What makes a cancer highly metastatic is not known. Here, we inquire on the metastatic potential (MP) of tumor cells, which reflects their probability to emigrate from the primary tumor to new sites to form secondary cancers. We determine the transcriptomic landscapes of single-cell-derived clones in hybrid EMT space and define metastatic potential gradient genes (MPGGs) that linearly track MP strength. Perturbation of selected MPGGs and linked processes reveals a dynamic cellular and molecular framework of what we define as “cell-state ensembles” underlying the emergence of high MPs. To test if MPGGs predict cancer recurrence, we build the MangroveGS machine-learning model with “gene signature ensembles”: MangroveGSMPGGs robustly predicts patient tumor recurrence and metastases, outperforms all other signatures and staging systems tested, and can be extended to multiple cancer types of epithelial nature. Our findings uncover an unsuspected shared strategy for the onset of metastases that underlies clinical outcome.

How metastatic cells arise is unclear. Here, we search for the induction of recently characterized pro -metastatic states as a surrogate for the origin of metastasis. Since cell-death-inducing therapies can paradoxically promote metastasis, we ask if such treatments induce pro-metastatic states in human colon cancer cells. We find that post-near-death cells acquire pro-metastatic states (PAMEs) and form distant metastases in vivo. These PAME ("let's go"in Greek) cells exhibit a multifactorial cytokine storm as well as signs of enhanced endoplasmic reticulum (ER) stress and nuclear reprogramming, requiring CXCL8, INSL4, IL32, PERK CHOP, and NANOG. PAMEs induce neighboring tumor cells to become PAME-induced migratory cells (PIMs): highly migratory cells that re-enact the storm and enhance PAME migration. Metastases are thus proposed to originate from the induction of pro-metastatic states through intrinsic and extrinsic cues in a pro metastatic tumoral ecosystem, driven by an impending cell-death experience involving ER stress modulation, metastatic reprogramming, and paracrine recruitment via a cytokine storm.

How cells with metastatic potential, or pro-metastatic states, arise within heterogeneous primary tumors remains unclear Here, we have used one index primary colon cancer to develop spiked-scRNAseq to link omics-defined single-cell clusters with cell behavior. Using spiked-scRNAseq we uncover cell populations with differential metastatic potential in which pro-metastatic states are correlated with the expression of signaling and vesicle-trafficking genes. Analyzing such heterogeneity, we define an anti-metastatic, noncell-autonomous interaction originating from non-/low-metastatic cells, and identify membrane VSIG1 as a critical mediator of this interaction. VSIG1 acts to restrict the development of pro-metastatic states autonomously and non-cell autonomously, in part by inhibiting YAP/TAZ-TEAD signaling. As VSIG1 re-expression is able to reduce metastatic behavior from multiple colon cancer cell types, the regulation of VSIG1 or its effectors opens new interventional opportunities. In general, we propose that crosstalk between cancer cells, including the action of VSIG1, dynamically defines the degree of pro-metastatic intra-tumoral heterogeneity.

The strategy of using existing drugs originally developed for one disease to treat other indications has found success across medical fields. Such drug repurposing promises faster access of drugs to patients while reducing costs in the long and difficult process of drug development. However, the number of existing drugs and diseases, together with the heterogeneity of patients and diseases, notably including cancers, can make repurposing time consuming and inefficient. The key question we address is how to efficiently repurpose an existing drug to treat a given indication. As drug efficacy remains the main bottleneck for overall success, we discuss the need for machine-learning computational methods in combination with specific phenotypic studies along with mechanistic studies, chemical genetics and omics assays to successfully predict disease-drug pairs. Such a pipeline could be particularly important to cancer patients who face heterogeneous, recurrent and metastatic disease and need fast and personalized treatments. Here we focus on drug repurposing for colorectal cancer and describe selected therapeutics already repositioned for its prevention and/or treatment as well as potential candidates. We consider this review as a selective compilation of approaches and methodologies, and argue how, taken together, they could bring drug repurposing to the next level.

