An in vitro model of cancer invasion with heterogeneous ECM created with droplet microfluidics.

Metastasis is a multi-step process that is critically affected by cues from the tumor micro-environment (TME), such as from the extracellular matrix (ECM). The role of the ECM in the onset of metastasis, invasion, is not yet fully understood. A further complicating factor is that the ECM in the TME is mostly heterogeneous, in particular presenting a basement membrane (BM) directly enveloping the tumor, which acts as a barrier to invasion into the surrounding stromal ECM. To systematically investigate the role of ECM in invasion, appropriate in vitro models with control over such ECM heterogeneity are essential. We present a novel high-throughput microfluidic approach to build such a model, which enables to capture the invasion of cancer cells from the tumor, through the BM and into the stromal tissue. We used a droplet-maker device to encapsulate cells in beads of a primary hydrogel mimicking BM, Matrigel, which were then embedded in a secondary hydrogel mimicking stromal ECM, collagen I. Our technology ultimately provides control over parameters such as tissue size, cell count and type, and ECM composition and stiffness. As a proof-of-principle, we carried out a comparative study with two breast cancer cell types, and we observed typical behavior consistent with previous studies. Highly invasive MDA-MB-231 cells showed single cell invasion behavior, whereas poorly invasive MCF-7 cells physically penetrated the surrounding matrix collectively. A comparative analysis conducted between our heterogeneous model and previous models employing a single type of hydrogel, either collagen I or Matrigel, has unveiled a substantial difference in terms of cancer cell invasion distance. Our in vitro model resembles an in vivo heterogeneous cancer microenvironment and can potentially be used for high throughput studies of cancer invasion.

Environmental stiffness restores mechanical homeostasis in vimentin-depleted cells.

Recent experimental evidence indicates a role for the intermediate filament vimentin in regulating cellular mechanical homeostasis, but its precise contribution remains to be discovered. Mechanical homeostasis requires a balanced bi-directional interplay between the cell's microenvironment and the cellular morphological and mechanical state-this balance being regulated via processes of mechanotransduction and mechanoresponse, commonly referred to as mechanoreciprocity. Here, we systematically analyze vimentin-expressing and vimentin-depleted cells in a swatch of in vitro cellular microenvironments varying in stiffness and/or ECM density. We find that vimentin-expressing cells maintain mechanical homeostasis by adapting cellular morphology and mechanics to micromechanical changes in the microenvironment. However, vimentin-depleted cells lose this mechanoresponse ability on short timescales, only to reacquire it on longer time scales. Indeed, we find that the morphology and mechanics of vimentin-depleted cell in stiffened microenvironmental conditions can get restored to the homeostatic levels of vimentin-expressing cells. Additionally, we observed vimentin-depleted cells increasing collagen matrix synthesis and its crosslinking, a phenomenon which is known to increase matrix stiffness, and which we now hypothesize to be a cellular compensation mechanism for the loss of vimentin. Taken together, our findings provide further insight in the regulating role of intermediate filament vimentin in mediating mechanoreciprocity and mechanical homeostasis.

SFAlab: image-based quantification of mechano-active ventral actin stress fibers in adherent cells.

Ventral actin stress fibers (SFs) are a subset of actin SFs that begin and terminate at focal adhesion (FA) complexes. Ventral SFs can transmit forces from and to the extracellular matrix and serve as a prominent mechanosensing and mechanotransduction machinery for cells. Therefore, quantitative analysis of ventral SFs can lead to deeper understanding of the dynamic mechanical interplay between cells and their extracellular matrix (mechanoreciprocity). However, the dynamic nature and organization of ventral SFs challenge their quantification, and current quantification tools mainly focus on all SFs present in cells and cannot discriminate between subsets. Here we present an image analysis-based computational toolbox, called SFAlab, to quantify the number of ventral SFs and the number of ventral SFs per FA, and provide spatial information about the locations of the identified ventral SFs. SFAlab is built as an all-in-one toolbox that besides analyzing ventral SFs also enables the identification and quantification of (the shape descriptors of) nuclei, cells, and FAs. We validated SFAlab for the quantification of ventral SFs in human fetal cardiac fibroblasts and demonstrated that SFAlab analysis i) yields accurate ventral SF detection in the presence of image imperfections often found in typical fluorescence microscopy images, and ii) is robust against user subjectivity and potential experimental artifacts. To demonstrate the usefulness of SFAlab in mechanobiology research, we modulated actin polymerization and showed that inhibition of Rho kinase led to a significant decrease in ventral SF formation and the number of ventral SFs per FA, shedding light on the importance of the RhoA pathway specifically in ventral SF formation. We present SFAlab as a powerful open source, easy to use image-based analytical tool to increase our understanding of mechanoreciprocity in adherent cells.

