ADAMTS-4 and ADAMTS-5 sequestration and activity in chondrocyte-agarose cultures

Introduction A primary event in the destruction of cartilage in arthritic diseases is the loss of aggrecan from the extracellular matrix of articular cartilage. During aggrecan breakdown, cleavage sites are utilized, which reside within the IGD of the aggrecan core protein. The Asn³⁴¹–Phe³⁴² bond is cleaved by members of the MMP family, whereas the second of the two cleavage sites, the Glu³⁷³–Ala³⁷⁴ bond, is cleaved by the aggrecanases which are all members of the A disintegrin and metalloproteinase with thrombospondin motifs (ADAMTS) family. Both ADAMTS‐4 and ‐5 have been shown to readily cleave aggrecan at this so‐called aggrecanase site ( Tortorella et al. 1999 ; Abbaszade et al. 1999 ; Sandy et al. 2000 ). An in vitro model of cartilage degradation has also shown that these enzymes are responsible for the loss of aggrecan from explant cultures of articular cartilage stimulated with IL‐1‐a or TNF‐a ( Tortorella et al. 2001 ). Both ADAMTS‐4 and ‐5 are thought to be synthesized as inactive zymogens that are activated via removal of their propeptide domain by the Golgi enzyme furin. The secreted active ADAMTS‐4 and ‐5 have predicted molecular weights of 67.9 and 73.6 kDa, respectively. The objective of this study was to monitor ADAMTS‐4 and ‐5 secretion and sequestration in the extracellular matrix of chondrocyte‐agarose cultures. Methods Porcine articular chondrocytes were isolated and embedded in agarose ( Hughes et al. 1997 ) before preculture in DMEM + 50 µg/ml gent. with 10% FBS and 25 µg/ml Phos.C for 21 days. The plates were washed, then cultured in serum‐free DMEM with or without IL‐1‐α for 96 h. GAG release to the medium was measured using the DMMB assay. Media samples were analysed by Western blotting for aggrecan metabolites using mAb BC‐3 to recognize the aggrecanase‐generated neoepitope ARGSV. The presence of ADAMTS‐4 in the media was analysed using the mAb anti‐TS‐4N, which recognizes the metalloproteinase domain, and commercially available polyclonal antibodies to the pro‐ and spacer domains of ADAMTS‐4. The presence of ADAMTS‐5 was detected using commercially available polyclonal antibodies to the pro‐ and spacer domains of ADAMTS‐5. Agarose plugs were extracted in detergent buffer and analysed by Western blotting for ADAMTS‐4 and ‐5 using the same monoclonal and polyclonal antibodies. Results In control cultures, only 20–30% of the total GAG was released into the medium after 96 h of culture. In contrast, 80–90% of the total GAG was released in cultures exposed to IL‐1. Western blot analysis showed aggrecanase‐generated aggrecan metabolites in the IL‐1‐treated cultures but none in the control cultures. Sequestered forms of both ADAMTS‐4 and ‐5 are present in the matrix prior to treatment in serum‐free conditions, and following treatment with or without IL‐1 for 96 h, there are no differences in the high molecular weight isoforms of the enzymes sequestered in the matrix. Western blots of partially purified media samples showed no differences in the zinc chelator‐bound isoforms of either ADAMTS‐4 or ‐5 between control and IL‐1‐treated cultures. However, the predominant heparin sepharose‐bound isoforms of ADAMTS‐4 and ‐5 co‐migrate at approximately 37 kDa. Each of the heparin‐bound 37‐kDa isoforms of ADAMTS‐4 and ‐5 are detected in increased amounts in IL‐1α‐treated cultures compared to controls. Discussion The increased amounts of the 37‐kDa isoforms of both ADAMTS‐4 and ‐5 in the IL‐1‐treated cultures suggest a role for these smaller isoforms in the increased aggrecanase activity seen in the IL‐1‐treated cultures compared to controls. This study has identified multiple isoforms of putative aggrecanase activity that could be responsible for increased aggrecan catabolism that leads to cartilage degradation in arthritis.     

Gene expression and matrix proteinase activity in osteochondrosis dessicans

Introduction Osteochondrosis dessicans (OCD) is a disorder of unknown aetiology where often a fragment of cartilage and subchondral bone separates from the articular surface. Previous studies have shown histological changes in glycosaminoglycan content in OCD cartilage compared to normal cartilage ( Koch et al. 1997 ). It has also been shown in equine OCD cartilage that there is excessive degradation of type‐II collagen compared to normal cartilage ( Laverty et al. 2002 ). The present study was undertaken to examine the gene expression in human OCD cartilage compared to its normal autologous articular cartilage and human osteoarthritic (OA) cartilage. Methods Cartilage from five OCD patients (18–34 years) was obtained at the time of surgery. Pieces of cartilage were either snap‐frozen (in preparation for RNA isolation) or the proteoglycans extracted with 4 m GuHCl. Total RNA was isolated from the cartilage using RNeasy minicolumns and reagents (Qiagen) according to the manufacturer's protocol. RT‐PCR was performed using an RNA PCR kit (Perkin‐Elmer) using a number of oligonucleotide primers. GuHCl‐extracted proteoglycan fragments were analysed using Western blotting with a number of antibodies to aggrecan metabolites, collagen metabolites and the small leucine‐rich proteoglycans. Results and discussion When OCD cartilage was compared to normal and human OA cartilage, there was an increase in aggrecan, collagen type‐II and collagen type‐X RNA expression. There was no change in RNA expression of link protein or type‐I collagen. The RNA expression of the aggrecanases (ADAMTS enzymes) was also different in the three different cartilage samples. Neither ADAMTS‐1, ‐4 or ‐5 was present in the normal cartilage. In contrast, in the OCD cartilage, there was expression of both ADAMTS‐1 and ‐4, whereas in the OA cartilage, there was expression of ADAMTS‐4 and ‐5. In the case of MMP RNA expression, MMP‐3 was decreased and MMP‐13 increased in OCD cartilage compared to both normal and OA samples. In addition, the expression of all three TIMP isoforms was increased in the OCD cartilage. Although inflammatory components are not expected in OCD pathology, expressions of inflammatory mediators such as COX‐2, IL‐1‐α and TNF‐α were all increased in the OCD cartilage when compared to normal, but expression of these mRNAs in the OA cartilage was higher. Analysis of proteoglycan fragments in the OCD cartilage by Western blotting showed the presence of aggrecan fragments containing the G1 domain, interglobular domain and the C‐terminal neoepitope generated by aggrecanase cleavage. There was also immunoreactivity for biglycan and link protein. Conclusion These results suggest that the phenotypic expression of chondrocytes at the site of the OCD lesion are markedly different from ‘normal’ articular cartilage and also pathological OA cartilage. Interestingly, the expression patterns of matrix proteinases and their natural inhibitors were also markedly different in OCD cartilage, again suggesting that there are specific biochemical expression patterns in OCD pathology, which may potentially be biomarkers of the disease process. Further studies are necessary to elucidate how the differences in gene expression and matrix protease activity may be involved in the aetiology of OCD.     

