SAR & lead optimization
70 articles
Why BRD9 PROTACs underperform in mice, and a humanized-E3 fix
- Widely used CRBN- and VHL-based BRD9 degraders such as dBRD9 and VZ185 are much less potent in mouse cells, even though BRD9 itself is highly conserved.
- The gap points to species differences in the E3 ligases, notably the Crbn I391 residue and structural variation in VHL, which likely hamper ternary complex formation.
- Expressing human CRBN or VHL, or introducing a CrbnI391V mutation, restored potency in mouse cells and improved BRD9 degradation in vivo.
Anyone who has watched a PROTAC look great in human lines and flat in a mouse efficacy study will recognize this. It is a reminder that degrader PK/PD in rodents can fail for E3 reasons rather than compound reasons, so a CrbnI391V or humanized-E3 model is worth setting up early. It also argues for profiling ternary complex formation across species before reading a negative in vivo result as a dead series.
Improving PROTAC-Mediated Degradation in Mouse Models: BRD9 as a Case Study ↗
Pyrroloquinoline antitubulin agents keep activity in paclitaxel- and daunorubicin-resistant cells
- A library of 1H-pyrrolo[3,2-c]quinolines was made and mapped by SAR, yielding compounds with low-nanomolar antiproliferative potency and very large selectivity over normal fibroblasts.
- The best derivatives kept activity in daunorubicin- and paclitaxel-resistant sublines and showed low efflux ratios, suggesting limited cross-resistance.
- A lead showed rapid absorption, moderate half-life and limited brain penetration, but zebrafish embryo toxicity pointed to a narrow therapeutic window.
Tubulin binders that dodge P-gp efflux are always of interest, and low efflux ratios plus retained activity in resistant lines is the right first filter. The zebrafish result is the honest sting in the tail: potent antimitotics often have a thin margin, so selectivity over fibroblasts in culture will not settle the safety question. I would look next at the binding site and how far the scaffold can be tuned toward a wider window.
Macrocyclic pyrazolotriazine CDK7 inhibitor combines kinome-wide selectivity with oral exposure
- Structure-based optimization of a 17-membered macrocyclic pyrazolo-triazine scaffold gave a potent CDK7 inhibitor.
- It showed no significant off-target activity across a panel of more than 400 kinases, addressing the poor selectivity of earlier CDK7 inhibitors.
- In cells it caused cell-cycle arrest, suppressed transcription and lowered c-Myc, especially in triple-negative breast cancer lines, with oral bioavailability and linear PK.
A clean result across 400+ kinases is rare for CDK7, where the covalent field has leaned on THZ-type warheads. A non-covalent macrocycle that is also orally exposed is the profile I'd actually want to pick up for a transcriptional-addiction indication. The open question is on-target toxicity, since selectivity among kinases won't spare you CDK7's role in normal transcription.
YKS-34: a PDE4B inhibitor with >1000-fold isoform selectivity for pulmonary fibrosis
- The authors set out to improve on the only approved PDE4B-preferential drug, nerandomilast, which is only about 9-fold selective over PDE4D.
- YKS-34 is a potent PDE4B inhibitor with more than 1000-fold selectivity over PDE4A, PDE4C and PDE4D.
- It suppressed TGF-β-driven fibroblast activation and LPS-induced TNF-α release in vitro, and showed favorable PK with accumulation in lung tissue.
PDE4D inhibition is what drives nausea and vomiting, so going from 9-fold to over 1000-fold is the right kind of optimization to chase. The pocket differences between B and D are tiny, which makes me curious what the binding mode looks like. I'd want to see this in a bleomycin model with a therapeutic dosing regimen before getting too excited.
Dialing hERG out of a DCN1–UBE2M inhibitor without losing potency
- The earlier tool compound NAcM-OPT blocked the DCN1–UBE2M interaction, was orally available and slowed xenograft growth, but carried significant hERG affinity that stalled development.
- The authors systematically modified the scaffold to cut hERG binding while keeping DCN1 engagement.
- Lead compound 110 matched the parent in biochemical and cellular potency, with better physicochemical properties and improved selectivity against hERG.
