**1. Introduction**

130 Rheumatoid Arthritis – Etiology, Consequences and Co-Morbidities

Yamanaka, H.; Makino, K., Takizawa, H., Fujimoto, N., Moriya, H. et al. (2000). Expression

*Rheum Dis,* Vol.59, No.6, pp. 455-461, ISSN: 0003-4967

and tissue localization of membrane-types 1, 2, and 3 matrix metralloproteinases in rheumatoid synovium. *Lab Invest,* Vol.80, No.5, pp. 677-687, ISSN: 0023-6837 Yoshihara, Y.; Nakamura, H., Obata, K., Yamada, H., Hayakawa, T., Fujikawa, K. & Okada,

Y. (2000). Matrix metalloproteinases and tissue inhibitors of metalloproteinases in synovial fluids from patients with rheumatoid arthritis or osteoarthritis. *Ann* 

> The dynamic milieu of synovial fluid is of particular interest for biomarker discovery of joint related diseases as it is composed not only of ultra-filtrated proteins originating in serum, but also proteins exclusively expressed and secreted by cells localized within the synovial membrane, fluid or cartilage. Lubricin (proteoglycan 4, prg4) is an abundant mucinous and secretory glycoprotein (~227 to 345 kDa) in synovial fluid (SF) and one of the factors considered responsible for boundary lubrication of diarthrodial joints (Swann et al., 1981; Swann et al., 1985; Jay, 1992). Lubricin is encoded by gene *PRG4* and synthesized in synovial fibroblasts (synoviocytes) and superficial zone chondrocytes. Different transcripts of *PRG4* have been referred to as superficial zone protein (SZP), megakaryocyte stimulating factor (MSF) precursor, camptodactyly arthropathy coxa vara pericarditis (CACP) protein, and hemangiopoietin (HAPO), which has recently been reviewed by Bao et al (Bao et al., 2011). As a primarily lubricating glycoprotein, lubricin has been found in SF, superficial layer of articular cartilage, tendons, and menisci (Schumacher et al., 1994; Schumacher et al., 1999; Rees et al., 2002; Rhee et al., 2005b; Schumacher et al., 2005; Sun et al., 2006). This tissuespecific distribution makes lubricin a potential biomarker during the exacerbation of chronic articular inflammation.

> Human synovial lubricin (1404 amino acids) has a large and central mucin-like domain characterized with 59 imperfect repeating units of EPAPTTPK which is subject to extensive *O*linked glycosylation. The abundance of negatively charged sugars in this domain contributes to the protein's boundary lubrication of the cartilage surface due to strong repulsive hydration forces (Jay, 1992). The mucin domain is flanked by a *C*-terminal hemopexin (PEX)-like domain and two somatomedin B (SMB)-like domains at its *N-*terminus (Flannery et al., 1999; Schumacher et al., 1999; Ikegawa et al., 2000). The two *N*-terminal SMB-like domains have 60% similarity to that of vitronectin, while *C*-terminal PEX-like domain also shows similarity to domains in vitronectin (40-50%) as well as to the matrix metalloproteinase (MMPs) family. Purified serum hemopexin has been showed to interact with hyaluronan, suggesting that the PEX-like domain in lubricin may also medicate the binding of lubricin to hyaluronan at or near cartilage surface (Hrkal et al., 1996). In addition to boundary lubrication, lubricin protects cartilage surfaces from protein deposition and cell adhesion (Rhee et al., 2005b).

<sup>\*</sup> Corresponding author

Glycoproteomics of Lubricin-Implication

degenerative joint diseases.

pannus (Rhee et al., 2005b).

