Photo-Induced Damages of Cytoplasmic and Mitochondrial Membranes by a [C 60 ]Fullerene Malonic Acid Derivative

Abstract

The biological activities of fullerene derivatives have attracted much attention in the last decade. In this paper, effects of dimalonic acid C₆₀ (DMA C₆₀) on cytoplasmic membrane, intracellular calcium concentration ([Ca²⁺]_i), and mitochondrial membrane in HeLa cells were studied by using laser scanning confocal microscopy together with fluorescent probes propidium iodide (PI), fluo-3 acetoxymethyl ester (fluo-3 AM), and tetramethyl rhodamine methyl ester (TMRM). The data showed that under laser irradiation produced by a Kr/Ar laser source with a low power less than 1 mW, DMA C₆₀ might induce damages against both cytoplasmic and mitochondrial membranes in a time- and dose-dependent manner. Prior to leakage of cytoplasmic membrane, a transient increase in [Ca²⁺]_i occurred due to influx of calcium from the culture medium. These data provided some novel clues to explain the mechanisms involved in the photo-induced cytotoxicity of fullerene derivatives.

In the past 10 years, the relationship between fullerenes and biology has attracted much attention (Bosi et al. 2003; Pantarotto et al. 2004). One achievement in this field is the discovery of the photo-induced cytotoxicity of various fullerene derivatives dissolved in aqueous solutions. Carboxylic acid C₆₀ derivatives were first evaluated the in vitro cytotoxicity against hepatic tumor cells HeLa S3 cell line by Tokuyama et al. When these compounds were incubated with the cells at 37°C for 72 h in the dark, no measurable activity was observed. However, distinctive inhibition of the cell growth was determined with a 6-W fluorescent light irradiation (total twice, every 24 h for 1 h each time). The IC₅₀ of these C₆₀ derivatives was about 6 μM, 1% of IC₅₀ of the potent cytotoxic agent mitomycin C (Tokuyama, Yamago, and Nakamura 1993). Recently Rancan et al. found that a tris-malonic acid C₆₀ were more toxic than a dendritic C₆₀ mono-adduct on a special line of human T-lymphocyte Jurkat cells under irradiation with both ultraviolet UVA and UVB light (Rancan et al. 2002). In addition, DNA photocleavage can be performed in the presence of fullerene derivatives from a considerable number of studies. The cleavage of nucleic acids was regarded to play a direct role in the photocytotoxicity of fullerene derivatives (Bosi et al. 2003).

In a previous study, we compared the photo-induced cytotoxicity of fullerene malonic acid derivatives with different numbers of adducts (n = 2~4) against human cervix uteri tumor–derived HeLa cells and found the dimalonic acid C₆₀ (DMA C₆₀) was most effective (Yang, Fan, and Zhu 2002). The structure of DMA C₆₀ was illustrated in Figure 1. As a successive study, here we investigated the photo-induced effects of this compound on both cytoplasmic and mitochondrial membranes as well as the intracellular calcium concentration ([Ca²⁺]_i) in HeLa cells, by the utilization of a laser scanning confocal microscopy (LSCM) together with fluorescent probes including propidium iodide (PI), tetramethyl rhodamine methyl ester (TMRM), and fluo-3 acetoxymethyl ester (fluo-3 AM). Evidence obtained from the present experiments showed that potent photo-induced damage was produced against both cytoplasmic and mitochondrial membranes, and calcium signal might be involved in the process.

MATERIALS AND METHODS

Preparation of DMA C₆₀

DMA C₆₀ was synthesized according to the previously described method (Cheng, Yang, and Zhu 2000). A stock of 4 mM was prepared with phosphate-buffered saline (0.2 M NaH₂PO₄/Na₂HPO₄, pH 8).

Cell Culture and Laser Treatment

HeLa cells used in all experiments were planted onto a glass cover of a Petri pool (Gene) with Delbecco’s modified Eagle medium (DMEM) (Gibco), supplemented with 15% activated fetal bovine serum (Sigma), 100 U/ml penicillin, and 100 μg/ml streptomycin, and placed in an incubator (Precision Scientific) providing a moisturized atmosphere of 37°C and 5% CO₂. Irradiation treatment was carried out with a Kr/Ar laser source (power ≤1 mW) equipped in a TCS NT LSCM (Leica), while the wavelength and time were changed when necessary.

Detection of Cytoplasmic Membrane Integrity

To cells was added 5 μg/mL PI (Sigma) in phosphate-buffered saline (138 mM NaCl, 6 mM KCl, 1 mM MgCl₂, 20 mM Hepes, 20 mM glucose, and 0.1 mM EGTA) with or without 1.1 mM CaCl₂, followed by the addition of DMA C₆₀. Successive irradiation and determination of cellular PI fluorescence were undertaken immediately by a LSCM with a low-power (≤1 mW) laser source. The excitation and emission wavelengths used for PI fluorescent determination were 567 and 630 nm, respectively. The relative PI fluorescent intensity (F _i ) at time point i was calculated according to the following formula: F _i = f _i /f ₈₀₀ × 100% (f _i is the original value at time point i, f ₈₀₀ is the original value at the 800th s, which was maximum).

[Ca²⁺]_i Determination

Cells were immersed with 80 μl of a calcium sensitive fluorescent dye fluo-3 AM (Biotum) at a final concentration of 5 μM in phosphate-buffered saline, and stained at 37°C for 60 min. Then DMA C₆₀ was added to cells, and successive irradiation and determination of cellular fluo-3 fluorescence were undertaken immediately by a LSCM with a low-power (≤1 mW) laser source. The excitation and emission wavelengths used for fluo-3 fluorescent determination were 488 and 526 nm, respectively (Hubmer et al. 1996).

Dual Determination of Cytoplasmic Membrane Integrity and [Ca²⁺]_i

Cells were stained with 5 μM Fluo-3-AM for 60 min, and incubated with 5 μg/ml PI and 40 μM DMA C₆₀. Successive irradiation and simultaneous determination of PI and fluo-3 fluorescence were undertaken immediately by a LSCM with a low-power (≤1 mW) laser source in a dual-channel way. The dual-excitation wavelengths were 488 (fluo-3) and 567 (PI) nm, whereas the dual emission wavelengths were 526 (fluo-3) and 630 (PI) nm, respectively.

Determination of Mitochondrial Membrane Integrity

Cells were stained with 1 μM TMRM (Biotium) at room temperature for 10 min. After addition of DMA C₆₀, successive irradiation and determination of PI fluorescence were undertaken immediately by a LSCM with a low-power (≤1 mW) laser source. The excitation and emission wavelengths used for TMRM fluorescent determination were 548 and 565 nm, respectively.

RESULTS

Photo-Induced Damage of Cytoplasmic Membrane by DMA C₆₀

PI is not able to pass through intact cytoplasmic membrane; however, it may flow into cells readily and produce fluorescence upon binding with cellular DNA while leakage of membrane takes place. By LSCM, the PI fluorescence was continuously determined immediately after addition of DMA C₆₀ (Figure 2). DMA C₆₀ was found to be able to cause damage against cytoplasmic membrane in HeLa cells irradiated by a laser with low-power (≤1 mW) and a wavelength of 567 nm. After DMA C₆₀ at a concentration range of 1~40 μM was added to cell culture containing PI, PI flowed into cells, resulting in a rapid increase in its fluorescence which reached maximum value within 750 s. T ₅₀, referred to the time-point when PI fluorescence reached 50% maximum, was used to evaluate the damage speed of cytoplasmic membrane. The data showed that T ₅₀ increased with the elevated concentration of DMA C₆₀. For example, T ₅₀ was about 60 s at the concentration of 40 μM, and it became 230 s or so when the concentration went down to 1 μM. In the absence of irradiation, PI fluorescence could not be observed under a fluorescent microscope even when cells were incubated with 40 μM DMA C₆₀ for up to 1 h (data not shown). Therefore, the present data indicated that the photo-induced damage of cytoplasmic membrane by DMA C₆₀ was time and dose dependent.

Effects of DMA C₆₀ on [Ca2+]_i

Unlike PI, fluo-3 AM readily diffuses through the cytoplasmic membrane and enters into cytosol, where intracellular esterases cleave its ester bond, producing free fluo-3. Fluo-3 is essentially nonfluorescent, but its fluorescent intensity at 526 nm increases at least 40 times upon Ca²⁺ binding. Therefore, a rise in [Ca²⁺]_i can be signaled in fluo-3–preloaded cells by a corresponding rise in fluorescent intensity. In the present experiment a double stain of fluo-3 and PI was used to study the involvement of intracellular calcium signal in photo-induced activity of DMA C₆₀. Figure 3 showed the continuous changes of PI and fluo-3 fluorescence in a single typical cell treated by 40 μM DMA C₆₀.[Ca²⁺]_i went up very sharply after DMA C₆₀ was added, reached the maximum at 36 s, then went down to a very low steady level at 60 s. Soon after fluo-3 fluorescence decreased, PI fluorescence increased and began to get plateau at 150 s (Figure 3). These data showed a transient rise in intracellular calcium level took place prior to the occurrence of membrane leakage, suggesting intracellular calcium signal might play a role in the photo-induced damage of DMA C₆₀ against the cytoplasmic membrane.

Using calcium-free phosphate-buffered saline instead of the calcium-containing buffer, calcium transport from the environmental medium into the cytosol was investigated. Figure 4 showed the time-course curves of fluo-3 fluorescent intensity after addition of 40 uM DMA C₆₀. In the absence of extracellular calcium, [Ca²⁺]_i decreased immediately (Figure 4, a), in comparison to an transient two fold increase and subsequent decrease in the presence of 1.1 mM extracellular calcium (Figure 4, b). It was concluded that transient increase in [Ca²⁺]_i induced by DMA C₆₀ was due to a calcium influx from the environmental medium other than a calcium release from the intracellular pools such as the endoplasmic reticulum.

Photo-Induced Damage of Mitochondrial Membrane by DMA C₆₀

As a mitochondria-selective fluorescent probe, TMRM accumulation within the mitochondria is positively correlated with the mitochondrial membrane potential. If mitochondrial membrane damage occurs, the membrane potential declines at once and thus causes a decrease or abolishment of TMRM fluorescence. Figure 5 showed the dynamic damage of the mitochondrial membrane in HeLa cells by DMA C₆₀ irradiated with a low-power (≤1 mW) laser source. The fluorescent intensity remained first at a relative steady level after addition of DMA C₆₀, and then declined rapidly. The start and duration time of the decline were both dependent on the compound concentration. At the concentration of 10 μM, the fluorescence began to decline at 130 s after addition of DMA C₆₀, and was completely abolished at about 300 s (Figure 5, a). While the concentration rose to 40 μM, the fluorescence started to fall at 50 s and declined at a much higher speed (Figure 5, b). These data indicated that DMA C₆₀ was able to stimulate the damage of mitochondrial membrane even when irradiated by a laser producer with very low power, and the stimulation was time and concentration dependent.

DISCUSSION

In the present paper the dynamic effects of DMA C₆₀ on both cytoplasmic and mitochondrial membranes, as well as intracellular calcium concentration, were determined for the first time by the utilization of the laser scanning confocal microscopic technique together with a set of fluorescent probes. The experimental data obtained showed that DMA C₆₀ could induce damages against both membranes in a time- and dose-dependent manner when irradiated by a laser producer with a low power of less than 1 mW (Figures 2, 5). In addition, a calcium influx and a transient increase in [Ca²⁺]_i took place prior to the membranous damage occurrence.

The photocytotoxicity of fullerene derivatives has been described in the previous papers, although the mechanisms involved still remain unclear (Cheng et al. 2001; Rancan et al. 2002; Yang, Fan, and Zhu 2002). Kotelnikova et al. have reported the membranotropic properties of the water-soluble alanine and alaylalanine C₆₀ (Kotelnikova et al. 1996). More recently, Foley et al. and Ali et al. have demonstrated the preferential subcellular localization of mono- and triadducts of malonic acid derivatives of C₆₀ to the mitochondria by using a mouse monoclonal antifullerene antibody (Foley et al. 2002; Ali et al. 2004). The subcellular localization of monomalonic acid fullerene was further confirmed by labeling the compound with ¹⁴C radioisotope. After the compound was incubated with cells for 24 h, the radioactivity was found mainly to locate in membranous compartments such as cytoplasmic membranes, mitochondria, and microsomes. Its highest portion was located within the mitochondria, whereas a small portion was within the cytosol (Foley et al. 2002). The present data revealed strong photo-induced damages of both cytoplasmic and mitochondrial membranes by DMA C₆₀ (Figures 2, 5). Furthermore, the damage of microsomal membranes separated from rat liver cells can also be caused by fullerol C₆₀(OH)₁₈ in the presence of irradiation by ultraviolet (UV) or tungsten lamps (Kamat et al. 2000; Sera, Tokiwa, and Miyata 1996). Taking these data together, it is concluded that various cellular membranous structures are all possible intracellular targets for the action of fullerene derivatives.

Based on the information mentioned above, a three-step pathway is speculated on the photo-induced cytotoxicity of fullerene derivatives related to cellular membrane damages. First, fullerene derivatives bind with various cellular membranous structures after they pass through intact cytoplasmic membranes. Then, membrane damages are induced in the case of light irradiation due to lipid peroxidation and protein oxidation (Kamat et al. 2000). Finally, irreversible cell death is produced by functional loss of the cell membrane system. Using various scavengers of reactive oxygen species (ROS), different ROS has been proven to mediate membrane damages by fullerene derivatives (Kamat et al. 2000; Rancan et al. 2002; Sayes et al. 2005). Taking the advantages of LSCM technique along with a specific fluorescent dye dihydroxyl rhodamine (DHR), intracellular ROS level is now being examined in our laboratory in order to confirm the putative roles of ROS in photocytotoxicity of fullerene derivatives.

Yonuschot has established that increased intracellular calcium was an early event in the photodynamic permeabilization of thymocytic cell membrane in the presence of a photosensitizer erythrosin B (Yonuschot 1991). Busch et al. showed that intracellular calcium level rose prior to the photolysis of the cytoplasmic membrane, and that this early rise in intracellular calcium was necessary for membrane rupture (Busch et al. 1998). Photoexcitation of DMA C₆₀ examined in the present experiments also promoted an influx of extracellular calcium and a transient increase in [Ca²⁺]_i prior to the membrane damage (Figures 3, 4), implying an involvement of calcium signal in the photocytotoxicity of fullerene derivatives. However, the details involved need to be further investigated. For example, whether [Ca²⁺]_i decrease resulted from the involvement of an active Ca²⁺ extrusion system or only from the leakage of dye through the damaged membranes remained to verify.

Finally, data from the present experiments also indicated that a potent photocytotoxicity by fullerene derivative could be reached when laser was used as the light source, because combination of a low compound concentration of less than a few micromolar, a short period of less than several minutes, and a low laser power of less than 1 mM were proven to be sufficient for induction of membrane damages (Figures 2, 5). Although no selective uptake by tumor cells has been demonstrated yet, fullerene derivatives have attracted considerable interests in the potential application in tumor photodynamic therapy due to their remarkable photocytotoxic activities (Tagmatarchis and Shinohara 2001). From this experiment, laser is recommended as an ideal candidate of light source engaged for this purpose.