This is the unglamorous but necessary second act: a validated PPI chemotype that was dead on cardiac safety, rescued by SAR alone. The takeaway for anyone with a basic, lipophilic PPI binder is that hERG can often be engineered out without giving up the pharmacophore, so check your own series before abandoning it.
Perspective: how structure-guided design finally delivered FGFR3-selective inhibitors
- Second-generation pan-FGFR inhibitors proved the concept but were limited by toxicities from inhibiting FGFR1, FGFR2 and FGFR4.
- Despite high structural homology across the family, structure-guided work has produced the first FGFR3-selective inhibitors to reach clinical evaluation.
- The Perspective covers the biological rationale, the design strategies behind the selectivity and emerging indications, including bladder cancer and achondroplasia.
Isoform selectivity inside a highly conserved kinase family is a good test of how far structure-based design really goes, and FGFR3 now has clinical-stage answers. The achondroplasia angle is the interesting one: a chronic, pediatric-adjacent indication raises the tolerability bar far above oncology, which is exactly where selectivity pays off.
A pyrazine-2-carboxamide antimalarial that cures mice in one dose has entered phase 1
- A P. falciparum growth-inhibition screen identified a new pyrazine-2-carboxamide series for blood-stage malaria.
- Lead optimization improved anti-plasmodial potency and metabolic stability and reduced CYP2D6 inhibition risk, giving IWY357.
- A single dose was fully curative in a Pf-humanized SCID mouse model, resistance selection showed a high barrier, and the compound has progressed through GLP toxicology into phase 1.
A phenotypic hit taken all the way to the clinic with a single-dose cure in the humanized-mouse model is what the field has wanted since resistance to front-line agents began eroding. The CYP2D6 fix matters for combinations, which is where a short-course regimen would actually live. I would watch the human PK, since a single-dose claim lives or dies on exposure.
Replacing saruparib's carboxamide yields an oral, brain-penetrant PARP1-selective inhibitor
- Aimed at CNS malignancies, the program replaced the carboxamide of saruparib with a series of heterocycles, targeting PARP1 over PARP2 to limit haematological toxicity.
- Caco-2 efflux, rather than the usual MDR1-MDCK assay, was used as a progression filter to select compounds with good unbound brain exposure.
- The lead compound was orally bioavailable and caused tumour regression in xenografts at a low dose, with a projected low clinical dose.
Swapping the primary carboxamide is a bold move, since that group anchors the nicotinamide pocket, and the heterocycles evidently kept the binding. The more useful lesson is the filter: if Caco-2 efflux predicts unbound brain exposure better than MDR1-MDCK, plenty of CNS programs may be discarding good compounds.
Pairwise-difference QSAR with quantile aggregation extrapolates beyond known potency
- Standard QSAR struggles to predict analogs more potent than anything in its training set, which is exactly the lead-optimization question.
- Pairwise-Difference QSAR learns activity changes from descriptor differences between compound pairs, then aggregates the predictions.
- On ChEMBL-derived benchmarks across about a hundred targets with the most potent compounds held out, a low-quantile aggregation beat mean and kNN aggregation, direct QSAR and Chemprop.
Every optimization team has felt this: the model ranks the series fine, then fails at the top end where you actually need it. Holding out the most potent decile is the right stress test. My caution is that public ChEMBL tails differ from a live program's SAR cliffs, so I'd trial it prospectively on one of our own series before trusting it to prioritize synthesis.
Breaking the Potency Barrier: A Practical Recipe for QSAR Extrapolation ↗
Adamantane-capped indole hydroxamate YP200 pushes HDAC11 selectivity into a lung cancer PDX model
- Starting from the earlier lead PB94, the authors optimized the core, linker and cap of an indole-based hydroxamic acid series against HDAC11.
- The best compound, YP200, combines a secondary-amine linker with an adamantane surface-recognition cap.
- It showed high potency for HDAC11 with strong selectivity over other HDAC isoforms, and suppressed tumor growth in a patient-derived NSCLC xenograft.
Hydroxamates are rarely clean across isoforms, so a bulky adamantane cap doing the selectivity work is the part I'd chase: it suggests HDAC11's narrow channel rewards cap size more than zinc-binder tweaks. Given the recent probe-validity debate around HDAC11 tools, I'd want kinetic isoform panels and cellular target engagement before trusting the efficacy story.