PRG‐4/SZP N‐ and C‐terminal domains: cloning,expression and characterization

Introduction Proteoglycan‐4 (PRG‐4), also known as superficial zone protein/proteoglycan (SZP), is an approximately 345‐kDa mucinous proteoglycan that has been detected in a variety of tissues including cartilage, tendon, bone, heart and liver ( Ikegawa et al. 2000 ). In the synovial joint, PRG‐4 is specifically synthesized by chondrocytes located in the superficial zone of articular cartilage and by some surface‐lining cells of the synovium ( Schumacher et al. 1994 ). Sequence analyses have shown that the N‐ and C‐terminal vitronectin‐like domains of PRG‐4 may impart interesting functions relevant to synovial joint metabolism ( Merberg et al. 1993 ; Flannery et al. 1999 ). The objective of this study was to investigate these potential functions, facilitated by the production of PRG‐4 N‐ and C‐terminal domains as recombinant proteins. Methods cDNAs for the human N‐terminal (exons 2–5) and bovine C‐terminal (exons 7–12) domains of PRG‐4 were obtained by RT‐PCR and cloned into the expression vector pMT‐BiP for inducible, secreted expression in Drosophila S2 cells. Proteins were purified using FLAG‐M2 antibody affinity chromatography and visualized by SDS‐PAGE and Western blotting with PRG‐4‐specific antibodies. The heparin‐binding properties of recombinant proteins were investigated using heparin affinity chromatography. The interactions of recombinant PRG‐4 domains with human plasminogen activator‐inhibitor (PAI)‐1 and bovine type‐II collagen were assayed using standard ELISA techniques. Results Stable cell lines have been generated that express human N‐terminal and bovine C‐terminal PRG‐4 domains. In both cases, two proteins have been purified, possibly due to a splice mechanism by the expression system. N‐terminal sequence data and Western blotting indicate that the two species in each case could represent full‐length and truncated proteins. Analyses of the two PRG‐4 N‐terminal domain species have confirmed the presence of a predicted heparin‐binding domain and indicate that the molecule can bind to PAI‐1, with binding activity localized towards its two somatomedin B domains. The somatomedin B domain of vitronectin is known to bind PAI‐1 ( Seiffert 1997 ). Analyses of the two PRG‐4 C‐terminal species have demonstrated self‐association under nonreducing conditions and binding to heparin and PAI‐1. Discussion The exact role of PRG‐4 in the synovial joint is yet to be elucidated. However, these results point towards the interaction of the N‐ and C‐terminal domains of PRG‐4 with structural molecules such as type‐II collagen and heparin, and functional molecules such as PAI‐1, a serpin that is involved in the fibrinolytic cascade and cell adhesion. These properties are in addition to the well‐documented boundary lubricating activity of the central mucinous region of PRG‐4.     

Molecular recognition and modulation of hepatocyte growth factor activity by heparan and dermatan sulfates

Introduction Hepatocyte growth factor/scatter factor (HGF/SF) is an unusual growth factor in that it binds both heparan sulfate (HS) ( Lyon et al. 1994 ) and dermatan sulfate (DS) ( Lyon et al. 1998 ) glycosaminoglycans (GAGs) with similar high affinities. Both these GAGs act as co‐receptors for HGF/SF in the activation of the Met receptor ( Lyon et al. 2002 ). Our aim was to determine the sequences in HS and DS that specifically interact with and modulate HGF/SF activity. Materials and methods A structurally unique DS, which possesses O‐sulfation at carbon‐6 of the hexosamine residue (and not carbon‐4 as in mammalian DS), was obtained from the sea cucumber, Ascidia nigra. A variety of HS‐ and DS‐like structures were also generated using various chemical modification procedures (specific desulfations and carboxyl reductions). The ability of these various GAG species to compete with cell surface GAGs for HGF/SF binding was tested using radiolabelled HGF/SF and MDCK cells. The modified GAG structures and the A. nigra DS are currently being tested for their ability to act as co‐receptors for the interaction between HGF/SF and Met by studying cell signalling and cellular response assays, using the sulfated GAG‐deficient CHO‐745 cell line. Results Unexpectedly, A. nigra DS was found to bind HGF/SF strongly with a KD of around 1 nm . This interaction is 20‐fold stronger than that of between HGF/SF and mammalian DS, but similar to that of with HS. A. nigra DS also stimulated HGF/SF‐mediated Erk activation and migration in CHO‐745 cells. Studies using the modified GAG species showed that, in the case of HS, 6‐O‐sulfate and N‐sulfate groups are most important for HGF/SF binding. For HGF/SF binding to DS, hexosamine O‐sulfate is most important. HGF/SF was also found to bind 6‐O‐sulfated GAGs more strongly than 4‐O‐sulfated ones. Discussion The data show that there is flexibility in the structures recognized by HGF/SF, and this explains the ability of the growth factor to bind both HS and DS. However, there are still observable preferences in GAG structure, such as 6‐O‐sulfation over 4‐O‐sulfation. Information on HGF/SF‐binding GAG structures is valuable for the design of HGF/SF antagonists that could be useful therapeutically in the treatment of solid tumours where HGF/SF‐Met activity is up‐regulated.     

Characterization of the promoter in ADAMTS-5 (aggrecanase-2)

Introduction Loss of aggrecan metabolites from the cartilage matrix into the surrounding synovial fluid have allowed for identification of two major cleavage sites within aggrecan: Asn³⁴¹–Phe³⁴² and Glu³⁷³–Ala³⁷⁴ peptide bonds. Proteolysis at the Glu³⁷³–Ala³⁷⁴ bond has been specifically attributed to ADAMTS‐4 and ADAMTS‐5. Genetic reporters are commonly used in cell biology to study mechanisms regulating gene expression. Firefly luciferase is widely used as such a reporter for the immediate availability of reporter activity, the sensitivity of the assay, and quickness and ease of use. By making time‐controlled cuts in the 5′‐noncoding region of ADAMTS‐4 and ‐5 and inserting the deletions into the genetic reporters, which are then transfected into a suitable cell line, activity can be measured. The project aim was to identify the cis and trans acting factors in the 5′‐noncoding regions that regulate ADAMTS‐5 expression using the Luciferase Assay System (Promega, Madison, WI, USA). Methods A 3‐kb fragment containing the 5′‐flanking region of ADAMTS‐5 was subcloned into pGL3‐Basic (Promega), a luciferase reporter vector lacking a eukaryotic promoter. Truncations have also been made to the full‐length construct using Erase‐a‐Base System (Promega) at restriction enzyme sites SacI at 11 bp from the origin and NheI at 21 bp from the origin, resulting in 200–2300 bp deletions on the TS‐5 vector. Adherent kidney embryonic fibroblast‐like cells (PEAK) were grown in RPMI 1640 media containing supplements (Gibco Invitrogen Corporation) at 37 °C in 5% CO₂. Cells were transfected using Fugene‐6 Reagent (Roche) as recommended by the manufacturer. 1.5 × 10⁵ cells were plated out in 12‐well plates and grown overnight. The next day, serum media was removed and RPMI 1640 media containing only glutamate was added. DNA complexes were prepared by mixing 97 µl of serum‐free medium, 3 µl of Fugene‐6, and 2 µg of DNA from Control, Basic, or TS‐5 full constructs and selected truncations and incubated for 30 min at RT. Hundred microlitres of the complex was pipetted directly into each well. After a 4–5‐h incubation, 100 µl of FBS was added to each well and left overnight at 37 °C in 5% CO₂. The next morning, the cells were harvested and washed twice in PBS, and resuspended in ×1 Passive Lysis Buffer. About 20 µl of the cell lysate was assayed for luciferase activity as described in the manufacturer's protocol (Promega). Results and discussion Full‐length and seven deletion fragments (numbering downstream from the ATG start codon) spanning the 5′‐noncoding region were selected from a library of 41 deletions prepared by the Erase‐a‐Base System: 3b.4 (?2322 bp), 4a.5 (?2110 bp), 5a.2 (1810 bp), 4a.2 (1427 bp), t3a.3 (?718 bp), 2a.1 (?433 bp) and 4a.1 (?278 bp). Firstly, an increase in activity occurs with 256 base pairs deleted from the full‐length vector (3b.4). With control considered 100% of activity, the full‐length vector produces greater than that amount indicating TS5 vector as having double the activity. A greater increase in activity (×6 that of the control) between the full length and 3b.4 suggests the presence of a suppressor somewhere in the first 256 base pairs of the 5′‐noncoding region. Activity then drops between 3b.4 and 4a.5. However, activity is still greater than that produced by the full length. A levelling‐off occurs at 5a.2, which continues until a sudden decrease at t3a.3 that is maintained in 2a.1 and 4a.1. Loss of activity between 4a.2 and t3a.3 is linked to loss of a TATA box at 1753–59 bp. There is a total of seven motifs existing in the first few hundred base pairs that could be possible suppressors including an H1 conserved site at 248–254 base pairs. By following the same methods using selected deletion fragments within ?2565 to ?2322, it will be possible to identify which motif is responsible for suppressing activity in this cell line. Similar experiments have been carried out in isolated articular chondrocytes.     

Anti-inflammatory actions of green tea catechins and ligands of peroxisome proliferator-activated receptors

Introduction Green tea catechins and peroxisome proliferator‐activated receptor (PPAR) agonists have been shown to reduce inflammation associated with collagen‐induced arthritis in mice ( Haqqi et al. 1999 ; Cuzzocrea et al. 2003 ) and to inhibit other markers of inflammation ( Vankemmelbeke et al. 2003 ; Sabatini et al. 2002 ; Singh et al. 2003 ; Jiang et al. 1998 ). We are investigating the mechanisms by which catechins and PPAR agonists interfere with the signalling pathways of the pro‐inflammatory cytokines, IL‐1 and TNF. Materials and methods Epigallocatechin gallate (EGCG), epicatechin gallate (ECG), epigallocatechin (EGC), epicatechin (EC) and fenofibrate were from Sigma Aldrich, Poole, UK. Rosiglitazone was from Cayman Chemical, Ann Arbor, MI, USA. Cytokines and IL‐8 ELISA were from NIBSC. IL‐6 ELISA from eBiosciences (San Diego, CA, USA). Cells used were primary fibroblasts from inflamed periodontal tissue or osteoarthritic chondrocytes from discarded tissue following surgery. Results We show the significant suppression of cytokine‐induced IL‐6 and IL‐8 release by EGCG, which is dose‐ and cell‐type dependent. The other three catechins have a much lesser effect. Preliminary results show that PPAR agonists also interfere with TNF‐induced IL‐8 production. Discussion Green tea has a long history of human consumption, and epidemiological studies have shown that it can reduce inflammatory diseases ( Greenwald et al. 2002 ). Our results confirm that EGCG has the strongest anti‐inflammatory effect, compared with the other three catechins. They also indicate that the mechanism of action of EGCG is cell‐type dependent. This could be because the compound targets components of pro‐inflammatory signalling pathways whose expression is restricted to certain differentiated cell types. Alternatively, the different sensitivities of cells to EGCG could reflect differences in their ability to metabolize the compound. Further experiments are needed to discriminate between these hypotheses.     

