Multimodal antiproliferative effects of oleanolic acid mitocans: In vitro and in vivo studies

Multimodal antiproliferative effects of oleanolic acid mitocans: In vitro and in vivo studies Adri´ an Puerta a , Aday Gonz´ alez-Bakker a , Eduardo Romanos b,c,d , Inmaculada Aguilar Domínguez e , M´ onica Martínez-Montiel e,f , Pen´ elope Merino-Montiel f , Raquel P. Herrera c,* , M. Concepci´ on Gimeno d , Jos´ e G. Fern´ andez-Bola˜ nos e , ´ Oscar L´ opez e,* , Jos´ e M. Padr´ on a,* a BioLab, Instituto Universitario de Bio-Org´ anica Antonio Gonz´ alez (IUBO-AG), Universidad de La Laguna, Apartado 456, E-38200 La Laguna, Spain b Servicio de Imagen M´ edica y Fenotipado, Instituto Aragon´ es de Ciencias de la Salud, Centro de Investigaci´ on Biom´ edica de Arag´ on (CIBA), Avda. San Juan Bosco, 13, Planta D, E-50009 Zaragoza, Spain c Laboratorio de Organocat´ alisis Asim´ etrica, Departamento de Química Org´ anica, Instituto de Síntesis Química y Cat´ alisis Homog´ enea (ISQCH), CSIC-Universidad de Zaragoza, C/Pedro Cerbuna 12, E-50009 Zaragoza, Spain d Departamento de Química Inorg´ anica, Instituto de Síntesis Química y Cat´ alisis Homog´ enea (ISQCH), CSIC-Universidad de Zaragoza, C/Pedro Cerbuna 12, E-50009 Zaragoza, Spain e Departamento de Química Org´ anica, Facultad de Química, Universidad de Sevilla, Apartado 1203, E-41071 Seville, Spain f Facultad de Ciencias Químicas, Benem´ erita Universidad Aut´ onoma de Puebla, Ciudad Universitaria, 72570 Puebla, PUE, Mexico ARTICLE INFO Keywords: Oleanolic acid Mitocan Triphenylphosphonium Antitumor agent Paraptosis Continuous live-cell imaging ABSTRACT Mitochondria-targeting drugs (mitocans) based on organic cations are emerging as powerful and selective cancer therapeutics. In this study, we have evaluated a novel series of oleanolic acid-derived mitocans, revealing nanomolar-range antiproliferative effects against human solid tumor cells. Continuous live-cell imaging revealed extensive cytoplasmic vacuolation, while mechanistic studies identified paraptosis as the dominant form of cell death. Remarkably, in vivo experiments demonstrated significant tumor growth inhibition in mice, with no detectable toxicity at therapeutic doses. These findings highlight the potential of oleanolic acid-derived mitocans as promising candidates for cancer therapy. 1. Introduction Mitochondria play a pivotal role in the regulation of cell death and survival in response to various stimuli [1]. The release of proteins located in the mitochondrial membrane –i.e. cytochrome c to the cytosol– serves as an effector of programmed cell death. In a similar way, mitochondria participate in autophagy through mitophagy. Programmed autophagy of this organelle ensures a correct turnover, allowing the degradation of non-functional mitochondria and facilitating the synthesis of new entities. Furthermore, alterations in mitochondrial functioning associate with tumor resistance to drugs via the development of apoptosis resistance, maintenance of metabolic requirements through exploitation of autophagy products, and enhanced biogenesis of the organelle by increased mitochondrial fission [2]. The relevance of mitochondrial metabolism in tumors appears in the literature extensively, with mutations occurring in p53, PI3K, myc or KRAS that influence the reprogramming of mitochondrial function to facilitate tumor development [3]. Given the role of mitochondria in cell death and metabolism and their –deeply demonstrated– relevance in tumors with high energy demands, targeting mitochondrial function has gained Abbreviations: 3-MA, 3-methyladenine; BA, betulinic acid; CCCP, carbonyl cyanide 3-chlorophenylhydrazone; CHX, cycloheximide; CI, confidence interval; CPT, camptothecin; CQ, chloroquine; cytarabine, Ara-C; DAPI, 4 ′ ,6-diamidino-2-phenylindole; DCF, dichlorofluorescein; DCFDA, 2 ′ ,7 ′ -dichlorodihydrofluorescein diacetate; FDA, Food and Drug Administration; GA, glycyrrhetinic acid; GI 50 , 50% growth inhibition; IC 50 , 50% inhibitory concentration, LC3, microtubule-associated protein light chain 3; MAPK, mitogen-activated protein kinase; MDC, monodansyl cadaverine; MMP, mitochondrial membrane potential; MoA, mode of action; OA, oleanolic acid; PBS, phosphate buffer saline; PTX, paclitaxel; Rf, resistance factor; RFU, relative fluorescence units; ROS, reactive oxygen species; SDS, sodium dodecyl sulfate; SEM, standard error of the mean; TMRE, tetramethylrhodamine ethyl ester; TPP, triphenylphosphonium; UA, ursolic acid; VP, verapamil. * Corresponding authors at: Instituto Universitario de Bio-Org´ anica Antonio Gonz´ alez, Universidad de La Laguna, PO Box 456, 38200 La Laguna, Spain. E-mail addresses: [email protected] (R.P. Herrera), [email protected] (´ O. L´ opez), [email protected] (J.M. Padr´ on). Contents lists available at ScienceDirect Biochemical Pharmacology journal homepage: www.elsevier.com/locate/biochempharm https://doi.org/10.1016/j.bcp.2025.116807 Received 25 October 2024; Received in revised form 11 February 2025; Accepted 13 February 2025 Biochemical Pharmacology 234 (2025) 116807 Available online 18 February 2025 0006-2952/© 2025 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). importance and centered efforts during the last few years in cancer treatment research [1–3], representing a way to selectively target tumor cells. Focusing on chemotherapy against mitochondria, there are multiple strategies that can be used to directly and/or indirectly inhibit its function. Intervention over metabolic processes that occur in the mitochondria is one of them. To impede the normal electron transport chain, one strategy is to inhibit the synthesis of biomolecules through the alteration of the tricarboxylic acid cycle, thus reducing cancer cell growth. The antidiabetic drug metformin has been widely studied preclinical and clinically, both alone and in combination with other drugs, with promising results [4]. Other strategies explored include direct inhibition of the tricarboxylic acid cycle (i.e. inhibition of mitochondrial glutaminase by telaglenastat or disruption or inhibition of pyruvate dehydrogenase kinase by dichloroacetate) as well as the inhibition of fatty acid synthesis [5]. Targeting mitochondrial metabolism is an effective form of combination therapy. Coupling inhibitors of the metabolic pathways occurring in this organelle with known FDAFig. 