In contrast, in prostaglandin E2 receptor EP2 subtype (EP2)-deficient mice, one of the receptors for PGE₂, developed significantly less tumors than respective wt controls [98]. Interestingly, microsomal PGE synthase mPGES-1 lung-specific (SPC-promoter driven) overexpressing mice that produce elevated levels of PGE₂, were not more susceptible to pulmonary carcinogenesis [99]. The finding from these two studies is perplexing, however, multiple isoforms of PGES exist, several PGE₂ receptors, as well as cell specificity of the models (epithelial versus whole body), may explain some of these discrepancies. A delicate balance between PGI₂:PGE₂ levels likely determines the inflammatory nature of this pathway. When both COX-1 and COX-2 are inhibited (either pharmacologically or genetically), several prostaglandins are inhibited (i.e., (PGI₂ and PGE₂) which may potentially nullify or decrease the effectiveness of the drug. However, as will be discussed below, it may depend on timing of treatment with inhibitors of these pathways.

In NSCLC cell lines, overexpression of PPARγ inhibited anchorage-independent growth and promoted differentiation [105]. In lung-specific PPARγ-overexpressing mice, pulmonary tumorigenesis, as well as suppression of COX-2 via NFκB, was significantly inhibited (75 % reduction in tumor multiplicity) [106]. However, if bone marrow from mice lacking PPARγ specific to myeloid cells (Lys-M-Cre+/PPARγ^flox/flox mice) was transplanted into wild-type mice, there was significant decrease in secondary tumors that were not altered by pioglitazone, a PPARγ agonist [107]. In addition, arginase-positive (a M2 macrophage marker) cells were reduced in the lungs supporting decreased M2 macrophages. Fewer metastases were also observed in an orthotopic mouse model after PPARγ deletion in the myeloid cells using LOX recombination [107]. These M2 alternative macrophages are hypothesized to promote tumor progression through induction of angiogenesis , breaking down the matrix, and increasing tumor cell motility, likely linked to the other two factors [107]. These findings suggest that PPARγ plays dual roles in lung tumorigenesis, both pro- and antitumorigenic, depending on tumor stage studied: antitumorigenic during early stages (initiation/promotion) and pro-tumorigenic during tumor progression and metastasis. We refer readers to go through a more complete review of PPARγ in lung cancer development.

Adiponectin (Acrp30) is an adipocyte-derived protein that is anti-inflammatory and important in regulating fat and glucose metabolism [108]. PPARγ is a central regulator of adiponectin [102]. In COPD, plasma levels of adiponectin are considered a biomarker for certain COPD phenotypes [109]. Adiponectin is also elevated in the BALF from COPD patients [110] and is protective against tobacco-induced inflammation and emphysema in Acrp30-deficient mice [111]. In Acrp30-deficient mice compared to wild-type mice injected with the Lewis Carcinoma cells, the deficient mice were found to promote tumor growth linked to reduction of macrophage recruitment to the site of the tumors [112]. Adiponectin is increased in some COPD patients that are underweight and does appear to be related inversely with body mass index (BMI) [65, 113]. Thus, the protective role of adiponectin in lung cancer may be a potential future therapeutic target.

