Reciprocal diurnal changes of phosphoenolpyruvate carboxylase expression and cytosolic pyruvate kinase, citrate synthase and NADP-isocitrate dehydrogenase expression regulate organic acid metabolism during nitrate assimilation in tobacco leaves

Authors

  • W.-R. Scheible,

    1. Botanisches Institut der Universität Heidelberg, Im Neuenheimer Feld 360, D-69120 Heidelberg, Germany
    Search for more papers by this author
    • *Present address: Carnegie Institution of Washington, Department of Plant Biology, 260 Panama Street, Stanford CA 94305, USA

  • A. Krapp,

    1. Botanisches Institut der Universität Heidelberg, Im Neuenheimer Feld 360, D-69120 Heidelberg, Germany
    Search for more papers by this author
  • M. Stitt

    1. Botanisches Institut der Universität Heidelberg, Im Neuenheimer Feld 360, D-69120 Heidelberg, Germany
    Search for more papers by this author

Correspondence: M.Stitt. Fax: + 49 6221 545859; e-mail: mstitt@botanik1.bot.uni-heidelberg.de

ABSTRACT

Diurnal changes of transcript levels for key enzymes in nitrate and organic acid metabolism and the accompanying changes of enzyme activities and metabolite levels were investigated in nitrogen-sufficient wild-type tobacco, in transfomants with decreased expression of nitrate reductase, and in nitrate-deficient wild-type tobacco. (i) In nitrogen-sufficient wild-type plants, transcript levels for nitrate reductase (NR, EC 1.6.6.1), nitrite reductase (NIR, EC 1.7.7.1) and phosphoenolpyruvate carboxylase (PEPC, EC 4.1.1.31) were high at the end of the night and decreased markedly during the light period. The levels of these three transcripts were increased and the diurnal changes were damped in genotypes with decreased expression of nitrate reductase. The levels of these transcripts were very low in nitrate-limited wild-type plants, except for a small rise after irrigation with 0·2 mM nitrate. (ii) The levels of the transcripts for cytosolic pyruvate kinase (PK, EC 2.7.1.40), mitochondrial citrate synthase (CS, EC 4.1.3.7) and NADP-isocitrate dehydrogenase (NADP-ICDH, EC 1.1.1.42) were highest at the end of the light period and beginning of the night. These three transcripts increase and the diurnal changes were damped in genotypes with decreased expression of NR. (iii) The diurnal changes of transcript levels were accompanied by changes in the activities of the encoded enzymes. The activities of NR and PEPC were highest in the early part of the light period, whereas the activities of PK and NADP-ICDH were highest later in the light period and during the first part of the night and CS activity was highest at the end of the night. Activity of PEPC, PK, CS and NADP-ICDH increased and the diurnal changes were damped in genotypes with low expression of NR. Activity of all four enzymes decreased in nitrate-limited wild-type plants. (iv) In the light, malate accumulated, citrate decreased, and about 30% of the assimilated nitrate accumulated temporarily as glutamine, ammonium, glycine and serine. These changes were reversed during the night. (v) It is proposed that the diurnal changes of expression facilitate preferential synthesis of malate to act as a counter-anion for pH regulation during the first part of the light period when NR activity is high, and preferential synthesis of 2-oxoglutarate to act as a nitrogen acceptor later in the day when large amounts of nitrogen have accumulated in ammonium, glutamine and other amino acids including glycine in the photorespiration pathway, and NR activity has been decreased.

INTRODUCTION

Photosynthetic metabolism in source leaves is subject to dramatic diurnal changes (Geiger & Servaites 1994). The most obvious is the alternation between photosynthesis in the light and respiration in the dark. Superimposed is a cycle of starch accumulation and mobilization which allows part of the assimilated carbon to be temporarily stored in the leaf to support sucrose export during the night (Stitt, Huber & Kerr 1987; Heineke et al. 1994; Geiger & Servaites 1994). The balance between carbon export and accumulation even shifts during the light period, with the early part of the light period being characterized by high sucrose phosphate synthase (SPS, EC 2.4.1.14) activity (Stitt et al. 1987; Stitt 1996) and the later part by lower SPS activity and higher rates of starch synthesis (Stitt et al. 1987; Stitt 1996). Analogous diurnal changes occur in nitrogen metabolism. Nitrate reductase (NR) activity is highest in the first part of the light period, declines in the later part of the light period, and is low or negligible at night (Galangau et al. 1988; Kaiser & Huber 1994; Scheible et al. 1997c), and amino acids accumulate during the day and decrease during the night (Scheible et al. 1997c).

The contribution of metabolite effectors and post-transcriptional modification to diurnal regulation of the Calvin cycle (Woodrow & Berry 1988; Stitt 1996), sucrose and starch synthesis (Stitt et al. 1987; Stitt 1996; Huber & Huber 1996) and nitrate assimilation (Champigny & Foyer 1992; Hoff, Truong & Caboche 1994; Huppe & Turpin 1994; Huber, Bachmann & Huber 1996; MacKintosh 1998) has been intensively studied. Transcriptional regulation and protein turnover have received less attention. Although there are diurnal changes of the levels of transcripts encoding proteins involved in photosynthetic carbon including chlorophyll binding proteins and the small subunit of Rubisco (Dean, Pichersky & Dunsmuir 1989), they are rarely accompanied by changes of the encoded protein, and evidence for a direct contribution to the regulation of metabolism is usually lacking.

Changes in transcription and protein turnover clearly contribute to the diurnal regulation of nitrate assimilation. Transcript for nitrate reductase (NIA) is high at the end of the night and decreases dramatically during the light period (Galangau et al. 1988; Scheible et al. 1997c) and NR protein rises to a maximum after 3–4 h in the light and then decreases (Scheible et al. 1997c). The decline of NIA transcript during the day is accompanied by a decline of nitrate and increase of glutamine, and the recovery of the NIA transcript level during the night is accompanied by an increase of nitrate and decrease of glutamine (Deng et al. 1991; Scheible et al. 1997c). Several lines of evidence show that nitrate induces (Pouteau et al. 1989; Cheng et al. 1991; Lin et al. 1994; Scheible et al. 1997b), and that glutamine or related downstream metabolites formed during nitrate or ammonium assimilation repress NIA (Vincentz et al. 1993; Hoff et al. 1994). The diurnal changes of NIA transcript and protein are attenuated in genotypes with low NR activity (Vaucheret et al. 1990; Scheible et al. 1997c), demonstrating that they are driven by signals deriving from nitrate and nitrate metabolism.

