Anion-Responsive Manganese Porphyrin Facilitates Chloride Transport and Induces Immunogenic Cell Death

Open AccessCCS ChemistryRESEARCH ARTICLE14 Jul 2022Anion-Responsive Manganese Porphyrin Facilitates Chloride Transport and Induces Immunogenic Cell Death Fang-Xin Wang†, Jie-Wei Liu†, Xiao-Qiao Hong†, Cai-Ping Tan†, Li Zhang, Wen-Hua Chen, Peter J. Sadler and Zong-Wan Mao Fang-Xin Wang† MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Department of Chemistry, Sun Yat-Sen University, Guangzhou 510275 Department of Chemistry, University of Warwick, Coventry CV4 7AL †F.-X. Wang, J.-W. Liu, X.-Q. Hong, and C.-P. Tan contributed equally to this work.Google Scholar More articles by this author , Jie-Wei Liu† MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Department of Chemistry, Sun Yat-Sen University, Guangzhou 510275 †F.-X. Wang, J.-W. Liu, X.-Q. Hong, and C.-P. Tan contributed equally to this work.Google Scholar More articles by this author , Xiao-Qiao Hong† School of Biotechnology and Health Sciences, Wuyi University, Jiangmen 529020 †F.-X. Wang, J.-W. Liu, X.-Q. Hong, and C.-P. Tan contributed equally to this work.Google Scholar More articles by this author , Cai-Ping Tan† MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Department of Chemistry, Sun Yat-Sen University, Guangzhou 510275 †F.-X. Wang, J.-W. Liu, X.-Q. Hong, and C.-P. Tan contributed equally to this work.Google Scholar More articles by this author , Li Zhang *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Department of Chemistry, Sun Yat-Sen University, Guangzhou 510275 Google Scholar More articles by this author , Wen-Hua Chen *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] School of Biotechnology and Health Sciences, Wuyi University, Jiangmen 529020 Google Scholar More articles by this author , Peter J. Sadler *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] Department of Chemistry, University of Warwick, Coventry CV4 7AL Google Scholar More articles by this author and Zong-Wan Mao *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] E-mail Address: [email protected] MOE Key Laboratory of Bioinorganic and Synthetic Chemistry, Department of Chemistry, Sun Yat-Sen University, Guangzhou 510275 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101212 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Chloride is the most abundant anion in living systems. Most natural or synthetic chloride anionophores function via hydrogen-bonding interactions. However, dynamic metal-anion coordination can also be an efficient way of transporting chloride across membranes. Here, we investigate anion transport by manganese(III) meso-tetraphenylporphyrin chloride {[Mn(TPP)Cl], TPP = meso-tetraphenylporphyrin} complex that exhibits labile axial coordination. [Mn(TPP)Cl] showed high chloride transport activity in a bilayer vesicle model with an EC50 value of 4.42 × 10−3 mol %. In living cells, [Mn(TPP)Cl] induced rapid chloride influx and autophagy. The release of Ca2+ and adenosine 5′-triphosphate (ATP), as well as the relocation of calreticulin, revealed that [Mn(TPP)Cl] caused immunogenic cell death. Proteomic analysis indicated that [Mn(TPP)Cl] impaired several physiological processes, including DNA synthesis, transcription, mitochondrial respiration, RNA translation, and immune response. Our study suggests that dynamic metal-anion interactions across membranes might provide a practical strategy for the interference of chloride homeostasis. Download figure Download PowerPoint Introduction Chloride exists in abundance, as high as ca. 120 mM in extracellular fluids and 4–60 mM in the cytosol. Chloride plays an important role in controlling membrane potentials, intracellular pH, and many physiological processes.1 Its transport across the membrane is regulated in an orderly manner by multiple chloride channels.2 One way to manipulate chloride transport is to incorporate ion channels in phospholipid