Designing degradable hydrogels for orthogonal control of cell microenvironments

Abstract

Degradable and cell-compatible hydrogels can be designed to mimic the physical and biochemical characteristics of native extracellular matrices and provide tunability of degradation rates and related properties under physiological conditions. Hence, such hydrogels are finding widespread application in many bioengineering fields, including controlled bioactive molecule delivery, cell encapsulation for controlled three-dimensional culture, and tissue engineering. Cellular processes, such as adhesion, proliferation, spreading, migration, and differentiation, can be controlled within degradable, cell-compatible hydrogels with temporal tuning of biochemical or biophysical cues, such as growth factor presentation or hydrogel stiffness. However, thoughtful selection of hydrogel base materials, formation chemistries, and degradable moieties is necessary to achieve the appropriate level of property control and desired cellular response. In this review, hydrogel design considerations and materials for hydrogel preparation, ranging from natural polymers to synthetic polymers, are overviewed. Recent advances in chemical and physical methods to crosslink hydrogels are highlighted, as well as recent developments in controlling hydrogel degradation rates and modes of degradation. Special attention is given to spatial or temporal presentation of various biochemical and biophysical cues to modulate cell response in static (i.e., non-degradable) or dynamic (i.e., degradable) microenvironments. This review provides insight into the design of new cell-compatible, degradable hydrogels to understand and modulate cellular processes for various biomedical applications.

Fig. 1. Overview. Degradable hydrogels can be used for orthogonal control of multiple properties in both two- and three-dimensional (2D and 3D) cellular microenvironments.

Fig. 2. Design considerations. The design of hydrogels for orthogonal property control in cellular microenvironments is dictated by the biocompatibility, crosslinking in presence of cells or proteins, mechanical properties, degradability, mass transport properties, and target microenvironment.

Fig. 3. Range of natural and synthetic polymer building blocks. Molecular structures of typical polymer repeat units used for preparation of cell compatible hydrogels: (A) hyaluronic acid, (B) chitosan, (C) heparin, (D) alginate, (E) linear poly(ethylene glycol) (PEG), (F) four-arm PEG, (G) poly(ethylene glycol)-b-poly(propylene oxide)-b-poly(ethylene glycol) (PEG-PPO-PEG), (H) poly(lactic acid-co-glycolic acid)-b-poly(ethylene glycol)-b-poly(lactic acid-co-glycolic acid) (PLGA-PEG-PLGA), and (I) poly(vinyl alcohol).

Fig. 4. Hyaluronic acid hydrogels for controlled release applications. (A) HA/heparin hydrogel particles were synthesised by inverse emulsion polymerization and amount of heparin in hydrogel particle was varied. BMP-2 was subsequently loaded. (B) The addition of heparin to HA hydrogel particles inflenced the in vitro release of BMP-2 from hydrogels with higher heparin conent, with less than 5% of loaded BMP-2 released over 13 days (HA-HPx, x = micrograms of heparin per milligram in hydrogel particles). Reprinted from Xu et al. with permission from Elsevier. Copyright (2011).

Fig. 5. Chitosan-based hydrogels for neural tissue engineering. (A) Schematic of synthesis of methacrylated chitosan. Methacrylated chitosan (0.5 to 2% w/w) hydrogels were crosslinked in the presence of cells by photoinitiated free radical polymerization (Irgacure photoinitiator with 365 nm light). (B) E-18 rat cortical neurons were immobilized within chitosan and agarose (Seaprep®) hydrogels for investigating neuronal survival and differentiation. The cells clumped into groups and displayed extensive neurite outgrowth in chitosan hydrogels (left), as compared to agarose hydrogels (right), indicating enhanced neuron function within the chitosan matrices (scale bar, 50 μm). (C) Neurite outgrowth quantification (p < 0.05). Reprinted from Valmikinathan et al. with permission from The Royal Society of Chemistry. Copyright (2012).

Fig. 6. Chemical functional groups for hydrogel formation. A wide range of functional groups is available for either hydrogel formation or modification post-polymerization. Functional group selection depends on several factors related to the application of interest, including the desired initiation mechanism, the specificity and speed of the reaction, and the stability of the resulting bond under various solution conditions.

Fig. 7. Diels–Alder click reaction for forming degradable hydrogels. (A) Schematic of hydrogel formation using Diels–Alder reaction between furan groups of HA and maleimide groups present on a PEG macromer. (B) Brightfield image of MDA-MB-231 cells (left), which are known to interact with HA via CD 44 receptor. Cells were seeded on HA/PEG hydrogels and after 14 days adopted a flattened or elongated morphology, indicating cell adhesion (scale bar, 20 μm). Cell viability was assessed using a live/dead assay (right, live cells in green, dead cells indicated by arrows) signifying a high level of cell survival (>98%), after 14 days (scale bar, 60 μm). Reprinted from Nimmo et al. with permission from American Chemical Society. Copyright (2011).

