Estimation of the molecular weights of proteins by Sephadex gel-filtration

Publisher Website
Google Scholar
Abstract
Correlation between elution volumes, Ve, and molecular weight was investigated for gel-filtration of proteins of known molecular weight on Sephadex G-75 and G-100 columns at pH 7.5. Graphs of Ve against log (molecular weight) were linear over the molecular-weight ranges 3000-35000 and 5000-60000 for Sephadex G-75 and G-100 respectively. Within these limits the gels give optimum separation of proteins according to molecular weight, but their useful working ranges extend to about 70000 and 150000 respectively. The lower molecular-weight limits for complete exclusion of proteins from Sephadex G-75 and G-100 were estimated as 110000 and 300000 respectively. The relationship between Ve and molecular weight was similar over the pH range 1.3-10.7. The use of Sephadex G-100 columns for estimating the molecular weights of proteins was described; 10 [mu]g. of protein with enzymic acitivity, or 100 [mu]g. of non-enzymic protein, may be sufficient for an estimation. Purification of the protein is unnecessary if it can be estimated in column effluents by a specific method. Molecular weights of enzymes estimated in this way were in agreement with literature values. In cases where molecular weights determined by other methods have not been published, Sephadex-gel-filtration estimates were: adenosine deaminase, 34000; yeast glucose 6-phosphate dehydrogenase, 110000; A. aerogenes glycerol dehydrogenase, 120000; milk alkaline phos-phatase, 148 000. Bovine haemoglobin showed very marked concentration-dependent dissociation on Sephadex G-100 columns, and dissociation of [beta]-lactoglobulin A into half-molecules also occurred. Good correlation was observed between distances migrated and molecular weights of proteins in thin-layer gel-filtration with Sephadex G-100. Potential sources of error in molecular-weight estimation by Sephadex gel-filtration were discussed. The maximum uncertainty in values estimated by the column procedure for globular proteins in the range 10000-150000 was put at [plus or minus] 10%. For other types of protein the possibility is that values will be less reliable, if globular proteins are used for calibration of the columns.