3D printable carboxylated cellulose nanocrystal-reinforced hydrogel inks for tissue engineering

SA (molecular weight: ∼50 kDa, M/G ratio: 1.67, viscosity: 100–300 cP for 2% solution at 25 °C), XG from Xanthomonas campestris (G1253, viscosity: 800–1200 cps), MCC, APS, and calcium chloride (CaCl₂) were purchased form Sigma-Aldrich Co. Ltd (USA).

Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), penicillin G-streptomycin and trypsin-EDTA were provided by Gibco (Carlsbad, USA). Glutaraldehyde solution (50%), phosphate-buffered saline (PBS), MTT assay and the Live/Dead assay kit were purchased from Sigma-Aldrich (Yongin, Korea). Human skin fibroblast cells (CCD-986Sk) were obtained from the American Type Culture Collection (ATCC; Manassas, USA). All chemicals were used as received and de-ionized water (DIW) was used throughout the experiments. In this study, an Inkredible⁺ 3D printer (Cellink Co. Ltd, Sweden) was used to print hydrogel filament constructs and 3D printed structures.

cCNCs were prepared by one-step oxidation of MCC using APS. In brief, 5 g MCC was added to 1 M APS 500 ml aqueous solution (pre-heated) under vigorous mechanical stirring for 16 h at 75 °C. Then, the suspension of CNCs was centrifuged and the obtained sample was washed with DIW at 10 000 rpm for 10 min until pH ∼4.0–5.0. This pH was further adjusted to 8 by adding 1 M sodium hydroxide (NaOH) and re-dispersed in 300 ml DIW by sonication for 20 min. The final suspension was stored in a refrigerator (at 4 °C) before further use. The yield of CNCs was measured as 45.6% based on gravimetric analysis [41].

Aqueous suspensions of cCNCs were prepared with different amounts as 0.350 g, 0.175 g, and 0.0875 g in 25 ml DIW, respectively. Then, various amounts (0.0375 g, 0.0750 g, and 0.150 g) of XG with SA (fixed amount 0.63 g) were added in these aqueous suspensions of cCNCs at a percentage ratio of 55/6, 28/12, and 14/24 (CNCs/XG) in accordance with the amount of SA, respectively. All hydrogel suspensions were homogenously mixed under constant mechanical stirring at 60 °C for 5 h. We fixed the concentration of SA at 2.5 wt% and only varied the amount of CNCs (from 55 wt% to 14 wt%) and XG (from 6 wt% to 24 wt%) relative to the total amount of SA. For comparative analysis, cell-free hydrogel inks of SA with CNCs (55 wt%) or XG (24 wt%) were also prepared. Finally, all cell-free hydrogel inks were subjected to five freeze–thaw cycles (room temperature to −20 °C to room temperature). The formulations were designated as BPC₅₅X₆, BPC₂₈X₁₂, BPC₁₄X₂₄, BPC₅₅, BPX₂₄, BP, where BP, C, X denote SA, cCNCs, and XG, respectively, and are reported in table 1). In addition, schematic diagrams of the printing of hydrogel ink and the suggested reaction mechanism with the synergistic effect of the components in the hydrogel network are shown in figure 1. All hydrogel inks were stored in sterile conditions in a refrigerator at 4 °C [42].

Zoom In Zoom Out Reset image size

In order to show the potential printing ability of hydrogel inks, we chose a low concentration of BP solution (2.5 wt%) and investigated the blend with XG (gelling matrix) and cCNCs (nanofillers) to allow stable extrusion as compared to the high concentration of BP solution. Here, we speculated that a higher concentration would likely make the hydrogel inks too viscous, making them unable to extrude and even difficult to remove trapped bubbles. Further, this also facilitated non-continuous printing of the hydrogel inks and high pressure would be required for printing, where consequent shear forces may result in a lower viability of encapsulated cells, if used, during printing [36].