Fig. 8. Michael-type addition reaction for hydrogel formation. (A) Schematic of hydrogel formation using the Michael-type addition reaction between vinyl sulfone groups of 4-arm PEG and cysteine residues present on the RLP. (B) Human aortic adventitial fibroblasts (AoAFs) were encapsulated during hydrogel formation and cell viability was evaluated via live/dead staining (fluorescent laser scanning confocal microscopy). AoAFs remained viable throughout the experiment, adopting a spread morphology (scale bar, 200 μm). Image reprinted from McGann et al. with permission from John Wiley and Sons publishing. Copyright (2013).

Fig. 9. Block copolymer assembly for hydrogel formation. Multiblock copolymer based hydrogels have been prepared with coacervate crosslinking by mixing equimolar dilute solution of negatively charged (sulfonate, carboxylate) and positively charged (ammonium, guanidinium) ionic ABA triblocks. Image reprinted from Hunt et al. with permission from John Wiley and Sons publishing. Copyright (2011).

Fig. 10. Degradation strategies for controlling hydrogel-based cell microenvironments. Chemically crosslinked hydrogels can be degraded via cleavage of (A) the polymer backbone, (B) crosslinker or (C) pendant group depending upon the chemistry used for hydrogel formation (choice of polymer, crosslinker, and crosslinking mechanism).

Fig. 11. Selection of labile groups to control degradation rates. Chemically crosslinked hydrogels can be engineered to degrade at a preprogrammed, cell-dictated, or user-defined rate with varying degrees of spatiotemporal control.

Fig. 12. Ultrasound-induced retro [3+2] cycloaddition of an embedded triazole moiety. Triazole bond formation can be reversed by the application of mechanical force, resulting in azide and alkyne functional groups. The generated azide and alkyne moieties of the functionalized poly(methyl acrylate) (PMA) were subsequently ‘clicked’ (Cu, CH₃CN) to form the triazole-based starting material. Image reprinted from Brantley et al. with permission from Nature publishing group. Copyright (2011).

Fig. 13. Selective cell release via photodegradation using differences in reactivity of o-nitrobenzyl groups. (A), (B) o-Nitrobenzyl linkers with different degradation kinetics were used to vary the degradation rate of adjacent hydrogels. (C) RFP-expressing hMSCs and GFP-expressing hMSCs were encapsulated within hydrogels made with (A) or (B), where the two hydrogels were in direct contact with each other (RFP = red fluorescent protein, GFP = green fluorescent protein). (D) The interface between hydrogels containing RFP- and GFP-expressing hMSCs was observed using optical microscopy. (E) Gels were exposed to light (10 mW cm^–2 at 365 nm, 30 minute total duration), resulting in a biased release of one cell population over another (R _GFP/R _RFP ∼ 2.4) which was consistence with the degradation rate constants of (A) and (B) (k _{app A}/k _{app B} ∼ 2.5). Image reprinted from Griffin et al. with permission from American Chemical Society. Copyright (2012).

Fig. 14. Spatiotemporal control of biochemical cues in 3D microenvironment. (A) Four-arm PEG functionalized with cyclooctyne was reacted with azide di-functionalized polypeptides via SPAAC reaction to form a hydrogel network via step-growth mechanism. A light-mediated thiol–ene reaction (cytocompatible 490–650 nm or 860 nm pulsed laser light) was used to immobilize cell adhesive thiol-functionalized peptides (RGD) using vinyl functionalities present on hydrogel network. Further, 3-D channels were degraded within the hydrogel using pulsed laser light (740 nm) via irreversible cleavage of nitrobenzyl ether moiety. (B) A cell-laden (3T3 fibroblasts) fibrin clot was encapsulated in the hydrogel (3D microenvironment). Biochemical (channel containing RGD noted by dashed polygon) and biophysical cues (photodegraded channel) were added to control 3T3 cell outgrowth in the presence of encapsulated hMSCs (right, top-down projection; left, 3D rendering). (Scale bar, 100 μm, hydrogel shown in red, F-actin in green and cell nuclei in blue) Image reprinted from DeForest et al. with permission from Nature publishing group. Copyright (2011).

Fig. 15. Modulation of cell adhesion and spreading in 2D microenvironment. (A) The hydrogels were prepared by copolymerizing acrylamide (Am) with acryloyl amino acid (AA) using a bis-acrylamide initiator. Depending upon the number of CH₂ groups on the AA pendant chain (n = 1 to 10, referred as C1, C2,…,C10), the interfacial hydrophobicity of the hydrogel varied, with water contact angle ranging from 26° to 85° (sessile drop method, 20 °C). (B), (C) Non-monotonic dependence on the monomer side chain length was observed in cell adhesion and spreading of hMSCs on C1–C10 hydrogels (scale bar, 400 μm). Image reprinted from Ayala et al. with permission from Elsevier. Copyright (2011).

