WO2022229915A1 - Fluid gel support medium for embedded printing, methods and uses thereof - Google Patents

Abstract

The present disclosure relates to a fluid gel support medium for embedded printing wherein the fluid gel support medium comprises 1 to 4 wt.% of negatively charged cellulose nanocrystals; and at least 0.5 mM of a cation; wherein the fluid gel support medium is self-assemblable into a fibrillar matrix by the addition of a hardening solution comprising 5 to 100 mM of cation. The use of the disclosed fluid gel as a support medium for embedded printing, and a method for embedded printing using said fluid gel support medium are also disclosed. The disclosure also relates to a kit comprising the fluid gel support medium for embedded printing, and to an organ-on-a-chip, tumor-on-a-chip, or tissue-on-a-chip model obtainable from the disclosed method for embedded printing.

Description

FLUID GEL SUPPORT MEDIUM FOR EMBEDDED PRINTING, METHODS

AND USES THEREOF

TECHNICAL FIELD

[0001] The present disclosure relates to a self-assembled fibrillar matrix for micro to macroscale biofabrication, in particular automated biofabrication of in vitro tissue/organ models, physiologically relevant ex vivo models, or other regenerative and personalized medicine applications.

BACKGROUND

[0002] Organ/tissue-on-chip (OoC) technology has seen increasing interest in the development of microphysiological systems with improved predictive power for their in vitro-in vivo extrapolations. This technology is developing rapidly in order to meet the increasing demands for reducing animal testing in basic research as well as in drug development processes. However, the complexities of organ functions are difficult to model in these systems. Many different fabrication approaches have been devised in terms of model complexity and device designs.

[0003] Among the most successful platforms currently translated in commercial products are those known as tissue plates, consisting of arrays of tissue chips for on- plate replicates. ^[1] Similarly to the majority of other OoC platforms, these plates are based on predefined sample housing circuits made of transparent polymeric materials such as poly(methyl methacrylate) or polydimethylsiloxane (PDMS). Despite mitigating the issue of sample replicate throughput, such platforms tend to sacrifice the potential to recreate the cell-cell and cell-extracellular matrix (ECM) cross-talk existing in living organisms.

[0004] In vivo, parenchyma and its corresponding stromal tissues are primarily made of non-cellular interstitial fibrillar ECM that support the function of the cellular constituents. ^[2] Cells and ECM are hierarchically organized by basement membranes (BM) that divide them into compartments. ^[3] Structurally, this specialized form of ECM consists of dense 3D fibrillar networks that, despite having a tissue-specific composition, conserves their nanoscale topographical features across anatomical locations. ^[4] Beyond the signaling derived from the biochemical nature of their components, the biophysical cues stemming from the nanoscale supramolecular organization of ECM and BM have a far more important biological function than simply providing structural support for cells. These fibrillar architectures of ECMs convey mechanical and topological signals that have many recognized roles in the regulation of important cellular mechanisms such as migration, mechanotransduction, or morphogenesis ^[2]. In addition to biophysical cues, the permeability allowed by the interstitial space of fibrous ECMs is also crucial for regulating cell signaling because it controls the permeation and mass transport of bioactive macromolecules, nutrients, and wastes.

[0005] A few examples of granular microgels ^[5] or shear-thinning macromolecular hydrogels systems based on specific chemical modifications for reversible crosslinking ^[6] have been proposed as support materials for the direct writing and long term maturation of embedded 3D cellular constructs. Although these systems could be considered as alternatives, they do not recapitulate the filamentous nature of the ECM.

[0006] The inability of current OoC models to recreate the 3D fibrillar nature of the compartmentalization barriers and the interstitial spaces where biological processes occur is a major limitation of these systems. Moreover, tissues and organs have 3D structures with architectures on different length scales crucial for their complex biological function, ^[7] which are difficult to reproduce on current OoC platforms usually based on circuits with fixed predefined geometries. Additionally, many conventional microfluidic devices often require cumbersome microfabrication manufacturing processes and complicated cell and biomaterial handling procedures that have not yet proven to be sufficiently efficient, feasible, or economical. Therefore, combinations of fabrication strategies and biomaterial platforms enabling the automated manufacture of 3D cellular constructions while providing the intrinsic barrier function of fibrillar ECMs and BMs would be a major advance in the field. By allowing the creation of OoC systems with more biomimetic cell-cell and cell-matrix crosstalk, such a platform would contribute to the construction of complex compartmentalized 3D tissue models with higher physiological relevance. [0007] A key requirement to enable the printing of high-resolution constructions applying matrix-assisted 3D bioprinting technologies is the use of support baths with adequate rheological behavior capable of providing a semisolid medium to print into.^[8] Self-assembling materials, of either natural or synthetic origin, offer unique advantages to provide a rich landscape of ECM-mimetic nanoscale filamentous structures, networks, and pores. ^[9] If rationally integrated with biofabrication technologies such as 3D bioprinting, these man-made supramolecular hydrogels can be leveraged for the bottom-up assembly of complex 3D structures with well-defined nano-to-macroscale hierarchical architectures.

[0008] These facts are disclosed in order to illustrate the technical problem addressed by the present disclosure.

GENERAL DESCRIPTION

[0009] The present disclosure relates to a bottom-up strategy for the direct manufacture of cell-laden devices that recreate the unique biophysical cues from the native fibrillar ECMs, while allowing the design of embedded bioengineered microtissues with arbitrary geometries.

[0010] In an embodiment, the disclosed platform combines the concepts of matrix- assisted 3D free-form bioprinting ^[8] with the controlled self-assembly of colloidal rod shaped cellulose nanocrystals (CNC) as building blocks to fabricate cell-laden constructs embedded within its own fibrillar hydrogel device. For that, hydrogel formation has to occur around cells and therefore fiber assembly and gelation mechanisms have to be cytocompatible.

[0011] In an embodiment, cellulose nanocrystals (CNC) in combination with commonly available 3D bioprinting can create microphysiological constructs embedded in a fibrillar support material with microfluidic channels, ECM biomimetic topography, and biomolecule permeability while maintaining optical transparency. The self-assembled CNC matrix with nanoscaled filamentous network allows liberty of choice of bioink independently of its viscosity, to host and combine different cell types, tailored permeability for biomacromolecules exchange, and exceptional structural stability in long-term culture to allow cell maturation.

[0012] In another embodiment, the self-assembled CNC matrix of the present disclosure is compatible with the fabrication of perfusable designs for in situ endothelization and compartmentalization, which allows tissue and organ-on-chip applications with the capability of real-time optical monitoring.

[0013] Current solutions for the direct writing and long term maturation of embedded 3D cellular constructs, including granular microgels ^[5] or shear-thinning macromolecular hydrogels systems ^[6], do not recapitulate the filamentous nature of the ECM, among other comparative disadvantages to the disclosed CNC-based hydrogels in terms of synthesis complexity and cost of implementation.

[0014] In an embodiment, the rheological properties of the CNC fluid gels can be optimized to meet the demands of matrix-assisted 3D bioprinting technology. The self- assembly of the CNC fluid gels into fibrillary hydrogel devices after printing can also be controlled to obtain the desired complex 3D structures. The combination of these hierarchical support material and fabrication technology for creating microfluidic channels and multicellular 3D microtissues embedded within a biomimetic fibrillar matrix is demonstrated, facilitating the rapid generation of large numbers of microphysiological systems that can be leveraged for in vitro modeling in different assay formats.

[0015] In an embodiment, the implementation of the disclosed supporting material was tested with various cell types and bioinks, and their biological performance was evaluated in both static and dynamic conditions. The range of comparative advantages offered by this system further provides a potential solution to scale-up for bioengineered tissue manufacture.

[0016] Compared to other nanocellulose-based paper devices fabricated applying similar matrix assisted 3D printing concepts, the disclosed supporting material avoids the need for post-printing construct drying to confer structural cohesion to the device or the need for high temperature treatments and extensive organic solvent washing for the removal of hydrophobic sacrificial inks. On the other hand, compared to PDMS- based microfluidic devices, the most popular platform used in the field, it bypasses the need for multiple and tedious microfabrication steps required for their manufacturing. Moreover, CNC are derived from renewable and green sources providing sustainable/green chemistry alternative to existing plastic polymers widely in used for fabricating microfluidic devices.

[0017] In an embodiment, the CNC-based platform described in the present disclosure supports high-resolution printing of perfusable microfluidic channels and embedded constructs with arbitrary freeform 3D shapes using different low viscosity hydrogel bioinks and cell types.

