February was a packed month for interesting molecules – there were nearly 20 molecules we thought about including in this list but couldn’t fit them all. There are several clinical candidates this month including mobocertinib, an exon 20 mutant inhibitor that was recently given Breakthrough Therapy Designation by the FDA. Also included are several interesting modes of action here including two degraders, multiple protein dimerizers, and a mechanism-based inhibitor that modifies its target’s substrate. Finally, chemists will appreciate some interesting structures on this list as well as the femtomolar inhibitor CPI-1328 with a reversible potency among the strongest ever recorded.
Small Molecules of the Month - Feb. 2021
send download link to:
- The Bristol Myers Squibb (BMS)/Celgene GSPT1 degrader (CC-90009) is a CRBN-based molecular glue (CELMoD). It is a clinical candidate in Ph. I for AML and MDS, and in contrast to CC-92480 highlighted last March, spares the primary targets of prior imide drugs (IKZF1/3) and instead selectively degrades GSPT1. The starting point was identified through a phenotypic screen against a panel of AML cell lines using their cereblon (CRBN) modulator library and the nontumor human epithelial cell line (THLE-2) as a counterscreen. Clinically, at a dose of 2.4 mg, 90% GSPT1 reduction was observed, representing an important proof-of-concept for expanding the scope of molecular glue-type degraders.
- The Novartis LTA4H metalloenzyme inhibitor, LYS006, is a selective oral agent in multiple Ph. II studies to treat inflammatory diseases including ulcerative colitis and NASH. The starting points for this picomolar inhibitor were identified through a fragment based approach using differential scanning fluorimetry (DSF) as an initial binding assay, confirming hits with X-ray crystallography. Structure based fragment-merging led to a remarkably potent amine lead, and early hERG and CYP inhibition signals were dealt with by introducing a carboxylic acid to the molecule.
- The Eisai STING agonist E7766 leads to tumor eradication in 9/10 treated animals in a syngeneic model when injected intratumorally, and is currently in a Ph. I study for advanced solid tumors. The molecule has an interesting macrocyclic structure, and behaves as a cyclic dinucleotide mimic with potent activity on all four human STING variants.
- The ARIAD/Takeda EGFR exon 20 insertion mutant (EGFRex20ins) inhibitor TAK-788 is a Breakthrough Therapy for patients with advanced non-small cell lung cancer (NSCLC) whose tumors harbor EGFR exon 20 insertion mutations. The appearance of exon 20 insertions is a common resistance mechanism to earlier generations of EGFR inhibitors including osimertinib, to which TAK-788 is structurally related. 1st and 2nd generation EGFR inhibitors partly depend on the fact that mutant forms of EGFR have destabilized inactive forms with reduced ATP affinity relative to wild-type (WT) EGFR, making the mutant forms easier to drug in cells. EGFR exon 20 mutants, however, have active sites that are very similar to WT, making tumors bearing these mutations difficult to drug without significant side effects due to WT inhibition. This oral covalent inhibitor is more potent against exon 20 mutants than WT in cells, and is also active against a range of other common EGFR mutations. While the selectivity is still modest and side effects that seem related to WT EGFR inhibition are observed clinically, the significant responses in exon 20 insertion-bearing tumors is encouraging.
- The Millenium/Takeda SUMO-activating enzyme (SAE) inhibitor TAK-981 is a mechanism-based inhibitor that covalently forms a SUMO substrate-inhibitor adduct that potently inhibits SAE, rather than reacting with SAE directly. The Takeda team has been able to successfully leverage this fascinating mechanism to develop potent inhibitors against a range of E1 enzymes including ATG7 and NEDD8-activating enzyme (NAE), one of which (pevonedistat) is currently in a Ph. III trial for AML. TAK-981 is currently in combination Ph. II trials for advanced solid tumors as an intravenous infusion. This Takeda platform is another great example of how chasing down mechanisms from a phenotypic screen can lead to a rich pipeline of new targets and clinical candidates beyond what was originally intended.
- The Bristol Myers Squibb (BMS) HPK1 kinase inhibitor, “compound 24” is an oral tool compound intended for cancer immunotherapy, with >50x selectivity against family members including GLK. The starting point was an IRAK4 kinase inhibitor identified from historical kinome selectivity data. A homology-model based on MST1 was used for optimization. Interestingly this uncharged kinase inhibitor possesses a cyclic lactone which binds to a backbone N-H in a co-crystal structure without hydrolizing and a primary alcohol which binds to a front-pocket aspartate. Oral BID dosing (100 mg/kg) in combination with an anti-PD-1 antibody led to a 100% cure rate in a syngeneic tumor model.
- The Arbutus PD-L1 inhibitor ARB-272572 is a preclinical molecule that potently induces PD-L1 dimerization and internalization and demonstrates oral activity (10 mg/kg QD) in a humanized mouse model. Biologic PD-1/PD-L1 inhibitors have had incredible success as cancer immunotherapy agents, and there has been significant accompanying interest in small molecule inhibitors. Given the crowdedness of the field, however, the Arbutus team also makes the case that small molecule PD-L1 inhibitors could be useful in alternate settings such as chronic hepatitis B, and demonstrate activity in an in vitro HBV assay as well.
- The EicOsis oral soluble epoxide hydrolase (sEH) inhibitor, EC5026, is a compound currently intended as a non-opioid analgesic and anti-inflammatory agent. It recently completed a Ph. I study in healthy volunteers without drug-related adverse events. Despite its small size, it has picomolar binding activity with the urea acting as an epoxide-opening transition-state mimic. The authors discuss a range of development considerations including their selection of a clinical path, IND-enabling studies, clinical findings, and even animal health development. This is a must-read for anyone interested in the overall drug discovery process beyond medicinal chemistry.
