We've updated our Privacy Policy to make it clearer how we use your personal data. We use cookies to provide you with a better experience. You can read our Cookie Policy here.

Advertisement

Fragment-Based Approach To Enhance Drug Discovery Productivity

Fragment-Based Approach To Enhance Drug Discovery Productivity content piece image
Listen with
Speechify
0:00
Register for free to listen to this article
Thank you. Listen to this article using the player above.

Want to listen to this article for FREE?

Complete the form below to unlock access to ALL audio articles.

Read time: 6 minutes

The ability to identify clinically valuable fragments is key to accelerating drug discovery.


Drug discovery is expensive and time-consuming. Productivity in drug discovery is tied to the ability to identify drug-like molecules that can be used to target clinically relevant pathways. In the last two to three decades, fragment-based drug discovery (FBDD) has emerged as a complementary approach to high-throughput screening (HTS) methods. FBDD is based on the theoretical concept of using small pieces of chemical fragments to bind to the active sites of protein targets, as individual chemical functional groups can affect the energy of binding between a drug and its target. Additionally, as a fragment has a simple structure, there is a lower probability of undesired interactions from side functional groups that can complicate actual binding affinities.


FBDD is also advantageous because fragment libraries sample a much larger chemical space than HTS and can regularly produce hits for optimization. Importantly, hit rates from fragments can be used to assess whether a biological target is potentially druggable, even for more complex targets like protein-DNA interactions. In this article, we will describe recent progress in the area of FBDD.

Validating hits from FBDD

Compared to HTS, fragments have weaker affinities. The equilibrium dissociation constant measures the propensity for the bound drug/target complex to dissociate to free drug and target. For FBDD, this is around 1 mM to 100 µM compared to that around 100 nM to 10 µM for HTS methods. Therefore, to accurately measure the affinities between fragments and their targets, sensitive biophysical methods are required. Some common examples include surface plasmon resonance (SPR), thermal shift affinity capture, X-ray crystallography and nuclear magnetic resonance (NMR).


X-ray crystallography is a powerful tool to obtain structures of proteins and complexes at high resolution. The resulting X-ray structures can be used to understand the binding mechanisms of inhibitors to the active sites of their targets as well as the formation of covalent bonds in active sites. It is routinely used for fragment hit identification and validation. Using this method, fragments must be crystallized to obtain their binding mechanisms at high resolution.


"When it comes to the choice of techniques such as X-ray crystallography, NMR and SPR for FBDD, it is about knowing what matters more: binding or relevance. With X-ray crystallography, the main advantage is that one can directly visualize a fragment with its target protein at the binding site. There is no need to infer this information, although it is important to question whether the binding is relevant. It is also important to note that unlike biophysical methods like SPR, X-ray crystallography does not offer differences in signal intensities which researchers use as a proxy to measure how strongly a fragment interacts with its target," says Prof. Frank von Delft, principal investigator at Oxford University.


"In my opinion, there are two main challenges in using X-ray crystallography for FBDD. First, it is the lack of systematic methodologies to obtain consistent and high-quality crystals of fragments to provide reliable structural information. Second, it is the need for computational tools which are powerful enough to analyze the structural information of crystals quickly and reliably. Despite these, I am optimistic about the use of X-ray crystallography for FBDD simply because of its ability to provide 'richness' of information to guide fragment study and modifications," adds von Delft.

Genomic Atlas Identifies Druggable Proteins for Neurological Disorders

Watch this webinar to hear Dr. Cruchaga discuss how existing drugs with the potential to slow or reverse brain damage were identified, the identification of causal proteins in Alzheimer’s disease, Parkinson’s disease and stroke and the first multi-tissue study yielding hundreds of novel protein quantitative loci

Watch Webinar

According to Dr. Julien Orts, assistant professor at the University of Vienna, while X-ray crystallography remains the most efficient 3D structure elucidation method available, it is not a reliable screening method. In his research, Orts develops NMR methods to improve drug discovery processes.


“NMR probes a larger chemical space, because NMR does not require crystallization to define whether a fragment is a true positive/a binder. NMR is the gold standard for screening, and it is also possible to obtain the structure of complexes using this approach.”


NMR is a highly sensitive technique that can identify fragments with different binding affinities from the nM to mM range, and also gives a lower rate of false positives. This is achieved by measuring the changes in NMR signals from either fragments/ligands or targets/proteins. While compound mixtures can be used in screening, the number of compounds that can be identified is limited due to signal overlaps. Various isotopes can be used in NMR for FBDD, including fluorine-19 and phosphorus-31. However, traditional NMR structure-calculation methods are not productive and cannot provide timely information for drug discovery timelines.


