Drug repurposing
10 articles
Approved anthracyclines and anthracenediones repurposed as single-molecule photosensitizers
- TD-DFT screening of a curated library of approved anticancer drugs picked out chromophore-bearing candidates with favourable photophysics.
- Mitoxantrone, pixantrone, daunorubicin and epirubicin generated reactive oxygen species on irradiation and showed enhanced cytotoxicity in breast cancer cells.
- One molecule thus supplies both the chemotherapy and the photodynamic effect, avoiding a two-component conjugate.
Starting from compounds with known PK and safety is the smart part: the CMC and tox story is already written. The catch is that these chromophores absorb at wavelengths that penetrate tissue poorly, and ROS from an anthracycline adds to cardiotoxicity worries. I'd want in vivo light-dose windows before getting excited.
Turning EGFR kinase inhibitors into BAX inhibitors with the kinase activity engineered out
- A virtual screen flagged EGFR inhibitors, including dacomitinib and gefitinib-like cores, as possible BAX binders.
- Systematic modification built a first SAR for BAX inhibition and reduced EGFR kinase activity considerably.
- The series showed better BAX inhibition in vitro, in cells and in iPSC-derived brain organoids, with drug-like properties.
Inhibiting an executioner protein like BAX is hard, so starting from a privileged kinase scaffold is pragmatic. The real test is whether the EGFR activity is gone enough: a cytoprotectant that still hits kinases is a poor neuroprotective. I'd like orthogonal proof of direct BAX engagement and a broader kinome panel.
Turning eltrombopag into DCL1191, an oral inhibitor of the 'undruggable' TFEB
- TFEB, a transcription factor long seen as undruggable, was approached starting from eltrombopag, which has weak TFEB inhibitory activity alongside its Tpo-R agonism.
- SAR optimisation gave DCL1191, which binds TFEB, disrupts TFEB–DNA binding and shows reduced Tpo-R and MITF activity; modelling and mutagenesis support the binding mode.
- It represses TFEB target genes, blocks autophagic flux and is orally bioavailable, with activity in autophagy-hyperactivated glioblastoma cells.
Starting from a marketed drug with a known off-target is a smart way into a transcription factor, because you inherit drug-like properties. The real test is the MITF/TFE3 family: if selectivity over close paralogues holds, this is a genuine tool compound rather than another autophagy-flavoured cytotoxin.
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.
Approved Ataxia Drug Repurposed to Degrade STING via 'Molecular Crowbar'
- Omaveloxolone is an FDA-approved Nrf2-activating drug, originally approved for Friedreich's ataxia, now repurposed as a STING-targeting agent.
- It binds the base of STING's ligand-binding domain, destabilizing the protein's folded structure.
- This triggers K27-linked polyubiquitination of STING and its subsequent degradation by the proteasome — a "molecular crowbar" mechanism.
- Unlike classical PROTAC-style bifunctional degraders, omaveloxolone acts as a single small molecule, directly destabilizing STING rather than using a ligand-linker-E3-ligase construct.
- In aged mice, omaveloxolone treatment reduced STING-pathway-driven inflammation, a hallmark of age-related "inflammaging."
A monovalent degrader that's neither a classic PROTAC nor a molecular glue — that's the interesting bit. "Molecular crowbar" reads like marketing, but mechanistically it sounds like ligand-induced conformational destabilization exposing a degron: closer to hydrophobic-tag logic than E3-recruitment chemistry, no linker or ternary-complex engineering required. Chemically elegant, if it holds up. The real draw is that omaveloxolone's safety and PK are already banked — repurposing means skipping a decade of tox work if this translates. But my triage instinct says this is serendipitous polypharmacology, not designed SAR; the Nrf2 pharmacophore was never built for STING's pocket. It's also secondary-sourced with no primary structural data yet — I'd want a co-crystal or HDX-MS confirming direct engagement before trusting the crowbar story over generic proteostasis stress.
AI Screening Pipeline Flags Repositioned CASP4 Inhibitor Candidates for Alzheimer's
- CASP4, an inflammatory caspase linked to Alzheimer's disease, is the target of a new computational drug-repositioning pipeline built to screen DrugBank compounds for potential inhibitors.
