Phosphoenolpyruvate Carboxylase Intrinsically Located in the Chloroplast of Rice Plays a Crucial Role in Ammonium Assimilation

Home , Nitrogen assimilation

Phosphoenolpyruvate carboxylase intrinsically located in the chloroplast of rice plays a crucial role in ammonium assimilation

Chisato Masumotoa,1, Shin-Ichi Miyazawaa,1, Hiroshi Ohkawaa,2, Takuya Fukudaa, Yojiro Taniguchia,3, Seiji Murayamaa,4, Miyako Kusanob, Kazuki Saitob, Hiroshi Fukayamaa,5, and Mitsue Miyaoa,6

aPhotobiology and Photosynthesis Research Unit, National Institute of Agrobiological Sciences, Kannondai, Tsukuba 305-8602, Japan; and bRIKEN Plant Science Center, Yokohama, Kanagawa 230-0045, Japan

Edited by Elisabeth Gantt, University of Maryland, College Park, MD, and approved February 4, 2010 (received for review November 12, 2009) Phosphoenolpyruvate carboxylase (PEPC) is a key enzyme of primary encoded by a small gene family (5). Treatments that lead to ele- metabolism in bacteria, algae, and vascular plants, and is believed to vation of the PEPC activity, such as enhanced nitrogen assimilation Oryza sativa be cytosolic. Here we show that rice ( L.) has a plant-type and Pi starvation in C3 plant leaves (6, 7), induce expression of PEPC, Osppc4, that is targeted to the chloroplast. Osppc4 was ex- PEPC kinase gene(s) and phosphorylation of PEPC (8, 9). These pressed in all organs tested and showed high expression in the leaves. mechanisms strictly regulate PEPC activity, which is essential for Its expression in the leaves was confined to mesophyll cells, and plants because of the irreversible nature of the reaction. Osppc4 accounted for approximately one-third of total PEPC protein In addition to the plant-type PEPC, vascular plants have in theleaf blade.RecombinantOsppc4 wasactivein thePEPCreaction, another isozyme, a bacterial-type PEPC that lacks the conserved V showing max comparable to cytosolic isozymes. Knockdown of phosphorylatable Ser residue (10). Whereas most PEPCs exist as Osppc4 expression by the RNAi technique resulted in stunting at homotetramers (class 1 PEPC) (1), bacterial-type PEPCs associate the vegetative stage, which was much more marked when rice plants with plant-type isozymes to form a heteromeric complex (class 2 were grown with ammonium than with nitrate as the nitrogen PEPC) in unicellular green algae (11) and castor oilseeds (12). It source. Comparison ofleaf metabolomesofammonium-grown plants has been recently shown that Class-2 PEPC is composed of four PLANT BIOLOGY suggested that the knockdown suppressed ammonium assimilation and subsequent amino acid synthesis by reducing levels of organic each of the bacterial- and plant-type isozymes (13). acids, which are carbon skeleton donors for these processes. We also Although PEPC genes and proteins have been extensively identified the chloroplastic PEPC gene in other Oryza species, all of studied in a variety of organisms, the enzymes reported so far are which are adapted to waterlogged soil where the major nitrogen all cytosolic (2). Here we report a previously undescribed plant- source is ammonium. This suggests that, in addition to glycolysis, type PEPC targeted to the chloroplast of rice. It is also unique in its the genus Oryza has a unique route to provide organic acids for primary structure, not belonging to any of known groups of the ammonium assimilation that involves a chloroplastic PEPC, and that plant-type PEPCs. This chloroplastic isozyme is one of the major this route is crucial for growth with ammonium. This work provides PEPCs in leaf mesophyll cells and plays a crucial role in ammo- evidence for diversity of primary ammonium assimilation in the nium assimilation in rice plants. leaves of vascular plants. Results amino acid synthesis | glycolysis | nitrogen assimilation | organic acid Rice Gene Encoding PEPC Targeted to Chloroplast. Searches of rice synthesis | Oryza genome databases identified six putative PEPC (Osppc) genes (Fig. S1). Five (Osppc1, 2a, 2b, 3, and 4) encode the plant-type hosphoenolpyruvate carboxylase (PEPC; EC 4.1.1.31) cata- PEPC that has the conserved phosphorylation domain and the C- Plyzes irreversible β-carboxylation of phosphoenolpyruvate terminal tetrapeptide QNTG (1, 3, 14), and the other (Osppc-b) − 2+ (PEP) in the presence of HCO3 and Mg to yield oxaloacetate encodes a bacterial-type enzyme with the C-terminal tetrapeptide (OAA) and inorganic phosphate (Pi). PEPC is widely distributed RNTG (15). As reported for Arabidopsis (10), all of the plant-type in all photosynthetic organisms including vascular plants, algae, Osppc genes have similar exon-intron structures that basically cyanobacteria, and photosynthetic bacteria, and also in non- contain 10 exons, whereas the bacterial type has a distinct structure photosynthetic bacteria and protozoa (1). In vascular plants, the (Fig. 1A), an indication of a different evolutionary lineage. The six reaction catalyzed by PEPC is the primary fixation step of photosynthetic CO2 assimilation in C4 photosynthesis and crassulacean acid metabolism (CAM). In most nonphotosynthetic tissues and in Author contributions: C.M., S.-I.M., H.O., H.F., and M.M. designed research; C.M., S.-I.M., the leaves of C3 plants, the primary function of PEPC is anapler- H.O., T.F., Y.T., S.M., M.K., K.S., H.F., and M.M. performed research; C.M., S.-I.M., M.K., otic, replenishing the tricarboxylic acid (TCA) cycle with inter- H.F., and M.M. analyzed data; and C.M., S.-I.M., M.K., H.F., and M.M. wrote the paper. mediates that are withdrawn for a variety of biosynthetic pathways The authors declare no conflict of interest. and nitrogen assimilation (1, 2). PEPC also plays roles in the This article is a PNAS Direct Submission. provision of malate in guard cells and legume root nodules (3). Freely available online through the PNAS open access option. Consistent with this wide range of roles, PEPC is encoded by a 1C.M. and S-I.M. contributed equally to this work. small gene family and expressed in various tissues of plants (4). 2Present address: Faculty of Agriculture and Life Science, Hirosaki University, Hirosaki 036- Most PEPCs are allosteric enzymes the activity of which is regu- 8561, Japan. lated by a variety of metabolic effectors (1). The activity of vascular 3Present address: National Institute of Crop Science, Tsukuba 305-8517, Japan. plant PEPC is further regulated through reversible phosphorylation 4Present address: National Agricultural Research Center for Hokkaido Region, Sapporo of a conserved Ser residue near the N terminus (3). Upon phos- 062-8555, Japan. phorylation, PEPC becomes less sensitive to its feedback inhibitor 5Present address: Graduate School of Agriculture, Kobe University, Kobe 657-8501, Japan. malate and more sensitive to the activator glucose 6-phosphate 6To whom correspondence should be addressed. E-mail: [email protected]. (Glu6P), thereby attaining higher activity (3). Phosphorylation of This article contains supporting information online at www.pnas.org/cgi/content/full/ PEPC is catalyzed by a dedicated PEPC kinase, which is also 0913127107/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.0913127107 PNAS Early Edition | 1of6 Downloaded by guest on September 27, 2021 A ppsO 1c a2cppsO bk7.3

