Increasing Phosphatidylinositol (4,5)-Bisphosphate Biosynthesis Affects Basal Signaling and Chloroplast Metabolism in Arabidopsis thaliana

One challenge in studying the second messenger inositol(1,4,5)-trisphosphate (InsP3) is that it is present in very low amounts and increases only transiently in response to stimuli. To identify events downstream of InsP3, we generated transgenic plants constitutively expressing the high specific activity, human phosphatidylinositol 4-phosphate 5-kinase Iα (HsPIPKIα). PIP5K is the enzyme that synthesizes phosphatidylinositol (4,5)-bisphosphate (PtdIns(4,5)P2); this reaction is flux limiting in InsP3 biosynthesis in plants. Plasma membranes from transgenic Arabidopsis expressing HsPIPKIα had 2–3 fold higher PIP5K specific activity, and basal InsP3 levels in seedlings and leaves were >2-fold higher than wild type. Although there was no significant difference in photosynthetic electron transport, HsPIPKIα plants had significantly higher starch (2–4 fold) and 20% higher anthocyanin compared to controls. Starch content was higher both during the day and at the end of dark period. In addition, transcripts of genes involved in starch metabolism such as SEX1 (glucan water dikinase) and SEX4 (phosphoglucan phosphatase), DBE (debranching enzyme), MEX1 (maltose transporter), APL3 (ADP-glucose pyrophosphorylase) and glucose-6-phosphate transporter (Glc6PT) were up-regulated in the HsPIPKIα plants. Our results reveal that increasing the phosphoinositide (PI) pathway affects chloroplast carbon metabolism and suggest that InsP3 is one component of an inter-organelle signaling network regulating chloroplast metabolism.

Fluctuations in cytosolic calcium occur in the light and dark and have a circadian rhythm [31][32][33][34][35]. Furthermore, in plants, the chloroplast is a major store of intracellular calcium and stromal calcium has been reported to change with light/dark transitions [35][36][37]. Chloroplast stromal calcium is low in the light and increases transiently for about 20 min at the end of day/beginning of dark. Photosynthetic electron transport is not required for dark-induced stromal calcium changes suggesting that proton motive force is not essential for the stromal calcium increase during the light/dark transition [37]. The transient increase in stromal calcium in the dark has been proposed to contribute to the down regulation of Calvin-Benson cycle enzymes such as fructose 1,5-bisphosphatase (FBPase) and seduloheptulose 1,7-bisphosphatase (SBPase) and to the dark deactivation of the ATP synthase [36,[38][39][40]. While there are several reports indicating a role for chloroplast calcium and changes in stromal calcium during the light/dark transition, the role of cytosolic calcium in regulating chloroplast metabolism remains a conundrum [40,41].
The earliest evidence for a role of the PI pathway and light signaling was from the work of Ruth Satter's laboratory using Samanea samman pulvini [42]. Subsequently, it was shown that increases in InsP 3 were associated with light-induced shrinking of flexor cells [43]. More recently, blue light signaling was correlated with changes in InsP 3 in Arabidopsis seedlings [12]. Notably, Chen et al. [12], found that in Arabidopsis seedlings, InsP 3 was higher in wild type seedlings in the light relative to the dark. Additional evidence that changes in InsP 3 correlate positively with light/dark transitions comes from two studies. In C4 plants, phosphoenolpyruvate phosphate carboxylase (PEPC) is activated in the light by phosphorylation by PEPC kinase. Coursol et al. [44] showed that an increase in InsP 3 preceded the increase in PEPC kinase activity. In a separate study, Arabidopsis plants with mutation in sac9, a PtdInsP 2 ptase, had increased InsP 3 [45] and were identified in a screen for plants with a delay in dark adapted deactivation of the ATP synthase [46]. These studies suggest that fluctuations in InsP 3 could contribute to light/dark regulation in the chloroplast.
It is difficult to identify events downstream of InsP 3 in planta. One approach that has been used is to remove or dampen the InsP 3 signal. Perera et al. [47] expressed the more active human InsP 5-ptase and lowered basal InsP 3 in Arabidopsis plants. These InsP 5-ptase transgenic plants revealed that InsP 3 -mediated responses were a component of gravitational signaling (the gravitational response in both roots and shoots was delayed) and contributed to about 30% of the stimulus-induced cytosolic, aequorin-sensitive calcium signal in response to salt or cold [15]. While dampening the InsP 3 signal revealed a decrease in response to gravity attributable to InsP 3 , the targets of InsP 3 -mediated signaling were not identified and the effects of InsP 3 on plant responses have been questioned [48].
In this work, to identify InsP 3 -mediated events, we increased the biosynthesis of InsP 3 . Our approach was to increase the synthesis of PtdInsP 2 , the flux limiting step in plant PI metabolism [49], by expressing a green fluorescent protein (GFP)-fusion construct of the human phosphatidylinositol phosphate 5-kinase1α (HsPIPKIα) in Arabidopsis plants. HsPIPKIα has a lower Km for PtdInsP and a higher Vmax making it more effective than the Arabidopsis PIPKs [50]. Plants expressing the HsPIPKIα had more than 2-fold increased PtdInsP 2 and InsP 3 in the leaves. There was a 10% decrease in total calcium suggesting a net efflux of calcium in response to increased InsP 3 as was found when HsPIPKIα was expressed in tobacco cells grown in suspension culture [49] suggesting that the InsP 3 -sensitive component of the organelle mobile calcium stores might be depleted in these cells. We found the HsPIPKIα expressing plants have higher starch both at the end of day and end of night suggesting decrease in transitory starch turnover and delay in the dark adaptation of the Calvin-Benson cycle. In addition, the HsPIPKIα plants were drought sensitive, but seedlings were more heat and light tolerant than the controls. While at first this seems counter intuitive, i.e., higher cytosolic InsP 3 should increase calcium signaling, it is possible that the constitutively increasing InsP 3 in the cytosol decreased the stores of cellular calcium and decreased or delayed dark adaptation and responses to other environmental cues. In summary, we demonstrate that increasing the flux through the PI pathway in plants affects chloroplast carbon metabolism and plant responses to environmental stress, and we hypothesize that InsP 3 -mediated signaling contributes to coordinating inter-organelle metabolism in plants. Future studies monitoring organelle calcium are needed to test this hypothesis.

