Abstract Background Hyperactivated protein arginine methyltransferases (PRMTs) are implicated in human cancers. Inhibiting tumor intrinsic PRMT5 was reported to potentiate antitumor immune responses, highlighting the possibility of combining PRMT5 inhibitors (PRMT5i) with cancer immunotherapy. However, global suppression of PRMT5 activity impairs the effector functions of immune cells. Here, we sought to identify strategies to specifically inhibit PRMT5 activity in tumor tissues and develop effective PRMT5i-based immuno-oncology (IO) combinations for cancer treatment, particularly for methylthioadenosine phosphorylase (MTAP)-loss cancer. Methods Isogeneic tumor lines with and without MTAP loss were generated by CRISPR/Cas9 knockout. The effects of two PRMT5 inhibitors (GSK3326595 and MRTX1719) were evaluated in these isogenic tumor lines and T cells in vitro and in vivo. Transcriptomic and proteomic changes in tumors and T cells were characterized in response to PRMT5i treatment. Furthermore, the efficacy of MRTX1719 in combination with immune checkpoint blockade was assessed in two syngeneic murine models with MTAP-loss tumor. Results GSK3326595 significantly suppresses PRMT5 activity in tumors and T cells regardless of the MTAP status. However, MRTX1719, a methylthioadenosine-cooperative PRMT5 inhibitor, exhibits tumor-specific PRMT5 inhibition in MTAP-loss tumors with limited immunosuppressive effects. Mechanistically, transcriptomic and proteomic profiling analysis reveals that MRTX1719 successfully reduces the activation of the PI3K pathway, a well-documented immune-resistant pathway. It highlights the potential of MRTX1719 to overcome immune resistance in MTAP-loss tumors. In addition, MRTX1719 sensitizes MTAP-loss tumor cells to the killing of tumor-reactive T cells. Combining MRTX1719 and anti-PD-1 leads to superior antitumor activity in mice bearing MTAP-loss tumors. Conclusion Collectively, our results provide a strong rationale and mechanistic insights for the clinical development of MRTX1719-based IO combinations in MTAP-loss tumors. Keywords: Immunotherapy, Combination therapy __________________________________________________________________ WHAT IS ALREADY KNOWN ON THIS TOPIC. WHAT THIS STUDY ADDS * MTAP loss results in altered arginine methylation and increased tumor sensitivity to PRMT5 inhibition and T cell killing. * Unlike GSK3326595 (1st-gen PRMT5i), MRTX1719 (2nd-gen PRMT5i) treatment achieves tumor-specific PRMT5 inhibition in MTAP-loss tumors with limited immunosuppressive effects. * MRTX1719 synergizes with anti-PD-1 to control tumor growth in syngeneic murine MTAP-loss tumor models. * Mechanistically, MRTX1719 successfully reduces activation of the PI3K pathway, a well-documented tumor intrinsic pathway contributing to immune resistance. HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY * This study suggests that MTA-cooperative PRMT5 inhibitors, such as MRTX1719, represent a promising therapeutic strategy to improve effectiveness of immune checkpoint inhibitors in cancer patients with MTAP loss. Background MTAP encodes methylthioadenosine phosphorylase (MTAP), a key enzyme required for the methionine salvage pathway. Homozygous MTAP deletion is commonly observed in human cancers bearing focal copy number deletion of 9p21.[69]^1 At this tumor suppressor locus, MTAP is frequently co-deleted with CDKN2A and CDKN2B, resulting in a complete loss of MTAP function.[70]^1 In addition, MTAP-selective deletion and epigenetic silencing of the MTAP promoter are also reported to lead to MTAP loss in solid tumors, including mesothelioma, non-small cell lung cancer (NSCLC), gliomas, and pancreatic cancer.[71]^2 Emerging evidence suggests that MTAP loss can impose both tumor-intrinsic and tumor-extrinsic effects on regulating tumor development and sensitivity to cancer treatment.[72]^3 Intrinsically, MTAP loss leads to intracellular accumulation of its substrate, 5′-deoxy-5′-methylthioadenosine (MTA), in cancer cells. MTA is an S-adenosylmethionine (SAM) analog, which can inhibit many methyltransferase enzymes, including protein arginine methyltransferases (PRMTs). The MTA-related metabolic change increases the sensitivity of MTAP-loss tumors to PRMT5 knockdown.[73]4,[74]6 Extrinsically, MTA can leak out of MTAP-loss cancer cells, leading to changes in the tumor immune microenvironment (TIME). Increased MTA in TIME negatively impacts the host’s antitumor immune responses.[75]^7 8 These unique metabolic and immunological features of MTAP-loss tumors offer new opportunities to develop personalized treatments for patients with inactivated MTAP. Targeting PRMT5 was recently reported to promote antitumor immune responses by activating the STING pathways.[76]^9 As 9p21 loss was one of the genetic biomarkers for immune resistance,[77]^10 it became attractive to develop PRMT5 inhibitor (PRMT5i)-based immuno-oncology (IO) combinations to take advantage of PRMT5 vulnerability brought by MTA accumulation in patients bearing 9p21 loss. However, PRMT5 also plays an indispensable role in maintaining T cell function.[78]^11 Inhibiting PRMT5 in T cells reduces interleukin-2 (IL-2) production and nuclear factor of activated T cells (NFAT)-driven gene expression, and impairs Th1 and Th17 responses.[79]^12 13 The majority of the first-generation (1^st-gen) PRMT5is, such as GSK3326595, JNJ64619178, PF-0693999, and PRT543, are SAM-competitive/cooperative and/or substrate-competitive.[80]14,[81]16 The 1^st-gen PRMT5i could directly inhibit T cells’ activation and effector function. Early phase clinical studies with these agents demonstrated mechanism-based toxicities, thus halting their further clinical development. Lack of tumor selectivity and significant bone marrow toxicities restrict the utilization of the 1^st-gen PRMT5is for IO combinations. Emerging clinic challenges of the 1^st-gen compounds created the need for developing better PRMT5 targeting therapeutics. Such efforts recently culminated with the development of second-generation (2^nd-gen) PRMT5 inhibitors. These compounds block PRMT5 activity when it is bound to MTA, but not SAM.[82]^17 Thus, 2^nd-gen PRMT5is are exquisitely selective for cells with accumulated MTA, such as MTAP-loss tumors. So far, four companies, including Mirati (MRTX9768/1719),[83]^17 18 Tango (TNG908, TNG462),[84]^19 Amgen (AMG193/9747)[85]^20 21 and AstraZeneca (AZD3470),[86]^22 have developed such inhibitors. Encouraging clinical activities of MRTX1719 were observed in patients with MTAP-loss melanoma, gallbladder adenocarcinoma, mesothelioma, and lung cancer.