Cyclic GMP–AMP (cGAMP) is a second messenger that activates the stimulator of interferon genes (STING) innate immune pathway to induce the expression of type I IFNs and other cytokines. Pharmacologic activation of STING is considered a potent therapeutic strategy in cancer. In this study, we used a cell-based phenotypic screen and identified podophyllotoxin (podofilox), a microtubule destabilizer, as a robust enhancer of the cGAMP–STING signaling pathway. We found that podofilox enhanced the cGAMP-mediated immune response by increasing STING-containing membrane puncta and the extent of STING oligomerization. Furthermore, podofilox changed the trafficking pattern of STING and delayed trafficking-mediated STING degradation. Importantly, the combination of cGAMP and podofilox had profound antitumor effects on mice by activating the immune response through host STING signaling. Together, these data provide insights into the regulation of cGAMP–STING pathway activation and demonstrate what we believe to be a novel approach for modulating this pathway and thereby promoting antitumor immunity.

Several immunomodulatory approaches to enhance antitumor T-cell responses have recently been developed. However, only a fraction of patients with cancer respond to these therapies. Harnessing innate immune systems to induce antitumor responses and connect them with adaptive immune activation creates new possibilities for cancer therapy (1, 2).

The cyclic GMP–AMP synthase (cGAS)–stimulator of interferon genes (STING) pathway is an important component of the innate immune response (3, 4). cGAS is a cytosolic DNA sensor that synthesizes 2′3′-cyclic GMP–AMP (cGAMP), a second messenger in the cGAS–STING pathway (5–7). cGAMP binds and activates the adaptor protein STING, which then activates tank-binding kinase 1 (TBK1) and transcription of interferon regulatory factor 3 (IRF3), ultimately initiating the expression of type I IFNs and other immune-modulating molecules (8, 9). These potent antiviral and antitumor cytokines can trigger downstream factors of adaptive immunity.

Emerging evidence suggests that the cGAS–STING pathway is critical in antitumor immunity (10, 11). Moreover, cancer cells often carry unstable genomes, with tumor cell nuclei or mitochondria leaking DNA that can be taken up by phagocytotic cells to activate the cGAS–STING pathway (12). cGAMP produced and synthesized by cancer cells can activate the STING pathway because cGAMP is an immune transmitter (13, 14). Thus, cGAMP and STING are important mediators in the antitumor response (15, 16).

STING trafficking is required for its signaling activation. Upon binding cGAMP, STING undergoes high-order oligomerization and is translocated from the endoplasmic reticulum (ER) through ER–Golgi intermediate compartments (ERGIC) to the Golgi (17, 18). Following activation, STING also triggers autophagy and is trafficked to lysosomes, where it is degraded (19–21). Because cGAMP–STING signaling can elicit a robust antitumor response, there has been tremendous interest in developing STING agonists (22–24), such as cGAMP analogues and nonnucleotide STING agonists. Moreover, the combination of cGAMP or other STING agonists with programmed cell death 1 ligand 1 (PD-L1) antibody displays stronger antitumor effects than either treatment alone (25, 26). We hypothesize that enhancing the cGAMP–STING signaling pathway with specific small molecules is a potential strategy to enhance the antitumoral effects of cGAMP. To test this, we used a cell-based chemical screening strategy, discovering that podofilox altered cGAMP–STING translocation patterns to significantly enhance the oligomerization and activation of STING. Our results reveal that upon cGAMP treatment, podofilox increased the total population of STING oligomers and STING–TBK1 complexes. We showed that podofilox strongly boosted the immune responses and antitumor effects mediated by cGAMP. The combination of cGAMP and podofilox triggered a robust immune response to tumors and the host STING pathway. Activating STING signaling in hosts was the dominant factor in the proposed antitumor immunotherapeutic strategy in vivo. Overall, these data provide insights into the mechanisms underlying activation of the cGAMP–STING pathway mediated by the compound podofilox in vitro and in vivo. These results have implications for utilizing the cGAS–STING pathway in cancer immune therapy.

Antibodies and reagents

Anti-GM130 antibody (catalog no. 610823) was purchased from BD Biosciences. Rabbit antibodies against STING (#13647), P-STING (#19781), P-TBK1 (#5483), P-IRF3 (#4947), and Stat1 (#14994) were purchased from Cell Signaling Technology. Rabbit antibody against β-tubulin (PA5068) was purchased from Abmart. Rabbit antibodies against p40 (#DF6820), p63 (#AF2379), and enzyme-labeled sheep anti-mouse/rabbit IGg Polymer (#PV6000D2) were purchased from OriGene. Mouse antibodies against CD3 (#TA506064), CD45 (#TA506046), CD4 (#TA500481), CD8 (#TA802376), and E-cadherin (#ZM-0092) were purchased from OriGene. Rabbit antibody against IFNAR2 (A1769) was purchased from ABclonal. IFNβ protein (10704-HNAS) was purchased from Sino Biological. Lipopolysaccharide (LPS; L5293) and polyinosinic-polycytidylic acid (Poly I:C; P1530) were purchased from Sigma-Aldrich. 2′3′-cGAMP (T10065) was purchased from TargetMol. 3′3′-cGAMP (tlrl-nacga), c-di-AMP (tlrl-nacda), and c-di-GMP (tlrl-nacdg) were purchased from InvivoGen. Podofilox (HY-15552), DMXAA (HY-10964), G10 (HY-19711), STING agonist-3 (HY-103665), RO8191 (HY-W063968), cycloheximide (HY-12320), nocodazole (HY-13520), combretastatin A4 (HY-N2146), vinblastine sulfate (HY-13780), ansamitocin P3 (HY-15739), 4′-Demethylepipodophyllotoxin (HY-17435), Colchicine (HY-16569), vincristine sulfate (HY-N0488), monomethyl auristatin E (HY-15162), paclitaxel (HY-B0015), topotecan (HY-13768), etoposide (HY-13629), docetaxel (HY-B0011), brefeldin A (HY-16592), and bafilomycin A1 (HY-100558) were purchased from MedChemExpress. The AlamarBlue Cell Viability Reagent (DAL1025) and Dead Cell Apoptosis Kit (V13242) were purchased from Invitrogen. Calcein-AM/PI Double Stain Kit (no. 40747ES76) was purchased from YEASEN.

Mice

Wild-type (WT) C57BL/6 mice and Balb/c mice were purchased from Shanghai SLAC Laboratory Animal Co., Ltd. Stinggt/gt mice were kindly provided by Dr. Liufu Deng from Shanghai Jiao Tong University School of Pharmacy (Shanghai, China). All mice were maintained under specific pathogen-free (SPF) conditions in a barrier-sustained facility and provided with sterile food and water. Animal experiments were carried out in accordance with regulations in the Guide for the Care and Use of Laboratory Animals issued by the Ministry of Science and Technology of the People's Republic of China. The protocol was approved by the Institutional Animal Care and Use Committee of Shanghai Jiao Tong University School of Medicine (Shanghai, China). All experiments were conducted with sex and age-matched mice.

Cell culture

ISG-THP1 cells (no. thpl-isg, InvivoGen) were obtained in 2015 and cultured in RPMI1640 medium (Gibco, Life Technologies) supplemented with 10% FBS (Thermo Fisher Scientific), 1% penicillin and streptomycin (100 IU/mL and 100 μg/mL, respectively, Gibco, Life Technologies). B16F10 (CRL-6475), 4T1 (CRL-2539), and L929 (CCL-1) cells were purchased from ATCC in 2015. BJ-5ta (CRL-4001), HEK293T (CRL-3216), and HeLa (CCL-2) cells were purchased from ATCC in 2015. These cell lines within 20 passages were stored in liquid nitrogen for long-term storage. mouse embryonic fibroblasts (MEF) were isolated from the WT ICR mice pregnant (Beijing Vital River Laboratory Animal Technology Co., Ltd.) for 13.5 to14 days. The above cells, including B16F10, 4T1, L929, BJ-5ta, HEK293T, and HeLa, were cultured in DMEM (Gibco, Life Technologies) supplemented with 10% FBS and 1% penicillin and streptomycin (100 IU/mL and 100 μg/mL, respectively, Gibco, Life Technologies) at 37°C in 5% CO2. All cell lines tested negative for Mycoplasma and rodent pathogens, and they have been passaged fewer than 20 times when used to start the experiments. Cell growth and evaluation based on morphology, growth speed, and immune response were carefully examined to avoid cross contamination.

Transfection and construction of stable cell lines

To generate stable cell lines, human STING (gene ID: 340061), mouse STING, STING-110GFP, and STING-GFP were cloned into the gateway cloning pCDH-CMV-MCS-IRES-blasticidin vector (#72299, Addgene) to generate lentivirus expression constructs. For CRISPR/Cas9 knockout, the sgRNA sequences targeting STING (sgRNA-forward: 5′-ACTCTTCTGCCGGACACTTG-3′ and reverse 5′-CAAGTGTCCGGCAGAAGAGT-3′), Stat1 (sgRNA-forward: 5′-CCTGATTAATGATGAACTAG-3′ and reverse 5′-CTAGTTCATCATTAATCAGG-3′), or Stat3 (sgRNA-forward: 5′-AGCTACAGCAGCTTGACACA-3′ and reverse 5′-TGTGTCAAGCTGCTGTAGCT-3′) were inserted into plentiCRISPR v2 (#52961, Addgene). ISG-THP1 cells (STING KO, Stat1 KO, Stat3 KO) were generated following a standard protocol (27). The viruses for stable cell lines were generated and collected following a standard protocol (27). HeLa and ISG-THP1 cells were infected with lentiviruses for 24 to 48 hours in the presence of 10 μg/mL polybrene (TR-1003, Sigma-Aldrich) before being selected with 2 μg/mL puromycin (ST551, Beyotime Biotechnology) and 5 μg/mL blasticidin (R21001, Thermo Fisher Scientific), according to the corresponding selection markers of the constructs. One week after selection, the cell pool was analyzed by Western blot (see Western Blotting) to determine the knockout efficiency of specific sgRNAs and the expression level of STING. Cells with the highest knockout (KO)/overexpressing efficiency were chosen for colonization of stable cell lines.

High-throughput screening

ISG-THP1 cells express the secreted luciferase (Lucia) reporter gene under the control of an IRF-inducible ISRE promoter. In the primary screen, ISG-THP1 cells were cultured in a 384-well plate (Nest, 761601) in an RPMI1640-based assay medium with cGAMP (0.5 μmol/L) and candidate compounds (10 μmol/L) at 5% CO2 and 37°C for 24 hours. Substrate buffer (50 mmol/L HEPES pH 7.0, 50 mmol/L NaCl, 0.05% CHAPS, 10 mmol/L EDTA, and 1 μmol/L coelenterazine) was added to the 384-well plate. The mixture was incubated in the dark for 5 minutes at room temperature. Luminescence was read using a Cytatio Cell Imaging Reader (BioTek). Natural Product Library (catalog no. HY-L021), Kinase inhibitor library (catalog no. HY-L009), FDA-Approved Drug Library (catalog no. HY-L022), and cell cycle/DNA Damage Compound Library (catalog no. HY-L004) were purchased from MedChemExpress. Natural Compound Library (L6000) was purchased from TargetMol. Preclinical/Clinical Compound Library (No. L3900), FDA-Approved Drug Library (No. L1300), and Express-Pick Library (catalog no. L3600) were purchased from Selleck. DiscoveryProbe FDA-approved Drug (No. L1021) and DiscoveryProbe Anticancer Compound Library (No. L1023) were purchased from APExBIO.

ELISA

ISG-THP1 cells and primary mouse bone marrow–derived macrophages (BMDM; obtained as described below in the section “In vitro culture and functional assay of BMDMs”) were separately seeded in 12-well culture plates at a concentration of 3 × 106 cells/mL. The cells were stimulated with cGAMP (0.5 μmol/L) or cGAMP (0.5 μmol/L) combined with indicated doses of podofilox for 24 hours or 6 hours, respectively. Triton X-100 was added to obtain lysates. The concentration of IFNβ was measured with VeriKine-HS Mouse Interferon Beta Serum ELISA Kit (No. 42410, PBL Assay Science) and ELISA Kit for human IFNβ (catalog no. SEA222Hu, Cloud-Clone Corp) in accordance with the manufacturers’ instructions.

