Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Osteoarthritis Cartilage. Author manuscript; available in PMC 2011 Feb 1.
Published in final edited form as:
PMCID: PMC2818349

Phenotype-related differential α-2,6- or α-2,3-sialylation of glycoprotein N-glycans in human chondrocytes

S Toegel, PhD, MPharmS,1,2,* M Pabst, MS,3 SQ Wu, MPharmS,4,5 J Grass, MS,3 MB Goldring, PhD,2 C Chiari, MD,6 A Kolb, MD,6 F Altmann, PhD,3 H Viernstein, PhD, MPharmS,4 and FM Unger, PhD4



Sialic acids frequently occur at the terminal positions of glycoprotein N-glycans present at chondrocyte surfaces or in the cartilage matrix. Sialic acids are transferred to glycoproteins in either α-2,3 or α-2,6 linkage by specific sialyltransferases (SiaTs) and can potentially affect cell functions and cell-matrix interactions. The present study aimed to assess the relationship between the expression of the human chondrocyte phenotype and the sialylation of chondrocyte glycoprotein N-glycans.


The transcription of 5 SiaT was quantified using real-time RT-PCR assays. N-glycan analysis was performed using LC-ESI-MS. Primary human chondrocytes were cultured in monolayer or alginate beads and compared to the chondrocyte cell lines C-28/I2 and SW1353. In addition, effects of interleukin-1β or tumor necrosis factor-α on primary cells were assessed.


Primary human chondrocytes predominantly express α-2,6-specific SiaTs and accordingly, α-2,6-linked sialic acid residues in glycoprotein N-glycans. In contrast, the preponderance of α-2,3-linked sialyl residues and, correspondingly, reduced levels of α-2,6-specific SiaTs are associated with the altered chondrocyte phenotype of C-28/I2 and SW1353 cells. Importantly, a considerable shift towards α-2,3-linked sialic acids and α-2,3-specific SiaT mRNA levels occurred in primary chondrocytes treated with IL-1β or TNF-α.


The expression of the differentiated chondrocyte phenotype is linked to the ratio of α-2,6- to α-2,3-linked sialic acids in chondrocyte glycoprotein N-glycans. A shift towards altered sialylation might contribute to impaired cell-matrix interactions in disease conditions.

Keywords: chondrocytes, extracellular matrix, cell - matrix interaction, glycoproteins, sialic acids, sialyltransferases, differentiation, phenotype


The extracellular matrix (ECM) of cartilage consists of numerous glycosylated proteins, which contribute to the maintenance of its specific functions [1]. Beside structural proteoglycans such as aggrecan, and adhesive proteins such as chondroadherin, glycosylated cell surface receptors, including CD44 and integrins, play critical roles in the mediation of chondrocyte-matrix interactions [13]. Protein glycosylation, i.e. the attachment of oligosaccharide chains to asparagine or serine/threonine residues, represents a common co-translational and post-translational modification of proteins, conferring protection against proteolytic degradation or controlling protein folding. It is known that carbohydrate chains vary according to tissue and cell type and are subject to change during development and oncogenesis [46]. In addition, various factors such as cytokines or differentiation factors can regulate cellular glycosylation and the expression of specific glycosyltransferases [711]. Thus, changes in the cellular glycophenotype can potentially affect cell functions such as cell adhesion, cell surface receptor activity, and the onset of apoptosis [12].

Sialic acids (Neu5Ac in humans) are negatively charged sugars typically found at the terminal positions of N- and O-linked oligosaccharides attached to cell surfaces or secreted glycoproteins. As a result of their non-reducing terminal position, sialic acids are involved in highly specific recognition phenomena, as well as in invasive properties of cancer cells [1316]. The transfer of sialic acids to penultimate galactose or N-acetylgalactosamine groups of glycoproteins is mediated by sialyltransferases (SiaTs). The diverse enzymes of the SiaT family share the cytidine monophosphate-sialic acid as donor substrate, but are distinguished by the oligosaccharide acceptor on which they act and by the linkage type they generate (e.g., α-2,3 or α-2,6 linkage in case of N-linked glycans in glycoproteins) [16]. The expression patterns of SiaTs show remarkable tissue specificity and appear to be regulated at the transcriptional level. In particular, the altered expression of α-2,3- or α-2,6-SiaTs and subsequent effects on cell surface sialylation have been demonstrated to influence osteoclastogenesis, endocytosis in dendritic cells, and adhesion properties of human colon cancer cells [1719].

Regarding inflammatory and degenerative joint diseases, recent studies have suggested a relationship between altered glycosylation and disease conditions. It is known that altered N-glycans of serum immunoglobulin G molecules contribute to the pathophysiology of rheumatoid arthritis [20]. In addition, initial evidence for altered N-glycosylation of cartilage tissue has been presented using an animal model of osteoarthritis [21]. Despite the significance of glycoproteins for ECM assembly in cartilage tissue, however, little is known about the regulation of the chondrocyte glycophenotype. Recently, we found that differences in the molecular phenotype between primary human chondrocytes and chondrocyte cell lines are reflected by specific cellular lectin-binding patterns [22,23]. In a further study, Yang and coworkers have shown that cytokine treatment of human and bovine chondrocytes results in specific glycosylation changes related to apoptosis and altered cell proliferation [12].

Owing to the potential implications of the chondrocyte glycophenotype in the pathophysiology of joint diseases such as osteoarthritis, we set out to examine the sialylation processes in human chondrocytes using a combination of gene expression and N-glycan structural analyses. We compared the sialylation pattern of primary human chondrocytes with those of immortalized chondrocytes C-28/I2 and chondrosarcoma cells SW1353, and investigated the effects of interleukin-1β (IL-1β) and tumour necrosis factor-alpha (TNF-α) on primary chondrocyte sialylation. Our findings provide new information on N-glycan sialylation in human chondrocytes and demonstrate that the status of the chondrocyte phenotype is critically linked to the ratio of α-2,6- to α-2,3- linked sialic acids of chondrocyte glycoprotein N-glycans.


