Development and quantitative characterization of the precursor rheology of hyaluronic acid hydrogels for bioprinting

Abstract

Bioprinting technologies have tremendous potential for advancing regenerative medicine due to the precise spatial control over depositing a printable biomaterial, or bioink. Despite the growing interest in bioprinting, the field is challenged with developing biomaterials for extrusion-based bioprinting. The paradigm of contemporary bioink studies relies on trial-and-error methods for discovering printable biomaterials, which has little practical use for others who endeavor to develop bioinks. There is pressing need to follow the precedent set by a few pioneering studies that have attempted to standardize bioink characterizations for determining the properties that define printability. Here, we developed a pentenoate-functionalized hyaluronic acid hydrogel (PHA) into a printable bioink and used three recommended, quantitative rheological assessments to characterize the printability: 1) yield stress, 2) viscosity, and 3) storage modulus recovery. The most important characteristic is the yield stress; we found a yield stress upper limit of ~1000 Pa for PHA. Measuring the viscosity was advantageous for determining shear-thinning behavior, which aided in extruding highly viscous PHA through a nozzle. Post-printing recovery is required to maintain shape fidelity and we found storage modulus recoveries above ~85% were sufficient for PHA. Two formulations had superior printability (i.e., 1.5 MDa PHA − 4 wt%, and 1 MDa PHA − 8 wt%), and increasing cell concentrations in PHA up to 9 × 10⁶ cells/mL had minimal effects on the printability. Even so, other factors such as sterilization and peptide modifications to enhance bioactivity may influence printability, highlighting the need for investigators to consider such factors when developing new bioinks.

1. Introduction

Bioprinting is a revolutionary technology that has empowered tissue engineers to recapitulate the native structure of tissues by allowing for spatial control over the deposition of printable materials, or bioinks, in addition to controlling cell distribution and the location of bioactive molecules within printed materials [1–5]. Conventional regenerative medicine research has investigated biomaterials containing homogeneous distributions of cells and biomolecules; however, native tissues are complex shapes with hierarchical structures from the macroscale down to the nanoscale, and homogeneous biomaterials that lack structured cues may not be capable of fully regenerating complex tissues. Bioprinting greatly increases the number of possibilities for physical arrangements of different biomaterials, cells, and biomolecules to mimic native tissues and enhance regeneration [6].

While there are a growing number of bioprinters commercially available, the bottleneck for research is the lack of printable bioinks [4,7,8]. Bioinks must satisfy traditional criteria for biomaterials in regenerative medicine such as promoting cell functions (e.g., migration, proliferation, differentiation), but must additionally ensure cell survival throughout the printing process and any post-modifications (e.g., hydrogel crosslinking). Along with the biological requirements, bioinks are challenging to develop because of their physical requirements [8,9]. Hydrogels are a commonly used platform for regenerative medicine because hydrogels facilitate cell survival and fulfill the biological requirements of a bioink; however, many hydrogel precursors are non-viscous liquids that are incapable of being deposited as a 3D structure without a mold, thus rendering them useless for extrusion-based bioprinting [10]. Fortunately, the development of new printable bioinks has emerged as a promising subdiscipline within the bioprinting community, but less fortunately, much of the research depends on trial-and-error of different material formulations to determine printability, with limited characterization of the material’s physical properties or correlation of the characterization to the printability [11–16]. Currently, there are no standardized methods to characterize a material’s printability [8,10], and the development of new bioinks is limited by time-consuming trial-and-error methods. Standardized characterization would stand to benefit from focusing on defining ‘printability windows’ for the properties that pertain to the requirements of the bioprinting process. For successful bioprinting, bioinks must 1) flow through a needle or nozzle as a stable filament, 2) ‘instantly’ retain its 3D shape after deposition on a platform, and 3) support the weight of additional deposited material without collapsing. A few of the rheological properties that relate to the process of bioprinting are the viscosity, yield stress, and storage modulus recovery, but only a handful of reviews and studies have attempted to standardize the characterization [17–23], and few new bioink papers have characterized their bioink with any of the suggested standardized methods, which include viscosity/shear thinning [10,21,24], yield stress [10,18,19], recovery [21,24], and storage/loss modulus [20]. Between the limited number of characterization studies, the viscosity has been well-characterized in a few ways to define quantitative ‘printable windows’ for the loss tangent or predicting printing parameters; however, the printability windows for yield stress and storage modulus recovery have not been quantitatively defined for universal comparison. Reproducible methodology is a key step toward developing protocols that can accelerate the development of new bioinks globally, and facilitate comparisons among groups, thus, there is a need for a continued effort to follow the precedent set by the aforementioned pioneering studies that began using uniform characterizations for bioinks [20,21].

Several types of bioinks are currently being investigated for extrusion-based bioprinting, such as hydrogels, cell aggregates, and decellularized extracellular matrix [3,10,25]. Hydrogels are advantageous because of their high-water content and biocompatibility, and there are several synthetic polymers (e.g., Pluronic^®, poly(ethylene glycol) (PEG)) and natural polymers (e.g., collagen, gelatin, hyaluronic acid (HA), chitosan, alginate, agarose) that have been bioprinted. Furthermore, there are chemically modified versions of both natural and synthetic hydrogels (e.g., methacrylated or diacrylated polymers) and blends of different types of hydrogels (e.g., gelatin/chitosan or PEG-diacrylate/alginate). HA is advantageous because it leverages benefits of both natural and synthetic hydrogels[26]. Similar to other natural polymers, HA is an extracellular matrix polymer found in several tissues (e.g., cartilage, epithelium, nerve) and is non-immunogenic. Similar to synthetic materials, HA is commercially available from microbial fermentation and can be easily chemically modified. The high viscosity of HA and chemical tunability for crosslinking capabilities enabled us to previously develop HA into a paste-like precursor for cartilage and bone regeneration [27–30], which had translational advantages for surgical placement. Our initial focus on surgical placement organically dovetailed into repurposing hydrogel precursor rheological characterization for identifying necessary physical properties for bioprinting [31]. In our previous work, we used a methacrylated HA hydrogel, but our work has evolved to using a pentenoate-functionalized HA (PHA) because of the rapid photocrosslinking of the thiol-ene click chemistry compared to the methacrylate chemistry [32].

