The printing channels of the standard “quill pen” surface patterning tools (SPTs) used with the Bioforce Nano eNabler^TM (NeN) are too narrow (approx. 5–40 µm in width) to accommodate the printing of mammalian cells (approximately 10 µm in diameter and larger), leading to the development of new SPTs with larger printing channels to permit cell passage. Polymeric SPTs were fabricated out of SU-8 negative photoresist using standard lithographic techniques and as described previously [27,37] and shown schematically in Figure 1. In short, the first SU-8 layer was spun onto a silicon wafer and exposed via photolithography to form the basal layer of the SPT, followed by a second resist layer that is exposed to form the reservoir and channel walls of the printing cantilever. Finally the SU-8 SPTs were released from the wafer during the SU-8 development. A micrograph of a completed SU-8 SPT is shown in Figure 2.

For our experiments, a Bioforce Nano eNabler^TM (Bioforce Nanosciences, Inc., Ames, IA, USA) was used for cell-based printing. The NeN benchtop molecular printing system was not originally optimized for live cell patterning (by the manufacturer). Some modifications to the system were required in order to maintain printed cell viability and enhance printing versatility. The NeN’s humidity sensor was removed to prevent the system from deactivating the humidifier while printing, avoiding cellular desiccation during the patterning process by maintaining >90% relative humidity within the printing enclosure. The standard SPT holder for the NeN was modified by elongating the holder’s arm to enable patterning onto the bottom of culture dishes and plates without crashing into the side walls of the dish or well. Due to this elongation, adjusting the field of view of the printing optic was required to properly image the lowered SPT during patterning. This was performed by manually lowering the objective in its housing until the proper focus at the SPT tip was achieved. When using this elongated SPT holder the tip’s angle of incidence is approximately 60° to the substrate, which allows the printing cantilever to deflect 200–400 µm after surface contact before tip failure.

Thiolated hyaluronan (Glycosil^®, 1% w/v), thiolated porcine gelatin (Gelin-S^®, 1% w/v) and polyethylene glycol diacrylate (PEGDA) MW 3400 (Extralink^®) were obtained from BioTime Inc. (Alameda, CA, USA). Irgacure^® 2959 was from BASF (Southfield, MI, USA). Polyethylene glycol (PEG) norbornene was derived from 4-arm PEG (MW 20,000) and was synthesized as previously described [35]. To form gels in 24-well tissue culture plates, all components (Glycosil, Gelin-S, and 1% w/v PEG norbornene) were reconstituted in 0.05% w/v Irgacure 2959 and mixed in a 2:2:1 volumetric ratio. The final concentrations (w/v) of Glycosil, Gelin-S, PEG norbornene, and Irgacure 2959 are 0.4%, 0.4%, 0.2%, and 0.05%, respectively. UV crosslinking was performed using a handheld compact UV lamp (365 nm, 4 W, 115 V, 0.16 A, Cat no. UVL-21; UVP, Upland, CA, USA). Oscillatory shear measurements of the elastic modulus, G′, and the viscous modulus, G′′, were obtained at room temperature using a Bohlin CVO rheometer (Malvern Instruments, Worcestershire, UK). Cells at the appropriate concentrations were mixed with the crosslinker in a 1:9 volumetric ratio and subsequently mixed with Glycosil and Gelin-S solutions and gelled as described above. For bioprinting, the hydrogel mixture (HyStem-C/PEGnor) was prepared as described above but with PEG-norbornene at 5% w/v initial concentration (1% w/v final concentration). The HyStem-C/PEGnor solution was then mixed with the cell-suspension at a 3:1 ratio to produce the final printing solution.

