Rapid phosphatidic acid accumulation in response to low temperature stress in Arabidopsis is generated through diacylglycerol kinase

Abstract

Phosphatidic acid (PtdOH) is emerging as an important signaling lipid in abiotic stress responses in plants. The effect of cold stress was monitored using (32)P-labeled seedlings and leaf discs of Arabidopsis thaliana. Low, non-freezing temperatures were found to trigger a very rapid (32)P-PtdOH increase, peaking within 2 and 5 min, respectively. In principle, PtdOH can be generated through three different pathways, i.e., (1) via de novo phospholipid biosynthesis (through acylation of lyso-PtdOH), (2) via phospholipase D hydrolysis of structural phospholipids, or (3) via phosphorylation of diacylglycerol (DAG) by DAG kinase (DGK). Using a differential (32)P-labeling protocol and a PLD-transphosphatidylation assay, evidence is provided that the rapid (32)P-PtdOH response was primarily generated through DGK. A simultaneous decrease in the levels of (32)P-PtdInsP, correlating in time, temperature dependency, and magnitude with the increase in (32)P-PtdOH, suggested that a PtdInsP-hydrolyzing PLC generated the DAG in this reaction. Testing T-DNA insertion lines available for the seven DGK genes, revealed no clear changes in (32)P-PtdOH responses, suggesting functional redundancy. Similarly, known cold-stress mutants were analyzed to investigate whether the PtdOH response acted downstream of the respective gene products. The hos1, los1, and fry1 mutants were found to exhibit normal PtdOH responses. Slight changes were found for ice1, snow1, and the overexpression line Super-ICE1, however, this was not cold-specific and likely due to pleiotropic effects. A tentative model illustrating direct cold effects on phospholipid metabolism is presented.

Keywords: abiotic stress; cold stress; diacylglycerol kinase; lipid signaling; phosphatidic acid; phosphoinositide; phospholipase; plant signaling.

Figure 1

Cold stress triggers the formation of ³²P-PtdOH in Arabidopsis seedlings. Five-days-old seedlings were metabolically radiolabeled O/N with ³²P_i and then incubated for 5 min at 0°C or maintained at 20°C. Lipids were extracted, separated by TLC, and visualized by phosphoimaging. Each lane represents an extract of two seedlings. Abbreviation: SPL, structural phospholipids.

Figure 2

Temperature-dependent accumulation of ³²P-PtdOH in Arabidopsis leaves and seedlings. (A) O/N ³²P-prelabeled seedlings were incubated for 5 min at the indicated temperatures. Lipids were then extracted, separated by TLC, and visualized by autoradiography. (B) Quantitation by phosphoimaging of ³²P-PtdOH formed at different temperatures in seedlings. (C) Formation of ³²P-PtdOH in leaf disks at different temperatures. Values are means of triplicates ±SD. Asterisks indicate highest temperatures giving rise to a significant (p < 0.05) increase in ³²P-PtdOH.

Figure 3

Kinetics of cold-induced ³²P-PtdOH accumulation in Arabidopsis seedlings and leaves. ³²P-prelabeled seedlings (A) or leaf disks (B) were incubated at 0°C (filled circles) or 20°C (control, open circles) for different periods of time. Lipids were then extracted, separated by TLC, and quantified by phosphoimaging. Data points (±SD) are from triplicate incubations.

Figure 4

Metabolic origin of the chilling-induced ³²P-PtdOH response in Arabidopsis seedlings. (A) In the presence of 0.5% n-butanol, accumulation of the transphosphatidylation product ³²P-PtdBut is used as measure of PLD activity. White bars, ³²P-PtdOH, gray bars, ³²P-PtdBut. (B) Seedlings were prelabeled with ³²P_i for 20, 60, or 180 min, to preferentially label the monoester-phosphates of lipids with high turnover rates. Subsequently, seedlings were transferred to cold (0°C) or kept at 20°C for an additional 15 min. Lipids were separated on TLC and visualized by phosphoimaging. (C) Dependence of ³²P-PtdOH levels in control (white bars) and cold conditions (gray bars) on the ³²P-prelabeling time. (D) Five-days old Col-0, plda1, pldd, and plda1/d knock-out seedlings were radiolabeled O/N with ³²P_i and then incubated for 5 min at 0°C or maintained at 20°C. ³²P-PtdOH increases are expressed as percentage of total ³²P-labeled lipids.

Figure 5

Cold stress-induced changes in ³²P-PtdOH vs. ³²P-PtdInsP. (A) TLC analysis of ³²P-phospholipids extracted from seedlings after 5 min exposure to the temperatures indicated. (B) Similar experiment as (A) Quantitation of radioactivity in the lipids was by phosphoimaging. Filled circles, ³²P-PtdOH; open circles, ³²P-PtdInsP. (C) A time course experiment at 0°C shows contrary changes in ³²P-PtdOH and ³²P-PtdInsP. All values are means of at least three samples containing two seedlings each from a representative experiment (error bars indicate SD).

