BMS202

Combination of Chidamide-Mediated Epigenetic Modulation with Immunotherapy: Boosting Tumor Immunogenicity and Response to PD-1/PD-L1 Blockade

Kun Tu,# Yulin Yu,# Yi Wang, Ting Yang, Qian Hu, Xianya Qin, Jingyao Tu, Conglian Yang, Li Kong, and Zhiping Zhang*

1. INTRODUCTION

Triple-negative breast cancer (TNBC) is a subtype of breast carcinoma with highly aggressive and metastatic characteristics. Owing to the absent expression of the progesterone receptor, estrogen receptor, and human epidermal growth factor receptor 2, the specific targeted therapy for TNBC has not been proposed at present.1 Many studies have been performed to explore the effective therapeutic approaches for TNBC, including immunotherapy and immunochemotherapy.2−4 Cancer immunotherapy, especially immune checkpoint block- ade (ICB) therapy, has made revolutionary clinical advances in a variety of cancers, while the application as a single therapeutic tool can only benefit a small subset of patients, which could be attributed to a “cold” immune microenviron- ment.5−7 TNBC has been considered as a “cold” tumor owing to barren tumor antigens and limited intratumoral infiltration of immune cells, which severely restricts the responsiveness and efficiency of ICB therapy.8,9 Therefore, how to enhance antitumor immunity is the key point for TNBC immunother- apy. Growing evidence has demonstrated that the occurrence of immunogenic cell death (ICD) within a tumor can induce an effective antitumor immune response, enhancing the treatment efficiency of ICB therapy.10,11 ICD is a type of programmed cell death, accompanied by the expression of tumor antigens and the emitting of highly immunostimulatory danger signals, such as high mobility group boX 1 (HMGB1) release, calreticulin (CRT) exposure, and adenosine triphos- phate (ATP) secretion.11,12 These danger signals play vital roles during the activation and enhancement of antitumor immunity, such as immune cell recruitment, the capture and presentation of the tumor antigen, and so forth.12,13 Never- theless, only a limited number of approaches [e.g., oXaliplatin (OXA), anthracyclines, radiotherapy, photodynamic therapy (PDT), and oncolytic viruses] were verified to induce ICD response.14 Thus, it is worth exploring other therapeutic agents to increase tumor immunogenicity and subsequently enhance the therapeutic efficacy of ICB therapy.

Histone deacetylase inhibitor (HDACi), an epigenetic modulator, has been demonstrated to exert multiple biological functions, such as inducing tumor apoptosis, inhibiting cell cycle and promoting cell differentiation.15,16 As a novel subtype-selective HDACi, chidamide (CHI) could selectively inhibit the activity of class I and class IIb HDAC with a low- concentration, which was applied to treat various cancers by monotherapy or in combination with other therapeutic agents.17−19 The reactive oXygen species (ROS)-mediated mitochondria apoptosis pathway is one of the apoptosis mechanisms induced by CHI, which is a common phenomenon in several prototypical ICD-induced agents, such as anthracyclines, PDT, and radiotherapy.14,20−22 Additionally, CHI could decrease the HMGB1 expression in acute myeloid leukemia cells, while the whereabouts of HMGB-1 were not described in detail.23 We conjectured that it might be released into the outside of the cells, which is an important feature of the ICD effect. Based on this consideration, we speculated that CHI might be a potential ICD inducer. Given the versatile antitumor activity and immunomodulatory capability of CHI, the compositional approach of CHI with ICB therapy could be a new cancer treatment strategy that is worth exploring.

As the main form of ICB therapy, anti-PD-1/PD-L1 monoclonal antibodies have been used to treat many kinds of cancers.24,25 However, several drawbacks still remained in antibody-based treatment strategies, such as poor tumor penetration, high treatment costs, and immune-related side effects.26 Therefore, developing a free-antibody approach for blocking the PD-1/L1 pathway possesses a significant clinical value, such as nucleic acid, a peptide, and a small molecule inhibitor.27−29 Among them, chemical small-molecule inhib- itors have attracted great research attention and interest owing to their good affinity, high stability, little immunogenicity, and low price. Bristol-Myers Squibb (BMS) has recently revealed a series of small-molecule compounds with good blocking activity for the PD-1/PD-L1 pathway, such as BMS-1001, BMS-202, and BMS-1166. As the representative of these inhibitors, BMS-202 could occupy the specific site of PD-L1 and induce protein dimerization to inhibit the PD-1/PD-L1 interaction.30 Related experiment results indicated that BMS-202 could inhibit tumor growth and improve antitumor immunity to some extent.

In this report, we described a liposome system to codeliver CHI and BMS-202 for the treatment of TNBC (Scheme 1). The CHI−F127 complex with Pluronic F127 was first prepared for increasing the water solubility of CHI and then encapsulated with BMS-202 into liposome with different drug- loaded spaces via the reverse evaporation method (CHI/BMS- 202@lipF). The liposome possessed good biocompatibility and security and endowed the therapeutics with favorable biodistribution. After intravenous administration, the liposome could passively accumulate in the tumor and steadily release drugs. The released CHI could induce tumor cell apoptosis and ICD response, enhance tumor immunogenicity, and promote the intratumoral infiltration of immune cells. Meanwhile, the released BMS-202 could induce PD-L1 protein dimerization on tumor cells to block PD-1/PD-L1 and then boost the antitumor immunity of immune cells. We expected that CHI/BMS-202@lipF could synergistically improve the tumor immunosuppressive microenvironment and realize effective antitumor efficacy. In summary, it may provide a referential approach and a new insight for clinical antitumor therapy based on HDACi combined with immunotherapy.

Scheme 1. Schematic Illustration of CHI/BMS-202@lipF- Mediated Synergistic TNBC Treatment; Briefly, CHI and BMS-202 Were Co-Loaded in Liposome through the Modified Reverse Evaporation Method; after Intravenous Administration, CHI/BMS-202@lipF Could Passively Accumulate in the Tumor and Sustainedly Release Drugs; the Released CHI Could Induce ICD Response within the Tumor, Including HSP Up-Regulation, CRT Exposure, ATP, and HMGB-1 Release; Additionally, CHI Increased the MHC I and MHC II Levels on Cancer Cells, Which Contributed to Antigen Presentation and T-Cell Recognition; Furthermore, CHI Was Capable of Promoting DC Maturation and NK Cell Infiltration and Activation within a Tumor; Meanwhile, the Released BMS-202 Blocked PD-1/PD-L1 Interaction, Followed by Restoring the Antitumor Immunity of T-Cells; Combined Therapy Using CHI and BMS-202 Effectively Enhanced Tumor Immunogenicity and Antitumor Immunity, Thereby Inhibiting Tumor Growth and Metastasis.

2. EXPERIMENTAL SECTION
2.1. Materials and Reagents. CHI of purity > 98% was purchased from Hunan Huateng Pharmaceutical Co., Ltd. BMS-202 was obtained from
Shanghai Biochempartner Co., Ltd. Cholesterol, 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC), and 1,2-distearo- yl-sn-glycero-3-phospho-ethanolamine-N-[methoXy (polyethylene gly- col)-2000] (DSPE-mPEG2k) were all supplied by Xi’an RuiXi Biological Technology Co. Ltd. Pluronic F127 was purchased from Sigma-Aldrich. 1,1-Dioctadecyl-3,3,3,3-tetramethylindotri-carbocyai- neiodide (DiR) was obtained from Absin Bioscience Inc. Recombi- nant mouse PD-L1 protein was bought from Sino Biological Inc. Anti- CD3e antibody and anti-CD28 antibody were provided by Boster Biological Technology Co., Ltd. The antibodies of anti-HMGB1 (3935S), anti-Bax (2772S), anti-Bcl-2 (3498S), anti-Mcl-1 (94296S), anti-HSP70 (4872S), anti-HSP90 (4877S), anti-GAPDH (5174S), and anti-PD-L1 (64988S) were obtained from Cell Signaling
Technology, Inc. PE-conjugated anti-CD86, fluorescein isothiocya- nate (FITC)-conjugated anti-CD80, antigen-presenting cell (APC)- conjugated anti-CD11c, PE-conjugated anti-DX5, PE-Cy7-conjugated anti-CD69, PerCP-Cy5.5-conjugated anti-CD3e, FITC-conjugated anti-CD8a, PE-conjugated anti-CD4, FITC-conjugated anti-I-A/I-E, APC-conjugated anti-H-2Kd, and PE-conjugated anti-PD-L1 were purchased from BD Pharmingen.

2.2. Cell Lines and Animals. 4T1 cells were provided by the Chinese Academy of Science and Technology Cell Bank and cultured in RMPI 1640 (Gibco) complete medium in the environment of 5% CO2 at 37 °C.

Mouse bone marrow-derived dendritic cells (BMDCs) were gained as previously described.34 Briefly, the BMDCs were isolated from mouse marrow cavities of tibias and femurs under aseptic conditions and cultured in the RPMI 1640 complete medium containing 20 ng/ mL recombinant granulocyte-macrophage colony-stimulating factor. After 7 days of continuous cultivation, BMDCs were collected for subsequent experiments.
Balb/c mice (female, 16−18 g) were provided by Hubei Provincial Center for Disease Control and Prevention and raised in the Animal Center of Huazhong University of Science and Technology. The experiments involving the animals were performed in compliance with the stipulations of the local Ethical Committee and Chinese law.

