Analysis of Dictyostelium discoideum Inositol Pyrophosphate Metabolism by Gel Electrophoresis

The social amoeba Dictyostelium discoideum was instrumental in the discovery and early characterization of inositol pyrophosphates, a class of molecules possessing highly-energetic pyrophosphate bonds. Inositol pyrophosphates regulate diverse biological processes and are attracting attention due to their ability to control energy metabolism and insulin signalling. However, inositol pyrophosphate research has been hampered by the lack of simple experimental procedures to study them. The recent development of polyacrylamide gel electrophoresis (PAGE) and simple staining to resolve and detect inositol pyrophosphate species has opened new investigative possibilities. This technology is now commonly applied to study in vitro enzymatic reactions. Here we employ PAGE technology to characterize the D. discoideum inositol pyrophosphate metabolism. Surprisingly, only three major bands are detectable after resolving acidic extract on PAGE. We have demonstrated that these three bands correspond to inositol hexakisphosphate (IP6 or Phytic acid) and its derivative inositol pyrophosphates, IP7 and IP8. Biochemical analyses and genetic evidence were used to establish the genuine inositol phosphate nature of these bands. We also identified IP9 in D. discoideum cells, a molecule so far detected only from in vitro biochemical reactions. Furthermore, we discovered that this amoeba possesses three different inositol pentakisphosphates (IP5) isomers, which are largely metabolised to inositol pyrophosphates. Comparison of PAGE with traditional Sax-HPLC revealed an underestimation of the cellular abundance of inositol pyrophosphates by traditional methods. In fact our study revealed much higher levels of inositol pyrophosphates in D. discoideum in the vegetative state than previously detected. A three-fold increase in IP8 was observed during development of D. discoideum a value lower that previously reported. Analysis of inositol pyrophosphate metabolism using ip6k null amoeba revealed the absence of developmentally-induced synthesis of inositol pyrophosphates, suggesting that the alternative class of enzyme responsible for pyrophosphate synthesis, PP-IP5K, doesn’t’ play a major role in the IP8 developmental increase.


Introduction
The model organism Dictyostelium discoideum, originally developed to study the transition to multicellularity, has subsequently been utilised in several areas of biology from chemotaxis [1] to transcriptional control [2].Upon exhaustion of nutrients, the Dictyosteliidae slime moulds are able to aggregate into multicellular forms, a process regulated by cAMP signaling [3].The aggregated slugs develop into fruiting bodies (or sporocarp) compromised of two main cell types; stalk cells and thousands of spore cells.Much of the early work with this amoeba focused on this fascinating behaviour.However, in the late 1980s this model organism began to offer insight into the metabolism of inositol phosphates [4].In fact, it was in D. discoideum that the synthesis of inositol hexakisphosphates (IP 6 ) through direct phosphorylation of inositol was discovered [5].
D. discoideum has also been instrumental in the discovery of inositol pyrophosphates (also known as diphosphoinositol phosphates) (For reviews see [6,7]) molecules containing highly energetic pyrophosphate moiety(ies) recently implicated into the regulation of cellular homeostasis [8,9,10].Inositol pyrophosphates were identified in 1993 in D. discoideum [11] and in mammalian cell [12].During the 1990s the synthesis of the inositol pyrophosphate IP 7 (diphosphoinositol pentakisphosphate or PP-IP 5 ) and, in particular, IP 8 (bisdiphosphoinositoltetrakisphosphate or (PP) 2 -IP 4 ) was linked to the D. discoideum developmental program [13].Furthermore, thanks to the high concentration of these molecules in this amoeba, NMR could be used to resolve the isomeric nature of IP 7 and IP 8 extracted from D. discoideum cells.The structure of these isoforms -the 5PP-IP 5 isomer of IP 7 and the 5,6(PP) 2 -IP 4 isomer of IP 8 [14] are, to date, the only resolved structures of inositol pyrophosphates extracted from cells.
Despite the influence of this organism, D. discoideum has faded from the attentions of inositol phosphate scientists over time.The last study demonstrating the importance of inositol pyrophosphate in regulating Dictyostelium chemotaxy was published over 10 years ago [15].This disengagement is in part due to the emergence of another experimental model, the yeast Saccharomyces cerevisiae [16,17] but also to the difficulty in promptly labelling the amoeba with tritium inositol ( 3 H-inositol) [18].Therefore the use of routine Sax-HPLC (Strong anion exchange chromatography) to resolve the different radiolabeled inositol phosphates becomes cumbersome and expensive to apply to amoeba cells.Thus, chromatographic separation of inositol phosphates in Dictyostelium is normally performed using metal dye detector post-colum derivatization (MDD-HPLC) [11,19] requiring a dedicated three pump HPLC apparatus and therefore is not a widespread technology.
The recent discovery that higher inositol phosphates can be resolved by polyacrylamide gel electrophoresis (PAGE) [26] and visualised by simple staining, bypassing the need to use radio labelled material, has enormously improved in vitro studies of inositol pyrophosphates [27].In particular, this has facilitated characterisation of the inositol pyrophosphate synthesizing kinases; the inositol hexakisphosphate kinases (IP6Ks in mammals, Kcs1 in yeast) and the Diphosphoinositol pentakisphosphate kinases (PP-IP5Ks in mammals, Vip1 in yeast) [26].
In the current work we applied this PAGE technology to samples obtained from live cells, allowing us to analyse the in vivo inositol phosphate metabolism by PAGE for the first time.We demonstrated the existence of different inositol pyrophosphate species by both DAPI and Toluidine Blue staining, and reveal a complex metabolism comprising inositol pyrophosphates derived from both IP 6 and inositol pentakisphosphates (IP 5 ).Furthermore, the analysis of inositol pyrophosphate metabolism during D. discoideum development revealed a far less dramatic increase in levels of IP 8 than has been previously described.

