EACR-MRS conference on Seed and Soil: In Vivo Models of Metastasis

Clinical & Experimental Metastasis, Mar 2018

New experimental tools are urgently required to better understand the metastatic process. The importance of such tools is underscored by the fact that many anti-cancer therapies are generally ineffective against established metastases. This makes a major contribution to the fact that metastatic spread is responsible for over 90% of cancer patient deaths. It was therefore timely that the recent “Seed and Soil: In Vivo Models of Metastasis” conference held in Berlin, Germany (27–29 of November 2017) aimed to give an in-depth overview of the latest research models and tools for studying metastasis, and to showcase recent findings from world-leading metastasis researchers. This Meeting Report summarises the major themes of this ground-breaking conference.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


EACR-MRS conference on Seed and Soil: In Vivo Models of Metastasis

EACR-MRS conference on Seed and Soil: In Vivo Models of Metastasis I. Teles Alves 0 1 2 5 6 7 8 9 10 N. Cohen 0 1 2 5 6 7 8 9 10 P. G. Ersan 0 1 2 5 6 7 8 9 10 R. Eyre 0 1 2 5 6 7 8 9 10 I. Godet 0 1 2 5 6 7 8 9 10 D. Holovanchuk 0 1 2 5 6 7 8 9 10 R. Jackstadt 0 1 2 5 6 7 8 9 10 L. Kyjacova 0 1 2 5 6 7 8 9 10 K. Mahal 0 1 2 5 6 7 8 9 10 A. Noguera‑Castells 0 1 2 5 6 7 8 9 10 L. Recalde‑Percaz 0 1 2 5 6 7 8 9 10 J. P. Sleeman 0 1 2 5 6 7 8 9 10 0 Departments of Molecular Biology and Genetics, Faculty of Science, Bilkent University , Ankara , Turkey 1 Department of Pathology, Sackler School of Medicine, Tel Aviv University , Tel Aviv , Israel 2 Department of Cell Biology and Biochemistry , Springer Science 3 I. Teles Alves 4 Business Media B.V. , Dordrecht , The Netherlands 5 Medical Faculty Mannheim, Centre for Biomedicine and Medical Technology Mannheim (CBTM), University of Heidelberg , Mannheim , Germany 6 Cancer Research UK, Beatson Institute , Glasgow , UK 7 Molecular Oncology group, Cancer Research UK, Manchester Institute, The University of Manchester , Manchester , UK 8 Department of Chemical and Biomolecular Engineering, The Johns Hopkins University , Baltimore, MD , USA 9 Department of Oncology, The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins University School of Medicine , Baltimore, MD , USA 10 Breast Biology Group, Breast Cancer Now Research Unit, Division of Cancer Sciences, Faculty of Biology, Medicine and Health, Manchester Cancer Research Centre, University of Manchester , Wilmslow Road, Manchester , UK New experimental tools are urgently required to better understand the metastatic process. The importance of such tools is underscored by the fact that many anti-cancer therapies are generally ineffective against established metastases. This makes a major contribution to the fact that metastatic spread is responsible for over 90% of cancer patient deaths. It was therefore timely that the recent “Seed and Soil: In Vivo Models of Metastasis” conference held in Berlin, Germany (27-29 of November 2017) aimed to give an in-depth overview of the latest research models and tools for studying metastasis, and to showcase recent findings from world-leading metastasis researchers. This Meeting Report summarises the major themes of this ground-breaking conference. Metastasis; Mouse models; Signaling Introduction The “Seed and Soil: In Vivo Models of Metastasis” conference was held in Berlin, Germany from the 27–29 of November 2017, and was the first ever joint conference held by the European Association for Cancer Research (EACR) and the Metastasis Research Society (MRS). Spear-headed by the conference chair Janine Erler with support from the scientific programme committee (Jonathan Sleeman, Ulrike Stein and Christof von Kalle), the conference focused on 8 9 10 Department of Medicine, University of Barcelona, Barcelona, Spain 11 Institut d’Investigacions Biomédiques August Pi i Sunyer, Barcelona, Spain 12 Department of Medicine, University of Barcelona, Barcelona, Spain 13 Medical Faculty Mannheim, University of Heidelberg, Mannheim, Germany 14 Institute of Toxicology and Genetics, Karlsruhe Institute for Technology (KIT), Karlsruhe, Germany amongst the attendees. With 13 invited speakers, 6 proffered papers and 103 posters, the 168 participants were treated to a rich and varied scientific menu of excellent quality. Keynote addresses were given by Robert Kerbel, Andreas Trumpp and Erik Sahai, the latter being kindly supported by a grant from the European Molecular Biology Organisation. Additionally, seven EACR-Worldwide Cancer Research Meeting Bursaries and five MRS Meeting Bursaries were awarded. The majority of the recipients of these awards have also coauthored the summary below that outlines the contents of this exciting conference. Opening keynote lecture and conference debate The joint congress began with keynote speaker, Robert S. Kerbel (University of Toronto, Toronto, Canada), focusing on the use of metastatic models for experimental therapeutics and introducing a new era of anti-vascular therapy. He reported that preclinical models of metastatic disease including genetically engineered mouse models (GEMMs) and patient derived xenografts (PDXs) are highly effective for testing treatments for primary tumors, but fail to predict treatment of advanced metastatic disease in phase III clinical trials [ 1 ]. His studies testing antiangiogenic tyrosine kinase inhibitor (TKI) drugs, i.e. sunitinib, alone or with chemotherapy recapitulated the negative phase III clinical results [ 2 ]. However, an antibody targeting the VEGF pathway (DC101) combined with paclitaxel