Photosynthetic Electron Transport Determines Nitrate Reductase Gene Expression and Activity in Higher Plants*

The influence of photosynthetic electron flow in chloroplasts on the expression and enzyme activity of the cytosolic nitrate reductase (NR) was studied. Using light sources that predominantly excite either photosystem I (PSI) or photosystem II (PSII), we modulated photosynthetic electron transport in tobacco, Arabidopsis, and Lemna sprouts. In all instances, oxidation of components of photosynthetic electron flow by PSI light correlated with an increase in NR activity and/or transcription. This is confirmed by experiments with electron transport inhibitors 3-(3 (cid:1) ,4 (cid:1) -dichlorophenyl)-1,1 (cid:1) -di-methyl urea and 2,5-dibromo-3-methyl-6-isopropyl- p -benzoquinone. In addition, a Lemna mutant deficient in the cytochrome b 6 /f complex failed to respond to the different light sources and exhibited a constitutively high level of NR activity. These data indicate that NR is activated by the oxidized state of an electron transport component located after the plastoquinone pool. An involvement of the cytoplasmic photoreceptor phytochrome A in this light regulation could be excluded, since an Arabidopsis phytochrome A mutant exhibited a wild-type like response. The observation that NR activity Promoter, Generation of Transgenic Plants, F 3 Generation— The NIA2 promoter from Arabidopsis was isolated using classical strategies. A genomic Arabidopsis library was constructed in (cid:3) gt10, and a positive phage was isolated using a NIA2 cDNA fragment as a probe. Sequence analysis confirmed that the phage contained the NIA2 sequence. Using PCR, we amplified (cid:3) 3.6 kbp of the 5 (cid:2) -flanking region. At the 5 (cid:2) -end we used a vector-specific oligonucleotide, at the 3 (cid:2) -end we designed a primer that ends directly before the ATG codon. After the complete sequence of the Arabidopsis genome became avail- able, our results were confirmed: NIA2 is located on chromosome 1 and our 5 (cid:2) -flanking region corresponds to the BAC F28L22 sequence 59370– 62005. The DNA fragment was inserted into the Sma I site of pBI101 (56) and transformed into Nicotiana tabacum (Samsun NN) via Agrobacterium tumefaciens (57). 35 transgenic lines were regenerated after selection on kanamycin (100 (cid:2) g/ml). F 3 seedlings of 6 lines that showed the highest expression level of the reporter gene were used for these studies. Physiological Characterization of Transgenic Lines— For the initial characterization of the transgenic lines, F 3 seedlings were grown on solidified half-strength MS media for 18 days in either or darkness. The media either not supplemented or supplemented nitrate m M ), cytokinin (N 6 -benzylaminopurine, 10 (cid:4) 5 M ), abscisic acid (10 (cid:4) M ), sucrose or (cid:4) M ). For electron port inhibitor experiments with DCMU or DBMIB as

Nitrate is the major source of nitrogen for all living organisms. It is reduced by nitrate reductase (NR), 1 the key enzyme of the nitrate-assimilating pathway. NR catalyzes the ratelimiting step in this process (1) and generates nitrite in the cytoplasm of a plant cell, which is translocated into the plastids for further reduction and metabolization (2)(3)(4)(5)(6). Expression of nitrate reductase and its activity is highly regulated by a variety of environmental and cell-internal factors, such as nitrate or ammonium (2,3,(7)(8)(9)(10)(11), the circadian rhythm (12)(13)(14)(15), sucrose or glucose (16 -18), the CO 2 concentration (19 -21), the developmental stage of the plastids (22)(23)(24)(25)(26), and especially light (12,(27)(28)(29)(30)(31). Most of the energy for nitrate assimilation in a cell derives from photosynthesis, and a few studies suggest that photosynthesis could be also involved in the light regulation of NR activity and NR gene expression (3,(32)(33)(34). However, our present understanding of this photosynthetic control is weak. Further complexity derives from the observation that NR is regulated at various levels. NR transcription is controlled via one or several responsive elements in the promoter (3,5,9,(35)(36)(37)(38)(39), and signal transduction mutants with abnormal NR transcription have been identified (40). Stimulation of NR gene expression leads to a rapid accumulation of the NR mRNA, followed by an increase in the protein level (41,42). The protein has a short half-life, which allows an efficient downregulation of the nitrate metabolism under unfavorable conditions (42,43). In addition, NR can be rapidly inactivated by phosphorylation (44 -47), and this scenario involves 14-3-3 proteins (48). Apparently the N-terminal part of the enzyme is essential for this regulation (49).
