Carbonylcyanide 4-trifluoromethoxyphenylhydrazone

We measured the mitochondrial respiratory capacity of the permeabilized EAT at 37 °C using a high-resolution respirometry (Oxygraph-2k, Oroboros Instruments, Innsbruck, Austria) as described^{15 (link),22 (link)}. After a careful manual dissection of the capillaries and connective tissues with the use of a magnifying glass, sample tissues (approx. 50 mg) were put into the respirometer chamber filled with 2 mL of MiR05 (in mmol/L: sucrose 110, K-lactobionate 60, EGTA 0.5, 0.1% BSA, MgCl₂ 3, taurine 20, KH₂PO₄ 10, HEPES 20, pH 7.1). Digitonin (2 μmol/L) was added to the chamber to permeabilize the tissue samples.
After the stabilization of baseline respiratory rates, the following respiratory substrates, ADP, and an uncoupler were added in the following order as described^{15 (link)}: (1) glutamate (final concentration, 10 mmol/l) and malate (2 mmol/L) (complex I-linked substrates); (2) ADP (5 mmol/L); (3) octanoyl-l-carnitine (0.15 mmol/L) (a fatty acid); (4) succinate (10 mmol/L) (a complex II-linked substrate); (5) cytochrome c (10 μmol/L); and (6) titration of carbonylcyanide p-trifluoromethoxyphenylhydrazone (FCCP; 0.5 µmol/L increments) (an uncoupler). The integrity of the outer mitochondrial membrane was tested by the addition of cytochrome c. An increase in oxygen consumption rate indicates damaged outer mitochondrial membrane because cytochrome c does not pass the intact outer mitochondrial membrane^{23 (link)}, but in this study, there was no increase in oxygen consumption rate after addition of cytochrome c in all patients. The respiratory rate (i.e., the O₂ consumption rate) values are expressed as the O₂ flux normalized to the permeabilized tissue mass (pmol/s/mg wet weight of EAT).

Mitochondrial respiration was determined using a Seahorse XFe96 extracellular flux analyzer (Agilent, Santa Clara, CA, USA) as previously described [8 (link)]. Single spheroids of a defined size and composition were formed by seeding 2.0 × 10⁴ cells into 96 well round bottom ultra-low adherence plates (Corning, Corning, NY, USA) and centrifuging for 5 min at 900 rpm. All cells were incorporated into small spheroids of approximately the same size by 12 h. To keep treatment times consistent between various times of adherence, the spheroids were transferred to the 96 well cell culture plate (Agilent) used for the assay at the appropriate time before the assay was run 24 h after the initial formation of all spheroids. Adherence time points included 0 h (for which spheroids were transferred directly to assay plates without time for adherence), 4, 8, and 12 h. For some experiments, cells were treated for 24 h and then allowed to adhere for 24 h prior to the assay run.
Preliminary optimization studies to achieve reproducible measurements of the metabolic responses of spheroids in the XFe96 were conducted to determine experimental protocol and inhibitor concentration. Experiments consisted of 3 min mixing, a 2 min wait step, and 3 min measurement cycles. Prior to the assay, the medium was replaced with 180 µL of serum-free, phenol-red-free, bicarbonate-free medium. Oxygen consumption rate (OCR) was measured under basal conditions for three cycles, then in the presence of ATP synthase inhibitor oligomycin (1.0 μmol/L), mitochondrial uncoupler carbonylcyanide-p-trifluoromethoxyphenylhydrazone FCCP (3.0 μmol/L, Sigma), and complex I inhibitor rotenone with cytochrome C oxidase inhibitor antimycin A (1.0 μmol/L rotenone + antimycin A) for five cycles each. All experiments were performed at 37 °C. Total traces as well as basal respiration, maximum respiration (after FCCP addition), spare respiratory capacity (basal OCR—maximum OCR), and ATP synthesis rate (basal OCR—oligo-stimulated OCR) were calculated. Non-mitochondrial respiration, represented by the OCR remaining after the addition of rotenone/antimycin A, was subtracted from all values to report only mitochondrial OCR. Following outgrowth imaging, spheroids were washed 1× with PBS and placed in 30 μL of RIPA buffer for protein analysis by Pierce^TM BCA Protein Assay (Thermo Fisher Scientific, Waltham, MA, USA). Data were presented as the mean ± SEM of three independent biological experiments performed in eight or more replicates.

