Organ-on-a-chip devices are powerful modeling systems that allow researchers to recapitulate the in vivo structures of organs as well as the physiological conditions those tissues are subject to. These devices are useful tools in modeling not only the behavior of a healthy organ but also in modeling disease pathology or the effects of specific drugs. The incorporation of fluidic flow is of great significance in these devices due to the important roles of physiological fluid flows in vivo. Recent developments in the field have led to the production of vascularized organ-on-a-chip devices, which can more accurately reproduce the conditions observed in vivo by recapitulating the vasculature of the organ concerned. This review paper will provide a brief overview of the history of organ-on-a-chip devices, before discussing developments in the production of vascularized organs-on-chips, and the implications these developments hold for the future of the field.

Organ-on-a-chip devices were born out of a search for a more effective model to use in drug-testing. Two-dimensional culture of cells, traditionally used for drug screening, is useful but limited in scope. In the 2D culture, cells are adhered in a monolayer to a rigid surface and grown in static conditions.1,2 Although cells grown in this manner can provide us with valuable information regarding their behavior and characteristics, researchers have recognized that static cell culture is limited in its ability to accurately reflect the processes that occur in vivo. While cheaper and easier to maintain, cells grown in the adherent monolayer culture cannot provide adequate information about the complex cellular activity that occurs under physiological conditions.3,4 Cells grown under static conditions are not exposed to the mechanical stresses that cells in the body, subject to the effects of fluid flow, are shaped by.5–7 Nor does static culture, generally composed of only one type of cell, accurately capture the diverse chemical and biochemical interactions that happen between cells in heterogeneous, three-dimensional tissue.8–11 As a result, cells grown under two-dimensional, static conditions differ from cells in vivo not only in their morphology but also in their activity.4 While this review will not delve into the complexities of the impact the growth environment has on the cell, there is a wealth of review articles and research studies that have detailed the observable differences between cells grown in the 2D vs 3D culture.12–16 The evidence suggests that 2D cell culture cannot be considered a fully accurate model of what occurs in the human body.

Three-dimensional cell culture was, therefore, conceived as a remedy for these shortcomings. Many articles detail the progression of the 3D cell culture from its beginnings in gel-based culture to the present.9,14,17 Today, the term 3D cell culture encompasses diverse paradigms, including scaffold models, organoids, and organs-on-chips,17,18 the latter of which are the focus of this paper. Organ-on-a-chip devices have been built upon the concept of microfluidic chips, which contain micro-channels through which fluids can flow.19 In recent years, advances in bioprinting and biomaterials have led to their incorporation in microfluidic devices, allowing for the creation of more complex systems. Organs-on-chips are devices that support the growth of cells or tissues to model human organs. Their initial development and progress have been covered extensively in other papers.20,21 Organ-on-a-chip devices allow for the establishment of controlled conditions, which can mimic the environment that cells experience in vivo.22 These devices are capable of supporting multiple cell types—which allows for recapitulation of cell–cell interactions that occur under physiological conditions—as well as modeling some of the biomechanical forces that influence cell differentiation and behavior, such as shear stress.23 Recent advances in technology have encouraged researchers to take on the challenge of creating vascularized organs-on-chips.

Vasculature is, of course, an important component of the organ structure in vivo. It supplies various organs with blood and thereby, ensures that they receive oxygen and the various nutrients available in circulation. Different organs can vary vastly in their vasculature. For example, the kidney is a highly vascularized organ designed to optimize its ability to filter blood. The kidney, along with the structure of the nephrons, is depicted in Fig. 1.

FIG. 1.

Illustration of the human kidney with vasculature, along with the structure of the nephron. Created in Biorender.

FIG. 1.

Illustration of the human kidney with vasculature, along with the structure of the nephron. Created in Biorender.

Close modal

The blood enters the kidney via the renal artery, which progressively branches, creating smaller arteries and eventually arterioles. These arterioles branch further into capillaries, which lead into glomeruli, which are networks of nerve endings, and capillaries, wherein waste is filtered from the blood. The human kidney processes 20%–25% of the cardiac output,24 receiving more blood than nearly any other organ. To support this high-volume input, the kidney has about 1 200 000 nephrons, each fed by the arterioles and supported by the capillaries. The millions of vessels and filtration centers allowing high volumes of blood to be quickly processed are essential to kidney function.25 A model that neglects to reflect this complexity would be incomplete. Vascularized organ-on-a-chip devices are, therefore, a further step toward achieving a model that articulates the structural complexity of physiological organs as accurately as possible.

Vascularized organs-on-chips are generally two-compartment devices, with one compartment reserved for the growth of endothelial cells, which will mature into vascular networks, and another compartment for the growth of the cells that make up the parts of the organ, which are distinct from its vasculature. Nevertheless, alternative methods of device organization are present, which we shall examine in further detail as we delve deeper into specific examples. In this review, we shall explore recent successes in the creation of vascularized organ-on-a-chip models, as well as the scientific and technological advancements that have made these successes possible. We shall make specific mention of those papers that have been most recently published or that describe a particular accomplishment. It should be noted that, for the purpose of this review, the term vascularization encompasses many distinct approaches, including not only the integration of microvasculature into a microfluidic system but also the application of interstitial flow or the co-culture of stromal cells with endothelial cells. We recognize that these methods have different merits, and we shall attempt to elucidate the differences between various approaches in the course of this review. Table I provides a brief overview of the different levels of vascularization encountered throughout the papers and the merits of each.

TABLE I.

Types of vascularization methods.

Vascularization methodBenefitsLimitations
Fluid flow through constructed, endothelial cell-lined channel that connects to the culture chamber • Relatively simple and easy to implement
• Allows for modeling of flow effects
• Allows for greater control of perfusion medium 
• Does not recapitulate complex, branched geometries present in vasculature in vivo 
Self-assembled microvasculature formed by endothelial cells under static conditions • Self-assembly and de novo vessel formation recreates processes that occur in vivo
• Greater likelihood of faithfully modeling in vivo vasculature 
• Lacks effects of fluid flow that influence endothelial cell orientation and morphology in vivo
• May be difficult to control flow in certain chip formats 
Artificial method to create a biomimetic network • Allows greater degree of researcher control over the vascularized network
• More complex geometries can be recapitulated
• Greater resemblance to in vivo vasculature than flow through microfluidic channels 
• Vessel formation is not de novo and may not possess complete fidelity to in vivo vasculature
• If synthetic material used is not sacrificial, its interactions with cells may influence results 
Self-assembled microvasculature formed by endothelial cells under flow conditions • Closest to recapitulating in vivo vasculature
• Models flow effects and self-assembly ensures recapitulation of in vivo vessel formation processes 
• May be difficult to successfully implement
• In multi-organ chips, it may be difficult to control perfusion of multiple media types 
Vascularization methodBenefitsLimitations
Fluid flow through constructed, endothelial cell-lined channel that connects to the culture chamber • Relatively simple and easy to implement
• Allows for modeling of flow effects
• Allows for greater control of perfusion medium 
• Does not recapitulate complex, branched geometries present in vasculature in vivo 
Self-assembled microvasculature formed by endothelial cells under static conditions • Self-assembly and de novo vessel formation recreates processes that occur in vivo
• Greater likelihood of faithfully modeling in vivo vasculature 
• Lacks effects of fluid flow that influence endothelial cell orientation and morphology in vivo
• May be difficult to control flow in certain chip formats 
Artificial method to create a biomimetic network • Allows greater degree of researcher control over the vascularized network
• More complex geometries can be recapitulated
• Greater resemblance to in vivo vasculature than flow through microfluidic channels 
• Vessel formation is not de novo and may not possess complete fidelity to in vivo vasculature
• If synthetic material used is not sacrificial, its interactions with cells may influence results 
Self-assembled microvasculature formed by endothelial cells under flow conditions • Closest to recapitulating in vivo vasculature
• Models flow effects and self-assembly ensures recapitulation of in vivo vessel formation processes 
• May be difficult to successfully implement
• In multi-organ chips, it may be difficult to control perfusion of multiple media types 

3D cell culture models of the heart allow researchers to explore not only its function but also the effects of various drugs. Previous review papers have extensively covered the evolution of heart-on-chip models.26–29 Organ-on-a-chip models of the heart so far have each tended to focus upon a specific feature. The myocardium, or muscular layer of the heart, is of special interest. Ellis et al. produced an organ-on-a-chip model of the human myocardium, which for the first time incorporated both human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) and induced pluripotent stem cell (iPSC)-derived endothelial cells, recapitulating not only the muscle architecture but the microvasculature [Fig. 2(a)].30 The authors used human induced pluripotent stem cells (iPSCs) derived from the human skin. The hiPSC-CMs were placed in the middle channel of the device and encapsulated using a photo-cross-linkable hydrogel. One day later, fibronectin was added to the side channels for 1 h, and iPSC-derived endothelial cells were added. They were kept under static conditions for the first 24 h to facilitate attachment, after which the side channels were perfused to exert shear stress on the cells. All cells used were obtained from the same source, and this model could thus be considered a proof-of-concept that could be applied to a personalized medicine approach. Another version of a heart-on-chip device was developed by King et al. wherein the chip supported co-culture of iPSC-CMs, human left ventricular fibroblasts, and human cardiac microvascular endothelial cells, cultured under 0.5 μl/min flow conditions, and supplemented over time with VEGF and Ang-1 [Fig. 2(b)]. This chip, too, was spontaneously vascularized and recapitulated a beating myocardium in addition to recreating the physiological environment through diverse cell culture.31 

FIG. 2.

Heart on-a-chip devices. (a) Schematic of on-chip model of myocardium using iPSCs. [Reproduced with permission from Ellis et al., Biomicrofluidics 11(2), 024105 (2017). Copyright 2017 AIP Publishing LLC.] (b) Composition of cell-laden hydrogel and schematic of heart-on-chip device with loaded cell-laden hydrogel. [Reproduced with permission from King et al., Cell. Rep. Methods 2(9), 100280 (2022). Copyright 2022 Imperial College London.]

FIG. 2.

Heart on-a-chip devices. (a) Schematic of on-chip model of myocardium using iPSCs. [Reproduced with permission from Ellis et al., Biomicrofluidics 11(2), 024105 (2017). Copyright 2017 AIP Publishing LLC.] (b) Composition of cell-laden hydrogel and schematic of heart-on-chip device with loaded cell-laden hydrogel. [Reproduced with permission from King et al., Cell. Rep. Methods 2(9), 100280 (2022). Copyright 2022 Imperial College London.]

Close modal

Meanwhile, Tang et al. created a device with a “sandwich” structure, which had layers with microchannels on the top and bottom and a third layer in the middle intended for the integrated culture. The device could be opened for easy administration of drugs in a drug screening application. hiPSC-CMs and primary human umbilical vein endothelial cells (HUVECs) were cultured in the device. The hiPSC-CMs were cultured in the bottom chamber of the device in order to obtain a model of the myocardium. HUVECs were cultured on a culture patch to then create the vascular endothelium, which was subsequently introduced into the middle layer. The results confirmed vascular architecture was more similar to that found in vivo under dynamic flow conditions and also confirmed the importance of the endothelial layer's presence in producing accurate results in drug-testing. By applying two different cardiac drugs with known effects to the device, the authors were able to study the difference in response elicited when administering the drug directly to the hiPSC layer vs when administering the drug through the vascular layer, demonstrating that the vascular layer served to significantly weaken the drugs' effects on the cardiomyocytes. Furthermore, comparison of the drugs' effects under flow conditions vs static conditions demonstrated that the presence of flow could increase the drugs' apparent effects. As such, the chip confirms the importance of accurate models which reflect vasculature and flow conditions in vivo when studying the safety, pharmacodynamics, and pharmacokinetics of drugs.32 

In general, the examples cited here created vascularized chips by culturing endothelial cells and cardiomyocytes under flow such that the microvasculature self-assembles. Nevertheless, there are certain distinctions which must be noted. Co-culture of cardiomyocytes and endothelial cells, as seen in Ellis et al., may produce differences in behavior and cell–cell interaction, as opposed to culture of different cell types in separate channels or on separate layers. Furthermore, administration of supplemental factors may also produce different results in the vascularization of the chip. These are factors that may be explored in greater detail in further studies. These three studies present three distinct, successful approaches to creating a vascularized heart-on-chip and demonstrate that the field has not resolved itself upon a singular approach to the development of an on-chip model.