Elevenostat is a multi-HDAC inhibitor biased to HDAC6, not an HDAC11-selective probe
- Re-testing Elevenostat (JB3-22) against recombinant HDAC isoforms with two orthogonal HDAC11 assays showed only weak HDAC11 inhibition.
- Full HDAC1–11 profiling found much stronger inhibition of HDAC6 and HDAC10, plus class I enzymes, with HDAC11 the least sensitive isoform.
- Cellular readouts (tubulin and histone acetylation, reduced viability) matched HDAC6 and class I engagement, so past HDAC11 conclusions drawn from this compound need re-reading.
This is the unglamorous but valuable kind of paper: a compound used as an HDAC11 tool turns out to hit HDAC6 hardest. Any phenotype published with Elevenostat now has an attribution problem. Before I reuse an inhibitor as a probe, I want full-panel selectivity and a matched inactive control, which this case argues for.
Aminopyrimidine NaV1.7 probe inhibits resting and inactivated states almost equally
- An automated patch-clamp screening cascade, built to deprioritize classic local-anesthetic pore binders, yielded an aminopyrimidine hit that was optimized into compound 6.
- Compound 6 combined balanced resting- and inactivated-state potency, selectivity over NaV1.5 and a tetrodotoxin-like isoform profile that includes NaV1.1.
- SAR pointed to the aminopyrimidine core and a methylene-linked aniline as key determinants of this profile.
State dependence is a leading suspect for why NaV1.7 programs keep failing clinically, so a probe that blocks resting and inactivated states alike is a good way to test that hypothesis. NaV1.1 activity is a flag for seizure-related liability, and I'd want a binding-site assignment before reading too much into the profile.
A Nontoxin Aminopyrimidine Probe for Near State-Independent Na V 1.7 Inhibition ↗
Scaffold-hopped BCL6 PROTAC degrades its target in DLBCL cells and works orally in xenografts
- Scaffold hopping from the BCL6 ligand BI-3802 gave a new tricyclic oxazinoquinolinone binder.
- Linker and CRBN-recruiter tuning produced compound 15, a PROTAC that degraded BCL6 quickly and potently, in a CRBN- and proteasome-dependent way.
- It strongly suppressed BCL6-dependent DLBCL cell growth, and oral dosing inhibited tumor growth in an OCI-LY1 xenograft without evident toxicity.
BI-3802 is a polymerizing degrader with a peculiar mechanism, so a conventional CRBN-recruiting PROTAC from a fresh tricyclic core is a useful alternative route to BCL6 removal. Oral efficacy matters most here; I'd want to see how the molecule balances size and permeability, and whether degradation depth tracks with tumor response across non-OCI-LY1 lines.
Bicyclic amine swap widens the G2019S window in indazole LRRK2 inhibitors
- Structure-guided hybridization produced an indazole-based LRRK2 series; the initial piperidine compounds were potent but barely distinguished wild-type from G2019S mutant kinase.
- Replacing the piperidine with an octahydrocyclopenta[c]pyrrole widened the selectivity window while keeping good potency; compound 34 represents the new chemotype.
- The authors pitch it as an alternative starting point for next-generation, Parkinson's-directed LRRK2 inhibitors.
Locking a floppy piperidine into a fused bicycle is the textbook way to sharpen selectivity, and it seems to have paid off here. Before getting attached I'd check what the extra carbon framework costs in lipophilicity and brain exposure, because a CNS target punishes both. Mutant-over-wild-type selectivity also needs to hold up in cells, not just in biochemical assays.
First thiazolamide PI4KIIIβ/HDAC dual inhibitors take a host-directed route against HCV
- Host-directed dual inhibition of PI4KIIIβ and HDACs had not previously been reported for hepatitis C, where resistance to direct-acting antivirals remains a concern.
- A series of thiazolamide-based dual inhibitors was designed, synthesised and evaluated.
- Lead compounds showed potent enzymatic inhibition of both targets together with anti-HCV activity.