- The workflow combined docking-based prioritization with ChemBERTa language-model embeddings and physicochemical descriptors to train machine-learning classifiers.
- A Random Forest classifier and regressor were used to identify likely active compounds and rank them by predicted potency.
- Downstream cheminformatics, PK-PD evaluation, and molecular docking narrowed the pool to five repositioning candidates.
- Molecular dynamics simulations and MM/PBSA free-energy analysis were then applied to assess binding stability, identifying two compounds (DB05316 and DB00519) with predicted binding compared favorably to the reference drug donepezil.
- The authors present the pipeline as a general strategy for prioritizing repurposed compounds against novel drug targets.
A five-stage funnel — docking, ChemBERTa embeddings, Random Forest, then MD/MM-PBSA — down to two compounds looks disciplined, but each stage has its own error rate, and stacking filters doesn't guarantee the errors cancel out. MM/PBSA rescoring of structurally unrelated ligands is notoriously unreliable for absolute ranking; chain enough filters and you can end up polishing false positives rather than eliminating them. Where this actually sits in the AI hit-finding wave: not novel chemistry, but faster triage of known-compound space, which is a real and growing niche. The genuine advantage is DrugBank sourcing — DB05316 and DB00519 already carry ADMET, safety, and formulation data, so there's no synthesis or fresh tox package standing between candidate and IND-enabling work. None of that substitutes for a binding assay. Until CASP4 activity is confirmed biochemically, this is a ranked hypothesis list, not a hit.
ChemBERTa-Guided Virtual Screening Identifies Repurposable CASP4 Inhibitors for Alzheimer's Disease
- Ligand-based virtual screening pipeline mined DrugBank for CASP4 inhibitor candidates, combining docking-score prioritization with SMILES-derived ChemBERTa embeddings and physicochemical descriptors.
- A Random Forest classifier trained on the hybrid ChemBERTa/physicochemical feature set separated active from inactive compounds with high accuracy and enriched actives among top-ranked hits.
- A companion Random Forest regressor, trained on experimental pIC50 values, ranked candidates by predicted potency.
- Cheminformatics filtering, PK-PD evaluation, and molecular docking narrowed the field to five DrugBank-listed candidates.
- Molecular dynamics simulations and MM/PBSA free-energy analysis confirmed stable binding poses for the top candidates.
- Positions the workflow as a drug-repositioning route toward CASP4-targeted Alzheimer's disease therapy, a caspase linked to neuroinflammation and pyroptosis.
What actually sells this to me is the DrugBank sourcing, not the screening funnel itself: the five candidates that survive docking, ChemBERTa/RF classification, and MM/PBSA already come with known ADMET and need no de novo synthesis to reach an animal model — that's a tractability shortcut a genuine hit-finding campaign rarely gets. CASP4's pocket is the usual caspase catalytic Cys-His dyad, so the real question the MD trajectories should answer is whether these scaffolds actually engage that dyad or just sit in a peripheral exosite; binding energy alone won't tell me that. Selectivity is the open flag — caspase-1/3/4 share fold homology, and off-target caspase hits in a neuroinflammation context could cut the wrong way. Methodologically, ChemBERTa embeddings layered onto classical descriptors for RF triage is a sane upgrade over pure docking-score ranking, though I'd want held-out chemotype validation before trusting the pIC50 regressor outside its training distribution.
HTS Screen Identifies CGP7930, a Known GABA-B Modulator, as a Potent Activator of the Cardiac Mitochondrial Calcium Uniporter
- Researchers screened a library of 1,280 known bioactive compounds in search of pharmacological activators of the mitochondrial calcium uniporter complex (MCUc), a target for which potent, selective agonists have been scarce.
- The screen identified CGP7930 — a compound previously characterized as a GABA-B receptor allosteric modulator — as a potent activator of mitochondrial calcium uptake, highlighting a drug-repurposing angle for a difficult target.
- Mechanistic follow-up showed the activating effect depends on MICU1, a key regulatory subunit of the uniporter complex, pointing to a specific mode of action rather than a nonspecific membrane effect.
- Enhanced mitochondrial calcium entry translated into improved cardiac contractile bioenergetics, suggesting relevance to conditions involving impaired heart energy metabolism.
- The work provides a valuable pharmacological tool compound and a starting point for developing more selective next-generation MCU activators.
