Lung cancer is the second most common cancer and the leading cause of cancer-associated mortality in the U.S.. The overall 5-year survival of lung cancer patients is less than 20%. The majority of patients are diagnosed with advanced stage disease. While progress has been made with targeted therapies, 5-year survival has so far improved in an incremental manner. Lung cancer is characterized by a prominent inflammatory tumor microenvironment, which in turn represents an important prognostic factor in patients. Unresolved inflammatory conditions promoted by smoking, chronic obstructive pulmonary disease (COPD) and interstitial lung diseases increase the risk of developing lung cancer. Therefore, tumor-promoting inflammation may play a significant role in cancer initiation and progression.

The first part of this dissertation focuses on the effect of the tumor suppressor LKB1 on promoting an inflammatory microenvironment in non-small cell lung cancers (NSCLC). Loss of function of LKB1/STK11 is evident in approximately 30% of primary NSCLC. In murine lung cancer models, Kras and Lkb1 double mutation generates highly metastatic lung tumors with different histological types. Although a variety of different mechanisms have been proposed to explain the tumor-promoting effects underlying LKB1 deficiency, no effective therapy has been applied clinically; loss of function mutations presents a therapeutic challenge. Most recently, studies have indicated that LKB1 loss favors an immune-suppressive microenvironment characterized by prominent inflammation, suggesting a new perspective for therapies. Utilizing normal human bronchial epithelial cells (HBECs) which were immortalized in the absence of viral onco-proteins, we find that knockdown of LKB1 elevates the production of multiple inflammatory proteins, among which CXCR2 ligands are the most abundantly secreted. Our data indicate that knockdown of LKB1 in HBECs leads to transcriptional and translational upregulation of CXCR2 ligands and conversely, forced expression of wild-type LKB1 in LKB1-null NSCLC tumor cells decreases CXCR2 ligand production. Non-supervised clustering analysis further reveals KRAS and LKB1 double mutation in human NSCLC cell lines predicts higher levels of CXCR2 ligands. In addition, gene expression analysis shows that CXCR2 ligands are also significantly elevated in murine KrasG12D; Lkb1-/- lung tumors compared to KrasG12D and KrasG12D; Tp53-/- tumors. Dissection of the underlying mechanisms reveals that the NF-κB and WNT pathways regulate CXCR2 ligands downstream of LKB1. Surprisingly, regulation of the NF-κB pathway by LKB1 is independent of AMPK, but requires the MARK family proteins. Knockdown of MARKs or inhibition of MARK function by a small chemical inhibitor in HBECs recapitulates LKB1 loss-induced NF-κB activation and subsequent CXCR2 ligand upregulation. CXCR2 ligands have been reported to play an important role in tumor initiation and progression via recruitment of immune cells and endothelial cells in a variety of cancer types including NSCLC. Therefore, our findings suggest that elevation of CXCR2 ligands by LKB1 deficiency facilitates tumor development by creating a tumor-favored microenvironment. Investigating the contribution of CXCR2 ligands to LKB1-dependent malignancy may aid in the development of novel prevention as well as therapeutic strategies against LKB1-null NSCLC.

The second part of this dissertation examines the impact of dysregulated inflammation on cancer progression. The plasticity of epithelial to mesenchymal transition (EMT) program has been considered to be an essential element regulating cancer metastasis. Cancer cells undergoing EMT need to maintain the mesenchymal phenotype during metastasis but revert back to epithelial phenotypes for successful outgrowth of clones at metastatic sites. However, the determinants of EMT plasticity are not yet clear and the underlying mechanisms have not been fully explored. Recently, we have found that a subset of NSCLC cells undergo EMT in the presence of cytokines including IL-1β, TNF-α and TGF-β (within 7 days), and this occurs concomitantly with increased cell migration and invasion. In addition, chronic exposure to these inflammatory cytokines leads to EMT memory, which refers to the phenomenon in which cells are able to maintain EMT despite withdrawal of the original stimulus. Intriguingly, in contrast to the acute EMT process, EMT memory uniquely depends on chronic cytokine exposure, and not on the signaling pathways (JNK/ERK) and transcription factors (fra-1/slug) mediating the acute EMT. Further studies demonstrate that E-cadherin is repressed via a dynamic alteration of histone modifications and subsequent DNA methylation during chronic IL-1β exposure. Furthermore, a pathway analysis of the RNA profile of these cells indicates that a large portion of the altered genes can be methylated. Phenotypically, EMT memory allows cancer cells to maintain highly migratory and invasive features during metastasis. These findings, for the first time, demonstrate that EMT memory is uniquely induced by chronic inflammation and identifies epigenetic modifications as its underlying mechanism. Better understanding of EMT will ultimately assist in the identification of targets for preventing and treating metastatic behaviors in lung cancer.