Extrusion-Based Bioprinting of Multilayered Nanocellulose Constructs for Cell Cultivation Using In Situ Freezing and Preprint CaCl2 Cross-Linking.

Extrusion-based bioprinting with a preprint cross-linking agent and an in situ cooling stage provides a versatile method for the fabrication of 3D structures for cell culture. We added varying amounts of calcium chloride as a precross-linker into native nanofibrillated cellulose (NFC) hydrogel prior to 3D bioprinting to fabricate structurally stable multilayered constructs without the need for a separate cross-linking bath. To further enhance their stability, we bioprinted the multilayered structures onto an in situ temperature-controlled printing stage at 25, 0, and -10 °C. The extruded and subsequently freeze-dried volumetric constructs maintained their structures after being immersed into a cell culture medium. The ability to maintain the shape after immersion in cell media is an essential feature for the fabrication of stem cell-based artificial organs. We studied the viability and distribution of mouse embryonic fibroblast cells into the hydrogels using luminescence technique and confocal microscopy. Adding CaCl2 increased the stability of the multilayered nanocellulose structures, making them suitable for culturing cells inside the 3D hydrogel environment. Lower stage temperature considerably improved the structural stability of the 3D printed structures, however, had no effect on cell viability.

Introduction

With the growing concerns over ethical issues surrounding animal-based testing, scientists are inclined toward mimicking human organs in laboratories using stem cell-based methods.[1−4] This has led to a boom in organ-on-chip applications and additive manufacturing of structures from natural hydrogels with infused live cells.[5−8] As mammalian cells need a soft and aqueous environment, it is challenging to fabricate cell-compatible volumetric structures (at least 10 × 10 × 10 mm3) via 3D printing.[9,10] Extracellular matrix properties required for cell attachment and shear thinning properties required for the fabrication of multilayered constructs through additive printing are not compatible with each other, making bioprinting more challenging.

For biomedical applications involving in vitro cell culture, 3D volumetric constructs are able to mimic the actual living cell environment better than 2D.[11−16] Nanofibrillated cellulose (NFC) is structurally similar to the extracellular matrix, providing a good support for 3D-culturing of cells.[17−20] Bioprinting involves printing of either hydrogels or cell-laden inks to fabricate computer-aided designs such as organs, tissues, vessels, etc. for various applications such as patient-specific organ printing and tissue engineering.[21] Bioprinting can be characterized into three groups: (1) inkjet, (2) laser, or (3) extrusion-based.[22] In the inkjet bioprinters, hydrogel beads loaded with cell aggregates are used as the prime building blocks of the structures but it lacks the precision for droplet placement. The absorbing layer expands into a bubble when it is vaporized by the laser and bursts into tiny droplets containing cells. These droplets are then forced toward the substrate.[23] However, laser-based 3D bioprinting is a costly and time-consuming process which limits the choice of bioinks to be used.[24]

In extrusion-based printing, hydrogels in the gel state are used as the printing material. Extrusion-based bioprinting allows individual cells or cell aggregates to be suspended in biopolymer gel-like nanocellulose. High viscosity bioinks can be printed using extrusion-based bioprinting with high cell density. Success in bioprinting depends on a number of parameters such as printability, biocompatibility, mechanical and structural properties, degradation of byproducts, and biomimicry ability of the biomaterial used for printing.[12] Extrusion-based printability further depends on five device-controlled parameters, namely, dispensing nozzle diameter, pneumatic pressure, flow rate of the dispensing material, temperature of the hydrogel, and writing speed.[25]

