Session A40: Focus Session: Protein Association I: Phase Separation, Crystallization, Self-Assembly

Phase Separation in Solutions of Monoclonal Antibodies

We report the observation of liquid-liquid phase separation (LLPS) in a solution of humanized monoclonal antibodies, IgG2, and the effects of human serum albumin, a major blood protein, on this phase separation. We find a significant reduction of phase separation temperature in the presence of albumin, and a preferential partitioning of the albumin into the antibody-rich phase. We provide a general thermodynamic analysis of the antibody-albumin mixture phase diagram and relate its features to the magnitude of the effective inter-protein interactions. Our analysis suggests that additives (HSA in this report), which have moderate attraction with antibody molecules, may be used to forestall undesirable protein condensation in antibody solutions. Our findings are relevant to understanding the stability of pharmaceutical solutions of antibodies and the mechanisms of cryoglobulinemia.   

Concentrated dispersions of equilibrium protein nanoclusters that reversibly dissociate into active monomers

Stabilizing concentrated protein solutions is of wide interest in drug delivery. However, a major challenge is how to reliably formulate concentrated, low viscosity (i.e., syringeable) solutions of biologically active proteins. Unfortunately, proteins typically undergo irreversible aggregation at intermediate concentrations of 100-200 mg/ml. In this talk, I describe how they can effectively avoid these intermediate concentrations by reversibly assembling into nanoclusters. Nanocluster assembly is achieved by balancing short-ranged, cosolute-induced attractions with weak, longer-ranger electrostatic repulsions near the isoelectric point. Theory predicts that native proteins are stabilized by a self-crowding mechanism within the concentrated environment of the nanoclusters, while weak cluster-cluster interactions can result in colloidally-stable dispersions with moderate viscosities. I present experimental results where this strategy is used to create concentrated antibody dispersions (up to 260 mg/ml) comprising nanoclusters of proteins [monoclonal antibody 1B7, polyclonal sheep Immunoglobin G and bovine serum albumin], which upon dilution in vitro or administration in vivo, are conformationally stable and retain activity.   

Aggregation behavior of bovine and porcine insulin in presence of inhibitors

Insulin aggregation can be problematic when the insulin is in pharmaceutical solution, as well as when aggregates form in insulin pumps or subcutaneously at port sites. Pharmaceutical insulins often include an added stabilizer, such as metacresol, but even with this added compound the protein may degrade, aggregate and become less effective. With greater understanding of the kinetics of aggregation associated with insulin aggregation, a more effective stabilizer may be identified to inhibit aggregation. Bovine and porcine insulin in pharmaceutically relevant concentrations of ~5 mg/ml were induced to denature and aggregate at elevated temperature and low pH and the aggregation behavior was characterized as a function of time and temperature by rheology and small angle x-ray scattering (SAXS). Aggregation behavior was probed in both neat solution and with the addition of known inhibiting compounds. The efficacy of the inhibiting compounds at retarding the protein aggregation was found to be related to the hydrophobicity of the compounds and potential inhibitors of interest were identified.   

Design rules for the self-assembly of a protein crystal

The need to crystallize proteins for X-ray studies has motivated theories of protein crystallization. These theories, based largely on the behavior of isotropic spheres that form close-packed crystals, predict that assembly is enhanced near the metastable critical point associated with phase separating into a vapor of proteins (protein-poor solution) and a liquid of proteins (protein-rich solution). However, most protein crystals are open structures stabilized by anisotropic interactions, and many assemble best above the critical point. Using theory and simulation, we show that optimal assembly of one such crystal, a model surface-layer protein crystal, is not predicted by the critical point but can be predicted by a combination of two design rules: the thermodynamic driving force must be on the order of the thermal energy, and interactions must be made as nonspecific as possible without promoting liquid-vapor phase separation. In experimental terms, our results suggest adjusting solution conditions in order to impose a defined supersaturation at the liquid-vapor critical point. Our findings suggest that self-assembly of open crystals is more akin to viral capsid self-assembly than to the crystallization of spherical colloids.   

Protein Crystal Nucleation and Growth

We have developed a microfluidic emulsion based technique to determine the homogeneous and heterogeneous nucleation rates of protein crystallization under conditions of high supersaturation. We utilize the fact that the nucleation rate is constant if no crystal nucleus is formed and count the number of protein droplets with no crystals with time, which decays exponentially with decay constant inversely proportional to nucleation rate and drop volume. We report results of experiments on nucleation and growth rates of lysozyme crystallization. The emulsions are placed on a temperature gradient stage allowing simultaneous measurement of rates as a function of temperature. We routinely scan 30,000 drops in each experiment.   

Water dynamics during the association of hiv capsid proteins studied by all-atom simulations

The C-terminal domain of the HIV-1 capsid protein (CA-C) plays an important role in the assembly of the mature capsid. We have used molecular dynamics simulations combined with enhanced sampling methods to study the association of two CA-C proteins in atomistic detail. In this talk we will discuss the dynamics of water during the association process. In particular, we will show that that water in the interfacial region does not undergo a liquid-vapor transition (de-wetting) during association of wild type CA-C. However, mutation of some hydrophilic residues does lead to a dewetting transition. We discuss the relationship between the arrangement of hydrophilic and hydrophobic residues and dewetting during protein association. For the HIV capsid protein, the arrangement of hydrophilic residues contributes to maintaining weak interactions, which are crucial for successful assembly.   

