Yasuaki Ura

Studies on circulating antibody against carcinoembryonic antigen (CEA) and CEA-like antigen in cancer patients

Characteristics of auto-antibodies for carcinoembryonic antigen (CEA) detected in sera from 3 cancer patients (2 colorectal and 1 breast cancer) were examined. The antibodies belonged to polyclonal immunoglobulin G (IgG). The binding of auto-antibodies with the labeled CEA was inhibited by not only the unlabeled CEA but also NCA-2 (feces and meconium). However, no binding with NCA was observed. Among these auto-antibodies the antibody directed against blood group Lewis determinants which are known to be present in many purified CEA preparations was not found. Previously we had suggested that CEA, NCA-2 and NCA may contain immune determinant in common with alpha 1-acid glycoprotein (AG). These auto-antibodies showed significantly enhanced reactivity for the labeled CEA preparation after purification by anti-AG affinity chromatography in spite of no immunological reaction with AG. These results suggest that auto-antibodies are raised against the common antigenic determinants of both CEA and NCA-2 which do not exist in NCA. These antibodies might be directed to common amino acid sequence shared by CEA and NCA-2, though not excluding the carbohydrate moiety. We surveyed about 500,000 cancer patients but could find only 3 patients who showed a difference in the values of CEA by the indirect and direct method. Thus, the existence of this type auto-antibody to CEA in cancer patients is a rare phenomenon.

Immunological study of tissue polypeptide antigen (TPA)—demonstration of keratin-like sites and blood group antigen-like sites on TPA molecules

We examined the immunological cross-reactivity between tissue polypeptide antigen (TPA) and keratin protein because of the reported sequence homology between these two proteins. TPA showed positive immuno-reactivity against both polyclonal and monoclonal antibodies for keratin. The binding of [125I]TPA with anti-keratin could be displaced dose-dependently by unlabeled keratin. However, no cross-reaction between keratin and anti-TPA was found. TPA also showed the positive immuno-reactivity with antibodies for blood group antigens (A, B and Lewis substances), but keratin did not. A standard solution of Lewis substance reacted with neither anti-TPA nor with anti-keratin. These data strongly suggest that TPA has an immunological similarity with both keratin protein and blood group antigens. When [125I]TPA and [125I]keratin were gel-filtered on a Sephadex G-200 column, the radioactivities of TPA and keratin were found mainly in the void volume fraction (MW greater than 200 000) and the MW of approximately 60 000, respectively. Chromatography on Sepharose 6B suggested that the MW of [125I]TPA was 320 000. When sera of cancer patients were gel-filtered on a Sephadex G-200 column, TPA activity was distributed mainly in the void volume fraction in all tested cases. This experiment suggests that TPA may be a glycoprotein (MW is 320 000) with both keratin-like and blood group antigen-like determinants.

Studies of Calmodulin-Like Portion in the TSH Receptor

Calmodulin (CaM) antagonists, such as W–7, W–5 chlorpromazine, and haloperidol, inhibited dose-dependently 125I–bTSH binding to its receptor. This inhibitory effect by CaM antagonist was diminished by the addition of EDTA. Not only anti–CaM antibody but also CaM inhibited dose–dependently 1251–bTSH binding to its receptor. These results may indicate the presence of a CaM–like structure in the membrane receptor for TSH.