1. Introduction
Osteoarthritis (OA) is a chronic degenerative joint disease characterized by cartilage degeneration. The main pathological features of OA are inflammation and attenuated cartilage regeneration [
1]. The excessive production of catabolic and pro-inflammatory mediators in OA-affected joints breaks down cartilage, exacerbates inflammation, dramatically reduces the number of chondrocytes, and limits the effect of anabolic growth factors [
2]. OA-affected chondrocytes have altered signaling activity and a tendency to shift to a hypertrophic phenotype, which increases the mechanical stress on cartilage and reduces the ability of cartilage to maintain its integrity [
3]. The currently available clinical treatments for OA include pharmacotherapy, surgery, and complementary therapy. However, none of these are effective permanent treatments [
4]. Recent advances in emerging therapies include those based on hydrogel scaffolds loaded with drugs, cells, genes, and bioactive molecules [
5]. Cell-based hydrogels can mimic tissue to a great extent
[6],
[7],
[8]. However, the implanted cells in these hydrogels do not survive for long periods, and clinical applications are further limited because of the high costs associated with individual patient differences and immunogenicity, as well as the cost of transport and preservation [
5]. Therefore, a more comprehensive treatment strategy is required to address the root cause of OA.
To reduce cartilage inflammation, promote chondrogenic differentiation, and maintain chondrocyte homeostasis, we focused on two small molecule therapeutics: dexamethasone (Dex) and kartogenin (KGN). Dex, one of the earliest and most readily available glucocorticoids, has anti-inflammatory properties. Dex is also a bone marrow stromal cell (BMSC) inducer, promoting differentiation of mesenchymal stem cells, and has recently been shown to modulate the early immune response by regulating the polarization of macrophages [
9,
10]. The polarization of macrophages has been implicated in cartilage repair and the progression of OA [
11]. The key to modulating the OA cartilage environment is the controlled release (time-dependent and concentration-dependent) and intracellular transport of Dex to control its anti-inflammatory effects and modulation of macrophage polarization. KGN promotes the differentiation of human mesenchymal stem cells (hMSCs) into chondrocytes in a dose-dependent manner [
12] and stimulates the expression of the runt-related transcription factor (RUNX) family, mainly that of RUNX1. RUNX1 has been shown to play a key role in chondrocyte formation, proliferation, and survival [
13,
14].
The cartilage tissue-like characteristics of hydrogels can provide a microenvironment with the appropriate geometry and mechanical parameters for hMSCs to adhere and then differentiate into chondrocytes and proliferate
[15],
[16],
[17],
[18]. Moreover, hydrogels can physically or chemically encapsulate small molecules, functional macromolecules, nanoparticles, and cells. By precisely designing the cross-linking density and degradability, hydrogels can achieve
in vivo controlled release of active ingredients over a period of days or months
[19],
[20],
[21].
Considering the pathological features of OA and the microenvironment that is required for cartilage regeneration, we used a hydrogel based on collagen, a natural material with good biocompatibility and degradability [
22]. Collagen hydrogel scaffolds support the growth, adhesion, and differentiation of mesenchymal stem cells and are used to construct cartilage
in vitro [
23,
24]. Silk protein, another natural protein, is multifunctional and easily processed [
25]. The treatment of silk protein with an organic solvent produces silk protein nanoparticles (SPNs) [
26], which have superior biocompatibility, degradability, and safety.
Our research aimed to address the difficulties of controlling inflammation and regenerating cartilage by designing a nanocomposite hydrogel scaffold with dual drug delivery. Different loading methods enabled different release mechanisms from the hydrogel of two therapeutic molecules (KGN and Dex molecules). The therapeutic effects of the molecules and the scaffolding effect of the nanocomposite hydrogel promoted the repair of OA defects through a spatiotemporal mechanism effecting anti-inflammation, immunomodulation, and promotion of cartilage differentiation (
Fig. 1). We envisage that this hydrogel-based system will offer three benefits over previously synthesized drug delivery systems. First, the two-protein (human-like collagen (HLC) and SPNs) drug delivery mechanism ensures that the hydrogel-based system is biocompatible, degradable, and safe. Second, the hydrogel-based system encapsulates and stabilizes biologically active drugs, and then provides for their controlled release through a two-protein drug delivery mechanism. The drug delivery mechanism controls the release of Dex in the early stages of treatment to modulate early immune responses and reduce inflammation. It also delivers KGN to cells with controlled release in the cell cytoplasm that promotes cell differentiation. Third, the hydrogel-based system is an integrated therapeutic system for OA that creates a microenvironment for cartilage regeneration. The scaffolding effect of the hydrogel and the therapeutic effect of the drugs synergistically induce rapid differentiation and regeneration of cartilage. Therefore, the insights that were generated from this study are of great significance for synthesizing advanced drug delivery systems with dual drug release mechanisms to repair articular cartilage tissue.
2. Materials and methods
2.1. Materials
The HLC (97 000 Da; China Patent No.: ZL01106757.8) was purchased from the Xi’an Juzi Biology Gene Technology Co., Ltd. (China), while the transglutaminase (TGase; 200 U∙g−1) was obtained from the Shanghai Yuanye Biotechnology Company (China). The KGN (relative molecular mass: 317.34 Da), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC), and N-hydroxysuccinimide (NHS) were obtained from Sigma-Aldrich (USA). Dex was purchased from Macklin Biochemical Technology Co., Ltd. (China).
2.2. Silk protein extraction
The cocoons were cleaned and subjected to degumming treatment by boiling them in a 5 g∙L−1 sodium carbonate solution for 30 min. Then, the dried degummed silk protein was dissolved in a ternary solution of calcium chloride/water/ethanol (molar ratio = 1:8:2) at 80 °C for 1 h. The solution was subsequently poured into a dialysis bag (molecular weight cutoff: 8 000–12 000 Da) and dialyzed in deionized water for 3 days to remove impurities. The dialyzed solution was then freeze-dried under vacuum conditions to obtain soluble silk protein. The extraction steps of the silk protein are illustrated in Fig. S1 in Appendix A.
2.3. Synthesis of KGN-loaded SPNs
The KGN (0.4 mg∙mL−1) was dissolved in dimethyl sulfoxide (DMSO), and EDC and NHS were added to the solution (the molar ratio of KGN/EDC/NHS was 1:1:1). The solution was stirred for 1 h and added dropwise to the silk protein solution (4 mg∙mL−1). After reacting for 2–4 h, the solution was centrifuged at 14 000 r∙min−1 for 5 min to remove any polymers that may have been generated. The excess impurities in the solution were removed by dialysis over the course of 24 h, and the solution was freeze-dried under vacuum conditions to obtain KGN-loaded silk protein. KGN-loaded SPNs (SPNs–KGN) were synthesized by preparing a 2% KGN-loaded silk protein solution and adding the solution dropwise to acetone at a volume ratio of 1:20 while stirring for 2–4 h. The mixture was centrifuged at 100 000 r∙min−1 for 30 min and washed with distilled water. The suspension was subsequently washed at least two times to produce a homogeneous suspension of SPNs–KGN, which was stored at 4 °C. The SPNs without KGN were prepared using pure acetone.