The purpose of the current study was to demonstrate the utility of standard protocols that can aid new bioink development, and we demonstrated the characterizations by developing PHA, a known biomaterial [32], into a bioink as an example. The current study examined the yield stress, viscosity, and storage modulus recovery as standard, quantifiable characterizations that can be compared between groups. Furthermore, we compared the parameters of the printable PHA formulations to the non-printable formulations in effort to define a window of printability for each parameter. Given that bioinks are commonly used to bioprint cells, we evaluated the rheological effects of different concentrations of rat bone marrow-derived mesenchymal stem cells (rBMSCs) within PHA, in addition to evaluating the cell viability of two cell types, rBMSCs and rat neural stem cells (rNSCs), after bioprinting. The goals of the experiments presented here were to develop a hyaluronic acid-based hydrogel precursor as a bioink and to characterize three vital rheological parameters to continue defining printability parameters to ultimately aid the development of new bioinks.

2. Materials and methods

2.1. Materials

Unless otherwise stated, all materials were purchased from Sigma-Aldrich (St. Louis, MO).

2.2. Synthesis of pentenoate-functionalized hyaluronic acid

PHA was synthesized from two different molecular weights of hyaluronic acid, 1.5 MDa and 1 MDa (actual MW: 1.55 MDa and 823 kDa; Lifecore Biomedical, Chaska, MN). PHA was synthesized as we described previously [29] except with the modification of the addition of a catalyst. Briefly, HA was dissolved in deionized water (DI) at a concentration of 0.5% (w/v) at room temperature before N,N-Dimethylformamide (DMF) was added at a 3:2 water to DMF ratio. In an additional step beyond our previous work, we added a catalyst, 4-(dimethylamino)pyridine, for the purpose of increasing the reaction efficiency of conjugating the pentenoate groups to the hydroxyls on the HA. Specifically, the catalyst was added to the HA in the water and DMF solution at a ratio of 0.25 g per g of HA and allowed to dissolve. 4-pentenoic anhydride was added in a 5-fold molar excess to HA and the pH was maintained between 8 and 9 with 1 M sodium hydroxide for approximately 1 h. After reacting overnight at room temperature, sodium chloride was added to achieve a final concentration of 0.5 M. Four times the reaction volume of acetone was added to precipitate the PHA and the solution was centrifuged at 7000×g to separate the PHA from the acetone. The centrifuged PHA pellets were collected and dissolved in DI water before being transferred to dialysis tubing (MWCO: 6–8 kDa, VWR, Radnor, PA) to remove impurities. PHA was dialyzed in DI water for 48 h with water exchanges every 12 h before the solution was brought to a physiological pH of 7.4, frozen, and lyophilized. Dry PHA was stored at −20 °C.

2.3. Characterization of PHA

The degree of substitution, or percent of repeating disaccharide units in PHA that were functionalized with a pentenoate group, was quantified using a Varian VNMRS-500 MHz NMR Spectrometer equipped with a 5 mm indirect detection room temperature probe (Varian, Palo Alto, CA). A PHA sample at 10 mM (i.e., concentration of the repeat unit) was prepared in deuterium oxide (D₂O; Cambridge Isotope Laboratories, Inc., Andover, MA) as a solvent. For all batches of PHA, a proton was collected at 80 °C with 16 scans, a recycle delay of 35 s, a 90-degree pulse width, and a 60-s pre-acquisition delay. NMR was performed at 80 °C to compare to the reported ¹H NMR spectrum of PHA, which was collected at 80 °C [32]. Furthermore, the higher temperature of 80 °C shifted the D₂O peak to 4.163 ppm [33], which limited overlap with the majority of the proton peaks of PHA, and after performing NMR on 1.5 MDa PHA at different temperatures (i.e., 25, 50, and 80 °C), proton peaks were more resolved at 80 °C than at 25 or 50 °C and provided more accurate integrations. The percent functionalization was determined by integration of the alkene peaks (2.46 and 2.29 ppm) on the pentenoate group, normalized to the acetyl methyl group (1.91 ppm) of the disaccharide repeat unit. 24–33% of the repeating disaccharide units of HA were functionalized with a pentenoate group for all the 1 MDa and 1.5 MDa batches of PHA synthesized and used in the current study.

2.4. Cell culture

All cell culture supplies were purchased from Thermo Fisher Scientific (Waltham, MA) unless otherwise stated. The rNSCs were purchased (Thermo Fisher Scientific; originally isolated from the cortex of Sprague Dawley rats at day 14 of gestation) and were cultured according to manufacturer protocols. Tissue culture flasks were coated with 1% CTS^™ CELLstart^™ Substrate in Dulbecco’s phosphate-buffered saline with calcium and magnesium (DPBS) for 1 h in a 37 °C, 5% CO₂ incubator. The rNSCs were cultured on coated flasks and in KnockOut DMEM/F-12 Basal Medium supplemented with StemPro Neural Stem Cell Serum Free Medium (NSC SFM) Supplement (2%), recombinant human basic fibroblast growth factor (bFGF; 20 ng/mL), recombinant human epidermal growth factor (EGF; 20 ng/mL), GlutaMAX^™-I supplement (2 mM), and penicillin-streptomycin (Pen/Strep; 1%). The rNSCs medium was changed every 2–3 days and cells were passaged at 80% confluency using StemPro^™ Accutase^™ Cell Dissociation Reagent.

The rBMSCs were purchased (ScienCell, Carlsbad, CA; originally isolated from the bone marrow from adult Charles River Sprague Dawley rats) and cultured in DMEM (low glucose, pyruvate) supplemented with fetal bovine serum (FBS; certified, US Origin; 10%) and Pen/Strep (1%). The rBMSCs medium was changed every 2–3 days and were passaged at 80% confluency using Trypsin-EDTA (0.25%) phenol red.

2.5. Hydrogel precursor preparation

Different PHA precursor formulations of each molecular weight of PHA were prepared for rheological testing by varying the concentration of PHA (i.e., 6–9 weight% (wt%) for 1 MDa PHA and 3–6 wt% for 1.5 MDa PHA). Additionally, precursor formulations with 1.5 MDa PHA (4 wt%) with varying concentrations of cells were prepared for rheological testing and cell viability testing. For all precursors, PHA was dissolved overnight at a 2× concentration in phosphate buffered saline (PBS), before being mixed with 0.05% (w/v) Irgacure 2959 (I2959) for a photoinitiator, dithiothreitol (DTT) at a 1:1 thiol:ene molar ratio for a crosslinker, and the remaining volume of PBS or PBS with cells. PHA precursors were left for 15–30 min in the dark to fully incorporate the added liquids. Due to the paste-like consistency of PHA precursors, letting the precursor mix completely was vital to achieve a homogeneous paste. PHA precursors were mixed once more and loaded into UV-shielding cartridges and centrifuged with a short-spin cycle for 30 s to eliminate air bubbles.

