Recent Advances in Understanding the Role of Cartilage Lubrication in Osteoarthritis

The remarkable lubrication properties of normal articular cartilage play an essential role in daily life, providing almost frictionless movements of joints. Alterations of cartilage surface or degradation of biomacromolecules within synovial fluid increase the wear and tear of the cartilage and hence determining the onset of the most common joint disease, osteoarthritis (OA). The irreversible and progressive degradation of articular cartilage is the hallmark of OA. Considering the absence of effective options to treat OA, the mechanosensitivity of chondrocytes has captured attention. As the only embedded cells in cartilage, the metabolism of chondrocytes is essential in maintaining homeostasis of cartilage, which triggers motivations to understand what is behind the low friction of cartilage and develop biolubrication-based strategies to postpone or even possibly heal OA. This review firstly focuses on the mechanism of cartilage lubrication, particularly on boundary lubrication. Then the mechanotransduction (especially shear stress) of chondrocytes is discussed. The following summarizes the recent development of cartilage-inspired biolubricants to highlight the correlation between cartilage lubrication and OA. One might expect that the restoration of cartilage lubrication at the early stage of OA could potentially promote the regeneration of cartilage and reverse its pathology to cure OA.


Introduction
Articular cartilage is an avascular, aneural and alymphatic connective tissue (which determines its very poor self-recovery ability) lining the bone ends of diarthrodial joints [1]. Combined with excellent load-bearing capacities, this cushion of articular cartilage, especially the outer surface of cartilage, provides extremely low friction with a friction coefficient as low as 10 −3 under a wide range of physiological pressures (even up to 100 atm) to maintain daily movements during a person's lifetime [2][3][4]. Recently, researchers noticed that the increase of cartilage friction plays a determining role in initiating the most common degenerative joint disability disease, that is osteoarthritis (OA), which is mainly characterized by the progressive but irreversible degradation of articular cartilage [5][6][7][8][9]. Briefly, aging-related changes or lesions usually lead to the compromisation of the outer surface of cartilage, which subsequently causes an increase of the friction coefficient. Chondrocytes, as the only cell type in cartilage, in return, up-regulate the secretion of the cartilage-degrading enzymes, such as matrix metalloproteinases (MMPs) and a disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) to degrade the type II collagen and aggrecan (the main components of cartilage matrix), respectively. Moreover, the cartilage degradation fragments are phagocytosed cells (such as macrophage and synovial fibroblasts) to inflame the synovium, promoting the production of MMPs and ADMATS to break down cartilage and deteriorate lubrication [10][11][12][13][14][15]. In this way, a positive feedback loop is formed due to the mutually reinforcing effect of increased friction and secretion of degradation enzymes, resulting in the progressive till total degradation of articular cartilage. Early, moderate, and late stages of OA can be classified mainly based on the degree of cartilage degradation. It is thought the breakdown of the type II collagen network initiates the point where the OA is considered irreversible [16][17][18]. Is it possible to modify and prevent the disease at its early stage especially considering the role of cartilage lubrication?
Among the population over 60 years old, 9.6% of men and 18.0% of women have OA symptoms, which made OA a serious global disease defined by the Osteoarthritis Research Society International in 2016 [19]. Cell-based therapies or cartilage tissue engineering for cartilage repair or regeneration have made significant advances recently [20,21], such as the adhesive peptide-based 3D scaffolds for cell culture [22], however, considerable efforts are still required for the tribological properties and durability of the neocartilage before clinical translation. Currently, the main nonsurgical options for OA treatment before the end of OA include using analgesics, anti-inflammatory drugs (such as acetaminophen), inhibitors (such as cyclooxygenase), or articular injection of hyaluronan and corticosteroids [23,24]. However, these nonsurgical options usually are highly controversial due to the nonuniversal effects when compared to that in state-of-the-art placebo-controlled [25,26]. Therefore, it is highly important to shed light on the remarkable lubrication of cartilage and the correlation between lubrication and cartilage regeneration, with the aim to improve the understanding of OA and encourage the development of approaches to alleviating and even treating it.
Intense research efforts have been made to deciphering the molecular mechanisms of cartilage lubrication since the 1930s [27]. A brief history of cartilage lubrication and the potential of injectable formulations to treat OA have been presented recently [18,28,29]. In this review, starting from the articular cartilage, we focus on discussing the boundary lubrication mechanism of cartilage, following by the shear-associated metabolism of chondrocytes, which could be fully taken advantage of to trigger cartilage regeneration, and finally, we review the recent biolubricants inspired by cartilage lubrication, a common theme of these inspired biolubricants is to resurface and restore the boundary lubrication of cartilage, and subsequently, alleviate the symptoms or even reverse the pathological progression of OA by promoting the regeneration of cartilage.

Articular Cartilage
A joint is a place where two or more bones meet, allowing the skeleton to move. Usually, joints differ in shape and structures according to the required movement and load, so we focus on discussing the diarthrosis knee joints. The knee joint consists of a joint capsule, ligaments, synovium, and the articular cartilage lining the ends of the opposing bones [1]. Synovial fluid within the capsule provides lubrication and nutrition, while the synovial membrane, a sac-like structure, surrounds the joint cavity and synovial fluid [30]. In daily life, it's essential to maintain the normal structure of the synovial joint.

