Myo-inositol hexakisphosphate, isolated from female gametophyte tissue of loblolly pine, inhibits growth of early-stage somatic embryos

Authors

  • Di Wu,

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author
  • M. Cameron Sullards,

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author
  • Charlie D. Oldham,

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author
  • Les Gelbaum,

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author
  • Jacob Lucrezi,

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author
  • Gerald S. Pullman,

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author
  • Sheldon W. May

    1. School of Chemistry and Biochemistry, Petit Institute for Bioengineering and Biosciences, School of Biology, and Institute of Paper Science and Technology, Georgia Institute of Technology, Atlanta, Georgia 30332, USA
    Search for more papers by this author

Author for correspondence:
Sheldon W. May
Tel: +1 404 894 4052
Email: sheldon.may@chemistry.gatech.edu

Summary

  • Myo-inositol hexakisphosphate (InsP6), abundant in animals and plants, is well known for its anticancer activity. However, many aspects of InsP6 function in plants remain undefined. We now report the first evidence that InsP6 can inhibit cellular proliferation in plants under growth conditions where phosphorus is not limited.
  • A highly anionic molecule inhibitory to early-stage somatic embryo growth of loblolly pine (LP) was purified chromatographically from late-stage LP female gametophytes (FGs), and then characterized structurally using mass spectrometry (MS) and nuclear magnetic resonance (NMR) analyses.
  • Exact mass and mass spectrometry-mass spectrometry (MS-MS) fragmentation identified the bioactive molecule as an inositol hexakisphosphate. It was then identified as the myo-isomer (i.e. InsP6) on the basis of 1H-, 31P- and 13C-NMR, 1H-1H correlation spectroscopy (COSY), 1H-31P heteronuclear single quantum correlation (HSQC) and 1H-13C HSQC. Topical application of InsP6 to early-stage somatic embryos indeed inhibits embryonic growth.
  • Recently evidence has begun to emerge that InsP6 may also play a regulatory role in plant cells. We anticipate that our findings will help to stimulate additional investigations aimed at elucidating the roles of inositol phosphates in cellular growth and development in plants.

Introduction

Myo-inositol-1,2,3,4,5,6-hexakisphosphate (InsP6), also called phytic acid, is myo-inositol with phosphate groups attached to each of its six carbons. myo-Inositol is a member of the B vitamins, that play a central role in growth and development (Abel et al., 2001). InsP6 is ubiquitous in animal and plant cells; it is the major form of phosphorus in seeds and also accumulates in other plant tissues and organs, including pollen, roots, tubers and turions (Cosgrove, 1980; Sasakawa et al., 1995; Raboy, 1997; Loewus & Murthy, 2000; Abel et al., 2001). It can represent from 1% to several per cent of a typical seed’s DW, c. 75 ± 10% of a seed’s total phosphorus, and normally > 90% of a mature seed’s total, acid-extractable inositol phosphates (Raboy, 1997). In mammalian cells, total cellular InsP6 concentration ranges from 10 to 100 μM, depending on cell type and developmental stage (Szwergold et al., 1987; Sasakawa et al., 1995; Shears, 2001).

Myo-inositol-1,2,3,4,5,6-hexakisphosphate was historically considered as an antinutrient in that it can decrease the bioavailability of essential dietary minerals and proteins in monogastric animals (chickens, swine, humans) as a result of its ability to bind them (Urbano et al., 2000; Palacios et al., 2007). Monogastric animals are unable to digest InsP6; their excreted InsP6 is transferred to surface waters via rain, drainage, surface runoff and wind erosion. Excreted phosphorus contributes to cyanobacterial blooms, hypoxia, and death of aquatic animals (Brinch-Pedersen et al., 2002; Turner et al., 2002; Vats et al., 2005).

In the late 1980s, InsP6 was shown to possess striking anticancer action. InsP6 inhibited the growth of all tested cell lines in a dose- and time-dependent manner (Shamsuddin & Vucenik, 2005). Reduction of the elevated rate of cell proliferation to normal values has been observed both in intact animals (Wattenberg, 1995; Challa et al., 1997; Gupta et al., 2003) and in cultures of human malignant cells (Shamsuddin et al., 1992; Sakamoto et al., 1993; Shamsuddin & Yang, 1995; Ferry et al., 2002; Singh et al., 2003). In terms of plants, InsP6 has been investigated for its effect on plant growth as a source of phosphorus under conditions where phosphorus is limited (Hayes et al., 2000; Richardson et al., 2007). However, reduction of proliferation in plant cells by InsP6, under conditions where phosphorus is not limited, has not been reported to date.

Loblolly pine (LP, Pinus taeda L.), a species of the pine family, constitutes the primary commercial species in the southern forests of the US covering 13.4 million ha. Somatic embryogenesis (SE) is a vegetative propagation system currently used to reproduce whole LP and other plants from suspension cultures (Jain et al., 1995; Pullman & Johnson, 2002; Pullman et al., 2003a). Well-controlled SE cannot only maintain the desirable genetic composition of the progeny, but can also improve the efficiency of propagation (Xu et al., 1997; Cairney et al., 2000). Unlike angiosperm embryos with attached cotyledons as seed storage organs, the diploid conifer embryo is surrounded by the unattached haploid female gametophyte (FG). This FG tissue is absent in LP embryogenic tissue culture.

In previous work, we have shown that extracts from early-stage FGs stimulate growth and multiplication of early-stage somatic embryos, whereas water extracts from late-stage FGs contain substance(s) inhibitory to early-stage somatic embryo growth (De Silva et al., 2008). We now report the identification of the inhibitory substance as InsP6 on the basis of liquid chromatography-mass spectrometry (LC-MS), liquid chromatography-mass spectrometry-mass spectrometry (LC-MS-MS), exact mass, and one- and two-dimensional 1H, 31P, and 13C NMR analyses. Our findings constitute the first evidence that InsP6 inhibits cell proliferation in plants under growth conditions where phosphorus is not limited. Many aspects of the function of InsP6 in plants have remained undefined (Turner et al., 2002; Raboy, 2003). We therefore anticipate that our results will help to stimulate additional investigations into the roles played by inositol phosphates in cellular growth and development in plants.