The strategy of using existing drugs originally developed for one disease to treat other indications has found success across medical fields. Such drug repurposing promises faster access of drugs to patients while reducing costs in the long and difficult process of drug development. However, the number of existing drugs and diseases, together with the heterogeneity of patients and diseases, notably including cancers, can make repurposing time consuming and inefficient. The key question we address is how to efficiently repurpose an existing drug to treat a given indication. As drug efficacy remains the main bottleneck for overall success, we discuss the need for machine-learning computational methods in combination with specific phenotypic studies along with mechanistic studies, chemical genetics and omics assays to successfully predict disease-drug pairs. Such a pipeline could be particularly important to cancer patients who face heterogeneous, recurrent and metastatic disease and need fast and personalized treatments. Here we focus on drug repurposing for colorectal cancer and describe selected therapeutics already repositioned for its prevention and/or treatment as well as potential candidates. We consider this review as a selective compilation of approaches and methodologies, and argue how, taken together, they could bring drug repurposing to the next level.

Targeting stemness promises new therapeutic strategies against highly invasive tumors. While a number of approaches are being tested, inhibiting the core transcription regulatory network of cancer stem cells is an attractive yet challenging possibility. Here we have aimed to provide the proof of principle for a strategy, previously used in developmental studies, to directly repress the targets of a salient stemness and pluripotency factor: NANOG. In doing so we expected to inhibit the expression of so far unknown mediators of pro-tumorigenic NANOG function. We chose NANOG since previous work showed the essential requirement for NANOG activity for human glioblastoma (GBM) growth in orthotopic xenografts, and it is apparently absent from many adult human tissues thus likely minimizing unwanted effects on normal cells. NANOG repressor chimeras, which we name NANEPs, bear the DNA-binding specificity of NANOG through its homeodomain (HD), and this is linked to transposable human repressor domains. We show that in vitro and in vivo , NANEP5, our most active NANEP with a HES1 repressor domain, mimics knock-down (kd) of NANOG function in GBM cells. Competition orthotopic xenografts also reveal the effectiveness of NANEP5 in a brain tumor context, as well as the specificity of NANEP activity through the abrogation of its function via the introduction of specific mutations in the HD. The transcriptomes of cells expressing NANEP5 reveal multiple potential mediators of pro-tumorigenic NANEP/NANOG action including intercellular signaling components. The present results encourage further studies on the regulation of context-dependent NANEP abundance and function, and the development of NANEP-based anti-cancer therapies.

How cells in primary tumors initially become pro-metastatic is not understood. A previous genome-wide RNAi screen uncovered colon cancer metastatic suppressor and WNT promoting functions of TMED3, a member of the p24 ER-to-Golgi protein secretion family. Repression of canonical WNT signaling upon knockdown (kd) of TMED3 might thus be sufficient to drive metastases. However, searching for transcriptional influences on other family members here we find that TMED3 kd leads to enhanced TMED9, that TMED9 acts downstream of TMED3 and that TMED9 kd compromises metastasis. Importantly, TMED9 pro-metastatic function is linked to but distinct from the repression of TMED3-WNT-TCF signaling. Functional rescue of the migratory deficiency of TMED9 kd cells identifies TGFα as a mediator of TMED9 pro-metastatic activity. Moreover, TMED9 kd compromises the biogenesis, and thus function, of TGFα. Analyses in three colon cancer cell types highlight a TMED9-dependent gene set that includes CNIH4, a member of the CORNICHON family of TGFα exporters. Our data indicate that TGFA and CNIH4, which display predictive value for disease-free survival, promote colon cancer cell metastatic behavior, and suggest that TMED9 pro-metastatic function involves the modulation of the secretion of TGFα ligand. Finally, TMED9/TMED3 antagonism impacts WNT-TCF and GLI signaling, where TMED9 primacy over TMED3 leads to the establishment of a positive feedback loop together with CNIH4, TGFα, and GLI1 that enhances metastases. We propose that primary colon cancer cells can transition between two states characterized by secretion-transcription regulatory loops gated by TMED3 and TMED9 that modulate their metastatic proclivities.

