INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Phytic acid, myo-inositol 1,2,3,4,5,6-hexakisphosphate, is an abundant component of plant seeds and is deposited in protein bodies as a mixed salt of mineral cations, such as K⁺, Mg²⁺, Ca²⁺, Zn²⁺, and Fe³⁺. Typically, 50% to 80% of the phosphorus in seeds is found in this compound. Phytic acid serves as a major storage form for myo-inositol, phosphorus, and mineral cations for use during seedling growth. The other known role of phytic acid is the control of inorganic phosphate (Pi) levels in both developing seeds and seedlings (Strother, 1980). In maize (Zea mays) kernels, nearly 90% of the phytic acid is accumulated in embryo and about 10% in aleurone layers. Maize endosperm contains only trace amount of phytic acid (O'Dell et al., 1972). In rice (Oryza sativa), barley (Hordeum vulgare), and wheat (Triticum aestivum), most of the phytic acid (approximately 90%) is found in the aleurone layers and only about 10% in embryo. Reduced phytic acid content in seeds is a desired goal for genetic improvement in several crops, including maize, rice, barley, wheat, and soybean (Glycine max). Because monogastric animals digest phytic acid poorly, animal feed is supplemented with Pi to meet the phosphorus requirement for animal growth. Undigested phytic acid is eliminated and is a leading phosphorus pollution source (Cromwell and Coffey, 1991). Although phytic acid as an antioxidant is suggested to have potential functions of reducing lipid peroxidation and some protective effects, phytic acid is considered to be an antinutritional substance in animal feed and human diets because it binds mineral cations and reduces their bioavailability (Zhou and Erdman, 1995). Low-phytic acid grain and legume in feed could reduce phosphorus pollution to environment and reduce amount of phosphorus supplementation required in animal feeds (Ertl et al., 1998). Such grain would also offer more available Fe and Zn for human nutrition (Mendoza et al., 1998).

Low-phytic acid mutants have been generated by mutagenesis in maize, rice, barley, and soybean (Rasmussen and Hatzack, 1998; Larson et al., 2000; Raboy et al., 2000; Wilcox et al., 2000) and used in genetic breeding (Raboy et al., 2001). Two types of recessive mutants affecting seed phytic acid are known. One mutant type, exemplified by maize lpa1, has low-phytic acid phenotype, but does not accumulate inositol polyphosphates. A second type, such as maize lpa2, also shows reduced phytic acid contents, but these seeds accumulate InsP₃, InsP₄, and InsP₅ (Raboy et al., 2000). Which genes are responsible for the low-phytic acid phenotype is not known.

In developing seeds, phytic acid is synthesized from Glc 6-P (Loewus and Loewus, 1983). 1D-myo-inositol 3-phosphate synthase (also known as 1L-myo-inositol 1-phosphate synthase) catalyzes the cyclization of Glc 6-P to produce 1D-myo-inositol 3-phosphate, Ins(3)P. A rice Ins(3)P synthase gene was cloned, and its mRNA is highly expressed in the aleurone layers, suggesting its role in phytic acid biosynthesis in developing rice seeds (Yoshida et al., 1999). Seeds also have a myo-inositol kinase activity, which catalyzes the Ins(3)P formation from myo-inositol and ATP (English et al., 1966). It has been proposed that a sequential ATP-dependent phosphorylation of Ins(3)P leads to the phytic acid production in developing seeds (Biswas et al., 1984).

Two inositol phosphate kinases were purified from germinating mung bean (Vigna radiata) seeds (Majumder et al., 1972; Majumder and Biswas, 1973; Chakrabarti and Biswas, 1981). The enzyme phosphorylated Ins(3)P to form InsP₅, but it could produce phytic acid when Ins(2)P was used as the substrate. Ins(1,3,4)P₃ 5-kinase, Ins(1,3,4,5)P₄ 6-kinase, and Ins(1,3,4,5,6)P₅ 2-kinase activities also were detected in immature soybean seeds (Phillippy et al., 1994; Phillippy, 1998). Using myo-inositol 2-P affinity chromatography, Bollmann at el. (1980) identified two inositol phosphate kinase activities in duckweed (Lemna gibba). One kinase catalyzed the phosphorylation of Ins(3)P to form inositol trisphosphate, and the other phosphorylated inositol trisphosphate to yield phytic acid. However, none of the genes encoding these enzymes has been cloned. In Arabidopsis, an Ins(1,3,4)P₃ 5/6-kinase gene was isolated (Wilson and Majerus, 1997). However, the role of this gene in phytic acid biosynthesis is not known.