Materials and Methods

Chemicals

Tris base, deuterium oxide (min 99.96 atom% D), Dowex resin 50WX8-200 and phytic acid dodecasodium salt hydrate from rice were obtained from Sigma (St. Louis, MO, USA). Phytic acid water solution (50% w/w) was obtained from Chromadex (Chromadex, Irvine, CA, USA). High-performance liquid chromatography-grade acetonitrile, sodium chloride, sodium phosphate monobasic and dibasic, ammonium acetate, and formic acid were purchased from Fisher Scientific (Pittsburgh, PA, USA). Muco-inositol hexakisphosphate (muco-InsP6) was generously gifted by Dr Alan Richardson from the collection of the late Dr Dennis Cosgrove. There is some uncertainty as to the barium stoichiometry in this muco-InsP6 material, and because of the very limited quantity available to us we were unable to carry out any experiments to resolve this uncertainty. The range of muco-InsP6 concentrations tested was such that, even if the barium stoichiometry in the sample was as low as zero or as high as six, the tested concentrations spanned the range over which InsP6 itself exhibits inhibitory activity.

Female gametophyte collection and water extraction

Loblolly pine cones were collected weekly throughout the sequence of embryo development from multiple open-pollinated mother trees over 10 yr. Seeds from Union Camp tree UC5-1036, located in a seed orchard near Bellville, GA, were collected during the years 1996–1998. Seeds from grafts of tree S4PT6 were collected in Louisiana in 1997 and 1998 and from Texas in 2002, 2003 and 2005. Cones were shipped on ice and received within 24–48 h of collection and processed within 2 wk of receipt. Cones were pried open, seeds were isolated, and FGs were collected and stored under liquid nitrogen. FG collection and water extraction were done using procedures described previously (Pullman & Buchanan, 2006; De Silva et al., 2008). Extracts were obtained from late-stage FGs (Stages 9.10–9.12). The resulting water extract solution was filtered with a 0.2 μm syringe filter (Pall, East Hills, NY, USA), and lyophilized overnight to dryness. In some experiments, FGs were obtained from full-term dry seeds from tree S4PT6 of year 2006; no differences in fractionation or bioactivity profiles were observed between FG extracts from these seeds and those from Stage 9.10–9.12 seeds.

Early-stage somatic embryogenic multiplication bioassay

A previously described staging system (Pullman & Webb, 1994) was used to evaluate morphological development in zygotic and somatic embryos. Somatic embryos at stage 2 were isolated by forceps from suspension culture and placed on 2 ml of multiplication medium 1250 contained in 24-well plates. Forty replicate bioassays were then carried out, for each fraction or control treatment tested, using the following protocol. Each FG extract fraction, corresponding to 20 FGs in our general bioassay protocol, was ultrafiltered, lyophilized, dissolved in 2 ml of deionized water, adjusted to pH 5.7, sterilized with a 0.2 μm syringe filter, and topically applied to the stage 2 somatic embryos at 50 μl per well. Embryos were grown in the dark at 23–25°C and after 4–7 wk the diameter of the embryogenic tissue was measured with a dissecting microscope using a calibrated eyepiece reticle. Typically a single embryo, c. 1 mm in size, grows into a mass of multiple embryos c. 5–9 mm in diameter depending on culture genotype, medium contents, and time. All the data were evaluated by multifactor analysis of variance. The significant differences between means of each treatment were determined by the multiple range test at 95% level of significance. Both analyses were performed using Statgraphics Plus Version 4.0 (Manugistics, Rockville, MD, USA). This bioassay system was used to study somatic embryogenesis in our earlier research (Pullman & Johnson, 2002; Pullman et al., 2003c, 2006; De Silva et al., 2008).

Liquid chromatography isolation

All chromatographic steps described here were carried out at 4°C. All collected fractions described here from a given column were pooled based on the absorbance monitored at 215, 254 and 280 nm. A Superdex 75 10/300 GL gel filtration column (10 × 300 mm, GE Healthcare, Piscataway, NJ, USA) was pre-equilibrated with 150 mM NaCl for 118 min at 0.5 ml min−1. Lyophilized powder corresponding to the water extract of 30 FGs was dissolved in 250 μl of 150 mM NaCl, centrifuged at 16 000 g for 10 min, and the supernatant was applied to this column and eluted at a flow rate of 0.5 ml min−1 using 150 mM NaCl. Fractions collected in multiple tubes (0.5 ml per tube) were pooled with the following designations: F1, from 15.0 to 27.0 min (6.0 ml); F2, from 27.0 to 34.0 min (3.5 ml); F3, from 34.0 to 55.0 min (10.5 ml); and F4, from 59.0 to 77.0 min (9.0 ml). All fractions were ultrafiltered (Millipore YM1 membrane, 1000 Da cutoff) and the retentates lyophilized. Fractions were bioassayed as described earlier.

A Mono S strong cation exchange column (5 × 50 mm, GE Healthcare) was used for the second chromatography step. Bioactive fraction ‘F2’ (i.e. the fraction eluting from the previous Superdex 75 column from 27.0 to 34.0 min) was dissolved in 300 μl of 20 mM sodium phosphate buffer, pH 2.5 (eluent A), applied to the Mono S column, and eluted at 1.5 ml min−1. The binary gradient elution was monitored spectrophotometrically. Eluent B was eluent A containing 2 M NaCl. The gradient started at 0% B for 9 min, 0– 50% B over 27 min, 50–100% B for 2 min, and 100% B for 18 min. Fractions collected in multiple tubes (1.0 ml per tube) were pooled with the designations: F2-S1, from 0.7 to 2.0 min (2.0 ml); F2-S2, from 2.0 to 3.3 min (2.0 ml); F2-S3, from 13.3 to 21.3 min (12.0 ml); F2-S4, from 21.3 to 27.3 min (9.0 ml); and F2-S5, from 37.5 to 41.5 min (6.0 ml). All fractions were ultrafiltered (YM1) and the retentates lyophilized. Fractions were bioassayed as described earlier.