To which extent and why plastids exert their photosynthetic control on NR activity in the cytosol and NR gene expression in the nucleus is mainly unknown. This study monitored for the first time the changes in expression of the chimeric NIA2 promoter::UIDA gene fusions in transgenic tobacco seedlings, the NIA2 transcript accumulation in Arabidopsis and NR activity in three different species (Lemna, Arabidopsis, and tobacco) in parallel after physiological modulation of photosynthetic electron transport. Using photosystem-specific excitation in combination with electron transport inhibitor experiments on whole seedlings we demonstrate that the efficiency of photosynthetic electron flow affects the transcriptional rate and transcript pool size of the nuclear gene as well as the NR enzyme activity in a coordinated manner underlining the importance of photosynthesis as regulator of nitrate assimilation pathway components in plastids, cytosol, and nucleus.

EXPERIMENTAL PROCEDURES
Growth of Tobacco and Arabidopsis Seedlings-All tobacco (wildtype and transgenic lines, Samsun NN) and Arabidopsis (var. Columbia, or Landsberg in comparison with the phytochrome A mutant in the same background) seedlings were grown in Petri dishes with 0.5ϫ strength Murashige and Skoog (MS) medium (50) supplemented with 2% sucrose in a temperature controlled growth chamber at 22°C. Seedlings were either kept in PSI or PSII light for 18 days or they were transferred from one light source to the other 96 h before harvest. The spectral quality and light quantity of the PSI-and PSII-light sources (51, 52) as well as the characterization of the acclimation process of the tobacco seedlings to the two light sources have been described earlier (53). The acclimation of Lemna sprouts and of Arabidopsis (Landsberg erecta) seedlings has been analyzed in this study (see below). White light control plants were illuminated continuously for 18 days with 30-watt white stripe lamps (OSRAM) with a photon flux density of 100 E. After 18 days, the seedlings were harvested and prepared for the determination of NR or ␤-glucuronidase (GUS) activity.
The electron transport inhibitors DCMU and DBMIB (Sigma) have been applied to wild-type and transgenic plants as described (53). Control seedlings were treated with the solvent without inhibitors. DBMIB treatment was repeated every 3-4 h during the last 96 h of experimentation. Stock solutions of DCMU were 10 mM in 50% ethanol and of DBMIB 100 mM in 10% Me 2 SO in ethanol. Sublethal effects of used inhibitor concentrations (10 M for DCMU and 25 M for DBMIB) have been determined and described earlier (53). The different inhibitor concentrations were prepared by dilution in sterile water directly prior to use.
For RNA isolation Arabidopsis wild-type seedlings were grown for 14 days under PSI or PSII light and were then switched to the respective other light source for 8 h. In inhibitor experiments plants were treated with 10 M DCMU or 25 M DBMIB just before this light switch. Untreated controls remained 14 days and 8 h under PSI or PSII light.
Growth of Lemna aequinoctialis-The propagation, growth conditions, growth media, and temperature were described by Appenroth et al. (54), except that the medium was supplemented with 50 mM glucose. Wild-type and mutant strains (1073, see Ref. 55) were treated identically. After 2 weeks in white light, sprouts were transferred to darkness for 3 days before shifting them to PSI or PSII light for 48 h. After acclimation, sprouts were transferred to the respective other light source for an additional 48 h, and NR activity was assayed.
Isolation of the NIA2 Promoter, Generation of Transgenic Plants, F 3 Generation-The NIA2 promoter from Arabidopsis was isolated using classical strategies. A genomic Arabidopsis library was constructed in gt10, and a positive phage was isolated using a NIA2 cDNA fragment as a probe. Sequence analysis confirmed that the phage contained the NIA2 sequence. Using PCR, we amplified ϳ3.6 kbp of the 5Ј-flanking region. At the 5Ј-end we used a vector-specific oligonucleotide, at the 3Ј-end we designed a primer that ends directly before the ATG codon. After the complete sequence of the Arabidopsis genome became available, our results were confirmed: NIA2 is located on chromosome 1 and our 5Ј-flanking region corresponds to the BAC F28L22 sequence 59370 -62005. The DNA fragment was inserted into the SmaI site of pBI101 (56) and transformed into Nicotiana tabacum (Samsun NN) via Agrobacterium tumefaciens (57). 35 transgenic lines were regenerated after selection on kanamycin (100 g/ml). F 3 seedlings of 6 lines that showed the highest expression level of the reporter gene were used for these studies.
