Animals, cell culture, and chemicals

Female BALB/c, male BALB/c-nu, and male C57 mice (6–8 weeks old) were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd (Beijing, China). All mice were housed in a specific pathogen-free environment at the Animal Center of Huazhong University of Science and Technology, China. All animal experiments were approved by the Animal Care Committee of Tongji Medical College, Hua Zhong University of Science and Technology ([2018] IACUC number: 3106) and conducted in accordance with the approved protocol, the ARRIVE guidelines [31] and the National Institutes of Health guide for the care and use of laboratory animals.

Lewis lung carcinoma (LLC) and murine colorectal carcinoma (CT26) cells were purchased from the China Center for Type Culture Collection (Wuhan, China). LLC cells were cultured in DMEM and CT26 cells in RPMI 1640 medium, which were supplemented with 10% fetal bovine serum (FBS), l-glutamine (2 mM), pyruvate (1 mM), penicillin (100 U/mL), and streptomycin (100 µg/mL).

Cisplatin, dihydroartemisinin, and integrated stress response inhibitor (ISRIB) were purchased from Selleck (Shanghai, China), while thapsigargin was obtained from Sigma-Aldrich (St. Louis, MO, USA). According to the manufacturer’s instructions, these chemicals were dissolved in dimethyl sulfoxide (DMSO) or N, N-dimethylformamide (DMF), respectively.

Quantification of apoptosis and cell death

Apoptosis and cell death were assessed by annexin V-fluorescein isothiocyanate/7-amino actinomycin D (7-AAD) staining (BD Biosciences, San Jose, CA, USA). Briefly, LLC and CT26 cells were treated with control solvent, CDDP, DHA, or C + D at a concentration of 100/150 µM CDDP or/and 10 µM DHA for 24 h, and the cells were collected, washed, and resuspended in 1× binding buffer. Subsequently, the anti-annexin V antibody and 7-AAD were co-stained at 4 ℃ for 15 min in the dark. All samples were analyzed using a BD LSR Fortessa X-20 flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA).

ROS generation

According to the manufacturer’s instructions, ROS levels were measured using a ROS Assay Kit (Beyotime Biotechnology, China). The LLC and CT26 cells, treated as described above, were harvested and stained with 10 µM 2’, 7’-dichlorofluorescein-diacetate (DCFH-DA) for 20 min at 37 ℃ in the dark. The stained cells were washed with phosphate-buffered saline and analyzed using BD LSR Fortessa X-20 flow cytometry (Becton Dickinson, Franklin Lakes, NJ, USA).

Analysis of CALR exposure

LLC and CT26 cells, cultured in the presence of the control solvent, CDDP, DHA, or C + D for 6 h, underwent analysis of extracellular CALR expression by flow cytometry as previously described [32]. Briefly, cells were fixed and incubated with an anti-CALR antibody (Abcam, ab2907), followed by incubation with a secondary antibody (Abcam, ab150077) for thirty minutes. 7-AAD was added to each sample after washing and incubating at 4 ℃ in the dark. The stained samples were analyzed by flow cytometry. Sample incubated without the primary anti-CALR antibody was used as isotype control, and the percentage of CALR exposure on the live (7-AAD stained negative) cells was determined.

ATP assays

LLC and CT26 cells were cultured with four different treatments for 24 h. Cell supernatants were collected and centrifuged at 400 × g for 5 min followed by a second centrifugation at 2,000 × g for 10 min at 4 ℃. The supernatants were collected and deproteinized using perchloric acid and potassium hydroxide, and ATP levels were measured using an ATP colorimetric/fluorometric assay kit (Abcam, ab83355) according to the manufacturer’s instructions. Fluorescence was assessed using a FLUOstar OPTIMA multilabel reader (BMG LabTech, Germany).

Detection of HMGB1 release

After treatment of LLC and CT26 cells for 24 h, cell supernatants were collected and centrifuged as described above. HMGB1 levels in the cell supernatants were quantified using an HMGB1 ELISA kit (Shino Test Corporation, Japan) following the manufacturer’s instructions and detected on a FLUOstar OPTIMA multilabel reader (BMG LabTech, Germany).

