AAZ2 induces mitochondrial-dependent apoptosis by targeting PDK1 in gastric cancer

AAZ2 induces mitochondrial-dependent apoptosis by targeting PDK1 in gastric cancer AAZ2可通过靶向PDK1介导线粒体依赖的凋亡 1 equal first-author 2 equal 1 1 1 1 1 1 3 1 1 corresp equal Tianhe HUANG, zn004593@whu.edu.cn Tianhe HUANG, zn004593@whu.edu.cn zn004593@whu.edu.cn https://orcid.org/0000-0002-2737-4016 1 corresp equal Yongchang WEI, weiyongchang@whu.edu.cn Yongchang WEI, weiyongchang@whu.edu.cn weiyongchang@whu.edu.cn https://orcid.org/0000-0001-7205-4663 武汉大学中南医院放化疗科， 湖北省肿瘤医学临床研究中心， 肿瘤生物学行为湖北省重点实验室， 中国武汉， 430071 武汉大学中南医院放化疗科， 湖北省肿瘤医学临床研究中心， 肿瘤生物学行为湖北省重点实验室， 中国武汉， 430071 Department of Radiation and Medical Oncology, Zhongnan Hospital of Wuhan University, Hubei Cancer Clinical Study Center & Hubei Key Laboratory of Tumor Biological Behaviors, Wuhan 430071, China Department of Radiation and Medical Oncology, Zhongnan Hospital of Wuhan University, Hubei Cancer Clinical Study Center & Hubei Key Laboratory of Tumor Biological Behaviors, Wuhan 430071, China 武汉大学化学与分子科学学院， 中国武汉， 430071 武汉大学化学与分子科学学院， 中国武汉， 430071 College of Chemistry and Molecular Science, Wuhan University, Wuhan 430072, China College of Chemistry and Molecular Science, Wuhan University, Wuhan 430072, China 武汉大学中南医院中西医结合科， 中国武汉， 430071 武汉大学中南医院中西医结合科， 中国武汉， 430071 Department of Integrated Chinese and Western Medicine, Zhongnan Hospital of Wuhan University, Wuhan University, Wuhan 430071, China Department of Integrated Chinese and Western Medicine, Zhongnan Hospital of Wuhan University, Wuhan University, Wuhan 430071, China Drastic surges in intracellular reactive oxygen species (ROS) induce cell apoptosis, while most chemotherapy drugs lead to the accumulation of ROS. Here, we constructed an organic compound, arsenical N -‍(4-(1,3,2-dithiarsinan-2-yl)phenyl)acrylamide (AAZ2), which could prompt the ROS to trigger mitochondrial-dependent apoptosis in gastric cancer (GC). Mechanistically, by targeting pyruvate dehydrogenase kinase 1 (PDK1), AAZ2 caused metabolism alteration and the imbalance of redox homeostasis, followed by the inhibition of phosphoinositide-3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway and leading to the activation of B-cell lymphoma 2 (Bcl2)/Bcl2-associated X (Bax)/caspase-9 (Cas9)/Cas3 cascades. Importantly, our in vivo data demonstrated that AAZ2 could inhibit the growth of GC xenograft. Overall, our data suggested that AAZ2 could contribute to metabolic abnormalities, leading to mitochondrial-dependent apoptosis by targeting PDK1 in GC. 大部分化疗药物可以促进活性氧（ROS）产生，同时ROS可以诱导细胞凋亡。本研究构建了一种有机砷化合物 N -(4-(1,3,2-dithiarsinan-2-yl)phenyl)acrylamide（AAZ2），其可通过促进ROS产生诱导胃癌线粒体依赖的细胞凋亡。具体机制为AAZ2通过靶向丙酮酸脱氢酶激酶1（PDK1）导致葡萄糖代谢改变和氧化应激，继而抑制PI3K/AKT/mTOR通路，最终激活半胱天冬酶（caspase）依赖的细胞凋亡。此外，体内实验也证实了AAZ2可以抑制胃癌移植瘤的生长。综上，在胃癌中，AAZ2可以通过靶向PDK1影响葡萄糖代谢及随后的氧化应激反应，抑制PI3K/AKT/mTOR信号通路，最终诱发线粒体依赖的细胞凋亡。 AAZ2 胃癌 活性氧（ROS） 细胞凋亡 丙酮酸脱氢酶激酶1（PDK1） 葡萄糖代谢 N -‍(4-‍(1,3,2-dithiarsinan-2-yl)phenyl)acrylamide (AAZ2) Gastric cancer Reactive oxygen species (ROS) Apoptosis Pyruvate dehydrogenase kinase 1 (PDK1) Glucose metabolism 1 Introduction 1 Gastric cancer (GC) is one of the leading causes of cancer-related deaths worldwide ( Smyth et al., 2020 Smyth et al., 2020 ; Joshi and Badgwell, 2021 Joshi and Badgwell, 2021 ). Our recent work discovered higher levels of reactive oxidative species (ROS) in GC patients, which initiate several signaling pathways to promote the development of GC ( Gu et al., 2018 Gu et al., 2018 ). Mild ROS levels trigger the proliferation of cancer cells, while high ROS content leads to cell apoptosis ( D'Souza et al., 2020 D'Souza et al., 2020 ; Harris and DeNicola, 2020 Harris and DeNicola, 2020 ). Thus, GC is potentially more vulnerable to oxidative stress-induced cell death when ROS-producing therapies are administered. Aberrant energy metabolism usually causes an imbalance of redox homeostasis and ROS accumulation ( Hayes et al., 2020 Hayes et al., 2020 ), suggesting that targeting the energy metabolism pathway may be an effective anti-tumor strategy. Some chemotherapy drugs could interrupt glucose metabolism and subsequently promote ROS production, leading to cancer cell death ( Hayes et al., 2020 Hayes et al., 2020 ). The pyruvate dehydrogenase complex/pyruvate dehydrogenase (PDH) kinase (PDC/PDK) axis could induce a “glycolytic shift” and has long been recognized as an anti-cancer therapeutic candidate ( Anwar et al., 2021 Anwar et al., 2021 ). PDK1 can inhibit the activity of PDH through the phosphorylation of the E1α subunit ( Anwar et al., 2021 Anwar et al., 2021 ; Wang et al., 2022 Wang et al., 2022 ). Studies have shown that PDK inhibitors like dichloroacetate (DCA), monochloracetate (MCA), and sodium pyruvate have a biochemical half maximal inhibitory concentration (IC 50 ) value ranging from 10 to 100 μmol/L ( Tataranni and Piccoli, 2019 Tataranni and Piccoli, 2019 ). Here, we speculated that acrylate (AA) and methacrylate (MAA) as pyruvate mimics could stimulate oxidative phosphorylation by inhibiting PDK1. Arsenic trioxide (ATO, As 2 O 3 ) is a compound that induces ROS production, causing oxidative stress, mitochondrial dysfunction, the activation of intracellular signaling pathways and DNA repair systems, and autophagy impairment ( Sanz et al., 2019 Sanz et al., 2019 ; Priya Wadgaonkar, 2021 Wadgaonkar, 2021 ), which was demonstrated as effective in solid tumors, including lymphoma, GC, liver cancer, and pancreatic cancer ( Wang QQ et al., 2020 Wang QQ et al., 2020 ). Researchers are increasingly attempting to construct organic arsenicals to reduce the side effects of ATO. Notably, compared with inorganic arsenicals, organic arsenicals like darinaparsin (DAR) ( Xu et al., 2019 Xu et al., 2019 ) and 4-‍( N -‍( S -glutathionylacetyl)amino) phenylarsonous acid (GSAO) ( Elliott et al., 2012 Elliott et al., 2012 ) have better structural diversity, stability, and bioavailability, and their