A-070 Development of high-affinity monoclonal antibodies specific to human thyroid stimulating hormone (TSH) that recognize the TSH variant R55G

Abstract Background Most current TSH assays are third-generation assays with a functional sensitivity of 0.01 µIU/mL or less. Immunoassays with higher sensitivity are needed to accurately determine TSH concentration in the serum of patients with severe hyperthyroidism. In addition, TSH assays should be able to detect different TSH variants that occur in the population. A novel mutation in TSH ß-subunit R55G was recently reported (2014, Drees et al.). This mutation disrupted the interaction of antibodies that recognized epitopes containing R55 with TSH. For accurate measurements, TSH assays should be able to detect wild-type (WT) TSH and the TSH variant R55G. The goal of this study was to provide high-affinity monoclonal antibodies that are suitable for the development of TSH assays with a functional sensitivity of 0.001 µIU/ml, which recognize the TSH variant R55G. Methods WT TSH and TSH variant R55G were produced in a mammalian cell line. Two sheep were immunized with WT TSH. B-cells were isolated and used for the development of recombinant antibodies. Full-size chimeric recombinant antibodies consisting of sheep IgG variable domains and human IgG1 constant domains were produced in a mammalian cell line. Results Twenty-nine sheep antibodies were developed which recognized WT TSH and the TSH variant R55G. We identified two groups of antibodies based on specificity. Antibodies of one group were specific to the ß-subunit, while antibodies of the second group were specific to the TSH dimer. None of the antibodies recognized recombinant a-subunit, luteinizing hormone (LH), follicle-stimulating hormone (FSH), or chorionic gonadotropin (hCG). Antibodies were tested in different combinations using a chemiluminescent particle-based sandwich immunoassay to find the antibody combinations with the best performance. Antibody combinations that demonstrated the highest signal-to-noise ratios were selected and five prototype immunoassays were developed. In the developed immunoassays, one antibody was specific to the TSH dimer and the second antibody was specific to the ß-subunit. Equilibrium dissociation constants (KD) were determined using the Octet for the best antibodies. KD were in the range 0.01-0.1 nmol. All of the prototype immunoassays recognized the TSH variant R55G along with WT TSH and native TSH purified from human pituitary glands. Cross-reactivity of immunoassays with recombinant LH, FSH (10 IU/ml), and hCG (1000 IU/ml) was absent for all immunoassays. LoD for immunoassays was determined using the 3rd International Standard for TSH. LoDs for all five immunoassays were 0.0008 µIU/ml. TSH concentrations in human serum samples (n = 50) were determined using prototype immunoassays and the reference immunoassay Roche Elecsys TSH. We observed a good correlation (Pearson*s correlation coefficient > 0.99) between TSH concentrations that were determined using developed immunoassays and the reference assay. Conclusion We developed sheep monoclonal antibodies that are suitable for the accurate quantitation of extremely low TSH concentrations in serum samples. Functional sensitivity of prototype immunoassays using new antibodies exceeded 0.001 µIU/ml. Antibodies recognized the TSH variant R55G, thus enabling the accurate detection of TSH in serum samples from individuals carrying this mutation. Neither individual antibodies nor immunoassays using these antibodies had cross-reactivity with other human glycoprotein hormones.