INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Perhaps the single most distinctive feature of plant inositol phosphate metabolism is the accumulation of inositol hexakisphosphate (InsP₆)² to levels up to several percent of dry weight in seed or storage tissues (Raboy and Dickinson, 1987) and in vegetative tissues to levels that are likely to be in excess of other inositol phosphates. It is likely that the inositol phosphates commonly found in plants are not related to the signaling molecule Ins(1,4,5) P₃, but are intermediates of the pathways of InsP₆ synthesis and breakdown.

We have described a number of inositol phosphates in the duckweed Spirodela polyrhiza (Brearley and Hanke, 1996a) and in barley aleurone tissue (Brearley and Hanke, 1996c). InsP₄ and InsP₅ species have been identified in mung bean (Stephens, 1990; Stephens et al., 1991) and soybean (Phillippy et al., 1994). Metabolic evidence (Brearley and Hanke, 1996b) suggests that some of those identified in S. polyrhiza are intermediates in InsP₆ biosynthesis. However, as there is some uncertainty surrounding the order of addition of the 1- and 5-Ps (fourth and fifth in the sequence proposed) to the inositol moiety of InsP₆, the validity of the proposed order of the metabolic sequence relies heavily on the nature of the inositol phosphates identified, and, paradoxically, on those present but not yet described. The identification of enzyme activities that phosphorylate endogenous inositol phosphates to products higher in the sequence is of crucial importance, therefore, in distinguishing between possible pathways. Furthermore, as the pathway described in plants differs from that described in Dictyostelium discoideum (Stephens and Irvine, 1990) (notwithstanding a report of an alternative nuclear pathway in this organism; Van der Kaay et al., 1995), while that in animals is unclear, the nature of the pathway operating in plants assumes greater significance because it may shed light on the pathways operating in other kingdoms.

Insofar as the patterns of isomers detected in plants are atypical of animal cells and may be indicative of functions for these compounds specific to plants, we have also set out to identify the range of InsP₃ species present in two experimental systems: frond tissue of the aquatic monocotyledonous plant S. polyrhiza (mesophyll cells predominantly) and root suspension cultures of the dicotyledonous plant Arabidopsis, which as a model experimental system in plant molecular genetics merits attention.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plant Material

Spirodela polyrhiza L. plants were labeled with myo-[2-³H]inositol (21 Ci/mmol, Amersham International, Buckinghamshire, UK) as described previously (Brearley and Hanke, 1996a). Root cell suspension cultures of Arabidopsis (ecotype Landsberg erecta) were obtained from Paul Duprée of the Department of Biochemistry, University of Cambridge. Stock cultures were maintained at 25°C on an orbital shaker in Gamborg's B5 medium (Sigma G-5893, Sigma Chemical, Poole, Dorset, UK) and further supplemented with 20 g/L Glc, 50 mg/L kinetin, 0.5 mg/L 2,4-dichlorophenoxyacetic acid, and 0.5 g/L 2-(N-morpholino)-ethanesulfonic acid (MES), pH 5.7 (22°C). Labeled cultures were initiated by the transfer of a small spatula load of cells from a 7-d-old stock culture to labeling medium comprising Gamborg's B5 basal salt mixture (Sigma G-5768) supplemented as above but also containing 1 mg/L nicotinic acid, 1 mg/L pyridoximine hydrochloride, and 10 mg/L thiamine hydrochloride. Cells 0.26 to 0.34 g fresh weight were removed from a subculture of cells labeled for 4 d in 10 mL of medium containing 20 µCi of myo-[2-³H]inositol. Commelina communis plants were grown from seed. Seed was originally obtained from Prof. T.A. Mansfield, University of Lancaster, UK.

Tissue Extraction

Labeled S. polyrhiza tissue and Arabidopsis suspension cultures were extracted with perchloric acid, neutralized and eluted from a strong anion-exchange HPLC column (Partisphere SAX, Whatman International, Maidstone, UK) with gradients of NaH₂PO₄. Peak fractions were desalted on Dowex AG1 X8 resin (formate form) columns (Brearley and Hanke, 1996a).

