Highly Porous Gas-Blown Hydrogels for Direct Cell Encapsulation with High Cell Viability

Abstract

Cell transplant therapies show potential as treatments for a large number of diseases. The encapsulation of cells within hydrogels is often used to mimic the extracellular matrix and protect cells from the body's immune response. However, cell encapsulation can be limited in terms of both scaffold size and cell viability due to poor nutrient and waste transport throughout the bulk of larger volume hydrogels. Strategies to address this issue include creating prevascularized or porous structured materials. For example, cell-laden hydrogels can be formed by porogen leaching or three-dimensional printing, but these techniques involve the use of multiple materials, long preparation times, and/or specialized equipment. Postfabrication cell seeding in porous scaffolds can result in inconsistent cell density throughout scaffold volumes and typically requires a bioreactor to ensure even cell distribution. In this study, we developed a highly cytocompatible direct cell encapsulation method during the rapid fabrication of porous hydrogels. Using sodium bicarbonate and citric acid as blowing agents, we employed photocurable polymers to produce highly porous materials within a matter of minutes. Cells were directly encapsulated within methacrylated poly(vinyl alcohol), poly(ethylene glycol), and gelatin hydrogels at viabilities as high as 93% by controlling solution variables, such as citric acid content, viscosity, pH, and curing time. Cell viability within the resulting porous constructs was high (>80%) over 14 days of analysis with multiple cell types. This work provides a simple, versatile, and tunable method for cell encapsulation within highly porous constructs that can be built upon in future work for the delivery of cell-based therapies.

Impact Statement

This simple method to obtain cell-laden hydrogel foams allows direct cell encapsulation within biomaterials without the need for porogens or microcarriers, while maintaining high cell viability. The successful encapsulation of multiple cell types into gas-blown hydrogels with varied chemistries shows the versatility of this method. While this work focuses on photocrosslinkable polymers, any quick gelling material could be used for foam fabrication in expansion of this work. The potential future impact of this work on the treatment of diseases and injuries that utilize cell therapies is wide-ranging.

Introduction

A major goal of tissue engineering is to successfully mimic the extracellular matrix (ECM) so that cells can grow and regenerate new, healthy tissue. To that end, many advances in the development of biomaterials with properties, such as cell attachment, electroconductivity, and environmental response, have been engineered.^1–4 Due to the large water content and soft nature of natural ECM, hydrogels are often used to recapitulate these environments. Hydrogels use a range of hydrophilic polymers, from completely synthetic materials, like poly(vinyl alcohol) (PVA) and poly(ethylene glycol) (PEG), to natural polymers, such as gelatin and fibrin.^5–9 Often, cell interactions are characterized on flat biomaterial surfaces in vitro. However, cells attached to two-dimensional constructs do not provide a realistic recreation of the natural ECM environment with cells embedded within three-dimensional (3D) tissues.

To enhance complexity, cells can be encapsulated within hydrogels.^10–12 In general, cell-encapsulated hydrogels are limited in size due to poor nutrient and waste transport to the center of bulk (solid) scaffolds.^13,14 Cell-laden hydrogels can be constructed at thicknesses of no greater than 2 mm before issues with diffusion-based nutrient/waste transport are observed.¹⁵ To combat the issue of poor nutrient and waste transport, materials have been engineered to contain either porous or prevascularized structures.^13,14,16,17 These materials can be created using methods such as 3D printing, particulate leaching, and gas foaming.^15,18,19 Use of 3D printing techniques, such as stereolithography, freeform reversible embedding of suspended hydrogels (FRESH), and two-photon polymerization, has been used to create vascular structures in hydrogels.^18,20,21

These methods are often only employed for small construct fabrication, due to the high resolution needed to make vascularized hydrogels where cells are only 100–200 μm away from a fabricated blood vessel.²² To address these limitations, advancements have been made, which allow rapid printing of high-resolution vascular networks using a specialized form of multiphoton printing called holographic 3D printing.²³ While these systems are promising, their high costs (e.g., Holograph-X by CELLINK costs over $1 million) limit their expanded use and overall translation.

By using highly porous hydrogel materials for cell encapsulation, low-cost scaffolds can be rapidly fabricated. Recently, particulate leached foams were developed using cytocompatible macroscopic porogens made from gelatin and PEG.^17,24 By changing the amount and size of the porogen used to fabricate the hydrogels, properties such as porosity and pore size were easily manipulated.¹⁷ The limitation of this method is the additional and tedious fabrication step to make the porogen. Also, particulate leached foams can suffer from poor interconnectivity between pores when fabricated with porogen concentrations <70%, which can limit mechanical property tuning and/or the potential ingrowth of vasculature after transplantation.^25,26

An alternative option is the use of gas-blown foams. Gas blowing has been previously used to create porous biomaterial foams for tissue-engineered scaffolds, traumatic wounds, and aneurysm treatments.^27–31 Gas-blown foams are produced by chemical blowing agents that release gases, such as carbon dioxide (CO₂), and/or the physical addition of gases. Gas bubbles get trapped as a polymer is crosslinked to create a porous scaffold. To use gas blowing for cell encapsulation, all foaming components must be cytocompatible.

Previous work attempted to encapsulate cells during gas blowing by introducing air bubbles into a gelatin methacrylate (GelMA) solution, while ultraviolet (UV) crosslinking. These foams showed improved cell viability over 7 days in comparison to the nonporous hydrogel controls, but a prolonged procedure involving multiple curing and cell addition steps was needed to successfully produce the constructs.¹⁹ Other work was completed by using sodium bicarbonate (NaHCO₃) and citric acid as chemical blowing agents during cell encapsulation. The reaction between citric acid and NaHCO₃ produces CO₂ as a by-product. Improved cell survival was seen in porous PEG dimethacrylate (PEGDMA) hydrogels compared to the nonporous controls at 7 days.³²

The major disadvantage of this method was that 50–70% of cells died during the encapsulation process due to the rapid pH changes, high concentrations of citric acid and NaHCO₃, and the cytotoxic crosslinking method.³² NaHCO₃ is a common component in cell culture media, making it a good candidate to use for gas blowing; however, the authors used NaHCO₃ at concentrations >50 times the amount used in standard culture media (e.g., Dulbecco's Modified Eagle Medium [DMEM]).

This work is based on the hypothesis that by rationally adjusting the solution content, components, and viscosity of the polymer solution, previous gas-blowing systems for direct cell encapsulation could be improved and used with a range of polymer types. We used principles such as viscosity tuning, buffering, and curing time to minimize the concentration of cytotoxic components required to form reproducible porous foams. This work provides a universal method for fabricating quick-gelling hydrogel foams with high viability of encapsulated cells. Using multiple photocurable hydrogel polymers, including PVA methacrylate (PVAMA), PEGDMA, and GelMA, we demonstrate the ability to obtain highly porous materials with increased viability of two encapsulated cell types (3T3 mouse fibroblasts and human mesenchymal stem cells [MSCs], Fig. 1). This adaptable system enables interchanging polymers and foaming materials, such as surfactants, thickeners, and photoinitiators, to provide highly tunable gas-blown foams for direct cell encapsulation with high cytocompatibility.

FIG. 1.

