Development of an in situ injectable hydrogel containing hyaluronic acid for neural regeneration

In this work, a novel enzymatically crosslinked injectable hydrogel comprising hyaluronic acid (HyA), dopamine (DA), and 3-(4-hydroxyphenyl) propionic acid (HPA) conjugates was successfully developed. To the best of our knowledge, it is the first time that HPA is conjugated to a HyA-based backbone. In situ hydrogelation of HyA-DA-HPA occurred in the presence of hydrogen peroxide (H2O2) as an oxidant and horseradish peroxidase (HRP) as a catalyst. Proton nuclear magnetic resonance and Fourier transform infrared spectroscopy were used to characterize the chemical reactions between HyA, DA, and HPA. Gel formation completed between 3 s to 5 min depending on the concentrations of polymer, HRP, and H2O2. Crosslinked HyA-DA-HPA gels acquired storage moduli ranging from ∼100 Pa to ∼20 000 Pa (at f = 2000 rad s−1). Biocompatibility of the hydrogels was examined with human mesenchymal stem cells (hMSCs) and human induced pluripotent stem cell-derived neural stem cells. The hydrogels made of 2.0 w/v% HyA-DA-HPA hydrogels, 0.24 U ml−1 HRP and ≤ 0.5 µmol ml−1 H2O2 were found biocompatible with hMSCs cultured on and encapsulated within the hydrogels. Since HyA serves as a backbone of the extracellular matrix in the central nervous system (CNS) and DA acquires the ability to restore dopaminergic neurons, use of this injectable HyA-DA-HPA hydrogel for stem cell transplantation is a potential treatment strategy for CNS repair and regeneration.


