1. A viscometer/rheometer capable of automatically measuring the dependence of viscosity of solutions and other macromolecules upon both concentration and shear rate was designed and constructed. The instrument was intensively validated by comparison of results obtained on solutions of small viscogens (glycerol, sucrose), synthetic polymers, and proteins, and published results in the literature, over a wide range of concentrations and viscosities. A U.S. government patent on this device has been applied for (A. Grupi &A. Minton). 2. The results of previously published measurements of the dependence of the viscosity of binary mixtures of monoclonal antibodies over a wide range of composition and total protein concentration have been successfully modeled using generalized forms of two semi-empirical equations, the extended Mooney equation and the Krieger-Dougherty equation. A further generalization to solutions of self- and/or hetero-associating proteins was proposed. The instrument outlined in part 1 above is currently being used to test the applicability of the proposed generalized relations to mixtures of other proteins, including proteins that are known to associate. 3. Our lab is currently developing a hybrid theoretical and numerical model for aggregate statistics of protein solutions as a function of ionic strength and pH. The goal is to create a coarse-grained description, a qualitative model that captures the fundamental properties for a wide class of systems. Our simulations concentrate on a coarse-grained field of inter-protein electrostatic interaction. There are numerous ways of doing this, all involving different levels of discretization and approximation to the fields and forces involved. We calculate the electrostatic field around a protein via numerical solution of the non-linear Poisson-Boltzmann equation. We attempt to match this field as closely as possible using a (small) set of discretized """"""""macro-charges"""""""", which vary with each ionic strength and pH. The hypothesis is that these """"""""toys"""""""" will give a rough idea of the thermodynamic quantities of a protein solution in bulk, pointing the way toward more detailed experiments both in vitro and in silico. Many typical treatments for a large number of particles consider only a isotropic contribution. By modeling the interprotein interactions using anisotropic electric fields we hope to capture more realism without sacrificing computational efficiency. Once the field has been approximated we will run a battery of Monte-Carlo experiments to determine thermodynamic variables (such as the average energy and the specific heat). Structural parameters such as the radial distribution function will be computed and used to calculate experimental properties such as light and X-ray scattering for comparison with experiment (Hoppe).
Showing the most recent 10 out of 12 publications
