Heterologous expression of a chloroplast outer envelope protein from Suaeda salsa confers oxidative stress tolerance and induces chloroplast aggregation in transgenic Arabidopsis plants

ABSTRACT

Suaeda salsa is a euhalophytic plant that is tolerant to coastal seawater salinity. In this study, we cloned a cDNA encoding an 8.4 kDa chloroplast outer envelope protein (designated as SsOEP8) from S. salsa and characterized its cellular function. Steady‐state transcript levels of SsOEP8 in S. salsa were up‐regulated in response to oxidative stress. Consistently, ectopic expression of SsOEP8 conferred enhanced oxidative stress tolerance in transgenic Bright Yellow 2 (BY‐2) cells and Arabidopsis , in which H₂O₂ content was reduced significantly in leaf cells. Further studies revealed that chloroplasts aggregated to the sides of mesophyll cells in transgenic Arabidopsis leaves, and this event was accompanied by inhibited expression of genes encoding proteins for chloroplast movements such as AtCHUP1, a protein involved in actin‐based chloroplast positioning and movement. Moreover, organization of actin cytoskeleton was found to be altered in transgenic BY‐2 cells. Together, these results suggest that SsOEP8 may play a critical role in oxidative stress tolerance by changing actin cytoskeleton‐dependent chloroplast distribution, which may consequently lead to the suppressed production of reactive oxygen species (ROS) in chloroplasts. One significantly novel aspect of this study is the finding that the small chloroplast envelope protein is involved in oxidative stress tolerance.

INTRODUCTION

Abiotic stresses such as high salinity, drought, high light and extreme temperature, greatly affect plant growth and crop productivity. Although these stresses may elicit different plant responses, a common incident caused by the majority of them is the accumulation of reactive oxygen species (ROS) (Apel & Hirt 2004; Sunkar et al . 2007). Under normal growth conditions, ROS act as important signalling molecules and control a range of biological processes, including plant growth, programmed cell death and stress responses (Mittler et al . 2004). Under stress conditions, however, the overproduction of ROS is highly toxic to cellular processes by causing oxidative damage to nucleic acids, lipid membranes and proteins (Polle 2001). In plant cells, ROS are mainly generated in chloroplasts, mitochondria and microbodies. The Mehler reaction and the triplet excited‐state chlorophyll are major sources of ROS in chloroplasts (Asada & Takahashi 1987; Mittler et al . 2004). The production of ROS is greatly affected by physiological and environmental factors. The ROS generation rate is enhanced when the absorbed light energy exceeds that required for CO₂ assimilation (Long, Humphries & Falkowski 1994; Niyogi 1999; Mittler et al . 2004).

Plants have developed various photoprotective mechanisms to protect cells from photodamage, including morphological responses such as movement of leaves and chloroplasts (Wada, Kagawa & Sato 2003), ROS scavenging (Mittler et al . 2004), and non‐photochemical quenching (Niyogi 1999; Muller, Li & Niyogi 2001). Chloroplasts change their intracellular distribution to regulate their light‐harvesting capacity. The locations of chloroplasts change dynamically in response to ambient light intensity (Wada et al . 2003). In addition, chloroplast aggregation was found to occur in response to various environmental factors such as mechanical stress (Sato, Kadota & Wada 1999; Sato, Wada & Kadota 2003), water stress, ABA treatment (Kondo et al . 2004), salinity (Yamada et al . 2009) and cold stress (Saltveit & Hepler 2004; Tanaka 2007). The cytoskeleton is shown to be crucial for chloroplast distribution, and the actin cytoskeleton is the main mediator of the process in higher plant cells (Takagi 2003; Wada et al . 2003). Chloroplast unusual positioning1 (CHUP1) (Oikawa et al . 2003, 2008; Schmidt von Braun & Schleiff 2008) and some motor proteins including myosin VI (Wang & Pesacreta 2004) and two kinesin‐like proteins (KAC1 and KAC2) (Suetsugu et al . 2011) were identified as critical linkers between the chloroplasts and the actin filaments.

In addition to controlling light absorption, reducing excessively absorbed light energy that results in overproduction of ROS is necessary to prevent photodamage. The ROS scavenging system includes a variety of low‐molecular weight antioxidants including Vitamin B6 (Havaux et al . 2009), ascorbic acid and glutathione (Noctor & Foyer 1998), and antioxidative enzymes, such as superoxide dismutase (SOD), ascorbate peroxidase (APX), glutathione S‐transferase (GST) and catalase (CAT) (Mittler 2002; Mittler et al . 2004; Foyer & Noctor 2005). In the chloroplast, water–water cycle contributes to dissipation of excess photons during oxidative stress by reducing dioxygen to water in photosystem I (PSI) (Asada 1999). In stroma, the ascorbate–glutathione cycle plays a role as the second defence against overproduced ROS (Noctor & Foyer 1998). Triplet excited‐state chlorophyll is one of the major sources of ROS. Non‐photochemical quenching process protects the reaction centres by transfer of excitation energy from antenna chlorophyll to xanthophyll cycle, which can then harmlessly dissipate excitation energy as heat (Muller et al . 2001).

The chloroplast is surrounded by a double‐layer envelope consisting of the outer and inner membranes. The outer envelope proteins (OEPs) were identified and characterized as protein import components, solute channel proteins and proteins involved in lipid metabolism (Inoue 2007). Recent research revealed some other regulation functions of OEPs in plastid division (Miyagishima, Froehlich & Osteryoung 2006), freezing tolerance (Fourrier et al . 2008), organelle movement (Oikawa et al . 2003, 2008; Schmidt von Braun & Schleiff 2008) and sugar signalling processes (Huang et al . 2006). In several extremophile plants, chloroplasts were found to respond to the abiotic stresses by changing their cellular distribution (Kondo et al . 2004). Although these observations suggest an important physiological significance of chloroplast movement in stress response in such plants, very little is known about the mechanisms associated with this cellular event.