Introduction
Hydrogels are one of the most promising polymeric biomaterials in tissue engineering for delivering cells, growth factors, peptides and drugs that aid tissue regeneration [1][2][3][4]. They facilitate efficient encapsulation and delivery of cells and/or soluble factors as they can be tailored to provide high oxygen permeability, supply nutrients and efficiently channel watersoluble metabolites through their hydrated matrix [4,5]. One of the greatest challenges for its application in neural tissue engineering is the effective entrapment and delivery of cells and growth factors to the target site. The soft nature of the tissue, its extremely limited regenerative capability, as well as the lack of nutrient and oxygen perfusion in the wound sites could hinder the delivery and growth of cells [6]. In this regard, in situ injectable hydrogels, which can be injected as a liquid and solidify in situ to form a substrate with ideal mechanical properties and the capability to support stem cell transplants, have been explored for neural repair [7]. Injectable hydrogels also allow a minimally invasive procedure, which may offer further benefits, including smaller incisions, reduced pain and scarring and short recovery times [8].
To date, a variety of natural and synthetic polymers, including alginate, collagen, and poly(lacticco-glycolic acid) (PLGA), as well as decellularized extracellular matrix (ECM), have been utilized as supporting grafts to facilitate stem cell transplantation (reviewed in Chan et al 2017, Tuladhar et al 2018, andNih et al 2016) [7,9,10]. For example, Ford et al achieved the generation of a functional and stable microvascular system by delivering co-cultures of neuronal progenitor cells and endothelial cells through a macro-porous polyethylene glycol (PEG)based hydrogel system in vivo [11]. A hydrogel made of self-assembling peptides has also been shown to support differentiation and maturation of human neural stem cells (NSCs) in 3D serum-free conditions in vitro and demonstrated its functional regenerative potential in rat spinal cord injuries [12].
Hyaluronic acid (HyA) is a linear polysaccharide serving as a major component in the central nervous system (CNS) ECM [13]. It binds to proteoglycans which further bind to tenascins, resulting in the structural backbone of the CNS [14]. Although HyA can be extracted from animal tissues, xeno-free HyA can be produced via bacterial fermentation [15]. In regard to its biocompatibility, its regulatory role in inflammation and glial scar formation, as well as the advantageous chemical tunability, HyA has become one of the most suitable materials for neural tissue engineering [16][17][18]. Previously, mechanically tunable HyA scaffolds for delivery of growth factors and cells have been shown to mediate and support survival and differentiation of neural cells both in vitro and in vivo [14,19,20]. To enhance cell-matrix interaction, adhesive molecules (i.e. collagen and RGD peptide) and heparin could be additionally functionalized into the HyA hydrogels to promote cell adhesion and sequestration of soluble factors [21].
Dopamine (DA) is one of the major neurotransmitters in the brain, playing a key role in brain aging and neurodegenerative diseases, particularly, Parkinson's disease [22]. In PC12 neuronal cells, it was reported that DA pre-treatment at a noncytotoxic concentration may serve as a potent inducer of cellular glutathione and NAD(P)H:quinone oxidoreductase 1, exerting neuroprotective effects [23]. Previously, the development of injectable hydrogel systems for local delivery of DA, DA-loaded microspheres, and DA-secreting cells has been reported [21,[24][25][26][27]. A hydrogel system made of chitosan, gelatin, and DA was shown to possess a sustained DA release profile without a burst release over the course of 500 h [25]. Implantation of a dextran dialdehyde crosslinked gelatin hydrogel loaded with DA could improve functional recovery in a 6-hydroxydopamine (6-OHDA)-rat hemiparkinsonian model, suggesting the potential application of DA-embedded hydrogel for treating Parkinson's disease [26]. Although there are a number of publications focusing on hydrogel systems for DA release, to the best of our knowledge, a hydrogel incorporating DA into its polymer backbone has never been reported.
In this study, we fabricated a novel hydrogel using HyA, 3-4-hydroxyphenylpropionic acid (HPA) and DA. While HyA is a strong hydrophilic anionic polysaccharide containing a large number of hydroxyl and carboxyl groups as hydrogen bond donors or acceptors, it can facilitate hydrogen bond formation and hydrogelation [28]. Previously, gelatinbased hydrogels formed by conjugating HPA into its backbone were found with higher adhesiveness compared to that of commercial fibrin glue [29].
It was suggested that a higher phenolic content in the polymer may result in higher cohesive and adhesive strength [29,30]. In this study, we aim to design an injectable hydrogel system that is able to improve cellular incorporation and tissue attachment, as well as neuroprotective effects by incorporating both HPA and DA into our system, which has never been reported previously. We used an enzymatic chemical crosslinking process that includes horseradish peroxidase (HRP) and hydrogen peroxide (H 2 O 2 ) to fabricate hydrogels of HyA, DA, and HPA. This crosslinking approach has been used to form fast gelling, non-toxic and stable hydrogels [31].
Our novel hydrogel was then tested for its biocompatibility. This hydrogel system based on HyA, the structural backbone of the CNS, and incorporating DA, which triggers some neuroprotective effects in the brain, can provide an alternative platform for delivering cells and growth factors to the CNS, particularly the brain, facilitating neural repair and regeneration.  To prepare a series of HyA-DA conjugates with catechol groups in various ratios, HyA was reacted with different concentrations of DA: 0.191 g (0.5 mmol), 0.574 g (1.5 mmol), 1.148 g (3.0 mmol), and 1.914 g (5.0 mmol), respectively. The reaction mixtures were stirred at room temperature for 4 h while maintaining pH 5.5. The solution was purified by dialysis for 3 d against acidified deionized water (pH 5.5) to inhibit oxidation of catechol groups and 1 d against neutral water for pH adjustment. The dialysis solution was refreshed every day.   HyA-DA-HPA gels were obtained by peroxidasecatalyzed gelation at 37 • C.

Characterization of polymer 2.3.1. Fourier transform infrared (FTIR) spectrometers
FTIR can determine the structure of molecules by measuring the molecule's characteristic absorption of infrared radiation [32]. After finishing the dialysis step, the HyA-DA-HPA polymer solutions were freeze-dried to obtain the cotton-like solids and these materials were then used for FTIR analysis. The successful synthesis of HyA-HPA and HyA-HPA-DA was determined by analyzing the spectra obtained by FTIR spectra (Tensor 27, Bruker, USA) using the OPUS Data Collection Program. Highly sensitive liquid nitrogen cooled MCT (mercury-cadmium-telluride) detector was used.