The three components of BP, XG, and cCNCs were blended to make multi-polysaccharide hydrogel formulations (see hydrogel inks in table 1) for printing a stable construct at room temperature. The prepared hydrogel inks were loaded into a clear syringe for printing the hydrogel constructs. The test was performed using a plastic cell culture petri dish (printing surface). Here, we analyzed the printability and stability while printing and after printing. We used the Inkredible⁺ bioprinter (Cellink, Sweden) to investigate the printability of hydrogel inks at 25 °C . For the extrusion, 25-gauge (25 G, inner diameter 410 μm) sterile standard conical bioprinting nozzles (Cellink, Sweden) were fixed to 3.0 ml plastic cartridges (2.0 ml hydrogel ink) for the printing. For constant extrusion, the required pneumatic pressure was set using the pressure control knob (PH1) and the distance from the needle to the printing surface was calibrated and optimized for consistent print line flow along with needle. For printing, a 5 × 5 grid (TissueModel_PCL.gcode; 10 × 10 × 1.2 mm, single PH, three layers, feed rate 60 mm min⁻¹) was used. The parameters for printing were set based on the printability of the hydrogel inks. The printing temperature and linear speed of the head were set to 25 °C and 1.0 mm s⁻¹, while the pressure was varied from 45 kPa to 95 kPa as per the continuous extrusion of the hydrogel inks. After printing, the printed constructs were subjected to crosslinking with 100 mM CaCl₂ aqueous solution, which was poured on it and retained for 5 min. Then, we digitally imaged each one at a time to show the stability of the printed construct.

2.5.1. Transmission electron microscopy (TEM)

2.5.2. FESEM

2.5.3. Fourier-transform infrared (FTIR) spectroscopy

The chemical changes in MCC before and after APS treatment and crosslinked hybrid hydrogels were evaluated by using FTIR spectroscopy (Perkin Elmer). The degree of oxidation (DO) of the cCNCs was also measured from FTIR spectra using the following equation [43]:

$$\begin{eqnarray*}&&{\rm{D}}{\rm{e}}{\rm{g}}{\rm{r}}{\rm{e}}{\rm{e}}\,{\rm{o}}{\rm{f}}\,{\rm{o}}{\rm{x}}{\rm{i}}{\rm{d}}{\rm{a}}{\rm{t}}{\rm{i}}{\rm{o}}{\rm{n}}\,\left(DO\right)=0.01+0.7\left({I}_{1608}/{I}_{1025}\right)\end{eqnarray*}$$

where I₁₆₀₈ is the intensity of the carboxylic (–COO⁻Na⁺) group and I₁₀₂₅ is the intensity of the highest absorbance peak from the structural IR spectra.

2.5.4. X-ray diffraction (XRD) analysis

The crystalline nature of the hybrid hydrogels was evaluated using XRD (Bruker AXS D8 Advance) at 40 kV and 30 mA with CuK radiation at a wavelength of 1.54 Å. The samples were scanned at diffraction angles in the range of 2θ = 10–70° with a scanning rate of 10 °C min⁻¹ at room temperature. The crystallinity (%) was calculated by using following equation [44]:

$$\begin{eqnarray*}&&C\,\left({\rm{ \% }}\right)=\left({I}_{002}-{I}_{{\rm{a}}{\rm{m}}}\right)/{I}_{002}\times 100\end{eqnarray*}$$

where I₀₀₂ is the maximum intensity peak at 2θ = 22.5° and I_am is the minimum intensity between the planes (101) and (002).

2.5.5. Rheological analysis

2.5.6. Dynamic mechanical analysis (DMA)

2.5.7. Fidelity and angle on edges

For printed constructs, the area of meshes was calculated with ImageJ 1.46 software (NIH). The fidelity on mesh area (%) was calculated by equation (1) [46]:

$$\begin{eqnarray}&&{\rm{F}}{\rm{i}}{\rm{d}}{\rm{e}}{\rm{l}}{\rm{i}}{\rm{t}}{\rm{y}}\,{\rm{o}}{\rm{n}}\,{\rm{m}}{\rm{e}}{\rm{s}}{\rm{h}}\,{\rm{a}}{\rm{r}}{\rm{e}}{\rm{a}}\,\left({\rm{ \% }}\right)=\left({A}_{{\rm{e}}}/{A}_{{\rm{t}}}\right)\times 100\end{eqnarray} \tag{ 1 }$$

where A_t denotes the theoretical mesh area and A_e denotes the experimental mesh area. All mesh areas were averaged for an individual crosslinked printed construct (before freeze-drying) and compared with the theoretical mesh area (100%).

