THE SIEVING OF ROD-SHAPED VIRUSES DURING AGAROSE-GEL ELECTROPHORESIS .1. COMPARISON WITH THE SIEVING OF SPHERES

The sieving of rod‐shaped viruses during agarose gel electrophoresis is quantitatively analyzed here with a previously proposed model [G. A. Griess et al. (1989) Biopolymers, 28, 1475–1484] that has one radius (PE) of the effective pore at each concentration of gel. By use of this model and an internal spherical size standard, a plot of electrophoretic mobility vs agarose percentage is converted to a plot of the radius of the effective sphere (effective radius) vs PE. Experimentally, when the concentration of the rod‐shaped bacteriophage, fd, is progressively increased, eventually the electrophoretic mobility of fd becomes dependent on its concentration. The concentration of fd at which this occurs decreases as the agarose concentration decreases. After avoiding this dependence on the concentration of sample, the effective radius of rod‐shaped particles, including bacteriophage fd, length variants of fd, and length variants of tobacco mosaic virus, is found to increase as PE increases until a plateau of approximately constant maximum effective radius is reached at P cE. In the region of this plateau, the effective sphere's measure that best approximates that of the rod is surface area. However, significant disagreement with the data exists for surface area; the maximum effective radius for fd varies as (length)0.69. For fd and its length variants, the value of 2·P cE/length increases from 0.21 to 0.86 as the length decreases from 2808 to 367 nm. The dependence of effective radius on PE and the proximity of 2·P cE to the length of the rod are explained by (a) random orientation of rods at PE values in the region of the plateau, and (b) increasingly preferential end‐first orientation (reptation) of the rod as PE decreases below P cE. This hypothesis of reptation is supported by a significant dependence of electrophoretic mobility on electrical potential gradient for a PE below, but not above, P cE. The dependence of 2·P cE/length on length is not rigorously understood, but is qualitatively explained by flexibility of the rods. This apparent flexibility has thus far prevented determination of a rod's axial ratio from quantitation of sieving during agarose gel electrophoresis. The electrical potential dependence of electrophoretic mobility is determined here by a procedure of two‐dimensional agarose gel electrophoresis. This procedure is also useful for detecting rod‐shaped particles in heterogeneous mixtures of predominantly spherical particles. Copyright © 1990 John Wiley & Sons, Inc.