The Phosphorylated Pathway of Serine Biosynthesis IsEssential Both for Male Gametophyte and EmbryoDevelopment and for Root Growth in ArabidopsisW

Borja Cascales-Miñana,a,1 Jesús Muñoz-Bertomeu,a,1,2 María Flores-Tornero,a Armand Djoro Anoman,a

José Pertusa,b Manuel Alaiz,c Sonia Osorio,d Alisdair R. Fernie,d Juan Segura,a and Roc Rosa,3

a Departament de Biologia Vegetal, Facultat de Farmàcia, Universitat de València, 46100 Burjassot (Valencia), SpainbDepartament de Biologia Funcional i Antropologia Física, Facultat de Biologia, Universitat de València, 46100 Valencia, SpaincGrupo de Componentes Bioactivos y Funcionales de Productos Vegetales, Departamento de Fisiología y Tecnología de ProductosVegetales, Instituto de la Grasa, Consejo Superior de Investigaciones Científicas, 41012 Seville, SpaindMax Planck Institut für Molekulare Pflanzenphysiologie, 14476 Potsdam-Golm, Germany

ORCID IDs: 000-0002-2099-3754 (JM-B); 0000-0002-9296-0070 (MF-T); 0000-0003-0043-2180 (ADA); 0000-0001-7774-2676 (JS);0000-0003-4254-8368 (RR).

This study characterizes the phosphorylated pathway of Ser biosynthesis (PPSB) in Arabidopsis thaliana by targetingphosphoserine phosphatase (PSP1), the last enzyme of the pathway. Lack of PSP1 activity delayed embryo development,leading to aborted embryos that could be classified as early curled cotyledons. The embryo-lethal phenotype of psp1mutantscould be complemented with PSP1 cDNA under the control of Pro35S (Pro35S:PSP1). However, this construct, which waspoorly expressed in the anther tapetum, did not complement mutant fertility. Microspore development in psp1.1/psp1.1Pro35S:PSP1 arrested at the polarized stage. The tapetum from these lines displayed delayed and irregular development. Theexpression of PSP1 in the tapetum at critical stages of microspore development suggests that PSP1 activity in this cell layer isessential in pollen development. In addition to embryo death and male sterility, conditional psp1 mutants displayed a short-root phenotype, which was reverted in the presence of Ser. A metabolomic study demonstrated that the PPSB plays a crucialrole in plant metabolism by affecting glycolysis, the tricarboxylic acid cycle, and the biosynthesis of amino acids. We provideevidence of the crucial role of the PPSB in embryo, pollen, and root development and suggest that this pathway is animportant link connecting primary metabolism with development.

INTRODUCTION

The amino acid L-Ser is essential for the synthesis of proteinsand other biomolecules required for cell proliferation, includingnucleotides and Ser-derived lipids such as phosphatidyl Ser andsphingolipids (see Supplemental Figure 1 online). Besides itsparticipation in metabolism, additional nonmetabolic functionsfor Ser have been described in mammals and plants. In mam-mals, the de novo synthesis of L-Ser is essential in the de-velopment and function of the central nervous system (Yoshidaet al., 2004). Furthermore, L-Ser is the precursor of D-Ser, a well-documented neuromodulator (Mothet et al., 2000). Similarly, inplants, D-Ser has recently been attributed a signaling role in malegametophyte-pistil communication (Michard et al., 2011).

Ser biosynthesis in plants proceeds via different pathways (Figure1). One is the glycolate pathway, which takes place in mitochondria

and is associated with photorespiration (Tolbert, 1980, 1997; Douceet al., 2001; Bauwe et al., 2010; Maurino and Peterhansel, 2010).Additionally, alternative nonphotorespiratory mechanisms of Serbiosynthesis have been postulated (Kleczkowski and Givan, 1988).In quantitative terms, Ser production through the glycolate pathwayis considered to be the most important, at least in photosyntheticcells (Tolbert, 1980; Douce et al., 2001). In this pathway, two Glymolecules are converted into one molecule of Ser in a reactioncatalyzed by two enzymes, the Gly decarboxylase complex and theSer hydroxymethyltransferase (Figure 1). Since the glycolate path-way is associated with photorespiration, it should be active mainlyin green tissues during daylight hours. It therefore follows that al-ternative pathways of Ser biosynthesis may be required in the darkand/or in nonphotosynthetic organs. However, the biological sig-nificance of the coexistence of several Ser biosynthetic pathways inplants is still not understood.A nonphotorespiratory pathway, the so-called glycerate

pathway, synthesizes Ser by the dephosphorylation of 3-phos-phoglycerate (3-PGA) (Kleczkowski and Givan, 1988) (Figure 1).This pathway includes a reversed sequence of reactions froma portion of the oxidative photosynthetic carbon cycle linking3-PGA to Ser (3-PGA-glycerate-hydroxypyruvate-Ser), with thesereactions catalyzed by enzymes such as 3-PGA phosphatase,glycerate dehydrogenase, Ala-hydroxypyruvate aminotransferase,and Gly hydroxypyruvate aminotransferase. The existence of en-zymatic activities of this pathway in plants has been demonstrated

1 These authors contributed equally to this work.2 Current address: Instituto de Biología Molecular y Celular de Plantas,Departamento Biotecnología, Universidad Politécnica de Valencia-ConsejoSuperior de Investigaciones Cientificas, C/Ingeniero Fausto Elio, 46022Valencia, Spain.3 Address correspondence to roc.ros@uv.es.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Roc Ros (roc.ros@uv.es).W Online version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.113.112359

This article is a Plant Cell Advance Online Publication. The date of its first appearance online is the official date of publication. The article has been

edited and the authors have corrected proofs, but minor changes could be made before the final version is published. Posting this version online

reduces the time to publication by several weeks.

The Plant Cell Preview, www.aspb.org ã 2013 American Society of Plant Biologists. All rights reserved. 1 of 18

(Kleczkowski and Givan, 1988). However, the extent to which thispathway could be functional in plants is as yet unknown, andgenes coding for the specific enzymes of the pathway have notbeen cloned and/or characterized.

A second nonphotorespiratory pathway, the phosphorylatedpathway of Ser biosynthesis (PPSB), synthesizes Ser via phos-phoserine from 3-PGA as a precursor (Handford and Davies,1958). Supporting evidence for the PPSB in plants derives fromthe isolation and characterization of the enzyme activities of thisroute (Slaughter and Davies, 1968; Larsson and Albertsson,1979; Walton and Woolhouse, 1986). This pathway, which isconserved in mammals and plants, defines a branch point for3-PGA from glycolysis and involves three enzymes catalyzingsequential reactions: 3-phosphoglycerate dehydrogenase,3-phosphoserine aminotransferase, and 3-phosphoserine phos-phatase (PSP) (Figure 1). In humans, the PPSB plays a crucialrole in cell proliferation control. This pathway can divert a sub-stantial fraction of the glycolytic flux (carbon metabolism) intoSer biosynthesis (nitrogen metabolism) and can contribute tocell proliferation and oncogenesis (Bachelor et al., 2011; Loca-sale et al., 2011; Pollari et al., 2011; Possemato et al., 2011). Forinstance, an enhanced PPSB results in an increased cell pro-liferation rate, which is associated with certain breast cancers(Locasale et al., 2011; Possemato et al., 2011). By contrast,downregulation of the PPSB causes a restriction in tumor cellproliferation (Possemato et al., 2011).

Unlike in mammals, the functional significance of the PPSB inplants is not yet known. Although some genes of the pathwayhave been cloned and the enzymes that they encode have beenbiochemically characterized in Arabidopsis thaliana (Ho et al.,1998, 1999a, 1999b; Ho and Saito, 2001), no genetic evidencefor the physiological functions of these genes has been providedto date. Synthesis of amino acids takes place mainly in matureroots and source leaves, which export N through the phloem-xylem system to supply sinks, such as flowers and seeds(Lam et al., 1996). Specifically, Ser is easily transported throughthe phloem (Riens et al., 1991; Hunt et al., 2010). This wouldimply that the Ser synthesized in photosynthetic cells throughthe photorespiratory pathway could be supplied to non-photosynthetic organs. Similarly, Ser synthesized through non-photorespiratory mechanisms in roots could contribute to theamino acid supply to sinks such as seeds and flowers. However,the relevance of each Ser biosynthetic pathway in different or-gans and under different environmental conditions remains to bedefined.In previous studies, we characterized the plastidial glycolytic

glyceraldehyde 3-phosphate dehydrogenase (GAPCp) familyin Arabidopsis (Muñoz-Bertomeu et al., 2009, 2010a, 2010b,2011a, 2011b). The gapcp1 gapcp2 double mutants showa drastic developmental phenotype, including arrested primaryroot growth, dwarfism, and male sterility. We concluded thatGAPCps play an important role in plant development as their

Figure 1. Schematic Representation of Ser Biosynthesis in Plants and Connections with the TCA Cycle.

The enzymes participating in each Ser biosynthetic pathway are as follows. Photorespiratory pathway (glycolate pathway): GDC, Gly decarboxylase;SHMT, Ser hydroxymethyltransferase. Glycerate pathway: PGAP, 3-phosphoglycerate phosphatase; GDH, glycerate dehydrogenase; AH-AT, Ala-hydroxypyruvate aminotransferase. Phosphorylated pathway: PGDH, 3-phosphoglycerate dehydrogenase; PSAT, 3-phosphoserine aminotransferase.Abbreviations used for metabolites are as follows: THF, tetrahydrofolate; 5,10-CH2-THF, 5,10-methylene-tetrahydrofolate; 3-PHP, 3-phosphohy-droxypyruvate; 3-PS, 3-phosphoserine.

2 of 18 The Plant Cell

activity affects Ser supply to roots (Muñoz-Bertomeu et al.,2009; Muñoz-Bertomeu et al., 2010a). We specifically hypothe-sized that GAPCp activity is essential for supplying the pre-cursor 3-PGA for the PPSB in the plastids of nonphotosyntheticorgans. From these results, it could be inferred that the PPSBmay play an important role in specific plant developmentalevents.

This study addresses the functional significance of the PPSBin plant metabolism and development by targeting the last en-zyme of this pathway, PSP1. Here, we show that PSP1 is es-sential for male gametogenesis and embryogenesis and that it isindispensable for postembryonic root development. Although itis clear that metabolic changes and developmental reprogram-ming are tightly related, the connecting links are not known. Thisstudy highlights the relevant function of the PPSB in connectingprimary metabolism with development.

RESULTS

Expression Pattern and Subcellular Localization of PSP

In The Arabidopsis Information Resource database (http://www.arabidopsis.org), we found a single gene (At1g18640) coding forPSP (PSP1). This gene was previously cloned, and the encodedprotein possesses in vitro PSP activity (Ho et al., 1999a). However,an exhaustive study of its expression pattern, which could giveclues about its functions, was lacking.

