Cell Chemical Biology

Transmission electron microscopy: PANC-1 cells were treated with active and inactive APD-CLD-derivatives under differential oxygen concentrations for 4 hours after which the cells were fixated in glutaraldehyde and analyzed by transmission electron microscopy. White arrows: collapsed mitochondria. Black arrows: regular mitochondria. N = 1. Fluorescence microscopy: MCF7 cells were treated with an APD-CLD-alkyne probe under hypoxia for 6 hours. The probe was conjugated to FAM-N3 via CuAAC. Subsequently, cells were stained with either rabbit anti-TOMM20 Ab (left panel) or rabbit anti-calnexin Ab (right panel) overnight followed by a goat anti-rabbit Ab conjugated with AlexaFluor647. Randomly selected pictures illustrate that the probe has strong co-localization with TOMM20. N = 1. MCF7 cells were treated with 2b, CCCP or DMSO under hypoxia or normoxia for 6 hours. Cells were stained with JC-1 (Cayman Chemical). Randomly selected pictures illustrate a hypoxia-selective loss of membrane potential after treatment with 2b. Viability assay of PANC-1 cells treated with CCCP: PANC-1 cells were treated with CCCP for 48 hours under hypoxia or normoxia and the viability was assessed with CellTiter Blue. The uncoupler did not display hypoxia-selective toxicity by itself. N = 3. Seahorse experiment: One-hour normoxic pretreatment of BxPC3 cells with BE-43547A2 does not alter the mitochondrial respiratory chain activity as measured by oxygen consumption. N = 7.

Figure 1b: Viability of PANC-1 cells treated with 1 or 2a for 48 hours under normoxia (red) or hypoxia (blue). Viability was assessed with CellTiter BlueTM (Promega). Figure 1d/Figure S1a: Normoxic treatment of PANC-1 cells with 2b for 8 or 24 hours followed by 24 hours of hypoxic incubation confirms the stability of the APD-CLDs in normoxic cell culture. N = normoxia, H = hypoxia. Viability was assessed as in (b). Figure 1e/Figure S1d: PANC-1 cells were treated with 2b under normoxia for 4.5, 8 or 24 hours before the medium was exchanged for fresh complete medium and cells were placed under hypoxia for additionally 24 hours. For comparison cells were kept under normoxia for 24 hours followed by 24 hours under hypoxia in presence of 2b (No wash). Figure S1c: PANC-1 cells were pretreated with hypoxia or normoxia for 47 hours before they were treated with rakicidin A under normoxia for 24 hours. N = 3. Figure S1f: Stability of rakicidin A (20 µM) in H2O/MeCN (1:1) containing either N-acetylcysteine (5 mM) and N-acetylglycine (5 mM). Injections of the mixtures were performed on a HPLC every third hour over the course of 1 day on a C18 column (Kinetex, 4.6 x 250 mm, C18, 5 µm, 100 Å, Phenomenex) equipped with a guard column and eluted with MeCN/H2O (40:60 --> 100:0 over 33 minutes). Retention time = 20-21 minutes. The absorbance of the APD chromophore was followed at 262 nm. N = 1.

Tumor volume and body weight: Tumor volume in mice bearing C3H mammary adenocarcinoma subcutaneous foot tumors, treated with vehicle (DMSO) or 5 mg/kg of BE-43547 for 7 consecutive days (indicated by arrows) starting when tumors reached 200 mm3 (day 1). Mice were sacrificed when the maximum allowable tumor volume was reached or when mice showed clear signs of morbidity, this was the case for two mice in the treated group. Body weight of the same mice can also be found in the Excel-file. 18FDG and HE staining of SiHa tumor slices: Visual comparison of intratumoral glycolytic activity (18FDG) and necrosis (HE) in representative SiHa tumor slices after treatment with a single IP injection of BE-43547 congeners at 10 mg/kg followed by one resting day, and then one IP injection on each 4 consecutive days at 5 mg/kg.One day after the last treatment, mice were injected IP with the glucose tracer analogue 18FDG. One hour later mice were sacrificed and tumors were removed and snap-frozen in pre-cooled isopentane. Subsequently, multiple 10 µm tumor sections were prepared using a cryostat, and analyzed for intratumoral glucose retention (FDG) with digital autoradiography. Following autoradiographic analysis, tissue sections were hematoxylin and eosin (H&E) stained for visualization of necrosis and digitalized using a Hamamatsu NanoZoomer slide scanner (Hamamatsu Photonics, Shizuoka, Japan).