Structure and Components of Articular Cartilage
Articular cartilage (also referred to as hyaline cartilage) is a highly hydrated glassy connective tissue comprised of chondrocytes (the only cell type embedded within cartilage) and the extracellular matrix (ECM) which is secreted and maintained by chondrocytes [1,31]. The ECM is predominantly composed of type II collagen bundles, negatively charged proteoglycans, non-collagenous proteins, water, and ions (primarily Na + and Cl − ). 11 types of collagens could be found in articular cartilage. Among them, type II collagen, rep-resenting 90-95% of all collagens in ECM and accounting for 60% of the dry weight of articular cartilage, forms a crosslinked core network, to enable the cartilage tensile and shear strength [32][33][34].
The type II collagen fibril networks interweave with proteoglycans (such as aggrecans), the second most abundant macromolecules in articular cartilage, which contribute to the lubrication and load-bearing properties of cartilage due to their strong hydration [12]. Proteoglycans are proteins covalently attached to glycosaminoglycans (long repetitive dimers of hexosamine and uronic acid). The most prevalent and largest in size of proteoglycans in articular cartilage are aggrecans, representing a bottle-brush structure with a polypeptide as backbone and chondroitin sulfate and keratan sulfate as the side chains. Usually, there are over 100 chondroitin sulfate and 20-40 keratan sulfate chains in one aggrecan molecule [35,36]. Therefore, the aggrecans are highly sulfated and negatively charged conferred by the sulfate groups in their side chains. These negative charges attract large water molecules to further strengthen the cartilage matrix. Hyaluronic acid or hyaluronan (HA), the only non-sulfated glycosaminoglycan, is built by the repeated dimers of b-D-(1,4)-N-acetylglucosamine and b-D-(1-3) glucuronic acid with a molecular weight up to 6 MDa [37]. HA and aggrecan form an extensive aggregate, comprising of a central HA to which about 100 aggrecan molecules are non-covalently attached via the link protein, thereby stabilizing this aggregation [38]. These aggregates further bind to the type II collagen fibers and have been demonstrated to play a significant role in cartilage lubrication (will be discussed later). The other proteoglycans could be found in the backbone of articular cartilage including biglycan, decorin, fibromodulin, lumican and perlecan. Non-collagenous proteins mainly include fibronectin, cartilage oligomeric matrix protein (COMP), laminin, chondronectin et al. [39,40].
Microscopically, three zones of the articular cartilage can be distinguished, that is the superficial zone (also referred to as the lamina splendens), the middle or transitional zone and the deep zone, as shown in Figure 1. These three zones of cartilage exhibit heterogeneity in the composition of ECM, which is reflected in the organization of collagen, size, phenotype, and metabolic activity of chondrocytes [41]. The superficial zone lies in the outer surface of articular cartilage, constitutes 10-20% of the full thickness of adult cartilage, characterized by two aspects, one is the type II collagen fibers, with a diameter of 30-35 nm, which are densely arranged and parallel to the articular surface, the other one is the long axis of flat and ellipsoidal chondrocytes parallel to the surface of the cartilage. Normally, the lubricating molecules are within the superficial zone. The middle zone is constituted of 40-60% of the thickness of total cartilage, the chondrocytes, exhibiting round or rectangular shape, are randomly distributed with their long axis perpendicular to the cartilage surface. The fibrils of type II collagen form an oblique transitional network and appear as arcades. The deep or radial zone constitutes the last 20-30% of the thickness of the cartilage. The shape of chondrocytes is round, the fibrils of type II collagen, with the largest diameter (40-80 nm), are perpendicular to the cartilage surface. The predominant biomechanical properties of the main three different zones of articular cartilage are summarized in Table 1.

Mechanotransduction of Chondrocytes
Mechanotransduction refers to the process of sensing and converting mechanical signals into biochemical signals to regulate cellular activities [52]. Located on the joint surfaces, a range of static and dynamic stresses (standing, walking, and jogging) are applied

Mechanotransduction of Chondrocytes
Mechanotransduction refers to the process of sensing and converting mechanical signals into biochemical signals to regulate cellular activities [52]. Located on the joint surfaces, a range of static and dynamic stresses (standing, walking, and jogging) are applied on articular cartilage. It is well documented the metabolism of chondrocyte is strongly regulated by the normal stress (compression), static loads were shown to be detrimental to the anabolic processes (biosynthesis of type II collagen and proteoglycans) while oscillatory loads with moderate frequencies and amplitudes (compression strain under 20%) have been shown to effectively promote the matrix accumulation and decrease the secretion of TNF-a and IL-6, which contribute to the degradation of matrix [53][54][55][56].
Unlike normal stress, shear stress gives rise to the shear strain of the cartilage, and then be transmitted to the chondrocytes, especially those within the superficial zone. Many previous studies suggested the shear stress activates chondrocytes and up-regulates the proinflammatory cytokines (TNF-a, and the family of interleukins) and MMPs [55,57], which elicit the degradation of cartilage, most of the underlying signaling pathways remain unclear, but we can summarize some of them in Figure 2 according to the previous reports [57][58][59][60][61][62][63]. The chondrocytes undergo a phenotypic switch to aberrantly express catabolic enzymes when the shear strain exceeds ≈1% estimated by Klein and Lin very recently under relative ideal circumstances considering the complexity of cartilage [64]. Moreover, the increased shear strain, or the shear stress or friction, induces chondrocyte apoptosis, which has been demonstrated by previous studies [65,66]. The mechanism regulating shear-mediated chondrocytes expression of IL-6 and MMPs and apoptosis is shown in Figure 2. The stimuli of high shear stress induce chondrocytes to express cyclooxygenase (Cox 2), which inhibits the activity of phosphatidylinositol 3-kinase (PI3-K), following decreases in antioxidant capacity to lead to chondrocyte apoptosis.
26, x FOR PEER REVIEW 5 of 20 on articular cartilage. It is well documented the metabolism of chondrocyte is strongly regulated by the normal stress (compression), static loads were shown to be detrimental to the anabolic processes (biosynthesis of type II collagen and proteoglycans) while oscillatory loads with moderate frequencies and amplitudes (compression strain under 20%) have been shown to effectively promote the matrix accumulation and decrease the secretion of TNF-a and IL-6, which contribute to the degradation of matrix [53][54][55][56]. Unlike normal stress, shear stress gives rise to the shear strain of the cartilage, and then be transmitted to the chondrocytes, especially those within the superficial zone. Many previous studies suggested the shear stress activates chondrocytes and up-regulates the proinflammatory cytokines (TNF-a, and the family of interleukins) and MMPs [57,58], which elicit the degradation of cartilage, most of the underlying signaling pathways remain unclear, but we can summarize some of them in Figure 2 according to the previous reports [57,[59][60][61][62][63][64]. The chondrocytes undergo a phenotypic switch to aberrantly express catabolic enzymes when the shear strain exceeds ≈1% estimated by Klein and Lin very recently under relative ideal circumstances considering the complexity of cartilage [65]. Moreover, the increased shear strain, or the shear stress or friction, induces chondrocyte apoptosis, which has been demonstrated by previous studies [66,67]. The mechanism regulating shear-mediated chondrocytes expression of IL-6 and MMPs and apoptosis is shown in Figure 2. The stimuli of high shear stress induce chondrocytes to express cyclooxygenase (Cox 2), which inhibits the activity of phosphatidylinositol 3-kinase (PI3-K), following decreases in antioxidant capacity to lead to chondrocyte apoptosis. High shear stress activates the Rac-1/Cdc42, which then transactivates MKK7 to regulate JNK2 activation, and this, in turn, triggers c-Jun phosphorylation which induces the overexpression of Cox-2. Cox-2 suppresses the activity of P13-K, which represses Nrf-2 to decrease the antioxidant capacity to permit disruption of the integrity of mitochondrial, activation of caspase-9, and the apoptosis of chondrocytes. The expression of Cox-2 also triggers the expression of PGE2, as well as the concomitant downstream expression of receptor EP2, as a result, IL-1b is rapidly and sustainably synthesized. Moreover, up-regulation of TLR4 due to high shear stress activates ERK1/2, P13-K and JNK Figure 2. Schematic of shear-induced cartilage matrix degradation and apoptosis of chondrocytes. High shear stress activates the Rac-1/Cdc42, which then transactivates MKK7 to regulate JNK2 activation, and this, in turn, triggers c-Jun phosphorylation which induces the overexpression of Cox-2. Cox-2 suppresses the activity of P13-K, which represses Nrf-2 to decrease the antioxidant capacity to permit disruption of the integrity of mitochondrial, activation of caspase-9, and the apoptosis of chondrocytes. The expression of Cox-2 also triggers the expression of PGE2, as well as the concomitant downstream expression of receptor EP2, as a result, IL-1b is rapidly and sustainably synthesized. Moreover, up-regulation of TLR4 due to high shear stress activates ERK1/2, P13-K and JNK pathways, which is also activated by IL-1b, then regulates NF-kB-dependent IL-6 and MMP synthesis. Abbreviations: Cyclooxygenase-2 (Cox-2), mitogen-activated protein kinase 7 (MKK7), nuclear factor-kB (NF-kB), prostaglandin E2 (PGE2), Interleukin-1b (IL-1b), Interleukin-6 (IL-6), phosphatidylinositol 3-kinase (PI3-K), c-Jun N-terminal kinase 2 (JNK2), NF-E2 related factor 2 (Nrf2), matrix metalloproteinases (MMPs), toll-like receptor 4 (TLR4), extracellular signal-regulated kinase (ERK1/2).
Collectively, favoring cartilage longevity requires dynamic normal stress combined with quite low or even zero shear stress. To maximize the regeneration of cartilage, the main way is to decrease the shear strain arising from the shear stress by reducing the friction coefficient of cartilage. Therefore, a scenario in which treating or healing OA at its early stage by restoring the lubrication of OA-damaged cartilage can be imagined.

