Therapeutic effect of novel drug candidate, PRG-N-01, on NF2 syndrome-related tumor

Abstract

Background

NF2-related schwannomatosis (NF2-SWN) is associated with multiple benign tumors in the nervous system. NF2-SWN, caused by mutations in the NF2 gene, has developed into intracranial and spinal schwannomas. Because of the high surgical risk and frequent recurrence of multiple tumors, targeted therapy is necessary. However, there are no approved drugs.

Methods

We examined the action mechanism of PRG-N-01, a candidate molecule for NF2-SWN, through the direct binding assay and mass spectrometry. For in vitro anti-proliferative experiments, primary cells derived from the NF2 mouse model and patient tumors were treated with PRG-N-01. The in vivo therapeutic and preventive efficacy was validated via intraperitoneal and oral administration in the NF2 mouse model (Postn-Cre; Nf2f/f). Gene expression profile in the DRG of the mouse model was explored by RNA sequencing. The pharmacological properties of PRG-N-01 were analyzed through the preclinical study.

Results

PRG-N-01 binds to the N-terminal extremity of TGFβR1 (TβR1) kinase domain, where TβR1 and RKIP interact, inhibiting the binding and preventing degradation of RKIP. In vivo administration in the mouse model suppressed schwannoma progression in the DRG. Early oral administration of the PRG-N-01 also demonstrated preventive effects on NF2-SWN. PRG-N-01 treatment suppressed tumor growth genes while upregulating genes related to for normal cell metabolism and Schwann cell differentiation in DRG. PRG-N-01 showed druggable properties through the preclinical study, including ADME, pharmacodynamics, pharmacokinetics, and toxicology.

Conclusions

Together, our study provides the rationale and critical data for a prospective clinical trial of PRG-N-01 in NF2-SWN patients indicating PRG-N-01 as a promising candidate for the treatment.

Graphical Abstract

Open in new tabDownload slide

NF2-related schwannomatosis (NF2-SWN) is a rare autosomal dominant genetic disorder, caused by loss-of-function mutations of the NF2 tumor suppressor gene encoding a protein called Merlin. NF2-SWN is characterized by the growth of multiple tumors in the nervous system, particularly those in the skull and spine¹ and is associated with intracranial and spinal schwannoma, meningioma, and ependymoma.² NF2-related tumors are benign, but their high risk of surgical removal and frequent occurrence of multiple recurrent tumors require targeted therapies.^3,4 Unfortunately, there are currently no targeted drugs approved by the FDA to treat NF2-SWN patients. As non-targeted treatment, Bevacizumab, a humanized monoclonal antibody that targets vascular endothelial growth factor A, has been shown to improve hearing and reduce tumor volume in about 40% of patients with progressive NF2-associated vestibular schwannomas (VS).⁵ However, the drug’s effects are not durable in patients who do respond, and bevacizumab may induce serious adverse effects, such as hypertension and proteinuria, that make long-term treatment difficult for some patients.⁶ Clinical phase 2 trials, such as Brigatinib and Selumetinib, are presently in progress to explore the effectiveness of targeted therapies in treating NF2-associated VS, yet the drug’s efficacy remains uncertain. Therefore, more effective and safer targeted therapies are urgently needed for NF2-associated tumors.

Merlin, a product of the NF2 gene, is a cell membrane protein that mediates contact-dependent inhibition of proliferation as a tumor suppressor.⁶ Loss of function of Merlin results in deregulation of numerous signaling pathways involved in tumorigenesis, including the Hippo pathway, Raf/MEK/ERK, PI3K/Akt, and mTOR signaling pathways.^7–10 Raf-kinase inhibitor protein (RKIP) has been identified as a negative regulator of the Raf-MEK-ERK pathway and is an important negative modulator of the PI3Kinase/Akt and mTOR signaling pathways involved in various biological processes, such as cell proliferation and differentiation.^11–13 Low or loss of RKIP expression is associated with tumor cell survival, proliferation, and metastasis in numerous human cancers, indicating that RKIP is an endogenous tumor suppressor protein.^14–16

Our previous study demonstrated that physical contact-induced reduction of RKIP leads to Merlin instability, resulting in tumorigenic progression in mesothelioma.¹⁷ Based on these findings, a novel pathomechanism of NF2-SWN was identified: Merlin deficiency decreases the stability of transforming growth factor-beta receptor 2 (TβR2), and physical stimuli such as pressure or heavy materials induce unnecessary interaction between transforming growth factor-beta receptor 1 (TβR1) and RKIP. This neo-interaction results in the phosphorylation and degradation of RKIP. RKIP reduction consequently mediates MAPK activation as well as Snail-mediated p53 suppression and occurrence of EMT in NF2-deficient cells.¹⁸ To obtain drugs that ameliorate this new NF2-SWN pathology, we screened candidate chemicals that selectively inhibit the aberrant interaction between TβR1 and RKIP, which induces a reduction of RKIP. A novel chemical, Nf18001, was selected finally and it inhibited proliferation and promoted differentiation of HEI-193 human schwannoma cells, and suppressed tumor growth in an allograft model without interfering with canonical TGF-β signaling.¹⁹

