The Mitochondrial Fission Regulator DRP3B Does Not Regulate Cell Death in Plants

Abstract

• Background and Aims Recent reports have described dramatic alterations in mitochondrial morphology during metazoan apoptosis. A dynamin-related protein (DRP) associated with mitochondrial outer membrane fission is known to be involved in the regulation of apoptosis. This study analysed the relationship between mitochondrial fission and regulation of plant cell death.

• Methods Transgenic plants were generated possessing Arabidopsis DRP3B (K56A), the dominant-negative form of Arabidopsis DRP, mitochondrial-targeted green fluorescent protein and mouse Bax.

• Key Results Arabidopsis plants over-expressing DRP3B (K56A) exhibited long tubular mitochondria. In these plants, mitochondria appeared as a string-of-beads during cell death. This indicates that DRP3B (K56A) prevented mitochondrial fission during plant cell death. However, in contrast to results for mammalian cells and yeast, Bax-induced cell death was not inhibited in DRP3B (K56A)-expressing plant cells. Similarly, hydrogen peroxide-, menadione-, darkness- and salicylic acid-induced cell death was not inhibited by DRP3B (K56A) expression.

• Conclusions These results indicate that the systems controlling cell death in animals and plants are not common in terms of mitochondrial fission.

INTRODUCTION

In healthy mammalian cells, fusion and fission events participate in regulating mitochondrial morphology (Shaw and Nunnari, 2002). Proteins controlling mitochondrial fission include Drp1 (DNM1 in yeast) and Fis1. Drp1 is a large GTPase similar to dynamin. Inhibition of Drp1 GTPase activity by a dominant negative protein defective in GTP binding (Drp1^K38A) results in elongated mitochondria (Pitts et al., 1999). Similarly, in Arabidopsis leaves and tobacco suspension cultured BY-2 cells, over-expression of a dominant-negative mutant of an Arabidopsis dynamin-related protein [DRP3B (K56A); DRP3B is also known as ADL2b] caused tubulation of mitochondria (Arimura and Tutsumi, 2002). Mitochondrial fusion in mammalian cells is controlled by the large GTPases Mfn1, Mfn2, Fzo1 and Opa1. Several reports have described dramatic alterations in mitochondrial morphology during the early stages of apoptotic cell death (Desagher and Martinou, 2000; Karbowski and Youle, 2003; Oakes and Korsmeyer, 2004). It was later reported that multiple components of the mitochondrial morphologenesis machinery can positively and negatively regulate apoptosis (Fannjiang et al., 2004; Lee et al., 2004). When mitochondrial fission is blocked by over-expression of a dominant negative mutant of Drp1 (Drp1^K38A), the downstream events of mitochondria-induced cell death, such as cytochrome c release, are blocked (Breckenridge et al., 2003). Moreover, down-regulation of mammalian Fis1 expression by RNA interference (RNAi) induces mitochondrial elongation and inhibits cell death induced by staurosporine, etoposide and other stimuli (Lee et al., 2004). By contrast, the down-regulation of Opa1 and Mfn1/2 in mammalian cells by RNAi leads to fragmentation of mitochondria, and Opa1 RNAi cells are more sensitive to apoptosis induced by various stimuli (Lee et al., 2004; Sugioka et al., 2004). In addition, over-expression of Fzo1 increases mitochondrial elongation and inhibits cytochrome c release and cell death (Sugioka et al., 2004). Thus, proteins that comprise the mitochondrial morphogenesis machinery can regulate programmed cell death.

