Cells

NCI-H1975, HeLa, HCT116, B16F10, SUNE1, 4T1, THP-1, MDA-MB-231, L929, CT26, HL-60, Jurkat, and HEK-293T cell lines were obtained from ATCC. STING−/− HeLa cells were gifts from Dr. Zhengfan Jiang (Peking University),74 and ATG5−/− HeLa cells were gifts from Dr. Feng Shao (Beijing Institute of Biological Science).54STING−/− NCI-H1975 is a monoclonal cell line generated in our laboratory. NCI-H1975, B16F10, 4T1, MDA-MB-231, L929, CT26, HEK-293T, HeLa, and derived cells were cultured in DMEM (Gibco). HL-60, Jurkat, SUNE1, and THP-1 cells were cultured in RPMI 1640 medium (Gibco). HCT116 cells were cultured in McCoy’s 5A medium (Gibco) with 10% FBS (ExCell Bio) and 100 U/mL penicillin-streptomycin and were maintained in a humidified, 5% CO2 atmosphere at 37 °C. All cell lines used in this study were authenticated using short-tandem repeat profiling less than 6 months before the project was initiated and were not cultured for more than 1 month.

Mouse BMDMs were generated from Dr. Xiaojun Xia’s laboratory (Sun Yat-sen University Cancer Center (SYSUCC)). BMDMs were isolated from C57BL/6 mouse femurs and maintained in L929 conditioned medium (DMEM supplemented with 10% FBS, 10% L929 supernatant, and 100 U/mL penicillin-streptomycin) for 5 days. The cells in suspension were discarded, and adherent BMDMs were used in the following experiments.

Reagents

Cisplatin (S1166), Adriamycin (S1208), Hydroxyurea (S1896), Rucaparib (S1098), 5-Fluorouracil (S1209), DMXAA (S1537), BFA (S7046), Rapamycin (S1039), and 3-MA (S2767) were from Selleck. Doxycycline hyclate (D9891), Phenylarsine oxide (P3075), GTP-Agarose suspension (G9768), Protein A agarose (P3476), mouse monoclonal Anti-Flag M2 Affinity Gel (A2220), and mouse monoclonal Anti-HA-Agarose antibody (A2095) were from Sigma-Aldrich. DiABZI STING agonist-1 (Tautomerism) (HY-112921) and Digitonin (HY-N4000) were from MCE. HaloTag® Ligands (GA1110) for super-resolution microscopy were from Promega. 2'3'-cGAMP (tlrl-nacga23-1) was from Invivogen. Pierce™ Anti-DYKDDDDK Magnetic Agarose (A36797) was from Thermo Fisher Scientific.

Antibodies

Anti-TIM23 (NBP180680) was from Novus Biologicals. Anti-CD63 (ab134045), CD3 (ab16669), Atg7 (ab133528), STING (ab181125) and RAB22A (ab137093) were from Abcam. Anti-FLOT2 (3436), CD9 (13403), β-tubulin (2128), β-actin (4970), V5-Tag (13202), Histone H3 (4499), Flag (14793), Flag (8146), Fibronectin (26836), EEA1 (3288), cGAS (15102), HA (3724), HA (2367), Atg5 (12994), STING (13647), phospho-STING (Ser366) (19781), TBK1 (3504T), phospho-TBK1 (Ser172) (5483), phospho-IRF-3 (Ser396) (4947), CD8 (98941) and Atg16L1 (8089T) were from Cell Signaling Technology. Anti-FIP200 (GTX129093), WIPI2 (GTX132453) and PI4KCA (GTX107441) were from GeneTex. Anti-GM130 (610822) was from BD Biosciences. STING (19851-1-AP), Syntenin-1 (22399-1-AP), PI4KCB (13247-1-AP), p62 (18420-1-AP), IRF3 (11312-1-AP), GAPDH (10494-1-AP), Alix (12422-1-AP), SAC1 (13033-1-AP), and rabbit IgG (30000-0-AP) were from Proteintech Group. Anti-CD81 (sc-166029), TBC1D2B (sc-398906), PI4K2A (sc-390026), HSP70 (sc-24), calnexin (sc-11397), CD63 (sc-5275), and mouse IgG (sc-2025) were from Santa Cruz Biotechnology. Anti-STING (MA5-26030) was from Thermo Fisher Scientific. Anti-Tsg101 (HPA006161), RAB22A (HPA066920), LC3B (L7543), and IgG conjugated to 10-nm gold particles (G7777, G7402) were from Sigma-Aldrich. Secondary Antibody Alexa Fluor-594 (A-11037, A-11037), -568 (A-11036), -647 (A-21236), -488 (A-11034, A32723) were from Invitrogen. HRP-conjugated antibody (W4011, W4021) was from Promega.