Here, we present a report on the biological characteristics of orally bioavailable PRG-N-01 (new code name of Nf18001). PRG-N-01 binds to the TβR1 kinase domain and inhibits the direct interaction between TβR1 and RKIP. In vivo administration of PRG-N-01 showed therapeutic and preventive effects by reducing the incidence of multiple sporadic tumorigenesis in NF2 model mice and inhibiting the development of schwannomas in the dorsal root ganglion (DRG). Additionally, DRG transcriptome analysis revealed that PRG-N-01 treatment decreased the cell division pathway and increased the oxidative phosphorylation pathway in tumor tissue. Furthermore, PRG-N-01 exhibited reasonable pharmacological properties as a drug in preclinical studies.

Materials and Methods

Mice

All experimental procedures involving laboratory animals were approved by the Animal Care Committee of Pusan National University (Approval number, PNU-2023-0113). NF2 (FVB/NJ) mice were obtained from Dr. D.W. Clapp (Indiana University, Indianapolis, IN). FVB/NJ mice were obtained from The Jackson Laboratory. See Supplementary Materials for details.

Image Acquisition of [18F] FDG PET/CT

Micro-PET images were acquired using an Inveon PET/CT scanner (Siemens Healthcare). [18F] FDG (580 ± 30 μCi) was administered intravenously through the tail vein. All of the micro-PET images were reconstructed using a 2-dimensional ordered-subset expectation maximum (OSEM) algorithm, and focal accumulations in the microPET images were quantified by region of interest analysis. Signal-to-background ratios were calculated using average counts per voxel on the coronal images.

Tumor Volume Quantitation

Tissue processing and quantification of dorsal root ganglion (DRG) volume were performed following previously established methods [21]. Spines were fixed with tissue fixative and then decalcified using 5% formic acid. DRGs and spinal nerves were meticulously isolated under a dissecting microscope (ZEISS Stemi 508). Tumor volume was determined by capturing DRG images with a microscope camera (ZEISS Axiocam 208). The tumor volume was calculated using the formula: volume = length × width × 0.52, where length and width measurements approximated the dimensions of a specific spheroid-shaped tumor.

Protein–Protein Interaction Analyses

GST-pull down assay and immunoprecipitation assays were performed to evaluate protein–protein interactions. For GST pull-down assay, agarose-bead–conjugated GST–TβR1 kinase domain aa 148-503 (KD) recombinant protein was incubated with HA-tagged RKIP- or FLAG-tagged Smad-2 or Smad3-transfected HEK 293 cell lysates or HA-tagged TβR2-transfected HCT116 cell lysates in PBS for overnight at 4°C. To evaluate the RKIP-MEK1/2 interaction, agarose-bead-conjugated GST-RKIP recombinant protein was incubated with HEK 293 cell lysates. For the immunoprecipitation assay, FLAG-tagged c-Raf or HA-tagged RKIP-transfected HEK 293 cell lysates in PBS were used. Whole lysates were incubated with appropriate primary antibodies overnight at 4°C and reacted with agarose-bead–conjugated protein A/G (Invitrogen) for 2 h. After centrifugation, the precipitates were washed twice with RIPA buffer and subjected to SDS-PAGE and western blotting analysis.

LC-MS/MS Analysis

The LC-MS/MS analysis was performed on a Waters UPLC Acquity I class plus coupled to a Qtrap 6500 plus (Applied Biosystems/Sciex). A Phenomenex, Aeris, 3.6 µm, 100×2.1 mm column was used for separation. The MS system was controlled by Analyst 1.6.1 software. See Supplementary Materials for details.

Prediction of TβR1-RKIP Complex and Molecular Docking

The amino acid sequences of the cytoplasmic domain (148-503) of TβR1 (P36897-1, Length 503) and the full-length RKIP (A0A1C6YP47, Length 190) were submitted as separate chains to the ColabFold (56) v1.5.2 (AlphaFold2 using MMseq2) using the default settings and the homodimer options: msa_mode = mmseq2_uniref_env, pair_mode = unpaired_paired, model_type = auto, num_recycles = auto, recycle_early_stop_tolerance = auto, max_msa = auto, num_seeds = 1. Predicted structures with best-ranked model 1 were selected for further analysis and visualized using Schrödinger Maestro software. Molecular docking experiments were performed using Schrödinger Glide software (version 2022-4). The predicted TβR1-RKIP model generated from ColabFold was imported to Maestro, and the protein was prepared for docking using the Protein Preparation Workflow with the OPL4 force field. The TβR1 portion was isolated, and the receptor grid for docking was defined as the centroid of 250Q, 251T, 255R, 265A with length 20A. PRG-N-01 was imported to Maestro in a SMILES format, and the compound was prepared for docking using the LigPrep module, which generated all possible states at a target pH of 7.0 ± 2.0. The prepared compounds were docked to the protein structure using the Glide software with extra precision (XP) settings.

