Bioinspired Hard–Soft Composite Scaffold with Excellent Lubrication and Osteogenic Properties for the Treatment of Osteochondral Defect

Natural articular cartilage is a typical self‐healing and superlubrication system capable of maintaining extremely low friction under physiological loadings. Cartilage wear and accidental trauma can cause irreversible defects to cartilage and subchondral bone with a significant decrease in intra‐articular lubrication, leading to the development of severe osteoarthritis and osteochondral defect. To address the important clinical problem of osteochondral defect, a bioinspired hard–soft (PEEK‐lubrication hydrogel) composite scaffold is designed and developed. The polymerization of polyethylene glycol diacrylamide (PEGDAA) and 2‐methacryloyloxyethyl phosphorylcholine (MPC) on polyetheretherketone (PEEK) substrate is achieved by UV initiation to form a strong interfacial bonding, and nano‐hydroxyapatite is deposited on porous PEEK substrate via polydopamine coating to improve osteogenic capability. Accordingly, the composite scaffold is successfully developed with lubrication and osteogenic activity. The tribological tests show that the lubrication performance of the composite scaffold is based on the hydration lubrication mechanism of the upper hydrogel layer, and the in vitro and in vivo experiments demonstrate that the composite scaffold is endowed with excellent biocompatibility and bioactivity. In conclusion, the bioinspired strategy for preparing a hard–soft composite scaffold shows a promising way in the treatment of osteochondral defect and provided a guideline for designing functional PEEK‐based biomaterials in tissue engineering scaffolds.


Introduction
With the aging of the population and the occurrence of inappropriate sports or accidents, the number of patients with osteochondral defects has been increasing, causing severe osteoarthritis and affecting the daily life of the patient.Typically, osteochondral defects involve the damage from the articular cartilage to the subchondral bone area. [1]While important progress has been made in the treatment of articular cartilage damage over the past few decades, the defect in the deep subchondral bone area has not been received sufficient attention, and osteochondral defects are now considered as a very important clinical problem.The complexity of treating osteochondral defects is twofold.On the one hand, there are no blood vessels in the cartilage tissue, and the physiological feature means that its regenerative capacity is significantly limited. [2,3]Even if regeneration is temporarily achieved, the regenerated cartilage tissue is fibrocartilage, which differs remarkably from the structure and function of natural hyaline cartilage. [4]On the other hand, the material structural properties, mechanical properties, and cellular distribution of articular cartilage and subchondral bone differ dramatically with each other. [5]Due to the heterogeneity of the osteochondral region, the application of a monophasic scaffold cannot accommodate the repair or replacement of tissues with appropriate functions and complex structures.In contrast, heterogeneous bilayer scaffolds can integrate materials with significantly different functional, structural, and mechanical properties, therefore biomimicking the natural organization of articular cartilage and subchondral bone. [6,7]n a healthy state, articular cartilage is a typical superlubrication system, which contains ≈80% interstitial water and 20% extracellular matrix permeated by a large number of charged macromolecules, proteins, chondrocytes, and other components. [8]The polyelectrolyte macromolecules are complexed with phosphatidylcholine lipids to expose phosphorylcholine groups on the cartilage surface to form a hydrated lubrication layer, allowing articular cartilage to exhibit an extremely low friction coefficient (COF, 0.001-0.01)over a wide range of shear rates and at the pressures in excess of 100 atm. [9,10]Similarly, hydrogels are 3D porous reticular materials that are formed through cross-linking hydrophilic water-soluble polymers or monomers under certain conditions.High water content endows the hydrogels with the characteristic of cartilage extracellular matrix, while the 3D network structure of hydrogels results in a suitable viscoelasticity similar to natural cartilage in terms of mechanical property, which means that hydrogels have promising applications in cartilage repair and replacement. [11,12]Similar to the lubrication mechanism of articular cartilage, hydrogels are able to attract hydration layer or store lubricant to achieve excellent lubrication behavior.For example, 2-methacryloyloxyethyl phosphorylcholine (MPC), which is structurally similar to the phosphatidylcholine lipids with the same zwitterionic phosphorylcholine groups and excellent biocompatibility, has been widely applied in the design of biomaterials for the purpose of enhancing lubrication performance. [13,14]olyetheretherketone (PEEK) is a thermoplastic polymeric engineering material, that has been widely investigated in the field of bone tissue engineering for clinical implant treatment, such as bone defects and spinal implants, due to its excellent biocompatibility, elastic modulus (3-4 GPa) similar to that of human cortical bone, chemical resistance, and radiation transmissibility. [15,16]owever, the inherent biological inertness of PEEK with relatively high hydrophobic surface, low surface energy, and weak cell and protein adsorption capacity significantly limits its osteogenic activity.Surface modification can effectively enhance the bioactivity of PEEK by improving its osseointegration ability. [17,18]Among the surface modification techniques, the use of dopamine as a universal method to achieve multifunctional coating on different substrates has been intensively reported. [19]The catechol groups in dopamine can be oxidized and self-polymerized in neutral or weakly alkaline media to form polydopamine (PDA) coating rich in catechol, quinone, and amine structures via covalent or hydrogen bonding and - interactions. [20,21]The PDA coating, in addition to being spontaneously deposited onto almost all types of substrates, can be covalently coupled to nucleophilic reagents, enabling the binding of biologically active molecules. [22,23]For example, Zhu et al. modified PDA coating on sulfonated PEEK and subsequently integrated peptides onto the material surface.The modified PEEK could promote cell proliferation, osteogenic differentiation, and greatly improve bone-like apatite formation in vitro. [24]Xu et al. prepared a composite scaffold using PDA coating combined with dexamethasone and minocycline liposomes on the PEEK surface, which significantly enhanced the antiinflammatory, antibacterial, and osseointegration performances and therefore had a great potential for application in orthopedic and dental fields. [25]Additionally, the excellent physicochemical properties and biocompatibility of polyethylene glycol (PEG) hydrogels enable their successful use in a wide range of biomedical applications. [26]Polyethylene glycol diacrylamide (PEGDAA) based on the presence of amide bonding has excellent degradation stability, which is considered beneficial for long-term applications in living organisms.