Given the complexity of the liver, designing an organ-on-a-chip that can recapitulate the organ structure in full is generally recognized as an ambition too lofty to be realistic given the current state of tissue engineering. Instead, researchers must, for the time being, restrict themselves to reproducing a particular facet of liver structure and/or function. Figure 3(a) depicts major structures incorporated into liver-on-a-chip devices. Ya et al. chose to focus on recapitulating the liver lobules. Liver lobules are structures that compose the two lobes of the liver. Roughly hexagonal in shape, the lobules are made up of hepatocytes arranged concentrically. The lobules are supplied with blood by the sinusoids, very small blood vessels, which deliver the necessary nutrients and oxygen to the hepatocytes.33,34 In order to model these structures, Ya et al. created a six-layer chip in the shape of a hexagonal prism, with channels intended to model the portal and central veins. A schematic of this device can be seen in Fig. 3(b). The chip was loaded with primary cells isolated from mouse livers and mixed with collagen. Culture media entered the chip through the portal vein and hepatic artery inlets on the first layer to travel through subsequent layers until it reached the coculture zone. Different flow rates were tested, and a rate of 50 μl/min was found to encourage the formation of in vivo-like sinusoids. Oxygen levels in the chip were regulated in the device by a regulating chip, ensuring that dissolved oxygen levels in the media were consistent with in vivo levels. Comparison to 2D and 3D static cultures found that the chip performed better in recapitulating physiological levels of substances such as albumin, urea, and bile acid. The chip was also demonstrated to be useful in drug screening applications. The self-assembled nature of the structures in the chip device constitute an improvement upon previous liver lobule models; however, the authors note the limitations of their chip, which include the fact that the cell culture cannot be maintained long-term due the high level of shear stress to which the cells are subjected.35 

FIG. 3.

Liver-on-a-chip devices. (a) Illustration of liver lobule and liver spheroid (created in Biorender). (b) Schematic of liver-on-chip device recapitulating liver lobules. [Reproduced with permission from Ya et al., ACS Appl. Mater. Interfaces 13(28), 32640–32652 (2021). Copyright 2021 American Chemical Society.] (c) Schematic of liver-on-chip device with bilayer microspheres. [Reproduced with permission from Liu et al., Front. Oncol. 12, 959299 (2022). Copyright 2022 Author(s), licensed under a Creative Commons Attribution License.

FIG. 3.

Liver-on-a-chip devices. (a) Illustration of liver lobule and liver spheroid (created in Biorender). (b) Schematic of liver-on-chip device recapitulating liver lobules. [Reproduced with permission from Ya et al., ACS Appl. Mater. Interfaces 13(28), 32640–32652 (2021). Copyright 2021 American Chemical Society.] (c) Schematic of liver-on-chip device with bilayer microspheres. [Reproduced with permission from Liu et al., Front. Oncol. 12, 959299 (2022). Copyright 2022 Author(s), licensed under a Creative Commons Attribution License.

Close modal

Another on-chip liver model employed a hexagonal shape similar to the model previously described, with supply channels being used to mimic the flow of blood from the hepatic artery and portal veins. However, this approach featured the use of bilayer microspheres created using hepatic stellate cells and hepatocytes, in order to recapitulate the tissue–tissue interfaces present in the liver. Liver sinusoidal endothelial cells were also incorporated in the chip co-culture, and self-assembled to create a vascular network, resulting in a liver microtissue that was fully vascularized [Fig. 3(c)]. The authors were able to confirm that the hepatocytes remained functional in the culture.36 

Meanwhile, Lee et al. developed a scalable, perfusable on-chip liver model intended for use in screening drugs or modeling disease. A cotton candy machine was used in order to obtain poly(N-isopropylacrylamide) fibers, which were mixed with gelatin hydrogels with cross-linking capability as well as with hepatic spheroids, or liver buds, before insertion into the chip. The fibers were intended to mimic a microvascular network and dissolved upon perfusion of the device with media, thus creating many small channels. While modeling late-stage fibrosis, the researchers compared both co-culture of hepatocytes and endothelial cells in the device to the effects of culturing hepatocytes alone. They found that signs of normal liver function, such as albumin secretion and expression of certain genes, were more pronounced in the presence of endothelial cells. The device was able to successfully model non-alcoholic steatohepatitis, as well as the inflammatory and liver toxicity responses to a hypocholesterolemia drug. The ability of the model to reproduce the results seen previously in animal models suggests that the model constitutes an acceptable alternative to animal testing.37 

Ya et al. were able to demonstrate de novo formation of blood vessels upon the application of a flow parameter in their chip. However, maintaining cell culture long-term in the chip is difficult. Liu et al. expanded upon what was accomplished in Ya et al., establishing a tissue–tissue interface as well as self-assembled vasculature. Meanwhile, Lee et al. have presented a novel method of creating a vascular network using sacrificial hydrogels. While de novo formation of vasculature has potential to accurately recapitulate in vivo characteristics of vasculature, the approach presented by Lee et al. nonetheless provides a degree of control over the vasculature, which may be preferable in certain applications.38 

The aforementioned models are particularly salient examples of the advances made in creating a vascularized liver-on-chip. However, this is not an exhaustive recounting of the strides made in the liver-on-a-chip research. Other review papers detail in greater depth the various advancements made with liver-on-a-chip models.39–41 

As mentioned in the Introduction of this review, the kidney is the organ responsible for filtration of the blood and is thus endowed with a highly complex vasculature, which processes high volumes of blood in a relatively short amount of time. The struggle to recapitulate this complexity in vitro is one of the major challenges confronting researchers in the field. The suggestion that a model with vascularized flow might yield better results than a static model was put forth in a study by Homan et al., which confirmed that kidney organoids cultured under interstitial flow conditions formed tubular compartments and demonstrated better gene expression, as well as contained more mature podocytes, when compared to organoids grown in static culture.42 This conclusion about the importance of flow was expanded upon in studies which established on-chip models of the kidney with applied flow. An example of a vascularized kidney on-chip model was developed by Lee et al. They created a polydimethylsiloxane (PDMS)-based device using a 3D printed mold. 300–500 μm kidney organoids were created from human pluripotent stem cell-derived kidney cells and were cultured in microwells located on the chip [Fig. 4(a)]. The substrate was coated with 1.5% Matrigel and 1.5% Matrigel also contained 100 ng/mL VEGF. The presence of VEGF was found to improve PECAM1 expression, suggesting it is important for the differentiation of the pluripotent stem cells into endothelial cells. Fluidic flow of 10 μl/min with shear stress of 1.31 × 10−5 dyne/cm3 was applied. PECAM1 expression increased under flow conditions, and ImageJ analysis of confocal images of the organoids showed significantly increased incidence of vascular cells in organoids cultured under flow conditions. The applicability of the device for drug-testing was tested by administrating tacrolimus, a drug used in preventing transplant rejection in kidney transplant patients. The study results showed that the organoid on the chip demonstrated lower viability than organoids in static culture, suggesting that static culture models underestimate the nephrotoxicity of such drugs.43 

FIG. 4.

Kidney-on-a-chip devices. (a) Kidney organoid-on-a-chip incorporating organoids derived from human pluripotent stem cells. [Reproduced with permission from Lee et al., Nano Converg. 8(1), 35 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 International License.] (b) Schematic of kidney glomerulus on-a-chip device comprised of a urinary channel and capillary channel. [Reproduced with permission from Roye et al., Micromachines 12(8), 967 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 International License.]

FIG. 4.

Kidney-on-a-chip devices. (a) Kidney organoid-on-a-chip incorporating organoids derived from human pluripotent stem cells. [Reproduced with permission from Lee et al., Nano Converg. 8(1), 35 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 International License.] (b) Schematic of kidney glomerulus on-a-chip device comprised of a urinary channel and capillary channel. [Reproduced with permission from Roye et al., Micromachines 12(8), 967 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 International License.]

Close modal

Another kidney on chip model chose to focus on recapitulating the kidney glomerulus—a key target in judging not only kidney function but also drug effects, due to the high volume of blood it filters. The glomerulus is composed of capillaries, which are lined by a fenestrated endothelium. The capillaries are surrounded by podocytes. Roye et al. for the first time demonstrated that induced pluripotent stem cell-derived vascular endothelial cells (vECs) could be used to recapitulate the kidney structure and function. Furthermore, their model was the first to obtain both epithelial and endothelial cells from the same source. Emulate pods were used to construct their device. The pods had two fluidic channels, which ran parallel to each other and lay one above the other in the device. The device used vascular endothelial cells (vECs) and intermediate mesoderm cells [Fig. 4(b)]. The device was primed using podocyte induction media and vascular endothelial media, with one media type used in each of the two channels. The cells were introduced into fluidic channels separated from each other by a porous membrane. The chip was placed in an automated vacuum regulator that was responsible for maintaining fluid flow and for controlling stretch-relaxation cycles. The chip was perfused at a rate of 60 μl/min, with shear stress of 0.0007 dyne/cm2 in the urinary channel and 0.017 dyne/cm2 in the microvascular channel. 10% cyclic strain at 4 Hz was also applied through the two channels. The mesoderm cells differentiated into podocytes while in the device over 5 days, before the induction media was replaced with CultureBoost-R. Culture was maintained for an additional week. The podocytes formed a barrier with a branching network, and layering of epithelium, basal membrane, and endothelium was observed in the capillary channel. By observing the filtration of substances such as insulin and albumin, functionality of the kidney cells was confirmed. The device was then perfused with a chemotherapeutic drug, Adriamycin, through the capillary channel. The endothelium began to detach under the influence of the drug, and damage to the podocytes was noted, thus confirming the applicability of the device in drug screening contexts.44 

While Homan et al. laid the foundation that flow is an important factor in vascularization of 3D cell culture models of the kidney, Lee et al. and Roye et al. expanded this concept with their models. Lee et al.'s model demonstrates not only the influence of flow, but the influence of growth factors important to vascularization on the differentiation of stem cells into kidney cells vs endothelial cells that make up the kidney vasculature. Roye et al.'s model, meanwhile, takes a step further in attempting to recapitulate a particular aspect of kidney architecture, expanding on that which is accomplished by Homan et al. and Lee et al. by incorporating co-culture and applying mechanical forces such as strain in addition to fluid flow. The result provides a model with advanced recapitulation of complex vasculature.

These are but a few examples of kidney-on-chip devices, which capture the directions in which vascularized kidney-on-chip research is proceeding. Others have reviewed in greater depth the developments occurring in the kidney-on-a-chip research.45–47 

The skin constitutes the largest organ in the human body. It consists of three main layers: the dermis, epidermis, and hypodermis, overlain by a layer of dead cells, which acts as a barrier.48  Figure 5(a) shows a labeled diagram of the skin and its layers.

FIG. 5.

Skin on-a-chip devices. (a) Diagram of normal human skin (created in Biorender). (b) Schematic of skin equivalent made by Mori et al., which Salameh et al. based their device on. [Reproduced with permission from Mori et al., Biomaterials 116, 48–56 (2017). Copyright 2017 Elsevier Ltd.] (c) Diagram of wounded skin with inflammatory response depicted (created in Biorender).

FIG. 5.

Skin on-a-chip devices. (a) Diagram of normal human skin (created in Biorender). (b) Schematic of skin equivalent made by Mori et al., which Salameh et al. based their device on. [Reproduced with permission from Mori et al., Biomaterials 116, 48–56 (2017). Copyright 2017 Elsevier Ltd.] (c) Diagram of wounded skin with inflammatory response depicted (created in Biorender).

Close modal

Models of the skin can provide valuable insight into processes such as wound healing as well as the ability of the skin barrier to act as a defense against chemical or pathogenic exposure. Furthermore, human skin models are useful in cosmetics and topical drug-testing applications, providing a higher level of accuracy than may be found in static, 2D cell culture, or animal models.49 Strides have been made toward creating vascularized organ-on-a-chip models of the skin that can recapitulate the various layers of the skin as well as the complexity of the microvasculature that supplies these layers with blood and oxygen and plays a role in skin permeability—an important factor in the ability of topical drugs, as well as pollutants and pathogens, to enter into the system through the skin. Mori et al. presented a way to create just such a vascularized device. They describe combining a “skin equivalent,” or skin model with an epidermal and dermal layer and vasculature consisting of vascular channels lined with endothelial cells, perfused using peristaltic pumps [Fig. 5(b)].49 Salameh et al. built upon this approach in order to create an improved skin equivalent, which was constituted of a differentiated epidermal layer along with a complex vasculature that not only encompassed the three perfusable channels, lined with endothelial cells which were induced to form angiogenic sprouts, but also a layer of HUVECs between the two layers of fibroblasts suspended in the collagen matrix, which formed a microvascular network that branched out and connected with the sprouts emanating from the perfusable channels. The model used normal human dermal fibroblasts and normal human epidermal keratinocytes in addition to HUVECs. Nylon wires were strung across the connectors and the devices were plasma etched after sterilization. Bovine type I collagen and normal human dermal fibroblasts were added to the devices and incubated at 37 °C until a layer was formed that resembled the dermis. Normal human dermal fibroblast 3D medium was used for the first five days, before being replaced with ECGM-2. The nylon wires were then removed, and the channels were filled with ECGM-2 medium containing HUVECs using a syringe pump, exerting a flow rate of 6.5 ml/h. The device was inverted for a half hour to coat the upper parts with HUVECs, before returning to the original position for incubation over 24 h to allow for adhesion of the cells. The dermal layer was then seeded with keratinocytes and immersed in DMEM/F12 immersion medium. Perfusion at 2–3 ml/h was applied. After 3 days, stratification was induced by exposure to the air–liquid interface. Vasculogenesis was established by adding neutralized collagen/fibroblasts onto the device until the nylon wires were reached, then incubating until the collagen gelled before adding HUVECs. More collagen/fibroblast mixture was poured over the HUVECs. The dermal layer was incubated for 5 days in a fibroblast/endothelial cell medium mix. The procedure was then repeated to seed the hollow channels. The applicability of the model for studying skin permeability to the chemicals caffeine and minoxidil was studied. While the constructed model was unable to obtain as low a skin permeation coefficient as observed in porcine skin, a commonly used reference model in experimental testing of chemical permeation, it was, nevertheless, able to obtain lower permeation coefficients than static models lacking perfusion and/or vascularization, demonstrating the importance of a vascularized and perfusable model for accuracy in drug and cosmetics testing.50 