In this study, we report the isolation of a cDNA encoding an 8.4 kDa chloroplast OEP (designated as SsOEP8) from Suaeda salsa , a halophytic plant in Northern China that can tolerate coastal seawater salinity. Expression of SsOEP8 was specifically induced by oxidative stress; ectopic overexpression of the gene conferred enhanced oxidative stress tolerance to tobacco Bright Yellow 2 (BY‐2) cells and Arabidopsis plants, in which H₂O₂ levels were considerably decreased; additionally, chloroplast aggregation and altered actin cytoskeleton organization were observed in the transgenic cells. The possible functions of SsOEP8 in stress tolerance and chloroplast distribution are discussed.

MATERIALS AND METHODS

Seedlings of S. salsa (Shandong ecotype, kindly provided by Prof. Hui Zhang, Shandong Normal University) and Arabidopsis thaliana (Columbia ecotype) were grown in a greenhouse at 24 ± 1 °C under 16 h photoperiod and watered weekly with either Hogland or Murashige and Skoog (MS) nutrient solutions, respectively. Light intensity was kept at 100 µ mol photons m⁻²s^−l. For expression pattern analysis, 5‐week‐old seedlings of S. salsa were subject to different abiotic stress treatments, including 400 mm NaCl, 400 mm LiCl, 500 mm mannitol, 10 µ m ABA, 20 µ m methyl viologen (MV), for varying times.

Real‐time PCR analysis

Total RNA was extracted from seedlings of S. salsa and Arabidopsis using a GT (Guanidine thiocyanate) method and RNase‐free DNase I was used to remove any contaminating genomic DNA. First‐strand cDNA was then reverse transcribed using SuperScript^TM III RNase H^– Reverse Trancriptase (Invitrogen, Carlsbad, CA, USA) from 1 µ g total RNA. For determination of the transcript levels of SsOEP8 , AtOEPs (named AtOEP7‐80 ), AtChup1 , AtPhot1 , AtPhot2 , AtPad3 and AtPOX , semi‐quantitative RT‐PCR was performed using the SYBR green PCR kit (TOYOBO, Osaka, Japan) with gene specific primers (Table 1).

Table 1. Primers used in real‐time PCR analyses

Gene name	Primer sequence
SsOEP8	Forward: 5′‐ GCC TGA GAC GAA AAC CGA TGA ‐3′;
	Reverse: 5′‐ GGA AAC AAT AAC CCC TAA CAA ATC A ‐3′
AtOEP7	Forward: 5′‐ GCG ATG GCG TTA GGA TGG TT ‐3′;
	Reverse: 5′‐ CCC TCT TTG GAT GTG GTT GC ‐3′
AtOEP16‐1	Forward: 5′‐ CCA TCT TCC TTT AGG GAC CAA ‐3′;
	Reverse: 5′‐ CAC ATC GGA CAA CTC GTT ATG‐3′
AtOEP16‐2	Forward: 5′‐ ATG ACA GAG GTT CGT GGA GGA G ‐3′;
	Reverse: 5′‐ GAA AGG AGA TTA GCA GCG GTG ‐3′
AtOEP16‐3	Forward: 5′‐ TGG TGC TGG GAC TAT TTA CGG ‐3′;
	Reverse: 5′‐ CGG GAT GCT CCT TGC TCT AT ‐3′
AtOEP37	Forward: 5′‐ TCT CCT TCA CCA ACC CAT CA ‐3′;
	Reverse: 5′‐ CCT CCA TAG CAA CTT CTC CTT G ‐3′
AtOEP80	Forward: 5′‐ TTT CCC AGT GAG AGG ACC G ‐3′;
	Reverse: 5′‐ ATT TAG TTC CGC AGA CCA ACC ‐3′
AtChup	Forward: 5′‐ CGT TAG CAA CTG AAG TCC GA ‐3′;
	Reverse: 5′‐ ACC AAT CAA CTG GAA TCC CGA A ‐3′
AtPhot1	Forward: 5′‐ AGA TGC TGT AAG ATT CTA TGC TG ‐3′;
	Reverse: 5′‐ ACC AAG AGC CCA CCA GTC TA ‐3′
AtPhot2	Forward: 5′‐ AAT CAC GGG TGC TGG TCA TA ‐3′;
	Reverse: 5′‐ CAC TCC ATC ATC TTC CCA CTT T ‐3′
AtPad3	Forward: 5′‐ GTT TAT GCG ATG GGT CGT GAT ‐3′;
	Reverse: 5′‐ GGT GAA GAA CTT GAA AGA AGG C ‐3′
AtPOX	Forward: 5′‐ ACA GTC CGA CCC AAG TAT CG ‐3′;
	Reverse: 5′‐ GAA TCC TCA CAT TTC CCA TCT T ‐3′
Ss18S	Forward: 5′‐ TGC CCA GCA GAT TGA CTA ‐3′;
	Reverse: 5′‐ CAC ACC AAG TAT CGC ATT TC ‐3′
AtEF1a	Forward: 5′‐ TTG GCG GCA CCC TTA GCT GGA TCA ‐3′;
	Reverse: 5′‐ ATG CCC CAG GAC ATC GTG ATT TCA ‐3′

The transcript level of either Ss18S (S . salsa 18S ribosomal RNA gene) or AtEF1a (Arabidopsis translation elongation factor) was used as quantitative control. PCR was conducted according to the following protocol: 30 s denaturation at 94 °C, 30 s annealing at 57 °C, 30 s elongation at 72 °C in 40 cycles. Fluorescence was detected at 80 °C. Samples were analysed in triplicate using independent cDNA samples and were quantified by the comparative cycle threshold method (Wittwer et al . 1997).

