Skip to Main Content
More Intelligent Procurement, Faster R&D

Go to Main Navigation

Advanced Screening Strategies for Molecular Glue Discovery

Scientist on April 30, 2026

This blog post was written by WuXi AppTec, a trusted partner and contributor to the pharmaceutical and life sciences industries, providing R&D and manufacturing services that help advance healthcare innovation. With operations across Asia, Europe, and North America, we offer integrated, end-to-end services through our unique CRDMO (Contract Research, Development, and Manufacturing Organization) platform.

Despite strides in small molecule drug development over the last century, approximately 85% of the proteome remains undruggable.1 This means that the majority of proteins evade targeting by conventional small molecule drugs with a straightforward inhibitor or activator mechanism of action. These proteins may lack available binding pockets for small molecule interactions. In other cases, a small molecule has been identified that targets the protein, but off-target effects prevent further development. As a result, many diseases cannot be treated with conventional small-molecule medicines. Furthermore, our increasing knowledge of the molecular mechanisms of disease provides opportunities for therapeutic intervention that cannot be approached with traditional small molecules.

One class of therapeutics addressing this challenge is molecular glues (Figure 1).2 “Molecular glues are molecules that can encourage two proteins to interact when they otherwise may not,” said Peichuan Xiang, Director at WuXi AppTec. “As a consequence, they can induce proximity to help stabilize existing interactions or create new, non-native interactions.” This mechanism is catalytic – meaning that it may induce the co-localization (and the functional consequence of that co-localization) between many pairs of target proteins, often resulting in efficacy at low concentrations.

This interaction could lead to degradation, stabilization, inhibition or activation of the target protein, ultimately resulting in the desired therapeutic outcome.3

For example, to promote degradation, a molecular glue can promote the binding between an E3 ligase (a ‘housekeeping’ protein responsible for tagging other proteins for degradation) and a second protein (the ‘target’ protein). By creating a favorable binding surface between these two proteins, the molecular glue brings them into proximity, allowing the ligase to ubiquitinate the target protein, marking it for degradation. An example of a non-degradative molecular glue is one that targets the RAS-MAPK signaling pathway, which is associated with many cancers. This molecular glue binds both RAF kinase and MEK (two proteins in the RAS-MAPK signaling pathway), preventing MEK phosphorylation and inhibiting downstream signaling.4


Figure 1: Molecular glues enable new interactions between proteins

Surprisingly, there are a number of drugs used to treat cancer and neurodegenerative disorders that were later discovered to function as molecular glues. For example, cyclosporin helps prevent transplant rejection, but the drug’s mechanism of action as a molecular glue was identified eight years after FDA approval.5 Another molecular glue, thalidomide, was used for decades in cancer treatments before its molecular glue activity was identified in the 2000s-2010s.6,7 As of January 2025, there are about a dozen molecular glues that have entered clinical studies.8

“The vast majority of the molecules that we now consider to be molecular glues were discovered serendipitously,” said David Madge, VP, Discovery Services from WuXi AppTec. “That’s starting to change, and there are a number of new platform technologies that are now facilitating molecular glue discovery.”

WuXi AppTec has developed numerous screening methods and libraries, as well as in vitro and in vivo assays, to identify and evaluate potential molecular glues. These screening methods have created a toolbox for more strategic and systematic discovery of molecular glues. This white paper describes the tools used to study molecular glue activity, including screening methods to identify potential molecular glues; in vitro assays to evaluate binding; protein degradation, selectivity, and functional outputs; and in vivo testing to validate in vitro findings and assess pharmacokinetic properties.

Molecular Glue Screening Methods

Molecular glue drugs rely on a three-part interaction between two proteins and the molecular glue to induce a therapeutic effect. Because they create a new surface and depend on interactions with two proteins rather than one, they are more complex than traditional active site inhibitors. Minor changes to the molecular glue’s structure can have outsized, poorly understood effects on the interaction, complicating structure-activity relationship modeling. This makes it even more important to identify promising hits from large-scale screening, given the limited options for downstream optimization.

Some of the most effective methods for molecular glue discovery include DNA-encoded libraries, affinity selection mass spectrometry and high-throughput screening, such as TR-FRET and spectral shift assays.

DNA encoded libraries

DNA-encoded libraries (DELs) are small molecule or peptide libraries constructed using combinatorial chemistry.9 These libraries are created using a split-and-pool approach, which is an iterative process used to prepare many molecules by first separating a mixture of compounds into distinct groups. Each group is then coupled to a different building block. These groups are then recombined before another round of splitting and coupling. The split-and-pool approach enables chemists to rapidly create millions to billions of compounds. For DELs, a unique DNA barcode attached to each compound (extended with each synthetic step) allows scientists to identify the hit compound after an affinity-based selection process.