How metastases develop is not well understood and no genetic mutations have been reported as specific metastatic drivers. Here we have addressed the idea that epigenetic reprogramming by GLI-regulated pluripotent stemness factors promotes metastases. Using primary human colon cancer cells engrafted in mice, we find that transient expression of OCT4, SOX2, KLF4 +/− cMYC establishes an enhanced pro-metastatic state in the primary tumor that is stable through sequential engraftments and is transmitted through clonogenic cancer stem cells. Metastatic reprogramming alters NANOG methylation and stably boosts NANOG and NANOGP8 expression. Metastases and reprogrammed EMT-like phenotypes require endogenous NANOG, but enhanced NANOG is not sufficient to induce these phenotypes. Finally, reprogrammed tumors enhance GLI2, and we show that GLI2high and AXIN2low, which are markers of the metastatic transition of colon cancers, are prognostic of poor disease outcome in patients. We propose that metastases arise through epigenetic reprogramming of cancer stem cells within primary tumors.

Human colon cancers commonly harbor loss of function mutations in APC, a repressor of the canonical WNT pathway, thus leading to hyperactive WNT-TCF signaling. Re-establishment of Apc function in mice, engineered to conditionally repress Apc through RNAi, resolve the intestinal tumors formed due to hyperactivated Wnt-Tcf signaling. These and other results have prompted the search for specific WNT pathway antagonists as therapeutics for clinically problematic human colon cancers and associated metastases, which remain largely incurable. This widely accepted view seems at odds with a number of findings using patient-derived material: Canonical TCF targets are repressed, instead of being hyperactivated, in advanced colon cancers, and repression of TCF function does not generally result in tumor regression in xenografts. The results of a number of genetic mouse studies have also suggested that canonical WNT-TCF signaling drives metastases, but direct in vivo tests are lacking, and, surprisingly, TCF repression can enhance directly seeded metastatic growth. Here we have addressed the abilities of enhanced and blocked WNT-TCF signaling to alter tumor growth and distant metastases using xenografts of advanced human colon cancers in mice. We find that endogenous WNT-TCF signaling is mostly anti-metastatic since downregulation of TCF function with dnTCF generally enhances metastatic spread. Consistently, elevating the level of WNT signaling, by increasing the levels of WNT ligands, is not generally pro-metastatic. Our present and previous data reveal a heterogeneous response to modulating WNT-TCF signaling in human cancer cells. Nevertheless, the findings that a fraction of colon cancers tested require WNT-TCF signaling for tumor growth but all respond to repressed signaling by increasing metastases beg for a reevaluation of the goal of blocking WNT-TCF signaling to universally treat colon cancers. Our data suggest that WNT-TCF blockade may be effective in inhibiting tumor growth in only a subset of cases but will generally boost metastases.

The WNT-TCF signaling pathway participates in adult tissue homeostasis and repair, and is hyperactive in a number of human diseases including cancers of the colon. Whereas to date there are no antagonists approved for patient use, a potential problem for their sustained use is the blockade of WNT signaling in healthy tissues, thus provoking potentially serious co-lateral damage. Here we have screened a library of plant and microorganism small molecules for novel WNT signaling antagonists and describe withanolide F as a potent WNT-TCF response blocker. This steroidal lactone inhibits TCF-dependent colon cancer xenograft growth and mimics the effects of genetic blockade of TCF and of ivermectin, a previously reported WNT-TCF blocker. However, withanolide F is unique in that it imposes a long-lasting repression of tumor growth, WNT-TCF targets and cancer stem cell clonogenicity after drug treatment. These findings are paralleled by its modulation of chromatin regulators and its alteration of overall H3K4me1 levels. Our results open up the possibility to permanently repress essential signaling responses in cancer cells through limited treatments with small molecules.

Alberto Gulino passed away on November 25, 2014. He will be dearly missed by family, friends and colleagues. Alberto was an enthusiastic scientist who provided important insights into several fields and critical support to young scientists. After graduating in medicine and surgery and specializing in endocrinology at the Catholic University in Rome (Italy), Alberto began his scientific career as a PhD student in biochemistry and developmental biology at the University of Paris. Following his doctoral degree he moved on to become a research associate at the Foundation for Hormone Research in Paris in the group of J Pasqualini, where he addressed the early role of estrogens and of selective estrogen receptor modulators. Alberto, subsequently, returned to his beloved Rome as an assistant professor at Sapienza University, where he continued work on steroid hormones, investigating the biochemical features of nuclear receptor complexes and their role in immunomodulation and tumorigenesis. Following a period at the National Cancer Institute in Bethesda, USA, Alberto was appointed associate professor at the University of L'Aquila (Italy), then full professor and, later on, dean of the Faculty of Medicine and Surgery. Alberto was also heavily involved in education being director of the PhD program in experimental medicine, director of the school of specialization in oncology, and director of the department of experimental medicine. During these years he guided the transformation of the Faculty of Medicine at L'Aquila into a modern and enterprising structure, favoring the return of talented abroad-trained Italian researchers. Back in Rome, at Sapienza University, Alberto's contagious enthusiasm for science and excellence led him to become director of the Laboratory of Molecular Oncology, director of the PhD Program in molecular medicine, and head of the imaging and treatment of brain cancer and the stem cell program of the Italian Institute of Technology at Sapienza University. His abilities to organize scientific events can also be highlighted by the superb and intense Hedgehog meeting he organized in Rome, at a fantastic small hotel facing Palazzo Farnese near Campo de' Fiori and Piazza Navona. The comment of the younger students was that it was ‘terrible' since the talks and discussions were so good that they did not have time to do all the things they had planned to do in Rome. Alberto Gulino's scientific contributions have been notable in different fields but we wish to highlight his work on the mechanism of acetylation and modulation of Gli proteins,1 the regulation of GLI by Numb,2 as well as his work on the regulation of HH signaling by miRNAs.3 In all three works Alberto and his team explored new territory and advanced the field by revealing intricate mechanisms using state of the art genetic, biochemical and cellular analyses. These findings now provide a fertile ground for continued progress in the understanding of the regulation and biological importance of HH signaling. Alberto was exceedingly generous and humble, keeping a healthy mixture of skepticism and trust. It will be impossible to forget his love of science and his style, where complex expositions crystallized in important insights and solid work. Alberto had a particular relational warmth and was admired by many of his coworkers, who saw in him a guiding and protecting figure. It will also be impossible to forget being a guest of Alberto in Rome, dining at superb small local restaurants where he and his wife, Isabella Screpanti, were known by name by the owners and the cooks, or visiting magnificent monuments or small streets and listening to Isabella tell stories of Rome, as only Romans can tell. Alberto was an important figure in Italian and European science, and a wonderful person. He will be missed at home and abroad by many friends and colleagues. We thank his students Enrico De Smaele, Elisabetta Ferreti, Lucia Di Marcotullio and Luca Canettieri for biographical information.

The progression of tumors to the metastatic state involves the loss of metastatic suppressor functions. Finding these, however, is difficult as in vitro assays do not fully predict metastatic behavior, and the majority of studies have used cloned cell lines, which do not reflect primary tumor heterogeneity. Here, we have designed a novel genome-wide screen to identify metastatic suppressors using primary human tumor cells in mice, which allows saturation screens. Using this unbiased approach, we have tested the hypothesis that endogenous colon cancer metastatic suppressors affect WNT-TCF signaling. Our screen has identified two novel metastatic suppressors: TMED3 and SOX12, the knockdown of which increases metastatic growth after direct seeding. Moreover, both modify the type of self-renewing spheroids, but only knockdown of TMED3 also induces spheroid cell spreading and lung metastases from a subcutaneous xenograft. Importantly, whereas TMED3 and SOX12 belong to different families involved in protein secretion and transcriptional regulation, both promote endogenous WNT-TCF activity. Treatments for advanced or metastatic colon cancer may thus not benefit from WNT blockers, and these may promote a worse outcome.