The third chromatography was achieved on a Mini Q strong anion exchange column (4.6 × 50 mm, GE Healthcare). The bioactive fraction ‘F2-S1’ (i.e. the fraction eluting from 0.7 to 2.0 min on the previous Mono S column) was dissolved in 300 μl of 20 mM Tris-Cl buffer at pH 8.0 (eluent A) and loaded on the pre-equilibrated Mini Q column. Eluent B was eluent A containing 2 M NaCl. The flow rate was 0.5 ml min−1, and the elution profile was monitored spectrophotometrically. A binary gradient was applied: 0% B for 23 min, 0–30% B over 58 min, 30–100% B for 4 min, and 100% B for 4 min. Fractions collected in multiple tubes (0.5 ml per tube) were pooled with the following designations: F2-S1-Q1, from 2.1 to 6.1 min (2.0 ml); F2-S1-Q2, from 6.1 to 11.1 min (2.5 ml); F2-S1-Q3, from 25.4 to 30.4 min (2.5 ml); F2-S1-Q4, from 30.4 to 36.4 min (3.0 ml); F2-S1-Q5, from 36.3 to 43.3 min (3.5 ml); and F2-S1-Q6, from 49.4 to 58.3 min (4.5 ml). Each fraction was ultrafiltered (Millipore YM3 membrane, 3000 Da cutoff) and the retentates were bioassayed as described earlier.

Final purification was accomplished on a Superdex Peptide 10/300 GL gel filtration column (10 × 300 mm, GE Healthcare) pre-equilibrated with 100 mM ammonium acetate buffer, pH 5.5, at 0.5 ml min−1 for 138 min before sample loading. The lyophilized (without ultrafiltration) bioactive fraction ‘F2-S1-Q6’ (i.e. the fraction that had eluted from 49.4 to 58.3 min on the previous Mini Q column) was dissolved in 150 μl of the same buffer, applied to the column, and eluted at a flow rate of 0.2 ml min−1 for 158 min. The elution was monitored spectrophotometrically. Fractions collected in multiple tubes (0.2 ml per tube) were pooled with the following designations: F2-S1-Q6-D1, from 62.1 to 80.1 min (3.6 ml); F2-S1-Q6-D2, from 80.1 to 90.1 min (2.0 ml); F2-S1-Q6-D3, from 90.1 to 95.1 min (1.0 ml); F2-S1-Q6-D4, from 95.1 to 98.1 min (0.6 ml); and F2-S1-Q6-D5, from 98.1 to 100.1 min (0.4 ml). All fractions were lyophilized and bioassayed as described earlier. The bioactive fraction ‘F2-S1-Q6-D2’ (i.e. the fraction eluting from 80.1 to 90.1 min) was further characterized by LC-MS and NMR. It should be noted that in our hands InsP6 behaves in dialysis at low ionic strength as if it had a higher molecular mass than predicted from its formula. Similar observations have been reported by others (Hanakahi et al., 2000; Irigoin et al., 2002), and this phenomenon has been studied in detail by Van der Kaay & Van Haastert (1995).

LC-MS and LC-MS-MS characterization under positive polarity

The mass spectrometer was an Applied Biosystems/MDS SCIEX 4000 QTRAP (Linear Ion Trap Quadrupole) LC-MS-MS system (Applied Biosystems, Foster City, CA, USA). The electrospray ionization (ESI) interface was operated in positive mode at 5500 V. A Vydac C8 capillary column (5 μ, 300 Å, 0.3 × 250 mm, Deerfield, IL, USA) was coupled to a Shimadzu LC-10ADVP. Full-scan mass spectra were recorded from m/z 500 to 2000 amu at 1000 amu s−1 by the linear ion trap as an enhanced mass spectrometric (EMS) scan, with declustering potential at 90 V. The total flow rate was 0.6 ml min−1, which was split to 5 μl min−1 to the column. Mobile phase A consisted of 2% acetonitrile and 98% water, and mobile phase B was 80% acetonitrile and 20% water, both containing 0.1% formic acid as the ion pair reagent. The elution protocol was a 1 min column pre-equilibration with 100% A, followed by a 10.0 μl sample injection, a 5 min sample load and washing with 100% A. The gradient was 105 min in total: 0% B for 12 min, 0–40% B over 63 min, 40–100% B for 5 min, 100% B for 8 min, back to 0% B for 5 min, and re-equilibration for 12 min.

Liquid chromatography-mass spectrometry-mass spectrometry spectra of the precursor ion at m/z 661 (obtained from F2-S1-Q6 ultrafiltered against an YM3 3000 Da cutoff membrane) under positive polarity were acquired subsequent to the LC-MS on the QTRAP. Enhanced product ion (EPI) scans were performed subsequent to the EMS scan at a scanning rate of 1000 amu s−1 over the mass range m/z 50–700 with collision energies at both 40 and 80 eV. Data acquisition and instrument control were performed by Analyst software (version 1.4.2).

Exact mass and MS-MS analyses under negative polarity

Exact mass analysis under negative polarity was obtained from the final active fraction from the Superdex Peptide column (F2-S1-Q6-D2).

Exact mass measurement under negative ion polarity was conducted on an Applied Biosystems/MDS SCIEX QSTAR® XL (hybrid quadrupole time-of-flight mass spectrometer, TOF-MS) LC-MS system, with F2-S1-Q6-D2 and phytic acid standard directly infused with an automated chip-based nanoelectrospray TriVersa NanoMate system (Advion, Ithaca, NY, USA). The NanoMate was controlled by the ChipSoft software (version 7.1.1). The ESI chip used was D-chip (4.1 μm ID). A low delivery pressure of 0.15 psi of nitrogen gas and a voltage of −1.6 kV were applied to generate a nanoelectrospray plume from a given nozzle when infusing the sample of interest. TOF-MS spectra were recorded over the mass range of m/z 300–2000. The declustering potential was held at −150 V and the focusing potential at −200 V. Data acquisition and QSTAR instrument control were performed by Analyst QS software (version 1.1).

Tandem mass spectra of the precursor ion at m/z 659 under negative polarity were acquired subsequent to the exact mass measurement on the QSTAR XL interfaced with the NanoMate system. The spectra were recorded from m/z 50–700 with collision energies at both 40 and 80 eV.

InsP6 standard preparation for NMR characterization

Phytic acid dodecasodium salt hydrate from rice (300.2 mg) was converted to InsP6 using the cation exchange resin Dowex 50WX8-200 hydrogen form. The eluate was collected, divided into 10 aliquots and lyophilized to dryness. Each aliquot of the lyophilized product was redissolved in 500 μl of deuterium oxide (min. 99.96 atom% D) and the pH adjusted to 8.22 by addition of concentrated NaOD.