Human chondrocyte cultures

Human articular cartilage was obtained during total knee replacement surgeries in patients with osteoarthritis with informed consent and in accordance with the terms of the ethics committee of the Medical University Vienna (EK-Nr.: 081/2005). Primary chondrocytes were isolated according to protocols published previously [22], seeded as high density monolayers (105 cells/cm2), and grown in an atmosphere of humidified 5% CO2/95% air at 37°C using Dulbecco's Modified Eagle Medium supplemented with 10% fetal calf serum (Biochrom AG, Germany), 2 µl/ml gentamycin, and 50 µg/ml ascorbate as culture medium (complete medium). Only freshly isolated cells were plated for all assays throughout the study. To induce inflammatory conditions, primary chondrocytes were treated with 10 ng/ml IL-1β or 40 ng/ml TNF-α for the time periods indicated. Three-dimensional (3D) chondrocyte cultures were established in alginate beads according to standard procedures and cultured for 3 weeks as described [24]. Immortalized chondrocytes C-28/I2, and chondrosarcoma cells SW1353, were seeded at 9.3×103 cells/cm2 and maintained in complete medium.

Quantitative real-time PCR

Chondrocytes were grown in duplicate wells in 12-well tissue culture plates (Iwaki, Japan) to 90% confluence prior to treatment with cytokines for 24 h. Total RNA was extracted using the NucleoSpin RNA II Kit (Macherey-Nagel, Germany). Each sample was run on the Agilent 2100 Bioanalyzer Nano LabChip for quality control and quantification of total RNA prior to reverse transcription into cDNA using the high capacity cDNA reverse transcription kit (Applied Biosystems, Austria).

SYBR-green based qPCR assays for the sialyltransferase transcripts ST6Gal1 (NM_003032), ST6Gal2 (NM_032528), ST3Gal3 (NM_174963), ST3Gal4 (NM_006278), and ST3Gal6 (NM_006100) were established. Amplification efficiencies of all primers were between 90 and 105% as evaluated using dilution series of cDNA prepared from chondrocyte mRNA. Primers and results of melting curve analyses are shown in Supplementary Table 1 and Supplementary Figure 1, respectively. qPCR reactions for transcripts from collagen type I (COL1), collagen type II (COL2) and aggrecan (AGC) were performed according to protocols published recently [25]. mRNA expression levels were calculated as relative copy numbers considering actual amplification efficiencies and with respect to that of glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) set at 1000. The regulation of a target gene by cytokines was calculated as quantities relative to the untreated control group using the MxPro real-time QPCR software, considering both amplification efficiencies and normalization to GAPDH.

Protein isolation and N-glycan preparation

Chondrocytes were seeded in 25 cm2 flasks (Greiner bio-one, Austria) and allowed to settle for 48 h. Then, the culture medium was changed and cells were maintained in the presence of 10 ng/ml IL-1β or 40 ng/ml TNF-α for 5 days. Cells were carefully washed 5 times with PBS buffer and scraped in the presence of 200 µl lysis buffer (containing protease inhibitor). After centrifugation for 5 min at 12,000 rpm, the pellet was washed once with PBS buffer and centrifuged again. The combined supernatants were mixed with three times the volume of acetone (−20°C for 30 min) to precipitate proteins. Further protein purification and glycan isolation was performed using the ‘in-gel release method‘ as described by Rendic et al [26]. Briefly, the precipitate was dissolved in SDS-PAGE sample buffer and subjected to SDS-PAGE, but only to the point when the unseparated protein zone had just entered the separation gel. Free N-glycans were isolated from this band by trypsin and PNGase F digestion. The released asparagine-linked glycans were reduced with 50 µL of 1% NaBH4 at room temperature for 4 h. To remove excess salt, the sample was further purified using a 10 mg HyperSep Hypercarb SPE cartridge (Thermo Scientific, USA), according to previous protocols [27].

Quantification of oligosaccharides using LC-ESI-MS

Recent work on fibrin N-glycans and immunoglobulin glycosylation revealed the ability of porous graphitic carbon (PGC) phases to separate isomeric glycans with a subsequent sensitive detection of both neutral and sialylated glycans by using ESI-MS [28,29]. The recently established retention time library of mammalian, diantennary glycans varying in the linkage of sialic acid was applied for the structural assignment of the peaks. LC-ESI-MS, using graphitized carbon as the separation phase (PGCC-ESI-MS), outperforms other glycan analysis methods including MALDI-TOF or NMR spectroscopy in terms of minimal sample consumption, minimal sample preparation, and the reliable qualitative and quantitative recording of neutral, mono- and di-sialylated structures with concomitant identification of isomeric structures [2628].

Analysis of borohydride-reduced oligosaccharides (from an aliquot equivalent to 2.5×105 cells) by positive-ion LC-ESI-MS was performed with a 100 × 0.32 mm PGC column (Thermo) and a flow rate of 5 µl/min maintained with a Dionex Ultimate 3000 cap flow system, as described recently [28,29]. Mass analysis was done with a Waters Q-TOF Ultima Global mass spectrometer with standard ESI-source and MassLynx V4.0 SP4 software for evaluation and quantification of obtained peak areas. Individual analyses were averaged, with a mean deviation for the relative glycan quantities of approximately 0.05 units.

Cell morphology

Primary chondrocytes were seeded on glass cover slips and cultured for 48 h in complete medium. Then, cells were exposed to 10 ng/ml IL-1β or 40 ng/ml TNF-α for up to 5 days. Using conventional light microscopy, cell morphological alterations were observed with regard to the formation of a spindle-shaped cell configuration, which is associated with a dedifferentiated fibroblast-like phenotype.