2.6. Rheological testing

The viscosity (n = 5), yield stress (n = 3), and storage modulus recovery (n = 3) of the PHA precursors without cells and the viscosity (n = 5) and yield stress (n = 5 for 7 × 10⁶ cells/mL and n = 3 for all other groups) of PHA precursors with rBMSCs (passage 9–12, 10⁶–9 × 10⁶ cells/mL) were evaluated on a Discovery Hybrid Rheometer-2 (TA Instruments, New Castle, DE) equipped with parallel 20-mm crosshatched plates. Cellink^® Start (Cellink^®, Gothenburg, Sweden) is a commercially-available, bioprintable support material and was used as a control group in rheological studies because of its optimized printability. Additionally, Colgate Total^® toothpaste (Colgate-Palmolive, New York City, NY) was used as a comparator group to put the results into a more understandable context. All tests were conducted at 25 °C and with a gap of 500 μm. Viscosity curves were determined by a logarithmic shear rate sweep from a shear rate of 0.1 s⁻¹–100 s⁻¹ with 3 points per decade. The yield stress was determined by an oscillatory shear stress sweep from 1 to 5000 Pa, and the yield stress was defined as the shear stress at the crossover point of the storage (G′) and loss (G″) moduli, as we have done previously [29,30,34,35]. The storage modulus recovery was determined by three phases of oscillatory shearing at a frequency of 1 Hz after a 5-min soak time where no stresses were applied, similar to what has been done previously in our group [28]. Materials underwent 5 min of a constant shear stress of 10 Pa to determine an initial storage modulus, 30 s of high shear stress of 3000 Pa (i.e., above the material’s yield stress to mimic extrusion and flow through a bioprinter nozzle), and another 5 min of a constant shear stress of 10 Pa. The initial storage modulus was defined by the average storage modulus of the initial 5 min of the material being exposed to 10 Pa of shear stress. The recovered storage modulus was defined by the storage modulus 5 s after the shear stress transitioned from 3000 Pa back to 10 Pa. The percent recovery was defined as the recovered storage modulus divided by the initial storage modulus and multiplied by 100%.

2.7. Printability assessment

The g-code for a grid structure was generated from a 10 × 10 × 1.2 mm (L × W × H) rectangular prism using Repetier-Host software (Hot-World GmbH & Co. KG, Willich, Germany). The object was sliced in Repetier-Host into three 0.4 mm layers, with 43% rectilinear pattern infill to create 6 parallel lines, no perimeter or shells, and a printhead speed of 10 mm/s to generate a distinct path and structure for assessing the printed materials. PHA precursors in bioprinter cartridges were loaded into a pneumatic, extrusion-based Inkredible + bioprinter (Cellink^®) and the 3-layer grid structure was bioprinted through a sterile 27G (i.e., 0.21 mm diameter) polypropylene conical nozzle (Cellink^®) onto a glass microscope slide (24 mm × 76 mm, L × W). The pressure used (40–180 kPa) to extrude the material was adjusted for each material to achieve a consistent flow and the best print possible. Images of printed materials were captured on a DSLR camera (Nikon D5500, B&H Photo Video, New York City, NY) with a macro lens (Nikon AF-S Micro-NIKKOR 60 mm f/2.8G ED Lens, B&H Photo Video).

The printability of each PHA precursor was qualitatively assessed by evaluating the shape fidelity of the printed grid within 1 min after printing prior to any crosslinking and scoring it on a scale of 0–3, based on a review presented by Malda et al. [10] where 0 was no shape fidelity, 1 was poor shape fidelity with low undefined structure, 2 was medium shape fidelity with intermediate irregular patterns, and 3 was high shape fidelity with well-defined building potential. Factors that affected shape fidelity were the edge shape of the individual printed strands (e.g., no edges, soft/undefined edges, clean/stable edges, fragmented/irregular edges), and overall stability of the printed structure shortly after it was printed (e.g., retainment of the grid structure after printing, or relaxation of the material into a puddle after printing). ImageJ was used to measure the strut size of bioprinted grid structures and the particle analysis feature in ImageJ was used to measure the pore areas, similar to Wilson et al. [22].

2.8. Cell viability analysis

To determine the cell viability after bioprinting, 1.5 MDa PHA was sterilized by ethylene oxide gas (AN74i, Anderson Anprolene, Haw River, NC) before a 4 wt% solution was prepared with 10⁶ - cells/mL of rBMSCs (passage 5; n = 3) or rNSCs (passage 3, n = 3) using the methods previously described. Hydrogel precursors of each cell type were bioprinted into rectangular prisms (8 × 6 × 0.3 mm, L × W × H), not the same grid structure as was used for the shape fidelity assessment, onto microscope slides (Fig. 1A). Each slide of hydrogels was crosslinked under a 312 nm handheld UV light at 9 mW/cm² (EB-160C, Spectroline, Westbury, NY) for 2 min (Fig. 1B). While 365 nm light is perceived to be safer than 312 nm light for cells, we used 312 nm light because the photoinitiator used (I2959) has an absorbance peak around 311 nm, which increased the crosslinking efficiency compared to 365 nm light. With 312 nm light, the amount of time needed for crosslinking PHA was only 2 min, which limited the amount of time cells were exposed to UV and did not have detrimental effects in previous studies [27,28,34,36]. An 8-well removable silicon chamber (Ibidi, Martinsried, Germany) was sealed on top of the printed and crosslinked hydrogels and each cell types respective medium was added to each well (Fig. 1C). Cells were cultured for 7 days with medium changes every other day. 2 h after bioprinting, and after 7 days of culture, hydrogels were stained with the LIVE/DEAD Viability/Cytotoxicity Kit for mammalian cells for 20 min using 2 mM calcein AM and 2 mM ethidium homodimer-1 (Thermo Fisher Scientific). Representative images of each group were taken on a Leica TCS SP8 confocal microscope (Leica Microsystems, Wetzlar, Germany) by taking a z-stack of 501 slices throughout the entire hydrogel. An approximate cell viability was determined from the images by comparing live and dead cell counts obtained with the Analyze Particle feature in ImageJ. We noted the cell densities between rBMSCs and rNSCs appeared to be slightly different in the confocal images taken. Based on the printability assessments of PHA with cells showing minor effects on printing from the addition of varying concentrations of cells, we did not anticipate a slight variation in cell density to significantly affect the viability. Therefore, the cell viability data still provided a reasonable representation of the effect of bioprinting on cell viability for both cell types.

