Development of Models to Study Breast Cancer In Vivo
Advances in the development of new therapeutics for cancer treatment have been hampered by lack of suitable preclinical in vivo models for drug testing. Two problems with current models may account for this. First, it has become apparent that the microenvironment of tumor cells is critical for cancer development and progression and that careful attention to reproducing the microenvironment in animals as it exists in the patient is critical for effective drug testing. Up until recently, preclinical drug testing has been carried out using human cell lines as xenografts in immunocompromised mice. The results have been less than satisfactory in that tumors produced by these lines grow slowly and rarely metastasize as expected. We believe that the mismatch between the human xenograft and the murine host is responsible in large part for the problem. This is an intractable problem that will, in all likelihood, be very difficult if not impossible to resolve. A second problem with these systems is the use of immunocompromised mice. Recent studies have shown that inflammation and the immune system play critical roles in tumorigenesis and metastasis. Signals produced by inflammatory and immune cells in response to the tumor have substantial effects on tumor cells and their interactions with the host. For these reasons, we have focused on the use of syngeneic mouse xenografts models and genetically engineered mice for our studies.
For development of syngeneic xenograft models, we have been working with a set of five genetically related mouse mammary tumor cells lines that include the metastatic 4T1 cell line. All five lines were obtained from a single spontaneously arising mammary tumor and differ in tumorigenic and metastatic characteristics. We have modified these lines for biophotonic imaging to facilitate the study of their growth and dissemination characteristics in vivo. We have also selected imagable variant 4T1 cell lines and pools that differ in propensity for metastasis to organ sites affected in breast cancer in human patients (i.e., lungs, liver, bone and brain). Gene expression profiles for these cell lines and pools have revealed a number of potential drug targets for therapeutic intervention of metastasis in late stage breast cancer and important new information regarding the networks of genes involved in regulating metastasis to specific organ sites (Figure 1).
Figure 1. Factors Released from Metastatic 4T1 Mammary Tumor Cells Induce Production and Recruitment of Host Inflammatory Cells to the Primary Tumor. The results shown in the figure were obtained by analysis of gene expression profiles obtained for the metastatic 4T1 cell line and two non-metastatic sister lines. Genes up-regulated in the 4T1 cell line are shown in red and down-regulated genes are shown in blue. Other genes involved in the pathways depicted are shown in green.
Epithelial-Mesenchymal Transition (EMT) and its Role in Late Stage Breast Cancer
In analyzing variant cells that metastasize to lungs more effectively we were able to identify the factor responsible for the more aggressive metastatic characteristics. The factor, called SIP1/Zeb2, is known to induce a process known as epithelial-mesenchymal transition (EMT) which occurs during early embryonic development. The transition involves conversion of immobile epithelial cells to ones with mesenchymal properties including increased motility and invasiveness characteristic of metastatic cells (Figure 2.).
Figure 2. Expression of EMT Markers E-Cadherin and Vimentin in Parental 4T1-12B and Variant L1-c12b Cell Lines. Parental (left) and variant (right) cells were stained with antibodies specific for E-cadherin (green) and vimentin (red). Cell nuclei were stained with DAPI (blue). Note increased vimentin and decreased E-cadherin staining indicative of EMT for the highly metastatic variant L1-c12b cell line.
We are also able to show that EMT induced by SIP1/Zeb2 is regulated by epigenetic changes produced by the unique conditions that exist within the microenvironment of the tumor and are actively involve in identify the nature of these changes.
Functional Genomics: A Novel Approach for Identification of Drug Targets for Breast Cancer
Previously, drug targets for cancer treatment have been identified on a gene by gene basis which has been laborious and time consuming. More recently, gene expression analysis using microarrays has produced gene expression signatures of prognostic value that have aided in identification of candidate genes for validation of drug target. Although microarray analysis provides a high throughput means of identifying genes whose expression is changed in more aggressive tumors and cell lines, most of the genes that are identified do not play a direct role into tumorigenesis or metastasis and have no therapeutic value. Functional genomics screens provide a more direct way of identify genes with therapeutic value since candidate genes are identified on a functional basis and will be more likely to participate directly in the process for which they are screened. All functional genomics screens that have been carried out to date have been done with tumor cell lines in vitro under artificial conditions. As a new paradigm for identification of cause causing genes, we have developed a novel genome wide functional genomics approach which we refer to as “Functional Genomics in Vivo” (Figure 3).
Figure 3. Identification of siRNA Sequences that Facilitate Metastasis of 4T1-12B Cells to Bone, Brain and Liver. A 4T1-12B siRNA cell library was administered to BALB/c mice via cardiac injection. After two weeks, siRNA sequences in bone, brain and liver were amplified by PCR and those enriched in the tissues were identified by hybridization to Affymetrix GeneChip arrays. Log plots of probe signals for the DNA amplified from the cell library vs. the tissues of interest are shown. siRNA with signals enriched >3-fold (i.e., those above the diagonal lines) for the duplicate experiments shown were analyzed further for the ability to facilitate metastasis.
The approach involves the uses of pooled siRNA libraries to augment the ability of cancer cells to grow or metastasize in vivo. Using this approach we have performed genome wide screens for genes that facilitate metastasis to three organs affected in breast cancer: liver, bone and brain and identified several new metastasis suppressor genes. Now that the utility of the approach has been demonstrated the technology is available for identifying new classes of oncogenes and metastasis suppressor genes that will help elucidate the molecular and pathological basis of late stage breast cancer and for identifying new targets for diagnosis, prognosis and treatment of the disease.