Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2020 Feb 12;182(4):2111–2125. doi: 10.1104/pp.19.01332

The Lipid Flippases ALA4 and ALA5 Play Critical Roles in Cell Expansion and Plant Growth1,[OPEN]

Jeffrey A Davis a, Randall B Pares a, Tilde Bernstein c, Stephen C McDowell a, Elizabeth Brown a, Jason Stubrich a, Alexa Rosenberg a, Edgar B Cahoon b, Rebecca E Cahoon b, Lisbeth R Poulsen c, Michael Palmgren c, Rosa L López-Marqués c, Jeffrey F Harper a,2,3
PMCID: PMC7140935  PMID: 32051180

A subclass of lipid flippases contribute to glycerolipid and sphingolipid homeostasis in Arabidopsis leaves, and play critical roles in cell expansion and vegetative growth.

Abstract

Aminophospholipid ATPases (ALAs) are lipid flippases involved in transporting specific lipids across membrane bilayers. Arabidopsis (Arabidopsis thaliana) contains 12 ALAs in five phylogenetic clusters, including four in cluster 3 (ALA4–ALA7). ALA4/5 and ALA6/7, are expressed primarily in vegetative tissues and pollen, respectively. Previously, a double knockout of ALA6/7 was shown to result in pollen fertility defects. Here we show that a double knockout of ALA4/5 results in dwarfism, characterized by reduced growth in rosettes (6.5-fold), roots (4.3-fold), bolts (4.5-fold), and hypocotyls (2-fold). Reduced cell size was observed for multiple vegetative cell types, suggesting a role for ALA4/5 in cellular expansion. Members of the third ALA cluster are at least partially interchangeable, as transgenes expressing ALA6 in vegetative tissues partially rescued ala4/5 mutant phenotypes, and expression of ALA4 transgenes in pollen fully rescued ala6/7 mutant fertility defects. ALA4-GFP displayed plasma membrane and endomembrane localization patterns when imaged in both guard cells and pollen. Lipid profiling revealed ala4/5 rosettes had perturbations in glycerolipid and sphingolipid content. Assays in yeast revealed that ALA5 can flip a variety of glycerolipids and the sphingolipid sphingomyelin across membranes. These results support a model whereby the flippase activity of ALA4 and ALA5 impacts the homeostasis of both glycerolipids and sphingolipids and is important for cellular expansion during vegetative growth.

Lipid flippases of the P4-type ATPase family are enzymes that hydrolyze ATP to flip lipids across a membrane bilayer toward cytosolic facing leaflets, whether that be from the outer leaflet of the plasma membrane (PM) or luminal leaflet of internal membranes (López-Marqués et al., 2014). Flippase activity and proper subcellular localization requires a β-accessory subunit (e.g. cell division control protein 50 in yeast; Saito et al., 2004). Flippases are proposed to be involved in multiple processes, such as flipping specific lipids to generate an asymmetric distribution between the leaflets of a membrane bilayer, generating membrane curvature, assisting vesicle formation, importing lipids from the environment, and facilitating lipid metabolism (Seigneuret and Devaux, 1984; Gall et al., 2002; Pomorski et al., 2003; Poulsen et al., 2015; Roland et al., 2019).

In plants, flippases are referred to as Aminophospholipid ATPases (ALAs). The β-accessory subunits of ALAs are referred to as ALA-interacting subunits (ALISs; Poulsen et al., 2008). The Arabidopsis (Arabidopsis thaliana) genome contains 12 ALAs that form five phylogenetic clusters (Baxter et al., 2003). Representatives from all five ALA clusters have been studied with either loss-of-function knockout (KO) mutants or RNA interference (RNAi) lines, revealing phenotypes that include increased sensitivity to cold stress (ALA1 RNAi, cluster 1; Gomès et al., 2000), disrupted stomatal regulation and lipid desaturation levels (ala10 KO, cluster 2; Poulsen et al., 2015; Botella et al., 2016), increased sensitivity to heat and heavy metals as well as constitutive male fertility defects (ala6 KO, ala4 KO, and ala6/7 double KO, cluster 3; McDowell et al., 2015; Niu et al., 2017; Sanz-Fernández et al., 2017), impaired pathogen defenses and reduced growth in rosettes and trichomes (ala3 KO, cluster 4; Zhang and Oppenheimer, 2009; McDowell et al., 2013; Underwood et al., 2017), and impaired viral defenses (ala2 KO, cluster 5; Guo et al., 2017; Zhu et al., 2017).

Lipid transport preferences have been characterized for four of the five ALA clusters, with fluorescent lipid uptake studies in yeast providing evidence that ALAs from different clusters have distinct transport activities. For example, ALA1 from cluster 1 transports phosphatidylserine (PS) > phosphatidylethanolamine (PE; Gomès et al., 2000), ALA10 from cluster 2 transports phosphatidylcholine (PC) > PE > PS > phosphatidylglycerol (PG) > lysophosphatidylcholine (LysoPC; Poulsen et al., 2015; Jensen et al., 2017), ALA3 from cluster 4 transports PE > PC > PS (Poulsen et al., 2008), and ALA2 from cluster 5 transports only PS (López-Marqués et al., 2010). Additionally, ALA10 (cluster 2) has been shown to facilitate the uptake of sphingolipids, including sphingomyelin (SM) and glucosylceramide (GlcCer; Jensen et al., 2017). To date, the lipid transport activities for a cluster-3 ALA has not been reported.