Generation and Growth of HsPIPKIα Transgenic Plants
Three independent transgenic Arabidopsis lines carrying the GFP fused human HsPIPKIα construct under the control of the cauliflower mosaic virus 35S promoter were generated as described by Im et al. [49] by Agrobacterium-mediated transformation using vacuum infiltration. GFP-HsPIPKIα plants (hereafter noted as HsPIPKIα plants) are smaller than WT and GFP alone under normal short-day growth condition (8 h of light/16 h of dark) ( Figure 1A). Transcripts were confirmed by reverse transcription (RT)-PCR using internal GFP forward and reverse primers and HsPIPKIα forward and reverse primers ( Figure 1B). No transcript was detected in the wild type using both primer sets. GFP transcripts were detected in GFP expressing lines (GFP alone and the HsPIPKIα lines). In the HsPIP5KIα plants, GFP fluorescence was localized at the plasma membrane ( Figure 2A). HsPIPKIα seedlings have shorter roots and stunted and bulged root hairs compared to WT and GFP plants ( Figure 2B,C). Seedlings were grown under short-day cycle in MS media. The different types of root hairs were counted for WT, GFP and two independent transgenic lines, HsPIPK5-8 and HsPIPK9-7 when seedlings were 6 days old ( Figure 2D). A similar root hair phenotype has been described by others overexpressing the plant PIPKs in planta and has been associated with defects in vesicle trafficking and cell wall biosynthesis in tip growing cells [

PtdInsP 2 and PIPK Specific Activity Increased in Seedlings and Leaves
The HsPIPKIα 9-7 lines had a more pronounced bulged root hair phenotype ( Figure 2D) and the highest PIP 5-kinase activity ( Figure 3A). Plasma membranes were isolated from young seedlings and leaves of 1 month-old plants, and PtdInsP 5-kinase specific activities were measured in vitro. Exogenous PtdIns(4)P was added to the reaction mixture so that the substrate would not be limiting. Expression of the HsPIPKIα in Arabidopsis increased the production of [ 32 P]PtdIns(4,5)P 2 15 to 25-fold more in young seedlings and 2 to 3-fold more in 1 month-old plants compared to WT and GFP lines ( Figure 3A,B). As indicated by the in vitro assays ( Figure 3A) and immunoblot of isolated proteins ( Figure 3C), GFP-HsPIPKIα was recovered primarily in the plasma membrane (upper phase) that was separated by aqueous two-phase partitioning.
Head group analysis was used to measure the endogenous PtdInsP 2 . For these experiments, lipids were extracted from plasma membranes of transgenic and control seedlings and the inositol head group was hydrolyzed with HCl. The total Ins(1,4,5)P 3 released was measured using an Ins(1,4,5)P 3 assay kit. The transgenic plants had 2 to 2.5-fold increased PtdIns(4,5)P 2 compared to WT and GFP plants ( Figure 3D). B. C.

D.
Control HsPIPKIα Figure 3. Membrane-associated PtdInsP 5-kinase specific activity from 17-d-old seedlings (A) and leaves from 1 month-old plants (B); The plasma membrane (PM) and lower phase fraction (LP) from wild-type and transgenic plants were separated by aqueous two-phase partitioning and analyzed for PtdInsP 5-kinase activity with added substrate, PtdIns(4)P.
The data are the mean of duplicate values ± SD; (C) Immunoblot of proteins from PM or LP of leaves from 1-month-old plants visualized using an antibody to GFP; (D) Mass measurement of PtdInsP 2 based on head group analysis of lipids extracted from isolated plasma membranes of leaves from the 1-month-old plants.
In order to determine if the increased PtdIns(4,5)P 2 changed the major phospholipids in HsPIPKIα seedlings, total lipids were extracted as described in the Experimental Section. The major composition of the phospholipids, such as PtdGro, PtdEtn, PtdIns, PtdCho, PtdSer and PtdOH, and galactolipids, such as MGDG and DGDG, was not significantly different between WT, GFP and HsPIPKIα plants ( Figure 4; the data are presented in Supplemental Data File 1).

