Different PA metabolizing pathways have been shown to contribute to PA production in response to abiotic stress (Munnik et al., 2000; Ruelland et al., 2002; Arisz et al., 2009; Bargmann et al., 2009; Li et al., 2009; Hong et al., 2010). The phospholipase D (PLD) enzyme hydrolyses primarily structural lipids such as phosphatidylcholine (PC), and phosphatidylethanolamine (PE), resulting in formation of PA and the remaining headgroup (Pappan et al., 1998). Phospholipase C (PLC) hydrolyses phosphatidylinositol lipids (PPIs) into water soluble inositol-bis or trisphosphate (IP₂, IP₃) and diacylglycerol (DAG), which remains in the membrane (Munnik and Vermeer, 2010). DAG can be subsequently phosphorylated to PA by DAG kinase (DGK; Arisz et al., 2009).

Twelve PLDs have been identified in the model plant species Arabidopsis, which were initially classified in two groups based on their N-terminal lipid-binding domain. These domains consist either of a pleckstrin homology (PH) and PHOX (PX) or a calcium dependent-lipid binding (C2) domain (Elias et al., 2002). Later, these classes were further subdivided into six classes based on sequence homology and in vitro enzymatic activity: three α-, two β-, three γ-, one δ-, and one ε- PLD with a C2 domain and two ζ-class PLDs that contain PH and PX domains (Qin and Wang, 2002; Bargmann and Munnik, 2006; Li et al., 2009).

In the Arabidopsis genome, nine PLCs and seven DGK genes have been identified. Initially, PLC/DGK derived PA was primarily implicated in responses to biotic stress (van der Luit et al., 2000; de Jong et al., 2004). However, abiotic stress, in particular cold stress, also induced an accumulation of PLC/DGK-mediated PA formation (Ruelland et al., 2002; Gomez-Merino et al., 2004). The DGKs involved in PA formation during cold stress have not been identified yet. T-DNA insertion lines of the seven DGKs did not alter PA formation in response to cold, which is likely due to redundancy (Arisz et al., 2013) and therefore the role of DGK-derived PA in response to abiotic stress remains largely unknown.

Roots are the primary site of perception of salt stress, drought and low nutrient availability. To cope with these conditions, plants adapt the growth and morphology of their roots. Several PLD isoforms were found to be involved in adjusting root system architecture during abiotic stress (Galvan-Ampudia and Testerink, 2011; Figure 1A).

PLDζ2 is involved in directional root growth in saline conditions. Exposing one side of the root to salt increased pin-formed (PIN) 2 internalization, effectively redistributing auxin in the root tip. This redistribution resulted in bending away from saline conditions, named halotropism. A pldζ2-KO (knock-out) mutant showed reduced clathrin-dependent PIN2 internalization and reduced primary root bending (Galvan-Ampudia et al., 2013). Expression of PLDζ2 increased under low phosphate availability (Oropeza-Aburto et al., 2011) and the pldζ2-KO showed increased root hair growth when deprived of phosphate (Cruz-Ramirez et al., 2006). In accordance, less PA was formed in low phosphate conditions in the pldζ1/2 double mutant (Li et al., 2006a) and this mutant showed reduced lateral root and increased primary root growth in low phosphate conditions (Li et al., 2006b). In addition, the pldζ2-KO mutant also exhibited decreased sensitivity to auxin and a reduced root gravitropic response (Li and Xue, 2007).

Nitrogen is another important nutrient for plants and Arabidopsis PLDε-OE lines displayed an increase in lateral root and root hair elongation and primary root growth in low nitrogen conditions. This effectively increased the dry weight of the plant under these conditions and indicated an important role for PLDε in growth and nitrogen signaling (Hong et al., 2009).