Scanning electron microscope (SEM)
Lyophilized samples were mounted on holders. The surface morphologies of the HyA-DA-HPA before and after gel formation were observed by SEM (Carl Zeiss Evo LS15 VP-SE, BSE, VPSE, EPSE detectors, Zeiss, Germany) at an accelerating voltage of 10 kV. Before the SEM investigation, the samples were coated with gold by SC7620 Mini Sputter (Quorum Technologies, UK).

Rheological measurement
Rheological measurements of the hydrogel formation were performed with a Diffusing Wave Spectroscopy system (LS instruments, CH). Hydrogel solutions with 200 nm diameter polystyrene tracer particles (10% w/v, Sigma-Aldrich) were used to determine how the gel network interaction modulates the tracer particles' thermal motion. The measurements were taken at 37 • C using 2 mm glass cuvettes in thickness (LS Instruments, CH). Solutions of HRP (0.24 U ml −1 ) and H 2 O 2 (0.25 µmol ml −1 ) were added to an aqueous solution of HyA-DA-HPA (0.5%, 1.0%, 1.5%, 1.75% and 2% (w/v)), prewarmed at 37 • C. The solution was vortexed and then immediately transferred into the cuvettes. The cuvettes were sealed with a lid and placed into the rheology system.

Gelation time
The HyA-DA-HPA hydrogels (400 µl) were prepared in 0.01 M PBS (pH 7.4) in 1 ml vials at room temperature. The polymer solution, the HRP, and H 2 O 2 solution were pre-heated at 37 • C before mixing and loading into the double syringe. Briefly, the polymer solution (200 µl, 4.0 w/v%) was mixed with HRP solution (100 µl of 0.96 U ml −1 of stock solution) and H 2 O 2 (100 µl of 4.0 µmol ml −1 of stock solution) by gentle shaking. The time taken for gel formation (gelation time) was determined using a test tube inversion method [33][34][35]. The gelation time was recorded as the time when the gel mass stopped flowing after inversion. The gelation time was measured at different concentrations of HRP, H 2 O 2 , and polymer.

Cell culture 2.4.1. Human GFP-MSC cells
Green fluorescence protein (GFP) was cloned into human mesenchymal stem cells (MSC, kindly provided from the Department of Pediatrics and Adolescent Medicine, LKS Faculty of Medicine, The University of Hong Kong). Briefly, primary mesenchymal cells obtained from unfractionated bone marrow mononuclear cells of a healthy donor were cultured for 2 months. Cells were infected with a VSV-G (expressing the G glycoprotein of the vesicular stomatitis virus) pseudotyped retroviral vector that contained the hTERT and GFP genes, separated by an internal ribosome entry site (IRES), under the control of the murine stem cell virus (MSCV) longterminal repeat (LTR). The GFP + and GFP -MSCs were then separated with a fluorescence-activated cell sorter (MoFlo, Cytomation, Fort Collins, CO, USA) [36]. MSC-GFP were cultured in Dulbecco's modified Eagle's medium (DMEM 1.0 mg l −1 of glucose, Gibco BRL, Gaithersburg, MD, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, Gibco BRL) and 0.1% (v/v) penicillin-streptomycin (PS, Gibco BRL).

Cellular responses of MSCs towards
HyA-DA-HPA hydrogels Hydrogels (400 µl) made with 2% (w/v) HyA-DA-HPA final concentration, 0.5 µmol ml −1 H 2 O 2 and 0.24 U ml −1 HRP (prepared in cell culture medium) were mixed with MSCs and injected into a 48-well culture plate and incubated for 1, 2 and 7 d. GFP expressing MSCs encapsulated within the hydrogels were imaged using a cooled CCD camera (EXI blue; QImaging, UK) attached to a Nikon Eclipse Ti-E inverted fluorescence microscope.