[87]^17 Since MTA accumulation does not occur in normal cells including T cells, 2^nd-gen PRMT5is are expected to circumvent potential immunosuppressive effects when combined with cancer immunotherapy. However, the direct effects of 2^nd-gen PRMT5is on immune cells have not been verified, and the efficacy of 2^nd-gen PRMT5i-based IO combinations in MTAP-loss tumors remains largely unexplored. Here, we generated multiple isogenic pairs of MTAP-wild-type (WT) and MTAP-knockout (KO) murine tumor cell lines. Using these lines, we characterized the impacts of MTAP loss on tumor sensitivity to T cell killing, and to GSK3326595 (1^st-gen) and MRTX1719 (2^nd-gen) treatment. Additionally, we compared the effects of both PRMT5is on T cell function and biology. Our results demonstrate that MRTX1719 achieves tumor-specific PRMT5 inhibition in MTAP-loss tumors with modest immunosuppressive effects. Mechanistically, transcriptomic and proteomic profiles of PRMT5i-treated tumors reveal that MRTX1719 reduces activation of the PI3K pathway, a well-documented immune-resistant pathway,[88]^23 highlighting the potential of MRTX1719 to overcome immune resistance in MTAP-loss tumors. Furthermore, we showed that combining MRTX1719 and anti-PD-1 results in heightened antitumor activities in mice bearing MTAP-loss tumors as compared with mice with single agent treatments. Collectively, our results provide a strong rationale for the clinical development of MRTX1719-based IO combinations in MTAP-loss tumors. Materials and methods Cell lines and mice The murine melanoma B16 cell line (ATCC, CRL-6475) was obtained from ATCC and licensed accordingly. Murine colon carcinoma MC38/gp100 cell line and pmel-1 T cells were generated in our previous studies.[89]^24 All tumor cells were maintained in RPMI1640 (Gibco, 11875119) with 10% heat-inactivated fetal bovine serum (FBS; R&D Bio-Techne, s11150) and 100 µg/mL of normocin (InvivoGen, ant-nr-2). Platinum-E cells were used to generate retrovirus for genetic modification and were maintained in Dulbecco's Modified Eagle Medium (DMEM) (Gibco, 11965118) with 10% heat-inactivated FBS and normocin (100 µg/mL). All listed cell lines were cultured at 37°C, 5% CO[2]. The mycoplasma detection kit (SouthernBiotech, 13100-01) was used to routinely monitor mycoplasma contamination of cultured cells. The maximum length of time for the in vitro cell culture between thawing and use in the described experiments was 2 weeks. Pmel-1/Thy1.1 transgenic mice on a C57BL/6 background were from in-house breeds. C57BL/6 mice (6–8 weeks old) were purchased from the Charles River Frederick Research Model Facility. All mice were maintained in a specific pathogen-free barrier facility at University of Houston (UH). Genetic modifications of tumor cell lines Lipofectamine CRISPRMAX Cas9 Transfection Reagent (Invitrogen, CMAX00001) was used to generate CRISPR KO cell lines according to the manufacturer’s recommendation. Alt-R Streptococcus pyogenes Cas9 Nuclease V3 (1081058) and Alt-R CRISPR-Cas9 tracrRNA, ATTO 550 (1075934) and mouse Mtap-specific crRNA (rGrGrArCrArArUrArGrUrCrArCrArArUrUrGrArGrGrUrUrUrUrArGrArGrCrUrArUrGrC rU) were purchased from Integrated DNA Technologies (IDT). Briefly, to generate Mtap-KO lines, tumor cells were seeded in a 96-well plate at a density of 60% confluency 1 day before the transfection. CRISPR-Cas9 protein and Mtap-specific crRNA:tracrRNA duplex were diluted in Opti-MEM Reduced Serum Medium (Gibco, 31985062) and were delivered into cells by CRISPRMAX Transfection Reagent. The next day, transfected cells were re-seeded into a 96-well plate at a density of one cell/well to generate Mtap-KO single colonies. At least 10 Mtap-KO single colonies were selected to evaluate MTAP expression through western blot analysis. The colonies with validated MTAP loss were used as stable Mtap-KO lines for further studies. Retrovirus-based gene delivery was used to ectopically express gene-of-interests (GOIs). dsDNA fragments encoding 3×FLAG-tagged mouse open reading frames of Mtap ([90]NM_024433.2) were inserted into the retroviral vector, RVKM-IRES-EGFP as previously described.[91]^25 To generate retroviral supernatants, Platinum-E cells were seeded for 16 hours, then were transfected with the retroviral vectors with and without encoding GOIs, along with retrovirus packaging plasmids, pCMV-VSV-G (Addgene, 8454) by the jetPRIME transfection reagent (VWR, 101000046) according to the manufacturer’s protocol. Viral supernatants were collected 72 hours post-transfection and filtered through 0.45 μm polyvinylidene fluoride or polyvinylidene difluoride (PVDF) syringe filter units (Millipore-Sigma, SLHV033NK) to remove cell debris. Designated titers of retrovirus were used to infect cells in the presence of 8 µg/mL hexadimethrine bromide (Sigma-Aldrich, 107689). Quantification of MTA levels in tumor cells and conditioned culture medium 1.5×10^6 of genetically modified MC38/gp100 and B16 tumor cells were seeded in one well in a six well-plate with 1.2 mL medium per well. After 24 hours, supernatant and cell pellets were collected for MTA quantification using an AB SCIEX QTRAP 5500 mass spectrometer coupled to a Shimadzu UPLC. For MTA measurement, pellets from 3 to 8 millions of cells were added to 50 µL of water, 10 µL of cell suspension was extracted with 200 µL of ice-cold acetonitrile/methanol (80/20, v/v) and sonicated for 10 min in ice-cold water bath, cell suspensions were centrifuged 13,000 rpm for 20 min at 4°C to precipitate protein. The supernatant was collected, and 10 µL was injected into the LC column. Similarly, 10 µL of cell culture media was extracted with 200 µL of precooled acetonitrile/methanol, and sonicated and centrifuged, 10 µL of supernatant was injected into the LC column. The chromatographic separation of metabolites was achieved using Waters Acquity UPLC reversed-phase columns (Waters Acquity BEH C18, 2.1×50 mm, particle size of 1.7 mm) with a flow rate of 0.6 mL/min under gradient conditions. Solvent A was 0.1% formic acid in water, and solvent B was 0.1% formic acid in acetonitrile (LC/MS grade from Thermo Fisher Scientific). The gradient started at 20% B and increased to 50% B over 2 min, changing to 80% B over 0.5 min, and maintained at 80% B for 1 min, then decreased to 20% B over 0.5 min. The LC-MS/MS assay was performed in electrospray ionization positive ion mode using multiple-reaction monitoring for MTA detection and quantification. The ion transition, collision energy (CE), delustering potential (DP), entrance potential (EP), and collision cell exit potential (CXP) are optimized (MTA m/z: 298.2→136.1, CE=21, DP=80, EP=9, CXP=19). The instrument IonSpray Voltage was set at 4500 V and a source temperature of 500°C. MTA levels were determined by comparing against a standard curve prepared using external reference standards in a mobile phase and normalized to cell count or medium volume. In vitro effects of PRMT5 inhibitors on cell growth GSK3326595 (CT-GSK332) and MRTX1719 (CT-MRTX1719) were purchased from ChemieTek. Both compounds were dissolved in dimethylsulfoxide (DMSO) (Thermo Fisher Scientific, BP231-100) at 10 mM for stock, and diluted into indicated working concentrations to treat tumor cells and T cells. To determine the effect of PRMT5 inhibitors on the in vitro growth of tumors and T cells, equal numbers of cells were