Anti-PD-1 (clone RMP1–14) and rat IgG2a isotype control antibody (clone 2A3) were purchased from BioXcell. APG-115 (Ascentage Pharma) was dissolved in DMSO (Sigma) to make a stock solution for in vitro use. MC38 cell line derived from a C57BL/6 murine colon adenocarcinoma and MH-22A cell line derived from C3H murine liver cancer were obtained from Sun Yat-Sen University Cancer Center (Guangzhou, China) and European Collection of Authenticated Cell Cultures, respectively. All cell lines were genetically authenticated and free of microbial contamination.

Six- to eight-week old female mice were obtained from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Mice were implanted subcutaneously with MC38 (0.5 × 10⁶, C57BL/6), MH-22A (5 × 10⁶, C3H), or Trp53 ^−/− MH-22A (5 × 10⁶, C3H) cells in 0.1 mL PBS per animal to establish syngeneic tumor models. When the average tumor size reached 50–100 mm³, tumor-bearing mice were randomly assigned into groups based on their tumor volumes. Trp53 ^−/− knockout C57BL/6 J mice were purchased from Biocytogen (Beijing, China).

APG-115 was formulated in a vehicle of 0.2% HPMC (Sigma Aldrich) and administered orally at 10 or 50 mg/kg daily or every other day (Q2D). Anti-PD-1 antibody was diluted in PBS and dosed intraperitoneally at 5 or 10 mg/kg twice a week (BIW). Vehicle plus isotype control antibody, or vehicle only were used as the control. Tumor volume (V) was expressed in mm³ using the following formula: V = 0.5 a × b²; where a and b were the long and short diameters of the tumor, respectively. As a measurement of efficacy, a T/C (%) value was calculated at a time point according to: T/C (%) = (T_RTV/C_RTV) × 100%; where T_RTV was the relative tumor volume (RTV) of the treatment group and C_RTV was the RTV of the control group. RTV = V_t/V₁; where V₁ and V_t were the average tumor volumes on the first day of treatment (day 1) and the average tumor volumes on a certain time point (day t), respectively. Additional measurements of response included stable disease (SD), partial tumor regression (PR), and complete regression (CR) were determined by comparing tumor volume change at day t to its baseline: tumor volume change (%) = (V_t-V₁/V₁). The BestResponse was the minimum value of tumor volume change (%) for t ≥ 10. For each time point t, the average of tumor volume change from t = 1 to t was also calculated. BestAvgResponse was defined as the minimum value of this average for t ≥ 10. The criteria for response (mRECIST) were adapted from RECIST criteria [32, 33] and defined as follows: mCR, BestResponse < − 95% and BestAvgResponse < − 40%; mPR, BestResponse < − 50% and BestAvgResponse < − 20%; mSD, BestResponse < 35% and BestAvgResponse < 30%; mPD, not otherwise categorized. SD, PR, and CR were considered responders and used to calculate response rate (%). Body weight of animals were monitored simultaneously. The change in body weight was calculated based on the animal weight of the first day of dosing (day 1). Tumor volume and changes in body weight (%) were represented as the mean ± standard error of the mean (SEM).

In re-challenge studies, naïve mice and CR mice were inoculated subcutaneously with 5 × 10⁶ MH-22A tumor cells per animal. Tumor growth was monitored for 3 weeks without further treatment.

Animal studies were conducted in the animal facility of GenePharma (Suzhou, China). The protocols and experimental procedures involving the care and use of animals were approved by the GenePharma Institutional Animal Care and Use Committee.

For analysis of TILs in the TME, isolated tumors were weighed and dissociated by gentle MACS buffer (Miltenyi) and then filtered through 70 μm cell strainers to generate single-cell suspensions. After counting viable cells, the samples were incubated with a live-dead antibody followed by FcγIII/IIR-blocking staining. Cells were then stained with fluorochrome-labeled antibodies against CD45 (Thermo Fisher Scientific, catalog #69–0451-82), CD4 (BD Biosciences, catalog #552775), CD8 (Thermo Fisher, catalog #45–0081-82), CD3 (Thermo Fisher, catalog #11–0032-82), CD49b (Thermo Fisher, catalog #48–5971-82), CD11b (Thermo Fisher, catalog #48–0112-82), F4/80 (Thermo Fisher, catalog #17–4801-82), CD206 (Thermo Fisher, catalog #12–2061-82), MHC-II (Thermo Fisher, catalog #11–5321-82), Gr-1 (Biolegend, catalog #108439), CD25 (BD, catalog #564370), NK1.1 (eBioscience, catalog #48–5941-82), and Foxp3 (Thermo Fisher, catalog #12–5773-82).