Extrusion-based bioprinting can be used to fabricate bioprinted organs, which will not only decrease the dependency on organ donation but also provide a promising alternative to animal testing. Current research regarding bioprinted constructs use combination of different types of cellulose or/and alginate, which are cross-linked with calcium chloride to form 3D structures.[26−32] In previous studies, the structures were printed inside a calcium chloride bath for postprinting cross-linkage,[27−31] few drops of calcium chloride were added during the printing process,[33] or a combination of both.[26] Rees et al. applied CaCl2 solution on the printed structure during the printing process.[33] Postprinting cross-linking or aerosol spray causes the multilayered structures to lose their structural integrity quickly when exposed to moisture or during cell culturing as the 3D structures are cross-linked at the surfaces only. The low percentage of dry-matter content in hydrogels makes it challenging to prevent the multilayered structure from collapsing.[34] Layer-by-layer stacking is challenging when the overhanging filaments collapse, affecting the overall structural stability. The resolution and the desired pore size of the printed 3D structures are also compromised.[35] Because of poor mechanical stability of hydrogels, the multilayered complex structures fail to maintain their structure postprinting.[36] In this paper, we explore an alternative approach to fabricate structurally stable multilayered structures using ink composed of nanocellulose and calcium chloride in aqueous media. The volumetric structures are bioprinted layer-by-layer directly on top of the cooling stage to enable instantaneous freezing, greatly increasing the structural fidelity of the printed structures. To the best of our knowledge, this is the first case of investigation of the fabrication of multilayered structures using nanocellulose cross-linked with calcium chloride before extrusion-based 3D printing.

Experimental Section

Materials

The concentration range of CaCl2 added in the NFC was taken from the literature[26−33] as the amount of CaCl2 that can be added depends on cell type. Previously, the 3D prints have been either directly printed into CaCl2 solution for postprinting cross-linking or in situ cross-linking with concentration ranging from 50 to 100 mM CaCl2, as shown in Table S1. In this study, three different concentrations of CaCl2 were added in NFC to find the best combination of NFC and CaCl2 for producing the structurally stable scaffold without negatively affecting cell culture of mouse embryonic fibroblast (WT-MEF) cells. Three different concentrations of 19.3, 57.9, and 96.5 mM CaCl2 were added into 1.15 weight percent (wt %) NFC to obtain NFC + 0.22 wt %, NFC + 0.67 wt %, and NFC + 1.08 wt % CaCl2, respectively.

NFC 1.15 wt % was provided by the Department of Bioproducts and Biosystems, Aalto University, Espoo, Finland. NFC was prepared by disintegrating never-dried and fully bleached birch fibers described elsewhere.[37] CaCl2 was purchased from Sigma-Aldrich.

Methods

Rheology Test

Rheology tests were performed using a discovery hybrid rheometer (HR2) with 20 mm parallel plate–plate configuration at 25 °C. Soak time of 300 s was kept constant for each sample and a solvent trap was used for each measurement to avoid drying. Amplitude sweep, frequency sweep, and flow sweep were conducted and described in detail later in the Results section. Five replicas of each sample were tested and then an average value was calculated for each sample to ensure the reliability of the test and to reduce error. An amplitude strain of 0.01–200% was applied which was used to obtain the linear region for performing frequency sweep. Frequency sweep was conducted between 0.1 and 100 Hz with a constant strain of 0.1%. Flow sweep ranging from 0.01 to 300 s–1 was performed to achieve the viscosity versus shear rate curve.

Bioprinter and In Situ Cold Stage

A Nordson EFD EV Series Automated Dispensing System was used to print the multilayered structures. The dispensing system consists of a stage and the dispenser having three printing heads. An Ultimus V High Precision Dispenser was connected to the printer. The stage can move in the y direction and the overhead dispenser attached to the robotic arm can be moved along x and z directions.

The movement is controlled via DispenseMotion computer software. The printer is equipped with a pneumatically driven microextruder (Nordson EFD E4). A liquid-cooled cold plate (Wakefield-Vette) provides heat sink to freeze the printed material directly on the substrate (9 cm × 9 cm glass plates). A liquid thermostat (LAUDA ECO RE 1050 SN) controls the temperature of the cold plate. The in-house bioprinter equipment has multiple nozzles, so the cooling stage can also be used for the fabrication of complex structures using multiple materials.[38,39]

The extrusion process plays a major role in the structural stability of the extruded structures.[40] For precise printing, the distance (z) between the nozzle tip and the printed surface needed to be approximately 0.5 mm. The hydrogel is printed directly on a substrate (glass plate) placed on the cooling plate to enable instantaneous freezing. The substrate is attached to the cold plate magnetically, via plastic fasteners with embedded permanent magnets. A 1 mm-thick iron plate is attached to the cold plate to provide freely adjustable positioning to the magnetic fasteners. Figure shows the printing setup. The addition of a thin iron plate did not have a significant effect on the difference between the temperature set on the cooling stage and the temperature of the glass plate on the cooling station. The difference between the cooling station setpoint temperature (−30 °C) and the temperature of the glass plate on the cooling stage was 11 °C without the iron plate and 12 °C with the iron plate. Hence, the addition of the iron plate enables a secure platform for printing without adversely affecting the low temperatures of the cold plate.

