Swelling Behaviors of 3D Printed Hydrogel and Hydrogel-Microcarrier Composite Scaffolds

Abstract

The present study sought to demonstrate the swelling behavior of hydrogel-microcarrier composite constructs to inform their use in controlled release and tissue engineering applications. In this study, gelatin methacrylate (GelMA) and GelMA-gelatin microparticle (GMP) composite constructs were three-dimensionally printed, and their swelling and degradation behavior was evaluated over time and as a function of the degree of crosslinking of included GMPs. GelMA-only constructs and composite constructs loaded with GMPs crosslinked with 10 mM (GMP-10) or 40 mM (GMP-40) glutaraldehyde were swollen in phosphate-buffered saline for up to 28 days to evaluate changes in swelling and polymer loss. In addition, scaffold reswelling capacity was evaluated under five successive drying–rehydration cycles. All printed materials demonstrated shear thinning behavior, with microparticle additives significantly increasing viscosity relative to the GelMA-only solution. Swelling results demonstrated that for GelMA/GMP-10 and GelMA/GMP-40 scaffolds, fold and volumetric swelling were statistically higher and lower, respectively, than for GelMA-only scaffolds after 28 days, and the volumetric swelling of GelMA and GelMA/GMP-40 scaffolds decreased over time. After 5 drying–rehydration cycles, GelMA scaffolds demonstrated higher fold swelling than both GMP groups while also showing lower volumetric swelling than GMP groups. Although statistical differences were not observed in the swelling of GMP-10 and GMP-40 particles alone, the interaction of GelMA/GMP demonstrated a significant effect on the swelling behaviors of composite scaffolds. These results demonstrate an example hydrogel-microcarrier composite system's swelling behavior and can inform the future use of such a composite system for controlled delivery of bioactive molecules in vitro and in vivo in tissue engineering applications.

Impact statement

In this study, porous three-dimensional printed (3DP) hydrogel constructs with and without natural polymer microcarriers were fabricated to observe swelling and degradation behavior under continuous swelling and drying–rehydration cycle conditions. Inclusion of microcarriers with different crosslinking densities led to distinct swelling behaviors for each biomaterial ink tested. 3DP hydrogel and hydrogel-microcarrier composite scaffolds have been commonly used in tissue engineering for the delivery of biomolecules. This study demonstrates the swelling behavior of porous hydrogel and hydrogel-microcarrier scaffolds that may inform later use of such materials for controlled release applications in a variety of fields including materials development and tissue regeneration.

Introduction

Hydrogels have gained significant popularity over the years as biomaterials in tissue engineering. In particular, gelatin methacrylate (GelMA) has been widely used in the development of scaffolds for cartilage, the osteochondral unit, and a variety of other tissues. GelMA is an attractive biomaterial for several reasons. It is inexpensive, highly biocompatible, offers several biochemical cues found in native tissues, and has tunable biochemical properties.^1–12 Important in material processing and scaffold fabrication, GelMA is both physically and chemically crosslinkable through temperature and UV light exposure in the presence of appropriate initiators, respectively. After UV crosslinking, GelMA constructs remain stable at physiological temperatures, making them attractive for in vitro and in vivo tissue engineering applications. GelMA has also become a popular biomaterial specifically for 3D printing applications.^13–19 GelMA can be printed at low temperatures and pressures, making fabrication of GelMA constructs accessible to a wide range of 3D printers with different operating capabilities and allowing for the incorporation of biological materials, such as cells and bioactive molecules, which might otherwise be adversely affected and/or inactivated by more severe processing conditions. The ability to physically and chemically crosslink GelMA is particularly valuable in a 3D printing context, permitting the development of structures with highly specific architectures that are stabilized by first controlling the temperature of the gel and then crosslinking the printed construct with UV light.

Micron-scale carriers have similarly been widely used in tissue engineering, specifically for the controlled delivery of encapsulated bioactive molecules, including ceramics and growth factors. In particular, the use of poly(lactic-co-glycolic acid) and gelatin-based microparticles (GMPs) in spatiotemporal growth factor delivery has been well-reported.^20–30 When embedded within a bulk polymer matrix, particularly in a 3D printing context, these microparticle carriers allow for location-specific distribution of bioactive molecules, which can be used to create spatial gradients in concentration and type of added factors. In addition, as has been demonstrated previously, varying the degree of crosslinking or block copolymer ratios of these particles can yield temporal control over the release of encapsulated factors.^20,23,27,31

As stated, GelMA has become an attractive material for 3D printing, as it is low cost, can be printed at low temperatures and pressures, be both physically and chemically crosslinked, and support a wide variety of additives. Similarly, the controlled release of loaded growth factors and other bioactive molecules from GMPs is well studied. However, to the authors' knowledge, few reports have examined the extended swelling behavior of porous hydrogel-based scaffolds and that of hydrogel-microcarrier composite scaffolds beyond 24 h, which would provide valuable insight relevant to use of such constructs in vivo. In addition, changes in swelling behavior of such printed scaffolds after cycles of drying and rehydration have not been widely examined, although capacity for reswelling is critical in the development, handling, and ease of use of tissue repair scaffolds in in vivo research and in the clinic. As a result, the focus of this work is the demonstration of both extended-term swelling and drying–reswelling behaviors of porous 3DP hydrogel and hydrogel-composite scaffolds.

In this study, porous hydrogel-based constructs were fabricated through extrusion 3D printing to determine the individual and combined swelling behavior of the bulk hydrogel and microparticle additive components. GelMA was synthesized as a bulk biomaterial ink for printing. In addition, GMPs were fabricated with two different degrees of crosslinking and incorporated within GelMA inks to produce two additional hydrogel-particle composite inks. Each material was evaluated for rheological properties, and scaffolds of each material were then evaluated for their swelling and degradation behavior. Specifically, this study focused on swelling and degradation of (i) printed porous hydrogel scaffolds and microparticle carriers; (ii) printed porous hydrogel-microparticle composite scaffolds; and (iii) hydrogel and hydrogel-composite scaffolds under cycles of drying and rehydration.