Based on the above considerations, in the present study, we designed and developed a bioinspired hard-soft (PEEK-lubrication hydrogel) composite scaffold, namely PEEK-PHA/PEGDAA-MPC hereafter, which biomimicked the articular cartilagesubchondral bone structure of the human joint. [27,28]In this composite scaffold as shown in Figure 1, the chemical linkages are formed by free radical polymerization between PEGDAA, MPC, and PEEK.Meanwhile, the mechanical locking is formed due to the embedding of hydrogel precursor solution in the porous PEEK and subsequent curing.Specifically, a lubricious hydrogel network consisting of PEGDAA and MPC formed on porous PEEK under UV light initiation.The charged phosphorylcholine groups (N + (CH 3 ) 3 and PO 4 -) in MPC attracted water molecules to produce a hydration layer for effective lubrication. [29,30]In addition, the porous PEEK substrate loaded with PDA-modified nano-hydroxyapatite (PHA) could greatly enhance the bioactivity and osteogenic capacity, which achieved the bone ingrowth and scaffold fixation purposes.It was anticipated that the composite scaffold with various structural, functional, and mechanical properties developed herein could be applied as a promising strategy for the replacement therapy of osteochondral defects in clinics.

Characterizations of Composite Scaffold
To verify the chemical composition of the hydrogels, Figure 2a demonstrates the Fourier transform infrared (FTIR) spectra of MPC, PEGDAA, and PEGDAA-MPC.The characteristic peaks of MPC are observed for PEGDAA-MPC.The absorption band of C═O appears at 1720 cm −1 , and the absorption peaks at ≈960, 1060, and 1240 cm −1 belong to the P─O, P─O─C, and O═P─O─ groups of the MPC phosphate groups, respectively. [31]The presence of the new peaks indicates that the PEGDAA-MPC hydrogel was successfully synthesized by the photopolymerization method.To further verify the composition of the hydrogels, the elemental composition of PEGDAA and PEGDAA-MPC hydrogels are analyzed using an X-ray photoelectron spectrometer (XPS).As can be seen in Figure 2b, the introduction of MPC results in the appearance of P 2s and P 2p signals in the spectra.In the high-resolution spectra in Figure 2c,d, the appearance of N 1s peaks at 398 and 400 eV is attributed to ─NH─ (present in both PEGDAA and MPC chemical structures) and quaternary amine groups (only from the ─N + (CH 3 ) 3 group in MPC).In addition, a new P 2p peak is observed at 131 eV for PEGDAA-MPC, which is attributed to the ─PO 4 -group in MPC, while no signal for the P element is observed for PEGDAA.These results can serve as further evidence for the successful introduction of MPC into the hydrogel network.
To check whether the nano-hydroxyapatite (nHA) surface is coated with PDA, the samples are examined by FTIR spectroscopy.As shown in Figure S1a (Supporting Information), from the FTIR spectrum of PHA, the broad peaks at 3220 cm −1 are attributed to the ─OH and ─NH─ groups in the PDA coating. [32]he characteristic peaks at 1620 cm −1 are assigned to the  bending vibration of N─H and the stretching vibration of the benzene ring, and the characteristic peaks at 1013 and 950 cm −1 corresponds to the ─PO 4 3-in nHA, which confirm the presence of PDA on the nHA surface.As indicated from the scanning electron microscope (SEM) image of PHA in Figure S1b (Supporting Information), the surface modified nHA tends to agglomerate with each other.