One-dimensional NMR characterization

Nuclear magnetic resonance spectra of the final purified active fraction (F2-S1-Q6-D2) were recorded in D2O at 500.13 MHz for 1H with 32 scans, 202.46 MHz for 31P with 64 scans and 125.77 MHz for 13C with 16 386 scans using a multinuclei probe maintained at 280 K on a Bruker DRX 500 spectrometer. F2-S1-Q6-D2 was dissolved in 500 μl of D2O (min. 99.96 atom% D) and lyophilized to dryness, repeated three times. The final lyophilized product was redissolved in 200 μl D2O (min. 99.96 atom% D) and transferred to a symmetrical 5 mm NMR tube (Shigemi, Allison Park, PA, USA) matched with D2O (insert 4.1 × 190 mm; outer tube 4.52 mm ID, 4.965 mm OD × 180 mm). A 5-mm-OD NMR tube (Wilmad, Buena, NJ, USA) designed for 500 MHz NMR was used for the InsP6 standard under the same conditions as those of F2-S1-Q6-D2, except for fewer scans. 1H, 31P, and 13C chemical shifts were reported in ppm and assignments for both F2-S1-Q6-D2 and the InsP6 standard were made with reference to the 4.67 ppm water peak. Coupling constants were determined using the MestRe-C software package and spectra were analyzed by MestReNova software (Mestrelab Research, Santiago de Compostela, Spain):

  • F2-S1-Q6-D2: 1H NMR analysis δ 4.59 (d, 1, H2), 4.09 (q, 2, H4/6), 3.82 (q, 1, H5), 3.76 (t, 2, H1/3). 31P NMR analysis δ 2.54 (1, P5), 2.29 (2, P1/3), 2.06 (1, P2), 1.63 (2, P4/6). 13C NMR analysis δ 77.40 (1, C5), 75.89 (2, C4/6), 74.51 (1, C2), 73.39 (2, C1/3).
  • InsP6 standard: 1H NMR analysis δ 4.56 (d, 1, H2), 4.08 (q, 2, H4/6), 3.79 (q, 1, H5), 3.74 (t, 2, H1/3). 31P NMR analysis δ 2.72 (1, P5), 2.54 (2, P1/3), 2.30 (1, P2), 1.75 (2, P4/6). 13C NMR analysis δ 77.47 (1, C5), 75.77 (2, C4/6), 74.72 (1, C2), 73.38 (2, C1/3).

Two-dimensional NMR characterization

1H-31P heteronuclear single quantum correlation (HSQC) experiments were carried out using the standard Bruker program hsqcetgpsi2 modified for 31P gradient strength. The parameters were 1.5 s recycle delay, 1024 data points in F2, 128 increments in F1, globally optimized alternating phase rectangular pulse (GARP) 31P decoupling during acquisition, and shifted sine-squared apodization before Fourier transformation.

1H-13C HSQC experiments were carried out using the standard Bruker program hsqcetgpsi2. The parameters were 1.5 s recycle delay, 1024 data points in F2, 128 increments in F1, GARP 13C decoupling during acquisition, and shifted sine-squared apodization before Fourier transformation.

Concentration of InsP6 dodecasodium salt equivalent to fraction F2-S1-Q6-D2

The F2-S1-Q6-D2 fraction used for the NMR analysis had a pH of 8.22 and an array of sodium adduct ions were observed downstream of the InsP6 peak in the LC-MS spectra. Eight of the 12 exchangeable hydrogens in InsP6 have a pKa below 8.22 (Barre et al., 1954; Isbrandt & Oertel, 1980), and the resulting InsP6-8Na salt has a molecular weight of 836 Da. F2-S1-Q6-D2 extracted from 100 FGs weighed 0.11 mg equivalent to 26.3 nmol from 20 FGs. Therefore, 26.3 nmol of the commercial phytic acid dodecasodium salt hydrate in 2 ml of water was used as the equivalent of fraction F2-S1-Q6-D2. In the bioassays, this was divided equally between 40 wells, and topically applied to the stage 2 somatic embryos at 50 μl per well, with each well already containing 2 ml of 1250 multiplication medium. The final concentration of 0.32 μM InsP6 in each well was calculated based on combined volume of the medium and the 50 μl topical application. A stock solution (0.53 mM) was made and several concentrations ranging from 0.032 to 3.2 μM were generated by dilution. The bioassays were performed as described earlier in ‘early-stage somatic embryogenic multiplication bioassay’.

Bioassays with muco-inositol hexakisphosphate

Somatic embryos at stage 2 were isolated by forceps from suspension culture and placed on 2 ml of multiplication medium 1250 contained in 24-well plates. Fifty replicate bioassays were then carried out, for each concentration tested, using the following protocol. A stock solution of muco-InsP6 (1.8 mg ml−1) was adjusted to pH 5.7 and sterilized with a 0.2 μm syringe filter. A series of dilutions were made such that, when topically applied to the stage 2 somatic embryos at 50 μl per well, the final concentration of muco-InsP6 ranged from 6.7 to 670 μg ml−1. Embryos were grown and measured and all data were analyzed as described in ‘early-stage somatic embryogenic multiplication bioassay’.

Norway spruce materials and experimental design

Briefly, seeds from F.W. Schumacher Co., Sandwich, MA, USA, were rinsed under cold tap water for 30 min, soaked overnight, and surface-sterilized. Shortly after sterilization, seeds were dissected and the integuments and nucellus removed. The FG was carefully split, the embryo removed, and the embryo and split gametophyte placed next to each other or the embryo was placed alone on 7 ml of initiation medium 1165 contained in six-well Petri plates. Sixteen replications of six seeds were tested per treatment with no FG present or with the FG placed next to the embryo and incubated at 23–24°C in 16 h of c. 30 μmol photons m−2 s−1 of cool white fluorescent light.

Results

Liquid chromatography isolation and identification

We established a purification protocol for the inhibitor of early-stage somatic embryo growth comprising water extraction, a Superdex 75 gel filtration column, a Mono S strong cation exchange column, two ultrafiltrations against 1000 Da cutoff membranes after the first two columns, a Mini Q strong anion exchange column, and a Superdex Peptide gel filtration column. For the first-step fast protein liquid chromatography (FPLC) (the Superdex 75 column), F2 was the only fraction that inhibited embryo growth in a statistically significant manner compared with the blank control (Fig. 1). The second fractionation was achieved on a Mono S column. F2-S1 eluting during the wash of unbound compounds was the only fraction that showed statistically significant inhibition to the embryo growth (Fig. 2), in exactly the same manner as that of the F2 control. Nonretention on a strong cation exchanger suggested that the inhibitor is highly anionic. Thus, Mini Q was selected for the third-step chromatography isolation. Bioassay results revealed that F2-S1-Q6 inhibited embryo growth in a statistically significant manner (Fig. 3). Therefore, F2-S1-Q6 was further purified on a Superdex Peptide column. Fig. 4 shows the bioassay results of the Superdex Peptide column. It is evident that F2-S1-Q6-D2 is the only subfraction that exhibits statistically significant inhibitory activity of the bioassay.