clinical application value is increasingly appreciated. A phase II study found that DAR could be safely administered with tolerable toxicity profiles in advanced hepatocellular carcinoma ( Wu et al., 2010 Wu et al., 2010 ). The anti-tumor effects of GSAO have been validated in a phase I trial ( Elliott et al., 2012 Elliott et al., 2012 ). In this work, we also generated an organic arsenical N -(4-(1,3,2-dithiarsinan-2-yl)phenyl)acrylamide (AAZ2) and showed that it could prompt ROS levels to trigger mitochondrial-dependent apoptosis in GC both in vivo and in vitro. Mechanistically, AAZ2 targeted PDK1, which triggered metabolism alteration and the imbalance of redox homeostasis, followed by the inhibition of phosphoinositide-3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway, thus leading to the activation of B-cell lymphoma 2 (Bcl2)/Bcl2-associated X (Bax)/caspase-9 (Cas9)/Cas3 cascades. 2 Materials and methods 1 2.1 Cell culture and reagents 2 HGC-27 and AGS (human GC cell lines) were purchased from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). 293T and GES1 (cell lines) were gifts from Wuhan University. Cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 (Gibco, Grand Island, NY, USA) medium containing 10% (volume fraction) fetal bovine serum (FBS; Gibco) with penicillin-streptomycin under 5% CO 2 at 37 ℃. ATO (202673) and N -acetylcysteine (NAC, 138061) were purchased from Sigma-Aldrich (St. Louis, MO, USA); benzyloxycarbonyl-Val-Ala-Asp-fluoromethylketone (Z-VAD-FMK, S7023) from Selleck (Houston, Texas, USA); insulin-like growth factor 1 (IGF1, 100-11) from Peprotech (Rocky Hill, NJ, USA); LY294002 (HY-10108) from MCE (NJ, USA); 2-deoxy-2-‍[(7-nitro-2,‍1,‍3-benzoxadiazol-4-yl)amino]‍‍-D-glucose (2-DGNB, 186689-07-6) from Cayman (Michigan, USA). The antibodies used were β‍-tubulin (10068-1-AP), mTOR (20657-1-AP), phosphorylated AKT (p-AKT, 66444-1-Ig), AKT (10176-2-AP), Bax (50599-2-Ig), Bcl2 (12789-1-AP), Cas3 (19677-1-AP), and Cas9 (10380-1-AP) from Proteintech (Wuhan, China); PDK1 (ab110025) from Abcam (Cambridge, MA, USA); p-PDH (AP1022) from ABclonal (Wuhan, China); poly(adenosine diphosphate (ADP)‍-ribose) polymerase (PARP, #9532) and cleaved-Cas3 (cle-Cas3, #9664) from the Cell Signaling Technology (Danvers, MA, USA); PDH (385512), p-PI3K (310163), p-mTOR (381557), cle-PARP (380374), and cle-Cas9 (381238) from ZEN-BIOIENCE (Chengdu, China); PI3K (AF6241) from Affinity (Sydney, AUS). 2.2 Synthesis and characterization of Z2 and AAZ2 2 4-(1,‍3,‍2-dithiarsinan-2-yl)aniline (Z2) is the precursor drug of AAZ2. The synthesis process of Z2 was as follows: p -arsanilic acid (5 g, 18.45 mmol) and 70% (volume fraction) ammonium thioglycolate (10 mL, 13 g, 120 mmol) were added into a 25-mL round flask containing a magnetic stir bar at 50 ℃. After 4 h, 1,‍3-propanedithiol (4 mL, 23 mmol) was added dropwise and the mixture was stirred overnight. After the reaction completed, the mixture was extracted with dichloromethane (DCM). The organic layer was dried over anhydrous Na 2 SO 4 and purified by column chromatography on silica gel (petroleum ether (PE)/DCM=1/4, volume ratio) to afford PAO-PDT in 50% yield. Subsequently, acyl chlorination was carried out to synthesize the final products. For the synthesis process and characterization of AAZ2, a solution of Z2 (0.274 g, 1 mmol) in DCM (30 mL) was added to trimethylamine (TEA) (0.25 mL, 2 mmol) under N 2 atmosphere. After half an hour, acryloyl chloride was added dropwise at 0 ℃ followed by stirring for 2 h. The reaction mixture was quenched with H 2 O and extracted with CH 2 Cl 2 (3×10 mL). The organic layer was washed with brine, dried over anhydrous Na 2 SO 4 , concentrated under reduced pressure, and purified by column chromatography (PE/DCM=1/50, volume ratio) to afford AAZ2 as a white solid in 56% yield. 2.3 Cell viability assay 2 Cells were incubated in a 96-well plate overnight. After treated with AAZ2 (dissolved in dimethyl sulfoxide (DMSO), diluted in RPMI medium) for 24, 48, and 72 h, cholecystokinin-8 (CCK8) (Dojindo, Japan) was added followed by incubation, and the absorbance was read by a microplate reader (SpectraMax M2, Molecular Devices, USA). 2.4 Colony formation assay 2 Cells were incubated in six-well plates overnight, followed by 24 h of AAZ2 treatment. Next, cells were detached, centrifuged, re-counted, and re-plated, followed by 8 d of culture in normal culture medium. After washing with phosphate-buffered saline (PBS), methanol fixation, and 30 min of crystal violet staining, the colony numbers were calculated by ImageJ software ( Schneider et al., 2012 Schneider et al., 2012 ). 2.5 Determination of intracellular ROS accumulation 2 The treated samples were collected and stained with 10 μmol/L 2',‍7'-dichlorofluorescin diacetate (DCFH-DA; S0033S, Beyotime, China) or 5 μmol/L dihydroethidium (DHE) probe at 37 ℃ for 30 min. The ROS levels were finally determined by flow cytometry (BD, USA). 2.6 Analysis of cell apoptosis and cell cycle changes 2 Samples were harvested, washed, and stained with Annexin V/fluorescein isothiocyanate (FITC) and propidium iodide (PI) dye (CCS012, MultiSciences, China) for 5 to 10 min. The cell cycle was determined by staining with DNA staining and permeabilization solution. The final apoptotic rate and cell cycle distribution were detected using flow cytometry. 2.7 Transmission electron microscopy (TEM) 2 Samples treated and untreated with AAZ2 (0.8 μmol/L) were fixed with 2.5% (volume fraction) glutaraldehyde solution (G5882, Sigma-Aldrich) at 4 ℃ overnight. The dehydrated cells were embedded and stained with uranyl acetate. Representative areas were chosen from ultrathin sectioning and detected by TEM (Quantum Design, CA, USA). 2.8 Mitochondrial membrane potential (MMP) assay 2 Samples were stained with 5,5',6,6'-tetrachloro-1,1',3,3'-tetraethyl-imidacarbocyanine iodide (JC-1) dye (Cat#C2006, Beyotime) for 30 min, washed twice, and evaluated by flow cytometry. 