Archives of Biochemistry and Biophysics 454 (2006) 114–122 www.elsevier.com/locate/yabbi
Avicins, natural anticancer saponins, permeabilize mitochondrial membranes 夽 Victor V. Lemeshko a,¤, Valsala Haridas b, Jairo C. Quijano Pérez a, Jordan U. Gutterman b,¤ a
b
Escuela de Física, Facultad de Ciencias, Universidad Nacional de Colombia, sede Medellín, AA3840 Medellín, Colombia Department of Molecular Therapeutics, The University of Texas M.D. Anderson Cancer Center, Houston, TX 77030, USA Received 8 July 2006, and in revised form 5 August 2006 Available online 23 August 2006
Abstract Avicins are a class of natural saponins with selective pro-apoptotic activity in cancer cells. In this work, we studied the inXuence of avicins on metabolic state of rat liver mitochondria. Avicin-treated mitochondria underwent a signiWcant decrease in oxygen consumption rate that was completely restored by addition of exogenous cytochrome c. On the other hand, avicins increased the rotenone-insensitive oxidation of external NADH in the presence of exogenous cytochrome c, long before high amplitude swelling of mitochondria was observed. The increase in external NADH oxidation was cyclosporin A-insensitive. Avicin G signiWcantly accelerated hydroperoxideinduced oxidation of mitochondrial endogenous NAD(P)H, the drop of the inner membrane potential and the high amplitude swelling of mitochondria. The obtained data might explain selective induction of apoptosis in tumor cells by avicins. Based on other studies showing that tumor cells generate hydroperoxides with a very high rate, avicins could provide a new strategy of anticancer therapy by sensitizing cells with high levels of reactive oxygen species to apoptosis. © 2006 Elsevier Inc. All rights reserved. Keywords: Mitochondria; Outer mitochondrial membrane; NADPH; Cytochrome c; Avicin; Hydroperoxides; Cancer; Apoptosis
Avicins represent a new class of plant stress metabolites that exhibit selective pro-apoptotic [1–4] and cytotoxic activity [5] in tumor cells, as well as anti-inXammatory [6,7] and antioxidant properties [8–10]. Our previous Wnding that avicins induce apoptosis in Jurkat cells by a direct perturbation of mitochondria [1], prompted us to study their eVects on the oxidative phosphorylation system of rat liver mitochondria. Most common eVects of anti-tumor drugs, by which they target mitochondrial structure and functions, are revealed in direct permeabilization of the outer mitochondrial mem-
夽 Financial support for this work was provided by the: Clayton Foundation for Research; Biomedical Research Foundation; Abraham J. and Phyllis Katz Foundation; Colciencias (Colombia) Grant #2213-05-16851. * Corresponding authors. E-mail addresses:
[email protected] (V.V. Lemeshko), jgutterm @mdanderson.org (J.U. Gutterman).
0003-9861/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2006.08.008
brane (OMM1) to cytochrome c, or in oxidative stress and mitochondrial swelling. The rupture or direct permeabilization of the OMM results in release of cytochrome c and other pro-apoptotic factors from mitochondria, as well as in disturbance of oxidative phosphorylation system [11,12]. Allosteric interactions of various factors with the voltagedependent anion channel that increase the probability of its closure under some physiological and pathological conditions, have been also considered as a possible way to induce tumor cells death [13,14]. Thus, mitochondria play a central role in cancer survival and are one of the main targets for developing anticancer drugs [11–15].
1 Abbreviations used: OMM, outer membrane of mitochondria; IMM, inner membrane of mitochondria; ROS, reactive oxygen species; TMPD,N,N,N⬘,N⬘-tetramethyl-p-phenylenediamine; FCCP, carbonyl-cyanide-p-triXuoromethoxy phenylhydrazone; DNP, 2,4-dinitrophenol; tBH, tert-butylhydroperoxide.
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
115
In this work, we demonstrate that anticancer drugs avicins induce the OMM permeabilization, which leads to release of cytochrome c and inhibition of respiratory chain in rat liver mitochondria. Another eVect of avicins is the creation of hypersensitivity of mitochondria to hydroperoxides. It was revealed in a faster exhaustion of endogenous NAD(P)H, accelerated swelling of mitochondria, and in a faster drop of the inner membrane potential induced by tert-butylhydroperoxide (tBH). Mitochondrial NADPHdependent metabolism of hydroperoxides [16,17] is coupled to the inner membrane potential through the energy-dependent trans-hydrogenase. Thus, the decreased generation of the inner membrane proton motive force, as result of cytochrome c release, should result in accumulation of hydroperoxides in the cells up to cytotoxic levels. The presence of high levels of reactive oxygen species (ROS) in most cancer cells [18–22] might sensitize tumor cells to the cytotoxic eVects of avicins. Avicins therefore, could have great potential as an anticancer drug, either by itself or in combination with other drugs, based on the synergistic eVect of avicins and hydroperoxides.
phosphate–Tris, pH 7.2, 400 M ADP, 2.5 M rotenone, 0.5 M antimycin A, 0.5 M myxothiazol, 0.5 M FCCP, 100 M DNP, 1 g/ml oligomycin, 20 M cytochrome c, and 0.5 mM TMPD were added to the medium wherever indicated.
Materials and methods
Monitoring of mitochondrial swelling
Materials
The swelling of rat liver mitochondria was determined by monitoring the apparent light absorbance at 640 nm using the SP-850 spectrophotometer (USA), additionally equipped with the magnetic stirrer and thermostatic chamber. The absorbance curve was recorded with the Linseis recorder (USA), connected to the spectrophotometer. Avicins (5 g/ml) were added to the SKH medium, supplemented with 6.5 mM succinate–Tris and
Avicin D and G were obtained from ground seedpods of Acacia victoriae as described earlier [23]. The solutions of 1 mg/ml in 10% DMSO were stored at ¡20 °C before use in experiments. Sucrose, D-mannitol, digitonin, Hepes, Trizma base, EDTA, EGTA, bovine heart cytochrome c, N,N,N⬘,N⬘-tetramethyl-p-phenylenediamine (TMPD), rotenone, antimycin A, myxothiazol, carbonyl-cyanide-p-triXuoromethoxy phenylhydrazone (FCCP), 2,4-dinitrophenol (DNP), -NADH, tert-butylhydroperoxide, succinic acid, glutamate, malate, and other salts were purchased from Sigma Chemical Co. (St. Louis, MO, USA). All chemicals were of analytical grade.