plated at the optimal seeding density, and were treated with either PRMT5i or vehicle control (DMSO) ranging from 0 to 10 μM for 3 days (T cell-related assay) or 5 days (tumor cell-related assays) at 37°C in 5% CO[2]. After treatment, cells were then cultured with CellTiter-Blue (CTB) (Promega, G8080) to determine live cell numbers according to the manufacturer’s instructions. Data were fitted with a four-parameter equation to calculate the half-maximal inhibitory concentration (IC50) on cell growth. At least three biological replicates were evaluated for each assay. Protein expression analysis Western blots were performed to determine the protein levels of MTAP, arginine methylation markers and phosphorylated AKT (pAKT) in tumor cells and T cells. An equal number of cells was collected and lysed in SDS containing Laemmli buffer (BIORAD, 1610747). Then, 10–30 µg of protein lysates was used for western blot analysis. For in vivo studies, tumor tissues were homogenized in lysis buffer using Precellys ceramic beads (Bertin Corp, 03961-1-003). Precellys 24 homogenizer (Bertin Corp, P002391-P24T0-A.0) was used to homogenize tissue samples at 6000 rpm under the program, including three times of the vertexing cycle using 30 s on and 45 s off. Detergent-compatible Bradford assay kits (Thermo Fisher Scientific, 23246) were used to determine protein concentration. Further, 10–20 µg of protein lysates was used for western blot analysis. Detailed information of antibodies used for western blot analysis is listed in [92]online supplemental table 1. Protein quantitation was performed using ImageJ 2 software (V.1.53). The area under each peak was calculated and normalized with β-actin expression. The ratios of normalized protein levels between the control groups and PRMT5i-treated groups were determined. In vitro T cell function assays The assays to evaluate T cell-mediated cytotoxicity were used as previously described.[93]^26 Briefly, gp100-expressing tumors were treated with PRMT5i for 3 days and labeled with CellTrace CFSE cell proliferation kits (Invitrogen, [94]C34554) according to the manufacturer’s protocol. Then, labeled tumor cells were co-cultured with pmel-1 T cells for 4 hours. The cell mixtures were followed by permeabilization with BD Cytofix/Cytoperm kit (BD Biosciences, 554714) for 20 min at room temperature and stained with an anti-cleaved caspase-3 monoclonal antibody ([95]online supplemental table 2). The percentage of tumor cells expressing cleaved caspase-3 was analyzed by LSRFortessa X20 (BD Biosciences). To induce the expression of major histocompatibility complex (MHC) I molecules, B16 tumors were treated with mouse interferon gamma (IFNγ) (200 ng/mL, BD Biosciences, 554587) at 37°C for 24 hours before the assay. Three biological replicates were included for all assays. In addition, enzyme-linked immunosorbent assay (ELISA) were used to evaluate the changes in cytokine production by T cells in response to PRMT5i treatment. Cultured pmel-1 T cells were treated with DMSO or PRMT5i at indicated concentrations for 3 days. Equal numbers of treated T cells were re-seeded into a 96-well plate and co-cultured with tumor cells at an E:T ratio=1 for 24 hours. The levels of IFNγ in conditioned media were determined by using the mouse IFNγ Quantikine ELISA Kit (R&D System, MIF00) according to the manufacturer’s instructions. Three biological replicates were included for all assays. Syngeneic mice studies To determine the antitumor activity of the combination of MRTX1719 with anti-PD-1, MC38/gp100 Mtap-KO or B16 Mtap-KO tumor cells were subcutaneously administrated in C57BL/6 mice 6–8 weeks old (1×10^6 cells/injection). Seven days after tumor inoculation, tumor-bearing mice were randomly assigned into four groups: control, MRTX1719 only, anti-PD-1 only, or combination treatment. MRTX1719 (30 mg/kg/day) was formulated in 1% methylcellulose (Sigma Aldrich, M0512) and was orally administrated once per day, whereas anti-mouse PD-1 (clone: 29F.1A12, BioxCell, BE0273) was intraperitoneally injected at a dose of 100 µg/per injection twice per week. Tumor size and health condition were monitored every 3 days. All experiments were performed in a double-blind manner. A portion of experimental mice were sacrificed at indicated time points to collect tissue samples for molecular and immune profiling. Molecular and immune profiling of tissue samples Spleen tissues from experimental mice were mechanically disrupted into single-cell suspensions, and Ammonium-Chloride-Potassium (ACK) lysis (Thermo Fisher Scientific, A1049201) was performed to remove red blood cells. Fresh tumor samples were either snap-frozen for molecular profiling or processed into single-cell suspensions for flow cytometry analysis. To prepare single-cell suspension, partial tumor samples were incubated in a triple-enzyme solution: collagenase type IV (Sigma Aldrich, C-5138), deoxyribonuclease type IV (Sigma Aldrich, D-5025), and hyaluronidase type V (Sigma Aldrich, H-6254) for 60 min at 37°C, and then were mechanically disrupted into single-cell suspensions. Single-cell suspensions were stained with a cocktail of antibodies targeting various cell surface and intracellular markers at 4°C for 30 min, according to the manufacturer’s instructions. For intracellular staining, cells were fixed and permeabilized either using a Foxp3 fixation/permeabilizing kit (Thermo Fisher Scientific, 00-5523-00) or a BD Cytofix/Cytoperm kit according to the manufacturer’s protocol. Stained spleen and tumor samples were analyzed using an LSRFortessa V.X-20. FACSDiva from BD Biosciences and Flowjo software (V.10.8.1) were used for data acquisition and analysis. Detailed information of antibodies used for immune profiling is listed in [96]online supplemental table 2. RNA sequencing and reverse-phase protein array Total RNA and protein were isolated from in vitro cell lines for RNA sequencing (RNA-seq) and reverse-phase protein array (RPPA), respectively. To isolate RNAs, tumor and T cells were directly lysed using Trizol (Life Technologies, 15596018). RNA-seq was used to determine mRNA expression levels. Quality check-up of raw reads of RNA-seq analysis was performed by FastQC (V.0.11.5) and summarized by MultiQC (V.1.7). FASTQ files were mapped to the mouse reference genome (GRCm39) using TOPHAT (V.2.0.10) and BOWTIE (V.2.1.0) with default parameter settings. Gene expression was quantified by transcripts per kilobase million (TPM) and transformed by log[2](TPM+1). Principal component analysis (PCA) was conducted on samples based on all genes to determine group clustering and identify outliers that are significantly separated from other biological replicates of the same experiment. Differentially expressed genes (DEGs) were identified based on the false discovery rate (FDR) and Log[2](fold change; FC) reported by DESeq2. Pathway enrichment analysis was performed using IPA (Ingenuity Pathway Analysis, V.90348151) to determine the functional profile of DEGs in canonical pathway sets.