The network porous microstructures of PEGDAA and PEGDAA-MPC hydrogels are observed by SEM and the elemental compositions are analyzed by energy dispersive spectroscopy (EDS) (Figure 3a,b).The interconnected and sparse porous structures are observed in PEGDAA, and only C and O elements are detected from the EDS energy spectrum.Compared with PEGDAA, the cross-linking degree of the hydrogel network is changed and the porous structure is much denser due to the introduction of MPC monomer, which facilitates the increase in mechanical properties.From the energy spectrum, not only the C and O elements are detected, but also more importantly, the presence of P element is observed, which apparently originates only from MPC.This result further verifies the successful polymerization of MPC.The cross-section of the composite scaffold is then observed (Figure 3c), and it is noteworthy that the hydrogel layer has been firmly bonded to the PEEK substrate, with no obvious interfacial delamination.In the present study, the "soft" and "hard" materials demonstrate a dual binding mechanism to ensure a strong interlocking of the bilayer structure.First, the prepared PEEK substrate is a porous structure with an unmodified upper surface.When the hydrogel is cured under UV initiation, part of the pre-gel solution enters the pores and cures, forming a mechanical interlocking between the hydrogel and the PEEK.In addition, upon UV excitation, free radicals generated on the PEEK chemically bond with the polymerized hydrogel, forming a chemical link.Based on the mechanical and chemical connections, a relatively strong hard-soft interface is successfully constructed.The substrate shows a porous structure with some residual NaCl in addition to C and O elements as seen in the EDS spectrum.Figure S2 (Supporting Information) displays the surface morphology of PEEK following surface modification, which demonstrates that the PHA coating is uniformly modified and the polishing scratches are masked by the coating.
The 3D structure of hydrogels can accommodate a large amount of water to support the network.Due to the different hydrophilicity and cross-linkage of hydrogels, the water absorption energy and swelling volume vary for different hydrogels.The swelling ratio is an important property that needs to be regulated for hydrogel applications.According to the swelling test result in Figure 4a,b, when the molar ratio of PEGDAA to MPC is larger than 1:2, the introduction of a large content of MPC can significantly improve the water absorption and hydrophilicity of the hydrogel.When the hydrogel is swollen in deionized water for 24 h, the bonding between the hydrogel and substrate is damaged and the hydrogel cannot maintain its shape and function.This is owing to the excessive water uptake of the hydrogel and increased volume expansion, resulting in the bonding at the interface between the substrate and the hydrogel being unable to bear excessive pressure.Consequently, the occurrence of obvious hydrogel fragmentation is observed.When the molar ratio of PEGDAA to MPC is less than or equals to 1:2 (including the PEGDAA group without MPC), after swelling in deionized water for 24 h, the bonding between the hydrogel and substrate is not affected, and a stable connection and original function can still be maintained, although the volume of hydrogel layer after swelling is slightly increased.Therefore, it is indicated that there is a limitation for the amount of MPC introduced to the hydrogel, and the content of MPC has a great influence on the property of the composite scaffold.
The calculation of the swelling ratio in Figure 4c also confirms the conclusion of the above experiments.Whether the hydrogel is placed in a simulated body fluid (SBF) solution or deionized water, the swelling ratio tends to become larger with the increase in MPC content.In both sets of liquid environments, the hydrogel is able to absorb water and swell rapidly within 6 h.After that, it reaches a relatively stable state, and the swelling ratio remains unchanged after incubation for 24 h.It is noteworthy that the swelling ratio of hydrogels is higher in deionized water than in SBF solution, which may be due to the high ion concentration in SBF solution.
As a potential solution for implant replacement for cartilagesubchondral bone defects, the degradation property of the mate-rial is an important factor influencing how long the material will function in the body and be able to avoid the pain associated with repeated surgeries.The SBF solution has an ionic concentration almost identical to that of human plasma and is utilized to simulate the body's fluid environment.In another control group, three different concentrations of type II collagenase are added according to the division of degradation time to assess the effect of enzymes on degradation.From Figure 4d,e and Figure S3 (Supporting Information), it can be concluded that both groups reach the dissolved state within 6 h.The SBF solution group shows a stable change in the quantity of the hydrogel during the nine-week test duration, with no obvious signs of degradation.Similarly, the group with the introduction of type II collagenase has no significant effect on the degradation of the hydrogel, and the quantity of the hydrogel remains very stable under the condition of increasing enzyme concentration.In accordance with the swelling experiment, it is clear from the degradation profile plots that the introduction of MPC has a potential effect on the change of hydrogel quantity.With the increase in MPC content, the water absorption of hydrogel is enhanced, the water content in the network becomes larger, and therefore the mass of hydrogel is increased.
MPC is a typical zwitterionic monomer with charged groups of N + (CH 3 ) 3 and PO 4 -that can attract water molecules, and the introduction of MPC can significantly improve the hydrophilicity of the hydrogel.From the water contact angle test result in Figure 4f, the contact angle of porous PEEK substrate is 78.7 ± 1.8°, which is a relatively hydrophobic surface.The surface-bound PEG-DAA hydrogel shows a relatively hydrophilic surface with a contact angle of 54.6 ± 1.6°.When the surface is combined with PEGDAA-MPC (1:2) hydrogel, the contact angle greatly reduces to a value as low as 15.0 ± 0.5°due to the high hydrophilicity of MPC.