ISRE-luciferase assay

ISG-THP1 cells were resuspended in serum-free RPMI1640 medium (catalog no. 61870036, Gibco, Life Technologies) at a density of 3 × 106 cells/mL and treated with 0.5 μmol/L cGAMP, 1 μg LPS, 10 μg Poly(I:C), or Sendai virus (MOI = 0.1; SeV, a gift from Zhengfan Jiang, Peking University, Beijing, China) and indicated doses of podofilox or vehicle (DMSO) for 24 hours. To evaluate expression of the luciferase reporter, 20 μL cell suspension and 60 μL substrate detection reagent (50 mmol/L HEPES pH 7.0, 50 mmol/L NaCl, 0.05% CHAPS, 10 mmol/L EDTA, and 1 μmol/L coelenterazine) were added to each well of a 96-well white plate. Luminescence was read using a Cytation Cell Imaging Reader (BioTek) set with an integration time of 0.1 seconds. Luminescence signals for test article samples were normalized to vehicle-treated samples and reported as relative fold change.

HEK293T cells IFNβ-luciferase reporter assay

HEK293T cells were transiently cotransfected with the secretory IFNβ-luciferase reporter plasmid (a gift from James Chen, University of Texas Southwestern Medical Center, Dallas, TX) and the different STING plasmids (STING WT, STING-110GFP, and STING-GFP). Cells were stimulated with 0.5 μmol/L cGAMP with or without 100 nmol/L podofilox 12 hours after transfection. After an additional 24-hour incubation, the luciferase assay was performed according to the standard protocol (see section “ISRE-luciferase assay” for details).

RNA extraction and qRT-PCR

Primary BMDM cells were resuspended in RPMI1640 medium (catalog no. 61870036, Gibco, Life Technologies) supplemented with 2% FBS at a density of 3 × 106 cells/mL and treated with 0.5 μmol/L cGAMP with or without 0.1 μmol/L podofilox. In total, 2.5 mL of cells were seeded into each well of a 6-well plate and incubated for 6 hours. Total RNA was extracted with an RNeasy Micro Kit (catalog no. 73934, QIAGEN) and 1 μg of purified RNA was reverse-transcribed into cDNA with Seniscript Reverse Transcription Kit (catalog no. 205313, QIAGEN). Real-time RT-PCR was performed with SsoFast EvaGreen Supermix (catalog no. 1725201, Bio-Rad) according to the manufacturer's instructions on a StepOne Plus (Applied Biosystems). Relative value of gene expression was calculated by the 2−ΔCt method. All PCR amplification was normalized to β-actin. Primers used for each gene were as follows: IFNβ, forward 5′-GCCTGGATGGTGGTCCGAGCA-3′ and reverse 5′-TACCAGTCCCAGAGTCCGCCTCT-3′; IFITM1, forward 5′-GCCACCACAATCAACATGCCTG-3′ and reverse 5′-ACCCACCATCTTCCTGTCCCTA-3′; CCL5, forward 5′-AGATCTCTGCAGCTGCCCTCA-3′ and reverse 5′-GGAGCACTTGCTGCTGGTGTAG-3′; CXCL10, forward 5′-TCCGGAAGCCTCCCCATCAGCACC-3′ and reverse 5′-TGCAGCGGACCGTCCTTGCGA-3′; IL6, forward 5′-CAAGAAAGACAAAGCCAGAGTC-3′ and reverse 5′-GAAATTGGGGTAGGAAGGAC-3′. β-actin, forward 5′- TGCTGACAGAGGCACCACTGAA-3′ and reverse 5′-CAGTTGTACGTCCAGAGGCATAG-3′.

cGAMP stimulation

Cell delivery of cGAMP was achieved by perfringolysin O (1 mg/mL, 1:20,000–1:40,000) in BJ-5ta, MEF, HEK293T, and HeLa cells. Perfringolysin O was not used for cGAMP stimulation in BMDM and ISG-THP1 cells. Perfringolysin O was overexpressed and purified as previously described (28).

Quantification of intracellular cGAMP by LC/MS-MS

ISG-THP1 cells were incubated with cGAMP (0.5 μmol/L) for 4 hours with or without podofilox. The cells were collected in 80% analytical pure MeOH and subjected to repeated freeze–thaw cycles with liquid nitrogen to obtain cell lysates. The lysates were loaded to LC/MS-MS for quantification. The LC/MS-MS analysis was performed on an AB SCIEX QTRAP 6500 triple quadrupole mass spectrometer in MRM and positive ionization mode. LC was separated in the Acquire XSelect HSS T3 analysis column (2.1 × 100 mm, 2.5 μm, Waters) with a LC-30A infinity binary pump. For gradient elution, methanol was used as solvent A. Ammonium acetate aqueous solution (10 mmol/L) was used as solvent B, and the flow rate was 0.2 mL/minute. The initial condition was 100% solvent B, and then it was reduced to 5% B in 3 minutes. After 2 minutes, it returned to the initial proportion until the end of 10 minutes. The sample was kept at 10°C and the column temperature was kept at 35°C. Injection volume was 5 μL. Mass spectrometry (MS) parameters were as follows: electrospray ionization (ESI) ion source temperature was 300°C, Curtain gas: 30 psi; collision activation dissociation (CAD) gas setting: medium, ion injection voltage: 5500 V, ion gas 1, and 2: 50 psi. cGAMP was determined as m/z 675.1, and secondary fragment ion m/z 524.1/506.1. Declustering potential (DP) 140 V, collision energy (CE) 30 V and collision exit voltage (CXP) 10 V. Data analysis was processed by AB SCIEX Analyst 1.63 Software (Applied Biosystems).

Western blotting

Activation of the STING pathway was assessed by Western blotting to analyze phosphorylation status and total protein levels of STING, TBK-1, and IRF3 using commercially available antibodies (see section “Antibodies and reagents” for details). Cells were treated with cell lysis buffer (catalog no. P0013, Beyotime Biotechnology) supplemented with protease inhibitors (Roche) on ice for 30 minutes. Subsequently, the cell lysates were centrifuged for 10 minutes at 13,000 × g and boiled with SDS loading buffer at 95°C for 5 minutes. Proteins were separated on 12% SDS-PAGE gels, immunoblotted onto transfer membranes (Immobilon-P: PVDF membrane; pore size: 0.45 μm; Merck Millipore), and subsequently incubated with different primary antibodies overnight (see section “Antibodies and reagents” for details). After incubation with horseradish peroxidase (HRP)-labeled secondary antibodies (HX2031/HX2032 Huaxingbio) for 1 hour at room temperature, the proteins were detected using enhanced chemiluminescence (ECL) substrates (MF074, Mei5 Biotechnology).

STING oligomerization assay

ISG-THP1 cells were seeded in 12-well culture plates at 3×106 cells/mL. The cells were incubated with cGAMP (0.5 μmol/L) or cGAMP (0.5 μmol/L) and podofilox (1 μmol/L or 5 μmol/L) for 4 hours. Cells were then resuspended in a buffer containing 25 mmol/L Tris pH 7.5, 5 mmol/L MgCl2, 1 mmol/L DTT, and 0.5 mg/mL Leupeptin (L8110, Solarbio) with a total volume of 300 μL. Then cells were incubated on ice for 15 minutes and sonicated using a micro-sonicator. Lysates were centrifuged at 800 × g for 5 minutes to generate the S1 supernatant. An equal volume of a buffer containing 20 mmol/L Tris pH 7.5, 150 mmol/L NaCl, 5 mmol/L MgCl2, 10% glycerol, 2% NP40, 1 mmol/L DTT, 5 mmol/L Na3VO4, and 0.5 mg/mL Leupeptin was added to 40 μL supernatant for native PAGE. Prepared samples were then mixed with 5 × native loading buffer (20 mmol/L Tris pH 7.5, 5 mmol/L MgCl2, 150 mmol/L NaCl, 1% NP40, 10% glycerol, 5 mmol/L Na3VO4, 1 mmol/L DTT, and 0.5 mg/mL Leupeptin), and run at constant power of less than 5W for 1 hour at 4°C in running buffer (cathode buffer: 25 mmol/L Tris, 192 mmol/L Glycine, 0.4% DOC-Na2, pH 8.5; anode buffer: 25 mmol/L Tris, 192 mmol/L Glycine). Samples were separated on 4%–15% Precast-GLgel Tris-Glycine (C661104, Sangon Biotech), and immunoblotted onto nitrocellulose membranes for further analysis (see section “Western Blotting” for details). The samples were also denatured and loaded onto SDS-PAGE to perform immunoblotting with indicated antibodies (as described in section “Western blotting” for details).

In vitro culture and functional assay of BMDMs

Femurs and tibias of C57BL/6 (WT and Singgt/gt) mice were collected and BM cells were flushed out with RPMI1640. The cells were differentiated into macrophages (BMDM) by culturing for 6 to 7 days in L929-cell conditioned media (LCCM, DMEM, 20% FBS, 30% L929 culture medium). A total of 2 × 106 cells were plated into a 12-well culture plate and cultured overnight. The cells were stimulated with cGAMP (0.5 μmol/L) or cGAMP (0.5 μmol/L) plus podofilox for subsequent experiments.

Immunofluorescence microscopy

HeLa cells stably overexpressing STING-110GFP or STING-GFP were cultured on confocal dishes (Corning). Then cells were stimulated with 2 μmol/L or 8 μmol/L cGAMP (with 50 ng/mL perfringolysin O delivery), 1 μmol/L brefeldin A, 1 μmol/L podofilox, and cGAMP or 20 μmol/L G10 for 2 hours. For immunostaining experiments, cells were seeded on coverslips in a 24-well plate. Cells were treated with 2 μmol/L or 8 μmol/L cGAMP (with 50 ng/mL perfringolysin O delivery) and/or 1 μmol/L podofilox. Cells were fixed with 4% paraformaldehyde (PFA; Solarbio) for 15 minutes. The blocked cells were incubated with anti-GM130, anti-tubulin, and anti-p-TBK1 primary antibodies at 4°C overnight, and then incubated with fluorescent dye–conjugated secondary antibodies (A11008, A11004, or A11011, Thermo Fisher Scientific) at room temperature for 2 hours according to manufacturer's instructions. The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI; C1002, Beyotime Biotechnology). Cell coverslips were mounted onto microscope slides using VECTASHIELD Mounting Medium (H-1000, Vector Laboratories) and fixed with a coat of clear nail polish. Confocal images were acquired on a Zeiss LSM980 Airyscan2 Confocal microscope using a 63× (NA 1.45) objective and processed in Zen Blue 3.1 software. Live-cell time-lapse imaging was acquired on a Nikon AX super-resolution confocal microscope using a 100× (NA 1.45) objective and processed in NIS-Elements software. Structured illumination microscope (3D-SIM) imaging was acquired using the highly intelligent and sensitive SIM (HIS-SIM) of Guangzhou Computational Super-resolution Biotech Co., Ltd, and Wiener deconvolution was used in reconstructed images.

Tumor organoids

Tumor organoid were generated from tumor tissue from 4 patients undergoing surgical resection of small-cell lung cancer (SCLC) at the Peking University People's Hospital (Beijing, China). The study was approved by the Tsinghua University Medical Ethics Committee and written informed consent was obtained from the patients and/or their authorized representatives. The method to process clinical samples has been described in detail in a previously published article (29). Briefly, the fresh tissue was minced as small as possible with surgical scissors and then suspended in AdDF+++ (Advanced DMEM/F12; catalog no. 12634028, Gibco by Life Technologies containing 1 × Glutamax; catalog no. 35050079, Gibco by Life Technologies, 10 mmol/L HEPES; catalog no. 11344041, Gibco by Life Technologies, and 1% penicillin & streptomycin). The suspension was sequentially filtered through membranes of 100 μm and 40 μm, and clusters with sizes between 40 to 100 μm were released into the organoid culture medium (OCM) created as described previously (29). Organoids were cultured in plates and on integrated superhydrophobic microwell array (InSMAR)-chips. The organoid suspension was centrifuged (500 × g, 5 minutes, 4°C) and then suspended in cold growth factor reduced Matrigel (catalog no. 354230, BD Biosciences). Drops of 60 μL Matrigel mixed with organoid suspension were inoculated into prewarmed ultra-low attachment 24-well flat-bottom cell culture plates (Corning). On InSMAR-chips, the organoids embedded in Matrigel were dispensed uniformly into each microwell with 5 to 10 organoids/400 μL. After Matrigel solidified, 400 μL OCM was added into each plate and 1 to 2 μL OCM was supplemented to microwells, then both plate and InSMAR-chip were transferred to a 37°C, 5% CO2 cell culture incubator. The OCM was refreshed depending on the growth of organoids.