2.5.8. In vitro cytocompatibility

2.5.9. Statistical analysis

Figure 2(A) shows the schematic process of the one-step preparation of cCNCs from MCC using APS as a mild oxidizing agent. The morphology of the prepared cCNCs was analyzed by TEM (see figure 2(B)). TEM images showed mixed spindle-like and sphere-like shapes with diameter in the range of 15–30 nm and the length in the range of 30–120 nm (as measured by ImageJ 1.46 software). Further, the structural changes in the preparation of cCNCs from MCC by APS chemical treatment were analyzed by FTIR and XRD analysis (as shown in figures 2(C) and (D)). In figure 2(C), MCC shows the characteristic peaks of cellulose at around 3338 cm⁻¹ (O–H stretching), 2898 cm⁻¹ (C–H stretching), 1425 cm⁻¹ (O–C–H and H–C–H deformation), 1372 cm⁻¹ (C–H vibrations), and 898 cm⁻¹ (C–H vibrations) [29, 47]. However, no significant changes were observed before and after APS treatment with MCC. In the case of cCNCs, a peak at around 1608 cm⁻¹ was observed due to the incorporation of carboxylic (–COOH) groups in the ionized form (–COO⁻Na⁺) after the neutralization of cCNCs with NaOH) [31, 48, 49]. The main chemical structure of MCC was not affected by APS treatment and only the peak position of –COOH groups (in cCNCs) was affected significantly by the neutralization with NaOH (surface deprotonating of –COOH groups at around 1728 cm⁻¹ due to C=O stretching) [50]. The DO of cCNCs was calculated at pH > 7 using the conductometric titration method as described elsewhere [50]. The DO value was 0.17 for cCNCs after the APS treatment [29, 51]. For confirmation, the DO of cCNCs was also calculated using the intensity of FTIR peaks at around 1608 cm⁻¹ and 1025 cm⁻¹ by the equation described in section 2.5.3 (FTIR analysis). The DO was calculated as 0.0975, which is similar to the value found using the chemical method.

Zoom In Zoom Out Reset image size

In support of the FTIR analysis, XRD analysis was performed to investigate the crystalline changes after the introduction of –COOH (APS treatment). In figure 2(D), MCC shows characteristic peaks at around 2θ = 15.1°, 16.5°, 22.5°, and 34.5° (040) which are assigned to the characteristic peaks of the native cellulose I structure and after APS treatment, cCNCs also showed similar characteristic peaks. However, the peak at around 34.5° was reduced. The crystallinity (%) of both MCC and cCNCs was observed to be almost similar (∼76.1% for MCC and ∼76.8% for cCNCs) [41].

To perform 3D printing of cell-free or cell-loaded hydrogel inks with post-printing fidelity it is a major challenge to keep other desirable properties of the native 3D microenvironment in tissue regeneration. Therefore, we prepared new compositions of cell-free hydrogel inks and characterized their rheological behaviors before 3D printing. Importantly, we analyzed how the rheological and mechanical behaviors can be tailored by the change in the composition, including their enhanced crosslinking reaction mechanism for printing ability or post-printing fidelity and geometric accuracy. For this development of hydrogel ink or bioink, the designed hydrogels should provide sufficient viscosity for dispensing (free-standing gel filament), mechanical strength, and stiffness to maintain post-printing shape fidelity [45].

To evaluate the printability of BP and BP with XG and/or cCNCs, the rheological properties (elastic and viscous behavior) of hydrogel inks were measured as function of frequency, shear rate, and temperature (figures 3(A)–(D)).