Figure 1.

Superdex 75 fast protein liquid chromatography elution profile for the fractionation of female gametophyte (FG) water extract eluted with 150 mM NaCl. Elution was monitored at 215, 254 and 280 nm, and the absorbance in milliabsorbance units (mAU) at 280 nm is shown on the chromatogram. Fractions collected in multiple tubes (0.5 ml per tube) were pooled and bioassayed with the following designations: F1, from 15.0 to 27.0 min (6.0 ml); F2, from 27.0 to 34.0 min (3.5 ml); F3, from 34.0 to 55.0 min (10.5 ml); and F4, from 59.0 to 77.0 min (9.0 ml). The inset shows the bioassay results (using genotype 500) for the fractions eluted from the Superdex 75 column. Forty replicate bioassays were carried out for each fraction and also for blank controls that contained only the bioassay components. All the data were evaluated by multifactor ANOVA. The significant differences between means of each treatment were determined by the multiple range test at 95% level of significance. F2 was the only fraction inhibiting the embryo growth in a statistically significant manner compared with the blank control. The chromatography and bioassays were repeated at least three times with similar results. LSD, least significant difference.

Figure 2.

Mono S fast protein liquid chromatography elution profile for the fractionation of the bioactive fraction ‘F2’ (from Fig. 1) eluted with 20 mM sodium phosphate buffer at pH 2.5. Elution was monitored at 215, 254 and 280 nm, and the absorbance in milliabsorbance units (mAU) at 215 nm is shown on the chromatogram. Fractions collected in multiple tubes (1.0 ml per tube) were pooled and bioassayed with the following designations: F2-S1, from 0.7 to 2.0 min (2.0 ml); F2-S2, from 2.0 to 3.3 min (2.0 ml); F2-S3, from 13.3 to 21.3 min (12.0 ml); F2-S4, from 21.3 to 27.3 min (9.0 ml); and F2-S5, from 37.5 to 41.5 min (6.0 ml). The inset shows the bioassay results (using genotype 500) for the fractions eluted from the Mono S column and the F2 control from the previous Superdex 75 column. Forty replicate bioassays were carried out for each fraction and also the blank control that contained only the bioassay components. All the data were evaluated by multifactor ANOVA. The significant differences between means of each treatment were determined by the multiple range test at 95% level of significance. F2-S1 was the only fraction that showed statistically significant inhibition to the embryo growth, in exactly the same manner as that of the F2 control, compared with the blank control. The chromatography and bioassays were repeated at least three times with similar results. LSD, least significant difference.

Figure 3.

Mini Q fast protein liquid chromatography elution profile for the fractionation of the bioactive fraction ‘F2-S1’ (from Fig. 2) eluted with 20 mM Tris-Cl buffer at pH 8.0. Elution was monitored at 215, 254 and 280 nm, and the absorbance in milliabsorbance units (mAU) at 215 nm is shown on the chromatogram. Fractions collected in multiple tubes (0.5 ml per tube) were pooled and bioassayed with the designations: F2-S1-Q1 from 2.1 to 6.1 min (2.0 ml), F2-S1-Q2 from 6.1 to 11.1 min (2.5 ml), F2-S1-Q3 from 25.4 to 30.4 min (2.5 ml), F2-S1-Q4 from 30.4 to 36.4 min (3.0 ml), F2-S1-Q5 from 36.3 to 43.3 min (3.5 ml), and F2-S1-Q6 from 49.4 to 58.3 min (4.5 ml). The inset shows the bioassay results (using genotype 500) for the fractions eluted from the Mini Q column and the F2-S1 control from the previous Mono S column. Forty replicate bioassays were carried out for each fraction and also the blank control that contained only the bioassay components. Both F2-S1-Q5 and F2-S1-Q6 had statistically significant inhibition to the embryo growth, with F2-S1-Q6 being the most significant and selected for further purification. The chromatography and bioassays were repeated at least three times with similar results. LSD, least significant difference.

Figure 4.

Superdex Peptide fast protein liquid chromatography elution profile for the fractionation of the bioactive fraction ‘F2-S1-Q6’ (from Fig. 3) eluted with 100 mM ammonium acetate buffer at pH 5.5. Elution was monitored at 215, 254 and 280 nm, and the absorbance in milliabsorbance units (mAU) at 215 nm is shown on the chromatogram. Fractions collected in multiple tubes (0.2 ml per tube) were pooled and bioassayed with the following designations: F2-S1-Q6-D1, from 62.1 to 80.1 min (3.6 ml); F2-S1-Q6-D2, from 80.1 to 90.1 min (2.0 ml); F2-S1-Q6-D3, from 90.1 to 95.1 min (1.0 ml); F2-S1-Q6-D4, from 95.1 to 98.1 min (0.6 ml); and F2-S1-Q6-D5, from 98.1 to 100.1 min (0.4 ml). The inset shows the bioassay results (using genotype 500) for the fractions eluted from the Superdex Peptide column, the blank control, the F2-S1 control from the Mono S column, and a buffer control labeled NH4Ac (6 ml of 100 mM ammonium acetate lyophilized to dryness, corresponding to the highest amount of ammonium acetate contained in any fraction). Forty replicate bioassays were carried out for each fraction and also the blank control that contained only the bioassay components. F2-S1-Q6-D2 is the only subfraction that exhibits statistically significant inhibitory activity in the bioassay. The chromatography and bioassays were repeated at least three times with similar results. LSD, least significant difference.