Materials and Methods

Experimental design

To evaluate the objectives of the study as outlined previously, eight different experimental groups (Supplementary Table S1) were examined for understanding the swelling and degradation behavior of 3DP hydrogel composites. Scaffolds were constructed solely of GelMA or of GelMA containing GMPs crosslinked with either 10 mM (GMP-10) or 40 mM (GMP-40) glutaraldehyde. To understand the individual and combined behaviors of GelMA and GMPs, all three scaffold formulations were examined under an extended swelling experiment (Fig. 1a) and cycles of drying and rehydration (Fig. 1b) as described hereunder. Moreover, GMPs were additionally examined in a separate swelling study to understand their inherent swelling behaviors when not confined within a printed GelMA network (Fig. 1c).

FIG. 1.

Schematics of swelling experiments performed in this study. (a) Summary of full-length study of scaffold swelling. GelMA-based scaffolds were fabricated using 3D printing and agitated in PBS at 37°C and 100 rpm for up to 28 days. At each timepoint (1, 7, 14, 21, and 28 days) scaffolds were removed from solution and weighed after removing surface water. Scaffolds were then dried in open air and under vacuum before obtaining final dry weight. (b) Summary of scaffold swelling cycle study. GelMA-based scaffolds were fabricated using 3D printing and agitated in PBS at 37°C and 100 rpm for five cycles of 24-hour swelling and drying. After each cycle, scaffolds were removed from solution and weighed after removing surface water before being returned to the swelling conditions. After cycle 5, scaffolds were dried in open air and under vacuum before obtaining final dry weight. (c) Summary of full-length study of particle swelling. GMP-10 and GMP-40 were agitated in PBS at 37°C and 100 rpm for up to 28 days. At each timepoint (1, 7, 14, 21, and 28 days) particles were removed from solution and weighed after removing surface water. Particles were then dried in open air and under vacuum before obtaining final dry weight. GelMA, gelatin methacrylate; GMP, gelatin microparticle.

Synthesis of GelMA and GMPs

Synthesis of GelMA was carried out using a modified protocol of that developed by Van Den Bulcke et al.³² In brief, gelatin type B (Nitta Gelatin, Inc., Osaka, Japan) was dissolved at 10% w/v in Dulbecco's phosphate-buffered saline (DPBS) under stirring at 50°C. Methacrylic anhydride (Sigma, St. Louis, MO) was then added at 1.25% v/v and allowed to react for 1 h, after which the vessel was diluted 5 × with warm (37°C) DPBS to quench. The gelatin solution was then dialyzed against ddH₂O for 7 days at 37°C using 10 kDa dialysis tubing. After dialysis, gelatin solutions were flash frozen in liquid nitrogen and lyophilized for 7 days to obtain the modified GelMA.

GMPs were also fabricated from gelatin type B following established procedures.^20,33 In brief, gelatin was dissolved at 10% w/v in ddH₂O at 60°C. The dissolved gelatin was added dropwise to 250 mL olive oil containing 0.5 wt% Span80 while stirring at 500 rpm. After stirring for 30 min, 100 mL of chilled acetone was added to the reaction vessel. After an additional 60 min, the microparticles were collected by filtration and washed with acetone. GMPs were then cross-linked in 0.1 wt% Tween 80 (Sigma) solution with either 10 mM (GMP-10) or 40 mM (GMP-40) glutaraldehyde (Sigma) while stirring at 500 rpm at 15°C. After 20 h, 25 mM glycine was added for 1 h to quench unreacted glutaraldehyde.

Cross-linked GMPs were then collected by filtration, washed with 5 ratios (100:0, 75:25, 50:50, 25:75, and 0:100) of chilled ddH₂O and acetone, and then frozen at −80°C for 1 h before being lyophilized overnight. Finally, dried GMPs were sieved to obtain particles of 50–100 μm in diameter. Owing to the yield of particles within this diameter range, GMP-10 and GMP-40 particles used in this study were each pooled from 10 separate batches of microparticle fabrication. To confirm microparticle formation, both groups of microparticles were swollen in excess PBS containing 0.1% fluorescein and imaged using confocal microscopy. Representative images of the microparticle groups are given in Figure 2.

FIG. 2.

Epifluorescence imaging of GMP-10 (left) and GMP-40 (right) loaded with fluorescein taken at 20 × magnification. Scale bar 200 μm.

Preparation and rheological characterization of print solutions

To create print solutions of each ink, GelMA was dissolved at 10% w/v in ddH₂O at 37°C. 2-hydroxy-1-[4-(2-hydroxyethoxy)phenyl]-2-methyl-1-propanone (Irgacure 2959; Sigma) was added as a photoinitiator at 0.5% w/v. For GelMA print solution, the dissolved gelatin solution was loaded into low-temperature print cartridges (EnvisionTEC, Gladbeck, Germany) and stored under refrigeration until printing. For GelMA/GMP composite print solutions, GMP-10 or GMP-40 particles were dispersed in the dissolved GelMA solution at 5% w/v before storage.

Rheological evaluations of each ink were performed with a Discovery Hybrid Rheometer-2 (TA Instruments) using a plate geometry (40 mm, 0.4 mm gap distance) with a solvent trap. Materials were evaluated for shear thinning behavior via a shear rate ramp in flow mode with shear rates ranging from 0.01 to 1000 s⁻¹. In addition, a shear stress ramp flow protocol ranging from 0.01 to 200 Pa was conducted for each material and the yield point determined from the intersection point of linear regressions of the plateau and viscosity drop regions of the resulting viscosity-shear stress diagrams. All measurements were performed at 26°C using n = 3 samples per material.

Fabrication of 3D Printed GelMA and GelMA/GMP composite scaffolds

Scaffolds were fabricated through extrusion 3D printing by adapting previous protocols in our research group.^16,34–36 In brief, cuboid scaffolds (8 × 8 × 4.3 mm) were designed in SketchUp (Trimble, Sunnyvale, CA) and sliced at a layer thickness of 0.4 mm in Perfactory Software (EnvisionTEC) for 3D printing. GelMA and GelMA/GMP composite inks were extruded according to the printing parameters in Table 1 through a 22G conical plastic needle at 0.4 mm offset using a 3D-Bioplotter (EnvisionTEC). Preflow and postflow refer to the duration of extrusion before starting or after completing each layer, respectively.

Table 1.

