[5]

The different components present in CS are distributed between the particulate and gaseous phases. For example, the aldehydes (formaldehyde, acrolein, and acetaldehyde) are present primarily during the gaseous phase and are associated with chronic pulmonary disease and lung toxicology; whereas the polycyclic aromatic hydrocarbons (PAHs), tobacco-specific nitrosamines, and metals (arsenic, cadmium, chromium) are present during the particulate phase and are most frequently associated with cancer.[6] In addition, the tar or particulate phase is rich in long-lived radicals, including semiquinone, which forms hydroxyl radicals and hydrogen peroxide when it reacts with the superoxide anion. Short-lived oxidants, such as superoxide anion and nitric acid, are more predominant during the gaseous phase, but they quickly react to form a highly reactive peroxynitrite.[7] It is believed that, under biologic conditions, a large amount of redox-cycling occurs within the aqueous portion of CS at the lung-lining fluid over a long period of time.[7] Studies also have demonstrated that, when combined, CS and ethanol work synergistically to induce colon tumors, because ethanol solubilizes the liposoluble CS components.[4] Furthermore, studies have indicated that heating condensates enhance their mutagenicity, and it is theorized that this occurs either by favoring the oxidation of promutagenic moieties into mutagenic components or by disruption of the inhibitors of mutagenicity present within CS condensates.[4]

The FTC uses a uniform, machine-based test method for measuring tar and nicotine yield in cigarettes. In agreement with efforts in the 1960s to reduce exposure to tobacco toxicants, the FTC mandated that cigarette producers reduce the tar and nicotine content present in CS. This effort changed the composition of cigarettes and reduced the amount of tar and nicotine present in CS from 38 mg tar and 2.7 mg nicotine in 1954 to 12 mg tar and 0.95 mg nicotine.[3] However, such efforts have not actually been effective in reducing the exposure to cigarette toxicants, because smokers have altered their smoking habits by trying to compensate for reduced nicotine levels and achieve the same biologic high or effect. Such measures may include, but are not limited to, covering the ventilation holes at the base of the cigarette filter to deliver more free nicotine and inhaling deeper into the lungs. The net effect of changes in design and smoking habits have actually led to alterations in disease patterns associated with tobacco use. For example, low-tar filtered cigarettes that were inhaled more deeply led to an increased risk of developing lung cancer with parallel changes in the location and histology of lung cancer from centrally located squamous cell cancers to more peripherally located adenocarcinomas.[8]

There is no single standard method used to generate CS; and, for the purpose of this review, smoke-collection methods are defined as follows: 1) total particulate matter (TPM) measured in smoke collected on a Cambridge glass fiber filter and eluted into dimethyl sulfoxide (DMSO), 2) CS extract (CSE) is measured in smoke that is bubbled into a solution (ethanol, chloroform, media, or phosphate-buffered saline, as indicated), and 3) whole smoke (WS) is measured in smoke that is directly exposed to cells or animals using a smoking chamber. Many of the studies using CS try to examine the carcinogenesis of smoking related cancer. Frequently, these studies expose immortalized “normal” tissues to various CS preparations to better understand the early mutations and phenotypic changes that contribute to cancer initiation. Although such analyses are pivotal in tracking the carcinogenesis of smoking-related cancer formation, they provide limited information regarding the effect of tobacco smoke on cancer biology.

Cigarette Smoke and Cancer Biology

Although substantial work has been done on the carcinogenic effects of CS, very few studies have investigated the effect of CS exposure on altering the cellular functions of cancer cells. Even fewer studies have examined the effect of whole cigarette smoke exposure on a biologic system rather than characterizing the effect of a single component of CS. In fact, far more studies that reported on the tumor-promoting properties of nicotine, nitrosamines, and nicotinic acetylcholine receptor (nAChR) agonists have demonstrated that nAChR or activation of the β-adrenergic receptor (β-AR) can increase cellular proliferation, angiogenesis, migration, or invasion and decrease cell death after exposure to cytotoxic agents, such as chemotherapy or radiotherapy.[9-14] Several excellent reviews are cited on this topic; however, very little is known about the importance of these individual components in the overall scope of CS exposure on cancer cells, such as the role of specific nAChRs in modulating the tumor-promoting effects of CS. The objective of this review was to explicitly discuss the effect of CS on cancer cells; therefore, any detailed discussion of the biologic effects of specific CS components like nAChR agonists are beyond the scope of this review.

