Periodic Reporting for period 1 - ProACCT (Proteomic Analysis of Cell communication in Tumors)
Berichtszeitraum: 2022-08-01 bis 2025-01-31
Cancer is an enormous biomedical challenge, in part due to the complexity of cancer cell dynamics. The combination of intrinsic genetic alterations and cues from the tumour microenvironment impacts the cancer phenotype and creates heterogeneous tumours that evolve in space and time. Cancer cells develop distinct phenotypes in diverse tumour regions and different metastatic locations. In addition, cancer progression, metastatic dissemination, and development of therapeutic resistance affect cellular interactions over time. Therefore, understanding tumour dynamics is critical for developing novel therapeutic approaches. Changes in protein expression are central determinants of cancer phenotypes; creating a detailed understanding of proteome dynamics would be an enormous scientific advancement in cancer. However, technological challenges, specifically the challenge of analysing single cells and small groups of cells, have delayed progress along these lines. Here, we push the boundaries of proteomic technologies to analyse single cells and small cell populations using mass spectrometry-based proteomics, aiming to better understand cell-cell communication in tumours and follow how these evolve during cancer progression in distinct metastatic locations. We will also develop spatial proteomic approaches to analyse tumours and metastases at a single-cell resolution to identify cellular communities and their functional interactions.
To achieve these goals, we combine technology development toward deep single-cell analyses, combined analysis of multiplexed imaging and spatial proteomics, and computational developments to elucidate cell-cell interactions.
Our research has already established robust and sensitive proteomic approaches that provide deep proteome coverage of single and small cell populations from mouse models and clinical samples. In addition, we developed a new algorithm to elucidate cellular interactions in the tumour microenvironment. These results form the basis for understanding cancer dynamics. We anticipate that these results will have a significant impact on cancer research and will potentially evolve to the development of novel therapeutic approaches in cancer.
To achieve these goals, we combine technology development toward deep single-cell analyses, combined analysis of multiplexed imaging and spatial proteomics, and computational developments to elucidate cell-cell interactions.
Our research has already established robust and sensitive proteomic approaches that provide deep proteome coverage of single and small cell populations from mouse models and clinical samples. In addition, we developed a new algorithm to elucidate cellular interactions in the tumour microenvironment. These results form the basis for understanding cancer dynamics. We anticipate that these results will have a significant impact on cancer research and will potentially evolve to the development of novel therapeutic approaches in cancer.
The ProACCT project combines the development of single-cell proteomics technologies with the analysis of tumour samples originating from distinct tumour environments. Since the commencement of the project, we have made substantial progress towards all of its aims.
Towards Aim 1: Develop a cutting-edge nano-proteomic pipeline to elucidate tumour proteogenomic diversity. We established a robust single-cell proteomic pipeline, including sample preparation miniaturization, automation and deep liquid-chromatography-mass spectrometry (LCMS) methodologies. We established an automated sample preparation method, which uses nanoliter volumes to increase overall sensitivity. We increased the throughput to enable the rapid processing of more than 1500 tumour cells. These methods allow routine identification of more than 2000 proteins per cell (Figure 1).
Towards Aim 2: Spatial proteomics of cancer microenvironment. We established methodologies that combine multiplexed imaging and single-cell-type proteomic analysis. Spatial proteomic analyses use deep learning-based image processing to automatically direct laser capture microdissection of single cells in their spatial context. To advance this method, we also combine it with Phenocycler multiplex imaging (CODEX), which provides high multiplexing capabilities. Our results show our ability to use the Phenocycler technology on PEN membrane slides and follow with MS-based proteomics that reaches deep coverage from areas of only 25-50 cells from archived tumour tissues (Figure 2).
Towards Aim 3: We established cancer mouse models to elucidate the dynamics of cell-cell communication in time. In accordance with the proposed research, we established mouse models of melanoma and breast cancer metastases to the liver and the lungs. We collect tumour samples and control organs in early and late time points and also compare these metastatic locations to the primary cancer site, namely skin for melanoma and mammary fat pad for breast cancer. Towards elucidation of cell-cell interactions, we isolated 10-14 cell populations from each mouse sample, including cancer and immune cells. We followed with proteomic analysis of each sample on the timsTOF SCP mass spectrometer. Overall, we identified 5000-7000 proteins per cell population, which serve as the basis for the elucidation of cellular dynamics and cell-cell interactions. We developed a novel computational algorithm that determines the strength of cellular interactions based on the protein expression differences between cell types within each tumour and organ, matching the CellTalk database and scoring the strength of interaction. These results will be used to select candidate regulators of cancer growth for further functional validation (Figure 3).