Printing Parameters for Hydrogel Materials

Material	Temperature (°C)	Pressure (bar)	Speed (mm/s)	Preflow (s)	Postflow (s)
GelMA	25.0	0.35	1.5	0.6	2.0
GelMA/GMP-10	26.5	0.65	2.0	0.8	0.55
GelMA/GMP-40	26.0	0.5	2.0	1.0	0.55

The printing pattern consisted of five continuous strands with 1.7 mm on-center spacing, as given in Figure 3. Subsequent layers were printed at a 90° angle with respect to the previous layer, up to a total height of 12 layers per scaffold. After depositing each layer, the print was crosslinked using UV light (365 nm) at 5 mW/cm² for 30 s. An additional wait time of 30 s was added before printing the next layer.

FIG. 3.

GelMA/GMP-40 composite fiber. Stitched image of single printed composite hydrogel/particle fiber taken using epifluorescence imaging. GMP-40 (observed as red dots within the bulk fiber) were stained overnight with Nile red and dried before loading at 5% w/v and printing. Scale bar 1 mm.

Swelling of 3DP hydrogel scaffolds

After printing, scaffolds were weighed (W_f) and then underwent swelling in DPBS by adapting previous protocols.³⁷ In brief, scaffolds were placed in scintillation vials with 10 mL sterile DPBS and swollen under agitation at 100 rpm and 37°C for up to 28 days. DPBS was replaced after 12 h and then at days 1, 4, 7, 10, 14, 18, 21, 24, and 28. At days 1, 7, 14, 21, or 28, scaffolds were removed from solution, excess liquid was removed using wax paper, and then scaffolds were weighed again to obtain the swollen weight (W_s). Scaffolds were then air dried for 24 h and then dried under vacuum for 24 h before obtaining the dry weight (W_d).

Separately, scaffolds were evaluated for changes in swelling under cycles of drying and rehydration. Scaffolds were weighed after printing and underwent swelling in DPBS for 24 h, then weighed, dried, and weighed again as described. After this initial process, each scaffold was subjected to an additional four cycles of reswelling and drying. Swelling of all scaffolds was conducted using n = 3 samples per group.

Swelling of GMPs

To evaluate swelling behavior of GMPs alone, 40 mg of GMP-10 and GMP-40 were swollen for up to 28 days in preweighed 40 μm cell strainers in six-well plates in 10 mL DPBS using the same timepoints as for scaffold swelling. At each timepoint, excess liquid was removed, the “wet” (swollen weight) of the particles and strainer was recorded, and then particles and strainer were dried as described previously, before recording final dry weight. Swelling of particles was conducted using n = 3 samples per group. In a previous study in our research group, GMPs were swollen in PBS, collected on filter paper, and then added to dry vials of known weight.³⁸

Calculation of swelling parameters

Swelling parameters (fold swelling, polymer loss, and volumetric swelling) were calculated according to previous methods.^37,38 The fold swelling ratio was determined according to Equation (1):

$$Foldswelling=\frac{W_s-W_d}{W_d}$$ (1)

where W_s is the wet (swollen) weight and W_d is the dry weight taken after vacuum drying. Polymer loss for scaffolds was determined using Equation (2a):

$$Polymerloss=\frac{W_{f,d}-W_d}{W_{f,d}}$$ (2a)

where W_d is again the dry weight, and W_f,d is the theoretical dry component of the printed weight (0.1W_f for GelMA scaffolds and 0.15W_f for GelMA/GMP scaffolds). Similarly, polymer loss for particles was determined according to Equation 2b:

$$Polymerloss=\frac{W_f-W_d}{W_f}$$ (2b)

where W_f represents the original weight of the particles added to the cell strainer. Finally, Volumetric swelling of scaffolds and particles was determined according to Equation 3:

$$Volumetricswelling=\frac{\left(1-PL\right)W_s}{W_f}$$ (3)

where PL is the polymer loss from Equation 2a or 2b, and W_s and W_f are the swollen and initial weights, respectively.

Statistical analysis

Prism 8.0 (GraphPad Software, San Diego, CA) was used to evaluate the data set using a two-way analysis of variance (ANOVA) test with post hoc analysis by Tukey's honest significant difference (HSD). Results were considered significant at p < 0.05. Results are presented as mean ± standard deviation.

Results

Rheological evaluation of print materials

Preprinting rheological evaluations of each hydrogel material were performed with a Discovery Hybrid Rheometer-2 using a plate geometry (40 mm, 0.4 mm gap distance) with a solvent trap. Using a shear rate ramp from 0.01 to 1000 s⁻¹, shear thinning behavior for all hydrogel inks was determined as given in Figure 4a. As expected, all three hydrogel inks demonstrated shear thinning behavior over the course of the ramp, indicating extrudability of all inks. In addition, the GelMA-only ink displayed a significantly lower viscosity at all shear rates than the GelMA/GMP inks.

FIG. 4.

Rheological behavior of GelMA and GelMA/GMP inks. (a) Shear thinning behavior of GelMA and GelMA/GMP inks. Both GelMA and GelMA/GMP composite inks demonstrate thinning behavior with increasing shear rate. As expected, viscosity of the GelMA only ink is consistently lower than of the composite inks. The GelMA/GMP-10 composite ink demonstrates higher viscosity than the GelMA/GMP-40 composite ink over the range of shear rates tested. (b) Yield point determination of GelMA and GelMA/GMP inks. The GelMA only ink demonstrates a significantly lower yield point than the composite inks at 14.7 ± 11.8 Pa. The yield points of the composite inks occur at 118.0 ± 12.9 Pa and 77.8 ± 16.2 Pa, respectively, for GelMA/GMP-10 and GelMA/GMP-40 composite inks. Data are reported as mean ± standard deviation for n = 3 samples/group.

In comparing the composite solutions, the GelMA/GMP-10 ink demonstrated a higher viscosity than the GelMA/GMP-40 ink across the range of shear rates tested, and also appeared to show more dramatic shear thinning behavior at higher shear rates relative to the GelMA or GelMA/GMP-40 inks. As given in Figure 4b, a shear stress ramp protocol was also conducted to determine the yield point of each material ink. Using the intersection of linear regressions of the plateau and viscosity drop regions of each curve, the yield point of each material was approximated to be 14.7 ± 11.9, 118.0 ± 12.9, and 77.8 ± 16.2 Pa, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40, with statistical differences across all groups at α = 0.05.