To better understand the phytic acid biosynthesis in developing seeds, we have focused on identifying the genes involved in the pathway. We have isolated maize cDNA clones that show sequence similarity to various inositol phosphate kinase genes from animals and yeast. Using Mutator insertion knockout technology, we were able to investigate functions. Reported in this paper is a maize inositol phosphate kinase gene found to be involved in phytic acid biosynthesis in developing seeds. After knocking out this gene, the phytic acid content in kernels was reduced and seeds accumulated myo-inositol, inositol phosphates, and Pi. The maize inositol phosphate kinase (ZmIpk) loss-of-function mutants are allelic to the low-phytic acid mutant lpa2. Cloning and sequencing of the ZmIpk gene from lpa2-2 showed that the lpa2-2 allele has a nucleotide mutation that causes immature termination of the ZmIpk open reading frame. In the lpa2-1 mutant, the genomic sequence was found rearranged in the ZmIpk locus and no mRNA expression was detected.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Maize Inositol Phosphate Kinase Sequence

Using human and Arabidopsis Ins(1,3,4)P₃ 5/6-kinase genes, we identified several maize cDNA clones from the DuPont/Pioneer EST database. Amino acid sequences of these clones show similarity at various levels to Ins(1,3,4)P₃ 5/6-kinase gene. One clone has an insert of 1.4 kb and encodes a protein of 342 amino acid residues. Its amino acid sequence is 56.7% identical to Arabidopsis Ins(1,3,4)P₃ 5/6-kinase 1 (GenBank Accession No. JC5401) and 34.0% to human Ins(1,3,4)P₃ 5/6-kinase (GenBank accession no. U51336). The maize gene is referred to as ZmIpk. The sequence alignment of ZmIpk, Arabidopsis, and human Ins(1,3,4)P₃ 5/6-kinase is shown in Figure 1.

View larger version (84K): [in this window] [in a new window]   Figure 1.   Amino acid sequence alignment of the maize ZmIpk, Arabidopsis, and human Ins(1,3,4)P3 5/6-kinase. Identical amino acids are shaded in black. AtItpk, Arabidopsis Ins(1,3,4)P3 5/6-kinase; HsItpk, human Ins(1,3,4)P3 5/6-kinase; AtItpk-1, GenBank accession number JC5401; AtItpk-2, T10544; AtItpk-3, NP_195103; ZmIpk, AY172635; and HsItpk, U51336.

PCR primers were designed from 5'- and 3'-untranslated region of the cDNA, and the genomic clones were amplified from maize B73 line. It was found that the maize ZmIpk genomic sequence has no intron. The gene encoding Arabidopsis Ins(1,3,4)P₃ 5/6-kinase 1 also lacks introns (The Institute for Genomic Research Arabidopsis database no. At5g16760).

ZmIpk Has Multiple Inositol Phosphate Kinase Activities

Glutathione S-transferase (GST)-tagged ZmIpk protein was expressed in Escherichia coli and purified with glutathione-Sepharose beads. To assay inositol phosphate kinase activity, [-³²P]ATP was used. A kinase would phosphorylate inositol phosphate substrates and produce a ³²P-labeled inositol phosphate product. The radioactive inositol phosphate product can be separated from [-³²P]ATP on a polyethylenimine (PEI)-cellulose-coated thin-layer chromatography (TLC) plate. This assay cannot identify the stereospecific structure of the inositol phosphate product, but it can demonstrate the inositol phosphate kinase activity of a protein of interest. With this assay, we tested several inositol monophosphates and polyphosphates. Under the assay condition as described in "Materials and Methods," the recombinant ZmIpk protein could phosphorylate Ins(1,3,4)P₃, Ins(3,5,6)P₃, and Ins(3,4,5,6)P₄, but not Ins(1,3,4,6)P₄ and Ins(1,3,5,6)P₄ (Fig. 2). The specific activity was higher on Ins(3,4,5,6)P₄ than Ins(1,3,4)P₃ and Ins(3,5,6)P₃ (Table I). The ZmIpk enzyme also could use Ins(1,2,5,6)P₄ as a substrate (data not shown). The ZmIpk protein displayed very weak kinase activity when incubated with Ins(1,3,4,5)P₄, and two products were produced (Fig. 2). The protein could not phosphorylate Ins(1,4,5)P₃, Ins(2,4,5)P₃, Ins(1,4)P₂, Ins(1)P, Ins(2)P, and Ins(4)P (data not shown).