Monitoring of the redox state of mitochondrial pyridine nucleotides The level of endogenous NAD(P)H of mitochondria was monitored Xuorimetrically (340 nm excitation, 450 nm Xuorescence). To minimize the inXuence of light dispersion on Xuorescence measurements, the exciting and emitting light beams were focused only on the 1.5 £ 1.5 mm corner of the cuvette [27] as shown in Fig. 1. Mitochondria were added at a concentration of 0.7 mg protein/ml to the incubation medium (100 mM sucrose, 75 mM KCl, 10 mM Hepes–Tris, and 6.5 mM succinate–Tris, pH 7.2). Avicins at Wnal concentration of 5 g/ml, 5 mM phosphate–Tris and 120 M tBH were added wherever indicated.
Measurement of the inner membrane potential The inner membrane potential of mitochondria was monitored using the potential-sensitive Xuorescent probe safranin O (520 nm excitation, 580 nm Xuorescence) [28]. Mitochondria were added at a concentration of 0.7 mg protein/ml to the incubation medium (100 mM sucrose, 75 mM KCl, 10 mM Hepes–Tris, pH 7.2, 6.5 mM succinate–Tris, 5 mM phosphate–Tris and 10 M safranin O) without or with avicins (5 g/ml).
Isolation of mitochondria Liver mitochondria from male Sprague–Dawley rats (4–6 month, starved overnight) were isolated according to the method of Schneider and Hogeboom [24] with modiWcations. Cooled liver was homogenized in medium containing 210 mM mannitol, 70 mM sucrose, 1 mM EGTA–Tris, 2.5 mM MgCl2, 10 mM Hepes–Tris, pH 7.2, at 2–4 °C. The homogenization medium contained MgCl2 to prevent cytochrome c adsorption on the mitochondria [25]. After washing mitochondria twice, they were Wnally suspended in medium containing 210 mM mannitol, 70 mM sucrose, 0.02 mM EGTA, and 0.3 mg/ml bovine serum albumin (fraction V, free fatty acids), 10 mM Hepes–Tris, pH 7.2, at 2–4 °C. Protein content was evaluated by the biuret method as described earlier [26] using bovine serum albumin as the standard and Na+-cholate.
Measurement of respiration The rate of oxygen consumption by rat liver mitochondria was measured using Clark-type oxygen electrode as described earlier [25]. Incubation medium was composed of 100 mM sucrose, 75 mM KCl, 10 mM Hepes–Tris, pH 7.2 (SKH medium). Succinate–Tris, pH 7.2 (6.5 mM), glutamate–Tris, pH 7.2 (4 mM) plus malate–Tris, pH 7.2 (1 mM), or NADH (0.6 mM) were added to the medium as substrates of oxidation, wherever indicated. After inhibition of glutamate–malate-dependent respiration, 10 mM ascorbate–Tris, pH 7.2, was added to initiate the cytochrome c- or TMPD-dependent respiration. Avicin D or avicin G (5 g/ml), 5 mM
Fig. 1. Two channel simultaneous monitoring of the Xuorescence and light dispersion using Aminoc-Bowman-2 luminescence spectrometer. To minimize the inXuence of light dispersion on Xuorescence measurements, the exciting and emitting light beams were focused only on the 1.5 £ 1.5 mm corner of the cuvette [27]. To follow mitochondrial swelling during Xuorimetric monitoring of mitochondrial NAD(P)H or of the inner membrane potential, the cuvette holder was equipped with a red light emitted diode (LED) and a photodiode (PD) allowing 90°-dispersion of light using the 2 £ 2 mm cuvette corner, which is diagonally opposite to the Xuorescence corner: the dispersed light, received by PD, is monitored through the Aux2 additional channel of the spectrometer.
116
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
5 mM phosphate–Tris, prior to the addition of mitochondria or 1 min after them. Calcium chloride (50 M) or the peptide BTM-P1 (4 g /ml) [28] were used as a positive control to induce high amplitude swelling of mitochondria. Mitochondrial swelling during Xuorimetric monitoring of mitochondrial NAD(P)H or the inner membrane potential was followed by using a modiWed cuvette holder of Aminoc-Bowman-2 luminescence spectrometer. Simultaneous measurements of 90°-dispersion of light emitted by red light emitted diode (LED) were made as shown in Fig. 1, using cuvette corner, 2 £ 2 mm, diagonally opposite to the Xuorescence corner. The signal from a photodiode (PD), receiving the dispersed light, was monitored through the Aux-2 additional channel of the spectrometer. All samples were constantly stirred with the magnetic stirrer and maintained at constant temperature of 30 °C.