Schematic representation of cell-encapsulated gas-blown hydrogel foam formation. Color images are available online.

Materials and Methods

IC50 values

3T3 mouse fibroblasts (ATCC) were used to assess half maximal inhibitory concentration (IC50) values (concentration at which 50% of cells remain alive after exposure) of citric acid, NaHCO₃, and Irgacure 2959 to be used for further solution optimization. Cells were cultured in DMEM with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (pen-strep) at 37°C with 5% CO₂. Cells were trypsinized and seeded at a density of 5000 cells per well in a 96-well plate. Wells (n = 3) were then exposed to solutions of citric acid, NaHCO₃, or Irgacure 2959, which were serially diluted in DMEM. After a 20-min exposure, solutions were removed, and cell viability was measured compared to an untreated control using a Resazurin assay (Alamar Blue) read by a plate reader. Cell viability was calculated as follows: $C e l l V i a b i l i t y (%) = \frac{A b s_{S a m p l e} - A b s_{B l a n k}}{A b s_{C o n t r o l} - A b s_{B l a n k}} \times 100 %$

The blank was a well with media only and no cells. Cell viability was plotted for each dilution, and IC50 values were determined as the point where the curves showed 50% cell viability.

Solution cytocompatibility

Using IC50 values as a guide for the concentration ceiling, the cytocompatibility of citric acid, NaHCO₃, and Irgacure 2959 was measured with adipose-derived MSCs at concentrations of 0.25, 1, and 0.2% (w/v) in DMEM, respectively. MSCs were seeded in 24-well plates at 2500 cells per well. Each component (n = 3) was added to a well and allowed to incubate for 30 min at 37°C and 5% CO₂. After incubation, cells were treated with a Resazurin assay for 2 h for cytocompatibility measurements or with Live/Dead assay stains and imaging for morphology analysis.

Buffering

Buffered solutions containing phosphate buffered saline (DPBS) or DMEM were supplemented with 1% NaHCO₃ and 50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES). To prepare the solutions, 10 × concentrated DPBS and DMEM (with NaHCO₃) solutions were made. HEPES (1M, pH 7) was added to a subset of solutions to provide 50 mM HEPES after diluting to 1 × . Then, an additional 1% NaHCO₃ was added. The solutions were equilibrated at room temperature for 1 h. Then citric acid was added at 0.25, 0.5, or 0.75%. A DMEM + HEPES control was also tested at 1%. Immediately after adding citric acid, the pH was measured (n = 3) using a pH probe (Thermo Scientific). Solutions were incubated at 37°C and 5% CO2 for 1 h, and then pH was retested.

Buffered solution cytocompatibility

Solutions containing DMEM or DPBS were supplemented with 50 mM HEPES, 1% NaHCO₃, and 0.2% Irgacure 2959. Then solutions were filtered through a 0.20 μm syringe filter. MSCs were seeded into 24-well plates at 2500 cells per well the night before testing. Buffered solutions were added (500 μL) to each well. Citric acid was then added to the wells at 0.25, 0.5, or 0.75% (n = 3). The plate was placed into a UV curing box (Analytik Jena CL-1000) with a wavelength of 365 nm for 5 min. The plate was removed and incubated at 37°C and 5% CO₂ for 20 additional minutes.

The samples were then stained with a Resazurin assay for cell viability measurement or with a Live/Dead assay for images and morphology assessments. After testing, fresh DMEM supplemented with 10% FBS and 1% pen-strep was added to each well and incubated for an additional 24 h. Live/dead and Resazurin assays were completed again to evaluate cell recovery.

Polymer synthesis

All chemicals were purchased from Fisher Scientific and used as received, unless otherwise stated. PVA was methacrylated using 2-isocyanatoethyl methacrylate (ICEMA) in DMSO (Fig. 2). Briefly, vacuum-dried PVA (25 kDa, 98% hydrolyzed) was dissolved in DMSO at 96°C to make a 10% solution under nitrogen. The solution was cooled to 60°C, and ICEMA was added dropwise at a 4% molar ratio of isocyanates to PVA hydroxyls. The reaction proceeded for 3 h at 60°C, and the product was precipitated in cold acetone, dried, redissolved in water, and then reprecipitated in cold acetone. The final product was vacuum dried for 24 h. Nuclear magnetic resonance (NMR) spectroscopy confirmed methacrylation of PVA by comparing the PVA backbone (δH 3.7–4.1 and 1.4–1.8 p.p.m.) and the methacrylate vinyl protons (δH 5.6 and 6.1 p.p.m). PVAMA had a degree of substitution of backbone pendant hydroxyls of 3%.

FIG. 2.

Methods for synthesis of PVAMA, PEGDMA, and GelMA used for hydrogel foams. GelMA, gelatin methacrylate; PEGDMA, poly(ethylene glycol) dimethacrylate; PVAMA, poly(vinyl alcohol) methacrylate. Color images are available online.

PEGDMA was synthesized using methacrylic anhydride (MA) with triethylamine (TEA) as a catalyst (Fig. 2). PEG (4 kDa) was dissolved in dichloromethane at 33 wt% with a small amount of hydroquinone. MA and TEA were added at a molar ratio of 2.2 MA: 2 TEA: 1 PEG. The reaction proceeded at room temperature under nitrogen for 3 days. The polymer was then precipitated in cold ether and vacuum dried. Methacrylation was characterized using NMR spectroscopy by comparing the adjacent proton to the methacrylate group (δH 4.3 p.p.m.) and the methacrylate vinyl protons (δH 5.6 and 6.1 p.p.m). PEGDMA had a degree of substitution of terminal hydroxyls of 97%.

GelMA was synthesized using MA (Fig. 2). Gelatin was dissolved in DMSO at 60°C. Methacrylic anhydride was then added to the solution at 0.1 g/g of gelatin. The reaction proceeded for 3 h at 60°C. The polymer was precipitated in cold ethanol, redissolved in water, and reprecipitated in ethanol. The final product was vacuum dried and characterized by Fourier transform infrared spectroscopy.

Rheology

Rheological measurements were performed using a TA Instruments Discovery Rheometer with a 40 mm plate geometry. Viscosity measurements were taken with a goal of matching viscosities of PVA-0L (Table 1), which produced highly porous foams. Viscosities of PEGDMA and GelMA hydrogel solutions were tuned with thickeners of poly(ethylene oxide) (PEO; 400,000 Da) and cornstarch, respectively. Tested samples included PVA-0L, PEG-0, PEG-3, and GelM-10 (Table 1). Solutions were conditioned at 37°C for 120 s. The samples were then tested using a flow sweep at shear rates between 1–100 radians per second.

Table 1.

