Freeform 3D printing using a continuous viscoelastic supporting matrix

L929 mouse fibroblasts (ATCC ® CRL-6364™) were seeded in 125 cm² t-flasks using Dulbecco's Modified Eagle's Medium Low Glucose (DMEM-LG, ThermoFisher Scientific), supplemented with 10% (v/v) fetal bovine serum (FBS, ThermoFisher Scientific) and (1% (v/v) antibiotic/antimycotic (ThermoFisher Scientific) in controlled atmosphere of 5% CO₂ and 37°C. Upon attaining confluence, cells were trypsinized and resuspended in alginate ink or XG-GMA supporting matrix.

An aqueous solution of 3.5% (w/v) sodium alginate (BioChemica, MM 10 000–600 000 g mol^–1) with 0.1% (w/v) of food dye was prepared to print complex freeform structures. To evaluate XG biocompatibility, L929 mouse fibroblasts (5 × 10⁶ cell ml^–1) were added in 2.0% (w/v) alginate, previously dissolved in complete cell culture medium.

Firstly, 50 mM CaCl₂ (BioChemica) was dissolved in distilled water. Afterwards, 1.5% (w/v) of XG (Doves Farm Foods Ltd, distributed by Bricer Unipessoal, Portugal) was slowly added to avoid the formation of lumps and kept under vigorous stirring overnight, at room temperature. For 3D bio-printing of cell laden filaments, CaCl₂ was dissolved in supplemented culture medium.

Rheological measurements of XG supporting matrix were performed on a Kinexus Lab + rheometer (Malvern Panalytical), by using a 20 mm diameter parallel plate geometry and 0.5 mm gap at 25 ºC. The variation of viscosity was measured with a continuously ramped shear rate (0.1–100 s^–1) to study shear-thinning properties. To determine the linear viscoelastic region (LVR), strain amplitude sweep measurements (0.1%–1000%) were performed at a frequency of 1 Hz. Oscillatory frequency sweep measurements (0.01–100 Hz) were then conducted at constant 3% strain amplitude to determine the storage (G') and loss (G'') moduli.

Complex freeform structures were printed using a bioplotter pneumatic dispensing system (3D-Bioplotter^®Developer Series, Envisiontec). Typically, extrusion was carried out through 25 G needle tips with a pneumatic pressure of 30 kPa and printing speed of 30 mm s^–1. Structures were designed with SolidWorks software and then exported to STL format files. These files were processed by Perfactory RP® Software Suite and sliced into 200 µm thick layers to generate G-code instruction for the bioplotter.

Methacrylated xanthan gum (XG-GMA) was synthesized as previously reported [24]. Briefly, 1.0 g of xanthan gum was dissolved at 0.5% (w/v) in distilled water under vigorous stirring, at room temperature. Afterwards, 8 ml of glycidyl methacrylate (TCI Chemicals) was added dropwise. The reaction was carried out at 60 ºC for 12 h. Afterwards, XG-GMA was precipitated with ethanol and then dissolved in water. The solution was dialyzed (MWCO 6–8 kDa regenerated celulose membrane) against water for 5 d, at RT, in the dark (changing water 3 times per day). The resulting solution was frozen at –80 ºC followed by freeze-drying (−86 ºC, Telstar Lyoquest) and stored at 4 °C, until further use. The obtained XG-GMA was characterized by proton nuclear magnetic resonance (¹ H NMR, D₂O solvent) and Fourier transformed infrared spectroscopy (FTIR-ATR). NMR spectra were recorded at 70 ºC on a Bruker Advance III 300 MHz spectrometer. FTIR spectra were obtained in a Brucker Tensor 27 spectrometer (256 scans, 4 cm^–1 resolution).

XG-GMA supporting matrix was prepared by dissolving 0.5% (w/v) of methacrylate XG in a phosphate-buffered saline solution (PBS, ThermoFischer Scientific) of 0.1% (w/v) 2-hydroxy-4'-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure 2959, Sigma-Aldrich), supplemented with 50 mM CaCl₂. Afterwards, L929 cells were mixed with XG-GMA supporting matrix, at a final density of 20 × 10⁶ cells ml^–1. After printing alginate single filaments, XG-GMA supporting matrix was photocrosslinked upon UV light exposure (OmniCure S2000, 320–500 nm filter, 60 s, 100 mW cm⁻²). Then, to liquefy the alginate filament crosslinked hydrogels were incubated in culture medium supplemented with 20 mM ethylenediaminetetraactetic acid (EDTA, Sigma-Aldrich) for 4 h.

To evaluate cell viability, cell-laden alginate filaments and cross-sections of cell-laden XG-GMA hydrogels were incubated in Calcein-AM/Propidium iodide (PI) (Live/Dead kit, ThermoFisher Scientific) for 20 min, according to the manufacturer's protocol. Widefield (Zeiss, Axio imager 2) and confocal laser scanning (Zeiss, LSM 880 Airy Scan) microscope systems were used to image the stained cell-laden structures. Acquired data was processed in Zeiss ZEN v2.3 blue edition software.

To illustrate the ability of XG to hold non-self-supporting structures, alginate constructs were 3D printed in a gravity-defying spatial arrangement into 1.5% (w/v) XG support matrix, supplemented with CaCl₂ (50 mM) for alginate ionic crosslinking and a green dye for better visualization. Firstly, we created a macroscale 3D asymmetric object in which the bottom of the structure (∼1 cm) was largely narrower than the top (>5 cm) (figures 2(A)(i)–(iii)), where the line of gravity falls outside the structure support base. Notably, during printing process, XG support material fluidizes locally owing to the shear force induced by the nozzle motion (movie S1(available online at stacks.iop.org/BF/12/035017/mmedia), supporting information), allowing the alginate ink deposition. When the printing nozzle moves away from the local extrusion, the shear force is locally removed and the extruded alginate is upheld within XG matrix, which recovers its initial viscosity by reestablishing intermolecular hydrogen bonding networks. Such noncovalent interactions are an intrinsic properties of XG polysaccharide, one of the greatest advantages over host-guest hydrogel supporting materials that requires chemical modifications for further supramolecular assembly.

Zoom In Zoom Out Reset image size

At the end of the 3D printing process, the 3D asymmetric hydrogel structure does not collapse by the gravity effect, highlighting the outstanding suspending properties of XG supporting matrix. Likewise, the continuous XG fluid matrix has also allowed the generation of a volumetric truncated tube with free-floating hollow vessels (figures 2(B)(i)–(iii)). Interestingly, despite the 3D artery-like tree being consecutively built on 2D patterned layers in a layer-by-layer fashion, the XG supporting matrix still enable the fabrication of out-of-plane branches, allowing to print structures beyond the commonly 2.5D. Moreover, this artery-like tree with out-of-plane branches was printed into the XG supporting matrix with significantly improved speed (30 mm.s^–1 for a centimeter scale construct), without compromising the printing resolution of the complex 3D constructs. This result is particularly attractive when compared with those obtained for the already reported supporting materials to be used for extrusion-based approach, ranging from 0.17–0.33 (spiral structures) [17, 18] to 3–10 mm.s^–1 (ventricle) [11] for supramolecular hydrogels and jammed gelatin microparticles, respectively. Therefore, XG matrix offers the unique possibility to create volumetric out-of-plane constructs with high speed and printing fidelity, paving the way for up-scaling construct dimensions to clinically relevant sizes, which is still an unmet challenge in the extrusion-based bio-printing field.

In a first attempt, these complex centimeter-scale constructs were printed with alginate ink, where crosslinking occurs within XG supporting matrix along with the extrusion process due to the compatibility of XG supporting matrix with CaCl₂ salt (movie S1, supporting information). According to the stability of XG solutions over broad salt concentrations, pH and presence of most enzymes [19], we envision that XG supporting matrix may act as an universal reservoir for several crosslinking agents, thus enabling 3D printing multiple materials that virtually undergo different gelation mechanisms (e.g. ionic, pH, enzymatic and through covalent bonds).

