The surface exposed amino acid residues of monomeric proteins determine the partitioning in aqueous two-phase systems

Electrostatic calculations and quantitative protein retention models for ion exchange chromatography

A novel set of protein descriptors has been developed to increase the understanding of protein behavior on chromatographic media. The protein descriptors are pH-dependent and based on electrostatic and hydrophobic properties of mainly the surface of the proteins as revealed by their three-dimensional structure. Interpretable and predictive quantitative structure property relationship (QSPR) models were then obtained for protein retention in ion exchange chromatography at different pH values. In most cases the calculated average surface potential could be directly related to retention times. Moreover, the high retention of human lactoferrin observed in cation exchange even at high pH values could be modeled by adding descriptors of the charge asymmetry.

Predicting the behaviour of proteins in hydrophobic interaction chromatography

This paper focuses on the prediction of the dimensionless retention time (DRT) of proteins in hydrophobic interaction chromatography (HIC) by means of mathematical models based on the statistical description of the amino acid surface distribution. Previous models characterises the protein surface as a whole. However, most of the time it is not the whole protein but some of its specific regions that interact with the environment. It seems much more natural to use local measurements of the characteristics of the surface. Therefore, the statistical characterisation of the distribution of an amino acid property on the protein surface was carried out from the systematic calculation of the local average of this property in a neighbourhood placed sequentially on each of the amino acids on the protein surface. This process allowed us to characterise the distribution of this property quantitatively using three main statistics: average, standard deviation and maximum. In particular, if the property considered is a hydrophobicity scale, these statistics allowed us to characterise the average hydrophobicity and the hydrophobic content of the most hydrophobic cluster or hotspot, as well as the heterogeneity of the hydrophobicity distribution on the protein surface. We tested the performance of the DRT predictive models based on these statistics on a set of 15 proteins. We obtained better predictive results with respect to the models previously reported. The best predictive model was a linear model based on the maximum. This statistic was calculated using an index of the mobilities of amino acids in chromatography. The predictive performance of this model (measured as the Jack Knife MSE) was 26.9% better than those obtained by the best model which does not consider the amino acid distribution and 19.5% better than the model based on the hydrophobic imbalance (HI). In addition, the best performance was obtained by a linear multivariable model based on the HI and the maximum. The difference between the experimental data and the prediction carried out by this model was smaller than those observed previously. In fact, this model obtained better predictive capacities than a previous linear multivariable model decreasing the Jack Knife MSE in 8.7%. In addition, this model allowed us to diminish the number of variables required, increasing, in this way, the degrees of freedom of the model.

Predicting the behaviour of proteins in hydrophobic interaction chromatography

This paper focuses on the prediction of the dimensionless retention time of proteins (DRT) in hydrophobic interaction chromatography (HIC) by means of mathematical models based on characteristics of the surface hydrophobicity distribution. We introduce a new parameter, called hydrophobic imbalance (HI), obtained from the three-dimensional structure of proteins. This parameter quantifies the displacement of the superficial geometric centre of the protein when the effect of the hydrophobicity of each amino acid is considered. This parameter is simpler and less expensive than those reported previously. We use HI as a way to incorporate information about the surface hydrophobicity distribution in order to improve the prediction of DRT. We tested the performance of our DRT predictive models in a set of 15 proteins. This set includes four proteins whose DRTs are known as very difficult to predict. By means of the variable HI, it was possible to improve the predictive characteristics obtained by models based on the average surface hydrophobicity (ASH) by 9.1%. Also, we studied linear multivariable models based on characteristics determined from the HI. By using this multivariable model, a correlation coefficient of 0.899 was obtained. With this model, we managed to improve the predictive characteristics shown by previous models based on ASH by 31.8%.