Erythrocyte Ghost Treatment of Inositol Phosphates

Desalted inositol phosphates were treated with erythrocyte ghosts under ionic conditions, 12.5 mM-HEPES, pH 7, 10 mM-MgCl₂, and 1 mM-EGTA, favoring the activity of inositol polyphosphate 5-phosphatase (Brearley et al., 1997).

Periodate Oxidation, Reduction, and Dephosphorylation of Inositol Phosphates to Polyols

We have used the periodate oxidation, reduction, and dephosphorylation technique originally devised by Clinton Ballou (Tomlinson and Ballou, 1961) and given a contemporary reading with myo-[³H]inositol labeled substrates (as described by Stephens [1990]) to determine the stereoisomerism of inositol phosphates identified in this study. Periodate attack on the inositol ring cleaves C---C bonds that bear cis-orientated hydroxyl groups. Subsequent reduction with borohydride yields an acyclic polyol phosphate that retains the stereochemistry of substitution of the inositol ring. Removal of phosphates with alkaline phosphatase then yields an acyclic polyol that can be resolved by HPLC and that retains the stereochemistry of substitution. Substitution of the hydroxyl functions of the parent inositol ring with phosphate can thus be seen to protect C---C bonds from periodate attack. Since within a particular class of inositol phosphates (e.g. trisphosphates, of which there are 20 possible non-cyclic stereoisomers) there are several stereoisomers that yield the same polyol, the identification of the parent(s) often relies on additional chromatographic information (i.e. the separation of parent inositol phosphates prior to oxidation).

Inositol phosphates were oxidized with sodium periodate, pH 4.0, reduced, and dephosphorylated (Stephens and Downes, 1990). After reduction, approximately 50 µg each of a mixture of unlabeled polyols: inositol, ribitol, D-arabitol, D-altritol, xylitol, D-glucitol, and L-iditol, was added to act as a carrier through subsequent processing steps. Polyols were resolved on either Polypore Pb²⁺ (Brownlee, Applied Biosystems, Santa Clara, CA) HPLC columns or NH₂ HPLC columns (Microsorb, Woburn, MA). For Polypore HPLC the column was held at 60°C, the injection volume was 30 to 50 µL, and the flow rate was 0.2 mL/min. The NH₂ column was maintained at ambient temperature. The sample was injected in 50 µL of acetonitrile:water (50:50, v/v) and the column eluted with 65% acetonitrile (v/v) in water (Brearley and Hanke, 1992). Throughout HPLC absorbance was monitored at 190 nm. Fractions were collected (see figures for details), and either aliquots or total fractions were counted for ³H and ¹⁴C.

Standards

[³²P]Ins(1,4,5) P₃ and [³²P]Ins(2,4,5) P₃ were prepared by alkaline hydrolysis of PtdIns(4,[³²P]5) P₂, the product of a purified PtdIns4P 5-kinase. Alternatively, [³²P]PtdIns(4,5) P₂ was prepared from [³²P]Pi-labeled turkey erythrocyte lipids. [³²P]Ins(2,4,5) P₃ and traces of [³²P]Ins(3,4,5) P₃ were also obtained by mild acid treatment of [³²P]Ins(1,4,5) P₃. [¹⁴C]Ins(3,4,6) P₃ was obtained from [¹⁴C]inositol-labeled S. polyrhiza (Brearley and Hanke, 1996a). [¹⁴C]Ins(1,2,3) P₃ and D- and/or L-[¹⁴C]Ins(1,2,6) P₃ were prepared by treatment of [¹⁴C]InsP₆ obtained from S. polyrhiza with a commercial preparation (Sigma) of wheat bran phytase essentially according to the method of Stephens (1990).