Shape-Morphing Photoresponsive Hydrogels Reveal Dynamic Topographical Conditioning of Fibroblasts.

The extracellular environment defines a physical boundary condition with which cells interact. However, to date, cell response to geometrical environmental cues is largely studied in static settings, which fails to capture the spatiotemporally varying cues cells receive in native tissues. Here, a photoresponsive spiropyran-based hydrogel is presented as a dynamic, cell-compatible, and reconfigurable substrate. Local stimulation with blue light (455 nm) alters hydrogel swelling, resulting in on-demand reversible micrometer-scale changes in surface topography within 15 min, allowing investigation into cell response to controlled geometry actuations. At short term (1 h after actuation), fibroblasts respond to multiple rounds of recurring topographical changes by reorganizing their nucleus and focal adhesions (FA). FAs form primarily at the dynamic regions of the hydrogel; however, this propensity is abolished when the topography is reconfigured from grooves to pits, demonstrating that topographical changes dynamically condition fibroblasts. Further, this dynamic conditioning is found to be associated with long-term (72 h) maintenance of focal adhesions and epigenetic modifications. Overall, this study offers a new approach to dissect the dynamic interplay between cells and their microenvironment and shines a new light on the cell's ability to adapt to topographical changes through FA-based mechanotransduction.

A multiscale computational model of arterial growth and remodeling including Notch signaling.

Blood vessels grow and remodel in response to mechanical stimuli. Many computational models capture this process phenomenologically, by assuming stress homeostasis, but this approach cannot unravel the underlying cellular mechanisms. Mechano-sensitive Notch signaling is well-known to be key in vascular development and homeostasis. Here, we present a multiscale framework coupling a constrained mixture model, capturing the mechanics and turnover of arterial constituents, to a cell-cell signaling model, describing Notch signaling dynamics among vascular smooth muscle cells (SMCs) as influenced by mechanical stimuli. Tissue turnover was regulated by both Notch activity, informed by in vitro data, and a phenomenological contribution, accounting for mechanisms other than Notch. This novel framework predicted changes in wall thickness and arterial composition in response to hypertension similar to previous in vivo data. The simulations suggested that Notch contributes to arterial growth in hypertension mainly by promoting SMC proliferation, while other mechanisms are needed to fully capture remodeling. The results also indicated that interventions to Notch, such as external Jagged ligands, can alter both the geometry and composition of hypertensive vessels, especially in the short term. Overall, our model enables a deeper analysis of the role of Notch and Notch interventions in arterial growth and remodeling and could be adopted to investigate therapeutic strategies and optimize vascular regeneration protocols.

Notch signaling regulates strain-mediated phenotypic switching of vascular smooth muscle cells.

Mechanical stimuli experienced by vascular smooth muscle cells (VSMCs) and mechanosensitive Notch signaling are important regulators of vascular growth and remodeling. However, the interplay between mechanical cues and Notch signaling, and its contribution to regulate the VSMC phenotype are still unclear. Here, we investigated the role of Notch signaling in regulating strain-mediated changes in VSMC phenotype. Synthetic and contractile VSMCs were cyclically stretched for 48 h to determine the temporal changes in phenotypic features. Different magnitudes of strain were applied to investigate its effect on Notch mechanosensitivity and the phenotypic regulation of VSMCs. In addition, Notch signaling was inhibited via DAPT treatment and activated with immobilized Jagged1 ligands to understand the role of Notch on strain-mediated phenotypic changes of VSMCs. Our data demonstrate that cyclic strain induces a decrease in Notch signaling along with a loss of VSMC contractile features. Accordingly, the activation of Notch signaling during cyclic stretching partially rescued the contractile features of VSMCs. These findings demonstrate that Notch signaling has an important role in regulating strain-mediated phenotypic switching of VSMCs.

Computational analysis of the role of mechanosensitive Notch signaling in arterial adaptation to hypertension.