BY‐2 cell culture and transformation

Tobacco (Nicotiana tabacum L.) cv BY‐2 cells were maintained and transformed as described previously (An 1985). BY‐2 suspension‐cultured cells were maintained in liquid Linsmaier and Skoog (LS) medium (pH 5.8) at 27 °C in the dark with shaking at 120 r.p.m. Cells were subcultured weekly into fresh medium.

The open reading frame (ORF) of SsOEP8 was amplified by PCR using gene‐specific primers (forward: 5′‐ CGG GAT CCG AGC TCA TGA AGA AAG AAG CAA CA ‐3′; reverse: 5′‐ TCC CCG CGG CCA CCT CTA GAA GAA CGT TGT TGA TCA TC ‐3′). The PCR product was digested with BamH I and Sac II and then fused to the 5′ end of the green fluorescent protein (GFP) coding region. The fused fragment was subcloned into the pPZP111 expression vector (Hajdukiewicz, Svab & Maliga 1994) under the control of the CaMV 35S promoter and the recombinant plasmid was designated as pSsOEP8‐GFP‐PZP.

BY‐2 cells subcultured for 4 d were transformed by Agrobacterium tumefaciens EHA105 containing respective plasmids and screened on LS solid medium containing 200 µ g mL^−l carbenicillin and 50 µ g mL^−l kanamycin. The resulting transformants were transferred onto solid medium containing 200 µ g mL^−l carbenicillin and 100 µ g mL^−l kanamycin, and maintained by monthly subculture.

Subcellular localization of SsOEP8 proteins

Arabidopsis was transformed using A. tumefaciens EHA105 harbouring pSsOEP8‐GFP‐PZP construct by the floral dip method (Clough & Bent 1998). Leaves of the transgenic Arabidopsis plants were visualized by a laser scanning confocal microscope (Leica, Solms, Germany). GFP fluorescence was excited at a wavelength of 488 nm, and chlorophyll fluorescence was excited at 543 nm.

The transgenic BY‐2 suspension cells were fluorescently labelled with 50 nm MitoTraker dye (Invitrogen, USA) for 30 min at 28 °C and washed twice with fresh medium. The labelled BY‐2 cells were visualized by a laser scanning confocal microscope (Leica). The GFP fluorescence was excited at a wavelength of 488 nm, and MitoTracker Red was excited at 543 nm.

Western blot analysis

Plastids of BY‐2 cells and Arabidopsis seedlings were extracted as previously described (Wang et al . 2008). Intact plastids were suspended in lysis buffer (10 mm Tris–HCl/pH 7.6, 50 mm KCl, 10 mm Mg acetate, 7 mm 2‐mercaptoethanol, 2% Triton X‐100) for 30 min and centrifuged at 30 000 g for 30 min. A portion of the supernatants corresponding to 80 µ g proteins was subjected to sodium dodecyl sulphate (SDS)‐polyacrylamide gel electrophoresis (PAGE) separation in a 10% (w/v) acrylamide gel and electro‐blotted onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Anti‐GFP antibodies were used to detect the SsOEP8‐GFP proteins, and anti‐rabbit IgG conjugated with alkaline phosphatase (Promega, Madison, WI, USA) was used as the secondary antibody. Immunoreactive bands were visualized by development with the BCIP/NBT detection system.

Oxidative stress tolerance analysis of transgenic BY‐2 cells and Arabidopsis plants

BY‐2 cells subcultured for 5 d were diluted to 50% (v/v) cell density with LS liquid medium. A 0.2 mL aliquot of the cell suspension was then transferred to 10 mL ddH₂O without or with 100 µ m MV and shaken for 20 h. Fluorescein diacetate (FDA) was used to estimate cell viability, as previously described (Ono et al . 1996; Nakayama et al . 2000). Aliquots of 1 mL of suspension cells were harvested and washed twice with fresh medium. Cells were resuspended in 1 mL medium with 8 µ L 0.5% FDA solution (w/v) in acetone for 20 min in the dark, washed three times and then observed under a fluorescence microscope (Leica) at 488 nm. To compare growth of tobacco plant cells on solid medium, 20 µ L cell aliquots were spotted on LS solid medium or medium containing 30 µ m MV. Growth status was observed 2 weeks later.

cDNA fragment containing the ORF of SsOEP8 was amplified by PCR using gene‐specific primers (forward: 5′‐ CGG GAT CCG AGC TCA TGA AGA AAG AAG CAA CA ‐3′; reverse: 5′‐ TCC AGA TCT TCA AGA ACG TTG TTG ATC ‐3′). The PCR product was digested with BamH I and Bgl II and then fused into plasmid pPZP111, and the resulting recombinant vector (designated as pSsOEP8‐PZP) was introduced into A. tumefaciens EHA105, which was used to transform Arabidopsis . Up to 30 transgenic lines were obtained and three lines with relatively higher levels of transgene expression were selected for further analyses. Five‐day‐old wild‐type and transgenic Arabidopsis seedlings (overexpressing SsOEP8::GFP or SsOEP8 ) grown on MS solid medium were transferred to MS solid medium (control) or medium supplemented with 1 µ m MV and grown vertically for 1 week. The MV‐treated seedlings were then recovered by transferring to MS solid medium and the root tip of each seedling was marked. Relative root elongation, seedling fresh weight and chlorophyll content were measured after recovery. Values of control plants were used as references. Plant chlorophyll concentration was measured following the method as previously described (Porra 2002) and the equations for quantification of the total chlorophyll are as follows:

All of the statistical analysis was performed with SigmaPlot tool version 9.0 (Sytat, USA).