2.3. Apoptosis Study of CHI on 4T1 Cells. The cytotoXicity was first evaluated by an MTT assay. Briefly, 5 × 103 4T1 cells were inoculated per well in a 96-well plate and cultured for 12 h. Then, different concentrations of CHI were added and incubated for 24 and
48 h. Then, the MTT reagent was added into the supernatant. Followed by incubation, the supernatant was replaced by DMSO. The absorbance was detected using a microplate reader (Multiskan MK3, Thermo, MA, USA) at 490 nm.

ROS generation and intracellular ATP level were evaluated using an assay kit (Beyotime Institute of Biotechnology). 1 × 105 4T1 cells were cultured per well in 12-well plates and cultured for 12 h. Then, a series concentration of CHI was added. After being incubated for 6 h, 10 μM dichlorodihydrofluorescein diacetate was added and incubated for 20 min at 37 °C for the detection of ROS by flow cytometry. To investigate the intracellular ATP level, CHI was added into plates and incubated for 24 h. Afterward, the cells were harvested and lysed to monitor the ATP using a luminometer following the methods of the ATP assay kit.

To detect the related protein (Bcl-2, Mcl-1, Bax, HSP70, and HSP90) expression after CHI treatment, 2 × 105 4T1 cells were cultured in siX-well plates and incubated for 12 h. After being treated with CHI for 24 h, the cells were lysed to extract the total protein for further western blot analysis. The calculation formula for the quantitative data is as follows.

2.4. Investigation of CRT Exposure, ATP, and HMGB1 Release In Vitro. 2 × 105 4T1 cells were cultured in siX-well plates. With 12 h of incubation, the cells were treated with cisplatin (CDDP, 1, 5, or 20 μg/mL), OXA (1, 5 or 20 μg/mL), and CHI (1, 5, or 20 μg/mL). After 24 h, the cell supernatant was collected for detecting ATP and HMGB-1. After being treated with drugs for 12 h, the cells were collected to evaluate the expression of CRT by flow cytometry. The visualization of HMGB-1 distribution and CRT exposure was performed by immunofluorescence (IF) analysis. 4T1 cells (1 × 105) were inoculated into a confocal dish and cultured overnight. The cells were then treated with CDDP (5 μg/mL), OXA (5 μg/mL), and CHI (5 μg/mL). To observe the distribution of HMGB-1, after being treated with drugs for 24 h, the cells were stained with the HMGB1 antibody and DAPI and visualized by CLSM. To view the CRT surface exposure, after being treated with drugs for 12 h, the cells were stained with the CRT antibody and DAPI for visualizing by CLSM.

2.5. In Vivo Antitumor Vaccination Response with CHI- Induced Dying Tumor Cells. The experiment schedule for vaccination is described in Figure 2A. Briefly, 4T1 cells were treated with CDDP (20 μg/mL), CHI (20 μg/mL), and OXA (20 μg/mL) for 24 h. Afterward, harvested cells were subcutaneously (s.c.) inoculated twice (on days −14 and −7) into the left flank of Balb/c mice at a density of 1 × 106. Then, 2 × 105 living 4T1 cells were s.c.injected into the contralateral (right) flank of mice on day 0. The contralateral tumor volume was monitored every 2 days and calculated in compliance with the formula of 1/2 × length × width2. On day 22, the tumor-bearing mice were sacrificed to collect the tumors for IF analysis. Moreover, the tumor-free mice were further re-challenged to investigate the immune memory.

2.6. Induced Maturation of BMDCs by CHI In Vitro. BMDCs were isolated and cultured as described above. The cells were incubated with CHI (0.5, 1, or 2 μg/mL) for 24 h. Afterward, the cells were collected and labeled with CD80, CD86, and CD11c antibodies. Then, the cells were detected by flow cytometry.

2.7. Expression of MHC I, MHC II, and PD-L1 on 4T1 Cells after Being Treated with CHI In Vitro. 1 × 105 4T1 cells were cultured in 24-well plates for 12 h. After being treated with CHI (0.5, 2, or 5 μg/mL) for 24 h, the cells were collected and stained with PD- L1, I-A/I-E, and H-2Kd antibodies. Subsequently, the expression levels were detected by flow cytometry. Cells incubated with IFN-γ were used as the positive control for PD-L1 detection.

2.8. Effects of BMS-202 on Proliferation and IFN-γ Secreted by Splenocytes In Vitro. Splenocytes were isolated from mouse spleen under aseptic conditions as per the manufacturer’s instructions. The splenocytes were resuspended with RPMI 1640 complete medium containing anti-CD28 (2 μg/mL) with or without the mouse PD-L1 protein (final concentration, 100 ng/mL) and seeded into 24-well plates that were pre-coated with anti-CD3 at a density of 5 × 105. Subsequently, BMS-202 (final concentrations, 1, 10, and 100 ng/mL) was added and treated for 72 h. The cells and supernatant were collected to detect CD3+ cells and IFN-γ by flow cytometry and an ELISA kit, respectively.

2.9. Preparation and Characterization of Liposomes.
2.9.1. Preparation of the CHI−F127 Complex. CHI (1 mg) and F127 (6 mg) were, respectively, dissolved in chloroform and deionized water (ddH2O). The miXture of CHI and F127 was then sonicated at 250 W for 1.5 min using an ultrasonic processor (JY92- IIN, Scientz, China). The chloroform was removed by rotary evaporation to form the CHI−F127 complex. The complex was freeze-dried for further use. The proportion of CHI in the complex was determined by HPLC (Ultimate 3000, Thermo Fisher Scientific, USA). Chromatographic separation was performed on a Kromasil 100-5 C18 column (150 × 4.6 mm, 5 μm) at 30 °C. The mobile phase consisted of acetonitrile (A) and 0.02 M KH2PO4 water solution (B, adjusted to pH 4.0 using phosphoric acid) at 1.0 mL/ min. The gradient elution program was executed as follows: 0−7.00
min, A/B = 30:70%; 7.00−7.01 min, A/B = 80:20%; 7.01−13.00 min,A/B = 80:20%; 13.00−13.01 min, A/B = 30:70%; 13.01−15.00 min,A/B = 30:70%. The detection wavelength was 258 nm. In addition, BMS-202 also could be detected using the method at a wavelength of 210 nm.

2.9.2. Preparation of CHI/BMS-202@lipF. Drug-loaded liposomes were prepared by the reverse evaporation method. In brief, 8 mg of DOPC, 7 mg of DSPE-mPEG2k, 2 mg of cholesterol, and 1 mg of BMS-202 were dissolved in 3 mL of chloroform. Afterward, 0.5 mL of CHI−F127 water solution containing 1.2 mg of CHI was miXed with the above organic phase. The miXture was sonicated at 200 W for 1 min to form water-in-oil emulsion. Then, the emulsion was transferred into a round-bottom flask and a gel film was formed by removing the organic solvent using a rotary evaporator. Subsequently, 1 mL of PBS was added into the flask for hydration to obtain the
drug-loaded liposome (CHI/BMS-202@lipF). For fluorescence visualization, BMS-202 was replaced by DiR to label liposome (DiR@lipF) in the absence of CHI.

2.9.3. Characterization of CHI/BMS-202@lipF. Dynamic light scattering (DLS, Brookhaven, USA) was performed to measure the size and zeta potential of liposomes. The morphology of liposomes was observed using a transmission electron microscope (JEM-1230,Japan). The stability of liposomes in ddH2O, PBS, and FBS was continuously monitored by DLS. The contents of CHI and BMS-202 in liposomes were detected using HPLC as established above. Afterward, the encapsulation efficiency (EE) and loading efficiency (LE) of CHI and BMS-202 were calculated using the following formulas: EE = (mass of loaded drug)/(initial mass of drug) × 100%; LE = (mass of loaded drug)/(mass of drug-loaded liposome) × 100%. To study the drug release behavior of liposomes in vitro, CHI/BMS- 202@lipF was placed in a dialysis bag with a molecular weight cut-off of 3500 Da and dialyzed with 0.01 M PBS (pH 7.4, 6.5, or 5.5) containing 0.5% Tween 80 accompanied by shaking at 37 °C. At different time points, the dialysate was renewed with fresh media. Subsequently, the dialysate was lyophilized and redissolved with methanol to detect the released content of CHI and BMS-202 using the HPLC method as mentioned above.

