Thoracic Oncology Lab Research

KRas Pathway

Mutations in the Ras genes occur frequently in human cancers, and with clear tumor-type specificity. For instance, whereas Hras mutations are frequent in papillary thyroid cancer and Nras mutations in hepatocellular carcinoma and melanomas, mutations in Kras are predominant in lung, pancreatic and colon cancers. The reason for this tissue-specificity has been a long standing question in the Ras field.

Minh To and colleagues demonstrated that mice completely lacking Kras protein, but express Hras under the regulation of the Kras locus, were highly susceptible to carcinogen induced lung tumors. Molecular analysis showed that the majority of lung tumors in these mice have acquired an oncogenic mutation in the Hras that was knocked into the Kras locus, while the endogenous Hras gene remained unaffected. Mutation-specificity preferences therefore involve regulatory elements specific to each Ras gene, rather than functional properties of the encoded proteins.

Kras is expressed as two isoforms, Kras4A and Kras4B, as a result of alternative splicing. Kras4B is the dominant isoform and is thought to be the major effector of the oncogenic activity ascribed to mutant Kras. However, mice expressing only the Kras4B isoform turned out to be highly resistant to lung tumor development, suggesting that Kras4A is the major determinant of carcinogenesis. These findings can have major implications for the design of targeted therapies, and are published in the October issue of Nature Genetics.

Lung Cancer

Wnt Signaling Pathway

Wingless-int (Wnt) proteins are important secreted signaling molecules that regulate numerous interactions in the cell and affect diverse biological processes from embryogenesis to tumorigenesis. When overexpressed, they appear to contribute to a cascade of cellular derangement that results in rampant cell proliferation and failure of defective cells to commit suicide.  Aberrant activation of Wnt signaling is strongly implicated in cancers such as non-small cell lung cancer (NSCLC, the most common type of lung cancer), mesothelioma, and esophageal, colorectal, and nasal‑pharyngeal carcinomas.  For the past nine years, the Thoracic Oncology Laboratory has intensely investigated the role of the Wnt signaling pathway in cancers that afflict the majority of patients in the Thoracic Oncology Clinic.

Integral to our translational ("benchtop-to-bedside") focus is the development of novel therapeutic agents targeting Wnt signaling. The Disheveled (Dvl) molecule, a key mediator of Wnt signaling and one the lab has shown to be overexpressed in NSCLC, is another key target.  When downregulated, Dvl appears to inhibit cell proliferation and selectively induce apoptosis in lung cancer.  Wnt activation appears to be mediated by PDZ domain interactions between Dvl and Frizzled (Frz) proteins, and it is believed that inhibiting these interactions in cancers overexpressing Dvl will prove an attractive targeted therapy strategy in NSCLC patients.  As such, the lab is developing an assay to inhibit PDZ domain interactions between Dvl and Frz proteins.

Previously, research in the Thoracic Oncology Laboratory demonstrated that hypermethylation silencing in the promoter region of two natural Wnt antagonists, Secreted Frizzled Related Protein (sFRP) and Wnt Inhibitory Factor 1 (WIF-1), is associated with aberrant Wnt pathway activation.  Conversely, functional restoration of sFRP and WIF-1 selectively induces apoptosis and suppresses tumor growth in cell lines associated with cancers of interest.  The working hypothesis is that the methylation status of Wnt antagonists could serve as a diagnostic marker for cancer, and the presence of such silenced antagonists in collections of tumor specimens obtained through the Thoracic Oncology Program is therefore being evaluated. The lab is also evaluating the anti-tumorigenic effects of recombinant WIF‑1 proteins in vitro and in mouse xenograft models of human cancer. 

