Arabidopsis Inositol Polyphosphate 6-/3-Kinase Is a Nuclear Protein That Complements a Yeast Mutant Lacking a Functional ArgR-Mcm1 Transcription Complex

Abstract

Inositol 1,4,5-trisphosphate 3-kinase, and more generally inositol polyphosphate kinases (Ipk), play important roles in signal transduction in animal cells; however, their functions in plant cells remain to be elucidated. Here, we report the molecular cloning of a cDNA (AtIpk2β) from a higher plant, Arabidopsis. Arabidopsis AtIpk2β is a 33-kD protein that exhibits weak homology (∼25% identical amino acids) with Ipk proteins from animals and yeast and lacks a calmodulin binding site, as revealed by sequence analysis and calmodulin binding assays. However, recombinant AtIpk2β phosphorylates inositol 1,4,5-trisphosphate to inositol 1,4,5,6-tetrakisphosphate and also converts it to inositol 1,3,4,5,6-pentakisphosphate [Ins(1,3,4,5,6)P₅]. AtIpk2β also phosphorylates inositol 1,3,4,5-tetrakisphosphate to Ins(1,3,4,5,6)P₅. Thus, the enzyme is a D3/D6 dual-specificity inositol phosphate kinase. AtIpk2β complements a yeast ARG82/IPK2 mutant lacking a functional ArgR-Mcm1 transcription complex. This complex is involved in regulating Arg metabolism–related gene expression and requires inositol polyphosphate kinase activity to function. AtIpk2β was found to be located predominantly in the nucleus of plant cells, as demonstrated by immunolocalization and fusion to green fluorescent protein. RNA gel blot analysis and promoter–β-glucuronidase reporter gene studies demonstrated AtIpk2β gene expression in various organs tested. These data suggest a role for AtIpk2β as a transcriptional control mediator in plants.

INTRODUCTION

In animal cells, inositol 1,4,5-trisphosphate 3-kinase, and more generally inositol polyphosphate kinases, play an important role in signal transduction by directly regulating the levels of inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] and inositol 1,3,4,5-tetrakisphosphate [Ins(1,3,4,5)P₄] (Communi et al., 1995b). Ins(1,4,5)P₃ is a critical second messenger responsible for generating localized transient Ca²⁺ increases (Berridge and Irvine, 1989; Berridge, 1997). Evidence suggests that Ins(1,3,4,5)P₄ itself may regulate intracellular Ca²⁺ concentration by promoting Ca²⁺ sequestration (Hill et al., 1988) or by acting synergistically with InsP₃ to mobilize intracellular Ca²⁺ stores (Morris et al., 1987). Other animal studies provide evidence for a direct second-messenger function of Ins(1,3,4,5)P₄ in Ca²⁺ release from intracellular stores (Lückhoff and Clapham, 1992). The discovery of specific Ins(1,3,4,5)P₄ receptors in various tissues further supports a role for Ins(1,3,4,5)P₄ in the inositol phosphate signaling pathway (Cullen, 1998).

Several cDNAs that encode Ins(1,4,5)P₃ kinase have been isolated and functionally characterized from rat (Choi et al., 1990; Thomas et al., 1994), human (Takazawa et al., 1991a, 1991b; Dewaste et al., 2000), and chicken (Bertsch et al., 1999). At least three distinct isoforms (A, B, and C) have been identified. The expression patterns of the A and B isoforms are tissue specific (Vanweyenberg et al., 1995). Various isoforms are regulated differentially by calcium/calmodulin, and the degree of stimulation of isoform C was much lower than that of Ins(1,4,5)P₃ kinase isoforms A and B (Woodring and Garrison, 1997; Dewaste et al., 2000).

Two main functional domains have been identified in Ins(1,4,5)P₃ kinase: a conserved C-terminal catalytic domain of ∼275 amino acids, and a nonconserved N-terminal regulatory domain of variable length (Takazawa and Erneux, 1991; Communi et al., 1993; Togashi et al., 1997). Animal Ins(1,4,5)P₃ kinases have been expressed functionally in Escherichia coli, either as maltose binding fusion proteins or β-galactosidase fusion proteins (Takazawa and Erneux, 1991; Thomas et al., 1996). Mammalian Ins(1,4,5)P₃ kinases are activated by Ca²⁺/calmodulin to different extents (Sim et al., 1990; Sims and Allbritton, 1998). Interestingly, the recently characterized Ins(1,4,5)P₃ kinase from the nematode Caenorhabditis elegans lacks the consensus calmodulin binding site and, not surprisingly, is insensitive to Ca²⁺/calmodulin (Clandinin et al., 1998). Phosphorylation experiments demonstrated that Ins(1,4,5)P₃ kinase isoforms A and B are substrates for protein kinase C– and protein kinase A–mediated phosphorylation in vitro (Sim et al., 1990; Wang et al., 1995; Communi et al., 1999).

An exciting new function for inositol 1,4,5-trisphosphate kinase was discovered recently. Thus, Odom et al. (2000) reported that a dual-specificity inositol polyphosphate kinase (Ipk2), which phosphorylates Ins(1,4,5)P₃ to inositol 1,4,5,6-tetrakisphosphate [Ins(1,4,5,6)P₄] and [Ins(1,4,5,6)P₄] to inositol 1,3,4,5,6-pentakisphosphate [Ins(1,3,4,5,6)P₅], is an indispensable component of the ArgR-Mcm1 transcriptional complex in yeast. The ArgR-Mcm1 complex consists of four proteins (Arg80p, Arg81p, Arg82p, and Mcm1p), all of which are required for the proper control of transcription (Bechet et al., 1970; Messenguy and Dubois, 1993). Arg82p is identical to Ipk2p. Inositol polyphosphate multikinase activity of the Arg82 protein (also called ArgRIII) was discovered independently by Saiardi et al. (1999)(2000). Arg82p and Mcm1p are pleiotropic regulators, whereas Arg80p and Arg81p serve as specific transcription factors in Arg metabolism. Mcm1 is a canonical member of the MADS (MCM1, Agamous, Deficiens, Serum Response Factor) box family of transcription factors. When the ARG82/IPK2 gene was disrupted in yeast, cell growth was slowed at 30°C and strongly impaired at 37°C.

To investigate whether or not the biosynthetic activity of Ipk2p is required for transcriptional regulation, a kinase-inactive Ipk2 mutant protein was generated. When the kinase-inactive Ipk2 protein was expressed in the ARG82/IPK2 deletion strain, formation of the ArgR-Mcm1 transcriptional complex still occurred, as shown by gel retardation experiments (binding to ArgR-Mcm1 target promoter elements). However, the kinase-inactive mutant was unable to complement the ARG82/IPK2 yeast mutant in a phenotypic growth assay that tests for cellular transcriptional activity, indicating that the synthesis of InsP₄/InsP₅ is necessary for gene regulation via the ArgR-Mcm1 complex in vivo. By contrast, the production of inositol hexakisphosphate (InsP₆) was not required for transcriptional activity. These experiments revealed a direct mechanism by which activation of the inositol phosphate signaling pathway controls gene expression (Odom et al., 2000).

