The effect of collagen hydrolysates on the activity of aggrecanases and the release of proteoglycans from human osteoarthritic articular cartilage

      Purpose: The aggrecanases ADAMTS4 (a disintegrin and metalloproteinase with a thrombospondin motif) and ADAMTS5 play a central role in the proteolytic degradation of osteoarthritic articular cartilage. Collagen hydrolysate (CH), consisting of type I collagen peptides are often used as dietary supplements for osteoarthritis. Our in silico calculations have revealed that the different collagen type I peptides bind to various docking sites on the aggrecanases so that their activities are affected differently. We also showed that small collagen peptides bind weakly to the α2A-domain of the integrin receptor, and that this represents another possible mechanism of action [Siebert et al. 2010 Stötzel et al. 2012]. Interestingly, our studies with CHs of bovine origin revealed considerable differences in their molecular composition and their effects on human chondrocytes [Schadow et al. 2013].
      The aim of this study was to test the extent to which two commercially available CHs are able firstly to influence the activity of ADAMTS4 and ADAMTS5 in vitro, and secondly to modulate the release of proteoglycans from human osteoarthritic cartilage explants.
      Methods: The differing composition of the CHs from hydrolytically generated collagen peptides was determined by MALDI-TOF mass spectrometry (MS). The activity of rhADAMTS4 and rhADAMTS5 was determined in the presence of the two CHs (Peptan™ F 5000 HD from Rousselot, and Mobiforte® from Astrid Twardy) using rhAggrecan-IGD as substrate, whereby the proteolytically released aggrecan peptide with the N-terminal sequence ARGSVIL was quantified in a microtitre plate using a mAb assay. The cartilage explants from the lateral condyle of the femur were obtained during a knee TEP implantation after written consent was obtained from the patients subsequent to approval by the local university ethics committee. After macroscopic evaluation of the arthritic changes using Collins’ procedure, explants were then prepared and cultivated. The extent of cartilage degradation was determined using explants after a 6 day treatment with 0-10 mg/ml CH where medium was changed after 3 days. Proteoglycan content was measured using the DMMB method, with TIMP-3 and MMP-3 content being determined by ELISA and NO content being measured using the Griess reaction. The vitality of the explants was determined by fluorescence microscopy. The results were compared with untreated control samples. Each experiment was repeated five times (n = 6). The data was statistically analysed by ANOVA analysis.
      Results: MALDI-TOF MS revealed striking differences between the different preparations. They differed both in terms of the composition of their mixtures regarding individual peptide CH fragments, and in terms of their mean molecular weight. For Peptan™ F 5000 HD the mean molecular weight was 3,600 Da, while for Mobiforte® it was slightly smaller at 3,300 Da. Only Mobiforte® significantly and concentration-dependently stimulated the activity of rhADAMTS4 and rhADAMTS5. This probably allosterically-induced stimulation was associated with a significantly increased release of proteoglycans and NO from human cartilage explants, while no effect was measured with respect to TIMP-3 levels. Peptan F™ 5000 HD concentration-dependently inhibited the activity of aggrecanases, while no effect was measured for the NO concentrations, proteoglycan release or the levels of TIMP-3.
      Conclusions: The CHs differ significantly regarding the molecular composition of the peptide mixtures, and also regarding their catabolic effects on human osteoarthritic articular cartilage. Before their use in patients, CHs should be thoroughly investigated regarding their effects first in vitro and then in vivo before being declared as safe and effective.