Other researchers have approached the fabrication of skin-on-chip models differently. For example, Sun et al. created a model for studying herpes simplex viral infection. Microfluidic channels were obtained by injection molding and soft lithography, after which primary human dermal endothelial cells were added. Fibroblast-containing collagen was added to the bottom well of the device, and the fibroblast medium was perfused through the inlet. The top well of the device was seeded with primary human dermal keratinocytes, while the growth medium was added to the inlet, and the device was constantly perfused with the endothelial cell medium. At 100% confluency of the keratinocytes, the keratinocyte differentiation medium was added to the top well and inlet. After overnight culture, the keratinocytes were exposed to air for differentiation over a period of 7–12 days. Interactions between the fibroblasts and endothelial cells helped to reshape the collagen matrix, and a chip that recapitulated skin vasculature as well as the epidermis was obtained. The study then went on to examine the effects of the application of HSV infectious particles on the surface of the epidermal layer. When the layer was unbroken, the development of a layer of dead skin cells was observed and keratinocytes began to demonstrate upward growth. Meanwhile, applying the same particles to a disrupted epidermis resulted in infection of the keratinocytes, and perfusion with neutrophils found that the microvasculature in the chip supports an inflammatory immune response. Perfusion of an anti-HSV drug showed the presence of a dose-dependent response and viral suppression. The chip device therefore highlights the importance of an unbroken skin barrier in defending against pathogens, as well as demonstrates its utility as a tool for studying pathogenesis and drug efficacy.51 Finally, Biglari et al. developed an organ-on-a-chip device for the purpose of examining wound-healing processes of the skin. They created a chip with two lateral channels, which supported the culture of dermal fibroblasts and M1 and M2 macrophages, as well as an inner channel which supported 3D co-culture of HUVECs mixed with Matrigel, allowing for the creation of a microvascular network. The chip was fabricated from PDMS using soft lithography to create micropatterning. Channels were all treated with collagen and fibronectin to enhance cell attachment. A diagram of a healing wound and inflammatory response can be seen in Fig. 5(c). In the absence of any immune cells, the authors were able to mimic an inflammatory response through the addition of Tumor Necrosis Factor-alpha (TNF-alpha) and macrophages. Their data suggests that their chip presents a valid representation of the wound-healing process. It was found that pro-inflammatory cytokine and chemokine levels increased in the presence of TNF-alpha and the macrophages, and that the reduction of this increase—such as by application of the drug Dexamethasone—was conducive to better vascularization in the HUVEC cells.52 

Both Salameh et al. and Sun et al. used a device construction approach that allowed them to layer different cell types in a way which mimics the layering of cells in the skin. Their approaches allowed them to obtain a device that models the skin as a barrier, able to demonstrate permeability and the effects of disruption. Meanwhile, Biglari et al. used a somewhat different approach, using lateral channels to model inflammation. All three approaches were able to successfully accomplish vascularization. The use of micropatterning by Sun et al. may provide a greater degree of control over the arrangement of cells in the device. Likewise, the co-culture in the Biglari et al. device may provide more opportunity for cellular cross-talk. Still, each approach achieves a unique aim and demonstrates a valid model of the skin.

Overall, researchers have made great progress in capturing the complexity of the structure of skin, its vasculature, and various factors that influence pathogenesis, permeation, and inflammation in the limits of the organ-on-a-chip device.

The endometrium is the lining of the uterus. With an epithelial layer over a stromal layer, the endometrium undergoes complex cyclical changes as it is shed and regenerated during the menstrual cycle. Endometrial thickness changes through the course of the menstrual cycle. Figure 6(a) depicts endometrial thickness over the course of one cycle, and Fig. 6(b) shows the hormonal fluctuations that influence these changes.

FIG. 6.

Endometrium-on-a-chip devices. (a) Fluctuation in endometrial thickness throughout the menstrual cycle (created in Biorender). (b) Changes in estrogen and progesterone levels throughout the cycle. [Reproduced with permission from Ahn et al., Human Reprod. 36(10), 2720–2731 (2021). Copyright 2021 Oxford University Press.] (c) Schematic diagram of endometrium-on-a-chip. [Reproduced with permission from Gnecco et al., Human Reprod. 34(4), 702–714 (2019). Copyright 2019 Oxford University Press.] The top chamber contains uterine microvascular endothelial cells (UtMVECs), and the bottom chamber contains stromal fibroblasts. (d) Schematic diagram of endometrium-on-a-chip. [Reproduced with permission from Ahn et al., Human Reprod. 36(10), 2720–2731 (2021). Copyright 2021 Oxford University Press.]

FIG. 6.

Endometrium-on-a-chip devices. (a) Fluctuation in endometrial thickness throughout the menstrual cycle (created in Biorender). (b) Changes in estrogen and progesterone levels throughout the cycle. [Reproduced with permission from Ahn et al., Human Reprod. 36(10), 2720–2731 (2021). Copyright 2021 Oxford University Press.] (c) Schematic diagram of endometrium-on-a-chip. [Reproduced with permission from Gnecco et al., Human Reprod. 34(4), 702–714 (2019). Copyright 2019 Oxford University Press.] The top chamber contains uterine microvascular endothelial cells (UtMVECs), and the bottom chamber contains stromal fibroblasts. (d) Schematic diagram of endometrium-on-a-chip. [Reproduced with permission from Ahn et al., Human Reprod. 36(10), 2720–2731 (2021). Copyright 2021 Oxford University Press.]

Close modal

Even though the endometrium appears to remain a relatively under-explored organ in the realm of organ-on-a-chip research, two striking studies have, nevertheless, stood out. Gnecco et al. created a two-chambered on-chip device, which recapitulated the endometrial structure and the vascular endothelium in order to obtain a better understanding of the process of decidualization, in which stromal cells in the endometrium produce proteins that support gestation in preparation for pregnancy. A schematic of this device can be seen in Fig. 6(c). Primary stromal cells and uterine microvascular endothelial cells (UtMVECs) were both used in the coculture of this device. The endothelial cells were kept under continuous perfusion using a syringe pump at 1 μl/min in order to subject them to the effects of shear stress, which not only had effects on the cell morphology but also on cell function. Over the course of two weeks, the devices were treated with estradiol and progestin methoxyprogesterone acetate (MPA) together to trigger decidualization, and vascularized models were compared with static models as well as models consisting solely of stromal cells. It was found that prolactin expression increased as a result of treatment with estradiol and MPA simultaneously (as opposed with estradiol alone), as well as in perfused co-culture models as opposed to static culture of stromal cells without an endothelium. Additionally, decidualization was more robust in co-cultures with continuous perfusion of the endothelial cell culture. Furthermore, it was found that shear stress on the endothelial cells induces the production of prostaglandin, which has further downstream interactions that induce the production of cyclic adenosine monophosphate (cAMP) from the stromal cells, which in turn is one of the factors necessary for decidualization. As such, the device created in this study allowed researchers to gain important insight into the critical role played by the vascular endothelium of the endometrium in the preparatory stages of pregnancy.53 

Ahn et al. were able to create a device that is able to capture the cyclical processes involving hormones and vasculature that occur in the endometrium in relation to the menstrual cycle. The device they created had three layers of tissue: the epithelial layer, the stromal layer, and the microvascular layer. The device had five parallel microchannels separated by micropillars. There were two central channels, one of which was loaded with HUVECs, endothelial stromal fibroblasts embedded in fibrin, and gel solution composed of 2.5 mg/ml fibrinogen and 0.15 U/ml aprotinin. The second channel was loaded with endothelial cells embedded in fibrin. The outermost channel of the device was used to culture additional endothelial stromal fibroblasts to produce biochemical cues that encourage angiogenesis [Fig. 6(d)]. Media was added to the upper reservoirs and aspirated from the lower reservoirs, before filling media channels 1 and 2 with media. Cell culture then proceeded under static conditions; vascular networks and lumens began to form after 3–4 days. After day 8, Ishikawa cells were introduced. Ishikawa cells are a line of human endometrial adrenocarcinoma cells which were derived from a 39-year-old woman. These cells have been shown to induce tumor formation in mice, and they possess estrogen and progesterone receptors.54 Once these cells were introduced, the device was tilted at a right angle and incubated for half an hour to allow for adhesion, before incubation for 4–6 days. They found that the vascularization of endothelial cells in the device was dependent on the presence of endometrial stromal fibroblasts, without which networks were unable to form. Administration of estradiol and progesterone in doses which mimic in vivo hormone levels as the menstrual cycle progresses showed that the epithelial layer changed in thickness in response, which is consistent with the process observed in vivo. Furthermore, administering an emergency contraceptive drug, levonorgestrel, to the endometrium-on-chip resulted in an increase in dead cells in the epithelial and stromal layers, as well as a decreased rate of blood vessel regression. It was also observed that the endometrium became unstable as a result of the drug's effects, with the stomal layer becoming leaky. This is consistent with the effects of the drug in vivo. The utility of the device for studying embryonic implantation, a complex process which requires further study, was demonstrated by coating microbeads with specific growth factors associated with implantation and adding them into the chip, where they showed specific binding to the endometrial epithelial layers, avoiding binding to other structures in the device.55 

Gnecco et al. and Ahn et al. demonstrate two distinct approaches to creating an endometrium-on-a-chip. The device created in Gnecco et al. demonstrates the effects of fluidic flow and resulting shear stress on endometrial processes such as decidualization, as well as expression of prolactin and prostaglandin. Meanwhile, Ahn et al. demonstrates the importance of cross-talk, despite lacking continuous perfusion in their model, showing the vascularization can be induced without flow. While Ahn et al. establishes the importance of cell–cell communication in recapitulating the processes and architecture present in vivo, Gnecco et al. go a step further by demonstrating the impact mechanical forces also have on these aspects of the organ model.

The blood–brain barrier is a network of vasculature that functions to control the movement of nutrients and other molecules into and out of the brain. The neurovascular unit is the network of blood vessels and smooth muscles responsible for controlling the barrier. The blood–brain barrier is of significance in drug modeling as it is necessary to determine whether a given drug will cross the barrier and enter the brain. On-chip models of the neurovascular unit attempt to provide insight into the function and structure of the barrier and may also be applied towards drug modeling purposes. There are many review papers, which cover the progress in the field in greater depth.56,57 Herein, we shall highlight a few examples of recent studies.

One example of a neurovascular unit on a chip is described by Wei et al., who developed a blood brain–barrier model using human neural stem cells derived from human embryos. The microfluidic device included a PDMS layer above a polycarbonate (PC) layer, followed by a lower PDMS layer and finally a PDMS-coated glass slide. The PDMS layers had channels for perfusion and chambers for cell culture, with the upper layer used for culture of endothelial cells and the lower layer used for culture of the neural cells. Human brain microvascular endothelial cells were seeded onto the PC membrane prior to chip assembly. The lower PDMS layer and the glass slide were treated with plasma to bond them together. The glass slide was sealed with the PC membrane with endothelial cells and the upper PDMS layer. Following assembly, continuous perfusion was begun using culture medium. Endothelial culture medium was perfused through the upper chamber. Neural stem cell neurospheres were then loaded into the lower chamber, where they adhered overnight. The chip was then perfused with differentiation medium for one week. Analysis confirmed the presence of astrocytes, neurons, and oligodendrocytes following the differentiation protocol. The brain microvascular cells had formed a continuous endothelium with characteristic morphology. Viability of the cells was confirmed with live-dead assays. Limited permeability of the barrier with FITC-dextran confirmed the integrity of the endothelium. The chip was used to model ischemic encephalopathy. The authors concluded that the chip demonstrated biocompatibility and recapitulated the blood brain barrier. However, they acknowledged that their model does not involve pericytes, and the ratios of the different types of neural cells did not reflect the ratio in vivo.58 