Permeabilization Experiments

Mesophyll protoplasts were prepared from C. communis and electroporated (Brearley et al., 1997) at a cell density of approximately 2 × 10⁷ cells/mL in 0.3 mL of medium containing 4.5 × 10⁵ dpm of [³H]Ins(3,4,5,6) P₄ obtained from S. polyrhiza (Brearley and Hanke, 1996a), supplemented in some cases with 20 mM MgATP. Glc 6-P was included at a 20 mM concentration to guard against the potential breakdown of added inositol phosphates by unspecified enzyme activities. Protoplasts were incubated for 7 min following electroporation and quenched with perchloric acid.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Multiple Isomers of InsP₃ in S. polyrhiza

Partisphere SAX HPLC of perchloric acid extracts from [³H]inositol-labeled S. polyrhiza resolved three peaks of InsP₃ (Fig. 1). Peak I eluted before [³²P]Ins(3,4,5) P₃, [³²P]Ins(1,4,5) P₃, and [³²P]Ins(2,4,5) P₃; peak II eluted after [³²P]Ins(3,4,5) P₃, with [³²P]Ins(1,4,5) P₃, and before [³²P]Ins(2,4,5) P₃; while peak III eluted after both [³²P]Ins(3,4,5) P₃ and [³²P]Ins(1,4,5) P₃ and with [³²P]Ins(2,4,5) P₃. We have previously shown that peak III contains Ins(3,4,6) P₃ (Brearley and Hanke, 1996a).

View larger version (12K): [in this window] [in a new window]   Figure 1.   Inositol trisphosphates in S. polyrhiza. Inositol trisphosphates from [3H]inositol-labeled S. polyrhiza were separated by HPLC and radioactivity monitored with a flow detector (Radiomatic series A-500, Canberra Packard, Pargbourne, Berks, UK). Separations such as this have been observed in more than 10 experiments.

Ins(1,2,3) P₃ Is Present in Vegetative Tissues of Monocots

The polyol products of periodate oxidation, reduction, and dephosphorylation of peak I were mixed with standards and resolved on Polypore Pb²⁺ HPLC (Fig. 2A). A single major peak of ³H-labeled material was detected and this eluted precisely with [¹⁴C]ribitol and before standards of arabitol, altritol, xylitol, glucitol, and iditol. However, because of the difficulty in obtaining separations of ribitol and inositol on Polypore Pb²⁺ columns, aliquots of peak fractions containing [¹⁴C]ribitol and ³H-label were pooled, freeze-dried, and re-run on an aminopropyl column (Brearley and Hanke, 1992). Under these conditions the ³H-label co-eluted precisely with [¹⁴C]ribitol and before [¹⁴C]inositol. This identifies the parent inositol phosphate as Ins(1,2,3) P₃.

View larger version (20K): [in this window] [in a new window]   Figure 2.   HPLC of polyols derived from InsP3s. Polyols from [3H]InsP3s peaks I, II, and III (Fig. 1) were mixed, respectively, with [14C]ribitol, xylitol, and glucitol (A); [14C]inositol, xylitol, and glucitol (B); or [14C]ribitol, xylitol, and glucitol (C), and resolved by HPLC. Fractions were collected and radioactivity in aliquots was determined. Unlabeled polyol standards, monitored at 190 nm, are indicated with arrows. 3H, ; 14C, . ins, Inositol; rib, ribitol; ara, arabitol; alt, altritol; xyl, xylitol; glu, glucitol; idi, iditol. Repeat analysis of polyols in [3H]InsP3 peaks II and III yielded similar results. Confirmation of the identity of the polyol derived from [3H]InsP3 peak I was provided by subsequent analysis on an aminopropyl column (see text).

Ins(1,4,5) P₃, if Present, Is Only a Minor Component

The identity of the inositol phosphate(s) in peak II was pursued in several ways. It was desalted and co-presented with an authentic standard of [³²P]Ins(1,4,5) P₃ to a preparation of human erythrocyte ghosts, and samples of the reaction products were withdrawn at intervals for HPLC. The results (not shown) revealed that the kinetics of metabolism of the ³H- and ³²P-labeled compounds to products with the chromatographic properties of InsP₂s, were very different. The principal [³H]InsP₂ product of dephosphorylation eluted after [³²P]Ins(1,4) P₂. Although this analysis cannot discount the possibility that the parent ³H peak contains some Ins(1,4,5) P₃, the indication is that the major component is not Ins(1,4,5) P₃. The analysis does suggest, however, that a component in the peak bears a phosphate in the five position.