[0018] In another embodiment, the controlled self-assembly of CNC after printing induces their fibrillation into networks that recreate the characteristic topography of native ECMs and allows the easy interstitial diffusion of macromolecules.

[0019] In another embodiment, the CNC-based platform enables the direct writing of microtissues in the 3D space of a perfusable bioinspired housing material that allows diffusion of signaling biomolecules for cell-cel I communications, using a simple extrusion 3D bioprinter without the need for specific microfabrication processes, equipment, or skills. The automated nature of the biofabrication platform further provides significant advantages of throughput, reproducibility, and scalability for the manufacturing of miniaturized multicellular systems with complex bioinspired 3D architectures.

[0020] In an embodiment, the disclosed CNC-based support medium is transparent for real-time monitoring and embedded tissues can be easily harvested under mild biocompatible conditions, by enzymatic digestion with cellulases for further offline processing.

[0021] The present disclosure relates to a fluid gel support medium for embedded printing wherein the fluid gel support medium comprises: 1 to 4 wt.% of negatively charged cellulose nanocrystals; and at least 0.5 mM of a cation; wherein the fluid gel support medium is self-assemblable into a fibrillar matrix by the addition of a hardening solution comprising 5 to 100 mM of a cation, preferably 6 to 10 mM of cation. [0022] In an embodiment, the cation is selected from a list comprising: Ca²⁺, Zn²⁺, Mg²⁺, Mn²⁺, Sr²⁺, Cu²⁺, Fe³⁺, Al³⁺ or mixtures thereof. Preferably, the cation is a bivalent cation. More preferably, the cation is Ca²⁺.

[0023] In an embodiment, the cellulose nanocrystals are grafted with sulfate groups.

[0024] In an embodiment, the fluid gel support medium is optically transparent. In another embodiment, the fibrillar matrix is optically transparent.

[0025] For the scope and interpretation of the present disclosure the term "optically transparent" is defined as the physical property of allowing light to pass through the material without appreciable scattering of light, resulting in a high transmit of the light that falls on them with little reflection.

[0026] In an embodiment, the cellulose nanocrystals are rod-shaped. In another embodiment, the cellulose nanocrystals have a size ranging from 70 to 2000 nm in length and 1 to 20 nm in width, preferably from 157 to 190 nm in length and 3 to 6 nm in width.

[0027] In an embodiment, the fibrillar matrix comprises fibrils with a diameter ranging from 10 to 100 nm, preferably from 21 to 32 nm.

[0028] In an embodiment, the pore size of the fibrillar matrix ranges from 40 to 300 nm, preferably from 74 to 96 nm.

[0029] In an embodiment, the fluid gel comprises 1.5 to 3 wt.% of cellulose nanocrystals, preferably 2.5 wt.% of cellulose nanocrystals; and 1.5 to 2.5 mM of Ca²⁺, preferably 2 mM of Ca²⁺; wherein the cellulose nanocrystals self-assemble into a fibrillar matrix by the addition of a hardening solution comprising 6 to 8 mM of Ca²⁺, preferably 7.5 mM of Ca²⁺.

[0030] In an embodiment, the fluid gel support medium further comprising an enzyme as a crosslinking agent of a printing ink.

[0031] The present disclosure also relates to a use of the disclosed fluid gel as a support medium for embedded printing, preferably embedded bioprinting.

[0032] An aspect of the present disclosure relates to a method for embedded printing using the disclosed fluid gel support medium, comprising the following steps: providing the fluid gel support medium comprising 1 to 4 wt.% of cellulose nanocrystals and 0.5 to 3 mM of a cation, preferably Ca²⁺; disposing, using an extrusion tip configured to travel along a predefined course through the fluid gel support medium, a volume of at least one polymeric ink into the fluid gel support medium, wherein the volume of the at least one polymeric ink disposed along the predefined course is retained within the fluid gel support medium; inducing the self-assembling of the fluid gel support medium into a fibrillar matrix by addition of a hardening solution comprising cations, preferably a 5 to 8 mM Ca²⁺ solution.

[0033] In an embodiment, the volume of the at least one polymeric ink is disposed using an extrusion printer coupled with a pneumatic-based printhead, a piston-based printhead, a screw-based printhead, or combinations thereof; preferably a pneumatic- based printhead.

[0034] In an embodiment, the volume of the at least one polymeric ink is disposed using a pressure ranging from 1 to 300 kPa. In another embodiment, the volume of the at least one polymeric ink is disposed using a speed ranging from 1-20 mm.s ¹.

[0035] In an embodiment, the at least one polymeric ink is selected from a list comprising: gelatin, alginate, methacrylated gelatin, Pluronic F-127, collagen, fibrin, hyaluronic acid, gellan gum, silk fibroin, platelet lysate, matrigel, polyethylene glycol), polyethylene glycol) diacrylate or mixtures thereof.

[0036] In an embodiment, the at least one polymeric ink further comprises a cell, a therapeutic agent, a growth factor, an enzyme, a metallic element, a cytokine, a contrasting agent, a cell receptor, a cell ligand, or combinations thereof.

[0037] In an embodiment, the polymeric ink is a sacrificial polymeric ink. In a further embodiment, the method further comprises the following steps allowing the volume of the sacrificial polymeric ink to at least partially solidify; and removing the volume of the at least partially solidified sacrificial polymeric ink from the fibrillar matrix to form one or more voids substantially similar in volume and form to the volume of the at least partially solidified sacrificial polymeric ink. [0038] In an embodiment, the method further comprises a step of digesting the fibrillar matrix by enzymatic digestion of the cellulose nanocrystals, preferably by enzymatic hydrolysis with cellulase at 37 °C.

[0039] The present disclosure also relates to a kit comprising the disclosed fluid gel support medium for embedded printing.

[0040] An aspect of the present disclosure relates to an organ-on-a-chip, tumor-on-a- chip, ortissue-on-a-chip model obtainable from the disclosed method.

BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The following figures provide preferred embodiments for illustrating the disclosure and should not be seen as limiting the scope of invention.

[0042] Figure 1: Schematic representation of an embodiment of the fabrication of bioprinted microphysiological systems embedded within the bioinspired hierarchical fibrillar matrix of the present disclosure. A) Colloidal suspensions of CNC with defined rheology are used as support fluid gels to fabricate SD constructs with arbitrary geometries. B) The printed constructs are then locked within a fibrillar matrix by inducing CNC self-assembly with addition of calcium ions. C) Bioprinted constructs embedded within the fibrillar matrix are cultured for in vitro maturation and bioassays. This platform enables possible application as: D) housing support materials for tissue and organ-on-chip microfluidic device; E) on-plate 3D bioprinted models for medium- to high-throughput assay replicates; F) temporary templating bioreactors for in vitro maturation of 3D bioprinted tissue constructs that can be harvested through bioorthogonal digestion.

[0043] Figure 2: Scanning electron microscopy images (SEM) of an embodiment of A) CNC showing its rod-shaped morphology, and B) self-assembled hydrogel showing hierarchical arrangements of CNC into fibrillar networks (scale bar: 200 nm).

[0044] Figure 3: Embodiment of an inversion test for gelation of CNC with variable concentration of calcium chloride (0, 0.5, 1.0, 1.5, 2.0 and 2.5 mM CaCh) A) CNC 1.5 wt. %; B) CNC 2 wt.%; C) CNC 2.5 wt. %. Dotted red square represents a colloidal suspension of CNC 2.5 wt. % with 2 mM Ca ²⁺, that does not form a firm gel like 2.5 mM Ca ² but also is not too liquid like 1.5 mM Ca ²⁺.

[0045] Figure 4: A) Printability window of CNC wt.% vs ionic strength and transmitted light images showing the resolution of a fabricated grid pattern (20 % infill) with gelatin 10 wt. % solution, printed in 2.5 wt.% CNC fluid gels at variable Ca²⁺ concentration and using a 27G needle (food color was added to aid visualization). B) Strain sweep test with Storage modulus (G') and loss modulus (G") of CNC self-assembled colloidal hydrogel 2.5 wt.% with 2 mM ion concentration (square) and 7.5 mM ion concentration (triangle) where (G') solid symbols (G") hollow symbols. C) Structural stability of CNC fluid gel crosslinked with 7.5 mM ion concentration to form a fibrillar hydrogel Day 1 vs Day 28 (the pink color results from CNC fibrillar matrix incubation in cell culture media with phenol red) (Scale bar: 1 mm).