For initial hydrogel biocompatibility experiments, adipose derived stem cells (ADSCs) were used. ADSCs were isolated from human tumescent lipoaspiration samples as previously described [38]. Briefly, stromal vascular fraction (SVF) was prepared from human lipoaspirate. The adherent cellular fraction (adipose derived stem cells) was isolated by incubating the SVF incubated overnight at 37 °C/5% CO₂ in control medium (Dulbecco’s modified Eagle’s Medium (DMEM), 10% fetal bovine serum (FBS), 1% antibiotic/antimycotic solution) on plastic tissue culture plates. Following incubation, the plates were washed well with 1× PBS. The remaining adherent cell population (adipose derived stem cells (ADSCs)) were cultured at 37 °C/5% CO₂ in non-inductive control medium at sub-confluent levels.

For cell printing experiments, a variety of cell types were used, including fibroblasts, macrophages, and stem cells. NIH-3T3 murine fibroblasts were grown in DMEM medium containing 10% fetal bovine serum (FBS), 2 mM L-glutamine, and 1% penicillin streptomycin. Murine alveolar macrophages (MHS) cells were transfected with the pEGFP-N1 plasmid to allow constitutive expression of roGFP-R12 within the cytosol, while their media was supplemented with G418 antibiotic to maintain the plasmid. The resultant MHS-roGFP-R12 cells were grown in RPMI-1640 medium (Gibco^®) supplemented with 10% FBS, 0.05 mM 2-mercaptoethanol, and 1 mg/mL G418 antibiotic. CCE mouse embryonic stem cells (mESCs) (StemCell Technologies, Vancouver, Canada) were grown on gelatin-coated tissue culture flasks in a maintenance medium consisting of Dulbecco’s Modified Eagle’s Medium (DMEM with 4.5 g/L D-glucose), supplemented with 15% (v/v) fetal bovine serum (FBS, StemCell Technology), 100 U/mL penicillin, 100 µg/mL streptomycin, 0.1 mM non-essential amino acids, 10 ng/mL murine recombinant leukemia inhibitory factor (StemCell Technology), 0.1 mM monothioglycerol, 2 mM L-glutamine, and 1 mM sodium pyruvate (Sigma-Aldrich) [39,40,41]. All cell types were incubated overnight at 37 °C/5% CO₂. Cells were harvested at 70–80% confluency, centrifuged at 3,000×g for 5 min, and re-suspended in fresh growth media via vortexing to generate a suspension of single cells.

Qualitative confirmation of our ability to pattern viable eukaryotic cells was performed by subsequent patterning, incubation and observation of printed NIH-3T3 fibroblasts. The SU-8 SPTs selected for this experiment possessed channel widths of 100 µm and were treated with UV/ozone for 30 min to enhance the hydrophilic nature of the printing channel. After treatment, the SU-8 SPT was loaded with 2 µL of the cell/hydrogel mixture containing NIH-3T3 cells and used to pattern a glass cover-slip with a 5 × 5 array consisting of columns and rows with 500 µm spacing. After the printing process, cell-patterned cover-slips were sealed in a dish and exposed to long-wave, 365 nm UV lamp for 2 min to crosslink the HyStem-C/PEGnor hydrogel and immobilize the patterned cells. The cover-slip was then submerged in fresh media and incubated at 37 °C/5% CO₂ until imaging was performed. Media was replaced every 24 h to maintain cell health. Bright-field imaging was performed at 18, 32, 54, 78, 100, and 194 h after patterning in order to monitor cell proliferation.

Mouse embryonic stem cells (mESCs) were patterned and fluorescently stained in order to visualize their growth within the three dimensional (3D) microenvironment of the crosslinked droplet. The Invitrogen LIVE/DEAD^® Viability/Cytotoxicity kit (calcein AM and ethidium homodimer-1) for mammalian cells was used for cell staining (Carlsbad, CA, USA). These stains were selected to both visualize cells during confocal microscopy, as well as qualitatively assess the viability of the patterned mESCs. The mESCs were suspended in HyStem-C/PEGnor and printed using UV/ozone treated SU-8 SPTs onto the bottom of a glass bottom dish (WillCo Wells B.V., Amsterdam, NL, USA) in a 5 × 5 array (500 µm spacings). The patterned dish was then sealed and exposed to UV light for 2 min, to crosslink the HyStem-C/PEGnor hydrogel and immobilize the patterned cells. The cell array was then submerged in fresh media and incubated for 72 h at 37 °C/5% CO₂, with media being refreshed every 24 h. At hour 72 confocal laser scanning microscopy (Leica SP-5, Leica Microsystems) was performed to gather three-dimensional information on the cell growth occurring within the printed spots.