Another study examined the effects of micropatterning in a neurovascular unit on a chip. Self-assembled monolayers of octadecanethiol were transferred onto gold-sputtered nanocoating in transwells and were used to immobilize anchoring units with similar purpose to collagen and fibronectin. PDMS stamps were molded with line and ring geometries used to replicate symmetry in blood vessels and were used in the transfer of the octadecanethiol into the transwells. Patterned surfaces were sterilized with ethanol, then were incubated with 25 μg/mL fibronectin. Flow with a rate of 15.5 ml/min was applied, exerting a 637 dynes/cm2 shear stress, using single-pump perfusion, with each transwell having its own inlet and outlet. While these are not physiological conditions—indeed, much lower shear stresses have been used in other models59—they allow for modeling the effects of mechanical forces on the cells. Cultures were then seeded into the transwells, with astrocytes added to the basolateral side and the endothelial cells added to the apical side. Co-culture under perfusion continued for 72 h prior to characterization. Symmetry of blood vessels was confirmed by cross sections and staining of actin and nuclei. Monoculture of endothelial cells was compared to the co-culture, and it was noted that ZO-1 expression was more prominent. Furthermore, transendothelial electrical resistance was higher and permeability more limited for models grown on micropatterning as opposed to endothelial monoculture on a non-patterned surface. Traction forces were elevated in the co-culture model, suggesting that astrocyte signaling plays an important role in control of the blood–brain barrier circulation. The comparison of the micropatterned model to culture on a non-patterned surface showed that mechanical stress is an important influence on the microenvironment and on the endothelial barrier properties, as well as on the morphology of endothelial cells. Furthermore, comparison of co-culture to endothelial cell mono-culture demonstrated the influence of astrocytes on the barrier, as well as on calcium signaling.60 

These studies build upon earlier studies such as those conducted by Campisi et al., in which a device with tri-culture of astrocytes, pericytes, and iPSC-derived endothelial cells was created. Soft lithography was used in the manufacturing of the PDMS chip, which had a single layer micro-channel and two fluid channels. To seed the cells, the cells were first resuspended in media containing thrombin, then mixed with fibrinogen before being added into the device's gel ports. Polymerization was carried out for 15 min at room temperature before the addition of media. The media was supplemented with VEGF for the first 4 days of culture. In the study, mono-culture of iPSC-ECs, co-culture of iPSC-ECs and pericytes, and tri-culture of pericytes, astrocytes, and iPSC-ECs were all tested and compared. The blood–brain barrier self-assembled through a process similar to vasculogenesis, with the presence of pericytes and astrocytes resulting in branched architecture with greater resemblance to that found in the in vivo blood–brain barrier. The cells in tri-culture also demonstrated the importance of cell–cell interactions. For example, paracrine and juxtacrine signaling was observed between the iPSC endothelial cells and the pericytes in the formation of the microvasculature. RT-PCR results also suggested that pericytes induced iPSC endothelial cells to mature into brain endothelial cells. Meanwhile, the presence of astrocytes led to greater expression of blood–brain barrier transporters and tight-junction proteins. The barrier formed from mixed culture was found to be more robust and difficult to penetrate than many in vitro models and bore resemblance to the permeability of the blood–brain barrier in rats. The interactions between the various cell types play an important role in recapitulating structure and function. Nevertheless, it must be noted that the model lacks neurons and microglia and may, therefore, not be a complete model of physiological processes.61 

One last study of note explores the results of using hypoxic conditions in order to enhance the differentiation of iPSCs into brain microvascular endothelial cells. Given that the blood–brain barrier develops in an oxygen-deprived environment in vivo, it was hypothesized that hypoxic induction conditions would lead to differentiation of iPSCs into endothelial cells with traits specific to brain microvascular endothelial cells. The chip in this study was also generated using soft lithography, creating a two-channel device made from PDMS. A CNS channel and a vascular microchannel were separated from each other by a PET membrane coated on either side with fibronectin and collagen IV. Astrocytes and pericytes were seeded into the CNS channel, and the iPSC-derived endothelial cells were seeded in the vascular channel. 5% oxygen was applied as a hypoxic condition, and 20% oxygen was used as a control; however, the authors note that small molecules can achieve a hypoxic effect without the need for special culture chambers. It was found that hypoxic conditions led to significantly greater mRNA expression of endothelial cell–cell adhesion molecules, barrier transporters, and factors and molecules like VEGF and PECAM-1. Additionally, Wnt7a mRNA expression increased over 25-fold, implicating upregulation of Wnta/beta-catenin signaling, which is important to the development of microvessels in the brain. Furthermore, the model produced a functional, low-permeability barrier that can maintain in vivo-like functionality for two weeks, which had not been accomplished in earlier models. Clearly, recreating not only cell–cell interactions but also in vivo developmental conditions can aid in achieving a more accurate model.62 

The aforementioned studies all present distinct approaches to creating a neurovascular unit on a chip. Many of these papers demonstrate the importance of co-culture in allowing for cellular cross-talk as well as recreating physiological effects such as mechanical stress. Wei et al., for example, demonstrates a more diverse co-culture but is still limited by its lack of pericytes and by the established cell ratio. Meanwhile, in Singh et al., the co-culture involves only astrocytes and endothelial cells, but it provides insight into the biomechanical effects of the presence of astrocytes specifically; similarly, while Campisi et al.'s chip also neglects to include neurons and microglia, it nevertheless provides a compelling basis for the importance of co-culture in recreating the cell–cell interactions, which occur in vivo. Finally, Park et al.'s hypoxia-enhanced chip provides insight into the impact the recreation of in vivo developmental and physiological conditions can have on the functionality of the chip. While perhaps imperfect in their recapitulation of the cellular diversity of the neurovascular unit, these studies provide valuable insight into different aspects of neural cell differentiation and cell–cell interaction. Future studies may continue to provide insight on the importance of specific cell types and will hopefully work towards a co-culture that more closely resembles in vivo tissue.

While the focus of this paper is on vascularized organ-on-a-chip devices designed primarily to model the function of healthy organs, it is impossible to overlook the preeminence of cancer models in the field. Several of the models described in previous sections have been found suitable for use in cancer studies. Indeed, vascularized on-chip models are an attractive paradigm with which to study the characteristics of tumors—including their formation, proliferation, and ability to coopt native vasculature—as well as the effectiveness and permeability of various anti-cancer drugs. Though it is possible to find models of tumors originating in nearly any organ one cares to name, we shall content ourselves here with the mention of a few notable examples.

A study in Small Structures describes the creation of a vascularized on-chip model of liver cancer. The chip was created to allow for further examination of the tumor microenvironment, in hopes of determining why cancer cells in hepatocellular carcinoma often survive after treatment. The device involved a heterogeneous culture, including hepatocytes, stellate cells, Kupffer-like macrophages, and endothelial cells. The device included a hollow chamber with microvasculature, along with a chamber for the extracellular matrix. Concentric layers of cells and collagen matrix were used to imitate liver sinusoids. Collagen type 1 was used to simulate the tumor microenvironment, at concentrations of 4 and 7 mg/ml to replicate liver stiffness for normal and cirrhotic conditions. The first concentric layer of the device was seeded with collagen I and HepG2 cells and HepG2 cells transfected with CYP3A4, a mutation of cytochrome P40 that affects chemotherapeutic outcomes, with high expression indicating a likelihood of poor outcomes. The collagen was polymerized around a needle, and a hollow lumen was created by removing the needle. The second layer was created in the same manner using stellate cells with collagen. Endothelial cells and TH1P macrophages were added to the lumen to create vasculature. The vasculature was preconditioned with flow. The model was able to effectively recapitulate the disease state, displaying inflammation and cirrhosis, along with increased vascular permeability.63 

Meanwhile, Park et al. developed a lung cancer model, which employed tumor spheroids composed of A549 cells (lung cancer cells), HUVECs, and human lung fibroblasts. The device used a hybrid hydrogel with a decellularized extracellular matrix base, which had pro-angiogenic and tissue-mimetic properties. Macro-channels, or large channels running through the hydrogel intended to mimic veins and arteries, were the sites of endothelial cell culture. The endothelial cells were encapsulated in the tissue-mimetic hydrogel and loaded onto the middle layer of the device, and after gelation a PDMS membrane was placed to prevent collapse of the hydrogel. The device was incubated at 37 °C overnight. The endothelial cells formed large vessels inside the channels, and angiogenic sprouting was observed, which led to vasculature forming to connect the various macro-channels. Additionally, new vasculature was observed growing from the macro-channels out towards the spheroids as the study went on. This allowed for a realistic example of the tumor microenvironment, as the device had recapitulated the formation of vascular networks, which support tumors in vivo. The model was additionally confirmed to be useful for drug-testing purposes.64 

A final intriguing example of an on-chip device from the tumor modeling sphere is a model of glioblastoma multiforme (GBM) developed by Silvani et al. The device was comprised of three compartments, of which one was devoted to tissue culture, and another was devoted to vascular flow. A biopsy punch was used to create the tissue compartment, and this compartment was loaded with a gel containing GBM cells. A bio-printed hydrogel was used in order to help create the concentric structures of the stromal cells observed in GBM. Functional tumor spheroids were grown using these bio-printed constructs. In order to model the blood brain barrier, the vascular channel of the device was coated with fibronectin and then seeded with HUVECs and human cerebral microvascular endothelial cells. Following continuous culture under perfusion with 9 dyne/cm2 shear stress for 72 h, the cells formed a monolayer lining the channel walls, and a perfusable lumen was formed within the channel. Immunofluorescence imaging confirmed co-localization of actin filaments and the presence of zonula occludens-1 protein, a tight-junction protein, at the border of the cells. By testing the permeability of the barrier, the authors were able to confirm that the barrier formed was more resistant to penetration than the purely HUVEC endothelia.65 The results suggest that a functional model of the blood brain barrier was obtained, a significant achievement given the role this barrier plays in blocking traffic of intravenously administered drugs to the brain, and the importance of its presence for any accurate model of brain cancers, and how they might respond to an administered therapeutic.

These three studies show different approaches to tumor modeling on a chip, each working with different types of tumor tissue. Many of these approaches used gel to recreate the tumor microenvironment. Co-culture of different cells is also used, allowing for cross-talk.

The aforementioned studies are merely a few examples of organ-on-a-chip devices developed for modeling cancers. Several review papers cover tumor-on-a-chip research in far more detail.66–68 The particular examples examined in this review are noted due to their focus on cancers in organs which are typically considered challenging to model, and their use of heterogeneous cell cultures to accurately recreate tumorous tissue as well as achieve vascularization. Indeed, on-chip models for tumors exist for a vast variety of cancers originating in many organs. However, on-chip devices in the tumor modeling realm are not confined to singular attempts at recapitulating pathogenesis in a specific organ; researchers have also endeavored to create devices suitable for high-throughput screening of cancer therapeutics. For example, Kim et al. created what they referred to as an “all-in-one microfluidic” device which could create a large number of vascularized tumor spheroids out of a number of different cell types.69 Additionally, multiple-organ-on-a-chip devices have been developed to study the phenomenon of cancer metastasis as well as to better understand the effects of anti-cancer drugs on different organs, given that many drugs may have side effects on organs unrelated to their target.70 

In short, the development of more accurate models of tumors is a subject of great interest among today's researchers and will no doubt continue to be a thriving source of progress in the world of organ-on-a-chip research. Nevertheless, we must acknowledge that the progress we see in cancer modeling, and in other, previously mentioned areas of organ-on-a-chip research, has been facilitated by many advances in technology. We shall discuss some of these innovations in greater detail in the next section.

Creating organotypic vasculature is critical to establishing models, which can recapitulate cellular events occurring in vivo.71 As noted in many studies, the endothelium differs from organ to organ and even displays heterogeneity within a single organ.72 The factors secreted by endothelial cells play a role in determining how the organ responds to stress, such as inflammation or fibrosis, and these factors also differ based on the type of organ-specific endothelial cell responsible for producing them.73,74 Furthermore, endothelial cells are important in maintaining the unique microenvironment of the organ.75 Abnormal behavior of endothelial cells may play a role in disease etiology.76–78 It is, therefore, essential that microfluidic models account for this diversity and strive to establish an endothelium, which can recreate this complexity. Vascularized organ-on-chip models attempt to accomplish this goal, and there have been many developments in the search for ways to create biomimetic vascular networks that can be integrated into a microfluidic system.

There are many factors affecting vascularization. For example, the presence of certain peptides or other chemicals may influence endothelial cell behavior.79,80 For example, RGD peptide, or arginine–glycyl–aspartic acid, is noted for promoting adhesion in endothelial cells when attached to various material surfaces.81 Similarly, the presence of certain growth factors, such as VEGF, may encourage migration and angiogenesis.82 The properties of the material onto which endothelial cells are seeded also has an impact.83 Finally, the cell source may impact the results as well. Endothelial cells derived from different vessels have different properties.84 Additionally, the use of primary cells vs iPSCs may also make a difference. For certain cell types, maturation of the cells from iPSCs may be challenging.85 Researchers have also sometimes found differences in primary cell and iPSC behavior,86 and iPSCs have demonstrated tumorigenicity in the long-term culture.87 iPSCs remain proliferative for longer than primary ECs, but it has not been confirmed whether their behavior reflects that of mature primary cells. Meanwhile, primary cells may be difficult to source, as they come from donors and may have a certain amount of variability from batch to batch.88,89 While this paper does not go into exhaustive detail regarding the various factors that may influence vascularization, there are many helpful articles that may be consulted for deeper understanding.90–97 

The growth in the field demonstrated by the articles reviewed in this paper thus far has been made possible by advances in biomaterials and techniques, specifically those pertaining to the creation of on-chip vasculature. We shall now examine these developments in greater detail.