Structure characterization by LC-MS and LC-MS-MS under positive polarity

Liquid chromatography-mass spectrometry was performed on each successive inhibitory fraction to investigate components, as well as to identify the inhibitor to be pursued. One distinct ion at m/z 661 was detected in all active fractions, including F2, F2-S1, F2-S1-Q6 (ultrafiltered against 3000 Da cutoff membrane), F2-S1-Q6-D2, and in F2-S1-Q5 (36.3–43.3 min) which was significantly more dilute and less active than F2-S1-Q6. Inactive fractions from both the Mini Q and the Superdex Peptide columns did not reveal the presence of this ion in their mass spectra. These results indicate that the ion at m/z 661 corresponds to the inhibitor to be pursued. LC-MS of F2-S1-Q6-D2 revealed that only the ion at m/z 661 was dominantly present, with an array of sodium adduct ions with mass difference of 22 amu (the mass difference between sodium Na+ and proton H+) being observed downstream of the ion at m/z 661. Additional LC-MS analyses of all fractions from the Mini Q column under negative polarity were acquired (data not shown). The m/z difference of 2 amu between the ion at m/z 661 from the positive scan and the ion at m/z 659 from the negative scan indicate that both ions are singly charged and the neutral monoisotopic mass of this molecule is 660 Da.

Liquid chromatography-mass spectrometry-mass spectrometry spectra of the precursor ion at m/z 661 under positive polarity were then acquired on the QTRAP. The precursor ion (M+H)+ at m/z 661 was subjected to collision energies of 40 and 80 eV. At 40 eV, the protonated ion at m/z 661 yielded a series of prominent ions at m/z 643, 625, 607, 581, 563, 545, 527, 483, 465, 447, 429, 385, 367, 349, 287, 269, 259, 189, 179, 161, and 99. At 80 eV, further dissociation intensified the low m/z fragment ions, whereas the abundances of the high m/z fragment ions decreased tremendously. Ions higher than m/z 447 were not observed.

One-dimensional NMR characterization

The 13C NMR spectra of F2-S1-Q6-D2 revealed four distinct peaks at 77.40, 75.89, 74.51 and 73.39 ppm with an intensity ratio of 1 : 2 : 1 : 2; this indicates the presence of six carbons with two independent pairs of equivalent carbons and two nonequivalent individual carbons. A search within the spectral database for organic compounds (National Institute of Advanced Industrial Science and Technology, 2007) based on 13C NMR spectra found only one match as myo-inositol. Similar patterns were observed in both 1H NMR (four distinct peaks at 4.59, 4.09, 3.82 and 3.76 ppm with an intensity ratio of 1 : 2 : 1 : 2, Fig. 5a) and 31P NMR (four distinct peaks at 2.54, 2.29, 2.06 and 1.63 ppm with an intensity ratio of 1 : 2 : 1 : 2) spectra of F2-S1-Q6-D2.

Figure 5.

1H NMR spectra of F2-S1-Q6-D2 and myo-inositol hexakisphosphate (InsP6) standard. (a) 1H NMR spectrum of F2-S1-Q6-D2; (b) 1H NMR spectrum of the InsP6 standard, both recorded at pH 8.22. (c) The structure of InsP6 with the carbons numbered.

The pH of F2-S1-Q6-D2 was measured at 8.22 and all its spectra were recorded at this value. The InsP6 standard was prepared from its dodecasodium salt through cation exchange resin and the pH adjusted to the same value using NaOD, in order to circumvent the generation of extra NaCl. The chemical structure of InsP6 is shown in Fig. 5(c). Both the chemical shifts and splitting patterns of F2-S1-Q6-D2 are identical to those of the InsP6 standard (Fig. 5b). Taken together, these results are consistent with the conclusion that the inhibitory molecule in F2-S1-Q6-D2 is InsP6. The 1H chemical shift assignments shown in Fig. 5 were made according to Bauman et al. (1999).

Exact mass analysis and MS-MS under negative polarity

Exact mass measurements and subsequent tandem MS under negative ion polarity of both F2-S1-Q6-D2 and the InsP6 standard were carried out on the QSTAR XL system interfaced with an automated chip-based nanoelectrospray TriVersa™ NanoMate system. As shown in the Fig. 6 (a) (inset), F2-S1-Q6-D2 exhibited exact mass (M-H) of 658.8506 with an error of negative 5.3 ppm with respect to the calculated molecular mass of 658.8541, and the InsP6 standard showed exact mass (M-H) of 658. 8563 with an accuracy of 3.3 ppm, as shown in the Fig. 6 (b) (inset).

Figure 6.

Exact mass analysis and MS-MS under negative polarity of both F2-S1-Q6-D2 and InsP6 standard. (a) Tandem mass spectrum of F2-S1-Q6-D2 at 40 eV and the exact mass (M-H). (b) Tandem mass spectrum and exact mass for the InsP6 standard.

Two collision energies of 40 and 80 eV were applied to the precursor ion (M-H) at m/z 659 from F2-S1-Q6-D2 and the InsP6 standard, respectively. MS-MS spectra of F2-S1-Q6-D2 (Fig. 6a) and the InsP6 standard (Fig. 6b) under both collision energies were identical (MS-MS at 80 eV not shown). Furthermore, MS-MS spectra acquired under negative polarity were consistent with those from positive mode, with each fragment ion lower by 2 amu.

Two-dimensional NMR characterization

Two-dimensional 1H-13C HSQC, 1H-31P HSQC and 1H-1H COSY experiments were obtained for both F2-S1-Q6-D2 and the InsP6 standard to assign the chemical shifts as well as to compare the correlation patterns. The chemical shift assignments and coupling constants are listed in Table 1 and the corresponding 1H-13C and 1H-31P HSQC spectra are shown in Figs 7 and 8. All two-dimensional NMR of F2-S1-Q6-D2 showed identical correlation patterns to those obtained from the InsP6 standard. Consistency in correlation patterns, chemical shifts and intensity ratios in all one-dimensional and two-dimensional NMR spectra unequivocally established that the inhibitory molecule isolated from FG tissue is myo-inositol hexakisphosphate.

Table 1. Myo-inositol hexakisphosphate (InsP6) 1H-, 31P- and 13C-NMR chemical shift assignments and coupling constants
 1H
 chemical shiftProton positionIntegrationMultiplicityJ (Hz)
F2-S1-Q6-D24.5921dJ2-P2 (8.80)
4.094/62qJ4-3,5,P4 (9.05)
J6-1,5,P6 (9.05)
3.8251qJ5-4,6,P5 (9.05)
3.761/32tJ1-6,P1 (8.30)
    J3-4,P3 (8.30)
InsP6 Standard4.5621dJ2-P2 (9.05)
4.084/62qJ4-3,5,P4 (9.54)
J6-1,5,P6 (9.54)
3.7951qJ5-4,6,P5 (9.54)
3.741/32tJ1-6,P1 (8.80)
    J3-4,P3 (8.80)
 31P
 Chemical ShiftPhosphate PositionIntegration  
F2-S1-Q6-D22.5451  
 2.291/32  
 2.0621  
 1.634/62  
InsP6 Standard2.7251  
 2.541/32  
 2.3021  
 1.754/62  
 13C
 Chemical ShiftCarbon PositionIntegration  
F2-S1-Q6-D277.4051  
 75.894/62  
 74.5121  
 73.391/32  
InsP6 Standard77.4751  
75.774/62  
74.7221  
73.381/32  
Figure 7.

