UC Berkley Finds New Molecular Glue Which May Stick Where Zytiga and Xtandi Fail
An Optimized RNF126-Targeting Covalent Handle for Molecular Glue Degraders | bioRxiv
A "Molecular Glue" Approach That May Finally Crack the AR-V7 Code
Scientists at UC Berkeley have engineered a safer, more refined version of a so-called "molecular glue" compound that can simultaneously destroy both the standard androgen receptor (AR) and its dangerous drug-resistant cousin, AR-V7, in prostate cancer cells. Unlike enzalutamide (Xtandi) or the PROTAC drug ARV-110 (avdegalutamide)—which cannot touch AR-V7—this new laboratory compound, called EST1140, works by recruiting a cellular "trash-disposal" protein called RNF126 to tag both AR and AR-V7 for destruction. While still in early laboratory (pre-clinical) research, this advance is part of a rapidly growing field that has already put multiple protein-degrader drugs into clinical trials and could eventually offer new hope to the large fraction of advanced prostate cancer patients for whom standard hormone therapies have stopped working because of AR-V7.
Why AR-V7 Is Such a Formidable Enemy
For most men with prostate cancer, the androgen receptor (AR) is the cancer's engine. Androgens—male hormones like testosterone—bind to the AR, switch it on, and drive tumor growth. Removing or blocking androgens with drugs like enzalutamide (Xtandi) or abiraterone (Zytiga) can suppress the cancer for years. But the cancer often fights back by producing a shortened, mutant version of the AR called AR-V7 (androgen receptor splice variant 7).
AR-V7 is missing the very region—called the ligand-binding domain—that enzalutamide and abiraterone attach to. Without that docking site, those drugs simply cannot grip it. AR-V7 remains permanently switched on, pumping out cancer-growth signals regardless of hormone levels or hormone-blocking drugs.
Because AR-V7 lacks the part of the protein that most drugs latch onto, it has long been considered "undruggable." No currently approved drug can block or destroy it. Taxane chemotherapy (docetaxel, cabazitaxel) remains an option, but it carries significant side effects and does not directly target AR-V7. Patients and their doctors have urgently needed a new strategy—and now, several research teams are pursuing it through a revolutionary technology called targeted protein degradation.
The Cell's Own Trash Disposal System
Our cells are remarkably efficient at getting rid of damaged or excess proteins. They use a system called the ubiquitin-proteasome pathway: a protein called ubiquitin is attached as a "destroy me" tag to an unwanted protein, which is then fed into a barrel-shaped molecular shredder called the proteasome. The shredder chews the tagged protein to bits.
The key gatekeepers that decide which proteins get the "destroy me" tag are enzymes called E3 ubiquitin ligases. Researchers have discovered that clever small molecules can recruit these gatekeepers to cancer-driving proteins the cell would otherwise leave alone—effectively tricking the cell into destroying its own cancer machinery.
Two main families of these drug-like molecules have emerged:
PROTACs (Proteolysis Targeting Chimeras) are bifunctional molecules with two "arms"—one grabs the target protein, and one grabs an E3 ligase. A chemical linker holds them together. They work catalytically: a single PROTAC molecule can destroy many copies of its target protein, one after another.
Molecular glue degraders are smaller, simpler molecules that also recruit an E3 ligase to destroy a target—but they do it as a single compact molecule, more like a natural drug. They tend to be easier to manufacture and have better prospects for oral dosing.
The New UC Berkeley Advance: An Optimized "Molecular Glue" Handle
A research group led by Professor Daniel Nomura at UC Berkeley, working in collaboration with Novartis, published an important new preprint in March 2026 describing a refined molecular glue approach specifically designed to degrade both AR and AR-V7. Their work builds on an earlier 2023 paper from the same lab that first demonstrated the concept.
The First-Generation Problem
In their 2023 work, the Nomura team discovered that a chemical fragment called a fumarate handle—a small chemical "latch"—could be appended to a known AR-binding ligand and cause the E3 ligase RNF126 to destroy both AR and AR-V7. The concept was a striking proof of principle. However, the fumarate handle was too reactive: it also bonded to glutathione, a natural cellular antioxidant, inactivating itself quickly and also generating dose-limiting cellular toxicity. It wasn't safe enough to develop into a drug.