- The Shionogi BACE1 aspartyl protease inhibitor, atabecestat (JNJ-54861911) is an oral, brain-penetrant compound that was advanced into the EARLY Ph. IIb/III clinical trial for treatment of preclinical Alzheimer’s disease (AD) patients. The starting point was identified from a phenotypic HTS, and was the first thiazine-based BACE1 disclosed in the patent literature. The Shionogi team discloses that the trial was discontinued due to liver enzyme elevation, and subsequent analysis showed that the treatment group actually had dose-related cognitive worsening, in agreement with other company’s findings. The Shionogi team shares a lot of helpful development details including their studies to de-risk an early 500 nM in vitro hERG signal.
- The Haisco voltage-gated calcium channel (VGCC) alpha-2-delta subunit inhibitor, HSK16149, is a potent and potentially longer-acting gabapentinoid drug candidate for chronic pain over pregabalin, with a potentially wider therapeutic index for the central nervous system (CNS). The molecule is planned for a Ph. II/III trial in China for diabetic peripheral neuropathic pain. Identifying novel chemical matter around small ligands like pregabalin is challenging, but the Haisco team appears to have found a nice niche in this brain-penetrant molecule’s interesting bridged cyclobutane structure. Chemists here may enjoy the structure of a recently approved-in-Japan competitor molecule from Daiichi Sankyo, mirogabalin as well.
- The EMD Serono/Merck KGaA toll-like receptors TLR7/TLR8 inhibitor M5049 is an oral, twice-daily anti-inflammatory agent currently in a Ph. I trial for Lupus. The molecule potently blocks both synthetic ligands and natural endogenous RNA ligands such as microRNA and Alu RNA. TLR7 and TLR8 selective modulators (both agonists and antagonists) have been of significant recent interest due to the TLR’s key roles in immune signaling but have been challenging to develop since TLR7/8 are membrane proteins that are difficult to express, purify, and crystallize. The Serono team however was able to generate an interesting crystal structure showing that the molecule binds to the interface between two monomers, stabilizing an inactive dimer, preventing receptor activation.
- The 2nd generation Constellation EZH2 histone methyltransferase inhibitor, CPI-1328, is an extremely potent tool compound with reversible femtomolar activity against EZH2, and is one of the most potent reversible small inhibitors ever reported. Previous analyses have suggested an empirical barrier for non-covalent inhibitors around 10 pM. The molecule builds on a finding from their previous paper that a thiomethyl group is critical for increased drug residence time. This activity translated to in vivo activity at 25 mg/kg QD despite the comprehensive target coverage thought to be needed against EZH2 to elicit an effect.
- The Yale BRAF-targeting PROTAC SJF-0628 is a VHL-based heterobifunctional degrader with in vivo activity in a xenograft model (50 mg/kg BID IP). Though mutant BRAF is a well-validated target, drugging mutant forms selectively over wild-type continues to be a challenge. It had been hypothesized that modalities such as targeted degradation might be a viable strategy to achieve functional selectivity for mutant forms of BRAF. Interestingly, the authors show here that BRAFWT is degraded to a far lower extent due to a weaker ternary complex between the compound and E3 ligase in cells, despite lower selectivity of the parent BRAF inhibitor, vemurafenib. However, under different cellular contexts, the authors show that wild-type still appears to be degraded, which may still present a safety issue if healthy human dividing cells are still affected in vivo, or in certain combination regimens. Like any important proof-of-concept, this article raises new questions for the modality, such as how to predict or optimize a therapeutic index preclinically given such mechanistic complexity.
- The Novartis pan-serotype dengue virus (DENV) NS4B protein inhibitor, NITD-688, is a preclinical candidate with strong activity against all four serotypes of DENV in vitro and excellent oral efficacy in a mouse model. The starting point was identified through phenotypic screening of 1.5M compounds and counterselection against host factors and cytotoxicity. The target was identified through resistance mutations and binding confirmed by 15N-NMR. Allometric scaling was used to predict a human efficacious dose of 35 mg QD. Non-GLP studies showed good tolerability in dog up to 80 mg/kg, and QTc prolongation was not observed despite partial inhibition of hERG in vitro. The authors allude to designing a trial to ensure patients are recruited within 48 h of fever onset to make sure the inhibitor can be administered early enough to affect outcomes.
Links to Articles:
- CC-90009 – Bristol Myers Squibb (BMS)/Celgene GSPT1 degrader
- LYS006 – Novartis LTA4H metalloenzyme inhibitor
- E7766 (STING) – Eisai STING agonist
- Mobocertinib (TAK-788) – ARIAD/Takeda EGFR exon 20 insertion mutant (EGFRex20ins) inhibitor
- TAK-981 – Millenium/Takeda SUMO-activating enzyme (SAE) inhibitor
- ”Compound 24” – Bristol Myers Squibb (BMS) HPK1 kinase inhibitor
- ARB-272572 – Arbutus PD-L1 inhibitor
- EC5026 – EicOsis oral soluble epoxide hydrolase (sEH) inhibitor
- Atabecestat (JNJ-54861911) – Shionogi BACE1 aspartyl protease inhibitor
- HSK16149 – Haisco voltage-gated calcium channel (VGCC) alpha-2-delta subunit inhibitor
- M5049 – EMD Serono/Merck KGaA toll-like receptors TLR7/TLR8 inhibitor
- CPI-1328 – 2nd generation Constellation EZH2 histone methyltransferase inhibitor
- SJF-0628 – Yale BRAF-targeting PROTAC
- NITD-688 – Novartis pan-serotype dengue virus (DENV) NS4B protein inhibitor
Explore drughunter.com for more drug discovery articles.