Identification of Inositol Phosphates Novel to Plants: D- and/or L-Ins(3,4,5) P₃ and D- and/or L-Ins(3,5,6) P₃

Another preparation of [³H]InsP₃ peak II was resolved on Partisphere SAX HPLC, desalted, and applied to an Adsorbosphere SAX HPLC column (Brearley and Hanke, 1996b) on which a single peak of ³H label was resolved. Periodate oxidation, reduction, and dephosphorylation of this peak yielded a major product that co-eluted precisely with xylitol, but after inositol, arabitol, and altritol, and before glucitol and iditol (Fig. 2B). Smaller amounts of ³H label co-eluted precisely with inositol, glucitol, and iditol. We have not determined the enantiomeric identity of the polyols obtained.

Because the parent peak co-eluted with Ins(1,4,5) P₃ but before both Ins(3,4,6) P₃ and Ins(2,4,5) P₃ on Partisphere SAX HPLC [see Stephens and Downes (1990) and Stephens and Irvine (1990) for separation of Ins(1,4,5) P₃ from Ins(3,4,6) P₃, and Chilvers et al., (1991), for separation of Ins(1,4,5) P₃ from Ins(2,4,5) P₃], the production of xylitol identifies the major component of this peak as D- and/or L-Ins(1,5,6) P₃. The glucitol-yielding component is D- and/or L-Ins(1,2,5) P₃, since D- and/or L-Ins(2,4,5) P₃, the only other isomers that yield glucitol (Stephens, 1990), elute after Ins(1,4,5) P₃. Because iditol, the third polyolobtained from the parent InsP₃ peak, is obtained from both D- and/or L-Ins(1,4,5) P₃ and D- and/or L-Ins(3,4,6) P₃, the elution of InsP₃ II before Ins(3,4,6) P₃ (the major peak in Figure 1), leaves D- and/or L-Ins(1,4,5) P₃ as the only possible parents for the iditol obtained. That this minor peak eluted before Ins(3,4,6) P₃, which was present in excess over the others, means that the iditol obtained from the second peak could not be derived from peak III.

D- and/or L-Ins(2,4,5) P₃ Is Present in Plants

We also performed an analysis of peak III in an attempt to identify any minor components in this peak. The peak was desalted, subjected to periodate oxidation, reduction, and dephosphorylation, and the products were resolved on HPLC (Fig. 2C). Among the potential products of oxidation of authentic InsP₃s, four peaks of ³H-polyols were detected. The first eluted just before [¹⁴C]ribitol and is likely to be [³H]inositol. Its presence is either the result of a failure in the oxidation of the InsP₃s in the parent peak (peak III) or is indicative of the presence of Ins(1,3,5) P₃ or Ins(2,4,6) P₃, which are both resistant to periodate oxidation. We are not aware of the identification of Ins(1,3,5) P₃ or Ins(2,4,6) P₃ in either plants or animals.

The second polyol, a minor component, co-eluted precisely with [¹⁴C]xylitol. The third, also a minor component co-eluted precisely with D-[¹⁴C]glucitol. No attempt was made to determine the enantiomerism of the [³H]glucitol product, whereas xylitol and myo-inositol are both meso-compounds. The fourth and major peak co-eluted with iditol. The enantiomerism of this product was characterized previously (Brearley and Hanke, 1996a). The parent inositol phosphate is Ins(3,4,6) P₃.

The presence of a trace of material that eluted with [¹⁴C]xylitol is perhaps more difficult to explain. The only myo-[2-³H]inositol-labeled InsP₃s that yield [³H]xylitol on oxidation are D/L-Ins(1,5,6) P₃ (Stephens, 1990). Xylitol was identified as the major product of oxidation of peak II, consequently the most likely explanation of the presence of [³H]xylitol in the products of oxidation of peak III is the incomplete resolution of peaks II and III on Partisphere SAX HPLC and on subsequent fractionation, i.e. the tailing of peak II into peak III.