Increased Flux through the Phosphoinositide Pathway in HsPIPKIα Transgenic Plants
To monitor the rate of PtdIns(4,5)P 2 biosynthesis in vivo, we labeled the seedlings with 32 Pi and harvested at each time point indicated ( Figure 5A). Lipids were extracted and separated by thin layer chromatography (TLC). The incorporation of 32 Pi into PtdIns(4,5)P 2 was 5 to 7-fold higher in HsPIPKIα plants compared to WT and GFP plants and saturated by 20 min when calculated as total [ 32 P]-labeled lipids ( Figure 5B). The incorporation of 32 Pi into PtdInsP was ~40% less in HsPIPKIα plants compared to WT and GFP plants. This is likely a result of the fast conversion of PtdIns(4)P to PtdIns(4,5)P 2 in HsPIPKIα plants ( Figure 5C). We also labeled the seedlings with [ 3 H]myo-inositol for 4 days to monitor the levels of intermediates of the PI pathway. In the WT, the ratio of total cellular [ 3 H]PtdIns(4)P to [ 3 H]PtdIns(4,5)P 2 was ≥20:1, whereas the ratio was reduced to 2:1 in the HsPIPKIα plants (Table 1). Note there was ~20% decrease in [ 3 H]PtdIns(4)P in HsPIPKIα plants which would be anticipated with an increase in PtdIns(4,5)P 2 biosynthesis. The data are consistent with previous work indicating that PIPK activity is a flux-limiting step in the plant PI pathway [49,52].     To determine how increased PtdIns(4,5)P 2 levels would affect the total cellular Ins(1,4,5)P 3 levels, we measured the total Ins(1,4,5)P 3 in the soluble fraction of WT and HsPIPKIα plants using an Ins(1,4,5)P 3 assay kit ( Figure 6A). The basal Ins(1,4,5)P 3 levels were increased 2 to 4-fold in the seedlings and in leaves of 2-month-old plants of HsPIPKIα compared to WT and GFP plants ( Figure 6B). These data combined with the radioisotope labeling data indicate that the HsPIPKIα plants had increased flux in the PI pathway and were producing more InsP 3 . While InsP 3 can generate calcium oscillations in vivo, it can also produce higher ordered InsPs [26,53,54]. We monitored the production of InsP 6 using both isotope labeling and mass measurement. The [ 3 H]myo-inositol labeling of seedlings revealed a significant increase in [ 3 H]-labeled InsP 6 indicating that the increase in InsP 3 had affected higher ordered InsP biosynthesis ( Figure 6C). To assess total InsP 6 , we used isocratic ion chromatography as described in the Experimental Section. In a preliminary experiment, we did not detect differences in total InsP 6 in seedlings (data not shown).
Because it was difficult to obtain enough material to do InsP 6 mass measurements on the seedlings, we analyzed seeds ( Figure 7A). InsP 6 is produced by two pathways: a lipid-mediated pathway resulting from the phosphorylation of lipid-generated Ins(1,4,5)P 3 and a non-lipid-dependent pathway, which involves de novo synthesis and the sequential phosphorylation of myo-1L-inositol phosphate [53]. The non-lipid-dependent pathway is the dominant pathway in storage tissue [55] and as shown in Figure 7A, the seeds from the HsPIPK1α plants had 40% less total InsP 6 . Typically seeds with low InsP 6 have high Pi [56,57] and this is what we found for the HsPIPK1α. The seed HOAc-soluble Pi in the HsPIPK1α lines was almost twice that of the controls ( Figure 7B). In contrast, the seedling HOAc-soluble Pi was about 20% less than wild type ( Figure 7C) and there was no significant difference in HOAc-soluble Pi in 2-month-old mature leaves ( Figure 7D). These data suggest that down regulation of the non-lipiddependent pathway was compensating for the increased flux through the PI pathway in seedlings and leaves, and that in seeds where the non-lipid pathway was dominant, down regulation led to a net decrease in InsP 6 . More extensive flux analyses of both the lipid-and non-lipid mediated pathway for InsP 6 biosynthesis are needed in order to determine whether the non-lipid pathway is down regulated in the leaves of the HsPIPK1α plants. Analysis of the seedlings using inductively coupled plasma (ICP) indicated that the total Pi, calcium and magnesium were slightly lower in the HsPIPKIα transgenics compared to controls ( Table 2). The decrease in total calcium would be anticipated if the increased flux through the PI pathway resulted in a constitutive signal such that there is a net efflux of calcium from the cells [49,58].