1. Mitochondria-targeting scaffolds: (A) Chemical structure of mitochondria-targeting vectors. Compiled from [9]; (B) Naturally occurring pentacyclic triterpenoids and representative TPP derivatives. The number refers to the numbering used in the corresponding original work. IC 50 values reported against human solid tumor cells. A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 2 approved drugs (e.g. mTOR inhibitor everolimus) have resulted in better efficacy [4]. Another mitochondria-targeting strategy is the induction of mitochondrial outer membrane permeabilization. This produces the release of the proteins located in the intermembrane space to the cytosol, which include activators of caspases and cytochrome c, subsequently inducing cell death [6]. The mitochondrial membrane consists of a double layer. Externally, and in contact with the cytoplasm, there is a porous membrane. Internally, it is a protein-rich membrane. This structure tightly regulates mitochondrial metabolism and prevents xenobiotics from crossing the mitochondrial membrane. Therefore, mitocans (i.e. drugs targeting the mitochondria) must be internalized in the organelles crossing the double-layered membrane. To achieve success, drug delivery to the mitochondria is of pivotal importance. The mitochondrial membrane potential of normal cells is set at 150–180 mV so lipophilic cations tend to accumulate in the mitochondrial matrix [7]. The mitochondria of cancer cells differ from those of healthy cells in membrane polarization, displaying −220 and −160 mV, respectively [8]. This difference allows for the fast and selective uptake of positively charged molecules. This property led to the development of diverse mitochondria-targeting vectors [9]. Vectorizing anticancer compounds with selected cationic derivatives represents an extensively explored strategy with promising results (Fig. 1A). Indubitably, the most studied of such cations is the triphenylphosphonium (TPP) ion. TPP has been widely used to obtain mitocans given its accessibility through chemical synthesis [10]. TPP possesses both a positive charge (delocalized over the three phenyl groups and stabilized by resonance) and lipophilicity to facilitate its internalization into the mitochondria [11]. These hydrophobic cations can diffuse through the lipid bilayers because their net charge is scattered along its surface. Moreover, TPP is easy to link and clinical trials demonstrated compatibility with physiological conditions and a lack of relevant toxicity, even at relatively high doses [12]. Chemotherapeutic drugs such as doxorubicin, tamoxifen, chlorambucil, docetaxel and gamitrinib, and other substances as honokiol, curcumin, betulinic acid (BA) and ursolic acid (UA) are just some examples of antiproliferative molecules bearing TPP ion to enhance their potency and selectivity [8]. Earlier, we reported the mitochondriotropic activity of masked phenolic scaffolds following this strategy [13]. These novel mitochondriotropic agents showed promising antiproliferative profiles, displaying remarkable potency and selectivity, including multidrug-resistant cells, upon co-administration with a pump-efflux inhibitor. TPP cations of pentacyclic triterpenoids showed potential anticancer activity. In particular, derivatives of glycyrrhetinic acid (GA, enoxolone) [14], UA [15] and BA [16–18] displayed activity against cancer cell lines (Fig. 1B). Encouraged by these results, we turned our attention to oleanolic acid (OA) with the idea to apply the same strategy to obtain new mitocans. OA is another pentacyclic triterpenoid present in more than 1620 plants (both, edible and medicinal). It is obtained in its pure form or with structural modifications. These pentacyclic triterpenoids –including OA– bear two functional groups that might be used for linking the TPP motif, a hydroxy group at position C-3, and a carboxylic acid on C-30 for GA, and C-28 for UA, BA and OA. The results of the biological evaluation revealed a better bioactivity profile for the carboxylic acid derivatives. Our main target pursued herein has been the development of a short series of OA-TPP hybrids as an example of organelle-directed anticancer therapy, and to study their mechanism in vitro, and their efficiency in a xenograft mice model. During the course of our investigations, Li and coworkers [19] reported a series of OA derivatives linked to TPP. In their study, they claimed that potential PI3K inhibition could be a mode of action (MoA) that induce the antiproliferative effects of the small molecules described. The lead compound was one of the compounds we were studying. In this work, we inform on the antiproliferative activity observed in vitro and the studies carried out to identify the MoA of a small family of OA-TPP hybrids. The lead compound underwent further in vivo studies in mice and its effect on tumor growth reduction will be shown. 2. Materials and methods 2.1. Reagents and chemicals OA (TCI Europe, Belgium), N,N-dimethylformamide (DMF, VWR, USA), α , ω -dibromoalkanes (Merck, Germany), potassium carbonate (K 2 CO 3 , Panreac, Spain), ethyl acetate (EtOAc, VWR, USA), hydrochloric acid (HCl, Panreac, Spain), sodium bicarbonate (NaHCO 3 , Panreac, Spain), sodium chloride (Panreac, Spain), sodium sulfate (Na 2 SO 4 , Panreac, Spain), cyclohexane (VWR, USA), acetonitrile (MeCN, VWR, USA), triphenylphosphine (Merck, Germany), RPMI 1640 (VWR, USA), fetal bovine serum FBS, (Biowest, France), L-glutamine (Lonza, Switzerland), penicillin G/streptomycin (Lonza, Switzerland), trypsin/ ethylenediaminetetraacetic acid (EDTA) (Lonza, Switzerland), sulforhodamine B (SRB, Merck, Germany), DMSO (VWR, USA), cytarabine (Ara-C, Merck, Germany), verapamil (VP, Merck, Germany), paclitaxel (PTX, Merck, Germany), vinblastine (VBL, Merck, Germany), phosphate buffer saline (PBS, Lonza, Switzerland), methanol (VWR, USA), crystal violet (Merck, Germany), sodium dodecyl sulfate (SDS, Merck, Germany), 3-chlorophenylhydrazone (CCCP, Merck, Germany), hydrogen peroxide (H 2 O 2 , Merck, Germany), tetramethylrhodamine ethyl ester (TMRE, Merck, Germany), glycerol (Merck, Germany), 2 ′ ,7 ′ -dichlorofluorescein diacetate (DCFDA, Merck, Germany), paraformaldehyde (PFA, Merck, Germany), ammonium chloride (NH 4 Cl, Merck, Germany), triton (Merck, Germany), 4 ′ ,6-diamidine-2 ′ -phenylindole dihydrochloride (DAPI, Merck, Germany), bis-benzimide H33342 trihydrochloride (Hoechst 33342, Merck, Germany), 3-methyladenine (3-MA, Merck, Germany), chloroquine (CQ, Merck, Germany), cycloheximide (CHX, Merck, Germany), Z-DEVD-FMK (Merck, Germany), luciferin (DLuciferin, potassium salt, DELTACLON S.L., Spain). 2.2. Chemistry 2.2.1. General procedure for the preparation of bromoalkyl derivatives I–IV To a solution of OA (1) (300 mg, 0.66 mmol) in DMF (5 mL), was added the corresponding α , ω -dibromoalkane (5.0 equiv.) and K 2 CO 3 (1.0 equiv.). The mixture was stirred at rt for 12 h. Then, it was diluted with H 2 O (30 mL) and extracted with EtOAc (2x15 mL). The combined organic fractions were washed with 1 M HCl (20 mL), sat. aq. NaHCO 3 (20 mL) and brine (20 mL), dried over Na 2 SO 4 and filtered. The filtrate was concentrated to dryness and purified by column chromatography (98:2 → 95:5 cyclohexane–EtOAc). 2.2.2. General procedure for the preparation of triphenylphosphonium salts 2–5 To a solution of bromoalkyl derivatives I–IV (1.0 equiv.) in MeCN was added PPh 3 (5.0 equiv.). Reaction was stirred at 115 ◦C (FisherPorter tube) for 24 h. Then, it was concentrated to dryness and the residue was purified by column chromatography. All compounds were characterized by spectroscopic ( 1 H and 13 C NMR) and spectrometric (HRMS) methods. 