Bronchioloalveolar Carcinoma

The Thoracic Oncology Laboratory also studies bronchioalveolar carcinoma (BAC), a subtype of NSCLC. BACs occur disproportionately in women and non smokers, respond to therapy in a highly nuanced way, and have a more favorable clinical prognosis. In 1999 the World Heath Organization (WHO) restricted the definition of BAC to its classic or "pure" form, a non invasive lung cancer growing along the alveolar walls of the lung. This pure form is relatively rare, comprising perhaps 3-4% of lung cancers, but the vast majority of BACs in clinical practice are mixed tumors exhibiting varying degrees of invasiveness. One of our group's goals is to identify aberrant Wnt signaling in BAC and to develop novel therapies against these anomalies. Preliminary data from work done here show marked overexpression of Wnt family genes in BAC tumor samples compared with in normal tissue. Microarray analyses are being used to measure the expression of Wnt signaling pathway genes in over 100 surgically resected BAC specimens. Several Wnt genes and signaling components are expected to be overexpressed as has been shown in other chemotherapy resistant cancers. Lab members are also working on correlating Wnt expression with disease stage or severity, with patient characteristics such as gender and smoking status, and with clinical endpoints including progression and survival. Furthermore, Wnt antibodies and RNA inhibitors in culture are being tested to determine their ability to kill BAC tumor cells, and novel targeting agents are being developed based on the most responsive compounds. Other interests in the lab involve correlating epidermal growth factor receptor (EGFR) mutations with histopathologic tumor characteristics in BAC, determining if BAC features predict long term survival in surgical patients, and establishing if mutations in the ERBB2 tyrosine kinase domain are associated with clinical characteristics of BAC.

Cancer Stem Cells

Abundant evidence suggests that Wnt signaling is active in both stem cell self-renewal and malignant proliferation of immature tumor cells.  This notion is particularly interesting in light of recent evidence supporting the existence of cancer stem cells.  So-called cancer stem cells are distinct populations of cells found within tumors and possess features typical of adult stem cells such as self-renewal and capacity for differentiation.  Cell surface markers specific to lung cancer stem cells have been identified in murine models, and our group is looking to discover lung cancer stem cells in human tumors.  After performing both in vitro and in vivo (mouse model) cell culture from surgically resected human lung tumors, we are using cell sorting methods to identify human lung cancer stem cells.

Future drug development will seek to identify compounds that are toxic to these cancer stem cells.  We recently proposed to screen for small molecule inhibitors of Wnt-2 transcription through our proprietary Wnt-2 reporter assay.  The results of this screen, we believe, will lead to therapies not only against terminally differentiated cancer cells-the traditional target of such agents-but also cancer stem cell populations, both driven by aberrant Wnt activation. In conjunction with several of our international collaborators, our lab has published a number of important review papers on the topic of cancer stem cells, and we are currently contributing to a book on the subject as well.

Inflammation in Lung Carcinogenesis

In collaboration with Dr. Lisa Coussens, a world-renowned expert on inflammation and Professor in the UCSF Department of Pathology, the Thoracic Oncology Laboratory is studying the role of inflammation in lung and esophageal carcinomas.  It is well-established that chronic inflammation contributes to the development of cancer.  Many studies have demonstrated that inflammatory leukocytes promote epithelial cancers by providing growth and survival factors to initiated cells and contribute to tissue remodeling and angiogenesis.

Hence, physiological processes necessary for tumor development (enhanced cell survival, tissue remodeling, and angiogenesis) are regulated, in part, by leukocytes and the soluble mediators they deliver. Molecular mechanisms mediating the dialogue between infiltrating immune cells with initiated epithelia have yet to be well-characterized. Moreover, the degree to which these interactions alter stem cell niches in neoplastic environments has not been explored.

Several cancers are likely associated with inflammation.  A majority of lung cancers are associated with tobacco smoking, and as such may arise from chronic inflammation and irritation due to smoke exposures.  Esophageal cancer often arises as a complication of chronic gastroesphageal reflux disease, and mesothelioma--cancer of the pleura, the thing covering of the lungs--is commonly associated with asbestos exposure.

In the long term, the group would like to define the lineages of functionally significant immune cells that potentiate cancer development and determine which lineages, if any, regulate Wnt and Hh signaling in lung and esophageal epithelia.  Moreover, the lab will investigate the role of the Wnt and Hh pathways in the continuum of carcinogenic events marking the progression esophageal cancer, and establish if, in regulating Wnt and Hh signaling, immune cells confer stem cell niche autonomy to initiated epithelial cells and thereby enhance tumorigenic potential in lung and esophagus.  Such knowledge will be crucial in developing new therapeutic targets, devising novel therapeutic agents, and identifying molecular markers associated with increased risk of dysplastic and malignant progression.