For detailed methodologies of Western blotting analysis, histology and immunohistochemistry, bulk RNA sequencing, preclinical study, and so on, see Supplementary Materials.

Statistical Analysis

Statistical differences were assessed using the Student’s t-test or 1-way ANOVAs combined with the Dunnett’s test for multiple comparisons (GraphPad Prism software, version 9.5.1). P-values < .05 were considered significant.

Results

Identification of the TGFβR1 Binding Site for RKIP

Previously, we reported that NF2 deficiency decreases the stability of TβR2 and induces unnecessary interaction between TβR1 and RKIP.^18,19 This interaction induces phosphorylation and degradation of RKIP. Considering that the decomposition of RKIP is influenced by the catalytic activation of TβR1,^18,19 it is expected that RKIP would bind to the kinase domain of TβR1. Therefore, TβR1 kinase domain (148-503) and RKIP recombinant protein were produced and GST-pulldown assay was conducted (Figure 1A). As shown in Figure 1B, RKIP binds to the kinase domain of TβR1 directly. To identify the more specific region responsible for the TβR1-RKIP interaction, we performed further pulldown assay using various TβR1 fragments (Figure 1A). RKIP strongly binds specifically to the N-terminal extremity of TβR1 kinase domain (230-322) (Figure 1C). Through in silico molecular docking analysis, 6 residues (225R, 250Q, 251T, 255R, 265A, and 323T) in TβR1 230-322 region, were estimated to have hydrogen bonds less than 1Å distance with RKIP (Figure 1D and  Supplementary Figure 1A). Consequently, the docking grid was generated centered around 250Q, 251T, 255R, and 265A of TβR1. Using mutants that substituted the specified residues, in vitro binding assay was conducted, revealing that A265 could be a critical residue of TβR1 for binding to RKIP (Figure 1E). Our result demonstrated that there is the neo-protein-protein interaction between the TβR1 kinase domain and RKIP. Therefore, it is essential to verify whether our drug, which inhibits the binding of TβR1 and RKIP, targets this specific binding site to enhance target accuracy, maximize therapeutic effects, and minimize side effects.

PRG-N-01 Selectively Inhibits the Interaction Between TβR1 and RKIP by Specifically Binding to the TβR1 kinase domain

As it had been demonstrated that the binding of a small molecule to the proteins affects the integrity of the protein–protein interaction through conformational changes or competes with the original protein partner,²⁰ we explore the binding mode of PRG-N-01 to TβR1-RKIP complex. To confirm that PRG-N-01 inhibits the interaction of TβR1 and RKIP through precise binding region, in vitro binding assays were performed using TβR1 kinase domain recombinant protein. PRG-N-01 interrupted the interaction of TβR1 kinase domain (148-503) and RKIP directly (Figure 2A and B). In addition, PRG-N-01 also inhibited interaction at a more precise binding site of TβR1 (230-322) and RKIP identified in Figure 1C (Figure 2C). To validate the accurate target of PRN-N-01 in the TβR1-RKIP complex, using LC-MS/MS analysis, in vitro binding assay between PRG-N-01 and recombinant proteins (GST-TβR1 kinase domain and GST-RKIP) was performed. The LC-MS/MS analysis showed that PRG- N-01 has an affinity for the TβR1 kinase domain, but not for the GST alone and GST-RKIP (Figure 2D and E). To explore the plausible binding mode of PRG-N-01 and gain further insight into how it can interfere with RKIP binding to TβR1, we conducted in silico molecular docking analysis. The molecular docking experiment yielded a plausible binding mode of PRG-N-01 to TβR1. The primary amine of the PRG-N-01 formed 2 hydrogen bonding interactions with the backbones of 250Q and 252V, and it was further stabilized with a pi-cation interaction between the central triazole and 255R (Figure 2F and  Supplementary Figure 1B). In in vitro binding assays, PRG-N-01 did not inhibit the binding of the TβR1 R255A mutant to RKIP, in contrast to its successful inhibition of the interaction between the TβR1 WT and RKIP. Moreover, the binding of the TβR1 T251A mutant to RKIP was inhibited partially by PRG-N-01. These findings suggest that the residues Thr251 and Arg255 of TβR1 could be critical sites for the binding of PRG-N-01 to TβR1 (Figure 2G). Accordingly, our results underscore that the neo-protein-protein interaction between the TβR1 kinase domain and RKIP could be a target site for the treatment inhibiting tumor formation under NF2 deficient conditions.²¹