1H-13C heteronuclear single quantum correlation (HSQC). (a) 1H-13C HSQC of F2-S1-Q6-D2. (b) 1H-13C HSQC of the myo-inositol hexakisphosphate (InsP6) standard. Carbon atoms are numbered as shown in Fig. 5(c).

Figure 8.

1H-31P heteronuclear single quantum correlation (HSQC). (a) 1H-31P HSQC of F2-S1-Q6-D2; (b) 1H-31P HSQC of the myo-inositol hexakisphosphate (InsP6) standard. Carbon atoms are numbered as shown in Fig. 5(c).

Concentration dependence of bioactivity on InsP6

Bioassays were carried out to confirm that the InsP6 standard inhibits the early-stage somatic embryo growth. The histories of the genotypes used in these experiments are as follows: genotype 500 was from a seed from mother tree 7–56 initiated in 2003; genotype 51 was from a seed from mother tree MWv-2 initiated in 2005; genotype 222 was from a seed from a high-value cross initiated in 2007; genotype 279 was from a seed from mother tree 7–56 initiated in 2006; and genotype 433 was from a seed from a high-value cross initiated in 2007.

Results for five genotypes (51, 222, 433, 279, 132) tested at none and five concentrations of InsP6 are averaged and shown in Fig. 9. The InsP6 standard at five concentrations was found to inhibit somatic embryo growth in a statistically significant manner. Furthermore, inhibition corresponding to the concentration of InsP6 actually isolated from female gametophytes (0.32 μM in the bioassay well) was the most significant. An additional two genotypes were tested at none and 0.32 μM. All seven genotypes tested showed reduced growth in the bioassay with application of InsP6 at 0.32 μM; differences were statistically significant at P = 0.05.

Figure 9.

Bioassay results averaged for five genotypes (51, 222, 433, 279, 132) tested at different concentrations of myo-inositol hexakisphosphate (InsP6) standard. Means of five genotypes are shown along with 95% least significant difference (LSD) intervals for each concentration of InsP6 tested.

Muco-inositol hexakisphosphate

Bioassays were carried out to test whether muco-InsP6, a stereoisomer of InsP6, also inhibits early-stage embryo growth. Results for one genotype (652) tested at none and six concentrations of muco-InsP6 are shown in Table 2. It is evident from the data that muco-InsP6 at six concentrations did not inhibit somatic embryo growth, whereas control experiments confirmed that, as expected, InsP6 itself did inhibit somatic embryo growth using this genotype.

Table 2.   Effect of muco-inositol hexakisphosphate (muco-InsP6) on early-stage somatic embryo growth
muco-InsP6 concentration (μg ml−1)Diameter (mm)Calculated muco-InsP6 concentration range (μM)1
  1. Diameter values are followed by their standard error. No treatment was statistically different from any other group, ANOVA > 0.05.

  2. 1Calculated muco-InsP6 molar concentration ranges for barium stoichiometry from zero to six (see ‘Materials and Methods’ section).

06.8 ± 0.30
6.75.7 ± 0.40.023–0.052
416.4 ± 0.40.14–0.32
506.4 ± 0.40.17–0.39
676.5 ± 0.40.23–0.52
926.3 ± 0.40.32–0.71
6706.0 ± 0.42.3–5.2

Norway spruce embryogenic tissue initiation

Several embryogenic tissue initiation protocols in conifers call for full-term seed FG to be placed next to the embryo on the initiation medium. With the finding that FG tissue contains InsP6 that is able to inhibit early-stage LP somatic embryo growth, the question arose if FG tissue might inhibit initiation when full-term seed embryos are used for initiation. To test this hypothesis, the procedure indicated in Pullman et al. (2003c) was used to initiate embryogenic tissue from mature Norway spruce embryos with FG tissue present or absent. After 9–10 weeks, explants were evaluated for the presence of embryogenic tissue (Table 3). Embryogenic tissue formation in Norway spruce was increased from 23.8 to 56.1% (differences were statistically significant) when the FG tissue was not present.

Table 3.   Effect of female gametophyte presence or absence on initiation of Norway spruce embryogenic tissue from full-term seeds
Treatment% initiation1
  1. FG, female gametophyte.

  2. 1Sixteen replications of six explants were tested per treatment. Analyses for initiation are based on arcsine transformation √(%). Values followed by the same letter are not statistically different by the multiple range test at = 0.05.

1165 + FG28.3 a
1165 No FG56.1 b

Discussion

We report here the first evidence of inhibition of plant somatic embryonic growth by InsP6. The highly anionic inhibitor was purified from late-stage LP FG tissue by water extraction, two gel filtrations and two ion exchange FPLC chromatographies, and the final bioactive molecule was then fully characterized as to structure and purity.

Liquid chromatography-mass spectrometry of the final active fraction indicated that the active molecule is homogeneous and has a neutral monoisotopic mass of 660 Da. The LC-MS-MS spectrum under positive mode of F2-S1-Q6 at a collision energy of 40 eV was identical to those obtained at 30 and 35 eV by Hsu et al. (2003). A series of prominent ions at m/z 643, 625, 607, 581, 563, 545, 527, 483, 465, 447, 429, 385, 367, 349, 287, 269, 259, 189, 179, 161, and 99 were generated by successive losses of H2O (18 amu), HPO3 (80 amu), or H3PO4 (98 amu) from the precursor ion at m/z 661, with some fragment ions arising from multiple pathways. Identical exact mass and fragmentation patterns obtained from high-resolution exact mass measurement and MS-MS analysis under negative mode clearly identified the purified compound from FG tissue as one of the isomers of inositol hexakisphosphate.