Enzymes further downstream in nitrate and ammonium metabolism including NII [encoding nitrite reductase (NIR)] (Wray 1993; Rastogi et al. 1993; Scheible et al. 1997b), GLN1 (encoding the cytosolic glutamine synthetase), GLN2 (encoding plastid glutamine synthetase) and GLU (ferredoxin-dependent glutamate synthase; Fd-GOGAT) (Redinbaugh & Campbell 1993; Scheible et al. 1997b) are also induced by nitrate. The levels and diurnal changes (Wray 1993; Lam et al. 1996) of these transcripts are modified when nitrate is supplied to mutants and transformants with little or no NR activity (Vaucheret et al. 1990; Kronenberger et al. 1993; Scheible et al. 1997b). The nitrate-induced change of transcripts leads to increased NR (Shaner & Boyer 1976; Galangau et al. 1988; Gowri et al. 1992; Lin et al. 1994; Scheible et al. 1997b), NIR (Scheible et al. 1997b), GS (glutamine synthetase) and Fd-GOGAT (Hecht et al. 1988; Hayakawa et al. 1992) activity.

Nitrate assimilation also requires alterations in carbon metabolism. Carbohydrate synthesis is decreased, and more carbon is converted via glycolysis to phosphoenolpyruvate (PEP) and enters organic acid metabolism (Stitt & Krapp 1999). Organic acid metabolism has two different functions during nitrate assimilation. (i) Phosphoenolpyruvate carboxylase (PEPC) operates with malate dehydrogenase to provide malate, that acts as a counter-anion and prevents alkalization during nitrate assimilation (Deng, Moreaux & Lamaze 1989; Martinoia & Rentsch 1994). The malate is exported to the roots, where it is decarboxylated (Martinoia & Rentsch 1994). (ii) PEPC also operates together with pyruvate kinase (PK), the mitochondrial citrate synthase (CS) and pyruvate dehydrogenase, and the cytosolic NADP-dependent isocitrate dehydrogenase (NADP-ICDH) (Heldt 1996; Scheible et al. 1997b; Stitt & Krapp 1999) to provide 2-oxoglutarate which is the primary carbon acceptor for inorganic nitrogen assimilation via nitrate reduction and the GOGAT pathway. Nitrate induces several genes that encode enzymes involved in organic acid synthesis, including PPC (encoding PEPC), PK (encoding cytosolic PK), CS (encoding mitochondrial CS) and ICDH1 (encoding NADP-ICDH) (Scheible et al. 1997b), and represses AGPS (Scheible et al. 1997b) which encodes the regulatory subunit of ADP-glucose pyrophosphorylase (AGPase), a key enzyme for the regulation of starch synthesis, leading to an increase of PEPC activity, accumulation of 2-oxoglutarate, malate, and other organic acids (Scheible et al. 1997b), a decrease of AGPase activity, and reduced starch accumulation (Scheible et al. 1997a, 1997b).

The following experiments investigate (i) whether the diurnal changes in NIA expression are accompanied by significant changes in the expression of genes encoding enzymes involved in organic acid metabolism (ii) whether these diurnal changes contribute to the diurnal regulation of leaf metabolism, and (iii) whether they are driven by nitrogen metabolism.

MATERIALS AND METHODS

Plant material and growth conditions

All tobacco (Nicotiana tabacum) genotypes were grown in sand culture in a 12 h light (600 μmol photons m−2 s−1) : 12 h dark regime in a plant growth chamber (Scheible et al. 1997a, 1997b, 1997c). Inorganic nitrogen was provided as a mixture of the 4 mM potassium nitrate and 4 mM magnesium nitrate. When nitrate was decreased to 0·2 mM it was replaced by a mixture of potassium and magnesium sulphate and chloride to maintain the same concentration of each cation and the same overall anion concentration. Plants were watered daily to field capacity with fresh nutrient solution, 2–3 h after the start of the light period. When the solution contained 12 mM nitrate, the sand contained at least 3 mM nitrate 24 h later (data not shown). When the solution contained 0·2 mM nitrate, the added nitrate was exhausted in the following hours (data not shown).

The youngest fully expanded source leaf was harvested from plants before internode elongation started, after 32 d for the 12 mM nitrate-fed wild-type plants and F23xNia30 mutants, or after 62 d for nitrate-limited wild-type plants and 12 mM nitrate-fed Nia30(145) transformants. Plant material was harvested at ambient photon flux into liquid nitrogen and subsequently stored at − 80 °C or in liquid nitrogen (e.g. for the determination of phosphorylated metabolites).

RNA gel blot analysis

Total RNA was isolated from leaf material after removing first, second and third order veins. RNA gel-blot analyses were performed as in Krapp et al. (1993), loading equal amounts (20 μg) of total RNA in each lane. Equal loading was routinely checked by ethidium-bromide staining of the RNA-gels and hybridization of the UV-crosslinked RNA blotting membranes (Hybond N, Amersham, Braunschweig, Germany) with an 18S-rRNA cDNA probe. Radioactively hybridized filters were washed with increasing stringency and depending on the used probe until the signal : background ratio was maximal. Autoradiography was carried out at –80 °C with Kodak X-Omat films using intensifying screens and radioactivity was quantified using Fuji BAS-III S screens, a Fujix BAS-1000 phosphoimager and the software TINA 2·09 (Raytest, Straubenhardt, Germany). Hybridized RNA gel-blots were stripped by washing up to three times 1 min in 0·1% (w/v) SDS at 90 °C. Quantitative removal of radiolabelled cDNA probes was checked before using the RNA membranes for another hybridization. Probes used for this work were derived from Nicotiana tabacum (NIA, Vincentz & Caboche 1991; NII, Kronenberger et al. 1993; PPC, Koizumi et al. 1991; PK, Gottlob-McHugh et al. 1992; ICDH1, Galvez et al. 1996) and Solanum tuberosum (CS, Landschütze, Müller-Röber & Willmitzer 1995).

Enzyme assays

Nitrate reductase activity was measured in the absence of magnesium as described in Scheible et al. (1997a, 1997b). PEPC activity was extracted and measured as in Scheible et al. (1997b). Pyruvate kinase, citrate synthase and NADP-ICDH were assayed in the same extracts as in Bergmeyer (1987).

