Microfluidic endothelium-on-a-chip development, from in vivo to in vitro experimental models.

In the last years, animal testing in medical research has been a controversial topic because of various reasons, such as ethical considerations and species differences. Therefore, more attention has been given to develop new technologies that can replace animal experiments and create in vitro models. Organ-on-a-chip (OOC) technology is a new and advanced technology based on microfluidic devices that can mimic the structure and function of entire organs and tissues as in vitro models. OOC models are miniature tissues and organs that assign characteristics for three-dimensional (3D) cell culture representation that resemble the original organs, together with their specific microenvironment microfluidic systems and specific biophysical processes, in order to mimic the normal physiological conditions and functionalities of the organs. Existing OOC models, such as liver, pancreas, heart, skin, brain, kidney, vessels, have been developed and designed for a specific function study. This review focuses on the main knowledge concerning OOC research and especially vascular endothelium-on-a-chip (EOC) model, developed in order to offer specific tools for studying vascular functions in physiological and pathological conditions. The field of OOC devices is still at the beginning, but in the future, this technology may have important roles in developing novel therapeutic approaches, offering new therapeutic molecules and providing the first step towards personalized medicine.

Introduction

Organ-on-a-chip (OOC) technology is a new developed technology designed to mimic the organs’ structure and create conditions for functional connections between cells by signaling molecules pathways. Therefore, OOC models assign features for three-dimensional (3D) representation of cellular complexity of the organs together with their natural microenvironments’ microfluidic systems and specific physical conditions, in order to recapitulate the as close as possible natural biology of the structures [1]. This in vitro advanced technology is able to replace animal experiments that are very important for preclinical screening of new therapeutic molecules, but which have various issues, such as time-consuming process, species differences and ethical considerations. Because cell culture systems are not exactly reproducing the entire organ function and they are also not in functional interaction with other types of cells, a system that permit multiple cell culture in a specific microenvironment and that are able to communicate through signaling molecules developed as a new advanced technology for experimental studies. This technology is able to offer new therapeutic molecules more rapid and that are highly reproducible in clinical studies. By cultivating patient’s cells-on-a-chip platform is also providing the first step to a personalized medicine [2,3]. Furthermore, the OOC systems are further developed for creating a multiorgan platform able to mimic the interactions between multiple organs [4]. Once created, multiorgan-on-a-chip (MOC) models can represent an ideal pharmacokinetic and pharmacodynamic tool for assessing the efficacy of pharmaceutical compounds on multiple organs [5]. A large number of MOC models were proposed but, because of the dynamic nature of human environment, the integration process of many organs into a chip remains a significant challenge in the future.

Despite the advantages of OOC compared with two-dimensional (2D) cultures and animal models, the major challenge to assign various OOC models consists in maintaining the cells viability as long as possible, for an adequate support of pathophysiological mechanism studies, but also to assess morphological and biochemical consequences associated with various disorders [6,7]. This advanced technology raised from the necessity to develop experimental models close related to natural conditions that can mimic multiple diseases and for a more rapid implementation of therapeutic strategies that can be transferred to clinical trials in a future personalized medicine. There are already serious concerns about the time to develop new therapeutic molecules. To meet these requests is necessary to create a model designed for a specific experiment that is testing a precise function of the organ by biomarkers. The parameters that serve as biomarkers have to have a precise assessing tool and have to offer the possibility to implementation in clinical studies. OOC already developed models correspond to this goal leading to a new technology in medical research field represented by laboratory-on-a-chip [8]. OOC models also permit the cell’s “cross-talking” by signalizing molecules, such as cytokines or chemokines, synthesized by various types of cells after they activation. The activation process can consist in triggering molecules for oxidative stress, inflammation, or apoptotic process. Existing OOC models, such as liver, pancreas, heart, skin, brain, kidney, are designed for a specific function study. After a better understanding of their functions, these models can serve to test various drugs in order to improve their delirious function resulting from a specific disorder [9]. Moreover, there is a growing interest in MOC models for drug testing due to the opportunity to study the systemic effects of new therapeutic molecules [10]. In this way, OOC can offer a viable solution that can contribute to a more realistic experimental model for a patient-specific disease and for a proper treatment. In this order, the cell culture on the chip can be represented by patient’s cells sample that are set on a chip in specific-organ microenvironment and the model can be used to test the efficiency of various drugs for each patient.