[0046] Figure 5: Embodiment of A) shear viscosity of the 2.5 wt.% CNC suspensions at variable concentration of Ca²⁺ ions: 0 mM (dotted line), 2.5 mM (full line) and 7.5 mM (dashed line); and B) evaluation of self-healing properties of CNC hydrogel by subjecting the 2.5 wt.% CNC with 2 mM Ca²⁺ to alternating high (200 %) and low (1 %) strain cycles ( G storage modulus, solid symbols, G": loss modulus, hollow symbols).

[0047] Figure 6: Embodiment of a CAD design and respective photographs of two layered linear patterns and a continuously extruded spiral printed in CNC colloidal matrix with 5 wt. % gelatin (GelTG, enzymatic crosslinking), 2 wt. % alginate (ionic-crosslinking) and 5 wt. % GelMA (photo-crosslinking).

[0048] Figure 7: Embodiment of confocal images comparing the resolution of fluorescently labeled (FITC) gelatin ink printed within A) 0.5 wt.% agarose slurry and B) 2.5 wt.% CNC colloidal suspension (scale bar: 250 pm). Embodiment of photographs of lattice structures with 10x10x1 mm and 20% rectilinear infill printed in C) 0.5 wt. % agarose slurry and D) 2.5 wt.% CNC colloidal suspension (scale bar:l mm). Embodiment of photographs showing the optical transparency of E) agarose bath, and F) CNC colloidal suspension. G) Resolution of FITC-labelled gelatin 5 wt.% ink printed at a constant pressure of 5 kPa and variable speed of B mm.s ¹, 5 mm.s ¹ and 8 mm.s ¹ with a 30G nozzle (scale bar: 150 pm). Images are representative of n = 3 independent experiments, data are means ± SD. [0049] Figure 8: Embodiment of the effect of printing speed. A) Print with increasing printing speed from left to right form 3 mm.s ¹, 5 mm. s ^1, and 8 mm.s ¹ in CNC colloidal matrix keeping the pressure constant to fabricate filaments with high resolution. B) Graphical illustration of the standard deviation of filament diameter plotted against the printing speed. Transmitted light images of the lattice with 20 % infill are printed where linear lines are with right-angle corners C) 0.5 wt.% Agarose bath D) Printing in 2.5 wt.% CMC. The images illustrate the transparency of the CNC colloidal matrix in comparison with the conventional microparticle support bath for embedded printing, allowing direct optical analysis post-printing. (Scale bar A, C, D: 500 p ).

[0050] Figure 9: A) Illustration of the process for direct 3D printing perfusable channels within CNC fibrillar matrix: i) print channel circuit with fugitive ink, ii) lock structure inducing self-assembly of the CNC with Ca²⁺, (iii) perfusion of the self-standing embedded channel after purging the sacrificial ink. B) Photograph of a printed and locked sinusoidal filament with Pluronic F-127 fugitive ink, C) its liquefication at 4 °C, and D) perfusion of the hollow channel with water (added food color for visualization). CAD models and respective perfusion of resulting microfluidic chip with E) bifurcated design,

F) independent inlets and bifurcations with common convergence point (water with different food colors is perfused to aid the visualization of each channel independently),

G) multilayered channels separated by a depth distance of 500 pm (water is perfused with blue food color in the bottom layer and red in the top layer to aid visualization, scale bars in B, C, D, E, F, G: 1mm). Time lapse fluorescence images of a hollow channel filled with dextran 4 kDa at H) 0 min and at I) 60 min post-injection. J) Plot of spatiotemporal fluorescence intensity for dextran 4 kDa throughout CNC matrix. Time- lapse fluorescence images of a hollow channel filled with dextran 250 kDa at K) 0 min and at L) 60 min post-injection. M) Plot of spatiotemporal fluorescence intensity for dextran 250 kDa throughout CNC matrix. Red and white dotted lines represent the limits of the fabricated channel and of detected fluorescence, respectively (scale bars in H, I, K, L: 200 pm). N) CAD model of the suspended spiral with 15 mm height, and O) respective embedded 3D print with multiple printheads. P) CAD model of miniaturized human right kidney, and Q) respective 3D printed structure with high volumetric infill (65 % concentric infill pattern, 25 layers with 200 pm layer height). R) CAD model of a butterfly, and S) respective 3D printed and locked structure (height 15 mm and width 12 mm, wings with 1.5 mm thickness on the bottom and 0.5 mm on top). T) Released structure immersed in PBS after enzymatic hydrolysis of CNC matrix with cellulase. O, Q, S structures were fabricated using GelTG ink and 27G nozzle. Food colors were added to aid visualization, (scale bars in Q, S, T: 1 mm).

[0051] Figure 10: A) 3D brightfield image reconstruction of CNC matrix embedding lattice structures printed with hASCs in Platelet Lysate (PL) bioinks (rectilinear pattern, 25% infill printed with 27G nozzle), and B) respective confocal laser microscopy (CLM) image sowing hASCs spatial distribution at day 0. C) 3D CLM image of hASCs cytoskeleton in PL bioinks after 4 weeks in culture. (A, B, C scale bar: 200 pm). hASCs viability (Live/Dead assay) in bioprinted constructs with D) GelTG; and E) PL hydrogel bioinks after 14 days of culture (D, E scale bar: 200 pm). F) Estimation of cell proliferation based on cell density by construct areas over a culture period of 14 days (ns p>0.05, * p<0.01, ** p<0.001, *** p<0.001, determined by two-way ANOVA followed by Tukey's post hoc tests). G) CAD design of miniaturized human femur along with a photograph of the respective embedded bioprinted construct using hASCs in PL bioink after 2 weeks of culture. H) CLM projection image of Live/Dead assay of the biofabricated femur embedded in CNC fibrillar matrix and I) its cross-section cut view (H and I scale bar: 1 mm). J) CAD design of 3D disc model with 5 mm diameter and 1 mm height. K) Photograph of the enzymatically (cellulose) released bioprinted disc-shaped construct after 7 days in culture, and L-M) its respective CLM images of cell cytoskeleton (scale bar L: 1 mm, M: 100 pm). Images are representative of n = 3 independent experiments, data are means ± SD.

[0052] Figure 11: A) Photograph of representative heterogeneous tissue model with the dimension of 10x10x0.4 mm and 200 pm layer height (2 layers) printed in a 12-well tissue culture plate (food colors were added into the ink to aid on visualization). B-C) Transmitted light microscopy images of embedded constructs printed with hASCs in PL bioinks, showing the transparency of the CNC fibrillar matrix enabling real-time optical analysis (scale bar B: 250 pm, and C: 150 pm). D) 3D reconstructions of CLM images of the heterogonous model of tissue interface, where hASCs were labeled with green or red cell tracker and bioink hydrogels used are PL or GelTG, respectively (scale bar: 1 mm). E) Bright-field tile scan image of heterogeneous circular construct illustrating cell compartmentalization, along with F) its respective CLM tile scan (scale bar: 1 mm).

[0053] Figure 12: Photograph of an embodiment of the perfusion chip A) 3D printed silicone gasket on glass B) 3D printed using poly-lactic acid (PLA) outer mold, C) assembly of perfusion chip with peristaltic pump, D) incubation of biofabricated sample enclosed in perfusion chip.

[0054] Figure 13: A) Schematic of the assembly of 3D bioprinted chips on printed polylactic acid (PLA) supports used for chip sealing and perfusion (left), and a photograph of 3D printed chip made of silicone gasket filled with CNC where HUVEC suspended in gelatin where bioprinted, scale bar: 500 pm (right). B) 3D reconstruction CLM images of HUVEC cytoskeleton (green) and nuclei (blue) within bioprinted channels where cells organize into a tubular monolayer after 2 weeks under constant perfusion (scale bar: 200 pm). C) Magnified projection image of the channel in B, along with 3D cross-section view of endothelialized channel showing an open lumen (scale bar: 100 pm). D) Higher magnification image of B showing endothelial cell monolayer (scale bar: 25 pm). E) Schematic of the process for printing the tumor-on-a-chip prototype using various bioinks and other materials to construct CNC embedded and compartmentalized structures and allow for culture under dynamic conditions. F) Photograph of the printed chip directly used for CLM analysis without the need to disassemble. G) CLM tile scan of the printed tumor model with cancer (MCF-7) cells in the inner circle and endothelial (EAy.h926) cells in the outer circle, showing actin staining after 7 days of perfusion culture, (scale bar: 1 mm). H) Immunofluorescence images of endothelial cells forming cell-cell contact. I) MCF-7 clusters in the tumor-on-a-chip co-culture conditions in comparison with J) its equivalent monoculture (H, I, J scale bar: 25 pm). K) Cell cluster size in monoculture and tumor-on-a-chip co-culture. p***< 0.0001, determined by unpaired t-test with Welch's correction. Images are representative of n = 3 independent experiments, data are means ± SD.