In the TGF-β signaling pathway, upon binding of the activated TGF-β ligand to TβR2, it recruits and activates TβR1,^22,23 which subsequently activates downstream mediators Smad2 and Smad3.^24,25 To evaluate the specificity of the PRG-N-01, we investigated the effect of PRG-N-01 on the interaction between TβR1 and its binding partners including TβR2, Smad2, and Smad3. We observed that PRG-N-01 had no appreciable effects on the interaction of TβR1-TβR2 (Supplementary Figure 2A), TβR1-Smad2 (Supplementary Figure 2B), and TβR1-Smad3 (Supplementary Figure 2B). Furthermore, PRG-N-01 did not affect the interaction between RKIP, a physiological endogenous inhibitor of the Raf-1/MEK/ERK pathway,²⁶ with MEK1/2 (Supplementary Figure 2C) and c-Raf (Supplementary Figure 2D). These results underscore that PRG-N-01 selectively targets the TβR1-RKIP interaction but no other interactions involving major partners that interact with TβR1 or RKIP. There is a concern that PRG-N-01 may affect the kinase activity of TβR1 due to its binding to the TβR1 kinase domain. However, it is suggested that PRG-N-01 does not directly affect the kinase active site at lysine 232,²¹ but rather binds around the active site (Figure 2F and  Supplementary Figure 1B). Our previous study has shown that PRG-N-01 does not affect Smad phosphorylation mediated by TβR1, unlike the TβR1 kinase inhibitor TEW7197.¹⁹ Therefore, it suggests that PRG-N-01 could minimize side effects originating from inhibition of the TβR1 kinase activity, not like other inhibitors.

Intraperitoneal Administration of PRG-N-01 Demonstrates Therapeutic Effects in Schwannoma of NF2 Model Mouse

In our previous study, PRG-N-01 showed the inhibitory effect on tumor growth in an allograft model with mouse schwannoma cells. To validate the efficacy of the drug, it is crucial to verify whether it is effective in transgenic mouse models. Prior to conducting in vivo efficacy testing, we assessed the anti-proliferative effect of PRG-N-01 on tumor cells derived from NF2 mouse model. We obtained primary Nf2-deficient tumor cells by culturing cells isolated from tumor tissues of dissected NF2 mouse model for one month, resulting in 2 types of homogeneous cell populations (#149 & #184) (Supplementary Figure 3A). Compare to mouse fibroblast Nor-10 and HEI-193 cells lacking both isoforms 1 and 2,²⁷ these cells did not express merlin. To evaluate the drug efficacy, we utilized #184 cells, which completely lack the merlin protein (Supplementary Figure 3B). Treatment with PRG-N-01 resulted in a dose-dependent reduction in the proliferation of #184 cells (Figure 3A), while the expression of Schwann cell differentiation markers, GFAP was increased (Figure 3B). These findings demonstrate that PRG-N-01 has the potential to prevent Nf2-deficient tumor cell proliferation in the NF2 mouse model. Restoration of NF2 expression also reduced proliferation of #184 cells (Supplementary Figure 3C  and  D). The NF2 mouse model used in this study is a genetically engineered Postin-Cre; Nf2^flox/flox mouse model.²⁸ These mice develop spinal, peripheral, and cranial nerve tumors that closely resemble human schwannomas by 10 months of age, with 100% penetrance.²⁹ To confirm the therapeutic effect of PRG-N-01 in this model mouse, PRG-N-01 was administered at the 9 months of age through intraperitoneal (i.p.) injection (Figure 3C). Postin-Cre; Nf2^flox/flox mice exhibit spontaneous Schwann cell hypertrophy or schwannoma in the dorsal root ganglion (DRG) from 5 months of age, and all mice at 15 months of age possess significantly enlarged DRG.²⁹ Based on this information, we checked the occurrence of schwannoma in NF2 mouse model at 9 months of age and the volumes of DRG from dissected spinal cords were increased (Figure 3D). After 5 months of treatment, the vehicle-treated group showed a noticeable increase in DRG size, whereas the PRG-N-01-treated group did not exhibit a significant increase in overall tumor size (Figure 3D and  Supplementary Figure 4A), particularly showing a marked reduction in the proportion of large tumors (> 0.8 mm) in lumbar regions (Figure 3D and  Supplementary Figure 4B). To further investigate in vivo therapeutic effects of PRG-N-01, we evaluated the expression level of Schwann cell differentiation markers, including myelin basic protein (MBP), SRY-Box Transcription Factor 10 (SOX10)³⁰ and RKIP, in tumor tissue following PRG-N-01 administration (Figure 3G). Immunohistochemistry showed increased expression of MBP (Supplementary Figure 4C), SOX10 (Supplementary Figure 4D) and RKIP (Supplementary Figure 4E) in the DRG in the Cre + group administered with PRG-N-01. Schwannoma of the Postin-Cre; Nf2^flox/flox exhibits similarities to NF2-SWN disease in humans, where histologically, Schwann cells proliferate and form whorls around the axon in DRG, eventually developing into a Schwannoma.²⁹ To investigate whether PRG-N-01 inhibits schwannoma formation in DRG, the number of whorl patterns formed by the proliferation of Schwann cells were counted. Compared to before treatment, the number of whorl patterns significantly increased in the 13-month Cre + vehicle-treated group, whereas the PRG-N-01-treated group showed a decrease (Supplementary Figure 4F). Analysis of the relationship between the number of RKIP-positive satellite cells and the number of Schwann cell whorls in the vehicle and PRG-N-01-treated group showed a negative correlation (P < .05) (Supplementary Figure 4G), suggesting that the reduced RKIP expression caused by NF2 deficiency, could be a pivotal pathological mechanism driving NF2-related tumorigenesis. These results demonstrate that in vivo administration of PRG-N-01 may have therapeutic potential for NF2 syndrome by inhibiting tumorigenic growth and promoting differentiation of NF2-deficient Schwann cells.