Esophageal Cancer and Barrett's Esophagus

Esophageal adenocarcinoma (EAC) is now the fastest growing malignancy in the Western world. It is extraordinarily difficult to treat and has an overall five-year survival of less than 15%. In recent years, there has been an alarming increase in the incidence of esophageal cancer, this against a backdrop where death rates from other cancers are dropping across the United States.

EAC is thought to result from inflammation caused by chronic gastroesophageal reflux disease (GERD). EAC is preceded by the replacement of the normal squamous epithelium in the distal esophagus with a metaplastic columnar epithelium, otherwise known as Barrett's esophagus (BE-a pre-malignant condition that arises as a complication of chronic GERD). In BE carcinogenesis, the early metaplastic epithelium develops sequentially into low-grade dysplasia, high-grade dysplasia, early adenocarcinoma, and finally invasive cancer. However, the pathogenesis of EAC is poorly understood, and the molecular mechanism by which esophageal epithelial cells undergo neoplastic transformation is largely unknown. That being said, the progression of EAC provides an exquisite multi-step model for the investigation of the intermediate events in carcinogenesis.

We have already shown that the Wnt signaling pathway is activated during BE carcinogenesis. Other studies have demonstrated that Hh is similarly activated in BE and esophageal carcinogenesis. Notably, both these Wnt and Hh activations occur early in the neoplastic progression of BE. Because BE is an inflammatory condition resulting from chronic GERD, the group hypothesizes that infiltrating immune cells may activate the Wnt and/or Hh signaling cascades in pre-malignant epithelial cells, potentiating neoplastic transformation. The group proposes to elucidate the role played by the Wnt and Hh pathways in BE carcinogenesis, to discover new molecular markers of increased EAC risk in BE patients, and to investigate whether immune cells, resulting from chronic inflammation, may be implicated in tumorigenesis through regulation of the Wnt and Hh pathways.

Mesothelioma

Mesotheliomas are tumors of the lining (pleural) cells of the thoracic cavity, pericardium, and peritoneum and are most commonly associated with prolonged asbestos exposure. Mesothelioma is a relatively uncommon but inexorably fatal carcinoma affecting about 3,000 new patients in the United States annually. Diffuse malignant mesothelioma (MM) comprises about 75% of mesotheliomas diagnosed. Despite advances in cancer treatment, the median survival rate remains low, and most patients die within 10-17 months of their first symptoms. The pathogenesis of the disease remains poorly understood, and the molecular mechanisms by which mesothelial cells undergo neoplastic transformation are largely unknown. The majority of studies on the pathogenesis of MM have focused on asbestos as the primary causative agent.

Although the pathogenesis of asbestos-induced malignancy is not well understood, current evidence suggests that the physicochemical properties of asbestos fibers play an important role in the initiation of an inflammatory response that in turn promotes epithelial cancers by providing growth and survival factors to initiated cells and contributes to tissue remodeling and angiogenesis. Thus, mesothelioma provides yet another platform to study the correlation between inflammation and cancer. Although there is a strong correlation between mesothelioma and asbestos exposure, up to one third of cases have no known association with asbestos-making predictions about the future prevalence of the disease difficult. Interestingly, simian virus-40 (SV40) antigens have been indicated as potential co-carcinogens or alternative carcinogens in MM, which may account for an increased susceptibility among patients with mesothelioma or perhaps a portion of cases with no known asbestos exposure. We are part of a nine-laboratory multicenter investigation into the correlation of mesothelioma with SV40.

Data from our lab have shown that the Wnt signaling pathway is activated in MM as evidenced by increased expression of cytosolic/nuclear beta-catenin and c-Myc, two downstream target genes in the activated pathway. Moreover, we have shown that sFRPs, potential endogenous inhibitors of Wnt, are down-regulated in MMs, and that Dvls, key mediators of the pathway, are overexpressed-both observations hitherto unseen in MM. In addition to illuminating the molecular mechanisms of Wnt signaling in MM, we are also developing molecular therapies to treat this malignancy.

