Molecular glue degraders have evolved from a serendipitous observation to one of the most dynamic and transformative fields in biomedical research. By harnessing the mechanism of induced protein–protein interactions, these small molecules have overcome long-standing barriers in drug discovery and opened unprecedented opportunities to target historically undruggable proteins. Characterized by low molecular weight, favorable druglike properties, and strong clinical efficacy in hematological malignancies, molecular glues are rapidly expanding their target space and therapeutic scope with the support of artificial intelligence and highefficiency screening platforms.

Molecular glue technology originated serendipitously in the 1950s with thalidomide, initially developed as a sedative and antiemetic but later withdrawn from the market worldwide due to severe congenital malformations and teratogenicity. Remarkably, this withdrawn drug was subsequently found to exert potent therapeutic effects in multiple myeloma and some other hematological malignancies, arousing widespread interest in its hidden mechanism of action.

In 2010, Japanese scientist Hiroshi Handa and his research team made a landmark discovery: thalidomide and its analogs (lenalidomide, pomalidomide, collectively called IMiDs) directly bind to cereblon (CRBN), the substrate-recognition subunit of the CRL4CRBN E3 ubiquitin ligase complex. This binding event remodels the substrate-binding pocket of CRBN, redirects the specificity of the E3 ligase, and promotes the ubiquitination and subsequent proteasomal degradation of downstream pathogenic proteins such as IKZF1/3 and CK1α.

This finding not only explained the clinical efficacy and adverse effects of thalidomide derivatives but also defined the core mechanism of molecular glues: inducing de novo protein–protein interactions between an E3 ubiquitin ligase and a target substrate, thereby driving selective and effective degradatioln of disease-causing proteins. Since then, molecular glue degraders have evolved from accidental findings to a rational, designable, and promising new class of drugs in targeted protein degradation.

Unlike traditional small-molecule inhibitors that occupy active sites to block protein function, molecular glue degraders eliminate target proteins entirely through the ubiquitin–proteasome system. This unique mode of action makes them particularly effective against non-enzymatic proteins such as transcription factors, which have been considered inaccessible to conventional drug development.

In recent years, molecular glues have emerged as a core technology in targeted protein degradation (TPD), serving as a complementary and transformative approach alongside conventional small molecules and proteolysis-targeting chimeras (PROTACs). Their unique mechanism of action and favorable properties endow them with distinct advantages that address critical limitations of other TPD strategies and traditional therapeutics. Their main strengths include:

Low molecular weight, strong membrane permeability, and superior oral bioavailability.

No requirement for high-affinity binding to E3 ligase or POI.

Greatly expanded druggable space by targeting previously intractable proteins.

Reduced risk of drug resistance compared to conventional inhibitors.

Favorable pharmacokinetic profiles in preclinical and early clinical studies.

Early molecular glues were mostly discovered by chance. With the structural elucidation of CRBN and other E3 ligases, clarification of molecular mechanisms, and the development of AIaided design and specialized compound libraries, the field has entered a new era of rational design. Modern strategies focus on precise modification of CRBN-binding scaffolds to enhance affinity and selectivity, high-throughput screening for POI-binding fragments that can induce favorable E3-POI interactions, and controlled induction of novel, specific interactions between E3 ligases and previously unrecognized substrates. Additionally, advances in structural biology techniques have enabled detailed visualization of ternary complexes, providing crucial insights for structure-guided optimization of molecular glues to improve their potency, selectivity, and druggability.

Despite rapid progress, molecular glue development faces substantial challenges that hinder its clinical translation and wide application.

Poor target selectivity and off-target degradation: Most molecular glue ligands lack sufficient binding specificity for the target E3 ligase–substrate interface, leading to off-target degradation of essential proteins, hematological toxicities, reduced therapeutic safety, and complicated clinical dose escalation.

Unintended neo-substrate degradation: Molecular glues often induce unintended E3 ligase–neo-substrate interactions, causing uncontrolled degradation of unrelated proteins, which impairs safety and limits therapeutic windows.

Trade-offs between degradative activity and drug-like properties: Chemical optimization often conflicts with degradation potency and druggability; enhanced degradation activity usually leads to increased molecular size, poor solubility, reduced membrane permeability, or unfavorable pharmacokinetics. Balancing strong degradation, favorable pharmacokinetics, and good oral bioavailability remains a major hurdle.

Difficulties in rational design and structural prediction: Ternary complex (E3 ligase–molecular glue–POI) formation is highly sensitive to subtle structural changes, making accurate assembly prediction challenging and rational design strategies more limited than traditional small-molecule drugs.

Limited understanding of structure–activity–selectivity relationships (SASR): The structural determinants of degradation selectivity are not fully clarified; minor molecular scaffold modifications can unexpectedly alter degradation profiles or target specificities, hindering efficient molecular optimization.

Molecular glues have achieved remarkable clinical progress, with multiple candidates advancing into clinical trials and showing promise in both hematological cancers and solid tumors.