Endothelium-on-a-chip (EOC) as a main model for specific functions study

One of the most important OOC is represented by vessels-on-a-chip. In order to create a model that can reproduce the specific functions of endothelial structures, efforts were made to mimic the structure fabrication related to specific functions. Even there is yet not possible to have all the functions and implied structures on the same chip, several characteristics on endothelial cells in accord with their functions were successful used for research in order to replace animals in biomedical experiments [11,12]. Vessel’s wall permeability changes can constitute one of the most important components of pathophysiological mechanisms involved in diseases from all pathology fields. On the other hand, endothelial cells constitute the most important structure involved in atherosclerosis, onset of thrombotic process or in cancer metastasis. Structure changes of vessels wall contribute essentially to platelet activation and adhesion to the vessel wall glycoproteins, and further will contribute to the coagulation cascade initiation due to specific surface receptors exposure. These phenomena occur consequently to exposure of extracellular membrane proteins, such as collagen and laminin, and release of von Willebrand factor (vWF) [13,14]. The entire thrombotic process is also requiring platelet participation by expression of their adhesion to extracellular membrane proteins via glycoprotein Ib (GPIb) and glycoprotein VI (GPVI) [15]. Platelet activation and release of adenosine diphosphate (ADP) and thromboxane A2 (TxA2) together with platelet aggregation via glycoprotein IIb/glycoprotein IIIa (GPIIb/GPIIIa) homodimers are the ultimate processes designated to the first step of hemostasis [16]. Platelet activation is a consequence of endothelial structural changes [16].

Creating a model for vessels-on-a-chip is frequently addressed only to endothelium cells that can maintain their viability for days, in order to study the effects of endothelial damage on platelet activation. They can also serve to study the permeabilization for various molecules that are released in circulation and can further influence the specific tissues’ structures or functions. The main functional systems that were considered to be important for endothelium involvement in specific disorders, in order to be reproduced on a chip, are presented in Table 1 [17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35].

Table 1

Pathological processes associated with endothelial structural damage and dysfunction

Function	Associated structures
Endothelial structure as a barrier	▪ Vasodilatation molecules: NO [17] is synthesized continuously in endothelial cells from L-arginine by the constitutive eNOS enzyme [18]; in physiological quantities, exerts a protective role by maintaining vasomotricity vessels equilibrium; synthesized in excess, it is a component of nitro-oxidative stress with harmful effects [19]. ▪ Soluble antiplatelet molecules (aggregation and adhesion inhibitors). ▪ PGI2 [20]; shear stress upregulates COX-2 activity and increase the production of PGI2; PECAM-1 regulates vascular inflammatory responses [21]. ▪ Anti-adhesive molecules in vessel wall structure – glycocalyx [17]; glycocalyx destruction is mediated by ROS, MMPs and other enzymes, such as heparanase and sialidases; glycocalyx degradation has as a result increasing of vessel wall permeability, leukocyte–endothelium interactions, thrombosis onset and vascular inflammation [22]. ▪ Anticoagulant molecules – thrombomodulin (anti-coagulation effect by binding to thrombin and inactivating factors Va and VIIIa by increasing the generation of activated protein C) [23]. ▪ Endothelial EP4 receptor (for PGE2) on endothelial cells has an antithrombotic role [24].
Increased shear stress	▪ Increased shear stress induces the nuclear accumulation of class I, class II HDACs and DNA methyltransferases in endothelial cells, promoting oxidative stress and inflammation with endothelial cells dysfunction as a consequence [25]. ▪ Activation of endothelial cells is associated with vWF factor releasing [26]; platelet-derived factor V released from activated platelets is an important contributor to arterial thrombosis, together with increased cGMP levels and overexpression of P-selectin on the platelet surface [27].
Platelet activation	▪ Endothelial cells lesion is leading to collagen exposure; in damaged vessels, vWF factor binds to subendothelial collagen fibers and to platelet GPIb–IX–V receptor complex, thus mediating platelets adhesion to the injury site [28]. ▪ Upon adhesion, the primary platelet activators (collagen, thrombin, ADP, TxA2 and epinephrine) will induce multiple platelet responses (platelet morphology changes from discoid shape to other forms, activation of αIIbβ3 integrin promoting platelet aggregation and synthesis of TxA2 [29]; in the following phase, an irreversible platelet activation will result by: secretion of platelet dense and alpha-granules, exposure of membrane PS on the platelet surface activating the coagulation cascade, as well as formation of dense fibrin networks and platelet-derived microparticles [29]. ▪ Activated platelet release vWf and factor V [30].
Oxidative stress and inflammation	▪ ROS molecules induce hyperadhesive vWF assembly [31]. ▪ Vascular injury induces recruitment of pro-inflammatory cytokines (TNFα), which increases expression of surface adhesion molecules (e.g., ICAM-1, VCAM-1) on endothelial cells and release of vWF [32].
Tumor metastasis, angiogenesis	▪ The metastatic process is initiated by destruction of extracellular matrix due to MMPs, released by tumor cells. These proteolytic enzymes are responsible for degradation of basement membrane components (collagen IV, laminin) [33,34]. ▪ Tumor angiogenesis is promoted by PDGF and VEGF [35].

