Interlinked Macroporous 3D Scaffolds from Microgel Rods

To study complex cooperative cell behavior in 3D constructs, scaffold platforms need to show consistent performance in reproducibility, have suitable geometry for cell migration, and, at the same time, allow certain flexibility in terms of parameter alteration to investigate their influence on the living tissue¹. In recent years, the concept of macroporous annealed particles (MAP), first described by Segura et al., developed into an efficient and versatile platform for 3D scaffold production². The tailored composition of the microgels, which are the building blocks of the final 3D scaffold, predefines properties such as the stiffness of the construct, the selective chemical reactivity of the gel network, and the final pore size of the scaffold^2,3,4,5,6. Cell adhesive peptides as cues for scaffold-cell interactions are incorporated into the polymer network of the microgels to allow for cell attachment and can be varied to investigate their specific effects on cells in culture. The 3D scaffolds are stabilized by interlinking of the annealed injectable microgels due to covalent or supramolecular bonds, resulting in robust and defined constructs for cell culture^2,3,5,7,8.

Microfluidics has established itself as one of the most accurate and adaptable methods for the preparation of defined granular hydrogels⁹. The possibility of producing larger quantities of the required building blocks in a continuous process while maintaining their chemical, mechanical, and physical monodispersity contributes substantially to the suitability of this process. Furthermore, the size and shape of the produced microgels can be manipulated by various methods such as batch emulsions, microfluidics, lithography, electrodynamic spraying, or mechanical fragmentation, which determine the geometry of the building blocks and, thus, the 3D structure of the final scaffold^1,10.

Recently, the concept of macroporous 3D scaffolds composed of functionalized microgel rods that rapidly interlink in aqueous solutions without further additives has been reported¹¹. The anisotropy of microgel rods resulted in higher porosities and pore distributions with larger pore sizes compared to employing spherical microgels in this study¹¹. In this way, less material creates larger pores with a variety of different pore geometries while maintaining the stability of the 3D scaffold. The system consists of two types of microgel rods with complementary primary amine and epoxy functional groups that are consumed within the interlinking reaction when coming in contact with each other. The functional groups that do not participate in the interlinking process remain active and can be used for selective post-modification with cell adhesive peptides or other bioactive factors. Fibroblast cells attach, spread, and proliferate when cultured inside the 3D scaffolds, first growing on the microgel surface and filling most of the macropores after 5 days. A preliminary co-culture study of human fibroblasts and human umbilical vein endothelial cells (HUVECs) showed promising results for the formation of vessel-like structures within the interlinked 3D scaffolds¹¹.

1. Required material and preparations for microfluidics

1. For the described microfluidic procedure, use 1 mL and 5 mL glass syringes and syringe pumps. On-chip droplet formation is observed via an inverted microscope equipped with a high-speed camera.
2. Create the microfluidic chip design (Figure 1B) using a computer-aided design software and produce a master template as already reported¹².
3. Achieve controlled UV-irradiation using a self-constructed UV-LED (λ = 365 nm, spot diameter ~4.7 mm) providing irradiance at 957 mW/cm² through a 0.13 mm thick cover glass.
   NOTE: Take into account possible local heat generation by the UV-LED during irradiation. If this is the case, ensure sufficient cooling by external airflow.

2. Microfluidic device production

NOTE: The microfluidic device production is based on a previous publication¹³.