Arteries grow and remodel in response to mechanical stimuli. Hypertension, for example, results in arterial wall thickening. Cell-cell Notch signaling between vascular smooth muscle cells (VSMCs) is known to be involved in this process, but the underlying mechanisms are still unclear. Here, we investigated whether Notch mechanosensitivity to strain may regulate arterial thickening in hypertension. We developed a multiscale computational framework by coupling a finite element model of arterial mechanics, including residual stress, to an agent-based model of mechanosensitive Notch signaling, to predict VSMC phenotypes as an indicator of growth and remodeling. Our simulations revealed that the sensitivity of Notch to strain at mean blood pressure may be a key mediator of arterial thickening in hypertensive arteries. Further simulations showed that loss of residual stress can have synergistic effects with hypertension, and that changes in the expression of Notch receptors, but not Jagged ligands, may be used to control arterial growth and remodeling and to intensify or counteract hypertensive thickening. Overall, we identify Notch mechanosensitivity as a potential mediator of vascular adaptation, and we present a computational framework that can facilitate the testing of new therapeutic and regenerative strategies.

Engineered patterns of Notch ligands Jag1 and Dll4 elicit differential spatial control of endothelial sprouting.

Spatial regulation of angiogenesis is important for the generation of functional engineered vasculature in regenerative medicine. The Notch ligands Jag1 and Dll4 show distinct expression patterns in endothelial cells and, respectively, promote and inhibit endothelial sprouting. Therefore, patterns of Notch ligands may be utilized to spatially control sprouting, but their potential and the underlying mechanisms of action are unclear. Here, we coupled in vitro and in silico models to analyze the ability of micropatterned Jag1 and Dll4 ligands to spatially control endothelial sprouting. Dll4 patterns, but not Jag1 patterns, elicited spatial control. Computational simulations of the underlying signaling dynamics suggest that different timing of Notch activation by Jag1 and Dll4 underlie their distinct ability to spatially control sprouting. Hence, Dll4 patterns efficiently direct the sprouts, whereas longer exposure to Jag1 patterns is required to achieve spatial control. These insights in sprouting regulation offer therapeutic handles for spatial regulation of angiogenesis.

Engineering tissue morphogenesis: taking it up a Notch.

Recreating functional tissues through bioengineering strategies requires steering of complex cell fate decisions. Notch, a juxtacrine signaling pathway, regulates cell fate and controls cellular organization with local precision. The engineering-friendly characteristics of the Notch pathway provide handles for engineering tissue patterning and morphogenesis. We discuss the physiological significance and mechanisms of Notch signaling with an emphasis on its potential use for engineering complex tissues. We highlight the current state of the art of Notch activation and provide a view on the design aspects, opportunities, and challenges in modulating Notch for tissue-engineering strategies. We propose that finely tuned control of Notch contributes to the generation of tissues with accurate form and functionality.

Mechano-regulated cell-cell signaling in the context of cardiovascular tissue engineering.

Cardiovascular tissue engineering (CVTE) aims to create living tissues, with the ability to grow and remodel, as replacements for diseased blood vessels and heart valves. Despite promising results, the (long-term) functionality of these engineered tissues still needs improvement to reach broad clinical application. The functionality of native tissues is ensured by their specific mechanical properties directly arising from tissue organization. We therefore hypothesize that establishing a native-like tissue organization is vital to overcome the limitations of current CVTE approaches. To achieve this aim, a better understanding of the growth and remodeling (G&R) mechanisms of cardiovascular tissues is necessary. Cells are the main mediators of tissue G&R, and their behavior is strongly influenced by both mechanical stimuli and cell-cell signaling. An increasing number of signaling pathways has also been identified as mechanosensitive. As such, they may have a key underlying role in regulating the G&R of tissues in response to mechanical stimuli. A more detailed understanding of mechano-regulated cell-cell signaling may thus be crucial to advance CVTE, as it could inspire new methods to control tissue G&R and improve the organization and functionality of engineered tissues, thereby accelerating clinical translation. In this review, we discuss the organization and biomechanics of native cardiovascular tissues; recent CVTE studies emphasizing the obtained engineered tissue organization; and the interplay between mechanical stimuli, cell behavior, and cell-cell signaling. In addition, we review past contributions of computational models in understanding and predicting mechano-regulated tissue G&R and cell-cell signaling to highlight their potential role in future CVTE strategies.

Ex Vivo Models to Decipher the Molecular Mechanisms of Genetic Notch Cardiovascular Disorders.