Physiological Characterization of Transgenic Lines-For the initial characterization of the transgenic lines, F 3 seedlings were grown on solidified half-strength MS media for 18 days in either light or darkness. The media were either not supplemented or supplemented with nitrate (15 mM), cytokinin (N 6 -benzylaminopurine, 10 Ϫ5 M), abscisic acid (10 Ϫ5 M), sucrose (2%), or norflurazon (10 Ϫ5 M). For electron transport inhibitor experiments seedlings were sprayed with DCMU or DBMIB as described above.
GUS Staining-Seedlings were harvested and immediately put into X-gluc solution (50 mg X-gluc; 1 ml of dimethylformamide; 4.9 ml of 50 mM sodium phosphate, pH 7.0; 250 l of Me 2 SO; 500 l of potassium hexacyanoferrate (III) (100 mM); 500 l of potassium hexacyanoferrate (II) (100 mM)) and incubated overnight at 37°C. After washing with water, the seedlings were incubated in 70% ethanol and stored at 4°C. For GUS staining of root hairs, seeds were germinated and seedlings were grown in liquid MS media to avoid hair damage.
Chl Fluorescence Measurements-In vivo Chla fluorescence parameters were measured with a pulse amplitude-modulated fluorometer (PAM101/103; Heinz Walz, Effeltrich, Germany). Arabidopsis seedlings and Lemna sprouts were arranged densely, so that the fluorescence of several seedlings/sprouts could be measured simultaneously under the emitter/detector unit. Fluorescence parameters were determined as described previously (53). The steady-state fluorescence Fs was calculated as F t Ϫ F o Ј ϭ F s . Fluorescence-quenching parameter qP (photochemical quenching) was calculated as qP ϭ (F m Ј Ϫ F s )/(F m Ј Ϫ F o ) (58). The effective quantum yield of PSII ( PSII) was calculated as PSII ϭ (F m Ј Ϫ F s )/F m Ј (59).
RNA Preparation and Quantitative RT-PCR-Total RNA from 3-5 g of leaf material was isolated following a protocol modified from Chomczynski and Sacchi (60) using the TRIzol reagent (Invitrogen Life Technologies). RT-PCR analysis was performed by reverse transcription of 5 g of total RNA with gene-specific reverse primers (see below) for Arabidopsis NIA2 and 18S rRNA genes using a first strand cDNA synthesis kit (K1631) (MBI Fermentas, St. Leon-Roth, Germany) followed by 20 PCR cycles. Gene-specific primer pairs for amplification of NIA2 and 18S cDNAs were as follows: NIA2: forward primer, 5Ј-ATG GCG GCC TCT GTA GAT AAT CGC CC-3Ј; reverse primer, 5Ј-CCT CGT GAC ATG GCG TCG TAA TCA CGG-3Ј; 18S: forward primer, 5Ј-GGT AGG CGA TTG GCT AAC ATT GTC TGC-3; reverse primer, 5Ј-GAG ACA CCA ACA GTC TTT CCT CTG CG-3Ј. PCR products were separated on 1.5% agarose gels and stained with ethidium bromide, and visualized bands were quantified with the ImageMaster Video system (Amersham Biosciences).
Enzyme Assays-The NR and GUS assays were described earlier (14,57). In both instances the system of reference was an equal amount of fresh weight.
Saccharose Determination-Saccharose was determined with an analytical kit from Roche Molecular Biochemicals (716260, Ingelheim, Germany).
Statistics-All NR activity data are based on seven independent experiments. The GUS values are based on independent experiments with the F 3 seedling populations of six independent lines; all experiments were repeated seven times.