Results
Resolving D. discoideum Cell Extract by PAGE Revealed the Presence of IP 6 , IP 7 and IP 8 Inositol polyphosphates are routinely extracted using strong acid solutions, usually Percloric Acid.If appropriately labelled with 3 Hinositol this cell extract can be neutralised and analysed by Strong anion exchange chromatography (Sax-HPLC) [28].The high levels of inositol pyrophosphates in D. discoideum prompted us to analyse a fraction (1/20 by volume) of the neutralised unlabeled cells extract, equivalent to 1-2 10 6 cells, by PAGE [26].Sample migration during PAGE is normal when just 20-40 microlitres of cell extract is loaded (Fig. 1).To our surprise, extract from vegetative Wild Type AX2 (WT) D. discoideum cells reveals the presence of three major bands by Toluidine staining.The fastest migrating band co-migrates with the commercially available IP 6 standard (Fig. 1).Staining with DAPI also reveals the same three major bands (Fig. 1).Interestingly, DAPI is heavily photobleached (resulting in negative staining) by the two slower migrating bands and not by that which comigrates with IP 6 .This method of staining reveals a further weaker band, which migrates still slower and is not always detectable by Toluidine.It was previously demonstrated that the ability to photobleach DAPI is a typical characteristic of the pyrophosphate moiety [26], however the large amount of IP 6 present in D. discoideum extract is able to induce some DAPI photo-bleaching, though at a much lower efficiency.Thus, besides IP 6 the other bands are expected to be IP 7 , IP 8 and, newly detected, endogenous IP 9, previously identified only in vitro [23].
To confirm the nature of these bands we use several approaches.First genetic; the analysis of acidic extract of Inositol Hexakisphosphate Kinase (IP6K) null amoeba (gene I6KA, DDB_G0278739) [15] reveals only the presence of a band comigrating with IP 6 and a weaker band, detectable only by DAPI, co-migrating with IP 7 (Fig. 1).The virtual absence of IP 7 , and the total deficiency of the respective IP 8 and IP 9 bands phenocopies the yeast ip6k mutant (kcs1D) that lack any detectable inositol pyrophosphates [16,29].Using traditional 3 H-inositol labelling and Sax-HPLC analysis, the absence of any inositol pyrophosphates in ip6k null amoeba has been previously verified [15].Interestingly, DAPI analysis reveals the presence of a new, retarded band in ip6k null amoeba (labelled nd in Fig. 1B).This band is of undetermined nature although DAPI photobleaching ability suggests the presence of pyrophosphate moieties.The analysis of phospholypase C (PLC) mutant cells reveals a pattern of bands similar to WT cells in striking contrast of the yeast plc1D strain that lacks the synthesis of any highly phosphorylated forms of inositol phosphates [17].However, our result is coherent with previous reports that demonstrate normal levels of inositol pyrophosphates in D. discoideum plc null cells (gene pipA, DDB_G0292736) [18,30] and with the ability of the amoeba to synthesize IP 6 directly from inositol independently from lipid cleavage [5].Secondly, we use enzymology to confirm the nature of these bands as genuine inositol phosphates.The treatment of WT extract with phytase, an enzyme capable of fully dephosphorylating IP 6 (also call Phytic Acid), resulted in the complete disappearance of the three major bands (Fig. 2A).The treatment of WT extract with DDP1 (Fig. 2B), a phosphatases that specifically degrades the pyrophosphate moiety, resulted in a almost complete degradation of IP 7 and IP 8 with the corresponding formation of IP 6 .
Phosphoanhydride bonds (the pyrophosphate moiety) are rapidly hydrolysed in acid at higher temperatures.Consequently, we also incubated the acidic extract at 90uC for 10 minutes before neutralization.This treatment (Fig. 2C) revealed the complete degradation of the IP 7 and IP 8 bands and the resultant formation of IP 6 as well as three further fast migrating bands.These three extra bands migrate as expected of IP 5 , which is almost undetectable in untreated cell extract (Fig. 1, 2, 3).This suggests the existence of an elaborate inositol pyrophosphate metabolism (see below).
Finally we use mass spectrometry to determine the mass of the purified bands (Fig. 3A).The analysis of the putative IP 7 band reveals, in the m/z range 500-1000, two major peaks at 738.822 and 760.793 m/z.These m/z values are well in accordance with the theoretical mass of deprotonated IP 7 and its sodium adduct respectively (Fig. 3B).The detection of an intense peak at +22 m/z with respect to the deprotonated molecule is an additional confirmation of the presence of phosphate moieties on the analyte.Similarly, the analysis of the putative IP 8 band reveals a major peak at 840.590 m/z, and a minor peak at 818.627 m/z (Fig. 3C).These values correspond to the theoretical mass of, respectively, the sodium adduct of deprotonated IP 8 and deprotonated IP 8 .The increased relative intensity of the sodium adduct is an additional confirmation of the increased number of phosphate moieties attached to the analyte in the putative IP 8 band compared to the putative IP 7 band.
Unfortunately, we observed a decrease in mass spectrometry ionization efficiency with increasing number of phosphate groups, as can be appreciated by comparing absolute intensities of IP 7 and IP 8 MALDI-TOF spectra in (Fig. 3B and C).Therefore, we were unable to determine the mass of the IP 9 band.Nevertheless, taken together these genetic and biochemical studies prove that the bands observable in D. discoideum extract are bona fide inositol pyrophosphates.