Analysis of the PSP1 promoter using the promomer tool(Winter et al., 2007) revealed that the promoter regions weresignificantly enriched in consensus sequences present in im-portant genes regulating anther development, such as the floralhomeotic genes AGAMOUS and AGAMOUS2 (see SupplementalTable 1 online). An in silico analysis of PSP1 expression based onmicroarray databases (http://www.bar.utoronto.ca/efp/cgi-bin/efpWeb.cgi and https://www.genevestigator.com/gv/) revealeda constitutive pattern of expression through the main Arabi-dopsis development stages (see Supplemental Figure 2 on-line). At the organ level, PSP1 displayed a high expressionlevel in hypocotyls, early stages of root, flower (stages 9 to 12),seed and embryo (globular stage) development, and in shootapices (vegetative, inflorescence, and transition apex). Weassessed the expression patterns of PSP1 by RT-PCR. At bothseedling and adult stages, PSP1 was expressed in all organsstudied, and no significant differences in expression betweenorgans were found (Figure 2A). When PSP1 expression wasstudied under different growth conditions, an 8-h exposure todarkness induced the gene expression in the aerial parts, butnot in roots, whereas longer exposures (24 h) repressed PSP1expression on both roots and aerial parts (Figure 2B). Ser, theproduct of PSP1 activity, did not repress PSP1 expression inthe light, but repressed it after 8-h exposure to darkness inaerial parts.

Analysis of promoter-b-glucuronidase (GUS) fusions con-firmed that PSP1 is expressed in all organs studied, but it dis-played a highly specific cell tissue pattern (Figure 3A). At theseedling stage, PSP1 was expressed mainly in the vegetativeapex, veins, the distal zone of cotyledons, and in stomata. In

roots, PSP1 was expressed mainly in the transition zone be-tween roots and shoots, along the root vasculature, in theproximal part of the meristems and in the root cap columella. Atthe adult stage, PSP1 transcripts were detected in all organsstudied, but once again, a very specific cell tissue pattern wasobserved. In flowers, GUS activity was associated mainly withanthers, stigma, and pollen grains (Figure 3A). In leaves, PSP1was expressed mainly in the distal zone and veins. To furtherconfirm and extend the analysis of the PSP1 expression pattern,we stably expressed a PSP1-GFP (for green fluorescent protein)protein fusion construct under the control of the PSP1 promoter(ProPSP:PSP1) in wild-type plants. GFP fluorescence corrobo-rated the cell tissue–specific PSP1 expression pattern in theroots and reproductive organs (Figure 3B). In seedling roots,PSP1 was clearly expressed in the entire meristematic zone. Inflowers, PSP1 was expressed in pollen, anthers, and carpels. Inlater developmental stages, PSP1 was expressed in siliquevalves and seeds, especially in embryos.Using a 35S-PSP1-GFP construct, Ho et al. (1999a) demon-

strated that PSP1 was localized in plastids. Our results

Figure 2. Expression Analysis of PSP1.

(A) RT-PCR analysis of PSP1 in the aerial parts and roots of 18-d-oldseedlings grown in one-fifth-strength MS medium (left) and in differentadult (30-d-old) plant organs grown in greenhouse conditions (right).Values are normalized to the expression in the aerial parts (left) andshoots (right), respectively.(B) Effect of different growth parameters on the relative PSP1 expressionin roots and aerial parts of 18-d-old seedlings grown in one-fifth-strengthMS medium 6 0.1 mM Ser.Values are the mean 6 SE (n = 3 independent biological replicates).*Significantly different compared with controls in MS medium in the light(P < 0.05).

Phosphorylated Pathway of Ser Synthesis 3 of 18

Figure 3. Tissue and Subcellular Localization of PSP1.

4 of 18 The Plant Cell

employing the native PSP1 promoter for intracellular visualiza-tion of PSP1 corroborated that it is localized in both chloroplastsand nongreen plastids in leaves (Figure 3C). In roots, PSP1 alsodisplayed a plastidial localization. Similar results were obtainedwhen the PSP1-GFP construct was expressed under the controlof the 35S promoter (see Supplemental Figure 3 online).

The Homozygous psp1 Mutation Is Embryo Lethal

In order to shed light on the in vivo function of PSP1 in Arabi-dopsis, a reverse genetic approach was followed. Two in-dependent T-DNA insertion mutant lines affecting PSP1 wereidentified in the mutant collections (Salk-062391 and GK877F12),but only the Salk-062391 line could be reconfirmed. This linewas named psp1.1. The presence and genomic location of theT-DNA insertion was verified by PCR amplification of genomicDNA and by sequencing of the PCR products. The T-DNAinsertion was localized in the fourth intron, 1153 nucleotidesdownstream of the start codon of the PSP1 gene (seeSupplemental Figure 4 online). Genotyping plants from theoriginal seed stock identified only wild-type (PSP1/PSP1) andheterozygous (PSP1/psp1.1) individuals, with no visual phe-notype seen. An analysis of the segregation of the mutantpsp1.1 allele was conducted in a population of 264 seedsobtained from heterozygous PSP1/psp1.1 plants. No homo-zygous individuals (psp1.1/psp1.1) could be rescued based onPCR genotyping (Table 1). The segregation analysis of thepopulation displayed a 1:2 ratio (PSP1/PSP1:PSP1/psp1.1;x2 = 1.38; P > 0.05), which is typical of a Mendelian segregationwith a lethal phenotype for psp1.1/psp1.1 individuals. To cor-roborate these results, a second segregation analysis wasconducted based on the kanamycin resistance conferred bythe T-DNA insertion in the mutant allele. Of 1183 seedlingsproduced from heterozygous PSP1/psp1.1 plants, 66.1% werekanamycin resistant and 33.9% were kanamycin sensitive(Table 1). These results do not match the expected 75% re-sistant individuals for viable psp1.1/psp1.1 mutants. On thecontrary, the observed results match a 1:2 segregation (PSP1/PSP1:PSP1/psp1.1; x2 = 0.17; P > 0.05), indicating a lethalphenotype associated with the mutant psp1.1 allele. PSP1/psp1.1 plants were visually indistinguishable from the wildtype, indicating the recessive nature of the mutant allele.

Further evidence for the role of PSP1 was obtained by ana-lyzing point mutations in psp1 alleles, which were identified bythe Seattle Arabidopsis TILLING (for targeting-induced local le-sions in genomes) Project (http://tilling.fhcrc.org/). A new psp1allele was identified (psp1.2), which resulted in the substitutionof Ser-178 for Phe (S178F). According to SIFT (for sorting

intolerance from tolerant) software analysis, this mutation ispredicted to affect protein functions (SIFT score < 0.05). Specifi-cally, the Ser-178-Phe mutation, with the highest position-specificscoring matrix score of all the putative mutants identified, is sit-uated in the PSP1 conserved active site, at an essential positionfor catalytic activity, suggesting that it could be lethal. In a pop-ulation of 58 individuals coming from heterozygous PSP1/psp1.2mother plants, no homozygous psp1.2/psp1.2 could be identifiedby genotyping. Once again, the segregation analysis matcheda 1:2 ratio (PSP1/PSP1:PSP1/psp1.2; x2 = 0.42; P > 0.05)corroborating the essential role of PSP1 in early Arabidopsisdevelopmental stages.To investigate whether the lethality associated with psp1 mu-

tations is due to the male gametophyte, female gametophyte, orembryo defects, reciprocal outcrosses of PSP1/psp1.1 plants asmale/female parent (donor/recipient) and the wild type as female/male parent (recipient/donor) were performed, and the segrega-tion of the mutant allele was studied based on the antibiotic re-sistance conferred by the T-DNA insertion (Table 1). The resultsindicate that the male and female mutant gametophytes weretransmitted with an efficiency of 79 and 86.4%, respectively,which provides evidence that psp1.1 triggers an embryo-lethalphenotype.To further confirm this, siliques from heterozygous PSP1/psp1.1

plants were dissected at different developmental stages. From 8d after pollination (DAP), we observed a population of abnormalseeds, which was randomly distributed along the length of thesilique (Figure 4A). At 15 DAP, mutant seeds were white andstarted to deflate. At 22 DAP, mutant seeds turned dark brown andwere completely deflated. Segregation analysis of a population of5064 seeds obtained from heterozygous PSP1/psp1.1 plants re-vealed a 1:4 ratio (20.9% mutant:79.1% normal seeds; x2 = 2.52;P > 0.05). This ratio agrees with the expected 20.4%:79.6% ratiofor psp1.1/psp1.1:PSP1PSP1+PSP1psp1.1 calculated accordingto the transmission efficiency for the PSP1 and psp1.1 alleles(Table 1). To characterize the nature of the seed nonviability, weexamined the embryos in developing siliques of heterozygousPSP1/psp1.1 plants (Figure 4B). At 2 DAP, all the embryos ex-amined reached a similar developmental stage (octant stage ac-cording to Capron et al., 2009). However, at 5 DAP, some of theembryos showed delayed development (early globular versus tri-angular stage). According to microarray databases, PSP1 ex-pression is maximal at the globular stage (see Supplemental Figure2 online). The developmentally delayed embryos could be as-signed to mutant seeds at 7 DAP (heart versus mid torpedo) and10 DAP (heart versus early cotyledon stage). The terminal abortedembryos (15 DAP) were albino and could be classified as earlycurled cotyledons (L1) according to the SeedGenes database

Figure 3. (continued).

(A) Left: Expression of GUS under the control of the PSP1 promoter in apical meristem (I), cotyledons (II), guard cells (III), root and shoot vasculature (IV,V, and VII), and root meristem (VI) of 6- to 10-d-old plants. Right: Expression of GUS in anthers (I and II), carpels (I and IV), pollen (III), siliques (V), andleafs (VI) of adult plants. Bars = 0.5 mm in I, II, IV, V,VII, VIII, IX, XI, XII, and XIII and 50 µm in III, VI, and X.(B) PSP expression in roots and floral organs visualized by PSP1-GFP expression under the control of PSP1 promoter. Pg, pollen grain. Bars = 0.25 mm.(C) Chloroplastic/plastidic localization of PSP1 by stable expression of PSP1-GFP fusion protein under the control of PSP1 promoter in mesophyll,stomata, and root cells. Bars = 50 µm.

Phosphorylated Pathway of Ser Synthesis 5 of 18

(http://www.seedgenes.org). PCR-based genotyping analysis re-vealed that the aborted embryos were psp1.1/psp1.1 individuals.

In order to corroborate the phenotype-genotype correlation ofthe psp1.1 mutation, we transformed PSP1/psp1.1 plants withthe construct ProPSP:PSP1. We were able to demonstratecomplementation of the embryo-lethal phenotype of psp1.1/psp1.1 individuals by obtaining homozygous psp1.1/psp1.1plants in the segregating population (Figure 5).