Synthesized Hydrogel Formulation Names and Solution Compositions

Name	Polymer	Surfactant	Initiator	Thickener	NaHCO₃ (% w/v)	HEPES (mM)	Citric acid (% w/v)	Media type
PVA-0	20% PVAMA	2% Gel	I2959 (0.2%)	N/A	1	50	0.25, 0.5, 0.75	DMEM, DPBS
PVA-1	20% PVAMA	2% Gel	I2959 (0.2%)	1% PVA 108k	1	50	0.25, 0.5, 0.75	DMEM, DPBS
PVA-2	20% PVAMA	2% Gel	I2959 (0.2%)	2% PVA 108k	1	50	0.25, 0.5, 0.75	DMEM, DPBS
PVA-3	20% PVAMA	2% Gel	I2959 (0.2%)	5% PVA 108k	1	50	0.25, 0.5, 0.75	DMEM, DPBS
PVA-0L	20% PVAMA	2% Gel	LAP (0.05%)	N/A	1	50	0.25, 0.5, 0.75	DMEM
PEG-0	20% PEGDMA	2% F-127	LAP (0.05%)	N/A	1	50	N/A	DMEM
PEG-3	20% PEGDMA	2% F-127	LAP (0.05%)	3% PEG 400K	1	50	0.25, 0.5, 0.75	DMEM
Gel-10	10% GelMA	N/A	LAP (0.05%)	10% Starch	1	50	0.25, 0.5, 0.75	DMEM

DMEM, Dulbecco's modified Eagle's medium; DPBS, phosphate buffered saline; Gel, gelatin; I2959, Irgacure 2959; LAP, lithium phenyl (2,4,6-trimethylbenzoyl) phosphinate; MA, methacrylate; PEG, poly(ethylene glycol); PVA, poly(vinyl alcohol).

Hydrogel foam fabrication

Hydrogel solutions were prepared by dissolving 1 g of PVAMA and 0.1 g of gelatin (surfactant) with or without PVA (108 k) thickeners in a solution containing 4.25 mL of DI water and 250 μL of 1M HEPES at 96°C. The solution was cooled to below 50°C, and 500 μL of 10 × DMEM or 10 × DPBS was added. NaHCO₃ and Irgacure 2959 were then added so that the final concentrations were 20% PVAMA, 2% gelatin, 1% NaHCO₃, 0.2% Irgacure 2959, and 50 mM HEPES. Thickeners were added at 0, 1, 2, and 5% (Table 1).

The effect of the photoinitiator on pore morphology was also characterized by replacing Irgacure 2959 with lithium phenyl (2,4,6-trimethylbenzoyl) phosphinate (LAP) at a concentration of 0.05% (w/v). Only hydrogels containing DMEM were examined for use with LAP (Table 1). PEGDMA and GelMA hydrogels were formulated using the same method, using surfactant and thickeners shown in Table 1. PEGDMA hydrogel solutions were all dissolved at room temperature, while GelMA was dissolved at 90°C.

To form hydrogels, 1 mL of solution was placed into a 24-well plate. Citric acid was then added from a 5% stock solution to a final concentration of 0.25, 0.5, or 0.75% (50, 100, or 150 μL, respectively). The plate was then vigorously shaken on a vortex mixer for 20 s and quickly transferred into a UV curing box. Solutions were irradiated for 5 min (Irgacure 2959) or 2 min (LAP). Samples were air dried, cut down the center, sputter coated with gold, and imaged using a scanning electron microscope (SEM, Jeol NeoScope JCM-5000) at 35 × magnification and 10 kV to qualitatively examine pore structures. In samples with visible porosity, pore size was examined by measuring 15 random pores per image (n = 30 pores) using the measurement tool in ImageJ.

Cell encapsulation

PVA-0L, PEG-3, and Gel-10 solutions were tested with 3T3s for cell encapsulation. PVA-0L was also used to encapsulate MSCs. Polymer solutions were sterile filtered, and then cells were added at 5 × 10⁴ cells/mL. Polymer-cell suspensions (500 μL) were placed into a 24-well plate (n = 3). Filter-sterilized 5% citric acid (75 μL) was added to each well, and the plate was vortexed for 30 s and then placed under UV light for 2 min. Upon completion of crosslinking, 1 mL of cell culture media was placed in each foam containing well for 20 min.

The media were then changed, and cells were allowed to incubate at 37°C for 24 h before Live/Dead staining. Cells were encapsulated in control samples without citric acid added to form nonporous hydrogels. At selected time points, samples were cut into three lateral cross-sections and placed in Live/Dead assay stain solution for 30 min. The samples were washed with PBS and then imaged. Cell viability was calculated as follows: $C e l l V i a b i l i t y (%) = \frac{L i v e C e l l s}{T o t a l C e l l s} \times 100 %$

Cell distribution was assessed by comparing cell numbers in images of a lateral image of the hydrogel upper portion and in a vertical cross-section of the lower portion of the foam.

Pore analysis

In samples with visible porosity, pore size was analyzed using the measurement tool in ImageJ on SEM images (n = 6). A horizontal line was drawn across the center of each image, and the diameters of all pores touching that line were quantified. Percent interconnectedness was estimated as previously described.²⁸ GNU Image Manipulation Program was used to measure the pixel area of interconnects (taken as portions of the image where openings between pores can be seen) compared to total foam pixel area in the image. Interconnectivity was calculated as follows: $I n t e r c o n n e c t i v i t y = \frac{p i x e l s_{i n t e r c o n n e c t s}}{p i x e l s_{i m a g e} - p i x e l s_{o u t s i d e f o a m}} \times 100 %$

Porosity was measured for PVA-0l foams with 0.75% CA, which were used in the 14-day cell encapsulation studies. The density of solid hydrogels (D_s) and foams (D_f) was measured using mass and volume measurements (n = 3). Porosity was determined as follows: $P o r o s i t y (%) = \frac{D_{s} - D_{f}}{D_{s}} \times 100 %$

Statistical analysis

Measurements are presented as mean ± standard deviation. Single factor analysis of variance with Tukey's post hoc was used determine significance between groups. A p < 0.05 was taken as statistically significant.

Experimental Results

IC50 values and solution cytocompatibility

The IC50 values for the critical foam blowing components, citric acid, Irgacure 2959, and NaHCO₃, were 0.23, 0.80, and 1.5%, respectively (Table 2). Using this data as a guide, the cytocompatibility of the critical blowing components (citric acid, Irgacure 2959, and NaHCO₃) at concentrations of 0.25, 0.2, and 1% (w/v, respectively) was examined with MSCs (Fig. 3). Both NaHCO₃ and Irgacure 2959 showed cytocompatibility above the 75% benchmark with values of 98 and 99% viability after 1 h of exposure and 95% and 99% after the 24-h recovery, respectively. Citric acid exposure resulted in reduced cell viability (51% and 21%) after both the 1-h treatment period and the 24-h recovery period, respectively.

FIG. 3.

Mesenchymal stem cell viability after 1 h of exposure to foam blowing components (1 wt% NaHCO₃, 0.2 wt% Irgacure 2959, and 0.25 wt% citric acid) and after 24-h recovery, measured using a Resazurin assay. Horizontal line = 75% viability, based on ISO 10993-5 standards for cytocompatibility. N = 3, mean ± standard deviation displayed. NaHCO₃, sodium bicarbonate. Color images are available online.

Table 2.

Half Maximal Inhibitory Concentration Values for Foaming Components Based on Testing with 3T3 Mouse Fibroblasts

Component	IC50 (% w/v)
Citric acid	0.23
Irgacure	0.80
NaHCO₃	1.5

NaHCO₃, sodium bicarbonate.

