Device and method for perfusion of 3D printed tissues

Abstract

Described herein is a pumpless recirculating platform for substantially unidirectional fluid flow. Also described herein is an epithelium model with patterned crypts and underlying perfusable microvasculature. A stamp for imprinting a patterned crypt on a crosslinked hydrogel scaffold is also described as well as an open chamber for cell culture, the chamber comprising a basolaterally perfused compartment and an apically perfused compartment. Related systems and methods are also described.

Description

[0001] DEVICE AND METHOD FOR PERFUSION OF 3D PRINTED TISSUES

[0002] FIELD

[0003] The present disclosure relates to tissue engineering, and in particular, to cell culture systems for modeling tissues.

[0004] BACKGROUND

[0005] Blood vessels play vital roles in organ function and development. Vascular systems not only supply tissues with oxygen and nutrients, but also participate in many biological processes, such as transporting immune cells in inflammatory response, trafficking cancer cells in tumor metastasis, establishing biochemical gradients, improving parenchymal tissue survival and function through paracrine signaling, and repairing tissues through angiogenesis, etc. The incorporation of vascular networks in biological models is integral to accurately model human diseases as well as to maintain proper tissue function in vitro. Microphysiological systems, also known as organ-on-a-chip, have been developed to model various cellular microenvironments in the human body1. For instance, Huh et al. developed a lung alveoli -on-a- chip, lung airway -on-a-chip, and placenta-on-a-chip, etc2-4. These organ chips consist of multi-layer microfluidic channels separated by a porous membrane that are lined with epithelium cells and human endothelial cells to model the epithelial and vascular interface of various organs. The membrane based vascular barrier is advantageous for tracking drug and biomolecular transports, immune cell extravasation, as well as simulating blood circulation.

[0006] Microfluidic systems have also been developed to culture microvascular networks to study angiogenesis, vasculogenesis and various vascular events by taking advantage of the ability of endothelial cells to self-assemble into a microvascular network given the right conditions5. This type of system allows for the intimate contact between the endothelial cells and other stromal and parenchymal cells and is particularly useful for studying the remodeling process of blood vessels in details in response to the chemical gradient, fluid flow, and drug treatments6, 7. However, these self-assembled vascular networks are often constrained inside a closed or semi-closed microfluidic channel that limits their physical integration upon implantation or with other larger tissues models such as organoids8, 9.

[0007] International Patent Application Publication No. WO 2021 / 223031 describes a cell culture system for constructing a perfusable network of self-assembled cells comprising a multi -well plate embedded with microchannels connecting a central well with at least one inlet well and at least one outlet well, the central well for culturing seeded cells within an extracellular matrix, wherein the perfusable network allows perfusion through the microchannels connecting the central well with at least one inlet well and at least one outlet well. The cell culture system allows the array of perfusable networks formed, connected, and perfused inside the multi-well plate to be accessible and / or extractable from the top of the central well. In aspects, the cell culture system can improve the experimental throughputs of organ-on-a-chip systems and expand the application of microphysiological systems to regenerative cell therapy. A perfusable network of self-assembled cells and method of making thereof using the cell culture system are also described. SUMMARY

[0008] In accordance with an aspect, there is provided a pumpless recirculating platform for substantially unidirectional fluid flow.

[0009] In an aspect, the fluid flow is driven by gravity.

[0010] In an aspect, the gravity -driven fluid flow is controlled by a rocker.

[0011] In an aspect, the rocker has a programmable dwell time at different tilt angles.

[0012] In an aspect, the platform comprises an inlet well, a tissue well, and an outlet well in substantially unidirectional fluid communication with one another.

[0013] In an aspect, a first channel connects the inlet well and the tissue well in fluid communication and a second channel connects the tissue well and the outlet well in fluid communication.

[0014] In an aspect, a slanted bridge connects the inlet well and the outlet well in fluid communication, wherein the slanted bridge is lower at the outlet well than at the inlet well to encourage unidirectional flow, such that: when the platform is tilted in a first direction, the higher end of the slanted bridge prevents fluid flow from the inlet well to the outlet well and encourages fluid flow from the inlet well, to the tissue well, to the outlet well; and when the platform is tilted in a second direction, the lower end of the slanted bridge encourages fluid flow from the outlet well to the inlet well over fluid flow in the opposite direction.

[0015] In an aspect, the slanted bridge has an angle of about 20°.

[0016] In an aspect, fluid flows from the inlet well to the tissue well, to the outlet well, and back to the inlet well.

[0017] In an aspect, the tissue well is bottomless.

[0018] In an aspect, the inlet well and / or the outlet well is bottomless.

[0019] In an aspect, the platform is provided in combination with a plate for holding the platform.

[0020] In an aspect, the plate holds a plurality of platforms.

[0021] In an aspect, the plate holds from about 8 platforms to about 128 platforms in standard 24 to 384 well format.

[0022] In an aspect, the platform is provided in combination with a sacrificial material for template production.

[0023] In an aspect, the sacrificial material is dissolvable.

[0024] In an aspect, the sacrificial material comprises gelatin, pluronic acid, alginate, or any combination thereof.

[0025] In an aspect, the sacrificial material is 3D printed on an adhesive sheet for application to an open bottom of the platform.

[0026] In an aspect, the sacrificial material creates one or more channels between the inlet well and the outlet well.

[0027] In an aspect, the one or more channels comprise vascular channels for seeding of endothelial cells. In an aspect, the platform is provided in combination with a substrate upon which to seed cells.

[0028] In an aspect, the substrate comprises fibrin gel, collagen, Matrigel, an extracellular matrix (ECM) hydrogel, or any combination thereof.

[0029] In an aspect, the platform is provided in combination with cells.

[0030] In an aspect, the cells form tubular tissues such as vascular tissues or renal proximal tubules.

[0031] In an aspect, the cells form spheroids.

[0032] In an aspect, the cells comprise immune cells, endothelial cells, cancer cells, epithelial cells, stromal cells, or any combination thereof.

[0033] In an aspect, the platform provides a shear stress of from about 0.7 to about 2 dyne / cm2.

[0034] In an aspect, the platform provides a fluid flow rate of from about 100 to 200 |iL / min.

[0035] In accordance with an aspect, there is provided a method for making the platform described herein, the method comprising applying a sacrificial template to an injection-molded bottomless well.

[0036] In accordance with an aspect, there is provided a method of perfusing cells, the method comprising seeding cells on the platform described herein and recirculating fluid substantially unidirectionally without the use of a pump.

[0037] In an aspect, cancer spheroids are seeded on the platform adjacent to a vascular tube, through which the fluid is recirculated, wherein the fluid comprises immune cells.

[0038] In an aspect, a tilt angle of from about - 20° to about - 60°, such as about - 45°, is used for a recirculation tilt angle and wherein a tilt angle of from about 5° to about 45°, such as about 15°, is used for a perfusion tilt angle.

[0039] In accordance with an aspect, there is provided an epithelium model with patterned crypts and underlying perfusable microvasculature.

[0040] In an aspect, the crypts are in an open well providing access to the apical surface of the epithelium.

[0041] In an aspect, the model is vascularized with self-assembled endothelial cells and fibroblasts to yield a tissue microenvironment with vascular-epithelial crosstalk.

[0042] In an aspect, the patterned crypts are imprinted on a hydrogel scaffold.

[0043] In an aspect, the model is an intestinal model, a skin hair follicle model, a lung alveoli model, or a mammary gland model.

[0044] In an aspect, the model is a colon model.

[0045] In accordance with an aspect, there is provided an immune-infiltrated cancer spheroid model made by continuously perfusing and recirculating immune cells with gravity -driven flow through a tubular blood vessel adjacent to a cancer spheroid.

[0046] In accordance with an aspect, there is provided a cancer-macrophage heterospheroid model comprising a macrophage population that continuously grows, renews, and / or self-evolves.

[0047] In accordance with an aspect, there is provided an immune-infiltrated cancer spheroid model comprising a vascular tube through which immune cells unidirectionally flow and a cancer spheroid adjacent to the vascular tube. In an aspect, the model further comprises stromal cells.

[0048] In an aspect, the model further comprises a cytokine gradient.

[0049] In accordance with an aspect, there is provided a stamp for imprinting a patterned crypt on a crosslinked hydrogel scaffold.

[0050] In an aspect, the stamp is 3D printed.

[0051] In an aspect, the stamp is formed from high temperature resin that is autoclavable.

[0052] In an aspect, the stamp comprises high aspect ratio features.

[0053] In an aspect, the high aspect ratio features comprise an array of pyramidal pin structures printed at the tip of the stamp.

[0054] In an aspect, the stamp surface is functionalized to reduce adherence of the stamp to the scaffold.

[0055] In an aspect, the stamp surface is polyethylene glycol (PEG) photografted or coated with an antiadherence rinsing solution.

[0056] In accordance with an aspect, there is provided a method of patterning a crosslinked hydrogel, the method comprising applying the stamp described herein to the crosslinked hydrogel and then removing the stamp.

[0057] In accordance with an aspect, there is provided an open chamber for cell culture, the chamber comprising a basolaterally perfused compartment and an apically perfused compartment.

[0058] In an aspect, the flow of fluid through both the basolaterally perfused compartment and the apically perfused compartment is gravity -driven.

[0059] In an aspect, the basolaterally perfused compartment comprises a basolateral inlet and a basolateral outlet for fluid flow.

[0060] In an aspect, the basolateral inlet leads to a basolateral inlet chamber and / or the basolateral outlet leads to a basolateral outlet chamber.

[0061] In an aspect, the basolateral inlet chamber and / or the basolateral outlet chamber are open or closed.

[0062] In an aspect, the apically perfused compartment comprises an apical inlet and an apical outlet for fluid flow.

[0063] In an aspect, the apical inlet leads to an apical inlet chamber and / or the apical outlet leads to an apical outlet chamber.

[0064] In an aspect, flow of fluid in the basolaterally perfused compartment is substantially orthogonal to the flow of fluid in the apically perfused compartment.

[0065] In an aspect, the apically perfused compartment is open for access to the apical compartment.

[0066] In an aspect, the open chamber comprises a hydrogel to separate the basolaterally perfused compartment and the apically perfused compartment.

[0067] In an aspect, the basolaterally perfused compartment and the apically perfused compartment are not separated by a membrane, such as a synthetic membrane.

[0068] In an aspect, the open chamber comprises a crypt-patterned base.

[0069] In accordance with an aspect, there is provided a platform comprising: a central open chamber for cell culture; a basolateral inlet chamber and a basolateral outlet chamber each adjacent to and in fluid communication with the open chamber; and an apical inlet chamber and an apical outlet chamber each adjacent to and in fluid communication with the open chamber; wherein the basolateral inlet and outlet chambers are not in fluid communication with the apical inlet and outlet chambers.

[0070] In an aspect, the platform comprises a second apical inlet chamber adjacent to and in fluid communication with the apical inlet chamber and a second apical outlet chamber adjacent to and in fluid communication with the apical outlet chamber.

[0071] In an aspect, the open chamber comprises a crypt-patterned base.

[0072] In accordance with an aspect, there is provided a multi-well plate comprising the open chamber or platform described herein.

[0073] Other features and advantages of the present disclosure will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples, while indicating embodiments of the disclosure, are given by way of illustration only and the scope of the claims should not be limited by these embodiments, but should be given the broadest interpretation consistent with the description as a whole.

[0074] DRAWINGS

[0075] Certain embodiments of the disclosure will now be described in greater detail with reference to the attached drawings in which:

[0076] FIGURE 1. Fabrication and setup of UniPlate, a) Schematic illustration and image of the UniPlate fabrication process, including manufacturing of customized bottomless wells (STEP 1) and gelatin printing on pressure-sensitive adhesive sheets (STEP 2). b) Schematic of the device assembly, c) Image of assembled UniPlate. Each plate contained eight devices grouped together using a 3D printed frame that fit inside a OneWell Plate. The inset shows the tissue well with 3D printed gelatin sacrificial template.

[0077] FIGURE 2. Injection molding, a) 3D printed master molds for manufacturing customized bottomless wells with injection molding. The master molds are designed into two pieces for easy release. One piece of master mold contains the inner features of the customized bottomless wells, and the other defines the outer frame, b) Photos of desktop injection molder and setup. The 3D printed master molds are designed to fit into an aluminum mold, c) The photos of manufactured bottomless wells before and after release from printed master molds.

[0078] FIGURE 3 Operation and recirculating flow on UniPlate, a) Schematic illustration of gelatin removal and perfusable channel formation after casting fibrin gel in the tissue wells, b) Single -channel UniPlate in two different designs: linear (I channel) and tortuous (S channel); c) Schematic illustration and time-lapse photos of a complete perfusion and recirculation cycle. Red food dye was added to the inlet well to better visualize the process.

[0079] FIGURE 4. a) The setting parameters of programable rocker for bidirectional and unidirectional flow conditions, b) In recirculation process, residual volume of UniPlate under various initial liquid volume in outlet wells at - 30° tilt angle; n=3 devices for each condition. Statistics significance was determined using one-way ANOVA. “ns” indicates not statistically significant, c) Confocal projected z-stack fluorescence images of vessels with bidirectional flow culture in UniPlate; Cells were stained for CD31 (red) and DAPI (blue). Scale bar: 200 pm. d) GFP -expressed endothelial cells-lined vessels formed after 7 days of culture under bidirectional flows with 5 min or 15 min dwell time; Scale bar: 1 mm.

[0080] FIGURE 5. Flow characterization and unidirectional perfusion of 3D tubular vessels on a UniPlate, a-b) Correlation of flow rate and wall shear stress with rocker perfusion tilt angles; n=3 devices for each condition, c) Correlation of residual volume in UniPlate outlet wells to recirculating tilt angles when the initial volume of media added to the outlet wells is 1 mL; n=3 devices for each tilt angle condition. Statistical significance was determined using one-way ANOVA. ***P < 0.001. d) Correlation of residual volume in UniPlate with initial media volume in outlet wells at - 40° recirculating tilt angle; n=3 devices for each condition. Statistical significance was determined using one-way ANOVA. “ns” indicates not statistically significant, e) Fluorescent images of blood vessels formed after seven days of culture under unidirectional and bidirectional flow; scale bar:l mm. f) Fluorescence images of a perfused (65 kDa dextran, red) blood vessel formed under unidirectional flow for permeability quantification on day 8; scale bar:l mm. g) Permeability coefficient quantification of vessels under unidirectional and bidirectional flow culture conditions; n=3 vessels per group. Statistical significance was determined using Kruskal -Wallis one-way ANOVA on ranks, “ns” indicates not statistically significant, h) Confocal projected z-stack fluorescence images of vessels with unidirectional flow on UniPlate. The cells were stained with CD31 (red) and DAPI (blue). Scale bar: 200 pm. i-j) Cell nucleus morphology and nucleus orientation quantification of endothelial cells in vessels after 7 days of culture with unidirectional or bidirectional flow; n=3 vessels analyzed for each culture condition. To quantify the cell nucleus orientation, 214 cells from three vessels were counted in the unidirectional group, and 238 cells from three vessels were counted in the bidirectional group. 0° indicates the cell nucleus orientation along the flow direction. Statistical significance was determined using Kruskal-Wallis one-way ANOVA on ranks, “ns” indicates not statistically significant.

[0081] FIGURE 6. A vascularized proximal tubule with recirculating flow can reabsorb glucose on UniPlate, a) Schematic of dual-channel recirculation flow setup; b) Top and bottom views of a 3D vascularized proximal tubule tissue perfused with blue and red food dyes; c) Images of a proximal tubule next to a tubular blood vessel (green) on day 4 after cell seeding; scale bar:l mm. d) Fluorescence images of dextran-perfused (65 kDa, red and green) vascularized proximal tubule tissue; scale bar:l mm. e-f) Confocal fluorescence images of vascularized proximal tubule tissue stained for VE-cadherin (red), SGLT2 (red), Na / K ATPase (green), and DAPI (blue). Scale bar: 200 pm. g) Quantification of glucose concentration in media perfusates from vascular and tubular channels on day 7 after reaching confluency. n=6 for each condition. Statistical significance was determined using one-way ANOVA. * indicates P < 0.05; “ns” indicates not statistically significant.

[0082] FIGURE 7. UniPlate enables long-term unidirectional recirculation of monocytes without pro- inflammatory activation, a) Schematic of monocyte (THP-1) recirculation in UniPlate and analysis, b) Time-lapse images of monocytes at the bottom of the inlet and outlet wells and inside the blood vessel channels after 24 h of static and recirculation cultures. Static and moving cells in the blood vessel channels are labeled with red and yellow squares, respectively. Scale bar: 200 pm. c) Quantification of cell viability before (0 h) and after 24 h of static or recirculation culture on UniPlate; n=3 for each condition, and cell viability at 0 h was used as the control. Statistical significance was determined using one-way ANOVA. “ns” indicates not statistically significant, d) Representative live / dead staining images of monocytes after 24 h of static or recirculating culture. Scale bar: 200 pm. e-f) Quantification of (e) total monocyte cell number and (f) monocyte cell numbers in inlet and outlet wells collected from media perfusates after 24 h of static culture or recirculation. n=3 for each condition. Statistical significance was determined using oneway ANOVA. *P < 0.05, ***P < 0.001, “ns” indicates not statistically significant, g) Quantification of cytokine secretion levels, including monocyte modulatory cytokines (h) GM-CSF and (i) MCP-1, in collected medium perfusates after 24 h of static and recirculating cultures. n=4 for each condition. Statistical significance was determined using one-way ANOVA. *P < 0.05, “ns” indicates not statistically significant.

[0083] FIGURE 8. Secretion of all analyzed inflammatory -related cytokines, including pro-inflammatory cytokines: IL-ip (a), IL-2 (b), IL-6 (c), IL-8 (d), TNF-a (e), IFN-y (f), IL-12p40 (g), IL-12p70 (h), and anti-inflammatory cytokines: IL-IRa (i), IL-4 (j), IL-5 (k), IL-10 (1), IL-13 (m) in static and recirculation groups. n=4 devices for each condition. Statistics significance was determined one-way ANOVA. “ns” indicates not statistically significant.

[0084] FIGURE 9. Development of a dynamic colon microenvironment in the IFlowPlate384. a, Conceptualization of the stepwise fabrication towards a colonic tissue model with vascularized crypts as well as apical and basolateral flow in the IFlowPlate384. b, Illustration on the design and operation of the IFlowPlate384. Scale bar, 1 mm.

[0085] FIGURE 10. Crypt-patterned colon barrier formation and characterization, a, Illustration of microstamping using the TopoStamp to imprint crypt surface topography on fibrin gels, b, Fluorescent confocal images of crypt-patterned colon barriers (top row) in F-actin (red), and DAPI (blue), as well as a depth colour gradient from combined F-actin and DAPI fluorescence (bottom). Scale bar, 200 pm. c, False coloured SEM images of the top and cross-section of a crypt-patterned colon epithelium barrier. Scale bars from left to right, 200, 50, 10, 10, 50, 200, and 200 pm. d, Timeline for colon crypt tissue biofabrication and assay quantification, e, Quantification of 4 kDa and 65 kDa dextran transport (n = 10;**P<0.01 ***P<0.001 ****P<0.0001). f, Fluorescent images of colon epithelium barriers stained with EdU (red) and DAPI (blue), with and without crypts, in the IFlowPlate384 and in a standard 384 wellplate, x-y plane scale bar, 1 mm. 3D plane scale bar, 500 pm. g-h, Quantification of proliferating cells (n = 3 - 5;**P<0.01 ****p<0.0001). All data is presented as mean ± s.d.

[0086] FIGURE 11. Surface functionalization of TopoStamp pin arrays. Black boxes outline patterning failure from gel warping

[0087] FIGURE 12. Caco-2 cultivation in various media and seeding densities, a, Brightfield images of Caco-2 proliferation in various cell media. Scale bar, 1 mm. b, Quantified Caco-2 proliferation in varying media (n = 5;*P<0.05 **P<0.01 ***P<0.001 ****P<0.000F). c, Brightfield images of Caco-2 expansion at various seeding densities. Scale bar, 1 mm. d, Quantification of Caco-2 confluence at varying seeding densities (n = 5; ****P<0.0001). All data is presented as mean ± s.d.

[0088] FIGURE 13. Chemotherapeutic drug dose response on crypt-patterned and planar colon epithelium tissues, a, Automated robotic liquid handling of fibrin gel casting and rapid crypt imprinting in the IFlowPlate384. b, Experimental timeline, c, Brightfield images of an array of crypt-patterned and planar colon epithelium on day 12. Tissues with defects were discarded and covered by black boxes. Scale bar, 3 mm. d, Permeability assay on day 12 (n = 55 - 61;****P<0.0001). e, Illustration of capecitabine prodrug mechanism of action, f, Quantification of dextran diffusion rates in crypt-patterned and planar tissues exposed to increasing doses of capecitabine before treatment (left), after treatment (middle), and after recovery (right). All data is presented as mean ± s.e.m.

[0089] FIGURE 14. Perfusable microvascular beds with crypt-patterned colon epithelium, a, Experimental timeline and schematic of tissue arrangement. Scale bar, 1 mm. b, Permeability of avascular and vascularized crypt-patterned colon epithelium as well as fibrin gel controls (n = 3 - ll;*P<0.05 **P<0.01). c, Histological cross-sections of vascularized crypts with hematoxylin (ECM, purple) and eosin (nuclei, blue), villin-1 (microvilli, dark brown), and CD31 (blood vessel, dark brown) staining. Scale bar, 200 pm. d, Confocal images of vascularized crypts stained with F-actin (red), DAPI (blue), and GFP- HUVEC (green). Scale bars, 500 pm. All data is presented as mean ± s.d.

[0090] FIGURE 15. IFlowPlate Duo with apical flow and use in IBD modelling, a, Image of the IFlowPlate Duo with basolaterally perfused compartments labelled with green dye and apically perfused compartments labelled with red dye. The plate is tilted 15° in the z-axis as well as rotated 45° in the x-y plane to achieve both apical and basolateral flow. It is secured in place with a holder on a programmable rocker, b, Quantification of 24 hour glucose consumption in different culture platforms (n = 5 - 10;****P<0.0001). c, Quantification of the apical velocity profile, d, Experimental timeline, e, Summary of cytokine gradients and upregulation before and after inflammation for avascular as well as vascularized colon epithelium with patterned crypts, f, Cytokine heatmap for all experimental conditions, g, Quantification of individual cytokine release profiles (n = 3 - 5; *P<0.05 **P<0.01 ***p<0.001). All data is presented as mean ± s.d.

[0091] FIGURE 16. IFlowPlate Duo dimensions, a, Top-down (left) and oblique (right) views, b, Bottom- up (left) and oblique (right) views, c, Side view. All dimensions are in millimetres. FIGURE 17. Device setup and flow measurements in UniPlate 24. a) Schematic overview of cancer microenvironments incorporating circulating monocytes and subsequent analysis protocols in the vascularized cancer spheroid model, b) Image of programmable rocker and UniPlate24 containing 8 independent units. Each unit is composed of three wells and a slanted bridge connecting the inlet well and the outlet well, c) Schematic illustration of rocker settings for one perfusion cycle with unidirectional recirculating flow, d) Quantification of blood vessel diameters in six different plates. n= 8 for each plate. Statistical significance was determined using one-way ANOVA. “ns” indicates not statistically significant, e) Flow rate changes in the printed vascular channel during the 5 min forward perfusion. The flow rate was measured every 30 s. n=6 tissues for plate A, and n=8 tissues for plate B and C. f, g) Quantification of average flow rate (e) and average wall shear stress (1) in the 5 min forward perfusion. n=6 tissues for plate A, and n=8 tissues for plate B and C. Statistical significance was determined using one-way ANOVA. “ns” indicates not statistically significant.

[0092] FIGURE 18. Blood vessels with continuous recirculating flow, a) Schematic illustration of vascular tissues with continuous recirculating flow, b) Stitched confocal maximum projected z-stack fluorescence images and 3D-reconstituted confocal fluorescence image of a blood vessel surrounded by fibroblasts and pericytes under recirculating flow. The cells were stained with VE-Cadherin (green), F- actin (red), and DAPI (blue). Scale bar: 200 pm. c) Permeability quantification of accellular vascular channel and vascular tissues with or without fibroblasts and pericytes. n=6 tissues for each group. Statistical significance was determined using one-way ANOVA with the Holm-Sidak method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. d) Immunofluorescent staining of blood vessels with or without fibroblasts and pericytes. The cells were stained with VE-Cadherin (green), F-actin (red), and DAPI (blue). Scale bar: 200 pm.

[0093] FIGURE 19. Design, dimension and operation of the printed stamp, a) The .stl file and the dimension of the stamp for incorporating spheroids, b) Image of a printed stamp made with high temperature resin, c) Image of a stamp applied to the plate unit. The stamp matches perfectly with the tissue well of the UniPlate.

[0094] FIGURE 20. Immune-infiltrated cancer spheroid model with continuous monocyte recirculation through a blood vessel, a) Schematic illustration of vascularized tumor model comprising a blood vessel, stromal cells and immune components within the tumor microenvironment, b) Schematic illustration of engineering a vascularized tumor model using a printed stamp, c) Experimental protocol for long term immune cells recirculation in our UniPlate devices, divided into two stages: formation of vascularized tumor tissue and long-term immune cell recirculation, d) Image of a vascularized tumor tissue with GFP- HUVECs (green) forming the tubular blood vessel, and fibroblasts and pericytes embedded in the gel matrix. Scale bar: 1 mm. e) Representative images of cell tracker -labeled monocytes (red) within blood vessels (green) under one-time monocyte perfusion or continuous monocyte recirculation over time. The edges of blood vessels are outlined with white lines. Scale bar: 1 mm. 1) Representative images of red cell tracker-labeled monocytes (red) in tumors under one-time monocyte perfusion or continuous monocyte recirculation over time. The dashed circle indicates the location of the cancer spheroid. Scale bar: 500 pm. g, h) Quantification of fluorescence signals (red) in blood vessels (g) and tumors (h) with one-time monocyte perfusion or continuous monocyte recirculation over time. n=8 tissues were quantified for fluorescence signals in blood vessels, and n=15 tumor sites were quantified for fluorescence signals in tumors.

