Discovery of Orally Available and Brain Penetrant AEP Inhibitors.

Alzheimer's Disease (AD) is the most widespread form of dementia, with one of the pathological hallmarks being the formation of neurofibrillary tangles (NFTs). These tangles consist of phosphorylated Tau fragments. Asparagine endopeptidase (AEP) is a key Tau cleaving enzyme that generates aggregation-prone Tau fragments. Inhibition of AEP to reduce the level of toxic Tau fragment formation could represent a promising therapeutic strategy. Here, we report the first orthosteric, selective, orally bioavailable, and brain penetrant inhibitors with an irreversible binding mode. We outline the development of the series starting from reversible molecules and demonstrate the link between inhibition of AEP and reduction of Tau N368 fragment both in vitro and in vivo.

Details Matter in Structure-based Drug Design.

Successful structure-based drug design (SBDD) requires the optimization of interactions with the target protein and the minimization of ligand strain. Both factors are often modulated by small changes in the chemical structure which can lead to profound changes in the preferred conformation and interaction preferences of the ligand. We draw from examples of a Roche project targeting phosphodiesterase 10 to highlight that details matter in SBDD. Data mining in crystal structure databases can help to identify these sometimes subtle effects, but it is also a great resource to learn about molecular recognition in general and can be used as part of molecular design tools. We illustrate the use of the Cambridge Structural Database for identifying preferred structural motifs for intramolecular hydrogen bonding and of the Protein Data Bank for deriving propensities for protein-ligand interactions.

Chemometrics in Protein Formulation: Stability Governed by Repulsion and Protein Unfolding.

Therapeutic proteins can be challenging to develop due to their complexity and the requirement of an acceptable formulation to ensure patient safety and efficacy. To date, there is no universal formulation development strategy that can identify optimal formulation conditions for all types of proteins in a fast and reliable manner. In this work, high-throughput characterization, employing a toolbox of five techniques, was performed on 14 structurally different proteins formulated in 6 different buffer conditions and in the presence of 4 different excipients. Multivariate data analysis and chemometrics were used to analyze the data in an unbiased way. First, observed changes in stability were primarily determined by the individual protein. Second, pH and ionic strength are the two most important factors determining the physical stability of proteins, where there exists a significant statistical interaction between protein and pH/ionic strength. Additionally, we developed prediction methods by partial least-squares regression. Colloidal stability indicators are important for prediction of real-time stability, while conformational stability indicators are important for prediction of stability under accelerated stress conditions at 40 °C. In order to predict real-time storage stability, protein-protein repulsion and the initial monomer fraction are the most important properties to monitor.

A high quality, industrial data set for binding affinity prediction: performance comparison in different early drug discovery scenarios.

We release a new, high quality data set of 1162 PDE10A inhibitors with experimentally determined binding affinities together with 77 PDE10A X-ray co-crystal structures from a Roche legacy project. This data set is used to compare the performance of different 2D- and 3D-machine learning (ML) as well as empirical scoring functions for predicting binding affinities with high throughput. We simulate use cases that are relevant in the lead optimization phase of early drug discovery. ML methods perform well at interpolation, but poorly in extrapolation scenarios-which are most relevant to a real-world application. Moreover, we find that investing into the docking workflow for binding pose generation using multi-template docking is rewarded with an improved scoring performance. A combination of 2D-ML and 3D scoring using a modified piecewise linear potential shows best overall performance, combining information on the protein environment with learning from existing SAR data.

Augmenting Structure-Based Design with Experimental Protein-Ligand Interaction Data: Molecular Recognition, Interactive Visualization, and Rescoring.

The previously introduced ratio of frequencies (RF ) framework provides statistically sound information on the relative interaction preferences of atoms in crystal structures. By applying the methodology to protein-ligand complexes, we can investigate the significance of interactions that are employed in structure-based drug design. Here, we revisit three aspects of molecular recognition in the light of the RF framework, namely stacking interactions of heteroaromatic rings with protein amide groups, interactions of acidified C-H groups, and interaction differences between syn and anti lone pairs of carboxylate groups. In addition, we introduce a highly interactive visualization tool that facilitates design idea generation in structure-enabled drug discovery projects. Finally, we show that applying the RF analysis as a simple rescoring tool after docking improves enrichment factors for the DUD-E diverse targets subset supporting the relevance of our approach.

Identification of Noncompetitive Protein-Ligand Interactions for Structural Optimization.

For efficient structure-guided drug design, it is important to have an excellent understanding of the quality of interactions between the target receptor and bound ligands. Identification and characterization of poor intermolecular contacts offers the possibility to focus design efforts directly on ligand regions with suboptimal molecular recognition. To enable a more straightforward identification of these in a structural model, we use a suitably enhanced version of our previously introduced statistical ratio of frequencies (RF) approach. This allows us to highlight protein-ligand interactions and geometries that occur much less often in the Protein Data Bank than would be expected from the exposed surface areas of the interacting atoms. We provide a comprehensive overview of such noncompetitive interactions and geometries for a set of common ligand substituents. Through retrospective case studies on congeneric series and single-point mutations for several pharmaceutical targets, we illustrate how knowledge of noncompetitive interactions could be exploited in the drug design process.

Study of the interaction between a novel, protein-stabilizing dipeptide and Interferon-alpha-2a by construction of a Markov state model from molecular dynamics simulations.