Fig. 1.

Graphic of the bioprinting, hydrogel formation, and cell culture. A) Hydrogel precursor containing pentenoate-functionalized hyaluronic acid (PHA), dithiothreitol (DTT), Irgacure 2959, and PBS with cells were mixed and loaded into UV-protected cartridges. PHA precursors were bioprinted in a 2 × 4 grid of rectangular prisms (8 × 6 × 0.3 mm, L × W × H) and onto a 24 × 76 mm glass microscope slide. B) Hydrogels were crosslinked after exposure to 312 nm UV light for 2 min. C) Silicon well chambers were sealed around the hydrogels and medium was added. Cells were cultured for 7 days.

2.9. Statistical methods

GraphPad Prism (Graphpad Software Inc, La Jolla, CA) was used to perform all statistical analyses. All groups were analyzed with a one-way analysis of variance (ANOVA), followed by Tukey’s post-hoc test. Results were considered significant at a level of p < 0.05. All results were reported as mean ± standard deviation.

3. Results

3.1. Printability assessment

All hydrogel precursors were printed into a 3-layer grid (Fig. 2A), given a shape fidelity score (Fig. 2B), and imaged from above (Fig. 2C). Images were analyzed in ImageJ for pore area and strut size (corresponding measurements below each image in Fig. 2C). Cellink^® Start was used as a control group because of its optimized printability and good shape fidelity (shape fidelity score = 3) after printing (Fig. 2A). Cellink^® Start printed with clean edges, was well-matched to the original grid file, and retained its shape after being printed. Toothpaste was included as a comparator group (Fig. 2A) to provide a more understandable context but failed to extrude through a 27G nozzle. Even with a larger 22G nozzle, toothpaste only printed with medium shape fidelity (score = 2). While the printed structure matched the model decently, it had soft edges (i.e., curved turns instead of 90° turns) and relaxed over time into a less defined structure. The pore area and strut size of Cellink^® Start were 0.13 mm² and 1.07 mm, respectively, while toothpaste had a pore area of 0.32 mm² and a strut size of 0.92 mm.

Fig. 2.

Shape fidelity of 1 and 1.5 MDa PHA at varying concentrations. The 1.5 MDa − 4 wt% and 1 MDa − 8 wt% formulations printed the best (outlined in red boxes). A) The printed object was a 3-layer grid with alternating layers of 6 horizontal or 6 vertical lines in a rectilinear pattern. Cellink Start had good shape fidelity after printing, while toothpaste only had medium shape fidelity due to quick material relaxation and loss of defined edges shortly after printing. B) Shape fidelity scores for 1 and 1.5 MDa PHA formulations. For both molecular weights, the shape fidelity increased, reached a peak, and decreased as the concentration of PHA increased. C) Macro images within 1 min after printing and before crosslinking from a top view to show the shape fidelity. Pore areas appeared to generally increase while strut sizes generally decreased with increasing PHA concentrations. Scale bars: 5 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

For the 1 MDa PHA groups (Fig. 2C, top row), the 5 wt% had no shape fidelity and immediately lost its shape after being deposited, resulting in a puddle (score = 0). The 6 wt% performed better than the 5 wt%, where the original structure of the 6 wt% was still visible after printing; however, the edges were soft, and the shape fidelity was still poor (score = 1). The 7 and 8 wt% groups both printed with clean and defined edges similar to the Cellink^® Start but given that the 7 wt% relaxed within a few minutes into an undefined shape, the 7 wt% had medium shape fidelity (score = 2) while the 8 wt% had good shape fidelity (score = 3). While the 9 and 10 wt% groups had structures that imitated the original model, both groups had sharper and fractured edges with printed lines that were irregular, which resulted in a medium shape fidelity (score = 2) for the 9 wt% group and a poor shape fidelity (score = 1) for the 10 wt% group. For pore area there generally appeared to be a trend of increased pore area with increased PHA concentration, ranging from 0.05 mm² (7 wt% PHA) to 0.71 mm² (10 wt% PHA), with the exception of the 6 wt% PHA (0.22 mm²). For strut size, there generally appeared to be a trend of decreased strut size with increased PHA concentration, ranging from 1.48 mm (6 wt% PHA) to 0.45 mm (10 wt% PHA). For context, the inner diameter of the nozzle used to bioprint the formulations was 0.21 mm and therefore, we anticipated the minimum strut size would be 0.21 mm. Out of the 1 MDa PHA groups, the 8 wt% appeared to have the closest pore area and strut size to Cellink^® Start.

For 1.5 MDa PHA groups (Fig. 2C, bottom row), the 3 wt% group was similar to the 1 MDa − 6 wt% group with no shape fidelity (score = 0) because of the immediate loss of structure upon deposition. The 4 wt% group was most similar to the 1 MDa − 8 wt%, with well-defined edges and good shape fidelity (score = 3). The 5 and 6 wt% groups were akin to the 1 MDa − 9 and 10 wt% groups, respectively, with medium (score = 2) and poor (score = 1) shape fidelity from the fractured edges and irregularity of the printed lines. Similar to the 1 MDa PHA, there appeared to be the same general trends of increased pore area and decreased strut size with increased PHA concentration for the 1.5 MDa PHA. The pore areas ranged from 0.14 mm² (4 wt% PHA) to 0.59 mm² (6 wt% PHA) and the strut sizes ranged from 1.04 mm (4 wt% PHA) to 0.51 mm (6 wt% PHA). Out of the 1.5 MDa PHA groups, the 4 wt% appeared to have the closest pore area and strut size to Cellink^® Start.

Shape fidelity scores were plotted against the PHA concentration for the 1 and 1.5 MDa PHA groups (Fig. 2B). For both molecular weights, as the concentration of PHA increased, the shape fidelity increased, reached a peak, and decreased. The groups that printed with medium, poor, or no shape fidelity had two distinct sets of characteristics, depending on if the PHA concentration was greater than or less than the previously mentioned ideal concentrations. The precursors that had a lower PHA concentration either had soft edges or relaxed into a puddle after printing. On the other hand, the precursors with a higher concentration had sharper or fractured edges that printed more irregularly. Overall, the concentration for each molecular weight of PHA that printed with best shape fidelity were 1 MDa − 8 wt% and 1.5 MDa − 4 wt% (Fig. 2B and C, red boxes).