It was recently demonstrated that various reactive oxygen species (ROS) and mammalian pro-apoptotic gene Bax treatments induced mitochondrial fragmentation in Arabidopsis thaliana (Yoshinaga et al., 2005a, b). Mitochondria showed morphological changes from a bacillus-like to a rounded shape, and their overall size halved. This indicates that mitochondrial fission occurred during plant cell death. Frank et al. (2001) demonstrated that over-expression of a dominant-negative mutant of Drp1 inhibited Bax-induced cell death in COS-7 cells. In the present study, transgenic Arabidopsis plants were generated possessing dominant-negative DRP3B (K56A), mitochondrial-targeted green fluorescent protein (mt-GFP) and dexamethasone (DEX)-inducible Bax. Using these transgenic plants, it was shown that over-expression of DRP3B (K56A) protein could not inhibit ROS- and Bax-induced cell death in plants, unlike for mammalian cells.

MATERIALS AND METHODS

Plant materials

The DEX-inducible mouse Bax transgenic line (Arabidopsis thaliana ecotype Col-0) was as described previously in Kawai-Yamada et al. (2001). To generate transgenic plants possessing the Cauliflower Mosaic Virus (CaMV) 35S promoter with mt-GFP and DRP3B (K56A), the plasmid described in Arimura and Tsutsumi (2002) was introduced into Arabidopsis thaliana ecotype Col-0 via Agrobacterium-mediated transformation. In this plant, mt-GFP and DRP3B (K56A) were co-expressed under CaMV 35S promoters. The Arabidopsis plant possessing mt-GFP was kindly provided by Dr Yasuo Niwa (Graduate School of Nutritional and Environmental Sciences, University of Shizuoka) (Niwa et al., 1999). Triple transgenic plants possessing DRP3B (K56A), mt-GFP and Bax were generated using pollen from the Bax transgenic plants to fertilize the flowers of mt-GFP- and DRP3B (K56A)-possessing plants. The F₂ population that grew on a medium containing hygromycin (20 µg mL^–1) was used for the present analysis. Transgenic plants were grown at 23 °C under continuous light conditions (60 µmol m^–2 s^–1).

Chemicals

Hydrogen peroxide (H₂O₂; Wako, Osaka, Japan) was diluted with distilled water to 50 or 100 mm. DEX (Sigma, St Louis, MO, USA), menadione (MD; Sigma) and salicylic acid (SA; nacalai tesque, Kyoto, Japan) were dissolved in dimethyl sulfoxide (DMSO). During treatment, the final DMSO concentration was never higher than 0·2 %, which had no effect on Arabidopsis leaves. In order to induce cell death, 10 μm DEX, 25 or 60 μm MD, and 300 or 400 μm SA were used.

Ion leakage measurement

Three leaf discs obtained from 3-week-old Arabidopsis plants were floated on distilled water. After addition of various chemicals, leaf discs were vacuumed for 5 min. The discs were incubated at 23 °C under continuous light (60 µmol m^–2 s^–1) without shaking. Electrolyte leakage was monitored using an electrical conductivity meter (Horiba, B-173, Kyoto, Japan) (Kawai-Yamada et al., 2004). Measured electrical conductivities of the medium were reported in μS cm^–1. This experiment was run in triplicate.

Microscopic observation

GFP fluorescence was examined at an excitation wavelength of 488 nm under a fluorescence microscope (DMRD, Leica, Wetzlar, Germany). In all experiments, epidermal cells of leaves were observed.

RESULTS

Bax-induced cell death in transgenic Arabidopsis with the dominant-negative mutant DRP3B (K56A)

To analyse the mechanisms of plant cell death in terms of mitochondrial regulation, an Arabidopsis homologue of Drp1 (DRP3B) was investigated using a Bax-induced plant cell death system (Kawai-Yamada et al., 2001; Yoshinaga et al., 2005a). Transgenic plants were generated possessing dominant-negative DRP3B (K56A) (Arimura and Tsutusmi, 2002), mt-GFP (Niwa et al., 1999) and Bax [DEX-inducible (Kawai-Yamada et al., 2001)], and the morphological changes observed in mitochondria were investigated.

As shown in Fig. 1A, expression of mt-GFP and dominant-negative DRP3B (K56A) protein in whole plants did not affect plant growth. Fluorescence microscopy revealed that mitochondria were longer in transgenic plants expressing mutant DRP3B (K56A) protein (Fig. 1B, DEX-).