Lubrication Mechanism of Articular Cartilage
The friction coefficients of normal articular cartilage can be as low as~0.002-0.02 [64]. The lubrication properties of articular cartilage have drawn attention since the 1930s and many theories have been proposed to claim the mechanisms behind the ultra-low friction of cartilage. Over the development of these decades, one could simplify these theories into fluid film lubrication and boundary lubrication [8].
In the context of fluid film lubrication, an early and classical lubrication mechanism purported that a fluid film that completely separates the opposing surfaces to prevent direct contact between surfaces is formed during articulation, and this fluid film enables the low friction coefficient [67][68][69]. Considering the "biphasic" structure (solid phase of the matrix and the liquid phase of the water within the cartilage matrix), the models to elucidate the fluid film lubrication mechanism were also further developed. When cartilage surfaces slide past each other under compression, the interstitial fluid within the deformed cartilages, on one hand, is extruded out to form a film between the cartilage surfaces, on the other hand, the interstitial fluid is pressurized (due to the nature of incompressible fluid) to support the major normal load, and consequently reduces the proportion of the normal load that supported by the cartilage matrix [70,71]. Thus, the cartilage exhibits low friction. One question that should be noticed here is the pressure used in these related experiments was not very high (about 0.2 MPa, especially compared with that exerted on the cartilage in vivo), so the formed fluid film might be squeezed out in the case of high pressure. In this situation, boundary lubrication is proposed.
In the context of boundary lubrication, friction arises from the partial contact between the opposing cartilage surface asperities. The friction coefficient is relatively independent of the sliding speed and normal load and is dominated by the outer surface of the articular cartilage and the synovial fluid, especially the molecules within the synovial fluid [72,73]. The main components in the synovial fluid from the human joint include hyaluronic acid, albumin, globulin, lubricin, phospholipids, and glycosaminoglycans [74]. Among these components, lubricin and phospholipids are commonly found at the surface of cartilage at high concentrations.
Typically, three main different research approaches were shown to explore which component in synovial fluid is responsible for the ultra-low friction of articular cartilage: selectively enzyme-degraded specific ones or more components in synovial fluid or cartilage [75], or selectively added ones or more purified (or artificially synthesized) components to the lubricating bath [72] or direct measurement of the lubrication abilities of one or more components at the molecular level [76][77][78]. Radin et al. found no difference in the friction coefficient when HA was removed from the synovial fluid using hyaluronidase [79]. It was also found that the lubrication ability of HA was quite poor when confined to HA molecules under limited space [80]. Collectively, HA, on its own, cannot explain the remarkable lubrication of articular cartilage. Lubricin, also known as proteoglycan 4, has been reported to significantly reduce the friction coefficient of cartilage in vitro, and its deficiency induces chondrocyte (especially those embedded within the superficial zone) apoptosis in vivo by knocking out the PRG 4 gene [65]. Therefore, it is thought lubricin is a key lubricant for the boundary lubrication of articular cartilage. Molecularly, Zappone et al. found the friction coefficient of surfaces bearing lubricin was relatively low (µ = 0.02-0.04) at the pressure of about 6 atm, while increased to ≈0.2 under higher pressure [81]. Additionally, when constructing the surfaces bearing lubricin and HA layer by layer, the friction coefficients are still quite high compared with those found in articular cartilage, especially under the pressure at the level of a few atmospheres (atm.) [80]. Phospholipids, such as dipalmitoyl-phosphatidylcholine, phosphatidylethanolamine and sphingomyelin, become active upon binding with the calcium ions [82], have also been reported to bind to the cartilage surface and act as the boundary biolubricants. Many studies have shown a strongly absorbed active phospholipid layer that can exhibit extremely low friction coefficients under pressures up to hundreds of atm., which is near to the level of pressure the articular cartilage is exposed to in vivo [83,84]. However, when treated with phospholipase or added phospholipids (containing HA and lubricin), there was little change in the friction coefficients of the synovial fluid [72,85]. Overall, how the components in the synovial fluid enable the extremely low friction coefficient, especially under high pressure, has not yet been identified.
Recently, as shown in Figure 3, a new picture emerged where it is the synergy between the molecules in the synovial fluid, especially those mentioned above, that determines the lubrication of articular cartilage under severe joint loading [86,87]. Specifically, HA associates with aggrecan via the link protein to form a bottle-brush structure where HA serves as the backbone and aggrecan as the side chain. HA and lubricin also form a complex, which is physically trapped on the surface and contributes to effectively eliminate the wear damage of the cartilage. HA also shows a high affinity with phospholipids according to the previous report [88]. Thus, Klein et al. pointed to a scenario in which HA (alone cannot bind to cartilage surface) anchors at the outer surface of cartilage with the assistance of lubricin, then further complexes with the phospholipids to act as an effective boundary lubricant to enable the remarkable lubrication of articular cartilage at high pressure, via the hydration mechanism [89][90][91]. In this way, lubricin serves as a "carrier" between the HA, phospholipids, and the outer surface of the articular cartilage.
Molecules 2021, 26, x FOR PEER REVIEW 7 of 20 apoptosis in vivo by knocking out the PRG 4 gene [66]. Therefore, it is thought lubricin is a key lubricant for the boundary lubrication of articular cartilage. Molecularly, Zappone et al. found the friction coefficient of surfaces bearing lubricin was relatively low (μ = 0.02-0.04) at the pressure of about 6 atm, while increased to ≈0.2 under higher pressure [82]. Additionally, when constructing the surfaces bearing lubricin and HA layer by layer, the friction coefficients are still quite high compared with those found in articular cartilage, especially under the pressure at the level of a few atmospheres (atm.) [81]. Phospholipids, such as dipalmitoyl-phosphatidylcholine, phosphatidylethanolamine and sphingomyelin, become active upon binding with the calcium ions [83], have also been reported to bind to the cartilage surface and act as the boundary biolubricants. Many studies have shown a strongly absorbed active phospholipid layer that can exhibit extremely low friction coefficients under pressures up to hundreds of atm., which is near to the level of pressure the articular cartilage is exposed to in vivo [84,85]. However, when treated with phospholipase or added phospholipids (containing HA and lubricin), there was little change in the friction coefficients of the synovial fluid [73,86]. Overall, how the components in the synovial fluid enable the extremely low friction coefficient, especially under high pressure, has not yet been identified.
Recently, as shown in Figure 3, a new picture emerged where it is the synergy between the molecules in the synovial fluid, especially those mentioned above, that determines the lubrication of articular cartilage under severe joint loading [87,88]. Specifically, HA associates with aggrecan via the link protein to form a bottle-brush structure where HA serves as the backbone and aggrecan as the side chain. HA and lubricin also form a complex, which is physically trapped on the surface and contributes to effectively eliminate the wear damage of the cartilage. HA also shows a high affinity with phospholipids according to the previous report [89]. Thus, Klein et al. pointed to a scenario in which HA (alone cannot bind to cartilage surface) anchors at the outer surface of cartilage with the assistance of lubricin, then further complexes with the phospholipids to act as an effective boundary lubricant to enable the remarkable lubrication of articular cartilage at high pressure, via the hydration mechanism [90][91][92]. In this way, lubricin serves as a "carrier" between the HA, phospholipids, and the outer surface of the articular cartilage.

Biolubricants Inspired by the Cartilage Lubrication with the Aim to Treat OA
As mentioned above, a layer of complex formed by the synergy of synovial fluid components acts as boundary lubricants to reduce the friction between opposing articular cartilage surfaces, especially at high pressure, and effectively ensure the nearly frictionless

Biolubricants Inspired by the Cartilage Lubrication with the Aim to Treat OA
As mentioned above, a layer of complex formed by the synergy of synovial fluid components acts as boundary lubricants to reduce the friction between opposing articular cartilage surfaces, especially at high pressure, and effectively ensure the nearly frictionless motion of knee joints. However, disruption of the integrity of cartilage surface and alteration in the composition of synovial fluid compromise this boundary lubricant layer, initiates a cascade of biological and biomechanical events lead to dysfunction of boundary lubrication and then triggers the development of OA [92,93]. This association between the loss of lubrication layer at the outer surface of the articular cartilage surface and OA pathogenesis has been demonstrated in OA animal models [94,95]. In the context of the correlation between dysfunction of cartilage lubrication and the stimuli-response of chondrocytes, especially the increased secretion of catabolic matrix degraded enzymes and the decreased secretion of cartilage matrix, researchers have focused on alleviating and treating OA by restoring the lubrication properties of damaged or OA affected articular cartilage. Hydration lubrication is thought to be the dominant lubrication mechanism in biological systems, so highly hydrophilic monomers are usually the prior choice used to design biolubricants [90] so that a robust hydration layer to reduce the friction could be formed once the designed biolubricants resurface the cartilage. The typical biolubricants inspired by cartilage lubrication with the aim to treat OA developed recently are summarized in Table 2, most of them are described in detail in the following sections.

Natural Lubricants and Their Derivatives
Hyaluronic acid, gives the synovial fluid high viscoelastic properties, was thought to ensure the lubrication of cartilage in the 1950s, especially when associated with significantly decreased viscoelasticity of synovial fluid from OA patients [111,112]. Therefore, articular injection of high molecular weight HA (usually above 1 MDa) has been used to alleviate the symptoms of OA clinically as viscosupplementation. However, the results are controversial due to the universal individual differences. The main reasons may be due to the short retention time in the joint cavity and the enzymatic degradation [96]. In this case, chemical modification and cross-linking are the main two options to try to bypass these two shortcomings. Elisseeff et al. developed a strategy to concentrate the HA to re-construct the boundary lubrication of OA-damaged cartilage by creating a HA-recruiting coating (called HABpep) on the outer surface of cartilage (Figure 4a) [96]. This method significantly increased the retention of HA in the joint and even improved the lubrication of OA-damaged cartilage to the level of healthy cartilage (Figure 4b,c). Based on these results, similarly, they further employed HA binding peptide-polymer systems (HABP2-8-arm PEG-COLBP) in vivo with the aim to reduce OA progression and treat OA. After localizing to the cartilage defects, these systems showed positive therapeutic results by decreasing the level of inflammatory factors and increasing the expression of aggrecan, and thus, slow down the cartilage degradation [97]. Recently, chondroitin sulfate was also used to be modified by HA binding peptide (used to recruit HA in synovial fluid) and type II collagen peptide (used to bind to the surface of cartilage) simultaneously, this library of peptide-modified chondroitin sulfate also resurfaced and then decreased the friction coefficients of articular cartilage without modification of the cartilage [98]. In addition to peptide modification, Zhang et al. grafted 2-methacryloyloxyethyl phosphorylcholine (MPC), a zwitterionic monomer with robust lubrication performance [113], to HA with low and high molecular weights (lHAMPC and hHAMPC) in order to enhance HA lubrication ability [99]. The friction coefficients of polystyrene microspheres versus silicon wafer lubricated by lHAMPC and hHAMPC aqueous solution were significantly decreased when compared with the control groups (water and HA solution). Additionally, HAMPC could upregulate the chondrocytes anabolic genes and downregulate the chondrocytes catabolic genes, especially the hHAMPC. However, the lubricating properties of these two polymers were not demonstrated by cartilage.
Very recently, recruiting the lubricin, a natural biolubricant, from the surrounding milieu emerged as another strategy to restore the lubrication of cartilage, as shown in Figure 5, Sharma et al. modified the chitosan with catechol (Chi-C) to prepare a mucoadhesive biopolymer, which can bind to the cartilage surface by electrostatic interaction and/or the chemical reaction between the oxidized derivative of catechol and amines exposed in degraded cartilage so that the sessile Chi-C recruited the lubricin (the oxidized catechol reacted with amines in lubricin) to restore the lubrication layer of damaged cartilage. Interestingly, this method reduced the friction coefficients of degraded cartilage to the level of healthy cartilage [100]. Additionally, they extended this thought to polycarbonate urethane, a biomaterial used for meniscus replacements, to reduce the friction coefficients when sliding against cartilage [114]. Very recently, recruiting the lubricin, a natural biolubricant, from the surrounding milieu emerged as another strategy to restore the lubrication of cartilage, as shown in Figure 5, Sharma et al. modified the chitosan with catechol (Chi-C) to prepare a mucoadhesive biopolymer, which can bind to the cartilage surface by electrostatic interaction and/or the chemical reaction between the oxidized derivative of catechol and amines exposed in degraded cartilage so that the sessile Chi-C recruited the lubricin (the oxidized catechol reacted with amines in lubricin) to restore the lubrication layer of damaged cartilage. Interestingly, this method reduced the friction coefficients of degraded cartilage to the level of healthy cartilage [101]. Additionally, they extended this thought to polycarbonate urethane, a biomaterial used for meniscus replacements, to reduce the friction coefficients when sliding against cartilage [115].