Buffering capacity

The DPBS solution containing 1% NaHCO₃ showed pH values of 8.33, 8.16, 8.03, and 7.41 after 1 h of treatment with 0, 0.25, 0.5, and 0.75% citric acid (Fig. 4A). DMEM with 1% NaHCO₃ had pH values of 7.99, 8.00, 7.90, and 5.64 after 1 h of incubation with increasing concentrations of citric acid (Fig. 4B). When supplemented with HEPES, the pH of DPBS solutions at 1 h (7.87, 7.62, 7.17, and 6.59) was lower than the corollary nonbuffered DPBS solutions (Fig. 4C). Similar trends were seen in HEPES-buffered DMEM solutions, except that pH was increased in HEPES-containing DMEM with 0.75% citric acid. The pH values of DMEM + HEPES solutions were 7.83, 7.79, 7.57, 7.35, and 7.05 (Fig. 4D). In general, DPBS solutions had higher pH compared to corollary DMEM solutions with and without HEPES.

FIG. 4.

Measured pH of solutions containing 1% NaHCO₃ in DPBS or DMEM without (A, B) or with (C, D) 50 mM HEPES supplementation. Measurements were taken directly after treatment with 0, 0.25, 0.5, or 0.75% citric acid and again after 1 h of equilibration. N = 3, mean ± standard deviation displayed. DMEM, Dulbecco's modified Eagle's medium; DPBS, phosphate buffered saline; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid. Color images are available online.

HEPES solution cell viability

All tested solutions had MSC viability above 75% before citric acid treatment initially (20 min) and after 24 h of recovery (Fig. 5A). DPBS solutions supplemented with citric acid showed cell viability <75%, with further decreases in viability at 24 h. DMEM solutions had high cell viability (>100%) after initial treatment with all concentrations of citric acid and >75% viability after the 24-h recovery, with 0.75% citric acid having the lowest MSC viability at 78%.

FIG. 5.

Mesenchymal stem cell viability after exposure to buffer solutions containing 1% NaHCO₃ and 50 mM HEPES treated with 0, 0.25, 0.5, and 0.75% citric at 20 min and after a 24-h recovery period. (A) Quantitative cell viability assessed by Resazurin assay and (B) visual assessment of cell morphology after straining with Live/Dead assay. N = 3, mean ± standard deviation displayed. White scale bar = 500 μm and applies to all images. *p < 0.05 relative to corollary DPBS solution. Color images are available online.

Cells treated with DPBS solutions underwent morphological changes, even when viability measurements were high (0% citric acid) (Fig. 5B). In general, cell rounding increased with increased citric acid. Fewer morphological changes were observed in DMEM-based solutions. MSCs appeared smaller in all testing solutions when compared to the DMEM control directly after exposure. Morphology returned to normal after the 24-h recovery in both DMEM and DPBS solutions, although DPBS wells had qualitatively lower cell density than DMEM samples after recovery.

Irgacure 2959-initiated hydrogel foam morphology

PVAMA-based hydrogels showed improved porous structure with increasing concentrations of both high molecular weight PVA (thickener) and citric acid. Hydrogels formed with DPBS were able to form pores without the use of a thickener (PVA-0; Fig. 6A). The high molecular weight PVA was added as a noncrosslinkable thickener that does not chemically incorporate into the PVA network. By adding a thickener, the hydrogel solution viscosity can be tuned independent of network chemistry to enable control over pore formation during crosslinking. Addition of a thickener (high molecular weight PVA) did not significantly affect overall pore size.

FIG. 6.

Scanning electron micrographs of (A) DPBS-based and (B) DMEM-based PVAMA foams with varied citric acid concentrations and varied concentrations of high molecular weight PVA (108 kDa) added as a thickener (PVA-0: 0%, PVA-1: 1%, PVA-2: 2%, and PVA-5: 5%). Scale bar = 500 μm and applies to all images. Text overlays include average pore sizes and approximate interconnectivities based on foam image analysis.

However, interconnectivity was increased with low concentrations of thickener (1%) and reduced with higher concentrations of thickener (2% and 5%) relative to the PVA-0 control. DMEM-based PVA solutions crosslinked with Irgacure were unable to form porous hydrogels without added thickener (images not included). While general trends of improved pores with increased thickener and citric acid were maintained in foams formed in DMEM, these scaffolds showed generally larger, less uniform pores; lower qualitative porosity; and reduced interconnectivity than corollary foams fabricated in DPBS (Fig. 6B).

It should also be noted that a nonporous “skin” layer is commonly formed on gas-blown foams, which results in spatial variations from the edge of samples into the centers. Ideally, a foam surface would have open pores when encapsulating cells to ensure adequate nutrient and waste transport. Qualitatively, this nonporous layer thickness is reduced as citric acid and thickener concentration is increased. Foams formed in DMEM had thicker visible skin layers.

LAP-initiated hydrogel foam morphology

Based on its faster initiation rates, LAP was used in place of Irgacure 2959 to study the ability to use these foaming methods with multiple polymer types. First viscosity was measured, based on the observations that the addition of noncrosslinkable high molecular weight thickeners (specified in Table 1 for each formulation) qualitatively affected both viscosity and foaming (better pore structures with increased thickener) in initial studies (Fig. 7A). At higher shear rates (100 radians/s), PEG-0 solutions (yellow line) showed the lowest viscosity of 0.05 Pa·s followed by PVA-0L (blue line) with a viscosity of 0.27 Pa·s. When thickeners were added to GelMA-10 (gray line) and PEG-3 (orange line) solutions, the viscosity was increased to 0.58 and 0.54 Pa·s, respectively, at 100 radians/s. PEG-0 exhibited the most shear thinning, with reduced viscosity as shear rate increased.

FIG. 7.

(A) Viscosity profiles of hydrogel solutions and (B) corresponding scanning electron micrographs of hydrogels based on PVA-0L, PEG-3, and GelMA-10 with 0.25, 0.5, and 0.75% citric acid. Scale bar of 500 μm applies to all images. Text overlays include average pore sizes and approximate interconnectivities based on foam image analysis. Color images are available online.

These solutions were utilized to form hydrogel foams with the addition of varying concentrations of citric acid (Fig. 7B). PEG-0 hydrogels were unable to form foams and were not imaged. Low porosity was observed with 0.25% citric acid, and increased porosity was qualitatively seen with increased citric acid content. The PVA-0L hydrogels showed highly porous structures with both 0.5% and 0.75% citric acid, and more uniform and interconnected pore structures were formed in comparison with those observed in the Irgacure-based PVA hydrogels shown in Figure 6.

In addition, PVA-0L foams did not have a visible nonporous skin layer. PEG-3 and GelMA-10 hydrogels also showed qualitatively increased porosity with increased citric acid, with more regular pores observed in PEG-3 hydrogels. A nonporous skin was observed on PEG-3 hydrogels with 0.5% citric acid, but the PEG-3 hydrogels with 0.75% citric acid were fully porous. The GelMA-10 hydrogels were less regular in overall appearance, but did not exhibit large sections of nonporous skin. In general, pore size and interconnects between pores increased with increased citric acid content in all three hydrogel systems. The PVA system provided the largest, most interconnected pores, PEG hydrogel foams had the smallest pores, and GelMA foams had the lowest interconnectivity.