chemotherapy has impact on treating advanced metastatic breast cancer [ 2 ]. It is known that the lack of efficacy of some anti-angiogenic drugs is due to acquired resistance [ 3 ]. One major explanation for resistance to anti-angiogenic drugs is “vessel co-option”, the ability of tumors to hijack the existing vasculature in organs, which results in the absence of sprouting angiogenesis [ 4 ]. To this end, Dr. Kerbel concluded with a discussion of the potential anti-angiogenic therapy targeting vessel co-option in tumors as a means to suppress or prevent metastasis. The opening keynote lecture was followed by the “Conference Debate”, in which Dr. Helmut Augustin and Dr. Yibin Kang discussed the hurdles and difficulties that need to be overcome if preclinical data is to be translated into successful clinical trials. Dr. Augustin described the process of metastasis as one of the last mysteries in cancer research, which is partially due to the fact that broadly used preclinical mouse models lack relevance to the clinics. He highlighted results from his recently published review describing an overview of the currently used mouse models in cancer [ 5 ]. In conclusion, most studies employ xenograft mouse models generated by cancer cell line injection, which fail to model crucial steps of the metastatic cascade. The use of cell lines that have been selected in culture for many years and which have lost the primary tumor characteristics represents a major reason why preclinical studies are difficult to translate into the clinics. To overcome these problems, Dr. Augustin suggested analyses and approaches which are more clinically relevant, e.g. to investigate metastatic spread after surgical removal of the primary tumor, and the use of autochthonous mouse models. Another improvement could be data standardization, meaning a uniform way to analyze and describe data, eventually helping to interpret results across different institutes and countries. Dr. Kang explained the importance of preclinical mouse models in efforts to understand metastasis, particularly because these mouse models help to understand the biological relevance of targets, stromal interactions, drug effects and to define/find therapeutic windows. He described the evolution of mouse models and the differences between immunocompetent and immunodeficient models. Furthermore, he highlighted recent developments in the field, such as humanized mouse models and CRISPR/Cas9 technology, which already improved current mouse models and will continue to do so. In a study where he investigated Metadherin (MTDH), he explained the need of multiple mouse models to fully describe the function of this and other molecules in cancer metastasis [ 6–8 ]. The points that Dr. Augustin and Dr. Kang discussed concordantly emphasize that one single mouse model is not sufficient to understand and challenge the process of metastasis to such a level that the results can be directly translated into the clinic. The researcher should always use several of the best-suited model systems available and combine those to overcome the potential limitations of each model and obtain more relevant and applicable data. New models of metastasis The biological complexity underlying metastasis requires complex and clinically relevant animal models for its study. The need for better models, and discussion as to what are the most relevant models available today were topics of intense debate during the first conference session. Yibin Kang (Princeton University, New Jersey, USA) identified miR-199a as a common regulator of stem cell activity both in normal mammary stem cells (MaSCs) and tumor-initiating cells (TICs)/cancer stem cells (CSCs) isolated from various breast cancer mouse models [ 9 ]. He showed that by suppressing a nuclear receptor corepressor LCOR, miR-199a protects normal MaSCs from interferonalpha (IFN-α)-mediated senescence and differentiation allowing the cells to retain self-renewal capacity. Importantly, Dr. Kang demonstrated that miR-199a-LCOR-IFN axis is active also in TIC/CSCs-rich ER- breast tumors, promotes tumor initiation and metastasis and predicts poor clinical outcome in ER- breast cancer patients. Altogether, Dr. Kang concluded that in mammary gland and immune cell-rich tumors, miR-199a-LCOR axis decreases the sensitivity of normal and malignant stem cells to autocrine/ paracrine interferon-mediated suppressive effects. Neta Erez (Tel Aviv University, Tel Aviv, Israel) presented a new model of spontaneous melanoma brain metastasis, which provides a platform to study the changes that occur in the brain metastatic niche and enables molecular detection of micrometastases [ 10 ]. Dr. Erez demonstrated that initiation of astrogliosis and neuroinflammation occur at early stages of metastases formation. Moreover, Dr. Erez showed that astrocytes play a major role in facilitating the growth of disseminated melanoma cells in the brain, presumably through the induction of a wound healing and astrogliosis response. This novel model can serve as a platform to study questions in the emergent field of the metastatic microenvironment, and to test new therapeutic combinations. Alberto Hernandez-Barranco (Spanish National Cancer Research Centre, Madrid, Spain) moved the focus to the role of exosomes in promoting lymph node (LN) metastases. He showed that exosomes derived from highly metastatic or lymphotropic melanoma tumor cells spread through the lymphatic system better than exosomes derived from low or non-metastatic tumor cells. Moreover, Hernandez-Barranco indicated that in the LN, most exosomes are incorporated into lymphatic endothelial cells (LECs) and macrophages. To elucidate the changes that follow exosome uptake, LECs were incubated with melanoma-derived exosomes. This resulted in an increase in the expression of lymphangiogenesis-related genes. With this, Hernandez-Barranco described a mechanism by which melanoma-secreted exosomes promote LN metastases formation via activation of the LN microenvironment. The focus of Jonathan P. Sleeman’s talk (University of Heidelberg, Mannheim, Germany) was the novel, lymphangiogenesis-independent role of the vascular endothelial growth factor-C (VEGF-C). Dr. Sleeman showed that VEGF-C is systemically increased in breast cancer patients compared to healthy controls. To mimic increased VEGF-C levels in circulation, he used adeno-associated viruses to stably elevate VEGF-C blood levels in various breast cancer animal models. He observed that systemic VEGF-C promotes metastasis formation by pre-metastatic conditioning of lungs via recruitment of CD11b + Gr1 + Ly6G + granulocytic MDSCs/neutrophils to the terminal bronchioles. Mechanistically, he showed that this process is dependent on VEGF-C receptor VEGFR-3 but not VEGFR-2. Collectively, Dr. Sleeman demonstrated that VEGF-C can foster metastasis independently of its effect on peritumoral lymphangiogenesis, changing our understanding of how VEGFC contributes to metastasis. Imaging metastasis Dr. Peter Friedl opened the “imaging metastasis” session by presenting important insights into cancer cell migration in vivo through the use of intravital multiphoton microscopy. By applying this advanced imaging technique to subcutaneous tumors of HT1080 (sarcoma) and MV3 (melanoma) cells in nude mice in a modified dorsal skin-fold chamber model, Dr. Friedl observed that the tumor invades 100–500 μm a day, and that in confined areas, the invading cancer cells undergo a jamming transition and move as individual cells or multicellular clusters [ 11 ]. Furthermore, during tumor progression, microniches existent in the tissue act as tracks that facilitate collective migration of cancer cells. Importantly, finger-like protrusions are also observed in human histology, matching these channel-like structures. By injecting dextran-FITC, he demonstrated that these channels are present in both healthy and cancerous tissues. In their experimental models, tumors were resected 3 weeks after tumor cells implantation, then spontaneous distant metastases were assessed on week 8. The three-step model of invasive cell migration states that tumor cells attach to the ECM, proteolytic degradation of the extracellular matrix (ECM) occurs, then tumor cells migrate through the remodelled ECM [ 12 ]. Contrary to this model, Friedl and co-workers found different types of invasion. They found that cancer cells expand against the tissue in a non-destructive way, suggesting that tissue geometry dictates the invasion pattern without a need for proteolytic degradation. They identified two types of collective invasion patterns in response to the connective tissue: linear collective invasion where the organ cells stay intact and confined; and broad or diffuse collective invasion, which is also confined but with higher perfusion. By using CLEM, an intravital fluorescence with 3D reconstruction, Dr. Friedl tracked expansion by leader cells that started invading out and spreading, while the tissue structure remained fully preserved. Dr. Friedl and his team applied pressure to the tumor to induce a passive position change, which resulted in the tumor invading into the surrounding tissue in the range of 500–600 μm, suggesting that the pre-formed tracks in the tissue act as highways that collective and single cell migration can follow in a process of guided migration. The in vivo imaging of PDX models would be the next step, but the fluorescent labeling is still a challenge. In addition to the environmental drive of cancer invasion and metastasis, Dr. Friedl also recognized the importance of cell adhesion. Although there is plenty of evidence about how integrins influence metastasis, there are cases where cells can move in an integrin-independent manner, such as leukocyte migration [ 13 ]. Dr. Friedl questioned whether integrins are required for attachment to the basement membrane. Since knock out of integrins in vitro is often lethal, they used a multi-targeted interference approach by combining RNAi and β1 and β3 integrins antibodies. Dr. Friedl and his team observed a reduction in adhesion force to collagen type I through atomic force spectroscopy. In vivo, tumor growth was critically affected. However, invasion was still possible. Thus, integrins are indispensable for tumor growth but dispensable for migration. Dr. Friedl highlighted that the speed of different migration modes remained mainly unperturbed. Absence of integrins has been widely related to a strong reduction in metastasis [ 14, 15 ], yet Dr. Friedl imaged cell dissemination and observed the same rate of metastatic events during lung colonization. Actually, when normalized by tumor mass, there is an enhanced probability of lung colonization after integrin targeting. Thus, reducing integrin availability potentially contributes to cell motility, metastasis and decreased tumor growth. These microenvironmental factors are important to improve anti-cancer therapies. For instance, MMP inhibitors may not efficiently prevent invasion, since the migratory channels are already shaped in the tissues. Combined anti-migration and cytotoxic therapies to combat metastatic transitions may be a better target. By definition, the extracellular matrix (ECM) is a collection of extracellular molecules that are secreted by