bilayer membranes. Chloride channel genes have been cloned and expressed successfully in biological membranes using classic molecular biological approaches.3 Short peptides containing the chloride channel domain, for example, the α-subunit glycine receptor, M2GlyR, have also been reported to exhibit good chloride transport activity.4,5 Barboiu et al.6 reported self-assembled columnar triazole quartets as artificial channels in lipid vesicles. However, whether successful self-assembly and chloride transport by these channels could be achieved under physiological conditions remains to be determined. The use of an anionophore (anion-selective ionophore) is another way to achieve chloride transport with high efficiency and selectivity. Prodigiosin is a natural secondary metabolite obtained from Serratia marcescens that facilitates H+/Cl− cotransport, and it is the most vigorous chloride anionophore reported thus far.7 Gale et al.8 reported that squaramide-based anionophores facilitated chloride transport into cells and caused apoptosis. Alfonso et al.9 designed pH-responsive pseudopeptide cages that encapsulate chloride in cavities. Sheppard et al.10 found bis-(p-nitrophenyl)ureidodecalin to be a potent transporter of chloride into living cells. Yang and co-workers11 mimicked natural ion channels with isophthalamide derivatives and recorded single-channel ion transport by patch clamping. In addition to organic molecules, metal complexes based on platinum and palladium have been found to exhibit anion exchange behavior.12,13 We have reported a cyclometalated iridium complex (Ir-biimidazole) that transports chloride via the formation of H-bonds to chloride and alkalinizes lysosomes to block autophagic flux.14 In general, most of these reported anionophores interact with chloride via H–Cl interactions, and several anionophores transport chloride via metal-anion interactions.15,16 Manganese porphyrins have been widely investigated in many fields, including chemical catalysis, materials science, printed electronics, and medical treatments. Notably, it has been reported that manganese (III) meso-tetraphenylporphyrin chloride ([Mn(TPP)Cl], TPP = meso-tetraphenylporphyrin, Figure 1) increased anion permeability in lung epithelial cells.17 The lipophilic macrocyclic structure and labile axial coordination satisfy the design principles for ionophores. We have shown that [Mn(TPP)Cl] facilitated chloride transport through bilayer membranes in vesicle models and living cells. It also induced autophagy and immunogenic cell death. Proteomics analysis showed that many physiological pathways were affected by [Mn(TPP)Cl], including DNA synthesis, RNA translation, mitochondrial respiration, and immune response. Collectively, our study demonstrates the potential of dynamic metal-anion interactions for achieving chloride transport and influencing physiological processes for therapeutic applications. Figure 1 | X-ray crystal structure of [Mn(TPP)Cl]. The non-hydrogen atoms are drawn in thermal ellipsoids at the 50% probability level. Download figure Download PowerPoint Experimental Methods Materials MnCl2·4H2O, TPP, N,N-dimethylformamide (DMF), dichloromethane, and ethyl acetate were obtained from Energy Chemical (Shanghai, China). N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES), NaCl, NaHCO3, Na2SO4, NaNO3, KCl, RbCl, LiCl, CsCl, and citrate were purchased from Aladdin (Shanghai, China). Methanol, acetonitrile, chloroform, cisplatin, acridine orange, Cell Counting Kit-8 (CCK-8) assay kit, egg yolk L-α-phosphatidylcholine (EYPC), Sephadex G-25, poly(vinylidene difluoride) (PVDF) membrane filter, Na-gluconate, K-gluconate, Ca-gluconate, and d-gluconate were purchased from Merck (Darmstadt, Germany). Fetal bovine serum (FBS), dialyzed FBS, l-lysine-2HCl (13C6, 15N2), l-lysine-2HCl, l-arginine-HCl (13C6, 15N4), l-arginine-HCl, penicillin/streptomycin, sulforhodamine B (SRB), and cell apoptosis detection Annexin