Metabolite analysis

Sucrose, glucose, fructose and starch were measured in the soluble and residual fraction of an ethanol–water extract (Scheible et al. 1997a, 1997b) as described in Stitt et al. (1989). Pyruvate, PEP, 3-PGA, α-OG, isocitrate, citrate and malate were measured in perchloric acid extracts as in Scheible et al. (1997b) and amino acids were quantified by high-performance liquid chromatography (HPLC) (Geigenberger et al. 1996). Total protein content was determined by two different methods, both described in Scheible et al. (1997a). Nitrate was determined from ethanol–water extracts by HPLC according to Gebauer, Melzer & Rehder (1984). Ammonia was measured enzymatically from 2% trichloroacetic acid extracts as in Bergmeyer (1987).

RESULTS

Experimental design

To perturb the rate and timing of nitrate assimilation and analyse the impact on diurnal gene expression and leaf metabolism we compared four sets of plant material: (i) wild-type tobacco growing on saturating (12 mM) nitrate, two genotypes with a (ii) small (F23xNia30) and (iii) large (Nia30(145)) decrease of NR activity growing on 12 mM nitrate, and (iv) wild-type tobacco growing on low (0·2 mM) nitrate. The plants were grown in a 12 h light : 12 h dark cycle and were irrigated with fresh nutrient medium 3–4 h into the light period every day. The following background information is helpful to understand the results. Wild-type tobacco contains four functional copies of the NIA gene at the NIA1 and NIA2 loci. F23xNia30 contains one functional and one dysfunctional gene copy at the NIA2 locus and two dysfunctional copies at the NIA1 locus. NR activity is reduced by about 50% during the first part of the light period, when NR activity is at the diurnal maximum in wild-type tobacco, but is similar or even higher than in wild-type tobacco at the end of the light period and during the dark period (Scheible et al. 1997c). Nia30(145) is a double null mutant transformed with a 12 kb construct containing the endogenous upstream promotor sequence, the structural gene and a downstream sequence from the tobacco NIA2 locus (Vaucheret et al. 1990). NR activity is present at a constitutive but low level (3–5% of that in wild-type plants, Scheible et al. 1997a). The non-functional genes in F23xNia30 and Nia30(145) carry point mutations and are transcribed (see Fig. 1a) and translated to produce catalytically inactive NR protein (data not shown, see Scheible et al. 1997b, 1997c). Wild-type tobacco and F23xNia30 grow at similar rates (Scheible et al. 1997a, 1997c), whereas Nia30(145) grows more slowly at a rate similar to that of nitrogen-limited wild-type tobacco (Scheible et al. 1997a, 1997c).

Figure 1a–m.

Diurnal changes of transcripts in source leaves. Transcripts were analysed by Northern blotting of total RNA prepared from the first fully expanded source leaf of wild-type tobacco (●), mutants with one instead of four functional NIA gene copies (F23xNia30; shaded circle) and null mutants re-transformed with a construct containing the structural NIA2 gene under the control of its endogenous promoter and expressing NR at 3–5% of the wild-type level (Nia30(145) (○) all growing on 12 mM nitrate, or from wild-type tobacco growing on 0·2 mM nitrate (▪). The plants were grown in sand culture in a 12-h light/12 h dark regime and were irrigated to field capacity daily with the appropriate 12 (circle) or 0·2 (square) mM nitrate nutrient solution. The youngest fully expanded leaves were harvested and extracted for total RNA after 4, 8, and 12 h illumination, and after 4, 8 and 12 h darkness. The shaded areas in panels g–m indicate the dark period. (a) and (g), nitrate reductase (NIA); (b) and (h), nitrite reductase (NII); (c) and (i), phosphoenolpyruvate carboxylase (PPC); (d) and (j), cytosolic pyruvate kinase (PK); (e) and (k), mitochondrial citrate synthase (CS); (f) and (l), NADP-dependent isocitrate dehydrogenase (ICDH1); (m) fumarase. The blots in panels a–f represent one typical result. The results in panels g–m are the mean ± SE of phospho imager scans from RNA-blots prepared from four different plants.

Diurnal changes of transcripts

The diurnal changes of transcript levels are shown in Fig. 1. Typical original data are shown in Fig. 1a–f, and the mean and standard error (SE) of four replicate preparations from four separate plants in Fig. 1g–m. The diurnal changes of the NIA transcript level (Fig. 1a & g) resemble those described previously (see above, also Vaucheret et al. 1990; Scheible et al. 1997c).

Transcripts for NII (Fig. 1b & h) and PPC (Fig. 1c & i) changed in parallel with the NIA transcript. In nitrogen-replete wild-type plants, levels of all three transcripts were high in the first 4 h of the light period, declined to very low levels in the second part of the light period, and recovered during the night. F23xNia30 contained increased levels of the NIA, NII and PPC transcripts, indeed the PPC transcript increased more than NIA (compare Fig. 1a & c). In Nia30(145), transcripts for NIA, NII and PPC were present at very high levels and the diurnal changes were partially damped (Fig. 1a–c). Transcripts for NIA, NII and PPC were present at very low levels in wild-type plants growing on 0·2 mM nitrate. They showed a small and transient increase in the first part of the light period after re-irrigation with 0·2 mM nitrate. This transient peak occured somewhat earlier for the NIA (Fig. 1a & g) and NII (Fig. 1b & h) transcripts than for the PPC transcript (Fig. 1c & i).

The transcripts for PK (Fig. 1d & j), CS (Fig. 1e & k) and ICDH1 (Fig. 1f & l) showed pronounced diurnal changes that were out of phase with those of the NIA, NII and PPC transcripts. The PK transcript rose to a maximum at the end of the light period and declined gradually during the dark period in nitrogen-replete wild-type plants (Fig. 1d & j). In Nia30(145) the absolute levels of the PK transcript were much higher, and the decrease during the night was delayed, as is already apparent in F23xNia30. The CS transcript level was low with a slight maximum towards the end of the dark period in nitrogen-replete wild-type plants (Fig. 1k). CS transcript was slightly higher in F23xNia30 and much higher in Nia30(145), where the diurnal changes were less marked because the CS transcript remained high in the light (Fig. 1e & k). The ICDH1 transcript level was lowest during the first part of the light period, rose to a maximum during the first part of the night, and declined later in the night in nitrogen-replete wild-type plants (Fig. 1f & l). A similar response was observed for F23xNia30. In Nia30(145), the ICDH1 transcript level was higher than in the wild type throughout the diurnal cycle, and reached a maximum later in the night (Fig. 1f & l).