Figure 1. CHI induced cell apoptotic and ICD effects in vitro. (A) Cell viability of 4T1 cells incubated with CHI for 24 h and 48 h (n = 5). (B) Intracellular ROS and (C) ATP levels of 4T1 cells incubated with CHI for 6 h (ROS) and 24 h (ATP), respectively (n = 3). (D) EXpression of Bcl- 2, Bax, Mcl-1, HSP70, and HSP90 on 4T1 cells incubated with CHI for 24 h as determined by western blot. (E) CLSM examination of CRT (red) and DAPI (blue) in 4T1 cells treated with PBS, CDDP (5 μg/mL), CHI (5 μg/mL), and OXA (5 μg/mL) for 12 h. (F) CRT+ 4T1 cells detected by flow cytometry (n = 3). (G) ATP and (H) HMGB-1 release from 4T1 cells exposed to different concentrations of CDDP, CHI, and OXA for 24
h. (I) CLSM examination of HMGB1 (red) release in 4T1 cells treated with PBS, CDDP (5 μg/mL), CHI (5 μg/mL), and OXA (5 μg/mL) for 24
h. Scale bar is 25 μm. Data were expressed as mean ± SD. *p < 0.05, **p < 0.01, ns: no significant difference. 2.10. In Vivo Biodistribution. 2 × 105 4T1 cells were implanted into the breast pad of a mouse to establish an orthotopic breast cancer model. When the tumor volume was 300−400 mm3, free DiR or DiR@lipF was intravenously (i.v.) injected at the dose of 0.5 mg/kg DiR per mouse. At preconcerted time points, the mice were scanned and imaged using an IVIS lumina imaging system (LICOR, Pearl Trilogy, USA). For investigation of the tissue distribution ex vivo, the tumor and major organs were collected, followed by imaging and quantified fluorescence intensity. 2.11. In Vivo Antitumor Effect. The 4T1 orthotopic tumor model was established as described above. At first, the antitumor efficacies of different dosages of CHI were investigated. When the tumor size reached 30−50 mm3, the mice were randomly separated into five groups (n = 5) and i.v. injected with saline or CHI@lipF (1, 2, 5, or 10 mg/kg of CHI) four times at 3-day intervals. The tumor growth and body weight were detected every 2 days. On day 19, all mice were sacrificed. Tumors were harvested for H&E and immunohistochemical (IHC) analysis. To evaluate the synergistic antitumor efficacy of CHI and BMS- 202, the mice were randomly separated into four groups (n = 8). When the tumor volume reached 30−50 mm3, the mice were i.v. injected with saline, CHI@lipF, BMS-202@lipF, or CHI/BMS-202@ lipF (CHI or BMS-202, 5 mg/kg). The treatment was performed four times at 3-day intervals. The tumor growth and body weight of mice were detected every 2 days. The activity and mental state of the mice were observed every day. When a clear death signal appeared or a large tumor size (>2000 mm3), the mouse was recorded as dead. In addition, the lungs were harvested on day 28 to perform the H&E analysis.

2.12. In Vivo Anti-metastasis Efficacy. To establish a mouse 4T1 lung metastasis model, 4T1 cells (5 × 105) were i.v. injected into female Balb/c mice (day 0). On day 1, the mice were randomly separated into four groups (n = 6) and i.v. injected with saline, CHI@ lipF, BMS-202@lipF, or CHI/BMS-202@lipF a total of four times at 3-day intervals. The dose was 5 mg/kg for CHI or BMS-202. On day 16, all mice were sacrificed. The lungs were collected and fiXed with bovine fiXative solution for 24 h, followed by counting the tumor nodules. In addition, the lungs were further analyzed by H&E staining.

2.13. Immune Response Analysis In Vivo. To investigate the proliferation and function of immune cells in vivo after combinational treatment, the spleen, lymph nodes, and the tumor were harvested from 4T1 tumor-bearing mice at 48 h post the last treatment. Subsequently, the harvested tissues were made into a single-cell suspension as previously reported.35 Anti-CD16/32 was used for blocking the nonspecific binding and then stained with corresponding fluorescein-conjugated antibodies: (1) dendritic cells (DCs) (CD11c+CD80+CD86+), (2) natural killer (NK) cells (CD3−DX5+),(3) activated NK cells (CD3−DX5+CD69+), (4) CD8 T cells (CD3+CD8+), and CD4 T cells (CD3+CD4+) according to the instructions of manufacturer. Cells were analyzed using a flow cytometer. In addition, H&E, IF (CD4 and CD8), and IHC (PD-L1) analyses were performed on the tumor.