PSP1 Expression in Anthers Is Required for MaturePollen Development

In the adult stage, transgenic psp1.1/psp1.1 ProPSP:PSP1 plantswere fertile and visually indistinguishable from the wild type. How-ever, when PSP1/psp1.1 plants were transformed with a PSP1-GFPcDNA under the control of the 35S promoter (Pro35S:PSP1), theresulting homozygous psp1.1/psp1.1 transgenic lines in the segre-gating population were sterile, producing small siliques with noseeds (Figures 5A and 5B). A PSP1 overdose effect was not thecause of the observed sterility phenotype since the wild typeexpressing Pro35S:PSP1 was fertile (Figures 5A and 5B).

As the 35S promoter exhibits very low expression, or none, inthe Arabidopsis tapetum (Grienenberger et al., 2009; Muñoz-Bertomeu et al., 2010b), we performed an ontogenic serial analysisof anther and pollen development from psp1.1/psp1.1 Pro35S:PSP1 plants compared with wild-type and psp1.1/psp1.1 ProPSP:PSP1 plants. For that purpose, floral buds were classified fromstages 7 to 12 according to the landmark events described bySmyth et al., (1990) and analyzed by transmission electron mi-croscopy (Figure 6A). In the psp1.1/psp1.1 Pro35S:PSP1 anthers,tetrads of microspores were formed (stage 7 according to Sanderset al., 1999) and progressed in their development till the polarizedmicrospore stage (stage 9). These initial steps of microsporedevelopment in psp1.1/psp1.1 Pro35S:PSP1 were similar tothose observed in the wild type. However, after the polarizedmicrospore stage, psp1.1/psp1.1 Pro35S:PSP1 microsporesinitiated a degeneration process characterized by detachment ofmicrospore protoplasm from the cell wall. At the end of thematuration process (stage 12), most of psp1.1/psp1.1 Pro35S:PSP1 pollen grains showed partially shrunken or completelycollapsed protoplasm. As in the wild type, pollen from psp1.1/psp1.1 ProPSP:PSP1 anthers developed and maturated normally

Table 1. Segregation of the psp1.1 Mutant Allele in the Progeny of Self-Crossed and Reciprocal Outcrossed Plants

Self-Crosses No. of Progeny

Genotypes of Progeny (PCR)

PSP1/PSP1 (%) PSP1/psp1.1 (%) psp1.1/psp1.1 (%)

PSP1/psp1.1 264 97 (36.7)a 167 (63.3)a 0 (0.0)a

PSP1/psp1.1 Pro35S:PSP1 192 54 (28.1)b 102 (53.1)b 36 (18.8)b

PSP1/psp1.1 ProPSP:PSP1 192 50 (26.0)c 99 (51.6)c 43 (22.4)c

psp1.1/psp1.1 ProPSP:PSP1 72 0 (0.0) 0 (0.0) 72 (100)

Self-Crosses No. of Progeny Antibiotic Resistance

Kans (%) Kanr (%)

PSP1/psp1.1 1183 401 (33.9)d 782 (66.1)d

PSP1/psp1.1 Pro35S:PSP1 1229 330 (23.9)e 899 (73.1)e

PSP1/psp1.1 ProPSP:PSP1 365 97 (26.6)f 268 (73.4)f

psp1.1/psp1.1 ProPSP:PSP1 479 0 (0.0) 479 (100)

Reciprocal Outcrosses No. of Progeny Antibiotic Resistance Transmission Efficiencyg

Kans (%) Kanr (%)Recipient 3 donor

PSP1/PSP1 3 PSP1/psp1.1 111 62 (55.9)h 49 (44.1)h 79.0PSP1/psp1.1 3 PSP1/PSP1 151 81 (53.6)i 70 (46.4)i 86.4

Kanr, kanamycin resistant; Kans, kanamycin sensitive.aSignificantly different from the expected 1:2:1 ratio for normal Mendelian segregation (x2 = 89.84; P < 0.001). Not significantly different from the 1:2ratio for embryo-lethal effect (x2 = 1.38; P > 0.05).bNot significantly different from the expected 1:2:1 ratio for normal Mendelian segregation (x2 = 4.13; P > 0.05).cNot significantly different from the expected 1:2:1 ratio for normal Mendelian segregation (x2 = 0.70; P > 0.05).dSignificantly different from the expected 1:3 ratio for normal Mendelian segregation considering antibiotic resistance phenotype (x2 = 49.94; P < 0.001).Not significantly different from the 1:2 ratio for embryo-lethal effect (x2 = 0.17; P > 0.05).eNot significantly different from the expected 1:3 ratio for normal Mendelian segregation considering antibiotic resistance phenotype (x2 = 2.25; P >0.05).fNot significantly different from the expected 1:3 ratio for normal Mendelian segregation considering antibiotic resistance phenotype (x2 = 0.48; P >0.05).gTransmission efficiency (%) = (mutant/wild type) 3 100.hNot significantly different from the expected 1:1 ratio for equal transmission efficiency (x2 = 1.52; P > 0.05).iNot significantly different from the expected 1:1 ratio for equal transmission efficiency (X2 = 0.80; P > 0.05).

6 of 18 The Plant Cell

(Figure 6C; see Supplemental Figure 5 online). As a consequenceof the pollen degeneration, anther locules from psp1.1/psp1.1Pro35S:PSP1 plants were almost empty at the late stages ofpollen development compared with those of the wild type orpsp1.1/psp1.1 ProPSP:PSP1 (see Supplemental Figure 6online). To confirm which step of gametophyte developmentwas affected by psp1.1 mutation, we performed Hoechststaining of microspore nuclei from psp1.1/psp1.1 Pro35S:PSP1 plants at different developmental stages (Figure 6B). For

this experiment, the most normally shaped microspores atthe last stages of development were selected since mostof them were seriously malformed. Nucleus staining cor-roborated that tetrad and polarized microspore developmentin psp1.1/psp1.1 Pro35S:PSP1 anthers was normal but thatsubsequent to the polarized microspore stage, develop-ment was arrested and no bicellular or tricellular pollen wasfound. These results indicate that microspores from psp1.1/psp1.1 Pro35S:PSP1 are unable to undergo the double mi-tosis necessary to reach the mature pollen stage. Scanningelectron microscopy revealed that mature pollen from thepsp1.1/psp1.1 Pro35S:PSP1 line was shrunken and col-lapsed and unable to germinate when cultured in vitro(Figure 6C).Analysis of anther development indicated that the tapetal cell

layer was formed in psp1.1/psp1.1 Pro35S:PSP anthers, buttapetum cells displayed irregular development, with partiallyshrunken protoplasm that was detached from the cell wall atcritical stages of pollen development (stages 9 and 10 accordingto Sanders et al., 1999) compared with wild-type cells (Figure6D). This may indicate a loss of turgor in the tapetum cells. It isimportant to note that pollen in psp1.1/psp1.1 Pro35S:PSP1anthers initiated the degeneration process at those developmentalstages (Figure 6A). A delay in the programmed cell death of thetapetum layer was also observed, since this layer was stillpresent at developmental stage 12 while it was absent in wild-type or psp1.1/psp1.1 ProPSP:PSP1 anthers (Figure 6D; seeSupplemental Figure 6 online).The observed morphological alteration in the tapetum of

psp1.1/psp1.1 Pro35S:PSP1 anthers correlated with the ex-pression of PSP1 in this cell layer, specifically at stage 9 (Figure6E; see Supplemental Figure 7 online). PSP1 was also clearlyexpressed in pollen, mainly in stage 12 and beyond (Figure 6E;see Supplemental Figure 7 online).We next conducted segregation analysis of self-fertilized PSP1/

psp1.1 plants that were homozygous for constructs Pro35S:PSP1and ProPSP:PSP1. Based on PCR genotyping, and unlike thesegregation of self-fertilized PSP1/psp1.1 plants, the segre-gation of PSP1/psp1.1 Pro35S:PSP1 displayed a typical Men-delian 1:2:1 ratio (PSP1/PSP1:PSP1/psp1.1:psp1.1/psp1.1; x2 =4.13, P > 0.05) (Table 1). The same was true for the segregationof self-fertilized PSP1/psp1.1 ProPSP:PSP1 (x2 = 0.7, P > 0.05;Table 1). Assuming a hypothesis of male sterility associated withthe psp1.1 allele, we should expect psp1.1/psp1.1 Pro35S:PSP1-GFP individuals to be sterile and psp1.1/psp1.1 ProPSP:PSP1 tobe fertile. In both cases, the observed results agreed with theexpected calculations according to the transmission efficiency forthe PSP1 and psp1.1 alleles (Figure 7). For instance, in the seg-regation analysis of PSP1/psp1.1 Pro35S:PSP1, we obtained18.8% sterile homozygous individuals, which agreed with theexpected 20.4%. In the segregation analysis of psp1.1/psp1.1ProPSP:PSP1, we obtained 26% fertile homozygous individuals,which almost perfectly matched the expected 25%. Based onkanamycin resistance, the segregation analysis of both self-fertilizedPSP1/psp1.1 Pro35S:PSP1 and PSP1/psp1.1 ProPSP:PSP1matched the 1:3 ratio (sensitive:resistant, x2 = 2.25 and 0.48,respectively, P > 0.05), unlike the 1:2 ratio observed in thePSP1/psp1.1 segregation analysis.

Figure 4. psp1 Homozygous Mutants Die at the Embryo Stage.

(A) Siliques from wild-type (WT) and heterozygous PSP1/psp1.1 plants at15 and 22 DAP observed with a binocular microscope. In the PSP1/psp1.1 silique, a population of mutant seeds is observed. The rightpicture shows a close-up of the mutant seeds. Bars = 1 mm.(B) The micrographs show the embryo development of the wild typeand homozygous psp1.1/psp1.1 from the same silique at differentstages observed with a differential interference contrast microscope(2, 5, 7, and 10 DAP) or with a binocular microscope (15 DAP). Bars =50 µm.

Phosphorylated Pathway of Ser Synthesis 7 of 18

Role of PSP1 in Vegetative Development

To study the function of PSP1 in the vegetative development,we obtained six independent conditional psp1.1/psp1.1 mu-tant lines from the segregating population of transformedPSP1/psp1.1 plants with a PSP1-GFP construct under the con-trol of the inducible heat shock promoter HS18.2 (ProHSP18.2:PSP1). Based on the induction pattern, two lines, whichshowed the lowest protein background expression level undernoninduced conditions, were selected for further analysis(Figure 8A).

In plates and under noninduced conditions, psp1.1/psp1.1ProHSP18.2:PSP1 (conditional psp1.1/psp1.1mutants) showed nosignificant difference in fresh weight (FW) of the aerial parts of theplant but the root FW was drastically reduced compared with thewild type (see Supplemental Figure 8 online). This reduction in rootFW was associated with a short-root phenotype (Figures 8B to 8D),which corresponds well with the specific expression pattern ofPSP1 in root meristems (Figures 3A and 3B). The root growth in-hibition of the conditional psp1.1/psp1.1 mutants was rescued byboth heat shock induction of PSP1 and Ser supply to the growthmedium (Figures 8B to 8D). Consequently, both treatments alsorecovered normal root FW values compared with controls (seeSupplemental Figure 8 online).