Starch Metabolism Is Altered HsPIPKIα Plants
In all plants, the PIP 5-kinase specific activity was higher in the leaves of the older plants compared to the seedlings and InsP 3 was higher in mature leaves in the afternoon versus morning. For these reasons and because there appeared to be less effect on leaf morphology than root morphology, we focused our studies on leaf metabolism.
Previously, we showed that increasing the flux through the PI pathway in tobacco cells grown in suspension culture resulted in increased sucrose uptake, increased respiration and with time, increased starch granules [49,59]. To visualize starch in HsPIPKIα plants, leaves from 6 week-old plants were stained with iodine ( Figure 8A). Leaves from all the lines had less starch in the morning (morning is defined as plants harvested in the dark at 9 AM, 1 h before the lights came on) than afternoon (plants harvested at 5 PM, 1 h before the lights went off); however, leaves from HsPIPKIα plants showed significantly more starch than wild type. The increased starch was evident in leaves harvested both in the morning and afternoon. To quantify the differences, starch was analyzed from leaves of 3-week-old seedlings. In the HsPIPKIα leaves the starch was 5-fold higher in the morning samples and 1.5-fold higher in the afternoon samples compared to WT and GFP ( Figure 8B). Since excessive starch accumulation can result in changes in chloroplast morphology, chloroplast structure was compared using EM. The chloroplasts of the HsPIPKIα plants appeared normal although swollen because of the large starch granules (Supplementary Figure 1) and total chlorophyll was not different in any of the lines (Supplementary Figure 2).
To further investigate how starch metabolism is altered in HsPIPKIα plants at the molecular level we monitored transcripts levels of genes that are involved in starch synthesis [60,61]. The ADP-glucose pyrophosphorylase 3 (APL3), which converts G1P and ATP to ADP-glucose and starch synthase (SS), responsible for elongating starch polymers, were up-regulated in both lines of the HsPIPKIα plants compared to WT and GFP ( Figure 9A). The glucose-6-phosphate transporter (Glc6PT), which imports Glc6P from the cytosol into the chloroplast where it is converted to glucose-1-phosphate, was upregulated in the Hs9-7 line but was only marginally increased (1.6 fold) in the Hs5-8 line. We also monitored transcript levels of genes that are involved in starch degradation such as glucan water dikinase (SEX1), phosphoglucan phosphatase (SEX4), α-amylase (DBE) and maltose transporter (MEX).
They were all highly expressed in HsPIPKIα plants compared to WT and GFP plants ( Figure 9B). Although starch degradation genes were higher in the kinase plants, the relative loss of starch during the night was only 50%-60% in the HsPIPKIα plants whereas 85%-87% of the starch was lost in the WT and GFP plants. These data suggest that the HsPIPKIα plants were not mobilizing all the starch during the night to sustain cellular metabolism. Light/dark regulation of starch metabolism is complex.
While starch metabolism appears to be under circadian control [62], plants are able to respond to environmental cues and adjust starch metabolism to compensate for day length [63]. Our data suggest that expression of HsPIPKIα affected the light/dark sensing that regulates starch metabolism. If carbon export from the leaves to root was affected, the increase in starch might have been accompanied by an increase in sucrose. We did not detect significant differences in soluble sugars in the HsPIPKIα 5-8 (Hs5-8) seedlings although there was a slight increase in sucrose in leaves of 3 week-old HsPIPKIα 9-7 (Hs9-7) seedlings growing on agar supplemented with 1% sucrose (Supplementary Figure 3A). If sucrose was limiting in the roots, we reasoned that adding sucrose would increase root growth. When seedlings were grown on agar with increased (3%) sucrose, root growth increased slightly (Supplementary Figure 3B). Although root growth was less inhibited at 6% sucrose in the HsPIPKIα seedlings, sucrose alone did not restore normal root growth. Increased anthocyanin biosynthesis is an indication of stress and a change in carbon flux. When anthocyanin levels were compared, the HsPIPK1α plants had more anthocyanin whether they were harvested morning or afternoon ( Figure 10).  week-old seedlings were harvested and anthocyanin was extracted and quantified as described in the Experimental Section. The data are the means ± SD of 3 biological replicates.

Constitutively Increasing PtdInsP 2 Biosynthesis and InsP 3 in Leaves did not Affect Photosynthetic Electron Transport
In response to changes in the environment and demands on energy and reductive power in the chloroplasts, plants can switch between cyclic and linear electron flow pathways [64]. Cyclic electron flow around photosystem I (CEF) also increases upon drought stress [65,66] and during light activation after a prolonged dark period [67]. Over the course of these experiments, we noticed that in leaves of the HsPIPKIα plants, the Ins(1,4,5)P 3 was about 2-fold higher in the afternoon compared to morning. In addition, others had reported that InsP 3 was higher in light than dark grown plants [12]. Furthermore, previous reports indicated that expressing HsPIPKIα in tobacco cells increased activity of ATP-dependent pumps and affected K + channels [49,68]. We reasoned that increasing biosynthesis of PtdInsP 2 should increase the demand for ATP and could potentially lead to activation of CEF.
To investigate whether the increased PtdInsP 2 biosynthesis reflected differences in photosynthetic electron flow, we quantified proton and electron transfer rates in 6 week-old plants to look for increased CEF. Figure 11 shows the steady state transthylakoid proton flux (v H + ) as a function of linear electron flow (LEF) rates at multiple light intensities. The H + /e − for LEF is fixed, and any increase in the slope of the v H + /LEF relationship would indicate an increase in proton translocation independent of LEF due to the activation of CEF [69]. Figure 11 shows no statistically significant increase in this slope (ANCOVA p > 0.05, n = 3) for either of the HsPIPKIα plants, suggesting no activation of CEF due to our calculated increase in ATP demand. This absence of CEF is further evidenced by comparison of pmf (expressed as total amplitude of electrochromic shift (ECS) during a dark interval (ECS t )) to pmf attributable to LEF alone (pmf LEF ) [69,70]. An increase in CEF should cause an upwards shift in the relationship of these parameters; however, we see no significant differences in either of the HsPIPKIα plants when compared to GFP (ANCOVA p > 0.05, n = 3). Taken together, Figures 9, 10 and 11 clearly show that the constitutive increase in the PI pathway affected chloroplast carbon metabolism and transcripts involved in starch biosynthesis while having little impact on photosynthetic electron transport. The data in Figure 11 show that any increase in demand for ATP imposed by expressing HsPIPKIα was not met by increasing cyclic electron flux. There was no significant difference in the total ATP  Figure 4A) and analysis of the NADPH/NADP ratio indicated that it was slightly less in the HsPIPKIα plants. The NADPH/NADP ratio in seedlings was 2.5 ± 0.07, 1.9 ± 0.15, 0.9 ± 0.24 and 0.9 ± 0.24 for the WT, GFP and Hs5-8 and Hs9-7 seedlings, respectively. (The NADPH value for the WT seedling was 2.7 μmol/g FW and for NADP was 1.0 μmol/g FW). The NADPH/NADP ratios for leaves of whole plants were 1.2, 0.8, and 0.5 for WT, Hs5-8 and Hs9-7, respectively. Based on these observations, it is likely that in order to maintain homeostasis there were changes in metabolic pathways (e.g., an increase in the malate s huttle [64,71] or increased mitochondrial respiration as reported for tobacco cells [49]) that provided any additional ATP needed as a result of the expression of HsPIPKIα.

