Identification and use of constitutive proteins for the normalization of high resolution electrophoretograms.

Quantitative inter-gel comparisons of proteins separated by high resolution two-dimensional protein electrophoresis present a number of problems. These problems may arise from: variations in pipetting and other mechanical manipulations of samples, protein loss during transfer from the first to the second gel dimension, variations in staining, and/or variations in film development during autoradiography, in the case of radioactively labeled proteins. This study presents a discussion of these issues and a normalization algorithm to deal with variations, which relies on a class of proteins present in most biological samples which by their nature may be considered internal standards. This class consists of proteins which are controlled by constitutive genes. Constitutive genes are genes that are expressed constantly. We have developed an algorithm which is currently available as a subroutine, 'FINDCONS', in the computerized densitometry and protein comparison analysis program, developed by Olson & Miller (1988). This algorithm identifies potentially 'constitutive' proteins. A normalization method employing these potentially 'constitutive' proteins was compared to several others by examining 2D-electrophoretograms of proteins from developing gypsy moths (Lymantria dispar L.) insect tissue. Following normalization, inter-gel comparisons of spots, which were 'identified' as 'constitutive', were observed to vary less in density than when no normalization method was used, or when normalization based on total integrated spot density was used. In addition to its use as a normalization tool, this algorithm and the subroutine FINDCONS may be useful as an aid in biological studies to identify 'constitutive' proteins.