Melanoma is the deadliest form of skin cancer, and virtually all patients progress on targeted therapies. Dysregulated metabolism has been shown to affect therapy response, so BRAF-mutated melanoma cell line models were used to connect cellular metabolism to therapeutic proliferative response. The data show that forcing a glycolytic metabolic strategy in the context of drug treatment enhances the antitumor effect. Anti-retrovirals, particularly zalcitabine, were shown to dramatically affect proliferation when combined with BRAF inhibitor. All in all, this dissertation provides an important contribution to response variability and assay development, the glycolytic biology in relation to BRAF inhibition, and a finer inspection of variability within a tumor.

It was been well established that the Activin A signaling pathway plays a pivotal role in the developing, adult, and diseased organism, commonly acting as a growth inhibitor. Though prevalent in embryonic tissues, Activin A secretion is downregulated in post-natal tissues. Contrary to its function, Activin A expression is often upregulated in cancer, and, of particular interest, in esophageal (ESCC) and head and neck squamous (HNSCC) cell carcinomas. Therefore, we investigated the function of Activin A in these contexts. Interestingly, we found that Activin A acts as an inhibitor of invasion and regulator of the extracellular matrix on dysplastic and ESCC cell lines that retain expression of Activin A receptor type IB, ACVRIB. When ACVRIB is lost, Activin A no longer exerts these effects, though the other components of the Activin A receptor complex remained intact. Consequently, we decided to further define the functional effects of loss of ACVRIB in HNSCC and ESCC cells using CRISPR/Cas9 and siRNA, respectively. In the absence of ACVRIB, we found increased proliferation, migration, and invasion. Using an organoptypic culture system and immunofluorescence staining, we observed altered expression of proteins responsible for cell-cell and cell-extracellular matrix adhesion. We conclude that ACVRIB-dependent Activin A signaling is necessary to regulate ESCC and HNSCC progression.

Despite recent progress, non-small cell lung cancer (NSCLC) remains the leading cause of cancer-related mortality in the United States and new therapeutic approaches are needed. Nuclear factor κB (NF-κB) is a master regulator of inflammatory signaling that is overexpressed in most solid tumors. Several mouse models of lung cancer have confirmed the requirement for NF-κB signaling in airway epithelial cells during lung tumorigenesis. However, despite these findings, inhibitors of NF-κB have been ineffective in treating NSCLC patients. Using genetic and pharmacologic inhibition of NF-κB signaling in murine lung cancer models, we found that blockade of NF-κB signaling in the myeloid inflammatory cell population paradoxically increases lung inflammation, airway epithelial cell proliferation, and lung tumorigenesis. We identified cathepsin G-mediated processing of IL-1β by neutrophils as a novel resistance mechanism of NSCLC to NF-κB inhibitors, and combined therapy with an NF-κB inhibitor and IL-1 receptor antagonist reduced tumorigenesis in mouse models of lung cancer. In NSCLC patients, plasma IL-1β concentration inversely correlated with progression-free survival and IL-1β levels were increased following treatment with an NF-κB inhibitor. These studies demonstrate that targeting common signaling pathways can have opposing effects in individual cell types during tumorigenesis; they support the use of rational, combined therapies to treat lung cancer.

Head and neck squamous cell carcinoma is a frequent and serious malignancy that accounts for more than 300,000 deaths world wide each year, the majority of which are the result of local invasion and lymph node metastasis. LIM-only protein 4 (LMO4) and LIM-domain binding protein 1 (LDB1), transcriptional adaptors that have important roles in normal epithelial cell differentiation, have also been associated with increased lymph node metastasis and decreased differentiation in carcinomas of the breast and of the head and neck. Here within, LMO4 and LDB1, in addition to single-stranded-binding proteins (SSBPs) are involved in regulating invasion, and provides an explanation for why LMO4 and LDB1 have concordant levels of expression in this disease. This work directly shows that VU-SCC-1729 cells can invade cellular monolayers and that loss of LDB1 in these cells not only reduces invasion, but also plays an unexpected role in the reduction of proliferation. A loss of LDB1 in VU-SCC-1729 cells also resulted in reduced tumor volume and vascularization in a xenograft tumor model. Genome-wide DNA occupancy studies evaluated LDB1 and SSBP2 binding patterns in head and neck cancer cells. Together, these data suggest that LMO4, LDB1, and SSBPs may have a collective role in regulating processes critical for progression of head and neck carcinoma including invasion, proliferation, and angiogenesis.