Swelling behavior of scaffolds up to 28 days

To examine swelling behavior of printed scaffolds, each sample was swollen in DPBS while shaking at 100 rpm and 37°C for up to 28 days. GelMA/GMP-40 scaffolds from various points in the study are given in Figure 5, which shows scaffolds after printing (upper left), after 7 (upper middle), 14 (upper right), 21 (lower left), and 28 (lower middle) days of swelling, respectively, and after drying (lower right). All scale bars are 5 mm.

FIG. 5.

GelMA/GMP-40 scaffolds (12 layers) used in swelling study. Scaffolds are shown after printing (upper left), 7 (upper middle), 14 (upper right), 21 (lower left), and 28 (lower middle) days of swelling, respectively, and drying (lower right). All scale bars 5 mm.

Results for fold swelling, which indicates the ratio of liquid to dry polymer content in each scaffold owing to solution uptake, are given in Figure 6a. As given in Figure 6a, there were no significant differences in the fold swelling of the GelMA, GelMA/GMP-10, and GelMA/GMP-40 scaffolds within each timepoint up to 21 days. At 28 days, there was a significant increase in the fold swelling of the GelMA/GMP-10 scaffold group relative to other groups, whereas there was a significant reduction in the fold swelling of the GelMA/GMP-40 scaffold group relative to other groups (8.6 ± 0.5, 10.3 ± 0.6, and 6.9 ± 0.1, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40 after 28 days). There did not appear to be statistically significant temporal effects on fold swelling of scaffolds.

FIG. 6.

Swelling of 3DP GelMA and GelMA/GMP composite scaffolds over 28 days. (a) Fold swelling over 28 days. There were no significant differences in the fold swelling of the GelMA, GelMA/GMP-10, and GelMA/GMP-40 scaffolds within each timepoint up to 21 days. At 28 days, there was a significant increase in the fold swelling of the GelMA/GMP-10 scaffold group relative to other groups, whereas there was a significant reduction in the fold swelling of the GelMA/GMP-40 scaffold group relative to other groups (8.6 ± 0.5, 10.3 ± 0.6, and 6.9 ± 0.1, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40 after 28 days). There did not appear to be statistically significant temporal effects on fold swelling of scaffolds. (b) Polymer loss over 28 days. At days 1, 7, 14, and 28, there was significantly higher polymer loss in GelMA/GMP-40 scaffolds compared with GelMA scaffolds, which was also significantly higher than for GelMA/GMP-10 scaffolds at days 1 and 28. At days 1, 7, and 28, polymer loss in GelMA/GMP-10 scaffolds was also significantly higher than in GelMA scaffolds. Some temporal effects were also observed for polymer loss. For GelMA scaffolds, there was a significant increase in polymer loss after 21 days relative to 1 and 7 days, whereas for GelMA/GMP-40 scaffolds, polymer loss was significantly increased at day 28 relative to days 1, 7, and 14. There were no significant temporal effects observed for GelMA/GMP-10 scaffolds. (c) Volumetric swelling over 28 days. Volumetric swelling of GelMA/GMP-10 scaffolds remained statistically higher than GelMA and GelMA/GMP-40 scaffolds at each timepoint other than day 7. GelMA and GelMA/GMP-40 scaffolds maintained statistically similar volumetric swelling at all timepoints up to 21 days. At 28 days, the volumetric swelling of GelMA/GMP-40 scaffolds was significantly reduced compared with other groups, consistent with the behavior observed in fold swelling and polymer loss at day 28 (0.8 ± 0.03, 1.1 ± 0.2, and 0.5 ± 0.1, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40). For GelMA/GMP-10 scaffolds, there were no statistically significant temporal effects on volumetric swelling. GelMA scaffolds demonstrated some temporal dependence, with decreased volumetric swelling after 21 days compared with 1 and 7 days. After 28 days, volumetric swelling of GelMA/GMP-40 scaffolds was significantly lower than all other timepoints (1, 7, 14, and 21 days), although no significant effects were observed between earlier timepoints in these scaffolds. W_s, swollen weight of scaffold at each time point; W_d, dry weight of scaffold; W_f,d, weight of polymer within scaffold at printing; PL, polymer loss; W_s, swollen weight of scaffold; W_f, printed weight of scaffold. Within each timepoint, groups that do not share the same letter (a, b, c) represent a statistically significant difference. Statistical differences within a scaffold group across timepoints are designated using the symbols shown. Data are reported as mean ± standard deviation for n = 3 samples/group.

Polymer loss of each of the scaffold groups is given in Figure 6b. At days 1, 7, 14, and 28, there was significantly higher polymer loss in GelMA/GMP-40 scaffolds compared with GelMA scaffolds, which was also significantly higher than for GelMA/GMP-10 scaffolds at days 1 and 28. At days 1, 7, and 28, polymer loss in GelMA/GMP-10 scaffolds was also significantly higher than that in GelMA scaffolds. Some temporal effects were also observed for polymer loss. For GelMA scaffolds, there was a significant increase in polymer loss after 21 days relative to 1 and 7 days, whereas for GelMA/GMP-40 scaffolds, polymer loss was significantly increased at day 28 relative to days 1, 7, and 14. There were no significant temporal effects observed for GelMA/GMP-10 scaffolds.

Volumetric swelling of scaffolds, which indicates ratio of the liquid component of the gel after swelling to that after printing, is given in Figure 6c. As shown, volumetric swelling of GelMA/GMP-10 scaffolds remained statistically higher than GelMA and GelMA/GMP-40 scaffolds at each timepoint other than day 7. GelMA and GelMA/GMP-40 scaffolds maintained statistically similar volumetric swelling at all timepoints up to 21 days.

At 28 days, the volumetric swelling of GelMA/GMP-40 scaffolds was significantly reduced compared with other groups, consistent with the behavior observed in fold swelling and polymer loss at day 28 (0.8 ± 0.03, 1.1 ± 0.2, and 0.5 ± 0.1, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40). For GelMA/GMP-10 scaffolds, there were no statistically significant temporal effects on volumetric swelling. GelMA scaffolds demonstrated some temporal dependence, with decreased volumetric swelling after 21 days compared with 1 and 7 days. After 28 days, volumetric swelling of GelMA/GMP-40 scaffolds was significantly lower than all other timepoints (1, 7, 14, and 21 days), although no significant effects were observed between earlier timepoints in these scaffolds.