Towards Aim 1: Develop a cutting-edge nano-proteomic pipeline to elucidate tumour proteogenomic diversity. We established a robust single-cell proteomic pipeline, including sample preparation miniaturization, automation and deep liquid-chromatography-mass spectrometry (LCMS) methodologies. We established an automated sample preparation method, which uses nanoliter volumes to increase overall sensitivity. We increased the throughput to enable the rapid processing of more than 1500 tumour cells. These methods allow routine identification of more than 2000 proteins per cell (Figure 1).
Towards Aim 2: Spatial proteomics of cancer microenvironment. We established methodologies that combine multiplexed imaging and single-cell-type proteomic analysis. Spatial proteomic analyses use deep learning-based image processing to automatically direct laser capture microdissection of single cells in their spatial context. To advance this method, we also combine it with Phenocycler multiplex imaging (CODEX), which provides high multiplexing capabilities. Our results show our ability to use the Phenocycler technology on PEN membrane slides and follow with MS-based proteomics that reaches deep coverage from areas of only 25-50 cells from archived tumour tissues (Figure 2).
Towards Aim 3: We established cancer mouse models to elucidate the dynamics of cell-cell communication in time. In accordance with the proposed research, we established mouse models of melanoma and breast cancer metastases to the liver and the lungs. We collect tumour samples and control organs in early and late time points and also compare these metastatic locations to the primary cancer site, namely skin for melanoma and mammary fat pad for breast cancer. Towards elucidation of cell-cell interactions, we isolated 10-14 cell populations from each mouse sample, including cancer and immune cells. We followed with proteomic analysis of each sample on the timsTOF SCP mass spectrometer. Overall, we identified 5000-7000 proteins per cell population, which serve as the basis for the elucidation of cellular dynamics and cell-cell interactions. We developed a novel computational algorithm that determines the strength of cellular interactions based on the protein expression differences between cell types within each tumour and organ, matching the CellTalk database and scoring the strength of interaction. These results will be used to select candidate regulators of cancer growth for further functional validation (Figure 3).
The technological development within the ProACCT project achieved several breakthroughs, which could potentially advance the proteomic and cancer fields.
1. High-throughput proteomics of single cells- Our results show our ability to increase the throughput of single-cell analyses to 1536-well plates. This forms a fourfold increase in the number of samples that can be analysed in a single experiment compared to other published methods. Such an increase significantly impacts the applicability of these techniques to tumour analyses, as lower throughput would not be sufficient to generate robust results.
2. Development of spatial proteomics combined with multiplexed imaging on the Phenocycler (CODEX). We established a protocol for multiplexed staining on membrane-coated slides, which are required for laser microdissection (LCM). This method allows advanced immune profiling of the tumour microenvironment, followed by deep analyses of the proteomes of cells of interest. This approach will provide a much deeper understanding of the dynamic differences that occur in cancer and immune cells in the context of their immediate microenvironment.
3. We comprehensively analysed cancer and immune cell populations from liver and lung metastases at different stages of metastasis formation. This is the largest proteomic dataset to date on immune and cancer cell populations. We use these data to infer cell-cell interactions, which are the basis for understanding the dynamics of cancer immune interactions and lead to the potential development of therapeutic approaches that may increase immunotherapy response in hard-to-treat liver metastases. We anticipate that the use of these data will have a critical impact on our understanding of cancer metastases in general.
1. High-throughput proteomics of single cells- Our results show our ability to increase the throughput of single-cell analyses to 1536-well plates. This forms a fourfold increase in the number of samples that can be analysed in a single experiment compared to other published methods. Such an increase significantly impacts the applicability of these techniques to tumour analyses, as lower throughput would not be sufficient to generate robust results.
2. Development of spatial proteomics combined with multiplexed imaging on the Phenocycler (CODEX). We established a protocol for multiplexed staining on membrane-coated slides, which are required for laser microdissection (LCM). This method allows advanced immune profiling of the tumour microenvironment, followed by deep analyses of the proteomes of cells of interest. This approach will provide a much deeper understanding of the dynamic differences that occur in cancer and immune cells in the context of their immediate microenvironment.
3. We comprehensively analysed cancer and immune cell populations from liver and lung metastases at different stages of metastasis formation. This is the largest proteomic dataset to date on immune and cancer cell populations. We use these data to infer cell-cell interactions, which are the basis for understanding the dynamics of cancer immune interactions and lead to the potential development of therapeutic approaches that may increase immunotherapy response in hard-to-treat liver metastases. We anticipate that the use of these data will have a critical impact on our understanding of cancer metastases in general.