Avidin-biotin affinity chromatography: application to the isolation of human placental insulin receptor.

Abstract
The ligand N.alpha.,B1-(6-biotinylamido)hexanoyl-insulin was attached noncovalently to Sepharose 4B immobilized succinoylavidin to form an insulin-affinity resin. This resin was used to isolate highly purified insulin receptor from human placental tissue by a 4-step process involving preparation of a crude membrane fraction, solubilization with Triton X-100, wheat germ agglutinin purification and insulin-affinity chromatography. NaDodSO4/PAGE [polyacrylamide gel electrophoresis] of the purified 125I-labeled receptor under nonreducing conditions showed the presence of a major component with a MW of .apprxeq. 350,000 and a minor component with a MW of .apprxeq. 166,000. Based on the assumption that the degree of labeling is comparable in both comonents, the material corresponding to the MW 350,000 peak reprsents .apprxeq. 94% of the receptor preparation as determined by scanning the autoradiograms. The specific insulin binding capacity of the preparation is 18 .+-. 6 .mu.g of 125I-labeled insulin/mg of protein as determined by the polyethylene glycol assay and analyzed by Scatchard plot. Insulin binding activity was stable at 4.degree. C and pH 7.6 for at least 12 wk but was destroyed by freezing and thawing. The availability of highly purified receptor afforded the opportunity to explore its precipitabiity by polyethylene glycol under assay conditions. Whereas trichloroacetic acid precipitated 95% of the 125I-labeled receptor, polyethylene glycol precipitated only 30%. If the specific activity of the receptor is corrected for incomplete precipitability by polyethylene glycol, the apparent specific binding would be 3.5 .+-. 1.2 mol of insulin/mol of receptor. These results are in disagreement with the current receptor model, which postulates that 1 mol of receptor (MW 350,000) binds 2 mol of insulin. Clearly, the problems associated with method available for determining insulin binding are sufficiently serious to preclude their use in determining receptor valence.

This publication has 21 references indexed in Scilit: