INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Di-myo-inositol-1,1'-phosphate (DIP) is an unusual inositol derivative that has been identified as a major solute in hyperthermophilic archaea including Pyrococcus woesei (22), Pyrococcus furiosus (16), Methanococcus igneus (5), and several eubacteria of the order Thermotogales (15). Intracellular DIP increases with increasing extracellular concentrations of NaCl in both M. igneus (5) and P. furiosus (16). DIP also increases dramatically at supraoptimal growth temperatures (>80°C for M. igneus and 98 to 101°C for P. furiosus). The unusual intracellular high concentration of K⁺ ions and the extreme optimal growth temperatures (100 to 104°C) of P. woesei (30) suggested the role of DIP as a main counterion of K⁺ with a possible thermostabilizing action. Scholz et al. (22) demonstrated that among several salts, the potassium salt of DIP provided optimum enzyme stabilization when the activity of glyceraldehyde-3-phosphate dehydrogenase of P. woesei was tested at 105°C under anaerobic conditions.

Since de novo synthesis of DIP occurs in response to external levels of NaCl and temperature, there must be regulatory biosynthetic mechanisms linked to osmotic pressure and temperature. To study the regulation, the enzymes and/or other proteins responsible for synthesis of this compatible solute must be isolated. This requires knowledge of the biosynthetic pathways involved in the synthesis of DIP. The sole known pathway for inositol biosynthesis in all other organisms is the conversion of D-glucose-6-phosphate to L-myo-inositol 1-phosphate (L-I-1-P) via L-myo-inositol 1-monophosphate (I-1-P) synthase and hydrolysis of I-1-P to myo-inositol via a specific phosphatase, I-1-P phosphatase (13, 14). Similar enzymes are likely to exist in methanogens. A logical pathway for the biosynthesis of DIP would then use myo-inositol and I-1-P as precursors. Activation of the I-1-P with CTP and condensation of the resultant CDP-inositol (CDP-I) with myo-inositol would generate DIP. As summarized in Fig. 1, DIP biosynthesis requires four key enzymes: I-1-P synthase (step 1), I-1-P phosphatase (step 2), CTP:I-1-P cytidylyltransferase (step 3), and DIP synthase (step 4). The enzymes that catalyze steps 1 and 2 have been well studied in plants, yeasts, and mammalian tissues. However, the enzymes invoked for steps 3 and 4 are novel activities, although based on similar chemical transformations in cells.

View larger version (22K): [in this window] [in a new window]   FIG. 1.   Proposed biosynthetic pathway for DIP showing the four key enzymatic activities. Based on similar transformations in other organisms, cofactors are indicated for several of the steps.

This work describes the use of ³¹P nuclear magnetic resonance (NMR) and colorimetric assays to verify the existence of three of these activities in cell extracts of M. igneus. Specific labeling of DIP with [¹³C]pyruvate was also used to probe the DIP biosynthetic pathway. The pattern of ¹³C label incorporation from [3-¹³C]pyruvate and [2,3-¹³C]pyruvate coupled with the known stereochemistry of DIP provided evidence that M. igneus also has enzymes of the pentose phosphate pathway (transaldolase and transketolase) that scramble label in glucose-6-phosphate.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. The sodium salts of [2,3-¹³C]pyruvate and [1-3¹³C]pyruvate were purchased from Cambridge Isotope Laboratories, Inc. D-Glucose-6-phosphate (G-6-P) disodium salt, I-1-P cyclohexylammonium salt (2-monophosphate isomer content, 27%), fructose-6-phosphate (F-6-P), -glycerophosphate, NAD⁺, CTP, CMP-morpholidate, and D₂O were obtained from Sigma. Ammonium sulfate and myo-inositol were obtained from Aldrich. CDP-I was chemically synthesized by the method of Roseman et al. (21) for nucleotide coenzymes. CMP-morpholidate and the trioctylamine salt of I-1-P (prepared by cation-exchange chromatography) were incubated in anhydrous pyridine. The CDP-I products (including the CDP adducts at C-1 of L-inositol and D-inositol as well as the CDP-I with the C-2 hydroxyl of inositol in the pyrophosphate bond) precipitated out of the pyridine solution. The precipitate was collected and purified by anion-exchange chromatography. The final mixture contained the CDP-I isomers, a small amount of UMP that cochromatographed with the pyrophosphate diesters, and the self-condensation product, di-cytidine-5,5'-pyrophosphate (DCPP), of CMP-morpholidate as monitored by ³¹P NMR spectroscopy. A ¹H-¹H total correlation spectroscopy (TOCSY) NMR experiment was used to confirm the presence of the inositol ring in CDP-I.

Cell growth and ¹³C labeling. M. igneus cell cultures were grown at 85°C in modified MGG medium (5) containing 0.3 M NaCl under an H₂-CO₂ (4:1) atmosphere. The pH of the medium was adjusted to a slightly acidic value (6.8 to 6.9) prior to sterilization by autoclaving. Once cells reached an optical density of 0.6 at 660 nm, they were harvested by centrifugation at 9,000 rpm for 30 min in a Sorvall centrifuge. Cell pellets were either (i) extracted three to five times with 70% ethanol (this lyses cells and precipitates protein and other macromolecular debris, leaving only small molecules in solution) as described previously (20) to prepare extracts containing DIP or (ii) resuspended and lysed by sonication (see below) to yield a cell-free protein extract. For ¹³C labeling experiments, 0.15 M stock solutions of the sodium salts of [2,3-¹³C]pyruvate and [3-¹³C]pyruvate, each enriched with 99% ¹³C, were made anaerobic by pressurizing for 10 s with argon (10 lb/in²) followed by 3 min of vacuum and repetition of the procedure three times. Because of the high growth temperature (85°C) and limited stability of pyruvate, the labeled sodium pyruvate was injected into M. igneus cultures (150 ml per bottle) in multiple doses of ~4 ml/h during exponential growth phase (a total of two or three injections) to ensure maximum uptake and utilization of the compound.

Isolation of DIP. Fifty milligrams of an M. igneus ethanol extract (prepared by incubating the cell pellet with 5 ml of 70% ethanol, centrifuging the sample, reextracting the pellet with ethanol three or more times, combining the ethanol fractions, and evaporating the solvent [20]) was dissolved in 2 ml of doubly distilled water. The solution pH was adjusted to 6.5 to 7.0 with HCl. The sample was then loaded onto a QAE-Sephadex (Sigma) column (bed volume, 15 ml) equilibrated with 50 mM ammonium acetate (pH 6.8). The column was eluted with 45 ml of 50 mM ammonium acetate. Fractions of 3 ml were collected, and 1-ml aliquots of these fractions were lyophilized, dissolved in D₂O, and analyzed by ¹H NMR. Fractions 8 to 11 containing pure DIP were pooled, lyophilized, and used for determination of optical activity of DIP.

Cell-free protein extracts. Since there was no apparent requirement for oxygen-sensitive cofactors in the DIP synthase enzymes, all protein extracts were made under aerobic conditions with the exception of extracts for assays of CDP-I synthase. An intact cell pellet (~1.0 g) was resuspended in 10 ml of standard buffer (50 mM Tris acetate, 1 mM EDTA, 50 mM 2-mercaptoethanol [pH 8.0]). The cells were lysed by sonication using a 30-s pulse-30-s off cycle repeated 10 times. The extract was then centrifuged at 9,000 rpm for 20 min to remove the cell debris. To the supernatant, solid ammonium sulfate was slowly added to 44% saturation. After a 20-min incubation, the suspension was centrifuged at 17,000 rpm. The precipitate was dissolved in standard buffer; this fraction is designated the 0 to 44% ammonium sulfate fraction. Solid ammonium sulfate was added to the supernatant to 85% saturation, and the precipitated protein was centrifuged and redissolved in standard buffer (the 44 to 85% ammonium sulfate protein fraction). A later fractionation used a 60 to 75% ammonium sulfate fraction for verification of I-1-P synthase activity. Ammonium sulfate fractions were dialyzed overnight at 4°C against 1,000 volumes of standard buffer. These protein fractions were used immediately or quickly frozen on dry ice and stored at 20°C.