Notch is an evolutionary, conserved, cell-cell signaling pathway that is central to several biological processes, from tissue morphogenesis to homeostasis. It is therefore not surprising that several genetic mutations of Notch components cause inherited human diseases, especially cardiovascular disorders. Despite numerous efforts, current in vivo models are still insufficient to unravel the underlying mechanisms of these pathologies, hindering the development of utmost needed medical therapies. In this perspective review, we discuss the limitations of current murine models and outline how the combination of microphysiological systems (MPSs) and targeted computational models can lead to breakthroughs in this field. In particular, while MPSs enable the experimentation on human cells in controlled and physiological environments, in silico models can provide a versatile tool to translate the in vitro findings to the more complex in vivo setting. As a showcase example, we focus on Notch-related cardiovascular diseases, such as Alagille syndrome, Adams-Oliver syndrome, and cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL). Impact statement In this review, a comprehensive overview of the limitations of current in vivo models of genetic Notch cardiovascular diseases is provided, followed by a discussion over the potential of microphysiological systems and computational models in overcoming these limitations and in potentiating drug testing and modeling of these pathologies.

Lateral induction limits the impact of cell connectivity on Notch signaling in arterial walls.

It is well known that arteries grow and remodel in response to mechanical stimuli. Vascular smooth muscle cells are the main mediators of this process, as they can switch phenotype from contractile to synthetic, and vice-versa, based on the surrounding bio-chemo-mechanical stimuli. A correct regulation of this phenotypic switch is fundamental to obtain and maintain arterial homeostasis. Notch, a mechanosensitive signaling pathway, is one of the main regulators of the vascular smooth muscle cell phenotype. Therefore, understanding Notch dynamics is key to elucidate arterial growth, remodeling, and mechanobiology. We have recently developed a one-dimensional agent-based model to investigate Notch signaling in arteries. However, due to its one-dimensional formulation, the model cannot be adopted to study complex nonsymmetrical geometries and, importantly, it cannot capture the realistic "cell connectivity" in arteries, here defined as the number of cell neighbors. Notch functions via direct cell-cell contact; thus, the number of cell neighbors could be an essential feature of Notch dynamics. Here, we extended the agent-based model to a two-dimensional formulation, to investigate the effects of cell connectivity on Notch dynamics and cell phenotypes in arteries. The computational results, supported by a sensitivity analysis, indicate that cell connectivity has marginal effects when Notch dynamics is dominated by the process of lateral induction, which induces all cells to have a uniform phenotype. When lateral induction is weaker, cells exhibit a nonuniform phenotype distribution and the percentage of synthetic cells within an artery depends on the number of neighbors.

Notch in mechanotransduction - from molecular mechanosensitivity to tissue mechanostasis.

Tissue development and homeostasis are controlled by mechanical cues. Perturbation of the mechanical equilibrium triggers restoration of mechanostasis through changes in cell behavior, while defects in these restorative mechanisms lead to mechanopathologies, for example, osteoporosis, myopathies, fibrosis or cardiovascular disease. Therefore, sensing mechanical cues and integrating them with the biomolecular cell fate machinery is essential for the maintenance of health. The Notch signaling pathway regulates cell and tissue fate in nearly all tissues. Notch activation is directly and indirectly mechanosensitive, and regulation of Notch signaling, and consequently cell fate, is integral to the cellular response to mechanical cues. Fully understanding the dynamic relationship between molecular signaling, tissue mechanics and tissue remodeling is challenging. To address this challenge, engineered microtissues and computational models play an increasingly large role. In this Review, we propose that Notch takes on the role of a 'mechanostat', maintaining the mechanical equilibrium of tissues. We discuss the reciprocal role of Notch in the regulation of tissue mechanics, with an emphasis on cardiovascular tissues, and the potential of computational and engineering approaches to unravel the complex dynamic relationship between mechanics and signaling in the maintenance of cell and tissue mechanostasis.

Vimentin regulates Notch signaling strength and arterial remodeling in response to hemodynamic stress.