[97]^27 To obtain reasonable numbers of DEGs for pathway enrichment analysis, we used the criteria to define DEGs as follows: |log[2]FC|>0.5 and the FDR<0.05. All samples for RPPA were prepared based on the guidance provided by the RPPA core facility at the MD Anderson Cancer Center. Relative protein expression levels were normalized by bidirectionally median centering values by antibody and sample to eliminate loading differences. Statistical analyses Summary statistics (eg, mean, SEM) of the data are reported. Assessments of differences in continuous measurements between two groups were made using unpaired two-sample t-test posterior to data transformation (typically logarithmic, if necessary) or Wilcoxon rank-sum test. Tumor growth among various treatments was evaluated using two-way analysis of variance with repeat measurements. The Kaplan-Meier method and log-rank test were used to compare survival between groups. p values <0.05 were considered significant. GraphPad Prism (V.10.3.0) was used for plotting graphs and statistical analyses. Results MTAP expression determines protein arginine methylation status and tumor sensitivity to T cell-mediated cytotoxicity Intrigued by recent preclinical and clinical studies supporting the association between 9p21 loss and immune resistance,[98]^1 10 28 we sought to evaluate the molecular and immunological impacts of MTAP loss in cancer. To this end, we selected two commonly used murine cell lines expressing a defined tumor antigen for T cell killing, MC38/gp100 and B16, and knocked out Mtap using Mtap-specific gRNAs to establish isogenic lines. Stable Mtap-KO lines were further exogenously expressed with Mtap to generate Mtap-OE lines. The parental lines (WT) and Mtap-KO lines expressing empty vectors (Vec) served as controls for Mtap-KO lines and Mtap-OE lines, respectively. Complete loss of MTAP expression and MTA accumulation were observed in Mtap-KO and Vec lines, whereas restored MTAP expression led to the reduction of MTA levels in Mtap-OE lines ([99]figure 1A, [100]online supplemental figure 1). Since MTA levels are tied to PRMT5 activity, we characterized the levels of methylated arginine including asymmetric dimethylarginine (ADMA), symmetric dimethylarginine (SDMA), and monomethylarginine (MMA) in these lines. As expected, knocking out Mtap significantly suppressed the generation of SDMA in both MC38/gp100 and B16 tumor lines, and re-introducing Mtap in Mtap-KO lines restored the levels of SDMA ([101]figure 1B). However, levels of ADMA and MMA among these isogenic lines are largely comparable with their corresponding controls ([102]figure 1C,D). Given that PRMT5 is the major enzyme catalyzing SDMA modification,[103]^29 these results suggest that MTAP loss can result in PRMT5-selective inhibition, as previously reported.[104]^2 4 30 Since both MC38/gp100 and B16 murine lines express gp100, a tumor antigen that can be specifically recognized by pmel-1 T cells in an H2-D^b restricted manner, we next examined whether MTAP expression modulates tumor sensitivity to T cell killing. Interestingly, when we co-cultured isogeneic tumor lines with paired tumor-reactive T cells, we observed that Mtap-KO cells exhibited significantly increased sensitivity to T cell killing ([105]figure 1E), whereas the Mtap-OE lines demonstrated decreased tumor death induced by T cells ([106]figure 1F). These results reveal that MTAP loss leads to reduced PRMT5 activity and increased tumor sensitivity to T cell killing, highlighting the therapeutic potential to develop PRMT5i-based IO combinations in MTAP-loss cancers. Figure 1. MTAP expression modulates SDMA levels and tumor sensitivity to T cell killing. (A) Western blot analysis of MTAP expression in murine tumor cell lines with Mtap perturbation. MC38/gp100 and B16 tumors were genetically modified to knockout Mtap (Mtap-KO). Mtap-KO lines were further modified to exogenously express 3×FLAG-tagged Mtap (Mtap-OE). The parental cell lines (WT) and Mtap-KO transduced with the empty vector (Vec) were used as controls for Mtap-KO and Mtap-OE, respectively. (B–D) Western blot analysis of changes in arginine methylation levels in Mtap-KO and Mtap-OE tumor lines. The levels of SDMA modification (B), ADMA modification (C), and MMA modification (D) in two sets of murine tumor cell lines with Mtap perturbation. (E, F) Changes in tumor sensitivity of Mtap-KO (E) and Mtap-OE (F) tumor lines to T cell killing. Genetically modified MC38/gp100 and B16 tumors were co-cultured with tumor reactive pmel-1 T cells. Following overnight incubation, number of live tumor cells were counted to calculate percentage of T cell killing (n=3 per group). At least two independent experiments were performed, and representative results were selected for illustration. *p<0.05, **p<0.01. ADMA, asymmetric dimethylarginine; KO, knockout; MMA, monomethylarginine; MTAP, methylthioadenosine phosphorylase; SDMA, symmetric dimethylarginine; WT, wild type. [107]Figure 1 [108]Open in a new tab MRTX1719 displays superior tumor-selective PRMT5 inhibition in MTAP-loss tumors To better guide the design of PRMT5i-based IO combinations, we next focused on comparing the effects of distinct types of PRMT5 inhibitors on tumors and T cells. We compared GSK3326595 (1^st-gen PRMT5i) and MTRX1719 (2^nd-gen PRMT5i) for subsequent experiments, as both were in active clinical development at the time of this study. The in vitro effects of both PRMT5 inhibitors on cell growth, protein arginine methylation, and cellular function were determined. As shown in [109]figure 2A,B, GSK3326595 treatment resulted in comparable growth inhibition in WT and Mtap-KO tumors. However, the half-maximal inhibitory concentrations (IC50) on cell growth of MTRX1719 in Mtap-KO cell lines are significantly lower than those in corresponding WT lines ([110]figure 2A,B). In both MC38/gp100 and B16-derived lines, GSK3326595 and MRTX1719 treatment consistently showed better capability to reduce SDMA levels in Mtap-KO lines when compared with WT lines ([111]figure 2C,D). Notably minimal SDMA inhibition was observed in WT lines when treated with MRTX1719 ([112]figure 2C,D), suggesting the requirement of MTA accumulation for on-target effects of MRTX1719. On the other hand, the impacts of GSK3326595 and MRTX1719 on ADMA and MMA levels in both WT and Mtap-KO lines are negligible ([113]figure 2C,D). Figure 2. Effects of PRMT5 inhibition on in vitro proliferation and methylarginine levels of tumor cells. (A, B) Growth inhibition of GSK3326595 (GSK) and MRTX1719 (MRTX) on WT or Mtap-KO cells derived from the MC38/gp100 line (A) and the B16 line (B). Equal numbers of tumor cells were treated with indicated PRMT5i ranging from 40 nM to 10 μM for 5 days. The CellTiter-Blue assay was used to determine the changes in cell numbers in response to PRMT5i (n=3 per group). The unit for IC50s listed in the figure is μM. (C, D) Representative western blot analysis of