Fumarase activity is required when the tricarboxylic acid cycle operates during respiration, but is not required when the tricarboxylic cycle operates in conjunction with PEPC to generate organic acids during nitrate assimilation. The transcript for fumarase did not show a marked diurnal rhythm, and was not markedly increased in F23xNia30 or Nia30(145) or decreased in nitrate-deficient wild-type plants (Fig. 1m).

Diurnal changes of enzyme activities

Figure 2 shows the diurnal changes of enzyme activities, related to total leaf protein. Compared to wild-type plants on 12 mM nitrate, total leaf protein was not significantly decreased in F23xNia30 growing on 12 mM nitrate, but was decreased by 50–60% in Nia30(145) growing on 12 mM nitrate and in wild-type plants growing on 0·2 mM nitrate (see legend of Fig. 2, also Scheible et al. 1997a). There were no significant and consistent diurnal changes in overall protein (data not shown).

Figure 2a–e.

Diurnal changes in the activities of enzymes in carbon and nitrogen metabolism. The activities were measured in extracts prepared from an aliquot of the leaf material used to investigate transcripts (Fig. 1). Enzyme activities were assayed in the first fully expanded source leaf of wild-type tobacco (●), mutants with one instead of four NIA gene copies (F23xNia30; shaded circle) and double null mutants that were retransformed with a construct containing the structural NIA2 gene under the control of its endogenous promoter (Nia30(145); ○) all growing on 12 mM nitrate, and in the first fully expanded leaf from wild-type tobacco growing on 0·2 mM nitrate (▪). The results are expressed on a protein basis to correct for the decrease in total leaf protein in Nia30(145) and in wild-type plants growing on low nitrate. The protein content was 23·8 ± 1·8, 22·9 ± 2·2, 9·8 ± 0·8, and 11·5 ± 0·7 mg g FW−1 in wild-types, F23xNia30 and Nia30(145) grown on 12 mM nitrate and in wild-types grown on 0·2 mM nitrate., respectively. The shaded area indicates the dark period. (a) NR (b) PEPC (c) pyruvate kinase (PK) (d) citrate synthase (CS) (e) NADP-isocitrate dehydrogenase (NADP-ICDH). The results are the mean ± SE (n = 4) of four separate plants.

NR was assayed in the absence of magnesium to override the effects of post-translational regulation on activity (Kaiser & Huber 1994). The diurnal changes of NR activity have been presented (Scheible et al. 1997c), and are included here for comparison with the other enzyme activities. In nitrate-replete wild-type plants, NR activity increased to a maximum after about 2–4 h light and decreased during the remainder of the light period (Fig. 2a). In F23xNia30, the diurnal changes were almost abolished. In Nia30(145), NR activity was low and did not show any diurnal changes. It is important to note that NR protein (which includes the product of the four point-mutated NIA genes) is present at a very high and fairly constant level in Nia30(145) (data not shown, see Scheible et al. 1997c). NR activity is very low in nitrate-limited wild-type plants but shows a transient rise after re-irrigation with 0·2 mM nitrate (Fig. 2a).

The diurnal changes of PEPC activity resemble those of NR, but are less marked. In wild-type plants on 12 mM nitrate, PEPC activity increased by 40% in the first part of the light period, declined during the remainder of the light period, and remained low during the night (Fig. 2b). The changes of PEPC activity follow, with a time lag of about 4 h, the changes of the PPC transcript level (Fig. 1c). This resembles the relation between the NIA transcript level and NR activity (see Figs 1a & 2a). The marked increase of the PPC transcript in Nia30(145) (see above) is accompanied by a two- to three-fold increase of PEPC activity (Fig. 2b) and a modification of the diurnal changes, with a maximum at the end of the night. This resembled the changes of the PPC transcript in Nia30(145). Wild-type plants growing on 0·2 mM nitrate contained very low PEPC activity (Fig. 2b), mirroring the very low level of the PPC transcript (Fig. 1c & i). The transient increase of the PPC transcript level after re-irrigation with 0·2 mM nitrate was followed by a slight increase of PEPC activity after 8–12 h illumination (Fig. 2b).

In nitrogen-sufficient wild-type plants, PK activity was highest during the dark period, started to decline towards the end of the dark period, continued to decrease during the first part of the light period to reach a minimum after 8 h illumination, and increased during the last 4 h of the light period (Fig. 2c). Plants contain isoforms of PK in the plastid and cytosol (Dennis, Kuang & Negm 1997). Because the plastid isoform represents only about 10% of the total activity in leaves (Stitt & ap Rees 1979), these large changes must be due to the cytosolic isoform. PK activity was increased in Nia30(145), in particular the decrease during the first part of the light period was abolished (Fig. 2c). The diurnal changes of PK activity in wild-type plants and Nia30(145) resemble the diurnal changes of PK transcript, which was low during the first part of the light period and high at the end of the light period in wild-type plants (Fig. 1d) and present at high and near-constitutive levels in Nia30(145) (Fig. 1d). In nitrate-limited wild-type plants, PK activity was reduced and the diurnal changes were abolished (Fig. 2c). The increase of PK activity in Nia30(145) and decrease of PK activity in nitrate-limited wild-type plants were much smaller than the corresponding changes in PEPC activity (Fig. 2b).

The CS activity in nitrogen-replete wild-type plants was highest at the end of the dark period, declined to low levels after 8–12 h light, and recovered gradually during the dark period (Fig. 2d). The timing of these changes of CS activity resembles the changes of the CS transcript, which was also highest during the dark period (Fig. 1e). The diurnal changes of CS activity were almost abolished in Nia30(145), which contained slightly higher CS activity than nitrogen-replete wild-type plants at the end of the night and start of the light period and much higher activity for remainder of the diurnal cycle (Fig. 2d). This mirrors the high and constant level of the CS transcript in Nia30(145) (Fig. 1e & k). Nitrate-deficient wild-type plants differed markedly to Nia30(145), having low CS activity except at the end of the night and start of the light period (Fig. 2d).