High light treatment and in vivo chlorophyll fluorescence analysis

High light treatment was performed by exposure of 1‐month‐old Arabidopsis plants to an irradiance of 1000 µ mol photons m⁻²s^−l at 23 °C for 24 h. Plants were then transferred to normal conditions (100 µ mol photons m⁻²s^−l, 23 °C) for 12 h. After 15 min of dark adaptation, the photosystem II activity was estimated by determination of photochemical yield (F _v/F _m) using a portable pulse‐modulated fluorometer (PAM‐2000, Walz, Effeltrich, Germany). Measurements were performed on the fully expanded leaves at 0, 6, 24 h after high light treatment and 6, 12 h after recovery.

H₂O₂ detection

One‐week‐old Arabidopsis seedlings germinated on MS medium were transferred to MS medium or medium supplemented with 10 µ m MV for 24 h. The seedlings were then incubated with 1 mg mL⁻¹ 3,3‐Diaminobenzidine (DAB) (dissolved in 50 mm TB buffer, pH 7.6) at room temperature in darkness for 6 h. Chlorophyll was removed with 100% ethanol.

Microscopy analyses of actin structure of BY‐2 cells

Actin filaments were labelled using Alexa‐488 phalloidin as previously described (Seagull 1990). BY‐2 cells subcultured for 5 days were harvested and resuspended in 66 nm Alexa 488‐conjugated phalloidin (Invitrogen, Carlsbad, CA, USA) in phosphate‐buffered saline containing 0.01% NP‐40 for 20 min in the dark. Actin structure was observed with confocal laser microscope (Leica), and the green fluorescence was excited at a wavelength of 488 nm.

RESULTS

Identification and structural characterization of SsOEP8 protein

SsOEP8 was isolated from a cDNA library of S. salsa by functional screening as previously described (Wang et al . 2008). The ORF of SsOPE8 encodes a small protein of 76 amino acids with a calculated molecular weight of 8.4 kDa. Four OEPs from different plant species were identified from a blast search and used for sequence alignment analysis. The result indicated that the SsOEP8 protein shares no less than 50% amino acid identity with each of AtOEP7 (Lee et al . 2001), Pea OEP14 (Li, Moore & Keegstra 1991) and SpOEP6.7 (Salomon et al . 1990) (Fig. 1b). Domain analysis with the TMHMM program (Sonnhammer, von Heijne & Krogh 1998; Krogh et al . 2001) shows that SsOEP8 contains a predicted α ‐helical transmembrane domain within 18 amino acids of the N‐terminus (Fig. 1b).

Figure 1

Sequence analysis of SsOEP8 . (a) Nucleotide and deduced amino acid sequences of the SsOEP8 gene. (b) Sequence alignment of outer envelope proteins (OEPs). The following protein sequences (with their accession numbers) were used for analysis: AtOEP7, Arabidopsis thaliana (NP_190810); Pea OEP14, Pisum sativum (AAA63414); SpOEP6.7, Spinacia oleracea (AAA34035); SsOEP8, Suaeda salsa (HQ660703). The predicted transmembrane domain (TMD) of SsOEP8 is shown in the box.

Expression profiles of SsOEP8 and AtOEPs under oxidative stress

As SsOEP8 gene was isolated while screening for salt tolerance‐related genes, we investigated the effect of salinity, drought and oxidants on the expression of SsOEP8 in S. salsa by real‐time PCR analysis. As seen in Fig. 2a, expression of SsOEP8 transcripts was significantly induced by oxidative stress. Accumulation of SsOEP8 mRNA occurred within 4 h of MV (a strong ROS production agent) treatment and reached the highest level 12 h after MV treatment.

Figure 2

Time course analysis of SsOEP8 and AtOEPs expression in response to oxidative stress. (a) Total RNA was isolated from 5‐week‐old Suaeda salsa treated with 20 µ m methyl viologen (MV; 0, 4, 8 and 16 h). Levels of SsOEP8 and Ss18S (accession number: AY556436) transcripts were measured by SYBR Green I real‐time PCR. (b) Total RNA was isolated from 4‐week‐old Arabidopsis plants treated with 20 µ m MV (0, 4, 8 and 16 h). Gene expression levels were measured using SYBR Green I real‐time PCR. AGI Codes: AtOEP7, AT3g52420; AtOEP16‐1, AT2g28900; AtOEP16‐2, AT4g16160; AtOEP16‐3, AT2g42210; AtOEP37, AT2g43950; AtOEP80, AT5g19620; AtEF1a, AT5g60390. The transcript level of either Ss18S or AtEF1a was used as quantitative control and the 0 h value was used as the reference. These experiments were done with three biological replicates and each with three technical repeats.

Apart from oxidative stress, the level of SsOEP8 mRNA did not show induction by salt, drought and ABA treatments (data not shown), indicating a specific function of SsOEP8 in the response to oxidative stress.

To find out whether expression of the OEP genes in glycophyte plants is also induced by oxidative stress, several OEP genes from Arabidopsis were selected and their expression patterns under MV stress were examined. As compared with SsOEP8 , the expression levels of AtOEPs were slightly induced by oxidative stress except for AtOEP6‐1 that showed a fourfold increase in mRNA amount at the time point of 4 h (Fig. 2b). These results demonstrated that expression of OEP genes responded to oxidative stress in higher plants and moreover, their expressions are enhanced to higher levels in the halophytic plant S . salsa than in the glycophytic plant Arabidopsis .

