Silk fibroin-based biomaterials for cartilage/osteochondral repair

1Osteoarthritis (OA) is a degenerative disease, which involves the whole joint (including articular cartilage, subchondral bone, ligaments, joint capsule, synovium, and muscles around the joint). The pathogenesis of OA is complex and its prevalence is correlated with age, gender and obesity status,. Patients suffer from severe pain and even amputation caused by OA, resulting in a sharp decline in quality of life. At present, about 250 million people around the world are suffered from OA. With the aging of the global population, increasing obesity, and joint injuries, the number will increase by about 50% in the next decade. Statistically, OA costs about $303 billion a year in medical expenses and lost income, which is a massive burden on suffered individuals, health systems, and the wider socio-economy. Typically, OA will occur when the dynamic balance between cartilage destruction and repair is broken, due to excessive mechanical load and pathological factors -. Furthermore, untimely repair of cartilage defects can lead to further deterioration of OA,. Conversely, under osteoarthritic conditions, the rising activity of multiple proteases gives rise to augmented cartilage destruction. In brief, the development of OA and cartilage destruction is a vicious cycle. Therefore, cartilage repair is the essential element in the treatment of OA,.

2Currently, the treatment strategies of cartilage repair include microfracture (bone marrow stimulation), autologous chondrocyte transplantation, and allogeneic/autologous cartilage transplantation,. Although these treatment strategies are widely used in the clinic, they still have several shortcomings. For example, microfractures often cause the formation of fibrocartilage. In addition, the application of allogeneic/autologous cartilage transplantation is limited by donor shortage, poor integration, and surgical infection,. Recently, tissue engineering has become a promising strategy for repairing damaged tissue, such as skin, heart, and bone -. In general, cartilage defects are caused by trauma, long-term wear, disease, and aging. The self-repairing ability of cartilage is limited due to the absence of blood vessels and the poor chondrocytes proliferation,. Therefore, introducing nutrients or stem cells from adjacent tissues is an effective strategy to accelerate cartilage repair. Subchondral bone is adjacent to and underneath the cartilage and is rich in vascular tissue and a variety of stem cells -. Therefore, constructing osteochondral defects (so that nutrients and a variety of stem cells in the subchondral bone can arrive at the cartilage region) to achieve integrated osteochondral repair is expected to accelerate cartilage repair,.

3Silk fibroin (SF) is a kind of natural protein material that has been used in the clinic for decades. Compared with synthetic and natural materials, SF exhibits superior toughness, biocompatibility, biodegradability, and thermal stability. Additionally, SF can also significantly promote bone and cartilage regeneration,. Based on these, SF-based biomaterials have been widely studied and used in the biomedical field, especially in tissue engineering. Moreover, SF-based biomaterials have more incomparable advantages than other materials in the cartilage/osteochondral repair, including: (1) Excellent cytocompatibility (suitable growth environment could accelerate cell proliferation and differentiation); (2) unique mechanical property (cell adhesion and growth require adequate mechanical support); (3) non-toxic degradation products and controllable biodegradability (the space left by the degradation of biomaterials is essential for the growth of regenerative tissues); (4) desirable processability (SF-based biomaterials are satisfactory for a variety of fabricating methods and modifications),. Therefore, SF-based biomaterials have great research value and application prospects in the field of cartilage/osteochondral repair (Figure 1).

4Herein, we review the research of SF-based biomaterials in cartilage/osteochondral repair in the past decade. We hope this review can provide a theoretical basis for the design and preparation of SF-based biomaterials for cartilage/osteochondral repair.

5Generally, silk is secreted by various animals, such as silkworms, spiders, scorpions, mites, and bees. Compared with other silk, silkworm silk is the only silk that can be produced on a large scale and has been applied in the clinic. Hence, this review focuses on the description and discussion of silkworm silk. As shown in Figure 2A, a silkworm silk consists of two SF fibers coated with sericin. To obtain pure SF, silkworm silk needs to be degummed, rinsed, and dried.

6SF has a complex molecular structure, which is composed of disulfide bond-linked light chain (molecular weight 26KDa) and heavy chain (molecular weight 390KDa), as shown in Figure 2B. The light chain, which has excellent elasticity, is formed of disordered amino acid sequences and does not participate in the formation of crystal structure. The heavy chain consists of several repetitive sequences, including Gly-Ala-Gly-Ala-Gly-Ser sequences, Gly-Ala-Gly-Ala-Gly-Tyr sequences, and Gly-Ala-Gly-Ala-Gly-Ser-Gly-Ala-Ala-Ser sequences. The heavy chain has remarkable tensile strength due to these sequences. Therefore, attributing to the heavy and light chains, SF has better mechanical properties than other natural and synthetic materials (maximum fracture elongation, ultimate Young's modulus, and toughness are 4%, 26%, 300-700MPa, and 70-78MJ/m3, respectively). In addition, the mechanical properties of SF can be adjusted by changing the size, quantity, orientation, and arrangement of crystalline (silk Ⅱ) and amorphous structures.

7The chemical composition and structure of SF endow it with excellent biocompatibility and biodegradability. The degradation products of SF (mainly amino acids and peptides) not only have no cytotoxicity, but also can provide nutrition for tissue regeneration. This makes SF have a good prospect of clinical application. Actually, surgical sutures and wound dressings prepared by SF have been widely used in biomedical fields. Currently, SF is attracting more and more attention in the field of cartilage/osteochondral repair.

8The articular cartilage is a multilayer tissue, including the hyaline cartilage layer and calcified cartilage layer. In addition, the subchondral bone is located directly below the articular cartilage (Figure 2). Hyaline cartilage, calcified cartilage, and subchondral bone form a massive complex with different components, structures, and properties.