2.9 Determination of lactic acid (LD) and adenosine triphosphate (ATP) production and glucose uptake level 2 Samples were harvested, disrupted, and centrifuged. Following the instructions of ATP detection kit (Cat# S0026, Beyotime), the supernatant and ATP working solution were incubated. Luminescence was measured by a microplate reader and quantified by relative light unit (RLU). The supernatants of the culture media were collected to evaluate the LD concentration following the detection kit (A019-2-1, Nanjing Jiancheng, China). As for glucose uptake capacity, cells were harvested and incubated with 100 mmol/L 2-DGNB (CAS186689-07-6, Cayman) for 30 min. Then, cells were washed, resuspended using glucose-free media, and detected by flow cytometry. 2.10 PDH activity 2 Samples were harvested and incubated with the reagents following the instructions of PDH reagent kit (PDH-1-Y, Suzhou Comin Biotechnology, China). The microplate reader (Molecular Devices) was used to detect PDK1 activity. 2.11 Molecular docking of AAZ2 to the PDK1 structural model 2 The simulation of docking interaction of AAZ2 and PDK1 was implemented using Schrödinger software ( https://www.schrodinger.com https://www.schrodinger.com ). The protein crystal structure of PDK1 (PBD code: 2Q8F) originated from the RCSB Protein Data Bank (PDB) database ( https://www.rcsb.org https://www.rcsb.org ) and was prepared by removing water, adding hydrogens, and performing energy minimization through the Maestro 11.0 platform of Schrödinger. The structure of AAZ2 was transformed, with energy minimized in Chem3D module and optimized in Schrödinger. The glide docking mode was applied to dock the prepared ligand and receptor. The default parameters were used to run the docking simulation. 2.12 Simulation of molecular dynamics (MD) 2 The complex of PDK1 and AAZ2 was simulated by Desmond 2020 ( https://www. https://www. ‍schrodinger.‍com/products/desmond). The third extended version of the optimized potentials for liquid simulations (OPLS3e) was chosen as the molecular force field, followed by energy minimization. Continuous systems were simulated by applying periodic boundary conditions. Long-range electrostatics were maintained by the particle mesh Ewald method. The system was equilibrated using an isothermal-isobaric ensemble (NPT) at a temperature of 300 K and a pressure of 1.0 bar (1 bar =0.1 MPa). The coupling of temperature-pressure parameters adopted the Berendsen coupling algorithm. The system ran for 100 ns at a time step of 1.2 fs. The track was recorded every 100 ps, and a total of 1000 frames were recorded. 2.13 Direct stochastic optical reconstruction microscopy (dSTORM) imaging 2 The treated and untreated samples were fixed with paraformaldehyde (15 min) and washed with PBS (three times). Next, cells were permeabilized with 0.2% (volume fraction) Triton X-100 (Sigma-Aldrich) for 10 min. After washing with PBS and blocking with 3% (volume fraction) bovine serum albumin (BSA; MCE) for 1 h, samples were incubated with primary antibody PDK1 at 4 ℃ overnight. The next day, samples were incubated with secondary antibodies for 40 min, followed by four times of washing with PBS. Prior to imaging, cells were soaked in standard dSTORM buffer (50 mmol/L Tris, pH 8.0, 10 mmol/L NaCl, 10 % (0.1 g/mL) glucose, 100 mmol/L mercaptoethylamine, 500 µg/mL glucose oxidase, and 40 µg/mL catalase). Then, cells were imaged using 405 nm laser under the same power setting. The exposure time was 10 ms for both the fluorescence and bright field (BF) images. Finally, QC-STORM ( Shang et al., 2021 Shang et al., 2021 ) was used to process the fluorescence images. 2.14 Western blot analysis 2 A total of 30 μg protein from each sample was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE; Shanghai Epizyme Biomedical Technology Co., Ltd., Shanghai, China) and transferred to a polyvinylidene fluoride (PVDF; Merck, Germany) membrane using wet transfer. After 60 min of blocking with 5% (0.05 g/mL) fat-free milk powder, the membrane was incubated with primary antibodies overnight at 4 ℃. The next day, the samples were washed with PBS and incubated with secondary antibodies for 1 h at room temperature. The final immune complex was detected using an enhanced chemiluminescence (ECL) detection kit (Bio-Rad, NY, USA). 2.15 Quantitative polymerase chain reaction (qPCR) 2 Evo M-MLV RT Kit (Thermo Fisher, MA, USA) was used to synthesize complementary DNA (cDNA). qPCR was carried out in CFX96 real-time PCR detection systems (CFX96) with SYBR Green reagent (MCE). The <math id="M1"><msup><mrow><mn mathvariant="normal">2</mn></mrow><mrow><mo>-</mo><mi mathvariant="normal">Δ</mi><mi mathvariant="normal">Δ</mi><msub><mrow><mi mathvariant="normal">C</mi></mrow><mrow><mi mathvariant="normal">T</mi></mrow></msub></mrow></msup></math> http://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=41072308&type= http://html.publish.founderss.cn/rc-pub/api/common/picture?pictureId=41072307&type= 7.61999989 2.96333337 method was applied to calculate the relative expression level of PDK1 to β-tubulin (Table S1). 2.16 Cell transfections for gene silencing or overexpression 2 In order to knock down the expression of PDK1, small interfering RNA (siRNA) targeted for PDK1 (GenePharma, Shanghai, China) (Table S1) and negative control were transfected using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). The media needed to be replaced with fresh normal culture media 6‍‒‍8 h later. After qPCR and western blot verification, cells could be used for subsequent experiments. To overexpress PDK1, a CMVenhancer-MCS-polyA-EF1A-zs-Green-sv40-puromycin vector with PDK1 was inserted and an empty plasmid vector (Genechem, Shanghai, China) was transfected with Lipofectamine 3000 (Invitrogen) and P3000 reagent (Invitrogen). 