For analysis of cytokines in TILs, single cell suspensions generated from isolated tumors were plated into six-well plates and stimulated with PMA (50 ng/mL) and ionomycin (500 ng/mL) for 4 hours. Two hours before the end of stimulation, protein transport inhibitor monensin (2 μM) was added. Cells were collected and incubated with a live-dead antibody followed by FcγIII/IIR-blocking staining. Cells were then stained with fluorochrome-labeled antibodies against CD45, CD4, CD8, CD3, IFN-γ (BD, catalog #554413) and TNF-α (BD, catalog #554419).

For PD-L1 expression, MH-22A cells were treated with indicated concentrations of APG-115 for 72 h. PD-L1 (BD, catlog #563369) expression and its fluorescence intensity were acquired on an Attune NxT flow cytometer (Life Technology) and analyzed using Flowjo software (BD).

Bone marrow cells were collected from two femurs of each mouse and plated in complete RPMI-1640 medium supplemented with 10% FBS, 100 ng/mL m-CSF (R&D, catlog # 416-ML-050), and 1% penicillin and streptomycin (Invitrogen). After 7 days, cells were harvested and evaluated by flow cytometry for expression of CD11b and F4/80. BMDMs were further treated with IL-4 (20 ng/mL, R&D) to induce alternative (M2) macrophage polarization with or without APG-115 (250 nM or 1 μM). Cells were then harvested and evaluated for the expression of M2 markers (MHC-II and CD206) by flow cytometry, expression of M2-related genes (Arg-1 and Retnla) by RT-qPCR, and p53, p21, c-Myc and c-Maf protein levels by Western blotting.

After treatment, BMDMs were harvested and mRNA was extracted using a RNAEasy mini plus kit (Qiagen). cDNA was retro-transcribed from 1 μg of RNA primed with random hexamers using cDNA reverse transcription kit (Takara), 3 ng of equivalent cDNA were amplified in a qPCR (SyBr) assay on an ABI7500 (Thermo Fisher) for the following genes: Arg-1 (SyBr green PCR using primers 5′- CATTGGCTTGCGAGACGTAGAC and 5′- GCTGAAGGTCTCTTCCATCACC) and Retnla (SyBr green PCR using primers 5′- CAAGGAACTTCTTGCCAATCCAG and 5′- CCAAGATCCACAGGCAAAGCCA). Relative gene expression was quantified with the 2-delta method, normalized to GAPDH housekeeping gene detected by SyBr green RT-PCR (5′-AACTTTGGCATTGTGGAAGG and 5′- GGATGCAGGGATGATGTTCT).

Cells were collected and lysed in RIPA lysis buffer (Yeasen, catalog #20101ES60) containing a protease inhibitor cocktail (Yeasen, catalog #20104ES08). Protein concentration was quantified by bicinchoninic acid assay (Thermal Fisher). Equal amounts of soluble protein were loaded and separated on a 10% SDS-PAGE, followed by transfer to nitrocellulose and then immunoblotting using primary antibodies, including p53 (CST, catalog #32532), p21 (abcam, catalog #ab109199), c-Myc (CST, catalog #13987 T), c-Maf (abcam, catalog #ab77071), p-STAT3 (CST, catalog #9145), t-STAT3 (CST, catalog #9139), PD-L1 (R&D, catalog #AF1019), Caspase 3 (CST, catalog #9665S), ZAP70 (CST, catalog #3165S), MDM2 (BD, catalog #556353), and β-actin (CST, catalog #3700S). HRP-conjugated secondary antibodies (Yeasen, catalog #33101ES60, catalog #33201ES60) were used at 1:5000 dilution.

CD4⁺ T cells were positively selected from mouse spleens using magnetic beads (Miltenyi, catalog #130–049-201) and stimulated with 10 μg/mL plate-bound anti-CD3 (eBioscience, catalog #16–0031-85) and 2 μg/mL anti-CD28 (eBioscience, catalog #16–0281-85) in the presence of 250 nM APG-115 or DMSO for 1 or 2 days. After treatment, cells were harvested and evaluated by flow cytometry for the expression of CD25 (BD, catalog #557192), CD62L (BD, catalog #553151), and Foxp3 (Thermo Fisher, catalog #12–5773-82). T cell activation were defined as CD25^highCD62L^low and enlarged cell size. CD25⁺Foxp3⁺T cells represented Treg population.