ADP: Adenosine diphosphate; cGMP: Cyclic guanosine monophosphate; COX-2: Cyclooxygenase-2; DNA: Deoxyribonucleic acid; eNOS: Endothelial nitric oxide synthase; GPIb–IX–IV: Glycoprotein Ib–IX–V; HDACs: Histone deacetylases; ICAM-1: Intercellular adhesion molecule-1; MMPs: Matrix metalloproteinases; NO: Nitric oxide; PDGF: Platelet-derived growth factor; PECAM-1: Platelet endothelial cell adhesion molecule-1; PGI2: Prostaglandin I2; PS: Phosphatidylserine; ROS: Reactive oxygen species; TNFα: Tumor necrosis factor alpha; TxA2: Thromboxane A2; VCAM-1: Vascular cell adhesion molecule-1; VEGF: Vascular endothelial growth factor; vWF: von Willebrand factor.

Microfluidic vessels models

In order to provide tools for the study of vascular phenomena, various in vitro microfluidic vessels models have been developed. They are complex in architecture and physiology, designed to offer solutions for artificial organ vascularization. The fabrication methods had to consider mimicking not only the histology of the vessels, but also to recreate an environment close to physiological conditions, corresponding to the organs or tissues where they were designated. Other models are designed to study the mechanisms associated with pathological conditions in order to understand the vessels contribution in onset or progression of various diseases. Using the advances in microfluidics, biomaterials and stem cells technology, the perfusable microvessels models can offer platforms that can imitate the tumor microenvironment, occluded microvessels or perfused capillary networks [36,37]. These microfluidic platforms can constitute an important advanced step in experimental studies, targeting not only the understanding of the vessel’s physiology and their contribution in various disorders, but also implementation of new therapeutic strategies. Due to a very complex architecture of human vessels, that have a particular structure and environments in each organ or tissue, there is not yet possible to reproduce the exact vessels’ models. Still, according with specific goals, different models are designed to study the effect of shear stress on vessels’ wall [38], the effects of exposure of the subendothelial structures to bloodstream or to shear stress with platelet activation and adhesion (thrombosis onset) [39], as well as the transendothelial transport phenomena [40].

