Characteristics and Mechanisms of Molecular Glues
Origins and Discovery of Molecular Glues
Small-molecule drugs are highly effective and constitute the most widely used class of
therapeutics in the medical industry.
With their low molecular weight, these compounds support oral bioavailability and
facilitate the crossing of biological barriers.
Additionally, these compounds are readily accessible through chemical synthesis and
amenable to structural modification.
Consequently, small-molecule drugs dominate the modern pharmaceutical market, accounting
for approximately 90% of global sales[1].
They typically act as enzyme inhibitors or receptor antagonists by occupying active
binding sites on target proteins.
However, only about 15% of pharmacologically relevant proteins encoded by the human
genome can be targeted by conventional
small molecules. The remaining proteins often feature shallow or broad active sites or
smooth surfaces with minimal binding pockets,
rendering them “undruggable”. A breakthrough in addressing this challenge is the
emergence of molecular glues—a novel class of small
molecules capable of inducing PPIs.
The concept of molecular glues arose from the therapeutic potential of modulating
intracellular protein interactions.
Unlike traditional inhibitors, molecular glues induce or stabilize PPIs, often leading
to the degradation of disease-associated
proteins via the
ubiquitin-proteasome system. Early serendipitous discoveries, such as
the immunosuppressants cyclosporin
A and FK506, which mediate novel interactions between
FKBP12 and
calcineurin, led to
their recognition as “molecular glues”.
A pivotal milestone was the finding that thalidomide and its analogs (lenalidomide,
pomalidomide) act as molecular glues by
binding to the E3 ligase cereblon (CRBN), thereby inducing degradation of
transcription
factors IKZF1 and
IKZF3—a mechanism
foundational to treatments for multiple myeloma and immunological disorders. These
discoveries highlighted the potential of
molecular glues to target proteins previously considered undruggable, including
transcription factors, scaffolding proteins,
and aggregation-prone species. Since 2000, research in this field has grown rapidly,
with particularly notable expansion after
2018, especially in oncology and drug discovery
[2-4].
Mechanisms and Biological Effects of Molecular Glues
Molecular glues can be classified into two main categories: PPI stabilizers and chemical
inducers of proximity (CIPs).
PPI stabilizers, a subset of PPI modulators, enhance the thermodynamic stability of
pre-existing protein interactions
(Figure 1A). A classic example is
paclitaxel, which stabilizes the interaction between
α-tubulin and β-tubulin in
microtubules.
In contrast, CIPs facilitate non-native PPIs by exploiting the principle of proximity
within biological systems (Figure 1B)
[5].
In many cases, molecular glues function by inducing or enhancing the proximity of two
proteins, thereby promoting or modulating
their interactions.
Figure 1. Schematic representation of molecular glues[5].
A: Molecular glues that stabilize protein-protein interactions (PPIs).
B: Schematic representation of chimeric inducer of proximity (CIPs).
These interactions can lead to a variety of biological effects, including:
(A) Targeted degradation, achieved by bringing a target protein into proximity with an
E3 ligase,
(B) Stabilization of a target-effector complex,
(C) Inhibition of target activity by obstructing its interaction with a native binding
partner, and
(D) Activation of target activity by facilitating interaction with an activating
regulatory protein (Figure 2).
Figure 2. Biological effects result from molecular glues[6].
Key Features of Molecular Glues and Comparison with PROTACs
Defining attributes of molecular glues include monovalency, selective binding, and the
capacity to stabilize or
induce novel protein complexes, distinguishing them from bifunctional compounds such as
PROTACs.
●Monovalency and Selectivity: Molecular
glues are typically small, monovalent molecules that bind to a single protein partner,
inducing a conformational change that enhances its affinity for a second protein. This
facilitates the formation or stabilization of a ternary complex, often in the absence of
measurable affinity between the second partner and the first protein without the
glue[7-9].
●Cooperativity: Molecular glues can increase
the affinity between protein partners, often elevating interactions from the low
micromolar to nanomolar range—a phenomenon known as positive cooperativity[5].