In plants, calcium is accepted as an important second messenger and InsP₃ is known to regulate Ca²⁺ mobilization (Munnik et al., 1998; Malho, 1999; Rudd and Franklin-Tong, 2001; Sanders et al., 2002). Recent findings suggest that nuclei from plant cells are capable of generating calcium signals independent of changes in cytosolic calcium ion concentration (Pauly et al., 2000). To date, however, very little is known about Ins(1,4,5)P₃ kinase function in higher plant cells, although Ins(1,4,5)P₃ 6-kinase activity has been detected in pea roots (Chattaway et al., 1992). Here, we describe the heterologous expression of an Arabidopsis inositol polyphosphate kinase (AtIpk2β) in E. coli, some of the biochemical properties of the recombinant enzyme, and the expression pattern of the AtIpk2β gene in Arabidopsis. Stevenson-Paulik et al. (2002) independently analyzed the biochemical characteristics of the AtIpk2β protein using the cDNA sequence cloned here. Furthermore, a role for AtIpk2β in regulating gene expression is demonstrated by functional complementation of the ARG82/IPK2 yeast mutant and a phenotypic assay of ArgR-Mcm1 transcriptional activity. Localization studies using green fluorescent protein (GFP) fusion protein and immunohistochemistry demonstrate that AtIpk2β is a nuclear protein in higher plants.

RESULTS

Cloning and Sequence Analysis of an Arabidopsis Inositol Polyphosphate 6-/3-Kinase

To identify plant genes that encode inositol polyphosphate kinase, animal Ins(1,4,5)P₃ kinase sequences were used to screen the GenBank database using the Basic Local Alignment Search Tool search algorithm. One of the genomic fragments identified is located on the P1 clone MAC9. This fragment, which is part of Arabidopsis chromosome 5 (http://www.kazusa.or.jp/kaos/), was amplified by PCR. It contains an open reading frame without introns of 903 nucleotides that encodes a protein of 300 amino acids with a calculated molecular mass of 33 kD. The amplified DNA then was used as a radioactively labeled probe to screen an Arabidopsis cDNA library. Three positive clones were identified. The longest cDNA, AtIpk2β (for nomenclature, see Stevenson-Paulik et al., 2002), was sequenced and found to hold the same coding region as its corresponding genomic counterpart. The complete AtIpk2β cDNA was ∼1.2 kb, including 5′ and 3′ untranslated regions. AtIpk2β shares 73% identity (84% similarity) with a second Arabidopsis inositol polyphosphate kinase, AtIpk2α, that we cloned and that was characterized enzymatically by Stevenson-Paulik et al. (2002).

The AtIpk2β protein exhibits only low sequence homology with Ins(1,4,5)P₃ kinase/inositol polyphosphate kinase from yeast and animals, with ∼25% identical and between 50 and 60% similar amino acid residues. Figure 1 shows an alignment of the AtIpk2β protein with inositol polyphosphate kinase sequences from yeast and animals. Although considerable sequence variation exists among these proteins, particularly in the N-terminal regions, several regions are conserved. The Arabidopsis protein lacks the N-terminal cal-modulin binding domain that is present in the human, rat, and chicken proteins, indicating that the plant protein, in contrast to its animal counterparts, is not a target for calmodulin regulation. In this respect, the Arabidopsis protein is more similar to the recently identified Ipk2 protein from Saccharomyces cerevisiae (Odom et al., 2000), with which it shares 22% identical amino acids (47% similarity). Therefore, the plant and yeast proteins may exhibit similar biological functions. The Ins(1,4,5)P₃ kinase protein from the nematode C. elegans also lacks a calmodulin binding site, although its length (461 amino acids) is more similar to those of the human Ins(1,4,5)P₃ kinase isoforms A and B, both of which contain calmodulin binding sites within their N-terminal domains.

Figure 1.

Alignment of Inositol Polyphosphate Kinase Amino Acid Sequences.

An alignment of inositol polyphosphate kinase proteins from human, nematode (C. elegans), Arabidopsis AtIpk2β, and yeast (S. cerevisiae) is shown. Amino acids that are identical in three or four proteins are shaded in black, and conservative amino acid substitutions are shaded in gray. Gaps are shown by dots. Amino acid residues crucial for binding Ins(1,4,5)P₃ (K) and ATP/Mg²⁺ (R and D), as well as a Trp residue (W) involved in calmodulin (CaM) binding (only in the human protein), are indicated by asterisks.

A series of experiments have demonstrated that the C-terminal 275 amino acids of rat brain Ins(1,4,5)P₃ kinase are sufficient to phosphorylate Ins(1,4,5)P₃, indicating that this region contains the catalytic domain (Takazawa and Erneux, 1991; Togashi et al., 1997). In this enzyme, the charged amino acids Arg-317 and Asp-414 are crucial for ATP/Mg²⁺ binding, as demonstrated by site-directed mutagenesis and chemical modification studies (Communi et al., 1993, 1995a), and Lys-262 is central to InsP₃ binding (Togashi et al., 1997). AtIpk2β possesses a catalytic domain similar to that of known Ins(1,4,5)P₃ kinases, including the amino acids essential for the binding of ATP/Mg²⁺ (Arg-135 and Asp-252) and InsP₃ (Lys-102) (Figure 1). Hydropathy analysis indicated that AtIpk2β is a hydrophilic protein (data not shown), consistent with its presence in a soluble plant cell fraction. Various consensus phosphorylation sites were identified (data not shown). Whether AtIpk2β is a target for protein phosphorylation in vivo remains to be determined.

AtIpk2β Does Not Bind Calmodulin

The AtIpk2β protein sequence lacks an obvious calmodulin binding domain. However, calmodulin binding domains are not highly conserved at the sequence level; therefore, they may escape detection by computational analysis alone. To investigate the calmodulin binding properties of AtIpk2β, two types of protein–protein interaction studies were performed. First, extracts from E. coli cells that express AtIpk2β were passed through a column containing calmodulin-conjugated agarose beads. No binding of AtIpk2β to calmodulin was observed, whereas the recombinant calmodulin binding Arabidopsis GAD1–positive control did bind to the column (data not shown), as described previously (Zik et al., 1998). In addition, a calmodulin-overlay assay was performed. An E. coli protein extract containing recombinant AtIpk2β was separated by SDS-PAGE, transferred to a nitrocellulose membrane, and probed with ³⁵S-labeled calmodulin. Calmodulin did not bind to AtIpk2β but interacted with the NtCBP4 control protein (data not shown), as described previously (Arazi et al., 2000). The inability of calmodulin to bind to AtIpk2β is in accordance with the lack of an obvious calmodulin binding domain.

AtIpk2β Encodes a Dual-Specificity Inositol Polyphosphate Kinase

To confirm that AtIpk2β encodes a functional enzyme, the AtIpk2β cDNA was subcloned into the expression vector pMal-c2 and expressed as a fusion to the E. coli maltose binding protein (MBP). Expression of this MBP-AtIpk2β fusion protein was demonstrated by protein gel blot analysis using an antibody directed against the MBP part of the fusion protein. Figure 2 demonstrates that E. coli cells produced the MBP-AtIpk2β fusion protein, which has a calculated molecular mass of 75 kD, after induction with isopropyl β-d-thiogalactopyranoside (IPTG). The MBP-AtIpk2β fusion protein was not detected in noninduced cells. Similarly, in control cells, MBP, which has a molecular mass of 42 kD, was produced after IPTG induction.

Figure 2.

Protein Gel Blot Analysis of the MBP-AtIpk2β Fusion Protein.

E. coli cells were transformed with either the empty vector pMAL-c2 (lane C) or plasmid pMAL-c2–AtIpk2β (lanes 1 and 2). Protein extracts were obtained from noninduced (−) or IPTG-induced (+) cells. Proteins (45 μg per lane) were detected using an antiserum that recognizes the MBP portion of the proteins. The positions of MBP (42 kD) and the MBP-AtIpk2β fusion protein (75 kD) are indicated by arrows.

