Abstract: TNG908 (Ralometostat) is a potent, selective, and brain-penetrant small molecule inhibitor of protein arginine methyltransferase 5 (PRMT5). It was specifically designed to exploit the synthetic lethal vulnerability of cancers harboring methylthioadenosine phosphorylase (MTAP) deletions, a genetic alteration frequently co-occurring with CDKN2A loss in 10-15% of all human cancers, including malignant peripheral nerve sheath tumors (MPNST), glioblastoma (GBM), and non-small cell lung cancer (NSCLC). Unlike first-generation PRMT5 inhibitors that lack tumor selectivity and cause dose-limiting hematological toxicities, TNG908 operates via a novel methylthioadenosine (MTA)-cooperative mechanism. By selectively binding to the PRMT5·MTA complex, TNG908 achieves a 15-fold selectivity for MTAP-null cancer cells over MTAP-wild-type (WT) normal cells. Preclinical studies demonstrate robust, dose-dependent antitumor activity and tumor regressions across diverse MTAP-deleted xenograft models. Furthermore, its optimized physicochemical properties enable significant blood-brain barrier penetration, supporting its clinical evaluation in Phase I/II trials for both central nervous system (CNS) and non-CNS MTAP-deleted solid tumors.
1. Introduction
The advent of large-scale cancer genome sequencing has identified critical genetic alterations that drive tumorigenesis, revealing novel therapeutic targets. One of the most prevalent synthetic lethal interactions in oncology is the dependence of MTAP-deleted cells on PRMT5 activity [1][2]. The MTAP gene is located on chromosome 9p21 in close proximity to the well-known tumor suppressor gene CDKN2A. Consequently, homozygous co-deletion of MTAP occurs in up to 90% of tumors with CDKN2A deletion, representing 10-15% of all human cancers [1]. This genetic profile is highly prevalent in several aggressive malignancies, including malignant peripheral nerve sheath tumors (MPNST), glioblastoma (GBM), mesothelioma, pancreatic adenocarcinoma, and non-small cell lung cancer (NSCLC) [1][2].
MTAP is a critical enzyme in the methionine salvage pathway. Its loss leads to the intracellular accumulation of its substrate, methylthioadenosine (MTA) [1]. Early efforts to target PRMT5 utilized S-adenosylmethionine (SAM)-competitive or SAM-uncompetitive inhibitors. However, these first-generation inhibitors inhibited PRMT5 equally in both MTAP-null and MTAP-WT cells, leading to a narrow therapeutic index and dose-limiting hematological toxicities [1][2]. To overcome this, TNG908 was developed as a next-generation, MTA-cooperative PRMT5 inhibitor that leverages the accumulation of MTA in MTAP-deleted cells to achieve selective tumor cell killing while sparing healthy tissues [2].
2. Pharmacological Activity
TNG908 exhibits potent and selective pharmacological activity against MTAP-deleted cancers. In vitro viability assays demonstrate that TNG908 inhibits cellular proliferation with an IC50 of 100–200 nM in MTAP-null cells, representing a 15-fold selectivity over isogenic MTAP-WT cells [1][2]. This selectivity is maintained across a broad panel of cancer cell lines independent of tumor histology [1].
In vivo, oral administration of TNG908 drives strong, dose-dependent pharmacodynamic target engagement and antitumor efficacy. In MTAP-deleted xenograft models representing GBM, NSCLC, and colorectal cancer, TNG908 achieved significant tumor growth inhibition and, in approximately 30% of models, durable tumor regressions [1][2]. Notably, sustained complete responses were observed even after the discontinuation of dosing in patient-derived xenograft (PDX) models [1].
A key differentiator of TNG908 is its ability to cross the blood-brain barrier (BBB). In non-human primates, TNG908 demonstrated a cerebrospinal fluid (CSF) to unbound plasma concentration ratio (Kp,uu,CSF) of 0.9, indicating excellent brain penetrance [1][2]. This translated to significant in vivo efficacy in a highly aggressive orthotopic GBM xenograft model, where TNG908 extended median survival by 3-fold [1].
Furthermore, TNG908 shows promising synergistic activity when combined with other targeted therapies. Combinations with MAT2A inhibitors (e.g., AG-270) and CDK4/6 inhibitors (e.g., abemaciclib, palbociclib) demonstrated robust combination benefits in MTAP-null models, providing a rationale for broad applicability across MTAP-deleted cancer types [1].
3. Molecular Mechanism of Action
PRMT5 is a type II protein arginine methyltransferase that uses the cofactor SAM to catalyze the symmetric dimethylation of arginine (SDMA) residues on proteins involved in transcription, splicing, and DNA damage repair [1][2]. TNG908 functions as an MTA-cooperative, substrate-competitive inhibitor of PRMT5. In MTAP-deleted cells, the accumulated MTA competes with SAM to form an inactive PRMT5·MTA complex [1].
