Gelatin-based composite hydrogels with biomimetic lubrication and sustained drug release

The occurrence of osteoarthritis is closely related to progressive and irreversible destruction of the articular cartilage, which increases the friction significantly and causes further inflammation of the joint. Thus, a scaffold for articular cartilage defects should be developed via lubrication restoration and drug intervention. In this study, we successfully synthesized gelatin-based composite hydrogels, namely GelMA-PAM-PMPC, with the properties of biomimetic lubrication and sustained drug release by photopolymerization of methacrylic anhydride modified gelatin (GelMA), acrylamide (AM), and 2-methacryloyloxyethyl phosphorylcholine (MPC). Tribological test showed that the composite hydrogels remarkably enhanced lubrication due to the hydration lubrication mechanism, where a tenacious hydration shell was formed around the zwitterionic phosphocholine headgroups. In addition, drug release test indicated that the composite hydrogels efficiently encapsulated an anti-inflammatory drug (diclofenac sodium) and achieved sustained release. Furthermore, the in vitro test revealed that the composite hydrogels were biocompatible, and the mRNA expression of both anabolic and catabolic genes of the articular cartilage was suitably regulated. This indicated that the composite hydrogels could effectively protect chondrocytes from inflammatory cytokine-induced degeneration. In summary, the composite hydrogels that provide biomimetic hydration lubrication and sustained local drug release represent a promising scaffold for cartilage defects in the treatment of osteoarthritis.


Introduction
From a biotribological viewpoint, osteoarthritis has been accepted as a lubrication deficiency-induced joint disease that is characterized by the breakdown of articular cartilage and inflammation of the joint. Therefore, a synergetic therapy integrating both lubrication and drug intervention is a promising approach for the treatment of osteoarthritis [1,2]. However, due to the insufficiency of blood vessels in articular cartilage, it is very difficult for nutritious agents/drugs to reach the joint via oral administration [3,4]. Thus, the design of a scaffold to simultaneously achieve enhanced lubrication and sustained drug † Kuan ZHANG and Jielai YANG contributed equally to this work. delivery is a good solution for cartilage defects, although this subject is still facing great challenges.
Hydrogels have been widely studied as an ideal substitute for articular cartilage and as a drug carrier. Great effort has been devoted to designing functional hydrogels with good biocompatibility [5], high mechanical strength [6], and low coefficient of friction (COF) [7,8]. Recently, many hydrogels have been developed to exhibit excellent physicochemical properties such as double-network hydrogels [9], mussel-inspired functional hydrogels [10][11][12], composite hydrogels [13], and hydrogen-bonded crosslinked hydrogels [14]. Gelatin, a single-chain derivative formed by collagen, is biocompatible and can be biodegraded in vivo. Additionally, the side chains of gelatin are rich in reactive groups (-COOH and -NH 2 ), and thus different gelatin hydrogels (GelMA) have been synthesized to repair skin, bone, and articular cartilage [15,16], in which methacrylic anhydride is introduced to initiate photopolymerization [17,18]. However, to the best of our knowledge, the lubrication properties of GelMA hydrogels have rarely been investigated.
Many methods have been reported for improving the mechanical properties of hydrogels, but the improvement in mechanical properties often compromises lubrication [19][20][21]. Naturally, articular cartilage has an ultra-low COF (at a level of 0.001-0.01) based on hydration lubrication mechanism [22][23][24], where polyelectrolyte biomacromolecules (such as hyaluronic acid, aggrecan, and lubricin) complex with phosphatidylcholine (PC) lipids form a lubricating boundary layer by exposing the hydrated phosphocholine groups (N + (CH 3 ) 3 and PO 4 − ) on the superficial surface [25]. Poly(2-methacryloyloxyethyl phosphorylcholine) (PMPC), a biocompatible polymer, has the same zwitterionic phosphocholine groups as PC lipids, and it has been widely used to enhance the lubrication properties of various biomedical materials through surface modification [26][27][28].
In this study, bioinspired by the hydration lubrication mechanism of articular cartilage, we developed gelatinbased composite hydrogels with biomimetic lubrication and sustained drug release (GelMA-PAM-PMPC). Specifically, acrylamide (AM) and MPC were introduced into GelMA to improve the mechanical and lubrication properties of the composite hydrogels. Additionally, diclofenac sodium (DS, a typical anti-inflammatory drug to relieve pain due to osteoarthritis) was encapsulated while preparing the composite hydrogels. It is hypothesized that the dual-functional composite hydrogels developed herein, as a scaffold for cartilage defects to achieve both lubrication enhancement and local drug delivery, may find applications in the treatment of osteoarthritis.

Synthesis of composite hydrogels
GelMA was synthesized according to previously described methods [29,30]. Gelatin (5 g) was added to phosphate buffer solution (PBS, 50 mL) and magnetically stirred at 50 °C until it was completely dissolved. Then, methacrylic anhydride (5 mL) was added dropwise to the above solution (0.5 mL/min) and reacted in a 50 °C isothermal water bath under magnetic stirring. After 4 h, the reaction solution was diluted with PBS (200 mL, 50 °C) and stirred to terminate the reaction. The above solution was placed 234 Friction 10(2): 232-246 (2022) | https://mc03.manuscriptcentral.com/friction in a dialysis bag (molecular weight cutoff: 8-14 kDa) and dialyzed in pure water at 50 °C for 6 d. The dialyzed solution was poured into a centrifuge tube for centrifugation (2,500 rpm), and the supernatant was transferred and lyophilized to obtain the GelMA product.
The synthesis of the composite hydrogels was achieved by photopolymerization. Briefly, GelMA (1 g), MPC (mass ratio to GelMA: 5%, 15%, 30%, and 50%), crosslinker (bis-acrylamide, 1%), and photoinitiator (I2959-Tos, 5 mg) were added to the AM solution to pre-polymerize the monomers at 37 °C. Subsequently, the pre-polymerized solution was transferred to a custom-made mold and irradiated using an ultraviolet spot light source (7.1 mW/cm 2 , 360-480 nm) for 5 min. Finally, the hydrogels (GelMA-PAM-PMPC) were taken out and soaked in PBS at 37 °C for 24 h to remove the uncrosslinked monomers. Pure GelMA-PAM and GelMA hydrogels were prepared using the same method. Unless mentioned otherwise, the GelMA-PAM-PMPC samples used in the following tests contained 30% MPC.

Characterization of composite hydrogels
1 H nuclear magnetic resonance (NMR) spectra of gelatin and GelMA were recorded using an Ascend 400 MHz NMR spectrometer (Bruker, USA) with D 2 O as the solvent. Fourier transform infrared (FTIR) spectra of PMPC, GelMA, and GelMA-PAM-PMPC were analyzed using a Nexus 670 spectrometer (Nicolet, USA) at 400-4,000 cm −1 . The water content of GelMA and GelMA-PAM-PMPC was measured via thermogravimetric analysis (TGA) (Q5000IR, TA Instruments, USA) from 25 to 300 °C at a heating rate of 10 °C /min.
To examine the swelling behavior of the hydrogels, GelMA and GelMA-PAM-PMPC were incubated in PBS at 37 °C and sampled at 0, 0.5, 1, 3, and 5 h after incubation. Then, the hydrogels were freeze-dried, and the swelling ratio (λ w ) and the relative volume (λ v ) were calculated using Eqs. (1) and (2).
Here, W i and W 0 are the swollen and dry weights of the hydrogels, respectively, which were measured in an equilibrium state. The swelling ratio (λ w ) was calculated as the ratio of the weight of swollen hydrogels to that of dry hydrogels (Eq. (1)). R i , R 0 , H i , and H 0 are the radius and height of the hydrogels (cylindrical samples) before and after swelling, respectively. The relative volume (λ v ) was calculated as the ratio of the volume of swollen hydrogels to that of dry hydrogels (Eq. (2)). The water contact angle of GelMA and GelMA-PAM-PMPC was obtained using an OCA25 contact angle goniometer (Dataphysics Instruments, Germany) by the sessile drop method. Distilled water (3 μL) was placed on the airside surface of the hydrogels at room temperature, and the static contact angle was collected after 10 s. The mean contact angle was calculated from the results of at least three measurements.
To depict the internal microstructures of the hydrogels, GelMA and GelMA-PAM-PMPC were freeze-dried to thoroughly remove water. Subsequently, the hydrogels were coated with Pt/Pd and examined using a Quanta 200 field emission scanning electron microscope (SEM, FEI, Eindhoven, Netherlands) under an accelerating voltage of 5 kV, which was coupled with energy dispersive spectroscopy (EDS) to enable elemental composition analysis. Image J software was used to quantify the pore distribution of the hydrogels based on a series of SEM images. The pores were randomly selected on the surface of the hydrogels. The pore size was nominated as the average pore diameter on the selected SEM images.
The mechanical properties of GelMA and GelMA-PAM-PMPC were evaluated using a Zwick Z020 universal testing machine (ZwickRoell, Germany) with a 0.25 kN load cell. The hydrogels used for the compressive performance test were cut into cylindrical shapes (diameter: 15 mm; length: 15 mm). During the compressive tests, the speed of the crosshead was maintained at 0.5 mm/min until the hydrogels failed. A series of five samples were evaluated to ensure the reproducibility of the data.