Fig. 1.

Phenotype of triple transgenic plants possessing DRP3B (K56A), mt-GFP and Bax. (A) Phenotypes of transgenic lines. All plants were 3-week-old seedlings grown under continuous light without Bax expression. (B) Morphology of mitochondria with or without Bax expression. Leaves obtained from 3-week-old seedlings were floated on distilled water with 0 (DEX-) or 10 μm dexamethasone (DEX+) for 1 d to induce Bax expression. Magnified images of DEX-treated leaves are shown at the bottom. To observe mt-GFP, epidermal cells were examined using a fluorescence microscope. Arrows indicate the clearly visible sites of mitochondrial connections.

It was recently reported that mammalian Bax could induce plant cell death (Kawai-Yamada et al., 2001; Yoshinaga et al., 2005a), which was accompanied by morphological changes in mitochondria (Yoshinaga et al., 2005a). As reported previously, punctiform mitochondria were observed in Bax-expressing plant cells (Fig. 1B, DEX+). However, in DRP3B (K56A)/mt-GFP/Bax plants, punctiform mitochondria were connected as a string-of-beads (Fig.1B, DEX+, arrow), although each mitochondrion was individually scattered in mt-GFP/Bax plants. This indicates that DRP3B (K56A) inhibits mitochondrial fission during Bax-induced plant cell death. However, etiolation was observed within 3 d after 10 μm DEX treatment in DRP3B (K56A)/mt-GFP/Bax plants as well as in the mt-GFP/Bax line (Fig. 2A). Ion leakage was measured to evaluate the rate of cell death (see Mitsuhara et al., 1999). As shown in Fig. 2B, mt-GFP/Bax and DRP3B (K56A)/mt-GFP/Bax plants exhibited increasing electrolyte leakage, although leakage was not noted in mt-GFP and DRP3B (K56A)/mt-GFP plants. A Bax transgenic plant without DEX treatment showed no clear ion leakage (data not shown).

Fig. 2.

Bax-induced cell death in dominant-negative mutant DRP3B (K56A) plants. (A) Leaf discs at 0, 3 and 5 d after dexamethasone (DEX) treatment. Three leaf discs obtained from wild-type (WT), mt-GFP/Bax and DRP3B (K56A)/mt-GFP/Bax plants were submerged in 10 μm DEX solution and incubated at 23 °C under continuous light. Leaf discs without DEX treatment were used as a control (Control). (B) Evaluation of ion leakage in transgenic Arabidopsis leaves after Bax expression. Electrolyte leakage from leaf discs prepared from 3-week-old seedlings was measured after 0, 1, 2 and 3 d in 10 μm DEX. Data shown are means ± s.e. of experiments run in triplicate.

Mitochondrial changes in DRP3B (K56A) plants under several stresses

To confirm whether similar mitochondrial changes are induced under other conditions causing plant cell death, DRP3B (K56A) plants were subjected to several stresses. After exposure of DRP3B (K56A) leaves to stresses such as 25 μm MD, 50 mm H₂O₂, 400 μm SA and dark treatment (Dark), mitochondia changed from tubular to a string-of-beads form (Fig. 3A). All mitochondria aligned together (Fig. 3A, arrows). These changes were also observed when Bax was expressed in DRP3B (K56A) plants (Fig. 1B). Moreover, ion leakage was observed after dark treatment (6 d), 60 μm MD, 100 μm H₂O₂ and 300 μm SA treatment in DRP3B (K56A) plants. No clear distinction was observed between mt-GFP and DRP3B (K56A)/mt-GFP plants (Fig. 3B).

Fig. 3.