1 ′ ’-{4 ′ -[(3 ′ ’β-Hydroxyolean-12 ′ ’–3n-oyl)oxy]butyl}triphenylphosphonium bromide (2). Bromide I (100 mg, 0.17 mmol), PPh 3 (222 mg, 0.85 mmol) and MeCN (7.5 mL) were used. Column chromatography (80:1 → 10:1 CH 2 Cl 2 –MeOH afforded 2 as a white solid. Yield: 110 mg (78 %). [ α ]25 D +32 (c 0.85, CH 2 Cl 2 ); 1 H NMR (300 MHz, CDCl 3 ) δ 7.92–7.67 (m, 15H, Ar-H), 5.14 (t, 1H, J 11 ′ ’ α ,12 ′ ’ =J 11 ′ ’ α ,12 ′ ’ =3.4 Hz, H12 ′ ’), 4.07 (m, 4H, CO 2 CH 2 , CH 2 P), 3.21 (m, 1H, H-3 ′ ’), 2.75 (brdd, 1H, J 18 ′ ’,19 ′ ’ α =14.0 Hz, J 18 ′ ’,19 ′ ’β =3.6 Hz, H-18 ′ ’), 2.12 (quint, 2H, J H,H = 7.4 Hz, CH 2 ), 1.89–1.77 (m, 4H, H-11 ′ ’ α ,β, H-16 ′ ’ α ,β), 1.71–1.06 (m, 20H, H-1 ′ ’β, H-2 ′ ’ α ,β, H-6 ′ ’ α ,β, H-7 ′ ’ α ,β, H-9 ′ ’, H-15 ′ ’ α ,β, H-19 ′ ’ α ,β, H22 ′ ’ α ,β, H-21 ′ ’ α ,β, 2CH 2 ), 1.09 (s, 3H, Me-27 ′ ’), 0.99 (s, 3H, Me-23 ′ ’), A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 3 0.91–0.87 (m, 1H, H-1 ′ ’ α ), 0.89 (s, 3H Me-30 ′ ’) 0.87, 0.85 (2 s, 3H each, Me-25 ′ ’, Me-29 ′ ’), 0.78, (s, 3H, Me-24 ′ ’), 0.70 (d, 1H, J 5 ′ ’,6 ′ ’β =10.2 Hz, H-5 ′ ’), 0.63 (s, 3H, Me-26 ′ ’) ppm; 13 C NMR (75.5 MHz, CDCl 3 ) δ 177.7 (C-28), 143.7 (C-13 ′ ’), 135.1 (d, 4 J C,P =3.0 Hz, Ar-Cp), 133.8 (d, 2 J C,P = 10.0 Hz, Ar-Co), 130.6 (d, 3 J C,P =12.6 Hz, Ar-Cm), 122.4 (C-12 ′ ’), 118.4 (d, 1 J C,P =85.9 Hz, Ar-Cipso), 79.0 (C-3 ′ ’), 63.1 (CO 2 CH 2 ), 55.3 (C-5 ′ ’), 47.6 (C-9 ′ ’), 46.7 (C-17 ′ ’), 45.8 (C-19 ′ ’), 41.7 (C-14 ′ ’), 41.4 (C-18 ′ ’), 39.3 (C-8 ′ ’), 38.8 (C-4 ′ ’), 38.5 (C-1 ′ ’), 37.1 (C-10 ′ ’), 33.9 (C-21 ′ ’), 33.2 (Me29 ′ ’), 32.8 (C-7 ′ ’), 32.5 (C-22 ′ ’), 30.7 (C-20 ′ ’), 29.5 (d, 2 J C,P =17.1 Hz, CH 2 -CH 2 P), 28.2 (Me-23 ′ ’), 27.7 (C-15 ′ ’), 27.3 (C-2 ′ ’), 25.9 (Me-27 ′ ’), 23.8 (Me-30 ′ ’), 23.5 (C-11 ′ ’), 23.1 (C-16 ′ ’), 22.5 (d, 1 J C,P =50.5 Hz, CH 2 P), 19.6 (d, 2 J C,P =3.3 Hz, CH 2 -CH 2 -CH 2 P), 18.4 (C-6 ′ ’), 17.1 (Me26 ′ ’), 15.7 (Me-24 ′ ’), 15.4 (Me-25 ′ ’) ppm; HRESI-MS m/z calcd. for C 52 H 70 O 3 P ([M–Br - ] + ): 773.5057, found: 773.5051. 1 ′ ’-{5 ′ -[(3 ′ ’β-Hydroxyolean-12 ′ ’–3n-oyl)oxy]pentyl}triphenylphosphonium bromide (3). Bromide II (150 mg, 0.25 mmol), PPh 3 (350 mg, 1.48 mmol) and MeCN (7.5 mL) were used. Column chromatography (80:1 → 5:1 CH 2 Cl 2 –MeOH) afforded 3 as a white solid. Yield: 186 mg (86 %). [ α ] 25 +33 (c 0.87, CH 2 Cl 2 ); 1 H NMR (300 MHz, CDCl 3 ) δ 7.88–7.66 (m, 15H, Ar-H), 5.23 (t, 1H, J 11 ′ ’a,12 =J 11 ′ ’b,12 =3.2 Hz, H12 ′ ’), 3.93 (t, 2H, J H,H =6.4 Hz, CO 2 CH 2 or CH 2 P), 3.86 (m, 2H, CO 2 CH 2 or CH 2 P), 3.20 (dd, 1H, J 2 ′ ’ α ,3 ′ ’ =4.9 Hz, J 2 ′ ’β,3 ′ ’ =10.6 Hz, H-3 ′ ’), 2.79 (dd, 1H, J 18 ′ ’,19 ′ ’ α =13.9 Hz, J 18 ′ ’,19 ′ ’β =4.1 Hz, H-18 ′ ’), 1.95–1.09 (m, 22H, H-1 ′ ’β, H-2 ′ ’ α ,β, H-6 ′ ’ α ,β, H-7 ′ ’ α ,β, H-9 ′ ’, H-15 ′ ’ α ,β, H-19 ′ ’ α ,β, H22 ′ ’ α ,β, H-21 ′ ’ α ,β, 3CH 2 ), 1.09 (s, 3H, Me-27 ′ ’), 1.03–0.92 (m, 1H, H1 ′ ’ α ), 0.97 (s, 3H, Me-23 ′ ’), 0.88, 0.87 (2 s, 3H each, Me-25 ′ ’, Me-29 ′ ’), 0.84 (s, 3H, CH 3 , Me-30 ′ ’), 0.76 (s, 3H, Me-24 ′ ’), 0.70 (d, 1H, J 5 ′ ’,6 ′ ’β = 10.4 Hz, H-5 ′ ’), 0.64 (s, 3H, Me-26 ′ ’) ppm; 13 C NMR (75.5 MHz, CDCl 3 ) δ 177.8 (C-28 ′ ’), 143.9 (C-13 ′ ’), 135.1 (d, 4 J C,P =2.9 Hz, Ar-Cp), 133.9 (d, 2 J C,P =10.0 Hz, Ar-Co), 130.6 (d, 3 J C,P =12.6 Hz, Ar-Cm), 122.4 (C12 ′ ’), 118.5 (d, 1 J C,P =85.8 Hz, Ar-Cipso), 79.1 (C-3 ′ ’), 63.6 (CO 2 CH 2 ), 55.3 (C-5 ′ ’), 47.7 (C-9 ′ ’), 46.7 (C-17 ′ ’), 46.0 (C-19 ′ ’), 41.8 (C-14 ′ ’), 41.4 (C-18 ′ ’), 39.4 (C-8 ′ ’), 38.8 (C-4 ′ ’), 38.5 (C-1 ′ ’), 37.1 (C-10 ′ ’), 34.0 (C21 ′ ’), 33.2 (Me-29 ′ ’), 32.8, 32.6 (C-7 ′ ’, C-22 ′ ’), 30.8 (C-20 ′ ’), 28.2 (Me23 ′ ’), 28.1 (CH 2 ), 27.8 (C-15 ′ ’), 27.3 (C-2 ′ ’), 26.7 (d, 2 J C,P =16.4 Hz, CH 2 -CH 2 P), 25.9 (Me-27 ′ ’), 23.8 (Me-30 ′ ’), 23.5 (C-11 ′ ’), 23.1 (C-16 ′ ’), 22.9 (d, 1 J C,P =49.3 Hz, CH 2 -P), 22.3 (d, 3 J C,P =4.2 Hz, CH 2 -CH 2 -CH 2 P), 18.4 (C-6 ′ ’), 17.2 (Me-26 ′ ’), 15.7 (Me-24 ′ ’), 15.5 (Me-25 ′ ’) ppm; HRESIMS m/z calcd. for C 53 H 72 O 3 P ([M−Br - ] + ): 787.5214, found: 787.5210. 1 ′ ’-{6 ′ -[(3 ′ ’β-Hydroxyolean-12 ′ ’–3n-oyl)oxy]hexyl}triphenylphosphonium bromide (4). Bromide III (100 mg, 0.16 mmol), PPh 3 (212 mg, 0.81 mmol) and MeCN (7.5 mL) were used. Column chromatography (80:1 → 10:1 CH 2 Cl 2 –MeOH) afforded 4 as a white solid. Yield: 120 mg (85 %). [ α ] 25 +28 (c 0.81, CH 2 Cl 2 ); 1 H NMR (300 MHz, CDCl 3 ) δ 7.91–7.66 (m, 15H, Ar-H), 5.23 (t, 1H, J 11a,12 =J 11b,12 =3.5 Hz, H-12), 3.93 (t, 2H, J H,H =6.7 Hz, CO 2 CH 2 ´ o CH 2 P), 4.02–3.89 (m, 2H, CO 2 CH 2 ´ o CH 2 P), 3.20 (dd, 1H, J 2 α ,3 =4.8 Hz, J 2β,3 =10.4 Hz, H-3), 2.82 (dd, 1H, J 18,19 α =14.1 Hz, J 18,19β =4.5 Hz, H-18), 1.93–1.21 (m, 26H), 1.17–0.99 (m, 24H, H-1 ′ ’β, H-2 ′ ’ α ,β, H-6 ′ ’ α ,β, H-7 ′ ’ α ,β, H-9 ′ ’, H-15 ′ ’ α ,β, H-19 ′ ’ α ,β, H-22 ′ ’ α ,β, H-21 ′ ’ α ,β, 4CH 2 ), 1.11 (s, 3H, Me-27 ′ ’), 0.98 (s, 3H, Me-23 ′ ’), 0.97–0.93 (m, 1H, H-1 ′ ’ α ), 0.89, 0.88 (2 s, 3H each, Me-25 ′ ’, Me-29 ′ ’), 0.85 (s, 3H, Me-30 ′ ’), 0.77 (s, 3H, Me-24 ′ ’), 0.71 (d, 1H, J 5 ′ ’,6 ′ ’β =11.6 Hz, H-5 ′ ’), 0.67 (s, 3H, Me-26 ′ ’) ppm; 13 C NMR (75.5 MHz, CDCl 3 ) δ 177.8 (C-28 ′ ’), 143.9 (C-13 ′ ’), 135.1 (d, 4 J C,P =2.9 Hz, Ar-Cp), 133.8 (d, 2 J C,P =10.0 Hz, Ar-Co), 130.6 (d, 3 J C,P =12.5 Hz, Ar-Cm), 122.3 (C-12 ′ ’), 118.5 (d, 1 J C,P =85.8 Hz, Ar-Cipso), 79.0 (C-3 ′ ’), 64.1 (CO 2 CH 2 ), 55.3 (C-5 ′ ’), 47.6 (C-9 ′ ’), 46.7 (C-17 ′ ’), 45.9 (C-19 ′ ’), 41.8 (C14 ′ ’), 41.4 (C-18 ′ ’), 39.4 (C-8 ′ ’), 38.8, 38.5 (C-1 ′ ’, C-4 ′ ’), 37.1 (C-10 ′ ’), 33.9 (C-21 ′ ’), 33.2 (Me-29 ′ ’), 32.8 (C-7 ′ ’), 32.5 (C-22 ′ ’), 30.8 (C-20 ′ ’), 30.1 (d, 2 J C,P =16.0 Hz, CH 2 -CH 2 -P), 28.3 (CH 2 ), 28.2 (Me-23 ′ ’), 27.7 (C-15 ′ ’), 27.2 (C-2 ′ ’), 26.0, 25.9 (Me-27 ′ ’, CH 2 ), 23.8 (Me-30 ′ ’), 23.5 (C11 ′ ’), 23.0 (C-16 ′ ’), 22.8 (d, 1 J C,P =49.7 Hz, CH 2 -P), 22.8 (d, 3 J C,P = 49.7 Hz, CH 2 -CH 2 -CH 2 -P), 18.4 (C-6 ′ ’), 17.1 (Me-26 ′ ’), 15.7 (Me-24 ′ ’), 15.4 (Me-25 ′ ’) ppm; HRESI-MS m/z calcd. for C 54 H 74 O 3 P ([M−Br - ] + ): 801.5370, found: 801.5357. 1 ′ ’-{12 ′ -[(3 ′ ’β-Hydroxyolean-12 ′ ’–3n-oyl)oxy]dodecyl}triphenylphosphonium bromide (5). Bromide IV (160 mg, 0.23 mmol), PPh 3 (298 mg, 1.15 mmol) and MeCN (8 mL) were used. Column chromatography (80:1 → 5:1 CH 2 Cl 2 –MeOH) afforded 5 as a white solid. Yield: 190 mg (85 %). [ α ]25 D +42 (c 0.82, CH 2 Cl 2 ); 1 H NMR (300 MHz, CDCl 3 ) δ 7.85–7.66 (m, 15H, Ar-H), 5.24 (t, 1H, J 11 ′ ’ α ,12 ′ ’ =J 11 ′ ’β,12 ′ ’ =3.3 Hz, H12 ′ ’), 3.96 (m, 2H, CO 2 CH 2 ), 3.71 (m, 2H, CH 2 P), 3.18 (dd, 1H, J 2 ′ ’ α ,3 ′ ’ = 5.1 Hz, J 2 ′ ’β,3 ′ ’ =10.3 Hz, H-3 ′ ’), 2.83 (dd, 1H, J 18 ′ ’,19 ′ ’ α =13.7 Hz, J 18 ′ ’,19 ′ ’β =4.4 Hz, H-18 ′ ’), 1.98–1.81 (m, 6H, H-11 ′ ’ α , H-11 ′ ’β, H-16 ′ ’ α , H-16 ′ ’β, CH 2 ), 1.66–1.34 (m, 19H, H-1 ′ ’β, H-2 ′ ’ α , H-2 ′ ’β, H-6 ′ ’ α , H-7 ′ ’ α , H-9 ′ ’, H-15 ′ ’β, H-19 α , H-21 ′ ’ α , H-22 ′ ’ α , H-22 ′ ’β, 4CH 2 ), 1.30–1.00 (m, 15H, H-6 ′ ’β, H-7 ′ ’β, H-15 ′ ’ α , H-19 ′ ’β, H-21 ′ ’β, 5CH 2 ), 1.11 (s, 3H, Me27 ′ ’), 0.95 (s, 3H, Me-23 ′ ’), 0.89 (s, 3H, Me-30 ′ ’), 0.92–0.84 (m, 1H, H1 α ), 0.86 (s, 6H, Me-25 ′ ’, Me-29 ′ ’), 0.74 (s, 3H, Me-24 ′ ’), 0.68 (s, 3H, Me26 ′ ’), 0.68 (d, J 5 ′ ’,6 ′ ’β =11.2 Hz, H-5 ′ ’) ppm; 13 C NMR (75.5 MHz, CDCl 3 ) δ 177.9 (C=O), 143.9 (C-13 ′ ’), 135.1 (d, 4 J C,P =2.9 Hz, Ar-Cp), 133.8 (d, 2 J C,P =9.9 Hz, Ar-Co), 130.6 (d, 3 J C,P =12.5 Hz, Ar-Cm), 122.4 (C-12 ′ ’), 118.5 (d, 1 J C,P =85.8 Hz, Ar-Cipso), 79.0 (C-3 ′ ’), 64.4 (CO 2 CH 2 ), 55.3 (C-5 ′ ’), 47.7 (C-9 ′ ’), 46.7 (C-17 ′ ’), 46.0 (C-19 ′ ’), 41.8 (C-14 ′ ’), 41.4 (C18 ′ ’), 39.4, 38.9 (C-4 ′ ’, C-8 ′ ’), 37.1 (C-10 ′ ’), 34.0 (C-21 ′ ’), 33.2 (Me-29 ′ ’), 32.8 (C-7 ′ ’), 32.6 (C-22 ′ ’), 30.8 (C-20 ′ ’), 30.6 (d, 2 J C,P =15.7 Hz, CH 2 - CH 2 P), 29.7 (x2) (2CH 2 ), 29.4(x2), 29.3 (3CH 2 ), 28.7 (CH 2 ), 28.2 (Me23 ′ ’), 27.7 (C-15 ′ ’), 27.3 (C-2 ′ ’), 26.2 (CH 2 ), 26.0 (Me-27 ′ ’), 23.7 (Me30 ′ ’), 23.5 (C-11 ′ ’), 23.1 (C-16 ′ ’), 22.9 (d, 1 J C,P =49.5 Hz, CH 2 P), 22.8 (d, 3 J C,P =4.6 Hz, CH 2 -CH 2 -CH 2 P), 18.4 (C-6 ′ ’), 17.1 (Me-26 ′ ’), 15.7 (Me-24 ′ ’), 15.4 (Me-25 ′ ’) ppm; HRESI–MS: calcd. for C 60 H 86 O 3 P ([M–Br - ] + ): 885.6309, found: 885.6293. 