●Contrast with Bifunctional Compounds:
Unlike PROTACs, which contain two distinct binding domains, molecular glues extend or
reshape protein-protein interfaces, frequently engaging solvent-exposed or shallow
surfaces rather than conventional ligand-binding pockets[6].
●Mechanistic Diversity:
Molecular glues are capable of stabilizing native interactions, inducing novel
interactions, or promoting protein aggregation, leading to outcomes such as targeted
degradation, complex stabilization, or modulation of the interactome[5].
Table 1. Comparison of molecular glues and PROTACs[10].
|
Molecular glue |
PROTAC |
| Mechanism |
Binds E3 or target protein induces PPI |
Binds target and E3 |
| Target protein |
To be determined |
Predictable |
| Discovery strategy |
Historically serendipitous discovery |
Rational design |
| Feature |
Monovalent |
Bivalent |
| Linker |
Without linker |
With linker |
| Molecular weight |
Lower |
Higher |
| Rule of five |
Typically within |
Beyond |
Binding pocket in the
target protein |
Nonessential |
Required |
Development Strategies of Molecular Glues
The discovery of molecular glues has evolved from serendipitous observations to
systematic and rational approaches.
Key strategies include (1) serendipitous discovery, (2)
high-throughput screening, (3) rational and structure-based
design, and (4) computational and in silico methods.
Serendipitous Discovery
Historically, numerous molecular glues, including thalidomide and its analogs, were
identified fortuitously during
unrelated drug development efforts. These discoveries often arose from unexpected
biological activities that were later
attributed to the induction of novel PPIs[10]. While
informative, this approach remains inefficient and unpredictable,
highlighting the need for more directed discovery strategies.
High-Throughput Screening
Deliberate discovery efforts increasingly rely on high-throughput chemical and
phenotypic screening, wherein large
compound libraries are evaluated for their ability to induce protein degradation or
modulate protein-protein interactions.
This strategy can be combined with multi-omics and morphological profiling to
elucidate targets and mechanisms.
For example, Cristina Mayor-Ruiz et al. identified compounds that promote
ubiquitination and degradation of cyclin K
through chemical screening in hyponeddylated cells, coupled with a multi-omics
target deconvolution campaign[11].
Various biochemical methods are employed in molecular glue screening
[6]:
(A)
DNA-encoded libraries: Compounds are tagged with unique DNA barcodes to
identify those that promote interactions.
(B)
Tethered fragment screening: An effector protein with an exposed cysteine
residue is incubated with a target peptide and a library of disulfide-containing
fragments. Affinity-based binding near the cysteine enables enrichment, which is
typically detected via
Mass Spectrometry (MS) or
Fluorescence Assay (FA).
(C)
TR-FRET: Target and effector proteins are tagged with donor and acceptor
fluorophores; proximity-induced FRET signals reflect interaction.
(D)
AlphaScreen: Bead-based binding transfers energy upon proximity between
partners, generating a chemiluminescent signal.
(E)
E3-driven microarrays: The target protein is immobilized on a surface,
and effector binding is tested in the presence or absence of putative glue
molecules.
Figure 3. Biochemical screening technologies[6].
Rational Design and Structure-Based Approaches
Advances in structural biology have enabled rational design of molecular glues by
revealing the interfaces and
conformational changes involved in ternary complex formation. Structure-guided
optimization and fragment-based drug
design are increasingly applied to enhance both potency and selectivity. For
instance, Ethan S. Toriki et al. developed
a strategy to convert protein-targeting ligands into covalent molecular glue
degraders, generating compounds capable of
degrading
BRD4,
BCR-ABL, c-ABL,
PDE5, AR,
AR-V7,
BTK,
LRRK2,
HDAC1/3, and SMARCA2/4
[12].
Figure 4. Rational chemical design of molecular glue degraders[12].
Rational design provides precision and enables targeting of previously undruggable
proteins, though it relies on detailed structural information.