Swelling behavior of scaffolds under rehydration cycles

To determine the swelling behavior of scaffolds after consecutive rehydration–drying cycles, additional scaffolds from each material group were printed and subjected to five cycles of swelling and drying under the same conditions as the original swelling study. As given in Figure 7a, no statistically significant effects of rehydration cycles were observed for the fold swelling of any of the scaffold groups. All three groups also maintained statistical similarity up to the third swelling cycle, after which the GelMA group displayed significantly higher fold swelling (7.8 ± 0.5) compared with other groups (6.2 ± 0.5 and 6.1 ± 0.4 for GelMA/GMP-10 and GelMA/GMP-40, respectively) for cycles 4 and 5. As given in Figure 7b, polymer loss was statistically similar between GelMA/GMP-10 and GelMA/GMP-40 scaffolds across all swelling cycles, and effects of swelling cycle number on polymer loss were not observed for these groups.

FIG. 7.

Swelling of 3DP GelMA and GelMA/GMP composite scaffolds over five swelling cycles. (a) Fold swelling over five swelling cycles. No statistically significant effects of rehydration cycles were observed for the fold swelling of any of the scaffold groups. All three groups also maintained statistical similarity up to the third swelling cycle, after which the GelMA group displayed significantly higher fold swelling (7.8 ± 0.5) compared with other groups (6.2 ± 0.5 and 6.1 ± 0.4 for GelMA/GMP-10 and GelMA/GMP-40, respectively) for cycles 4 and 5. (b) Polymer loss over five swelling cycles. Polymer loss was statistically similar between GelMA/GMP-10 and GelMA/GMP-40 scaffolds across all swelling cycles, and effects of swelling cycle number on polymer loss were not observed for these groups. Polymer loss of GelMA scaffolds was statistically lower than for both GelMA/GMP groups for swelling cycles 1–3, and then was statistically similar in cycles 4 and 5. (c) Volumetric swelling over five swelling cycles. All groups demonstrated statistically similar volumetric swelling ratios within each swelling cycle up to cycle 4. After the fifth swelling cycle, GelMA scaffolds showed lower volumetric swelling (0.6 ± 0.09) than GelMA/GMP-10 (0.7 ± 0.07) or GelMA/GMP-40 (0.7 ± 0.07) scaffolds. In addition, effects of swelling cycle number were observed for all groups. Specifically, the volumetric swelling ratios after cycles 3, 4, and 5 were each statistically lower than at cycle 1 for all groups. In addition, the GelMA and GelMA/GMP-40 groups each demonstrated significantly reduced volumetric swelling at cycle 5 compared with cycles 2 and 3. Within each cycle, groups that do not share the same letter (a, b) represent a statistically significant difference. Statistical differences within a scaffold group across cycles are designated using the symbols shown. Data are reported as mean ± standard deviation for n = 3 samples/group.

Polymer loss of GelMA scaffolds was statistically lower than that for both GelMA/GMP groups for swelling cycles 1–3, and then was statistically similar in cycles 4 and 5. Finally, volumetric swelling of these scaffold groups is given in Figure 7c. All groups demonstrated statistically similar volumetric swelling ratios within each swelling cycle up to cycle 4. After the 5th swelling cycle, GelMA scaffolds showed lower volumetric swelling (0.6 ± 0.09) than GelMA/GMP-10 (0.7 ± 0.07) or GelMA/GMP-40 (0.7 ± 0.07) scaffolds. In addition, effects of swelling cycle number were observed for all groups. Specifically, the volumetric swelling ratios after cycles 3, 4, and 5 were each statistically lower than at cycle 1 for all groups. The GelMA and GelMA/GMP-40 groups each demonstrated significantly reduced volumetric swelling at cycle 5 compared with cycles 2 and 3.

Swelling behavior of particles up to 28 days

To evaluate the swelling behavior of GMP-10 and GMP-40 particles separately from composite scaffolds, GMPs were swollen in DPBS using 40 μm cell strainers for up to 28 days. As given in Figure 8a, there were no significant differences observed in fold swelling between GMP-10 and GMP-40 at each time point (4.1 ± 0.5 and 4.0 ± 0.3 for GMP-10 and GMP-40, respectively, after 28 days). In addition, there were no temporal effects on fold swelling observed aside from significantly higher swelling in GMP-10 particles at day 14 relative to day 1.

FIG. 8.

Swelling of GMP-10 and GMP-40 particles over 28 days. (a) Fold swelling over 28 days. No significant differences in fold swelling were observed in fold swelling between GMP-10 and GMP-40 groups at each time point (4.1 ± 0.5 and 4.0 ± 0.3 for GMP-10 and GMP-40 particles, respectively, after 28 days), as designated by “ns.” In addition, there were no temporal effects on fold swelling observed aside from significantly higher swelling in the GMP-10 group at day 14 relative to day 1. (b) Polymer loss over 28 days. There were no significant differences in polymer loss observed between GMP-10 and GMP-40 groups at each time point, as designated by “ns.” However, both groups demonstrated significantly lower polymer loss at day 21 relative to day 1. (c) Volumetric swelling over 28 days. Similar to fold swelling, there were no significant differences observed in volumetric swelling ratio between GMP-10 and GMP-40 at each timepoint, as designated by “ns,” whereas both groups showed increased swelling at day 14 (5.5 ± 0.1 and 5.4 ± 0.2) relative to day 1 (3.8 ± 0.6 and 3.8 ± 0.6). Within each timepoint, groups that do not share the same letter represent a statistically significant difference. Statistical differences within a scaffold group across timepoints are designated using the symbols shown. Data are reported as mean ± standard deviation for n = 3 samples/group.

Polymer loss results from GMPs are given in Figure 8b. There were no significant differences in polymer loss observed between GMP-10 and GMP-40 at each time point. However, both groups demonstrated significantly lower polymer loss at day 21 relative to day 1. Finally, volumetric swelling of GMPs is given in Figure 8c. Similar to fold swelling, there were no significant differences observed in volumetric swelling ratio between GMP-10 and GMP-40 at each timepoint, whereas both groups showed increased swelling at day 14 (5.5 ± 0.1 and 5.4 ± 0.2) relative to day 1 (3.8 ± 0.6 and 3.8 ± 0.6).