Zoom In Zoom Out Reset image size

3.2.1. Frequency- and temperature-dependent rheological analyses

3.2.2. Evaluation of shear-thinning behavior

To analyze this behavior, shear stress analysis for shear-thinning behavior was measured as a function of shear rate (see figures 4(A) and (B)). At higher shear rate, BP hydrogel inks bear high shear stress in dispensing the hydrogel filament through the extruder (see figure 4(A)), whereas shear stress slightly decreased when XG was incorporated in the BP matrix as in BPX₂₄ hydrogel ink. Similar behavior was also observed in BPC₅₅ and BPC₂₈X₁₂ hydrogels, while this effect was not so significant in the case of BPC₅₅X₆ and BPC₁₄X₂₄ hydrogel inks due to the synergistic behavior via restricted alignment (entanglement through extensive physical crosslinking) of SA/XG chains in BPC₁₄X₂₄ hydrogel and low shear-thinning effect of cCNCs in BPC₅₅X₆ hydrogel [4]. Further, figure 4(B) shows a similar trend to figure 3(B). The viscosity (cP) of the hydrogel inks reduced and the slope of all formulations of hydrogel inks, especially hydrogel inks with cCNCs and/or XG, became steeper as the shear rate (1/s) increased.

Zoom In Zoom Out Reset image size

This shear-thinning behavior can be understood using the suggested schematic mechanism of the behavior of hydrogel inks via the extrusion nozzle tip model, as shown in figure 4(C). Before entering into the conical section, polymer chains and nanofillers are randomly arranged in the hydrogel inks, whereas in the conical section, hydrogel inks bear more shear forces. In this case, the formed hydrogel network is broken up under shear forces and the polymer chains as well as nanofillers aligned to result in reduced viscosity. Hydrogel inks exhibit increased shear-thinning behavior (i.e. reduction in viscosity) from the upper section to the lower section of the nozzle tip. When shear stress was removed (outside the extruder tip), this aligned hydrogel network of polymeric chains and nanofillers again recovered to normal form (i.e. initial viscosity) due to thixotropic behavior and instantly exhibited a stable printed filament line at the printing surface. This is possibly due to the shear-thinning property of XG as well as cCNCs of high aspect ratio that aligned parallel when extruded out and reversed to normal viscosity via anisotropic behavior of cCNCs or percolation effect (figure 4(C)).

The structural changes in chemical functionalities after hydrogel formation are shown in figures S1(A) and (B) (available online at stacks.iop.org/BF/12/025029/mmedia). Further, figure S2(A) shows a digital image of the 3D bioprinter (Cellink Inkredible⁺ model, Sweden) used for printing BP hydrogel ink in a three-layer construct. The initial printing of the hydrogel construct (i.e. BP) can be seen in figure S2(B). To understand the printing process and chemical crosslinking, a schematic of the dispensing of cell-free BPCX hydrogel ink is given in figure 5(A). In addition, a suggestive reaction mechanism is provided before and after crosslinking with CaCl₂. Before crosslinking, there is only extensive hydrogen bonding, while after crosslinking both hydrogen bonding and ionic crosslinking are facilitated, which make it a physically stable hydrogel network with dynamic mechanism under external shear forces. The viscosity of the hydrogel inks was low enough to avoid high printing pressures to allow extrusion, and high enough to print stable filament lines or printed constructs. Figure 5(B) shows the printed ear model to analyze the stability of the construct using BP (5.0 wt%) hydrogel ink. However, BP (2.5 wt%) hydrogel ink was also used for the printing of this ear model, but this printed construct was not as stable (image is not shown here) and collapsed very quickly, diminishing the ear structure before crosslinking. In addition, to investigate the printing stability in 3D constructs, a human nose model (Human nose_scale 50) was also printed using BPC₅₅X₆ hydrogel ink with 128 kPa pneumatic pressure (fast printing) (see figure 5(C)) for the printed nose model, figure 5(D) for oven-dried, and figure 5(E) for the fully swollen nose model in PBS). This swelling of the dried nose model demonstrates that the printed nose model construct did not lose its architecture or anatomy largely. In addition, we have also printed some hydrogel formations by free-hand without using the Cellink Incredible⁺ bioprinter (as shown in figure S3).