View larger version (102K): [in this window] [in a new window]   Figure 2.   Inositol phosphate kinase activity assay with [-32P]ATP. A, Radioautograph of inositol phosphate kinase assay. The reaction mixtures contained 5 µL of the recombinant ZmIpk protein, 40 µM inositol phosphate as indicated in each lane, 40 µM ATP, and 0.5 µL of [-32P]ATP (3,000 Ci mmol1) in a total volume of 20 µL. The reaction product was separated on a PEI-cellulose-coated TLC plate. Lane 1, Ins(1,3,4)P3, the enzyme was inactivated by boiling before added to the reaction mixture; lane 2, Ins(1,3,4)P3; lane 3, Ins(1,3,4,5)P4; lane 4, Ins(1,3,4,6)P4; lane 5, Ins(1,3,5,6)P4; lane 6, Ins(3,4,5,6)P4; and lane 7, Ins(3,5,6)P3. B, Authentic inositol phosphates purchased from Sigma-Aldrich (St. Louis) were loaded on the TLC plate and developed with 0.5 N HCl as in A. The plate then was sprayed with ammonium molybdate/HClO4/HCl solution to show the migration of the inositol phosphates. Lane 1, Ins(1,3,4,5)P4; lane 2, Ins(1,3,4,6)P4; lane 3, Ins(1,3,5,6)P4; lane 4, Ins(3,4,5,6)P4; and lane 5, Ins(1,3,4,5,6)P5.

View this table: [in this window] [in a new window]   Table I.   Inositol phosphate kinase activity of the ZmIpk The reaction mixtures contained 5 µL of the recombinant ZmIpk protein, 40 µm inositol phosphate, 40 µm ATP, and 0.5 µL of [-32P]ATP (3,000 Ci mmol1) in a total volume of 20 µL. The reaction products were separated on a PEI-cellulose-coated TLC plate. Radioactivity was quantified by liquid scintillation counting.

Screening for ZmIpk Mu Insertion Mutants

To elucidate the function of ZmIpk, we used a reverse genetics approach to generate loss-of-function alleles. The ZmIpk-specific PCR primers were paired with a Mutator (Mu) primer to screen a population of plants carrying Mutator transposable elements. Four independent lines with Mu inserted into the ZmIpk gene, designated as ZmIpk-mum, were identified from a collection of about 40,000 Mu insertion lines.

The Mu insertion site in the ZmIpk gene was mapped by sequencing the Mu-ZmIpk junction region (Fig. 3). The Mu insertion in ZmIpk-mum1, ZmIpk-mum2, and ZmIpk-mum3 was localized to the 5'-terminal region at nucleotide positions 237 (amino acid position 61), 245 (amino acid position 64), and 366 (amino acid position 104), respectively. The Mu insertion of ZmIpk-mum4 occurred in the 3'-terminal region at the nucleotide position 872 (amino acid position 273). The Mu insertion disrupted the open reading frame of the ZmIpk gene.

View larger version (27K): [in this window] [in a new window]   Figure 3.   Mu transposon insertion sites in ZmIpk-mum mutants and nucleotide mutation in lpa2-2 allele. The Mu-ZmIpk junction region in ZmIpk-mum alleles and ZmIpk gene in lpa2-2 allele were PCR-amplified and sequenced. The rectangular box represents the open reading frame of the ZmIpk gene. The triangle marks the Mu insertion site. *, The site of mutated nucleotide in lpa2-2 allele. The underlined sequences represent the characteristic duplication associated with Mu insertion.

ZmIpk Mu Insertion Mutants Show Low-Phytic Acid Phenotype

Lines carrying the ZmIpk-mum alleles displayed a recessive low-phytic acid phenotype. F₂ seeds of the ZmIpk Mu insertion lines were analyzed for phytic acid and Pi. About 25% of the F₂ seeds had on average 30% less phytic acid and about 3-fold more Pi. The remaining 75% of the F₂ seeds expressed the wild-type phenotype for phytic acid and Pi contents (Table II). Similar phenotypes in four independent Mu insertion lines and the segregation ratio support the assumption that the low-phytic acid phenotype is caused by Mu insertion in the ZmIpk gene. This assumption was confirmed by genotyping F₂ seeds. PCR was conducted with ZmIpk gene-specific primers flanking Mu insertion sites. A PCR product of 1.3-kb fragment is expected to be amplified from the intact ZmIpk gene, but not from ZmIpk-mum alleles. The seeds with genotype of ZmIpk/ZmIpk and ZmIpk/ZmIpk::Mu will yield the 1.3-kb PCR fragment, whereas ZmIpk::Mu/ZmIpk::Mu will not. It was found that the low-phytic acid kernels did not contain an intact copy of the ZmIpk gene (Table II). The presence of ZmIpk::Mu alleles in these low-phytate kernels was confirmed by PCR using Mu primer and gene-specific primer. In the kernels with normal phytic acid and Pi contents, the 1.3-kb DNA fragment was amplified from intact ZmIpk gene (Table II), indicating that at least one ZmIpk copy did not have a Mu insertion. F3 and subsequent generations showed that all four Mu insertion lines had the low-phytic acid and high-Pi phenotype. These experiments demonstrated that the Mu insertion in the maize ZmIpk gene causes the low-phytic acid and high-Pi phenotype.

In addition to the change in kernel phytic acid and Pi, ZmIpk Mu insertion lines accumulated myo-inositol, InsP₃, InsP₄, and InsP₅ in kernels (Table III; Fig. 4). Embryos were isolated from mutant and nonmutant control seeds, and myo-inositol content was measured. ZmIpk Mu insertion mutant had higher myo-inositol content in embryos than nonmutant control (Table III).

