Differentiation of normal and tumor cells is controlled by regulatory networks enforced by lineage‐determining transcription factors (TFs). Among them, TFs such as FOXA1/2 bind naïve chromatin and induce its accessibility, thus establishing new gene regulatory networks. Pancreatic ductal adenocarcinoma (PDAC) is characterized by the coexistence of well‐ and poorly differentiated cells at all stages of disease. How the transcriptional networks determining such massive cellular heterogeneity are established remains to be determined. We found that FOXA2, a TF controlling pancreas specification, broadly contributed to the cis‐regulatory networks of PDACs. Despite being expressed in both well‐ and poorly differentiated PDAC cells, FOXA2 displayed extensively different genomic distributions and controlled distinct gene expression programs. Grade‐specific functions of FOXA2 depended on its partnership with TFs whose expression varied depending on the differentiation grade. These data suggest that FOXA2 contributes to the regulatory networks of heterogeneous PDAC cells via interactions with alternative partner TFs.

Tandem repeats (TRs) are generated by DNA replication errors and retain a high level of instability, which in principle would make them unsuitable for integration into gene regulatory networks. However, the appearance of DNA sequence motifs recognized by transcription factors may turn TRs into functional cis-regulatory elements, thus favoring their stabilization in genomes. Here, we show that, in human cells, the transcriptional repressor ZEB1, which promotes the maintenance of mesenchymal features largely by suppressing epithelial genes and microRNAs, occupies TRs harboring dozens of copies of its DNA-binding motif within genomic loci relevant for maintenance of epithelial identity. The deletion of one such TR caused quasi-mesenchymal cancer cells to reacquire epithelial features, partially recapitulating the effects of ZEB1 gene deletion. These data demonstrate that the high density of identical motifs in TRs can make them suitable platforms for recruitment of transcriptional repressors, thus promoting their exaptation into pre-existing cis-regulatory networks.

No abstract

No abstract

Pancreatic ductal adenocarcinoma (PDAC), one of the most highly lethal tumors, is characterized by complex histology, with a massive fibrotic stroma in which both pseudo‐glandular structures and compact nests of abnormally differentiated tumor cells are embedded, in different proportions and with different mutual relationships in space. This complexity and the heterogeneity of the tumor component have hindered the development of a broadly accepted, clinically actionable classification of PDACs, either on a morphological or a molecular basis. Here, we discuss evidence suggesting that such heterogeneity can to a large extent, albeit not exclusively, be traced back to two main classes of PDAC cells that commonly coexist in the same tumor: cells that maintained their ability to differentiate toward endodermal, mucin‐producing epithelia and epithelial cells unable to form glandular structures and instead characterized by various levels of squamous differentiation and the expression of mesenchymal lineage genes. The underlying gene regulatory networks and how they are controlled by distinct transcription factors, as well as the practical implications of these two different populations of tumor cells, are discussed.

ObjectivePancreatic ductal adenocarcinomas (PDACs) include heterogeneous mixtures of low-grade cells forming pseudoglandular structures and compact nests of high-grade cells organised in non-glandular patterns. We previously reported that low-grade PDAC cells display high expression of interferon regulatory factor 1 (IRF1), a pivotal transcription factor of the interferon (IFN) system, suggesting grade-specific, cell-intrinsic activation of IFN responses. Here, we set out to determine the molecular bases and the functional impact of the activation of IFN-regulated responses in human PDACs.DesignWe first confirmed the correlation between glandular differentiation and molecular subtypes of PDAC on the one hand, and the expression of IRF1 and IFN-stimulated genes on the other. We next used unbiased omics approaches to systematically analyse basal and IFN-regulated responses in low-grade and high-grade PDAC cells, as well as the impact of IRF1 on gene expression programmes and metabolic profiles of PDAC cells.ResultsHigh-level expression of IRF1 in low-grade PDAC cells was controlled by endodermal lineage-determining transcription factors. IRF1-regulated gene expression equipped low-grade PDAC cells with distinctive properties related to antigen presentation and processing as well as responsiveness to IFN stimulation. Notably, IRF1 also controlled the characteristic metabolic profile of low-grade PDAC cells, suppressing both mitochondrial respiration and fatty acid synthesis, which may in part explain its growth-inhibiting activity.ConclusionIRF1 links endodermal differentiation to the expression of genes controlling antigen presentation and processing as well as to the specification of the metabolic profile characteristic of classical PDAC cells.