Mouse strains that are susceptible to lung carcinogenesis, such as FVB or BALB/c, have significantly elevated levels of NFkB in response to urethane in the airway epithelium, type II cells, and macrophages compared to a resistant strain, B6 [122]. The same study also used mice overexpressing, by a Clara cell-specific promoter (CC10), IKBα, an inhibitor of NFκB, in the airway epithelial cells using a Clara cell-specific promoter (CC10) to demonstrate that the inhibition of NFκB greatly decreased lung tumor multiplicity in urethane-induced carcinogenesis compared to those observed in wt mice. In this study, the authors also demonstrated that inflammatory cells (macrophages, PMNs) and cytokines (TNF and IL12p70) were inhibited in the lungs of mice overexpressing IKBα. It is therefore expected that other pro-inflammatory mediators downstream of NFκB are involved in this mechanism (see below). In a second model, transgenic mice overexpressing constitutive IκB kinase β (IKKβ) in the airway epithelium (IKTA mice) were used to provide continual activation of NFκB in the airways of both FVB and B6 backgrounds [123]. These mice exhibited chronic inflammation, PMNS, macrophages, and lymphocytes, without stimulus. Following a single urethane injection, the IKTA mice developed more tumors and inflammation (macrophages, lymphocytes) compared to WT mice, but no differences in tumor size or BrDu uptake, reflecting proliferative rates, were observed [123]. However, increased numbers of early lesions, atypical adenomatous hyperplasias [124], were observed, supported by increased PCNA staining and decreased caspase 3, in the IKTA mice compared to WT mice. T regulatory (Treg) cells were upregulated in the IKTA mice along with the NFKB-induced chemokines that regulate Treg recruitment (CCL17, 20, and 22) [123]. In addition, when the MCA/BHT two-stage model was used with these mice on an FVB background, IKTA mice were found to significantly promote tumor development above that observed with MCA/BHT alone in WT mice. These IKTA mice were also able to promote tumors in the absence of BHT, suggesting that NFκB can act as a tumor promoter [123]. Thus, active NFκB signaling in the lung can alter the lung to create a pro-tumorigenic microenvironment. Interestingly, bortezomib, a proteosome inhibitor used to inhibit NFκB activation, increased tumor number and size in response to urethane while inhibiting NFκB expression in respiratory epithelium and macrophages, 4 months following urethane [66]. These results were surprising, however, likely due to a continuation of inflammation, characterized by elevated macrophages, PMNs, and lymphocytes, as well as multiple cytokines/chemokines, in chronic bortezomib-treated animals. Bortezomib was evaluated in clinical trials for NSCLC and was found to have little effect [125, 126]. An important point here is that this drug is not specific to NFκB and has alternative target pathways.

Additional studies examined the role of NKFB in metastasis. LLC cells metastasized more extensively in the lungs of mice administered adenoviral vectors encoding active NFκB, with increased NFκB activity and hence enhanced lung inflammation prior to tail vein injection of the LCC cells [127].

The roles of NRF2 and KEAP1 in lung cancer are controversial [132, 137–140]. Lung cancer differs from findings in most other organs and pulmonary diseases, such as respiratory syncytial virus infection [141], emphysema [142], and hyperoxia [143]. In the aforementioned respiratory diseases, symptoms significantly worsen in the absence of NRF2. Interestingly, the tumor responses appear unique in mechanism compared to other diseases. Other mouse models (including colon, bladder, liver, and mammary) have demonstrated that a lack of NRF2 increases the potential for carcinogenesis [144–147].

Animal studies on the NRF2 pathway have used NRF2-deficient mice on multiple backgrounds, Keap1-deficient mice, metastasis models, as well as NRF2-deficient mice crossed to a K-Ras ^G12D mouse. These models will be discussed next. Urethane-induced tumorigenesis resulted in significantly less tumor development in NRF2-deficient mice (^−/−) than the NRF2 ^+/+ mice (WT; BALB background strain) [148]. However, NRF2 ^−/− mice had enhanced inflammatory cell infiltrates including monocytes, macrophages, and lymphoctyes, hyperpermeability, and elevated myeloperoxidase, that was suggestive of increased numbers of PMNs, compared to the NRF2 ^+/+ mice 11 weeks following urethane. Additionally, NRF2 ^+/+ mice had early adenomatous lesions 12 weeks following urethane above that observed in the NRF2 ^−/− mice, while the NRF2 ^−/− mice had increased apoptotic cells compared to the wild-type mice [148]. Finally, cell death markers, such as LDH, support the hypothesis that the urethane-initiated epithelial cells, such as the type II cells, were more susceptible to cell death in the mice lacking Nrf2. The NRF2 ^+/+ mice, therefore, have both a growth advantage and an increased cytoprotection for tumorigenesis.

Transcriptome analysis revealed differences between the NRF2 ^−/− and NRF2 ^+/+ mice involved in these responses to urethane at an early and late time point. Early (12 weeks) NRF2-modulated genes involved glutathione metabolism, cell–cell signaling, oxidative stress, and immune responses [148]. At the advanced stage, the NRF2-dependent genes associated with cell cycle/proliferation and cell death, which correlated in direction and magnitude with the increased death of initiated cells in the NRF2 ^−/− mice. In this primary mouse cancer model, NRF2 promotes survival properties and therefore supports the human studies demonstrating resistance to anticancer drugs as well as increased malignancy. NRF2 protection may be both carcinogen and stage dependent since studies using the two-stage MCA/BHT model with NRF2 ^−/− mice did not demonstrate differences in tumor numbers between strains (unpublished data, A. K. Bauer and H. Y. Cho).

