Cell Printing in Complex Hydrogel Scaffolds

Abstract

Advancements in the microfabrication of soft materials have enabled the creation of increasingly sophisticated functional synthetic tissue structures for a myriad of tissue engineering applications. A challenge facing the field is mimicking the complex microarchitecture necessary to recapitulate proper morphology and function of many endogenous tissue constructs. This study describes the creation of PEGDA hydrogel microenvironments (microgels) that maintain a high level of viability at single cell patterning scales and can be integrated into composite scaffolds with tunable modulus. PEGDA was stereolithographically patterned using a digital micromirror device to print single cell microgels at progressively decreasing length scales. The effect of feature size on cell viability was assessed and inert gas purging was introduced to preserve viability. A composite PEGDA scaffold created by this technique was mechanically tested and found to enable dynamic adjustability of the modulus. Together this approach advances the ability to microfabricate tissues that better mimic native constructs on cellular and subcellular length scales.

I. Introduction

As demand for functional, synthetic tissue-alternatives far outpaces supply, innovative means of their production has increasingly become a focus of biomaterial research [1]. Although emerging technologies such as 3D printing enable the creation of increasingly detailed tissue scaffolds, current methods are unable to mimic the complex microarchitecture of endogenous constructs [2]. This capability is critical for applications such as organ engineering where heterogeneous cell organization and scaffold modulus are key to proper morphological development [3-5].

A challenge in developing microscale synthetic hydrogel (microgel) tissue structures is selecting materials that enable patterning on single cell length scales while maintaining cell viability and desired mechanical characteristics [6]. One popular synthetic tissue scaffold, polyethylene glycol diacrylate (PEGDA), has been widely used in tissue engineering due to its good biocompatibility, tunable mechanical properties, and customizable chemistry [7-9]. Facile spaciotemporal manipulation at scales approaching a single micron [10] and success as a 3d printing bioink [11] also make PEGDA well suited to micropatterning tissue architecture. However, despite these advantages, PEGDA has shown adverse cell viability effects that compound at decreasing microenvironment length scales [12].

This study aimed to address the problem by developing PEGDA hydrogel microenvironments that maintain a high level of viability at single cell length scales and integrate these into a composite scaffold with tunable modulus.

Single cells were encapsulated in cylindrical PEGDA hydrogel features within a microfluidic nitrogen purged device. The width and height of the posts were systematically varied and the viability of cells assessed after encapsulation. Cytotoxic byproducts of polymerization were modeled with a reaction diffusion model. PEGDA was polymerized in the continuous phase of microfluidic devices containing microgel features to form composite scaffolds and their modulus was measured by tensile testing. Nitrogen-purged PEGDA microstructures exhibited excellent viability at single cell length scales compared to those fabricated under ambient conditions. Tensile testing of composite scaffolds demonstrated dynamically adjustable modulus with varying microgel polymerization time. Excellent viability coupled with tunable modulus properties suggests feasibility of this platform for a variety of tissue microfabrication and tissue-on-a-chip applications.

II. Methodology

A. Device Preparation

Sylgard 184 Polydimethylsiloxane (PDMS) (Dow Corning) was poured onto silicon wafers molds with straight microfluidic channels of varying depth photopatterned onto their surface. The molds were vacuumed to remove air bubbles and cured at 70 °C for 3 hours. Microfluidic devices were removed from the molds using a scalpel to cut around the features. The PDMS microfluidic devices were punched with an inlet and outlet hole, then plasma bonded with a PDC-002 plasma cleaner (Herrick) to No. 1 glass slides. Nitrogen purge chambers were constructed in an analogous manner and were plasma bonded to the top of devices directly over the microfluidic channel. Completed devices were placed on a 37 °C hot plate until use.

B. Cell Preparation

MDCK cells (ATCC) at 80% confluency where trypsinized for 12 minutes, then 4 ml of DMEM (Genesee) containing 10% FBS (GIBCO) was added to neutralize the trypsin. That volume was then spun down at 200 RPM for 5 minutes, the media decanted, and then resuspended in 1ml of PBS (Fisher) containing 3% FBS.