DETAILED DESCRIPTION [0055] The present disclosure relates to a fluid gel support medium for embedded printing, a kit comprising said fluid gel support medium, a method for obtaining the fluid gel support medium; and a method for obtained an embedded print matrix.

[0056] CNC colloidal stability in aqueous media depends on its synthesis method but is often based on electrostatic repulsions provided by surface charged groups. On the other hand, due to their abundant surface hydroxyl groups, CNC can establish extensive hydrogen bonding and form shear-thinning physical gels at high nanoparticle concentration (usually above ~ 7 wt. %). Shielding of CNC surface charges with mono or bivalent biocompatible ions (e.g. Na⁺ or Ca²⁺) has been applied not only to control the rheological behavior of these colloidal suspensions and produce viscous gels at much lower nanoparticle concentration (lower limit of about 1 wt. %) but also to induce the further self-assembly of these building-blocks into bioinspired fibrillar hydrogels with high structural cohesion and optical transparency.

[0057] In an embodiment, CNC was produced from microcrystalline cellulose by sulfuric acid hydrolysis, a method that leads to negatively charged nanoparticles grafted with sulfate groups on their surface. CNC colloidal suspension was produced via acid hydrolysis of microcrystalline cellulose (MCC) powder (Sigma-Aldrich). Sulphuric acid (95-98% from Sigma-Aldrich) was added to achieve a final concentration of 64 wt.% in the aqueous solution of microcrystalline cellulose. The solution was heated at 60 °C for 40 minutes at 500rpm. Acid hydrolysis was stopped by adding an excess of cold water (5 times the initial volume). Afterwards, the solution was left to decant at 4°C. Then, the supernatant was discarded, and the remaining suspension was centrifuged for 10 minutes at 8603G and 5°C (Sigma 2-16K, Sigma-Aldrich), the supernatant was replaced with ultrapure water and subjected to further centrifugation cycles until the supernatant becomes turbid. The resulting suspension was collected and dialyzed using a cellulose dialysis tubing membrane (MWCO: 12-14kDa, 0-76mm width, Sigma-Aldrich) against deionized water until neutral pH. After dialysis, the content was removed from the membranes, and subjected to 5 sonication cycles of 5 minutes (VCX-130PB-220, Sonics), using an ultrasound probe (Horn ½" SOLID vc 70/13c 3 -0561) at 60 % of amplitude output, under ice-cooling to prevent overheating. Then, the suspension was centrifuged for 10 minutes at 8603G at 5°C (Centrifuge 5810 R, Eppendorf) to remove

IB big particles, after which the solution is degassed with a vacuum pump. Finally, the final supernatant containing CNC is stored at 4°C until further use. For the production of CNC fluid gel support bath, an initial 3 wt.% CNC colloidal solution was sonicated for 1 min at 40 % of amplitude output and then diluted to make nanoparticles suspensions with the desired concentration (1.5 - 2.5 wt. %).

[0058] In an embodiment, the morphology of CNC was analyzed via STEM (Auriga Compact, Zeiss). For sample preparation a drop of 0.0015 wt.% CNC solution was placed on TEM grids (Carbon Type B, 400 M Cu, Monocomp). Images are acquired with an acceleration voltage from 25 to 30 kV. The results show rod-shaped nanocrystals with the dimension of 173.4 ± 16.1 nm in length and 4.9± 1.8nm in width (Figure 2A).

[0059] In an embodiment, the viscoelastic behavior of CNC suspensions as a function of nanoparticle and calcium ions concentration was then characterized. To limit the window of CNC vs ionic strength that leads to fluid gels which could be explored to assist the printing process, a qualitative evaluation by a simple inversion test was performed (Figure 3). The crosslinking agent i.e., calcium chloride solutions (Sigma-Aldrich), was added to CNC at varying concentrations (1.5, 2.0, 2.5, and 7.5 mM) to develop the fluid gel. After the addition of calcium chloride, the colloidal suspension is sonicated again for 1 min at 40 % of amplitude output to ensure homogenous mixing. 500 pi of each prepared colloidal suspension was transferred to 1.5 ml Eppendorf tubes for the inversion test. After 10 minutes at rest, the tubes were inverted upside down to visually inspect the change in viscosity of the fluid gels formed. The results showed that the viscosity of CNC suspensions has a mutual nanoparticle - ionic strength dependence: they start to become viscous enough to resist gravity-driven flow at CNC concentrations of 1.5 wt. % with the addition of 2.5 mM of Ca²⁺ and result in viscous gels that do not flow (within the time frame of the test) at 2.5 wt. % CNC and 2 mM Ca²⁺.

[0060] In an embodiment, preliminary printability tests were performed using a 10 wt. % gelatin ink, showing that the obtained gels can be applied as support matrices to assist in the embedded bioprinting process. Figure 4 shows that the lower limit of parameters that ensured printability is 2 wt. % CNC and 1.5 mM Ca²⁺. While keeping the CNC concentration constant at 2.5 wt. % and varying the ionic strength of the suspension (1.5 - 7.5 mM Ca²⁺), 2 mM Ca²⁺ showed the best performance in retaining the printed structure's shape and fidelity (grid pattern, 20% infill, 27G needle as a nozzle) without leaving noticeable crevices and raised peaks in the CNC support matrix originated by the movement of the nozzle, which in contrast could be well noticed in the 7.5 mM Ca²⁺ formulation. Indeed, it can be observed that at lower ionic concentration (1.5 mM) although the pattern is printable, the printed filaments display width spreading, which indicates the insufficient capacity of the fluid gel to maintain the shape fidelity of printed structure. On the other hand, above 2 mM Ca ²⁺ the higher self-assembly of CNC leads to stiffer gels, which leaves visible needle track mark and does not efficiently self-heal to maintain the structural fidelity of the fabricated structure. Rheology tests were then performed to further assess the impact of ionic strength on the viscoelastic behavior of CNC suspensions and confirm if they exhibit the required properties to be used as a support matrix for embedded bioprinting. The shear rate dependence of viscosity of 2.5 wt. % CNC suspensions at different electrolyte concentrations is shown in Figure 5 and its corresponding strain sweeps are shown in Figure 4B. In agreement with the results of the qualitative assays, increasing Ca²⁺ gradually increases the viscosity of the CNC suspensions as a result of nanoparticle surface charge screening, which compresses the thickness of the electrical double layer and promotes their lateral aggregation¹¹⁰¹. During this process CNC establishes enough network percolation, transitioning from low viscosity liquids to solid-like hydrogels of increasing storage moduli following the increase of suspension's ionic strength. These results are in good agreement with previous studies where the dependence of CNC hydrogel stiffness on ionic strength has been demonstrated, a phenomenon attributed to enhanced side-by-side self-assembly of CNC promoted by the effects of cationic ions addition. ^[11,12] Also, the results illustrate that although the hydrogel with 7.5 M ion concentration is stronger than the one with 2 M ion concentration it breaks at a much lower strain in comparison with the colloidal CNC matrix formed at lower ion concentration.

[0061] In the state of the art, the rheological properties may be measured by many methods. In the present disclosure, to measure the viscoelastic properties of the 2.5 wt. % CNC colloidal suspension (0 and 2.0 mM Ca²⁺) and fibrillar matrix (7.5 mM Ca²⁺), 320 pi of each tested gel was poured on a parallel plate setup equipped with a geometry of 20 mm diameter and final gap of 1 mm between plates, using a Kinexus Pro Rheometer (Malvern Instruments, United Kingdom). The temperature was set at 37 °C and frequency 1 Hz. Shear viscosity was measured in response to shear rate (0.001 to 100 s ¹). Strain-dependent oscillatory shear rheology was determined at a fixed frequency of 1 Hz and strains between 0.1 % to 200 %. The self-healing properties of the CNC gels (2.0 mM Ca²⁺) were measured by alternate strain cycles of 1 % and 200 % (n=3 for all rheological measurements).