The presence of [³H]glucitol in the products of peak III is unlikely to be explained by the tailing of peak II into peak III, since glucitol was only a minor component relative to xylitol in the products of treatment of peak II (Fig. 2B), whereas it was present in approximately equal amounts to xylitol in the products of treatment of peak III. We conclude that the [³H]glucitol derived from peak III is indicative of the presence of D- and/or L-Ins(2,4,5) P₃, which elutes with D/L-Ins(1,4,6) P₃ on Partisphere SAX HPLC (14). D- and/or L-Ins(2,4,5) P₃ has been detected in avian erythrocytes (Radenberg et al., 1989), in which it is one of the major isomers in terms of chemical mass, and possibly in rat mammary tumor cells (Wong et al., 1992).

InsP₃s in Arabidopsis: Identification of Ins(1,2,3) P₃ and D- and/or L-Ins(3,4,6) P₃

When extracts from labeled Arabidopsis cultures were mixed with [¹⁴C]Ins(1,2,3) P₃, D- and/or L-[¹⁴C]Ins(1,2,6) P₃, and [¹⁴C]Ins(3,4,6) P₃, three distinct peaks of [³H]InsP₃ were resolved (Fig. 3A). The first ³H peak co-eluted precisely with [¹⁴C]Ins(1,2,3) P₃ and, given the diagnostic very early-eluting nature of the latter compound on Partisphere SAX HPLC, is likely to be the same. The second peak of ³H label eluted after [¹⁴C]Ins(1,2,3) P₃ and before [¹⁴C]Ins(1,2,6) P₃, while the third co-eluted precisely with [¹⁴C]Ins(3,4,6) P₃ after [¹⁴C]Ins(1,2,6) P₃. In a separate HPLC run (Fig. 3B) the first peak eluted before [³²P]Ins(1,4,5) P₃, [³²P]Ins(2,4,5) P₃, and also before what we assume to be a trace of [³²P]Ins(3,4,5) P₃, which elutes before Ins(1,4,5) P₃ on SAX columns (Stephens and Downes, 1990). The second peak of ³H label co-eluted approximately with [³²P]Ins(1,4,5) P₃ and before [³²P]Ins(2,4,5) P₃, while the third ³H peak eluted after [³²P]Ins(2,4,5) P₃.

View larger version (27K): [in this window] [in a new window]   Figure 3.   Inositol trisphosphates in Arabidopsis root suspension cultures. Extracts from [3H]inositol-labeled Arabidopsis were mixed with standards and resolved by HPLC. Radioactivity in column fractions was determined by dual-label scintillation counting (A: 3H, ; 14C, ), and by on-line scintillation counting (B: 3H, ; 32P, ). The data in B were five-point-smoothed by the flow-detector software. The peaks identified are the internal standards. Separation of three InsP3 peaks was confirmed in an independent experiment.

At this level of analysis the indications are that root cell suspension cultures of Arabidopsis contain InsP₃s with the chromatographic properties of Ins(1,2,3) P₃ and Ins(3,4,6) P₃. The other peak was not identified, but we can exclude Ins(1,2,3) P₃, D/L-Ins(1,5,6) P₃, D/L-Ins(2,4,5) P₃, D/L-Ins(1,2,6) P₃, D/L-Ins(1,4,6) P₃, and probably also D/L-Ins(1,3,4) P₃, which elutes before D/L-Ins(1,5,6) P₃ on SAX columns (Stephens and Downes, 1990), as potential identities.