Presence of a Complex IP 5 Derived Inositol Pyrophosphates Metabolism
The appearance of bands migrating faster that IP 6 after acidic hydrolysis prompted us to perform further analysis with the aim to determine their exact nature.We split a cell extract (from 5 ml culture growth at a density of 2-4610 6 cells/ml) into two halves and subjected one half to acidic hydrolysis.After neutralisation this half as well as the untreated half were analysed by PAGE.We employed IP 6 as a standard as well the six IP 5 isomers (Fig. 4A).Loading a larger amount of cell extract allowed us to detect fast migrating three weak bands in the untreated sample lane.These bands co-migrate with distinct IP 5 standards.The acidic treated samples reveals a robust increase of both IP 6 , due to the conversion of IP  species and thus cannot be directly detected in untreated WT cell extract.
Densitometry analysis of treated IP 6 and IP 5 s versus untreated counterpart reveals that ,60% of the IP 6 pool is converted to IP 7 and IP 8 while $90% of IP 5 pool is converted to inositol pyrophosphate species (Fig. 4B).This indicates that pyrophosphates derived from IP 5 and IP 6 have differing metabolism and turnover.The extraordinary ability of PAGE to resolve different IP 5 isomer and the densitometry analysis of the three IP 5 bands reveal that the IP 5 pool of D. discoideum cell is distributed as follows, ,10% I(1,2,3,4,6)P 5 ; ,30% I(2,3,4,5,6,)P 5 and/or I(1,2,4,5,6)P 5 ; ,60% I(1,3,4,5,6)P 5 and/or I(1,2,3,5,6)P 5 and/or I(1,2,3,4,5)P 5 .The presence of at least three IP 5 isomers is confirmed by an early report [5,31].However these earlier studies, which relied on HPLC revealed differing relative distributions of the IP 5 isomeric species [31].It is important to note that strong acidic conditions (such as those used in HPLC analysis) can induce phosphate groups to move to adjacent hydroxyl positions, altering the isomeric nature of inositol phosphates, a well known phenomena during phosphoinositide (inositol lipid) purification [32].High temperature and acidity are also able to induce movement of phosphate groups around the hydroxyl groups of the inositol ring in IP 5 (Figure S1).However, the presence of the three IP 5 species in untreated samples (Fig. 4A) is supportive of the genuine existence of at least three different IP 5 isomers in D. discoideum.The fact that all of these IP 5 species are enriched after pyrophosphate hydrolysis indicates that multiple IP 5 isomers are precursors of inositol pyrophosphate species indicating a complex isomeric mixture of pyrophosphates derived from IP 5 .

