The Plant Journal (2003) 35, 332±341
doi: 10.1046/j.1365-313X.2003.01809.x
Transgenic tobacco plants expressing antisense ferredoxin-NADP(H) reductase transcripts display increased susceptibility to photo-oxidative damage Javier F. Palatnik1,y, Vanesa B. Tognetti1, Hugo O. Poli1, Ramiro E. RodrõÂguez1, NicolaÂs Blanco1, Martha Gattuso2, Mohammad-Reza Hajirezaei3, Uwe Sonnewald3, Estela M. Valle1 and NeÂstor Carrillo1, 1 Instituto de BiologõÂa Molecular y Celular de Rosario (IBR), Universidad Nacional de Rosario, Suipacha 531, S2002-LRK Rosario, Argentina, 2 Area BiologõÂa Vegetal, Facultad de Ciencias BioquõÂmicas y FarmaceÂuticas, Universidad Nacional de Rosario, Suipacha 531, S2002-LRK Rosario, Argentina, and 3 Institut fuÈr P¯anzengenetik und Kulturp¯anzenforschung, Corrensstrasse 3, 06466 Gatersleben, Germany Received 11 October 2002; revised 29 April 2003; accepted 7 May 2003. For correspondence (fax 54 341 439 0465; e-mail
[email protected]). y Present address: Department of Molecular Biology, Max Planck Institute for Developmental Biology, Spemannstrasse 37-39, D-72076 TuÈbingen, Germany.
Summary Ferredoxin-NADP(H) reductase (FNR) catalyses the ®nal step of the photosynthetic electron transport in chloroplasts. Using an antisense RNA strategy to reduce expression of this ¯avoenzyme in transgenic tobacco plants, it has been demonstrated that FNR mediates a rate-limiting step of photosynthesis under both limiting and saturating light conditions. Here, we show that these FNR-de®cient plants are abnormally prone to photo-oxidative injury. When grown under autotrophic conditions for 3 weeks, specimens with 20±40% extant reductase undergo leaf bleaching, lipid peroxidation and membrane damage. The magnitude of the effect was proportional to the light intensity and to the extent of FNR depletion, and was accompanied by morphological changes involving accumulation of aberrant plastids with defective thylakoid stacking. Damage was initially con®ned to chloroplast membranes, whereas Rubisco and other stromal proteins began to decline only after several weeks of autotrophic growth, paralleled by partial recovery of NADPH levels. Exposure of the transgenic plants to moderately high irradiation resulted in rapid loss of photosynthetic capacity and accumulation of singlet oxygen in leaves. The collected results suggest that the extensive photo-oxidative damage sustained by plants impaired in FNR expression was caused by singlet oxygen building up to toxic levels in these tissues, as a direct consequence of the over-reduction of the electron transport chain in FNR-de®cient chloroplasts. Keywords: ferredoxin-NADP(H) reductase, antisense, transgenic tobacco plants, oxidative stress, photoinhibition.
Introduction Ferredoxin-NADP(H) reductases (FNR, EC 1.18.1.2) constitute a functional group of ¯avoproteins that catalyse the reversible electron transfer between two molecules of obligatory one-electron carriers such as ferredoxin (Fd) and a single molecule of NADP(H) (Shin and Arnon, 1965). 2 Fd:Fe2 NADP H ! 2 Fd:Fe3 NADPH
1
Flavoenzymes displaying FNR activity can be found in plastids, mitochondria and bacteria (Arakaki et al., 1997). Even though the plant-type and mitochondria-type FNR isoforms are structurally unrelated and apparently evolved 332
from independent origins, all FNR proteins possess some common features. They are hydrophilic monomers of 27±42 kDa made up of FAD- and NADP(H)-binding domains that can split electrons between obligatory one- and twoelectron carriers (Bruns and Karplus, 1995; Deng et al., 1999; Morales et al., 2000; Ziegler and Schulz, 2000). In heterotrophic tissues, the reaction depicted in Equation 1 is driven toward Fd reduction, providing low-potential redox equivalents for metabolic pathways as diverse as steroid hydroxylation, nitrate reduction, hydrogen and nitrogen ®xation, methane oxidation and fatty acid ß 2003 Blackwell Publishing Ltd
Photoinhibition in FNR-deficient tobacco plants desaturation (Carrillo, 1998; Fillat et al., 1995; Ziegler and Schulz, 2000). In chloroplasts and in vegetative cells of cyanobacteria, where Fd is reduced by a photochemical reaction at the level of photosystem I, FNR mediates electron transfer to NADP, providing the NADPH required for CO2 ®xation and other biosynthetic pathways (Shin and Arnon, 1965). Different modes of membrane association regulate FNR activity in vivo, as changes in the ratio of loosely and tightly bound enzyme correlate with NADP photoreducing activity and affect FNR kinetic parameters (Carrillo and Vallejos, 1987). Furthermore, oxidative stress causes FNR solubilisation from the thylakoid membranes, thus favouring the NADPH-oxidising activities of the reductase (Palatnik et al., 1997). Until recently, it has therefore been dif®cult to assess the relative contribution of FNR to electron transport and photosynthesis. To address this question, transgenic tobacco plants expressing an FNR-speci®c antisense transcript were constructed (Hajirezaei et al., 2002). Analysis of these transformants revealed that FNR activity becomes limiting for photosynthesis under both limiting and saturating illumination, suggesting that the amount of FNR constrains photosynthesis under normal growth and might possibly restrict biomass production (Hajirezaei et al., 2002). In the current paper, we extend our previous characterisation of FNR de®ciency in transgenic plants. In addition to the predictable impairment of photosynthesis (Hajirezaei et al., 2002), FNR antisense plants were extremely sensitive to light and displayed symptoms of oxidative stress even under moderate irradiation. The extent of damage correlated with both the light intensity and the degree of FNR depletion. We report here that FNR-de®cient transgenic lines were abnormally prone to photoinactivation, presumably resulting from over-reduction of the photosynthetic electron transport chain and accumulation of singlet oxygen in illuminated leaves. Chloroplast pigments and membranes were early targets of photoinhibition, whereas stromal proteins were affected after several weeks of autotrophic growth and only in the most FNR-depleted lines.