3.2. Rheological testing

3.2.1. Viscosity evaluation

All of the materials displayed decreasing viscosities with increasing shear rates from 0.1 to 100 s⁻¹ (i.e., shear thinning) (Fig. 3A and B), with viscosities ranging from 410 to 11,600 Pa-s at 0.1 s⁻¹ and from 1 to 18 Pa-s at 100 s⁻¹. The Cellink^® Start appeared to have a linear decrease in viscosity from 1510 Pa-s at 0.1 s⁻¹ to 5 Pa-s at 100 s⁻¹ on the log-log plot of viscosity versus shear rate, but the toothpaste viscosity, which ranged from 570 at 0.1 s⁻¹ to 13 Pa-s at 100 s⁻¹, appeared to be more similar to the PHA groups with a gradual decrease at lower shear rates and a sharper decrease at higher shear rates. Overall, increasing the PHA concentration generally appeared to have increased the viscosity behavior but at the highest shear rate, all materials had viscosities under 18 Pa-s. Toothpaste had viscosity behavior similar to the PHA groups, but the toothpaste shear thinning behavior appeared to not decrease as quickly as the PHA groups.

Fig. 3.

Rheological testing of varying PHA concentrations of 1 and 1.5 MDa PHA precursors. Viscosity data by itself did not indicate printability; however, the printable formulations had a yield stress between 0 and 1000 Pa and recovered 85% or greater of their initial storage modulus. A) The viscosities of the 1 MDa PHA increased as the PHA concentration increased; however, all groups’ viscosities decreased to below 18 Pa-s at the highest tested shear rate and demonstrated shear thinning behavior. B) Similar to the 1 MDa PHA groups, the viscosity behavior of the 1.5 MDa PHA was shear thinning and increased as the PHA concentration increased. C) 1.5 MDa − 3 and 4 wt% had similar yield stresses to Cellink^® Start and were less than 1000 Pa. The 1 MDa PHA groups did not have measurable yield stresses. D) Cellink Start^®, 1.5 MDa − 3 wt%, and 1.5 MDa − 4 wt% recovered greater than 85% of their initial storage moduli after being disrupted by a high shear stress. *p < 0.05, ^**p < 0.01, ^***p < 0.001, ^**** p < 0.0001.

3.2.2. Yield stress evaluation

None of the 1 MDa PHA groups had a yield stress, but all the 1.5 MDa PHA groups did have a yield stress (Fig. 3C). For reference, toothpaste had a yield stress of 105 ± 6 Pa. For the 1.5 MDa PHA, there was a trend of increased yield stress with increased PHA concentration. Specifically, the yield stresses of the 3 and 4 wt% groups were 320 ± 90 Pa and 380 ± 17 Pa, respectively, and were not significantly different from Cellink^® Start, which had a yield stress of 217 ± 5 Pa. The 5 wt% group had a yield stress of 1140 ± 150 Pa that was approximately 5.3 times greater than Cellink^® Start and 3.5 times greater than the 3 wt% group (p < 0.005). the yield stress of the 6 wt% group was 2010 ± 360 Pa, which was greater than all other groups (p < 0.005).

3.2.3. Storage modulus recovery evaluation

Because the 1 MDa PHA groups did not have a yield stress and the loss modulus, G′, was higher than the storage modulus, G″, at all shear rates, there would be no loss or recovery of the storage modulus after shearing and the recovery was not evaluated. All references to PHA here refer specifically to the 1.5 MDa groups. Cellink^® Start and PHA 3 wt% and 4 wt% all had recovered storage moduli that were not significantly different from their initial storage moduli (Fig. 3D). Specifically, Cellink^® Start had an 87 ± 1% storage modulus recovery, 3 wt% had an 89 ± 2% recovery, and 4 wt% had an 85 ± 6% recovery. Toothpaste and the 5 wt% and 6 wt% had decreased recovered storage moduli from their initial storage moduli (p < 0.0001), where the percent recoveries of the storage modulus were 32 ± 2%, 76 ± 5%, and 68 ± 5%, respectively.

3.3. Printability and rheological assessments of PHA with encapsulated cells

The 1.5 MDa − 4 wt% PHA precursor with varying cell concentrations (i.e., 10⁶, 3 × 10⁶, 5 × 10⁶, 7 × 10⁶, and 9 × 10⁶ cells/mL) was evaluated for shape fidelity after bioprinting (Fig. 4A), yield stress (Fig. 4B), and viscosity (Fig. 4C). The bioprinted 3-layer grid of all PHA groups with cells, despite the increasing cell concentrations, printed with high shape fidelity (i.e., score = 3) with clean and defined edges that maintained their structure after printing. The pore areas and strut sizes of the PHA with varying cell concentrations were similar to PHA without cells and fell within the range of 0.09–0.22 mm² for pore area and 0.94–1.14 mm for strut size (listed under each respective image in Fig. 4A). The yield stress of the PHA with 10⁶ cells/mL (580 ± 15 Pa) was approximately 1.5 times greater than PHA with no cells (380 ± 17 Pa), 1.4 times greater than the PHA with 5 × 10⁶ cells/mL (410 ± 20 Pa), and 1.75 times greater than the PHA with 9 × 10⁶ cells/mL (330 ± 60 Pa). The yield stresses of the PHA with 3, 5, 7 and 9 × 10⁶ cells/mL were 430 ± 50, 410 ± 20, 310 ± 70, and 330 ± 60 Pa, respectively, and were not different from the yield stress of PHA without cells (380 ± 17 Pa). The viscosities of all the PHA with cells groups at shear rates ranging from 0.1 to 100 s⁻¹ generally appeared to show similar shear thinning behavior to PHA without cells. Overall, the PHA groups with cells displayed similar shear thinning behavior compared to the PHA alone.

Fig. 4.

The printability, yield stress, and viscosity of varying cell concentrations in a 1.5 MDa − 4 wt% PHA precursor. The printability, yield stress, and viscosity were minimally affected by increasing cell concentrations, up to 9 × 10⁶ cells/mL. A) The shape fidelity of the PHA precursor remained high for all cell concentrations tested (1–9 × 10⁶ cells/mL). All cell concentrations had similar pore areas and strut sizes. B) The yield stress of PHA with 1 × 10⁶ cells/mL was increased compared to PHA without cells; however, all higher concentrations had similar yield stresses to PHA without cells. C) The viscosity of the PHA groups with cells all showed similar shear thinning behavior to that of the PHA without cells. Scale bar: 5 mm. *p < 0.05.