ppsO b2c bk1 3cppsO

4cppsO

-cppsO b

B pStRSLBL C Ro to epyt liatxoF psO p 4c Osppc1 25 tellim eziaM psO p 1c ppsO c2a 52 toor 001 1.0 M zia Ce 3 psO cp b2 Whe ta 2525 47 100 77 enacraguS sO pp b2c 69 001 100 C ppsO c3 30 001 01 0 4 psO p a2c 1 00 59 muhgroS ppsO c4 52 C4 enacraguS CeziaM 4 b-cppsO C 03 3 3cppsO C4 epyt C epyt cARsO 1 3 F xo t ia l hR odes 52 im ll te C4 rg ass

Fig. 1. Comparison of rice PEPC (Osppc) genes. (A) Exon–intron structures. The coding and noncoding regions are represented by white/colored and black boxes, respectively. Thin lines show splitting/conjunction of exons among plant-type genes. (B) Expression in rice organs. Total RNA was subjected to RT-PCR. Expression of the rice actin gene (OsRAc1) was examined as a control. The numbers of PCR cycles are indicated on the right. LB, leaf blade; LS, leaf sheath; Rt, root; Sp, spikelet. (C) Phylogenetic relationships of plant-type PEPCs from Gramineae. Presence or absence of the N-terminal extension of Osppc4 does not affect the result. Bootstrap values are shown at nodes. Details in SI Text.

putative PEPC genes are all expressed in rice plants, but with Osppc4, a Major PEPC in Leaf Mesophyll Cells. RT-PCR analyses (Fig. different organ specificities (Fig. 1B). 1B) showed that Osppc4 was expressed in all organs tested, with the The deduced amino acid sequences of the plant-type genes are highest expression in the leaf blade. The cellular specificity of highly homologous to one another (Fig. S1), sharing amino acid Osppc4 expression was examined by histochemical staining of identities above 73%. One apparent difference among them is a β-glucuronidase (GUS) activity in transgenic rice introduced with deletion of 42 amino acid residues in the N-terminal region of the Osppc4::GUS chimeric gene (Fig. 2C). In the leaf blade and the Osppc3. Another notable feature is an extension of about 40 leaf sheath, Osppc4 expression was confined to green parenchymal amino acid residues at the N terminus of Osppc4. The plant-type cells (mesophyll cells), and no expression was detected in epidermis, fi PEPCs studied so far are classi ed into three groups, the C3,C4, vascular bundles, or guard cells. We also detected expression in and root types (2). Phylogenetic analysis of plant-type PEPCs from green cross cells of the ovary, although the expression was low. No Gramineae (Fig. 1C) indicates that Osppc2a, 2b, and 3 belong to the C3 type and Osppc1 to the root type, whereas Osppc4 does not belong to any of known types. To examine whether the N-terminal extension of Osppc4 acts as A B stsalporolhC a transit peptide, we expressed a protein fusion between the N- aDk SMT fl 511 terminal portion of Osppc4 and green uorescent protein (GFP) 111 in dayflower epidermal cells. GFP fluorescence completely over- 701 201 lapped with chlorophyll (Chl) fluorescence in a guard cell (Fig. GlPF hCegreM 2A), indicating that the extension acts as a chloroplast targeting CPEP-itna peptide. We used immunoblotting to examine the presence of C i PEPC protein in isolated chloroplasts. We detected four immu- iii noreactive bands in total leaf soluble protein, but one of the four bands was highly concentrated in a stromal fraction of the chloroplast preparation (Fig. 2B). These results indicate that rice has a PEPC protein inside the chloroplast. ii Recombinant Osppc4 lacking its predicted transit peptide was active in the PEPC reaction (Table 1). Its Vmax values at pH 7.3 and iv 8.0 were comparable to those of recombinant Osppc2a, a major cytosolic isozyme in rice leaves. The affinity for the substrate PEP and the sensitivity to the inhibitor malate, however, were lower than Fig. 2. Subcellular and cellular localization of Osppc4. (A) Expression of the fl those of the C3 and root types, but rather comparable to C4-type Osppc4::GFP protein fusion in guard cells of day ower epidermis. (B) Detec- values. It was also much less sensitive to the amino acid inhibitors tion by immunoblotting of a PEPC protein in intact chloroplasts isolated from Asp and Glu and the activator Glu6P as compared with Osppc2a. rice leaf blades. M and S, membrane and stromal fractions, respectively, of the Taking these results together, we conclude that Osppc4 enc- chloroplast preparation; T, total leaf soluble protein. Numbers on left indicate estimated relative molecular masses. (C) Expression of Osppc4::GUS.(i–iii) odes a functional PEPC protein that is targeted to the chlor- Cross-sections of leaf blade, leaf sheath, and an ovary 5 d after flowering, oplast. Other types of plastids may also have PEPC, because respectively. (Bar, 100 μm.) (iv) Stoma of leaf epidermis. (Bar, 20 μm.) GUS Osppc4 was also expressed in roots (Fig. 1B) and etiolated leaves. reaction times were 2, 2, 12, and 12 h for i, ii, iii, and iv, respectively.