The joint is a vital weight-bearing part of the body, and excellent mechanical strength is one of the key factors required for joint implant materials. [33]The upper layer of the hydrogel needs to provide sufficient support in the early implantation process before adapting to the surrounding cartilage tissues.The stressstrain curve in Figure 4g shows that the maximum compressive strength of PEGDAA hydrogel is ≈1.27MPa.After MPC polymerization into the hydrogel network, the change of the cross-linking degree makes the network structure of the hydrogel much denser, and therefore the maximum compressive strength increases to 1.38 MPa.Meanwhile, in comparison with PEGDAA, PEGDAA-MPC can bear a greater strain, and the elasticity of the hydrogel increases.
Figure 4h displays the result of the rheological test for the hydrogels.It is clear that with the angular frequency from 1-100 rad s −1 during the shearing process, the energy storage mod-ulus G′ of the samples with different MPC contents is always larger than the loss modulus G″, indicating that the samples are in the state of the hydrogel.The energy storage modulus G′ also increases with the introduction and content of MPC, indicating that the mechanical shear resistance of the hydrogel is improved to some extent.

Lubrication Property of Composite Scaffold
A series of tribological tests are conducted to investigate the lubrication property of the hydrogel layer of the composite scaffold and the interfacial stability of the hard-soft bilayer.As shown in Figure 5a, the tribological tests are performed in a reciprocal motion mode with a ball-on-plate contact.The polytetrafluoroethylene (PTFE) sphere is used as the upper specimen, and the composite scaffold is attached at the bottom of a surface dish where a sufficient amount of SBF solution is maintained as the fluid environment during the test.Figure 5c shows the friction coefficient of the material for the parameters of 1 N load and 0.5 Hz frequency.When the PEEK substrate is not modified by the hydrogel layer, it is a porous and relatively rough surface with a COF value of ≈0.214.When the upper layer is polymerized with a PEGDAA hydrogel layer, the COF value greatly reduces to 0.063.This is due to the fact that the hydrogel is a hydrophilic material with a large amount of water in the network, and the surface hydration layer is effective in reducing the COF at the interface. [34,35]More importantly, the lubrication property of the hydrogel is highly dependent on the MPC content, and the COF values of the hydrogel further decrease to 0.038 and 0.022 along with the increase in the MPC content.
As a viscoelastic material, the lubrication performance of hydrogel is highly dependent on the applied load and frequency of the tribological test. [36,37]Consequently, the COF is investigated under different conditions by conducting the tribological tests with variable loads and frequencies for the PEGDAA-MPC group (PEGDAA:MPC = 1:2).The surface roughness of PEGDAA-MPC before the friction experiment is 87 nm (Figure 5b).When the applied load is 1 N and the reciprocating frequency is increased from 0.25 to 1 Hz, there is no obvious change on the COF value, which is slightly increased from 0.020 to 0.035 (Figure 5d).When the frequency is 0.5 Hz and the applied load is increased from 1 to 3 N, the COF value increases from 0.022 to 0.043 (Figure 5e).This is attributed to the elastic deformation of the hydrogel as the normal load increases.The contact stress and indentation depth also increase under high applied load, thus leading to a significant increase in the transverse friction force F x and correspondingly to an increase in the COF value. [38]It is noteworthy that the upper hydrogel layer is able to bond stably with the substrate following the tribological tests with different loads and frequencies, and no damage is observed at the interface for the composite scaffold.This indicates that the interfacial bonding strength based on mechanical interlocking and chemical grafting is very strong and stable.
The reason that the introduction of MPC to the hydrogel can significantly improve the lubrication property is attributed to the charged zwitterionic phosphorylcholine groups (N + (CH 3 ) 3 and PO 4 − ).Due to the interactions between the water dipole and the enclosed zwitterionic charges, the phosphorylcholine groups can attract water molecules to form a tenacious hydration layer around the charges.Based on the mechanism of hydration lubrication, the hydration layer is not only able to bear pressure without being squeezed out from the interface but also behaves in a fluid-like manner under shear, thus greatly reducing friction under various conditions. [39,40]Additionally, it is also owing to the interconnected porous structure and excellent water absorption capacity of the hydrogel, which contributes to maintaining good lubrication properties.

Biocompatibility, Osteogenic Ability, and Biomineralization
Excellent biocompatibility of implantable material is a prerequisite to achieve potential application in clinics.Cytotoxicity experiments are performed in two aspects to verify the biocompatibility of the hydrogel and PEEK substrate, respectively.Specifically, the biocompatibility is verified by co-culturing the chondrocytes with an extract from the hydrogel.6b shows that most of the cells are present as live cells and only a small amount of red fluorescence is observed, which is considered as a normal apoptotic phenomenon of the cells.
Figure 7a shows the relative survival of MC3T3-E1 osteoblasts co-cultured with PEEK, PDA, and nHA.The relative cell survival rate is more than 90% at all three-time points, indicating that the materials used are non-toxic and allow the cells to proliferate within 5 d.Alizarin red staining is used to characterize the ability of osteoblasts for producing calcareous nodules.Figure 7b shows the results of alizarin red staining of the specimens after coculturing with the cells for 7 d.The surface-modified PEEK displays significant deposition of calcareous nodules compared with bare PEEK.The PHA deposited on the surface by polydopamine coating provides the material with good osteogenic induction and bioactivity, significantly improving the original biological inertness of PEEK.