Cellular responses of human iPSC-derived NSCs
towards HyA-DA-HPA hydrogels Hydrogels (400 µl) made with 2% (w/v) HyA-DA-HPA final concentration, 0.5 µmol ml −1 H 2 O 2 and 0.24 U ml −1 HRP were injected into a 48-well culture plate. Human iPSC-derived NSCs were seeded on top of the prepared hydrogels and control substrates (tissue culture polystyrene (TCP) plates) with a cell density of 100 k cells cm −2 in Neural Basal Medium supplemented with 10 µg ml −1 laminin and 10 ng ml −1 BDNF (PeproTech). Cell viability was evaluated using Calcein AM (Abcam), which stained viable cells with green fluorescence through the reaction of Calcein AM with intracellular esterase, and DRAQ5™ (Abcam), which is a cell permeable far-red fluorescent DNA dye staining cell nuclei. The viability was determined by the percentage of viable cells to total cell nuclei on day 1 and day 3.

Statistical analysis
The MSC experiments were conducted with 4 independent replicates and the human iPSC-derived NSC experiments were conducted with 3 independent experiments and 9 replicates in total. Statistical analysis of the MSC experiments was performed with one-way analysis of variance (ANOVA) with Tukey's honest significant difference post hoc test and the statistical analysis of the human iPSC-derived NSC experiments was performed with two-sample t-tests. A value of p < 0.05 was considered statistically significant.

Formation of HyA-DA-HPA polymer
In the functionalized HyA hydrogels, HPA is used to aid both the attachment of incorporated cells as well as the attachment of the scaffold to the site so that the scaffold is less likely to become detached after transplantation [29,30]. DA, on the other hand, has been used to treat Parkinson's disease patients or to reduce the death of dopaminergic neurons, such that this hydrogel may function to aid neural viability [22,23]. To combine these three compounds together, the synthesis of HyA-DA-HPA was achieved via a two-step reaction ( figure 1). Firstly, HyA-DA was conjugated from HyA and DA through a general carbodiimide/active ester-mediated coupling reaction with different ratios of catechol groups. After purification of the HyA-DA conjugates by dialysis, instead of lyophilizing to obtain HyA-DA conjugates, HPA was introduced to the purified HyA-DA conjugate solution for synthesizing the HyA-DA-HPA conjugate. This reaction was carried out between amine groups of HyA and succinimide-activated HPA. Finally, a freeze-drying step was carried out to obtain the lyophilized HyA-DA-HPA polymer.