[0062] In an embodiment, the microstructure of the CNC fibrillar matrix was observed by high-resolution SEM (JSM-6010LV, JEOL, Japan). Before analysis, CNC hydrogels were critical point dried after solvent exchange from water to ethanol with ethanol gradient from 10 to 99.9 vol.%, and then ethanol was replaced with liquefied CO2 using critical point dryer (Autosamdri-815 Series-A Tousimis). To expose the inner structure of dried samples, they were freeze-fractured after immersion in liquid nitrogen. Before analysis, the samples were sputter-coated with platinum (Cressington Scientific Instruments). The images were collected with an acceleration voltage from 2 to 10 kV. Image analysis was performed using Image J software. Results are provided as mean ± standard deviation (n = 3, independent experiments). Figure 2B shows an embodiment of the hierarchical fibrillar architecture of CNC hydrogels gelled with 7.5 mM Ca²⁺. Remarkably, the hydrogel is composed of entangled fibrils with a mean diameter of 27.1± 5.5 nm and a mean pore size of 85.3 ± 10.5 nm, falling within the corresponding range of dimensions found in native ECMs. In vivo, the ECM fibrils of parenchyma and stromal tissues vary from 10 nm to 230 nm, depending on the tissue type, ^[2] while the dense 3D networks of BM have a pore size ranging from about 10-130 nm.

[0063] In an embodiment, the shear-thinning (Figure 5A) and stress-yielding (Figure 5B) responses of the CNC hydrogels reflects the shear-induced disruption of the non- covalent nanoparticle assemblies formed by the addition of Ca²⁺. After disruption by an external mechanical stimulus, CNC hydrogels are able to rapidly recover from a predominantly viscous to a predominantly elastic state (Figure 5B), demonstrating self- healing potential. This characteristic is essential as support matrix material in order to "lock" the extruded ink behind the nozzle after printing and prevent crevice formation. The tunable rheology of CNC suspensions mediated by Ca²⁺ ions confers to this material excellent properties as a support matrix for printing with different inks, including liquid build constructions based on different post-printing solidification approaches.

[0064] In an embodiment, the embedded 3D printing experiments were performed using a 2.5 wt. % CNC colloidal suspension with 2.0 mM Ca²⁺ as a support matrix, based on their shear-thinning and self-healing rheological properties. This fluid gel was first produced and then poured into desired support mold or plate to be used for the printing process. BioX (Cellink, Sweden) with pneumatic printheads was used for 3D (bio)printing experiments. Computer-aided designs (CAD) were created using the free online software TINKERCAD and saved in stl (stereolithography) file format. The human femur was downloaded from NIH 3D Print Exchange (Model ID 3DPX-000168) while human right kidney was acquired from Bioverse. co, both models were scaled down prior to their conversion into G-code. The stl to G-code conversion program PursaSlic3r 2.1 software was then used to slice the models into layers and translate the coordinates into commands. Cartridges of 3 ml were loaded with (bio)inks and 27G and 30G blunt needles were used as nozzles. The prints were performed applying pressures between 3-5 kPa with the printing speed of 5-8 mm.s ¹ (unless otherwise specified for specific ink or print). After 2-3 minutes post-printing, the support fluid gel was converted into a fibrillar matrix by inducing the hierarchical self-assembly of CNC through the addition of an excess of 7.5 mM Ca²⁺ solution on top of the constructs, hardening it into stable hydrogels and locking the embedded 3D structures in place. After 30 minutes, the Ca²⁺ solution was removed and changed by phosphate buffer saline (PBS, Sigma-Aldrich), where the embedded prints were maintained until further analysis or cell-cultured.

[0065] The present disclosure is more particularly described in the following examples that are intended as illustrative only, since numerous modifications and variations are possible and will be apparent to those skilled in the art.

Example 1:

[0066] The disclosed CNC-based support medium was used to print various low viscosity inks based on hydrogel precursors with different crosslinking mechanisms (Figure 6): 5 wt.% gelatin crosslinked by microbial transglutaminase (mTG) diffusing from CNC fluid gel as representative enzymatic crosslinking (GelTG), 2 wt.% alginate as representative of ionic crosslinking, and 5 wt.% methacrylated gelatin (GelMA) as representative of photo-crosslinking.

[0067] In an embodiment, a solution of 10 wt.% and 5 wt.% gelatin (type-A, porcine skin, Sigma-Aldrich) were prepared in PBS (Sigma-Aldrich) and their pH was adjusted to 7.4 before further use. For the preparation of enzymatically crosslinked gelatin 5 wt.% ink (GelTG), the colloidal CNC suspension was supplemented with 2 wt.% transglutaminase enzyme (Ajinomoto, Japan). To prepare the alginate ink, sodium alginate (Sigma-Aldrich) 2 wt.% was dissolved in PBS and stirred at room temperature for 48 hours. GelMA was synthesized following the pre-established protocol by reacting gelatin (type-A, porcine skin, Sigma-Aldrich) with methacrylic anhydride (Sigma-Aldrich). For the preparation of the photocrosslinkable ink, 5 wt.% GelMA was dissolved in PBS with 0.5% Irgacure 2959 (Sigma-Aldrich). Inbuilt 365 nm light module in BioX printer was used for 3 minutes and at a distance of 3 cm from the light module for GelMA photocuring after printing. To aid the visualization of printed structures in photographs and transmitted light microscopy images during the optimization and characterization steps, food color dyes were added into the inks, while for confocal laser microscopy (CLM) fluorescein isothiocyanate- dextran 250 kDa (Sigma-Aldrich) was added into the ink prior to printing.

[0068] In an embodiment, 2.5 wt.% CNC suspensions with 2 mM Ca²⁺ were used as support matrix to assist the printing process and a 7.5 mM Ca²⁺ solution was used to induce the post-printing self-assembly of CNC, embedding the printed structures within the fibrillary matrix. As shown in Figure 6, the resolution and fidelity of the printed patterns were well maintained, irrespective of the nature of the ink polymer or its crosslinking mechanism. This versatility is particularly important for the field of embedded bioprinting in general as it confers low restriction on the type of ink material that can be explored for each particular application.

Example 2:

[0069] The CNC-based matrix of the present disclosure was compared with an agarose microparticle fluid gel, a widely used support bath, regarding printing fidelity and resolution. [0070] In an embodiment, the agarose bath was produced by autoclaving 0.5 wt.% Agarose (SeaKem, Lonza, USA) with 11 mM CaCU (Sigma-Aldrich) followed by cooling at constant magnetic stirring at 700 rpm to form agarose microparticles. Then, the solution was poured into the desired mold to be used for embedded printing. The desired structures were directly fabricated in the agarose support bath following the same printing protocols described for CNC fluid gel.

[0071] As shown in Figure 7A-B, 5 wt.% gelatin ink printed at the same speed and pressure with a 27G nozzle into these two different support materials result in filaments with diameters of 904.6 ± 34.1 pm in the agarose bath (Figure 7A) and of 181.2± 11.3 pm in our CNC matrix (Figure 7B). Further, while the irregular agarose microparticles create tortuosity on the print bed path that leads to roughened filaments with a variable diameter (Figure 7C), a problem shared by the other microparticle-based support materials, no major visible distortion or particle footprints are seen on the surface of the filaments printed in our colloidal CNC fluid (Figure 7D). Due to the rapid self-healing behavior of the CNC matrix, the filament resolution can be easily tuned by varying the printing speed without negatively affecting the dispersion and distribution of the ink, ranging from 172 ± 4.2 pm at 3 mm.s ¹ down to 37 ± 5.1 pm at 8 mm.s ¹ (Figure 7G and Figure 8A-B). Similarly, unlike microparticulate baths that are typically translucent but not completely transparent compromising its macro and microscopic post-printing optical analysis (Figure 7E, Figure 8C), the disclosed fibrillar CNC matrix is highly transparent (Figure 7F, Figure 8D).