Ins(3,4,5,6) P₄ 6-Phosphatase and Ins(3,4,5,6) P₃ 4-Phosphatase Activities in Mesophyll Protoplasts

To determine the origins of D- and/or L-Ins(1,5,6) P₃ and D- and/or L-Ins(1,4,5) P₃ components of peak II, and bearing in mind the identification of Ins(3,4,5,6) P₄ in S. polyrhiza (Brearley and Hanke, 1996a), we considered the possibility that D- and/or L-Ins(1,5,6) P₃ [L-Ins(1,5,6) P₃ = D-Ins(3,4,5) P₃] and D- and/or L-Ins(1,4,5) P₃ [L-Ins(1,4,5) P₃ = D-Ins(3,5,6) P₃] might be the product(s) of dephosphorylation of Ins(3,4,5,6) P₃. We permeabilized preparations of mesophyll protoplasts in the presence of [³H]Ins(3,4,5,6) P₄. We rationalized the choice of experimental material in that the inositol phosphates identified in S. polyrhiza are probably those of the frond mesophyll tissue, which represents the major part of this much reduced and highly specialized plant. We prepared mesophyll protoplasts from the leaves of the monocot C. communis from which, unlike S. polyrhiza, the abaxial epidermis is easily stripped, easing the preparation of protoplasts.

Mesophyll protoplasts permeabilized in the presence of [³H]Ins(3,4,5,6) P₄ gave two small peaks of InsP₃ (Fig. 4A). Supplementation of the basic electroporation medium (ATP concentration, 60 µM) with 20 mM MgATP stimulated production of an InsP₅. The generation of InsP₃s in other experiments in which we did not detect phosphorylation of InsP₄ suggests that the InsP₃s detected are the products of direct dephosphorylation of Ins(3,4,5,6) P₄. Of the potential products of direct dephosphorylation of Ins(3,4,5,6) P₄, three [Ins(3,4,5) P₃, Ins(3,5,6) P₃, and Ins(3,4,6) P₃] can be resolved in order of increasing retention time on Partisphere SAX HPLC (Stephens and Downes, 1990; Stephens and Irvine, 1990; Brearley and Hanke, 1996a, 1996b). Ins(4,5,6) P₃ also elutes after Ins(1,4,5) P₃, the enantiomer of Ins(3,5,6) P₃ (Stephens and Downes, 1990). By inclusion on Partisphere SAX HPLC of a standard of [¹⁴C]Ins(3,4,6) P₃, we found (Fig. 4B) that the two InsP₃ products of metabolism of [³H]Ins(3,4,5,6) P₄ eluted before [¹⁴C]Ins(3,4,6) P₃, which eluted before a trace of late-eluting material. The HPLC run also included ¹⁴C-labeled standards of Ins(1,2,3) P₃ and D- and/or L-Ins(1,2,6) P₃. According to the known chromatographic properties of InsP₃s on SAX columns, we identified the first two InsP₃s as Ins(3,4,5) P₃ and Ins(3,5,6) P₃, in order of increasing elution time, and the third as Ins(4,5,6) P₃. The identification of the same InsP₃s, albeit as stereoisomers and not individual enantiomers, in S. polyrhiza provides strong evidence that D- and/or L-Ins(1,5,6) P₃ and D- and/or L-Ins(1,4,5) P₃ found therein are products of metabolism of Ins(3,4,5,6) P₄. The foregoing also explains the origins of D- and/or L-Ins(1,4,5) P₃ in S. polyrhiza and suggests that the isomer might be the L-enantiomer [D-Ins(3,5,6) P₃].

View larger version (16K): [in this window] [in a new window]   Figure 4.   Metabolism of Ins(3,4,5,6) P4 by permeabilized mesophyll protoplasts. Extracts from protoplasts permeabilized with [3H]Ins(3,4,5,6) P4, with or without supplemental ATP, were resolved by HPLC (A). The traces are normalized with respect to the InsP4 peak. HPLC traces of extracts mixed with [14C]InsP3 and InsP5 standards are shown in B and C. 3H, ; 14C, . The peaks identified are the internal standards. Phosphorylation of Ins(3,4,5,6) P4 to InsP5 was repeated in two additional experiments. Two peaks of InsP3 were generated in all four experiments.