Figure 1

Diagrammatic representation of (a) extrusion-based bioprinter with an in situ cold stage and iron plate held by magnetic fasteners and (b) parameters used for freeze-drying.

All multilayered constructs were printed using a 0.15 mm Nordson stainless-steel precision tip. The pneumatic pressure used for extrusion was between 100 and 200 kPa. Percentage of relative humidity (RH) around the bioprinter was kept between 10 and 15% RH. The printing speed was 10 mm/s and the line width of each printed strand was 1 mm with a constant line spacing of 1 mm. All structures were printed at 25 °C and only the stage temperature was varied. Figure shows the images of the heights and planar dimensions of all multilayered structures.

Figure 3

Images of area (A) and height (H) of 2, 4, and 8 layers (from right to left) of NFC with added amount of CaCl2 printed on varying temperature of the cold stage (25, 0, and −10 °C from left to right).

Freeze Dryer

To obtain stable and dry structures containing multiple layers at room temperature, freeze-drying was used after bioprinting. Printed and frozen structures were immediately transferred to the Martin Christ Epsilon 2-4 LSCplus freeze dryer to maintain structural stability. Freeze-drying, also known as lyophilization, is the process of removing water using sublimation, that is, conversion of water directly into vapor from the frozen state (ice).[41] The vapor is collected on the surface of a condenser at low temperature. The system used here has an inbuilt ice condenser capable of operating at temperatures around −88 °C and the stage temperature can be lowered down to −75 °C as silicon oil is used as the coolant. When energy is added while keeping the vacuum pressure inside the chamber below 6.11 mbar, sublimation (i.e., direct conversion from liquid state to gaseous state) of water occurs. This allows gentle extraction of water from the frozen volumetric structure, resulting in a porous aerogel structure, which retains its shape at room temperature. The multilayered structures in this study were freeze-dried for 20 h and the parameters are summarized in Figure b.

Cell Culture

To test cell viability, five replicas of each sample type were applied in an ultralow attachment 96-well plate (CorningTM CostarTM Ultra-Low Attachment), which was placed on the cold stage under similar environmental conditions that were used for multilayered structures. When the plate reached the desired temperature, the hydrogels were applied by micropipetting. The samples were then freeze-dried for 20 h to form porous aerogels using the freeze-drying process mentioned above (Figure b). Varying amounts of samples (Table S2) were pipetted considering their shrinkage percentage during freeze-drying. As mentioned earlier, the scaffolds prepared by applying the nanocellulose hydrogel directly into the well plates are referred to as 3D structures unlike the multilayered structures, which were extruded using the bioprinter.

MEFs were maintained in high-glucose Dulbeccós modified eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and in a humidified 37 °C, 5% CO2 incubator. For the cell viability assay, cells were cultured in a low-attachment 96-well plates (CorningTM CostarTM Ultra-Low Attachment plates) covered with printed NFCs and let to grow for 48 h. Cells were rinsed twice with PBS and 100 μL of Cell titer Glo (CellTiter-Glo 3D Cell Viability Assay, Promega) reagent was added. As a control, the NFCs in medium without cells were used. The luminescence signal was measured using an EnVisionTM Multilabel Plate Reader. The test was performed in five replicas and in three independent occasions.

For the fluorescent imaging, cells were transiently transfected with LifeAct plasmid (pCMV/pCAG-LifeAct plasmid, Ibidi) using the Neon transfection system (Thermo Fisher Scientific) according to the manufacturer’s instructions. Transfected cells were cultured on glass-bottomed 96-well plates covered with different types of NFCs in triplicate. Cells were let to grow for 24 h, after which the media was changed to FluoroBright DMEM supplemented with 10% FBS. Samples were imaged with a Zeiss Cell Observer Z1 inverted microscope and LSM 780 confocal unit (Zeiss, Oberkochen, Germany) using 40×/1.14, WD 0.62 mm water immersion objective. Representative images were made with the aid of an ImageJ 3D viewer.