In spite of recent advances in therapy, cancer remains a leading cause of death worldwide. Therapy response is often unpredictable and relapse frequently occurs. In many cases, this therapy resistance is attributed to subsets of therapy resistant cancer cells and surrounding stromal cells that support a resistant phenotype. A better understanding of cellular heterogeneity in cancer is therefore crucial in order to develop novel therapeutic strategies and improve patient outcomes. Experimental technologies like mass cytometry (CyTOF) allow for high-content, multi-parametric single-cell analysis of human tumor samples. However, analytical tools and workflows are still needed to standardize and automate the process of identifying and quantitatively describing cell populations in the resulting data. This dissertation presents a novel workflow for automated discovery and characterization of novel and rare cell subsets, quantification of cellular heterogeneity, and characterization of cells based on population-specific feature enrichment. First, a modular workflow is described that combines biaxial gating, dimensionality reduction, clustering, and hierarchically clustered heatmaps to maximize rare population discovery and to create an interpretable visualization of cell population characteristics. Next, a novel method is introduced for quantifying cellular heterogeneity based on two-dimensional mapping of cells in phenotypic space using tSNE analysis. Finally, an algorithmic method termed Marker Enrichment Modeling (MEM) is introduced that automatically quantifies population-specific feature enrichment and generates descriptive labels for cell populations based on their feature enrichment scores. MEM analysis is shown to identify features important to cell identity across multiple datasets, and MEM labels are effectively used to compare populations of cells across tissue types, experiments, institutions, and platforms. Going forward, the tools presented here lay the groundwork for novel computational methods for machine learning of cell identity and registering cell populations across studies or clinical endpoints. Automated methods for identifying and describing cell populations will enable rapid discovery of biologically and clinically relevant cells and contribute to the development of novel diagnostic, prognostic, and therapeutic approaches to cancer and other diseases.

Oral Squamous Cell Carcinoma (OSCC) is the sixth most common cancer worldwide. OSCC commonly invades into the lymph nodes and mandible, which correlate with increased rates of recurrence and lower overall survival. Tumors that infiltrate mandibular bone proliferate rapidly, cause large amounts of bone destruction and require extensive surgeries. Unfortunately, the molecular mechanisms of OSCC invasion into the mandible are not well understood and survival rates have not significantly improved for over 30 years. In a syngeneic model of murine OSCC, Parathyroid Hormone-related Protein (PTHrP), has been shown to be required for OSCC invasion into the mandible. Existing studies have identified the Hedgehog (Hh) transcription factor, Gli2 as the regulator of PTHrP and our previous work in breast cancer metastasis to bone suggest that TGFB may regulate Gli2 transcription. Here we demonstrate that Hh and TGFB signaling concomitantly regulate Gli2 and subsequently PTHrP in OSCC. Additionally, we have elucidated a mechanical signaling mechanism that controls Gli2 activation levels, where extra-cellular matrix rigidities similar to bone, but not basement membrane, causes an increase in ciliogenesis. We also demonstrate the feasibility of targeting Gli2 in vivo using GANT58 loaded microspheres to prevent bone destruction.

Lung cancer is the leading cause of cancer deaths worldwide. More than 40% of lung cancers are classified as lung adenocarcinoma (LUAD), which is defined by pathological characteristics. Oncogenic mutations in the epidermal growth factor receptor (EGFR) tyrosine kinase domain (TKD) are found in a significant portion of LUAD. Such mutations, which present most commonly as small in-frame deletions in exon 19 (ex19del) or as point mutations in exon 21 (L858R), allow for EGFR activation, sensitivity to EGFR tyrosine kinase inhibitors (TKIs), and molecularly targeted therapy in this malignancy. By analyzing the tumors of patients with LUAD, we identified two EGFR alterations that had not previously been reported in lung cancer: the EGFR kinase domain duplication (EGFR-KDD) and the EGFR-RAD51 fusion. While knowledge about common EGFR mutations (L858R or ex19del) has allowed for rational treatment of specific cohorts of LUAD, nothing was known about the EGFR-KDD and EGFR-RAD51 in LUAD. These mutationsâwhich we call EGFR rearrangementsâmaintain the wild type sequence of the EGFR TKD and could have mediated either sensitivity or primary resistance to EGFR-TKI therapy. The goal of these studies was to fill in this knowledge gap: to characterize a new type of EGFR alteration in LUAD, to determine the sensitivity of these rearrangements to therapy, and, along the way, to uncover more about the biology of EGFR. Collectively, the research in this dissertation used in vitro and in silico models, as well as patient-derived data, to address a previously unrecognized potential mechanism of oncogenesis in lung cancer: EGFR rearrangements. These studies established EGFR-KDD and EGFR-RAD51 as bona fide recurrent oncogenes and therapeutic targets in LUAD. The work is already proving clinically relevant as these alterations have been identified in additional lung cancer patientsâwho may now have more targeted treatment options.