Figure 6

Cold-induced ³²P-PtdOH induction in known Arabidopsis cold response mutants. Five-days-old seedlings were prelabeled O/N with ³²P_i and subsequently incubated at 0°C or kept at room temperature for 15 min. Lipids were then extracted, separated by TLC, and quantified by phosphoimaging. ³²P-PtdOH levels are expressed as percentage of the total ³²P-lipid. Values are means from triplicate incubations from a typical experiment; error bars indicate SD. White bars, control; gray bars, 0°C. (A) The mutants fry1, hos1, los1, and their wt background, C24RD29A-LUC. (B) The snow1 mutant and the wt control, Col-0.

Figure 7

³²P-PtdOH responses in seedlings of the ice1 mutant and ICE1 overexpression transgenic line (Super-ICE1). Five-days-old seedlings were prelabeled O/N with ³²P_i and incubated at 0°C or with 300 mM NaCl for 15 min. ³²P-PtdOH levels are expressed as percentage of the total ³²P-lipid (average ±SD). ³²P-PtdOH was enhanced due to cold and salt in all genotypes (p < 0.025), but salt-induced ³²P-PtdOH was decreased in ice1 compared to wildtype (*p = 0.008).

Figure 8

Model illustrating potential early effects of cold stress on phospholipid metabolism and de novo synthesis in Arabidopsis. The main route to rapid cold-induced PtdOH formation is suggested to be based on the phosphorylation of PLC-generated DAG from PtdInsP (reactions 1/3). The activity of PECT, which produces the precursor of the polar head of PtdEtn, CDP-Etn, is proposed to be down regulated by low ambient temperature (2). This would lead to reduced PtdEtn formation, and potentially, to DAG accumulation, which might cause PtdOH to accumulate as a result from phosphorylation of DAG by a DGK (3), or due to product inhibition of PAP by DAG (4). The major pathway of PtdCho synthesis depends on methylation of EtnP to ChoP by PEAMT, which could be inhibited by PtdOH (5). Note that the model only highlights immediate effects of cold temperature; longer exposure to cold induces a myriad of metabolic changes which impact lipid biosynthesis in different ways. Abbreviations: Acyl-CoA, acyl-coenzyme A; CDS, CDP-DAG synthase; EK, ethanolamine kinase; EPT, CDP-ethanolamine phosphotransferase; GPAT, glycerol 3-phosphate acyltransferase; LPAAT, lysophosphatidic acid acyltransferase; lyso-PtdOH, lysophosphatidic acid.

Figure A1

Two pathways with the potential to generate DAG and PtdOH at the expense of PtdIns4P. PtdIns4P is suggested to be the substrate of cold-induced PLC activity which not only generates DAG, but at the same time releases InsP₂ that can be converted to InsP₆ and/or Ins. The latter products may have a functional relevance in the stress response because InsP₆ is a signaling compound in plants, and Ins is a precursor to compatible solutes (Vermeer and Munnik, 2010). Alternatively, inositolphosphorylceramide synthase (IPCS) generates DAG while transferring the InsP headgroup from PtdIns to ceramide (Cer), generating inositolphosphorylceramide (IPC). The PtdIns substrate in this conversion can be derived from PtdIns4P dephosphorylation, as, in yeast, through Sac1 activity. DAG generated via either of these pathways may subsequently be phosphorylated by DGK to generate PtdOH.

Figure A2

Cold temperatures suppress ³²P-PtdEtn accumulation in Arabidopsis leaf disks. After 30 min ³²P-labeling and 5 min incubation at the indicated temperatures, phospholipids were quantitatively analyzed by phosphoimaging. Values are in arbitrary units (AU) representing means (±SD) of the radioactivity levels.

Figure A3

Genotyping of the Arabidopsis dgk T-DNA insertion lines. DNA was isolated from leaf discs from 3 to 4 independent plants and zygosity determined by PCR using specific wt- and mutant-primer sets as listed in Table A1. Gel sample order (from left to right): Size marker; Col-0 (wt), with first primer set, LP/RP (= wt band) and second primer set, RP/LB (= insertion band); Three (or four) independent dgk plants, with first primer set, LP/RP (= wt band) and second primer set RP/LB (= insertion band); Last lane, primer control (−). Some gels end with the size marker. Results clearly show that all lines used are homozygous T-DNA insertion lines since Col-0 was the only line giving wt bands and all insertion lines gave only bands with the second primer set. For the SALK lines LBb1.3 was used as LB primer, for the SAIL lines LB3 was used as LB primer (see primer list, Table A1).

Figure A4

Determination of DGK KO/KD-expression in Arabidopsis T-DNA insertion mutants by RT-PCR. RNA was isolated from 9-days old roots (DGK1,−3,−5,−7) or flowers (DGK2,−4,−6) since the latter genes did not reveal expression in the root (not shown). Predicted band Sizes: DGK1, 702 bp; DGK3, 592 bp; DGK5, 534 bp; DGK6, 434 bp; DGK7, 791 bp; SAND (reference gene), 244 bp. Abbreviations: ND, not detectable; NAC, non-amplification control (test for genomic DNA contamination), i.e., RT reaction without RT enzyme on Col-0 RNA; NTC, no target control (test for contamination + primer dimers), i.e., RT reaction with water sample.