Example 3:

To test the capability of the CNC fibrillary matrix to withhold perfusable channels, an integral feature of microphysiological systems built on microfluidic OoC platforms, the thermoresponsive polymer Pluronic F-127 (25 wt. %, Sigma Aldrich)) was used as fugitive ink to create an open channel (Figure 9). After printing the desired structures (Figure 9B) and locking the CNC with calcium ions (Figure 9C), the perfusability of the created channels was demonstrated by injection of colored dye solutions after removing the sacrificial material (Figure 9D). Perfusable microchannels can be easily fabricated within the disclosed system irrespective of their design complexity, as showed by a series of structures with different arbitrary patterns: a common microfluidics chip design with bifurcated channels (Figure 9E), a design with three different inlets and a convergence point in the middle (Figure 9F), and multi-layered channels (Figure 9G). The capacity for automated fabrication of perfusable microchannel networks with structural integrity using this system shows that it can be used as a microfluidic chip platform, with the advantage of allowing the direct 3D writing of heterogeneous microphysiological systems requiring compartmentalization and perfusion for in vitro tissue/organ and disease modeling. Furthermore, these fabrication steps can be performed under biocompatible conditions, an aspect that has several advantages over competing technologies, beyond the inherent benefits derived from its potential to emulate the physical cues of the ECM fibrillar structure.

Example 4:

[0072] To demonstrate the permeability of the CNC hydrogels for biomolecules of different sizes, it was evaluated the diffusion of FITC-labelled dextrans (4 kDa and 250 kDa, used as model biomolecules) through the fibrillar matrix from a sacrificially printed 200 pm microchannel over the time course of 60 minutes.

[0073] In an embodiment, hollow channels were fabricated using pluronic F-127 (25 wt. %) as fugitive ink. The printed channels were locked by the addition of 7.5 mM calcium chloride solution and the fugitive ink was purged after liquefaction by placing the fabricated structure at 4 °C. Then, aqueous solutions (25 pg.mL ¹) of 4 kDa or 250 kDa dextran-FITC (Sigma-Aldrich, USA) were separately injected into the channel, and samples were immediately observed by fluorescence microscopy (Axio Observer, Zeiss, Germany). The diffusion of the dextran with different molecular sizes over time (up to 60 min) was analyzed by quantification of spatial changes in mean fluorescence intensity on the acquired images plotted against the distance from the center of the channel.

[0074] The results show that the low molecular weight dextran (Figure 9H-J) diffuses faster than the corresponding high molecular weight (Figure 9K-M). Surprisingly, the interstitial porosity of the CNC matrix allows an easy but controlled spatiotemporal diffusion of biomolecules, creating concentration gradients through the hydrogel volume. This permeability is in different aspects remarkable for support and barrier materials of microphysiological systems. First, it ensures that the compartmentalized 3D cellular constructs embedded in the matrix receive the nutrients they need to survive while allowing cellular signaling crosstalk during the in vitro maturation and functional assessment steps. Second, by generating concentration gradients of biomolecules, which play key roles in the regulation of a wide range of biological processes including e.g. development, wound healing or cancer metastasis, it improves the biological significance of the proposed models.

[0075] Indeed, a fundamental requirement for developing functional in vitro tissue models is ensuring the availability of oxygen, nutrients, and signaling molecules to support cell activities. Thus, controlling their diffusion and permeability through cellular constructs is a key factor to be addressed while developing these systems. BM and ECM play critical roles in the regulation of this mass transport phenomena, which are essential for cell function but are poorly recreated by the solid polymeric materials and microporous membranes used as barriers for cellular compartmentalization in microfluidic devices. The fibrillar nature of CNC hydrogels can better mimic the interstitial permeability of the ECM for biomolecule diffusion, contributing to improve the biological relevance of the fabricated microphysiological systems.

Example 5:

[0076] To demonstrate the capability of the disclosed CNC matrix to support the easy fabrication of delicate freeform structures using low viscosity inks, a 15 mm high multicomponent coil using GelTG and multiple printheads was printed. As shown in Figure 9N-0, the CNC matrix allowed high-resolution printing of the coil, without noticeable distortions caused by the repeated movements of the nozzle in 3D space during the printing process. Moreover, the system also enables the fabrication of dense volumetric structures with excellent resolution and precision, as demonstrated by the printing of a miniaturized right human kidney (5 mm height, 25 layers, and 65% infill, Figure 9P-Q).

[0077] Interestingly, nanocelluloses are biocompatible materials that are not biodegradable in mammalian organisms but can be bioorthogonally degraded by cellulase enzymes (the resulting by-product being just glucose) Thus, the CNC matrix can be used as a template bioreactor for the in vitro maturation of complex bioprinted constructs and then allow their mild release by enzymatic digestion of the support material for further in vitro biological analysis or even for its potential in vivo transplantation. In an embodiment, a complex 3D butterfly of 15 mm height, 12 mm width, and variable wing thickness ranging from 1.5 mm at the bottom down to 0.5 mm on top was printed with GelTG ink (CAD design in Figure 9R). After printing, the structure was locked in the CNC matrix by addition of calcium ions (Figure 9S), allowed to crosslinking and then released from the support materials by digestion of CNC with celullase under cell culture conditions (cell culture media at 37 °C). Unlike most of the granular and microparticles based slurries typically used as support baths, there are no major visible residues of the support material on the released structure and the fidelity of the 3D printed model was well preserved (Figure 9T).

Example 6:

[0078] In this example, the performance of the CNC matrix as a platform for building living microphysiological systems was evaluated. Different low viscosity bioinks were tested in combination with human adipose-derived stem cells (hASCs), as a representative source of stromal/stem cells widely applied on the fabrication of tissue- engineered constructs.

[0079] In an embodiment, Human Adipose stem cells (hASCs) were obtained from lipoaspirate samples of the abdominal region after informed consent of patients undergoing plastic surgery, under the scope of protocols established with Hospital da Prelada (Porto, Portugal) with the approval of the Hospital and University of Minho Ethics Committees. hASCs were isolated according to the previously optimized protocol. Alpha Modified Eagle Medium (alpha-MEM) supplemented with 10 vol. % fetal bovine serum (FBS, Thermo Fisher Scientific) and 1 vol. % penicillin/streptomycin solution (Sigma-Aldrich) was used to maintain cells in culture. hASCs were not used beyond passage 8. Human Umbilical Vein Endothelial Cells (HUVEC) were obtained from ThermoFisher Scientific and maintained in culture using EndoGRO-MV media (Sigma- Aldrich). HUVEC were not used beyond passage 6. MCF-7 breast cancer cell line was purchased from Sigma-Aldrich and maintained in culture using high glucose Dulbecco's Modified Eagle Medium (DMEM, Gibco) supplemented with 10 % FBS (Thermo Fisher Scientific) and 1 % antibiotic/antimycotic solution (Sigma-Aldrich). The endothelial cell line EA.hy926 (ATCC^® CRL-2922 ^") was also commercially obtained from (ATCC, LGC Standards, UK) and maintained in culture using low glucose Dulbecco's Modified Eagle Medium (DMEM, Gibco) supplemented with 10 % FBS (Thermo Fisher Scientific) and 1 % antibiotic/antimycotic solution (Sigma-Aldrich).

[0080] In an embodiment, two different hydrogel precursors were tested: gelatin (5 wt.%, GelTG) that is a widely used bioink material, and platelet lysate (PL), a bioactive human-derived liquid formulation that gels in the presence of calcium ions, promotes fast cellular colonization and remodeling of bioprinted constructs, but lacks the physical properties to maintain its structural integrity. The GelTG bioink was prepared by adding cell suspension to a gelatin 10 wt.% in PBS (Sigma-Aldrich) and adjusting the final gelatin concentration to 5 wt. %. For the preparation of PL bioinks, platelet concentrates were obtained from healthy human blood donors, provided by Servigo de Imuno- Hemoterapia of Centro Hospitalar de S. Joao, EPE (Porto, Portugal) under an approved institutional board protocol (ethical commission of CHSJ/FMUP approved at 18/13/2018). PL was produced according to a previously established protocol. Briefly, the samples of platelet concentrate were pooled from 12 healthy human donors and subjected to three freeze-thaw cycles consisting of freezing by immersion in liquid nitrogen followed by heating at 37 °C in the water bath. The produced PL was aliquoted and stored at -80 °C until further use. The frozen aliquots were thawed at 37 °C for 5 minutes, centrifuged at 4000 G for 5 min (Centrifuge 5810 R, Eppendorf) and then filtered with a 0.45 pm pore membrane filter prior to being mixed with the trypsinized cell pellets, resulting in the final PL bioinks used for bioprinting.