[0095] FIGURE 21. Tissue-specific macrophage recruitment from the blood vessel to the cancer microenvironment, a) Representative images of fluorescence signals in three types of spheroid models: fibroblasts, Huh7 cells, and MDA-MB-231 cells under continuous monocyte recirculation over time. Red cell tracker-labeled monocytes were perfused in tissues and replenished daily. The dashed circle indicates the location of the cancer spheroids. Scale bar: 1 mm. b, c) Quantitative analysis of fluorescence signals in blood vessels (c) and in tissue spheroids (d) with 7 days of continuous monocyte recirculation. The inset figures show monocyte signals in three experimental groups after 3 days of recirculation. n=6 tissues were quantified for fluorescence signals in blood vessels. n=15 tumor sites from 6 tissues were quantified for fluorescence signals in tumors, d-g) Secretion levels of representative pro-inflammatory cytokines, including GM-CSF (d), IFN y (e), IL-6 (f), and IL-8 (g) in collected media perfusates from the three tissuespecific groups. n=4 tissues for each group. Statistical significance was determined using one-way ANOVA or Kruskal-Wallis one-way ANOVA on ranks with the Holm-Sidak method method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. “ns” indicates not statistically significant, h-j) Secretion levels of representative MMP cytokines, including MMP-3 (h), MMP-9 (i), and TIMP-4 (j) in collected media perfusates from the three tissue-specific groups. n=4 tissues for each group. Statistical significance was determined using one-way ANOVA or Kruskal-Wallis one-way ANOVA on ranks with the Holm-Sidak method method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. “ns” indicates not statistically significant, k) Histological section of immune-infiltrated cancer spheroid tissues after 7 days of continuous monocyte recirculation with fibroblast, Huh7 and MDA-MB-231 spheroids stained for hematoxylin and eosin (H&E, nuclei stains are blue; ECM and cytoplasm stains are pink), CD 68, and CD 163 (dark brown). The spheroids are outlined with white dotted circles. Scale bar: 100 pm. 1) Vessel sprouting in an MDA- MB-231 cancer spheroid model on day 2 (before monocyte perfusion) and day 7 after co-culturing with cancer spheroids. The cell tracker-labelled monocytes (red) were introduced on day 2. Scale bar: 1 mm. m) Permeability changes in the blood vessel in the MDA-MB-231 tumor model over coculture time. TRITC- dextran (65 KDa) was perfused through blood vessels for permeability analysis. n=8 tissues for each group. Statistical significance was determined using one-way with the Holm-Sidak method method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. “ns” indicates not statistically significant.

[0096] FIGURE 22. Secretion levels of all analyzed inflammatory -related cytokines, including IL-IRa, IL-1 , IL-2, IL-4, IL-12p40, IL-12p70, IL-13, MCP-1, TNF-a, and MMPs cytokines, including MMP-2, MMP-7, MMP-8, MMP-10, MMP-12, MMP-13, TIMP-1, TIMP-2, TIMP-3 in collected media perfusates from three cancer spheroid models with fibroblasts, MDA-MB-231, or Huh7 spheroids. n=4 for each condition. Statistical significance was determined using one-way ANOVA or Kruskal-Wallis one-way ANOVA on ranks with the Holm-Sidak method method. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. “ns” indicates not statistically significant.

[0097] DETAILED DESCRIPTION

[0098] In aspects, the present disclosure describes simple-to-use microphysiological cell culture systems in the format of conventional multi-well plates for fabricating, culturing, perfusing and optionally testing multiple vascularized microenvironments within an “open-top” well connected by channels to an inlet and outlet well. The formed vascular barrier guides fluid perfusion in the vascular lumen from an inlet well to an outlet well while the interstitial space outside of the vascular network can be accessed from the top of the well. Without the physical constraints of traditional microfluidic channels, micro-tissue spheroids and / or organoids of various sizes and quantities can be incorporated and vascularized by the microvascular bed formed in the well. The large array of perfusable vascular networks can be tested in a high-throughput fashion or easily extracted from the wells for use outside the cell culture device, such as for in vivo transplantation.

[0099] In specific aspects, described herein is a platform that recirculates fluid in a substantially unidirectional manner without the use of a pump. This is typically gravity -driven and replicates physiological shear stress without the requirement of a pump to drive fluid flow.

[0100] In other specific aspects, described herein is a cell culture model for epithelial structures that form crypts, such as in the intestinal epithelium, glandular structures, hair follicles, or lung alveoli for example. The crypts are formed so as to have apical access and an underlying perfusable microvasculature. A stamp for imprinting the crypts is also described and is useful in high throughput methods. Similarly, an open well comprising both an apically perfused compartment and a basolaterally perfused compartment is described. These compartments are in aspects separate and distinct from one another, the way they would be in a physiological environment, encouraging physiologically relevant structures.

[0101] I, Definitions

[0102] Unless otherwise indicated, the definitions and embodiments described in this and other sections are intended to be applicable to all embodiments and aspects of the present disclosure herein described for which they are suitable as would be understood by a person skilled in the art.

[0103] In understanding the scope of the present disclosure, the term “comprising” and its derivatives, as used herein, are intended to be open ended terms that specify the presence of the stated features, elements, components, groups, integers, and / or steps, but do not exclude the presence of other unstated features, elements, components, groups, integers and / or steps. The foregoing also applies to words having similar meanings such as the terms, “including”, “having” and their derivatives. The term “consisting” and its derivatives, as used herein, are intended to be closed terms that specify the presence of the stated features, elements, components, groups, integers, and / or steps, but exclude the presence of other unstated features, elements, components, groups, integers and / or steps. The term “consisting essentially of’, as used herein, is intended to specify the presence of the stated features, elements, components, groups, integers, and / or steps as well as those that do not materially affect the basic and novel characteristic(s) of features, elements, components, groups, integers, and / or steps.

[0104] Terms of degree such as “substantially”, “about” and “approximately” as used herein mean a reasonable amount of deviation of the modified term such that the end result is not significantly changed. These terms of degree should be construed as including a deviation of at least ±5% of the modified term if this deviation would not negate the meaning of the word it modifies.

[0105] As used in this disclosure, the singular forms “a”, “an” and “the” include plural references unless the content clearly dictates otherwise.

[0106] In embodiments comprising an “additional” or “second” component, the second component as used herein is chemically different from the other components or first component. A “third” component is different from the other, first, and second components, and further enumerated or “additional” components are similarly different.

[0107] The term “and / or” as used herein means that the listed items are present, or used, individually or in combination. In effect, this term means that “at least one of’ or “one or more” of the listed items is used or present.

[0108] The term “perfusable” as used herein refers to an interconnected network or series of channels through which a liquid medium can flow or circulate, for example a perfusable vascular network that provides oxygen and nutrients to a three-dimensional cell culture.

[0109] The term “vascular network” or “vascular bed” as used herein refers to a complex network of blood vessels that supplies oxygen and other nutrients to the tissue or organ.

[0110] The term “organoid” as used herein refers to complex three-dimensional structures derived from stem cells that can represent multiple cell lineages that can mimic the complex organ physiology.

[0111] The term “spheroid” as used herein refers to three-dimensional structures derived from cell lines which represents a specific tissue component depending on the source of the cell.

[0112] The term “crypt” as used herein refers to an epithelial structure that is three dimensional and forms cavernous structures. Examples of crypts include but are not limited to the intestinal epithelium, glands, such as mammary glands, salivary glands, sweat glands, etc., the alveoli in the lungs, and hair follicles.

[0113] II. Cell Culture Systems and Methods of the Disclosure

[0114] Disclosed herein is a pumpless recirculating platform for substantially unidirectional flow. By “substantially unidirectional,” it will be understood that the fluid flows primarily in a singular direction, from an inflow, to an outflow, and back to the inflow, and so on. It is contemplated that there could be some small level of backflow limited to up to about 15% of the total fluid per cycle, such as up to about 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, or 15% and any ranges therein, which would then remix with the circulating fluid and continue cycling.

[0115] The platform is described as being pumpless, meaning that there is no pump actively driving the fluid to flow continuously. In typical aspects, the fluid is gravity -driven, through use of a rocker, so that the motion of the rocker causes the fluid to move back and forth around the platform. Due to the structure of the platform, as will be described, fluid is encouraged to flow mostly in a single direction and backflow is discouraged. Typically, the rocker can be programmed to have specific dwell times at specific tilt angles in order to control the length of time and angle used to drive the fluid flow. These parameters can be adjusted to arrive at a desired shear stress and / or fluid flow rate.

[0116] For example, typically, the desired shear stress is from about 0.7 to about 2 dyne / cm2, such as from about 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, or 1.9 to about 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, or 2.0 dynes / cm2 or any value or range thereinbetween. Likewise, typically the desired fluid flow rate is from about 100 to 200 pL / min, such as from about 105, 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, or 195 to about 110, 115, 120, 125, 130, 135, 140, 145, 150, 155, 160, 165, 170, 175, 180, 185, 190, 195, or about 200 pL / min, or any value or range thereinbetween.

[0117] The platform is typically constructed so as to comprise an inlet well, a tissue well, and an outlet well, all of which are in substantially unidirectional fluid communication with one another, such that the fluid flows from the inlet well to the tissue well, then to the outlet well, then back to the inlet well. Typically the inlet well and the tissue well are in fluid communication with one another through a first channel and the tissue well and the outlet well are in fluid communication with one another through a second channel.

[0118] In order to control the unidirectional flow of fluid, a slanted bridge connects the outlet well to the inlet well. The slanted bridge is lower at the outlet well than at the inlet well. In this way, tilting of the platform in a first direction, such that the inlet well is higher than the outlet well, results in fluid flow from the inlet well into the tissue well and then to the outlet well. Because the slanted bridge is higher at the inlet well, fluid can substantially only pass through the first channel from the inlet well into the tissue well and then through the second channel into the outlet well and cannot pass from the inlet well directly to the outlet well via the slanted bridge.

[0119] Likewise, when the platform is tilted in a second direction, such that the inlet well is lower than the outlet well, most fluid will follow the route of least resistance and flow out of the outlet well via the slanted bridge into the inlet well. There may be a minor amount of backflow at this point, but this is small because it is easier for fluid to flow over the slanted bridge than it is to backflow through the second channel.

[0120] The slanted bridge can have any desired angle to accomplish this flow and this can be varied depending upon the length of the bridge and the desired tilt angle of the rocker driving the fluid flow. In typical aspects, the slanted bridge has an angle of from about 5° to about 45°, such as from about 5°, 10°, 15°, 20°, 25°, 30°, 35°, or 40° to about 10°, 15°, 20°, 25°, 30°, 35°, 40°, or 45°, or any single value or range thereinbetween. Typically, the slanted bridge has an angle of about 20°.

[0121] Typically, the tissue well is bottomless and, optionally, the inlet well and / or the outlet well are also bottomless. The bottomless platform allows for a template for cell seeding to be provided on the bottom surface of the well(s). It will be understood that the wells need not be bottomless and a template could instead be provided on the bottom of the wells. The template in aspects is sacrificial and is 3D printed onto the surface of a sheet that can be attached to the bottom of the wells. For example, the sheet may be adhesive and the sacrificial material may be any such material that is dissolvable, for example, at 37°C. For example, the sacrificial material typically comprises gelatin, pluronic acid, alginate, or any combination thereof. In aspects, the sacrificial material, once dissolved, leaves behind one or more channels between the inlet well and the outlet well. In particular aspects, the one or more channels comprise vascular channels for seeding of endothelial cells.

[0122] The platform is typically placed on a plate and, in aspects, a plurality of such platforms can be placed on a single plate for a high throughput system. For example, the plate in aspects holds from about 8 platforms to about 128 platforms in standard 24 to 384 well format.

[0123] Typically, the platform is provided alone or in a system or a kit comprising one or more of the aforementioned components. In further aspects, the platform is provided in combination with a substrate upon which to seed cells, such as fibrin gel, collagen, Matrigel, an extracellular matrix (ECM) hydrogel, or any combination thereof. Further, the platform may be provided in combination with cells, such as cells that form tubular tissues. Examples of tubular tissues include but are not limited to vascular tissues or renal proximal tubules. Cells that form these tissues include endothelial cells or kidney cells for example. Other examples of cells include spheroids, such as cancer spheroids. Combinations of cells, such as heterospheroids, are also contemplated. Further, the platform may be provided in combination, alternatively or additionally, with cells such as immune cells, endothelial cells, cancer cells, epithelial cells, stromal cells, or any combination thereof.

[0124] Also described are methods of making such platforms. The methods typically comprise applying a sacrificial template to an injection -molded bottomless well.

[0125] Also described are methods of perfusing cells using the platforms described herein. The methods comprise seeding cells on the platform, typically in the tissue well, and then recirculating fluid substantially unidirectionally. As described, this recirculation is typically accomplished without the use of a pump. Typically, the recirculation is accomplished through gravity, which is applied to the cells by tilting a rocker at certain tilt angles so as to drive fluid flow as described. A recirculation tilt angle will cause fluid to flow from the outlet well to the inlet well via the slanted bridge and a perfusion tilt angle will cause fluid to flow from the inlet well through the tissue well and into the outlet well via the first and second channels. Typically, a tilt angle of from about -20° to about -60°, such as from about -20°, -25°, -30°, -35°, or -40° to about -25°, -30°, -35°, -40°, or -45°, such as about -45°, is used for a recirculation tilt angle. Typically, a tilt angle of from about 5° to about 45°, such as from about 5°, 10°, 15°, 20°, 25°, 30°, 35°, or 40° to about 10°, 15°, 20°, 25°, 30°, 35°, 40°, or 45° such as about 15°, is used for a perfusion tilt angle.

[0126] In some aspects, cancer spheroids are seeded on the platform adjacent to a vascular tube, through which the fluid is recirculated. The fluid typically comprises immune cells.

[0127] Also described herein is an epithelium model. The model has patterned crypts and underlying perfusable micro vasculature. Typically, the crypts are provided in an open well so that access to the apical surface of the epithelium is provided. The perfusable microvasculature is typically vascularized with endothelial cells and, optionally, fibroblasts. Due to the construction of the crypts and the underlying perfusion, the endothelial cells and optional fibroblasts will typically self-assemble, thereby providing a tissue microenvironment with vascular-epithelial crosstalk.

[0128] While typically the model is a colon epithelium model, it will be understood that the model can be used for other types of tissues comprising crypt-like structures where there is an apical and basolateral surface, such as other parts of the intestine, alveolar tissues, glandular tissues, hair follicles, etc.

[0129] Typically, the patterned crypts are imprinted on a hydrogel scaffold, which is typically crosslinked. The hydrogel may be made from any biocompatible material but typically comprise a fibrin matrix, fibrin, Matrigel, collagen I, a decellularized matrix, or a combination thereof.

[0130] Thus, also provided is a stamp for imprinting the patterned crypt on the crosslinked hydrogel scaffold. Conventional processes typically involve impressing a patterned stamp into a hydrogel and crosslinking the hydrogel while the stamp is in place. It has been found herein that the stamp can be imprinted directly into the crosslinked hydrogel very rapidly so that the process can be carried out in a high throughput manner. By using an already-crosslinked hydrogel, the process is much faster and more efficient.

[0131] The stamp can be made by any method but is typically 3D printed. The stamp is typically made from high temperature resin that is capable of being autoclaved or otherwise sterilized. The stamp typically comprises high aspect ratio features, such as pyramidal pin structures printed at the tip of the stamp. It will be understood that the stamp can be designed to have any desired features that resemble the type of crypt that is being represented in the model in question.

[0132] Typically, the stamp surface is functionalized to reduce adherence of the stamp to the scaffold. For example, the stamp surface is in aspects polyethylene glycol (PEG) photografted. In additional or alternative aspects, the stamp surface is coated with an anti-adherence rinsing solution.

[0133] Thus also provided are methods of patterning a crosslinked hydrogel. The methods comprise applying the stamp described herein to the already -crosslinked hydrogel and then removing the stamp.

[0134] Also described herein is an immune -infiltrated cancer spheroid model. The model is typically made by continuously perfusing and recirculating immune cells with gravity -driven flow through a tubular blood vessel adjacent to a cancer spheroid.

[0135] Also described herein is a cancer-macrophage heterospheroid model comprising a macrophage population that continuously grows, renews, and / or self-evolves.

[0136] Also described herein is an immune -infiltrated cancer spheroid model comprising a vascular tube through which immune cells unidirectionally flow and a cancer spheroid adjacent to the vascular tube.

[0137] In aspects, the model further comprises stromal cells and / or a cytokine gradient.

[0138] Also disclosed herein is an open chamber for cell culture. The chamber is open and is not considered a closed system. Typically, the chamber is open at the top of the chamber, allowing access to the cells being cultmed therein. A removable covering such as a lid may be included in order to reduce evaporation or a lid may be explicitly excluded from the open chamber described herein. The chamber is not sealed and is open and accessible to the operator.

[0139] For example, the chamber is typically an open cylinder. In aspects, the chamber comprises an open top and a closed bottom, where the open top and closed bottom have substantially equal diameters. The closed bottom may be an integral part of the chamber or it may be provided by means of a separate base. For example, the chamber may be provided as a well in a multi-well plate, where the plate may be open- or closed-bottomed and optionally provided with a base to enclose the bottom end of the chamber.

[0140] In typical aspects, the chamber comprises a basolaterally perfused compartment and an apically perfused compartment and both compartments are typically independent of one another and can have distinct perfusion solutions that do not intermix. Typically, the flow of fluid through both the basolaterally perfused compartment and the apically perfused compartment is gravity -driven, substantially as described above, such that a pump is not necessarily required.

[0141] Typically, the basolaterally perfused compartment comprises a basolateral inlet and a basolateral outlet for fluid flow. In typical aspects, the basolateral inlet leads to a basolateral inlet chamber and / or the basolateral outlet leads to a basolateral outlet chamber, which are both independently open or closed.

[0142] Likewise, typically the apically perfused compartment comprises an apical inlet and an apical outlet for fluid flow. The apical inlet typically leads to an apical inlet chamber and / or the apical outlet leads to an apical outlet chamber.

[0143] While flow of fluid in the compartments can be in any direction, typically flow of fluid in the basolaterally perfused compartment is substantially orthogonal to the flow of fluid in the apically perfused compartment.

[0144] Typically, the apically perfused compartment is open for access to the apical compartment and, typically, a hydrogel separates the basolaterally perfused compartment and the apically perfused compartment. In some specific aspects, the use of a membrane is excluded herein for separating the basolaterally perfused compartment and the apically perfused compartment. In other words, in some aspects, these compartments are not separated by a membrane, such as a synthetic membrane.

[0145] In typical aspects, the open chamber comprises a crypt-patterned base, as described herein, which separates the basolaterally perfused compartment and the apically perfused compartment.

[0146] Also described herein is a platform. The platform comprises a central open chamber for cell culture. A basolateral inlet chamber and a basolateral outlet chamber each adjacent to and in fluid communication with the open chamber are provided. Similarly, an apical inlet chamber and an apical outlet chamber each adjacent to and in fluid communication with the open chamber are provided. However, the basolateral inlet and outlet chambers are not in fluid communication with the apical inlet and outlet chambers, such that both communicate independently with the open chamber, the apical inlet and outlet chambers being in fluid communication with the apically perfused compartment above seeded cells and the basolateral inlet and outlet chambers being in fluid communication with the basolaterally perfused compartment below perfused cells. In this way, the platform is useful for modeling cells that grow in vivo with an apical surface and a basolateral surface. Due to the basolateral flow, vasculature can be modeled under the epithelium.

[0147] In some aspects, the platform comprises a second apical inlet chamber adjacent to and in fluid communication with the apical inlet chamber and a second apical outlet chamber adjacent to and in fluid communication with the apical outlet chamber.

[0148] In aspects, the open chamber comprises a crypt-patterned base, separating the apical and basolateral compartments, as described herein.

[0149] Also described herein is a multi-well plate comprising the open chamber or platform described herein. The multi-well plate may comprise any number of wells. In some embodiments, the multi-well plate comprises 6 wells, 12 wells, 24 wells, 96 wells, 384 wells, or 1536 wells. In some embodiments, the multi-well plate is a 384-well plate. In some embodiments, the multi-well plate has a cap to cover the opentop wells.

[0150] In some embodiments, the open chambers or platforms described herein may be used as a high- throughput drug testing platform. In some embodiments, the open chambers or platforms are used to improve the experimental throughputs of organ-on-a-chip and / or organoid systems. In some embodiments, the open chambers or platforms are used in vitro for research and development. In some embodiments, the perfusable network is used in vivo for cell therapy. In some embodiments, the device allows tissue extraction for applications such as, but not limited to, downstream analysis (e.g. gene expression analysis, histopathological analysis, etc.) or implantation for regenerative therapy.

[0151] In some embodiments, the seeded cells include cells such as hepatocytes, pancreatic Islet cells, fibroblasts, chondrocytes, osteoblasts, endothelial cells, exocrine cells, smooth or skeletal muscle cells, myocytes, adipocytes, ectodermal cells, ductile cells, kidney cells, intestinal cells, parathyroid cells, thyroid cells, nerve cells, ocular cells, integumentary cells, pluripotent cells and stem cells, or combinations thereof.

[0152] EXAMPLES

[0153] The following non-limiting examples are illustrative of the present disclosure:

[0154] EXAMPLE 1, PUMP-LESS, RECIRCULATING PLATFORM ENABLES UNIDIRECTIONAL PERFUSION OF 3D PRINTED TUBULAR TISSUES

[0155] Abstract

[0156] The direction and pattern of fluid flow affect vascular structure and function, in which vessellining endothelial cells exhibit variable cellular morphologies and vessel remodeling by mechanosensing. To recapitulate this microenvironment, some approaches have been reported to successfully apply unidirectional flow on endothelial cells in organ-on-a-chip systems. However, these platforms either require bulky pumps or are limited to 2D monolayer cell cultures. Herein, we demonstrate a pumpless recirculating platform with an open-well design (UniPlate) that enables unidirectional media recirculation through 3D printed tubular tissue constructs on a programmable rocker. To ensure that the platform is scalable for industrial manufacturing, the device was made of polystyrene via injection molding in combination with 3D printed gelatin that serves as a sacrificial template. We first engineered tubular blood vessels with unidirectional perfusion. Then we expanded the design to incorporate duo -recirculating flow for culturing vascularized renal proximal tubules with glucose reabsorption function. In addition to media recirculation, human monocyte recirculation in engineered blood vessels was also demonstrated for over 24 h, with minimal loss of cells, cell viability, and inflammatory activation. UniPlate can be a valuable tool to more precisely control the cellular microenvironment of organ-on-a-chip systems for drug discovery.

[0157] Introduction

[0158] Blood vessels are routinely exposed to dynamic mechanical stimuli, including shear stress exerted by blood flow and cyclic strain from pulsatile pressure generated by cardiac contraction. Blood flow remains a significant factor in regulating the structure and function of vascular tissues.1For instance, in the presence of luminal shear stress, endothelial cells tend to stay quiescent and form tighter barriers, while vessel sprouting is activated when the interstitial flow is predominant.2The permeability of blood-brain barriers was found to decrease when the vessels are exposed to pulsatile pressure.3In addition to blood vessels, fluid flow also plays an important role in regulating tubular epithelial tissues. In renal proximal tubules, luminal perfusion was reported to facilitate the formation of microvilli brush border, which is a key cellular apparatus for sensing fluid flow.4High fluid shear stress has been shown to accelerate cell differentiation and apoptosis at the tips of intestinal villi in 3D small intestinal models, thereby reproducing the cell turnover process on native intestinal.5

[0159] In addition to the presence of flow, the specific direction of flow has also been documented to affect cell morphology and inflammatory signaling activation in endothelial cells.6, 7Under unidirectional perfusion, endothelial cells preferentially display an elongated morphology and align in the flow direction compared to the bidirectional flow condition.8Unidirectional flow is also beneficial for reducing mitochondrial reactive oxygen species production and protecting the mitochondrial DNA integrity of endothelial cells.9Anatomically, blood flow in the human body is unidirectional and re -circulatory. Nutrient-rich blood is transported and distributed to all tissues by arteries, and blood with metabolic byproducts is returned to the heart via the veins. The liver lobule has a specific hierarchical anatomy with blood flowing from the hepatic artery to the portal vein unidirectionally, which leads to the formation of multiple gradients, such as nutrients, oxygen and hormones.10Under the stimulation of unidirectional flow, vascularization of kidney organoids was enhanced with increasing vessel coverage, vessel junctional density, and vessel length, which further promoted kidney organoids matmation.11On the contrary, bidirectional flow is found to preferentially present in regions of atherosclerotic lesions with activated endothelial nitric oxide synthase signaling, which is a high-risk factor in many cardiovascular events related to endothelial dysfunction and atherosclerosis.12, 13

[0160] Existing approaches established to realize unidirectional flow in organ-on-chip models tend to use a syringe or peristaltic pump to perfuse and circulate media.14 16However, bulky supporting systems, tedious tubing connections, the risk of trapping bubbles, and the presence of dead volumes in tubing often limit the usability and scalability of such systems. To overcome problems derived from pumps and tubing, gravity -driven pumpless organ-on-a-chip devices with unidirectional flow are attractive alternatives.17 19The basic principle is to introduce a bypass channel connecting the inlet and outlet reservoirs for media recirculation.20By elaborately designing connections between the reservoirs, Shuler et al. demonstrated a microfluidic Unichip device to obtain continuous unidirectional flow on a rocking system.21Similarly, unidirectional flow can be achieved with a rotating platform.22However, these existing platforms are generally limited to 2D monolayer cell cultures in closed microfluidic systems. Closed microfluidic systems are usually PDMS-based and do not integrate well with techniques, such as 3D bioprinting, to produce more complex tissue structures. Moreover, immune cell circulation is yet to be demonstrated in gravity -driven perfusion platforms with recirculating flow. Therefore, a pumpless recirculating system that supports 3D bioprinted tissues and recirculation of immune cells in an open-well format is much needed.

[0161] In this study, we developed a pumpless recirculating platform that enabled unidirectional fluid flow through 3D printed vascular / tubular tissues. Unidirectional flow and recirculation were controlled by a rocker with a programmable dwell time at different tilt angles. We demonstrated for the first time the formation of 3D tubular blood vessels with different geometries and vascularized renal proximal tubules with active glucose transport functions under a unidirectional recirculating flow driven by gravity. This platform can also recirculate immune cells for at least 24 h with minimal activation, making it suitable for investigating immunological surveillance in diseased blood vessels and distal tissues.