Enzyme activity was tested using HPLC analysis (see Methods). Enzyme activity was detected in both crude protein extracts from E. coli expressing the MBP-AtIpk2β fusion protein and in amylose resin–purified fusion protein. By contrast, no activity was detected in extracts from cells transformed with the empty pMal-c2 vector.

Four different ³H-inositol–labeled substrates—inositol 1,3,4- trisphosphate [Ins(1,3,4)P₃], Ins(1,4,5)P₃, Ins(1,3,4,5)P₄, and Ins(1,3,4,5,6)P₅—were incubated either with extracts from bacteria expressing MBP-AtIpk2β or with affinity-purified protein obtained therefrom. Ins(1,4,5)P₃ was converted into products (Figure 3A) that, subjected to HPLC on a Partisphere Strong Anion Exchange (SAX) column with a gradient (Brearley and Hanke, 1996a) that resolves different classes of inositol phosphate product (InsP₃, InsP₄, InsP₅, and InsP₆), were identified as InsP₄ and InsP₅. Among the other potential substrates tested, only Ins(1,3,4,5)P₄ was phosphorylated, yielding an InsP₅ product that coeluted precisely with the InsP₅ product of the phosphorylation of Ins(1,4,5)P₃ (Figures 3A and 3B). Further experiments on this column (described below) indicate that the InsP₅ product eluted with ¹⁴C-Ins(1,3,4,5,6)P₅ (Figure 4A) and before d/l-¹⁴C-Ins(1,2,3,4,5)P₅, d/l-¹⁴C-Ins(1,2,4,5,6)P₅, and Ins(1,2,3,4,6)P₅ (data not shown) (see Brearley and Hanke [1996a][1996b] for the resolution of InsP₅ stereoisomers on this column).

Figure 3.

Analysis of AtIpk2β Enzyme Activity.

Inositol phosphate substrates were incubated with enzyme extracts from E. coli cells expressing the MBP-AtIpk2β fusion protein. The products of the assay were resolved on a Partisphere SAX HPLC column eluted with a gradient derived from buffers A (water) and B (2.5 M NaH₂PO₄) mixed as follows: 0 min, 0% B; 5 min, 0% B; 65 min, 100% B; and 75 min, 100% B.

(A) Ins(1,4,5)P₃ was used as a substrate. Note that InsP₄ and InsP₅ were produced.

(B) Ins(1,3,4,5)P₄ was used as a substrate. Note the synthesis of InsP₅.

No inositol polyphosphate kinase activity was detected in extracts from noninduced E. coli cells (data not shown). Similar results were obtained with affinity-purified protein (data not shown).

Figure 4.

Identification of InsP₄ and InsP₅ Products of MBP-AtIpk2β Action on Ins(1,4,5)P₃.

Inositol phosphate substrates were incubated with affinity-purified MBP-AtIpk2β fusion protein. The products of the assay were mixed with ¹⁴C-labeled standards (detection indicated by the dashed traces) in (A) and (B) and with ³H- and ¹⁴C-labeled standards in (C). Products and standards were resolved on a Partisphere SAX HPLC column eluted with a gradient derived from buffers A (water) and B (2.5 M NaH₂PO₄) as follows: 0 min, 0% B; 5 min, 0% B; 65 min, 100% B; 75 min, 100% B (A); or on an Adsorbosphere SAX HPLC column eluted with a gradient derived from buffers A (water) and B (1 M NH₄H₂PO₄, pH 3.35, with H₃PO₄) mixed as follows: 0 min, 0% B; 120 min, 100% B ([B] and [C]). Radioactivity was monitored by dual-label scintillation counting online (12-s integration interval) or after the collection of 0.5-min fractions (Brearley and Hanke, 1996a, 1996b). Arrows indicate the elution of ³H-Ins(1,4,5)P₃ substrate, ³H-Ins(1,3,4,5)P₄ standard, and various ¹⁴C-labeled standards.

To determine the identity of the InsP₄ product(s) of the phosphorylation of Ins(1,4,5)P₃, we mixed the products of assays with d/l-¹⁴C-Ins(1,4,5,6)P₄ and ¹⁴C-Ins(2,4,5,6)P₄ and repeated the HPLC. Such analysis yielded a broad peak of InsP₄, the tail of which eluted with the internal standard of d/l-¹⁴C-Ins(1,4,5,6)P₄ [note that Ins(1,4,5,6)P₄ and Ins(3,4,5,6)P₄ are enantiomers and therefore cannot be resolved by nonchiral HPLC] but before the ¹⁴C-Ins(2,4,5,6)P₄ standard (Figure 4A). This elution profile suggested the presence of a second InsP₄ reaction product with chromatographic properties not very dissimilar to those of Ins(1,4,5,6)P₄. All subsequent attempts to resolve the InsP₄ peak into more than one component, using a variety of elution conditions, were unsuccessful on the Partisphere SAX column. Nevertheless, inclusion of a standard of ¹⁴C-Ins(1,3,4,5,6)P₅ confirmed the precise coelution of the InsP₅ product with this particular stereoisomer of InsP₅ (Figure 4A). In striking contrast to the results presented in Figures 3A, 3B, and 4A, we observed no phosphorylation of Ins(1,3,4)P₃ in six of six independent assays. In all of these assays, positive controls using Ins(1,4,5)P₃ as a substrate yielded InsP₄ and InsP₅ products in 40 to 90% yield. Additional assays showed that Ins(1,3,4,5,6)P₅ was not a substrate in this assay. These experiments clearly demonstrate that AtIpk2β phosphorylates Ins(1,4,5)P₃ successively, to add phosphates to both the D-6 and D-3 positions. Therefore, AtIpk2β represents a dual-specificity inositol polyphosphate kinase, as described recently for Ipk2p from S. cerevisiae (Odom et al., 2000).

In our attempts to distinguish the nature of the InsP₄ product(s) of AtIpk2β action on Ins(1,4,5)P₃, we used another anion-exchange HPLC column with different selectivity from the Partisphere SAX columns. Separation of the assay products with Ins(1,4,5)P₃ on Adsorbosphere SAX columns (Brearley and Hanke, 1996b) eluted with NH₄H₂PO₄ yielded an InsP₄ product that coeluted precisely with ¹⁴C-Ins(3,4,5,6)P₄ and hence its enantiomeric partner, Ins(1,4,5,6)P₄ (Figure 4B), but after ³H-Ins(1,3,4,5)P₄ when this standard was added in combination with ¹⁴C-Ins(3,4,5,6)P₄ to the reaction products (Figure 4C). Inclusion of a standard of ¹⁴C-Ins(1,3,4, 5,6)P₅ in similar HPLC runs further confirmed the generation of Ins(1,3,4,5,6)P₅ reaction products (data not shown). These key experiments show that AtIpk2β lacks 3-kinase activity against Ins(1,4,5)P₃ but possesses 6-kinase activity against Ins(1,4,5)P₃ and 3-kinase activity against Ins(1, 4,5,6)P₄.