With DAAs curing most patients, the case for a host-directed HCV drug is hard to make commercially, and a dual PI4KIIIβ/HDAC profile adds tox risk. I read this as a good exercise in multitarget design, but would want to see cytotoxicity and kinase selectivity before calling the window real.
Millepachine hybrid hits tubulin and HDAC8 together in ovarian cancer models
- The natural product millepachine was fused with a key HDAC-inhibitor pharmacophore to build dual tubulin/HDAC inhibitors.
- Lead compound 6 was broadly antiproliferative across 12 cancer lines, inhibited tubulin polymerisation and HDAC8, and was selective for HDAC8 over HDAC1, 4 and 6.
- It suppressed tumour growth in an ovarian cancer xenograft with a favourable safety profile.
HDAC8 is a less crowded choice than the usual HDAC6/class I pairing, and pairing it with a tubulin binder is mechanistically coherent. My worry is the classic one for dual hybrids: the two activities rarely scale together, so I'd want evidence that HDAC8 inhibition, not just the tubulin arm, contributes to the in vivo effect.
Fixing solubility and CLK4 off-target liability in STK33 male-contraceptive indazoles
- SAR and property work on the 3-(biphenylethynyl)-1H-indazole scaffold behind CDD-2807 targeted its poor kinetic solubility.
- CDD-3094 kept nanomolar cellular potency while gaining solubility, lowering hERG liability and showing a clean broad kinase panel.
- A co-crystal structure guided changes against the main off-target CLK4, giving a strongly selective biochemical tool whose cellular potency dropped.
This is a textbook potency-versus-properties trade: solubility fixed in CDD-3094, then selectivity bought at the price of cell activity. For a contraceptive, where the safety bar is far higher than in oncology, the CLK4 discrimination matters more than raw potency; the next step is recovering permeability.
Donepezil-derived dual BuChE/FAAH inhibitors for multi-target Alzheimer's therapy
- Donepezil-based derivatives were designed and made as dual inhibitors of human butyrylcholinesterase and fatty acid amide hydrolase.
- Lead G15 inhibited both enzymes in a balanced way, with sub-micromolar to low-micromolar activity.
- It showed measurable selectivity over the carboxylesterases CES1 and CES2, a common liability for serine-hydrolase inhibitors.
Dual-target hybrids rarely land balanced potency on both enzymes, so G15 is a useful starting point. Selectivity over carboxylesterases is only modest, though, and I'd want in vivo exposure data before believing the multi-target rationale.
Brain-penetrant colchicine-site tubulin ligand nominated for glioma
- Only temozolomide and vorasidenib are approved brain-permeable, orally available drugs for brain cancers, so the team sought new brain-penetrant options.
- SAR work yielded a new class of highly brain-penetrant tubulin ligands; co-crystal structures confirm colchicine-site binding, with the tubulin βT7 and αT5 loops flipped relative to the colchicine-bound state.
- SAR showed high sensitivity to substitution with few indispensable moieties, leading to compound 40 as the clinical candidate.
Tubulin binders that actually cross the blood-brain barrier usually die on P-gp efflux or neurotoxicity, so a tight series with strong brain exposure is notable. The flipped βT7/αT5 loops hint at a binding mode worth docking my own colchicine-site scaffolds against. Chronic neurotoxicity is still the question.
Chimeric Hsp90β-selective inhibitors keep selectivity while tuning PK
- Pan-Hsp90 inhibitors stalled clinically because of their broad inhibition, so Hsp90β-selective indazolone compounds have been pursued, though earlier ones needed long optimization to become drug-like.
- This work modifies the northern indazolone fragment and the southwestern appendage at the same time, testing whether the best traits of each region can be combined.
- The resulting chimeric inhibitors retain Hsp90β affinity and selectivity while shifting key pharmacokinetic parameters.
Merging two independently optimized vectors is a sensible, low-risk move, and selectivity surviving the merge suggests the β-specific pocket tolerates more change than I feared. What would decide it for me is tissue exposure and tolerability, since that is where pan-inhibitors failed.
DEL hit becomes brain-penetrant reversible MAGL inhibitor MAGLi 432
- A bicyclopiperazine hit from a DNA-encoded library was optimized by structure-guided design into compound 1.5 (MAGLi 432), a reversible MAGL inhibitor.