Subcellular localization of SsOEP8 proteins

The similarity in the amino acid sequence of SsOEP8 with AtOEP7, which is a prominent outer envelope protein, suggested that SsOEP8 may have the same cellular distribution as AtOEP7. To verify this possibility, GFP was fused to the C‐terminal end of SsOEP8 and a plasmid construct containing SsOEP8::GFP under the control of the 35S promoter was introduced into BY‐2 suspension cells, as well as into Arabidopsis plants by Agrobacterium ‐mediated transformation. MitoTraker dye (Invitrogen, USA) was used to label the plastids and mitochondria in BY‐2 cells (Nebenfuhr, Frohlick & Staehelin 2000). As shown in Fig. 3, SsOEP8–GFP fusion proteins were targeted into the membranes of mitochondria and plastids in BY‐2 cells that do not possess chloroplast (Fig. 3a); in leaf cells of Arabidopsis , SsOEP8–GFP fusion proteins were specifically targeted to the envelope membrane of chloroplast (Fig. 3b).

Figure 3

Subcellular localization of SsOEP8. (a) SsOEP8‐GFP fusion proteins in transgenic tobacco BY‐2 cells. MitoTraker dye (Invitrogen, Carlsbad, CA, USA) was used to label plastids and mitochondria in BY‐2 cells (Nebenfuhr et al . 2000). (b) SsOEP8‐GFP fusion proteins in transgenic Arabidopsis leaf cells. Photos were taken with 20× (upper panel), 63× (middle panel) and 100× (lower panel) objectives. Bars = 10 µ m.

Ectopic expression of SsOEP8 confers enhanced oxidative stress tolerance to transgenic BY‐2 cells and Arabidopsis plants

Transgenic tobacco BY‐2 cells expressing SsOEP8::GFP cDNA were generated to investigate the role of SsOEP8 proteins against oxidative stress in plant cells. Approximately 20 independent transgenic lines were generated by Agrobacterium ‐mediated transformation. After verification of SsOEP8 expression in these transgenic cells, two lines (OEP8‐2, OEP8‐4) showing relatively higher expression levels of the transgene were selected for subsequent analysis. As expression of SsOEP8 is responsive to oxidative stress, we tested the oxidative tolerance of the transgenic cells by MV treatment. BY‐2 cells subcultured for 5 d were treated with 100 µ m MV for 20 h and stained with FDA to determine cell viability. As Fig. 4 shows, transgenic BY‐2 cells gained a highly improved tolerance to oxidative stress. Wild‐type and transgenic cells containing empty vector showed little staining with FDA after MV treatment, whereas more than 70% of cells of the two SsOEP8 transgenic lines survived the treatment (Fig. 4b). Such enhanced oxidative stress tolerance of the transgenic cells was clearly seen when cells were grown on solid medium containing 30 µ m MV (Fig. 4c).

Figure 4

Overexpression of SsOEP8‐green fluorescent protein (GFP) improves oxidative stress tolerance in Bright Yellow 2 (BY‐2 cells). (a) Western blot analysis. A sample of 80 µ g of protein was analysed per lane, and anti‐GFP antibodies were used to detect the SsOEP8‐GFP proteins. (b) Fluorescein diacetate (FDA) fluorescence viability test of BY‐2 cells after methyl viologen (MV) treatment (100 µ m) for 20 h. Wild‐type (WT) cells and BY‐2 cells transformed with empty vector were used as controls. (c) WT and transgenic BY‐2 cells (line 4) were grown on the Linsmaier and Skoog (LS) solid medium or medium containing 30 µ m MV (LS + MV) for 2 weeks. These experiments were done with three biological replicates and each with three technical repeats.

To examine the function of SsOEP8 in the whole plant, Arabidopsis plants were transformed with the SsOEP8::GFP construct, and two independent transgenic lines (OEP8‐G‐1 and OEP8‐G‐3) that exhibited different expression levels of the transgene were selected for further analysis (Fig. 5b). Wild‐type and transgenic plants were subjected to oxidative stress treatment with MV. As seen in Fig. 5a, growth of the wild‐type plants was arrested by MV stress; however, transgenic plants showed reduced symptoms of damage after MV treatment as well as a better recovery from the treatment. Between the two transgenic lines, the line with higher expression level of SsOEP8‐GFP showed stronger tolerance to the MV stress. To evaluate this difference in more detail, root elongation, fresh weight, and chlorophyll concentration of wild‐type and transgenic plants were compared after recovery from MV treatment for 3 weeks. For wild‐type plants, the elongation of the main root of stressed plants was less than 5%, their fresh weight was less than 15% and their chlorophyll concentration was no higher than 20% of those values recorded for plants grown on MS medium. Nevertheless, the said values of transgenic lines were significant higher and positively correlated with the expression levels of SsOEP8 proteins (Fig. 5c–e). To exclude the possibility that GFP fused to SsOEP8‐GFP could influence the phenotypes of transgenic plants, the cDNA fragment containing the ORF of SsOEP8 was cloned into pPZP111, and the resulting vector was introduced into Arabidopsis plants by Agrobacterium ‐mediated transformation. Up to 30 transgenic lines were obtained and three lines (OEP‐1, OEP‐5 and OEP‐7) with relatively higher levels of transgene expression (Supporting Information Fig. S1a) were selected for further analyses. In accord with SsOEP8‐GFP overexpression plants, all the transgenic plants overexpressing SsOEP8 alone showed increased tolerance to oxidative stress (Supporting Information Fig. S1b).