Generation of bone marrow-derived dendritic cells

BMDCs were differentiated from bone marrow precursors according to previously published protocols [33]. Briefly, 2 × 106 bone marrow cells were cultured in the complete RPMI 1640 medium, supplemented with 10% heat-inactivated FBS and mGM-CSF (20 ng/mL, Pepro Tech, NJ, USA). On the third day, a fresh complete culture medium was added. On the sixth and eighth days, half of the cultural supernatant was refreshed with the complete medium. On the ninth day, the BMDCs in the culture supernatant were harvested for further analysis.

Phagocytosis assay

CT26 and LLC cells were labeled with 1 µM 1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate (DiL, Beyotime Biotechnology, China). After 20 h of treatment with control, DHA, CDDP, or C + D, the tumor cells were refreshed in a complete culture medium without drugs for 4 h. Subsequently, the cells were collected, washed, and co-cultured with 3,3’-dioctadecyloxacarbocyanine perchlorate (DiO, Beyotime Biotechnology, China)-labeled BMDCs or CD11c-labeled BMDCs for 14 h in a 1:2 ratio. The stained cells were visualized with the confocal laser scanning microscope (CLSM, Nikon-si-A1, Japan) or analyzed by flow cytometry.

Western blot immunoassay

Protein was extracted by radioimmunoprecipitation assay buffer supplemented with phosphatase and protease inhibitors (Thermo fisher Scientific, Waltham, MA, USA). The protein was denatured at 100 ℃ for 10 min after adding a “protein Loading buffer” (Boster Biological Technology, China). Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to the polyvinylidene difluoride membrane. After being blocked with 5% skim milk powder in tris-buffered saline with 0.1% tween 20 (TBST), membranes were incubated with the following primary antibodies overnight at 4 ℃: anti-p-eIF2α (Cell Signaling Technology, 3398, 1:1,000), anti-eIF2α antibody (Cell Signaling Technology, 5324, 1:1,000), and anti-PERK antibody (Cell Signaling Technology, 3194, 1:1,000). Membranes were washed with TBST and HRP-conjugated secondary antibodies were incubated at room temperature for 1 h. After 3 washes with TBST, blots were detected using electrochemiluminescence (Thermo fisher Scientific, Waltham, MA, USA). β-actin was used as an internal reference.

siRNA knockdown of EIF2ΑK3

According to the manufacturer’s instructions, 10 nM siGENOME mouse EIF2ΑK3 siRNA (Horizon, M-044901-01-0005) or siGENOME scrambled siRNA (Horizon, D-001206-13-05) was mixed with Lipofectamine™ RNAiMAX transfection reagent (Thermo fisher Scientific, Waltham, MA, USA) and diluted with Opti-MEM (Gibco, Grand Island, NY, USA). After treatment with non-coding or coding siRNA for 48 h, LLC and CT26 cells were seeded in appropriate plate with fresh culture medium for another 24 h, following different drug treatment.

Vaccination assay of dying CT26 cells in vivo

CT26 cells were cultured with control solvents or chemotherapeutic drugs for 20 h, followed by 4 h of fresh complete culture medium without chemotherapeutic drugs to eliminate the residual drugs. For inhibiting the phosphorylation of eIF2α, CT26 cells were transfected with EIF2AK3 siRNA or scramble siRNA for 48 h before chemotherapeutic drug administration. Next, BALB/c-nu or BALB/c mice were randomized into three groups and subcutaneously inoculated with 3 × 106 CT26 single tumor cell vaccines in 100 µl on the left flank, and an equal volume PBS without tumor cells was used as vaccine-negative control. All the tumor cell vaccines used in our vaccination model did not form primary tumors. One week later, untreated 5 × 105 live CT26 cells in 100 µl were injected into the right flank. Tumor incidence and growth was observed for at least 30 days. Tumor-bearing mice were then euthanized and the tumors were stored for immunohistochemical staining after fixing, embedding, and sectioning.

