Drug Discovery Literature Update July 2026

Orion Pharma, Pfizer, Johnson & Johnson, Bristol Myers Squibb, Gilead, Zoetis, Teva Pharmaceuticals
https://doi.org/10.1021/acs.jmedchem.6c01042

Fluorine is one of the most useful atoms in the medicinal chemist's toolbox for modulating the properties of a molecule. It is particularly useful for improving metabolic stability at soft spots and increasing the permeability of compounds by adding lipophilicity. Terminal CF₃ and CF₂ placed within chains or rings are by far the most common groups used.

Updated definitions of what comprises Per- and polyfluoroalkyl substances (PFAS) include these CF₃ and CF₂ groups. This is concerning, as it means that PFAS regulations could also apply to pharmaceutical research and production, which could delay or prevent new medicines from reaching patients.

This article was written by the IQ Consortium, which contains members from across the pharmaceutical and biotech sectors. They surveyed medicinal chemistry departments across 15 member companies on the use of these groups and how new regulations would impact drug research and development. The majority of responses indicate that these groups play a vital role and would not be easy to replace.

Amgen
https://doi.org/10.1021/acs.jmedchem.6c01357

Amgen's sotorasib was the first KRAS G12C inhibitor to enter clinical trials and was approved by the FDA in May 2021. Since then, many other compounds have entered clinical trials, including the approved adagrasib. Patients with KRAS G12C NSCLC often develop brain metastases, but many KRAS G12C inhibitors show low CNS penetration in preclinical models, thought to arise from their physicochemical properties.

To address this, Amgen applied the concept of CNS multiparameter optimisation to DNA-encoded library design. This allowed them to identify inhibitors with more favourable properties to reach the CNS. One compound was optimised, increasing potency by ~200-fold while balancing the unbound brain-to-plasma concentration ratio.

Eindhoven University of Technology, University of Dundee, Ambagon Therapeutics
https://doi.org/10.1038/s41467-026-75333-w

PROTACs rely on the availability of a ligand that binds the protein of interest. However, some proteins lack well-defined pockets or contain disordered regions, which makes identifying ligands challenging.

In this work, the authors developed a molecular glue-PROTAC hybrid to induce ubiquitination of the 14-3-3/Estrogen Receptor α (ERα) complex. They used 3'-deacetylated fusicoccin N-acetyl (FC-NAc), a molecular glue which stabilises the interaction between the hub protein 14-3-3 and the disordered C-terminal F-domain of ERα. A series of chemical linkers was used to attach a VHL ligand to FC-NAc. Cooperativity was linker-dependent, with an 11-atom alkyl linker giving the strongest VHL recruitment, and cryo-EM revealed a de novo interface between VHL and 14-3-3.

Wei Du, Weill Cornell Medicine
https://doi.org/10.1038/s41467-026-75591-8

Degrader efficacy depends on many factors, but how these combine is poorly defined. In this work, the authors built a quantitative model of degrader efficacy using eight parameters. Some of these are degrader-independent, such as endogenous target protein turnover and E3 ligase abundance. Others are degrader-specific, such as target affinity and complex cooperativity. Most of them are available from public proteomics data, or can be measured using well-defined methods.

The model was first applied to well-studied BET degraders (dBET1 and MZ1). Extending it to promiscuous kinase degraders then allowed testing across a much broader set of targets. The framework is also used to flag targets predicted to be highly degradable, including transcription factors, epigenetic regulators and GPCRs with long protein half-lives.