Further evidence for the role of PSP1 in root developmentis provided by the phenotypic analysis of transgenic lineswith reduced PSP1 expression. For this purpose, we usedartificial microRNA (amiRNA) lines. These lines phenocopiedthe short-root phenotype (Figure 9) and the Ser rescue ofroot growth of conditional psp1.1/psp1.1 mutants (data notshown).At the adult stage, the six conditional psp1.1/psp1.1 mutant

lines analyzed showed a sterile phenotype before induction andwere fertile after heat shock treatment (Figure 5), thus confirmingthe essentiality of PSP1 for embryo and/or anther development.However, unlike the short-root phenotype, it was not possible tocomplement the sterile phenotype of these conditional psp1.1/psp1.1 mutants by Ser supply to the growth medium (data notshown).In order to further test the function of PSP1 in embryo develo-

pment, ovules of noninduced conditional psp1.1/psp1.1 mutantswere cross-pollinated with pollen from heterozygous PSP1/psp1.1plants. As in heterozygous PSP1/psp1.1 plants, a population ofaborted seeds with abnormal embryo development could be ob-served. In this case, the segregation analysis of 435 seeds revealeda 1:1 ratio (48% mutant:52% normal seeds; x2 = 0.66; P > 0.05) asexpected.

Figure 5. Phenotypic Characterization of psp1.1 Mutants and Complemented Lines.

(A) Morphology of adult shoots from wild-type plants (WT), the wild type transformed with a PSP1-GFP cDNA under the control of the 35S promoter(PSP1/PSP1 Pro35S:PSP1), heterozygous PSP1/psp1.1 and homozygous psp1.1/psp1.1 lines complemented with a PSP1-GFP cDNA under thecontrol of the endogenous PSP1 promoter (ProPSP:PSP1), the 35S promoter (Pro35S:PSP1), or the heat shock–inducible promoter (ProHS18.2:PSP1).Homozygous psp1.1/psp1.1 lines transformed with ProHS18.2:PSP1 are shown before and after heat shock treatment. Bars = 1.5 cm.(B) Close-up of siliques from the same lines described in (A). Bars = 1.5 cm.

8 of 18 The Plant Cell

Figure 6. PSP1 Expression Is Essential for Proper Tapetum and Pollen Development.

(A) Developmental analysis by transmission electron microscopy of pollen (from stages 7 to 12 according to Sanders et al., 1999) in wild-type plants(WT) and homozygous psp1.1/psp1.1 plants transformed with PSP1-GFP cDNA under the control of the 35S promoter (Pro35S:PSP1). Bars = 20 mm.

Phosphorylated Pathway of Ser Synthesis 9 of 18

Metabolic Characterization of Lines Overexpressing andUnderexpressing PSP1

PSP activity was increased by 75% in the wild type expressingPro35S:PSP1 (Oex) and reduced by 57% in amiRNA linescompared with controls (58.5 6 2.2, 14.5 6 1.2, and 33.3 6 4.8µmol Pi mg protein21 h21 in Oex, amiRNA, and the wild type,respectively). The root growth rate in the Oex did not signifi-cantly differ from that of the wild type (Figure 9). However,metabolomic analysis revealed a clearly altered metabolitecontent in both amiRNA and Oex plants compared with the wildtype (Table 2; see Supplemental Table 2 online). In the aerial partof both amiRNAs and Oex plants, total amino acid increasedcompared with the wild type, but Oex plants showed thegreatest increase (27% versus 61%, respectively). In roots,metabolite changes differed from those found in the aerialparts and even occasionally, showed an inverse trend. In thisorgan, amiRNA lines displayed a greater increase in totalamino acid levels compared with the wild type (32% versus15% increase, respectively). Given that no perfect inversecorrelation of metabolite levels between Oex and amiRNA lineswas found, we focused on either those metabolites that sig-nificantly changed more than 1.5 times in only one of the lines(overexpressed or underexpressed) compared with the wildtype or those changes in which the Oex/amiRNA ratio washigher than 1.5 or lower than 0.5.

In aerial parts, significant increases in Pro and Orn were notedin Oex lines, while Trp increased in amiRNA lines compared withthe wild type (Table 2). A clear significant increase was alsofound in organic acids participating in either glycolysis or in thetricarboxylic acid (TCA) cycle (citric, glyceric, fumaric, and malic)in the amiRNA lines (but not the Oex lines) in comparison to thewild type. Regarding sugars and sugar alcohols, galactinol,myoinositol, maltose, and raffinose were significantly increased inthe Oex lines compared with the wild type. By contrast, 1-O-methylglucopyranoside significantly increased in amiRNA lines comparedwith the wild type.

In roots, a change in the Trp levels was observed, being sig-nificantly increased in amiRNA but decreased in Oex lines com-pared with the wild type (Table 2). A drastic reduction in theglyceric acid level was found in the Oex plants, while the level ofthis metabolite was not significantly altered in the amiRNA plants.

By contrast, other glycolytic intermediates, such as Fru-6-P andGlc-6-P, increased in the Oex lines. Finally, Ser levels increased inboth aerial parts and roots of both amiRNAs and Oex plants.Neither the total soluble protein and starch contents of roots

and aerial parts of the Oex lines nor the total soluble proteincontent of the amiRNA lines in roots and aerial parts were sig-nificantly altered in comparison to the wild type (Table 2).However, the starch content of amiRNA lines was increased by40% in the aerial parts in comparison to the wild type.

DISCUSSION

The existence of the PPSB in plants has been known since the1950s. However, the quantitative contribution and physiologicalsignificance of this pathway had not been defined in plants,mainly due to the coexistence of two other pathways with thesame metabolic function. By following a gain- and loss-of-function approach in Arabidopsis, we provide here both molecularand genetic evidence for the essential role of the PPSB in plants.Our data provide strong evidence concerning PSP1 expression

in different cell types and under differing environmental con-ditions, which affords clues as to its function. We have shown thatPSP1 is constitutively expressed in all Arabidopsis organs butdisplays a highly specific cell-type pattern in reproductive organsand roots, with expression in the tapetum, root cap columella,meristematic zone, and root elongation zone. This expressionpattern corresponds with the proposed function of PSP1 in em-bryo, root, and pollen development. It is interesting to note thatwe found a high PSP1 expression level in the aerial parts ofseedlings in comparison to that in roots. These results could in-dicate a role of the PPSB in green organs, even during the lightperiod. According to our expression data, we suggest that thePPSB may play a more ubiquitous role in different plant organsand under different environmental conditions than initially be-lieved. However, the essentiality of the pathway may be highlycell-type specific.Inhibiting or enhancing PSP1 expression drastically altered the

primary metabolism of Arabidopsis. Although it is difficult to givea complete picture of the metabolic changes associated withPSP1 activity, some conclusions can be drawn from our data.Silencing PPSB in the aerial parts resulted in elevated glyceric

Figure 6. (continued).

(B) Developmental analysis of pollen (from tetrad to tricellular stage) in wild-type and psp1.1/psp1.1 Pro35S:PSP1 plants visualized by Hoecht stainingof microspore nucleus. Bars = 25 mm.(C) Scanning electron micrographs of pollen grains (left) and light microscopy images of pollen germination assays (right) from the wild type, the wildtype transformed with Pro35S:PSP1 (PSP1/PSP1 Pro35S:PSP1), and the endogenous PSP1 promoter (PSP1/PSP1 ProPSP:PSP1); heterozygousPSP1/psp1.1 and homozygous psp1.1/psp1.1 plants transformed with Pro35S:PSP1 and the endogenous PSP1 promoter (ProPSP:PSP1). Bars = 20mm (left pictures) and 200 mm (right pictures).(D) Analysis by transmission electron microscopy of anther development (from stages 7 to 12 according to Sanders et al., 1999) in wild-type plants andhomozygous psp1.1/psp1.1 plants transformed with Pro35S:PSP1. Top images in each stage: cross sections of anthers. Bottom images in each stage:close-up of tapetum cells. Arrows point to tapetum cells. Bars = 20 mm.(E) PSP1 expression in the tapetum and pollen at different anther development stages (stages 9 and 12 in anther and pollen, respectively) visualized byconfocal microscopy of a PSP1-GFP fusion protein under the control of the endogenous PSP1 promoter expressed in the wild type. GFP fluorescenceand the merge images are presented. Bars = 75 mm.

10 of 18 The Plant Cell

acid, whose phosphate derivatives are important biochemical in-termediates in glycolysis, specifically 3-PGA, where glycolysis andthe PPSB diverge. The increase in glyceric acid in the PSP1-silenced lines was accompanied by an increment in all intermediatesof the TCA cycle that we measured. In the Oex lines these inter-mediates did not increase and some were even reduced in com-parison to the wild type. It thus seems reasonable to assume thatinhibition of the PPSB in the aerial parts is able to increase theglycolytic flux to the TCA cycle, which would in turn have im-portant consequences for plant growth and development.

In roots, silencing the PPSB did not produce the same effecton glycolytic metabolites and the TCA cycle as in the aerialparts, although Oex plants clearly showed a lowered glycericacid level, hinting at a depletion of 3-PGA in these lines asa result of enhancing the PPSB. The different behavior of theTCA cycle in roots and aerial parts could be related to the lack

of photosynthetic and photorespiratory activity in roots, whichprovide respiratory substrates and Ser to the aerial parts. Insupport of this hypothesis is the considerable body of evi-dence for regulatory links between photorespiration and theTCA cycle (Timm et al., 2012).Altering the PPSB also clearly changed the biosynthesis of

several amino acids. The amino acid Trp proved to be the com-mon link between aerial parts and root metabolism in both PSP1-underexpressing and -overexpressing lines. The Oex/amiRNA Trpratio was 0.2 in both aerial parts and roots. This amino acid de-rives from phosphoenolpyruvate, a metabolite downstream of3-PGA in the glycolytic pathway.As expected, we observed that the Ser levels of both aerial

parts and roots of Oex plants increased. It is conceivable that allof the other alterations observed in the metabolite profiling ofthese lines could be a consequence of this increase in Ser. We

Figure 7. Synopsis of the Genetic Evaluation of the psp1.1 Allele from Segregation Analysis of Heterozygous PSP1/psp1.1 Plants Transformed or Notwith a Construct Carrying PSP1 under the Control of Either the 35S Promoter (Pro35S:PSP1) or the Endogenous PSP1 Promoter (ProPSP:PSP1).

Allelic frequencies were estimated based on the transmission efficiencies for the psp1.1 allele calculated in Table 1. White boxes refers to wild-typeplants (PSP1/PSP1); gray boxes denote heterozygous PSP1/psp1.1 plants; striped boxes depict nonviable individuals; black boxes represent ho-mozygous psp1.1/psp1.1 plants complemented with either Pro35S:PSP1 (sterile) or ProPSP:PSP1 (fertile). Observed values were obtained by PCRgenotyping or antibiotic resistance conferred by the T-DNA insertion in the mutant allele. Expected values were calculated based on the allelicfrequencies obtained.