Results Avicins decrease the rate of respiration of rat liver mitochondria Avicins were shown to selectively induce apoptosis in cancer cells [2]. In the case of Jurkat cells, this eVect of avicin G, for example, was accompanied with signiWcant inhibition of cell respiration (data not shown). To elucidate the mechanism by which avicins inhibit cell respiration, we studied their inXuence on respiration of rat liver mitochondria using diVerent substrates of oxidation. As shown in Fig. 2, avicin D (Fig. 2b) and avicin G (Fig. 2c) had no essential eVect on state 4 respiration with glutamate–malate as substrates of oxidation. However, respiration rates in state 3 and uncoupled state (State 3u) were decreased by 70–73% upon avicin treatments (Fig. 2b and c) as compared to untreated mitochondria (Fig. 2a). An intermediate eVect of inhibition was obtained with avicins at concentration of 2 g/ml (data not shown). In contrast to the observed inhibition of glutamate– malate–dependent respiration, the ascorbate-cytochrome c-dependent oxygen consumption was enhanced by 81% in avicin D- (Fig. 2b), and by 105% in avicin G-treated (Fig. 2c) mitochondria, as compared to untreated control (Fig. 2a). However, the rate of respiration with ascorbate– cytochrome c-TMPD was not aVected by avicins, indicating that complex IV is not inhibited (Fig. 2). Succinate-dependent respiration rates in state 3 and uncoupled state were also inhibited by avicins, by 72–74% (Table 1). State 4 respiration was inhibited by a less extent (by 26–28%) (Table 1). The maximum activity of respiratory chain was completely recovered by the addition of 20 M exogenous cytochrome c (Table 1). These results indicate that avicins cause the release of cytochrome c from rat liver mitochondria that is consistent with our previous Wnding that cytochrome c is rapidly released in avicintreated Jurkat cells, triggering the apoptotic cascade [1]. Avicins increase rotenone-insensitive oxidation of external NADH by mitochondria It has been shown that the rate of external NADH oxidation by rat liver mitochondria depends on the OMM
Fig. 2. The inXuence of avicins on the oxidative phosphorylation system of rat liver mitochondria with glutamate plus malate as substrates of oxidation. Mitochondria (Mc), 1 mg protein/ml, were added to SKH incubation medium supplemented with 4 mM glutamate, 1 mM malate, and 5 mM phosphate–Tris, pH 7.2, without (a) or with 5 g/ml of avicin D (b) and 5 g/ml of avicin G (c). ADP, 400 M ADP; Ol, 1 g/ml oligomycin; DNP, 100 M 2,4-DNP; M, 0.5 M myxothiazol, AA, 0.5 M antimycin A; Asc, 10 mM ascorbate; C,20 M cytochrome c; T, 0.5 mM TMPD. Numbers on curves— respiration rate, in natoms of oxygen/min per 1 mg of mitochondrial protein. *p < 0.01 in comparison to the untreated control.
integrity [25 and references therein]. The oxidation of external NADH can be strongly activated by damage of the OMM allowing cytochrome c shuttling between the outside facing electron transport system of the outer membrane and cytochrome c oxidase of the inner membrane [25,26]. FCCP-uncoupled mitochondria, when treated with avicin G, demonstrated more than 3-time increase in the rate of cytochrome c-dependent oxidation of external NADH (Table 2) that was insensitive to cyclosporin A, a speciWc inhibitor of mitochondrial permeability transition pore (data not shown). In these experiments, the internal pathway of NADH oxidation was inhibited by rotenone at the complex I, and by antimycin A and myxothiazol at the complex III. Further increase in the rate of oxidation was observed when the concentration of avicin G was increased
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
117
Table 1 The inXuence of avicin D and G on the respiration of rat liver mitochondria with succinate as substrate of oxidation in isotonic medium with inorganic phosphate Metabolic state
Control
Avicin D treated
Avicin G treated
State 4 State 3 State 3u State 3u, plus 20 M cytochrome c
39 § 3 170 § 11 213 § 20 207 § 11
28 § 2¤ 44 § 4¤ 54 § 6¤ 195 § 16
29 § 2¤¤ 47 § 3¤ 59 § 5¤ 191 § 11
Mitochondria, 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2, and 5 mM phosphate– Tris, pH 7.2, without (control) or with 5 g/ml of avicin D or avicin G. State 3, addition of 400 M ADP; state 3u, addition of 100 M DNP. Respiration rate—in natoms of oxygen/min per 1 mg of mitochondrial protein, means § SEM, n D 4. ¤ p < 0.01. ¤¤ p < 0.05 in comparison to the control. Table 2 The inXuence of avicin G on the rate of rotenone-insensitive oxidation of external NADH by rat liver mitochondria Additions (1) — (2) 20 M cytochrome c (3) 20 M cytochrome c plus Avicin G (5 g/ml) (4) 20 M cytochrome c plus Avicin G (15 g/ml)
Respiration rate
p
5§1 29 § 1 93 § 5
— p2–1 < 0.001 p3–2 < 0.001
178 § 9
p4–3 < 0.001 p4–2 < 0.001
Mitochondria, 0.7 mg protein/ml, and RAMF (2.5 M rotenone, 0.5 M antimycin A, 0.5 M myxothiazol, and 0.5 M FCCP) were added to SKH incubation medium supplemented with 0.6 mM NADH. After 2 min incubation, 20 M cytochrome c was added, followed by the addition of avicin G to the Wnal concentrations of 5 and 15 g/ml. Respiration rate—in natoms of oxygen/min per 1 mg of mitochondrial protein, mean § SEM, n D 6–7.
from 5 to 15 g/ml (Table 2), although it remained lower than the rate of TMPD-cytochrome c-dependent oxidation of ascorbate (Fig. 2c). The acceleration of external NADH oxidation in the presence of exogenous cytochrome c is normally caused by the rupture of the OMM under high amplitude swelling of rat liver mitochondria, or by its direct permeabilization to cytochrome c [25,26 and references therein]. The earlier eVects of avicins on mitochondrial respiration are not due to high amplitude swelling of mitochondria To determine if mitochondrial swelling is involved in the earlier avicin-induced inhibition of mitochondrial respiration (Fig. 2, Table 1), we studied the inXuence of these substances on apparent light absorbance by mitochondria at 640 nm, as well as on the light dispersion. As shown in Fig. 3, no signiWcant swelling of mitochondria was observed in avicin D (Fig. 3a) or avicin G-treated (Fig. 3b) mitochondria. However, low amplitude swelling was observed immediately upon the addition of avicins to mitochondria. This eVect was saturated (Fig. 3c and d, respectively for avicin D and avicin G), probably due to a decrease in the
Fig. 3. Avicins induce low amplitude swelling of rat liver mitochondria as compared with the following high amplitude swelling induced by the peptide BTM-P1. Mitochondria (Mc), 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2, with (a and b) or without 5 mM phosphate–Tris, pH 7.2, (c and d). D5, 5 g/ml of avicin D; G5, 5 g/ml of avicin G; D, 2.5 g/ml of avicin D; G, 2.5 g/ml of avicin G and P, 4 g/ml of the peptide BTM-P1.
oncotic pressure of the intermembrane space, which in turn could be caused by the OMM permeabilization to proteins. In comparison, the addition of the peptide BTM-P1, which was recently shown to permeabilize the inner mitochondrial membrane (IMM) [28], caused high amplitude swelling of mitochondria (Fig. 3a and b). High amplitude swelling of avicin-treated mitochondria was also induced by Ca2+ in the presence of inorganic phosphate (Fig. 4), factors, which are known to induce opening of the permeability transition pore in the IMM. These data also demonstrate that mitochondria in the presence of avicins have somewhat higher sensitivity to calcium-phosphate in undergoing high amplitude swelling. A 10-min incubation with avicins G in medium containing potassium and inorganic phosphate, resulted in a small decrease in the IMM potential (Fig. 5b) in comparison with untreated control (Fig. 5a). However, a signiWcant drop of the IMM potential was observed following a longer incubation with this drug. Interestingly, the avicin G-induced swelling of mitochondria (Fig. 5d) occurred long before the IMM potential dropped and faster than in the control (Fig. 5c). It is possible that the observed swelling is related to a potassium permeability of the IMM [29]. Avicin G-induced inhibition of respiration was observed in medium without inorganic phosphate in Wrst minutes of incubation (Table 3), suggesting that the observed permeabilization of the OMM to cytochrome c and inhibition of respiratory chain by avicins (Tables 1 and 2) is not a result of high amplitude phosphate-dependent swelling of mitochondria. No essential swelling of mitochondria was observed during at least 20 min of incubation with or
118
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
without avicin G in this medium (data not shown). Even in the presence of inorganic phosphate, high amplitude swelling of avicin G-treated mitochondria was initiated after 6– 7 min of incubation (Fig. 5). No marked decrease in the IMM potential was detected in the used phosphate-containing medium with avicins at concentration of 2 g/ml, or in a sucrose isotonic medium with avicins at concentration of 5 g/ml (data not shown). These data are consistent with no decrease in the IMM potential in Jurkat cells during Wrst several hours of application of avicins (2 g/ml) [1]. In general, the results allow suggestion that avicins initially permeabilize the OMM, with no gross alteration of the IMM permeability, thus causing the inhibition of respiratory chain due to the release of cytochrome c. Avicins signiWcantly accelerate the permeabilization of the inner mitochondrial membrane by hydroperoxides Fig. 4. Avicins slightly increase the high amplitude swelling of rat liver mitochondria induced by calcium. Mitochondria (Mc), 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2, and 5 mM phosphate–Tris, pH 7.2, without avicins (a), with 5 g/ml of avicin D (b), or with 5 g/ml of avicin G (c). Ca, 50 M CaCl2.
Fig. 5. The inXuence of avicin G on the inner membrane potential (b) of rat liver mitochondria and light dispersion (d), in comparison with untreated control (curves a and c, respectively). Mitochondria (Mc), 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2, 5 mM phosphate–Tris, pH 7.2, and 10 M safranin O; G5, 5 g/ml of avicin G.
Table 3 The inXuence of avicin G on the respiration of rat liver mitochondria with succinate as substrate of oxidation in isotonic medium without inorganic phosphate Metabolic state
Control
Avicin G treated
State 4 State 3u
26 § 1 217 § 6
20 § 1 p < 0.01 56 § 4 p < 0.01
Mitochondria, 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate and 5 g/ml avicin G. After 2-min incubation (state 4 respiration), 100 M DNP was added for state 3u. Respiration rate—in natoms of oxygen/min per 1 mg of mitochondrial protein, means § SEM, n D 4–5.
Hydroperoxides have been shown to increase the calcium-phosphate-dependent high amplitude swelling of mitochondria [27,30–33]. As well, inorganic phosphate has been shown to signiWcantly increase mitochondrial hydrogen peroxide generation [34]. The main enzymatic antioxidant defense system of mitochondria, catabolising hydroperoxides, includes glutathione peroxidase, glutathione reductase, thioredoxin peroxidase and thioredoxin reductase, and use NADPH as donator of electrons to Wnely reduce hydroperoxides [16,17]. This enzyme system is coupled to generation of the IMM proton motive force through the trans-hydrogenase, which catalyzes the energydependent reduction of NADP+ using NADH as reducing equivalent (Fig. 6). In the presence of avicins, the generation of the IMM proton motive force should be less eVective due to the cytochrome c release from mitochondria, thus avicins might aVect the oxidative stress defense system of mitochondria. As shown in Fig. 7b, avicin G-treated energized mitochondria maintained essentially decreased level of NAD(P)H in comparison to the control (Fig. 7a), especially in the presence of exogenous hydroperoxide tBH. No eVect was seen on the mitochondrial light dispersion at least for several minutes of incubation (Fig. 7d). Subsequent addition of inorganic phosphate dramatically accelerated the drop of the NAD(P)H level in avicin G-treated mitochondria in comparison to untreated control. The data also demonstrates the powerful eVect of inorganic phosphate on the swelling of avicin G-treated mitochondria (Fig. 7d), in comparison to untreated control (Fig. 7c). The swelling was observed with a short lag-period after an induction of NAD(P)H oxidation by inorganic phosphate (Fig. 7b), meanwhile the level of NAD(P)H in control mitochondria initially increased (Fig. 7a). The addition of tBH to avicintreated mitochondria in the medium containing inorganic phosphate, signiWcantly accelerated the drop of the NAD(P)H level (Fig. 8b), in comparison to untreated control (Fig. 8a), and caused faster drop of the inner membrane
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
Fig. 6. Schematic presentation of energy-dependent NAD(P)H-GSH/thioredoxin-mediated enzymatic antioxidant system of mitochondria. GSH , reduced glutathione; TR, thioredoxin; GR, glutathione reductase; TRR, thioredoxin reductase; GSSG, oxidized glutathione; TSST, oxidized thioredoxin; under succinate oxidation, high ratios NADPH/NADP+, GSH/ GSSG, TR/TSST are maintained by the energy-dependent reverse electron transport (RET) from succinate to NAD+ and by the energy-dependent trans-hydrogenase. The decreased activity of the respiratory chain due to the avicin-induced cytochrome c release results in a diminished capacity to generate proton motive force and the energy-dependent recovery of NADPH Wnely used to reduce hydroperoxides; thus avicins signiWcantly aVect the oxidative stress defense system of mitochondria. THG, trans-hydrogenase.
Fig. 7. The inXuence of sequential additions of tBH and inorganic phosphate on the redox state of mitochondrial pyridine nucleotides (a and b) and light dispersion (c and d) in the absence (a and c) and presence (b and d) of avicin G. Mitochondria, 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2; G5, 5 g/ml of avicin G; tBH, 120 M tBH; Pi, 5 mM phosphate–Tris, pH 7.2.
potential (Fig. 9, traces b and a, respectively) inducing high amplitude swelling of mitochondria (Figs. 8 and 9, traces d and c, for avicin G-treated and control mitochondria, respectively).
119
Fig. 8. Avicin G accelerates tBH-induced oxidation of mitochondrial NAD(P)H (b) and high amplitude swelling (d) in comparison with untreated control (a and c, respectively) under succinate oxidation. Mitochondria, 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2, and 5 mM phosphate– Tris, pH 7.2; G5, 5 g/ml of avicin G; tBH, 120 M tBH.
Fig. 9. Avicin G accelerates tBH-induced drop of the inner membrane potential (b) and high amplitude swelling (d) of rat liver mitochondria in comparison with untreated control (a and c, respectively) under succinate oxidation. Mitochondria (Mc), 0.7 mg protein/ml, were added to SKH incubation medium supplemented with 6.5 mM succinate–Tris, pH 7.2, 5 mM phosphate–Tris, pH 7.2, and 10 M safranin O; G5, 5 g/ml of avicin G; tBH, 120 M tBH.
Discussion In eukaryotic cells, mitochondria play a crucial role in inducing cell death via apoptosis, besides energy production
120
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
and regulating many other metabolic pathways. Mitochondria are also key players in metabolic changes related to the pathogenesis of cancer [35]. Recent experimental studies showed that cells with greater tumorigenic potential consume more oxygen, but exhibit decreasing ATP yield per unit of consumed oxygen [36] by an unknown mechanism. These cells are believed to use the mitochondrial oxidative phosphorylation system for purposes other than ATP generation, which could include generation of heat or ROS. However, mitochondria in cancer cells are characterized by resistance to oxidative stress and to other pro-apoptotic factors that seems to arise from either reduced synthesis of proteins like Bax and Bid, which favor the outer mitochondrial membrane permeabilization to cytochrome c, and/or from the over-expression of proteins like Bcl-2 and Bcl-xL preventing cytochrome c release [37–39]. Targeting the energy generation system of tumor cells, which would include both the mitochondrial and glycolytic pathways, can prove to be successful strategy for anticancer therapy. Avicins, recently discovered class of plant stress metabolites induce apoptosis in a variety of human tumor cells by a direct action on mitochondria in a p53-independent manner [1,3,4]. In Jurkat cells, besides inducing apoptosis, avicins cause inhibition of the Jurkat cells respiration (data not shown). To understand how avicins regulate mitochondrial metabolism, we studied here the eVect of avicin D and avicin G on respiration of relatively highly coupled rat liver mitochondria. The obtained data demonstrate signiWcant inhibition of respiration in metabolic state 3 and uncoupled state, by avicins, which could be completely restored by the addition of exogenous cytochrome c. The observed reactivation of respiratory chain (Table 1), as well as an increase in cytochrome c-dependent oxidation of ascorbate (Fig. 2b and c) and of exogenous NADH (Table 2) seem to not be related to a possible increase of the aYnity of the outer side of the IMM to cytochrome c, because the concentration of added cytochrome c was several-times higher than Km of mitochondrial cytochrome c oxidase for cytochrome c [25]. The obtained data rather indicate that avicin-induced inhibition of respiration is due to the release of cytochrome c from mitochondria. With this respect, avicins actuate like Bax, a pro-apoptotic agent, which has been shown to inhibit mitochondrial respiration subsequently reactivated by exogenous cytochrome c [40]. Cytochrome c is an ancient molecule critical for electron transport in mitochondria. It may be instrumental in maintaining the redox balance of the cell and could act as an antioxidant [41]. In the last decade, the critical role of cytochrome c in inducing programmed cell death has been elucidated. Thus, the function of mitochondria and speciWcally cytochrome c are critical factors in life and death decisions by cells. Cytochrome c, released from mitochondria, can be maintained in the cytoplasm in reduced state by the cellular reducing equivalents such as glutathione [6] or NADH [25]. The reduced form of cytochrome c is believed to be non-apoptotic [42]. In rat liver mitochondria, cytochrome c-dependent oxidation of external NADH is signiWcantly
increased by avicins, due to cytochrome c shuttling between the OMM of all mitochondria and the IMM cytochrome c oxidase of mitochondria with the permeabilized or damaged OMM [25,43]. The released cytochrome c could also alter the redox state of the cytoplasm. Increased steady state level of NAD+, as a result of increased oxidation of cytoplasmic NADH, could aVect various redox-signaling and NAD+-dependent processes in the cells. It has been hypothesized, for example, that the redox state of cytoplasmic pyridine nucleotides is a crucial regulator of the energy metabolism in prostate cancer cells and could be a potential target for prostate cancer therapy [44]. Liver mitochondria are characterized by a very high NADH–cytochrome c reductase activity due to the presence of Xavoprotein–cytochrome b5 electron transport system in the OMM and in the membranes of endoplasmic reticulum [45]. Therefore, as long as NADH in the cytoplasm is maintained at high level by glycolysis, cytochrome c should be kept in its reduced state [25], thus delaying apoptosis. How cytochrome c is released from mitochondria in response to avicin treatment is not clear. Recently, it was shown that avicin D and avicin G permeabilize artiWcial lipid membranes, as well as the plasma membrane of red blood cells [46]. The ability of avicin G, but not of avicin D, to form channels in artiWcial lipid membranes signiWcantly decreased in the absence of cholesterol. Both avicin G and avicin D self-assemble into channels with an estimated pore diameter of 2.2 nm, which is not large enough to be permeable to cytochrome c [46]. A theoretical model of a 10strand channel formed by avicin G, suggested by authors, has a hole diameter, which is consistent with the experimentally estimated size of the pore. On the other hand, according to the earlier electron microscopic observations [47,48], saponin forms uniform holes or pits of approximately 8 nm in diameter, each surrounding by a ring of a 3 nm in width, in the Rous sarcoma virus membrane, in the cell membrane from chicken liver and in human and guinea pig erythrocytes. It was estimated that approximately 20 molecules of saponin-type glycosides with the lipid–soluble triterpene system and Wve water-soluble sugar molecules, as occurs in many saponins, surround each 8-nm hydrophilic hole and interact through their peripheral lipophilic groups with cholesterol in a 1:1 molar ratio [48]. Although the holes formed by saponins seems to be suYcient to be permeable to cytochrome c, the sugar moieties of saponin molecules would be exposed into the hole, according to the 8-nm model [48], thus the diameter of the most narrow part of the hole, the size of a Wlter, might be signiWcantly less and even close to that estimated by Li et al. [46]. Although the holes formed by saponin in biological membranes were shown to be insensitive to trypsin or pepsin treatments, and can be also formed in a simple layer of cholesterol on the surface of distilled water without any proteins [48], theoretically it cannot be ruled out that avicins induce a protein or protein–lipids oligomerization in the OMM, leading to formation of signiWcantly larger pores permeable to cytochrome c. For example, BH3 domain-
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
dependent homo-oligomerization of Bax and Bak [49–53] or BH3 domain-independent homo-oligomerization of tBID [50], have been shown to permeabilize the OMM to cytochrome c and to other mitochondrial pro-apoptotic factors by a mechanism that does not involve permeability transition pore activation in the inner membrane. Recently, it has been demonstrated that Bax undergoes a transient conformational change in a lipid membrane surface before being inserted into the membrane: this Bax–lipid interaction was shown to be a prerequisite for tBid-induced Bax oligomerization and pore formation [54]. Similarly, the permeabilization of red blood cells by natural toxin lysenin [55] strongly depends on sphingomyelin, which has been demonstrated to be indispensable for lysonin oligomerization in the membrane. Mitochondria of tumor and normal cells, for example, signiWcantly diVer in the cholesterol content in their OMM (see [46] for references), but a diVerence in other lipids, as well as in the content of membrane proteins, might contribute in a higher eYciency of avicins in tumor cell. While avicins induce the release of cytochrome c, causing inhibition of respiration of both Jurkat cells and rat liver mitochondria, the Jurkat cells undergo apoptosis, whereas rat hepatocytes were resistant to avicins (data not shown), indicating that some additional events besides cytochrome c release are also involved in killing of Jurkat cells. The inhibition of respiratory chain by avicins should decrease an eYciency of NAD(P)H-GSH/thiredoxindependent enzymatic antioxidant system (Fig. 6) followed by increase in the sensitivity of mitochondria to the IMM permeabilization by tBH (Figs. 7 and 8). As result, tBH in the presence of avicin G caused accelerated drop of the IMM potential and high amplitude swelling of mitochondria (Fig. 9). This is consistent with Crompton’s suggestion [56] that the IMM depolarization would decrease the capacity of the energy-linked trans-hydrogenase to maintain high NADPH/NADP+ ratio thus allowing decreasing GSH/GSSG ratio and oxidation of the IMM vicinal thiols that has been demonstrated to open permeability transition pore [16,57,58]. The decreased eYciency of the NAD(P)H-GSH/thiredoxin-dependent enzymatic antioxidant system in the presence of avicins might explain the observed selectivity in killing tumor cells by these natural saponins. Tumor cells have been shown to generate hydrogen peroxide and other ROS with the rate signiWcantly higher than in normal cells, as well as to maintain signiWcantly higher concentrations of this toxic compounds, close to the threshold of cytotoxicity [18–22]. Most of anticancer drugs actuate inducing elevated amounts of intracellular ROS in tumor cells [17,22,59,60]. Recently, it was discovered that mitochondrial factor P66Shc, localized within the mitochondrial intermembrane space and normally associated in a silent form with TIM– TOM import complex between the OMM and IMM, may be liberated by various pro-apoptotic factors that converts it into a very active generator of ROS using fully reduced cytochrome c as donator of electrons [61]. The permeabili-
121