Ins(3,4,5,6) P₄ 1-Kinase Activity in Mesophyll Protoplasts

The InsP₅ product of metabolism of [³H]Ins(3,4,5,6) P₄ was chromatographed (Fig. 4C) against internal standards of [¹⁴C]InsP₅, as described previously (Brearley and Hanke, 1996a, 1996b). The order of elution of InsP₅s on Partisphere SAX columns is Ins(1,3,4,5,6) P₅ < D/L-Ins(1,2,3,4,5) P₅ < Ins(1, 2, 3, 4, 6) P₅ < D/L-Ins(1,2,4,5,6) P₅ (Stephens et al., 1991). Thus, the precise co-elution of the [³H]InsP₅ product with an internal standard of [¹⁴C]Ins(1,3,4,5,6) P₅ and before standards of D/L-[¹⁴C]Ins(1,2,3,4,5) P₅ and D/L-[¹⁴C]Ins(1,2,4,5,6) P₅ identifies the product as Ins(1,3,4,5,6) P₅ and so defines Ins(3,4,5,6) P₄ 1-kinase activity in mesophyll protoplasts. This provides a convincing explanation of the metabolic origin of Ins(1,3,4,5,6) P₅ in plants and S. polyrhiza in particular.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Inositol Trisphosphates in Plants

Detailed analysis of inositol trisphosphates in two plants has revealed a complex array of isomers. None of the isomers detected is unique to plants though D/L-Ins(2,4,5) P₃ and Ins(1,2,3) P₃ have only occasionally been reported in vivo in any kingdom (Brearley and Hanke, 1992, 1996b; Barker et al., 1995). Although Ins(1,4,5) P₃ is one of only a few InsP₃s identified to date, it appears that this isomer is only a minor component of the inositol phosphate complement of higher plants.

A schematic diagram of the metabolic relationships of InsP₃ species identified in mesophyll cells is indicated in Figure 5. In this context, it is worth remembering that the storage tissue of the turion of S. polyrhiza is a specialized form of the mesophyll of the frond.

View larger version (23K): [in this window] [in a new window]   Figure 5.   Inositol phosphates and a scheme for their metabolism in mesophyll cells. Steps for which there is direct metabolic evidence are indicated with solid arrows. Reactions that have been demonstrated in other experimental systems in vitro are indicated with dashed arrows. Asterisks (*), InsP3s common to monocots and dicots.

Metabolic Origins of Inositol Trisphosphates in Plants: InsP₃s as Products of Catabolism

The existing literature provides clues to the origins of InsP₃s in plants. The work of Johnson and Tate (1969), Lim and Tate (1971, 1973), and Tomlinson and Ballou (1961, 1962) reviewed by Cosgrove (1980), suggests that both Ins(1,2,3) P₃ and Ins(1,2,6) P₃ are products of InsP₆ breakdown in vitro. The identification of these two isomers in aleurone tissue and of Ins(1,2,3) P₃ in S. polyrhiza and Arabidopsis in the present study suggests that these isomers are products of InsP₆ metabolism (cytosolic or not) in vivo.

If D- and/or L-Ins(1,5,6) P₃ in S. polyrhiza turns out to be the L-enantiomeric form [D-Ins(3,4,5) P₃], then an explanation of the origin of this compound is provided by our demonstration of metabolism of Ins(3,4,5,6) P₄ to Ins(3,4,5) P₃ in mesophyll protoplasts. Other investigators (Radenberg et al., 1989) have speculated that the D- and/or L-Ins(1,5,6) P₃, which they identified in avian erythrocytes, may be a product(s) of dephosphorylation of Ins(3,4,5,6) P₄ and Ins(1,3,4,5) P₄. Our approach affords an experimental test of this. Whereas Ins(3,4,5,6) P₄ is present in higher plants, there is no evidence yet for Ins(1,3,4,5) P₄.