The NADP-ICDH activity was high during the first part of the dark period, started to decrease towards the end of the dark period, continued to decrease during the first 8 h of the light period, and started to recover during the last 4 h of the light period in wild-type plants growing on 12 mM nitrate (Fig. 2e). These changes occur in parallel with or slightly earlier than the changes of the ICDH1 transcript level, which started to increase between 8 and 12 h into the light period and was at a maximum during the first part of the dark period in well fertilized wild-type plants (Fig. 1f & m). In Nia30(145), NADP-ICDH activity at the start of the light period was as high as in nitrogen-sufficient wild-type plants, and the decrease during the first part of the light period was abolished. This matches the higher levels of the ICDH1 transcript in Nia30(145) (Fig. 1f & l). NADP-ICDH activity in wild-type plants on 0·2 mM nitrate resembled well-fertilized plants, except that the recovery was delayed until the first part of the night (Fig. 2e).

The results in Figs 1 & 2 show that the enzymes fall into two blocks with respect to the timing of the diurnal changes of transcripts and enzyme activities. NR and PEPC show a maximum for their transcripts at the start of the light period and a maximum for the enzyme activities early in the light period, whereas PK, CS and NADP-ICDH show a maximum in their transcripts and activities towards the end of the light and during the dark period.

Diurnal changes of nitrate, ammonium and amino acids in source leaves

When wild-type tobacco was grown on 12 mM nitrate, nitrate decreased during the light period and recovered during the dark period (Fig. 3a, see also Scheible et al. 1997c). In F23xNia30, nitrate was two-fold higher but still showed diurnal changes. In Nia30(145), nitrate levels were eight-fold higher and the diurnal changes were abolished. Nitrate-limited wild-type plants contained very low nitrate. (Fig. 3a and inset). Following re-irrigation with 0·2 mM nitrate nutrient solution 2–3 h into the light period, nitrate increased transiently in the leaves (from 0·11 just prior to irrigation to 0·20 μmol g FW−1 after 4 h into the light phase). Nitrate also showed a transient 10-fold increase (from 0·10 to 1·07 μmol g FW−1, data not shown) in the roots.

Figure 3a–x.

Diurnal changes of nitrate, ammonium, amino acids, carbohydrates, glycolytic intermediates and organic acids. The measurements were carried out in aliquots from the material used to determine transcripts and enzyme activities in Figs 1 & 2. See Fig. 2 for symbols. Results in panels a–i, m, o, p, q, s, t and v are given in μmoles g FW−1, panels j, k, l show ratios, results in panels n, r, w and x are given in nmoles g FW−1, and starch, in panel u, is expressed in mmoles hexose equivalents g FW−1. The shaded area indicates the dark period. (a), Nitrate (NO3), note enlarged insert for wild type grown on 0·2 mM nitrate; (b), glutamine (Gln); (c), asparagine (Asn); (d), glycine (Gly); (e), ammonium (NH4+); (f), glutamate (Glu); (g), aspartate (Asp); (h), serine (Ser); (i), sucrose; (j), glutamine/glutamate ratio; (k), asparagine/aspartate ratio; (l), glycine/serine ratio; (m), glucose; (n), glucose-6-phosphate (Glc6P); (o), malate; (p), citrate; (q), fructose; (r), glucose-1-phosphate (Glc1P); (s), isocitrate; (t), 2-oxoglutarate (αOG); (u), starch; (v), glycerate-3-phosphate (3PGA); (w), phosphoenolpyruvate (PEP); (x), pyruvate. The results are the mean ± SE (n = 4 separate plants).

Glutamine and ammonium levels varied dramatically, depending on the time of day, the genotype, and the nitrogen regime. In wild-type tobacco growing on 12 mM nitrate, large amounts of glutamine (Fig. 3b) and ammonium (Fig. 3e) accumulated in leaves during the day and were re-mobilized at night. Accumulation of glutamine and ammonium during the light period was reduced in F23xNia30, and abolished in Nia30(145) and nitrate-deficient wild-type plants. Compared with wild-type plants, glutamine and ammonium levels at the end of the light period were about 50% lower in F23xNia30, and more than 95% lower in Nia30(145) (Fig. 3b, e). In nitrate-deficient wild-type plants glutamine was very low, but did show a slight transient increase after re-supplying 0·2 mM nitrate (48, 98, 190 and 122 nmol g FW−1 after 0, 4, 8, and 12 h illumination, respectively; Fig. 3b).

Glutamate levels did not show any marked diurnal changes in nitrogen-replete wild-type plants (Fig. 3f). Similar levels were found in F23xNia30, and marginally lower levels in Nia30(145) (Fig. 3f). Glutamate also remained high in nitrate-deficient wild-type plants. This stable level of glutamate contrasts with the large changes of ammonium and glutamine (see above) and the large changes of amino acids lying downstream of glutamate (see below). The glutamine : glutamate ratio rose markedly during the light period and decreased during the dark period in wild-type plants growing on 12 mM nitrate (Fig. 3j). These changes of the glutamine : glutamate ratio were less marked in F23xNia30, and were abolished in Nia30(145), and in nitrate-deficient wild-type plants.

Asparagine (Fig. 3c) and aspartate (Fig. 3g) were present at similar levels in nitrogen-sufficient wild-type plants and F23xNia30, and at much lower levels in Nia30(145) and nitrate-deficient wild-type plants during the light period. Asparagine increased during the first part of the dark period in wild-type plants, whereas it decreased in F23xNia30 and Nia30(145) (Fig. 3c). Asparagine also rose slightly in the dark in nitrate-deficient wild-type plants. Aspartate rose in the dark in all four sets of plant material (Fig. 3g). The asparagine : aspartate ratio in F23xNia30 and Nia30(145) resembled that in wild-type plants in the light, but was much lower in the dark (Fig. 3k). The asparagine : asparatate ratio was 10- to 40-fold lower than the glutamine : glutamate ratio (compare Fig. 3j & k).

Glycine rose 10-fold during the day in wild-type plants growing on 12 mM nitrate (Fig. 3d). There was a smaller increase of serine (Fig. 3h). As a result, the glycine : serine ratio increased during the day (Fig. 3l). The increase of glycine was less marked in F23xNia30, and negligible in Nia30(145) and nitrate-deficient wild-type plants. When the four sets of plant material are compared, the magnitude and timing of the diurnal changes of glycine correlate with the diurnal changes of glutamine and ammonium (Fig. 3d and Fig. 3b & e).

