This was a loaded month for pioneering drug discovery programs! Among the highlights are a reversible-covalent inhibitor from Novartis, the hotly followed SHP2 inhibitor from Novartis, several innovative compounds from Pfizer including an oral GLP-1R agonist with a cryo-EM structure, a liver-targeted ACC inhibitor, and an oral KHK inhibitor. Remarkably, half of the molecules are negatively charged or pro-drugs of anions, and several have oral bioavailability beyond what a chemist 20 years ago would have thought probable. Most of these don’t look like they would be low-clearance, high bioavailability drugs, but they are! Detailed comments and links below. Also by popular request, SMILES strings of the molecules are now included in the subscriber downloads/e-mails, for those of you that don’t want to spend your weekends drawing “compound 9” like me. 🙂

Small Molecules of the Month - Sep. 2020

Subscribe to get a link to download a high-resolution PDF of September 2020's "Small Molecules of the Month" and a list of the molecules' SMILES strings.

send download link to:

I confirm that I have read and agree to the Terms of Use and Privacy Policy.

FGF401 (roblitinib) is a highly selective, reversible-covalent inhibitor of the kinase FGFR4, a receptor that can drive progression of certain cancers. The Novartis team found that the FGFR4 protein has a high re-synthesis rate and therefore would require significant drug exposure to have a meaningful effect, so they deprioritized irreversible covalent starting points and pursued a potentially safer reversible-covalent strategy instead. The aldehyde of roblitinib targets a unique Cys by the hinge region of the protein (GK+2) that is not present in the other FGFR’s and is present in only 4 other kinases. It is remarkable that they were able to optimize an aldehyde for clinical oral dosing, and given the recent approval of another aldehyde drug, voxelotor, it is certain that aldehydes will become a more widely tested drug motif in the future. This Novartis work is certain to be regarded as a modern classic in drug discovery for some time.

TNO155 is an allosteric inhibitor of the SHP2 phosphatase, which has been a hot target in oncology recently due to its promotion of mutant KRAS activity as well as its potential role in cancer immune evasion. Phosphatases have been notoriously difficult to drug however, with reported phosphatase active site inhibitors typically having low potency, low selectivity, and poor properties due to the conserved, polar nature of the active site. Hence the Novartis team pursued allosteric inhibitors, and discovered a lead with a remarkable binding mode that stabilizes an inactive SHP2 conformation through interactions between three distinct protein domains. The primary amine in the molecule displaces a structural water and makes direct interactions with 3 different residues. TNO155 is the first allosteric SHP2 inhibitor to enter clinical trials and will be closely watched by many.

BMS-986143 is a potent, reversible inhibitor of the Bruton’s tyrosine kinase (BTK), which has been successfully targeted by a number of irreversible drugs for oncology. The safety bar is much higher for inflammatory conditions, however, which led the BMS team to pursue reversible inhibitors. The two defined axial stereocenters of compounds like 986413 were introduced as a way to address earlier observed in vivo safety issues with a prior lead that had free rotation about both axes and could be more promiscuous. I found it surprising that these molecules are not only configurationally stable but also stable to metabolism and hydrolysis of the pyridopyrimidinedione heterocycle in vivo. Amazingly, the BMS team took a related molecule into clinical efficacy studies, the >200 kg-scale, atropstereoselective synthesis of which we highlighted previously.

AZD9833 is a potent, orally available selective estrogen receptor degrader (SERD) and antagonist in development for breast cancer. The only approved SERD is the widely used fulvestrant, which is given by monthly intramuscular injection and may have submaximal antitumor activity. A safe, oral alternative has been highly sought after and the oncology community looks forward to clinical efficacy data from the AZD team.

AL-611 is a prodrug to a nucleoside inhibitor of HCV NS5B polymerase. While there have been very successful nucleoside drugs (e.g. sofosbuvir), several nucleoside inhibitors of HCV previously failed in the clinic due to severe toxicities, in some cases due to incorporation of the nucleosides into host RNA. The Janssen Biopharma (formerly Alios) team tried to navigate safety issues observed with other nucleosides by increasing the complexity of the sugar moiety. The presence of two fully-substituted, fluorine-bearing stereogenic carbons caught my eye. AL-611 is orally available in dogs and was selected as a drug candidate but discontinued due to the effectiveness of already approved drugs.

PF-06882961 is an oral small molecule agonist of the glucagon-like peptide-1 receptor (GLP-1R). GLP-1R agonists increase the production of insulin and are highly useful in the treatment of diabetes. However, until the recent introduction of an oral formulation of semaglutide, GLP-1 agonists were all peptides requiring inconvenient injections. Small molecule agonists with convenient oral dosing are therefore desirable, but have been difficult to identify despite significant industry effort. The Pfizer team was able to identify weak agonist starting points (~70% @ 20 uM) using an innovative positive allosteric modulator-sensitized cell-based assay, which they would not have identified without sensitization. Significant optimization led to the ~100,000x more potent agonist PF-06882961 (13 nM in non-sensitized cell line). Despite additional challenges including the fact that a key residue difference in GLP-1 across species led to preclinical efficacy only being observed in monkeys, the Pfizer team was able to take PF-06882961 into development and demonstrate GLP-1R agonist activity in humans for the first time with a small molecule!