Expression and Regulation of the Arabidopsis NIA2
Promoter::UIDA Gene Fusion in Tobacco-To analyze the regulation of NR gene expression under various environmental conditions as well as its spatial expression characteristics in whole plants, we generated transgenic tobacco plants with a chimeric Arabidopsis NIA2 promoter::UIDA gene fusion (see "Experimental Procedures"). The isolated 3.6-kbp promoter region differed only in 3 positions in the far upstream region from the sequence, which is now available in the data bank (Gen-Bank TM acc. no. F28L22, see "Experimental Procedures"). Initial primer extension analyses identified 3 major transcription start sites (59, 88, and 152-bp upstream of the ATG codon), the most prominent one being located 88-bp upstream of the ATG codon (data not shown), which is consistent with previous results (36). F 3 seedlings from six independent transgenic lines were used for initial physiological and histological studies (Table I and Fig. 1). GUS gene expression was significantly stimulated by light. The application of cytokinin, abscisic acid, sucrose, norflurazon, DCMU, and DBMIB had only small effects on the NIA2 promoter-driven reporter gene expression in the dark, only nitrate could activate it significantly. Additional illumination, however, led to a stimulation of promoter activity in each case. This becomes especially apparent when combinations of nitrate, sucrose, and cytokinin treatments were used (data not shown). As only exception, norflurazon treatment of seedlings, which completely prevents plastid biogenesis (61), resulted in an inhibitory effect either when applied alone or in combination with other positive regulators (data not shown). These results demonstrate that the transgene responds similarly to the applied regulatory compounds or signals as described in other studies with NR (see the Introduction) and confirm at least partial transcriptional regulation of NR. Further GUS staining experiments indicate that the promoter is active in shoots and roots (Fig. 1). If seedlings are grown under conditions that allow extensive root formation and subsequent analysis of GUS staining (see "Experimental Procedures"), the high expression level in root hairs becomes obvious. Irrespective of the system of reference (fresh weight or protein content), ϳ20% of the total GUS activity in white light-grown seedlings is found in roots, while more than 40% is detectable in roots of etiolated seedlings. More detailed studies uncovered that the activity in shoots is up-regulated by light and the functional stage of the plastids while the GUS level in the roots is more or less constitutively expressed (data not shown). This suggests that light regulation of NR in shoots is limited to photosynthetic active tissue. Therefore, all subsequent physiological studies were performed solely with cotyledons from seedlings grown under different light conditions. Transgenic tobacco lines with PSI promoter::UIDA reporter gene constructs have been demonstrated to represent a useful tool to investigate photosynthetic redox signaling pathways between plastids and the nucleus (53). Therefore the tobacco lines with the NIA2 promoter::UIDA gene fusion offered the opportunity to determine in an identical experimental setup whether light regulation of the Arabidopsis NIA2 promoter activity is coupled to photosynthetic electron transport. In this setup transgenic tobacco seedlings were grown under light sources favoring either PSI or PSII. Such light sources induce imbalances in the excitation of the two photosystems, and shifts between them can be used to generate oxidation or reduction signals from photosynthetic electron transport (51,52,62). Furthermore, in this system exogenous application of sublethal concentrations of electron transport inhibitors DCMU and DBMIB, which do not block, but limit the photosynthetic electron transport allow to confirm a coupling of photosynthetic electron transport and transgene promoter activity (53). By performing such experiments with the tobacco lines carrying the NIA2 promoter::UIDA gene fusion we determined that PSI light stimulated and PSII light inhibited the Arabidopsis NIA2 promoter activity (Fig. 2). Conversely, PSI light grown plants (PSI plants) which were shifted to PSII light (PSI 3 II plants) showed a decrease in the NIA2 promoter activity while PSII light-grown plants (PSII plants) exhibited the opposite reaction after a shift to PSI light (PSI 3 I plants). We then added DCMU and DBMIB in the same way as described previously (53) in order to manipulate the redox signal generated by the photosynthetic electron transport. Both inhibitors had no or only small effects on the NIA2 promoter activity in PSI plants but activates it significantly in PSII plants. These data are consistent with that obtained in the light-shift experiments and demonstrate that NR transcription is coupled to photosynthetic electron transport.
Redox Regulation of NIA2 Transcript Accumulation-To analyze if the regulation of the NIA2 promoter activity is reflected in transcript abundance and if the transgene expression in tobacco correlates with the situation in Arabidopsis we monitored the NIA2 transcript pool size by RT-PCR in Arabidopsis seedlings, which were grown under the same conditions as the transgenic tobacco lines. The NIA2 RT-PCR product (Fig. 3, 1120 bp) exhibited essentially the same regulation as the promoter::reporter gene constructs; under PSI light the seedlings contain high amounts of NIA2 transcripts (Fig. 3, lane 2), a shift to PSII light for only 8 h decreased the NIA2 transcript abundance (Fig. 3, lane 3) whereas pretreatment with DCMU and DBMIB abolished this repression (Fig. 3, lanes 4 and 5). In PSII light-grown plants only small amounts of NIA2 transcripts could be detected (Fig. 3, lane 7) while a shift to PSI light increased the transcript pool size (Fig. 3, lane 8). A preceding inhibitor treatment had no negative effect on this activation (Fig. 3, lanes 9 and 10).
Redox Regulation of Nitrate Reductase Enzyme Activity-Nitrate reductase is also highly regulated at the level of enzyme activity. To elucidate if the observed redox regulation extends to the enzyme activity tobacco has been grown under the different light regimes and NR activity was determined (Fig. 4). Under PSI light NR activity was higher than under PSII light. Furthermore shifts between these light sources resulted in corresponding changes of NR activity, i.e. a PSI 3 II light shift repressed the enzyme activity whereas a PSII 3 I light shift increased it. These results are consistent with those obtained at the transcriptional level.
To obtain additional support for the coupling of NR activity to photosynthetic electron transport, which did not rely on inhibitor experiments, we included a mutant of L. aequinoctialis (1073) that lacks the cytochrome b 6 f complex because of a mutation in the Rieske protein (55). In this mutant all redoxreactive compounds located after the plastoquinone pool are always oxidized. Furthermore, light regulation of NR activity was shown to be mediated by phytochrome A (phyA) (63). To prove the independence of the observed photosynthetic control from the cytosolic photoreceptor we decided to test the response of a phyA-deficient Arabidopsis mutant to the PSI and PSII light sources. Since both organisms have not been used in our experimental setup before we first had to determine the regular response of the respective wild-types to PSI and PSII light treatments. We checked the respective acclimation of wild-type Arabidopsis seedlings and L. aequinoctialis sprouts by determining characteristic changes in Chl fluorescence parameters, which were obtained by standard pulse amplitude modulated fluorescence measurements (Table II) as reported in previous studies with mustard and tobacco seedlings (52,53). No significant changes in the maximal quantum yield were observed after acclimation of the plants to the different conditions. However, PSI plants show a higher level in steady state fluorescence Fs than PSII plants, which is known as a typical acclimation response (53). The Fs/Fm ratio therefore is high in PSI plants and decreases after acclimation to PSII light, and the opposite reaction was observed for PSII plants after acclimation to PSI light. As a consequence the reduction state of the first electron acceptor of PSII, Q A , expressed as 1-qP, is significantly higher in PSI and PSII 3 I plants than in PSII and PSI 3 II plants. By contrast, the effective quantum yield PSII is higher in PSII plants than in PSI plants and changes accordingly after the shift to the alternate light source. These data indicate that Arabidopsis and Lemna readily acclimate to light quality.
We then determined the NR activity in these organisms after acclimation to the PSI and PSII light sources. In L. aequinoctialis wild-type sprouts NR was found to be substantially decreased under PSII light as compared with PSI seedlings (Fig.  5). Shift experiments from one light source to the other demonstrated that 2 days are sufficient to acclimate the NR activity to the new light conditions. The Lemna mutant 1073, however, failed to respond to the different light conditions and showed a constitutively high level of NR activity under all illumination conditions, comparable to the activity level detectable in PSI light acclimated wild-type sprouts. Thus, oxidation of compounds associated with the photosynthetic electron flow after the PQ pool appears to be accompanied by an increase in the NR activity in the cytoplasm. Fig. 6 demonstrates that this is also observed for Arabidopsis seedlings. NR activity was always stimulated by PSI light and repressed by PSII light. For the shifted seedlings, the redox state of the plastoquinone pool was also modulated by the inhibitors DCMU or DBMIB (53), which were applied directly before the light shifts. In all instances, NR activity increased to values comparable to that from the PSI seedlings (data not shown). It is concluded that the redox state of the plastoquinone pool is not involved in NR activity in the cytoplasm and that redox-reactive compounds located after the plastoquinone pool in the electron transport chain are crucial for NR regulation. The same experiments performed with the phyA mutant led to comparable results (Fig. 6), again illumination with PSI light activated NR activity while PSII light repressed it. Since this response pattern is identical to the wild-type seedlings it is concluded that the observed redox regulation of NR activity is independent from phyA-mediated light regulation. DISCUSSION We showed that acclimation of three different plant species to light sources favoring either PSI or PSII excitation dramatically affected NR activity and/or gene expression. In all instances oxidation of photosynthetic electron transport components after the plastoquinol oxidation site, either by the PSI light source, by the application of inhibitors, or by a mutation  in the cytochrome b 6 f complex, resulted in an increase in NR activity in the cytoplasm. The