Globular protein gelation - theory and experiment

Heat-set globular protein gel networks are discussed in relation to protein charge and screening. Homogeneous 'fine-stranded' gels form when electrostatic repulsion is high, with a transition to microphase-separated structures as repulsion falls. Fine-stranded networks showing density fluctuations occur close to this transition and probably arise from a combination of kinetic trapping and a drive towards only limited phase separation. For the most uniform structures (pH far from pl, and low salt content) network building appears to involve three main stages: initial protein unfolding, linear fibrillar aggregation, and random cross-linking of the fibrils. A mean field model is described which incorporates these features and includes the possibility of cooperative linear aggregation (nucleation and growth). Application of this to cure data for acid beta -lactoglobulin gels was only partially successful, however, a higher order (n = 4) nucleation process being required to explain gel time-concentration data, while only a lower second order process could reproduce the shapes of the cure curves. As uniform gels give way to phase-separated structures network building becomes still more complex. Here solution demixing of unfolded monomers, and/or the initial aggregates, must be included in the model. This seems beyond the current mean-field approach, and simulation is likely to be required. This is true, even for the more homogeneous structures, when gel properties of interest extend from the linear elastic, to the time-dependent, and non-linear. (C) 2001 Elsevier Science Ltd. All rights reserved.