A novel keratanase-generated keratan sulfate antibody and its application to structural analysis of skeletal and corneal keratan sulfate

Introduction The objective of this study was to make monoclonal antibodies specific for keratanase‐generated neoepitopes in keratan sulfate (KS) and to use them along with existing KS monoclonal antibodies (e.g. 5D4, IB4) to investigate KS sulfation pattern motifs in connective tissue proteoglycans during development, ageing and disease. Methods Bovine nasal cartilage aggrecan (BNC A1D1) was trypsin digested, generating a range of GAG‐peptide fragments. The sample was then subjected to anion‐exchange and size exclusion chromatography to separate KS peptides from CS attachment domain fragments. Fractions were analysed by Western blotting for positive immunoreactivity for KS, then pooled and keratanase digested to generate ‘KS stub’ antigens. Immunization and fusions were carried out as previously described ( Caterson et al. 1983 ; Hughes et al. 1992 ). Screenings involved the use of a range of antigens; including keratanase vs. keratanase II‐digested bovine cartilage aggrecan and bovine corneal KS‐PGs. A new monoclonal antibody, BKS‐I, was identified that specifically recognized a keratanase‐generated neoepitope on both skeletal and corneal KS. This novel monoclonal antibody was used along with existing KS monoclonal antibodies 5D4 and 1B4 to investigate KS structure. Results and discussion Bovine trypsin‐generated aggrecan KS‐peptides were chondroitinase ABC treated and either keratanase or keratanase II treated. The digests were run on SDS‐PAGE and immunolocated with monoclonal antibody 5D4 (that recognizes linear disulfated N‐acetyl lactosamine disaccharide‐containing segments in KS) and the new ‘KS‐stub’ monoclonal antibody BKS‐I. Our results indicated that there was reduced monoclonal antibody 5D4 immunostaining after keratanase pretreatment. However, keratanase II digestion completely removed all 5D4 structural epitopes. In contrast, BKS‐I showed no immunostaining on the untreated KS‐peptides but strong staining on keratanase treated samples and no staining after keratanase II digestion. Similar patterns of immunoreactivity were observed with Western blot analysis of untreated, keratanase treated and keratanase II treated corneal KS‐PGs. Conclusion These data indicate that monoclonal antibody BKS‐I recognizes a nonreducing terminal neoepitope‐containing sulfated N‐acetylglucosamine adjacent to a nonsulfated lactosamine disaccharide. We also conclude that skeletal KS must have a structure with four possible variations opposed to the generic structures, proposed as being made of disulfated disaccharides at the nonreducing end, followed by a series of monosulfated disaccharides at the middle and nonsulfated disaccharides nearer the linkage region. 5D4 staining, observed after keratanase digestion, indicates that there must be a minimum structure of a pentasulfated hexasaccharide remaining on the KS chain ‘stubs’ near the linkage region of skeletal and corneal KS. The BKS‐I monoclonal antibody can be used to demonstrate differential substitution of KS GAG chains in the CS attachment region of cartilage aggrecan with ageing. It has also proven useful for immunohistochemical analyses identifying the sites of KS–PG association with collagen lamellae of cornea.     

A Metric Linkage Disequilibrium Map of a Human Chromosome

We used LDMAP ( Maniatis et al. 2002 ) to analyse SNP data spanning chromosome 22 ( Dawson et al. 2002 ), to obtain a whole‐chromosome metric LD map. The LD map, with map distances analogous to the centiMorgan scale of linkage maps, identifies regions of high LD as plateaus (‘blocks’) and characterises steps which define the relationship between these regions. From this map we estimate that block regions comprise between 32% and 55% of the euchromatic portion of chromosome 22 and that increasing marker density within steps may increase block coverage. Steps are regions of low LD which correspond to areas of variable recombination intensity. The intensity of recombination is related to the height of the step and thus intense recombination hot‐spots can be distinguished from more randomly distributed historical events. The LD maps are more closely related to the high‐resolution linkage map ( Kong et al. 2002 ) than average measures of ρ with recombination accounting for between 34% and 52% of the variance in patterns of LD (r = 0.58 – 0.71, p = 0.0001) . Step regions are closely correlated with a range of sequence motifs including GT/CA repeats. The LD map identifies holes in which greater marker density is required and defines the optimal SNP spacing for positional cloning, which suggests that some multiple of around 50,000 SNPs will be required to efficiently screen Caucasian genomes. Further analyses which investigate selection of informative SNPs and the effect of SNP allele frequency and marker density will refine this estimate.     

Increased turnover of proteoglycan aggregates in tendon vs. cartilage

Introduction Although the function of proteoglycans (PGs) within the tendon extracellular matrix are not fully understood, changes in their turnover have been associated with tendinopathies ( Riley et al. 1994 ). In contrast to cartilage, aggrecanases are constitutively expressed and active in tendon ( Rees et al. 2000 ), indicative of a high rate of aggrecan turnover. Clinical trials investigating the use of active site MMP inhibitors have been confounded by side effects, which involve tendonitis and ‘musculoskeletal syndrome’. Such side effects may relate to nonspecific inhibition of tendon aggrecanases required to maintain normal metabolic homeostasis. The purpose of this study, therefore, was to compare the rate of turnover of tendon and cartilage PGs derived from the same joint and to determine the effect of MMP inhibitors (actinonin and marimastat) on aggrecan catabolism. Materials and methods Deep digital flexor tendon explants from compressed and tensional regions were dissected from young and mature bovine. Explants were precultured and then cultured for a further 4 days with or without marimastat (0–2 µm ) or actinonin (0–200 µm ). PG and lactate quantification, Western blot analysis of degradation products and RT‐PCR analyses were performed. In a separate experiment for measurement of PG turnover, explants were set up as described above and then pulse chase labelled with [³⁵S] sulfate. The rate of turnover of ³⁵S‐labelled PGs from the matrix of tendon (and articular cartilage obtained from the same animal) was subsequently calculated from the amount of ³⁵S‐labelled macromolecules appearing in the medium each day and that remaining in the matrix of explants at the termination of culture. Results PG turnover (presumably predominantly aggrecan) was markedly higher in tendon vs. cartilage. This difference was apparent in tendons from all regions and ages. Both marimastat and actinonin inhibited aggrecanase‐mediated PG catabolism in both tendon and cartilage explants. As expected, mRNA expression for the aggrecanases, MMPs and TIMPs was unaffected by the addition of these inhibitors to the culture medium. Discussion Aggrecan turnover in tendon is higher than that of articular cartilage, which may be attributed to distinct physiological properties of this PG in tendon. Importantly, immunohistochemical staining for aggrecan in tendon indicates its presence in between collagen fibres and fibril bundles ( Vogel et al. 1999 ), and thus aggrecan aggregates may dissipate resultant compressive loads by resisting the flow of water in these locations. In addition, aggrecan may facilitate the sliding of fibrils during the small amount of elongation of the tendon whilst under tension. Thus, the half‐life of tendon aggrecan is significantly reduced because it constantly participates in repeated resistance to compression. Our data also demonstrates that both marimastat and actinonin can inhibit aggrecanase‐mediated PG catabolism in tendon cultures. This suggests that the occurrence of ‘musculoskeletal syndrome’ in clinical trial patients may be due to the fact that these inhibitors affect the activity of aggrecanases in tendon, thus preventing them from playing their normal role in tendon aggrecan turnover and consequently perturbing normal physiological function.     