The active molecule was then identified as the myo-isomer of inositol hexakisphosphate on the basis of 1H-, 31P- and 13C-NMR, 1H-1H COSY, 1H-31P HSQC and 1H-13C HSQC. In the one-dimensional NMR spectra, the chemical shifts, splitting patterns and intensity ratios were identical to those of the InsP6 standard, and a search within spectral database for organic compounds based on 13C NMR spectra found only one match as myo-inositol. The 13C-, 1H-, and 31P- NMR spectra all exhibited four distinct peaks with an intensity ratio of 1 : 2 : 1 : 2, which is the pattern expected for the myo-isomer (Fig. 5c). Indeed, Angyal & Odier (1982) compared the 13C NMR spectra of a series of diastereomeric inositols, and only the myo- stereoisomer exhibited this 1 : 2 : 1 : 2 intensity ratio. All the two-dimensional NMR of the isolated molecule showed identical correlation patterns to those obtained from the myo-inositol hexakisphosphate standard. Taken together, these results unequivocally establish that the inhibitory molecule isolated from FG tissue is myo-inositol hexakisphosphate. Furthermore, we find here that topical application of the muco-InsP6, a stereoisomer of InsP6, to early-stage somatic embryos does not inhibit embryonic growth in a statistically significant manner, demonstrating the stereo specificity of InsP6’s inhibition.

We demonstrate here that topical application of InsP6 to early-stage somatic embryos from several different genotypes varying in genetic background indeed inhibits embryonic growth in a statistically significant manner. Further, we demonstrate a practical application of our findings, in that presence of full-term seed FG tissue inhibits embryogenic tissue initiation and should be removed to facilitate culture initiation. Medium 1250 contains 9.33 mg l−1 Na2EDTA to buffer divalent metal ions. Therefore, we do not believe that the inhibitory activity of InsP6 is the result of metal chelation. Further, when Pullman et al. (2003b) modified a similar salt mixture for improvement of embryo development and maturation by comparing control and raised medium phosphorous through addition of extra KH2PO4, somatic embryo yield differences were not statistically significant between treatments.

Myo-inositol hexakisphosphate is ubiquitous and the most abundant inositol phosphate derivative in eukaryotic cells. It is known for its anticancer activity in reducing the proliferation of malignant cells (Shamsuddin et al., 1992; Shamsuddin & Yang, 1995; Ferry et al., 2002). Additionally, InsP6 increases differentiation of malignant cells leading to reversion to the normal phenotype with decreased production of tumor markers (Shamsuddin & Vucenik, 2005). Some evidence has begun to emerge that InsP6 may also function as a signaling molecule in plant cells. Lemtiri-Chlieh et al. (2000) reported that the plant stress hormone, abscisic acid, increases InsP6 in intact guard cells of Solanum tuberosum and that InsP6 inhibits the inward rectifying K+ current of S. tuberosum and Vicia faba guard cell protoplasts in a Ca2+-dependent manner. Subsequently (Lemtiri-Chlieh et al., 2003), they showed by laser uncaging of InsP6, in V. faba guard cell protoplasts loaded with calcium-sensitive dye, that InsP6 causes release of Ca2+ from internal stores. It should also be noted that Tan et al. (2007) have recently reported that InsP6 is a cofactor in the transport inhibitor response 1 protein (TIR1) that senses and becomes activated by the phytohormone auxin. However, reduction of proliferation in plant cells by InsP6 has not been reported to date, and many aspects of the function of InsP6 in plants have remained undefined (Turner et al., 2002; Raboy, 2003). Our findings constitute the first report that InsP6 inhibits cell proliferation in plants.

Is it possible that inhibition of somatic embryo growth in plants by InsP6 and InsP6’s anticancer activity occur via similar mechanisms? In JB6 epidermal cells, it has been shown that InsP6 inhibits epidermal growth factor-induced phosphatidylinositol-3 kinase (PtdIns 3-kinase), thereby impairing epidermal growth factor- or phorbol ester-induced cell transformation and activator protein 1 activation (Huang et al., 1997). PtdIns 3-kinases are widely distributed in eukaryotic cells, and they are involved in a number of cellular processes, including activation of intracellular signaling molecules such as rac, ras, rab, mitogen-activated protein kinase, protein kinase B/Akt (Vanhaesebroeck & Waterfield, 1999), protein kinase C and JNK/p38 kinase (Leevers et al., 1999; Meijer & Munnik, 2003; Amin et al., 2007). Turning to plants, PtdIns 3-kinase homologs have been cloned in soybean (Hong & Verma, 1994), Arabidopsis thaliana (Welters et al., 1994) and Brassica napus (Das et al., 2005), and expression of antisense PtdIns 3-kinase AtVPS34 mRNA results in severe inhibition in growth and development of second-generation transformed plants. Recently, both PtdIns 3-kinase and PtdIns 4-kinase activities have been observed during the induction of somatic embryogenesis in Coffea arabica (Ek-Ramos et al., 2003), and the products of both kinase activities were detected in the somatic-embryo extracts. Moreover, growth of these somatic embryos was inhibited when a kinase inhibitor was included in the induction medium during the first differentiated stage (Ek-Ramos et al., 2003). Taken together, these facts are not inconsistent with the notion that inhibition of PtdIns kinase may be a common feature of InsP6’s activity as an inhibitor of somatic embryo growth in plants and as an anticancer agent, but at this point the evidence must be regarded as circumstantial. In this regard, we have carried out a BLAST database search on an expressed sequence tag library of LP somatic embryos (Cairney et al., 2006), and we have identified one singleton (Gene Bank number DR688191) that shows 83% identity in amino acid sequence to that of PtdIns 3-kinase AtVPS34. Clearly, additional studies will be needed to fully elucidate the mechanisms by which InsP6, (and perhaps other inositol phosphates as well) regulate cellular growth and development in plants. Such studies could well lead to significant improvements in the technology of somatic embryogenesis in plants.

Acknowledgements

This work was supported by the National Research Initiative of the USDA Cooperative State Research, Education and Extension Service, grant number 2003-35103-12924, and also through The Consortium for Plant Biotechnology Research, Inc. by DOE Prime Agreement No. DEFG36-02GO12026, and by the Monsanto Company as a member of the CPBR. We thank the Institute of Paper Science and Technology at Georgia Tech for a fellowship award to D.W. We also thank Drs Kristi Burns, Yanfeng Chen, Zhen Zhou, and Michael Foster for many technical discussions and recommendations. Muco-InsP6 was provided as a generous gift by Dr Alan Richardson from the collection of the late Dr Dennis Cosgrove.

Ancillary