This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 146/4/1553    most recent pp.107.112698v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Cossegal, M. Articles by Rogowsky, P. Search for Related Content PubMed PubMed Citation Articles by Cossegal, M. Articles by Rogowsky, P. Agricola Articles by Cossegal, M. Articles by Rogowsky, P.

BIOCHEMICAL PROCESSES AND MACROMOLECULAR STRUCTURES

Transcriptional and Metabolic Adjustments in ADP-Glucose Pyrophosphorylase-Deficient bt2 Maize Kernels^1,[W]

Reproduction et Développement des Plantes, UMR 879 INRA-CNRS-ENSL-UCBL, IFR128 BioSciences Lyon-Gerland, F–69364 Lyon cedex 07, France (M.C., P.C., S.M., P.R.); Unité de Recherche en Génomique Végétale, UMR 1165 INRA-CNRS-UEVE, F–91057 Evry cedex, France (S.B., M.-L.M.-M.); Pôle Métabolome de la Plateforme Génomique Fonctionnelle Bordeaux, IFR BVI, F–33883 Villenave d'Ornon, France (A.M., C.D.); Biogemma SAS, Laboratoire de Biologie Cellulaire et Moléculaire, F–63028 Clermont-Ferrand cedex 2, France (V.G., P.P.); and INRA UMR AgroParisTech/INRA MIA, F–75231 Paris cedex 05, France (M.-L.M.-M.)

	ABSTRACT

TOP ABSTRACT RESULTS DISCUSSION CONCLUSION MATERIALS AND METHODS LITERATURE CITED

 
During the cloning of monogenic recessive mutations responsible for a defective kernel phenotype in a Mutator-induced Zea mays mutant collection, we isolated a new mutant allele in Brittle2 (Bt2), which codes for the small subunit of ADP-glucose pyrophosphorylase (AGPase), a key enzyme in starch synthesis. Reverse transcription-polymerase chain reaction experiments with gene-specific primers confirmed a predominant expression of Bt2 in endosperm, of Agpsemzm in embryo, and of Agpslzm in leaf, but also revealed considerable additional expression in various tissues for all three genes. Bt2a, the classical transcript coding for a cytoplasmic isoform, was almost exclusively expressed in the developing endosperm, whereas Bt2b, an alternative transcript coding for a plastidial isoform, was expressed in almost all tissues tested with a pattern very similar to that of Agpslzm. The phenotypic analysis showed that, at 30 d after pollination (DAP), mutant kernels were plumper than wild-type kernels, that the onset of kernel collapse took place between 31 and 35 DAP, and that the number of starch grains was greatly reduced in the mutant endosperm but not the mutant embryo. A comparative transcriptome analysis of wild-type and bt2-H2328 kernels at middevelopment (35 DAP) with the 18K GeneChip Maize Genome Array led to the conclusion that the lack of Bt2-encoded AGPase triggers large-scale changes on the transcriptional level that concern mainly genes involved in carbohydrate or amino acid metabolic pathways. Principal component analysis of ¹H nuclear magnetic resonance metabolic profiles confirmed the impact of the bt2-H2328 mutation on these pathways and revealed that the bt2-H2328 mutation did not only affect the endosperm, but also the embryo at the metabolic level. These data suggest that, in the bt2-H2328 endosperms, regulatory networks are activated that redirect excess carbon into alternative biosynthetic pathways (amino acid synthesis) or into other tissues (embryo).

Genetic studies of the maize kernel started at the beginning of the last century. Together with Drosophila, maize was the favorite model of geneticists because the several hundred kernels on a single segregating ear allowed easy scoring of traits concerning the color, size, or shape of the kernel. Not surprisingly, some of these mutants turned out to lack genes involved in the synthesis of the reserve substances that constitute the nutritional interest of the maize kernel. Kernel mutants were originally obtained spontaneously (Mains, 1949), later on by ethyl methanesulfonate mutagenesis (Neuffer and Sheridan, 1980), and, more recently, by insertional mutagenesis (Clark and Sheridan, 1991; Scanlon et al., 1994). Mutant characterization was first limited to cytological descriptions, later on to biochemical quantifications of metabolites or enzyme activities, and only more recently to the cloning of the underlying genes. For example, the shrunken2 (sh2) mutant was isolated in 1949 (Mains, 1949). The lack of starch in the mature kernel was demonstrated in 1953 (Laughnan, 1953) and the absence of ADP-Glc pyrophosphorylase (AGPase) activity was demonstrated in 1966 (Tsai and Nelson, 1966), whereas the isolation of the mutated gene had to wait until 1990 (Bhave et al., 1990).

In addition to a dozen mutants involved in starch synthesis, several developmental mutations have been cloned. Mutations in the two plastid ribosomal proteins L35 (Magnard et al., 2004) and S9 (Ma and Dooner, 2004) lead to lethality of the early embryo, but do not influence endosperm development. The characterization of empty pericarp (emp) mutants showed that loss of the negative regulator of the heat shock response Emp2 (Fu et al., 2002) or the mitochondrion-targeted pentatricopeptide repeat protein Emp4 (Gutierrez-Marcos et al., 2007) affects both embryo and endosperm development. Lesions in the Cys proteinase Dek1 (Lid et al., 2002) cause a defective kernel (dek) phenotype similar to the one observed after loss of the transmembrane protein Lachrima (Stiefel et al., 1999). Finally, the analysis of etched (et) mutants indicated that the zinc ribbon protein Etched1 (da Costa e Silva et al., 2004) is needed to avoid fissures and cracks in the maize kernel. Many more genes transcribed in the maize kernel have been characterized in great detail. Even though in many cases rather precise functions can be attributed to these genes by sequence homology to proteins of known function, these putative functions need to be demonstrated by the analysis of corresponding mutants or transgenic knockouts.

Among all the developmental and biosynthetic pathways of the maize kernel, the one leading to starch is certainly the best characterized. Most genes involved in the major steps of starch synthesis have been identified by mutant analysis. Along the pathway leading from Suc via activated hexoses to the starch polymer we find, in order, Miniature1 (Mn1) coding for the cell wall invertase IncW2, Shrunken1 (Sh1) and Sucrose synthase1 (Sus1) coding for Suc synthase, Sh2 and Brittle2 (Bt2) coding for AGPase, Bt1 coding for an ADP-Glc transporter, Waxy1 (Wx1) coding for the granule-bound starch synthase GBSSI, Sugary2 (Su2) and Dull1 (Du1) coding, respectively, for the soluble starch synthases SSIIa and SSIIIa, Amylose extender1 (Ae1) coding for branching enzyme BEIIb, and Su1 coding for the debranching enzyme ISAI (Hannah, 2005, and refs. therein).

The rate-limiting step in starch synthesis is the synthesis of ADP-Glc from Glc-1-P and ATP by AGPase (EC 2.7.7.27; Russell et al., 1993). AGPase is a heterotetramer composed of two large subunits encoded by Sh2 and two small subunits encoded by Bt2. Homozygous sh2 or bt2 kernels collapse (shrink) during the maturation process, becoming angular (brittle) at maturity when they contain very little starch and strongly increased amounts of Suc (Laughnan, 1953; Cameron and Teas, 1954). Whereas the presence of two different subunits is typical for higher plants, both photosynthetic and nonphotosynthetic bacteria have only a single AGPase gene. All genes have evolved from a common ancestor and substantial sequence similarity remains between Sh2 and Bt2, although it is smaller than the similarity between Bt2 and genes encoding small subunits in other species (Smith-White and Preiss, 1992). Both maize genes individually have small but detectable catalytic activity in Escherichia coli (Burger et al., 2003), questioning earlier reports on potato (Solanum tuberosum) AGPase genes that attributed catalytic activity only to the small subunit (Ballicora et al., 1995).

The presence of residual AGPase activity in sh2 (12%) and bt2 (17%) kernels was the first indication for additional AGPase genes in the maize genome (Dickinson and Preiss, 1969) and led ultimately to the isolation of two additional genes from an embryo cDNA library (Giroux and Hannah, 1994). Agp1 (large subunit) and Agp2 (small subunit) show predominant expression in the embryo in northern blots, contrary to Sh2 and Bt2, which are specifically expressed in the endosperm (Giroux and Hannah, 1994). Finally, a third small subunit gene, L2, was identified in a leaf cDNA library (Prioul et al., 1994). The genes Agp2 and L2 were renamed Agpsemzm and Agpslzm, respectively, when the intron/exon structure of Agpslzm was determined (Hannah et al., 2001). Although the three Agps genes clearly have a common ancestor, their physiological roles have diverged after gene duplication. Whereas Agpslzm is responsible for the transient accumulation of starch during the daytime in the source organ leaf, Agpsemzm and Bt2 take care of the permanent storage of starch in the sink organs embryo and endosperm.

Another point of divergence between the three Agps genes is their subcellular localization. Although it has been a matter of long-standing debate, there is now growing consensus that the Bt2 protein is located in the cytoplasm rather than the plastids of endosperm cells. The gene lacks transit peptide sequences, the protein does not seem to be processed (Giroux and Hannah, 1994), 95% of the AGPase activity of endosperm cells is found in the cytoplasmic fraction (Denyer et al., 1996), and Bt2-GFP fusions locate to the cytoplasm of transgenic tobacco (Nicotiana tabacum) plants (Choi et al., 2001). The cytoplasmic localization of Bt2 is shared by its counterparts in barley (Hordeum vulgare; Thorbjornsen et al., 1996a), wheat (Triticum aestivum; Burton et al., 2002) and rice (Oryza sativa; Ohdan et al., 2005), but differs fundamentally from the plastid localization of dicot AGPS proteins and monocot nonendosperm AGPS proteins (Beckles et al., 2001). The authors speculate that a cytosolic AGPase has the advantage of committing carbon to starch synthesis rather than other metabolic pathways in the presence of high Suc levels.