3.4. Viability assessment

1.5 MDa − 4 wt% PHA was separately bioprinted into the rectangular scaffolds (8 × 6 × 0.3 mm) with 100% infill, not the grids previously used for shape fidelity assessments, and were crosslinked with encapsulated rBMSCs or rNSCs (Fig. 1). Live/Dead staining showed live cells in green and dead cells in red on the day hydrogels were bioprinted (day 0) and after 7 days of culture (Fig. 5). From the cell counts of the representative image of each group, the rBMSCs had a viability around 100% after the bioprinting and hydrogel crosslinking process. After 7 days of culture, the viability only decreased to 83%. On the other hand, the rNSCs had a viability of about 87% after the bioprinting and crosslinking processes. After 7 days of culture, the viability dropped to approximately 55%. We noted that the size of the rNSCs at day 7 appeared larger than they were at day 0, and we speculate the increased cell size could be due to a small amount of cell spreading or spontaneous differentiation.

Fig. 5.

Viability of rat bone marrow mesenchymal stem cells (rBMSCs) and rat neural stem cells (rNSCs) after bioprinting into rectangular scaffolds (8 × 6 × 0.3 mm) with 100% infill after 0 days and 7 days of culture. The viability was high for both rBMSCs and rNSCs after bioprinting and cells survived for 7 days in culture. B) After rBMSCs or rNSCs were bioprinted in PHA hydrogels (day 0), there was high viability of both cell types (>87%) (green, live cells; red, dead cells). Over 7 days of culture, rBMSCs maintained a high viability (~85%) but the rNSC viability decreased (~55%). Scale bar: 200 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

4. Discussion

We have established a quantitative range of printability, or printability window, of the yield stress and storage modulus recovery, using a PHA bioink as an example. In the experiments performed here, we have not only developed a known crosslinkable hyaluronic acid hydrogel into a printable material, but we quantified three rheological assessments that are universally comparable (i.e., yield stress, storage modulus recovery, and viscosity) and determined two printability windows of yield stress and storage modulus recovery for PHA. While the ranges of printability determined from the experiments presented here are specific for PHA, the application of the methods are standard rheological experiments that can be employed by any group to define the printability of other materials and develop new bioinks. Toothpaste and a commercially available bioink, Cellink^® Start, were tested alongside the PHA here to further demonstrate the applicability of the rheological experiments, but one limitation of the studies here was that other potentially printable biomaterials were not tested. Even so, the data provided here provides a quantitative point of comparison or starting point for the development of new bioinks, especially for other HA-based materials or similar natural polymers.

To provide a link between the qualitative trial-and-error methods and the quantitative rheological measures of viscosity, yield stress, and storage modulus recovery, a shape fidelity assessment was employed to determine the printability of different PHA formulations. From previous work, we found that lower molecular weight (i.e., 60 kDa) PHA precursor solutions had a low viscosity, even at higher concentrations (i.e., 10 wt%); therefore, we investigated higher molecular weights of PHA as potential bioinks (i.e., 1 and 1.5 MDa). From the qualitative shape fidelity assessments, the formulations that printed with good shape fidelity were the 1 MDa − 8 wt% PHA and the 1.5 MDa − 4 wt%. Comparing the two groups that printed with good shape fidelity, the 1.5 MDa − 4 wt% contained half the concentration of PHA compared to the 1 MDa − 8 wt%, which made the higher molecular weight PHA a more practical material source for a bioink because the lower concentration uses less material. In the current study, we used a simple qualitative scale for evaluating shape fidelity. There are other methods of quantifying the shape fidelity, such as using image analysis software to measure pore area, circularity of pores [37], strut size [22], printed area [14], or filament fusion/filament collapse [19]. As Ribeiro et al. [19] emphasized, there is a need for a universal testing method of bioink shape fidelity, or correlation of shape fidelity with another quantitative measure (e.g., yield stress) to compare between different bioinks.

The viscosity of a bioink is an important parameter to characterize because it determines how much pressure is necessary to keep the material extruding at a desired flow rate. Higher viscosity materials or materials with a yield stress maintain their shape after printing better than low viscosity materials or materials without a yield stress; however, a material experiences high shear forces when flowing through a needle, and viscous bioinks necessitate shear thinning behavior, or having lower viscosities at high shear rates, to be able to flow through a needle [22]. The relevant shear rates for needles have been estimated to be 100–500 s⁻¹ all along the length of the needle, whereas tapered nozzles had higher estimated shear rates around 200–3000 s⁻¹ localized to just the nozzle tip [21,38]. In the current study, PHA failed to extrude through a 22G blunt-tip needle (i.e., 0.413 mm diameter), but was extrudable through a 27G nozzle tip (i.e., 0.21 mm diameter), making nozzles a more practical choice for extrusion of materials with high viscosity.

As others have found [10,18–21,24], the viscosity measurement alone is insufficient for determining printability. In these studies, we found that the PHA formulations that printed the best did not have a distinct viscosity behavior that differentiated them from the groups that did not print well. Comparing the viscosity curve of the Cellink^® Start to the printable 1.5 MDa − 4 wt% group, both groups had shear thinning behavior and converged to viscosities below 18 Pa-s at a shear rate of 100 s⁻¹, which is in the neighborhood of the relevant shear rates through a nozzle; however, the Cellink^® Start followed a different, more linear decrease with increasing shear rates when compared to the PHA groups. Therefore, there was not a “printable” viscosity range and the viscosity alone could not be used to determine printability. The most important information the viscosity data provided was determining whether the material was shear thinning, which was vital for the extrusion of viscous materials through a nozzle. Others, such as Paxton et al. [21] found similar results and demonstrated additional uses for viscosity data in predicting other printing parameters. Specifically, they used the viscosity data to determine the shear thinning coefficients from a Power Law regression and predicted the printing parameters ranges for the printing pressure and nozzle geometry. They found that the higher viscosity materials would require higher pressures and larger diameter nozzles for successful extrusion.