AtIpk2β Complements the Yeast ARG82/IPK2 Mutant

Odom et al. (2000) discovered that the yeast Arg82 protein, a regulator of the transcriptional complex ArgR-Mcm1, is identical to Ipk2p, an InsP₃-InsP₄ kinase (Saiardi et al., 2000). The biochemical studies described here demonstrate that AtIpk2β encodes a dual-specificity inositol polyphosphate kinase. Therefore, it was tempting to speculate that Arabidopsis AtIpk2β can complement the ARG82/IPK2 yeast mutant by forming a functional ArgR-Mcm1–like transcription complex. To test this hypothesis, yeast complementation experiments and a phenotypic assay for ArgR-Mcm1 transcriptional activity were used (essentially according to the protocol of Odom et al. [2000]). To this end, the AtIpk2β coding region was inserted into the yeast expression vector pYES2 and transformed into a ARG82/IPK2 deletion strain. Expression of the Arabidopsis protein in yeast cells was tested by protein gel blot analysis using an antibody raised against AtIpk2β. As is shown in Figure 5, AtIpk2β was detectable in yeast cells only when transformed with the AtIpk2β cDNA (lane 4). A protein of similar size (33 kD) also was detected in crude protein extracts obtained from Arabidopsis leaves (lane 5), indicating that the AtIpk2β cDNA used for yeast complementation contained the complete AtIpk2β coding region.

Figure 5.

Protein Gel Blot Analysis of AtIpk2β Expression in the Yeast ARG82/IPK2 Mutant.

AtIpk2β was detected using an anti-AtIpk2β antibody. Lane 1, wild-type yeast; lane 2, ARG82/IPK2 mutant; lane 3, mutant transformed with the empty vector pYES2; lane 4, mutant transformed with the AtIpk2β cDNA (pYES2-AtIpk2β); lane 5, crude protein extract from Arabidopsis leaves.

As seen in Figure 6, wild-type yeast cells grew at 37°C, whereas ARG82/IPK2 yeast mutant cells did not grow because Arg82p/Ipk2p is essential for cellular growth at this temperature (Dubois and Messenguy, 1994). Growth of mutant yeast cells was restored successfully by AtIpk2β expression but not by transformation with the empty pYES2 vector (Figure 6A). A similar result was obtained when cells were cultured in liquid medium (Figure 6B). Furthermore, both the yeast mutant transformed with AtIpk2β and the wild-type control strain grew at 37°C on medium containing Arg as a sole nitrogen source, whereas the mutant strain and the mutant transformed with the empty vector did not grow under these conditions (data not shown). These results indicate that AtIpk2β forms a functional transcription complex with yeast Arg80-Arg81-Mcm1 proteins in vivo and that the Arabidopsis protein is targeted to the yeast nucleus; otherwise, as shown previously (Dubois and Messenguy, 1994; Odom et al., 2000), no phenotypic complementation would have been observed.

Figure 6.

Arabidopsis AtIpk2β Complements the Yeast ARG82/IPK2 Mutant.

(A) Yeast cells were streaked on synthetic defined minimal medium plus uracil, Leu, His, and Met. Top, the wild type; right, ARG82/IPK2 mutant; left, mutant transformed with the empty vector pYES2; bottom, mutant transformed with the AtIpk2β cDNA (pYES2-AtIpk2β). Cells were incubated for 3 days at 37°C.

(B) Growth curves obtained for yeast cells cultured in liquid medium at 37°C. M, mutant; WT, wild type.

Odom et al. (2000) have shown that Ipk2-mediated production of InsP₄/InsP₅ is required in vivo for the function of the ArgR-Mcm1 complex. Therefore, we concluded that AtIpk2β has the capacity to synthesize InsP₄/InsP₅ in vivo, as demonstrated in vitro with recombinant AtIpk2β protein expressed in E. coli (Figure 3).

AtIpk2β Is a Nuclear Protein

The complementation experiments in yeast suggest that AtIpk2β is targeted to the nucleus. To determine whether AtIpk2β is located in the nucleus in plant cells, we generated a fusion protein between AtIpk2β and GFP and expressed it transiently in tobacco BY2 protoplasts. Fluorescence microscopy revealed the accumulation of GFP fluorescence predominantly in the nucleus (Figures 7A and 7B). A lower but significant amount also was detected in the cytosol. GFP itself has been reported to have a tendency to enter the nucleus, and the same may be true for the AtIpk2β-GFP fusion protein, although it has a much higher molecular mass (60 kD). To further investigate the nuclear localization of AtIpk2β, immunohistochemical studies were performed on Arabidopsis leaf sections using the anti-AtIpk2β primary antibody and fluorescein isothiocyanate (FITC)–conjugated secondary antibody. As shown in Figures 7D and 7F, green FITC fluorescence was observed in the nucleus, as confirmed by 4′,6-diamidino-2-phenylindole staining (Figures 7C and 7E). We tested whether AtIpk2β contains a known nuclear localization signal (NLS) using PSORT prediction software (http://psort.nibb.ac.jp/) and a recently established NLS data set (Cokol et al. 2000). No NLS was predicted for the Arabidopsis protein. These tools also failed to discover an NLS in yeast Arg82p/Ipk2p.

Figure 7.

Localization of AtIpk2β in Plant Nuclei.

(A) and (B) Tobacco BY2 protoplast expressing the AtIpk2β-GFP fusion protein. Nuclei are indicated by arrows. A bright-field image (A) and GFP fluorescence (B) are shown.

(C) to (F) Immunolocalization of the AtIpk2β protein in an Arabidopsis leaf cross-section ([C] and [D]) (epidermal cell layer and underlying mesophyll cells) and in a pair of guard cells ([E] and [F]). Nuclei are indicated by arrows. Nuclei were visualized by 4′,6-diamidino-2-phenylindole staining in (C) and (E). AtIpk2β in nuclei was visualized by an FITC-labeled secondary antibody and fluorescence microscopy.

AtIpk2β Gene Expression

The expression pattern of the AtIpk2β gene was tested by RNA gel blot analysis. A single transcript of ∼1.2 kb was detected in flowers, roots, stems, and leaves (Figure 8A). No significant variation in transcript level was apparent in the tissues tested. AtIpk2β transcript levels of leaves from plants treated briefly with NaCl, abscisic acid, mannitol, water, and drought were not altered greatly (3-h treatments; Figure 8B, top gel), whereas the control gene rd29A responded strongly to these treatments (Figure 8B, bottom gel), as demonstrated previously (Yamaguchi-Shinozaki and Shinozaki, 1993).

Figure 8.

RNA Gel Blot Analysis of AtIpk2β Gene Expression.

(A) RNA isolated from greenhouse-grown Arabidopsis plants. AtIpk2β transcript was detected in various tissues. Lane 1, flower; lane 2, root; lane 3, stem; lane 4, leaf.

(B) RNA isolated from leaves subjected to one of the following treatments (3 h of incubation): untreated control (lane 1); 100 μM abscisic acid (lane 2); water (lane 3); drought (lane 4); 200 mM NaCl (lane 5); and 500 mM mannitol (lane 6). The control gene rd29A was induced under these conditions. A total of 50 μg of RNA was loaded per lane.

To visualize the transcriptional regulation of the AtIpk2β promoter, a 1.1-kb AtIpk2β 5′ regulatory fragment was isolated, fused to the E. coli β-glucuronidase (GUS) reporter gene (AtIpk2β-GUS), and introduced into the Arabidopsis nuclear genome by Agrobacterium tumefaciens–mediated transformation. Twenty-one transgenic lines were screened for GUS activity in young seedlings and plants (Figure 9L), and lines with moderate and strong GUS staining were used for further studies (Figures 9A to 9K). The AtIpk2β-GUS gene was not expressed in immature pollen grains, but it was expressed in mature pollen (Figures 9A to 9C and 9F). In addition, GUS activity was observed in other flower tissues, including the vascular strands (Figure 9D), stigma cells (Figures 9B and 9C), and the abscission zones of fully elongated siliques (Figure 9E). In roots, GUS activity was detected mainly along the central cylinder and the root tip (Figures 9G and 9H). Although AtIpk2β promoter activity was strongest in vascular bundles (Figure 9K), GUS staining generally was seen in other cells as well, including mesophyll cells (Figures 9I and 9J), especially in strongly expressing lines or when incubation times were extended.