**2.1 Materials and methods** 

fluid.

of Important Biological Glyco- and Peptide-Epitopes in Synovial Fluid 133

cathepsin B, and MMPs (Jones et al., 2003; Elsaid et al., 2005). MMPs are an enzyme family of calcium-dependent zinc-containing endopeptidase which is known to play important roles in tissue remodeling during physiological as well as pathological processes. In cartilage, MMPs are the principal proteases capable of degrading a wide variety of the extracellular matrix components (Nagase & Woessner, 1999). The released fragment of lubricin together with other synovial residual proteins and cartilage matrices floating in synovial fluid may be detected by biochemical or immunochemical assay. The profile of proteins or fragments within SF may represent diagnostic or prognostic biomarker for

Defect of lubricin function leads to CACP syndrome in human, which is a rare and Mendelian genetic arthropathy causing juvenile-onset, inflammatory, precocious joint failure (Marcelino et al., 1999). Although *Prg4*-/- mice did not have noticeably reduced fertility or life span, with aging knockout mice underwent synovial hyperplasia, subintimal fibrosis, proteinaceous deposits on the cartilage surface, irregular cartilage surface and endochondral growth plates, and ultimate invasion of the cartilage surface by synoviocytes reminiscent of human CACP and the cartilage invasion of RA joints by the inflammatory

As all these studies indicate, it is reasonable to speculate that inflammation-induced alterations of both the level, degradation and glycosylation of lubricin that occur in the joints of patients with RA and OA may accelerate the destruction of joints and exacerbate the disease. Monitoring new glyco-epitopes and/or proteolytic fragments of lubricin may serve as a potential biomarker for advanced diagnosis of early stage. To perform this, it is necessary to fully characterize the lubricin molecule by glycoproteomics. In this study, we used various biotinylated lectins or anti-carbohydrate antibodies together with MS to characterize glyco-epitopes on lubricin. The results confirm that lubricin contains immunologically important *O*-linked oligosaccharide epitopes that are capable of binding selectins and galectins. Proteomic analysis indicated that not all repeat units are occupied with *O*-linked oligosaccharides and also revealed several fragments of lubricin in synovial

It is known that joint damage may progress despite decreased inflammatory activity and erosions may develop in patients with few signs of inflammation by conventional assessments (Flato et al., 2003). Therefore, predicting the progression and consequences of inflammatory pathology are essential for optimal clinical management. The ideal biomarker of persistent inflammation in arthritis should fulfill a number of criteria including: detectable levels in early disease, expression which coincides with each inflammatory episode and expression that is restricted to the inflamed joint. The identification of differentially expressed proteins contributes to understanding the molecular factors of the disease better and paves the way for new diagnostic and prognostic markers, and eventually

Synovial fluid samples from RA patients were collected during therapeutic joint aspiration at the Rheumatology Clinic, Sahlgrenska University Hospital (Gothenburg, Sweden). All patients gave informed consent and the procedure was approved by the Ethics Committee

to novel targets in the development of therapeutic strategies.

**2.1.1 Enrichments of lubricin from synovial fluid** 

**2. Glycoproteomic characterization of synovial lubricin** 

During inflammation, glycosylation properties such as sialylation, sulfation, and fucosylation, are regulated to manipulate cell adhesion, differentiation, maturation, and activation in the case of immune cells. Bone and cartilage cells like osteoblasts, synovial fibroblasts and chondrocytes have been shown to possess the enzymes necessary for the synthesis of *N*- and *O*-glycans of glycoproteins, among which some activities are regulated by cytokines found in inflamed joints (Brockhausen & Anastassiades, 2008). Based on the established biosynthetic pathways, it was reported that human joint glycoproteins mainly had complex bi-antennary *N*-glycans and *O*-glycans with core 1 and the branched core 2 structures (Brockhausen & Anastassiades, 2008). In our previous study, the *O*-linked oligosaccharides of lubricin were characterized (Estrella et al., 2010). On lubricin, core 1 *O*linked oligosaccharides are the predominant structures. Removal of sialic acid and core 1 oligosaccharides caused loss of boundary lubrication (Jay, 1992; Jay et al., 2001), showing that these structural elements are sufficient for providing lubricating property of lubricin. With aid of liquid-chromatography-mass spectrometry (LC-MS), small proportion of sialylated core 2 oligosaccharides were also found on lubricin both with and without sulfation. This indicates that lubricin glycosylation also have other task requiring complex *O*-glycosylation. In summary, both core 1 and core 2 glyco-epitopes on lubricin have the potential of excessive interactions with glyco-binding proteins, such as selectins and galectins, to facilitate inflammation.