[0162] Results

[0163] UniPlate were fabricated with injection molding and 3D printing

[0164] We developed a pumpless recirculating platform in the form of a standard 24-well plate that supports recirculating unidirectional perfusion of 3D printed vascular / tubular tissues on a programmable rocking system. The platform is modular and partially injection-molded from polystyrene, which is the industry -standard material for microtiter plates, thus minimizing the nonselective drug absorption issues associated with traditional PDMS-microfluidic devices.23Each UniPlate contains eight fully assembled devices secured on a customized frame. Each device was made of two components: a customized bottomless well and a thin pressure-sensitive adhesive sheet with 3D printed gelatin sacrificial template. The device was fabricated using injection molding and 3D printing (Figure la). Customized bottomless wells made of polystyrene were manufactured by injection molding, and bioprinting was used to produce the gelatin template on pressure-sensitive adhesive sheets. The assembled device contained three separate wells (inlet, tissue, and outlet wells) and a slanted bridge that was directly connected to the inlet and outlet wells (Figure lb). The slanted bridge was designed to slant at 20° angle. The three wells were connected by two microchannels (400 pm x 400 pm) (Figure 2). These channels serve as inlets and outlets to the tissue well where the 3D printed tissues will be contained. One UniPlate can hold eight devices independently removable from the plate for analysis (Figure 1c). To produce perfusable tissues, a subtractive manufacturing method was adopted with some modifications.24Specifically, gelatin was selected as the sacrificial material and could be printed with different patterns on the pressure-sensitive adhesive sheet using an extrusion-based bioprinter. The device was then assembled by hand-pressing the bottomless wells with pressure-sensitive adhesive sheets containing 3D printed gelatin. Parts of the printed gelatin template were designed to align with the inlet and outlet channels after assembly to establish flow connections. To produce the tissues, fibrin gel was cast over the gelatin template inside the tissue wells (Figure 3a). After gelation, gelatin can be dissolved at 37 °C, leaving an open channel that can be populated with endothelial or epithelial cells. Native blood vessels and epithelial tubules have diverse geometries. For example, the uterine spiral artery and proximal tubule vessels are curved and tortuous.4, 25To demonstrate the flexibility offered by the extrusion printing process in tissue customization, we produced two designs to model vessels with linear and tortuous shapes, termed I and S channels, respectively (Figure 3b).

[0165] For unidirectional perfusion and recirculation, the plate was placed on a programmable rocker (Figure 4a). When the rocker was tilted at a positive angle (e.g., 15°), the media level in the inlet well was higher than that in the outlet wells; thus, the media flowed through the tubular channel in the fibrin gel to the outlet wells (Figure 3c). The media can only perfuse through channels embedded inside the fibrin gel and accumulate slowly in outlet wells because the slanted bridge allows only one-way perfusion from the outlet to inlet wells. After 5 min of forward perfusion, the rocker stage was flipped to a negative tilt angle (e.g., -40°) to allow the accumulated media in the outlet wells to flow back to the inlet wells immediately via the slanted bridge, which offers a much lower flow resistance than the tubular channel inside the fibrin gel. The rocker held this tilt position for only 10 s, which is more than sufficient to allow media recirculation. Finally, the rocker changed the tilt direction again and recovered to the original positive tilt angle to start a new cycle. Red food dye was added to the inlet wells to better visualize the complete recirculation cycle on UniPlate.

[0166] Rocker tilt angle regulates flow rate and residual volume in UniPlate

[0167] The flow rates in the single -channel design were evaluated by collecting and measuring the volume of media perfusate from the inlet to the outlet. The flow rates can be adjusted from 100 to 200 |iL / min by increasing with the tilt angle (Figure 5a). The calculated wall shear stress within the tubular blood vessels ranges from 0.7 to 2 dyne / cm2when the tilt angle changes from 10° to 30° (Figure 5b), which is comparable with that of venous in vivo (1 - 4 dyne / cm2).26With the connecting bridge, the majority of media in outlet wells could flow back to inlet wells immediately. However, some media was retained in the outlet wells. To maximize recirculation, we evaluated the residual media volume in the outlet wells at different recirculation tilt angles. The minimum residual media volume was achieved at - 40° tilt angle, which is the largest accessible tilt angle for the rocker. Approximately 100 |iL of media, 10% of the initial volume, were not recirculated (Figure 5c). Changing the initial media volume in the outlet well did not affect the volume of residual media (Figure 5d, Figure 4b). It is important to note that the residual media will mix with the media from the next perfusion cycle and eventually circulate back to the inlet wells. However, to achieve the highest recirculation efficiency, - 40° tilt angle was chosen as the recirculation tilt angle, whereas the perfusion tilt angle was set to 15°.

[0168] Tubular blood vessels can be perfused unidirectionally in UniPlate

[0169] To produce a blood vessel model with unidirectional flow, GFP -expressing human umbilical vein endothelial cells (GFP-HUVECs) were seeded in tubular channels embedded in fibrin gel. After seven days of culture, endothelial cells proliferated and formed a confluent lumen under both unidirectional and bidirectional conditions (Figure 5e). Engineered blood vessels showed an open lumen surrounded by endothelial cells (Figure 5h, Figure 4c). To evaluate the vessel permeability, TRITC-conjugated dextran (65 kDa) was added to the inlet wells. The blood vessels formed under unidirectional flow showed a permeability of 3.5 x 10-6cm / s, which is comparable to those of native blood vessels (0.15 x 10-6cm / s) and other reported blood vessel chip models (4.1 x 10-6cm / s) (Figure 5f, g).27, 28The permeabilities of the blood vessels under unidirectional and bidirectional flows were also compared and surprisingly showed no significant difference under our current experimental condition.

[0170] Cell nuclei elongation was evaluated by comparing nuclei circularity and orientation of endothelial cells in blood vessels cultured under unidirectional and bidirectional conditions. However, there was no significant difference in cell nuclei circularity under unidirectional and bidirectional conditions (Figure 5i), indicating that endothelial cells in 3D tubular blood vessels do not show apparent cell elongation under unidirectional or bidirectional flow under shear stress at approximately 1 dyne / cm2. There was also no significant difference in cell nucleus orientation between unidirectional and bidirectional conditions (Figure 5j). Endothelial cells tended to orientate at an angle ranging from - 75° to - 15° to the flow direction in both conditions. Unlike the 2D culture condition on closed PDMS / PMMA-based devices, we found that endothelial cells in 3D perfusable vascular tissues responding to unidirectional flow did not show significant differences in cell nuclei elongation and permeability compared to the bidirectional flow condition at the cellular level. These results could be explained by the differences in matrix stiffness between the fibrin gel (1 - 8 kPa) and substrates used to support 2D endothelial monolayer cultures (0.8- 10 MPa for PDMS) in prior studies.29-31Shear stress could also be a contributing factor for the difference in observation. In a previous study, the enclosed PDMS / PMMA devices had a higher shear stress in the range of 5 -10 dyne / cm2which could promote endothelial cell alignment.32Furthermore, mechanical stimuli in 3D blood vessels are much more complex than those in 2D endothelium. In 3D, the endothelium was also exposed to transendothelial interstitial flow into the gel, which could have an impact on the cells. The underlying matrix could also be sufficiently soft to deform under flow.33The impact of these additional mechanical factors on cellular morphology must also be considered in this 3D model. For instance, we noticed that vessel sprouting could be induced in our model by adjusting the dwell time of the rocker from 5 to 15 min (Figure 4d). Based on the current flow rate, it takes about 8 min for 1 m of culture media to flow into the outlet well. In the 15 -min settings, as the media drains from the inlet wells, the flow rate and fluid shear stress will reduce significantly over time, which makes transendothelial interstitial flow from the interstitial space more dominant over time. Previous studies have demonstrated that interstitial flow can induce vessel sprouting, which is consistent with our observations.2Therefore, the 5-min setting helped to maintain continuous luminal flow, which suppresses vascular sprouting.

[0171] Vascularized renal proximal tubule with recirculating flow demonstrated glucose reabsorption function

[0172] Next, we expanded the single -channel UniPlate design to incorporate two individually perfusable networks inside the same gel matrix to reproduce a vascular-tubular network that emulated vascularized proximal tubule complexes in the kidney (Figure 6a). The two independent channels were perfused with two sets of inlet and outlet wells, and two slanted bridges for recirculation. Blood vessels and proximal tubule tissues were produced by seeding endothelial and renal proximal tubule epithelial cells into their corresponding channels. The media were perfused unidirectionally through both the blood vessels and proximal tubule and visualized with red and blue dyes (Figure 6b). We found that vascular and proximal tubule networks formed confluent tubular structures after four days of culture (Figure 6c), and both blood vessels and proximal tubule networks demonstrated a high barrier function that confines large proteins (Figure 6d). Similar to the single -channel tubular vessel, a 3D luminal structure of the vessel and proximal tubule was observed. VE-cadherin staining of the blood vessels showed continuous intercellular junctions, indicating the formation of a tight endothelial barrier.34

[0173] Glucose reabsorption is one of the primary functions of the renal proximal tubule, which transports glucose from the proximal tubule to the neighboring blood vessels. In this process, sodium-glucose cotransporters are involved in the assistance of Na / K ATPase to maintain osmotic equilibrium and membrane potential in cells.35The proximal tubule contains the glucose transporters SGLT2 and Na / K ATPase (Figure 6f). To model glucose transport, both the blood vessel and proximal tubule were treated with a medium containing approximately 100 mg / dL glucose, which is in the range of normal blood glucose levels for non-diabetics. After 24 h of unidirectional flow, the glucose levels in perfusates from each channel were measured. Compared to the initial concentration of glucose in fresh media, perfusates from the blood vessels showed an increase in glucose concentration with reduced glucose levels in the proximal tubule (Figure 6g), indicating the active transport of glucose from the proximal tubule to blood vessels. Moreover, 3D vascularized proximal tubule tissues with or without endothelial cells exhibit similar glucose levels in the vessel perfusates, indicating that the proximal tubule is the dominant factor in reabsorbing glucose, while the blood vessel passively takes up glucose from the interstitial space, which is consistent with previously published studies.36

[0174] Immune cells can recirculate on UniPlate

[0175] Blood is a multicomponent mixture containing fluids, plasma, blood cells, and platelets, and is an important part of any physiological model that involves blood vessels. The long-term recirculation of whole blood or circulating immune cells in engineered blood vessel models using gravity -driven flow remains challenging because of cell sedimentation issues in long-term cultures. Cell settlement and aggregation occur when the cells are cultured under relatively static conditions with minimal convective mixing. However, with excessive agitation, inflammatory activation of circulating immune cells may also occur owing to mechanically induced stress. To investigate the feasibility of using UniPlate for re -circulating immune cells, human monocytes (THP-1) were recirculated for 24 h on the platform. Cell number and viability were evaluated by live / dead staining, and media perfusates were collected for cytokine analysis (Figure 7a). Under static conditions, as expected, most circulating monocytes were found to move in the vessel channel at 0 h but completely settled to the bottom of the inlet / outlet wells after 24 h of static culture (Figure 7b). The cells left inside the blood vessels were also found to be static. Compared to the static culture group, no apparent difference was detected in the recirculation group at 0 h, and all cells moved from the inlet well to the outlet well through the vessel channel. However, after 24 h of recirculation, many cells continued to flow inside the blood vessel, and cells in the inlet and outlet wells remained suspended in the medium with good dispersion. These results demonstrate the feasibility of monocyte recirculation in UniPlate. Furthermore, after 24 h of recirculation, the viability of monocytes was comparable to that of cells at 0 h (Figure 7c, d). The overall cell numbers in the perfusates collected from the static and recirculation groups were comparable, indicating no significant cell loss during recirculation (Figure 7e). However, a more even cell distribution in the inlet and outlet wells was observed in the recirculation group, with an increased cell number in the outlet wells and a reduced cell number in the inlet wells, indicating that monocytes were successfully perfused, recirculated, and continuously mixed in the media throughout the entire device (Figure 7f). To investigate whether circulating monocytes were activated by mechanical stress during recirculation, we analyzed cytokine secretion. As expected, we did not observe any significant changes in the secretion levels of the inflammatory cytokines (Figure 7g). Specifically, GM-CSF and MCP-1 are key cytokines responsible for recruiting circulating monocytes to sites of inflammation. Both static and recirculation groups showed similar secretion levels of GM-CSF lower than 0.6 pg / mL (Figure 7h). Interestingly, the secretion of MCP-1 (Figure 7i) was significantly higher under static conditions than under recirculation conditions, indicating that there was a certain level of inflammation in the culture, likely from endothelial cells, after 24 h of static culture. There was no significant difference in the secretion levels of other key pro-inflammatory cytokines such as IL-6, IL-8, TNF-a, and IFN-y (Figure 8), indicating that monocytes were not significantly activated by recirculation flow. Overall, these results indicate that UniPlate supports monocyte recirculation.

[0176] Discussion

[0177] Microphysiological systems aim to recapitulate the key features of human tissue functions by precisely controlling how individual cells interact with each other and the surrounding microenvironment. Among the many different types of interactions between cells and their microenvironments, dynamic fluid flow is often an important factor affecting cellular behavior and function.37’38In this study, we demonstrated a pumpless recirculating platform that can produce a long-term recirculating unidirectional flow of culture media and circulating immune cells in 3D tubular tissues, such as blood vessels and renal proximal tubules, embedded in fibrin gels. Here, we achieved unidirectional flow with a programmable rocker and customized well-plate design. This gravity -driven perfusion approach eliminates the issues of tubing connectivity, air bubbles, and dead volume in conventional microfluidic fluid control systems. Platforms that adopt gravity -driven flows are highly scalable. Each plate can contain multiple devices and can be stacked on the rocker to increase the experimental throughput. The dimensions and layout of the UniPlate were designed to match those of standard commercial 24-well plates, making it compatible with commercial plate readers and standard microscope imaging setups. We selected a 24-well plate format owing to the physical constraints of our benchtop injection molder. However, in future work, based on the same design principle, the platform can be scaled up to a 384-well plate format or other alternate formats to further increase the experimental throughput and reduce the reagent usage per tissue.

[0178] The fabrication of UniPlate is cost-effective, scalable, and compatible with 3D extrusion bioprinting techniques, which can be used to easily customize the tissue structures within the device. In this study, we demonstrated only very simple 3D printed structures as a proof-of-concept. However, in future studies, much more complex 3D tissue structures can be printed using gelatin, alginate, or Phironic acid as sacrificial templates.24, 39To further increase the model complexity, stromal cells, spheroids, or organoids could be embedded inside or placed on top of the gel.40As stromal cells and parenchymal tissues are incorporated, future studies could explore the formation of oxygen and nutrient gradients that spontaneously arise from unidirectional perfusion. However, built-in oxygen sensors may be required to detect such gradients directly. The body of the device was made of polystyrene using the injection molding method, which is consistent with the manufacture of industry-standard microtiter plates. This device eliminates the nonselective drug absorption issue in PDMS devices, which can absorb up to 90% of hydrophobic molecules in solution.23

[0179] An important advancement of this platform is the continuous recirculation of circulating immune cells in engineered tubular tissue models. The circulating monocytes were maintained in suspension and recirculated. As they were perfused through engineered blood vessels, each monocyte would have passed through the blood vessels and interacted with the endothelium, on average, 250 times in 24 h. Previously, monocyte perfusion experiments were almost always performed by perfusing monocytes once through a tissue within a short period of time. Without sufficient exposure time, the extent of monocyte recruitment and the immune response can be severely underestimated. However, continuous recirculation allows monocytes and other immune cells to accumulate and interact with injured tissues, making it possible to model chronic tissue inflammation, repair and regeneration.41Monocytes and circulating immune cells are crucial components in immunological surveillance, and more accurately recapitulating monocyte recirculation in engineered tissues will provide better guidance in disease modeling, clinical disease treatment and drug discovery.42

[0180] Conclusions

[0181] In summary, we developed a pumpless recirculating platform that allows unidirectional flow through 3D printed vascular / tubular tissues. Single-channel and duo-channel designs were created to model blood vessels and vascularized proximal tubules with unidirectional recirculating flow driven by gravity for the first time. Additionally, monocytes were successfully perfused and recirculated for at least 24 h without impairing cell viability or triggering inflammatory activation, indicating the potential of this platform for culturing circulating immune cells in a complex organ-on-a-chip environment under unidirectional recirculating flow. This platform opens up possibilities for modeling and studying chronic inflammatory diseases by prolonging the interaction of circulating immune cells with endothelium and epithelial tissues. Ultimately, this improved our ability to better control the microenvironment in microphy siological systems for pathological modeling and drug discovery.

[0182] Materials and Methods

[0183] Unidirectional plates fabrication

[0184] All designs of customized bottomless wells and gelatin patterns were created using the AutoCAD software. The master molds for injection molding were printed using a Formlab printer (Formlabs, Form3B) following the supplier’s instructions. Considering the high working temperature of the injection molding process, high-temperature resin (Formlabs, cat. # RS-F2-HTAM-02) was selected as the master mold. After printing, two post-processing steps were performed to completely remove the non-crosslinked resin. Briefly, the print was immersed in isopropanol (ACS reagent, >99.5%, Sigma-Aldrich, cat# 190764- 4L) and sonicated for 1 h. Next, the print was transferred to a UV chamber (Formlabs, FH-CU-01) and cured by UV exposure (light wavelength: 405 nm) for 2.5 h at 65 °C. The printed master molds were ready for use. A benchtop injection-molding machine (Galomb, Model S-100) was used to manufacture the customized bottomless wells (Figure 2). The working temperature of the injection molder was set to 237 °C which is the maximum temperature that the heating unit allows. Ease release 205 (Sculpture Supply Canada) was applied to the master mold for the easy release of injection-molded parts. Next, polystyrene resin pellets (Galomb, cat# RSN-003-1) with a melting point at 221-232 °C were loaded to the cartridge. Once the polystyrene pellets became molten plastic in approximately 3 min, the handle was manually pulled down. In this process, molten polystyrene is pressed into master molds. After pressing for 20 s, the parts were released from the mold. The plastic parts were then soaked in isopropanol for Ih to removing the residual release reagent on the surface.

[0185] To fabricate the base of the device, 40% (w / v) gelatin (Sigma-Aldrich, cat# G9391-100G) was printed on a pressure-sensitive adhesive sheet (Strouse, cat# 3M-9795R) using an extrusion-based bioprinter (Cellink™, BIO X). The bioprinter was set to work under a working pressure of 15 KPa and printing speed of 5 mm / s using a 27-gauge (inner diameter:0.2 mm) needle. With the bioprinter, gelatin fibers with around 200 |im diameter were patterned on the adhesive sheets. The adhesive sheets were airdried for an hour after gelatin printing. Before combining the adhesive sheet with the injection-molded bottomless wells, the bottom side of the bottomless well was treated with a corona treater (Electro -Technic Products Inc., Model BD-20) for 2 min. The pressure-sensitive adhesive sheet with printed gelatin was hand-pressed onto bottomless wells. The pressure-sensitive adhesive becomes sticky only in the region that is exposed to compression forces. Therefore, this material does not affect the area inside the wells.

[0186] The entire process was repeated eight times to create eight devices. The devices were then assembled in a OneWell Plate (VWR, cat# 30617-434) and secured in place using a 3D printed frame. The frame was printed using a FormLab printer. Then, one full UniPlate device was packaged in a sealing bag and stored at 4 °C until use. All plates were sterilized using an ultraviolet chamber (Benchmark Scientific, UV Clave™) for 20 min before use in cell culture experiments. All the UniPlate devices were used only once.

[0187] Quantification of flow rate and shear stress

[0188] To calculate the volumetric flow rates (Q) through the blood vessel model, 700 |iL and 50 |il of PBS was added to the inlet and outlet wells respectively. Then the plate was placed on a rocker tilted at an angle (10°, 15°, 20°, 24°, and 30°). After 5 min, the media accumulated in the outlet wells were collected, and the media volume was measured. Wall shear stress (T) was calculated using equation (1):

[0189] The diameter of the tubular vessel was measured from brightfield images using ImageJ, assuming that the vessel has a circular cross-section. The viscosity of water at 25 °C r] = 8.9 X 10-3dyne ■ s / cm2) was used in calculation.

[0190] Cell culture

[0191] Green fluorescent protein-expressing human umbilical vein endothelial cells (GFP-HUVECs, Angio-Proteomie, cat# CAP-0001GFP) were cultured in Endothelial Cell Growth Medium 2 (ECGM2, Sigma-Aldrich, cat# C-22011) according to the manufacturer’s protocol. Human renal proximal tubular epithelial cells (RPTEC-TERT1, Everycyte, cat# CHT-003-0002) were cultured in ready -to-use ProxUp medium (Evercyte, cat# MHT-003). THP-1 monocyte cells were cultured in RPMI 1640 medium (Cedarlane Labs, cat# 30-2001) supplemented with 10% FBS and 2 -mercaptoethanol (0.05 mM) in nontreated multiple well plates. To prepare for further experiments, cells were plated in T75 cell culture flasks and grown in an incubator with 5% CO2 at 37 °C. The endothelial cells used in this study were between passages 3-6. Proximal tubule epithelial cells were cultured until passage eight.

[0192] Three-dimensional tissues formation and plates operation

[0193] Fibrin gel was prepared by mixing fibrinogen (Sigma Aldrich, Cat# F3879-1G) and thrombin (Sigma Aldrich Cat#, T6884- 100UN). A stock solution of thrombin was prepared at a concentration of 10 U / ml in 0.1% BSA (Sigma-Aldrich, Cat# A9205-10ML). Fibrin gel was prepared by mixing 100 pL of fibrinogen (10 mg / ml) and 20 pL of thrombin (10 U / ml) and then immediately cast into the tissue wells to encapsulate the gelatin template. After gelation at room temperature for 15 min, the plate was transferred to the incubator and incubated with PBS overnight at 37 °C. As gelatin has a melting point in the range of 31.7-34.2 °C, gelatin will melt and dissolve in PBS in the incubator. Thus, an open lumen was generated inside the hydrogel after aspirating the gelatin dissolved in PBS, followed by rinsing with the complete medium. To prevent fibrin gel degradation over time, all cell culture media were supplemented with 1% (v / v) aprotinin (Sigma Aldrich, Cat# 616370-100MG-M). The cell culture medium was changed daily.

[0194] For cell seeding, 200 |iL of cell suspension (2 million cells / ml) was added in both inlet and outlet wells, where the cells will flow into the tubular channels in fibrin gel. The plate was placed on a level surface in an incubator under static condition to allow cells attachment for Ih. After 1 h, the plate was moved to a rocker to initiate unidirectional or bidirectional perfusion. ECGM2 medium with 1% aprotinin was used to culture the vascular tissues. For the duo-channel design, endothelial cells and renal proximal tubule epithelial cells were seeded at the same time. Generally, 200 |iL of endothelial cells suspension and 200 |1L of RTPECs suspension were added in their corresponding inlet and outlet wells. ProxUp medium containing 1% aprotinin was used for the proximal tubule until it reached 100% confluency. The medium was then changed to a modified DMEM medium (no glucose, Thermo Fisher Scientific, Cat# 11966025) supplemented with 100 mg / dL D-(+)-glucose (Sigma-Aldrich, Cat# G5767-500G) and ProxUp supplements (Cat# MHT-003-S).

[0195] Permeability assay

[0196] The barrier function of the tubular blood vessel model was measured by perfusing a TRITC-labeled dextran solution (65 kDa, Sigma Aldrich, cat# T1162) through the vessel. Specifically, 300 |iL and 100 |iL of culture medium were added to the tissue well and outlet well respectively and 500 |iL of TRITC-labeled dextran (400 |ig / ml) solution was added in the inlet well. Time-lapse images were captured using a microplate reader (BioTek, Cytation 5) at 15 min intervals. From the time-lapse images, the permeability and Pd for each vascular tissue were calculated using Equation (2):

[0197] In the Equation (2), / / and / , represent average intensities at the final and initial time point, while h represents the average intensity for background. At is the time interval between images, and d is average diameter of vascular tissues.

[0198] Glucose transport studies

[0199] For the glucose transport experiment, ECGM2 media and modified DMEM were mixed at a 1:1 v / v ratio and then added to both vascular and proximal tubule inlet / outlet wells on day 7 after the proximal tubule reached full confluency. Media perfusates from the proximal tubule and vasculature outlets were collected after 24 h of perfusion. Glucose levels in all collected perfusates were measured using a glucose meter (Contour® Next One Meter).

[0200] Monocytes recirculation and cytokines analysis

[0201] To demonstrate the feasibility of recirculating suspended cells on UniPlate, 700 |iL of human monocytes (THP-1) suspension (1.5 million cells / mL) was added in the inlet well and 200 |1L of culture medium were added to the tissue and outlet wells. For the static condition, the plate was kept leveled in an incubator. For the recirculating condition, the plate was placed on the rocker. After 24 h, the culture media and circulating cells from both static and recirculating conditions were collected. To evaluate cell loss, cell numbers in both the inlet and outlet wells were counted using a hemocytometer. To evaluate the impact of mechanical stress from recirculation, the culture media was analyzed for inflammatory cytokines. Cytokine analysis was performed by Eve Technologies using the Human Cytokine Array Proinflammatory Focused 15-plex panel (cat# HDF15).

[0202] Immunostaining and live / dead staining Before fixation, all tissues were washed three times with fresh culture medium to remove dead cell debris. The tissues were then fixed with 4% (w / v) paraformaldehyde solution for 20 min on a rocker at room temperature. Next, the tissues were washed with PBS buffer three times and blocked using a blocking solution (10% FBS in PBS) for an hour on a rocker at room temperature. After removing the blocking solution, tissues were incubated with primary antibodies, anti-CD31 (Abeam, cat# ab9498), anti-VE- Cadherin (Abeam, cat# ab33168), anti-Sodium Potassium ATPase (Abeam, cat# ab76020), or anti-SGLT2 (Abeam, cat# ab85626) overnight at 4 °C. A blocking buffer containing 0.1% Triton X-100 was used to stain VE-cadherin, Na / K ATPase, and SGLT2. The tissues were washed with PBS three times to remove residual primary antibodies, followed by incubation with fluorochrome -conjugated secondary antibodies and DAPI (Sigma-Aldrich, cat# D9542) for 1 h at room temperature. All primary and secondary antibodies were diluted in phosphate -buffered saline (PBS) containing 2% FBS. The tissues were washed again with PBS three times prior to imaging using a confocal microscope (ZEISS, 3i Marianas LightSheet). Live / dead staining was performed using 5 -Carboxyfluorescein diacetate (CFDA, Sigma- Aldrich, cat# C4916-25MG) and propidium iodide solution (PI, Sigma-Aldrich cat# P4864-10ML), following the manufacturer’s instructions.