Discussion

In this study, porous hydrogel constructs were fabricated by extrusion 3D printing to determine the individual and combined swelling behavior of the bulk hydrogel and microparticle additive components. Specifically, this study focused on swelling and degradation of (i) printed porous hydrogel scaffolds and microparticle carriers independently; (ii) printed porous hydrogel-microparticle composite scaffolds; and (iii) hydrogel and hydrogel-composite scaffolds under cycles of drying and rehydration.

To investigate these objectives, constructs of three biomaterial inks (GelMA, GelMA with 10 mM GMPs, and GelMA with 40 mM GMPs) were fabricated using 3DP and evaluated for swelling behavior under continuous swelling and drying–rehydration conditions. The results of this study demonstrated distinct differences in the swelling behavior of GelMA constructs depending on the incorporation and crosslinking density of GMPs within the bulk gel. In particular, after 28 days of swelling the incorporation of GMP-10 led to increased fold and volumetric swelling relative to GelMA, whereas conversely the incorporation of GMP-40 led to decreases in these values. Under cycles of drying and rehydration, GelMA/GMP scaffolds behaved similar to each other after five cycles, with both groups having decreased fold and increased volumetric swelling, respectively, compared with GelMA.

In evaluating the rheological properties of GelMA and GelMA/GMP composite materials, all three hydrogel formulations demonstrated shear thinning behavior, which is to be expected from gelatin-based materials. In addition, as expected, the incorporation of GMPs increased the viscosity of composite inks at all shear rates. In comparing the composite solutions, the GelMA/GMP-10 ink consistently demonstrated a higher viscosity than the GelMA/GMP-40 ink and a more dramatic shear thinning behavior at higher shear rates.

A shear stress ramp protocol was also conducted to determine the yield point of each material ink. Consistent with the previous results, the GelMA/GMP inks demonstrated higher yield points than the GelMA-only ink, with GelMA/GMP-10 also being higher than GelMA/GMP-40. The initial conclusions drawn from these results were, whereas both GMPs would swell, increasing viscosity relative to the GelMA-only ink, GMP-10 in particular would demonstrate higher swelling compared with GMP-40 as has been shown previously,^37–39 leading to the highest viscosity in that group. However, as discussed hereunder in particle swelling experiments, the true mechanism may be more complex.

In evaluating swelling behavior under extended swelling conditions, there were no significant effects of swelling time on the fold swelling of any of the scaffold groups, indicating that under swelling in DPBS at 37°C, these scaffolds reach equilibrium fold swelling after 24 h and maintain such equilibrium until at least day 28. Although there were no differences between each group at prior timepoints, after 28 days, there was a significant increase in the fold swelling of the GelMA/GMP-10 scaffold group relative to other groups, whereas there was a significant reduction in the fold swelling of the GelMA/GMP-40 scaffold group relative to other groups (8.6 ± 0.5, 10.3 ± 0.6, and 6.9 ± 0.1, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40 after 28 days). Although few reports in the literature evaluate the fold swelling of scaffolds out to 28 days, determinations of fold swelling are common at the 24-h timepoint. The fold swelling ratios for GelMA, GelMA/GMP-10, and GelMA/GMP-40 scaffolds, respectively, after 24 h were 8.0 ± 1.4, 8.8 ± 0.6, and 8.3 ± 0.4. The 24-h fold swelling ratio for GelMA observed in this study is relatively consistent with prior literature.

For example, Celikkin et al. and Zhao et al. reported fold swelling ratios of ∼8.0 and 7.2 ± 0.2, respectively, for 10% w/v GelMA produced under similar synthesis and crosslinking conditions, whereas Nichol et al. and Krishnamoorthy et al. reported slightly higher fold swelling ratios of ∼15 and ∼12, respectively, for similar procedures but with lower amounts of added methacrylic anhydride and less UV intensity for crosslinking.^2,8–10 Hoch et al. similarly reported fold swelling ratios between ∼5.5 and ∼8.0 for GelMA constructs created using four different amounts of methacrylic anhydride.¹¹ The general similarity of these results suggest that although mass swelling ratio may not be an intrinsic material property—one that is independent of dimension or configuration—the extent of methacrylation and subsequent extent of crosslinking are larger driving forces in swelling behavior. This can be observed when considering that in studies where either the amount of added methacrylic anhydride or crosslinking intensity and time are relatively high (e.g., crosslinking in this study and Celikkin et al., MAA addition in Zhao et al. and Hoch et al.), consistently lower swelling ratios are observed than for the procedures used by Nichol et al. and Krishnamoorthy et al., which yielded higher swelling ratios presumably owing to lower crosslinking.

In a separate study of unmodified type A gelatin swelling, D'Agostino et al. reported a swelling ratio of ∼4.5 for 10% gelatin after 24 h, with decreasing fold swelling up to 75 h. Two factors likely influenced the different observations here. First, the authors swelled unmodified gelatin as opposed to GelMA in this study, likely influencing the trend of swelling behavior after 24 h owing to the lack of chemical crosslinking within the gel. Second, although methacrylation may reduce overall swelling capacity,^7,8 the authors also swelled constructs at 25°C, which likely led to the scale of the swelling ratio being much lower given that 10% w/v gelatin is largely gelled at 25°C compared with 37°C, which is widely used to approximate physiological temperature.⁴⁰

To the authors' knowledge, there have been no earlier studies of GelMA bulk-GMP microcarrier composite swelling; however, corollaries may exist in the literature. For example, Hutson et al. evaluated the swelling behavior of PEG-GelMA composite gels. Although there were no GelMA-only scaffolds evaluated, a corollary for the GelMA-only scaffold in this study may be their 10% PEG/0% GelMA group, which demonstrated a swelling ratio of ∼14. Similarly, the nearest corollary to the GelMA/GMP groups used in this study from that of Hutson et al. may be the 5% PEG/10% GelMA, which had a swelling ratio of ∼13 compared with 8.8 or 8.3 for GelMA/GMP-10 and GelMA/GMP-40 in this study.¹² The mass swelling from their study was also determined as wet weight over dry weight; recalculation of fold swelling values from this study would yield 9.0 ± 1.4, 9.8 ± 0.6, and 9.3 ± 0.4, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40 scaffolds.