3D drug assays on InSMAR-chip

For drug sensitivity assays, organoids were harvested and suspended with 30 μL Matrigel containing 500 to 1,000 organoids, then inoculated on InSMAR-chips (named organoid chip). OCM was dispensed on silylated glass using a standard microarray robotic spotter that we have employed before (30). After Matrigel solidified, the glass carrying the OCM array was aligned and sandwiched upside down to enable contact of the InSMAR-chip with Matrigel droplets on it (named feeding process). The InSMAR-chip was placed into a 37°C, 5% CO2 incubator. OCM medium was refreshed every two days by the recycling feeding process.

On day 3, the first cell-viability assays were conducted using alamarBlue Cell Viability Reagent (AB1# for short; catalog no. DAL1025, Invitrogen). According to the feeding process, alamarBlue was dispensed on a silylated glass and formed alamarBlue droplet arrays with a volume of 400 nL of each droplet at a final concentration of 10% and fed onto the organoid chip. The organoid chip overlaid with alamarBlue was transferred to an incubator for 2 to 3 hours of incubation. After that, the whole organoid chip was scanned using Olympus IX83 inverted fluorescence microscope. To eliminate background noise from the fluorescence signal of alamarBlue, we incubated alamarBlue on Matrigel (microwells without organoids) as a negative control (NC). After scanning, alamarBlue was removed by sweeping a filter paper over the arrays and replaced with 1.6 μL of drug-containing OCM each microwell. To consider the pre-existing volume of Matrigel in the organoid chip, the drug compound concentration was calculated as follows:
formula
Organoids were treated with drugs for 72 hours before second cell viability was assessed (AB2# for short). At the end of the treatment, the drug-containing OCM was removed and alamarBlue was fed on the organoid chip again. After 2 to 3 hours of incubation, the whole organoid chip was scanned again. Percentage of viable cells after drug treatment was analyzed by normalizing alamarBlue signals with that of vehicle control (0.1% DMSO, VC) and negative control (NC). The cell viability fraction was calculated using the following formula:
formula

(OAB1# & OAB2# = Fluorescence values of drug-treated organoids from first and second cell-viability assays; NCAB1# & NCAB2# = Background noise values of NC from first and second cell-viability assays; VCAB1# & VCAB2# = Fluorescence values of VC from first and second cell-viability assays; n = number of repeat microwells per drug concentration).

Drug dose–response curves were plotted as cell-viability fraction against the logarithm of drug concentrations in μmol/L and were fitted to estimate the half-inhibitory concentration (IC50).

Histology and immunostaining

The harvested organoids were washed with cold PBS and suspended in 40 μL of 10 mg/mL Fibrinogen solution (#9001–32–5, Sigma-Aldrich), then immediately mixed with 20 μL of Thrombin reagent (T8021, Solarbio) to facilitate fibrinogen solidification. Afterwards, the organoid fibrinogen pellets were fixed in 1 mL 4% PFA (Sigma-Aldrich). Matched tissue (fixed in 4% PFA) and organoid fibrinogen pellets were dehydrated and embedded in paraffin, cut at 5 μm, and stained using a standard hematoxylin and eosin (H&E) protocol. In brief, samples were dewaxed with xylene and washed with different concentrations of ethanol: 100% ethanol, 95% ethanol, 80% ethanol, 75% ethanol. Slides were fixed in 4% PFA at room temperature for 15 minutes and washed with PBS for 5 minutes. Stained with hematoxylin for 10 minutes, differentiation medium for 30 seconds and eosin for 2 minutes in sequence, washed with tap water and mounted. For IHC staining, the paraffin slides were first baked at 72°C for 30 minutes, then deparaffinized in xylene and rehydrated through a graded ethanol series. After antigen retrieval, the endogenous peroxidase was blocked in 3% hydrogen peroxide deionized water for 15 minutes. The slides were blocked with 5% goat serum (catalog no. 16210072, Thermo Fisher Scientific) and incubated with primary antibodies (see section “Antibodies and reagents” for details) at 4°C overnight. After being washed with PBS, slides were incubated with secondary antibodies (PV6000D2, OriGene) at room temperature for 20 minutes and developed with 3,3′Diaminobenzidine (DAB) chromogenic solution (PV-6000D4, OriGene) following the manufacturer's instructions. Cell nuclei were restained with hematoxylin and dehydrated using gradient ethanol. Then slides were hyalinized with xylene and sealed for microscope snapshots. Antibodies were diluted in blocking buffer at an appropriate dilution ratio, and more detailed information of antibodies can be found in the manufacturer's protocol. Bright-field images were obtained with an Olympus IX83 inverted fluorescence microscope. H&E and IHC images were acquired using the 3DHISTECH Panoramic SCAN system. The series of images were viewed by 3DHISTECH CaseViewer software.

Calcein AM/PI double staining

Organoids were treated with 10 μmol/L gemcitabine–cisplatin, 10 μmol/L cGAMP (with 30 ng/mL perfringolysin O delivery) with or without podofilox (10 μmol/L), and vehicle (0.1% DMSO and the same perfringolysin O background) for 72 hours, and incubated with Calcein AM/PI reagent for 1-hour following the manufacturer's protocol (catalog no. 40747ES76, YEASEN). Bright-field and IF images were obtained with an Olympus IX83 inverted fluorescence microscope.

Tumor organoid apoptosis assay

Organoids were harvested out of Matrigel after treatment with gemcitabine–cisplatin (10 μmol/L), cGAMP (10 μmol/L, with 30 ng/mL perfringolysin O delivery) with or without podofilox (10 μmol/L), and vehicle (0.1% DMSO and the same perfringolysin O background) for 72 hours. The organoids were washed with PBS and trypsinized into single cells with TrypLE (catalog no. A1285901, Life Technologies), then the single cells were collected and resuspended in a binding buffer (catalog no. V13242, Invitrogen). The cells were incubated with Annexin V-FITC and PI (catalog no. V13242, Invitrogen) for 10 minutes at room temperature in the dark and analyzed using a BD FACSCalibur. The data were analyzed using FlowJo (version 10.8.1, TreeStar).

Tumor inoculation, treatment, and measurement

B16F10 melanoma tumor cells (CRL-6475, ATCC) and 4T1 breast cancer cells (CRL-2539, ATCC) were grown in DMEM containing 10% FBS. A total of 1 × 106 tumor cells in 100 μL PBS were injected subcutaneously into the right dorsal flanks of the C57BL/6 mice. Seven days after tumor inoculation, mice were treated with vehicle, 10-μg cGAMP, podofilox (0.7 mg/kg), or 10-μg cGAMP along with podofilox (0.7 mg/kg) by intratumoral injection on days 7, 10, and 13. In some experiments, 200 μg of PD-L1–specific antibody was injected intraperitoneally. Tumors were measured every three days and the tumor sizes were calculated using the following formula: π/6 × length × width × height.

Tumor-infiltrating leukocyte separation and staining

For analyses of tumor-infiltrating leukocytes, tumors from B16F10 tumor–bearing mice were collected, then minced and filtered through a 70-μm strainer to obtain single-cell suspensions. Red blood cells were lysed with red blood cells (RBC) lysis buffer. The cells were further blocked and stained with different antibodies: Pacific Blue anti-mouse CD45 (catalog no. 157212, BioLegend), PerCP/Cy5.5 anti-mouse CD3 (catalog no. 100218, BioLegend), PE/Cy7 anti-mouse CD4 (catalog no. 100422, BioLegend), Alexa Fluor 700 anti-mouse CD8α (catalog no. 100730, BioLegend), Brilliant Violet (BV) 711 anti-mouse CD69 (catalog no. 104537, BioLegend), BV510 anti-mouse CD25 (catalog no. 102042, BioLegend), FITC anti-mouse PD-1 (catalog no. 135214, BioLegend), APC anti-mouse/human CD44 (catalog no. 103012, BioLegend), PE anti-mouse CD62 L (catalog no. 161204, BioLegend), and PE anti-mouse NK-1.1 (catalog no. 156504, BioLegend). Stained cells were analyzed with an LSRII instrument (BD Biosciences) and the data were analyzed using FlowJo (version 10.8.1, TreeStar).

Intracellular flow cytometry

To determine the intracellular expression of cytokines, tumor-infiltrating cells were stimulated with a cell stimulation cocktail (catalog no. 423303, BioLegend) for 4 hours in the presence of brefeldin A (catalog no. 423303, BioLegend). Cells were first stained with antibodies against cell surface antigens and then fixed with 4% PFA and permeabilized with Intracellular Staining Perm Wash Buffer (catalog no. 421002, BioLegend) for 30 minutes. After staining with BV 711 anti-mouse IFNγ (catalog no. 505836, BioLegend), BV 510 anti-mouse TNFα (catalog no. 506339, BioLegend), and FITC anti-human/mouse Granzyme B (catalog no. 515403, BioLegend), flow cytometry was performed using an LSRII (BD Biosciences) and the results were analyzed using FlowJo (version 10.8.1, TreeStar).

Statistical analysis

Statistical comparisons were performed with unpaired t tests to compare two datasets. One-way ANOVA was used for more than two datasets, with Tukey multiple comparisons test to determine within-group difference. Two-way ANOVA was used to evaluate differences between treatment groups over time for in vivo tumor growth. The survival among different treatment groups was analyzed using the Kaplan–Meier method. A value of P < 0.05 was considered statistically significant. GraphPad Prism 8 software (GraphPad Software, Inc.) was used for the analysis.

Data availability

The data generated in this study are available in the article and its Supplementary Data files and all materials and data are available from the corresponding authors on request.

Discovery of the cGAMP–STING signaling enhancer podofilox

Recent studies have suggested cGAMP-induced innate immunity could provide a novel strategy for cancer treatment (31, 32). Pharmacologic activation of the cGAMP–STING signaling pathway is thus under investigation. In this study, we used THP1-Lucia ISG cells (ISG-THP1 cells), which contain a luciferase reporter construct under the control of an IRF-inducible ISRE promoter, to identify small molecules from a collection of approximately 150,000 commercially available structurally diverse drug like small molecules that regulate innate immune pathways (Supplementary Fig. S1). We identified a series of microtubule destabilizers, including nocodazole, combretastatin A4, ansamitocin P-3, 4′-Demethylepipodophyllotoxin, colchicine, vinblastine sulfate, vinorelbine ditartrate, vincristine sulfate, and podofilox, that enhanced the IRF-induced immune response in ISG-THP1 cells induced by cGAMP (Fig. 1AC; Supplementary Fig. S2A–S2H). We also investigated the effect of monomethyl auristatin E (MMAE), which is widely used in antibody–drug conjugation (ADC), showing that it enhanced the activation of the cGAMP–STING signaling pathway (Supplementary Fig. S2I). To further define the modulation mechanism of these microtubule destabilizers, we investigated the effects of microtubule stabilizers. The results showed that microtubule stabilizers (paclitaxel and docetaxel) and DNA topoisomerase inhibitors (etoposide and topotecan) were not able to enhance cGAMP–STING signaling, whereas podofilox robustly enhanced the IRF-induced immune response upon cGAMP treatment in a dose-dependent manner (Supplementary Fig. S3A–S3C). These results suggested that the enhancement effect of podofilox was not through mitotic arrest and related cell death.

Figure 1.