2.3. Cell lines, growth and maintenance The human solid tumor cell lines from the different tissues used in this study were obtained from the American Type Culture Collection (ATCC, Manassas, VA). The cell lines were tested for their authentication by PCR profiling using short tandem repeats, which was performed by BaseClear (Leiden, The Netherlands). All cells were grown in RPMI 1640 medium supplemented with heat inactivated 5 % FBS, 2 mM L-glutamine, 100 U/mL of penicillin G and 0.1 mg/mL of streptomycin. Cell cultures were maintained in 60 mm or 100 mm tissue culture-treated dishes (depending on the desired number of cells) placed in a 37 ◦C, 5 % CO 2 and 95 % humidified air incubator. For passaging, cells were detached from the plate, washed twice with PBS without calcium and magnesium and then incubated for 4 min at 37 ◦C with a solution of trypsin/EDTA (170000 U trypsin/L, 200 mg/L EDTA). Fresh medium was added to neutralize the action of trypsin and the cells were counted and seeded in a new dish at the desired concentration to maintain their growth or perform an experiment. 2.4. Antiproliferative assays in 96-well plates An in-house version of the SRB assay of the National Cancer Institute (NCI, USA) was the preferred method to assess the effects of the small molecules under study on cell population growth [20]. Cell seeding density was 2,500–10,000 cells/well depending on the doubling time of the cell line. After 24 h of incubation, serial ten-fold dilutions of compounds in the cell culture medium were added. The maximum test concentration was 100 µM. As a negative control, the cells were treated with the maximal percentage of DMSO in the test wells (0.25 % v/v). After 48 h of exposure to the compounds, the SRB protocol [21] was applied and the optical density (OD) of each well was measured at 530 nm using a microplate reader (PowerWave XS, BioTek, USA). Interferences from the plate materials were reduced by measuring each well in the dual mode at 620 nm. This dual wavelength method was preferred in order to avoid volume differences that would affect the path length of the beam of light. For background correction, wells containing A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 4 cell culture medium (blank) were used. 2.4.1. Antiproliferative activity The percentage growth (PG) was calculated with respect to untreated control cells (C) at each drug concentration based on the difference in OD at the beginning (T 0 ) and the end of drug exposure (T), according to NCI formulas. Thus, 50 % growth inhibition (GI 50 ) represents the drug concentration at which PG is +50. 2.4.2. Efflux chemoresistance The cell line-based assay to predict whether a certain chemotherapeutic agent could develop chemoresistance when behaving as a P-gp substrate requires a wild type cell line (SW1573), and its variant overexpressing P-gp (SW1573/Pgp). The method involves the calculation of the GI 50 values (as described above) for each compound under study, and in the absence or presence of VP, which is known to inhibit P-gp. The cell culture medium containing VP was prepared by adding VP at final concentration of 10 µM. The standard P-gp substrates PTX and VBL were used as the positive controls. In this assay, the resistance factor (Rf) for a certain compound is the ratio of the GI 50 value in the cell line overexpressing P-gp (mutant, mt) and in the wild type (wt). Rf is calculated in the absence and presence of VP. 2.5. Cell colony formation assay A range from 450 to 800 cells were seeded onto a six-well plate, depending on the proliferative profile the cell line. After 24 h, the medium was refreshed and compounds were added at the desired concentration. Cells were incubated for at least seven days until a high number of colonies could be observed microscopically in the control wells, defining a colony as a group of at least 50 cells. The medium was removed and the cells were rinsed with PBS and fixed with methanol for 2 min. Excess of methanol was rinsed with another wash of PBS and the cells were stained using crystal violet (0.5 % w/v, deionized water). Colonies were counted using AutoCellSeg, a MATLAB implementation for automatic segmentation. The OD of each well was determined dissolving the stained colonies in 1 % SDS in PBS and subsequently measuring the absorbance at 590 nm using a BioTek PowerWave XS Absorbance Microplate Reader. 2.6. Mitochondrial membrane potential disruption To evaluate changes in the mitochondrial membrane potential (MMP), TMRE was used as a fluorescent probe in both fluorescence microscopy and microplate reader measurements. For fluorescence microscopy, cells were seeded onto coverslips, treated, and incubated for the indicated times. Then, 30 min before the ending point of the experiment, the medium was replaced with fresh medium containing TMRE (0.5 µM) and incubated until reaching the endpoint at 37 ◦C under dark conditions. Then, cells were rinsed twice with PBS to avoid cell dislodgement. Given that the use of fixative agents is not compatible with TMRE because of the reduction of the TMRE signal, coverslips were mounted directly onto microscopy slides over a drop of glycerol solution (10 % in PBS). At least 6 random field fluorescence images were taken using a Leica DM 4000B® with a Qwin software (Leica Microsystems) equipped with N2.1 excitation filter (bandpass [BP] filter =515–560 nm) at 40X magnification. For microplate reader measurements, cells were seeded in 96 well dark-wall clear-bottom plates at a density of 10,000 cells/well. Treatment and incubation with TMRE dye followed as previously described for fluorescence microscopy assay. Then, cells were rinsed very carefully with PBS, 100 µL of PBS was added, and the fluorescence was measured using a Varioskan Lux Multimode plate reader (Thermo Fisher Scientific, USA) using 549/575 nm excitation/emission wavelengths. Fluorescence units were normalized to the cell number in every well using crystal violet staining. Briefly, the cells were fixated with methanol for 1 min, rinsed with PBS and stained with crystal violet (0.5 % w/v, deionized water). Dye was resuspended using SDS (1 % in PBS) and measured at 590 nm using BioTek’s PowerWave XS Absorbance Microplate Reader. The TMRE standard curve was obtained by measuring different concentrations of the fluorescent dye in RPMI medium without phenol red. Every concentration was added in triplicates to Sterilin™ Microtiter™ 96 well black bottom microplates in a total volume of 100 µL/well. The fluorescence intensity was measured by reading from the top, as previously described. 2.7. Reactive oxygen species detection Changes in the formation of reactive oxygen species (ROS) were identified using DCFDA. Cells were seeded onto 96 well dark-wall clearbottom plates as described previously. Prior to the exposure to the selected compounds, the medium was removed and the cells were incubated with 40 µM of DCFDA in RPMI without phenol red for 30 min. Then, the compounds were added to the medium containing DCFDA considering the final concentration desired. Fluorescence intensity was measured using a Varioskan Lux Multimode plate reader at 492/520 nm excitation/emission wavelengths. 2.8. DNA staining Prior to cell seeding, sterile coverslips were placed onto a 6-well plates. Then, cells were plated on top of the coverslip at a density of 100,000 cells/well. After overnight incubation, the compounds were added to the appropriate concentration. At the desired experimental time stop, cells were fixed using PFA solution (1 mL, 4 % in PBS) for 10 min at room temperature. PFA solution was removed and cells were washed with NH 4 Cl solution (50 mM in PBS) for 15 min (to reduce PFA background) and rinsed with PBS afterwards. A Triton solution (0.1 % in PBS) was used to permeabilize the cells. After 10 min of incubation, the cells were rinsed 2–3 times with PBS for 5 min each. Staining was performed using a DAPI solution (750 µL, 1 µg/mL in methanol) for 10 min in the absence of light and moderate oscillation. The DAPI solution was recovered and cells were rinsed again 2–3 times maintaining darkness conditions. Coverslips were mounted onto a microscopy slide over a drop of Mowiol mounting medium and sealed using nail polish to avoid evaporation. At least 6 random fields were imaged using 20X and 40X magnifiers in Leica DM 4000 B® microscope (Leica Microsystems) with the A excitation filter (BP =340–380). Quantification of the mean fluorescence of DAPI-stained cells was performed using FIJI/ImageJ software (NIH, USA). 2.9. Continuous live-cell imaging Drug-induced phenotypic changes related to the diverse MoA of compounds subjected to study were observed in a continuous manner using a live cell microscopy CX-A label-free cell imaging system, (Nanolive S.A, Switzerland). Cells were seeded on 35 mm high glassbottom µ-dish (IBIDI, Gr¨ afelting, Munich, Germany) at a concentration of 80,000–100,000 cells per mL, depending on the cell type. After 24 h, the medium was replaced with RPMI 1640 without phenol red since the pH indicator would interfere with the refractive index measurements. Compounds were added to cells at the desired concentrations and experimental settings were configured depending on the type of experiment (i.e. number of cycles and their duration). Areas where the number of cells was representative were selected for image acquisition. Data were transferred to FIJI (NIH, USA) for visual inspection. After the images were taken for the desired period, EVE software (Nanolive S.A Switzerland) was used for cell segmentation and analysis of the refractive index (RI) based parameters. These 11 phenotypic parameters are divided into two groups: a) Morphology related parameters: cell area (%), cell area ( μ m 2 ), cell eccentricity, cell perimeter, cell A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 5 extent, cell compactness, cell form factor and cell granularity; and b) Content related parameters: mean refractive index, dry mass (pg) and average dry mass density (pg/ μ m 3 ). Live Cell Death Assay (LCDA) software (Nanolive S.A., Switzerland) was used to integrate refractive index parameters to obtain apoptosis and necrosis kinetics. 2.10. Fluorescence labelling in live cell imaging Hoechst 33,342 was used to stain the nuclear compartments over time. Briefly, cells were seeded as described for the refractive index live cell imaging. Assay concentration of Hoechst was prepared from a 10 mg/mL stock solution (1:2000 dilution) and incubated under dark conditions for 10 min. The medium was removed and the cells were washed once with PBS. The compounds under study were added to RPMI without phenol red at the desired concentration. Refractive index acquisition was performed as previously described. For fluorescence imaging, the exposure time and signal amplification were selected empirically for each experiment and maintained at the same time for all the samples measured. Fluorescence images were obtained using a DAPI-UV filter every 10, 20 or 30 cycles, depending on the experiment. Fluorescence and RI images were mounted and analysed using STEVE software (Nanolive S.A., Switzerland). 2.11. Animals and housing All experiments followed the PI87/20 research procedures approved by the ethics committee for animal experiments of the University of Zaragoza. Animal care and use was carried out in accordance with the Spanish policy for animal protection RD53/2013 and the European Union directive 2010/63 for the protection of animals used for experimental and other scientific purposes. For the in vivo toxicology study, 14 male RjOrl:SWISS were used. Eighteen adult female Rj:ATHYMFoxn1 nu/nu mice (15 weeks) were used in the in vivo oncological efficacy studies. All animals were purchased from Janvier Laboratories. Up to a maximum of 5 animals were kept in 30 x 20 x 15 cm boxes, with access to osmotized and autoclaved water and irradiated food ad libitum (2914 Teklad Global 14 % protein rodent maintenance diet). The room temperature was maintained at 23 ±1 ◦C with a 12 h light cycle (starting at 8:00 a.m.). 2.12. In vivo toxicology assay Before testing compound 3 in an in vivo efficacy study, we evaluated the in vivo toxicity of the most promising compound in vitro. For such a purpose, an acute oral (PO) and intraperitoneal (IP) toxicity test was performed based on the OECD Test No. 425: Acute Oral Toxicity: Upand-Down Procedure. The highest dose tested was 17.5 mg/kg and the starting dose was 0.55 mg/kg, applying a slope of 2. The test was performed in male RjOrl: SWISS mice of 11–14 weeks with dosing intervals of 48 h. Then, toxicity was evaluated after a single administration. Drugs were administered oral (PO) to conscious mice with a standard volume of 10 mL kg −1 body weight. as a suspension in physiological saline (5 % DMSO). Animals were monitored individually at least once during the first 30 min after injection, periodically during the first 24 h (with special attention to the first 4 h) and daily thereafter for a total of 14 days. Health status of the animals was monitored using a modified scale based on the proposal by Morton and Griffiths. General appearance (0–3), weight (0–2), spontaneous behavior (0–3), and provoked behavior (0–3) were scored. 