Next-Generation CRBN Modulators Show Promise Against IMiD Resistance in PROTAC Therapy

In the rapidly evolving landscape of precision oncology, few strategies have captured the imagination of researchers quite like proteolysis targeting chimeras, or PROTACs. Unlike conventional inhibitors that merely silence malfunctioning proteins, PROTACs harness the cell’s own ubiquitin-proteasome system to completely destroy disease-causing targets. At the heart of many of these degraders lies cereblon (CRBN), the substrate receptor of the CRL4 E3 ubiquitin ligase complex, which is recruited by small-molecule ligands such as thalidomide, lenalidomide, and pomalidomide—the so-called immunomodulatory drugs, or IMiDs. But a persistent clinical challenge has emerged: mutations in CRBN itself can undermine the therapeutic efficacy of these drugs, driving resistance in patients with relapsed or refractory cancers, particularly multiple myeloma. Now, a new computational study published in Results in Chemistry offers a compelling case for next-generation CRBN E3 ligase modulators, or CELMoDs, as potential alternatives capable of withstanding these resistance-causing mutations.

The research, led by E. Rosette Celsiya Mary and C. Sudandiradoss from Vellore Institute of Technology, systematically compared the binding resilience of first-generation IMiDs against next-generation CELMoDs—including iberdomide, mezigdomide, CC-885, CC-122, and CC-3060—across a panel of structurally prioritized deleterious CRBN variants. The team began by retrieving 409 missense single nucleotide polymorphisms (SNPs) of CRBN from the NCBI dbSNP database and then deployed a battery of functional prediction tools, including PolyPhen-2, PANTHER, MutationTaster, and SNPs & GO, alongside structural stability predictors DynaMut2 and DUET. From this comprehensive screen, six variants located within the CRBN Ubiquitin Ligase Targeting (CULT) domain—residues 318 to 426—emerged as particularly damaging: G344R, N351T, H353R, H357Q, S379I, W380R, and P382R. These residues sit squarely within the thalidomide-binding pocket, the very region that IMiDs and CELMoDs must engage to recruit CRBN into a functional ternary complex with the target protein.

What makes these findings clinically resonant is that several of the identified variants overlap with mutations experimentally documented in IMiD-treated patients. The P382R variant, for instance, has been reported in an IMiD-refractory or relapsed patient, while W380R sits within the most significant thalidomide-binding pocket. Other experimentally characterized variants such as C326Y, associated with pomalidomide refraction, and A347V, linked to lenalidomide resistance, further underscore the vulnerability of this domain. The researchers also noted that G344R was the only high-scoring variant found in the ClinVar database, flagged as a variant of uncertain significance—a detail that reinforces its potential clinical relevance even as most other high-impact variants remain absent from clinical genomic repositories.

To interrogate how these mutations affect ligand binding, the team constructed a homology model of CRBN using MODELLER with the UCSF Chimera 1.19 interface, anchored on a high-resolution crystal structure (PDB ID: 7BQU, resolved at 1.9 Å) and refined against an AlphaFold 2-predicted template. The resulting model achieved 100% sequence identity with the hybrid template, and Ramachandran plot analysis confirmed that approximately 89 to 90 percent of residues occupied favorable regions, with only 5 percent in disallowed zones. With this validated structural foundation in place, the researchers performed molecular docking using GNINA 1.0 integrated with AutoDock Vina, leveraging a convolutional neural network-based scoring function to predict binding affinities across all eight ligands against both wild-type and mutant CRBN proteins.

The docking results were striking. Among all tested compounds, CC-885—a next-generation CELMoD—demonstrated the most potent and consistent binding, with affinity scores ranging from −9.34 to −10.40 kcal/mol across all CRBN variants. Its best predicted affinity of −10.40 kcal/mol against wild-type CRBN substantially outperformed thalidomide (−7.80 kcal/mol), lenalidomide (−7.36 kcal/mol), and pomalidomide (−7.15 kcal/mol). Although CC-885’s ligand efficiency (−0.34) was lower than that of the smaller IMiDs—a consequence of its larger molecular size distributing binding energy across more heavy atoms—its absolute binding strength was unmatched. Mezigdomide and iberdomide occupied an intermediate tier, showing improved performance over traditional IMiDs but falling short of CC-885’s dominance. Notably, mezigdomide exhibited enhanced binding affinity in the S379I (−8.24 kcal/mol) and P382R (−8.15 kcal/mol) mutants compared to wild-type (−7.89 kcal/mol), hinting at some mutation-resilient properties among certain CELMoDs.

Interaction profiling using the Protein Ligand Interaction Profiler (PLIP) revealed that CC-885 engaged in a richer network of non-covalent contacts than its competitors, combining hydrogen bonds, hydrophobic interactions, and π-cation interactions across wild-type and mutant complexes. Residues such as ASP50, SER272, SER339, LEU261, TRP264, ASP265, LEU268, LEU273, and LEU342 featured prominently in these interactions, with LYS269 contributing a π-cation contact. By contrast, pomalidomide relied primarily on hydrogen bonds with ASP50, THR51, ASP265, and LYS269, while mezigdomide showed more variable and sometimes absent interaction patterns across mutants. This richer interaction landscape suggested that CC-885’s larger scaffold might provide a structural buffer against mutation-induced perturbations in the binding pocket.