responsible redox control parameter could not be identified, however the inhibitor studies clearly demonstrate that the redox state of the plastoquinone pool cannot be the origin of the signal, because application of DBMIB which blocks the electron transport after the plastoquinone pool, has the same effect on NR and GUS activities as DCMU which blocks before the plastoquinone pool. Therefore, the redox sensor should be located after the Q o site of the cytochrome b 6 /f complex and must activate NR activity in a more oxidized state. One might argue that in planta treatments with DBMIB may lead to wrong results since DBMIB in very high concentrations is known to bind at the Q B site of PSII, however, in our experiments this is unlikely since the used concentration was shown to inhibit electron transport only partially (53). This is supported by the constitutively high and unregulated NR activity in the Lemna mutant lacking the cytochrome b 6 f complex (55), which suggests that the redox signal(s) originate even downstream from the cytochrome b 6 f complex. Alternatively, one might speculate that if the redox signal originates from a component in the cytochrome b 6 f complex, the lack of this complex in the mutant leads to the complete loss of any regulation resulting in permanent high NR activity.
Application of DCMU and DBMIB usually results in a decline of the response of interest, which makes it difficult to exclude side effects. In our studies, DCMU and DBMIB treatment of dark-grown seedlings had only little effect on NIA2 promoter activity suggesting no major side effects of the inhibitors at least at the concentrations used here. However, NR activity and expression are stimulated by both inhibitors. In addition, also in the Lemna mutant, in which the electron flow is completely blocked, a significant increase in the NR activity can be observed in comparison to the isogenic wild type. Stimulation rather than inhibition of the NR activity and/or transcription in response to the inhibitors therefore indicates that the effects are specific. Interestingly, DCMU and DBMIB had no effect on GUS gene expression when dark-grown tobacco seedlings were set into light. This could mean that the lightinduced NIA2 promoter activity during greening of seedlings does not depend on photosynthesis and is mediated by other light receptors such as phytochromes. Another possibility, however, could be that the intensity of the white light source (which is 3-5 times of that of the PSI or PSII light sources) is so high that sublethal inhibitor concentrations does not generate a significant signal. Treatment of seedlings with 100 M DCMU resulted in a 30% increase in GUS activity (data not shown) suggesting that this second possibility might be true.
In Arabidopsis, NR activity is the result of two gene products, NIA1 and NIA2. It remains to be determined whether both genes respond identically to the redox signal. Our data demonstrate that the total NR activity is regulated comparably to the expression of the NIA2::UIDA gene fusion, and previous studies have shown that NIA1 and NIA2::reporter gene fusions in transgenic tobacco respond similarly to exogenously applied agents such as nitrate (36). Thus it is very likely that the  regulatory mechanisms controlling the expression of both genes are similar.
The importance of light for NR regulation has been investigated in many studies (63,64), however, until now there is still confusion as to whether light is an absolute requirement and what the primary photoreceptor is for such a regulation. Gowri and Campbell (8) have shown that nitrate can induce NR mRNA in etiolated and light-grown maize leaves, and Cheng et al. (16) proposed that light can be replaced by sucrose. Thus it was argued that light is not obligatory when nitrate and sufficient carbohydrates are available. Furthermore, in many studies it remained open whether phytochromes or components deriving from photosynthesis are crucial for NR regulation (63). Our data clearly demonstrate that phytochrome A in Arabidopsis is not involved in the NR regulation analyzed in this study and that the role of the photosynthetic electron flow for NR expression is probably more significant than anticipated thus far. It remains to be determined whether NR activity also responds to different irradiances. The PSI light used here has a lower photosynthetic active radiation than the PSII light, but nonetheless has an activating effect. Regulation of NR differs from other nuclear-encoded redox-controlled genes investigated in this system in that the PSI light source activates NR. Pfannschmidt et al. (53) have recently demonstrated that light sources preferentially exciting PSII stimulate the expression of nuclear genes for PSI components. These differences might be explained by the fact that the nitrate assimilation pathway is predominantly dependent upon the availability of reduction equivalents, whereas expression of PSI genes is regulated by an acclimation process, which optimizes light harvesting under unfavourable PS stoichiometry.