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.0913127107 Masumoto et al. Downloaded by guest on September 27, 2021 Table 1. Kinetic characteristics of recombinant PEPC proteins Inhibition/activation at −1 Vmax (unit·mg ) Km (PEP) (mM) pH 7.3 (fold) Protein I50 (malate) at pH 7.3 (mM) pH 7.3 pH 8.0 pH 7.3 pH 8.0 Asp Glu Glu6P

Osppc2a 16.9 21.0 0.16 0.04 0.06 0.07 0.07 2.51 Osppc4 18.0 19.1 2.00 1.03 0.7 0.86 0.71 1.17 Maize root type 21.8 28.3 0.05 0.04 0.24 —– —– 1.08

Maize C4 type 18.2 23.0 1.48 0.59 0.82 —– —– 2.44 Flaveria C3 type —– 27 —– 0.06 0.08* —– —– —–

Flaveria C4 type —– 29 —– 0.65 1.2* —– —– —–

Recombinant Osppc2a containing the entire protein and recombinant Osppc4 lacking the predicted transit peptide portion (43 amino acid residues at N terminus) were subjected to PEPC activity assay. Activities in presence of metabolite effector (5 mM Asp, 5 mM Glu, or 10 mM Glu6P) were presented as fold changes over those with no additives. Data represent means of two to three measure- − ments. Maize and Flaveria data are taken from references 16 and 17, respectively. One unit corresponds to 1 μmol min 1. *Measured at pH 7.6.

+ − expression was detected in roots even after a prolonged GUS with NH4 and NO3 , respectively. The stunting was characterized reaction, probably because of very low expression levels in this by a reduction in lamina area, which was more marked than the organ. Osppc2a was also highly expressed in the leaf blade (Fig. 1B). decrease in dry weight (Fig. 3B). The growth analysis (Fig. S4) Its expression was detected in leaf mesophyll cells as well as vascular showed that the knockdown signiﬁcantly decreased the leaf area bundles and guard cells and in various tissues of the ovary (Fig. S2). ratio. By contrast, the net assimilation rate was not signiﬁcantly Such diverse expression of Osppc2a agrees with the wide range of affected. The photosynthetic CO2 assimilation rate on the leaf area roles suggested for the cytosolic PEPC (2, 3). basis, measured from gas exchange, was not at all affected by the The level of Osppc4 was estimated from PEPC activities of knockdown irrespective of the nitrogen source used. Photosynthetic

transgenic rice plants, termed 4i, in which expression of Osppc4 characteristics such as dependence on intercellular CO2 concen- PLANT BIOLOGY was knocked down by the RNAi technique. The maximum PEPC tration and light intensity in both ambient (21%) and low (2%) O2 activity of the leaf blade, measured in the presence of Glu6P, was remained unchanged, indicating that the knockdown did not affect decreased in the knockdown to about two-thirds of the activity in either CO2 assimilation or photorespiration. Levels of chlorophyll the wild type (Fig. S3), indicating that Osppc4 accounted for and soluble protein on the leaf area basis remained unaffected. about one-third of total PEPC protein. Similarly, Osppc2a was These results suggest that the stunting results from the reduction in estimated to account for about one-half of total PEPC protein. lamina area rather than changes in photosynthetic properties of the leaf blade. Effects of Osppc4 Knockdown on Growth. To study the function of Nitrogen is highly invested in the photosynthetic machinery Osppc4, we compared nontransgenic rice and the Osppc4 knock- (18), and the leaf blade has a much higher nitrogen concentration down lines. Stunting at the vegetative stage was a phenotypic change than the leaf sheath and the root in rice plants (19). It is possible visible in the knockdown lines (Fig. S3). To characterize the stunting that the Osppc4 knockdown limits enlargement of the lamina area more precisely, we performed a growth analysis using plants grown by suppressing nitrogen uptake or assimilation. In fact, the nitro- + − hydroponically with either ammonium (NH4 ) or nitrate (NO3 )as gen uptake rate, determined from the nitrogen content of the the nitrogen source. We used the knockdown line 4i-2 showing a whole plant, was slightly reduced by the knockdown (Fig. S4). To larger reduction in the leaf PEPC activity (Fig. S3). examine whether the knockdown affected uptake or assimilation + + + + Rice plants preferentially use NH4 and thrive better with NH4 of NH4 by the root, we compared NH4 concentrations in the − ﬁ + than with NO3 . The knockdown signi cantly inhibited the growth shoot xylem sap (Fig. 3C). When plants grown with NH4 were + − + with NH4 and, to a much lesser extent, with NO3 (Fig. 3A). Plant transferred to a fresh culture medium, the NH4 concentration dry weight after cultivation for 25 days was reduced by 19% and 6% increased from 0.05 to 0.3–0.4 mM in both nontransgenic rice and

04 0 A B C 6.0 2 * ** ) 2 003 4.0 002 ** 1 **

Dry weight (g) weight Dry 0 2. Lamina area (cm 001 * in shoot xylem sap(mM) + * 4 0 0 * NH 0 42812160 42812160 retfAerofeB Ti em (day) iT ad(em y) refsnart