Age-related changes in the glycosaminoglycans of human meniscal aggrecan

Introduction Meniscal damage and degradation, which are strongly correlated with subsequent OA, have been identified in approximately 60% of people over 60 years of age. Age‐related changes in articular cartilage glycosaminoglycans (GAGs) have been described, and used to facilitate the study of pathology‐related changes ( Plass et al. 1998 ). However, such data do not yet exist for the meniscus. Materials and methods Undamaged human menisci were obtained following leg amputations, and the vascular and avascular zones of each lateral and medial meniscus were extracted into 4 m GuHCl. Aggrecan was recovered in the A1 fraction following CsCl density gradient centrifugation, and the relative abundance of chondroitin, dermatan and keratan sulphates (CS, DS and KS) was examined by NMR spectroscopy at 400 MHz and 43 °C. Results Human meniscal aggrecan was shown to contain CS, DS and KS, and our data show age‐related changes in the relative abundance of these GAGs. The change was similar for medial and lateral menisci and for the vascular and avascular zones within these. The KS abundance in aggrecan from young menisci (<15 years) was found to be 15–20% of the total GAGs. However, in older samples, it comprised only 7–12% of the GAGs. We have confirmed the presence of DS in human meniscal aggrecan and show that the abundance of DS gradually falls from approximately 16% at 10 years to 2–4% at 75 years. There is some variability between humans, although the trend is clear and for each human there is good agreement between medial and lateral menisci and vascular and avascular locations. The levels of CS comprise the remainder of the GAG attached to aggrecan and contribute the remainder of the GAG abundance. This can be seen to increase from 67 to 72% at 10 years to approximately 90% at 75 years. Discussion Our data show a clear age‐related change in the relative abundance CS, DS and KS from human meniscal aggrecan. The data show a decrease in the abundance of KS and DS and a concomitant increase in CS levels. These observations differ from those widely seen for articular cartilage, in which the levels of CS are seen to fall with age. We have confirmed that DS is a component of human meniscal aggrecan in agreement with previous work ( McNicol & Roughley 1980 ). However, previously reported levels of DS, approximately 20%, are those found only in younger menisci. Absolute levels of these GAGs have not yet been determined, and hence the mechanisms which bring about this relative increase in CS with age may include either changes in biosynthetic output and/or widespread GAG loss in which KS and DS loss increases with age.     

Conformation of glycosaminoglycans by ion mobility mass spectrometry

Introduction Glycosaminogycans (GAGs) are polysaccharides found on most animal cell surfaces and extracellular matrices. They constitute an important class of biologically active macromolecules that are implicated in the biological activity of morphogens, growth factors, cytokines, chemokines and enzymes. Intervention in signalling pathways they are involved in, represents a valuable therapeutical opportunity for cancer treatment, wound healing and specific organ regeneration. GAGs are linear polysaccharides with alternating uronic acid and hexosamine residues. Their polysaccharide precursors are extensively modified (epimerization, N‐deacetylation N‐, O‐sulfation) creating a wide heterogeneity to their structures. The binding specificity of GAGs is encoded in their primary structures but ultimately depends on how this basis set of functional groups is presented to a protein in three‐dimensional space. Methods As a part of our efforts to characterize the conformation of GAGs, we present here our preliminary results of conformational studies of heparin‐derived oligosaccharides by ion mobility mass spectrometry. This MS technique is designed to study conformation of mass‐selected gas‐phase ions. When ions drift through an inert gas under the influence of a weak electric field, their arrival time depends on their collisional cross sections. From this data, the experimental cross sectional areas are easily derived. These are then compared with those obtained from candidate geometries generated from molecular modelling, and hence conformation(s) of studied compounds can be determined. Results and discussion We have measured the cross sectional areas of four heparin oligosaccharides: one disaccharide and three tetrasaccharides. Amongst these are two fully sulfated species and also two tetrasaccharides each with one sulfate group missing. Our samples were prepared by enzymatic digestion of heparin and contain a nonreducing terminal unsaturated glycosyluronic acid ring. We employ the GLYCAM 2000 force field (Woods) and the AMBER ( Case et al. 2002 ) suite of programs to model the gas‐phase conformations of the oligosaccharides. We have generated AMBER parameters for sulfate groups using ab initio calculations and are in the process of generating structures for which we calculate the cross sectional areas using a projection approximation ( Wyttenbach et al. 1997 ). We present a comparison of experimental cross sections with those of calculated models. We also compare them with cross sections calculated using oligosaccharide fragments extracted from X‐ray and NMR structures of heparin oligosaccharides – protein complexes.     