PLDα1 and PLDδ are involved in different responses to abiotic stress including reactive oxygen species (ROS) signaling in response to ABA (Sang et al., 2001; Zhang et al., 2003; Zhang et al., 2009; Uraji et al., 2012) studied in stomata and leaves. The same phospholipases were also shown to play distinct roles in freezing tolerance (Welti et al., 2002; Li et al., 2004). Expression of PLDδ was elevated in response to dehydration and high salt stress (Katagiri et al., 2001). Salt stress induced formation of PA through PLDα1 and PLDδ, where both single mutants showed a reduction in primary root growth in saline conditions and during dehydration. This was even clearer in the pldα1/δ double mutant (Bargmann et al., 2009). A similar reduction in root growth was observed in pldα3, which was more susceptible to salinity and water deficiency. In hyperosmotic-stress conditions, the pldα3 mutant displays a reduction in primary root growth and a reduction in lateral roots (Hong et al., 2008).

Together, these studies show that PLDs are important for maintaining root growth in saline and hyperosmotic stress conditions amongst other functions. PA is not limited to its function as a signaling lipid; it is also an important intermediate during lipid-turnover. Therefore it is hard to separate the role of PA in lipid-turnover from its role in signaling and protein recruitment (Testerink and Munnik, 2011). To discern between the different roles of PA it is important to identify which proteins interact with PA and how these mediate the response that eventually leads to the acclimation to different stresses (Testerink and Munnik, 2005). In contrast to other signaling lipids such as phosphoinositides, no consensus PA-binding domain has been identified, which hampers the identification of new PA targets. A number of PA binding proteins have been identified involved in diverse cellular processes (Table 1).

A central mediator in metabolism, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), is targeted to PA in response to salt in Arabidopsis roots (McLoughlin et al., 2013). The best-described role of GAPDH is the conversion of glyceraldehyde-3-phosphate to D-glycerate 1,3-bisphosphate in the glycolytic breakdown of glucose. PA-binding does not alter the activity of GAPDH dramatically in vitro but adding PA to seedlings did induce proteolytic cleavage of glyceraldehyde-3-phosphate dehydrogenase C2 (GAPC2; Kim et al., 2013). Adding exogenous PA also reduced primary root growth, which was more severe when GAPDH was over-expressed whilst knock-out mutants showed less reduction in growth (Kim et al., 2013) indicating that the effect of PA on root growth was partially mediated by proteolytic degradation of GAPDH. Although GAPDH has been described to be involved in different non-metabolic processes, promoting its degradation might mediate energy conservation and arrest of root growth, which are immediate and relevant responses to any osmotic stress including saline conditions (Munns and Tester, 2008).

Phosphatidic acid has also emerged as an important regulator of microtubules and actin organization and re-organization. Microtubule reorganization is crucial for plants to adapt to saline conditions (Wang et al., 2007). In the pldα1 mutant background, the microtubule disorganization was more severe in response to salt and could not be recovered after the removal of salt. This effectively resulted in the plant being more salt sensitive in a microtubule-associated proteins 65-1 (MAP65-1) dependent manner. PA directly interacted with MAP65-1 and promoted the interaction of MAP65-1 with microtubules, effectively promoting the polymerization of cortical microtubules (Zhang et al., 2012).

In addition to microtubules, PA levels were also shown to be important for the behavior of actin filaments through the regulation of actin capping proteins (Li et al., 2012). PA specifically interacts with an actin capping protein (CP; Huang et al., 2006), which prevents actin filament from annealing and elongating. PA-binding reduced the activity of CP, effectively promoting actin reorganization and promoting cytoskeleton dynamics which are required for adaptation to adverse conditions.

In mammalian cells, PA and PA-generating enzymes such as PLD and DGK have been implicated in various aspects of vesicle transport (Manifava et al., 2001; Corda et al., 2002; Jang et al., 2012), but so far, little evidence is present that suggest a similar role in plants. Recently, clathrin heavy chain and clathrin assembly units were shown to recruit to the membrane in Arabidopsis roots in response to salt and to bind to PA-beads (McLoughlin et al., 2013). This likely represents an important aspect of the molecular mechanism of salt-induced PIN2 internalization which controls directional root bending in saline conditions (Galvan-Ampudia et al., 2013).