zation of the OMM to cytochrome c would facilitate maintenance of cytochrome c in a fully reduced form by using cytoplasmic NADH as donator of electrons [25,43,45]. On the other hand, avicins signiWcantly increased mitochondrial sensitivity to hydroperoxides, revealed in signiWcant acceleration of tBH-induced high amplitude swelling of mitochondria. Mitochondrial swelling cause the rupture of the OMM and massive release of mitochondrial pro-apoptotic factors [62]. Any factors and drugs increasing hydrogen peroxide generation in tumors seem to potentiate anticancer eYciency and cancer selectivity of avicins. In conclusion, the avicins represent a novel class of natural electrophilic metabolites that aVect the pro-oxidant/ antioxidant balance as well as the life and death decisions at several nodal points [1,6]. We demonstrate that avicins might lower cell energy metabolism by increasing the OMM permeability to cytochrome c, which in turn results in inhibition of respiration and in decreased eYciency of mitochondrial NAD(P)H-GSH/thiredoxin-dependent enzymatic antioxidant system, leading to increased sensitivity of cells to oxidative stress. Thus, the primary eVect of avicins at the mitochondrial level is the permeabilization of the OMM to cytochrome c, and the secondary eVect is the creation of hypersensitivity of mitochondria to oxidative stress. Based on this synergistic eVect of avicins and hydroperoxides, we speculate that avicins could have great potential when used with other anticancer drugs, which are known to increase ROS generation in tumor cells, thus allowing more selective and eYcient killing of cancers. The obtained results open up exciting possibilities for future biological and clinical exploration of avicins in the treatment of cancers. Acknowledgments Research was supported by the Clayton Foundation for Research. Additional support was provided by the Biomedical Research Foundation, the Abraham J. and Phyllis Katz Foundation, and the Colciencias (Colombia) Grant #221305-16851. We thank Drs. Craig B. Thompson, Eyal Gottlieb, Marco Colombini and Neil Blackstone for their valuable suggestions and help. References [1] V. Haridas, M. Higuchi, G.S. Jayatilake, D. Bailey, K. Mujoo, M.E. Blake, C.J. Arntzen, J.U. Gutterman, Proc. Natl. Acad. Sci. USA 98 (2001) 5821–5826. [2] K. Mujoo, V. Haridas, J.J. HoVmann, G.A. Wachter, L.K. Hutter, Y. Lu, M.E. Blake, G.S. Jayatilake, D. Bailey, G.B. Mills, J.U. Gutterman, Cancer Res. 61 (2001) 5486–5490. [3] A. Gaikwad, A. Poblenz, V. Haridas, C. Zhang, M. Duvic, J.U. Gutterman, Clin. Cancer Res. 11 (2005) 1953–1962. [4] N. Mitsiades, C.J. McMullan, V. Poulaki, J., Negri, J. Gutterman, K.C. Anderson, C.S. Mitsiades, Blood 104 (2004) abstract #2403. [5] J.U. Gutterman, H.T. Lai, P. Yang, V. Haridas, A. Gaikwad, S. Marcus, Proc. Natl. Acad. Sci. USA 102 (2005) 12771–12776.
122
V.V. Lemeshko et al. / Archives of Biochemistry and Biophysics 454 (2006) 114–122
[6] M. Hanausek, P. Ganesh, Z. Walaszek, C.J. Arntzen, T.J. Slaga, J.U. Gutterman, Proc. Natl. Acad. Sci. USA 98 (2001) 11551–11556. [7] V. Haridas, C.J. Arntzen, J.U. Gutterman, Proc. Natl. Acad. Sci. USA 98 (2001) 11557–11562. [8] V. Haridas, M. Hanausek, G. Nishimura, H. Soehnge, A. Gaikwad, M. Narog, E. Spears, R. Zoltaszek, Z. Walaszek, J.U. Gutterman, J. Clin. Invest. 113 (2004) 65–73. [9] V. Haridas, S-O. Kim, G. Nishimura, A. Hausladen, J.S. Stamler, J.U. Gutterman, Proc. Natl. Acad. Sci. USA 102 (2005) 10088–10093. [10] N.W. Blackstone, M.M. Kelly, V. Haridas, J.U. Gutterman, Proc. R. Soc. Lond. B. Biol. Sci. 272 (2005) 527–531. [11] N. Dias, C. Bailly, Biochem. Pharmacol. 70 (2005) 1–12. [12] J.S. Armstrong, Br. J. Pharmacol. 147 (2006) 239–248. [13] T.K. Rostovtseva, W. Tan, M. Colombini, J. Bioenerg. Biomembr. 37 (2005) 129–142. [14] J.J. Lemasters, E. Holmuhamedov, Biochim. Biophys. Acta 1762 (2006) 119–181. [15] L. Bouchier-Hayes, L. Lartigue, D. Newmeyer, J. Clin. Invest. 115 (2005) 2640–2647. [16] A.J. Kowaltowski, R.F. Castilho, A.E. Vercesi, FEBS Lett. 495 (2001) 12–15. [17] M.P. Rigobello, A. Folda, G. Scutari, A. Bindoli, Arch. Biochem. Biophys. 441 (2005) 112–122. [18] T.P. Szatrowski, C.F. Nathan, Cancer Res. 51 (1991) 794–798. [19] A. Gupta, S.F. Rosenberger, G.T. Bowden, Carcinogenesis 20 (1999) 2063–2073. [20] N.W. Schoene, K.S. Kamara, Free Radic. Biol. Med. 27 (1999) 364–369. [21] A.L. Jackson, L.A. Loeb, Mutat Res. 477 (2001) 7–21. [22] A. Laurent, C. Nicco, C. Chereau, C. Goulvestre, J. Alexandre, A. Alves, E. Levy, F. Goldwasser, Y. Panis, O. Soubrane, B. Weill, F. Batteux, Cancer Res. 65 (2005) 948–956. [23] G.S. Jayatilake, D.R. Freeberg, Z. Liu, S.L. Richheimer, M.E. Blake, D.T. NietoBailey, V. Haridas, J.U. Gutterman, J. Nat. Prod. 66 (2003) 779–783. [24] W.C. Schneider, G.H. Hogeboom, J. Biol. Chem. 183 (1950) 123–128. [25] V.V. Lemeshko, J. Biol. Chem. 277 (2002) 17751–17757. [26] V.V. Lemeshko, Arch. Biochem. Biophys. 388 (2001) 60–66. [27] V.V. Lemeshko, Biochem. Biophys. Res. Commun. 291 (2002) 170–175. [28] V.V. Lemeshko, M. Arias, S. Orduz, J. Biol. Chem. 280 (2005) 15579– 15586. [29] H.T. Facundo, M. Fornazari, A.J. Kowaltowski, Biochim. Biophys. Acta 1762 (2006) 202–212. [30] R.F. Castilho, A.J. Kowaltowski, A.R. Meinicke, E.J. Bechara, A.E. Vercesi, Free Radic. Biol. Med. 18 (1995) 479–486. [31] T.E. Gunter, D.R. PfeiVer, Am. J. Physiol. 258 (1990) C755–C786. [32] A.-L. Nieminen, A.M. Byrne, B. Herman, J.J. Lemasters, Am. J. Physiol. 272 (1997) C1286–C1294. [33] A.M. Byrne, J.J. Lemasters, A.L. Nieminen, Hepatology 29 (1999) 1523–1531. [34] A.J. Kowaltowski, L.E. Netto, A.E. Vercesi, J. Biol. Chem. 273 (1998) 12766–12769.