PAGE Analysis Revealed High Levels of Inositol Pyrophosphates
Previous studies aiming to analyse the level of inositol pyrophosphate during D. discoideum development have estimated the levels of IP 7 and IP 8 in vegetative stage cells to be 3-6% of the levels of IP 6 , a ratio comparable to that observable in mammalian cells.This data, obtained using MDD-HPLC [13], was subsequently confirmed by traditional metabolic 3H-inositol labelling and Sax-HPLC technology [15].Strikingly, however, our extraction and PAGE analysis reveals substantially higher levels of inositol pyrophosphates during vegetative state growth (Fig. 1, 2, 3, 4).DAPI analysis reveals a markedly darker stain of the IP 8 band over IP 6 (Fig. 1B).This can be attributed to the favourable ability of the pyrophosphates moiety to photobleach DAPI [26].The monoaminic Toluidine, however, stains the single phosphates groups with similar efficiency.Therefore a molecule of IP 8 possessing 8 phosphate groups, compared to just 6 on a molecule of IP 6 groups should stain more intensely than IP 6 .Experimentally this value has been calculated to be 1.27+/20.08(see material and methods for details and Figure S2).However, even taking into account this correction factor, densitometry measurement of PAGE analysis revealed the ratio of IP 8 to IP 6 in the vegetative state to be in the range 30-40% (Fig. 1, 2, 3, 4).Therefore, traditional HPLC technology substantially underestimates the level of cellular inositol pyrophosphates.It is likely that this effect is due to the fact that pyrophosphate (phosphoanhydride) bonds are acid labile and prone to degradation during acidic HPLC running conditions.
To further confirm this observation we ran cultures in parallel; one labelled with 3 H-inositol and analysed by HPLC, while the second was run by PAGE and analysed by Toluidine staining.We rapidly extracted the inositol phosphates at 4uC to minimise the duration and effect of the acidic conditions.This parallel analysis, reveals that the IP 8 level as ratio over IP 6 was 27.5%+/26.9(n = 4) and 36.3%+/24.7 (n = 4) analysed by Sax-HPLC or PAGE respectively.Therefore traditional HPLC analysis results in a substantial 1/4 underestimation of IP 8 cellular levels.
Application of PAGE to this in vivo system for the first time allows us to determine the intracellular concentration of highly phosphorylated inositol phosphates by simple densitometry (using IP 6 concentration standards to simply calculate a linear regression curve).Our study reveals that in vegetative state, estimating a cell volume of 0.20 pL, the concentration of IP 6 , IP 7 and IP 8 are ,520, 60 and 180 mM respectively.Interestingly, the IP 6 value is in accordance with previous estimates [5,33].

Inositol Pyrophosphates Cellular Levels Increase during Development
Dictyostelium development occurs upon exhaustion of food supply.This starvation response can be induced by shifting vegetative D. discoideum cells to agar plates made with a simple phosphate buffer (see material and methods).The regulation of inositol pyrophosphate metabolism during the slime mould's developmental program has been previously investigated [13].This study revealed the most dramatic cellular concentration change in IP 7 and IP 8 so far reported, with a 25-fold increase of IP 8 level [13].However, our obseravtions that IP 7 and IP 8 are present at high levels in vegetative cells (Fig. 1-4) led us to question the scale of this dramatic increase.Therefore we subjected WT and ip6k null cells to starvation, inducing the developmental program.
Cells were grown to a density of 2610 6 cells/ml, washed in phosphate buffer and then plated onto 20 mM phosphate buffer agar plates.Cells were collected at 5 time points; time zero, whilst still in the vegetative state; after one hour of starvation; upon first visual signs of aggregation (6-9 hr depending on strain); during the ''slug'' stage 15-17 hs after induction of starvation and finally after 24 hr as mature fruiting bodies.Analysis of acidic cell extract shows a clear increase of IP 8 (in comparison to IP 6 ) of 2,6 fold during the developmental time course.Therefore, although we observe a clear and substantial increase in IP 8 levels (Fig. 5A), it is in the region of three fold, well below the 25-fold seen previously by HPLC analysis [13].
We also performed developmental study of ip6k null stain (Fig. 1) [15].This analysis revealed the lack of induction of any inositol pyrophosphate forms (Fig. 5B).This data indicates that the D. discoideum PP-IP 5 K homologous gene (DDB_G0284617) does not play any major role in the developmental increase of inositol pyrophosphates.
The D. discoideum development program is elicited by cAMP signal and it was reported that cAMP stimulation induced a rapid (within minutes) threefold increase in inositol pyrophosphate levels [15].We repeated these studies and failed to see any significant change in IP 7 and IP 8 levels in response to cAMP when analysed by PAGE (Fig. 6).