2.17 RNA sequencing (RNA-seq) 2 Total RNA from samples was extracted using the Trizol reagent (15596026, Thermo Scientific, MA, USA) under strict quality control with RNA integrity number (RIN) values of 6.10‍‒‍9.00. Subsequently, the ribosomal RNA (rRNA)-depletion method was selected to purify the samples, and RNA-seq libraries were constructed followed by an Illumina protocol (Illumina, San Diego, CA, USA). The Illumina HiSeq X Ten (PE150) platform was employed to sequence the libraries. 2.18 Immunohistochemistry (IHC) and staining with hematoxylin and eosin (HE) 2 The collected tumor tissues were fixed in 10% (volume fraction) formalin and embedded in paraffin. Specimens were sectioned at 5 µm thickness and stained using immunohistochemical techniques with primary antibodies against cle-Cas3 (1:100) and Ki67 (1‍:‍100). Secondary antibodies and diaminobenzidine were used for detection. The organs were stained with HE for further observation. 2.19 Animals and treatments 2 Five-week-old female BALB/c nu/nu mice (18‍‒22 g, n =21) were purchased from Vital River Laboratories (Beijing, China). The mice were randomly divided into three groups ( n =7): vehicle, oxaliplatin (L-OHP), and AAZ2. A total of 4.5×10 6 HGC27 cells were injected into the right flank in 200 µL of serum-free RPMI-1640. When the average tumor volume reached about 100 mm 3 , the mice were administered 10 mg/kg L-OHP, 10 mg/kg AAZ2, and vehicle every 5 d through intraperitoneal (i.p.) injection. The treatment duration lasted for 15 d. The tumor volume calculation formula was length×width 2 ×0.5. At the end of the study, all mice were sacrificed, and samples were harvested and processed for the subsequent assays. 2.20 Statistical analysis 2 Data from three independent experiments were presented as mean±standard deviation (SD). The GraphPad Prism 8.0 software (San Diego, USA) was applied for statistical analysis. When comparing more than two groups of data, one-way analysis of variance (ANOVA) followed by Dunnett’s post-hoc test was used; when comparing multiple independent groups, one-way ANOVA and non-parametric Kruskal-Wallis test were performed followed by Dunnett’s post-hoc test; when comparing two groups, unpaired Student’s t -test was adopted. Statistical significance was defined as P -value of < 0.05. 3 Results 1 3.1 Synthesis and cytotoxicity of AAZ2 2 We designed and synthesized three organic arsenicals: AAZ2, N -(4-(1,‍3,2-dithiarsinan-2-yl)phenyl)methacrylamide (MAZ2), and N -‍(4-(1,‍3,‍2-dithiarsinan-2-yl)phenyl)acetamide (ACZ2) ( Fig. 1 Fig. 1 a). All of these compounds were characterized by nuclear magnetic resonance (NMR) spectroscopy and full-scan mass spectrometry (MS) (Fig. S1). AAZ2 was chosen as the candidate for its larger cytotoxicity to GC cells ( Fig. 1 Fig. 1 b). The AGS and HGC27 were sensitive to AAZ2 ( Fig. 1 Fig. 1 c). Thus, these two cell lines were selected for subsequent experiments. In addition, compared to the same concentration of ATO (the classical inorganic arsenic), AAZ2 was more cytotoxic to GC cells ( Fig. 1 Fig. 1 d) and less cytotoxic to normal cells ( Fig. 1 Fig. 1 e), including 293T and GES1. AAZ2 could also dramatically reduce the colony numbers (Figs. 1 1 f and 1 1 g) and induce G2/M phase arrest (Figs. 1 1 h and 1 1 i) in a dose-dependent manner. 1;2; 3.2 Mitochondrial-dependent apoptosis induced by AAZ2 via ROS generation 2 Next, we attempted to explore the potential mechanism of AAZ2-induced cell death. Since ATO induced cell apoptosis through ROS accumulation, we first detected whether our organic arsenical AAZ2 could increase the intracellular ROS levels in GC cells. By means of DCFH-DA staining, we confirmed that AAZ2 could increase cellular ROS in a time- and dose-dependent manner ( Fig. 2 Fig. 2 a), and ROS scavenger NAC could rescue ROS generation ( Fig. 2 Fig. 2 b). Using TEM, the apoptotic bodies could be observed after AAZ2 treatment ( Fig. 2 Fig. 2 c). Since ROS play a significant role in determining mitochondrial-dependent apoptosis, we detected the effects of AAZ2 on the MMP and the mitochondrial-dependent apoptosis cascades. The flow cytometry results showed that AAZ2 treatment significantly decreased MMP, which could be attenuated by NAC, as shown by JC-1 staining ( Fig. 2 Fig. 2 d). By means of Annexin V-FITC/PI staining, we also determined that AAZ2-induced apoptosis could be rescued by NAC ( Fig. 2 Fig. 2 e). Moreover, AAZ2 treatment altered the mitochondrial-mediated apoptotic pathways by upregulating the expression of cle-PARP, cle-Cas9, and cle-Cas3, and downregulating Bcl2 and Bax expression. Importantly, NAC could diminish the effect of AAZ2 on mitochondrial-mediated apoptotic pathways ( Fig. 2 Fig. 2 f). Moreover, the pan-caspase inhibitor Z-VAD-FMK could attenuate AAZ2-induced apoptosis (Figs. 2 2 g and 2 2 h). Our data so far suggested that AAZ2 induced mitochondrial-dependent apoptosis via ROS generation. 3.3 ROS generation triggered by AAZ2 through targeting PDK1-mediated metabolism alteration 2 Later, we tried to figure out how AAZ2 promoted ROS generation. PDKs are major regulatory enzymes in glucose metabolism for their negative control of PDC by phosphorylation ( Anwar et al., 2021 Anwar et al., 2021 ), and AAZ2 was initially designed to target PDK1. Molecular simulation and molecular dynamics (MD) simulation were performed to verify the interaction between AAZ2 and PDK1. The docking models showed that AAZ2 could bind to PDK1 directly with a binding energy of -6.61 kcal/mol ( Fig. 3 Fig. 3 a), which was better than the interaction between PDK1 and DCA (a commonly used PDK1 inhibitor) with a binding energy of -5.15 kcal/mol. Besides, MD simulation found that the root mean square deviation (RMSD) curve was less than 2.1 Å and reached dynamic equilibrium within 10 ns ( Fig. 3 Fig. 3 b), demonstrating a robust and well-matched binding interaction between AAZ2 and PDK1. The western blot experiment verified that AAZ2 dramatically lowered the protein levels of PDK1, phosphorylated E1α subunit of PDH (p-PDH E1α) ( Fig. 3 Fig. 3 c) and hypoxia-inducible factor-1α (HIF-1α) (Fig. S2a), and slightly increased the non-phosphorylated PDH E1α in a dose-dependent manner in GC cells ( Fig. 3 Fig. 3 c). Using dSTORM, we discovered that the PDK1 density was 12.6–32.1 molecules/µm 2 before treatment and decreased to 9.4–28.8 molecules/µm 2 after 24 h of AAZ2 exposure (Figs. 3 3 d and 3 3 e). More importantly, AAZ2 treatment could increase the activity of PDH, which is downstream of PDK1 ( Fig. 3 Fig. 3 f). These data advocated that PDK1 was a direct downstream target of AAZ2 in GC cells. 