The bioactivity of surface-coated PEEK was inextricably associated with the formation of apatite-like deposits by biomineralization in simulated body fluids.Accordingly, this method is used to evaluate the biomineralization ability of PEEK and PEEK-PHA.The SEM images in Figure 7c show the results of biomineralization of the samples.After 7 d of incubation in SBF, no mineralization products are found on the surface of porous PEEK.In contrast, a large amount of mineralization products are clearly formed on the surface of PEEK-PHA.The reason may be due to the modification of PDA imparting a large number of free catechol groups on the surface of the specimens, which promotes interfacial integration, and the introduction of nHA is also considered to contribute to the mineralization process. [41,42]For PEEK-PHA, the implantation site is located the subchondral bone, and the material surface has an excellent ability for mineralization, which is important to the interfacial integration between the implant and surrounding bone tissues.
Furthermore, the in vivo biocompatibility of the composite scaffold implanted in rabbits is shown in Table 1 and Figure S4 (Supporting Information).It is observed that one and three months after surgery, the histopathological surroundings in the treatment group demonstrate no significant differences compared with those in the control group (all p-values > 0.05).The total scores of the treatment group are 0.60 and 1.60 higher than those of the control group at two different time points, respectively.However, the differences are not statistically significant, and it indicates that the postoperative histological circumstances around the composite scaffold are comparable to the physiological environment in the self-healing process of osseous tissues.Benefited from the great progress in materials science in recent years, the PEEK scaffold is currently applicable in biomedical fields, especially in orthopedic surgeries.For example, a PEEK cage manufactured by 3D printing is applied in lumbar interbody fusion operation, which is regarded as substantial proof for its biocompatibility. [43,44][47] Based on the pathological results of the present study, it is considered that the composite scaffold has good biocompatibility and is not toxic to the human body.

Conclusion
In this study, we developed a hard-soft heterogeneous bilayer composite scaffold, that PEEK-PHA/PEGDAA-MPC, for osteochondral defect replacement therapy.Based on oxidative covalent polymerization and non-covalent self-assembly, nHA was loaded on porous PEEK through a polydopamine coating.Subsequently, a hydrogel consisting of PEGDAA and MPC was polymerized on the top surface of the PEEK-PHA material by UV initiation to form a strong interfacial bonding by mechanical interlocking and chemical grafting.The tribological tests showed that the hydrogel was able to maintain relatively small COF values at different MPC contents and experimental conditions, which was attributed to the hydration lubrication mechanism of the zwitterionic phosphorylcholine groups.Even after a long period of friction, the strong bonding between the hydrogel and substrate at the interface was not greatly affected, and the morphology and function were well maintained.In addition, the material biocompatibility was verified by the in vitro cellular experiments using the hydrogel and PEEK substrate and the in vivo animal test using the composite scaffold.More importantly, the surface modification of the PEEK substrate improved the biological inertness and significantly enhanced the osteogenic ability as well as biomineralization ability.In conclusion, the integrated bionic scaffold prepared in this study was endowed with excellent lubrication, osteogenic, and load-bearing functions and thus may provide a promising solution for the clinical treatment of osteochondral defects.

Experimental Section
Materials and Reagents: MPC was purchased from Joy-Nature Co., Ltd, Nanjing, China.PEEK powder was purchased from Joinature Polymer Co., Ltd, Jilin, China.PEGDAA was purchased from Ponsure Biotechnology Co., Ltd, Shanghai, China.Tris-HCl (pH 8.5), SBF (pH 7.4), and alizarin red staining were purchased from Solarbio Biotech Co., Ltd, Beijing, China.Dopamine hydrochloride (98%) and nHA were purchased from Aladdin Bio-Chem Technology Co., Ltd, Shanghai, China.Chondrocytes and MC3T3-E1 cells were purchased from iCell Bioscience Inc., Shanghai, China.Other chemical and biological reagents were purchased from the Chemical Store of Tsinghua University, Beijing, China.
Preparation of Porous PEEK: The NaCl particles were ground in a mortar and sieved with a sieve to obtain a particle size less than 300 μm.Then the NaCl particles were mixed thoroughly with the PEEK powders at a mass ratio of 2:1.Then the mixed product was placed in a custommade stainless steel mold (20 mm × 15 mm) and compressed using a compressor at a pressure of 3 MPa.The prepared sheet specimens (size: 20 mm × 15 mm × 5 mm) were sintered using a muffle furnace (SX-G07123, CTJZH, China) at 365 °C for 6 h.Subsequently, the specimens were polished with 2000 grit sandpaper and cleaned with ethanol in an ultrasonic bath for 6 h and in deionized water for 48 h to adequately filter out the NaCl particles.Finally, the specimens were dried in an oven at 60 °C for 48 h to obtain porous PEEK.