Mitochondrial changes under various stresses in DRP3B (K56A) plants. (A) Morphological changes in mitochondria under reactive oxygen species (ROS) stress. Epidermal cells of leaf discs obtained from transgenic Arabidopsis [DRP3B (K56A)/mt-GFP] plants were examined under a fluorescence microscope. Leaf discs were treated with 25 μm menadione (MD), 50 mm hydrogen peroxide (H₂O₂) or 400 μm salicylic acid (SA) and incubated for 1 d at 23 °C under continuous light. For dark treatment (Dark), leaf discs floated on distilled water were kept under continuous dark for 4 d. Arrows indicate the clearly visible sites of mitochondrial connections. Leaf discs treated with distilled water are shown as a control. Elongated mitochondria were observed in control cells. (B) Ion leakage from stressed leaves of DRP3B (K56A) plants. Three leaf discs prepared from 3-week-old plants were treated with 6 d of darkness (Dark), distilled water (control), 60 μm MD, 50 mm H₂O₂ or 300 μm SA for 1 d. mt-GFP plants were used as a control.

DISCUSSION

In mammalian cells, various key apoptotic events involve mitochondria, including the release of cytochrome c and loss of mitochondrial transmembrane potential (Green and Reed, 1998). In plants, mitochondria are also important regulators of ROS generation and cellular ATP levels (Maxwell et al., 2002; Yao et al., 2002). Recently, mitochondrial fragmentation during apoptosis has been described (Desagher and Martinou, 2000; Karbowski and Youle, 2003). Moreover, mitochodrial fragmentation has been demonstrated during the early stages of ROS- and Bax-induced plant cell death, suggesting the possibility of mitochondrial fission during plant cell death (Yoshinaga et al., 2005b). In addition, it has been reported that cell death was inhibited or promoted by artificially controlling the protein participating in mitochondrial fission and fusion in mammalian cells and yeast (Frank et al., 2001; Breckenridge et al., 2003; Fannjiang et al., 2004). The down-regulation of human Fis1 (hFis1) causes extensive mitochondrial fusion, and the cells become resistant to apoptosis. By contrast, down-regulation of Opa1 causes extensive mitochondrial fragmentation, and cells become more sensitive to death. Thus, the components of the mitochondrial fission–fusion machinery can positively and negatively regulate apoptosis. Likewise, in plants, low oxygen pressure induces elongation of mitochondria (Van Gestel and Verbelen, 2002). This indicates a correlation between stress and mitochondrial morphology.

Little is currently known about higher plant mitochondrial dynamics. However, recently, several genes involved in the control of plant mitochondrial dynamics have been identified (Logan et al., 2003, 2004; Sheahan et al., 2005). Arimura and Tsutsumi (2002) revealed that DRP3B, an Arabidopsis homologue of Drp1, controlled mitochondrial fission in Arabidopsis leaves and tobacco BY-2 suspension cultures. Moreover, DRP3B (K56A) is a dominant-negative protein, and plant cells expressing DRP3B (K56A) exhibit longer but fewer mitochondria. Punctate morphology of mitochondria has been shown to be one of the indicators in the early stage of ROS stress-induced cell death (Yoshinaga et al., 2005b). Using DRP3B dominant-negative plants, mitochondrial fission was shown to be incomplete, and mitochondria aligned in a string-of-beads form under ROS stress- and Bax-induced cell death. DRP3B (K56A) blocks fission of the outer mitochondrial membrane by reducing the affinity for GTP to DRP3B (Pitts et al., 1999; Arimura and Tutsumi, 2002). This indicates that mitochondrial fragmentation during plant cell death is not simply collapse, as controlled by DRP3B.

Frank et al. (2001) demonstrated that inhibition of Drp1 by over-expression of a dominant-negative mutant blocked Bax-induced cell death in COS-7 cells. However, in the present study, cell death was not inhibited in transgenic plants over-expressing DRP3B (K56A), which differs from observations in mammals and yeast (Frank et al., 2001; Fannjiang et al., 2004). The function of the mitochondrial fission regulator during cell death is different between animals and plants.