For T cell proliferation, CD4⁺ T and CD8⁺ T cells were positively selected from mouse spleens using magnetic beads (Miltenyi, catalog #130–049-201 and #130–096-495) and then stimulated with a series of concentrations of plate-bound anti-CD3 and 2 μg/mL anti-CD28 in the presence of 250 nM APG-115 or DMSO. After 72 h, relative cell numbers were determined using CellTiter-Glo luminescent cell viability assay (Promega, catalog #G7571) and normalized to unstimulated cultures treated with DMSO control.

OT-I splenocytes were stimulated with 2 μg/mL OVA peptide (SIINFEKL, GL Biochem, catalog #53698) and 10 ng/mL rmIL-2 (R&D, catalog #402-ML-500) for 72 h in the complete RPMI-1640 medium supplied with vehicle, 50 nM, 250 nM, or 1 μM APG-115. Cells were harvested after the treatment. EL4 cells (target cells, T) were labeled with 50 nM CellTrace Far-Red dye (Invitrogen, catalog #C34564) and then pulsed with 20 μg/mL OVA peptide for 30 min at 37 °C in the complete RPMI-1640 medium. Labeled EL4 cells (2 × 10⁴) were seeded in each well of a 96-well plate. OT-I CD8⁺ T cells (effector cells, E) treated with four different conditions were seeded with targeted EL4 cells in an E:T ratio of 0:0, 0.5:1, 2:1, or 8:1. The effector and target cells were co-cultured overnight. PI dye was added to the mixed cell solutions at 1:10000 and incubated for 10 min. Percentage of target cell lysis were analyzed using FACS LSRFortessa (BD).

One-way ANOVA followed by Bonferroni’s posttest was applied to assess the statistical significance of differences between multiple treatment groups. All data were analyzed in SPSS version 18.0 (IBM, Armonk, NY, U.S.A.). Prism version 6 (GraphPad Software Inc., San Diego, CA, U.S.A.) was used for graphic presentation.

Considering the essential role of p53 in M1 function and M2 polarization [28–30], we first explored how APG-115-mediated p53 activation affected M1 and M2 macrophages. Briefly, BMDMs were generated and confirmed by expression of CD11b⁺F4/80^hi using a flow cytometer (Fig. 1a, left panel). After stimulated under a M2-polarizing condition in the presence of 20 ng/mL IL-4 for 24 h, a substantial population (30.6%) of CD206⁺MHC-II⁻ M2 macrophages were induced (Fig. 1a, middle panel). Concurrent treatments with 250 nM or 1 μM APG-115 (IL-4 + APG-115) inhibited M2 polarization, resulting in only 11 and 12% M2 macrophages, respectively (Fig. 1a, right panels). RT-PCR analysis showed that mRNA expression of M2-associated genes (i.e., Arg-1 and Retnla) were substantially upregulated after IL-4 treatment for 48 h (Fig. 1b). Under concurrent treatment with APG-115, IL-4-induced mRNA expression of M2 associated genes was significant suppressed. These results demonstrate that APG-115 suppresses M2 macrophage polarization in vitro.Fig1Fig. 1

APG-115 suppresses alternative M2 macrophages polarization in vitro and increases M1 macrophages in vivo through activation of p53 pathway. a BMDMs were generated under the treatment with m-CSF for 7 days and then treated with IL-4 (20 ng/mL) to induce alternative macrophage polarization (M2) for 24 h in the absence or presence of APG-115. Cells were then harvested for detection of M2 macrophages (CD206⁺MHC-II^low) by flow cytometry. b the mRNA expression levels of Arg-1 and Retnla in the above BMDMs induced by the treatment with IL-4 (20 ng/mL) with or without APG-115 were analyzed by RT-qPCR. Duplicated samples were tested. c Western blot analysis of p53, p21, c-Myc and c-Maf total proteins in BMDMs treated with IL-4 (20 ng/mL) with or without APG-115 (1 μM) for 0, 4, or 24 h, or sequentially treated with IL-4 and then APG-115 for 24 h each (24 h + 24 h). d Quantification of C. 0 (black bars), 4 (blue bars), or 24 h (green bars), or sequentially treated with each agent for 24 h (24 h + 24 h, red bars). e naïve BALB/c mice were treated with APG-115 (10 mg/kg, Q2D × 2 doses; n = 5). Two days after the last dose, spleens were collected, dissociated into single-cell suspensions, and stained with macrophage markers for flow cytometry analysis. Macrophages were defined as CD11b⁺ F4/80^hi, and further analyzed for M1 macrophages by expression of MHC-II. Pooled data of percentages of macrophages gated on CD45⁺CD3⁻ live cells (f) and percentages of M1 macrophages gated on macrophages (g) from five mice were plotted

