Search

More than 300 sera from patients with a connective tissuedisease were analyzed with the immunoblotting technique. Thepresence of autoantibodies against an 86,000-mol wt markerantigen for diffuse scleroderma (Scl-86) was found in 14 outof 33 patients with scleroderma. The presence of anti-Scl-86antibodies seemed to correlate with the diagnosis of diffusescleroderma since they were found in 13 out of 22 diffusescleroderma patients and in only one out of 11 patients withlimited scleroderma. All scleroderma sera (33 patients' seraand 13 reference sera) were also tested for the presence ofanti-Scl-70 antibodies. It was found that all anti-Scl-70 positivesera (n = 25) contained anti-Scl-86 antibody as well, suggestinga relationship between these two antigens. However, the Scl86antigen was shown to be an extremely insoluble nonchromosomalprotein, resistant to boiling in sodium dodecyl sulfate.This contrasts with the Scl-70 antigen, which has been describedas a thermolabile, soluble antigen present in the chromatinfraction. Together, our results are consistent with the idea thatScl-70 is a degradation product of Scl-86.The Scl-86 antigen is present in freshly prepared rabbitthymus, spleen, and liver nuclei as well as in nuclei fromvarious cultured cell lines, but is not detectable in extractablenuclear antigen from rabbit thymus.In a limited retrospective study, the anti-Scl-86 antibodieswere found in two sera from patients with Raynaud's phenomenonbefore the development of diffuse scleroderma. Therefore,it is possible that screening of patients' serum for this antibodymight predict the development of diffuse scleroderma.

The murine monoclonal antibody (MAb), designated DF3, reacts with a 300,000-mol wt mammary epithelial antigen. A sequential double-determinant radioimmunoassay (RIA) has been developed to monitor circulating DF3 antigen. Using this assay, we have demonstrated that 33 of 36 normal women had plasma RIA antigen levels less than 150 U/ml. In contrast, 33 of 43 patients (76%) with metastatic breast cancer had RIA DF3 antigen levels greater than or equal to 150 U/ml. The difference between these two groups was statistically significant (P less than 0.001). Similar results have been obtained with a double-determinant enzyme-linked immunoassay (EIA). Only 6 of 111 age-matched normal subjects had EIA DF3 antigens levels greater than or equal to 30 U/ml, while 42 of 58 patients (72%) with breast cancer had levels equal to or above this value. Thus, similar patterns of specificity are obtained with the EIA or RIA. The elevation of circulating DF3 antigen levels in breast cancer patients has been confirmed by transfer blot assays. MAb DF3 reactivity occurred predominantly with circulating antigens of three different molecular weights ranging from 300,000 to approximately 400,000 mol wt. We also demonstrate that patients with both primary and metastatic breast cancer who were free of detectable disease at the time of sampling have DF3 antigen levels that are similar to those obtained from normal subjects. While patients with hepatoma (27%) and ovarian carcinoma (47%) also had elevated circulating DF3 antigen levels, the results suggest that DF3 antigen levels may be useful in distinguishing breast cancer patients from those with esophageal, gastric, colorectal, pancreatic, and lung carcinomas. Furthermore, the results of the RIA, EIA, and transblot analyses demonstrate that the measurement of circulating DF3 antigen levels provides a new and potentially useful marker to follow the clinical course of patients with metastatic breast cancer.

Two O-hydroxylysyl glycosides, Hyl-Gal-Glc and Hyl-Gal, have been isolated from normal human urine and shown to be identical to two glycosides isolated from alkaline hydrolysates of collagen. A relatively sample and reproducible analytical procedure has been devised to measure the levels of these glycosides in human urine. By the use of this procedure it was shown that a normal diet has only a small effect on 24-hr urinary excretion levels of these glycosides indicating an endogenous origin. Urinary glycoside levels appear to be highest in children, roughly paralleling collagen turnover as indicated by urinary hydroxyproline levels. Collagen turnover equivalents calculated from urinary hydroxylysyl glycoside levels were found to be significantly larger than collagen turnover equivalents calculated from urinary hydroxyproline levels. This suggests that urinary glycosides are more quantitative indicators of collagen metabolism than urinary hydroxyproline. The ratio of Hyl-Gal-Glc to Hyl-Gal was measured in urines of diseased as well as normal individuals and a bimodal distribution was found. Alkaline hydrolysates of different human connective tissue collagens showed that only bone collagen, of the collagens examined, had a low ratio of Hyl-Gal-Glc to Hyl-Gal compared to human urine. Other collagens examined had higher ratios than found in human urine. On the basis of these results it is postulated that the bimodal distribution of glycoside ratios represents two populations of collagen turnover, the lower ratio population having a high bone collagen turnover, the lower ratio population having a high bone collagen turnover relative to the second population. Examination of the types of subjects making up the two populations supports this hypothesis. These data suggest that urinary O-hydroxylysyl glycoside excretion, in addition to providing a more quantitative estimate of collagen turnover than urinary hydroxyproline, may prove to be of value as a specific means of studying the metabolism of bone collagen.