JQ-1

A serine/threonine kinase largely overlooked in myeloma biology drives autophagy-dependent tumor cell survival and proteasome inhibitor resistance, according to a preclinical study published August 25 in Nature Communications by researchers at Tampa, Florida-based H. Lee Moffitt Cancer Center and Research Institute. The work identifies ULK3 (Unc-51 like kinase 3) as a disease-stage-correlated dependency in multiple myeloma and describes a second-generation small-molecule inhibitor, MA9-060, that reduced tumor burden, suppressed osteolytic bone disease, and improved survival in mouse xenograft models. Corresponding author Conor Lynch and colleagues analyzed RNA-sequencing data from 813 Moffitt patient samples spanning disease stages from monoclonal gammopathy of undetermined significance (MGUS) through late relapsed/refractory multiple myeloma (LRMM), and validated findings in an independent cohort of 615 patients. ULK3 expression rose progressively across disease stages and was elevated in proteasome inhibitor-resistant cell lines, a pattern not observed for the related kinases ULK1 or ULK2. Tissue microarray analysis confirmed elevated ULK3 protein in bone marrow biopsies from newly diagnosed (n=14) and relapsed/refractory (n=11) patients compared with 15 healthy donor controls. CRISPR-Cas9 knockdown of ULK3 in U266 and MM.1S cell lines reduced proliferation and impaired autophagic flux, and lentiviral re-expression of ULK3 in knockdown cells restored proliferative capacity, confirming on-target specificity. In NSG mice bearing luciferase-labeled U266 myeloma cells, ULK3 knockdown significantly reduced tumor burden by bioluminescence imaging and improved survival compared with non-targeting controls (n=20 per group). Critically, pharmacologic treatment with MA9-060 provided no additional survival benefit in mice bearing ULK3-knockout tumors, directly confirming on-target activity in vivo. The mechanistic basis centers on the ULK3–ATG13–FIP200 autophagy initiation complex. Co-immunoprecipitation and LC-MS/MS proteomics confirmed that ULK3 physically associates with ATG13 and FIP200 in myeloma cells but not in healthy B cells, suggesting a disease-specific dependency. ULK3 appears to operate through a stress-induced, mTORC1-independent mechanism, making it relevant under the proteotoxic stress conditions generated by proteasome inhibitor treatment — precisely the context in which myeloma cells upregulate autophagy as an adaptive survival response. The researchers optimized an initial multikinase tool compound into MA9-060, which inhibited ULK3 with an IC50 of approximately 227 nM and showed around fivefold selectivity over ULK1. X-ray crystallography confirmed binding within the ULK3 ATP-binding pocket, while experiments using selective JAK2 and BRD4 inhibitors supported ULK3 inhibition as the principal driver of the anti-myeloma activity. In NSG mouse xenograft experiments (n=6–10 per group), MA9-060 administered at 10 mg/kg/day subcutaneously on a Monday-to-Friday schedule significantly reduced bioluminescence tumor signal versus vehicle and improved overall survival, according to the paper. Faxitron X-ray and histological analysis also showed protection against myeloma-associated osteolytic bone loss in treated animals, an effect not observed with JQ1 or chloroquine, suggesting a bone-protective dimension beyond direct tumor cytotoxicity. Ex vivo validation on primary patient material used the EMMA platform, a live-cell imaging system that cultures CD138+ myeloma cells in autologous bone marrow stroma. In a first cohort of 15 patients, MA9-060 reduced primary cell viability. In a second cohort of 36 patients, activity was observed across both disease settings, including samples from patients with prior proteasome inhibitor exposure. A paired RNA-sequencing and EMMA analysis in 44 patients showed that higher ULK3 expression correlated with greater ex vivo sensitivity to MA9-060, supporting ULK3 as a potential predictive biomarker. The authors also reported that MA9-060 combined with carfilzomib deepened cytotoxic responses in both newly diagnosed and relapsed/refractory patient samples. Differential activity between malignant myeloma cells and healthy donor CD19+ B cells and normal bone marrow plasma cells was observed in vitro, and MA9-060 was tolerated in NSG mice at the tested dose without reported gross toxicity. Formal preclinical toxicology studies have not been reported. No dedicated clinical program exists for ULK3 inhibition in any indication as of August 2026. All existing ULK kinase inhibitor programs, including an undisclosed ULK1/2 dual inhibitor in preclinical development at Hayward, California-based Arcus Biosciences targeting RAS-mutant solid tumors, focus on ULK1 and/or ULK2 rather than ULK3. Of note, Moffitt's Office of Innovation lists ULK3 inhibitors as patented and available for licensing. Spot something wrong? Report an issue with this article

In recent years, E3 ligases have become the focus of researchers developing modern targeted therapy strategies. The use of natural degradation mechanisms enables the selective removal of proteins responsible for the development of diseases such as cancer. Particularly promising results have been achieved in the treatment of multiple myeloma, where protein degradation–based therapies have demonstrated high efficacy and safety in clinical trials. CRBN and VHL currently form the foundation of targeted protein degradation (TPD) technology. However, to fully exploit the potential of this therapeutic strategy, further research is needed - not only focused on optimizing degraders themselves, but also on expanding the search for new E3 ligases that could complement the TPD toolkit. Broadening the set of available ligases could improve the selectivity and effectiveness of new therapies and help overcome treatment resistance that sometimes occurs with the use of the CRBN ligase. In our new publication titled Ligand-induced Conformational Plasticity of the CTLH E3 Ligase