Hydroquinone reconstitution. Lyophilized hydroquinones were reconstituted in dimethyl sulfoxide (DMSO). For the experimental trials, 0.4% DMSO was used as a vehicle, which corresponds to a non-toxic concentration in vitro [[43] (link), [44] (link), [45] (link)].
Human platelet purification. The blood sample was obtained from voluntary donors (10 days without antiplatelet medication) who agreed to participate after signing informed consent (Scientific Ethics Committee of the University of Talca N° 04–2022) [40 (link),46 (link)]. Briefly, blood anticoagulated with acid-citrate-dextrose (ACD) was obtained in a 4:1 v/v ratio. The sample was centrifuged for 12 min at 250 g to obtain platelet-rich plasma (PRP). Subsequently, the PRP was centrifuged for 8 min at 900 g to obtain the platelet pellet. The plasma supernatant was removed, and the platelets were resuspended in Tyrodes buffer without calcium plus ACD in a ratio of 5:1 v/v. The platelets were centrifuged once more for 8 min at 900 g and the pellet was resuspended in Tyrodes buffer without calcium [47 (link)]. Finally, the platelets were counted (Mindray BC-3000 Plus hematology counter, Japan), adjusted to the desired concentration, and used within 3 h [37 (link)].
Cellular cytotoxicity by the release of lactate dehydrogenase (LDH). Washed platelets (200-250 × 10⁶ platelets/mL) were incubated with the compounds under study or vehicle for 15 min at 37 °C. Then, the platelets were centrifuged at 900 g for 8 min to collect the supernatant. The supernatant was mixed in equal parts with the working reagent of the lactate dehydrogenase cytotoxicity kit (Cayman Chemical, Ann Arbor, MI, USA) and left to react for 30 min at 37 °C. The absorbance was analyzed at 490 nm (Microplate Reader Thermo Scientific Multiskan Go, Finland) [31 (link)].
Computational modeling. Atomistic models of a mitochondrial membrane were constructed, both in water and in the presence of the study compounds NP4 and NPO4, as described previously [17 (link)]. Briefly, the membrane, comprising 72 lipids per leaflet, was generated using VMD 1.94, with the lipid proportions as follows: 40% POPC, 30% POPE, 15% cardiolipin, 2.5% POPA, 2.5% palmitoyl-oleoyl phosphatidylserine, and 10% 1,2-diacyl-sn-glycero-3-phosphoinositol. Molview [48 (link)] and Open Babel [49 (link)] were employed to build the 3D structures of NP4 and NPO4. The models were optimized using HF/6-31G*, and single-point calculations with the polarized continuum model were performed to obtain the electrostatic molecular potential for charge matching. CHELP charges were calculated for each atom, with +1 added for each compound [50 (link)]. Gaussian 09 was used for ab-initio calculations, and Paramchem was utilized for assigning the parameters for molecular dynamics (MD) calculations [51 (link)].
Boxes were prepared for each molecule, with the membrane, and TIP3P water molecules [52 (link)] were used as a solvent, with NaCl added to neutralize and reach a concentration of 0.15 M. NAMD v2.14 and the CHARM36 force field was employed for MD simulations [53 (link)]. The box underwent a 30,000-step minimization, followed by a 10 ns equilibration in an isobaric-isothermal ensemble at 300 K and 1 atm. Soft harmonic constraints were applied and gradually reduced from 10 to 0 kcal mol⁻¹ Å⁻². Particle Mesh Ewald [54 (link)] was used with a cutoff of 9 Å. The time step was 4 fs, applying the hydrogen mass partitioning method [55 (link)].
Steered molecular dynamics (SMD) [[56] (link), [57] (link), [58] (link)] was applied to study how NP4 and NPO4 compounds progress through a path crossing the lipid bilayer at a speed of 20 Å/ns over 20 ns. The reaction coordinate was defined considering the distance between the center of mass of the heavy atoms of the compounds and the nitrogen atoms of the lipids. Free energy profiles were calculated for each compound by selecting equispaced coordinates along the SMD path and minimizing them for 1000 steps. For each system studied, 35 windows separated by 1.0 Å were used with a polarization harmonic constraint with a strength of 2.5 kcal mol⁻¹ Å⁻². Each window spanned 5 ns, yielding 175 ns of sampling per compound. The potential mean force (PMF) was obtained using the weighted histogram analysis method (WHAM) [59 (link)].