arginine methylation from at least two independent experiments of MTAP-intact (WT) and MTAP-loss (Mtap-KO) tumors derived from MC38/gp100 (C) and B16 (D) treated with PRMT5i. The levels of SDMA modification, ADMA modification, and MMA modification in tumor cells receiving 5-day PRMT5i treatment were determined. GSK3326595 (GSK) and MRTX1719 (MRTX) at either the low concentration (0.1 μM, the left part of the triangle) or the high concentration (0.5 μM, the right part of the triangle) were used for treatment. Tumor cells treated with an equal amount of DMSO serve as controls (Ctrl). ADMA, asymmetric dimethylarginine; KO, knockout; MMA, monomethylarginine; MTAP, methylthioadenosine phosphorylase; PRMTs, protein arginine methyltransferases; SDMA, symmetric dimethylarginine; WT, wild type. [114]Figure 2 [115]Open in a new tab Next, we treated cultured CD8^+ T cells derived from pmel-1 mice with PRMT5is. The IC50s of GSK3326595 and MRTX1719 on in vitro T cell growth are 0.1 and 3.3 μM, respectively ([116]figure 3A). T cells treated with GSK3326595 at both high and low concentrations showed significantly reduced SDMA levels ([117]figure 3B). However, treatment of MRTX1719 only at a high concentration can achieve SDMA reduction in T cells ([118]figure 3B). Similar to tumor cells, no significant changes in ADMA and MMA levels were observed in T cells treated with either PRMT5i ([119]figure 3B). Furthermore, we also evaluated the changes in cytokine production and cytotoxicity of T cells in response to PRMT5i treatments. Gp100-reactive CD8^+ T cells were pretreated with PRMT5 inhibitors. Equal numbers of treated T cells were co-cultured with MC38/gp100 tumors. The IFNγ levels in the conditioned medium and the percentages of cleaved caspase-3 tumor cells were analyzed to evaluate cytokine production and cytotoxicity of PRMT5i-treated T cells, respectively. Our data demonstrated that GSK3326595, but not MRTX1719, can significantly impair cytokine production and cytotoxicity of CD8^+ T cells ([120]figure 3C,D). Figure 3. Effects of PRMT5 inhibitors on in vitro proliferation, methylarginine levels, and effector function of T cells. (A) Growth inhibition of GSK3326595 (GSK) and MRTX1719 (MRTX) on murine CD8^+ T cells. Cultured CD8^+ T cells derived from pmel-1 mice were treated with indicated PRMT5i ranging from 40 nM to 10 μM for 3 days. The CellTiter-Blue assay was used to determine the changes of T cell numbers in response to PRMT5i. The unit for IC50s listed in the figure is μM. (B) Representative western blot analysis of arginine methylation from at least two independent experiments of murine CD8^+ T cells treated with PRMT5 inhibitors. The levels of SDMA modification, ADMA modification and MMA modification in pmel-1 T cells receiving 3-day PRMT5i treatment were determined. GSK3326595 (GSK) and MRTX1719 (MRTX) at either the low concentration (0.1 μM, the left part of the triangle) or the high concentration (0.5 μM, the right part of the triangle) were used for treatment. T cells treated with an equal amount of DMSO serve as controls (Ctrl). (C, D) Changes in cytokine production (C) and cytotoxicity (D) of CD8^+ T cells treated with PRMT5i. Cultured CD8^+ T cells derived from pmel-1 mice were treated with indicated PRMT5i at either the low concentration (0.1 μM, the left part of the triangle) or the high concentration (0.5 μM, the right part of the triangle) for 3 days. Equal numbers of treated pmel-1 T cells were co-cultured with MC38/gp100 tumors. The IFNγ levels in a conditioned medium from overnight culture were determined by ELISA and used to evaluate cytokine production by T cells. The percentages of cleaved caspase-3 tumor cells after T cell co-culture were determined by flow cytometry and used to evaluate T cell cytotoxicity. Triplicate samples for each group were analyzed. *p<0.05, **p<0.01, ****p<0.0001. ADMA, asymmetric dimethylarginine; KO, knockout; MMA, monomethylarginine; MTAP, methylthioadenosine phosphorylase; PRMTs, protein arginine methyltransferases; SDMA, symmetric dimethylarginine; WT, wild type. [121]Figure 3 [122]Open in a new tab Next, we compared the in vivo on-target effects of PRMT5 inhibitors on tumor tissues (WT and Mtap-loss tumor) and peripheral lymphoid tissues (spleen). Consistent with our observations from in vitro experiments, significantly decreased SDMA levels were found in MTAP-loss tumor tissues from mice receiving either GSK3326595 or MRTX1719 ([123]figure 4A). WT tumor tissues and splenocytes from GSK3326595-treated mice, but not MRTX1719-treated mice, exhibited lower SDMA levels ([124]figure 4A). Compared with tissues from vehicle-treated mice, there were insignificant changes in ADMA and MMA levels in tumors bearing Mtap-WT or Mtap-KO, and corresponding spleen tissues from PRMT5i-treated mice ([125]figure 4B,C). These results suggest that MRTX1719 can efficiently inhibit PRMT5 function in MTAP-loss tumors and circumvent suppressive effects on normal cells including immune cells, whereas effects of GSK3326595 are independent of MTAP expression. Figure 4. In vivo effects of PRMT5 inhibitors on methylarginine levels in tumor and spleen tissues. C57BL/6 mice were subcutaneously inoculated with either WT or Mtap-KO tumors derived from MC38/gp100 (1×10^6 cells per injection). Tumor-bearing mice were orally treated with DMSO (Ctrl), GSK3326595 (GSK), or MRTX1719 (MRTX). The dose of PRMT5i is 30 mg/kg/day. Tumor tissues and spleen tissues were collected from mice receiving 7-day treatment. The levels of SDMA modification (A), ADMA modification (B), and MMA modification (C) were determined. Data from samples collected from two mice/group are shown. ADMA, asymmetric dimethylarginine; KO, knockout; MMA, monomethylarginine; SDMA, symmetric dimethylarginine; WT, wild type. [126]Figure 4 [127]Open in a new tab MRTX1719 and GSK3326595 have distinct molecular effects in tumors and T cells Next, we sought to use transcriptomic and proteomic analysis to compare the global effects of GSK3326595 and MRTX1719 treatment on regulating gene expression and subsequent pathway activity among tumors with or without NTAP-loss and T cells. Tumor cells and T cells were treated with PRMT5i at two different concentrations. RNA and protein samples harvested from treated cells were analyzed by RNA-seq and RPPA for transcriptomic and proteomic profiling, respectively. The unsupervised principal component analysis (PCA) showed that samples within each treatment group clustered together, confirming the reproducibility of RNA-seq results ([128]online supplemental figure 2). The changes in gene expression between PRMT5i-treated groups and their corresponding control groups were calculated. By using the cut-off at |log[2]FC|>0.5 and FDR<0.05, differentially expressed genes (DEGs) in response to PRMT5i treatment were identified in all three tested cell types. In both WT tumors and T cells, GSK3326595 treatment resulted in a higher number of DEGs (n=577/2170 in WT tumor/T cells) than