[0081] In an embodiment, hASCs suspended in GelTG or PL hydrogels at a cell density of 6xl0⁶ cells. ml ¹ were used as bioinks for the fabrication of representative microphysiological systems with selected 3D configuration. Both multiwell plates (for tissue plate demonstration) or cell culture p-Dish (Ibidi, Germany) were used as a support platform for the prints, depending on the assay. After printing, the fabricated structures were locked with 7.5 mM calcium chloride solution, kept in a cell incubator for 30 minutes, and then excess calcium chloride solution was replaced by culture media. All the constructs printed with GelTG and PL bioinks were cultured up to four weeks under standard culture condition (alpha-MEM with 10% FBS) except the miniatured human femur bone biofabricated with PL bioinks, which was maintained in culture using alpha-MEM without supplementation of FBS following the xenofree culture conditions previously proposed. The media was changed every alternate day. (n = 3 for biologically independent experiments).

[0082] Figure 10A-B illustrates an embodiment of filaments obtained with both bioinks, showing good resolution and homogenous cell distribution without noticeable dispersion into the support material, confirming that CNC fluid gel provides the required flow restriction for the extruded solutions, and its further self-assembly into a fibrillar matrix by Ca²⁺ ions locks the printed structures in space during bioink gelation. Moreover, the CNC matrix supports the in vitro maturation of embedded 3D printed constructs, maintaining high cell viability (Figure 10D-E) and the original filament shape resolution (Figure IOC). Interestingly, although no significant difference was observed in cell density at day 1, PL-based bioink showed a two-fold increase in cell density compared to gelatin at day 14 (Figure 10F), demonstrating its highly inductive cell proliferation properties. The high cell density, viability, and shape fidelity are not only maintained in simple constructs with low infill but also in complex volumetric 3D bioprinted structures (miniaturized structure of human femur bone, 60% infill, Figure lOG-l). This combination of PL-based bioink and support device might therefore be an interesting fabrication option to reduce the time required to generate mature 3D constructs with organotypic cellular density, as preferred for improving functional in vitro modeling.

[0083] In an embodiment, to demonstrate that the maturated tissue can be harvested from the support materials on-demand after the culture steps, disc-shaped constructs (rectilinear pattern, 25% infill, 27G nozzle) were printed with GelTG, cultured within the CNC fibrillar matrix for 7 days, and then retrieved via enzymatic hydrolysis with cellulase. In a further embodiment, Cellulase from Trichoderma reesei (Sigma-Aldrich) at a concentration of 23 g.L ¹ in PBS or culture media was used for enzymatic hydrolysis of CNC matrix and thus harvest the acellular or bioprinted constructs, respectively. The enzyme was sterile filtered prior to use for releasing the bioprinted construct. After 24 hours of incubation at 37 °C, the released structure was washed with PBS and processed for further analysis.

[0084] Figure 10J-M shows an embodiment of the harvesting of a 3D printed object, where a densely cellularized construct kept its 3D printed shape and microfilament resolution after release, demonstrating that the reverse templating process is stable and effective.

Example 7:

[0085] In an embodiment, the disclosed CNC-based matrix allows the rapid generation of large numbers of highly reproducible heterogeneous constructs with arbitrary 3D shapes on standard cell culture multiwell plates (Figure 11A). 12 replicates of a two- component squared structure with dimensions of IO^cIO^c 0.4 mm and 20 % rectilinear infill pattern with a 27G nozzle were printed in less than 5 minutes, allowing immediate real-time monitoring of embedded cell patterns without compromising the integrity of fabricated constructs (Figure 11B-C).

[0086] In another embodiment, the CNC platform enables the manufacturing of multicellular and multicomponent 3D systems with cell compartmentalization, or the simulation of tissue interfaces typically found in most OoC platforms. As shown in Figure 11D-F using hASCs labeled with a fluorescent dye and encapsulated in different bioink hydrogels (PL and GelTG for demonstration), heterogeneous 3D constructs with different geometries and microscale precision can be easily printed without requiring specialized expertise for the fabrication or operation of the device.

Example 8:

[0087] In the present example it was explored the potential of multi-material biofabrication approaches to manufacture microphysiological systems incorporating different types of cells within its own microfluidic perfusion bioreactor. In an embodiment, the CNC platform allows the printing of dense volumetric structures and allow the incorporation of sacrificial reservoirs and fluid circuits for gravity-driven flow systems. In another embodiment, the CNC platform is connected to external pumps to provide a finer fluid flow control.

[0088] In an embodiment, outer silicone gaskets were 3D printed on glass slides with DOWSIL TORAY SE 1700 (DOW, USA) in a polymer and curing agent proportion of 10:1 using BioX (25G plastic conical needle as a nozzle, pressure of 120 kPa, printing speed of 20 mm. sec^-1). After printing, the silicone gaskets were kept at 100 °C for 4 hours for curing. The different bioinks on CNC fluid gel were printed at 5 kPa and 5.0 mm.s ¹ using 27G blunt needles as a nozzle. For fabrication of perfusable vascular channels, HUVECs in gelatin 5 wt.% solutions dissolved in EndoGRO-MV at a density of 8 x 10⁶ cells per mL were used as bioink to print a single channel of 18 mm length and 200 pm diameter. After 2-3 minutes, the structure was locked with 7.5 mM calcium chloride solution, sealed with a glass slide on top of the built construction that was afterward fixed on a custom support frame 3D printed using poly-lactic acid (PLA, Mitsubishi Chemicals Performance Polymers, USA) in a B2X300 3D printer (BEEVERYCREATIVE, Portugal), and placed in a cell incubator for liquefication of gelatin. The perfusion chip was flipped every 10 minutes for 30 minutes to avoid cell sedimentation. After 2 hours of incubation, the channel was perfused with EndoGRO-MV media by connecting a peristaltic pump to metallic connectors (attached to the perfusion chip as inlet and outlet for media) with a flow rate of 40 pL.min ¹. The perfusion chip was kept in dynamic culture for a period of two weeks (Figure 12).

[0089] In an embodiment, vascular channels, a key component of many tissue interface models, were created. Silicone gaskets were first printed on a glass slide demarcating the outer border of the 3D tissue chip that was then filled with CNC fluid gel (Figure 13A). HUVEC cells suspended in gelatin was used as bioink for direct printing the vascular channel and simultaneously achieve its in-situ endothelization. After printing, the channel structure was connected to hollow metal pins interfaced through the gasket walls, the CNC bath was fibrillated with Ca²⁺ and then the system was sealed and perfused for a period of one week to allow cell culture. Surprisingly, cells showed homogeneous distributions and started to organize into a tubular monolayer (Figure 13B) with an open lumen covering the inner surface of the channel (Figure 13C-D). No invasion of cells into the fibrillar CNC matrix was noticed, showing that the perfusable vascular channels are able to maintain their structural integrity during cell culture under dynamic conditions. The disclosed CNC platform allows exceptional cell proliferation with in situ endothelization whilst maintaining the self-standing structural integrity of the channel, in a simple and integrated single process.

Example 9:

[0090] In an embodiment, the CNC platform of the present disclosure was used as a tumor-on-a-chip model. In an embodiment, permeable chip gaskets were first printed on glass slides with a silicone elastomer ink, in which CNCs fluid gel was introduced for printing the cellular model (Figure 13E-G). After printing the external gasket and fill it with CNC fluid gel, a circular shaped model was fabricated by first printing the inner ring with MCF-7 cancer cells at a density of 6 x 10⁶ cells. ml ¹ in PL, and then EA.hy926 endothelial cells suspended in gelatin 5 wt.%/ PL (3:1 v/v) at a density of 8 x 10⁶ cells. ml ¹ was used as bioink to print the in outer cell rings. After printing the cell structures, gelatin 5 wt.% solution was used as fugitive ink to print an open external channel allowing the circulation of culture media in the periphery of the fabricated cellular construct. The CNC was then fibrillated, and the bioprinted chip was sealed and connected to a peristaltic pump for perfusion. After incubating the chip at 37 °C, the liquefied fugitive ink was purged and then the system was kept in dynamic culture over the period of one week at a flow rate of 40 pL.min ¹. (n = 3, biologically independent experiments).

[0091] Figure 131 depicts an embodiment of the immunofluorescence images, showing the development of MCF-7 clusters. Endothelial cells tend to form extensive cell-to-cell contacts (Figure 13H), although well-developed networks were still not evident at this time point. Interestingly, the self-clustering of MCF-7 overtime to form 3D microtissues has been correlated with improved differentiation of these breast cancer cells^[13] and is consistent with the signaling of angiocrine factors that regulate tumor growth in co culture models. The effects of this paracrine crosstalk in co-culture are well evident when comparing it with equivalent MCF-7 monoculture OoC systems (Figure 13J), where the size of formed cell clusters is significantly lower (Figure 13K). The results show that this OoC can be a promising platform for modeling tumor environments.

