| Markers of collagen metabolism in the assessment of rheumatoid arthritis.: With special reference to cross-linked carboxyterminal telopeptide of type I collagen (ICTP) | ||
|---|---|---|
| Prev | Chapter 2. Review of the literature | Next |
4-hydroxyproline, an imino acid formed in a posttranslational enzyme reaction from proline, is a necessary requirement for the stability of the triple-helical conformation of the collagen molecule at body temperature. It makes up about 12% of the weight of the collagen molecule, and can thus be used as a measure of the tissue collagen content. Acid hydrolysis followed by a chemical colour reaction to assess this imino acid can be used both for tissue samples and for urine.
Once freed from the helical part of a collagen molecule, hydroxyproline can no longer be incorporated into a new protein; yet, exogenous hydroxyproline is absorbed from the diet. Most of the imino acid (up to 90%) is metabolized in the liver, but some is always passed into the urine. The increase of hydroxyproline correlates with growth velocity in children and with the presence of bone degradation processes, e.g. bone metastases of cancers, in adults (Risteli & Risteli 1999). A generally enhanced bone metabolic rate also increases urinary hydroxyproline excretion. Some hydroxyproline is directly derived from collagen synthesis, since it is present in the helical domain of PINP and PIIINP. Although urinary hydroxyproline has been used as a marker of bone resorption for a long time, it is relatively non-specific owing to the fact that it is present in many types of collagen and that only about 10% of hydroxyproline is excreted into urine. Furthermore, in inflammatory conditions, the complement component C1q may elevate the urinary hydroxyproline levels since it contains collagen-like amino acid sequences in the form of a triple helix as part of its structure (Robins 1982, Reid 1982).
In contrast to the 4-hydroxyproline content, which is similar in all type I collagens, the extent of other posttranslational modifications varies from one situation and tissue to another. Such modifications include the hydroxylation of lysine and the hydroxylation of proline at position 3, leading to the formation of 3-hydroxyproline. The presence of a half-way glycosylated hydroxylysine, galactosylhydroxylysine, has been suggested to be characteristic of adult bone tissue, and the excretion of this amino acid into urine has been used as a marker of bone collagen degradation (Bettica et al. 1996). Nevertheless, it is not suitable for routine measurements because high-pressure liquid chromatography (HPLC), a complex and demanding technique, must be used (Cormier 1995).
The biological function of fibrillar collagen is to provide the tissue with tensile strength. The tensile strength is due to the covalent bonds, known as collagen cross-links, that develop between individual collagen molecules in a collagen fiber. The cross-links are formed in a complicated series of partially alternative chemical reactions that gradually lead, through divalent cross-links joining two polypeptide chains, to multivalent, i.e. tri- or even tetravalent, cross-links (Eyre et al 1984). The continuation of cross-linking explains the fact that older collagen in soft tissues is progressively less soluble when studied in various ways, e.g. by digestion with pepsin or bacterial collagenase. In addition to the intermolecular cross-links essential for tensile strength, intramolecular cross-links also exist (Risteli & Risteli 1999). In soft tissues, conditions associated with rapid remodelling of tissue seem to involve a different route of collagen degradation than does turnover under steady-state conditions (Everts et al. 1996).
A characteristic approach in resolving the structure of collagen cross-links has been to isolate these compounds from hydrolyzed tissue samples, i.e. to destroy the information of their original locations in collagen molecules. According to this tradition, the concentrations of either free or total cross-links can be measured in different body fluids, usually urine. This approach is more specific for assessing the breakdown of fibrillar collagen than the measurement of hydroxyproline, as the cross-links can only be derived from the degradation of collagen molecules that have participated in collagen fibers. Furthermore, pyridinoline cross-links are not absorbed from the diet, which means that their excretion is related only to their endogenous formation (Risteli & Risteli 1999). The urinary excretion of cross-links has been validated as a marker of the bone resorption rate (Seibel et al. 1992), but a substantial day-to-day variation in the deoxypyridinoline concentrations has been reported (Seyedin et al. 1993). Because urinary cross-link excretion decreases by 30% between 8:00 and 11.00 a.m. (Schlemmer et al. 1992), sampling time is important for reproducible results.