TNG908 selectively binds to this PRMT5·MTA complex. Biochemical assays reveal that TNG908 binds apo-PRMT5 with a Ki of 3.2 nM, but its potency increases approximately 12-fold in the presence of MTA, achieving a Ki of 0.26 nM for the PRMT5·MTA complex [1][2]. This cooperative binding mechanism ensures that TNG908 selectively suppresses SDMA marks in MTAP-null cells (IC50 = 9 nM) compared to MTAP-WT cells (IC50 = 180 nM) [1]. The inhibition is highly specific to PRMT5, as TNG908 does not modulate asymmetric dimethylarginine (ADMA) marks conferred by Type I PRMTs and shows no significant off-target activity against a panel of 38 other methyltransferases [1][2].
4. Structure-Activity Relationship (SAR)
The discovery of TNG908 originated from a high-throughput screening (HTS) campaign using a peptide displacement assay, which identified a low molecular weight hit compound featuring an oxamide core [2]. The oxamide moiety was prioritized for optimization due to its stability and non-electrophilic nature compared to ketoamides [2].
Structure-based drug design (SBDD) guided the optimization process. X-ray co-crystal structures revealed that the pyridine ring of the hit compound π-stacked with Phe327 and Trp579. Substitution at the 6-position of the pyridine with an amino group or a carboxamide significantly improved potency. These groups formed critical hydrogen bonds with the backbone carbonyl of Glu435 and the side chain of Lys333, effectively locking Glu435 into a rotamer that sterically blocks SAM binding, thereby enhancing selectivity for the MTA-bound state [2].
Exploration of the piperidine ring revealed that a trans-2-phenyl-5-methyl substitution provided a 20-fold increase in biochemical potency [2]. Further structural analysis showed that the phenyl ring approached the backbone carbonyl of Ser310. While adding a para-hydroxyl group (phenol) improved potency via hydrogen bonding, it detrimentally increased efflux and reduced permeability, hindering BBB penetrance [2].
To balance potency with optimal CNS pharmacokinetic properties, traditional hydrogen-bond donors were replaced with isosteres. The incorporation of a benzothiazole group successfully engaged the Ser310 carbonyl lone pair through a non-traditional C-S σ* orbital interaction. This modification maintained high potency while lowering the polar surface area (PSA), resulting in TNG908's high passive permeability (MDCKII-WT A-B = 16.4 cm/s) and low efflux ratio (efflux ratio = 3), making it highly suitable for CNS targeting [2].
5. Current Limitations
Despite its robust preclinical profile, there are current limitations and challenges associated with TNG908. First, while TNG908 drives strong single-agent activity, complete tumor regressions were observed in only about 30% of the MTAP-deleted xenograft models tested. The specific factors and potential additional biomarkers that dictate complete sensitivity versus partial resistance to PRMT5 inhibition remain unknown and require further investigation [1].
Second, while combination therapies show promise, they carry potential risks. For instance, the combination of an MTA-cooperative PRMT5 inhibitor with a MAT2A inhibitor demonstrated strong synergy in MTAP-null models, but synergy was also observed in MTAP-WT cells at high concentrations. This suggests that such combinations might be limited by overlapping on-target toxicities in healthy tissues [1].
Finally, there are complexities in modeling brain penetrance preclinically. TNG908 showed a reduced unbound brain partition coefficient in mice (Kp,uu ~0.15) compared to non-human primates (Kp,uu ~0.6-0.9) due to species differences in efflux transporter expression. While non-human primates are considered better predictors of human BBB penetrance, this discrepancy complicates the interpretation of efficacy in rodent orthotopic models [1].
6. Future Perspectives
TNG908 represents a significant advancement in precision oncology, successfully translating the concept of collateral lethality into a viable therapeutic strategy. It is currently being evaluated in a Phase I/II clinical trial (NCT05275478) to assess its safety, tolerability, and preliminary antitumor activity in patients with MTAP-deleted advanced or metastatic solid tumors, including glioblastoma [1][2].
The exceptional brain penetrance of TNG908 positions it as a highly promising candidate not only for primary CNS malignancies like GBM but also for MTAP-deleted tumors that frequently metastasize to the brain, such as NSCLC and melanoma [1]. Future clinical development will likely explore rational combination strategies. Because nearly all MTAP-deleted tumors also harbor CDKN2A deletions, combining TNG908 with brain-penetrant CDK4/6 inhibitors (e.g., abemaciclib) offers a strong biological rationale and has shown clear preclinical benefit [1]. Continued research into the adaptive responses to PRMT5 inhibition and the identification of predictive biomarkers will further refine patient selection and maximize the clinical utility of TNG908 across diverse oncology indications, including sarcomas and MPNSTs.