2.14. Statistical Analysis. Results were presented as means ± standard deviations (SDs). One-way analysis of variance was performed to analyze the significant difference of data through GraphPad Prism 7 software. The significant differences were set at *p < 0.05, **p < 0.01, and ***p < 0.001. 3. RESULTS AND DISCUSSION 3.1. CHI Induced Mitochondrial Dysfunction and ER Stress In Vitro. The antitumor molecular mechanism of HDACi was mainly mediated through regulating gene transcription by changing the chromosome structure, further affecting the downstream antitumor signaling pathways.16 As a novel HDACi, CHI has been demonstrated to inhibit tumor cell growth through multiple pathways, including cell cycle arrest, apoptosis, autophagy, differentiation, and so forth. For simplicity, we first used the MTT assay to test the cytotoXicity of CHI on 4T1 cells. CHI could inhibit cell viability in both dose- and time-dependent manners (Figure 1A). The IC50 values for CHI were 19.6 ± 1.3 and 4.1 ± 0.3 μg/mL at 24 and 48 h, respectively. Previous studies have demonstrated that HDACi could increase the intracellular ROS level and cause cell apoptosis.20,21 Intracellular ROS plays an important role in tumor immunogenicity.36 Thereby, we further detected the ROS level of 4T1 cells treated with CHI for 6 h. As shown in Figure 1B, intracellular ROS increased in a dose-dependent manner. The intracellular ROS of the group treated with a high concentration of CHI (20 μg/mL) increased nearly twofold compared with the control group. It was reported that ROS generation was usually accompanied by mitochondrial malfunction, and the mitochondria-associated signaling path- way was an important point in apoptosis, induced by CHI.37,38 In Figure 1C, it exhibited the change of intracellular ATP (a characteristic indicator of mitochondrial function) on 4T1 cells treated with CHI for 24 h. The intracellular ATP level of the CHI-treated groups was much lower than that of the control group, especially in groups with a high concentration. Collectively, the activated mitochondria−apoptosis pathway induced by ROS generation and mitochondrial malfunction might be one of the antitumor mechanisms of CHI. Bcl-2 family proteins could adjust the permeability of the mitochondrial membrane, which could interfere with mito- chondrial cytochrome C releasing, caspase 3 and caspase 9 activating, and then regulate cell apoptosis.39,40 For further confirmation of the mitochondrial-mediated apoptosis path- way, 4T1 cells were chosen and treated by CHI for 24 h. Then, we detected the expression level of the Bcl-2 family protein using the western blot, including Bax, Bcl-2, and Mcl-1. As shown in Figure 1D, the expression of Bax was significantly increased with the enhancement of the drug dose, whereas the expression of Bcl-2 was sharply inhibited even at a low concentration (1 μg/mL). More importantly, Mcl-1 was also down-regulated in a dose-dependent manner. Mcl-1 played a critical role in the process of resisting cell apoptosis caused by endoplasmic reticulum (ER) stress.41,42 Interestingly, ER stress is a typical initiator for inducing ICD, which could enhance tumor immunogenicity and antitumor immune response.43 Certain classic ICD inducers (such as anthracyclines) were often reported with the down-regulation of Mcl-1.44,45 Therefore, based on the results of CHI-induced ROS generation, mitochondrial apoptosis, and ER stress, we hypothesized that CHI might be a potential ICD inducer. 3.2. Identification of CHI as a Potential ICD Inducer.So far, only a few cancer-treatment therapeutics (e.g., OXA, anthracyclines, radiotherapy, oncolytic viruses, and PDT) have been reported to own the ability of inducing an ICD response.14 ICD is one kind of programmed cell death. When a cell was the in ICD stage, it would release many kinds of damage-related molecular patterns (DAMPs) and could elicit an effective immune response against cancer. Several pivotal DAMP molecules have been identified, including heat shock proteins (HSPs), CRT, HMGB1, and ATP. These molecules were also considered as predominant hallmarks of ICD evaluation. Many research studies have reported that when cells were in the ICD stage, they would over-express HSP70 and/or HSP90.13 As a protein with chaperone activity, an HSP could facilitate the transportation of the MHC I−peptide complex to the extracellular and/or cell surface,meanwhile enhancing the capability of DCs, such as migration and phagocytosis to apoptotic cells.46 As shown in Figure 1D, HSP70 was sharply increased even at a low treating concentration of CHI (1 μg/mL), while HSP90 exhibited moderate up-regulation in a dose-dependent manner. The results indicated that CHI could promote the expression of an HSP on tumor cells, especially for HSP70, which contributed to enhancing the antitumor immune response. When cancer cells were treated with ICD inducers, they would express CRT, a decisive biomarker of ICD, on the cell surface, which could attract DCs to phagocytose dying tumor cells.47 Therefore, we detected CRT on 4T1 cells after treating with drugs for 12 h by CLSM and flow cytometry. CHI was compared with OXA (a positive ICD inducer) and CDDP (a negative ICD inducer). As shown in Figure 1E, there was no visible expression of CRT in the cells treated with CDDP, whereas obvious CRT exposure was seen in cells treated with OXA or CHI. Besides, the quantitative results measured by flow cytometry showed that the CRT+ cells in the groups of OXA and CHI were, respectively, 59.7 ± 1.5 and 47.9 ± 2.1%, which were significantly higher than that of CDDP (17 ± 1.7%) and PBS (9.2 ± 0.8%) (Figure 1F). Figure 2. CHI induced ICD response to elicit effective antitumor vaccination on 4T1 breast cancer. (A) EXperiment schedule of the vaccination. Dying tumor cells were collected from the 4T1 cells treated with CDDP, CHI, or OXA for 24 h. Then, the healthy female Balb/c mice were treated with 2 rounds of s.c. injection of dying tumor cells (1 × 106 per mouse) on days −14 and −7, followed by s.c. injection of live 4T1 cells (2 × 105 per mouse) on the contralateral flank (day 0) (n = 8). The contralateral tumors were collected on day 22 for IF analysis. In addition, the tumor-free mice were further re-challenged with 2 × 105 live 4T1 cells per mouse and monitored up to day 35. (B) 4T1 tumor growth curves of each mouse at the contralateral site. (C) Image and (D) weight of tumor tissues collected from the contralateral site on day 22. (E) Percentage of tumor-free mice at the contralateral site. (F) CD4+ and CD8+ T cells on the contralateral tumor. Scale bar is 100 μm. Data were expressed as means ± SD. ***p < 0.001. The extracellular release of ATP and HMGB-1 mainly occurred in the mid and late stages of ICD. ATP, a “find me” signal, was capable of attracting monocytes, including DCs, to the site of apoptosis, while HMGB-1 could promote DC maturation by activating toll-like receptor 4.12 As shown in Figure 1G,H, the release of ATP and HMGB-1 in cells treated with OXA or CHI for 24 h increased in a dose-dependent way and was obviously higher than that of the CDDP-treated group. Furthermore, we evaluated the process of HMGB-1 release by CLSM. Figure 1I shows that HMGB-1 was mainly located in the nucleus of cells treated with PBS and CDDP. In contrast, HMGB1 was almost totally released to the cytoplasm or outside of cells in the OXA- and CHI-treated groups. Collectively, these results demonstrated that HSP up- regulation, CRT exposure, and ATP and HMGB-1 release in CHI-treated cells verified that CHI was capable of inducing an ICD response in vitro. 3.3. In Vivo Antitumor Vaccination Response with CHI-Induced Dying Tumor Cells. The vaccination response in a syngeneic animal model was regarded as the gold standard for identifying an ICD inducer.11 In brief, s.c. vaccination of dying tumor cells caused by an ICD inducer could arouse an antitumor immune response in immunocompetent mice, followed by impeding the onset and progression of rechallenged tumors which were seeded into vaccinated mice with the homologous live tumor cells. As seen in the description in Figure 2A, dying tumor cells were collected from the 4T1 cells that were treated with CDDP, CHI, or OXA for 24 h and then s.c. inoculated to one flank of healthy female Balb/c mice on two occasions (7-day interval). Subsequently, the mice were implanted with 4T1 cells in the contralateral flank by s.c. injection. The mice that vaccinated with OXA- and CHI-treated 4T1 dying cells showed significantly inhibited the growth of contralateral tumor, while the tumor progress in CDDP-treated group was similar to that the control group (Figure 2B−D). Notably, 4 and 2 out of 8 mice without tumorigenesis occurred in OXA- and CHI-treated groups, respectively. The contralateral tumors were harvested on day 22 for immunological analysis. Furthermore, re-challenging these tumor-free mice with live 4T1 cells in these tumor-free mice, the mice remained without tumor recurrence, which indicated the establishment of immune memory (Figure 2E). It is revealed in Figure 2F that in the mice vaccinated with OXA- and CHI-treated 4T1 dying cells, CD4+ and CD8+ T cells were remarkably infiltrated in the contralateral tumor, while CDDP-treated group showed a negligible change compared with the control. Figure 3. Immunomodulatory effects of CHI and BMS-202 in vitro. The expression of (A) PD-L1, (B) MHC I, and (C) MHC II on 4T1 cells treated with CHI for 24 h (n = 3). (D) Quantitative analysis and (E) representative flow cytometry images of CD80+CD86+ (gate on CD11c+) in BMDCs treated with CHI for 24 h (n = 3). (F) Schematic immune regulation effects of BMS-202. (G) Number of CD3+ cells and (H) IFN-γ secretion in splenocytes with different treatments. Data were expressed as means ± SD. *p < 0.05, **p < 0.01, ***p < 0.001. Therefore, based on the above results, we confirmed the conclusion that CHI can work as an ICD inducer and was capable of enhancing tumor immunogenicity and antitumor immunity through inducing an ICD response. 3.4. Immunomodulatory Effects of CHI In Vitro.Beyond direct cytotoXicity and an ICD-inducing effect, we of a tumor to anti-PD-1/PD-L1 therapy (Figure 3A). Moreover, CHI was also capable of up-regulating the expression of MHC-related molecules on tumor cells to promote antigen presentation and T-cell recognition. As shown in Figure 3B, CHI obviously increased the expression of MHC I on 4T1 cells even at a relatively low concentration (0.5 μg/mL), which contributed to the recognition of tumor cells by CD8+ T cells. Additionally, the MHC II of tumor cells was also up-regulated in a dose-dependent manner to the benefit of CD4+ T cell recognition (Figure 3C).EXcept for tumor cells, HDACi could also directly regulate the function of immune cells. In order to explore the mechanism of DC maturation induced by CHI, we treated BMDCs with CHI for 24 h. As shown in Figure 3D,E, the expression of costimulatory molecules of CD80 and CD86 on BMDCs was sharply increased in a dose-dependent manner.also explored other immunomodulatory functions of CHI, such as increasing PD-L1 and major histocompatibility complex (MHC)-related molecular expression on tumor cells and DC maturation. We demonstrated that the expression of PD-L1 on 4T1 cells was increased in a dose-dependent way after being treated with CHI for 24 h, which could promote the sensitivity.The ratio of CD80+CD86+ cells in the group treated with CHI (2 μg/mL) increased 2.5-fold compared to the control, which indicated that CHI was capable of effectively promoting DC maturation. 3.5. Blockade of PD-1/PD-L1 Interaction by BMS-202 In Vitro. It is generally accepted that PD-1/PD-L1 interaction could inhibit the proliferation and cytokine secretion of T-cells. BMS-202, a small-molecular inhibitor of PD-L1, can inhibit PD-1/PD-L1 interaction by inducing PD-L1 protein dimeriza- tion.30 To directly evaluate the blocking ability of BMS-202 on PD-1/PD-L1 interaction at the cellular level, we isolated lymphocytes from mice spleen as the research object. As shown in Figure 3F−H, the splenic lymphocyte stimulated by the CD3 and CD28 antibodies (costimulatory signal for T cell activation) exhibited an obvious increase of CD3+ T cells and IFN-γ secretion. However, in the presence of PD-L1, the activated effect can be heavily inhibited. Interestingly, the inhibition effect caused by the PD-L1 protein could be relieved after the simultaneous introduction of BMS-202 at a low concentration. The results indicated that BMS-202 was capable of relieving the immunosuppressive effect of PD-L1 on activated T cells, thereby restoring proliferation and activation of T cells. Figure 4. Characterization of CHI/BMS-202@lipF. (A) Water solubility of CHI in free or CHI−F127 complex form. (B) EE and LE of CHI and BMS-202 in CHI@lipF, BMS-202@lipF, and CHI/BMS-202@lipF. (C) Particle size, polydispersity, and zeta potential of CHI/BMS-202@lipF, which were measured by DLS. (D) TEM image of CHI/BMS-202@lipF, the scale bar is 0.5 μm. (E) Stability of CHI/BMS-202@lipF in the media of ddH2O, PBS, and FBS. (F,G) Drug release behaviors of CHI and BMS-202 from CHI/BMS-202@lipF in different pH media, n = 3. Data are shown as means ± SD. ***p < 0.001. Besides the PD-1/PD-L1 blockade activity, BMS-202 was also reported to induce tumor cell death directly.31 As shown in Figure S1, BMS-202 inhibited cell survival in a dose- dependent manner after being incubated with 4T1 cells for 24 h. The IC50 value was 5.2 ± 0.4 μg/mL. Collectively, BMS-202 could possess dual antitumor effects, including PD-1/PD-L1 blockade activity and direct cytotoXicity. 3.6. Preparation and Characterization of Liposomes. The above results have demonstrated that CHI is capable of regulating tumor immunogenicity, boosting antitumor im- munity, and up-regulating the expression of immunosuppres- sive PD-L1 on cancer cells. The up-regulated PD-L1 could sensitize the tumor to BMS-202-mediated PD-L1 blockade therapy. Therefore, the combination of CHI and BMS-202 is a synergetic antitumor strategy worth exploring. Both CHI and BMS-202 are poorly water-soluble compounds, especially CHI. To improve their solubility, CHI and BMS-202 were co- delivered through a liposome delivery system. It was found through early exploration via the conventional film-hydration method that the EE of CHI in liposomes was low (<20%). This might be ascribed to the fact that the log P value of CHI (∼2.5) was outside the scope of the drug type suitable for liposome encapsulation (log P < −0.3 or log P > 4.5).48
First, we increased the water solubility of CHI from 0.00236 to 3.75 mg/mL by forming a complex with F127 (CHI-f127), which enhanced the water solubility nearly 1600-fold compared with free CHI (Figure 4A, Table S1). Subsequently, the CHI−F127 complex together with BMS-202 was encapsulated into liposome via the reverse evaporation method. The CHI−F127 complex tended to distribute into hydrophilic cavity, while BMS-202 was inserted into the hydrophobic phospholipid layer. The resultant drug-loaded liposome CHI/BMS-202@lipF displayed the superior encap- sulation of CHI and BMS-202 with EE of 85.3 ± 2.8 and 98.3 ± 3.1%, respectively (Figure 4B). DLS and TEM examination exhibited that the hydrodynamic diameter and surface charge were, respectively, 144.7 ± 2.4 nm and −31.04 ± 2.31 mV with a spheroid morphology for CHI/BMS-202@lipF (Figure 4C,D). In addition, the particle size of CHI/BMS-202@lipF exhibited negligible changes in the presence of salt and serum by DLS determination, which implies favorable stability for further application in vivo (Figure 4E). The drug release behaviors of CHI and BMS-202 in CHI/BMS-202@lipF were investigated in the environment of different pH values (pH 5.5, 6.5, and 7.4), which mimicked the environment of blood, a tumor matriX, and the endosome of tumor cells, respectively. As shown in Figure 4F,G, the release of BMS-202 showed moderate pH dependence, while a negligible impact was observed for CHI. It might be attributed to the amino on BMS-202, which could be protonated under low pH conditions. In addition, the release rates of CHI and BMS- 202 were both relatively gentle without burst release, which contributed to relieving the systemic toXicity. Collectively, CHI/BMS-202@lipF realized effective drug encapsulation with good stability and biocompatibility.

