RESUMEN
There has been a remarkable increase in the number of biologics, especially monoclonal antibodies, in the market over the last decade. In addition to attaining the desired binding to their targets, a crucial aspect is the 'developability' of these drugs, which includes several desirable properties such as high solubility, low viscosity and aggregation, physico-chemical stability, low immunogenicity and low poly-specificity. The lack of any of these desirable properties can lead to significant hurdles in advancing them to the clinic and are often discovered only during late stages of drug development. Hence, in silico methods for early detection of these properties, particularly the ones that affect aggregation and solubility in the earlier stages can be highly beneficial. We have developed a computational framework based on a large and diverse set of protein specific descriptors that is ideal for making liability predictions using a QSPR (quantitative structure-property relationship) approach. This set offers a high degree of feature diversity that may coarsely be classified based on (1) sequence (2) structure and (3) surface patches. We assess the sensitivity and applicability of these descriptors in four dedicated case studies that are believed to be representative of biophysical characterizations commonly employed during the development process of a biologics drug candidate. In addition to data sets obtained from public sources, we have validated the descriptors on novel experimental data sets in order to address antibody developability and to generate prospective predictions on Adnectins. The results show that the descriptors are well suited to assist in the improvement of protein properties of systems that exhibit poor solubility or aggregation.
Asunto(s)
Productos Biológicos , Desarrollo de Medicamentos , Estudios Prospectivos , Relación Estructura-Actividad Cuantitativa , SolubilidadRESUMEN
Structural heterogeneity often constrains the characterization of aggregating proteins to indirect or low-resolution methods, obscuring mechanistic details of association. Here, we report progress in understanding the aggregation of Adnectins, engineered binding proteins with an immunoglobulin-like fold. We rationally design Adnectin solubility and measure amide hydrogen/deuterium exchange (HDX) under conditions that permit transient protein self-association. Protein-protein binding commonly slows rates of HDX; in contrast, we find that Adnectin association may induce faster HDX for certain amides, particularly in the C-terminal ß-strand. In aggregation-prone proteins, we identify a pattern of very different rates of amide HDX for residues linked by reciprocal hydrogen bonds in the native structure. These results may be explained by local loss of native structure and formation of an inter-protein interface. Amide HDX induced by self-association, detected here by deliberate modulation of propensity for such interactions, may be a general phenomenon with the potential to expose mechanisms of aggregation by diverse proteins.
Asunto(s)
Amidas/química , Deuterio/química , Hidrógeno/química , Unión Proteica , Secuencia de Aminoácidos , Enlace de Hidrógeno , Modelos Moleculares , Proteínas/química , SolubilidadRESUMEN
Accurate modeling of the effects of mutations on protein stability is central to understanding and controlling proteins in myriad natural and applied contexts. Here, we reveal through rigorous quantitative analysis that stability prediction tools often favor mutations that increase stability at the expense of solubility. Moreover, while these tools may accurately identify strongly destabilizing mutations, the experimental effect of mutations predicted to stabilize is actually near neutral on average. The commonly used "classification accuracy" metric obscures this reality; accordingly, we recommend performance measures, such as the Matthews correlation coefficient (MCC). We demonstrate that an absurdly simple machine-learning algorithm-a neural network of just two neurons-unexpectedly achieves high classification accuracy, but its inadequacies are revealed by a low MCC. Despite the above limitations, making multiple mutations markedly improves the prospects for achieving a stabilization target, and modest improvements in the precision of future tools may yield disproportionate gains.