Surface Functionalization of PEEK: In this part of experimental manipulation, the upper surface of the PEEK specimen was protected from any surface modification.First, porous PEEK (except the upper surface) was immersed in 0.5 m Tris-HCl buffer (pH 8.5) containing dopamine hydrochloride (2 mg mL −1 ).The samples were collected after 24 h of reaction at 37 °C and 100 rpm in a shaker (ZQPW-250, BLABOTERY, China) in the dark.The unattached dopamine was washed with deionized water, and the samples were lyophilized to obtain polydopamine (PDA)-modified porous PEEK (PEEK-PDA).Subsequently, 0.1 g of PDA-decorated nHA (PHA, the material preparation process was provided in Supporting Information) was uniformly dispersed in deionized water, and PEEK-PDA (except the upper surface) was immersed in the suspension and stirred vigorously under dark conditions at room temperature for 12 h to allow the complete adsorption of PHA onto PEEK-PDA.Finally, the samples were washed sufficiently with deionized water and freeze-dried to obtain PEEK-PHA.
Preparation of PEEK-Hydrogel Composite Scaffold: Briefly, 20 wt.% of PEGDAA, 0.5 wt.% of photo-initiator I2959, and different amounts of MPC (molar ratio PEGDAA:MPC = 2:1, 1:1, and 1:2) were dissolved in deionized water and ultra-sonicated for 30 min.Subsequently, nitrogen was passed for 30 min to discharge oxygen from the solution and to obtain the gel precursor solution.The PEEK-PHA substrate was fixed on a custom-made mold and the precursor solution was added dropwise to the untreated PEEK upper surface.After irradiation with a UV light source (Omnicure-2000, EXCELITAS, China) at a wavelength of 365 nm and an intensity of 80 mW cm −2 for 7 min, the solution was completely cross-linked to obtain the composite scaffold of PEEK-PHA/PEGDAA-MPC. Finally, the material was placed in deionized water for 24 h to remove unreacted monomers from the hydrogel.
Material Characterizations: FTIR spectra of MPC, PEGDAA, and PEGDAA-MPC were analyzed at a wavelength from 400 to 4000 cm −1 using a spectrometer (Nexus 670, Nicolet, USA).To observe the morphology of the microscopic network structure of the hydrogels, PEGDAA, PEGDAA-MPC, and the composite scaffold PEEK/PEGDAA-MPC were freeze-dried.Subsequently, the materials were observed by a field emission SEM (Quanta 200, FEI, Netherlands) at an accelerating voltage of 5 kV, which was associated with EDS to examine the elemental compositions.To further verify the compositional structure of the hydrogels, the elemental analysis of the lyophilized PEGDAA and PEGDAA-MPC hydrogels was performed by XPS (PHI Quantera II, Ulvac-Phi Inc., Japan) equipped with a 15 kV Mg K radiation source.
To analyze the swelling behavior of the hydrogels, PEGDAA and PEGDAA-MPC with different molar ratios were incubated in the SBF solution and deionized water at room temperature, respectively.The samples were taken out at three-time points of 0, 6, and 24 h, and water was wiped dry before being equilibrated and weighed.Subsequently, the hydrogels were lyophilized and weighed, and the swelling ratio (Q) was calculated by the following equation. [48]In this equation, W wet and W dry are the wet weight and lyophilized weight at equilibrium at the three-time points, respectively.The swelling ratio was calculated as the ratio of the mass of the hydrogels after swelling to the mass of the lyophilized hydrogels at a certain time point.
To examine the degradation stability of the hydrogels in different liquid environments, two typical test conditions were set up including SBF solution and SBF solution with type II collagenase at different concentration gradients (50 μg mL −1 , 200 μg mL −1 , and 1 mg mL −1 , the concentration was increased every 3 weeks).The overall duration of the degradation experiment was 9 weeks, which was performed in a shaker maintained at a constant temperature of 37 °C and 80 rpm.The solution was changed every 24 h.The hydrogels with different components were weighed immediately after preparation, noted as W 0 .The samples were removed at the indicated time points and weighed after wiping the surface water to keep the hydrogels in equilibrium, noted as W t .The weight change was calculated by the following Equation ( 2): [49] Weight change% = The static water contact angle was measured by the sessile drop method using a contact angle goniometer (OCA-20, Dataphysics Instruments, Germany).The sample was tested at least three times and the average value was calculated.
The compressive properties of the hydrogels were evaluated using a universal material testing machine (ZWICKZ020, ZWICK, Germany) equipped with a 20 kN pressure transducer under the compression mode.The hydrogels used for the compression test were made into a cylindrical shape (14 mm in diameter and 15 mm in height) using a custommade mold.During the test, the downward compression rate was kept consistently at 0.5 mm min −1 until the hydrogels ruptured and stressstrain curve was obtained.The test was repeated at least three times for the same condition to ensure data reliability.
To evaluate the rheological properties of the hydrogels, the prepolymerized hydrogels were dropped into a custom-made mold and formed into discs with a diameter of 8 mm and a thickness of 1 mm to fulfill the test requirements.The hydrogels with different amounts of MPC contents were selected and tested on a rheometer (MCR-302, ANTON PAAR, Austria) for modulus changes in the angular frequency variation range of 1-100 rad s −1 under the preload of 1 N and shear strain of 1% at room temperature.