In a study led by Orts, Torres et al. developed the NMR molecular replacement method (NMR2) that can reduce the time needed to generate ligand-protein complex structures by using published structures of the target protein and relegating the observed nuclear Overhauser effect as an unimportant restraint. This method was first published in 2020 to solve the complex between three derivatives of a fragment and the protein receptor PIN1, a peptidyl-prolyl cis/trans isomerase overexpressed in several cancers which recognizes phospho-serine/threonine-proline motifs that contribute to tumor initiation and growth. Recently, the same group adapted the NMR2 method to understand the bindings models to bromodomains which recognize the posttranslational modifications of BRD4 and TRIM24, proteins that are also implicated in cancer.


“FBDD is providing better access to the chemical space right from the start of a project because fragments are simple and diverse. Any modification starting from a fragment towards a more complicated molecule reduces the chemical space greatly. So, by starting with simple ‘blocks’, it is possible to cover the largest possible chemical space and narrow down the search only if it is meaningful.”


Orts is hoping to apply his NMR2 method to other targets: “KRAS is the target I am currently trying to drug with fragments. KRAS doesn’t crystallize well, so X-ray crystallography isn’t helpful in this case and presents a major problem. KRAS is also very dynamic and the flexible regions (loops) play a crucial role in its functions. NMR is by nature better suited to studying flexible receptors. We are solving NMR2 structures of KRAS in complex with fragments to accelerate drug design for this oncoprotein,” adds Orts.

Kinetic Curvature: Assessment of Small Molecule Kinetics and Affinity

The detection and characterization of binding events is facilitated by sensitive biophysical technologies. Surface plasmon resonance (SPR) is a core technology used in many pharma and biotechnology settings for this purpose, however traditional initial SPR screening of weak affinity compounds using a single fixed concentration injection can present numerous challenges. Download this app note to discover how OneStep® injections can help you.

View App Note

Connecting fragments

After a fragment hit is found, the fragment can be grown into compounds with greater potency. This strategy is called fragment growing. Another way to create a potent compound is to make use of fragment merging when the fragments have overlapping binding sites which can be determined through X-ray crystallography. This can also be facilitated by analyzing the binding modes of fragments to the active sites of their targets using NMR and molecular docking. The most powerful approach is fragment linking, in which two or more fragments are connected to dramatically improve binding affinities. For instance, linking two fragments with binding affinities in the mM range will result in a compound with an affinity in the nM range. However, for this approach to work, the target usually has a relatively large binding pocket for fragments to bind to different regions within the pocket. Extensive structural information to understand molecular interactions is also important.


Mycobacterium abscessus (Mab) is a rapidly growing species of non-tuberculous mycobacteria (NTM) and a major threat to individuals with cystic fibrosis. Phosphoribosylaminoimidazole succinocarboxamide synthetase, or PurC, is an essential enzyme involved in de novo purine biosynthesis in bacteria. As PurC has distinct structural differences from its human ortholog, PAICS, it is an attractive target for antimicrobial drug discovery. Charoensutthivarakul et al. made use of FBDD approaches such as fragment growing and fragment merging to discover a new class of 4-amino-6-(pyrazol-4-yl)pyrimidine-based inhibitors.


Screening two fragment libraries, 35 fragments were identified as hits, in which 60% of them were found to bind to the adenosine triphosphate (ATP) site of Mab PurC by X-ray crystallography. Guided by structural information, the authors then made use of the addition and deletion of various functional groups to improve binding affinities. Importantly, the authors integrated fragment growing and merging strategies using hit fragments 1 and 2 to generate two compounds with significantly better binding affinities in the 50–150 nM range from a > 300 uM range. This method highlights the strength of FBDD to accelerate the overall hit-to-lead progression in drug discovery. Similar strategies were recently employed by Smith et al. to discover MRTX1719, a synthetic inhibitor to treat cancers with MTAP deletions.


“In this paper, we used both fragment growing and merging in the development of potent inhibitors of PurC. These strategies arose from the screening of an in-house fragment library as well as a fragment library at the Diamond synchrotron. We used the structural data from both of these screens to develop these inhibitors,” says Dr. Anthony Coyne, principal research associate at the University of Cambridge.


“While fragment growing is the most common method of elaboration, fragment merging has also been used in previous medicinal chemistry campaigns. However, a key aspect of these is the availability of high-quality structural data. In our work on PurC we were able to develop robust crystals for X-ray crystallography and during this project we were fortunate to solve over thirty structures that were crucial to our strategies. We have been able to develop high-affinity inhibitors that have been informed by structural biology across a wide range of different targets such as RAD51-BRCA2, MurB and TrmD. The interaction between medicinal chemistry along with structural biology is key to all the projects mentioned above,” Coyne adds.

Moving forward

FBDD has the potential to increase productivity in drug discovery, enabling more drugs to be approved for clinical use to benefit patients. However, fragments bind weakly to their targets, and continual progress is required in techniques such as X-ray crystallography and NMR to provide high-resolution structural information and analyze ligand-receptor interactions. New chemistry methods to synthesize, grow, merge and connect fragments will also facilitate faster hit-to-lead optimization. FBDD can also benefit from increased interest in artificial intelligence, in which using large datasets to predict key properties of fragments and their bioactivity and even creating virtual fragments may be able to accelerate drug discovery.