2.13. In vivo oncological efficacy assay 4T1 mouse mammary tumor cells expressing luciferase was implanted subcutaneously in mice and visualized using optical imaging (IVIS LUMINA 100) to monitor tumor growth non-invasively in a longitudinal study. The study was carried out in 2 experimental groups: control and treated (compound 3). Each experimental group consisted of 9 female mice of the Rj:ATHYM-Foxn1 nu/nu strain (Janvier Labs) with an age of 15 weeks and a mean weight of 38.36 g (±2.46 g). On day 0, all animals (18) were implanted by subcutaneous administration in the lumbar region with 1x10 6 tumor cells in a volume of 200 µL of phosphate buffered saline (PBS). On day 2, after optical imaging follow-up, the experimental groups were established. At that time, the mean radiance of each experimental group was set between 2.83 x 10 7 and 3.15 x 10 7 photon/sec/cm 2 /sr. The start of treatment was on day 2 once the groups had been established. The trial had a total duration of 22 days from the day of inoculation of the tumor cells. Optical imaging and caliper follow-up, weight change and treatment were performed on days 2, 6, 8, 10, 13, 15, 17 and 20, with a dose of 0.55 mg/Kg (3 times/week). Treatment was performed orally with a sterile metal feeding tube 22ga x 25 mm (FTSS-22S-25, Instech) at the end of optical imaging follow-up at a concentration of 1 mg/kg. Drugs were administered orally (PO) to conscious mice with a standard volume of 10 mL kg −1 body weight as a suspension in physiological saline (5 % DMSO). Five minutes before optical imaging acquisition, a volume of 10 mL kg −1 body weight of a luciferin solution (15 mg/mL) was administered intraperitoneally to each mouse. Through the caliper measurements, the mean tumor volume was calculated with the following formula (Length*Width 2 /2). 3. Results 3.1. Synthesis and antiproliferative effects of OA-TPP hybrids In the design of OA-TPP hybrids, we envisioned the linkage between OA and the TPP fragment using a flexible and lipophilic hydrocarbon chain. To connect both key structural motifs we used linear alky chains containing 4, 5, 6 or 12 methylene groups. The general formula of these hybrid molecules is OA-(CH 2 ) n -TPP. OA-TPP hybrids were accessed in a two-step synthetic pathway (Table 1): 1) chemoselective alkylation of the carboxylic moiety of OA with an excess of α , ω -dibromoalkanes with different lengths under basic conditions (K 2 CO 3 ), 2) followed by nucleophilic displacement of the terminal bromine atom with PPh 3 in a Fisher-Porter tube. As a first step, the OA-TPP hybrids 2–5 as well as pure OA (1), were evaluated as antiproliferative agents against a panel of human solid tumor cell lines (Table 1). Although the structural variation of the compounds was very subtle –presenting only differences in the number of methylene units that conform to the linker– a preliminary structure–activity relationship was established. The length of the linker significantly affected the potency of the compounds. Minor differences were observed in the antiproliferative effects of the C 4 , C 5 and C 6 compounds (2, 3 and 4, respectively), which exhibited potencies within the nanomolar range. However, mitocan 5 (bearing a C 12 linker) showed a clear loss of potency compared to the other hybrids. Hence, compounds with short chains (n =4–6) showed GI 50 values in the nanomolar range, whereas long alkyl chains (n =12) produced a significant decrease in potency, with GI 50 values at the micromolar level. This is consistent with the reported predominant antiproliferative effects of pure OA on cancer cells [22]. The set of OA-TPP hybrids was also evaluated in the non-tumor immortalized cell line BJ-hTERT (Table 1). Interestingly, none of the compounds exerted antiproliferative effects on this cell line (GI 50 >10 5 nM), showing remarkable selectivity against cancer cell lines. The absence of an effect on cell growth in non-tumor human fibroblasts indicated promising selectivity towards healthy cells. This is a relevant result because one of the main drawbacks of the use of mitocans in therapy is their intrinsic toxicity [9]. The absence of antiproliferative activity in a healthy cell line agreed with other reports in the literature regarding the absence of toxic effects when using OA or TPP separately [10,22]. In the case of the OA-TPP hybrids presented here, the selectivity A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 6 between tumor and non-tumor cell lines was even larger. Consistent with the antiproliferative data, compound 5 was discarded at this point. 3.2. OA-TPP hybrids are not P-gp substrates Next, we checked whether the OA-TPP hybrids 2–4 could be substrates of the efflux pump P-gp, a known mechanism of drug resistance in cancer cells. The assay comprises a wild type cell line (SW1573) and a mutant variant that overexpresses P-gp (SW1573/Pgp), and PTX functions as positive control. Both cell lines were exposed to the compounds under evaluation. The GI 50 values were computed after 48 h of exposure in the absence or presence of 10 µM verapamil (VP) in cell culture (Table 2). In this assay, the resistance factor (Rf) for any given compound is defined as the ratio of its GI 50 in the cell line overexpressing Pgp (mutant, mt) and the wild type (wt). Rf was calculated in the absence and presence of VP. From the results obtained in the absence of VP, we can infer from the Rf values that the mitocans 2–4 were 10–15 times less potent against the mutant cells. To our surprise, the co-treatment with VP sensitized both the wild type and the resistant cancer cell lines. This is an unusual result for this type of assay. VP is used in this assay because it is a P-gp blocker, which avoids drug efflux and consequently enhances the activity. However, this result was not consistent with the increased potency observed for the OA-TPP hybrids. It has been reported that TPP, benzodiazepines, as well as several Ca 2+ antagonists –like VP– inhibit mitochondrial Na + /Ca 2+ exchange [23]. Thus, we speculated that the effects observed in Table 2 for the co-treatment with VP were just a result of the synergistic interaction between the OA-TPP hybrids and VP. Furthermore, these results agree with the premise that compounds 2–4 are mitocans. 3.3. Effects on reproductive viability Given the potent effect observed during the antiproliferative screening, we decided to evaluate the ability of OA-TPP hybrids 2–4 to interfere with the survival and growth ability of cells after long-term exposure. To achieve this goal, clonogenic assays were performed by incubating HeLa cells with the compounds for 8 days. Considering this extended exposure time, only the GI 50 and GI 50 /3 concentrations were assayed (Table 1). All three compounds were able to produce a decrease in the number of cell colonies formed when compared to untreated cells, and in a concentration dependent manner. In addition, the compounds were unable to reduce colony formation when the cells were exposed to a lower dose. Interestingly, the first differences between the compounds were observed in these assays. HeLa colonies were statistically significant less in the treatment with GI 50 dose for compounds 3 and 4 but not for 2. Since the clonogenic assay considers a colony as a group of 50 cells, to uniquely evaluate the number of colonies, relevant differences could be concealed. The images taken depict differences in the size of the colonies between the control and the treatment groups (Fig. 2). The differences after crystal violet staining measurement between the untreated and treated samples were more evident. Through this process, all three compounds significantly reduced the colony staining at the higher dose, and derivatives 3 and 4 were able to do so at the lower dose. These results demonstrated that the OA-TPP hybrids were able to hinder cell growth even when a single administration was maintained for 8 days. Simultaneously, exposure to the compounds was able to reduce not only the number of colonies but also the number of the cells conforming them. Table 1 Synthesis and antiproliferative activity (GI 50 , nM) of OA and OA-TPP hybrids against human solid tumors cell lines and non-tumor cell line BJ-hTERT. Cell line (origin) 1(H) 2 (n =4) 3 (n =5) 4 (n =6) 5 (n =12) Ara-C A549 (lung) 7167 ±1136 17 ±1.5 21 ±2.4 34 ±13 6418 ±2759 75 ±17 HBL-100 (breast) 23803 ±7173 31 ±13 31 ±7.4 57 ±11 6845 ±2910 642 ±100 HeLa (cervix) 13827 ±9121 45 ±8.1 36 ±1.7 41 ±8.5 5558 ±1952 59 ±7.8 T-47D (breast) 35847 ±7209 32 ±7.6 26 ±1.9 47 ±7.5 >10 5 9437 ±3705 WiDr (colon) 31684 ±8087 18 ±8.9 13 ±1.5 38 ±9.3 >10 5 20296 ±6973 BJ-hTERT (fibroblasts) >10 5 >10 5 >10 5 >10 5 >10 5  Values represent the mean ±standard deviation of two to five independent experiments. Ara-C was used as the positive control. Table 2 Antiproliferative activity (GI 50 , nM) of OA-TPP hybrids and PTX against SW1573 and SW1573/Pgp cell lines being co-incubated with or without VP. w/o Verapamil w Verapamil Compound SW1573 SW1573/ Pgp Rf SW1573 SW1573/ Pgp Rf 224 ±3.6 240 ±27 10 0.12 ± 0.02 14 ±2.4 117 315 ±2.2 230 ±20 15 0.88 ± 0.20 18 ±4.8 21 421 ±4.9 220 ±26 10 0.46 ± 0.04 18 ±1.9 38 PTX 0.053 ± 0.022 300 ±110 564 0.46 ± 0.21 0.31 ± 0.15 0.7 Values represent the mean ±standard deviation of at least three independent experiments. A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 7 3.4. OA-TPP hybrids produce mitochondrial membrane depolarization and reactive oxygen species production The mitochondrial membrane potential and reactive oxygen species (ROS) of HeLa cells after treatment with the OA-TPP derivatives was evaluated through diverse commonly used methods. For comparison purposes, carbonyl cyanide 3-chlorophenylhydrazone (CCCP) and hydrogen peroxide were used as positive controls, respectively. CCCP acts as a protonophore that dissipates the proton gradient necessary for maintaining the mitochondrial membrane potential (MMP) [24], thus producing a depolarization effect. To monitor changes in the MMP, we first used tetramethylrhodamine ethyl ester (TMRE). This lipophilic cationic dye acts as a fluorescent probe for mitochondrial polarization, accumulates in healthy polarized mitochondria but releases with membrane depolarization. The reduction in the fluorescence signal pointed out MMP depolarization, with TMRE being released into the cytosol and subsequently washed out (Fig. 3A–B). Pure OA (1) was included in the assay to show the relevance of the TPP ion fragment in the MoA of the compounds under study. Although a very modest effect was observed, there was no significant reduction in relative fluorescence Fig. 2. Effects on reproductive viability: (A) Representative images of the colony formation assay. HeLa cells were exposed to derivatives 2–4 at two concentrations (GI 50 and GI 50 /3) for 8 days and stained with crystal violet; (B) Relative number of colonies compared to control; (C) Absorbance of crystal violet staining resuspended in PBS with 1 % SDS relative to the control. *p <0.05; **p <0.005. Fig. 3. Effect of OA-TPP hybrids on mitochondria: (A) Representative fluorescence images of untreated HeLa cells or treated with 3 at 5X GI 50 (180 nM) and incubated for 30 min with TMRE. Scale bar: 10 µm; (B) Fluorescence intensity of HeLa cells stained with TMRE when exposed for 30 min to CCCP and 1–4 at 5X their GI 50 (70.000, 225, 180 and 200 nM, respectively). Error bars: SD of at least three independent experiments. *p <0.05; **p <0.01; ***p <0.001; (C) Effect of OA-TPP hybrids on ROS induction. Fluorescence signal of DCF expressed as RFU relative to that of the control. Hydrogen peroxide served as positive control. Error bars represent standard deviation of at least three independent experiments. *p <0.05; **p <0.01. A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 8 units (RFU) by 1 when compared to the untreated samples after 18 h of exposure. However, a significant decrease in the fluorescent signal was observed after treatment with the three selected compounds, and in the (decreasing) order 3 >4 >2. In addition, hybrid 3 was able to produce a significant reduction in MMP, which was comparable to that of the positive control CCCP. This result was confirmed using fluorescence microscopy. As depicted in Fig. 3A–B, untreated cells incubated for 30 min with TMRE showed broad fluorescence staining, while cells treated with compound 3 leaked the dye into the extracellular medium, resulting in a loss of signal. These results suggest that there is an effect of OA-TPP hybrids on mitochondrial activity, and the TPP ion motif in the molecule plays a key role in promoting this effect. In addition, the length of the linker influences this specific activity given the differences observed between the three hybrids, highlighting the five-carbon linker (3) as the one that exerted a higher reduction in MMP. The next step was the evaluation of reactive oxygen species (ROS) formation because of exposure to the compounds. Unbalanced ROS equilibrium is a clear sign of mitochondrial dysfunction in an oxidative stress environment, preceding apoptotic cell death through the mitochondrial pathway [25]. Considering that the compounds produced a reduction in the MMP, variations in the formation of ROS at short times of exposure should happen. Therefore, the induction of ROS after 3 h was measured in HeLa cells using the widely known DCFDA assay. Hydrogen peroxide served as positive control. Alike the results obtained with the TMRE assay, the highest dichlorofluorescein (DCF) signal was observed after exposure to compound 3, while OA-TPP hybrids 2 and 4 showed weaker fluorescence signals (Fig. 3C). Hence, positive DCF fluorescence –independently of whether the increased signal depends on total ROS accumulation or cytochrome c release– suggests perturbation of the mitochondrial function by the addition of the OA-TPP derivatives in the same direction as the reduction in MMP, as both processes are indicative of damage induced to this organelle. Interestingly, coadministration with ROS inhibitor ascorbic acid alongside with 3 did not modify the antiproliferative effects of 3 on HeLa cells. 