The objective of our research program is the identification of genes causing an emp phenotype of the maize kernel and the phenotypic characterization of the corresponding mutants. We report here the isolation of a novel insertional allele of bt2, as well as a detailed molecular and phenotypic characterization of the bt2-H2328 mutant, including gene-specific reverse transcription (RT)-PCR, transcriptome and metabolome analyses never performed before on bt2 mutants.

	RESULTS

TOP ABSTRACT RESULTS DISCUSSION CONCLUSION MATERIALS AND METHODS LITERATURE CITED

 

Isolation and Molecular Characterization of a Mutator Insertion in Bt2

During a forward genetics approach aimed at the systematic cloning of monogenic recessive mutations responsible for kernel phenotypes in the Mutator (Mu)-induced Biogemma mutant collection (Martin et al., 2006), the flanking sequences H6P (EU137670) and H6P' (EU137671) showed cosegregation with the kernel phenotype of mutant H2328. Genetic mapping of H6P on chromosome 4 in BIN 4.04/4.05 between markers UMC49d and PHI026 incited us to test whether mutant H2328 was able to complement the bt2 mutation, a well-known kernel mutation with a very similar map position. Absence of complementation led us to the conclusion that there was close linkage between H6P, H6P', and Bt2 and that a mutation in Bt2 rather than H6P' was responsible for the phenotype of H2328.

In the search for a molecular lesion in the Bt2 gene of H2328, gene-specific primers were used either in pairs to yield four PCR fragments that covered nearly all of the Bt2 gene or in combination with a Mu primer in a more specific search for Mu insertions (see "Materials and Methods"). PCR on genomic DNA of wild-type, heterozygous, and homozygous mutant plants revealed a Mu insertion in one of the four fragments. Cloning and sequencing of the flanking sequences of the insertion showed that the insertion was located in exon 6 of the Bt2 gene generating a 9-bp duplication of bases TGATGTGAC (position 4,422–4,431 in accession no. AF334959). Perfect cosegregation between this insertion and the kernel phenotype of mutant H2328 was found on a segregating population of 66 individuals. From here on, this novel bt2 allele will be referred to as bt2-H2328.

Phenotypic Analysis

Mature bt2 kernels contain essentially a seemingly normal embryo, whereas the surrounding endosperm is completely collapsed. Numerous phenotypic analyses have been performed on bt2 mutants, but most of them focused on mature kernels or biochemical features. For cytological and molecular analysis of bt2 kernels throughout development, it seemed important to have an easy means to distinguish wild-type from mutant kernels. Following up on earlier observations on sh2 kernels, we marked the two size classes visible on immature 30-DAP kernels with black dots (bigger, plumper) or red dots (smaller, slight depression; Fig. 1A ). Observation at maturity showed that all kernels with black dots were mutant, whereas kernels with red dots were wild type (Fig. 1B). Similar observations were made on segregating ears of mutants H182, H816, bt2, and, to a lesser extent, sh2 (data not shown).

View larger version (70K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. Macroscopic phenotype of bt2-H2328. A and B, On a segregating ear of bt2-H2328, wild-type kernels were marked with a black dot and mutant kernels with a red dot. The ear was photographed at 30 (A) and 60 DAP (B). Mutant kernels were plumper than wild-type kernels at 30 DAP and collapsed at 60 DAP. C to F, The same ear of a self-pollinated bt2-h2328 homozygous mutant plant was photographed at 31 (C), 35 (D), 40 (E), and 45 DAP (F). Black arrows, Mutant kernels already collapsed at 31 DAP; red arrows, mutant kernels still plump at 40 DAP.

Cytological observations of bt2-H2328 mutant kernels before kernel collapse confirmed that important differences existed between mutant and wild-type kernels at 30 DAP (Fig. 2, A and B). There was a strong reduction of starch grain number and size in endosperm cells of bt2-H2328 mutant kernels (Fig. 2, B and E). Only the endosperm seemed to be affected and not the embryo (Fig. 2, C and F). In fact, the overall aspect of mutant endosperm cells was reminiscent to that of mutant and wild-type embryo cells. At 30 DAP, the differences in the endosperm were strongest in the apical (Fig. 2, I and M) and central (Fig. 2, J and N) parts and less pronounced in the basal (Fig. 2, G and K) and lateral parts (Fig. 2, H and L) of the endosperm.

View larger version (123K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Cytological observations of bt2-H2328 kernels at 30 DAP. Wild-type and homozygous mutant kernels from a segregating ear were collected at 30 DAP. Paraffin sections were stained by the periodic acid-Schiff procedure coloring starch grains in pink. Global views (A and D), high magnifications (B, C, E, and F), and low magnifications (G–N) of embryo (emb) and endosperm tissue (en) are presented.

There has been a long-standing debate concerning the subcellular localization of AGPase in maize endosperm. Both cytoplasmic (Denyer et al., 1996) and plastid localization (Echeverria et al., 1988) have been reported. In addition, alternative splicing of a single gene has been known to be at the origin of cytoplasmic and plastid isoforms in barley (Thorbjornsen et al., 1996b), wheat (Burton et al., 2002), and rice (Ohdan et al., 2005). Interestingly, our in-depth analysis of the corresponding genomic sequences showed that the first intron of Bt2 contained a 255-bp sequence with 89% homology to the end of exon 1 of Agpslzm (Fig. 3 ). Using a start codon at the beginning and a splice site at the end of the conserved sequence, we derived a theoretical amino acid sequence of BT2 that carried an alternative N terminus (Fig. 3, BT2b). An interrogation of the TargetP Web site (http://www.cbs.dtu.dk/services/TargetP) predicted BT2b to be targeted to the plastid with a score of 0.971, just like AGPSLZM (score 0.946) or APSEMZM (score 0.862), whereas BT2a was predicted to remain in the cytosol (score 0.184 for plastid, score 0.770 for other). Very similar findings have been published recently by other authors (Rosti and Denyer, 2007).

View larger version (42K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 3. Gene structure of maize Agps genes. Top, Schematic drawing of the intron/exon structure of Agpsemzm, Agpslzm, Bt2b (plastidial form), and Bt2a (cytoplasmic form) with introns as narrow green rectangles and exons as wide green rectangles. Middle, Summary of exon lengths of Agps genes and sequence similarities between Agps exons. Bottom, Alignment of the deduced N-terminal amino acid sequences of AGPSEMZM, AGPSLZM, BT2b, and BT2a. Conserved residues are shaded in black; amino acids predicted to be part of chloroplast target sequences are written in red.

Anchoring of the Agps genes on the physical map of maize allowed the establishment of synteny between the regions encompassing Bt2 (chromosome 4, BIN 4.05, ctg 165) or Agpslzm (chromosome 1, BIN 1.07, ctg 44) and a single region in the rice genome (chromosome 8, Os08g0345800), on one hand, and between the region containing Agpsemzm (chromosome 2, BIN 2.06, ctg 92) and a second region in the rice genome (chromosome 9, Os09g0298200), on the other hand. The two rice regions corresponded to the only two Agps genes present in the rice genome, which code for OsAgps1 (Os09g0298200) targeted to the plastid and OsAgps2 (Os08g0345800) with dual plastidial and cytoplasmic targeting (Lee et al., 2007). These data confirmed on the physical map recent observations made on the genetic map (Rosti and Denyer, 2007), and supported the conclusion of the authors that Bt2 and Agpslzm arose from gene duplication after the separation of maize and rice.

Expression of Agps Genes

We designed gene-specific primers to gain further insight into the expression level of the three Agps genes in major maize tissues, in different kernel compartments, and at various stages of kernel development. Control experiments demonstrated the specificity of the primer pairs by cloning and sequencing of PCR products and the absence of genomic contamination by the use of unrelated intron-spanning control primers (see "Materials and Methods"). In wild-type plants, the three Agps genes were not tissue specific, although they showed preferences in terms of their spatial and temporal expression pattern (Fig. 4 ). As expected, Bt2 was strongly expressed in kernels, but moderate expression was also found in ears and tassels and weak expression in leaves. Separate analysis of Bt2a (cytoplasmic) and Bt2b (plastidial) allowed attribution of most of the kernel transcript to Bt2a and most of the transcript seen in other tissues to Bt2b. For Agpslzm, we confirmed much stronger expression in leaves than in kernels and demonstrated additional expression of comparable strength in ears and tassels. Agpsemzm indeed showed moderate expression in the kernel, but even higher expression was discovered in ears. It was weakly expressed in leaves, stems, and even roots (Fig. 4A).

View larger version (76K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 4. Expression of Apgs gene RT-PCR studies with gene-specific primers for Bt2a (5' end), Bt2b (5' end), Bt2 (3' end of both Bt2a and Bt2b), Agpslzm (3' end), and Agpsemzm (3' end) were carried out on the tissues indicated. Lby, Leaf blade young; Lsy, leaf sheath young; Lba, leaf blade adult; Lsa, leaf sheath adult; Rs, root seedling; Rm, root mature; Sa, seedling aerial parts; Ei, ear immature; Em, ear mature; Ti, tassel immature; Tm, tassel mature; Si, silk; St, stem; K12, kernel at 12 DAP; 12, 12 DAP; 35, 35 DAP; 50, 50 DAP; 70, 70 DAP.

Transcript levels in the bt2-H2328 mutant were assessed both in leaves of wild-type and mutant sister plants and in 35-DAP kernels of segregating ears (Fig. 4C). The mutant was not a complete knockout because Bt2 was expressed in mutant kernels albeit at lower levels than in wild-type controls. The reduction concerned both splice forms Bt2a and Bt2b. Unexpectedly, the expression of Agpslzm and Agpsemzm was also affected in the bt2-H2328 mutant (Fig. 4C). Quantification by quantitative RT-PCR indicated down-regulation by a factor of 14 (Bt2), a factor of 3 (Agpslzm), or a factor of 2 (Agpsemzm) in bt2-H2328 kernels (data not shown).