Considerable amounts of reduced nitrogen accumulate in the source leaves of nitrogen-sufficient wild-type plants during the light period and are re-mobilized during the dark period. About 150 μmol nitrate g FW−1 is assimilated per day in tobacco in these growth conditions (Scheible et al. 1997c). Of this, about 34 μatoms nitrogen g FW−1 is temporarily retained as ammonium (Fig. 3e), glutamine (Fig. 3b), glycine (Fig. 3d) and serine (Fig. 3h) in the leaves and is re-mobilized during the night. This is equivalent to about 25% of the assimilated nitrate. In comparison with wild-type tobacco, only half as much (17 μatoms nitrogen g FW−1) accumulates in the leaves of F23xNia30 during the light phase. Negligible amounts of assimilated nitrogen accumulate in Nia30(145) or in nitrate-deficient wild-type plants

Diurnal changes of carbohydrates

In nitrogen-sufficient wild-type plants, sucrose accumulated during the light period and decreased during the dark period (Fig. 3i). Glucose (Fig. 3m) and fructose (Fig. 3q) increased to a maximum during the first 4 h of the light period, decreased during the second part of the light period, and remained low during the night. Starch accumulated rapidly during the second part of the light period, and was almost fully re-mobilized during the night (Fig. 3u).

Sucrose levels were slightly lower in F23xNia30, and much lower in Nia30(145) (Fig. 3i). The transient accumulation of glucose and fructose was reduced in F23xNia30, and almost abolished in Nia30(145) (Fig. 3m & q). The rate of starch accumulation in F23xNia30 resembled that in nitrogen-replete wild-type plants (Fig. 3u), whereas it was much slower in Nia30(145) (Fig. 3u). Nitrate-limited wild-type plants contained similar levels of sucrose (Fig. 3i), glucose (Fig. 3m) and fructose (Fig. 3q) to nitrogen-replete plants at the start of the light period, but less sucrose accumulated during the light period (Fig. 3i) and the transient accumulation of glucose and fructose in the first part of the light period was abolished (Fig. 3m & q). Starch was high and did not show any marked diurnal changes (Fig. 3u, see also Scheible et al. 1997a, 1997b).

Diurnal changes of phosphorylated intermediates

In nitrogen-sufficient wild-type plants, glucose-6-phosphate (Glc6P, Fig. 3n), fructose-6-phosphate (data not shown), and glucose-1-phosphate (Fig. 3r) were present at similar levels in the light and dark. Glycerate-3-phosphate (3PGA, Fig. 3v) and phosphoenolpyruvate (PEP, Fig. 3w) were much higher in the light period than in the dark period, as seen for other species (Stitt et al. 1987; Stitt 1996).

In comparison with wild-type plants, for most of the diurnal cycle hexose phosphates were slightly reduced in F23xNia30 and markedly reduced in Nia30(145) (Fig. 3n & r). In the light, 3PGA and PEP were slightly lower in F23xNia30 and much lower in Nia30(145) (Fig. 3v & w). The low 3PGA in Nia30(145) correlates with the slow rate of starch accumulation (see Fig. 3u). In the dark, 3PGA and PEP in F23xNia30 and Nia30(145) were not markedly lower than in wild-type plants (Fig. 3v & w).

Diurnal changes of organic acids

In nitrogen-replete wild-type plants, malate and citrate were present at 100–200 fold higher levels than the other organic acids (compare Fig. 3o, p, s & t). During the light period, malate rose by about 40 μmol g FW−1 and citrate decreased by about 10 μmol g FW−1. These changes were reversed during the dark period. The diurnal fluctuations of malate and citrate are substantial, compared to the estimated rate of nitrate assimilation (about 150 μmol nitrate g FW−1 per day, see above) and the amount of nitrogen temporarily retained in ammonium, glutamine and other amino acids (about 30 μatoms N g FW−1, see above). Other organic acids were present at much lower levels, and showed smaller diurnal changes. Pyruvate increased by approximately 50% during the day and decreased during the night (Fig. 3x). Isocitrate increased gradually during the day, and decreased during the night (Fig. 3s). 2-Oxoglutarate increased by about 30% after illumination and decreased at the end of the light period (Fig. 3t).

In F23xNia30, the overall levels of organic acids resembled those in wild-type plants, but the diurnal changes were subtly modified. Malate was slightly higher at the end of the light period and for the most of the dark period (Fig. 3o), pyruvate was lower for most of the light period (Fig. 3x), citrate decreased further during the light period (Fig. 3p), and isocitrate was higher during the dark period (Fig. 3s). 2-Oxoglutarate levels were similar to those in wild-type plants (Fig. 3t). Nia 30(145) shows dramatic changes in the levels and diurnal changes of the organic acids. Pyruvate (Fig. 3x), citrate (Fig. 3p), isocitrate (Fig. 3s) were decreased, 2-oxoglutarate was much higher (Fig. 3t), and malate was far lower (Fig. 3o) than in wild-type tobacco.

DISCUSSION

Expression of genes that encode enzymes involved in organic acid metabolism shows marked diurnal changes in leaves that are strongly modified in response to changes in nitrogen metabolism

There are large diurnal changes of the levels of the transcripts for PPC, PK, CS and ICDH1 in tobacco leaves, which are accompanied by changes of the activities of the encoded enzymes. This temporal agreement establishes that the diurnal changes of transcript levels contribute to the changes of enzyme activity, but obviously does not exclude further contributions from post-transcriptional regulation and protein turnover, as are known to occur for NIA (Nussaume et al. 1995; Weiner & Kaiser 1999; Stitt & Krapp 1999).

The diurnal changes of PPC, PK, CS and ICDH1 expression are driven by signals derived from nitrate and/or the metabolism of nitrate. This link is demonstrated by the modified diurnal changes in nitrate-limited wild-type plants and their suppression in genotypes with low NR activity (see Figs 1 & 2). It is especially revealing to compare nitrate-limited wild-type plants with Nia30(145). Both have similar rates of growth and levels of amino acids, protein and sugars, but differ with respect to the timing of nitrate assimilation (Fig. 3 and Scheible et al. 1997c). In nitrate-limited wild-type plants, the daily addition of 0·2 mM nitrate is followed by a transient increase of NR activity, a slight transient increase of PPC transcript and PEPC activity, and a transient decrease of CS and NADP-ICDH activity. Nia30(145) has low NR activity and high nitrate and low glutamine throughout the day and night, and has generally higher activities of PPC, CS and NADP-ICDH which do not show these transient changes.