Computational and In Silico Methods
Emerging computational tools—such as molecular docking, free energy calculations,
and AI-driven design—are increasingly applied to predict and
optimize molecular glue candidates before synthesis and experimental testing. For
example, Balint Dudas et al. employed an efficient
computational approximation of cooperativity to establish a protocol for identifying
potent molecular glues from large libraries. By applying
cooperative binding principles in ternary complex formation, this method predicts
ligand-induced PPIs and their potential for target
degradation[13].
Figure 5. Possible paths leading to ternary complex formation from the isolated
components: the two proteins (A and B) and the ligand (L)[13].
In silico approaches accelerate discovery and reduce experimental workload but
require validation and further methodological refinement.
Table 2. Representative strategies used in molecular glue development[10].
| Molecular glue |
Ligase |
Target protein |
Discovery strategy |
| Thalidomide |
CRL4CRBN |
IKZF1, IKZF3 |
Serendipity |
| Lenalidomide |
CRL4CRBN |
IKZF1, IKZF3, CK1α |
Serendipity |
| CC-885 |
CRL4CRBN |
GSPT1 |
Rational design |
| CC-122 |
CRL4CRBN |
IKZF1, IKZF3, ZFP91 |
Rational design |
| Indisulam |
CRL4DCAF15 |
RBM39 |
Serendipity |
| BI-3802 |
SIAH1 |
BCL6 |
Serendipity |
| AN1 |
LC3 |
mHTT |
HTS |
| 10O5 |
LC3 |
mHTT |
HTS |
| Asukamycin |
UBR7 |
TP53 |
Covalent binding |
| dCeMM2 |
DDB1 |
cyclin K |
Scalable chemical profiling |
| dCeMM3 |
DDB1 |
cyclin K |
Scalable chemical profiling |
| (R)-CR8 |
DDB1 |
cyclin K |
Data mining |
| HQ461 |
DDB1 |
cyclin K |
Chemical genetics |
| NRX-103094 |
SKP1β-TrCP |
β-catenin peptide |
Rational design |
Clinical Progress and Challenges of Molecular Glues
Clinical Progress of Molecular Glues
Molecular glues have rapidly evolved from conceptual tools in chemical biology to
promising clinical candidates, particularly for diseases involving previously
undruggable targets. Their unique ability to induce or stabilize PPIs has driven the
development of both degradative and non-degradative therapeutics, with a number of
candidates advancing into clinical trials. To date, nearly a dozen molecular glues
have entered clinical studies (Table 3).
Table 3. Selected examples (non-exhaustive) of molecular glue degraders in clinical
studies[7].
| Molecular glue |
Target |
Target function |
Company |
Indication |
Clinical status |
| MRT-2359 |
GSPT1 |
GTPase, translation termination factor |
Monte Rosa |
MYC-driven cancer |
Phase 2; discontinued |
| CC-90009 |
Bristol Myers Squibb (BMS) |
| HbF-activating CELMoD |
NA |
Transcription factor |
BMS |
Sickle cell disease |
Phase 1 |
| NA (glue) |
WIZ |
Transcription factor |
Novartis |
Sickle cell disease |
Preclinical |
| Helios CELMoD |
IKZF2 |
Transcription factor |
BMS |
Cancer |
Phase 1 |
| DKY709 |
Novartis |
Cancer |
Discontinued |
| NA |
RBM39 |
Splicing factor |
Seed Therapeutics |
Cancer |
Phase 1 in 2025 |
| MRT-8102 (glue) |
NEK7 |
Kinase |
Monte Rosa |
Inflammation |
Preclinical |
| MRT-6160 |
Vav1 |
Guanine nucleotide exchange factor |
Monte Rosa |
Autoimmunity |
Phase 1 |
Note: MCE can provide products for research use only. We do not sell to patients.