In previous studies, Holland et al., Kim et al., and Lam et al. each evaluated the swelling behavior of oligo(poly(ethylene glycol) fumarate) (OPF) bulk gels with GMP-10 and GMP-40 incorporated.^30,37,41 Holland et al. observed apparent increases in swelling ratio of OPF, OPF/GMP-10, and OPF/GMP-40 scaffolds at day 28 relative to day 1, but statistical similarity between groups at day 28.³⁷ The authors also observed polymer loss of ∼80–90% for all groups after 28 days under enzymatic conditions, compared with 10–40% under standard conditions in this study. Kim et al. demonstrated, under standard PBS conditions, no temporal differences in fold swelling of bilayered OPF/GMP (GMP-10 in top layer and GMP-40 in bottom layer) constructs over 28 days, as well as statistically similar sol fraction values between ∼40 and 45% over 28 days in standard PBS.³⁰

The latter result highlights some behavioral differences in the bulk polymer interaction with GMPs and effects of mixing GMPs within a construct, which was not carried out in this study, as GelMA and GelMA/GMP-40 constructs demonstrated temporal differences in polymer loss over 28 days, and polymer loss was shown to increase (i) with the presence of GMPs and (ii) with GMP crosslinking over 28 days. Separately, Lam et al. found that for any given bulk polymer formulation, the swelling ratio of GMP-10 and GMP-40-incorporating gels was statistically similar after 28 days. In both cases, the behavior of gel/GMP composites differs from that observed in this study, where GelMA/GMP-10 scaffolds had significantly higher swelling ratios than GelMA and GelMA/GMP-40 scaffolds after 28 days. Overall, comparison of these collective results and those of further studies not examined here provides valuable information about material and composite swelling properties, which should inform the selection of materials for tissue engineering, particularly in a 3D printing context. In this way, researchers can better tailor their choice of material components to specific intended applications.

Evaluation of the drying and rehydration behavior of hydrogel-based biomaterial scaffolds is critical in their development and use in in vivo research and in clinical applications. As the hydrated scaffold dries, there may be conformational changes in the gel as water is removed and the remaining polymer chains interact, resulting in different final swelling properties than might be observed by swelling without drying. In addition, understanding of rehydration cycles is particularly valuable in this study as the scaffolds were produced by extrusion 3D. As a result, the scaffold may experience changes in the physical orientation of fibers between swelling cycles in addition to the conformational changes noted previously.

Temenoff et al. previously evaluated the swelling behavior of OPF hydrogel disks under 4 drying–rehydration cycles. In constructs that were swollen immediately after fabrication (analogous to this study), the authors demonstrated that the fold swelling ratio for both high and low molecular weight formulations was highest in the first swelling period (14.76 ± 0.16 and 6.37 ± 0.35, respectively, for high and low MW OPF), decreasing significantly in cycles 2–4 and maintaining statistical similarity in those cycles,⁴² highlighting the possibility of conformational changes after drying. By contrast, in this study, there were no significant differences observed in any group as a function of rehydration cycle, with GelMA, GelMA/GMP-10, and GelMA/GMP-40 fold swelling ratios of 7.7 ± 0.8, 7.4 ± 0.7, and 7.2 ± 0.5 after one cycle and 7.8 ± 0.5, 6.2 ± 0.5, and 6.1 ± 0.4 after five cycles, respectively.

Temenoff et al. also observed no significant differences in polymer loss (referred to as sol fraction) across cycles within each group (0.4 ± 0.04 and ∼0.3 for high and low MW after four cycles); this behavior was observed in this study for both GelMA/GMP composites and GelMA scaffolds between cycles 4 and 5 (0.17 ± 0.1, 0.17 ± 0.1, and 0.19 ± 0.1, respectively, for GelMA, GelMA/GMP-10, and GelMA/GMP-40 after five cycles).⁴² Although an analogous group was not considered here, the authors in an earlier study also demonstrated that for scaffolds that were dried immediately after fabrication, the final swelling ratio was significantly reduced compared with scaffolds swollen immediately after fabrication, again highlighting the need to consider processing conditions of biomaterial scaffolds.⁴²

Hutson et al. also evaluated rehydration of dried PEG/GelMA scaffolds, as a ratio of rehydrated weight to original wet weight. The reswelling ratios for their 10% PEG/0% GelMA and 5% PEG/10% GelMA formulation were ∼0.5 and ∼0.75, respectively.¹² Corollary ratios for GelMA, GelMA/GMP-10, and GelMA/GMP-40 scaffolds in this study were 0.9 ± 0.02, 0.9 ± 0.3, and 0.9 ± 0.3, respectively. Both studies indicate the improved ability of GelMA and GelMA/GMP composite constructs to reswell compared with OPF and PEG/GelMA composite constructs. Similar to material considerations for desired swelling behavior as discussed previously, consideration of rehydration results here should also inform material selection in tissue repair applications, as the capacity of a construct for reswelling is critical in the development, handling, and ease of use of tissue repair scaffolds in vivo.

In evaluating the individual swelling behavior of GMP-10 and GMP-40 particles separately from composite scaffolds, GMPs were swollen in DPBS using 40 μm cell strainers for up to 28 days. In this study, no significant differences were observed in fold swelling between GMP-10 and GMP-40 particles at each time point (4.1 ± 0.5 and 4.0 ± 0.3 after 28 days for GMP-10 and GMP-40 particles, respectively). This result was somewhat surprising from previous expectations; it was expected that GMP-10 would demonstrate higher fold swelling than GMP-40 over time. The expected behavior would provide context for why the inclusion of GMP-10 led to significantly higher swelling in the GelMA/GMP-10 group compared with the other groups, and why the GelMA/GMP-10 solution had the highest viscosity of the three solutions in rheological tests.