3.5. Exploring the mode of cell death One of the hallmarks of cell death caused by drugs is nuclear shape modification. DAPI staining of cell exposed to OA-TPP hybrids 2–4 for 24 h allowed observing reduction in the number of stained entities –as expected by their antiproliferative effects–, nuclear flattening and the appearance of apoptotic features, such as nuclear size decrease (Fig. 4A). Although some features of DNA fragmentation were observed, especially in the treatment with 2, there was no significant number of cells manifesting this effect. While DAPI staining helped visualizing nuclear shape modifications, live cell imaging of Hoechst pre-stained cells allowed to follow in time the features of cell death (Fig. 4B, Supplementary Videos S1a-c). Cell shrinkage, membrane blebbing and overall cellular compaction –typical apoptotic nuclear phenotypes– were observed after eight hours of exposure to mitocans 2–4. Kinetics of the status of the cell culture showed how the three compounds induced cell death, mainly through apoptosis. Based on image analysis, OA-TPP induced an apoptotic mode of death with a late transition to necrotic phenotypes (slight membrane swelling) in some of the cells conforming the population under study. Moreover, all the compounds induced massive cytoplasmic vacuolation. Interestingly, time-lapse observation of the cell fate showed how cells that were more vacuolated needed more time to show apoptotic phenotypes and die later than those that presented less vacuoles in total. Cytoplasmic vacuoles were easily observed in the space surrounding the nucleus. 3.6. Studies on vacuole formation As mitochondrial function should be altered by MMP reduction, it could be feasible that compensatory mechanisms would be triggered to maintain the metabolic requirements of cells [26]. To explain these effects produced by the OA-TPP hybrids, multiple hypotheses were formulated. Since all three derivatives (2–4) were able to induce vacuole formation regardless of the size of the linker between the two moieties, we speculated if the inclusion of the TPP was determinant for this Fig. 4. Exploring the mode of cell death: (A) Representative images of DAPI staining in HeLa cells after exposure to OA-TPP hybrids for 24 h. Yellow arrows: Representative condensed nuclei. Red arrows: DNA fragmentation. 40x magnification. Scale bar: 10 µm; (B) Representative snapshots of HeLa cells treated with 200 nM of OA-TPP hybrids. Nuclear staining was performed using Hoechst 33,342 and tracked over time using continuous live-cell imaging. Scale bar: 20 µm. A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 9 cyclooxygenase-2 cDNA, and their combination on the growth of human bladder cancer cells, Biomed. Pharmacother. 63 (2009) 241–248, https://doi.org/10.1016/ j.biopha.2008.04.007. [30] R. Schoeman, N. Beukes, C. Frost, Cannabinoid combination induces cytoplasmic vacuolation in MCF-7 breast cancer cells, Molecules 25 (2020) 4682, https://doi. org/10.3390/molecules25204682. [31] T. Qi, B. Chen, Z. Wang, H. Du, D. Liu, Q. Yin, B. Liu, Q. Zhang, Y. Wang, A pHactivatable nanoparticle for dual-stage precisely mitochondria-targeted photodynamic anticancer therapy, Biomaterials 213 (2019) 119219, https://doi. org/10.1016/j.biomaterials.2019.05.030. [32] J. Meng, Z. Jin, P. Zhao, B. Zhao, M. Fan, Q. He, A multistage assembly/ disassembly strategy for tumor-targeted CO delivery, Sci. Adv. 6 (2020) eaba1362, https://doi.org/10.1126/sciadv.aba1362. [33] D. Lee, I.Y. Kim, S. Saha, K.S. Choi, Paraptosis in the anti-cancer arsenal of natural products, Pharmacol. Ther. 162 (2016) 120–133, https://doi.org/10.1016/j. pharmthera.2016.01.003. [34] W.-B. Wang, L.-X. Feng, Q.-X. Yue, W.-Y. Wu, S.-H. Guan, B.-H. Jiang, M. Yang, X. Liu, D.-A. Guo, Paraptosis accompanied by autophagy and apoptosis was induced by celastrol, a natural compound with influence on proteasome, ER stress and Hsp90, J. Cell. Physiol. 227 (2012) 2196–2206, https://doi.org/10.1002/ jcp.22956. [35] F. Fontana, M. Raimondi, M. Marzagalli, A. Di Domizio, P. Limonta, The emerging role of paraptosis in tumor cell biology: perspectives for cancer prevention and therapy with natural compounds, Biochim. Biophys. Acta 1873 (2020) 188338, https://doi.org/10.1016/j.bbcan.2020.188338. [36] L. Ma, X. Xuan, M. Fan, Y. Zhang, G. Yuan, G. Huang, Z. Liu, A novel 8-hydroxyquinoline derivative induces breast cancer cell death through paraptosis and apoptosis, Apoptosis 27 (2022) 577–589, https://doi.org/10.1007/s10495-02201737-w. [37] S. Sperandio, K.S. Poksay, B. Schilling, D. Crippen, B.W. Gibson, D.E. Bredesen, Identification of new modulators and protein alterations in non-apoptotic programmed cell death, J. Cell. Biochem. 111 (2010) 1401–1412, https://doi.org/ 10.1002/jcb.22870. [38] S. Sperandio, K. Poksay, I. de Belle, M.J. Lafuente, B. Liu, J. Nasir, D.E. Bredesen, Paraptosis: mediation by MAP kinases and inhibition by AIP-1/Alix, Cell Death Differ. 11 (2004) 1066–1075, https://doi.org/10.1038/sj.cdd.4401465. [39] X. Chen, X. Chen, X. Zhang, L. Wang, P. Cao, V. Rajamanickam, C. Wu, H. Zhou, Y. Cai, G. Liang, Y. Wang, Curcuminoid B63 induces ROS-mediated paraptosis-like cell death by targeting TrxR1 in gastric cells, Redox Biol. 21 (2019) 101061, https://doi.org/10.1016/j.redox.2018.11.019. [40] A. Boice, L. Bouchier-Hayes, Targeting apoptotic caspases in cancer, Biochim. Biophys. Acta 1867 (2020) 118688, https://doi.org/10.1016/j. bbamcr.2020.118688. [41] Q. Cui, J.-h. Yu, J.-n. Wu, S.-i. Tashiro, S. Onodera, M. Minami, T. Ikejima, P53mediated cell cycle arrest and apoptosis through a caspase-3independent, but caspase-9-dependent pathway in Oridonin-treated MCF-7 human breast cancer cells, Acta Pharmacol. Sin. 28 (2007) 1057–1066, https://doi.org/10.1111/j.17457254.2007.00588.x. [42] Q. Sun, T. Chen, X. Wang, X. Wei, Taxol induces paraptosis independent of both protein synthesis and MAPK pathway, J. Cell. Physiol. 222 (2010) 421–432, https://doi.org/10.1002/jcp.21982. [43] B.M. Ram, G. Ramakrishna, Endoplasmic reticulum vacuolation and unfolded protein response leading to paraptosis like cell death in Cyclosporine A treated cancer cervix cells is mediated by Cyclophilin B inhibition, Biochim. Biophys. Acta 1843 (2014) 2497–2512, https://doi.org/10.1016/j.bbamcr.2014.06.020. [44] W.A. Maltese, J.H. Overmeyer, Methuosis: nonapoptotic cell death associated with vacuolization of macropinosome and endosome compartments, Am. J. Pathol. 184 (2014) 1630–1642, https://doi.org/10.1016/j.ajpath.2014.02.028. [45] P. Weerasinghe, L.M. Buja, Oncosis: an important non-apoptotic mode of cell death, Exp. Mol. Pathol. 93 (2012) 302–308, https://doi.org/10.1016/j. yexmp.2012.09.018. [46] G.N. Li, X.J. Zhao, Z. Wang, M.S. Luo, S.N. Shi, D.M. Yan, H.Y. Li, J.H. Liu, Y. Yang, J.H. Tan, et al., Elaiophylin triggers paraptosis and preferentially kills ovarian cancer drug-resistant cells by inducing MAPK hyperactivation, Signal Transduct. Target. Ther. 7 (2022) 317, https://doi.org/10.1038/s41392-022-01131-7. [47] H. Montecino-Garrido, D. M´ endez, R. Araya-Maturana, J.P. Millas-Vargas, S. Wehinger, E. Fuentes, Triphenylphosphonium-based compounds (Honokiol, Lonidamine, and Atovaquone) on the platelet function and cytotoxic activity, Front. Pharmacol. 13 (2022) 893873, https://doi.org/10.3389/ fphar.2022.893873. [48] H.C. Ong, J.T.S. Coimbra, G. Kwek, M.J. Ramos, B. Xing, P.A. Fernandes, F. García, Alkyl vs. aryl modifications: a comparative study on modular modifications of triphenylphosphonium mitochondrial vectors, RSC Chem. Biol. 2 (2021) 1643–1650, https://doi.org/10.1039/D1CB00099C. A. Puerta et al. Biochemical Pharmacology 234 (2025) 116807 16