In vitro drug loading and release
Initially, a calibration curve of the drug DS in PBS at various concentrations (5, 10, 15, 20, and 25 μg/mL) was obtained by measuring the absorbance value using a UV-vis spectrophotometer (UV-8000s, Metash Instruments, China) at 276 nm, as displayed in Fig. S1 in the Electronic Supplementary Material (ESM). To load the drug, GelMA and GelMA-PAM-PMPC (100 mg) were added to the DS solution in PBS (20 mL) and uniformly dispersed by ultrasound. The mixture was stirred for 12 h, and the hydrogels were removed and washed with deionized water several times. The DS remaining in PBS was monitored according to the calibration curve. Similarly, the DS released from the hydrogels at various intervals was evaluated using a UV-vis spectrophotometer. The test for drug release was performed until the solution concentration remained nearly constant. The drug loading capacity (LC), encapsulation efficiency (EE), and cumulative drug release of the hydrogels were calculated using Eqs. (3)- (5). Each test was repeated three times from which the mean value was calculated.
Here, M t is the amount of DS released from the hydrogels at time t, while M a is the DS encapsulated in the hydrogels.

Lubrication properties
The lubrication properties of the hydrogels were evaluated by a tribological test, which was performed on a UMT-3 universal material tester (Center for Tribology Inc., USA) operated in a pin-on-disk rotating mode (rotation radius: 2 mm). The hydrogels (GelMA-PAM and GelMA-PAM-PMPC) were fixed to the platform of the tester with cyanoacrylate glue. The contact pair was a sphere (bearing steel GCr15) with a radius of 3 mm. Tribological tests were conducted at 25 °C under various experimental conditions: the content of MPC (0%, 5%, 15%, 30%, and 50%), normal load (0.1, 0.2, 0.5, 1, and 2 N), and rotation frequency (0.5, 1, 2, and 5 Hz). Tribological tests were performed for 10 min, and sufficient deionized water was added to the surface of the hydrogels as a lubricant. The tribological test under each condition was performed at least three times to obtain reliable data, and the mean COF was then recorded. After the tribological test, the surface topology of the hydrogels and the steel spheres was evaluated using an optical interferometer (NeXView, ZYGO, USA), and the surface roughness values (Sa) of the samples were obtained from at least three measurements at random positions on the surface [31].

Primary chondrocytes isolation and culture
Articular cartilage was collected from mice (C57B/L6, male, 4-7 days old, Ruijin Hospital, School of Medicine, Shanghai Jiao Tong University). The pieces of articular cartilage were dissected and separated aseptically from the underlying bone and connective tissues. The cartilage was then cut into small pieces, washed with PBS, and digested using collagenase type II (0.2%) at 37 °C for 4 h. Afterward, the cartilage tissues were suspended and seeded into tissue culture plates in an incubator (37 °C and 5% CO 2 ). The culture medium was Dulbecco's modified eagle medium (DMEM)/ nutrient mixture F-12, which was supplemented with 1% penicillin/streptomycin antibiotics and 10% fetal bovine serum. The cells were passaged using 0.25% trypsin-EDTA solution when the confluence reached approximately 80%-90%. Leach solution of hydrogel materials was used for all the following cell experiments.