PRG-N-01 Exhibits Preventive Effect Against Schwannoma in NF2 Mouse model via Oral Administration

Given that NF2-SWN manifests in adolescence and persists lifelong, assessing the preventive efficacy of PRG-N-01 via oral administration is essential. PRG-N-01 was administered beginning at the 11 weeks of age (Figure 4A). To evaluate the prevention potential of PRG-N-01, mice were orally administered 25 or 50 mg/kg of the drug mixed with their diet for 9 or 13 months. The control group received a diet mixed with vehicle (solvent only) under the same conditions (Figure 4A). PET/CT with 18F-FDG enables the detection of glycolytic activity and radiotracer accumulation in tumor areas,³¹ which reflects the intensity of tumor glucose metabolism.³² In this study, the decreased patterns of tumor location numbers and radioactivity uptake intensities in the mice, to which PRG-N-01 was orally administered for 9 months, were examined (Figure 4B, C and  Supplementary Figure 5A). Noticeably, the strongest reduction of radiotracer accumulation was observed in the periorbital area of the PRG-N-01-treated group (Figure 4D). These findings suggest that PRG-N-01 may effectively suppress spontaneous multiple schwannoma formation in the NF2 mouse model. The DRG volume in Cre + PRG-N-01-treated group was significantly reduced by 49.4% compared to the vehicle-treated group (Supplementary Figure 5B  and  C) in both the thoracic (Supplementary Figure 5D) and lumbar regions (Supplementary Figure 5E). The number of whorl patterns was not observed at all in the untreated Cre- group, whereas it was significantly increased in the vehicle-treated Cre+ group (Figure 4E). However, in comparison to number of whorl patterns in the vehicle-treated group, their number in the PRG-N-01-treated Cre+ groups was significantly reduced by 55.9% and 54.9%. (Supplementary Figure 5F). These findings suggest that PRG-N-01 can effectively suppress Nf2-deficient Schwann cell proliferation in DRG. In DRG tissue, the expression level of Schwann cell differentiation markers, MBP and SOX10, was increased (Figure 4F, Supplementary Figure 5G, H). Furthermore, the number of RKIP-positive satellite cells wrapping around axons was decreased in the Cre+ group compared to the control group (Supplementary Figure 5I) and the reduction was rescued by the administration of PRG-N-01 (Supplementary Figure 5J). The number of RKIP-positive satellite cells and Schwann cell whorls in the vehicle and PRG-N-01-treated group showed a negative correlation (p < 0.05) (Supplementary Figure 5K). These results indicate that administering PRG-N-01 at an early stage demonstrates a disease-preventive effect for NF2-SWN.