Investigation of protein retention and selectivity in HIC systems using quantitative structure retention relationship models

In the present work, the effect of stationary phase resin chemistry and protein physicochemical properties on protein binding affinity in hydrophobic interaction chromatography (HIC) was investigated using linear gradient chromatography and quantitative structure-retention relationship (QSRR) modeling. Linear gradient experiments were carried out for a set of model proteins on four different HIC resins having different backbone and ligand chemistry. The retention data exhibited significant differences in protein binding affinity, not only across the phenyl and butyl ligand chemistries, but also for the different backbone chemistries found in the Sepharose (cross-linked agarose) and the Toyopearl 650 M (polymethacrylate) series of resins. QSRR models based on a Support Vector Machine (SVM) approach were developed for the linear retention data using molecular descriptors based on protein crystal structure and primary sequence information as well as a set of new hydrophobicity descriptors based on the solvent accessible protein surface area. The results indicate that the QSRR models were successfully able to capture and selectivity predict the changes observed in these systems. Furthermore, the new descriptors resulted in physically interpretable models of protein retention and provided insights into the factors influencing protein affinity in these different HIC systems. The approach put forth in this study provides a framework for developing predictive tools and for gaining insight into protein selectivity in hydrophobic interaction chromatography.

New approaches for predicting protein retention time in hydrophobic interaction chromatography

Hydrophobic interaction chromatography (HIC) is an important technique for the purification of proteins. In this paper, we review three different approaches for predicting protein retention time in HIC, based either on a protein's structure or on its amino-acidic composition, and we have extended one of these approaches. The first approach correlates the protein retention time in HIC with the protein average surface hydrophobicity. This methodology is based on the protein three-dimensional structure data and considers the hydrophobic contribution of the exposed amino acid residues as a weighted average. The second approach, which we have extended, is based on the high correlation level between the average surface hydrophobicity of a protein's hydrophobic interacting zone and its retention time in HIC. Finally, a third approach carries out a prediction of the average surface hydrophobicity of a protein, using only its amino-acidic composition, without knowing its three-dimensional structure. These models would make it possible to test different operating conditions for the purification of a target protein by computer simulations, and thus make it easier to select the optimal conditions, contributing to the rational design and optimization of the process.

Nanoparticle Conjugation Increases Protein Partitioning in Aqueous Two-Phase Systems

We describe the effect of bioconjugation to colloidal Au nanoparticles on protein partitioning in poly(ethylene glycol) (PEG)/dextran aqueous two-phase systems (ATPS). Horseradish peroxidase (HRP) was conjugated to colloidal Au nanoparticles by direct adsorption. Although HRP alone had very little phase preference, HRP/Au nanoparticle conjugates typically partitioned to the PEG-rich phase, up to a factor of 150:1 for conjugates of 15-nm colloidal Au. Other protein/Au nanoparticle conjugates exhibited partitioning of greater than 2000:1 to the dextran-rich phase, as compared with approximately 5:1 for the free protein. The degree of partitioning was dependent on polymer concentration and molecular weight, nanoparticle diameter, and in some instances, nanoparticle concentration in the ATPS. The substantial improvements in protein partitioning achievable by conjugation to Au nanoparticles appear to result largely from increased surface area of the conjugates and require neither chemical modification of the proteins or polymers with affinity ligands, increased polymer concentrations, nor addition of high concentrations of salts. Adsorption to colloidal particles thus provides an attractive route for increased partitioning of enzymes and other proteins in ATPS. Furthermore, these results point to ATPS partitioning as a powerful means of purification for biomolecule/nanoparticle conjugates, which are increasingly used in diagnostics and materials applications.