Plasmids

The constitutively active RAB GTPase library, FLOT1-HA, FLOT2-HA, and SBP-RILP constructs were generated as previously described.18 GFP-LC3, GFP-RFP-LC3 constructs were gifts from Dr. Min Li (Sun Yat-sen University),75,76 HaloTag plasmids were gifts from Dr. Dong Li (Chinese Academy of Sciences),77 and the EGFP-PHOSBP construct was a gift from Shunji Jia (Tsinghua University).78 The cDNAs of RAB22A and STING were obtained by PCR and cloned into a pSIN vector with or without a tag (HA, Flag, V5, GFP, Halo, or mCherry). STINGV155M, STINGV155M/LIR467, STINGR281Q, and Atg16L1 (with K/R changed to A) constructs were generated by PCR using forward primers containing mutated sites step by step. L-Flag-STINGV155M-GFP was cloned by PCR from STINGV155M-GFP with a Flag-coding sequence inserted between aa 112 and aa 113. Truncation mutants of Atg16L1 (aa 211–230) were generated by PCR from WT Atg16L1. The fusion cassettes of GFP-FKBP-INPP5E, GFP-FKBP-SAC1, FRB-Flag-RAB22AQ64L, and 3× Flag-Atg16L1(211–230)-GFP were generated by PCR with a One Step cloning kit (Vazyme). Tet-on GFP-RAB22AQ64L was obtained by PCR and cloned into a pSIN vector with a tet-on transactivator. The shRNA expression constructs were inserted into a pLKO.1-puro backbone. Tet-on shRNA targeting RAB22A was cloned into a pLKO.1-puro backbone with a tet-on transactivator. sgRNA sequences were designed with the web application GUIDES (http://guides.sanjanalab.org/)79 and inserted into lentiCRISPRv2 plasmids. Sequences of sgRNA used in this study are shown in Supplementary information, Table S1.

All the mentioned constructs were fully verified by Sanger sequencing.

Transfection, lentivirus, and stable cell line construction

For transient transfection, cells were transfected with plasmids using polyethylenimine (PEI) (Polysciences) or Lipofectamine 3000. The culture medium was refreshed after 6 h. Cells were transfected with siRNAs (Ribobio) using RNAiMAX (Thermo Fisher), and the culture medium was refreshed after 6 h. The mRNA level of the target gene was quantified 72 h after transfection.

Lentiviral production for gene overexpression, sgRNA, and Cas9 expression was performed as follows. HEK-293T cells were co-transfected with 3 μg of lentiCRISPRv2-sgRNA/pSin-EF2-cDNA, 2 μg of psPAX2 (gag, pol), and 1 μg of pMD2G using 24 μL of PEI (2 mg/mL) after being seeded for 24 h. Viral supernatants were collected 48 h after transfection and filtered through 0.45-μm PVDF filters (Millipore). Cells in six-well plates were infected with a virus at appropriate viral titers in the presence of 10 μg/mL polybrene (Sigma) and centrifuged at 2000 rpm (800× g) for 60 min at 37 °C. Stable cell lines were selected using 0.5 μg/mL puromycin.

EVP isolation

EVP-free FBS was prepared by overnight ultracentrifugation at 100,000× g. Cells were washed twice with PBS and cultured in DMEM supplemented with 10% EVP-free FBS for 48 h. The supernatants of the cultured cells were collected and subjected to sequential centrifugation steps (600× g for 10 min and 2000× g for 30 min) at 4 °C to eliminate cells and cellular debris. Next, the supernatants were transferred to new tubes and centrifuged at 100,000× g for 2 h to precipitate the EVPs. All pellets were washed in PBS and recentrifuged at the same speed before being resuspended in sterile PBS. The concentrations of EVPs were quantitated by BCA assay, and the EVPs were stored at –80 °C.