Challenges and Future Directions
Despite their transformative potential, the development of molecular glues faces
several scientific and practical challenges:
●Unpredictable Discovery and Low Hit
Rates: Most molecular glues have been identified serendipitously, and
systematic discovery remains difficult due to the complexity of the PPIs they
stabilize. High-throughput screens generally yield low hit rates, making the process
resource-intensive and slow.
●Limited Structural and Mechanistic
Insight: Incomplete understanding of the interfaces and mechanisms by which
glues induce or stabilize protein complexes hinders rational design and
optimization, especially for undruggable targets[9].
●Validation and Optimization:
Establishing specificity, efficacy, and safety of candidate glues requires advanced
biochemical, biophysical, and cellular assays, alongside comprehensive
pharmacokinetic and toxicity studies[14].
●Resistance and Off-Target Effects:
Similar to other targeted therapies, molecular glues may encounter resistance
mechanisms and trigger off-target effects through modulation of non-target
proteins[5].
Nevertheless, ongoing advances in structural biology, computational modeling, and
high-throughput screening are paving the way for broader and more effective
applications of molecular glues.
Summary
Molecular glues represent a paradigm shift in drug discovery, modulating or
degrading proteins previously considered “undruggable”
by stabilizing protein-protein interactions. Their monovalent, selective, and
cooperative properties enable formation of ternary
complexes, resulting in targeted degradation, complex stabilization, or modulation
of protein activity. These features distinguish
them from bifunctional compounds like PROTACs and expand the range of
pharmacologically accessible proteins, providing innovative
strategies for therapeutic development.
Molecular glues have been discovered and developed through serendipitous findings,
high-throughput screening, rational structure-based
design, and computational approaches. Some have advanced into clinical trials,
showing promise in oncology and other disease areas.
Despite ongoing challenges in specificity, discovery efficiency, mechanistic
understanding, and potential off-target effects,
they offer a versatile platform for mechanistic studies, drug development, and
translational applications, highlighting their
transformative potential in modern drug discovery.
Recommended Molecular Glue Degraders and Compound Library:
| Product Name |
Catalog Number |
Target(s) |
Key Features |
| Iberdomide |
HY-101291 |
Cereblon (CRBN) |
Orally active and potent CRBN E3 ligase modulator (CELMoD) |
| Pomalidomide |
HY-10984 |
Ikaros (IKZF1/3) |
Third-generation immunomodulatory agent |
| dCeMM2 |
HY-144971 |
Cyclin K |
Molecular glue-type degrader that targets cyclin K |
| dCeMM4 |
HY-144977 |
Cyclin K |
Molecular glue degrader of cyclin K; induces ubiquitination via
CDK12-cyclin K/CRL4B ligase interaction |
| HQ461 |
HY-144981 |
Cyclin K |
Molecular glue targeting CDK12-DDB1 to degrade cyclin K |
| FPFT-2216 |
HY-145319 |
PDE6D, IKZF1/3, CK1α |
Multi-target degrader |
| ALV1 |
HY-145776 |
IKZF1, IKZF2 |
Molecular glue degrader for Ikaros (IKZF1) and Helios (IKZF2) |
| TMX1 |
HY-153385 |
BRD4BD2 |
Covalent molecular glue degrader targeting BRD4 bromodomain 2 (via DCAF16) |
| MMH1 |
HY-156827 |
BRD4BD2 |
Novel BRD4 bromodomain 2 degrader (via CUL4-DCAF16 E3 ligase) |
| BMS-986397 |
HY-159646 |
CK1α |
CRBN-based molecular glue degrader of CK1α |
| EM12-FS |
HY-164891 |
NTAQ1 |
Dual-function: High-affinity CRBN ligand (His353 binding) & molecular glue degrader of NTAQ1 |
| Molecular Glue Compound Library |
HY-L137 |
/ |
Compound library containing various molecular glue compounds |
LinkedIn
Researchgate
Facebook
X
Instagram
Mail
Characteristics and Mechanisms of Molecular
Glues
Development Strategies of Molecular Glues
Clinical Progress and Challenges of Molecular
Glues
