Erica Golemis, PhD

In disease states such as cancer, tumors reprogram their signaling to support abnormal growth processes. Our laboratory is interested in defining the changes in cell signaling that occur as tumors initiate, progress, and develop resistance to drugs, with the ultimate goal of inhibiting these processes. We hope through these studies to better define the interactions of signaling pathways in malignant versus normal cells, allowing improvements in cancer diagnosis and treatment. In specific projects, our work divides between a body of translational studies focused on the optimal use of therapeutic drugs, and more basic investigations into fundamental cell signaling mechanisms.  

One focus of laboratory interest is in the study of smoking-related cancers. Work by Alexander Deneka addresses head and neck cancer (HNC), a disease commonly arising from chronic exposure to tobacco smoke and ethanol use, which cause damaging mutations in tumor suppressor gene TP53. Cells with TP53 mutations have impaired cell cycle checkpoints, and evade radiation- and cisplatin-induced cellular damage, resulting in rapid emergence of therapeutic resistance. In collaboration with the laboratory of Barbara Burtness (Yale Cancer Center), we have explored synergistic combinations of drugs targeting cell cycle transitions, identifying several effective combinations which will soon be examined in early phase clinical trials.  Dr. Deneka is also collaborating with researchers at Caris to analyze patterns of mutation in HNC that provide potential biomarkers for use of these and other drug combinations under preclinical and clinical evaluation. The goal of this work is to develop new and effective modality for treating TP53-mutated cancers.

In other work, Dr. Deneka has been identifying new mechanisms regulating invasiveness in lung cancer and evaluating new therapeutic approaches for this disease. Non-small cell lung cancer (NSCLC) has a low survival rate, with metastasis contributing to the vast majority of deaths. Dr. Deneka has used knockout mouse models and complementary studies in human NSCLC cell models to define a role for the scaffolding protein NEDD9 as a regulator of NSCLC growth and therapeutic response.  Mechanistically, he has shown that NEDD9 regulates autophagy and glycolysis, metabolic processes that fuel the early growth of NSCLC tumors (Deneka et al, submitted), suggesting that NEDD9 expression may be a potential biomarker for in vivo response to drugs targeting tumor metabolism.

In collaboration with the laboratory of Igor Astsaturov, Anna Lilly is exploring the function of sterol response element binding proteins (SREBPs) and SREBP cleavage-activating protein (SCAP) in regulating pancreatic epithelial differentiation, with the ultimate goal of better understanding how lipid profile affects pancreatic cancer risk. To this end, Lilly is using engineered mouse models with conditional pancreatic loss of SCAP, a critical chaperone regulating SREBP stability and activation.  In pilot experiments with one such model (Pdx1-Cre;Scap^f/f), mice lacking Scap expression during pancreatic development had selective and severe impairment of acinar differentiation, but not on ductal architecture, or integrity of endocrine islets.  In mice genetically predisposed to pancreatic cancer, this phenotype was linked to a significant change in the frequency and gross presentation of tumors. Our data suggest SCAP function is essential in controlling differentiation or survival of specific cell populations within the pancreas, potentially affecting the availability of progenitor populations subject to tumorigenesis; we are investigating this idea.

Compromised barrier function of colon epithelial tissue is associated with inflammatory bowel diseases (IBDs), and increases the risk of colorectal cancer (CRC) through a process of tumor-elicited inflammation (TEI). Tight junction (TJ) proteins, including specific claudins, are critical regulators of epithelial barrier function and paracellular permeability. Changes in claudin composition in the kidney are strongly linked to the etiology of autosomal dominant polycystic kidney disease (ADPKD), an inherited disease affecting ~ 1 in 1000 individuals. ADPKD-inducing mutations in PKD1 causing non-leaky barriers that can withstand high hydrostatic pressure within renal cysts a hallmark of this common disease. Intriguingly, a large population study suggested decreased incidence of CRC in ADPKD patients, while other studies implicated PKD1 in control of epithelial-mesenchymal transition (EMT) and invasion. After years of working in both ADPKD and cancer, we had become struck by the fact that many of the signaling networks activated in ADPKD were extremely similar to those activated and oncogenic in cancer: but ADPKD did not result in increased incidence of tumors. Anna Nikonova is investigating the role of PKD1 in the initiation and progression of colorectal cancer, based on a model in which the PKD1 gene plays a central role in controlling barrier integrity and function in colonic epithelia, regulating cancer risk and progression. This work may reveal specific therapeutic combinations for use in limiting tumor growth, which could help in prevention and treatment for individuals with CRC. 

Reciprocally, Dr. Nikonova is using insights from her studies of PKD1 signaling at cilia with the goal of developing safe and effective therapeutic options for patients with ADPKD. This work follows directly from our extensive prior studies of ciliation, and are aimed both at identifying new and potentially beneficial therapeutic combinations for ADPKD, and assessing the risks versus benefits to ADPKD patients of therapies employed for cancer (as it is estimated that 35-40% of all individuals, including patients with ADPKD, will develop cancer at some point in their lifetime). Hence, study of the relationship of ADPKD and cancer signaling is important not only to potentially improve outcomes in ADPKD patients, but also to avoid inducing harm.

Ilya Serebriiskii has led collaborative studies with Joshua Meyer and Foundation Medicine that have investigated mutational profiles in CRC.  This work has focused on differences in mutations associated with early- versus late-onset CRC, and in the colon versus rectum tumor subsite.  Initial studies analyzed data from 18,000 tumors to establish overall mutational frequencies, and to analyze in detail the frequency of distinct disease-associated mutations in the oncogenic drivers KRAS and NRAS. This work is ongoing to elucidate additional trends in RAS mutational profiles using a larger set of 34,000 tumors, and also is expanded to analyze other oncogenes and tumor suppresors pertinent to CRC, such as PTEN.

Beyond CRC, mutational hyperactivation of KRAS, NRAS, or HRAS occurs in 25% of human cancers overall, and in a subset of developmental syndromes. Mitchell Parker, a joint graduate student with the laboratory of Roland Dunbrack, has been working with large datasets to refine understanding of the structural determinants of RAS activation, inactivation, and inhibition. His analysis is based on use of a machine learning approach to dissect structural features of the 646 human KRAS, NRAS, and HRAS structures deposited in the Protein Data Bank (PDB) from 1990-2021, including RAS WT and mutated structures in complex with various entities, such as biological ligands, metal ions, drug candidates, chemical compounds, and bound proteins or peptides. The goal is to develop a unified structural classification of RAS conformations that can guide the design of targeted inhibitors.