Tribological Tests: The lubrication performance of the hydrogels was evaluated by a series of tribological experiments while the bonding strength between the PEEK substrate and hydrogel was observed.The tribological tests were performed using a universal material tester (UMT-5, Center for Tribology Inc., USA) under reciprocating mode.A PTFE sphere with a diameter of 8 mm was utilized as the upper specimen.The lower specimen, which was the composite scaffold to be tested, was fixed in a surface dish (attached with the test bench) by cyanoacrylate adhesive.Subsequently, SBF solution was added to the surface dish as the lubricant to simulate the physiological environment, and sufficient aqueous solution was maintained in the surface dish during the test.The test conditions were set as a stroke length of 4 mm and a test time of 15 min.To examine the COF of the materials under different conditions of modulation, the tribological experiments were performed with different MPC contents, different loads (1, 2, and 3 N), and different sliding frequencies (0.25, 0.5, and 1 Hz).Each test condition was repeated at least three times, and then the average COF values were recorded.
Cell Culture: Chondrocytes were used for in vitro cellular experiments to verify the biocompatibility of the upper hydrogel layer.Cell culture was performed in a thermostatic incubator with 5% CO 2 and 37 °C environment using -MEM as the specific medium.The cell culture medium was renewed every 24 h, and the cells were passaged when they reached ≈80% confluence.Additionally, mouse osteoblasts (MC3T3-E1) were used to verify the biocompatibility of the underlying PEEK substrate.The cell growth medium consisted of -MEM and fetal bovine serum at a 9:1 volume ratio, and the cells were cultured in a thermostatic incubator with 5% CO 2 and 37 °C environment.The cell culture medium was renewed every 48 h, and similarly, the cells were passaged when they reached ≈80% confluence.
Chondrocytes Proliferation and Morphology: Briefly, 10 mg of hydrogel sample was prepared and then sterilized with UV light for both the front and back sides.10 mL of complete medium was added, extracted after 24 h at 37 °C, filtered, and sterilized for use.The extracted product was diluted to the working solution concentration using a complete medium.The experimental groups were divided into the control group, PEGDAA group, and PEGDAA-MPC group (PEGDAA:MPC = 1:2).Regarding the control group, 0.5 mL well −1 of complete medium was added, and 0.5 mL well −1 of working solution with a concentration of 0.05 mg mL −1 was used for the remaining two groups, respectively.Three replicate wells were set up for each sample.The samples were inoculated into 48-well plates at 1 × 10 4 cells per well and incubated at 5% CO 2 and 37 °C for 24, 72, and 120 h.Afterward, the medium was removed and the wells were washed three times with PBS and incubated at 100 μL/well containing 10% CCK-8 medium at 5% CO 2 and 37 °C for 2 h.The absorbance values at 450 nm were measured by an enzyme marker (SPARK 10 m, TECAN, China) to calculate the relative cell viability.
Subsequently, 1 mL well −1 of complete medium was added to the control group, and 1 mL well −1 of working solution at a concentration of 0.05 mg mL −1 was added to each of the remaining two groups.The cells were seeded in 24-well plates at 2 × 10 4 cells per well and incubated at 5% CO 2 and 37 °C for 120 h.Reagent A (Calcein-AM) and Reagent B (PI) were diluted ten times with dye diluent, respectively.Then 985.5 μL of serumfree medium was mixed with 10 μL of diluted Reagent A and 4.5 μL of Reagent B. After that an appropriate amount of cell culture medium was added in the confocal dish, and the cells were washed in the dish once with PBS to remove excess serum.The staining solution was added at 1000 μL well −1 , incubated at room temperature for 15 min in the dark, and finally observed by a laser scanning confocal microscope (LSM-800, ZEISS, Germany).
Osteoblast Proliferation: The PEEK samples were sterilized in 75% ethanol solution for 2 h and then irradiated with UV light for 2 h.The CCK-8 assay was used to determine cell proliferation.Briefly, the MC3T3-E1 cell suspension was co-cultured with PEEK, PDA, and nHA at a density of 5 × 10 3 cells per well in 96-well plates.After incubation for 1, 3, and 5 d, the medium was removed and gently washed three times with PBS.The incubation was continued for 2 h.Then, 100 μL of working solution containing 10% CCK-8 was added to each well, and the absorbance values were measured at 450 nm using the enzyme marker.
Osteogenic Capability: Alizarin red staining was utilized to assess the osteogenic capacity of PEEK and PEEK-PHA.The samples were sterilized by UV light, immersed in the medium for 2 h, and then co-cultured with cell suspension at a density of 8 × 10 4 cells per well for 7 d at 37 °C and 5% CO 2 .The samples were rinsed three times with PBS at indicated time intervals and fixed with 4% paraformaldehyde at room temperature for 30 min.After removal of the fixative, the samples were rinsed three times with PBS.Subsequently, the cells were stained with 0.2% alizarin red solution for 30 min.After that, the specimens were rinsed three times with deionized water and dried in the air before observation using an optical microscope (Axiovert-40, ZEISS, Germany).