We have demonstrated that WIF-1 expression is down-regulated in both MM cell lines and primary tissue when compared to normal mesothelial cell lines and adjacent pleura, respectively. Our data also suggest that WIF-1 silencing is an important mechanism underlying the constitutively activated Wnt signaling in mesothelioma, and therapies targeting inhibition of the Wnt pathway through WIF-1 might be promising for future treatment of MM.

Furthermore, our lab has discovered that up-regulation of Wnt2 protein is a common event in MM, and we have shown that inhibition of Wnt2 induces apoptosis in the diseased cells. We were able to achieve substantial inhibition of Wnt2 in mesothelioma cell lines using a combination of Wnt2 antibody and Alimta, one of the current standard MM treatments. We thus propose that inhibition of Wnt2 is of therapeutic interest in the development of more effective treatments for MM. In another effort to stymie MM oncogenesis, we were able to down-regulate the expression of Dvl using a reformed type of small interfering RNA (siRNA), stealth RNAi. Dvl stealth RNAi down-regulated the expression of Dvl in mesothelioma cells and induced cell cycle aberration causing suppression of cell growth. Moreover, Dvl stealth RNAi in combination with the chemotherapeutic agent cisplatin suppressed cell growth synergistically. Our group has, additionally, discovered a small molecule inhibitor of Dvl, which inhibits Wnt signaling in lung cancer, and we are currently testing its efficacy in the treatment of mesothelioma.

Despite environmental restrictions on asbestos use and exposure, the 30-40 year latency of this highly lethal disease will not result in a decline in incidence for several more decades. In the mean time, study into its pathogenesis and treatment are vital to improving the lives of patients and elucidating mechanisms of cancer in general.

Expression Profiling and Diagnostics Development

Another research focus in the Thoracic Oncology Program is on gene expression-the output of gene products within a cancer cell. A common measure of gene expression is the level of messenger RNA (mRNA) being produced. An mRNA is a molecule that carries the blueprint for production of cellular proteins. The amount of mRNA present suggests what genes are active within the cell. The unique pattern of gene activity serves as a "genetic signature" that can be correlated with clinical outcomes and can thereby drive treatment decisions.

An interesting feature of lung cancer is that patients who survive the illness for at least three years without recurrence tend not to succumb to it at all. Although histopathological data are unable to distinguish between the more aggressive form of the cancer and its milder counterpart, we proposed that there is a genotypic distinction between the two forms. Quantitative PCR (qPCR) is a method by which levels of mRNA can be measured relative to a "housekeeping" gene--a gene whose mRNA levels are expressed consistently across tumor and normal cells. We recently completed a large-scale microarray and qPCR study comparing specimens from Stage I adenocarcinoma patients with no neoadjuvant chemotherapy who survived the three year mark without recurrence to similar patients for whom the disease proved lethal to determine whether any genes commonly overexpressed in cancer are specifically implicated in one group of patients versus the other. We found a distinct four gene signature overexpressed in patients with the more aggressive lung cancer. Based upon these results, we hope to develop a clinically relevant assay correlating long-term patient survival with the mRNA expression levels of these four genes of interest. Such an assay might aid in developing a more individualized course of treatment, perhaps avoiding physiologically taxing chemotherapy and radiation treatments for those patients who fit the profile for milder disease. Our group aims to develop similar expression assays to correlate staging and gene expression in other thoracic carcinomas.

Lung Cancer System Genetics

The Thoracic Oncology Lab has, in collaboration with UCSF's Dr. Allan Balmain, Ph.D., FRSE., one of the world's leading molecular geneticists, embarked upon an ambitious project that seeks to make possible an individualized approach to lung cancer therapies.  To do so, the tumor of each patient in the study will be characterized at the molecular level to pinpoint the diagnosis, and consequently the best treatment. The processing of these tumors utilizes state-of-the-art technology from leaders in the analysis of complex genetic information for biomedical purposes.