2.4. Characterization of SPNs–KGN
The concentration of KGN in KGN-loaded silk protein was determined using high-performance liquid chromatography (HPLC; Alliance e2695, Waters Corporation, USA). The linear regression equation (Fig. S2 in Appendix A) for the KGN concentration and peak area of the drug was first determined. The pre-reaction KGN solution and the post-reaction dialysate were collected, and the concentration of unloaded KGN was determined using HPLC and calculated from the regression equation. The encapsulation efficiency and drug loading of KGN were then calculated (Eqs. (S1) and (S2) in Appendix A). The details are described in the Supplementary data in Appendix A.
The particle size distribution, zeta potential, and polydispersity of the SPNs–KGN were measured using a dynamic light scattering spectrophotometer (Zetasizer Nano ZS, Malvern Panalytical, UK). The circular dichroism of the nanoparticles was determined using a circular dichroism spectrometer (BRIGHTTIME Chirascan, Applied Photophysics, UK) with a scan range of 190–250 nm and wavelength step of 2 nm. The SPNs–KGN nanoparticles were dried at 25 °C and coated with gold before being observed using scanning electron microscopy (SEM; JSM-IT800, JEOL Ltd., Japan). 1H nuclear magnetic resonance (NMR, Avance III HD 400, Bruker Corporation, Germany) spectroscopy was used to characterize the surface chemistry of the nanoparticles.
2.5. Synthesis of drug-loaded nanocomposite hydrogels
Dex was dissolved in deionized water at a concentration of 5 mg∙mL−1. The Dex solution was added to HLC (20 mg∙mL−1), and the mixture was heated in a 37 °C water bath to dissolve the HLC completely. After fully reacting, the solution was dialyzed to remove uncross-linked small molecules and freeze-dried under vacuum conditions to obtain HLC(Dex). The loading of Dex on HLC was determined through HPLC using a C18 reversed-phase column. The linear regression equation (Fig. S2) for the Dex concentration and peak area of the drug was first determined. The pre-reaction Dex solution and the post-reaction dialysate were collected, and the concentration of unloaded Dex was determined using HPLC and calculated from the regression equation. The encapsulation efficiency and drug loading equations were similar to those for KGN. The details are described in the Supplementary data in Appendix A. SPNs–KGN (1 mg∙mL−1) were dissolved completely in deionized water, and 100 mg∙mL−1 of HLC(Dex) was added to the solution. After being thoroughly mixed, a TGase solution (60 U∙g−1 of protein) was added, and the mixture was reacted at 4 °C for 6–10 h to obtain the HLC(Dex)–SPNs–KGN hydrogel. The following hydrogels were prepared for this study: HLC–SPNs hydrogel, which did not contain Dex or KGN; HLC hydrogel, which did not contain Dex, KGN, or SPNs; HLC–SPNs–KGN hydrogel, which was formed when HLC and SPNs–KGN were mixed and cross-linked using TGase; and HLC(Dex)–SPNs hydrogel, which was formed when HLC(Dex) and SPNs were mixed and cross-linked using TGase.
2.6. Characterization of HLC(Dex)–SPNs–KGN hydrogel
HLC(Dex)–SPNs–KGN hydrogels were frozen in liquid nitrogen immediately before cryo-electron microscopy (PP3000T Cryoprep transfer system (Quorum, UK), FEI Quanta 450 environmental scanning electron microscope (Field Electron and Ion Company, USA)) to observe the internal pore structure.
To characterize the swelling behavior of the hydrogel, the hydrogel was immersed in phosphate-buffered saline (PBS) and removed at certain time intervals. The hydrogel was weighed until a constant weight of the hydrogel was obtained. The swelling rate of the hydrogel was calculated using Eq. (S3) in Appendix A.
The rheological properties of the hydrogel were tested using a rheometer (MCR302, Anton Paar, Austria) set to oscillatory mode at a 5% strain, 1 rad∙s−1 frequency, and 25 °C temperature. The cyclic compression test was performed on the hydrogel using an INSTRON 5565 testing machine (Instron, USA). The modulus data of the cylindrical hydrogel (1 cm diameter and 1 cm thickness) were automatically calculated by the testing machine.
To determine the degree of gel formation, the gel fraction of the hydrogel was calculated by measuring the weight of the hydrogel before and after formation. The steps were as follows: The components of the hydrogel were mixed to determine the mass of the whole component, after which the mixture was stirred to form a hydrogel. The hydrogel was removed, the surface water was absorbed with filter paper, and the hydrogel was weighed again. The gel fraction was then calculated (Eq. (S4) in Appendix A).
2.7. In vitro drug release studies
HLC(Dex)–SPNs–KGN hydrogel (100 mg) and SPNs–KGN (10 mg) were immersed in PBS (1 mL, pH 7.4) under shaking (100 r∙min−1) at room temperature. The HLC(Dex)–SPNs–KGN hydrogel (100 mg) and SPNs–KGN (10 mg) were also immersed in PBS (1 mL, pH 7.4) with 1 U∙mL−1 collagenase (S10053, Shanghai Gene Biotechnology Co., Ltd., China) under shaking (100 r∙min−1) at room temperature. At predetermined time intervals, 0.1 mL of the release solution was withdrawn, and the peak area was determined using HPLC to calculate the total release rate. Then, 0.1 mL of fresh PBS or PBS with 1 U∙mL−1 collagenase was added to the solution to maintain a constant total volume, and the dialysis test continued. The amount of KGN and Dex released from the hydrogel was determined using HPLC analysis.
2.8. Cytocompatibility studies
The cytotoxicity of the HLC(Dex)–SPNs–KGN hydrogel was determined using a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. To obtain an extract of the hydrogel, the hydrogel was immersed in culture medium at a concentration of 100 mg∙mL−1 for 24 h at 37 °C. Human articular chondrocytes (HACs) and hMSCs were seeded in a 96-well plate at a density of 1 × 104 cells per well and treated with the hydrogel extract. After 24 h of incubation, the extract was replaced with PBS containing the MTT reagent and DMSO. The absorbance of each well was measured at 490 nm. Cell viability was determined using Eq. (S5) in Appendix A. For acridine orange/ethidium bromide (AO/EB) staining of HACs and hMSCs that were treated with the extracts of different hydrogel samples (HLC hydrogel, HLC–SPNs hydrogel, HLC(Dex)–SPNs hydrogel, and HLC(Dex)–SPNs–KGN hydrogel), HACs and hMSCs (1 × 104 mL−1) were plated onto a 96-well plate. After 24 h of incubation, the extracts of the hydrogel samples were added to the plate. The cells were incubated for 3 days with the extracts, then the extracts were removed, and the wells were cleaned with PBS. HACs and hMSCs were stained using an AO/EB cell viability kit and photographed using a fluorescence microscope. To evaluate the adhesion of HACs and hMSCs on the HLC(Dex)–SPNs–KGN hydrogel, the hydrogel was inserted into a 48-well plate and soaked in cell culture medium for 24 h. HACs and hMSCs (1 × 105 mL−1) were seeded in each well and incubated at 37 °C. After 2 days, the hydrogel was washed with PBS, 2.5% glutaraldehyde was added, and the hydrogel was fixed for 10 h. An alcohol gradient was used to dehydrate the hydrogel, and the hydrogel was observed using SEM.
2.9. Cell endocytosis of SPNs–KGN
Mesenchymal stem cells were cultured at a concentration of 1 × 105 cells per well. The HLC(Dex)–SPNs–KGN hydrogel (100 mg) was added to each well for co-culture. After 1 and 3 h, cells were collected by digestion and centrifugation, and a transmission electron microscopy (TEM; Tecnai G2 Spirit, Thermo Fisher Scientific, USA) fixative was added. The morphology of the cells was observed using TEM. The details are described in the Supplementary data in Appendix A.