But static docking alone cannot capture the dynamic reality of protein-ligand recognition. To address this, the team turned to 200-nanosecond molecular dynamics simulations using GROMACS 2025.4 with the CHARMM36 force field, solvating each complex in a dodecahedral box with TIP3P water and 0.15 M NaCl to maintain physiological ionic strength. They focused on the wild-type, S379I, and P382R variants bound to pomalidomide, mezigdomide, and CC-885. The results were unambiguous: CC-885-bound complexes maintained superior conformational stability across triplicate independent simulations, with protein RMSD values ranging from approximately 0.39 to 0.56 nm, radius of gyration values consistently within 2.38 to 2.42 nm, and solvent-accessible surface area values remaining in a narrow band of roughly 230 to 270 nm². In contrast, pomalidomide complexes exhibited higher fluctuations and greater solvent exposure, particularly in the P382R mutant, while mezigdomide complexes showed intermediate behavior with notable ligand mobility in the S379I variant.

Ligand RMSD analysis further cemented CC-885’s advantage. The mean ligand RMSD values for CC-885 complexes ranged from approximately 0.26 to 0.57 nm across replicates, with the P382R mutant displaying particularly consistent ligand positioning. Pomalidomide and mezigdomide, by comparison, exhibited greater structural deviations and increased ligand mobility, with the P382R mutant showing the greatest fluctuation in the pomalidomide-bound complex. Hydrogen bond analysis revealed that CC-885 maintained persistent, continuous interactions throughout all simulation replicates, whereas pomalidomide and mezigdomide showed more intermittent bonding patterns. Molecular Mechanics Poisson-Boltzmann Surface Area (MM/PBSA) calculations confirmed these observations, with CC-885 complexes exhibiting the most favorable binding free energies—reaching approximately −35 to −40 kcal/mol in certain replicates—compared to the weaker affinities of pomalidomide and mezigdomide complexes.

Perhaps most intriguingly, the S379I variant, which the variant impact analysis had classified as stabilizing rather than destabilizing, was confirmed through simulation to enhance protein compactness, as evidenced by consistently lower SASA values and favorable radius of gyration measurements across all ligand-bound complexes. Yet even this stabilizing mutation did not compromise CC-885’s binding performance. The P382R variant, clinically associated with IMiD resistance, showed clear signs of destabilization in the pomalidomide-bound complex—elevated RMSD and SASA values—but remained well-accommodated within the CC-885 binding environment. This differential response underscores a central thesis of the study: the choice of E3 ligase-recruiting ligand can determine whether resistance-causing mutations translate into therapeutic failure or remain structurally tolerated.

The implications of this work extend beyond the immediate finding that CC-885 outperforms first-generation IMiDs in mutation-resilient binding. By systematically mapping the structural consequences of CRBN variants within the CULT domain and demonstrating that next-generation CELMoDs can maintain stable ternary complex architecture despite these perturbations, the study provides a rational framework for future PROTAC design. Rather than treating resistance as an inevitable clinical endpoint, researchers can now consider ligand selection as a proactive strategy to preempt mutation-induced failure. The authors acknowledge that their findings remain at the computational stage and that experimental validation and clinical investigation will be essential to translate these insights into therapeutic practice. Nevertheless, the convergence of variant impact prediction, homology modelling, deep-learning-enhanced docking, and extended molecular dynamics simulation represents a powerful methodological template for identifying mutation-resilient degraders. As PROTAC therapy continues to gain momentum in cancer treatment since its clinical resurgence in 2015, studies like this one illuminate a path toward designing next-generation molecular glues and degraders that can withstand the evolutionary pressure of tumor mutation—turning the very mutations that once conferred resistance into targets for strategic ligand engineering.

Subject of Research: Computational evaluation of next-generation CRBN E3 ligase modulators as alternatives to immunomodulatory drugs for overcoming CRBN mutation-driven resistance in PROTAC therapy

Article Title: CRBN E3 ligase modulators (CELMoDs) as potential alternative for immunomodulatory drug (IMiD) resistance in CRBN-based targeted therapy

Article References: Mary, E. R. C., & Sudandiradoss, C. (2026). CRBN E3 ligase modulators (CELMoDs) as potential alternative for immunomodulatory drug (IMiD) resistance in CRBN-based targeted therapy. Results in Chemistry, 31, Article 103972. https://doi.org/10.1016/j.rechem.2026.103972

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103972

Keywords: PROTACs, CRBN, CELMoDs, IMiDs, drug resistance, molecular docking, molecular dynamics, CC-885, mezigdomide, CULT domain, multiple myeloma, targeted protein degradation