cells, and its structure as well as its mechanical and physical properties influence cell fate and behaviour. Extensive changes in the cellular microenvironment can promote cancer development and contribute to metastatic dissemination. As Dr. Chris Madsen (Lund University, Sweden) highlighted, it is crucial to study and understand changes in tissue-specific ECM composition as this represent important an important regulatory component in cancer development and progression. He introduced a fast and efficient method to improve the analysis of native versus tumor ECM architecture, the so called ‘in situ decellularization of tissue’ (ISDoT) approach, which allows removal of all cells in a tissue, only leaving an extracellular matrix scaffold for further analysis. This technique is based on the delivery of decellularization reagents through the cardiovascular system of the mouse to any organ or anatomical area [ 16 ]. These 3D decellularized tissues maintain their original tissue-specific structure and typical composition, and can be used for high-resolution fluorescence imaging and for comparative quantitative proteomic analyses. This approach allows the comparison of the structure and composition of native, healthy ECM of a specific organ with the destructive, dramatically changed ECM found in macro-/micrometastatic tissue of the same organ. It also shows the vast structural differences between intratumoral fibers and the decellular ECM fibers in the adjacent stroma of macrometastatic lung tissue observed through fluorescence imaging. In summary, the ISDoT technique allows the preservation of typical ECM and basement membrane (BM) integrity of any organ of interest and can be used to quantify its biochemical ECM and BM composition. In addition, ISDoT can be used to map the 3D topology of the ECM at different time points of metastatic dissemination and organ colonisation, which will fundamentally improve the study and the understanding of ECM remodelling during the metastatic process. Proffered Paper 2 was presented by Adam Marcus and was entitled “using an image-guided genomics platform to probe the phenotypic heterogeneity of collective cancer invasion”. With the aim of investigating the fascinating phenotypic heterogeneity of cancer cell populations, Dr. Adam Marcus and his team created an image-guided genomics technique named spatiotemporal genomic and cellular analysis (SaGA). This advanced technique consists of a precise selection through the optical highlighting of cells of interest, followed by their analysis. Collective invasion (Ecad, SP-C) was observed in lung tissue from the Gilbert-Ross mouse model Kras, which is a Kkb1 metastatic lung adenocarcinoma GEMM using LV-Cre that promotes lung adenocarcinoma in 90% of the mice. The same collective invasion phenomenon was also observed in 3D spheroids [ 17 ]. In a video of an invading spheroid, Dr. Marcus showed that some cells initiate invasion by breaking through the matrix and seem to be followed by the other cells, forming cellular streams. When these starting cells move too far and lose contact with the spheroid body, the adjacent cells instantly stop following in 84% of the cases [ 17 ]. Therefore, invasion seems to be regulated by a specific class of cells that Dr. Marcus termed “leaders”. These leader cells can move back to the spheroid, allowing the follower cells to re-attach and follow them again [ 17 ]. To further investigate this hypothesis, Dr. Marcus and his team applied SaGA and after dissociating leader and follower cells by FACS, they observed that both populations retain their morphology and properties. While spheroids made of leader cells are highly invasive, follower cells remain packed and hardly invaded. When mixing both cell populations, leader cells remain leaders and rescue follower cells promoting their migration again [ 17 ]. Leader cells also appear to be more drug resistant. Interestingly, the two cell populations also behave differently in vivo. Leader cells are unable to establish a tumor at the injection site but appear to promote metastatic disease, whereas follower cells establish tumors and show significantly less metastasis. Subsequent RNASeq on these cell populations suggested an atypical VEGF-based vasculogenesis signaling that promotes the recruitment of follower cells, and a focal adhesion kinase-fibronectin signaling that facilitates the leader cell motility. The model proposed by Dr. Marcus is a symbiotic ecosystem model, where leader cells provide an escape mechanism to follower cells and in return, follower cells provide leader cells with increased growth and survival. Moreover, the phenotype of leader cells seems to be maintained through epigenetic reprogramming of the follower cells. Dr. Marcus is currently working towards applying this methodology and hypothesis to patient-derived organoids, with the aim of pursuing new therapies targeting this cooperation. Veronika te Boekhorst presented Proffered Paper 3, which had the title “Calpain-2 induced shutdown of β1 integrin function controls amoeboid reprogramming and dissemination under hypoxic challenge”. During metastatic dissemination, cancer cells can use a variety of single cell or collective migration strategies [ 18 ]. Plasticity of tumor cell migration can be affected by changes in cell adhesion, mechano-signalling adaptations, proteolysis and other microenvironment challenges [ 19 ]. The team of Dr. Peter Friedl has recently identified tumor hypoxia as a promoter of the dissemination of amoeboid-moving single cells from otherwise collectively invasive epithelial cells, the so called collective-to-amoeboid transition [ 19 ]. Dr. Veronika te Boekhorst (MD Anderson Cancer Center, Houston, USA) described how hypoxic stress and elevated levels of HIFs (hypoxia-inducible factors) in a 3D spheroid model of epithelial breast and squamous carcinoma cells induced a switch from collective invasion to an amoeboid-type of single cell movement. These cells are characterized by blebby or actin-rich pseudopodal protrusions that can initiate invasion. β1 integrins at the leading edge confer adhesion and contractility and reduced β1 integrin therefore is rate limiting for the switch from an elongated to an amoeboid migration phenotype. It was also observed that the cysteine protease calpain-2 controls β1 integrin shutdown as a consequence of hypoxia and HIF stabilization and mediates the blebby amoeboid migration switch. In vitro as well as in human squamous carcinoma xenografts, calpain-2 levels are high in reprogrammed and disseminated amoeboid-rounded cells upon HIF stabilization. The inhibition of calpain-2 expression or function can reverse the amoeboid to a mesenchymal migration type. Dr. te Boekhorst concluded that the HIF-induced and calpain-2-mediated integrin shutdown may represent a physiologically important escape mechanism of cells upon metabolic changes in the surrounding tissue areas. Overall, these results improve our understanding of this clinically relevant mechanism that is conserved in epithelial cancer cells and allows plasticity reprogramming by the microenvironment, eventually driving and promoting metastatic dissemination. Transgenic models of metastasis Mouse models to study cancer biology and metastatic dissemination have evolved significantly since the first transplantable tumor model in nude mice 50  years ago [ 20 ]. Currently, genetically engineered mouse models (GEMMs) are gaining favor over transplantable and cell inoculation models. Indeed, as opposed to models involving immunodeficient mice, GEMMs recapitulate all the steps in the metastatic process, including the interactions with the microenvironment and the immune system, which makes them appealing to use in preclinical studies. Dr. Karin de Visser (The Netherlands Cancer Institute, Amsterdam, Netherlands), in collaboration with Prof. Jos Jonkers’ group is using a conditional knockout breast cancer mouse model (K14::cre; Cdh1loxP/loxP; p53loxP/loxP) [ 21 ] to study tumor-induced systemic inflammation and metastasis. One of the limitations of GEMMs is the appearance of multiple tumors and tumor regrowth after excision of the primary, which hampers the study of metastasis. In order to uncouple tumor initiation from metastasis, primary tumors can be excised and implanted into new immunocompetent mice without losing their metastatic ability. This is known as a GEMM-derived syngraft. Using both GEMMs and GEMM-derived syngrafts, Dr Visser’s group has observed increased levels of circulating neutrophils in both models, which correlated with the metastatic tumor burden [ 22 ]. More importantly, clinical data show that breast cancer patients with elevated levels of neutrophils have worse prognosis. Conversely, neutrophil depletion with anti-Ly6G antibody significantly reduced lung metastases formation in mice. Hence, in an effort to discover the cancer cell genetic background responsible for neutrophil activation, they profiled a panel of 16 GEMMs with different tumor-initiating genetic modifications and observed elevated levels of neutrophils in Tp53 deficient mice. Further in vitro studies led to the discovery of interplay between Tp53 deficient tumor cells and macrophages, which trigger macrophages to release pro-inflammatory cytokines. The biggest challenge the metastasis research field currently faces is translating pre-clinical findings successfully into the clinic, as exemplified by the failure of several clinical trials with anti-angiogenic drugs. This was emphasized by Prof. Augustin, Prof. Kerbel and others. Prof. Augustin (German Cancer Research Center, Heidelberg, Germany) advocates the use of GEMMs with fully functioning immune systems, and studying the mechanisms that drive metastasis in a spatio-temporal manner, rather than focusing on endpoints. Through this approach, and in collaboration with Prof. Clare Isacke’s group, Prof. Augustin described how endosialin-positive pericytes facilitate breast cancer cells’ intravasation and consequent metastatic dissemintation [ 23 ]. With a series of examples of new GEMM-derived models, Prof. Augustin explained how our understanding of metastasis and the efficacy of clinical trials can be improved. The first strategy presented used a GEMMderived syngraft model (derived from the MT/ret mouse model of melanoma), which was employed to study antiangiogenic treatment with anti-angiopoietin2 antibody (anti-Ang2) in a neoadjuvant setting, as compared to adjuvant treatment [ 24 ]. Approximately 60% of mice treated in a neoadjuvant setting were completely free of metastases, while the remaining 40% did not show signs of lymph node dissemination. This suggests that anti-Ang2 prevents cancer cells from colonizing lymph nodes during tumor growth. This observation can be recapitulated by removal of draining lymph nodes upon primary tumor removal. The second strategy for focal tumor formation was local electroporation, which can be used to induce tumorigenesis. This was demonstrated using the BrafV600E; PtenloxP/loxP mouse melanoma model [ 25 ] and the local delivery of plasmids expressing Tyr:Cre. The tyrosinase promoter on these plasmids can be activated with UV light or depilation. The tumors can be removed and these mice will develop metastases several weeks after