Figure 5. Biodistribution of liposomes. (A) Representative fluorescence images of mice at different time points; the red circle shows the site of the tumor. (B) Average signal quantification of a tumor site in vivo, n = 3. (C) Representative fluorescence images and (D) the semi-quantitative result of major organs and tumor tissues collected from mice at different time points, n = 3. Data are shown as means ± SD.

3.7. In Vivo Biodistribution Study. To verify the tumor accumulation and biodistribution of liposome in vivo, the 4T1- bearing mice were i.v. injected with free DiR or DiR@lipF at an identical dosage of DiR (0.5 mg/kg), followed by monitoring the fluorescence intensity of integral mice (in vivo) or excised tissues (ex vivo) using IVIS. As shown in Figure 5A,B, DiR@lipF exhibited considerable accumulation and a long retention time in tumor tissue compared to free DiR. Furthermore, the quantitative data of the excised tissues revealed that DiR@lipF also demonstrated high distribution in the liver and spleen except for the tumor, which might contribute to regulating the function of immune cells in the spleen (an immune organ) owing to the immunomodulatory activity of CHI and BMS-202 (Figure 5C,D).

3.8. In Vivo Antitumor Effect against Murine Orthotopic Breast Cancer. The antitumor formulation of CHI, as an HDACi, was mainly administrated orally or intraperitoneally, which usually needed a high dosage such as 25 mg/kg.49 In this study, we achieved a decreased dosage for CHI through a liposome delivery system via systematic administration, which was expected to attenuate the potential
toXicity. To investigate the dose-dependent antitumor efficiency, 4T1 tumor-bearing mice were i.v. injected with CHI@lipF at different CHI doses. As shown in Figure S2A, CHI@lipF could inhibit tumor growth to different extents. A photograph of the excised tumor and tumor weight also exhibited a similar tendency (Figure S2B,C). The tumor inhibitory rates were 48.5, 52.5, 71.9, and 77.3% when the doses were 1, 2, 5, and 10 mg/kg, respectively. Moreover, the body weight loss in all groups was negligible, which indicated good biosecurity of liposome-mediated drug delivery (Figure S2D). It should be noted that there was no significant improvement in the antitumor efficacy in the 10 mg/kg group compared with the 5 mg/kg group. Considering the cost, potential toXicity, and low-dose chemotherapeutics-mediated immunologic adjuvant effect, 5 mg/kg of CHI was chosen for subsequent studies. Afterward, H&E staining of the tumor tissue indicated that CHI@lipF was capable of inducing tumor cell apoptosis and necrosis (Figure S3). Furthermore, the IHC analysis exhibited an obvious up-regulation of PD-L1 in the tumor after being treated with CHI@lipF in a dose-dependent manner, which could sensitize a tumor to BMS-202-mediated PD-L1 blockade therapy.

To validate whether the combination strategy of HDACi and anti-PD-1/PD-L1 therapy could synergistically inhibit tumor growth, a treatment study was performed (Figure 6A). Briefly, the tumor-bearing mice were injected with saline, CHI@lipF (CHI, 5 mg/kg), BMS-202@lipF (BMS-202, 5 mg/ kg), or CHI/BMS-202@lipF (CHI, 5 mg/kg; BMS-202, 5 mg/ kg) when the tumor grew to 30−50 mm3. As shown in Figure 6B,C, the CHI/BMS-202@lipF-treated group exhibited a higher tumor growth suppression advantage over individual therapeutic-treated groups. In addition, the CHI/BMS-202@ lipF group did not exhibit significant body weight loss and normal organ damage, which indicated the good biosafety of CHI/BMS-202@lipF (Figures 6D and S4). The average tumor volumes of the CHI/BMS-202@lipF group at day 24 were 3.86-, 2.21-, and 1.61-fold smaller than those of saline, BMS- 202@lipF, and CHI@lipF, respectively. Furthermore, the median survival time of tumor-bearing mice treated with CHI/BMS-202@lipF was prolonged to 47 days, which was longer than that of other groups, including saline (30 days), BMS-202@lipF (38 days), and CHI@lipF (42 days) (Figure 6E). The examination of H&E, TUNEL, and cleaved-caspase 3 revealed that the drug-treated groups were capable of inducing tumor apoptosis and necrosis, in which CHI/BMS-202@lipF was most notable (Figures 6F and S5). The IHC analysis of PD-L1 showed that the up-regulated PD-L1 expression of tumor in CHI-treated mice was relieved by BMS-202. Lungs were harvested to investigate spontaneous metastasis of the 4T1 orthotopic tumor. Obviously, invasive lung tumor nodules were observed in groups of saline and BMS-202, but they were significantly decreased in CHI@lipF and CHI/BMS-202@lipF, which might be attributed to the anti-metastasis activity of CHI through up-regulating E-cadherin and down-regulating epithelial−mesenchymal transition (EMT).