1;2;3; Given that the PDC/PDK axis can induce a “glycolytic shift,” we next analyzed whether AAZ2 changed the metabolism of GC. It was found that AAZ2 could decrease glucose uptake, LD generation, and ATP level (Figs. 3g‒‍3i), while lower AAZ2 levels (0.1 and 0.2 μmol/L) could rarely induce gastric cell death but could also induce this kind of shift ( Figs. 2 Figs. 2 b‒2d). These data indicated that AAZ2-mediated PDK inhibition contributed to the glucose metabolism shift from glycolysis and LD generation to glucose oxidation ( Fig. 3 Fig. 3 j). Moreover, we tested the effects of ATO in GC cells. ATO failed to lead to a metabolic shift, as shown by the stable glucose uptake level and ATP and LD production (Figs. S2b‒S2d). Correspondingly, ATO showed no effect on PDK1 (Fig. S2e) and PDH activity (Fig. S2f), suggesting that AAZ2 showed stronger anti-tumor effects than ATO, and that AAZ2 can function by targeting the specific target PDK1. However, the relationship between ROS generation and AAZ2/PDK1-induced metabolism alteration remained to be explored. Thus, we built PDK1 overexpressing GC cells (Figs. 3 3 k and 3 3 l). Overexpression of PDK1 could reverse AAZ2-induced ROS generation ( Fig. 3 Fig. 3 m), desensitize the GC to AAZ2 ( Fig. 3 Fig. 3 n), and abate AAZ2-incuded apoptosis rates ( Fig. 3 Fig. 3 o). Conversely, the knockdown of PDK1 (Figs. 3 3 p and 3 3 q) enhanced AAZ2-triggered ROS accumulation ( Fig. 3 Fig. 3 r) and sensitized cells for the cytotoxic effect of AAZ2 (Figs. 3 3 s and 3 3 t). Taken together, our data supported that, by targeting PDK1, AAZ2 induced glycolytic phenotype alteration and subsequently led to ROS generation. 3.4 Mitochondrial-dependent apoptosis activated by AAZ2 through the inhibition of ROS-mediated PI3K/AKT/mTOR pathway 2 In order to better understand how ROS induce mitochondrial-dependent apoptosis, we performed RNA-seq for cells treated or untreated with AAZ2. AAZ2-treated cells showed a different gene profiling pattern ( Fig. 4 Fig. 4 a). Gene set enrichment analysis (GSEA) demonstrated that differential genes mainly concentrated on apoptosis and glucose starvation ( Fig. 4 Fig. 4 b). The PI3K/AKT/mTOR signaling pathway is one of the most well-known ROS-regulated pathways ( Perillo et al., 2020 Perillo et al., 2020 ), playing a vital role in tumor progression through proliferation enhancement, apoptosis inhibition, and glycolysis activation ( Fattahi et al., 2020 Fattahi et al., 2020 ; Tewari et al., 2022 Tewari et al., 2022 ). As expected, AAZ2 treatment decreased the protein levels of p-AKT, p-PI3K, and p-mTOR, and the inhibitory effects could be attenuated by NAC ( Fig. 4 Fig. 4 c), suggesting that AAZ2/PDK1/ROS-induced cell apoptosis is mediated by the suppression of PI3K/AKT/mTOR pathways. In addition, the PI3K/AKT/mTOR pathway activator IGF1 and inhibitor LY294002 were used to test the specific effect of AAZ2. Pre-treatment with IGF1 slightly decreased the AAZ2-induced apoptosis (Figs. 4 4 d and 4 4 e) and increased cell viability ( Fig. 4 Fig. 4 f). Correspondingly, the inhibitor LY294002 further enhanced apoptosis (Figs. 4 4 g and 4 4 h) and reduced cell viability ( Fig. 4 Fig. 4 i). Collectively, we concluded that AAZ2/PDK1 induced the metabolism alteration and subsequent ROS accumulation, followed by suppressing the PI3K/AKT/mTOR signaling pathway to mediate mitochondrial-dependent apoptosis. 1;2; 3.5 HGC27 xenograft growth inhibited by AAZ2 in vivo 2 The in vivo effect of AAZ2 on GC was determined using a HGC27 xenograft model ( Fig. 5 Fig. 5 a). AAZ2 treatment could reduce both the volume and weight (Figs. 5 5 b and 5 5 c) of HGC27 xenograft tumors. No significant differences were detected in the body weight ( Fig. 5 Fig. 5 d), hematological and biochemical parameters, including white blood cell (WBC), hemoglobin (HGB), platelet (PLT), alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine, or urea nitrogen ( Fig. 5 Fig. 5 e), as well as organ morphogenesis of liver, heart, and kidney ( Fig. 5 Fig. 5 f) of mice, suggesting that AAZ2 had no significant side effects on nude mice. Consistent with the in vitro results, IHC staining indicated that AAZ2 decreased the expression level of Ki67 and increased that of cle-Cas3 ( Fig. 5 Fig. 5 g). When using tumor tissue, we obtained similar trends as in vitro, including the change of PDK1 and mitochondrial-mediated apoptosis pathways ( Fig. 5 Fig. 5 h). By analyzing the public dataset, we discovered that PDK1 was highly expressed in many cancers, including in GC. High levels of PDK1 in GC patients were positively correlated with poor overall survival (Figs. 5 5 i and 5 5 j). Thus, by targeting PDK1, AAZ2 showed promising anti-tumor activity with low toxicity to other organs and could be selected as a candidate reagent for cancer treatment. 