Cell proliferation

Primary chondrocytes of 2 donors were seeded into 96-well microplates (Iwaki, Japan) and cultured for 48 h. Then, cells were treated in quadruplet with 10 ng/ml IL-1β or 40 ng/ml TNF-α for 24 h. The total number of cells was determined using the CyQuant assay kit (Molecular Probes, USA) pursuing a modified protocol [30]. Moreover, the rate of proliferation of cells was assessed by incorporation of 2-bromodeoxyuridine into the DNA of dividing cells using the colorimetric BrdU assay (Roche, Germany), according to the manufacturer’s instructions.

Sulfated glycosaminoglycan (sGAG) assay

Primary chondrocytes were seeded in triplicate into 6-well tissue culture plates (Iwaki, Japan) and cultured for 48 h prior to treatment with 40 ng/ml TNF-α for 5 days. An improved 1,9-dimethyl-methylene blue (DMMB) colorimetric assay was performed according to Barbosa et al. [31] to assess both cell-associated sGAG and sGAG released into the culture medium. Absorbance of the samples was measured at 656 nm and related to equivalent cell numbers, which were determined using the CyQuant assay kit. The experiment was repeated twice using cells from different donors.


Data were exported to the GraphPad Prism statistics software package (GraphPad Prism Software, USA). The Gaussian distribution of the data was verified using the Kolmogorov-Smirnov test. Statistics were performed using one-way ANOVA with post-hoc Tukey tests, cross-comparing all study groups (95% confidence interval) consisting of independent mean values. P-values <0.05 were considered significant.


Expression profiles of SiaT genes in human chondrocyte models

The mRNA levels of 5 SiaTs were determined with respect to GAPDH in human chondrocytes using qPCR. Table 1 shows the SiaT expression levels in primary chondrocytes (PC) cultured as monolayers (ML) or in alginate beads (3D), as well as in SW1353 and C-28/I2 cell lines. In general, SiaTs were expressed at low levels in all cell models. In monolayers of primary chondrocytes (n=10 donors), ST6Gal1 showed the highest abundance of all evaluated SiaTs. In comparison to ST6Gal1, ST6Gal2 was expressed at more than 400-fold lower levels. In addition, the mRNA levels of the α-2,3-SiaTs (ST3Gal3, ST3Gal4, and ST3Gal6) were all low compared to ST6Gal1. Among the α-2,3-SiaTs, ST3Gal4 was the gene with the highest expression level. All in all, the SiaT expression profile of primary chondrocytes predicts a sialylation pattern in which α-2,6-linked sialylation predominates over α-2,3-sialylation. When cultured in alginate beads, primary chondrocytes expressed comparable levels of α-2,6-SiaTs and slightly reduced levels of ST3Gal3 and ST3Gal4 (1.6-fold and 1.9-fold, respectively). When comparing the chondrosarcoma cell line SW1353 with primary chondrocytes, we observed remarkably diminished (2.5-fold) ST6Gal1 expression in SW1353 cells, whereas mRNA levels of ST3Gal4 were slightly increased. The SiaT profile of C-28/I2 chondrocytes was characterized by almost completely suppressed ST6Gal1 mRNA levels, whereas the α-2,3-SiaT profile only showed minor alterations in comparison to primary chondrocytes. Taken together, primary human chondrocytes are characterized by a predominant expression of α-2,6-SiaT, predicting the preponderant presence of α-2,6-linked terminal sialic acid residues on primary chondrocyte glycoprotein N-glycans. In contrast, immortalized and chondrosarcoma cell models feature a SiaT expression profile shifted towards α-2,3-SiaTs.

Table 1
mRNA expression of N-glycan related sialyltransferases SiaTs)

Sialylation of chondrocyte glycoprotein N-glycans in human chondrocyte models

Sialylation patterns and linkage-types of chondrocyte glycoprotein N-glycans were determined using LC-ESI-MS. Figure 1A shows the structure of the diantennary H5N4F glycan and its possible sialylated isomers analyzed in this study. Both the abbreviated composition and the full structural description according to the proglycan system are presented [29,32]. These glycans were selected for quantification since they represent a common modification of cellular proteins and are among the most abundant glycan structures in the human chondrocyte glycome (own unpublished result). Moreover, they are practically absent in fetal calf serum-containing cell culture medium minimizing the risk of medium-dependent glycan structure contamination.

Figure 1
Analysis of sialylated and free H5N4F1 glycans in human chondrocytes

Using LC-ESI-MS online analysis via PGC column, we quantified free H5N4F1 structures as well as H5N4F1 glycans carrying one or two sialic acid residues in α-2,6- or/and α-2,3-linkage and compared them across various primary chondrocyte models. As shown in Figure 1B, the proportional distribution of the different H5N4F1 glycans was strongly dependent on cell type and culture conditions. In primary chondrocytes cultured as monolayers, the majority of diantennary glycans carried one α-2,6-linked sialic acid (45.4%), whereas only 11.3% were decorated by one α-2,3-linked sialic acid residue. Among the S2H5N4F1 glycans, most structures (21.2%) carried both one α-2,6-linked and one α-2,3-linked sialic acid. In comparison to this mixed type, S2(α-2,6)H5N4F1 and S2(α-2,3)H5N4F1 structures were generated at 3-fold lower levels. In 3D-culture systems, primary chondrocytes expressed a sialylation pattern characterized by increased levels of H5N4F1 glycans carrying one or two α-2,6-linked sialic acids. Concomitantly, those structures decorated with one or two α-2,3-linked sialic acids were produced at reduced levels. Of note, the sialylation pattern of N-glycans of SW1353 chondrosarcoma cells showed an inverse trend towards significantly reduced S1(α-2,6)H5N4F1 and S2(α-2,6)H5N4F1 structures and increased S1(α-2,3)H5N4F1 and S2(α-2,3)H5N4F1 glycans. This trend was even intensified in the C-28/I2 cell line, where α-2,6-linked sialylation was practically absent, while the majority of H5N4F1 glycans (87.8%) carried terminal sialic acids in α-2,3-linkage. In summary, sialylation patterns of chondrocyte N-glycans showed alteration towards increased α-2,6-/α-2,3-sialylation ratios in 3D-culture and decreased α-2,6-/α-2,3-sialylation ratios in SW1353 and C-28/I2 cells.