Adult T-cell leukemia/lymphoma (ATLL) is an incurable peripheral T-cell malignancy where most patients succumb within the first year of diagnosis. Development of ATLL requires human T-cell lymphotrophic virus type 1 (HTLV-I) retroviral infection followed by accumulation of somatic mutations and changes in gene expression. The discovery of genes and pathways involved in the initiation of ATLL may provide novel therapeutic targets for treatment of this fatal disease. Since the IL-2 signaling pathway plays an important role in ATLL development, mutational analysis of IL-2 signaling pathway components should yield a better understanding of disease progression and outcome. Janus kinase 3 (JAK3), a nonreceptor tyrosine kinase is a key kinase upstream in the IL-2 signaling pathway. Activating somatic mutations in JAK3 have been described in leukemias and lymphomas including acute megakaryoblastic leukemia and natural killer/T-cell lymphoma. Three mutations in the regulatory FERM domain of JAK3 were identified in four of the thirty-six ATLL patients screened and no mutations were found in the twenty-four screened ethnically matched controls. These somatic, missense mutations occurred in the amino terminal regulatory FERM domain. JAK3 FERM domain may play an autoregulatory role by inhibiting kinase activity in the absence of IL-2. In cell culture assays all three FERM domain mutations induce gain of function in JAK3 and can be inhibited with a JAK specific kinase inhibitor, tofacitinib. One of the JAK3 FERM domain mutations (E183G) was characterized in vivo and found to be oncogenic in cooperation with the loss of cell cycle regulatory proteins p16(INK4a) and p14(ARF). These findings emphasize the importance of JAK3 activation in ATLL development and offer a novel therapeutic target for this incurable disease.

LZAP (Cdk5rap3, C53) is a putative tumor suppressor lost in 30% of human head and neck squamous cell carcinoma. LZAP depletion enhances cell invasion, xenograft tumor growth and angiogenesis. Mechanistically, LZAP inhibits activity of NF-¦ÊB and Chk1 and Chk2 through binding and decreasing phosphorylation of critical proteins. LZAP has no known enzymatic activity, implying that its biological functions are likely mediated by protein-protein interactions. The focus of this dissertation is to better understand LZAP activity through discovery of LZAP-associated proteins. Immunoprecipitation followed by liquid chromatography¨Cmass spectrometry identified members of protein phosphatase 2C family, PPM1A and PPM1B, as LZAP binding proteins Wip1/PPM1D, a known RelA phosphatase is a member of this family. Here we show that PPM1A dephosphorylates S536 and S276 of RelA and selectively inhibits NF-¦ÊB transcriptional activity resulting in decreased expression of cytokines implicated in cancer metastases. In human prostate cancer, metastatic deposits had lower PPM1A mRNA compared to primary tumors without metastases. In xenograft models, PPM1A expression inhibited bony metastases, whereasPPM1A depletion enhanced tumor growth. Interestingly, activities of LZAP and PPM1A to inhibit RelA inhibition are mutually dependent. LZAP contains motifs predicted to bind the docking domain (D domain) of mitogen activated protein kinases (MAPKs). LZAP binds and inhibit p38 phosphorylation, alters p38 cellular localization, and inhibits basal and cytokine-stimulated p38 activity. LZAP binds the p38 phosphatase, Wip1, and facilitates Wip1 binding and dephosphorylation of p38. These data suggest that LZAP inhibition of p38 phosphorylation and activity depends, at least partially, on Wip1. Regulation of LZAP activity is not well understood; however, we found that LZAP protein is modified by phosphorylation and ubiquitination, forms dimers/oligomers, and may be a target of PPM family members. These findings provide insight into LZAP activity and may lead to mechanisms to restore LZAP activity. Together, this work suggests a common mechanism of LZAP activity to regulate important cancer-related proteins, namely interaction with and regulation of PPM phosphatase family members.

The overall survival for neuroblastoma remains dismal, in part due to the emergence of resistance to chemotherapeutic drugs resulting in aggressive, refractory disease. As a neuroendocrine tumor, neuroblastomas secrete a number of peptides; one such being the gastrin-releasing peptide (GRP). GRP antagonists have been used to inhibit the proliferation of cancer cells. Here, I demonstrate that GRP silencing induced apoptosis in neuroblastoma cells and, in combination, allowed the usage of sublethal doses of chemotherapeutic drugs to elicit responses similar to lethal doses of the same drugs when used alone. GRP silencing also decreased tumorigenesis in vitro and suppressed liver metastasis in vivo. Moreover, GRP silencing increased PTEN levels with a simultaneous inhibition of AKT/mTOR/FAK activation in neuroblastoma cells. Interestingly, PTEN overexpression inhibited GRP-mediated neuroblastoma progression in vitro. This placed PTEN as a critical negative regulator of the oncogenic effects of GRP in neuroblastoma progression. This study provides a rationale for the use of GRP antagonists in patients with aggressive, refractory neuroblastomas.