Cell proliferation
Cell cytotoxicity was evaluated using a CCK-8 kit with reference to the manufacturer's instructions. Briefly, primary chondrocytes were cultured in 96-well plates at 1×10 4 cells per well. The culture plates were placed in a humidified atmosphere of 37 °C and 5% CO 2 during incubation. After co-culturing with GelMA and GelMA-PAM-PMPC hydrogels for 1, 3, and 5 days, the cells were washed with PBS, and 10 μL of the CCK-8 solution was added to each well, before being further cultured at 37 °C for another 2 h. Subsequently, the solution absorbance was measured at 450 nm using a multiskan spectrum microplate photometer (Thermo Scientific, Finland). The group treated with untreated chondrocytes was assigned to the blank group. Three replicates of each group were evaluated and shown as optical density, which directly correlated to the number of viable cells.

Cell morphology and viability
The chondrocytes were seeded on glass coverslips and then co-cultured with the GelMA and GelMA-PAM-PMPC hydrogels for 1, 3, and 5 days. The effect of GelMA and GelMA-PAM-PMPC hydrogels on the morphology and viability of chondrocytes was evaluated using a live/dead cell kit (Life Tech, USA). After co-culturing with the hydrogels for 1, 3, and 5 days, ethidium homodimer-1 (2 μL) and calcein acetomethoxy (0.5 μL) were mixed with DMEM (1 mL) for staining the cells in the dark for 30 min. Following incubation, the constructs were washed with PBS three times and observed using a fluorescent microscope (Axio Imager M1, ZEISS, Germany). The group treated with untreated chondrocytes was assigned to the blank group.

Phalloidin staining
The chondrocytes were plated on glass coverslips and then co-cultured with the GelMA and GelMA-PAM-PMPC hydrogels for 1, 3, and 5 days following the procedure described above. The attached cells were fixed using paraformaldehyde (4%) and permeabilized with Triton X-100 (0.2%) for 10 min. Afterward, the cells were stained with 100 nM Alexa Fluor 488-conjugated phalloidin (A12379, Thermo Fisher) for 30 min in the dark and then fixed with paraformaldehyde (4%) for 20 min. After washing with PBS, the stained cells were observed using a laser scanning confocal microscope (LSM800, ZEISS, Germany). The group treated with untreated chondrocytes was assigned to the blank group.

Real-time quantitative polymerase chain reaction (RT-qPCR) assay
To investigate the protection of the hydrogels for inflammation-induced chondrocyte degeneration, the RT-qPCR assay was used to analyze the expression levels of cartilage-specific genes such as COL2A1, aggrecan, ADAMTS5, and MMP13. The chondrocytes were seeded in 6-well plates at 1×10 5 cells per well, stimulated with TNF-α or IL-1β (concentration: TNF-α (5 ng/mL) or IL-1β (10 ng/mL)), and treated with GelMA and GelMA-PAM-PMPC hydrogels simultaneously for 24 h. The total RNA of chondrocytes was extracted from the cells using TRIzol reagent (Invitrogen, USA). RNA (1 μg) was reverse transcribed to synthesize complementary DNA (cDNA). For RT-qPCR, 10 μL of reaction volume was applied, that is, 5 μL of 2X SYBR Master Mix, 4.5 μL of diluted cDNA, and 0.25 μL of each primer. The parameters of RT-PCR were set as follows: 95 °C for 10 min, followed by 40 cycles at 95 °C for 15 s and 60 °C for 1 min. The procedure was conducted using an ABI 7500 Sequencing Detection System (Applied Biosystems, CA, USA). The values of cycle threshold (Ct) were obtained and normalized to β-actin. The relative mRNA level of each gene was evaluated based on the 2 -ΔΔCt method [33]. The primers were designed with the aid of the NCBI Primer-Blast Tool, as shown in Table 1. The group using chondrocytes treated with TNF-α or IL-1β was assigned to the blank group.

Statistical analysis
Quantitative data are presented as mean ± standard deviation. Independent tests were repeated at least three times to verify the results. One-way analysis of variance was performed to detect significant differences between separate groups. Statistical analysis was conducted using GraphPad Prism software (GraphPad Software Inc., USA). The level of significance was displayed as *P < 0.05, **P < 0.01.