[0092] In the state of the art, the cell staining with fluorescent dyes may be accomplished by many methods. In the present disclosure, all fluorescent stainings were performed on cell constructs embedded within the CNC matrix except for the samples in perfused systems. In this case, similar staining protocols were followed but instead of adding reagents in static conditions, all the reagents were perfused stepwise via a peristaltic pump. Cell viability on the bioprinted constructs was assessed by staining cells with Calcein AM (Thermo Fisher Scientific) 1:500 dilution in culture media and propidium iodide (Thermo Fisher Scientific) 1:1000 dilution in PBS to stain live and dead cells, respectively, on day 1, 3, 7 and 14 for hASCs and day 14 for HUVECs. To track hASCs within different compartments after printing, cells were labeled with CellTracker™ Deep Red Dye or CellTracker™ Green CMFDA Dye (ThermoFisher Scientific) following the manufacturer's instructions prior to suspending it in the respective bioink hydrogel to aid in visualization after bioprinting. For cell cytoskeleton staining and immunolabeling, the samples were first washed with PBS three times and 4% paraformaldehyde (Thermo Fisher Scientific) was used to fix the cells for 30 minutes at room temperature. Then cell membrane was permeabilized with 0.2% Triton X-100 in PBS for 1 h at room temperature. Afterward, the samples were blocked in 1% (w/v) BSA in PBS for 1 h at room temperature. F-actin fibers were stained with phalloidin conjugated with rhodamine (Sigma-Aldrich, 1:200 dilution) in PBS and nuclei were stained with 4', 6- diamidino-2-phenylindole (DAPI, Sigma-Aldrich, 1:1000 dilution) in PBS. In the tumor- on-chip, endothelial cells were immunolabeled with anti-CD31(PECAM-l)-APC antibody (Thermo Fisher Scientific, 1:200 dilution) in 1 wt. % BSA solution for 1 h at room temperature. Then F-actin fibers were stained with phalloidin conjugated with fluorescein isothiocyanate (Sigma-Aldrich) and nuclei were counterstained with DAPI. Confocal laser microscope TCS SP8 (Leica Microsystems, Germany) was used for fluorescence imagining acquired directly from the constructions embedded in CNC matrix. ImageJ was used to quantify the proliferative density of cells and for the characterization of the size distribution of cell aggregates. Proliferative density was estimated from corresponding cells' area occupied in selected regions of interest (ROI). The area of cell aggregates was manually defined by their outer boundaries. Results are provided as mean ± standard deviation (n = 3, independent experiments). For those skilled in the art, other embodiments will be obvious.

[0093] In an embodiment, the statistical analysis was performed in GraphPad Prism 6 software. Two-way ANOVA followed by Tukey's post hoc tests was used for analysis of variance between multiple groups while unpaired two-tailed Student's t tests with Welch's correction was used for comparisons between two experimental groups. Results are provided as mean ± standard deviation of n ³ 3 independent experiments. Values were considered significant when p < 0.05. [0094] The term "comprising" whenever used in this document is intended to indicate the presence of stated features, integers, steps, components, but not to preclude the presence or addition of one or more other features, integers, steps, components or groups thereof.

[0095] The disclosure should not be seen in any way restricted to the embodiments described and a person with ordinary skill in the art will foresee many possibilities to modifications thereof. The above described embodiments are combinable.

[0096] The following claims further set out particular embodiments of the disclosure.

Claims

C L A I M S

1. A fluid gel support medium for embedded printing wherein the fluid gel support medium comprises:

1 to 4 wt.% of negatively charged cellulose nanocrystals; and at least 0.5 mM of a cation; wherein the fluid gel support medium is self-assemblable into a fibrillar matrix by the addition of a hardening solution comprising 5 to 100 mM of cation.

2. The fluid gel according to the previous claim wherein the cation is selected from a list comprising: Ca²⁺, Zn²⁺, Mg²⁺, Mn²⁺, Sr²⁺, Cu²⁺, Fe³⁺, Al³⁺ or mixtures thereof.

3. The fluid gel according to any of the previous claims wherein the cation is Ca²⁺.

4. The fluid gel according to any of the previous claims wherein the cellulose nanocrystals are grafted with sulfate groups.

5. The fluid gel according to any of the previous claims wherein the fluid gel support medium is optically transparent.

6. The fluid gel according to any of the previous claims wherein the fibrillar matrix is optically transparent.

7. The fluid gel according to any of the previous claims wherein the cellulose nanocrystals are rod-shaped with a size ranging from 70 to 2000 nm in length and 1 to 20 nm in width, preferably from 157 to 190 nm in length and 3 to 6 nm in width.

8. The fluid gel according to any of the previous claims wherein the fibrillar matrix comprises fibrils with a diameter ranging from 10 to 100 nm, preferably from 21 to 32 nm.

9. The fluid gel according to any of the previous claims wherein the pore size of the fibrillar matrix ranges from 40 to 300 nm, preferably from 74 to 96 nm.

10. The fluid gel according to any of the previous claims comprising:

1.5 to 3 wt.% of cellulose nanocrystals, preferably 2.5 wt.% of cellulose nanocrystals; and 1.5 to 2.5 mM of Ca²⁺, preferably 2 mM of Ca²⁺; wherein the cellulose nanocrystals self-assemble into a fibrillar matrix by the addition of a hardening solution comprising 6 to 8 mM of Ca²⁺, preferably 7.5 mM of Ca²⁺.

11. The fluid gel according to any of the previous claims further comprising an enzyme as a crosslinking agent of a printing ink.

12. Use of a fluid gel as described in any of the previous claims as a support medium for embedded printing, preferably embedded bioprinting.

13. A method for embedded printing using a fluid gel support medium as described in any of previous claims 1-11, comprising the following steps: providing the fluid gel support medium comprising 1 to 4 wt.% of cellulose nanocrystals and 0.5 to 3 mM of a cation, preferably Ca²⁺; disposing, using an extrusion tip configured to travel along a predefined course through the fluid gel support medium, a volume of at least one polymeric ink into the fluid gel support medium, wherein the volume of the at least one polymeric ink disposed along the predefined course is retained within the fluid gel support medium; inducing the self-assembling of the fluid gel support medium into a fibrillar matrix by addition of a hardening solution comprising cations, preferably a 5 to 8 mM Ca²⁺ solution.

14. The method according to the previous claim wherein the volume of the at least one polymeric ink is disposed using an extrusion printer coupled with a pneumatic- based printhead, a piston-based printhead, a screw-based printhead, or combinations thereof, preferably a pneumatic-based printhead.

15. The method according to the previous claims 13-14 wherein the volume of the at least one polymeric ink is disposed using a pressure ranging from 1 to 300 kPa.

16. The method according to the previous claims 13-15 wherein the volume of the at least one polymeric ink is disposed using a speed ranging from 1-20 mm.s ¹.

17. The method according to the previous claims 13-16 wherein the at least one polymeric ink is selected from a list comprising: gelatin, alginate, methacrylated gelatin, Pluronic F-127, collagen, fibrin, hyaluronic acid, gellan gum, silk fibroin, platelet lysate, matrigel, polyethylene glycol), polyethylene glycol) diacrylate or mixtures thereof.

18. The method according to the previous claim wherein the at least one polymeric ink further comprises a cell, a therapeutic agent, a growth factor, an enzyme, a metallic element, a cytokine, a contrasting agent, a cell receptor, a cell ligand, or combinations thereof.

19. The method according to the previous claims 13 - 18 wherein the polymeric ink is a sacrificial polymeric ink.

20. The method according to the previous claim further comprising the following steps: allowing the volume of the sacrificial polymeric ink to at least partially solidify; and removing the volume of the at least partially solidified sacrificial polymeric ink from the fibrillar matrix to form one or more voids substantially similar in volume and form to the volume of the at least partially solidified sacrificial polymeric ink.

21. The method according to the previous claims 13-20 further comprising a step of digesting the fibrillar matrix by enzymatic digestion of the cellulose nanocrystals, preferably by enzymatic hydrolysis with cellulase at 37 °C.

22. A kit comprising a fluid gel support medium for embedded printing as described in previous claims 1-11.

23. An organ-on-a-chip, tumor-on-a-chip, or tissue-on-a-chip model obtainable from a method as described in claims 13-21.