1. Prepare 20 g of a 10:1 (by mass) mixture of polydimethylsiloxane (PDMS) and curing agent. Mix vigorously for 3 min.
2. Mix 60 mg of Oil Red O in 2.0 g of toluene. Add 50 µL of Oil Red O in toluene solution to the PDMS mixture. Mix until there is homogeneous color distribution to decrease undesired light scattering and keep the spot size focused during on-chip gelation.
3. Remove the bubbles by placing the mixture into a desiccator equipped with a vacuum pump.
4. Cast the PDMS mixture into the master template to a height of 5-5.5 mm and degas again.
5. Allow the PDMS to cure for 18 h at room temperature to avoid diazo dye degradation.
6. Cut out the PDMS structure and punch inlet and outlet holes into the structure using a line core sampling tool (0.77 mm inner diameter, 1.07 mm outer diameter).
7. Wash the PDMS and cover glass with isopropanol and deionized water repeatedly 5x and, subsequently, remove the liquid via pressured air or nitrogen flow after each washing step.
8. Treat the dry glass and PDMS together in a plasma oven at 0.2 mbar, with an oxygen flow of 20 mL/min for 60 s at 100 W, and connect directly to bind the glass and PDMS together to form the microfluidic device.
   NOTE: Avoid bending the PDMS structure during the binding process to minimize channel structure deformation.
9. To render the channels of the microfluidic chip hydrophobic, place the device together with 50 µL of trichloro-(1H,1H,2H,2H-perfluoroctyl)-silane in a desiccator under vacuum overnight (close the connection to the pump after decreasing the vapor pressure). Rinse the outside of the device with hydrofluoroether.
   CAUTION: Perform these steps in a fume hood and avoid any contact with the perfluoro silane. Use a glass vacuum desiccator sealed with grease. Clean the desiccator thoroughly before using for other purposes.

3. Solution preparation for microfluidics

1. For the continuous oil phase, mix paraffin oil and hexadecane (1:1 v/v) and add 8% (w/w) of a non-ionic surfactant. Fill one 1 mL (first oil, Figure 1) and one 5 mL (second oil, Figure 1) glass syringe.
2. Prepare the prepolymer solutions for the disperse phase in brown glass vials to prevent photoinitiator decomposition and unintended gelation.
   1. Amine component prepolymer solution containing GRGDS-PC
      1. For a 300 µL solution, weigh out 33.3 mg of star-polyethylene glycol-acrylate (sPEG-AC) and 3.03 mg of lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP).
      2. Dissolve 18.74 mg of 2-aminoethyl methacrylate hydrochloride (AMA) in 1.5 mL of deionized water and pass the solution through a syringe filter (0.20 µm pore size).
         NOTE: AMA is hygroscopic and should be discarded as soon as the solid forms lumps in the storage container.
      3. Prepare sterile 36 mM GRGDS-PC aliquots in water and keep at −20 °C until use.
      4. Use 291.7 µL of the AMA solution to dissolve sPEG-AC and LAP, and add 8.3 µL of the GRGDS-PC solution. Take up the solution with a 1 mL glass syringe and protect from light using aluminum foil.
   2. Epoxy component prepolymer solution
      1. For a 300 µL solution, weigh 33.3 mg of star-polyethylene glycol-acrylate (sPEG-AC), 3.03 mg of LAP and dissolve in 300 µL of deionized water. Add 2.4 mg of glycidyl methacrylate (GMA).
         NOTE: Vigorous shaking accelerates GMA dissolving. Take up the solution with a 1 mL glass syringe and protect from light using aluminum foil.

4. Production and purification of amine and epoxy functionalized microgel rods

Figure 1: Arrangement of the microfluidic on-chip gelation assembly. (A) Front view and angled view of the component arrangement during microfluidics. (B) Microfluidic chip design used for on-chip gelation of microgel rods. (1) PE tube to the first oil inlet. (2) Light-protected PE tube to the disperse phase inlet. (3) PE tube to the second oil inlet. (4) PE tube from the outlet to the product collection container. (5) UV lamp and irradiation location on the straight 80 µm channel near the outlet. (6) Microscope objective/observation position. (7) Colored PDMS component of the microfluidic device. (8) Cover glass bonded to the PDMS. Please click here to view a larger version of this figure.