2.10. In vitro anti-inflammatory activity studies
In vitro anti-inflammatory activity experiments were conducted using mouse monocyte/macrophage-like cells (RAW264.7 cells). First, RAW264.7 cells were inoculated onto sterile coverslips and incubated for 24 h, and the supernatant was carefully removed. The treatment for each group was added to the cells, and the cells were incubated for 24 h. The cells were washed with PBS, fixed with 4% paraformaldehyde, subjected to immunofluorescence staining (interleukin (IL)-10 (20850-1-AP) and tumour necrosis factor (TNF)-α (ab1793), Abcam, UK), and observed using a laser confocal microscope (A1, Nikon, Japan). The experimental groups included the normal RAW264.7 cell (control) group; lipopolysaccharide (LPS) group (positive control, 10 ng∙mL−1 LPS solution); HLC–SPNs, HLC(Dex)–SPNs, and HLC(Dex)–SPNs–KGN hydrogel groups (treated with 50 mg∙mL−1 hydrogel extract); and LPS + HLC–SPNs, LPS + HLC(Dex)–SPNs, and LPS + HLC(Dex)–SPNs–KGN hydrogel groups (treated with 10 ng∙mL−1 LPS solution and 50 mg∙mL−1 hydrogel extract).
2.11. In vitro assays on chondrogenic differentiation of hMSCs
Sterilized hydrogels (HLC–SPNs hydrogel, HLC(Dex)–SPNs–KGN hydrogel, HLC–SPNs–KGN hydrogel, and HLC(Dex)–SPNs hydrogel) were immersed in an induction medium, and hMSCs and the hydrogels were incubated together for 10 days. The experimental groups were the blank control group, which was the induction medium alone; the positive control group, which was the induction medium with 600 nmol∙L−1 of KGN; the HLC–SPNs hydrogel group, which was the induction medium incubated with 100 mg∙mL−1 of HLC–SPNs hydrogel; the HLC(Dex)–SPNs–KGN hydrogel group, which was the induction medium incubated with 100 mg∙mL−1 of HLC(Dex)–SPNs–KGN hydrogel; the HLC–SPNs–KGN hydrogel group, which was the induction medium incubated with 100 mg∙mL−1 of HLC–SPNs–KGN hydrogel; and the HLC(Dex)–SPNs hydrogel group, which was the induction medium incubated with 100 mg∙mL−1 of HLC(Dex)–SPNs hydrogel. The hMSCs were fixed in 4% paraformaldehyde at 37 °C for 30 min, stained with toluidine blue (Solarbio, China), and observed using a microscope. Immunofluorescence staining was performed after the hMSCs were fixed. The antibodies were against type II collagen (Col II), cartilage oligomeric matrix protein (COMP), aggrecan, and SOX-9. The hMSCs were observed using a confocal fluorescence microscope (Nikon A1).
The Co-60-irradiated sterilized hydrogels (diameter of 10 mm and height of 2 mm; HLC–SPNs hydrogel, HLC(Dex)–SPNs hydrogel, HLC–SPNs–KGN hydrogel, and HLC(Dex)–SPNs–KGN hydrogel) were immersed in chondrogenic differentiation medium for 6 h, and then, the hMSCs were seeded in each hydrogel at 2 × 104 cells. The hydrogels with cells were cultured in chondrogenic differentiation medium for 5 and 10 days. The cells were digested and collected for enzyme-linked immunosorbent assay (ELISA) and quantitative real-time polymerase chain reaction (qPCR) assay for chondrogenic-related proteins and genes, respectively. The details are described in the Supplementary data in Appendix A.
2.12. qPCR assay for detecting gene expression in three-dimensional cultured cells
Cells were cultured according to the method described in Section 2.11, the total cellular RNA was extracted using TRIzol reagent (Takara, Japan) according to the manufacturer’s instructions. First-strand complementary DNA (cDNA) was synthesized with a Prime Script RT reagent kit (Takara). qPCR was performed using a Bio-Rad qPCR system (Bio-Rad, USA) for COMP, COL2A1, ACAN, and SOX9. The values obtained were normalized to the negative control as the fold change. Details on the experimental methods are available in the Supplementary data in Appendix A.
2.13. In vitro degradation and anti-inflammatory effects of the hydrogels
All animal experiments were reviewed by the Northwest University Laboratory Animal Management and Ethics Committee. The experiments complied with the legal and institutional guidelines related to animal ethics and were approved by the Ethics Committee (NWU-AWC-20210810R). The degradation of the hydrogel samples was studied by implanting the samples under the skin (i.e., subcutaneous) on the backs of rabbits on week 1, 5, and 10. The skin tissues were collected and fixed in 4% neutral paraformaldehyde. The degraded hydrogel samples were freeze-dried under vacuum conditions, and the structure of the samples was observed using SEM. The freeze-dried, degraded hydrogel samples were also weighed to calculate the degradation rate. Hematoxylin and eosin (H&E) staining and immunohistochemical staining for TNF-α were performed on the skin tissues. Details on the experimental methods are available in the Supplementary data in Appendix A.
2.14. Establishment of inflammatory cartilage defects and implantation of the hydrogels in rabbit knee joints
First, 0.3 mL of 4% papain saline solution was injected into each rabbit’s hip joint on day 1, 4, and 7. After 7 days, the rabbits were anesthetized, and inflammatory cartilage defects (with a 5.0 mm diameter and 2.5 mm thickness) were established in the rabbits’ hindlimbs using a surgical drill. Twenty-seven rabbits were divided into the control group, HLC–SPNs hydrogel group, and HLC(Dex)–SPNs–KGN hydrogel group. The cartilage defects on the knee joints of rabbits in the control group were untreated. The cartilage defects on the knee joints of rabbits in the hydrogel groups were filled with the respective hydrogel samples to the surface of the native cartilage, which was followed by ligament repositioning and wound closure (suturing the periosteum as well as the skin, Fig. S3 in Appendix A). The details are described in the Supplementary data in Appendix A.
2.15. Evaluation of cartilage repair
Rabbit knee joints were collected for observation and sample analysis. The International Cartilage Repair Society (ICRS) classification system was used to determine the repair status of the surgical sites on the knee joints (Table S2 in Appendix A) by assessing the appearance of the defects at the surgical sites and the connection between the defects and their surrounding areas. A micro-computed tomography (micro-CT) scanner (SKYSCAN 1172, Bruker, USA) was used to perform qualitative and quantitative analyses of tissue regeneration at the surgical sites. The measurement parameters were bone mineral density (BMD), bone volume/total volume ratio (BV/TV), trabecular thickness (Tb.Th), and trabecular spacing (Tb.Sp). For the histological evaluation, H&E staining was used to assess the distribution of cells and fibrous tissues on the knee joints, and safranin O staining was used to detect the presence of glycosaminoglycans in the tissues of the knee joints. H&E and safranin O staining images were used to perform ICRS visual histological scoring (Table S3 in Appendix A). Immunohistochemical staining was performed for TNF-α, a disintegrin and metalloproteinase with thrombospondin motifs 5 (ADAMTS5), and IL-6. Immunofluorescence staining was performed for bone morphogenetic protein (BMP)2, transforming growth factor (TGF)-β1, cluster of differentiation (CD)68, CD86, CD206, RUNX1, SOX-9, Col II, aggrecan, COMP, CD90, and CD73. The displacement–force curve, modulus of elasticity, and hardness of the cartilage damage sites were detected by nanoindentation (Step 500-NHT3/NHT2, Anton Paar, Austria). The details are described in the Supplementary data in Appendix A.