Discussion
The recently developed PAGE technology to resolve and visualise inositol phosphates has been previously employed to characterise in vitro enzymatic reaction [26,27].Here we show the huge potential of this technology to study inositol pyrophosphate metabolism in D. discoideum.The high abundance of this class of molecules in D. discoideum, coupled with the ease of analysis by PAGE has allowed us to re-evaluate the regulation of inositol pyrophosphate metabolism during the amoeba development.This re-evaluation has revealed a 3-fold increase in IP 8 levels reached in mature fruiting bodies a value far below the 25-fold increase that was previously determined by MDD-HPLC [13].MDD-HPLC technology requires the extracted samples to be resolved using an elution buffer containing 0.1M Hydrochloric Acid [13].This condition is likely to result in the hydrolysis of pyrophosphate bonds and thus increased variability between samples.Therefore, the difference of IP 8 induction during Dictyostelium development between our PAGE analysis and the previous study [13] is most likely due to the acidic sensitivity of the pyrophosphate moiety and its degradation during the strong acidic conditions associated with MDD-HPLC.
In agreement with this, analysis of 3 H-inositol labelled inositol pyrophosphates by Sax-HPLC, a technology that requires less acidic conditions (Ammonium Phosphate buffer at pH 3.8) reveals that the IP 8 ratio over IP 6 is ,27% higher than previous studies suggested [13].However, pyrophosphate hydrolysis still occurs in the mildly acidic Sax-HPLC running conditions.In fact, when the IP 8 ratio over IP 6 ratio was measured by PAGE and densitometry (sample resolved in 1XTBE, buffer pH8.0) this value was still higher at ,36%.Therefore both in vitro study [26] and also the current in vivo PAGE analysis suggest that HPLC analysis underestimates the cellular levels of inositol pyrophosphates.Unfortunately, technical problems still preclude the application of PAGE analysis to mammalian cells.As such, HPLC analysis remains, at least for now, the only viable method for this system.We have also demonstrated that PAGE can resolve several of the different IP 5 isomeric species (Fig. 4A).This has allowed us to observe at least three different IP 5 isomers in vegetative D. discoideum consistent with early reports [5,31].This is not a surprise as mammalian cells also possess multiple IP 5 species [34,35].Unexpectedly, all the IP 5 s species are precursors of inositol pyrophosphates.Therefore, the inositol pyrophosphates derived from IP 5 are quite complex in compositions, with multiple isomeric forms of PP-IP 4 and (PP) 2 -IP 3 existing in vivo.The same enzymes that generate IP 7 from IP 6 , the IP6Ks, also generate these inositol pyrophosphates species in vitro [16,29].Thus the relative cellular abundance between IP 7 /IP 8 and PP-IP 4 /(PP) 2 -IP 3 might depend on the levels of IP 5 versus IP 6 .While in D. discoideum IP 6 is far more abundant of IP 5 this is not the case the majority of mammalian cell lines where the cellular concentration of these two inositol polyphosphates are similar and regulated in neuron by neurotrofine signal [36].Consequently inositol pyrophosphate derived from both IP 5 and IP 6 precursors are likely to have similar cellular abundance and physiological importance.More attention needs to be invested to study the functions of the IP 5 derived inositol pyrophosphates.The fact that PP-IP 4 is often undetected on Sax-HPLC (due to co-migration with IP 6 ), is neither an indicator of its absence nor of a lack of physiological roles.
The simple, inexpensive and reliable PAGE analysis leads to clear qualitative and quantitative information by using simple densitometry.Classical 3 H-inositol labelling and Sax-HPLC analysis of highly phosphorylated inositol perhaps retains the advantage of a higher dynamic range to calculate the relative abundance of the inositol polyphosphates.However, the use of radioactive material and HPLC apparatus has limited the implementation of inositol phosphate research to the large majority of cell biology laboratories.Here we have demonstrated the huge potential of PAGE technology to study D. discoideum inositol phosphate metabolism.PAGE analysis coupled with the generation of the knockout mutant strains for the several inositol phosphate kinases present in the amoeba genome will create a genetic system that will easily surpass the S. cerevisiae model, due to the higher complexity and greater similarity to the mammalian system.
The obvious future objective is to apply this PAGE technology to analyse inositol pyrophosphate metabolism in mammalian experimental models.However, because of the low abundance of inositol polyphosphate in mammalian cells, the direct application of PAGE technology to this system is not yet possible (Saiardi lab unpublished result).On the other hand, the effortless nature of PAGE technology should encourage further effort towards this goal, thereby opening new avenues for investigation.

