A high-sensitivity ELISA for detection of human FGF18 in culture supernatants from tumor cell lines.

Fibroblast growth factor 18 (FGF18) is elevated in several human cancers, such as gastrointestinal and ovarian cancers, and stimulates the proliferation of tumor cells. This suggests that FGF18 may be a promising candidate biomarker in cancer patients. However, the lack of a high-sensitivity enzyme-linked immunosorbent assay (ELISA) does not permit testing of this possibility. In this study, we generated monoclonal antibodies against human FGF18 and developed a high-sensitivity ELISA to measure human FGF18 at concentrations as low as 10 pg/mL. Of the eight tumor cell lines investigated, we detected human FGF18 in culture supernatants from four tumor cell lines, including HeLa, OVCAR-3, BxPC-3, and SW620 cells, albeit the production levels were relatively low in the latter two cell lines. Moreover, the in-house ELISA could detect murine FGF18 in sera from mice overexpressing murine Fgf18 in hepatocytes, although the sensitivity in detecting murine FGF18 was relatively low. This FGF18 ELISA could be a valuable tool to validate FGF18 as a potential biomarker for cancer patients and to test the contribution of FGF18 for various disease models invivo and in vitro.

Structural basis for antigen recognition by methylated lysine-specific antibodies.

Proteins are modulated by a variety of posttranslational modifications including methylation. Despite its importance, the majority of protein methylation modifications discovered by mass spectrometric analyses are functionally uncharacterized, partly owing to the difficulty in obtaining reliable methylsite-specific antibodies. To elucidate how functional methylsite-specific antibodies recognize the antigens and lead to the development of a novel method to create such antibodies, we use an immunized library paired with phage display to create rabbit monoclonal antibodies recognizing trimethylated Lys260 of MAP3K2 as a representative substrate. We isolated several methylsite-specific antibodies that contained unique complementarity determining region sequence. We characterized the mode of antigen recognition by each of these antibodies using structural and biophysical analyses, revealing the molecular details, such as binding affinity toward methylated/nonmethylated antigens and structural motif that is responsible for recognition of the methylated lysine residue, by which each antibody recognized the target antigen. In addition, the comparison with the results of Western blotting analysis suggests a critical antigen recognition mode to generate cross-reactivity to protein and peptide antigen of the antibodies. Computational simulations effectively recapitulated our biophysical data, capturing the antibodies of differing affinity and specificity. Our exhaustive characterization provides molecular architectures of functional methylsite-specific antibodies and thus should contribute to the development of a general method to generate functional methylsite-specific antibodies by de novo design.