No abstract

Lactoferrin (LF) and β-lactoglobulin (BLG), strongly basic and weakly acidic bovine milk proteins, form optically clear coacervates under highly limited conditions of pH, ionic strength I, total protein concentration C(P), and BLG:LF stoichiometry. At 1:1 weight ratio, the coacervate composition has the same stoichiometry as its supernatant, which along with DLS measurements is consistent with an average structure LF(BLG2)2. In contrast to coacervation involving polyelectrolytes here, coacervates only form at I < 20 mM. The range of pH at which coacervation occurs is similarly narrow, ca. 5.7-6.2. On the other hand, suppression of coacervation is observed at high C(P), similar to the behavior of some polyelectrolyte-colloid systems. It is proposed that the structural homogeneity of complexes versus coacervates with polyelectrolytes greatly reduces the entropy of coacervation (both chain configuration and counterion loss) so that a very precise balance of repulsive and attractive forces is required for phase separation of the coacervate equilibrium state. The liquid-liquid phase transition can however be obscured by the kinetics of BLG aggregation which can compete with coacervation by depletion of BLG.

Lactoferrin (LF) and β-lactoglobulin (BLG) are among the protein pairs that exhibit heteroprotein coacervation, a unique and relatively unexamined type of liquid-liquid phase separation (LLPS). In prior work we found that LF and BLG undergo coacervation at highly constrained conditions of pH, ionic strength and protein stoichiometry. The molar stoichiometry in coacervate and supernatant is LF : BLG2 1 : 2 (where BLG2 represents the 38 kDa BLG dimer), suggesting that this is the primary unit of the coacervate. The precise balance of repulsive and attractive forces among these units, thought to stabilize the coacervate, is achieved only at limited conditions of pH and I. Our purpose here is to define the process by which such structural units form, and to elucidate the forces among them that lead to the long-range order found in equilibrium coacervates. We use confocal laser scanning microscopy (CLSM), small angle neutron scattering (SANS), and rheology to (1) define the uniformity of interprotein spacing within the coacervate phase, (2) verify structural unit dimensions and spacing, and (3) rationalize bulk fluid properties in terms of inter-unit forces. Electrostatic modeling is used in concert with SANS to develop a molecular model for the primary unit of the coacervate that accounts for bulk viscoelastic properties. Modeling suggests that the charge anisotropies of the two proteins stabilize the dipole-like LF(BLG2)2 primary unit, while assembly of these dipoles into higher order equilibrium structures governs the macroscopic properties of the coacervate.

The aggregation of β-lactoglobulin (BLG) near its isoelectric point was studied as a function of ionic strength and pH. We compared the behavior of native BLG with those of its two isoforms, BLG-A and BLG-B, and with that of a protein with a very similar pI, bovine serum albumin (BSA). Rates of aggregation were obtained through a highly precise and convenient pH/turbidimetric titration that measures transmittance to ±0.05 %T. A comparison of BLG and BSA suggests that the difference between pHmax (the pH of the maximum aggregation rate) and pI is systematically related to the nature of protein charge asymmetry, as further supported by the effect of localized charge density on the dramatically different aggregation rates of the two BLG isoforms. Kinetic measurements including very short time periods show well-differentiated first and second steps. BLG was analyzed by light scattering under conditions corresponding to maxima in the first and second steps. Dynamic light scattering (DLS) was used to monitor the kinetics, and static light scattering (SLS) was used to evaluate the aggregate structure fractal dimensions at different quench points. The rate of the first step is relatively symmetrical around pHmax and is attributed to the local charges within the negative domain of the free protein. In contrast, the remarkably linear pH dependence of the second step is related to the uniform reduction in global protein charge with increasing pH below pI, accompanied by an attractive force due to surface charge fluctuations.