Finally, in consideration of the range of InsP₃ isomers in plants, it is apparent that Ins(1,2,3) P₃ is common to monocots (S. polyrhiza and barley) and dicots (Arabidopsis). It occurs in non-highly differentiated cells (mesophyll cells) in the fronds and turions of S. polyrhiza, in root cortex cells of Arabidopsis, and in terminally differentiated cells in barley aleurone. It is likely that the route of Ins(1,2,3) P₃ generation is shared in monocots and dicots and perhaps also in animal cells (Barker et al., 1995), where it has been shown that Ins(1,2,3) P₃ is a product of InsP₆ breakdown by cell homogenates. Clearly, Ins(1,2,3) P₃ is not restricted to specialized cells or storage tissues in plants, so we should consider a general "housekeeping" function for this isomer. One such suggestion is that Ins(1,2,3) P₃ is an inhibitor of hydroxyl free radical formation (Barker et al., 1995; Phillippy and Graf, 1997).

InsP₃s as Intermediates in Synthetic Pathways

Ins(3,4,6) P₃ is one of the few InsP₃s for which there is direct evidence in vivo of the identity of its metabolic neighbors. In avian erythrocytes, Ins(3,4,6) P₃ is the precursor of Ins(3,4,5,6) P₄, which in turn is the precursor of Ins(1,3,4,5,6) P₅ (Stephens and Downes, 1990). A caveat is necessitated by the possibility, however hypothetical, of substrate cycles involving these inositol phosphates, which could compromise the interpretation of non-equilibrium labeling studies. The operation of substrate cycles for these specific isomers has not been tested directly or indirectly. Ins(3,4,6) P₃ is also an intermediate in a pathway of InsP₆ biosynthesis in D. discoideum (Stephens and Irvine, 1990).

In plants (Brearley and Hanke, 1996b), Ins(3,4,6) P₃ is an intermediate in a biosynthetic sequence to InsP₆, which shares the partial sequence Ins(3,4,6) P₃  Ins(3,4,5,6) P₄  Ins(1,3,4,5,6) P₅, which has been described in avian erythrocytes (Stephens and Downes, 1990) and may represent steps in a route of synthesis of InsP₆ in the animal kingdom. An alternative possibility (Shears, 1996) places Ins(1,3,4) P₃ at a branch point in inositol phosphate metabolism leading either to inositol through the action of inositol phosphate phosphatases or to InsP₆ via Ins(1,3,4,6) P₄ and Ins(1,3,4,5,6) P₅. Thus, the recent cloning of human (Wilson and Majerus, 1996) and Arabidopsis (Wilson and Majerus, 1997) genes whose products, tested against a limited number of substrates, show Ins(1,3,4) P₃ 5/6-kinase activity in vitro is particularly interesting.

Given the lack of consensus on the route(s) of InsP₆ synthesis in animals, plants and D. discoideum, our demonstration that Ins(3,4,5,6) P₄ is a substrate for Ins(3,4,5,6) P₄ 1-kinase activity in mesophyll protoplasts of monocots is entirely consistent with the earlier proposal that in plants the 1-P of InsP₆ is added after the 4- and 5-Ps (Brearley and Hanke, 1996b). Because the 3-P and not the 1-P is added first, such observations discount the possibility that Ins(1,4,5) P₃ or Ins(1,3,4) P₃ are precursors of InsP₆ in S. polyrhiza. We found no evidence for the presence of Ins(1,4,5) P₃-kinase activity in mesophyll protoplasts (Brearley et al., 1997) under similar experimental conditions to those described here.

Moreover, an inositol polyphosphate 5/6-kinase has been cloned from Arabidopsis (H.W. Xue, C.A. Brearley, and B. Mueller-Roeber, unpublished data) that is identical to that previously reported in Arabidopsis (Wilson and Majerus, 1997). The product of our gene shows Ins(3,4,6) P₃ 5-kinase activity, consistent with the precursor-product relationship of Ins(3,4,6) P₃ and Ins(3,4,5,6) P₄ in a pathway to InsP₆ (Brearley and Hanke, 1996a, 1996b), and which, considering the identification of D- and/or L-Ins(3,4,6) P₃ in Arabidopsis, might suggest a common pathway in monocots and dicots. It is, however, quite possible that there are multiple routes to InsP₆ in plants. Phillippy (1998) has recently identified separate inositol 1,3,4-trisphosphate 5-kinase and inositol 1,3,4,5-tetrakisphosphate 6-kinase activities in immature soybean plants.