View larger version (16K): [in this window] [in a new window]   Figure 4.   Inositol phosphate accumulation in ZmIpk Mu insertion mutants. HPLC analysis of inositol phosphates in mature seeds of ZmIpk Mu insertion mutants (A) and nonmutant control (B).

ZmIpk Mu Insertion Mutants Are Allelic to lpa2

lpa2 is a recessive, low-phytic acid mutant. The mutant accumulates InsP₃, InsP₄, InsP₅, and Pi in seeds. We found the embryo of lpa2 mutant seeds also accumulated myo-inositol (Table III), as the ZmIpk-mum mutants. It is not known what gene is responsible for the lpa2 mutation. We were interested to see how the lpa2 mutation and the ZmIpk Mu insertion knockout are related because they have a very similar phenotype. The homozygous plants carrying the ZmIpk-mum-3 allele were crossed with the recessive lpa2-1 and lpa2-2 alleles. Individual F₁ seeds from both crosses displayed the high-Pi phenotype in a rapid Pi test. Ten F₁ kernels from each ear were pooled and assayed quantitatively for phytic acid and Pi. The F₁ seeds showed low-phytic acid and high-Pi phenotype (Table IV). Control crosses of the ZmIpk-mum-3 allele and a nonmutant line produced F₁ seeds with normal phytic acid and Pi contents. These experiments demonstrated that the lpa2 mutants carry mutation in the ZmIpk gene.

The ZmIpk gene was amplified using PCR from the lpa2-2 allele. A mutation of C to T was found at the nucleotide position 158 (Fig. 3). The nucleotide change introduced a stop codon at amino acid position 35 in the place of Gln of the wild-type ZmIpk. The immature polypeptide has only 34 amino acid residues and thus is too short to possess inositol phosphate kinase activity.

PCR amplification of the ZmIpk gene from the lpa2-1 allele using the primers flanking the open reading frame did not produce any product. Southern analysis revealed different band patterns between lpa2-1 mutant and nonmutant near-isogenic lines (Fig. 5). The probe used for the Southern blotting covered the nucleotide positions 367 to 1,088. This region contains a BamHI restriction site, and two bands were detected in the nonmutant line in the Southern analysis, as expected. In the lpa2-1 mutant line, two bands also were seen, but the fragments are larger. The restriction enzyme EcoRI, EcoRV, HindIII, and XbaI do not cut in the probe region, and a single band was detected in the nonmutant line, as expected. For the lpa2-1 mutant line, XbaI digestion revealed two bands, and HindIII digestion produced two fragments of about 0.7 and 1.6 kb which do not exist in the nonmutant ZmIpk gene. This result indicated that a rearrangement of the genomic sequence occurred in the ZmIpk locus of lpa2-1 mutant and that an intact ZmIpk gene is absent in the lpa2-1 mutant. The ZmIpk mRNA expression in immature seeds was examined with reverse transcriptase-PCR, and no expression was detected in the lpa2-1 mutant.

View larger version (27K): [in this window] [in a new window]   Figure 5.   Southern-blot analysis of the ZmIpk gene in the lpa2-1 mutant. A, Diagram of the ZmIpk gene restriction sites. The rectangular box represents the open reading frame of the ZmIpk gene. B, Southern-blot analysis of the lpa2-1 mutant line (lanes 1, 3, 5, 7, and 9) and near-isogenic nonmutant line (lanes 2, 4, 6, 8, and 10).

ZmIpk Expression

Northern-blot analysis was conducted with total RNA from various tissues and at different developmental stages. The maize ZmIpk gene was expressed in the embryo of 15 d after pollination (Fig. 6). The gene was expressed in kernels at earlier stages but very low levels. ZmIpk gene expression in the embryo peaked at 15 d after pollination and then declined. No expression was detected from endosperm and vegetative tissues by Northern analysis of total RNA.

View larger version (67K): [in this window] [in a new window]   Figure 6.   Northern-blot analysis of the ZmIpk gene. Total RNA from different tissues of B73 was separated on 1% (w/v) gel. After transfer, the blot was probed with the maize ZmIpk. Ethidium bromide-stained gel was shown at bottom as a control for loading. Each lane contained 10 µg of total RNA.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

By knocking out the maize ZmIpk gene with Mu insertion, we demonstrated that the ZmIpk gene is involved in phytic acid biosynthesis in developing maize seeds. The ZmIpk Mu insertion mutant seeds have a reduced phytic acid content and accumulate Pi and myo-inositol, as well as inositol phosphates, which were not detectable in nonmutant seeds. The ZmIpk protein possesses inositol phosphate kinase activity, and its mRNA is expressed in embryo where phytic acid accumulates in seeds. These results provided evidence that the ZmIpk is one of the kinases responsible for the phosphorylation of Ins(3)P to InsP₆. By allelic tests, Southern-blot analysis, cloning, and sequencing of the ZmIpk gene from low-phytic acid lpa2 mutants, we found that the mutant lpa2 carries a mutation in the ZmIpk gene.