Discovery and characterization of podofilox as a cGAMP–STING signaling enhancer. A, Chemical structure of podofilox. B, C, and H, Cell-based activity of podofilox in WT and STING KO ISG-THP1 cells after treatment with cGAMP (0.5 μmol/L) and/or podofilox (indicated concentrations) for 24 hours; fold change of ISRE luciferase activity was normalized to DMSO-treated cells. D and E, Immunoblotting analysis of indicated proteins in WT or STING KO ISG-THP1 cells stimulated with cGAMP (0.5 μmol/L) and/or podofilox for indicated times (D) or 4 hours (E). F and G, IFNβ production in ISG-THP1 cells (WT and STING KO) and BMDMs (WT and Stinggt/gt) in response to cGAMP (0.5 μmol/L) and/or podofilox treatment (24 hours for ISG-THP1; 6 hours for BMDMs, n = 3). I, ISG-THP1 cells were treated with cGAMP (0.5 μmol/L), LPS (1 μg), and Sendai virus (SeV, MOI = 0.1) for 24 hours with or without podofilox, and fold change of luminescence was normalized to DMSO-treated cells. J, A model showing whether podofilox potentiation is dependent on direct STING signaling or the indirect IFN and IFNAR effect. K, ISG-THP1 cells were pretreated in the absence or presence of anti-IFNAR2 antibody (10 μg/mL or 20 μg/mL) for 12 hours, and stimulated with cGAMP (0.5 μmol/L) or IFNβ (200 pg/mL) for 24 hours with or without podofilox (1 μmol/L). Fold change of luminescence was normalized to DMSO-treated cells. Data are representative of three independent experiments; each bar represents mean ± SEM. One-way ANOVA was applied for more than two data sets. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 1.

Discovery and characterization of podofilox as a cGAMP–STING signaling enhancer. A, Chemical structure of podofilox. B, C, and H, Cell-based activity of podofilox in WT and STING KO ISG-THP1 cells after treatment with cGAMP (0.5 μmol/L) and/or podofilox (indicated concentrations) for 24 hours; fold change of ISRE luciferase activity was normalized to DMSO-treated cells. D and E, Immunoblotting analysis of indicated proteins in WT or STING KO ISG-THP1 cells stimulated with cGAMP (0.5 μmol/L) and/or podofilox for indicated times (D) or 4 hours (E). F and G, IFNβ production in ISG-THP1 cells (WT and STING KO) and BMDMs (WT and Stinggt/gt) in response to cGAMP (0.5 μmol/L) and/or podofilox treatment (24 hours for ISG-THP1; 6 hours for BMDMs, n = 3). I, ISG-THP1 cells were treated with cGAMP (0.5 μmol/L), LPS (1 μg), and Sendai virus (SeV, MOI = 0.1) for 24 hours with or without podofilox, and fold change of luminescence was normalized to DMSO-treated cells. J, A model showing whether podofilox potentiation is dependent on direct STING signaling or the indirect IFN and IFNAR effect. K, ISG-THP1 cells were pretreated in the absence or presence of anti-IFNAR2 antibody (10 μg/mL or 20 μg/mL) for 12 hours, and stimulated with cGAMP (0.5 μmol/L) or IFNβ (200 pg/mL) for 24 hours with or without podofilox (1 μmol/L). Fold change of luminescence was normalized to DMSO-treated cells. Data are representative of three independent experiments; each bar represents mean ± SEM. One-way ANOVA was applied for more than two data sets. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

In addition, we validated that podofilox promoted activation of the STING–TBK1–IRF3 signaling cascade upon cGAMP treatment by analyzing the phosphorylation of TBK1, STING, and IRF3 (Fig. 1D; Supplementary Fig. S3D–S3G). Potentiation had already been observed within two hours after stimulation was initiated (Fig. 1D), but no activation was observed in STING-KO cells (Fig. 1E). These results further support that the podofilox-mediated enhancement of the IRF-induced immune response in ISG-THP1 cells was not caused by mitotic arrest or cytotoxicity. Moreover, podofilox significantly enhanced cGAMP-induced production of interferonβ (IFNβ) in ISG-THP1 WT cells but not in ISG-THP1 STING KO cells (Fig. 1F). Similarly, podofilox boosted cGAMP-triggered IFNβ production by WT mouse BMDMs but not Stinggt/gt BMDMs (Fig. 1G).

We further confirmed STING-dependent activity, which was augmented by podofilox, by performing a cell-based luciferase assay with ISG-THP1 cells (Fig. 1H) and assessing the expression of CCL5, CXCL10, IL-6, and IFITM1 in BMDMs (Supplementary Fig. S4A). To verify the specificity of podofilox action through the cGAMP–STING pathway but not the Toll-like receptor (TLR) signaling or RNA-sensing innate immune pathways, ISG-THP1 cells were stimulated with LPS, SeV, Poly(I:C) combined with podofilox. We found that podofilox enhanced the immune response after cGAMP treatment but not after stimulation of TLR- or RNA-sensing innate immune pathways under the test conditions (Fig. 1I; Supplementary Fig. S4B–S4E). Together, these experiments revealed that podofilox robustly and specifically enhanced the cGAMP–STING signaling pathway.

Because the ISRE-driven reporter is also downstream of type I IFNs and interferon-α/β receptor (IFNAR), we tested whether the ISRE-driven immune response was induced by direct STING signaling or the indirect IFN and IFNAR effect (Fig. 1J). We used an anti-IFNAR2 to block the secondary effect and found that human IFNβ (hIFNβ)-induced ISRE-reporter activity was blocked by the antibody (Fig. 1K). In contrast, the ISRE-driven immune response induced by cGAMP was not affected by anti-IFNAR2, indicating that direct STING–IRF3 signaling was induced by cGAMP treatment. Moreover, the effect of podofilox to enhance the cGAMP-induced ISRE response was still significant with anti-IFNAR2 treatment (Fig. 1K). These results suggested that podofilox enhanced cGAMP-mediated ISRE-driven immune activity mainly in a direct way. Importantly, podofilox was not able to enhance the ISRE-driven immune response when stimulated with IFNβ (Fig. 1K). We also found that podofilox only showed a tendency to enhance the ISRE-driven immune response when stimulated with RO8191, which is an agonist that directly binds to IFNAR2 and activates the expression of ISGs (Supplementary Fig. S4F). Phosphorylation of STING and IRF3 was direct evidence of cGAMP–STING signaling activation. We found that phosphorylation signaling with cGAMP treatment was enhanced by podofilox, but was not regulated by RO8191 or Stat1/Stat3 KO (Supplementary Fig. S4G–S4J). Together, these results suggested that podofilox directly boosted the cGAMP–STING signaling pathway.

Podofilox changed the trafficking pattern of STING induced by cGAMP

We explored the regulatory mechanism underlying the effects of podofilox on the cGAMP–STING pathway. One important feature of the STING pathway is dynamic STING trafficking upon activation mediated by cGAMP. Therefore, we examined whether podofilox can enhance the cGAMP–STING pathway by regulating STING trafficking. To examine the trafficking behavior of STING, we generated HeLa cells stably expressing STING-GFP or STING-110GFP. Because GFP that was attached to the C-terminus of STING (STING-GFP) impaired the interaction of STING with TBK1, we generated a STING-110GFP construct, which contained a GFP tag in the lumen region of STING (Supplementary Fig. S5A). Consistent with findings of our previous studies, cGAMP induced STING translocation and perinuclear puncta formation (Fig. 2A; Supplementary Fig. S5B and Supplementary Movie; refs. 8, 33). In contrast, podofilox changed the cGAMP–STING trafficking pattern, which led to the development of dispersed STING perinuclear puncta in small vesicles scattered throughout the cytoplasm (Fig. 2A; Supplementary Fig. S5B and Supplementary Movie). G10, which is a human-specific STING agonist (34), induced tiny STING puncta. It seems that these puncta were generated through a different pathway compared with cGAMP treatment, and they were not be changed by podofilox (Supplementary Fig. S5B and Supplementary Movie).

Figure 2.

Podofilox alters STING trafficking and promotes STING activity. A, HeLa cells stably expressing STING-110GFP were stimulated with cGAMP (2 μmol/L) and/or podofilox (1 μmol/L) for 2 hours. A specific inhibitor of protein trafficking, brefeldin A (BFA, 1 μmol/L), was used to inhibit STING trafficking. Nuclei were stained with Hoechst (blue). Representative confocal imaging of STING is shown. Scale bar, 10 μm. B, ISG-THP1 cells were stimulated with cGAMP (0.5 μmol/L) and/or podofilox for 4 hours. STING oligomerization was analyzed by native PAGE. Indicated proteins were detected by immunoblotting, and the results are representative of three independent biological replicates. C, HeLa cells (STING deficient) and HeLa hSTING cells stably expressing human STING were stimulated with cGAMP (0.5 μmol/L) and/or podofilox at the indicated time points. Indicated proteins were analyzed by immunoblotting. D, ISG-THP1 cells were treated with cGAMP (0.5 μmol/L) and/or podofilox (1 μmol/L) for 4 hours, intracellular cGAMP was quantified by LC/MS-MS. There was no significant difference of intracellular cGAMP level between cGAMP and cGAMP + podofilox group. E, HeLa cells stably expressing STING-GFP or STING-110GFP were treated with cGAMP (STING-GFP; 8 μmol/L, STING-110GFP; 2 μmol/L) and/or podofilox (1 μmol/L) for 2 hours. The STING (green) puncta are shown as orthogonal and 3D projections of Z-stack images. Scale bar, 5 μm. Quantifications of STING puncta volume are shown on the right of images (n = 20). F, BJ-5ta and MEF cells were stimulated with cGAMP (0.5 μmol/L) alone, or a combination of cGAMP and podofilox with or without brefeldin A (BFA, 1 μmol/L) for 2 hours, and whole cell lysates were harvested for immunoblotting. G and H, HeLa cells (STING-GFP or STING-110GFP) were treated (as in E), fixed, permeabilized, and stained for GM130 (a Golgi protein, red) and tubulin (red). Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI; blue). Scale bars, 10 μm. Quantitation of colocalization was calculated as Pearson correlation coefficient (r), and is shown on the right of each row of images (n = 50). Dashed white boxes in each main image indicate enlarged areas of interest shown below. All 3D-SIM images are z-stack images G, Image processing and quantitative analysis were performed in Imaris (version 9.7) and Image J (NIH, Research Service Branch, Bethesda, MD) software. cGAMP was delivered by 50 ng/mL perfringolysin O (A, C, E, F, H, G). Data are represented as means ± SEM, and are representative of three independent biological replicates. One-way ANOVA was applied for more than two data sets. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 2.