Another important but largely unstudied topic is the effect of direct CS exposure compared with indirect CS exposure. For example, in the aerodigestive tract, mucosal surfaces may be exposed to all or nearly all components present in CS. However, exposure to CS in other tissues relies on the systemic absorption, metabolism, and distribution of a portion of the chemicals present in CS. The effects of CS are further complicated by the generation of reactive oxygen species (ROS) and other effects that may affect cancer cell biology. Unfortunately, these potential effects are not well known, and future exposure models should consider defining these types of exposure.

Cell Proliferation

It has been demonstrated that the deregulation of normal cell proliferative pathways is central to cancer initiation and progression; however, the effect of CS on cell proliferative pathways in cancer cells remains unclear and needs to be further evaluated (Fig. 1). Hussain et al evaluated the effect of 10-day TPM exposure on A549 and Calu-6 lung cancer cells and observed that it increased the tumorigenicity of xenografts in nude mice.[15] More long-term (1-year) TPM exposure studies in A549 cells also demonstrated enhanced tumorigenicity.[16] Further analysis indicated that TPM decreased the formation of the SMA and MAD-related protein (SMAD) family member 3 (Smad3)/Smad4 transcription complex induced by transforming-growth factor-β (TGF-β).[16] This decrease in Smad3/Smad4 complex formation increased cell viability and was attributed to CS-induced reduction of Smad3 expression.[16] Thus, it was observed that chronic CS exposure inhibited TGF-β signaling and conferred a more malignant and tumorigenic phenotype.[16] Conversely, a study conducted by Tsuji et al provided evidence that CSE bubbled in DMSO induced a senescence phenotype in A549 cells.[17] Specifically, cells acquired a flat and enlarged appearance and exhibited other features characteristic of senescence, including increased senescence-associated β-galactosidase activity, increased lipofusion, increased p21CIP1/WAF1/Sdi1, and irreversible induction of growth arrest.[17] That study also demonstrated that the CSE-induced senescence was mediated in part by ROS formation.[17] Currently, the balance between proliferation and senescence is not well defined based on exposure to different forms or preparations of CS.

Figure 1.

The network of proteins involved in mediating cancer cell proliferation associated with cigarette smoke exposure is illustrated. ROS indicates reactive oxygen species; EGFR, epidermal growth factor receptor; β1-AR, β1-adrenergic receptor; Dkk-1, Dikkopf-1; Wnt, wingless; FZD, frizzled; SRC, v-src avian sarcoma viral oncogene homolog; Cox-2, cyclooxygenase-2; 5-Lox, 5-lipoxygenase-activating protein; LRP-6, low-density lipoprotein receptor-related protein 6; MEK, mitogen-activated protein kinase kinase; PI3K, phosphoinositide 3-kinase; PGE1, prostaglandin E1; LTB4, leukotriene B4; SirT1, sirtuin-1; Ezh1, enhancer of zeste1; Dvl-2, disheveled 2, JNK, c-Jun N-terminal kinase; ERK1/2, extracellular signal-regulated kinases 1 and 2; AKT, protein kinase B; AHR, aryl hydrocarbon receptor; VEGFR, vascular endothelial growth factor receptor; MMP-2, matrix metalloproteinase-2; MMP-9 matrix metalloproteinase-9; CYP1A1 cytochrome P450 enzyme 1A1; CYP181, cytochrome P450 enzyme 171; ADM, adrenomedullin; ABCG2, a xenobiotic pump that is up-regulated by aryl hydrocarbon receptor signaling; miR-487b, microRNA 487b; BMI1, BMI1 polycomb ring finger oncogene; MYC, v-myc myelocytomatosis viral oncogene homolog; KRAS, Kirsten rat sarcoma viral oncogene homolog; SUZ12, polycomb protein SUZ12; WNT5A, wingless-type MMTV integration site family member 5A.