MRTX1719 treatment (n=11/732 in WT tumor/T cells), indicating a stronger global change in gene expression induced by GSK3326595 in WT tumors and T cells ([129]figure 5A). Among all comparisons, the highest number of DEGs (n=3403) was observed in Mtap-KO tumors treated with MRTX1719 ([130]figure 5A). Furthermore, the number of DEGs in MRTX1719-treated T cells is less than half of the number in GSK3326595-treated T cells ([131]figure 5A). Remarkably, unlike GSK3326595, MRTX1719 dramatically alters gene expression in Mtap-KO cells but not in T cells ([132]online supplemental figure 3). These results not only support increased vulnerability to PRMT5i due to MTAP loss but also reveal increased selectivity of MRTX1719 in MTAP-loss tumors. In addition, we found that the changes in gene expression following MRTX1719 treatment in Mtap-KO cells highly correlate with those induced by GSK3326595 (1096 overlapping genes, 91% of DEGs in GSK3325595), and MRTX1719 displays more prominent impacts at the same dosage levels ([133]online supplemental figure 4A,B). Next, we performed the pathway enrichment analysis of DEGs identified in MRTX1719-treated cells. Due to the limited numbers of DEGs observed in WT tumors, the pathway enrichment analysis was only performed in Mtap-KO tumors and T cells. The canonical pathways enriched by DEGs (p<1×10^-5) in either Mtap-KO cells or T cells are illustrated in [134]figure 5B, and [135]online supplemental figure 4C. We found that the PI3K/AKT pathway is one of the pathways that are significantly enriched only in Mtap-KO cells, but not in T cells ([136]figure 5B). Tumor-selective inhibition of the PI3K/AKT pathway by PRMT5i in Mtap-KO cells was further confirmed by RPPA results, demonstrating reduced phosphorylation on AKT ([137]figure 5C). To validate these findings, we conducted western blot analysis on proteins isolated from tumor and immune cells treated with PRMT5i. Both in vitro and in vivo results confirmed that PRMT5i can reduce the activation of the PI3K/AKT pathway in MC38/gp100 Mtap-KO tumors, with MRTX1719 treatment displaying stronger suppressive effects on PI3K/AKT activation than GSK3326595 in vivo ([138]figure 5D,E, [139]online supplemental figure 5A,B). Furthermore, phosphorated AKT levels in splenocytes from PRMT5i-treated mice remained unchanged ([140]figure 5E, [141]online supplemental figure 5B). Figure 5. Characterization of molecular changes in PRMT5i-treated tumor cells and T cells. MC38/gp100 WT and Mtap-KO tumors and cultured CD8^+ T cells derived from pmel-1 mice were treated with 0.5 μM of GSK3326595 (GSK) or MRTX1719 (MRTX). The transcriptomic changes between DMSO-treated cells and PRMT5i-treated cells were determined. Differentially expressed genes (DEGs) in responses to PRMT5i treatment are defined as the genes with |log[2]FC|>0.5 and FDR<0.05. (A) Venn diagram showing the numbers of overlapped DEGs among WT tumors, Mtap-KO tumors, and T cells. (B) Bubble plot illustrating DEG-enriched pathways. DEGs identified from MRTX1719-treated Mtap-KO tumors and CD8^+ T cells were selected for IPA analysis. The top five significantly enriched pathways in Mtap-KO tumors and the top three significantly enriched pathways in T cells are listed. The size of each dot represents the number of overlapping DEGs within the corresponding pathway. The gray line indicates p value=0.01. (C) Heatmap showing proteomic changes in PRMT5i-treated tumor cells. MC38/gp100 WT and Mtap-KO tumors were treated with indicated PRMT5i at either the low concentration (0.1 μM, the left part of the triangle) or the high concentration (0.5 μM, the right part of the triangle). Tumor cells treated with DMSO serve as controls. The proteomic changes between DMSO-treated cells and PRMT5i-treated cells were determined by using mean-centered data. The proteins differentially expressed (|Log[2]FC|>1) in at least one comparison were selected for illustration. (D) In vitro effects of PRMT5i on AKT phosphorylation in MTAP-loss tumor cells. MC38/gp100 Mtap-KO tumors were treated with indicated PRMT5i at either the low concentration (0.1 μM, the left part of the triangle) or the high concentration (0.5 μM, the right part of the triangle). Tumor cells treated with DMSO serve as controls. (E) In vivo effects of PRMT5i on AKT phosphorylation in MTAP-loss tumors and splenocytes. Mice bearing MC38/gp100 Mtap-KO tumors were orally treated with DMSO (Ctrl), GSK3326595 (GSK), or MRTX1719 (MRTX). The dose of PRMT5i is 30 mg/kg/day. Tumor tissues and spleen tissues were collected from mice receiving 7-day treatment. Data from samples collected from two mice/group are shown. KO, knockout; WT, wild type. [142]Figure 5 [143]Open in a new tab MRTX1719 promotes T cell-mediated antitumor activity in mice bearing MTAP-loss tumors Previously, we reported that oncogenic activation of the PI3K pathway plays a critical role in the development of immune resistance, and tumor-specific PI3K inhibition by a PI3Kβ inhibitor improves the efficacy of immunotherapy in PTEN-loss tumors.[144]^23 31 As MRTX1719 treatment results in tumor-specific inhibition on the PI3K activation, we hypothesized that MRTX1719 could sensitize MTAP-loss tumors to immune attacks, particularly those mediated by T cells. To test this hypothesis, WT and Mtap-KO cells derived from MC38/gp100 and B16 tumors were pretreated with either DMSO or PRMT5i, followed by co-culture with tumor-reactive T cells. As shown in [145]figure 6A, both GSK3326595 and MRTX1719 consistently increased the percentage of apoptotic Mtap-KO tumor cells induced by tumor-reactive T cells. However, limited effects of PRMT5i on the sensitivity of WT tumors to T cell killing were observed ([146]figure 6A,B). Next, we evaluated the antitumor efficacy of MRTX1719 in combination with immune checkpoint blockade (ICB) in two preclinical tumor models with MTAP loss (MC38/gp100 and B16). Mice bearing Mtap-KO tumors were treated with vehicle, MRTX1719, anti-PD-1 antibody, or its combination 7 days after tumor inoculation. We found that the combination treatment achieved better antitumor efficacy when compared with the groups receiving single agent treatment ([147]figure 6C). Immune profiles of tumor and spleen tissues from treated mice are largely comparable with those from control mice ([148]figure 6D, [149]online supplemental table 3 and 4). However, when we used Ki67, a biomarker expressed by proliferative cells, to evaluate the proliferation of tumor-infiltrating immune cells in MC38/gp100 Mtap-KO tumors. Compared with the control treatment, both the anti-PD-1 treatment and the combination treatment significantly increased percentages of Ki67^+ cells in tumor-infiltrating CD8^+ cells. These results suggest that MRTX1719 has limited negative impacts on the effect of anti-PD-1 on promoting CD8^+ T proliferation ([150]figure 6D). Reduced pAKT levels in tumors were also observed in mice receiving either MRTX1719 treatment or combination treatment ([151]figure 