[0203] Statistical analysis

[0204] SigmaPlot software was used for testing the normality and equality of variance. One-way ANOVA or Kruskal-Wallis one-way ANOVA on ranks were used for determining statistical significance. Data in all graphs are plotted as mean with standard deviation (SD), and at least three independent samples (n > 3) were used in each condition to perform quantitative analysis.

[0205] References

[0206] 1. Dessalles, C. A.; Ramon-Lozano, C.; Babataheri, A.; Barakat, A. I., Luminal flow actuation generates coupled shear and strain in a microvessel-on-chip. Biofabrication 2021, 14 (1).

[0207] 2. Song, J. W.; Munn, L. L., Fluid forces control endothelial sprouting. Proc Natl Acad Sci U SA 2011, 108 (37), 15342-7.

[0208] 3. Partyka, P. P.; Godsey, G. A.; Galie, J. R.; Kosciuk, M. C.; Acharya, N. K.; Nagele, R. G.; Galie, P. A., Mechanical stress regulates transport in a compliant 3D model of the blood-brain barrier. Biomaterials 2017, 115, 30-39.

[0209] 4. Homan, K. A.; Kolesky, D. B.; Skylar-Scott, M. A.; Herrmann, J.; Obuobi, H.; Moisan, A.; Lewis, J. A., Bioprinting of 3D Convoluted Renal Proximal Tubules on Perfusable Chips. Scientific Reports 2016, 6 (1).

[0210] 5. Costello, C. M.; Phillipsen, M. B.; Hartmanis, L. M.; Kwasnica, M. A.; Chen, V.; Hackam, D.; Chang, M. W.; Bentley, W. E.; March, J. C., Microscale Bioreactors for in situ characterization of GI epithelial cell physiology. Sci Rep 2017, 7 (1), 12515.

[0211] 6. Chiu, J. J.; Chien, S., Effects of disturbed flow on vascular endothelium: pathophysiological basis and clinical perspectives. Physiol Rev 2011, 91 (1), 327-87. 7. Wang, C.; Baker, B. M.; Chen, C. S.; Schwartz, M. A., Endothelial cell sensing of flow direction. Arterioscler Thromb Vase Biol 2013, 33 (9), 2130-6.

[0212] 8. Boonyaphon, K.; Li, Z.; Kim, S.-L, Gravity -driven preprogrammed microfluidic recirculation system for parallel biosensing of cell behaviors. Analytica Chimica Acta 2022, 1233.

[0213] 9. Shin, J.; Hong, S. G.; Choi, S. Y.; Rath, M. E.; Saredy, J.; Jovin, D. G.; Sayoc, J.; Park, H. S.; Eguchi, S.; Rizzo, V.; Scalia, R.; Wang, H.; Houser, S. R.; Park, J. Y., Flow-induced endothelial mitochondrial remodeling mitigates mitochondrial reactive oxygen species production and promotes mitochondrial DNA integrity in a p53-dependent manner. Redox Biol 2022, 50, 102252.

[0214] 10. Trefts, E.; Gannon, M.; Wasserman, D. H., The liver. Curr Biol 2017, 27 (21), R1147-R1151.

[0215] 11. Homan, K. A.; Gupta, N.; Kroll, K. T.; Kolesky, D. B.; Skylar-Scott, M.; Miyoshi, T.; Mau, D.; Valerius, M. T.; Ferrante, T.; Bonventre, J. V.; Lewis, J. A.; Morizane, R., Flow-enhanced vascularization and maturation of kidney organoids in vitro. Nat Methods 2019, 16 (3), 255-262.

[0216] 12. Niu, N.; Xu, S.; Xu, Y.; Little, P. J.; Jin, Z. G., Targeting Mechanosensitive Transcription Factors in Atherosclerosis. Trends Pharmacol Sci 2019, 40 (4), 253-266.

[0217] 13. Simmons, R. D.; Kumar, S.; Jo, H., The role of endothelial mechanosensitive genes in atherosclerosis and omics approaches. Arch Biochem Biophys 2016, 591, 111-31.

[0218] 14. Herland, A.; Maoz, B. M.; Das, D.; Somayaji, M. R.; Prantil-Baun, R.; Novak, R.; Cronce, M.; Huffstater, T.; Jeanty, S. S. F.; Ingram, M.; Chalkiadaki, A.; Benson Chou, D.; Marquez, S.; Delahanty, A.; Jalili-Firoozinezhad, S.; Milton, Y.; Sontheimer-Phelps, A.; Swenor, B.; Levy, O.; Parker, K. K.; Przekwas, A.; Ingber, D. E., Quantitative prediction of human pharmacokinetic responses to drugs via fluidically coupled vascularized organ chips. Nat Biomed Eng 2020, 4 (4), 421-436.

[0219] 15. Schneider, S.; Bubeck, M.; Rogal, J.; Weener, H. J.; Rojas, C.; Weiss, M.; Heymann, M.; van der Meer, A. D.; Loskill, P., Peristaltic on-chip pump for tunable media circulation and whole blood perfusion in PDMS-free organ-on-chip and Organ-Disc systems. Lab Chip 2021, 21 (20), 3963-3978.

[0220] 16. Sung, J. H.; Wang, Y. L; Narasimhan Sriram, N.; Jackson, M.; Long, C.; Hickman, J. J.; Shuler, M. L., Recent Advances in Body-on-a-Chip Systems. Anal Chem 2019, 91 (1), 330-351.

[0221] 17. Sung, J. H.; Kam, C.; Shuler, M. L., A microfluidic device for a pharmacokinetic - pharmacodynamic (PK-PD) model on a chip. Lab Chip 2010, 10 (4), 446-55.

[0222] 18. Chen, Y.; Chan, H. N.; Michael, S. A.; Shen, Y.; Chen, Y.; Tian, Q.; Huang, L.; Wu, H., A microfluidic circulatory system integrated with capillary-assisted pressure sensors. Lab Chip 2017, 77 (4), 653-662.

[0223] 19. Lee, H.; Kim, D. S.; Ha, S. K.; Choi, L; Lee, J. M.; Sung, J. H., A pumpless multi-organ-on-a- chip (MOC) combined with a pharmacokinetic-pharmacodynamic (PK-PD) model. Biotechnol Bioeng 2017, 114 (2), 432-443. 20. Lee, D. W.; Choi, N.; Sung, J. H., A microfluidic chip with gravity -induced unidirectional flow for perfusion cell culture. Biotechnol Prog 2019, 35 (1), e2701.

[0224] 21. Wang, Y. I.; Shuler, M. L., UniChip enables long-term recirculating unidirectional perfusion with gravity -driven flow for microphysiological systems. Lab Chip 2018, 18 (17), 2563-2574.

[0225] 22. Yang, Y.; Fathi, P.; Holland, G.; Pan, D.; Wang, N. S.; Esch, M. B., Pumpless microfluidic devices for generating healthy and diseased endothelia. Lab Chip 2019, 19 (19), 3212-3219.

[0226] 23. Toepke, M. W.; Beebe, D. J., PDMS absorption of small molecules and consequences in microfluidic applications. Lab Chip 2006, 6 (12), 1484-6.

[0227] 24. Rajasekar, S.; Lin, D. S. Y.; Zhang, F.; Sotra, A.; Boshart, A.; Clotet-Freixas, S.; Liu, A.; Hirota, J. A.; Ogawa, S.; Konvalinka, A.; Zhang, B., Subtractive manufacturing with swelling induced stochastic folding of sacrificial materials for fabricating complex perfusable tissues in multi-well plates. Lab Chip 2022, 22 (10), 1929-1942.

[0228] 25. Cui, Y.; Wang, W.; Dong, N.; Lou, J.; Srinivasan, D. K.; Cheng, W.; Huang, X.; Liu, M.; Fang, C.; Peng, J.; Chen, S.; Wu, S.; Liu, Z.; Dong, L.; Zhou, Y.; Wu, Q., Role of corin in trophoblast invasion and uterine spiral artery remodelling in pregnancy. Nature 2012, 484 (7393), 246-50.

[0229] 26. Gray, K. M.; Stroka, K. M., Vascular endothelial cell mechanosensing: New insights gained from biomimetic microfluidic models. Semin Cell Dev Biol 2017, 77, 106-117.

[0230] 27. Yuan, W.; Lv, Y.; Zeng, M.; Fu, B. M., Non-invasive measurement of solute permeability in cerebral microvessels of the rat. Microvasc Res 2009, 77 (2), 166-73.

[0231] 28. Zheng, Y.; Chen, J.; Craven, M.; Choi, N. W.; Totorica, S.; Diaz-Santana, A.; Kermani, P.; Hempstead, B.; Fischbach-Teschl, C.; Lopez, J. A.; Stroock, A. D., In vitro microvessels for the study of angiogenesis and thrombosis. Proc Natl Acad Sci U SA 2012, 109 (24), 9342-7.

[0232] 29. Kim, J. S.; Kim, T. H.; Kang, D. L.; Baek, S. Y.; Lee, Y.; Koh, Y. G.; Kim, Y. L, Chondrogenic differentiation of human ASCs by stiffness control in 3D fibrin hydrogel. Biochem Biophys Res Commun 2020, 522 (1), 213-219.

[0233] 30. Seghir, R.; Arscott. S., Extended PDMS stiffness range for flexible systems. Sensors and Actuators A: Physical 2015, 230, 33-39.

[0234] 31. Chakraborty, A.; Chakraborty, S.; Jala, V. R.; Haribabu, B.; Sharp, M. K.; Berson, R. E., Effects of biaxial oscillatory shear stress on endothelial cell proliferation and morphology. Biotechnol Bioeng 2012, 109 (3), 695-707.

[0235] 32. Dewey, C. F., Jr.; Bussolari, S. R.; Gimbrone, M. A., Jr.; Davies, P. F., The dynamic response of vascular endothelial cells to fluid shear stress. J Biomech Eng 1981, 103 (3), 177-85.

[0236] 33. Dessalles, C. A.; Leclech, C.; Castagnino, A.; Barakat, A. L, Integration of substrate- and flow- derived stresses in endothelial cell mechanobiology. Commun Biol 2021, 4 (1), 764.

[0237] 34. Vestweber, D., VE-cadherin: the major endothelial adhesion molecule controlling cellular junctions and blood vessel formation. Arterioscler Thromb Vase Biol 2008, 28 (2), 223-32. 35. Ferrannini, E., Sodium-Glucose Co-transporters and Their Inhibition: Clinical Physiology. Cell Metah 2017, 26 (1), 27-38.

[0238] 36. Lin, N. Y. C.; Homan, K. A.; Robinson, S. S.; Kolesky, D. B.; Duarte, N.; Moisan, A.; Lewis, J. A., Renal reabsorption in 3D vascularized proximal tubule models. Proc Natl Acad Sci U SA 2019, 776 (12), 5399-5404.

[0239] 37. Leung, C. M.; de Haan, P.; Ronaldson-Bouchard, K.; Kim, G.-A.; Ko, J.; Rho, H. S.; Chen, Z.; Habibovic, P.; Jeon, N. L.; Takayama, S.; Shuler, M. L.; Vunjak-Novakovic, G.; Frey, O.; Verpoorte, E.; Toh, Y.-C., A guide to the organ-on-a-chip. Nature Reviews Methods Primers 2022, 2 (1).

[0240] 38. Zheng, F.; Fu, F.; Cheng, Y.; Wang, C.; Zhao, Y.; Gu, Z., Organ-on-a-Chip Systems: Microengineering to Biomimic Living Systems. Small 2016, 72 (17), 2253-82.

[0241] 39. Kolesky, D. B.; Truby, R. L.; Gladman, A. S.; Busbee, T. A.; Homan, K. A.; Lewis, J. A., 3D bioprinting of vascularized, heterogeneous cell-laden tissue constructs. Adv Mater 2014, 26 (19), 3124-30.

[0242] 40. Rajasekar, S.; Lin, D. S. Y.; Abdul, L.; Liu, A.; Sotra, A.; Zhang, F.; Zhang, B., IFlowPlate-A Customized 384-Well Plate for the Culture of Perfusable Vascularized Colon Organoids. Adv Mater 2020, e2002974.

[0243] 41. Hohsfield, L. A.; Ammann. C. G.; Humpel, C., Inflammatory status of transmigrating primary rat monocytes in a novel perfusion model simulating blood flow. J Neuroimmunol 2013, 258 (1-2), 17-26.

[0244] 42. Tatham, K. C.; O'Dea, K. P.; Romano, R.; Donaldson, H. E.; Wakabayashi, K.; Patel, B. V.; Thakuria, L.; Simon, A. R.; Sarathchandra, P.; Harefield, P. i.; Marczin, N.; Takata, M., Intravascular donor monocytes play a central role in lung transplant ischaemia-reperfusion injury. Thorax 2018, 73 (4), 350-360.

[0245] EXAMPLE 2. COLON EPITHELIUM BARRIER WITH VASCULARIZED CRYPTS TO MODEL INFLAMMATORY BOWEL DISEASE

[0246] Abstract

[0247] Inflammatory bowel disease (IBD) is a complex, debilitating illness primarily affecting the colon. To better study the disease and develop treatments, engineered in vitro models with colon-specific physiological features are needed. Current colon models lack integration of colonic crypt structures with underlying perfusable vasculature, where vascular-epithelial crosstalk is affected by disease progression. We present a colon epithelium barrier model with vascularized crypts that recapitulates relevant cytokine gradients in both healthy and diseased conditions. Using our previously published IFlowPlate384 platform, we first imprinted crypt topography and populated the patterned scaffold with colon cells. Proliferative colon cells spontaneously localized to the crypt niche while differentiating into epithelial barriers with a tight brush border. Toxicity of the chemotherapeutic drug, capecitabine, was tested and showed a dosedependent response and recovery from crypt-patterned colon epithelium only. Perfusable microvasculature was then incorporated around the colon crypts followed by treatment with pro-inflammatory TNFa and IFNy cytokines to model IBD conditions. We found in vivo-like stromal basal-to-apical cytokine gradients in tissues with vascularized crypts and gradient reversals upon inflammation. Taken together, we demonstrated crypt topography integrated with underlying perfusable microvasculature has significant value for emulating colon physiology and in advanced disease modelling.

[0248] Introduction

[0249] The colon performs several vital functions including absorption, secretion, motility, and digestion of luminal compounds. The importance of these roles makes the colon a major target for pharmaceutical drug screening and disease modeling1 2. The colon epithelium is a single columnar layered cell sheet that serves as a selective and protective barrier3-4. This epithelial barrier curves inward to present a series of organized invaginations called crypts5. Self-renewal of the colon epithelium during homeostasis is made possible by the crypts housing resident LGR5+ adult stem cells that migrate unidirectionally up the crypt to proliferative transit-amplifying cells1-6-7. Colon cell turnover is the highest in the body as differentiated cells shed into the lumen every 4-5 days2. Colonic crypt structures vary between individuals but are approximately 500 pm deep, 125 pm wide with approximately 200 pm spacing between crypts8. The epithelial polarity with proliferative cells in crypts and differentiated quiescent cells on the luminal surface is maintained by in vivo endogenous growth signal gradients5. Stromal crosstalk from cell types like vasculature and fibroblasts at the base of the crypt introduce factors such as Notch, R-spondin, bone morphogenic protein (BMP), prostaglandins, Hedgehog, and Wnt-3A that span the length of the crypt axis to balance stem cell self-renewal and differentiation7. Perturbations to the tightly regulated crypt microenvironment can lead to pathologies like inflammatory bowel disease (IBD)9. IBD primarily affects the colon epithelium and disrupts barrier function causing a “leaky gut” through the release of inflammatory cytokines10. Over time, the inflammation can also lead to structural changes in the colon, such as strictures (narrowing of the colon) or fistulas (abnormal connections between the colon and other organs), which can result in complications such as bowel obstruction or infection11. The exact causes of IBD are not fully understood, but it is thought to be related to a combination of exogenous and endogenous factors.

[0250] Current animal models of the human colon present many differences in anatomy, expression of intestinal drug metabolizing enzymes and transporters, cellular composition, and immune response. These factors ultimately compromise transferability of preclinical results to human clinical trial responses1213. To better emulate human physiology, recent studies have focused on human colon organoids derived from primary colon resident stem cells or isolated primary crypts seeded into hydrogel scaffolds such as Matrigel domes14 18. These multicellular constructs, usually 100-200 pm in diameter, can recapitulate in vivo colon form and function through crypt budding, differentiation and cell renewal1214. Organoids also conserve donor phenotype with human specific drug metabolizing enzymes and transporters12. However, the formation and dimensions of colon organoids can be highly stochastic19. Moreover, enclosed organoid lumens prevent apical access and limits transport studies, as drug addition is exclusively delivered to the basolateral cell-matrix interface20. Most colon organoid cultures are also static, with no apical or basal flow, and lack vasculature with supporting stromal cells18.

[0251] Recent advances in microfluidic technology and 3D bioprinting now allow for more precise control over the colon tissue geometry and microenvironment1 21 23. Several groups to date have successfully biofabricated colon tissue with organoids or Caco-2 cell lines to form differentiated monolayers emulating in vivo colonic epithelium. Wang et al. cultivated primary colon organoid derived tissues on crypt- patterned transwell inserts24. The work demonstrates localization of stem cells to crypt bases with migrating cells differentiating at the apical surface, as seen in-vivo. Nikolaev et al. 3D bioprinted crypt-patterned colon tubules with apical perfusion24. They found innate cell-fate patterning along crypts and demonstrated wound healing response to dextran sodium sulfate that would normally cause organoids to collapse. Patterning in their model not only instructed cell localization, but also influenced tissue response to external stimuli. Taebnia et al. 3D bioprinted micropillar arrays as villi mimetics for Caco-2 culture25. Their results demonstrated a similar propensity for colon cells to recognize villi architecture and polarize along the villus axis into microvilli expressing barriers. Interestingly, decreasing villi density produced less differentiated ZO-1 expressing monolayers at villi bases. The results revealed villi spacing could impact cell phenotype. A similar study by Rudolph et al., used inverse molding to create a scaffold with both crypts and villi26. The study also used Caco-2 cells to form patterned intestinal epithelium and demonstrated increased expression of enterocyte specific sucrase isomaltase marker in the patterned scaffold. Taken together, the recent works highlighted the ability for colon organoid and Caco-2 cells to recognize colon-specific patterning in scaffolds. However, one missing theme in these works is that no group developed a colon epithelium barrier model with both patterned crypts and underlying perfusable microvascular networks. Tissue-relevant crypt-patterning combined with stromal and vascular biochemical gradients along the crypt axis can significantly impact homeostasis and disease progression of colon tissues.

[0252] We report for the first time a colon epithelium model with patterned crypts and underlying perfusable microvasculature in a previously published high-throughput 384-wellplate (IFlowPlate384) for drug screening and disease modelling. Previously, we showed the ability to vascularize embedded colon organoids in the IFlowPlate38416. We improve upon this work by guiding colon cells to assemble in engineered crypt structures, within an open well that enables access to the apical surface of the colon epithelium. An array of colon epithelial tissues can be rapidly patterned with crypts by inserting a 3D printed stamp in fibrin gels. To demonstrate the scalability of our model, a robotic liquid handler cast 120 gels which were then imprinted with TopoStamps in less than 5 minutes. Planar and crypt-patterned colon barriers were treated with the chemotherapeutic drug capecitabine, and revealed a dose -dependent response only in the crypt-patterned group. The patterned colon crypts were vascularized with self-assembled endothelial cells and fibroblasts to yield a tissue microenvironment with vascular -epithelial crosstalk. The IFlowPlate384 platform was also modified to support pulsatile apical flow over the colon epithelium with basolateral perfusion through the vascular beds. Lastly, the dynamic model was treated with TNFa and IFNy to induce IBD conditions with elevated pro-inflammatory cytokine release profdes. We discovered the presence of vasculature set a more physiologically relevant cytokine gradient milieu before treatment and reversed the gradients post-treatment. The results implicate the importance of vasculature in IBD modelling.

[0253] Results

[0254] Colon epithelial barriers with proliferative crypts can be patterned with a TopoStamp in the IFlowPlate384

[0255] We sought to develop a model containing key features of the native colon in a high-throughput platform with apical and basolateral flow. This can be conceptualized as (1) developing a colon barrier on a naturally derived fibrin extracellular matrix (ECM), (2) imparting crypt topology, (3) co-culturing human umbilical vein endothelial cells (HUVECs) and fibroblasts (FBs) to create perfusable microvascular networks, and (4) engineering apical as well as basolateral flow to the vascularized colon tissues to model IBD in response to pro-inflammatory cytokine triggers (Figure 9a). To achieve this, we used a previously published customized 384-wellplate (IFlowPlate384) which contains 128 individual units, where each unit is made of three wells interconnected by two- microchannels (Figure 9b)16,27. This is described in WO 2021 / 223031, incorporated by reference herein in its entirety. The inlet and outlet wells serve as medium reservoirs that provide interstitial flow through the fibrin gel in the centre well by a hydrostatic pressure head when subjected to a 15° z-axis tilt on a programmable IFlowRocker. The centre well comprises a fibrin gel with colon cells seeded on the surface forming an epithelium barrier.

[0256] Patterned crypt structures have been shown to drive colon tissue maturation in vitro8,24,26,28. However, existing methods relying on PDMS molds or stereolithography printing are not scalable to 384 wellplate formats26,28. Our aim was to use a 3D printed stamp (TopoStamp) to rapidly imprint an array of hydrogel scaffolds inside microwells (Figure 10a). Because of the simple procedure and minimal manual manipulation involved in this method, it is suitable for the small well size of a 384-wellplate. TopoStamps were 3D printed with high temperature resin that is autoclavable. High aspect ratio features such as an array of pyramidal pin structures (—130 pm in diameter, 580 pm spacing) were printed at the tip of the stamp (Figure 10a). Untreated TopoStamps, however, were not suitable for micromolding because fibrin gel would adhere to the stamp and become distorted after imprinting. To overcome this issue, we functionalized the stamp surface using several methods (Figure 11). Silanization made the stamp surface more hydrophobic which caused increased adherence to the gel. Air plasma treatment through temporary hydroxyl deposition made the surface more hydrophilic but was not effective enough to prevent adhesion29,30. Both polyethylene glycol (PEG) photografting and anti-adherence rinsing solution coating produced intact crypt-patterned gels. PEG photografting involved more complex and time-consuming steps, so we finalized on using simple antiadherence coating solution.

[0257] Caco-2 cells were seeded on the patterned fibrin gel surface as they are known to spontaneously differentiate into mature enterocyte brush border epithelium once confluent31. Cell seeding density on the gel surface and media formulation were optimized (Figure 12a-d). Confluent colon epithelium barriers were formed in 7 days in all medium formulations and are fully matured by 14 days (Figure lOd). Fibrin gel remodelling was minimal during culture. Patterned crypts remained hollow, and the surfaces were entirely populated with cells (Figure 10b). SEM confirmed the presence of characteristic brush border enterocytes with microvilli by day 14 (Figure 10c). Microvilli are present on both apical and basal surfaces of the crypt cross section, indicating mature Caco-2 tissues are spread throughout the length of the crypt-axis, as seen in other works26,32. Interestingly, we also observed fibrin gel compaction at the cell-matrix interface around the patterned crypts as opposed to the more porous matrix in the bulk gel. This is likely due to physical compression of the gel in the stamping process. Gel compaction could increase gel stiffness for the colon cells growing inside the crypts. Stiffer gels are also known to increase cell proliferation33. Next, we conducted permeability assays using 4 kDa and 65 kDa tracer dextran molecules (Figure lOe). On day 3 post-seeding, no barrier resistance to mass transfer was observed. As expected, 4 kDa dextran tracers diffused over two times faster than 65 kDa dextran molecules. On day 8, barrier formation was statistically different compared to fibrin-only controls, suggesting the formation of confluent colon epithelium in the IFlowPlate384. Dextran diffusion further decreased by day 14 indicating further maturation and development of epithelial junctions in both planar and crypt-patterned conditions. The data corroborates SEM and confocal imaging (Figure 10b,c) showing leak-tight barriers by day 8 post seeding and continual differentiation to day 14.

[0258] We then examined the impact of the crypt-pattern on the localization of proliferating cells. We compared planar and crypt-patterned tissues in the IFlowPlate384 with basolateral flow and crypt-patterned tissue in a standard 384 wellplate without interstitial perfusion. More proliferating cells were present in the crypt-patterned tissues than planar tissues in the IFlowPlate384 and these proliferating cells also tend to aggregate inside the crypts (Figure lOf-h). Surprisingly, static crypt-patterned tissue in a standard 384 wellplate with no basolateral flow showed few proliferative cells, suggesting the lack of basolateral medium supply to crypt bases could cause starvation and influence proliferation of the colonic epithelium. Native intestinal tissue comprises proliferative crypts aiding in high epithelium turnover6. These results showed the importance of both crypt-like niche and basolateral interstitial flow on the maintenance of proliferating cell niche inside crypts. Similar findings have been reported with Caco-2 derived crypt structures in a membranebased gut-on-chip device with basolateral flow proving crucial to crypt base sternness34.