Aryl Hydrocarbon Receptor

The aryl hydrocarbon receptor (AHR) is a basic helix-loop-helix transcription factor that is involved in maintaining a wide range of homeostatic functions, including, but not limited to, cell proliferation, gene transcription, cell motility and migration, and inflammation.[18] AHR activation results in the transcription of three types of detoxifying genes: 1) phase I drug-metabolizing cytochrome P450 (CYP) enzymes, including CYP1A1, CYP1A2, CYP1B1, and CYP2S1; 2) phase II enzymes such as uridine diphosphate (UDP)-glucuronosyl transferase-1AG6 and several glutathione-S-transferases; and 3) phase III transporters, including multidrug resistance-associated proteins and P-glycoproteins.[18] It has been demonstrated that dysregulation of AHR contributes to multiple aspects of cancer, including initiation, promotion, and progression.[18] The AHR functions as the primary mediator of xenobiotic metabolism; chemical carcinogens like those present in CS (including benzo[a]pyrene and PAH) act as ligands and directly bind to the receptor.[18] Specifically, exposure of A549 cells to CSE induced activation of the AHR, which then would bind specific DNA xenobiotic response elements and drive the expression of both CYP1A1 and adrenomedullin (ADM), a proto-oncogene that acts as a growth factor.[19] It also was demonstrated that ADM mediated CS-induced tumor growth in subcutaneous A549 tumors.[19] Studies performed by Uppstad et al indicated that TPM exposure induced both CYP1A1 and CYP1B1 in 10 additional lung cancer cell lines.[20] Exposure of lung and esophageal cancer cells to TPM also up-regulated the expression of ABCG2, a xenobiotic pump that is up-regulated by AHR signaling; and it was demonstrated that the inhibition of AHR partially abrogated TPM-induced increased ABCG2 expression.[21] This increase in ABCG2 expression results from activation of specificity protein 1 (Sp1) sites within the ABCG2 promoter; treatment with mithramycin, an Sp1 inhibitor, reduced the expression of TPM-induced ABCG2 expression and inhibited cell growth in vitro and in vivo.[21] TPM-induced ABCG2 expression increased a side population of Calu-6 and A549 cancer cells, suggesting an increase in the population of pluripotent tumor cells.[21] Furthermore, knockdown of TPM-induced ABCG2 expression decreased cell proliferation, clonogenicity, and migration, suggesting that ABCG2 may contribute to a more malignant, invasive phenotype.[21]

Nonconical Wnt Signaling Pathway

The wingless (Wnt) signaling pathway is involved in governing the maintenance, self-renewal, and differentiation of mammalian adult tissues.[22] It is also believed that Wnt signaling is involved in maintaining cancer stem cells, and it was recently demonstrated that different aspects of the pathway are activated in response to CS exposure. Hussain et al examined the effect of CS on epigenetic changes and observed that 10-day TPM exposure of A549 and Calu-6 lung cancer cells increased the tumorigenicity of xenografts in nude mice.[15] Moreover, TPM exposure induced the down-regulation of Dickkopf-1 (Dkk-1), a Wnt signaling inhibitor that is frequently silenced by methylation in cancer.[15] It also was observed that this Dkk-1 repression modulated nonconical Wnt signaling by increasing T-cell factor reporter activity, increasing cyclin D, and phosphorylating LDL receptor-related protein-6 (LRP-6), dishevelled-2 (Dvl-2), and c-Jun N-terminal kinase (JNK).[15] Hussain et al also observed that TPM exposure induced irreversible recruitment of the polycomb machinery to the Dkk-1 promoter and that knockdown of histone-lysine N-methyltransferase (EZH2) and Sirtuin1 (SirT1) abrogated this induced repression.[15] Expanding on these findings, Xi et al evaluated the effect of CS on altering the microRNA (miR) transcriptome in lung cancer cells and observed that TPM-induced repression of miR-487b up-regulated several target messenger RNAs (mRNAs) involved Wnt signaling.[23] Characterization of miR-487b indicated that it mediates cell signaling arrest and senescence in lung cancer cells by directly targeting the following mRNA sequences: 1) BMI1 polycomb ring finger oncogene (BMI1), 2) polycomb protein SUZ12 (SUZ12), 3) wingless-type MMTV integration site family member 5A (WNT5A), 4) v-myc myelocytomatosis viral oncogene homolog (MYC), and 5) Kirsten rat sarcoma viral oncogene homolog (KRAS).[23] Of these mRNA targets, BMI1 and SUZ12 encode for core components of the polycomb repressor complexes 1 and 2, respectively, WNT5A is a nonconical Wnt ligand, and MYC and KRAS are involved in cell proliferation.[23] Collectively, these studies provide evidence that CS-induced cell proliferation may be governed in part by epigenetic alterations in the Wnt signaling pathway; however, the mechanism by which this occurs and its relevance to the maintenance of cancer stem cells needs to be further evaluated.