2.16. Statistical analysis
Data are presented as means ± standard deviations. Statistical analysis was performed using SPSS software. One-way or two-way analysis of variance (ANOVA) was used to analyze significant differences at different thresholds: *p < 0.05; **p < 0.01; and ***p < 0.001.
3. Results
3.1. Characterization of SPNs–KGN
The carboxyl group in the molecular structure of KGN is activated by EDC and NHS to couple with the amino group of silk protein (
Fig. 2(a)) to obtain KGN-loaded silk protein. The drug loading and encapsulation efficiency of KGN were adjusted by controlling the concentrations of KGN and silk protein (Table S4 in Appendix A). The selected concentrations of silk protein and KGN were 4 and 0.4 mg∙mL
−1, respectively. The drug loading and encapsulation efficiencies of KGN of the corresponding KGN-loaded silk protein were 2.05% ± 0.32% and 20.30% ± 4.76%, respectively. Organic solvents such as acetone can cause the molecular chain of silk protein to undergo conformational transformation into a β-fold structure and form nanoscale assemblies (
Fig. 2(b)). The circular dichroism of the SPNs–KGN (
Fig. 2(c)) showed negative ellipticity at 217, indicating a conformational transformation of the silk protein from a random conformation to a β-fold structure. The inset of
Fig. 2(c) shows the colloidal characteristics (the Tyndall effect) of the SPNs–KGN. The
1H NMR spectrum of KGN displayed characteristic peaks of aromatic protons, which were located between 7.3 and 7.9 ppm (
Fig. 2(d-i), peaks 1–13). The
1H NMR spectrum of the silk protein displayed the characteristic peaks of Tyr–CH
2, which were located between 2.74 and 2.89 ppm (
Fig. 2(d-ii)); the characteristic peaks of Gly/Ser–CH
2, which were located between 3.41 and 3.54 ppm (
Fig. 2(d-iii)); and the characteristic peak of R–NH
2, which were located at 2.36 ppm (
Fig. 2(d-iv)). The
1H NMR spectrum of the KGN-loaded silk protein displayed the major peaks of the
1H NMR spectrum of KGN and silk protein. Moreover, the carboxyl group of KGN conjugated with the amino group of the silk protein and formed an amide bond. Thus, the
1H NMR spectrum of the KGN-loaded silk protein revealed the major peaks of ArCONHR, which were located between 6.22 and 6.35 ppm (
Fig. 2(d-v)). The average particle size and zeta potential of SPNs–KGN were (211.80 ± 8.23) nm and (6.25 ± 0.36) mV, respectively (
Fig. 2(e) and Table S5 in Appendix A). The nanoparticle morphology was observed by SEM and appeared irregularly spherical, as shown in
Fig. 2(f).
3.2. Characterization of HLC(Dex)–SPNs–KGN hydrogel
3.2.1. Morphology and fundamental performance
After Dex and HLC were mixed, Dex was cross-linked with the HLC network through hydrogen bonding. The HLC(Dex) compound was mixed with SPNs–KGN, and the mixture was cross-linked using TGase to form the HLC(Dex)–SPNs–KGN hydrogel. The phosphate group of Dex reacted with the –OH group of the HLC network (
Fig. 3(a)). As shown in
Fig. 3(b), in comparison with the Fourier-transform infrared spectroscopy (FTIR) spectrum of HLC, the FTIR spectrum of HLC(Dex) showed decay of the peak correlating with the stretching vibration of –OH (3200–3700 cm
–1), decay of the peak correlating with the stretching vibration of the –OH of carboxyl groups (2915 cm
–1), and the appearance of a peak correlating with the P–O stretching vibration (1090 cm
–1). The peaks correlating with the P–O bending vibrations (1777 cm
–1) and the symmetrical P–O stretching vibration suggested that the hydroxyl groups of Ser, Tyr, Thr, and Arg in HLC or the carboxyl groups of Asp and Glu in HLC reacted with the phosphate group of Dex to form hydrogen bonds.
The
1H NMR spectra of HLC(Dex) showed characteristic peaks of H at different positions of Dex, indicating formation of an HLC(Dex) complex
(Fig. S4 in Appendix A
). The encapsulation efficiency and drug loading of Dex were 4.12% ± 0.18% and 1.15% ± 0.21%, respectively. Thus, there was approximately 1 mg of Dex per gram of HLC(Dex)–SPNs–KGN hydrogel. The HLC(Dex)–SPNs–KGN hydrogel was formed by mixing HLC, Dex, and SPNs–KGN, with TGase and incubating the mixture at 4 °C for 12 h. The milky and translucent hydrogel had an irregular porous structure (
Fig. 3(c)).
Fig. 3(d) shows the pore structure within the hydrogel in the wet state obtained using cryo-electron microscopy. The pore size of the HLC(Dex)–SPNs–KGN hydrogel was primarily distributed between 10 and 30 μm, which supported an appropriate oxygen tension and nutrient supply. The magnified view shown in
Fig. 3(d) demonstrated that there were many SPNs–KGN within the structure of the HLC(Dex)–SPNs–KGN hydrogel.
3.2.2. Release of KGN and Dex
As shown in
Fig. 3(e), the HLC(Dex)–SPNs–KGN hydrogel demonstrated an immediate release of Dex followed by a sustained release of KGN, which was controlled by the loading mechanisms of Dex and KGN. On the first day, unstable hydrogen bonds between HLC and Dex on the external surface of the hydrogel were broken, leading to the release of Dex. The swelling process also led to the rapid release of a small amount of Dex inside the hydrogel when hydrogen bonds were broken. This small but sudden release of Dex ended after the hydrogel swelling reached equilibrium. However, a similar sudden release of KGN, which was controlled by the HLC network and SPNs–KGN, was not significant. As the HLC backbone of the hydrogel degraded, approximately 80% of the encapsulated Dex was released within 40 days. In comparison, 40% of the KGN was released within 40 days. The gel fraction of the nanocomposite hydrogel was 95.15% ± 0.39%, with a small amount of liquid remaining during gel formation, indicating that there was some drug loss. Therefore, the release rate in the drug release experiment was less than 100%. The release behavior of KGN and Dex in the presence of type I collagenase was similar to that in PBS, except that the rate of drug release from the individual carriers became faster with time (
Fig. 3(f)) as the hydrogel degradation released more drug. More than 90% of the KGN in the SPNs–KGN was released in 4 days, nearly 100% of the Dex was released from the HLC(Dex)–SPNs–KGN hydrogel in 10 days, and more than 80% of the KGN in the HLC(Dex)–SPNs–KGN hydrogel was released in 30 days. The release of KGN required first the release of the SPNs–KGN, as the nanoparticles were covalently cross-linked with HLC molecules, and therefore their release followed the degradation of the HLC backbone of the HLC(Dex)–SPNs–KGN hydrogel. KGN molecules were released by the degradation of the SPNs–KGN, and therefore the release of KGN was significantly slower than that of Dex. Thus, the release of KGN from HLC(Dex)–SPNs–KGN was slowed as it required the degradation of the HLC gel network and the SPNs. The HLC backbone protected the SPNs–KGN, which were released by the degradation of the HLC molecular backbone. Then, the SPNs were further degraded and released KGN molecules (
Fig. 3(g)). The release of KGN from the HLC(Dex)–SPNs–KGN hydrogel was slow because covalent bonds are more resistant to hydrolysis than physical interactions, such as van der Waals and electrostatic interactions.