Researchers have taken pains to identify the circumstances under which the formation and growth of biomimetic vascular networks might be first induced and then encouraged. For example, Mykuliak et al. attempted to determine whether adipose or bone-marrow derived mesenchymal stem cells would be better suited for vasculogenesis on a microfluidic chip device. The study found that bone-marrow derived stem cells had qualities that made them more attractive compared to the adipose-derived stem cells, such as higher pericytic character, as well as an increased ability to form basement membrane and produce mature networks.98 Another recent study examined angiogenesis on various biomaterials, both natural and synthetic. Using an angiogenesis on-a-chip device, the authors evaluated the relative merits of collagen, fibrin, and dextran vinyl sulfone hydrogels. To encourage cell adhesion and enable the incorporation of heparin-binding proteins collagen and fibrinogen, dextran vinyl sulfone hydrogels were functionalized with arginine–glycyl–aspartic acid (RGD) and heparin-binding peptide. The authors were able to confirm that increasing matrix density of collagen and fibrin hydrogels resulted in increased endothelial cell invasion and multicellular angiogenic sprouting, whereas the synthetic dextran vinyl sulfone hydrogel demonstrated less sprouting. Additionally, they were able to link the permeability of the matrix with the amount of endothelial cell invasion. Nevertheless, it must be noted that while the authors determined many mechanical properties of the various matrices, such as Young's modulus and the porosity, the study did not vary these properties independently in its examination of angiogenic sprouting, so it cannot be determined which of these properties was most impactful in the sprouting process.99 There have also been attempts to study vasculogenesis and vasculature itself through the use of on-chip devices and in vitro techniques meant to model vascular networks and their creation, rather than a specific organ.100 For example, Fathi et al. developed an on-chip model of a lymphatic vessel and demonstrated its potential to recapitulate both healthy and diseased pathology, as well as examined the effects of shear stress on the formation of the endothelial layers.101 Additionally, Yang et al. demonstrated that they were able to achieve shear stress and flow rates on par with physiological values in a pumpless device with unidirectional and bidirectional flow, in which they cultured HUVECs. The cells formed a monolayer as expected. Non-unidirectional flow induced inflammatory response in the endothelial cells, showing the device is capable of recapitulating healthy and diseased states of the endothelial barrier. It was suggested that such a device could be incorporated into more complex tissue-on-chip models in order to account for the effects of the endothelium.102 Although a full review of these efforts is out of the scope of this paper, these models have been extensively reviewed by others.103–105 However, such studies, which investigate both vascular structure and composition as well as the underlying chemical and biological processes involved in vasculogenesis, have helped provide useful insights that have aided not only in better understanding vasculature and how it works, but also in engineering vasculature in organ-on-a-chip devices.

In addition to studying vasculature itself, efforts have been made to discover new and creative ways to create vasculature composed of vessels similar in structure to what is observed in vivo: cylindrical, with hollow lumens. Previous review papers have discussed more conventional methods of engineering vasculature, such as lithographic techniques106 and scaffolds.103 However, many new techniques with potential have been put forward. Novel methods of creating biomimetic vasculature include such innovations as viscous fingering and the use of leaf venation as a blueprint. Viscous fingering is a lithographic technique that makes use of the phenomenon of spontaneously formed ink patterns, which had previously been dismissed as defective printing and which are produced when a liquid transfer that replaces a higher-viscosity liquid with a lower-viscosity one occurs. The resulting patterns, caused by instability in the air–liquid interface, demonstrate branching and have a “finger-like” appearance, which have the advantage of resembling the random arrangement of vasculature in vivo. Researchers have found a way to exploit this phenomenon by printing such viscous fingerprints onto biocompatible polymers, like gelatin.107 Konopka et al. employed this approach in creating an on-chip device modeling primary blood vessels. Using viscous fingering, they were able to create cylindrical lumens, which could be loaded with HUVECs. The device showed angiogenic activation in the endothelial cells.108 

Meanwhile, leaf venation as a reference point for the creation of biomimetic vasculature is the approach taken by Mao et al. in their study entitled “Human-On-Leaf Chip: A Biomimetic Vascular System Integrated with Chamber-Specific Organs.” Using a computer-aided design (CAD) drawing of the leaf vein patterns present in the species Osmanthus fragrans, the researchers were able to transfer the pattern in CAD onto a PDMS chip, which had multiple chambers for the creation of different organ niches. Endothelial cells were added to the vascular channels, while HUVEC cells were seeded onto the chambers, allowing for the creation of microvascular networks in each organ niche that connected with the channels. The channels were perfused, resulting in connection between the various chambers, and parenchymal cells were later added to the chambers. The device was able not only to demonstrate successful vascularization through the use of this novel technique but also was able to serve as a model for metastatic cancer of the pancreas, which had spread to the liver and the bone.109 

Another notable method of creating vascular channels is a subtractive manufacturing technique described by Rajashekar et al. wherein the properties of various materials were exploited to create channels within a hydrogel. In this technique, alginate was first patterned onto a polystyrene sheet, after which the fibers were dried and arranged in a bottomless 384-well plate. A hydrogel solution was then added in order to rehydrate the dried alginate, and the plate was then incubated. During incubation, the alginate network swelled and changed shape within the hydrogel, pulling away from the polystyrene. The hydrogel was cross-linked in order to hold the alginate in place. Calcium was then extracted from the alginate in order to dissolve it, creating a perfusable network where the fibers had been. Like viscous fingering and leaf venation, this method, too, allows for the formation of biomimetic networks, although it allows for perhaps a greater degree of control over the placement of the networks.110 

Aside from these advances in knowledge and techniques in regard to vascularization, much attention has also been given to the creation of high-throughput and scalable chip devices and manufacturing techniques, which can be used for drug screening and related purposes. A Lab on a Chip paper described a laser-cutting technique that allows for efficient patterning of hydrogel. The microfluidic device created with this technique was able to support the growth of vascularized spheroids.111 In a separate study, Phan et al. described the creation of a device, which repurposes a 96-well plate, with 12 microfluidic device units in between two PDMS layers, and three tissue chambers connected to microfluidic channels. CD31+ endothelial cells isolated from human cord blood were loaded into the tissue chambers, and vasculature was observed as having formed within seven days. The networks were perfusable, and the device was shown to be useful in anti-cancer drug screening. The device allows for the scalable production of vascularized micro-tissue and can be successfully used in drug-testing applications.112 A more recent paper confirmed the utility of this device, using it to culture colon cancer cells and study the associated vasculature. The authors of this paper also compared 2D monolayer culture to 3D culture, as well as non-vascularized to vascularized culture in the context of drug-testing, and found that vasculature allows for better and more accurate modeling.113 

Finally, multi-organ-on-a-chip devices, in which microfluidic models of various organs are fluidically linked into one system, are also an important advancement that cannot be ignored. Multiple organ-on-a-chips can be used to study the physiological interactions between organs, as well as model complex diseases that affect multiple organs and evaluate whole-body effects of drugs.114 There are many interesting examples of such devices. For example, Herland et al. developed a device with liver, gut, and kidney models linked by perfusable vascular channels lined with endothelial cells, along with an arteriovenous compartment used to recapitulate physiological flow. The endothelial–parenchymal interface was confirmed to be present—crucial due to its role in influencing pharmacokinetics and pharmacodynamics. Testing of nicotine on the device gave data that match known pharmacokinetics, although renal clearance was unlikely to be accurate due to the lack of glomeruli in the mode. Nevertheless, liver clearance observed in the model was found to be more accurate to human physiology than the results provided by animal models.115 An even more ambitious multi-organ chip is described by Ronaldson-Bouchard et al., who created a device with matured tissue niches of the human heart, liver, bone, and skin, all linked by vascular flow, and each cultured in a compartment with organ-specific media that is shielded from the general vascular flow by a selectively permeable endothelial barrier.116 Meanwhile, Novak et al. presented an intriguing model which fluidically links ten vascularized organs-on-chips together. The model was able to maintain cell culture for weeks and was successful in modeling small molecule transport between organs. The model was found to recapitulate known clinical profiles and pharmacokinetics for nicotine and cisplatin.117 As vasculature increases the accuracy of organ-on-a-chip models, multi-organ chip devices elevate the model's ability to reflect interactions and processes that occur in vivo. In such devices, vascularization becomes more important than ever, as the vascularized nature of the organ-on-a-chips linked within the system, as well as the vascular flow that connects them, must be biomimetic and as accurate as possible in order to yield a model, which is of use in a clinical setting.

The field of organ on chips is diverse and ever-evolving, with researchers continuing to develop and improve upon manufacturing techniques and device design in order to increase the utility of their models. This review has examined publications that concern a variety of techniques, devices, and organs, showing that there are many different ideas about how best to obtain vascularization and biomimetic tissue in microfluidic devices. It is probable that, as research continues and new information comes to light concerning the success of various techniques and the complex factors influencing vascularization, certain techniques will become favored over others. The process of manufacturing vascularized organ-on-a-chip devices may then become more standardized. Undoubtedly, this would lead to much interesting work, as scalable manufacturing becomes more common, and more complex systems become easier to model. Nevertheless, we must take into account key barriers, which must be addressed if the field is to make progress. One challenge may be maintaining long-term culture in an on-chip device due to the effects of shear stress on the cells as well as the increasing potential for unstable genotype as culture continues. Another challenge is determining a standard for the establishment of biomimetic vasculature. There are now many approaches to vascularization, but it is difficult to determine which method is most effective. The lack of standardization in the field makes comparison and evaluation of different devices difficult. Finally, as single organ models are not the most accurate models of in vivo function, and multi-organ chips become more common, researchers will have to grapple with how to supply multiple different vascularized organs with the medium suitable to the unique cultures. Despite these challenges, the current state of the field promises a fascinating future with the development of more complex and physiologically relevant models.

The authors thank Dr. Kaitlyn Sadtler. This research was supported by the Intramural Research Program of the NIH, National Institute of Biomedical Imaging and Bioengineering.

The contents of this publication are the sole responsibility of the authors and do not necessarily reflect the views, opinions, or policies of the NIH and the Department of Health and Human Services (HHS). Mention of trade names, commercial products, or organizations does not imply endorsement by the U.S. Government.

The authors have no conflicts to disclose.