NIA, NII and PPC differ from PK, CS and ICDH with respect to their diurnal changes of expression and their sensitivity to the nitrogen supply

The genes can be divided into two blocks, that differ with respect to the timing of the diurnal change, and the sensitivity with which expression responds to changes in nitrogen metabolism. The first group includes NIA (see also Introduction), NII (see also Kronenberger et al. 1993) and PPC, and shows two main characteristics. First, the transcript level is highest at the start of the light period and the encoded enzyme activity is highest during the first part of the light period (Figs 1a–c, 2a & b). Second, the transcript level and activity of the encoded enzyme responds very strongly to changes in nitrogen metabolism. NIA, NII and PPC transcript levels (Fig. 1a–c), PEPC activity (Fig. 2b) and NR protein (Scheible et al. 1997c) are dramatically increased in Nia30(145), and very low in nitrate-limited wild-type plants.

The second group includes PK, CS, and ICDH1. For this group, transcript levels and encoded activities are highest at the end of the light period and first part of the dark period (Figs 2c–e & 1d–f). Also, although expression is affected by manipulation of nitrogen metabolism, the changes are less dramatic than for the first group of enzymes. Thus, although transcript levels for PK, CS, and ICDH1 and activities of PK, CS and IDCH are higher in Nia30(145) than in nitrogen-sufficient wild-type plants, this is due to attenuation of the decrease of their transcripts during the day rather than a general increase of transcript levels. A similar picture holds for the changes of enzyme activity. During the night, Nia30(145) and nitrogen-replete wild-type plants have comparable activities of PK, CS and NADP-ICDH. During the light period, Nia30(145) has about two-fold higher activities of PK, CS and NADP-ICDH than nitrogen-replete wild-type plants, because the activities do not decrease in the light period in Nia30(145) whereas they do in wild-type plants.

The diurnal changes in expression are accompanied by accumulation of malate, depletion of citrate and accumulation of proximal nitrogen metabolites during the day, and reversal of these changes during the night

The differing diurnal rhythms of PPC expression and PK, CS, and ICDH expression are accompanied by marked changes in metabolism. During the first part of the light period, when NR and PEPC activity are at a maximum and PK, CS and NADP-ICDH activity are declining (Fig. 2), malate accumulates (Fig. 3o), whereas citrate decreases (Fig. 3p) and glutamine and ammonium accumulate (Fig. 3b & e). Later in the light period and during the dark period when NR and PEPC activity are lower and PK, CS and NADP-ICDH activity are at a maximum, malate decreases, citrate increases and ammonium and glutamine decrease.

Further evidence that the diurnal changes of enzyme activities affect metabolism is provided by the altered diurnal responses of organic acids and amino acids in F23xNia30 and Nia30(145). In F23xNia30, PPC transcript is increased at the start of the light period (Fig. 1c), and the decline of the transcripts for PK, CS, and ICDH1 at the end of the dark period is delayed and less marked than in wild-type plants (Fig. 1d–f & j–l). During the day, 3PGA, PEP and pyruvate are decreased (Fig. 3v–x), citrate is depleted more rapidly (Fig. 3p) and malate accumulates at a similar or slightly faster rate (Fig. 3o) than in wild-type plants. During the night, there is a more sustained increase of isocitrate (Fig. 3s) than in wild-type plants. In Nia30(145), where all four enzymes are strongly expressed (Fig. 2), there is a marked decrease of the glycolytic precursors and a dramatic increase of 2-oxoglutarate (Fig. 3o, p, s, t & v–x).

The shift in metabolism that accompanies the reciprocal changes of PPC expression and PK, CS, and ICDH1 expression in wild-type plants allows part of the demand for carbon during nitrate assimilation to be shifted to a later time in the diurnal cycle. The decrease of the citrate pool during the day is equivalent to about 10% of the total carbon required during the conversion of nitrate to amino acids. The accompanying increase of ammonium and glutamine allows about 30% of the assimilated nitrate to be temporarily accumulated in a form that contains no, or relatively low, carbon.

Coordinate expression of PPC and NIA expression favours malate synthesis for pH regulation when nitrate assimilation is rapid

The reciprocal changes of PPC expression and PK, CS and ICDH1 expression also have interesting implications for pH regulation. Nitrate assimilation leads to alkalinization that (see INTRODUCTION) is counteracted by synthesis of malate, which is exported to the roots and decarboxylated (Martinoia & Rentsch 1994). During nitrate assimilation when the tricarboxylic acid cycle operates in the anapleurotic mode, malate synthesis occurs via PEPC and does not require PK, CS and NADP-ICDH activity. Indeed, net malate synthesis will be prevented if these enzymes are present at high activity relative to PEPC, because they convert PEP and C-4 acids into citrate, isocitrate and 2-oxoglutarate.

Due to the low buffer capacity in the leaf, processes that contribute to pH regulation will have to be tightly tied to the current rate of nitrate assimilation. On the other hand, provision of 2-oxoglutarate may not have to be so tightly linked because a temporary imbalance can be buffered via changes in the pools of ammonium, glutamine and the major organic acids (see above). This could provide one reason why the diurnal changes of NIA, NII and PPC transcripts and encoded activities are tightly linked, and are out of phase with diurnal changes of the enzymes that are required for 2-oxoglutarate synthesis but compete with malate synthesis. These reciprocal changes channel carbon with high priority into malate to prevent alkalization when nitrate is being rapidly assimilated.

Some aspects of our results cannot be explained in terms of the changes in expression and the overall enzyme activities. The high activity of PEPC in Nia30(145) is accompanied by a low rate of malate accumulation. This indicates that high PEPC activity in vivo requires further signals, which are generated as a result of nitrate assimilation and presumably activate PEPC via ‘fine’ regulation or post-translational modification. Possible candidates include changes of pH, amino acids, organic acids or protein phosphorylation (Huppe & Turpin 1994; Chollet, Vidal & O'Leary 1996).

Increased expression of PK, CS and NADP-ICDH allows remobilization of accumulated ammonium and glutamine later in the day

The gradual accumulation of glutamine and ammonium during the light period (Fig. 3b & e) is accompanied by a parallel accumulation of glycine (Fig. 3d). There is a good correlation between the accumulation of ammonium and glutamine and the accumulation of glycine when genotypes with different levels of NR, or nitrate-limited and nitrogen-replete plants are compared (Fig. 3e, b & d). This indicates that de novo nitrate assimilation interacts closely with the recycling of ammonium released by glycine decarboxylase during photorespiration.