surgery. This approach can be extended to other models, such as hepatocellular carcinoma to study intra-hepatic metastases. It is worth mentioning that the two aforementioned strategies are not mutually exclusive but can rather be used in parallel. The advantage of the syngraft model is the capacity for scaling up the tumor material and allowing for biobanking, while the electroporation model can be employed to study the metastatic process from tumor initiation onward. In addition, understanding the mechanisms of how disseminated tumor cells grow in a specific secondary organ is another hurdle that GEMMs are managing to overcome. Metastases need a reciprocal adaptation between tumor and stromal cells at the metastatic niche that will allow the initial metastatic colonization and the progression of the disease. Dr. Shani, an MD PhD student in Dr. Neta Erez’ group (Tel Aviv University, Tel Aviv, Israel) has focused her research on cancer associated fibroblasts from the metastatic microenvironment that are the most abundant stromal cells in the lung [ 26 ]. She demonstrated that lung fibroblasts are essential for permissive metastatic-niche formation using an innovative transgenic mice model. Combining the MMTV-PyMT model of mammary carcinogenesis and the YFP-Collagen-1α (Col1α) expressing mouse model they generated the PyMT-Col1α-YFP model that develops an autochthonous mammary tumor with spontaneous lung metastases. Using these mice, they isolated lung fibroblasts at different stages of metastasis and found different gene expression patterns using RNA-sequencing. HIF1α, c-myc and EGF, among others, changed their expression in the early stage fibroblasts, whereas metastasis-associated fibroblasts presented changes in genes associated with inflammation, extracellular matrix remodeling and stress responses. Throughout this study, Dr. Shani and colleagues showed a dynamic co-evolution of lung fibroblasts in the metastatic microenvironment during breast cancer metastasis, emphasizing the importance of studying the gene-expression modifications in fibroblasts and the contribution these pathways may have in breast cancer lung metastasis development. Finally, reinforcing the importance of the metastatic stroma in metastasis evolution, Dr. Gelman (Roswell Park Cancer Institute, Buffalo, NY, USA) and his group studied the metastasis-regulating gene SSeCKS/AKAP12 which is down-regulated in many metastases compared to the primary tumor. To study the role of the kinase scaffolding protein SSeCKS/AKAP12 [ 27 ] in melanoma Dr. Gelman used a knock-out (KO) mouse model for SSeCKS where B16F10 melanoma cells (Brafwt) or their metastatic variant SM1WT1-LM3 (bearing Braf human melanoma mutation, BrafT1799A) were inoculated. They showed that the KOmice had increased peritoneal metastasis, which could be due to the increased levels of chemo-attractants in the peritoneal-fluid (PF). Two of the most abundant chemokines in the PF were CXCL9 and CXCL10, both CXCR3 ligands, and loss of the receptor inhibited both the chemotaxis and the metastatic ability in KO-mice. These results suggested that KO-mice peritonea secrete high levels of CXCR3 ligands, thereby facilitating melanoma cells chemoattraction and thus metastasis. Besides, they determined that these chemoattractants were secreted by KO peritoneal membrane fibroblasts (PMF) that showed a senescence-associated secretory phenotype (SASP). Finally, using SSeCKS scaffolding-site mutants and different kinase inhibitors, Dr. Gelman defined PKC, PKA and PI3K/AKT signalling pathways as CXCL10 expression regulators in KO PMF. With this work, Dr. Gelman and his group described the role of stromal SSeCKS protein as a metastasis-suppressor [ 28 ]. Loss of SSECKS/ AKAP12 increased secretion of chemokines and senescence factors by PMF that form the pre-metastatic niche and promoted the peritoneal metastatic chemotaxis of B16F10 melanoma cells. In summary, this session presented different mouse models that allow a more appropriate study of metastatic tumours due to their similarity with human disease, and reinforced the importance of using such models in cancer research. Modeling metastasis The final session of the meeting was dedicated to metastatic models and their relevance in answering specific research questions. First, Bin-Zhi Qian (University of Edinburgh, UK) discussed the mononuclear phagocyte system and its role in breast cancer metastasis, where they have previously shown the importance of host macrophages for lung metastasis [ 29, 30 ]. Dr Qian focused on breast cancer bone metastasis, where they found that the CCL2/CCR2 axis was critical in recruiting macrophages derived from circulating inflammatory monocytes. This event promoted the ability of breast cancer cells to metastasize to the bone. Furthermore, they found two sub-locations of macrophages in the bone marrow: the perivascular niche with resident macrophages, and the endosteal niche with osteal tissue macrophages that are important in metastatic seeding. These extravasating macrophages secrete VEGF, which increases vascular permeability, enabling tumor cells to extravasate and metastasize to the bone. The gene expression profile of inflammatory monocytes and macrophages showed that there are several macrophages subpopulations with different profiles. As it is well established that tumor cells do not work alone but need to interact with surrounding stromal cells such as fibroblasts. Erik Sahai (The Francis Crick Institute, London) focused on the interactions between fibroblasts and cancer cells in the primary tumor niche. Fibroblasts and cancer cells are mechanically coupled (through