Previous studies in our research group have suggested that the composite scaffold swelling behavior in this study may be specifically owing to the interaction between the particles and bulk GelMA. For example, Holland et al. demonstrated 24-h fold swelling ratios of 8.3 ± 0.4 and 5.6 ± 1.3 for GMP-10 and GMP-40, respectively (compared with 3.7 ± 0.3 and 3.8 ± 0.3 in this study), but statistically similar fold swelling ratios and polymer loss of OPF, OPF/GMP-10, and OPF/GMP-40 constructs over 28 days.³⁷ In a previous study, the authors observed similar behavior: fold swelling ratios of 12.2 ± 0.9 and 10.5 ± 0.4 for GMP-10 and GMP-40 particles, respectively, but no statistical differences in composite swelling.³⁸ In a later study, Vo et al. demonstrated no effect of GMP-10 incorporation on the fold swelling of poly(N-isopropyl acrylamide) composite gels.³⁹

Although the underlying mechanisms for the results of this study were not specifically investigated in detail, there are a few possible explanations for the observed behaviors. One explanation for the composite swelling results observed could be that although the particles alone maintain statistically similar swelling over 28 days, the different GMP formulations may have different interactions with the bulk polymer to differences in the crosslinking of each component, leading to different swelling behaviors. As described, previous studies have shown differences in swelling between GMP formulations but no resulting differences in composite swelling. In this study, it may instead be the case that the particles individually show similar swelling behavior but have different effects on the composite. The GMP-10 particles would have a lower degree of crosslinking and may uptake additional solution within the bulk gel when swollen compared with either GelMA or GelMA/GMP-40 scaffolds, whereas GMP-40s would be expected to demonstrate lower swelling owing to their higher degree of crosslinking. This would provide context for the 28-day fold swelling and volumetric swelling results.

By contrast, GMP-10 may more easily dispel solution taken up during swelling, whereas GMP-40 may retain any water, as well as possible degradation products, longer, accelerating degradation relative to the other groups and leading to the higher polymer loss observed in this study. Differences between the swelling of composite scaffolds in this study and in the earlier studies discussed previously may also be owing to electrostatic interactions present between GMPs and the bulk GelMA matrix, which would not be present in an OPF-based system. In addition, the incorporation of the particles may inherently create points of reduced crosslinking of the bulk gel because of their physical presence and scattering of the UV light used for crosslinking. Within the GelMA-only solution, the bulk phase exists in one continuous phase, whereas in the GelMA/GMP groups, GMPs are dispersed throughout the bulk phase, possibly disrupting continuity and impacting the degree of crosslinking of the gel.

Because formulation and temporal differences in behavior largely did not occur until day 28, it may be that significant breakdown of the bulk gel does not begin to occur until after day 21, after which the properties of the embedded particles would more strongly drive future swelling behavior; it would be interesting to examine a longer timepoint, such as 35 or 42 days, to confirm this. The proposed mechanism(s) for this behavior may create the opportunity for more detailed study in the future.

We acknowledge several areas in which similar future studies could be improved. Although scaffolds were evaluated under extended swelling conditions in DPBS, it may also be of value to examine their swelling behavior under enzymatic conditions. Although outside the scope of this study, enzymatic conditions would yield insight into swelling behavior of the scaffolds under a stressed or wound healing environment.³⁷ In addition, as swelling and polymer loss calculations were made using scaffold and particle weights, scaffold degradation was not measured directly. It may be valuable to, under standard or enzymatic conditions, measure degradation directly such as through a protein quantification assay.

From the results although it is proposed that the interaction of GMPs and the bulk GelMA hydrogel influenced the swelling of composite scaffolds, morphological changes in the particles embedded within fibers were not evaluated owing to the difficulty of separating these particles from the bulk gel. It may be valuable in a future experiment to compare the swelling of free and embedded particles through image analysis for comparison. Likewise, future studies could more closely investigate the interactions between GMPs and bulk GelMA, to demonstrate the underlying mechanism for the behaviors observed in this study.

Finally, future work might also focus on factorial analyses of different concentrations of both GelMA and GMPs in solution. Although the focus of this study was the specific context of pure GelMA and the inclusion and crosslinking density of GMPs, it would be valuable in a partial factorial study to evaluate the swelling properties of 3DP scaffolds with varying concentrations of bulk GelMA and of each crosslinking density of GMPs. Although other research groups have made some of these determinations for bulk GelMA-only constructs, such a study could inform the selection of a printable biomaterial more tailored to a specific intended application. Similarly, bilayered scaffolds as have been used previously in our research group, with both GMP-10 and GMP-40 in the same construct, would be valuable groups for future study, as these types of construct are attractive for the spatiotemporal delivery of multiple bioactive molecules, yet differences in their swelling behaviors compared with monophasic scaffolds remain unstudied.

Conclusion

This study demonstrates the extended swelling behaviors of GelMA constructs on incorporation of gelatin microcarriers of different crosslinking densities, and likewise the swelling behavior of these constructs under drying–rehydration cycles. Incorporation of GMPs did not affect fold swelling, but did affect polymer loss and volumetric swelling, up to day 21, whereas after 28 days, GMP inclusion led to significantly increased fold and volumetric swelling for GelMA/GMP-10 and decreases in both values for GelMA/GMP-40. Although it was expected that this behavior would be paralleled in particle-only swelling, GMP-10 and GMP-40 particles did not demonstrate significantly different swelling over 28 days. However, it is clear that the interaction of each type of GMP with the bulk hydrogel led to significant changes in composite swelling behavior, which may be owing to the differences in crosslinking between the two GMPs and the bulk hydrogel.

In examining the reswelling capacity of these scaffolds, the number of drying and rehydration cycles did not have a significant effect on the fold swelling of any individual scaffold group over the course of the study. Statistical differences between different formulations did appear after cycle 4, where GelMA scaffolds had increased fold swelling and decreased volumetric swelling compared with both GMP groups. Although in the context of material handling and use in the clinic, it is likely that fewer than four cycles of drying would be required; this behavior is nonetheless an important consideration in material selection.

Overall, the observations from this study demonstrate the differences in swelling behavior of an example hydrogel/microcarrier system under short-term (24 h) and extended (28 day) swelling and under cycles of drying and rehydration. The conclusions from these experiments are valuable in the selection of materials in composite systems for controlled delivery of bioactive molecules in vitro and in vivo in future experiments.

Disclosure Statement

No competing financial interests exist.

Funding Information

Support was provided by the National Institutes of Health (P41 EB023833 and R01 AR068073) in the preparation of this work. S.M.B., H.A.P., and M.M.S. acknowledge support from the National Science Foundation Graduate Research Fellowship Program. M.M.S. also acknowledges support from a Ford Foundation Pre-Doctoral Fellowship. K.J.H. acknowledges support from the National Institute of Dental and Craniofacial Research (F31 DE030333).

Supplementary Material

Supplementary Table S1