MHS-roGFP-R12 was developed essentially as described by Melillo et al. [42]. MHS-roGFP-R12 cell-laden hydrogel mixture was loaded into UV/ozone treated SU-8 SPTs and deposited in a 5 × 5 array (500 µm spacings) onto the bottom of a glass bottom dish (WillCo Wells B.V.). The patterned dish was then sealed and UV-crosslinked for 2 min to immobilize cell deposits. After exposure, the dish was opened and the cell array was rinsed three times with 1× HBSS buffer (Gibco^®), after which fresh media was added to the dish, submerging the printed array. The added media lacked any FBS to prevent neutralization of ROS in solution during the detection trials. Following addition of media, the dish was mounted on a confocal laser scanning microscope (Leica SP5, Leica Microsystems) and the cell array was imaged by fixing the emission wavelength at 515 nm while the excitation wavelength was toggled between 405 and 488 nm, to quantify oxidization and reduction fluorescence shifts, respectively. After base-line images were gathered at 1 min intervals for 5 min, the cells were exposed to 500 µM of hydrogen peroxide and imaged at 1 min intervals for a period of 8 min. Cells were then exposed to 10 mM of Dithiothreitol (DTT) in order to neutralize oxidizing activity and return cellular roGFP to a fully reduced state, followed by imaging at 1 min intervals for 4 min. Results gathered from individual cells were quantified using Leica LAS image processing software and plotted to determine if shifts in roGFP excitation would directly correspond to the addition of oxidizers and reducers to the environment.

Initial bioprinting experiments with HyStem^®-C showed that we did not have the requisite temporal control of gelation for bioprinting on glass cover-slips using the modified SPT in the Nano eNabler device. More specifically, HyStem-C gelled too slowly after spotting. This constraint required close attention to maintenance of chamber humidity; in addition, spotting experiments were slowed since we were required to wait for gelation before adding media. Another limitation was the hydrogel’s spontaneous crosslinking. Since crosslinking typically occurs via the Michael Addition reaction between the hydrogel thiols and the PEGDA acrylates, the reaction gels in 10–20 min and can potentially clog the SPT channels mid-experiment. The solution was to substitute the polyethylene glycol (PEG)-diacrylate crosslinker with PEG-norbornene [35], which can be photo-coupled to thiol-containing molecules using radical-initiated “thiol-ene” chemistry [43]. Specifically, radically initiated thiyl radicals form from the cysteines of the thiolated hyaluronic acid (Glycosil) and porcine gelatin (Gelin-S) components in the presence of photoinitiator (Irgacure 2959) and ultraviolet light. The thiyl groups couple through the alkene bond of the norbornene adducts on the ends of four-armed PEG molecule (Figure 3).

Using thiol-ene chemistry, hydrogels were formed within thirty seconds and attained a shear elastic modulus (G′) of up to 550 Pa after 15 min with only a final concentration of 0.2% w/v PEG norbornene (Figure 4(A)). The gelation occurred at every ratio of thiols and alkenes (from thiol:alkene molar ratios of 31 down to 1.94 depending on the PEG norbornene concentration used (Figure 4(B)). Importantly, gel stiffness could also be modulated with increasing concentration of PEG norbornene to a shear elastic modulus greater than 2 kPa (Figure 4(C)). The optimized Irgacure 2959 concentration is 0.05% (Figure 4(C)). Therefore, photoinitiated thiol-ene chemistry enabled rapid gelation of thiolated hyaluronan/gelatin hydrogels, thus providing a matrix that is well-suited for bioprinting.