The intermediate filament (IF) cytoskeleton has been proposed to regulate morphogenic processes by integrating the cell fate signaling machinery with mechanical cues. Signaling between endothelial cells (ECs) and vascular smooth muscle cells (VSMCs) through the Notch pathway regulates arterial remodeling in response to changes in blood flow. Here we show that the IF-protein vimentin regulates Notch signaling strength and arterial remodeling in response to hemodynamic forces. Vimentin is important for Notch transactivation by ECs and vimentin knockout mice (VimKO) display disrupted VSMC differentiation and adverse remodeling in aortic explants and in vivo. Shear stress increases Jagged1 levels and Notch activation in a vimentin-dependent manner. Shear stress induces phosphorylation of vimentin at serine 38 and phosphorylated vimentin interacts with Jagged1 and increases Notch activation potential. Reduced Jagged1-Notch transactivation strength disrupts lateral signal induction through the arterial wall leading to adverse remodeling. Taken together we demonstrate that vimentin forms a central part of a mechanochemical transduction pathway that regulates multilayer communication and structural homeostasis of the arterial wall.

A Supramolecular Platform for the Introduction of Fc-Fusion Bioactive Proteins on Biomaterial Surfaces.

Bioorthogonal chemistry is an excellent method for functionalization of biomaterials with bioactive molecules, as it allows for decoupling of material processing and bioactivation. Here, we report on a modular system created by means of tetrazine/trans-cyclooctene (Tz/TCO) click chemistry undergoing an inverse electron demand Diels-Alder cycloaddition. A reactive supramolecular surface based on ureido-pyrimidinones (UPy) is generated via a UPy-Tz additive, in order to introduce a versatile TCO-protein G conjugate for immobilization of Fc-fusion proteins. As a model bioactive protein, we introduced Fc-Jagged1, a Notch ligand, to induce Notch signaling activity on the material. Interestingly, HEK293 FLN1 cells expressing the Notch1 receptor were repelled by films modified with TCO-protein G but adhered and spread on functionalized electrospun meshes. This indicates that the material processing method influences the biocompatibility of the postmodification. Notch signaling activity was upregulated 5.6-fold with respect to inactive controls on electrospun materials modified with TCO-protein G/Fc-Jagged1. Furthermore, downstream effects of Notch signaling were detected on the gene level in vascular smooth muscle cells expressing the Notch3 receptor. Taken together, our results demonstrate the successful use of a modular supramolecular system for the postprocessing modification of solid materials with functional proteins.

Influence of the Assembly State on the Functionality of a Supramolecular Jagged1-Mimicking Peptide Additive.

Expanding the bioactivation toolbox of supramolecular materials is of utmost relevance for their broad applicability in regenerative medicines. This study explores the functionality of a peptide mimic of the Notch ligand Jagged1 in a supramolecular system that is based on hydrogen bonding ureido-pyrimidinone (UPy) units. The functionality of the peptide is studied when formulated as an additive in a supramolecular solid material and as a self-assembled system in solution. UPy conjugation of the DSLJAG1 peptide sequence allows for the supramolecular functionalization of UPy-modified polycaprolactone, an elastomeric material, with UPy-DSLJAG1. Surface presentation of the UPy-DSLJAG1 peptide was confirmed by atomic force microscopy and X-ray photoelectron spectroscopy analyses, but no enhancement of Notch activity was detected in cells presenting Notch1 and Notch3 receptors. Nevertheless, a significant increase in Notch-signaling activity was observed when DSLJAG1 peptides were administered in the soluble form, indicating that the activity of DSLJAG1 is preserved after UPy functionalization but not after immobilization on a supramolecular solid material. Interestingly, an enhanced activity in solution of the UPy conjugate was detected compared with the unconjugated DSLJAG1 peptide, suggesting that the self-assembly of supramolecular aggregates in solution ameliorates the functionality of the molecules in a biological context.

MDA-MB-231 Breast Cancer Cells and Their CSC Population Migrate Towards Low Oxygen in a Microfluidic Gradient Device.

Most cancer deaths are caused by secondary tumors formed through metastasis, yet due to our limited understanding of this process, prevention remains a major challenge. Recently, cancer stem cells (CSCs) have been proposed as the source of metastases, but only little is known about their migratory behavior. Oxygen gradients in the tumor have been linked to directional migration of breast cancer cells. Here, we present a method to study the effect of oxygen gradients on the migratory behavior of breast CSCs using a microfluidic device. Our chip contains a chamber in which an oxygen gradient can be generated between hypoxic (<1%) and ambient (21%) conditions. We tracked the migration of CSCs obtained from MDA-MB-231 breast cancer cells, and found that their migration patterns do not differ from the average MDA-MB-231 population. Surprisingly, we found that the cells migrate towards low oxygen levels, in contrast with an earlier study. We hypothesize that in our device, migration is exclusively due to the pure oxygen gradient, whereas the effects of oxygen in earlier work were obscured by additional cues from the tumor microenvironment (e.g., nutrients and metabolites). These results open new research directions into the role of oxygen in directing cancer and CSC migration.