[0259] Colon epithelium with patterned crypts reveal dose-dependent response and recovery to capecitabine treatment

[0260] To demonstrate our capability of scaling up tissue production using the IFlowPlate384 and TopoStamp, we used an automated robotic liquid handler to cast 120 fibrin hydrogels in the IFlowPlate384 (Figure 13a). After gelation, we used a TopoStamp to imprint 64 gels with crypt-topography in less than 5 minutes and left the remaining 64 gels as planar tissue controls. Following this procedure (Figure 13b), we showed that we can fabricate 120 tissues with over 96% success rate (Figure 13c). As expected, the presence of crypt structures introduced ~60% greater surface area leading to 2.5-5.0 times increased dextran diffusion rates in patterned tissues compared to planar tissues. However, there was also more variation in crypt- patterned colon tissue compared to planar tissue for dextran transport (Figure 13d), likely due to some variation introduced in the additional stamping step and the increase in surface area. Despite the larger variation, the permeability of the crypt-patterned barrier is still approximately ten-fold lower than gel-only controls which indicates significantly tight barriers in both planar and crypt-patterned conditions.

[0261] We examined the dose response and recovery from a chemotherapeutic prodrug, capecitabine to demonstrate the utility of the colon model for drug toxicity screening. Capecitabine has a mode of action of inhibiting DNA formation in proliferative cells and cancer cells leading to apoptosis (Figure 13e)35. Specifically, it is converted into 5 -fluorouracil (5-FU) in cancerous cells, which inhibits thymidylate synthase (TS)36. TS is an enzyme that plays a critical role in DNA synthesis by converting deoxyuridylate (dUMP) to deoxy thymidine monophosphate (dTMP) for the biosynthesis of thymine nucleic acids. Inhibition of TS leads to the arrest of DNA production, affecting cancerous and proliferating cells. Additionally, capecitabine is administered orally which is analogous to apical drug delivery36. We evaluated capecitabine apical drug dose response on both crypt-patterned and planar tissues by treating the tissues with the drug for three days followed by recovery for another three days (Figure 13b). We found dextran diffusion rates remained consistent throughout untreated, treated, and recovery phases for planar Caco-2 tissues. The unchanged diffusion rates revealed no quantifiable effect on the barrier integrity of the planar colon epithelium tissues from the drug treatments (Figure 13g). However, in crypt-patterned tissues, dextran transport showed a clear dose-response effect after three days of treatment, where barrier permeability remained higher at 0.1, 1 and 5 pM concentrations. After three days of recovery, all groups showed some sign of recovery, but barrier permeability still remained higher in the 1 and 5 pM conditions. The results are interesting, but not surprising since we know capecitabine preferentially impacts proliferating cells. Gonzalez-Dominguez et al. demonstrated capecitabine is toxic to proliferating Caco-2 cells starting at 2.8 pM dosage while having no effect on non-proliferative Caco-2 cells37. Our model suggests crypt-patterned tissues would be more sensitive to the drug since we showed crypt-patterning induced the formation of more proliferative cells that aggregate in crypts. This result highlights the importance of considering the crypt domain in drug toxicity screening, especially drugs that preferentially affect proliferating cells.

[0262] Crypt-patterned colon epithelium can be vascularized in the IFlowPlate384

[0263] Native colon tissue is highly vascularized, particularly around the colon crypts. Close proximity of stroma to crypts allows for bidirectional paracrine signalling1618. These biochemical signals form a gradient maintaining the crypt niche and sustaining colon homeostasis7. As vascularization of crypt-patterned epithelium has yet to be achieved, we sought to incorporate perfusable microvasculature around the patterned crypts in the IFlowPlate384. We adopted a previously published method by mixing HUVECs and FBs in fibrin gels (Figure 14a (1)). After gelation, crypt topography was then imprinted (Figure 14a(2)). Over the next five days, microvasculature self-assembled and anastomosed to the inlet / outlet channels. On day 5, colon epithelial cells were subsequently seeded on top of the fibrin gel in the centre wells. Vasculature perfusion was tested with 65 kDa TRITC dextran (TRITC-D) on day 10 (Figure 14a(3)). Once a confluent colon monolayer formed, Dulbecco's Modified Eagle Medium (DMEM) was compartmentalized to centre wells and Endothelial Cell Growth Medium 2 (EGM2) was compartmentalized to inlet and outlet wells. On day 14, epithelium barrier permeability was tested, and tissues were stained. The permeability assay showed significant colon barrier function, but the presence of vasculature did not significantly impact the overall mass transport (Figure 14b). As Caco-2 cells are known to form very tight intercellular junctions, the colon epithelium barrier is likely the dominating factor limiting mass transport. Not surprisingly, the overall mass transport of dextran is molecular weight dependent, where higher molecular weight (65 kDa) dextran showed a much slower transport rate. Histological cross-sections of vascularized colon crypts revealed a ring of polarized and differentiated colon cells with brush border microvilli protein on the apical surface (Figure 14c). Interestingly, the gel around the crypt circumference showed a darker purple hue from the hematoxylin dye compared to the bulk gel, indicating gel compaction consistent with our SEM images (Figure 10c). Paracrine signalling in epithelial tissue is known to occur with spatial separation of tens to hundreds of micrometers38,39. As evidenced by CD31+ vessel histological cross sections, the physical separation between vessels and crypts is less than 70 pm, indicating cell-cell crosstalk is likely to occur around the crypts. Notably, the self-assembled microvasculature near the bottom of the well can sprout towards the epithelialized crypts as visualized in 3D confocal images (Figure 14d), consistent with the histological crosssections. Overall, we showed evidence of close epithelia-stroma interactions resembling the colon in vivo microenvironment in the IFlowPlate384.

[0264] IFlowPlate Duo enabled apical flow on colon epithelium

[0265] The IFlowPlate384 allows basolateral perfusion but lacks apical flow31. Apical perfusion was introduced by modifying the IFlowPlate384 to join 4 additional reservoir wells with the centre well. The wall height between these connected wells was reduced to 5 mm for increased mixing (red dye, Figure 15a, Figure 16). Basolateral perfusion is still driven by hydrostatic pressure head generation from 15° z-axis tilting on a programmable IFlowRocker (green dye, Figure 15a). A 45° plate slant in the x-y plane was also introduced for apical flow between the centre wells and upper and lower reservoir wells (Figure 15a). The platform is named IFlowPlate Duo because of the presence of both basolateral and apical flow. This new configuration significantly increased apical media reservoir volume. The benefit of this volume increase is apparent when we quantified glucose consumption and found glucose was significantly depleted by 64.0 ± 4.87% in a 384-wellplate and 27.4 ± 8.17% in the IFlowPlate384 in just 24 hours, but with minimal depletion of 0.14 ± 2.36% in the IFlowPlate Duo. This is a valuable feature for future glucose transport studies that require a glucose gradient to be maintained for at least 24 hours. Glucose depletion is also associated with apoptosis or necrosis of human cells through activation of autophagy40. The IFlowPlate Duo could help to mitigate these glucose withdrawal stresses.

[0266] Apical flow is an important mechanical stimulus to colon epithelium that experience peristaltic movement in vivo. Caco-2 maturation through mucus production and spontaneous villi formation has been shown to be enhanced by apical flow32. We added 1 pm spherical beads to track flow profdes (Figure 15c) and observed pulsatile apical flow on the surface of the epithelium. Velocities pulse immediately after tilting, and rapidly decrease to a steady state flow. The maximum shear stress during the velocity pulse is ~1.3-1.9 dyne cm'2which can clear accumulated cell debris and waste removal. The luminal flow then remains slow at steady state with a shear stress range of ~0.03-0.08 dyne cm'2, which is in the range seen in native colon tissue of 0.002-0.08 dyne cm'2 ll 13. Low shear stress in these ranges alone have been reported to differentiate Caco-2 cells with increased cell height and polarization in less culture time11 13. The IFlowPlate duo can emulate in vivo physical stresses by applying both apical stimulus to epithelium in addition to basolateral perfusion.

[0267] Vasculature around colon crypts impacts cytokine gradients in healthy and IBD colon models

[0268] Cytokine gradients play a crucial role in both healthy and diseased colons. We sought to investigate changing cytokine gradients under both healthy and inflammatory conditions in colon tissues with and without vasculature. In colon tissue, epithelial (IL-1 p44) and stromal (IL-6, MCP-145, IL-4, IL-5, IL-8, IL-10, IL-12, IL-13, GM-CSF, IFNy44, IL-2, TNFa46, and IL-IRa47) cytokines are more concentrated in the lamina propria and diffuse down a concentration gradient to the epithelium. Our vascular model recapitulated IL- 1 , IL-4, IL-6, IL-8, IL-12p40, IL-13, GM-CSF, IFNy, and TNFa gradients compared to the avascular condition (Figure 15e-g). Healthy colon cytokine gradients are distinct in patients with IBD and contribute to the pathology by altering epithelium homeostasis48. We induced IBD conditions by treating the tissue with TNFa and IFNy pro-inflammatory cytokines (50 ng mL4) on the apical side as they are reported to be produced in significant quantities in inflamed colon tissue49. Stimulation with TNFa and IFNy upregulated more cytokines in the apical compartment in both the avascular and vascularized tissues. Avascular tissues present only apical-to-basal pro-inflammatory cytokines gradients (IL-1 P and IL-2) post-stimulation. Vascularized tissues reveal a wider range of pro-inflammatory cytokine gradients in both apical-to-basal (IL- 1 P, IL-6, IL-8, GM- CSF) and basal-to-apical (IL-12p70, IFNy, TNFa) directions. IL-ip, IL-6, and IL-8 help to recruit immune cells to colonic epithelium in IBD3”32. GM-CSF regulates immune cell phenotype during activation53-54, while IL-12p70 modulates T-helper cell differentiation to Thl (T-helper 1) cells to produce IFNy at inflammation sites55. Taken together, these statistically significant gradients in vascularized tissues instruct immune cell recruitment to inflamed epithelium (IL-ip, IL-6, IL-8), polarize macrophages at the endothelial interface (GM-CSF), and can further fine-tune the extent of inflammation at the site of injury (IL-12p70). The complex and instructive cytokine niche during IBD is absent in the avascular colon tissues, demonstrating the crucial need for underlying vasculature to accurately mimic the disease process in vitro.

[0269] Discussion

[0270] We demonstrate for the first time in vitro co-culture of perfusable microvascular beds with crypt- patterned colon epithelium. Vascularized colon barriers with crypt topography can be valuable for modeling complex disease such as IBD56, where much of the epithelium -vascular interactions in disease progression remain to be explored. Our IBD model recreated the diseased colon microenvironment with altered cytokine gradients between the colonic epithelium, and underlying microvasculature57-58. IL-8, IL- ip, IL-6, TNFa, Thl -associated cytokines (GM-CSF), and IL-12 (p40 and p70 subunits), are all common cytokines found in higher concentrations in patients with a form of IBD39 61. The cytokine profile of our model recapitulated most of these reported cytokines. Specifically, IL- 1 p, IL-6, IL-8, GM-CSF have basal- to-apical gradients and IL-12p70, TNFa have the reverse gradient in our vascularized colon tissues. Avascular colon tissues only demonstrate an IL- 1 p apical-to-basal gradient from the reported cytokines. A study by Park et al., also noted increased TNFa, IL-6, and IL-8 secretion in vascularized IBD organoid compared to non-vascularized organoids18. However, previous works on vascularized IBD organoids cannot differentiate between apical and basolateral cytokine expression1618.

[0271] Tissue and matrix remodeling is another important feature in disease progression. Kim et al., recreated the intestinal-vascular niche with a dual channel membrane -based microfluidic platform62. Caco- 2 upper channel epithelium was exposed to 0.02 dyne cm'2flow and lower channels were lined with endothelial cells with perfusion. Basolateral secretion of IL- 1 , IL-6, IL-8, and TNFa were demonstrated under inflammation. Although epithelial and basal sampling is possible in the IBD model, tissue -relevant crypt architectures were missing. Some colon epithelial folding and invaginations were observed under fluid flow on the membrane-based platform, but the dimensions of the self-assembled colon crypts are not physiological and do not reach in vivo-like depths as we have demonstrated with our patterning technique. One feature of IBD is the remodelling of crypts leading to crypt dysplasia63. Dysplastic crypts are often characterized with loss of cell differentiation in crypts, irregular cell nuclei shape, atypical crypt branching or budding, and irregular crypt sizes or shapes. Synthetic membranes cannot undergo remodelling by diseased colon cells, whereas natural ECM can be remodelled to capture diseased conditions such as crypt dysplasia. In our model, histological sectioning of diseased crypts in fibrin is possible and could reveal cellular and structural changes of crypts during disease progression. On the basolateral side, our selfassembled microvascular beds can also more accurately emulate vascular remodeling and angiogenesis, not possible in endothelial cell lined microfluidic channels64. Angiogenesis requires matrix remodelling for sprouting from pre-existing vessels. IBD severity has been correlated with increased angiogenic activity through the release of angiogenic growth factors; vascular endothelial growth factor (VEGF)-A and basic fibroblast growth factor (bFGF) during angiogenesis activation56. VEGF and bFGF are reported to synergize to worsen inflammation during pathogenesis. HUVEC vessels have been found to secrete both VEGF and bFGF65. Worsening inflammation from angiogenic factor secretion is currently not well understood in IBD, and warrants future work to investigate the underlying mechanism. Therefore, epithelial-vascular co-culture in ECM gels enable matrix remodelling that is crucial to capture angiogenesis-driven inflammation in addition to crypt dysplasia for IBD modelling.

[0272] Knowing the benefit of crypt-patterning in colon culture, scaling the patterning technique up for high-throughput drug testing is important in preclinical drug testing. Wang et al used crypt patterned colon tissues in 12 well transwell inserts for a drug screen, but required 3 plates to test 36 samples28. We tested capecitabine toxicity on 64 tissues using only a portion of the IFlowPlate384. To the best of our knowledge, this study was the first to test capecitabine drug dose response in crypt -patterned colon barriers. In place of traditional 5-FU intravenous dosing, capecitabine is orally administered with near 100% oral bioavailability35. With our open-lumen tissues, we could administer capecitabine apically as opposed to standard 96 wellplate colon spheroids with enclosed lumens66. Once administered, a complex series of enzymatic modifications convert the prodrug to its active moiety 5-FU to act selectively within tumour and proliferative cells36,67. 5-FU inhibits DNA replication in the proliferative cells to arrest cell metabolism and expansion. Over extended capecitabine treatment, damaged DNA triggers apoptotic p-53 signalling pathways leading to apoptosis. The clinical regimen is 3 -week cycles with 2 weeks of treatment followed by 1 week of recovery. Our 72-hour treatment was shorter than the clinical regimen. However, an increased dose showed acute cellular stress with increased barrier permeability. As capecitabine targets proliferative cells, colon models with crypts are more sensitive to the drug in vitro compared to planar tissues. Coupling our results with the known mode of action, we believe capecitabine could preferentially impact proliferative colon stem cells in vivo and cause colon injury.

[0273] Despite the progress we made here, there are also some limitations12. Caco-2 cells are a commonly used cell line, but have limited functionality and do not capture the diversity of colon epithelial cells in vivo. Primary colon organoids could be a better cell source and would allow us to study tissues with all native colon cell types, such as LGR5+ adult stem cells, transit amplifying cells, goblet cells, enteroendocrine cells, and colonocytes that are absent in colon cell lines12,20. Primary colon organoids will also enable us to investigate sex-specific response in disease and therapy with male and female primary cells, which is not possible with male-only Caco-2 cells. However, co-culture of colon organoids with vasculature will require optimization of matrix formulation. We have previously shown that a combination of Matrigel and fibrin can support the culture of vascularized colon organoids16. However culturing colon organoids as a monolayer barrier on the gel surface might require additional matrix optimization to finetune the stiffness of the gel. In our study, we included fibroblasts to support vasculature formation. Fibroblasts are also important by depositing collagen which could impact matrix composition over time68,69. Further studies are warranted to investigate how fibroblasts remodel the fibrin gel and deposit collagen with potential to support colon organoid culture. Circulating immune cells are also a key cell type in the colon microenvironment, especially in IBD. Immune cells are well known to exacerbate colon inflammation in IBD70. In line with this notion, future IBD models could circulate immune cells like human peripheral blood mononuclear cells through our perfusable vasculature to study extravasation to colon tissues and remodelling of colon crypts upon infiltration71. Overall, the presented colon model with vascularized crypts sets the basis for more in-depth colon and IBD studies.

[0274] Conclusion

[0275] We have developed vascularized crypt-patterned colon barriers in a high-throughput custom wellplate. Crypt topography on colon barriers was enhanced with built-in perfusable vasculature that surrounds the crypts in close proximity. By incorporating vasculature, we established more in vivo-like basal-to-apical cytokine gradients as opposed to avascular tissues. Upon application of TNFa and IFNy to trigger inflammation, the cytokine gradients reversed. Interestingly, a chemokine pathway for immune cell recruitment and infiltration developed only in the vascularized colon model. These findings re-emphasize the crucial role of underlying stromal endothelium in creating a complex, instructive cytokine niche along the crypt axis. Overall, our vascularized colon model represents a promising tool for investigating human colon physiology, toxicology assessments, and gastrointestinal pathologies.

[0276] Materials and Methods

[0277] Cell culture

[0278] Human colorectal adenocarcinoma cell-line was purchased from American Type Culture Collection (Caco-2 BBel, male, ATCC, Cat. # CRL-2102TM) and cultured in Dulbecco's Modified Eagle Medium (DMEM, Gibco; Thermo Fisher Scientific, Cat. # 11995123) supplemented with 10% fetal bovine serum (FBS, Gibco; Thermo Fisher Scientific, Cat. # 12484028), 1% HEPES (1 M, Gibco; Thermo Fisher Scientific, Cat. # 15630080), and 1% penicillin-streptomycin (100X, Sigma-Aldrich, Cat. # P4333). Green fluorescent protein expressing human umbilical vein endothelial cells (GFP-HUVECs, multiple donors, Angio-Proteomie, Cat. # CAP-0001GFP) were cultured in 0.2% w / v gelatin (Sigma-Aldrich, Cat. # G9391- 100G) coated flasks with Endothelial Cell Growth Medium 2 (EGM2, Sigma-Aldrich, Cat. # C-22111). Human primary lung fibroblasts (FB, lot-specific sex, ATCC, Cat. # PCS-201-013) were cultured in DMEM with 10% FBS, 1% HEPES, and 1% penicillin-streptomycin. All cells were expanded in T-75 cell culture flasks within an incubator at 37°C and 5% CO2 until 80% confluency before harvesting for seeding within the microfluidic device. All cells were seeded between passages 3 - 9.

[0279] TopoStamp fabrication and imprinting

[0280] The crypt-pattern insert was first generated using computer aided design (CAD) software AutoDesk Inventor 2022. The file was exported to a Formlabs Form 3B+ resin 3D printer (Cat # PKG- F3B-WSVC-MSP-BASIC). TopoStamps were printed in high temperature resin (Formlabs, Cat. # RS-F2- HT AM-02) at 25 pm resolution with subsequent 99.5% anhydrous 2-proponal (Formlabs, Cat. # 278475- 1L) washing for 30 minutes in a Form Wash basin (Formlabs, Cat. # FH-WA-01). Stamps were transferred to a Form Cure (Formlabs, Cat. # FH-CU-01) for post-curing of parts at 45 °C for 40 minutes, enabling crypt pins to completely stiffen. Prior to gel insertion, TopoStamp pins were coated with Anti-Adherence Rinsing Solution (StemCell Technologies, Cat. # 07010) for 30 minutes followed by 30 minutes of ambient air drying. After fibrin crosslinking, TopoStamps were inserted carefully for 5 minutes and removed to generate crypt topography.

[0281] IFlowPlate™ hydrogel casting, cell seeding, and operation

[0282] 25 pL of sterile distilled (DI) water was first added to inlet and outlet wells before gel casting in the centre wells. Fibrinogen (Sigma Aldrich, Cat. # F3879) at 10 mg mL'1was used as a hydrogel scaffold for cell culture. Aliquots of fibrinogen at 125 pL were mixed with 25 pL thrombin (1 U mL1, Sigma Aldrich, Cat. # T6884) to initiate gelation. The solution was quickly cast at 25 pL into each centre well of the IFlowPlate384 (OrganoBiotech, Cat. # A001) and IFlowPlate Duo. Plates were gently tapped to ensure gels dropped and spread. Crosslinking fibrin was left at room temperature for 30 minutes to gel. For crypt- patterned gels, TopoStamps were inserted into the gels as described in the TopoStamp fabrication and imprinting method section. Next, DI water was aspirated and replaced with 90 pL of medium in inlet and outlet wells. IFlowPlate384 centre wells were filled with 50 pL of medium and IFlowPlate Duo with 500 jxL of medium. All medium used was supplemented with 20 pg ml;1aprotinin (Sigma Aldrich, Cat. # 616370-100MG-M) to prevent cells from degrading fibrin. In the FlowPlate384, 65 pL Caco-2 cell suspensions at 1.2 million cells ml;1were seeded on fibrin gels in centre wells. In the IFlowPlate Duo, 32.5 pL of cell suspension at 2.4 million cells ml;1were seeded. The plates were left on the IFlowRocker™ (OrganoBiotech, Cat. # B001) in an incubator for long-term culture and medium was changed every 2 days.

[0283] If vasculature was included, 1.5 million HUVEC mL'1and 0.3 million FB mL'1were first suspended in 125 pL fibrinogen aliquots of which 25 pL were cast into the centre well. On day 2, 110 pL of cell suspension at 1 million HUVEC mL'1were seeded in inlet / outlet wells to form an endothelial lining to anastomose with the vasculature in the centre well. On day 3, the plates were placed on the IFlowRocker for long-term culture. On day 5, Caco-2 cells were seeded in centre wells and cultured in EGM2. On day 13, the centre well medium was substituted to DMEM with EGM2 remaining in inlet and outlet wells. Culture medium was changed daily in all compartments. The programmable IFlowRocker was set to tilt at a 15° angle. The tilt direction was programmed to change every 15 minutes to maintain perfusion. The IFlowPlate Duo utilized a custom adapter to orient the plate at a 45° angle in the x-y plane during tilting.

[0284] Permeability and perfusability assay

[0285] Barrer function was quantified in permeability assays using fluorescein isothiocyanate -dextran (FITC, 0.123 mg mL1, average Mw 4 kDa, Sigma-Aldrich, Cat. # 46944- 100MG-F) at 1:1 equimolar concentration with tetramethylrhodamine isothiocyanate-dextran (TRITC, 2 mg mL1, average Mw 65 kDa, Sigma-Aldrich, Cat. # T1162-100MG) in DMEM. Fluorescent dextran solutions were added to centre wells at 65 pL in the IFlowPlate384 and 500 pL in the IFlowPlate Duo. Inlet and outlet wells contained dextran free medium to serve as dextran sinks. A dextran concentration ladder was created with 1: 10 serial dilutions. Plates were imaged and fluorescent reading were collected in a Cytation5 multimode reader (BioTek Instruments Inc.) immediately after dextran addition into the IFlowPlates. Plates were incubated overnight on the IFlowRocker and re-imaged 24 hours post-dextran addition to quantify dextran accumulation in inlet and outlet wells. Vascular perfusability was visualized by adding dextran solution in EGM2 in inlet wells at 90 pL, no medium in outlet wells, and 50 pL dextran-free medium in the centre well. Vasculature was immediately imaged for dextran flow through vascular networks to outlet wells.

[0286] Proliferation assay

[0287] A Click-iT™ EdU Alexa Fluor™ 647 Imaging Kit (Thermo Fischer Scientific, Cat. # Cl 0340) was used to incorporate 5-ethynyl-2'-deoxyuridine (EdU) in tissue DNA. On day 1 EdU at 10 pM in DMEM was dispensed to all tissue culture compartments. All volumes used throughout the assay are 90 pL in inlet / outlet wells, and 50 pL in centre wells. EdU was pulsed for 24 hours with plates left on the IFlowRocker in an incubator. On day 2 the tissues were washed 3 times in Dulbecco's phosphate buffered saline IX (D-PBS, Thermo Fischer Scientific, Cat. # 14190144) and fixed with 3.7% paraformaldehyde (Electron Microscopy Sciences, Cat. # EMS 15710-S) overnight at 4°C on an IFlowRocker. On day 3 tissues were washed 3 times with D-PBS and left overnight to block in 3% Bovine Serum Albumin solution (BSA, Sigma-Aldrich, A9205-10ML) in D-PBS at 4°C on a programmable rocker. On day 4 tissues were permeated overnight at 4°C under perfusion with 10% Fetal Bovine Serum (FBS, Thermo Fisher Scientific, Cat. # 12484028) containing 0.1% Triton™ X-100 (Sigma-Aldrich, X100-500ML). On day 5 tissues were washed again 3 times in D-PBS and left in 3% BSA under perfusion overnight. On day 6 click-iT® reaction mix was created following vendor instructions and added to the IFlowPlate384 under perfusion at 4°C. On day 7, washing and 3% BSA blocking was again performed prior to 4',6-Diamidino-2-phenylindole dihydrochloride (DAPI, Sigma-Aldrich, Cat. # D9542-5MG) nuclei staining at 1:500 dilution in 10% FBS infused D-PBS solution. On day 8, tissues were washed 3 times and left overnight for imaging on day 9. Tissues were imaged in a Cytation5 multimode reader and a Marianas lightsheet / confocal microscope (3i Intelligent Imaging Innovations). Negative control tissue without an EdU pulse (Supplementary Figure 3) validated no occurrences of click-iT reactions in the absence of EdU. Acquired fluorescence images were processed with ImageJ software to extract EdU fluorescence from known well areas and plotted as normalized intensity to area values.

[0288] High-throughput robotic gel casting

[0289] An automatic liquid handling system (Hamilton Company, NIMBUS4, NMBSBK4670) cast fibrin hydrogels into the IFlowPlate384. For casting hydrogels, four tubes of distilled water, four microtubes of 10 mg mL’1fibrinogen and four microtubes of 1 U mL-1thrombin were prepared and placed in the NIMBUS. Empty microtubes were loaded for mixing fibrinogen and thrombin. The NIMBUS first added distilled water to inlet and outlet wells in the IFlowPlate384. The NIMBUS then aliquoted fibrinogen to the empty microtubes, mixed thrombin with fibrinogen, and cast gels in centre wells of 120 tissue subunits of the IFlowPlate384 leaving 8 wells blank. TopoStamps were inserted after crosslinking to generate crypt topography on 64 gels and leaving 56 planar. The NIMBUS then aspirated water from inlet and outlet wells, then added 20 pL of DMEM to inlet and outlet wells and 90 pL to centre wells. The plate was placed in an incubator until cell seeding.