- The compound combines high potency, selectivity over other serine hydrolases and CNS drug-like properties, with a favorable oral half-life in rats.
- It inhibited MAGL in human brain homogenate and worked in mouse neuroinflammation models; the campaign was steered by a translational "Five-Star Matrix" framework.
Reversibility is the part I'd want here, since irreversible MAGL agents ran into CB1 desensitization on chronic dosing. A non-covalent DEL chemotype that already shows target engagement in human brain tissue is a good template, and I'd steal the habit of testing in human tissue early rather than at the end.
Lipophilicity-controlled optimization delivers PERK inhibitor APL-11056
- Starting from a literature-derived PERK inhibitor of moderate potency but good kinase selectivity and PK, the team ran a lead optimization campaign steered by log D and lipophilic ligand efficiency.
- This produced APL-11056 (compound 49), a potent and selective PERK inhibitor with excellent in vivo pharmacokinetics.
- High permeability is linked to a fluorine alpha to the core benzamide, which appears to limit efflux-transporter recognition.
Letting LLE drive the SAR instead of chasing potency is the discipline that makes this readable, and the alpha-fluoro benzamide trick for dodging efflux is a handy one for my own amide series. PERK's ISR biology has bitten inhibitors before, so the tolerability data will matter more than the PK table.
Chiral imidazoquinolines show stereochemistry-dependent TLR7/8 agonism and antagonism
- A library of chiral imidazoquinolines was made and profiled in HEK-Blue hTLR7 and hTLR8 reporter assays.
- The SAR identified a potent dual TLR7/8 agonist and, from the same family, a dual antagonist.
- Activity depended on stereochemistry, likely because the two enantiomers interact differently in the TLR7/8 binding pocket.
Seeing both an agonist and an antagonist come out of one imidazoquinoline family, steered by stereochemistry, is a good argument for making both enantiomers early. For an adjuvant or autoimmune program, that switch is useful, but reporter-cell data alone only gets it to a hit; cytokine profiling in primary cells would tell me whether it holds up.
4-Aminoquinazoline PKMYT1 inhibitors cut H-bond donors for CCNE1-amplified tumors
- A hydrogen-bond donor-to-acceptor conversion was used to lower the donor count of a 4-aminoquinazoline PKMYT1 series and improve drug-likeness.
- Lead 12f was a nanomolar enzyme inhibitor, potent in OVCAR-3 ovarian cells and synergistic with PARP inhibitors.
- With acceptable oral PK, it showed efficacy in an OVCAR-3 xenograft and was well tolerated.
Trimming HBDs is a clean, transferable tactic for improving permeability and oral exposure in a kinase series. The PARP-inhibitor synergy is the commercially interesting part, but PKMYT1 is crowded, so I'd check selectivity against other kinases and how 12f compares with the clinical frontrunners.
Istradefylline-derived thienopyrimidinones give an oral NLRP3 inhibitor for colitis
- The anti-Parkinsonian drug istradefylline showed modest NLRP3 inhibition, and optimizing it gave a thienopyrimidinone series.
- Lead L22 was roughly tenfold more potent than istradefylline, bound the NLRP3 NACHT domain directly and blocked inflammasome assembly.
- It was orally bioavailable, eased symptoms in a DSS colitis model and looked safe in acute and subacute toxicity studies.
Starting from a marketed drug's off-target activity gives a head start on drug-likeness, and direct NACHT-domain binding is what you want to see. NLRP3 is a busy field, so the real test is how L22 compares with clinical-stage inhibitors on potency and brain or gut exposure.
Pyridazine nicotinamides as a new FTO inhibitor chemotype for AML
- A 67-compound SAR campaign on pyridazine-containing nicotinamides gave E6 as the most active FTO inhibitor, with binding confirmed by MST, CETSA and DARTS.
- E6 was active against MOLM-13 cells, raised global m6A, upregulated ASB2 and RARA and lowered c-Myc.
- It showed acceptable CYP safety and PK and inhibited MOLM-13 xenograft growth.