Previous studies showed that the phytic acid biosynthesis follows a pathway of Ins(3)P  Ins(3,4)P₂  Ins(3,4,6)P₃  Ins(3,4,5,6)P₄  Ins(1,3,4,5,6)P₅  InsP₆ in the duckweed Spirodela polyrhiza (Brearley and Hanke, 1996a) and Ins(3)P  Ins(3,6)P₂  Ins(3,4,6)P₃  Ins(1,3,4,6)P₄  Ins(1,3,4,5,6)P₅  InsP₆ in the slime mold Dictyostelium discoideum (Stephens and Irvine, 1990; van Haastert and van Dijken, 1997). In both species, the phosphorylation of Ins(3)P to InsP₆ was described as a linear and stepwise process through a single pathway. If maize seeds have a similar pathway, a lack of phytic acid synthesis would be expected in ZmIpk total-knockout alleles. In the ZmIpk Mu insertion mutants and lpa2 alleles, we observed only 30% phytic acid reduction. One possible explanation could be that these mutants lost only part of the ZmIpk activity. However, this is very unlikely because the Mu was inserted in the 5' region of ZmIpk gene, which corresponds to the N-terminal region of ZmIpk protein. The disrupted open reading frame encodes only 60, 63, and 103 amino acid residues in the lines carrying ZmIpk-mum1, ZmIpk-mum2, and ZmIpk-mum3 alleles, respectively. In the lpa2-2 allele, the immature ZmIpk peptide has only 34 amino acid residues. These peptides would be too short to have any inositol phosphate kinase activity. In the lap2-1 allele, the genomic sequence is rearranged in the ZmIpk locus and ZmIpk mRNA expression is undetectable. An alternative explanation could be the existence of one or more salvage pathways, as proposed in barley developing grain (Hatzack et al., 2001). However, the phenotype of ZmIpk-mum and lpa2 alleles is also consistent with a hypothesis that maize seeds have multiple inositol phosphate phosphorylation pathways for phytic acid biosynthesis. In addition to the ZmIpk gene reported here, we have isolated other genes that have sequence similarity to Ins(1,3,4)P₃ 5/6-kinase genes ( J. Shi, H. Wang, and Y. Wu, unpublished data).

Cloning the ZmIpk gene and identification of the ZmIpk Mu insertion mutants provided tools to dissect further the Ins(3)P phosphorylation process. Previous studies of the maize lpa2-1 allele have identified the inositol phosphates accumulated in the mutant seeds as Ins(1,2,4,5,6)P₅, Ins(1,4,5,6)P₄, and Ins(1,2,6)P₃ (Raboy et al., 2000). In addition to the inositol phosphates, we also observed the accumulation of myo-inositol in the ZmIpk-mum alleles. These compounds could be intermediates of phytic acid biosynthesis. They also could be metabolic products of the intermediates because developing seeds have many inositol phosphate kinase and phosphatase activities (Biswas et al., 1984; Brearley and Hanke, 1996b; Hatzack et al., 2001). The existence of recombinant ZmIpk protein will allow testing if the ZmIpk can use these compounds and other inositol phosphates as substrates. The preferred substrate of the ZmIpk may be identified by in vitro analysis, such as kinetic study and enantiomeric identification of the phosphorylation products. In addition, the production of low-phytic acid mutants and their double or triple combination would be helpful in understanding the pathway. Along these lines, we have identified other mutants with a low-phytic acid phenotype.

The maize inositol phosphate kinase gene belongs to the Ins(1,3,4)P₃ 5/6-kinase gene family with members found in mammals, higher plants, and Entamoeba histolytica, but not in yeast (Wilson and Majerus, 1996, 1997; Field et al., 2000). Although the protein sequence similarity is low among the members from distanced species, the homologous region extends over the entire sequence. This group was initially considered solely as Ins(1,3,4)P₃ 5/6-kinases. However, E. histolytica Ins(1,3,4)P₃ 5/6-kinase was found to have Ins(1,4,5)P₃ 3-kinase activity also (Field et al., 2000). Recent studies on human Ins(1,3,4)P₃ 5/6-kinase revealed that the enzyme also has Ins(3,4,5,6)P₄ 1-kinase activity and inositol phosphatase activity (Yang and Shears, 2000; Ho et al., 2002). The maize ZmIpk can use Ins(1,3,4)P₃, Ins(3,5,6)P₃, Ins(3,4,5,6)P₄, and Ins(1,2,5,6)P₄ as substrates. This enzyme is likely to phosphorylate the 1-hydroxyl group of Ins(3,5,6)P₃ and Ins(3,4,5,6)P₄ because the products comigrated on TLC plates with Ins(1,3,5,6)P₄ and Ins(1,3,4,5,6)P₅, respectively, but this interpretation of the data needs to be supported by enantiomeric structural analysis. We also noticed that the ZmIpk protein produced minor amounts of InsP₄ and InsP₅ when incubated with Ins(1,3,4,5)P₄. It is possible that the ZmIpk protein has a phosphatase activity on Ins(1,3,4,5)P₄. InsP₄ may be generated through rephosphorylation of the product from the phosphatase reaction.