[0290] Drug testing

[0291] Colon barriers were cultmed in the Nimbus assisted IFlowPlate384 until day 12 when the quality control permeability assay was performed. Four out of the 120 tissues presented no barrier function due to physical defects and were discarded. Capecitabine (Cancer Panel - Plate 4893) was dissolved in DMEM at 5, 1, 0.1, 0.01 pM and 0.05% DMSO vehicle controls were apically administered. Daily medium changes were performed with capecitabine and vehicle controls for 72 horns. A permeability assay was performed immediately after 72-hour treatment and then all compartments were replenished with fresh DMEM for recovery. Another permeability assay was completed at the end of the 3 -day barrier recovery period.

[0292] Glucose consumption

[0293] Fresh DMEM was added to 14-day old crypt-patterned colon tissue in the IFlowPlate384, IFlowPlate Duo, and a standard 384 wellplate for 24 hours. After 24 hours of culture in each plate, DMEM in the centre well was removed and analyzed for glucose molarity using a calibrated glucometer (Contour® Next One meter) and compared to the glucose concentration in fresh DMEM.

[0294] IFlowPlate Duo apical shear stress Fibrin hydrogels were cast in centre wells of the IFlowPlate Duo and perfused with fluorescent FITC polystyrene beads (1.0 pm, Sigma-Aldrich, Cat. # L1030), diluted at a ratio of 1:5 in D-PBS. The plates were loaded on a Nikon brightfield microscope and tilted 15° with a 45° diagonal slant to emulate culture conditions. Apical bead flow on the surface of fibrin gels was recorded during perfusion. Microparticle perfusion was analyzed using a PIV plugin from ImageJ

[0295] (https: / / sites.uooule.com / site / qinpzong senu / piv) to generate velocity vector fields. Shear stress was calculated using the Newtonian fluid shear stress formula72dv T = 11 — dz

[0296] The dynamic viscosity (p) of D-PBS was approximated as water at room temperature (1.0016 mPa«s). Maximum velocities from vector fields were used at peak and steady state flows for the velocity gradient (dv) and an assumed value of 10 pm distance of spherical beads from the apical surface of the fibrin gel (dz). The polystyrene beads are 1 pm in diameter and aggregate in D-PBS from intrinsic hydrophobicity to larger complexes. Single beads and colloidal aggregates are assumed to flow a few microns higher than their respective diameters to not contact the apical gel surface.

[0297] Cytokine assay and colon inflammation in the IFlowPlate Duo

[0298] Supernatants on day 13 of culture were collected from all compartments in the IFlowPlate Duo and stored in -80°C. The collected supernatant was sent to Eve Technologies for quantifying pro- and antiinflammatory cytokines using Human Cytokine Array Pro-inflammatory Focused 15-plex panel (Cat. # HDF15). Vascularized and avascular colon tissues were stimulated apically with 50 ng mL-1TNFa (Sigma- Aldrich, Cat. # SRP3177-50UG) and IFNy (Sigma-Aldrich, I17001-100UG) on day 13. A cytokine assay was subsequently performed on day 14 after stimulation to quantify tissue inflammation.

[0299] Scanning electron microscopy

[0300] Tissue samples were washed 3 times in D-PBS and fixed in 4% paraformaldehyde. The samples were subsequently rinsed twice in buffer solution, post-fixed in 1% osmium tetroxide in 0.1 M phosphate buffer for 1 hour and then dehydrated through a graded ethanol series (50%, 70%, 70%, 95%, 95%, 100%, 100%). The samples were kept immersed in 100% ethanol, and transferred to the chamber of a Leica EM CPD300 critical point dryer (Leica Mikrosysteme GmbH, Wien, Austria). The chamber was sealed and then flushed 12 times with liquid CO2. The CO2 filled chamber was heated to 35°C and pressure increased in chamber to above 1100 psi. The gas was vented slowly from the chamber until atmospheric pressure was reached and the samples were dehydrated without surface tension damage. The dried samples were mounted onto scanning electron microscopy (SEM) stubs. The samples on the stubs were then placed in the chamber of a Polaron Model E5100 sputter coater (Polaron Equipment Ltd., Watford, Hertfordshire) and approximately 15 nm of gold was deposited onto the stubs. The samples were viewed in a Tescan Vega II LSU scanning electron microscope (Tescan USA, PA) operating at 20 kV at the Canadian Centre for Electron Microscopy. Immunostaining and histology

[0301] The tissues were preserved in a 4% paraformaldehyde solution at 4°C overnight on the IFlowRocker after being rinsed three times with D-PBS. On day 2, fixed tissues were blocked with 10% FBS and permeated with 0.1% Triton X-100 overnight at 4°C. This was followed by a day 3 D-PBS wash. On day 4, cytoskeletal (Phalloidin-iFluor™ 594 Conjugate, F-actin, 1:1000, Cedarlane Labs, Cat. # 20553- 300) and cell nuclei (DAPI, 1:500, Sigma-Aldrich, Cat. # D9542-5MG) stains were mixed in 10% FBS in PBS solution and incubated for 24 hours at 4°C with perfusion. Before imaging with a Cytation5 multimode reader and a Marianas lightsheet / confocal microscope, the tissues were rinsed on day 5 with D-PBS for 24 hours at 4°C. 3D confocal renderings were post-processed using NIS-Element Nikon software at the McMaster Centre for Advanced Light Microscopy. For histology, tissues were preserved in 10% formalin solution (Sigma-Aldrich, Cat. #HT501128-4L) for 48 hours prior to being processed at McMaster University's Histology Facility for sectioning. Histological sections were stained for endothelial vessels (CD31, primary, rabbit, histology, 1:75, Abeam, Cat. # ab28364), microvilli (villin, primary, rabbit, histology, 1:200, Abeam, Cat. # abl30751), cell nuclei (eosin), and extracellular matrix (ECM, hematoxylin). Histology slices were imaged on a Nikon brightfield microscope.

[0302] Statistical analysis

[0303] The mean and standard deviation of all data are plotted unless stated otherwise in the figure caption. GraphPad was used to assess for equality and normality. In all tests, a p-value of 0.05 or lower was regarded as statistically significant. Comparing two normal data populations was calculated with unpaired parametric two tailed t-tests (Figure 14b, 15g) and paired parametric two tailed t-tests (Figure lOh). Non-normal data sets were calculated with unpaired two tailed nonparametric Mann-Whitney t-tests (Figure 13d, 14b, 15g). Statistical analyses on 3 or more sets of experimental data groups were performed with unmatched one-way ANOVA with mean ranks compared using Holm-Sidak’s test on normal data (Figure 10g, 15b, Figure 12b). Non-normal data utilized unpaired nonparametric test with mean ranks compared using Dunn’s test (Figure lOe, 13h, Figure lOd).

[0304] References

[0305] 1. Bein, A. et al. Microfluidic Organ-on-a-Chip Models of Human Intestine. Cmgh 5, 659- 668 (2018).

[0306] 2. Meng, F. et al. Mechanical stretching boosts expansion and regeneration of intestinal organoids through fueling stem cell self-renewal. Cell Regeneration 11, 39 (2022).

[0307] 3. Nelson, C. M. The mechanics of crypt morphogenesis. Nat Cell Biol doi: 10.1038 / s41556- 021-00703-z.

[0308] 4. Fang, G. et al. Enabling peristalsis of human colon tumor organoids on microfluidic chips. Biofabrication (2021) doi:10.1088 / 1758-5090 / ac2ef9.

[0309] 5. Perez-Gonzalez, C., Ceada, G., Matejcic, M. & Trepat, X. Digesting the mechanobiology of the intestinal epithelium. Curr Opin Genet Dev 72, 82-90 (2022). 6. Martin-Alonso, M. et al. Smooth muscle-specific MMP17 (MT4-MMP) regulates the intestinal stem cell niche and regeneration after damage. Nat Commun 12, 6741 (2021).

[0310] 7. Santos, A. J. M., Lo, Y. H., Mah, A. T. & Kuo, C. J. The Intestinal Stem Cell Niche: Homeostasis and Adaptations. Trends in Cell Biology vol. 28 1062-1078 Preprint at https: / / doi.Org / 10.1016 / j.tcb.2018.08.001 (2018).

[0311] 8. Wang, Y. et al. Formation of Human Colonic Crypt Array by Application of Chemical Gradients Across a Shaped Epithelial Monolayer. CMGH 5, 113-130 (2018).

[0312] 9. Gajendran, M. et al. A comprehensive review and update on ulcerative colitis,. Disease- a-Month vol. 65 Preprint at https: / / doi.Org / 10.1016 / j.disamonth.2019.02.004 (2019).

[0313] 10. Michielan, A. & D’lnca, R. Intestinal Permeability in Inflammatory Bowel Disease: Pathogenesis, Clinical Evaluation, and Therapy of Leaky Gut. Mediators of Inflammation vol. 2015 Preprint at https: / / doi.org / 10.1155 / 2015 / 628157 (2015).

[0314] 11. Gajendran, M. et al. A comprehensive review and update on ulcerative colitis,. Disease- a-Month vol. 65 Preprint at https: / / doi.Org / 10.1016 / j.disamonth.2019.02.004 (2019).

[0315] 12. Zietek, T., Boomgaarden, W. A. D. & Rath, E. Drug Screening , Oral Bioavailability and Regulatory Aspects : A Need for Human Organoids. 1-24 (2021).

[0316] 13. Jozani, K. A., Kouthouridis, S., Hirota, J. A. & Zhang, B. Next generation preclinical models of lung development, physiology and disease.

[0317] 14. Carberry, B. J. et al. 3D printing of sacrificial thioester elastomers using digital light processing for templating 3D organoid structures in soft biomatrices. Biofabrication (2021) doi:10.1088 / 1758-5090 / aclc98.

[0318] 15. Engevik, A. C. Using Microfluidics to Model Mucus. Cmgh 9, 551-552 (2020).

[0319] 16. Rajasekar, S. et al. IFlowPlate — A Customized 384-Well Plate for the Culture of Perfusable Vascularized Colon Organoids. 2002974, 1-12 (2020).

[0320] 17. Han, H. et al. A Bioprinted Tubular Intestine Model using a Colon-Specific Extracellular Matrix Bioink. Adv Healthc Mater 2101768 (2021) doi: 10.1002 / adhm.202101768.

[0321] 18. Park, S. E. et al. Geometric engineering of organoid culture for enhanced organogenesis in a dish. Nat Methods (2022) doi:10.1038 / s41592-022-01643-8.

[0322] 19. Tissue geometry drives deterministic organoid patterning. 9021, (2022).

[0323] 20. Wang, Y. et al. Self -renewing Monolayer of Primary Colonic or Rectal Epithelial Cells. Cmgh . 165-182. e7 (2017).

[0324] 21. Hinman, S. S., Wang, Y., Kim, R. & Allbritton, N. L. In vitro generation of self-renewing human intestinal epithelia over planar and shaped collagen hydrogels. Nat Protoc 16, 352-382 (2021).

[0325] 22. Rajasekar, S. et al. Subtractive manufacturing with swelling induced stochastic folding of sacrificial materials for fabricating complex perfusable tissues in multi-well plates. Lab Chip 482-487 (2022) doi:10.1039 / dllc01141c. 23. Sotra, A. & Zhang, B. Seeding A Growing Organ. Trends in Biotechnology vol. 39 753- 754 Preprint at https: / / doi.Org / 10.1016 / j.tibtech.2021.05.002 (2021).

[0326] 24. Nikolaev, M. et al. Homeostatic mini-intestines through scaffold-guided organoid morphogenesis. Nature 585, 574-578 (2020).

[0327] 25. Taebnia, N. et al. Dual-Material 3D-Printed Intestinal Model Devices with Integrated Villi-like Scaffolds. ACS Appl Mater Interfaces acsami.lc22185 (2021) doi:10.1021 / acsami.lc22185.

[0328] 26. Rudolph, S. E. et al. Crypt-Villus Scaffold Architecture for Bioengineering Functional Human Intestinal Epithelium. ACS Biomater Sci Eng (2022) doi: 10.1021 / acsbiomaterials.2c00851.

[0329] 27. Lin, D. S. Y., Rajasekar, S., Marway, M. K. & Zhang, B. From model system to therapy: Scalable production of perfusable vascularized liver spheroids in ‘open-top’ 384-well plate. ACS Biomater Sci Eng 7, 2964-2972 (2021).

[0330] 28. Wang, Y., Sims, C. E. & Allbritton, N. L. Human 2D Crypt Model for Assaying Intestinal Stem Cell Proliferation and Differentiation. Anal Chem (2022) doi:10.1021 / acs.analchem.2c00905.

[0331] 29. Jeon, H. J. et al. Enhanced Osteoblast Adhesion and Proliferation on Vacuum Plasma- Treated Implant Surface. Applied Sciences (Switzerland) 12, (2022).

[0332] 30. Sasmazel, H. T., Alazzawi, M. & Alsahib, N. K. A. Atmospheric pressure plasma surface treatment of polymers and influence on cell cultivation. Molecules vol. 26 Preprint at https: / / doi.org / 10.3390 / molecules26061665 (2021).

[0333] 31. Natoli, M., Leoni, B. D., D’Agnano, L, Zucco, F. & Felsani, A. Good Caco-2 Cell Culture Practices. Toxicology in Vitro 26, 1243-1246 (2012).

[0334] 32. Kim, H. J. & Ingber, D. E. Gut-on-a-Chip microenvironment induces human intestinal cells to undergo villus differentiation. Integrative Biology (United Kingdom) 5, 1130-1140 (2013).

[0335] 33. Ge, H., Tian, M., Pei, Q., Tan, F. & Pei, H. Extracellular Matrix Stiffness: New Areas Affecting Cell Metabolism. Frontiers in Oncology vol. 11 Preprint at https: / / doi.org / 10.3389 / fonc.2021.631991 (2021).

[0336] 34. Shin, W. & Kim, H. J. 3D in vitro morphogenesis of human intestinal epithelium in a gut- on-a-chip or a hybrid chip with a cell culture insert. Nat Protoc (2022) doi:10.1038 / s41596-021-00674-3.

[0337] 35. Walko, C. M. & Lindley, C. Capecitabine: A Review. Clin Ther 27, (2005).

[0338] 36. Schellens, J. H. M. Capecitabine. Oncologist 12, 152-155 (2007).

[0339] 37. Gonzalez-Dominguez, J. M. et al. Intrinsic and selective activity of functionalized carbon nanotube / nanocellulose platforms against colon cancer cells. Colloids Surf B Biointerfaces 212, (2022).

[0340] 38. Yamamoto, S., Hotta, M. M., Okochi, M. & Honda, H. Effect of vascular formed endothelial cell network on the invasive capacity of melanoma using the in vitro 3D co-culture patterning model. PLoS One 9, (2014).

[0341] 39. Piard, C. et al. 3D printed HUVECs / MSCs cocultures impact cellular interactions and angiogenesis depending on cell-cell distance. Biomaterials 222, 119423 (2019). 40. Ramirez-Peinado, S. et al. Glucose-starved cells do not engage in prosurvival autophagy.

[0342] Journal of Biological Chemistry 288, 30387-30398 (2013).

[0343] 41. Fois, C. A. M., Schindeler, A., Valtchev, P. & Dehghani, F. Dynamic flow and shear stress as key parameters for intestinal cells morphology and polarization in an organ-on-a-chip model. Biomed Microdevices 23, 1-12 (2021).

[0344] 42. Kim, S. W. et al. Shear stress induces noncanonical autophagy in intestinal epithelial monolayers. Mol Biol Cell 28, 3043-3056 (2017).

[0345] 43. Lindner, M., Laporte, A., Block, S., Elomaa, L. & Weinhart, M. Physiological shear stress enhances differentiation and mucus-formation of intestinal epithelial cells in vitro . 1-19 (2020).

[0346] 44. Mahapatro, M., Erkert, L. & Becker, C. Cytokine -Mediated Crosstalk between Immune Cells and Epithelial Cells in the Gut. (2021).

[0347] 45. Sun, H., Tan, J., Chen, H., Wu, N. & Su, B. Immune niches orchestrated by intestinal mesenchymal stromal cells lining the crypt-villus. 1-8 (2022) doi: 10.3389 / fimmu.2022.1057932.

[0348] 46. Andrews, C., McLean, M. H. & Durum, S. K. Cytokine tuning of intestinal epithelial function. Front Immunol 9, (2018).

[0349] 47. McEntee, C. P., Finlay, C. M. & Lavelle, E. C. Divergent roles for the IL-1 family in gastrointestinal homeostasis and inflammation. Frontiers in Immunology vol. 10 Preprint at https: / / doi.org / 10.3389 / fimmu.2019.01266 (2019).

[0350] 48. Luissint, A.-C., Parkos, C. A. & Nusrat, A. Inflammation and the Intestinal Barrier: Leukocyte-Epithelial Cell Interactions, Cell Junction Remodeling, and Mucosal Repair. HHS Public Access. Gastroenterology 151, 616-632 (2016).

[0351] 49. Nava, P. et al. Interferon-y Regulates Intestinal Epithelial Homeostasis through Converging P-Catenin Signaling Pathways. Immunity 32, 392-402 (2010).

[0352] 50. Th, C. D. et al. IL-ip mediates chronic intestinal inflammation by promoting the accumulation of IL-17A secreting innate. 209, 1595-1609 (2012).

[0353] 51. Grivennikov, S. et al. NIH Public Access. 15, 103-113 (2010).

[0354] 52. Kvedaraite, E. T cell crosstalk in inflammatory bowel disease, doi: 10.1111 / imm.13391.

[0355] 53. Castro-dopico, T. et al. Article GM-CSF Calibrates Macrophage Defense and Wound Healing Programs during Intestinal Infection and Inflammation 11 GM-CSF Calibrates Macrophage Defense and Wound Healing Programs during Intestinal Infection and Inflammation. (2020) doi: 10.1016 / j .celrep.2020.107857.

[0356] 54. Lee, K. M. C., Achuthan, A. A. & Hamilton, J. A. GM-CSF : A Promising Target in Inflammation and Autoimmunity. 225-240 (2020).

[0357] 55. Peluso, I. et al. Interleukin- 12 and Thl immune response in Crohn ’ s disease : Pathogenetic relevance and therapeutic inplication. 12, 5606-5610 (2006). 56. Stiirzl, M., Kunz, M., Krug, S. M. & Naschberger, E. Angiocrine Regulation of Epithelial Barrier Integrity in Inflammatory Bowel Disease. Frontiers in Medicine vol. 8 Preprint at https: / / doi.org / 10.3389 / fmed.2021.643607 (2021).

[0358] 57. Kondreddy, V., Keshava, S., Esmon, C. T., Pendurthi, U. R. & Rao, L. V. M. A critical role of endothelial cell protein C receptor in the intestinal homeostasis in experimental colitis. Sci Rep 10, (2020).

[0359] 58. Cromer, W. E., Mathis, J. M., Granger, D. N., Chaitanya, G. v. & Alexander, J. S. Role of the endothelium in inflammatory bowel diseases. World J Gastroenterol 17, 578-593 (2011).

[0360] 59. Pearl, D. S. et al. Cytokine mucosal expression in ulcerative colitis, the relationship between cytokine release and disease activity. J Crohns Colitis 7, 481 489 (2013).

[0361] 60. Nakase, EL, Sato, N., Mizuno, N. & Ikawa, Y. The influence of cytokines on the complex pathology of ulcerative colitis. Autoimmunity Reviews vol. 21 Preprint at https: / / doi.Org / 10.1016 / j.autrev.2021.103017 (2022).

[0362] 61. Nemeth, Z. H. et al. Crohn’s Disease and Ulcerative Colitis Show Unique Cytokine Profiles. Cureus (2017) doi:10.7759 / cureus.H77.

[0363] 62. Kim, H. J., Li, EL, Collins, J. J. & Ingber, D. E. Contributions of microbiome and mechanical deformation to intestinal bacterial overgrowth and inflammation in a human gut-on-a-chip. Proc Natl Acad Sci USA 113, E7-E15 (2016).

[0364] 63. Choi, W. T. et al. Hypermucinous, Goblet Cell-Deficient and Crypt Cell Dysplasias in Inflammatory Bowel Disease are Often Associated with Flat / Invisible Endoscopic Appearance and Advanced Neoplasia on Follow-Up. J Crohns Colitis 16, 98-108 (2022).

[0365] 64. Zhang, S. et al. Interstitial Flow Promotes the Formation of Functional Microvascular Networks In Vitro through Upregulation of Matrix Metalloproteinase-2. Adv Fund Mater 2206767 (2022) doi: 10.1002 / adfm.202206767.

[0366] 65. Tan, A. W. et al. Enhanced in vitro angiogenic behaviour of human umbilical vein endothelial cells on thermally oxidized TiO 2 nanofibrous surfaces. Sci Rep 6, 1-10 (2016).

[0367] 66. Johnson, P. A. et al. A rapid high throughput bioprinted colorectal cancer spheroid platform for in vitro drug- and radiation-response. Biofabrication 15, 014103 (2023).

[0368] 67. Shaaban, S., Negm, A., Ibrahim, E. E. & Elrazak, A. A. Hepatocellular carcinoma chemotherapeutic agents: Efficacy and mode of action. Oncology Reviews vol. 8 Preprint at https: / / doi.org / 10.4081 / oncol.2014.246 (2014).

[0369] 68. Gomes, R. N., Manuel, F. & Nascimento, D. S. The bright side of fibroblasts: molecular signature and regenerative cues in major organs. NPJ Regen Med 6, 1-12 (2021).

[0370] 69. Wang, Y., Kumar, R., Ou, C., Chen, A. & Yang, Z. Bioactive Materials Advances in hydrogel-based vascularized tissues for tissue repair and drug screening. Bioact Mater (2021) doi: 10.1016 / j .bioactmat.2021.07.005. IQ. Neurath, M. F. Targeting immune cell circuits and trafficking in inflammatory bowel disease. Nature Immunology vol. 20 970-979 Preprint at https: / / doi.org / 10.1038 / s41590-019-0415-0 (2019).

[0371] 71. Jessen, B. et al. Level of Tumor Necrosis Factor Production by Stimulated Blood Mononuclear Cells Can Be Used to Predict Response of Patients With Inflammatory Bowel Diseases to Infliximab. Clinical Gastroenterology and Hepatology 19, 721-731. el (2021).

[0372] 72. Pisapia, F., Balachandran, W. & Rasekh, M. Organ -on-a-Chip: Design and Simulation of Various Microfluidic Channel Geometries for the Influence of Fluid Dynamic Parameters. Applied Sciences (Switzerland) 12, (2022).

[0373] EXAMPLE 3, IMMUNE-INFILTRATED CANCER SPHEROID MODEL WITH VASCULAR RECIRCULATION REVEALS TEMPORALLY DEPENDENT AND TISSUE-SPECIFIC MACROPHAGE RECRUITMENT

[0374] Abstract

[0375] Immune cells infiltration in tumor has been reported to contribute to tumor progression and affecting clinical outcomes in cancer treatment. Considerable efforts have been made to elucidate the interactions between tumors and the immune system. However, current models for understating tumor- immune interactions are either not comprehensive or limited to short-term studies. Recognizing the tumor- immune interaction being a dynamic and long-term process, we developed an immune-infiltrated cancer spheroid model by continuously perfusing and recirculating immune cells with gravity -driven flow through a tubular blood vessel adjacent to a cancer spheroid. Pericytes and fibroblasts were also embedded in the gel matrix to better support the endothelial cells and enhanced the vascular barrier. Through continuous recirculation of monocytes, we successfully demonstrated the monocyte adhesion within blood vessels, transendothelium migration, differentiation and macrophage recruitment into cancer spheroids in breast carcinoma and hepatoma models over a week. The macrophage recruitment process was temporally dependent and tissue-specific. It naturally led to the formation of cancer -macrophage heterospheroids with a macrophage population that was continuously growing, renewing, self-evolving. The elevated levels of IL-6 and IL-8 in the breast carcinoma model was observed, suggesting a pro -inflammatory environment that can promote tumor progression and metastasis. This platform provides a valuable framework for investigating immune cell infiltration and differentiation within the tumor microenvironment, which is essential for the advancement of cancer immunotherapies.

[0376] Introduction

[0377] Cancer is the leading cause of death worldwide, accounting for nearly 10 million deaths in 2020, or nearly one in six deaths. Cancer immunotherapy has become a focus point of research in the field of tumor immunology over the past decades.1Our immune system can suppress tumor growth by destroying cancer cells, but it can also promote tumor growth by aiding cancer cells in evading immune surveillance.2The interactions between cancer and immune systems are complex, making it challenging to devise effective cancer treatment strategies. Modeling these cancer-immune interactions could be beneficial for better understanding the underlying mechanisms of tumor immunology. In the tumor microenvironment, several factors, such as the endothelium barrier, mural cells and stromal cells, circulating immune cells and physical blood flow, have been reported to impact tumor progression, angiogenesis, metastasis, and prognostic.3,4For instance, endothelial cells function as the gatekeeper of immunity, interacting with circulating immune cells in blood stream and controlling their extraversion from the blood vessels to the parenchyma by regulating the expression of adhesion proteins such as vascular cell adhesion molecule I.5Pericytes have been implicated as mediators of tumor angiogenesis and metastasis, where they are activated and detach from the vessel wall, enabling endothelial cell migration and invasion into the surrounding matrix to form new blood vessels with increased vessel permeability.6Recently, immune infiltration in tumors has been reported to play a key role in the tumor progression and is regarded as a promising target for novel cancer therapeutic strategies.7,8Therefore, a proper cancer model that integrates those key components of tumor microenvironments is needed to fully recapitulate the physiological conditions of cancer and understand the interactions between immune system and cancer for clinical cancer therapy.