Taken together, these results showed that Bt2a, Agpslzm, and Agpsemzm were preferentially but not specifically expressed in endosperm, leaf, and embryo, respectively. Bt2b was expressed in almost all tissues tested and had a very similar expression pattern as Agpslzm. The results suggested also that the residual activity of AGPase in mutant kernels of many bt2 alleles could be readily explained by the expression of Agpslzm, Agpsemzm, and/or residual Bt2 expression.

Transcriptome Analysis

Despite a wealth of knowledge on the enzymology and allosteric regulation of maize AGPase in wild-type and bt2 kernels, only very limited information was available on the consequences of AGPase dysfunction on gene expression (Giroux et al., 1994). To investigate in a nonbiased fashion whether the block of starch synthesis and the subsequent accumulation of Suc caused adjustments of metabolic pathways on the transcriptional level, a microarray experiment using the 18K GeneChip Maize Genome Array (Affymetrix) was carried out. Whole kernels (pericarp + endosperm + embryo) from self-pollinated ears of heterozygous plants (+/bt2-H2328) were collected at 35 DAP and visually divided into pools of wild-type and mutant kernels. Two different ears were used to obtain a biological duplicate.

The statistical analysis resulted in a large number of differentially expressed genes. A total of 2,345 of the 17,734 probe sets present on the array were declared differentially expressed (Bonferroni P value <0.05). These genes showed either increased or decreased expression with an absolute value of log₂ ratio >1.08. We then asked the question whether certain Gene Ontology (GO) terms were overrepresented among differentially expressed probe sets in comparison to all probe sets. Not surprisingly, polysaccharide metabolic processes (GO:0005976) and related terms showed a statistically significant (P value <0.01) increase among differentially expressed probe sets (Table I ). Interestingly, amino acid metabolic processes (GO:0006520) and related terms were even more frequent in the list of differentially expressed probe sets (Table I). On the other hand, there was no significant difference in the case of lipid metabolic processes (GO:0006629), protein metabolic processes (GO:0019538), or transcription (GO:0006350). A further dissection of amino acid metabolic processes revealed that the changes in the transcriptome concerned the synthesis, rather than the catabolism, of amino acids and that most amino acid families were concerned, even though at various degrees (Table I).

View larger version (52K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 5. Summary of transcriptional and metabolic changes in bt2-H2328 kernels. The scheme shows selected pathways of the carbohydrate and amino acid metabolism. Glycolyis, starch synthesis, TCA, and PPP are surrounded by circles; the different amino acid biosynthetic pathways are not. Solid arrows indicate a single enzymatic step, dashed arrows more than one enzymatic step. Enzymes are numbered according to Table II. Enzymes encoded by genes over- or underexpressed in bt2-H2328 kernels are highlighted in green and pink, respectively. Yellow highlighting indicates that at least one corresponding gene was overexpressed and at least one corresponding gene is underexpressed. Metabolites showing higher accumulation in bt2-H2328 than in wild-type kernels, endosperms, or embryos are highlighted in green.

To see whether these changes in amino acid metabolism had consequences on the transcriptional regulation of storage proteins, we checked the expression of 17 Zein genes as well as of a Legumin and a Globulin gene as classified earlier (Woo et al., 2001). Among the 51 corresponding sense probe sets on the array, not a single one was up-regulated in bt2-H2328 kernels, whereas 23 were down-regulated, 23 showed no change, and five were indistinguishable from background (data not shown).

Taken together, these data indicate that the lack of Bt2-encoded AGPase triggers large-scale changes on the transcriptional level that concerns mainly genes involved in carbohydrate or amino acid metabolic pathways.

Metabolome Analysis

To provide further insight into the impact of the bt2-H2328 mutation on the metabolome of the developing maize kernel, a quantitative analysis by ¹H NMR spectroscopy was carried out on wild-type and mutant kernels of three segregating ears at 35 DAP. For each ear, a pool of 10 wild-type kernels and a pool of 10 mutant kernels were constituted to have a biological triplicate. For each sample, two extractions were made in parallel to obtain a technical duplicate. In parallel, 15 additional wild-type kernels and 15 additional mutant kernels of each ear were microdissected into embryo, on one hand, and endosperm (plus adhering pericarp), on the other, and pooled.

In the first instance, the NMR signatures (see Supplemental Figs. S1–S3) were analyzed by principal component analysis (PCA) of 0.04-ppm intervals of the entire spectra with the exclusion of the water and ethanol regions (see Supplemental Figs. S4 and S5). Comparing wild-type and bt2-H2328 endosperm, the first two PCA components explained 94% of total variability. PC1 with PC2 clearly separated the wild-type from the mutant samples. Examination of PC1 and PC2 loadings suggested that the difference between the wild-type and bt2-H2328 samples involved Suc, Fru, malate, and Ala. In the embryo, the first two PCA components explained 90% of total variability. PC1 with PC2 clearly separated the mutant from the wild-type samples. Examination of PC1 and PC2 loadings suggested that the difference between the wild-type and mutant samples involved lactic acid, Suc, citrate, lipids, and Pro. These data clearly demonstrated that the bt2-H2328 mutation did not only affect metabolite levels in the endosperm, but also in the embryo.

In parallel, 18 to 21 individual metabolites identified in the NMR signatures were quantified (see Supplemental Tables S1–S3). t tests allowed the pinpointing of metabolites with statistically significant differences between wild-type and bt2-H2328 kernels, endosperms, and embryos (Tables III–V ). Whereas increased levels of Suc and reducing sugars had been described repeatedly for bt2 kernels (Cameron and Teas, 1954), many of the other listed metabolites had never been measured before. Increases in citrate, malate, and fumarate levels indicated an influence of the bt2-H2328 mutation on the TCA cycle, whereas increases in Ala, Val, and Tyr levels revealed compensation in amino acid metabolism. Whereas these changes mainly concerned the endosperm, the embryo did not only share some of the changes (citrate, choline, Suc), but also showed its proper adjustments, in particular in lipid metabolism.

	DISCUSSION

TOP ABSTRACT RESULTS DISCUSSION CONCLUSION MATERIALS AND METHODS LITERATURE CITED

This study also highlights the difficulties of forward genetics with Mu-induced populations. A systematic transposon display approach on 40 kernel mutants yielded only one insertion cosegregating with the phenotype, which in the end turned out to be only tightly linked to, but not causal, for the phenotype. Whereas technical problems related to the high copy number and high somatic transposition rate of Mu certainly contributed to this very limited success, we also hypothesize that the active Mu stock used for the initial mutagenesis may have promoted the transposition of other elements not detected by our Mu-oriented transposon display, increased the frequency of single base mutations, or provoked epigenetic alterations.

Agps Expression in Maize Tissues

Our gene-specific RT-PCR expression data of the three Agps genes in major maize tissues and in kernel compartments throughout development enlarge and complement earlier studies based on enzymological or immunological differences of AGPase proteins in certain tissues or size differences of Agps transcripts in mutant kernels (Preiss et al., 1971; Giroux and Hannah, 1994). Whereas expression is strong in the tissues, where Bt2 (endosperm), Agpsemzm (embryo), and Agpslzm (leaf) were originally isolated, this expression is by no means specific. For example, we reveal substantial expression of all three genes in immature and mature ears and tassels. In the kernel, we document expression of Agpslzm in the embryo and of Agpsemzm in the endosperm. Taken together, these data indicate that Agps genes may play important roles in organs other than the classical source (leaf) and sink organs (kernel) and that residual AGPase activity found in bt2 kernels can be attributed not only to Agpsemzm, but also to Agpslzm.

Agps Gene Structure and Evolution

Alternative splicing of the first intron of Agps genes leading to either cytoplasmic or plastid-targeted isoforms of AGPase is well documented in barley (Thorbjornsen et al., 1996b), wheat (Burton et al., 2002), and rice (Lee et al., 2007). We confirmed and extended very recent findings of alternative splicing of Bt2 in maize (Rosti and Denyer, 2007) by splice variant-specific RT-PCR in all major maize tissues. The expression of Bt2a, the variant coding for the cytoplasmic isoform, was limited to the kernel and, more precisely, to the developing endosperm up to the dehydration stage, substantiating earlier hypotheses that the cytoplasmic isoform, which is not found outside the grass tribe, is a particularity of grass endosperm, where it may have a functional role in partitioning large amounts of carbon into starch when Suc is abundant (Beckles et al., 2001). In contrast, Bt2b was expressed at various levels in almost all tissues analyzed and its overall expression pattern was very similar to Agpslzm. This is additional support for the model of Rosti and Denyer (2007), which claims that Bt2 and Agpslzm arose by gene duplication from a common ancestor after the separation of maize from rice, wheat, and barley, that Agpslzm subsequently lost the capacity to code for the cytoplasmic isoform and specialized in coding for the plastid isoform, whereas Bt2 maintained the capacity to code for both isoforms, but to some extent specialized in coding for the cytoplasmic one. A closer look at the upstream sequences of Agpslzm and Bt2b reveals substantial sequence similarity of 90% (not counting the numerous gaps) in the 200 bp immediately upstream of the putative ATGs, which drops to 48% in the 1,800 bp further upstream. Based on a clone from the maize full-length cDNA project (accession no. DR969810), the 5'-UTR of Agpslzm is 130 bp. Consequently, conserved cis-elements are either located very close to the transcription initiation site in the highly conserved region or in small clusters of conserved nucleotides found in the region with lower overall homology. In addition, the newly available genomic sequence of Agpsemzm allows confirmation on the level of intron sizes and sequences that the similarity between Agpslzm and Bt2 is substantially higher than between Agpsemzm and either gene. Anchoring of the Agps genes on the physical maize map indicates that the genomic regions of Bt2 and Agpslzm show synteny with a single region in rice around OsAgps2, a gene exhibiting alternative splicing and targeting to both the cytoplasm and plastid (Lee et al., 2007). On the other hand, Agpsemzm shows synteny with OsAgps1, which is targeted exclusively to the plastid (Lee et al., 2007).