Figure 6. CHI/BMS-202@lipF-mediated antitumor effect in the orthotopic 4T1 tumor model. (A) Treatment schedule for CHI/BMS-202@lipF- mediated antitumor combination therapy. (B) Spaghetti curves, (C) average tumor volume, and (D) body weight change of mice with different treatments (n = 8). (E) Survival rate of mice with various treatments (n = 8). (F) H&E and PD-L1 expression analysis of tumor and H&E analysis of lungs in mice with different treatments, the scale bar is 200, 100, and 250 μm, respectively. Data are expressed as means ± SD. *p < 0.05, **p < 0.01. Figure 7. Mechanism study of CHI/BMS-202@lipF-mediated antitumor effect in an orthotopic 4T1 tumor model. (A) Representative flow cytometric images and (B) quantification results of CD80+CD86+ in CD11c+ cells in a tumor (n = 3). (C) Representative flow cytometric images and (D) quantification of CD3+CD8+ T cells in a tumor (n = 3). (E) Representative flow cytometric images and (F) quantification of CD3+CD4+ T cells in a tumor (n = 3). (G) Representative flow cytometric images and (H) quantification of CD3−DX5+ NK cells in a tumor (n = 3). (I) Representative flow cytometric images and (J) quantification of activated NK cells (DX5+CD69+ in CD3−) in a tumor (n = 3). Data are shown as means ± SD. *p < 0.05, **p < 0.01. 3.9. Immunological Mechanism of the CHI/BMS-202@ lipF-Mediated Antitumor Effect in Murine Orthotopic Breast Cancer. The influence of epigenetic alteration on cancer immunology has been extensively discussed else- where.51 As a predominant category of epigenetic regulatory drug, HDACi has been demonstrated to sensitize tumors to intervention by immune cells, including NK cells, CD8+ T cells, DCs, and so forth.52 Besides the universal immunomo- dulatory capacity of HDACi, we also demonstrated that CHI could induce an ICD effect on tumor cells, which could further enhance the tumor immunogenicity and antitumor immune response. Furthermore, BMS-202-mediated PD-1/PD-L1 blockade therapy was capable of restoring the antitumor immunity of cytotoXic T lymphocytes (CTLs). Therefore, it is necessary to evaluate the underlying antitumor immune mechanism mediated by CHI/BMS-202@lipF in vivo. DCs, professional APCs, are responsible for activating and regulating the immune response. Immature DCs are capable of engulfing foreign substrates, such as tumor antigens and apoptotic cell fragments, and then migrate to tumor-draining lymph nodes (TDLNs) for maturation. Mature DCs are vital for antitumor immune response, which presents the tumor antigen to the T cell receptor for promoting activation and intratumoral infiltration of T cells. Therefore, after various treatments, the TDLNs and the tumor were harvested to examine t he po pulation of mature DCs (CD11c+CD80+CD86+). As shown in Figure 7A,B, the mature DCs in the tumor were increased to different extents after being treated with drug-loaded liposomes. Therefore, the CHI/BMS-202@lipF group exhibited an obvious advantage in the promotion of DC maturation over other groups, which was 1.81-, 1.38-, and 1.21-fold more efficient than saline, BMS-202@lipF, and CHI@lipF, respectively. The mechanism of BMS-202@lipF-promoted DC maturation may be related to PD-1/PD-L1 blockade.53,54 Additionally, owing to the enhancement of tumor immunogenicity after various treat- ments, the release of the tumor antigen and danger signals increased, which could promote the antigen capture by DCs and their subsequent migration to TDLNs. Figure S6 further confirms this process. Figure 8. CHI/BMS-202@lipF-mediated anti-metastasis effect in the 4T1 lung metastasis model. (A) Treatment schedule for CHI/BMS-202@lip. (B) Image of excised lung tissue (n = 6). (C) Number of lung nodules (n = 6). (D) H&E image of representative lung tissue chosen from B. The scale bar is 5 mm and 250 μm, respectively. Data are expressed as means ± SD. ***p < 0.001. The intratumoral infiltration of T cells is vital for immunotherapy. As shown in Figure 3B, in vitro studies have demonstrated that CHI could promote the expression of MHC-related molecules on tumor cells for enhancing the recognition and killing sensitivity of CTLs to tumor cells. In addition, a CHI-induced ICD response and a BMS-202- mediated PD-L1 blockade could promote innate and adaptive immunity. We investigated the population of CD4+ and CD8 T+ cells in the tissues of the spleen and tumor after different treatments. As shown in Figure 7C,D, CHI/BMS-202@lipF groups displayed an obvious increase of CD8+ T cells within a tumor, which was 2.42-, 1.20-, and 1.72-fold higher than that of saline, BMS-202@lipF, and CHI@lipF, respectively. Addition- ally, the intratumoral CD4+ T cells in the CHI/BMS-202@lipF group were 2.51-fold higher than that in the saline group, 1.79- fold higher than that in BMS-202@lipF, and 1.56-fold higher than that in CHI@lipF, respectively (Figure 7E,F). Further- more, similar tendencies of intratumoral CD4+ and CD8+ T cells were observed by CLSM examination (Figure S7). A systemic immune response was equally important for exerting antitumor immunity except for tumorous local immunity. As shown in Figure S8, CHI/BMS-202@lipF increased the proportion of CD4+ and CD8+ T cells in the spleen to 21.7 ± 1.0 and 8.1 ± 0.2%, which were 1.9- and 2.3-fold higher than that of the saline group, respectively. NK cells, as killer immune cells, play a vital role in antitumor immunity. Previous studies have shown that CHI sensitized a tumor to NK cell killing by increasing the expression of NKG2D on tumor cells.17 In addition, CHI was capable of directly promoting NK activation and proliferation as well. As shown in Figure 7G,H, the group treated with CHI/BMS- 202@lipF showed an obvious NK cell (CD3−DX5+) infiltration in the tumor, which was 2.41-, 1.81-, and 1.24- fold higher than that of saline, BMS-202@lipF, and CHI@lipF, respectively. Additionally, treating with CHI/BMS-202@lipF further increased the amount of activated NK cells (CD3−DX5+CD69+) in a tumor, which indicated the validity of the antitumor immune response (Figure 7I,J). A similar tendency was shown in the spleen except for the number of activated NK cells (Figure S9). Based on the above results, the potential antitumor process of CHI/BMS-202@lipF could be summarized as follows. First, CHI showed direct cytotoXicity and induced the ICD response within tumor cells. Second, CHI up-regulated the expression of MHC I, II on tumor cells, which contributed to the antigen presentation and T-cell recognition. Third, CHI could regulate the function of immune cells, such as promoting DC maturation, intratumoral infiltration and activation of T cells and NK cells, and so forth. At last, CHI promoted PD-L1 expression on tumor cells, while BMS-202 could induce PD-L1 protein dimerization to block the PD-1/PD-L1 interaction and subsequently restored the antitumor immunity of CTLs. Collectively, CHI/BMS-202@lipF could synergistically im- prove tumor immunogenicity, activate the immune system, relieve the immunosuppressive microenvironment, and finally realize effective antitumor efficacy. 3.10. CHI/BMS-202@lipF-Mediated Anti-metastasis Effect in a 4T1 Lung Metastasis Model. Metastasis has been known as the main cause of high cancer mortality. TNBC is severely aggressive, with a high tendency for distant spread and metastasis. Circulating tumor cells (CTCs) play a critical role during tumor metastasis, which spread to distant normal organs through blood circulation, followed by invading and depositing to form metastatic nodules.55 Normally, the CTCs could be recognized and eliminated by peripheral immune cells, such as NK cells, CTLs, and so forth. However, CTCs could further promote the expression of immunosuppressive receptors (CD47, PD-L1, etc.) to evade immunological surveillance. The expression level of PD-L1 on CTCs is higher than that of primary tumor cells, which indicates that CTCs might be more vulnerable to PD-L1 blockade therapy.56,57 Additionally, HDACi has been reported to down-regulate EMT for decreasing tumor invasiveness.50,58 Furthermore, HDACi is also capable of modulating the peripheral immune system to attack CTCs except for direct cytotoXicity. As we mentioned above, CHI/BMS-202@lipF could inhibit spontaneous tumor metastasis on the 4T1 orthotopic breast cancer model. To further clarify the potential of CHI/BMS- 202@lipF for the inhibition of CTCs and metastatic cancer, the establishment of a 4T1 lung metastatic model and treatment schedule was performed as described in Figure 8A. After being treated with various formulations for four rounds, the lungs were collected and fiXed to count the metastasizing tumor nodules and for H&E examination. As shown in Figure 8B,C, a number of nodules were observed on the surface of the lungs in the saline group with 181.7 ± 21.9, which indicated the successful establishment of the metastasizing tumor model. Nevertheless, after being treated with BMS-202@lipF, CHI@ lipF, and CHI/BMS-202@lipF, the metastasizing nodules could be dramatically decreased to 23.5 ± 7.3, 12.7 ± 4.3, and 8.2 ± 3.5, respectively. Additionally, a similar tendency was observed in H&E examinations of the lungs (Figure 8D). In summary, CHI/BMS-202@lipF revealed a favorable metastasis suppression ability. 