4 Discussion 1 In this study, we constructed an organic arsenical, AAZ2, and demonstrated that it could prompt the ROS to trigger mitochondrial-dependent apoptosis in GC both in vitro and in vivo. By targeting PDK1, AAZ2 induced glycolytic phenotype alteration and subsequent ROS generation. Increased ROS suppressed the PI3K/AKT/mTOR signaling pathway, which induced mitochondrial-dependent apoptosis ( Fig. 6 Fig. 6 ). PDK1 is highly expressed and associated with poorer overall survival in patients with GC ( Lin et al., 2019 Lin et al., 2019 ). PDK1 has long been regarded as a potential therapeutic target that can negatively regulate PDC activity followed by the metabolic shift in cancer cells ( Missiroli et al., 2020 Missiroli et al., 2020 ). Pai et al. (2021) Pai et al. (2021) reported that BX795, a PDK1 inhibitor, could enhance the efficacy of cisplatin and radiotherapy in oral squamous cell carcinoma. Anwar et al. (2021) Anwar et al. (2021) demonstrated that the inhibition of PDK1 could lead to the removal of the glycolytic energy source and mandate the cancer cells for using mitochondrial resources, which induced toxic ROS and made cells prone to death signals. Here, docking and MD simulation showed a good binding capacity between PDK1 and AAZ2. Meanwhile, dSTORM, a single-molecule localization-based super-resolution microscopy ( Nerreter et al., 2019 Nerreter et al., 2019 ; García-Guerrero et al., 2021 García-Guerrero et al., 2021 ), indicated that AAZ2 inhibited PDK1 expression in GC cells. AAZ2 also increased PDH activity and inhibited glucose metabolism. Furthermore, knockdown or overexpression of PDK1 could potentiate or attenuate the AAZ2-induced ROS generation and apoptosis, respectively. These data support that AAZ2 could inactivate PDK1, followed by glycolytic phenotype alteration and subsequent ROS generation. It is worth noting that AAZ2 could simultaneously decrease PDK1 expression and activity, indicating that inhibition is not the only mechanism by which AAZ2 reduces the PDK1 function. It has been reported that DCA (a classical PDK1 inhibitor) could reduce the protein levels of PDK1 in cancer cells ( Velpula et al., 2013 Velpula et al., 2013 ; Sradhanjali et al., 2017 Sradhanjali et al., 2017 ) and promote the degradation of HIF-1α, which in turn downregulated the expression of PDK1, suggesting a mechanism whereby DCA reduces the abundance of PDK1 via HIF-1α ( Škorja Milić et al., 2021 Škorja Milić et al., 2021 ). Similar to DCA, AAZ2 is a pyruvate mimic and could inhibit the expression of HIF-1α (Fig. S2a). Excessive intracellular levels of ROS can damage the syntheses of lipids, proteins, and DNA, induce drug resistance ( Cui et al., 2018 Cui et al., 2018 ), and even cause cell death, including autophagy and apoptosis ( Weng et al., 2018 Weng et al., 2018 ; Perillo et al., 2020 Perillo et al., 2020 ). Jin et al. (2020) Jin et al. (2020) reported that a PDK1 inhibitor, hemistepsin A, bound to the lipoamide-binding site of PDK1 and subsequently increased ROS generation and glycolysis inhibition, which induced ROS-dependent mitochondrial apoptotic both in vivo and in vitro. In this study, AAZ2 caused the accumulation of ROS as early as 15 min after AAZ2 treatment. Meanwhile, decreased MMP and increased apoptosis rates were observed, which could almost be attenuated by ROS scavengers NAC and Z-VAD-FMK, suggesting that AAZ2-mediated apoptosis was caspase-dependent in GC cells. Collectively, it can be reasonably concluded that AAZ2 exerts an anti-GC effect that is dependent on PDK1 inhibition and the subsequent accumulation of excessive ROS. The PI3K/AKT/mTOR pathway is one of the major cellular signaling pathways that mediate essential intracellular functions and is tightly associated with ROS ( Fattahi et al., 2020 Fattahi et al., 2020 ; Wang WP et al., 2020 Wang WP et al., 2020 ). In our study, RNA-seq suggested that differential genes were enriched in apoptosis and glucose starvation. Therefore, the PI3K/AKT/mTOR pathway was evaluated for AAZ2-mediated anti-GC effects. The western blot showed that AAZ2 could downregulate the levels of PI3K/AKT/mTOR pathway-related proteins. NAC could reverse the inhibitory effects, suggesting that ROS are involved in inhibiting the PI3K/AKT/mTOR pathway. The PI3K/AKT/mTOR pathway inhibitor LY294002 could potentiate AAZ2-mediated apoptosis, and activator IGF1 could partially attenuate apoptosis. All these data supported that ROS-mediated PI3K/AKT/mTOR pathway inhibition was involved in the process of AAZ2-mediated apoptosis. 5 Conclusions 1 In the present study, a novel organic arsenical, AAZ2, was constructed, which could inhibit the proliferation and growth of GC both in vitro and in vivo. By targeting PDK1, AAZ2 caused the metabolism alteration and imbalance of redox homeostasis, followed by the inhibition of PI3K/AKT/mTOR pathway, thus leading to the activation of mitochondrial-dependent apoptosis cascades. Therefore, AAZ2 may serve as a potential treatment agent for GC. References Anwar S , Shamsi A , Mohammad T , et al. , 2021 . Targeting pyruvate dehydrogenase kinase signaling in the development of effective cancer therapy . Biochim Biophys Acta Rev Cancer , 1876 ( 1 ): 188568 . https://doi.org/10.1016/j.bbcan.2021.188568 https://doi.org/10.1016/j.bbcan.2021.188568 Targeting pyruvate dehydrogenase kinase signaling in the development of effective cancer therapy Cui QB , Wang JQ , Assaraf YG , et al. , 2018 . Modulating ROS to overcome multidrug resistance in cancer . Drug Resist Updat , 41 : 1 - 25 . https://doi.org/10.1016/j.drup.2018.11.001 https://doi.org/10.1016/j.drup.2018.11.001 Modulating ROS to overcome multidrug resistance in cancer D'Souza LC , Mishra S , Chakraborty A , et al. , 2020 . Oxidative stress and cancer development: are noncoding RNAs the missing links? Antioxid Redox Signal , 33 ( 17 ): 1209 - 1229 . https://doi.org/10.1089/ars.2019.7987 https://doi.org/10.1089/ars.2019.7987 Oxidative stress and cancer development: are noncoding RNAs the missing links? Elliott MA , Ford SJ , Prasad E , et al. , 2012 . Pharmaceutical development of the novel arsenical based cancer therapeutic GSAO for Phase I clinical trial . Int J Pharm , 426 ( 1-2 ): 67 - 75 . https://doi.org/10.1016/j.ijpharm.2012.01.024 https://doi.org/10.1016/j.ijpharm.2012.01.024 Pharmaceutical development of the novel arsenical based cancer therapeutic GSAO for Phase I clinical trial Fattahi