The comparison of the remaining genomic sequences of Bt2, Agpslzm, and Agpsemzm reveals a conserved structure composed of 10 exons and nine introns. All three genes share the peculiarity that the last intron is situated beyond the coding sequence in the 3'-UTR. Consequently, it is not surprising that the sequence similarity in exon 10 (52% to 81%) is considerably lower than in all preceding exons (81% to 97%), with the exception of exon 1. Sequence similarity is highest in exons 5 and 6, hinting possibly at some functionally important features in this part of the protein in addition to the four well-characterized domains, ATP site (junction exon 2/exon 3), catalytic site (exon 3), Glc-1-P site (exon 3), and activator site (exon 9; Crevillen et al., 2003).

Influence of bt2-H2328 on the Expression of Genes Involved in Polysaccharide and Amino Acid Metabolism

A comparison of the transcriptome of bt2-H2328 and wild-type kernels revealed a widespread influence of the mutation on transcription because it affected 13% of the probe sets present on the array. Expression differences at 35 DAP concerned primarily genes involved in polysaccharide or amino acid metabolism, but not in lipid or protein metabolism. Whereas the 18K Affymetrix chip admittedly does not represent the entire maize genome and many of the probe sets present on the chip do not carry an annotation allowing a link to GO terms, the data nevertheless strongly support the idea that a block in starch synthesis triggers adjustments on the transcriptional level that favor the flow of excess carbon into glycolysis, TCA, and amino acid synthesis.

A more precise dissection of the metabolic pathways that are turned on or off in mutant kernels is hampered by several factors. (1) In many cases, it is impossible to determine the subcellular localization of the encoded enzymes because the probe sets on the array are frequently based on EST assemblies that only partially cover a given gene; (2) only in rare cases is information available on the expression pattern of a given gene in the different compartments of the maize kernel; (3) in many metabolic pathways, not all the enzymes involved are represented on the array; (4) most enzymes are encoded by multigene families and individual members do not always show the same trends of expression; and (5) many of the encoded enzymes participate in more than one metabolic pathway and/or can function in two directions. For example, phosphoglucomutase (4 in Fig. 5) is needed in Suc biosynthesis and Suc degradation, in starch synthesis and starch degradation, and can be found both in the cytoplasm and plastids. Consequently, it is difficult to draw any conclusion from the down-regulation of one of the three genes present on the array.

Nevertheless, some information can be gained concerning the fate of Suc. There are three enzymes involved in Suc synthesis and/or cleavage: invertase (IncW, Ivr; 1 in Fig. 5), Suc synthase (SuSy; 2 in Fig. 5), and Suc phosphate synthase (SPS; 6 in Fig. 5). Among the six characterized maize invertase genes (Kim et al., 2000), only IncW1 is up-regulated. IncW1 is expressed in the entire endosperm and is the major invertase transcript found during the filling stage. However, its biological role has been elusive in contrast to the IncW2-encoded invertase, which is expressed only at the base of the endosperm and is involved in Suc unloading (Chourey et al., 2006). Because IncW1 is a cell wall-bound invertase, its overexpression leads to increased, irreversible Suc cleavage in the apoplast and consequently to increased extracellular levels of Glc and Fru. Among the three characterized SuSy genes (Carlson et al., 2002), only Sus1 is down-regulated in bt2-H2328 kernels. Sus1 activity in the kernel is smaller than that of Sh1 and both enzymes are thought to contribute mainly to cellulose, rather than starch, biosynthesis. The concomitant down-regulation of genes encoding UDP-Glc dehydrogenase (3 in Fig. 5) and UDP-glycosyltransferase (7 in Fig. 5) points to a general repression of pathways using Suc for cell wall biosynthesis. Finally, there is contrasting regulation for genes encoding SPS, which is mainly used in leaves to synthesize Suc but has also been found in the maize kernel, where it is supposed to catalyze the resynthesis of Suc in the basal part of the endosperm (Im, 2004). Whereas the expression patterns of seven SPS genes have been reported in maize, the authors make it impossible to relate their proprietary data to public sequences or genes (Lutfiyya et al., 2007). Our own RT-PCR data indicate that the up-regulated gene is a constitutive gene, whereas the two down-regulated genes are primarily expressed in leaves (data not shown). One might speculate that the down-regulation is a feedback mechanism normally operating in the leaf, whereas the induction could be due to a lack of intermediates in cell wall synthesis.

Because knowledge is a lot more limited concerning gene families encoding enzymes in glycolysis, TCA, PPP, and amino acid biosynthesis, it is difficult to discuss individually all the up- and down-regulated genes shown in Figure 5. Nevertheless, some general trends emerge. Only up-regulation is found for genes involved in the PPP and the synthesis of aromatic amino acids, and considerably more up- than down-regulation for genes in the TCA cycle (Fig. 5), whereas the picture is more complex for glycolysis and the synthesis of other types of amino acids. This confirms on the gene level the conclusion of the global analysis, that excess Suc is used for the synthesis of amino acids rather than starch.

In addition, our data corroborate the overall conclusion of two previous, much more limited studies, that there is coordinated transcriptional regulation of storage product genes in the maize endosperm (Doehlert and Kuo, 1994; Giroux et al., 1994). However, whereas the authors of the latter study had found an increase in Sh1, Sus, Sh2, Wx1, Zein (small and large), Agp1, and Agpsemzm genes at 22 and 30 DAP of the bt2-7480 mutant, we found either constant expression (Sh1, Wx1) or decreased expression (Sh2, Zein, Sus, Agpsemzm) at 35 DAP in the bt2-H2328 mutant. These differences may be caused in part by technical differences; our microarray data (15 probes/gene) and RT-PCR data obtained with wild-type and mutant kernels stemming from the same segregating ear were probably more specific and more homogeneous than the previous northern data with wild-type and mutant kernels harvested from sibling plants. Biological explanations include the differences in the bt2 allele, the genetic background, and the developmental stage. Because there are important differences and sometimes inversions in the relative abundance of gene expression between 14, 22, and 30 DAP, one does not necessarily expect similar results at 30 (previous study) and 35 DAP (our study). Finally, our data are easier to reconcile with a decrease in Sh2 protein in bt2 kernels at 30 DAP (Giroux et al., 1994) and a 3-fold decrease in zein content in mature bt2 kernels (Tsai et al., 1978).

Suc is a prime candidate to play a role in the mechanism, which translates the strong reduction of AGPase activity into transcriptional activation or repression because it is strongly accumulated in the bt2-H2328 mutant and known to act as a signal molecule (Borisjuk et al., 2004; Koch, 2004). Suc response elements have been pinpointed in promoters and trans-acting factors binding to theses cis-elements have been identified (Sun et al., 2003). The transcript levels of Sh1, Sh2, and Bt2 have been shown to vary when maize kernels are exposed to different Suc concentrations (Giroux et al., 1994), and other genes coding for enzymes in carbohydrate or amino acid metabolism may be regulated in the same way. The precise mechanism of the signal transduction downstream of Suc remains to be elucidated, but two functionally redundant protein kinases have recently been implicated in this pathway in Arabidopsis (Baena-Gonzalez et al., 2007). It would be interesting to assess the role of the orthologous proteins on global transcriptional regulation in the maize kernel.

Metabolic Adjustments in bt2-H2328 Kernels

¹H NMR metabolomics confirmed several of the predictions based on the transcriptome data of bt2-H2328 kernels. Increased levels of citrate, fumarate, and malate correlate well with up-regulation of genes in the TCA cycle, increased levels of Tyr, Val, and Ala with an overall effect on the transcriptome of amino acid synthesis. In addition, stimulation of glycolysis in bt kernels is corroborated by earlier results in a nondefined bt mutant, which demonstrated a slight increase in Glc-1-P and Glc-6-P, no influence on Fru-6-P or Fru-1,6-bisP, and doubling up in dihydroxyacetone phosphate, Glc-3-P, and Fru-2,6-bisP (Tobias et al., 1992). The accumulation of lactate found in this study may also be indicative of this stimulation. The overall data also fit well with the finding that the bt2 and sh2 mutations, but not other mutations in starch synthesis (incW2, sh1, sus1, ae, wx1, or su1), trigger significantly increased fluxes of carbohydrates through glycolysis, PPP, TCA cycle, and gluconeogenesis (Spielbauer et al., 2006). Increased levels of the TCA cycle metabolites malate and citrate had also been found in transgenic potato tubers with strongly reduced AGPase activity due to an antisense construct of the small subunit gene Agp-B (Geigenberger et al., 1999). This suggests that at least some of the metabolic adjustments made in sink organs with reduced AGPase activity are similar in monocot and dicot.

Surprisingly, the metabolic changes in bt2-H2328 kernels were not restricted to the endosperm, where Bt2 is preferentially expressed, but concerned also the embryo, where the nonaffected Agpsemzm gene is the main contributor to AGPase activity. Because Suc levels are also increased in the embryo, one may hypothesize that part of the excess Suc is transported from the endosperm to the embryo, where it triggers similar metabolic adjustments as in the endosperm. These findings on the metabolic level are new, even though differences between AGPase activities in wild-type and bt2 kernels had not only been established in endosperm but also in the embryo but not the pericarp (Dickinson and Preiss, 1969).