Receptor GID4, we present research results indicating the high potential of a much less characterized protein, GID4 ligase, in targeted protein degradation. In this work we describe the development of its ligands. GID4 is distinguished by high structural plasticity and the ability to recognize a wide range of molecules, which makes it a particularly interesting target for the design of new drugs. E3 ligases are enzymes that play a key role in protein degradation by attaching ubiquitin to proteins destined for removal. The GID4 protein is a receptor subunit of the E3 ligase complex known as the CTLH complex. Its primary role is to recognize characteristic protein motifs called degrons. In particular, this ligase binds so-called Pro/N-end degrons - protein sequences that begin with an N-terminal proline. This mechanism was first described in yeast, where it controls the degradation of gluconeogenesis enzymes. In humans, however, the functions of GID4 are much broader. The protein is involved in lipid metabolism regulation, influences cell migration, and participates in processes occurring in the cell nucleus, such as RNA splicing and chromatin remodeling. Structurally, GID4 adopts a characteristic β-barrel fold - a structure resembling a cylindrical barrel composed of β-sheets. Inside this barrel lies a deep hydrophobic pocket that serves as the binding site for degrons or synthetic ligands. Access to this pocket is controlled by four flexible structural loops. These can be compared to molecular “doors,” because depending on the type of molecule that binds, they can open or close the entrance to the protein interior. Loops L2 and L3 play a particularly important role - their movements largely determine the shape and volume of the binding pocket. Our crystallographic studies identified amino acids that are crucial for ligand interactions. One of the most important is the residue Glu237, which stabilizes the ligand within the binding pocket. Mutation of this residue completely abolishes the ability to bind these molecules. This type of pocket organization enables very precise recognition of specific protein motifs while remaining flexible enough to adapt to new ligands. Understanding this architecture enabled further research into the structural dynamics of GID4. One of the most important results of our work was the determination of as many as 49 crystal structures of GID4. Analysis of distances between the loops allowed us to distinguish three main conformations of the protein: Closed form – most often observed in complexes with natural peptides. Loops L2 and L3 move closer together, significantly reducing the volume of the binding pocket. Semi-open form – an intermediate state, usually induced by ligands containing aromatic rings. Open form – loops L2 and L3 are clearly separated, greatly increasing the pocket volume. This conformation proved particularly interesting from the perspective of small-molecule drug design. Work on GID4 ligands began with a small chemical fragment with relatively weak affinity of approximately 78 µM. Using a structure-based drug design strategy, the molecule was gradually optimized to reach an affinity of 23–30 nM. This represents more than a 2,500-fold increase in binding strength. A key feature of GID4 ligands is their ability to exploit regions of the protein pocket that are not occupied by natural substrates. Modifications introduced into the initial molecule extend toward additional hydrophobic residues, increasing the contact surface and strengthening interactions between the ligand and the protein residues. The new ligands were also tested in cellular systems using a thermal stability assay. The results showed that the molecules not only bind the protein in vitro, but also penetrate the cell membrane and stabilize GID4 inside cells. An increase in the protein aggregation temperature of approximately 3.6°C was observed, providing strong evidence for effective ligand–protein interaction in the cellular environment. The effect of the molecules translates into a clear improvement in protein stability, confirming the effectiveness of the ligands in binding to GID4. The next stage involved the creation of PROTAC molecules - chimeras that combine an E3 ligase ligand with a ligand for a target protein. In this case, the bromodomain inhibitor JQ1 was used, which binds the BRD4 protein selected as a potential degradation target. Although in cellular tests the PROTAC molecule did not yet lead to BRD4 degradation, biochemical experiments demonstrated the formation of a ternary complex involving GID4 and BRD4 mediated by the PROTAC molecule. This represents an important proof of concept showing that GID4 could in the future serve as a recruiter in TPD systems. Structural analyses revealed that the open conformation of GID4 has parameters very similar to the cereblon ligase, which is currently one of the best-characterized proteins used in the design of molecular glues. Molecular glues stabilize direct contact between an E3 ligase and the protein destined for degradation (the so-called neosubstrate). For this mechanism to work, the ligase must possess an appropriate combination of binding pocket volume and protein–protein interaction surface. The open form of GID4 meets these conditions almost perfectly, suggesting that this protein could also become an attractive target for the design of new drugs based on this mechanism. The significance of research on GID4 goes beyond understanding the structure of a single protein. The development of strong ligands for this ligase significantly expands knowledge about protein degradation technologies and may in the future influence the therapeutic effectiveness of TPD in medicine, enabling access to safer treatment methods. Exploring the potential of new ligases can contribute to the development of new therapies, and with further progress there is a real possibility that in the coming years GID4 will become part of a new generation of protein-degrading targeted therapies.