Animal experiments

Animal care and experiments were performed in strict accordance with the Guide for the Care and Use of Laboratory Animals and the Principles for the Utilization and Care of Vertebrate Animals and were approved by the Animal Research Committee of SYSUCC (Approval no. 19120D). Female BALB/c mice were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. 4T1 cells (5 × 105) were subcutaneously injected into the right inguinal area of BALB/c mice (n = 6, aged 6 weeks). Six days post implantation, the mice were randomly assigned into five treatment groups: the PBS, –diABZI EVP, +diABZI EVP, STING−/−–diABZI EVP, and STING−/−+diABZI EVP groups. Mice were administered 200 μL of PBS, 5 µg of EVPs derived from WT or STING−/− 4T1 cells treated with or without 10 µM diABZI three times a week via tail vein injection. Tumor sizes were measured by calipers three times a week. Tumor volume was calculated using the formula length × width2/2 (mm3). Mice were sacrificed 19 days after implantation, and then, the tumors were dissected and evaluated.

Vector, RAB22A, shNC, or shRNA-PI4K2A stable 4T1 cells (5 × 105) were injected into the mammary fat pads through surgery in BALB/c mice. Six days post implantation, mice were treated with or without 1.5 mg/kg (vector vs RAB22A) or 3.0 mg/kg (shNC vs shRNA-PI4K2A) of diABZI three times per week by tail intravenous injection, as indicated, for another 14 days. Tumor size was measured using calipers every other day. Tumor volume was calculated using the formula length × width2/2 (mm3). The mice were sacrificed 20 days after implantation, and then, the tumors were dissected and evaluated.

Fluorescence-activated cell sorting

For analysis of cytotoxic T cell infiltration in mouse tumors, mouse tumors were digested at 37 °C at 100 rpm for 1 h in digestion buffer (RPMI 1640 medium containing 0.4 mg/mL Collagenase IV (Sigma-Aldrich) and 50 U/mL DNase I (Sigma-Aldrich). The digested cell suspension was filtered by a 70-μm cell strainer to obtain single cell suspension and washed with staining buffer (PBS containing 2% FBS) twice and PBS once. Lymphocytes were obtained from peripheral blood lymphocyte isolates (TBD, LTS1077). Then cells were stained with Zombie Aqua™ Fixable Viability Kit (bv510) (Biolegend, 423101) in PBS for 15 min to stain the non-viable cells. After that, cells were washed with staining buffer twice and stained with APC anti-CD45 (Biolegend, 103111), FITC anti-CD3ε (Biolegend, 100203), and PE/cy7 anti-CD8α (Biolegend, 100721) antibodies at 4 °C for 30 min. Flow cytometry analysis was performed after cells were washed twice with staining buffer.

Immunoblot and immunoprecipitation

For western blotting, the cells were washed once with cold PBS and then lysed on ice in RIPA buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, and 1% NP40) supplemented with Protease Inhibitor Cocktail Set I (Calbiochem; 539131) and Phosphatase Inhibitor Cocktail Set II (Calbiochem; 524625). The lysates were cleared by centrifugation at 12,000 rpm for 20 min at 4 °C. For immunoprecipitation assay, the antibody beads were washed three times with RIPA buffer. Subsequently, 20 μL of the beads were added and incubated with the lysates for 6 h at 4 °C. For endogenous immunoprecipitation assay, protein A or protein A/G agarose beads were washed three times with RIPA buffer, and then, an antibody or the control rabbit or mouse IgG was added to the cell lysates with the washed agarose beads, followed by incubation for 6 h at 4 °C. The beads were then washed five times with RIPA buffer. The immunoprecipitates and cell lysates were then boiled in gel loading buffer for 10 min and loaded onto sodium dodecyl sulfate-polyacrylamide gels. The proteins in the gels were transferred to Immobilon-P PVDF membranes (Millipore), which were then blocked in PBS with 5% nonfat milk and 0.1% Tween-20 and probed with primary antibodies overnight at 4 °C. Secondary HRP-conjugated antibodies were then added, and Clarity ECL substrate (Bio-Rad) or High-sig ECL substrate (Tanon) was used for detection by MiniChmei Chemiluminescence imager (SAGECREATION, Beijing).