Figure 9.

Expression of AtIpk2β-GUS in Transgenic Arabidopsis.

(A) Young flower of a greenhouse-grown plant. Note the absence of GUS staining in pollen grains.

(B) Flower at a more developed stage. Note GUS activity in stigma cells (arrow).

(C) Fully developed flower. GUS activity is present in stigma cells (black arrow). The strongest GUS activity is detected in pollen grains (white arrow) at this stage of flower development.

(D) GUS activity in vascular bundles of the flower.

(E) to (J) Dark-field images.

(E) Fully elongated silique. The arrow indicates strong GUS activity in the abscission zone.

(F) Anther with mature pollen grains.

(G) Root of a 1-week-old seedling. Note strong GUS staining at the root tip and weaker staining in the central cylinder.

(H) GUS activity was observed frequently along the roots.

(I) Cross-section of a stem (8 μm). Strong GUS staining (pink) is seen in the vascular bundle (arrow) and in parenchymatous cells.

(J) Cross-section of a leaf (8 μm). GUS activity (pink) is visible in the vascular bundle (arrow) and in mesophyll cells.

(K) Leaf of a young plantlet 3 days after transfer to soil. Note the strong GUS activity in the vascular tissue.

(L) Collection of several AtIpk2β-GUS lines (indicated by numbers) and a wild-type control (WT).

DISCUSSION

Here, we describe the cloning and characterization of a plant inositol polyphosphate kinase. The AtIpk2β protein shares on average only 25% identical residues with animal Ins(1,4,5)P₃ kinases. However, AtIpk2β possesses a catalytic domain conserved among known Ins(1,4,5)P₃ kinases that includes the Lys-102 residue critical for InsP₃ binding and the Arg-135 and Asp-252 residues critical for ATP and Mg²⁺ binding, respectively. When expressed in E. coli as an MBP-AtIpk2β fusion protein, AtIpk2β shows dual-specificity InsP₃-InsP₄ kinase activity, successively phosphorylating Ins(1,4,5)P₃ at the D6 and D3 positions and Ins(1,3,4,5)P₄ at the D6 position, to yield Ins(1,3,4,5,6)P₅. AtIpk2β is similar to the recently identified yeast enzyme encoded by the IPK2 gene, which functions as a transcriptional regulator, Arg82 (ArgRIII), and also displays dual-specificity InsP₃-InsP₄ kinase activity (Odom et al., 2000). By contrast, mammalian Ins(1,4,5)P₃ kinases phosphorylate the inositol ring only at the D3 position.

AtIpk2β displays not only significant homology with inositol polyphosphate kinases from animals and yeast but also with the animal inositol hexakisphosphate kinase (IP6K) (Schell et al., 1999). Residue Lys-262 of the rat Ins(1,4,5)P₃ kinase is conserved in the IP6K sequence and also appears in the yeast IP6K (KCS1) and ArgRIII proteins. Therefore, this sequence could be a consensus sequence for inositol polyphosphate kinase (Saiardi et al., 1999). As suggested by Schell et al. (1999), IP6K and Ins(1,4,5)P₃ kinase form a distinct family with enzymes that phosphorylate hydroxyl groups on an already partly phosphorylated inositol ring, such as Ins(1,3,4)P₃ 5/6-kinase (Wilson and Majerus, 1997) and Ins(1,3,4,5,6)P₅ 2-kinase (York et al., 1999). We recently cloned another cDNA from Arabidopsis that encodes a protein 73% identical to AtIpk2β. This protein, AtIpk2α, also sequentially phosphorylates Ins(1,4,5)P₃ to generate Ins(1,3,4, 5,6)P₅ (Stevenson-Paulik et al., 2002).

From a metabolic perspective, it is not possible to say at present whether AtIpk2β contributes solely to nuclear inositol phosphate metabolism or makes a more general contribution to the metabolism of inositol in the cell as a whole. Certainly, in yeast, lesions in ARG82/IPK2 manifest as major changes in the profile of inositol phosphates extracted from labeled cells. The relationship of AtIpk2β with pathway(s) of InsP₆ synthesis and turnover remains to be determined. Our cloning of AtIpk2β and demonstration of 3/6-kinase activity of the enzyme potentially affords a molecular genetic explanation for a number of biochemical studies of inositol phosphate kinase activity in plants. Thus, Chattaway et al. (1992) described an activity with Ins(1,4,5)P₃ 6-kinase activity in extracts of pea roots. Similarly, Ins(1,3,4,5)P₄ 6-kinase activity was detected in immature soybean seeds (Phillippy, 1998). Perhaps more significantly, Hatzack et al. (2001) characterized a mutant of barley that shows reduced InsP₆ levels in the grain and increased levels of d- and/or l-Ins(1,3,4,5)P₄. If AtIpk2β constitutes part of a pathway to InsP₆ in plants, then the production of Ins(1,3,4,5,6)P₅, observed in our in vitro assay, but not InsP₆, demands the activity of an Ins(1,3,4,5,6)P₅ 2-kinase such as that described from soybean (Phillippy et al., 1994) and in in vivo experiments (Brearley and Hanke, 1996a, 1996b). Clearly, the contribution of AtIpk2β to InsP₆ biosynthesis remains to be tested in vivo.

Previously, we described a metabolic pathway accounting for InsP₆ accumulation in turions of Spirodela polyrhiza, Ins-Ins3P-Ins(3,4)P₂-Ins(3,4,6)P₃-Ins(3,4,5,6)P₄-Ins(1,3,4,5, 6)P₅- InsP₆, which is distinguished from the potential pathways involving Ins(1,4,5)P₃ discussed above by the addition of the 1-phosphate very late in the metabolic sequence (Brearley and Hanke, 1996a, 1996b). We have further identified Ins(3,4,5,6)P₄ 1-kinase activity in mesophyll cells (Brearley and Hanke, 2000). It seems likely that there are multiple routes to InsP₆ in different cellular locations and in different tissues associated with different functions. Recent work on guard cells ascribes a special function to InsP₆ as a physiological regulator of the inward-rectifying channel (Lemtiri-Chlieh et al., 2000). We can differentiate on a time scale between the accumulation of InsP₆ in storage tissues such as turions of Spirodela or barley grains and the rapid (within minutes) increase of InsP₆ levels in guard cells. It is quite plausible that different pathways operate in these different physiological/developmental processes. The cloning and characterization of AtIpk2β described here affords hitherto unparalleled opportunity to define the contributions of specific kinases to InsP₆ biosynthesis in a variety of contexts.