One of the most used models for transendothelial transport mechanisms assessment consists in analysis of a fluorescent tracer that can across the tight junctions of endothelial cells [41]. The above properties of the vessels have a common important function that refers to the role of the barrier exerted by vessels wall. Thus, barrier integrity constitutes one of the most important features to study, in almost each microfluidic vessel model. Targeting this goal, efforts were made in order to develop instruments that are able to quantify the permeability of vessels wall. The existing technique offers the possibility for quantitative assessment of the permeability of vessels wall by measuring transepithelial/transendothelial electrical resistance (TEER), as well as paracellular diffusion of radiolabeled permeability markers (e.g., Mannitol) [40]. For qualitative assessment of membrane structure, different methods can be used: freeze-fracture electron microscopy of transmembrane fibrils or immunohistochemical analysis of tight junction-related protein expression [40]. As for the study of transendothelial transport mechanisms, the permeability of vessels’ wall can be study by measuring the ability of a well-known size molecule that serve as a fluorescent tracer. By assessment the tracer passage in surrounding medium, the leakage can be estimated and is expressed in terms of permeability (P) (cm/s) [42]. Typical molecules used in these models are fluorescent-labeled Dextran or fluorescent small molecule dye that can further be identified into the interest region, and florescence intensity measured according to the time interval [43]. There are conditions where the permeability offers valuable information about the intensity of the pathological process, e.g., an inflammation after an acute toxic aggression on endothelium, where the permeability is expected to be increased compared with normal endothelium function. The increased permeability can be transitive or permanent, and this characteristic can be also studied by assessment of vascular permeability in different time intervals. The same phenomenon is expected to be present when vessels’ architecture is anarchic, e.g., in novel vessels formation due to ischemic condition or due to neoplastic processes. The leakage can express the intensity of ischemia [vascular endothelial growth factor (VEGF) synthesis is related to ischemia magnitude] or may depend on tumor characteristics [36]. One of the most important transport functions studied on microvessel models is the penetrability of endothelium barrier by drugs or by drug delivery systems, such as nanocarriers [44]. Drugs’ penetrability through microvessels on in vitro models are most studied mechanisms for tumors treatments or for cancer metastasis prevention [44,45].

According to the literature data, the main studies focusing on structural and functional properties of vessels and their modeling in order to study various phenomena are presented in Table 2 [36, 41, 42, 46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67].

