Interactions of Alpha-fetoprotein and Other Proteins with Concanavalin A

The chromatography of alpha-fetoprotein (AFP) containing sera on concanavalin A-Sepharose 4 B immunoadsorbant (Con A) yields two fractions, one which does and the other which does not bind the Con A. It is thus possible to calculate for various specimens the ratios Mean ratios of three categories of sera whose AFP titres varied between 14,000 and 430,000 IU/ml were the following: eight cord sera, m = 19.1 ± 4.2%; seven primary liver carcinoma sera (PLC), m = 15.9 ± 2.4%; three teratoblastoma sera, m = 37.5 ± 5.5%. The comparison of mean values by the Wilcoxon test indicated that there were statistically significant differences (p<0.05) between cord sera and teratoblastoma sera on the one hand, and between PLC and teratoblastoma sera on the other hand. It was also found that the ratios did not depend upon levels of AFP in sera. Variations of serum glycoproteins such as those reported in cancer patients are unlikely to account for these differences, especially since they are of much smaller magnitude than those achieved in in vitro experiments. Establishing the existence of patterns of heterogeneity of AFP linked to the origin of serum specimens will however require, in order to eliminate possible interferences by other glycoproteins, the use of purified AFPs. Should the existence of such patterns be confirmed, it will be interesting to see if they are related to differences such as those recently reported between fetal and tumour AFP when inhibition of lymphocyte stimulation was used.