+ Fig. 3. Effects of the Osppc4 knockdown on growth and NH4 concentration of the shoot xylem sap. (A and B) Seedlings of nontransgenic rice (closed + − symbols) and homozygous knockdown line of T4 generation (4i-2; open symbols) were transferred to culture solution containing either NH4 (circles) or NO3 (triangles) (day 0) and cultivated for 25 days at the ﬁnal nitrogen concentration of 1 mM. Five plants were harvested every 6–7 days after the transfer, and plant dry weight (A) and lamina area (B) were determined. Data represent means ± SD. (C) Nontransgenic rice (closed bars) and the knockdown line 4i-2 (open + + bars) were grown hydroponically with 1 mM NH4 until the ninth leaf had emerged, transferred to a fresh culture solution containing 2 mM NH4 at 11:00, and incubated for 6 h in the light. Shoot xylem sap was collected from a stump of the main culm before and after transfer. Data represent means of two plants before transfer and means ± SD of four plants after transfer. *P < 0.05; **P < 0.01 between nontransgenic rice and 4i-2 by Student’s t test.

Masumoto et al. PNAS Early Edition | 3of6 Downloaded by guest on September 27, 2021 the knockdown line. These results indicate that the knockdown The reason for increased levels of the photorespiratory inter- + does not substantially affect NH4 uptake and assimilation by the mediates Gly, Ser, and glycerate is obscure, as we did not observe root, and suggest that the stunting caused by the knockdown any detectable effects of the knockdown on photosynthetic prop- + results from some defect in NH4 assimilation in the shoot. erties measured from gas exchange. In photorespiration, glyoxylate is converted to Gly by two different enzymes, glutamate:glyoxylate Effects of Osppc4 Knockdown on Leaf Metabolism. We compared the aminotransferase (GGAT) and serine:glyoxylate aminotransferase metabolomes of the leaf blade at noon between nontransgenic rice (SGAT), the former of which uses Glu as the substrate. It is likely that and two knockdown lines. Among 101 compounds successfully depletion of Glu shifts the balance to the SGAT reaction. Such an annotated, the levels of 49 showed significant differences between imbalance in the photorespiratory reactions might allow the accu- nontransgenic rice and each of the two knockdown lines (P < 0.05; mulation of photorespiratory intermediates and related compounds. Table S1). Four notable changes in the levels of metabolites were observed (Fig. 4): (i) increased levels of secondary metabolites Chloroplastic PEPC Gene in Other Oryza Species. In database searches, fi related to the shikimate pathway, (ii) decreased levels of organic we identi ed the chloroplastic PEPC gene only in the Japonica- acids, (iii) opposite changes in levels of Gln and Glu involved in type rice cultivar Nipponbare, of which the entire genome has been NH + assimilation by the glutamine synthetase (GS)/glutamate completely sequenced (21). We therefore examined whether the 4 ′ synthase (GOGAT) cycle, and (iv) increased levels of photo- corresponding gene is present in other plant species by 5 -RACE respiratory intermediates. using degenerate primers and RNA from leaf blades. No cDNA – for PEPC with an N-terminal extension was detected in barley or The shikimate level was increased by 50 60% in the knockdown fi lines (Fig. S5), an indication of an increased level of PEP, the maize, but it was successfully ampli ed in all four Oryza species tested, the Indica-type rice cultivar Kasalath (O. sativa;AA precursor to shikimate as well as the substrate of PEPC. Among fi organic acids, the largest change was observed in the malate level, genome) and three wild rice species, O. ru pogon (AA), O. eichingeri (CC), and O. rhizomatis (CC), an indication that the which was decreased by one-half. This likely reflects a decreased chloroplastic PEPC gene is present and expressed in these species. level of the product of PEPC, OAA, which is readily converted to The deduced amino acid sequences of cDNAs from the same malate by NADP-malate dehydrogenase (MDH) in the chlor- genome-type species are almost completely the same (Fig. S6). oplast under illumination. Levels of citrate and isocitrate were also Even between different genome types, amino acid identities decreased, albeit to a lesser extent. By contrast, the levels of exceed 95%, making it likely that the chloroplastic PEPC gene was fumarate and succinate remained unchanged. Although the level present before the emergence of the genus Oryza. of 2-oxoglutarate (2-OG) did not show a significant difference, these results suggest that the metabolic flow from malate to 2-OG Discussion is largely suppressed by the knockdown of Osppc4. This study shows that rice has a plant-type PEPC targeted to the The level of Gln was notably increased, being almost doubled in chloroplast, Osppc4 (Figs. 1 and 2). Osppc4 was highly expressed in the line 4i-2, whereas that of Glu was decreased by about 20%. the leaves, accounting for about one-third of total PEPC protein in Because Gln and Glu are the substrate and the product, respectively, the leaf blade (Fig. S3). Because Osppc4 expression was confined to of GOGAT, the increased level of Gln suggests substantial sup- leaf mesophyll cells (Fig. 2C), the protein abundance in these cells pression of the GOGAT reaction. The reduced supply of 2-OG, would be higher than this estimate. We also identified the chloro- another substrate of GOGAT, could be the main cause for the plastic PEPC gene in other Oryza species belonging to two different suppression. Glu produced by the GS/GOGAT cycle is readily genome types, suggesting that it might be common among the genus converted to Asp in leaf cells. The level of Asp was reduced to about Oryza, but we did not find an ortholog of Osppc4 in barley or maize, two-thirds, suggesting suppression of the reaction of aspartate ami- and it is also absent from the Arabidopsis genome. Arabidopsis has 4 notransferase (AspAT), a key enzyme of amino acid synthesis, which PEPC genes, three plant types and one bacterial type (10). As uses Glu and OAA as substrates. Together, these findings lead us to judged from organ specificities of expression (10) (Fig. 1B), the rice + conclude that the Osppc4 knockdown suppresses NH4 assimilation plant–type genes Osppc1, Osppc2a/2b,andOsppc3 correspond to and the subsequent amino acid synthesis by depleting organic acids. Atppc1, Atppc2,andAtppc3, respectively, of Arabidopsis. Osppc2a