PF-05221304 is a liver targeted inhibitor of acetyl-CoA carboxylase (ACC). ACC is a key regulator of lipid metabolism, and there has been significant interest in ACC inhibitors for the treatment of multiple indications including NASH and diabetes. Pfizer previously took a systemic ACC inhibitor into the clinic (PF-05175157), which was found to exert the desired reduction on liver-based lipogenesis but was also found to have a side effect of reducing platelet counts due to activity in the bone marrow. To identify a potentially safer liver targeting molecule, the Pfizer team looked for molecules that were likely to be recognized by the liver uptake transporters, OATP1B1 and OATP1B3, and used a non-human primate model to confirm evaluate the therapeutic index. PF-05221304 was well-tolerated and showed desired activity in healthy human volunteers without significant effects on platelet counts after 2 weeks of dosing. Readers may recognize a familiar name on two highlights this month, including this impressive work!

PF-06835919 is an oral first-in-class clinical candidate for NAFLD/NASH targeting ketohexokinase (KHK), an enzyme which initiates the metabolism of fructose. The authors describe a very elegant structure-based design campaign to improve a modestly active, quickly cleared lead molecule into this highly potent, permeable, bioavailable, and metabolically stable candidate. Key to success included the displacement of a water molecule and removal of three(!) hydroxyl groups from the original lead. It’s interesting that the molecule contains both an azetidine and azabicyclohexane motif, features that were once rare but now seem to be increasingly found in clinical molecules.

“compound 9” is an investigational prodrug of pibrentasvir (ABT-530), a key component of AbbVie’s marketed HCV combination Mavyret. Pibrentasvir (PIB) is the largest approved oral small molecule since 1983 (MW = 1113, not including the peptide semaglutide at MW = 4113), and came with some solubility issues (aq. sol. pH 7 = <0.0001 mg/mL). The poor solubility of PIB requires a high excipient/drug ratio for suitable absorption, and is the main reason Mavyret is given in a 3-pill, QD regimen. The AbbVie team investigated solubilizing prodrugs of PIB to enable a single pill formulation with bioequivalence to Mavyret, arriving at the significantly more soluble and bioavailable compound 9. Compound 9 is a trimethyl-lock prodrug, which does not depend on esterase activity for drug release but on a spontaneous intramolecular cyclization reaction after dephosphorylation by intestinal alkaline phosphatase.

GLPG1205 is a potent, selective, and orally bioavailable negative allosteric modulator of the GPCR GPR84, a receptor associated with inflammation. Despite possessing both a dioxane ring and an relatively unhindered alkyne, the molecule has low clearance and is highly bioavailable in preclinical species (>100% F). The molecule was taken into a Ph. IIa trial initially for IBD, but failed to meet its primary efficacy endpoint. Currently, it is being evaluated in a Ph. II study in idiopathic pulmonary fibrosis (IPF).

OP-5244 is a potent (250 pM), orally bioavailable inhibitor of the nucleotidase CD73, in contrast to the previously highlighted IV inhibitor in July. I thought it was remarkable that the ORIC team was able to identify an orally available, low clearance, and permeable compound with phosphonic acid motif and four other hydrogen bond donors (incl. 3 free hydroxyl groups). The authors label OP-5244 a tool compound, but it will be interesting to see if an oral candidate emerges against the backdrop of CD73 antibodies in development.

SCO-267 is a full agonist of the GPCR GPR40, whose activation stimulates secretions of insulin and incretin. Since no crystal structures were available, the authors used conformational modeling to rationally improve the ligand from a weak lipophilic initial starting point (0.4 uM, clogP 9.4) to a more lipophilically efficient candidate (12 nM, clogP = 5.8). The compound is orally active in a rodent obesity model and was advanced to a Ph. I study.

“compound 23” is a selective CRBN-based degrader of the pseudokinase, IRAK3, with in vitro activity in primary human cells. A volcano plot shows remarkable separation between IRAK3 degradation vs. most other proteins, with very little activity against family members IRAK1 and IRAK4. The authors were able to quickly generate this degrader from existing kinase inhibitors in their compound collection, showing how rapidly useful tools to assess structural roles of kinases can be generated through the degradation platform.

TAS-120 (futibatinib) is an orally available, irreversible FGFR1-4 inhibitor with a unique binding mode, targeting a unique cysteine in the P-loop of the FGFRs, rather than the Cys in the hinge region of FGFR4 as for roblitinib (above) or BLU9931 and fisogatinib. It is currently the most advanced irreversible FGFR inhibitor in clinical development with a Ph. III trial on-going.

Links to Articles:

  1. FGF401 (roblitinib)
  2. TNO155
  3. BMS-986143
  4. AZD9833
  5. AL-611
  6. PF-06882961
  7. PF-05221304
  8. PF-06835919
  9. “compound 9”
  10. GLPG1205
  11. OP-5244
  12. SCO-267
  13. “compound 23”
  14. TAS-120 (futibatinib)

Hope this was helpful. Explore for more.