Isaiah J. Fidler, D.V.M., Ph.D.
R. E. “Bob” Smith Distinguished Chair in Cell Biology
Professor of Cancer Biology and Urology
Director, Cancer Metastasis Research Center
Operational Director, Smith Research Building
Research in my laboratory continues to focus on the biology and therapy of cancer metastasis. Most recently, we have focused on the development and progression of brain metastasis. A large number of cancer patients develop metastasis to the brain. For untreated patients, the median survival is one to two months, and conventional radio-chemotherapy can extend the median survival to four to six months. The outcome of metastasis in general and brain metastasis in particular depends on the interaction of specific metastatic cells with host factors in the organ microenvironment (the “seed and soil” principle).
Histological examination of clinical specimens of human breast, lung, melanoma and colon brain metastases demonstrates that the lesions are surrounded and infiltrated by activated astrocytes expressing glial fibrillary acidic protein (GFAP). GFAP-positive astrocytes are also associated with experimental brain metastases produced in mice by lung, brain, melanoma and colon cancers. We isolated astrocytes from the brain of the “ImmortoMouse” and established the cells in culture. Multiple in vitro studies conclude that astrocytes cocultured with tumor cells protect the tumor cells from chemotherapeutic agents (Taxol, VCR, VBL, 5-FU). Establishment of a gap junction between astrocytes and tumor cells is required for this chemoprotection. Coculture of tumor cells with other tumor cells or fibroblasts does not protect the cells from chemotherapeutic drugs. Microarray experiments for cross-species hybridization (human tumor cells cocultured with mouse astrocytes or mouse fibroblasts) identified upregulation of survival genes in tumor cells cocultured with astrocytes, but not with fibroblasts. Significantly, these survival genes were uniformly upregulated in different tumor cells (breast, lung) cocultured with astrocytes.
The role of astrocytes in physiology is to supply glucose and oxygen to neurons to assure their survival. Unfortunately, tumor cells that can proliferate in the brain parenchyma exploit astrocytes for these same functions. Since the development of metastases depends on the interactions of tumor cells with host factors, treatment of brain metastasis must be directed against both the tumor cells and the organ microenvironment.
Ongoing Research Support
5-P30-CA 16672 (Fidler, PI) 06/30/08-07/01/09 – 5%
CCSG Histopathology Laboratory
To provide histologic services to investigators at the MD Anderson Cancer Center.
Fidler, PI 01/01/09-12/31/09 – 20%
Actelion Pharmaceuticals, Ltd.
“The Role of Endothelin in Progressive Growth and Metastasis of Neoplasms”
To investigate the role of endothelin in the progressive growth and metastasis of neoplasms.
Selected Recent Key Publications
Fidler IJ, Kim S-J, Langley RR. The role of the organ microenvironment in the biology and therapy of metastasis. J Cell Biochem 101: 927-936, 2007.
Nakamura T, Kuwai T, Kitadai Y, Sasaki T, Fan D, Coombes KR, Fidler IJ. Zonal heterogeneity for gene expression in human pancreatic carcinoma. Cancer Res 67: 7597-7604, 2007.
Nakamura T, Kuwai T, Kim J-S, Fan D, Kim SJ, Fidler IJ. Stromal MMP-9 is essential to angiogenesis and progressive growth of orthotopic human pancreatic cancer in parabiont nude mice. Neoplasia 9(11): 979-986, 2007.
Kuwai T, Nakamura T, Kim SJ, Sasaki T, Kitadai Y, Langley RR, Fan D, Hamilton SR, Fidler IJ. Intratumoral heterogeneity for expression of tyrosine kinase growth factor receptors in human colon cancer surgical specimens and orthotopic tumors. Amer J Pathol 172(2): 358-366, 2008.
Weihua Z, Tsan R, Huang W-C, Wu Q, Chiu C-H, Fidler IJ, Hung M-C. Survival of cancer cells is maintained by EGFR independent of its kinase activity. Cancer Cell 13: 385-393, 2008.
Kuwai T, Nakamura T, Sasaki T, Kitadai Y, Kim JS, Langley RR, Fan D, Wang X, Do K-A, Kim SJ, Fidler IJ. Phosphorylated epidermal growth factor receptor on endothelial cells in colon cancer is the primary target for therapy with tyrosine kinase inhibitors. Neoplasia 10: 489-500, 2008.
Menashe Bar-Eli, Ph.D.
Professor of Cancer Biology
The focus of my laboratory is to study the molecular biology of melanoma metastasis. The molecular changes associated with the transition of melanoma cells from radial growth phase (RGP) to vertical growth phase (VGP, metastatic phenotype) are not very well defined.