Inflammatory cytokines induce chondrocyte dedifferentiation and a shift of α-2,6/α-2,3-sialylation

Figure 2 confirms that primary chondrocytes undergo dedifferentiation under treatment with pro-inflammatory cytokines IL-1β and TNF-α [33]. Morphological changes of chondrocytes towards a spindle-shaped fibroblast-like phenotype were observed after exposure to cytokines (Figure 2A). Moreover, both IL-1β and TNF-α reduced chondrocyte proliferation (Figure 2B) and expression of chondrocyte marker genes AGC and COL2 (Figure 2D), resulting in decreased COL2/COL1 ratios [34]. In addition, TNF-α decreased the production of both cell-associated and secreted sGAG, further supporting cellular dedifferentiation (Figure 2C).

Figure 2
Phenotypic changes of primary human chondrocytes under exposure to inflammatory cytokines

To test whether the linkage-type of chondrocyte N-glycans was altered under inflammatory conditions, SiaT mRNA expression and N-glycan sialylation were determined in primary chondrocytes (n=6 donors) after treatment with pro-inflammatory cytokines. Figure 3 A and B show the log-fold change of SiaT mRNA levels following IL-1β and TNF-α treatment, respectively. ST6Gal1 mRNA levels were markedly reduced under the influence of the cytokines (4.7-fold in case of IL-1β and 3.3-fold in case of TNF-α), whereas the expression of ST3Gal4 was enhanced under the same conditions (2.8-fold and 3.2-fold, respectively). ST6Gal2, ST3Gal3, and ST3Gal6 were also subject to reduction in the presence of IL-1β or TNF-α. Given the different steady-state expression levels of the single SiaTs (Table 1), the contribution of each gene to the entire SiaT pool is displayed graphically in Figure 3C. It is obvious that the α-2,3-SiaTs predominate after exposure of chondrocytes to IL-1β or TNF-α. In general, diminished ST6Gal1 and increased ST3Gal4 levels appear to be responsible for this shift towards reduced α-2,6-/α-2,3-SiaT mRNA ratios under inflammatory conditions.

Figure 3
mRNA levels of N-glycan related sialyltransferases in human chondrocytes exposed to pro-inflammatory cytokines

Analysis of N-glycan sialylation using LC-ESI-MS confirmed the predicted differential expression of α-2,6- and α-2,3-sialic acid residues after cytokine treatment. Interestingly, the described shift in SiaT mRNA levels was not immediately followed by a preponderance of α-2,3-linked sialyl residues in the total of glycoprotein glycans, since no significant alteration was detected by LC-ESI-MS after 24 h of cytokine treatment (data not shown). However, following a 5-day exposure of the cells to TNF-α or IL-1β, a corresponding shift of α-2,6-towards α-2,3-linked sialylation was observed. Figure 4 presents detailed information on the alteration of sialylated H5N4F1 structures in chondrocytes from 3 donors, indicating that both S2(α-2,6)H5N4F1 and S1(α-2,6)H5N4F1 were significantly decreased, whereas those structures containing α-2,3-linked sialic acids (S1(α-2,3)H5N4F1, S2(α-2,6/α-2,3)H5N4F1, S2(α-2,3)H5N4F1) were enhanced by both cytokines. The most striking differences were observed in shift of S1(α-2,6)H5N4F1 and S2(α-2,3)H5N4F1 glycans under TNF-α exposure, resulting in a 0.6-fold decrease and 2.5-fold increase, respectively. A representative LC-ESI-MS chromatogram of S2H5N4F1 glycans under the influence of TNF-α or IL-1β is provided as Supplementary Figure 2 and illustrates the underlying analytic principle, used to compare untreated primary chondrocytes with cytokine-treated cells. Taken together, alterations of the chondrocyte phenotype induced by TNF-α or IL-1β were accompanied by decreasing α-2,6-/α-2,3-SiaT mRNA ratios and corresponding shifts of sialylated N-glycans towards α-2,3-sialylated H5N4F1 structures.

Figure 4
Regulation of H5N4F1 glycan sialylation under IL-1β and TNF-α exposure


Sialylation of cell surface glycoproteins plays a key role in many biological processes, including cell-matrix interactions. Although disturbed cell-matrix interactions are of fundamental importance in the onset of degenerative joint diseases including osteoarthritis [1,35,36], little is known about the sialylation patterns in human chondrocytes and their involvement, if any, in cartilage breakdown processes. This gap in our knowledge may result, at least in part, from the methodological challenges due to low SiaT transcription levels and linkage-type specific glycoprotein N-glycan analysis in human chondrocytes. Up to the present, evidence for the sialylation of chondrocytes has been generated using lectin histochemistry [37], lectin binding studies [22], enzyme-linked lectin assays [12,38], and enzyme activity assays [12]. In this study, we have employed an advanced approach to characterize the chondrocyte glycophenotype. Linking together qPCR and LC-ESI-MS techniques, we correlated quantitatively the gene expression of α-2,6- and α-2,3-specific SiaTs with the presence of α-2,6- and α-2,3-linked sialyl groups in chondrocyte glycoprotein N-glycans.

We here present 2 lines of evidence indicating that the chondrocyte phenotype is linked to specific sialylation patterns of chondrocyte glycoprotein N-glycans. Firstly, the sialylation of various in vitro chondrocyte models with well known phenotypic differences was investigated. Secondly, the sialylation of primary human chondrocytes was evaluated after treatment with IL-1β or TNF-α, two cytokines with pathophysiological significance in inflammatory joint diseases.