Interfacial Microrheology of Lysozyme Layers During Formation at the Air-Water Interface

Proteins can adsorb to the air-water interface to form viscoelastic layers. Characterizing the rheology of such layers is challenging, due to the confined geometry, the fragility of the layers, and the possibility of mesoscale spatial heterogeneity. Passive microrheology --- using the thermal motion of colloidal probes to interrogate the mechanical response of the surrounding medium --- is a suitable technique for addressing these difficulties. In particular, this approach sheds light on the properties of incipient protein layers that are characterized by modest interfacial viscosities. We describe microrheology studies of lysozyme layers at the air-water interface, in which we determine the evolving interfacial shear response through the viscoelastic transition that signifies layer formation. Spatial heterogeneity in the interfacial rheology is identified and discussed within the framework of layer formation as a gel transition. Layers formed by adsorption of protein from the aqueous subphase and by spreading protein directly onto the interface are compared and studied across a range of concentrations, demonstrating the sensitivity of layer properties to the rate and manner of protein accretion.   

Measurements of Diffusion within Concentrated Bovine $\alpha$-Crystallin Suspensions

$\alpha -$Crystallin is a major protein component of the vertebrate eye lens. The chaperone-like behavior of these water soluble proteins play a key role in maintaining lens transparency by preventing condensation of other lens proteins. We report photon correlation spectroscopy measurements, both X-ray Photon Correlation Spectroscopy (XPCS) and Dynamic Light Scattering (DLS), indicating protein diffusion within suspensions of $\alpha $-Crystallin. Measurements were carried out at 2$^{\circ}$C, 10$^{\circ}$C and 35$^{\circ}$C, over a wide range of concentrations from the diluted limit to the regime close to the physiological lens concentration. In the diluted regime, DLS measurements can be modeled by a single exponential fit indicating a single relaxation mode and at higher concentrations two relaxation modes can be identified by fitting the data to a double exponential decay function, a clear indication of the ploydispersed nature of the concentrated samples. XPCS measurements show dynamics at the highest concentration but cannot resolve the faster dynamics (below 20ms) at lower concentration. We also provide estimates for the viscosity of $\alpha $-Crystallin suspensions as a function of temperature and protein volume fraction using the falling ball method.   

Master Equation Approach to Protein Assembly -- Degradation of Protein Aggregates

Protein aggregation is dependent on several microscopic processes such as nucleation, elongation and fragmentation of aggregates. Applying simple chemical kinetics to these processes allows one to derive master equations describing the entire system. In almost all cases they take the form of highly non-linear coupled differential equations for which no exact analytical solutions can be derived. Nonetheless an analytical description of the problem is absolutely essential to determine the relative importance of different microscopic processes and develop a rational approach to finding cures. Using a self consistent approach, my group has recently made headway in finding approximate analytical solutions for several systems and successfully applied them to explain a wide range of experimental observations (Knowles et al., Science 326, 1533 (2009)). I have generalised this description to include degradation of aggregates by various cellular processes. These degradation processes are thought to play an important role in vivo in determining when aggregation speed becomes critical and leads to disease.   

Intermediate-range order in protein suspensions

Intermediate-range order (IRO) has been widely observed in vitreous silica, water ice, metallic glass, and even in ionic liquids. Our recent work demonstrates that there is IRO present in a colloidal suspension, such as protein solution, when both a short-range attraction and long-range repulsion are present. We have verified this experimentally using lysozyme solutions, where a peak (IRO peak) seen in small angle neutron scattering (SANS) has been mistakenly called a cluster peak as it has once been considered an indication of a cluster rich phase in solution. By combining both SANS and neutron spin echo (NSE) techniques, we clearly show that there is no direct relation between cluster formation and the presence of an IRO peak. By investigating the short time dynamics using NSE, we show that the formation of clusters is still indeed possible at high concentrations.   

An energy landscape approach to protein aggregation

Protein aggregation into ordered fibrillar structures is the hallmark of a class of diseases, the most prominent examples of which are Alzheimer's and Parkinson's disease. Recent results (e.g. Baldwin et al. J. Am. Chem. Soc. 2011) suggest that the aggregated state of a protein is in many cases thermodynamically more stable than the soluble state. Therefore the solubility of proteins in a cellular context appears to be to a large extent under kinetic control. Here, we first present a conceptual framework for the description of protein aggregation ( see AK Buell et al., Phys. Rev. Lett. 2010) that is an extension to the generally accepted energy landscape model for protein folding. Then we apply this model to analyse and interpret a large set of experimental data on the kinetics of protein aggregation, acquired mainly with a novel biosensing approach (see TPJK Knowles et al, Proc. Nat. Acad. Sc. 2007). We show how for example the effect of sequence modifications on the kinetics and thermodynamics of human lysozyme aggregation can be understood and quantified (see AK Buell et al., J. Am. Chem. Soc. 2011). These results have important implications for therapeutic strategies against protein aggregation disorders, in this case lysozyme systemic amyloidosis.