Physiological Characteristics of the HsPIPK1α Plants
Although one might reason that a constitutive InsP 3 -mediated signal would make the plants stress tolerant, it was also possible that stress-induced changes in basal metabolism would render the plants stress sensitive [52]. Specifically, the constitutive increase in InsP 3 -mediated signaling should deplete InsP 3 -sensitive calcium stores of the organelles and render the remaining calcium tightly bound. If this were true, then the total cellular calcium would be reduced and stress responses that require an increase in stored (organelle) calcium might be compromised. As shown in Table 2 above, there was a 10% decrease in overall calcium in the HsPIPKIα plants. We used several approaches to test for stress tolerance.
Perera et al. [15] showed that plants with constitutively low InsP 3 had increased tolerance to the withdrawal of water for up to 12 days. The authors concluded that the plants with constitutively low InsP 3 -mediated signaling had induced compensatory pathways that rendered the plants drought tolerant. To test the drought tolerance of the HsPIPKIα plants, we withheld water for 9 days. As shown in Figure 12A, the HsPIPKIα plants were more drought sensitive than WT plants. In addition, the HsPIPKIα plants had increased leaf water loss in a detached leaf assay ( Figure 12B). The phenotype is the opposite of the plants with low InsP 3 reported by Perera et al. [15]. The data could be interpreted as indicating that InsP 3 -mediated responses were not involved in stomatal closure. However, it is possible that a decrease in organelle calcium stores or extracellular calcium affected the ability of the guard cells to close. We did not measure extracellular calcium per se, but InsP 3 has been shown to increase in response to added extracellular calcium, and this response requires the presence of the chloroplast thylakoid calcium binding protein, CAS [72,73]. If the guard cells in the HsPIPKIα plants had depleted chloroplast calcium stores or extracellular calcium, in theory, the stomata should not have closed as rapidly and the plants should be more sensitive to water loss. It is also possible by increasing the flux through the PI pathway and increasing PtdInsP 2 but decreasing PtdIns4P, we affected membrane biogenesis and/or plasma membrane pumps and channels such that stomatal closure was impaired and the plants wilted faster [14,16,68]. More extensive studies of guard cell calcium stores and membrane trafficking are needed to determine the underlying mechanisms rendering the HsPIPKIα plants drought sensitive. It should be noted that this phenotype of the HsPIPKIα plants is in contrast to what was reported for the sac9 (PtdInsP 2 ptase) mutants which have increased PtdInsP 2 . The sac9 mutants were reported to have constitutively closed stomata [46]. These differences in the phenotype of the sac9 mutant and HsPIPKIα plants may reflect differences in the levels of PtdInsP 2 in the leaves of the sac9 and HsPIPKIα or other effects of the sac9 mutation that may have more direct effects on membrane biogenesis and cell wall deposition [74]. Several labs have reported changes in InsP 3 in response to abscisic acid (ABA); however, these genetic approaches to increase or decrease InsP 3 by altering the expression of phospholipase C or InsP ptases have had mixed results [28]. Ectopic expression of endogenous InsP Ptase1 to lower InsP 3 decreased ABA-induced stomatal closure, while lowering InsP 3 by expressing the human InsP 5-ptase increased the ABA-sensitive stomatal closure. The phenotypes of the loss of function mutants are more consistent. Mutations in InsP Ptase1 and 2 resulted in increased InsP 3 and increased sensitivity to ABA in seed germination assays [75], and plants with a mutation in InsP ptase12, an InsP ptase [76] have pollen that germinates precociously and are hypersensitive to ABA [77]. As predicted from these mutants, the HsPIPKIα seeds germinated quickly and were more sensitive to ABA than wild type in germination assays ( Figure 12C). The HsPIPKIα seedlings were also similar to the InsP ptase mutants in that they had an incomplete venation pattern in the cotyledons [78] (data not shown).
Regulation of cytosolic calcium is important for heat tolerance [79][80][81] and heat has been shown to increase PtdInsP kinase activity and PtdInsP 2 in tobacco cells [82]. We asked whether the HsPIPKIα, which have increased PtdInsP 2 would be heat tolerant. The HsPIPKIα seedlings were grown in the dark for 2.5 days, exposed to 48 °C for 30 min and then placed in the light for 24 h. As shown in Figure 13

C.
A.

B.
HsPIPKIα seedlings were more heat tolerant. Survival was quantified by measuring chlorophyll recovery after 24 h. Figure 13. HsPIPKIα seedlings are more tolerant of heat and light. (A) Seedlings were grown in the dark for 2.5 days, exposed to 48 °C for 30 min and placed in the light 24 h or kept in the light at room temperature (control); (B) Chlorophyll was extracted in acetone and absorbance was measured at 663 nm. The data are reported as % of the control chlorophyll recovered per g FW. The data are the means ± SD for 3 biological replicates consisting of 25 seedlings.

Very Few Differences in Transcript Levels Were Detected Using Microarray Analysis
In an attempt to gain some insight into what affects expressing HsPIPKIα had on plant gene expression, we did a microarray analysis of cDNA from three-week-old seedlings harvested just before the lights came on (morning). Table 3 reveals the results of the analysis of both HsPIPKIα 9-7 and 5-8 lines compared to the WT controls. Supplementary Figure 4B shows a heat map of changes detailed in Table 3. Some of the transcript changes may reflect systemic changes in vascular transport and cell wall structure associated with up-regulation of the PI pathway [51,74,[83][84][85]. Transcripts of PIPKs were first reported associated with vasculature [51,86] and these transcript changes may reflect tissue specific sensitivity to the expression of the HsPIPKIα or a reflection of effects on long distance signaling by InsP 3 -mediated events [87]. In addition, the transcript changes may reflect an up-regulation of pathogen responses or endocytic pathways associated with changes in phosphoinositides induced during symbiosis (e.g., PR1, Thioredoxin h8 [18,19,88,89]). We did not detect significant changes in the starch biosynthetic transcripts in the array. This may be due to differences in sensitivity of the standard microarrays compared to qPCR. Additional studies of tissue specific, targeted gene expression, cell wall structure and pathogen response are necessary to understand the impact of increased flux through the PI pathway induced in these studies. Table 3. Genes with greater than two-fold change in expression in the HsPIPKIα 9-7 and 5-8 lines compared to WT. Notations in parentheses indicate the specific line (9-7 or 5-8).  1.00 (9-7) 1.66 (5-8)

Generation and Selection of HsPIPKIα Transgenic Plants
The gene encoding the human PIPKIα (NM_003557) was cloned into pK7WGF2 (Functional Genomics Division, Department of Plant Systems Biology, Ghent University, Ghent, Belgium as previously described [49]. Recombinant plasmids were transformed into Agrobacterium tumefaciens EHA105 using the freeze-thaw method [90] and then transformed into Arabidopsis (Arabidopsis thaliana ecotype Columbia) by the floral dip method [91]. Four independent transformed lines were further selected. Stable expression of the transgene was monitored by RT-PCR and immunoblotting as described below.

Plant Growth Conditions
Wild-type (ecotype Columbia) and HsPIPKIα transgenic Arabidopsis thaliana plants were grown under short-day conditions (8 h of light/16 h of dark) at 21 °C with a light intensity of ~150 µmol· m -2 ·s -1 in the North Carolina State University Phytotron in a growth chamber. For all soil-grown experiments, a large batch of soil mix (Promix PGX; Hummert International, Earth City, MO, USA) was moistened well with water and the pots were filled with an equal amount of soil prior to planting the seeds. For experiments using seedlings, seeds were surface-sterilized by first incubating in 70% ethanol for 1 min, then incubating in a mixture of 30% (v/v) commercial bleach and 0.1% Triton X-100, with occasional agitation for 12 min and then washed with sterilized dH 2 O for 7 times and stratified for 48 h at 4 °C prior to plating on Murashige and Skoog medium (Caisson Labs, North Logan, UT, USA) containing 1% sucrose and 0.8% agar type M (Sigma-Aldrich, St Louis, MO, USA). Plates were incubated vertically in a growth chamber under short-day conditions as described above. For root and hypocotyl elongation measurements, 4 day after germination plates were covered and placed in the dark and growth was monitored every 24 h for a 3-to 4-day period.

Seed Germination Assays
Surface-sterilized, stratified seeds were plated on Murashige and Skoog plates containing different concentrations of ABA as indicated. Germination was counted as the emergence of green cotyledons at 3 days after plating.

RNA Extraction, RT-PCR, and qRT-PCR Analysis
RNA was isolated from harvested leaves using the plant RNeasy Mini kit (Qiagen Sciences Inc., Frederick, MD, USA) with the on-column RNase-free DNase I treatment. RT was carried out to generate cDNA using Omniscript reverse transcriptase enzyme (Qiagen Sciences Inc.) and random primers according to the manufacturer's instructions (Qiagen Sciences Inc.). For RT-PCR, cDNAs were amplified using HotStar Taq DNA Polymerase (Qiagen Sciences Inc.) and gene-specific primers. q RT-PCR was carried out using Full Velocity SYBR-Green QPCR Master Mix (Stratagene, La Jolla, CA, USA) on an MX3000P thermocycler (Stratagene). Gene-specific primers for select genes were designed with the help of AtRTPrimer, a database for generating specific RT-PCR primer pairs [92], and are shown in Supplementary Table 1. PCR was optimized, and reactions were performed in duplicate. Transcript levels were standardized based on cDNA amplification of the reference gene ACTIN2/8 and/or PP2A. Relative gene expression data were generated using the 2 - Ct method [93] using the wild-type as the reference.

Protein Isolation and Immunoblotting
Total protein extract was obtained from plants frozen in liquid N 2 or seedlings grown as described by Weigel and Glazebrook [94]. Microsomal fraction proteins were obtained by two-phase partitioning as described previously [49]. Protein concentrations were quantified as described by Bradford [95]. Protein was separated by 10% (w/v) SDS-PAGE and transferred to PVDF membrane by electroblotting and membranes were incubated with antibodies (anti-mouse GFP [Clonetech Lab, Mountain View, CA, USA]), and incubated with horseradish peroxidase-conjugated anti-mouse or anti-rabbit. Immunoreactivity was visualized by incubating the blot in SuperSignal West Pico Chemiluminescent substrate (Pierce Protein Products, Thermo Fisher Scientific, Rockford, IL, USA) and exposure to X-ray film. After chemiluminescence detection, total protein was visualized by staining the blots with Amido black (Sigma-Aldrich, St Louis, MO, USA). Following chemiluminescence detection, total protein was visualized by staining the blots with Amido black (Sigma-Aldrich, St Louis, MO, USA).

PtdInsP 5-Kinase Assays
In vitro lipid kinase assays were performed using plasma membrane proteins (2 µg) and endomembrane fraction protein (30 µg). The standard assay was as previously described [49] with the following modifications. Reactions were performed either in the absence or presence of substrate 125 µM PtdIns(4)P from porcine brain (Avanti Polar Lipids) at room temperature for 10 min in a total volume of 50 µL. After incubation, phospholipids were extracted and separated by TLC as described [96].