Pyridinoline (PYD) and deoxypyridinoline (DPYD), chemically correctly known as hydroxylysylpyridinoline and lysylpyridinoline, respectively, are the two nonreducible trivalent cross-links that stabilize type I collagen chains and are released during the degradation of mature collagen fibrils. Pyridinoline is abundant in bone and cartilage, whereas deoxypyridinoline is largely, although not entirely, confined to bone. Type III collagen also contains pyridinoline cross-links at the aminoterminus. Total PYD and DPYD are measured in hydrolyzed urine samples by fluorimetric detection after reversed-phase HPLC. This technique requires extensive sample pretreatment and purification steps, which limit its clinical application. It is possible to measure only free fractions of pyridinoline by avoiding the hydrolysis steps. During bone degradation, only about 40% of the cross-links are excreted in a free form. The remaining 60% continue to be peptide-bound (Cormier 1995).
The earlier studies to measure collagen degradation in RA with assays of PYD and DPYD have shown controversial results as to the association with radiological progression in hands and feet (Robins et al. 1986, Seibel et al. 1989).
It has been established that type I collagen has two cross-link forming sites, one in the aminoterminal peptide region and another in the carboxyterminal region of the molecule. Immunoassays to detect both N-telopeptide of type I collagen and carboxyterminal telopeptide of type I collagen (ICTP) have been described. The use of N-telopeptide of type I collagen as assayed in 24-hour urine samples (NTx) has been suggested to be valuable for monitoring changes in bone turnover in individual patients, for instance, after antiresorptive therapy (Cormier 1995). NTx is an assay recognizing small cross-linked peptides from urine that are derived from type I collagen. The assay is said to be a specific and responsive index of bone resorption activity (Garnero et al. 1994). No studies of RA patients had addressed urinary NTx excretion, until St. Clair et al. (1998) recently found urinary NTx excretion to be more abundant in patients with RA than in healthy controls.
The ICTP antigen prepared in vitro is a cross-linked peptide derived from mature type I collagen fibres. A similar, though somewhat larger, antigen seems to be liberated during the turnover of type I collagen fibres in vivo and can be found in human serum. ICTP contains a mature trivalent collagen cross-link and adjacent peptide material from three polypeptide chains. Two of these are the carboxyterminal ends of the α1 chains of one type I collagen molecule and the third is derived from either an α1 or an α2 chain of the helical region of another molecule. This structure has been verified by N-terminal protein sequencing. A compound of this kind can only be derived from the degradation of type I collagen molecules which have participated in the formation of collagen fibres (Risteli et al. 1993).
The standard and tracer antigens of the assay are cross-linked ICTP collagen liberated by digestion with bacterial collagenase or trypsin from decalcified human femoral bone. The ICTP assay has been shown to be a reliable marker for increased type I collagen degradation in situations that include local destruction of bone tissue, e.g. multiple myeloma (Elomaa et al. 1992), bone metastases from carcinomas (Aruga et al. 1997), and both early and advanced RA (Hakala et al. 1993b, Kotaniemi et al. 1994, Paimela et al. 1994). On the other hand, the circulating ICTP antigen levels do not reflect accelerated or retarded physiological bone resorption, such as seen in the postmenopausal state or during the use of estrogen replacement therapy (Hassager et al. 1994) or short-term treatment with bisphosphonates (Garnero et al. 1994). To determine the reason for this discrepancy, Sassi et al. (submitted) designed a study to characterize the antigenic determinant of the ICTP assay. They measured the immunoreactivity of the cleavage products of the major osteoclastic enzyme, cathepsin K. They showed that cathepsin K cleaves the trivalent ICTP structure between the phenylalanine-rich region and the cross-link, destroying the reactivity with ICTP antibodies. They postulated that, in vivo, the above process is part of physiologic and postmenopausal bone resorption, whereas the increased circulating concentrations of immunoreactive ICTP found in other clinical situations, as in multiple myeloma and RA, are not primarily derived from physiological cathepsin-K-mediated, osteoclastic bone turnover.
In a recent cross-sectional analysis, where 5 different markers of collagen degradation were tested in patients with advanced RA, serum ICTP together with urinary pyridinoline (PYD) were found superior to the other tested markers (urinary deoxypyridinoline, N-telopeptide and Crosslaps` assay for the C-telopeptide cross-linking domain) in discriminating patients with RA from healthy controls (St. Clair et al. 1998). These measures also had minimal short-term, day-to-day variability, and they were hence suggested to be useful in assessing the effect of potentially disease-modifying drug therapies (Cortet et al. 1998).