In vitro assays for I-1-P, inositol, and DIP biosynthesis. Two different methods, ³¹P NMR spectroscopy and a colorimetric inositol monophosphate assay (1), were used to measure the formation of I-1-P. The ³¹P NMR assays monitored substrate (G-6-P) conversion to I-1-P. The assay mixtures (0.5 ml) contained 5 mM G-6-P, 1 mM NAD⁺, 15 mM NH₄Cl, 50 mM Tris acetate (pH 8.0), 50 µl of enzyme extract, and 20% D₂O (for the spectrometer lock). Samples were incubated at 37°C for 38 h, 55°C for 16 h, and 85°C for 2 to 4 h. Good signal-to-noise spectra were acquired within 30 min, and at room temperature (<25°C) no significant product was generated during this time period. The ammonium sulfate fraction also contained G-6-P isomerase activity, which converts G-6-P to F-6-P. All of these different phosphorylated compounds gave rise to distinct resonances in the ³¹P NMR spectrum (Fig. 1) that could be easily distinguished on the basis of chemical shift and ¹H coupling pattern (the -CH₂OPO₃⁼ unit yields a triplet for each anomer of G-6-P [these overlap to form an asymmetric quartet] with J_HP ~ 6.0 Hz and a triplet for F-6-P with J_HP ~ 5.0 Hz, while the -CHOPO₃⁼ of I-1-P occurs as a doublet with J_HP ~ 8.5 Hz). The ³¹P spectrum in Fig. 1 has not been ¹H decoupled to emphasize the different ¹H-³¹P coupling patterns for substrates, products, and side reactions. The I-1-P doublet is about 0.47 ppm upfield of those for G-6-P and 0.14 ppm downfield of F-6-P in the pH range of 7.5 to 8.5. Confirmation of the identity of a ³¹P resonance as belonging to a particular phosphorylated compound was also provided by the pH dependence of the ³¹P chemical shift and comparison to known phosphate compounds. Some of an authentic sugar phosphate thought to be responsible for a given resonance was added to the in vitro assay mixture at room temperature. If the added species augmented a resonance observed in the sample (i.e., if the two gave rise to intensity at the same chemical shift), and if this persisted over a wide range of pH values (i.e., only a single peak is observed and not two separate resonances over a wide pH range), the sugar phosphate in the sample was likely to be the known compound. It is highly unlikely that two phosphorylated sugars would have the same chemical shifts and identical pK_as. Thus, in identifying products in these in vitro assays, ³¹P spectra were examined over a range of pH values in the absence and presence of known sugar phosphates. The J_HP coupling constants are also characteristic of different sugar compounds and aided in identification of phosphates (e.g., J_HP = 6 Hz for G-6-P, J_HP = 5 Hz for F-6-P, and J_HP = 8.5 Hz for I-1-P).