Effector T cells play a critical role in antitumor immunity. Consequently, the effect of MDM2 inhibitor on T cells could impact antitumor immune responses occurring in the context of MDM2 inhibitor-mediated tumor cell death. To investigate how MDM2 inhibitor influences T cells, we exposed CD4⁺ T cells and CD8⁺ T cells isolated from mouse spleens to APG-115 or DMSO control for 72 h. The results showed that APG-115 had a significant effect on T cells, leading to a substantial increase in T cell numbers after 72 h (P < 0.05 for 5 and 10 μg/mL, Fig. 2a). This effect was dependent on sufficient stimulation and was not observed under unstimulated or weakly stimulated conditions.Fig2Fig. 2

APG-115 increase mouse T cell proliferation and enhances mouse CD4⁺ T cell activation. a CD4⁺ T and CD8⁺ T cells were positively selected from mouse spleens using magnetic beads and then stimulated with indicated concentrations of plate-bound anti-CD3 and 2 μg/mL anti-CD28 in the presence of 250 nM APG-115 or DMSO. After 72 h, relative cell numbers were determined using CellTiter-Glo luminescent cell viability assay (Promega) and normalized to unstimulated cultures treated with DMSO control. * P < 0.05. b immunoblots for the expression of caspase 3, cleaved caspase 3, and Zap-70 (loading control) in total cell lysates of anti-CD3/CD28-stimulated CD4⁺ T cells exposed to APG-115 or solvent control DMSO for 3, 6, or 24 h (h). c CD4⁺ T cells were positively selected from mouse spleens using magnetic beads and then stimulated with 10 μg/mL plate-bound anti-CD3 and 2 μg/mL anti-CD28 in the presence of 250 nM APG-115 or DMSO for the indicated periods of time. T cell activation markers (CD25 and CD62L) were determined by flow cytometry. CD25^high CD62L^low T cells represented an activated population. d an increase in cell size was shown after APG-115 treatment

Earlier studies suggest that p53 is also involved in the regulation of PD-L1 expression [37]. We then evaluated if APG-115 influenced PD-L1 expression on tumor cells besides its effects on immune cells. After in vitro treatment of MH-22A cells with APG-115, expression of p53 and p-STAT3 proteins were upregulated in a dose-dependent manner, indicating activation of p53 and STAT3 signaling pathway in these tumor cells (Fig. 3a). As a downstream component of STAT3 pathway, PD-L1 levels were elevated accordingly. Flow cytometry analysis further revealed that APG-115 treatment resulted in a dose-dependent increase in the surface expression of PD-L1 on tumor cells (Fig. 3b and c). The data suggests that induction of PD-L1 expression on tumor cells by APG-115 may sensitize these cells to anti-PD-1 therapy.Fig3Fig. 3

APG-115 upregulates PD-L1 expression on MH-22A tumor cells. MH-22A mouse tumor cells were treated with indicated concentrations of APG-115 for 72 h in vitro. a expression levels of MDM2, p53, total STAT3 (t-STAT3), phosphorylated STAT3 (p-STAT3), PD-L1, and β-actin (loading control) were determined by Western blotting. b PD-L1 expression levels which were reflected by fluorescence intensity were determined by flow cytometry and the same results were shown as a bar chart (c)

The above data suggest that p53 activation by APG-115 regulates immune responses, potentially including both adaptive and innate immunity. We then asked if the combined therapy of APG-115 and PD-1 blockade synergistically enhances antitumor immunity in vivo. Presumably, APG-115 activates immune response through the immune cells in the TME and, most likely, its immunological effect is independent of Trp53 status of tumors. Therefore, syngeneic tumor models with various Trp53 status, including Trp53 ^wt MH-22A, Trp53 ^mut MC38, and Trp53 ^−/− MH-22A models, were used to test the hypothesis.