The biological role of other metabolites with increased levels in bt2-H2328 kernels is less straightforward. Choline is primarily used for the synthesis of the vital membrane lipid phosphatidyl choline, but its increase in mutant kernels may rather be linked to its additional role as a precursor for the synthesis of betaine. Maize belongs to the plant species that accumulate betaine in response to high salinity, cold, and drought (Rhodes and Hanson, 1993), and the synthesis of this osmoprotectant could possibly counteract the osmotic stress exerted by high Suc levels. GABA is a nonprotein amino acid that also accumulates in response to biotic and abiotic stresses (Bouche and Fromm, 2004). Its synthesis via the GABA shunt bypasses two steps of the TCA cycles, among them succinyl-CoA ligase (19 in Fig. 5), the only one showing a clear transcriptional down-regulation in bt2-H2328 kernels. This suggests that the TCA cycle not only is activated in mutant kernels, but also that it functions in a modified manner.

	CONCLUSION

TOP ABSTRACT RESULTS DISCUSSION CONCLUSION MATERIALS AND METHODS LITERATURE CITED

 
The antagonistic relationship between the starch and protein content of the maize kernel is a well-known phenomenon among maize breeders and global analyses have shown that increases in starch were generally linked to decreases in protein (Duvick and Cassman, 1999). Mutants in starch synthesis are a means to investigate the starch/protein balance. We show here that, in the bt2-H2328 mutant, important changes take place in the transcription of genes involved in amino acid synthesis and that there is an increase in the content of certain amino acids. Whereas at first sight these data could be interpreted as a shift in the type of reserve substances from starch to protein, things are certainly a lot more complex. In fact previous data document that the bt2 mutation causes a strong decrease in zeins, the major storage proteins in the maize kernel, whereas the content of non-zein proteins is increased (Tsai et al., 1978). Taken together with the synthesis of stress-related compounds in mutant kernels, the data suggest that at least some of the transcriptomic and metabolic adjustments reflect a response to an emergency situation rather than a concerted shift to the synthesis of alternative reserve substances.

	MATERIALS AND METHODS

TOP ABSTRACT RESULTS DISCUSSION CONCLUSION MATERIALS AND METHODS LITERATURE CITED

 

Plant Material

The bt2-H2328 mutant was isolated from a Mu-based mutant collection established by Biogemma SAS (Martin et al., 2006). Wild-type lines used were F252 (gift of A. Charcosset, INRA, Moulon) for back crosses of the mutant collection and A188 (Gerdes and Tracy, 1993) for molecular work. Mutants sh1 (stock 912A), sh2-N391B (stock 318E), bt1-N1992 (stock 515E), bt2 (stock 414A), su1 (stock 407D), and su2 (stock 609D) were provided from the Maize Genetics Cooperation Stock Center (University of Illinois). Plants were grown in open air in Lyon (France) or in a greenhouse with a 16-h illumination period at 24°C/19°C (day/night) without control of the relative humidity.

Allelism Tests

Homozygous mutant plants from the Maize Stock Center (bt1, bt2, sh1, sh2, su1, su2) and from our collection (E2416, H0182, H0816, H2328) were grown in the greenhouse. At least two plants of each mutant were crossed with the mutant H2328. In control experiments, at least one plant of each mutant was self-pollinated and at least one plant of each mutant was crossed with inbred line F252.

Amplification of Mu-Flanking Sequences

The Bt2 gene of family H2328 was scanned for Mu insertions by PCR reactions with primers Bt2-353U (5'-AATGGTTTGCTATGGCTTCACTC-3') and Bt2-1692 L (5'-GTCGGGGTCCAGGCAGGTCTG-3') for segment a, Bt2-1609U (5'-GCGGAGTGTCCATCTCGAT-3') and Bt2-2770 L (5'-TGAGGAAATCAGAAAGACACAACAG-3') for segment b, Bt2-2634U (5'-CTCTGCTTCCCTCAACCGTCAC-3') and Bt2-3481 L (5'-TCCTTGCTGCCGTCCCCTTGG-3') for segment c, and Bt2-3503U (5'-CGCAAGTCAAAGGATAAAAAGATT-3') and Bt2-6404 L (5'-ATGAGAAATGCCGCTGCCATAGAA-3') for segment d, either as couples or in conjunction with primer AIMS2-Mu2 (5'-GCBCTCTTCGTCYATAATGGCAATTATCTC-3').

Cosegregation Analysis

One hundred twenty normal looking kernels (wild type or heterozygous) from one to four ears segregating for the mutation under investigation were sown in the field. After germination, leaf samples were taken from each plant for gDNA extraction and genotyping. The following primers were used either as a couple (band in wild-type and heterozygous plants) or in combination with the Mu-specific primer AIMS2Mu2 (band only in heterozygous plants): H6P-L634cos3 (5'-TGGGCCAAAGAGCAAGTCTG-3') and H6P-L634cos4 (5'-GTTCAGAAATGGAAGGGCACTG-3') for the insertion in H6P, H6P'-L635cos1 (5'-TGCCGCATATCGATCAGATTC-3') and H6P'-L635cos2 (5'-TATTCGACTGTATCCGTTCCGTT-3') for the insertion in H6P', and Bt2S5 (5'-CAATACCAGATTTCAGGTATGCTTTC-3') and Bt2S4 (5'-GCTTACTTTAATAACACATCCTTCACCA-3') for the insertion in Bt2. At flowering, all the plants were selfed for phenotyping. Fully wild-type ear and segregating mutant ear were indicative of a wild-type plant and a heterozygous plant, respectively.

Cytological Analysis

Kernels were harvested at 30 DAP and cut into three pieces of equal width along the longitudinal axis. The central slice was fixed in paraformaldehyde, dehydrated, embedded in paraffin, sectioned, and stained as described previously (Opsahl-Ferstad et al., 1997). The periodic acid-Schiff procedure stains insoluble carbohydrates, including starch and cellulose, in pink (Vozzo and Young, 1975).

Genetic Mapping

The H6P locus was mapped by PCR with primers H6P-L634cos3 and H6P-L634cos4 (see section on cosegregation analysis) in the intermated B73/Mo17 mapping population (Lee et al., 2002) exploiting a size polymorphism between the parental lines B73 and Mo17.

Agps Reference Sequences

The analysis of genomic sequences was based on a B73 bacterial artificial chromosome (BAC) sequence for Bt2 (AC193357), which fully included the previously published B73 sequence (AF334959), and B73 BAC sequences for Agpsemzm (AC177860) and Agpslzm (AC209218). A 46-bp 5'-UTR was defined for Bt2 by similarity with EST CO455578. The full-length cDNA sequences EC896031 and DR788076 as well as EST contigs BT016913 and AY105915 were the basis for a 140-bp 5'-UTR and a 245-bp 3'-UTR of Agpsemzm. A 130-bp 5'-UTR and a 233-bp 3'-UTR were defined for Agpslzm based on full-length cDNA sequences DR969809 and DR969810. The flanking sequence tags of the linked insertions H6P and H6P' were given the accession numbers EU137670 and EU137671, respectively.

RT-PCR

Total RNA was isolated using the following procedure. Tissues were ground to powder under liquid nitrogen and transferred to a tube containing equal volumes of extraction buffer (200 mM Tris-HCl, pH 9, 400 mM KCl, 200 mM Suc, 35 mM MgCl₂, 25 mM EGTA) and phenol/chloroform (pH 8) and vortexed for 30 s. The aqueous phase resulting from a 5-min centrifugation at 18,000g was reextracted twice with phenol/chloroform. RNA was precipitated by addition of 1 M acetic acid (1/10 volume) and ethanol (2.5 volumes). The RNA pellet was washed with 3 M sodium acetate (pH 6) and resuspended in water. A second acetic acid/ethanol precipitation was performed before final resuspension in water.

RNA was treated with RNase-free DNase and the DNase inactivated according to the instructions of the supplier (Ambion). The RNA was quantified in a spectrophotometer at 260 nm. Approximately 5 µg of total RNA were reverse transcribed using random hexamers (Amersham Pharmacia Biotech) and reverse transcriptase without RNaseH activity (Fermentas) in a final volume of 20 µL. Then 2.5 x 10⁵ copies of GeneAmplimer pAW109 RNA (Applied Biosystems) were added to the RT reaction. The cDNA was diluted 50 times and 5 µL used for amplification by PCR in a volume of 20 µL.

The following primers were used for gene-specific amplification: Bt2-cDNA-F2 (5'-GCTGATAAAAAACTCCTTGCCGAAAAA-3') and Bt2-cDNA-R3 (5'-CACAGCTGCATGTCGCACGTTCA-3') for Bt2, leaf-cDNA-F (5'-GCTGACAAGAAACTCCTTGCCGAAAAT-3') and leaf-cDNA-R2 (5'-CTGGCTTGCATGTCGCACATCCG-3') for Agpslzm, and BEF2 (5'-TATCCACTGACGAAGAAGAGGGCG-3') and BER2 (5'-AAGGACATTCTTATCATTCTCA-3') for Agpsemzm. The specificity of the primers was shown by cloning PCR products in tissues with low relative expression (Bt2 in leaf, Agpslzm in 35-DAP kernel, Agpsezm in 35-DAP kernel) and sequencing of 10 clones per gene. In all three cases, all 10 clones were identical to the reference sequences AF334959 (Bt2), AF334960 (Agpslzm), or AC177860 (Agpsezm). The alternatively spliced 5' ends of Bt2 were detected with primers Bt2-F21 (5'-GCCGCTGCAAATGATTCAACATACC-3') for Bt2a and primer Bt2-F11 (5'-ATGGCCGCGATAGCCTCAGCTT-3') for Bt2b, both in conjunction with the same reverse primer Bt2-R11 (5'-GCAGGCTTGGCACGCTTCTTTG-3'). The primer Bt2-F11 had been designed to hybridize only to Bt2b and not to Agpslzm, and this specificity was demonstrated by sequencing 10 cloned PCR products obtained with the combination Bt2-F11/Bt2-R11 in both 20-DAP kernels and leaves. All 20 sequences corresponded to the Bt2b reference sequence.