Results and discussion Preparation of clonally propagated tobacco plants deficient in chloroplast ferredoxin-NADP(H) reductase Tobacco plants transformed with an FNR antisense DNA displayed decreased levels of the ¯avoenzyme (Hajirezaei et al., 2002). For the present studies, we selected specimens displaying FNR amounts ranging from 15 to 45% of those found in the parental lines (Figure 1a; Table 1) and propagated them clonally. In all cases, the residual levels of the suppressed ¯avoprotein showed a good correlation with ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
333
Figure 1. Phenotypes of transformed tobacco plants de®cient in ferredoxinNADP(H) reductase (FNR). (a) FNR levels in leaves of wild-type (wt) or antisense (as) transgenic lines. Leaf extracts from 3-week-old plants were fractionated by SDS±PAGE and blotted onto nitrocellulose membranes for immunodetection of FNR. Twenty micrograms of protein was loaded on each lane. The electrophoretic mobility of FNR is indicated by an arrow. (b) Development of control (wt) and transformed tobacco lines grown under autotrophic conditions. Tissue-culture-derived seedlings were rooted in soil and grown for additional 8 weeks at 200 mmol quanta m 2 sec 1 with a 16-h photoperiod. The inset illustrates the impairment of leaf development in FNR-de®cient plants. Leaves were taken from the third node of the plants depicted in panel (b).
Table 1 Ferredoxin-NADP(H) reductase (FNR) and NADP(H) levels in FNR antisense tobacco plants FNR level (%)
NADPH fractions (%)
Genotype
3 weeks
8 weeks
3 weeks
8 weeks
Wild-type as-40 as-30 as-20
100 41 9 27 8 15 6
100 41 5 31 7 18 5
38 27 19 12
36 37 31 24
4 7 5 6
5 11 7 8
Seedlings were cultured for 3 weeks in MS-agar at 200 mmol quanta m 2 sec 1, transferred to soil and grown for the indicated times under the same light regime. Relative FNR levels were calculated from immunoblots as those of Figure 1(a), and expressed as percentages of the FNR contents present in nontransformed control plants. NADP(H) amounts were determined by enzyme cycling as indicated in the Experimental procedures. The NADPH fractions were calculated as [NADPH] 100/ ([NADPH] [NADP]). Total NADP(H) levels varied between 27 and 37 nmol g 1 of FW in different specimens. Mean values and standard errors of three independent replicates are given.
334 Javier F. Palatnik et al. their enzymatic activities, as assayed in total leaf extracts by measuring the FNR-speci®c cytochrome c reductase reaction (data not shown, but see Hajirezaei et al., 2002). The variabilities in FNR protein contents and activities were similar in transgenic plants recovered from successive rounds of propagation under tissue culture conditions and in the nontransformed controls (20%), suggesting that the transgene was neither lost nor silenced during propagation. When rooted and grown autotrophically under greenhouse conditions, development of the transgenic lines was increasingly impaired and their leaves displayed spreading chlorosis that eventually covered the entire leaf surface (Figure 1b). The antisense phenotype became evident within the ®rst week of transfer, and the pace of its progress correlated with the degree of FNR de®ciency (Figure 1b). Plants with less than 20±25% extant FNR were viable for only a few weeks under autotrophic conditions, and could only be maintained by tissue culture propagation. FNR-deficient plants are abnormally susceptible to lightdependent pigment degradation and membrane damage Impaired CO2 assimilation because of FNR de®ciency likely causes most phenotypic manifestations observed in the transformants, as already demonstrated for similar lines using control analysis (Hajirezaei et al., 2002). However, chlorophyll de®ciencies of the antisense plants were less severe in the developing foliar tissue ± where requirements for synthetic reactions were more demanding ± than in fully expanded leaves (Figure 1b). Chlorosis progressed as leaves aged, suggesting that declines in chlorophyll pools were related to accelerated pigment degradation rather than to diminished synthesis. Most features of FNRde®cient leaves indeed resembled those of senescing tissues. Finally, damage became more severe as the light intensities at which plants were grown was raised, suggesting that the FNR-de®cient transformants might display abnormal sensitivity to high irradiation. To probe this possibility, the effects of different light regimes were studied in the various transgenic lines. Leaves from FNR-de®cient plants were reported to contain decreased levels of both chlorophyll a and b (Hajirezaei et al., 2002). Exposure of leaf discs obtained from as-20 transformants (approximately 80% reduction in FNR content) to high ¯uence rates led to accelerated pigment degradation (Figure 2a) and lipid peroxidation (Figure 2b), as compared to wild-type controls. Membrane damage, measured as electrolyte leakage, was also evident, depending on both FNR depletion and irradiation time (Figure 2c). At the functional level, exposure of leaves from FNR-de®cient tobacco to 1000 mmol quanta m 2 sec 1 resulted in the inhibition, relative to non-transformed siblings, of the photosynthetic rates measured subsequently at 600 mmol quanta m 2 sec 1 (Figure 2d).
