Despite significant efforts, genetic alterations that specifically promote metastatic phenotypes have not been identified. However, cancers with high propensity to metastasize have distinct transcriptional profiles. This suggests that metastatic programming in primary tumours is largely caused by aberrations in transcriptional control. Nevertheless, how metastatic transcriptional programs arise has remained elusive. We will use experimental metastasis models to identify critical transcriptional control mechanisms of metastatic progression. First, we will use established clones of varying metastatic potential to perform chromatin and transcriptional profiling (ATAC-seq, RNA-seq and ChIP-seq). Using computational methods with clinical validation, we will identify candidate gene regulatory regions and their regulators underlying metastasis. We will then use experimental genetics (gene and gene regulatory element loss-of-function and gain-of-function) to identify functional drivers of metastatic transcriptional programs.
To form metastasis, tumour cells need to leave the primary tumour, reach the blood flow, colonize distant tissues, and establish secondary tumours. Using state of the art lineage tracing in mouse models of skin SCCs presenting spontaneous lung and lymph node metastasis and epithelial to mesenchymal transition (EMT), we will assess the importance of EMT and mesenchymal to epithelial transition (MET) in regulating the development of metastasis. Using single cell RNA and ATAC sequencing, we will unravel the tumour states present in the primary tumours, circulating tumour cells (CTCs) and at the metastatic sites. We will define the functional role of these different tumour states and the molecular mechanisms regulating their transition using genetic gain and loss of function and pharmacological inhibition of the pathways uncovered by these analyses. This study will provide key insights into the molecular mechanisms that regulate tumour transition states controlling metastasis.
Outcome of patients with HR+/Her2- metastatic breast cancer is highly heterogeneous, with a subset of patients who are long term survivors, and at the other extreme, a group of patients who die within two years. The objective of this project is to characterise this group with very poor outcome. Specific Aim 1: to identify a genomic profile associated with poor outcome. We have already performed whole exome sequencing of 617 metastatic breast cancer patients (Bertucci, Nature, 2019). We will have sequenced overall 1500 patients by the end of 2020. We believe this cohort will represent patient heterogeneity sufficiently well to develop useful statistical models of patient outcomes. Genomic data are often characterized by complex interactions, so we plan to use graph-based genomic data analysis to model these interactions (Pirayre 2017). Specific Aim 2: to identify a proteomic profile associated with poor outcome. In order to address this question, we will profile the same samples mentioned in specific aim 1 in a proteomic facility (Liquid chromatography-mass spectrometry, n=6 000 prot, Astra Zeneca). We plan to profile 1 000 samples with a high degree of precision. From the statistical point of view, using so many variables to characterize a relatively low number of patients may incur bias. We will specifically use bias-reduction techniques from AI techniques, such as bagging and boosting methods (Bühlman 2012), to produce more reliable models and profiles. Specific Aim 3: to integrate data from genomic and proteomics with methods of representation learning for multi-modal data and to develop a predictor of patients outcome with artificial intelligence models. A special focus will be devoted to the identification of biological markers with a strong influence on patient outcomes thanks to techniques of explainable AI. Combining data from these two very different sources is a challenge from the data analytics point of view. We plan to leverage the power of graph-based neural-networks to this end (Kipf 2017). Specific Aim 4: to validate the predictor in an independent dataset (PADA1 trial, n=300 for the validation set).
Luminal cells (LCs) act as the cancer cell of origin in the most common BCa, the luminal subtype. It remains unclear whether the heterogeneity found in luminal-derived tumours and metastasis post treatment arises from a pre-existing heterogeneity within LCs. Recently developed technology allows for primary and metastasis BCa human/mouse organoid lines, which broadly recapitulate the diversity of the disease (Morphology, Histopathology, ER status, etc.) and are consistent with in vivo xeno- transplantations and patient responses. These will be used to identify in vitro drug resistance and clonal selection mechanisms and its relationship with LC heterogeneity. In this context, we will genetically define the various multipotent and differentiated populations using luminal markers and use lineage tracing to assess whether lineage-restricted tumor-initiating cells (TIC) maintain the ER+ lineage. We have devised a strategy based on editing the genomes of patient-derived tumor organoids using CRISPR/Cas9 technology to integrate reporter cassettes at desired marker genes. We aim to describe the gene programs that define these populations, and how they evolved upon treatment selection and metastasis. Lineage-tracing experiments will confirm the capacity of the distinct populations to self-renew and generate progeny over periods of metastasis and drug treatment selection.
Understanding how the immune system affects cancer progression remains one of the fundamental questions in cancer biology and is crucial for the development of effective therapies. To dissect the molecular mechanisms of immune evasion and their emergence during tumour dormancy and metastatic colonization, we will use established metastasis models combined with a novel technique we have developed, utilizing CRISPRa to trace clones in heterogeneous cell populations (CaTCH; Umkehrer et al. submitted). CaTCH enables lineage tracing of millions of cells with stably integrated DNA barcodes (BCs). Importantly, activation of a reporter allows FACS-based isolation of clones with distinct phenotypes of interest (i.e. tumour cells that stay dormant or escape dormancy in an immune-competent or -deficient background), enabling their mechanistic workup. We will assess the immune-contexture, mutational load and the transcriptional programs in the selected cancer cell clones and their pre-selection counterparts. By integrating clinical data, we will identify candidate mediators of these phenotypes, probe their role in functional metastasis assays (gain/loss of function) and dissect the underlying mechanisms.
We will study the contribution of the vascular compartment in the castration-naïve metastatic PCa, encompassing a tumour type that is metastatic at the time of first diagnosis. Specifically, we aim to: 1) identify critical molecular changes in the stroma of metastatic prostate cancer through computational and clinical analysis; and 2) conduct co-culture analysis with modify stroma and transformed epithelium, as well as in vivo modification of the stroma in murine models of prostate cancer. The nature of the stromal contribution to the progression of the disease will be mechanistically deconstructed and exploited for the development of therapies.
Emerging evidence indicates that immune cells are mobilized and activated in the tumour microenvironment (TME) during metastatic cancer progression. The TME is also altered following various anticancer therapies, which can paradoxically contribute to a lack of response/acquired resistance to treatment. In particular, inflammation can promote metastatic progression. We will systematically focus on different immune cell types whose levels vary across systemic inflammation, in specific tissues – including the brain, and whose variations are further exacerbated by the presence of a primary tumour. We will determine whether this translates to increased cancer metastasis to this site, and the potential dependency on specific cytokines or factors. Subsequent genetic, chemical or biological studies will follow to confirm the causality of these mechanisms to the metastatic phenotype. The underlying mechanistic contribution of the microenvironment to metastasis and therapeutic resistance will be based on a range of complementary techniques including mouse models of cancer, 3D co-culture systems, computational approaches, and analysis of patient samples in collaboration with our clinical colleagues.
Invasive lobular breast cancer (ILC) represents 10-15% of all newly diagnosed breast cancers and has a distinct pattern of metastatic spread, with a greater propensity to develop ovarian, GI and leptomeningeal metastasis where the tumour spread is typically over the serosal surfaces rather than infiltrating into the parenchyma. In addition, lobular breast cancers have poor long-term outcome indicative of a latent metastatic phenotype. The genetics of ILC have been well studied, however the biology of ILC is poorly understood due to the paucity of clinically relevant models. To identify therapeutic vulnerabilities associated with ILC, we will develop improved in vitro, ex vivo and in vivo models and interrogate the biology of ILC metastatic colonisation – focusing on the pattern of metastatic spread in the peritoneum and leptomeninges. In parallel we will collect liquid biopsies (ascitic fluid and cerebrospinal fluid) from ILC patients for ultra-low-pass and whole exome sequencing to identify potential drivers of metastatic spread, determine the clonal evolution of ILC metastasis and for the development of diagnostic and disease monitoring biomarkers.
We build on our previous hypotheses/work to investigate and unravel how MSK1 contributes to BCa dormancy. We will determine how chromatin is organized within dormant tumour cells (DTCs), focusing on histone H3 to map genomic and transcription changes associated with metastatic escape from non-adaptive, anti-tumor immune activities. We will test the metastasis cascade of events to pinpoint how to therapeutically intervene in crosstalk between signalling pathways and/or the epigenetic regulations that support metastatic escape. We will characterize, at the genetic, biochemical, and cellular levels, whether MSK1 in DTC cells controls mediators of the non-adaptive immune response. To establish the contribution of MSK1 in attenuating the NK functions in metastatic cells, we will use both stable gain-of-function (gene overexpression) and loss-of-function (gene knockdown/knockout via shRNA and CRISPR/Cas9). We will test the metastatic activities of these transfectants with in vitro experiments as well as inoculation into immunocompromised mice of various backgrounds with immune deficiencies, including Balb/c Nude athymic mice (FOXN1nu) that retain functional innate system, and the more immune compromised NOD/SCID gamma mice with neither adaptive nor innate immunity.
We aim to understand the contribution of endothelial cells and the angiogenic switch to support expansion of indolent latent metastatic cells. We will determine how endothelial cells are organized within latent metastasis by focusing on genomic, transcription and translation changes associated with overt metastasis. We will test several functions that support metastasis to identify where, when and how endothelial cells contribute to support exit from latency. To establish the contribution of endothelial cells in attenuating or supporting the expansion of latent metastasis, we will use a series of genetically engineered mouse models with reduced and enhanced angiogenesis through the manipulation of the PI3K pathway specifically in the endothelium. This will be back-crossed in Balb/c Nude athymic mice (FOXN1nu) background to subsequently introduce a series of human breast cancer latent metastasis models. Next, we will screen for communication signals and gene expression patterns among cancer and endothelial populations. Subsequent mechanistic analysis will be performed to understand the molecular dependencies among the cellular populations. Finally, confocal microscopy of in bone events will be performed post intra iliac injection and bone growth in vitro.
Tumour dormancy is characterized by a reversible cell cycle arrest of tumour cells and is thought to be the cause of metastasis relapse years after remission. The mechanisms that regulate tumour dormancy are poorly understood. We will use a model of skin squamous cell carcinoma, in which fluorescently labelled tumour cells (TCs) spontaneously form metastasis in the lung, to decipher the mechanisms that control metastasis dormancy. In this model, we have found that isolated quiescent TCs could be identified. To uncover the mechanisms that control metastasis dormancy, we will micro-dissect overt metastases, FACS isolate fluorescently quiescent and proliferative TCs, and perform single cell RNA seq to identify the gene signature of dormant and proliferative metastatic cells. We will then use pharmacological approaches to assess the function of candidate signalling pathways that are likely to mediate dormancy and the function of these genes or pathways in the regulation of lung metastasis proliferation will be assessed. This study will be important to uncover new regulators of metastasis dormancy with important implications for cancer therapy.
To identify genomic alterations and signalling pathways associated with treatment failure in patients with metastatic HER2-enriched (HER2-E) hormone receptor-positive (HR+) breast cancer (BC): DNA and RNA of >400 HR+ samples (including 140 HER2-E) from 4 metastatic cohorts will be extracted and analyzed for genomic alterations. Massive parallel sequencing (MPS; MiSeq Illumina platform) will be used to detect somatic mutations on a panel of 60 oncogenes and tumour suppressor genes, covering most frequently mutated exons including PIK3CA, TP53, CDH1, GATA3 and ESR1 as well as tumour mutational burden. The nCounter Breast Cancer 360 Panel (Nanostring Technologies) will be used to determine the expression of 752 genes across 23 key BC pathways and processes. Intrinsic subtypes and biologically relevant gene signatures will be determined using R software. Subsequent phenotypic characterization will take place.
The majority of the world population smokes or has smoked, and a further proportion is or has been exposed to environmental tobacco smoke. Although recent evidence consistently supports an association between smoking and an increased risk of breast cancer development, progression to metastasis and death, the underlying mechanisms are poorly understood. Here we will assess the effects of exposure to cigarette smoke using mouse models of the most frequent gain-of-function mutation in breast cancer (i.e., mutant phosphatidylinositol 3-kinase (PI3K)). We will directly assess how smoking contributes to breast cancer metastasis in a cell-autonomous (i.e., within cancer cells) or non-cell-autonomous (i.e., immune cells within metastatic sites) manner, and identify and validate the oncogenic molecular and cellular mechanisms involved. We will use a battery of cell biological assays, as well as multiphoton intravital imaging to dissect the mechanisms underlying cigarette smoke-evoked metastasis.
The nucleic acid sensor STING is an important regulator in innate immune cells to protect the body against invading pathogens and malignant cells. Currently, synthetic STING agonists are tested in clinical trials for their ability to induce a de novo antitumor immune response in metastatic tumour patients. Besides activating immune cells, we discovered that STING agonists can induce cell death in both cancer and immune cells, although at different concentration ranges. In preliminary experiments, we discovered that some cancer cells die via the immunogenic cell death pathway. Understanding if and how STING induces immunogenic cancer cell death and activates immune cells while maintaining their viability will be crucial. For this project, we will investigate which pathways regulate STING induced cell death in cancer and immune cells using CRISPR, proteomics and high throughput drug screening platforms. Determine if the identified cell death pathways are immune activating or immune silent and test which of these pathways are differentially regulated in cancer and immune cells.
Today, many cancers are often diagnosed at a late stage resulting in poor treatment options and survival rates. One of the most difficult cancers to detect early on is pancreatic cancer, which has one of the lowest 5-year relative survival rates at only 9%. Symptoms of pancreatic cancer, such as weight loss, abdominal discomfort and occasionally diabetes, are often mistaken for signs of less severe illnesses and overlooked in clinical practice. Other examples of cancer types diagnosed in a late stage are lung, colorectal, ovarian, stomach and liver cancer. Recently, deep learning has demonstrated to be a highly effective methodology to learn complex data structures, such as healthcare data. Danish health registries are some of the largest and most comprehensive healthcare datasets in the world. These comprise disease history, clinical notes, laboratory measurements, drugs and diagnostic images. Due to the linkability by the Central Person Register (CPR) identification number, it is possible to achieve individual-level linking across all data types.