heterotypic interactions by E-cadherin and N-cadherin) enabling fibroblasts to pull cancer cells [ 31 ]. Dr Sahai also focused on the interaction between disseminated tumor cells and the stroma in the lung metastatic niche. Using the D2.0R and D2.A1 dormancy models they performed several co-cultures with different stromal populations of the lung. They focused on alveolar cells type I and II cells to try and identify growth arrest genes or survival repressors, and proliferation genes or survival inducers. They continue to investigate the critical role of the cross talk with epithelial cells in the lung parenchyma in lung metastasis. There is evidence that breast cancer metastases are seeded by cancer stem cells (CSCs), as Rachel Eyre (University of Manchester, UK) discussed in the final proffered paper of the meeting. Dr Eyre presented data from in vitro and in vivo models that demonstrated that IL1β-Wnt signalling promotes bone colonization by breast CSC. The inhibition of IL1β-Wnt signalling prevented bone metastases in vivo. This observation suggests that clinically available drugs that target this pathway should be considered as adjuvant therapy in breast cancer patients. Identifying biomarkers that predict the ability of tumors to metastasize is key goal in cancer research. Ulrike Stein (MDC, Germany) summarized extensive research carried out in her lab on MACC1, a novel gene identified as a potential target for metastasis therapy [ 32 ]. Since its discovery, it has been shown that MACC1 expression can be used as a prognostic marker for progression, metastasis and survival in several solid tumors [ 33 ], and MACC1 has been considered an anti-metastasis target. As Lovastatin (a member of the statin family) has been identified as an inhibitor of MACC1, it might be appropriate to use this drug for the treatment of patients with MACC1-induced metastasis who are at high risk for shorter survival. The closing keynote lecture was given by Andreas Trumpp (DKFZ, Germany), and was dedicated to new metastasis models and their uses in key areas of cancer research. Professor Trumpp focused initially in pancreatic adenocardinoma (PDAC), a poor prognosis cancer, and his use of PDX models to develop a stratification system for molecular subclassification [ 34 ]. Using these models, clinical outcome can be predicted, and CYP3A5 was identified as a key factor mediating drug resistance in PDAC. Next, Professor Trumpp described leukapheresis as a technique that can be used to isolate large numbers of circulating tumors cells (CTCs) from metastatic breast cancer patients. CTCs can be isolated for single cell RNAseq, and both cellular heterogeneity and response to therapy can be observed at the single cell level. He concluded that new technologies will prove powerful tools for preventing and treating cancer metastasis. Conclusions and perspectives The Seed and Soil conference 2017 was clearly an outstanding success, and exceeded the expectations of the organisers. In many conferences, the auditorium is full for the first couple of sessions then starts to empty out. This was not the case for the Seed and Soil conference, and the auditorium was packed throughout the meeting right up to the last talk. The quality of the questions and discussion during the sessions was highly informative and of excellent quality. The amount of interest in poster sessions is always a good measure of the success and relevance of the conference, and it is therefore also notable that the poster sessions were uniformly crammed full of attendees eager to talk to poster presenters, with discussions going well over the allotted time frame and into the night. The interest and enthusiasm engendered by the conference clearly illustrates the importance of the topic for the cancer research community. With this in mind, the conference organisers are already making plans for another Seed and Soil conference in 2 years time. 1 3 1. Francia G et al ( 2011 ) Mouse models of advanced spontaneous metastasis for experimental therapeutics . Nat Rev Cancer 11 ( 2 ): 135 - 141 2. Guerin E et al ( 2013 ) A model of postsurgical advanced metastatic breast cancer more accurately replicates the clinical efficacy of antiangiogenic drugs . Cancer Res 73 ( 9 ): 2743 - 2748 3. Kuczynski EA et al. ( 2016 ) Co-option of liver vessels and not sprouting angiogenesis drives acquired sorafenib resistance in hepatocellular carcinoma . J Natl Cancer Inst . https://doi. org/10.1093/jnci/djw030 4. Donnem T et al ( 2013 ) Vessel co-option in primary human tumors and metastases: an obstacle to effective anti-angiogenic treatment? Cancer Med 2 ( 4 ): 427 - 436 5. Gengenbacher N , Singhal M , Augustin HG ( 2017 ) Preclinical mouse solid tumour models: status quo, challenges and perspectives . Nat Rev Cancer 17 ( 12 ): 751 - 765 6. Hu G et al ( 2009 ) MTDH activation by 8q22 genomic gain promotes chemoresistance and metastasis of poor-prognosis breast cancer . Cancer Cell 15 ( 1 ): 9 - 20 7. Hu G , Wei Y , Kang Y ( 2009 ) The multifaceted role of MTDH/ AEG-1 in cancer progression . Clin Cancer Res 15 ( 18 ): 5615 - 5620 8. Wan L et al ( 2014 ) MTDH-SND1 interaction is crucial for expansion and activity of tumor-initiating cells in diverse oncogene- and carcinogen-induced mammary tumors . Cancer Cell 26 ( 1 ): 92 - 105 9. Celià-Terrassa T , Liu DD , Choudhury A , Hang X , Wei Y , Zamalloa J , Alfaro-Aco R , Chakrabarti R , Jiang YZ , Koh BI , Smith HA , DeCoste C , Li JJ , Shao ZM , Kang Y ( 2017 ) Normal and cancerous mammary stem cells evade interferon-induced constraint through the miR-199a-LCOR axis . Nat Cell Biol 19 ( 6 ): 711 - 723 10. Schwartz H , Blacher E , Amer M , Livneh N , Abramovitz L , Klein A , Ben-Shushan D , Soffer S , Blazquez R , Barrantes-Freer A , Müller M , Müller-Decker K , Stein R , Tsarfaty G , Satchi-Fainaro R , Umansky V , Pukrop T , Erez N ( 2016 ) Incipient melanoma brain metastases instigate astrogliosis and neuroinflammation . Cancer Res 1 76 ( 15 ): 4359 - 4371 11. Haeger A , Krause M , Wolf K , Friedl P ( 2014 ) Cell jamming: collective invasion of mesenchymal tumor cells imposed by tissue confinement . Biochim Biophys Acta Gen Subj 1840 ( 8 ): 2386 - 2395 . https://doi.org/10.1016/j.bbagen. 