E4P Calvin Tryptamine Oxalate DHAP Glycolate Shikimate cycle 5-Hydroxy- Trp E4P Glycolysis tryptamine pathway Glycolysis Glyoxylate Tyr Shikimate PEP Glycerate Glycerol HCO - Photorespiration Caffeate Phe 3 HydroxyPyr Fig. 4. Changes in leaf metabolite levels after the Glycerol-3P PEPC Osppc4 knockdown. Nontransgenic rice (NT) and Quinate two homozygous knockdown lines (4i-2, 4i-10; T3 Leu Malate OAA Ser Gly Ile generation) were grown hydroponically with 1 Chlorogenate + Pyr Val mM NH4 . Midportions of the 10th leaf blade were Cys Met Thr harvested at noon and subjected to a metabolome Acetyl-CoA Malate analysis (Table S1). Metabolites annotated and OAA Asn Homoserine quantiﬁed are framed, and those whose levels Citrate Citrate Fumarate showed signiﬁcant differences (P < 0.05 by the Ala Asp Lys Student’s t test) in both knockdown lines are TCA cycle Aconitate Aconitate Pro Fold Succinate framed in color. In leaves, 2-oxoglutarate (2-OG) over NT Ornithine Isocitrate Isocitrate Glu synthesized from citrate in the cytosol is mainly > 1.3 + Succiny-CoA + used for NH4 assimilation by the GS/GOGAT cycle 1.1-1.3 GS/GOGAT NH4 2-OG 2-OG (20). DHAP, dihydoxyacetone phosphate; E4P, 0.9-1.1 cycle Putrescine erythrose 4-phosphate; Glycerol-3P, glycerol 3- 0.7-0.9 Gln < 0.7 phosphate; HydroxyPyr, hydroxypyruvate; Pyr, -Ala Uracil Arg pyruvate.

