Cell Biochem Biophys (2010) 58:169–179 DOI 10.1007/s12013-010-9104-1
ORIGINAL RESEARCH
Effects of Menadione, Hydrogen Peroxide, and Quercetin on Apoptosis and Delayed Luminescence of Human Leukemia Jurkat T-Cells Irina Baran • Constanta Ganea • Agata Scordino • Francesco Musumeci • Vincenza Barresi • Salvatore Tudisco • Simona Privitera • Rosaria Grasso • Daniele F. Condorelli • Ioan Ursu • Virgil Baran • Eva Katona • Maria-Magdalena Mocanu Marisa Gulino • Raluca Ungureanu • Mihaela Surcel • Cornel Ursaciuc
•
Published online: 10 August 2010 Springer Science+Business Media, LLC 2010
Abstract Menadione (MD) is an effective cytotoxic drug able to produce intracellularly large amounts of superoxide anion. Quercetin (QC), a widely distributed bioflavonoid, can exert both antioxidant and pro-oxidant effects and is known to specifically inhibit cell proliferation and induce apoptosis in different cancer cell types. We have investigated the relation between delayed luminescence (DL) induced by UV-laser excitation and the effects of MD, hydrogen peroxide, and QC on apoptosis and cell cycle in human leukemia Jurkat T-cells. Treatments with 500 lM I. Baran (&) C. Ganea E. Katona M.-M. Mocanu R. Ungureanu Department of Biophysics, ‘‘Carol Davila’’ University of Medicine and Pharmacy, 8 Eroii Sanitari, 050474 Bucharest, Romania e-mail:
[email protected] A. Scordino F. Musumeci S. Tudisco S. Privitera R. Grasso M. Gulino Laboratori Nazionali del Sud, Istituto Nazionale di Fisica Nucleare, 62 S. Sofia, Catania, Italy A. Scordino F. Musumeci S. Tudisco S. Privitera R. Grasso M. Gulino Dipartimento di Metodologie Fisiche e Chimiche per l’Ingegneria, Universita` di Catania, 95125 Catania, Italy V. Barresi D. F. Condorelli Dipartimento di Scienze Chimiche, Sezione di Biochimica e Biologia Molecolare, Universita` di Catania, 6 A. Doria, 95125 Catania, Italy I. Ursu V. Baran IFIN-HH, 407 Atomistilor Str., 077125 Magurele Bucharest, Romania M. Surcel C. Ursaciuc Department of Immunology, ‘‘Victor Babes’’ National Institute, 99-101 Spl. Independentei, Bucharest, Romania
H2O2 and 250 lM MD for 20 min produced 66.0 ± 4.9 and 46.4 ± 8.6% apoptotic cell fractions, respectively. Long-term (24 h) pre-exposure to 5 lM, but not 0.5 lM QC enhanced apoptosis induced by MD, whereas shortterm (1 h) pre-incubation with 10 lM QC offered 50% protection against H2O2-induced apoptosis, but potentiated apoptosis induced by MD. Since physiological levels of QC in the blood are normally less than 10 lM, these data can provide relevant information regarding the benefits of flavonoid-combined treatments of leukemia. All the three drugs exerted significant effects on DL. Our data are consistent with (1) the involvement of Complex I of the mitochondrial respiratory chain as an important source of delayed light emission on the 10 ls–10 ms scale, (2) the ability of superoxide anions to quench DL on the 100 ls– 10 ms scale, probably via inhibition of reverse electron transfer at the Fe/S centers in Complex I, and (3) the relative insensitivity of DL to intracellular OH• and H2O2 levels. Keywords Apoptosis Delayed luminescence Oxidative stress Flavonoids Mitochondrial respiratory chain Leukemia
Introduction Menadione (vitamin K3) is a clinically important chemotherapeutic agent used in the treatment of leukemia and other cancer types [1, 2]. Menadione (MD) participates in redox cycling reactions catalyzed by a number of flavoenzymes [3–5], thus producing intracellularly large amounts of superoxide anion (O-• 2 ). MD reduction at Complex I of the mitochondrial respiratory chain (MRC) [3, 5], which accounts for 50% of MD metabolism [5], can
170
readily divert the electron flow from Complex I [3] and therefore interfere with mitochondrial respiration. Quercetin (QC) is a natural flavonoid ubiquitously occurring in fruits, vegetables, and tea. This well-investigated phytochemical can display both antioxidant and prooxidant properties and is known to specifically inhibit cell proliferation and induce apoptosis in different cancer cell types [6–13]. Numerous studies have indicated that malignant cells are more susceptible than normal cells to QC cytotoxicity [6, 9, 14], and this property could be exploited to increase the efficiency of leukemia chemotherapies by employing QC-combined treatments. In human acute leukemia Jurkat T-cells, QC accumulates in large quantities inside the mitochondria, where it is stored in a biologically active form by binding to QC-binding proteins [15] and can act as an activator or inhibitor of the mitochondrial permeability transition pore (mPTP), depending on its pro- or anti-oxidant character, respectively [16]. QC is able to inhibit MRC Complexes I and III [17] and participate in quinone redox cycling [9, 18]. Therefore, high doses of QC increase cellular H2O2 and -• O-• 2 production [8, 14, 16]. O2 can be subsequently dismutated to H2O2 by cytosolic or mitochondrial superoxide dismutases, and additional OH• can be produced from H2O2 through Fe/Cu-dependent Fenton reactions. For short periods of treatment, QC effectively decreases the cellular H2O2 content, whereas in long-term administration it depletes the intracellular pool of the endogenous antioxidant GSH (reduced glutathione), leading to further accumulation of reactive oxygen species (ROS) and toxic metabolites of QC [8, 10, 11, 14, 16]. MD, hydrogen peroxide, and QC can activate the apoptotic program via a Ca2?-dependent mitochondrial pathway, by promoting elevation of cytosolic Ca2? levels, mPTP opening, collapse of the mitochondrial membrane potential (Dwm) and release of cytochrome c from mitochondria [1, 4, 6, 12, 16, 17, 19–26]. Commitment to apoptosis or necrosis depends on whether the blockade of ATP production and the opening of mPTP are transitory or prolonged, respectively [1, 22, 26, 27]. In Jurkat cells, MD and H2O2 at low/medium concentrations preferentially induce cell death through apoptosis, whereas high concentrations promote necrosis by disrupting ATP production [1, 17, 19, 21, 26]. At the moment, the current available data on the effects of QC on apoptosis induced by MD and hydrogen peroxide in Jurkat T-cells are extremely limited. In this paper we investigate these effects under treatments of different time and dosage of the flavonoid and provide a series of useful data regarding the dual, pro- and anti-oxidant character of QC in this cellular system. Moreover, our data could aid in the better understanding of the molecular mechanisms underlying the mode in which this flavonoid acts in
Cell Biochem Biophys (2010) 58:169–179
leukemia cells and could contribute to the characterization of QC as a clinical chemotherapeutic agent. Importantly, our investigations indicate the ability of QC to enhance MD-induced apoptosis in Jurkat leukemia cells even at physiological levels of 5–10 lM. Our studies also provide relevant information regarding the optimization of the time and dosage of QC administration in a flavonoid–MD combined treatment of leukemia. Having in view the growing interest of using delayed luminescence (DL) spectroscopy in clinical applications [28–32], a second goal of our study was to gain new insights into the biochemical mechanisms responsible for DL of living cells, as well as to provide new data regarding the relation between DL and the cell status. Thus, we have investigated, for the first time to our knowledge, the correlation between apoptosis, oxidative stress, and DL. DL (also called delayed fluorescence) represents a very weak, long-time scale light emission following exposure to pulsed light or UV radiation. At the moment, the origin of DL is still a matter of debate, but it is generally thought that DL in biological systems may arise from a variety of possible reactions and sources, such as direct emitters like flavins, carbonyl derivatives and aromatic compounds, molecular oxygen and its species, the DNA, as well as collective molecular interactions, e.g., triplet–triplet annihilation, collective hydrolysis, charge recombination within the mitochondrial/chloroplast electron transport system, or the cytoskeleton [33–41]. We took advantage of the specific ability of QC to accumulate inside the mitochondria of Jurkat cells, as well as of the fact that both QC and MD appear to interact robustly with the mitochondria and to induce apoptosis through the mitochondrial pathway. Thus, we could probe whether DL is connected to the mitochondrial metabolism. In addition, the characteristic growth of Jurkat cells in suspension makes this cell type more suitable for DL investigation, as it greatly minimizes the sample preparation time and the cell stress during probation for DL, and any artifactual interference of a supplementary trypsinization step with the cellular metabolism is thus avoided. We evaluated the effects of MD, hydrogen peroxide, and QC on DL of Jurkat cells and found a strong anti-correlation between apoptosis and DL on a specific time scale (100 ls–1 ms after UV-excitation of the cell sample). Moreover, our data are consistent with an important role of MRC Complex I in delayed photoemission and indicate that superoxide is a powerful quencher of DL. Nevertheless, further studies are needed to characterize the exact relation between DL and the mitochondrial respiratory system, but the progresses achieved in promoting DL spectroscopy as a valuable tool in the diagnosis of mitochondrial disorders or cancer [28–32] point to the critical need for such investigations.
Cell Biochem Biophys (2010) 58:169–179
Materials and Methods Cell Cultures Human leukemia Jurkat T cell lymphoblasts were cultured in suspension in MegaCell RPMI 1640 medium (Sigma M3817) supplemented, according to the manufacturer, with 5% fetal bovine serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100 lg/ml streptomycin, at 37C in a humidified incubator with a 5% CO2 atmosphere. We used hydrogen peroxide 30% solution (Sigma H1009) and stock solutions of menadione sodium bisulphite (Sigma M2518) dissolved in phosphate buffer saline (PBS), or dihydrated QC (Sigma Q0125) dissolved in dimethyl sulfoxide (DMSO). Logarithmic cells were pre-incubated with QC or the vehicle for 24 or 1 h, as indicated. DMSO was 0.1% (v/v) in all cultures. MD and H2O2 were added directly to the cultures, which were further incubated for the indicated times. After the treatment, cells were washed twice with PBS and resuspended in PBS (for DL samples, *40 9 106 cells/ml) or in complete medium (for proliferation/apoptosis assessment, *0.2 9 106 cells/ml). DL samples were analyzed immediately by DL spectroscopy. Cell density, viability, and morphology were examined with a CCD camera Logitech QuickCam Pro 4000, connected to an Olympus CK30 phase contrast microscope, giving an 80009 overall magnification. For cell density assessment, 25 ll aliquots of the DL samples were diluted in PBS, stained with 0.4% trypan blue solution and *1500–2000 cells were imaged on a Burker haemocytometer at the time of the DL assay. The cells that excluded the dye were considered as living. Cell count evaluation was performed both during DL experiments, directly by visual inspection under the microscope, and later on, by analyzing the recorded photographs with the ImageJ software. Additional trypan blue exclusion tests were performed 24 h after the treatments.
Flow-Cytometry 24 and 48 h after the treatment, samples containing 106 cells were fixed in 70% ethanol and frozen at -20C. For flow-cytometer determinations, the ethanol-fixed samples were washed with PBS, incubated with a propidium iodide PI/RNAse staining buffer (PHARMINGEN 550825) for 30 min at 37C in the dark and analyzed with a Becton– Dickinson FACS Calibur flow-cytometer. For data acquisition and analysis we used the CellQuest, WinMDI 2.8, and Cylchred software. Apoptosis was evaluated as the fraction of hypodiploid cell fragments (the sub-G0/G1 cell fraction). The G0/G1, S and G2/M cell fractions were calculated for the non-apoptotic cell population, by excluding
171
the hypodiploid events from cell-cycle analysis. For analysis of the flow-cytometer distributions (number of counts vs. PI fluorescence) we also developed an alternative, simple least-square fitting procedure in which the distributions of events corresponding to the sub-G0/G1 or S-phases, which appeared as relatively wide distributions, could be excellently reproduced by a sum of three Gaussian distributions, whereas the distributions corresponding to the G0/G1 or G2/M-phases, which appeared each as a distinct, relatively narrow peak, were very well reproduced by a single Gaussian distribution. The results provided by this analysis were in fact closely similar to those obtained with the WinMDI 2.8/Cylchred software. For each treatment, the 24/48-h data (at least six per treatment) were used collectively to calculate the mean and the standard error of the mean (s.e.m.). For the 50 lM QC-treatment we also evaluated the apoptosis and cell-cycle distributions in samples fixed in ethanol 9 h after QC removal.
DL Spectroscopy We used an improved version of the ARETUSA set-up [37]. This highly sensitive equipment is able to detect single photons and has a very low background signal; moreover it presents a high efficiency in collecting the luminescence originating from the cell sample, and a low gap-time (11 ls) between the end of the excitation pulse and the beginning of signal acquisition. The cell samples were excited by a Nitrogen Laser source (Laser Photonics LN 230C; wavelength 337 nm, pulse-width 5 ns, energy 100 ± 5 lJ/pulse). A multi-alkali photomultiplier tube (Hamamatsu R-7602-1/Q) was used as a detector for photoemission signals with wavelengths in the range 400–800 nm, in single photon counting mode. The laser power was reduced, in some cases, to prevent the dimpling of the photomultiplier. The detected signals were acquired by a Multi-channel Scaler (Ortec MCS PCI) with a minimum dwell-time of 200 ns. DL measurements were done on at least 3 different drops from each cell sample (drop volume 15–25 ll) at room temperature (20 ± 1C). PBS luminescence was subtracted from all the recordings. Photoemission was recorded between 11 ls and 0.1 s after laser-excitation; however, due to large fluctuations in the signal-to-noise ratio at the end of the decay curves, only emission up to 10 ms was analyzed. DL intensity (I) was obtained as the number of photons recorded within a certain time interval divided to that time interval and to the number of living cells in the drop. The quantum yield was calculated in three time domains of the DL intensity curve: 11–100 ls (denoted DL-I), 100 ls–1 ms (DL-II), and 1–10 ms (DL-III), as the ratio between the I-integral and the energy of the laser.
172
Statistics All the data presented in this paper were obtained from at least three different experiments, and are expressed as mean ± s.e.m. Statistically significant differences were determined using Student’s t-test compared to control or between the indicated groups. A level of P \ 0.05 was considered significant in all statistical tests.
Results Effects of MD, Hydrogen Peroxide, and QC on Apoptosis and the Cell Cycle Flow-cytometer measurements (Fig. 1) and trypan blue exclusion tests indicated that QC-treatment with 10 lM for 1 h or 0.5 lM for 24 h did not exert cytotoxic effects on Jurkat cells. However, 5 lM QC-treatment for 24 h raised the apoptotic rate to 14.5 ± 2.3% (vs. 9.2 ± 3.1% in control cells), and 24-h incubation with 50 lM QC robustly increased the apoptotic cell fraction to 74.1 ± 3.4% (Fig. 1a). Our data are consistent with other studies on QC cytotoxicity in Jurkat cells, which indicated, based on MTT cell viability assays, an IC50 * 8 lM for treatment durations C24 h [6, 7, 12]. We also obtained a good agreement with previous assessment of the apoptotic fraction shortly after the 50 lM QC-treatment for 24 h (51.9 ± 1.4% after 9 h from drug removal in this work, as compared with &45% immediately after the treatment [6]). This treatment activated the G2/M checkpoint and arrested the cell cycle in the G2/M-phase for a certain period (G2/M cell fraction 9 h after drug removal: 39.3 ± 1.4% in 50 lM QC-treated cells vs. 16.8 ± 3.2% in control cells). The ability of QC to halt the cell cycle at the G2/M-phase was reported in Jurkat and other cell types [7, 13] where appears to operate via down-regulation of cyclin B1 [13]. However, the non-apoptotic cells in QC-treated cultures presented normal cell-cycle distributions 24–48 h after release (Fig. 1b-d), suggesting that the cell-cycle effects of QC in these cells were reversible. In favor of this, we obtained clonogenic survival *40% in cells treated with 50 lM QC for 24 h (not shown), indicating that the surviving cells eventually removed the G2/M block and continued proliferation. 25 lM MD for 20 min increased the apoptotic cell fraction to 14.7 ± 4.1%, in agreement with other reports [21, 23], whereas longer treatments (4 h) increased the apoptotic rate to a consistently higher value, 36.7 ± 3.6% (Fig. 1a). In the latter case there was a partial blockage in the G2/M-phase of the cell cycle (G2/M fraction 25.2 ± 2.3% in MD-treated cells vs. 16.8 ± 3.3% in control cells, P \ 0.01), indicating the presence of DNA
Cell Biochem Biophys (2010) 58:169–179
damage and activation of the G2/M checkpoint (Fig. 1d), probably via inhibition of cyclin B/cdk1 complex formation [2]. MD at high concentration (250 lM for 20 min) produced 46.4 ± 8.6% apoptotic cells, but did not change the cell-cycle distribution after C24 h from the treatment (Fig. 1b–d). Pre-incubation with 5 lM QC for 24 h or 10 lM QC for 1 h significantly enhanced apoptosis induced by 250 lM MD (apoptotic fractions: 64.7 ± 10.4 and 70.6 ± 8.6%, respectively), suggesting an increased ROS production in combination treatments as compared with MD alone. In the latter treatment (denoted ‘‘QC10* ? MD’’) we observed an increase in the fraction of S-phase cells (Fig. 1c), which may be due to the presence of persistent mitochondrial DNA damage in some cells [42]. Treatment with 500 lM H2O2 for 20 min potently induced apoptosis in 66.0 ± 4.9% cells (Fig. 1a), in agreement with previous reports [19]. Pre-incubation with 0.5 lM QC for 24 h increased the apoptotic cell fraction to 73.7 ± 3.1%, whereas pre-exposure to 10 lM QC for 1 h protected cells from H2O2-induced apoptosis (33.4 ± 8.3% apoptotic ratio). The latter effect appears to be associated with the activation of the G2/M checkpoint (Fig. 1d): the G2/M cell fraction at 24 h after the H2O2 challenge increased from 17.8 ± 4.7% (with no QC-preincubation) to 57.9 ± 1.7% (with QC-preincubation). Moreover, the G2/M fraction decreased thereafter to 41.7 ± 4.9% at 48 h after the ‘‘QC10* ? H’’ treatment, indicating the onset of release from the G2/M arrest and resumption of cell proliferation. Interestingly, a lower dose of H2O2 (100 lM for 20 min) produced 34.8 ± 9.9% apoptotic cells and a G2/M ratio of 34.6 ± 4.5% 24–48 h after the treatment (not shown), whereas the highest dose used (500 lM for 20 min) produced a moderate increase in the S-phase cell fraction (Fig. 1c). Since H2O2 can induce cell-cycle arrest at G2/M by expression of p21Cip1, a cyclin-dependent kinase inhibitor [43], it is possible that in our experiments Jurkat cells were unable to cope with the extensive oxidative damage produced by the highest dose of H2O2 and thus could not complete the current cell-cycle phase in less than 48 h. In favor of this, the damage repair following a milder stress appeared to be quite slow (as discussed above, the G2/M blockade persisted for [48 h in both treatments with 100 lM H2O2 and 500 lM H2O2 ? 10 lM QC-preincubation). In addition, hemocytometer counts indicated (not shown) that there was no net growth in cell cultures 24–48 h after the most severe oxidative treatments (namely ‘‘MD’’ and ‘‘H’’). Together, these results suggest that under excessive oxidative stress induced by MD or hydrogen peroxide, Jurkat cells accumulate severe damage and therefore are stopped in all the phases of the cell cycle. Under our experimental conditions, a large part of the arrested cells initiated the apoptotic program; however, it
Cell Biochem Biophys (2010) 58:169–179
173
A
B
Apoptotic cells (%) 100 90 80 70 60 50 40 30 20 10 0
G0/G1 cells (%) 60
a
a
a,b
a,b
a,c
a
50
a
40
a a,c
a
a
a
a
a
30 20
a,b
a
10
C
D
S cells (%)
G2/M cells (%)
60
a
a
a
H 0. 5+ H Q C 10 *+ H C
Q
M D 0. 5+ M D Q C 5+ M Q D C 10 *+ M D Q C
Q C 10 M * D 25 x2 M 0' D2 5 x4 h
5
50
C Q
Q C
l
5 0.
Ct r
Q C
Q C
H 0. 5+ Q H C 10 *+ H
M D 0. 5+ M D Q C 5+ Q M C D 10 *+ M D
M
Q C
Q C 10 * D 25 x2 M 0 ' D 25 x4 h
5
50 Q C
0. 5
Q C
Q C
C tr l
0
a,c
60
50
50
40
40
10 *+ H
H 0. 5+ H Q C
Q C
M D 0. 5+ M D Q C 5+ M Q D C 10 *+ M D
M
Q C
Q C 1 D2 0* 5 x2 M 0' D 25 x4 h
50
5 Q C
Q C
C tr l
Q C
H 0. 5+ H Q C 10 *+ H Q C
M D 0. 5+ M D Q C 5+ M Q D C 10 *+ M D Q C
Q C 10 M * D 25 x2 0' M D 25 x4 h
50 Q C
Q C
C tr
5
0 0. 5
10
0 l
20
10
Q C
a
30
20
0. 5
a,c
30
Fig. 1 Apoptosis (a) and cell-cycle (b–d) distributions after treatment of Jurkat cells with 0.5, 5 or 50 lM quercetin for 24 h (QC0.5, QC5, QC50), 10 lM quercetin for 1 h (QC10*), menadione (MD25: 25 lM for 20 min or 4 h as indicated, MD: 250 lM for 20 min), 500 lM H2O2 for 20 min (H), or after combined treatments
(quercetin pre-incubation followed by addition of 250 lM menadione or 500 lM H2O2 for 20 min). a—Significantly different from Control (Ctrl), b—significantly different from MD (presented for the menadione-group only, black bars), c—significantly different from H (presented for the H2O2-group only, hatched bars)
appeared that the surviving cells eventually managed to repair the damage and continued to proliferate after [48 h from exposure (we obtained clonogenic survival *40% for the treatment with 250 lM MD, not shown). Finally, it is worth mentioning that trypan blue viability tests indicated that necrosis was low (\15%) under all treatments, which is consistent with other reports [19, 21] and with the clonogenic survival data mentioned above. Next we investigated the characteristics of DL of Jurkat cells after the treatments described in Fig. 1, and considered that apoptosis evaluated at 24–48 h after exposure was an indicative measure of the level of oxidative stress induced by the drugs.
shown as it goes beyond the scope of this paper). Figure 2a presents an example where the time course of UV-induced light emission of untreated cells is compared with that of 50 lM QC-treated cells. To better evidence the effects of various chemical treatments on DL kinetics, we also represented the time course of the photoemission intensity relative to the DL intensity of control cells (e.g., Fig. 2b–d). This depiction revealed that there are three distinct time domains in which DL manifests different characteristics, as will be discussed below in more detail. Both QC and MD inhibited DL in a dose-dependent manner (Fig. 2b, c). Surprisingly, we found that QC and MD at high doses exhibited virtually identical effects on DL over a wide time interval, from 100 ls to 10 ms after laser-excitation (Fig. 2d). Having in view the molecular interactions that are common to both chemicals, it is most likely that their similar effects on DL were caused by: (1) superoxide and/or H2O2 production, (2) inhibition of Complex I of the mitochondrial respiratory chain, (3) Dwm collapse induced by mPTP opening, (4) damaged DNA or (5) cell-cycle arrest. The latter option appears to be highly unlikely, since the DL assays were performed within 30–50 min after the treatments and hence the time elapsed
Effects of MD, Hydrogen Peroxide, and QC on DL The DL spectra of Jurkat cells presented a complex, multicomponent decay (Fig. 2). The experimental kinetic curves of the photoemission intensity have the appearance of a relatively straight decrease when displayed in log–log scale (e.g., Fig. 2a), which could be fitted by assuming linear combinations of at least six exponential components for any of the treatments used in this study (this analysis is not
174
Cell Biochem Biophys (2010) 58:169–179
A
B 1.E+06 QC50
1.E+05
Control 1.E+04
I (a.u.)
1.E+03 1.E+02 1.E+01 1.E+00 1.E-01 0.00001
0.0001
0.001
0.01
t (s)
C
D
I/ICtrl
1
0.1
M25 MD25 Q0.5 + M250 QC0.5 + MD250 Q5 + M250 QC5 + MD250 M250 MD250 0.01 0.00001
0.0001
0.001
0.01
t (s) Fig. 2 Kinetics of DL emission of Jurkat cells after various treatments, labeled as in Fig. 1. In b–d the intensity of light emission (I) is normalized to the DL intensity of untreated cultures (ICtrl)
was not sufficient for the cells to accumulate in a certain phase of the cell cycle. In the following we discuss these aspects in more detail and analyze the correlation with the apoptosis data presented in the previous section. DL-I photoemission (time scale 11–100 ls) appeared to be relatively insensitive to QC (Fig. 2a, b) but considerably reduced by MD (Fig. 2c). DL-I exhibited a noticeable insensitivity even at high levels of QC (50 lM), in samples where *40% of the cells were in advanced stages of apoptosis (not shown), arguing against an essential role of ROS in DL-I. MD at concentrations C25 lM produced a consistent and fairly even (*60%) reduction in the DL-I quantum yield, irrespective of the treatment time or the prior administration of QC (Fig. 3a), indicating a saturation effect. 500 lM H2O2 produced a moderate (*40%) reduction in the DL-I quantum yield (Fig. 3a). In our experiments, the necrotic cell fraction appeared to be low (\15% after trypan blue staining) under all treatments, indicating that ATP production and Dwm were only transiently altered after exposure to QC, MD, or H2O2. So, we can infer that the cells treated with QC for 24 h had normal
respiratory activity and normal Dwm at the moment of the DL assay, which is consistent with the finding that the inhibitory effects of QC on mitochondrial respiration and Dwm are rapid (15 min) [17]. Long et al. [24] found that in H2O2-treated cardiomyocytes the activity of MRC complexes II and IV decline within 20 min of exposure, followed by Dwm dissipation after 20 min of exposure, whereas the decrease in complex I activity manifests later, after 60 min. Godar [23] has determined in Jurkat cells that 25 lM MD applied for 20 min or 4 h, respectively, decreases Dwm in about 10 and 70% of the cells immediately after the treatment. Other studies also indicate that Dwm disruption is one of the earliest steps in apoptosis and takes place immediately after the apoptosis induction treatments, including MD stress [25]. Collectively, these observations indicate that Dwm is re-established after an initial, rapid decline following the treatments applied in our study to Jurkat cells. Since our DL measurements were done within 50–70 min from exposure to MD or H2O2, it is therefore unlikely that the changes we observed in DL emission were caused by alteration of Dwm. Moreover,
Cell Biochem Biophys (2010) 58:169–179
DNA involvement as a source of DL fails to explain our data, since QC50-samples presented high levels of DNA fragmentation, and the protective QC10*-pretreatment did not improve the DL-I decrease produced by H2O2 (Fig. 3a). An additional argument comes from our studies on a different cell type, the budding yeast Saccharomyces cerevisiae, which presented similar kinetics of DL photoemission in untreated cells (not shown) and it appeared that DL is not correlated with the existence of DNA strand breaks or microtubule damage [41]. Taken together, all these findings and observations support the hypothesis that DL-I is related to the respiratory activity of the mitochondria, and that DL-I reduction by MD and H2O2 was most likely due to a temporary inhibition of forward electron transport in MRC complex I. This idea could also explain the moderate correlation between DL-I and apoptosis obtained from all treatments (Fig. 3d). The relatively moderate reduction of DL-I observed in H2O2 treatments may simply reflect the delay in complex I inhibition, similarly to the cardiomyocytes case [24], so that we probably did not capture the maximal inhibition of complex I. The presumed timing of inhibition of mitochondrial respiration and ATP production (*60 min after exposure) is also consistent with other reports [24, 26, 27]. Overall, our data obtained with MD-treated cells (Fig. 3a) suggest an important (*60%) contribution to DL-I of forward electron transfer in Complex I. The regions DL-II (time scale 100 ls–1 ms) and DL-III (1–10 ms) were highly sensitive to both QC and MD, in a dose-dependent manner (Figs. 2, 3). There is a strong negative correlation between the DL-II quantum yield and the apoptotic cell fraction under all treatments (Fig. 3e), suggesting that all the three drugs (MD, H2O2, and QC) affect DL-II by a common mechanism. For reasons detailed above, our results indicate that inhibition of forward reactions in the respiratory chain and mitochondrial depolarization may participate to this effect only to a small extent. Likely candidates for explaining this behavior are the reactive oxygen species OH•, H2O2, and O-• 2 , which can be produced inside the cells by all of these drugs (H2O2 could indirectly cause an increase in the O-• level via 2 inhibition/damage of respiratory complexes [22]). However, the QC10*-pretreatment, which offered consistent protection against the H2O2 insult, failed to restore DL-II/ III emission (Fig. 3b,c), suggesting that DL-II/III could be • correlated to the O-• 2 level rather than to the OH /H2O2 levels. In favor of this, inhibition of DL-II/III emission in cells treated with the superoxide generators MD and QC was dose-dependent and appeared to be strikingly similar at high concentrations of the drugs (Fig. 2d). In the ‘‘QC10* ? H’’ treatment, the dominant effect on DL-II/III emission was exerted by QC (not shown). This observation can be explained by the flavonoid reduction of OH•/H2O2
175
levels being accompanied by a transient increase in mitochondrial O-• 2 production via inhibition of the respiratory chain. Furthermore, the relatively reduced (*30–40%) decrease of DL-II/III emission in QC10*-treated cells (Fig. 3b,c) could be explained if the deleterious effects of QC-induced superoxide production were promptly annihilated by activation of endogenous antioxidants, an interpretation substantiated by the lack of cytotoxicity of the QC10*-treatment. By assuming that the contribution of forward electron transfer in Complex I is [ 0% and \ 15% in the QC50-treatment, we could estimate from the corresponding data (&60% and &90% maximal inhibition in Fig. 3b, c, respectively) that the contribution of the O-• 2 sensitive states is &45–60% to DL-II and &75–90% to DL-III. The higher contribution of these states to DL-III vs. DL-II can explain why the H2O2 treatments appeared to weaken the correlation between DL-III and apoptosis, whereas the correlation remained strong in treatments with MD, QC, or MD/QC-combination (Fig. 3e, f).
Discussion This work adds more data to the growing body of evidence that emphasize the need to accurately establish the therapeutic range and hence the proper dosage in QC supplementation. In human normal lymphocytes QC can produce • noticeable quantities of O-• 2 and OH at levels C50 lM within 30 min of treatment [14]. After intake, QC is rapidly metabolized and normal levels in the plasma remain below 10 lM [9], but can increase considerably in human tissues, in particular at the inflammatory sites [15]. In this work, a 24-h treatment with a physiological level (5 lM) of QC enhanced apoptosis induced by MD in human leukemia Jurkat T-cells, whereas a short-term treatment with 10 lM QC reduced apoptosis induced by hydrogen peroxide. While long-term administration of QC could improve significantly the MD-based treatment of leukemia, it is important that normal cells remain unexposed to noxious levels of the flavonoid. The fact that malignant cells are more susceptible than normal cells to QC cytotoxicity could therefore represent an additional benefit in a flavonoid-combined treatment. Moreover, our studies provide some new insights into the relation between the cell status and DL. With treatments of varying time and dosage of three different proapoptotic agents (MD, hydrogen peroxide, and QC) we obtained a strong anti-correlation (r = -0.79) between apoptosis of human leukemia Jurkat cells and UV-induced delayed photoemission on a specific time interval ranging from 100 ls to 1 ms after UV-excitation of the cell samples. More specifically, the MD/QC/MD ? QC treatments targeted at increasing the intracellular superoxide level
176
Cell Biochem Biophys (2010) 58:169–179
A
D DL-I relative quantum yield (%)
10 µs – 100 µs
DL-I Relative quantum yield (% of control) 120 100
a 80
a
60
a
a
a
a
a a
40 20
120 80 40 0 0
C
Q
100 µs – 1 ms
120 100
a
80
a
60
a
a
a
a a
40
a
a
a
20
DL-II relative quantum yield (%)
E
a,b
80
100
QC MD/MD+QC H/H+QC
120
80
40
0 20
H 0. 5+ H Q C 10 *+ H
M D 0. 5+ M D Q C 5 +M Q D C 10 *+ M D
40
60
80
100
C
Apoptotic cells (%)
Q C
Q
x4 h
x2 0' 5
D2 5 M
50 M D2
10 *
Q C
5 C
Q C
Q
60
160
0 0. 5
40
rall = -0.79 rMQ = -0.90
0 Q C
20
Apoptotic cells (%)
B DL-II Relative quantum yield (% of control)
QC MD/MD+QC H/H+QC
160
H 0. 5+ H Q C 10 *+ H
M D 0. 5+ M Q D C 5 +M Q D C 10 *+ M D Q C
Q C 5 M D2 0 5 x2 M D2 0' 5 x4 h
5
10 *
C
Q C
Q
Q C
0. 5
0
rall = -0.59 rMQ = -0.57
F
C DL-III Relative quantum yield (% of control)
120
b a
a
a
100
a
a,c
60
a
40
a
20
a
a
a a H 10 *+
Q C
H 0. 5+ H C Q
M D 0. 5+ M D Q C 5 +M Q D C 10 *+ M D C
Q
5 M D2 0 5 x2 M 0' D2 5 x4 h
10 * C
Q C
5 Q
C Q
Q C
0. 5
0
DL-III relative quantum yield (%)
140
80
rall = -0.64 rMQ = -0.84
1 ms – 10 ms
QC MD/MD+QC H/H+QC
160
120
80
40
0 0
20
40
60
80
100
Apoptotic cells (%)
Fig. 3 Relative DL-quantum yield (a–c) on different time scales (boxes) and its correlation to the apoptotic cell fraction (d–f) under the treatments indicated in Fig. 1; the treatment with 100 lM H2O2 for 20 min was also included in the analysis. Significant differences between treatments are represented as in Fig. 1. QC, MD/MD ? QC,
and H/H ? QC denote single QC treatments, MD treatments with or without QC-preincubation, and H2O2 treatments with or without QCpreincubation, respectively. The Pearson correlation coefficients are shown for all treatments (rall) and for MD/MD ? QC/QC treatments (rMQ)
were associated with a strong inverse correlation between both DL-II and -III and apoptosis (rMQ = -0.90 and -0.84, respectively). A potential application of DL
spectroscopy can thus be envisioned as a rapid and accurate method to assess the pro-apoptotic capacity of certain chemical treatments for acute leukemia.
Cell Biochem Biophys (2010) 58:169–179
Our data also indicate the DL-quenching ability of superoxide. Early studies have suggested that singlet oxygen (1O2) is an important factor in mitochondrial luminescence [36]. 1O2 can increase luminescence of organic chromophores by energy transfer to electronically excited states of the chromophore [44]. In the presence of water, 1 O2 can be produced from superoxide; however, at high levels O-• 2 appears to quench singlet oxygen [44], which could explain the DL reduction by superoxide generators. Spontaneous photoemission by mitochondria was associated with lipid peroxidation that occurs inside the mitochondrial membrane and produces 1O2 [36]. However, we did not detect any DL increase in our experiments with varying levels of superoxide generators. Moreover, H2O2 appears to exert its cytotoxic effects on Jurkat cells without producing significant levels of lipid peroxides [45], and in our work the QC10*-pretreatment, which offered substantial protection against H2O2 and hence would prevent the lipoperoxidation effects of H2O2 (if any), did not ameliorate the DL reduction observed after the H2O2 treatment. Taken together, these findings suggest that DL in Jurkat cells is not related appreciably to lipid peroxidation or production of singlet oxygen. Earlier works reported that electron transport inhibitors reduce light emission from mitochondria and chloroplasts [33, 36, 39]. Some recent models [38, 40] describe DL of plants and thylakoid membranes on the sub-second scale as deriving from radiative charge recombination in Photosystem II (PSII) reaction centers, and propose the existence of three light-emitting PSII-states. Their lifetimes, determined by the rates of electron transfer between different redox states, were predicted to be *100 ls, *600 ls, and *2–3 ms, respectively [38]. There is a notable similarity to our observations regarding the kinetics of DL in Jurkat cells and the role of mitochondrial Complex I in delayed light emission. PSII, the first component in the photosynthetic electron transfer chain, represents the counterpart of MRC Complex I, in which the two electrons delivered by NADH to flavine mononucleotide (FMN) are individually transferred between eight consecutive iron–sulfur (Fe/S) clusters and eventually reach the ubiquinone. In Complex I, two different time scales (&90 ls and &1–2 ms) of the electron flow kinetics were resolved in the time domain 90 ls–8 ms [46], which are similar to the DL time scales revealed by our measurements on Jurkat cells. Keeping in line with all these observations, our results support the notion that FMN can absorb UV radiation thus producing excited singlet states that may either decay to the ground state by prompt fluorescence [22] or undergo intersystem crossing to long-lived triplet states [47] which can further relax to metastable intermediate states [47]. In analogy with the PSII case [38, 40], the long lifetime of the triplet- or metastable-state species allows a series of
177
photochemical reactions to occur in MRC Complex I and produce secondary excitations, thus giving rise to DL. In this framework, our data suggest, as a possible scenario, that charge recombination in the Fe/S sites occurring during successive electron transfer steps in MRC Complex I can proficiently modulate the intensity and kinetics of DL on a long-time scale (up to at least 10 ms). Indeed, charged sulfur was found to significantly enhance the rate of flavine triplet decay [47]. The observed DL-II and DL-III reduction by MD, QC, and H2O2 treatments suggests inhibition by superoxide of reverse electron transfer at the two extreme Fe/S centers N1a and N2 (for DL-II, decay times *100 ls [46]), or at the remaining centers (for DL-III, decay times *1 ms [46]), which requires the reduced Fe/S sites to react with O-• 2 [46]. According to the higher sensitivity of DL-III over DL-II, it would mean that the six intermediary Fe/S centers are, in their reduced form, more readily accessible to, or have higher affinity for O-• 2 than the two extreme Fe/S clusters. In conclusion, our data are consistent with the hypothesis that DL of Jurkat cells originates in major part from Complex I of the mitochondrial respiratory chain, and that DL-I is determined mainly by forward electron transfer reactions, whereas DL-II and DL-III are determined by reverse electron transfer reactions within Complex I. Certainly, further investigations are needed to clarify these aspects, but it is growingly clear that the use of DL spectroscopy in clinical diagnostic applications or in prospective studies of cancer treatments can offer the benefit of accuracy and rapidity of the tests. Acknowledgments This work was partially supported by the Romanian Ministry of Education and Research under CNCSISUEFISCSU Grant PNII-IDEI no. 1138/2009, code 1449/2008, and CNMP Grant PNII-Partnership no. 71-073/2007.