V/propidium iodide (PI) kit were purchased from Thermo Fisher (MA, USA). N-(Ethoxycarbonylmethyl)-6-methoxyquinolinium bromide (MQAE), 3-methyladenine (3-MA), z-VAD-fmk, necrostatin-1 (Nec-1), and ferrostatin-1 (Fer-1) were purchased from MedChemExpress (NJ, USA). Anti-p62 antibody (ab91526), anti-LC3B antibody (ab192890), anti-β actin antibody (ab8226), and Alexa Fluor® 647 anticalreticulin antibody (ab196159) were purchased from Abcam (Cambridge, UK). An adenosine 5′-triphosphate (ATP) detection assay kit, Fluo-4, and the materials (solutions, buffers, and markers) used in western blotting and immunofluorescence assays were purchased from Beyotime (Shanghai, China). Equipment Proton nuclear magnetic resonance (1H NMR) spectra were recorded on a Bruker Avance 500 NMR spectrometer (Bruker, MA, USA). High-resolution mass spectra were recorded on a Thermo LTQ Orbitrap XL mass spectrometer (Thermo Fisher, MA, USA). Liquid chromatography mass spectrometry (LC-MS) and high-performance liquid chromatography (HPLC) spectra were recorded on a Q Exactive Hybrid Ultimate 3000 mass spectrometer (Thermo Fisher, MA, USA). Elemental analysis was carried out using an Elemental Vario EL CHNS analyzer (Elementar, Hanau, Germany). Single-crystal X-ray diffraction measurements were performed on a Bruker Smart 1000 CCD diffractometer (Bruker, MA, USA). UV–vis spectra were recorded on a Varian Cary 300 spectrophotometer (Varian, NV, USA). EYPC vesicles were prepared by extrusion through 100 nm Whatman nuclepore track-etched polycarbonate membranes using an Avanti Mini-Extruder (Avanti, AL, USA). Chloride efflux was measured using a Mettler–Toledo PerfectIon™ chloride ion-selective electrode assembled with a Mettler–Toledo Seven Compact S220 ionometer (Mettler Toledo, OH, USA). The conductivity was measured with a Jenway 4510 conductivity meter (Jenway, Staffordshire, UK). Cells were monitored by a Zeiss LSM 880 confocal laser scanning fluorescence microscope (Zeiss, Oberkochen, Germany). Absorption and luminescence were measured by an Infinite M200pro microplate reader (Tecan, Männedorf, Switzerland). Flow cytometry was conducted on a BD FACSCalibur (BD Biosciences, NJ, USA). Chloride transport in EYPC model Precisely, 1.18 mL of a 25 mM HEPES buffer (500 mM NaNO3, pH 7.0) was added to a vesicle dispersion (0.3 mL), followed by the addition of varying concentrations of [Mn(TPP)Cl] in dimethyl sulfoxide (DMSO) (20 μL). Then the chloride efflux was monitored as a function of time using a chloride-selective electrode. After 5 min, 5 wt % aqueous Triton X-100 (50 μL) was added. The relative chloride efflux (%) was calculated using the following equation: ([Cl−] – [Cl−]0)/([Cl−]total – [Cl−]0) × 100, where [Cl−]0, [Cl−], where [Cl−]total represent the concentrations of chloride anions at the initial time, after a period, and after the addition of 5 wt % aqueous Triton X-100 (50 μL). The lipid concentration was 1.32 mM. Chloride transport in a U-tube assay [Mn(TPP)Cl] (1 mM) and TBAPF6 (2 mM) were dissolved in chloroform (5 mL) with TBAPF6 as the supporting electrolyte. The measuring conditions for the chloride ion receiving phase (3 mL) entailed using 500 mM NaNO3 and 25 mM HEPES (pH 7.0), with 500 mM NaCl and 25 mM HEPES were in the chloride-donating aqueous phase (3 mL). The chloride ion concentration of the chloride-receiving aqueous phase was recorded every 12 h. Cell line and cell culture conditions Human cervical carcinoma HeLa cells were obtained from the Experimental Animal Centre of Sun Yat-Sen University (Guangzhou, China). RFP-LC3 transfected HeLa cells, as autophagy reporter cells (RFP-LC3 = autophagy tandem lentiviral biosensor), were obtained from Guangzhou Institute of Biomedicine and Health, Chinese Academy of Sciences (Guangzhou, China). HeLa cells were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS, 100 μg/mL streptomycin, and 100 U/mL penicillin G. Under normoxia, cells were grown at 37 °C in a 5% CO2 humidified atmosphere. For hypoxic conditions, cells were incubated at 37 °C in a 1% O2/5% CO2 humidified atmosphere. All detailed experimental methods are available in the Supporting Information. Results and Discussion Synthesis and characterization of [Mn(TPP)Cl] [Mn(TPP)Cl] was synthesized according to a reported procedure.18 MnCl2·4H2O (10 mmol) and H2TPP (1 mol) were refluxed in 100 mL DMF for 6 h, during which Mn2+ was air-oxidized to Mn3+. The crude product was purified by silica column chromatography with CH2Cl2 and ethyl acetate as the eluent. [Mn(TPP)Cl] was characterized by electrospray ionization mass spectrometry (ESI-MS), 1H NMR, HPLC, LC-MS, and elemental analysis ( Supporting Information Figures S1–S4). X-ray diffraction studies of a crystal formed in CHCl3 revealed that [Mn(TPP)Cl] crystallized in a tetragonal space group I4/m ( Supporting Information Table S1). The asymmetric unit consisted of one flat molecule, in which Mn3+ is coordinated to four N atoms of deprotonated TPP and an axial Cl, forming a distorted square-pyramidal geometry. The porphyrin lies on an asymmetric mirror plane, which caused a twofold symmetric disorder of the central Mn and Cl (Mn–Cl 2.349(18) Å, Mn–N 2.007(2) Å, Mn–Por 0.113(18) Å, where Por is the plane center of the porphyrin complex). The bond lengths of Mn–Cl and Mn–N reported in the literature fell between 2.30–2.66 Å and 2.002–2.019 Å, respectively, similar to the bond lengths observed here ( Supporting Information Table S2). Axial ligand exchange of [Mn(TPP)Cl] Anion exchange in the axial position is vital for ion response. First, the conductivity of [Mn(TPP)Cl] was determined in different solvents. In polar-protic methanol (dipole moment 1.69 D), the molar conductivity, Λ, of [Mn(TPP)Cl] was 65 S·cm2·mol−1, which indicated dissociation of the chloride anion. In contrast, Λ values for the polar-aprotic acetonitrile (3.92 D) and nonpolar-aprotic chloroform (1.04 D) were 2.2 and 0.1 S·cm2·mol−1, respectively, which indicated the presence of an intact Mn–Cl bond in these aprotic solvents. The intactness of the Mn–Cl bond was also investigated by UV–vis absorption spectroscopy. According to the Gouterman four-orbital model, [Mn(TPP)Cl] showed characteristic absorption peaks, an intense Soret band (S0 → S2), and two Q bands (α and β, S0→S1).19–21 As shown in Supporting Information Table S3, the wavelengths of the Soret band were 466 nm in methanol and 467 nm in water (containing 1% methanol), which indicated that the chloride in the axial position tended to dissociate from the Mn(III) center. However, the Soret band redshifted to 476 and 478 nm in CH3CN and CHCl3, respectively. Also, Q bands redshifted in these solvents with ε(β band)/ε(α band) > 1, confirming that the chloride anion was coordinated to the Mn(III) center in aprotic solvents. Exchange reactions of the axial ligands on the manganese porphyrin were studied by UV–vis spectroscopy. [Mn(TPP)Cl] was dissolved in chloroform to a 1 mM concentration and then mixed with an equal volume of an aqueous sodium salt solution (100 mM). The samples were stirred vigorously for 1 day, and then the organic phase was separated for analysis. The Mn species [Mn(TPP)X] (X = axial anion) showed different absorption bands in the near-UV region and visible region (Figure 2a). The Soret band of [Mn(TPP)Cl] is blueshifted from 478 to 471 nm after anion exchange with bicarbonate. Bicarbonate, as a weak base, generated an excess of hydroxide that could also bind in the axial position. The nitrate-containing species coordinated to the metal center, showing stronger absorbance in the near-UV and Q band regions, and the Soret band shifted to 477 nm, while the Soret band shifted to 476 nm when the Mn(III) center was coordinated by sulfate anions. Figure 2 | (a) UV–vis absorption spectra of [Mn(TPP)X] in chloroform. The absorbances are normalized to the Soret band (set to 1.0). The inset shows the