3.2.3. Swelling and mechanical properties of the HLC(Dex)–SPNs–KGN hydrogel
Fig. 3(h) shows the swelling rates of the HLC, HLC–SPNs, and HLC(Dex)–SPNs–KGN hydrogels in PBS. The hydrogel samples underwent three stages of swelling in PBS before achieving equilibrium. There were no significant differences in the swelling rates of the hydrogel samples, which may be because they had the same cross-linking patterns. The three hydrogels reached swelling equilibrium in 120 min, with equilibrium swelling rates lower than 400%.
As cartilage is constantly subjected to mechanical stress, the viscoelasticity and fatigue resistance of the HLC(Dex)–SPNs–KGN hydrogel were evaluated in dynamic rheological and cyclic compression experiments, respectively. The fatigue resistance of the HLC(Dex)–SPNs–KGN hydrogel was evaluated by cyclic compression tests. As shown in
Fig. 3(i) (inset figure), the hydrogel returned to its initial shape after the compressive load was released; however, the cyclic compression curve was a non-linear closed curve, indicating hysteresis. After 20 compression cycles, the compression modulus of the hydrogel ((102.52 ± 8.42) kPa) did not significantly differ from the initial compression modulus of the hydrogel, indicating that the hydrogel had good fatigue resistance and may withstand the cyclic mechanical stresses of daily activities.
Fig. 3(j) shows the changes in the storage modulus of the HLC, HLC–SPNs, and HLC(Dex)–SPNs–KGN hydrogels as a function of time. The storage modulus of the HLC–SPNs hydrogel ((146.45 ± 2.34) Pa) and HLC(Dex)–SPNs–KGN hydrogel ((147.57 ± 2.37) Pa) were higher than that of the HLC hydrogel ((104.29 ± 3.64) Pa), which suggested that SPNs enhance the mechanical properties of the hydrogel samples due to the covalent cross-linking between SPNs and HLC molecules.
3.3. In vitro cytocompatibility and anti-inflammatory properties
Fig. 4(a) shows that the HLC(Dex)–SPNs–KGN hydrogel promoted the proliferation of HACs and hMSCs. The results of AO/EB staining of the cells (
Fig. 4(b)) showed that the HLC(Dex)–SPNs–KGN hydrogel had a significantly higher cell density than that of the control. Moreover, the morphology of the cells of the HLC(Dex)–SPNs–KGN hydrogel was similar to that of the cells of the control, which indicated that the hydrogel had good cytocompatibility. Cellular implantation is an important method for evaluating the cytocompatibility of cartilage scaffolds.
Fig. 4(c) shows the cytocompatibility of the internal structure of the HLC(Dex)–SPNs–KGN hydrogel for accommodating the three-dimensional (3D) growth of hMSCs and HACs. The SEM images show the uniform distribution of hMSCs and HACs in the hydrogel, in which the hMSCs and HACs adhered to the surface and pores of the hydrogel and maintained their cell morphology, indicating the cytocompatibility of the hydrogel as a cartilage scaffold with good adhesion of the cells to the hydrogel. We investigated how the SPNs–KGN released from the HLC(Dex)–SPNs–KGN hydrogel entered the cells, with the results shown in
Figs. 4(d) and
(e). The TEM results showed that at 1 h, the SPNs–KGN were distributed around the cell membrane, and after 3 h, endocytic vesicles wrapped around the SPNs–KGN in the cytoplasm, which indicated that SPNs–KGN entered the cytosol of the cell.
To confirm the anti-inflammatory activity of the HLC(Dex)–SPNs–KGN hydrogel, we added an extract of the hydrogel to RAW264.7 macrophages in which an anti-inflammatory response had been activated. LPS is an endotoxin that elicits a cellular immune response and induces the differentiation of RAW264.7 macrophages into pro-inflammatory phenotypes. The inflammatory factor TNF-α was more abundantly expressed in macrophages in the LPS group than in the normal control environment (
Figs. 4(f) and
(g); a single channel picture is shown in Fig. S5 in Appendix A). We found that Dex-loaded HLC(Dex)–SPNs and HLC(Dex)–SPNs–KGN hydrogel extracts significantly induced the production of the anti-inflammatory factor IL-10 and reduced the expression of the pro-inflammatory factor TNF-α by RAW264.7 macrophages in an inflammatory environment.
3.4. Promotion of chondrogenic differentiation of hMSCs
To confirm whether the HLC(Dex)–SPNs–KGN hydrogel could induce the differentiation of hMSCs into chondrocytes (
Fig. 5(a)), we performed toluidine blue staining on the hMSCs that were induced with an extract of the hydrogel. When combined with the acidic materials in chondrocytes, the cations in the toluidine blue staining solution produce a puce color. As shown in
Fig. 5(b), no puce acidic mucopolysaccharides were observed in the hMSCs of the control on day 10, whereas a portion of the hMSCs with the HLC(Dex)–SPNs–KGN hydrogel extracts developed a puce color on day 5 (the arrows point puce indicating the production of acidic mucopolysaccharides). The areas that were covered by the purple color expanded by day 10. In contrast, the induction period for chondrogenic differentiation of the hMSCs of the control was 21 days, which indicated that the hydrogel promoted the differentiation of hMSCs into chondrocytes. The differentiation of hMSCs into chondrocytes upon induction by the hydrogel extracts was further examined by immunofluorescence staining for the expression of COMP, Col II, aggrecan, and SOX-9 proteins.
Figs. 5(c)–(f) show that the expression of COMP, Col II, aggrecan, and SOX-9 proteins in the control hMSCs and the HLC–SPNs hydrogel extract-treated hMSCs were very low. In contrast, the proteins were significantly expressed in hMSCs treated with KGN and the HLC(Dex)–SPNs, HLC–SPNs–KGN, and HLC(Dex)–SPNs–KGN hydrogel extracts. The increased expression of the proteins in the hMSCs treated with HLC–SPNs–KGN and HLC(Dex)–SPNs–KGN hydrogel extracts was more significant than that in the hMSCs treated with KGN (
Fig. 5(g)). Additionally, we cultured hMSCs directly on the hydrogels (
Fig. 5(h)). The cells were digested and collected on day 5 and day 10, and their differentiation was examined using ELISA kits and qPCR. The qPCR assay results are shown in
Fig. 5(i). Similar to the immunofluorescence results, KGN-loaded hydrogels (HLC(Dex)–SPNs–KGN and HLC–SPNs–KGN) significantly enhanced
COMP,
COL2A1,
ACAN, and
SOX9 gene expression. HLC(Dex)–SPNs hydrogel exerted a pro-differentiation effect similar to that of the HLC–SPNs hydrogel, which indicated the synergistic effect by which Dex and KGN induced chondrogenic differentiation of hMSCs. ELISA kits assay results (Fig. S6 in Appendix A) showed the same trend.
3.5. In vivo cartilage repair studies on HLC(Dex)–SPNs–KGN hydrogel
We established an O model (inflammatory cartilage defects) and implanted hydrogels in rabbit knee joints.
Fig. 6(a) shows the efficacy of the HLC–SPNs and HLC(Dex)–SPNs–KGN hydrogels for promoting cartilage regeneration in experimental OA rabbits. After the first month, the control group joints showed large depressions with little regenerating tissue. In contrast, the HLC–SPNs and HLC(Dex)–SPNs–KGN hydrogel groups showed integration of the hydrogel scaffolds into the host cartilage tissue. After the second month, the control defect remained visible, and the borders around the defects were well defined. In contrast, the borders around the defects in the HLC–SPNs hydrogel group disappeared, and a large amount of new granulation tissue was visible, while the defects in the HLC(Dex)–SPNs–KGN hydrogel group were completely filled with new cartilage tissue. By the third month, the control defect remained partially empty, and a gap between adjacent areas of cartilage tissue was evident. The HLC–SPNs hydrogel defect showed a lack of continuity between new and native cartilage tissue and had a slightly rough surface. The boundary between the implanted HLC(Dex)–SPNs–KGN hydrogel and the surrounding tissue disappeared, and the hydrogel was integrated into the native cartilage tissue. The repaired cartilage surface had a smooth surface. The cartilage tissue of the knee joints was assessed using the ICRS cartilage repair assessment criteria, and the results are shown in
Fig. 6(b). The control and HLC–SPNs hydrogel groups showed grade III (abnormal) cartilage repair. In contrast, the HLC(Dex)–SPNs–KGN hydrogel group showed grade II (near normal) cartilage repair.
Micro-CT analysis was performed on the articular knee joints, and two-dimensional projection images and 3D reconstruction images obtained by tomographical analysis were used to assess the cartilage repair. The results are shown in
Fig. 6(c). After the first month, no new cartilage tissue was observed on the defects in the control group. However, the HLC(Dex)–SPNs–KGN hydrogel group showed some initial cartilage repair. By the third month, the defects of the HLC(Dex)–SPNs–KGN hydrogel group were completely filled and well-integrated into the surrounding tissue. The new cartilage tissue had a similar structure to the native cartilage tissue. Furthermore, as shown in
Figs. 6(d)–(h), the BMD, trabeculae number (Tb.N), Tb.Th, and BV/TV of the HLC(Dex)–SPNs–KGN hydrogel group were significantly higher than those of the HLC–SPNs hydrogel and control groups. Moreover, the Tb.Sp of the HLC(Dex)–SPNs–KGN hydrogel group was significantly lower than that of the control group. These results suggested that the HLC(Dex)–SPNs–KGN hydrogel induced a significantly higher degree of cartilage formation and cartilage metabolism than the HLC–SPNs hydrogel and control.
3.6. Histological and immunohistochemical studies
To further investigate the efficacy of the HLC(Dex)–SPNs–KGN hydrogel to promote OA repair, we performed H&E and safranin O staining and assessed the following histological characteristics on the basis of the 11 parameters of the O’Driscoll system tissue scale: the formation of new cartilage tissue, degradation of hydrogel samples, and production of subchondral bone. After the first month, little fibrous tissue was found in the control. However, the HLC–SPNs hydrogel group showed a mixture of fibrous and cartilage-like tissue (
Fig. 7(a)). A large amount of cartilage-like extracellular matrix was observed in the HLC(Dex)–SPNs–KGN hydrogel group, and new bone tissue was observed in the lower part of the cartilage. By the third month, the control group had more fibrous tissue than after the first month. The HLC–SPNs hydrogel group and HLC(Dex)–SPNs–KGN hydrogel group displayed a mixture of fibrous and cartilage-like tissue at the defect sites (Fig. S9 in Appendix A). The HLC(Dex)–SPNs–KGN hydrogel group had a higher density of cartilage tissue than HLC–SPNs hydrogel group.
Safranin O was used to differentiate cartilage tissue from bone tissue, assess the integrity of the articular surfaces of the cartilage matrix, and evaluate the aggregation between chondrocytes and glycosaminoglycans. The results showed that there was no visible cartilage formed in the control group, while cartilage tissue was relatively intact in the HLC–SPNs and HLC(Dex)–SPNs–KGN hydrogel groups (
Fig. 7(a)). The HLC(Dex)–SPNs–KGN group had a higher content of red-labeled anionic polysaccharides than the HLC–SPNs group had. These results indicated that HLC(Dex)–SPNs–KGN treatment promoted cartilage repair.
Histological scoring was performed on the H&E and safranin O staining images of the defective joints. The overall filling of the defects indicated the generation of new tissue in the defects (
Fig. 7(b)). The cartilage score (
Fig. 7(c)) assessed the integrity of the articular surface of the cartilage matrix and aggregation between chondrocytes and glycosaminoglycans. The HLC(Dex)–SPNs–KGN hydrogel group showed a higher overall filling and cartilage score than the control and HLC–SPN hydrogel groups. These results suggested that the HLC(Dex)–SPNs–KGN hydrogel supported cartilage formation.
An immunohistochemical analysis of three pro-inflammatory factors (IL-6, TNF-α, and ADAMTS5) was performed at the site of cartilage damage (
Fig. 7(d)). Positive reactions to IL-6, TNF-α, and ADAMTS5 were evident in the control, and the relative areas increased over time (
Fig. 7(d)), which indicated that the inflammatory response continued to increase in the defective joints, suggesting that the inflammation in the cartilage required clinical interventions. Interestingly, we found that the inflammatory response in the HLC–SPNs hydrogel group was low. The anti-inflammatory activity of the HLC–SPNs hydrogel, which did not contain anti-inflammatory drugs, suggested that the hydrogel itself may provide conditions that promoted tissue repair. As shown in
Fig. 7(e), the HLC(Dex)–SPNs–KGN hydrogel group had significantly lower levels of IL-6, TNF-α, and ADAMTS5 than the control and HLC–SPNs hydrogel groups, which indicated that the HLC(Dex)–SPNs–KGN hydrogel effectively controlled the inflammation.
3.7. In vivo effect of HLC(Dex)–SPNs–KGN hydrogel on macrophage polarization
Fig. 8(a) shows the results of the immunofluorescence staining of CD68-labeled macrophages, CD86-labeled M1 macrophages, and CD206-labeled M2 macrophages. After the first month, the number of macrophages (M1 and M2) in the HLC(Dex)–SPNs–KGN hydrogel group and HLC–SPNs hydrogel group were two times higher than the number of macrophages in the control group.
As shown in
Figs. 8(b) and
(c), after the first month, the M1 macrophage population of the HLC(Dex)–SPNs–KGN hydrogel group (6.25%) was lower than that of the control group (31.25%), and the M2 macrophage population of the HLC(Dex)–SPNs–KGN hydrogel group was higher (46%) than that of the control group (6.23%). Moreover, the M1 macrophage population of the HLC(Dex)–SPNs–KGN hydrogel group was 85.6% lower than that of the HLC–SPNs hydrogel group. In contrast, the M2 macrophage population of the HLC(Dex)–SPNs–KGN hydrogel group was 91.6% higher than that of the HLC–SPNs hydrogel group. The M2/M1 ratio of the HLC(Dex)–SPNs–KGN hydrogel group was significantly higher than that of the control group and HLC–SPNs hydrogel group, which was attributed to the release of Dex from the HLC(Dex)–SPNs–KGN hydrogel. Moreover, by the third month, the M2/M1 ratio of the HLC(Dex)–SPNs–KGN hydrogel group remained significantly higher than that of the control group and HLC–SPNs hydrogel group (
Fig. 8(c)), which suggested that M2 macrophages had a significant role in the late stages of cartilage repair.
Figs. 8(d) and
(e) show that the HLC(Dex)–SPNs–KGN hydrogel group had a significantly higher secretion of BMP2 and TGF-β1 than the control group and the HLC–SPNs hydrogel group. These results suggested that the polarization of M1 macrophages into M2 macrophages, which was mediated by HLC(Dex)–SPNs–KGN hydrogel, facilitated chondrogenic differentiation and cartilage repair.
3.8. Effect of HLC(Dex)–SPNs–KGN hydrogel on chondrogenic differentiation
The effect of the HLC(Dex)–SPNs–KGN hydrogel on the differentiation of hMSCs into chondrocytes was investigated using an
in vivo model of articular cartilage injuries. The images of the immunofluorescence staining for chondrocyte-specific proteins (COMP, Col II, aggrecan, and SOX-9), shown in
Figs. 9(a) and
(b), indicated that the HLC(Dex)–SPNs–KGN hydrogel induced chondrogenic differentiation of hMSCs. The levels of COMP, Col II, aggrecan, and SOX-9 in the HLC(Dex)–SPNs–KGN hydrogel group were significantly higher than those in the control group and the HLC–SPNs hydrogel group. Moreover, the expression of RUNX1 was highest in the HLC(Dex)–SPNs–KGN hydrogel group. A summary of the regulatory effects of the HLC(Dex)–SPNs–KGN hydrogel throughout the different stages of cartilage repair is shown in
Fig. 9(c). In the early stages of cartilage repair, a large amount of Dex was released, which downregulated the expression of pro-inflammatory factors and promoted the polarization of M1 macrophages into M2 macrophages. In the middle and late stages of cartilage repair, KGN, which was released from the hydrogel scaffold by a sustained release, induced the differentiation of hMSCs into chondrocytes and maintained chondrocyte stability. Therefore, KGN and Dex acted synergistically to induce cartilage repair.
4. Discussion and conclusions
Cartilage healing occurs primarily in three consecutive and partially overlapping stages, which are the inflammatory stage, cartilage differentiation stage, and endochondral ossification or remodeling stage. To provide targeted treatment for these three phases, a nanocomposite hydrogel was designed to provide controlled release and local delivery of Dex and KGN to induce cartilage regeneration. KGN was covalently grafted onto silk protein (
Fig. 2(d)), and the KGN-loaded silk protein was treated with acetone to form SPNs–KGN (
Fig. 2(f)). HLC was mixed with Dex, and the phosphate groups of Dex cross-linked with the Tyr, Ser, Thr, and Arg residues of HLC to form hydrogen bonds (
Fig. 3(b)), resulting in the formation of an HLC(Dex) compound. The HLC(Dex) compound was mixed with SPNs–KGN, and the mixture was enzymatically cross-linked using TGase to form an engineered nanocomposite hydrogel. TGase catalyzes peptide bonds between glutamine and basic amino acids, such as lysine, which results in the formation of a hydrogel network [
16,
27]. In the present system, TGase not only catalyzed the cross-linking between HLC molecules, but also covalently cross-linked SPNs–KGN with the HLC network (
Fig. 3(d)). The SPNs provided a spatial barrier to control the release of KGN (
Figs. 3(e) and
(f)) and transport hydrophobic KGN molecules to a hydrophilic environment. In addition, it should be noted that the silk proteins in this study were not subjected to an efficient purification process, and further decontamination and purification steps are required for clinical applications. The porous structure of the HLC hydrogel promoted cell proliferation and differentiation (
Figs. 4(a)–(c)), increasing the oxygen tension and nutrient supply and creating a conducive environment for the inward growth of cartilage tissue [
28].
For decades, Dex has been incorporated into intra-articular glucocorticoid injections to treat inflammation and pain; however, repeated injections at low doses can increase the pain, while repeated injections at high doses can adversely affect the growth of bone cells and carries a risk of systemic resorption. At a concentration of 1 nmol∙L
−1, Dex protects cartilage from the damage induced by TNF-α [
29], and at a concentration between 100 nmol∙L
−1 and 10 μmol∙L
−1, Dex inhibits the loss of glycosaminoglycans from cartilage. This concentration range of Dex (100 nmol∙L
−1–10 μmol∙L
−1) is significantly lower than that of the injectable Dex used in clinical practice (7.7 mmol∙L
−1) [
30]. The amount of Dex that was added to the HLC(Dex)–SPNs–KGN hydrogel was carefully designed on the basis of the release rate and effective concentration of Dex. The concentration of Dex in the HLC(Dex)–SPNs–KGN hydrogel was approximately 1.9 mmol∙L
−1, and the average daily release of Dex from the hydrogel in a PBS solution to achieve satisfactory anti-inflammatory effects on cartilage and joints and to promote cartilage repair was approximately 38 nmol∙L
−1 (
Figs. 3(e) and
(f)). The average daily release of Dex was calculated over a 40-day period.
To alleviate inflammation during the early stages of cartilage repair, the degraded hydrogel released Dex at a controlled dose and rate. Inflammation is a complex process that involves a series of immune cells migrating to the site of inflammation. Dex regulates the polarization of M1 macrophages into M2 macrophages in response to the transition from an inflammatory phase to a regenerative phase [
9]. Dex was effectively released and, as an anti-inflammatory drug, produced anti-inflammatory effects (
Figs. 4(f) and
(g)), which was expected to reduce the excessive inflammation that characterizes the early stage of cartilage repair. Even in the absence of KGN and Dex, TNF-α was significantly decreased in the HLC–SPNs hydrogel (
Fig. 4(g)) because of the anti-inflammatory effects of HLC [
31,
32].
The median effective concentration of KGN is reported to be 100 nmol∙L
−1 [
14], and the calculated average daily release of KGN from the HLC(Dex)–SPNs–KGN hydrogel on the basis of a 40% release rate over a 40-day period was 630 nmol∙L
−1 in a PBS solution (
Figs. 3(e) and
(f)). The calculated average daily release of KGN was greater than the median effective concentration of KGN that effectively promoted cartilage differentiation. KGN was released from the HLC(Dex)–SPNs–KGN hydrogel as the hydrogel and the SPNs–KGN degraded (
Fig. 3(g)). These results indicated that the HLC(Dex)–SPNs–KGN hydrogel delayed the release of Dex and KGN, which then underwent controlled release from the hydrogel through two distinct release mechanisms.
KGN, a hydrophobic molecule, is poorly dispersed in culture medium, which reduces its cellular uptake. In contrast, the KGN molecules in the SPNs–KGN were carried by the hydrophilic SPNs, which improved the dispersion, solubility, and cellular uptake rate of the KGN molecules (
Figs. 4(d) and
(e)). The release of KGN was controlled by both the HLC network and SPNs. KGN and SPNs were cross-linked by amide bonds.
In vivo, amidases such as glutaminase and urease are versatile tools for amide bond cleavage. Proteases such as trypsin, chymotrypsin, and papain also hydrolyze the amide bonds of amino acids. SPNs–KGN is usually degraded by these enzymes to release KGN. The progressive degradation of the HLC network released SPNs–KGN, which transported KGN to the inner cell membrane for degradation to release KGN molecules. Thus, KGN molecules were released progressively. KGN regulated chondrogenic differentiation of hMSCs mainly during the mid-to-late stage of cartilage repair. The function of HLC(Dex)–SPNs–KGN hydrogel in cartilage differentiation assays was confirmed by using cartilage marker proteins to analyze the degree of cell differentiation (
Fig. 5). COMP is a cartilage-specific protein that is expressed in the matrix surrounding chondrocytes, while aggrecan and Col II are the main functional proteins of chondrocytes. SOX-9, which is crucial for chondrogenic differentiation of hMSCs, regulates multiple stages of cartilage differentiation [
33]. KGN-loaded hydrogels (HLC(Dex)–SPNs–KGN and HLC–SPNs–KGN) significantly enhanced COMP, Col II, aggrecan, and SOX-9 expression (
Figs. 5(c)–(g)). Moreover, expression of chondrocyte-associated proteins in the hMSCs in the HLC(Dex)–SPNs hydrogel was promoted. Dex has been shown to effectively induce chondrogenic differentiation of hMSCs [
34]. The expression of chondrocyte-related proteins in the hMSCs of the HLC(Dex)–SPNs–KGN hydrogel was significantly higher than that in the hMSCs of the HLC(Dex)–SPNs hydrogel and the HLC–SPNs–KGN hydrogel, which indicated the synergistic effect by which Dex and KGN induced chondrogenic differentiation of hMSCs (
Fig. 5(g)). Thus, the hMSCs of the HLC(Dex)–SPNs–KGN hydrogel showed a significantly higher degree of chondrogenic differentiation than the hMSCs of the other groups.
In an animal model of cartilage injury, the HLC(Dex)–SPNs–KGN hydrogel showed the greatest repair effect (
Fig. 6) through anti-inflammatory (
Figs. 7(d) and
(e)) and immunomodulatory mechanisms, and by promoting pro-chondrocyte differentiation and maintenance of chondrocyte homeostasis.
Cartilage repair mediated by biomaterials requires host immune cells, host osteoblasts, and appropriate biomaterials [
28]. Implanted biomaterials are recognized by the immune system of the host as foreign bodies that trigger an immune response-mediated biological behavior of mesenchymal stem cells, cartilage, and osteoblasts [
35]. Among the immune cells, macrophages have a central role in the inflammatory response and host defenses. The two phenotypes of macrophages, M1 and M2, and the interconversion of the two macrophage phenotypes have different effects on cartilage repair [
36,
37]. The HLC(Dex)–SPNs–KGN and HLC–SPNs hydrogels recruited macrophages by promoting the adsorption of proteins (fibronectin, fibrinogen, and complement components) that are recognized by immune cells. The HLC(Dex)–SPNs–KGN hydrogel was synthesized from HLC, which has a variety of amino acids that bind to proteins in tissue. The HLC-bound proteins were recognized by macrophages and promoted their recruitment.
Dex is not only an effective anti-inflammatory agent, but also induces the polarization of macrophages in cartilage tissue, increasing the M2/M1 ratio, and thereby attenuating the inflammatory response and facilitating the transition from a pro-inflammatory to a pro-regenerative state (
Figs. 8(a)–(c)). In conjunction with the immunohistochemical analysis of pro-inflammatory factors, the effect of macrophage polarization on cartilage repair mediated by the HLC(Dex)–SPNs–KGN hydrogel was also analyzed. In the early stages of cartilage repair, M1 macrophages amplify the inflammatory response by promoting tissue fibrosis [
28]. Macrophages may then exacerbate the inflammatory response and destroy native cartilage tissue (which is partly responsible for OA); or, macrophages may suppress the inflammatory response and initiate tissue repair. Therefore, modulating the early immune response and polarization of M1 macrophages into M2 macrophages promotes cartilage regeneration. The HLC(Dex)–SPNs–KGN hydrogel modulated the polarization of M1 macrophages into M2 macrophages, thus reducing the persistence of M1 macrophages that may otherwise lead to the formation of fibrous capsules (Fig. S10 in Appendix A). The polarization of M1 macrophages into M2 macrophages decreases the levels of pro-inflammatory factors in tissue. Here, M2 macrophages expressed high levels of BMP2 and TGF-β1 (
Figs. 9(d) and
(e)), which promotes the proliferation and differentiation of periosteal mesenchymal cells, thus influencing chemotaxis in cartilage remodeling [
37].
The regeneration of cartilage is difficult because cartilage lacks nerves and vascular tissue. We labeled the stem cells in the damaged area by immunofluorescence staining (CD90 and CD73), which showed the presence of a large number of stem cells (Fig. S11 in Appendix A), indicating that the HLC–SPNs and HLC(Dex)–SPNs–KGN hydrogels served as scaffolds that facilitated the proliferation and adhesion of cells, such as BMSC from bone marrow synovial mesenchymal stem cells from the synovium, and chondrocytes from surrounding healthy cartilage. In addition to the contribution of KGN to the regulation of chondrogenic differentiation and cartilage homeostasis at the mid-to-late stage of cartilage repair, the transcription factors secreted by M2 macrophages and the scaffolding effect of the HLC network synergistically induced cartilage regeneration (
Figs. 9(a) and
(b)).
Previous studies have demonstrated that RUNX1 promotes the proliferation of chondrocytes and the production of cartilage [
38], and stabilizes the degree of cartilage differentiation; inhibiting RUNX1 promotes osteogenesis and terminal chondrocyte differentiation (chondrocyte hypertrophy, chondrocalcification), and it has been suggested that RUNX1 is a downstream effector of KGN [
39]. The expression of RUNX1 was highest in the HLC(Dex)–SPNs–KGN hydrogel group (
Figs. 9(a) and
(b)), indicating that the HLC(Dex)–SPNs–KGN hydrogel not only regulated chondrocyte differentiation but also stabilized chondrocytes. The mechanical properties of the HLC(Dex)–SPNs–KGN hydrogel group were closer to those of the normal group than to those of the control and the HLC–SPNs hydrogel group (Fig. S12 in Appendix A), indicating that the effect of the HLC(Dex)–SPNs–KGN hydrogel on damage repair was better than that of the other treatments.
In this study, we synthesized a protein-based nanocomposite hydrogel with dual drug delivery mechanisms that delivered Dex and KGN
in situ to reduce inflammation and enhance cartilage regeneration. The hydrogel provided controlled release of Dex and KGN using two release mechanisms, with corresponding modulating effects through the different stages of cartilage repair. First, as a scaffold, the hydrogel provided an appropriate microenvironment to promote the adhesion and proliferation of macrophages and hMSCs. Second, the hydrogel controlled the release of Dex in the early stages of treatment, during which Dex reduced inflammation by downregulating the expression of the pro-inflammatory factors TNF-α, IL-6, and ADAMTS5, and by inducing the polarization of M1 macrophages into M2 macrophages, which promoted chondrogenic differentiation through immunomodulation (upregulation of BMP2 and TGF-β1 in M2 macrophages). Third, the release of KGN regulated chondrogenic differentiation mainly during the mid-to-late stage of cartilage repair by upregulating the expression of Col II, SOX-9, COMP, and aggrecan, and maintained chondrocyte homeostasis and avoided hypertrophy by upregulating the expression of RUNX1 (
Fig. 9(c)). In conclusion, this dual drug-loaded nanocomposite hydrogel is a promising therapeutic option for promoting cartilage formation and osteoarthritic defect repair.
CRediT authorship contribution statement
Huan Lei: Writing – original draft, Validation, Supervision, Software, Methodology, Formal analysis, Data curation. Daidi Fan: Writing – review & editing, Investigation, Funding acquisition, Conceptualization.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was supported by the National Key Research and Development Program of China (2019YFA0905200).
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.eng.2025.05.010.
PDF
(10516KB)