Podofilox alters STING trafficking and promotes STING activity. A, HeLa cells stably expressing STING-110GFP were stimulated with cGAMP (2 μmol/L) and/or podofilox (1 μmol/L) for 2 hours. A specific inhibitor of protein trafficking, brefeldin A (BFA, 1 μmol/L), was used to inhibit STING trafficking. Nuclei were stained with Hoechst (blue). Representative confocal imaging of STING is shown. Scale bar, 10 μm. B, ISG-THP1 cells were stimulated with cGAMP (0.5 μmol/L) and/or podofilox for 4 hours. STING oligomerization was analyzed by native PAGE. Indicated proteins were detected by immunoblotting, and the results are representative of three independent biological replicates. C, HeLa cells (STING deficient) and HeLa hSTING cells stably expressing human STING were stimulated with cGAMP (0.5 μmol/L) and/or podofilox at the indicated time points. Indicated proteins were analyzed by immunoblotting. D, ISG-THP1 cells were treated with cGAMP (0.5 μmol/L) and/or podofilox (1 μmol/L) for 4 hours, intracellular cGAMP was quantified by LC/MS-MS. There was no significant difference of intracellular cGAMP level between cGAMP and cGAMP + podofilox group. E, HeLa cells stably expressing STING-GFP or STING-110GFP were treated with cGAMP (STING-GFP; 8 μmol/L, STING-110GFP; 2 μmol/L) and/or podofilox (1 μmol/L) for 2 hours. The STING (green) puncta are shown as orthogonal and 3D projections of Z-stack images. Scale bar, 5 μm. Quantifications of STING puncta volume are shown on the right of images (n = 20). F, BJ-5ta and MEF cells were stimulated with cGAMP (0.5 μmol/L) alone, or a combination of cGAMP and podofilox with or without brefeldin A (BFA, 1 μmol/L) for 2 hours, and whole cell lysates were harvested for immunoblotting. G and H, HeLa cells (STING-GFP or STING-110GFP) were treated (as in E), fixed, permeabilized, and stained for GM130 (a Golgi protein, red) and tubulin (red). Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI; blue). Scale bars, 10 μm. Quantitation of colocalization was calculated as Pearson correlation coefficient (r), and is shown on the right of each row of images (n = 50). Dashed white boxes in each main image indicate enlarged areas of interest shown below. All 3D-SIM images are z-stack images G, Image processing and quantitative analysis were performed in Imaris (version 9.7) and Image J (NIH, Research Service Branch, Bethesda, MD) software. cGAMP was delivered by 50 ng/mL perfringolysin O (A, C, E, F, H, G). Data are represented as means ± SEM, and are representative of three independent biological replicates. One-way ANOVA was applied for more than two data sets. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

Brefeldin A, a widely used inhibitor of ER-to-Golgi protein trafficking, blocks the exit of STING from the ER (20). We showed that brefeldin A treatment effectively inhibited STING translocation (Fig. 2A). To verify the different STING trafficking patterns mediated by cGAMP and podofilox, we systematically examined the effects of STING trafficking and microtubule destabilizers that we identified in our screening experiments (Supplementary Fig. S2). We confirmed that all tested compounds led to dispersal of STING puncta induced by cGAMP (Supplementary Fig. S5C). We reasoned that STING signaling was sustained and enhanced in vesicles with accumulated STING. To determine the relevance of dispersed vesicles to STING signaling, we collected the cell membrane and examined STING oligomerization and activation using a native PAGE assay. We found that podofilox induced many more cGAMP-dependent STING oligomers than cGAMP treatment alone (Fig. 2B). Moreover, phosphorylated STING signaling was enhanced with podofilox (Fig. 2B). In addition, we generated HeLa cells stably expressing human STING or mouse STING. These cells were treated with cGAMP or DMXAA (a specific agonist of mouse STING; ref. 35). We found that podofilox promoted the phosphorylation of TBK1, STING, and IRF3 (Fig. 2C; Supplementary Fig. S5D).

To exclude the possibility that podofilox enhanced STING activation by mediating increased cGAMP entry into cells, we quantified intracellular cGAMP by MS (LC-MS/MS) showed that podofilox did not affect the entry of cGAMP into the cytosol (Fig. 2D). Moreover, we treated ISG-THP1 cells with different STING agonists, including 3′3′-cGAMP, c-di-GMP, c-di-AMP, and a nonnucleotide STING agonist described in previous publications (22, 36, 37). Podofilox enhanced the STING activation induced by these agonists (Supplementary Fig. S5E and S5F). In contrast, G10 did not activate STING–IRF3 signaling in ISG-THP1 cells that express the HAQ variant of human STING, because hSTING with the HAQ mutation does not respond to G10 stimulation, consistent with a previously published article (34). Accordingly, cotreatment with podofilox and G10 failed to induce enhancement of STING signaling in ISG-THP1 cells (Supplementary Fig. S5G and H). These results indicated that podofilox enhanced the activation of STING signaling triggered by multiple STING agonists, including cyclic dinucleotides (CDN).

We then assessed the number of STING puncta formed upon treatment with cGAMP with or without podofilox. The number of STING puncta induced by cGAMP with podofilox treatment was significantly greater than that induced by cGAMP treatment alone (Fig. 2E). These results suggested that podofilox induced more STING oligomers to constantly activate STING signaling. The effect of podofilox was confirmed in BJ-5ta and MEF cells, which express endogenous STING (Fig. 2F). In addition, studies have shown that microtubules transport a broad range of organelles and vesicles to maintain their proper distributions and functions (38). STING is a membrane protein distributed in the ER. After binding cGAMP, STING exits the ER and translocates to the Golgi (Fig. 2G; Supplementary Fig. S6A and S6B). We suspected that podofilox might lead to redistribution of STING vesicles by manipulating microtubule populations. Immunostaining showed that podofilox destroyed microtubules and Golgi apparatus (Fig. 2G and H; Supplementary Fig. S6B; ref. 39). We tested whether podofilox blocked the transfer of STING vesicles from ER to Golgi apparatus (Fig. 2H). Our results showed that podofilox only slightly attenuated the colocalization of STING vesicles with Golgi apparatus after cGAMP treatment (Fig. 2G; Supplementary Fig. S6B). The accumulated small STING puncta were distributed with the Golgi marker throughout the cytoplasm after cotreatment of podofilox and cGAMP. Together, these data suggested that podofilox changed cGAMP–STING trafficking and induced accumulation of STING puncta, thus, the number of STING oligomers increased, promoting the activation of STING signaling.

Podofilox augmented activation of the cGAMP–STING–TBK1 complex and delayed STING degradation

We then characterized the STING–TBK1 complex activation induced by cGAMP and podofilox in dispersed vesicles with STING accumulation. We transfected HEK293T cells with STING, STING-110GFP, and STING-GFP. We found that STING and STING-110GFP, but not STING-GFP, activated IFNβ expression and phosphorylation of STING and TBK1 upon cGAMP treatment (Fig. 3A and B). Podofilox enhanced immune signaling in this assay (Fig. 3A and B). These results suggested that cells expressing STING-110GFP could be used to study the dynamics of STING trafficking and activation of the STING–TBK1 complex. Upon cGAMP stimulation, phosphorylated TBK1 showed some overlap with STING puncta (Fig. 3C). STING degradation was detected over time, and the level of TBK1 phosphorylation decreased during this process (Fig. 3C; 120–240 minutes). In contrast, colocalization of small STING puncta with phosphorylated TBK1 was sustained after cGAMP and podofilox treatment (Fig. 3C). We also quantified the number of STING puncta and the level of phosphorylated TBK1. Podofilox significantly augmented and prolonged STING–TBK1 complex signaling (Fig. 3C).

Figure 3.

Podofilox augments cGAMP–STING–TBK1 signaling and delays STING degradation. A, HEK293T cells were transiently cotransfected with the secretory IFNβ-luciferase reporter plasmids and various STING plasmids (STING WT, STING-110GFP, and STING-GFP) for 12 hours. Cells were stimulated with cGAMP (0.5 μmol/L) with or without podofilox (100 nmol/L) for 24 hours. B, Immunoblotting was carried out to examine the phosphorylation levels of STING and TBK1 proteins (as in A). * indicates a nonspecific band. C, HeLa cells stably expressing STING-110GFP were stimulated with cGAMP (8 μmol/L) with or without podofilox (1 μmol/L) for indicated times, and then subjected to immunostaining for STING (green, 488 nm), p-TBK1 (red, 568 nm), and DAPI (blue). Scale bars, 5 μm. Statistics of fluorescence intensity were performed in Imaris software version 9.7 (n = 20). D and F–I, Immunoblotting analysis of STING degradation in BJ-5ta cells treated with cGAMP (8 μmol/L) with or without podofilox (1 μmol/L, pretreated for 30 minutes), brefeldin A (BFA, 1 μmol/L), and bafilomycin A1 (BafA1, 100 nmol/L) for 2 hours D, and STING stability was determined by immunoblotting at indicated time points in the absence or presence of cycloheximide (CHX, 50 μg/mL) F–I. Quantitative analysis of STING total protein was performed in Image J software (no pretreatment in H; pretreated with podofilox and CHX for 30 minutes in I). E, HeLa cells (STING-110GFP) were stimulated with cGAMP (2 μmol/L) and/or podofilox for 2 hours, and stained for LysoTracker Red. Scale bars, 10 μm. All 3D-SIM images are z-stack images. cGAMP was delivered by 50 ng/mL perfringolysin O (AI). Data are represented as means ± SEM, and are representative of three independent biological replicates. One-way ANOVA was applied for more than two data sets. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 3.

Podofilox augments cGAMP–STING–TBK1 signaling and delays STING degradation. A, HEK293T cells were transiently cotransfected with the secretory IFNβ-luciferase reporter plasmids and various STING plasmids (STING WT, STING-110GFP, and STING-GFP) for 12 hours. Cells were stimulated with cGAMP (0.5 μmol/L) with or without podofilox (100 nmol/L) for 24 hours. B, Immunoblotting was carried out to examine the phosphorylation levels of STING and TBK1 proteins (as in A). * indicates a nonspecific band. C, HeLa cells stably expressing STING-110GFP were stimulated with cGAMP (8 μmol/L) with or without podofilox (1 μmol/L) for indicated times, and then subjected to immunostaining for STING (green, 488 nm), p-TBK1 (red, 568 nm), and DAPI (blue). Scale bars, 5 μm. Statistics of fluorescence intensity were performed in Imaris software version 9.7 (n = 20). D and F–I, Immunoblotting analysis of STING degradation in BJ-5ta cells treated with cGAMP (8 μmol/L) with or without podofilox (1 μmol/L, pretreated for 30 minutes), brefeldin A (BFA, 1 μmol/L), and bafilomycin A1 (BafA1, 100 nmol/L) for 2 hours D, and STING stability was determined by immunoblotting at indicated time points in the absence or presence of cycloheximide (CHX, 50 μg/mL) F–I. Quantitative analysis of STING total protein was performed in Image J software (no pretreatment in H; pretreated with podofilox and CHX for 30 minutes in I). E, HeLa cells (STING-110GFP) were stimulated with cGAMP (2 μmol/L) and/or podofilox for 2 hours, and stained for LysoTracker Red. Scale bars, 10 μm. All 3D-SIM images are z-stack images. cGAMP was delivered by 50 ng/mL perfringolysin O (AI). Data are represented as means ± SEM, and are representative of three independent biological replicates. One-way ANOVA was applied for more than two data sets. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

To further define the process of STING degradation, we used a high concentration of cGAMP to trigger robust STING degradation (Fig. 3D; Supplementary Fig. S6C). Bafilomycin A1 (BafA1) and BFA blocked STING degradation. With podofilox treatment, the level of total STING was sustained and accumulated (Fig. 3D; Supplementary Fig. S6C). Moreover, perinuclear puncta of STING were surrounded and engulfed by lysosomes as a result of cGAMP treatment. In contrast, cotreatment with podofilox and cGAMP changed lysosome localization patterns (Fig. 3E; Supplementary Fig. S6D). Consistent with these data, in the absence or presence of cycloheximide, podofilox significantly delayed the degradation of STING stimulated with cGAMP over time in BJ-5ta cells with or without pretreatment (Fig. 3FI). Together, these experiments revealed that podofilox mediated a stronger and longer cGAMP–STING–TBK1 signaling cascade.

Podofilox enhanced STING activation and the antitumor effect of cGAMP in tumor organoids

We employed human patient tumor organoids as in vitro tumor models to evaluate whether podofilox enhances cGAMP-mediated immune response. Using the mechanical sample processing method developed previously (29), we first established lung cancer organoids from samples obtained during clinical surgical resection (Supplementary Fig. S7A and S7B). H&E staining demonstrated that the tumor organoids closely resembled the morphologies of the corresponding tumor tissues (Supplementary Fig. S7C). In addition, no immune cells, blood vessels, or other mesenchymal components were found in conjunction with the tumor cells (Supplementary Fig. S7B). The expression of p40 and p63, markers of lung squamous cell carcinoma, in the parental tumor tissue samples was similar to that in the matched tumor organoids (Supplementary Fig. S7C). Immunoblot results revealed that the lung cancer organoids responded to cGAMP stimulation, which activated the phosphorylation and signaling of TBK1 and STING (Supplementary Fig. S7D). Next, we employed an InSMAR chip developed previously to perform a drug sensitivity test (Supplementary Fig. S8A–S8E). Our previous study suggested that the results of drug tests performed on the chip were consistent with the data of patients’ genetic mutations, the clinical efficacy of treatment, and patient-derived xenograft (PDX) assessment results (29).