Lipid Profiling
To determine the effects of HsPIPKIα expression on total glycerol lipid profile, we extracted lipids from leaves from 3 week-old seedlings in the protocol as described by the Kansas Lipidomics Facility [97] and lipid analysis and quantification were performed as described [49] at the Kansas Lipidomics Facility.

In Vivo Labeling Studies
For short-term labeling studies with 32 Pi, 13 or 17-day-old seedlings (~10 seedlings per well) were transferred to a multi-well plate containing 800 µL of 0.5× Murashige and Skoog medium. The seedlings were incubated overnight with gentle rotation. In the morning, 50 µCi of carrier-free [ 32 P] Pi (~62 µCi mL -1 ) was added to each well and seedlings were harvested at the indicated time points by immediate transfer to 500 µL of cold 20% (v/v) PCA and incubated on ice for ~20 min. The PCA treated seedlings were then washed with cold water twice, and lipids were extracted, separated by TLC, and 32 P-labeled lipids were quantified with a Bioscan Imaging Scanner.

Labeling Studies with [ 3 H]myo-Inositol
One-week-old seedlings (~10 seedlings per well) were transferred to a multiwell plate containing 800 µL of 0.5× Murashige and Skoog medium containing 45 µCi of [ 3 H]myo-inositol. Plates were incubated in a growth chamber under long-day conditions with gentle rotation to ensure aeration for 4 days. After incubation, the seedlings were quickly blotted on tissue and ground in liquid N 2 . The frozen ground powder was incubated in 0.75 N HCl containing 0.2% phytate (as carrier) on ice for 20 min. The pellet and supernatant were separated by centrifugation, the pellet was washed with cold water twice, and the [ 3 H] myo-inositol labeled lipids were extracted from the pellet. The lipids were separated by TLC and quantified with a Bioscan Imaging Scanner. [ 3 H] inositol hexaphosphate was also identified from the supernatant based on the coelution of standard Ins(1,2,3,4,5,6)P 6 using ion chromatography. For these analyses, fifty microliters of the HCL extract were diluted to 1 mL with 0.375 N HCl and filtered through a 0.45 μm nylon filter. Fifty and one hundred microliter aliquots were analyzed by isocratic ion chromatography using 0.25 N HNO 3 eluant and Dionex AG7/AS7 columns as previously described [98,99]. Twelve 1 mL fractions were collected at 1 min intervals and counted with 5 mL EcoLume in plastic scintillation vials. InsP 6 was calculated as the cpm in fraction 8 divided by the total cpm of the 12 fractions times 100%. Two biological replicates were analyzed to give a total of two wild-type and two HsPIPKIα line extracts. Two analyses (50 μL and 100 μL) were performed on each of the four diluted extracts.

Determination of Total InsP 6 in Seeds
InsP 6 analysis of seeds was a modification of a previous method described by Bentsink et al. [100]. Specifically, dry seeds (4-5 mg) were rehydrated in 500 μL 0.5 N HCl for 60 min at 55 °C. The mixture was ground with a plastic pestle and centrifuged 5 min at 15,000 ×g. The supernatant solutions were filtered through a 0.45 μm pore size 17 mm nylon filter, diluted with an equal volume of water, and InsP 6 was determined by isocratic ion chromatography using 0.25 N HNO 3 eluant and Dionex AG7/AS7 columns as previously described [98,99]. Triplicate biological samples were analyzed.

Quantification of Soluble Pi
Leaves of 3-week-old seedlings were harvested, immediately frozen in liquid N 2 , and ground to a fine powder. Soluble Pi was extracted by adding 10 times 1% [v/v] HOAC of sample weight. The extracted sample was analyzed for Pi as described by Bartlett [101] measuring A660.

Determination of Anthocyanin and Chlorophyll A
Anthocyanin content was determined as describe in Teng et al. [102]. Frozen samples from 3 week-old seedlings were homogenized in 1% [v/v] HCl in MeOH at 4 °C and incubated overnight. After centrifugation at 15,000 ×g for 15 min, the absorbance of supernatants was measured at 530 and 657 nm and anthocyanin was calculated using the formula A 530 − 0.25 × A 657 and corrected for the volume and sample weight.
For chlorophyll a measurements, the samples (25 seedlings/treatment) were extracted in ethanol (100% v/v). Chlorophyll was quantified by measuring the absorbance at 665 nm (Eb665) and 750 nm (Eb750). After the reading, the samples were acidified by adding 10 μL of 2N HCl directly to the cuvette, mixed well, incubated for 5 min, read at 665 nm (Ea665) and 750 nm (Ea750). Chlorophyll was calculated using the formula 29.6 × [(Eb 665 − Eb 750 ) − (Ea 665 −E a 750 )] and reported as % of the control (non-heat treated samples) for each line or per g FW as indicated.

Staining and Quantification of Starch
For starch staining, leaves were harvested from 6 week-old plants at the end of the day and at the end of the night. Chlorophyll was removed with 80% EtOH and stained with IKI solution (1% [w/v] iodine, 2% [w/v] potassium iodine) for 1 min and rinsed with dH 2 O and imaged by scanner. For starch quantification, frozen samples from 3 week-old seedlings were homogenized in 80% (v/v) EtOH and boiled for 3 min and centrifuged at 3,000 ×g for 10 min. Insoluble fraction was determined by measuring the amount of glucose released by treatment with α-amylase and amyloglucosidase, as described by Smith and Zeeman [103].