Fully-Synthetic Polymer Lubricants
Many fully synthetic biolubricants were designed and synthesized and exhibited remarkable lubrication at the molecular level measured by a very typical and sensitive technique, surface force balance, or surface force apparatus [115][116][117]. In this section, we only focus on discussing the fully synthetic biolubricants developed recently, which clearly claimed to restore the lubrication of cartilage.
Benetti et al. designed and synthesized a series of bottle-brush graft-copolymers where linear poly-2-methyl-2-oxazoline (LPMOXA) was alternatively tethered together with aldehyde-bearing hydroxybenzaldehyde (HBA) to the polyglutamic acid (PGA) (PGA-LPMOXA-HBA), as shown in Figure 6, these copolymers anchored on the surface of degraded cartilage via Schiff base between aldehyde and amino groups (presented in collagen), thus a boundary lubrication layer was formed and the part of PMOXA provided lubricity, the lubrication test showed the friction coefficients of cartilage resurfaced by PGA-LPMOXA-HBA were significantly reduced [101]. Additionally, due to the biopassivity provided by PMOXA, protein adsorption was prevented by the films of PGA-LPMOXA-HBA. Followingly, they replaced the LPMOXA with cyclic poly(2-methyl-2-oxazoline) (CPMOXA) to synthesis another series of biolubricants with different densities of side chains, called PGA-CPMOXA-HBA [102]. After chemosorbing on the surface of cartilage via Schiff bases, PGA-CPMOXA-HBA generated denser lubricious films due to the absence of chain ends of the exposed CPMOXA when compared with PGA-LPMOXA-HBA, so PGA-CPMOXA-HBA reduced the friction coefficients of treated cartilage even under the higher load (Figure 6b). Noteworthy, the films of PGA-CPMOXA-HBA effectively prevent the penetration of catabolic enzymes (MMPs and ChABC) and thus, reduced cartilage degradation up to 50% (Figure 6c,d).

Fully-Synthetic Polymer Lubricants
Many fully synthetic biolubricants were designed and synthesized and exhibited markable lubrication at the molecular level measured by a very typical and sensitive te nique, surface force balance, or surface force apparatus [116][117][118]. In this section, we o focus on discussing the fully synthetic biolubricants developed recently, which clea claimed to restore the lubrication of cartilage.
Benetti et al. designed and synthesized a series of bottle-brush graft-copolym where linear poly-2-methyl-2-oxazoline (LPMOXA) was alternatively tethered togeth with aldehyde-bearing hydroxybenzaldehyde (HBA) to the polyglutamic acid (PG (PGA-LPMOXA-HBA), as shown in Figure 6, these copolymers anchored on the surf of degraded cartilage via Schiff base between aldehyde and amino groups (presented collagen), thus a boundary lubrication layer was formed and the part of PMOXA provid under the higher load (Figure 6b). Noteworthy, the films of PGA-CPMOXA-HBA effectively prevent the penetration of catabolic enzymes (MMPs and ChABC) and thus, reduced cartilage degradation up to 50% (Figure 6c,d). Additionally, Lawrence et al. designed a library of bottle-brush copolymers with different size of backbone and side chains, and density of side chains to mimic lubricin. Highly hydrophilic polyethylene glycol (PEG) was tethered to polyacrylic acid (PAA) to build the biolubricant pAA-g-PEG, and a thiol terminus was anchored so that pAA-g-PEG could bind to the cartilage surface. These copolymers could resurface the cartilage within 39 min and then effectively lubricated the cartilage [104,106]. Recently, they further evaluated the lubrication of these copolymers using the cartilage harvested from surgery-induced rats 8 weeks after articular injection. There were no differences in the changes of friction coefficients in all treated groups, which may be due to the influence caused by the limited size of rat cartilage. However, the other histological results may demonstrate the correlation between restoring lubrication and attenuating the progression of OA [105].

Nanospheres or Nanoparticles and Nano (Micro) Gels
Nanospheres or nanoparticles and microgels modified with highly hydrated polymers were also developed as biomimetic lubricants recently. The major advantage of these biomimetic lubricants is to combine lubrication with drug-releasing so that they present dual functions. In recent two years, many such biomimetic lubricants emerged. Zhou et Additionally, Lawrence et al. designed a library of bottle-brush copolymers with different size of backbone and side chains, and density of side chains to mimic lubricin. Highly hydrophilic polyethylene glycol (PEG) was tethered to polyacrylic acid (PAA) to build the biolubricant pAA-g-PEG, and a thiol terminus was anchored so that pAA-g-PEG could bind to the cartilage surface. These copolymers could resurface the cartilage within 39 min and then effectively lubricated the cartilage [103,105]. Recently, they further evaluated the lubrication of these copolymers using the cartilage harvested from surgeryinduced rats 8 weeks after articular injection. There were no differences in the changes of friction coefficients in all treated groups, which may be due to the influence caused by the limited size of rat cartilage. However, the other histological results may demonstrate the correlation between restoring lubrication and attenuating the progression of OA [104].