Immunohistochemistry and immunofluorescence

Freshly collected tumors were fixed with 4% paraformaldehyde and embedded in paraffin before they were sectioned into 4-μm thickness. Then, the slices were subjected to deparaffinization, antigen retrieval, and blocking. For Immunohistochemistry analysis, the slices were incubated with anti-RAB22A antibody for 2 h at room temperature followed by incubation with anti-rabbit IgG secondary antibody and DAB reagent. The sections were viewed under a microscope (EVOS FL Auto Cell Imaging System, USA).

For immunofluorescence staining, cells were fixed with 4% paraformaldehyde for 15 min, further permeabilized with 0.5% Triton X-100 (Sigma-Aldrich) for 15 min, and blocked with goat serum (ZSGB-BIO, ZLI-9056) for 30 min at room temperature. After blocking, the cells were stained with primary antibody for 2 h at room temperature or overnight at 4 °C and washed three times with PBS. Then, the cells were stained with a secondary antibody for 1 h followed by Hoechst 33342 (Invitrogen, H3570) for 2 min. After staining, the cells were washed three times with PBS. Then, the cells were mounted with an antifade mounting medium (Beyotime, P0128M) and imaged by a confocal microscope (Zeiss LSM880 with Airyscan). Structured illumination microscopy (SIM) super-resolution images were taken using a Nikon N-SIM system with a 100× oil immersion objective lens, 1.49 NA (Nikon). Images were captured using Nikon NIS-Elements and were reconstructed using slice reconstruction in NIS-elements.

Live-cell imaging

Time-lapse images of autophagosomes/MVB-like structures fusing with early endosomes were taken by a Nikon Ti2 spinning disk microscope. HeLa cells stably expressing tet-on GFP-RAB22AQ64L were seeded on a 35-mm glass-bottom dish. Twenty-four hours later, they were transfected with either calnexin-mCherry or STINGV155M-Halo and treated with 50 ng/mL doxycycline for another 24 h. Before imaging, cells transiently expressing STINGV155M-Halo were pre-incubated with 30 nM HaloTag Ligands for 30 min at 37 °C, and then the culture medium was refreshed. Time-lapse movies were recorded using images captured at a 10–20 s interval.

Cell stimulation with 2’3’-cGAMP

Cells were treated with 2 µM 2’3’-cGAMP in digitonin permeabilization solution (50 mM HEPES, pH 7.0, 100 mM KCl, 3 mM MgCl2, 0.1 mM dithiothreitol, 85 mM sucrose, 0.2% BSA, 1 mM ATP, and 10 µg/mL digitonin) for 30 min at 37 °C. Cells were then incubated in a fresh medium for 1 h before western blot analysis or in an EVP-free medium for 48 h before EVP isolation.

Establishment of a heterozygous STINGV155M/WT knock-in HeLa cell line

The establishment of endogenous heterozygous STINGV155M/WT knock-in HeLa cell line was processed by an improved method based on homology-directed repair, as indicated before.80 Briefly, an efficient sgRNA (5’-AUCGAGAAAUGGGGGCAGAG-3’) targeting STING intron 3 was cloned into pLenti-CRISPR-V2. A donor DNA template containing cytomegalovirus promoter-driven DsRed-expressing cassette flanked by the left and right homology arms (HAs) was used. The desired V155M mutation was located on the left arm. Because the donor template carried an independent expressing cassette, the expression of dsRed does not affect STING expression. Donor template was amplified by PCR and purified. Then the sgRNA-Cas9 and donor template were co-transfected into HeLa cells. Six days later, dsRed-positive cells were sorted by flow cytometry, and positive clones were selected. cDNA of these cells was analyzed by Sanger sequencing to identify clones that have been successfully edited.

