6.2. Detection of procollagens in serum, BALF and ELF

Procollagens are produced intracellularly and secreted to extracellular space, where procollagen propeptides are then cleaved from procollagen molecules by specific proteases (Prockop et al. 1979). Cleavage of carboxyterminal propeptide is essential for the formation of collagen fibres (Miyahara et al. 1982), but type III collagen retains commonly part of the aminoterminal propeptide in the collagen molecule, forming pN-collagen, which has been detected in mature collagen fibres (Karttunen et al. 1989). When these fibrils are later degraded, the antigenic material is liberated into the extracellular space (Risteli & Risteli 1986). As known, commercial radioimmunoassay kits for PIIINP employing bovine antigen measure partly type III pN-collagen as collagen fibrils degrade (Risteli & Risteli 1990). All except one of the previous studies have used commercial radioimmunoassay kits from two manufacturers, employing bovine antigen for the analysis of ­PIIINP in BALF and in serum. In the present study, human antibodies were used for the assays. The information of BALF-PICP in sarcoidosis, fibrosing alveolitis or asbestos-exposed patients is limited to one preliminary study without further information of the methodology (Tukiainen et al. 1994).

In the present study, the original serum methods for analysing PIIINP and PICP were modified to better detect the low propeptide concentrations in BALF, and the concentrations were also calculated in ELF as estimated by the urea method (Rennard et al. 1986) for the first time. Commercial kits used for serum samples are usually adjusted to measure levels within and above the normal range, and their reliability may diminish at concentrations close to the detection limit. An essential problem in investigating soluble constituents in BALF is the question of the detection limit because of the highly diluted samples. Some of the previous publications have used various methods to concentrate BALF (table 1). Most studies have used ultrafiltration. The concentration factor may vary even in the same study between 10-and 100- fold (Low et al. 1992). In some studies (Low et al. 1983, Bjermer et al. 1986, O’Connor et al. 1989, Low et al. 1992, Schaberg et al. 1994), the concentration of PIIINP has been adjusted to the albumin or protein concentration, which also affects the comparability of the previous studies. The only study using the same commercial radioimmunoassay method and employing the same human antigens as the present study did not report any difference in the levels of PIIINP in BALF or serum between sarcoidosis patients and controls. The BALF analyses were made with the kit for serum without any pre-assay processing of BALF. (Milman et al. 1995).

Some further reasons why the previous studies on procollagen propeptides in serum and BALF have been so contradictory may include the clinical limitations of the utility of BALF itself (Davis 1994) and the fact that there has been a variety of methods for analysing procollagen peptides in BALF and serum as well as the pre-assay concentration of BALF. There is a large range of normal values for the soluble constituents and differential counting of cell populations in BALF, depending on the method used or, for example, the smoking habits. Thus, abnormalities can rarely be considered specific. Airspace cells and secretions may not reflect interstitial processes, and the composition of epithelial lining fluid may change during bronchoalveolar lavage. (Davis 1994)

Some of the contradictory results of many previous studies can be explained partly with the above restrictions of the BALF assay and the technical variation of processing BALF as well as with the variable methods used to detect procollagen peptides. Based on the above considerations, some of the methodological problems have been avoided in the present study.