2014 . 03 .020 12. Liotta LA ( 2016 ) Adhere, degrade, and move: the three-step model of invasion . Cancer Res . https://doi.org/10.1158/ 0008 - 5472 .CAN- 16 -1297 13. Lämmermann T , Bader BL , Monkley SJ , Worbs T , Wedlich-Söldner R , Hirsch K , Sixt M ( 2008 ) Rapid leukocyte migration by integrin-independent flowing and squeezing . Nature 453 ( 7191 ): 51 - 55 . https://doi.org/10.1038/nature06887 14. Grzesiak JJ , Cao HST , Burton DW , Kaushal S , Vargas F , Clopton P , Bouvet M ( 2011 ) Knockdown of the β1 integrin subunit reduces primary tumor growth and inhibits pancreatic cancer metastasis . Int J Cancer 129 ( 12 ): 2905 - 2915 . https://doi.org/10.1002/ ijc.25942 15. Ju JA , Godet I , Ye IC , Byun J , Jayatilaka H , Lee SJ , Gilkes DM ( 2017 ) Hypoxia selectively enhances integrin α 5 β 1 receptor expression in breast cancer to promote metastasis . Mol Cancer Res 15 ( 6 ): 723 - 734 . https://doi.org/10.1158/ 1541 - 7786 .MCR- 16 -0338 16. Mayorca-Guiliani AE , Madsen CD , Cox TR , Horton ER , Venning FA , Erler JT ( 2017 ) ISDoT: in situ decellularization of tissues for high-resolution imaging and proteomic analysis of native extracellular matrix . Nat Med 23 : 890 - 898 17. Konen J , Summerbell E , Dwivedi B , Galior K , Hou Y , Rusnak L , Marcus AI ( 2017 ). Image-guided genomics of phenotypically heterogeneous populations reveals vascular signalling during symbiotic collective cancer invasion . Nat Commun . https://doi. org/10.1038/ncomms15078 18. Friedl P , Locker J , Sahai E , Segall JE ( 2012 ) Classifying collective cancer cell invasion . Nat Cell Biol 14 : 777 - 783 19. Lehmann S , te Boekhorst V , Odenthal J , Bianchi R , van Helvert S , Ikenberg K , Ilina O , Stoma S , Xandry J , Jiang L , Grenman R , Rudin M , Friedl P ( 2017 ) Hypoxia induces a HIF-1-dependent transition from collective-to-amoeboid dissemination in epithelial cancer cells . Curr Biol 27 : 392 - 400 20. Kersten K , de Visser KE , van Miltenburg MH , Jonkers J ( 2017 ) Genetically engineered mouse models in oncology research and cancer medicine . EMBO Mol Med 9 : 137 - 153 21. Derksen PWB et al ( 2006 ) Somatic inactivation of E-cadherin and p53 in mice leads to metastatic lobular mammary carcinoma through induction of anoikis resistance and angiogenesis . Cancer Cell 10 : 437 - 449 22. Coffelt SB et al ( 2015 ) IL-17-producing γδ T cells and neutrophils conspire to promote breast cancer metastasis . Nature 522 : 345 - 348 23. Viski C et al ( 2016 ) Endosialin-expressing pericytes promote metastatic dissemination . Cancer Res 76 : 5313 - 5325 24. Srivastava K et al ( 2014 ) Postsurgical adjuvant tumor therapy by combining anti-angiopoietin-2 and metronomic chemotherapy limits metastatic growth . Cancer Cell 26 : 880 - 895 25. Dankort D et al ( 2009 ) BrafV600E cooperates with Pten loss to induce metastatic melanoma . Nat Genet 41 : 544 - 552 26. Cohen N , Shani O , Raz Y , Sharon Y , Hoffman D , Abramovitz L et al ( 2017 ) Fibroblasts drive an immunosuppressive and growthpromoting microenvironment in breast cancer via secretion of Chitinase 3-like 1 . Oncogene 36 ( 31 ): 4457 - 4468 27. Su B , Gao L , Meng F , Guo L , Rothschild J , Gelman IH . Adhesionmediated cytoskeletal remodeling is controlled by the direct scaffolding of Src from FAK complexes to lipid rafts by SSeCKS/ AKAP12 . Oncogene 32 ( 16 ): 2016 - 2026 28. Muramatsu M , Gao L , Peresie J , Balderman B , Akakura S , Gelman IH ( 2017 ) SSeCKS/AKAP12 scaffolding functions suppress B16F10-induced peritoneal metastasis by attenuating CXCL9/10 secretion by resident fibroblasts . Oncotarget 8 ( 41 ): 70281 - 70298 29. Qian BZ et al ( 2011 ) CCL2 recruits inflammatory monocytes to facilitate breast-tumour metastasis . Nature 475 ( 7355 ): 222 - 225 30. Qian BZ et al ( 2015 ) FLT1 signaling in metastasis-associated macrophages activates an inflammatory signature that promotes breast cancer metastasis . J Exp Med 212 ( 9 ): 1433 - 1448 31. Labernadie A et al. ( 2017 ) A mechanically active heterotypic E-cadherin/N-cadherin adhesion enables fibroblasts to drive cancer cell invasion . Nat Cell Biol 19 ( 3 ): 224 - 37 32. Stein U et al ( 2009 ) MACC1, a newly identified key regulator of HGF-MET signaling, predicts colon cancer metastasis . Nat Med 15 ( 1 ): 59 - 67 33. Rohr UP et  al ( 2017 ) Prognostic value of MACC1 and proficient mismatch repair status for recurrence risk prediction in stage II colon cancer patients: the BIOGRID studies . Ann Oncol 28 ( 8 ): 1869 - 1875 34. Noll EM et al. ( 2016 ) CYP3A5 mediates basal and acquired therapy resistance in different subtypes of pancreatic ductal adenocarcinoma . Nat Med 22 ( 3 ): 278 - 287

This is a preview of a remote PDF: https://link.springer.com/content/pdf/10.1007%2Fs10585-018-9886-x.pdf

I. Teles Alves, N. Cohen, P. G. Ersan, R. Eyre, I. Godet, D. Holovanchuk, R. Jackstadt, L. Kyjacova, K. Mahal, A. Noguera-Castells, L. Recalde-Percaz, J. P. Sleeman. EACR-MRS conference on Seed and Soil: In Vivo Models of Metastasis, Clinical & Experimental Metastasis, 2018, 449-456, DOI: 10.1007/s10585-018-9886-x