Mammalian Ins(1,4,5)P₃ kinases can be activated by calmodulin in a Ca²⁺-dependent manner (Yamaguchi et al., 1988; Communi et al., 1994; D'Santos et al., 1994). Their ability to bind calmodulin was exploited to purify Ins(1,4,5)P₃ kinase from mammals (Takazawa et al., 1989; Erneux et al., 1993). Studies have demonstrated that Tyr-165 is necessary for calmodulin binding in rat Ins(1,4,5)P₃ kinase (Erneux et al., 1993). However, AtIpk2β is smaller than animal Ins(1, 4,5)P₃ kinases, and sequence comparisons have revealed that AtIpk2β lacks the consensus calmodulin binding domain(s) located at the N termini of animal Ins(1,4,5)P₃ kinases. Calmodulin-agarose affinity experiments and calmodulin overlay assays performed in the presence of a positive control indicate that AtIpk2β does not bind calmodulin, confirming the sequence information that AtIpk2β does not have a calmodulin binding site. Interestingly, the recently characterized inositol polyphosphate kinases from C. elegans and S. cerevisiae also lack the consensus calmodulin binding site (Clandinin et al., 1998). Apparently, the same is true for the Ins(1,4,5)P₃ kinase from the unicellular green alga Chlamydomonas eugamentos, because its enzymatic activity is not affected by calmodulin (Irvine et al., 1992). Therefore, it is likely that plant, nematode, and yeast inositol polyphosphate kinases are regulated differently than mammalian Ins(1,4,5)P₃ kinases. Our preliminary experiments (data not shown) indicate that AtIpk2β can be phosphorylated by protein kinase C in vitro. Further experiments will be needed to determine whether phosphorylation is physiologically relevant to the regulation of AtIpk2β.

In rat, Ins(1,4,5)P₃ kinase (isoform B) has been demonstrated to exist as a peripheral membrane protein tightly associated with the cytosolic face of the extended endoplasmic reticulum and as a cytosolic protein, and it has been implicated in membrane traffic and Ca²⁺ homeostasis (Soriano et al., 1997). One of the most intriguing observations that we have made is that AtIpk2β has the ability to complement the yeast ARG82/IPK2 deletion mutant. Complementation demonstrated that AtIpk2β affects transcriptional control over Arg metabolism in yeast mutants, because the mutant itself lacks a functional ArgR-Mcm1 transcription complex. Our GFP localization and immunolocalization studies provided further direct evidence for the presence of AtIpk2β in the plant nucleus, as has been shown for yeast Ipk2p. Together, these data indicate a previously unanticipated function for AtIpk2β in gene regulation in higher plants. However, recent studies indicated that this might not be a function of mammalian Ins(1,4,5)P₃ kinase (Dewaste et al., 2000). Arg82p/Ipk2p (ArgRIII) is a transcriptional regulator that also has inositol polyphosphate multikinase activity, which suggests a link between the turnover of inositol phosphates and the regulation of gene expression (Odom et al., 2000; Saiardi et al., 2000). Although the exact functional significance of a nuclear phosphoinositide signaling pathway is not well understood, the increasing evidence from animals and yeast reveals many phosphoinositides, and their corresponding lipid kinases are present in the nucleus (York et al., 1999; Divecha et al., 2000). Furthermore, it has been reported in plants that another enzyme, phosphatidylinositol 3-kinase, is located in the nucleus (Bunney et al., 2000). This work suggests that nuclear phosphoinositide signaling may play an important role in transcriptional control. In light of these results, it will be interesting to identify proteins that interact with AtIpk2β in planta and to reveal target genes that are regulated by such AtIpk2β-modulated transcription complexes.

InsP₃ is well known for its ability to induce the release of Ca²⁺ from internal stores. Therefore, an important question is whether and to what extent AtIpk2β exerts control over Ca²⁺ release in plant cells. Recent evidence suggests that plant nuclei are not passively permeable to Ca²⁺ and that they control the concentration of Ca²⁺ independently of the cytosol (Pauly et al., 2000). It may be rewarding to study the functional relationship between nuclear calcium homeostasis and inositol phosphate metabolism in the future.

The involvement of the InsP₃ signal transduction pathway in plant reproductive processes such as pollen tube inhibition, reorientation, and pollen-mediated self-incompatibility has been demonstrated in great detail (Franklin-Tong et al., 1993, 1996; Malho et al., 1995). Here, we have extended our analysis of AtIpk2β function by analysis of AtIpk2β gene expression monitored with promoter-GUS fusion constructs. GUS activity was detected in various tissues at different stages of plant development but was absent from immature pollen grains, whereas strong GUS activity was observed in mature pollen and stigma cells at a more developed stage. The specific temporal nature of AtIpk2β promoter activity likely implies a function for AtIpk2β in plant reproduction. The foregoing identification of specific inositol phosphate kinase activity and the nuclear localization of the AtIpk2β gene product coupled with the functional complementation of the ARG82/IPK2 yeast mutant provide molecular genetic evidence for inositol phosphate kinase activity that has not been detected previously in higher plants. That plant inositol phosphate kinases might be involved in transcriptional regulation adds considerably to our understanding of inositol phosphate function.

METHODS

Materials

DNA restriction enzymes were purchased from Boehringer Mannheim (Mannheim, Germany) and New England Biolabs (Danvers, MA). Oligonucleotides were obtained from TibMolbiol (Berlin, Germany). DNA sequencing was performed by Seqlab (Göttingen, Germany). α-³²P-dCTP was obtained from ICN (Meckenheim, Germany). Calmodulin-agarose beads were obtained from Sigma (Deisenhofen, Germany). Unless indicated otherwise, other chemicals were purchased from Boehringer Mannheim, Merck (Darmstadt, Germany), or Sigma.

Bacteria and Plants

Escherichia coli strain XL-1 Blue (Stratagene, Heidelberg, Germany) was used for DNA vector construction and cDNA library screening. Agrobacterium tumefaciens strain GV3101 was used for plant transformation. Arabidopsis thaliana C24 plants were grown in a phytotron in soil (GS90; Gebr. Patzer, Sinntal Jossa, Germany) with a 16-h-light (22°C)/8-h-dark (15°C) regimen.

Isolation and Sequence Analysis of AtIpk2β

DNA manipulations were performed using standard protocols described by Sambrook et al. (1989). For the isolation of AtIpk2β genomic DNA, based on a Basic Local Alignment Search Tool (Altschul et al., 1990) search of Arabidopsis sequences, a gene that showed similarity to animal inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] kinase sequences was identified on the P1 clone MAC9. The coding region (∼0.9 kb) of this putative Arabidopsis inositol polyphosphate kinase (AtIpk2β) was amplified by PCR using genomic DNA as a template and the primers 3K-f (5′-GATCGAATTCATGCTCAAGGTCCCTGA-ACACC-3′) and 3K-r (5′-GTCACTCGAGCTAGCGCCCGTTCTCAAGTAGG-3′). These primers introduced, respectively, an EcoRI site at the 5′ end and a XhoI site (underlined) at the 3′ end of the amplified fragment. The PCR product was cloned into vector pCR2.1 (Invitrogen, Leek, The Netherlands), resulting in plasmid pCR-K-2, and sequenced (Seqlab) to confirm its identity.

Isolation of AtIpk2β cDNA

The 0.9-kb genomic DNA fragment described above was used to screen 10⁶ recombinant phages of an Arabidopsis λcDNA library (Stratagene) under stringent hybridization conditions. Purified phage plaques were rescued as pBluescript SK− derivatives by in vivo excision, and cDNA inserts were sequenced. The plasmid with the longest cDNA was named pmIP3K. Computational analysis was performed with the help of the programs of the Wisconsin Genetics Computer Group (GCG package, version 8.1; Devereux et al., 1984). The FASTA (Pearson and Lipman, 1988) and Basic Local Alignment Search Tool (Altschul et al., 1990) search programs were used for sequence comparisons on DNA and amino acid sequences in the GenBank (http://www.ncbi.nlm.nih.gov/), EMBL (http://www.ebi.ac.uk/embl/), and SWISS-PROT (http://www.ebi.ac.uk/swissprot/index.html) databases. Sequence alignments were performed using the BESTFIT program of the GCG package.