The Second-Generation Solution
The new 2026 paper describes the team's success in redesigning the chemical handle. By incorporating a trans-cyclobutane linker—a rigid, four-carbon ring structure—into the molecule, they more than doubled the compound's stability (measured by how long it takes glutathione to neutralize it, from 29 minutes for the old design to 68 minutes for the new one). Crucially, this improved stability translated to dramatically reduced toxicity to cells across the tested concentration range.
The optimized handle, when attached to the BET bromodomain inhibitor JQ1 (a well-studied cancer research tool compound), produced a new degrader called J594 that selectively and potently eliminated BRD4, a cancer-related protein, in cells—while showing minimal harm to normal cell viability. Critically, this degradation required RNF126, confirming the mechanism is working as intended.
Tackling AR and AR-V7 in Prostate Cancer Cells
The team then attached the same optimized handle to a ligand called VPC-14228 that binds to the DNA-binding domain of the androgen receptor—a region shared by both full-length AR and AR-V7. This produced a new compound called EST1140.
When tested in the 22Rv1 prostate cancer cell line, which expresses both AR and AR-V7 and serves as a standard laboratory model of androgen-independent, treatment-resistant prostate cancer, the results were compelling:
- EST1140 caused dose-dependent, proteasome-dependent degradation of both AR and AR-V7.
- Degradation was confirmed to require RNF126—cells engineered without RNF126 largely resisted it.
- Mass spectrometry analysis of the entire cellular protein landscape showed that AR and AR-V7 were the dominant proteins reduced, indicating high selectivity with minimal collateral damage to other proteins.
- EST1140 destabilized both AR and AR-V7 proteins (shown by cellular thermal shift assay), suggesting the drug engages these targets directly in cells.
- EST1140 suppressed AR-driven gene activity (transcriptional activity) with a half-maximal effective concentration (EC₅₀) of approximately 8 µM—and achieved near-complete suppression, whereas enzalutamide at even the highest tested doses achieved less than 50% suppression (because it cannot affect AR-V7).
- By contrast, the PROTAC drug ARV-110 (avdegalutamide) eliminated only full-length AR, not AR-V7.
Context: A Rapidly Expanding Field
This work is part of an extraordinarily active global research effort. Protein degradation as a therapeutic strategy has moved with remarkable speed from the laboratory to clinical trials. Some important landmarks for patients to know about:
PROTACs in Clinical Trials for Prostate Cancer
ARV-110 (avdegalutamide / bavdegalutamide), developed by Arvinas, was the first oral PROTAC to enter clinical trials for prostate cancer in 2019. It targets full-length AR and has shown clinical activity in men with specific AR mutations, but like all current AR-targeting drugs, it does not eliminate AR-V7. It is currently in Phase II clinical development.
BMS-986365 (gridegalutamide / CC-94676), developed by Bristol Myers Squibb, is an AR degrader that has become the second PROTAC worldwide and the first AR-targeting PROTAC to enter Phase III clinical trials. Preclinical data suggest it is approximately 100-fold more potent than enzalutamide at suppressing AR-driven gene transcription.
Overall, more than 40 PROTAC drug candidates are currently being evaluated in clinical trials for various cancers, with three compounds now in Phase III trials as of early 2026, according to a comprehensive review in MedComm. The entire timeline from first clinical trial entry (2019) to first regulatory submission (2025) spanned only six years—an unusually rapid trajectory for a new drug modality.
Parallel Molecular Glue Research for AR-V7
The Nomura lab is not alone in pursuing molecular glue-based degradation of AR-V7. At Weill Cornell Medicine, a team funded by the Daedalus Fund for Innovation identified a hit compound in a high-throughput screen of approximately 170,000 compounds that can simultaneously degrade both full-length AR and AR-V7 through a mechanism that targets the N-terminal domain (NTD) of the androgen receptor—a region that is present in both full-length AR and AR-V7. Results were presented at the 2025 AACR Annual Meeting in Chicago and described as potentially "first-in-class."
Separate work published in the Journal of Medicinal Chemistry in 2025 identified a compound called BWA-6047 that acts as both a PROTAC and a molecular glue simultaneously, degrading AR, AR-V7, and also a separate cancer-related protein called GSPT1 with extraordinary potency (DC₅₀ values in the low nanomolar range), showing efficacy in prostate cancer mouse models without obvious toxicity.