Phosphorylated Pathway of Ser Synthesis 11 of 18

also saw that the Ser level increased in the amiRNA plants,which was somewhat unanticipated. The elevated Ser level inthe aerial parts of amiRNA plants could be explained by anactivation of the other pathways for Ser biosynthesis, espe-cially the glycolate pathway, in order to compensate for thelack of PPSB activity in nonphotosynthetic organs. Ser pro-vided through the glycolate pathway could be transported toamiRNA roots through the phloem, thus accounting for thehigh Ser levels noted in these organs. However, a cause/effect

correlation between Ser deficiency and the root developmentalarrest observed in amiRNA lines cannot be rationalized unlessother parameters are taken into account. Ser measurementsreflect the mean amino acid content in the whole organ. Thus, itis possible that Ser deficiency is restricted only to specific celltypes, such as root meristematic and columella cells wherePSP1 is specifically expressed, and that such local Ser defi-ciencies provoke root growth inhibition. Thus, the high Ser levelin whole amiRNA roots could result from the inhibition of root

Figure 8. Ser Rescues the Arrested Root Development in Conditional psp1.1/psp1.1 Mutants.

(A) Immunoblot showing PSP1 expression in three representative conditional mutants (psp1.1/psp1.1 ProHSP18.2:PSP1) with (37°C 45 min) or without(control) heat shock induction compared with the wild type expressing PSP1 under the control of the endogenous PSP1 promoter (PSP1/PSP1 ProPSP:PSP1). Protein gel blot analysis was performed using anti-GFP antibodies. The arrow indicates PSP1-GFP. Ponceau staining of the blot is shown as theloading control in the bottom panel.(B) to (D) Seeds from wild-type (WT) and conditional psp1.1/psp1.1 mutants were germinated on one-fifth-strength MS medium for 8 to 10 d and werethen transplanted to a medium with (0.1 mM) or without (Control) Ser for 10 additional days. The same experiment was also done with heat shock–treated plants (37°C 45 min). Bar = 1 cm.(C) Final root length.(D) Growth rate before and after transplanting. Values are the mean 6 SE of two independent transgenic lines (n $ 36 plants). *Significantly differentcompared with the wild type (P < 0.05). Arrowhead indicates significant difference compared with plants grown in control medium without heat shock(P < 0.05).

12 of 18 The Plant Cell

growth, which in turn could mask Ser deficiency in specific celltypes. One likely consequence of root growth inhibition inamiRNA lines is that most of the amino acids determined werealso accumulated in this organ. In fact, the level of 19 out of 23measured amino acids increased in amiRNA roots comparedwith the wild type. Similarly, the starch level in the aerial partsincreased, which might also reflect the low demand of assim-ilates for growth in roots.

One important question that remains unanswered is why theglycerate or glycolate pathways cannot compensate for the lackof PPSB activity in roots and reproductive organs. It could bethat some cells (i.e., tapetum cells) are devoid of Ser trans-porters, that others (i.e., meristematic cells) are not convenientlyconnected to the vasculature, or a combination of both of thesehypotheses. Hence, the only supplies of this amino acid forthese cells would originate from intracellular biosynthesis. For allthese scenarios, our results suggest that other Ser biosyntheticpathways, such as the glycerate pathway or the activity ofthe Gly decarboxylase complex:Ser hydroxymethyltransferasecoupled reactions, are of little biological relevance in cells from

tapetum, embryo, or root meristems and that only the PPSB isresponsible for Ser supply to these cells.We found that PSP1 activity is essential for embryo de-

velopment. Homozygous psp1.1/psp1.1 mutants show a delayin embryo development starting at the globular stage, which fi-nally led to abortion. At early stages of embryogenesis, manynutrients are supplied to the embryo by the endosperm and/ormaternal tissues, while, at later stages, most metabolites aresynthesized autonomously by the embryo itself. At the globular-heart transition stage, many metabolic pathways are establishedin the embryo. Specifically, the first step of differentiation ofproplastids, where PSP1 is localized, occurs at the globular-to-heart transition phase (Devic, 2008). The delay in psp1.1/psp1.1mutant embryo development at the globular stage is probablyrelated to the establishment of the PPSB in the embryo.The role of Ser and/or Ser derivatives in plant reproduction

has been recently demonstrated (Michard et al., 2011; Yamaokaet al., 2011). Michard et al. (2011) showed that D-Ser is involvedin a plant signaling mechanism between male gametophyteand pistil, affecting pollen tube growth and morphogenesis.Yamaoka et al. (2011) showed that mutants of phosphatidylserinesynthase (PSS1), an enzyme involved in phosphatidylserinebiosynthesis, exhibit similar phenotypes as psp1.1/psp1.1 mu-tants, that is, diminished fertility due to inhibition of pollenmaturation and to a high embryo abortion rate. However, wediscovered differences between the mutants. For instance, re-ciprocal crossing revealed that the male mutant allele pss1 hasa decreased transmission rate. We found no significant differ-ence in the transmission efficiency of the psp1.1 male and fe-male alleles compared with the wild-type allele, indicating nogametophytic defect unless gametes are produced by homo-zygous anthers (psp1.1/psp1.1 Pro35S:PSP1-GFP). The segrega-tion analysis of heterozygous PSP1/psp1.1 and cross-pollinatedovules of noninduced conditional psp1.1/psp1.1 mutants withpollen from heterozygous PSP1/psp1.1 plants corroborated thatmale psp1.1 alleles are viable if produced by heterozygous antherswhere PSP is expressed. According to our data, the only possibilitythat defects in the male psp1.1 gametophyte could be specificallyassociated with PSP activity in pollen is if premeiotic gene ex-pression in diploid heterozygous microsporocytes enables allmutant gametophytes to survive the loss of PSP1 in homozygousgametes, as has already been suggested for other essential genes(Muralla et al., 2011). In our case, this possibility is improbable,since we were unable to detect PSP1 expression in premeioticsporogenous cells.Several of the genes regulating tapetum development have

been found to affect pollen viability and/or development(Colcombet et al., 2005; Yang et al., 2007). Accordingly, our resultsindicate a temporal correlation between PSP expression in thetapetum (stages 9 and 10 according to Sanders et al., 1999) andthe developmental alterations in psp1.1/psp1.1 Pro35S:PSP1-GFP tapetum and microspores. These stages are known to becritical for pollen maturation (Ge et al., 2010). PSP1 was alsoexpressed in pollen but in later stages (stages 11 to 14) wherepsp1.1/psp1.1 Pro35S:PSP1-GFP mutant microspores alreadydisplayed altered development.Cells from the tapetum and microspores develop as “com-

municating” neighbors. Early in development, tapetum cells

Figure 9. Root Phenotype of PSP1-Silenced and OverexpressingPlants.

Root length (A) and root growth rate (B) of two independent PSP1-silenced (amiRNA) and overexpressing (Oex) plants compared with wild-type plants (WT). Seeds were germinated on one-fifth-strength MSmedium and grown for 18 d. Values are the mean 6 SE (n $ 24 plants).*Significantly different compared with the wild type (P < 0.05).

Phosphorylated Pathway of Ser Synthesis 13 of 18

are involved in the secretion of molecules into the locule formicrospore maturation. We propose that the absence of PSP1activity affects the proper development of both cell types,which in turn affects pollen maturation. The developmentalmodifications observed in the psp1.1/psp1.1 Pro35S:PSP1-GFP tapetum along with the specific PSP1 expression in thiscell layer during the critical stages of pollen maturation suggestthat an essential step in microspore development is PSP1expression during the post-tetrad stages in the tapetum andnot in the microspore itself. Proper timing of programmed celldeath in the tapetum is required for normal microsporogenesis(Kawanabe et al., 2006). Since we found delayed degenerationof the tapetum cell layer in the psp1.1/psp1.1 Pro35S:PSP1-GFP anthers, we contend that it is possible that incorrect co-ordination of the timing between the degeneration of tapetumand microspore development could ultimately have affectedpollen maturation.

During embryo and anther development, cells actively divide.The root apical meristem, where PSP1 is clearly expressed, alsoundergoes intense mitotic activity. Our results indicate that thePPSB plays an essential role in these nonphotosynthetic, ac-tively dividing cells. In the last years, considerable informationlinking the PPSB with the regulation of cell proliferation andoncogenesis in mammals has appeared (Bachelor et al., 2011;Locasale et al., 2011; Pollari et al., 2011; Possemato et al.,2011). Similarly, the function of this pathway in plant embryo,

anther, and root development could be manifested via the reg-ulation of cell proliferation. In both mammals and plants, Ser iscrucial for methyl group transfer by providing one carbon unitsto tretrahydrofolate metabolism, which is, in turn, essential toDNA and RNA metabolism (Hanson and Roje, 2001; Hansonand Gregory, 2011; Kalhan and Hanson, 2012). In mitogenicallystimulated lymphocytes, the demands for increased nucleotidebiosynthesis for DNA replication during cellular proliferationare matched by a corresponding increase in Ser utilization(Eichler et al., 1981). Thus, Ser could be the link connectingprimary metabolism with the regulation of cell cycle progression.Given the evidence for the one-carbon metabolism takingplace in Arabidopsis plastids (Zhang et al., 2010), the arrestedroot and embryo development in the psp1.1/psp1.1 mutantwould suggest that the PPSB is required to supply Ser for theproper functioning of one-carbon metabolism in some non-photosynthetic cells. If this idea proves correct, it opens upthe possibility of evolutionarily conserved signaling mechanismsassociated with primary metabolism in plants and animals.