Zoom In Zoom Out Reset image size

Figure 6(A) shows the digital and FESEM images of the post-printing stability of the various hydrogel inks (with or without crosslinking). After 30 min, the uncrosslinked printed construct of BP converted to a bulk hydrogel-type construct and no clear filament lines were observed. Further, BPX₂₄ also converted to a bulk hydrogel-type construct, but filament lines could be observed inside the printed construct. However, the incorporation of cCNCs showed a stable printed construct (see BPC₅₅) showing almost clear printed filament lines. Further, the incorporation of cCNCs and XG together as in BPC₅₅X₆ hydrogel ink also synergistically showed good printing stability with almost clear printing filament lines. Additionally, BPC₂₈X₁₂ and BPC₁₄X₂₄ also showed similar behavior. However, in all cases, the filament lines somewhat mingled with each other due to the flow of hydrogel inks. In addition, the crosslinked printed constructs exhibited good post-printing stability and all formulation types showed continuous and clear printing of hydrogel constructs, but the filament line width was observed to be higher compared to the extrusion nozzle tip inner diameter. However, the printed constructs were influenced and dwindled a little (∼10%–15% reduction in their printed size dimensions) by chemical crosslinking as compared to the uncrosslinked printed constructs. Here, an extensive hydrogen-bonds build up in hybrid hydrogels might strengthen before and after printing of hydrogels, and formed stable 3D printing of hydrogel ink. Further, the digital images of freeze-dried printed constructs show the morphology of BP and BP with cCNCs and/or XG at lower magnification (scale bar: 1 mm). However, the crosslinked printed constructs did not show significant changes in the overall printed structure after freeze-drying. The fidelity on mesh area (%) of the printed constructs was also measured, as shown in figure 6(B). BP exhibits a much smaller mesh area (∼33.6%) compared to the theoretical mesh area (100%), where it was hard to keep clear and stable filament lines in the printed construct. Further, the incorporation of XG in SA matrix (i.e. BPX₂₄) increased in the mesh area (∼39.6%) with stable filament lines, whereas the incorporation of cCNCs in SA matrix (i.e. BPC₅₅) showed a somewhat lower mesh area (∼36.6%) than that of BPX₂₄. When both cCNCs and XG were added to the SA matrix (i.e. BPC₅₅X₆), clear printed filament lines with a higher mesh area (∼82.6%) was observed compared to BP, BPX₂₄, and BPC₅₅ printed constructs, but lower than that of the theoretical mesh area of the model construct. However, the mesh area was decreased to ∼61.8% and ∼39.4% for BPC₂₈X₁₂ and BPC₁₄X₂₄, respectively. This is due to the gelling property of XG and high water absorption ability which forms thick filament lines after the extrusion of hydrogel ink. Therefore, the performance of BPC₅₅X₆ hydrogel ink was good and the mesh area was closer to the theoretical mesh area (100%) compared to other hydrogel inks. In addition, BPC₅₅X₆ and BPC₂₈X₁₂ hydrogel inks showed the lowest deviation from the theoretical edge angle (0–90°), while other compositions showed almost round mesh areas (figure S4).

Zoom In Zoom Out Reset image size

These FESEM images showed a homogenous microstructure with a consistent macroporous structure and maintained their integrity of printed filaments, but almost collapsed filament lines were observed. This is possibly due to freeze-drying, where finely printed constructs collapsed and very thin printed filament lines did not hold the pressure balance (i.e. applied vacuum pressure) during the freeze-drying process. However, a porous microstructure of the printed filaments can be observed even after the collapsing of printed filament lines, especially in BPC₅₅, BPC₅₅X₆, BPC₂₈X₁₂, and BPC₁₄X₂₄ printed constructs at high magnification.

The dynamic behavior of the biomaterial under a 3D microenvironment (i.e. in vivo shear forces) is an important factor in the regeneration of tissue via cell attachment, proliferation, migration, and differentiation. In the bulk hydrogels, this dynamic behavior is varied from the outer part to inner part. Therefore, we analyzed the mechanical behavior of fully swollen bulk hydrogels under compression. Figure 7(A) shows the digital images of only BPC₅₅X₆ hydrogel of various sizes. For hydrogel response, the compressive mechanical properties of the crosslinked hydrogels were evaluated by DMA under unconfined compression (see figure 7(B)), where the composition and crosslinking effect might have a synergistic effect on the strength and stiffness of hydrogels. In figure 7(B), BP hydrogel shows a good stability (109.4 ± 3.5 kPa) of the hydrogel network due to extensive physical crosslinking (five freeze–thaw cycles) and chemical crosslinking (ionic interactions). However, the incorporation of XG in BP matrix (BPX₂₄) decreased the compressive strength (42.02 ± 3.1 kPa) and this might be due to the weak gel behavior of XG and the broken intra- and intra-molecular hydrogen bonding of the BP hydrogel network under unconfined compression, making overall a weak hydrogel network compared to BP hydrogel. Further, the incorporation of cCNCs in the BP matrix (BPC₅₅) increased the compressive strength (246.9 ± 5.8 kPa) tremendously compared to BP and BPX₂₄ hydrogels, due to the high elastic modulus and effective load transfer ability of cCNCs [58].