Fig. 16. Dual growth factor delivery by sequestration in controlled cell microenvironments. (A) Human endothelial cells from the umbilical cord vein (HUVECs) on RGD-modified hydrogel substrates were presented with varying amounts of basic fibroblast growth factor (FGF-2) and vascular endothelial growth factor (VEGF) via sequestration. The enhanced cell survival and typical spindle-like morphology on the hydrogel substrate comprising the combination of FGF-2 and VEGF highlights the synergistic activity of both growth factors (fluorescence microscopy images after live/dead staining). (B) Further, HUVEC proliferation was enhanced by the dual presentation (MTT assay, day 3). (C) These hydrogels were placed onto the developing chicken embryo chorioallantoic membrane (CAM) from embryonic day 8 until day 12 to study the effect of growth factors on vascularization. An increased number of vessels within the site of gel transplantation was observed; (D) representative images indicate substantial angiogenic response to combined FGF-2 and VEGF delivery. Reprinted from Zieris et al. with permission from Elsevier. Copyright (2011).

Fig. 17. Hydrogels for delivering cargo biomolecules (therapeutic proteins, growth factors, and drugs) with pre-programmed or user-defined release behavior. (A) A degradable silk fibroin hydrogel for pre-programmed release was formed in the presence of BMP-2, using a sonication-induced gelation process, and was injected into a bone defect. (B) Enhanced bone formation was observed in vivo in SASCO Sprague Dawley rats injected with the silk hydrogel containing BMP-2 as compared to the control group (no growth factor) (week 12, X-ray radiography). A, B reprinted from Diab et al. with permission from Elsevier. Copyright (2011). (C) Photodegradable hydrogel microparticles were synthesized by an inverse suspension polymerization of a PEG-diphotodegradable acrylate with a PEG tetrathiol (PEG4SH) via base-catalyzed Michael addition. (D) A model cargo protein (Annexin V) was loaded (right), and its release to 3T3 cells (left) was triggered by cytocompatible irradiation (1 min of 13.5 mW cm^–2 at 365 nm). C, D reprinted from Tibbitt et al. with permission from John Wiley and Sons publishing. Copyright (2012).

Fig. 18. Modulation of substrate stiffness in a 2D dynamic microenvironment. (A) Incorporation of a photodegradable nitrobenzyl moiety in a hydrogel network, prepared using polyacrylamide acryl hydrate (PAAH) and a 4-bromomethyl-3-nitrobenzoic acid (BNBA) crosslinker, enabled gel degradation and softening using cytocompatible doses of UV light (365 nm). (B) Softening of the posterior substratum (rear) did not significantly alter cell spreading (left column). However, softening of the anterior substratum (front) led to reverse polarity (central column) or trapping in the softened region (right column; scale bar, 20 μm). Reprinted from Frey et al. with permission from The Royal Society of Chemistry publishing group. Copyright (2009).

Fig. 19. Stress gradients within hydrogels influence cell differentiation in three dimensions. (A) Schematics of the process for creating three-dimensional multicellular hydrogels and encapsulating human mesenchymal stem cells (hMSCs). Briefly, prepolymer type I collagen was added to PDMS molds and hMSCs were suspended, before polymerizing at 37 °C. Liquid agarose was added to the mold to encase the collagen hydrogel at 4 °C. (B) Phase image of hMSCs in three-dimensional structures at day 0. (C) The hydrogel constructs with encapsulated hMSCs were suspended in mixed media and after 14 days, the cells at the edge of the constructs differentiated down an osteogenic lineage (blue) and those at the center underwent adipogenesis (red) (oil droplets, alkaline phosphatase staining) (D) Longitudinal section and (E) cross-sections confirmed the patterning of lineage specification in a tension-dependent manner (scale bar, 250 μm). Reprinted from Ruiz et al. with permission from John Wiley and Sons publishing. Copyright (2008).

Fig. 20. Spatiotemporal manipulation of biophysical cues in 3D cell microenvironments. (A) Hydrogels were prepared by reaction of an acrylated PEG macromer with thiol-functionalized, degradable and cell-adhesive peptides using a Michael-type addition reaction (–UV hydrogel, degradable peptide crosslinks), in the presence of an inactive photoinitiator. Using photolithography (4 min with 10 mW cm^–2 at 365 nm), remaining acrylate groups on the PEG macromers in select regions were reacted by photoinitiated free radical polymerization, forming non-degradable covalent crosslinks (+UV hydrogel) (confocal microscopy images show top and bottom surface photopatterned with 250 μm stripes). (B) Encapsulated hMSCs (day 14 stained with calcein) spread only in –UV regions, where the degradable peptide was used as a crosslinker (scale bar, 100 μm). Reprinted from Khetan et al. with permission from Elsevier. Copyright (2010).

Prathamesh M. Kharkar

Kristi L. Kiick

April M. Kloxin