A novel keratanase-generated keratan sulphate antibody and its applications

Introduction The sequencing of the genome has provided us with important information regarding the primary structure of many matrix proteins. This in turn has lead to advances in studies of the functions of post‐translational modifications on connective tissue proteoglycans (PGs). Changes in GAG structure with ageing and disease have been well documented ( Thonar et al. 1986 ; Brown et al. 1998 ). However, little is known about the exact sites of and differential substitution of GAGs on the aggrecan core protein and how these substitutions facilitate normal function or the changes seen with disease. The CS : KS ratio of substitution change significantly, with KS levels increasing with age and decreasing with the onset of disease. Objective was to produce monoclonal antibody (MAb) reagents to keratanase (k'ase) generated stub epitopes, that could be used to help identify, characterize and quantify sites of KS substitution on PGs, providing the potential to determine how the arrangement of such substitutions change with development, ageing and pathology. Methods Bovine Nasal Cartilage aggrecan (BNC A1D1) was trypsin digested, generating a range of glycosaminoglycan (GAG) fragments. The sample was then subjected to anion‐exchange and size exclusion chromatography to separate KS from CS fragments. Fractions collected were analysed by SDS‐PAGE and Western blotting. Fractions positive for KS were pooled and kinase digested to expose the KS stub antigens. Immunization and fusions were carried out as previously described ( Nieduszynski et al. 1990 ). Initial screenings were carried out using ELISA. Briefly, 96‐well microtitre plates were coated with the immunizing antigen overnight at 37 °C. The plates were then blocked prior to the addition of hybridoma media for 1–2 h at 37 °C. Binding was detected using an alkaline phosphatase‐conjugated secondary antibody for 1 h at 37 °C prior to the addition of the substrate. Positive wells were further screened by ELISA and SDS‐PAGE using the immunizing antigen, chondroitinase‐digested BNC and an A1D1 BNC preparation to establish the kinase stub specificity of the hybridomas. Further screenings by Western blotting was carried out on positive hybridomas selected. Antigens used included keratanase‐digested bovine corneal KS‐PGs, keratanase‐II‐digested KS‐PGs and a nonkeratanase‐digested corneal KS‐PG sample. Results Screening: Screening identified two positive hybridomas, B‐KS‐I and B‐KS‐II, which were specific for kinase‐generated KS stub. On screening, these antigens showed reactivity specifically for kinase‐digested BNC abc core, with no reactivity to the nonkinased linear KS GAG epitopes. Reactivity to kinase‐digested corneal KS‐PGs indicated that the MAbs generated were indeed to a stub structure in the KS chain and not to some linkage region epitope, amino acid sequence or oligosaccharide present on the core protein. Application: Immunohistochemistry utilizing B‐KS‐I was used to localize KS in a range of tissues along side anti‐KS 5D4. In human articular cartilage engineered grafts, labelling showed B‐KS‐I and 5D4 to have broadly overlapping labelling patterns for KS; however, label for B‐KS‐I had a much more restricted and subtle tissue distribution than that of antibody 5D4. Discussion These new KS stub MAbs have potential to be used in many different areas of research. They may be used in analysis of trypsin‐digested purified aggrecan from cattle joints of different ages to determine sites of KS substitution, which remain common or change with development and ageing. They may also be used in analysis of cartilage explant culture metabolites to assess KS substitution on the aggrecan fragments generated after stimulation of these cultures with cytokines such as IL‐1 or TNF‐α. Collectively it will provide important new information on the changing pattern of KS substitution in connective tissue PGs with development, ageing and the onset of pathology.     

The importance of elastin and fibulin-5 on spine development: a study of elastin KO and fibulin-5 KO on the development of vertebral body and intervertebral disc

Introduction Idiopathic scoliosis is the most common scoliosis and generally develops during juvenile or adolescent growth spurt ( Stehbens 2003 ). It was reported that elastic fibre system might play a role in some idiopathic scoliosis patients ( Hadley‐Miller et al. 1994 ). Indeed, transgenic mice with defects in elastic fibre system (including elastin null and fibulin‐5 null) appear to develop severe kyphoscoliosis. The aim of this study was to understand how such defects exert their effect on the spinal column. Materials and methods Newborn elastin KO and 16‐week‐old fibulin‐5 KO mice spines were fixed in 10% formalin. Paraffin‐embedded sections (20 µm) were stained with haematoxylin & eosin (H&E) and alcian blue after the sections were dewaxed and rehydrated. The sections were examined by light microscopy. Results H&E staining revealed that those newborn elastin KO mice seem to have a delayed ossification in vertebral bodies compared to that of wild‐type. Also, cell morphology in IVD appears very much different. Cells in outer annular (OA) and endplate regions are much round‐like comparing fibroblast‐like cells in wild‐type. In addition, GAG expressions showed by alcian blue staining appear much sparse and irregular in the matrix of IVD in newborn elastin KO mice. Fibulin‐5 KO mice (16 weeks) seem to have many cell clusters or clones in the growth plate, which is an indication of abnormal growth. Our results reveal the importance of elastin and fibulin‐5 on the development of spine. Discussion Kypho‐scoliosis is a spinal deformity. Several different spinal tissues, e.g. muscle, vertebrae, IVD and ligament, are involved in the stability and load carriage of the spinal column. Some gene defects in these load‐bearing structures can lead to the scoliotic deformity, e.g. elastin null, fibulin‐5 null, collagen‐II null ( Aszodi et al. 1998 ), perlecan null ( Costell et al. 1999 ) and LTBP3‐null ( Dabovic et al. 2002 ) as well as a mutant in a muscle‐specific protein ( Blanco et al. 2001 ), but other defects [such as collagen‐IX KO's ( Kimura et al. 1996 )] do not. The relationship between defects in these structures and development of scoliotic is still unclear.     

Changes in mucin-type O-glycosylation in breast cancer: implications for the host immune response