Quantitative reverse transcription (qRT)-PCR assays

mRNA was extracted with an RNA extraction kit (TIANGEN) according to the manufacturer’s instructions. cDNA for each sample was generated using 1 μg of total mRNA and RT SuperMix for qPCR (Vazyme). Then, quantitative PCR assays were performed using SYBR Color qPCR Master Mix (Vazyme) with a LightCycler 480 (Roche). For the details of the qRT-PCR primers used in this study, please see Supplementary information, Table S2.

TEM

Cells with confluence at 80%–90% were centrifugated, then the cell pellet was pre-fixed using 2.5% neutral glutaraldehyde (Ala Aesar, A17876) diluted in 0.1 M phosphate buffer for 2 h at 4 °C. After that, the pellet was rinsed with 0.1 M phosphate buffer six times for 30 min each. Then the pellet was post-fixed with 1% osmic acid (TED PELLA, 18456) diluted in 0.1 M phosphate buffer for 1 h at 4 °C followed by washing with 0.1 M phosphate buffer three times for 5 min each. Pellets were then dehydrated in gradient ethanol (30%, 50%, 70%, 90%, and 100%) for 5 min each. Pellets were infiltrated with a 1:1, 1:2, and 0:1 mixture respectively of acetone and Epon812 resin (TED PELLA, GP18010) at 38 °C for 3–4 h, and then were embedded in Epon812 resin. After being polymerized at 37 °C, 45 °C, and 60 °C for 12 h, 12 h, and 48 h, respectively, the resin was ultra-thin sectioned using a 70 nm-thick ultramicrotome (UC-7). Then the sections were double stained with 2% uranyl acetate (EMS, 22400) for 30 min and lead citrate (TED PELLA, 19314) for 15 min. After being dried in the air, samples were observed with a transmission electron microscope (Japan Electron Optics Laboratory Co., Ltd., JEM-1400 PLUS) at a voltage of 100 kV. Images were captured using a CCD camera (EMSIS, MORADA).

IEM

Cells were prepared for IEM with LR White resin (14381-UC, ELECTRON MICROSCOPY SCIENCES) as previously described.18 Briefly, cells were pelleted at 150× g for 8 min and fixed in a solution containing 2% paraformaldehyde, 0.05% glutaraldehyde, and 0.1 M PBS (pH 7.4) for 90 min at 4 °C. The fixed pellets were washed three times with 0.1 M PBS (pH 7.4) for 10 min each time at 4 °C and then dehydrated with a 30%, 50%, 70%, and 90% graded ethanol series at –20 °C; each ethanol dehydration step lasted for 20 min, and the 30% ethanol dehydration step was performed at 4 °C. Samples were infiltrated with 40%, 70%, and 100% LR white-ethanol series at –20 °C for 1 h per step, followed by infiltration with 100% LR white overnight at –20 °C. The resin-containing sample was then polymerized in a PCR tube by UV irradiation (360 nm) at –20 °C for 72 h and at room temperature for 48 h. Immunolabeling was performed with a rabbit anti-HA antibody (1:20; Cell Signaling; 3724) for 2 h at 37 °C, followed by incubation with goat anti-rabbit IgG conjugated to 10-nm gold particles (1:20; Sigma; G7402) as the secondary antibody for 2 h at 37 °C. The samples were visualized at an accelerating voltage of 120 kV with a JEOL JEM-1400 transmission electron microscope with an AMT XR41 digital imaging system.

High-resolution (12%–36%) iodixanol density gradient fractionation

A 400-mL cultured supernatant sample of HeLa cells was subjected to ultrafiltration to obtain the concentrated extracellular matter. The concentrated extracellular matter was isolated by iodixanol density gradient fractionation as previously described.67 Briefly, iodixanol (OptiPrep) density media (Sigma-Aldrich, D1556) was prepared in ice-cold PBS immediately before use to generate discontinuous step (12%–36%) gradients. The concentrated extracellular matter was resuspended in ice-cold PBS and mixed with ice-cold iodixanol/PBS for a final 36% iodixanol solution. The suspension was added to the bottom of a centrifugation tube, and the solutions with descending concentrations of iodixanol in PBS were carefully layered on top, creating a complete gradient. The bottom-loaded 12%–36% gradient layer was subjected to ultracentrifugation at 120,000× g for 15 h at 4 °C with an SW41 TI rotor (Beckman Coulter). Twelve individual 1-mL fractions were collected from the top to bottom of the gradient. Each 200-μL fraction was transferred to new tubes, 50 μL of 5× loading buffer was added, and the tubes were boiled for 10 min. Each 20-μL sample was subjected to SDS-PAGE and western blotting.