The progressive accumulation of ammonium, glutamine and glycine during the day in nitrogen-replete wild-type plants correlates with decreasing expression of NIA and PPC and increasing expression of PK, CS and ICDH1. These changes of expression will restrict de novo nitrate assimilation and ammonium formation, and divert carbon away from malate synthesis and towards synthesis of 2-oxoglutarate to support metabolism of the ammonium and glutamine. As will be shown elsewhere (P. Matt, A. Krapp and M. Stitt, unpublished), transcript levels for GLN2 (encoding plastid glutamine synthetase) and total glutamine synthetase activity change in parallel with PK, CS and NADP-ICDH being low in the morning and highest at the end of the light period. This indicates that the changes of expression of the enzymes in organic acid metabolism are part of a larger programme that allows coordinate regulation of de novo nitrate and ammonium metabolism and its integration with the larger flows of ammonium associated with photorespiration.

Ammonium is assimilated via reductive amination of glutamate followed by transfer of an amino group from glutamine to 2-oxoglutarate to form two molecules of glutamate. The glutamate (Fig. 3f) and 2-oxoglutarate (Fig. 3t) pools are remarkably stable throughout the diurnal cycle and between the various genotypes and growth conditions, except for the dramatic increase of 2-oxoglutarate in Nia30(145). This contrasts with the large changes of metabolite pools upstream and downstream of glutamate and 2-oxoglutarate, including glutamine (Fig. 3b), pyruvate, malate, citrate and isocitrate (Fig. 3x, o, p & s), aspartate, asparagine, glycine and serine (Fig. 3g, c, d & h). Maintenance of adequate pools of 2-oxoglutarate and glutamate may be important for effective operation of the GOGAT pathway.

Possible signals underlying the diurnal changes in expression

The diurnal changes of the NIA, NII and PPC transcript levels (Fig. 1a–c) show a robust correlation with leaf nitrate levels during the diurnal rhythm, between different genotypes, and when nitrate-limited and nitrogen-sufficient plants are compared (Fig. 3a). Although PPC expression is also increased by ammonium and glutamine (Hirel et al. 1987; Sugiharto & Sugiyama 1992; Sugiharto et al. 1992) this cannot play a major role in the diurnal regulation of PPC expression in tobacco growing on nitrate. The PPC transcript and PEPC activity are inversely correlated to ammonium and glutamine during the diurnal cycle in wild-type plants (compare Figs 1c & 2b with Fig. 3e & b, and in genotypes with different levels of NR (compare Figs 1e & 2b with Fig. 3e & b).

NIA, NII and PPC expression (Figs 1 & 2) decrease during the light period in nitrogen-replete wild-type plants. This decrease is much less marked in Nia30(145), implying it is due to feedback signals generated during nitrate assimilation. Although the repression of NIA and NII is usually explained as a consequence of the glutamine accumulation (see INTRODUCTION), this may not be the sole explanation as these transcripts decrease in a similar manner in F23xNia30 (Fig. 1g & h) even though glutamine was lower and increased more slowly during the light period in the F23xNia30 than in wild-type plants (Fig. 3b). A similar discrepancy is found when wild-type plants are compared in ambient and elevated carbon dioxide (Geiger et al. 1998). Further, even though GOGAT-deficient Arabidopsis mutants accumulate large amounts of glutamine, NIA expression is similar to wild-type plants (Dzuibany et al. 1998). The decrease of the PPC transcript during the light period can also hardly be explained via glutamine (see above). Our results therefore indicate that other downstream metabolites formed during nitrate assimilation inhibit NIA, NII and PPC expression, or that the decrease of nitrate in the leaves is large enough to account for the decline of expression during the day.

Three lines of evidence indicate that the diurnal changes of PK, CS and ICDH1 expression are related to downstream events in nitrate assimilation. First, the diurnal changes of the three transcripts in wild-type plants are almost inversely related to leaf nitrate content. Second, there is a time lag of several hours between the diurnal peak of NIA expression and the peak for PK, CS and ICDH1 expression (compare Fig. 1a–c & d–f, Fig. 2a, b & c–e). Third, and most important, when wild-type plants are compared with F23xNia30 and Nia30(145) there is a negative correlation between PK, CS and ICDH1 expression (Fig. 1f–h) during the first part of the light period and NR activity at this time (Fig. 2a). This indicates that some event during nitrate assimilation represses PK, CS and ICDH, and that this repression is reversed later in the light period.

One intriguing possibility is that NIA, NII and PPC and PK, CS and ICDH1 respond in a reciprocal manner to changes of pH and/or malate. NIA and PPC transcript levels (Fig. 1) are often negatively and PK, CS and ICDH1 transcript levels often positively correlated with malate (Fig. 3o). These two sets of genes may also be differentially affected by ammonium, glutamine or glycine. Although NIA (Cheng et al. 1991; Vincentz et al. 1993; Morcuende et al. 1998), NII (Kronenberger et al. 1993), PK, CS and ICDH1 (Krapp et al. 1993; Fieuw et al. 1995; Koch 1996) are induced by sugars, the levels of the transcripts for NIA, NII, PPC, PK, CS and ICDH1 do not correlate with sugar levels during the diurnal cycle in nitrogen-replete wild-type plants (Fig. 3i, m & q) or in genotypes with different NR activities.

In conclusion, four important enzymes in organic acid metabolism show marked diurnal rhythms in their transcript levels and the activities of the encoded enzymes. The absolute level of expression and the diurnal rhythms are modified in nitrate-deficient wild-type plants and in genotypes with decreased NR activity, showing that the diurnal changes in expression are driven by signals deriving from nitrate and nitrate assimilation. The enzymes fall into two broad groups with respect to the timing of the diurnal changes. PPC changes in parallel with NIA and NII, being maximal in the early part of the light period, whereas expression of PK, CS and ICDH1 is highest at the end of the light period and during the dark period. Reciprocal expression of these genes allows the carbon demand associated with nitrate assimilation to be spread over a longer time period, prioritizes malate synthesis to allow pH regulation during the first part of the day when NIA expression is maximal, and prioritizes synthesis of 2-oxoglutarate later in the diurnal cycle when large amounts of ammonium and glutamine have accumulated.

ACKNOWLEDGMENTS

This work was supported by the Deutsche Forschungsgemeinschaft (Sonderforschungsbereich 199). We are grateful to David Dennis, Bertrand Hirel, Michel Hodges, Bernd Müller-Röber, Uwe Sonnewald, and Yasuyuki Yamada for providing us with several of the cDNA-probes.

Ancillary

Advertisement