Conclusions

PSP1 is expressed in photosynthetic and nonphotosyntheticorgans but plays an essential function only in very specific celltypes, probably because the other Ser pathways in these celltypes cannot compensate for the lack of PPSB activity. We

Table 2. Metabolite Levels in the Aerial Parts and Roots of Wild-Type, PSP1-Silenced (amiRNA), and PSP1-Overexpressing Plants (Oex)

Aerial Part Roots

Metabolites amiRNAs Oex The Wild TypeOex/amiRNAs amiRNAs Oex

The WildType

Oex/amiRNAs

Amino AcidsOrn 1.08 6 0.11 1.79 6 0.12 1 6 0.04 1.7 1.13 6 0.07 0.75 6 0.03 1 6 0.06 0.7Pro 1.01 6 0.05 3.08 6 0.39 1 6 0.07 3.1 1.70 6 0.15 1.20 6 0.05 1 6 0.05 0.7Ser 1.57 6 0.19 1.86 6 0.08 1 6 0.05 1.2 1.53 6 0.09 1.13 6 0.02 1 6 0.03 0.7Trp 3.58 6 1.08 0.89 6 0.07 1 6 0.12 0.2 1.73 6 0.17 0.43 6 0.05 1 6 0.11 0.2

Organic AcidsCitric acid 1.72 6 0.21 0.94 6 0.03 1 6 0.05 0.5 0.98 6 0.05 1.00 6 0.04 1 6 0.02 1.0Fumaric acid 2.03 6 0.37 0.90 6 0.03 1 6 0.03 0.4 0.67 6 0.03 0.84 6 0.04 1 6 0.05 1.2Glyceric acid 1.46 6 0.14 1.05 6 0.06 1 6 0.09 0.7 0.86 6 0.10 0.38 6 0.01 1 6 0.07 0.4Malic acid 1.74 6 0.27 0.89 6 0.03 1 6 0.03 0.5 0.79 6 0.01 0.82 6 0.02 1 6 0.03 1.0

Sugars and Sugar AlcoholFru-6-P 0.97 6 0.08 1.09 6 0.07 1 6 0.11 1.1 0.99 6 0.05 1.44 6 0.06 1 6 0.03 1.5Galactinol 1.31 6 0.20 2.69 6 0.24 1 6 0.12 2.1 2.13 6 0.25 2.26 6 0.06 1 6 0.1 1.1Glc-6-P 1.21 6 0.14 1.17 6 0.14 1 6 0.11 1.0 1.04 6 0.06 1.66 6 0.05 1 6 0.03 1.6Glucopyranoside, 1-O-methyl 2.18 6 0.30 1.38 6 0.19 1 6 0.17 0.6 1.07 6 0.13 1.34 6 0.13 1 6 0.08 1.2Inositol, myo 1.18 6 0.11 2.11 6 0.12 1 6 0.05 1.8 1.60 6 0.22 1.58 6 0.03 1 6 0.06 1.0Maltose 1.05 6 0.09 1.99 6 0.41 1 6 0.12 1.9 1.36 6 0.06 1.19 6 0.03 1 6 0.04 0.9Raffinose 1.10 6 0.14 3.39 6 0.34 1 6 0.09 3.1 1.30 6 0.10 1.67 6 0.03 1 6 0.08 1.3

TotalsAmino acids 1590.6 6 60.5 2015.4 6 100.0 1249.4 6 41.2 1.3 1599.9 6 37.5 1384.8 6 25.7 1204.2 6 51.6 0.9Organic acids 1249.4 6 51.7 1355.3 6 51.7 1020.5 6 24.9 1.1 971.3 6 19.4 950.1 6 14.8 734.6 6 47.4 1.0Sugars and sugar alcohol 144.1 6 11.6 157.4 6 7.4 120.9 6 5.1 1.1 581.8 6 23.4 594.8 6 9.8 515.7 6 9.6 1.0Proteins (n = 6) 8.93 6 0.61 9.71 6 0.76 8.66 6 0.69 1.1 3.52 6 0.38 4.09 6 0.24 4.75 6 0.37 1.2Starch (n = 3) 1.30 6 0.04 0.78 6 0.09 0.912 6 0.00 0.6 0.08 6 0.01 0.081 6 0.004 0.11 6 0.02 1.0

Data are relative values normalized to the mean response calculated for the wild type. Values represent the mean 6 SE of 12 independent determinations.For Oex and amiRNA, data correspond to samples of two independent transgenic plants. The Oex-to-amiRNA ratio (Oex/amiRNA) is also presented. Totalsrepresent the sum of all metabolites in each group in arbitrary units. Protein and starch contents are given in absolute values (mg g21 FW). Those values thatare significantly different to the wild type are shown in bold; P < 0.05. The full data set is shown in Supplemental Table 2 online.

14 of 18 The Plant Cell

provided molecular, morphologic, physiological, and geneticevidence for the essential role of the PPSB in male gametophyteand embryo development. We have also proven that this path-way is indispensable for proper root growth. It is well known thatmetabolism is intimately related to development, but the con-necting links between these processes remain unknown. Al-though it is difficult to separate the metabolic and regulatoryfunctions of a metabolic pathway, the PPSB might be one of themain actors connecting primary metabolism with development.

Finally, we provided tools that will allow the future study of thispathway under different environmental conditions and its inter-actions with other Ser biosynthetic pathways.

METHODS

Plant Material and Growth Conditions

Arabidopsis thaliana seeds (ecotype Columbia 0, Col-er 105) were suppliedby the European Arabidopsis Stock Center (Scholl et al., 2000). Unlessstated otherwise, seeds were sown on 0.8% agar plates containing one-fifth-strength Murashige and Skoog (MS) medium with Gamborg vitaminsas previously described (Muñoz-Bertomeu et al., 2009). Some plates weresupplemented with 0.1 mM Ser as indicated in the legends to the figures.The optimal concentration of Ser was previously determined (Muñoz-Bertomeu et al., 2009). For the selection of PSP1 transgenic plants, half-strengthMS plates supplementedwith 0.5%Suc and appropriate selectionmarkers were used. Some plantlets and seeds were also grown ingreenhouse conditions as described elsewhere (Muñoz-Bertomeu et al.,2009).Where indicated, conditional psp1.1/psp1.1mutants were treated for45 min at 37°C daily before and after pollination.

Primers

All primers used in this work are listed in Supplemental Table 3 online.

Mutant Isolation and Characterization

The psp1.1 (SALK_062391) allele of the At1g18640 gene was identified inthe SIGnAL Collection database at the Salk Institute (Alonso et al., 2003).Mutants were identified by PCR genotyping using gene-specific primersand left border primers of the T-DNA insertion. PSP_RP and PSP-LP werethe gene-specific primers for psp1.1 allele. The T-DNA insertion wasconfirmed by sequencing the fragment amplified by the T-DNA internalborders (LBb1.3) and gene-specific primers.

Tilling mutants were provided through the Seattle TILLING Project(http://tilling.fhcrc.org/), which performed a high-throughput reverse ge-netic screening to identify ethyl methanesulfonate–induced mutations inthe Col-er 105 background (Till et al., 2003). As a result, a psp1 allele(psp1.2) with a S178F amino acid substitution was identified. Tillingmutants were identified by PCR amplification of the target sequences withprimers S178Fr2 and S178FRev and by restriction analysis of the PCRproduct with the enzyme EcoRV, which differentiates between the wild-type and the mutant alleles.

Cloning and Plant Transformation

For gene promoter-reporter fusions, a 1.4-kb fragment corresponding tothe native PSP1 promoter was fused to the GUS gene in pCAMBIA1303,giving ProPSP-GUS. The cDNA corresponding to PSP1 (At1g18640) wasplaced under the control of three different promoters; the constitutive 35Spromoter (Pro35S), the native PSP1 promoter (ProPSP), and the heatshock–inducible promoter of gene At5g59720 (ProHSP18.2) (Matsuhara

et al., 2000) giving constructs Pro35S:PSP1, ProPSP:PSP1, andProHSP18.2:PSP1, respectively. The PSP1 cDNA was cloned in thepCR8/GW/TOPO plasmid and subcloned in the plasmid pMDC83 (Curtisand Grossniklaus, 2003). This plasmid allowed us to clone PSP1 in framewith a GFP cDNA at the C-terminal position (PSP-GFP), giving thePro35S:PSP1 vector. This vector was used to obtain constructs ProPSP:PSP1 and ProHSP18.2:PSP1. The 35S promoter of Pro35S:PSP1 wasexchanged with the native PSP1 promoter of ProPSP-GUS, giving theProPSP:PSP1 vector. The 35S promoter of Pro35S:PSP1 was alsoexchanged for the promoter region of gene HSP18.2, giving theProHSP18.2:PSP1 vector. The HSP18.2 promoter was obtained fromPTT101 plasmid (Matsuhara et al., 2000).

AmiRNAs were generated to target PSP1 using the Web microRNAdesigner (http://wmd2.weigelworld.org/cgi-bin/mirnatools.pl). The amiRNAswere cloned according to the protocol by Rebecca Schwab in DetlefWeigel’s laboratory (http://wmd2.weigelworld.org/themes/amiRNA/pics/Cloning_of_artificial_microRNAs.pdf). All PCR-derived constructs wereverified by DNA sequencing. Additional cloning information is provided inSupplemental Table 4 online.

Arabidopsis plants were transformed with the different constructsusing the floral dipping method (Clough and Bent, 1998) with Agro-bacterium tumefaciens carrying pSOUP. For the amiRNA and GUS lines,the wild type was used. As psp1.1/psp1.1 is embryo lethal, whenevernecessary, the progeny of heterozygous plants (PSP1/psp1.1) wastransformed with the different constructs. Transformants were selectedby antibiotic selection, while homozygous psp1.1/psp1.1, heterozygousPSP1.1/psp1.1, and the wild type were identified by PCR genotypingusing gene-specific primers (PSP_RP and PSP-LP) and left border pri-mers of the T-DNA insertions (LBb1.3). At least four independent singleinsertion homozygous T3 lines were obtained for all different constructs.After characterization by RT-PCR, two to five different lines were selectedfor further analysis depending on the experiment. We used both syngenicwild-type lines as well as wild-type Columbia-0 as controls for our studies.Syngenic plants were the wild-type progeny from the segregation ofheterozygous transgenic PSP1/psp1.1 plants. For amiRNAs, we used ascontrols the wild type used for transformation with the amiRNAs.

RT-PCR

Total RNA was extracted from seedlings and adult plants using theNucleoSpin RNA II kit (Macherey-Nagel). RNA (0.5 to 1 µg) was reversetranscribed using polyT primers and the first-strand cDNA synthesis kit forquantitative RT-PCR (Fermentas) according to the manufacturer’s in-structions. Real-time quantitative PCR was performed using a 5700 se-quence detector system (Applied Biosystems) with the Power SYBR GreenPCR Master Mix (Applied Biosystems) according to the manufacturer’sprotocol. Each reaction was performed in triplicate with 1 mL first-strandcDNA in a total volume of 25 µL. Data are the mean of three biologicalsamples. PCR amplification specificity was confirmed with a heat disso-ciation curve (from 60 to 95°C). Efficiency of PCR was calculated anddifferent internal standards were selected (Czechowski et al., 2005) de-pending on the efficiency of the primers. Relative mRNA abundance wascalculated using the comparative cycle thresholdmethod according to Pfaffl(2001). Primers used for PCRs are listed in Supplemental Table 3 online.

Pollen Germination and GUS Activity Assays

In vitro pollen germination assays were done using the optimized solidmedium described by Boavida and McCormick (2007) as previouslydescribed (Muñoz-Bertomeu et al., 2010b).

For GUS activity assays, plant organs were incubated in GUS buffer (100mM sodium phosphate, pH 7.0, 10 mM EDTA, 0.1% Triton X-100, 0.5 mMpotassium ferricyanide, 0.5mMpotassium ferrocyanide, and 2mM5-bromo-4-chloro-3-indonyl-b-D-glucuronic acid [X-GlcA; Duchefa]) overnight at

Phosphorylated Pathway of Ser Synthesis 15 of 18

37°C. Plant material was cleared in 70% ethanol before microscopyobservation. At least six independent transgenic lines showed identicalGUS-staining patterns, which differed only in the expression level of GUS.Images were acquired with a Leica DM1000 microscope and a LeicaDC350 digital camera.