The Cycloxygenase-2/5-Lipoxygenase–Activating Protein Pathway

Several studies have examined the link between CS-induced up-regulation of the arachidonic acid cascade and cancer progression. Li et al examined the role of cycloxygenase-2 (COX-2) and the β-adrenergic receptors in the pathogenesis of smoking-related esophageal squamous cell carcinomas and observed that exposure of EC109 cells to either a chloroform fraction of CS (CSE-C) or an ethanol fraction of CS (CSE-E) stimulated cell proliferation.[24] CSE-C also increased mRNA expression of β1-adrenergic and β2-adrenergic receptors and COX-2, whereas CSE-E induced increased expression of β1-adrenergic and β2-adrenergic receptors but did not alter COX-2 expression.[24] Furthermore, Li and colleagues observed that β1-adrenergic and β2-adrenergic receptor antagonists and COX-2 inhibitor eliminated CSE-C–induced cell proliferation but not CSE-E–induced cell proliferatoin.[24] Thus, their observations suggest that the proliferative action of chloroform extract in EC109 squamous esophageal cells is mediated through a β1-aderenergic and β2-adrenergic receptor-dependent and COX-2–dependent mechanism, whereas the pathway stimulated by the ethanol extract of CS needs to be further investigated.

It also was reported that treatment with WS promoted the formation of inflammation-associated adenomas in the colons of dextran sulfate sodium (DSS)-treated mice.[25] Specifically, the combination of DSS and WS exposure increased tumor incidence from 12.5% to 87.5%, and these tumors were characterized by increased vascularization and elevated expression of 5-lipoxygenase-activating protein (5-LOX), vascular endothelial growth factor (VEGF), and matrix metalloproteinase-2 (MMP-2).[25] In vitro studies performed in SW1116 colon cancer cells demonstrated that exposure to CSE-E similarly induced increased cell proliferation; however, unlike in EC109 cells, CSE-E also increased COX-2 expression in a dose-dependent manner.[26] SW1116 cells incubated with ethanol-CSE for 18 hours before subcutaneous implantation into BALB/c nude mice increased tumor growth compared with controls, and this effect was decreased by COX-2 inhibition.[26] The combination of ethanol and chloroform CSEs also stimulated the proliferation of SW1116 cells with increased expression of 5-LOX (a cell proliferation promoter) and its downstream product leukotriene B4 (LTB4), but it had no effect on COX-2 or prostaglandin E2 (PGE2) levels.[27, 28] Chromatin immunoprecipitation (ChIP) analysis revealed that the increase in 5-LOX expression and subsequent cell proliferation resulted from a loss of methylation in the CpG-dense region at nucleotides 13 through 121 of the 5-LOX promoter.[27] In addition, CSE increased the protein levels of the angiogenic signaling molecules VEGF, MMP-2, and MMP-9.[28] Pre-exposure of cells to CSE also induced a 3-fold increase in tumor xenografts, and tumors exhibited increased cell proliferation, decreased apoptosis, and increased levels of 5-LOX, LTB4, COX-2, and PGE2.[27] Inhibition of either 5-LOX or COX-2 reduced tumor size and decreased LTB4 and PGE2 levels, respectively.[27] Moreover, inhibition of 5-LOX partially blocked CSE-induced proliferation and reduced the induced VEGF, MMP-2, and MMP-9 expression; whereas inhibition of MMP-2 and MMP-9 reduced VEGF but did not alter 5-LOX expression in vitro.[28] Collectively, these studies suggest that COX-2 and 5-LOX play a role in mediating CS-induced cell proliferation and tumor growth.