Cell encapsulation

3T3 fibroblast encapsulation was initially assessed with PVA-0L, PEG-3, and GelMA-10 hydrogels. Cells encapsulated inside PVA-0L with 75% citric acid showed low cell viability (28 ± 9%; Fig. 8B). Adjustment of initial pH to 7.2 by adding 1M HCl to PVA0-L hydrogel solutions significantly improved cytocompatibility to 89 ± 10%. Encapsulated 3T3s had high cytocompatibility in PEG3 and GelMA-10 hydrogels formed with 0.75% citric acid (85%).

FIG. 8.

(A) Schematic representation of hydrogel sample preparation for Live/Dead imaging. (B) Live/Dead images of 3T3 fibroblasts encapsulated in hydrogels with quantified viabilities. n = 3, mean ± standard deviation displayed. (C) SEM image of 0.75% CA PVA-0L hydrogel adjusted to pH 7.2 with corresponding pore size, interconnectivity, and porosity. (D) 3T3 fibroblast and MSC viability after encapsulation in PVA-0L nonporous and porous foam hydrogels (N = 9: three samples per formulation, three images per sample) measured using counts from images. (E) Representative images after Live/Dead staining of center cross-sections over 14 days. Scale bar = 100 μm. *p < 0.05 between nonporous (solid) and porous (foam) hydrogels. MSC, mesenchymal stem cell. Color images are available online.

Using this information on the importance of initial pH, 3T3 and MSC viability were further assessed over up to 14 days in PVA-0L solid and porous hydrogels. First, it was confirmed that cell distribution was consistent between the top horizontal cross-section and the lower vertical cross-section of foams, with cell density values of 9.6 ± 2.1 and 9.0 ± 0.3 cell/mm, respectively.

There was no significant difference in viability between 3T3s or MSCs encapsulated in nonporous gels or porous foams up to day 3 (Fig. 8D, E). On day 7, foams showed higher 3T3 viability in comparison to solid, nonporous hydrogels, while MSC viability was comparable between the non-porous and porous hydrogels. Further declines in 3T3 viability were seen in nonporous gels at 14 days, with high viability maintained in porous foams. MSCs showed no significant difference in cytocompatibility between the nonporous solid and porous foam hydrogels at any time point.

Discussion

Since 50–70% of encapsulated cells were not viable in previous work to obtain gas-blown hydrogel foams,³² obtaining IC50 values of critical blowing components (citric acid, Irgacure 2959, and NaHCO₃) and using that information to improve viability were important to increase viability of encapsulated cells. The calculated IC50 values for citric acid and NaHCO₃ were 5–10 × lower than the concentrations used in previous work.³² The IC50 value of Irgacure 2959 was 0.8%, which is well above concentrations that are typically used for cell encapsulation in photocurable polymers.^32–34

Based on these studies, 1 wt% NaHCO₃, 0.2 wt% Irgacure 2959, and 0.25 wt% citric acid were selected as initial starting values for foam development. Citric acid was used at a concentration that is slightly higher than the IC50 value of 0.23 wt%, as it is the component that is required for the production of CO₂ to form porous foams during crosslinking.

While 3T3s are good model cells for initial characterization, MSCs were employed to assess cytotoxicity with a therapeutic cell type. Both Irgacure 2959 and NaHCO₃ had MSC cytocompatibility above the benchmark of 75% set by ISO 10993-5.³⁵ Citric acid was the only major blowing component with observed toxic effects on MSCs, which was expected based on the initially measured IC50 values.

Citric acid has previously shown to limit cell growth and attachment, but most prior studies do not adjust the pH when introducing citric acid.^36–38 To that end, we examined the effect of gaining tighter control over the solution pH by buffering. Buffers commonly used in cell culture are NaHCO₃ and HEPES. The current foaming solution already employs close to cytotoxic concentrations of NaHCO₃, so HEPES, a zwitterion, was used to supplement the blowing solutions and improve pH control. We measured the pH of foam solutions in DPBS and DMEM containing 1% NaHCO₃ with or without 50 mM HEPES to test this hypothesis.

In general, DPBS-based solutions containing only 1% NaHCO₃ had a higher pH than corollary solutions in DMEM. Immediately after the addition of citric acid, the solution pH decreased, and then it gradually increased. Higher concentrations of citric acid generally reduced pH values. HEPES-supplemented solutions had smaller differences between initial and equilibrated pH values in comparison to NaHCO₃ alone, except for 0.75% citric acid solution in DMEM. HEPES supplementation increased the initial pH in comparison to NaHCO₃ alone, except for 0.75% citric acid in DPBS. pH values above 7.8 or below 6.8 can cause cell necrosis and apoptosis.³⁹ Solutions supplemented with only NaHCO₃ without the addition of citric acid had pH values outside of this range (7.99: DMEM and 8.33: DPBS), demonstrating that the addition of citric acid is important for neutralizing the added NaHCO₃.

As an extra step, we measured pH values of DMEM solutions containing HEPES with 1% citric acid. This solution had an initial pH of 6.07, which is lower than the ideal range for cell culture. However, the equilibrated 1-h pH was increased to a physiologically relevant level of 7.05. In general, this study shows that HEPES increases the buffering capacity of solutions to make them more resistant to drastic pH changes, which could increase cell viability, while allowing the use of higher concentrations of citric acid to improve bubble formation and gas blowing.

Once we understood how to control foaming solution pH, we characterized cytocompatibility of buffered foam solutions with MSCs. We chose a 20-min incubation as an estimate for the maximum amount of time that cells would be stored in an unreacted hydrogel solution to examine practical working times that would be needed for clinical applications. The 24-h time point provides information on potential longer term exposure effects. Improved MSC viability was observed in the presence of 0.25% citric acid in buffered solutions compared with initial IC50 value testing with 3T3s.

At 24 h, DPBS solutions showed 2 × higher MSC viability (vs. 3T3 viability), but viability was still below the 75% benchmark. DMEM solutions had cell viability >75% for citric acid concentrations up to 0.75%. Initial viability after the addition of citric acid was consistently higher than the 24-h viability, which is attributed to citric acid's role in the Krebs cycle; namely, the Resazurin assay used to measure cell viability is a metabolic assay, and cell metabolism is increased in the presence of citric acid.⁴⁰

To qualitatively confirm initial cell viabilities and evaluate effects of foaming solutions on cell morphology, Live/Dead images were taken at 20 min and 24 h after treatment. After 20 min, all solutions with or without added citric acid showed cell shrinking. This shrinking is attributed to the hypertonic concentration of NaHCO₃, causing cells to lose water to equilibrate the surrounding solution. Cells exposed to DPBS-based solutions showed increased cell rounding, detachment, and death compared to those in corollary DMEM solutions. After 24 h of recovery, cells regained size compared to the control solution. Overall, using DMEM-based foam solutions resulted in better cellular characteristics for use in the development of gas-blown foams with encapsulated cells.