CGP7930 landing as an MCU activator is the kind of hit that's simultaneously a gift and a headache. MICU1-dependence is the real finding here — it means the compound isn't just permeabilizing membranes or brute-forcing the pore, it's engaging a regulatory interface, which is exactly what you want if you're hunting for a selectivity handle to build SAR around. But this molecule already has a day job as a GABA-B PAM, so before I'd touch it as a lead I'd want a counter-screen and a clean SAR series that walks GABA-B affinity down while holding MCU activation up — classic polypharmacology triage, not a bonus. As a repurposed known compound it buys you PK/tox data, not real IP or optimized potency; it's a starting point, not a lead. Given how thin the MCU tool-compound cupboard has been, though, any validated chemical matter here is worth chasing.
Tideglusib Identified as First Chemical Probe for NatA Acetyltransferase via Novel Covalent Mechanism
- NatA catalyzes a large share of eukaryotic Nα-acetylation, a co-translational modification important for protein stability, localization and function; no validated small-molecule NatA probe had previously been reported.
- Screening an in-house ~700-compound library and optimizing a thiadiazolidinedione scaffold identified tideglusib (YD3139) — a clinical-stage GSK-3β inhibitor — as a hit, repurposing a known drug toward NatA.
- Mechanistic studies showed tideglusib first binds NatA allosterically, after which CoA released during catalysis chemically modifies it into a reversible covalent adduct, a previously undescribed catalysis-dependent inhibition mechanism.
- The compound showed low-micromolar potency against yeast and human NatA with good selectivity over other NAT family members, and proteomics confirmed on-target activity in yeast cells, validating it as a mechanistic chemical probe.
What earns this a second look is the mechanism: an allosteric binder that stays non-covalent until the enzyme's own turnover releases CoA to staple it in place — a genuinely new trick, not just another warhead. Whether it generalizes beyond NatA's odd CoA-release chemistry is unclear, but it's worth filing away as a template for other CoA- or cofactor-dependent enzymes. The selectivity over other NATs despite a conserved active site is the more actionable lesson: skip the catalytic pocket, go allosteric, and you dodge the conservation problem for free. As probe discovery, repurposing tideglusib beats de novo synthesis for speed, but it also means every existing GSK-3β dataset on this molecule is now confounded — nobody should read across to NatA biology from tideglusib phenotypes without new chemical matter. NatA itself looks like a legitimate, underexplored target; this is a probe paper, not a drug paper, but it's the right first step.
Kura Oncology Spins Off Caspian Therapeutics, Repurposing a Menin Inhibitor for Diabetes
- Kura Oncology spun off Caspian Therapeutics, a separately financed company pursuing menin inhibition for diabetes; Kura retains about 50% ownership, and CEO Troy Wilson chairs Caspian's board.
- Caspian launched with $50 million in financing led by BVF Partners, joined by Eli Lilly, T1D Fund, Invus, Montanova and Kura itself.
- Its lead candidate, KO-7246, targets the same menin-KMT2A interaction as Kura's FDA-approved leukemia drug ziftomenib, but is optimized for chronic metabolic disease rather than AML.
- Menin represses genes that limit pancreatic beta-cell growth; blocking it expanded beta-cell mass and boosted insulin production in preclinical type 1 and type 2 diabetes models.
- Caspian plans to file an IND and begin early trials assessing KO-7246's effects on beta-cell function and insulin output.
The spinout structure is the real story here: Kura offloads capital-intensive diabetes development onto a separately financed entity while keeping ~50% upside — a capital-efficient way to pursue a second indication without diluting the oncology balance sheet. Lilly's participation is the tell: not just funding, but a diabetes-franchise incumbent buying an option on a genuinely novel beta-cell-mass-expansion mechanism, distinct from GLP-1s and existing insulin-secretagogue classes. That novelty cuts both ways, though — menin inhibition is proliferative by design, and driving beta-cell expansion in pancreatic tissue carries an obvious tumorigenicity/hyperplasia overhang preclinical models won't fully resolve; watch how the IND-enabling tox package addresses it. Broader signal: oncology target biology migrating into metabolic disease is a repeatable playbook other biotechs with approved oncology assets may start mining.
Kura cancer drug discovery spurs launch of a diabetes biotech ↗