The Epidermal Growth Factor Receptor Pathway

The epidermal growth factor receptor (EGFR) is a key component in mediating cell proliferation, survival, and differentiation during development.[29] The deregulation of EGFR has been intensively studied within the context of cancer biology, and there is evidence that EGFR signaling may be modulated by CS exposure. Preliminary examinations indicated that WS exposure of H292 mucoiepidermoid pulmonary carcinoma cells promoted cell proliferation and that pretreatment with AG1478, an EGFR kinase inhibitor, reduced this effect.[30] Several studies also reported an increase in EGFR phosphorylation after CS exposure, thus suggesting involvement of the EGFR pathway in mediating CS-stimulated cell proliferation in lung cancer cells. There are two schools of thought regarding how EGFR activation occurs, and below we examine both ligand-dependent and ligand-independent EGFR activation.

Ligand-dependent EGFR activation

Canonical EGFR activation involves the binding of several identified EGFR ligands, including epidermal growth factor (EGF), transforming growth factor-α (TGF-α), heparin-binding EGF-like growth factor (HB-EGF), amphiregulin (AREG), β-cellulin, epiregulin, and epigen.[29] These ligands remain anchored along the outer surface of the plasma membrane in their “pro-” or inactive form until the cell membrane-anchored metalloprotease A disintegrin and metalloproteinase (ADAM) proteins activate the ligands. This process is known as ligand “shedding” and occurs in response to specific physiologic signals, which essentially solubilize the EGFR ligands. These mature ligands can then bind to the EGFR receptor, inducing phosphorylation and dimerization of the EGFR subunits and resulting in subsequent activation of the v-ras oncogene homolog/v-raf murine leukemia viral oncogene/microtubule associated protein kinase (Ras/Raf/MAPK), phosphoinositide 3-kinase/protein kinase B (PI3K/AKT), and signal transducer and activator of transcription (STAT) signaling cascades.

It has been demonstrated that exposure to CS induces EGFR activation in nonsmall cell lung cancer (NSCLC) cell lines, although there is no known exogenous EGFR ligand present in CS.[30, 31] The earliest event in the subsequent CS-induced signaling cascade is the generation of de novo ROS by membrane-bound nicotinamide adenine dinucleotide phosphate oxidase.[30] Generation of ROS then induces phosphorylation of v-src avian sarcoma viral oncogene homolog (Src) kinase; and Src, in turn, activates the ε-isoform of protein kinase C (PKCε), which then directly interacts with the tumor necrosis factor-convertase (TACE)/ADAM17 metalloproteinase, phosphorylates it at serine/threonine residues,[32] and TACE/ADAM17 facilitates the cleavage and release of EGFR ligands.[30, 31] Studies disagree on which EGFR ligand is shed from the cell membrane, but reports suggest that CS induces HB-EGF,[33, 34] AREG,[30, 33, 35] and TGF-α[31, 33, 35] to activate EGFR. Although all of those studies used the same H292 lung carcinoma model system, they did not use the same cigarette types or CS exposure preparations (see Table 1). Regardless of which CS-induced ligand binds to EGFR, activation of the receptor appears to trigger the Ras/Raf/ mitogen-activated protein kinase kinase/extracellular signal-regulated kinase (Ras/Raf/MEK/ERK) and PI3K/AKT signaling cascades, resulting in the transcription of several downstream target genes, including MMP-1,[34] interleukin 8 (IL-8),[33-35] and mucin 5AC (MUC5AC).[31, 34, 35]

Table 1. Referenced Methodologies of Cigarette Smoke Preparations Demonstrating Epidermal Growth Factor Receptor Ligand Shedding
StudyCS PrepCig TypeTar, mg/CigNicotine, mg/CigEGFR Ligand
  1. Abbreviations: AREG, amphiregulin; Cig, cigarette; CS Prep, cigarette smoke preparation; CSE, cigarette smoke extract; EGFR, epidermal growth factor receptor; HB-EGF, heparin-binding epidermal growth factor-like growth factor, TGF-α, transforming growth factor-α; TPM, total particulate matter; WS, whole smoke 1R4F, Kentucky Reference Cigarette 1R4F; 2R1, Kentucky Reference Cigarette 2R1; 1R1, Kentucky Reference Cigarette 1R1; 2R4F, Kentucky Reference Cigarette 2R4F.