Researchers can screen DELs either in suspension as a pooled library or with multiple copies of each molecule attached to an individual bead. In either case, the goal is to identify compounds that induce complex formation between two target proteins.

  • To screen with in-solution DELs, the first target protein is immobilized on a matrix. The other protein is pre-incubated with the DEL and then added to the first target for interaction. If the second target protein does not interact with the first, it will be washed away. Any interacting molecules that bind the first target can then be released using heat and identified by their unique DNA barcode (Figure 2).


Figure 2: Screening workflow for in-solution DEL screening

  • On-bead screening uses fluorescently labeled target proteins, each with a different label. These proteins are incubated with the on-bead DEL consisting of DNA-tagged beads attached to the molecule library. If the molecule forms a ternary complex, the complex can be identified using a fluorescence-based readout, and molecular glue hits can be identified by their corresponding DNA barcodes.

Screening molecular glues with DELs provides a large screening capacity and can be completed quickly.

Affinity selection mass spectrometry

Like screening DELs, affinity selection mass spectrometry (ASMS) also identifies molecular glue hits based on affinity. The first step of ASMS is to incubate a pooled-molecule library (for example, 400,000 small molecules arrayed into 2000 ‘pools’ of 200 compounds each) with target proteins. The ternary complexes formed will be larger than individual proteins or compounds, allowing them to be separated by chromatography. Because the library used in ASMS does not contain DNA barcodes, hits must then be identified using mass spectrometry (Figure 3).

Using mass spectrometry to identify compounds has the advantage that the compounds appear in their native format during screening and are not subject to potential interference from a DNA barcode. However, screening with ASMS is not as high throughput as DELs and requires specialized equipment.


Figure 3: The ASMS workflow involves first incubating target proteins with compounds to identify binding molecules. These molecules are then separated and detected.

High-throughput screening

Molecular glues can also be screened using a one-compound-per-well high-throughput screening (HTS) approach. This method uses proximity-based assays, such as time-resolved Förster resonance energy transfer (FRET) or spectral shift technology, to detect ternary complex formation and molecular glue activity.

The HTS approach has several advantages. There is only one compound per well, which eliminates any interference that could occur in a pooled well of compounds (e.g., aggregation). Additionally, TR-FRET, spectral shifts and degradation assays, if incorporated, provide much more information than a yes/no binding result, which is the outcome of other screening methods. For example, the Crelux spectral shift technology from WuXi AppTec uses fluorophores attached to a target protein to measure the strength of the interaction. The basis of this technology is that the polarity of the environment around the fluorophore changes in response to binding events. When binding occurs, the fluorophore’s emission at 650 nm and 670 nm changes, and this shift can be dependent on molecular glue concentration (Figure 4).


Figure 4: Spectral shift technology measures a change in the fluorophore emission spectrum upon binding or conformational change of a protein.

Screening libraries from WuXi AppTec (Table 1):

  1. A diversity DEL, which covers the entire known drug space, based on over 6,000 bioactive scaffolds. This library comprises over 200 sub-libraries, each constructed from a different synthetic route, totaling approximately 50 billion structures. This library is ideal for targets that lack reported molecular ligands or binding sites.
  2. Glue-focused DEL based on previously known molecules with molecular glue activity. These libraries begin with reported small molecule ligands or core scaffolds of molecular glues and expand chemical space through extensive chemical modifications. This type of library is useful for targets with existing molecular glues, helping increase screening performance and hit rates. Examples of this type of library include an immunomodulatory imide drug (IMID) focused library targeting the E3 ligase cereblon (CRBN).
  3. Small molecule library of over 370,000 compounds for affinity selection mass spectrometry and high throughput screening.


Table 1: Screening capabilities available from WuXi AppTec.

Validation and optimization of screening hits

Once the top molecular glue hits have been identified, in vitro and in vivo assays are used to understand how they bind, how they affect their target protein and what functional consequence they produce. Xiang added that these assays are extremely helpful in the early stage for hit triage.

Binding

Molecular glue binding can be assessed either with each target protein individually or with both proteins in complex. Assessing each target individually confirms binding to each target, while comparing the results from binary and ternary binding can show cooperativity (Figure 2). There are several assays used to understand binding:

  • Biophysical assays measure physical interactions to understand complex formation. These assays provide quantitative readouts of binding affinity and association and dissociation kinetics. Biophysical assays include surface plasmon resonance, microscale thermophoresis, temperature-related intensity change and biolayer interferometry. These assays can help scientists study how binding kinetics change with the concentration of the hit compound and how binding alters the surface properties of the proteins.
  • Biochemical assays, such as TR-FRET, evaluate properties including proximity, enzyme kinetics and protein-protein interactions. These assays typically use labeled substrates, enzymes or energy transfer to provide quantitative readouts, such as fluorescence or luminescence.
  • Cellular assays, such as NanoBiT and NanoBRET, provide data on molecular glue - target binding, proximity and complex formation within the context of living cells (unlike assays that are done using purified proteins).