Expression of the Maltose Binding Protein–AtIpk2β Fusion Protein in E. coli

The AtIpk2β coding sequence was excised from plasmid pCR-K-2 using EcoRI-XhoI restriction enzymes and subcloned into the expression vector pMAL-c2 (New England Biolabs, Schwalbach, Germany), which had been cut previously with EcoRI and SalI. E. coli cells transformed with the resulting plasmid (pMAL-c2-AtIpk2β) were cultured at 37°C in 10 mL of YT broth medium (8 g of peptone from casein, 5 g of yeast extract, 5 g of NaCl per liter, pH 7.0) supplemented with 50 μg/mL ampicillin. Protein production was induced during the OD₆₀₀ 0.5 to 0.6 phase by the addition of isopropyl β-d-thiogalactopyranoside at a final concentration of 10 mM followed by incubation of the cultures for an additional 3 h. Bacteria were harvested by centrifugation and resuspended in 200 μL of Laemmli buffer (50 mM Tris-HCl, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol) or 1 mL of resuspension buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM DTT, and 0.1% Triton X-100) supplemented with 1 μg/mL leupeptin, 1 μg/mL pepstatin, and 2 μg/mL aprotinin (Boehringer Mannheim). Bacteria were lysed by sonication and centrifuged. Supernatants were analyzed by SDS-PAGE (Laemmli, 1970), assayed for inositol polyphosphate kinase activity or calmodulin binding, or used for further protein purification. Fusion protein was purified at 4°C using amylose-resin according to the manufacturer's instructions (New England Biolabs) and eluted with buffer (20 mM Tris-HCl, pH 7.4, 200 mM NaCl, 10 mM mercaptoethanol, 1 mM EDTA, and 10 mM maltose). Fusion protein purity was >80% as estimated from Coomassie Brilliant Blue R250–stained SDS-PAGE protein gels. Protein concentrations were determined by the Bradford (1976) method using BSA as a standard.

SDS-PAGE and Protein Gel Blot Analysis

Proteins were separated on 10% SDS–polyacrylamide gels. Protein gel blot analysis was performed essentially as described previously (Landschütze et al., 1995). The antiserum raised against maltose binding protein (MBP)–KCO1 (K. Czempinski and B. Mueller-Roeber, unpublished data) was used at a 1:300 dilution in TBST (20 mM Tris-HCl, pH 7.5, 500 mM NaCl, 0.1% Tween 20, and 1% BSA). Alkaline phosphatase–conjugated secondary antibody (anti-rabbit IgG; Promega, Mannheim, Germany) was used at a 1:3000 dilution in TBST. Blots were developed using nitroblue tetrazolium in conjunction with alkaline phosphatase, using 5-bromo-4-chloro-3-indolyl phosphate as a substrate.

To detect AtIpk2β in protein extracts from yeast and plant, a polyclonal antibody against AtIpk2β was raised in rabbits (Dr. J. Pineda Antibody Service, Berlin, Germany) using the MBP-AtIpk2β fusion protein purified from E. coli. Antiserum was purified by IgG affinity purification. Crude protein extracts from yeast of plants were separated (45 μg per lane) on a 10% SDS–polyacrylamide gel. The anti MBP-AtIpk2β antiserum was used at a 1:1000 dilution in TBST. Horseradish peroxidase–conjugated secondary antibody from the SuperSignal West Pico Detection Kit (Pierce, Rockford, IL) was used at a 1:10,000 dilution in TBST. Blots were developed using SuperSignal Substrate (Pierce).

Inositol Polyphosphate Kinase Assay

E. coli cells expressing the MBP-AtIpk2β protein were harvested by centrifugation at 3500g for 10 min at 4°C. The pellet from a 10-mL culture of induced cells was resuspended in 200 μL of resuspension buffer (as described above) and sonicated (six 10-s bursts) on ice with a probe-type sonicator (MSE Soniprep 150; Curtison Matheson Scientific, Houston, TX). The sonicated suspension was centrifuged at 100,000g for 25 min. Assays were started by the addition of 10 μL of supernatant to 10 μL of assay buffer {100 mM Hepes, pH 7.5, 40 mM Glc-6-P in some assays [to guard against phosphatase activity], 20 mM EGTA, 10 mM MgATP, 50 mM mercaptoethanol, 20 μM unlabeled substrate [except Ins(1,3,4,5,6)P₅, see below], and 2 mg/mL BSA}. Four different ³H-labeled substrates—Ins(1,3,4)P₃, Ins(1,4,5)P₃, Ins(1,3,4,5)P₄, and Ins(1,3,4,5,6)P₅—were included in the assay. ³H-Ins(1,3,4)P₃, ³H-Ins(1,4,5)P₃, and ³H-Ins(1,3,4,5)P₄ (21 Ci/mmol) were obtained from DuPont–New England Nuclear. ³H-Ins(1,3,4,5,6)P₅ (of unknown specific activity) was obtained from ³H-inositol–labeled plant tissue (Brearley and Hanke, 1996a, 1996b). Assays were terminated after incubation (45 min at 37°C) by the addition of 4 μL of 70% HClO₄. Supernatants obtained by centrifugation (14,000g for 5 min) were diluted to 1 mL with water and resolved by HPLC on Partisphere (Whatman, Maidstone, UK) or Adsorbosphere Strong Anion Exchange (Alltech Associates, Carnforth, UK) columns (Brearley and Hanke, 1996a, 1996b). ¹⁴C-labeled standards, Ins(3,4,5, 6)P₄ and Ins(1,3,4,5,6)P₅, were obtained directly from ¹⁴C-inositol–labeled plant tissue (Brearley and Hanke, 1996a), whereas others, d/l-Ins(1,4, 5,6)P₄ and Ins(2,4,5,6)P₄, and d/l-Ins(1,2,4,5,6)P₅ and d/l-Ins(1,2,3, 4,5)P₅, were obtained by acid hydrolysis (Brearley and Hanke, 1996a) of ¹⁴C-Ins(3,4,5,6)P₄ and ¹⁴C-Ins(1,3,4,5,6)P₅, respectively.

Calmodulin Binding Assays

Calmodulin-Agarose Affinity Assay

E. coli lysates (50 μL) containing the MBP-AtIpk2β fusion protein were diluted into 1 mL of incubation buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5 mM CaCl₂, and 0.1% Triton X-100) and incubated for 30 min with preequilibrated calmodulin-agarose beads (Sigma). The beads were washed three times in washing buffer (incubation buffer without Ca²⁺) and then eluted in washing buffer with 20 mM EDTA. Aliquots from each step were analyzed by SDS-PAGE.

Calmodulin-Overlay Assay

Recombinant protein samples were separated by SDS-PAGE and electrotransferred to a nitrocellulose membrane using a semidry blotting apparatus. The membrane was stained with Ponceau Red and preincubated for 30 min in calmodulin binding buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 1 mM CaCl₂) containing 1% nonfat dry milk powder. Purified recombinant petunia calmodulin labeled with ³⁵S-Met (Fromm and Chua, 1992) was added to a concentration of 0.2 μg/mL. The membrane was incubated by shaking gently for 16 h at room temperature, washed twice for 10 min with calmodulin binding buffer, dried, and exposed to Kodak X-Omat AR film.

