Scaffold-Based Drug Design

Scaffold-Based Drug Design: Replacing an Unwanted Molecular Motif

Let's start by discussing what an unwanted scaffold might be. Firstly, an unwanted scaffold could be a structural component that, while forming the pharmacophore of a molecule—the part responsible for the biological activity and necessary for binding to the target—could simultaneously be the part responsible for toxicity within a compound series.

Unfortunately, it often happens that insights into whether a compound may cause undesirable in vivo effects are gained quite late in the R&D process. This is particularly frustrating when all previous findings suggest that everything was on the right track and that the compound had the potential to become a blockbuster drug. One way out of this dead end could be to investigate a new series of compounds structurally related to the current one. This is the principle of 'scaffold hopping': replacing a pharmacophore with a non-identical motif. This can start with the substitution of a single heavy atom, which is sometimes insufficient, as we will discuss shortly, extending up to a complete replacement of the core scaffold that maintains a similar arrangement of molecular functionalities.

Unwanted scaffolds can also include FDA-approved drugs. The reason behind this is that these drugs are so successful that it’s challenging to profit from them because their critical scaffold is protected by a Markush structure patent. Generally, this issue is addressed extensively at the start of an R&D campaign by focusing only on structures that are novel enough to be patentable. However, it’s still possible to derive new chemotypes from known structures that target the same blockbuster mechanism of action.

Basic Principles of Structure-Based Drug Design

Scaffold-based drug design works best when layered with as much information as possible, similar to other computational methods. This means that while 2D methods can indeed yield success, 3D approaches provide the essential refinement needed for optimal results. Especially when attempting to overcome the limitations of a scaffold, structural modifications become crucial; purely 2D approaches often fall short in such cases.

For example, molecular fingerprints—a common method in 2D similarity assessments—are quite sensitive to even minor modifications in a molecule’s structure. Significant structural changes can lead to drastically lower similarity scores, which may result in related molecules being erroneously mixed with entirely unrelated structures in the middle or lower rankings of a dataset.

The scaffold hopping example on the left was reported in this publication.

Introducing Fuzzyness while Keeping Functionality

The most efficient method to achieve a scaffold hop during a scaffold-based drug design campaign is to introduce a wild card parameter that retains the core essence of the compound while delivering structurally distinct motifs. This form of fuzziness allows one to escape the gravitational field of similarity associated with a molecule and its space within the chemical space, while still generating results that maintain similar functionalities to the template molecule.

By subsequently combining with other orthogonal methods (e.g., 3D alignment, molecular fingerprints), one can ultimately identify compounds that are structurally related to the original molecule through multiple approaches. This becomes particularly interesting when applying these techniques to a series of active compounds with different scaffolds, as it enables the identification of a consensus among results from the custom libraries. This strategy helps in pinpointing candidates that are not only similar by one metric but also consistently show relatedness across various analytical dimensions.

Our developed algorithm for this purpose is called FTrees, which also serves as the engine for the Scaffold Hopper Mode in infiniSee.

BioSolveIT software for Chemical Space exploration:

infiniSee: Chemical Space navigation platform with a graphical user interface.
infiniSee retrieves relevant chemistry from ultra-large Chemical Spaces containing billions or even trillions of compounds based on their similarity to a query compound. Results are synthetically accessible per design in one or two steps and, in the case of our partners' Chemical Spaces, can be ordered directly to your table.
The Scaffold Hopper Mode can be run to retrieve compounds related by pharmacophore features. Local constraints can be applied to fine tune your results (e.g., keeping important decorations of the molecule).
The Analog Hunter Mode can be used to perform subsequent molecular fingerprint-based similarity screening.
infiniSee xREAL: Exclusive platform to screen Enamine's largest compound catalog featuring trillions of compounds.
infiniSee xREAL contains all features of infiniSee and supports all three Chemical Space exploration search modes.

Command-line tools for Chemical Space exploration:

FTrees: Pharmacophore-based similarity screen. Algorithm behind the Scaffold Hopper Mode.
SpaceLight: Retrieves close analogs based on molecular fingerprints. Algorithm behind the Analog Hunter Mode.

Enriching Actives with 3D Methods

As an independent or complementary method, 3D molecule alignment can add the necessary refinement to results, enabling the identification of even more precisely similar pharmacophoric arrangements.
It is particularly important to incorporate key project insights into this process. For example, if multiple key features that define the pharmacophore are known, constraints can be applied to the template molecule to ensure that the resulting compounds maintain these functionalities in a 3D arrangement.

3D molecule alignment and superposition-focused virtual screening in the context of scaffold-based drug design can be achieved with following BioSolveIT platform:

SeeSAR's Similarity Scanner Mode: Ligand-based virtual screening.

Command-line tools for 3D superpositioning:

FlexS: 3D compound alignment.

Structure-Based Core Replacement

Furthermore, scaffold hopping can also be applied in the context of a structure-based approach. This method involves selecting a portion of the molecule to be replaced using vectors while keeping the decorations (side chains) intact. A database search then identifies replacements that fit the specified 3D criteria. The search can be further refined by using additional pharmacophore constraints to ensure the proposed scaffolds meet specific project requirements.

The algorithm behind this process is called ReCore. ReCore leverages empirical data (e.g., from the PDB or CCDC) to suggest substitutions for molecular motifs, ensuring that new cores preserve the desired binding interactions and spatial orientations.

SeeSAR's Inspirator Mode: Runs ReCore to replace undesired core parts of a molecule. Several 3D databases are available.

Excited for more drug discovery solutions?

Keep on exploring!

Scaffold-Based Drug Design