Culture Condition and cAMP Treatment
Amoeba cells were inoculated at a density of 1610 5 cells/ml in HL5 in a glass flask, and incubated with shaking at 22uC, 120 RPM.To keep the cells in the vegetative state, stock cultures were diluted every 2-3 days such that cell density didn't surpass 5-6610 6 cells/ml.Every 2-3 weeks new D. discoideum were started from DMSO stock.Treatment with cAMP was performed on active growing vegetative stage cell.cAMP was added to a final concentration of 50 mM to100 ml of cells and the treatments were terminate by adding 100 ml of Percloric Acid 2 M to initiate inositol phosphate extraction.

Inositol Phosphates Extraction
The inositol polyphosphate extraction procedure is an adaptation of the yeast protocol previously described [28].D. discoideum cells were collected during the exponential growth phase (1-3610 6 cells/ml) washed twice with KPO 4 H buffer 20 mM pH 6.0 and centrifuged at 1500 RPM on a Sorval RC-3C centrifuge for 3 min.The cell pellets were transferred to eppendorf tubes, resuspended in 1 M Perchloric acid, vortexed for 5 min at 4uC and centrifugated at 14000 RPM at 4uC for 5 min.The supernatants were transferred to a new tube and neutralised using 1M Potassium Carbonate containing 3 mM EDTA.The samples were placed on ice for 2-3 hours and subsequently spun for 10 min.The supernatants were transferred to new tubes and stored at 4uC.If required, the supernatants volume was reduced using a speed vacuum.

PAGE Analysis and Band Intensity Analysis
To resolve inositol phosphates we used 2461660.1 cm glass plates, using 35% polyacrylamide in 1XTBE.Samples were mixed with 66Dye (0.01% Orange G or Bromophenol Blue; 30% glycerol; 10 mM TrisHCl pH 7.4; 1 mM EDTA).Gels were prerun for 30 min at 300 V and run at 600 V 6 mA overnight at 4uC until the Orange G had run through 2/3 of the gel.Gels were stained with DAPI or Toluidine Blue as described previously [26].After scanning, the Tiff format file, band densitometry was performed using ImageJ software (http://rsbweb.nih.gov/ij/).To determine the differential Toluidine Blue staining efficiency pure IP 8 and IP 7 were converted to IP 6 by acid hydrolysis and resolved by PAGE.The different densitometry intensity of IP 7 and IP 8 untreated samples versus the generated IP 6 was then calculated (Supporting Fig. S2).IP 8 is 1.27+/20.08(average +/2 standard deviation, n = 5) times more strongly labelled that than corresponded generated IP 6 , in good accordance with the theoretical value of 1.33.To determine the amount of inositol phosphates present in D. discoideum, cell extracts were run together with IP 6 concentration standards from 1 nM to 8 nM.By determining the densitometry value of the IP 6 standard a linear regression curve was calculated.The densitometry value of the IP 6 present in the cell extract was calculated from the linear regression curve to determine its molar amount.The values for IP 7 and IP 8 were calculated determining the densitometry of the respective bands normalised for the Toluidine staining efficiency (1.27 for IP 8 and 1.15 for IP 7 ).The cellular concentration of inositol phosphate was then calculated estimating a cell volume of 0.20 pL.

Enzymatic Reactions and Acid Idrolysis
Neutralised D. discoideum cell extract, or purified inositol phosphate, were incubated in 30 ml enzymatic reactions containing 5XBuffer (100 mM Hepes 6.8; 250 mM NaCl; 30 mM MgSO4; 5 mM DTT; 5 mM NaF), 2 ml of recombinant purified Ddp1 (10-2-ng) or Phytase (Sigma).Reactions were incubated at 37uC for 2 hr or overnight and stopped by the addition of 2 ml EDTA (100 mM).Acidic pyrophosphate hydrolysis was performed by incubating the cell extract at 90uC for 20 min prior to neutralisation.After this treatment the samples were neutralised using Potassium Carbonate as described above.

Mass Spectrometry
Inositol phosphates from D. discoideum cell extract were purified as described above and previously [28] and directly subjected to mass spectrometry [37].Matrix-assisted laser desorption ionization (MALDI) mass spectrometry was performed on a Voyager DE-STR (Applied Biosystems, Framingham, MA), equipped with a MALDI ion source and a time-of-flight mass analyzer (MALDI-TOF).9-aminoacridine (9-AA, Sigma-Aldrich) was used as matrix, due to its superior performance in revealing acidic analytes in negative ion mode [37].A double deposition sample preparation procedure was adopted.Typically, 0.5 mL of matrix solution, consisting of 7 mg/mL 9-AA in a 1:1 mixture (v/v) of acetonitrile and water was spotted on the stainless steel MALDI sample stage and air-dried.Then, 0.5 mL of the analyte solution, either pure or diluted 1:5 (v/v) in water, was spotted on to the matrix crystals and allowed to dry.Mass spectra were acquired in delayed extraction, reflectron negative ion mode using the following settings: accelerating voltage 20,000 V, grid voltage 73%, extraction delay time 300 nsec, acquisition mass range 300-1,500 m/z.Each spectrum was the average of 400-500 individual laser shots acquired in series of 100 consecutive shots.