Protein ubiquitination and degradation

For molecular glues targeting the degradative machinery, such as E3 ligases, a first step in understanding their action is to examine whether binding triggers protein ubiquitination and degradation. This can be achieved using biochemical assays, such as TR-FRET, which use fluorescently labeled ubiquitin, or cellular assays, such as a pull-down assay, to capture ubiquitinated proteins from cells.

Western blots, including near-infrared western blots, can be used to quantify protein levels and determine the extent and kinetics of degradation. By testing different concentrations of the molecular glue, one can observe how its concentration affects protein degradation (Figure 5). These high-throughput western blot technologies enable scientists to quickly evaluate multiple proteins and compounds at various concentrations in an automated manner.


Figure 5: Western blot quantifying VAV1 degradation in Jurkat cells at different compound concentrations.

In-cell western blots can also quantify endogenous protein levels directly in cultured cells using immunofluorescence. For CRBN-related molecular glues, a CRBN knockout cell line can be used to compare inherent protein instability to molecular glue-induced degradation in a CRBN-expressing cell line.

Selectivity, functional assays, and in vivo pharmacology

Demonstrating binding and degradation of degradative molecular glues is a first step toward validating and prioritizing molecular glue hits. Selectivity testing, functional assays and in vivo pharmacology are then used to characterize hits further and identify those that should proceed to development.

It is critical that molecular glues target only the intended proteins, as off-target interactions can cause unwanted biological effects. For example, researchers often use Western blots to determine whether related proteins are degraded; however, this method cannot provide a complete picture of the entire proteome. Mass spectrometry can help scientists understand the effect of molecular glue on the whole proteome. For some molecular glues, such as CRBN-related molecular glues, degradation profiling panels are available to help assess unwanted degradation of common unintended targets.

Once it is established whether hits induce any off-target activity, the next step is to determine whether the outcome of molecular glue binding has the desired downstream effect. These assays will differ depending on the molecular glue’s function (induced degradation or another proximity-induced functional effect) and on the target protein. For degradative molecular glues, this can mean observing the consequences of degradation on the target proteins. If the hypothesis is that degrading the target protein changes the proliferation rate of a specific cell type, for example, assays should be designed to measure proliferation rates. For non-degradative molecular glues, if the hypothesis is that bringing two proteins together changes the transcription levels of downstream genes, then quantifying those transcripts would be an appropriate assay.

After candidates pass through in vitro studies, the next step is to evaluate their behavior in a living organism. While these studies are not unique to molecular glues and are common to other modalities, they help scientists understand the molecular glue activity in animal models. This includes absorption, distribution, metabolism and excretion (ADME) and drug metabolism and pharmacokinetic (DMPK) testing. For example, in vivo pharmacokinetic studies conducted in animal models can help drug developers understand how the body processes the drug, while metabolite identification assays aid in assessing its efficacy and safety profile. As mentioned earlier, due to their mechanism of action, molecular glues that induce degradation can exert effects at lower concentrations than other small-molecule drugs and can exhibit long-lasting inhibition of the target protein.10

These types of in vivo studies help evaluate molecular glue hits for further development towards the clinical stage (Phase I, II and III trials) and ultimately, approval.

Conclusions

Molecular glue discovery is a complex process that involves screening thousands or millions of compounds and then confirming their binding properties and functions both in vitro and in vivo (Figure 6). Having the right screening strategies to identify the most promising hits is essential.


Figure 6: Workflow representation for advancing molecular glue drug discovery

Molecular glues open new opportunities for targeting undruggable proteins, potentially having a tremendous impact on patients with limited treatment options. By providing opportunities to engage with disease pharmacology in ways that traditional small-molecule drugs cannot, molecular glues can help turn undruggable proteins into targets, paving the way for treatments for diseases once considered untreatable.

WuXi AppTec has developed a comprehensive toolbox to facilitate the entire molecular glue discovery process, including:

  • Screening libraries such as DEL, HTS and small molecule libraries
  • Protein production and crystallography capabilities
  • Full testing support, including biochemical, biophysical and cellular assays
  • Cell lines, tumor models and animal models to support in vitro and in vivo assays
  • In vivo pharmacology studies to assess safety, efficacy and toxicity

View our recent webinar on molecular glue drug discovery HERE

Download our poster showcasing our HTS platform for molecular glue drug discovery HERE

WuXi AppTec is a trusted partner and contributor to the pharmaceutical and life sciences industries, providing R&D and manufacturing services that help advance healthcare innovation. With operations across Asia, Europe, and North America, we offer integrated, end-to-end services through our unique CRDMO (Contract Research, Development, and Manufacturing Organization) platform.

Contact us HERE to get your questions answered by one of our experts!

Connect with WuXi AppTec on Scientist.com