Prediction of retention times of proteins in hydrophobic interaction chromatography using only their amino acid composition

This paper focuses on the prediction of the dimensionless retention time of proteins (DRT) in hydrophobic interaction chromatography (HIC) by means of mathematical models based, essentially, only on aminoacidic composition. The results show that such prediction is indeed possible. Our main contribution was the design of models that predict the DRT using the minimal information concerning a protein: its aminoacidic composition. The performance is similar to that observed in models that use much more sophisticated information such as the three-dimensional structure of proteins. Three models that, in addition to the amino acid composition, use different assumptions about the amino acids tendency to be exposed to the solvent, were evaluated in 12 proteins with known experimental DRT. In all the cases analyzed, the model that obtained the best results was the one based on a linear estimation of the aminoacidic surface composition. The models were adjusted using a collection of 74 vectors of aminoacidic properties plus a set of 6388 vectors derived from these using two mathematical tools: k-means and self-organizing maps (SOM) algorithms. The best vector was generated by the SOM algorithm and was interpreted as a hydrophobicity scale based partly on the tendency of the amino acids to be hidden in proteins. The prediction error (MSE(JK)) obtained by this model was almost 35% smaller than that obtained by the model that supposes that all the amino acids are completely exposed and 40% smaller than that obtained by the model that uses a simple correction factor considering the general tendency of each amino acid to be exposed to the solvent. In fact, the performance of the best model based on the aminoacidic composition was 5% better than that observed in the model based on the three-dimensional structure of proteins.

Structure and Thermodynamics of Protein−Polymer Solutions: Effects of Spatially Distributed Hydrophobic Surface Residues

Protein-polymer association in solution driven by a short-range attraction has been investigated using a simple coarse-grain model solved by Monte Carlo simulations. The effect of the spatial distribution of the hydrophobic surface residues of the protein on the adsorption of weakly hydrophobic polymers at variable polymer concentration, polymer length, and polymer stiffness has been considered. Structural data on the adsorbed polymer layer and thermodynamic properties, such as the free energy, energy, and entropy, related to the protein-polymer interaction were calculated. It was found that a more heterogeneous distribution of the surface residues promotes adsorption and that this also applies for different polymer concentrations, polymer chain lengths, and polymer flexibilities. Furthermore, the polymer adsorption onto proteins with more homogeneous surface distributions displayed larger sensitivity to polymer properties such as chain length and flexibility. Finally, a simple relation between the adsorption probability and the change in the free energy was found and rationalized by a simple two-state adsorption model.

Is it possible to predict the average surface hydrophobicity of a protein using only its amino acid composition?

Hydrophobicity is one of the most important physicochemical properties of proteins. Moreover, it plays a fundamental role in hydrophobic interaction chromatography, a separation technique that, at present time, is used in most industrial processes for protein purification as well as in laboratory scale applications. Although there are many ways of assessing the hydrophobicity value of a protein, recently, it has been shown that the average surface hydrophobicity (ASH) is an important tool in the area of protein separation and purification particularly in protein chromatography. The ASH is calculated based on the hydrophobic characteristics of each class of amino acid present on the protein surface. The hydrophobic characteristics of the amino acids are determined by a scale of aminoacidic hydrophobicity. In this work, the scales of Cowan-Whittaker and Berggren were studied. However, to calculate the ASH, it is necessary to have the three-dimensional protein structure. Frequently this data does not exist, and the only information available is the amino acid sequence. In these cases it would be desirable to estimate the ASH based only on properties extracted from the protein sequence. It was found that it is possible to predict the ASH from a protein to an acceptable level for many practical applications (correlation coefficient > 0.8) using only the aminoacidic composition. Two predictive tools were built: one based on a simple linear model and the other on a neural network. Both tools were constructed starting from the analysis of a set of 1982 non-redundant proteins. The linear model was able to predict the ASH for an independent subset with a correlation coefficient of 0.769 for the case of Cowan-Whittaker and 0.803 for the case of Berggren. On the other hand, the neural model improved the results shown by the linear model obtaining correlation coefficients of 0.831 and 0.836, respectively. The neural model was somewhat more robust than the linear model particularly as it gave similar correlation coefficients for both hydrophobicity scales tested, moreover, the observed variabilities did not overcome 6.1% of the mean square error. Finally, we tested our models in a set of nine proteins with known retention time in hydrophobic interaction chromatography. We found that both models can predict this retention time with correlation coefficients only slightly inferior (11.5% and 5.5% for the linear and the neural network models, respectively) than models that use the information about the three-dimensional structure of proteins.

Effect of surface hydrophobicity distribution on retention of ribonucleases in hydrophobic interaction chromatography

The effect of surface hydrophobicity distribution of proteins on retention in hydrophobic interaction chromatography (HIC) was investigated. Average surface hydrophobicity as well as hydrophobic contact area between protein and matrix were estimated using a classical thermodynamic model. The applicability of the model to predict protein retention in HIC was investigated on ribonucleases with similar average surface hydrophobicity but different surface hydrophobicity distribution. It was shown experimentally that surface hydrophobicity distribution could have an important effect on protein retention in HIC. The parameter "hydrophobic contact area," which comes from the thermodynamic model, was able to represent well the protein retention in HIC with salt gradient elution. Location and size of the hydrophobic patches can therefore have an important effect on protein retention in HIC, and the hydrophobic contact area adequately describes this.

A theory of protein–resin interaction in hydrophobic interaction chromatography

Docking simulations were performed in order to investigate surface area of interaction between several ribonucleases and a reduced model for the hydrophobic moiety used in Phenyl Sepharose using the program AutoDock 3.0. For each ribonucelase, 80 independent simulations with populations consisting of 100 random structures were performed and from these the most probable docked protein-ligand conformations were obtained. A new methodology was used to select the most probable conformations, based on qualitative and quantitative considerations. The interacting amino acids in each protein were identified. The average surface hydrophobicity of the interfacial zone (local hydrophobicity, LH) was determined. The LH showed a high correlation level (r2 = 0.99) with the "hydrophobic contact area" (HCA) experimentally determined for the different ribonucleases as well as with the dimensionless retention time (r2 = 0.90). This study allowed us to identify the zones on the protein surface most probably involved in protein retention in HIC, without tedious experimental work. Given the good correlation level obtained, this new methodology may constitute a novel approach that could be used to predict protein behavior in HIC.

Hydrophobic peptide tags as tools in bioseparation

Hydrophobic interactions are highly selective, and differences in surface hydrophobicities between proteins can be used as an efficient handle to facilitate protein isolation. Aromatic amino acid residues are of particular importance for molecular recognition because they have a key role in several biological functions. The hydrophobicity of a protein can easily be altered with minor genetic modifications, such as site-directed mutagenesis or fusions of hydrophobic peptide tags. An important advantage of hydrophobic peptide tags over traditional affinity tags is the possibility of exploring simple and inexpensive bioseparation materials. Recent results demonstrate the potential of hydrophobic interaction chromatography and aqueous two-phase systems as tools to study relative hydrophobicities of recombinant proteins with only minor alterations. This review focuses on hydrophobic peptide tags as fusion partners, which can be used as important tools in bioseparation.

Alternative bioseparation operations: life beyond packed-bed chromatography

Chromatography is undoubtedly the workhorse of downstream processes, affording high resolution for bioseparations. At the same time, it has the notoriety of being the single largest cost center in downstream processing and of being a low-throughput operation. Consequently, 'chromatography alternatives' are an attractive proposition, even if only a reduction in the extent of use of packed beds can be realized. This paper reviews the current state of unit operations posing as chromatography alternatives--including membrane filtration, aqueous two-phase extraction, three-phase partitioning, precipitation, crystallization, monoliths and membrane chromatography--and their potential to do the unthinkable.

Binding of amino acids to "smart" sorbents: where does hydrophobicity come into play?