Asunto(s)
Mutación , Proteínas/química , Bases de Datos de Proteínas , Aprendizaje Automático , Modelos Moleculares , Pliegue de Proteína , Estabilidad Proteica , Proteínas/genéticaRESUMEN
Isotropic chemical shifts measured by solution nuclear magnetic resonance (NMR) spectroscopy offer extensive insights into protein structure and dynamics. Temperature dependences add a valuable dimension; notably, the temperature dependences of amide proton chemical shifts are valuable probes of hydrogen bonding, temperature-dependent loss of structure, and exchange between distinct protein conformations. Accordingly, their uses include structural analysis of both folded and disordered proteins, and determination of the effects of mutations, binding, or solution conditions on protein energetics. Fundamentally, these temperature dependences result from changes in the local magnetic environments of nuclei, but correlations with global thermodynamic parameters measured via calorimetric methods have been observed. Although the temperature dependences of amide proton and nitrogen chemical shifts are often well approximated by a linear model, deviations from linearity are also observed and may be interpreted as evidence of fast exchange between distinct conformational states. Here, we describe computational methods, accessible via the Shift-T web server, including an automated tracking algorithm that propagates initial (single temperature) 1 H15 N cross peak assignments to spectra collected over a range of temperatures. Amide proton and nitrogen temperature coefficients (slopes determined by fitting chemical shift vs. temperature data to a linear model) are subsequently calculated. Also included are methods for the detection of systematic, statistically significant deviation from linearity (curvature) in the temperature dependences of amide proton chemical shifts. The use and utility of these methods are illustrated by example, and the Shift-T web server is freely available at http://meieringlab.uwaterloo.ca/shiftt.
Asunto(s)
Biología Computacional/métodos , Proteínas/química , Calorimetría , Enlace de Hidrógeno , Espectroscopía de Resonancia Magnética , Modelos Moleculares , Conformación Proteica , Termodinámica , Navegador WebRESUMEN
Accurately predicting changes in protein stability upon amino acid substitution is a much sought after goal. Destabilizing mutations are often implicated in disease, whereas stabilizing mutations are of great value for industrial and therapeutic biotechnology. Increasing protein stability is an especially challenging task, with random substitution yielding stabilizing mutations in only â¼2% of cases. To overcome this bottleneck, computational tools that aim to predict the effect of mutations have been developed; however, achieving accuracy and consistency remains challenging. Here, we combined 11 freely available tools into a meta-predictor (meieringlab.uwaterloo.ca/stabilitypredict/). Validation against â¼600 experimental mutations indicated that our meta-predictor has improved performance over any of the individual tools. The meta-predictor was then used to recommend 10 mutations in a previously designed protein of moderate thermodynamic stability, ThreeFoil. Experimental characterization showed that four mutations increased protein stability and could be amplified through ThreeFoil's structural symmetry to yield several multiple mutants with >2-kcal/mol stabilization. By avoiding residues within functional ties, we could maintain ThreeFoil's glycan-binding capacity. Despite successfully achieving substantial stabilization, however, almost all mutations decreased protein solubility, the most common cause of protein design failure. Examination of the 600-mutation data set revealed that stabilizing mutations on the protein surface tend to increase hydrophobicity and that the individual tools favor this approach to gain stability. Thus, whereas currently available tools can increase protein stability and combining them into a meta-predictor yields enhanced reliability, improvements to the potentials/force fields underlying these tools are needed to avoid gaining protein stability at the cost of solubility.
Asunto(s)
Biología Computacional/métodos , Modelos Moleculares , Mutación Puntual , Ingeniería de Proteínas , Proteínas Recombinantes/química , Algoritmos , Sustitución de Aminoácidos , Curaduría de Datos , Bases de Datos de Proteínas , Enlace de Hidrógeno , Interacciones Hidrofóbicas e Hidrofílicas , Internet , Cinética , Aprendizaje Automático , Conformación Proteica , Pliegue de Proteína , Estabilidad Proteica , Proteínas Recombinantes/genética , Proteínas Recombinantes/metabolismo , Reproducibilidad de los Resultados , Programas Informáticos , Solubilidad , Propiedades de Superficie , TermodinámicaRESUMEN
Aggregation can be thought of as a form of protein folding in which intermolecular associations lead to the formation of large, insoluble assemblies. Various types of aggregates can be differentiated by their internal structures and gross morphologies (e.g., fibrillar or amorphous), and the ability to accurately predict the likelihood of their formation by a given polypeptide is of great practical utility in the fields of biology (including the study of disease), biotechnology, and biomaterials research. Here we review aggregation/solubility prediction methods and selected applications thereof. The development of increasingly sophisticated methods that incorporate knowledge of conformations possibly adopted by aggregating polypeptide monomers and predict the internal structure of aggregates is improving the accuracy of the predictions and continually expanding the range of applications.
