Quantitative estimation of cellular retinoic acid-binding protein activity in normal, dysplastic, and neoplastic human breast tissue.

A technique for reproduction and quantitative determination of human cellular retinoic acid-binding protein (CRABP) activity in breast tissue specimens is described. A multiphasic polyacrylamide disc gel electrophoresis system (operative at pH 10.2) was adapted for this purpose. This technique allows, after incubation with tritiated retinoic acid (RA) overnight, the separation of the specific CRABP activity from the nonspecific serum-originated binding activity and from the free RA. Previous purification of the tissue cytosols is therefore not necessary. The same assay method was also used for the determination of the molecular weight (Ferguson plot, m.w. 13,000) and the dissociation constant Kd (2.5 x 10(-7) M) of mammary CRABP. The activity in tissue cytosol, stored at -70 degrees, was found to be stable for at least 3 months. Results from 88 breast tissue specimens of different pathological degree are presented. CRABP activity was found in all tissue categories with progressively increasing amounts from normal tissue to breast cancer. The activity in the cancer tissues (14.85 +/- 12.05 pmol RA bound per mg soluble protein: N = 27) was significantly different (p less than 0.001) from the activity determined in tissue with simple dysplasia without epithelial proliferation [4.3 +/- 2.2 (S.D.) pmol RA bound per mg protein; N = 30]. It is possible that in the cases where high amounts of CRABP activity are found in dysplastic and preneoplastic tissue, a high risk for breast cancer development exists. Therefore, CRABP is tentatively proposed as a dedifferentiation and/or proliferation marker.