Colloidal protein stability and surface coverage by trehalose: bridging molecular dynamics and experimental work

Abstract

The aim of this study was to assess to what extent a simulated trehalose monolayer can predict the minimum molar trehalose to protein ratios required to preserve the colloidal stability of lyophilized lactate dehydrogenase, catalase, and β-galactosidase. We employed an iterative molecular dynamics (MD) approach to quantify the number of trehalose molecules required to saturate the protein’s accessible surface area. These in-silico values were compared to experimentally determined colloidal stability thresholds (tetramer recovery) in lyophilized formulations, defined as the minimum molar ratio required to maintain hydrodynamic volume, as assessed by size exclusion chromatography following storage at 30°C and 43% relative humidity for a week. A quantitative, protein-specific correlation was observed between the simulated surface coverage and the experimental colloidal stability thresholds. This result shows that MD simulations can serve as a powerful pre-formulation tool for predicting minimum excipient requirements and provides a foundation for a more rational approach to protein stabilization.

Introduction

Lyophilization, or freeze-drying, is a commonly used method to increase protein stability and to extend their shelf life. By removing water and bringing the protein in a dry state molecular motion is reduced, generally hampering degradation pathways [1], [2]. Even though freeze drying is considered as a gentle drying method, several stresses can still occur during the freezing and drying phases of this process. To overcome the harmful effects of these stresses to the protein, excipients can be added to protein formulations. For example, disaccharides such as sucrose and trehalose are often used as lyoprotectants to increase protein stability during lyophilization [3], [4]. Several mechanisms have been proposed to explain how these excipients preserve protein integrity [5], [6]. When properly formulated and lyophilized, effective protein stabilizing excipients form a highly viscous glassy matrix, which greatly reduces translational molecular mobility of the system. As a result, degradation processes such as denaturation and aggregation are significantly slowed down [7]. This kinetic stabilization mechanism is often referred to as vitrification. However, thermodynamic stabilization mechanisms also play a role in stabilization by lyoprotectants. Disaccharides are able to substitute water in the protein’s hydration shell during drying, due to their hydrogen bonding capacity. By doing so, the hydrogen bond network of a protein with its surroundings is maintained, allowing maintenance of the native state [8], [9]. This is often referred to as the water replacement theory. Over the past decades, these two theories (i.e., vitrification and water replacement) have been continuously refined. For example, it is argued that water is not only replaced, but small amounts of residual moisture “anchor” the protein to the amorphous matrix [6]. This would effectively result in coupling of the ‘slow’ ⍺-relaxation of the matrix, characterized by a high glass transition temperature (Tg), to ‘fast’ β-relaxation, or local mobility of the incorporated protein. β-relaxation has been shown to correlate well with protein stability [10], [11]. Nevertheless, a combination of multiple mechanisms is considered essential for effective protein stabilization, especially to endure challenging storage conditions and improve protein storage stability. This is exemplified in multi-excipient formulations, where a polymer increases viscosity of the matrix and provides kinetic stability and small excipients like disaccharides or amino acids provide effective water replacement and suppression of local mobility [12], [13]. Ultimately, the packing densities and interaction potential of excipients with the protein surface likely play a role in the relative contribution of each of these mechanisms, which remains an active subject of debate to this day [14].Fig. 1.Fig. 2.Fig. 3..

Fig. 1. Snapshots showing the lactate dehydrogenase (left), catalase (middle) and β-galactosidase as ribbon diagram (top) and populated with trehalose after its final packing and equilibration step (bottom), with protein in purple and trehalose in green.

Despite growing experimental molecular-level insight into solid-state protein stabilization [6], [9], [15], formulation design remains largely empirical, underscoring the need for predictive tools. Molecular dynamics (MD) simulations offer such potential, since they can provide atomistic-level insights into the interactions between formulation components [6], [15], [16], [17], [18], [19]. While powerful, these simulations are not intended to replace experimental work, rather, both should be considered in a complementary strategy. Once validated by experimental work, simulations might contribute to more informed decision-making during product development. Ultimately, such a combined approach may lead to a more rational and efficient methodology for formulation development and reduce the required experimental efforts. As such, we combined simulations and experiments to investigate how trehalose packs on the surfaces of various proteins and its correlation to colloidal storage stability. To this end, we performed MD simulations to estimate the number of trehalose molecules required for a monolayer on the surface of three model proteins, i.e., lactate dehydrogenase, catalase and β-galactosidase. These were selected for their broad range in molecular weight (MW), from 144 to 540 kDa, while trehalose was chosen because it represents one of the most widely utilized and extensively characterized lyoprotectants in biopharmaceutical formulations [20]. Experimentally we determined, via size exclusion chromatography (SEC), which trehalose to protein ratios were required to maintain stability of the hydrodynamic volume for the three proteins. Then, using large-scale MD simulations we showed that the trehalose monolayer coverage strongly correlated with the experimentally derived minimum number of trehalose molecules required for colloidally stable formulations. Our data demonstrate that a quantitative, protein-specific correlation between computed excipient surface coverage and experimental molar excipient to protein ratio to maintain SEC recovery exists, providing mechanistic insight on the molecular driving forces that are responsible for the surface-covering hydrogen bond networks at the atomistic scale.

Download the full article as PDF here Colloidal protein stability and surface coverage by trehalose

or continue reading here

Materials

Catalase, sodium dihydrogen phosphate dihydrate, disodium hydrogen phosphate dihydrate and potassium carbonate were obtained from Merck (Rahway, NJ, United States). Lactate dehydrogenase and 41-piperazineethanesulfonic acid (HEPES) and sodium chloride were purchased from Sigma-Aldrich (Zwijndrecht, The Netherlands). β-Galactosidase was obtained from Sorachim (Lausanne, Switzerland). BioHale©️ Trehalose dihydrate was kindly provided by DFE Pharma (Goch, Germany).

D. Zillen, V. Adupa, K.T.T. Nguyen, H.W. Frijlink, P.R. Onck, W.L.J. Hinrichs, Colloidal protein stability and surface coverage by trehalose: bridging molecular dynamics and experimental work, European Journal of Pharmaceutics and Biopharmaceutics, 2026, 115242, ISSN 0939-6411, https://doi.org/10.1016/j.ejpb.2026.115242.