Despite the value of chondrocytic cell lines, such as the C-28/I2 and SW1353 cell lines used here, that are used frequently as surrogates for primary human chondrocytes in vitro, they differ markedly from primary chondrocytes in their molecular phenotype [22,39,40]. For instance, the reduced expression of chondrocyte marker genes such as AGC or COL2, as well as altered reactivity to IL-1β, may be related to their high state of proliferation and continuous synthesis of proteins associated with log-phase growth. The present study clearly demonstrates that the presence of α-2,6-linked sialyl residues constitutes a typical feature of the differentiated phenotype of primary human chondrocyte cultures, whereas the abundance of α-2,3-linked sialyl residues is associated with the altered phenotype of chondrocytic cell lines. This result further supports our recent observation that immortalized cell lines are characterized by a reduced binding of Sambucus nigra agglutinin (SNA) [22]. Importantly, differences in N-glycan sialylation among the chondrocyte cell models were in correlation with differences in SiaT expression levels, indicating a crucial role of ST6Gal1 transcription for the phenotype-specific sialylation of chondrocyte glycoproteins. Our finding on the preponderant α-2,6-sialylation of human chondrocytes is in agreement with a previous lectin histochemical study of human knee joints reporting that chondrocytes in zones I – IV of normal hyaline cartilage were stained to a higher degree with SNA lectin than with Maackia amurensis agglutinin [37]. Interestingly, this feature distinguishes articular chondrocytes from their progenitor cells, the mesenchymal stem cells (MSC), which were previously shown to carry sialic acids predominantly in α-2,3-linkage [41], corresponding to the phenotype of the proliferating chondrocyte cell lines studied here .

It is tempting to speculate about the physiological reasons underlying the predominant α-2,6-sialylation of chondrocyte glycoprotein N-glycans. Homeostasis in cartilage tissue is maintained by chondrocyte–matrix interactions and the signals that chondrocytes receive in contact with matrix components. The glycoprotein CD44 is the major receptor for the glycosaminoglycan hyaluronan (HA), anchoring the proteoglycan network of cartilage to the surface of chondrocytes. Detailed studies have shown that at least two of the five N-glycosylation sites within the HA binding domain of CD44 can modify binding activity and that both carry terminal α-2,3-linked sialic acids [42,43]. Furthermore, modification of CD44 glycosylation by inhibition of N-glycan biosynthesis or enzymatic cleavage of N-glycans or terminal sialic acids can trigger the activation of CD44-mediated HA binding in various cells [4346]. Therefore, the presence of terminal α-2,3-linked sialic acid on N-glycans of the CD44 receptor inhibits the binding to HA [46]. Given the observation that α-2,6-SiaTs and α-2,3-SiaTs compete for the same limited pool of N-glycan substrates [47,48], it is conceivable that the predominant expression of α-2,6-SiaTs in human chondrocytes facilitates the activation of CD44 receptors, which in fact is required for a functional matrix assembly of cartilage. In contrast, the dominant α-2,3-sialylation of MSC [41] may be a physiological requirement for migration and homing of MSC to sites of condensation and cartilage formation in the embryo. A similar physiological mechanism has been proposed recently for hematopoietic stem and progenitor cells (HSPC) [48]. In that study, an increased relative abundance of α-2,3-sialylation was found in HSPC whereas α-2,6-sialylation was prevalent in peripheral blood-derived CD34+ and CD34 cells. Thus, the authors suggested that the enriched α-2,3-sialylation of HSPC might constitute a mechanism to facilitate migration of HSPC from bone marrow into the circulation.

In recent years, the inflammatory component of osteoarthritis has been recognized as a decisive factor for the progression of the disease. The release of cytokines such as IL-1β or TNF-α from both the synovium and the chondrocytes themselves may be involved in the induction of inflammatory factors, including inducible nitric oxide synthetase, cyclooxygenase 2, and phospholipase A2 [4951], as well as proteinases that promote the destruction of cartilage matrix. The present work shows that IL-1β and TNF-α induce alterations of the differentiated chondrocyte phenotype and that these alterations are associated with a remarkable reorganization of chondrocyte N-glycan sialylation. Strongly enhanced ST3Gal4 mRNA levels and the parallel drop of ST6Gal1 expression result in reduced α-2,6- to α-2,3-sialylation ratios in chondrocyte glycoprotein N-glycans. We found that this linkage-type remodeling occurred at both S1H5N4F1 and S2H5N4F glycans, i.e. those H5N4F structures carrying one or two sialic acid residues. Of note, this finding is in line with recent in vivo experiments studying N-glycans in rabbit cartilage after ligament transection [20]. There, it was shown that the amount of S1(α-2,6)H5N4F1 structures significantly decreased in cartilage tissues excised from animals with induced osteoarthritis as compared to those of healthy animals. It appears noteworthy that the shift towards reduced α-2,6- to α-2,3-sialylation ratios in human chondrocytes under exposure to IL-1β and TNF-α, as observed in our study, is not a general cellular phenomenon. As reported recently, human endothelial cells respond to IL-1β, TNF-α, and lipopolysaccharide (LPS) with increased α-2,6-SiaT mRNA levels, increased α-2,6-SiaT enzyme activity, and enhanced cellular binding of SNA lectin [52]. In contrast, stimulation of human monocyte maturation with TNF-α, IL-1β, LPS, or IFN-γ leads to downregulation of both ST6Gal1 and ST3Gal4 transcription [19]. Thus, it can be concluded that the effect induced by pro-inflammatory molecules on cellular sialylation is strongly dependent on cell and tissue type. For human chondrocytes we propose a novel mechanism involving α-2,6-linked sialic acids that represent a typical structure of the differentiated chondrocyte phenotype, whereas the enhanced expression of α-2,3-linked sialic acids accompanies cellular dedifferentiation induced by IL-1β and TNF-α. In light of the significance of sialic acids in CD44-matrix interactions, a shift towards α-2,3-sialylation under inflammatory conditions may modify the binding ability of CD44 to the HA substrate, influencing cellular anchoring of matrix components and affecting signal transduction and gene expression [53,54]. Similarly to CD44, the β1 subfamily of integrins is substrate of several glycosyltransferases including ST6Gal1 [5558]. Although controversial and highly dependent on the cell type, those studies have demonstrated the relationship between sialylation of β1 integrins and adhesion to ECM components. Since chondrocytes express a number of different integrins including α1β1, α2β1, α3β1, α5β1, and α6β1 integrins [59], the altered N-glycan sialylation in the presence of TNF-α or IL-1β may potentially influence the binding affinity of these receptors for their matrix substrates. This hypothesis, however, still remains to be experimentally proven.