C. Hydrogel Prepolymer Preparation

10,000 molecular weight PEGDA (Sigma) was added to 1x PBS at 15 wt% and vortexed for 5 min. Lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) was synthesized [13] and added to 1x PBS at 1 wt% and vortexed for 2 min. MDCK cells at a concentration of 10⁵ cells per ml in PBS were obtained and kept at a temperature of 37 °C via water bath until ready for use. Aliquots of PEGDA 10,000 and LAP were added to the cell solution to achieve 8 wt% and 0.014 wt% by total mass, respectively, then vortexed gently for 10 seconds.

D. Microgel Fabrication

The microgel fabrication apparatus consisted of a Polygon 400 digital micromirror device (DMD) (Mightex) attached to the rear aperture of an IX 81 inverted microscope (Olympus) with a Retiga 2000R camera (Q Image). MetaMorph software (Molecular Devices) was used to control image visualization and stage automation. Prepolymer cell solution described in C. was immediately added to a pre-warmed microfluidic channel device and placed back on a 37 °C hotplate for 10 min or until cells settled to the bottom of the channel. The device was then placed on the microscope stage and a 20x LUCPlan FLN objective (Olympus) was used to focus at the bottom of the channel. Tygon tubing connected to a nitrogen source was inserted into the inlet of the purge chamber and pressurized to 7 psi. Mightex software was used to create a circular exposure area of desired size and the file loaded to the Polygon. The microscope stage was manually controlled to position cells in the center of the exposure area and exposed with 365 nm light at an intensity of 335 mW/cm2 for 3 seconds. This was repeated for as many cells as could be exposed in 4 minuets. At the end of the 4-minute period the device was removed from the stage and flushed with 0.5 mL of LIVE/DEAD solution (Fisher) warmed to 37 °C at standard concentration in PBS. This protocol was repeated in triplicate for channels of varying height at a constant feature size of 250μm, and features of varying diameter at a constant channel height of 125μm. It was then repeated in its entirety without the nitrogen purge.

E. Viability Analysis

Devices were placed in an Galaxy 170 S incubation chamber (New Brunswick) immediately after flushing and incubated for 20 minuets. After incubation they were removed and imaged using an IX 71 inverted microscope (Olympus). An X-Cite 120 LEDmini fluorescent illumination system (Excelitas Technologies) was used to provide white light, which was filtered through a FITC or Texas Red filter cube for visualizing fluorescence of live and dead stains, respectively. Images of encapsulated cells were captured in brightfield, FITC and Texas Red. Image J software was used to analyze viability of images using the color combining feature. Data was plotted in IGOR (Wavemetrics).

F. Modeling

A reaction diffusion model of PEGDA microgels was created using COMSOL Multiphysics software. Cytotoxic reactive oxygen species (ROS) generated during microgel polymerization were modeled for features of varying width and height, and the data correlated with experimental viability results.

G. Composite Scaffold Fabrication

700 MW PEGDA (Sigma) prepolymer solution prepared in the same manner as PEGDA 10,000 described previously, was added to 25mm × 5mm × 1mm deep channel devices constructed as described above, but not plasma bonded. Channels were attached to no. 1 glass slides with a thin layer of PDMS. Arrays of 200 μm diameter microgels were polymerized using a X-Cite 200 DC Multispectral UV source (Excelitas Technologies) attached to the rear aperture of an IX 71 inverted microscope (Olympus) with a 20x LUCPlan FLN objective (Olympus) and adjustable iris. A total of 500 microgel were fabricated in each channel device. After fabrication channels were rinsed with 1mL 1x PBS and refilled with prepolymer solution. Filled devices were then exposed to multispectral UV light in bulk for 3 seconds via an OmniCure S2000 (Excelitas Technologies). Completed composite scaffolds were removed from the channel devices and soaked in 1x PBS.

H. Tensile Testing

Scaffolds were tested using a DMA Q800 (TA Instruments). Samples were placed between two tensile grips and loaded in tension until failure. Both ends of the samples were wrapped with 100-grit Aluminum Oxide sandpaper (3M) to avoid slippage while maintaining low clamping force. The tests were conducted in ambient condition and the loading rate was chosen to be 0.5 N/min yielding failure within approximately 2 minutes. The force and displacement was recorded by the DMA Q800 and corresponding stress and strain were calculated by the default instrument software.

III. Results

A. Viability with Decreasing Channel Height

Cells encapsulated in 250 μm diameter microgels showed constant viability of 100% with decreasing channel height until a critical height of 25 μm was reached and viability dropped to 0% (Fig. 1). COMSOL modeling results showed this depth corresponded with a 63% increase in ROS generation.

Fig. 1.

A) Plot of percent cell viability and percent increase in reactive oxygen species (ROS) after microgel fabrication vs. the channel height in which the microgel feature was fabricated. B) Brightfield image of cells fabricated in 250 μm diameter microgels within a 55 μm deep channel. C) Color combined image of LIVE/DEAD stain for B) showing full viability after microgel fabrication.

B. Viability with Decreasing Microgel Width

Cells encapsulated in 125 μm depth channels showed a trend of decreasing viability from 100% to 9 ± 3% with decreasing microgel diameter (Fig 2). COMSOL modeled ROS concentration showed a trend of increasing with decreasing microgel diameter.

Fig. 2.

A) Plot of percent cell viability and percent increase in reactive oxygen species (ROS) after microgel fabrication vs. microgel width. B-E) Brightfield images of cells fabricated in microgels of 100 μm, 60 μm, 45 μm, and 30 μm, respectively.

C. Viability with Nitrogen Purging

Cells encapsulated in 30 μm diameter microgels within a 100 μm deep nitrogen purged channel showed a 76 ± 6% increase in viability compared to the trial at ambient conditions (Fig. 3). Cells encapsulated in 250 μm diameter microgels within a 25 μm deep nitrogen purged channel showed an 88 ± 10% increase in viability over the trial at ambient conditions.

Fig. 3.

A) Plot of percent cell viability after encapsulation in a 30 μm diameter microgel under ambient atmospheric conditions and under nitrogen purge (+ N2). B) Plot of percent cell viability after microgel encapsulation in a 25 μm deep channel under ambient atmospheric conditions and under nitrogen purge (+N2). C) Brightfield image of cells fabricated in 30 μm diameter microgels within a 100 μm deep nitrogen purged channel. D) Color combined image of LIVE/DEAD stain for B) showing full viability after microgel fabrication with nitrogen purge.

D. Composite Scaffold Modulus

Scaffolds fabricated with microgel features exposed for 30 milliseconds had a modulus (stress vs. strain) of 1.87, versus 0.84 for scaffolds with microgel features exposed for 15 milliseconds. The modulus of controls without microgel patterning was measured at 3.55.

IV. Discussion

Microfluidic channel devices were fabricated from highly gas permeable PDMS. Although this characteristic is advantageous for long-term cell culture and a standard in tissue engineering [14], oxygen present during the PEGDA polymerization reaction is preferentially converted to cytotoxic ROS [12]. High oxygen permeability through the PDMS channel top coupled with relatively low oxygen and ROS permeability through the PEGDA prepolymer solution contributed to the sharp drop in viability observed in the 25 μm height microgel (Fig 1), but a trend of steadily decreasing viability with decreasing microgel width (Fig 2). Purging the channel with nitrogen during polymerization mitigated the flux of atmospheric oxygen into the microgel and dramatically improved viability in 25 μm height and 30 μm width features (Fig 3). Tensile testing of composite scaffold samples suggested dynamic adjustability of the modulus by changing the microgel polymerization time (Fig 4). Longer UV exposure of microgels resulted in the surrounding matrix to have increased interfacial bonding due to enhanced grafting and chain entanglement. The stronger interfacial bond between the microgel and the matrix led composite samples with 30ms UV exposure to possess higher elastic modulus than those exposed for 15ms.

Fig. 4.

Stress vs. strain plotted for composite scaffolds with microgel features exposed for 15 milliseconds, 30 milliseconds, and solid hydrogel control.

V. Conclusion

This study demonstrated the ability to fabricate high viability microgels on a single cell scale and control mechanical properties of complex hydrogel scaffolds. It is anticipated that these capabilities will enable the engineering of superior synthetic tissues in the future.