Sax-HPLC Analysis
D. discoideum were cultured in inositol free SIH media containing 50 mCi/ml [ 3 H]inositol.Cell culture (6 ml) were seeded at 1610 5 cells/ml and grown at 22uC for 3-4 days to get a cell density of 2-3610 6 /ml.Cells were collected and washed once with KPO 4 H buffer 20 mM pH 6.0.Inositol phosphates were extracted as described above and resolved by HPLC as previously described [28].

D. discoideum Development
D. discoideum were cultured in HL/5 media to a density of 2.0610 6 cells/ml The cells were washed twice with KPO 4 H buffer 20 mM pH 6.0 and resuspended at 1610 7 cells/ml in the same buffer Cells were then transferred in solution to 35 mm, 20 mM phosphate agar plates such that each plate contained 1610 7 cells.The cells were allowed to settle before aspirating the phosphate buffer.The cells were then allowed to develop in a humidity chamber at 22 C. 10 plates (equivalent to a 1610 8 cells at the start of the time course) were harvested from plates at 5 time points; 0 hr during vegetative state; after 1 hr starvation; upon first signs of aggregation (6-9 hr); during the ''slug'' stage (15 hr); and finally as mature fruiting bodies (24-25 hr).Cells pellets were frozen at 280uC.Inositol polyphosphates were extracted as described above, normalised by protein concentration and analysed by PAGE.

Supporting Information
Figure S1 IP 5 isomerisation by acid treatment.To verify that acid treatment of IP 5 can induce movement of phosphate groups around the inositol ring we incubated two nanomols of IP 6 and two nanomols of I(1,3,4,5,6)P 5 with 1M Percloric acid for 30 min in ice as well as for 5 and 30 minutes at 900C.IP 6 is totally unaffected by these treatments.Untreated I(1,3,4,5,6)P 5 (lane 2) is 95% pure as demonstrated by its migration as a major single band.Low temperature acid treatment has no effect on I(1,3,4,5,6)P 5 , whilst high temperature induces rapid isomerisation.Just five minute at high temperature are sufficient to substantially convert I(1,3,4,5,6)P 5 into other IP 5 isomeric forms.Densitometry analysis confirmed that the total IP 5 Toluidine staining did not change upon acid treatment, indicating the absence of acid-induced IP 5 degradation to lower inositol phosphates.(PDF) Figure S2 Differential Toluidine Blue staining capability of IP 8 and IP 6 .To ascertain the relative efficiency of staining of IP 8 and IP 6 by Toluidine blue serial amounts of IP 8 from 1 nmol (A) to 16 nmol (E) were incubated in the presence of 1M Percloric acid in 20 ml (sample from A9 to E9) for 30 min at 90uC.Untreated (from A to E) and acid-treated (from A9 to E9) samples were resolved on 35% PAGE.To avoid loss of material during the neutralization step, acid-treated samples were directly loaded on the gel causing a slight retardation in migration in these lanes (as shown by the different migration of Bromophenol blue (BBF) between treated and untreated samples).Once stained with Toluidine blue, the gel was analysed with ImageJ software.Densitometry analysis enabled each pair of samples (treated and untreated) to be plotted on a graph.The areas of the peaks in these graphs correspond to the relative staining of the IP 6 and IP 8 bands on the gel.Depicted are the analyses of samples D-D9 and E-E9.Dividing the densitometry derived values for untreated IP 8 by those for the acid-generated IP 6 indicates the difference in staining efficiency of the two molecules by Toluidine blue.On average IP 8 is stained 1.27+/20.08(+/2 SD) better that IP 6 .A virtually identical result was obtained from a second, independent experiment also run in quintuplicate.The experimentally calculated value of 1.27 is in good accordance with the theoretical value of 1.33 reflecting the presence of eight phosphates groups in IP 8 rather than the six in IP 6 (8/6 = 1.33).(PDF)