Design of composite hydrogels
We successfully developed gelatin-based composite hydrogels with the properties of biomimetic lubrication and sustained local drug release (GelMA-PAM-PMPC) that could function as a cartilage substitutional scaffold for treating osteoarthritis. As shown in Fig. 1(a), the composite hydrogels were synthesized via photopolymerization of GelMA, AM, and MPC, using I2959-Tos as the initiator, which was developed in our previous study [34]. The crosslinked network of the composite hydrogels included physical (e.g., entangled chains and hydrogen bonding in gelatin) and chemical (e.g., covalent bonds between the monomers) crosslinking, which enhanced the mechanical properties of the composite hydrogels. As shown in Fig. 1(b), the lubrication property of the composite hydrogels was assessed by a series of tribological tests, which were performed on a ball-on-disk rotating tribometer. The zwitterionic phosphocholine groups (N + (CH 3 ) 3 and PO 4 − ) in PMPC, bioinspired by the hydration lubrication mechanism of articular cartilage, contributed to the lubrication enhancement. Additionally, the composite hydrogels were encapsulated with the drug DS during preparation to endow the hydrogels with sustained drug release behavior. The chondroprotective potential of the composite hydrogels was verified by analyzing the mRNA expression levels of cartilage-specific genes following co-culturing with the cytokine-treated chondrocytes.

Characterization of composite hydrogels
The   Fig. 2(d). The water content of GelMA and GelMA-PAM-PMPC is 61.8% and 73.5%, respectively. The higher water content of GelMA-PAM-PMPC, compared with GelMA, is attributed to the introduction of PMPC, which is regarded as a superhydrophilic material in the presence of phosphocholine groups [35]. The relative volume, swelling ratio, and water contact angle of GelMA and GelMA-PAM-PMPC are illustrated in Figs. 2(e)-2(h). For the same reason, the swelling ratio and relative volume of GelMA-PAM-PMPC are higher than those of GelMA, and the water contact angle of GelMA-PAM-PMPC (43.7° ± 0.5°) is much lower than that of GelMA (104.8° ± 1.3°). With the incubation of the hydrogels in PBS, the swelling ratio of GelMA and GelMA-PAM-PMPC increases greatly during the initial 1 h and reaches equilibrium after 3 h. However, the relative volume of GelMA and GelMA-PAM-PMPC remains unaffected after incubation, which is beneficial when the hydrogels are applied as a scaffold for cartilage defects. The cross-sectional microstructure, pore size distribution, and elemental composition of GelMA and GelMA-PAM-PMPC examined by SEM and EDS are shown in Figs. 2(i)-2(n). GelMA has a highly interconnected and loose porous network with a pore size of approximately 30 μm. Compared with GelMA, the microstructure of GelMA-PAM-PMPC is denser, and the pore size reduces dramatically (to ~6 μm), which is attributed to the increase in the covalent crosslinking of the hydrogels. Additionally, the www.Springer.com/journal/40544 | Friction detection of P for GelMA-PAM-PMPC further confirms that the development of composite hydrogels has been successful.
Mechanical strength is one of the key factors for cartilage substitutional scaffolds, especially when the hydrogels should provide sufficient mechanical support during the early stage of implantation before adapting to surrounding cartilage tissues. The compressive strength of GelMA and GelMA-PAM-PMPC is shown in Fig. 2(o). Compressive strength of GelMA-PAM-PMPC (0.788 MPa at 42% compression strain) is significantly improved than that of GelMA (0.084 MPa at 14% compression strain). This result is attributed to the chemically covalent crosslinking and the corresponding increase in the density of the microstructure of the hydrogels. Generally, gelatin forms physically crosslinked hydrogels by its own hydrogen bonding, and the introduction of methacrylic anhydride, AM, and MPC generates chemically crosslinked hydrogels by UV-induced photopolymerization.

In vitro drug loading and release
The in vitro drug loading and release of the hydrogels was investigated using DS as a nonsteroidal antiinflammatory drug for the treatment of osteoarthritis at 37 °C. The LC and EE of GelMA-PAM-PMPC (10.9%, 49.2%) are slightly higher than that of GelMA (9.5%, 42.1%). The drug release profile of DS-loaded GelMA and GelMA-PAM-PMPC is shown in Fig. 2(p). Both curves display an initial rapid drug release, which is followed by a plateau stage. After 10 days, 88.7% of DS is released from GelMA, which is much higher than that of GelMA-PAM-PMPC (64.5%). DS released from GelMA-PAM-PMPC is lower at each time interval, compared with GelMA, which indicates that the composite hydrogels can achieve a sustained drug release of DS.

