Background

Gel filtration separates molecules of different sizes and shapes based on their relative abilities to penetrate a bed of porous beads with well-defined pore sizes, which identifies the fractionation range. Molecules larger than the fractionation range, which are completely excluded from entering the pores, flow quickly through the column and elute first at the void volume (V₀), which is the interstitial volume outside the support particles. Molecules smaller than the fractionation range, which are able to diffuse into the pores of the beads, have access to the total volume available to the mobile phase, therefore they move through the bed more slowly and elute last. Molecules with intermediate dimensions will be eluted with an elution volume (V_e) comprised between the void volume and the total volume available to the mobile phase (the smaller the molecule, the greater its access to the pores of the matrix, and thus the greater is its V_e).

The molecular weight of a protein can be determined by comparison of its elution volume parameter K_d, which represents the distribution of a given solute between the stationary and mobile phases (see Data analysis below), with those of different known calibration standards.

If the protein of interest has the same shape (generally globular) as the standard calibration proteins, the gel filtration experiment provides a good estimate of its molecular weight. However, because the shape of proteins can vary significantly and may be not known for an unknown protein, care must be taken in the determination of molecular size from elution volume. For example, a protein with an elongated shape could elute at a position that does not correspond to its dimension and which is significantly different from the position of a spherical protein having the same molecular weight. This is the case for some Ca²⁺ sensor proteins, e.g., calmodulins from different organisms (Sorensen and Shea, 1996; Sorensen et al., 2001; Astegno et al., 2014; Astegno et al., 2016; Vallone et al., 2016), which have anomalous migrations in gel filtration, resulting in a defined overestimation of the molecular weight due to their highly extended conformation. Thus, it is clear that in a gel filtration experiment the elution profile of proteins is closer to their Stokes radius (R_s) rather than to their molecular weight. R_s is a hydrodynamic value indicative of the apparent size of the dynamic solvated/hydrated protein.

For this reason, a SEC-based approach has been employed to resolve Ca²⁺-induced changes in the hydrated shape of Ca²⁺ sensor proteins by determination of their R_s in apo- and Ca²⁺-bound conditions. Ca²⁺ binding usually causes a decrease in the R_s (Sorensen and Shea, 1996; Sorensen et al., 2001; Astegno et al., 2016). The same SEC-based approach may be applicable to the detection of other protein-small molecule (e.g., other metals) interactions that cause changes in the structure of the protein to a less or more extended conformation (Asante-Appiah and Skalka, 1997; Bagai et al., 2007; De Angelis et al., 2010).

The values of Rs have been reported for a large number of proteins; in particular, some proteins are especially convenient for calibration of gel filtration columns (le Maire et al., 1986; Uversky, 1993) (Table 1). A gel filtration column can determine the hydrodynamic size, R_s, of a sample protein by comparison with the R_s of these water-soluble calibration proteins.

Table 1. Standards for calibrating analytical gel filtration