George Milne

Turnover rates of different collagen types measured by isotope ratio mass spectrometry

The rates of collagen turnover in different tissues have been estimated in growing rats previously exposed to gaseous 18O2. The abundance of the stable isotope was measured using isotope ratio mass spectrometry following combustion of isolated collagen-derived hydroxyproline. Using this method, problems of label reutilization associated with radiolabelling methods are avoided. In general the results confirm the slow turnover rates with half-lives of total collagen in skin, muscle and gut of 74, 45 and 244 d, respectively. The use of cyanogen bromide digests of whole tissues followed by isolation of collagen type-specific peptides has allowed the comparison of turnover rates of collagen types I and III, indicating that collagen type III is turned over more rapidly than type I.

Isolation and characterization of a metalloproteinase secreted by rat glioma cells in serum-free culture.

The isolation of a metalloproteinase secreted by a rat glioma cell line (BT5C) in serum-free media is described. After affinity purification, the activity was present as a double band with Mr 86000 and 76000 both of which required CaCl2 for activity. The enzyme was able to degrade gelatin but not casein. It was unable to degrade native types I, III, IV and V collagens but their denatured counterparts were degraded. Using a radiolabel release assay the enzyme was inhibited by EDTA, 1:10 phenanthroline and TIMP confirming that it belongs to the family of metalloproteinases. Its activity was not affected by either serine or cysteine protease inhibitors. The proteinase was activated by APMA but was unaffected by trypsin treatment.

Evidence for Renal Tubular Resorption of Collagen Fragments from Immunostaining of Rat Kidney with Antibodies Specific for Denatured Type I Collagen

Antibodies to rat collagens I and III were raised in sheep. The antisera to collagen I were separated by affinity chromatography into components specific for either native or denatured forms. Immunolabelling of rat kidney sections with antibodies to native collagen I showed staining only of the interstitial matrix. By contrast, antibodies to denatured collagen I revealed the presence of immunoreactive material primarily in the upper part of the proximal tubules, detected in both fixed and cryostat sections. In fixed material, the granular appearance of staining in the region of the brush border was shown to be distinct from the protein droplets counterstained by toluidine blue. Collagen III antibodies stained the interstitial matrix in a similar pattern to that for native collagen I, but no proximal tubule staining was observed despite the fact that antibodies to denatured collagen III were shown to be present. No material reactive with denatured collagen I antibodies was detected in urine or serum by an inhibition ELISA technique. The results are discussed in terms of renal tubular resorption of collagen degradation products.

A radiolabel-release microwell assay for proteolytic enzymes present in cell culture media

A modified method for the measurement of proteolytic enzyme activity in cell culture-conditioned media has been developed. Using the release of 3H-labeled peptides from 3H-labeled gelatin the method is performed in microwell plates. The substrate is insolubilized and attached to the wells by glutaraldehyde treatment, thus eliminating the need for a precipitation step at the end of the assay. The assay is sensitive, reproducible, and convenient for small sample volumes. The effect of different protease inhibitors on activity can be assessed rapidly allowing an early characterization of the enzyme. It can also be adapted to microplate spectrophotometric analysis by staining residual substrate with Coomassie blue.

Identification of lysine-derived crosslinks in porcine collagen type X from growth plate and newly mineralized bone

Intact collagen type X cannot readily be extracted from the growth plate. Both the use of pepsin to release this molecule from tissue and the relative solubility of collagen type X following treatment of chick embryos with beta-aminopropionitrile (Chen et al., 1992) suggest that the insolubility may by brought about by the formation of lysine-derived crosslinks. By immunocytochemical labelling using antibodies specific for collagen type X, we have shown that this collagen type persists in the cartilaginous spicules present in metaphyseal bone and appears to be colocalized with collagen type II. The combined concentration of the reducible bifunctional crosslinks, dihydroxylysinonorleucine and monohydroxylysinonorleucine, in collagen type X isolated from the premineralized and newly mineralized growth plate was about 0.6 residues/ molecule, a level which might explain the relative intractability of collagen type X. Pyridinoline and deoxypyridinoline were present in very small amounts in collagen type X; this suggests that, unlike the situation in other types of collagen, few of the bifunctional crosslinks undergo maturation to pyridinium compounds. Although it is clear that collagen type X contains lysinederived crosslinks, work is in progress to establish which molecule also participates in the formation of these crosslinks.