One tumor cell property essential for metastasis is the expression of cell surface adhesion molecules that mediate cell-to-cell or cell-to-matrix interactions. MCAM/MUC18 is a 113-kDa cell surface glycoprotein originally discovered on malignant melanoma cells. Our laboratory has shown that the expression of MCAM/MUC18 by human melanoma cells directly correlates with their metastatic potential in nude mice, and that so far no MCAM/MUC18-negative cell lines have been found to be metastatic. Furthermore, enforced expression of MCAM in MCAM-negative cells rendered them highly tumorigenic and increased their metastatic potential in vivo.
While upregulation of the MCAM gene consists of gain of function, we have recently demonstrated that the progression of melanoma is associated with loss of c-KIT. We found that re-expression of the c-KIT receptor in highly metastatic cells inhibited their tumorigenic and metastatic potential in nude mice. Moreover, the ligand for c-KIT, SCF, inhibited the growth and induced apoptosis in melanoma cell lines expressing c-KIT but not in melanocytes, under both in vitro and in vivo conditions. These data imply that melanoma cells expressing c-KIT may allow malignant melanoma cells to escape SCF/c-KIT-mediated apoptosis, hence contributing to tumor growth and eventually metastasis.
In other studies we have shown that transition of melanoma cells to the metastatic phenotype is associated with increased activity of the metalloproteinase (MMP-2) that can be regulated by IL-8.
We found that the highly metastatic cells do not express the transcriptional factor AP-2. Because all the above three genes (MCAM, c-KIT and MMP-2) contribute to the metastatic phenotype, and since all the three are regulated by AP-2, and since other important genes involved in the progression of human melanoma such as E-cadherin, HER-2, VEGF, FAS/APO-1, bcl-2 and Kai-1, are also regulated by AP-2, we hypothesized that loss of AP-2 could be a "major switch" in the development of malignant melanoma. We were able to demonstrate that loss of AP-2 expression resulted in loss of c-KIT and upregulation of MUC18. Furthermore, re-introduction of AP-2 into the highly metastatic cells caused inhibition of tumor growth and significant reduction in their metastatic potential in nude mice. Using cDNA microchip, we recently identified the thrombin receptor PAR-1 to be a target for regulation by AP-2. We found that loss of AP-2 resulted in overexpression of PAR-1 in metastatic melanoma cells, which in turn contributes, to invasion and metastasis. Based on our data, we propose the notion that AP-2 serves a key regulator of melanoma metastasis.
In other studies we have recently demonstrated that dominant-negative CREB can inhibit growth and metastasis of melanoma via regulation of MMP-2 and MUC18 gene expression. In addition, we also demonstrated that CREB and its associated proteins act as survival factors for human melanoma cells, thus, providing a mechanism, for the first time, on how overexpression of CREB in melanoma cells may contribute to the acquisition of the metastatic phenotype.
In recent studies, we found that PAF, which is secreted by cells in the tumor microenvironment, stimulates the phosphorylation and activation of CREB in metastatic melanoma cells. Based on our data on the involvement of MUC18 and IL-8 in the progression of human melanoma, we recently developed two fully humanized antibodies to target these molecules. Treatment of melanoma bearing nude mice with fully human IL-8 (ABX-IL-8) or fully human anti-MUC18 (ABX-MA1) reduced melanoma growth and inhibited their metastatic potential.
Douglas D. Boyd, Ph.D.
Professor of Cancer Biology
My laboratory investigates the transcriptional control of genes that contribute to the spread of cancer. One of these genes, the urokinase receptor (u-PAR), is overexpressed in colon cancer. We are using transgenic mice and DNaseI hypersensitivity assays to identify novel regions driving tissue-specific u-PAR expression both in healthy mice and animals genetically induced for colon cancer, the latter allowing an analysis of transcriptional requirements in colon cancer in vivo.
We also study the regulation of the expression of the MMP-9 metalloproteinase since this gene product also contributes to cancer spread. We reported that MTA1, a multiprotein complex with chromatin remodeling activity, binds directly to the MMP-9 promoter and represses its expres-sion. To further investigate the role of the chromatin environment, we have employed recombi-nant technology to genomically integrate MMP-9 promoter-reporter constructs in a site-specific manner. This technology creates a refined system to investigate the contribution of the chromatin environment to MMP-9 gene expression. Moreover, we are currently using this system to screen compound libraries for candidate agents that repress MMP-9 expression.
For both MMP-9 and u-PAR transcription, we are employing expression cloning and gene trap strategies to identify novel regulators of both u-PAR and MMP-9 expression, and, to date, we have identified two hitherto unknown regulators of their expression, i.e., KLF4 and SM22, respectively.
Finally, we have identified a novel gene (ZNF306) encoding a transcription factor that contributes to colon cancer progression based on our data with human cancer cells forced to over-express, or conversely, knocked down for this DNA-binding protein. Presently, we are using animal models to determine if the ZNF306 transgene causes malignancy in APC +/Min mice which normally only form adenomas and whether mice null for this zinc finger protein become resistant to azoxymethane-induced carcinogenesis. Using cyclic amplification and selection of targets (CAST-ing), we have identified a ZNF306 DNA-binding site and we are currently determining if ZNF306 is a “master” regulator of an expressed gene program that contributes to the progression of this malignancy.