GSPT1-directed molecular glue degraders represent yet another approach. Monte Rosa Therapeutics' compound MRT-2359, a GSPT1 degrader, has entered Phase I/II clinical trials for several solid tumors including prostate cancer. Interim Phase 1 data reported in December 2025 showed encouraging results in heavily pretreated mCRPC patients, with a 100% disease control rate in AR-mutant patients when combined with enzalutamide.
What Makes This Different From Existing Options?
It helps to compare what each type of drug does—and does not—accomplish:
Important Caveats: We Are Still in the Laboratory
It is vital for patients and families to understand that EST1140 is a laboratory research compound, not a drug. It has only been tested in cell cultures—it has not yet been tested in animals, and it is far from clinical trials in humans. Several important challenges remain:
- The compound still needs to be optimized for potency, oral bioavailability (the ability to be taken as a pill and absorbed), and metabolic stability in living systems.
- Cell-based potency of approximately 8 µM (micromolar) is relatively modest for a drug candidate—most clinical drugs work at nanomolar concentrations, orders of magnitude lower.
- The compound needs extensive safety testing before it could be administered to humans.
- Many promising laboratory compounds fail to translate into effective drugs due to challenges with delivery, toxicity, or efficacy in whole-organism models.
The researchers themselves note that future work will focus on improving the compound for in vivo (living system) applications. The modular nature of the design—a "handle" that can be swapped onto different targeting ligands—gives reason for optimism that further optimization is achievable.
The Novartis–UC Berkeley Partnership: Why Industrial Backing Matters
A critical detail in this paper is easy to overlook in the fine print: this is not purely academic research. The study was conducted within the Novartis-Berkeley Translational Chemical Biology Institute—a formal joint institute between UC Berkeley and Novartis BioMedical Research, not merely a one-time grant. Multiple co-authors are listed as employees of Novartis BioMedical Research in Cambridge, Massachusetts and Basel, Switzerland. The work was explicitly co-funded by Novartis Biomedical Research alongside federal NIH and NSF grants.
This structural relationship is significant for several reasons. Academic laboratories, however brilliant, typically lack the resources to advance a compound through the lengthy and expensive process of drug development on their own. Novartis brings exactly what is needed for the next phase: large teams of medicinal chemists who can synthesize and screen hundreds of molecular variants to optimize potency and drug-like properties; sophisticated ADMET (absorption, distribution, metabolism, excretion, and toxicology) screening platforms; animal pharmacology capabilities; and the regulatory and manufacturing infrastructure required for clinical trials. When a major pharmaceutical company is embedded in the research from the beginning rather than licensing it afterward, the transition from discovery to development is considerably smoother.
Professor Nomura himself has deep commercial ties that further accelerate translational potential. He is a co-founder and equity holder in Frontier Medicines and Zenith—two biotechnology companies focused on covalent chemistry and targeted protein degradation respectively—and serves on the scientific advisory boards of Photys Therapeutics, Ten30 Biosciences, and others. He is also an Investment Advisory Partner at the venture capital firm Andreessen Horowitz Bio (a16z Bio). These relationships mean that even if Novartis were not to advance a compound directly, the scientific network around this work is dense with commercial partners who could do so.
The Road to Clinical Trials: What Comes Next
For patients wondering when a drug based on this research might reach clinical trials, it helps to understand the formal sequence of steps required. The path is well-defined—and, compared to historical norms, is moving faster in the protein degrader field than in almost any other area of drug discovery.
Step 1: Lead Optimization (Ongoing, likely 1–3 years)
The compound EST1140 described in the paper is a proof-of-concept molecule, not a development candidate. Its potency of approximately 8 micromolar needs to improve by roughly 1,000-fold before it could be viable as a drug. Novartis medicinal chemists will systematically modify the molecule's structure—tweaking the targeting ligand, the RNF126 handle, and the linker between them—to improve potency, selectivity, oral bioavailability, and metabolic stability. This generates large libraries of analogs, most of which are discarded. The modular "handle and ligand" design of the platform actually accelerates this process compared to conventional drug discovery, because improvements to the handle can be rapidly transferred across different targeting ligands.
Step 2: Pre-Clinical In Vivo Studies (1–2 years)
Once a sufficiently optimized compound is identified, it must be tested in animal models—typically mice bearing human prostate cancer tumors. These experiments establish that the compound reaches the tumor in adequate concentrations, degrades AR and AR-V7 in the tumor itself (not just in a dish), and produces tumor shrinkage or growth inhibition without unacceptable toxicity. This is the stage where many otherwise promising compounds fail.