Similar to the enhancement effect of podofilox on the cell lines, the phosphorylation-related activation of STING signaling cascades induced by cGAMP was enhanced by podofilox treatment in lung cancer organoids (Supplementary Fig. S7E). The combination of cGAMP and podofilox had a strong inhibitory effect on controlling the growth of lung cancer organoids compared with clinical first-line chemotherapy regimens, and a weaker antitumor effect was observed with podofilox treatment alone (Supplementary Fig. S7F and S7H). Moreover, similar results were obtained with calcein AM staining (Supplementary Fig. S7G). In contrast, clinical chemotherapeutic agents showed strong drug resistance at a 1 μmol/L concentration in the lung cancer organoids, which partially explains the poor prognosis in cancer patients (Supplementary Fig. S7F). Our experiments showed that cGAMP and podofilox directly induced cytotoxicity by activating the STING pathway in tumor organoids (Supplementary Fig. S7F–S7H). Consistent with previous studies, prolonged activation of STING caused mitochondria or ER stress–mediated apoptosis in malignant immune cells and hepatocytes (40, 41). In addition, one study showed that slow accumulation of phosphorylated IRF3 triggered apoptosis during mitotic aberrations (42). Collectively, these results demonstrated that activation of the cGAMP–STING signaling pathway in tumor organoids derived from human patient samples could be regulated by podofilox.

Podofilox enhanced the antitumor effects of cGAMP through host STING activation in vivo

To test whether podofilox could enhance the antitumor activity of cGAMP in vivo, we subcutaneously inoculated immunocompetent WT C57BL/6 mice with B16F10 melanoma cells. Once tumors were established, different treatments were administered as indicated. Podofilox showed a negligible effect, while cGAMP treatment significantly prevented tumor growth (Fig. 4A and B), consistent with findings from our previous study and other studies (26, 43). A combination treatment of podofilox and cGAMP further reduced the tumor burden (Fig. 4A and B). As a result of reduced tumor growth, cGAMP-treated mice lived longer than those treated with vehicle or podofilox, while combination treatment with podofilox and cGAMP significantly prolonged mouse survival compared cGAMP treatment alone, as determined by survival curve analysis (Fig. 4C).

Figure 4.

Podofilox enhances cGAMP antitumor effects in vivo. A–C, WT C57BL/6 mice were inoculated with 1 × 106 B16F10 tumor cells and treated with PBS, cGAMP (10 μg/mice), podofilox (0.7 mg/kg), or cGAMP along with podofilox by intratumoral injection on days 7, 10, and 13 (n = 5). A, Tumor growth was measured every 3 days. B, Spider plot and groups of tumor growth curves are shown. C, Survival curves of B16F10 tumor-bearing mice after treatments are shown and were analyzed using the Kaplan–Meier method. D, WT mice were inoculated with 1 × 106 B16F10 tumor cells and treated on days 7 and 10. Tumors were collected 5 hours after treatment on day 10. RNA was extracted and gene expression of IFNβ and ISG was quantified by qPCR. E and F, Both WT (n = 5) and Stinggt/gt mice (n = 5) were inoculated with 1 × 106 B16F10 tumor cells and treated with vehicle or cGAMP (10 μg/mice) along with podofilox (0.7 mg/kg) by intratumoral injection on days 7, 10, and 13. Tumor growth was measured every 3 days. Survival curves of WT and Stinggt/gt tumor-bearing mice after treatments is described in F. G, WT C57BL/6 mice (n = 5) were inoculated with 1 × 106 B16F10 tumor cells (WT or STING KO) and treated with vehicle, cGAMP (10 mg/mice), or cGAMP along with podofilox (0.7 mg/kg) by intratumoral injection on days 7, 10, and 13. Tumor growth was measured every 3 days. Mice were humanely euthanized if severe ulceration occurred or tumor diameter reached 20 mm. Data are shown as means ± SEM and analyzed by two-way ANOVA followed by Tukey multiple comparisons test (in vivo tumor growth). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 4.

Podofilox enhances cGAMP antitumor effects in vivo. A–C, WT C57BL/6 mice were inoculated with 1 × 106 B16F10 tumor cells and treated with PBS, cGAMP (10 μg/mice), podofilox (0.7 mg/kg), or cGAMP along with podofilox by intratumoral injection on days 7, 10, and 13 (n = 5). A, Tumor growth was measured every 3 days. B, Spider plot and groups of tumor growth curves are shown. C, Survival curves of B16F10 tumor-bearing mice after treatments are shown and were analyzed using the Kaplan–Meier method. D, WT mice were inoculated with 1 × 106 B16F10 tumor cells and treated on days 7 and 10. Tumors were collected 5 hours after treatment on day 10. RNA was extracted and gene expression of IFNβ and ISG was quantified by qPCR. E and F, Both WT (n = 5) and Stinggt/gt mice (n = 5) were inoculated with 1 × 106 B16F10 tumor cells and treated with vehicle or cGAMP (10 μg/mice) along with podofilox (0.7 mg/kg) by intratumoral injection on days 7, 10, and 13. Tumor growth was measured every 3 days. Survival curves of WT and Stinggt/gt tumor-bearing mice after treatments is described in F. G, WT C57BL/6 mice (n = 5) were inoculated with 1 × 106 B16F10 tumor cells (WT or STING KO) and treated with vehicle, cGAMP (10 mg/mice), or cGAMP along with podofilox (0.7 mg/kg) by intratumoral injection on days 7, 10, and 13. Tumor growth was measured every 3 days. Mice were humanely euthanized if severe ulceration occurred or tumor diameter reached 20 mm. Data are shown as means ± SEM and analyzed by two-way ANOVA followed by Tukey multiple comparisons test (in vivo tumor growth). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

4T1 breast cancer cells are resistant to anti–PD-1/PD-L1 immune checkpoint immunotherapy and show minor responses to cGAMP treatment (44). However, the combination treatment of podofilox and cGAMP showed superior efficacy in inhibiting tumor growth compared with the corresponding monotherapies (Supplementary Fig. S9A). cGAMP binds STING to mediate type I IFN production and antitumor activity in vivo. Type I IFN expression was very low in vehicle- or podofilox-treated mice, while cGAMP treatment induced IFNβ and ISG expression (Fig. 4D). Podofilox further enhanced cGAMP-induced IFNβ and ISG expression in vivo (Fig. 4D).

We further investigated whether host STING was required for the antitumor effect of podofilox and cGAMP treatment. Although combined treatment with cGAMP and podofilox significantly reduced the tumor burden in WT mice, the antitumor efficacy was much weaker in Stinggt/gt mice, indicating that the antitumor effects depended on host STING and the host immune response (Fig. 4E and F). We then investigated whether STING in tumor cells contributed to the antitumor effect. WT and STING KO B16F10 melanoma cells were subcutaneously injected into immunocompetent WT C57BL/6 mice. cGAMP and cGAMP-podofilox treatments induced an effective response against both WT and STING KO B16F10 cells (Fig. 4G). However, the combination regimens showed slightly different effectiveness in inhibiting WT and STING-deficient B16F10 tumor growth. On the basis of the evidence presented above, we speculate that host STING signaling is the dominant factor promoting the antitumor effect of combination treatment with podofilox and cGAMP and that the contribution of tumor STING is relatively minor.

Podofilox promoted the cGAMP-mediated antitumor immune microenvironment

The tumor microenvironment has been widely implicated in antitumor immune responses. We investigated the impact of podofilox and cGAMP treatment on tumor-infiltrating lymphocytes (TIL), immune cell activation, and immunologic mechanisms using the B16F10 model. Intratumoral injection of cGAMP induced both CD8+ and CD4+ T-cell infiltration in tumors, while podofilox treatment induced only a few T cells to infiltrate tumors (Fig. 5A). The combination treatment of podofilox and cGAMP increased the number of CD8+ and/or CD4+ T cells in tumors to a greater extent than cGAMP treatment alone (Fig. 5A; Supplementary Fig. S9B). Infiltrated CD8+ and CD4+ T cells showed higher expression of CD69 after cGAMP treatment, while podofilox treatment further enhanced CD69 expression (Fig. 5B; Supplementary Fig. S9C), indicating that podofilox boosted cGAMP-induced activation of T cells within tumors. Combined treatment of podofilox and cGAMP also increased the presence of PD1 on tumor-infiltrating CD8+ and CD4+ T cells to the highest levels, suggesting that podofilox enhanced cGAMP-induced T-cell activation (Fig. 5C; Supplementary Fig. S9D). In addition, more natural killer (NK) cells infiltrated tumors when podofilox and cGAMP were administered together (Fig. 5D).

Figure 5.

Podofilox promotes cGAMP-mediating antitumor immunity. A–F, WT C57BL/6 mice were inoculated with 1 × 106 B16F10 tumor cells and treated with PBS, cGAMP (10 μg/mice), podofilox (0.7 mg/kg), or cGAMP combined with podofilox by intratumoral injection on days 7, 10, and 13 (n = 5). Cells were isolated from the tumor (tumor-infiltrating leukocytes; TIL) on day 18 and analyzed by flow cytometry. A, The population of CD3+CD8+ T cells in CD45+ leukocytes was measured. B and C, CD69 and PD1 expression on CD8+ T cells were assessed by flow cytometry. D, The population of natural killer (NK) cells in CD45+ leukocytes was measured. E, IFNγ, TNFα, and Granzyme B production in CD8+ T cells were tested. F, Granzyme B and IFNγ production in NK cells were tested. Dot plots are shown as means ± SEM and were analyzed by one-way ANOVA followed by Tukey multiple comparisons test. G, WT C57BL/6 mice were inoculated with 1 × 106 B16F10 tumor cells and treated with PBS, cGAMP (10 μg/mice), podofilox (0.7 mg/kg), and cGAMP combined with podofilox on days 7, 10, and 13 by intratumoral injection, while anti–PD-L1 (200 μg/mice) was given on days 7, 10, and 13 by intraperitoneal injection. Tumor growth was measured every 3 days (n = 5) and further analyzed by two-way ANOVA followed by Tukey multiple comparisons test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 5.

Podofilox promotes cGAMP-mediating antitumor immunity. A–F, WT C57BL/6 mice were inoculated with 1 × 106 B16F10 tumor cells and treated with PBS, cGAMP (10 μg/mice), podofilox (0.7 mg/kg), or cGAMP combined with podofilox by intratumoral injection on days 7, 10, and 13 (n = 5). Cells were isolated from the tumor (tumor-infiltrating leukocytes; TIL) on day 18 and analyzed by flow cytometry. A, The population of CD3+CD8+ T cells in CD45+ leukocytes was measured. B and C, CD69 and PD1 expression on CD8+ T cells were assessed by flow cytometry. D, The population of natural killer (NK) cells in CD45+ leukocytes was measured. E, IFNγ, TNFα, and Granzyme B production in CD8+ T cells were tested. F, Granzyme B and IFNγ production in NK cells were tested. Dot plots are shown as means ± SEM and were analyzed by one-way ANOVA followed by Tukey multiple comparisons test. G, WT C57BL/6 mice were inoculated with 1 × 106 B16F10 tumor cells and treated with PBS, cGAMP (10 μg/mice), podofilox (0.7 mg/kg), and cGAMP combined with podofilox on days 7, 10, and 13 by intratumoral injection, while anti–PD-L1 (200 μg/mice) was given on days 7, 10, and 13 by intraperitoneal injection. Tumor growth was measured every 3 days (n = 5) and further analyzed by two-way ANOVA followed by Tukey multiple comparisons test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Close modal

Effective cytokine production by TILs is very important for mediating antitumor immune effects. cGAMP or podofilox treatment alone induced TNFα production in CD8+ and CD4+ T cells but had no effect on IFNγ production, while combination treatment induced more TNFα- and IFNγ-producing CD8+ and CD4+ T cells (Fig. 5E; Supplementary Fig. S9E and S9F). Granzyme B produced by CD8+ T cells and NK cells is essential for tumor-killing effects. Although cGAMP treatment alone induced granzyme B expression in CD8+ T cells and NK cells, when combined with cGAMP treatment, podofilox further enhanced the production of granzyme B (Fig. 5E and F).