[0378] Many tumor models have been developed, enabling the investigation of tumor progression, angiogenesis, and metastasis processes and mechanisms. These include animal models and organ -on-chip tumor models.9,10Although animal cancer models can maintain the biological characteristics of primary tumors, they fall short in representing the full complexity and functionality of human immune system, with limited development of human mature innate immune cell populations.11,12Humanized animal cancer models are time-consuming and costly to create and maintain, with a relatively short experimental time.133D tumor spheroids models remain a popular approach for exploring tumor responses to anticancer drugs, yet they inadequately represent blood vessel-tumor and immune-tumor interactions.14In contrast, organ-on-chip models offer advantages in more precisely controlling the tumor microenvironment and can replicate complex biological processes or interactions, such as tumor-blood vessel interactions, tumor angiogenesis, tumor-immune cell interactions or biochemical gradient.15 17For instance, Meng et al. developed a vascularized tumor model in fibrin gel that demonstrated tumor angiogenesis and tumor metastasis by temporally and spatially manipulating vascular endothelial growth factor (VEGF) or epidermal growth factor (EGF) gradients.18Neufeld et al. reported a glioblastoma model that mimicked cell diversity and tumor-stroma interactions in the glioblastoma microenvironment by incorporating pericytes and astrocytes into glioblastoma spheroids and generating blood vessels surrounded by pericytes. This model even exhibited tumor growth curves similar to those in mouse models.19Recent advancements in tumor modeling have started to integrate immune components into the tumor microenvironment as we now know that the development of a tumor is closely associated with the immune cell recruitment and differentiation process.20-23However, current models struggle to replicate this intricate process accurately. A significant hurdle is the continuous recruitment of immune cells from the circulation into the cancer microenvironment, a persistent process that contrasts with the standard practice of one-time cell seeding typically used in tissue production. Current methods mostly uses external peristaltic pumps to support long-term perfusion and recirculation of immune cells.24,25But the dead volume within the tubing connections and the mechanical stress imposed on cells by these pumps can introduce variability that may compromise cell viability and function. Therefore, there is a pressing need to create a comprehensive tumor model that facilitates ongoing recruitment of immune cells from the bloodstream into the tumor microenvironment.

[0379] To recapitulate tumor microenvironments with continuous immune cell circulation (Figure 17a), we report an immune -infiltrated cancer spheroid model. This model is created by initially seeding endothelial cells in a vascular channel embedded in an extracellular matrix (ECM) hydrogel, then placing cancer spheroids next to the tubular blood vessel, followed by circulating immune cells (monocytes in this work) under flow using a pumpless platform - UniPlate™, as we previously published.26Flow rate characterization confirms that there is a continuous unidirectional flow from the inlet to the outlet well via the blood vessels. Stromal cells, especially fibroblasts and pericytes, which contribute to the barrier function of blood vessels, were included in the cancer spheroid model to reconstitute the interactions between the blood vessel and stroma. Moreover, using the UniPlate, specific monocyte migration and infiltration into cancer spheroids were demonstrated for the first time by continuously recircuiting immune cells using a pumpless recirculating platform. This setup effectively models tissue specific monocyte migration and infiltration, showcasing the dynamic process of immune cells responding to cancer. We believe this immune-infdtrated cancer spheroid model, which recapitulates the interactions among blood vessels, stromal cells, tumors, and circulating immune cells, is a valuable tool for elucidating tumor-immune interactions.

[0380] Results and discussion

[0381] UniPlate setup and unidirectional perfusion

[0382] The UniPlate devices were manufactured by injection molding and 3D printing, as we have previously described.26In this work we introduced a funnel-shaped design for the inlet and outlet wells of the UniPlate device (Figure 17b). This design aims to eliminate dead zones in the corners of wells, thereby enhancing cell recirculation efficiency. The subtractive manufacturing method was used to create perfusable tubular tissues embedded in a hydrogel matrix.27The printed gelatin template serves as a sacrificial scaffold for the tubular tissues. After removing the gelatin, the open lumen is populated with endothelial cells to produce a tubular blood vessel. For continuous medium perfusion and recirculation, we placed the plate on a programmable rocker. For each perfusion cycle, the rocker first tilts to 15° for 5 min, and then it shifts to the opposite tilt direction and remain at -30° for 10 s to allow the culture medium to flow back (Figure 17c). A single tubular channel design with the average diameter of about 530 pm was used in this work. The coefficient of variances (CV) among different plates was between 2.61% and 6.98% (Figure 17d), indicating a high degree of consistency in the dimensions of the engineered tubular channels across different UniPlate devices. We also evaluated the flow condition by collecting medium accumulated in the outlet well over a perfusion cycle (30 seconds to 5 minutes). The flow rates gradually changed from 19.1 to 69.9 |iL / min in the 5-min interval as the culture media drain from inlet wells to outlet wells but reset at the beginning of each cycle (Figure 17e). The average flow rate was 36.8 ± 4.4 |iL / min (Figure 17f) with an average wall shear stress of 0.41 ± 0.03 dyne / cm2(Figure 17g). Although the shear stress observed in our study is lower than that found in in arterial (4-30 dyne / cm2) and venous (1-4 dyne / cm2) blood flow, it aligns with values reported in prior studies.28-30

[0383] Stromal cells improve vascular barrier in 3D tubular blood vessels

[0384] To construct tubular blood vessels on UniPlate, we embedded stromal cells, specifically fibroblasts and pericytes, into a fibrin gel matrix that envelops a tubular channel seeded with endothelial cells (Figure 18a). Under conditions of continuous recirculating flow, the tube developed a contiguous endothelial lining on its inner surface (Figure 18b). The engineered blood vessels, supported by stromal cells, displayed a strong barrier function with a permeability coefficient of 9.75 x 10-7cm / s, which is comparable to that of rat cerebral microvessels (1.5 x 10-7cm / s), for 70 KDa dextran.31Blood vessels with endothelial linings showed significantly lower permeability than the acellular controls, underscoring the function of the endothelial layer in maintaining vascular integrity (Figure 18c). Notably, vessels containing both fibroblasts and pericytes had even lower permeability compared to those lined with endothelial cells alone, indicating that the presence of stromal cells leads to tighter vascular junctions. The importance of fibroblasts and pericytes in the development and strengthening of blood vessels is supported in previous studies.32,33The presence of intercellular junctions in vascular tissues, both with and without stroma cells, was further validated by VE-Cadherin expression (Figure 18d), confirming the establishment of tight junctions in the engineered blood vessels.

[0385] Macrophage recruitment to cancer spheroids is temporally dependent

[0386] Building on the tubular blood vessel, a vascularized tumor model was developed by incorporating cancer spheroids next to the blood vessel using a specially designed printed stamp for precise placement (Figure 18a,b). The stamp was designed with a taper tip (Figure 19) and printed with a high temperature resin using a FormLab printer, making it suitable for autoclaving and subsequent use in cell culture. Its sharp tip is engineered to penetrate the fibrin gel from above, leaving behind a cavity that is then filled with cancer spheroids. Our vascularized cancer spheroid model, complete with endothelial lining, stromal support, and immune elements, was established through two phases: initial formation of the tumor with its vascular network, and subsequent long-term cycling of immune cells (Figure 20c). To visualize immune cell movement, monocytes were marked with a red cell tracker and circulated within the blood vessel.

[0387] We explored the impact of sustained immune cell circulation on tumor microenvironments by comparing two scenarios: one-time monocyte perfusion and continuous monocyte recirculation. In this study, one-time monocyte perfusion refers to allowing monocytes to pass through the blood vessels a single time during the entire experimental procedure. In contrast, continuous monocyte recirculation involves cycling immune cells through the blood vessels every five minutes and introducing new monocytes each day. This mimics the average lifespan of human monocytes, which is typically one day.34In a natural setting, a monocyte circulates through the vascular system multiple times daily, with each loop taking about a minute.35Our model could theoretically replicate this, allowing for up to 288 circulation cycles in a day, akin to the in vivo frequency. The extent of monocyte adhesion within blood vessels and infiltration into the cancer spheroids was then measured (Figure 20d). The results indicated that continuous monocyte recirculation led to increased monocyte adhesion to the vessel walls over time (Figure 20e). However, the one-time monocyte perfusion showed minimal monocyte attachment, indicating that continuous monocyte recirculation is necessary for effective immune cell-tumor interactions. Additionally, a weak fluorescence signal of monocytes was initially detected 2 hours after introducing monocytes into the one-time monocyte perfusion group, but it gradually diminished over time. This result suggests that the interaction between immune cells and tumor blood vessels, after a onetime perfusion with monocytes, is not stable enough to persist.

[0388] To further examine the dynamics of monocyte -cancer interactions, we monitored the fluorescently labeled monocytes at tumor sites during monocyte perfusion (Figure 20f). A notable increase in fluorescence signal at tumor sites was observed with continuous monocyte recirculation, indicative of monocyte migration and infiltration into the tumor mass. Conversely, tumor sites subjected to the one-time monocyte perfusion exhibited no significant increase in fluorescence intensity, once again demonstrating the importance of continuous monocyte recirculation to effectively study monocyte behaviors and immune responses within tumors. Interestingly, no apparent fluorescence signal was detected at tumor sites following 2 hours of continuous monocyte recirculation. However, after one day of sustained monocyte recirculation, signal became detectable. We quantitatively analyzed the relative fold changes in fluorescence intensity within both blood vessels and tumors to evaluate monocyte adhesion and transmigration (Figure 20g, h). In line with the fluorescence time-lapse images, the relative change in monocyte signal within both blood vessels and tumors for the one-time monocyte perfusion group did not vary significantly over time. In contrast, for the continuous monocyte recirculation group, the fluorescence signal within the blood vessels experienced a 6-fold increase during the initial three days, stabilizing between day 2 and 4. Moreover, the fluorescence signal within the tumor increased continuously over the 5-days observation period. This increase suggests that cytokines released by cancer spheroids might be constantly attracting monocytes, prompting their sustained recruitment to the tumor site.

[0389] Macrophage recruitment from vascular circulation is tissue-specific

[0390] The behavior of monocytes within the tumor microenvironment is multifaceted. Circulating monocytes are activated and drawn to the tumor site in response to a cocktail of inflammatory chemokines, cytokines, and growth factors present within the tumor milieu.36,37The recruitment process involves endothelium activation, the monocytes adhering to the endothelial layer of blood vessels, then transmigrating across the endothelial barrier into the parenchyma, differentiating into macrophages and ultimately homing to the tumor location.38Once situated in the tumor microenvironment, macrophage can further polarize and eventually become tumor-associated macrophages.39Different tumor models with immune cell infiltration can be developed using organ-specific cancer cells. Hepatoma cells (Huh7) were used for modelling liver tumors, and MDA-MB-231 cells were selected to model the more aggressive breast tumors, while fibroblast spheroids were prepared as a non-cancerous control (Figure 21a).40Compared to the non-cancerous fibroblasts control, both tumor models exhibited significant macrophage signaling just one day after monocyte recirculation, indicating selective monocyte migration into tumor microenvironments. We quantified monocyte / macrophage signals in blood vessels and tumors for fibroblast, Huh7 and MDA-MB-231 spheroids (Figure 21b, c). After 3 days of monocyte recirculation, fluorescence intensity in MDA-MB-231 and Huh7 spheroids showed 5- and 7- fold increase, respectively, significantly higher than that in the fibroblast group.

[0391] Cancer cells produce distinct cytokine profiles that promote macrophage recruitment

[0392] To identify the chemoattractants responsible for monocyte recruitment and migration, medium perfusates after 7 days of monocyte recirculation were collected for cytokine analysis (Figure 21d-g). GM-CSF and IFN y, involved in monocyte activation, recruitment and migration, were secreted significantly more in the MDA-MB-231 group, with an upward trend observed in the Huh 7 group compared to the non-cancerous fibroblast control.41,42This finding correlates with the increased monocyte migration observed in tumor conditions. Tumor models also exhibited higher secretions of inflammatory cytokines IL-6 and IL-8, which facilitate monocyte transendothelial migration, in contrast to the non-cancerous fibroblast control.43,44The MDA-MB-231 group, in particular, displayed significantly more IL-8 secretion than the Huh7 group. High levels of IL-8 secretion can support tumor metastasis by enhancing the motility and invasiveness of cancer cells, consistent with the known increased metastatic potential of MDA-MB-231 cancer cells and IL-8’s role in establishing a metastasis niche.45,46Further analysis of matrix metalloproteinases (MMPs), which impact the extracellular matrix degradation and immune cell migration, was conducted (Figure 21 h-j) . MMP-3, associated with cancer metastasis and tumor growth in various cancers including breast cancer, showed elevated levels in the MDA-MB-231 group versus the Huh7 group, further confirming the aggressive nature of MDA-MB-231 cancer cells. No significant difference of MMP-2 secretion (Figure 22) was found between non- cancerous group and cancerous groups, likely due to the complex context of tumor microenvironment where both tumor cells and fibroblasts can secrete MMPs, with levels and activity varying significantly. However, a significantly increased level of TMIP-4, which binds to MMP-2 to inhibit the degradation of ECM, was observed in MDA-MB-231 group. Lower of MMP-9 levels in the tumor groups compared to non-cancerous fibroblast group suggest that fibroblasts might be the primary source of MMP-9 production in tissues containing fibroblast spheroids.

[0393] Assessing macrophage polarization and vascular responses in cancer spheroids

[0394] To evaluate macrophage presence and polarization in cancer spheroid models undergoing continuous monocyte recirculation, we performed histological analyses (Figure 21k). Notably, we observed positive CD68 staining, a biomarker for all macrophage types, in both Huh7 and MDA-MB-231 spheroids. This confirms the transformation of monocytes into macrophages and their subsequent infiltration into the tumor microenvironment. In contrast, staining for CD 163 — a marker typically associated with M2 -like macrophages — did not show positive results, suggesting the infiltrated macrophages are likely Ml-type, known for their anti-tumor activities, which are beneficial during the early stages of cancer progression.47Histological analysis reveals that CD68(+) macrophages constitute a substantial portion of the overall cell population within the tissue spheroids. However, contrary to expectations, these macrophages have not led to an increase in spheroid size but have instead contributed to increased density over time. Moreover, there is a tendency for these cells to cluster rather than distribute evenly, leading to a spatial pattern reminiscent of the heterogeneity observed in tumor microenvironments .48

[0395] In addition to the impact of macrophages on the cancer spheroids, the recruited macrophages could also influence the blood vessel. We occasionally observed vessel sprouting from the pre-existing vascular tubule towards cancer spheroids in the MDA-MB-231 cancer model, a characteristic of tumor microenvironments (Figure 211). This sprouting occurred prior to monocyte perfusion and the length of the sprout significantly increased following monocyte recirculation. This increase is likely driven by a VEGF gradient produced by cancer spheroids and infiltrated macrophages, which are known to promote angiogenesis.49,50Even in the absence of sprouting, we consistently observed a decrease in barrier function of blood vessels over time in the presence MDA-MB-231 spheroids, similar to what has been reported previously (Figure 21m).51The increased permeability of blood vessels could be attributed to the high levels of MMP-3 secretion in MDA-MB-231 models, which disrupts ECM protein and intercellular junction and preparing the blood vessel for angiogensis.52Collectively, our immune- infiltrated tumor models successfully replicate the essential interactions within the tumor microenvironment, particularly the interaction between endothelium, stroma, cancer cells and immune components.

[0396] In this study we demonstrated that UniPlate can be used to engineer vascularized cancer spheroids with infiltrated immune cells through recirculating flow. The open-well format of UniPlate enables the easy placement of cancer spheroids to the designated sites using a stamp. This feature allows for the addition of spheroids or organoids at any time point after formation of blood vessels. We showed the inclusion of fibroblasts and pericytes enhances the barrier function of blood vessels. Furthermore, we achieve long-term perfusion of monocytes which can circulate through the engineered blood vessels up to 288 times in a day on average, with a daily replenishment of fresh cells, reflecting the natural lifespan and circulation frequency of monocytes in vivo.38The continuous perfusion and recirculation of immune cells through these vessels leads to the creation of immune -infiltrated tumor models that emulate the complex interactions between blood vessels, stromal cells, and circulating immune cells within tumor microenvironments. Currently, the UniPlate platform accommodates eight tissues per plate, but for higher throughput, platforms such as UniPlate96 and UniPlate384 with more compact well designs could be devised. Unlike platforms that rely on pumps, the design of our platform is scalable. After the addition of cancer spheroids, the blood vessels became inflammatory, showing increased permeability. Immune cells inside these vessels responded to inflammatory signals, such as pro-inflammatory cytokines IFN y, exhibiting rolling, adhesion, and transmigration behaviors. IL-6 and IL-8 are considered critical biomarkers for certain cancer types.53For example, in melanoma patients, elevated levels of IL-6 and IL-8 have been linked to decreased median overall survival.54These cytokines are implicated in the growth, recruitment, and activation of myeloid-derived suppressor cells, which suppress anti-tumor immune responses and aid in the tumor's evasion of immune detection. Our study showed a tissue-specific migration of monocytes towards tumor sites, with an increased secretion of IL-6 and IL-8 in cancer tissues versus the non-cancerous fibroblast control. This selective monocyte trafficking indicates that our immune -infiltrated cancer spheroid models simulated key features of tumor microenvironments. The monocyte differentiation and macrophage infiltration in tumor microenvironments in our models were confirmed by histological analysis, and the infiltrated macrophages are confirmed to be Ml macrophages. Ml macrophages are typically reported to have antitumor effects and present in the early stage of cancer progression. However, they can be transitioned into M2 macrophages promoting cancer progression and metastasis in tumor microenvironments, owing to the plasticity of macrophages.55In this study, THP-1 cells were used to model circulating immune cells. In the future, other immune cells, such as neutrophils, primary peripheral blood mononuclear cells, and T cells that capable of recognizing and eliminating tumor cells can also be included to fully replicate the diverse immune cell population in circulation.

[0397] Conclusions

[0398] In summary, we developed immune -infiltrated cancer spheroid models with long term vascular perfusion and continuous recirculation of immune cells using the UniPlate. The incorporation of stromal cells contributes to the engineering of blood vessels with enhanced barrier function, and a vascularized tumor model was developed by integrating tumor spheroids. Through the perfusion and recirculation of immune cells in vascularized tumor tissues, we confirmed that continuous recirculation of immune cells is crucial for the study of immune responses, offering more opportunities for immune cells to traverse blood vessel and achieve effective immune responses. Additionally, tissue-specific monocyte migration and macrophage infiltration towards tumor sites was demonstrated using three types of spheroids. Overall, our immune-infiltrated cancer spheroid model with long-term recirculating vascular perfusion could be used to recapitulate the essential components of cancer microenvironment for applications in testing immunotherapeutic.

[0399] Materials and Methods

[0400] UniPlate device manufacturing and cell culture

[0401] The UniPlate devices (A004, OrganoBiotech, Inc.) were manufactured through injection molding and 3D printing, as we previously detailed.26Each unit consists two components: a bottomless plate featuring three connecting wells, and an adhesive sheet with printed gelatin templates. Eight units were combined to form a full plate, which was subsequently packaged in a sealing bag for sterilization. Green fluorescent protein-expressing human umbilical vein endothelial cells (GFP-HUVECs, Angio-Proteomie, cat# CAP-0001GFP) were cultured in Endothelial Cell Growth Medium 2 (ECGM2, Sigma- Aldrich, cat# C-22011) according to the manufacturer’s protocol. Human umbilical vein endothelial cells immobilized with telomerase reverse transcriptase (HUVEC / TERT 2, Evercyte, cat# CHT-006-0008) were cultured in ECGM2 medium supplemented with 20 pg / mL G418 solution (InVivoGen. cat# 108321-42-2). Human pericytes from placenta (hPC-PL, Sigma-Aldrich, cat# C- 12980) were cultured in pericytes growth medium 2 (Sigma-Aldrich, cat# C-28041) following the manufacturer’s protocol. Human lung fibroblasts were cultured in Dulbecco's Modified Eagle Medium (DMEM, Thermo Fisher Scientific, cat# 11995065) supplemented with 10% fetal bovine serum (FBS, Thermo Fisher Scientific, cat# 12484028), 1% HEPES (IM, Thermo Fisher Scientific, cat# 15630080), and 1% penicillin-streptomycin glutamine (100X, Thermo Fisher Scientific, cat# 10-378-016). Human breast carcinoma cells (MDA-MB-231, ATCC), kindly provided by Dr. Fei Geng at McMaster University, were cultured in DMEM supplemented with 10% FBS, and 1 % penicillin-streptomycin, and 1% HEPES solution. Human hepatoma cells line (Huh7) cells, a gift from Dr. Shinichiro Ogawa at University of Toronto, were cultured in DMEM supplemented withl0% FBS, 1% penicillin-streptomycin and 1% HEPES solution. THP-1 monocyte cells were cultured in RPMI 1640 medium (Cedarlane Labs, cat# 30-2001) supplemented with 10% FBS and 0.05 mmol / L 2-mercaptoethanol (Sigma-Aldrich, cat# M3148-25ML) in nontreated 6-well plates. For cell passaging, Accutase™ cell dissociation reagent (Thermo Fisher Scientific, cat# Al 110501) was used for detaching pericytes during cell passaging, while 0.05% trypsin-EDTA (Thermo Fisher Scientific, cat# 25300120) was utilized for all other adherent cells. The endothelial cells, fibroblasts, and pericytes employed in this research were between passages 3-7. MDA-MB-231 cells were used at passage numbers ranging from 13-20. For further experimental preparations, cells were plated in cell culture flasks and grown in an incubator with 5% CO2 at 37°C.

[0402] Spheroids generation

[0403] Commercial Aggrewell™ 800 plates were used to generate uniform tumor spheroids, following the supplier’s instructions. Before cell addition, the Aggrewell plates were treated with an anti-adherent rinsing solution (Stemcell Technologies, cat# 07010) for 5 min at room temperature. In this work, 0.6 million MDA-MB-231 cells and 0.15 million fibroblasts were introduced in single well of Aggrewell, resulting in each spheroid comprising a total of 2500 cells, including 2000 MDA-MB-231 cells and 500 fibroblasts. After centrifuging the cells at 200 rpm for 3 min, the Aggrewell plates were placed in the incubator to allow spheroids formation. Within 2 days of incubation, tightly formed spheroids were achieved and subsequently harvested for further use. Huh7 spheroids were produced following the identical procedure with the same number of cells. Fibroblast spheroids, however, were created in the same manner but solely with fibroblasts. The day after spheroid formation, fresh culture medium was added to the Aggrewells to ensure an adequate nutrient supply. All spheroids from a single well were collected into an Eppendorf tube, and after allowing 5 min for static settlement, the supernatant was removed. The spheroids were then washed once with fresh medium before being resuspended in 200 pL of a fibrinogen solution (10 mg / mL).

[0404] Vascular tissue and vascularized tumor tissue culture

[0405] The development of vascular tissues followed the protocol from our previous work, with the modification of including stromal cells within the gel. Fibroblasts and pericytes were mixed in equal parts and resuspended in 10 mg / mL fibrinogen solution (Sigma Aldrich, Cat# F3879-1G) achieving a final density of 0.05 million cells / mL. To prepare the fibrin gel, 100 pL of the fibrinogen solution containing stromal cells was mixed with 20 |1L of thrombin solution (10 U / mL Sigma Aldrich Cat#, T6884-100UN). The mixture was promptly cast into one of the tissue wells, ensuring the printed gelatin template was fully enclosed within the gel. Following overnight incubation at 37 °C, an open lumen was created inside the hydrogel by aspirating the gelatin, which had dissolved in PBS, and then the hydrogel was rinsed with complete medium. Endothelial cells were subsequently seeded into both inlet and outlet wells by adding 50 |1L of cell suspension at a density of 2 million cells / mL to form vascular tissues.

[0406] For the construction of vascularized tumor models, a stamp was employed to create cavities atop the gel for cancer spheroids. The stamp, designed as a cylinder with a 600 pm diameter and a sharp tip measuring 200 pm in diameter, was fabricated using Formlab 3B printer with high temperature resin at a 25 pm printing resolution. The utilization of high temperature resin enabled the sterilization the stamps through autoclaving. The stamp was inserted into the gel 15 min after the fibrin gel was cast. The stamp remained in the tissue well for an additional 20 min to allow further gelation. The gap between the stamp tip and the bottom of the well was precisely set to 40 pm to address the issue of delamination at close distances and the challenges of monocyte migration at larger distances. To avoid hydrogel adhesion to the stamp, the stamps were treated with an anti -adherent rinsing solution for Ih and then air-dried for 30 min. Upon removing the stamp, cavities for the spheroids were successfully created.

[0407] After the formation of confluent blood vessels, the collected spheroids were introduced to the tissues. A fibrin gel, prepared by mixing fibrinogen (10 mg / mL) with the spheroids and thrombin (10 U / mL), was layer over the vascular tissue. Immediately following this addition, the plate was centrifuged at 300 rpm for 10 s to ensure the spheroids settled into the pre-formed spheroid cavities. The gel was then allowed to set at room temperature for 20 min, after which fresh medium containing 1% (v / v) aprotinin (Sigma Aldrich, Cat# 616370-100MG-M) was added to the plate. The plate was subsequently placed on the rocker for culturing. To inhibit the degradation of the fibrin gel over time, all cell culture media were supplemented with 1% (v / v) aprotinin. The medium for all tissues was refreshed on a day basis.

[0408] Flow rate measurements

[0409] To quantify the flow rate in tubular channels, the volume of liquid flowing through the tubular channel to the outlet well over a specified period was collected and measured. Initially, 450 pL of PBS solution containing 1% Aprotinin was added to the inlet well, and 200 pL was added to the tissue well. The plate was then positioned on a rocker set to 15° tilt angle. The liquid accumulated in the outlet well after intervals of 30, 60, 90, 120, 150, 180, 210, 240, 270, 300 s during forward perfusion was collected separately. The dynamic flow rate at each 30-second was subsequently measured.

[0410] Permeability characterization

[0411] To evaluate the permeability of the engineered blood vessels, we perfused the vessels with TRITC-labeled dextran (65 kDa, Sigma Aldrich, cat# T1162). We introduced 200 |iL of culture media into the tissue well and 500 |iL of a TRITC-labeled dextran (1 mg / mL) solution into the inlet well. To visualize the perfusion of the dextran molecule through the blood vessels, a microplate reader (BioTek, Cytation 5) was used to capture time-lapse perfusion images at 30-minute intervals. Subsequently, the diffusive permeability Pa for each blood vessel was calculated using a formula previously reported in the literature.56In this formula, / / and / / denote the average fluorescence intensities at the final and initial time points, while h is the average background fluorescence intensity. At is the time interval used in the experiment, and d is the average diameter of vessel channel.