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.0913127107 Masumoto et al. Downloaded by guest on September 27, 2021 and 2b are highly homologous to each other with respect to their of suppression of malate synthesis in the leaves would be much exon–intron structure (Fig. 1A) and nucleotide sequence, and they larger if translocation of malate were taken into account. It is thus are likely to be paralogous to each other. evident that the Osppc4 metabolic pathway acts as a primary route + Osppc4 is also unique in its primary structure and does not belong for organic acid synthesis when rice plants are grown with NH4 . to any of the known groups of the plant-type PEPC (Fig. 1C). The It has been demonstrated using a variety of plant species that − catalytic and substrate-binding domains are almost the same in all when leaves are supplemented with NO3 after nitrogen starvation, plant-type Osppc proteins, whereas Osppc4 has a phosphorylation the cytosolic PEPC is activated through phosphorylation to meet the + domain distinct from the others (Fig. S1). Although it is unknown demand for carbon skeletons for NH4 assimilation (6). This is also whether Osppc4 can be recognized by PEPC kinases, regulatory the case in rice leaves (9). The cytosolic PEPC in leaf mesophyll cells phosphorylation of Osppc4 inside the chloroplast seems unlikely, might play roles in providing organic acids when rice plants are because no PEPC kinase gene encoding a protein targeted to the − − + grown with NO3 . Reduction of NO3 to NH4 requires seven chloroplast has been found in the rice genome (9). It is likely that the electron equivalents of reducing power per molecule, and it takes in vivo activity of Osppc4 is regulated by concentrations of metab- part in maintaining a balance of production and consumption of – olite effectors and the stromal pH. The stromal pH is 7.2 7.3 in ATP and NADPH during photosynthesis. It is possible that Osppc4, ∼ darkness and rises under illumination to 8, a pH at which PEPC together with NADP-MDH, plays an additional role in exporting becomes less sensitive to the feedback inhibitor malate (2). reducing power from the chloroplast, which would be beneficial for The enzyme characteristics of Osppc4 were similar to those of the + growth with NH4 . C4-type enzymes rather than to the C3 and root types, showing a low Our results clearly show that Osppc4 is crucial for the growth of affinity to PEP and a low sensitivity to malate (Table 1). These + rice plants with NH4 . Plants of the genus Oryza, which includes features seem favorable for reactions in the stroma. From the Km rice, are characterized by their ability to grow in waterlogged soil, values for PEP of C type isozymes, the concentration of PEP in the 3 which is a reductive environment, so that all of the inorganic leaf mesophyll cytosol of C plants is considered to be much below 3 nitrogen is converted to NH +. It is likely that Oryza species that of C plants, ∼3 mM in maize (22). The PEP concentration in 4 4 acquired the chloroplastic PEPC to adapt to such a habitat. Storage the stroma seems high enough for Osppc4 to react. Assuming that proteins and secondary metabolism show great diversity in vascular all leaf PEP is distributed homogenously in the cytosol and the fi fi stroma, the PEP concentration is calculated to be ∼1mMbothinthe plants, and many species-speci c genes have been identi ed (21). light and in darkness, using the PEP content of rice leaf blades [50– By contrast, it has long been believed that primary metabolism, − 1 with the exception of photosynthetic CO2 assimilation and nitro- 60 nmol (mg Chl) ] (23) and the cytosol and stromal volumes of PLANT BIOLOGY − fi barley [24.9 and 34.8 μL(mgChl) 1, respectively] (24). If all PEP is gen xation, is essentially the same. We propose that Oryza species have a unique route for organic acid synthesis and that primary localized to the stroma, the estimate increases to 1.8 mM. Con- + centrations of malate are 1 mM in the cytosol of spinach leaves, and NH4 assimilation is not necessarily the same in all vascular plants. levels in the stroma are 3 and 1 mM in the light and darkness, Materials and Methods respectively (25). If the elevation of stromal pH under illumination is taken into consideration, the presence of 3 mM malate in the light Plant Materials and Chloroplast Preparation. Rice plants (Oryza sativa L. cv. may not substantially prevent the Osppc4 reaction. Nipponbare) were grown in soil or in hydroponics under natural light conditions in a temperature-controlled greenhouse (SI Text). Seeds of wild rice species were We were able to elucidate the function of Osppc4 by comparing the generous gifts from D.A. Vaughan [Gene Bank, National Institute of Agro- Osppc4 knockdown lines with nontransgenic rice. The knockdown led + biological Sciences (NIAS), Japan]. Intact chloroplasts were prepared from leaf to stunting resulting from suppression of NH4 assimilation in the blades of 14-day-old rice seedlings by differential centrifugation (28). shoot (Fig. 3). Comparison of leaf metabolomes (Fig. 4) suggests that the Osppc4 knockdown greatly reduces levels of organic acids and Full-Length cDNA Clones, Gene Construction, and Transformation of Rice. Full- + thereby suppresses NH4 assimilation by the GS/GOGAT cycle and length cDNAs for Osppc2a and Osppc4 (accession nos. AK073703 and subsequent amino acid synthesis. It is generally accepted that organic AK066653, respectively) were kindly provided from the Rice Genome Resource acids are synthesized from photosynthates through glycolysis and the Center, NIAS. All primers used for gene construction are listed in Table S2.To TCA cycle, and that 2-OG synthesized from citrate in the cytosol is construct the Osppc4::GFP protein fusion, the sequence corresponding to the + N-terminal portion of the Osppc4 protein (amino acid residues 1–56) was used mainly for NH4 assimilation in leaf cells (20). These results suggest that Osppc4 constitutes another route for organic acid syn- amplified from the full-length cDNA by PCR and cloned into a GFP expression vector [CaMV35S-sGFP(S65T)-nos3′; a generous gift from Y. Niwa (University of thesis: It is likely that OAA produced by the action of Osppc4 is ′ fl readily converted to malate by NADP-MDH and is exported to the Shizuoka, Japan] (29). To construct promoter::GUS chimeric genes, 5 - anking regions of Osppc2a (from −2,240 to +1,945, numbered from the transcription cytosol by the malate-OAA shuttle under illumination (26). Once − fi “ ” start site) and Osppc4 (from 2,106 to +635) were ampli ed from the rice exported, malate takes a conventional route, being converted to 2- genomic DNA by PCR and cloned upstream of GUS in the binary vector pBI-Hm OG, imported to the chloroplast, and used by the GS/GOGAT cycle (a generous gift from M. Matsuoka, Nagoya University, Japan) (30). For in the stroma. A low sensitivity to Glu and Asp of Osppc4 (Table 1) knockdown of Osppc2a and Osppc4,3′ untranslated sequences (313 and 277 supports this hypothesis. We presume that PEP might be synthesized bp, respectively) were amplified from their respective full-length cDNAs by PCR from the Calvin cycle intermediate inside the chloroplast. 3-Phos- and cloned into the RNAi vector for rice, pANDA (a generous gift from phoglycerate, the product of Rubisco, is an intermediate of glycolysis K. Shimamoto, Nara Institute of Science and Technology, Japan) (31). Rice and can be converted to PEP via two enzymatic reactions catalyzed by transformation was performed via Agrobacterium-mediated gene transfer. 3-phosphoglycerate mutase and PEP enolase. This scenario is plau- sible because database searches indentified genes encoding these RT-PCR. RNA extraction and RT-PCR were performed as described (9, 32) using enzymes with transit peptides in the rice and Arabidopsis genomes. If the primers listed in Table S2. Total RNA had been treated with DNase before use. The DNA polymerase was first activated at 95 °C for 9 min, and PCR was so, photosynthates can be converted to organic acids without passing – through glycolysis. Osppc4 may also play a role in supplying carbon carried out for 25 30 cycles of 15 s at 95 °C, 30 s at 55 °C, and 1 min at 72 °C, + followed by a final extension step for 7 min at 72 °C. skeletons for NH4 assimilation/amino acid synthesis in other tissues/ organs such as leaf vascular bundles and roots, when malate is Subcellular and Cellular Localization of Osppc. The construct for the Osppc4:: translocated from leaf mesophyll cells. GFP protein fusion was introduced into dayflower (Commelia communis L.) In many cases, reductions of enzyme levels do not lead to sig- leaf segments by bombardment using 1.0-μm gold particles and a gene nificant metabolic perturbations, probably due to induction of delivery system (PDS-1000; Bio-Rad). After incubation in darkness at 25 °C for compensatory pathway(s) (27). In the Osppc4 knockdown lines, 16 h, the epidermis was peeled from the bombarded leaf segment, and the leaf malate level was reduced by one-half (Fig. S5). The extent fluorescence was observed with a fluorescence micrometer (AX70; Olympus)