Yeast Complementation and Phenotypic Assay for ArgR-Mcm1 Transcriptional Activity

The AtIpk2β cDNA from plasmid pmIP3K was inserted via BamHI-XhoI into vector pYES2 (Invitrogen) to generate the plasmid pYES2-AtIpk2β. Yeast transformation was performed with the FastTrack Yeast Transformation Kit (Geno Technology, St. Louis, MO). The ARG82/IPK2 deletion strain (derived from BY4741; Mata; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0; YDR173c::kanMX4) was obtained from EUROSCARF (http://www.rz.uni-frankfurt.de/FB/fb16/mikro/euroscarf). Wild- type strain BY4741 (EUROSCARF) was used for control experiments. Yeast transformants were selected on uracil-minus synthetic complete medium (Invitrogen) by growing them at 30°C for 4 days. Yeast complementation experiments were performed on synthetic defined medium supplemented with uracil, Leu, His, and Met. The phenotypic assay for ArgR-Mcm1 transcriptional activity was performed basically as described by Odom et al. (2000). Yeast cells were streaked on synthetic defined minimal medium containing Arg as the sole nitrogen source, supplemented with limiting amounts (20 μg/mL) of nitrogen compounds for which the strain was auxotrophic, and incubated for 3 days at 37°C. For further complementation analysis, a yeast growth curve (OD₆₀₀ over time) was generated. Yeast was cultured in liquid SD medium supplemented with uracil, Leu, His, and Met.

Expression of AtIpk2β–Green Fluorescent Protein Fusion Protein in Tobacco BY2 Protoplasts

The coding region of AtIpk2β was amplified by PCR with oligonucleotide primers 3KGFP-f (5′-CCTCGAGATGCTCAAGGTCCCTGA-ACACC-3′) and 3KGFP-r (5′-ACTAGTGCGCGCCCGTTCTCAAGT-AGG-3′). The primers added XhoI (5′ end) and SpeI (3′ end) restriction sites (underlined in the primer sequences) to the amplified fragment. The PCR product was inserted via XhoI-SpeI sites into vector p35S-GFP-JFH-1 (S65-T) (Hong et al., 1999), producing an in-frame fusion between the AtIpk2β coding region and the green fluorescent protein (GFP) coding region. The AtIpk2β-GFP fusion was inserted downstream of the 35S promoter of Cauliflower mosaic virus in a pUC18 derivative. The final plasmid, IP3K-tGFP, was used for transient expression in tobacco (Nicotiana tabacum) BY2 cells (Negrutiu et al., 1987). Protoplasts were analyzed for AtIpk2β-GFP expression with an Olympus AX70 microscope (Tokyo, Japan).

Immunolocalization of AtIpk2β in Arabidopsis Leaf Sections

Leaves from Arabidopsis C24 plants overexpressing AtIpk2β were excised and fixed for 2 h at room temperature by immersion in 4% paraformaldehyde in a buffer of 50 mM Na-phosphate, pH 7.2. After three washes with PBS (14 mM Na₂HPO₄, 3 mM NaH₂PO₄, and 150 mM NaCl, adjusted to pH 7.4 with 1 M HCl), 30-μm sections were cut using a microtome (Cryostat HM5000, Microm, Walldorf, Germany). The sections were blocked for 1 h in PBS containing 2% BSA, incubated for 1 hr at room temperature with primary antibody raised against the MBP-AtIpk2β fusion protein, and diluted 1:50 in PBS. After three washes, the sections were incubated for 1 h with fluorescein isothiocyanate (FITC)–conjugated goat anti-rabbit IgG (1:50 dilution; Pierce). The samples were washed three times and incubated with 4′,6-diamidino-2-phenylindole. Tissues were examined with an Axiophot fluorescence microscope (Zeiss, Jena, Germany) to visualize 4′,6-diamidino-2-phenylindole and FITC fluorescence. No FITC fluorescence was detected in control experiments, in which the primary antibody was omitted.

RNA Gel Blot Analysis

Total RNA from different Arabidopsis tissues was prepared according to the protocol of Logemann et al. (1987). For induction experiments, young leaves were cut and incubated for 3 h in different conditions: 100 μM abscisic acid, 200 mM NaCl, 500 mM mannitol, water, or drought. RNA (50 μg per lane) was separated electrophoretically on denaturing 15% (v/v) formaldehyde and 1.5% (w/v) agarose gels and blotted onto Hybond-N membranes (Amersham Pharmacia Biotech, Freiburg, Germany). RNA was fixed to the membrane via UV cross-linking (Stratalinker; Stratagene). The 0.9-kb AtIpk2β genomic DNA fragment of plasmid pCR-K-2 was used as a probe. Hybridizations were performed at 62°C. The membranes were washed at the same temperature in SSC/SDS (1× SSC is 0.15 M NaCl and 0.015 M sodium citrate), with a final wash in 0.2× SSC and 0.5% SDS. Blots were exposed to Kodak X-Omat AR films between intensifying screens for 1 to 2 days at −70°C.

Isolation of the AtIpk2β Promoter Fragment and β-Glucuronidase Assay

A 1.1-kb AtIpk2β promoter fragment in front of a start codon was obtained by PCR with the oligonucleotide primers IP3KP-f (5′-CGC-GTCGACCACAAACTTACGTTTTTGAAACGACCC-3′) and IP3KP-r (5′-CTCTAGACTTCTTTGTGGAAGAGTATATATAC-3′). These primers introduced a SalI site at the 5′ end and a XbaI site at the 3′ end of the amplified fragment (underlined). The PCR product was cloned into the TA cloning vector (pCR 2.1) and sequenced. The promoter fragment was excised via SalI-XbaI restriction sites and subcloned into pGPTV-HPT (Becker et al., 1992) carrying the β-glucuronidase (uidA) reporter gene and the nopaline synthase (nos) terminator sequence. The resulting plasmid was first introduced into Agrobacterium by electroporation and then used to transform Arabidopsis by vacuum infiltration (modified from Grant et al., 1995). The transformed plants were screened on selective AM medium (half-concentrated Murashige and Skoog medium [Duchefa, Haarlem, The Netherlands], supplemented with 10% sucrose) and grown subsequently in the greenhouse. β-Glucuronidase activity was analyzed in the F1 and F2 generations of transgenic plants using 5-bromo-4-chloro-3-indolyl β-d-glucuronic acid (Duchefa, Haarlem, The Netherlands) as a substrate (Jefferson et al., 1987). Tissues were vacuum-infiltrated (30 min) and then incubated overnight at 37°C. When required, stained tissue was fixed in hydroxyethylmethacrylate Technovit 7100 (Heraeus-Kulzer, Wehrheim, Germany). Tissue sections (8 μm) were cut with a microtome (Leica Instruments, Nussloch, Germany) and viewed with an Olympus AX70 microscope.

Upon request, all novel materials described in this article will be made available in a timely manner for noncommercial research purposes.

Accession Numbers

The nucleotide sequences presented in this article have been submitted to the GenBank/EMBL Data Bank under accession numbers AJ243592 (AtIpk2β genomic DNA) and AJ245521 (AtIpk2β cDNA). Accession numbers for other sequences mentioned in this article are as follows: AB010069 (Arabidopsis P1 clone MAC9); AJ404678 (AtIpk2α); NP_002211 (human 1D-myo-inositol-trisphosphate 3-kinase A); AF045613 (Caenorhabditis elegans inositol trisphosphate 3-kinase form 3); and NP_010458 (Saccharomyces cerevisiae, Arg82 protein).