NMR spectroscopy. ¹H WALTZ-decoupled ¹³C NMR (125.7-MHz) spectra of ethanol extracts were obtained by using a Varian Unity 500 spectrometer and a 5-mm broadband probe. Samples were dissolved in 0.5 ml of D₂O. ¹³C spectral accumulation parameters included 25,000-Hz sweep width, 65,024 datum points, 90° pulse width (11 µs), and a 1.0-s delay time between acquisitions. Typically, 20,000 to 28,000 transients were acquired, and the free induction decays were processed with 2-Hz line broadening. The absolute amounts of ¹³C incorporated into DIP carbons in extracts from samples incubated with [3-¹³C]pyruvate were estimated by comparing ¹H intensities and ¹³C intensities for DIP and the glutamate isomers. The reductive partial tricarboxylic acid cycle used by this organism labels L--glutamate at C-3 and C-4 but not at C-2. The intensities of the DIP carbons were compared to that of L--glutamate C-2. The bulk ratio of DIP to L--glutamate was then obtained from the ¹H NMR spectrum of the extract and was used to obtain enrichment levels for the DIP carbons.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Chirality of DIP isolated from M. igneus. The DIP synthesized by Pyrococcus species is chiral and has an L,L configuration of the inositols (26). However, it is possible that the DIP isolated from M. igneus is synthesized in a meso form (L,D configuration of inositols esterified at C-1, alternatively called di-myo-inositol-1,3-phosphate) or with a D,D configuration of the inositol units. Therefore, the optical activity of this compound was measured. Moderate amounts of DIP were separated from the - and -glutamate isomers in M. igneus ethanol extracts by anion-exchange chromatography on QAE-Sephadex (the DIP is not significantly retained by the column, whereas the two glutamate isomers stick). The specific rotation of this material (using a concentration of 8 mg/ml) was determined to be []_D²⁰ = 2.39°. This value can be compared to those for the three different optical isomers of DIP that have been synthesized: L,L, +2.4°; L,D, 0.0; and D,D, 2.8° (26). Thus, the DIP produced by the thermophilic methanogen has the same chirality (L,L) as the DIP produced by the archaeon Pyrococcus.

In vitro assays for enzymes of inositol biosynthesis. ³¹P NMR spectroscopy can be used to identify intermediates and products of inositol biosynthesis by monitoring the phosphorylated solutes produced from the incubation of known phosphocompounds (and appropriate cofactors) with protein extracts of M. igneus. The chemical shifts of substrates, products, and related compounds such as G-6-P (4.7 ppm), F-6-P (4.0 ppm), I-1-P (4.2 ppm), I-2-P (4.8 ppm), and P_i (2.6 ppm) are well separated and quite distinct from those for NAD⁺ (9 to 11 ppm), a cofactor in the I-1-P synthase activity. Both I-1-P synthase and I-1-P phosphatase have been well studied in plants and other organisms (2, 6, 17, 24), and the data obtained provide a background for evaluating these activities in methanogen cell extracts. Two ammonium sulfate fractions (0 to 44% and 44 to 85%, the latter more tightly defined as a 60 to 75% fraction in later experiments) from lysed M. igneus showed evidence of these enzymes. A summary of the cofactors needed for the different reactions identified in the DIP biosynthetic pathway, enzyme specific activities in these relatively crude extracts, and estimated activation energies are provided in Table 1.

1

1

1

1

1

1

1

1

1

1

1

1

View larger version (8K): [in this window] [in a new window]   FIG. 2.   1H-coupled 31P NMR spectrum (202.3 MHz) of an I-1-P synthase assay mixture containing 5 mM G-6-P, 1 mM NAD+, 50 mM Tris acetate (pH 8.2), and 100 µl of 60 to 75% ammonium sulfate protein extract that had been incubated at 85°C for 4 h. Note that only I-1-P is a doublet in the coupled spectrum.

1

1

View larger version (15K): [in this window] [in a new window]   FIG. 3.   1H-decoupled 31P spectra (202.3 MHz) for analysis of specific I-1-P phosphatase activities from incubations (55°C, 16 h) of mixtures containing 6 mM MgCl2, 50 mM Tris acetate (pH 8.0), and the following: 2 mM I-1-P and the 0 to 44% fraction (A), 2 mM I-1-P and the 44 to 85% fraction (B), 2 mM G-6-P and the 0 to 44% fraction (C), and 2 mM G-6-P and the 44 to 85% fraction (D).

Synthesis of DIP from myo-inositol and CDP-I. The proposed synthesis of DIP from I-1-P and myo-inositol requires activation of the I-1-P for condensation with myo-inositol. In lipid metabolism, CTP is used as a carrier molecule in phosphoryl and nucleotidyl transfers (25, 27). The nucleophilic character of I-1-P makes it a good candidate for a phosphoryl transfer reaction. The reaction carried out by CTP:I-1-P cytidylyltransferase (CDP-I synthase) would yield the intermediate, CDP-I (Fig. 1). DIP biosynthesis then proceeds similarly to the biosynthesis of phosphatidylserine. CDP-I is attacked by a myo-inositol hydroxyl group with the assistance of a general base provided by an enzyme. Such an activity would represent what we will refer to as a DIP synthase.