Lubrication property of composite hydrogels
A series of tribological tests were performed to reveal the lubrication properties of the composite hydrogels, as shown in Fig. 3(a). Before the tribological tests, the surface roughness of GelMA-PAM and GelMA-PAM-PMPC is measured to be 218 and 284 nm, respectively (Figs. 3(b) and 3(c)). The COF of GelMA-PAM-PMPC with various MPC contents is displayed in Fig. 3(d) (load: 0.5 N; frequency: 2 Hz). The lubrication of the composite hydrogels is highly dependent on the MPC content, and with the increase in the MPC content, the COF value decreases significantly from 0.052 (0%) to 0.011 (30%). A further increase in the MPC content to 50% does not improve the lubrication. Consequently, this setup is applied in the tribological tests under different conditions. Hydrogels are viscoelastic and the surface physicochemical properties can result in a complex lubrication performance [36,37]. The lubrication of hydrogels is related to the applied load and rotation frequency, which is investigated and illustrated in Figs. 3(e) and 3(f). The COF values of GelMA-PAM and GelMA-PAM-PMPC increase greatly from 0.042 to 0.063 and from 0.014 to 0.041 when the normal load changes from 0.1 to 2 N. This result is attributed to the effect of the normal load on the contact stress and deformation of the hydrogels, as schematically shown in Fig. 3(g). Under a higher normal load, with the increase in the contact stress and indentation depth, the lateral friction force will be greatly increased, thus resulting in a larger COF value. Additionally, the COF value of GelMA-PAM-PMPC slightly decreases from 0.018 to 0.011 when the rotation frequency increases from 0.5 to 5 Hz, and a similar trend is obtained for GelMA. The short contact time of slip at a high rotation frequency produces an effective hydration interface between the tribopairs, which can result in a reduced COF value [38,39]. Under each experimental condition, the COF value of GelMA-PAM-PMPC is lower than that of GelMA-PAM, although it has a relatively high surface roughness. This indicates that the enhanced lubrication of the composite hydrogels is due to the introduction of the PMPC.
The zwitterionic phosphocholine groups (N + (CH 3 ) 3 and PO 4 − ) in PMPC are the same as that in the PC lipids, which form a complex with the polyelectrolyte biomacromolecules and dominate the superlubrication of articular cartilage based on hydration lubrication. Phosphocholine groups can attract water molecules to form a tenacious hydration shell around the charges as a result of the interaction between the water dipole and enclosed zwitterionic charges [40]. The hydration shell not only supports high pressures 240 Friction 10(2): 232-246 (2022) | https://mc03.manuscriptcentral.com/friction without being squeezed out but also behaves in a fluidlike manner under shear, leading to a significant reduction in the interfacial friction under various test conditions [41]. Therefore, the composite hydrogels GelMA-PAM-PMPC can enhance lubrication owing to the hydration lubrication mechanism. Additionally, the excellent water trapping capability of the composite hydrogels due to interconnected porous microstructure also contributes to maintaining good lubrication performance.
The optimal scaffold for repairing articular cartilage defects should maintain a low COF value after multiple testing cycles [42]. To examine the durability of the composite hydrogels, hard steel and soft polydimethylsiloxane (PDMS) balls were employed as tribopairs to slide against the hydrogels under a normal load of 0.5 N and a rotation frequency of 5 Hz, with an extended duration of 10,000 cycles. The surface roughness of the steel ball and PDMS ball is approximately 9 and 30 nm, respectively, as shown in Figs. 3(h) and 3(i). It is indicated from Figs. 3(j) and 3(k) that the COF value remains relatively unchanged during the test. A lower COF value is obtained using steel ball as the contact tribopair, compared with the PDMS ball, although the contact stress between the steel ball and the hydrogels is much larger than that between the PDMS ball and the hydrogels (as the elastic modulus of steel is much higher than that of PDMS). The larger COF value using the PDMS ball as the contact tribopair is attributed to its higher surface roughness. The above results indicate that the composite hydrogels can sustain low friction under extended loading cycles, especially when sliding against a steel ball.