Table 2

Vessels-on-a-chip models: structural and functional properties

Endothelial structures models	After the ECs are seeded into the channels, they are incubated for a period of time (between 1–3 days) in order to get confluences. The formation of the monolayer of ECs with confluences can be faster if the microenvironment of the cells is provided with fibronectin, prior the introduction of ECs. The ECs monolayers are growing in a cylindrical channel within collagen gels and their diameters can increase to 75–150 μm after maturation. In this model, the density of the cells is approximately 107 cells/mL and offers a five days viable endothelial barrier [42]. Adherence between ECs can be tested by identification of specific junction protein, such is VE–cadherin. There are specific models that reproduce human ECs barrier that can support repeated measurements for permeability coefficient and where focal gaps can be detected [41]. Most used human ECs lines are HUVECs and HDMECs [36].
Vascular flow, pressure, and stability	Mimicking the vascular shear stress represents one of the greatest challenges for scientists, due to the complex system of forces that can be a modeling factor for endothelial structures. Unidirectional shear stress with a constant pressure produces endothelial lesion that are reproducing the real human conditions. Improving the stability and longevity of the ECs has to be based on pulsatile flow conditions. Shear stress and transmural pressure are the dominant mechanical signals that impose the rhythm of cells proliferation [46]. Pulsatile flow creates conditions for reorganization of the cytoskeleton and junctions modeling [41]. The intensity of pulsatile flow can be adapted also to various pathologies in order to mimic various disorders, such as atherosclerosis or stenosis conditions. Vascular cell adhesion molecules geometries correlate with shear stress on the vessels wall [47]. The variation of the pressure can be obtained by computer-controlled valve system that can modulate the recirculating fluid system physical properties [41]. More complex approaching is needed to create endothelial models that can recapitulate the human BBB. Using a commercially available microfluidic platform, Brown et al. developed a human microfluidic BBB model, using co-culture of primary human astrocytes and hCMEC/D3, resembling the human brain capillaries [48]. The model allows to study the junction between ECs under a physiological shear stress, acting for more than 24 hours, by assessing the phenotypical tight junction biomarkers [48].
Barrier function	The ECs barrier function is tested by permeability coefficient that represents the quantity of molecules with a certain size that are allowed to pass through endothelial junctions [36]. The barrier function is tested with fluorescent substances, such as Dextran. In vivo vessels permeability coefficients have normal range between 10−8 to 10−6 cm/s for 70 kDa Dextran [36]. The permeability of endothelial barrier is expected to be normal in physiological conditions and increased in pathological conditions (inflammation, apoptosis or necrosis), that lead to an activated endothelium with increased cells metabolic activity [49]. There are also significant differences between the permeability of the vessels with normal architecture and new vessels with anarchic architecture where the increased leakage is one of the most important features. The neovessels anarchic architecture in tumors contributes to the tumor cells spreading and metastasis [50]. Tumor’s vessels have an irregular diameter, abnormal branching, without typical characteristics of arterioles, capillaries or venules, being unable to be classified. Their basal membrane is incomplete and has abnormal pericytes coat resulting large-caliber with thin walls [51]. The expression of various receptors, integrins or growth factors is also abnormal [52]. Cationic liposomes are able to avid binding to pathological endothelium (associated with inflammatory process or tumors). This characteristic facilitates therapies with liposomes as carriers for various drugs [53]. Building OOC BBB is useful for quantification of drug targeting brain tissue study and is more complex due to its particular structure. The tight junction between ECs and the presence of polarized efflux pumps on luminal surface has to be modeled [54]. OOC BBB needs to be integrated in a functional network, known as neurovascular unit and contains neurons, pericytes, astrocytes, microglial cells, and ECM [55]. The BBB permeability is increasing in all the pathological situations of the central nervous system associated with cerebral edema. Disruption of BBB is associated with various neurological diseases, such as epilepsy, central nervous system tumors, ischemic stroke, multiple sclerosis or Alzheimer’s disease [55].
Cell motility and morphological parameters assessment	The morphology of the ECs is an important parameter to evaluate cell viability. In normal conditions, ECs are elongated, with the long diameter in the direction of shear stress. Any transition from the normal morphology to a pathological shape can indicate a dysfunction of the ECs. This can be from a mild dysfunction, such as the transition from cobblestone shape to a spindle-like morphology under shear stress, to severe changes that could indicate apoptosis or necrosis [56,57]. ECs in brain vessels resist to shear stress and do not elongate under these mechanical forces. This is an important characteristic that was demonstrated on BBB-on-a-chip model, in a specific microfluidic environment. This may indicate that ECs in BBB are programmed to resist to elongation under the shear stress due to their special and unique phenotype [38]. Tight junctions are forming in static conditions that will decrease the motility of the cells that prevent up regulation of proteins synthesis. Increasing the structural protein synthesis rate is associated with cell’s structural and functional changes [58]. If the shear stress conditions are increased due to hypertension or other systemic vascular diseases, will occur barrier function impairment, which could be reversed by therapeutic interventions [59]. The effect of share stress on dhBMECs was evaluated using live cells imaging, observing the cell morphology, cells speed, proliferation and apoptosis rate. Confluent monolayer is maintained in microfluidic conditions, on OOC device, and proteins and gene expression serve as biomarkers for various functions assessments (immunofluorescence method) [58]. Tight junctions can be evaluated by identification of several highly express adhesion proteins, such as VE-cadherin, ZO-1, occludin and claudin-5 [60]. Cells motility increase under the expression of mitogenic growth factors, such as FGF, EGF, VEGF, and can be considered as a parameter of cells activity. Cells speed can be evaluated by imaging of the cells and tracking the position of individual cell [38]. Moreover, ECs-on-a-chip model can be arranged as microfluidic network of channels containing proper constituents for red blood cells adding. In this way, a more complex platform results and this platform can also offer conditions for an in vitro model of in vivo circulation with communication between different cell types [61]. OOC models containing ECs that mimic vessels wall can be used as microfluidic platforms for studying the effect of drugs on morphological and structural changing in pathological conditions. There is also the possibility to integrate different drugs concentration on these platforms, in order to assess the dose-dependent effects. These microfluidics systems have all the advantages of lab-on-a-chip technology that allowed cell operations, such as seeding, culture, stimulation and staining into a chip [62]. In order to study liver drug toxicity and metabolism or the efficacy of anticancer molecules, microfluidics-based in vitro systems were realized [63,64,65]. The proliferation rate assessment can be used as a parameter to evaluate the efficiency of anticancer drugs [66].
Capillary perfusion	To study the function of capillary network, the capillary perfusion can be evaluated. Capillary perfusion can be defined as the ratio of fluorescence intensity of the substance in vascularized tissues compared with unvascularized tissues [36]. A variety of pathological conditions can be assessed by this method, such as increased capillary perfusion due to hyperemia associated with inflammation or decreased capillary perfusion in ischemic conditions [36]. The fluorescence method is using fluorescence-labeled Dextran as a molecule that has the property to remain in the vessel lumen, if there is no increased permeability, or to increase the fluorescence, if is accumulating in an ischemic capillary network [42].
ECs activation	Using live-cell microscopy, it is possible to analyze the cells by tracking their positions in a time-defined interval. Image analysis can provide also a quantitative assessment of mitosis rate, apoptosis process or motility of the cells [36].
Neoangiogenesis	Engineered tissue technique aims to create platforms that are able to mimic original structures in human body. Starting with ECs culture, microfluidic technology can create an environment-on-a-chip that can recapitulate the vessels structure. By integration in an ECM and by adding fibroblasts, this platform can create proper conditions for branching and anastomosis. Finally, a capillary network will result that can be designed to study the physiological and pathological conditions details. The angiogenesis resulting from this technique is a valuable tool for the study of various triggers, such as shear stress or cytokine influence [67].

BBB: Blood–brain barrier; dhBMECs: Derived human brain microvascular endothelial cells; ECM: Extracellular matrix; ECs: Endothelial cells; EGF: Epidermal growth factor; FGF: Fibroblast growth factor; hCMEC/D3: Human cerebral microvascular endothelial cell line; HDMECs: Human dermal microvascular endothelial cells; HUVECs: Human umbilical vein endothelial cells; OOC: Organ-on-a-chip; VE: Vascular endothelial; VEGF: Vascular endothelial growth factor; ZO-1: Zonula occludens-1.

Future perspectives of EOC models in pathology

Two of the most important pathological conditions, ischemia and uncontrolled cells proliferation during tumor formation, are the main targets for EOC platforms development. Therefore, these models are the most advanced models and the technology implemented is the most complex one.

The role of EOC models designed for tumor growth and metastasis studies

Tumor hallmarks include several characteristics represented mainly by continuous producing of growth molecules by tumor cells, evading molecules that function as suppressors, decreased apoptotic signals, facilitating replicative immortality, angiogenesis and enabling metastasis [68,69]. Tumors are also characterized by genetic instability and associated inflammation. They are also recruiting normal cells that contribute to the tumor microenvironment. Every type of tumor can recruit a specific repertoire of normal cell that will contribute to the specific metabolic feature of the tumor microenvironment [68]. To survive and growth, tumor provide neovessel formation. Neovessels serves for tumor nutrition but they are also one of the most important pathways for metastasis. Therefore, one of the main objectives in antitumor therapeutic strategies is to suppress neovessel proliferation and function. Metastatic tumor cells are spreading to the other organs or tissues through two steps: intravasation and extravasation [68]. The process of intravasation needs macrophages recruitment (tumor-associated macrophages that facilitate transendothelial migration of cancer cells) [70]. This process is very complex and incomplete understood, taking place at the interface between tumor and neovessels and having as participants numerous types of cells and signaling mechanisms. The goal of in vitro vascular models is to create a specific microenvironment that is able to mimic tumor environment and to offer a tool for studying details of metastatic process cascade [71,72]. Is already known that neovessels have an abnormal structure (less smooth muscle cells and pericytes), and consequently, an increased permeability, that facilitate tumor metastasis [73]. There is a linear relationship between the neovessels network density and the metastatic process, as well as between leakages size and tumor invasion [73].