1

1

View larger version (14K): [in this window] [in a new window]   FIG. 4.   31P spectrum (121.4 MHz) for analysis of DIP synthase activity. Fifty microliters of CDP-I stock solution (30 mM), 300 µl of dialyzed cell extract, and 5 mM myo-inositol were incubated at 65°C for 2 h. The insets show expansions of the 3- to 5-ppm region, the phosphate monoester region, and the DIP region at ~1 ppm. Each phosphorus resonance in the spectrum is labeled.

View larger version (34K): [in this window] [in a new window]   FIG. 5.   1H-1H TOCSY NMR spectrum of DIP partially purified from the DIP synthase assay mixture. The vertical dotted line identifies all six DIP inositol protons in the ring spin system.

¹³C labeling of DIP and the pentose phosphate pathway. ¹³C label incorporation from specific precursors has provided information on a wide array of biosynthetic pathways (e.g., reductive or oxidative segment of the tricarboxylic acid cycle, amino acid biosynthesis, and -amino acid osmolyte biosynthesis) in methanogens (7-11, 18, 19, 23). Use of a double-labeled molecule such as [2,3-¹³C]acetate or [2,3-¹³C]pyruvate as a precursor allows easy identification of small amounts of label incorporation since a ¹³C-¹³C doublet flanks the singlet for ¹³C-¹²C (9) when the ¹³C₂ unit is incorporated intact. Label incorporation into DIP can be predicted based on its proposed biosynthetic pathway (Fig. 6). The doubly labeled precursor should label four of the DIP carbons (C-1, C-2, C-5, and C-6), while [3-¹³C]pyruvate should label only two, C-1 and C-6. The carbons for these species have been identified in natural-abundance ¹³C NMR spectra of ethanol extracts of M. igneus (5). In the ¹H-decoupled ¹³C NMR spectrum of DIP (Fig. 7A) obtained from labeling of M. igneus cultures with [2,3-¹³C]pyruvate, ¹³C-¹³C coupled doublets clearly flank the natural-abundance peaks of the C-1, C-4, and C-5 nuclei. For C-1, each resonance is also split by a small ¹³C-³¹P coupling so that each line in the multiplet is a doublet as well. The 71.8- to 73.0-ppm region of the spectrum is characterized by an extensive overlap of the resonances that correspond to the nuclei of the C-6, C-2, and C-3 atoms. Although the ¹³C-¹³C splitting pattern flanking ¹³C-¹²C singlets is harder to decipher, the complexity in the region suggests that ¹³C enrichment occurred for all carbon atoms.

View larger version (19K): [in this window] [in a new window]   FIG. 6.   Predicted 13C labeling for DIP using [2,3-13C]pyruvate: *, label from C-3 of pyruvate; , label from C-2 of pyruvate.

View larger version (22K): [in this window] [in a new window]   FIG. 7.   1H-decoupled 13C-NMR spectra of DIP obtained from labeling of M. igneus cultures with [2,3-13C]pyruvate (A) and [3-13C]pyruvate (B). Identities of the different DIP carbons are indicated by numbers above the resonances. Relative enrichment of DIP carbons in panel B is monitored by the integrated intensities.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