Anagha Rama Varma: Conceptualization (equal); Investigation (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (equal). Parinaz Fathi: Conceptualization (equal); Funding acquisition (lead); Project administration (lead); Resources (lead); Supervision (lead); Visualization (supporting); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
S.
Breslin
and
L.
O’Driscoll
, “
Three-dimensional cell culture: The missing link in drug discovery
,”
Drug Discov. Today
18
(
5
),
240
249
(
2013
).
2.
R.
Edmondson
,
J. J.
Broglie
,
A. F.
Adcock
, and
L.
Yang
, “
Three-dimensional cell culture systems and their applications in drug discovery and cell-based biosensors
,”
ASSAY Drug Dev. Technol.
12
(
4
),
207
218
(
2014
).
3.
M.
Abbas
,
F.
Moradi
,
W.
Hu
,
K. L.
Regudo
,
M.
Osborne
,
J.
Pettipas
,
D. S.
Atallah
,
R.
Hachem
,
N.
Ott-Peron
, and
J. A.
Stuart
, “
Vertebrate cell culture as an experimental approach—Limitations and solutions
,”
Comp. Biochem. Physiol. Part B: Biochem. Mol. Biol.
254
,
110570
(
2021
).
4.
M.
Kapałczyńska
,
T.
Kolenda
,
W.
Przybyła
,
M.
Zajączkowska
,
A.
Teresiak
,
V.
Filas
,
M.
Ibbs
,
R.
Bliźniak
,
Ł.
Łuczewski
, and
K.
Lamperska
, “
2D and 3D cell cultures—A comparison of different types of cancer cell cultures
,”
Arch. Med. Sci. AMS
14
(
4
),
910
919
(
2018
).
5.
N.
Wang
, “
Review of cellular mechanotransduction
,”
J. Phys. D: Appl. Phys.
50
(
23
),
233002
(
2017
).
6.
B.
Li
,
F.
Li
,
K. M.
Puskar
, and
J. H.-C.
Wang
, “
Spatial patterning of cell proliferation and differentiation depends on mechanical stress magnitude
,”
J. Biomech.
42
(
11
),
1622
1627
(
2009
).
7.
G. H.
Altman
,
R. L.
Horan
,
I.
Martin
,
J.
Farhadi
,
P. R. H.
Stark
,
V.
Volloch
,
J. C.
Richmond
,
G.
Vunjak-Novakovic
, and
D. L.
Kaplan
, “
Cell differentiation by mechanical stress
,”
FASEB J.
16
(
2
),
1
13
(
2002
).
8.
M.
Mirbagheri
,
V.
Adibnia
,
B. R.
Hughes
,
S. D.
Waldman
,
X.
Banquy
, and
D.
Kun Hwang
, “
Advanced cell culture platforms: A growing quest for emulating natural tissues
,”
Mater. Horiz.
6
(
1
),
45
71
(
2019
).
9.
D.
Huh
,
G. A.
Hamilton
, and
D. E.
Ingber
, “
From three-dimensional cell culture to organs-on-chips
,”
Trends Cell Biol.
21
(
12
),
745
754
(
2011
).
10.
C. M.
Kofron
and
U.
Mende
, “
In vitro models of the cardiac microenvironment to study myocyte and non-myocyte crosstalk: Bioinspired approaches beyond the polystyrene dish
,”
J. Physiol.
595
(
12
),
3891
3905
(
2017
).
11.
R.
Liu
,
X.
Meng
,
X.
Yu
,
G.
Wang
,
Z.
Dong
,
Z.
Zhou
,
M.
Qi
,
X.
Yu
,
T.
Ji
, and
F.
Wang
, “
From 2D to 3D co-culture systems: A review of co-culture models to study the neural cells interaction
,”
Int. J. Mol. Sci.
23
(
21
),
13116
(
2022
).
12.
K.
Duval
,
H.
Grover
,
L.-H.
Han
,
Y.
Mou
,
A. F.
Pegoraro
,
J.
Fredberg
, and
Z.
Chen
, “
Modeling physiological events in 2D vs. 3D cell culture
,”
Physiology
32
(
4
),
266
277
(
2017
).
13.
J.
Caleb
and
T.
Yong
, “
Is it time to start transitioning from 2D to 3D cell culture?
,”
Front. Mol. Biosci.
7
, 33 (
2020
).
14.
Y.
Li
and
K. A.
Kilian
, “
Bridging the gap: From 2D cell culture to 3D microengineered extracellular matrices
,”
Adv. Healthc. Mater.
4
(
18
),
2780
2796
(
2015
).
15.
M.
Sun
,
A.
Liu
,
X.
Yang
,
J.
Gong
,
M.
Yu
,
X.
Yao
,
H.
Wang
, and
Y.
He
, “
3D cell culture—Can it be as popular as 2D cell culture?
,”
Adv. NanoBiomed. Res.
1
(
5
),
2000066
(
2021
).
16.
B. M.
Baker
and
C. S.
Chen
, “
Deconstructing the third dimension—How 3D culture microenvironments alter cellular cues
,”
J. Cell Sci.
125
(
13
),
3015
3024
(
2012
).
17.
D. V.
Krysko
,
R.
Demuynck
,
I.
Efimova
,
F.
Naessens
,
O.
Krysko
, and
E.
Catanzaro
, “
In vitro veritas: From 2D cultures to organ-on-a-chip models to study immunogenic cell death in the tumor microenvironment
,”
Cells
11
(
22
),
3705
(
2022
).
18.
J.
Hoarau-Véchot
,
A.
Rafii
,
C.
Touboul
, and
J.
Pasquier
, “
Halfway between 2D and animal models: Are 3D cultures the ideal tool to study cancer-microenvironment interactions?
,”
Int. J. Mol. Sci.
19
(
1
),
181
(
2018
).
19.
A.-G.
Niculescu
,
C.
Chircov
,
A. C.
Bîrcă
, and
A. M.
Grumezescu
, “
Fabrication and applications of microfluidic devices: A review
,”
Int. J. Mol. Sci.
22
(
4
),
2011
(
2021
).
20.
D.
Singh
,
A.
Mathur
,
S.
Arora
,
S.
Roy
, and
N.
Mahindroo
, “
Journey of organ on a chip technology and its role in future healthcare scenario
,”
Appl. Surf. Sci. Adv.
9
,
100246
(
2022
).
21.
C. M.
Leung
,
P.
de Haan
,
K.
Ronaldson-Bouchard
,
G.-A.
Kim
,
J.
Ko
,
H. S.
Rho
,
Z.
Chen
,
P.
Habibovic
,
N. L.
Jeon
,
S.
Takayama
,
M. L.
Shuler
,
G.
Vunjak-Novakovic
,
O.
Frey
,
E.
Verpoorte
, and
Y.-C.
Toh
, “
A guide to the organ-on-a-chip
,”
Nat. Rev. Methods Primer
2
(
1
),
1
29
(
2022
).
22.
A. E.
Danku
,
E.-H.
Dulf
,
C.
Braicu
,
A.
Jurj
, and
I.
Berindan-Neagoe
, “
Organ-on-a-chip: A survey of technical results and problems
,”
Front. Bioeng. Biotechnol.
10
,
840674
(
2022
).
23.
L. A.
Low
,
C.
Mummery
,
B. R.
Berridge
,
C. P.
Austin
, and
D. A.
Tagle
, “
Organs-on-chips: Into the next decade
,”
Nat. Rev. Drug Discov.
20
(
5
),
345
361
(
2021
).
24.
R. G.
Evans
,
D. W.
Smith
,
C.-J.
Lee
,
J. P.
Ngo
, and
B. S.
Gardiner
, “
What makes the kidney susceptible to hypoxia?
,”
Anat. Rec.
303
(
10
),
2544
2552
(
2020
).
25.
M. A.
Wallace
, “
Anatomy and physiology of the kidney
,”
AORN J.
68
(
5
),
799
820
(
1998
).
26.
Y. S.
Zhang
,
J.
Aleman
,
A.
Arneri
,
S.
Bersini
,
F.
Piraino
,
S. R.
Shin
,
M. R.
Dokmeci
, and
A.
Khademhosseini
, “
From cardiac tissue engineering to heart-on-a-chip: Beating challenges
,”
Biomed. Mater.
10
(
3
),
034006
(
2015
).
27.
Y.
Zhao
,
N.
Rafatian
,
E. Y.
Wang
,
Q.
Wu
,
B. F. L.
Lai
,
R. X.
Lu
,
H.
Savoji
, and
M.
Radisic
, “
Towards chamber specific heart-on-a-chip for drug testing applications
,”
Adv. Drug Deliv. Rev.
165–166
,
60
76
(
2020
).
28.
Q.
Yang
,
Z.
Xiao
,
X.
Lv
,
T.
Zhang
, and
H.
Liu
, “
Fabrication and biomedical applications of heart-on-a-chip
,”
Int. J. Bioprinting
7
(
3
),
370
(
2021
).
29.
V.
Paloschi
,
M.
Sabater-Lleal
,
H.
Middelkamp
,
A.
Vivas
,
S.
Johansson
,
A.
van der Meer
,
M.
Tenje
, and
L.
Maegdefessel
, “
Organ-on-a-chip technology: A novel approach to investigate cardiovascular diseases
,”
Cardiovasc. Res.
117
(
14
),
2742
2754
(
2021
).
30.
B. W.
Ellis
,
A.
Acun
,
U. I.
Can
, and
P.
Zorlutuna
, “
Human iPSC-derived myocardium-on-chip with capillary-like flow for personalized medicine
,”
Biomicrofluidics
11
(
2
),
024105
(
2017
).
31.
O.
King
,
D.
Cruz-Moreira
,
A.
Sayed
,
F.
Kermani
,
W.
Kit-Anan
,
I.
Sunyovszki
,
B. X.
Wang
,
B.
Downing
,
J.
Fourre
,
D.
Hachim
,
A. M.
Randi
,
M. M.
Stevens
,
M.
Rasponi
, and
C. M.
Terracciano
, “
Functional microvascularization of human myocardium in vitro
,”
Cell. Rep. Methods
2
(
9
),
100280
(
2022
).
32.
Y.
Tang
,
F.
Tian
,
X.
Miao
,
D.
Wu
,
Y.
Wang
,
H.
Wang
,
K.
You
,
Q.
Li
,
S.
Zhao
, and
W.
Wang
, “
Heart-on-a-chip using human iPSC-derived cardiomyocytes with an integrated vascular endothelial layer based on a culture patch as a potential platform for drug evaluation
,”
Biofabrication
15
(
1
), 015010 (
2022
).
33.
G.
Janani
,
S.
Priya
,
S.
Dey
, and
B. B.
Mandal
, “
Mimicking native liver lobule microarchitecture in vitro with parenchymal and non-parenchymal cells using 3D bioprinting for drug toxicity and drug screening applications
,”
ACS Appl. Mater. Interfaces
14
(
8
),
10167
10186
(
2022
).
34.
Z.
Kmieć
, “
Introduction—Morphology of the liver lobule
,” in
Advances in Anatomy Embryology and Cell Biology
, Cooperation of Liver Cells in Health and Disease, edited by
Z.
Kmieć
(
Springer
,
Berlin
,
2001
), pp.
1
6
.
35.
S.
Ya
,
W.
Ding
,
S.
Li
,
K.
Du
,
Y.
Zhang
,
C.
Li
,
J.
Liu
,
F.
Li
,
P.
Li
,
T.
Luo
,
L.
He
,
A.
Xu
,
D.
Gao
, and
B.
Qiu
, “
On-chip construction of liver lobules with self-assembled perfusable hepatic sinusoid networks
,”
ACS Appl. Mater. Interfaces
13
(
28
),
32640
32652
(
2021
).
36.
J.
Liu
,
C.
Feng
,
M.
Zhang
,
F.
Song
, and
H.
Liu
, “
Design and fabrication of a liver-on-a-chip reconstructing tissue-tissue interfaces
,”
Front. Oncol.
12
, 959299 (
2022
).
37.
J. B.
Lee
,
J. S.
Park
,
Y. M.
Shin
,
D. H.
Lee
,
J.-K.
Yoon
,
D.-H.
Kim
,
U. H.
Ko
,
Y.
Kim
,
S. H.
Bae
, and
H.-J.
Sung
, “
Implantable vascularized liver chip for cross-validation of disease treatment with animal model
,”
Adv. Funct. Mater.
29
(
23
),
1900075
(
2019
).
38.
T.
Amemiya
,
N.
Hata
,
M.
Mizoguchi
,
R.
Yokokawa
,
Y.
Kawamura
,
R.
Hatae
,
Y.
Sangatsuda
,
D.
Kuga
,
Y.
Fujioka
,
K.
Takigawa
,
Y.
Akagi
,
K.
Yoshimoto
,
K.
Iihara
, and
T.
Miura
, “
Mesenchymal glioblastoma-induced mature de-novo vessel formation of vascular endothelial cells in a microfluidic device
,”
Mol. Biol. Rep.
48
(
1
),
395
403
(
2021
).
39.
J.
Deng
,
W.
Wei
,
Z.
Chen
,
B.
Lin
,
W.
Zhao
,
Y.
Luo
, and
X.
Zhang
, “
Engineered liver-on-a-chip platform to mimic liver functions and its biomedical applications: A review
,”
Micromachines
10
(
10
),
676
(
2019
).
40.
D. Y.
No
,
K.-H.
Lee
,
J.
Lee
, and
S.-H.
Lee
, “
3D liver models on a microplatform: Well-defined culture, engineering of liver tissue and liver-on-a-chip
,”
Lab. Chip
15
(
19
),
3822
3837
(
2015
).
41.
A.
Özkan
,
D.
Stolley
,
E. N. K.
Cressman
,
M.
McMillin
,
S.
DeMorrow
,
T. E.
Yankeelov
, and
M. N.
Rylander
, “
The influence of chronic liver diseases on hepatic vasculature: A liver-on-a-chip review