1. Insert the needles into the PE tubes and remove the gas from the syringe and the tube.
2. Insert a PE tube into the outlet for product collection.
3. Place all the glass syringes in the syringe pumps and insert each tubing end into the corresponding inlet (Figure 1).
   NOTE: Protect the prepolymer tubing from light via aluminum foil or a black tube to avoid unintentional gelation.
4. Focus the microscope on the oil-water cross-section.
5. Start the first oil syringe pump (flow rate = 100-200 µL/h) to fill the channel with oil first to prevent channel wetting by the disperse phase.
6. Decrease the first oil flow rate to 30 µL/h and start the prepolymer syringe pump (flow rate = 100-200 µL/h) until the dispersed aqueous phase can be observed at the cross-section.
7. Set the prepolymer flow rate to 30 µL/h and focus the microscope on the outlet.
8. Start the second oil syringe pump (flow rate of 300 µL/h) and wait until the flow regime is stable.
9. Place the outlet tube in a collection container.
   NOTE: Place the end of the tube in the upper part of the container to avoid a pressure increase due to the accumulation of the product over time.
10. Set the UV irradiation system such that the irradiance is in the range of 900-1000 mW/cm² (used irradiance = 957 mW/cm²) and the irradiation spot is in the straight channel part before the outlet (Figure 1B).
    NOTE: Make sure to not irradiate near the oil-water cross-section to avoid clogging the channel. For additional protection, cover the channel structures prior to the irradiation spot on the top of the unit.
11. Before UV irradiation, adjust the flow rates of the prepolymer and the first oil to achieve the desired aspect ratio in the range of 3.0 to 4.5, and set the irradiation time of the dispersed phase to ~2.3 s, depending on the size of the irradiation spot.
12. Start UV irradiation and, if necessary, adjust the flow rates again according to the previous subsection.
    NOTE: Observe the production at the beginning and monitor for stable flow behavior in the outlet and the uniformity of the microgel rods. The second oil flow can be adjusted to optimize product transport within the outlet.
13. Change the collection container and note the product collection start time and flow rates.
    NOTE: Protect the product from dust. Stop collecting before any syringe runs low on solution.
14. To end collection, remove the collection container, noting the time. Stop irradiation and all the syringe pumps.
15. Wash the product subsequently 5x each with n-hexane, isopropanol, and deionized water. Remove the supernatant after rod sedimentation.
    NOTE: After each solution addition, disperse the product carefully and wait 10 min before replacing the solvent, considering molecular diffusion. Replace the isopropanol gradually with water to prevent the rods from floating to the surface of the liquid. Multiple additional washing steps with water decrease the remaining isopropanol traces.

5. Macroporous scaffold formation

1. Determine the number of microgel rods per dispersion volume for the epoxy and amine functionalized samples.
2. Adjust the number of epoxy and amine functionalized samples by dilution or concentration to a similar value between 1,000-5,000 rods/100 µL.
   NOTE: Use a centrifuge at ~2,000 x g for 5-10 s to speed up the sedimentation of the microgel rods. If the number of rods per dispersion volume is low, rod-interlinking may result in smaller rod clusters instead of one stable construct. If the number of rods per dispersion volume is high, the mixing quality of the two components will be compromised.
3. Transfer the first component (~1,200 rods) dispersion into a conical 1.5 mL or 2 mL transparent vial.
4. Add the second component in a controlled manner in a continuous operation (100 µL pipette). After addition, mix the contents directly using the pipette to take up liquid and add it again. The interlinked structure forms within seconds during mixing.
   NOTE: If multiple clusters form instead of one coherent structure, recheck the number of microgel rods per volume or provide more controlled mixing of the two components.

6. Cell adhesive post-modification

1. Calculate the theoretical number of epoxy groups in the interlinked structure based on the flow rate of the dispersed polymer phase, the number of microgels collected during a specific time, and the dilution factor of the microgel dispersion used for scaffold formation.
   NOTE: Approximate the theoretical number of epoxy groups per scaffold by the following equation.
   
   
   n_th: Theoretical amount of substance
   c_GMA: GMA concentration in prepolymer solution
   Q: Flow rate of prepolymer solution
2. Add GRGDS-PC solution to the interlinked structure to modify all remaining epoxy groups with the cell adhesive peptide bearing a free amine and thiol (n[theoretical epoxy groups]/n[GRGDS-PC] = 1). Leave at room temperature overnight.
3. Remove the unreacted molecules by washing with deionized water and removing the supernatant.