We recently reported the discovery of a novel protein stabilizing dipeptide, glycyl-D-asparagine, through a structure-based approach. As the starting hypothesis leading to the discovery, we postulated a stabilizing effect achieved by binding of the dipeptide to an aggregation prone region on the protein's surface. Here we present a detailed study of the interaction mechanism between the dipeptide and Interferon-alpha-2A (IFN) through the construction of a Markov state model from molecular dynamics trajectories. We identify multiple binding sites and compare these to aggregation prone regions. Additionally, we calculate the lifetime of the protein-excipient complex. If the excipient remained bound to IFN after administration, it could alter the protein's therapeutic efficacy. We establish that the lifetime of the complex between IFN and glycyl-D-asparagine is extremely short. Under these circumstances, stabilization by stoichiometric binding is consequently no impediment for a safe use of an excipient.

Exploring Chemical Space for New Substances to Stabilize a Therapeutic Monoclonal Antibody.

The physical stability of therapeutic proteins is a major concern in the development of liquid protein formulations. The number of degrees of freedom to control a given protein's stability is limited to pH, ionic strength, and type and concentration of excipient. There are only very few, mostly similar excipients currently in use, restricted to the list of substances generally recognized as safe for human use by the U.S. Food and Drug Administration. Opposed to this limited number of available excipients, there is the vast chemical space, which is hypothesized to consist of 1060 compounds. Its potential to stabilize proteins has never been explored systematically in the context of the formulation of therapeutic proteins. Here we present a screening strategy to discover new excipients to further improve an already stable formulation of a therapeutic antibody. The data are used to build a predictive model that evaluates the stabilizing potential of small molecules. We argue that before worrying about the hurdles of toxicity and approval of novel excipient candidates, it is mandatory to assess the actual potential hidden in the chemical space.

Advancing Therapeutic Protein Discovery and Development through Comprehensive Computational and Biophysical Characterization.

Therapeutic protein candidates should exhibit favorable properties that render them suitable to become drugs. Nevertheless, there are no well-established guidelines for the efficient selection of proteinaceous molecules with desired features during early stage development. Such guidelines can emerge only from a large body of published research that employs orthogonal techniques to characterize therapeutic proteins in different formulations. In this work, we share a study on a diverse group of proteins, including their primary sequences, purity data, and computational and biophysical characterization at different pH and ionic strength. We report weak linear correlations between many of the biophysical parameters. We suggest that a stability comparison of diverse therapeutic protein candidates should be based on a computational and biophysical characterization in multiple formulation conditions, as the latter can largely determine whether a protein is above or below a certain stability threshold. We use the presented data set to calculate several stability risk scores obtained with an increasing level of analytical effort and show how they correlate with protein aggregation during storage. Our work highlights the importance of developing combined risk scores that can be used for early stage developability assessment. We suggest that such scores can have high prediction accuracy only when they are based on protein stability characterization in different solution conditions.

Structure-based discovery of a new protein-aggregation breaking excipient.

Reducing the aggregation of proteins is of utmost interest to the pharmaceutical industry. Aggregated proteins are often less active and can cause severe immune reactions in the patient upon administration. At the same time the biopharmaceutical market is pushing for high concentration formulations and products that do not require refrigerated storage conditions. For a given protein, the only solution pH, ionic strength and concentration of a very limited number of excipients are the only parameters that can be varied to obtain a stable formulation. In this work, we present a structure-based approach to discover new molecules that successfully reduce the aggregation of proteins and apply the approach to the model protein Interferon-alpha-2a.

Zn2+-triggered self-assembly of Gonadorelin [6-D-Phe] to produce nanostructures and fibrils.

A synthetic derivative, GnRH [6-D-Phe], stable against enzymatic degradation, self-assembles and forms nanostructures and fibrils upon a pH shift in the presence of different concentrations of Zn2+ in vitro. Attenuated Total Reflection Fourier Transform Infrared spectroscopy (ATR-FTIR) revealed the existence of higher order assembly of Zn2+: GnRH [6-D-Phe]. Nuclear Magnetic Resonance spectroscopy (NMR) indicated a weak interaction between Zn2+ and GnRH [6-D-Phe]. Atomic Force Microscopy (AFM) showed the existence of GnRH [6-D-Phe] oligomers and fibrils. Molecular Dynamic (MD) simulation of the 10:1 Zn2+: GnRH [6-D-Phe] explored the interaction and dimerization processes. In contrast to already existing short peptide fibrils, GnRH [6-D-Phe] nanostructures and fibrils form in a Tris-buffered pH environment in a controlled manner through a temperature reduction and a pH shift. The lyophilized Zn2+: GnRH [6-D-Phe] assembly was tested as a platform for the sustained delivery of GnRH [6-D-Phe] and incorporated into two different oil vehicle matrices. The in vitro release was slow and continuous over 14 days and not influenced by the oil matrix.

Towards electroenzymatic processes involving old yellow enzymes and mediated cofactor regeneration.

Old yellow enzymes are able to catalyze asymmetric C=C reductions. A mediated electroenzymatic process to regenerate the NADPH in combination with an old yellow enzyme was investigated. Due to the fact that the overall process was affected by a broad set of parameters, a design of experiments (DoE) approach was chosen to identify suitable process conditions. Process conditions with high productivities of up to 2.27 mM/h in combination with approximately 90% electron transfer efficiency were identified.