Figure 5

Effect of methyl viologen (MV) on wild‐type and transgenic Arabidopsis plants. (a) Five‐day‐old wild‐type (WT) and SsOEP8::GFP transgenic seedlings were transferred to Murashige and Skoog (MS) solid medium or medium supplemented with 1 µ m MV for 1 week and then recovered by transferring them to MS solid medium. Photographs were taken 3 weeks after recovery. (b) Western blot analysis of SsOEP8‐GFP protein levels. A sample of 80 µ g of protein was analysed per lane, and anti‐green fluorescent protein (GFP) antibodies were used to detect the SsOEP8‐GFP proteins. (c–e) Statistical analysis of damage to the MV‐stressed seedling. Plants shown in (a) were used to calculate the relative root elongation (expressed as percentage of root elongation on MS medium) (c), fresh weight (expressed as percentage of fresh weight on MS medium) (d) and chlorophyll content (expressed as percentage of chlorophyll content on MS medium) (e). In (c–e), data are expressed as means of three independent experiments with 30 plants each. Bars indicate SD. * above columns denote significant difference (P  < 0.05).

In addition to oxidative stress caused by MV treatment, the high‐light stress tolerance of the wild‐type and SsOEP8 overexpressing plants was also analysed. As Fig. 6 shows, 12 h after recovery, the leaves of control plants withered and curled badly, while such severe injury was not observed in the leaves of the transgenic plants (OEP‐1, OEP‐5 and OEP‐7). Consistently, the chlorophyll fluorescence was higher in the transgenic plants as compared with the wild‐type control.

Figure 6

Effect of high light stress on wild‐type (WT) and transgenic Arabidopsis plants. (a) WT and SsOEP8 transgenic Arabidopsis plants (OEP‐1, OEP‐5 and OEP‐7) were illuminated at high light (1000 µ mol m⁻²s^−l) for 24 h then recovered under a light intensity of 100 µ mol m⁻²s^−l. Photographs were taken before stress treatment and 12 h after recovery. (b) The PS II activity (F _v/F _m) in the leaves of Arabidopsis plant was measured after dark adaptation for 15 min. The data are the mean value ± SD of three individual experiments.

Together, these results indicate that heteroexpression of SsOEP8 gene from halophytic S. salsa could enhance the ability of the glycophytic plant cells against oxidative stress.

Ectopic overexpression of SsOEP8 was associated with reduced H₂O₂ content

It has been frequently reported that improved oxidative stress tolerance is accompanied by higher ROS scavenging ability. Thus, H₂O₂ levels in Arabidopsis plants were analysed by DAB staining. As shown in Fig. 7, a low level of H₂O₂ was detected in both wild‐type and transgenic plants without stress treatment. When treated with 10 µ m MV, all plants accumulated higher amounts of H₂O_2. However, the levels of H₂O₂ in transgenic plants were much lower than those in wild‐type plants. These results suggest that the overexpression of SsOEP8 enhanced the oxidative tolerance of transgenic plants by reducing the ROS contents.

Figure 7

Effects of methyl viologen (MV) stress on the accumulation of H₂O₂ in the leaves of wild‐type (WT) and SsOEP8::GFP transgenic Arabidopsis plants. One‐week‐old seedlings were transferred to Murashige and Skoog (MS) solid medium (a), or medium supplemented with 10 µ m MV (b) for 24 h. H₂O₂ accumulation in leaves was detected as brown areas. The experiment was repeated more than three times.

Heteroexpression of SsOEP8 led to chloroplast aggregation in Arabidopsis leaf cells

The localization of chloroplasts is influenced by environmental stimuli such as high light (Zurzycki 1957; Park, Chow & Anderson 1996; Kasahara et al . 2002; Wada et al . 2003), mechanical stress (Sato et al . 1999, 2003), water stress (Kondo et al . 2004) and cold stress (Saltveit & Hepler 2004; Kodama et al . 2008). The chloroplast localization of SsOEP8 and the increased high light tolerance of the transgenic plants suggest that SsOEP8 may have a contribution to chloroplast distribution. To pinpoint this possibility, we first conducted the white band assay that was used to test the deficiency in chloroplast movement under high light condition (Oikawa et al . 2003). Figure 8d shows that a clear white band formed in the wild‐type leaf upon high light treatment, but no band was observed with the leaf of the transgenic plant under the same condition, indicating that a variation of chloroplast distribution occurred in the mesophyll cells of the transgenic plants. To verify this result, we next carried out the histological analysis with the leaves of the wild‐type and transgenic plants. As Fig. 8 shows, in wild‐type plant chloroplasts were distributed along the edge of mesophyll cells under normal light intensity (100 µ mol photons m⁻²s^−l). In contrast, most of the chloroplasts aggregated at the side of mesophyll cells in the transgenic plant (Fig. 8a,b). The phenomenon of chloroplast aggregation was also found in mesophyll cells of SsOEP8 transgenic lines (Fig. 8c), indicating that the chloroplast clumping was not caused by GFP–GFP interaction.

Figure 8

Microscopic analysis of chloroplast distribution in Arabidopsis leaf cells and actin structure in Bright Yellow 2 (BY‐2) cells. Mesophyll cells (a) and cross‐sections (b) of wild‐type (WT) and SsOEP8::GFP transgenic plant leaves under low light intensity. (c) Mesophyll cells of WT and SsOEP8 transgenic plants under low light intensity. (d) White band assay of WT and SsOEP8 transgenic plants. Photo was taken 20 h after high light treatment. (e) Actin cytoskeleton in mesophyll cells of the WT and transgenic BY‐2 cells. Bars = 30 µ m.