Using this information, the effects of polymer type, citric acid content, and viscosity on foam porosities were characterized to determine whether porous foams could be fabricated using cytocompatible foaming solutions. Initial attempts to fabricate foams using the PVA-0 formulation only showed successful pore formation at a concentration of 0.75% citric acid in DPBS. No foam was formed using DMEM and the PVA-0 formulation. Due to the poor cell viability of DPBS solution with 0.75% citric acid, the solution's rheological properties were tuned to improve foaming outcomes. Thickeners are often used to increase the viscosity of gas-blown foam solutions and tune porosity.^41,42 Thus, we hypothesized that adding thickeners in the form of high molecular weight polymers could be used to achieve higher porosities at lower concentrations of citric acid.

The PVA-1 formulation with 1% PVA (108 kDa) included as a noncrosslinkable thickener to tune hydrogel solution viscosity formed porous structures at a concentration of 0.5% citric acid in both DMEM- and DPBS-based solutions. In PVA-2 and PVA-5 formulations with 2% and 5% PVA (108 kDa) added as a thickener in DPBS, porous structures begin to form at even lower concentrations of citric acid (0.25%). DMEM solutions with 0.25% citric acid showed limited pore formation, which is attributed to the higher buffering capacity of DMEM. Namely, at a citric acid concentration of 0.25%, the solutions do not reach the acidic pH required for the release of CO₂ by NaHCO₃.^32,43

While adding noncrosslinkable high molecular weight PVA thickeners to increase solution viscosity improved pores, visible pore content in DMEM-based foams was inferior to that in DPBS foams. The pore morphology of 0.75% citric acid foams in DMEM show more elongated and larger pores, which could potentially be caused by slow gelling or crosslinking.¹¹ We hypothesize that hydrogel crosslinking may be slower in DMEM than that in DPBS due to interactions with DMEM components, such as antioxidants like phenol red. While further studies can be done to characterize hydrogel crosslinking over time in DMEM versus DPBS, the speed of gelling could be increased to better trap CO₂ air bubbles in DMEM-based foams by using a more efficient photoinitiator.

Many studies have shown LAP to have more efficient initiation compared to Irgacure 2959, even at lower concentrations.^44,45 Another benefit of using LAP is that it can be initiated using more cell-friendly visible light, which could be harnessed in future work to eliminate the need for UV light.^44,46,47 Due to the more efficient formation of free radicals by LAP, extended curing times can increase cytotoxicity; thus, irradiation times were decreased to 2 min in LAP-initiated hydrogels based on previous work showing efficient crosslinking within this time frame.⁴⁴

Substitution of Irgacure 2959 with LAP resulted in drastic improvements to pore structure, even at lower initiator concentrations and shorter curing times. The PVA-0L foam formed with LAP was a highly porous scaffold with citric acid concentrations of 0.5% and 0.75%, without the need for thickeners. This result shows the importance of balancing both gelling and blowing processes to achieve ideal foam properties. These hydrogels also showed interconnected pore structures, which is ideal for implantable hydrogels to allow tissue and vascular ingrowth.⁴⁸ Future studies will focus on how these pore structures affect mechanical properties of synthesized foams and how properties such as stiffness and swelling could be tuned based on polymer chemistry.

While this system was initially designed using PVAMA, the concepts are translatable to other polymers. To demonstrate the adaptability of this approach, PEGDMA and GelMA were employed to fabricate porous hydrogels. Initially, the PEG-0 formulation was utilized, but this composition did not yield a porous foam. Thus, we aimed to increase the viscosity of the PEGDA solution to values that are similar to or slightly higher compared with PVA-0L. Using the PEG-3 solution with 3% PEO (400 kDa) added as a noncrosslinkable thickener, we increased viscosity to obtain porous hydrogels. The same principle was applied to the GelMA-10 formulation, wherein a natural, biodegradable thickener, cornstarch, was added at 10% to provide highly porous hydrogel foams.

Utilizing a completely natural polymer system for cell encapsulation could allow cellular remodeling of the construct during healing. It should be noted that the addition of thickeners has the potential to affect mechanical properties,^49,50 but cornstarch could be quickly removed by treatment with the cytocompatible enzyme, amylase. While all three polymers successfully formed foams, pore morphology differed between the materials. Other factors, such as surfactant type, concentration, and polymer interactions can impact foaming.⁵¹ Future studies can examine how adjusting these parameters can be used to optimize foam structures for selected polymers and applications and how the resulting scaffold architectures affect mechanical properties.

Initially, cell encapsulation was completed using 3T3 cells as a proof of concept for the adaptability of this system. Encapsulated 3T3s had low viability in the PVA-0L hydrogel foams, but PEG-3 and GelMA-10 showed low toxicity 24 h after encapsulation. The PEG-3 and GelMA-10 solutions had initial pH of 7.2, while PVA-0L had an initial pH of 7.8. Thus, pH of the PVA-0L solution was adjusted to 7.2, and 3T3 encapsulation was repeated to result in significantly increased cell viability. Further study on the specific pH range required to achieve high cell viability is needed in future work, but a pH of 7.2 was utilized in the remaining studies.

Both 3T3s and MSCs were then encapsulated in nonporous hydrogel and porous hydrogel foams (PVA-0L) and characterized over 2 weeks to assess long-term viability. Both cell types had similarly high viabilities after 24 h, and at 3 days of culture, foams and solid gels showed no statistical difference in cell viability.

However, the continued culture of 3T3s for 14 days showed a large decline in viability in the nonporous hydrogels, while foams maintained cytocompatibility over 80%. This result confirms previous work stating that encapsulation within hydrogels without adequate channels for nutrient waste transport decreases viability over extended culture times.^13,14,17,32 MSCs showed no significant difference in cell viability in foam and solid gels, but were successfully encapsulated within a PVA-0L with a viability of 89%. It should be noted that the same volume of hydrogel solution was used for nonporous solid hydrogels and porous foam hydrogel synthesis, resulting in a thinner solid scaffold relative to corollary foams. We hypothesize that differences between cell viabilities over time would be more pronounced in a thicker nonporous hydrogel.

This work focused on the development of a versatile platform system for rapid encapsulation of cells within porous scaffolds with high cell viability. Beyond cytocompatibility, the metabolic and proliferation activity of cells are also important. The sustained viability of encapsulated cells in this study provides a foundation for additional work on more detailed cellular investigations. Based on factors such as material selection, scaffold stiffness, cell adhesive factors, and material biodegradation, future studies can be completed to quantify and improve cell metabolism and proliferation, signaling, and/or differentiation within porous biomaterial scaffolds.

Conclusions

This simple method to obtain cell-laden hydrogel foams allows for direct encapsulation within biomaterials without the need for porogens or microcarriers, while maintaining high cell viability. The design of this cytocompatible gas foam blowing system involves four major components: cytocompatible blowing agents, buffering agents, thickeners, and quick gelling polymers. Improving the cell viability of biological blowing components at higher concentrations by buffering increases the amount of gas that can be produced (and therefore the pores that can be formed) before displaying cytotoxic effects. By controlling the viscosity of these solutions with added thickeners, gas trapping increased at lower concentrations of blowing components, allowing for potential use of these methods with more sensitive cell types in future work. Finally, adjusting the gelling rate of polymers improves the porosity and pore morphology.