Characterization of HyA-DA-HPA polymer
Functional groups and chemical bonds of HyA-DA conjugates were identified by FTIR, shown as supplementary figure S1 (available online at stacks.iop.org/BMM/15/055005/mmedia). A broadband centered at 3300 cm −1 was assigned to the vibration of the hydroxyl groups of HyA. The absorption of HyA-DA conjugates exhibited additional peaks consistent with the main vibration modes of DA, such as the C-H vibration at 2850 cm −1 and the C = C ring stretching at 1520 cm −1 . The absorption peaks located at 1720, 1630 and 1410 cm −1 were attributed to C = O, amide bond, and C-O stretching vibrations, respectively [40].
It was found that the FTIR spectrum of HyA-DA-HPA showed peaks at 1750 cm −1 (C = O stretching) and 1200 cm −1 (C-N stretching) (figure 2). These results correspond to the expected structure of the chemical composition, suggesting that DA successfully conjugated to HyA via amide bond and the HyA-DA conjugated to HPA via amine groups of the HyA and succinimide ester active groups of the HPA. The chemical structure of HyA-DA conjugates with various concentrations of DA was characterized using 1 H NMR, as shown in figure S2. The N-acetyl peak of HyA backbone appeared at around 2 ppm. Multiples from 2.8 to 3.7 ppm are associated with the disaccharide unit and anomeric protons composed of dglucuronic acid and N-acetyl-d-glucosamine in the HyA while the position between 6.7 and 7.2 ppm corresponds to the catechol ring of DA. Moreover, the area under this peak increased with an increasing DA feed ratio [25]. Figure 3(a) shows the 1 H NMR spectrum of HyA. The peak at δ 1.9 (number 1) ppm and the group peaks around δ 3.0-4.0 (number 2) ppm and δ 4.8 (number 3) ppm correspond to acetamidemethyl protons and protons from sugar ring and anomatic protons, respectively. The 1 H NMR spectrum of DA is shown in figure 3(b). The chemical shift assignments are based on the literature values [41,42]. The lower-field portion of the spectrum (right) contains 2 triplets (number 1, 2) corresponding to the two methylene groups on the ethylamine portion of the molecule. The high field region (left) shows the aromatic protons (numbers 3, 4 and 5). The typical 1 H NMR spectra of HPA are shown in figure 3(c).  Pores in the HyA-DA-HPA gel can serve as communication and transportation channels for cells and nutrients [6]. The estimated pore size based on the images ranges from 50 to 300 µm. HyA-DA-HPA hydrogel formation is shown in figure 4(c). The HyA-DA-HPA hydrogel formation occurred via the HRPmediated coupling of the introduced phenol moieties. Coupling of phenols could take place either via a carbon-carbon bond at the ortho positions or via a carbon-oxygen bond between the carbon atom at the ortho and the phenoxy oxygen positions [43]. This is similar to the crosslinking of hyaluronic acid-tyramine (HyA-Tyr) conjugates via the oxidative coupling of Tyr moieties using HRP and H 2 O 2 . HRP has been frequently used as a catalyst for the oxidative coupling of phenol derivatives under mild reaction conditions [31]. In this case, the oxidative coupling of phenol proceeded at the C-C and C-O positions between phenols [44,45].

Rheological measurement
The mechanical properties of the hydrogel were characterized at 37 • C using rheological measurement at various concentrations of HyA-DA-HPA: 0.5%, 1.0%, 1.5%, 1.75% and 2.0% (w/v) (figure 5). Higher concentrations of HyA-DA-HPA had higher storage moduli (G'), resulting in higher mechanical stability. Crosslinked HyA-DA-HPA gels exhibited storage moduli ranging from~100 Pa to~20 000 Pa (at the frequency of 2000 rad s −1 ) which are typical of a soft hydrogel and similar to native soft tissues [46]. Previously, with magnetic resonance elastography, the storage moduli (G') of the grey matter and white matter in the adult human brain are~3.1 kPa and 2.7 kPa, respectively [47]. Substrate elasticity also affects behaviors and fates of cells; for example, cell proliferation and differentiation of NSCs cultured in 3D alginate hydrogels varied with different elastic moduli, where the softest hydrogel with a modulus of~180 Pa increased cell proliferation and resulted in the greatest enhancement in neuronal differentiation [48]. Adult NSCs also exhibited a higher neuronal differentiation on soft gels (E = ∼0.1-0.5 kPa) and an enhanced glial differentiation on stiffer gels (E = ∼1-10 kPa) [49]. Our results showed that as the concentration of the HyA-DA-HPA increased, the storage moduli increased due to the enhanced crosslinking. Hydrogels with storage moduli <500 Pa were excluded from the selection as these soft gels were too difficult to handle and they degraded within a few days even in in vitro cell cultures. Thus, HyA-DA-HPA with a concentration higher than 1.5 w/v% was preferred to use for in vitro studies.