Histology and immunohistochemical (IHC) staining

Hematoxylin and eosin (H&E) staining was performed using standard procedures, and images were captured by light microscopy. Immunohistochemistry was conducted and analyzed as previously described [34]. Briefly, tumor tissue sections were dewaxed with xylene and rehydrated using different concentrations of ethanol. Antigen retrieval was achieved through a 20-minute microwave pretreatment in a sodium citrate solution, followed by cooling at room temperature. Endogenous peroxidase activity was blocked using 3% H2O2. After a 30-minute block using 5% BSA buffer, slides were incubated with primary anti-CD8 antibody (Abcam, ab209775, 1:50) or anti-Ki67 antibody (Servicebio, GB121141, 1:600), horseradish peroxidase-labeled secondary antibody, and diaminobenzidine substrate at the optimal concentration for IHC staining using an immunohistochemistry kit (Sangon Biotech, China), following the manufacturer’s instructions. Positive cells were counted in five fields on each slide using ImageJ software (http://imagej.nih.gov/ij/).

Anti-tumor efficacy in mice bearing established tumors

Mice were anesthetized with 1% pentobarbital sodium, and a total of 5 × 105 CT26 cells resuspended in 100 µl were subcutaneously inoculated into the right lower flank of BALB/c or BALB/c-nu mice. C57 mice were subcutaneously injected with 5 × 105 LLC cells to generate flank-localized tumors. On the eighth day after-tumor injection, mice were randomized into four groups and intraperitoneally dosed with the solvent or chemotherapeutic drugs (3 mg/kg CDDP, 40 mg/kg DHA) once every other day for a total of four doses. For the eIF2α phosphorylation inhibition, vector or 2.5 mg/Kg ISRIB dissolved in corn oil was intraperitoneal (i.p.) injected once a day for two days before the C + D treatment, and then with ISRIB alone or the chemotherapy combination every other day from the eighth day. Body weight was monitored during treatment; tumor growth was measured with a digital caliper; and tumor diameter was calculated using the formula: V = 1/2a × b2 (where ‘a’ is the tumor length and ‘b’ is the tumor width). On the 18th day, tumor-bearing BALB/c mice were euthanized, and tumor tissue was stripped, weighed, and stored for TME analysis. After fixing, embedding, and sectioning the tumors, tumor slices were used for H&E, IHC, and terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) staining. For survival analysis, mice were considered at the endpoint when tumor volume reached 2000 mm3.

TUNEL assay

Tumor tissue slices were treated using an in situ cell death detection kit (Roche, Switzerland) according to the manufacturer’s instructions. In this experiment, apoptotic cells were stained green and analyzed through CLSM (Nikon-si-A1, Japan).

Flow cytometry analysis for TME

To assess the proportion of immunocytes in the TME, flow cytometry analysis was performed as previously described [19]. Tumors were dissected and incubated in a buffer containing 2% FBS, DNAse I (0.02 mg/mL), and collagenase IV (2 mg/mL) (all from Sigma-Aldrich, St. Louis, MO, USA) for 1 h at 37 °C. Spleens and inguinal draining lymph nodes were homogenized. Red blood cells were lysed and the single-cell suspension was obtained through a 70 μm cell strainer. For surface staining, single cells were blocked with anti-CD16/CD32 antibodies and stained with the surface antibodies for 30 min at 4 ℃. Dead cells were excluded from all stained samples using the Zombie Aqua Fixable Viability Cell Staining Kit.

For intracellular detection of interferon-gamma (IFN-γ) in tumor-infiltrating lymphocytes (TILs), cells were cultured with phorbol myristate acetate (50 ng/mL, Sigma Aldrich), ionomycin (1 µg/mL, Sigma-Aldrich), and brefeldin A (1 µl/mL, BioLegend, San Diego, CA, USA) for 5 h at 37 °C. For intracellular staining, surface-labeled cells were fixed and permeabilized with Foxp3/Transcription Factor Staining Buffer Set (eBioscience, San Diego, CA, USA) and stained with intracellular antibodies. The stained cells were acquired using a BD LSR Fortessa X-20 flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA). The data were analyzed using FlowJo_V10 software. All antibodies were purchased from BD Pharmingen, eBioscience, or BioLegend; the details of these antibodies are shown in Table S1.

Statistical analysis

All statistical analyses were performed using GraphPad Prism software (v.8.0, San Diego, CA, USA). All statistics were two-tailed. The Shapiro–Wilk test was used to test whether the data followed a normal distribution, and Levene’s F test was used to test whether the data exhibited variance homogeneity. The data that do not follow the normal distribution were log-transformed to base 10 logarithms (log10) to approach a normal distribution.