The successful encapsulation of multiple cell types into these gas-blown hydrogels shows the potential versatility of this method. This system could be used in a range of applications where cell delivery is required. Furthermore, expanding to other therapeutic cell types, such as pancreatic islets for the treatment of type 1 diabetes or vascular endothelial cells for the study of scaffold vascularization, could provide new approaches to improving their long-term viability and efficacy.

While this work focuses on the use of photocrosslinkable polymers, any quick gelling method could be used for foam fabrication in the future expansion of this work, including thermoresponsive materials, Schiff's base, and ionic crosslinking methods. Due to the high tunability of this process, components like thickeners could be swapped for polymers with desired properties in future work. The potential future impact of this method on the treatment of diseases and injuries that utilize cell therapies is wide-ranging.

Footnotes

Authors' Contributions

H.T.B.: conceptualization, methodology, validation, formal analysis, Investigation, data curation, writing—original draft, writing—review and editing, and visualization; and M.B.B.M.: conceptualization, validation, formal analysis, resources, writing—review and editing, visualization, supervision, project administration, and funding acquisition.

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by funding from the Crohn's and Colitis Foundation Award #823117.

References

Navaei

, Saini

, Christenson

, et al. Gold nanorod-incorporated gelatin-based conductive hydrogels for engineering cardiac tissue constructs. Acta Biomater, 2016; 41:133–146; doi: 10.1016/J.ACTBIO.2016.05.027

Park

, Choi

, Kim

, et al. Micropatterned conductive hydrogels as multifunctional muscle-mimicking biomaterials: Graphene-incorporated hydrogels directly patterned with femtosecond laser ablation. Acta Biomater, 2019; 97:141–153; doi: 10.1016/J.ACTBIO.2019.07.044

, Fu

, Wu

, et al. pH-sensitive peptide hydrogel for glucose-responsive insulin delivery. Acta Biomater, 2017; 51:294–303; doi: 10.1016/J.ACTBIO.2017.01.016

Nuttelman

, Mortisen

, Henry

, et al. Attachment of fibronectin to poly(vinyl alcohol) hydrogels promotes NIH3T3 cell adhesion, proliferation, and migration. J Biomed Mater Res A, 2001; 57:217–223; doi: 10.1002/1097-4636

Beaman

, Howes

, Ganesh

, et al. Shape memory polymer hydrogels with cell-responsive degradation mechanisms for Crohn's fistula closure. J Biomed Mater Res A, 2022; 110(7):1329–1340; doi: 10.1002/JBM.A.37376

Browning

, Cosgriff-Hernandez

. Development of a biostable replacement for PEGDA hydrogels. Biomacromolecules, 2012; 13(3):779–786; doi: 10.1021/bm201707z

Lih

, Lee

, Park

, et al. Rapidly curable chitosan-PEG hydrogels as tissue adhesives for hemostasis and wound healing. Acta Biomater, 2012; 8(9):3261–3269; doi: 10.1016/j.actbio.2012.05.001

Burmeister

, Roy

, Becerra

, et al. In Situ delivery of Fibrin-based hydrogels prevents contraction and reduces inflammation. J Burn Care Res, 2017; 39(1):40–53; doi: 10.1097/BCR.0000000000000576

Zhan

. Effect of matrix stiffness and adhesion ligand density on chondrogenic differentiation of mesenchymal stem cells. J Biomed Mater Res A, 2020; 108(3):675–683; doi: 10.1002/jbm.a.36847

10.

Kar

, Vernon Shih

Y-R

, Velez

, et al. Poly(ethylene glycol) hydrogels with cell cleavable groups for autonomous cell delivery. Biomaterials, 2016; 77:186–197; doi: 10.1016/j.biomaterials.2015.11.018

11.

Choh

S-Y

, Cross

, Wang

. Facile synthesis and characterization of disulfide-cross-linked hyaluronic acid hydrogels for protein delivery and cell encapsulation. Biomacromolecules, 2011; 12(4):1126–1136; doi: 10.1021/BM101451K

12.

Bryant

, Davis-Arehart

, Luo

, et al. Synthesis and characterization of photopolymerized multifunctional hydrogels: Water-soluble poly(vinyl alcohol) and chondroitin sulfate macromers for chondrocyte encapsulation. Macromolecules, 2004; 37(18):6726–6733; doi: 10.1021/MA0499324/ASSET/IMAGES/LARGE/MA0499324F00007.JPEG

13.

Shao

, Gao

, Xie

, et al. Synchronous 3D bioprinting of large-scale cell-laden constructs with nutrient networks. Adv Healthc Mater, 2020; 9(15):1901142; doi: 10.1002/ADHM.201901142

14.

Griffith

, Miller

, Sainson

RCA

, et al. Diffusion limits of an in vitro thick prevascularized tissue. Tissue Eng, 2005; 11(1–2):257–266; doi: 10.1089/TEN.2005.11.257

15.

Park

, Chung

, Lee

, et al. Microporous cell-laden hydrogels for engineered tissue constructs. Biotechnol Bioeng, 2010; 106(1):n/a-n/a; doi: 10.1002/bit.22667

16.

Yang

, Shridhar

, Gerwitz

, et al. An in vitro vascular chip using 3D printing-enabled hydrogel casting. Biofabrication, 2016; 8(3):035015; doi: 10.1088/1758-5090/8/3/035015

17.

Wang

, Lu

, Lam

, et al. Fabrication of cell-laden macroporous biodegradable hydrogels with tunable porosities and pore sizes. Tissue Eng C Methods, 2015; 21(3):263–273; doi: 10.1089/TEN.TEC.2014.0224/ASSET/IMAGES/LARGE/FIGURE7.JPEG

18.

Chan

, Zorlutuna

, Jeong

, et al. Three-dimensional photopatterning of hydrogels using stereolithography for long-term cell encapsulation. Lab Chip, 2010; 10(16):2062; doi: 10.1039/c004285d

19.

Salvador

, Oliveira

, Mano

. Leachable-free fabrication of hydrogel foams enabling homogeneous viability of encapsulated cells in large-volume constructs. Adv Healthc Mater, 2020; 9(20):2000543; doi: 10.1002/ADHM.202000543

20.

Lee

, Hudson

, Shiwarski

, et al. 3D bioprinting of collagen to rebuild components of the human heart. Science, 2019; 365(6452):482–487; doi: 10.1126/SCIENCE.AAV9051/SUPPL_FILE/AAV9051S9.MP4

21.

Zhu

, Ma

, Gou

, et al. 3D printing of functional biomaterials for tissue engineering. Curr Opin Biotechnol, 2016; 40:103–112; doi: 10.1016/j.copbio.2016.03.014

22.

Carmeliet

, Jain

. Angiogenesis in cancer and other diseases. Nat 2000 4076801, 2000; 407(6801):249–257; doi: 10.1038/35025220

23.

, Fan

, Zhu

. Three-dimensional printing: The potential technology widely used in medical fields. J Biomed Mater Res Part A, 2020; 108(11):2217–2229; doi: 10.1002/JBM.A.36979

24.