Expression of genes involved in chloroplast movement and oxidative stress in transgenic plants

Previous studies revealed that CHUP1 (Oikawa et al . 2003, 2008; Schmidt von Braun & Schleiff 2008), PHOT1 (Jarillo et al . 2001; Kagawa et al . 2001; Sakai et al . 2001) and PHOT2 (Sakai et al . 2001) play important roles in chloroplast movement. To see if chloroplast aggregation in transgenic cells was associated with expression changes in these genes, real‐time PCR analysis was conducted. Results indicated that transcription of these genes was significantly down‐regulated in transgenic plants and negatively correlated with the expression of SsOEP8 (Fig. 9). These results implicate that the aggregation of chloroplast in transgenic plants may be a result of the suppressed expression of the genes involved in chloroplast movement.

Figure 9

Genes identified to be differentially expressed in wild‐type (WT) and transgenic Arabidopsis plants. Total RNA was isolated from 4‐week‐old plants. Gene expression levels were measured using SYBR Green I real‐time PCR. AGI Codes: chup1 , At3g25690; phot1, AT3G45780; phot2, AT5G58140; pad3, AT3g26830; pox, AT5g19880. The transcript level of AtEF1a was used as quantitative control and the WT value was used as the reference. These experiments were done with three biological replicates and each with three technical repeats.

Pad3 (encoding a cytochrome P450 monooxygenase) and POX (encoding a putative peroxidase) are two oxidative stress‐related genes in Arabidopsis reported to be up‐regulated in chup1 mutants (Oikawa et al . 2003). Interestingly, we observed that the expression levels of these two genes were also increased in SsOEP8 transgenic plants (Fig. 9) in which chup1 expression was down‐regulated, revealing a link between the aggregation of chloroplasts and increases in oxidative stress tolerance.

Ectopic overexpression of SsOEP8 resulted in altered actin cytoskeleton organization in transgenic BY‐2 cells

Actin filaments are critical for chloroplast movement and positioning (Kandasamy & Meagher 1999; Oikawa et al . 2003; Schmidt von Braun & Schleiff 2008). The aggregation of chloroplasts in transgenic Arabidopsis leaf cells prompted us to investigate if SsOEP8 overexpression had an effect on actin cytoskeleton. For better visualization of actin filaments, we chose transgenic BY‐2 cells for the analysis. Wild‐type and transgenic BY‐2 cells were labelled using phalloidin. Figure 8e shows that the actin architecture and distribution varied between the two types of cells. In wild‐type BY‐2 cells, actin filaments formed a complicated network consisting of actin arrays and thick bundles which orientated either obliquely or vertical to the axis of cell elongation in the cortical region; however, actin filaments appeared denser, and many thinner and shorter actin cables were observed in transgenic cells.

DISCUSSION

SsOEP8 is localized on the outer envelope of chloroplast

The amino acid sequence of SsOEP8 showed 58 and 53% amino acid sequence identity with OEP7 of Arabidopsis (Lee et al . 2001) and OEP6.7 of spinach (Salomon et al . 1990), which are considered as models to study the mechanism of protein insertion into the chloroplast outer envelope. Many chloroplast proteins are encoded by the nuclear genome and contain N‐terminal transit peptides, which are removed in chloroplast after translocation. In contrast, most proteins of the chloroplast outer envelope do not have cleavable transit sequences and have a different import pathway (Soll & Tien 1998). A seven–amino acid region in the C‐terminal region and the transmembrane domain (TMD) in N‐terminal were reported to be determinants for AtOEP7 to target into the chloroplast outer envelope membrane (Lee et al . 2001). Several topology prediction programmes, such as TMHMM (Sonnhammer et al . 1998; Krogh et al . 2001) and chloroplast transit peptides prediction programmes such as ChloroP (Emanuelsson, Nielsen & Von Heijne 1999), were used to examine the possible membrane topology of SsOEP8. Results showed that SsOEP8 contains a putative TMD near the N‐terminus at positions 15–32, and no typical transit peptide was found. GFP microscopy analysis showed that SsOEP8‐GFP fusion proteins were targeted to the chloroplast envelope in transgenic Arabidopsis leaf cells, and the localization pattern was very similar to that of known Arabidopsis OEPs such as AtOEP7 (Lee et al . 2001), AtToc64 (Lee et al . 2004) and CHUP1 (Oikawa et al . 2008). Therefore, these data demonstrate that SsOEP8 is a chloroplast OEP.

SsOEP8 is involved in chloroplast aggregation

Chloroplasts change their intracellular distribution in response to environmental factors to adjust energy absorption and protect cells from photodamage (Wada et al . 2003; Kondo et al . 2004; Saltveit & Hepler 2004; Tanaka 2007; Yamada et al . 2009). Kondo et al . (2004) demonstrated chloroplast aggregation in the leaves of succulent plants under combined light and water stress conditions and ABA treatment. It was proposed that such an unusual distribution of chloroplasts is a particular morphological strategy used by succulent plants to protect themselves against light stress in combination with other stresses (Kondo et al . 2004). In this study, we observed that chloroplast aggregation occurred in the transgenic plants, and this event was correlated with the expression levels of SsOEP8 . In the transgenic line OEP‐G‐1 that expressed higher level of SsOEP::GFP , dense chloroplast clumping was found. On the other hand, the chloroplasts in leaf cells of transgenic line OEP‐G‐3, which expressed lower level of SsOEP::GFP , were less aggregated (data not shown). As S . salsa possesses succulent leaves, such an incident of chloroplast aggregation may reflect the physiological role of the SsOEP8 in this expremophile plant. Consistence with this notion, we found that expression of SsOEP8 was remarkably induced by MV stress (Fig. 2a) but expression of most Arabidopsis OEP genes was only slightly enhanced under same condition (Fig. 2b). It should be mentioned that the phenomenon of chloroplast aggregation was also found in mesophyll cells of SsOEP8 transgenic lines, indicating that the unusual chloroplast distribution in SsOEP8‐GFP transgenic plants was a results of SsOEP8 overexpression, other than GFP‐GFP interactions.