Colin P. N. Dinney, M.D.
Professor of Urology
Professor of Cancer Biology
I will continue my research into the regulation of bladder cancer metastasis by EGFR signaling pathways in collaboration with Drs. Menashe Bar-Eli and David McConkey of Cancer Biology. Together with Drs. David McConkey of Cancer Biology and Randall Millikan of GU Medical Oncology, I am evaluating the role for interferon therapy for bladder cancer, and specifically how interferon modifies death receptor expression. We are also collaborating on studies to develop proteasome inhibitors as therapy for metastatic bladder cancer. As part of collaboration with Dr. William Benedict of GU Medical Oncology and and Dr. Liana Adam of Urology, I am evaluating the development of interferon gene therapy for superficial bladder cancer. This will be translated into a clinical trial of adenoviral mediated interferon gene therapy.
Lee M. Ellis, M.D.
Professor of Cancer Biology
Chair ad interim, Department of Cancer Biology
Professor of Surgery
William C. Liedtke, Jr., Chair in Cancer Research
The overall goal of our laboratory is to investigate mechanisms of tumor growth and metastasis of gastrointestinal (GI) malignancies with a focus on angiogenesis. The primary tumor system under study in our laboratory is metastatic colorectal cancer, although we also study other tumor types including pancreatic adenocarcinoma, gastric cancer and carcinoid tumors.
For the past 14 years, our laboratory has investigated the role of VEGF in tumor growth, metastasis and angiogenesis. We have recently focused our efforts on determining the role of VEGF receptors on tumor cells. We are currently investigating the role of VEGF receptor-1, neuropilin-1 and neuropilin-2 on colon cancer cells. Investigations in these areas may help elucidate mechanisms of action of anti-VEGF therapy, which is currently approved for several tumor types by the FDA for patients with metastatic disease. In addition, we are currently investigating angiogenic mechanisms in specific organs relevant to colon cancer metastasis.
Continuing along our translational themes, we are also investigating mechanisms of resistance to standard chemotherapy for colon cancer. We have established chemotherapy resistant colon cancer and gastric cancer cell lines. Initial investigations demonstrated that oxaliplatin resistant colon cancer cells led to epithelial to mesenchymal transition (EMT).
Although targeted therapy for GI malignancies has demonstrated promise in clinical trials, we strongly believe that it is important to validate new targets for the next generation of anti-neoplastic regimens. We are also investigating the role of cancer stem cells in chemoresistance. Lastly, we have developed a number of neuroendocrine tumor cell lines in order to develop targeted therapies or metastatic carcinoid tumors.
Gary E. Gallick, Ph.D.
Professor of Cancer Biology
Research in my laboratory focuses on of the role of protein tyrosine kinases in promoting cellular survival and angiogenesis, and the result of aberrant activation of these enzymes on tumor progression and metastasis of colon, prostate and pancreatic cancers. Using a combination of in vitro techniques and animal models, we have demonstrated that increases in activity of Src family kinases are common in solid tumors and causative of increased metastatic potential. We are dissecting the signal transduction pathways responsible for these critical events in tumor progression with an emphasis on regulation of Src regulation of pro-angiogenic factors and migration and invasion. In pancreatic cancer, we are focusing on Src-regulated signal transduction pathways leading to constitutive and hypoxia-inducible expression of vascular endothelial growth factor (VEGF). Specifically, Src regulates VEGF transcription through HIF-1α, STAT3 and ID2. We have demonstrated that ID2 is a HIF-regulated gene at the transcriptional level and then regulates HIF-1a stability in an autocrine "feed-forward" loop. Decreased ID2 decreases VEGF expression in vitro, and tumor growth and angiogenesis in orthotopic mouse models of pancreatic cancer in vivo. Roles of ID2 in angiogenesis and its potential as a novel target for cancer therapeutic agents are under investigation.
In prostate cancer, we are determining roles of Src activation in both tumor cells and host endothelial cells in the development of lymph node metastases. These studies use a combination on in vitro and in vivo work to test the hypothesis that Src activation in the tumor cell leads to Src activation in the endothelial cell. We have demonstrated that increased Src activity in the tumor is sufficient to promote lymph node metastases. To examine the role of Src in the host in this process, we have developed src-/- nude mice to compare tumor growth and metastatic potential with "conventional" nude mice.