Results and Discussion

Titration of Native Nanocellulose

Nanocellulose is biocompatible and nontoxic.[42−45] Nanocellulose consists of fibers with at least one dimension in nanoscale.[46] This makes NFC a suitable candidate for in vitro cell studies as NFC mimics the ECM which promotes the cell growth.[47−49] This together with the ability to form highly porous structures makes nanocellulose a promising candidate for the fabrication of lightweight 3D structures, films, and membrane and its properties can be easily processed in the aqueous state.[46]

Titration of native nanocellulose with 0.050 M NaOH showed that the native nanocellulose used in this study has 3.325 mmol/g of the carboxylic group present (Supporting Information). This made it possible for calcium ions to ionically cross-link with the carboxylate groups in nanocellulose, thus allowing surface manipulation of NFC.

Rheology Test Analysis

Other cellulose derivatives, such as methylcellulose, show a critical variance in the viscoelastic properties when exposed to a temperature between 40 and 50 °C because of shrinkage experience by the polymers because of hydrophobic attractions. NFC shows minor change in viscoelastic properties because of the alignment of the macromolecules to the crystalline part of the nanofibrillated structure. Hence, polymers of NFC are capable of maintaining their structure when subjected to a wide range of pH, temperatures, and ionic strength.[50] At low concentrations, NFC gel has high viscosity, which allows cells to stay suspended in the gel in the 3D shape. Moreover, NFC exhibits a change in viscosity when it is under shear stress (shear thinning). This might be because of the temporary disruption of the fibrillated network of the gel enabling the viscoelastic gel restoration within a few seconds after the removal of the shear stress force.[50] This ability of NFC to vary its viscosity with shear stress allows easy dispensing through a nozzle to print specific shapes and incorporation of cells into the gel. To study the effect of the addition of CaCl2, three different mechanical tests were carried out, namely, frequency sweep, amplitude sweep, and flow sweep using Discovery-HR2. Graphs for storage modulus, loss modulus, and viscosity are shown in Figure a–c.

Figure 2

Graphical representation of rheological properties of the NFC with various amounts of CaCl2. (a) Storage moduli, (b) loss moduli vs angular frequency, and (c) viscosity with respect to the shear rate for native NFC and added CaCl2 amounts. For simplicity, it is referred as 0 wt % in the graphical representation to show that it is native NFC with 0, 0.22, 0.67, and 1.08 wt % added CaCl2. All measurements were performed at 25 °C. Average values based on five parallel measurements are shown.

The results obtained for storage modulus G′ (Figure a) and the loss modulus G″ (Figure b) versus angular frequency followed a similar trend for native NFC and for the hydrogels containing NFC plus the varying amounts of CaCl2. Similarly, viscosity versus shear rate graphs for all NFC with and without CaCl2 produced a similar trend (Figure c). Interestingly, the amount of CaCl2 added in NFC was not directly proportional to the change in moduli. The addition of CaCl2 lowered both the storage and loss modulus compared to native NFC. However, the highest amount of CaCl2 increased both loss and storage modulus which means that the hydrogel became stiffer. All hydrogels experienced noticeable shear thinning; viscosity significantly decreased with an increase in the shear rate from 0.01 to 300 s–1. Hence, all these hydrogels are suitable for extrusion printing.

Fabrication of the Multilayered Structures

The bioprinter was programmed to extrude two, four, and eight layers with areas of 100 mm2 and a height of 2, 4, and 8 mm, respectively. In this paper, we refer to extruded volumetric structures as multilayered structures, and the pipetted structures are called 3D structures. As the structures were printed on the cooling stage, they are referred as frozen structures even for 25 °C to discriminate them from the freeze-dried structures.

We observed decreased area and increased height as the temperature of the cooling stage was decreased from 25 to −10 °C. Therefore, the cooling station was found to play a significant role in the fabrication of structurally stable volumetric constructs. Figure shows the dimensions achieved by each scaffold printed on the varying cold stage temperature for each amount of CaCl2 added in the native NFC.