References 1. Laux, I., & Nel, A. (2001). Evidence that oxidative stress-induced apoptosis by menadione involves Fas-dependent and Fas-independent pathways. Clinical Immunology, 101, 335–344. 2. Matzno, S., Yamaguchi, Y., Akiyoshi, T., Nakabayashi, T., & Matsuyama, K. (2008). An attempt to evaluate the effect of vitamin K3 using as an enhancer of anticancer agents. Biological and Pharmaceutical Bulletin, 31, 1270–1273. 3. Brie`re, J. J., Schlemmer, D., Chretien, D., & Rustin, P. (2004). Quinone analogues regulate mitochondrial substrate competitive oxidation. Biochemical and Biophysical Research Communications, 316, 1138–1142. 4. Criddle, D. N., Gerasimenko, J. V., Baumgartner, H. K., Jaffar, M., Voronina, S., Sutton, R., et al. (2007). Calcium signalling and pancreatic cell death: Apoptosis or necrosis? Cell Death and Differentiation, 14, 1285–1294. 5. Floreani, M., & Carpenedo, F. (1992). One- and two-electron reduction of menadione in guinea-pig and rat cardiac tissue. General Pharmacology, 23, 757–762.
178 6. Chen, D., Daniel, K. G., Chen, M. S., Kuhn, D. J., Landis-Piwowar, K. R., & Dou, Q. P. (2005). Dietary flavonoids as proteasome inhibitors and apoptosis inducers in human leukemia cells. Biochemical Pharmacology, 69, 1421–1432. 7. De Vincenzo, R., Ferlini, C., Distefano, M., Gaggini, C., Riva, A., Bombardelli, E., et al. (2000). In vitro evaluation of newly developed chalcone analogues in human cancer cells. Cancer Chemotherapy and Pharmacology, 46, 305–312. 8. Ferraresi, R., Troiano, L., Roat, E., Lugli, E., Nemes, E., Nasi, M., et al. (2005). Essential requirement of reduced glutathione (GSH) for the anti-oxidant effect of the flavonoid quercetin. Free Radical Research, 39, 1249–1258. 9. Jeong, J. H., An, J. Y., Kwon, Y. T., Rhee, J. G., & Lee, Y. J. (2009). Effects of low dose quercetin: Cancer cell-specific inhibition of cell cycle progression. Journal of Cellular Biochemistry, 106, 73–82. 10. Kim, G. N., & Jang, H. D. (2009). Protective mechanism of quercetin and rutin using glutathione metabolism on H2O2induced oxidative stress in HepG2 cells. Annals of the New York Academy of Sciences, 1171, 530–537. 11. Kim, B. M., Choi, Y. J., Han, Y., Yun, Y. S., & Hong, S. H. (2009). N, N-dimethyl phytosphingosine induces caspase-8dependent cytochrome c release and apoptosis through ROS generation in human leukemia cells. Toxicology and Applied Pharmacology, 239, 87–97. 12. Rao, Y. K., Geethangili, M., Fang, S. H., & Tzeng, Y. M. (2007). Antioxidant and cytotoxic activities of naturally occurring phenolic and related compounds: A comparative study. Food and Chemical Toxicology, 45, 1770–1776. 13. Zhang, Q., Zhao, X. H., & Wang, Z. J. (2009). Cytotoxicity of flavones and flavonols to a human esophageal squamous cell carcinoma cell line (KYSE-510) by induction of G2/M arrest and apoptosis. Toxicology in Vitro, 23, 797–807. 14. Yen, G. C., Duh, P. D., Tsai, H. L., & Huang, S. L. (2003). Prooxidative properties of flavonoids in human lymphocytes. Bioscience, Biotechnology, and Biochemistry, 67, 1215–1222. 15. Fiorani, M., Guidarelli, A., Blasa, M., Azzolini, C., Candiracci, M., Piatti, E., et al. (2010). Mitochondria accumulate large amounts of quercetin: Prevention of mitochondrial damage and release upon oxidation of the extramitochondrial fraction of the flavonoid. The Journal of Nutritional Biochemistry, 21, 397–404. 16. De Marchi, U., Biasutto, L., Garbisa, S., Toninello, A., & Zoratti, M. (2009). Quercetin can act either as an inhibitor or an inducer of the mitochondrial permeability transition pore: A demonstration of the ambivalent redox character of polyphenols. Biochimica et Biophysica Acta, 1787, 1425–1432. 17. Dorta, D. J., Pigoso, A. A., Mingatto, F. E., Rodrigues, T., Prado, I. M., Helena, A. F., et al. (2005). The interaction of flavonoids with mitochondria: Effects on energetic processes. ChemicoBiological Interactions, 152, 67–78. 18. Metodiewa, D., Jaiswal, A. K., Cenas, N., Dickancaite´, E., & Segura-Aguilar, J. (1999). Quercetin may act as a cytotoxic prooxidant after its metabolic activation to semiquinone and quinoidal product. Free Radical Biology and Medicine, 26, 107–116. 19. Barbouti, A., Amorgianiotis, C., Kolettas, E., Kanavaros, P., & Galaris, D. (2007). Hydrogen peroxide inhibits caspase-dependent apoptosis by inactivating procaspase-9 in an iron-dependent manner. Free Radical Biology and Medicine, 43, 1377–1387. 20. Chien, S. Y., Wu, Y. C., Chung, J. G., Yang, J. S., Lu, H. F., Tsou, M. F., et al. (2009). Quercetin-induced apoptosis acts through mitochondrial- and caspase-3-dependent pathways in human breast cancer MDA-MB-231 cells. Human and Experimental Toxicology, 28, 493–503. 21. Dumont, A., Hehner, S. P., Hofmann, T. G., Ueffing, M., Dro¨ge, W., & Schmitz, M. L. (1999). Hydrogen peroxide-induced
Cell Biochem Biophys (2010) 58:169–179
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35. 36.
37.
38.
apoptosis is CD95-independent, requires the release of mitochondria-derived reactive oxygen species and the activation of NF-kappaB. Oncogene, 18, 747–757. Foster, K. A., Galeffi, F., Gerich, F. J., Turner, D. A., & Mu¨ller, M. (2006). Optical and pharmacological tools to investigate the role of mitochondria during oxidative stress and neurodegeneration. Progress in Neurobiology, 79, 136–171. Godar, D. E. (1999). UVA1 radiation triggers two different final apoptotic pathways. Journal of Investigative Dermatology, 112, 3–12. Long, X., Goldenthal, M. J., Wu, G. M., & Marı´n-Garcı´a, J. (2004). Mitochondrial Ca2? flux and respiratory enzyme activity decline are early events in cardiomyocytes response to H2O2. Journal of Molecular and Cellular Cardiology, 37, 63–70. Macho, A., Hirsch, T., Marzo, I., Marchetti, P., Dallaporta, B., Susin, S. A., et al. (1997). Glutathione depletion is an early and calcium elevation is a late event of thymocyte apoptosis. Journal of Immunology, 158, 4612–4619. Saito, Y., Nishio, K., Ogawa, Y., Kimata, J., Kinumi, T., Yoshida, Y., et al. (2006). Turning point in apoptosis/necrosis induced by hydrogen peroxide. Free Radical Research, 40, 619–630. Lyamzaev, K. G., Izyumov, D. S., Avetisyan, A. V., Yang, F., Pletjushkina, O. Y., & Chernyak, B. V. (2004). Inhibition of mitochondrial bioenergetics: The effects on structure of mitochondria in the cell and on apoptosis. Acta Biochimica Polonica, 51, 553–562. Ortner, M. A., Ebert, B., Hein, E., Zumbusch, K., Nolte, D., Sukowski, U., et al. (2003). Time gated fluorescence spectroscopy in Barrett’s oesophagus. Gut., 52, 28–33. Musumeci, F., Applegate, L. A., Privitera, G., Scordino, A., Tudisco, S., & Niggli, H. J. (2005). Spectral analysis of laserinduced ultraweak delayed luminescence in cultured normal and tumor human cells: Temperature dependence. Journal of Photochemistry and Photobiology B: Biology, 79, 93–99. Kim, H. W., Sim, S. B., Kim, C. K., Kim, J., Choi, C., You, H., et al. (2005). Spontaneous photon emission and delayed luminescence of two types of human lung cancer tissues: Adenocarcinoma and squamous cell carcinoma. Cancer Letters, 229, 283–289. Kemmner, W., Wan, K., Ru¨ttinger, S., Ebert, B., Macdonald, R., Klamm, U., et al. (2008). Silencing of human ferrochelatase causes abundant protoporphyrin-IX accumulation in colon cancer. FASEB Journal, 22, 500–509. Mik, E. G., Johannes, T., Zuurbier, C. J., Heinen, A., HoubenWeerts, J. H., Balestra, G. M., et al. (2008). In vivo mitochondrial oxygen tension measured by a delayed fluorescence lifetime technique. Biophysical Journal, 95, 3977–3990. Felker, P., Izawa, S., Good, N. E., & Haug, A. (1973). Effects of electron transport inhibitors on millisecond delayed light emission from chloroplasts. Biochimica et Biophysica Acta, 325, 193–196. Popp, F. A., Nagl, W., Li, K. H., Scholz, W., Weingartner, O., & Wolf, R. (1984). Biophoton emission. New evidence for coherence and DNA as source. Cell Biophysics, 6, 33–52. Slawinski, J. (1988). Luminescence research and its relation to ultraweak cell radiation. Experientia, 44, 559–571. Hideg, E., Kobayashi, M., & Inaba, H. (1991). Spontaneous ultraweak light emission from respiring spinach leaf mitochondria. Biochimica et Biophysica Acta, 1098, 27–31. Tudisco, S., Scordino, A., Privitera, G., Baran, I., & Musumeci, F. (2004). ARETUSA – Advanced research equipment for fast ultraweak luminescence analysis: New developments. Nuclear Instruments and Methods in Physics Research Section A, 518, 463–464. Goltsev, V., Chernev, P., Zaharieva, I., Lambrev, P., & Strasser, R. J. (2005). Kinetics of delayed chlorophyll a fluorescence
Cell Biochem Biophys (2010) 58:169–179
39.
40. 41.
42.
registered in milliseconds time range. Photosynthesis Research, 84, 209–215. Katsumata, M., Takeuchi, A., Kazumura, K., & Koike, T. (2008). New feature of delayed luminescence: Preillumination-induced concavity and convexity in delayed luminescence decay curve in the green alga Pseudokirchneriella subcapitata. Journal of Photochemistry and Photobiology B: Biology, 90, 152–162. Guo, Y., & Tan, J. (2009). A kinetic model structure for delayed fluorescence from plants. BioSystems, 95, 98–103. Baran, I., Ganea, C., Ursu, I., Musumeci, F., Scordino, A., Tudisco, S., et al. (2009). Effects of nocodazole and ionizing radiation on cell proliferation and delayed luminescence. Romanian Journal of Physics, 54, 557–567. Koczor, C. A., Shokolenko, I. N., Boyd, A. K., Balk, S. P., Wilson, G. L., & Ledoux, S. P. (2009). Mitochondrial DNA damage initiates a cell cycle arrest by a Chk2-associated mechanism in mammalian cells. The Journal of Biological Chemistry, 284, 36191–36201.
179 43. Seomun, Y., Kim, J. T., Kim, H. S., Park, J. Y., & Joo, C. K. (2005). Induction of p21Cip1-mediated G2/M arrest in H2O2treated lens epithelial cells. Molecular Vision, 11, 764–774. 44. Khan, A. U. (1978). Activated oxygen: Singlet molecular oxygen and superoxide anion. Photochemistry and Photobiology, 28, 615–626. 45. Chen, Z. H., Saito, Y., Yoshida, Y., & Niki, E. (2008). Effect of oxygen concentration on free radical-induced cytotoxicity. Bioscience, Biotechnology, and Biochemistry, 72, 1491–1497. 46. Verkhovskaya, M. L., Belevich, N., Euro, L., Wikstro¨m, M., & Verkhovsky, M. I. (2008). Real-time electron transfer in respiratory complex I. Proceedings of the National Academy of Sciences of the United States of America, 105, 3763–3767. 47. Swartz, T. E., Corchnoy, S. B., Christie, J. M., Lewis, J. W., Szundi, I., Briggs, W. R., et al. (2001). The photocycle of a flavine-binding domain of the blue light photoreceptor phototropin. The Journal of Biological Chemistry, 276, 36493–36500.