The yield stress is the most important parameter to characterize because a yield stress is required for a material to hold its shape after bioprinting and support the weight of added layers. Comparing the yield stresses of the PHA formulations that had good shape fidelity to the formulations that did not, printable materials had yield stresses below a maximum threshold value of around 1000 Pa, where PHA with a yield stress above the threshold printed with fractured and irregular lines. The upper limit of the yield stress determined from our results specifically applies to extrusion of PHA through a 27G (i.e., 0.21 mm) nozzle at the pressures used here, and the yield stress limits for different size nozzles or materials may deviate from that value (see Table 1). For example, Paxton et al. [21] found that the printable formulation of poloxamer 407 through a 0.25 mm nozzle had a yield stress of 227 Pa, but the formulation with a 348 Pa yield stress was not printable, indicating a yield stress limit around 300 Pa. While there may be different yield stress limits for different materials to print with good shape fidelity, a material with high viscosity may be sufficient for printing with good shape fidelity under the right conditions. Among these conditions for good shape fidelity for high viscosity materials, the most important is time, and specifically the time elapsed from printing from the nozzle to crosslinking for a given layer. Given enough time, even the highest viscosity bioink in theory can flow (i.e., change shape) if it does not possess a sufficient yield stress.

Table 1.

Yield stresses of different printable bioinks.

Material	Nozzle diameter (mm)	Yield stress (Pa)	Notes	Author
Pluronic F127/diacrylated Pluronic F127 (PF17/AC3)	0.30	151	No other printable or non-printable formulations.	Muller et al. [17]
25 wt% poloxamer 407	0.25	227	15, 20, and 30 wt% formulations were not printable.	Paxton et al. [21]
8% alginate/1% CaCl2	0.25	166	No other alginate/CaCl2 formulations were tested.	Paxton et al. [21]
3–10% methacrylated gelatin/ 0–0.5% gellan gum (gelMA/gellan)	0.34	0.13–48.2	3/0.5%, 10/0%, and 10/0.5% gelMA/gellan were all printable. 3/0.2% was not printable, but had a yield stress of 0.22 Pa.	Mouser et al. [18]
20–30% polaxamer/1–4% poly(ethylene glycol) (px/PEG)	0.34	~125 to 425	All formulations were printable.	Ribeiro et al. [19]
2.5% kappa-carrageenan/4–6% nanosilicates (κCA/nSi)	0.34	4.2–33	Pure κCA had the highest yield stress and was not printable. At least 4% nSi were needed for printability.	Wilson et al. [22]

In the current study, none of the 1 MDa PHA groups had yield stresses, but the 1 MDa − 8 wt% printed with good shape fidelity. The 8 wt% group had a high enough viscosity (i.e., 4 times greater than the 1.5 MDa − 4 wt%) to compensate for the lack of a yield stress within the time period given for crosslinking. Similarly, Paxton et al. [21] found a printable formulation of alginate-gelatin that did not have a yield stress, but was still able to be printed with good shape fidelity, which was accomplished by lowering the print bed temperature to increase the viscosity of the material upon deposition. While a high viscosity is not a substitute for a yield stress, we found that a material with a high viscosity may delay the movement of the material after printing, and aid in shape retainment, especially if a crosslinking mechanism is employed directly or soon after printing. Overall, the existence of a yield stress is the most important parameter to characterize for evaluating whether the material can retain a printed shape and the yield stress limit for a particular material and nozzle size can be used to determine printable formulations.

Given that the yield stress does not account for a material’s exposure to high shear stresses during the extrusion process, the storage modulus recovery is necessary to characterize to determine how well the yield stress and original properties are recovered after printing. In the current study, we demonstrated that PHA formulations that printed with good shape fidelity and the Cellink^® Start material had storage modulus recoveries above 85%, similar to others [17,39]. We propose a minimum threshold for storage modulus recovery of 85% for printable PHA formulations. While the recovery data may not be entirely relevant for other polymer systems, we believe the 85% target is a reasonable starting point and guideline for novel bioinks entering the literature. We additionally found that higher viscosities could mitigate the impact on the shape fidelity of materials with low recovery. Even though the 1.5 MDa − 3 wt% and 4 wt% PHA had similar recoveries of their storage modulus, it is possible that because the 3 wt% was less viscous, the shape fidelity started out with soft edges and deteriorated to no edges, where the 4 wt% printed with rigid edges and relaxed to normal defined edges. Crosslinkable hydrogels can additionally overcome low recoveries by initiating crosslinking throughout the printing process to conserve the printed structure immediately upon deposition [40]. The PHA hydrogels used here are advantageous for overcoming low recoveries because of the quick crosslinking time (~2 min) compared to other previously used methacrylated HA chemistries with longer crosslinking times (~5–30 min) [27,28]. Overall, for the PHA formulations, groups with good shape fidelity had storage modulus recoveries above 85%.

Given that many bioinks are used to bioprint cells in constructs that mimic native tissues or even print multiple cell types in different locations on the same structure, it is important to investigate the effects that cells have on the printability of bioinks and how well cells survive the printing and post-printing processes (e.g., UV crosslinking). To illustrate and quantify this point, we demonstrated that the increasing cell concentration in 1.5 MDa − 4 wt% PHA with 1 × 10⁶ cells/mL increased the yield stress but otherwise had minimal effects at higher cell concentrations up to 9 × 10⁶ - cells/mL on the printed shape fidelity, yield stress, and viscosity when compared to PHA with no cells. Other studies found opposite trends of decreasing bioink viscosity when 2.5 × 10⁶ cells/mL or less were encapsulated [37,38]. Given the discrepancy, for development of future bioinks, it will be important to take into account the effects of encapsulated cells on the printability of the bioinks and characterize bioinks accordingly. Given that the tested cell concentrations did not diminish printability, cell viability studies were conducted with 10⁶ cells/mL. Even though exposure to UV light may be harmful to cells, we have previously shown high viability of cells after hydrogel encapsulation and exposure to 312 nm light [27,28,36]. PHA has a fast thiol-ene crosslinking chemistry and only requires a short duration of UV exposure (i.e., 2 min) and a low concentration of photoinitiator I2959 (i.e., 0.05% w/v) for sufficient crosslinking, thus, we did not anticipate the UV exposure to significantly increase cell death. In support of our hypothesis, there was good viability of rBMSCs and rNSCs (i.e., greater than 87%) after exposure to UV light and additionally being extruded through a nozzle. Altogether, the high cell viabilities indicated that the 1.5 MDa − 4 wt% PHA precursor met the minimum biological requirement of keeping cells alive throughout the bioprinting process. The rBMSCs maintained a good viability after 7 days of culture; however, the rNSCs began to lose viability. Even though rNSCs initially survived the stresses of bioprinting, the rNSCs may not be as robust as the rBMSCs and experienced negative long-term effects from the bioprinting process. Several studies showed that cell viability after printing decreases as the viscosity of the material increases and as the nozzle diameter decreases because of the high shear stresses [41,42]. The rNSCs may have been more susceptible to damage from the high shear stresses compared to the rBMSCs. Low pressures and larger nozzles are favorable for printing cells [41,42], but larger nozzles reduce the resolution of the print. Low-viscosity crosslinkable materials may be able to overcome the limitations by simultaneously depositing and crosslinking the material to form stable structures [40].