The temperature of the cooling stage had an immense impact on the structural integrity of the multilayered structures. Lower temperature not only enabled the fabrication of structurally stable prints but also allowed the printing of additional layers without compromising the desired dimensions. Eight layers of NFC without any added CaCl2 spread the most and therefore compromised the height reached. It was not possible to visualize the difference between the 0.22 and 0.67 CaCl2 scaffolds with the naked eye. However, the addition of 1.08 wt % CaCl2 made the multilayered structures fragile because of the lowest nanocellulose fiber content and high CaCl2-to-nanocellulose ratio. Both the addition of CaCl2 and lowering the temperature of the cooling stage helped in the fabrication of stable multilayered volumetric constructs (Figure ). Multilayered structures containing 0.22 wt % CaCl2 content which are printed at −10 °C achieved most swelling height after swelling in cell media for 24 h (Figures , 5 and 6). Moreover, out of all samples containing 0.22 wt % CaCl2, those printed at −10 °C had the highest change in mass after swelling (Figure c).

Figure 4

Effect of the cooling stage temperature and CaCl2 concentration on height. Height of 2, 4, and 8 layers of (a) multilayered extruded samples on the cooling stage at three different temperatures, (b) freeze-dried samples, and (c) after swelling in cell media for 24 h. The height of the volumetric structures is measured using a vernier calliper after printing, freeze-drying, and swelling test. Two-way Anova analysis with Tukey test for p-value <0.05 is performed with three replicas and represented in a–c. Each sample type at 25 °C was taken as the reference value for the values obtained at 0 and −10 °C and the p-values were calculated in relation to the value obtain for 25 °C. 0.0332 (*P), 0.0021 (**P), 0.0002 (***P), and <0.0001 (****P).

Figure 5

Effect of cooling stage temperature and CaCl2 concentration on the area of the printed structures. Area of 2, 4, and 8 layers of (a) multilayered extruded samples on the cooling stage at three different temperatures, (b) freeze-dried samples, and (c) after swelling in cell media for 24 h. The area (mm2) for all three states, that is, frozen, freeze-dried, and swollen scaffolds was calculated after measuring the length using vernier callipers. Two-way Anova analysis with Tukey test was conducted for two replicas. P-values <0.05 are shown on the graphs with respect to the respective sample’s area of the volumetric structure printed at 25 °C. 0.0332 (*P), 0.0021 (**P), 0.0002 (***P), and <0.0001 (***P).

Figure 6

Effect of the cooling stage temperature and the CaCl2 concentration on the swelling degree and water uptake. Graphical representation of 2, 4, and 8 layers after swelling in cell media for 24 h with respect to the freeze-dried multilayered samples for (a) percentage of the swelling degree of height SDH (%) using eq , (b) percentage of the swelling degree of area SDA (%) using eq , and (c) percentage of change in mass ΔMass (%) after swelling in cell media for 24 h using eq . Two-way Anova analysis with Tukey test was conducted for two replicas. P-values <0.05 are considered significant and shown on the graphs with respect to the respective sample’s height of the volumetric structure printed at 25 °C. 0.0332 (*P), 0.0021 (**P), 0.0002 (***P), and <0.0001 (****P).

For two-layered frozen structures, the maximum height was achieved with 0.67 wt % CaCl2 printed at −10 °C. 1.08 wt % four-layered structures printed at −10 °C were the highest and were closely followed by 0.67 wt % (−10 °C), without any significant height difference between them. Both 0.67 wt % (0 °C) and 0.67 wt % (−10 °C) yielded the highest eight-layered frozen structures.

After freeze-drying the samples, their dimensions were measured using a vernier calliper. In general, a decrease in the overall height was observed for all the samples (Figure b). The highest (1.67 mm) two-layered freeze-dried structure was 0.22 wt % (25 °C) closely followed by 0.67 wt % (−10 °C) (1.46 mm). 0.67 wt % (0 and −10 °C) were the highest four-layered freeze-dried structures (2.75 and 2.82 mm, respectively). 0.67 wt % (0 °C) and 0.22 wt % (0 °C) were the highest eight-layered freeze-dried structures (6.23 and 6.17 mm) with 0.67 wt % (−10 °C) shortly falling behind (5.86 mm). Overall, not much variation in swelling height (Figure c) was observed and 0.67 wt % (0 and −10 °C) showed the highest structures. The change in height during freeze-drying was calculated and is described later (Figure a).