7. Sterilization and transfer into cell culture media

1. Reduce the water level to just submerge the interlinked scaffold.
2. Open the vial and irradiate with UV light of λ = 250-300 nm. Close the vial and transfer the vial onto a clean bench. Wash 1x with sterile water.
3. Replace the water in the vial with cell culture media and allow for equilibration for 5 min. Repeat this with fresh cell culture media 2x.
4. Transfer the macroporous scaffold into a cell culture well plate for the experiment by pouring or using a spatula.

Figure 2: Macroporous crosslinked scaffold structure. (A) 3D projection of a 500 µm confocal microscopy Z-stack of the interlinked macroporous scaffold. Scale bar represents 500 µm. (B) Interlinked scaffold composed of ~10,000 microgel rods on a cover glass taken directly out of water. Scale bar represents 5 mm. Please click here to view a larger version of this figure.

This protocol results in a stable 3D macroporous construct composed of interlinked amine and epoxy functionalized microgel rods (Figure 2A). The construct should exhibit a compact geometry if the described type of mixing is used, which is formed within 2 s or 3 s (Figure 2B).

The interlinked construct stability depends on the microgel rod building blocks of which it is composed. The amine functionalized microgel rods exhibit an average stiffness of 2.0 ± 0.2 kDa, determined by nanoindentation (Figure 3A). If the rods are too soft, the interlinked macroporous structure may not be achieved due to deformation of the building blocks. To detect active functional groups, fluorescein isothiocyanate (FITC) can be used to visualize primary amino groups, and fluorescein amine isomer I can be employed to label epoxy groups (Figure 3B,C). The amine microgel rods have dimensions with an average length of 553 µm ± 29 µm and an average width of 193 µm ± 7 µm in deionized water, resulting in an aspect ratio of ~3.0 and a reduction in volume (collapse) by ~73% of their size in cell culture media¹¹.

Figure 3: Microgel properties. (A) Effective Young's modulus of amine and epoxy microgel rods along with amine and epoxy microgel spheres measured by nanoindentation. Data displayed as a box plot extending from the 25th to 75th percentile, with the whiskers reaching from the 5% to 95% quantiles. The lines inside the boxes represent the medians, the empty squares indicate the means, and the black squares represent outliers (n = 40; p-values are calculated using one-way ANOVA with Bonferroni correction, **p < 0.01, ****p < 0.0001). (B) Top: Confocal microscopy image of an amine microgel rod functionalized with FITC and an epoxy microgel rod functionalized with fluorescein amine isomer I. Bottom: Corresponding bright field images. All scale bars represent 100 µm. Please click here to view a larger version of this figure.

As described in the related publication, sphere-like microgels produced via the same method lead to multiple interlinked clusters rather than one stable macroporous scaffold¹¹. The higher aspect ratio of the microgel rods allows for better overall stability due to more efficient structure bridging in 3D (Figure 4).

Figure 4: Influence of the aspect ratio on the structure formation. (A) Bright field image of an interlinked construct composed of microgel rods. (B) Bright field image of interlinked clusters composed of sphere-like microgels. Scale bars represent 500 µm. Please click here to view a larger version of this figure.

The mean values of the macropores in the scaffolds composed of microgel rods are 100 µm, with 90% of the pore sizes ranging from 30 µm to over 150 µm¹¹. Sphere-like microgels result in clusters with pore sizes between ~10 µm to 55 µm, with a mean value around 22 µm¹¹. This is consistent with the reported numbers by other studies preparing MAPs based on spherical microgels^2,4,14.