A biomimetic microfluidic model to study signalling between endothelial and vascular smooth muscle cells under hemodynamic conditions.

Cell signalling and mechanics influence vascular pathophysiology and there is an increasing demand for in vitro model systems that enable examination of signalling between vascular cells under hemodynamic conditions. Current 3D vessel wall constructs do not recapitulate the mechanical conditions of the native tissue nor do they allow examination of cell-cell interactions under relevant hemodynamic conditions. Here, we describe a 3D microfluidic chip model of arterial endothelial and smooth muscle cells where cellular organization, composition and interactions, as well as the mechanical environment of the arterial wall are mimicked. The hemodynamic EC-VSMC-signalling-on-a-chip consists of two parallel polydimethylsiloxane (PDMS) cell culture channels, separated by a flexible, porous PDMS membrane, mimicking the porosity of the internal elastic lamina. The hemodynamic EC-VSMC-signalling-on-a-chip allows co-culturing of human aortic endothelial cells (ECs) and human aortic vascular smooth muscle cells (VSMCs), separated by a porous membrane, which enables EC-VSMC interaction and signalling, crucial for the development and homeostasis of the vessel wall. The device allows real time cell imaging and control of hemodynamic conditions. The culture channels are surrounded on either side by vacuum channels to induce cyclic strain by applying cyclic suction, resulting in mechanical stretching and relaxation of the membrane in the cell culture channels. The blood flow is mimicked by creating a flow of medium at the EC side. Vascular cells remain viable during prolonged culturing, exhibit physiological morphology and organization and make cell-cell contact. During dynamic culturing of the device with a shear stress of 1-1.5 Pa and strain of 5-8%, VSMCs align perpendicular to the given strain in the direction of the flow and EC adopt a cobblestone morphology. To our knowledge, this is the first report on the development of a microfluidic device, which enables a co-culture of interacting ECs and VSMCs under hemodynamic conditions and presents a novel approach to systematically study the biological and mechanical components of the intimal-medial vascular unit.

Mechanosensitivity of Jagged-Notch signaling can induce a switch-type behavior in vascular homeostasis.

Hemodynamic forces and Notch signaling are both known as key regulators of arterial remodeling and homeostasis. However, how these two factors integrate in vascular morphogenesis and homeostasis is unclear. Here, we combined experiments and modeling to evaluate the impact of the integration of mechanics and Notch signaling on vascular homeostasis. Vascular smooth muscle cells (VSMCs) were cyclically stretched on flexible membranes, as quantified via video tracking, demonstrating that the expression of Jagged1, Notch3, and target genes was down-regulated with strain. The data were incorporated in a computational framework of Notch signaling in the vascular wall, where the mechanical load was defined by the vascular geometry and blood pressure. Upon increasing wall thickness, the model predicted a switch-type behavior of the Notch signaling state with a steep transition of synthetic toward contractile VSMCs at a certain transition thickness. These thicknesses varied per investigated arterial location and were in good agreement with human anatomical data, thereby suggesting that the Notch response to hemodynamics plays an important role in the establishment of vascular homeostasis.

Metastasis in context: modeling the tumor microenvironment with cancer-on-a-chip approaches.

Most cancer deaths are not caused by the primary tumor, but by secondary tumors formed through metastasis, a complex and poorly understood process. Cues from the tumor microenvironment, such as the biochemical composition, cellular population, extracellular matrix, and tissue (fluid) mechanics, have been indicated to play a pivotal role in the onset of metastasis. Dissecting the role of these cues from the tumor microenvironment in a controlled manner is challenging, but essential to understanding metastasis. Recently, cancer-on-a-chip models have emerged as a tool to study the tumor microenvironment and its role in metastasis. These models are based on microfluidic chips and contain small chambers for cell culture, enabling control over local gradients, fluid flow, tissue mechanics, and composition of the local environment. Here, we review the recent contributions of cancer-on-a-chip models to our understanding of the role of the tumor microenvironment in the onset of metastasis, and provide an outlook for future applications of this emerging technology.