Grain with low phytic acid and high Pi can reduce the environmental impact of animal waste. Such grain is also beneficial to human and animal nutrition because it has more available Fe and Zn. The low-phytic acid mutants from ethyl methane-sulfonate (EMS) mutageneses are being used to genetically breed low-phytic acid crops. These mutants are loss-of-function mutation, and the low-phytic acid accumulation is a recessive trait. Cloning ZmIpk gene provides an alternative choice to manipulate phytic acid content in grain. Silencing the ZmIpk gene through transformation could reduce the phytic acid biosynthesis and produce a dominant trait of low phytic acid.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Material

Maize (Zea mays) low-phytic acid mutant allele lpa2-1 was identified in an EMS-mutagenized population, although it appears to have been a spontaneous mutation preceding the mutagenesis treatments (Raboy et al., 2000). The lpa2-2 allele was identified in an EMS-mutagenized population (V. Raboy, personal communication). The lpa2-1 stock was backcrossed twice with inbred line PHJ90, and the lpa2-2 stock was backcrossed four times with PHN46. Both backcross progeny were self-pollinated to homozygosity and were used in the allelism test crosses with the Mu insertion lines.

Identification and Cloning of Maize Inositol Phosphate Kinase Gene

Human and Arabidopsis Ins(1,3,4)P₃ 5/6-kinase gene (GenBank accession nos. U51336 and JC5401) were used to search the Pioneer/DuPont EST database using the BLAST program. An EST from a cDNA library of B74 anther had an overall similarity with Ins(1,3,4)P₄ 5/6-kinase genes. The insert of the clone was sequenced using a fluorescent sequencer (model 377, PerkinElmer Instruments, Norwalk, CT). To isolate genomic clones, two primers (5' primer, 5'-ATTCCTCCCGAACCCGACCCGATGGC-3', and 3' primer, 5'-AGCTCGTTTTTCATTAGAATTCCG-3') were used to conduct PCR with B73 genomic DNA as templates. The PCR products were cloned into plasmid vector pCR2.1 (Invitrogen, Carlsbad, CA).

Identification of ZmIpk Mu Insertion Mutants

A mutant containing a Mu-transposable element insertion was identified in a collection of indexed mutagenized F₂ families derived from several Mu active stocks (Bensen et al., 1995). The mutant was identified using a Mu-specific primer (5'-AGAGAAGCCAACGCCA(A/T) CGCCTC(C/T) ATTTCGTC-3') and ZmIpk gene-specific primers (5'-CCGAAGAAGCAGCAAAGCTTCATCCAG-3' and 5'-TGGTTTGGAAAGAGCTAGGAGGTCCTC-3'). To reduce Mu copy number in the background, the mutant line carrying ZmIpk-mum3 allele was backcrossed with inbred line PHP38. The ZmIpk-mum3 allele was tracked by monitoring the low-phytate phenotype of the corresponding self-pollinated ears. After three backcrosses, the ZmIpk Mu insertion line was self-pollinated to produce ZmIpk-mum3 homozygotes. The presence of the ZmIpk-mum3 allele in the homozygous plants was confirmed by PCR.

Determination of Phytic Acid and Pi

Phytic acid and Pi in dry, mature seeds were assayed according to modifications of the methods described by Haug and Lantzsch (1983) and Chen et al. (1956), respectively. Single kernels were ground using a Geno/Grinder2000 (Sepx CertiPrep, Metuchen, NJ). Twenty-five- to 35-mg samples were placed into 1.5-mL Eppendorf tubes. One milliliter of 0.4 N HCl was added, and the tubes were shaken on a gyratory shaker at room temperature for 3.5 h. The tubes were then centrifuged at 3,900g for 15 min. Supernatants were transferred into fresh tubes and used for both phytic acid and Pi determinations. Measurements were performed in duplicate. For phytic acid assay, 35 µL of each extract was placed into wells of a 96-well microtiter plate. Thirty-five microliters of distilled water and 140 µL of 0.02% (w/v) ammonium iron (III) sulfate-0.2 N HCl were added to each sample. The plate was covered with a rubber lid and heated in a thermal-cycler at 99°C for 30 min. The plate was cooled to 4°C, kept on an ice-water bath for 15 min, and then left at room temperature for 20 min. The plate was sealed with sticky foil and centrifuged at 3,900g at 24°C for 30 min. Eighty microliters of each supernatant was placed into wells of a fresh 96-well plate, 120 µL of 1% (w/v) 2,2'-bipyridine-1% (v/v) thioglycolic acid was added to each well, and then absorbance was recorded at 519 nm using a VERSAmax microplate reader (Molecular Devices, Sunnyvale, CA). Phytic acid content is presented as phytic acid phosphorus. Authentic phytic acid (P-7660, Sigma-Aldrich) served as a standard. The phytic acid assay may also measure InsP₅ and InsP₄ if they present in samples. To determine Pi, 200 µL of each extract was placed into wells of a 96-well microtiter plate. One hundred microliters of 30% (w/v) aqueous trichloroacetic acid was added to each sample, and the plates were shaken and centrifuged at 3,900g for 10 min. Fifty microliters of each supernatant was transferred into a fresh plate and 100 µL of 0.42% (w/v) ammonium molybdate-1 N H₂SO₄:10% (w/v) ascorbic acid (7:1) was added. The plates were incubated at 37°C for 30 min and then A₈₀₀ was measured. Potassium phosphate was used as a standard. Pi content is presented as Pi phosphorus.

A rapid test also was used to assay Pi content in kernels. Individual kernels were placed in a 25-well plastic tray and crushed at 2,000 psi using a hydraulic press. Two milliliters of 1 N H₂SO₄ was added into each sample and incubated at room temperature for 2 h. Four milliliters of 0.42% (w/v) ammonium molybdate-1 N H₂SO₄:10% (w/v) ascorbic acid (6:1) was added. In the case of an increased Pi content, blue color developed in 20 min. Nonmutant kernels served as a negative control, and mutant lpa2 kernels served as a positive control.

Determination of Seed myo-Inositol

myo-Inositol was quantified in dry, mature seeds and excised embryos. Tissue was ground as above and mixed thoroughly. One hundred-milligram samples were placed into 7-mL scintillation vials. One milliliter of 50% (v/v) aqueous ethanol was added, and the vials were shaken on a gyratory shaker at room temperature for 1 h. Extracts were decanted through a 0.45-µm nylon syringe filter attached to a 1-mL syringe barrel. Residues were re-extracted with 1 mL of fresh 50% (v/v) aqueous ethanol, and the second extracts were filtered as before. The two filtrates were combined in a 10- × 75-mm glass tube and evaporated to dryness in a speedvac (Savant Instruments, Holbrook, NY). The myo-inositol derivative was produced by redissolving the residues in 50 µL of pyridine and 50 µL of trimethylsilylimadazole:trimethylchlorosilane (100:1; Tacke and Casper, 1996). The silylation reaction is compromised if a precipitate appears at this stage. The tubes were capped and incubated at 60°C for 15 min. One milliliter of 2,2,4-trimethylpentane and 0.5 mL of distilled water were added to each sample, and each was vortexed and then centrifuged at 1,000g for 5 min. The upper organic layers were transferred with Pasteur pipettes into a 2-mL glass autosampler vial and crimp capped. myo-Inositol, as hexa-trimethylsilyl ethers, was quantified with an gas chromatograph (model 5890, Agilent, Palo Alto, CA) coupled with an mass spectrometer (model 5972, Agilent). Measurements were performed in triplicate. One-microliter samples were introduced in the splitless mode onto a 30-m × 0.25-mm i.d. × 0.25-µm film thickness 5MS column (Agilent). The initial oven temperature of 70°C was held for 2 min, after which it was increased at 25°C min¹ to 170°C, then at 5°C min¹ to 215°C, and finally at 25°C min¹ to 250°C, at which it was held for 5 min. The inlet and transfer line temperatures were 250°C. Helium at a constant flow of 1 min¹ was the carrier gas. Electron impact mass spectra from m/z 50 to 560 were acquired at 70 eV after a 5-min solvent delay. The myo-inositol derivative was well resolved from other peaks in the total ion chromatograms. Authentic myo-inositol standards in aqueous solutions were dried, derivatized, and analyzed at the same time. Regression coefficients of four-point calibration curves were typically 0.999 to 1.000.