Rubisco degradation in tobacco plants expressing subphysiological levels of FNR When grown in soil for 3 weeks under moderate irradiation (200 mmol quanta m 2 sec 1), total protein pro®les of transgenic lines showed no signi®cant differences with respect to wild-type plants, despite the ample variations in FNR and chlorophyll contents as well as in photosynthetic capacities (see also Hajirezaei et al., 2002). Following the sixth week of autotrophic growth, however, protein contents per leaf area began to drop in plants containing 20% or less of wild-type FNR levels (Figure 3a). A number of individual proteins have been shown to be especially prone to degradation under oxidative stress or senescing conditions. In wheat plants, the chloroplastic enzymes Rubisco and glutamine synthetase (GS) have been recently singled out as early targets for oxidative cleavage induced by either overirradiation, redox cycling or oxidants (Palatnik et al., 1997, 1999; Stieger and Feller, 1997). A similar decline in Rubisco and other photosynthetic enzymes has been observed during the process of leaf senescence in various plant species (Humbeck et al., 1996; Wingler et al., 1998). The steady-state levels of GS and the large subunit of Rubisco were hardly affected in FNR-de®cient tobacco during the ®rst weeks of autotrophic growth in the greenhouse (data not shown). Following that phase, and concomitant with the generalisation of chlorosis, the amounts of Rubisco declined sharply, whereas those of GS doubled (Figure 3b,c, lane 2). The corresponding activities followed a similar trend, although Rubisco inactivation was already evident in 3-week-old plants, before any signi®cant decline in Rubisco polypeptides could be detected (Figure 3d). Interestingly enough, a novel 66-kDa species that crossreacted with a Rubisco antiserum became apparent at this later stage of development (Figure 3b, lane 2). The 66-kDa band was recognised by antisera speci®c for either the large or the small subunits (data not shown), indicating that both components of the Rubisco holoenzyme were present in this protein variant. In a previous report (Hajirezaei et al., 2002), we have shown that FNR de®ciency could lead to extensive oxidation of the NADPH pool. This effect was attributed to faulty regeneration (by FNR) of the NADPH consumed in the Calvin cycle. When Rubisco levels began to decrease in the senescing leaves of severely depleted FNR antisense plants, the NADPH pool was partially restored to wild-type values (Table 1), despite extensive inactivation of the photosynthetic electron transport chain. Ultrastructural changes in FNR-deficient tobacco plants As chlorosis advanced, the decline in photosynthetic pigments was re¯ected in the structure of the senescing cells and plastids. The micrograph of Figure 4(a) shows that ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
Photoinhibition in FNR-deficient tobacco plants
335
Figure 2. Light-dependent damage to leaf pigments, membranes and photosynthetic activities in ferredoxin-NADP(H) reductase (FNR)-de®cient tobacco. Plants that had been grown in soil for 6 weeks were used in all cases, except for the photosynthetic experiments of panel (d), in which 4-week-old specimens were employed. (a) Total chlorophyll (circles) and carotenoids (squares) were determined in discs from wild-type (open symbols) or transformed plants of line as-20 (closed symbols), after 18 h of incubation under the light regimes shown in the abscissa. Total chlorophyll and carotenoid contents corresponding to 100% were 27.66 0.85 and 4.15 0.07 mg cm 2 of leaf tissue, respectively, for wild-type plants, and 14.20 0.90 and 2.46 0.04 mg cm 2 of leaf tissue, respectively, for as-20 antisense plants. (b) The extent of lipid peroxidation was evaluated by measuring the malondialdehyde (MDA) levels in leaf homogenates. Discs from control (wt) and as-20 plants were exposed to 1200 mmol quanta m 2 sec 1 for 8 h at 258C prior to MDA determination. (c) Ion leakage, estimated as the increase in conductivity of the medium, was determined in discs from control (*), as-40 (~), as-30 (&) and as-20 ( ) leaves irradiated at 800 mmol quanta m 2 sec 1 for the times indicated. (d) Light-dependent inhibition of photosynthesis in leaves from FNR antisense tobacco lines. The rates of CO2 assimilation, corrected for dark respiration, were estimated in the youngest fully expanded leaf. Plants were exposed to 1000 mmol quanta m 2 sec 1 for the times indicated in the abscissa, and then were assayed for photosynthesis as described under Experimental procedures. CO2 ®xation rates were expressed as the percentage of the initial rates exhibited by the same plants. They were in mmol CO2 m 2 sec 1: non-transformed controls (*), 11.0 1.2; as-40 (~), 7.5 0.8; as-30 (&), 4.0 0.8; and as-20 ( ), 1.2 0.3. All experiments were carried out at 258C, and the averaged data represent the means SE of replicate independent assays performed on six to nine individual plants.
wild-type plants displayed normal leaf anatomy, with dorsiventral compression and well-developed cells and stomata (S arrow). Large spherical nuclei containing dispersed chromatin were observed in the mesophyll. Cells in the one-layer palisade parenchyma were rich in chloroplasts, ®rmly pressed against the plasmalemma and wall by a large central vacuole (Figure 4a, C arrow). Electron microscopic analysis con®rmed that most of the cell volume is occupied by the vacuole, surrounded by a narrow band of cytoplasm containing the nucleus, chloroplasts and mitoß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
chondria with well-de®ned cristae (Figure 4b). Chloroplasts in cells from both the palisade and spongy parenchyma displayed extensive thylakoid stacking, with a few small osmiophilic globules (plastoglobuli) (Figure 4c), while the tonoplast exhibited a single membrane layer. Similar images in chlorotic areas of as-20 antisense plants revealed swelling and deterioration in cells and plastids, corresponding to leaf tissues that were no longer able to carry on photosynthesis (data not shown). The pale green sectors of young developing leaves from this
336 Javier F. Palatnik et al. different from those of wild-type plants even at a pre-symptomatic stage. Mitochondria, on the other hand, maintained their ultrastructural integrity (Figure 4e, inset). A slight decline in cell number per leaf section was accompanied by a compensating increase in cell size (Table 2). Chloroplast contents diminished steadily as chlorosis progressed, especially in the palisades cells, to reach about half of wildtype numbers in this severely FNR-de®cient line (Table 2). FNR deficiency leads to singlet oxygen accumulation in illuminated leaves
Figure 3. Rubisco degradation is accelerated in autotrophically grown ferredoxin-NADP(H) reductase (FNR)-de®cient antisense plants. Regenerated transgenic and control tobacco plants were grown in the greenhouse for 3 or 8 weeks, and extracts were prepared from the youngest fully expanded leaves as described before by Palatnik et al. (1997). Cleared lysates obtained from non-transformed (lanes 1 and 3) or from FNR antisense plants of the as-20 line (lanes 2 and 4) were subjected to SDS±PAGE and either stained with Coomassie Brilliant Blue (a), or blotted onto nitrocellulose membranes for immunodetection of Rubisco (b) and glutamine synthetase (GS) (c). In panel (a), samples corresponding to 15 mg of leaf protein (lanes 1 and 2) or 28 mm2 of leaf tissue (lanes 3 and 4) were used, whereas 1 or 15 mg of protein was loaded in panels (b) and (c), respectively. The arrows show the positions of the Rubisco large subunit (LSU), GS and a low-mobility protein (R) reacting with antisera raised against Rubisco. (d) Rubisco and GS activities were determined in soluble extracts from wildtype (wt) and as-20 plants, grown for 3 weeks (closed bars) or 8 weeks (open bars) under autotrophic conditions. One activity unit is de®ned as the amount of enzyme catalysing the conversion of 1 mmol of substrate per hour under the conditions of the experiments.