Figure 1 .
Figure 1.PAGE analysis of D. discoideum cell extract reveal the presence of three major bands.Inositol phosphates were extracted from 10 ml culture growth of wild type (WT) D. discoideum (AX2 strain) and the IP 6 -Kinase (ip6k null) and phospholipase C mutant (plc null) grown at a density of 2-4610 6 .About 30-40 microliters of neutralised cell extract (equivalent to 1/20 of the total volume) was resolved on 35.5% PAGE [26] and visualized with Toluidine blue (A) and DAPI staining (B).The figure shows the result of a representative experiment that was repeated three times.doi:10.1371/journal.pone.0085533.g001 7 and IP 8 to IP 6 , and the three IP 5 isomers.This indicates presence of inositol pyrophosphate generated from IP 5, such as PP-IP 4 and likely also (PP) 2 -IP 3 , which are converted back to IP 5 by acidic treatment.The inositol pyrophosphates PP-IP 4 and (PP) 2 -IP 3 possessing six and seven phosphates groups migrate very closely (or co-migrate) with the more abundant IP 6 and IP 7

Figure 2 .
Figure 2. Treatment by Phytase, Ddp1 and acidic degradation define IP 6 , IP 7 , and IP 8 , in D. discoideum cell extract.Wild type D. discoideum cell extract (-) was incubated with phytase (Phy) (A), recombinant diphosphoinositol polyphosphate phosphohydrolase (DDP1) (B) or treated with acid at high temperature (C).The inositol phosphate nature of the three major bands detectable by Toluidine stain is demonstrated by the Phytase treatment (A), an enzyme able to remove the phosphate group from any position of the inositol rings.The pyrophosphate nature of the two slower migrating bands is demonstrated by their disappearance after DDP1 treatment (B) and by the well known acidic sensitivity of the phosphoanhydride bond (C).The figure shows the result of a representative experiments repeated three to four times.doi:10.1371/journal.pone.0085533.g002

Figure 3 .
Figure 3. Mass spectrometry analysis of inositol pyrophosphates purified from D. Discoideum cell extract.Gel purified inositol pyrophosphates (A) were subjected to mass spectrometry (B,C).The comparison of the m/z spectrum of IP 7 (B) and IP 8 (C) is shown.The peaks in the spectra describing inositol pyrophosphates purified from D. discoideum are in agreement with the theoretical values for molecular weight that are deduced to be 738.82Da and 818.78 Da respectively.doi:10.1371/journal.pone.0085533.g003

Figure 4 .
Figure 4. Characterization of D. discoideum IP 5 species.Half of the acidic cell extract (from 5 ml culture) of WT D. discoideum was incubated on ice (-) while the second half was incubated at 90uC for 20 min (Acid).Both samples were thenneutralised and resolved on 35.5% PAGE together with the six possible IP 5 isomers.Inositol phosphates were visualised by Toluidine staining.Densitometry analysis of treated versus untreated sample was performed and IP 6 and IP 5 s bands intensity compared.(A).Acidic treatment reveals the distinct presence of three IP 5 species, which are otherwise barely detectable, indicating that D. discoideum possesses a complex IP 5 -derived inositol pyrophosphate metabolism.(B) Schematic representation of inositol pyrophosphate metabolism in D. discoideum.The gray arrow to (PP) 2 -IP 3 indicates a likely potentiallye step.The dashed arrow from inositol to IP 3 indicates uncharacterized enzymatic steps.The figure shows the result of a representative experiment that was repeated three times.doi:10.1371/journal.pone.0085533.g004

Figure 5 .
Figure 5. PAGE analysis of inositol pyrophosphate during D. Discoideum development.Amoeba development program was induced as described in material and methods.The inositol phosphates extracted at the indicated time points were resolved on 35% PAGE and visualised with Toluidine.(A) The analysis of wild type (WT) D. discoideum developmental program reveal a 2.6fold increase in the IP 8 /IP 6 ratio at the late stage of development, as quantified by densitometry quantified (Bottom), average +/2 SD of four independent experiments.(B) To the contrary inositol pyrophosphates are not induced during IP 6 -Kinase (ip6k null) developmental program.The figure shows the result of a representative experiment that was repeated four times for the WT and two times for ip6k1 null.doi:10.1371/journal.pone.0085533.g005

Figure 6 .
Figure 6.No alteration of IP 7 and IP 8 metabolism after cAMP treatment.Vegetative growing D. Discoideum were incubated for the indicated time with 50 mM cAMP.The incubation was terminated with equal volume of 2M Percloric acid to extract the inositol phosphates.These were resolved on 35% PAGE and stained with Touilidine blue.Two independent experiments are shown with short (left) and long (right) cAMP incubation time.The figure shows the result of a representative experiment that was repeated three times.doi:10.1371/journal.pone.0085533.g006