Cell viability by Calcein-AM. Washed platelets (200-250 × 10⁶ platelets/mL) were labeled with 0.1 μM Calcein-AM and incubated for 15 min at 37 °C protected from light. Then were incubated with the compound NP4 (1, 2, 5, 10 and 20 μM) for another 15 min at 37 °C in the dark. Viability was analyzed on the BD Facs Lyric Flow Cytometer (BD Biosciences, USA). The percentage of Calcein-negative platelets (without fluorescence) in the CD61-PE-positive subpopulation (BD Biosciences, San José, CA, USA) was identified as non-viable platelets. As a control of cell damage, Triton X-100 0.1% was used [31 (link)]. Representative dot plots are available in Supplementary Fig. 15. For all Flow Cytometry experiments, platelet purity (>99%) was confirmed using anti-CD61-FITC or anti-CD61-PE antibodies (Supplementary Fig. 7).
Cellular apoptosis (externalization of phosphatidylserine). Washed platelets (200-250 × 10⁶ platelets/mL) were incubated with the NP4 compound (1, 2, 5, 10 and 20 μM) for 15 min at 37 °C. Then 15 μL of the reaction was mixed with 25 μL of annexin V binding buffer (1x) and labeled with 2 μL of annexin V FITC (BD Pharmingen, FITC Annexin V Apoptosis Detection Kit I) [47 (link)]. The sample was acquired with the BD FACS Lyric flow cytometer (BD Biosciences, USA) and the platelets were identified as the CD61-PE positive population. As a control for phosphatidylserine externalization, platelets were stimulated with 2 μg/mL Collagen and 10 μM TRAP-6 [60 (link)].
Platelet aggregation. Washed platelets (200-250 × 10⁶ platelets/mL) were preincubated with 2 mM CaCl₂ and then with the compounds under study for 5 min at 37 °C in the aggregometer (Agg RAM- Helena Biosciences). Aggregation was initiated with the agonists: collagen 2 μg/mL; convulxin 20 ng/mL; TRAP-6 (Thrombin receptor activator peptide 6) 5 μM; PMA (Phorbol myristate acetate) 100 nM or arachidonic acid 250 μg/mL. The aggregation curve was evaluated for 5 min at 37 °C in shaking at 1000 rpm [61 (link)]. Half maximal inhibitory concentration (IC₅₀) was calculated in the compounds that showed over 50% inhibition of platelet aggregation [62 (link)].
Platelet aggregation in platelet-rich plasma. The blood sample was obtained from voluntary donors (10 days without antiplatelet medication) who agreed to participate after signing informed consent (Scientific Ethics Committee of the University of Talca N° 04–2022) [40 (link),46 (link)]. Venous blood was collected by phlebotomy into 3.2% sodium citrate tubes. The tubes were centrifuged at 250 g for 12 min to obtain PRP. After reserving the PRP, the tubes were centrifuged again at 900 g for 10 min to obtain platelet-poor plasma (PPP). For the aggregation reaction, PRP was adjusted to 200-250 × 10⁶ platelets/mL with PPP [[63] (link), [64] (link), [65] (link)]. Platelet counting was performed in a counter (Mindray BC-3000 Plus hematology counter, Japan). The samples were preincubated with the compounds under study for 5 min at 37 °C in the aggregometer (Agg RAM- Helena Biosciences). Aggregation was initiated with collagen 2 μg/mL and the reaction was evaluated for 5 min at 37 °C in shaking at 1000 rpm [61 (link)].