Analysis of ATP and NADP(H) and NAD(H)
ATP was assayed using a bioluminescence assay kit (Sigma-Aldrich) according to the manufacturer's directions. NADP(H) and NAD(H) were extracted and assayed as described by Matsumura and Miyachi using an enzyme cycling assay measuring the absorbance at 570 nm [104].

Sugar Analysis
Soluble sugars and inositol were analyzed by gas chromatography-mass spectrometry. Leaf tissue of 3 week-old seedlings was ground in a cold 60:40 (v/v) methanol:H 2 O solution, mixed with acetonitrile, and dried under vacuum. Samples were analyzed at the Metabolomics and Proteomics Laboratory at North Carolina State University. The sugars were converted to trimethylsilyl derivatives, and gas chromatography-mass spectrometry was performed using a ThermoTrace GC Ultra gas chromatograph coupled to a Thermo DSQ II mass spectrometer. The mass spectrometer was operated with an electron-impact source in positive mode monitoring m/z 191, 204, 217, 361, and 437. Quantitation was conducted by comparing peak areas obtained for trimethylsilyl derivatives of fructose, glucose, and sucrose in the samples with a series of reference standards analyzed concurrently, and data were processed using Thermo's Xcalibur software. Data presented are averages from three independent biological replicates.

ICP Analysis
All elements except Cu were analyzed on a Perkin Elmer inductively coupled plasma-optical emission spectrometer (ICP-OES). 50 mg of pre-weighted dried leaves of 3 week-old seedlings were digested with 4 mL of conc. HNO 3 (Trace Metal Grade) and 2 mL of 30% H 2 O 2 (ACS reagent grade). Due to the low concentration of Cu in sample digestates (ppb), Cu was determined by inductively coupled plasma mass spectrometry using a Varian-820 Quadrupole ICP-MS.

In Vivo Spectroscopic Analysis
Photosynthetic parameters were measured using an in-house constructed spectrophotometer/fluorimeter modified from [105] with humidified air supplied to the underside side of the leaf, as described in [70,[106][107][108]. LEF rates were calculated as: Where ϕ II is the yield of photosystem II calculated using chlorophyll a fluorescence changes from steady state to saturating light [109,110], and i is the actinic light intensity. The thylakoid proton circuit was monitored using dark interval relaxation kinetics of the electrochromic shift (ECS) of absorption at 520 nm of the carotenoids in response to the transthylakoid electric field [69]. Total light induced pmf was estimated as the total ECS from light to dark (ECS t ). Light induced transthylakoid proton flux (v H + ) was estimated from the initial slope of the ECS from light to dark. The pmf attributable to LEF (pmf LEF ) was calculated as: Where is the lifetime of the ECS decay [69,70]. Data analysis was performed in, and descriptive statistics and figures were generated with Origin 9.0 (Microcal Software). Statistical analysis was performed using MATLAB R2012a (The Mathworks). Statistical significance was set at p < 0.05.

RNA Isolation for Microarray Analysis
For the microarray analysis, leaf samples were collected from 3 week-old seedlings harvested in the dark just before the lights came on and immediately ground in liquid N 2 . Three biological replicates were performed for the wild type, GFP, and two independent transgenic lines (Hs5-8 and Hs9-7). RNA was isolated using the Plant RNeasy kit (Qiagen Sciences Inc., Frederick, MD, USA), and biotinylated target cRNA was synthesized using the 3' IVT Express kit (Affymetrix, Santa Clara, CA, USA). RNA quality was monitored on an Agilent 2100 bioanalyzer. Arabidopsis arrays (ATH1 from Affymetrix) were hybridized, and the data acquisition and analysis were performed by Expression Analysis using the Affymetrix fluidics station and GCOS software.

Conclusions
Leaves of plants expressing HsPIPKIα had increased PtdInsP 2 biosynthesis and increased total InsP 3 . Our focus was to characterize the effects of increasing the flux through the PI pathway in leaves. Compared to WT and GFP-expressing plants, the leaves of the HsPIPKIα plants had increased starch and anthocyanin both at the end of day and end of night. InsP 3 levels were highest in the afternoon in the HsPIPKIα plants, correlating positively with photosynthesis. Although chloroplast carbon metabolism was affected, photosynthetic electron transport was not different in the HsPIPKIα plants compared to the WT or GFP controls. There are many reports indicating a role for calcium in the chloroplast and specifically for changes in stromal calcium during the light/dark transition [31,[34][35][36][37]. Johnson et al. [36] suggested that cytosolic calcium might be the source of calcium for the stromal increase during the light/dark transition and showed that photosynthetic electron transport was not required for the dark-induced stromal calcium changes; however, the role of cytosolic calcium in regulating chloroplast and organelle metabolism remains a conundrum [40,41]. Based on previous work and the data presented in this paper, we hypothesize that InsP 3 is one of the components of cytosolic signaling which affects chloroplast calcium homeostasis and that InsP 3 likely contributes to coordinating organelle calcium signaling during basal metabolism, as well as light/dark transitions and stress-induced responses. While more extensive studies with tissue and organelle-specific calcium probes [111][112][113] are needed to determine whether a constitutive InsP3 signal can affect chloroplast calcium and or light/dark calcium fluctuations, the HsPIPKIα plants, which have increased flux through the PI pathway, provide a platform for these studies.
for the ICP analysis. The lipid analyses described in this work were performed at the Kansas Lipidomics Research Center Analytical Laboratory. Kansas Lipidomics Research Center was supported by National Science Foundation (EPS 0236913, MCB 0455318, DBI 0521587), Kansas Technology Enterprise Corporation, K-IDeA Networks of Biomedical Research Excellence (INBRE) of National Institute of Health (P20RR16475), and Kansas State University.