Interlaboratory reproducibility of yeast protein patterns analyzed by immobilized pH gradient two-dimensional gel electrophoresis.

Abstract

An interlaboratory comparison was conducted on the positional and quantitative reproducibility of yeast proteins resolved by two-dimensional polyacrylamide gel electrophoresis (2-D PAGE) using isoelectric focusing with immobilized pH gradient (pH 4-7) in the first dimension. The basic experimental set-up was as follows: one laboratory prepared and distributed a [35S]methionine-labeled total yeast protein extract (Göteborg, Sweden), another laboratory prepared the IPG strips to be used by all labs in this study (Munich, Germany), the third laboratory (Aarhus, Denmark) circulated the protocols and coordinated the modest attempts to unify them. Samples were run horizontally in the first dimension and vertically in the second. The gels were sent to Göteborg for processing by phosphoimager technology and computerized image analysis (PDQuest), and the 2-D PAGE resolved proteins were located and quantified automatically. A subset of 470 spots was manually matched in all gels out of an average of 1328 resolved proteins. The positional interlaboratory comparison revealed great pattern reproducibility, the correlation coefficient in no case being less than 0.9994. In absolute terms an average deviation of 2.8 mm (x-position) and 1.8 mm (y-position) were obtained for all nine gels (three gels per lab). The interlaboratory comparison of protein quantitation displayed higher variability, and the best correlation coefficient generated was 0.975. An average standard deviation of 34.5% was calculated for protein quantitation including all three labs, a value slightly higher than the intralaboratory variation (range 20-28%). Thus, despite differences in protocols, chemicals and equipment, the immobilized pH gradient technology gave extremely high positional and quantitative reproducibility. This will greatly facilitate the exchange of data and the establishment of multi-user image-based 2-D gel databases.