Biocompatibility of the HyStem-C/PEG norbornene hydrogels was explored using human adipose-derived stem cells (ADSCs) isolated from human lipoaspirates. Human ADSCs were chosen for these experiments due to their clinical potential (e.g., for tissue engineering and cell therapy [44,45]). ADSCs were cultured on top of either HyStem-C/PEG norbornene or HyStem-C hydrogels (the latter as a control). The proliferation rates between cells at two different cell numbers was similar for both hydrogels after one week (Figure 5), indicating that neither the UV exposure nor presence of PEG norbornene and Irgacure 2959 presented cytotoxicity issues at the UV intensities used.

For cell printing experiments, we first attempted printing of a robust fibroblast line, namely NIH-3T3 cells. NIH-3T3 cells were printed in the HyStem-C/PEG norbornene hydrogel and imaged at several time points during their growth, starting at 18 h and ending at 192 h, when the cells began to overgrow the surface (Figure 6). Directly after patterning, cells possessed a rounded morphology, similar to that displayed by unattached cells suspended in solution. However, images at 18 h show cells beginning to spread and form long pseudopodia, indicative of cell attachment to either the glass surface or the 3D HyStem-C/PEGnor matrix (Figure 6). As growth progressed through 32 h, cells began proliferating, and at 100 h started to escape from the printed matrix and spread across the glass substrate. By 192 h the cells had completely over-grown the entire array area (4 mm²), obscuring the array and discouraging further imaging. At 192 h the original depositions remained visible, due to the higher density of cells within the 3D matrix, as compared to the confluent cells growing on the interstitial, two dimensional glass surface.

To assess the 3D nature of printed spots of cells in HyStem-C/PEG norbornene, and further assess biocompatibility with other cell lines, we printed mouse embryonic stem cells (mESCs). Patterned mESCs were subjected to live/dead staining. After staining, confocal laser scanning microscopy was used to reveal the cell positions within the printed matrix material by producing an optical cross-section of the deposition from the complied image planes (Figure 7). Examination of the deposition cross-section confirms that the cells growing within the matrix reach a height of approximately 20–25 µm from the glass surface. Like many eukaryotic cells, the average mESC cell possesses a diameter of roughly 10 µm, suggesting that the growth within the microenvironment of the printed matrix is two to three cell layers in height, which is indicative of 3D cell growth. Additionally, none of the cells within the examined deposition were positive for ethidium homodimer-1 (which stains the nucleus red for dead cells), demonstrating that high survivability of mESCs was maintained when submitted to our patterning process.

The ROS sensing capability of the printed MHS-roGFP-R12 cells was tested by exposing the cell array to oxidizing agent (hydrogen peroxide), followed by exposure to reducing agent (DTT). Cell response was quantified by selecting three separate cells within a single printed spot for observation. Utilizing confocal scanning laser microscopy in conjunction with Leica LAS image processing software, the resulting intensity at 515 nm was measured for each cell at 1 min intervals during exposure. The MHS-roGFP-R12 cells demonstrate maximum fluorescence emission at 515 nm when excited by 405 nm, if the roGFP-R12 protein is in the oxidized form. Likewise, they demonstrate maximum excitation at 488 nm when in the reduced state. Thus, by exciting cells at both 405 and 488 nm wavelengths and then taking the ratio of resulting emission at 515 nm (for both excitation wavelengths), one can deduce the relative state (oxidized vs. reduced) for the roGFP-R12 protein [14]. Ratiometric comparison of oxidized to reduced roGFP-R12 was performed for each of three cells, individually, which were then averaged together in order to quantify the overall response from the spot as a whole (Figure 8). Examining these data reveals that the printed cells were able to respond to both oxidizing and reducing environments rapidly and effectively. Results from all three cells demonstrate an oxidative response directly upon addition of hydrogen peroxide (minute 5), which increases toward a maximal level until a reductive response is elicited immediately upon addition of reductive DTT (minute 13). This result confirms the proper functioning of the patterned MHS-roGFP-R12 cells, as well as the capability to print a functioning and reversible live-cell-based ROS sensor.