Masumoto et al. PNAS Early Edition | 5of6 Downloaded by guest on September 27, 2021 through either a U-MNIVA or WIG filter (Olympus). Histochemical analysis of from halogen lamps at a photosynthetically active photon flux density of − − the GUS activity was performed using tissue sections 20–60 μm thick (33). 500 μmol·m 2·s 1 in a growth chamber. A main culm was cut 6–7 cm above the base, and the xylem sap exuded from the stump was blotted onto filter paper Recombinant Osppc Proteins. Recombinant Osppc2a and Osppc4 proteins His- in a plastic microtube for 30–40 min. Then 10 mM formic acid was added to the + + tagged at their N-termini were produced as follows. Sequences encoding the tube to prevent NH4 from volatilizing. The NH4 concentration was deter- Osppc2a protein and the mature portion of Osppc4 were amplified from their mined fluorometrically after reaction with o-phthalaldehyde (35) using an respective full-length cDNAs by PCR using the primers listed in Table S2 and HPLC detection system (474 Scanning Fluorescence Detector; Waters). inserted into pET32 and pET28 (Novagen), respectively. The resultant plasmids were transformed into Escherichia coli expression strain BL21 (DE3) cells (Nova- Metabolome Analysis. Hydroponically grown plants were used. A total of 24 gen). The transformed bacteria were grown in LB medium containing an seedlings, 8 each of nontransgenic rice and the two knockdown lines, were appropriate antibiotic at 30°C for 16 h, and expression of recombinant protein grown in a 30-L container with a culture solution containing 1 mM NH Cl. β 4 was induced by incubation with 1 mM isopropyl- -D-thiogalactopyranoside at 30 ° Segments of ∼3cm(∼30 mg fresh weight) were harvested from the mid- fi C for 3 h. Recombinant protein was puri ed from a cell lysate by chromatography section of the uppermost fully expanded leaf (10th leaf) at noon on a sunny fi on His-Bind resin (Clontech) and desalted by gel ltration (NAP-6, GE Healthcare). day and were immediately frozen in liquid nitrogen. Metabolite extraction, sample preparation, and gas chromatography–time-of-flight/mass spec- Protein and Enzymatic Analyses. Soluble protein was extracted from rice leaf trometry (GC-TOF/MS) were performed as described previously (36). Samples blades (32). Protein determination, SDS/PAGE, and immunoblotting were corresponding to 5 mg fresh weight were subjected to the analysis performed as described (34). An antiserum raised against the root nodule PEPC of soybean was a generous gift of C. P. Vance (University of Minnesota). The ACKNOWLEDGMENTS. We are grateful to Dr. Duncan A. Vaughan [National PEPC activity was measured by an enzyme-coupled spectrophotometric Institute of Agrobiological Sciences (NIAS)], Japan, and to Profs. Yasuo Niwa method at 30°C (23). The maximum activity was measured at pH 7.5 in the (University of Shizuoka, Japan); Makoto Matsuoka (Nagoya University, Japan); presence of 5 mM Glu6P. Kinetic parameters of recombinant PEPC proteins Ko Shimamoto (Nara Institute of Science and Technology, Japan); and Carroll P. were determined according to the method of Dong et al. (16) (SI Text). Vance (University of Minnesota) for generous gifts of wild rice seeds, vectors, and antibodies. We are also thankful to the following individuals: Drs. Sayaka Physiological Analyses. Gas exchange was measured at a leaf temperature of Moritaand ToshiakiOhkura(National Institute ofAgro-Environmental Sciences, 27°C with an open gas-exchange system (23). For growth analysis, plants were Japan) and Ei-ichi Minami (NIAS, Japan), for allowing us to use the NC analyzer grown hydroponically (SI Text). A total of 40 seedlings, 20 each of non- and the HPLC system; Mr. Makoto Kobayashi and Ms. Naomi Hayashi, RIKEN Plant Science Center, Japan, for their technical assistance; Drs. Atsushi Fukush- transgenic rice and the knockdown line, were transplanted into a 30-L con- ima and Masanori Arita for providing the custom program of metabolome tainer with a culture solution containing either NH Cl or KNO as the nitrogen 4 3 analysis; and Drs. Par Jonsson, Hans Stenlund, and Thomas Moritz (Umeå Plant fi source and cultivated for 25 days at the nal nitrogen concentration of 1 mM. Science Center, Sweden), for kindly providing the custom software for data To collect the shoot xylem sap, plants were grown hydroponically with 1 mM pretreatment. This work was supported by a Grant-in-Aid for Basic Research NH4Cl. At the 40th day, which corresponded to the 6th day after the last from the Ministry of Education, Culture, Sports, Science, and Technology of renewal of the culture solution, plants were transferred to a fresh solution Japan (18370022) and a grant from the Ministry of Agriculture, Forestry, and containing 2 mM NH4Cl, and were incubated for 6 h at 28°C under illumination Fisheries of Japan (Genomics for Agricultural Innovation, GPN0006; to M.M.).

1. Izui K, Matsumura H, Furumoto T, Kai Y (2004) Phosphoenolpyruvate carboxylase: A phosphoenolpyruvate carboxylases of Flaveria trinervia (C4) and Flaveria pringlei (C3). new era of structural biology. Annu Rev Plant Biol 55:69–84. Eur J Biochem 246:452–460. 2. Chollet R, Vidal J, O’Leary MH (1996) Phosphoenolpyruvate carboxylase: A ubiquitous, 18. Makino A, Sakuma H, Sudo E, Mae T (2003) Differences between maize and rice in N-use highly regulated enzyme in plants. Annu Rev Plant Physiol Plant Mol Biol 47:273–298. efﬁciency for photosynthesis and protein allocation. Plant Cell Physiol 44:952–956.