Poly(N-isopropylacrylamide) (PNIPA)-based sorbents have been successfully used as sorbents in temperature-sensitive chromatography. Yet, the mechanisms controlling the binding of biochemicals to these sorbents and, therefore, the separation process are not fully understood. In the current work, the role of hydrophobic interactions in the binding of amino acids of different hydrophobicities to PNIPA microgels was studied. Binding experiments were conducted both below (25 degrees C) and above (37 degrees C) the volume-phase transition temperature of the gel. At 25 degrees C, no straightforward correlation between the partition coefficient and the hydrophobicity could be suggested for low hydrophobicity values. Contrary, at higher hydrophobicities the partition coefficient increases with increasing hydrophobicity. This correlation holds for the whole hydrophobicity range at 37 degrees C; however, the binding data suggests two different binding mechanisms of the hydrophilic amino acids and the hydrophobic ones. Isothermal titration calorimetry measurements confirmed this suggestion: The binding of hydrophobic amino acids seems to be driven by hydrophobic interactions, as evident from the positive binding enthalpy and the clear correlation between the amino acid's hydrophobicity and the binding entropy. Contrary, the binding of the hydrophilic amino acids was exothermic, implying a binding mechanism based on specific interactions, most probably hydrogen bonding.

Methodology for predicting the separation of proteins by hydrophobic interaction chromatography and its application to a cell extract

Hydrophobic interaction chromatography (HIC) is widely used in the downstream processing of proteins. Resolution of HIC is very good, but sometimes not as high as expected. Resolution values could be increased if good operating conditions were selected. In this paper we present a methodology for selecting good operating conditions. First, it is necessary to predict the dimensionless retention time (DRT) of each protein in the mixture. We use a correlation such as DRT = A + Bphi + Cphi2, where phi is the superficial hydrophobicity of the protein, which is calculated considering the hydrophobicity of the superficial amino acids using the Miyazawa-Jernigan scale. Considering that there was little interaction amongst proteins in a mixture at the concentrations investigated (2 g/l of each protein), the behaviour of the proteins in the mixture was considered to be similar to that of the individual proteins. Using simulations it was possible to test different operating conditions for the purification of a target protein from a mixture of proteins and it was possible to select ideal conditions. The methodology developed was also tested for the purification of a recombinant protein from a fermentation extract of yeast producing human superoxide dismutase and the results have been satisfactory.

Mathematical correlations for predicting protein retention times in hydrophobic interaction chromatography

This paper reports how a novel method for determining the surface hydrophobicity of a protein can be used to predict the dimensionless retention time (DRT) in hydrophobic interaction chromatography. The methodology has three steps. First, it is necessary to know the three-dimensional structure of the target protein. Then, it is necessary to calculate the surface hydrophobicity of the protein, considering that each amino acid has a relative contribution to the surface properties, using an appropriate equation. This will depend largely on the relative scale used to evaluate the hydrophobicity of each amino acid. Forty-two scales were investigated and two gave the best correlation, the Miyazawa-Jernigan and the Cowan-Whittaker scales. Finally, a "quadratic model" (DRT = a + b phi(surface) + c phi2(surface)) is used to predict the retention time of the target protein. The methodology was able to correlate adequately the retention data for monomeric proteins not included in the generation of the model.

Improved partitioning in aqueous two-phase system of tyrosine-tagged recombinant lactate dehydrogenase

The partitioning of Bacillus stearothermophilus lactate dehydrogenase (LDH) in an aqueous two-phase system was studied. Particularly, the influence of tyrosine tags on the partitioning was evaluated. The hydrophobic effect, caused by the addition of tyrosine residues, was determined in a system based on dextran and the thermoseparating ethylene oxide-propylene oxide random copolymer (EO30PO70). Five different LDH variants were constructed with N-terminal tags containing tyrosines (Y3 and Y6), tyrosines and prolines (Y3P2 and Y6P2), and only prolines (P2). LDH fused with tags containing tyrosines increased the partitioning coefficient, and the more tyrosines added to the protein, the larger improvement in partitioning. When prolines were added between the tyrosine-rich tag and the protein, a further increased partitioning was obtained. The enhanced partitioning was attributed to the rigid structure of the proline, which in turn led to an increase in the exposure of the tag to the surroundings. The best tyrosine tag, Y6P2, increased the partition coefficient four times and additionally, a higher thermostability was observed.