Degenerative joint disease and joint injury are associated with increased turnover of articular cartilage proteins, inflammation, and alterations to other joint tissue proteins (Goldring & Goldring, 2007). So far, several synovial joint-specific biomarkers have been identified in adults, such as calgranulin A, B, and C (Sinz et al., 2002; Liao et al., 2004), fibrinogen β-chain, fructose bisphosphonate aldolase A, alpha-enolase (Tilleman et al., 2005), tenascin-C (Hasegawa et al., 2004), serum amyloid A (SAA), and broader inflammatory biomarkers, such as C-reactive protein (Kuhn et al., 2004) and haptoglobin (Sinz et al., 2002; Kantor et al., 2004). Lubricin as one important synovial component to monitor the state of a joint is less investigated, despite its highly relevant function as a biolubricant. Because of the size and posttranslational modifications of lubricin, it is not readily detectable by traditional two-dimensional electrophoresis (2-DE). However, a decreased expression of lubricin together with increased degradation of lubricin have been associated with more aggressive rheumatoid arthritis (RA) and osteoarthritis (OA). This strongly indicates that lubricin may be a good joint-specific biomarker. For example, *in vitro* boundary lubricating test indicated that SF from chronic inflammatory RA patients had decreased lubricating ability in comparison with SF from acute knee joint synovitis patients and cartilage transplant donors (Elsaid et al., 2005). According to the expression level of lubricin in synovium, RA patients could be classified into two groups, of where lower expression level of lubricin was associated with a more aggressive disease stage (Ungethuem et al., 2010). As for OA, animal models of OA also feature reduced levels of lubricin, particularly in the early stage of the disorder (Young et al., 2006; Elsaid et al., 2007). Also, when applied exogenous lubricin in an animal model of OA, it appears to be chondroprotective and to reduce structural damage (Flannery et al., 2009; Teeple et al., 2011). It has been demonstrated that lubricin expression is down-regulated by proinflammatory cytokines (e.g., interleukin (IL)-1β, tumor necrosis factor α (TNFα), and IL-6) (Flannery et al., 1999; Rhee et al., 2005b; Young et al., 2006; Schmidt et al., 2008). Decreased synovial lubricin level may be caused by degradation with neutrophil elastase,

During inflammation, glycosylation properties such as sialylation, sulfation, and fucosylation, are regulated to manipulate cell adhesion, differentiation, maturation, and activation in the case of immune cells. Bone and cartilage cells like osteoblasts, synovial fibroblasts and chondrocytes have been shown to possess the enzymes necessary for the synthesis of *N*- and *O*-glycans of glycoproteins, among which some activities are regulated by cytokines found in inflamed joints (Brockhausen & Anastassiades, 2008). Based on the established biosynthetic pathways, it was reported that human joint glycoproteins mainly had complex bi-antennary *N*-glycans and *O*-glycans with core 1 and the branched core 2 structures (Brockhausen & Anastassiades, 2008). In our previous study, the *O*-linked oligosaccharides of lubricin were characterized (Estrella et al., 2010). On lubricin, core 1 *O*linked oligosaccharides are the predominant structures. Removal of sialic acid and core 1 oligosaccharides caused loss of boundary lubrication (Jay, 1992; Jay et al., 2001), showing that these structural elements are sufficient for providing lubricating property of lubricin. With aid of liquid-chromatography-mass spectrometry (LC-MS), small proportion of sialylated core 2 oligosaccharides were also found on lubricin both with and without sulfation. This indicates that lubricin glycosylation also have other task requiring complex *O*-glycosylation. In summary, both core 1 and core 2 glyco-epitopes on lubricin have the potential of excessive interactions with glyco-binding proteins, such as selectins and

