Microheterogeneity of transthyretin in serum and ascitic fluid of ovarian cancer patients

Abstract

Background: Transthyretin (TTR), a traditional biomarker for nutritional and inflammatory status exists in different molecular variants of yet unknown importance. A truncated form of TTR has recently been described to be part of a set of biomarkers for the diagnosis of ovarian cancer. The main aim of the study was therefore to characterize differences in microheterogeneity between ascitic fluid and plasma of women affected with ovarian cancer and to evaluate the tumor site as the possible source of TTR. Methods: Subjects were 48 women with primary invasive epithelial ovarian cancer or recurrent ovarian carcinoma. The control group consisted of 20 postmenopausal women. TTR and retinol-binding protein (RBP) levels were measured by enzyme-linked immunoassay (ELISA) and C-reactive protein (CRP) levels by a high-sensitivity latex particle turbidimetric assay. The molecular heterogeneity of TTR was analysed using immunoprecipitation and matrix-associated laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS). Presence of TTR in tumor tissue was determined with indirect peroxidase immunostaining. Results: TTR and RBP (μg/ml) levels in serum were 148.5 ± 96.7 and 22.5 ± 14.8 in affected women compared to 363.3 ± 105.5 and 55.8 ± 9.3 in healthy postmenopausal women (p < 0.01). In ascitic fluid, levels were 1.02 ± 0.24 and 4.63 ± 1.57 μg/ml, respectively. The mean levels of TTR and RBP in serum showed a tendency to decrease with the severity of the disease and were lower in affected women whose CRP levels were > 40 mg/ml (p = 0.08 for TTR; p < 0.05 for RBP). No differences in TTR microheterogeneity were observed between TTR isolated from serum of affected and healthy women or from ascitic fluid. TTR occurred rather consistently in four variants. Mass signals were at 13758 ± 7, 13876 ± 13 (greatest intensity), 13924 ± 21 and 14062 ± 24 Da, representing native, S-cysteinylated, S-cysteinglycinylated and glutathionylated TTR, respectively. Serum of healthy and affected women as well as ascitic fluid contained the truncated fragment of TTR (12828 ± 11 Da). No immunoreactive TTR was observed in the tumor sites. Conclusion: The severity of the cancer associated catabolism as well as the inflammation status affect serum TTR and RBP levels. Neither TTR nor its truncated form originates from tumor tissue and its occurrence in ascites may well reflect the filtration from blood into ascitic fluid. © 2005 Gericke et al., licensee BioMed Central Ltd.