Biomineralization: The in vitro biomineralization activity of PEEK specimens were assessed by incubating them in SBF solution to produce bonelike apatite.Specifically, the porous PEEK and PEEK-PHA specimens were immersed in 20 mL of SBF solution at room temperature and incubated for 7 d.Then the samples were removed, rinsed with deionized water, and freeze-dried.The mineralization products on the surface were observed using the SEM.
In Vivo Evaluation: Twenty New Zealand adult rabbits were prepared for the in vivo experiments to evaluate the biocompatibility of the scaffold, which was approved by the Ethics Committee of Beijing Tsinghua Changgung Hospital (No.23459-6-01).On the day of scaffold implantation, the rabbits were anesthetized, and the left hindlimbs were disinfected with tinctures of iodine and ethyl alcohol.The knee joint was exposed via medial parapatellar approach, and lateral patella luxation made the whole distal femur completely visible.A reamer with 3 mm in diameter was used to drill into the lateral femoral condyle to create a 3 mm diameter and 10 mm depth socket for implantation of the scaffold.Ten rabbits were randomly assigned to the treatment group with the implantation of a scaffold into the socket.Another ten rabbits were classified into the control group and no further treatment (implantation) was performed after the socket was created.The articular capsule and skin were closed in a traditional manner.The soft bandage was placed on the operated limb and maintained for 48 h after surgery.The suture was removed two weeks later.Postoperative daily activity was not limited during the study.One month after surgery, five rabbits in the treatment group and five in the control group were euthanized, and their distal femurs were removed and immersed in 10% formalin for further pathological analysis.The remaining rabbits were euthanized three months after surgery, and the samples were harvested and prepared in a similar manner.The femoral condyles were dehydrated via a series of graded ethyl alcohol and then embedded in polymethyl methacrylate.The slices (100 μm thickness) consisting of the scaffold and surrounding tissues were prepared through microtome, and then stained by hematoxylin and eosin.An experienced histologist assessed the biocompatibility based on the presence or absence of associated necrosis, inflammatory infiltration, fibrosis, giant cells, foreign body debris, and fatty infiltration, and graded each aspect as 0 point (no reaction), 1 point (mild reaction), 2 points (medium reaction), and 3 points (severe reaction).The summation of points presented the biocompatibility of the scaffold and lower scores associated with better biocompatibility.Independent samples t-tests were used to evaluate the significant difference between various groups and a p-value of <0.05 was considered statistically significant.

Figure 1 .
Figure 1.The schematic diagram for the preparation of integrated PEEK-lubrication hydrogel composite scaffold biomimicking the articular cartilagesubchondral bone structure.

Figure 3 .
Figure 3. Surface morphology and elemental composition of the prepared hydrogels.SEM images associated with the EDS analysis of a) PEGDAA, b) PEGDAA-MPC, and c) Cross-section of PEEK/PEGDAA-MPC.

Figure 4 .
Figure 4. Material characterizations: a,b) Schematic diagram of the swelling stability of PEEK/PEGDAA-MPC.c) Swelling ratio of the hydrogels with different MPC contents.d,e) Resistance to degradation of the hydrogels with different MPC contents under the liquid environment of SBF and SBF & collagenase II.f) The water contact angle of PEEK, PEGDAA, and PEGDAA-MPC.g) Compressive strength of PEGDAA and PEGDAA-MPC hydrogels.h) Rheological property of hydrogels with different MPC contents.

Figure 5 .
Figure 5. Lubrication performance of the hard-soft composite scaffold: a) Schematic diagram of the tribological test and corresponding lubrication mechanism.b) Surface topography of PEGDAA-MPC examined by an optical interferometer.c) COF values of the hydrogels with different MPC contents.d,e) COF values of PEEK/PEGDAA-MPC (1:2) under different applied loads and frequencies.

Figure 6 .
Figure 6.In vitro biocompatibility of the hydrogels: a) Relative viability of chondrocytes co-cultured with PEGDAA and PEGDAA-MPC extracts at 1, 3, and 5 d using CCK-8 assay.b) Live and dead staining of the cells examined by laser scanning confocal microscopy.
Figure 6a displays the relative cell viability of PEGDAA and PEGDAA-MPC (1:2) hydrogel extracts after incubation with chondrocytes.The optical density values are examined at three time points 1, 3, and 5 d.The results indicate that the relative survival rate of cells in both hydrogels is higher than 90%, indicating that the material is able to support the proliferation of chondrocytes for 5 d with good biocompatibility.Subsequently, the live and dead staining of the cells is observed utilizing laser confocal photography at the end of culturing for 5 d.Calcein AM-labeled live cells are shown in green fluorescence, and propidium iodide (PI) labeled dead cells are demonstrated in red fluorescence.The fluorescence image in Figure

Figure 7 .
Figure 7.In vitro biomineralization of PEEK with surface modification.a) CCK-8 assay to determine the relative viability of MC3T3-E1 cells with PEEK, PDA, and nHA on 1, 3, and 5 d.b) Optical images of alizarin red staining of calcium nodules after culturing for 7 d.c) SEM images of PEEK and PEEK-PHA after biomineralization for 7 d.

Table 1 .
The biocompatibility of the composite scaffold is based on the comparison of histopathological surroundings with the control group.