Primers AIMS2-Mu2 (see above) and Bt2-cDNA-R3 (see above) allowed amplification of the bt2-H2328 allele downstream of the Mu insertion. Wild-type, +/bt2-H2328, and bt2-H2328/bt2-H2328 plants were identified both by PCR genotyping and by self-pollination followed by phenotypic analysis of the ears. Genomic contamination was shown to be negligible by the use of intron-spanning control primers ART1 (5'-GTCAAGTTCTGGTTCCAGAACCG-3') and ART4 (5'-CCGTGCCCAACGGGCTAGACACA-3') amplifying ZmOCL1 (Khaled et al., 2005).

The constitutively expressed 18S rRNA gene (primers 5'-CCATCCCTCCGTAGTTAGCTTCT-3' and 5'-CCTGTCGGCCAAGGCTATATAC-3') was used as an internal control of RNA quantity and GeneAmplimer pAW109 RNA (primers 5'-CATGTCAAATTTCACTGCTTCATC-3' and 5'-TGACCACCCAGCCATCCTT-3') as positive control of the RT-PCR efficiency. To get semiquantitative results, the number of cycles of the PCR reactions was adjusted for each gene to obtain barely visible bands in agarose gels. Aliquots of the PCR reaction were loaded on agarose gels and stained with ethidium bromide.

Transcriptome Analysis

Whole kernels (pericarp + endosperm + embryo) without glumes from greenhouse-grown ears of heterozygous, self-pollinated plants were collected at 35 DAP and visually divided into pools of phenotypically normal-looking kernels (genotype +/+ or +/bt2-H2328) and pools of phenotypically mutant kernels (genotype bt2-H2328/bt2-H2328). Two different ears were used for a biological duplicate. Total RNA was extracted using the following procedure. Samples were ground to powder under liquid nitrogen and transferred to a tube containing 4.5 mL of extraction buffer (0.1 M NaCl, 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, pH 8, 1% SDS) and 3 mL phenol/chloroform (pH 8), and agitated for 10 min. The aqueous phase resulting from a 20-min centrifugation at 5,000g was reextracted with phenol/chloroform. RNA was precipitated by addition of 3 M sodium acetate, pH 5.2 (1/10 volume), and ethanol (2.5 volumes). The RNA pellet was resuspended in water. Contaminating DNA was removed using a DNase set (Qiagen) and RNA purified using the Min Elute kit (Qiagen). All RNA samples were checked for their integrity on the Agilent 2100 bioanalyzer according to the protocols of Agilent Technologies.

Two micrograms of total RNA were used to synthesize biotin-labeled cRNAs with the one-cycle cDNA synthesis kit (Affymetrix). SuperScript II reverse transcriptase and T7-oligo(dT) primers were used to synthesize single-stranded cDNA at 42°C during 1 h, followed by synthesis of double-stranded cDNA using DNA ligase, DNA polymeraseI, and RNaseH during 2 h at 16°C. After cleanup of the double-stranded cDNA with the Sample Cleanup Module (Affymetrix), in vitro transcription was performed in the presence of biotin-labeled UTP using the GeneChip IVT labeling kit (Affymetrix). The labeled cRNA was purified with the Sample Cleanup Module (Affymetrix) and quantified with RiboGreen RNA quantification reagent (Turner Biosystems). Fragmentation of 15 µg of labeled cRNA was carried out for 35 min at 94°C, followed by hybridization during 16 h at 45°C to Affymetrix GeneChip Maize Genome Array representing approximately 14,850 maize transcripts, corresponding to 13,339 genes. After hybridization, the arrays were washed with two different buffers (stringent: 6x SSPE, 0.01% Tween 20; and nonstringent: 100 mM MES, 0.1 M Na⁺, 0.01% Tween 20) and stained with a complex solution including Streptavidin R-Phycoerythrin conjugate (Molecular Probes) and anti-Streptavidin biotinylated antibody (Vectors Laboratories). The washing and staining steps were performed in a GeneChip Fluidics Station 450 (Affymetrix). The Affymetrix GeneChip maize genome arrays were finally scanned with the GeneChip Scanner 3000 7G piloted by GeneChip Operating Software. All this steps were performed on Affymetrix platform at INRA-URGV in Evry.

The raw CEL files were imported in the Bioconductor software package in R for data analysis (Gentleman et al., 2004). The data were normalized with the gcrma algorithm (Irizarry et al., 2003) available in the Bioconductor package. To determine differentially expressed genes, we performed a usual two-group t test that assumes equal variance between groups. The variance of the gene expression per group is a homoscedastic variance, where genes displaying extremes of variance (too small or too large) were excluded. The raw P values were adjusted by the Bonferroni method, which controls the familywise error rate (Dudoit et al., 2003). A gene is declared differentially expressed if the Bonferroni P value <0.05. All raw and normalized data are available from the Gene Expression Omnibus repository at the National Center for Biotechnology Information (Barrett et al., 2007) under accession number GSE7030.

Overrepresentation of GO terms among differentially expressed genes was assessed with the GO Browser tool of the Functional Genomics suite of the Spotfire Decision Site program (Spotfire). GO annotations of the probe sets were downloaded from the NetAffx Web site (http://www.affymetrix.com; Liu et al., 2003).

¹H NMR Spectroscopy

Homozygous mutant bt2-H2328/bt2-H2328 plants were pollinated by heterozygous +/bt2-H232 plants and kernels without glumes were collected at 35 DAP and visually divided into pools of phenotypically wild-type kernels (genotype +/bt2-H2328) and phenotypically mutant kernels (genotype bt2-H2328/bt2-H2328). Three different ears were used for a biological triplicate. In addition to pools of 10 whole kernels (pericarp, endosperm, and embryo), pools of 15 embryos and 15 endosperm + pericarp were established for wild-type and mutant kernels of each of the three ears. Microdissected embryos and remaining endosperm + pericarp were from the same kernels. From each sample pool, two extractions were made in parallel to obtain a technical replicate.

Kernel or kernel parts were frozen in liquid nitrogen, ground with a mixer mill MM300 (Qiagen), and lyophilized for 5 d. Soluble metabolites were extracted with a series of hot ethanol extractions (80%–20%, 50%–50%, 100%–0%, ethanol-water [v/v]) according to Stitt and Ap Rees (1978), with slight modifications described previously (Moing et al., 2004), dried under vacuum, and lyophilized. The dried extracts were resuspended in 100 mM phosphate buffer, pH 6.0, in D₂O and analyzed at 500.162 MHz on a Bruker spectrometer (Bruker BiospinAvance). Special care was taken to allow absolute quantification of the individual metabolites through addition of EDTA sodium salt solution (1 mM final concentration in the NMR tube) to improve the resolution and quantification of organic acids, such as malate and citrate, adequate choice of the NMR acquisition parameters (pulse angle 90°, relaxation delay 15 s), and use of an electronic reference (ERETIC mode; Akoka et al., 1999) calibrated with Suc and Glu sodium salt solutions as described previously (Moing et al., 2004). ¹H NMR profiles were analyzed with AMIX-VIEWER and AMIX Statistics v.3.6.8 software (Bruker Biospin). PCA of the 36 respective samples was used on the entire NMR signatures after data reduction to spectral domains of 0.04 ppm with exclusion of the water and ethanol regions and normalization by dividing by the sum of intensities of all domains. In addition, 18 to 21 individual metabolites, depending on the tissue, were identified using published data (Moing et al., 2004) and acquisition of NMR spectra of reference compounds in the same conditions of solvent. They were quantified using the metabolite mode of AMIX software (Bruker v.3.7.2) and the number of protons of the corresponding resonance to calculate the concentration in the NMR tube, and then the concentration in the sample using the sample dry weight. Glu, citrate, and malate are expressed as g of the acid form. The concentration of NMR unknown compounds (named according to the form of the resonance, S for singlet, M for multiplet, and its frequency in ppm) was calculated on the assumption that the measured resonance corresponded to one proton and using an arbitrary molecular mass of 100 D. Then t tests were performed to pinpoint statistically significant differences between samples.

Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers EU137670 and EU137671.

Supplemental Data

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

Supplemental Figure S1. Representative quantitative ¹H NMR (500 MHz) profiles of extracts of wild-type embryo (A), endosperm (B), or whole kernel (C).

Supplemental Figure S2. Comparison of representative wild-type (A) and bt2-H2328 mutant (B) ¹H NMR (500 MHz) profiles of embryo extract.

Supplemental Figure S3. Comparison of representative wild-type (A) and bt2-H2328 mutant (B) ¹H NMR (500 MHz) profiles of endosperm extract.

Supplemental Figure S4. PCA of ¹H NMR signatures of endosperm extracts of wild type and bt2-H2328 mutant.

Supplemental Figure S5. PCA of ¹H NMR signatures of embryo extract of wild type and bt2-H2328 mutant.

Supplemental Table S1. Metabolite levels in wild-type and bt2-H2328 kernels at 35 DAP.

Supplemental Table S2. Metabolite levels in wild-type and bt2-H2328 endosperms at 35 DAP.

Supplemental Table S3. Metabolite levels in wild-type and bt2-H2328 embryos at 35 DAP.