Asunto(s)
Proteínas/química , Secuencia de Aminoácidos , Humanos , Modelos Moleculares , Estructura Cuaternaria de Proteína , Estructura Terciaria de Proteína , SolubilidadRESUMEN
Protein design is still a challenging undertaking, often requiring multiple attempts or iterations for success. Typically, the source of failure is unclear, and scoring metrics appear similar between successful and failed cases. Nevertheless, the use of sequence statistics, modularity and symmetry from natural proteins, combined with computational design both at the coarse-grained and atomistic levels is propelling a new wave of design efforts to success. Here we highlight recent examples of design, showing how the wealth of natural protein sequence and topology data may be leveraged to reduce the search space and increase the likelihood of achieving desired outcomes.
Asunto(s)
Ingeniería de Proteínas/métodos , Proteínas/química , Proteínas/genética , Secuencia de Aminoácidos , Biología ComputacionalRESUMEN
The production of recombinant proteins in Escherichia coli frequently results in the formation of insoluble protein aggregates called inclusion bodies (IBs). The determinants of IB formation remain poorly understood and are of much interest for biotechnological and research applications, as well as offering insight into disease-related in vivo protein aggregation. Here we investigate a set of engineered target-binding proteins based upon the fibronectin type III domain, and we find that variations in sequence at just three positions in a solvent-exposed loop greatly alter the extent of IB formation. The loop is analogous to the third complementarity-determining region of immunoglobulin variable domains and has been shown to be conformationally mobile. In contrast to studies of other proteins, the extent of IB formation is not explained by differences in thermal stability measured by differential scanning calorimetry. Instead, IB formation is correlated with the average local stability of the FG loop, as modeled by an ensemble of structures generated using Rosetta's kinematic closure loop reconstruction method. This correlation suggests that loop instability may promote local unfolding, exposing aggregation-prone surfaces. Consistent with this mechanism, sequence-based predictions of aggregation propensity produced by Zyggregator are also correlated with IB formation, though not with modeled loop stability. The combination of average model energy scores with sequence-based aggregation predictions accounts for the variation in IB formation remarkably well (R(2)=0.8). The comparison with experimental data validates the ensemble modeling approach, which may be applicable to dynamic protein loops involved in a wide range of phenomena.
Asunto(s)
Fibronectinas/metabolismo , Agregado de Proteínas , Proteínas Recombinantes/metabolismo , Escherichia coli/metabolismo , Fibronectinas/genética , Conformación Proteica , Estabilidad Proteica , Proteínas Recombinantes/genéticaRESUMEN
The design of stable, functional proteins is difficult. Improved design requires a deeper knowledge of the molecular basis for design outcomes and properties. We previously used a bioinformatics and energy function method to design a symmetric superfold protein composed of repeating structural elements with multivalent carbohydrate-binding function, called ThreeFoil. This and similar methods have produced a notably high yield of stable proteins. Using a battery of experimental and computational analyses we show that despite its small size and lack of disulfide bonds, ThreeFoil has remarkably high kinetic stability and its folding is specifically chaperoned by carbohydrate binding. It is also extremely stable against thermal and chemical denaturation and proteolytic degradation. We demonstrate that the kinetic stability can be predicted and modeled using absolute contact order (ACO) and long-range order (LRO), as well as coarse-grained simulations; the stability arises from a topology that includes many long-range contacts which create a large and highly cooperative energy barrier for unfolding and folding. Extensive data from proteomic screens and other experiments reveal that a high ACO/LRO is a general feature of proteins with strong resistances to denaturation and degradation. These results provide tractable approaches for predicting resistance and designing proteins with sufficient topological complexity and long-range interactions to accommodate destabilizing functional features as well as withstand chemical and proteolytic challenge.