Another FTO series, but the useful part is the new scaffold: FTO chemotypes are still thin, and most earlier ones lean on the same few motifs. Potency is modest, so I'd treat E6 as a starting point and look for selectivity over ALKBH5 and other AlkB family members before reading too much into the cell phenotype.
One atom flips a CRBN degrader from NSD2-selective to GSPT1-selective
- Starting from LLC0424, an NSD2 degrader that also engages GSPT1, a single-atom change gave two close analogues: 424-ND (NSD2-selective) and 424-GD (GSPT1-selective).
- 424-ND suppressed androgen receptor signaling in prostate cancer cells, while 424-GD triggered the integrated stress response.
- Biolayer interferometry showed distinct ternary-complex cooperativity for each target, and simulations suggested different low-energy conformations explain the bias.
Degrader selectivity here tracks ternary-complex cooperativity, not binary affinity, which is exactly what I'd want to see in an SAR table. A matched pair this tight is a ready-made probe set for NSD2 and GSPT1. I'd check proteome-wide selectivity and whether the cooperativity trend holds in cells.
Single-Atom Modification Enables Reversal of Degradation Selectivity between NSD2 and GSPT1 ↗
Scope library: a chemical space between fragments and full molecules for tractable hits
- Scope defines a chemical space between fragments and elaborated low-molecular-weight library compounds, emphasizing compound quality, ligand efficiency and selected property ranges to keep hits tractable for medicinal chemistry.
- Scope ring systems (ring-X-ring) organize the space and guide intentional functionalization toward low- to mid-micromolar potency; a roughly 10k-compound set suits medium-throughput SPR, native MS, DSF and biochemical screens.
- Case studies on trypanosomal CLK1, a 14-3-3/ERα molecular glue project and SARS-CoV-2 Mpro show comparable primary hit rates to diversity sets, with better ligand efficiency after counter-screening.
Honestly, the pitch I like here is hits you can actually start working on: a library sitting between fragments and full-size compounds, built around ligand efficiency, with ring-X-ring cores that tell you where to functionalize next. A 10k set fits SPR, native MS and DSF, so a smaller group can run it. The open question is whether primary hit rates on par with diversity sets, with better LE after counter-screening, hold up beyond three targets, and how much the scaffold constraint narrows novelty.
Structure-guided design yields selective wild-type KIT inhibitors that avoid class III RTK off-targets
- Close ATP-site similarity across class III receptor tyrosine kinases makes selective KIT inhibition hard; this work tackles it with structure-guided optimization.
- A triazolopyridine hinge binder gave a selective start, and a central linker change led to a reverse arylacetamide series with better potency and pharmacokinetics.
- A pyrazole linker and hinge-motif edits kept potency in human serum, while terminal aryl tuning reduced CYP time-dependent inhibition and sharpened kinome selectivity.
- Crystal structures showed that linker stereoelectronics near the gatekeeper can drive selectivity, yielding a lead suited to chronic therapy.
The linker, not the hinge binder, is what caught my eye: stereoelectronics near the gatekeeper apparently buy selectivity that ATP-site similarity across class III RTKs would otherwise deny. That is a handle worth stealing for other crowded kinase families. I also like that CYP time-dependent inhibition and human-serum potency were fixed in the same campaign, which reads like a lead meant for chronic dosing.
A Quinazoline Nav1.7 Lead, Absolute Configuration Fixed by MicroED, Shows Oral Analgesia in Rodents
- Starting from a modest focused-library quinazoline hit, SAR across about twenty analogs gave a chiral lead with sub-micromolar potency and clear selectivity over Nav1.5.
- The absolute configuration was assigned by MicroED, so the active enantiomer is unambiguous without large crystals.
- The lead had acceptable oral exposure, reduced pain behavior in two rodent models, and was clean in a 44-target off-panel; Nav1.7-knockout neuron controls support on-target action.
Nav1.7 has burned a lot of programs on selectivity and translation, so a clean off-target panel plus knockout-neuron controls is what I'd look at first. Potency is still modest and the dose needed is high, so this is a starting scaffold rather than a candidate. The practical takeaway is MicroED: assigning chirality from nanocrystals of an early lead saves weeks of enantiomer chasing.