DIP is a novel osmolyte in archaea and closely related eubacteria that appears to track thermophily. Understanding this correlation on a molecular level requires a detailed understanding of DIP biosynthesis in these organisms. The in vitro assays and ¹³C labeling presented in this work confirm a pathway for DIP synthesis in M. igneus that consists of the following five steps: (i) production of G-6-P by gluconeogenesis and scrambling of glucose label via enzymes of the pentose phosphate pathway; (ii) generation of L-I-1-P by the enzymatic action of I-1-P synthase; (iii) conversion of I-1-P to myo-inositol by a specific phosphatase, I-1-P phosphatase; (iv) conversion of L-I-1-P to CDP-I with CTP as the carrier molecule in the phosphoryl transfer of this reaction; and (v) nucleophilic attack of the free hydroxyl group of C-1 of myo-inositol on CDP-I to produce DIP. Steps ii and iii have been elucidated in detail in plants and other organisms. Analogs of step iv occur in many cells as part of phospholipid biosynthesis, although this is not the way phosphatidylinositol is synthesized in cells (CDP-diacylglycerol is the activated species). Step v, the novel DIP synthesis reaction, is based on phosphatidylinositol synthesis in both eukaryotic and bacterial cells with myo-inositol attacking CDP-I instead of the CDP-diacylglycerol. Step v is the direct synthesis of DIP, and this reaction was shown to occur with chemically synthesized CDP-I and free myo-inositol in crude extracts of M. igneus. Understanding the mechanistic details of how steps iv and v are carried out will require purifying the activities from cell extracts.

One of the intriguing results of the in vitro studies is that the temperature dependence of the different enzymes sheds light on why DIP is accumulated only in M. igneus grown above 80°C. The synthesis of DIP from CDP-I and myo-inositol has quite a high activation energy, ~120 kJ · mol¹ (Table 1). For comparison, the start of the DIP biosynthetic pathway, I-1-P synthase, exhibits an activation energy of 60 to 70 kJ · mol¹ (Table 1). The I-1-P phosphatase in crude mixtures has an even lower activation energy (<50 kJ · mol¹), although since multiple phosphatases exist, this value may not be very accurate. A purified I-1-P phosphatase from M. jannaschii expressed in Escherichia coli has an activation energy of 54 kJ · mol¹ (3). At lower growth temperatures, any I-1-P generated may be hydrolyzed to myo-inositol, possibly for incorporation into lipids. I-1-P cannot accumulate to levels needed for conversion to CDP-I and eventual DIP synthesis. As the growth temperature is increased, however, not only is more I-1-P generated, but DIP synthase activity is enhanced. Cells need ~0.15 µmol of DIP per mg of protein for osmotic balance when grown in normal medium containing 0.3 M NaCl (5). The rate of DIP synthesis that is extrapolated for growth at 85°C, ~0.01 µmol · min¹ · mg¹, easily provides this amount in the doubling time of the organism. Whether the DIP really has a thermoprotective role or whether accumulation of DIP represents a way for the cell to use excess inositol (by storing it in a compound used for osmotic balance) in this organism remains to be determined.

As a by-product of analyzing the pathway for biosynthesis of DIP by ¹³C labeling, we have found that M. igneus must have many of the enzymes of the pentose phosphate pathway. Yu et al. (28) have suggested that a nonoxidative pathway for pentose production may be advantageous for an autotroph since it would avoid release of CO₂. However, the existence of transaldolase and transketolase activities in this methanococcus is in direct contrast to what has been observed with a number of other methanogens. Representatives of other genera (e.g., Methanobacterium thermoautotrophicum and Methanospirillum hungatei) do not show evidence of pentose phosphate enzymes, and it has been suggested that pentoses in these methanogens are formed by the oxidative decarboxylation of hexoses (7). Other methanococci have been examined by either ¹³C NMR or enzymatic assays specifically for these activities, with mixed results. ¹³C labeling of M. jannaschii (23), a hyperthermophilic organism, indicated that pentose phosphate enzymes were not present since no ¹³C label scrambling was observed: [3-¹³C]pyruvate labeled only C-1 and C-6 of hexoses. Another mesophilic methanococcus, M. maripaludis, showed significant transaldolase, transketolase, and appropriate epimerase activities (28). However, no ¹³C labeling experiments were carried out to verify that hexose labels would be scrambled. These limited comparisons show no correlation between thermophily and a nonoxidative pathway for pentose biosynthesis, since M. igneus appears more like the mesophilic M. maripaludis in possessing active transaldolase, transketolase, and epimerase activities.