Realtime apoptosis-necrosis assay: PANC-1 cells were treated with BE-43547A2 (3 µM) or the apoptosis-inducer raptinal (10 µM) under either hypoxia or normoxia while the membrane integrity and phosphatidylserine translocation was followed in real-time using a membrane impermeable DNA stain and an annexin V-luciferase fusion protein, respectively (RealTime-Glo™ Annexin V Apoptosis and Necrosis Assay, Promega, Cat#: JA1011). A delay between phosphatidylserine translocation and loss of membrane integrity designates a canonical apoptotic cell death. Data is represented in relative fluorescent and luminescent units. N = 3. Inhibition of cell death pathways: BE-43547A2-induced cell death in hypoxic PANC-1 cells cannot be blocked by the pan-caspase inhibitor Q-VD-OPh (24 h). As a positive control normoxic cells treated with the apoptosis-inducer raptinal could be rescued with Q-VD-OPh. Toxicity was measured by CellTox Green (Promega) and RFUs were normalized to the average RFU of untreated cells. N = 3. PANC-1 cells were treated with BE-43547A2 (100 nM) in combination with serial dilutions of necrostatin-5 (Nec-5) or DPQ for 48 hours under hypoxia. Toxicity was measured by CellTox Green. N = 3. Modulatory profiling: PANC-1 cells were treated with lethal concentrations of APD-CLDs or ferroptosis-inducers in combination with a panel of small molecule modulators of various cellular processes. Each data point represents the mean of three replicates. Viability was assessed with CellTiter Blue (Promega, Cat# G8080). Hits in the modulatory profiling were confirmed by treating PANC-1 cells with serial dilutions of APD-CLDs or ferroptosis-inducers in the absence or presence of the hit compound for 48 hours under hypoxia. N = 3. PANC-1 cells treated with APD-CLDs can be rescued by the ER stress inhibitor azoramide. N = 3. Full dose-response curves of the modulation of ferroptosis with the ferroptosis-inhibitors NHI-2 and antimycin A. N = 3.

Table S1: PANC-1 cells were treated with 2a or DMSO for 12 hours under hypoxia after which metabolites were extracted and identified by LC-qTOF-MS. Statistically significant changes (p < 0.05, VIP > 1) in metabolite levels are plotted relative to DMSO-treated cells. N = 10. Figure 3c: GSH-depletion in PANC-1 cells by APD-CLDs were confirmed by GSH-GloTM assay (Promega) after 12 hours under hypoxia or normoxia. Sidak’s multiple comparisons test. N = 2. Likewise, the ROS levels was quantified by fluorescence of the ROS-activated fluorophore H2DCFDA (Molecular Probes) in normoxic versus hypoxic PANC-1 cells treated with 1. Data is normalized to the average of the two lowest dosings of rakicidin A. N = 3. Unpaired t-test. Figure 3d + GSEA: PANC-1 cells were treated as in described for table S1 and after 12 hours total RNA was isolated. Transcripts were sequenced, quantified and the combined gene expression of 1 and 2a-treated cells was plotted relative to DMSO-treated cells. N = 3. See full RNAseq dataset via link below. Figure 3f: RT-qPCR on CHAC1 transcription after treatment with erastin, 1 and 2b for 12 hours under hypoxia or normoxia. All data is normalized to normoxic DMSO treatments. Sidak’s multiple comparisons test (hypoxia vs. normoxia). N = 3.

Figure 2b: PANC-1 cells were treated with an APD-CLD alkyne probe or an APD-CLD-deficient alkyne probe (125-1000 nM, 2-fold dilution) for 12 hours under hypoxia. The lysates were tagged with rhodamine-N3 via CuAAC, separated by electrophoresis and visualized by excitation of rhodamine B. Figure 2c: Protein targets in living PANC-1 cells were enriched by CuAAC to biotin-PEG3-N3 followed by pulldown on streptavidin-coated Dynabeads. The enriched proteins were separated by electrophoresis and visualized by SimplyBlue Safe Stain. Figure S2a: In-gel fluorescence analysis of covalent binding partners for the APD-CLD alkyne probe. PANC-1 cells were treated with siRNA against RTN3, RTN4, CYB5B or non-targeting control siRNA, and subsequently treated with an APD-CLD alkyne probe for 12 hours under hypoxia. Then cells were lysed and lysates were tagged by CuAAC to rhodamine B azide. The proteins were separated by SDS-PAGE and analyzed for fluorescence in a CCD camera. Subsequently, the gel was stained by SimplyBlue SafeStain for total protein visualisation. Figure S2c,d,e,f: See [Viability data.xlsx]. Cells were treated with either the APD-CLD alkyne probe or electrophilic acrylamide RTN4 ligands for 48 hours under normoxia or hypoxia. Viability was assessed by CellTIter Blue (Promega) and is given as RFUs.

Raw data for publication