Because high expression of PD1 in T cells was induced by the combination treatment of podofilox and cGAMP, we further tested the effects of blocking PD1 signaling in combined podofilox and cGAMP treatment. Adding anti–PD-L1 treatment to combination podofilox and cGAMP treatment further reduced the tumor burden in both the B16F10 and 4T1 tumor models (Fig. 5G; Supplementary Fig. S9A), suggesting that blocking PD1 signaling in T cells further improved the tumor immune therapy mediated by cGAMP and podofilox. More control experiments, including combining anti–PD-L1 with podofilox or cGAMP, are needed to clearly characterize the effect of cGAMP/podofilox/PD-L1 in antitumor therapy. Taken together, our findings suggest that podofilox altered the cGAMP-mediated antitumor microenvironment, boosted antitumor immunity, and directly induced cytotoxicity of tumor organoids, suggesting a potential immunotherapeutic approach for cancer treatment by administering podofilox in combination with cGAMP and/or anti–PD-L1.

Here, we combined chemical screening, biochemical experiments, and multiple cell models to identify that a microtubule destabilizer, podofilox, enhanced the cGAMP–STING signaling pathway. The cGAS–STING pathway is critical DNA-sensing machinery in innate immunity. STING activation triggers multiple signaling cascades and leads to the expression of type I IFNs and other cytokines. The generation of type I IFN is important to tumor-specific T cells (45, 46), which renders STING a potential target for cancer immunotherapy. Despite the existence of productive cGAS–STING agonists, their antitumor efficacy is lacking due to hydrophilicity, negative charges, and susceptibility to enzymatic degradation. Moreover, STING signaling quickly turns off after it is activated by the agonists. Therefore, if a strategy could enhance the STING immune response and decrease the negative regulation of STING activation at the same time, it would have potential for cancer immunotherapy. Our results show that podofilox enhanced cGAMP-mediated immune responses by increasing STING puncta and promoting the oligomerization of STING. Imaging and immunoblotting evidence also showed that podofilox delayed STING degradation and prolonged STING–TBK1 cascades (Fig. 6). Moreover, combined treatment of podofilox and cGAMP increased the number of CD8+ and/or CD4+ T cells in a mouse B16F10 tumor model. These results reveal that podofilox could serve as a tool to boost the cGAMP–STING pathway. We also revealed that other commonly used microtubule destabilizers, including MMAE, colchicine, and vinblastine, enhanced the activation of cGAMP–STING signaling. As a proof of concept, we demonstrated that podofilox and cGAMP combination therapy boosted STING immune signaling and antitumor effects in vivo. We speculate that other microtubule destabilizers approved for clinical antitumor therapy, such as vinca alkaloids and MMAE, might provide remarkable antitumor effects by promoting activation of the cGAMP–STING pathway. Further studies are needed to determine the effects of different microtubule inhibitors in antitumor immunity. Together, this work sets the stage for investigating host adaptive antitumor immunity via pharmacologic activation of STING.

Figure 6.

Model of effects of podofilox on cGAMP signaling proposed as a result of this study: Podofilox enhances cGAMP-mediated immune response by increasing STING-containing membrane puncta and the oligomerization level of STING. STING binds with cGAMP and translocates it from the endoplasmic reticulum to the ERGIC and Golgi, and forms puncta in the perinuclear region. Podofilox disperses the STING-containing puncta to generate many STING-accumulated vesicles, where stronger STING oligomerization and phosphorylation of TBK1, STING, and IRF3 are induced. Moreover, podofilox delays/inhibits STING degradation, but the inhibition of STING degradation is not proved to be necessary to enhance STING signaling. Our data demonstrated that podofilox promoted STING signaling activation to induce more type I IFNs and stronger antitumor immunity.

Figure 6.

Model of effects of podofilox on cGAMP signaling proposed as a result of this study: Podofilox enhances cGAMP-mediated immune response by increasing STING-containing membrane puncta and the oligomerization level of STING. STING binds with cGAMP and translocates it from the endoplasmic reticulum to the ERGIC and Golgi, and forms puncta in the perinuclear region. Podofilox disperses the STING-containing puncta to generate many STING-accumulated vesicles, where stronger STING oligomerization and phosphorylation of TBK1, STING, and IRF3 are induced. Moreover, podofilox delays/inhibits STING degradation, but the inhibition of STING degradation is not proved to be necessary to enhance STING signaling. Our data demonstrated that podofilox promoted STING signaling activation to induce more type I IFNs and stronger antitumor immunity.

Close modal

Microtubule-targeting agents (MTA) are highly effective chemotherapeutic drugs used in the treatment of lung cancer, breast cancer, and hematologic malignancies (47, 48). MTAs are classified as microtubule destabilizers and microtubule stabilizers on the basis of the regulation of microtubule dynamics. Microtubule destabilizers, including nocodazole, vincristine sulfate, and podofilox, which were identified in our compound screening, promote microtubule depolymerization, whereas microtubule stabilizers, including paclitaxel and docetaxel, promote the polymerization of microtubules (49). In this study, we identified a series of microtubule destabilizers that regulate the cGAMP–STING signaling pathway through high-throughput cell-based phenotypic screening. We found that only the microtubule destabilizers (podofilox) but not stabilizers (paclitaxel and docetaxel) robustly enhanced the cGAMP–STING signaling cascade. This suggested that podofilox-induced enhancement of the cGAMP–STING pathway cannot solely be attributed to their shared antimitotic and cytotoxic effects. Moreover, our results revealed that podofilox directly regulated cGAMP–STING signaling, but not the indirect IFNAR signaling pathway. This is different from studies demonstrating that both microtubule destabilizers and stabilizers could promote activation of the cGAS–STING pathway through mitochondrial DNA release and micronucleus formation (50–52). Direct activation of the cGAS–STING pathway was relatively minor when cells were stimulated with microtubule inhibitors, likely due to the inhibitory effect of apoptosis on cGAS–STING signaling. In comparison, we found that podofilox robustly enhanced the cGAMP-mediated STING signaling pathway by changing the STING translocation pattern and increasing STING puncta (Fig. 6). Our biochemical and functional experiments revealed that podofilox promoted activation of the STING–TBK1–IRF3 signaling cascade in a cGAS-independent mode. Together, these findings provided mechanistic insights into how podofilox regulates innate immune signaling independent of its antimitotic or cytotoxic effects.

Currently, a major focus of immunotherapy is augmentation of innate immunity and fostering a sustained CD8+ T-cell-rich tumor environment. Pharmacologic activation of the cGAMP–STING signaling pathway is under investigation for this purpose with the aim of converting nonresponsive, ‘‘cold’’ tumors into responsive, ‘‘hot’’ tumors (53). Unfortunately, in clinical trials, two CDN-based STING agonists administered as a monotherapy or in combination with anti-PD1 were only effective in a minority of patients (54). In particular, 4T1 breast cancer cells with poor immunogenicity and a high metastatic rate are difficult to cure due to their resistance to immune checkpoint blockade (55). Consistent with our study, cGAMP has shown no antitumor effect on tumor organoids or in 4T1 tumor-bearing mouse models, which explains the failure of CDNs (including cGAMP) in clinical trials. Encouragingly, in mouse melanoma and breast tumor models, the combination of cGAMP and podofilox inhibited tumor growth via IFN-dependent T-cell activation and boosted T-cell infiltration or PD-L1 expression inside tumors, which may have the potential to convert a “cold” tumor into a “hot” tumor, thereby promoting the tumor response to PD-1/PD-L1 blockade and expanding the benefits of immunotherapies.

We previously reported that cGAMP can activate dendritic cells (DC) in vitro and in vivo and further boost antitumor effects (26). Therefore, although we cannot exclude other cell functions, it is highly possible that treatment with cGAMP and podofilox enhanced DCs antigen cross-presentation, costimulatory molecule expression, and cytokine release and that these effects further activate T cells in antitumor immune responses. Overall, our findings suggested that podofilox enhanced cGAMP–STING pathway–mediated type I IFN signaling, which in turn activated the tumor microenvironment and elicited antitumor immunotherapeutic effects in mouse tumor models.

No disclosures were reported.

J. Han: Data curation, software, investigation, methodology, writing–original draft, writing–review, and editing. S. Hu: Writing–original draft. Y. Hu: Writing–original draft. Y. Xu: Resources and formal analysis. Y. Hou: Formal analysis and methodology. Y. Yang: Methodology. H. Su: Formal analysis and methodology. Z. Zhang: Data curation, formal analysis, and methodology. P. Liu: Supervision. X. Sun: Data curation and supervision. C. Zhang: Data curation, supervision, project administration, writing–review, and editing.

This work is supported in part by the National Natural Science Foundation of China (project approval number: 32070875 and 82241074, to C.G. Zhang and 31970587 and 32170609 to X. Sun), and grants from the Tsinghua-Peking Center for Life Sciences and the Foundation of Shanghai Oriental Scholar (TP2018045). The study is also supported by the Beijing Natural Science Foundation (Z220018). We thank Xuedong Liu (CU-Boulder) for critically reading and providing suggestions to improve the manuscript, Hang Yin lab for sharing reagents, Juanjuan Du for critical discussion, and Liufu Deng for providing Stinggt/gt mice. High-throughput drug screening was performed at the Center of Pharmaceutical Technology, Tsinghua University.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.

Note: Supplementary data for this article are available at Cancer Immunology Research Online (http://cancerimmunolres.aacrjournals.org/).