3. Vidal J, Chollet R (1997) Regulatory phosphorylation of C4 PEP carboxylase. Trends 19. Makino A, Harada M, Sato T, Nakano H, Mae T (1997) Growth and N allocation in rice – Plant Sci 2:230–237. plants under CO2 enrichment. Plant Physiol 115:199 203. 4. Gehrig H, Heute V, Kluge M (2001) New partial sequences of phosphoenolpyruvate 20. Hanning I, Heldt HW (1993) On the function of mitochondrial metabolism during carboxylase as molecular phylogenetic markers. Mol Phylogenet Evol 20:262–274. photosynthesis in spinach (Spinacia oleracea L.) leaves: Partitioning between 5. Nimmo HG (2003) Control of the phosphorylation of phosphoenolpyruvate respiration and export of redox equivalents and precursors for nitrate assimilation carboxylase in higher plants. Arch Biochem Biophys 414:189–196. products. Plant Physiol 103:1147–1154. 6. Champigny M-L (1995) Integration of photosynthetic carbon and nitrogen 21. International Rice Genome Sequencing Project (2005) The map-based sequence of the – metabolism in higher plants. Photosynth Res 46:117–127. rice genome. Nature 436:793 800. 7. Gregory AL, et al. (2009) In vivo regulatory phosphorylation of the phosphoenolpyruvate 22. Leegood RC (1985) The intercellular compartmentation of metabolites in leaves of – carboxylase AtPPC1 in phosphate-starved Arabidopsis thaliana. Biochem J 420:57–65. Zea mays L. Planta 164:163 171. 8. Chen Z-H, Nimmo GA, Jenkins GI, Nimmo HG (2007) BHLH32 modulates several biochemical 23. Fukayama H, et al. (2003) Activity regulation and physiological impacts of the maize fi and morphological processes that respond to Pi starvation in Arabidopsis. Biochem J 405: C4-speci c phosphoenolpyruvate carboxylase overproduced in transgenic rice plants. – 191–198. Photosynth Res 77:227 239. 9. Fukayama H, et al. (2006) Characterization and functional analysis of phosphoenolpyruvate 24. Winter H, Robinson DG, Heldt HW (1993) Subcellular volumes and metabolite concentrations in barley leaves. Planta 191:180–190. carboxylase kinase genes in rice. Plant J 47:258–268. 25. Heineke D, et al. (1991) Redox transfer across the inner chloroplast envelope 10. Sánchez R, Cejudo FJ (2003) Identification and expression analysis of a gene encoding membrane. Plant Physiol 95:1131–1137. a bacterial-type phosphoenolpyruvate carboxylase from Arabidopsis and rice. Plant 26. Scheibe R (2004) Malate valves to balance cellular energy supply. Physiol Plant 120:21–26. Physiol 132:949–957. 27. Hodges M (2002) Enzyme redundancy and the importance of 2-oxoglutarate in plant 11. Rivoal J, Trzos S, Gage DA, Plaxton WC, Turpin DH (2001) Two unrelated phos- ammonium assimilation. J Exp Bot 53:905–916. phoenolpyruvate carboxylase polypeptidesphysically interact in thehigh molecular mass 28. Nakatani HY, Barber J (1977) An improved method for isolating chloroplasts retaining isoforms of this enzyme in the unicellular green alga Selenastrum minutum. JBiolChem their outer membranes. Biochim Biophys Acta 461:500–512. 276:12588–12597. 29. Chiu W, et al. (1996) Engineered GFP as a vital reporter in plants. Curr Biol 6:325–330. 12. Gennidakis S, et al. (2007) Bacterial- and plant-type phosphoenolpyruvate carboxylase 30. Sato Y, et al. (1999) Loss-of-function mutations in the rice homeobox gene OSH15 polypeptides interact in the hetero-oligomeric Class-2 PEPC complex of developing affect the architecture of internodes resulting in dwarf plants. EMBO J 18:992–1002. – castor oil seeds. Plant J 52:839 849. 31. Miki D, Shimamoto K (2004) Simple RNAi vectors for stable and transient suppression ’ 13. O Leary B, Rao SK, Kim J, Plaxton WC (2009) Bacterial-type phosphoenolpyruvate of gene function in rice. Plant Cell Physiol 45:490–495. carboxylase (PEPC) functions as a catalytic and regulatory subunit of the novel Class-2 32. Fukayama H, et al. (2001) Significant accumulation of C4-specific pyruvate, PEPC complex of vascular plants. J Biol Chem 284:24797–24805. orthophosphate dikinase in a C3 plant, rice. Plant Physiol 127:1136–1146. 14. Kai Y, Matsumura H, Izui K (2003) Phosphoenolpyruvate carboxylase: Three- 33. Matsuoka M, Sanada Y (1991) Cis-acting elements in the pyruvate, orthophosphate – dimensional structure and molecular mechanisms. Arch Biochem Biophys 414:170 179. dikinase gene in maize. Mol Gen Genet 228:143–152. 15. Mamedov TG, Moellering ER, Chollet R (2005) Identification and expression analysis 34. Tsuchida H, et al. (2001) High level expression of C4-specific NADP-malic enzyme in leaves of two inorganic C- and N-responsive genes encoding novel and distinct molecular and impairment of photoautotrophic growth of a C3 plant, rice. Plant Cell Physiol 42:138–145. forms of eukaryotic phosphoenolpyruvate carboxylase in the green microalga 35. Husted S, Hobbern C, Mattsson M, Schjoerring J (2000) A critical experimental – + Chlamydomonas reinhardtii. Plant J 42:832 843. evaluation of methods for determination of NH4 in plant tissue, xylem sap and 16. Dong L-Y, Masuda T, Kawamura T, Hata S, Izui K (1998) Cloning, expression, and apoplastic fluid. Physiol Plant 109:167–179. characterization of a root-form phosphoenolpyruvate carboxylase from Zea mays: 36. Kusano M, et al. (2007) Applicationofa metabolomic method combining one-dimensional

Comparison with the C4-form enzyme. Plant Cell Physiol 39:865–873. and two-dimensional gas chromatography-time-of-ﬂight/mass spectrometry to metabolic 17. Svensson P, Bläsing O, Westhoff P (1997) Evolution of the enzymatic characteristics of phenotyping ofnaturalvariants inrice.JChromatogrB AnalytTechnolBiomedLife Sci 855:

C4 phosphoenolpyruvate carboxylase. A comparison of the orthologous PPCA 71–79.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.0913127107 Masumoto et al. Downloaded by guest on September 27, 2021