In Trp53 ^wt MH-22A hepatoma syngeneic model, APG-115 single agent showed no antitumor activity, whereas the anti-PD-1 antibody effectively reduced tumor volume by exhibiting a T/C (%) value of 22% on d15 (Fig. 4a). Addition of 10 mg/kg or 50 mg/kg APG-115 to PD-1 blockade enhanced antitumor activity by showing T/C (%) values of 17 and 6%, respectively. Because the tumors had reached the maximum allowable size, the animals in vehicle and two APG-115-treated groups were sacrificed on d15 while the remaining three groups continued on treatment. At the end of treatment (d22), one out of eight animals treated with the anti-PD-1 antibody showed SD (i.e., 12.5% response rate). In the combination groups, one SD and one CR appeared under 10 mg/kg APG-115 (i.e., 25% response rate) and one SD and two CR occurred under 50 mg/kg APG-115 treatment (i.e., 37.5% response rate). The tumor growth curves were continually monitored for additional 21 days after drug withdrawal. On d42, the response rates for anti-PD-1, APG-115 (10 mg/kg) plus anti-PD-1, and APG-115 (50 mg/kg) plus anti-PD-1 were 12.5% (1 PR), 25% (2 CR), and 62.5% (2 SD, 1 PR, 2 CR), respectively.Fig4Fig. 4

APG-115 enhances anti-PD-1 antibody mediated tumor suppression in Trp53 ^wt , Trp53 ^mut and Trp53 ^−/− syngeneic mouse tumor models. APG-115 was tested alone and in combination with anti-PD-1 antibody in mice subcutaneously implanted with Trp53 ^wt MH-22A (a-d; n = 8), Trp53 ^mut MC38 (e-g; n = 10), or Trp53 ^−/− MH-22A (h-j; n = 10) tumor cells. APG-115 was orally administered every day in Trp53 ^wt MH-22A models or every other day in both Trp53 ^mut MC38 and Trp53 ^−/− MH-22A models. Anti-PD-1 antibody was administered intraperitoneally BIW. Treatments were conducted for 3 weeks in Trp53 ^wt MH-22A and Trp53 ^mut MC38 models, and for 12 days in Trp53 ^−/− MH-22A model. Data representing at least two independent experiments were presented as the mean of tumor volumes of mice in each group (A, E, H) or tumor volumes for individual mice (B, C, D, F, G, I and J). The control groups were treated with APG-115 vehicle (A) or isotype antibody plus vehicle (I + V; E and H)

To investigate the mechanism underlying enhanced antitumor activity of the combined therapy, we next assessed TILs in the TME by flow cytometry. In Trp53 ^wt MH-22A syngeneic tumors, in comparison with the control, treatment with anti-PD-1 alone only slightly increased the proportions of CD45⁺ cells, CD3⁺ T cells, and cytotoxic CD8⁺ T cells without reaching statistical significance (P > 0.05, Fig. 6a), whereas the combined therapy exerted a more significant effect of increasing infiltration of these cells (P < 0.01). There were approximately 1.5 to 2-fold increases relative to the control. Additionally, M1 macrophages was significantly increased by either anti-PD-1 antibody or combined therapy in comparison with the control (P < 0.01); however, there was no significant difference between these two treatments (P > 0.05). Most strikingly, M2 macrophages was significantly decreased by the combined therapy in comparison with both control (P < 0.01) and anti-PD-1 monotherapy (P < 0.05).Fig6Fig. 6

Flow cytometry analysis of TILs in the TME of syngeneic tumors with wild-type (a) or mutant (b) Trp53 . Mice with established MH-22A or MC38 tumors were treated with 10 mg/kg APG-115 (a and b), 10 mg/kg (a) or 5 mg/kg (b) anti-PD-1 antibody, or the combination as described in the legend of Fig. 4. The control group was treated with isotype control antibody and APG-115 vehicle (I + V). On day 14, syngeneic tumors were harvested, dissociated into single-cell suspensions, and stained for flow cytometry analysis. Percentages of CD45⁺, CD3⁺ T cells, CD8⁺ T cells, M1 and M2 macrophages in the tumors under the different treatments were assessed. Data were representative of two (a) or three (b) independent experiments and shown as dot plots (n = 5 or 10). ****P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05, by one-way ANOVA with Bonferroni post-test. I + V indicates isotype control and vehicle of APG-115