Microscopy

For transmission electron microscopy of anther and pollen, inflo-rescences containing buds at different developmental stages from wild-type, psp1.1/psp1.1 Pro35S:PSP1-GFP, and psp1.1/psp1.1 ProPSP:PSP1-GFP plants were collected and classified into different de-velopment groups according to Smyth et al. (1990). Buds from the samegroup were removed from plants, fixed with 2.5% glutaraldehyde, andpostfixed in 1% osmium. Tissue was dehydrated in an ethanol series andinfiltrated with LR-White resin. Ultrathin sections (60 nm) were examinedwith a JEOL_1010 transmission electron microscope at 60 kV equippedwith a MegaView III digital camera with Analysis software.

For differential interference contrast microscopy, siliques from het-erozygous PSP1/psp1.1 plants were dissected longitudinally at 2, 5, 7,and 10 DAP. Ovules from individual siliques were morphologically clas-sified, cleared in Hoyer’s solution, and observed with a Nikon ECLIPSEE800 compound microscope equipped with Nomarski differential in-terference contrast optics and a Nikon DXM1200F digital camera withNikon ACT-1 software. Bright-field photographs of 15-d-old hand-dissected embryos submerged inwater were obtained using a stereoscopiczoom microscope Nikon SMZ 1500 equipped with a DS-Fi1 digital sightcamera.

For scanning electron microscopy, pollen was mounted on standardstubs and coated with gold-palladium prior to observation under a fieldemission microscope (Hitachi S-4100). GFP fluorescence was observedwith a confocal microscope (Leica TCS-SP).

For nucleus staining with Hoechst, inflorescences containing budsat different developmental stages from wild-type and psp1.1/psp1.1Pro35S:PSP1-GFP plants were collected and classified in different groupsas described for transmission electron microscopy. Buds from the samegroup were dissected on a microscope slide, and pollen grains releasedfrom anthers were stained with 10 µg/mL Hoechst dye. Samples wereobservedwithNikon Eclipse E800 compoundmicroscope equippedwith anepifluorescence module and a Nikon DXM1200F digital camera with NikonACT-1 software.

Metabolite Determination and PSP Activity

Aerial parts and roots of 18- to 21-d-old wild-type, Oex, and amiRNA lineswere used to determine metabolite content. Starch was determined withthe Enzytec starch kit (ATOM). The protein content was quantified usingthe Bio-Rad protein assay kit. The levels of other metabolites were de-termined in derivatized methanol extracts by gas chromatography–massspectrometry using the protocol defined by Lisec et al. (2006).

PSP activity was measured in at least six biological replicates of wild-type, Oex, and amiRNA root extracts according to Ho et al. (1999a).

Immunoblotting

For immunoblots, crude protein extracts from whole wild type andconditional psp1.1/psp1.1 mutants were obtained by harvesting 0.5 gplant material in liquid nitrogen and grinding in 1 mL ice-cold homoge-nization buffer (140 mM NaCl, 8 mM Na2HPO4$7H2O, 2 mM KH2PO4, and10 mM KCl, pH 7.4) with 1% protease inhibitor cocktail (Sigma-Aldrich;P-9599), followed by two successive centrifugations (15,000g for 15 minat 4°C). Protein samples were quantified with Bradford reagent (Bio-Rad)using BSA as a standard, and then 100 µg total proteins was separatedin 8% SDS-PAGE. Proteins were electrotransferred onto immunoblot

nitrocellulose membranes (Bio-Rad) using the Mini Tran-Blot Cell (Bio-Rad) for 1 h at 100 V with transfer buffer (17 mM Tris, 192 mM Gly, and20% [v/v] methanol). Immunoblots were probed with anti-GFP antibodies(Molecular Probes; ref. A-11122) at a final dilution of 1:20,000. Ponceau S–stained membranes are shown as loading controls. Cross-reacting bandswere detected using the ECL selectWestern-blotting detection reagent kit(Amersham Biosciences).

Statistics

Experimental values represent mean values and SE; n represents thenumber of independent samples. P values were calculated with a Stu-dent’s t test (two-tailed) using Microsoft Excel. The level of significancewas fixed at 5% (0.05), representing the probability of error if the hy-pothesis of a significant difference betweenmean values was accepted. Inthe segregation analysis, data were tested to evaluate if the observedvalues agreed with Mendelian proportions using the x2 test. P valueshigher than 0.05 indicate that the observed values are not significantlydifferent from the expected ratio. P values lower than 0.001 indicate thatthe observed values differ are significantly from the expected ratio.

Accession Number

Sequence data from this article can be found in the Arabidopsis GenomeInitiative or GenBank/EMBL databases under accession numberAt1g18640 (PSP).

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure 1. Simplified Scheme Showing the MainMetabolites Synthesized from Serine.

Supplemental Figure 2. In Silico Analysis of PSP1 ExpressionPattern.

Supplemental Figure 3. Subcellular Localization of PSP1 by StableExpression of PSP1-GFP Fusion Construct under the Control of the35S Promoter in Arabidopsis.

Supplemental Figure 4. Genomic Organization of psp1.1 T-DNAMutant Line.

Supplemental Figure 5. Developmental Analysis of Pollen in Homo-zygous psp1.1/psp1.1 Plants Transformed with PSP1-GFP cDNAunder the Control of the Endogenous PSP1 Promoter by TransmissionElectron Microscopy.

Supplemental Figure 6. Light Microscopy Images of SectionedAnthers from Wild-Type and Homozygous psp1.1/psp1.1 PlantsTransformed with PSP1 under the Control of Either the 35S Promoter(Pro35S:PSP1) or the Endogenous PSP1 Promoter (ProPSP:PSP1) atStage 12.

Supplemental Figure 7. PSP1 Expression in the Anther and Pollen atDifferent Anther Development Stages.

Supplemental Figure 8. Fresh Weight of Wild-Type and Conditionalpsp1.1/psp1.1 Mutant Lines under Different Growth Conditions.

Supplemental Table 1. Significantly Enriched Consensus Sequencesin the Promoter Regions of PSP1.

Supplemental Table 2.Metabolite Levels in the Aerial Parts and Rootsof Wild-Type, PSP1-Silenced, and PSP1-Overexpressing Plants.

Supplemental Table 3. List of Primers Used in This Work.

Supplemental Table 4. Constructs.

16 of 18 The Plant Cell

ACKNOWLEDGMENTS

This work was funded by the Spanish Government: Projects BFU2009-07020 and BFU2012-31519, Juan de la Cierva to J.M.-B., Formación deProfesorado Universitario fellowship to B.C.-M., Agencia Española deCooperación Internacional fellowship to A.D.A; by the Valencian Re-gional Government: PROMETEO/2009/075, GVACOMP/2011/244, andISIC/2013/004; and by the University of Valencia: V Segles fellowship toM.F.-T. We thank Servei Central de Suport a la Investigación Experi-mental of the Universitat de València for technical assistance. We alsothank Hellen Warburton for the language review.

AUTHOR CONTRIBUTIONS

B.C.-M., J.M.-B., A.D.A., M.F.-T., M.A., and S.O. performed the research.J.S., J.P., J.M.-B., and R.R. designed the research. A.R.F. and S.O.analyzed the data. J.M.-B., A.R.F., and R.R. wrote the article.

Received April 17, 2013; revised April 17, 2013; accepted May 22, 2013;published June 14, 2013.

REFERENCES

Alonso, J.M., et al. (2003). Genome-wide insertional mutagenesis ofArabidopsis thaliana. Science 301: 653–657.

Bachelor, M.A., Lu, Y., and Owens, D.M. (2011). L-3-phosphoserinephosphatase (PSPH) regulates cutaneous squamous cell carcinomaproliferation independent of L-serine biosynthesis. J. Dermatol. Sci.63: 164–172.

Bauwe, H., Hagemann, M., and Fernie, A.R. (2010). Photorespiration:Players, partners and origin. Trends Plant Sci. 15: 330–336.

Boavida, L.C., and McCormick, S. (2007). Temperature as a determinantfactor for increased and reproducible in vitro pollen germination inArabidopsis thaliana. Plant J. 52: 570–582.

Capron, A., Chatfield, S., Provart, N., and Berleth, T. (2009). Em-bryogenesis: Pattern formation from a single cell. The ArabidopsisBook 7: e0126, doi/10.1199/tab.0126.

Clough, S.J., and Bent, A.F. (1998). Floral dip: A simplified method forAgrobacterium-mediated transformation of Arabidopsis thaliana.Plant J. 16: 735–743.

Colcombet, J., Boisson-Dernier, A., Ros-Palau, R., Vera, C.E., andSchroeder, J.I. (2005). Arabidopsis SOMATIC EMBRYOGENESISRECEPTOR KINASES1 and 2 are essential for tapetum development andmicrospore maturation. Plant Cell 17: 3350–3361.

Curtis, M.D., and Grossniklaus, U. (2003). A Gateway cloning vectorset for high-throughput functional analysis of genes in planta. PlantPhysiol. 133: 462–469.

Czechowski, T., Stitt, M., Altmann, T., Udvardi, M.K., and Scheible,W.R. (2005). Genome-wide identification and testing of superiorreference genes for transcript normalization in Arabidopsis. PlantPhysiol. 139: 5–17.

Devic, M. (2008). The importance of being essential: EMBRYO-DEFECTIVE genes in Arabidopsis. C. R. Biol. 331: 726–736.

Douce, R., Bourguignon, J., Neuburger, M., and Rébeillé, F. (2001).The glycine decarboxylase system: A fascinating complex. TrendsPlant Sci. 6: 167–176.

Eichler, H.G., Hubbard, R., and Snell, K. (1981). The role of serinehydroxymethyltransferase in cell proliferation: DNA synthesis from serinefollowing mitogenic stimulation of lymphocytes. Biosci. Rep. 1: 101–106.

Ge, X., Chang, F., and Ma, H. (2010). Signaling and transcriptionalcontrol of reproductive development in Arabidopsis. Curr. Biol. 20:R988–R997.

Grienenberger, E., Besseau, S., Geoffroy, P., Debayle, D., Heintz,D., Lapierre, C., Pollet, B., Heitz, T., and Legrand, M. (2009). ABAHD acyltransferase is expressed in the tapetum of Arabidopsisanthers and is involved in the synthesis of hydroxycinnamoylspermidines. Plant J. 58: 246–259.

Handford, J., and Davies, D.D. (1958). Formation of phosphoserinefrom 3-phopho-glycerate in higher plants. Nature 182: 532–533.