FNR-de®cient line, on the other hand, already had many abnormal characteristics, including irregular cell morphology, disorganisation of the epidermis and parenchymatous structures, and increase of the intercellular space with distortion of the cell wall (Figure 4d). Cytoplasm became thinner, and there was a signi®cant decline in the number and pigmentation of chloroplasts (Figure 4d; Table 2). Many of the features observed in the leaf cross-sections at electronic resolution resembled those present in senescing tissues, with the symptoms becoming more evident towards the inner mesophyll region (Figure 4d,e). They include reduction and condensation of the cellular cytoplasm, as well as nuclear condensation and degradation. Cells in these sections contained rudimental or agranal chloroplasts (Figure 4f) ®lled with large plastoglobuli (Figure 4e, P arrow), indicating that these plastids were
Photoinhibition usually results from kinetic reasons whenever the rate of photon absorption exceeds the rate of photon utilisation. Excess light energy leads to over-reduction of the photosynthetic electron transport chain, which in turn causes diversion of electrons to adventitious acceptors, most conspicuously, oxygen. As FNR catalyses the last step of photosynthetic electron transport, over-reduction of the intersystem electron carriers can be taken for granted in FNR-de®cient plants. The surplus of reducing equivalents might then favour propagation of toxic radical species in situ by at least two recognised mechanisms. One of them results from electron misrouting to soluble oxygen, a process occurring preferentially at the reducing side of photosystem I with production of superoxide radicals and hydrogen peroxide in the chloroplast stroma (Asada et al., 1998; Palatnik et al., 2002). A second route involves formation of excited triplet chlorophyll at the antennae, especially that of photosystem II, followed by energy transfer to ground-state molecular oxygen to generate excited singlet oxygen (1O2). The two mechanisms therefore produce different oxygen-centred derivatives, and can also be distinguished on the basis of their preferred targets: stromal components in the ®rst case, membrane lipids and proteins in the second. Singlet oxygen production during exposure of tobacco leaves to moderate light intensities was assessed by imaging quenching of the ¯uorescence of DanePy (dansyl2,2,5,5-tetramethyl-2,5-dihydro-1H-pyrrole). It has been previously demonstrated that this double ¯uorescent and spin probe could penetrate leaf cells and chloroplasts (Fryer et al., 2002; Hideg et al., 2001). Detached leaves from FNRpro®cient and -de®cient plants were in®ltrated with DanePy and then irradiated at 300 mmol quanta m 2 sec 1 for various times. Leaf tissue from non-transformed plants failed to show any signi®cant decrease in the ¯uorescence emission of the probe (Figure 5a), even after prolonged treatment (data not shown), indicating that 1O2 accumulation was negligible under these mild conditions of irradiation (Hideg et al., 2001). At variance, DanePy quenching was prominent in leaves of the as-20 antisense line after only 60 min of illumination (Figure 5b). When samples were in®ltrated with dansyl chloride, a reagent that contains ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
Photoinhibition in FNR-deficient tobacco plants
337
Figure 4. Ferredoxin-NADP(H) reductase (FNR) de®ciency causes structural changes in the chloroplasts of autotrophically grown antisense plants. Leaf sections from wild-type plants (a±c) and from pale green regions of as-20 antisense plants (d±f) were prepared and stained as indicated under Experimental procedures. (a, d) Sections of 6 mm were photographed at 40 magni®cation and standardised illumination intensity. The arrows show a stomata (S) and chloroplasts (C) around the central vacuole. The remaining pictures depict chloroplasts in 0.06mm leaf sections analysed by transmission electron microscopy at two different resolution levels. The sections presented also contain mitochondria (M) and a plastoglobulus (P). Bars correspond to 0.5 mm (b, e) and 0.05 mm (c, f). The inset in panel (e) shows a mitochondrion with well-de®ned inner structure.