GSPT1 is a key regulator of the cell cycle, and its dysregulation is closely associated with hematologic malignancies such as acute myeloid leukemia (AML). At present, the development of GSPT1-targeted molecular glues has entered a critical stage of clinical translation, with multiple candidates advancing to Phase I/II clinical trials.

To reduce off-target degradation of unintended proteins, the CRBN-binding scaffold has been expanded beyond conventional phthalimide cores to include novel scaffolds such as benzotriazinones, benzimidazoles, and oximes, thereby improving target selectivity and degradation activity. In addition, modification of substituents on benzene and heterocyclic rings enhances binding affinity to both CRBN and GSPT1, stabilizes the ternary complex, and strengthens anti-proliferative activity.

Current optimization strategies for GSPT1 molecular glues focus on high selectivity and combination therapy to support clinical translation and application.

DEG6498 is the world’s first HuR-targeted molecular glue degrader to enter clinical development globally in 2025, and is currently in Phase I clinical trials. The ongoing trials focus on advanced solid tumors including colorectal cancer, lung cancer, and hepatocellular carcinoma.

This compound was developed using the GlueXplorer® platform, from which lead molecules were identified via high-throughput screening of tens of thousands of molecular glue compounds. Subsequent medicinal chemistry optimization focused on improving molecular hydrophobicity, introducing heterocyclic structures to enhance oral bioavailability, and imposing conformational constraints to boost target selectivity.

To date, the compound has demonstrated favorable tolerability in patients with no dose-limiting toxicities observed. It holds great potential to fill the unmet medical need for the treatment of multiple refractory solid tumors.

In 2024, Novartis and collaborative groups reported a new class of CRBN-recruiting molecular glue degraders targeting transcription factor WIZ. From phenotypic screening of a CRBN-biased compound library, two lead molecules, dWIZ-1 and dWIZ-2, were discovered. These agents upregulate fetal hemoglobin (HbF) expression without impairing erythroid proliferation and differentiation, thereby ameliorating sickle cell disease (SCD) and thalassemia.

Both degraders assemble canonical CRBN ternary complexes with the ZF7 domain of WIZ. The β-hairpin glycine G876 of WIZ accommodates the phthalimide moiety, which occupies the CRBN tri-tryptophan pocket. Mechanistically, dWIZ-1 forms direct hydrogen bonds with WIZ(ZF7) V874 and reinforces interprotein hydrogen-bonding networks between CRBN and WIZ, further stabilizing their interface.

dWIZ-1 suffers from poor in vivo bioavailability and difficult stereoisomeric separation due to its chiral methyl group, which triggers metabolic isomerization. These drawbacks are eliminated in dWIZ-2 via methyl group removal.

Subsequent structure–activity relationship (SAR) optimization adopted a piperidine cyclization approach based on dWIZ-2 to restrict molecular flexibility. Site-specific methyl substitution on the piperidine ring was applied to boost WIZ degradation efficacy while maintaining selectivity against SALL4 through steric hindrance. The resultant optimized analogs displayed superior potency, excellent selectivity, and favorable pharmacokinetic properties.

Currently, Novartis, Bristol-Myers Squibb (BMS), and other pharmaceutical companies are advancing related drug candidates, with some already in clinical development. These agents represent a novel oral therapeutic strategy for the treatment of related hematological disorders.

The molecular glue field is rapidly evolving toward multi-target, multi-indication, and technology-integrated development. Targets have expanded from classic transcription factors (IKZF1/3) to kinases (CK1α), RNAbinding proteins (HuR), GTPases (RAS), and many other classes. Therapeutic indications have broadened from hematological malignancies to solid tumors, autoimmune diseases, neurodegenerative disorders, and metabolic diseases.

Several key technologies are driving this expansion

High-efficiency screening methods, including DNA-encoded libraries, affinity selection mass spectrometry (ASMS), and high-throughput screening (HTS), enable rapid identification of new targets and degraders;

Deep learning models such as DeepTernary use SE(3)-equivariant architectures to accurately predict ternary complex formation, greatly shortening the research cycle;

Structure-based computational mining, such as the algorithm developed by Monte Rosa Therapeutics, has identified more than 1,600 human proteins containing CRBN-compatible G-loop motifs, vastly expanding the potential target space for molecular glues.

Molecular glues represent a revolutionary approach to drug discovery with demonstrated therapeutic potential in oncology, immunology, neurodegeneration, and genetic diseases.

Current optimization strategies focus on modular design: dissecting the ternary complex into E3 ligase-binding and POI-binding modules, then using structure–activity relationship analysis to improve potency, selectivity, and pharmacokinetic properties. Supported by advanced screening platforms, AI-aided design, and specialized compound libraries, molecular glues are poised to continue driving innovation and deliver breakthrough therapies for unmet medical needs.

MCE provides a comprehensive suite of specialized libraries to support every stage of molecular glue research and development. These resources enable researchers to accelerate hit discovery, lead optimization, and clinical translation, fully unlocking the potential of molecular glue technology.