A microfluidic device for investigating epithelial and / or endothelial barrier functions, such as in placental-like or breast duct-like 3D microvascular tissues

Abstract

The present invention relates to a microfluidic device configured to study epithelial and / or endothelial barrier functions in a human 3D cellular microtissue model, while also allowing the application of mechanical stimuli, such as a luminal and / or lateral flow. Embodiments comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP) representative of a maternal-fetal interface (placental barrier), as well as a mammary duct interface, that is comprised of human primary mammary-derived endothelial cells, fibroblasts and inclusion of an epithelial layer (MCF10). The model can be extended to include tumors (cell line derived or patient-derived). The present invention further relates to a method for producing the human 3D vascular microtissue model using the microfluidic device according to the invention, as well as uses of the human 3D vascular microtissue models, for example to study cellular barrier function (permeability to solutes), the effects of interstitial, luminal and / or lateral flow and to study epithelial and / or endothelial barrier function (permeability to solutes), and extravascular matrix properties (diffusivity, stiffness, transport of molecules, antibodies and / or cells, and matrix proteins). The present invention further relates methods for representing a pre-eclamptic (disease) model of the placental barrier, and means for testing drug treatments.

Description

[0001] A microfluidic device for investigating epithelial and / or endothelial barrier functions, such as in placental-like or breast duct-like 3D microvascular tissues

[0002] The present invention relates to a microfluidic device configured to study epithelial and / or endothelial barrier functions in a human 3D cellular microtissue model, while also allowing the application of mechanical stimuli, such as a luminal and / or lateral flow. Embodiments comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP) representative of a maternal-fetal interface (placental barrier), as well as a mammary duct interface, that is comprised of human primary mammary-derived endothelial cells, fibroblasts and inclusion of an epithelial layer (MCF10). The model can be extended to include tumors (cell line derived or patient-derived). The present invention further relates to a method for producing the human 3D vascular microtissue model using the microfluidic device according to the invention, as well as uses of the human 3D vascular microtissue models, for example to study cellular barrier function (permeability to solutes), the effects of interstitial, luminal and / or lateral flow and to study epithelial and / or endothelial barrier function (permeability to solutes), and extravascular matrix properties (diffusivity, stiffness, transport of molecules, antibodies and / or cells, and matrix proteins). The present invention further relates methods for representing a pre-eclamptic (disease) model of the placental barrier, and means for testing drug treatments.

[0003] Background of the invention

[0004] Human placentae are highly vascularized organs that undergo constant vascular growth and remodeling to ensure effective exchange between maternal blood and fetal circulatory systems (1). The majority of the essential solute exchange (oxygen, nutrients, hormones, antibodies) takes place across a thin layer of epithelial cells of the terminal villous trees, which contains a network of fetoplacental capillaries. This dense capillary network develops from the elongation and ramification of pre-existing blood vessels formed by vasculogenesis (new vessel formation) by approximately 6 weeks of gestation. From 25 weeks post-conception until term, villous vascular growth switches from branching angiogenesis to non-branching angiogenesis generating coiled capillary structures that reside within the extremities of the fetoplacental vascular trees (2). Formation of a proper vascular tree in the early placenta is crucial to ensure optimal blood volume loading, preventing placental insufficiency and adverse fetal outcomes. Structural villous microvascular network abnormalities are common features of gestational disorders associated with perinatal morbidity and mortality such as in pre-eclampsia and fetal growth restriction (3,4). To determine the etiology of these alterations it is crucial to understand the mechanisms underlying fetoplacental vasculogenesis and regulation.

[0005] Studying fetoplacental vascular development is challenging, mainly due to ethical considerations and inaccessibility of the tissue, particularly in early pregnancy. Current knowledge on villous vascular network development and function comes from observations of human placenta explants obtained at different stages of gestation (5,6), ultrasound examinations (7) and in silico modelling (8,9). However, these methods cannot provide a direct measurement of the physical and chemical cues occurring during placental vasculogenesis and angiogenesis, such as mechanical shear stress and associated mechanotransduction. In fact, developing vessels in vivo are exposed to constant mechanical stimuli including fluid shear stress induced by interstitial flow (IF) and intraluminal blood flow (10).

[0006] Several in vivo and in vitro studies have documented the effects of flow-induced shear stress on vascular morphogenesis, demonstrating flow-induced regulation of angiogenic sprouting and lumen formation by endothelial cell migration, alignment, and apical deformation (11-14). Hemodynamic forces trigger vessel remodeling through cellular rearrangement and regulate vascular permeability via modulation of endothelial adherence and tight junctions (15-18). The role of human placental hemodynamics during pregnancy remains poorly understood since in vivo experimentation has mainly focused on umbilical circulation (19), lacking proper characterization of the placental micro-hemodynamics occurring at the fetoplacental interface. Although the placenta at term constitutes a valid tool for ex vivo perfusion studies (20), an understanding of early placental vascular circulations is limited, requiring the employment of biomimetic models.

[0007] Three-dimensional human tissue-specific microvascular networks have recently contributed to the inventor’ s understanding of flow-induced vascular growth and remodeling, by overcoming the limitations of simpler monolayer systems (14, 16, 21, 22). For instance, in another difficult- to-study tissue, it has been demonstrated that interstitial flow enhances vessel formation, function and longevity in brain-specific microvasculature comprised of endothelial cells, pericytes, and astrocytes (23).

[0008] Deng P, et al. (in: Fluidic Flow Enhances the Differentiation of Placental Trophoblast-Like 3D Tissue from hiPSCs in a Perfused Macrofluidic Device. Front Bioeng Biotechnol. 2022 Jun 30;10:907104. doi: 10.3389 / fbioe.2022.907104. PMID: 35845423; PMCID: PMC9280037) describe an in vitro placental trophoblast-like model via self-organization of human induced pluripotent stem cells (hiPSCs) in a perfused 3D culture macrofluidic device. This device allowed cell seeding, in situ trophoblast lineage differentiation, and formation of trophoblastlike tissues from hiPSCs in a biomimetic microenvironment. It incorporated extracellular matrix (ECM) and fluid flow in a single device. After trophoblast lineage differentiation, Deng et al. were able to generate the 3D clusters with major cell types of the human placenta, including trophoblast progenitor cytotrophoblasts (CTBs), differentiated subtypes, syncytiotrophoblasts (STBs), and extravillous trophoblasts (EVTs) under long-term 3D culture (-23 days). Moreover, the formed tissues exhibited enhanced expressions of CTB-, STB-, and EVT-related markers at the level of genes and proteins under a dynamic culture compared with static conditions. RNA-seq analysis revealed the higher expression of trophoblast-specific genes in 3D tissues, indicating the essential role of fluid flow to promote the trophoblast differentiation of hiPSCs. The device as disclosed does not permit application of flow, such as a lateral flow, and cannot be used to efficiently study a vascular and / or barrier function of the placental tissue.

[0009] Nishiguchi A, et al. (in: In vitro placenta barrier model using primary human trophoblasts, underlying connective tissue and vascular endothelium. Biomaterials. 2019 Feb; 192: 140-148. doi: 10.1016 / j.biomaterials.2018 08.025. Epub 2018 Aug 25. PMID: 30448698) disclose a 3D- vascularized human primary placental barrier model (transwell model) comprising a layer of laminin and collagen-coated trophoblasts (either BeWo or primary cytotrophoblasts) onto a thick layer of self-assembled capillary networks, formed from primary fibroblasts (normal human dermal fibroblasts (NHDFs)) and human umbilical vein endothelial cells (HUVECs) in a fibrin hydrogel. The perfusion of vasculature is not possible with this model.

[0010] Pu, et al. (in: A 3-dimensional microfluidic platform for modeling human extravillous trophoblast invasion and toxicological screening. Lab Chip. 2021 Feb 9;21 (3):546-557. doi: 10.1039 / d01c01013h. PMID: 33166377; PMCID: PMC8212566) describe a placenta-on-a-chip platform that enables the evaluation of trophoblast invasion with intraluminal flow into a nonlumen vascular center within an engineered PDMS 3D microfluidic chip, further requiring a geometrically defined outer layer supported by posts.

[0011] WO 2016-154319A1 discloses a microfluidic device that can simulate capillary blood flow on a fetal side of the device and pooled blood on a maternal side of the device (i.e., intervillous space) separated by a permeable membrane. The microfluidic device can reconstitute the maternal-fetal interface, can expand the capabilities of cell culture models, and can provide an alternative to current maternal-fetal transfer models. Similarly, EP3404093B1 generally relates to an organ mimic device with microchannels and methods of use and manufacturing thereof.

[0012] Despite the development of novel in vitro placental models (reviewed in (24)), at present, no models investigate the effect of flow-associated placental vascular development and at the same time allow for a testing of prospective and existing pharmaceuticals or other bio-relevant compounds for use in human pregnancy. Human placental explant tissues are difficult to maintain, perfuse, and usually represent only the end stages of pregnancy. Recently, several humanized in vitro models have been designed to mimic aspects of the placental barrier (105); however, they lack the combination of the two crucial adjacent barriers regulating maternal- fetal exchange, namely the epithelial-like barrier of trophoblasts and the underlying fetal vasculature.

[0013] Furthermore, several in vitro placental barrier models have been developed to measure the transfer of substances between the mother and fetus (82). The different technologies that currently exist may be summarized as follows.

[0014] In the transwell technology monolayers of trophoblast cells are grown on plates or semipermeable membranes to investigate placental barrier properties such as hormone secretion, transcellular transport of glucose, environmental toxicants, and susceptibility to parasite infection (83-85).

[0015] In the micromesh cell culture technology, a multilayer placental barrier model using endothelial cells, fibroblast and iPSC-derived trophoblast cells in order to study drug permeability (86). In the microfluidic technology, dual 2D layers of endothelial cells and trophoblast are seeded either side of a semipermeable membrane, recapitulating the transport barrier from the maternal to fetal circulations (see, for example, Rabussier G, et al. Healthy and diseased placental barrier on-a-chip models suitable for standardized studies. Acta Biomater. 2023 Jul l;164:363-376. doi: 10.1016 / j.actbio.2023.04.033. Epub 2023 Apr 26. PMID: 37116636). Fluid flow through separate microfluidic channels either side of this cell barrier enables both exposure to cell specific medium, and quantification of transport across the layer. This allows for an exchange of continuous flowing nanoparticles across trophoblast cell lines (87, 88), and / or a quantification of transport across the layer (89-91).

[0016] Finally, 3D and more complex microfluidic models comprise culturing of cells in three- dimensions to better recapitulate the complex microenvironment and cellular cross-talk involved in placental development. These are found as 3D trophoblast organoid models capable of self-renewal, expansion and differentiation (92, 93), rotating wall bioreactor system to develop a 3D-based cell coculture (trophoblast and microvascular endothelial cells) model that that both simulate fluid flow and recreate the suspended microgravity environment found within in vivo tissues (94), flow culture systems for placental villous explants to study the impact of shear stress impact on tissue structural integrity (95), and finally “placenta-on-chip” that integrates vascular networks (96, 105) or the complex geometry of the villi structure using gelatin-based hydrogel material (97).

[0017] Previously, the inventors established a 3D model of terminal villi microvasculature was established using a triculture of stromal and endothelial cells ((25) incorporated herein by reference for the purposes of the invention in its entirety). This model demonstrated that placental pericytes contribute to growth restriction, which was largely dependent on VEGF and angiopoietin / Tie2 signaling. Fibroblasts, on the other hand, contributed to increased vasculogenesis; however, a limitation of the former model was the use of non-specific fibroblasts.

[0018] All in vitro models that have been designed so far in order to mimic aspects of the placental barrier; however, still lack the combination of the two crucial adjacent barriers that are involved in maternal-fetal exchange and do not comprise all the required components, namely endothelium, stromal tissue and trophoblasts. It is therefore an object of the present invention to provide an improved human placental 3D vascular model that reflects the conditions in vivo and can be used to properly investigate the effect of cellular barrier functions, such as flow-associated vascular and epithelial barrier functions and morphology. Other objects and advantages will become apparent to the person of skill when further studying the present disclosure.

[0019] In a first aspect of the present invention, the present invention solves the above object by providing a microfluidic device, comprising a) an elongated central culture channel for holding mammalian cells in a suitable gel, b) at least one first media channel extending at least in part thereof arranged in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel, configured to allow a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, and c) at least one third media channel extending at least in part thereof in parallel to a third side of the elongated central culture channel, wherein optionally the at least one third media channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one third media channel, while preparing the device. This enables containment of the cell and gel mixture in the central channel, and once dissolved allows for an adjacent cell monolayer to be formed.

[0020] Preferably, the microfluidic devices according to the present invention advantageously are a post-less system, which renders them less complicated to make and allows for direct contact between media channel and gel, i.e. does not provide a physical barrier.

[0021] Preferred is the microfluidic device according to the present invention, wherein the mammalian cells in the culture chamber are selected from human cells, and preferably comprise cells and / or cell lines, such as cancer and / or immortalized cell lines, derived from epithelial tissues having an epithelial layer, vascular cells, cells of the respiratory system, cells of the digestive system, bladder cells, mammary cells, ductal cells and cancer cells thereof, and preferably comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP), and a layer of human trophoblasts, located on the side of the cell culture that is in contact with the at least one third media channel, and further preferably comprises a combination of mammary cells and epithelial cells.

[0022] In a second aspect of the present invention, the present invention solves the above object by providing a microfluidic device, comprising a) an elongated central culture channel for holding human cells in a suitable gel, b) at least one first media channel extending at least in part thereof in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel, configured to allow a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, wherein the mammalian cells in the gel comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP).

[0023] Preferably, this microfluidic device according to the present invention advantageously also is a post-less system, which renders it less complicated to make, and allows for direct contact between media channel and gel, i.e. does not provide a physical barrier.

[0024] Preferred is the microfluidic device according to the present invention, wherein the device furthermore comprises a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel.

[0025] The benefit of the above novel designs allows for transport to be explored across the entire barrier inclusive of trophoblasts, fetal vasculature and its stroma. The devices further allow for assays for examining transport across the maternal-fetal interface. The model is preferably designed to be representative of the physical barrier between mother and fetus and preferably includes trophoblasts and fetal vasculature derived from human cell sources. This system mimics the role of the placenta which controls transfer of solutes, antibodies, or other foreign- fetal material.

[0026] Safety testing of pharmaceuticals or other bio-relevant compounds for use in human pregnancy has been extremely limited for ethical reasons as well as the lack of a relevant and suitable model. For instance, animal models lack the species-specific placentation of humans and alternatively human placental explant tissues are difficult to maintain, perfuse, and usually represent only the end stages of pregnancy. Recently, as mentioned above, several humanized in vitro models have been designed to mimic aspects of the placental barrier in vivo; however, these lack the combination of the two crucial barriers (here, an appropriate amount of trophoblasts and placental fetal vasculature) regulating maternal-fetal exchange. The benefit of the present novel design allows for transport to be explored across the entire barrier such as the one inclusive of trophoblasts, fetal vasculature and its stroma.

[0027] Following from the above-mentioned development of fetal-microvessels, the inventors have further developed new designs (Figures 9, 10 and 11) that in the preferred embodiment thereof integrates trophoblasts into the system, generating the critical epithelial layer that forms a barrier between the maternal blood and fetal capillaries. These models may act as the barrier in the terminal villi of the placenta. The general method for producing the new design is to (1) introduce a temperature-dependent dissolvable gel into the maternal' channel of the device, (2) seed fetal microvessels into the "fetal' side including the culture chamber, and (3) dissolve the gel and about 48 hours later seed an appropriate amount of barrier cells, such as, for example, more than 1.0 x 106cells / mL, such as approx. 1.5 x 106cells / mL of trophoblasts into the maternal' channel (immortalized trophoblast cell lines, BeWo and HTR-8 / SVneo , and primary isolated trophoblasts (see Fig. 13), may be used as preferred examples), or (3a) as in (3) dissolve the gel in the third or fourth media channel and establish barrier cells on the third and / or fourth side of the culture chamber as described herein.

[0028] The inventive design is unique allowing the development of a combined trophoblast layer and perfusable fetal vasculature without the need for a separational mesh. Furthermore, perfusion of both the maternal and fetal compartments is possible. The design has the possibility to generate pressure gradients across the maternal-fetal interface and allows for the ability to examine both trophoblast barrier and fetal vessel barrier properties.

[0029] Furthermore, in a preferred embodiment thereof, the design has the capacity to integrate a variety of cells, e g., of primary tissue from patients, and / or is adopted to be used with other tissues / organs with an epithelial layer as disclosed herein.

[0030] Preferred is the microfluidic device according to the present invention, wherein the HPF and HPP are cultured at a ratio of about 10:1 of endothelial: stromal cells HPF / HPP, preferably wherein the HPF and HPP are at a ratio of 1 : 1. Further preferred is the microfluidic device according to the present invention, wherein the device is positioned on a solid support, such as a biochip, coverslip, and / or wherein several or a multitude of devices are grouped on said solid support.

[0031] In a third aspect of the present invention, the present invention solves the above object by providing a method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to the present invention, here in the “gel-barrier version” comprising the gel in at least one of the medium channels, b) incubating the cells in the culture chamber in a suitable culture medium, and c) dissolving of at least one of the gel barrier(s) as described herein and seeding suitable mammalian barrier cells as described herein.

[0032] Preferred is a method for producing a placental-derived human 3D vascular microtissue model according to the present invention, wherein in b) the tri-culture is incubated for about 48h in a suitable culture medium under essentially static conditions, bl) a lateral flow is established, for example an intermittent interstitial flow (IF), from at least one first media channel to at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of preferably between about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of preferably between about 0.20 pm / s and 1.25 pm / s across the central culture channel, and wherein the flow is established between about day 2 and about day 7 of the tri-culture.

[0033] In a fourth aspect of the present invention, the present invention solves the above object by providing a method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to the present invention, here designated as the “vascular version”, b) incubating the tri-culture for about 48h in a suitable culture medium under essentially static conditions, c) establishing a lateral flow, for example an intermittent interstitial flow (IF), from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of preferably between about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of preferably between about 0.20 pm / s and 1.25 pm / s across the central culture channel, wherein the flow is established between about day 2 and about day 7 of the tri-culture, and d) optionally, testing the perfusion rate following step c).

[0034] Preferred is the method according to the present invention, further comprising the step of introducing a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel (the “vascular barrier version”).