Constitutive activation of canonical WNT ‐ TCF signaling is implicated in multiple diseases, including intestine and lung cancers, but there are no WNT ‐ TCF antagonists in clinical use. We have performed a repositioning screen for WNT ‐ TCF response blockers aiming to recapitulate the genetic blockade afforded by dominant‐negative TCF . We report that Ivermectin inhibits the expression of WNT ‐ TCF targets, mimicking dn TCF , and that its low concentration effects are rescued by direct activation by TCF VP 16 . Ivermectin inhibits the proliferation and increases apoptosis of various human cancer types. It represses the levels of C‐terminal β‐ CATENIN phosphoforms and of CYCLIN D1 in an okadaic acid‐sensitive manner, indicating its action involves protein phosphatases. In vivo , Ivermectin selectively inhibits TCF ‐dependent, but not TCF ‐independent, xenograft growth without obvious side effects. Analysis of single semi‐synthetic derivatives highlights Selamectin, urging its clinical testing and the exploration of the macrocyclic lactone chemical space. Given that Ivermectin is a safe anti‐parasitic agent used by > 200 million people against river blindness, our results suggest its additional use as a therapeutic WNT ‐ TCF pathway response blocker to treat WNT ‐ TCF ‐dependent diseases including multiple cancers.

Canonical Hedgehog (HH) signaling leads to the regulation of the GLI code: the sum of all positive and negative functions of all GLI proteins. In humans, the three GLI factors encode context-dependent activities with GLI1 being mostly an activator and GLI3 often a repressor. Modulation of GLI activity occurs at multiple levels, including by co-factors and by direct modification of GLI structure. Surprisingly, the GLI proteins, and thus the GLI code, is also regulated by multiple inputs beyond HH signaling. In normal development and homeostasis these include a multitude of signaling pathways that regulate proto-oncogenes, which boost positive GLI function, as well as tumor suppressors, which restrict positive GLI activity. In cancer, the acquisition of oncogenic mutations and the loss of tumor suppressors – the oncogenic load – regulates the GLI code toward progressively more activating states. The fine and reversible balance of GLI activating GLIA and GLI repressing GLIR states is lost in cancer. Here, the acquisition of GLIA levels above a given threshold is predicted to lead to advanced malignant stages. In this review we highlight the concepts of the GLI code, the oncogenic load, the context-dependency of GLI action, and different modes of signaling integration such as that of HH and EGF. Targeting the GLI code directly or indirectly promises therapeutic benefits beyond the direct blockade of individual pathways.

Spanish MINECO; Marie Curie FP6 (RTN035528-2); FP7(ITN238186); Fundacion Areces; Juntad e Andalucia (Project P08-FQM-4267); Swiss National Science Foundation; the University of Geneva; European Research Council; Leenaards foundation; European Molecular Biology Organization (younginvestigatorprogram); Cantonet Republique de Geneveto

A central question in biology is how secreted morphogens act to induce different cellular responses within a group of cells in a concentration-dependent manner. Modeling morphogenetic output in multicellular systems has so far employed linear diffusion, which is the normal type of diffusion associated with Brownian processes. However, there is evidence that at least some morphogens, such as Hedgehog (Hh) molecules, may not freely diffuse. Moreover, the mathematical analysis of such models necessarily implies unrealistic instantaneous spreading of morphogen molecules, which are derived from the assumptions of Brownian motion in its continuous formulation. A strict mathematical model considering Fick's diffusion law predicts morphogen exposure of the whole tissue at the same time. Such a strict model thus does not describe true biological patterns, even if similar and attractive patterns appear as results of applying such simple model. To eliminate non-biological behaviors from diffusion models we introduce flux-limited spreading (FLS), which implies a restricted velocity for morphogen propagation and a nonlinear mechanism of transport. Using FLS and focusing on intercellular Hh-Gli signaling, we model a morphogen gradient and highlight the propagation velocity of morphogen particles as a new key biological parameter. This model is then applied to the formation and action of the Sonic Hh (Shh) gradient in the vertebrate embryonic neural tube using our experimental data on Hh spreading in heterologous systems together with published data. Unlike linear diffusion models, FLS modeling predicts concentration fronts and the evolution of gradient dynamics and responses over time. In addition to spreading restrictions by extracellular binding partners, we suggest that the constraints imposed by direct bridges of information transfer such as nanotubes or cytonemes underlie FLS. Indeed, we detect and measure morphogen particle velocity in such cell extensions in different systems.