6E, [152]online supplemental figure 5C). Taken together, our in vitro and in vivo results suggest that MRTX1719 synergizes with T cell-mediated immunotherapy, particularly anti-PD-1, by enhancing tumor sensitivity to immune attacks. Figure 6. MRTX1719 enhances T cell-mediated immune responses in MTAP-loss tumors. (A, B) In vitro effects of PRMT5i on tumor sensitivity to T cell-mediated killing. WT and Mtap-KO cells derived from the MC38/gp100 line (A) and the B16 line (B) were pretreated by either 0.1 μM or 0.5 μM of PRMT5i or DMSO (Ctrl). Treated tumor cells were co-cultured with tumor-reactive T cells. Tumor sensitivity to T cell killing was evaluated by the percentage of cleaved caspase-3 of tumor cells after 4-hour coculture. (C) Growth curves and Kaplan-Meier survival curves of tumor-bearing mice receiving MRTX1719 and/or anti-PD-1. MC38/gp100 Mtap-KO cells or B16 Mtap-KO cells were inoculated into C57BL/6 mice. 7 days after tumor inoculation, mice were randomly treated with MRTX1719 (30 mpk), anti-PD-1 (100 µg), or a combination of both. The endpoints include tumors reaching 1.5 cm in diameter, tumor ulceration with bleeding, or tumor ulceration exceeding 2 mm in diameter. Representative results from two independent experiments were shown (n=5–7 per group). (D) Abundancy and proliferation of T cells in Mtap-KO tumor tissues from mice receiving MRTX1719 and/or anti-PD-1. Tumor tissues were collected either on day 7 or day 14 after treatments for flow cytometry analysis (n=5 per group). (E) Levels of AKT phosphorylation in Mtap-KO tumor tissues from mice receiving MRTX1719 and/or anti-PD-1. Tumor tissues were collected from two mice/group 7 days after treatments. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. KO, knockout; WT, wild type. [153]Figure 6 [154]Open in a new tab Discussion Cancer immunotherapy has changed the treatment landscape and improved outcomes for advanced cancer treatment.[155]32,[156]34 However, dysfunction in tumor intrinsic pathways enables tumor cells to meet the need for hyperproliferation, metastasis and immune escape. Our previous results have demonstrated that targeting oncogenic pathways and/or metabolic pathways could lead to new therapeutic strategies to improve cancer immunotherapy.[157]^23 31 35 Here, we explored the immunological impacts of targeting PRMT5, one of the epigenetic enzymes, and evaluated the therapeutic potential of combining PRMT5 inhibitors with cancer immunotherapy. As the major type II arginine methyltransferase, PRMT5 was first identified as a transcriptional repressor by modifying histones H3 and H4. In the context of cancer, PRMT5 was reported to repress the expression of tumor suppressor genes and modulate pathways critical for tumor development, such as the epithelial-mesenchymal transition pathway and growth factor receptor signaling pathways.[158]^36 Furthermore, upregulated PRMT5 has been implicated in most cancer types, and negative associations between PRMT5 expression and patient survival were reported in multiple types of cancer.[159]^37 With these results, PRMT5 is becoming a popular therapeutic target for cancer treatment. Recently, inhibition of PRMT5 was reported to increase expression levels of several interferon response genes (ISGs), including Ifnb1, Ccl5 and Cxcl10 in B16 tumors following dsDNA stimulation by regulating the activation of the cGAS/STING pathway.[160]^9 However, our results demonstrate that ISGs are moderately downregulated in both T cells and Mtap-KO tumor cells after MRTX1719 treatment ([161]online supplemental figure 4C). These data suggest that the overall impacts of PRMT5 inhibition on the activation of the IFN pathway could be multifaceted and dependent on cell types and stimulation types. Future studies are warranted to dissect regulatory mechanisms of PRMT5 on the IFN pathway in distinct types of tumor cells, particularly in 9p21-loss tumors which could harbor deletion of IFN genes. Given that activation of the IFN pathways is critical for T cell-mediated antitumor activity, it is unlikely that the synergistic effect of MRTX1719 and anti-PD-1 observed in our studies depends on the role of PRMT5 in modulating the IFN pathways. Additionally, via the KEAP1-NRF2 axis, PRMT5 regulates the ferroptosis pathway in cancer cells. In triple-negative breast cancer (TNBC) lines, PRMT5 promotes ferroptosis resistance, and inhibiting PRMT5 potentiates the antitumor activity of cancer immunotherapy.[162]^38 However, the effect of PRMT5 on regulating ferroptosis was not observed in non-TNBC lines,[163]^38 further confirming that regulatory roles of PRMT5 could be cell-type dependent. In preclinical lung cancer models, inactivation of PRMT5 was reported to increase the expression of PD-L1, a key immune inhibitory molecule.[164]^39 PRMT5 is indispensable for T cell activation, proliferation, and differentiation.[165]^40 These results imply a double-edged effect of PRMT5 inhibition in determining T cell-mediated antitumor immune responses and point out the challenges of combining PRMT5 inhibitors with cancer immunotherapy. A search of published and patent literature on PRMT5 inhibitors reveals at least five types of PRMT5 inhibitors at different developmental stages based on their mechanisms of actions (MOAs). These MOAs include (1) SAM-competitive, (2) SAM-cooperative and substrate-competitive, (3) SAM and substrate- competitive, (4) proteolysis targeting chimera (PROTAC)-based, and (5) MTA-cooperative and SAM-competitive ones.[166]^14 16 41 42 Most of the early-developed PRMT5 inhibitors (1^st-gen) are either SAM-cooperative or SAM-competitive. Among them, several inhibitors including GSK3326595 were evaluated in early phase clinical studies. Partial responses were observed in a portion of patients with cervical cancer and adenoid cystic carcinoma enrolled in these phase I/II trials.[167]^16 43 The initial results showed that dose-limiting toxicities (DLTs) of 1^st-gen PRMT5i are largely manageable.[168]^16 43 However, thrombocytopenia, anemia and neutropenia were reported as common DLTs, alluding to the challenges of combining 1^st-gen PRMT5i with immunotherapy which is mediated by cells exclusively derived from hematopoietic stem cells. With the unique metabolic feature associated with MTA accumulation in MTAP-loss tumors, the ideal PRMT5 inhibitors for MTAP-loss tumors would be the 2^nd-gen or MTA-cooperative PRMT5 inhibitors. Such 2^nd-gen inhibitors have been recently developed by multiple pharmaceutical companies.[169]^17 19 44 The early clinical data from phase I studies with MRTX1719 as monotherapy showed multiple partial responses in patients with MTAP-loss including epithelioid malignant mesothelioma, adenocarcinoma, melanoma, and NSCLC.[170]^17 As MTA accumulation is minimal in MTAP-intact cells including T cells, we expect that 2^nd-gen PRMT5i could overcome the direct on-target suppressive effects on T cells. DLTs on hematopoietic cells in MTRX1719-treated patients have not been observed at any evaluated dose level (up to 800 mg, once per day), supporting this expectation.