Platelet activation markers. Washed platelets (200-250 × 10⁶ platelets/mL) were incubated with the compounds (1, 5 and 10 μM), then activated with collagen 2 μg/mL and incubated for 5 min at 37 °C. Aliquots of the reaction were then collected and labeled separately with human antibodies (BD Biosciences, San José, CA, USA) for P-selectin, CD63, activated GPIIb/IIIa (PAC-1), and fibrinogen to assess the levels of each marker on the platelet membrane. The sample was acquired using the BD Facs Lyric flow cytometer (BD Biosciences, USA). Representative dot plots are available in Supplementary Figures 8-11 and 20-23. In the case of the activation markers P-selectin and bound fibrinogen in PRP, the same procedure for washed platelets was performed with representative dot plots available in Supplementary Figs. 15 and 16. The platelet population was identified with anti-CD61-PE or anti-CD61-FITC (BD Biosciences, San José, CA, USA) [66 (link),67 (link)].
Mitochondrial membrane potential (ΔΨm). Washed platelets (50 × 10⁶ platelets/mL) were labeled with a 100 nM tetramethylrhodamine methyl ester (TMRM) probe and incubated with compounds (1, 5 and 10 μM) for 20 min at 37 °C. The sample was acquired in the BD Facs Lyric flow cytometer (BD Biosciences, USA). Representative dot plots are available in Supplementary Figs. 12, 24, and 25. Carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP) 1 μM was used as a control for decreasing the mitochondrial membrane potential [31 (link),38 (link)].
Mitochondrial ROS (mtROS) Levels. The MitoSOX probe was used to identify mitochondrial ROS production specifically [68 (link)]. Washed platelets (50 × 10⁶ platelets/mL) were labeled with the MitoSOX® Red probe (Invitrogen, Carlsbad, CA, USA) 10 μM and incubated for 20 min at 37 °C protected from light [69 (link)]. After, compounds (1, 5, and 10 μM) were added to the platelets and incubated for 10 min at 37 °C. The sample was acquired in the BD Facs Lyric flow cytometer (BD Biosciences, USA) (Representative dot plots available in Supplementary Figs. 14 and 19). As a positive control for mtROS, antimycin A 20 μM was used.
Intraplatelet ROS levels. Washed platelets (50 × 10⁶ platelets/mL) were labeled with the dihydroethidium (DHE) probe 10 μM and incubated for 20 min at 37 °C protected from light. After, compounds (1, 5, and 10 μM) were added to the platelets and incubated for 10 min at 37 °C. The sample was acquired in the BD Facs Lyric flow cytometer (BD Biosciences, USA). Representative dot plots are available in Supplementary Figs. 18 and 27. As a positive control for ROS, antimycin A 20 μM was used [38 (link),66 (link)].
Intraplatelet calcium levels. Washed platelets (200-250 × 10⁶ platelets/mL) were labeled Fluo-4-AM probe with 0.5 μM and incubated for 20 min at room temperature protected from light. Subsequently, platelets were adjusted to a count of 50 × 10⁶ platelets/mL, and compounds (1, 5, and 10 μM) were added and incubated for 10 min at 37 °C. The sample was acquired in the BD Facs Lyric flow cytometer (BD Biosciences, USA). Representative dot plots are available in Supplementary Figs. 13 and 26. As a control for increased intracellular calcium, P-trifluoromethoxyphenylhydrazone carbonylcyanide (FCCP) 1 μM was used [60 (link)].
Oxygen consumption and extracellular acidification rate. Oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were measured with a Seahorse XFe24 extracellular flux analyzer (Agilent, Santa Clara, CA, USA). Briefly, 100 μL of platelets washed in modified Tyrode-HEPES buffer were deposited (20–25 × 10⁶ cells/well) and then centrifuged at 300 g for 10 min. Platelets were incubated with NP4 10 μM for 5 min, and then Tyrode's-HEPES buffer was removed and Seahorse medium (8.3 g/L DMEM, 1.85 g/L NaCl, 5 mM glucose, 1 mM pyruvate, 2 mM glutamine, 5 mM HEPES, pH 7.4) was added to a final volume of 600 μL [70 (link)]. OCR was measured before and after the sequential addition of 3 μg/mL collagen, 2.5 μM oligomycin, 1.4 μM FCCP, and 2 μM/2 μM antimycin A/rotenone [71 (link)]. Non-mitochondrial OCR was subtracted from all measurements. Respiratory parameters obtained with Mito Stress Test (Supplementary Table 1) were calculated as follows: Baseline (baseline OCR), collagen (OCR after collagen addition), activation (collagen-basal), ATP-independent or proton leak (OCR resistant to oligomycin addition), ATP-linked respiration (basal-proton leak), maximum (OCR obtained after addition of FCCP), and spare (maximal-basal), coupling efficiency (OCR Basal – OCR ATP-indep)/OCR Basal) [72 (link)]. Respiratory parameters were calculated according to Refs. [66 (link),70 (link),73 (link)].