shift of the Soret band. (b) 1H NMR spectra of β-pyrrole protons in Mn porphyrin species in CDCl3. Download figure Download PowerPoint NMR was also used to investigate the chemical environment of the β-pyrrole protons, as they are sensitive to the axial ligand effects on π interactions between Mn(III) and the porphyrin. The 1H NMR peak was relatively broad due to the presence of the paramagnetic high-spin d4 Mn3+. [Mn(TPP)Cl] when dissolved in methanol (1 mM, 5 mL) and mixed with solutions containing excess sodium salts (100 mM, 45 mL). After stirring for 1 day, the precipitate was filtered and dissolved in CDCl3 at ca. 2 mM. As Mn–X bonds were expected to be intact in chloroform, the β-pyrrole proton chemical shift depended on the axial coordination (Figure 2b). The chemical shift of the β-pyrrole protons in [Mn(TPP)Cl] was −22.7 ppm in CDCl3. For the species formed in a solution with excess NaNO3, the signal shifted to −29.5 ppm, which indicated a stronger π interaction between Mn(III) and porphyrin. With the species formed in excess Na2SO4 solution, the β-pyrrole 1H signal shifted downfield (δ = −20.6 ppm). In the bicarbonate solution, two peaks were detected, probably due to the coordination of the hydroxide (δ = −23.0 ppm) and bicarbonate (δ = −16.0 ppm) anions. The different β-pyrrole proton signals implied the interactions of different anions with the Mn(III) center. Chloride transport activity in the lipid vesicle model and U-tube assay Chloride transport was evaluated in the lipophilic EYPC vesicle model to mimic bilayer lipid membranes (Figure 3a). Real-time chloride efflux into extravesicular buffer was monitored by a chloride-selective electrode. [Mn(TPP)Cl] exhibited a high chloride transport activity in a concentration-dependent manner (Figure 3b). Chloride efflux mediated by [Mn(TPP)Cl] reached maximum at 300 s at a molar ratio of 15.6 mmol % ([Mn(TPP)Cl]:lipid = 1:6400), with all chloride inside the EYPC vesicles transported to the extravesicular buffer. Hill analysis showed that the transport efficacy EC50 of [Mn(TPP)Cl] was 4.42 × 10−3 mol % ( Supporting Information Figure S5), which was eightfold higher than that for the Ir-biimidazole complex previously reported by us.14 Besides, [Mn(TPP)Cl] was more potent than the squaramide-based compound SA-3 (EC50 = 2.4 × 10−2 mol %) and the pyridine diamide-strapped calix[4]pyrrole DSC4P-1 (EC50 = 3.52 mol %) that carried chloride through hydrophobic liposome membranes via the formation of H-bonds to chloride, and their transport activities correlated with the number of H-bonds.22,23 Figure 3 | (a) EYPC model for the determination of chloride transport. Intravesicular buffer solution: 25 mM HEPES, 500 mM NaCl, pH 7.0. Extravesicular buffer solution: 25 mM HEPES, 500 mM sodium salt (NaNO3, NaHCO3, or Na2SO4), pH 7.0. (b) Chloride efflux facilitated by [Mn(TPP)Cl] at different molar ratios in the EYPC model. (c) Chloride efflux facilitated by [Mn(TPP)Cl] (7.8 mmol %) with different anions in the extravesicular buffer at pH 7.0. (d) Chloride efflux facilitated by [Mn(TPP)Cl] (0.0078 mol %) in the presence of various chloride salts (500 mM) inside the EYPC vesicles. (e) U-tube apparatus for evaluation of ion carrier activity. (f) Chloride transport by [Mn(TPP)Cl] and DMSO (blank) in the U-tube assay. Download figure Download PowerPoint In a nitrate-containing buffer, the relative rate of chloride efflux induced by [Mn(TPP)Cl] was significantly higher by 22% at 300 s, compared with the bicarbonate-containing buffer, whereas the rate in the sulfate-containing buffer was the lowest (Figure 3c). This indicated that chloride transport could occur via Cl−/NO3− or Cl−/HCO3− antiport and not via Cl−/SO42− antiport. Although sulfate and manganese center coordination stabilized the structure, sulfate was not feasible for dynamic anion exchange. Alkali metal cations (Li+, Na+, K+, Rb+, and Cs+) showed little effect on this chloride transport (Figure 3d), suggesting that monovalent cations might not be involved in Mn(III)-anion interactions. The chloride efflux measurements at varying pH values (pH 5.0/6.0/7.0, adjusted by citrate) were not significantly different ( Supporting Information Figure S6). A U-tube apparatus with an aqueous-organic-aqueous phase system that mimics biofluids and hydrophobic lipid membranes (Figure 3e) was used to determine the activity of [Mn(TPP)Cl] as an ion carrier. A total of 57 mmol chloride ion was transported by [Mn(TPP)Cl] through chloroform (5 mL) from the chloride-resource aqueous phase (3 mL) to the chloride-receiving aqueous phase (3 mL) in 48 h at 20 °C (Figure 3f). This chloride transport rate was much faster than that reported for the squaramide-based anionophores.24 [Mn(TPP)Cl] facilitates chloride transport and disrupts chloride homeostasis Subsequently, we determined whether [Mn(TPP)Cl] could facilitate chloride transport into living cells. MQAE is a commercial chloride probe with high selectivity that shows a negative linear correlation between fluorescence emission intensity and chloride concentration.8 With human cervical carcinoma HeLa cells, a dose-dependent decrease in MQAE emission intensity was detected upon incubation with [Mn(TPP)Cl] for 10 min (Figure 4a), confirming that [Mn(TPP)Cl] acted as a chloride anionophore and disrupted the intracellular chloride homeostasis. Figure 4 | (a) Relative emission intensity of MQAE in HeLa cells. HeLa cells were preloaded with MQAE at 5 mM for 30 min, and then incubated with [Mn(TPP)Cl] at 0–40 μM for 10 min (n = 4). (b) Cell viability after treatment with [Mn(TPP)Cl] at 0–20 μM for 24 h (n = 4). HeLa cells were cultured in HBSS and chloride-free HBSS with FBS (10%, v/v) at pH 7.4. Download figure Download PowerPoint As a consequence, [Mn(TPP)Cl] showed potent cytotoxicity (IC50 = 1.7 ± 0.1 μM, 24 h; IC50 = 1.0 ± 0.1 μM, 48 h) toward the HeLa cells, determined by CCK-8 assay. Hypoxia is a common condition detected in several cancerous regions due to rapid oxygen consumption and limited oxygen diffusion.25 Interestingly, [Mn(TPP)Cl] retained good anticancer capability under hypoxia (IC50 = 2.1 ± 0.1 μM, 24 h, 1% pO2), whereas the classical metallodrug cisplatin was less effective (IC50 = 63.1 ± 1.7 μM) under hypoxia. We measured the effect of chloride transport on cell viability in Hank’s balanced salt solution (HBSS) and chloride-free HBSS (chloride replaced by gluconate) using the SRB assay. We found that the cytotoxicity of [Mn(TPP)Cl] was higher in HBSS than that measured in chloride-free HBSS in a dose-dependent manner (Figure 4b), which implied that chloride homeostasis played an important role in cell death. [Mn(TPP)Cl] causes cytoprotective autophagy and immunogenic cell death We employed transmission electron microscopy (TEM) to observe the ultrastructural alterations occurring in HeLa cells after treatment with [Mn(TPP)Cl] (Figure 5a). We observed the formation of extensive large vacuoles in the cytoplasm after treatment with 5 μM [Mn(TPP)Cl] for 6 h. Additionally, the endoplasmic reticulum (ER) expanded widely with mitochondrial swelling and the disappearance of cristae. Moreover, the percentages of annexin V- and PI-positive cells increased upon treatment with [Mn(TPP)Cl] (5 μM) for 12 h (Figure 5b), indicating an increase in necrotic/late apoptotic cells. Distinct from the apoptotic cells, the [Mn(TPP)Cl] treated swelled to a larger size, reflected by forward-scattered light (FSC) that passed around cells (Figure 5c). The increasing signal in the side-scattered light (SSC) showed that more light was reflected by intracellular particulates, suggesting the complexity of cellular changes upon [Mn(TPP)Cl] treatment (Figure 5c). Figure 5 | (a) Representative TEM images of HeLa cells treated with [Mn(TPP)Cl] (5 μM, 6 h). The enlarged images show the morphology of mitochondria (I), ER (II) and vacuoles (III). (b) Cell death detected