The transport of the redox signal from the thylakoid membrane to the cytoplasm at present is difficult to explain and may involve additional factors. In any case, triose phosphates generated during photosynthesis appear to be transported into the cytoplasm and provide the reducing equivalents for NR activity through their oxidation by triose phosphate dehydrogenase. Two systems, the dihydroxyacetone phosphate/phosphoglycerate shuttle and the malate/oxalacetate shuttle, have been proposed to provide NADH for nitrate reduction (see Ref. 64). Whether these pools are plastid target sites for regulation is presently unclear. An additional putative transmitter could be sucrose. Cheng et al. (16) have shown that 2% sucrose can replace light in eliciting an increase of NR mRNA accumulation in dark-adapted green Arabidopsis plants and that a 2.7-kbp region of the 5Ј-flanking sequence of the NIA1 gene is sufficient to confer the light or sucrose response. Our physiological conditions differ from those of Cheng et al. (16) in that we kept the seedlings continuously on 2% sucrose under light conditions. In addition, we could not detect any significant difference in the internal sucrose concentrations under our conditions (PSI plants: 1.19 Ϯ 0.07 mg/g fresh weight; PSII plants: 1.22 Ϯ 0.05 mg/g fresh weight). Thus, the involvement of bulk sucrose in the plastids and cytoplasm as a specific transmitter for the efficiency of the photosynthetic electron transport appears unlikely. In addition, Oswald et al. (34) reported that a plastidderived redox signal can override the sugar-regulated expression of nuclear-encoded photosynthesis genes suggesting that photosynthetic redox signals may act independently from the sugar status of the cell. However, they found no significant effect of DCMU on NR transcript accumulation in an Arabidopsis cell culture. In the study NR mRNA abundance was found to be high under 3% sucrose and low after sugar repletion, and application of DCMU did not affect this regulation. It is difficult to reconcile this observation with our finding that DCMU application raises NR expression and activity. How-ever, a conceivable explanation could be that 3% sucrose results in such a high NR induction that it masks any effect of the inhibitor even after sugar repletion whereas the constant presence of 2% sucrose in our study allows the variation of NR expression and activity by photosynthetic redox signals. This suggests that in the case of NR sugar signals can override photosynthetic redox signals, which represents the opposite type of regulation as observed for nuclear photosynthetic genes. This is consistent with the fact that NR is positively regulated by sugar signals while photosynthesis genes are negatively regulated. The exact range in which these signals cooperate or inhibit each other has to be analyzed in the future in a well defined physiological system.
What is the reason for the photosynthetic control of NR expression and activity by the PSI and PSII lights used in this study? Recently, Wollman (65) proposed that the progress in the field of State I/State II transitions offers a new view of photosynthesis as a flexible energy conversion system in which State I behaves as a carbon fixation device whereas State II operates more likely as an ATP generator. This view provides an attractive explanation for our observations. In State I (which is reached under PSI light) linear electron flow and generation of reducing equivalents are promoted thus allowing the reduction of assimilated nitrate beside CO 2 . In State II (which is reached under PSII light) ATP generation is preferred, and nitrate (and CO 2 ) reduction is decreased.
It has been previously reported that NR activity and transcription increases with increasing CO 2 concentrations suggesting that nitrate and CO 2 reduction are correlated. It was interpreted that a simultaneous increase in CO 2 fixation and nitrate assimilation allows for a faster plant growth (26). Our investigations show that nitrate assimilation appears to be correlated to the light energy distribution between the photosystems that provide an acclimation of the plant cell metabolism to the energy supply by photosynthesis. The amount of reduced NADPH does not seem to be the signaling parameter since the Lemna mutant lacking the cytochrome b 6 f complex is not able to produce this compound but shows increased nitrate activity. This is consistent with observations in barley where a high nitrate reductase activity was found to be not strictly linked to a reduced ferredoxin pool nor to high Calvin cycle activity (33). Therefore we propose as working hypothesis that the signal is generated between the PQ pool and the reducing side of PSI, which regulates nitrate reductase activity and/or expression under low light conditions when photosynthetic energy and reducing power supply are limited and therefore have to be used economically. Further investigations that will clarify these interactions are under way.