4. CONCLUSIONS In summary, we presented a combined approach of epigenetic regulation and PD-1/PD-L1 blockade therapy through lip- osomes against tumor growth and metastasis. The liposomes realized effective drug encapsulation with well stability, biocompatibility, and security, which could effectively accu- mulate into a tumor and show long retention after systematic administration. Additionally, the released CHI was capable of inducing an ICD effect on the tumor to elicit an anticancer immune response. The combination with BMS-202 further enhanced T cell-mediated antitumor immunity and effectively inhibited the growth of primary tumors and metastatic tumors. This strategy might provide a referential and alternative approach for clinical antitumor therapy based on HDACi- combined immunotherapy. ■ ASSOCIATED CONTENT *sı Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.1c08290. Cell viability of 4T1 cells treated with BMS-202 for 24 h; antitumor effect of CHI@lipF in 4T1 tumor-bearing mice; analysis of H&E and PD-L1 in 4T1 tumor tissue treated with CHI@lipF; H&E examination of major organs in mice treated with CHI/BMS-202@lipF; TUNEL and cleaved-caspase 3 expression analysis of a tumor in 4T1 tumor-bearing mice with different treatments; flow cytometric analysis of DC maturation in TDLNs treated with CHI/BMS-202@lipF; IF analysis of CD4+ and CD8+ T cells in tumor tissue treated with CHI/BMS-202@lipF; flow cytometric analysis of CD4+ and CD8+ T cells in the spleen treated with CHI/BMS-202@lipF; flow cytometric analysis of NK cells in the spleen treated with CHI/BMS-202@ lipF; and EE and LE of CHI in the CHI−F127 complex (PDF) ■ AUTHOR INFORMATION Corresponding Author Zhiping Zhang − Department of Oncology, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030 Hubei, China; National Engineering Research Center for Nanomedicine and Hubei Engineering Research Center for Novel Drug Delivery System, Huazhong University of Science and Technology, Wuhan 430030, China; orcid.org/0000-0002-9235-5321; Phone: +86-027-83601832; Email: zhipingzhang@ mail.hust.edu.cn Authors Kun Tu − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Yulin Yu − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Yi Wang − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Ting Yang − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Qian Hu − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Xianya Qin − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Jingyao Tu − Department of Oncology, Tongji Hospital, Huazhong University of Science and Technology, Wuhan 430030, China Conglian Yang − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Li Kong − Tongji School of Pharmacy, Tongji Medical College, Huazhong University of Science and Technology, Wuhan 430030, China Complete contact information is available at: https://pubs.acs.org/10.1021/acsami.1c08290 Author Contributions #K.T. and Y.Y. contributed equally to this research. Notes The authors declare no competing financial interest. ■ ACKNOWLEDGMENTS We would like to thank the program for the HUST Academic Frontier Youth Team (2018QYTD13) and the National Natural Science Foundation of China (81872810) for support to this work. In addition, we also appreciate the Tongji Medical College of HUST and the Analytical and Testing Center of HUST for the help with TEM characterization, CLSM, and FACS analysis. ■ REFERENCES (1) Dawood, S. Triple-Negative Breast Cancer Epidemiology and Management Options. Drugs 2010, 70, 2247−2258. (2) Katz, H.; Alsharedi, M. Immunotherapy in Triple-Negative Breast Cancer. Med. Oncol. 2017, 35, 13. (3) Shi, J.; Liu, F.; Song, Y. Progress: Targeted Therapy, Immunotherapy, New Chemotherapy Strategies in Advanced Triple- Negative Breast Cancer. Canc. Manage. Res. 2020, 12, 9375−9387. (4) Kagihara, J. A.; Andress, M.; Diamond, J. R. Nab-Paclitaxel and Atezolizumab for the Treatment of PD-L1-Positive, Metastatic Triple- Negative Breast Cancer: Review and Future Directions. Expert Rev. Precis. Med. Drug Dev. 2020, 5, 59−65. (5) Postow, M. A.; Callahan, M. K.; Wolchok, J. D. Immune Checkpoint Blockade in Cancer Therapy. J. Clin. Oncol. 2015, 33, 1974−1982. (6) Zhao, X.; Subramanian, S. Intrinsic Resistance of Solid Tumors to Immune Checkpoint Blockade Therapy. Cancer Res. 2017, 77, 817−822. (7) O’Donnell, J. S.; Teng, M. W. L.; Smyth, M. J. Cancer Immunoediting and Resistance to T Cell-Based Immunotherapy. Nat. Rev. Clin. Oncol. 2019, 16, 151−167. (8) Song, X.; Zhou, Z.; Li, H.; Xue, Y.; Lu, X.; Bahar, I.; Kepp, O.; Hung, M.-C.; Kroemer, G.; Wan, Y. Pharmacologic Suppression of B7-H4 Glycosylation Restores Antitumor Immunity in Immune-Cold Breast Cancers. Canc. Discov. 2020, 10, 1872−1893. (9) Telli, M. L.; Stover, D. G.; Loi, S.; Aparicio, S.; Carey, L. A.; Domchek, S. M.; Newman, L.; Sledge, G. W.; Winer, E. P. Homologous Recombination Deficiency and Host Anti-Tumor Immunity in Triple-Negative Breast Cancer. Breast Canc. Res. Treat. 2018, 171, 21−31. (10) Pfirschke, C.; Engblom, C.; Rickelt, S.; Cortez-Retamozo, V.; Garris, C.; Pucci, F.; Yamazaki, T.; Poirier-Colame, V.; Newton, A.; Redouane, Y.; Lin, Y.-J.; Wojtkiewicz, G.; Iwamoto, Y.; Mino- Kenudson, M.; Huynh, T. G.; Hynes, R. O.; Freeman, G. J.; Kroemer, G.; Zitvogel, L.; Weissleder, R.; Pittet, M. J. Immunogenic Chemotherapy Sensitizes Tumors to Checkpoint Blockade Therapy. Immunity 2016, 44, 343−354. (11) Kroemer, G.; Galluzzi, L.; Kepp, O.; Zitvogel, L. Immunogenic Cell Death in Cancer Therapy. Annu. Rev. Immunol. 2013, 31, 51−72. (12) Krysko, D. V.; Garg, A. D.; Kaczmarek, A.; Krysko, O.; Agostinis, P.; Vandenabeele, P. Immunogenic Cell Death and DAMPs in Cancer Therapy. Nat. Rev. Canc. 2012, 12, 860−875. (13) Xia, T. Critical Role of Immunogenic Cell Death in Cancer Therapy. Sci. Bull. 2017, 62, 1427−1429. (14) Zhou, J.; Wang, G.; Chen, Y.; Wang, H.; Hua, Y.; Cai, Z. Immunogenic Cell Death in Cancer Therapy: Present and Emerging Inducers. J. Cell Mol. Med. 2019, 23, 4854−4865. (15) Schrump, D. S. CytotoXicity Mediated by Histone Deacetylase Inhibitors in Cancer Cells: Mechanisms and Potential Clinical Implications. Clin. Canc. Res. 2009, 15, 3947−3957. (16) Martínez-Iglesias, O.; Ruiz-Llorente, L.; Sánchez-Martínez, R.; García, L.; Zambrano, A.; Aranda, A. Histone Deacetylase Inhibitors: Mechanism of Action and Therapeutic Use in Cancer. Clin. Transl. Oncol. 2008, 10, 395−398. (17) Ning, Z.-Q.; Li, Z.-B.; Newman, M. J.; Shan, S.; Wang, X.-H.; Pan, D.-S.; Zhang, J.; Dong, M.; Du, X.; Lu, X.-P. Chidamide (CS055/HBI-8000): A New Histone Deacetylase Inhibitor of the Benzamide Class with Antitumor Activity and the Ability to Enhance Immune Cell-Mediated Tumor Cell CytotoXicity. Canc. Chemother. Pharmacol. 2012, 69, 901−909. (18) Lu, X.; Ning, Z.; Li, Z.; Cao, H.; Wang, X. Development of Chidamide for Peripheral T-Cell Lymphoma, the First Orphan Drug Approved in China. Intractable Rare Dis. Res. 2016, 5, 185−191. (19) Jiang, Z.; Li, W.; Hu, X.; Zhang, Q.; Sun, T.; Cui, S.; Wang, S.; Ouyang, Q.; Yin, Y.; Geng, C.; Tong, Z.; Cheng, Y.; Pan, Y.; Sun, Y.; Wang, H.; Ouyang, T.; Gu, K.; Feng, J.; Wang, X.; Wang, S.; Liu, T.; Gao, J.; Cristofanilli, M.; Ning, Z.; Lu, X. Tucidinostat Plus EXemestane for Postmenopausal Patients with Advanced, Hormone Receptor-Positive Breast Cancer (ACE): a Randomised, Double- Blind, Placebo-Controlled, Phase 3 Trial. Lancet Oncol. 2019, 20, 806−815. (20) Gong, K.; Xie, J.; Yi, H.; Li, W. CS055 (Chidamide/HBI-8000), a Novel Histone Deacetylase Inhibitor, Induces G (1) Arrest, ROS- Dependent Apoptosis and Differentiation in Human Leukaemia Cells. Biochem. J. 2012, 443, 735−746. (21) Zhou, H. X.; Han, L.; Wang, H.; Wei, J.; Guo, Z. G.; Li, Z. H. Chidamide Inhibits Glioma Cells by Increasing OXidative Stress via the miRNA-338-5p Regulation of Hedgehog Signaling. Oxid. Med. Cell. Longevity 2020, 8, 1−17. (22) Xie, Y.; Tang, P.; Xing, X.; Zhao, Y.; Cao, S.; Liu, S.; Lu, X.; Zhong, L. In Situ EXploring Chidamide, a Histone Deacetylase Inhibitor, Induces Molecular Changes of Leukemic T-Lymphocyte Apoptosis Using Raman Spectroscopy. Spectrochim. Acta Mol. Biomol. Spectrosc. 