The major epithelial mucin, which is expressed by more than 90% of breast carcinomas and epithelial breast cancer cell lines, is MUC1. The MUC1 membrane mucin has a relatively simple structure, where a tandem repeat (TR) region carrying multiple O‐glycans forms the major part of the extracellular domain. The gene shows a size polymorphism, due to variation in the number of tandem repeats, each of which has five potential glycosylation sites. MUC1 has been found to be a good model for studying mucin‐type O‐glycosylation both in vitro and in vivo. Changes in the pattern of glycosylation of MUC1 have been observed in breast cancer ( Burchell et al. 2001 ) and have led to an interest in MUC1 as a possible target antigen for immunotherapeutic intervention ( Taylor‐Papadimitriou et al. 2000 ). The mucin expressed at high levels in the normal lactating mammary gland carries mainly core 2‐based O‐glycans (2–3 per tandem repeat). In breast cancers, the level of the ST3 Gal 1 enzyme is elevated ( Burchell et al. 1999 ) and this is associated with increased numbers of Sialyl T O‐glycans being added, as predicted from in vitro studies, showing competition for the core 1 substrate between this sialyl transferase and the relevant core 2 enzyme ( Dalziel et al. 2001 ). Moreover, the TR domain is more densely glycosylated. While the Sialyl T O‐glycan is commonly found on normal cells (e.g. resting T cells), it is cancer associated in the mammary gland. However, several lines of evidence indicate that this MUC1 glycoform, as expressed on breast cancers, can be immunosuppressive, maybe reflecting an inhibitory interaction with a lectin, possibly a siglec ( Nath et al. 1999 ), on immune effector cells. In 25–30% of breast cancers, O‐glycosylation is terminated very early so that GalNAc (Tn) or NeuAc α2,6 GalNAc (Sialyl Tn) glycans appear. Sialyl Tn is highly tumour‐specific and is detected immunohistochemically only when the enzyme ST6GalNAcI is expressed. The Tn and Sialyl Tn will bind to different lectins, including MGL and could stimulate the immune response. Preliminary studies in mouse models suggest that an immunization protocol which includes MUC1‐carrying Sialyl Tn O‐glycans can protect against tumour growth, while clinical studies in breast cancer patients using immunogens based on the Sialyl Tn glycan are underway.     

Analysis of CD44 hyaluronan-binding domain mutants by NMR

Introduction Hyaluronan (HA) is a ubiquitous high molecular mass glycosaminoglycan composed of a repeating disaccharide. CD44, the major cell surface receptor for HA, has a HA‐binding domain (CD44_HABD) at the N‐terminus of the protein, the 3D structure of which has been determined by both NMR and X‐ray crystallography ( Teriete et al. 2004 ); NMR spectra collected on the protein in complex with HA oligosaccharides has allowed us to predict how they may thread across the interaction surface. Amino acids previously implicated in HA binding include R41, Y42, R78 and Y79 ( Peach et al. 1993 ; Bajorath et al. 1998 ), which form a cluster on the surface of the Link module‐like region ( Teriete et al. 2004 ), as well as residues in the C‐terminal extension (R150, R154, K158 and R162) ( Peach et al. 1993 ). The position of the putative‐binding residues in the C‐terminal segment, and NMR data, led to the hypothesis of two modes of HA binding ( Teriete et al. 2004 ). Here, this hypothesis is tested by NMR studies of single‐site mutants in the context of the CD44_HABD construct. Materials and methods Four mutants of CD44_HABD were made, each with a single residue substitution (R150A, R154A, K158A and R162A). These constructs were expressed as ¹⁵N‐labelled proteins in Escherichia coli, refolded and purified to homogeneity. ¹H–¹⁵N HSQC spectra were acquired on the mutants in the presence of varying concentrations of HA hexasaccharide (HA6) and compared to the wild‐type construct to determine changes of protein fold and ligand binding. Results The mutants R150A, R154A and K158A have similar HSQC spectra to wild‐type CD44_HABD except for local chemical shift perturbations around the altered residue. Conversely, the R162A mutant has widespread chemical shift differences compared to wild‐type indicating that this mutation disrupts the fold. On binding HA6, the R150A, R154A and K158A mutants all experience shift perturbations similar to that seen with the wild‐type protein. Discussion The interaction of HA6 with wild‐type CD44_HABD has been found to cause a significant conformational change in the protein ( Teriete et al. 2004 ). The results here indicate that this ligand‐induced rearrangement can also occur in the R150A, R154A and R158A mutants. Therefore, these mutations do not seem to affect the binding of CD44 to HA6. It has been shown previously that the R162A mutation has reduced affinity for HA compared to wild‐type protein ( Peach et al. 1993 ). This loss of function may be due to the perturbation of the protein fold, and it is possible therefore that R162 does not participate directly in binding. Work is in progress to test the functional activity of these CD44_HABD mutants using ELISA‐like assays and to investigate their chemical shift perturbation in the presence of longer oligosaccharides.     

Hyaluronan: a simple polysaccharide with structural plasticity and functional diversity

Hyaluronan (HA) is a high‐molecular weight glycosaminoglycan that is involved in an extracellular matrix (ECM) organization and cell adhesion, essential to a wide range of normal physiological processes (e.g. development, immunology and reproduction). Its diverse biological activities may seem surprising for a simple linear polysaccharide composed solely of a repeating disaccharide of glucuronic acid and N‐acetylglucosamine. However, HA is likely to be able to take up many different conformational states, which are transient and rapidly interchanging in solution ( Day & Sheehan 2001 ). It has been suggested that the interaction of HA with different HA‐binding proteins (hyaladherins) may stabilize particular conformations of the polysaccharide leading to HA‐protein complexes with distinct architectures and unique biological properties ( Day & Sheehan 2001 ). Consistent with this hypothesis, recent studies on TSG‐6 ( Blundell et al. 2003 ) and CD44 ( Teriete et al. 2004 ) and molecular modelling of other members of the Link module superfamily reveal significant structural diversity in the HA‐binding domains in these related hyaladherins. Furthermore, there is evidence emerging that the complexes formed between polymeric HA and proteins can exhibit a wide range of higher order structures depending on the hyaladherin(s) involved. In some cases, these complexes may play mainly a structural role in ECM (e.g. the link protein/aggrecan/HA aggregates that provide load bearing function to cartilage), while in other cases they act as cell binding/activation sites (e.g. HA cables formed during inflammation). There is evidence that the function of HA can be modulated by its interaction with proteins. For example, preincubation of HA with TSG‐6 enhances/induces the binding of HA to cell surface CD44 on constitutive/inducible lymphocyte backgrounds (Lesley et al.); these activities are probably caused by the formation of cross‐linked HA fibrils that lead to receptor clustering. HA can also become modified by the covalent attachment of heavy chains (HCs), derived from the interα‐inhibitor, where TSG‐6 plays an essential role in this process (Mukhopadhyay et al.); this reaction is critical for the formation of a HA‐rich ECM required for successful ovulation and fertilization. HC‐HA complexes are also formed during inflammation but their function is not yet clear.     