,”
Micromachines
11
(
5
),
487
(
2020
).
42.
K. A.
Homan
,
N.
Gupta
,
K. T.
Kroll
,
D. B.
Kolesky
,
M.
Skylar-Scott
,
T.
Miyoshi
,
D.
Mau
,
M. T.
Valerius
,
T.
Ferrante
,
J. V.
Bonventre
,
J. A.
Lewis
, and
R.
Morizane
, “
Flow-enhanced vascularization and maturation of kidney organoids in vitro
,”
Nat. Methods
16
,
255
262
(
2019
).
43.
H. N.
Lee
,
Y. Y.
Choi
,
J. W.
Kim
,
Y. S.
Lee
,
J. W.
Choi
,
T.
Kang
,
Y. K.
Kim
, and
B. G.
Chung
, “
Effect of biochemical and biomechanical factors on vascularization of kidney organoid-on-a-chip
,”
Nano Converg.
8
(
1
),
35
(
2021
).
44.
Y.
Roye
,
R.
Bhattacharya
,
X.
Mou
,
Y.
Zhou
,
M. A.
Burt
, and
S.
Musah
, “
A personalized glomerulus chip engineered from stem cell-derived epithelium and vascular endothelium
,”
Micromachines
12
(
8
),
967
(
2021
).
45.
N.
Ashammakhi
,
K.
Wesseling-Perry
,
A.
Hasan
,
E.
Elkhammas
, and
Y. S.
Zhang
, “
Kidney-on-a-chip untapped opportunities
,”
Kidney Int.
94
(
6
),
1073
1086
(
2018
).
46.
J.
Lee
and
S.
Kim
, “
Kidney-on-a-chip: A new technology for predicting drug efficacy, interactions, and drug-induced nephrotoxicity
,”
Curr. Drug Metab.
19
(
7
),
577
583
(
2018
).
47.
R.
Paoli
and
J.
Samitier
, “
Mimicking the kidney: A key role in organ-on-chip development
,”
Micromachines
7
(
7
),
126
(
2016
).
48.
J. A.
McGrath
and
J.
Uitto
, “
Anatomy and organization of human skin
,” in
Rook’s Textbook of Dermatology
(
Wiley-Blackwell
,
Oxford
,
2010
), pp.
1
53
.
49.
N.
Mori
,
Y.
Morimoto
, and
S.
Takeuchi
, “
Skin integrated with perfusable vascular channels on a chip
,”
Biomaterials
116
,
48
56
(
2017
).
50.
S.
Salameh
,
N.
Tissot
,
K.
Cache
,
J.
Lima
,
I.
Suzuki
,
P. A.
Marinho
,
M.
Rielland
,
J.
Soeur
,
S.
Takeuchi
,
S.
Germain
, and
L.
Breton
, “
A perfusable vascularized full-thickness skin model for potential topical and systemic applications
,”
Biofabrication
13
(
3
),
035042
(
2021
).
51.
S.
Sun
,
L.
Jin
,
Y.
Zheng
, and
J.
Zhu
, “
Modeling human HSV infection via a vascularized immune-competent skin-on-chip platform
,”
Nat. Commun.
13
(
1
),
5481
(
2022
).
52.
S.
Biglari
,
T. Y. L.
Le
,
R. P.
Tan
,
S. G.
Wise
,
A.
Zambon
,
G.
Codolo
,
M.
De Bernard
,
M.
Warkiani
,
A.
Schindeler
,
S.
Naficy
,
P.
Valtchev
,
A.
Khademhosseini
, and
F.
Dehghani
, “
Simulating inflammation in a wound microenvironment using a dermal wound-on-a-chip model
,”
Adv. Healthc. Mater.
8
(
1
),
1801307
(
2019
).
53.
J. S.
Gnecco
,
T.
Ding
,
C.
Smith
,
J.
Lu
,
K. L.
Bruner-Tran
, and
K. G.
Osteen
, “
Hemodynamic forces enhance decidualization via endothelial-derived prostaglandin E2 and prostacyclin in a microfluidic model of the human endometrium
,”
Human Reprod.
34
(
4
),
702
714
(
2019
).
54.
M.
Nishida
, “
The Ishikawa cells from birth to the present
,”
Hum. Cell
15
(
3
),
104
117
(
2002
).
55.
J.
Ahn
,
M.-J.
Yoon
,
S.-H.
Hong
,
H.
Cha
,
D.
Lee
,
H. S.
Koo
,
J.-E.
Ko
,
J.
Lee
,
S.
Oh
,
N. L.
Jeon
, and
Y.-J.
Kang
, “
Three-dimensional microengineered vascularised endometrium-on-a-chip
,”
Human Reprod.
36
(
10
),
2720
2731
(
2021
).
56.
S. I.
Ahn
and
Y.
Kim
, “
On-chip physiological mimicry of neurovascular unit: Challenges and perspectives
,”
Neural Regen. Res.
19
(
3
),
499
500
(
2024
).
57.
D. J.
Alcendor
,
F. E.
Block III
,
D. E.
Cliffel
,
J. S.
Daniels
,
K. L. J.
Ellacott
,
C. R.
Goodwin
,
L. H.
Hofmeister
,
D.
Li
,
D. A.
Markov
,
J. C.
May
,
L. J.
McCawley
,
B.
McLaughlin
,
J. A.
McLean
,
K. D.
Niswender
,
V.
Pensabene
,
K. T.
Seale
,
S. D.
Sherrod
,
H.-J.
Sung
,
D. L.
Tabb
,
D. J.
Webb
, and
J. P.
Wikswo
, “
Neurovascular unit on a chip: Implications for translational applications
,”
Stem Cell Res. Ther.
4
(
1
),
S18
(
2013
).
58.
W.-J.
Wei
,
Y.-C.
Wang
,
X.
Guan
,
W.-G.
Chen
, and
J.
Liu
, “
A neurovascular unit-on-a-chip: Culture and differentiation of human neural stem cells in a three-dimensional microfluidic environment
,”
Neural Regen. Res.
17
(
10
),
2260
2266
(
2022
).
59.
Y. I.
Wang
,
H. E.
Abaci
, and
M. L.
Shuler
, “
Microfluidic blood–brain barrier model provides in vivo-like barrier properties for drug permeability screening
,”
Biotechnol. Bioeng.
114
(
1
),
184
194
(
2017
).
60.
A. V.
Singh
,
V.
Chandrasekar
,
P.
Laux
,
A.
Luch
,
S. P.
Dakua
,
P.
Zamboni
,
A.
Shelar
,
Y.
Yang
,
V.
Pandit
,
V.
Tisato
, and
D.
Gemmati
, “
Micropatterned neurovascular interface to mimic the blood–brain barrier’s neurophysiology and micromechanical function: A BBB-on-CHIP model
,”
Cells
11
(
18
),
2801
(
2022
).
61.
M.
Campisi
,
Y.
Shin
,
T.
Osaki
,
C.
Hajal
,
V.
Chiono
, and
R. D.
Kamm
, “
3D self-organized microvascular model of the human blood-brain barrier with endothelial cells, pericytes and astrocytes
,”
Biomaterials
180
,
117
129
(
2018
).
62.
T.-E.
Park
,
N.
Mustafaoglu
,
A.
Herland
,
R.
Hasselkus
,
R.
Mannix
,
E. A.
FitzGerald
,
R.
Prantil-Baun
,
A.
Watters
,
O.
Henry
,
M.
Benz
,
H.
Sanchez
,
H. J.
McCrea
,
L. C.
Goumnerova
,
H. W.
Song
,
S. P.
Palecek
,
E.
Shusta
, and
D. E.
Ingber
, “
Hypoxia-enhanced blood-brain barrier chip recapitulates human barrier function and shuttling of drugs and antibodies
,”
Nat. Commun.
10
(
1
),
2621
(
2019
).
63.
A.
Özkan
,
D. L.
Stolley
,
E. N. K.
Cressman
,
M.
McMillin
,
T. E.
Yankeelov
, and
M.
Nichole Rylander
, “
Vascularized hepatocellular carcinoma on a chip to control chemoresistance through cirrhosis, inflammation and metabolic activity
,”
Small Struct.
4
(
9
),
2200403
(
2023
).
64.
S.
Park
,
T.
Kim
,
S.
Kim
,
S.
You
, and
Y.
Jung
, “
Three-dimensional vascularized lung cancer-on-a-chip with lung extracellular matrix hydrogels for in vitro screening
,”
Cancers
13
(
16
),
3930
(
2021
).
65.
G.
Silvani
,
C.
Basirun
,
H.
Wu
,
C.
Mehner
,
K.
Poole
,
P.
Bradbury
, and
J.
Chou
, “
A 3D-bioprinted vascularized glioblastoma-on-a-chip for studying the impact of simulated microgravity as a novel pre-clinical approach in brain tumor therapy
,”
Adv. Ther.
4
(
11
),
2100106
(
2021
).
66.
L.
Wan
,
C. A.
Neumann
, and
P. R.
LeDuc
, “
Tumor-on-a-chip for integrating a 3D tumor microenvironment: Chemical and mechanical factors
,”
Lab Chip
20
(
5
),
873
888
(
2020
).
67.
S. J.
Hachey
and
C. C. W.
Hughes
, “
Applications of tumor chip technology
,”
Lab Chip
18
(
19
),
2893
2912
(
2018
).
68.
X.
Liu
,
J.
Fang
,
S.
Huang
,
X.
Wu
,
X.
Xie
,
J.
Wang
,
F.
Liu
,
M.
Zhang
,
Z.
Peng
, and
N.
Hu
, “
Tumor-on-a-chip: From bioinspired design to biomedical application
,”
Microsyst. Nanoeng.
7
(
1
),
50
(
2021
).
69.
Y.
Kim
,
J.
Ko
,
N.
Shin
,
S.
Park
,
S.-R.
Lee
,
S.
Kim
,
J.
Song
,
S.
Lee
,
K.-S.
Kang
,
J.
Lee
, and
N. L.
Jeon
, “
All-in-one microfluidic design to integrate vascularized tumor spheroid into high-throughput platform
,”
Biotechnol. Bioeng.
119
(
12
),
3678
3693
(
2022
).
70.
V.
Carvalho
,
M.
Banobre-Lopez
,
G.
Minas
,
S. F. C. F.
Teixeira
,
R.
Lima
, and
R. O.
Rodrigues
, “The integration of spheroids and organoids into organ-on-a-chip platforms for tumour research: A review,”
Bioprinting
27
, e00224 (2022).
71.
L. L.
Bischel
,
K. E.
Sung
,
J. A.
Jiménez-Torres
,
B.
Mader
,
P. J.
Keely
, and
D. J.
Beebe
, “
The importance of being a lumen
,”
FASEB J.
28
(
11
),
4583
4590
(
2014
).
72.
W. C.
Aird
, “
Endothelial cell heterogeneity
,”
Cold Spring Harb. Perspect. Med.
2
(
1
),
a006429
(
2012
).
73.
J. M.
Gomez-Salinero
,
T.
Itkin
, and
S.
Rafii
, “
Developmental angiocrine diversification of endothelial cells for organotypic regeneration
,”
Dev. Cell
56
(
22
),
3042
3051
(
2021
).
74.
D. J.
Nolan
,
M.
Ginsberg
,
E.
Israely
,
B.
Palikuqi
,
M. G.
Poulos
,
D.
James
,
B.-S.
Ding
,
W.
Schachterle
,
Y.
Liu
,
Z.
Rosenwaks
,
J. M.
Butler
,
J.
Xiang
,
A.
Rafii
,
K.
Shido
,
S. Y.
Rabbany
,
O.
Elemento
, and
S.
Rafii
, “
Molecular signatures of tissue-specific microvascular endothelial cell heterogeneity in organ maintenance and regeneration
,”
Dev. Cell
26
(
2
),
204
219
(
2013
).
75.
H. G.
Augustin
and
G. Y.
Koh
, “
Organotypic vasculature: From descriptive heterogeneity to functional pathophysiology
,”
Science
357
(
6353
),
eaal2379
(
2017
).
76.
D. T.
Paik
,
L.
Tian
,
I. M.
Williams
,
S.
Rhee
,
H.
Zhang
,
C.
Liu
,
R.
Mishra
,
S. M.
Wu
,
K.
Red-Horse
, and
J. C.
Wu
, “
Single-cell RNA sequencing unveils unique transcriptomic signatures of organ-specific endothelial cells
,”
Circulation
142
(
19
),
1848
1862
(
2020
).
77.
M. A.
Erickson
and
W. A.
Banks
, “
Blood–brain barrier dysfunction as a cause and consequence of Alzheimer’s disease
,”
J. Cereb. Blood Flow Metab.
33
(
10
),
1500
1513
(
2013
).
78.
A.
Minagar
,
A. H.
Maghzi
,
J. C.
McGee
, and
J. S.
Alexander
, “
Emerging roles of endothelial cells in multiple sclerosis pathophysiology and therapy
,”
Neurol. Res.
34
(
8
),
738
745
(
2012
).
79.
Y.
Chen
,
Y.
Qi
, and
W.
Lu
, “
Endogenous vasoactive peptides and vascular aging-related diseases
,”
Oxid. Med. Cell. Longev.
2022
,
1534470
(
2022
).
80.
K. M.
Malinda
,
M.
Nomizu
,
M.
Chung
,
M.
Delgado
,
Y.
Kuratomi
,
Y.
Yamada
,
H. K.
Kleinman
, and
M. L.
Ponce
, “
Identification of laminin Α1 and Β1 chain peptides active for endothelial cell adhesion, tube formation, and aortic sprouting
,”
FASEB J.
13
(
1
),
53
62
(
1999
).
81.
M. H.
Fittkau
,
P.
Zilla
,
D.
Bezuidenhout
,
M. P.
Lutolf
,
P.
Human
,
J. A.
Hubbell
, and
N.
Davies
, “
The selective modulation of endothelial cell mobility on RGD peptide containing surfaces by YIGSR peptides
,”