  2. a

    These were cigarettes with the filters removed.

Richter 200233Aqueous CSE1R4Fa9.20.8TGF-α, AREG, HB-EGF
Lemjabbar 200330WS1R4F9.20.8AREG
Shao 200431Aqueous CSE2R132.92.19TGF-α
Baginski 200635Aqueous CSE1R132.91.98TGF-α
Newland 200834TPM2R4F9.20.8HB-EGF

Ligand-independent EGFR activation

There is also evidence that CS oxidative stress stimulates ligand-independent EGFR activation. Studies in A549 lung cancer cells have demonstrated that oxidative stress from whole CS induced EGFR autophosphorylation at residues Y1068, Y1173, and Y845; but this activation did not generate receptor subunit dimerization, and pretreatment with an EGFR inhibitor did not alter CSE-induced EGFR phosphorylation at these sites.[36] The Y845 residue is a specific Src-dependent phosphorylation site, and analyses have indicated that Src is recruited to EGFR and phosphorylated at Y416.[36] Src generally is not involved in traditional EGF/EGFR signal transduction, thus indicating that the conformation of EGFR induced by CS may be distinct from that induced by EGF binding.[36] Additional analyses determined that the innate kinase activity is neither necessary nor essential for EGFR to interact with Src.[37] Despite the failure of EGFR to properly dimerize because it is in the wrong conformation, the ERK1/2 and AKT signaling cascades were activated, and pretreatment with an EGFR inhibitor did not prevent the observed cascade activation.[36, 38] Further exposure of tyrosine kinase inhibitor (TKI)-sensitive cells to CS induced treatment resistance, and inhibition of Src restored sensitivity to TKIs.[36]

Modulating Cell Death

Evasion of cell death is an essential hallmark of cancer; however, only a few studies have evaluated the impact of CS exposure on the ability of cancer cells to evade death-related pathways. Chronic TPM exposure in A549 cells demonstrated that the observed decrease in TGF-β-induced Smad3/Smad4 complex formation also was correlated with a reduction in apoptosis, which resulted from B-cell chronic lymphocytic leukemia/lymphoma 2 (Bcl-2) up-regulation.[16] Chronic exposure of SCaBER, a bladder cancer cell line, to CSC vapor altered cell tumor biology in several ways; in vitro, CSC vapor induced reduced mitochondrial-resistant protein adenylate kinase-3 levels, decreased mitochondria membrane potential, increased intracellular ROS, and increased cisplatin resistance.[39] In addition, 6-month CSC vapor exposure of SCaBERs increased tumorigenicity, induced cisplatin resistance, and elevated B-cell lymphoma-extra large (Bcl-xL) and Bcl-2 protein expression in subcutaneous xenografts compared with CSC vapor-naive control tumors.[39]

Ratovitski examined the effect of different smoke types on altering cell biology of HNSCC, a head and neck squamous cell carcinoma cell line.[40] The author demonstrated that both mainstream smoke extract and sidestream smoke extract induced the expression of ΔNp63α and nitric oxide synthase (NOS)−2 through the regulation of interferon regulatory factor 6 (IRF6) in the NOS-2 promoter region.[40] Furthermore, mainstream smoke extract induced the cleavage of the autophagic marker LC3B; and knockdown of ΔNp63α, IRF6, and NOS-2 using small-interfering RNAs also modulated the CS-induced autophagic response.[40] Thus, these observations provide evidence that the regulation of NOS-2 expression by ΔNp63α/IRF6 interplay governs the induction of autophagy in response to CS exposure in HNSCC cancer cells.[40] Changes in autophagy and decreased apoptosis associated with CS exposure support the hypothesis that CS confers a prosurvival phenotype, but the dose-dependent and time-dependent effects of CS on cancer cell survival and response to cytotoxic agents are largely unexplored.