Hanson, A.D., and Gregory, J.F., III., (2011). Folate biosynthesis,turnover, and transport in plants. Annu. Rev. Plant Biol. 62: 105–125.

Hanson, A.D., and Roje, S. (2001). One-carbon metabolism in higherplants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52: 119–137.

Ho, C.L., Noji, M., and Saito, K. (1999a). Plastidic pathway of serinebiosynthesis. Molecular cloning and expression of 3-phosphoserinephosphatase from Arabidopsis thaliana. J. Biol. Chem. 274: 11007–11012.

Ho, C.L., Noji, M., Saito, M., and Saito, K. (1999b). Regulation ofserine biosynthesis in Arabidopsis. Crucial role of plastidic 3-phosphoglycerate dehydrogenase in non-photosynthetic tissues. J.Biol. Chem. 274: 397–402.

Ho, C.L., Noji, M., Saito, M., Yamazaki, M., and Saito, K. (1998).Molecular characterization of plastidic phosphoserine aminotransferasein serine biosynthesis from Arabidopsis. Plant J. 16: 443–452.

Ho, C.L., and Saito, K. (2001). Molecular biology of the plastidicphosphorylated serine biosynthetic pathway in Arabidopsis thaliana.Amino Acids 20: 243–259.

Hunt, E., Gattolin, S., Newbury, H.J., Bale, J.S., Tseng, H.M.,Barrett, D.A., and Pritchard, J. (2010). A mutation in amino acidpermease AAP6 reduces the amino acid content of the Arabidopsis sieveelements but leaves aphid herbivores unaffected. J. Exp. Bot. 61: 55–64.

Kalhan, S.C., and Hanson, R.W. (2012). Resurgence of serine: Anoften neglected but indispensable amino acid. J. Biol. Chem. 287:19786–19791.

Kawanabe, T., Ariizumi, T., Kawai-Yamada, M., Uchimiya, H., andToriyama, K. (2006). Abolition of the tapetum suicide program ruinsmicrosporogenesis. Plant Cell Physiol. 47: 784–787.

Kleczkowski, L.A., and Givan, C.V. (1988). Serine formation in leavesby mechanisms other than the glycolate pathway. J. Plant Physiol.132: 641–652.

Lam, H.M., Coschigano, K.T., Oliveira, I.C., Melo-Oliveira, R., andCoruzzi, G.M. (1996). The molecular-genetics of nitrogen assimilationinto amino acids in higher plants. Annu. Rev. Plant Physiol. Plant Mol.Biol. 47: 569–593.

Larsson, C., and Albertsson, E. (1979). Enzymes related to serinesynthesis in spinach chloroplasts. Physiol. Plant. 45: 7–10.

Lisec, J., Schauer, N., Kopka, J., Willmitzer, L., and Fernie, A.R.(2006). Gas chromatography mass spectrometry-based metaboliteprofiling in plants. Nat. Protoc. 1: 387–396.

Locasale, J.W., et al. (2011). Phosphoglycerate dehydrogenasediverts glycolytic flux and contributes to oncogenesis. Nat. Genet.43: 869–874.

Matsuhara, S., Jingu, F., Takahashi, T., and Komeda, Y. (2000).Heat-shock tagging: A simple method for expression and isolation ofplant genome DNA flanked by T-DNA insertions. Plant J. 22: 79–86.

Maurino, V.G., and Peterhansel, C. (2010). Photorespiration: Currentstatus and approaches for metabolic engineering. Curr. Opin. PlantBiol. 13: 249–256.

Michard, E., Lima, P.T., Borges, F., Silva, A.C., Portes, M.T.,Carvalho, J.E., Gilliham, M., Liu, L.H., Obermeyer, G., and Feijó,J.A. (2011). Glutamate receptor-like genes form Ca2+ channels inpollen tubes and are regulated by pistil D-serine. Science 332: 434–437.

Mothet, J.P., Parent, A.T., Wolosker, H., Brady, R.O., Jr., Linden,D.J., Ferris, C.D., Rogawski, M.A., and Snyder, S.H. (2000). D-serine isan endogenous ligand for the glycine site of the N-methyl-D-aspartatereceptor. Proc. Natl. Acad. Sci. USA 97: 4926–4931.

Phosphorylated Pathway of Ser Synthesis 17 of 18

Muñoz-Bertomeu, J., Anoman, A.D., Toujani, W., Cascales-Miñana, B., Flores-Tornero, M., and Ros, R. (2011a). Interactionsbetween abscisic acid and plastidial glycolysis in Arabidopsis. PlantSignal. Behav. 6: 157–159.

Muñoz-Bertomeu, J., Bermúdez, M.A., Segura, J., and Ros, R.(2011b). Arabidopsis plants deficient in plastidial glyceraldehyde-3-phosphate dehydrogenase show alterations in abscisic acid (ABA)signal transduction: interaction between ABA and primary metabolism. J.Exp. Bot. 62: 1229–1239.

Muñoz-Bertomeu, J., Cascales-Miñana, B., Alaiz, M., Segura, J.,and Ros, R. (2010a). A critical role of plastidial glycolytic glyceraldehyde-3-phosphate dehydrogenase in the control of plant metabolism anddevelopment. Plant Signal. Behav. 5: 67–69.

Muñoz-Bertomeu, J., Cascales-Miñana, B., Irles-Segura, A.,Mateu, I., Nunes-Nesi, A., Fernie, A.R., Segura, J., and Ros, R.(2010b). The plastidial glyceraldehyde-3-phosphate dehydrogenaseis critical for viable pollen development in Arabidopsis. PlantPhysiol. 152: 1830–1841.

Muñoz-Bertomeu, J., Cascales-Miñana, B., Mulet, J.M., Baroja-Fernández, E., Pozueta-Romero, J., Kuhn, J.M., Segura, J., andRos, R. (2009). Plastidial glyceraldehyde-3-phosphate dehydrogenasedeficiency leads to altered root development and affects the sugar andamino acid balance in Arabidopsis. Plant Physiol. 151: 541–558.

Muralla, R., Lloyd, J., and Meinke, D. (2011). Molecular foundations ofreproductive lethality in Arabidopsis thaliana. PLoS ONE 6: e28398.

Pfaffl, M.W. (2001). A new mathematical model for relative quantificationin real-time RT-PCR. Nucleic Acids Res. 29: e45.

Pollari, S., Käkönen, S.M., Edgren, H., Wolf, M., Kohonen, P., Sara,H., Guise, T., Nees, M., and Kallioniemi, O. (2011). Enhancedserine production by bone metastatic breast cancer cells stimulatesosteoclastogenesis. Breast Cancer Res. Treat. 125: 421–430.

Possemato, R., et al. (2011). Functional genomics reveal that the serinesynthesis pathway is essential in breast cancer. Nature 476: 346–350.

Riens, B., Lohaus, G., Heineke, D., and Heldt, H.W. (1991). Aminoacid and sucrose content determined in the cytosolic, chloroplastic,and vacuolar compartments and in the phloem sap of spinachleaves. Plant Physiol. 97: 227–233.

Sanders, P.M., Bui, A.Q., Weterings, K., McIntire, N., Hsu, Y.-C.,Lee, P.Y., Truong, M.T., Beals, T.B., and Goldberg, R.B. (1999).Anther developmental defects in Arabidopsis thaliana male-sterilemutants. Sex. Plant Reprod. 11: 297–322.

Scholl, R.L., May, S.T., and Ware, D.H. (2000). Seed and molecularresources for Arabidopsis. Plant Physiol. 124: 1477–1480.

Slaughter, J.C., and Davies, D.D. (1968). The isolation andcharacterization of 3-phosphoglycerate dehydrogenase from peas.Biochem. J. 109: 743–748.

Smyth, D.R., Bowman, J.L., and Meyerowitz, E.M. (1990). Earlyflower development in Arabidopsis. Plant Cell 2: 755–767.

Till, B.J., et al. (2003). Large-scale discovery of induced point mutationswith high-throughput TILLING. Genome Res. 13: 524–530.

Timm, S., Mielewczik, M., Florian, A., Frankenbach, S., Dreissen,A., Hocken, N., Fernie, A.R., Walter, A., and Bauwe, H. (2012).High-to-low CO2 acclimation reveals plasticity of the photorespiratorypathway and indicates regulatory links to cellular metabolism ofArabidopsis. PLoS ONE 7: e42809.

Tolbert, N.E. (1980). Photorespiration. In The Biochemistry of plants,D.D. Davies, ed (New York: Academic Press), pp. 488–525.

Tolbert, N.E. (1997). The C2 oxidative photosynthetic carbon cycle.Annu. Rev. Plant Physiol. Plant Mol. Biol. 48: 1–25.

Walton, N.J., and Woolhouse, H.W. (1986). Enzymes of serine andglycine metabolism in leaves and nonphotosynthetic tissues of Pisumsativum L. Planta 167: 119–128.

Winter, D., Vinegar, B., Nahal, H., Ammar, R., Wilson, G.V., andProvart, N.J. (2007). An “Electronic Fluorescent Pictograph”browser for exploring and analyzing large-scale biological data sets.PLoS ONE 2: e718.

Yamaoka, Y., Yu, Y., Mizoi, J., Fujiki, Y., Saito, K., Nishijima, M., Lee, Y.,and Nishida, I. (2011). PHOSPHATIDYLSERINE SYNTHASE1 is requiredfor microspore development in Arabidopsis thaliana. Plant J. 67:648–661.

Yang, C., Vizcay-Barrena, G., Conner, K., and Wilson, Z.A. (2007).MALE STERILITY1 is required for tapetal development and pollenwall biosynthesis. Plant Cell 19: 3530–3548.

Yoshida, K., Furuya, S., Osuka, S., Mitoma, J., Shinoda, Y., Watanabe, M.,Azuma, N., Tanaka, H., Hashikawa, T., Itohara, S., and Hirabayashi, Y.(2004). Targeted disruption of the mouse 3-phosphoglyceratedehydrogenase gene causes severe neurodevelopmental defectsand results in embryonic lethality. J. Biol. Chem. 279: 3573–3577.

Zhang, Y., Sun, K., Sandoval, F.J., Santiago, K., and Roje, S. (2010).One-carbon metabolism in plants: Characterization of a plastidserine hydroxymethyltransferase. Biochem. J. 430: 97–105.

18 of 18 The Plant Cell

DOI 10.1105/tpc.113.112359; originally published online June 14, 2013;Plant Cell

Pertusa, Manuel Alaiz, Sonia Osorio, Alisdair R. Fernie, Juan Segura and Roc RosBorja Cascales-Miñana, Jesús Muñoz-Bertomeu, María Flores-Tornero, Armand Djoro Anoman, José

ArabidopsisEmbryo Development and for Root Growth in The Phosphorylated Pathway of Serine Biosynthesis Is Essential Both for Male Gametophyte and

 This information is current as of May 23, 2018

 

Supplemental Data /content/suppl/2013/05/24/tpc.113.112359.DC1.html

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists