Skip to main content

Discovery of breast cancer aggressiveness markers using topo-proteomics mapping

Periodic Reporting for period 3 - PROTEOMICAN (Discovery of breast cancer aggressiveness markers using topo-proteomics mapping)

Reporting period: 2018-07-01 to 2019-12-31

Breast cancer is the most prevalent cancer type in women and one of the most common death causes in women worldwide. In the last decades, breast tumors have been classified into 3-4 subtypes, in order to assign the appropriate treatment modality and to determine patient prognosis. However, in many of the cases, patient do not respond to these treatments, or develop resistance and tumor relapse. One of the underlying causes of this resistance is the heterogeneity of the tumors. It is now recognized that most tumors are not actually composed of a single subtype, but are rather a heterogeneous mass of cells, of different clones, each one with its distinct characteristics. These distinct populations are affected by their mutational profiles that evolved with tumor development, and by the microenvironment of healthy cells adjacent to the cancer cells. We hypothesize that characterization of the various clones within single tumors will highlight the key regulators of tumor development. Furthermore, determining which of the clones drive tumor development and progression can suggest the means of eradicating the tumors despite their heterogeneous nature. This study aims to understand tumor heterogeneity in breast cancer, using state of the art proteomics technologies. Mass spectrometry-based proteomics monitors the global profiles of protein changes. Determination of protein levels reflects the cellular phenotype much more closely than genomic approaches, and is therefore expected to reveal tumorigenic mechanisms that cannot be identified by other approaches. To achieve our major research goal, we combine state of the art proteomics, with histopathological analysis of tumors, and advanced computational analysis to create the first of its kind topological protein mapping of breast cancer. The major part of the project separated the tumors based on their histopathological profiles, followed with microscopy-guided laser microdissection of each region, and their mass spectrometry-based proteomic analysis. Overall, we analyze hundreds of tumor regions, to understand the diversity of the proteomic profiles within tumors and between patients. Advanced computational analysis of the proteomic data will suggest which of these clones is more aggressive, and propose potential drugs that can target them. We then examine the aggressiveness of these clones and follow with biological examination of their aggressiveness. Altogether, this research will unravel yet unknown features of breast cancer that are overlooked by routine clinical analyses, and by all previous proteomic and genomic studies. The success of this research can then highlight novel drug targets, and combination therapies that target distinct cancer clones, therefore reduce tumor resistance and increase patient survival.
To address the goals of the project, we selected breast tumors with heterogeneous characteristics, and performed the entire workflow, from histopathological analysis, to state of the art proteomics, on 338 tumor regions. Notably, the largest published proteomic cancer dataset to date includes approximately 100 tumors, and these typically average the entire tumor sample, rather than address the different phenotypes in a microscopy-guided way. Therefore, our results stretch far beyond current studies. Due to the small size of the tumor regions, we had to overcome major challenges to increase the sensitivity of the workflow, and optimize various techniques of sample preparation and quantitative analysis. Furthermore, to increase the throughput of the analysis, we are implementing the first of its kind automated, high throughput, high yield technique for sample preparation. This technique will undoubtedly transform routine proteomics analyses in a large range of applications. We have so far analyzed a partial dataset of the analyzed tumors, aiming to identify patterns related to the heterogeneous nature of the tumors. Initial data analysis of the proteomic profiles showed fundamental differences between tumor regions, even within individual patients. Interestingly, examination of low-grade breast cancer showed dominance of the histological subtype, which was much stronger than the impact of the inter-patient variance. In contrast, examination of high-grade tumors showed lower level of heterogeneity, and larger separation between the patients. The dominant factor in these cases was the expression of the known breast cancer receptors. In each of these comparisons, we identified the proteins and processes that are associated with the individual region characteristics, and suggested cancer regulators that are associated with that region’s characteristics.
Following the proteomic characterization of the cancer cell populations, we aim to investigate which of these regions has more aggressive features that potentially drive progression of the entire tumors. To that end, we have optimized the necessary methodologies that allow extraction of individual cells from tumors, and their growth in culture. Growth rate and invasiveness of these different colonies will reflect the aggressiveness of the extracted cells. Integration with the proteomics of these cells will associate between the proteomic profile and the phenotype. Overall, our results and technological advances within the project so far have set the basis for determination of the contribution of specific proteins to the aggressive nature of various tumor regions in breast cancer. These will all be integrated in the second half of the project.
To achieve our scientific goals, we had to push the proteomic technology to its limit in terms of sample amounts and throughput. We tackled this challenge by combining method development with the biological aims that we are addressing. We improved the yield 4-fold using a novel sample preparation technique, and implemented an automated robot for sample preparation, to increase the throughput. This combination has never been done in proteomics, and it opens new opportunities for large clinical proteomic projects. The combination of our proteomic and biological expertise will provide unique breast cancer topological mapping and open the way to individualized therapy assignment that would lead complete therapeutic response.