The transcription factor TCF7L2 is indispensable for intestinal tissue homeostasis where it transmits mitogenic Wnt/ β-Catenin signals in stem and progenitor cells, from which intestinal tumors arise. Yet, TCF7L2 belongs to the most frequently mutated genes in colorectal cancer (CRC), and tumor-suppressive functions of TCF7L2 were proposed. This apparent paradox warrants to clarify the role of TCF7L2 in colorectal carcinogenesis. Here, we investigated TCF7L2 dependence/independence of CRC cells and the cellular and molecular consequences of TCF7L2 loss-of-function. By genome editing we achieved complete TCF7L2 inactivation in several CRC cell lines without loss of viability, showing that CRC cells have widely lost the strict requirement for TCF7L2. TCF7L2 deficiency impaired G1/S progression, reminiscent of the physiological role of TCF7L2. In addition, TCF7L2-negative cells exhibited morphological changes, enhanced migration, invasion, and collagen adhesion, albeit the severity of the phenotypic alterations manifested in a cell-line-specific fashion. To provide a molecular framework for the observed cellular changes, we performed global transcriptome profiling and identified gene-regulatory networks in which TCF7L2 positively regulates the proto-oncogene MYC, while repressing the cell cycle inhibitors CDKN2C/CDKN2D. Consistent with its function in curbing cell motility and invasion, TCF7L2 directly suppresses the pro-metastatic transcription factor RUNX2 and impinges on the expression of cell adhesion molecules. Altogether, we conclude that the proliferation-stimulating activity of TCF7L2 persists in CRC cells. In addition, TCF7L2 acts as invasion suppressor. Despite its negative impact on cell cycle progression, TCF7L2 loss-of-function may thereby increase malignancy, which could explain why TCF7L2 is mutated in a sizeable fraction of colorectal tumors.

Micronuclei are DNA-containing structures separate from the nucleus found in cancer cells. Micronuclei are recognized by the immune sensor axis cGAS/STING, driving cancer metastasis. The mitochondrial apoptosis apparatus can be experimentally triggered to a non-apoptotic level, and this can drive the appearance of micronuclei through the Caspase-activated DNAse (CAD). We tested whether spontaneously appearing micronuclei in cancer cells are linked to sub-lethal apoptotic signals. Inhibition of mitochondrial apoptosis or of CAD reduced the number of micronuclei in tumor cell lines as well as the number of chromosomal misalignments in tumor cells and intestinal organoids. Blockade of mitochondrial apoptosis or deletion of CAD reduced, while experimental activation CAD, STING-dependently, enhanced aggressive growth of tumor cells in vitro. Deletion of CAD from human cancer cells reduced metastasis in xenograft models. CAD-deficient cells displayed a substantially altered gene-expression profile, and a CAD-associated gene expression ‘signature’ strongly predicted survival in cancer patients. Thus, low-level activity in the mitochondrial apoptosis apparatus operates through CAD-dependent gene-induction and STING-activation and has substantial impact on metastasis in cancer.

Epithelial-mesenchymal transition (EMT) fosters cancer cell invasion and metastasis, the main cause of cancer-related mortality. Growing evidence that SNAIL and ZEB transcription factors, typically portrayed as master regulators of EMT, may be dispensable for this process, led us to re-investigate its mechanistic underpinnings. For this, we used an unbiased computational approach that integrated time-resolved analyses of chromatin structure and differential gene expression, to predict transcriptional regulators of TGFβ1-inducible EMT in the MCF10A mammary epithelial cell line model. Bioinformatic analyses indicated comparatively minor contributions of SNAIL proteins and ZEB1 to TGFβ1-induced EMT, whereas the AP-1 subunit JUNB was anticipated to have a much larger impact. CRISPR/Cas9-mediated loss-of-function studies confirmed that TGFβ1-induced EMT proceeded independently of SNAIL proteins and ZEB1. In contrast, JUNB was necessary and sufficient for EMT in MCF10A cells, but not in A549 lung cancer cells, indicating cell-type-specificity of JUNB EMT-regulatory capacity. Nonetheless, the JUNB-dependence of EMT-associated transcriptional reprogramming in MCF10A cells allowed to define a gene expression signature which was regulated by TGFβ1 in diverse cellular backgrounds, showed positively correlated expression with TGFβ signaling in multiple cancer transcriptomes, and was predictive of patient survival in several cancer types. Altogether, our findings provide novel mechanistic insights into the context-dependent control of TGFβ1-driven EMT and thereby may lead to improved diagnostic and therapeutic options.

A cancer of unknown primary (CUP) is a metastatic cancer for which standard diagnostic tests fail to locate the primary cancer. As standard treatments are based on the cancer type, such cases are hard to treat and have very poor prognosis. Using molecular data from the metastatic cancer to predict the primary site can make treatment choice easier and enable targeted therapy. In this article, we first examine the ability to predict cancer type using different types of omics data. Methylation data lead to slightly better prediction than gene expression and both these are superior to classification using somatic mutations. After using 3 data types independently, we notice some differences between the classes that tend to be misclassified, suggesting that integrating the data might improve accuracy. In light of the different levels of information provided by different omics types and to be able to handle missing data, we perform multi-omics classification by hierarchically combining the classifiers. The proposed hierarchical method first classifies based on the most informative type of omics data and then uses the other types of omics data to classify samples that did not get a high confidence classification in the first step. The resulting hierarchical classifier has higher accuracy than any of the single omics classifiers and thus proves that the combination of different data types is beneficial. Our results show that using multi-omics data can improve the classification of cancer types. We confirm this by testing our method on metastatic cancers from the MET500 dataset.

Manifold learning algorithms do not extract the structure of datasets in an abstract form or they do not have high performance for complex data.In this paper, a method for Learning an Inductive Riemannian Manifold in Abstract form (LIRMA) is presented in which the structure of patterns is determined by solving the embedded dynamical system of the patterns. In order to model corresponding system, the true sequence of patterns is estimated using a topology preserving method.LIRMA has the advantage of being an inductive method with low complexity. Additionally, it is a topology preserving method with respect to quantitative measures.