In conclusion, the present report has revealed that the expression ratio of α-2,6- to α-2,3-linked terminal sialic acids constitutes a signature of the differentiation status of human chondrocytes. Whereas α-2,6-linked sialic acids are the typical structure in primary human chondrocytes, the altered phenotype observed in chondrocytic cell lines or induced by IL-1β and TNF-α is associated with abundant α-2,3-linked sialic acids. This finding may shed light on the significance of sialylated N-glycan structures for cartilage homeostasis and enhance our understanding of cartilage pathophysiology. Future studies clarifying the molecular basis, as well as the functional consequences of shifting sialylation patterns in human chondrocyte N-glycans, are therefore anticipated.

Supplementary Material


Supplementary Figure 1:

Melting curves were generated by heating amplified qPCR products stepwise from 55°C to 95°C while continuously monitoring the fluorescence. Melting curves confirmed the absence of primer dimers and verified the specific amplification of single qPCR products of ST6Gal1, ST6Gal2, ST3Gal3, ST3Gal4, and ST3Gal6.


Supplementary Figure 2:

Differential regulation of α-2,6-linked and α-2,3-linked sialyl residues in S2H5N4F1 glycans by IL-1β and TNF-α.

The chromatographic separation of isobaric sialylated S2H5N4F1 glycans (taken from one representative experiment) using a porous graphitic column is presented for illustrative purposes. Trace 1 (lower lane) shows glycans of untreated primary human chondrocytes. Traces 2 (middle lane) and 3 (upper lane) show glycans of chondrocytes exposed to IL-1β and TNF-α, respectively.


Supplementary Table 1:

Primer sequences for qPCR assays.


The authors thank Franz Gabor and Michael Wirth for their constant support and interest. Verena Plattner and Thomas Dalik are gratefully acknowledged for technical assistance. Reinhard Zimmermann and Rénaud Leonard are acknowledged for helpful advice. Dr. Goldring’s research is supported by NIH grant R01-AG022021.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