[171]^17 Although genetic inhibition of PRMT5 displays synthetic lethality with MTAP-loss in tumors,[172]^4 45 there are limited results to directly compare IC50s of 1^st-gen PRMT5i on cell viability using isogenic pairs of tumors with and without MTAP loss. Recent results from a 10-day viability assay demonstrate that IC50s of GSK3326595 in HCT116 with and without MTAP deletion are comparable (200 and 164 nM, respectively).[173]^17 These results are consistent with the data we reported here. However, results from our group and others demonstrate that PRMT5 inhibition can induce more dramatic reductions of SDMA levels and more transcriptional changes in MTAP-loss tumors, when compared with WT tumors. These results confirmed that MTAP-loss tumors are more sensitive to PRMT5 inhibitors. More importantly, our data suggest that a significantly higher concentration of second-gen PRMT5i is required to achieve on-target effects of on MTAP-intact tumor cells and T cells when compared with MTAP-loss tumors. Furthermore, we showed that MRTX1719 treatment sensitizes MTAP-loss tumors to in vitro T cell killing and in vivo anti-PD-1 treatment. Multi-omic analyses of MTAP-loss tumors and T cells in response to MRTX1719 showed reduced phosphorylation of AKT in MTAP-loss tumors and not in T cells. These findings are consistent with a recent study which identified AKT as one of the PRMT5 substrates, showing that PRMT5 promotes AKT activation by methylating AKT1 at arginine 391 (R391).[174]^46 Given that phosphorylation of AKT is an important trigger for activation of the PI3K pathway, an oncogenic pathway with immunosuppressive roles, our data provide a strong rationale for combining MRTX1719 with T cell-mediated immunotherapy, such as anti-PD-1 in MTAP-loss patients. MTAP loss commonly occurs in patients with cancer with 9p21 deletion.[175]^1 10 47 Immune resistance associated with deletions of this loci has been implicated in head and neck squamous cell carcinoma, TNBC, high-grade serous ovarian carcinoma, melanoma, renal cell carcinoma and other rare solid tumors.[176]^1 10 47 48 Most of these studies showed that immune resistance is associated with co-deletion of CDKN2A/B and MTAP, but not with CDKN2A/B deletion, implying the contribution of MTAP to immune resistance. However, knocking out MTAP in two murine tumor cells promotes tumor apoptosis induced by tumor-reactive T cells, and restoring MTAP expression in Mtap-KO cells reduces their sensitivity to T cell killing. These effects could be explained by MTA being a natural PRMT5 inhibitor. Thus, the intrinsic effect of MTAP loss might not directly drive the immune-resistant phenotype observed in patients with 9p21 deletion. This observation is further supported by a recent paper showing that the immune evasion associated with 9p21 deletion is mediated by deletion of the IFN gene cluster.[177]^49 Meanwhile, MTAP loss in cancers with 9p21 deletion creates a unique vulnerability to 2^nd-gen PRMT5 inhibitors, supporting the potential of 2^nd-gen PRMT5 inhibitors for overcoming immune resistance in this type of cancer. As the models used in this report are limited to perturbation of the MTAP gene, they might not entirely mirror patients with 9p21 deletion, in which the codeletion of MTAP, CDKN2A/B and IFN-related genes commonly occurs. We are actively developing murine tumor models bearing deletions of 4C4, the murine ortholog to human 9p21. These models will be valuable to validate the antitumor activity of 2^nd-gen PRMT5i-based IO combinations in 9p21 cancers. Besides the intrinsic effect of MTAP loss, MTAP-loss tumors could impose extrinsic effects by altering the MTA levels in the tumor microenvironment. We observed increased MTA levels in the culture medium in Mtap-KO cells. Enhanced MTA levels in the tumor microenvironment have been reported to impair the effector function of T cells.[178]^7 8 Due to the limited duration of co-culture of tumor and T cells in reported cytotoxicity assays, the impacts of MTA accumulation on T cell cytotoxicity might be underappreciated. Enzyme-mediated depletion of MTA by PEG-MTAP was reported to restore T cell function in MTAP-loss cancer.[179]^28 Via complementing of 2^nd-gen PRMT5i on intrinsic effects associated with MTAP loss, it is worthwhile to evaluate whether the administration of PEG-MTAP can further improve the efficacy of MRTX1719-based IO combinations. In conclusion, our studies demonstrated that MTAP loss results in a series of tumor intrinsic changes, including increased MTA level, reduced SDMA level and enhanced sensitivity to T cell killing. We further showed that MRTX1719 achieves tumor-specific PRMT5 inhibition by leveraging high MTA environments created by MTAP loss. A synergistic effect of MRTX1719 and anti-PD-1 was observed in preclinical MTAP-loss tumor models, supporting the application of 2^nd-gen PRMT5i-based IO combinations to overcome immune resistance in MTAP-loss tumors. supplementary material online supplemental file 1 [180]jitc-12-9-s001.pdf^ (8.1MB, pdf) DOI: 10.1136/jitc-2024-009600 Footnotes Funding: This work was supported in part by the Department of Defense (TC, WP&HX, ME220215P1), by American Cancer Society (WP, RSG-23-1155993-01-MM), by the following National Institutes of Health grants: R35GM153387 (MTB), R01CA240700 (ES) and P50CA221703 (MDACC Melanoma SPORE), by Melanoma Research Alliance Young Investigator Award (WP, 558998); and by Cancer Prevention and Research Institute of Texas (WP, RP200520). HX is a CPRIT scholar of cancer research. Provenance and peer review: Not commissioned; externally peer reviewed. Patient consent for publication: Not applicable. Ethics approval: All animal studies followed the protocols approved by The Institutional Animal Care and Use Committee at UH. Contributor Information Si Chen, Email: schen63@cougarnet.uh.edu. Jiakai Hou, Email: jhou21@central.uh.edu. Roshni Jaffery, Email: rtjaffer@cougarnet.uh.edu. Ashley Guerrero, Email: amguerr9@cougarnet.uh.edu. Rongjie Fu, Email: rfu1@mdanderson.org. Leilei Shi, Email: lshi1@mdanderson.org. Ningbo Zheng, Email: nzheng4@central.uh.edu. Ritu Bohat, Email: rbohat@mdanderson.org. Nicholas A Egan, Email: Naegan92@gmail.com. Chengtai Yu, Email: cyu6@mdanderson.org. Sana Sharif, Email: ssharif6@cougarnet.uh.edu. Yue Lu, Email: ylu4@mdanderson.org. Wei He, Email: whe3@mdanderson.org. Shuyue Wang, Email: swang21@mdanderson.org. Donjeta Gjuka, Email: donjetagjuka@gmail.com. Everett M Stone, Email: stonesci@utexas.edu. Pooja Anil Shah, Email: PShah3@mdanderson.org. Jordi Rodon Ahnert, Email: JRodon@mdanderson.org. Taiping Chen, Email: tchen2@mdanderson.org. Xinli Liu, Email: xliu65@central.uh.edu. Mark T Bedford, Email: mtbedford@mdanderson.org. Han Xu, Email: hxu4@mdanderson.org. Weiyi Peng, Email: wpeng2@central.uh.edu. Data availability statement Data are available upon reasonable request. References