One of the critical steps in this protocol is the quality of the 2-aminoethyl methacrylate (AMA) used as the comonomer for primary amine functionalization. The AMA should be a fine-grained and preferably colorless powder delivered in a gas-tight brown glass container. One should avoid using greenish and lumpy material, as it significantly impairs the gelation reaction and negatively affects the reproducibility of the results. In case of poor gelation and unstable microgel rods, one can consider changing the supplier.

If the mixing of amine and epoxy microgel rods leads to multiple interlinked clusters rather than to one stable structure, check the number of rods in each stock dispersion and set it to a similar value in the range of 1,000-5,000 rods/100 µL for both samples. If the dimensions of the microgel differ significantly from the values mentioned here, adjust the number of microgel rods to the volume fraction. Increase the number per volume if the gels are smaller and decrease the number in the opposite case.

The method described did not focus on optimizing the mixing process to have a more controlled assembly of the two complementary functionalized mixing components. Since interlinking occurs within a few seconds, the total volume of the dispersion to form one interlinked construct and the volume fractions of the microgels used are adjusted to obtain stable macroporous structures by simply pipetting the two components one after the other. In the future, it would be advantageous to analyze the flow and mixing properties of the microgel rod dispersions and gain additional insight into the structure formation of the scaffolds. A more controlled mixing of the two interlinking microgel components could enable automated and high-throughput formation of these macroporous scaffolds and allow for incremental construct growth.

Since interlinking proceeds very rapidly, the pores of the scaffold are created during mixing. If the microgel rods were first completely sedimented before chemically interlinking, a much higher stacking to jamming ratio would be expected. Therefore, the interlinking kinetics are likely to have a large effect on MAP formation and the resulting porosity and pore sizes. Cell-adhesive functionalization by post-modification or incorporation into the gel network during microgel preparation has demonstrated that L929 and human fibroblast cells attach and spread on the microgel rods' surface first and, subsequently, fill most of the macropores after 5-7 days in culture¹¹. Due to the pore sizes predominantly ranging from 30 µm to above 150 µm and the interconnected pore structure, confirmed by confocal microscopy, seeded cells can easily enter the interlinked scaffold¹¹. So far, these microgel rod-based scaffolds have been used to grow fibroblast and HUVEC cells in co-culture. HUVECs seeded after 14 days of fibroblast cell culture formed first elongated vessel-like structures within the following 16 days¹¹. The study of other cell types in co-cultures remains to be addressed in the future. To allow cells to be added directly to the system during scaffold formation, fast rod interlinking in cell-compatible buffers or culture media is required, which would enable this system to be extended to bioprinting platforms.

Non-injectable macroporous 3D scaffolds can also be created by various alternative methods like solvent casting, freeze-drying, gas foaming particulate leaching, electrospinning, or 3D printing^1,15,16. Cell migration and proliferation can easily occur without the need to degrade the scaffold beforehand, as long as the required pore geometry and permeability have been achieved throughout the network. 3D printing utilizing jammed, extrudable bioinks from PEG-, agarose-, and norbornene-modified hyaluronic acid-based microgel spheres combines the MAP and 3D printing technologies, controlling the porosity between the annealed microgels and also the geometry of the overall printed structure¹⁷.

Compared to alternative existing methods, the resulting scaffolds exhibit a wide variety of 3D macropore geometries due to the higher aspect ratio of the microgel building blocks and their interlinked two-component configuration¹¹. The utilization of microgel rods creates more porosity and, thus, more space for cell-cell interactions compared to the assembly of spherical microgels. This is central for biological tissue formation as it enhances cell-cell interactions. At the same time, less material can be used to produce macroporous scaffolds of the same volumes, while maintaining the stability of the construct. The reduction of scaffold material is beneficial as more open space is provided for tissue formation, and less material has to be degraded, which is important for both in vitro and in vivo applications. Biofunctionalization of the microgels can be extended to other peptide types and bioactive factors. The resulting stable macroporous scaffolds require less material to create more porosity. Together with the tunability of the system, this method offers high potential as a versatile platform for cell culture in 3D.