[0035] Preferred is the method according to the present invention, wherein the hydrostatic pressure gradient (AP) as established in step c) is at about less than 10 mm H2O, preferably at between about: 2 and 7 mm H2O. Further preferred is the method according to the present invention, wherein the pressure gradient is at 3 mm H2O or ~30 Pa or a mean lateral flow velocity of 0.30 ± 0.12 pm / s on day 2.

[0036] In a fifth aspect of the present invention, the present invention solves the above object by providing a placental-derived human 3D vascular microtissue model, produced according to a method according to the present invention. Preferably, the placental-derived human 3D vascular microtissue model according to the present invention exhibits a laminar flow across the model and / or a barrier function as described herein.

[0037] In a sixth aspect of the present invention, the present invention solves the above object by providing the use of the microfluidic device according to the present invention or the placental- derived human 3D vascular microtissue model according to the present invention to study placental barrier function(s), for drug screenings, for assessing novel dmg targets against disfunction, for example in pre-eclampsia, to study trophoblast cell invasion, or for toxicological screening of substances and stresses that interfere with trophoblast cell invasion, to study placental branching angiogenesis, the impact of nutrients, hormones, and environmental factors on placental development, gestational disorders of fetal vasculature, and / or vascular-associated gestational disorders, and / or to study epithelial and / or endothelial barrier function (permeability to solutes), transport of solutes, pollutants, antibodies, immune cells and drugs across an epithelial cell barrier and / or a vascular cell barrier and extravascular matrix properties (diffusivity, stiffness, and matrix proteins), and in particular the effects of luminal, interstitial and / or lateral flow on endothelial barrier function (permeability to solutes) and extravascular matrix properties (diffusivity, stiffness, and matrix proteins).

[0038] As mentioned above, in a first aspect thereof, the present invention provides a microfluidic device, comprising a) an elongated central culture channel for holding mammalian cells in a suitable gel, b) at least one first media channel extending at least in part thereof arranged in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel, configured to allow a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, and c) at least one third media channel extending at least in part thereof in parallel to a third side of the elongated central culture channel, wherein optionally the at least one third media channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one third media channel. The preferred suitable gel is a fibrinogen / thrombin gel (see examples below). The preferred dissolvable gel barrier comprises a temperature-dependent dissolvable gel, such as a gelatin.

[0039] Preferred is the microfluidic device according to the present invention, wherein said device comprises a fourth media channel extending at least in part thereof in parallel to a fourth side of the elongated central culture channel, wherein optionally the at least one third media channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one third media channel.

[0040] Further preferred is the microfluidic device according to the present invention, wherein the mammalian cells in the culture chamber are selected from human cells, and preferably comprise cells and / or cell lines, such as cancer and / or immortalized cell lines, derived from epithelial tissues having an epithelial layer, vascular cells, cells of the respiratory system, cells of the digestive system, bladder cells, mammary cells, ductal cells and cancer cells thereof, and preferably comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP), and and a layer of human trophoblasts, located on the side of the cell culture that is in contact with the at least one third or fourth media channel. In a second aspect thereof, the present invention provides a microfluidic device, comprising a) an elongated central culture channel or chamber for holding human cells in a suitable gel. In the case of the present placental model, a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP) is performed in this chamber. The preferred gel is a hydrogel, such as a fibrinogen / thrombin gel (see examples below).

[0041] The device according to the invention then comprises b) at least one first media channel extending at least in part thereof in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel. The channels (at least two are present in the device) allow a controlled flow of medium through the device, and in parallel to the elongated central culture channel or chamber (for a preferred embodiment, see Figure 9, 12 or 13. As a medium, any suitable medium for culturing the cells as involved can be used, such as endothelial medium (e.g., VascuLife medium) and / or fibroblast medium and / or pericyte growth medium (see examples below).

[0042] Importantly, the device according to the invention is configured to allow a lateral flow, such as an interstitial or luminal or sideways flow, from the at least one first media channel to the at least one second media channel through the central culture channel. In the context of the present invention, the term “lateral flow” shall also include to mean a pressure gradient that is established across a substantial part of the gel region of the culture chamber. This allows, amongst others, to study the impact of the interstitial flow (preferred embodiment of the lateral flow) or luminal flow on vascular and extravascular remodeling, as described herein.

[0043] The device may also comprise other mammalian cells in the gel of the culture chamber. These comprise any suitable cells and / or cell lines, such as cancer and / or immortalized cell lines, derived from epithelial tissues having an epithelial layer, vascular cells, cells of the respiratory system, cells of the digestive system, bladder cells, mammary cells, ductal cells and cancer cells thereof, and mixtures thereof. As mentioned, preferred is a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP). In a preferred microfluidic device according to the present invention, the device furthermore comprises a barrier layer of cells, preferably of epithelial cells, more preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel.

[0044] In contrast to the state of the art, placental fibroblasts are now included in a completely placental-derived human 3D vascular model, which is used to study the impact of interstitial or luminal flow on vascular and extravascular remodeling. The use of the placental fibroblasts advantageously provides a model that is as close to the situation in vivo as possible. Furthermore, the system can now include epithelial barrier cells and / or layers (see above). Therefore, as an alternative approach to animal testing, the invention comprises new devices and assays that allow for measurement of transfer of material across the maternal-fetal interface. As previously described, a human fetal-like vasculature is grown on a microfluidic chip in that human umbilical vein endothelial cells (HUVEC) were co-cultured in the presence of stromal cells (fibroblasts and pericytes), which form interconnected microvessels capable of perfusion in 5-7 days.

[0045] Subsequent integration of trophoblasts cells into the system generates the desired and critical epithelial layer that forms a barrier between the maternal blood and fetal capillaries (see Figure 9). The trophoblast monolayer was seeded adjacent to the fetal vessels first in the chip design after generation of the fetal vasculature, using in this example two immortalized lines, BeWo and HTR-8 / SVneo, and primary isolated trophoblasts (see Fig. 13).

[0046] The inventor’s results as obtained in the context of the present invention show that growth and remodeling of fetal microvessels, as well as study epithelial and / or endothelial barrier function (permeability to solutes) and extravascular matrix properties (diffusivity, stiffness, and matrix proteins) are impacted by the presence of flow. Moreover, flow-conditioned fetal microvessels exhibit significant changes in the expression of angiogenic and inflammatory cytokines. Computational fluid dynamics reveals an increasing trend of both luminal and interstitial velocities and shear stresses with increasing flow (Figure 5). Interstitial flow (in the extravascular region) promotes early connection of networks which remain viable and perfusable for long (weeks) culture periods. The cells and the microfluidic device(s) can be cultured under standard conditions at 37 °C and 5% CO2 in a standard incubator, or can be cultured in a hypoxic incubator.

[0047] Now incorporating placental fibroblasts, as opposed to co-culture with lung fibroblasts as done previously (25), the inventive model required adjustment. Preferably, HPF and HPP are cultured at a ratio of about 10:1 of endothelial: stromal cells HPF / HPP, preferably wherein the HPF and HPP are at a ratio of 1 :1. Furthermore, the device preferably comprises a gel comprising a hydrogel, such as a fibrinogen / thrombin gel (fibrin gel) at a concentration at from about 3 to about 3.5mg / ml. Nevertheless, other suitable gel systems may be used as well, such as collagen and alginates or the like. Also, an interstitial flow was introduced in order to achieve perfusable vessels. In this tri-culture placental fibroblasts associated with microvessels, as do pericytes. This shows that stromal cells have a clear role in contributing to fetoplacental vascular growth and remodeling.

[0048] In a preferred embodiment thereof, the microfluidic device according to the present invention further comprises media reservoirs, preferably 3D-printed tailored media reservoirs, connected to at least one of the at least one media channels. The reservoirs can be connected to the at least one media channel via inlets, and are used in order to generate a lateral flow based on hydrostatic pressure gradients (AP). For a preferred example, see examples below. A preferred gradient as established in c) of the device is at about less than 10 mm H2O, preferably at between about: 2 and 7 mm H2O.

[0049] Further preferred is the microfluidic device according to the present invention, wherein the device is positioned on a solid support, such as a biochip or coverslip, and / or wherein several or a multitude of devices are grouped on said solid support. In this embodiment, the microfluidic device according to the present invention is used as an array, for example to study (or test) properties of the placental-derived human 3D vascular microtissue model as present on the device. The test can then be done in an automated manner, e.g. by using robots, and automated optical analysis.

[0050] While being a preferred embodiment of the invention, the seeding of the monolayer on the side of the gel of the culture chamber left room for even more improvement. Gel retraction and cell aggregation at least to some extent prevented the formation of a connected monolayer (see Figures 9 to 11), as was found in case of using the BeWo cell line. Furthermore, poor vessel formation and no perfusability was found due to a blocking of the media distribution to the whole area of the gel, as was found in case of using the cells HTR- 8 / SVneo.

[0051] Therefore, in order to overcome these disadvantages, the inventors generated an alternate design that include an extra fluidic chamber for the epithelial interface (Figure 9A and 12 and 13). The inventors’ preliminary data show that in this system the vessel can form a proper connected network, and the epithelial barrier resembles physiological permeability properties (i.e., not permeable to 70kDa molecules, see Figure 11).

[0052] Methods of how to produce a microfluidic device according to the present invention are described herein and in reference (25) (incorporated by reference in its entirety), and are further disclosed in the literature as cited herein and known to the person of skill in the art.

[0053] As a material for the device according to the invention, any suitable biocompatible polymer can be used, preferred is Silicone Elastomer, such as, for example poly dimethyl siloxane (PDMS) elastomer (see below). Examples are known to the person of skill, and can be found in the literature (e.g. Okoshi, M , Yoshida, T. Fabrication of Silicone Rubber-Based Biochip for Disinfection Under t / / / Deep“UVttt pLight) Laser-Induced

[0054] Photodissociation. EZec / zOM. Mater. Lett. 17, 68^73 (2021)).

[0055] Another important aspect of the present invention then relates to a method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to the present invention as above.

[0056] In the next step of method for producing a placental-derived human 3D vascular microtissue model of the present invention, b) the cells in the culture chamber are incubated in a suitable culture medium. These comprise any suitable cells and / or cell lines, such as cancer and / or immortalized cell lines, derived from epithelial tissues having an epithelial layer, vascular cells, cells of the respiratory system, cells of the digestive system, bladder cells, mammary cells, ductal cells and cancer cells thereof, and mixtures thereof. As mentioned, preferred is a triculture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP) that is incubated for about 48h in a suitable culture medium under essentially static conditions. In the context of the present invention, “essentially static conditions” shall mean that no substantial (forced or induced) flow of medium occurs in the device or at least that no (forced or induced) flow, such as lateral flow, of medium takes place in or through the central culture channel. This is done in order to achieve a seeding of the cells in the central culture channel.

[0057] In the next step of method for producing a placental-derived human 3D vascular microtissue model of the present invention, c) a lateral flow is established, for example an intermittent interstitial flow (IF) or luminal flow, from the at least one first media channel to the at least one second media channel across the central culture channel. This is achieved by applying a suitable pressure gradient of, for example, between about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of between about 0.20 pm / s and 1.25 pm / s across the central culture channel, wherein the flow is established between about day 2 and about day 7 of the tri-culture. Preferred is a method according to the present invention, wherein the flow is re-stablished daily by replenishing the reservoir media volumes during 5 days following step b). Nevertheless, other intervals may be used, if desired, e.g. twice daily or every two days.

[0058] As the next step, for the barrier-versions of the devices, in the first embodiment a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel is established.

[0059] In a second embodiment of the barrier-version, a barrier layer of cells, preferably of trophoblasts, is established, by dissolving the dissolvable gel barrier of the device, and seeding of the cells as described herein. Other methods to achieve this are known to the person of skill or described in the art.

[0060] Preferred is the method according to the present invention, wherein the hydrostatic pressure gradient (AP) as established in step c) is at less than about 10 mm H2O, preferably at between about: 2 and 7 mm H2O. Also preferred is the method according to the present invention, wherein the pressure gradient is at 3 mm H2O or ~30 Pa or a mean lateral flow velocity of 0.30 ± 0.12 pm / s on day 2. In order to follow-up development of vasculature in the 3D model and / or to detect changes in the phenotype(s) of the cells as involved in the model, another aspect of the method according to the present invention, further comprises the step of detecting changes in the expression of angiogenic and inflammatory cytokines and chemokines during or after step c) and / or d). Respective tests are described herein and in the literature.

[0061] As a next step, d), optionally the perfusion rate can be tested. Methods for such tests are described herein (see examples), described in (2), and in the literature. The test can also be done during the tri-culture, in order to follow-up development of vasculature in the 3D model.

[0062] As a next step, d), optionally the permeability of substances can be tested in the barrier-variants of the present invention. Methods for such tests and others are described herein (see examples), and in the literature. The test can also be done during the development of the barrier cell layer, in order to follow-up development of vasculature in the 3D model.

[0063] Another important aspect of the present invention then relates to a method for producing a breast-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to the present invention as above, b) seeding human mammary vascular endothelial cells (HMVEC) or iPSC-derived endothelial cells and human mammary fibroblasts (HMF) in the elongated central culture channel for about 8h in a suitable culture medium under essentially static conditions, c) introducing an epithelial cell (MCF10) monolayer into the third media channel, optionally followed by the insertion of tumor spheroids or patient-derived organoids on day 1 after seeding, d) establishing a lateral flow, for example an intermittent interstitial flow (IF) or luminal flow, from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of 3mm H2O 24 hours post-seeding, and e) optionally, testing the perfusion rate following step d). Except for the cells, this model is generally established like the other models, and also tested and used as described herein.

[0064] Yet another important aspect of the present invention then relates to a placental-derived human 3D vascular microtissue model or a breast-derived human 3D vascular microtissue model, produced according to a method according to the present invention. As mentioned above, these models comprise and are based on a preferred design to study the barrier function of cells, in particular epithelial and endothelial cells.

[0065] In another embodiment, the model comprises a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP). In contrast to the state of the art, placental fibroblasts are here included in a completely placental- derived human 3D vascular model, which may be used to study the impact of interstitial flow or luminal flow on vascular and extravascular remodeling. The use of the placental fibroblasts advantageously provides a model that is as close to the situation in vivo as possible.

[0066] Preferred is the placental -derived human 3D vascular microtissue model or the breast-derived human 3D vascular microtissue model according to the present invention, exhibiting a laminar flow (see also above), in particular a vascular flow, across the model.

[0067] In some embodiments, the model further comprises a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel.

[0068] Advantageously the microfluidic device according to the present invention and / or the placental- derived human 3D vascular microtissue model and / or the breast-derived human 3D vascular microtissue model according to the present invention enable a direct measurement of physical and chemical cues, such as mechanical stimuli, such as shear stress and corresponding cytokines and chemokines, as disclosed for example herein below.

[0069] Yet another important aspect of the present invention then relates to the use of the microfluidic device according to the present invention or the placental-derived human 3D vascular microtissue model according to the present invention to study barrier functions, such as the placental barrier function(s) as disclosed herein, for drug screenings, for assessing novel drug targets against disfunction, for example in pre-eclampsia, or the breast-derived human 3D vascular microtissue model according to the present invention to study breast duct function(s).

[0070] Importantly, endothelial barrier function significantly improved in flow-conditioned vessels (Fig. 2E), compared to static cultures. The inventor’s results demonstrate the necessity for flow and a magnitude-dependence to ensure development of perfusable fetoplacental-like vessels with improved barrier function. Vascular abnormalities were found in fetal growth restriction- affected placentae (45), which are known to be characterized by high vascular resistance and fetoplacental hypoperfusion.

[0071] Yet another important aspect of the present invention then relates to the use of the microfluidic device according to the present invention or the placental-derived human 3D vascular microtissue model according to the present invention to study trophoblast cell invasion, such as, for example, for toxicological screening of substances and stresses that interfere with trophoblast cell invasion. Recently, Park et al. (in: A microphysiological model of human trophoblast invasion during implantation. Nat Commun 13, 1252 (2022). https: / / doi.org / 10.1038 / s41467-022-28663-4) described a microenigneered system to model a complex sequence of orchestrated multicellular events that plays an essential role in early pregnancy. Their implantation-on-a-chip was capable of reconstructing the three-dimensional structural organization of the maternal-fetal interface to model the invasion of specialized fetal extravillous trophoblasts into the maternal uterus. Using primary human cells isolated from clinical specimens, they demonstrated in vivo-like directional migration of extravillous trophoblasts towards a microengineered maternal vessel and their interactions with the endothelium necessary for vascular remodeling. Through parametric variation of the cellular microenvironment and proteomic analysis of microengineered tissues, they showed the important role of decidualized stromal cells as a regulator of extravillous trophoblast migration.

[0072] Yet another important aspect of the present invention then relates to the use of the microfluidic device according to the present invention or the placental-derived human 3D vascular microtissue model according to the present invention or the breast-derived human 3D vascular microtissue model according to the present invention to study placental or breast branching angiogenesis, breast cancer and related angiogenesis, the impact of nutrients, hormones, and environmental factors on placental development, gestational disorders of fetal vasculature, and / or vascular-associated gestational disorders. Respective test and assays are described herein, and are known in the art.

[0073] Yet another important aspect of the present invention then relates to the use of the microfluidic device according to the present invention or the placental-derived human 3D vascular microtissue model according to the present invention or the breast-derived human 3D vascular microtissue model according to the present invention to study epithelial and / or endothelial barrier function (permeability to solutes), drug screenings and testing of novel drug targets, transport of solutes, pollutants, antibodies, immune cells and drugs across an epithelial cell barrier and / or a vascular cell barrier and extravascular matrix properties (diffusivity, stiffness, and matrix proteins), and in particular the effects of interstitial and / or lateral and / or luminal flow on the study of epithelial and / or endothelial barrier function (permeability to solutes) and extravascular matrix properties (diffusivity, stiffness, and matrix proteins). Respective assays are known in the art, and the relevant literature.

[0074] Tun, et al. (in: Differences in placental capillary shear stress in fetal growth restriction may affect endothelial cell function and vascular network formation. Sci Rep 9, 9876 (2019). https: / / doi.org / 10.1038 / s41598-019-46151-6) describe that fetal growth restriction (FGR) placentae frequently exhibit poor vascular branching. They hypothesized that vascular structural malformation at the organ level alters microvascular shear stress, impairing angiogenesis. Endothelial cells cultured under predicted FGR shear stresses migrated significantly slower and with greater persistence than in shear stresses predicted in normal placentae. These cell behaviors suggest a dominance of vessel elongation over branching. Taken together, these results suggest (1) poor vascular development increases vessel shear stress, (2) increased shear stress induces cell behaviors that impair capillary branching angiogenesis, and (3) impaired branching angiogenesis continues to drive elevated shear stress, jeopardizing further vascular formation.