The concept of ROS reduction leading to increased carcinogenesis is against the normal dogma [149]. However, oncogenes K-Ras ^G12D or B-Raf ^V619E (mutated, activated) expressed in vitro in murine NIH3T3 or mouse embryonic fibroblasts (MEFs) resulted in decreased ROS levels [150]. In contrast, Nrf2 activity was elevated while the ratio of reduced to oxidized levels of glutathione (GSH/GSSH) increased [150]. K-Ras and B-Raf signaled through MEK, ERK MAP Kinase, and AP-1, eventually activating Nrf2, leading to antioxidant responses [150]. In vitro studies were validated in several in vivo mouse models (lung and pancreatic cancer). NRF2 ^−/− and NRF2 ^+/+ mice were bred to the K-Ras ^G12D mouse (B6/129/SJL background strain) in lung tumor studies which demonstrated that the NRF2 ^−/− mouse had a significant reduction in K-Ras ^G12D-initiated lung tumors compared to WT mice [150]. Ki67 staining was also reduced, indicating less proliferation and overall decrease in adenomas, adenomatous alveolar hyperplasias, and bronchiolar hyperplasias [150]. Thus, these studies suggest that Nrf2 can be regulated by specific oncogenes (Kras, B-raf) to enhance tumorigenesis by reduction of ROS creating a more favorable cellular microenvironment. The findings described here support the other primary mouse study which suggests that initiated cells are reduced in mice lacking NRF2. Less initiated cells lead to a decreased number of tumors. Thus, the cellular environment in the Nrf2^−/− mice lacks the favorable protection of mice with sufficient NRF2 [148].

Lastly, NRF2-deficient mice on a C57/BL6/J background developed a significantly higher number of metastatic LLC lung nodules than wild-type mice [151]. Inflammation, both pulmonary and bone marrow (myeloid-derived suppressor cells; MDSCs), was also elevated in the NRF2-deficient mice with metastatic tumors in the lung. KEAP1 mutant mice with increased levels of Nrf2 protein were resistant to metastasis [151] and demonstrated reduced levels of ROS. In metastasis, Nrf2 appears to be preventive, which is the reverse of findings observed in the primary mouse lung cancer studies. However, the mechanisms regulating metastasis differ from the earlier stages of carcinogenesis. Another difference between studies was the strain backgrounds which are known to differ in many phenotypes, including the polarity of their immune systems, which could influence responsiveness [152].

Interleukin 6 (IL6) can be elevated in human COPD and is considered a pro-inflammatory cytokine linked to lung cancer [157, 158]. Il6-deficient mice bred to the CCLR mice (Kras^G12D) were used alone and in the NThi COPD-like model [69]. These mice with no stimulus developed fewer tumors than the CCLR mice (WT) alone. Additionally, when exposed to NThi, these mice exhibited significant reductions in promotion, suggesting IL6 is a critical cytokine in the inflammation observed in response to NThi, a COPD-like model.

IL10 is an anti-inflammatory cytokine, based on suppression of macrophage and dendritic cell (DC) function, including production of pro-inflammatory cytokines and antigen presentation, Th2 cytokine [159]. However, feedback mechanisms of both the Th1 and Th2 responses can result from suppression. In UV-induced skin carcinogenesis, IL10-deficient mice were found to be resistant [160]. IL10 is immunosuppressive to CD8 + T cell effector functions and DC maturation, which can lead to an impaired antitumor responses [159, 161]. Clinically, recombinant IL10 suppressed endotoxin and resulted in immune stimulatory capacities as well as anti-inflammatory roles [159]. NSCLC patients expressing elevated IL10 levels in tumor-associated macrophages (TAM) were associated with a worse prognosis [162]. In urethane-induced carcinogenesis, Il10 heterozygous (−/+) mice enhanced lung tumor multiplicity in female mice only compared to wt controls [163]. Therefore, IL10 appears to both suppress pro-inflammatory pathways, such as TNF, and impair antitumor responses. Additional IL10 tumor models demonstrated that lower IL10 levels appear immunosuppressive while higher IL10 levels may induce immune activation events, such as tumor rejection [159]. Gender also appears to be involved in lung IL10 responses; the mechanism behind gender involvement is unclear. Thus, the tumor microenvironment in addition to the amount of IL10 may determine the outcome of the response.