[35] J.L. GriYn, J.P. Shocker, Nat. Rev. Cancer 4 (2004) 551–561. [36] A. Ramanathan, C. Wang, S.L. Schreiber, Proc. Natl. Acad. Sci. USA 102 (2005) 5992–5997. [37] N. Zamzami, G. Kroemer, Nat. Rev. Mol. Cell. Biol. 2 (2001) 67–71. [38] G. Brumatti, R. Weinlich, C.F. Chehab, M. Yon, G.P. AmaranteMendes, FEBS Lett. 541 (2003) 57–63. [39] A.S. Don, P.J. Hogg, Trends Mol. Med. 10 (2004) 372–378. [40] F. Appaix, K. Guerrero, D. Rampal, M. Izikki, T. Kaambre, P. Sikk, D. Brdiczka, C. Riva-Lavieille, J. Olivares, M. Longuet, B. Antonsson, V.A. Saks, Biochim. Biophys. Acta 1556 (2002) 155–167. [41] V.P. Skulachev, FEBS Lett. 423 (1998) 275–280. [42] Z. Pan, D.W. Voehringer, R.E. Meyn, Cell Death DiVer. 6 (1999) 683– 688. [43] P. Bernardi, G.F. Azzone, J. Biol. Chem. 256 (1981) 7187–7192. [44] P.W. Hochachka, J.L. Rupert, L. Goldenberg, M. Gleave, P. Kozlowski, BioEssays 24 (2002) 749–757. [45] C. Schnaitman, J.W. Greenawalt, J. Cell. Biol. 38 (1968) 158–175. [46] X.X. Li, B. Davis, V. Haridas, J.U. Gutterman, M. Colombini, Biophys. J. 88 (2005) 2577–2584. [47] R.R. Dourmashkin, R.M. Dougherty, R.J. Harris, Nature 194 (1962) 1116–1119. [48] A.D. Bangham, R.W. Horne, A.M. Glauert, J.T. Dingle, J.A. Lucy, Nature 196 (1962) 952–955. [49] B. Antonsson, S. Montessuit, B. Sanchez, J.-C. Martinou, J. Biol. Chem. 276 (2001) 11615–11623. [50] M. Grinberg, R. Sarig, Y. Zaltsman, D. Frumkin, N. Grammatikakis, E. Reuveny, A. Gross, J. Biol. Chem. 277 (2002) 12237–12245. [51] V. Mikhailov, M. Mikhailova, K. Degenhardt, M.A. Venkatachalam, E. White, P. Saikumar, J. Biol. Chem. 278 (2003) 5367–5376. [52] X. Yi, X.-M. Yin, Z. Dong, J. Biol. Chem. 278 (2003) 16992–16999. [53] M.G. Annis, E.L. Soucie, P.J. Dlugosz, J.A. Cruz-Aguado, L.Z. Penn, B. Leber, D.W. Andrews, EMBO J. 24 (2005) 2096–2103. [54] J.A. Yethon, R.F. Epand, B. Leber, R.M. Epand, D.W. Andrews, J. Biol. Chem. 278 (2003) 48935–48941. [55] A. Yamaji-Hasegawa, A. Makino, T. Baba, Y. Senoh, H. KimuraSuda, S.B. Sato, N. Terada, S. Ohno, E. Kiyokawa, M. Umeda, T. Kobayashi, J. Biol. Chem. 278 (2003) 22762–22770. [56] M. Crompton, Biochem. J. 341 (1999) 233–249. [57] B.V. Chernyak, P. Bernardi, Eur. J. Biochem. 238 (1996) 623–630. [58] G.P. McStay, S.J. Clarke, A.P. Halestrap, Biochem. J. 367 (Pt 2) (2002) 541–548. [59] P.X. Petit, M.C. Gendron, N. Schrantz, D. Metivier, G. Kroemer, Z. Maciorowska, F. Sureau, S. Koester, Biochem. J. 353 (2001) 357–367. [60] L. Zhou, Y. Jing, M. Styblo, Z. Chen, S. Waxman, Blood 105 (2005) 1198–1203. [61] M. Giorgio, E. Migliaccio, F. Orsini, D. Paolucci, M. Moroni, C. Contursi, G. Pelliccia, L. Luzi, S. Minucci, M. Marcaccio, P. Pinton, R. Rizzuto, P. Bernardi, F. Paolucci, P.G. Pelicci, Cell 122 (2005) 221–233. [62] P. Bernardi, V. Petronilli, F. Di Lisa, M. Forte, Trends Biochem. Sci. 26 (2001) 112–117.