S , Amjadi-Moheb F , Tabaripour R , et al. , 2020 . PI3K/AKT/mTOR signaling in gastric cancer: epigenetics and beyond . Life Sci , 262 : 118513 . https://doi.org/10.1016/j.lfs.2020.118513 https://doi.org/10.1016/j.lfs.2020.118513 PI3K/AKT/mTOR signaling in gastric cancer: epigenetics and beyond García-Guerrero E , Götz R , Doose S , et al. , 2021 . Upregulation of CD38 expression on multiple myeloma cells by novel HDAC6 inhibitors is a class effect and augments the efficacy of daratumumab . Leukemia , 35 : 201 - 214 . https://doi.org/10.1038/s41375-020-0840-y https://doi.org/10.1038/s41375-020-0840-y Upregulation of CD38 expression on multiple myeloma cells by novel HDAC6 inhibitors is a class effect and augments the efficacy of daratumumab Gu HF , Huang TH , Shen YC , et al. , 2018 . Reactive oxygen species-mediated tumor microenvironment transformation: the mechanism of radioresistant gastric cancer . Oxid Med Cell Longev , 2018 : 5801209 . https://doi.org/10.1155/2018/5801209 https://doi.org/10.1155/2018/5801209 Reactive oxygen species-mediated tumor microenvironment transformation: the mechanism of radioresistant gastric cancer Harris IS , DeNicola GM , 2020 . The complex interplay between antioxidants and ROS in cancer . Trends Cell Biol , 30 ( 6 ): 440 - 451 . https://doi.org/10.1016/j.tcb.2020.03.002 https://doi.org/10.1016/j.tcb.2020.03.002 The complex interplay between antioxidants and ROS in cancer Hayes JD , Dinkova-Kostova AT , Tew KD , 2020 . Oxidative stress in cancer . Cancer Cell , 38 ( 2 ): 167 - 197 . https://doi.org/10.1016/j.ccell.2020.06.001 https://doi.org/10.1016/j.ccell.2020.06.001 Oxidative stress in cancer Jin L , Kim EY , Chung TW , et al. , 2020 . Hemistepsin A suppresses colorectal cancer growth through inhibiting pyruvate dehydrogenase kinase activity . Sci Rep , 10 : 21940 . https://doi.org/10.1038/s41598-020-79019-1 https://doi.org/10.1038/s41598-020-79019-1 Hemistepsin A suppresses colorectal cancer growth through inhibiting pyruvate dehydrogenase kinase activity Joshi SS , Badgwell BD , 2021 . Current treatment and recent progress in gastric cancer . CA Cancer J Clin , 71 ( 3 ): 264 - 279 . https://doi.org/10.3322/caac.21657 https://doi.org/10.3322/caac.21657 Current treatment and recent progress in gastric cancer Lin JX , Xie XS , Weng XF , et al. , 2019 . UFM1 suppresses invasive activities of gastric cancer cells by attenuating the expression of PDK1 through PI3K/AKT signaling . J Exp Clin Cancer Res , 38 : 410 . https://doi.org/10.1186/s13046-019-1416-4 https://doi.org/10.1186/s13046-019-1416-4 UFM1 suppresses invasive activities of gastric cancer cells by attenuating the expression of PDK1 through PI3K/AKT signaling Missiroli S , Perrone M , Genovese I , et al. , 2020 . Cancer metabolism and mitochondria: finding novel mechanisms to fight tumours . eBioMedicine , 59 : 102943 . https://doi.org/10.1016/j.ebiom.2020.102943 https://doi.org/10.1016/j.ebiom.2020.102943 Cancer metabolism and mitochondria: finding novel mechanisms to fight tumours Nerreter T , Letschert S , Götz R , et al. , 2019 . Super-resolution microscopy reveals ultra-low CD19 expression on myeloma cells that triggers elimination by CD19 CAR-T . Nat Commun , 10 : 3137 . https://doi.org/10.1038/s41467-019-10948-w https://doi.org/10.1038/s41467-019-10948-w Super-resolution microscopy reveals ultra-low CD19 expression on myeloma cells that triggers elimination by CD19 CAR-T Pai S , Yadav VK , Kuo KT , et al. , 2021 . PDK1 inhibitor BX795 improves cisplatin and radio-efficacy in oral squamous cell carcinoma by downregulating the PDK1/CD47/Akt-mediated glycolysis signaling pathway . Int J Mol Sci , 22 ( 21 ): 11492 . https://doi.org/10.3390/ijms222111492 https://doi.org/10.3390/ijms222111492 PDK1 inhibitor BX795 improves cisplatin and radio-efficacy in oral squamous cell carcinoma by downregulating the PDK1/CD47/Akt-mediated glycolysis signaling pathway Perillo B , di Donato M , Pezone A , et al. , 2020 . ROS in cancer therapy: the bright side of the moon . Exp Mol Med , 52 ( 2 ): 192 - 203 . https://doi.org/10.1038/s12276-020-0384-2 https://doi.org/10.1038/s12276-020-0384-2 ROS in cancer therapy: the bright side of the moon Sanz MA , Fenaux P , Tallman MS , et al. , 2019 . Management of acute promyelocytic leukemia: updated recommendations from an expert panel of the European LeukemiaNet . Blood , 133 ( 15 ): 1630 - 1643 . https://doi.org/10.1182/blood-2019-01-894980 https://doi.org/10.1182/blood-2019-01-894980 Management of acute promyelocytic leukemia: updated recommendations from an expert panel of the European LeukemiaNet Škorja Milić N , Dolinar K , Miš K , et al. , 2021 . Suppression of pyruvate dehydrogenase kinase by dichloroacetate in cancer and skeletal muscle cells is isoform specific and partially independent of HIF-‍1α . Int J Mol Sci , 22 ( 16 ): 8610 . https://doi.org/10.3390/ijms22168610 https://doi.org/10.3390/ijms22168610 Suppression of pyruvate dehydrogenase kinase by dichloroacetate in cancer and skeletal muscle cells is isoform specific and partially independent of HIF-‍1α Smyth EC , Nilsson M , Grabsch HI , et al. , 2020 . Gastric cancer . Lancet , 396 ( 10251 ): 635 - 648 . https://doi.org/10.1016/s0140-6736(20)31288-5 https://doi.org/10.1016/s0140-6736(20)31288-5 Gastric cancer Sradhanjali S , Tripathy D , Rath S , et al. , 2017 . Overexpression of pyruvate dehydrogenase kinase 1 in retinoblastoma: a potential therapeutic opportunity for targeting vitreous seeds and hypoxic regions . PLoS ONE , 12 ( 5 ): e0177744 . https://doi.org/10.1371/journal.pone.0177744 https://doi.org/10.1371/journal.pone.0177744 Overexpression of pyruvate dehydrogenase kinase 1 in retinoblastoma: a potential therapeutic opportunity for targeting vitreous seeds and hypoxic regions Schneider CA , Rasband WS , Eliceiri KW , 2012 . NIH Image to ImageJ: 25 years of image analysis . Nat Methods , 9 ( 7 ): 671 - 675 . https://doi.org/10.1038/nmeth.2089 