Effect of HRP, H 2 O 2 , and polymer concentrations on gelation time
The gelation time was shown to be controllable by changing the concentration of HRP, H 2 O 2 , and polymer ( figure 6). The gelation time decreased from 186 to 11 s as the HRP concentration increased from 0.05 to 0.24 U ml −1 at a constant H 2 O 2 concentration (1.0 µmol ml −1 ) and polymer concentration (2.0 w/v%) ( figure 6(a)). HRP allows the phenol substrate in the HyA-DA-HPA conjugate to couple each other via the carbon-carbon bond at the ortho position or via a carbon-oxygen bond between the carbon at the ortho position and the phenoxy oxygen in an aqueous solution [33,50]. Thus, gelation time decreased as HRP concentration increased since the rate of creating phenoxy radicals increased in the coupling reaction. In contrast, as H 2 O 2 concentration increased from 0.15 to 1.0 µmol ml −1 , the gelation time slightly increased from 6.2 to 12.8 s ( figure 6(b)). In the case of 0.15 µmol ml −1 H 2 O 2 , a longer gelation time of the hydrogel occurred compared to 0.25 µmol ml −1 . This result was potentially due to the fact that a small number of phenoxyl radicals were produced at a lower concentration of the catalyst, resulting in a delay of the crosslinking reaction. From 0.25 µmol ml −1 H 2 O 2 to 1.0 µmol ml −1 H 2 O 2 , the gelation time increased due to the generation of an inactivated form of HRP caused by excessive H 2 O 2 . Faster gelation time was observed for higher polymer concentrations as shown in figure 6(c). This can be explained by a higher concentration of reactive phenol groups available with a higher polymer concentration and it increased the overall crosslinking reaction rate, facilitating gel formation.

Effect of H 2 O 2 concentration on in vitro cell viability of MSCs
We selected MSCs as our cell model to test the biocompatibility of the hydrogels. The effect of H 2 O 2 concentration on cell viability of MSCs was evaluated for finding the highest acceptable concentration of crosslinking reaction initiator. To achieve liable     (b2)). However, on day 7 (figure 7(b3)), the cells performed spread morphology. It is possible that on day 1 and day 2, cells remained rounded within the hydrogel while they were embedded in a newly formed dense polymer network with low porosity. After days of culture, viable and functional MSCs degraded the matrix, attached and bound to the hydrogel scaffold, presenting a spread and elongated morphology. These results suggested that the cells could survive and that there was no cytotoxicity in the hydrogel matrix after 7 d. It also demonstrated our hydrogel platform as a promising material for MSC delivery since it can not only support the growth of cells encapsulated within but also display remodeling properties upon interaction with cells. This remodeling process can facilitate the integration of hydrogel with the ECM secreted by the encapsulated cells and further bridge with the natural ECM in the transplanted sites, resulting in improved tissue adhesion and host integration [51].