9Hyaline cartilage is a non-blood supply tissue composed mainly of water, type II collagen and aggregates. The core protein of aggrecan has covalent binding and powerfully negatively charged glycosaminoglycan side chain. It is non-covalently linked to hyaluronic acid through connexin to form a proteoglycan polymer. Highly cross-linked type II collagen fibers form an organized network that captures negatively charged proteoglycan aggregates and interacts with other collagen, small proteoglycan, and other cartilage-specific/non-specific matrix proteins. Specially, chondrocytes are the only cell type in cartilage, accounting for only 1-2% of the total volume of cartilage. However, chondrocytes maintain basic metabolism and proliferate slowly due to the insufficient mitotic activity of mature articular chondrocytes.

10Calcified cartilage is a thin layer of tissue between hyaline cartilage and subchondral bone. It is produced by endochondral ossification and still exists after the growth plate is closed. Calcified cartilage is denser and more mineralized than the adjacent cartilage. Besides, there is a histologically defined tidal mark (tidemark) between calcified cartilage and articular cartilage. In addition, the biological composite formed by calcified cartilage and subchondral bone is especially suitable for transmitting and distributing mechanical force under physiological load. Compared with cartilage, most of the mechanical load is borne by the composite.

11Subchondral bone is located directly below the calcified cartilage with unique structural, biological, and mechanical properties. It forms a plate-like structure similar to cortical bone in other skeletal parts. In comparison to cortical bone, subchondral bone has more porous and metabolic activity. In addition, the bones in the subchondral bone layer combine to form a trabecular network (also known as cancellous bone). The trabeculae in cancellous bone are oriented in different directions according to their location, and they provide a unique structural network to adapt to local mechanical effects. Moreover, the periosteum forming the edge of the joint is in direct contact with the joint capsule and synovium. Under the joint capsule and synovium, ligaments and tendons are inserted into the bone to form a unique structure at the end.

12In brief, hyaline cartilage, calcified cartilage and subchondral bone have different characteristics. It is worth noting that the species of silkworm affects the structure of SF molecule, and then affects the biological properties of SF-based biomaterials. For example, Saha et al. compared the performance of mulberry SF and non-mulberry SF in the field of bone and cartilage repair. Under similar conditions, they found that mulberry SF was more osteoinductive, while non-mulberry SF was more chondroinductive. Meanwhile, Singh et al. compared cartilage repair properties of mulberry SF hydrogels and non-mulberry SF hydrogels. Compared with mulberry SF hydrogels, non-mulberry SF hydrogels showed better-promoting effects on sulfated glycosaminoglycans and collagen deposition. This proved that non-mulberry SF is more suitable for cartilage repair. Additionally, the characteristics of hyaline cartilage, calcified cartilage and subchondral bone should be taken into account in the design of SF-based biomaterials for cartilage/osteochondral repair. For the cartilage layer, there is only one kind of chondrocyte with a gradient distribution. On the one hand, the SF-based biomaterial should be modified to promote the proliferation and differentiation of chondrocytes. On the other hand, the SF-based biomaterial should be endowed with a gradient structure to ensure the gradient distribution of chondrocytes (similar to primary cartilage). For the calcified cartilage layer, SF-based biomaterials should be designed as dense structure. The dense structure could prevent the migration of angiogenic cells from the subchondral bone layer to the cartilage layer and avoid cartilage layer fibrosis. For the subchondral bone layer, the biomineralization and antibacterial ability of SF-based biomaterials could be enhanced by adding metal ions -. It is worth mentioning the synergistic role of osteogenesis and angiogenesis in promoting subchondral bone repair. Therefore, SF-based materials could add angiogenic factors to accelerate the repair of subchondral bone.

133D bioprinting is a controllable rapid prototyping and additive manufacturing technology,. It allows personalized adjustment of size parameters and layer-by-layer printing of biomaterials with complex structures via computer assistance. In addition, 3D printed biomaterials show great potential in meeting the mechanical, structural and biological requirements for cartilage/osteochondral repair -. Generally, printing equipment (biological printer) and ink (biological ink) are the two main components of 3D bioprinting technology.

14Currently, the main types of 3D bioprinting equipment include inkjet printing, extrusion printing, and laser-assisted printing. Among them, the extrusion strategy is suitable for a wide range of materials and curing methods. Furthermore, compared with the inkjet printing method and laser-assisted method, the extrusion printing technology is able to process higher-resolution patterns and is more cost-effective. Additionally, extrusion 3D printing can effectively overcome the problem that SF-based biomaterials fail to be commercialized due to the insufficiency of standardization and industrialization in the preparation process. For instance, Trucco et al. constructed and verified the analytical model of a 3D extrusion bioprinting system based on extrusion to predict the width of deposited SF filaments. The model took into account the key printing process parameters (pressure and speed of biological ink) to predict the SF filament width from the various rheological properties of different hydrogels. This model helped researchers to define the most appropriate SF printing parameters to maximize the fidelity of the manufactured structure to the design parameters. In addition, in order to improve the affinity of SF-based biomaterials for chondrogenic cells, other natural macromolecular materials and stem cell-specific affinity peptides can be added when printing SF-based biomaterials. For example, Thunsiri et al. prepared bilayer bioactive biomaterial scaffolds with polylactic acid (PLA), polycaprolactone (PCL), hydroxyapatite (HA), chitosan (CS), and SF by 3D extrusion bioprinting. The experimental results showed that the scaffolds had the potential to treat full-thickness articular cartilage defects. Meanwhile, Shi et al. constructed SF-gelatin-E7 (SFG-E7) scaffolds by integrating SF with gelatin in combination with bone mesenchymal stem cell (BMSC)-specific-affinity peptide using 3D bioprinting technology (Figure 4A). They found that E7 BMSCs specific affinity peptides significantly promoted the proliferation and differentiation of BMSC and the production of extracellular matrix (ECM). These works proved that extrusion 3D bioprinting had strong controllability and laid a foundation for industrialized and standardized production of SF-based biomaterials.