With respect to migration, we have developed and characterized novel models in which cells selected for increased migration in vitro are greatly increased in metastatic potential in nude mouse models. These biologically selected cells are altered in expression of regulators of the Rac pathway, a "G" protein known to be important in migration. In addition, we are examining mediators of focal adhesion structure and their role in migration and tumor progression. We are the first to demonstrate that actin filament associated protein 110 (AFAP-110), a known Src substrate) is increased in progressive stages of prostate cancer, and by siRNA technology, that this protein is important in tumorigenicity of prostate cancer cells. AFAP-110 regulates integrin stability and Src/Focal Adhesion Kinase (FAK)-mediated signaling, a process we are investigating.
In the last two years, Src inhibitors have reached clinical trial for solid tumors. As a result, we work closely with clinicians conducting these trials to identify both predictive markers for efficacy of Src family inhibitors and markers for success of these inhibitors. We are also involved in testing novel inhibitors in pre-clinical studies. These clinically related studies are highly integrated into our fundamental laboratory work. All the projects in my laboratory involve trainees, including graduate students, postdoctoral fellows, residents and clinical fellows.
Curtis A. Pettaway, M.D.
Associate Professor of Urology
Associate Professor of Cancer Biology
The theme of my research is to define the molecular phenotype of advanced/metastatic prostate cancer in animal models and patient specimens of prostate cancer. The studies performed over the past year build upon our knowledge gained in characterizing the molecular events associated with prostate cancer progression in an orthotopic mouse model of human prostate cancer progression. Expression of matrix metalloproteinases, vascular endothelial growth factor E-cadherin and several other genes have been evaluated in radical prostatectomy specimens and found to be highly associated with advanced prostate cancer, subsequent to radical prostatectomy. Over the past year, we have evaluated the expression of two other genes associated with angiogenesis, basic fibroblast growth factor (bFGF) and interleukin-8 (IL-8). We have correlated expression with Gleason sum and pathologic stage and found a significant association in the case of IL-8 but not bFGF. A preliminary study has also been completed comparing the expression of MMP-2, MMP-9 and E-cadherin expression in prostate biopsies with their corresponding expression in the radical prostatectomy specimens. We have found that the correlation is good and that biopsy gene expression could enhance the prediction of pathologic stage compared with Gleason score, clinical stage or serum prostate specific antigen (PSA) levels. We are currently evaluating racial variation in the expression of these same genes.
A second project has studied the functional effects of overexpression of prostate specific antigen in human prostate carcinoma cells over the past year. Overexpression of PSA subsequent to transfection resulted in enhanced tumor growth but not metastasis. This effect was correlated with a higher proliferative index in cells with high PSA expression. A third project is related to exploring the role of Bcl-2 in androgen independent prostate cancer cell proliferation. We have shown over the past year that Bcl-2 contributes to androgen independent proliferation in LNCaP cells by enhancing the transcriptional activity of androgen-regulated genes. Thus, we describe a novel function for Bcl-2 in prostate cancer progression.
Janet E. Price, D.Phil.
Associate Professor of Cancer Biology
Metastases are the result of interactions between tumor cells and tissue environments mediated through cell-to-cell and cell-to-matrix contacts and through the release of cytokines and growth factors. The primary focus of my laboratory is to investigate how these interactions affect the development of metastases, using nude mouse models of human cancer metastasis, with a special focus on breast cancer.
Breast cancer metastases are commonly found in bone and the brain. Breast cancer is the second most common cause of brain metastasis after lung cancer, with a clinical diagnosis in 10-15% of patients. Bone metastases are found in a higher percentage of patients, up to 90% of women with metastatic breast cancer. Although it is well known that breast cancers can metastasize to the brain and bone, relatively little is known about how these metastases form and the phenotypes of breast cancers that grow in these organ environments. Without such information, the rational design of new therapies to prevent or control the growth of metastases is impossible. In large part, the progress in understanding the biology of breast cancer metastases has been limited by the lack of suitable cell lines and experimental models. We propose to develop experimental models based on the nude mouse models we have already established and to use them to study the pathogenesis of breast cancer metastasis in bone and the brain. For the brain metastasis model. we will focus on the role of angiogenic factors, including vascular endothelial growth factor, and use a model of injection of cells in the intracarotid artery to simulate dissemination to the brain.
Expression array analysis of isogenic variants of the GI101A human breast cancer cell line, which differed in metastatic potential in nude mice, identified a relatively small number of genes differentially expressed in the more aggressive cells. High expression of four genes identified in the array analyses and measured on tissue microarrays of clinical breast cancer specimens was significantly associated with lymph node metastasis, and shorter patient survival was associated with high expression of three of these genes. The results show that the xenograft model can provide information about the biology of human breast cancer metastasis that is relevant to the clinical progression of the disease. The results also demonstrate, as anticipated, that many different gene products expressed by the breast cancer cells may potentially contribute to their metastatic potential. Rather than focusing on the roles of individual genes, current research seeks to identify common pathways that regulate expression of groups of metastasis-associated genes, notably those mediating tumor-stromal interactions.