RAB22AQ64L-positive sub-organelle immunoprecipitation

HeLa cells stably expressing Flag-RAB22A or Flag-RAB22AQ64L were seeded on a 15-cm dish for 24 h. Upon reaching 90% confluence, the cells were washed and scraped into ice-cold KPBS buffer (136 mM KCl and 10 mM KH2PO4 at pH 2.5 adjusted with KOH) containing protease inhibitor, and then, the cells were pelleted by centrifugation at 2000× g for 1 min at 4 °C. The pellets were resuspended in 200 μL of KPBS buffer, and 10 μL of this suspension was used as the whole-cell control. Then, the remaining cells were homogenized with 40 strokes in a 2-mL Dounce homogenizer (Sigma, D8938). Next, the homogenate was centrifuged at 3000× g for 3 min at 4 °C. Supernatants containing anti-DYKDDDDK beads (Thermo Scientific™, A36797) washed three times with KPBS were incubated on a rotary shaker for 30 min at 4 °C. The mixture was placed on a magnetic rack to remove the supernatants, and then, the beads were washed three times with 1 mL of KPBS buffer. Beads containing immunoprecipitated samples were added to 40 μL of lysis buffer and boiled at 100 °C for use in western blot analysis.

The ELISA for STING

The serum of a patient was added to 0.5% NP40 for 1 h on ice to lyse the vesicles completely, and the level of STING was detected following the instructions of an ELISA kit (Cloud-Clone Corp, HEN011Hu). In brief, first, 100 μL of serum sample was added to a 96-well strip plate and incubated for 1 h at 37 °C. Then, the liquid in each well was removed, 100 μL of reagent A working solution was added, and the plate was incubated for 1 h at 37 °C. The solution was aspirated, and the well was washed with 350 μL of 1× wash solution using a multichannel pipette three times. Next, 100 μL of reagent B working solution was added and incubated for 30 min at 37 °C. The wash process was repeated five times as described above. Ninety microliter of substrate solution was added to each well and incubated for 10–20 min at 37 °C until the liquid sample turned blue. Then, 50 μL of stop solution was added and mixed on a shaker gently. A microplate reader was used to immediately measure the solution at 450 nm. The level of STING was analyzed based on a standard curve made on the basis of a standard solution.

GTP-binding assay

Cells were harvested upon reaching confluency in a culture of 90% and were suspended in binding buffer (20 mM HEPES, pH 8, 150 nM NaCl, and 10 mM MgCl2) containing a cocktail of protease and phosphatase inhibitors, lysed using three freeze–thaw cycles, and then centrifuged at 12,000 rpm for 20 min. The supernatants were incubated with 100 μL of GTP-agarose suspension (Sigma-Aldrich, G9768) for 2 h with rotation at 4 °C. The beads were pelleted by centrifugation, washed three times in binding buffer, and suspended in 40 μL of SDS-PAGE sample buffer. The proteins were boiled and subjected to SDS-PAGE and western blotting.

Lyso-Tracker Red probe

HeLa cells stably expressing GFP-RAB22A or GFP-RAB5A were plated on glass-bottom plates. After 24 h of growth to confluence, the medium was replaced with a medium containing 50 nM Lyso-Tracker Red (Beyotime Institute, C1046), and the cells were treated for 30 min with 5% CO2 at 37 °C. The Lyso-Tracker Red was removed, and a complete medium was added. Images were obtained using a fluorescence microscope.

Inducible recruitment of phospholipid phosphatases

In brief, we established a HeLa cell line stably expressing FRB-Flag-RAB22AQ64L that was split onto slides overnight and then transiently transfected with GFP-FKBP-INPP5E or GFP-FKBP-SAC1 with Lipofectamine 3000 in Opti-MEM medium. The medium was changed to a regular medium 8 h after transfection. After culturing overnight, the cells were treated with rapamycin (1 μM) for 80 min before immunostaining. Images were obtained using a fluorescence microscope.