IL17 is a CD4 helper T cell-derived pro-inflammatory cytokine that is involved in upregulating cytokines and chemokines [164]. Increases in IL17 have been linked to asthma, COPD, and several types of tumors, including lung ADC. In the CC-LR and CC-LR-NTHi mouse models, IL17 was significantly elevated [165]. IL17-deficient mice crossed to the CC-LR (mutant K-ras) mice in the presence or absence of NTHi developed far fewer tumors than the wild-type mice [165]. Inflammatory responses in these Il17-deficient mice were also greatly reduced [165] as well as the cytokines regulated by IL17, such as Il6, Cxcl2, Ccl2, and Arg1. Additionally, l17-deficient mice had reduced levels of myeloid-derived suppressor cells (MDSCs) or Gr1⁺Cd11B⁺cells, which were mostly Ly6G⁺, indicating mostly PMN-like MDSCs. Lastly, CC-LR mice administered anti-Gr1 also resulted in reduced numbers of tumors and inflammation, indicating these PMN-like MDSCs are involved in the regulation of pro-tumorigenic inflammation.

Lung cancer risk is significantly decreased in those individuals exposed to endotoxin, such as farm and textile workers [175–179]. The primary receptor that binds endotoxin (lipopolysaccharide; LPS) is TLR4 [167], thus it is likely involved in the protection observed with endotoxin exposure. BHT-induced inflammation (macrophages and lymphocytes) and injury was significantly higher in Tlr4-deficient mice than the wt controls [180]. Additionally, Tlr4-deficient mice were more susceptible to tumor multiplicity compared to wt mice 20 weeks following the MCA/BHT protocol. Thus, in the MCA/BHT model, TLR4 appears to protect the pulmonary epithelium possibly by reducing the inflammation and pulmonary injury elicited by the oxidative metabolites of BHT. In other types of cancers, other TLR agonists, such as for TLR7 or TLR8 (synthetic imidazoquinoline, imiquimod, resiquimod, etc.), are approved for treatment of basal cell carcinoma and potentially for breast and melanoma [181]. Thus, these innate immune receptors should not be overlooked for future therapeutic strategies in lung.

Interferon γ is a Th1 cytokine and a key immunoregulatory mediator involved in eliciting innate and adaptive responses in host defense [182, 183]. IFNγ can activate macrophages, upregulate MHC class II molecules, and stimulate natural killer cells [184]. Specifically, IFNγ increases nitric oxide synthase [167], indoleamine (2,3)-dioxygenase(tryptophan degradation), and reactive oxygen species (ROS) [184]. MCA-induced fibrosarcoma was significantly enhanced in Ifnγ-deficient mice compared to wt controls [185]. Additionally, some human tumors are unresponsive to IFNγ [185], which supports a more general antitumorigenic role of this cytokine. Ifnγ –deficient mice developed significantly more and larger tumors than wt mice in response to the MCA/BHT model (A. K. Bauer and S. R. Kleeberger, unpublished results). BHT-induced chronic inflammation in the Ifnγ-deficient mice was also significantly elevated compared to wt controls. In a separate study, Ifnγ-deficient mice stimulated tumor growth in urethane-induced carcinogenesis [186]. In human cancers including NSCLC, IFNγ upregulates antiangiogenic chemokines such as monokine-induced by IFNγ (MIG; CXCL9) and interferon-activated p10 (IP10; CXCL10) [187–190]. These studies on specific innate immune pathways imply that individuals with defective innate immune systems may be more susceptible to pulmonary neoplasia.

M2 alternative activation by IL4 and IL13 results in increased Th2 responses and increased proliferation, or tumor promotion [193, 194]. Il4 receptor α-deficient mice (Il4Rα ^−/−) in response to urethane-induced carcinogenesis, developed smaller tumors than those observed in the wild-type BALB/cJ mice [186]. Phenotying the macrophages in urethane-exposed lungs revealed that M2 macrophages predominate in the pre-malignant lesions. In contrast, those lungs bearing carcinomas contain only M1 macrophages. Bone marrow monocytes adopted these expression patterns before entering the circulation. These different phenotypes of macrophages are likely involved in regulating the Th₁ versus Th₂ cytokine response in the tumor microenvironment and appear to influence tumor development, as was seen in the Ifnγ− and IL4Rα-deficient mouse models [186]. Because of the bone marrow involvement, these two cytokine/cytokine receptors may be early biomarkers for pulmonary AC.