Degenerative joint disease and joint injury are associated with increased turnover of articular cartilage proteins, inflammation, and alterations to other joint tissue proteins (Goldring & Goldring, 2007). So far, several synovial joint-specific biomarkers have been identified in adults, such as calgranulin A, B, and C (Sinz et al., 2002; Liao et al., 2004), fibrinogen β-chain, fructose bisphosphonate aldolase A, alpha-enolase (Tilleman et al., 2005), tenascin-C (Hasegawa et al., 2004), serum amyloid A (SAA), and broader inflammatory biomarkers, such as C-reactive protein (Kuhn et al., 2004) and haptoglobin (Sinz et al., 2002; Kantor et al., 2004). Lubricin as one important synovial component to monitor the state of a joint is less investigated, despite its highly relevant function as a biolubricant. Because of the size and posttranslational modifications of lubricin, it is not readily detectable by traditional two-dimensional electrophoresis (2-DE). However, a decreased expression of lubricin together with increased degradation of lubricin have been associated with more aggressive rheumatoid arthritis (RA) and osteoarthritis (OA). This strongly indicates that lubricin may be a good joint-specific biomarker. For example, *in vitro* boundary lubricating test indicated that SF from chronic inflammatory RA patients had decreased lubricating ability in comparison with SF from acute knee joint synovitis patients and cartilage transplant donors (Elsaid et al., 2005). According to the expression level of lubricin in synovium, RA patients could be classified into two groups, of where lower expression level of lubricin was associated with a more aggressive disease stage (Ungethuem et al., 2010). As for OA, animal models of OA also feature reduced levels of lubricin, particularly in the early stage of the disorder (Young et al., 2006; Elsaid et al., 2007). Also, when applied exogenous lubricin in an animal model of OA, it appears to be chondroprotective and to reduce structural damage (Flannery et al., 2009; Teeple et al., 2011). It has been demonstrated that lubricin expression is down-regulated by proinflammatory cytokines (e.g., interleukin (IL)-1β, tumor necrosis factor α (TNFα), and IL-6) (Flannery et al., 1999; Rhee et al., 2005b; Young et al., 2006; Schmidt et al., 2008). Decreased synovial lubricin level may be caused by degradation with neutrophil elastase,

galectins, to facilitate inflammation.

cathepsin B, and MMPs (Jones et al., 2003; Elsaid et al., 2005). MMPs are an enzyme family of calcium-dependent zinc-containing endopeptidase which is known to play important roles in tissue remodeling during physiological as well as pathological processes. In cartilage, MMPs are the principal proteases capable of degrading a wide variety of the extracellular matrix components (Nagase & Woessner, 1999). The released fragment of lubricin together with other synovial residual proteins and cartilage matrices floating in synovial fluid may be detected by biochemical or immunochemical assay. The profile of proteins or fragments within SF may represent diagnostic or prognostic biomarker for degenerative joint diseases.

Defect of lubricin function leads to CACP syndrome in human, which is a rare and Mendelian genetic arthropathy causing juvenile-onset, inflammatory, precocious joint failure (Marcelino et al., 1999). Although *Prg4*-/- mice did not have noticeably reduced fertility or life span, with aging knockout mice underwent synovial hyperplasia, subintimal fibrosis, proteinaceous deposits on the cartilage surface, irregular cartilage surface and endochondral growth plates, and ultimate invasion of the cartilage surface by synoviocytes reminiscent of human CACP and the cartilage invasion of RA joints by the inflammatory pannus (Rhee et al., 2005b).

As all these studies indicate, it is reasonable to speculate that inflammation-induced alterations of both the level, degradation and glycosylation of lubricin that occur in the joints of patients with RA and OA may accelerate the destruction of joints and exacerbate the disease. Monitoring new glyco-epitopes and/or proteolytic fragments of lubricin may serve as a potential biomarker for advanced diagnosis of early stage. To perform this, it is necessary to fully characterize the lubricin molecule by glycoproteomics. In this study, we used various biotinylated lectins or anti-carbohydrate antibodies together with MS to characterize glyco-epitopes on lubricin. The results confirm that lubricin contains immunologically important *O*-linked oligosaccharide epitopes that are capable of binding selectins and galectins. Proteomic analysis indicated that not all repeat units are occupied with *O*-linked oligosaccharides and also revealed several fragments of lubricin in synovial fluid.

It is known that joint damage may progress despite decreased inflammatory activity and erosions may develop in patients with few signs of inflammation by conventional assessments (Flato et al., 2003). Therefore, predicting the progression and consequences of inflammatory pathology are essential for optimal clinical management. The ideal biomarker of persistent inflammation in arthritis should fulfill a number of criteria including: detectable levels in early disease, expression which coincides with each inflammatory episode and expression that is restricted to the inflamed joint. The identification of differentially expressed proteins contributes to understanding the molecular factors of the disease better and paves the way for new diagnostic and prognostic markers, and eventually to novel targets in the development of therapeutic strategies.