	ACKNOWLEDGMENTS

 
We thank Laurent Decousset for mapping of the H6P locus, Daniel Pissaloux for the annotation pipeline, Domenica Manicacci for the design of genomic Bt2 primers, Hervanne Cassagnet and Géraldine Brunoud for help with the cosegregation analysis of H2328, and Frédérique Rozier for technical assistance in molecular biology. Isabelle Desbouchages, Alexis Lacroix, and Armand Guillermin are acknowledged for plant culture, and Claudia Bardoux and Hervé Leyral for media and buffer preparation.

Received November 6, 2007; accepted February 15, 2008; published February 20, 2008.

	FOOTNOTES

 
¹ This work was supported by the Association Nationale de la Recherche Technique (Ph.D. grant from Conventions Industrielles de Formation par la Recherche). The work was carried out in the framework of the project PMG² financed by Biogemma SAS.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Peter Rogowsky (peter.rogowsky{at}ens-lyon.fr).

^[W] The online version of this article contains Web-only data.

www.plantphysiol.org/cgi/doi/10.1104/pp.107.112698

^* Corresponding author; e-mail peter.rogowsky{at}ens-lyon.fr.

	LITERATURE CITED

TOP ABSTRACT RESULTS DISCUSSION CONCLUSION MATERIALS AND METHODS LITERATURE CITED

 
Akoka S, Barantin L, Trierweiler M (1999) Concentration measurement by proton NMR using the ERETIC method. Anal Chem 71: 2554–2557

Baena-Gonzalez E, Rolland F, Thevelein JM, Sheen J (2007) A central integrator of transcription networks in plant stress and energy signalling. Nature 448: 938–942[CrossRef][Medline]

Ballicora MA, Laughlin MJ, Fu Y, Okita TW, Barry GF, Preiss J (1995) Adenosine 5'-diphosphate-glucose pyrophosphorylase from potato tuber. Significance of the N terminus of the small subunit for catalytic properties and heat stability. Plant Physiol 109: 245–251[Abstract]

Barrett T, Troup DB, Wilhite SE, Ledoux P, Rudnev D, Evangelista C, Kim IF, Soboleva A, Tomashevsky M, Edgar R (2007) NCBI GEO: mining tens of millions of expression profiles—database and tools update. Nucleic Acids Res 35: D760–765[Abstract/Free Full Text]

Beckles DM, Smith AM, ap Rees T (2001) A cytosolic ADP-glucose pyrophosphorylase is a feature of graminaceous endosperms, but not of other starch-storing organs. Plant Physiol 125: 818–827[Abstract/Free Full Text]

Bhave MR, Lawrence S, Barton C, Hannah LC (1990) Identification and molecular characterization of shrunken-2 cDNA clones of maize. Plant Cell 2: 581–588[Abstract/Free Full Text]

Borisjuk L, Rolletschek H, Radchuk R, Weschke W, Wobus U, Weber H (2004) Seed development and differentiation: a role for metabolic regulation. Plant Biol 6: 375–386[CrossRef][Medline]

Bouche N, Fromm H (2004) GABA in plants: just a metabolite? Trends Plant Sci 9: 110–115[CrossRef][Web of Science][Medline]

Burger BT, Cross JM, Shaw JR, Caren JR, Greene TW, Okita TW, Hannah LC (2003) Relative turnover numbers of maize endosperm and potato tuber ADP-glucose pyrophosphorylases in the absence and presence of 3-phosphoglyceric acid. Planta 217: 449–456[CrossRef][Web of Science][Medline]

Burton RA, Johnson PE, Beckles DM, Fincher GB, Jenner HL, Naldrett MJ, Denyer K (2002) Characterization of the genes encoding the cytosolic and plastidial forms of ADP-glucose pyrophosphorylase in wheat endosperm. Plant Physiol 130: 1464–1475[Abstract/Free Full Text]

Cameron JW, Teas HJ (1954) Carbohydrate relationships in developing and mature endosperms of brittle and related maize genotypes. Am J Bot 41: 50–55[CrossRef][Web of Science]

Carlson SJ, Chourey PS, Helentjaris T, Datta R (2002) Gene expression studies on developing kernels of maize sucrose synthase (SuSy) mutants show evidence for a third SuSy gene. Plant Mol Biol 49: 15–29[CrossRef][Web of Science][Medline]

Choi SB, Kim KH, Kavakli IH, Lee SK, Okita TW (2001) Transcriptional expression characteristics and subcellular localization of ADP-glucose pyrophosphorylase in the oil plant Perilla frutescens. Plant Cell Physiol 42: 146–153[Abstract/Free Full Text]

Chourey PS, Jain M, Li QB, Carlson SJ (2006) Genetic control of cell wall invertases in developing endosperm of maize. Planta 223: 159–167[CrossRef][Web of Science][Medline]

Clark JK, Sheridan WF (1991) Isolation and characterization of 51 embryo-specific mutations of maize. Plant Cell 3: 935–951[Abstract/Free Full Text]

Crevillen P, Ballicora MA, Merida A, Preiss J, Romero JM (2003) The different large subunit isoforms of Arabidopsis thaliana ADP-glucose pyrophosphorylase confer distinct kinetic and regulatory properties to the heterotetrameric enzyme. J Biol Chem 278: 28508–28515[Abstract/Free Full Text]

da Costa e Silva O, Lorbiecke R, Garg P, Muller L, Wassmann M, Lauert P, Scanlon M, Hsia AP, Schnable PS, Krupinska K, et al (2004) The Etched1 gene of Zea mays (L.) encodes a zinc ribbon protein that belongs to the transcriptionally active chromosome (TAC) of plastids and is similar to the transcription factor TFIIS. Plant J 38: 923–939[CrossRef][Web of Science][Medline]

Denyer K, Dunlap F, Thorbjornsen T, Keeling P, Smith AM (1996) The major form of ADP-glucose pyrophosphorylase in maize endosperm is extra-plastidial. Plant Physiol 112: 779–785[Abstract]

Dickinson DB, Preiss J (1969) Presence of ADP-glucose pyrophosphorylase in shrunken-2 and brittle-2 mutants of maize endosperm. Plant Physiol 44: 1058–1062[Abstract/Free Full Text]

Doehlert DC, Kuo TM (1994) Gene-expression in developing kernels of some endosperm mutants of maize. Plant Cell Physiol 35: 411–418[Abstract/Free Full Text]

Dudoit S, Shaffer JP, Boldrick JC (2003) Multiple hypothesis testing in microarray experiments. Stat Sci 18: 71–103[CrossRef][Web of Science]

Duvick DN, Cassman KG (1999) Post-green revolution trends in yield potential of temperate maize in the north-central United States. Crop Sci 39: 1622–1630[Abstract/Free Full Text]

Echeverria E, Boyer CD, Thomas PA, Liu KC, Shannon JC (1988) Enzyme activities associated with maize kernel amyloplasts. Plant Physiol 86: 786–792[Abstract/Free Full Text]

Fu S, Meeley R, Scanlon MJ (2002) Empty pericarp2 encodes a negative regulator of the heat shock response and is required for maize embryogenesis. Plant Cell 14: 3119–3132[Abstract/Free Full Text]

Geigenberger P, Muller-Rober B, Stitt M (1999) Contribution of adenosine 5'-diphosphoglucose pyrophosphorylase to the control of starch synthesis is decreased by water stress in growing potato tubers. Planta 209: 338–345[CrossRef][Web of Science][Medline]

Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry J, et al (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5: R80[CrossRef][Medline]

Gerdes JT, Tracy WF (1993) Pedigree diversity within the Lancaster surecrop heterotic group of maize. Crop Sci 33: 334–337[Abstract/Free Full Text]

Giroux MJ, Boyer C, Feix G, Hannah LC (1994) Coordinated transcriptional regulation of storage product genes in the maize endosperm. Plant Physiol 106: 713–722[Abstract]

Giroux MJ, Hannah LC (1994) ADP-glucose pyrophosphorylase in shrunken-2 and brittle-2 mutants of maize. Mol Gen Genet 243: 400–408[Web of Science][Medline]

Gutierrez-Marcos JF, Dal Pra M, Giulini A, Costa LM, Gavazzi G, Cordelier S, Sellam O, Tatout C, Paul W, Perez P, et al (2007) empty pericarp4 encodes a mitochondrion-targeted pentatricopeptide repeat protein necessary for seed development and plant growth in maize. Plant Cell 19: 196–210[Abstract/Free Full Text]

Hannah LC (2005) Starch synthesis in the maize endosperm. Maydica 50: 497–506

Hannah LC, Shaw JR, Giroux MJ, Reyss A, Prioul JL, Bae JM, Lee JY (2001) Maize genes encoding the small subunit of ADP-glucose pyrophosphorylase. Plant Physiol 127: 173–183[Abstract/Free Full Text]

Im KH (2004) Expression of sucrose-phosphate synthase (SPS) in non-photosynthetic tissues of maize. Mol Cells 17: 404–409[Web of Science][Medline]

Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, Speed TP (2003) Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249–264[Abstract]

Khaled AS, Vernoud V, Ingram GC, Perez P, Sarda X, Rogowsky PM (2005) Engrailed-ZmOCL1 fusions cause a transient reduction of kernel size in maize. Plant Mol Biol 58: 123–139[CrossRef][Web of Science][Medline]

Kim JY, Mahe A, Guy S, Brangeon J, Roche O, Chourey PS, Prioul JL (2000) Characterization of two members of the maize gene family, Incw3 and Incw4, encoding cell-wall invertases. Gene 245: 89–102[CrossRef][Web of Science][Medline]

Koch K (2004) Sucrose metabolism: regulatory mechanisms and pivotal roles in sugar sensing and plant development. Curr Opin Plant Biol 7: 235–246[CrossRef][Web of Science][Medline]