[0075] In the context of the present invention, the term “about” shall mean a deviation of + / - 10% from a given value, unless indicated otherwise.

[0076] Limited access to human placentae, particularly at early stages of pregnancy, impedes the elucidation of mechanisms associated with fetoplacental vascular development. To circumvent this challenge, the present invention in one embodiment thereof employs a 3D placental- specific microvascular tissue on-chip. Building on the inventors’ previous model (25), the inventors generated an all-placental vascular microtissue and incorporated a flow reservoir, allowing them to characterize vasculogenesis and extravascular remodeling processes in response to flow conditions. Now incorporating placental fibroblasts, as opposed to co-culture with lung fibroblasts as done previously (25), the inventive model required adjustment, including the endothelial to stromal cell ratio (lowered from 5:1 to 10: 1), fibrin gel concentration (from 3 to 3.5mg / mL) and introduction of interstitial flow in order to achieve perfusable vessels. In this tri-culture placental fibroblasts associated with microvessels, as do pericytes, as also shown in the previous model (25). This shows that stromal cells have a clear role in contributing to fetoplacental vascular growth and remodeling.

[0077] Besides the stroma, little is known about placental hemodynamics in early pregnancy, particularly in the villous vasculature. Ultrasound imaging has shown that blood flow velocity waveforms of the umbilical artery and its branches change with advancing gestation and correlate with the development of the placental villous trees and capillary networks (39). Despite the presence of a complete vascular network within the villi, it is believed that the chorionic circulation is not fully established until the end of the first trimester. This flow is progressively established in the third month of gestation, together with the perfusion of the maternal blood into the intervillous space (40). The precise mechanisms behind the formation and remodeling of the villi vascular network remain unknown. Nevertheless, the inventors hypothesized that hemodynamic forces and associated signaling could play a role in directing vasculogenesis. Previous work has shown that interstitial flow promotes early vessel connectivity in 3D vasculature on-a-chip (16). Herein, static, low and high interstitial flow velocities ranging from ~0.1-1.2 pm / s (day 2 measurements) were applied via reservoirs. To the best of the inventor’s knowledge, there are no reports on human fetoplacental interstitial fluid velocity; however, the inventor’ s measurements fall within the physiological range found in most soft tissues (from 0.1 to 4.0 pm / s) (41, 42). For these placental-like vessels, interstitial flow applied at early timepoints is necessary to establish fully perfusable vascular networks, which transition to luminal flow by day 5 after seeding (Fig. IE). Flow-conditioning has a clear impact on vessel morphology, resulting in earlier connectivity, reduced branch density and larger vessel diameters (Fig. 2). Importantly, endothelial barrier function significantly improved in flow-conditioned vessels (Fig. 2E), compared to static cultures. Other studies conducted in perfused placental models have also demonstrated that shear stress (20 dyn cm2or 2 Pa) promotes vasodilatation (nitric oxide release) and decreases vascular resistance (43, 44). The inventor’s results demonstrate the necessity for flow and a magnitude-dependence to ensure development of perfusable fetoplacental-like vessels with improved barrier function. In line with this, vascular abnormalities were found in fetal growth restriction-affected placentae (45), which are known to be characterized by high vascular resistance and fetoplacental hypoperfusion. The inventor’s findings may implicate the need for flow early in fetal vessel development to prevent gestational complications. Mechanical forces imparted by blood flow induce changes in vascular function and remodeling, by promoting inflammatory and angiogenic responses (46, 47); however, insight into flow- induced signaling in the placenta remains limited. Pro-inflammatory chemokines, including IL- 8 and MCP-1 are associated with angiogenesis, as evidenced by various studies (48, 49). Sustained levels of these chemokines are produced by the placenta to stimulate the immune response against potential infections during pregnancy (50, 51), but it remains unclear what role, if any, these chemokines play in the development of placental vasculature. The inventor’s findings indicate that the expression of IL-8 and MCP-1 during vessel formation in the inventor’s system is strongly influenced by flow, confirming previous reports of their regulation by shear stress (52-54). The human placenta has also been found to be a significant source of locally produced angiogenic factors which include members of the VEGF family, FGF family, and angiopoietins, among others (55). The inventors analyzed Ang-2 levels in the inventor’s fetoplacental vessels, which is highly expressed in early gestation and promotes vascular remodeling in the presence of VEGF (56, 57). The inventor’s findings indicate increased Ang- 2 expression due to flow at early timepoints (Fig. 3D), which contrasts previous research showing a decrease in shear-stress dependent expression of Ang-2 in experiments with flow exposed HUVEC monolayers (58, 59). It is worth noting that previous studies investigating Ang-2 have primarily focused on laminar flow conditions (resulting in WSS>6 dyn cm-2or 0.6 Pa), whereas other studies have demonstrated that Ang-2 expression is upregulated at lower shear stress levels (1 dyn cm'2) (60). Notably, the expression of Ang-2 in fibroblasts has been previously found to elicit vascular growth, while concurrently maintaining a balance in the quiescent action exerted by Angiopoietin-1 (Ang-1), as produced by placental pericytes (25). These observations support the inventor’s findings and highlight the need for further research to fully understand the complex relationship between shear stress and Ang-2 expression. In particular, VEGF and its receptors are expressed in trophoblasts and villous fetal vessels at early developmental stages, suggesting their involvement in the initiation and progression of vasculogenesis (61). Herein, the inventors observed a noteworthy elevation of VEGF levels at later stages (day 7) within static cultures, proposed as a compensatory mechanism to counteract the absence of flow. In line with this hypothesis, VEGF expression was found elevated in fetal growth restriction placentas, suggesting that a decrease in fetal-maternal blood circulation during placentation enhances the expression of the angiogenic factor (62).

[0078] Vessel stabilization is driven by hemodynamic forces through regression and pruning of branches exposed to low blood flow, along with the maintenance of vessel connections experiencing a threshold level of flow (63). After two weeks of culture under intermittent flow conditions, placental-like microvessels were perfusable and functional (maintained a relatively low permeability) (Fig. 4), consistent with observations in a brain 3D microvascular system (23). Although vessel branching did not differ, flow significantly preserved vessel diameter and prevented vascular narrowing, as opposed to the vessel constriction as observed in static cultures (Fig. 4H), which decreased from 74.5% at day 7 to -59% area coverage at day 14). Previous observations in HUVEC and lung co-cultures have also shown that continuous flow (at a low WSS of < 1 Pa) maintains a stable vascular diameter in formed vessels

[0016] , At this late stage of culture, very few static-cultured vessels were perfusable, and interestingly those few vessels maintained a high barrier function (Fig. 4F). However, flow-conditioned vessels exhibited a slight decrease of endothelial barrier capacity with respect to day 7 (Fig. 4F, layered data plot in orange) that the inventors attribute to the limited stimuli experienced by the vasculature upon reestablishment of the hydrostatic pressure (every other day) from day 7 onward. In addition, flow-conditioning was essential in maintaining microvessels connectivity over 4 weeks in culture, highlighting the importance of implementing flow for long-term in vitro microvessel studies.

[0079] Luminal flow in the vessels was measured by particle tracking and demonstrated a significant increase in the mean velocity in flow-conditioned vessels by day 7. To obtain a more accurate prediction, vascular network flow was modeled in COMSOL (Fig. 5). Heterogeneous flow distributions were apparent in the networks, which depend on the formation of open lumens in the media channels. Shear stress for flow-conditioned vessels was predicted to be 0.32 Pa (3.2 dyne / cm2), whereas shear stress within the EVM was O.OOlPa (0.01 dyne / cm2). CFD results deviated from experimental bead measurements by approximately one order of magnitude. Note that velocities and shear stress predictions are obtained using 2D segmented projections and do not account for the exact spatial complexity of 3D vessels, nor the wall effects or size-dependent effects of beads. It was reported previously that the presence of common additives in the media (64), tracer beads size, and their properties (65), such as deformability, contribute to elastic noslip boundaries at the fluid interfaces influence the velocities.

[0080] The inventors postulated that differences in vascular morphogenesis and long-term stability, observed in the presence of flow, could be due to extravascular rearrangements and changes in the ECM, as previously reported (66). In line with this hypothesis, recent findings showed that flow-conditioning reduces proteolytic activity of cathepsins prolonging in vitro microvessel stability (67). Flow-conditioned vessels, as opposed to the static ones, exhibited significantly reduced extravascular diffusivity and overall increased matrix / tissue rigidity (Fig. 6). Despite evidence showing that presence of stromal cells can contribute to tissue stiffening (28), the flow-dependent increase in stiffness is not attributable to alterations in stromal cell population, since they remain stable (close to 30% of the total cell population) over time (Fig. 6E). Examination of proteins in the EVM, by immunostaining, revealed that flow promotes increased deposition of collagen I, fibronectin, and laminin (Fig.7A-B). Several studies have reported the effect of shear stress in regulating matrix deposition and remodeling (68-70), but none in the context of placental tissue. Collagens, fibronectin, and laminins are abundant in the placental stroma and basement membrane, and their expression increases with advancing gestation to support the developing tissue structure and function (71). Low expression of these ECM proteins, observed in the absence of flow, could explain their increased destabilization over time. Supporting this, a recent study reported that endothelial cells plated on matrices derived from villous stromal fibroblasts where fetal growth restriction was clinically observed (with reduced expression of collagen I and fibronectin respect to control) exhibited impaired proliferation and migration, suggesting that matrix composition is crucial for fetoplacental vascular development (72). A more in-depth examination of the gel / tissue composition utilizing mass spectrometry, revealed an abundance of EVM-related proteins deposited by the cells over the culture period (Fig. 7C). As expected, tissue protein enrichment is influenced by the presence of the flow, revealing changes in protein expression associated with EVM organization and vascular homeostasis. Consistent with the immunofluorescence findings, MS analysis revealed enrichment in laminin in response to flow, indicating flow-induced adaptation and alterations in cell-matrix adhesion, which ultimately resulted in matrix remodeling. In addition, multimerin 2, an EVM molecule that enhances vascular stability and regulates permeability by stabilizing endothelial junctions [73, 74], was found enriched in flow- conditioned tissues suggesting a protective function in maintaining vessel function and stability under mechanical shear stress. The substantial increase of prothrombin, which serves as a precursor to thrombin and facilitates fibrin formation

[0075] , provides further evidence of flow- induced enhancements in EVM rigidity and subsequent matrix stabilization. Flow conditions also resulted in the depletion of nicotinamide phosphoribosyltransferase (NAMPT), an essential coenzyme that plays a critical role in energy production, DNA repair, and signaling pathways. The overexpression of NAMPT has been associated with inflammatory processes and the development of various human conditions, including acute lung injury, atherosclerosis, and cancer

[0076] , While direct evidence of the specific effects of flow on NAMPT expression is lacking, our findings suggest that the protein is mechanosensitive, indicating a potential protective role in response to flow mechanical forces. Moreover, flow has an impact on actin dynamics as indicated by the reduced levels of profilin and fascin in flow-conditioned tissue. Based on previous reports [77-79], we propose that the depletion of these actin-bundling proteins can alter the organization of the cytoskeleton and hinder cell migration. Consequently, once the perfusable vascularized tissue is formed, the absence or reduction of cellular movement plays a role in ensuring the functionality and stability of the fetoplacental vascular barrier.

[0081] The inventors examined flow on the formation of fetoplacental microvessels - enabled by the inventor’s 3D model. The inventor’s triculture system is more physiological than previous methods; however, the applied intermittent flow only partially reproduces the hemodynamic forces to which the in vivo villous capillary networks are subjected (19, 80). Moreover, the inventor’s model lacks the branched villous structure and critical epithelial layer (trophoblasts) that forms a barrier between the maternal blood and fetal capillaries. Trophoblasts are mechanosensitive to shear stress and release angiogenic factors that are involved in the regulation of placental vascular formation and morphogenesis (81). The role of trophoblasts in the process of maternal vascular remodeling has been the subject of extensive investigation, yet the implications of these cells in the development of fetal villous blood vessels remain largely unexplored. Integration of trophoblasts into the inventor’s system resulted in the formation of a fetal-maternal interface.

[0082] Failure in placenta vascularization is closely linked to pregnancy pathologies, but the inventor’s understanding of these disorders is limited by the lack of suitable models. Animal models and human explants both have limitations and recent in vitro models lack functional vasculature. To address this issue, the inventors present a 3D on-chip model of the human placenta terminal villi, which includes the fetal mesenchyme and the vascular endothelium. Fully perfusable fetal microvessels are generated by co-culturing HUVEC, placental fibroblasts and pericytes, in a microfluidic chip with attached flow reservoir. Through regulation of pressure gradients across the microtissue, flow is shown to play a crucial role in the growth and remodeling of fetoplacental microvessels. Flow-conditioning results in the early formation of interconnected placental vascular networks and maintains their viability for prolonged periods in culture (> 2 weeks). Increasing flow results in enlargement of vessel diameters concomitant with increasing shear stress acting within vessels (~1.5 Pa), leading to enhanced barrier function. Moreover, shear forces measured within the interstitial space (~0.3 Pa), as predicted by computational fluid dynamics simulations, result in increased stiffness, associated protein deposition, and decreased diffusivity, all of which contributes to significant placental vascular remodeling. Altogether, the inventive model provides a means to infer complex in vivo parameters including shear stress on developing vascularized placental tissue and holds promise for providing insight into vascular-associated gestational disorders. Furthermore, fluid dynamics play a crucial role in regulating the development and remodeling of fetal vessels, suggesting that poor or restrictive flow conditions can negatively impact fetoplacental vasculogenesis. The inventor’s findings serve as a basis for further research into the mechanisms of placental vascular defects in pregnancy-related disorders related to flow insufficiency.

[0083] The inventor’s in vitro placental vascular model is crucial for understanding the mechanisms behind fetoplacental vasculogenesis and regulation. It allows for mimicking a part of the complex in vivo tissue in a controlled in vitro setting, enabling direct measurement of physical and chemical cues, such as shear stress and corresponding cytokines and chemokines. This model holds promise for studying pregnancy-associated diseases and can provide insight into gestational disorders of fetal vasculature, and vascular-associated gestational disorders. In the more general approach of the barrier function version comprising a dissolvable gel, also cellular barrier functions involved in other conditions or diseases, such as cancer and the like may be tested using the model

[0084] The present invention provides models of the maternal-fetal interface and breast duct-on-chip models both in healthy and disease states as unique devices, and provides methods for drug screens / targeted therapies and or tox / safety testings.

[0085] The present invention relates to the following items:

[0086] Item 1. A microfluidic device, comprising a) an elongated central culture channel for holding mammalian cells in a suitable gel, b) at least one first media channel extending at least in part thereof arranged in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel, configured to allow a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, and c) at least one third media channel extending at least in part thereof in parallel to a third side of the elongated central culture channel, wherein optionally the at least one third media channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one third media channel.

[0087] This enables containment of the cell and gel mixture in the central channel, and once dissolved allows for an adjacent cell monolayer to be formed.

[0088] Preferably, the microfluidic devices according to the present invention advantageously are a post-less system, which renders them less complicated to make and allows for direct contact between media channel and gel, i.e. does not provide a physical barrier.

[0089] Item 2. The microfluidic device according to Item 1, wherein said device comprises a fourth channel extending at least in part thereof in parallel to a fourth side of the elongated central culture channel, wherein optionally the fourth channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one third media channel.

[0090] Item 3. The microfluidic device according to Item 1 or 2, wherein the at least one gel barrier comprises a temperature-dependent dissolvable gel.

[0091] Item 4. The microfluidic device according to any one of Items 1 to 3, wherein the mammalian cells in the culture chamber are selected from human cells, and preferably comprise cells and / or cell lines, such as cancer and / or immortalized cell lines, derived from epithelial tissues having an epithelial layer, vascular cells, cells of the respiratory system, cells of the digestive system, bladder cells, mammary cells, ductal cells and cancer cells thereof, and preferably comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP), and a layer of human trophoblasts, located on the side of the cell culture that is in contact with the at least one third or fourth media channel.

[0092] Item 5. A microfluidic device, comprising a) an elongated central culture channel for holding human cells in a suitable gel, b) at least one first media channel extending at least in part thereof in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel, configured to allow a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, wherein the mammalian cells in the gel comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP).

[0093] Item 6. The microfluidic device according to Item 4 or 5, wherein the HPF and HPP are cultured at a ratio of about 10:1 of endothelial: stromal cells HPF / HPP, preferably wherein the HPF and HPP are at a ratio of 1 : 1.

[0094] Item 7. The microfluidic device according to any one of Items 1 to 6, wherein the gel of the culture channel comprises a hydrogel, such as a fibrinogen / thrombin (fibrin) gel.

[0095] Item 8. The microfluidic device according to any one of Items 1 to 7, further comprising added media reservoirs, preferably 3D-printed tailored media reservoirs, connected to at least one of the at least one media channel.

[0096] Item 9. The microfluidic device according to any one of Items 1 to 8, wherein the device furthermore comprises a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel.

[0097] Item 10. The microfluidic device according to any one of Items 1 to 9, wherein the device is positioned on a solid support, such as a biochip or coverslip, and / or wherein several or a multitude of devices are grouped on said solid support.

[0098] Item 11. A method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to any one of claims 1 to 4 and 8 to 10, b) incubating the cells in the culture chamber in a suitable culture medium, and c) dissolving of at least one of the gel barrier(s) as described herein and seeding suitable mammalian barrier cells. Item 12. A method for producing a placental-derived human 3D vascular microtissue model according to Item 11, wherein in b) the tri-culture is incubated for about 48h in a suitable culture medium under essentially static conditions, bl) a lateral flow is established, for example an intermittent interstitial flow (IF) or luminal flow, from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of between preferably about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of preferably between about 0.20 pm / s and 1.25 Lini / s across the central culture channel, and wherein the flow is established between about day 2 and about day 7 of the tri-culture.

[0099] Item 13. A method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to any one of items 5 to 10, b) incubating the tri-culture for about 48h in a suitable culture medium under essentially static conditions, c) establishing a lateral flow, for example an intermittent interstitial flow (IF) or luminal flow, from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of preferably between about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of preferably between about 0.20 pm / s and 1.25 pm / s across the central culture channel, wherein the flow is established between about day 2 and about day 7 of the tri-culture, and d) optionally, testing the perfusion rate following step c).

[0100] Item 14. The method according to Item 12 or 13, wherein the hydrostatic pressure gradient (AP) as established in step c) is at less than about 10 mm H2O, preferably at between about: 2 and 7 mm H2O.

[0101] Item 15. The method according to any one of Items 12 to 14, wherein the pressure gradient is at 3 mm H2O or ~30 Pa or a mean lateral flow velocity of 0.30 ± 0.12 pm / s on day 2. Item 16. The method according to any one of Items 12 to 15, wherein the flow is re-stablished daily by replenishing the reservoir media volumes during 5 days following step b).

[0102] Item 17. The method according to any one of Items 12 to 16, further comprising the step of introducing a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel.

[0103] Item 18. The method according to any one of Items 12 to 17, further comprising the step of detecting changes in the expression and / or amount of cellular analytes, such as, for example, angiogenic and inflammatory cytokines during step c) and / or d).

[0104] Item 19. A placental-derived human 3D vascular microtissue model, produced according to a method according to any one of Items 11 to 18.

[0105] Item 20. The placental -derived human 3D vascular microtissue model according to Item 19, exhibiting a laminar flow across the model.

[0106] Item 21. The placental -derived human 3D vascular microtissue model according to Item 19 or 20, exhibiting an epithelial barrier function.

[0107] Item 22. Use of the microfluidic device according to any one of Items 1 to 5 and 8 to 10 or the placental-derived human 3D vascular microtissue model according to Item 21 to study placental barrier function(s).

[0108] Item 23. Use of the microfluidic device according to any one of Items 1 to 10 or the placental- derived human 3D vascular microtissue model according to any one of Items 19 to 21 to study trophoblast cell invasion, such as, for example, for toxicological screening of substances and / or stresses that interfere with trophoblast cell invasion.

[0109] Item 24. Use of the microfluidic device according to any one of Items 1 to 10 or the placental- derived human 3D vascular microtissue model according to any one of Items 19 to 21 to study placental branching angiogenesis, the impact of nutrients, hormones, and environmental factors on placental development, gestational disorders of fetal vasculature, and / or vascular-associated gestational disorders. Item 25. Use of the microfluidic device according to any one of Items 1 to 10 or the placental- derived human 3D vascular microtissue model according to any one of Items 19 to 21 to study epithelial and / or endothelial barrier function (permeability to solutes), for drug screenings, for assessing novel drug targets against disfunction, for example in pre-eclamp si a, transport of solutes, pollutants, antibodies, immune cells and drugs across an epithelial cell barrier and / or a vascular cell barrier and extravascular matrix properties (diffusivity, stiffness, and matrix proteins), and in particular the effects of interstitial and / or lateral flow and / or luminal flow on the study of epithelial and / or endothelial barrier function (permeability to solutes) and extravascular matrix properties (diffusivity, stiffness, and matrix proteins).

[0110] The invention will now be described further in the following examples with reference to the accompanying figures, nevertheless, without being limited thereto. For the purposes of the present invention, all references as cited are incorporated by reference in their entireties.

[0111] Figure 1 shows that interstitial flow promotes network perfusion in placental microvessels. A) Schematic demonstrating the culture timeline and protocol. B) Immunofluorescence image of tri-culture placental vessels with orthogonal view showing open vascular lumen. C) Interstitial flow conditions (depicted graphically) promote complete vascular bed perfusion. Perfusion is shown by FITC labelled dextran introduced into the RFP-labelled vessels at day 7. Scale bar is 1000 pm. D) Percentage of perfused vessels is quantified across static and interstitial flow conditioned vessels at day 7. Shown are box plots with outer edge as standard error and bars with standard deviation. Significance is measured by One-way ANOVA and indicated *** < 0.001 for Tukey means comparison test. E) Pressure drop measured over time across different days in culture. Shown is mean and standard deviation of 3 separate experiments with >6 devices.

[0112] Figure 2 shows that interstitial flow promotes vessel connectivity and barrier function. A) Confocal images of microvessels (HUVEC with cytoplasmic RFP) at day 3 and 7 of culture under static and flow conditions. Morphology is measured by segmentation of images of RFP- cytoplasmic labeled HUVEC. Morphologic parameters were measured across days 3 and 7 for B-D. B) Connectivity is quantified as a ratio of ratio of vessel junctions to endpoints. C) Branch density is the number of branches per area. D) Effective diameter is measured as vascular area coverage over length. E) Permeability to 70kDa dextran is shown for static and flow conditions at day 7. Hashed lines for the static condition indicate that very few vessels were perfused and thus measured. Significance is measured by One-way ANOVA and indicated by *<0.05, **<0.01, and *** < 0.001 for Tukey means comparison test and # represents significance across days.

[0113] Figure 3 shows that interstitial flow impacts inflammatory and angiogenic signaling at early timepoints in vessel development. Concentrations of inflammatory molecules A) IL8 and B) MCP1 were measured by ELISA across various time points from 3D microvessels. Angiogenic signaling molecules C) Ang2 and D) VEGF were significantly different from statically cultured vessels at early and later time points, respectively. Significance is measured by One-way ANOVA and indicated by *<0.05, **<0.01, and *** < 0.001 for Tukey means comparison test; # represents significance across day 5 and day 3 and $ across day 7 and day 3.

[0114] Figure 4 shows that flow promotes long-term stability of placental microvessels. A) Confocal images demonstrate differences in vascular density between static and flow-conditioned vessels at day 14 in culture. Both vessel B) area coverage and C) effective diameter are significantly increased by flow-conditioning. D) Branch density and E) length were not significantly impacted by flow. F) Vascular permeability is measured for fully perfused flow-conditioned vessels and the few perfusable vessels that remain in the static condition. Flow sustains vessel perfusion capacity after 14 days in culture (G) and higher area coverage H) overtime than in static condition. Significance is measured by One-way ANOVA and indicated by *<0.05 and **<0.01 for Tukey means comparison test.

[0115] Figure 5 shows CFD simulation predictions of fetoplacental microvascular tissue. A) Pipeline for computational predictions of velocities and shear stresses within microvessels embedded in porous gel. Shown are 4 independent predictions from 7 mm H2O condition for B) velocity and C) shear stress distributions within microvessels and gel regions.

[0116] Figure 6 shows that interstitial flow alters extravascular remodeling and tissue-level stiffness. A) FRAP measurements performed on extravascular regions. ROI is bleached region of interest, EVS and IVS are extravascular and intravascular space respectively. B) Diffusivity measurements calculated from FRAP experiments for static and flow-conditioned microvessels at day 7. C) Schematic representation of nanoindentation on microtissues. D) Stiffness measurements performed across static and flow-conditioned samples. E) Populations of stromal cells from microvessels corresponding to different culture times. Significance is measured by One-way ANOVA and indicated by *<0.05, **<0.01, and *** < 0.001 for Tukey means comparison test.

[0117] Figure 7 shows that flow induces changes in EVM protein deposition and composition. A) Confocal images at day 7 of static and flow-conditioned microvessels (red-labeling shown in the comer image insets) immunolabeled with collagen I, laminin and fibronectin (greenlabeling). B) Fluorescence intensity of collagen I, fibronectin and laminin was measured and normalized to vessel area. Significance is measured by One-way ANOVA and indicated by *<0.05 for Tukey means comparison test. C) EVM protein composition based on the total peptide count after 7 days of culture, and ranked in a decreasing order of abundance (50 most abundant proteins). Proteins are categorized according to 5 ontology terms selected for their relevance to vascularization and matrix remodeling. D) PCA performed on mass spectrometrybased proteomics normalized data obtained from flow-conditioned samples for seven days. E) K-means clustering and heatmap of the 13 hit and candidate proteins, showing differential expression across the static and flow conditions (shown as the Log.2 ratio of protein abundance in each 3 and 7mm H2O sample relative to the average protein abundance in 0mm H2O samples). F) Volcano plots of the p values vs. the log2 protein abundance differences between flow and static conditions. Red dots, hits: FDR <0.05, FC >30% Blue dots, candidates: FDR <0.05, no FC threshold, p values are calculated from moderated t-test (Limma).

[0118] Figure 8 shows an example of the human fetal-like vasculature on a microfluidic chip (in accordance with the present invention), including the “vasculature barrier” function. Human umbilical vein endothelial cells (HUVEC) were co-cultured in the presence of stromal cells (fibroblasts and pericytes), which formed interconnected microvessels capable of perfusion in 5-7 days. This was followed by integration of trophoblasts cells into the system, generating the critical epithelial layer that forms a barrier between the maternal blood and fetal capillaries. In this embodiment, the trophoblast monolayer was seeded adjacent to the fetal vessels first in the chip design established for generation of the fetal vasculature using two immortalized lines, BeWo and HTR-8 / SVneo (Fig. 13).

[0119] Figure 9 shows that in another embodiment of the device according to the present invention with a temperature-dissolvable gel barrier (bottom medium channel) and comprising trophoblasts, the permeability of trophoblasts is limited. A) Schematic overview over model, demonstrating fetal and maternal components in this configuration. B) cellular structure at gel barrier (left) and DAPI staining (right). C) shows the selective permeability of the model as shown in Figure B using Dextran particles of different sizes. Molecules larger than 1 kDa do not readily cross the barrier (as expected in vivo).

[0120] Figure 10 shows that perfusion of the model in Figure 9 with sFLT-1 results in a pre-eclamptic (PE) phenotype of the trophoblast barrier. The dysfunction of the trophoblast barrier with increasing concentrations of sFLT-1 enables transport of dextran from the maternal component to the fetal vasculature.

[0121] Figure 11 shows that perfusion of the model in Figure 9 with sFLT-1 results in a pre-eclamptic (PE) phenotype of the microvessels. A) Microvessels regress with increasing concentrations of sFLT-1 and show significant loss of normal morphology B) at concentrations relevant to plasma levels of patients with PE.

[0122] Figure 12 shows a 3 -channel post-less chip design according to the invention in a CAD drawing demonstrating major features of the 3 channel chip.

[0123] Figure 13 shows an overview of the methods for producing the complete 2-barrier model of the maternal-fetal interface according to the invention. A) View of the fabricated microfluidic 3- channel device. The size of the device is compared to a 1 Euro coin. B) Experimental cell-based approach to generate maternal-fetal interface-on-chip and image showing the interface between microvessels and trophoblast layer. C) The photos indicate that the placental barrier on-chip is permeable to a fluorescent analog of glucose (2-NDBG). D) The placental barrier on-chip shows reduced permeability for dextran sizes larger than 0.4 kDa. E) The placental barrier on- chip shows low a permeability to immunoglobulins (G and M) and insulin after 24 hours from the perfusion. Dashed lines indicate corresponding maternal concentrations. F) Almost pure population of primary cytotrophoblasts (cytokeratin 7 positive) was obtained by dissecting the villi from term placenta explants. G) Isolated primary cytotrophoblasts spontaneously fuse and differentiate into syncytiotrophoblasts after 72 hours in culture. H) Image of maternal-fetal interface generated with the integration of isolated primary cytotrophoblasts (G).

[0124] Figure 14 shows a trophoblast barrier function assay according to the invention. A) Schematic showing timeline and assay (below) to measure trophoblast barrier permeability. Dextran was added to the maternal component of the chip. Permeability of the trophoblast barrier to 0.4 kDa dextran was monitored by time-lapse imaging. Scale bar 200 microns. B) Permeability measurements to 0.4kDa dextran demonstrated a significant increase (barrier dysfunction) when cultured under hypoxic conditions.

[0125] Figure 15 shows the feto-placental microvessel-morphology and -barrier assays. Since the fetoplacental microvessels can act as an indicator of developmental health, according to the invention a method was established to measure changes in the vascular morphology and barrier function. A) Schematic of experimental timeline for analyzing microvessel morphology and barrier function. The lower diagram demonstrates where molecules (sFLTl and sEng) were perfused in the maternal compartment and Dextrans perfused in the fetal compartment for measurement of microvessels barrier function. To mimic pre-eclampsia on-chip, the molecules sEng and sFltl were supplemented to normal culture media, or alternatively the entire chip was cultured under hypoxic conditions. B) Images of fetal microvessels at day 7, shown for untreated, and those treated with PE-related factors sFLTl, sEng, or under 1% O2 conditions. All scalebars are 200 microns. C) Morphologic assessment was carried out by imaging vessels in bright-field and using a machine-learning (ML) pipeline in order to segment vessels, and output morphologic mappings of various features (as in 98). D) One parameter, mean microvessel diameter, is shown, as measured in C). Vessel diameter was significantly reduced under all PE-mimicking conditions. E) The microvessel barrier function was measured as in A) and B) and shown for 0.4 kDa dextran to be altered by sEng. Vessels were not perfusable by hypoxia-conditioning.

[0126] Figure 16 shows an overview of secondary use-case of the 3-channel chip design according to the invention as a duct-on-chip model. A) Image of 3-channel device with reservoir and experimental cell-based approach to generate a duct-on-chip. B) Image of interface between microvessels and epithelial monolayer. C) Image showing fluorescent image of microvessels and E-cadherin (epithelial marker). D) Schematic of experimental timeline for inclusion of hormones, tumors, drugs. E) Inclusion of a tumor in the ductal compartment. F) Barrier of the duct-on-chip epithelial layer is impermeable for Dextran sizes larger than 1 kDa. G) Permeability to various dextran sizes measured in response to no treatment (NT), addition of hormones (Fol = follicular levels of estrogen and progesterone) and hormones + tamoxifen (Fol + TAM). All scalebars are 250 microns, unless indicated otherwise. Table 1 shows quantitative analysis of simulation outcomes in reference to Fig. 5. Surface averaging is applied to estimate the velocities and shear stresses (mean ± SD) within combined vessels + gel (tissue), vessels only and gel only regions.

[0127] Examples

[0128] Materials and methods

[0129] Cell culture

[0130] Human umbilical vein endothelial cells (HUVEC) were purchased from Lonza and cultured in endothelial media (VascuLife, Lifeline cell systems) on 50 pg / ml rat tail collagen I (Roche) coated T75 or T150 flasks. HUVEC were transduced to stably express cytoplasmic RFP (LentiBrite RFP Control Lentiviral Biosensor, Millipore Sigma Aldrich), and were used between passages 6 and 9. Primary human placental fibroblasts (HPF) were purchased from ATCC, cultured in fibroblast media on uncoated T75 flasks (FibroLife, Lifeline cell systems) and used between passages 4 and 8. Unlabeled and GFP-labeled primary human placental microvascular pericytes (HPP) were acquired from Angioproteomie, cultured in pericyte growth medium according to manufacturer's protocols and used between passage 6 and 9. All cells were cultured under normal conditions at 37 °C and 5% CO2 in a standard incubator, dissociations were carried out using TiypLE Express (Gibco), and media was completely refreshed every other day. Device fabrication

[0131] Devices were fabricated as previously described (25). Briefly, PDMS (SYLGARD™ 184 Silicone Elastomer Kit, Dow) was mixed at a 10:1 elastomer to cross-linker ratio according to manufacturer's protocols, degassed, and poured onto a mold. Following further degassing, PDMS was placed in a 60°C oven overnight and single devices were cut, punched using 1- and 2-mm diameter biopsy punches for the gel and media ports, respectively, and air-plasma bonded (Harrick systems) to clean glass slides. Assembled devices were then incubated at 60°C overnight to restore the native hydrophobic state. All devices were sterilized under ultraviolet light for at least 30 min prior to cell seeding.

[0132] Device seeding and microvessels formation under interstitial flow

[0133] Fibrinogen derived from bovine plasma (Sigma) was reconstituted in phosphate-buffered saline (PBS) to a working concentration of 7 mg / ml before use. Thrombin (Sigma) stock solution (100 U / ml in 0.1% w / v bovine serum albumin solution) was diluted to a 4 U / ml working solution in cold VascuLife medium. Endothelial cells and stromal cells were cultured until near confluence prior to detachment and re-suspended in thrombin, separately, to concentrations of 24 million endothelial cells / ml and 2.4 million stromal cells / ml (a combination of fibroblasts and pericytes). Cell suspensions were mixed 1:1 by volume and then combined with fibrinogen solution to produce a final concentration of 6 million endothelial cells / ml and 0.6 million stromal cells / ml, in a 10: 1 ratio within fibrin (3.5 mg / ml). The cell-gel mixture was injected into the device central channel and allowed to polymerize for 15-20 min at 37°C in a humidified chamber. VascuLife media was added to each media channel (150 ul total) and devices were initially cultured under static conditions. An intermittent interstitial flow (IF) across the hydrogel channel was generated 48 hours after seeding by adding 3D-printed tailored media reservoirs (Form3 printer, Formlabs). Equal total volumes of media were differentially added to the reservoirs compartments to generate three hydrostatic pressure gradients (AP): 0, 3 and 7 mmFEO. The flow was re-stablished daily replenishing the reservoir media volumes during 5 days following seeding. From day 8 onwards, with the vessels fully perfusable, hydrostatic pressure gradients were restored every other day.

[0134] Imaging and vessel morphology quantification

[0135] All images were acquired on a Stellaris 8 confocal microscope using LAS X software (Leica).

[0136] Confocal z-stacks were obtained for all time-points and used to quantify the morphology of the microvascular networks cultured under different AP, as previously described (25). Briefly, a custom macro (ImageJ, NIH) was generated to process the images as follows: projection of maximum intensity of the RFP channel in the z-direction, Gaussian filter smoothing, removal of outliers and conversion to binary images. The analyze particles and 2D skeletonize built-in ImageJ plugins were finally performed to determine branch density, network connectivity and effective diameter (vessel area / total length). An alternative label-free method can be used as described in (98).

[0137] Permeability measurements

[0138] Microvessels cultured under different flow conditions were perfused at day 7 or 14 with 70 kDa FITC labelled dextran (Merck) by applying a hydrostatic pressure drop across the central gel channel. Briefly, media was removed and replaced by adding 40 ul of fluorescent perfusate solution (FITC 0.1 mg / ml) into one channel. After perfusion into and through the microvessels, convective flow was stopped by applying an equal volume of dextran solution in the opposite channel. Following stabilization (-=2—3 min), time-lapse confocal images were captured (3 >< 5 min intervals), from which permeability measurements were made, as reported previously (25).

[0139] Cytokine release

[0140] Supernatants were collected from n = 3 device reservoirs per AP condition of two or three independent experiments at days 3, 5 and 7 and were frozen until use. Concentration of VEGF, Ang-2 and MCP-1 were measured from individual samples following manufacturer’s protocols of respective Quantikine ELISA Kits (R&D systems, DVE00, DANG20 and DCP00). Cytokines expression profiles in 2D cultures were assessed using a human angiogenesis antibody array (Abeam). Supernatants from lung fibroblasts, placental fibroblasts and pericytes were collected 48 hours after seeding and processed according to manufacturer's instructions. Cytokine profiles on the array membranes were detected by chemiluminescence using the Fusion FX Spectra (Vilber, France). The relative (semi-quantitative) expression of intensity was normalized to the positive controls.

[0141] Fluid dynamics characterization by bead-tracking

[0142] Evaluation of fluid velocity of static and flow-conditioned vessels was assessed at day 7 through perfusion of 2.0 pm fluorescent beads (Fluorescent blue latex beads, Sigma Aldrich) into the microvascular networks using reservoirs with applied pressure gradients of 5 mm H2O. Timelapse images (1024x1024 pixels at a resolution of 0.64 pm / pixel with a 0.06 sec frame rate) were acquired for 25 seconds using a fluorescent microscope (Thunder Imager - DMi8 microscope, Leica). Tracking of each individual bead (distance over time) was done using the particle tracking plugin - TrackMate (NTH ImageJ) (25). First, individual beads are detected based on size through the “LoG detector” method with an “Estimated blob diameter” of 5 pm, and an intensity threshold of 2 AU. Then, the “simple LAP Tracker” method with “Linking max distance” of 15 pm, “Gap-closing max distance” of 15 pm, and “Gap-closing max frame gap” of 2 frames, is used to identify the same object over time. Upon application of a filter for “Track displacement” to easily remove non-moving objects, the function “Analysis” is used to obtain the beads mean speed.

[0143] Computational fluid dynamics simulation

[0144] In order to estimate the range of fluid velocities and shear stresses generated by the application of pressure-induced flows within microvessels and surrounding gel domains, binary masks were obtained from confocal microscopy images and converted to drawing exchange format (dxf) in Adobe Illustrator 2021. These .dxf files were imported into COMSOL multiphysics software (Version 6.0). Using the laminar flow physics module with porous domain enabled physics setting, the spatial profile of luminal and interstitial fluid velocities and shear stresses were computed for 70 Pa applied pressure (a pressure drop, AP, of 7 mm H2O) between the inlet and outlet of four microfluidic chips containing placental microvessels. A computationally efficient physics-controlled extra fine mesh was applied to the model, resulting in an overall computational time of ~2 hours.

[0145] Diffusivity measurements and initial fluid velocity tracking

[0146] Diffusivity was assessed in the extravascular space (adjacent but outside of vessels observable in images) for each flow condition using FRAP measurements on a Stellaris 8 confocal microscope (Leica). Devices were perfused with 70 kDa FITC labelled dextran and incubated at 37°C for twelve hours to allow total diffusion throughout the gel matrix (extravascular space). Small regions (30 pm 0) within the matrix were bleached and, immediately after photobleaching, time-lapse images were recorded every 0.4 sec to capture the fluorescence recovery. Analysis of diffusivity was performed using the Matlab frap analysis plugin (27), as previously described (28). Similarly, the inventors determine the initial IF velocity (day 2 after seeding), using FRAP as done previously (16). Briefly, pressure gradients were generated across the gel and cell containing devices by adding FITC dextran to the corresponding reservoir media volumes. Following dextran saturation, a region of 30 pm 0 was bleached and consecutive time-lapse images were collected to observe the fluorescence recovery following the direction of fluid flow. The IF velocity was estimated using Matlab frap analysis plugin tracking the movement of the centroid of the bleached region.

[0147] Microtissue stiffness assessment

[0148] Mechanical resistance of fetal placental vascularized microtissues cultured under different flow conditions was measured on day 7 using a Chiaro Nanoindenter (Optics 11, Amsterdam, Netherlands), as previously described (28). Briefly, gels were cut and extracted from the devices, placed in a petri dish and fully submerged in VascuLife media. Nanoindentation measurements were performed using a spherical probe tip with a radius of 25 pm and a stiffness of 0.027 N / M. After proper calibration of the probe, 12 pm depth indentations were applied to the gels. The effective Young’s modulus values were derived from load-indentation curves by fitting to the standard Hertz model and assuming a Poisson’s ratio of 0.5, using the manufacturer’s data analysis plug-in. An average of ~80 measurements were performed per condition, for two independent experiments.

[0149] Flow cytometry

[0150] Characterization of stromal cell population was performed at day 3 and 7 after seeding by flow cytometry. To extract the cells from the devices, the gels were resected and digested in a solution of Accutase (Gibco) and 50 FU / ml Nattokinase (Japan Bioscience Ltd) for 15-20 min at 37°C. An average of 6 gels was pooled for each time point. Single cells were then stained with the pericyte marker CD 140b PerCP-Cy™5.5 (BD Bioscience) for 2 hours at 4 °C, washed with PBS, analyzed on a BD LSR II and later processed using FlowJo v.10.8.1 software. The stromal cell population was gated using CD140b+and by exclusion of the RFP+cell cluster (expressed only in HUVEC).

[0151] Immunostaining

[0152] On day 7, devices were washed with PBS and fixed with 4% paraformaldehyde for 15 min. Samples were then permeabilized and blocked for non-specific binding with 0.1% v / v Triton X-100, 5% w / v BSA and 1% v / v serum (same source of secondary antibody) in PBS overnight at 4 °C on an orbital shaker. Primary antibodies were diluted in 0.5% w / v BSA PBS and added to samples overnight at 4 °C on an orbital shaker. Primary antibodies used in this study are S100A4 (FSP-1) (1:100, Abeam), Collagen I (1:200, Abeam), Fibronectin (1 :200, Abeam), and Laminin (1:200, Abeam). The following day, samples were washed with 0.1% v / v Triton X- 100 PBS and then incubated with the appropriate secondary antibody (1:200, Alexa Fluor 488 or 647, Invitrogen) and DAPI counterstain diluted in PBS overnight at 4°C on an orbital shaker. Samples were washed with PBS and stored at 4 °C before imaging. To ensure effective staining and washing of unbound reagents, all washes and incubations steps were performed applying a pressure gradient across the gel.

[0153] Mass spectrometric characterization of matrix derived proteins

[0154] Tryptic digestion of hydrogel / tissue samples for mass spectrometry (MS) analysis was performed at day 7 after seeding as described by L.A. Sawicki et al.

[0029] , Two experiments were conducted, with each experiment having > 2 replicates per static- or flow-conditioned devices. Samples were first decellularized to remove cellular structures and protein contribution. Briefly, extracted gel matrices were washed with PBS and wash buffer (100 mM Na2HPO4, 2 mM MgC12, 2 mM EGTA) prior incubation to a lysis buffer (8 mM Na2HPO4, 1% NP-40) for 90 min at 37 °C. Following additional washes (300 mM KC1, 10 mM Na2HPO4), decellularized gels were degraded with collagenase (Gibco, 50 U / mL in HBSS) for 1 h at 37° C to release all retained proteins. Gels were then dissolved by vigorous pipetting and stored at -80°C prior to lyophilization (Labconco FreeZone 2.5). Lyophilized samples were reconstituted in 25 mM NFLHCCh, reduced with DTT and Cysteine residues were alkylated by the addition of iodoacetamide

[0029] , Proteins were digested over night by the addition of trypsin (Promega) and samples were acidified by using formic acid according to

[0029] , Digested samples were applied to spin column concentrators (Corning) with a 10 kDa cut off in order to remove trypsin and collagenase. Mixed peptides were subjected to a reverse phase clean-up step (OASIS HLB 96-well pElution Plate, Waters). Peptides were dried and reconstituted in 10 pl of 400 mM Hepes / NaOH, pH 8.5 and reacted for 1 h at room temperature with 80 pg of TMT18plex (Thermo Scientific) dissolved in 4 pl of acetonitrile. Peptides were subjected to a reverse phase clean-up step prior their analysis by LC-MS / MS on an Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific).

[0155] To this end, peptides were separated using an Ultimate 3000 nano RSLC system (Dionex) equipped with a trapping cartridge (Precolumn C18 PepMaplOO, 5 mm, 300 pm i.d., 5 pm, 100 A) and an analytical column (Acclaim PepMap 100. 75 x 50 cm C18, 3 mm, 100 A) connected to a nanospray -Flex ion source. The peptides were loaded onto the trap column at 30 pl per min using solvent A (0.1% formic acid) and eluted using a gradient from 2 to 80% Solvent B (0.1% formic acid in acetonitrile) over 2 h at 0.3 pl per min (all solvents were of LC-MS grade). The Orbitrap Fusion Lumos was operated in positive ion mode with a spray voltage of 2.2 kV and capillary temperature of 275° C. Full scan MS spectra with a mass range of 375-1.500 m / z were acquired in profile mode using a resolution of 120.000 with a maximum injection time of 50 ms, AGC operated in standard mode and aRF lens setting of 30%. Fragmentation was triggered for 3 s cycle time for peptide like features with charge states of 2-7 on the MS scan (data- dependent acquisition). Precursors were isolated using the quadrupole with a window of 0.7 m / z and fragmented with a normalized collision energy of 34%. Fragment mass spectra were acquired in profile mode and a resolution of 30,000. Maximum injection time was set to 94 ms and AGC target to custom. The dynamic exclusion was set to 60 s.

[0156] Acquired data were analyzed using FragPipe

[0030] and a Uniprot Homo sapiens database (UP000005640, ID9606 with 20594 entries, October 26th 2022) including common contaminants. The following modifications were considered: Carbamidomethyl (C, fixed), TMT18plex (K, fixed), Acetyl (N-term, variable), Oxidation (M, variable) and TMT18plex (N- term, variable). The mass error tolerance for full scan MS spectra was set to 10 ppm and for MS / MS spectra to 0.02 Da. A maximum of 2 missed cleavages were allowed. A minimum of 2 unique peptides with a peptide length of at least seven amino acids and a false discovery rate below 0.01 were required on the peptide and protein level

[0031] ,

[0157] MS data processing

[0158] The raw output files of FragPipe (protein.tsv - files, (30)) were processed using the R programming language (ISBN 3-900051-07-0). Contaminants were filtered out and only proteins that were quantified with at least two unique peptides were considered for the analysis. 333 proteins passed the quality control filters. Log2 transformed raw TMT reporter ion intensities were first cleaned for batch effects using the ‘removeBatchEffects’ function of the limma package (32) and further normalized using the vsn package (variance stabilization normalization, (33)). Proteins were tested for differential expression using the limma package. The replicate information was added as a factor in the design matrix given as an argument to the TmFif function of limma. A protein was annotated as a hit with a false discovery rate (fdr) smaller than 5 % and a fold-change of at least 30 % and as a candidate with a fdr below 5 % with no fold-change threshold. Hit and candidate proteins were clustered into 2 clusters (method kmeans) based on the Euclidean distance between normalized TMT intensities divided by the 0 mm H2O data point. Statistics

[0159] Statistical significance was analyzed using OriginPro v.9.85 performing one-way ANOVA followed by the post-hoc Tukey’s test for multiple comparison. Differences were considered statistically significant with p <0.05. Data shown here are from at least 2 independent experiments with n > 3 devices, with at least 3 measurements per device, unless otherwise specified.

[0160] As an alternative approach to animal testing, the inventors propose a new device and assays which allow for measurement of transfer of material across the maternal-fetal interface. Following the previous development of a method for growing human fetal -like vasculature on a microfluidic chip (25), now human umbilical vein endothelial cells (HUVEC) are co-cultured in the presence of stromal cells (fibroblasts and pericytes), which form interconnected microvessels capable of perfusion in 5-7 days. Then, trophoblasts cells are integrated into the system, generating the critical epithelial layer that forms a barrier between the maternal blood and fetal capillaries (Fig. 9, 12, and 13). In this embodiment, the trophoblast monolayer was seeded adjacent to the fetal vessels in the chip design established for generation of the fetal vasculature (25) using two immortalized lines, BeWo and HTR-8 / SVneo, or primary isolated trophoblasts (se Fig. 13).

[0161] The general method for producing the new preferred design according to the invention was:

[0162] (1) Introducing of a temperature-dependent dissolvable gel into the ■'maternal " channel of the device,

[0163] (2) Seeding of fetal microvessels (common protocol) into the "fetal" channel, and

[0164] (3) About 48 hours later, seeding of approx. 1.5 x 106cells / mL of trophoblasts into the "maternal" channel (see Figure 9), or in the embodiment with three channels, into the third medium channel (see Figure 10).

[0165] So far, the immortalized trophoblast cell lines, BeWo and HTR-8 / SVneo, and primary isolated trophoblasts (see Fig. 13) have been tested.

[0166] The inventive system, while initially setup as a maternal-fetal interface, can be used with other tissues. For instance, any tissue that has an epithelial layer and underlying vasculature can be generated on-chip. This allows for a use as a model also for respiratory or digestive tracks, for organs including the bladder, or for certain (epithelial) cancer types, such as, for example, breast cancer, in particular ductal carcinomas (DCIS).

[0167] Using the HTR-8 / svNeo trophoblasts, an impermeability to larger molecules (molecular weights of 70 and 4 kDa) and a low permeability to 0.4 kDa Dextrans over a 20 minute measurement was demonstrated (Figure 9B, Figure 14A). Immediate permeability of a glucose homolog (2-DG-labelled with GFP) was shown as expected (Figure 13C). In order to make this model system more pre-eclamptic (PE) like, oxidative stress (by incubation in low oxygen concentrations - 1% O2) or incubation with the known PE-associated circulating factor (sFLT- 1) was used (Figures 10, 11, 14, 15). These treatments led to the expected negative changes both in vascular density and trophoblast barrier function.

[0168] Production protocol for generating a 3-channel post-less chip design (see Figure 12 and 13A)

[0169] An inverse mold was generated according to the outlined design (Figure 12) either by 3D printing with a biocompatible resin or by laser cutting, as outlined in Cherubini and Haase (99). For laser cutting, individual devices were cut from 0.5 mm thin sheets of acrylic and acetone- bonded to a larger thicker (>1 mm) piece of acrylic. After generation of mold is complete, devices were generated by preparing poly-dimethyl siloxane (PDMS) in a 10:1 ratio of elastomer to crosslinker, which was cured within the mold. After 24 hours at 65°C, the PDMS was removed from the mold and inlet holes punched, similarly to Cherubini et al. (99). The PDMS devices were air-plasma bonded to clean glass coverslips, and the devices were left to return to hydrophobic state before seeding with cells (Figure 13 A).

[0170] Methods for generating and characterizing a model of the maternal-fetal interface

[0171] Human umbilical vein endothelial cells (HUVEC), human placental fibroblasts and human placental pericytes were seeded in the central gel channel to establish a placenta-specific microvasculature, as previously established in Cherubini et al. (100). Initially, the bottom third channel was filled with a dissolvable gelatin solution (10% gelatin in PBS, solid below 37°C) to allow the hydrogel forming solely in the central channel. After 48 hours once the gelatin was dissolved at 37°C, a *trophoblast cell (HTR-8 / SVneo) monolayer was introduced into the bottom channel (Figures 9 and 13B). A concentration of 800,000 cells / mL was used for HTR- 8 / SVneo cells. The device was tilted at 90° for 2 hours to encourage development of a confluent monolayer. Interstitial flow was then established across the central channel using a pressure gradient of 7mm H2O within a 3D resin printed reservoir established at 48 hours post-seeding. The media was changed, and the pressure gradient re-established daily. The placental barrier function can be assessed after 7 days of culture by perfusing the tissue with fluorescent dextrans of various molecular weights, immunoglobulins such as IgG and IgM, insulin, and glucose (using the fluorescent analog 2-NBDG) (Figure 13). The permeability of fluorescent dextrans and 2-NBDG was measured by perfusing these molecules into the maternal compartment and monitoring the fluorescent signal at the interface over approximately 12 minutes using a confocal microscope. Permeability values were calculated as in Shin et al. (101) (Figure 13D). The permeability of immunoglobulins and insulin was measured by perfusing these molecules into the maternal compartment and assessing their concentration in the supernatant collected from the fetal compartment using ELISA assays after 24 hours.

[0172] * Primary cytotrophoblast cells can be isolated from term placenta explants to form the epithelial barrier (Figure 13 F-H). These cells are obtained after dissecting chorionic villi and performing sequential digestion steps as described by Li and Schust (102).

[0173] Methods for generating and characterizing a model of pre-eclampsia (PE) using the maternal-fetal interface:

[0174] The PE phenotype in the maternal-fetal interface on-chip was generated using various approaches: by culturing the devices under hypoxic conditions (1% O2) for up to 7 days, or by perfusing the maternal compartment with PE-associated factors, such as soluble endoglin (sEng) or soluble fms-like tyrosine kinase-1 (sFltl), at concentrations similar to those found in the plasma of pre-eclamptic patients. For the hypoxic condition, all the cells (HUVEC, stromal and trophoblast cells) were preconditioned to 1% O2 (or to the reduced oxygen concentration of interest) prior to integration into the 3 -channel microfluidic system. PE-associated factors, such as sEng (30 ng / mL) or sFltl (50 ng / mL to match plasma levels) were perfused into the maternal compartment from day 5 in culture for the duration of 48 hours (Figures 14 and 15).

[0175] Methods for measurement at day 7 (or later)

[0176] The microvessel permeability was measured as previously described (100). The microvessel morphologic parameters (diameter, branch density, branch length, etc.) were measured by imaging at day 7 (other time points are possible). Bright field images were subjected to a machine-learning pipeline for vascular segmentation and spatial mapping of morphometric parameters (98) (Figure 15C). The trophoblast barrier permeability was measured by perfusing the tissue with fluorescent dextrans of various molecular weights and performing timelapse imaging at the interface between the maternal and fetal compartments. Trophoblast permeability values were calculated as in Shin et al. (101). Alternatively, perfusion of analytes (antibodies or small molecules) in the maternal compartment can be measured to cross the trophoblast barrier by collection in the fetal compartment media channels using an ELISA assay or LCMS.

[0177] Method in brief for generating a model of a breast duct-on-chip

[0178] To generate a breast duct model (Figure 16), human mammary vascular endothelial cells (HMVEC) and human mammary fibroblasts (HMF) were seeded in the central gel channel to establish mammary-specific microvasculature, as previously established in Moccia et al. (103). Alternatively, HMVEC can be replaced with iPSC-derived endothelial cells. After 8 hours, an epithelial cell (MCF 10) monolayer was introduced into the bottom channel. A concentration of 50,000 cells / mL was used for MCF 10 cells. Patient-derived epithelial cells can also be used after additional optimization. The device was tilted at a 90° for 1 hour to encourage development of a confluent monolayer. Interstitial flow was established using a pressure gradient of 3mm H2O within a 3D resin printed reservoir established at 24 hours post-seeding. *Optionally tumor spheroids or patient-derived organoids can be inserted into the bottom channel on day 1 after seeding, when the reservoir is added. Media was changed and the pressure gradient re-established daily. The volume of media in the reservoir chambers were 600 pL and 240 pL, respectively. Measurements of epithelial barrier function were made following several days of culture (or when microvessels were established) by perfusion with dextrans of various molecular weights.

[0179] Interstitial flow enables placental microvasculature network formation and function

[0180] The inventors previously presented a 3D in vitro model of placental terminal villi microvessels capable of recapitulating several aspects of placental vasculopathies (25). Now, to better approximate the physiological cellular composition of placental fetal tissue, the inventors sourced primary placental fibroblasts (HPF) that were integrated together with endothelial cells (HUVEC) and placental pericytes (HPP) to generate microvascular networks on a chip. Cells were cultured with an endothelial to stromal cell ratio of 10:1 in a 3.5 mg / ml fibrin gel, as opposed to 5:1 and 3 mg / ml previously (25), within a single-gel channel PDMS device (Fig. 1A). These differences in culture are in-part attributed to the different secretome of stromal cells isolated from placental and lung tissues, which show dissimilarities in angiogenic profiles. In particular, inflammatory molecules including PDGF-BB and MCP-1 are increased for placental fibroblasts in contrast to lung fibroblasts. The formation of connected microvascular networks (with this new tri-culture) occurs over approximately one week (Fig. IB); however, networks are fully perfusable only when an interstitial flow (IF) was applied across the hydrogel through the addition of a 3D printed media reservoir (shown in Fig. 1A top right). Placental microvessels cultured under static conditions self-assemble by day 4; however, they were not perfusable, showing loss of connectivity and vessel pruning by day 7.

[0181] Pressures gradients (AP) representative of static (0 mm H2O), low (3 mm H2O or ~30 Pa) and high pressure gradient (7 mm H2O or ~70 Pa) culture conditions were applied to microvessels (Fig 1C), with mean interstitial fluid flow velocity measured as 0.13 ± 0.06 pm / s, 0.30 ± 0.12 pm / s and 1.23 ± 0.32 pm / s, respectively at day 2. On day 7, perfusion with FITC-dextran shows complete vessel connectivity in devices cultured under flow, whereas partial perfusion was observed in its absence (Fig. 1C), quantified as perfused vessel area (Fig. ID).

[0182] Vessels form over several days in culture to produce open perfusable lumen, hence there is a shift from interstitial to luminal flow over time. This transition was measured by the fluid volumetric drop (every »4 hours). The pressure gradient decays exponentially at earlier timepoints but demonstrates a linear decrease over 24 hours by day 5. This suggests the presence of open-lumen and a switch from interstitial to primarily intraluminal flow at this timepoint in culture (Fig. IE).

[0183] Interstitial flow promotes early placental vasculogenesis

[0184] Placental microvessels become increasingly perfusable when subjected to daily interstitial flow, thus corresponding morphologic and functional barrier properties were assessed quantitatively. Development of the vessels was monitored at specific timepoints by confocal microscopy under static and flow conditions (Fig. 2A). In early phases of culture (day 3), IF conditioned vessels resulted in significant increases in network connectivity (Fig. 2B), reduced branch density (Fig. 2C) and larger vessel diameters (Fig. 2D). At day 7, connectivity remains unchanged between flow and static conditions. At this later time point, branch density was reduced in all cases compared to day 3 under the same conditions (Fig. 2C). Although diameter continues to increase over time (from day 3 to 7) across both conditions, flow conditioned microvessels are significantly larger in diameter compared to static cultured vessels (Fig. 2D). Next, the influence of static or flow conditioning on endothelial barrier function was examined by perfusion on day 7 with 70 kDa FITC-dextran. Vessels cultured under high flow conditions resulted in significantly decreased permeability values (increased barrier function) (Fig. 2E), indicating that an intermittently applied pressure gradient is sufficient to influence the phenotype of the endothelium over time.

[0185] Interstitial flow enhances inflammatory signaling in developing placental vessels

[0186] Vessel formation and remodeling is intimately linked to inflammation and growth factor signaling, and little is known about signaling in human fetoplacental vasculogenesis. In light of this, the inventors examined the release of various cytokines from the inventor’ s static and flow- conditioned placental microvessels. Both, inflammatory cytokines including interleukin 8 (IL8) and monocyte chemoattractant protein-1 (MCP-1), as well as the pro-angiogenic factors vascular endothelial growth factor (VEGF) and Angiopoietin 2 (Ang-2) were quantified by ELISA (Fig. 3). Generally, flow-conditioning resulted in increased inflammatory signaling across all timepoints, particularly for IL8, whereas for MCP1 levels dropped over time (although this was shown for both conditions). Ang-2 is also upregulated in case of flow- conditioned vessles, VEGF levels are significantly increased in statically cultured vessels over time (Fig. 3D).

[0187] Flow sustains placental microvascular stability

[0188] The inventors examined whether flow-conditioning could maintain structure and functionality in long-term culture of placental microvessels (> 7 days). Confocal images acquired at 14 days after seeding demonstrated distinct differences in vessel coverage between static and flow conditions (Fig. 4A). Microvessels maintained under static culture exhibited significantly reduced area coverage (Fig. 4B) and correspondingly narrow effective diameters (Fig. 4C). Branch density and length remained unchanged between static and flow conditions (Fig. 4D,E).

[0189] Perfusion with 70kDa FITC-dextran at day 14 allowed for characterization of barrier function at this late stage of culture. Reduced leakiness was observed for higher (7 mm H2O) pressure gradients compared to 3 mm H2O. Permeability of statically cultured vessels was lower overall; however, only a small fraction of vessels was perfused in this case (see Fig. 4G). Over time, vessel area and barrier function decrease in comparison to earlier timepoints (see Fig. 4H and orange dots from day 7 in Fig. 4F). Overall, flow-conditioning results in increased vascular coverage and allows for culture of connected vessels past 3-4 weeks. Characterization of velocity and shear stress distributions in placental microvessels

[0190] First, as a proxy for experimental measurements of fluid velocities inside the placental microvessels, fluorescent beads were introduced into the microvessels at day 7 and tracked by time-lapse microscopy. Mean velocities are significantly increased for flow-conditioned vessels, corresponding to an increased effective diameter (Fig. 2C).

[0191] Next, to accurately map velocity and shear stress distributions within the microvessels, large sections of 70 kDa FITC-dextran perfused devices were imaged and converted into binary masks for import as vectors into computational fluid dynamics (CFD) software. Using COMSOL, and the Brinkman equation physics module for flow through porous media, both the vessels and extravascular regions were modeled as separate domains. Vessels are treated as open pores with permeability of 1 and with simplified fluid parameters of water (density = 1000 kg / m3, viscosity = le'4Pa*s) and the extravascular matrix (EVM) as a gel (density = 985 kg / m3, viscosity = le-2 Pa*s) with a porosity of 0.3 and hydraulic permeability k=le'15m2, with fibrin parameters taken from the following references [33, 34], Note, hydraulic permeability of the acellular fibrin gel is ~k=le'13m2, but for the inventor’s simulations k=le'15m2was used as an estimate, to account for the cellular component (placental fibroblasts and pericytes). A pressure gradient (50 Pa or 5 mm H2O) was simulated across the gel region, a as a mid-range value between the 3 and 7 mm H2O used in the experimental setup. From these simulations both velocity and shear stress distributions are predicted (Fig. 5). The preferential flow patterns are shown to be heterogeneous and depend on lumen formation at the interface of the gel and media channels.

[0192] Interstitial flow alters biophysical properties of the placental microvascular tissue

[0193] Extracellular matrix (ECM) provides a basic structure in which hollow blood vessels reside and remodel. ECM provides mechanical and chemical cues that contribute to vascular network formation and barrier integrity (35). The inventors previously demonstrated significant changes in tissue stiffness and diffusional properties in response to co-culture with specific fibroblasts (28). Here, the inventors assessed whether flow conditioning alone can drive these tissue-level changes. First, the inventors examined diffusivity in extravascular regions from vessels cultured under static or flow conditions at day 7. FRAP measurements (Fig. 6A) were performed from which diffusion coefficients were calculated. Flow-conditioned vessels result in significantly reduced diffusivity in extravascular regions (Fig. 6B). Next, the inventors assessed the impact of flow-conditioning on microvessel tissue stiffness. Apparent Young's moduli were assessed by nanoindentation at day 7 by exposing the tissue for probing as shown in the schematic in Fig. 6C. Microvascular tissues cultured under flow are significantly stiffer than the static ones (Fig. 6D), with stiffness being dependent on the magnitude of the pressure gradient (7 versus 3 mm H2O).

[0194] Since ECM is dynamic, constantly undergoes remodeling and can activate cell proliferation and tissue morphogenesis (36), the inventors assessed whether changes in tissue-level properties are associated with changes in the stromal cell population over time. To this aim, cells were extracted from the hydrogel matrix of static and flow conditioned devices 7 days after seeding to compare the cell composition over time, cells were analyzed by flow cytometry across days; however, no differences in the percentage of endothelial and stromal cells were found (Fig. 6E).

[0195] Flow promotes significant ECM protein deposition and remodeling

[0196] Tissue-level stiffness increase corresponded with decreased diffusivity in extravascular regions of flow-conditioned placental vessels. Since there are many identified ECM proteins in the developing placenta, the inventors next examined the impact of flow on protein production. After 7 days in culture, microvessels were fixed and stained for collagen I, laminin and fibronectin (Fig. 7A). Normalized mean intensities, measured in comparison to static conditions, demonstrates that all ECM proteins are more significantly expressed under flowconditions (Fig. 7B). To validate these findings, the inventors performed mass spectrometry analysis to measure compositional changes of the matrix under static and flow-conditions for vessels after 7 days in culture. Specifically, individual samples from each condition were subjected to tandem mass tag (TMT) labeling, followed by multiplexing, and ultimately measured in a mass spectrometry-based experiment. As a result, the inventors identified a total of 804 proteins, of which 333 were quantified, with many showing relevance to vascularization, extracellular matrix (ECM) composition and organization, as classified by gene ontology (Fig. 8A). For each protein, fold changes expression between conditions was calculated and ratios of protein expression was obtained by dividing the fold change values for flow conditioned samples by the corresponding fold change value for the static counterparts. Moreover, PCA showed that static and flow-conditioned sample replicates were spaced and clustered, consistently with their experimental condition. Thus, the inventors identified 26 proteins as hits and candidates, which the inventors clustered into six different groups based on their expression similarities and behaviors under different conditions (Fig. 8B). Several hits and candidates from the protein list were found to be associated with ECM composition and remodeling. For instance, matrix constituent laminin (LAMC1) appears significantly increased under flow conditions (Fig 8C) indicating major protein deposition due to fluid mechanical stress. On the other end, protein such as Inter-alpha-trypsin inhibitor heavy chain H2 (ITIH2) and Pentraxin 3 (PTX3), which are known to be regulators of ECM stability (37, 38), display reduced expression (statistically significant in ITIH2) suggesting flow-induced matrix destabilization and consequent remodeling.

[0197] Generating a maternal-fetal interface model on-chip

[0198] Based on the development and characterization of the perfusable feto-placental microvessels (100), the inventors further developed a model inclusive of a trophoblast layer, representative of the maternal-fetal interface. This required the design of a novel device with a central gel channel and 3 separate media channels - each separated only by a reduced wall acting as a phase-guide (Figures 12 and 13). By including a dissolvable gel (gelatin) in the third channel (with no mirroring channel), the cell and fibrin hydrogel is seeded within the frame of the middle channel (Figure 13). Human umbilical vein endothelial cells (HUVEC), human placental fibroblasts and human placental pericytes were seeded in the central gel channel to establish a placental-specific microvasculature, as in (100). Subsequently, following dissolution of the gelatin, trophoblast cells are added to the third channel 24-48 hours later. HTR-8 / SVneo are shown introduced into the bottom channel, which form a tight monolayer. Interstitial flow was established at 48 hours post-seeding using a pressure gradient of 7mm H2O within a 3D resin printed reservoir.

[0199] This design allows for separate perfusion of a fetal and maternal compartment and enables simultaneous growth of the 2 barriers critical to represent placental transport, namely the trophoblast layer and underlying feto-placental microvessels.

[0200] Characterizing the healthy maternal-fetal interface

[0201] To assess whether or not the maternal -fetal interface behaves as expected in vivo, a number of assays were performed, including assessing the capacity of the trophoblast barrier to limit or permit transport of known solutes. Normal trophoblast barrier function was assessed after 7 days of culture by perfusion of the maternal channel with a glucose analog (fluorescent 2- NBDG). As expected (104), 2-NBDG passed the trophoblast barrier readily (Figure 13C). Next, fluorescent dextrans of various molecular weights were perfused into the maternal channel, and permeability of the trophoblast barrier assessed. The trophoblast barrier was capable of limiting transport to molecules larger than IkDa (Figure 9B) and shows a size dependent effect on transport - less transport of larger molecules (Figure 13 D).

[0202] To further support physiologic placental barrier function, immunoglobulins including IgG and IgM, and insulin, were perfused into the maternal compartment and assessed at 24 hours for transport by ELISA measurements of collected supernatants (Figure 13 E). As expected, transport was limited, particularly for insulin, which does not cross the maternal -fetal interface.

[0203] A similar trophoblast barrier can also be developed using primary trophoblasts (isolated cytotrophoblasts that can fuse into syncytiotrophoblasts in vitro). These have been shown to develop a barrier in the model (Figure 13 F-H).

[0204] Mimicking a pre-eclamptic state in the maternal-fetal interface model

[0205] The inventors have shown that either by perfusion of soluble molecules relevant in preeclampsia (sFLT-1 and sEng) or by culture in hypoxic conditions (1% O2), according to the timeline in Figure 14, that normal trophoblast barrier function is disrupted. Permeability (measured as fluorescent dextran flux) is measured according to the different treatments (Figure 14 B). Hypoxic culture conditions result in a significant dysfunction of the trophoblast barrier.

[0206] In addition to measuring trophoblast barrier function, the microvessels were assessed under these conditions (Figure 15). Imaging of microvascular networks were performed 2 days following treatments, or at 7 days following 1% O2 culture. Using a machine-learning pipeline, microvascular networks are segmented, and morphologic features quantified in a spatial manner (Figure 15 C). Average changes in specific vascular features can be reported, such as mean vessel diameter (Figure 15 D). Using this approach, morphometric data demonstrates the significant effect of all treatments - hypoxia and sFLT-1 and sEng negatively impact development of feto-placental microvessels. Moreover, microvessel permeability can be measured, as done previously (and shown in Figure 15). While hypoxia results in complete microvessel regression and lack of perfusion, sEng results in leakier microvessels.

[0207] Use of the 3-channel model as a mammary duct-on-chip In addition to generating a maternal-fetal interface using the post-less 3 channel design, the inventors have also demonstrated its use as a duct-on-chip model, useful for measuring transport across a breast duct (Figure 16). Inclusion of mammary-derived microvessels (as in 103) into the central channel, is followed by inclusion of an epithelial layer. The epithelial barrier function can be measured with or without the presence of a tumor in the duct channel and also shows responsiveness to hormones (follicular levels of combined estrogen and progesterone, as in (103)) as well as tamoxifen (Figure 16G).

[0208] References as cited

[0209] [1] K.L. Thornburg, S. Louey, Uteroplacental circulation and fetal vascular function and development, Curr Vase Pharmacol 11(5) (2013) 748-57.

[0210] [2] P. Kaufmann, T.M. Mayhew, D.S. Charnock-Jones, Aspects of human fetoplacental vasculogenesis and angiogenesis. II. Changes during normal pregnancy, Placenta 25(2-3) (2004) 114-26.

[0211] [3] J. Kingdom, B. Huppertz, G. Seaward, P. Kaufmann, Development of the placental villous tree and its consequences for fetal growth, Eur J Obstet Gynecol Reprod Biol 92(1) (2000) 35- 43.

[0212] [4] T.M. Mayhew, D.S. Charnock-Jones, P. Kaufmann, Aspects of human fetoplacental vasculogenesis and angiogenesis. III. Changes in complicated pregnancies, Placenta 25(2-3) (2004) 127-39.

[0213] [5] R. Demir, P. Kaufmann, M. Castellucci, T. Erbengi, A. Kotowski, Fetal vasculogenesis and angiogenesis in human placental villi, Acta Anat (Basel) 136(3) (1989) 190-203.

[0214] [6] T.M. Mayhew, Fetoplacental angiogenesis during gestation is biphasic, longitudinal and occurs by proliferation and remodelling of vascular endothelial cells, Placenta 23(10) (2002) 742-50.

[0215] [7] J.S. Abramowicz, E. Sheiner, Ultrasound of the placenta: a systematic approach. Part II: functional assessment (Doppler), Placenta 29(11) (2008) 921-9.

[0216] [8] A.R. Clark, M. Lin, M. Tawhai, R. Saghian, J.L. James, Multiscale modelling of the fetoplacental vasculature, Interface Focus 5(2) (2015) 20140078.

[0217] [9] R. Plitman Mayo, D.S. Charnock-Jones, G.J. Burton, M L. Oyen, Three-dimensional modeling of human placental terminal villi, Placenta 43 (2016) 54-60.

[0218]

[0010] P. Campinho, A. Vilfan, J. Vermot, Blood Flow Forces in Shaping the Vascular System: A Focus on Endothelial Cell Behavior, Frontiers in Physiology 11 (2020).

[0011] V. Gebala, R. Collins, I. Geudens, L.K. Phng, H. Gerhardt, Blood flow drives lumen formation by inverse membrane blebbing during angiogenesis in vivo, Nat Cell Biol 18(4) (2016) 443-50.

[0219]

[0012] P.A. Galie, D.H. Nguyen, C.K. Choi, D.M. Cohen, P.A. Janmey, C.S. Chen, Fluid shear stress threshold regulates angiogenic sprouting, ProcNatl Acad Sci U S A 111(22) (2014) 7968- 73.

[0220]

[0013] Y. Abe, M. Watanabe, S. Chung, R.D. Kamm, K. Tanishita, R. Sudo, Balance of interstitial flow magnitude and vascular endothelial growth factor concentration modulates three- dimensional microvascular network formation, APL Bioeng 3(3) (2019) 036102.

[0221]

[0014] S. Kim, M. Chung, J. Ahn, S. Lee, N.L. Jeon, Interstitial flow regulates the angiogenic response and phenotype of endothelial cells in a 3D culture model, Lab Chip 16(21) (2016) 4189-4199.

[0222]

[0015] F. Orsenigo, C. Giampietro, A. Ferrari, M. Corada, A. Galaup, S. Sigismund, G. Ristagno,

[0223] L. Maddaluno, G.Y. Koh, D. Franco, V. Kurtcuoglu, D. Poulikakos, P. Baluk, D. McDonald,

[0224] M. Grazia Lampugnani, E. Dejana, Phosphorylation of VE-cadherin is modulated by haemodynamic forces and contributes to the regulation of vascular permeability in vivo, Nat Commun 3 (2012) 1208.

[0225]

[0016] K. Haase, F. Piatti, M. Marcano, Y. Shin, R. Visone, A. Redaelli, M. Rasponi, R.D. Kamm, Physiologic flow-conditioning limits vascular dysfunction in engineered human capillaries, Biomaterials (2021) 121248.

[0226]

[0017] E. Kochhan, A. Lenard, E. Ellertsdottir, L. Herwig, M. Affolter, H.G. Belting, A.F. Siekmann, Blood flow changes coincide with cellular rearrangements during blood vessel pruning in zebrafish embryos, PLoS One 8(10) (2013) e75060.

[0227]

[0018] A. Lenard, E. Ellertsdottir, L. Herwig, A. Krudewig, L. Sauteur, H.G. Belting, M. Affolter, In vivo analysis reveals a highly stereotypic morphogenetic pathway of vascular anastomosis, Dev Cell 25(5) (2013) 492-506.

[0228]

[0019] G. Acharya, S.-E. Sonesson, K. Flo, J. Rasanen, A. Odibo, Hemodynamic aspects of normal human feto-placental (umbilical) circulation, Acta Obstetricia et Gynecologica Scandinavica 95(6) (2016) 672-682.

[0229]

[0020] N. Kupper, E. Pritz, M. Siwetz, J. Guettler, B. Huppertz, Placental Villous Explant Culture 2.0: Flow Culture Allows Studies Closer to the In Vivo Situation, Int J Mol Sci 22(14) (2021).

[0230]

[0021] C. Mandrycky, B. Hadland, Y. Zheng, 3D curvature-instructed endothelial flow response and tissue vascularization, Sci Adv 6(38) (2020) eabb3629.

[0022] A. Marturano-Kruik, M M. Nava, K. Yeager, A. Chramiec, L. Hao, S. Robinson, E. Guo, M.T. Raimondi, G. Vunjak-Novakovic, Human bone perivascular niche-on-a-chip for studying metastatic colonization, Proc Natl Acad Sci USA 115(6) (2018) 1256-1261.

[0231]

[0023] M.A. Winkelman, D.Y. Kim, S. Kakarla, A. Grath, N. Silvia, G. Dai, Interstitial flow enhances the formation, connectivity, and function of 3D brain microvascular networks generated within a microfluidic device, Lab Chip 22(1) (2021) 170-192.

[0232]

[0024] M. Cherubini, S. Erickson, K. Haase, Modelling the Human Placental Interface In Vitro- A Review, Micromachines (Basel) 12(8) (2021).

[0233]

[0025] K. Haase, M.R. Gillrie, C. Hajal, R.D. Kamm, Pericytes Contribute to Dysfunction in a Human 3D Model of Placental Microvasculature through VEGF-Ang-Tie2 Signaling, Advanced Science 6(23) (2019) 1900878.

[0234]

[0026] J.-Y. Tinevez, N. Perry, J. Schindelin, G.M. Hoopes, G.D. Reynolds, E. Laplantine, S.Y. Bednarek, S.L. Shorte, K.W. Eliceiri, TrackMate: An open and extensible platform for singleparticle tracking, Methods 115 (2017) 80-90.

[0235]

[0027] P. Jonsson, M.P. Jonsson, J.O. Tegenfeldt, F. Hook, A method improving the accuracy of fluorescence recovery after photobleaching analysis, Biophys J 95(11) (2008) 5334-48.

[0236]

[0028] A. Akinbote, V. Beltran-Sastre, M. Cherubini, R. Visone, C. Hajal, D. Cobanoglu, K. Haase, Classical and Non-classical Fibrosis Phenotypes Are Revealed by Lung and Cardiac Like Microvascular Tissues On-Chip, Frontiers in Physiology 12 (2021)

[0237]

[0029] L.A. Sawicki, L.H. Choe, K.L. Wiley, K.H. Lee, A.M. Kloxin, Isolation and Identification of Proteins Secreted by Cells Cultured within Synthetic Hydrogel-Based Matrices, ACS Biomater. Sci. Eng. 4(3) (2018) 836-845.

[0238]

[0030] M.M. Savitski, M. Wilhelm, H. Hahne, B. Kuster, M. Bantscheff, A Scalable Approach for Protein False Discovery Rate Estimation in Large Proteomic Data Sets, Mol Cell Proteomics 14(9) (2015) 2394-2404.

[0239]

[0031] M.E. Ritchie, B. Phipson, D. Wu, Y. Hu, C.W. Law, W. Shi, G.K. Smyth, limma powers differential expression analyses for RNA-sequencing and microarray studies, Nucleic Acids Res 43(7) (2015) e47.

[0240]

[0032] W. Huber, A. von Heydebreck, H. Sultmann, A. Poustka, M. Vingron, Variance stabilization applied to microarray data calibration and to the quantification of differential expression, Bioinformatics 18 Suppl 1 (2002) S96-104.

[0241]

[0033] O. Moreno-Arotzena, J.G. Meier, C. Del Amo, J.M. Garcia-Aznar, Characterization of Fibrin and Collagen Gels for Engineering Wound Healing Models, Materials (Basel) 8(4) (2015) 1636-1651.

[0034] Y.T. Ho, G. Adriani, S. Beyer, P.T. Nhan, R.D. Kamm, J.C.Y. Kah, A Facile Method to Probe the Vascular Permeability of Nanoparticles in Nanomedicine Applications, Sci Rep 7(1) (2017) 707.

[0242]

[0035] D J. LaValley, C.A. Reinhart-King, Matrix stiffening in the formation of blood vessels, Advances in Regenerative Biology 1(1) (2014) 25247.

[0243]

[0036] C. Bonnans, J. Chou, Z. Werb, Remodelling the extracellular matrix in development and disease, Nat Rev Mol Cell Biol 15(12) (2014) 786-801.

[0244]

[0037] A. Hamm, J. Veeck, N. Bektas, P.J. Wild, A. Hartmann, U. Heindrichs, G. Kristiansen, T. Werbowetski-Ogilvie, R. Del Maestro, R. Knuechel, E. Dahl, Frequent expression loss of Inter- alpha-trypsin inhibitor heavy chain (ITIH) genes in multiple human solid tumors: A systematic expression analysis, BMC Cancer 8(1) (2008) 25.

[0245]

[0038] N.S. Baranova, A. Inforzato, D.C. Briggs, V. Tilakaratna, J. J. Enghild, D. Thakar, C.M. Milner, A. J. Day, R.P. Richter, Incorporation of Pentraxin 3 into Hyaluronan Matrices Is Tightly Regulated and Promotes Matrix Cross-linking, Journal of Biological Chemistry 289(44) (2014) 30481-30498.

[0246]

[0039] L.T. Merce, M.J. Barco, S. Bau, Color Doppler sonographic assessment of placental circulation in the first trimester of normal pregnancy, Journal of Ultrasound in Medicine 15(2) (1996) 135-142.

[0247]

[0040] V.H.J. Roberts, T.K. Morgan, P. Bednarek, M. Morita, G.J. Burton, J O. Lo, A.E. Frias, Early first trimester uteroplacental flow and the progressive disintegration of spiral artery plugs: new insights from contrast-enhanced ultrasound and tissue histopathology, Human Reproduction 32(12) (2017) 2382-2393.

[0248]

[0041] S.R. Chary, R.K. Jain, Direct measurement of interstitial convection and diffusion of albumin in normal and neoplastic tissues by fluorescence photobleaching, Proc Natl Acad Sci U S A 86(14) (1989) 5385-9.

[0249]

[0042] H. Dafni, T. Israely, Z.M. Bhujwalla, L.E. Benjamin, M. Neeman, Overexpression of vascular endothelial growth factor 165 drives peritumor interstitial convection and induces lymphatic drain: magnetic resonance imaging, confocal microscopy, and histological tracking of triple-labeled albumin, Cancer Res 62(22) (2002) 6731-9.

[0250]

[0043] S. Jones, H. Bischof, I. Lang, G. Desoye, S.L. Greenwood, E.D. Johnstone, M. Wareing, C.P. Sibley, P. Brownbill, Dysregulated flow-mediated vasodilatation in the human placenta in fetal growth restriction, The Journal of Physiology 593(14) (2015) 3077-3092.

[0044] K.M. Wieczorek, A S. Brewer, L. Myatt, Shear stress may stimulate release and action of nitric oxide in the human fetal-placental vasculature, American Journal of Obstetrics and Gynecology 173(3, Part 1) (1995) 708-713.

[0251]

[0045] T.O. Junaid, R.S. Bradley, R.M. Lewis, J.D. Aplin, E.D. Johnstone, Whole organ vascular casting and microCT examination of the human placental vascular tree reveals novel alterations associated with pregnancy disease, Scientific Reports 7(1) (2017) 4144.

[0252]

[0046] M.T. Bryan, H. Duckies, S. Feng, S.T. Hsiao, H R. Kim, J. Serbanovic-Canic, P C. Evans, Mechanoresponsive networks controlling vascular inflammation, Arterioscler Thromb Vase Biol 34(10) (2014) 2199-205.

[0253]

[0047] J. Ando, K. Yamamoto, Flow detection and calcium signalling in vascular endothelial cells, Cardiovasc Res 99(2) (2013) 260-8.

[0254]

[0048] K.H. Hong, J. Ryu, K.H. Han, Monocyte chemoattractant protein- 1 -induced angiogenesis is mediated by vascular endothelial growth factor-A, Blood 105(4) (2005) 1405-7.

[0255]

[0049] J. Heidemann, H. Ogawa, M.B. Dwinell, P. Rafiee, C. Maaser, H.R. Gockel, M.F. Otterson, D.M. Ota, N. Lugering, W. Domschke, D.G. Binion, Angiogenic Effects of Interleukin 8 (CXCL8) in Human Intestinal Microvascular Endothelial Cells Are Mediated by CXCR2*, Journal ofBiological Chemistry 278(10) (2003) 8508-8515.

[0256]

[0050] K. Shimoya, N. Matsuzaki, T. Taniguchi, T. Kameda, M. Koyama, R. Neki, F. Saji, O. Tanizawa, Human placenta constitutively produces interleukin-8 during pregnancy and enhances its production in intrauterine infection, Biol Reprod 47(2) (1992) 220-6.

[0257]

[0051] P. Toti, F. Arcuri, Z. Tang, F. Schatz, E. Zambrano, G. Mor, T. Niven-Fairchild, V.M. Abrahams, G. Krikun, C.J. Lockwood, S. Guller, Focal increases of fetal macrophages in placentas from pregnancies with histological chori oamnionitis: potential role of fibroblast monocyte chemotactic protein-1, Am J Reprod Immunol 65(5) (2011) 470-9.

[0258]

[0052] M. Cheng, J. Wu, X. Liu, Y. Li, Y. Nie, L. Li, H. Chen, Low shear stress-induced interleukin-8 mRNA expression in endothelial cells is mechanotransduced by integrins and the cytoskeleton, Endothelium 14(6) (2007) 265-73.

[0259]

[0053] H. Yu, Y. Zeng, J. Hu, C. Li, Fluid shear stress induces the secretion of monocyte chemoattractant protein- 1 in cultured human umbilical vein endothelial cells, Clin Hemorheol Microcirc 26(3) (2002) 199-207.

[0260]

[0054] T. Aoki, K. Yamamoto, M. Fukuda, Y. Shimogonya, S. Fukuda, S. Narumiya, Sustained expression of MCP-1 by low wall shear stress loading concomitant with turbulent flow on endothelial cells of intracranial aneurysm, Acta Neuropathol Commun 4(1) (2016) 48.

[0055] L.P. Reynolds, D.A. Redmer, Angiogenesis in the Placental, Biology of Reproduction 64(4) (2001) 1033-1040.

[0261]

[0056] E.G. Zhang, S.K. Smith, P.N. Baker, D.S. Chamock- Jones, The Regulation and Localization of Angiopoietin-1, -2, and Their Receptor Tie2 in Normal and Pathologic Human Placentae, Molecular Medicine 7(9) (2001) 624-635.

[0262]

[0057] P.C. Maisonpierre, C. Suri, P.F. Jones, S. Bartunkova, S.J. Wiegand, C. Radziejewski, D. Compton, J. McClain, T.H. Aldrich, N. Papadopoulos, T.J. Daly, S. Davis, T.N. Sato, G.D. Yancopoulos, Angiopoietin-2, a Natural Antagonist for Tie2 That Disrupts in vivo Angiogenesis, Science 277(5322) (1997) 55-60.

[0263]

[0058] M. Dixit, E. Bess, B. Fisslthaler, F.V. Hartel, T. Noll, R. Busse, I. Fleming, Shear stress- induced activation of the AMP-activated protein kinase regulates FoxOla and angiopoietin-2 in endothelial cells, Cardiovasc Res 77(1) (2008) 160-8.

[0264]

[0059] S. Chlench, N. Mecha Disassa, M. Hohberg, C. Hoffmann, T. Pohlkamp, G. Beyer, M. Bongrazio, L. Da Silva- Azevedo, O. Baum, A.R. Pries, A. Zakrzewicz, Regulation of Foxo-1 and the angiopoietin-2 / Tie2 system by shear stress, FEBS Letters 581(4) (2007) 673-680.

[0265]

[0060] W. Goettsch, C. Gryczka, T. Korff, E. Ernst, C. Goettsch, J. Seebach, H.J. Schnittler, H.G. Augustin, H. Morawietz, Flow-dependent regulation of angiopoietin-2, J Cell Physiol 214(2)

[0266] (2008) 491-503.

[0267]

[0061] R. Demir, Expression of VEGF receptors VEFGR-1 and VEGFR-2, angiopoietin receptors Tie-1 and Tie-2 in chorionic villi tree during early pregnancy, Folia Histochem Cytobiol 47(3)

[0268] (2009) 435-45.

[0269]

[0062] F. Bamt, A. Barut, B.D. Gun, N.O. Kandemir, M I. Harma, M. Harma, E. Aktunc, S.O. Ozdamar, Intrauterine growth restriction and placental angiogenesis, Diagnostic Pathology 5(1)

[0270] (2010) 24.

[0271]

[0063] M. Ouame, A. Pena, C.A. Franco, From remodeling to quiescence: The transformation of the vascular network, Cells & Development 168 (2021) 203735.

[0272]

[0064] C. Deroy, N. Stovall-Kurtz, F. Nebuloni, C. Soitu, P.R. Cook, E.J. Walsh, Predicting flows through microfluidic circuits with fluid walls, Microsystems & Nanoengineering 7(1) (2021) 93.

[0273]

[0065] J. Cappello, J. Rivero-Rodriguez, Y. Vitry, A. Dewandre, B. Sobac, B. Scheid, Beads, bubbles and drops in microchannels: stability of centred position and equilibrium velocity, Journal of Fluid Mechanics 956 (2023) A21.

[0066] T.A. Russo, A.M.M. Banuth, H.B. Nader, J.L. Dreyfuss, Altered shear stress on endothelial cells leads to remodeling of extracellular matrix and induction of angiogenesis, PLoS One 15(11) (2020) e0241040.

[0274]

[0067] S.A. Douglas, K. Haase, R.D. Kamm, M.O. Platt, Cysteine cathepsins are altered by flow within an engineered in vitro microvascular niche, APL Bioeng 4(4) (2020) 046102.

[0275]

[0068] R.M. Delaine-Smith, S. MacNeil, G.C. Reilly, Matrix production and collagen structure are enhanced in two types of osteogenic progenitor cells by a simple fluid shear stress stimulus, Eur Cell Mater 24 (2012) 162-74.

[0276]

[0069] T. Yamane, N. Yamaguchi, Y. Yoshida, M. Mitsumata, Regulation of extracellular matrix production and degradation of endothelial cells by shear stress, International Congress Series 1262 (2004) 407-410.

[0277]

[0070] R.L. Steward, C.-M. Cheng, J.D. Ye, R.M. Beilin, P.R. LeDuc, Mechanical stretch and shear flow induced reorganization and recruitment of fibronectin in fibroblasts, Scientific Reports 1(1) (2011) 147.

[0278]

[0071] B.B. O'Connor, B.D. Pope, M.M. Peters, C. Ris-Stalpers, K.K. Parker, The role of extracellular matrix in normal and pathological pregnancy: Future applications of microphy si ologi cal systems in reproductive medicine, Exp Biol Med (Maywood) 245(13) (2020) 1163-1174.

[0279]

[0072] S. Ji, D. Gumina, K. McPeak, R. Moldovan, M.D. Post, E.J. Su, Human placental villous stromal extracellular matrix regulates fetoplacental angiogenesis in severe fetal growth restriction, Clin Sci (Lond) 135(9) (2021) 1127-1143.

[0280]

[0073] R. Pellicani, E. Poletto, E. Andreuzzi, A. Paulitti, R. Doliana, D. Bizzotto, P. Braghetta, R. Colladel, G. Tarticchio, P. Sabatelli, F. Bucciotti, G. Bressan, R.V. lozzo, A. Colombatti, P. Bonaldo, M. Mongiat, Multimerin-2 maintains vascular stability and permeability, Matrix Biology 87 (2020) 11-25.

[0281]

[0074] A. Fejza, E. Poletto, G. Carobolante, L. Camicia, E. Andreuzzi, A. Capuano, E. Pivetta, R. Pellicani, R. Colladel, S. Marastoni, R. Doliana, R.V. lozzo, P. Spessotto, M. Mongiat, Multimerin-2 orchestrates the cross-talk between endothelial cells and pericytes: A mechanism to maintain vascular stability, Matrix Biol Plus 11 (2021) 100068.

[0282]

[0075] J.W. Weisel, R.I. Litvinov, Fibrin Formation, Structure and Properties, in: D A D Parry, J.M. Squire (Eds.), Fibrous Proteins: Structures and Mechanisms, Springer International Publishing, Cham, 2017, pp. 405-456.

[0076] L.Q. Zhang, D.P. Heruth, S.Q. Ye, Nicotinamide Phosphoribosyltransferase in Human Diseases, J Bioanal Biomed 3 (2011) 13-25.

[0283]

[0077] G. Mouneimne, S.D. Hansen, L.M. Selfors, L. Petrak, M.M. Hickey, L.L. Gallegos, K.J. Simpson, J. Lim, F.B. Gertler, J.H. Hartwig, R.D. Mullins, J.S. Brugge, Differential remodeling of actin cytoskeleton architecture by profilin isoforms leads to distinct effects on cell migration and invasion, Cancer Cell 22(5) (2012) 615-30.

[0284]

[0078] M. Nejedla, Z. Li, A.E. Masser, M. Biancospino, M. Spiess, S.D. Mackowiak, M.R. Friedlander, R. Karlsson, A Fluorophore Fusion Construct of Human Profilin I with Non- Compromised Poly(L-Proline) Binding Capacity Suitable for Imaging, J Mol Biol 429(7) (2017) 964-976.

[0285]

[0079] N. Elkhatib, Matthew B. Neu, C. Zensen, Kurt M. Schmoller, D. Louvard, Andreas R. Bausch, T. Betz, Danijela M. Vignjevic, Fascin Plays a Role in Stress Fiber Organization and Focal Adhesion Disassembly, Current Biology 24(13) (2014) 1492-1499.

[0286]

[0080] N.S. Dellschaft, G. Hutchinson, S. Shah, N.W. Jones, C. Bradley, L. Leach, C. Platt, R. Bowtell, P.A. Gowland, The haemodynamics of the human placenta in utero, PLoS Biol 18(5) (2020) e3000676.

[0287]

[0081] G.E. Lash, K. Naruse, B.A. Innes, S.C. Robson, R.F. Searle, J.N. Bulmer, Secretion of Angiogenic Growth Factors by Villous Cytotrophoblast and Extravillous Trophoblast in Early Human Pregnancy, Placenta 31(6) (2010) 545-548.

[0288]

[0082] Cherubini M, Erickson S, Haase K (2021) Modelling the Human Placental Interface In Vitro-A Review. Micromachines (Basel) 12 (8). doi: 10.3390 / mil2080884

[0289]

[0083] Rothbauer M, Patel N, Gondola H, Siwetz M, Huppertz B, Ertl P (2017) A comparative study of five physiological key parameters between four different human trophoblast-derived cell lines. Scientific Reports 7 (1):5892. doi: 10.1038 / s41598-017-06364-z

[0290]

[0084] Almeida MPO, Ferro EAV, Briceno MPP, Oliveira MC, Barbosa BF, Silva NM (2019) Susceptibility of human villous (BeWo) and extravillous (HTR-8 / SVneo) trophoblast cells to Toxoplasma gondii infection is modulated by intracellular iron availability. Parasitology Research 118 (5)11559-1572. doi:10.1007 / s00436-019-06257-2

[0291]

[0085] Widhalm R, Ellinger I, Granitzer S, Forsthuber M, Bajtela R, Gelles K, Hartig P-Y, Hengstschlager M, Zeisler H, Salzer H, Gundacker C (2020) Human placental cell line HTR- 8 / SVneo accumulates cadmium by divalent metal transporters DMT1 and ZIP ! 4’>‘. Metallomics 12 (11)11822-1833. doi:10.1039 / d0mt00199f

[0086] Li Z, Kurosawa O, Iwata H (2020) A Novel Human Placental Barrier Model Based on Trophoblast Stem Cells Derived from Human Induced Pluripotent Stem Cells. Tissue Eng Part A 26 (13-14):780-791. doi: 10.1089 / ten.TEA.2019.0342

[0292]

[0087] Blundell C, Yi Y-S, Ma L, Tess ER, Farrell MJ, Georgescu A, Aleksunes LM, Huh D (2018) Placental Drug Transport-on-a-Chip: A Microengineered In Vitro Model of Transporter- Mediated Drug Efflux in the Human Placental Barrier. Adv Healthc Mater 7 (2). doi : 10.1002 / adhm .201700786

[0293]

[0088] Schuller P, Rothbauer M, Kratz SRA, Holl G, Taus P, Schinnerl M, Genser J, Bastus N, Moriones OH, Puntes V, Huppertz B, Siwetz M, Wanzenbock H, Ertl P (2020) A lab-on-a-chip system with an embedded porous membrane-based impedance biosensor array for nanoparticle risk assessment on placental Bewo trophoblast cells. Sensors and Actuators B: Chemical 312: 127946. doi: 10.1016 / j .snb.2020.127946

[0294]

[0089] Blundell C, Tess ER, Schanzer ASR, Coutifaris C, Su EJ, Parry S, Huh D (2016) A microphy si ologi cal model of the human placental barrier. Lab Chip 16 (16):3065-3073. doi:10.1039 / c61c00259e

[0295]

[0090] Lee JS, Romero R, Han YM, Kim HC, Kim CJ, Hong J-S, Huh D (2016) Placenta-on-a- Chip: A novel platform to study the biology of the human placenta. J Matem Fetal Neona 29 (7): 1046-1054. doi: 10.3109 / 14767058.2015.1038518

[0296]

[0091] Mosavati B, Oleinikov AV, Du E (2020) Development of an Organ-on-a-Chip-Device for Study of Placental Pathologies. International Journal of Molecular Sciences 21 (22). doi:10.3390 / ijms21228755

[0297]

[0092] Haider S, Meinhardt G, Saleh L, Kunihs V, Gamperl M, Kaindl U, Ellinger A, Burkard TR, Fiala C, Pollheimer J, Mendjan S, Latos PA, Knofler M (2018) Self-Renewing Trophoblast Organoids Recapitulate the Developmental Program of the Early Human Placenta. Stem Cell Reports 11 (2):537-551. doi: 10.1016 / j. stemcr.2018.07.004

[0298]

[0093] Turco MY, Gardner L, Kay RG, Hamilton RS, Prater M, Hollinshead MS, McWhinnie A, Esposito L, Fernando R, Skelton H, Reimann F, Gribble FM, Sharkey A, Marsh SGE, O'Rahilly S, Hemberger M, Burton GJ, Moffett A (2018) Trophoblast organoids as a model for maternal- fetal interactions during human placentation. Nature 564 (7735):263-267. doi:10.1038 / s41586- 018-0753-3

[0299]

[0094] McConkey CA, Delorme-Axford E, Nickerson CA, Kim KS, Sadovsky Y, Boyle JP, Coyne CB (2016) A three-dimensional culture system recapitulates placental syncytiotrophoblast development and microbial resistance. Sci Adv 2 (3):el501462. doi:10.1126 / sciadv.1501462

[0095] Kupper N, Pritz E, Siwetz M, Guettler J, Huppertz B (2021) Placental Villous Explant Culture 2.0: Flow Culture Allows Studies Closer to the In Vivo Situation. Int J Mol Sci 22 (14). doi:10.3390 / ijms22147464

[0300]

[0096] Nishiguchi A, Gilmore C, Sood A, Matsusaki M, Collett G, Tannetta D, Sargent IL, McGarvey J, Halemani ND, Hanley J, Day F, Grant S, Murdoch-Davis C, Kemp H, Verkade P, Aplin JD, Akashi M, Case CP (2019) In vitro placenta barrier model using primary human trophoblasts, underlying connective tissue and vascular endothelium. Biomaterials 192: 140- 148. doi: 10.1016 / j . biomaterials.2018.08.025

[0301]

[0097] MandtD, Gruber P, Markovic M, TromayerM, RothbauerM, Kratz SRA, Ali SF, Hoorick IV, Holnthoner W, Miihleder S, Dubruel P, Vlierberghe SV, Ertl P, Liska R, Ovsianikov A (2018) Fabrication of biomimetic placental barrier structures within a microfluidic device utilizing two-photon polymerization. Int J Bioprint 4 (2): 144. doi:10.18063 / IJB.v4i2.144

[0302]

[0098] Rappez L, Akinbote A, Cherubini M, Uhlmann V, Haase K, Label-free phenotyping of human microvessel networks, bioRxiv (2024) 2024.02.20.581133.

[0303]

[0099] Cherubini M., Haase K., A Bioengineered Model for Studying Vascular-Pericyte Interactions of the Placenta, Methods Mol Biol 2608 (2023) 409-423.

[0304]

[0100] Cherubini M., Erickson S., Padmanaban P., Haberkant P., Stein F., Beltran-Sastre V., Haase K., Flow in fetoplacental-like microvessels in vitro enhances perfusion, barrier function, and matrix stability, Science Advances 9(51) (2023) eadj 8540.

[0305]

[0101] Shin Y., Choi S.H., Kim E., Bylykbashi E., Kim J.A., Chung S., Kim D.Y., Kamm R.D., Tanzi R.E., Blood-Brain Barrier Dysfunction in a 3D In Vitro Model of Alzheimer's Disease, Advanced Science 6(20) (2019) 1900962.

[0306]

[0102] Li L , Schust D.J., Isolation, purification and in vitro differentiation of cytotrophoblast cells from human term placenta, Reprod Biol Endocrinol 13 (2015) 71.

[0307]

[0103] Moccia C., Cherubini M., Fortea M., Akinbote A., Padmanaban P , Beltran-Sastre V., Haase K., Mammary Microvessels are Sensitive to Menstrual Cycle Sex Hormones, Adv Sci (Weinh) 10(35) (2023) e2302561.

[0308]

[0104] Joshi N.P., Mane A.R., Sahay A S., Sundrani D.P., Joshi S.R., Yajnik C.S., Role of Placental Glucose Transporters in Determining Fetal Growth, Reprod Sci 29(10) (2022) 2744- 2759.

[0309]

[0105] Kouthouridis S, Sotra A, Khan Z, Alvarado J, Raha S, Zhang B. Modeling the Progression of Placental Transport from Early- to Late-Stage Pregnancy by Tuning Trophoblast Differentiation and Vascularization. Adv Healthc Mater. 2023 Dec;12(32):e2301428. doi: 10.1002 / adhm.202301428. Epub 2023 Nov 7. PMID: 37830445.

Claims

Claims1. A microfluidic device, comprising a) an elongated central culture channel for holding mammalian cells in a suitable gel, b) at least one first media channel extending at least in part thereof arranged in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at least in part thereof in parallel to an opposite second side of the elongated central culture channel, configured in a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, and c) at least one third media channel extending at least in part thereof in parallel to a third side of the elongated central culture channel, wherein optionally the at least one third media channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one third media channel, wherein preferably the device comprises a fourth media channel extending at least in part thereof in parallel to a fourth side of the elongated central culture channel, wherein optionally the fourth media channel is filled with a dissolvable gel barrier suitable to block cells from entering into the at least one fourth media channel, and wherein preferably the at least one dissolvable gel barrier comprises a temperature-dependent dissolvable gel.

2. The microfluidic device according to claim 1, wherein the mammalian cells in the culture chamber are selected from human cells, and preferably comprise cells and / or cell lines, such as cancer and / or immortalized cell lines, derived from epithelial tissues having an epithelial layer, vascular cells, cells of the respiratory system, cells of the digestive system, bladder cells, mammary cells, ductal cells and cancer cells thereof, and preferably comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP), and a layer of human trophoblasts, located on the side of the cell culture that is in contact with the at least one third or fourth media channel.

3. A microfluidic device, comprising a) an elongated central culture channel for holding human cells in a suitable gel, b) at least one first media channel extending at least in part thereof in parallel to a first side of the elongated central culture channel, and at least one second media channel extending at leastin part thereof in parallel to an opposite second side of the elongated central culture channel, configured in a lateral flow connection from the at least one first media channel to the at least one second media channel through the central culture channel, wherein the mammalian cells in the gel comprise a tri-culture of human umbilical vein endothelial cells (HUVECs), human placental fibroblasts (HPF), and human placental pericytes (HPP).

4. The microfluidic device according to claim 2 or 3, wherein the HPF and HPP are cultured at a ratio of about 10: 1 of endothelial :stromal cells HPF / HPP, preferably wherein the HPF and HPP are at a ratio of 1 : 1.

5. The microfluidic device according to any one of claims 1 to 4, wherein the gel of the culture channel comprises a hydrogel, such as a fibrinogen / thrombin gel.

6. The microfluidic device according to any one of claims 1 to 5, further comprising added media reservoirs, preferably 3D-printed tailored media reservoirs, connected to at least one of the at least one media channel.

7. The microfluidic device according to any one of claims 1 to 6, wherein the device furthermore comprises a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel.

8. The microfluidic device according to any one of claims 1 to 7, wherein the device is positioned on a solid support, such as a biochip, and / or wherein several or a multitude of devices are grouped on said solid support.

9. A method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to any one of claims 1 to 8, b) incubating the cells in the culture chamber in a suitable culture medium, and c) dissolving of at least one of the gel barrier(s) as described herein, and d) seeding suitable epithelial and / or epidermal mammalian barrier cells, wherein preferably in b) the tri-culture is incubated for about 48h in a suitable culture medium under essentially static conditions,bl) a lateral flow is established, for example an intermittent interstitial flow (IF) or luminal flow, from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of between preferably about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of preferably between about 0.20 pm / s and 1.25 Lim / s across the central culture channel, and wherein the flow is established between about day 2 and about day 7 of the tri-culture.

10. A method for producing a placental-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to any one of claims 3 to 8, b) incubating the tri-culture for about 48h in a suitable culture medium under essentially static conditions, c) establishing a lateral flow, for example an intermittent interstitial flow (IF) or luminal flow, from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of preferably between about ~20 Pa and ~70 Pa between the at least one first media channel and the at least one second media channel and / or by applying a mean lateral flow velocity of preferably between about 0.20 pm / s and 1.25 pm / s across the central culture channel, wherein the flow is established between about day 2 and about day 7 of the tri-culture, and d) optionally, testing the perfusion rate following step c).

11. The method according to claim 9 or 10, wherein the hydrostatic pressure gradient (AP) as established in step c) is at less than about 10 mm H2O, preferably at between about: 2 and 7 mm H2O, and / or wherein the flow is at 3 mm H2O or ~30 Pa or a mean lateral flow velocity of 0.30 ± 0.12 pm / s on day 2, and / or wherein the flow is re-stablished daily by replenishing the reservoir media volumes during 5 days following step b).

12. The method according to any one of claims 9 to 11, further comprising the step of introducing a barrier layer of cells, preferably of trophoblasts, that is positioned on the side of either the first or second side of the elongated central culture channel, and / or further comprising the step of detecting changes in the expression and / or amount of cellular analytes, such as, for example, angiogenic and inflammatory cytokines during step c) and / or d).

13. A method for producing a breast-derived human 3D vascular microtissue model, comprising the steps of a) providing the microfluidic device according to any one of claims 1, 2 or 8, b) seeding human mammary vascular endothelial cells (HMVEC) or iPSC-derived endothelial cells and human mammary fibroblasts (HMF) in the elongated central culture channel for about 8h in a suitable culture medium under essentially static conditions, c) introducing an epithelial cell (MCF10 or other cell line) monolayer into the third media channel, optionally followed by the insertion of tumor spheroids or patient-derived organoids on day 1 after seeding, d) establishing a lateral flow, for example an intermittent interstitial flow (IF) or luminal flow, from at least one first media channel to the at least one second media channel across the central culture channel by applying a hydrostatic pressure gradient of 3mm H2O 24 hours post-seeding, and e) optionally, testing the perfusion rate following step d).

14. A placental-derived human 3D vascular microtissue model, produced according to a method according to any one of claims 9 to 12 or a breast-derived human 3D vascular microtissue model produced according to a method according to claim 13, preferably exhibiting a laminar or luminal flow across the model, and / or exhibiting an epithelial barrier function.

15. Use of the microfluidic device according to any one of claims 1 to 8 or the placental -derived human 3D vascular microtissue model to study placental barrier function(s) or the breast- derived human 3D vascular microtissue model according to claim 14 to study breast duct function(s).

16. Use of the microfluidic device according to any one of claims 1 to 8 or the placental -derived human 3D vascular microtissue model according to claim 13 to study trophoblast barrier function, such as, for example, for toxicological screening of substances and / or stresses that interfere with trophobl st cell invasion.

17. Use of the microfluidic device according to any one of claims 1 to 8 or the placental -derived human 3D vascular microtissue model according to claim 13 or the breast-derived human 3D vascular microtissue model according to claim 14 to study placental or breast branchingangiogenesis, breast cancer and related angiogenesis, the impact of nutrients, hormones, and environmental factors on placental or breast development, gestational disorders of fetal vasculature, and / or vascular-associated gestational disorders.

18. Use of the microfluidic device according to any one of claims 1 to 8 or the placental -derived human 3D vascular microtissue model according to claim 13 or the breast-derived human 3D vascular microtissue model according to claim 14 to study epithelial and / or endothelial barrier function (permeability to solutes), for drug screenings, for assessing novel drug targets against disfunction, for example in pre-eclamp si a, transport of solutes, pollutants, antibodies, immune cells and drugs across an epithelial cell barrier and / or a vascular cell barrier and extravascular matrix properties (diffusivity, stiffness, and matrix proteins), and in particular the effects of luminal, interstitial and / or lateral flow on study epithelial and / or endothelial barrier function (permeability to solutes) and extravascular matrix properties (diffusivity, stiffness, and matrix proteins).

Images