https://doi.org/10.1038/nmeth.2089 NIH Image to ImageJ: 25 years of image analysis Shang MT , Zhou ZW , Kuang WB , et al. , 2021 . High-precision 3D drift correction with differential phase contrast images . Opt Express , 29 ( 21 ): 34641 - 34655 . https://doi.org/10.1364/OE.438160 https://doi.org/10.1364/OE.438160 High-precision 3D drift correction with differential phase contrast images Tataranni T , Piccoli C , 2019 . Dichloroacetate (DCA) and cancer: an overview towards clinical applications . Oxid Med Cell Longev , 2019 : 8201079 . https://doi.org/10.1155/2019/8201079 https://doi.org/10.1155/2019/8201079 Dichloroacetate (DCA) and cancer: an overview towards clinical applications Tewari D , Patni P , Bishayee A , et al. , 2022 . Natural products targeting the PI3K-Akt-mTOR signaling pathway in cancer: a novel therapeutic strategy . Semin Cancer Biol , 80 : 1 - 17 . https://doi.org/10.1016/j.semcancer.2019.12.008 https://doi.org/10.1016/j.semcancer.2019.12.008 Natural products targeting the PI3K-Akt-mTOR signaling pathway in cancer: a novel therapeutic strategy Velpula KK , Bhasin A , Asuthkar S , et al. , 2013 . Combined targeting of PDK1 and EGFR triggers regression of glioblastoma by reversing the Warburg effect . Cancer Res , 73 ( 24 ): 7277 - 7289 . https://doi.org/10.1158/0008-5472.can-13-1868 https://doi.org/10.1158/0008-5472.can-13-1868 Combined targeting of PDK1 and EGFR triggers regression of glioblastoma by reversing the Warburg effect Wadgaonkar P , Chen F , 2021 . Connections between endoplasmic reticulum stress-associated unfolded protein response, mitochondria, and autophagy in arsenic-induced carcinogenesis . Semin Cancer Biol , 76 : 258 - 266 . https://doi.org/10.1016/j.semcancer.2021.04.004 https://doi.org/10.1016/j.semcancer.2021.04.004 Connections between endoplasmic reticulum stress-associated unfolded protein response, mitochondria, and autophagy in arsenic-induced carcinogenesis Wang P , Jin JM , Liang XH , et al. , 2022 . Helichrysetin inhibits gastric cancer growth by targeting c-Myc/PDHK1 axis-mediated energy metabolism reprogramming . Acta Pharmacol Sin , 43 ( 6 ): 1581 - 1593 . https://doi.org/10.1038/s41401-021-00750-0 https://doi.org/10.1038/s41401-021-00750-0 Helichrysetin inhibits gastric cancer growth by targeting c-Myc/PDHK1 axis-mediated energy metabolism reprogramming Wang QQ , Jiang Y , Naranmandura H , 2020 . Therapeutic strategy of arsenic trioxide in the fight against cancers and other diseases . Metallomics , 12 ( 3 ): 326 - 336 . https://doi.org/10.1039/c9mt00308h https://doi.org/10.1039/c9mt00308h Therapeutic strategy of arsenic trioxide in the fight against cancers and other diseases Wang WP , Dong XX , Liu Y , et al. , 2020 . Itraconazole exerts anti-liver cancer potential through the Wnt, PI3K/AKT/mTOR, and ROS pathways . Biomed Pharmacother , 131 : 110661 . https://doi.org/10.1016/j.biopha.2020.110661 https://doi.org/10.1016/j.biopha.2020.110661 Itraconazole exerts anti-liver cancer potential through the Wnt, PI3K/AKT/mTOR, and ROS pathways Weng MS , Chang JH , Hung WY , et al. , 2018 . The interplay of reactive oxygen species and the epidermal growth factor receptor in tumor progression and drug resistance . J Exp Clin Cancer Res , 37 : 61 . https://doi.org/10.1186/s13046-018-0728-0 https://doi.org/10.1186/s13046-018-0728-0 The interplay of reactive oxygen species and the epidermal growth factor receptor in tumor progression and drug resistance Wu J , Henderson C , Feun L , et al. , 2010 . Phase II study of darinaparsin in patients with advanced hepatocellular carcinoma . Invest New Drugs , 28 ( 5 ): 670 - 676 . https://doi.org/10.1007/s10637-009-9286-9 https://doi.org/10.1007/s10637-009-9286-9 Phase II study of darinaparsin in patients with advanced hepatocellular carcinoma Xu XH , Wang HB , Li HY , et al. , 2019 . S -Dimethylarsino-glutathione (darinaparsin ® ) targets histone H3 . 3 , leading to TRAIL-induced apoptosis in leukemia cells. Chem Commun , 55 ( 87 ): 13120 - 13123 . https://doi.org/10.1039/c9cc07605k https://doi.org/10.1039/c9cc07605k S-Dimethylarsino-glutathione (darinaparsin®) targets histone H3 Study design: Yi LI and Wenyan SHE; Data collection: Yi LI, Wenyan SHE, Xiaoran XU, Yixin LIU, Sheng TIAN, and Shiyi LI; Data analysis: Yi LI, Yixin LIU, and Xinyu WANG; Data interpretation: Yi LI, Wenyan SHE, Xinyu WANG, and Miao WANG; Writing of manuscript: Yi LI; Critical revision: Yi LI, Chaochao YU, and Pan LIU; Final approval: Yi LI, Yongchang WEI, and Tianhe HUANG; Overall supervision: Tianhe HUANG and Yongchang WEI. All authors have read and approved the final manuscript, and therefore, have full access to all the data in the study and take responsibility for the integrity and security of the data. This study was supported by the Wuhan University Zhongnan Hospital Translational Medicine and Interdisciplinary Research Joint Fund (No. ZNJC201910), China. 10.1631/jzus.B2200351 1673-1581（2023）03-0232-16 李怡,佘文妍,徐笑然等.AAZ2可通过靶向PDK1介导线粒体依赖的凋亡[J].浙江大学学报（英文版）（B辑：生物医学和生物技术）,2023,24(03):232-247. 李怡,佘文妍,徐笑然等.AAZ2可通过靶向PDK1介导线粒体依赖的凋亡[J].浙江大学学报（英文版）（B辑：生物医学和生物技术）,2023,24(03):232-247. This study was supported by the Wuhan University Zhongnan Hospital Translational Medicine and Interdisciplinary Research Joint Fund (No. ZNJC201910), China. 2022-07-01 2022-12-04 2023-03-15 2023-03-14 2023-02-09 Copyright © Zhejiang University and Springer-Verlag GmbH Germany, part of Springer Nature 2022 2023 浙江大学出版社 Yi LI, Wenyan SHE, Xiaoran XU, Yixin LIU, Xinyu WANG, Sheng TIAN, Shiyi LI, Miao WANG, Chaochao YU, Pan LIU, Tianhe HUANG, and Yongchang WEI declare that they have no conflict of interest. The study was approved by the Medical Ethics Committee Zhongnan Hospital of Wuhan University (No. ZN2021006) and performed in accordance with the Declaration of Helsinki. Compliance with ethics guidelines Supplementary information 1 Table S1; Figs. S1 and S2 浙江大学学报（英文版）（B辑：生物医学和生物技术） 3 24