[0412] Long-term monocyte recirculation

[0413] Tissues were supplied with fresh CellTracker -labeled monocytes daily during the medium change. To demonstrate long-term perfusion and recirculation of immune cells using the UniPlate, 0.4 million of human monocytes (THP-1) were introduced into the inlet well for each tissue. Prior to their addition, THP-1 cells were labeled with the cell tracker red CMTPX (Invitrogen, cat#C34552) following the supplier’s instructions. Typically, THP-1 cells were incubated with a 10 |1M dye solution for 30 min in the incubator, followed by a wash with PBS. Then, the labelled monocytes were resuspended in fresh medium and incubated for additional 5 h before their introduction to the tissues. For the one-time monocyte perfusion condition, the plate was left undisturbed for 1 h after adding the monocytes. Then, the cell suspension in both inlet and outlet wells was aspirated and replaced with fresh medium. Afterwards, the plate was moved to the rocker for subsequent culturing, with daily medium changes for the tissues. Conversely, for continuous monocyte perfusion and recirculation, the plates were maintained on the rocker throughout the experiment. Tissues were supplied with fresh cell tracker -labeled monocytes daily during the medium change.

[0414] Immunostaining

[0415] All tissues were fixed using a 4% (w / v) paraformaldehyde solution for 1 h at room temperature. Next, the tissues were washed three times with PBS buffer and blocked with a blocking solution (10% FBS in PBS) containing 0.1% Triton X-100 for an hour at room temperature. After removal of the blocking buffer, tissues were incubated overnight at 4°C with the primary antibody, anti-VE-Cadherin (Abeam, cat# ab33168), on a rocker. Before imaging with a confocal microscope (ZEISS, 3i Marianas LightSheet), the tissues underwent another 1-2 days of washing with PBS. To ensure the complete removal of all residual primary antibodies, the tissues were washed with PBS over 2 days with the PBS refreshed daily. Subsequently, the tissues were incubated for an hour at room temperature with fluorochrome-conjugated secondary antibodies, F-actin (Cedarlane Labs, cat# 20553-300) and DAPI (Sigma-Aldrich, cat# D9542). Both primary and secondary antibodies were diluted in phosphate -buffered saline (PBS) containing 2% FBS. F-actin was diluted at a 1:1000 ratio in a 2% FBS solution, and DAPI was diluted at a 1:100 ratio in a 2% FBS solution. Before imaging with a confocal microscope (ZEISS, 3i Marianas LightSheet), the tissues underwent another 1-2 days of washing with PBS.

[0416] For histological analysis, tissues after 7 days of continuous monocyte recirculation were fixed in a 10% formalin solution for 48 h before stained and embedded in paraffin wax. Hematoxylin and eosin (H&E) staining was performed, and the selected slides were further stained with CD68 (Agilent, cat# M087601- 2) and CD163 (Abeam, cat# abl82422) antibodies.

[0417] Cytokine analysis

[0418] Cytokine analysis was conducted by Eve Technologies using two specific panels: Human Cytokine Array Proinflammatory Focused 15-plex panel (cat# HDF15) and Human MMPTIMP 13-Plex Discovery Assay (cat# HMMP / TIMP-C,0). Medium perfusates collected from tissues containing fibroblasts, MDA-MB-231, and Huh7 spheroids after 7 days of continuous monocyte recirculation were prepared for cytokine analysis. The collected medium was centrifuged at 1000 G for 10 min at 4°C to separate the supernatants, which were then utilized for the cytokine analysis.

[0419] Data quantification and statistical analysis

[0420] The fold change in fluorescence intensity in blood vessels and at tumor sites was calculated by analyzing the histogram intensity of fluorescence within the selected regions of interest (ROI) for timelapse images using ImageJ software. For statistical analysis, Prism 9 (GraphPad, USA) was employed to test the normality and equality of variance. Depending on these tests' outcomes, either One-way ANOVA or Kruskal-Wallis one-way ANOVA on ranks, combined with Holm-Sidak method at a significant level of p < 0.05, was used to determine statistical significance. Data are presented in all graphs as scatter plots with standard deviation (SD) using Prism 9. To ensure robust quantitative analysis, at least four independent samples (n > 4) were used for each experimental condition.

[0421] References

[0422] 1 Dougan, M. & Dranoff, G. Immune therapy for cancer. Annu Rev Immunol 27, 83-117, doi: 10.1146 / armurev.immunol.021908.132544 (2009).

[0423] 2 Zhang, Y. & Zhang, Z. The history and advances in cancer immunotherapy: understanding the characteristics of tumor-infiltrating immune cells and their therapeutic implications. Cell Mol Immunol 17, 807-821, doi:10.1038 / s41423-020-0488-6 (2020).

[0424] 3 Hendry, S. A. et al. The Role of the Tumor Vasculature in the Host Immune Response: Implications for Therapeutic Strategies Targeting the Tumor Microenvironment. Front Immunol 7, 621, doi: 10.3389 / fimmu.2016.00621 (2016).

[0425] 4 Wu, T. & Dai, Y. Tumor microenvironment and therapeutic response. Cancer Lett 387, 61-68, doi:10.1016 / j.canlet.2016.01.043 (2017). 5 Amersfoort, J., Eelen, G. & Carmeliet, P. Immunomodulation by endothelial cells - partnering up with the immune system? Nat Rev Immunol 22, 576-588, doi:10.1038 / s41577-022-00694-4 (2022).

[0426] 6 Raza, A., Franklin, M. J. & Dudek, A. Z. Pericytes and vessel maturation during tumor angiogenesis and metastasis. Am J Hematol 85, 593-598, doi: 10.1002 / ajh.21745 (2010).

[0427] 7 Motz, G. T. & Coukos, G. Deciphering and reversing tumor immune suppression. Immunity 39, 61-73, doi: 10.1016 / j.immuni.2013.07.005 (2013).

[0428] 8 Shiao, S. L., Ganesan, A. P., Rugo, H. S. & Coussens, L. M. Immune microenvironments in solid tumors: new targets for therapy. Genes Dev 25, 2559-2572, doi:10.1101 / gad.l69029.111 (2011).

[0429] 9 Cekanova, M. & Rathore, K. Animal models and therapeutic molecular targets of cancer: utility and limitations. Drug Des Devel Ther . 1911-1921, doi:10.2147 / DDDT.S49584 (2014).

[0430] 10 Trujillo-de Santiago, G. et al. The Tumor-on-Chip: Recent Advances in the Development of Microfluidic Systems to Recapitulate the Physiology of Solid Tumors. Materials (Basel) 12, doi:10.3390 / mal2182945 (2019).

[0431] 11 Chuprin, J. et al. Humanized mouse models for immuno-oncology research. Nat Rev Clin Oncol 20, 192-206, doi:10.1038 / s41571-022-00721-2 (2023).

[0432] 12 Rosato, R. R. et al. Evaluation of anti-PD-1 -based therapy against triple-negative breast cancer patient-derived xenograft tumors engrafted in humanized mouse models. Breast Cancer Res 20, 108, doi:10.1186 / sl3058-018-1037-4 (2018).

[0433] 13 Li, Z. et al. Application of Animal Models in Cancer Research: Recent Progress and Future Prospects. Cancer ManagRes 13, 2455-2475, doi:10.2147 / CMAR.S302565 (2021).

[0434] 14 Prince, E. et al. Microfluidic Arrays of Breast Tumor Spheroids for Drug Screening and Personalized Cancer Therapies. Adv Healthc Mater, e2101085, doi:10.1002 / adhm.202101085 (2021).

[0435] 15 Miller, C. P., Shin, W., Ahn, E. H., Kim, H. J. & Kim, D. H. Engineering Microphysiological Immune System Responses on Chips. Trends Biotechnol 38, 857-872, doi:10.1016 / j.tibtech.2020.01.003 (2020).

[0436] 16 Kumar, V. & Varghese, S. Ex Vivo Tumor-on-a-Chip Platforms to Study Intercellular Interactions within the Tumor Microenvironment. Adv Healthc Mater 8, el 801198, doi: 10.1002 / adhm.201801198 (2019).

[0437] 17 Sobrino, A. et al. 3D microtumors in vitro supported by perfused vascular networks. Sci Rep 6, 31589, doi:10.1038 / srep31589 (2016).

[0438] 18 Meng, F. et al. 3D Bioprinted In Vitro Metastatic Models via Reconstruction of Tumor Microenvironments. Adv Mater 31, M806899, doi:10.1002 / adma.201806899 (2019).

[0439] 19 Neufeld, L. et al. Microengineered perfusable 3D-bioprinted glioblastoma model for in vivo mimicry of tumor microenvironment. Sci Adv 7, doi: 10.1126 / sciadv.abi9119 (2021). 20 Lee, S. W. L. et al. Integrated in silico and 3D in vitro model of macrophage migration in response to physical and chemical factors in the tumor microenvironment. IntegrBiol (Camb) 12, 90-108, doi: 10.1093 / intbio / zyaa007 (2020).

[0440] 21 Wang, H. F. et al. Tumor-Microenvironment-on-a-Chip for Evaluating Nanoparticle-

[0441] Loaded Macrophages for Drug Delivery. ACS Biomater Sci Eng 6, 5040-5050, doi : 10.1021 / acsbiomate rials .0c 00650 (2020) .

[0442] 22 Aung, A., Kumar, V., Theprungsirikul, J., Davey, S. K. & Varghese, S. An Engineered Tumor-on-a-Chip Device with Breast Cancer-Immune Cell Interactions for Assessing T-cell Recruitment. Cancer Res 80, 263-275, doi: 10.1158 / 0008-5472.CAN-19-0342 (2020).

[0443] 23 Boussommier-Calleja, A. et al. The effects of monocytes on tumor cell extravasation in a 3D vascularized microfluidic model. Biomaterials 198, 180-193, doi: 10.1016 / j. biomaterials.2018.03.005 (2019).

[0444] 24 Sharifi, F. et al. A Foreign Body Response-on-a-Chip Platform. Adv Healthc Mater 8, 61801425, doi:10.1002 / adhm.201801425 (2019).

[0445] 25 Chi, C. W. et al. Enabling continuous immune cell recirculation on a microfluidic array to study immunotherapeutic interactions in a recapitulated tumour microenvironment. Lab Chip 24, 396-407, doi:10.1039 / d31c00662j (2024).

[0446] 26 Zhang, F., Lin, D. S. Y., Rajasekar, S., Sotra, A. & Zhang, B. Pump-Less Platform Enables Long-Term Recirculating Perfusion of 3D Printed Tubular Tissues. Adv Healthc Mater, e2300423, doi: 10.1002 / adhm.202300423 (2023).

[0447] 27 Rajasekar, S. et al. Subtractive manufacturing with swelling induced stochastic folding of sacrificial materials for fabricating complex perfusable tissues in multi-well plates. Lab Chip 22, 1929- 1942, doi: 10.1039 / dllc01141c (2022).

[0448] 28 Klarhofer, M., Csapo, B., Balassy, C., Szeles, J. C. & Moser, E. High-resolution blood flow velocity measurements in the human finger. Magn Reson Med 45, 716-719, doi:10.1002 / mrm.l096 (2001).

[0449] 29 Mathur, T. et al. Organ-on-chips made of blood: endothelial progenitor cells from blood reconstitute vascular thromboinflammation in vessel-chips. Lab Chip 19, 2500-2511, doi: 10.1039 / c91c00469f (2019).

[0450] 30 Homan, K. A. et al. Bioprinting of 3D Convoluted Renal Proximal Tubules on Perfusable Chips. Scientific Reports 6, doi: 10.1038 / srep34845 (2016).

[0451] 31 Yuan, W., Lv, Y., Zeng, M. & Fu, B. M. Non-invasive measurement of solute permeability in cerebral microvessels of the rat. Microvasc Res 77, 166-173, doi:10.1016 / j. mvr.2008.08.004 (2009).

[0452] 32 Kosyakova, N. et al. Differential functional roles of fibroblasts and pericytes in the formation of tissue-engineered microvascular networks in vitro. NPJ Regen Med 5, 1, doi: 10.1038 / s41536- 019-0086-3 (2020). 33 Herland, A. et al. Distinct Contributions of Astrocytes and Pericytes to Neuroinflammation Identified in a 3D Human Blood-Brain Barrier on a Chip. PLoS One 11, e0150360, doi: 10.1371 / joumal.pone.0150360 (2016).

[0453] 34 Patel, A. A. et al. The fate and lifespan of human monocyte subsets in steady state and systemic inflammation. J Exp Med 214, 1913-1923, doi: 10.1084 / jem.20170355 (2017).

[0454] 35 Tarr, L., Oppenheimer, B. S. & Sager, R. V. The circulation time in various clinical conditions determined by the use of sodium dehydrocholate. American Heart Journal 8, 766-786 (1933).

[0455] 36 Ley, K., Laudanna, C., Cybulsky, M. I. & Nourshargh, S. Getting to the site of inflammation: the leukocyte adhesion cascade updated. Nat Rev Immunol 7, 678-689, doi:10.1038 / nri2156 (2007).

[0456] 37 Kratofd, R. M., Kubes, P. & Deniset, J. F. Monocyte Conversion During Inflammation and Injury. Arterioscler Thromb Vase Biol 37, 35-42, doi:10.1161 / ATVBAHA.H6.308198 (2017).

[0457] 38 Yona, S. et al. Fate mapping reveals origins and dynamics of monocytes and tissue macrophages under homeostasis. Immunity 38, 79-91, doi:10.1016 / j.immuni.2012.12.001 (2013).

[0458] 39 Akins, E. A., Aghi, M. K. & Kumar, S. Incorporating Tumor-Associated Macrophages into Engineered Models of Glioma. iScience 23, 101770, doi: 10.1016 / j.isci.2020.101770 (2020).

[0459] 40 Parekh, A. et al. Bioimpedimetric analysis in conjunction with growth dynamics to differentiate aggressiveness of cancer cells. Sci Rep 8, 783, doi: 10.1038 / s41598-017-18965-9 (2018).

[0460] 41 Gocher, A. M., Workman, C. J. & Vignali, D. A. A. Interferon-gamma: teammate or opponent in the tumour microenvironment? Nat Rev Immunol 22, 158-172, doi: 10.1038 / s41577-021- 00566-3 (2022).

[0461] 42 Shi, Y. et al. Granulocyte -macrophage colony -stimulating factor (GM-CSF) and T-cell responses: what we do and don't know. Cell Res 16, 126-133, doi:10.1038 / sj.cr.7310017 (2006).

[0462] 43 Meniailo, M. E. et al. Interleukin-8 favors pro-inflammatory activity of human monocytes / macrophages. Int Immunopharmacol 56, 217-221, doi:10.1016 / j.intimp.2018.01.036 (2018).

[0463] 44 Weissenbach, M. et al. Interleukin-6 is a direct mediator of T cell migration. Eur J Immunol 34, 2895-2906, doi:10.1002 / eji.200425237 (2004).

[0464] 45 Fousek, K., Hom, L. A. & Palena, C. Interleukin-8: A chemokine at the intersection of cancer plasticity, angiogenesis, and immune suppression. Pharmacol Ther 219, 107692, doi:10.1016 / j.pharmthera.2020.107692 (2021).

[0465] 46 Steele, C. W. et al. CXCR2 Inhibition Profoundly Suppresses Metastases and Augments

[0466] Immunotherapy in Pancreatic Ductal Adenocarcinoma. Cancer Cell 29, 832-845, doi:10.1016 / j.ccell.2016.04.014 (2016).

[0467] 47 Chanmee, T., Ontong, P., Konno, K. & Itano, N. Tumor-associated macrophages as major players in the tumor microenvironment. Cancers (Basel) 6, 1670-1690, doi: 10.3390 / cancers6031670 (2014). 48 Yuan, Y. Spatial Heterogeneity in the Tumor Microenvironment. Cold Spring Harb PerspectMed , doi:10.1101 / cshperspect.a026583 (2016).

[0468] 49 Ferrara, N. VEGF and the quest for tumour angiogenesis factors. Nat Rev Cancer 2, 795- 803, doi:10.1038 / nrc909 (2002).

[0469] 50 Lamagna, C., Aurrand-Lions, M. & Imhof, B. A. Dual role of macrophages in tumor growth and angiogenesis. J Leukoc Biol 80, 705-713, doi: 10.1189 / jlb.1105656 (2006).

[0470] 51 McDonald, D. M. & Baluk, P. Significance of blood vessel leakiness in cancer. Cancer Res 62, 5381-5385 (2002).

[0471] 52 Kugeratski, F. G., Santi, A. & Zanivan, S. Extracellular vesicles as central regulators of blood vessel function in cancer. Science Signaling 15, eaaz4742 (2022).

[0472] 53 Tobin, R. P. et al. IL-6 and IL-8 Are Linked With Myeloid-Derived Suppressor Cell Accumulation and Correlate With Poor Clinical Outcomes in Melanoma Patients. Front Oncol 9, 1223, doi: 10.3389 / fonc.2019.01223 (2019).

[0473] 54 Sanmamed, M. F. et al. Changes in serum interleukin-8 (IL-8) levels reflect and predict response to anti-PD-1 treatment in melanoma and non-small-cell lung cancer patients. Ann Oncol 28, 1988- 1995, doi: 10.1093 / annonc / mdxl90 (2017).

[0474] 55 Boutilier, A. J. & Elsawa, S. F. Macrophage Polarization States in the Tumor Microenvironment. IntJMol Sci 22, doi: 10.3390 / ijms22136995 (2021).

[0475] 56 Kolesky, D. B., Homan, K. A., Skylar-Scott, M. A. & Lewis, J. A. Three-dimensional bioprinting of thick vascularized tissues. Proc Natl Acad Sci U S A 113, 3179-3184, doi: 10.1073 / pnas.1521342113 (2016).

Claims

CLAIMS:

1. A pumpless recirculating platform for substantially unidirectional fluid flow.

2. The platform of claim 1, wherein the fluid flow is driven by gravity.

3. The platform of claim 2, wherein the gravity -driven fluid flow is controlled by a rocker.

4. The platform of claim 3, wherein the rocker has a programmable dwell time at different tilt angles.

5. The platform of any one of claims 1 to 4, comprising an inlet well, a tissue well, and an outlet well in substantially unidirectional fluid communication with one another.

6. The platform of claim 5, wherein a first channel connects the inlet well and the tissue well in fluid communication and a second channel connects the tissue well and the outlet well in fluid communication.

7. The platform of claim 5 or 6, wherein a slanted bridge connects the inlet well and the outlet well in fluid communication, wherein the slanted bridge is lower at the outlet well than at the inlet well to encourage unidirectional flow, such that: when the platform is tilted in a first direction, the higher end of the slanted bridge prevents fluid flow from the inlet well to the outlet well and encourages fluid flow from the inlet well, to the tissue well, to the outlet well; and when the platform is tilted in a second direction, the lower end of the slanted bridge encourages fluid flow from the outlet well to the inlet well over fluid flow in the opposite direction.

8. The platform of claim 7, wherein the slanted bridge has an angle of about 20°.

9. The platform of any one of any one of claim 5 to 8, wherein fluid flows from the inlet well to the tissue well, to the outlet well, and back to the inlet well.

10. The platform of any one of claims 5 to 9, wherein the tissue well is bottomless.

11. The platform of any one of claims 5 to 10, wherein the inlet well and / or the outlet well is bottomless.

12. The platform of any one of claim 1 to 11 in combination with a plate for holding the platform.

13. The platform of claim 12, wherein the plate holds a plurality of platforms.

14. The platform of claim 13, wherein the plate holds from about 8 platforms to about 128 platforms in standard 24 to 384 well format.

15. The platform of any one of claims 1 to 14 in combination with a sacrificial material for template production.

16. The platform of claim 15, wherein the sacrificial material is dissolvable.

17. The platform of claim 15 or 16, wherein the sacrificial material comprises gelatin, pluronic acid, alginate, or any combination thereof.

18. The platform of any one of claims 15 to 17, wherein the sacrificial material is 3D printed on an adhesive sheet for application to an open bottom of the platform.

19. The platform of any one of claims 15 to 18, wherein the sacrificial material creates one or more channels between the inlet well and the outlet well.

20. The platform of claim 19, wherein the one or more channels comprise vascular channels for seeding of endothelial cells.

21. The platform of any one of claims 1 to 20, in combination with a substrate upon which to seed cells.

22. The platform of claim 21, wherein the substrate comprises fibrin gel, collagen, Matrigel, an extracellular matrix (ECM) hydrogel, or any combination thereof.

23. The platform of any one of claims 1 to 22 in combination with cells.

24. The platform of claim 23, wherein the cells form tubular tissues such as vascular tissues or renal proximal tubules.

25. The platform of claim 23 or 24, wherein the cells form spheroids.

26. The platform of claim 23 or 24, wherein the cells comprise immune cells, endothelial cells, cancer cells, epithelial cells, stromal cells, or any combination thereof.

27. The platform of any one of claims 1 to 26 providing a shear stress of from about 0.7 to about 2 dyne / cm2.

28. The platform of any one of claims 1 to 27, providing a fluid flow rate of from about 100 to 200 |iL / min.

29. A method for making the platform of any one of claims 1 to 28, the method comprising applying a sacrificial template to an injection -molded bottomless well.

30. A method of perfusing cells, the method comprising seeding cells on the platform of any one of claims 1 to 28 and recirculating fluid substantially unidirectionally without the use of a pump.

31. The method of claim 30, wherein cancer spheroids are seeded on the platform adjacent to a vascular tube, through which the fluid is recirculated, wherein the fluid comprises immune cells.

32. The method of claim 30 or 31, wherein a tilt angle of from about - 20° to about - 60°, such as about - 45°, is used for a recirculation tilt angle and wherein a tilt angle of from about 5° to about 45°, such as about 15°, is used for a perfusion tilt angle.

33. An epithelium model with patterned crypts and underlying perfusable micro vasculature.

34. The model of claim 33, wherein the crypts are in an open well providing access to the apical surface of the epithelium.

35. The model of claim 33 or 34, vascularized with self-assembled endothelial cells and fibroblasts to yield a tissue microenvironment with vascular-epithelial crosstalk.

36. The model of any one of claims 33 to 35, wherein the patterned crypts are imprinted on a hydrogel scaffold.

37. The model of any one of claims 33 to 36, wherein the model is an intestinal model, a skin hair follicle model, a lung alveoli model, or a mammary gland model.

38. The model of claim 37, wherein the model is a colon model.

39. An immune-infiltrated cancer spheroid model made by continuously perfusing and recirculating immune cells with gravity -driven flow through a tubular blood vessel adjacent to a cancer spheroid.

40. A cancer-macrophage heterospheroid model comprising a macrophage population that continuously grows, renews, and / or self-evolves.

41. An immune-infiltrated cancer spheroid model comprising a vascular tube through which immune cells unidirectionally flow and a cancer spheroid adjacent to the vascular tube.

42. The model of claim 41, further comprising stromal cells.

43. The model of any one of claims 33 to 42, comprising a cytokine gradient.

44. A stamp for imprinting a patterned crypt on a crosslinked hydrogel scaffold.

45. The stamp of claim 44, wherein the stamp is 3D printed.

46. The stamp of claim 44 or 45, wherein the stamp is formed from high temperature resin that is autoclavable.

47. The stamp of any one of claims 44 to 46, wherein the stamp comprises high aspect ratio features.

48. The stamp of claim 47, wherein the high aspect ratio features comprise an array of pyramidal pin structures printed at the tip of the stamp.

49. The stamp of any one of claims 44 to 48, wherein the stamp surface is functionalized to reduce adherence of the stamp to the scaffold.

50. The stamp of claim 49, wherein the stamp surface is polyethylene glycol (PEG) photografted or coated with an anti-adherence rinsing solution.

51. A method of patterning a crosslinked hydrogel, the method comprising applying the stamp of any one of claims 44 to 50 to the crosslinked hydrogel and then removing the stamp.

52. An open chamber for cell culture, the chamber comprising a basolaterally perfused compartment and an apically perfused compartment.

53. The open chamber of claim 52, wherein the flow of fluid through both the basolaterally perfused compartment and the apically perfused compartment is gravity -driven.

54. The open chamber of claim 52 or 53, wherein the basolaterally perfused compartment comprises a basolateral inlet and a basolateral outlet for fluid flow.

55. The open chamber of claim 54, wherein the basolateral inlet leads to a basolateral inlet chamber and / or the basolateral outlet leads to a basolateral outlet chamber.

56. The open chamber of claim 55, wherein the basolateral inlet chamber and / or the basolateral outlet chamber are open or closed.

57. The open chamber of any one of claims 52 to 56, wherein the apically perfused compartment comprises an apical inlet and an apical outlet for fluid flow.

58. The open chamber of claim 57, wherein the apical inlet leads to an apical inlet chamber and / or the apical outlet leads to an apical outlet chamber.

59. The open chamber of any one of claims 52 to 58, wherein flow of fluid in the basolaterally perfused compartment is substantially orthogonal to the flow of fluid in the apically perfused compartment.

60. The open chamber of any one of claims 52 to 59, wherein the apically perfused compartment is open for access to the apical compartment.

61. The open chamber of any one of claims 52 to 60, comprising a hydrogel to separate the basolaterally perfused compartment and the apically perfused compartment.

62. The open chamber of any one of claims 52 to 61, wherein the basolaterally perfused compartment and the apically perfused compartment are not separated by a membrane, such as a synthetic membrane.

63. The open chamber of any one of claims 52 to 62, wherein the open chamber comprises a crypt- patterned base.

64. A platform comprising: a central open chamber for cell culture; a basolateral inlet chamber and a basolateral outlet chamber each adjacent to and in fluid communication with the open chamber; and an apical inlet chamber and an apical outlet chamber each adjacent to and in fluid communication with the open chamber; wherein the basolateral inlet and outlet chambers are not in fluid communication with the apical inlet and outlet chambers.

65. The platform of claim 64, comprising a second apical inlet chamber adjacent to and in fluid communication with the apical inlet chamber and a second apical outlet chamber adjacent to and in fluid communication with the apical outlet chamber.

66. The platform of claim 64 or 65, wherein the open chamber comprises a crypt-patterned base.

67. A multi-well plate comprising the open chamber of any one of claims 52 to 63 or platform of any one of claims 64 to 66.

Images