Laughnan JR (1953) The effect of the sh2 factor on carbohydrate reserves in the mature endosperm of maize. Genetics 38: 485–499[Free Full Text]

Lee M, Sharopova N, Beavis WD, Grant D, Katt M, Blair D, Hallauer A (2002) Expanding the genetic map of maize with the intermated B73 x Mo17 (IBM) population. Plant Mol Biol 48: 453–461[CrossRef][Web of Science][Medline]

Lee SK, Hwang SK, Han M, Eom JS, Kang HG, Han Y, Choi SB, Cho MH, Bhoo SH, An G, et al (2007) Identification of the ADP-glucose pyrophosphorylase isoforms essential for starch synthesis in the leaf and seed endosperm of rice (Oryza sativa L.). Plant Mol Biol 65: 531–546[Medline]

Lid SE, Gruis D, Jung R, Lorentzen JA, Ananiev E, Chamberlin M, Niu XM, Meeley R, Nichols S, Olsen OA (2002) The defective kernel 1 (dek1) gene required for aleurone cell development in the endosperm of maize grains encodes a membrane protein of the calpain gene superfamily. Proc Natl Acad Sci USA 99: 5460–5465[Abstract/Free Full Text]

Liu G, Loraine AE, Shigeta R, Cline M, Cheng J, Valmeekam V, Sun S, Kulp D, Siani-Rose MA (2003) NetAffx: Affymetrix probesets and annotations. Nucleic Acids Res 31: 82–86[Abstract/Free Full Text]

Lutfiyya LL, Xu N, D'Ordine RL, Morrell JA, Miller PW, Duff SM (2007) Phylogenetic and expression analysis of sucrose phosphate synthase isozymes in plants. J Plant Physiol 164: 923–933[CrossRef][Web of Science][Medline]

Ma ZR, Dooner HK (2004) A mutation in the nuclear-encoded plastid ribosomal protein S9 leads to early embryo lethality in maize. Plant J 37: 92–103[CrossRef][Web of Science][Medline]

Magnard JL, Heckel T, Massonneau A, Wisniewski JP, Cordelier S, Lassagne H, Perez P, Dumas C, Rogowsky PM (2004) Morphogenesis of maize embryos requires ZmPRPL35-1 encoding a plastid ribosomal protein. Plant Physiol 134: 649–663[Abstract/Free Full Text]

Mains EB (1949) Heritable characters in maize—linkage of a factor for shrunken endosperm with the A1 factor for aleurone color. J Hered 40: 21–24[Free Full Text]

Martin A, Lee J, Kichey T, Gerentes D, Zivy M, Tatout C, Dubois F, Balliau T, Valot B, Davanture M, et al (2006) Two cytosolic glutamine synthetase isoforms of maize are specifically involved in the control of grain production. Plant Cell 18: 3252–3274[Abstract/Free Full Text]

McElver J, Tzafrir I, Aux G, Rogers R, Ashby C, Smith K, Thomas C, Schetter A, Zhou Q, Cushman MA, et al (2001) Insertional mutagenesis of genes required for seed development in Arabidopsis thaliana. Genetics 159: 1751–1763[Abstract/Free Full Text]

Moing A, Maucourt M, Renaud C, Gaudillere M, Brouquisse R, Lebouteiller B, Gousset-Dupont A, Vidal J, Granot D, Denoyes-Rothan B, et al (2004) Quantitative metabolic profiling by 1-dimensional H-1-NMR analyses: application to plant genetics and functional genomics. Funct Plant Biol 31: 889–902[CrossRef]

Neuffer MG, Sheridan WF (1980) Defective kernel mutants of maize: I. Genetic and lethality studies. Genetics 95: 929–944[Abstract/Free Full Text]

Ohdan T, Francisco PB Jr, Sawada T, Hirose T, Terao T, Satoh H, Nakamura Y (2005) Expression profiling of genes involved in starch synthesis in sink and source organs of rice. J Exp Bot 56: 3229–3244[Abstract/Free Full Text]

Opsahl-Ferstad HG, Le Deunff E, Dumas C, Rogowsky PM (1997) ZmEsr, a novel endosperm-specific gene expressed in a restricted region around the maize embryo. Plant J 12: 235–246[CrossRef][Web of Science][Medline]

Preiss J, Lammel C, Sabraw A (1971) A unique adenosine diphosphoglucose pyrophosphorylase associated with maize embryo tissue. Plant Physiol 47: 104–108[Abstract/Free Full Text]

Prioul JL, Jeannette E, Reyss A, Gregory N, Giroux M, Hannah LC, Causse M (1994) Expression of ADP-glucose pyrophosphorylase in maize (Zea mays L.) grain and source leaf during grain filling. Plant Physiol 104: 179–187[Abstract]

Rhodes D, Hanson AD (1993) Quaternary ammonium and tertiary sulfonium compounds in higher-plants. Annu Rev Plant Physiol 44: 357–384[CrossRef][Web of Science]

Rosti S, Denyer K (2007) Two paralogous genes encoding small subunits of ADP-glucose pyrophosphorylase in maize, Bt2 and L2, replace the single alternatively spliced gene found in other cereal species. J Mol Evol 65: 316–327[CrossRef][Web of Science][Medline]

Russell DA, DeBoer DL, Stark DM, Preiss J, Fromm ME (1993) Plastid targeting of β-glucuronidase and ADP-glucose pyrophosphorylase in maize (Zea mays L.) cells. Plant Cell Rep 13: 287–292

Scanlon MJ, Stinard PS, James MG, Myers AM, Robertson DS (1994) Genetic analysis of 63 mutations affecting maize kernel development isolated from Mutator stocks. Genetics 136: 281–294[Abstract]

Smith-White BJ, Preiss J (1992) Comparison of proteins of ADP-glucose pyrophosphorylase from diverse sources. J Mol Evol 34: 449–464[CrossRef][Web of Science][Medline]

Spielbauer G, Margl L, Hannah LC, Romisch W, Ettenhuber C, Bacher A, Gierl A, Eisenreich W, Genschel U (2006) Robustness of central carbohydrate metabolism in developing maize kernels. Phytochemistry 67: 1460–1475[CrossRef][Web of Science][Medline]

Stiefel V, Becerra EL, Roca R, Bastida M, Jahrmann T, Graziano E, Puigdomenech P (1999) TM20, a gene coding for a new class of transmembrane proteins expressed in the meristematic tissues of maize. J Biol Chem 274: 27734–27739[Abstract/Free Full Text]

Stitt M, ap Rees T (1978) Pathways of carbohydrate oxidation in leaves of Pisum sativum and Triticum aestivum. Phytochemistry 17: 1251–1256[CrossRef][Web of Science]

Sun C, Palmqvist S, Olsson H, Boren M, Ahlandsberg S, Jansson C (2003) A novel WRKY transcription factor, SUSIBA2, participates in sugar signaling in barley by binding to the sugar-responsive elements of the iso1 promoter. Plant Cell 15: 2076–2092[Abstract/Free Full Text]

Thorbjornsen T, Villand P, Denyer K, Olsen OA, Smith AM (1996a) Distinct isoforms of ADPglucose pyrophosphorylase occur inside and outside the amyloplasts in barley endosperm. Plant J 10: 243–250[CrossRef][Web of Science]

Thorbjornsen T, Villand P, Kleczkowski LA, Olsen OA (1996b) A single gene encodes two different transcripts for the ADP-glucose pyrophosphorylase small subunit from barley (Hordeum vulgare). Biochem J 313: 149–154[Web of Science][Medline]

Tobias RB, Boyer CD, Shannon JC (1992) Alterations in carbohydrate intermediates in the endosperm of starch-deficient maize (Zea mays L.) genotypes. Plant Physiol 99: 146–152[Abstract/Free Full Text]

Tsai CY, Larkins BA, Glover DV (1978) Interaction of the opaque-2 gene with starch-forming mutant genes on the synthesis of zein in maize endosperm. Biochem Genet 16: 883–896[CrossRef][Web of Science][Medline]

Tsai CY, Nelson OE (1966) Starch-deficient maize mutant lacking adenosine diphosphate glucose pyrophosphorylase activity. Science 151: 341–343[Abstract/Free Full Text]

Vozzo JA, Young RW (1975) Carbohydrate, lipid and protein distribution in dormant, stratified, and germinated Quercus nigra embryos. Bot Gaz 136: 306–311

Woo YM, Hu DW, Larkins BA, Jung R (2001) Genomics analysis of genes expressed in maize endosperm identifies novel seed proteins and clarifies patterns of zein gene expression. Plant Cell 13: 2297–2317[Abstract/Free Full Text]

This article has been cited by other articles:

				K. Weigelt, H. Kuster, T. Rutten, A. Fait, A. R. Fernie, O. Miersch, C. Wasternack, R. J. N. Emery, C. Desel, F. Hosein, et al. ADP-Glucose Pyrophosphorylase-Deficient Pea Embryos Reveal Specific Transcriptional and Metabolic Changes of Carbon-Nitrogen Metabolism and Stress Responses Plant Physiology, January 1, 2009; 149(1): 395 - 411. [Abstract] [Full Text] [PDF]

				T. L. Slewinski, Y. Ma, R. F. Baker, M. Huang, R. Meeley, and D. M. Braun Determining the Role of Tie-dyed1 in Starch Metabolism: Epistasis Analysis with a Maize ADP-Glucose Pyrophosphorylase Mutant Lacking Leaf Starch J. Hered., November 1, 2008; 99(6): 661 - 666. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 146/4/1553    most recent pp.107.112698v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Cossegal, M. Articles by Rogowsky, P. Search for Related Content PubMed PubMed Citation Articles by Cossegal, M. Articles by Rogowsky, P. Agricola Articles by Cossegal, M. Articles by Rogowsky, P.