Tunable Bicontinuous Macroporous Cell Culture Scaffolds via Kinetically Controlled Phase Separation

Abstract

3D scaffolds enable biological investigations with a more natural cell conformation. However, the porosity of synthetic hydrogels is often limited to the nanometer scale, which confines the movement of 3D encapsulated cells and restricts dynamic cell processes. Precise control of hydrogel porosity across length scales remains a challenge and the development of porous materials that allow cell infiltration, spreading, and migration in a manner more similar to natural ECM environments is desirable. Here, a straightforward and reliable method is presented for generating kinetically-controlled macroporous biomaterials using liquid-liquid phase separation between poly(ethylene glycol) (PEG) and dextran. Photopolymerization-induced phase separation resulted in macroporous hydrogels with tunable pore size. Varying light intensity and hydrogel composition controlled polymerization kinetics, time to percolation, and complete gelation, which defined the average pore diameter (Ø = 1-200 µm) and final gel stiffness of the formed hydrogels. Critically, for biological applications, macroporous hydrogels are prepared from aqueous polymer solutions at physiological pH and temperature using visible light, allowing for direct cell encapsulation. Human dermal fibroblasts in a range of macroporous gels are encapsulated with different pore sizes. Porosity improved cell spreading with respect to bulk gels and allowed migration in the porous biomaterials.

Keywords: Biomaterials; Phase separation; Porosity; hydrogels; kinetic control.

Figure 1

Composition controls phase‐separation in polysaccharide containing, photopolymerized thiol–ene hydrogels. a) Simplified schematic of a phase diagram for liquid–liquid phase separation. Phase separation in aqueous solutions of PEG polymers (φ, red) and polysaccharides (green) can be induced through polymerization of the macromolecular components. Phase separation can proceed via binodal or spinodal decomposition mechanisms depending on the mixture composition (φ) and temperature (T). b) Macromolecular components of precursor solution for macroporous system include hydrogel phase: 8‐arm 20 kDa PEG‐norbornene, di‐thiol (dithiothreitol; PEG‐SH; thiolated peptides), and excluded phase: dextran and hyaluronic acid. c) Irradiation of hydrogel forming solution (λ = 405 nm, I = 0.5 mW cm⁻²) triggered thiol–ene polymerization and phase‐separation into a bicontinuous network via spinodal decomposition that was arrested by gelation. Hydrogel network phase (red) and polysaccharide phase (black). Scale bar, 100 µm. d) Representative images of phase‐separated hydrogels (red) perfused with dextran‐FITC (green), demonstrating the formation of a bicontinuous and perfusable network. Scale bar, 100 µm. e) Representative images of the porosity in 3 wt.% PEG hydrogels generated at different dextran concentrations (0.5–2.5 wt.%; I = 1.0 mW cm⁻²). Scale bar, 50 µm. f) Representative images of the porosity in 3 wt.% PEG hydrogels generated at different HA concentrations (0.5–3.0 wt.%, I = 1.0 mW cm⁻²). Scale bar, 50 µm. g) Pore size in 3 wt.% PEG hydrogels as a function of dextran concentration (0.0–3.0 wt.%; I = 1.0 mW cm⁻²; n = 3). Experimental values are represented as means ± SD (n = 3). h) Pore size in 3 wt.% PEG hydrogels as a function of interpolated solution viscosity (0.0–3.0 wt.% HA; I = 1.0 mW cm⁻²; Fitted curve: Pore size [µm] = 562 * Viscosity[mPa*s]⁻¹]). Experimental values represent single measurements (n = 3).

Figure 2

Kinetic control of pore size in macroporous hydrogels. a) In phase‐separating, PEG–dextran hydrogels the crossover time, τ _c , decreased with increasing irradiation intensity, with a scaling based on a power law fit of τ _c ∼ I ^−0.85 or R_p ∼ I ^0.85. Experimental values are represented as single data points (n = 2). b) Pore size in macroporous hydrogels decreased with increasing polymerization intensity, as described by the physical model. Experimental values are represented as mean ± SD (n = 3). The solid line represents the scaling law from the physical model (Equation 3). c) Pore evolution in phase‐separating PEG–dextran hydrogels at low (0.5 mW cm²), medium (1.0 mW cm⁻²), and high (10.0 mW cm⁻²) light intensities. The light irradiation started at t = 0 s. d) Rhodamine‐labeled gel phase (red) and dextran‐FITC (M _n ∼ 500 kDa) in the pore phase (green) in gels made at increasing light intensity from left to right. Scale bar,100 µm. e) DLP‐patterned hydrogel with stripes of low to high intensity, from left to right, generated pores of different sizes as seen in the images of each stripe region and the interface regions. The gel phase was labeled with rhodamine and imaged with confocal microscopy. Scale bars: 500 µm in the overview and 100 µm in insets.

Figure 3

Cell spreading and migration in macroporous hydrogels. a) Macroporous hydrogels facilitated spreading of the encapsulated hDFs as compared with nanoporous gels. Transmitted light images (above) show hydrogel porosity in PEG–MMP hydrogels and fluorescence images (below) show the encapsulated cells 7 days after encapsulation in the corresponding gels with labeled actin cytoskeleton (green), showing varying degrees of cell spreading depending on pore size (nano Ø ∼ 10 nm, small Ø = 6.0 ± 1.0 µm, medium Ø = 15.0 ± 1.5 µm, and large Ø = 45.0 ± 5.6 µm pores). Scale bar, 100 µm. b) Fluorescently labeled cells in macroporous PEG–MMP hydrogels with medium porosity (Ø = 15 ± 1.48 µm). Scale bar, 50 µm. c) Number of cell–cell connection in nanoporous and macroporous gels after 7 days (AngioTool). d) Percentage of cell area in the hydrogels for both nanoporous and macroporous hydrogels (AngioTool). e) Average hDF length in both nanoporous and macroporous gels (AngioTool). Trajectories of hDF migration in macroporous hydrogels with small pores (f), medium pores (g), and large pores (h). i) Distances (i.e., total length of the traveled path or path length) of hDF migration in macroporous hydrogels increased with increasing pore size. j) Velocities of hDF migration in macroporous hydrogels with small, medium, and large porosities. k) Fluorescence images of cells in nanoporous and macroporous hydrogels with increasing pore sizes. The morphology of the spread cells and the expression of alpha smooth muscle actin (αSMA) varied between different porosity hydrogels. DAPI (blue, actin (green), αSMA (magenta). Scale bar, 100 µm. l) Expression of αSMA in the nanoporous and macroporous hydrogels. m) Cell spreading and growth within macroporous gels with large porosity after 11 days in culture. DAPI (blue), actin (green), αSMA (magenta). Scale bar, 50 µm.