Discoveries of cancer susceptibility genes have already been made in breast cancer (BRCA genes 1 and 2) and in colon cancer (abnormal Rb gene in HNPCC) but not in lung cancer.  The focus here is the analysis of inherited or germline DNA, DNA present in every cell in the body at birth. The project's goal is to identify small variations in the germline, known as "single nucleotide polymorphisms" or SNPs, by analyzing the normal tissue of lung cancer patients. When SNP analyses are overlaid on clinical data, insights can be gained as to which individuals are at elevated risk for developing the disease.  The Thoracic Oncology Group is performing SNP analysis using Affymetrix Molecular Inversion Probes (MIP).  This work has clear relevance to early detection research because it can help define the population of individuals who should be screened.

As a person ages, the accumulation of "somatic" mutations in cells can reach a tipping point and result in the development of lung cancer. In genetics, "somatic" refers to cells or tissue in the body that reside outside the germline, cells that are constantly being regenerated. Somatic mutations can result from environmental causes, i.e. smoking, air pollution, or radon exposure, or can occur sporadically such as when cells malfunction during gene replication. Known mutations and the pathways within which they reside are logical therapeutic targets for discovery. When paired with clinical outcome data, mutations can serve as diagnostic indicators, predictors of survival, and biomarkers for tumor aggressiveness. Mutations can also predict response to therapy, i.e. presence of EGFR mutations predicts response to the cancer drug Tarceva.   The group is employing Affymetrix Mismatch Repair Detection (MRD) and Affymetrix Mutation Sorters (MS) to better characterize lung cancer tumors.

Extra copies of identical genes and/or missing stretches of DNA in lung cancer tumors may be important even when no mutations are involved. CNVs (copy number variations), like mutations, appear to be useful as prognostic biomarkers and predictors of response to therapy. In non-small cell lung cancer (NSCLC), for example, some scientists believe a patient's EGFR copy number is more clinically significant than the EGFR mutation.  Variation in gene copy numbers is also being explored by microarray-based comparative genomic hybridization (a-CGH) using Affymetrix Molecular Inversion Probe (MIP). 

Another aspect of the project focuses on gene expression-the output of gene products within a cancer cell. A common measure of gene expression is the level of messenger RNA (mRNA) being produced. mRNA is a molecule that carries the blueprint for production of cellular proteins. The amount of mRNA present suggests what genes are active within the cell. The unique pattern of gene activity serves as a "genetic signature" that can be correlated with clinical outcomes and can thereby drive treatment decisions.

Gene or DNA sequencing is the process of defining the contours of a stretch of DNA as a distinct subunit, i.e. a gene, and then mapping its function.  Some of the work above will generate clues as to which regions in the cell might contain genes important in lung cancer. The investigators will then sequence genes already known to be active in the disease as well as any others discovered during the research.

To date, there have been no attempts to link both germline and somatic approaches to lung cancer in a concerted systems approach. Other labs have concentrated on one or two of these platforms at most. This project is unique because it leverages all platforms, combining analysis of germline SNPs with knowledge of somatic mutations, gene copy number abnormalities, and gene expression changes in patients' tumors, all with the goal of developing predictive molecular biomarkers. By using multiple datasets derived from a patient's lung tumor and normal tissue, a more accurate readout of the tumor becomes possible.

The resulting data will be mined using advanced computational methods and mathematical algorithms pioneered at UCSF. These datasets will provide an unprecedented opportunity to develop new diagnostics, biomarkers for risk assessment and prognosis, and novel combinations of therapeutic targets. The ultimate objective is to establish the wiring diagram of the lung cancer cell-the network of genetic variants and their expression patterns that influence individual lung cancer susceptibility, risk of progression, and response to therapy.

Cancer treatment is rapidly proceeding towards the era of personalized medicine where treatment is based on the distinctive molecular characteristics of a patient's tumor. The data from this project will be used to construct networks that will include gene expression profiles of lung tumors. These networks will capture the complexity of somatic events intrinsic to the tumor layered over the genetic background that is inherent to the individual. This knowledge will help to more accurately predict disease outcome so that patients at high risk of relapse will receive the most aggressive treatment. It will also allow patients to receive novel combinations of therapies that will afford maximum treatment.

The research is supported by a generous grant from The  Bonnie J. Addario Lung Cancer Foundation, the nation's largest philanthropy devoted exclusively to eradicating lung cancer, through research, early detection, education, prevention, and treatment.