1. Knudson CB, Knudson W. Cartilage Proteoglycans. Semin Cell Dev Biol. 2001;12:69–78. [PubMed]
2. Dudhia J. Aggrecan, aging and assembly in articular cartilage. Cell Mol Life Sci. 2005;62:2241–2256. [PubMed]
3. Nicoll SB, Barak O, Csóka AB, Bhatnagar RS, Stern R. Hyaluronidases and CD44 undergo differential modulation during chondrogenesis. Biochem Biophys Res Commun. 2002;292:819–825. [PubMed]
4. Varki A. Biological roles of oligosaccharides: all of the theories are correct. Glycobiology. 1993;3:97–130. [PubMed]
5. Varki A. Nothing in glycobiology makes sense, except in the light of evolution. Cell. 2006;126:841–845. [PubMed]
6. Wang PH, Leeb WL, Juanga CM, Yang YH, Lo WH, Lai CR, et al. Altered mRNA expressions of sialyltransferases in ovarian cancers. Gynecol Oncol. 2005;99:631–639. [PubMed]
7. Brockhausen I, Schutzbach J, Kuhns W. Glycoproteins and their relationship to human disease. Acta Anat. 1998;161:36–78. [PubMed]
8. Delmotte P, Degroote S, Lafitte JJ, Lamblin G, Perini JP, Roussel P. Tumor Necrosis Factor alpha increases the expression of glycosyltransferases and sulfotransferases responsible for the biosynthesis of sialylated and sulfated Lewis × epitopes in the human bronchial mucosa. J Biol Chem. 2002;277:424–431. [PubMed]
9. Howard S, Anastassiades T. Differential effects of bone associated factors on newly synthesized anionic glycoconjugates by articular chondrocyte cultures from adult and immature bovines. J Rheumatol. 1993;20:2083–2094. [PubMed]
10. Wagers AJ, Waters CM, Stoolman LM, Kansas GS. Interleukin 12 and interleukin 4 control T cell adhesion to endothelial selectins through opposite effects on alpha1, 3-fucosyltransferase VII gene expression. J Exp Med. 1998;188:2225–2231. [PMC free article] [PubMed]
11. Yang X, Lehotay M, Anastassiades T, Harrison M, Brockhausen I. The effect of TNFalpha on glycosylation pathways in bovine synoviocytes. Biochem Cell Biol. 2004;82:559–568. [PubMed]
12. Yang X, Yipa J, Anastassiades T, Harrison M, Brockhausen I. The action of TNFα and TGFβ include specific alterations of the glycosylation of bovine and human chondrocytes. Biochim Biophys Acta. 2007;1773:264–272. [PubMed]
13. Varki A. Glycan-based interactions involving vertebrate sialic-acid-recognizing proteins. Nature. 2007;446:1023–1029. [PubMed]
14. Chiricolo M, Malagolini N, Bonfiglioli S, Dall’Olio F. Phenotypic changes induced by expression of β-galactoside alpha2,6 sialyltransferase I in the human colon cancer cell line SW948. Glycobiology. 2006;16:146–154. [PubMed]
15. Malagolini N, Chiricolo M, Marini M, Dall’Olio F. Exposure of alpha2,6-sialylated lactosaminic chains marks apoptotic and necrotic death in different cell types. Glycobiology. 2009;19:172–181. [PubMed]
16. Dall’Olio F, Chiricolo M. Sialyltransferases in cancer. Glycoconjugate J. 2001;18:841–850. [PubMed]
17. Higai K, Ishihara S, Matsumoto K. NFkB-p65 Dependent Transcriptional Regulation of Glycosyltransferases in Human Colon Adenocarcinoma HT-29 by Stimulation with Tumor Necrosis Factor a. Biol Pharm Bull. 2006;29:2372–2377. [PubMed]
18. Takahata M, Iwasaki N, Nakagawa H, Abe Y, Watanabe T, Ito M, et al. Sialylation of cell surface glycoconjugates is essential for osteoclastogenesis. Bone. 2007;41:77–86. [PubMed]
19. Videira PA, Amado IF, Crespo HJ, Algueró MC, Dall’Olio F, Cabral MG, et al. Surface α2-3- and α2-6-sialylation of human monocytes and derived dendritic cells and its influence on endocytosis. Glycoconjugate J. 2008;25:259–258. [PubMed]
20. Watson M, Rudd PM, Bland M, Dwek RA, Axford JS. Sugar printing rheumatic diseases: a potential method for disease differentiation using immunoglobulin G oligosaccharides. Arthritis Rheum. 1999;42:1682–1690. [PubMed]
21. Matsuhashi T, Iwasaki N, Nakagawa H, Hato M, Kurogochi M, Majima T, et al. Alteration of N-glycans related to articular cartilage deterioration after anterior cruciate ligament transection in rabbits. Osteoarthr Cartilage. 2008;16:772–778. [PubMed]
22. Toegel S, Plattner VE, Wu SQ, Chiari C, Gabor F, Unger FM, Goldring MB, et al. Lectin binding patterns reflect the phenotypic status of in vitro chondrocyte models. In Vitro Cell Dev Biol Anim. 2009 in press. [PubMed]
23. Toegel S, Harrer N, Plattner VE, Unger FM, Viernstein H, Goldring MB, Gabor F, et al. Lectin binding studies on C-28/I2 and T/C-28a2 chondrocytes provide a basis for new tissue engineering and drug delivery perspectives in cartilage research. J Control Release. 2007;117:121–129. [PubMed]
24. DeCeuninck F, Lesur C, Pastoureau P, Caliez A, Sabatini M. Culture of chondrocytes in alginate beads. In: Sabatini M, Pastoureau P, DeCeuninck F, editors. Cartilage and Osteoarthritis. Totowa: Humana Press; 2004. pp. 15–22. [PubMed]
25. Toegel S, Wu SQ, Piana C, Unger FM, Wirth M, Goldring MB, et al. Comparison between chondroprotective effects of glucosamine, curcumin, and diacerein in IL-1β stimulated C-28/I2 chondrocytes. Osteoarthr Cartilage. 2008;16:1205–1212. [PubMed]
26. Rendic D, Wilson IBH, Lubec G, Gutternigg M, Altmann F, Léonard R. Adaptation of the “in-gel release method” to N-glycome analysis of low-milligram amounts of material. Electrophoresis. 2007;28:4484–4492. [PubMed]
27. Packer NH, Lawson MA, Jardine DR, Redmond JW. A general approach to desalting oligosaccharides released from glycoproteins. Glycoconjugate J. 1998;15:737–747. [PubMed]
28. Pabst M, Bondili JS, Stadlmann J, Mach L, Altmann F. Mass + Retention Time = Structure: A Strategy for the Analysis of N-Glycans by Carbon LC-ESI-MS and Its Application to Fibrin N-Glycans. Anal Chem. 2007;79:5051–5057. [PubMed]
29. Stadlmann J, Pabst M, Kolarich D, Kunert R, Altmann F. Analysis of immunoglobulin glycosylation by LC-ESI-MS of glycopeptides and oligosaccharides. Proteomics. 2008;14:2858–2871. [PubMed]
30. Piana C, Gull I, Gerbes S, Gerdes R, Mills C, Samitier J, et al. Influence of surface modification on vitality and differentiation of Caco-2 cells. Differentiation. 2007;75:308–317. [PubMed]
31. Barbosa I, Garcia S, Barbier-Chassefière V, Caruelle JP, Martelly I, Papy-García D. Improved and simple micro assay for sulfated glycosaminoglycans quantification in biological extracts and its use in skin and muscle tissue studies. Glycobiology. 2003;13:647–653. [PubMed]