Cell cytotoxicity and protection for chondrocytes degeneration
To evaluate the potential clinical application of the composite hydrogels, we investigated the in vitro cytotoxicity of GelMA and GelMA-PAM-PMPC on primary mouse chondrocytes and performed tests to determine whether GelMA and GelMA-PAM-PMPC can protect against chondrocyte degeneration. The hydrogel samples used in the following tests were encapsulated with DS. The live/dead assay and CCK-8 test were conducted to examine the cytotoxicity of the hydrogels on cell viability and proliferation of primary mouse chondrocytes. Figures 4(a)-4(c) show the results of the live/dead assay after co-culturing the chondrocytes with GelMA and GelMA-PAM-PMPC for 1, 3, and 5 days, where the dead cells are labeled red and living cells green. Most of the seeded chondrocytes are alive during culturing, and the density of the cells increases from day 1 to day 5. The viability of the chondrocytes co-cultured with the hydrogels is almost the same as that of the blank group for all incubation times, indicating that the hydrogels are highly biocompatible with no detrimental effect on the chondrocytes. Overall, there are no significant differences in the live/dead cells among the blank, GelMA, and GelMA-PAM-PMPC groups at different time intervals. Phalloidin staining was used to observe the fibrous actin of the cytoskeleton for the chondrocytes, and the results are shown in Fig. 4(d). The morphology of the cells in the three groups is complete and normal, which indicates that the hydrogels have good biocompatibility. The tetrazolium salt of CCK-8 is cleaved to a soluble formazan using live cells, and thus the absorbance is directly related to the number of viable cells. As displayed in Fig. 4(e), the CCK-8 test indicates that there are no significant differences between the experimental groups (GelMA and GelMA-PAM-PMPC) and the blank group at each time interval, and the number of viable cells is greatly increased on days 3 and 5 for GelMA and GelMA-PAM-PMPC. This further indicates that cell proliferation is not affected by co-culturing the chondrocytes with the hydrogels. In summary, the live/dead assay, phalloidin staining, and CCK-8 test all indicate that the hydrogels are biocompatible and have no cytotoxicity to chondrocytes.
Multiple factors are involved in the pathogenesis of osteoarthritis, such as reactive oxygen species, mechanical loading, and inflammatory cytokines (e.g., TNF-α and IL-1β). It is considered that chondrocyte degeneration, which is accepted to be the most significant feature of osteoarthritis, is closely related to the inflammatory cytokines. For example, inflammatory cytokines play an important role in the development of osteoarthritis, and the inflammatory environment contributes to extracellular matrix degradation and chondrocyte hypertrophy. The manifestation of osteoarthritis at the cellular level is due to the increase in catabolic genes and the degradation of anabolic genes. In this study, IL-1β and TNF-α were used to treat chondrocytes to mimic the symptoms of osteoarthritis. The protective effect of the drug-loaded hydrogels on chondrocyte degeneration was examined after co-culturing for 24 h. The mRNA expression levels of anabolic genes (aggrecan and COL2A1) and catabolic genes (MMP13 and ADAMTS5) were examined by RT-qPCR analysis. As displayed in Fig. 5, the mRNA expression of aggrecan and COL2A1 in the hydrogel samples (GelMA and GelMA-PAM-PMPC) is significantly higher than that of the blank group, and the mRNA expression of MMP13 and ADAMTS5 in the hydrogel samples is significantly lower than that of the blank group. These results indicate that the DS-loaded hydrogels have chondroprotective potential for inflammatory cytokine-induced chondrocytes, and consequently can be an effective scaffold for cartilage defects in the treatment of osteoarthritis.

Conclusions
In this study, we successfully synthesized gelatin-based composite hydrogels, namely GelMA-PAM-PMPC, with biomimetic hydration lubrication and sustained drug release via photopolymerization, which could be used as a scaffold for cartilage defects in the treatment of osteoarthritis. The lubrication test indicated that the composite hydrogels maintained a relatively low COF under different experimental conditions and extended duration, which was attributed to the hydration lubrication mechanism of the zwitterionic phosphocholine headgroups. The drug release test showed that the composite hydrogels efficiently encapsulated the anti-inflammatory drug of DS and achieved a sustained release behavior. Additionally, the in vitro test demonstrated that the composite hydrogels were biocompatible and protected the chondrocytes from inflammatory cytokine-induced degeneration, upregulating the mRNA expression levels of anabolic genes and downregulating that of catabolic genes. In summary, the composite hydrogels prepared herein, with the dual functions of biomimetic hydration lubrication and sustained drug release, provide a promising approach for repairing cartilage defects in the treatment of osteoarthritis.