Tumor metastasis process is also correlated with the microvessel diameter, beginning in microvessels larger than 30 μm [73]. The tumor cells can be added to an extracellular matrix (ECM) component in order to observe the interface between ECM and new formed tumor microvessels [71]. Detachments of tumor cells, their transendothelial migration and invasion process can be observed [74]. The spreading of the tumor into surrounding tissues involves few steps, represented by detachment of the tumor cells from primary tumor, degradation of ECM and migration process [75,76]. Tumor microenvironment plays a crucial role for metastasis rate [70]. Using real-time imaging was reported that vascular permeability is transient, depending on tumor microenvironment, macrophages cooperation, and signaling molecules synthesis, which finally will influence tumor cell dissemination [70]. The metastasis is also time depending on neovessels formation, angiogenesis being a fast process designed to tumor cells growth, proliferation and surviving [68]. The observation of interaction between endothelial cells and tumor cells is made by live cells microscopy fluorescent-labeled proteins being used to identify various pathological aspects, such as increased permeability or intensity of various signaling pathways [77]. OOC models comprise the structures required for a specific function. A special attention is paid to recreate the specific tumor microenvironment and to maintain a viable cells network, in order to offer a valuable tool for tumoral processes that controls metastasis cascade. Understanding the underlying mechanisms of tumor metastasis is an important step for therapeutic strategies applied to those patients who already present a metastatic tumor.

Ischemia and EOC models

Occluded microvessels models are important in vitro platform for vascular diseases, such as atherosclerosis, myocardial infarction or ischemic stroke. The complexity of such a model resides from the complexity of the targeting organs structures. Creating an entire OOC model could bring these platforms closer to the real pathogenetic processes associated with vascular diseases. Therefore, the efforts are made toward heart-on-a-chip or brain-on-a-chip models fabrication that can mimic the structure and biological process inside the organ [78,79]. The most difficult task is to create a proper microfluidic environment for a correct cell signaling pathways [80,81,82]. The endothelial cells play an important role in these models due to their crucial contribution to the barriers that compose the interface between the blood and a specific tissue. One of the most complex structures that assign the barrier function is the blood–brain barrier (BBB). The endothelial cells morphology, function, specific protein expressions and the microenvironment has to recapitulate the brain real physiological conditions of flow dynamics. After the building of the proper model, the microvessels can be obstructed manually or specific condition for thrombosis onset can be reproduced [55]. Perfusable tissue-engineered microvessel models can serve for studying the morphology and function of endothelial cells studying under the turbulent flow associated with atherosclerosis [47]. There is already possible to create in vivo models that are able to reproduce the specific 3D geometry of aneurysms, stenosis, bifurcations, in order to study the hemodynamic complex consequences of different pathological shear stress patterns [47]. It is believed that thrombotic mechanism onset is conditioned by hemodynamic particularities, endothelial cells morphology changes and platelet activation [83]. Mannino et al. created a vasculature model by seeding a monolayer of endothelial cells on a support that respect a circular distribution geometry, able to decrease the platelet aggregability; by introducing a disturbing flow factor, the platelet adhesivity and aggregability can be increased [47]. This model can serve as an in vitro support to study the processes associated with platelet activation and thrombosis onset. In vitro models can also create the possibility for studying endothelial injury under occluded and non-occluded microvessels [84]. Already existing models can also reproduce flow conditions in order to study platelets–endothelial cells interactions, that is a dynamic process yet incompletely understood [85,86]. The potential of microfluidic device technologies in this field bring hopes for a better understanding of pathophysiological mechanism and for a more efficient therapeutic strategies implementation.

Conclusions

The existing EOC models were developed in order to assign specific tools for vascular functions in physiological and pathological conditions, such as cells viability, transport properties, mitosis rate, proliferation, angiogenesis, apoptosis, permeability and metastasis. No models can perfect mimic the human conditions, but they offer the possibility to assess endothelial functions by replacing animal experiments with in vitro studies. The challenge is to fabricate EOC models as close as possible to in vivo conditions. EOC technology can be also used as a tool for therapeutic molecules efficiency researches. By incorporating the patient’s cell lines in these models of chip platform, there are already creating conditions for the first steps in a precise personalized medicine.

Conflict of interests

The authors declare that they have no conflict of interests.