33. Lefebvre V, Peeters-Joris C, Vaes G. Modulation by interleukin 1 and tumor necrosis factor alpha of production of collagenase, tissue inhibitor of metalloproteinases and collagen types in differentiated and dedifferentiated articular chondrocytes. Biochim Biophys Acta. 1990;1052:366–378. [PubMed]
34. Martin I, Jakob M, Schaefer D, Dick W, Spagnoli G, Heberer M. Quantitative analysis of gene expression in articular cartilage from normal and osteoarthritic joints. Osteoarthr Cartilage. 2001;9:112–118. [PubMed]
35. Goggs R, Carter SD, Schulze-Tanzil G, Shakibaei M, Mobasheri A. Apoptosis and the loss of chondrocyte survival signals contribute to articular cartilage degradation in osteoarthritis. Vet J. 2003;166:140–158. [PubMed]
36. Peters JH, Loredo GA, Benton HP. Is osteoarthritis a ‘fibronectin-integrin imbalance disorder’? Osteoarthr Cartilage. 2002;10:831–835. [PubMed]
37. Lyons TJ, Stoddart RW, McClure SF, McClure J. Lectin and other histochemical studies of the articular cartilage and the chondro-osseous junction of the normal human knee joint. J Mol Hist. 2007;38:13–23. [PubMed]
38. Przybysz M, Maszczak D, Borysewicz K, Szechiński J, Kątnik-Prastowska I. Relative sialylation and fucosylation of synovial and plasma fibronectins in relation to the progression and activity of rheumatoid arthritis. Glycoconjugate J. 2007;24:543–550. [PubMed]
39. Gebauer M, Saas J, Sohler F, Haag J, Söder S, Pieper M, et al. Comparison of the chondrosarcoma cell line SW1353 with primary human adult articular chondrocytes with regard to their gene expression profile and reactivity to IL-1β Osteoarthr Cartilage. 2005;13:697–708. [PubMed]
40. Finger F, Schörle C, Zien A, Gebhard P, Goldring MB, Aigner T. Molecular Phenotyping of Human Chondrocyte Cell Lines T/C-28a2, T/C-28a4, and C-28/I2. Arth Rheum. 2003;48:3395–3403. [PubMed]
41. Heiskanen A, Hirvonen T, Salo H, Impola U, Olonen A, Laitinen A, et al. Glycomics of bone marrow-derived mesenchymal stem cells can be used to evaluate their cellular differentiation stage. Glycoconjugate J. 2008;26:367–384. [PubMed]
42. Teriete P, Banerji S, Noble M, Blundell CD, Wright AJ, Pickford AR, et al. Structure of the Regulatory Hyaluronan Binding Domain in the Inflammatory Leukocyte Homing Receptor CD44. Mol Cell. 2004;13:483–496. [PubMed]
43. English NM, Lesley JF, Hyman R. Site-specific De-N-glycosylation of CD44 Can Activate Hyaluronan Binding, and CD44 Activation States Show Distinct Threshold Densities for Hyaluronan Binding. Cancer Res. 1998;58:3736–3742. [PubMed]
44. Lesley J, English N, Perschl A, Gregoroff J, Hyman R. Variant Cell Lines Selected for Alterations in the Function of the Hyaluronan Receptor CD44 Show Differences in Glycosylation. J Exp Med. 1995;182:431–437. [PMC free article] [PubMed]
45. Katoh S, Zheng Z, Oritani K, Shimozato T, Kincade PW. Glycosylation of CD44 Negatively Regulates Its Recognition of Hyaluronan. J Exp Med. 1995;182:419–429. [PMC free article] [PubMed]
46. Skelton TP, Zeng C, Nocks A, Stamenkovic I. Glycosylation Provides Both Stimulatory and Inhibitory Effects on Cell Surface and Soluble CD44 Binding to Hyaluronan. J Cell Biol. 1998;140:431–446. [PMC free article] [PubMed]
47. Breen KC, Potratz A, Georgopoulou N, Sandhoff K. The generation and characterization of a rat neural cell line overexpressing the alpha2,6(N) sialyltransferase. Glycoconjugate J. 1998;15:199–202. [PubMed]
48. Hemmoranta H, Satomaa T, Blomqvist M, Heiskanen A, Aitio O, Saarinen J, et al. N-glycan structures and associated gene expression reflect the characteristic N-glycosylation pattern of human hematopoietic stem and progenitor cells. Exp Hematol. 2007;35:1279–1292. [PubMed]
49. Hedbom E, Häuselmann HJ. Molecular aspects of pathogenesis in osteoarthritis: the role of inflammation. Cell Mol Life Sci. 2002;59:45–53. [PubMed]
50. Goldring MB. The role of the chondrocyte in osteoarthritis. Arth Rheum. 2000;43:1916–1926. [PubMed]
51. Lotz M. The role of nitric oxide in articular cartilage damage. Rheum Dis Clin North Am. 1999;25:269–282. [PubMed]
52. Hanasaki K, Varki A, Stamenkovic I, Bevilacqua MP. Cytokine-induced beta-galactoside alpha-2,6-sialyltransferase in human endothelial cells mediates alpha 2,6-sialylation of adhesion molecules and CD22 ligands. J Biol Chem. 1994;269:10637–10643. [PubMed]
53. Knudson W, Casey B, Nishida Y, Eger W, Kuettner KE, Knudson CB. Hyaluronan oligosaccharides perturb cartilage matrix homeostasis and induce chondrocytic chondrolysis. Arth Rheum. 2000;43:1165–1174. [PubMed]
54. Puré E, Cuff CA. A crucial role for CD44 in inflammation. Trends Mol Med. 2001;7:213–221. [PubMed]
55. Pochec E, Litynska A, Amoresano A, Casbarra A. Glycosylation profile of integrin alpha 3 beta 1 changes with melanoma progression. Biochim Biophys Acta. 2003;1643:113–123. [PubMed]
56. Seales EC, Jurado GA, Singhal A, Bellis SL. Ras oncogene directs expression of a differentially sialylated, functionally altered beta1 integrin. Oncogene. 2003;22:7137–7145. [PubMed]
57. Seales EC, Jurado GA, Brunson BA, Wakefield JK, Frost AR, Bellis SL. Hypersialylation of β1 integrins, observed in colon adenocarcinoma, may contribute to cancer progression by upregulating cell motility. Cancer Res. 2005;65:4645–4652. [PubMed]
58. Bellis SL. Variant glycosylation: an underappreciated regulatory mechanism for beta1 integrins. Biochim Biophys Acta. 2004;1663:52–60. [PubMed]
59. Loeser RF, Sadiev S, Tan L, Goldring MB. Integrin expression by primary and immortalized human chondrocytes: evidence of a differential role for alpha1beta1 and alpha2beta1 integrins in mediating chondrocyte adhesion to types II and VI collagen. Osteoarthr Cartilage. 2000;8:96–105. [PubMed]
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Compound
    PubChem chemical compound records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records. Multiple substance records may contribute to the PubChem compound record.
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence and PMC links.
  • MedGen
    Related information in MedGen
  • Nucleotide
    Primary database (GenBank) nucleotide records reported in the current articles as well as Reference Sequences (RefSeqs) that include the articles as references.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...