The Arabidopsis phosphoinositide-dependent kinase 1 (PDK1) binds several phosphoinositides and PA through its PH domain (Deak et al., 1999). PA activates PDK1 and indirectly, its downstream target AGC2-1 (OXI1) (Anthony et al., 2004, 2006). This signaling cascade induces a respiratory burst required for the full activation of mitogen activated protein kinase 6 (MPK6; Rentel et al., 2004), which was also shown to bind PA (Yu et al., 2010). MPK6 knock-out plants showed reduced growth in the primary root in saline conditions and the activation of MPK6 in response to salt is abolished in the pldα1 mutant background. In addition, MPK6 can phosphorylate the Na⁺/H⁺ antiporter salt overly sensitive (SOS1) in vitro and might therefore play a direct role in sodium homeostasis in roots (Yu et al., 2010). Another phosphorylation target of PDK1, PINOID (PID), was shown to bind phosphoinositides and PA. PID is involved in the asymmetric distribution of PIN auxin transporters, which are key regulators of root development (Zegzouti et al., 2006). Additionally, PA was also shown to bind the subunit of protein phosphatase 2A (PP2A); roots curl in NPA 1 (RCN1; Testerink et al., 2004). Recruitment to the membrane increased the activity of PP2A (Gao et al., 2013). PP2A in turn dephosphorylates the auxin transporter PIN1, changing its polarity and effectively altering the auxin distribution in the root, which is necessary for normal root development (Gao et al., 2013).

In a proteomics screen to identify PA targets during salt stress in Arabidopsis roots, the potassium channel β subunit KAB1 was shown to recruit to the membrane in response to salt and bind to PA (McLoughlin et al., 2013). The Arabidopsis genome encodes a single potassium channel β subunit 1 (Tang et al., 1995), which, as a tetramer, associates with the transmembrane α subunits of KAT1 channels (Tang et al., 1996). KAT1 is inactivated rapidly and internalized in response to ABA in guard cells. The inactivation occurs more rapidly than internalization, which suggests an additional mechanism. PA-recruitment could play a role in the inactivation of KAT1 by competing for KAB1, but this remains to be investigated. Since AtKAT1 expression is prevalent in leaves rather than roots, KAB1 is speculated to bind not exclusively to KAT1 but also to a different α subunit, viz. AKT1, which is selectively expressed in Arabidopsis roots (Tang et al., 1996). Recruitment of KAB1 by PA opens the possibility that PA might function as a regulator of potassium homeostasis in different parts of the plant.

The sucrose non-fermenting 1-related protein kinase 2 proteins (SnRK2s), is a family of osmotic stress-activated protein kinases (Kulik et al., 2011). SnRK2.10, a member of the class 1 subfamily, which is activated by salt and water stress but not by ABA, was identified in a PA affinity screen (Testerink et al., 2004). More recently, both SnRK2.10 and its close homolog SnRK2.4 were shown to bind PA directly. Arabidopsis snrk2.4 and 2.10-KO mutants exhibited reduced primary root length and lateral root density in saline conditions, respectively (McLoughlin et al., 2012). SnRK2.4 was shown be involved in regulating the ROS levels in roots in response to cadmium, but remarkably snrk2.4-KO mutants displayed an increase in primary root length when exposed to cadmium (Kulik et al., 2012). Overexpression of the SnRK2.4 wheat ortholog (TaSnRK2.4) in Arabidopsis caused an increase in primary root growth and resulted in more drought tolerant plants. This was explained by stronger water retention ability in these plants compared to wild-type (Mao et al., 2010). Overexpression of SAPK4 (a rice class 1 SnRK2 ortholog) increased tolerance to salinity and oxidative stress (Diedhiou et al., 2008).