Oikawa et al . (2008) reported that overexpression of the N‐terminal hydrophobic region (NtHR) of CHUP1 fused with GFP (NtHR‐GFP) and chloroplast OEP 7‐GFP induced a chup1 ‐like phenotype. The authors proposed that the GFP proteins may occupy the surface of the outer envelope and replace the existing CHUP1, resulting in an aggregation of chloroplasts. Based on this speculation, SsOEP8 proteins may compete with CHUP1 proteins for the space on the outer envelope, and some of the CHUP1 proteins may be displaced from the surface. Furthermore, the expression of CHUP1 , shot1 and shot2 , which were reported to play critical roles in chloroplast movement, was down‐regulated in SsOEP8::GFP transgenic plants. Decreased expression of CHUP1 could be due to reduced CHUP1 abundance on the outer envelope or increase in the cytoplasm, which might consequently lead to suppression of CHUP1 gene expression as a feedback/coordinate response. Likewise, suppressed expression of phot1 and phot2 may also result from a regulatory event for the coordinated expression of proteins responsible for chloroplast movement.

Overexpression of SsOEP8 conferred increased oxidative tolerance in transgenic BY‐2 cells that do not contain chloroplasts. SsOEP8 may exert its function in other subcellular organelles in these cells. It is demonstrated that a plant chloroplast‐localized protein, AtGRXcp, was targeted to the mitochondria when expressed in yeast (Cheng et al . 2006). We observed that SsOEP8‐GFP fluorescence was co‐localized in BY‐2 cells with the MitoTracker Red which is used as a probe for mitochondria and plastids (Nebenfuhr et al . 2000). As mitochondrion is an important cellular compartment involved in ROS production and scavenging and plastids are the precursors of chloroplasts, it is possible that SsOEP8 rendered the enhanced oxidative tolerance to the transgenic BY‐2 cells by reducing ROS generation in these cellular localizations.

Possible link between oxidative stress tolerance and chloroplast aggregation involving SsOEP8

The intracellular concentration of ROS is determined by both the ROS‐generation system and the ROS‐scavenging system. Chloroplast is one of the major sources of ROS production in plant cells. Chloroplasts change their intracellular distribution in response to the environmental factors to adjust the energy absorption and control the generation of ROS (Wada et al . 2003; Kondo et al . 2004; Saltveit & Hepler 2004; Tanaka 2007; Yamada et al . 2009). In this study, we detected an inverse correlation between SsOEP8 expression and H₂O₂ abundance in the leaves of transgenic plants under oxidative stress. To understand the mechanism by which overexpression of SsOEP8 affected the ROS levels in the transgenic plants, we detected the ROS‐scavenging ability of SsOEP8 protein but no enzyme activity was found. Then we compared the activities of several ROS‐scavenging enzymes, including SOD, APX, POD, CAT, GR and NDH, but no significant difference in enzyme activities was detected between the wild‐type and transgenic plants either (data not shown). Although two oxidative stress‐related genes were up‐regulated in transgenic plants, it is more likely that the improved oxidative stress tolerance of the transgenic plants was due to the reduction in ROS generation rather than enhanced ROS scavenging.

Functional association of SsOEP8 with actin cytoskeleton

Actin filaments were shown to be crucial for chloroplast distribution (Wada et al . 2003). Chloroplasts migrate along the longitudinally thick actin cable by using motor molecules or fine microfilaments attached to the cables. Partly or fully disrupted actin filaments showed different degrees of chloroplast aggregation (Kandasamy & Meagher 1999; Oikawa et al . 2003; Schmidt von Braun & Schleiff 2008). Compared with wild‐type cells, the majority of actin filaments are thin and transverse the cell's long axis in transgenic cells and the vertical arrays of thick actin cables almost disappear in these cells. These results indicate that aggregation of chloroplasts in transgenic plants might be due to changes in actin cytoskeleton organization. The actin binding activity of SsOEP8 was therefore analysed to further investigate the possible interaction between SsOEP8 and the actin cytoskeleton. Results showed that SsOEP8 could not bind to actin filament (Supporting Information Fig. S2). Base on these observations and the facts that (1) CHUP1 can interact with actin and regulate actin polymerization (Schmidt von Braun & Schleiff 2008); (2) the physiological phenotypes of SsOEP8 overexpressing lines are very similar with CHUP1 mutant; and (3) the expression level of CHUP1 in SsOEP8 overexpressing lines are lower than those in wild‐type plants, we suggest that SsOEP8 may indirectly influence actin cytoskeleton organization by affecting other actin‐binding components such as CHUP1. Further investigations on the relationships of SsOEP8 and cytoskeleton organization must be carried out to confirm this speculation.

ACKNOWLEDGMENTS

This work was supported by grants from the National Natural Science Foundation (Grant No. 30771162) and the Ministry of Agriculture of China for transgenic research (Grant No. 2009ZX08009‐096B).