PI4P binding assay

Pull-down using PI4P-coated beads was performed using cell lysates of 293T cells transfected with HA-Atg16L or HA-Atg16L 6A plasmid, in accordance with the manufacturer’s instructions (Echelon, P-B004A-2).

Liposome preparation

Liposome was prepared as described in published paper81,82 with some modifications. A total of 900 μg lipids (500 μg POPC, 102 μg cholesterol, 182 μg DPPE, 53 μg DOPS, and 63 μg PI4P) was dissolved in CH3Cl:CH3OH:1N HCl (2:1:0.01) mixture in a glass bottle. The solution was dried under pressurized N2 and then vacuumed for 40 min to completely remove the solvent. 300 μL HEPES butter (50 mM HEPES, 100 mM KCl, pH 7.5) was added to the vial. The vial was kept in liquid nitrogen to freeze the buffer and then transferred to room temperature water to melt the ice. The procedure was repeated until the solution became clear. The mixture was then extruded 18 times through an 800-nm polycarbonate membrane (Whatman) using an extruder.

Liposome-binding assay

HEK-293T cells transiently transfected with HA-Atg16L or HA-Atg16L 6A mutant were lysed in RIPA buffer as described above and the proteins were immunoprecipitated by HA agaroses. Proteins were eluted by 300 μg/mL HA peptide in HEPES buffer. The same amount of the two proteins was incubated with 30 μg liposome at room temperature for 30 min on a rotating wheel. Liposome was pelleted by centrifugation at 50,000 rpm using TLA110 rotor (Beckman) and then washed twice with HEPES buffer. The final protein-binding liposome was lysed in 20 μL RIPA buffer, and the amount of binding protein was analyzed by western blotting.

Clinical specimens and evaluation

A total of 100 paraffin-embedded NPC patient tissue samples were obtained from SYSUCC. These patients were pathologically diagnosed between January and December 2013 and received radiotherapy with or without chemotherapy. The overall survival time was calculated from the day of diagnosis to the end of the follow-up day in November 2019 or the date of death because of recurrence and/or metastasis. An anti-RAB22A rabbit antibody (Sigma, HPA066920) was used for immunostaining. The staining index (SI) was determined by multiplying the score obtained for the staining intensity with the score obtained for the positive area. The intensity score was defined as follows: 0, negative; 1, weak; 2, moderate; and 3, strong. Tissue sections were examined and scored separately by two independent investigators blinded to the clinic pathological data. X-tile software (V.3.6.1; Yale University, New Haven, Connecticut, USA) was used to generate the optimal cutoff values for determining high and low RAB22A density in the tumor areas and the relationship of this density to overall survival. Patients with a high SI (≥ 2) were assigned to the RAB22A high-expression group, and those with a low SI (0 and 1) were assigned to the RAB22A low-expression group. Survival analyses were performed using the Kaplan–Meier method, and differences were compared by log-rank test.

The NPC patient serum samples described in this paper were collected during the course of chemotherapy or radiotherapy between April and November 2020 in SYSUCC. Then, ELISA was performed to detect the level of STING in the patient serum samples. Approximately 4–6 weeks after chemotherapy or at the end of radiotherapy, treatment responses were assessed radiographically by two senior clinicians specializing in NPC using Response Evaluation Criteria in Solid Tumors guidelines (version 1.1).83 Then, we compared the STING levels in the serum of NPC patients between the two RAB22A expression groups using t-tests.

This retrospective study was approved by the Ethics Committee of SYSUCC (Approval no. B2021-096-01) in accordance with the Declaration of Helsinki.

MTT assay

A 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was used to measure cell viability. Briefly, 4T1 cells were seeded at a density of 800 cells per well in a 96-well microplate. The cells were incubated with MTT for 4 h, and the optical density was detected at 490 nm with the microplate reader once per day for 4 days.

Statistical analysis

GraphPad Prism (version 8.3.0) software was used for all statistical analyses. The significance of the differences was assessed using the two-tailed Student’s t-test. The correlations between RAB22A expression and overall survival curves were assessed using Kaplan–Meier plots and compared with the log-rank test. Data are presented as means ± SEM. Differences were considered significant when P values were < 0.05.