Cellular responses of human iPSC-derived
NSCs on HyA-DA-HPA hydrogels Although MSCs have been shown to differentiate into neural lineages upon transplantation, their mechanisms could rely more on the trophic effects for neurogenesis, angiogenesis, and inflammatory and immune responses [52]. For example, transplantation of MSCs can increase expression levels of BDNF, insulin-like growth factor 1 (IGF-1), basic fibroblast growth factor (bFGF), vascular endothelial growth factor (VEGF), epidermal growth factor (EGF), and interleukin-10 (IL-10), etc, through direct secretion or activation of endogenous host cells to mediate cell apoptosis, survival, angiogenesis, and anti-inflammation to improve functional recovery [53][54][55][56][57]. Human iPSC-derived NSCs, which can bypass certain ethical issues compared to embryonic stem cell-derived progeny and can potentially lower the risk of immune rejection, have been reported as an alternative to treat brain injuries [58]. Compared to MSCs' trophic effects, these cells facilitate neuroregeneration and repair via neuronal protection, reducing microglia activation, and most importantly, long-term integration of differentiated neurons and glial cells [59][60][61][62]. We selected HyA-DA-HPA hydrogels (2.0 w/v%) made with 0.24 U ml −1 HRP and 0.5 µmol ml −1 H 2 O 2 to assess its biocompatibility with human iPSC-derived NSCs, which express typical markers of NSCs (e.g. Pax6 and Nestin) and acquire the capability to differentiate into βIII-Tubulin positive neurons (figures 8(a) and (b)). As the NSC culture media is serum free, we added 10 µg ml −1 laminin and 10 ng ml −1 BDNF in the media to promote cell binding to the substrate and neuronal differentiation, respectively. On both day 1 and day 3, human iPSC-derived NSCs on the hydrogels exhibited a round shape morphology while the cells on the TCP control were flattened and spread with typical neurite morphology (figures 8(c) and (d)). The cells cultured on the 2D hydrogel substrate exhibited significantly lower cell viability, 51.3 ± 4.6% and 33.3 ± 3.7% on day 1 and day 3, respectively, compared to the TCP control, which is 79.2 ± 1.8% and 64.0 ± 4.3% on day 1 and day 3, respectively (figures 8(e) and (f)). The discrepant cellular responses observed in the transformed MSC cell lines and human iPSC-derived NSCs might due to that iPSC-derived cultures are inherently more sensitive to their culture environment. Subtle effects of molecules released after gel degradation can affect the iPSC-derived NSCs while the MSCs remain unresponsive. We also observed that after incubation, the hydrogel acquired a large volume expansion, resulting in increased surface area. While in the MSC culture, the addition of 10% (v/v) FBS can sufficiently cover the internal structure of the expanded hydrogel, the provided laminin coating (10 µg ml −1 ) in the iPSC-derived neural cultures, on the other hand, might no longer be able to supply adequate ECM protein after gel expansion and thus affect cell binding, attachment, viability, and growth. In the future, cell adhesion molecules can potentially be functionalized into our hydrogel polymer to improve cell attachment, proliferation, and growth. A higher concentration of BDNF can be incorporated into the hydrogel to support cell survival in the early phase of cell delivery.
DA as a neurotransmitter known to control movement and emotions has been used for local delivery to treat Parkinson's disease [63]. Application of hydrogel for delivery of DA or its agonists with or without incorporation of cells has been reported [24-26, 64, 65]. In general, the controlled release of molecules modulated via physical interactions (e.g. electrostatic interactions and hydrophobic association) possess a relatively short-term release compared to chemical conjugation methods, where the polymer immobilizes the drug and it is only released when the network degrades [66]. In our study, we chemically conjugated DA into our polymer backbones. Although we have not yet examined the release of DA from the hydrogel, the covalent linkages between DA and the backbone polymer might lead to a slower release compared to physical incorporation methods. As most neurological diseases require longterm therapy, this hydrogel might facilitate the treatment of these diseases. It has been shown that chemical modifications of a drug molecule might affect drug activity or property if it modifies molecules' active sites or causes a conformation change [67]. It is worth understanding whether our hydrogel could release functional DA molecules after degradation and if there are any side products formed after material degradation. Future works will focus on the characterization of the released products, their release profile, and their effects on biocompatibility and cellular behaviors.

Conclusion
We have synthesized a novel hydrogel material based on HyA-DA-HPA. The gelation time of the HyA-DA-HPA hydrogel can be controlled from 3 s to 5 min by changing the concentration of HRP, H 2 O 2 , and polymer. For specific applications, the hydrogel can be rapidly formed within 3 s under physiological conditions using HRP and H 2 O 2 as catalysts for coupling. The effect of H 2 O 2 concentration on cell viability was evaluated using MSCs. The synthesized hydrogel acquires no cytotoxicity, except for the hydrogels formed with 0.75 and 1.0 µmol ml −1 of H 2 O 2 due to the cytotoxicity of the residual H 2 O 2 . We also examined the biocompatibility of the hydrogel with human clinically relevant cell populations, human iPSC-derived NSCs. A decreased viability was observed and the results could be attributed to the release of the degradation products and the lack of cell binding sites. Initial in vitro studies revealed the potential of the encapsulation of MSCs inside the hydrogel. Future studies will include the detailed analysis of the effects of HRP, H 2 O 2 , and polymer concentrations on the gel's stiffness and degradation time. Examination of the formation of any side products and their effects on biocompatibility, cell behaviors, and long term cell viability will be conducted to improve its biocompatibility with human clinically relevant cells.