Figure shows that 0.22 and 0.67 wt % containing scaffolds exhibit a significant reduction in area when the temperature of the cooling stage was reduced. This shows that the area of 0.22 and 0.67 wt % containing structures is the most concise and best mimic the programmed dimensions for both the frozen (Figure a) and freeze-dried (Figure b) states for all two, four, and eight layers. For frozen structures (Figure a), 0.22 wt % (25 °C) had the smallest area for the two-layered structures. Four-layered structures 0.22 wt % (0 °C) and 0.67 wt % (−10 °C) maintained the smallest area. The eight-layered scaffold (0.67 wt %) printed at −10 °C achieved the smallest area. This shows that the addition of CaCl2 prevented the spreading of the extruded scaffolds.

Like height, area of the scaffolds also reduced after freeze-drying. The smallest area for all two-, four-, and eight-layered freeze-dried structure was achieved by 0.67 wt % extruded at both 0 and −10 °C (Figure b). After leaving the structures to swell for 24 h, the area of all structures stayed close to that of their freeze-dried area, except for only NFC scaffolds containing 0 wt % CaCl2 (Figure c). This shows that lower temperature of the cooling stage enables the fabrication of structurally stable multilayered structures and the addition of CaCl2 in the hydrogels prior to printing enables the structures to maintain their shape during swelling. The change in the swelling area with respect to the freeze-dried area was calculated and analyzed in detail (Figure b).

Swelling Degree and Water Uptake

To analyze the effect of temperature at which the structures are printed on the change in height/area between freeze-dried and the swollen structures (after 24 h), the percentage of change in height, area, and the water uptake with respect to change in mass, eqs , 2, and 3 were used to calculate percentage of swelling degree height (SDH), swelling degree area (SDA), and change in mass (ΔMass).

In eqs and 2, Hhydrated and Ahydrated are the swollen sample’s height and area, respectively, after it is allowed to swell in the cell media for 24 h. Hdry and Adry are the height and area of the freeze-dried structure. SDH (%) is the percentage of the swelling degree according to the height and SDA (%) is the percentage of the swelling degree in relation to the area of the 3D scaffolds.

Water uptake is calculated using eq .where ΔMass (%) is the percentage of water uptake with respect to mass (g) and Mhydratyed and Mdry are the mass (g) after immersion in cell media solution for 24 h and mass of freeze-dried structures, respectively.

Figure a shows that NFC + 0 wt % Ca experienced significant SDH (%) because of the significant decrease in height with colder stage temperature. The addition of Ca did not allow spreading or collapsing of the negative SDH (%) which shows that if the height decreased, the positive value depicts an increase in the height. The only significant trend was observed by two- and four-layered structure containing 0.67 and 1.08 wt %. No significant trend was visible in change of height for eight layers. Like SDH (%), SDA (%), in Figure b, the trend was also not significant. This shows that most samples maintained their shapes after swelling for 24 h.

Overall, no significant trend was visible for SDH (%) (Figure a) except for 0.67 and 1.08 wt % (25 °C). Figure b shows that the SDA (%) of scaffolds containing 0.22 wt % Ca decreased by less than 30%. 0.22 wt % scaffolds could absorb the most (Figure c) without compromising the structural dimensions and shape of the multilayered scaffold. Higher amount of CaCl2 might have made the scaffolds too rigid to absorb high amount of water as the scaffolds containing 1.08 wt % CaCl2 absorbed the least amount of water.

Hence, the data shown in Figures –6 proved that the multilayered structures made of nanocellulose with 0.22 wt % CaCl2 and printed at −10 °C cold stage are the best combination for the swelling which was further proved by the biocompatibility tests.

The added CaCl2 does not cause instantaneous gelation of NFC but rather enhances the printability and the ability of the printed scaffolds to maintain the desired shape. This allows the long-term storage of the NFC plus CaCl2 without having to prepare a new batch each time before printing, hence eradicating any batch variations between prints ensuring repeatability. This was noticeable in the multilayered printed scaffolds with no visible difference in the printability and the structural stability of the volumetric constructs printed using the same batch over a six-month period.

