Precision of sodium dodecyl sulfate-polyacrylamide-gel electrophoresis for the molecular weight estimation of a membrane glycoprotein: studies on bovine rhodopsin.

Criteria for assessing the precision and accuracy of methods for estimation of molecular weight for proteins using sodium dodecyl sulfate-polyacrylamide-gel electrophoresis have been applied to rhodopsin from bovine visual cell outer segment membranes. Various methods of preparing this hydrophobic protein for electrophoresis differ in their ability to solubilize and disaggregate polypeptide constituents of the outer segment membrane, with resultant variations in the pattern of protein bands and the apparent molecular weight of rhodopsin. Even with optimal solubilization and disaggregation, the behavior of rhodopsin relative to a series of standard proteins is such that the apparent molecular weight decreases systematically from 40,400 to 34,500 as the acrylamide concentration increases from 4 to 10%. As demonstrated by Ferguson plots of logR_f vs gel concentration and split gel experiments, this discrepancy is explained by the fact that the extrapolated R_f for zero gel concentration (Y₀) for rhodopsin is significantly lower than the Y₀'s for the soluble proteins used as molecular weight standards. In such cases, a possibly more reliable molecular weight estimate is obtained by plotting the retardation coefficient (K_R) vs molecular weight. This method yields a value of 29,500 ± 1000 for bovine rhodopsin if only the errors in measurement of R_f are considered and a quadratic relationship between K_R and molecular weight is used. Using weighted linear regression for K_R vs molecular weight, we obtain a molecular weight estimate of 32,700 ± 5000 when the uncertainty in the calibration curve is considered. Because of uncertainties regarding the detergent-binding properties of rhodopsin and the relationship of its Stokes radius to its molecular weight by comparison with the soluble protein standards, these values must be viewed with caution.