Importantly, in embedded 3D bio-printing, the final printed constructs must be removed from the supporting matrix without jeopardizing its structural integrity or even the viability of encapsulated cells. Herein, the truncated tube with out-of-plane free-floating hollow vessels was easily removed from XG support matrix by immersing the container in water. The rapid and efficient dilution of XG solution released the 3D printed construct without damages, preserving the hollow parts (figure 2(B)(iv), movie S2 supporting information). Furthermore, after removal, the alginate layers of the asymmetric truncated tube were well fused together, without delamination. In the presence of cells, 3D structures can be immersed in culture medium to wash the hybrid construct, maintaining cell viability. With this approach complex cell-laden structures can be easily released from XG matrix in mild conditions in contrast to other current embedded bio-printing approaches, which comprise chemical and/or enzymatic degradation [10, 12].

The rheological properties of XG support matrix are critical during the deposition and fixation of the extrudate. The XG supporting matrix exhibited pseudo-plasticity, in which the fluid viscosity decreased exponentially with increasing shear rate (figure 3(A)(i)). Such gel-like behavior was predominantly elastic, at a strain of 3% (figure 3(A)(ii) and figure S1, supporting information). However, with the strain increasing, the loss modulus (G'') surpassed the elastic one (G') at approximately 126% strain, where XG matrix underwent a transition from a gel-like elastic state to a liquid-like state (figure S1, supporting information). At this point, the linear viscoelastic region is vanished and G' becomes strain dependent. We therefore hypothesized this shear-yielding is similar to the shear force applied by the nozzle motion within XG supporting matrix, allowing it to fluidize with little resistance for the extrusion of the alginate ink.

Zoom In Zoom Out Reset image size

To demonstrate the freedom of design provided by XG, a continuous filament was freely drawn in any direction of the 3D space (figure 3(B) and movie S3 in supporting information), remaining stable within XG. supporting matrix, even without a bottom layer sustaining its own weight. Such demonstration highlights the unique ability of XG as a continuous phase to provide physical support for the construction of nonplanar 3D arbitrary shapes. This is particularly attractive for open-source 3D printers that offers the possibility of printing in all orthogonal directions.

To demonstrate the cell-friendly environment of XG, mouse fibroblasts L929 cells (density 5 × 10⁶ cells ml^–1) were encapsulated in a 2.0% (w/v) alginate ink and 3D printed as cell-laden hydrogel filaments into a 1.5% (w/v) XG culture medium sepplemented with 10% FBS. After 1 h, these filaments were removed from XG supporting matrix and incubated in growth medium for 1, 3 and 7 d. Live/Dead analysis showed that L929 cells were viable until a period of 7 d in culture (figure 3(C)), confirming the biocompatible nature of the XG supporting material, as well as their tolerance to the shear forces exerted during the bioprinting process.

To further illustrate the versatility of XG support matrix, we also developed a photocurable XG gel reservoir to design thick perfusable tissue engineering constructs, thereby addressing the recurrent challenge of vascularization and perfusion of nutrients [9, 11, 17, 25, 26]. To achieve this, XG was firstly modified with methacrylate functional groups (XG-GMA) (figure 4(A) and figure S2, supporting information), a common chemical strategy to produce photocrosslinkable hydrogels for biomedical applications [24, 26–30]. Afterwards, an alginate hydrogel filament was printed within 0.5% (w/v) methacrylated XG matrix, which was subsequently crosslinked upon UV light exposure (60 s, 100 mW cm⁻²), enclosing the alginate filament within XG-GMA hydrogel matrix (figure 4(B)). Such alginate filament was then liquified with EDTA, leaving behind an open channel that was easily perfused with a blue dye flowing through the well defined spiral microchannel (figure 4(C) and movie S4, supporting information).

Zoom In Zoom Out Reset image size

To demonstrate the effectiveness of microchannels for regulating mass transport into thick hydrogel constructs, a single alginate filament (∼350 μm, figure S3, supporting information) was printed into a XG-GMA supporting matrix laden with L929 cells at a density 20 × 10⁶ per ml. Upon photocrosslinking and removal of the sacrificial alginate ink, perfused thick cell-laden XG-GMA hydrogels (figure S4, supporting information) were cultured in vitro up to 7 d. Cross-section analysis of perfused hydrogels (figure 4(D)) clearly shown viable cells at different time points, with increased viability and cell clusters at 5 and 7 d, when compared to the bulk controls. Looking ahead, this embedded printing approach may be readily extended to create more complex and hierarchical vascular-mimetic networks for addressing pratical biomedical applications such as in vitro models and organ-on-a-chip platforms.