15According to the reports, choosing the appropriate biological ink is the key to successful biological printing. Biological inks need to meet several basic standards, such as excellent bioactive, appropriate viscosity, suitable flexibility and great stability,. Generally, modification of SF biological ink is required to strengthen the bioactivity of SF when preparing SF-based biomaterials using 3D bioprinting. For example, Li et al. fabricated SF-based biological ink with platelet-rich plasma (PRP), and used 3D bioprinting technology to prepare SF-PRP scaffolds. The addition of PRP effectively enhanced the bioactivity of SF-based biological ink and accelerated cartilage repair. However, SF-based biomaterials still face the problem that the mechanical strength is inadequate to meet the needs of cartilage/osteochondral repair. In order to overcome this problem, SF biological ink can be modified with dual functions. For instance, Deng et al. designed SF-parathyroid hormone (SF-PTH) biological ink and mechanically modified SF-methacrylic anhydride (SF-MA) biological ink. Then, successfully constructed the integrated gelatin methacryloyl (GM) + SF-PTH/GM + SF-MA biphasic scaffold with excellent biological activity and a mechanical gradient via 3D bioprinting technology. In addition, in vivo experiments showed that the scaffold had desirable osteochondral repair ability.

16Electrospinning is a highly controllable technology that allows fine-tuning of multiple parameters. Compared with other fiber spinning processes, electrospinning is able to produce long fibers with a smaller diameter and a higher ratio of surface area to volume. In addition, electrospun scaffolds prepared by assembling electrospun fibers have the advantages of low cost, simple process, and similar nanofibrous structure to ECM with interconnected pores compared with other tissue scaffolds. Moreover, due to the highly controllable of electrostatic spinning technology, it is available to prepare SF-based scaffolds with a wide range of pore sizes. For example, Huang et al. prepared SF scaffolds with gradient pore size under appropriate relative humidity by adjusting fiber diameter (concentration of electrospinning solution) and collector temperature. The experimental data showed that the large pore layer scaffolds with 37.2±12.9 μm pore size significantly promoted the migration of cells into the scaffold. The medium pore layer scaffold with 11.6±1.4 μm pore size was more beneficial to cell proliferation. Additionally, the small pore layer scaffold with a micropore size of 5.9±1.4 μm had higher mechanical properties.

17In recent years, electrospinning has been widely used in the field of cartilage tissue engineering,. For instance, Li et al. constructed acellular three-dimensional meniscus scaffolds composed of PCL, SF, and strontium by wet electrospinning (Figure 4B). The functional scaffolds had been successfully used for meniscus regeneration in the rabbit meniscectomy model. This study suggested that electrospinning could mix a variety of substrates and combine the characteristics of substrates to fit the needs of cartilage tissue engineering. Furthermore, SF-based scaffolds for full-thickness repair of hyaline cartilage, calcified cartilage and subchondral bone could also be prepared by electrospinning technique. For example, Liu et al. designed and synthesized CS/SF/hydroxyapatite three-layer scaffolds via electrospinning. Results confirmed that this CS/SF/hydroxyapatite three-layer scaffolds could mimic the chondral layer, calcified layer, and subchondral bone layer respectively to promote osteochondral repair. Generally, ordinary electrospinning usually produces two-dimensional films with tightly packed layers which hinder osteoid cells migration. To break through this limitation, electrospinning could be combined with other preparation methods to improve the cartilage/osteochondral repair effect of SF-based biomaterials. For instance, Chen et al. prepared three-dimensional/hyaluronic acid/SF scaffold (3DHAS) in a stepwise design by combining electrospinning, gas foaming and freeze-drying techniques. They found that 3DHAS had the characteristics of low density, large pore size, high porosity, strong water absorption capacity, and good mechanical stability. Importantly, there was excellent cartilage regeneration after subcutaneous transplantation of 3DHAS. Moreover, electrospinning could maintain the elasticity of SF (which is essential for cartilage/osteochondral tissue engineering). For instance, Christakiran et al. reported a simple, scalable, and repeatable strategy to develop electrospun double-layer bioglass/SF composites for the repair of osteochondral defects. The biphasic composites exhibited both an elastic region pertinent for cartilage tissue and a stiff compression resistant region simulating the bone phase.

18Freeze-drying is a flexible and controllable technology for the preparation of biomaterials. Freeze-drying has attracted wide attention because it could control the growth of ice nuclei to adjust the pore size of biomaterials,. This method has shown certain advantages, such as process adaptability, excellent pore-making capacity, and environmental friendliness.

19It is reported that the pore shape and orientation of biomaterials can be controlled by adjusting the freezing direction,. For example, Fan et al. prepared biomimetic anisotropic 3D SF scaffolds with the radial channel via directional freeze-drying technology (Figure 4C). During the preparation process, heat is transmitted from the center to all around. Therefore, the growth direction of ice crystals is from the periphery to the center. Compared with other SF scaffolds, the freeze-drying scaffolds had a higher cell capture and growth promotion ability. More importantly, the scaffolds also regulated cell migration, arrangement, elongation, and interaction. In addition, the unidirectional pore could be prepared by changing the heat transfer direction from radial direction to the single direction via freeze-drying technology. For instance, Zhang et al. fabricated biomimetic SF-based scaffolds with cartilage ECM-like configurations (horizontally aligned structure of the surface layer, vertically aligned structure of the bottom layer and random pore structure of the middle layer) by directional freezing. The above two examples showed that freeze-drying technology can control the morphology of SF-based biomaterials. The radial pore structure and directional pore structure will be discussed in detail in 4.3.1. Directional structure.

20Notably, the process of preparing biomaterials by freeze-drying technique causes the solvent to freeze and then sublimate, while having little effect on the solute. Therefore, freeze-drying is also convenient for SF-based biomaterials to carry drugs and growth factors. For instance, Sun et al. successfully prepared oriented microtubule collagen/CS/SF-transforming growth factor-β1 (TGF-β1) scaffolds with good hydrophilicity by freeze-drying technique. The scaffolds had good physical and chemical properties, mechanical properties, and biocompatibility. Furthermore, the scaffolds induced cartilage and subchondral bone regeneration by releasing TGF-β1.