Another consideration that may affect printability is the effect of different sterilization techniques on material composition and mechanical performance. Ethylene oxide was used as a sterilization technique for PHA to be used in cell viability studies and because ethylene oxide can participate in addition reactions with amines, carboxylic acids, hydroxyls, and sulfhydryl groups [43], preliminary studies with NMR showed there was most likely residual ethylene oxide in the sample, which could be eliminated by sample aeration after sterilization. Further, we found there was most likely an addition reaction of the ethylene oxide onto the carboxylic acids or hydroxyls present on the PHA backbone. Other studies found similar addition reactions occurring on materials with carboxylic acids [43,44]. In terms of rheological performance, preliminary studies with sterilized PHA showed that the ethylene oxide sterilization could have affected the yield stress and viscosity, but further verification will need to be performed in future studies. A few other studies found effects on other scaffolds after ethylene oxide sterilization in terms of yield strength, swelling, pore morphology, and molecular weight, but the effects were not the same among all the materials tested [43,45,46], which highlights the need for each material to be characterized individually, as the effects vary. Overall, we emphasize the need to characterize the effects on compositional and mechanical performance from the sterilization technique used because of the potential influences on the printability.

We have developed the printability of the PHA material in these studies; however, HA materials lack motifs for cell attachment, reducing the applicability for regenerative medicine. In preliminary studies, we found that rNSCs do not adhere or spread well when encapsulated or are seeded on PHA hydrogels. The same challenges arise with other inert synthetic polymers, but polymer modifications with adhesive peptides are used to overcome this challenge. Similar to synthetic polymers, PHA can be modified with thiolated bioactive factors, such as adhesive peptides, which can enhance the bioactivity for better regeneration of tissues. While the overall bioactivity may be enhanced, such modifications could influence or alter the printability. Preliminary studies involving bioprinted PHA modified with laminin-based adhesive peptides for enhancing rNSC adhesion resulted in similar printability, but the specific rheological effects need to be further investigated. We speculate the rheological performance could be dependent on peptide concentration, size, and interactions. At lower concentrations, the rheology could be dominated by the long polymer chains of PHA and similar to that of PHA alone; however, the conjugation of a high concentration or longer peptides may cause more peptide “branches” on PHA and thus, more physical entanglements and a potentially higher yield stress. On the other hand, the ability of peptides to interact with other peptides or the polymer via charge or hydrogen bonding may influence the interactions with neighboring peptides, causing aggregations or repulsions, and further influencing the bulk rheological properties and printability. From the immense variety of peptides that can be synthesized, the influence of peptide modifications to polymers on printability is another topic in need of further investigation by the bioprinting field.

5. Conclusions

There are several physical properties, printing parameters, and biological requirements that all influence the printability of a material from an extrusion-based bioprinter. In terms of physical properties, we assessed the printability of varying formulations of a hyaluronic acid-based hydrogel precursor, PHA, and characterized the formulations with three rheological tests: viscosity, yield stress, and storage modulus recovery. We found that the viscosity alone did not determine printability, but the viscosity behavior was valuable in determining if PHA had shear thinning behavior and would be extrudable through a nozzle during bioprinting. Further, possession of a yield stress was a critical characteristic for maintaining a printed shape; however, the yield stress upper limit for PHA through a 27G (i.e., 0.21 mm diameter) nozzle was 1000 Pa, where PHA formulations with yield stresses beyond 1000 Pa printed with irregular lines. While the yield stress could be a direct indicator of printability, a material could print with high shape fidelity if the viscosity was high enough to delay the material relaxation and a crosslinking mechanism was employed directly or soon after printing to retain the printed shape. Similarly, good storage modulus recovery after shearing was necessary for preventing material relaxation and maintaining the printed shape fidelity after being extruded, but materials with high viscosity could reduce the relaxation effect of PHA formulations that had low recovery through use of a quick UV crosslinking mechanism. Bioprinting has significant advantages for controlling the spatial location of different cell types to mimic the native hierarchy of tissues, but bioprinting cells comes with additional challenges. We found minimal effects of cell concentration on the printability of PHA, but others have found that it decreased viscosities. Additionally, we found that the carboxylic acid and hydroxyl-containing PHA may be susceptible to compositional changes from ethylene oxide sterilization, which may in turn influence the printability. Another parameter that may influence printability is the use of peptide modifications on PHA or other inert polymers to overcome the lack of motifs for cell attachment. Overall, cell concentration, sterilization methods, and peptide modifications are important considerations to account for during future bioink development. To demonstrate the feasibility of using PHA as a bioink, we found that two stem cell types each had good viabilities after printing. Overall, in these studies we developed a hyaluronic acid-based hydrogel into a bioink, characterized different formulations with three quantifiable rheological assessments, and determined printability windows for the yield stress and storage modulus recovery of PHA. We hope to accelerate future bioink development by recommending the rheological evaluation of the viscosity, yield stress, and storage modulus recovery to promote the use of standardized characterizations for more efficient, quantitative, and reliable methods to develop new printable bioinks.

Statement of significance.

Bioprinting has potential for regenerating damaged tissues; however, there are a limited number of printable biomaterials, and developing new bioinks is challenging because the required material physical properties for extrusion-based printing are not yet known. Most new bioinks are developed by trial-and-error, which is neither efficient nor comparable across materials. There is a need for the field to begin utilizing standard methods proposed by a few pioneering studies to characterize new bioinks. Therefore, we have developed the printability of a hyaluronic acid based-hydrogel and characterized the material with three quantitative rheological tests. The current work impacts the bioprinting field by demonstrating and encouraging the use of universal bioink characterizations and by providing printability windows to advance new bioink development.