Introduction to the themed collection on fragment-based drug discovery

When fragment-based drug discovery (FBDD) began more than 25 years ago it was met with scepticism and many questions: will very small molecules (MW 150–250 Da) really bind specifically to proteins? How will such weak binders be detected? And how long will it take to increase the affinity from millimolar to nanomolar? But there were compelling reasons to try. Smaller molecules sample chemical space much more effectively than typical high-throughput screening (HTS) libraries, so fewer compounds are screened and the hit rate is higher. Furthermore, if a fragment can be observed by X-ray crystallography, there is an opportunity to utilise structure-based drug design (SBDD) to incorporate new binding interactions and increase affinity through fragment growing, merging, or linking, whilst paying close attention to overall physicochemical properties.

Today, FBDD is widely established throughout pharma, biotech, and universities for discovering new biological tool molecules and medicines. Six marketed drugs come from FBDD, including four kinase inhibitors for oncology, a Bcl-2 protein–protein interaction (PPI) disruptor, venetoclax, and a covalent KRAS inhibitor, sotorasib. This number looks set to increase because the FBDD pipeline has delivered over 50 compounds to the clinic, and around 25 fragment-to-lead (F2L) success stories are published annually.

This journey from disruptive technology to established tool has involved close collaboration between medicinal chemists, computational chemists, X-ray crystallographers, biophysicists, chemical biologists, biochemists, and other bioscientists. The most widely used biophysical screening methods include NMR, ITC, SPR, thermal shift, and mass spectrometry. Medicinal chemists utilise automation and high-throughput experimentation for F2L synthesis, and new challenges have emerged such as selective C–H bond activation in the presence of the unprotected polar functional groups required for fragment–protein binding.

One potential limitation of FBDD is the range of targetable protein families. The many examples of kinases and other enzymes as well as PPIs reflect the availability of high-resolution protein structures, but relatively few membrane proteins have been targeted by FBDD. This may be about to change because cryo-electron microscopy has led to breakthroughs for transmembrane proteins. The technique has been shown to detect fragment binding and may therefore allow FBDD to be applied more widely. As the pool of targets and protein families increases, we may find out if there are certain properties regarding the binding pocket required for successful FBDD.

Of course, informatics and artificial intelligence (AI) are also having an impact, for instance to understand the properties of a good fragment. Computational methods like free energy perturbation to design target molecules and protein–ligand structure predictions, which allow SBDD for targets without an experimentally determined high-resolution 3D structure, are also making great strides.

The articles in this special themed collection nicely capture many of these trends: ¹⁹F NMR spectroscopy to study fragment–protein complexes, next-generation fragment screening libraries, FBDD-derived CK2α inhibitors with a novel mechanism of action, fragments binding to Y220C mutated p53, photoactivatable DNA-encoded fragments, and open-source AI for fragment library design.

In addition to illustrating the multidisciplinary nature of FBDD, these articles also demonstrate the worldwide interest in the subject, with contributors from Germany, Japan, the Netherlands, the People's Republic of China, Singapore, South Korea, Sweden, the United Kingdom, the United States, and elsewhere. FBDD is truly a global effort, and its practitioners – regardless of their specialty – form a close-knit community. If you are new to the field, welcome. We hope you enjoy reading these articles as much as we did!

---