1.
Gajewski
TF
,
Schreiber
H
,
Fu
YX
.
Innate and adaptive immune cells in the tumor microenvironment
.
Nat Immunol
2013
;
14
:
1014
22
.
2.
Demaria
O
,
Cornen
S
,
Daëron
M
,
Morel
Y
,
Medzhitov
R
,
Vivier
E
.
Harnessing innate immunity in cancer therapy
.
Nature
2019
;
574
:
45
56
.
3.
Cai
X
,
Chiu
YH
,
Chen
ZJ
.
The cGAS-cGAMP-STING pathway of cytosolic DNA sensing and signaling
.
Mol Cell
2014
;
54
:
289
96
.
4.
Ishikawa
H
,
Barber
GN
.
STING is an endoplasmic reticulum adaptor that facilitates innate immune signalling
.
Nature
2008
;
455
:
674
8
.
5.
Sun
L
,
Wu
J
,
Du
F
,
Chen
X
,
Chen
ZJ
.
Cyclic GMP-AMP synthase is a cytosolic DNA sensor that activates the type I interferon pathway
.
Science
2013
;
339
:
786
91
.
6.
Gao
P
,
Ascano
M
,
Wu
Y
,
Barchet
W
,
Gaffney Barbara
L
,
Zillinger
T
, et al
.
Cyclic [G(2′,5′)pA(3′,5′)p] is the metazoan second messenger produced by DNA-activated cyclic GMP-AMP synthase
.
Cell
2013
;
153
:
1094
107
.
7.
Wu
J
,
Sun
L
,
Chen
X
,
Du
F
,
Shi
H
,
Chen
C
, et al
.
Cyclic GMP-AMP is an endogenous second messenger in innate immune signaling by cytosolic DNA
.
Science
2013
;
339
:
826
30
.
8.
Zhang
C
,
Shang
G
,
Gui
X
,
Zhang
X
,
Bai
XC
,
Chen
ZJ
.
Structural basis of STING binding with and phosphorylation by TBK1
.
Nature
2019
;
567
:
394
8
.
9.
Zhang
X
,
Shi
H
,
Wu
J
,
Zhang
X
,
Sun
L
,
Chen
C
, et al
.
Cyclic GMP-AMP containing mixed phosphodiester linkages is an endogenous high-affinity ligand for STING
.
Mol Cell
2013
;
51
:
226
35
.
10.
Woo
SR
,
Fuertes
MB
,
Corrales
L
,
Spranger
S
,
Furdyna
MJ
,
Leung
MY
, et al
.
STING-dependent cytosolic DNA sensing mediates innate immune recognition of immunogenic tumors
.
Immunity
2014
;
41
:
830
42
.
11.
Woo
SR
,
Corrales
L
,
Gajewski
TF
.
Innate immune recognition of cancer
.
Annu Rev Immunol
2015
;
33
:
445
74
.
12.
Deng
L
,
Liang
H
,
Xu
M
,
Yang
X
,
Burnette
B
,
Arina
A
, et al
.
STING-dependent cytosolic DNA sensing promotes radiation-induced type I interferon-dependent antitumor immunity in immunogenic tumors
.
Immunity
2014
;
41
:
843
52
.
13.
Marcus
A
,
Mao
AJ
,
Lensink-Vasan
M
,
Wang
L
,
Vance
RE
,
Raulet
DH
.
Tumor-derived cGAMP triggers a STING-mediated interferon response in non-tumor cells to activate the NK cell response
.
Immunity
2018
;
49
:
754
63
.
14.
Carozza
JA
,
Böhnert
V
,
Nguyen
KC
,
Skariah
G
,
Shaw
KE
,
Brown
JA
, et al
.
Extracellular cGAMP is a cancer cell-produced immunotransmitter involved in radiation-induced anti-cancer immunity
.
Nat Cancer
2020
;
1
:
184
96
.
15.
Corrales
L
,
Glickman
LH
,
McWhirter
SM
,
Kanne
DB
,
Sivick
KE
,
Katibah
GE
, et al
.
Direct activation of STING in the tumor microenvironment leads to potent and systemic tumor regression and immunity
.
Cell Rep
2015
;
11
:
1018
30
.
16.
Demaria
O
,
De Gassart
A
,
Coso
S
,
Gestermann
N
,
Di Domizio
J
,
Flatz
L
, et al
.
STING activation of tumor endothelial cells initiates spontaneous and therapeutic antitumor immunity
.
Proc Natl Acad Sci U S A
2015
;
112
:
15408
13
.
17.
Dobbs
N
,
Burnaevskiy
N
,
Chen
D
,
Gonugunta
VK
,
Alto
NM
,
Yan
N
.
STING activation by translocation from the ER is associated with infection and autoinflammatory disease
.
Cell Host Microbe
2015
;
18
:
157
68
.
18.
Mukai
K
,
Konno
H
,
Akiba
T
,
Uemura
T
,
Waguri
S
,
Kobayashi
T
, et al
.
Activation of STING requires palmitoylation at the Golgi
.
Nat Commun
2016
;
7
:
11932
.
19.
Gui
X
,
Yang
H
,
Li
T
,
Tan
X
,
Shi
P
,
Li
M
, et al
.
Autophagy induction via STING trafficking is a primordial function of the cGAS pathway
.
Nature
2019
;
567
:
262
6
.
20.
Konno
H
,
Konno
K
,
Barber
GN
.
Cyclic dinucleotides trigger ULK1 (ATG1) phosphorylation of STING to prevent sustained innate immune signaling
.
Cell
2013
;
155
:
688
98
.
21.
Gonugunta
VK
,
Sakai
T
,
Pokatayev
V
,
Yang
K
,
Wu
J
,
Dobbs
N
, et al
.
Trafficking-mediated STING degradation requires sorting to acidified endolysosomes and can be targeted to enhance anti-tumor response
.
Cell Rep
2017
;
21
:
3234
42
.
22.
Ramanjulu
JM
,
Pesiridis
GS
,
Yang
J
,
Concha
N
,
Singhaus
R
,
Zhang
SY
, et al
.
Design of amidobenzimidazole STING receptor agonists with systemic activity
.
Nature
2018
;
564
:
439
43
.
23.
Pan
BS
,
Perera
SA
,
Piesvaux
JA
,
Presland
JP
,
Schroeder
GK
,
Cumming
JN
, et al
.
An orally available non-nucleotide STING agonist with antitumor activity
.
Science
2020
;
369
:
eaba6098
.
24.
Chin
EN
,
Yu
C
,
Vartabedian
VF
,
Jia
Y
,
Kumar
M
,
Gamo
AM
, et al
.
Antitumor activity of a systemic STING-activating non-nucleotide cGAMP mimetic
.
Science
2020
;
369
:
993
9
.
25.
Fu
J
,
Kanne
DB
,
Leong
M
,
Glickman
LH
,
McWhirter
SM
,
Lemmens
E
, et al
.
STING agonist formulated cancer vaccines can cure established tumors resistant to PD-1 blockade
.
Sci Transl Med
2015
;
7
:
283ra52
.
26.
Wang
H
,
Hu
S
,
Chen
X
,
Shi
H
,
Chen
C
,
Sun
L
, et al
.
cGAS is essential for the antitumor effect of immune checkpoint blockade
.
Proc Natl Acad Sci U S A
2017
;
114
:
1637
42
.
27.
Tan
Y
,
Kagan
JC
.
Innate immune signaling organelles display natural and programmable signaling flexibility
.
Cell
2019
;
177
:
384
98
.
28.
Feil
SC
,
Rossjohn
J
,
Rohde
K
,
Tweten
RK
,
Parker
MW
.
Crystallization and preliminary X-ray analysis of a thiol-activated cytolysin
.
FEBS Lett
1996
;
397
:
290
2
.
29.
Hu
Y
,
Sui
X
,
Song
F
,
Li
Y
,
Li
K
,
Chen
Z
, et al
.
Lung cancer organoids analyzed on microwell arrays predict drug responses of patients within a week
.
Nat Commun
2021
;
12
:
2581
.
30.
Zhang
P
,
Zhang
J
,
Bian
S
,
Chen
Z
,
Hu
Y
,
Hu
R
, et al
.
High-throughput superhydrophobic microwell arrays for investigating multifactorial stem cell niches
.
Lab Chip
2016
;
16
:
2996
3006
.
31.
Wang
Z
,
Chen
J
,
Hu
J
,
Zhang
H
,
Xu
F
,
He
W
, et al
.
cGAS/STING axis mediates a topoisomerase II inhibitor-induced tumor immunogenicity
.
J Clin Invest
2019
;
129
:
4850
62
.
32.
Hu
M
,
Zhou
M
,
Bao
X
,
Pan
D
,
Jiao
M
,
Liu
X
, et al
.
ATM inhibition enhances cancer immunotherapy by promoting mtDNA leakage and cGAS/STING activation
.
J Clin Invest
2021
;
131
:
e139333
.
33.
Shang
G
,
Zhang
C
,
Chen
ZJ
,
Bai
XC
,
Zhang
X
.
Cryo-EM structures of STING reveal its mechanism of activation by cyclic GMP-AMP
.
Nature
2019
;
567
:
389
93
.
34.
Banerjee
M
,
Middya
S
,
Shrivastava
R
,
Basu
S
,
Ghosh
R
,
Pryde
DC
, et al
.
G10 is a direct activator of human STING
.
PLoS One
2020
;
15
:
e0237743
.
35.
Kim
S
,
Li
L
,
Maliga
Z
,
Yin
Q
,
Wu
H
,
Mitchison
TJ
.
Anticancer flavonoids are mouse-selective STING agonists
.
ACS Chem Biol
2013
;
8
:
1396
401
.
36.
Woodward
JJ
,
Iavarone
AT
,
Portnoy
DA
.
c-di-AMP secreted by intracellular Listeria monocytogenes activates a host type I interferon response
.
Science
2010
;
328
:
1703
5
.
37.
Burdette
DL
,
Monroe
KM
,
Sotelo-Troha
K
,
Iwig
JS
,
Eckert
B
,
Hyodo
M
, et al
.
STING is a direct innate immune sensor of cyclic di-GMP
.
Nature
2011
;
478
:
515
8
.
38.
Barlan
K
,
Gelfand
VI
.
Microtubule-based transport and the distribution, tethering, and organization of organelles
.
Cold Spring Harb Perspect Biol
2017
;
9
:
a025817
.
39.
Cole
NB
,
Sciaky
N
,
Marotta
A
,
Song
J
,
Lippincott-Schwartz
J
.
Golgi dispersal during microtubule disruption: regeneration of Golgi stacks at peripheral endoplasmic reticulum exit sites
.
Mol Biol Cell
1996
;
7
:
631
50
.
40.
Tang
CH
,
Zundell
JA
,
Ranatunga
S
,
Lin
C
,
Nefedova
Y
,
Del Valle
JR
, et al
.
Agonist-mediated activation of STING induces apoptosis in malignant B cells
.
Cancer Res
2016
;
76
:
2137
52
.
41.
Petrasek
J
,
Iracheta-Vellve
A
,
Csak
T
,
Satishchandran
A
,
Kodys
K
,
Kurt-Jones
EA
, et al
.
STING-IRF3 pathway links endoplasmic reticulum stress with hepatocyte apoptosis in early alcoholic liver disease
.
Proc Natl Acad Sci U S A
2013
;
110
:
16544
9
.
42.
Zierhut
C
,
Yamaguchi
N
,
Paredes
M
,
Luo
JD
,
Carroll
T
,
Funabiki
H
.
The cytoplasmic DNA sensor cGAS promotes mitotic cell death
.
Cell
2019
;
178
:
302
15
.
43.
Li
T
,
Cheng
H
,
Yuan
H
,
Xu
Q
,
Shu
C
,
Zhang
Y
, et al
.
Antitumor activity of cGAMP via stimulation of cGAS-cGAMP-STING-IRF3 mediated innate immune response
.
Sci Rep
2016
;
6
:
19049
.
44.
Ohkuri
T
,
Kosaka
A
,
Ishibashi
K
,
Kumai
T
,
Hirata
Y
,
Ohara
K
, et al
.
Intratumoral administration of cGAMP transiently accumulates potent macrophages for anti-tumor immunity at a mouse tumor site
.
Cancer Immunol Immunother
2017
;
66
:
705
16
.
45.
Hemann
EA
,
Green
R
,
Turnbull
JB
,
Langlois
RA
,
Savan
R
,
Gale
M
., Jr.
Interferon-lambda modulates dendritic cells to facilitate T cell immunity during infection with influenza A virus
.
Nat Immunol
2019
;
20
:
1035
45
.
46.
Motwani
M
,
Pesiridis
S
,
Fitzgerald
KA
.
DNA sensing by the cGAS-STING pathway in health and disease
.
Nat Rev Genet
2019
;
20
:
657
74
.
47.
Tsai
AC
,
Wang
CY
,
Liou
JP
,
Pai
HC
,
Hsiao
CJ
,
Chang
JY
, et al
.
Orally active microtubule-targeting agent, MPT0B271, for the treatment of human non-small cell lung cancer, alone and in combination with erlotinib
.
Cell Death Dis
2014
;
5
:
e1162
.
48.
Dumontet
C
,
Jordan
MA
.
Microtubule-binding agents: a dynamic field of cancer therapeutics
.
Nat Rev Drug Discov
2010
;
9
:
790
803
.
49.
Steinmetz
MO
,
Prota
AE
.
Microtubule-targeting agents: strategies to hijack the cytoskeleton
.
Trends Cell Biol
2018
;
28
:
776
92
.
50.
Fermaintt
CS
,
Takahashi-Ruiz
L
,
Liang
H
,
Mooberry
SL
,
Risinger
AL
.
Eribulin activates the cGAS-STING pathway via the cytoplasmic accumulation of mitochondrial DNA
.
Mol Pharmacol
2021
;
100
:
309
18
.
51.
Mackenzie
KJ
,
Carroll
P
,
Martin
CA
,
Murina
O
,
Fluteau
A
,
Simpson
DJ
, et al
.
cGAS surveillance of micronuclei links genome instability to innate immunity
.
Nature
2017
;
548
:
461
5
.
52.
Lohard
S
,
Bourgeois
N
,
Maillet
L
,
Gautier
F
,
Fetiveau
A
,
Lasla
H
, et al
.
STING-dependent paracriny shapes apoptotic priming of breast tumors in response to anti-mitotic treatment
.
Nat Commun
2020
;
11
:
259
.
53.
Galon
J
,
Bruni
D
.
Approaches to treat immune hot, altered and cold tumours with combination immunotherapies
.
Nat Rev Drug Discov
2019
;
18
:
197
218
.
54.
Gajewski
TF
,
Higgs
EF
.
Immunotherapy with a sting
.
Science
2020
;
369
:
921
2
.
55.
Kim
K
,
Skora
AD
,
Li
Z
,
Liu
Q
,
Tam
AJ
,
Blosser
RL
, et al
.
Eradication of metastatic mouse cancers resistant to immune checkpoint blockade by suppression of myeloid-derived cells
.
Proc Natl Acad Sci U S A
2014
;
111
:
11774
9
.