Biocompatibility

To study the biocompatibility of the printed structures, adherent MEF cells were cultured on the nanocellulose scaffolds for 48 h. Cells were cultured in high concentration (100,000 cells/well) in order to observe the effect of the 3D living space provided by the scaffold, rather than the cell proliferation rate in general. As seen from the cell viability assay (Figure a), the cells were viable only when grown with NFC, as the culture plates used were nonadherent and could not support the growth of the cells. However, as cells plated in wells containing the NFCs extruded together with different CaCl2 concentrations were compared, it was obvious that adding the CaCl2 within the hydrogel increased cell viability even further, which can be observed as increased luminescence. Out of the three CaCl2 concentrations used (0.22, 0.67, and 1.08 wt %), the smallest CaCl2 concentration was the most optimal for the cells, when NFC was printed either in −10 °C or +25 °C (Figure a). Whether Ca can dissolve from the NFC during incubation is not known, and to control that, we tested how adding CaCl2 straight into the growth media would affect the cell growth. We added 10 mM CaCl2 (0 + Ca), which equals to 0.11 wt % of CaCl2 within the NFC and observed no change in cell viability (Figure a), suggesting that the Ca was affecting to cell viability via the structure of the substrate, rather than affecting cell metabolism as a soluble ion.

Figure 7

Cross-linking of CaCl2 in nanocellulose improved cell viability. (a) Cell viability measured as the relative amount of ATP within each sample. The NFC without CaCl2 cross-linking is marked as 0 and when CaCl2 is added in different concentrations (wt %), samples are marked as 0.22, 0.67, and 1.08. In the 0 + CaCl2 sample, 10 mM CaCl2 was added to the medium. Controls (ctrl) are samples without cells. No NFC refers to the sample containing cells in culture medium without nanocellulose. Data are normalized to the NFC without CaCl2 (marked as 0). The statistical significance of all results was analyzed by one-way ANOVA and Bonferroni test: *P < 0.05, **P < 0.01, and ***P < 0.001 when compared to the NFC without Ca (marked as 0). (b) 3D rendering of the z-stacks of the fluorescence microscopy images of the MEF cells cultured on different types of nanocellulose for 48 h. High population of MEF cells is observed in CaCl2 cross-linked NFC samples. Glass coverslip (2D) was used as the control. Scale bar (Z) is 50 μm.

Our hypothesis for the addition of CaCl2 was that the porosity/enhanced structure caused by the added calcium would enhance the cells’ ability to reach higher into the 3D matrix of the NFC. This hypothesis was further reinforced with the notion that cell viability was increased in NFCs printed with CaCl2, suggesting that cells had more space to grow into. To visualize whether the cells are able to reach higher into NFC, we transfected the cells with LifeAct-GFP that expresses an actin-binding peptide fused with green fluorescent protein leading to fluorescent actin fibers that can be monitored within live cells. We let the cells settle into the NFC for 24 h and visualized the cell morphology with a confocal microscope and obtaining z-stack images (Figures b and S2). As the images show, NFC without calcium shows cells on the bottom of the container, whereas with increased calcium concentration cells are able to inhabit the NFC higher in the z-direction, suggesting that the porosity provided by CaCl2 addition improved the biocompatibility of the NFC.

Conclusions

The addition of the cooling stage to the bioprinter enhances the structural stability of multilayered printed structures. Although the temperature of the cooling stage did not affect the biocompatibility of the 3D structures, the addition of CaCl2 had a significant impact on the structural stability of the 3D structures and hence improved the cell viability of MEF cells cultured within the scaffold. This study provides a protocol for the fabrication of stable volumetric constructs without the need for a calcium bath using in situ cold stage and adding CaCl2 in the printing ink. This also provides the recipe for improving the biocompatibility of the scaffolds. These results can be used to fabricate precise multilayered structures which maintain their structure during cell culture for tissue regeneration and organ-on-chip applications. In addition to nanocellulose, the proposed methodology can be utilized for other printable hydrogels which do not have instantaneous gelation.