Table 2 Reduction in the number of chloroplasts per cell in ferredoxin-NADP(H) reductase (FNR)-de®cient antisense tobacco plants Genotype
Cell area (mm2)
Cells (mm 2)
Chloroplasts per cell
Wild-type Palisade parenchyma Spongy parenchyma
1.36 0.29 0.83 0.08
261 21 360 22
86.8 8.0 41.7 4.8
as-20 Palisade parenchyma Spongy parenchyma
1.44 0.29 1.03 0.15
206 22 306 42
40.7 6.1 29.1 5.2
Samples of leaf tissue from wild-type and as-20 transgenic plants were fixed and contrasted as indicated under Experimental procedures. The numbers of cells and chloroplasts, as well as the cell areas, were estimated by using the IMAGE-PRO PLUS 3.0 Package. Results are the means SE obtained from 20 sections as those in Figure 4(a,d).
the same ¯uorescent group as DanePy, there was no quenching during photoinhibition (Figure 5a,b). This shows that the observed behaviour of DanePy was caused by the conversion of its pyrrole group into nitroxide through reaction with 1O2, and not by direct quenching of the dansyl moiety (Hideg et al., 2001). The collected results suggest that excess excitation energy cannot be ef®ciently dissipated as heat or photochemistry in the antisense plants, leading to abnormal build-up of singlet oxygen in the illuminated leaves. Superoxide accumulation was also monitored in leaves subjected to the same light treatment by imaging purple formazan deposits that result from reduction of nitroblue tetrazolium (NBT). The levels of the oxygen-centred derivative appeared to be similar in wild-type plants and FNRde®cient specimens (Figure 5c,d). Interpretation of in situ superoxide detection using NBT is fraught with certain dif®culties caused by the limitations of the procedures ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
employed. Although superoxide is reported to be the major species responsible for the reduction of NBT to formazan (Maly et al., 1989), the actual extension of insoluble deposits also depends on the stock of antioxidants present in the tissue that might vary widely among different lines. However, we observed only marginal differences in as-20 plants and their wild-type siblings with respect to the contents of glutathione and ascorbate, and the activities of superoxide dismutase (SOD) and ascorbate peroxidase (APX), all major components of the antioxidant cell barrier (data not shown). Moreover, the patterns of isoenzymes displaying SOD and APX activities were nearly identical in FNR-replete and -de®cient plants (Figure 6), even after exposure to moderate irradiation (data not shown). These results strongly suggest that wild-type levels of superoxide accumulation observed in as-20 leaves were not caused by increased antioxidant contents or scavenging activity in this line, relative to non-transformed specimens.
338 Javier F. Palatnik et al. Concluding remarks
Figure 5. Imaging of DanePy ¯uorescence and nitroblue tetrazolium (NBT) reduction in in®ltrated tobacco leaves after irradiation at 300 mmol quanta m 2 sec 1. Leaves from control (a, c) and FNR-de®cient plants from line as-20 (b, d) were excised, in®ltrated and illuminated as described under Experimental procedures. (a, b) The images of ¯uorescence emission from DanePy (left side of the leaf) and dansyl chloride (right side of the leaf) were photographed 60 min after the beginning of the light treatment. DanePy ¯uorescence quenching re¯ects singlet oxygen production. The spots on the leaf surface are the sites of in®ltration. (c, d) NBT-treated leaves were placed in 96% (v/v) ethanol to remove chlorophyll, visualise formazan deposits and preserve tissue integrity (Jabs et al., 1996).
Figure 6. Effect of ferredoxin-NADP(H) reductase (FNR) de®ciency on the levels of superoxide dismutase and ascorbate peroxidase isoforms in tobacco leaves. Control (lanes 1 and 3) and transgenic plants of line as-20 (lanes 2 and 4) were grown in soil for 4 weeks at 200 mmol quanta m 2 sec 1. Cleared lysates corresponding to 70 mm2 (lanes 1 and 2 of panel (a)) or 280 mm2 of leaf tissue (lanes 1 and 2 of panel (b)), or to 25 mg (lanes 3 and 4 of panel (a)) or 100 mg of total soluble protein (lanes 3 and 4 of panel (b)) were resolved by non-denaturing PAGE, and assayed for superoxide dismutase (SOD) (a) and ascorbate peroxidase (APX) (b) activities by following established procedures. The identities of the various isoenzymes were assigned according to Sen Gupta et al. (1993) and Mittler and Zilinskas (1993), with pAPX and cAPX representing plastidic and cytosolic forms, respectively.
The dynamic changes observed in autotrophically grown FNR antisense plants closely resembled the syndrome elicited by photoinhibition. Those changes include the presence of agranal chloroplasts (Figure 4f), the light-dependent degradation of chlorophyll and carotenoids (Figure 2a), inactivation of photosynthesis (Figure 2d), and augmented ion leakage indicative of membrane damage (Figure 2c). Lipid peroxidation resulting from oxidative stress is likely to initiate this membrane deterioration (Figure 2b). Then, the consequences of FNR depletion in antisense plants go beyond its role in catalysing a rate-limiting step in photosynthesis (Hajirezaei et al., 2002). FNR-de®cient plants are particularly vulnerable to photoinhibitory damage, and this sensitivity contributes signi®cantly to the bleached phenotype of the de®cient plants (Figure 1b). In general, photoinhibition occurs when irradiation is too high or when CO2 assimilation is limited by environmental stress, both situations leading to accumulation of NADPH and other reduced electron carriers (Palatnik et al., 2002). A similar situation presumably arises in chloroplasts of FNR-de®cient antisense plants, although in this case the blockade is established at the last step of the photosynthetic electron transport chain, so that NADPH remains low (Table 1). Illuminated leaves from FNR-de®cient plants displayed abnormal accumulation of singlet oxygen, while superoxide amounts remained at wild-type levels under the same conditions (Figure 5). Singlet oxygen has been implicated in self-perpetuating lipid peroxidation by direct reaction with polyunsaturated fatty acids and, in general, oxidative degradation of membrane proteins and pigments (Fryer et al., 2002). Most of the early damage undergone by the FNR antisense plants, i.e. agranal chloroplasts, augmented electrolyte leakage, lipid peroxidation and pigment degradation, was indeed con®ned to membranous regions, whereas soluble proteins such as Rubisco were affected at a later stage of development when chlorosis was well advanced. Even then, Rubisco degradation did not conform to the typical patterns of oxidative breakdown mediated by hydroxyl radicals (the end-product of superoxide propagation), especially with respect to the formation of the 37-kDa cleavage intermediate described in a number of systems (Palatnik et al., 1997, 1999; Stieger and Feller, 1997). Rather, both the inactivation and decline of this enzyme and the increase in GS protein and activity conform to a premature leaf senescence syndrome in FNR-depleted plants undergoing photoinhibitory stress (Humbeck et al., 1996; Masclaux et al., 2000; Wingler et al., 1998). Accumulation of the 66-kDa Rubisco adduct visualised in Figure 3(b) is also symptomatic of photoinactivation processes. A photoproduct of similar mass, made up of both the large and the small subunits of Rubisco, has been detected in leaves of various C3 species ± including tobacco ± exposed to ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
Photoinhibition in FNR-deficient tobacco plants increased levels of solar ultraviolet radiation (Wilson et al., 1995). Then, our collected results provide strong support for the participation of singlet oxygen, rather than superoxide, in the progress of the photoinactivation condition occurring in FNR-de®cient plants. Experimental procedures Plant material Construction and characterisation of FNR-de®cient tobacco plants (Nicotiana tabacum cv. Samsun NN) have been described elsewhere (Hajirezaei et al., 2002). Seeds from nine independent transformation events were grown on MS-agar media (Murashige and Skoog, 1962) supplemented with 2% (w/v) sucrose and 100 mg ml 1 kanamycin, and the resulting seedlings were selected for FNR levels by using SDS±PAGE and immunoblotting. FNR contents varied widely within the progeny (T1 generation) of each primary transformant, from wild-type levels to less than 5% of the amounts present in the parental plants (Hajirezaei et al., 2002). T1 specimens displaying FNR levels ranging from 15 to 45% of the contents found in the non-transformed siblings were selected for further studies. To ensure the maintenance of stable FNR levels in the transformants, internodes were cut from specimens containing four to ®ve nodes and the cuttings were transferred to MS-agar and allowed to regenerate the missing shoot to recover clonal seedlings. Phenotypes of these propagated plants were indistinguishable from those of their T1 parents. Antisense lines containing low levels of FNR could therefore be maintained by these procedures. Plants were cultured in MS-agar supplemented with sucrose and kanamycin for a minimum of 3 weeks, and then were transferred to soil. To favour optimal growth conditions, plants were watered daily with 50±100 ml of a nutrient solution prepared according to Geiger et al. (1999). In all cases, plants were illuminated at 200 mmol quanta m 2 sec 1 to provide a 16-h/8-h photoperiod.
Stress treatments Leaf discs (12-mm diameter) were punched from the 1±2 youngest fully expanded leaves of plants grown for 3±8 weeks in soil, by using a cork borer. For light treatments, two to three discs per leaf were taken from various specimens of the same line, weighted and ¯oated individually, top side up, in 1 ml of water in 24-well plates. Plant ages, light intensities and incubation times are indicated in the legends to Figures. Electrolyte leakage from the leaf discs was measured as the increase in conductivity of the medium, using a Horiba B-173 conductivity meter. To estimate the total ion content of the leaf tissue, some discs were autoclaved after the experiment and the conductivity of the resulting solution was determined as above.
Structural analysis of leaf tissue For the observations at low resolution, leaves were harvested from plants grown in soil for 3±4 weeks, and ®xed for 24 h at 258C in a mixture of formaldehyde, acetic acid and 50% (v/v) ethanol (1 : 1 : 18 by volume). Samples were dehydrated through a graded n-butanol series, and ®nally embedded in paraf®n wax. Transversal leaf sections, 6±8 mm thick, were cut and contrasted by periodic acid-Schiff (PAS) staining (O'Brien and McCully, 1981). Images were analysed by using the Image-Pro Plus 3.0 Package (Media Cybernetics, Silver Spring, MD, USA). ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
339
For electron microscopic observations, leaves were ®xed for 2 h at 48C in 2% (v/v) glutaraldehyde buffered with 20 mM cacodylate, pH 7.2, and post-®xed under the same conditions in 1% (w/v) osmium tetroxide. Samples were dehydrated in a graded ethanol series and propylene oxide and then embedded in Spurr's resin (Spurr, 1969). Ultrathin sections (60±90 nm thick) were cut with a glass knife on an ultramicrotome and mounted on Formvardcoated copper grids. The sections on the grids were stained for 30 min at 378C with 3% (w/v) uranyl acetate, then incubated for 10 min at 308C with lead citrate (0.13 M lead nitrate, 0.2 M trisodium citrate dehydrate), and ®nally examined with a JEOL-JSM100CXII (JEOL Inc., USA) electron microscope.
Photosynthetic measurements FNR antisense plants and their parental siblings, grown in soil for 4 weeks, were irradiated at 258C and 1000 mmol quanta m 2 sec 1 for the times indicated in Figure 2(d). Net CO2 uptake rates were then determined according to Von Caemmerer and Farquhar (1981), using a Qubit Systems Inc. infra-red gas analyser (Kingston, Canada), at 600 mmol quanta m 2 sec 1. The CO2 concentration of the air entering the leaf chamber and the leaf temperature were adjusted to 400 mmol mol 1 and 258C, respectively.
In situ detection of singlet oxygen and superoxide radicals in irradiated leaves Singlet oxygen was detected by in®ltrating leaves with 10 mM DanePy dissolved in 0.5% (v/v) ethanol and measuring the lightdependent quenching of the probe ¯uorescence (Hideg et al., 2001). Leaves from 6-week-old plants displaying equivalent development were excised at the petiole, in®ltrated with the reagent on the abaxial side with the aid of a needle-less syringe, and then illuminated at 300 mmol quanta m 2 sec 1. Leaves were allowed to transpirationally imbibe water during the stress period. Usually, two to three interveinal patches were in®ltrated in the anterior region of the leaf on one side of the major central vein. As a control, an equivalent surface on the other side was treated with 10 mM dansyl chloride (Hideg et al., 2001). DanePy ¯uorescence was excited in situ with 345-nm radiation produced from a UV source (Photodyne Inc., USA) and photographed using a Nikon digital Coolpix 950 (Nikon, USA) camera with the lens protected with a 545-nm band pass ®lter. The accumulation of superoxide radicals in leaves subjected to the same light regime was visualised by using the NBT reduction assay as described by Jabs et al. (1996).