Nanospheres or Nanoparticles and Nano (Micro) Gels
Nanospheres or nanoparticles and microgels modified with highly hydrated polymers were also developed as biomimetic lubricants recently. The major advantage of these biomimetic lubricants is to combine lubrication with drug-releasing so that they present dual functions. In recent two years, many such biomimetic lubricants emerged. Zhou et al. modified the polystyrene (PS) nanospheres with poly (3-sulfopropyl methacrylate potassium salt) (PSPMA) via atom transfer radical polymerization [106]. Another charged chitosan nanoparticles modified with hydrophilic sulfate groups via ring-open reaction and then loaded aspirin (an anti-inflammatory drug) were prepared as dual-functional lubricants to treat OA [107]. These series of biolubricants, combined with the previous analogous work [118][119][120], greatly decreased the friction coefficient due to the formed thick hydration layer around the nanospheres or nanoparticles. However, the potential of such biolubricants being used to treat OA requires further studies since they mainly focus on the lubrication properties.
From this point of view, as shown in Figure 7, Zhang et al. grafted PMPC onto the surface of mesoporous silica nanospheres (MSN) via photopolymerization to prepare biomimetic lubricated nanospheres (MSNs-NH 2 @PMPC), meanwhile, diclofenac sodium (DS), an anti-inflammatory drug, was encapsulated in MSNs-NH 2 @PMPC to prepare dual-functional (lubrication and drug delivery) biolubricants MSNs-NH 2 @PMPC-DS. They found MSNs-NH 2 @PMPC (10 mg/mL in aqueous) were adsorbed and formed a thin film on the surface of Ti6Al4V, and hence acted as a boundary lubrication layer to significantly reduce the friction coefficient when PE sphere pin sliding against Ti6Al4V lubricated by MSNs-NH 2 @PMPC solution. Further in vivo treatment of OA was evaluated after confirming the biocompatibilities of MSNs-NH 2 @PMPC and MSNs-NH 2 @PMPC-DS. The results indicated that the MSNs-NH 2 @PMPC-DS best inhibited the degradation of cartilage, and promoted the expression of type II collagen, which was attributed to the synergy of lubrication and drug delivery [108]. By replacing the PMPC with 3-dimethyl-[2-(2methylprop-2-enoyloxy)ethyl]azaniumyl]propane-1-sulfonate polymer (pSBMA), the same dual-functional biomimetic lubricants were also prepared [109]. tassium salt) (PSPMA) via atom transfer radical polymerization [107]. Another charged chitosan nanoparticles modified with hydrophilic sulfate groups via ring-open reaction and then loaded aspirin (an anti-inflammatory drug) were prepared as dual-functional lubricants to treat OA [108]. These series of biolubricants, combined with the previous analogous work [119][120][121], greatly decreased the friction coefficient due to the formed thick hydration layer around the nanospheres or nanoparticles. However, the potential of such biolubricants being used to treat OA requires further studies since they mainly focus on the lubrication properties.
From this point of view, as shown in Figure 7, Zhang et al. grafted PMPC onto the surface of mesoporous silica nanospheres (MSN) via photopolymerization to prepare biomimetic lubricated nanospheres (MSNs-NH2@PMPC), meanwhile, diclofenac sodium (DS), an anti-inflammatory drug, was encapsulated in MSNs-NH2@PMPC to prepare dual-functional (lubrication and drug delivery) biolubricants MSNs-NH2@PMPC-DS. They found MSNs-NH2@PMPC (10 mg/mL in aqueous) were adsorbed and formed a thin film on the surface of Ti6Al4V, and hence acted as a boundary lubrication layer to significantly reduce the friction coefficient when PE sphere pin sliding against Ti6Al4V lubricated by MSNs-NH2@PMPC solution. Further in vivo treatment of OA was evaluated after confirming the biocompatibilities of MSNs-NH2@PMPC and MSNs-NH2@PMPC-DS. The results indicated that the MSNs-NH2@PMPC-DS best inhibited the degradation of cartilage, and promoted the expression of type II collagen, which was attributed to the synergy of lubrication and drug delivery [109]. By replacing the PMPC with 3-dimethyl-[2-(2methylprop-2-enoyloxy)ethyl]azaniumyl]propane-1-sulfonate polymer (pSBMA), the same dual-functional biomimetic lubricants were also prepared [110].  Figure 7b, n=5 and the data were presented as mean ± SD, **P < 0.01, MSNs as the control group, # P < 0.05, MSNs-DS as the control group. Reprinted with permission from ref. [109]. Copyright 2020 Elsevier.  Figure 7b, n = 5 and the data were presented as mean ± SD, ** p < 0.01, MSNs as the control group, # p < 0.05, MSNs-DS as the control group. Reprinted with permission from Ref. [108]. Copyright 2020 Elsevier.
Very recently, considering such biolubricants were articularly injected into the joint of OA model rats, a new constructing method was reported in order to enhance the injectability [110]. As shown in Figure 8, Methacrylate gelatin (GelMA) hydrogel microspheres were firstly prepared by microfluidic fabrication, then copolymer polymerized by the monomer methacrylate dopamine (DMA) and MPC was absorbed onto the surface of GelMA to generate biolubricants GelMA@DMA-MPC via the robust adhesiveness of dopamine. Subsequently, DS was encapsulated in GelMA@DMA-MPC (GelMA@DMA-MPC@DS) to enable lubrication and drug-delivery simultaneously. As the core of GelMA@DMA-MPC was a microgel, the injectability was thought to be enhanced. The friction coefficients were most decreased by~29.6% when PTFE sliding against silicon wafer. Additionally, when co-cultured