Migration and Invasion

Several studies have evaluated the effect of CS on mediating invasion and metastasis in cancer cells (Fig. 2). One such study demonstrated that long-term aqueous CSE exposure conferred a more mesenchymal and invasive phenotype to MCF7 breast cancer cells; the cells took on a spindly, fibroblast-like appearance and exhibited enhanced anchorage-independent cell growth and increased migration through Matrigel-coated filters.[41] Similar phenotypic alterations were observed in both orthotopic and ectopic animal models. Chronically CSE-exposed MCF7 cells established highly invasive and highly metastatic tumors when implanted into the mammary fat pads of immunodeficient nonobese diabetic/severe combined immunodeficiency γ (NSG) mice.[41] In that study, 7 days after implantation, the mock-treated MCF7 cells had formed several intraductal masses, whereas the masses formed by the chronic CSE cells had invaded beyond the confines to the duct and into the surrounding stroma.[41] Furthermore, chronic CSE altered the characteristic of ectopic MCF7 tumors; although the subcutaneous tumors established from the chronic CSE cells were smaller than the MCF7 mock treatment tumors, they also had metastasized to the lungs and liver.[41] In addition, aqueous CSE modulated the expression of metastasis tumor antigen 1 (MTA1), a subunit of the NuRD nuclear remodeling complex thought to be involved in mediating the epithelial-to-mesenchymal transition (EMT).[42] Specifically, NSCLC cell lines exposed to CSE demonstrated increased cell invasion as assessed by Matrigel invasion assay.[42] Further examination indicated that the enhanced metastatic potential was associated with increased expression of both MTA1 mRNA and MTA1 protein.[42]

Figure 2.

The network of proteins involved in mediating cancer cell migration and invasion associated with cigarette smoke exposure is illustrated. α7nAChR indicates α7 nicotinic acetylcholine receptor; ROS, reactive oxygen species; MTA-1, metastasis tumor antigen 1; JAK2, Janus kinase 2; Stat3, signal transducer and activator of transcription 3; Na,K-ATPase, Na pump α subunit; MUC4, mucin 4.

Exposure of A549 cells to CS impaired the function of Na,K-ATPase, an ATP-dependent pump used to maintain the sodium gradient across membranes.[43] Specifically, CS induced ROS production, which reduced Na,K-ATPase pump activity and reduced NaK-α1 levels at the cell surface.[43] The effect of CS-induced decrease in Na,K-ATPase activity on cancer progression needs to be further investigated; however, it is postulated that loss of Na,K-ATPase activity disrupts tight junctions, alters cell polarity, and may be involved in early EMT events.[43] Further examination of the effect of CS on the expression of claudin, a central protein component of tight junctions, in lung carcinoma cell lines demonstrated an early CSE-induced increase in the expression of several claudins followed by a subsequent decrease in mRNA levels.[44] These CS-induced changes in claudin expression may be important in lung cancer biology, because tight junction dysfunction and claudin alterations can cause decreased cell adhesion, loss of differentiation, uncontrolled cell proliferation, loss of cohesion, and invasiveness, all of which contribute to cancer progression and EMT.[44] Aqueous CSE also induced time-dependent and dose-dependent expression of MUC4 mucin in well differentiated pancreatic cancer cell lines.[45] In vivo studies demonstrated that WS exposure of orthotopic pancreatic cancer xenografts increased pancreatic tumor weight and significantly increased the occurrence of metastasis in the liver, stomach, spleen, kidney, peritoneal wall, and mesenteric lymph nodes of nude mice.[45] Furthermore, that study also indicated that CS exposure was correlated with an up-regulation of MUC4 and α7nAChR and increased the phosphorylation of STAT3 (Y705) in primary tumor xenografts.[45] These observations suggest that CS increases MUC4 mucin production in pancreatic cancer through activation of the α7nAChR/Janus kinase 2 (JAK2)/STAT3 signaling cascade, thereby promoting metastasis.[45] These data suggest that CS increases migration and invasion, leading to increased metastasis.