21Salt leaching is widely used to prepare SF-based biomaterials due to its high effectiveness and low cost. This method induces SF aggregation by adding salt particles to the SF solution. Porous SF biomaterials can be obtained by removing the undissolved salt particles (Figure 4D). The properties of SF-based materials fabricated by salt-leaching are influenced by the size of salt particles and the concentration of SF solution. In general, salt-leaching is combined with other methods to fabricate SF-based biomaterials. For example, Ribeiro et al. prepared a novel SF-based material by combining salt-leaching and freeze-drying methods. This material had suitable porosity and mechanical properties for cartilage repair. In addition, Zhou et al. prepared a SF-CS material by combining salt-leaching and cross-linking methods. This material could maintain the chondrocyte phenotype, attenuate the interleukin (IL)-1β-induced inflammatory response of chondrocytes in vitro, and accelerate osteochondral repair in vivo.

22As mentioned in section 2, SF contains hydrophobic residue repetitive sequences and a large number of polar groups. Therefore, SF has excellent cross-linking properties. The commonly used cross-linking methods for preparing SF-based hydrogels include enzymatic cross-linking, addition of cross-linker, photo cross-linking and ultrasonication (Figure 4E). Enzymatic cross-linking is a chemical cross-linking method which induce cross-linking reaction through activating certain groups by utilizing the catalysis of enzymes. This method can maintain the mechanical properties and biocompatibility of SF due to its mild reaction condition. For example, Zhang et al. constructed a novel enzymatically cross-linked SF-laponite (LAP) nanocomposite hydrogel. This hydrogel had good biocompatibility and excellent mechanical properties. Both the in vitro and in vivo experiments showed that this hydrogel could stimulate osteogenesis and chondrogenic differentiation of BMSCs, and accelerate the regeneration of osteochondral defects. Meanwhile, Li et al. fabricated a SF-tyramine-substituted gelatin (GT) hydrogel by combining enzymatic cross-linking method and 3D bioprinting. This hydrogel had good biocompatibility and could be used as a cell delivery vehicle to accelerate cartilage repair. However, the high price of enzyme limits the development of enzymatic cross-linking. Cross-linkers are a series of substances that can bond multiple linear molecules into a network structure. In addition, cross-linkers are cheaper than enzymes and can enhance the mechanical properties of SF-based hydrogels by forming stable covalent bonds. For example, Wang et al. constructed a SF/diglycidyl ether (BDDE) hydrogel with high elasticity and mechanical stability by using BDDE as cross-linker. The hydrogel had remarkable biocompatibility and the ability of long-term retention in the articular cavity. Photo cross-linking refers to control the cross-linking process through using ultraviolet, visible light or gamma rays by adding photo-initiators. Compared with other methods, the photo cross-linking has the fastest cross-linking rate. For instance, Piluso et al. prepared a cell-laden SF hydrogel in a few minutes by using riboflavin as photo-initiators. This hydrogel showed excellent cytocompatibility with cartilage derived progenitor cells, mesenchymal stem cells, and dental pulp derived stem cells and was expected to be a promising tissue engineering scaffold material for cartilage repair. However, the residues of photo-initiators and cross-linkers may lead to the cytotoxicity of SF-based biomaterials. Ultrasonication is a physical cross-linking method. This method accelerates cross-linking by increasing local temperature and shear-force, which can avoid the cytotoxicity problems caused by photo-initiators and cross-linkers. For instance, Yuan et al. prepared an ultrasonication-induced SF hydrogel. This hydrogel showed satisfactory physicochemical and biomechanical properties and could promote cartilage regeneration.

23Hydrogel is a hydrophilic three-dimensional network polymer material formed by physical or chemical cross-linking -. The physical properties of the hydrogels are similar to biological tissues, such as high water content, rubber texture, and low interfacial tension. Therefore, hydrogels have excellent biocompatibility with surrounding tissues due to their physical properties are similar to natural tissues. Additionally, hydrogels are ideal for imitating the complex and organized network of natural ECM,. Besides, the natural EMC-like microenvironments of hydrogels are suitable for loading cells to promote cartilage/osteochondral repair. For example, Cui et al. presented a rapid and cyto-compatible photo cross-linking process to produce cell-encapsulated SF hydrogels cross-linked Through the formation of di-tyrosine bonds. The high density of chondrocytes in this hydrogel effectively accelerated cartilage repair.

24It was reported that the mechanical properties, shape, and swelling properties of hydrogels could change with the alternation of external conditions, such as temperature, pH value, and ion concentration. The environment-responsive hydrogel designed based on this characteristic can realize the intelligent release of chondrogenic and osteochondrogenic drugs,. For example, Xu et al. prepared a series of stimulus-responsive composite hydrogels using SF and CS as raw materials and 1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide/N-Hydroxysuccinimide as cross-linking agent. The swelling properties of hydrogels showed stimulus-response under different pH and ion concentrations. This work indicated that hydrogels have strong plasticity.

25Importantly, the injectability exhibited by hydrogels is very practical in tissue engineering. This feature allows the hydrogel to fit tightly at the interface of cartilage defects to elevate the integration effect. Besides, injectable hydrogels as drug carriers adapt easily to defects of any size or shape and allow drugs to be evenly distributed,. For example, Dong et al. designed an injectable p-hydroxybenzene propanoic acid-modified chitosan/SF hydrogel with loading kartogenin (KGN)/poly(lactic-co-glycolic acid) microsphere (Figure 5A). The hydrogel was injected in situ to recruit endogenous stem cells and induce them to differentiate into chondrocytes. In addition, after injection of the hydrogel into the cartilage defect site, there was a transition from solution to gel state. Unfortunately, the complicated procedures and toxicity of physicochemical transition methods still hinder further applications of injectable hydrogels. To overcome this problem, Yuan et al. developed an injectable SF hydrogel in a novel one-step ultrasonication cross-linking method. In this way, the toxic side effects caused by the gel transformation of injectable hydrogels in cartilage/osteochondral defects could be avoided.

26Scaffolds are 3D porous matrices that promote cartilage/osteochondral repair and regeneration by providing a specific microenvironment. Ideally, scaffolds should be able to: (1) Promote transport of nutrients, and cell survival, proliferation, and differentiation; (2) provide suitable structural mechanical support; (3) degrade at a controllable rate; (4) present minimal inflammation and toxicity in vivo -.

27Nowadays, SF scaffolds have been widely used in cartilage/osteochondral engineering,. Compared with hydrogels, the scaffolds have fixed shapes and higher mechanical strength. In addition, the stable surface of the scaffolds could be modified by functional coating to improve cartilage/osteochondral repair performance,. For example, Li et al. prepared SF/graphene oxide (GO) meniscus scaffolds with tannic acid/Sr2+ functional coating (Figure 5B). According to the results of tissue staining and the Osteoarthritis Research Society International scoring system, the expression of inflammatory cytokines (such as IL-6, IL-8, and MMPs) in the rat knee joint tissue was significantly down-regulated. Meanwhile, cartilage degeneration and OA injury were also inhibited. In addition, scaffolds could be designed into different 3D shapes to fit the notch of cartilage/osteochondral. For instance, Gao et al. have successfully fabricated porous decellularized cartilaginous matrix/silk fibroin (DCM/SF) scaffolds (Figure 5C). It had been confirmed that DCM/SF scaffolds supported cartilage ring regeneration of BMSCs both in vivo and in vitro. Compared with DCM scaffolds and SF scaffolds, DCM/SF scaffolds significantly promoted the chondrogenesis of implanted BMSCs. In a rabbit model, the scaffold accomplished the restoration of the cricoid tracheal cartilage.

28Scaffolds could serve as an adjunct to cell therapy to promote cell attachment, growth and differentiation. For example, Yang et al. synthesized CS/SF porous scaffolds loaded with C-type natriuretic peptide gene-modified BMSCs to accelerate the repair of articular cartilage defects. Generally, it is difficult to preserve scaffold-cell composites for a long time, which limits the applications of SF-based scaffolds. To address this difficulty, ran et al. developed a SF scaffold loaded with cartilage stem/progenitor cells. The SF scaffold-cell composites could be preserved in liquid nitrogen, and their activity, migration, and chondrogenic capacity were unaffected for up to 3 months.

29Microcarriers usually refer to small spherical scaffolds suitable for cell culture, growth, and transport,. High specific surface area (promote cell growth and maintain cell differentiation phenotype) and injectability (realize tissue regeneration by directly injecting into the target site) enable microcarriers to accelerate cartilage/osteochondral repair,. In particular, porous microcarriers could provide a protected environment for seed cells attachment, proliferation, migration, nutrient exchange, and metabolic waste excretion. For example, Galuzzi et al. compared the activity of chondrocytes under the three models (pellet, alginate beads, and SF/alginate microcarriers). The experiments proved that chondrocytes in the SF/alginate microcarriers were the most active and secreted the most ECM. Meanwhile, Fang et al. prepared strontium-containing SF porous microcarriers with the dual functions of bioactive factor and cell delivery (Figure 5D). Cytobiological analysis showed that the porous microcarriers allowed attachment and proliferation of seeded cells. Besides, the released Sr2+ could stimulate the osteogenic differentiation of BMSCs.

30To circumvent the problem of scarce cell sources, microcarriers could be loaded with growth factors to promote homing, adhesion, and growth of chondrogenic/osteogenesis cells. For instance, He et al. synthesized the SF/CS microcarriers by loading TGF-β1 to accelerated chondrogenesis. In addition, microcarriers could be loaded with active polypeptides and drugs to stimulate cartilage/osteochondral repair. For example, Zhang et al. prepared functional SF nanospheres microcarriers with peptides E7 (7-amino acid (EPLQLKM) small peptide) and KGN (Figure 5E). The microcarriers rapidly release peptides E7 in the early stage to recruit BMSCs and provide relatively slow and sustained KGN release to induce cartilage formation of BMSCs, thus promoting osteochondral regeneration.

31At present, different kinds of SF-based biomaterials for cartilage repair have their characteristics and advantages. It is worth mentioning that the role of the internal structure of SF-based biomaterials in cartilage/osteochondral repair cannot be ignored. Furthermore, SF-based biomaterials were designed in different structures, including directional structure, layered structure, and gradient structure. For cartilage/osteochondral tissue engineering, these elaborate structures were excellent in terms of improving the repair effects.

32In recent years, bioactive SF-based scaffolds with radial or axial pore structures have been developed to enhance cell migration and infiltration in vivo and in vitro, and further accelerate cartilage/osteochondral regeneration,. For example, Yang et al. combined SF and decellularized cartilage ECM by temperature gradient-guided thermal induced phase separation to produce composite scaffolds (axial pore structures scaffolds). The results showed that the axial pore structures endowed the scaffolds with remarkable mechanical properties and were suitable for adipose stem cell attachment and proliferation. Meanwhile, Chen et al. prepared three-dimensional SF/bacterial cellulose nanoribbon (SF/BCNR) composite scaffolds with a radial layered structure by directional freezing technology (Figure 6A). The radial interconnected pores promoted the flow of nutrients and waste to maintain cells vitality and provided robust mechanical performance. Therefore, the radial pore structures SF/BCNR scaffolds had high application potential in the field of cartilage/osteochondral regeneration.

33Although there have been numerous studies demonstrating that the directional structure of scaffolds could accelerate cartilage/osteochondral regeneration. However, the underlying mechanism of directional structure promoting cartilage/osteochondral repair is still unclear. At present, only comparative studies between different structures are available. For instance, Feng et al. constructed three kinds of biomimetic SF/collagen composite scaffolds with different pore structures (random pore, radial pore, and axially arranged pore), and evaluated their ability to regenerate osteochondral defects in the rabbit model (Figure 6B). The experimental data showed that the radial and axially arranged scaffolds had structural and mechanical anisotropy. Due to their high porosity, appropriate elastic modulus, good biocompatibility, and favorable three-dimensional microenvironment, these scaffolds enhanced cell activity in vitro and in vivo. Compared with random scaffolds, these properties contributed to a faster regeneration rate and better regeneration effect of osteochondral defects. Furthermore, in the comparison between radial pore structures and axial structures, the radial pore structures had a more significant effect on accelerating osteochondral repair.

34For cartilage/osteochondral tissue engineering, the key to the preparation of suitable scaffolds lies in the design of the structure. Meanwhile, scaffolds should closely simulate the physiological structure and environment of natural osteochondral tissue. In addition, on the basis of the physiological structure of natural subchondral bone and cartilage, many researchers have designed double-layer scaffolds with a cartilage layer and a subchondral bone layer to regenerate cartilage and subchondral bone at the same time -. For example, Ribeiro et al. constructed a new type of double-layer composite scaffold with multiple functional modifications for osteochondral regeneration (Figure 6C). The double-layer scaffolds included horseradish peroxidase-cross-linked SF (HRP-SF) cartilage layer, HRP-SF/ZnSr-doped β-tricalcium phosphate (β-TCP) subchondral bone-like layer (HRP-SF/dTCP layer).

35They found that different layers in the bilayer scaffolds stimulated the expression of different genes in cells to promote the repair of osteochondral. In order to overcome the poor integration of regenerated tissue and primary tissue, Wu et al. constructed a kind of double-layer SF composite scaffolds with a peripheral package (Figure 6D). The scaffold consisted of a dense and smooth biomimetic cartilage layer, a porous layer loaded with bone morphogenetic protein-2 (BMP-2) and a hydrogel coating loaded with transforming growth factor-β3 (TGF-β3). The experimental results showed that the double-layer SF scaffolds accelerated the regeneration of osteochondral, and the encapsulation hydrogel strengthened the integration of tissues. This special multi-layer design pointed out that researchers could achieve better cartilage/osteochondral repair by combining different types of SF-based biomaterials.

36With the deepening of the study of double-layer scaffolds, the inherent defects in osteochondral repair become more apparent. The poor integration of the repaired cartilage layer and subchondral bone layer leads to a clear boundary between them,. As we mentioned earlier in 3. Articular cartilage and subchondral bone, the calcified cartilage layer is an important physical barrier between cartilage and bone in different oxygen and nutrient environments. Without the calcified cartilage layer, the formation of regenerated cartilage and osteochondral tissue will be damaged. In order to closely simulate the natural osteochondral structure and regeneration environment, three-layer scaffolds have become available,. For instance, Ding et al. have successfully prepared biomimetic three-layer osteochondral scaffolds composed of SF/HA (Figure 6E). Significantly, the intermediate layer could play a role in preventing the cells within the chondral and the bony layers from mixing with each other. Meanwhile, the intermediate layer enhanced the integration of articular cartilage and subchondral bone.

37Many natural tissues, including bones, have multiple gradient structures to regulate cell behavior and guide tissue function -. Generally, if biomaterials misfit with the structure of cartilage/osteochondral, the long-term repair effect will be unsatisfactory,. To mimic the progressive osteochondral structure, researchers could design SF-based scaffolds with gradient pore sizes,. For example, Xiao et al. designed and prepared SF/CS/nano-hydroxyapatite (NHA) osteochondral scaffolds with gradient pore sizes. The gradient pore size structure achieved better proliferation and differentiation of the stem cells to complete the repair of osteochondral defects.

38The gradient modification of SF-based biomaterials could promote the gradient differentiation and distribution of cells to accelerate the repair of osteochondral. For example, Guo et al. prepared gradient silicified SF R5 peptide (GSSR5) composite biomaterials (Figure 6F). The GSSR5 composite biomaterials regulated the osteogenic differentiation of hBMSCs in the bone induction environment. The material facilitated the differentiation of cells in a manner consistent with the R5 gradient distribution. This cellular differentiation and distribution consistent with native tissue facilitated osteochondral repair.

39Generally, introducing gradient cues when preparing biomaterials requires special instruments and methods. Moreover, these instruments and methods are suitable for special materials only. To break through this limitation, Xu et al. prepared multifunctional building block hydrogels with tunable gradients using a simple low voltage electric field (Figure 6G). This novel approach enables this hydrogel to exhibit multiple gradients. These gradient structures drove cells to differentiate in different directions and accelerated osteochondral repair.

40Articular cartilage is a special load-bearing tissue that provides friction-free movement in synovial joints such as the knee joint, hip joint, or shoulder. The elastic modulus of articular cartilage is between 1-10kPa, which requires the biomaterials to repair cartilage need to match its mechanical properties. Therefore, when preparing biomaterials for osteochondral repair, researchers should regulate the mechanical properties of SF. For example, Chameettachal et al. studied the effect of adding micronized CS into freeze-dried porous SF scaffolds. According to the experimental data, micronized CS significantly enhanced the mechanical properties of SF scaffolds when the ratio of SF to CS microparticles was 2: 1 (the compressive strength of the non-reinforced SF scaffolds and reinforced scaffolds for SF:CS=1:2, 1:1 and 2:1 were 14.40 ± 3.2, 25.40 ± 5.4, 28.10 ± 6.3 and 30.10 ± 7.5 MPa, respectively). Meanwhile, Mirmusavi et al. added multi-walled carbon nanotubes (MWNTs) based on CS/SF scaffolds. Characterization found that CS and MWNTs improved the tensile strength compared with pure SF scaffolds (the ultimate tensile strength of pure SF scaffolds, unsaturated CS/SF scaffolds, saturated CS/SF scaffolds and saturated CS-MWNTs/SF scaffolds were 28.32 ± 3.27, 39.21 ± 2.06, 41.18 ± 2.75 and 46.68 ± 2.20 MPa, respectively).

41Normally, the articular cartilage is in a constant cycle of compression-rebound. In addition, during the process of cartilage/osteochondral repair, chondrocytes are more active in the dynamic environment. To prevent SF-based biomaterials premature damage after implantation in a dynamic environment of cartilage/osteochondral defect sites, SF-based biomaterials should exhibit excellent anti-fatigue ability. For instance, Huang et al. modified the SF with cholesterol or β-cyclodextrin, and prepared a kind of SF hydrogel with high mechanical strength (the compressive stress of SF hydrogels and modified-SF hydrogels were 0.1 and 3.16 MPa, respectively), high toughness and remarkable fatigue resistance (the modified-SF hydrogels remained in the original state without any deformation or strength degradation after 10 loading-unloading cycles at 60% strain). Interestingly, the hydrogel could achieve self-repair after stress damage. This indicated that the self-repair hydrogels had great potential in cartilage/osteochondral repair.

42The degradation properties of biomaterials directly affect the speed and quality of cartilage/osteochondral repair. Specifically, in the early stage of implantation, biomaterials should be stable to provide mechanical support for cells and tissues. After that, biomaterials need to be gradually degraded to match the new tissue growth.

43As a kind of protein material, the degradation rate of SF-based biomaterials is mainly affected by proteases. Most proteolytic enzymes tend to degrade non-crystalline SF. This suggests that SF-based biomaterials with controlled degradability can be prepared by changing the content of crystalline structures. For instance, Jin et al. constructed water-stable SF films through increasing the content of crystalline structures by methanol treatment. Additionally, the degradation rate of SF-based materials can also be regulated by the introduction of proteinase inhibitors or protease binding sequence SF. For example, Pritchard et al fabricated a slowly degraded SF-based biomaterial by incorporating protein inhibitors. This material can realize the controlled release of drugs. On the contrary, Huang et al. prepared a novel kind of transgenic SF by introducing matrix metalloproteinase-2-sensitive sequences into SF. The transgenic SF had a quicker degradation rate and was non-toxic to BMSCs.

44At present, tissue engineering has been considered as the primary therapeutic strategy for osteochondral regeneration and cartilage defect repair. However, the therapeutic effect of SF-based biomaterials on cartilage/osteochondral defects is still far from the expected effect. Consequently, it is necessary to functional modification of SF to improve the repair ability of SF-based biomaterials (Table 1).

45It was reported that SF-based biomaterials could enhance cartilage/osteochondral repair by loading bioactive materials. For example, Wang et al. synthesized SF/HA scaffolds functionalized with non-immunogenic aptamer (Apt19s). The Apt19s significantly improved the cell recruitment ability of SF/HA scaffolds and accelerated the repair of osteochondral. In order to promote cell adhesion, Cheng et al. hybridized SF with chitin to prepare 3D composite scaffolds (Figure 7A). The hybrid treatment of SF formed a rough surface that well mimicked the ECM to facilitate cell attachment and proliferation. In addition, SF-based biomaterials are able to carry active molecules to treat cartilage/osteochondral defects. For example, Li et al. fabricated PCL/SF scaffolds loaded with synovium-derived mesenchymal stem cells-specific affinity peptide (L7) via 3D bioprinting technology (Figure 7B). L7 peptide promoted the proliferation and differentiation of endogenous synovium-derived mesenchymal stem cells (SMSCs). Meanwhile, the L7 peptide stimulated the ECM production to provide a favorable microenvironment for meniscus regeneration and cartilage protection.

46Cartilage/osteochondral regeneration is regulated by a variety of growth factors. SF-based biomaterials loaded with growth factors can enhance the interaction with chondrocytes and osteoblasts. For example, Wu et al. fabricated biodegradable SF-gelatin porous scaffolds loading with ginsenoside Rb1 and TGF-β1. The results showed that the scaffolds loaded with Rb1 and TGF-β1 could create a good microenvironment for cartilage regeneration by promoting cartilage formation and inhibiting inflammation in vivo. Meanwhile, Chen et al. prepared SF-porous gelatin scaffolds loading with SDF-1α and TGF-β1. They found the scaffolds could accelerate cartilage repair by sustained release of SDF-1α and TGF-β1.

47SF-based biomaterials also could carry different cells to achieve the repair of different tissues. For example, Chen et al. designed a new type of elastin-like peptide modified SF composite scaffolds through simple and green dehydration treatment (Figure 7C). They found that the SF composite scaffolds enhanced the adhesion, proliferation, and differentiation of BMSCs or chondrocytes in vitro. Additionally, the SF composite scaffolds containing BMSCs or chondrocytes could promote bone repair and cartilage repair, respectively.

48In addition to bioactive molecules, growth factors and cells, loading exosomes can enhance the interaction between SF-based biomaterials and cells -. For instance, Zhang et al. prepared a composite SF hydrogel loaded with BMSC derived exosomes (Figure 7D). It was experimentally demonstrated that exosomes enhanced the recruitment of chondrocytes and accelerated cartilage repair. Furthermore, it is feasible to improve the performance of SF-based biomaterials for cartilage/osteochondral repair by loading a variety of nanoparticles (NPs),. For instance, Li et al. prepared SF hydrogels containing BMP-2/TGF-β1@CS NPs. The experimental results showed hydrogels released BMP-2/TGF-β1 from CS NPs to accelerate articular cartilage repair. Meanwhile, Min et al. contrasted alginate-poloxamer/SF dual network hydrogels loading with hyaluronic acid/chitosan-poly(dioxanone) complex nanoparticles (HA/CH-PDO NPs). They found that HA/CH-PDO NPs could slow-release bone morphogenic protein-7(BMP-7) to promote SMSCs differentiation into chondrocytes.

49Cartilage and osteochondral repair have always been a clinical challenge. In recent years, tissue engineering techniques have shown certain advantages in cartilage/osteochondral repair. As a natural macromolecular material with unique mechanical properties, superior processability and excellent biocompatibility, SF is regarded as an excellent tissue engineering material. This paper reviews the research on SF-based biomaterials for cartilage/osteochondral repair. From these studies, it is concluded that SF-based biomaterials have promising applications in cartilage/osteochondral repair. For example, due to the amorphous feature of SF-based hydrogels, they could closely fit irregular cartilage/osteochondral defects. In addition, the controllable degradation of SF-based hydrogels enables them to adapt to different pathological conditions and provides the basis for personalized treatment. Besides, SF-based scaffolds possess more robust mechanical properties compared to other biomaterial scaffolds, which could support cell adhesion and growth better. In addition, the cartilage/osteochondral repair ability of SF-based biomaterials can also be improved by loading polypeptides, cells, exosomes, nanoparticles and growth factors. These functional modifications combined with bioactive substances enhance the ability of SF-based biomaterials to recruit endogenous stem cells or chondrogenesis/osteogenesis. The functional modification of SF-based biomaterials also includes the strengthening of SF itself. For example, the tyrosine-specific diazonium coupling chemistry is used to design the surface chemistry and hydrophilicity of SF for promoting the adhesion, proliferation and differentiation of BMSCs on SF-based biomaterials. Genetic modification of silkworms by means of gene editing can also achieve the purpose of functional modifying SF. For example, insert RGD gene sequence into mulberry silkworm SF gene to produce mulberry silkworm SF containing RGD sequence.

50Currently, the clinical application researches (data from www.ClinicTrials.gov) demonstrate that the clinical applications of SF-based biomaterials are relatively mature in surgical sutures and wound dressings. In addition, SF-based scaffolds applied to breast reconstruction have shown remarkable improvements. It is worth noting that clinical studies on the use of SF-based scaffolds for meniscal cartilage repair have also been conducted in recent years (Table 2). Although SF-based biomaterials have promising applications, they still face several challenges in sericulture. On the one hand, the production standardization of cocoon (SF source) is insufficient due to the lack of international testing standards. On the other hand, the quality of silkworm cocoons is affected by several production factors, including the silkworm species, breeding method, temperature, humidity, light and hygiene conditions of the breeding environment, transportation and storage methods. Additionally, the inadequate mechanization of the sericulture industry is caused by the complex breeding conditions of silkworms and the seasonal supply of mulberry leaves. Meanwhile, the low degree of mechanization requires a lot of manual operation and leads to the high price of SF. Besides, in the aspects of SF-based biomaterials preparation, there are several factors that inhibit their clinical transformation and commercialization for cartilage/osteochondral repair. (1) The preparation process of SF-based biomaterials is tedious. The LiBr dissolution system is the most commonly used system for dissolving SF. However, the dissolution of SF by this system is time-consuming (SF is slow to dissolve and requires further dialysis, ultrasound and centrifugation after dissolution), and the obtained SF solution is insufficiently concentrated. (2) The forms and preparation methods of completed SF-based biomaterials are undiversified. (3) Most SF-based biomaterials are non-intelligent and difficult to re-intervene after implantation. (4) The entire field of cartilage/osteochondral repair is at a bottleneck stage, and it is difficult to make significant breakthroughs by only aiming to construct new SF-based biomaterials. Fortunately, some scholars have explored and discussed these limiting factors, and made certain improvements. For example, Wang et al. used CaCl2/formic acid solution (weight ratio of 1:20) instead of LiBr solution to dissolve SF. By this dissolution system, the stable and homogeneous SF solution could be obtained in 5 min. This work significantly optimized the preparation process of SF-based biomaterials. To promote the diversification of SF-based biomaterials, it is necessary to combine the advantages of various preparation methods. For instance, Chen et al. combined the advantages of electrostatic spinning, freeze-drying, and gas foaming to prepare SF-based scaffolds with excellent performance. Meanwhile, combining the advantages of various types of SF-based biomaterials is equally important to promote diversification. For example, Wu et al. have combined SF-based scaffolds and SF-based hydrogels to construct new SF-based composite biomaterials that integrate the advantages of scaffolds and hydrogels. In addition, designing SF-based biomaterials with magnetic field response is an effective way to advance the intelligence process. For instance, Wang et al. prepared a thermosensitive and pH-sensitive SF-based biomaterial that would morphological transform in alternating magnetic fields. Notably, Chen et al. discussed the bottleneck of bone tissue repair and the construction of bone organoids. Similarly, SF-based biomaterials may construct the cartilage/osteochondral-like tissues in vitro, which could be transplanted to achieve cartilage/osteochondral repair.

51In conclusion, SF-based biomaterials have been widely studied in the field of cartilage/osteochondral repair. In the future, Researchers should develop special mechanical equipment suitable for silkworm rearing and promote mechanization to reduce the production cost of SF-based biomaterials. Meanwhile, it is also necessary to improve the preparation process of SF-based biomaterials. With the assistance of a variety of preparation methods and functional modifications, SF-based biomaterials should be more diversified and intelligent. It is important that SF-based biomaterials transform from 3D biomaterials to 4D biomaterials in combination with the fourth dimension "time". With the progress of cartilage/osteochondral repair process, 4D SF-based biomaterials can be adaptively adjusted to ensure the rapid repair of cartilage/osteochondral. We believe that SF-based biomaterials will be promising for clinical applications in cartilage/osteochondral repair.