Scott

, Nichols

, Kuntz-Willits

, et al. Modular scaffolds assembled around living cells using poly(ethylene glycol) microspheres with macroporation via a non-cytotoxic porogen. Acta Biomater, 2010; 6(1):29–38; doi: 10.1016/J.ACTBIO.2009.07.009

25.

Mikos

, Thorsen

, Czerwonka

, et al. Preparation and characterization of poly(l-lactic acid) foams. Polymer (Guildf), 1994; 35(5):1068–1077; doi: 10.1016/0032-3861(94)90953-9

26.

Somo

, Akar

, Bayrak

, et al. Pore interconnectivity influences growth factor-mediated vascularization in sphere-templated hydrogels. Tissue Eng C Methods, 2015; 21(8):773; doi: 10.1089/TEN.TEC.2014.0454

27.

Çetin

, Sera Kahraman

, Gümüşderelioğlu

, et al. Novel scaffolds based on poly(2-hydroxyethyl methacrylate) superporous hydrogels for bone tissue engineering. J Biomater Sci, 2011; 22(9):1157–1178; doi: 10.1163/092050610X501704

28.

Petryk

, Haas

, Vakil

, et al. Shape memory polymer foams with tunable interconnectivity using off-the-shelf foaming components. J Biomed Mater Res A, 2022; doi: 10.1002/JBM.A.37383

29.

Beaman

, Shepherd

, Satalin

, et al. Hemostatic shape memory polymer foams with improved survival in a lethal traumatic hemorrhage model. Acta Biomater, 2022; 137:112–123; doi: 10.1016/J.ACTBIO.2021.10.005

30.

Kuang

, Yuk

, Huh

. Polysaccharide-based superporous hydrogels with fast swelling and superabsorbent properties. Carbohydr Polym, 2011; 83(1):284–290; doi: 10.1016/j.carbpol.2010.07.052

31.

Singhal

, Rodriguez

, Small

, et al. Ultra low density and highly crosslinked biocompatible shape memory polyurethane foams. J Polym Sci Part B Polym Phys, 2012; 50(10):724–737; doi: 10.1002/polb.23056

32.

Desai

, Tang

, Ross

, et al. Critical factors affecting cell encapsulation in superporous hydrogels. Biomed Mater, 2012; 7(2):024108; doi: 10.1088/1748-6041/7/2/024108

33.

, Casillas

, Krishnamoorthy

, et al. Effects of Irgacure 2959 and lithium phenyl-2,4,6-trimethylbenzoylphosphinate on cell viability, physical properties, and microstructure in 3D bioprinting of vascular-like constructs. Biomed Mater, 2020; 15(5):055021; doi: 10.1088/1748-605X/AB954E

34.

Kang

, Armstrong

, Lee

, et al. Optimizing photo-encapsulation viability of heart valve cell types in 3D printable composite hydrogels. Ann Biomed Eng, 2017; 45(2):360–377; doi: 10.1007/S10439-016-1619-1/TABLES/1

35.

ISO

. ISO 10993-5:2009 Biological Evaluation of Medical Devices—Part 5: Tests for in Vitro Cytotoxicity. Int Organ Stand Geneva, 2009; 34.

36.

Lan

, Lan

, Chan

, et al. The effects of extracellular citric acid acidosis on the viability, cellular adhesion capacity and protein synthesis of cultured human gingival fibroblasts. Aust Dent J, 1999; 44(2):123–130; doi: 10.1111/J.1834-7819.1999.TB00213.X

37.

Chen

, Lv

, Liu

, et al. Effect of food additive citric acid on the growth of human esophageal carcinoma cell line EC109. Cell J, 2017; 18(4):493; doi: 10.22074/CELLJ.2016.4716

38.

Shi

, Bi

, Zhang

, et al. The effect of citric acid on the structural properties and cytotoxicity of the polyvinyl alcohol/starch films when molding at high temperature. Carbohydr Polym, 2008; 74(4):763–770; doi: 10.1016/j.carbpol.2008.04.045

39.

Michl

, Park

, Swietach

. Evidence-based guidelines for controlling pH in mammalian live-cell culture systems. Commun Biol, 2019; 2(1):1–12; doi: 10.1038/s42003-019-0393-7

40.

Icard

, Coquerel

, Wu

, et al. Understanding the central role of citrate in the metabolism of cancer cells and tumors: An update. Int J Mol Sci, 2021; 22(12); doi: 10.3390/IJMS22126587

41.

Duong

HTT

, Burford

. Effect of foam density, oil viscosity, and temperature on oil sorption behavior of polyurethane. J Appl Polym Sci, 2006; 99(1):360–367; doi: 10.1002/APP.22426

42.

Singhal

, Small

, Cosgriff-Hernandez

, et al. Low density biodegradable shape memory polyurethane foams for embolic biomedical applications. Acta Biomater, 2014; 10(1):67–76; doi: 10.1016/j.actbio.2013.09.027

43.

Lookabill

, Dulaney

, Shepherd

, et al. Sodium bicarbonate. Crit Care Toxicol Diagnosis Manag Crit Poisoned Patient, 2021; 2967–2986; doi: 10.1007/978-3-319-17900-1_169

44.

Fairbanks

, Schwartz

, Bowman

, et al. Photoinitiated polymerization of PEG-diacrylate with lithium phenyl-2,4,6-trimethylbenzoylphosphinate: Polymerization rate and cytocompatibility. Biomaterials, 2009; 30(35):6702–6707; doi: 10.1016/J.BIOMATERIALS.2009.08.055

45.

Nguyen

, Goering

, Reipa

, et al. Toxicity and photosensitizing assessment of gelatin methacryloyl-based hydrogels photoinitiated with lithium phenyl-2,4,6-trimethylbenzoylphosphinate in human primary renal proximal tubule epithelial cells. Biointerphases, 2019; 14(2):021007; doi: 10.1116/1.5095886

46.

Greene

, Lin

, Andrisani

, et al. Comparative study of visible light polymerized gelatin hydrogels for 3D culture of hepatic progenitor cells. J Appl Polym Sci, 2017; 134(11); doi: 10.1002/APP.44585

47.

Monteiro

, Thrivikraman

, Athirasala

, et al. Photopolymerization of cell-laden gelatin methacryloyl hydrogels using a dental curing light for regenerative dentistry. Dent Mater, 2018; 34(3):389–399; doi: 10.1016/J.DENTAL.2017.11.020

48.

Xia

, Luo

. Vascularization in tissue engineering: The architecture cues of pores in scaffolds. J Biomed Mater Res Part B Appl Biomater, 2022; 110(5):1206–1214; doi: 10.1002/JBM.B.34979

49.

Dragan

. Design and applications of interpenetrating polymer network hydrogels. A review. Chem Eng J, 2014; 243:572–590; doi: 10.1016/j.cej.2014.01.065

50.

Merlin

, Sivasankar

. Synthesis and characterization of semi-interpenetrating polymer networks using biocompatible polyurethane and acrylamide monomer. Eur Polym J, 2009; 45(1):165–170; doi: 10.1016/j.eurpolymj.2008.10.012

51.

Maimouni

, Cejas

, Cossy

, et al. Microfluidics mediated production of foams for biomedical applications. Micromachines, 2020; 11(1):1–20; doi: 10.3390/mi11010083