Role of the EGF-like domains in mammalian tolloid (mTLD) secretion and procollagen C-proteinase activity

Introduction Bone morphogenetic protein (BMP)‐1 and its larger splice variant mammalian tolloid (mTLD) belong to the tolloid group of astacin‐like metalloproteinases that are fundamental to tissue patterning and extracellular matrix assembly. BMP‐1 and mTLD exhibit similar substrate specificity in vitro; however, BMP‐1 is a much better procollagen C‐proteinase than mTLD. mTLD consists of a prodomain (which is cleaved by a furin‐like enzyme) ( Leighton & Kadler 2003 ), a zinc metalloproteinase domain and a C‐terminal part comprising five CUB domains thought to be important for protein–protein interactions ( Hartigan et al. 2003 ), and two EGF‐like domains, which in other proteins are involved in calcium ion binding. BMP‐1 lacks the most C‐terminal two CUB domains and one EGF‐like domain. mTLD activity is known to be calcium ion dependent, as demonstrated for the chick homologue ( Hojima et al. 1985 ). In our current work, we are studying the role of the EGF‐like domains in the secretion and procollagen C‐proteinase activity of mTLD, and the contribution that these domains made to calcium ion dependency. Materials and methods We designed proteins lacking EGF1, EGF2 or both. NotI sites were introduced by PCR at the borders of the EGF domain of a cDNA clone encoding a V5‐His mTLD. Restriction enzyme digestion was used to delete individual domains. The mutant constructs in pCEP4 were stably transfected into 293‐EBNA cells. Expression was analysed by Western blot. The wild‐type and the mutant enzymes were purified on a nickel ion column, and their activity was determined by cleavage of type‐I procollagen in the presence or absence of 5 mm CaCl₂. Results We showed that (1) the mTLD proteins lacking EGF1, EGF2 or EGF1 + EGF2 were poorly secreted into the culture medium compared to mTLD and (2) the EGF deletion mutants remained calcium ion dependent, but some differences were seen. Most notably, the ΔEGF2 and ΔEGF1 + ΔEGF2 mutants were found to be better C‐proteinases than the wild‐type enzyme in the presence of calcium ions. Conclusion From these preliminary data, we concluded that (1) the EGF domains are necessary for efficient secretion (2) both EGF1 and EGF2 domains contribute to the calcium ion dependency of mTLD and (3) the EGF2 domain might be a Ca²⁺‐activated hinge that ‘swings’ the CUB‐4 and CUB‐5 domains away from the active site. The ?EGF2 mTLD might be expected to have an open conformation, thereby making it a better C‐proteinase than the wild‐type enzyme, and (?4) Ca²⁺ ions are bound by other domains in mTLD and not only by the EGF‐like domains.     

ADAMTS13 mutations and polymorphisms in congenital thrombotic thrombocytopenic purpura

Congenital thrombotic thrombocytopenic purpura (TTP) (also known as Upshaw‐Schulman syndrome, USS) is a rare, life‐threatening disease characterized by thrombocytopenia and microangiopathic hemolytic anemia, associated with the deficiency of the von Willebrand factor‐cleaving protease (ADAMTS13) due to mutations in the corresponding gene. The spectrum of clinical phenotype in congenital TTP is wide, encompassing neonatal‐onset disease and adult‐onset disease, forms with a single disease episode and chronic‐relapsing forms. We review ADAMTS13 gene variants associated with inherited ADAMTS13 deficiency and congenital TTP. To date, 76 mutations of ADAMTS13 are reported in the literature. Missense mutations, which constitute nearly 60% of ADAMTS13 mutations, preferentially localize in the 5′‐half of the gene encoding the N‐terminal half of the protein, where the domains that are indispensable for ADAMTS13 catalytic function are situated. In vitro expression studies in cell cultures have shown that defects in protein secretion and catalytic activity are the main mechanisms responsible for the deficiency of ADAMTS13 in congenital TTP patients. Even if data from the literature suggest the existence of genotype–phenotype correlations, a clear relationship between the type and the effect of ADAMTS13 genetic defects with disease manifestations remains to be established. Hum Mutat 30:1–9, 2009. © 2009 Wiley‐Liss, Inc.     

Up-regulated lactate, TGF-β and collagen synthesis in the ischaemic skin of patients with peripheral vascular disease

Introduction We have previously shown that dermal hypoxia alters collagen turnover in chronically ischaemic skin; however, no mechanism for this has been determined. In cultured human dermal fibroblasts, hypoxia causes an up‐regulation of collagen synthesis, probably mediated by TGF‐β ( Falanga et al. 2002 ) in the presence of increased lactate. Here, we examine whether these processes are present in vivo in chronically ischaemic skin. Materials and methods Paired biopsies of uninjured skin were harvested at below knee amputation from 16 patients with a history of peripheral vascular disease (PVD), following the quantification of ischaemia, using the ankle brachial pressure index (ABPI) and by lactate measurement. Nonischaemic samples were taken proximally from the amputation resection margin and ischaemic samples from a predetermined distal site. Site‐matched biopsies were taken for control at total knee replacement and varicose vein operations. Lactate levels were measured using enzymatic determination (Sigma, UK), collagen type‐I synthesis was determined by immunoassay for released C‐terminal propeptide (PICP) (Prolagen C, Quidel) and TGF‐β by ELISA. TGF‐β RI and RII were localized using immunohistochemistry. Results The ABPI in all patients with PVD was <0.4 indicating severe ischaemia. Levels of lactate were elevated in the ischaemic tissue of these patients when compared to nonischaemic samples (P < 0.001), and an up‐regulation of collagen type‐I synthesis was demonstrated in the ischaemic samples (P < 0.01). Levels of TGF‐β were also raised (P < 0.05). TGF‐β RI and RII were expressed on dermal fibroblasts, keratinocytes and endothelial cells. Discussion Increased lactate levels resulting from hypoxic metabolism have been demonstrated in skin flaps of animal models ( Hoopes & Im 1978 ); however, lactate levels in human tissue, as a direct assessment of chronic ischaemia, have not previously been reported. Hypoxia, lactate and TGF‐β have been shown to stimulate collagen synthesis in vitro ( Falanga et al. 2002 ; Cerbon‐Ambriz et al. 1991 ) but not in vivo in PVD. These findings are consistent with the hypothesis that chronic hypoxia leads to changes in the ECM of uninjured but ischaemic skin and may predispose it to dermal failure.