Biological screen for novel, clinically active peptides in the treatment of cancer
The overall goal of the project is to design a high-throughput screening system for small molecules that will affect the p53 pathway in cells from the central nervous system. Since p53 has been implicated in cell death pathways in both cancer and neurodegeneration, molecules that both increase and decrease p53 activity are sought after.p53 function is both good and bad. It is nature’s most powerful defense against cancer, but the same input to p53 in a normal cell type in a vulnerable cellular compartment, such as stem cells, will lead to aging or degenerative effects in an organism. Compounds that either activate or ameliorate p53 function, therefore, have therapeutic relevance in different disease states. We report here on a high-throughput assay designed to screen for molecules that affect p53 function both positively and negatively in a variety of cell types. In this assay, the green fluorescent protein (GFP) is used as the reporter gene for p53 function so that changes in its function can be observed over time in live cells. Since the promoter reporter gene construct is inserted into a lenti-virus, it can easily be integrated into the genome of any cell type, including growth arrested differentiated cell types. To demonstrate p53 transactivation of the reporter gene, infected cells were treated with etoposide, a chemotherapeutic agent known to stimulate p53, and Nutlin-3, an inhibitor of the interaction between MDM2 and p53. Results obtained by Western blot demonstrated that the protein levels of the reporter gene are induced by both agents. Interestingly, although p53 was induced within 2 hours in a glioma cell line, GFP did not increase before 24 hours. A glioma cell line harbouring the construct was plated at 10,000 cells per well in a 96-well plate in RP 1640 medium without riboflavin and phenol red. After 24 hours, etoposide and Nutlin-3 were added to the medium. Again, a six-fold increase in fluorescence was detected by 24 hours. We also infected tumor cells from a patient glioma as well as normal human fibroblasts. These cells exhibited enhanced fluorescence when treated with etoposide and Nutlin-3. These results indicated that at least some tumor cells that harbor wild type p53 respond to the drugs at the transcriptional level of p53. We were able to observe changes in fluorescence microscopically that went undetected by the plate reader. In this case, we decided to read the same cells by the BD bioimager. The bioimager has the capacity to detect levels of fluorescence in cells and to take qualitative data such as cell shape. We are also able to view fluorescence in individual cells. A 96-well plate was used and cells were plated both in normal and RP 1640 media. Cells were exposed to etoposide and Nutlin-3 and fluorescence was measured at 24 hours. We found the bioimager to be efficient at detecting changes in both fluorescence and qualitative data. The bioimager was used to test extracts from marine bacteria. In this experiment 1:4 dilution of the original extract was placed in triplicate on the cells and read at 24 and 48 hours. We did not observe any changes in fluorescence or cell shape. The function of p53 has not been thoroughly investigated in cells of the central nervous system, especially tumor cells. Here, chemotherapeutic agents are extremely limited in their efficacy. Many of these tumors still contain wild type p53. Our most immediate goals are to search for drugs that activate p53 function in such cells, but we can also screen molecular libraries containing RNAi or cDNA. This assay is biologically based, that is the outcome is linked directly with a known pathway. As personalized profiles may become part of the routine of a patient’s treatment, p53 function in glioma cells could become an important treatment option.
The overall goal of the project is to design a high-throughput screening system for small molecules that will affect the p53 pathway in cells from the central nervous system. Since p53 has been implicated in cell death pathways in both cancer and neurodegeneration, molecules that both increase and decrease p53 activity are sought after.Several goals have been addressed over the year for this project: 1. to design in a lenti-virus a reporter construct that responds to p53 function; 2. to infect various cell types with this construct and assess function; 3. to screen for effective compounds from different molecular libraries to demonstrate utility of the assay. Aim 1: A commercially available construct was introduced into a human glioma cell line, U87, that contains wild type p53. This construct contains a GFP reporter gene under the control of a synthetic p53 binding sequence that represents the most commonly occurring bases at each position of the canonical binding sequence. The idea is to detect changes in fluorescence in response to p53 activity. We have shown response to regulators of p53, nutlin-3 and etoposide, both by Western blot and a high-throughput screen to detect fluorescence on a 96-well plate assay. Problems: One of the problems we have encountered is that the construct has background that is read by the plate reader. This is problematic where p53 activity increases because we need a way to associate increase in fluorescence with cell number. To overcome this first problem, we have designed a new construct that encompasses a p53-binding site from its normal genomic milieu. This construct has also been placed into U87 cells. It has a lower background activity so that when p53 is activated, increases in fluorescence are quite clear on the microplate reader (Victor3; Perkin Elmer, California, USA). Although we tend to prefer the new construct for drug screens focused on cancer, the commercial construct can be used to identify agents that decrease p53 activity that would subsequently be applicable to neurodegenerative disease states. A second problem encountered is that normal medium (10% DMEM) also yields a high level of fluorescence that interferes with signal emanating from the cells. We are addressing this problem in two ways. Firstly, we are using a custom made medium that no longer contains the fluorescing agent (Biological Industries, Israel). In this medium, we have been able to measure unobstructed fluorescence directly from the cells. Secondly, we have tested an alternative system to collect fluorescence data. Imaging equipment (BD Pathway 855 High-Content Bioimager; BD Biosciences, California, USA) is available in the Institutt for biomedisin precisely for high throughput screening. Preliminary results suggest that this would be the preferred method of screening because we can obtain imaging on live cells in addition to fluorescence data. Aim 2: The second major goal was to test function of these constructs in different cell types. The U87 cells are an established cell line in culture that will serve well as a preliminary screen. However we are mostly interested in how cells direct from patient tumor material will respond to compounds screened and whether normal cell types will be affected also. Therefore, we have put the construct into primary glioma cells from a patient tumor as well as a population of normal fibroblasts. By the plate reader assay, the construct responds efficiently in both of these cell types as well. Aim 3: The third goal is to test molecular libraries for compounds that will affect p53. We have had access to a collection of natural product extracts. We have been generously allowed to screen 12 samples purely to demonstrate proof of principle. None of them appears to activate p53 transcriptional properties. However, the screening method with the bioimager works and can be scaled up to screen libraries of our choice. Finally, an additional reagent has been developed in this time period. Once we have some putative p53 inducing compounds, we would like to test them for specificity to the p53 pathway. To test the requirement for p53 function, we will investigate how the same cells respond where p53 has been knocked out. We have put an RNAi construct against p53 into U87 cells in order to do this. We have shown by western blot and functional tests that these cells are now without p53 function. Our focus now is to streamline this process, to use the constructs to identify p53 positive human tumors, and to use molecular libraries to probe its function in an effort to identify new therapeutic reagents.