Immune Modulation

Evidence indicates that CS can induce changes in cancer cell immune responses. Analysis of CS composition has revealed that each puff contains from 1014 to 1016 oxidants and that these oxidants can cause prolonged redox-cycling at the lung lining.[7] Vitamin E (α-tocopherol) is an antioxidant present in the alveolar lining fluid of the lung and is responsible for protecting the tissue from oxidant-related damage.[46] Studies have demonstrated that exposure to CS interferes with the ability of A549 lung cancer cells to control tocopherol levels.[46] Specifically, CS induced the degradation and redistribution of SR-B1, the primary receptor involved in regulating tocopherol uptake and decreasing oxidant-induced damage.[46] These observations suggest a mechanism by which CS oxidants intrinsically enhance the carcinogenicity of CS by inhibiting the uptake of vitamin E into lung tumors. Oxidative stresses identified in CS also induced the expression of a novel long-noncoding RNA (lncRNA), termed smoke and cancer-associated lncRNA-1 (SCAL1), in multiple lung cancer cell lines.[47] In addition, SCAL1 expression is regulated by nuclear factor (erythroid-derived 2)-like 2 (NRF2), a transcription factor activated by chemical and oxidative stresses and known to alter the expression of several protective antioxidant genes.[47] Knockdown of SCAL1 with CSE exposure reportedly increased oxidative toxicity, suggesting that SCAL1 may play a role in mediating the cytoprotective function of NRF2 in response to oxidative stress in the lung.[47]

Data also have implicated nuclear transcription factor-κB (NF-κB) signaling as a modulator of immune response by CS in cancer models. NF-κB is a ubiquitous transcription factor that exists in the cytoplasm as a heterotrimer (p50, p65 [RelA], c-Rel, p52, and RelB subunits) and is kept inactive as a dimer in the cytoplasm by the IκB inhibitory protein (IκB isoforms IκBα, IκBβ, IκBγ, IκBδ, IκBε, and Bcl-3). The NF-κB pathway is triggered when an external stimulus interacts with specific receptors and activates the I-κB kinase (IKK) complex made up of 2 catalytic subunits, IKKα and IKKβ, and 1 regulatory subunit, IKKγ. Activated IKK phosphorylates IκB and thereby facilitates IκB degradation, which, in turn, enables NF-κB to translocate into the nucleus and induce target gene expression. NF-κB is involved in the regulation of inflammation and is activated by many different stimuli, such as the cytokines IL-1 and tumor necrosis factor or extracellular stressors like H2O2 and CS.[48] Exposure with TPM induced NF-κB activation in multiple cancer cell types, including U937 cells (human histiocytic lymphoma), HeLa cells (human epithelial adenocarcinoma), Jurkat cells (human T cell), H1299 cells (human NSCLC carcinoma), and 14B and 1483 cells (human head and neck squamous cell carcinoma).[49] Activation of NF-κB by TPM exposure depended on IKKα-mediated degradation of I-κBα.[49] Maity et al identified an alternate NF-κB signaling axis; they determined that, in resting A549 cells, NF-κB in the form of c-Rel/p50 heterodimers is held in complex with I-κBε and, thus, is inactivated.[50] However, the administration of aqueous CSE enables IKKβ-mediated phosphorylation and degradation of I-κBε, which allows the c-Rel/p50 heterodimer to translocate to the nucleus and transcribe NF-κB target genes.[50] Discrepancies between the IKKα-I-κBα-p50/p65 and IKKβ-I-kBε-c-Rel/p50 NF-κB signaling cascades, as reported by Anto et al and Maity et al, respectively, may be caused by differences in CS exposure preparations; Anto et al used the particulate phase of CS, which was collected on a Cambridge filter and dissolved into DMSO, whereas Maity et al examined the role of aqueous CSE in mediating NF-κB signaling.[49, 50] Together, these data suggest that CS can induce the modulation of several different pathways involved in modulating tumor-associated immune responses.


Data suggest that the exposure of cancer cells to CS increases proliferation, migration, invasion, metastasis, and angiogenesis and leads to the activation of immunomodulatory pathways. Data also suggest that CS modulates cell death, leading to a prosurvival phenotype. However, interpretation is limited in part by studies using a variety of CS preparations and on the relative paucity of data reporting on the effects of CS in cancer cells. Significant work is needed to better understand the biologic effects of active CS on cancer cell phenotype. Given the strong clinical correlates demonstrating that continued smoking by cancer patients is associated with increased mortality, toxicity, and recurrence, much work is needed to understand whether the biologic effects of CS in cancer patients can be reversed through smoking cessation or whether critical biologic mediators can be identified to help develop more efficacious cancer treatments.


This work was supported in part by the American Cancer Society (grant MRSG-11-031-01-CCE to G.W.W.).


The authors made no disclosures.