Determination of Seed Inositol Phosphates

The presence of significant amounts of inositol phosphates in mature seeds was determined by HPLC according to the Dionex Application Note AN65: Analysis of Inositol Phosphates (Dionex Corporation, Sunnyvale, CA). Tissue was ground and mixed as above. Five hundred-milligram samples were placed into 20-mL scintillation vials to which 5 mL of 0.4 M HCl was added. The samples were shaken on a gyratory shaker at room temperature for 2 h and then allowed to sit at 4°C overnight. Extracts were centrifuged at 1,000g for 10 min and filtered through a 0.45-µm nylon syringe filter attached to a 5-mL syringe barrel. Just before HPLC analysis, 600-µL aliquots were clarified by passing through a 0.22-µm centrifugal filter. A Dionex DX 500 HPLC with a Dionex model AS3500 autosampler was used. Twenty-five microliter samples were introduced onto a Dionex 4- × 250-mm OmniPac PAX-100 column. Dionex 4- × 50-mm OmniPac PAX-100 guard and ATC-1 anion trap columns were used. Inositol phosphates were eluted at 1 mL min¹ with the following mobile phase gradient: 68% A (distilled water)/30% B (200 mM NaOH) for 4.0 min, 39% A/59% B at 4.1 through 15.0 min, and return to initial conditions at 15.1 min. The mobile phase contained 2% C (50% [v/v] aqueous isopropanol) at all times to maintain column performance. A Dionex conductivity detector module II was used with a Dionex ASRS-Ultra II anion self-regenerating suppressor set up in the external water mode and operated with a current of 300 mA. Although quantitative standards were available, InsP₃, InsP₄, and InsP₅ were partially but clearly resolved from each other and InsP₆.

Recombinant Protein Expression and Purification

The ZmIpk coding region was reverse transcriptase-PCR-amplified from seed RNA using the following primers: 5'-ATTCCTCCCGAACCCGACCCGATGGC-3' and 5'-CATCTTATTTCACGACAACATGGTTG-3'. The PCR product was cloned into the vector pCR2.1 (Invitrogen), and the insert was sequenced to confirm its sequence and then digested with EcoRI. The EcoRI fragment was cloned into the pGEX.4T-1 expression vector (Amersham Biosciences Inc., Piscataway, NJ). The methods to transform Escherichia coli (strain DH5), to induce with isopropyl--D-thiogalactopyranoside and to purify the GST-tagged protein were all according to the manufacturer's recommendations. In brief, a single colony of E. coli transformed with the GST fusion construct was grown at 37°C overnight, diluted with fresh media, and cultured with vigorous shaking until the OD₆₀₀ reached 0.6. Isopropyl--D-thiogalactopyranoside was added to induce the expression of the GST-ZmIpk fusion protein. Cells were harvested by centrifugation, and the pellets were resuspended in ice-cold bacterial lysis buffer (50 mM Tris-HCl, pH 7.4, 100 mM NaCl, and 100 µM phenylmethylsulfonyl fluoride). Bacterial cells were lysed by sonication. Triton X-100 was added to a final concentration of 1%. After incubation on ice for 1 h, the cell lysate was centrifuged at 10,000g for 10 min at 4°C. Glutathione-Sepharose 4B beads (Amersham Biosciences Inc.) were added to the supernatant and incubated with gentle shaking for 45 min at 4°C. The Sepharose beads were washed four times with the cell lysis buffer and twice with phosphate-buffered saline. The GST-ZmIpk fusion protein was eluted with 10 mM glutathione in 50 mM Tris-HCl and 100 mM NaCl. For every 500 mL of the bacterial culture, 200 µL of the elution buffer were used to elute the ZmIpk protein. The eluted protein was used for inositol phosphate kinase assay.

Inositol Phosphate Kinase Assay

Inositol phosphate kinase activities were assayed according to Wilson and Majerus (1996) with modifications. Assays were carried out in 20 µL of reaction mixture containing 5 µL of enzyme preparation, 20 mM HEPES (pH 7.2), 6 mM MgCl₂, 10 mM LiCl, 1 mM dithiothreitol, 40 µM ATP, and 0.5 µL of [-³²P]ATP (3,000 Ci mmol¹), 40 µM inositol phosphates (Sigma-Aldrich). After incubation at 30°C for 30 min, reactions were stopped by adding 10 µL of 0.2 N HCl. The samples were taken to dryness using a Savant-Vac concentrator, and then resuspended in 20 µL of 0.3 M HCl-0.2 M KH₂PO₄. One microliter of each sample was loaded on a PEI-cellulose-coated TLC plate (Merck, Gibbstown, NJ) to separate the products according to Spencer et al. (1990). The plates were developed in 0.5 N HCl, dried in air at 60°C, and autoradiographed with x-ray film (Eastman Kodak, Rochester, NY). Radioactivity was quantified in excised areas of the plate corresponding to radioactively labeled inositol phosphate by liquid scintillation counting.

Southern- and Northern-Blot Analysis

Genomic DNA were isolated from leaf tissues. Ten micrograms of DNA was digested with various restriction enzymes, resolved on 1% (w/v) agarose gel, and transferred to nylon membrane (Bio-Rad, Hercules, CA). Total RNA from various tissues and developmental stages were prepared using Purescript RNA isolation kit (Gentra, Minneapolis). Ten micrograms of RNA was resolved on 1% (w/v) agarose/formaldehyde/MOPS gel and transferred to nylon membrane. Preparation of ZmIpk probe, hybridization, and washing were carried out according to the manufacturer's instructions.