with chondrocytes, it was found GelMA@DMA-MPC@DS contributed to protecting the chondrocytes from degeneration associated with inflammation by increasing the expression of type II collagen and down-regulating the level of MMP13 and ADAMTS5. Relatively excellent lubrication ability and biocompatibility permitted GelMA@DMA-MPC to be further investigated to examine the in vivo therapeutic effects in OA model rats. The in vivo results revealed GelMA@DMA-MPC@DS delayed the degradation of cartilage and showed the best therapeutic effects. Very recently, considering such biolubricants were articularly injected into the join of OA model rats, a new constructing method was reported in order to enhance the inject ability [111]. As shown in Figure 8, Methacrylate gelatin (GelMA) hydrogel microspheres were firstly prepared by microfluidic fabrication, then copolymer polymerized by the monomer methacrylate dopamine (DMA) and MPC was absorbed onto the surface o GelMA to generate biolubricants GelMA@DMA-MPC via the robust adhesiveness of do pamine. Subsequently, DS was encapsulated in GelMA@DMA-MPC (GelMA@DMA MPC@DS) to enable lubrication and drug-delivery simultaneously. As the core o GelMA@DMA-MPC was a microgel, the injectability was thought to be enhanced. The friction coefficients were most decreased by ~29.6% when PTFE sliding against silicon wa fer. Additionally, when co-cultured with chondrocytes, it was found GelMA@DMA MPC@DS contributed to protecting the chondrocytes from degeneration associated with inflammation by increasing the expression of type II collagen and down-regulating the level of MMP13 and ADAMTS5. Relatively excellent lubrication ability and biocompati bility permitted GelMA@DMA-MPC to be further investigated to examine the in vivo therapeutic effects in OA model rats. The in vivo results revealed GelMA@DMA MPC@DS delayed the degradation of cartilage and showed the best therapeutic effects. The plots of friction coefficients versus time using PTFE pin and a silicon wafer as friction pairs; (c) Representative images of histological staining assay showed the disruption of cartilage with different treatments. The black arrow shows the fissures within cartilage, the black triangle indicates tissue cellularity with cloning. Reprinted with permission from ref. [111]. Copyright 2021 KeAi publishing.
One may notice that the lubrication performance of this series of biomimetic lubri cants was evaluated by some specific polymers (PDMS, PTFE, PE) or alloys (Ti6Al4V rather than cartilage though it was emphasized with the potential or aim to restore the lubrication of cartilage to treat OA. However, there is no direct evidence that restoration of cartilage lubrication can reverse the pathologic progression of OA at its early stage (though it is most expected), so drug intervention should be taken into consideration es pecially at the middle or late stage of OA. Particularly, it should be noted that the anima models used in these studies are rats, the anatomic structure of the stifle joint of rats (the size of joint, the thickness of articular cartilage and the mechanical loading) are signifi cantly different from humans, more clinically relevant data are required from the large One may notice that the lubrication performance of this series of biomimetic lubricants was evaluated by some specific polymers (PDMS, PTFE, PE) or alloys (Ti6Al4V) rather than cartilage though it was emphasized with the potential or aim to restore the lubrication of cartilage to treat OA. However, there is no direct evidence that restoration of cartilage lubrication can reverse the pathologic progression of OA at its early stage (though it is most expected), so drug intervention should be taken into consideration especially at the middle or late stage of OA. Particularly, it should be noted that the animal models used in these studies are rats, the anatomic structure of the stifle joint of rats (the size of joint, the thickness of articular cartilage and the mechanical loading) are significantly different from humans, more clinically relevant data are required from the large animal models of OA (including dog, sheep/goat, and horse) to further evaluate the validities and potential therapeutics of such biomimetic lubricants.

Conclusions and Future Perspectives
The synergy between the phosphocholine lipids, HA and lubricin to form a highly hydrated boundary lubrication layer on the outer surface of cartilage is now generally accepted to explain the remarkable lubrication performance of articular cartilage, the integrity of this lubrication layer is vitally critical to maintaining the homeostasis of articular cartilage which is determined by the only embedded cells within it, chondrocytes. Otherwise, the initiation of OA is induced because of the increased friction of cartilage. Inspired by the molecular understanding of cartilage lubrication, the mechanosensitive nature of chondrocytes may give us the confidence to restore the impaired lubrication of articular cartilage at the early stage of OA to prevent the degradation of cartilage and promote the accumulation of the main components of ECM (type II collagen and aggrecan) to regenerate cartilage by regulating the metabolism of chondrocytes. It is, therefore, dual or multi-functional biomimetic lubricants (combined with drugs or inhibitors or gene regulators or autophagy) that may continuously attract attention in the future.

Conflicts of Interest:
The authors declare no conflict of interest.