Gene Expression Profile of Dorsal Root Ganglia Schwannoma of PRG-N-01-treated NF2 Mouse Model

To explore the effect of PRG-N-01 on gene expression in the mouse model, we conduct RNA sequencing in the dorsal root ganglia (DRG) of vehicle-treated (G3) and PRG-N-01-treated NF2 mouse model (G4). Our analysis revealed a distinct group of 180 differentially expressed genes (DEGs) in the DRGs, with 157 genes being down-regulated and 23 genes up-regulated upon PRG-N-01 treatment. These DEGs are well divided into G3 and G4 groups (Supplementary Figure 6A  and  B). Gene set enrichment analysis (GSEA) indicated 23 statistically significant Hallmark gene sets. In the PRG-N-01-treated group, mitotic spindle gene sets were depleted; in contrast, oxidative phosphorylation (OXPHOS) and fatty acid metabolism gene sets were enriched (Supplementary Figure 6C). This suggests that PRG-N-01 treatment inhibits tumor growth drivers in the DRG of the NF2 mouse model while promoting gene sets associated with oxidative phosphorylation and lipid metabolism. Importantly, the enhancement of OXPHOS in PRG-N-01 treated groups signifies a reversal of the typical metabolic reprogramming observed in various cancer cells, where glycolysis is enhanced and OXPHOS capacity is diminished.³³ Upregulation of fatty acid gene sets suggests that PRG-N-01 promotes Schwann cell differentiation by facilitating the metabolic processes essential for myelin production.³⁴ Notably, the leading-edge genes in the oxidative phosphorylation and the fatty acid metabolism gene sets were significantly upregulated, while those in the mitotic spindle assembly sets were significantly downregulated (Supplementary Figure 6D  and  E). RT-qPCR validation revealed that PRG-N-01 induced an increase in oxidative phosphorylation-related genes, such as NDUFB8 and NDUFA4 and lipid metabolism–related genes, such as S100A10 and EPHX1 in HEI-193 cells (Supplementary Figure 6F-I). These findings strongly support our in vivo results, suggesting that PRG-N-01 not only suppresses Schwannoma formation but also promotes the differentiation of tumor cells into Schwann cells in the NF2 mice model (Figures 3 and 4).

Comparative Study of PRG-N-01, Brigatinib, and Selumetinib for the Treatment of Nf2-deficient Schwannoma Cells

For the treatment of NF2-SWN, several targeted therapies used in various cancers are being applied and clinical phase 2 trials are currently underway. One trial involves the use of Brigatinib (NCT04374305), a small molecule inhibitor that targets FAK1 and EphA2 in the NF2-associated signaling cascade,³⁵ while the other involves Selumetinib (NCT03095248), a MEK1 inhibitor that targets the Ras/ERK signaling cascade.³⁶ We compared the in vitro anti-proliferative effects of PRG-N-01 with that of Brigatinib and Selumetinib in NF2-deficient tumor cells. Both PRG-N-01 and Brigatinib dose-dependently reduced the proliferation of HEI-193 cells, whereas Selumetinib did not exhibit any effect (Figure 5A). Furthermore, PRG-N-01 showed the most significant growth inhibition in primary mouse Nf2-deficient tumor cells (Figure 5B). However, Brigatinib showed toxicity in human normal fibroblasts even at low concentration, while PRG-N-01 did not exhibit any significant cytotoxic effect even at high concentrations (Figure 5C). Brigatinib induced cytochrome c release from mitochondria³⁷ (Figure 5D) and activated caspase-3,³⁸ and inactivated PARP³⁹ in normal fibroblast, suggesting apoptosis (Figure 5E). Caspase-3 activation and PARP inactivation were also observed in HEI-193 cells treated with Brigatinib (Figure 5F). While, treatment with PRG-N-01, regardless of the concentration used, decreased the expression of stem cell factor SOX2 but increased the expression of a Schwann cell differentiation marker SOX10, MBP,²⁸ and cell cycle negative regulators p21⁴⁰ and p27⁴¹ in Hei-193 cells (Figure 5G and  Supplementary Figure 7A-F). Brigatinib decreased the expression of all these markers. In contrast to Brigatinib, treatment with PRG-N-01 did not affect the total cell count, but the number of S100B-positive cells, representing Schwann cell, decreased in NF2-SWN patients derived to primary human vestibular schwannoma (VS) cells (Supplementary Figure 8A-C). Additionally, PRG-N-01 downregulated SOX2, a stem cell factor, but upregulated Schwann cell differentiation markers SOX10, MPZ, and GFAP (Supplementary Figure 8D). These findings demonstrate the selective anti-cell growth effect of PRG-N-01 on human primary VS cells, supporting its potential as a therapeutic agent targeting NF2-deficient schwannoma.

