Comparison of humanized, chimeric, and mouse JAA-F11 anti-Thomsen-Friedenreich Antigen antibodies
MetadataShow full item record
The Thomsen-Friedenreich Antigen (TF-Ag), the disaccharide Galβ1-3 GalNAcα, that is present on the surface of more than 90% of carcinomas including breast cancers, is a potential cancer therapeutic target. The pancarcinoma expression of TF-Ag, combined with the evidence of a mechanistic role for TF-Ag in metastasis, show that this target would have clinical utility. JAA-F11, a mouse monoclonal antibody, with discriminating specificity for the TF-Ag alpha-linked tumor marker, has been shown to decrease metastasis and increase survival in an in vivo mouse breast cancer model. However, mouse antibodies in human therapy have been shown to be immunogenic and to have short half-lives, limiting their translation to the clinic. To address this problem, we hypothesize that utilizing complementarity determining regions (CDRs) defined by Kabat and Chothia methods, along with the crystal structure and computational carbohydrate threading data and framework regions selected by three different modified homology approaches, will result in a humanized IgG1 variant of JAA-F11 that will have decreased immunogenicity, while retaining the important discriminating biological and chemical specificity, relative to chimeric IgG1 or to the parental mouse JAA-F11. Using CDR-grafting, we have created four humanized JAA-F11 antibodies which bind to TF-Ag which are of low immunogenicity when tested by the T20 scoring method. Using a glycan array containing over 600 saccharides, the humanized JAA-F11 antibodies were shown to maintain or have improved the chemical specificity compared to the mouse JAA-F11, binding the TF-Ag alpha-linked structures and not bind TF-Ag beta-linked structures, or common elongations of either sialic acid or N-acetyl glucosamine at the 3-hydroxyl of the galactose seen on normal tissue. The humanized antibodies maintained or have improved affinity for TF-Ag compared to the mouse JAA-F11. Different humanized JAA-F11 antibodies internalize as well as the mouse JAA-F11 antibody into a breast tumor cell line, indicating potential as an antibody-drug conjugate, which can deliver an antibody bound toxin to selectively kill cancer cells. Two humanized JAA-F11 induced antibody directed cellular cytotoxicity (ADCC) which is not induced by the mouse JAA-F11, indicating potential of the humanized antibodies as direct cancer therapeutics. Similar to the mouse JAA-F11, none of the humanized antibodies induced complement dependent cellular cytotoxicity (CDC). Similar to the mouse antibody, the humanized variants bind TF-Ag on human breast cancer cell lines, including triple-negative cell lines, indicating the potential to treat all types of breast cancers, regardless of receptor status. Unlike some anti-TF-Ag antibodies made by others, humanized JAA-F11 and chimeric antibodies also do not cause proliferation of cancer cells in vitro , similar to the mouse JAA-F11. Together, the results support the conclusion that humanized IgG1 variants of JAA-F11 have decreased immunogenicity, while retaining the important discriminating biological and chemical specificity for TF-Ag, relative to chimeric IgG1 or to the parental mouse JAA-F11 antibody. The results also demonstrate that humanized antibodies to the TF-Ag have the potential to provide an adjunct therapy to treat cancer, including triple negative breast cancer which currently has no targeted therapy.