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest types of cancer and is characterized by a high intra-tumor heterogeneity, with the coexistence of well- and poorly-differentiated cells in virtually all the tumor cases. FOXA1/2 play a fundamental role in pancreas development and are believed to act as pioneer factors, namely to create accessible chromatin by displacing nucleosomes early in development and making genomic regulatory elements available to other transcriptional regulators. Therefore, we asked whether FOXA1/2 also control differentiation in human PDAC. Whereas FOXA1 showed a remarkable propensity to be expressed selectively by well-differentiated tumor areas and cell lines, FOXA2 showed a broad profile of expression throughout PDAC stages. Through the combination of genomic and functional assays we found that FOXA2 contributed to activate distinct transcriptional programs in tumor cells of different grades. Such complexity in FOXA2 activity specifically depended on its interaction with distinct transcription factors selectively expressed in high- or low-grade tumor cells. While in low-grade (well-differentiated) PDAC cells FOXA2 binding specificity relied on the interaction with the epithelial transcription factor HNF1beta, in high-grade cells it depended on the cooperation with HOXB8 and HOXB9. Knocking out these transcription factors was sufficient to abolish differential recruitment of FOXA2 to grade specific genomic sites. In summary, we found that FOXA2 is broadly express throughout PDAC stages and, through the interaction with different transcription factors, shapes the transcriptional program of the different tumor grades in pancreatic cancer. These findings help to shed light on the complexity of PDAC heterogeneity and on the circuitries that maintain PDAC differentiation and grading. Citation Format: Marta Milan, Gioacchino Natoli. Role of the FOXA transcription factors in shaping distinct transcriptional programs in human pancreatic cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr LB-202.

Introduction An increasing number of recent studies suggest that tumour response to chemotherapy cannot be fully described in terms of only interaction of the drug with the main cellular target (e.g. DNA or microtubules) but can include multiple physiological and physicochemical changes. Insight into drug-induced functional alterations in cancer cells and tissues is crucially important for understanding of mechanisms of a drug action and for development new approaches to enable monitoring of the early-treatment response. Material and methods In our studies, we focus on multiparameteric analysis of tumour responses to chemotherapy using advanced fluorescence imaging techniques. Previously, we developed methodologies for in vivo probing several parameters, including cytosolic pH, metabolic status, and viscosity of plasma membrane, in mouse tumour models. Mapping of cytosolic pH in tumours is performed using ratiometric genetically encoded sensor SypHer2 and fluorescence whole-body imaging [Shirmanova et al. BBA-GS 2015]. Imaging of cellular metabolism is based on the visualisation of fluorescence intensities and lifetimes of intrinsic metabolic cofactors NAD(P)H and FAD [Shirmanova et al. Sci.Rep. 2017]. Viscosity is measured using molecular BODIPY-based rotor and fluorescence lifetime imaging microscopy (FLIM) [Shimolina et al. Sci.Rep. 2017]. Results and discussions We showed that acidification of cytosolic pH occurs after therapy with cisplatin in vivo, and this, likely, favours metabolic reorganisation of cells. Treated tumours exhibited a decreased relative contribution from free (cytosolic, protein-unbound) NAD(P)H, indicating a metabolic shift from glycolysis towards oxidative metabolism. It is interesting that optical metabolic imaging could allow early detection of tumour response to chemotherapy, before there are changes in tumour size, and metabolic changes were the same for therapeutic agents with different mechanisms of action (cisplatin, paclitaxel). Moreover, tumours treated with cisplatin displayed decreased plasma membrane viscosity. Presumably, this alteration of viscosity participates in cisplatin-induced apoptosis. Similar changes of pH, energy metabolism and viscosity were previously reported for cisplatin-treated cultured cancer cells. Conclusion These results, therefore, suggest that all investigated parameters play a role in the cytotoxicity of the drugs and may provide a useful approach for monitoring tumour responses to chemotherapy. The study was supported by the Russian Science Foundation (#14-25-00129). Introduction Pancreatic Ductal Adenocarcinoma (PDAC) is the most frequent neoplasia of the exocrine pancreas. This tumour is and is characterised by a pervasive heterogeneity, with the coexistence of a range of histological grades, from epitheliallike to mesenchymal-like features. We previously dissected the transcriptional and epigenetic networks underlying PDAC grading. We identified the association of low grade phenotypes with a cell-autonomous interferon-related signature. Therefor...