Preclinical Profiling of PRG-N-01

To determine whether PRG-N-01 could successfully proceed to clinical trials, we conducted preclinical research. PRG-N-01 was soluble up to 100 µM in 0.3% DMSO in extracellular solution (ECS) with a pH of 6.9, and stable in plasma from CD-1 mice, SD rats, beagle dogs, cynomolgus monkeys, and humans after 120 min (Supplementary Table 1). It showed high permeability in Caco-2 cells without being a significant BCRP substrate and demonstrated moderate permeability across MDR1-MDCK II cells, indicating it is not a strong P-glycoprotein substrate (Supplementary Table 2). PRG-N-01 underwent high metabolism in CD-1 mouse and beagle dog hepatocytes, moderate metabolism in SD rat, cynomolgus monkey, and human hepatocytes (Supplementary Table 1), and showed moderate to slow metabolism in liver microsomes across these species (Supplementary Table 1). Pharmacokinetic studies revealed low clearance rates (29.3 to 66 mL/min/kg CL), long elimination half-lives (1.03 to 2.03 h), and low to moderate distribution volumes (1.3 to 5 liter/kg) post-intravenous injection (Figure 6A–D). Oral administration led to moderate to high bioavailability (52.3% in mice, 54.9% in rats, 68.8% in dogs) with peak blood concentration (T_max) between 0.67 and 24.7 h (Figure 6A–D). Dose-related increases in maximum blood concentration and exposure were noted (Figure 6D), indicating good oral availability and pharmacokinetic properties of PRG-N-01 in preclinical species.

To establish the safety of PRG-N-01, particularly for NF2-SWN patients who may require combination therapy and have additional health complications, a comprehensive panel of in vitro safety assays was performed. PRG-N-01 was not found to be a substrate for the transporters OATP1B1, OATP1B3, OAT1, OAT3, OCT2, MATE1, and MATE2-K under tested conditions. However, it inhibited OCT2, MATE1, and MATE2-K with IC50 values of 0.663, 1.30, and 1.25 μM, respectively, indicating potential drug-drug interaction (DDI) risks (Supplementary Table 2). Importantly, PRG-N-01 did not inhibit any major cytochrome (CYP) enzymes (CYP 1A2/2B6/2C8/2C9/2C19/2D6/3A), although CYP2C8 was identified as a major metabolic enzyme for metabolite M5 formation, with CYP3A also potentially involved (Supplementary Tables 2, 3, and  6). Further in-depth studies are currently underway to assess DDI risks more comprehensively.

In evaluating the effects of PRG-N-01 on human ether-à-go-go related gene (hERG) potassium channels via automated patch clamp in HEK293 cells, it exhibited low cardiotoxicity with an IC50 of 15.66 µM (Supplementary Table 4). Additionally, PRG-N-01 showed low genetic toxicity risk in mini Ames and micronucleus assays, indicating minimal mutagenicity and chromosomal damage (Supplementary Table 5). Testing against a panel of receptors, channels, and enzymes, PRG-N-01 significantly affected Alpha2A by 60.94% at 10 μM, with negligible effects on 33 other targets (Supplementary Table 6).

In assessing the safety of PRG-N-01, single-dose toxicity tests showed maximum tolerated doses of 300 mg/kg in rats and 1000 mg/kg in dogs (Supplementary Table 5). Follow-up studies, including 2 and 4-week repeated oral toxicity tests, aimed to evaluate long-term effects and established the no-observed-adverse-effect level (NOAEL) at 25 mg/kg in rats and 80 mg/kg in dogs. These results indicate a promising safety profile of PRG-N-01 for clinical development (Supplementary Table 5).

Discussion

The majority of NF2-SWN patients harbor NF2 gene mutations, resulting in an imbalance of TGF-β receptors that promotes TβR1-mediated phosphorylation and degradation of RKIP.¹⁸ In a previous study, we developed a novel NF2-SWN therapeutic candidate, PRG-N-01 (new code name of Nf18001), which inhibits TβR1-mediated RKIP degradation without disrupting canonical TGF-β signaling.¹⁹ Understanding more detailed molecular mechanisms of PRG-N-01 and validating efficacy in model mice and obtaining preclinical data is essential for developing clinical treatment for patients with NF2-SWN. PRG-N-01 effectively disrupts the binding between TβR1 and RKIP by targeting the amino acids 230-322 of TβR1, which are crucial for RKIP binding (Figures 1C, 1D and 2C). The in vitro binding assay further confirms these findings by showing that PRG-N-01 readily binds to the TβR1 kinase domain but not to RKIP (Figure 2D and E). Based on the inhibitory effect of PRG-N-01 on tumor growth in an allograft model with mouse schwannoma cells, we assessed the efficacy of PRG-N-01 using the NF2 mouse model (Postn-Cre; Nf2^f/f) in this study. The therapeutic effect in the mouse model was confirmed by administering PRG-N-01 via i.p. injection after tumor formation (Figure 3). The volume of DRG and the number of whorl patterns reflecting the occurrence of Schwann cell hypertrophy or schwannoma, was decreased in the PRG-N-01-treated group. Besides, the expression of schwann cell differentiation markers and RKIP was increased (Figure 3  and  Supplementary Figure 4). PRG-N-01 also has an NF2-SWN prevention effect via early oral administration by demonstrating inhibition of schwannoma-genesis and inducing tumor cell differentiation in the Nf2-deficient mice model (Figure 3). Transcriptome analysis demonstrates the tumor growth inhibitory effect and the normal metabolism and cell differentiation promoting effect by PRG-N-01 (Supplementary Figure 6), strongly supporting our in vivo results (Figures 2 and 3). The in vitro antitumor efficacy of PRG-N-01, Brigatinib, and Selumetinib in NF2-deficient tumor cells was compared (Figure 5A and B). Notably, PRG-N-01 showed no significant cytotoxicity against human normal fibroblasts in contrast to Brigatinib (Figure 5C and F). Stem cell factor SOX2 was decreased, and Schwann cell differentiation marker and cell cycle negative regulators were increased by PRG-N-01 treatment (Figure 5G).