Proton pumping and activity of reconstituted complex I. Purification of complex I from Y. lipolytica mitochondrial membranes and reconstitution of complex I into Proteoliposomes was performed as previously described [74 (link),75 (link)]. Briefly, complex I was reconstituted in proteoliposomes (Cxl), and then ACMA (9-Amino-6-Chloro-2-Methoxyacridine) was added to quantify the proton pumping activity. Fluorescence changes were monitored at 25 °C in a Fluorolog-3 spectrofluorometer (Horiba Scientific). Settings: λ_ex = 430 nm, λ_em = 475, slit 5 nm, time increment 0.2 s. Proteoliposomes (15 μg protein) were diluted in 2 ml assay buffer (20 mM Mops pH 7.2, 80 mM KCl, 0.5 μM valinomycin). After starting the measurement, 0.5 μM ACMA, 70 μM decylubiquinone (DBQ), 125 μM NADH, NP4 (5; 25 or 50 μM)/vehicle, and Carbonyl cyanide m-chlorophenyl hydrazone (CCCP) 5 μM were added after 50, 100, 150, 200 and 300 s, respectively [76 (link)]. The measurements were done in triplicate for 5, 15, 30, 45, 60 and 100 μM NP4. The Average over the time after the addition of NADH/NP4/CCCP till the next addition was taken from each measurement. The quenching after NADH the addition was taken as the zero % value and the dequench with CCCP was set to 100%. The zero value was subtracted from the dequenching with NP4 and the max dequenching with CCCP. The DMSO negative control was measured in duplicate.
Bleeding time in vitro Innovance PFA-200 system. Blood was obtained in 3.2% citrate tubes which were allowed to settle for 30 min before the test. Whole blood was mixed with NP4 10 μM and incubated for 10 min. Subsequently, the sample was loaded into the Collagen/Epinephrine or Collagen/ADP cartridge and the closure time was measured in the Innovance PFA-200 system (Siemens Healthcare Diagnostics Products, Munich, Germany). Eptifibatide 10 μM was used as a control [77 (link)].
Clot retraction. Blood was obtained in 3.2% citrate tubes, which were centrifuged for 12 min at 250 g to obtain platelet-rich plasma (PRP). The assay was prepared in Khan tubes by adding 750 μL of tyrodes buffer without calcium, 200 μL of PRP, and 5 μL of red blood cells; the mixture was treated with NP4 10 μM or vehicle and incubated for 10 min. A glass rod was positioned in the center of the tube (as a support for clot adhesion) and clot retraction was initiated by adding 50 μl of 10 Units/mL thrombin (final concentration 0.5 Units/mL). Photographs were recorded at 0; 15; 30; 60; 90; and 120 min time intervals. The clots formed after 2 h were weighed to quantify retraction [78 (link)].

Mitochondrial respiration was determined using the XFe96 extracellular flux analyzer (Agilent, Santa Clara, CA, USA) as previously described [23 (link)]. Single spheroids were formed in 96-well ultra-low adherence plates at 1.5 × 10⁴ density for 24 h under the indicated conditions. Single spheroids were transferred to a XFe96 cell culture plate 4 h before the assay to allow for slight adherence of spheroids to the assay plate to avoid disturbance by the injections of inhibitors. Prior to the assay, the medium was changed to serum-free, bicarbonate-free medium. Experiments consisted of 3 min mixing, 2 min wait, and 3 min measurement cycles. Oxygen consumption rate (OCR) was measured under basal conditions and upon addition of mitochondrial inhibitors oligomycin (1.0 μmol/L), carbonylcyanide-p- trifluoromethoxyphenylhydrazone FCCP (3.0 μmol/L), and a combination of rotenone and antimycin A (1.0 μmol/L). These conditions were determined in preliminary optimization studies to achieve reproducible metabolic responses in 3D culture. All experiments were performed at 37 °C.