Biomaterials
26
(
2
),
167
174
(
2005
).
82.
M.
Morales-Ruiz
,
D.
Fulton
,
G.
Sowa
,
L. R.
Languino
,
Y.
Fujio
,
K.
Walsh
, and
W. C.
Sessa
, “
Vascular endothelial growth factor–stimulated actin reorganization and migration of endothelial cells is regulated via the serine/threonine kinase Akt
,”
Circ. Res.
86
(
8
),
892
896
(
2000
).
83.
Y.
Zhang
,
X.
Wang
,
Y.
Zhang
,
Y.
Liu
,
D.
Wang
,
X.
Yu
,
H.
Wang
,
Z.
Bai
,
Y.
Jiang
,
X.
Li
,
W.
Zheng
, and
Q.
Li
, “
Endothelial cell migration regulated by surface topography of poly(ε-caprolactone) nanofibers
,”
ACS Biomater. Sci. Eng.
7
(
10
),
4959
4970
(
2021
).
84.
D.
Bouïs
,
G. A. P.
Hospers
,
C.
Meijer
,
G.
Molema
, and
N. H.
Mulder
, “
Endothelium in vitro: A review of human vascular endothelial cell lines for blood vessel-related research
,”
Angiogenesis
4
(
2
),
91
102
(
2001
).
85.
T.
Tricot
,
C. M.
Verfaillie
, and
M.
Kumar
, “
Current status and challenges of human induced pluripotent stem cell-derived liver models in drug discovery
,”
Cells
11
(
3
),
442
(
2022
).
86.
S.-P.
Peng
and
S.
Copray
, “
Comparison of human primary with human iPS cell-derived dopaminergic neuron grafts in the Rat model for Parkinson’s disease
,”
Stem Cell Rev. Rep.
12
,
105
120
(
2016
).
87.
R.
Konoe
and
R.
Morizane
, “
Strategies for improving vascularization in kidney organoids: A review of current trends
,”
Biology
12
(
4
),
503
(
2023
).
88.
Y. K.
Kurokawa
,
R. T.
Yin
,
M. R.
Shang
,
V. S.
Shirure
,
M. L.
Moya
, and
S. C.
George
, “
Human induced pluripotent stem cell-derived endothelial cells for three-dimensional microphysiological systems
,”
Tissue Eng. Part C: Methods
23
(
8
),
474
484
(
2017
).
89.
C. C.
Kennedy
,
E. E.
Brown
,
N. O.
Abutaleb
, and
G. A.
Truskey
, “
Development and application of endothelial cells derived from pluripotent stem cells in microphysiological systems models
,”
Front. Cardiovasc. Med.
8
, 625016 (
2021
).
90.
S.
Baiguera
and
D.
Ribatti
, “
Endothelialization approaches for viable engineered tissues
,”
Angiogenesis
16
(
1
),
1
14
(
2013
).
91.
R.
Zheng
,
F.
Li
,
F.
Li
, and
A.
Gong
, “
Targeting tumor vascularization: Promising strategies for vascular normalization
,”
J. Cancer Res. Clin. Oncol.
147
(
9
),
2489
2505
(
2021
).
92.
M.
Lovett
,
K.
Lee
,
A.
Edwards
, and
D. L.
Kaplan
, “
Vascularization strategies for tissue engineering
,”
Tissue Eng. Part B: Rev.
15
(
3
),
353
370
(
2009
).
93.
E. S.
Bayrak
,
B.
Akar
,
N.
Xiao
,
H.
Mehdizadeh
,
S. I.
Somo
,
E. M.
Brey
, and
A.
Cinar
, “
Agent-based modeling of vascularization in gradient tissue engineering constructs
,”
IFAC Pap.
48
(
8
),
1240
1245
(
2015
).
94.
Y.-C.
Chiu
,
M.-H.
Cheng
,
H.
Engel
,
S.-W.
Kao
,
J. C.
Larson
,
S.
Gupta
, and
E. M.
Brey
, “
The role of pore size on vascularization and tissue remodeling in PEG hydrogels
,”
Biomaterials
32
(
26
),
6045
6051
(
2011
).
95.
M. T.
Poldervaart
,
H.
Gremmels
,
K.
van Deventer
,
J. O.
Fledderus
,
F. C.
Öner
,
M. C.
Verhaar
,
W. J. A.
Dhert
, and
J.
Alblas
, “
Prolonged presence of VEGF promotes vascularization in 3D bioprinted scaffolds with defined architecture
,”
J. Controlled Release
184
,
58
66
(
2014
).
96.
I.
Freeman
and
S.
Cohen
, “
The influence of the sequential delivery of angiogenic factors from affinity-binding alginate scaffolds on vascularization
,”
Biomaterials
30
(
11
),
2122
2131
(
2009
).
97.
I. E.
Hoefer
,
B.
den Adel
, and
M. J. A. P.
Daemen
, “
Biomechanical factors as triggers of vascular growth
,”
Cardiovasc. Res.
99
(
2
),
276
283
(
2013
).
98.
A.
Mykuliak
,
A.
Yrjänäinen
,
A.-J.
Mäki
,
A.
Gebraad
,
E.
Lampela
,
M.
Kääriäinen
,
T.-K.
Pakarinen
,
P.
Kallio
,
S.
Miettinen
, and
H.
Vuorenpää
, “
Vasculogenic potency of bone marrow- and adipose tissue-derived mesenchymal stem/stromal cells results in differing vascular network phenotypes in a microfluidic chip
,”
Front. Bioeng. Biotechnol.
10
,
764237
(
2022
).
99.
W. Y.
Wang
,
R. N.
Kent
,
S. A.
Huang
,
E. H.
Jarman
,
E. H.
Shikanov
,
C. D.
Davidson
,
H. L.
Hiraki
,
D.
Lin
,
M. A.
Wall
,
D. L.
Matera
,
J.-W.
Shin
,
W. J.
Polacheck
,
A.
Shikanov
, and
B. M.
Baker
, “
Direct comparison of angiogenesis in natural and synthetic biomaterials reveals that matrix porosity regulates endothelial cell invasion speed and sprout diameter
,”
Acta Biomater.
135
,
260
273
(
2021
).
100.
A. R.
Henderson
,
H.
Choi
, and
E.
Lee
, “
Blood and lymphatic vasculatures on-chip platforms and their applications for organ-specific in vitro modeling
,”
Micromachines
11
(
2
),
147
(
2020
).
101.
P.
Fathi
,
G.
Holland
,
D.
Pan
, and
M. B.
Esch
, “
Lymphatic vessel on a chip with capability for exposure to cyclic fluidic flow
,”
ACS Appl. Bio Mater.
3
(
10
),
6697
6707
(
2020
).
102.
Y.
Yang
,
P.
Fathi
,
G.
Holland
,
D.
Pan
,
N. S.
Wang
, and
M. B.
Esch
, “
Pumpless microfluidic devices for generating healthy and diseased endothelia
,”
Lab Chip
19
(
19
),
3212
3219
(
2019
).
103.
K.
Haase
and
R. D.
Kamm
, “
Advances in on-chip vascularization
,”
Regen. Med.
12
(
3
),
285
302
(
2017
).
104.
R. A.
Nelson
,
E. K.
Rhee
,
M.
Alaeddine
, and
M.
Nikkhah
, “
Advances in biomaterials for promoting vascularization
,”
Curr. Stem Cell Rep.
8
(
4
),
184
196
(
2022
).
105.
A. M. A. O.
Pollet
and
J. M. J.
den Toonder
, “
Recapitulating the vasculature using organ-on-chip technology
,”
Bioengineering
7
(
1
),
17
(
2020
).
106.
A.
Aazmi
,
H.
Zhou
,
Y.
Li
,
M.
Yu
,
X.
Xu
,
Y.
Wu
,
L.
Ma
,
B.
Zhang
, and
H.
Yang
, “
Engineered vasculature for organ-on-a-chip systems
,”
Engineering
9
,
131
147
(
2022
).
107.
P.
Brumm
,
A.
Fritschen
,
L.
Doß
,
E.
Dörsam
, and
A.
Blaeser
, “
Fabrication of biomimetic networks using viscous fingering in flexographic printing
,”
Biomed. Mater.
17
(
4
),
045012
(
2022
).
108.
J.
Konopka
,
D.
Kołodziejek
,
M.
Flont
,
A.
Żuchowska
,
E.
Jastrzębska
, and
Z.
Brzózka
, “
Exploring endothelial expansion on a chip
,”
Sensors
22
(
23
),
9414
(
2022
).
109.
M.
Mao
,
H. P.
Bei
,
C. H.
Lam
,
P.
Chen
,
S.
Wang
,
Y.
Chen
,
J.
He
, and
X.
Zhao
, “
Human-on-leaf-chip: A biomimetic vascular system integrated with chamber-specific organs
,”
Small
16
(
22
),
2000546
(
2020
).
110.
S.
Rajasekar
,
D. S. Y.
Lin
,
F.
Zhang
,
A.
Sotra
,
A.
Boshart
,
S.
Clotet-Freixas
,
A.
Liu
,
J. A.
Hirota
,
S.
Ogawa
,
A.
Konvalinka
, and
B.
Zhang
, “
Subtractive manufacturing with swelling induced stochastic folding of sacrificial materials for fabricating complex perfusable tissues in multi-well plates
,”
Lab Chip
22
(
10
),
1929
1942
(
2022
).
111.
Q.
Li
,
K.
Niu
,
D.
Wang
,
L.
Xuan
, and
X.
Wang
, “
Low cost rapid prototyping and assembly of an open microfluidic device for a 3D vascularized organ-on-a-chip
,”
Lab Chip
22
,
2682
2694
(
2022
).
112.
D. T. T.
Phan
,
X.
Wang
,
B. M.
Craver
,
A.
Sobrino
,
D.
Zhao
,
J. C.
Chen
,
L. Y. N.
Lee
,
S. C.
George
,
A. P.
Lee
, and
C. C. W.
Hughes
, “
A vascularized and perfused organ-on-a-chip platform for large-scale drug screening applications
,”
Lab Chip
17
(
3
),
511
520
(
2017
).
113.
Y.
Liu
,
C.
Sakolish
,
Z.
Chen
,
D. T. T.
Phan
,
R. H. F.
Bender
,
C. C. W.
Hughes
, and
I.
Rusyn
, “
Human in vitro vascularized micro-organ and micro-tumor models are reproducible organ-on-a-chip platforms for studies of anticancer drugs
,”
Toxicology
445
,
152601
(
2020
).
114.
D. E.
Ingber
, “
Human organs-on-chips for disease modelling, drug development and personalized medicine
,”
Nat. Rev. Genet.
23
(
8
),
467
491
(
2022
).
115.
A.
Herland
,
B. M.
Maoz
,
D.
Das
,
M. R.
Somayaji
,
R.
Prantil-Baun
,
R.
Novak
,
M.
Cronce
,
T.
Huffstater
,
S. S. F.
Jeanty
,
M.
Ingram
,
A.
Chalkiadaki
,
D.
Benson Chou
,
S.
Marquez
,
A.
Delahanty
,
S.
Jalili-Firoozinezhad
,
Y.
Milton
,
A.
Sontheimer-Phelps
,
B.
Swenor
,
O.
Levy
,
K. K.
Parker
,
A.
Przekwas
, and
D. E.
Ingber
, “
Quantitative prediction of human pharmacokinetic responses to drugs via fluidically coupled vascularized organ chips
,”
Nat. Biomed. Eng.
4
(
4
),
421
436
(
2020
).
116.
K.
Ronaldson-Bouchard
,
D.
Teles
,
K.
Yeager
,
D. N.
Tavakol
,
Y.
Zhao
,
A.
Chramiec
,
S.
Tagore
,
M.
Summers
,
S.
Stylianos
,
M.
Tamargo
,
B. M.
Lee
,
S. P.
Halligan
,
E. H.
Abaci
,
Z.
Guo
,
J.
Jacków
,
A.
Pappalardo
,
J.
Shih
,
R. K.
Soni
,
S.
Sonar
,
C.
German
,
A. M.
Christiano
,
A.
Califano
,
K. K.
Hirschi
,
C. S.
Chen
,
A.
Przekwas
, and
G.
Vunjak-Novakovic
, “
A multi-organ chip with matured tissue niches linked by vascular flow
,”
Nat. Biomed. Eng.
6
(
4
),
351
371
(
2022
).
117.
R.
Novak
,
M.
Ingram
,
S.
Marquez
,
D.
Das
,
A.
Delahanty
,
A.
Herland
,
B. M.
Maoz
,
S. S. F.
Jeanty
,
M. R.
Somayaji
,
M.
Burt
,
E.
Calamari
,
A.
Chalkiadaki
,
A.
Cho
,
Y.
Choe
,
D. B.
Chou
,
M.
Cronce
,
S.
Dauth
,
T.
Divic
,
J.
Fernandez-Alcon
,
T.
Ferrante
,
J.
Ferrier
,
E. A.
FitzGerald
,
R.
Fleming
,
S.
Jalili-Firoozinezhad
,
T.
Grevesse
,
J. A.
Goss
,
T.
Hamkins-Indik
,
O.
Henry
,
C.
Hinojosa
,
T.
Huffstater
,
K.-J.
Jang
,
V.
Kujala
,
L.
Leng
,
R.
Mannix
,
Y.
Milton
,
J.
Nawroth
,
B. A.
Nestor
,
C. F.
Ng
,
B.
O’Connor
,
T.-E.
Park
,
H.
Sanchez
,
J.
Sliz
,
A.
Sontheimer-Phelps
,
B.
Swenor
,
G.
Thompson
,
G. J.
Touloumes
,
Z.
Tranchemontagne
,
N.
Wen
,
M.
Yadid
,
A.
Bahinski
,
G. A.
Hamilton
,
D.
Levner
,
O.
Levy
,
A.
Przekwas
,
R.
Prantil-Baun
,
K. K.
Parker
, and
D. E.
Ingber
, “
Robotic fluidic coupling and interrogation of multiple vascularized organ chips
,”
Nat. Biomed. Eng.
4
(
4
),
407
420
(
2020
).