Furthermore, the pharmacokinetics (Figure 6), metabolic stability (Supplementary Table 1) and safety profile (Supplementary Tables 4 and 5) of PRG-N-01 were investigated both in vitro and in vivo, demonstrating its druggability is promising. NF2-SWN manifests during adolescence and has a wide range of severity, necessitating long-term treatment, which makes the development of orally administered medications essential. Our preclinical data indicate that PRG-N-01 has good metabolic stability, favorable pharmacokinetics upon oral administration, and a bioavailability (BA) exceeding 50% in all cases. Therefore, the development of an oral formulation for PRG-N-01 appears to be highly promising. Based on repeated oral toxicity tests and following cautious dosing recommendations for the first administration in humans, the starting dose was set at 25 mg, and it will be increased to a maximum of 50 mg through an accelerated titration design. Phase 1/2a study to find dose and to evaluate the efficacy and safety of PRG-N-01 in patients with NF2-SWN are currently underway in South Korea.

The dysregulation of the TGF-β signaling pathway is believed to play a pivotal role in the development of colon cancer, affecting various TGF-β-mediated processes like growth inhibition, apoptosis, differentiation, and other TGF-β-controlled functions.^42,43 The most prevalent mechanism contributing to TGF-β signaling dysregulation in colon cancers involves the mutational inactivation of TβR2. This inactivation of TβR2 within colon cancer cells contributes to the malignant characteristics of the disease through multiple pathways, including Wnt-β-catenin, Hippo, and MAPK.⁴⁴ Previously, our group showed that reduction of TβR2 expression leads to the non-canonical activation of TβR1, thereby promoting cell proliferation through the activation of MAPK and Snail.^18,19 In this study, we provide evidence that PRG-N-01 could bind TβR1 kinase domain but not affect its kinase activity, thereby suppressing schwannoma formation in an NF2-SWN mouse model. Therefore, it is worth investigating whether the binding capacity of PRG-N-01 to TβR1 could be linked to pathology of the TβR2-inactivated colon cancer.

In conclusion, our study provides preliminary evidence that PRG-N-01, a first-in-class drug, may represent a promising treatment option for NF2-SWN. PRG-N-01 exhibits several unique features that differentiate it from other NF2-SWN therapies currently under investigation. We look forward to the upcoming clinical trials for healthy volunteers and patients with NF2-SWN-related tumors. While additional studies are needed to confirm its safety and efficacy in humans, the development of PRG-N-01 as a targeted therapy for NF2-SWN possess significant potential for improving the lives of NF2-SWN patients.

Supplementary material

Supplementary material is available online at Neuro-Oncology (https://dbpia.nl.go.kr/neuro-oncology).

Acknowledgments

We thank Dr. In Seok Moon at the Department of Otorhinolaryngology, Yonsei University College of Medicine, for providing vestibular schwannoma tissue of the NF2-SWN patient.

Conflict of interest statement

Yeon-Ho Chung, Bae-Hoon Kim, Yeongseon Ji, Tae-Gyun Woo, Songyoung Baek, Eunbyeol Shin, Yi Jin Song, Yuju Kim, Jin Han, and Bum-Joon Park are employees of PRG S&T Co., Ltd.

Funding

This work was supported by National Research Foundation of Korea (NRF) grants funded by the Korean government (MSIT) (RS-2024-00399681, RS-2024-00339289 and RS-2024-00442484).

Authorship Statement

BJP, BHK, and YHC conceptualized the study, designed the study methodology, and supervised the study. SYP, YHC, BHK, ML, JL, HRK, and JC wrote the original draft of the manuscript. ML and JC conducted the RNA-seq and data analysis. JL and HRK conducted prediction of protein complex and molecular docking. YHC, JYS, SP, TGW, SB, ES, YIS, YJH, YK and HJ performed the experiment. SYP and MK were responsible for the administration of the preclinical study.

Data Availability

The data generated in this study are available within the article and its supplementary data files. All other data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Author notes