2020, 241, 118669. (23) Liu, L.; Zhang, J.; Zhang, X.; Cheng, P.; Liu, L.; Huang, Q.; Liu, H.; Ren, S.; Wei, P.; Wang, C.; Dou, C.; Chen, L.; Liu, X.; Zhang, H.; Chen, M. HMGB1: an Important Regulator of Myeloid Differ- entiation and Acute Myeloid Leukemia as well as a Promising Therapeutic Target. J. Mol. Med. 2021, 99, 107−118. (24) Rizvi, N. A.; Hellmann, M. D.; Snyder, A.; Kvistborg, P.; Makarov, V.; Havel, J. J.; Lee, W.; Yuan, J.; Wong, P.; Ho, T. S.; Miller, M. L.; Rekhtman, N.; Moreira, A. L.; Ibrahim, F.; Bruggeman, C.; Gasmi, B.; Zappasodi, R.; Maeda, Y.; Sander, C.; Garon, E. B.; Merghoub, T.; Wolchok, J. D.; Schumacher, T. N.; Chan, T. A. Mutational Landscape Determines Sensitivity to PD-1 Blockade in Non-Small Cell Lung Cancer. Science 2015, 348, 124−128. (25) Pardoll, D. M. The Blockade of Immune Checkpoints in Cancer Immunotherapy. Nat. Rev. Canc. 2012, 12, 252−264. (26) Postow, M. A.; Sidlow, R.; Hellmann, M. D. Immune-Related Adverse Events Associated with Immune Checkpoint Blockade. N. Engl. J. Med. 2018, 378, 158−168. (27) Geng, Q.; Jiao, P.; Jin, P.; Su, G.; Dong, J.; Yan, B. PD-1/PD-L1 Inhibitors for Immuno-Oncology: from Antibodies to Small Molecules. Curr. Pharm. Des. 2018, 23, 6033−6041. (28) Zhan, M.-M.; Hu, X.-Q.; Liu, X.-X.; Ruan, B.-F.; Xu, J.; Liao, C. From Monoclonal Antibodies to Small Molecules: the Development of Inhibitors Targeting the PD-1/PD-L1 Pathway. Drug Discov. Today 2016, 21, 1027−1036. (29) Zarganes-Tzitzikas, T.; Konstantinidou, M.; Gao, Y.; Krzemien, D.; Zak, K.; Dubin, G.; Holak, T. A.; Dömling, A. Inhibitors of Programmed Cell Death 1 (PD-1): a Patent Review (2010-2015). Expert Opin. Ther. Pat. 2016, 26, 973−977.Stimulates DCs and Reverts PEL-Induced Immune Suppression.Oncotarget 2015, 6, 29543−29554. (46) Murshid, A.; Gong, J.; Calderwood, S. K. The Role of Heat Shock Proteins in Antigen Cross Presentation. Front. Immunol. 2012, 3, 63. (47) Obeid, M.; Tesniere, A.; Ghiringhelli, F.; Fimia, G. M.; Apetoh, L.; Perfettini, J.-L.; Castedo, M.; Mignot, G.; Panaretakis, T.; Casares, N.; Métivier, D.; Larochette, N.; van Endert, P.; Ciccosanti, F.; (30) Zak, K. M.; Grudnik, P.; Guzik, K.; Zieba, B. J.; Musielak, B.;Piacentini, M.; Zitvogel, L.; Kroemer, G. Calreticulin EXposure Dömling, A.; Dubin, G.; Holak, T. A. Structural Basis for Small Molecule Targeting of the Programmed Death Ligand 1 (PD-L1). Oncotarget 2016, 7, 30323−30335. (31) Ashizawa, T.; Iizuka, A.; Tanaka, E.; Kondou, R.; Miyata, H.; Maeda, C.; Sugino, T.; Yamaguchi, K.; Ando, T.; Ishikawa, Y.; Ito, M.; Akiyama, Y. Antitumor Activity of the PD-1/PD-L1 Binding Inhibitor BMS-202 in the Humanized MHC-Double Knockout NOG Mouse. Biomed. Res. 2019, 40, 243−250. (32) Shang, Q.; Zhou, S.; Jiang, Y.; Wang, D.; Wang, J.; Song, A.; Luan, Y. Rational Design of a Robust Antibody-Like Small-Molecule Inhibitor Nanoplatform for Enhanced Photoimmunotherapy. ACS Appl. Mater. Interfaces 2020, 12, 40085−40093. (33) Zhang, R.; Zhu, Z. Y.; Lv, H. Y.; Li, F. T.; Sun, S. Q.; Li, J.; Lee, C. S. Immune Checkpoint Blockade Mediated by a Small-Molecule Nanoinhibitor Targeting the PD-1/PD-L1 Pathway Synergizes with Photodynamic Therapy to Elicit Antitumor Immunity and Anti- metastatic Effects on Breast Cancer. Small 2019, 15, 1903881. (34) Hu, M.; Zhang, J.; Yu, Y.; Tu, K.; Yang, T.; Wang, Y.; Hu, Q.; Kong, L.; Zhang, Z. Injectable Liquid Crystal Formation System for Reshaping Tumor Immunosuppressive Microenvironment to Boost Antitumor Immunity: Postoperative Chemoimmunotherapy. Small 2020, 16, 2004905. (35) Hu, Q.; Shang, L. H.; Wang, M. M.; Tu, K.; Hu, M.; Yu, Y. L.; Xu, M. W.; Kong, L.; Guo, Y. Y.; Zhang, Z. P. Co-Delivery of Paclitaxel and Interleukin-12 Regulating Tumor Microenvironment for Cancer Immunochemotherapy. Adv. Healthcare Mater. 2020, 9, 1901858. (36) Kotsafti, A.; Scarpa, M.; Castagliuolo, I.; Scarpa, M. Reactive OXygen Species and Antitumor Immunity-from Surveillance to Evasion. Cancers 2020, 12, 1748. (37) Inoue, S.; Riley, J.; Gant, T. W.; Dyer, M. J. S.; Cohen, G. M. Apoptosis Induced by Histone Deacetylase Inhibitors in Leukemic Cells is Mediated by Bim and NoXa. Leukemia 2007, 21, 1773−1782. (38) Inoue, S.; Walewska, R.; Dyer, M. J. S.; Cohen, G. M. Downregulation of Mcl-1 Potentiates HDACi-Mediated Apoptosis in Leukemic Cells. Leukemia 2008, 22, 819−825. (39) Reed, J.; Jurgensmeier, J.; Matsuyama, S. Bcl-2 Family Proteins and Mitochondria. Biochim. Biophys. Acta Bioenerg. 1998, 1366, 127− 137. (40) Martinou, J.-C.; Youle, R. J. Mitochondria in Apoptosis: Bcl-2 Family Members and Mitochondrial Dynamics. Dev. Cell 2011, 21, 92−101. (41) Jiang, C. C.; Lucas, K.; Avery-Kiejda, K. A.; Wade, M.; deBock, C. E.; Thorne, R. F.; Allen, J.; Hersey, P.; Zhang, X. D. Up-Regulation of Mcl-1 is Critical for Survival of Human Melanoma Cells upon Endoplasmic Reticulum Stress. Cancer Res. 2008, 68, 6708−6717. (42) Verfaillie, T.; van Vliet, A.; Garg, A. D.; Dewaele, M.; Rubio, N.; Gupta, S.; de Witte, P.; Samali, A.; Agostinis, P. Pro-Apoptotic Signaling Induced by Photo-OXidative ER Stress is Amplified by NoXa, not Bim. Biochem. Biophys. Res. Commun. 2013, 438, 500−506. (43) Rufo, N.; Garg, A. D.; Agostinis, P. The Unfolded Protein Response in Immunogenic Cell Death and Cancer Immunotherapy. Trends Canc. 2017, 3, 643−658. (44) Wei, G.; Margolin, A. A.; Haery, L.; Brown, E.; Cucolo, L.; Julian, B.; Shehata, S.; Kung, A. L.; Beroukhim, R.; Golub, T. R. Chemical Genomics Identifies Small-Molecule MCL1 Repressors and BCL-XL as a Predictor of MCL1 Dependency. Canc. Cell 2012, 21, 547−562. (45) Granato, M.; Gilardini Montani, M. S.; Filardi, M.; Faggioni, A.; Cirone, M. Capsaicin Triggers Immunogenic PEL Cell Death,Dictates the Immunogenicity of Cancer Cell Death. Nat. Med. 2007, 13, 54−61. (48) Nii, T.; Ishii, F. Encapsulation Efficiency of Water-Soluble and Insoluble Drugs in Liposomes Prepared by the Microencapsulation Vesicle Method. Int. J. Pharm. 2005, 298, 198−205. (49) Zhang, W.; Niu, J.; Ma, Y.; Yang, X.; Cao, H.; Guo, H.; Bao, F.; Haw, A.; Chen, Y.; Sun, K. The Synergistic Antitumor Activity of Chidamide in Combination with Bortezomib on Gastric Cancer. OncoTargets Ther. 2020, 13, 3823−3837. (50) Lin, S.-H.; Wang, B.-Y.; Lin, C.-H.; Chien, P.-J.; Wu, Y.-F.; Ko, J.-L.; Chen, J. J. W. Chidamide Alleviates TGF-Beta-Induced Epithelial-Mesenchymal Transition in Lung Cancer Cell Lines. Mol. Biol. Rep. 2016, 43, 687−695. (51) Hogg, S. J.; Beavis, P. A.; Dawson, M. A.; Johnstone, R. W. Targeting the Epigenetic Regulation of Antitumour Immunity. Nat. Rev. Drug Discov. 2020, 19, 776−800. (52) Conte, M.; De Palma, R.; Altucci, L. HDAC Inhibitors as Epigenetic Regulators for Cancer Immunotherapy. Int. J. Biochem. Cell Biol. 2018, 98, 65−74. (53) Carenza, C.; Calcaterra, F.; Oriolo, F.; Di Vito, C.; Ubezio, M.; Della Porta, M. G.; Mavilio, D.; Della Bella, S. Costimulatory Molecules and Immune Checkpoints are Differentially EXpressed on Different Subsets of Dendritic Cells. Front. Immunol. 2019, 10, 1325. (54) Sun, N.-Y.; Chen, Y.-L.; Lin, H.-W.; Chiang, Y.-C.; Chang, C.- F.; Tai, Y.-J.; Chen, C.-A.; Sun, W.-Z.; Chien, C.-L.; Cheng, W.-F. Immune Checkpoint Ab Enhances the Antigen-Specific Anti-Tumor Effects by Modulating Both Dendritic Cells and Regulatory T Lymphocytes. Canc. Lett. 2019, 444, 20−34. (55) Yu, M. Metastasis Stemming from Circulating Tumor Cell Clusters. Trends Cell Biol. 2019, 29, 275−276. (56) Zeng, Z.; Shi, F.; Zhou, L.; Zhang, M. N.; Chen, Y.; Chang, X. J.; Lu, Y. Y.; Bai, W. L.; Qu, J. H.; Wang, C. P.; Wang, H.; Lou, M.; Wang, F. S.; Lv, J. Y.; Yang, Y. P. Upregulation of Circulating PD-L1/ PD-1 is Associated with Poor Post-Cryoablation Prognosis in Patients with HBV-Related Hepatocellular Carcinoma. PLoS One 2011, 6, No. e23621. (57) Mazel, M.; Jacot, W.; Pantel, K.; Bartkowiak, K.; Topart, D.; Cayrefourcq, L.; Rossille, D.; Maudelonde, T.; Fest, T.; AliX- Panabier̀es, C. Frequent EXpression of PD-L1 on Circulating Breast Cancer Cells. Mol. Oncol. 2015, 9, 1773−1782. (58) Su, Y.; Hopfinger, N. R.; Nguyen, T. D.; Pogash, T. J.; Santucci-Pereira, J.; Russo, J. Epigenetic Reprogramming of Epithelial Mesenchymal Transition in Triple Negative Breast Cancer Cells BMS202 with DNA Methyltransferase and Histone Deacetylase Inhibitors. J. Exp. Clin. Canc. Res. 2018, 37, 314.