Step 3: IND-Enabling Studies (1–2 years)
Before any human can receive a new compound, the FDA requires a formal package of safety data called an Investigational New Drug (IND) application. This includes detailed toxicology studies in multiple animal species, pharmacokinetic profiling, and demonstration of manufacturing consistency and purity. Preparing this package is time-consuming and expensive, but it is a well-understood process for a company of Novartis's size.
Step 4: Phase I Clinical Trial (First in Humans)
The first human study primarily tests safety and determines the appropriate dose range. For a prostate cancer drug of this type, Phase I would likely enroll men with heavily pretreated mCRPC—exactly the population for whom current options have been exhausted. Phase I trials typically enroll 20–80 patients and take 1–3 years. For the molecular glue field as a whole, the precedent set by PROTACs is encouraging: ARV-110 entered Phase I in 2019 and reached Phase II results within three years.
Realistic Timeline Estimate
One important nuance: the paper describes a platform as much as a single drug. Novartis may ultimately advance a compound that looks chemically quite different from EST1140, while using the same core RNF126-targeting mechanism. The clinical candidate that eventually enters trials may not carry the name EST1140 at all—but it will be the direct descendant of this work.
It is also worth noting that patients do not necessarily need to wait for this specific program. The broader question—can targeted protein degradation address AR-V7-driven resistance?—is being pursued by multiple independent groups. Any one of several currently active programs (the Weill Cornell molecular glue effort, the BWA-6047 PROTAC/glue hybrid, or others not yet publicly disclosed) could reach clinical trials first and validate the concept for patients sooner.
The Bigger Picture for Patients
For men with advanced, castration-resistant prostate cancer—particularly those who have progressed through enzalutamide or abiraterone, or who test positive for AR-V7—the options today are limited. The advances described here represent a genuinely new approach: instead of trying to outcompete AR-V7 for a binding site it doesn't have, or ignore it, these molecular glue degraders instruct the cancer cell's own machinery to eliminate the AR-V7 protein entirely.
Multiple independent research groups in the United States, Europe, and Asia are converging on this strategy from different angles—different chemical designs, different E3 ligases, different parts of the AR protein being targeted—giving the field a diversity and resilience that bodes well for eventual clinical success. The speed with which the broader PROTAC/molecular glue field has moved from laboratory to clinical trials (just six years to Phase III) suggests that once a sufficiently optimized compound is identified, clinical development could proceed relatively quickly.
Patients who are interested in clinical trials related to AR-V7 or targeted protein degradation in prostate cancer are encouraged to discuss options with their oncologist and to search ClinicalTrials.gov using terms such as "AR-V7," "PROTAC prostate cancer," or "molecular glue prostate cancer."
Verified Sources and Formal Citations
https://doi.org/10.64898/2026.03.06.709959
https://pubs.acs.org/doi/10.1021/acscentsci.2c01317
https://www.nejm.org/doi/full/10.1056/NEJMoa1315815
https://aacrjournals.org/cancerres/article/85/8_Supplement_1/1645/761030
https://pubs.acs.org/doi/10.1021/acs.jmedchem.5c00443
https://pubs.acs.org/doi/10.1021/acs.jmedchem.4c02205
https://www.stocktitan.net/news/GLUE/
https://doi.org/10.1039/d1cb00011j
https://pmc.ncbi.nlm.nih.gov/articles/PMC12114078/
https://pmc.ncbi.nlm.nih.gov/articles/PMC12495453/
https://www.scienceopen.com/hosted-document?doi=10.15212/AMM-2024-0075
https://www.biochempeg.com/article/434.html
https://doi.org/10.1038/s41589-024-01668-4
https://ascopubs.org/doi/10.1200/JCO.2019.37.7_suppl.259
https://doi.org/10.3390/ijms22115515
https://pmc.ncbi.nlm.nih.gov/articles/PMC11499039/
https://pmc.ncbi.nlm.nih.gov/articles/PMC5735831/
https://www.oncozine.com/aacr-2025-a-first-in-class-ar-v7-ar-fl-molecular-glue-degrader-may-revolutionize-prostate-cancer-treatment/
https://blog.crownbio.com/targeted-protein-degradation-with-protacs-and-molecular-glues
https://doi.org/10.1038/s41573-025-01316-z
https://www.stocktitan.net/news/GLUE/
https://doi.org/10.64898/2026.03.06.709959

Comments
Post a Comment