Zoom In Zoom Out Reset image size

In addition, the incorporation of both cCNCs (high) and XG (low) content in the BP matrix (BPC₅₅X₆) increased the compressive strength (223.2 ± 2.6 kPa), but less than that of BPC₅₅ hydrogels, and further, this strength value decreased as the content of cCNCs decreased and the content of XG increased in BPC₂₈X₁₂ (173.4 ± 3.9 kPa) and BPC₁₄X₂₄ (165.5 ± 6.2 kPa) hydrogels, respectively.

Dynamic stiffness values were also measured and showed complex behavior during the compression stress (see figure 7(C)). Therefore, dynamic stiffness values were measured at different strains (10, 20, 30, and 40%), as reported in table 2. Dynamic stiffness values followed a similar trend of compressive strength up to BPC₅₅ hydrogel, but decreased significantly in case of BPC₅₅X₆ and further gradually increased for BPC₂₈X₁₂ and BPC₁₄X₂₄ hydrogels, respectively, at 10% and 20% of strains. In BPCX hydrogels, we speculate that the initially negligible load transfer of cCNCs (that were far apart) might be facilitated in hydrogel, and when more stress was applied, they came together to participate in the load transfer mechanism (behaving as a stiffer hydrogel network). Further, XG chains provide more entanglement of polymeric chains (due to freeze–thaw cycles) and fitness with cCNCs, which facilitate slightly more stiffness to the overall hydrogel network. The hydrogel network behaves differently under applied force during low and high unconfined compression (as can be seen in figures 7(D) and (E)). The synergistic effect of XG and cCNCs controls the rheological properties of the hydrogel inks during printing, and promisingly in an in vivo 3D microenvironment. The actual moduli values of the produced hydrogels are in 100 s of kPa which may be quite stiff for musculoskeletal tissues in vivo. However, the parameters are optimized for the print fidelity and some fine-tuning might be required to achieve physiologically relevant rigidity of the hydrogels. Here, we speculate that the dynamic stiffness values may have a positive impact on the initial cell adherence and growth, and further during tissue formation (in vivo dynamic 3D microenvironment).

3.6.1. Cell morphology

Morphological analysis of cells on hydrogel or scaffold is an important factor in analyzing biochemical functions involved in tissue formation. These biochemical functions in a native 3D microenvironment are dependent on several factors (e.g. density, hydrophilicity, stiffness, elastic modulus, and degradation of the hydrogel) [59, 60]. FESEM images (as shown in figure 8) showed cells spreading with round morphology to hydrogel surfaces after 1 day of culture and good spreading was observed. In general, BP (i.e. SA) hydrogel is biodegradable and biocompatible, but is not able to interact specifically with mammalian cells [61, 62]. However, the cells on BP hydrogel might have little cell adherence on the hydrogel surface due to scratches or irregular surface characteristics, or be physically adsorbed on hydrogel surface followed by the aggregation of cells (by forming clusters of cells), due to more cell–cell interactions than cell–matrix interactions [63]. Further, large numbers of cells were tightly adhered to the hydrogel surfaces and had migrated through pores after 7 days of culture. Additionally, highly populated cell adherence, spreading, and aggregation to make stratified cell layers by covering almost all hydrogel surfaces was observed after 14 days of cell culture.