by annexin V-fluorescein isothiocyanate/propidium iodide (annexin V-FITC/PI) double staining. HeLa cells were treated with [Mn(TPP)Cl] (5 μM) or DMSO for 12 h. (c) FSC and SSC signals of HeLa cells after treatment with [Mn(TPP)Cl] (5 μM) or DMSO for 12 h. (d) LC3 expression in the RFP-LC3 transfected HeLa cells. Cells were treated with [Mn(TPP)Cl] (10 μM) for 2 h. Scale bar: 20 μm. (e) Expression of autophagy-related proteins LC3-I/II and p62 in HeLa cells. β-Actin was used as the loading control protein. (f) Effect on cell viability after treatment with cell death inhibitors. Cells were pre-treated with inhibitors (2 mM 3-MA, 20 μM z-VAD-fmk, 20 μM Nec-1, or 200 nM Fer-1) for 2 h, and then incubated with [Mn(TPP)Cl] (0–10 μM) for 24 h (n = 5). (g) Extracellular ATP level in DMEM culture medium. HeLa cells were treated with [Mn(TPP)Cl] at 0–10 μM for 6 h. (h) Ca2+ levels in HeLa cells, detected by flow cytometry. Cells were treated with [Mn(TPP)Cl] at 0–10 μM for 6 h and then stained with the Ca2+ probe Fluo-4. (i) The translocation of calreticulin in HeLa cells. Cells were treated with DMSO or [Mn(TPP)Cl] (5 μM) for 12 h and then stained with anticalreticulin antibody (red) and Hoechst 33342 (blue). Scale bar: 20 μm. FSC, forward scatter; SSC, side scatter. Download figure Download PowerPoint In some vacuoles, features like double-membrane structures with incompletely degraded substrates were observed, typical of autophagy. Acridine orange accumulated in the protonated form trapped in acidic vesicular organelles and emitted red fluorescence after irradiation.26 The emission intensity of acridine orange was dose-dependent in HeLa cells treated with [Mn(TPP)Cl] ( Supporting Information Figure S7), indicating an increase in the formation of acidic structures. Microtubule-associated protein 1 light chain 3 (LC3) is the hallmark protein of autophagosome formation, and the conversion of LC3-I to LC3-II indicated autophagy.27 Large quantities of red fluorescent dots representing LC3-II appeared in HeLa cells transfected with RFP-LC3 plasmid after treatment with [Mn(TPP)Cl] (10 μM, 2 h) (Figure 5d). Western blot analysis also confirmed the conversion of LC3-I to LC3-II and increased expression of p62, an LC3-binding protein (Figure 5e).28 Moreover, the viability of HeLa cells declined upon pretreatment with the autophagy inhibitor, 3-MA (Figure 5f), which implied that autophagy is a protective response to the cellular stress initiated by [Mn(TPP)Cl]. Treatment of cells with Nec-1 (a RIP1 kinase inhibitor for necroptosis),29 z-VAD-fmk (a pan-caspase inhibitor),30 or Fer-1 (a ferroptosis inhibitor)31 failed to prevent cell death. Some anionophores (e.g., prodigiosin, Ir-biimidazole complex, and squaramide SA-3) have also been reported to induce autophagy in cells.8,14 We assume that chloride homeostasis plays an essential role in lysosome function, perturbing pH control, enzyme activity, and signaling cascades. Interestingly, we noticed that [Mn(TPP)Cl] induced cellular responses involved with immunogenic cell death. ATP is expelled to the extracellular biofluid, where it functions as a powerful short-range “find me” signal in immunogenic cell death.32 After [Mn(TPP)Cl] treatment, the ATP level in the DMEM culture medium increased in a dose-dependent manner (Figure 5g). ER is involved in various cellular signaling processes, and it plays a crucial role as a dynamic Ca2+ store.33 The intracellular Ca2+ level elevated by ca. fourfold in [Mn(TPP)Cl]-treated HeLa cells (Figure 5h). Moreover, calreticulin was exposed to the cell membrane after treatment of [Mn(TPP)Cl] (Figure 5i), which has been identified as a key feature in immunogenic cell death.34 Altogether, these results showed that [Mn(TPP)Cl] could induce an immune response. SILAC-based proteomics and protein–protein interaction network Further, we elucidated the mechanism of cell death induced by [Mn(TPP)Cl], stable isotope labeling of amino