Analytical procedures Leaf extracts were prepared by homogenisation of the 1±2 youngest expanded leaves of control and transgenic plants grown in soil for 3±8 weeks, in a buffer containing 50 mM Tris±HCl, pH 7.5, 5 mM MgCl2, 1 mM dithiothreitol and 1 mM phenylmethylsulphonyl ¯uoride. Proteins were determined in cleared lysates by the method of Peterson (1977), and resolved by SDS±PAGE on 12% (w/v) polyacrylamide gels essentially as reported previously by Palatnik et al. (1997). Experimental details on Western blotting and immunoreaction with speci®c antisera are also described there. Antibodies speci®c for the large and small subunits were puri®ed from a serum raised against the Rubisco holoenzyme by the method of Plaxton (1989). Chlorophylls and carotenoids were determined spectrophotometrically after extraction with 80% (v/v) aqueous acetone, followed by an additional step with pure acetone to guarantee quantitative recovery of chlorophylls a and b (Lichtenthaler,
340 Javier F. Palatnik et al. 1987). The extent of lipid peroxidation was estimated in total leaf homogenates by the procedures of Templar et al. (1999), measuring the reaction of the accumulated malondialdehyde with thiobarbituric acid. NADP(H) contents were determined by enzyme cycling essentially as described before (Hajirezaei et al., 2002). The activities of Rubisco (Ruuska et al., 2000) and glutamine synthetase (Shapiro and Stadtman, 1971) were assayed by published procedures. Isoforms of the antioxidant enzymes SOD (Sen Gupta et al., 1993) and APX (Mittler and Zilinskas, 1993) were visualised in cleared leaf extracts of control and transgenic plants after non-denaturing PAGE and activity staining.
Acknowledgements The authors wish to thank Dr Kalman Hideg (University of PeÂcs, Hungary) for the generous gift of DanePy. This research was supported by grant BID OC/AR 1201 PICT99-5105 from the National Agency for the Promotion of Science and Technology (ANPCyT, Argentina). H.O.P., E.M.V. and N.C. are members of the National Research Council (CONICET, Argentina), and J.F.P., V.B.T., N.B. and R.E.R. are Fellows of the same Institution.
References Arakaki, A.K., Ceccarelli, E.A. and Carrillo, N. (1997) Plant-type ferredoxin-NADP reductases: a basal structural framework and a multiplicity of functions. FASEB J. 11, 133±140. Asada, K., Endo, T., Mano, J. and Miyake, C. (1998) Molecular mechanism for relaxation of and protection from light stress. In Stress Responses of Photosynthetic Organisms (Satou, K. and Murata, N., eds). Amsterdam: Elsevier, pp. 37±52. Bruns, C.M. and Karplus, P.A. (1995) Re®ned crystal structure of spinach ferredoxin-NADP oxidoreductase at 1.7 AÊ resolution: oxidized, reduced and 20 -phospho-50 -AMP bound states. J. Mol. Biol. 247, 1125±1145. Carrillo, N. (1998) Partition of ferredoxin-NADP oxidoreductase between photosynthesis and antioxidant metabolism. In Photosynthesis: Mechanisms and Effects, Vol. 3 (Garab, G., ed.). Dordrecht: Kluwer, pp. 1511±1516. Carrillo, N. and Vallejos, R.H. (1987) Ferredoxin-NADP oxidoreductase. In The Light Reactions. Topics in Photosynthesis, Vol. 8 (Barber, J., ed.). Amsterdam: Elsevier, pp. 527±560. Deng, A., Aliverti, A., Zanetti, G., Arakaki, A.K., Ottado, J., Orellano, E.G., Calcaterra, N.B., Ceccarelli, E.A., Carrillo, N. and Karplus, P.A. (1999) A productive NADP binding mode of ferredoxin-NADP reductase revealed by protein engineering and crystallographic studies. Nat. Struct. Biol. 6, 847±853. Fillat, M.F., Peleato, M.L., Razquin, P. and GoÂmez-Moreno, C. (1995) Effects of iron de®ciency in photosynthetic electron transport and nitrogen ®xation in the cyanobacterium Anabaena: ¯avodoxin induction as adaptive response. In Iron Nutrition in Soils and Plants (Abadia, J., ed.). Dordrecht: Kluwer, pp. 315± 321. Fryer, M.J., Oxborough, K., Mullineaux, P.M. and Baker, N.R. (2002) Imaging of photo-oxidative stress in leaves. J. Exp. Bot. 53, 1249±1254. Geiger, M., Haake, V., Ludewig, F., Sonnewald, U. and Stitt, M. (1999) The nitrate and ammonium supply have a major in¯uence on the response of photosynthesis, carbon metabolism, nitrogen metabolism and growth to elevated carbon dioxide in tobacco. Plant Cell Environ. 22, 1177±1199. Hajirezaei, M.R., Peisker, M., Tschiersch, H., Palatnik, J.F., Valle, E.M., Carrillo, N. and Sonnewald, U. (2002) Small changes