Zoom In Zoom Out Reset image size

3.6.2. Cell viability and proliferation

Cell viability and proliferation analysis (Live/Dead assay and MTT assay) also supported the cell morphological analysis (FESEM). The Live/Dead assay as a qualitative analysis was performed on 3D bulk hydrogels to visualize the distribution of live (green color) and dead (red color) cells after 1, 7, and 14 days of cell culture (see figure 9). In this analysis, calcein-AM stains live cells a green color by facilitating the reaction of intracellular esterase, whereas ethidum homodimer 1 (EthD-1) stains dead cells in red color by binding to the DNA of dead membrane compromised cells. For this, CCD-986Sk cells (1 × 10⁵ cells) were cultured in bulk hydrogels for different time periods of 1, 7, and 14 days. As can be seen in figure 9(A), Live/Dead assay showed a distribution of viable cells homogenously in all hydrogel types and dead cells were hardly observed.

Zoom In Zoom Out Reset image size

The number of live cells was observed to be much higher than that of dead cells after 14 days of cell culture and very few dead cells or negligible death of cells were observed (see figure 9(B)). It was observed that cellular activity remained stable and showed no cytotoxicity on metabolic activity of the CCD-986Sk cells in hydrogels. Further, as reported in literature, viable cells can have the ability to remodel the hydrogel network and deposit their own ECM [64, 65]. Moreover, hydrogel surfaces were well adhered and covered with live (green) proliferating cells at higher density (see figure 9) and reflected the behavior as described by FESEM.

Here, we speculate that the density in the hydrogel network was increased due to ionic crosslinking and hydrogen bonding, while providing proper living space for CCD-986Sk cells and efficient transport of nutrients to the cells. The incorporation of cCNCs and/or XG in BPCX hydrogels is assumed to provide the required stiffness for cell adherence and growth (stiff substrate-based cell mechanics can be seen in figures S5(A) and shear-thinning effect in dynamic 3D microenvironment for cell viability compared to only BP hydrogel [60]. For comparison, a printed hydrogel construct (e.g. BPC₅₅X₆) was also used for Live/Dead assay after 14 days of cell culture showing stable cell viability with no cytotoxicity (see figures S5(B)–(D)).

Cell viability in hydrogels was further examined by MTT assay (see figure 9(C)). Compared to the 2D culture plate, hydrogels (3D) provide a larger space and higher surface area for cell proliferation and viability, and an extended proliferative activity (i.e. ability to survive) for a longer cell culture time period. After 1 day of incubation, BP showed good cell viability and further increased with the increasing time duration (from 7 days to 14 days). Overall, cells were observed to be viable and metabolically more active in all hydrogels and no significant difference was observed after 1 day of incubation. However, BPC₅₅X₆ hydrogel shows more significant cell viability apart from other hydrogel compositions for 1, 7, and 14 days of incubation. This is possibly due to the increased hydrogel stiffness which has a significant impact on cell fate [30].

In this study, the extrusion behavior and printing stability of hydrogel inks via a single gauge (25 G) was investigated by the incorporation of XG and/or cCNCs. Here, we only evaluated the synergistic effect of XG and cCNCs on rheological behavior, printing ability and shape fidelity, mechanical properties of bulk hydrogels, and their in vitro cell cytocompatibility behaviour (adherence and viability). Furthermore, these biomaterials need more combinations for finding the optimized formulation for post-printing stability and fidelity, even without a crosslinker. For better understanding, the use of different gauge sizes (e.g. each blend with different gauge needles) may give a better insight into the hydrogel inks. The optimization of pneumatic pressure (i.e. shear stress) and increased concentration of the SA matrix by involving a high content of XG and/or cCNCs should be explored. Future studies may involve the effect of concentration, crosslinking agents (e.g. types, time), printing with pore size, compression testing with printed constructs (not bulk hydrogels), and in vitro degradation kinetics (in PBS or DMEM) of the printed constructs. Additionally, cell-loaded hydrogel inks (i.e. bioinks) must be analyzed for realistic behavior of the cells during printing. In this case, a cell survival study via the printing process for short-term and long-term behavior of the cells should be performed. These parameters may help in better understanding the printability and cell viability in bioinks. Therefore, further testing is needed to evaluate the proper printing of natural polymer-based hydrogel inks and bioinks [36, 66], by taking the benefits from our findings.