in the activity of chloroplastic NADP-dependent ferredoxin oxidoreductase lead to impaired plant growth and restrict photosynthetic capacity of transgenic tobacco plants. Plant J. 29, 281±293. Hideg, EÂ., Ogawa, K., KaÂlal, T. and Hideg, K. (2001) Singlet oxygen imaging in Arabidopsis thaliana leaves under photoinhibition by excess photosynthetically active radiation. Physiol. Plant. 112, 10±14. Humbeck, K., Quast, S. and Krupinska, K. (1996) Functional and molecular changes in the photosynthetic apparatus during senescence of ¯ag leaves from ®eld-grown barley plants. Plant Cell Environ. 19, 337±344. Jabs, T., Dietrich, R.A. and Dang, J.L. (1996) Initiation of runaway cell death in an Arabidopsis mutant by extracellular superoxide. Science, 273, 1853±1856. Lichtenthaler, H.K. (1987) Chlorophylls and carotenoids: pigments of photosynthetic biomembranes. Meth. Enzymol. 148, 350±382. Maly, F.E., Nakamura, M., Gauchat, J.F., Urwyler, A., Walker, G., Dahinden, C.A., Cross, A.R., Jones, O.T.G. and Weck, A.L. (1989) Superoxide-dependent nitroblue tetrazolium reduction and expression of cytochrome b247 components by human tonsillar lymphocytes and B cell lines. J. Immunol. 142, 1260± 1267. Masclaux, C., Valadier, M.-H., BrugieÂre, N., Morot-Gaundry, J.-F. and Hirel, B. (2000) Characterization of the sink/source transition in tobacco (Nicotiana tabacum L.) shoots in relation to nitrogen management and leaf senescence. Planta, 211, 510±518. Mittler, R. and Zilinskas, B.A. (1993) Detection of ascorbate peroxidase activity in native gels by inhibition of the ascorbatedependent reduction of nitroblue tetrazolium. Anal. Biochem. 212, 540±546. Morales, R., Charon, M.H., Kachalova, G., Serre, L., Medina, M., GoÂmez-Moreno, C. and Frey, M. (2000) A redox-dependent interaction between two electron-transfer partners involved in photosynthesis. EMBO Rep. 1, 271±276. Murashige, T. and Skoog, F. (1962) A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant. 15, 473±497. O'Brien, P.T. and McCully, M.E. (1981) The Study of Plant Structure: Principles and Selected Methods. Melbourne: Termacarphi Pyt Ltd, pp. 6.91±6.92. Palatnik, J.F., Valle, E.M. and Carrillo, N. (1997) Oxidative stress causes ferredoxin-NADP reductase solubilization from the thylakoid membranes in methyl viologen-treated plants. Plant Physiol. 115, 1721±1727. Palatnik, J.F., Carrillo, N. and Valle, E.M. (1999) The role of photosynthetic electron transport in the oxidative degradation of chloroplastic glutamine synthetase. Plant Physiol. 121, 471±478. Palatnik, J.F., Valle, E.M. and Carrillo, N. (2002) Oxidative stress and damage in chloroplasts from dawn to dusk. In Advances in Plant Physiology, Vol. 4 (Hemantaranjan, A., ed.). Jodhpur: Scienti®c Publishers, pp. 75±88. Peterson, G.L. (1977) A simpli®cation of the protein assay method of Lowry et al. which is more generally applicable. Anal. Biochem. 83, 346±356. Plaxton, W.C. (1989) Molecular and immunological characterisation of plastid and cytosolic pyruvate kinase isozymes from castor-oil-plant endosperm and leaf. Eur. J. Biochem. 181, 443±451. Ruuska, S.A., Andrews, T.J., Badger, M.R., Price, G.D. and Von Caemmerer, S. (2000) The role of chloroplast electron transport and metabolites in modulating Rubisco activity in tobacco. ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
Photoinhibition in FNR-deficient tobacco plants Insights from transgenic plants with reduced amounts of cytochrome complex or glyceraldehyde 3-phosphate dehydrogenase. Plant Physiol. 122, 491±504. Sen Gupta, A., Heinen, J.L., Holaday, A.S., Burke, J.J. and Allen, R.D. (1993) Increased resistance to oxidative stress in transgenic plants that overexpress chloroplastic Cu/Zn superoxide dismutase. Proc. Natl. Acad. Sci. USA, 90, 1629±1633. Shapiro, B.M. and Stadtman, E.R. (1971) Glutamine synthetase (Escherichia coli). Meth. Enzymol. 17A, 910±922. Shin, M. and Arnon, D.I. (1965) Enzymic mechanisms of pyridine nucleotide reduction in chloroplasts. J. Biol. Chem. 240, 1405±1411. Spurr, A.R. (1969) A low-viscosity epoxy resin embedding medium for electron microscopy. J. Ultrastruct. Res. 26, 31±43. Stieger, P.A. and Feller, U. (1997) Requirements for the lightstimulated degradation of stromal proteins in isolated pea (Pisum sativum L.) chloroplasts. J. Exp. Bot. 48, 1639±1645. Templar, J., Kon, S.P., Millgan, T.P., Newman, D.J. and Raftery, M.J. (1999) Increased plasma malondialdehyde levels in
ß Blackwell Publishing Ltd, The Plant Journal, (2003), 35, 332±341
341
glomerular disease as determined by a fully validated HPLC method. Nephrol. Dial. Transplant. 14, 946±951. Von Caemmerer, S. and Farquhar, G.D. (1981) Some relationships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta, 153, 376±387. Wilson, M.I., Ghosh, S., Gerhardt, K.E., Holland, N., Babu, T.S., Edelman, M., Dumbroff, E.B. and Greenberg, B.M. (1995) In vivo photomodi®cation of ribulose-1,5-bisphosphate carboxylase/ oxygenase holoenzyme by ultraviolet-B radiation. Formation of a 66-kilodalton variant of the large subunit. Plant Physiol. 109, 221±229. Wingler, A., von Schaewen, A., Leegood, R.C., Lea, P.J. and Quick, P. (1998) Regulation of leaf senescence by cytokinin, sugars, and light. Effects on NADH-dependent hydroxypyruvate reductase. Plant Physiol. 116, 329±335. Ziegler, G.A. and Schulz, G.E. (2000) Crystal structures of adrenodoxin reductase in complex with NADP and NADPH suggesting a mechanism for the electron transfer of an enzyme family. Biochemistry, 39, 10986±10995.