In this study, we created a double-KO mutant for the two cluster-3 ALAs expressed in vegetative tissues, specifically ALA4 and ALA5. The ala4/5 mutations resulted in dwarfed plants with underlying defects in cellular expansion. This growth reduction could be restored by the expression of either ALA4 or ALA5, and partially restored by the ectopic expression of ALA6, another cluster-3 ALA normally expressed only in pollen. Lipidomic profiling of ala4/5 rosettes revealed perturbations in both glycerolipids and sphingolipids, including relative increases in the abundance of phosphatidic acid (PA), lysophosphatidylethanolamine (LysoPE), and GlcCers, as well as decreases in glycosylinositolphosphoceramides (GIPCs). Transport assays using ALA5 expressed in yeast provided evidence that cluster-3 ALAs can transport both glycerolipids (PC > PE > PS) and a phosphate-containing sphingolipid SM. Together, these results support a model whereby ALA4 and ALA5 function as lipid flippases that contribute to lipid homeostasis, cellular expansion, and plant growth.

RESULTS

In Arabidopsis, transcript expression profiles of genes encoding cluster-3 ALAs (ALA4–ALA7) generated from public databases (Winter et al., 2007; Huang et al., 2017) indicate that ALA4 and ALA5 are expressed primarily in vegetative structures, whereas ALA6 and ALA7 are expressed primarily in pollen (Fig. 1, A and B). A double KO of ALA6 and ALA7 was previously shown to result in pollen tubes that were short, slow growing, and hypersensitive to a heat stress (McDowell et al., 2015). To investigate the functions of ALA4 and ALA5 in vegetative tissues, plant lines harboring T-DNA disruptions in these genes were obtained from publicly available mutant collections (Fig. 1C; Sessions et al., 2002; Alonso et al., 2003; Woody et al., 2007).

Figure 1.

Figure 1.

Open in a new tab

Expression patterns of Arabidopsis cluster-3 ALAs and genetic organization of ALA4 and ALA5 T-DNA disruptions. A, Phylogenetic tree of Arabidopsis ALA family and other P-type ATPases generated in the Arabidopsis Heat Tree Viewer (http://arabidopsis-heat-tree.org), with ALA clusters (Baxter et al., 2003) bracketed by gray rectangles and thicker lines representing bootstrap values of 98 or greater. B, Database expression profile of ALA4, ALA5, ALA6, and ALA7 genes in rosette, root, hypocotyl, trichomes, and mature pollen, generated from the Arabidopsis eFP browser (Winter et al., 2007), with root hair data from Huang et al. (2017). Error bars represent ± sd; n = 3 expression replicates. Asterisks indicate not detected [ND]. C, Gene diagram of ALA4 and ALA5 with exons in black, introns in white, and triangles indicating position of T-DNA insertion. Allele name and accession number are detailed above gene schematics. D, Protein topology representing ALA4 and ALA5 with transmembrane domains labeled and represented by black boxes. Predicted truncated protein products are detailed below the protein schematic.

A Double Genetic Disruption of ALA4 and ALA5 Results in Dwarfism

Two independent sets of ala4/5 mutants, namely ala4-2/5-3 and ala4-3/5-1, were generated as homozygous double disruptions (Fig. 1C). The T-DNA insertions for ala4-3, ala5-1, and ala5-3 were located within exon-coding sequences and are predicted to disrupt the expression of full-length proteins (Fig. 1D), with any truncated protein products lacking essential features of a typical ATPase (e.g. active site, transmembrane domains). By contrast, the ala4-2 insertion in the second set of alleles (ala4-2/5-3) was located within an intron, which has the potential to be spliced out to allow production of full-length proteins. A qualitative reverse transcription-PCR (RT-PCR) analysis indicated that the ala4-3/5-1 allele set provides a complete KO for both ala4-3 and ala5-1 (Fig. 2G). However, the ala4-2/5-3 set represents a mixed combination of a KO for ala5-3 but only a knockdown (KD) for ala4-2. The ala4-2/5-3 set is hereafter referred to as ala4/5 KD, whereas ala4/5 KO refers to ala4-3/5-1.

Figure 2.

Figure 2.

Open in a new tab

Loss of ALA4 and ALA5 results in reduced vegetative growth. A, Representative image of a growth comparison assay. Scale bar = 1 cm. B, Rosette diameter quantification of wild-type (WT) Col-0, ala4-2/5-3 KD (ss948), ala4-3/5-1 KO (ss1270), Rescue 1 (35S::ALA4 in KO, ss2145, ps1712), and Rescue 2 (35S::ALA5 in KO, ss2146, ps1713) plants; n = 8 growth assays. C, Root length quantification from light-grown seedlings sampled over 7 d; n = 8 seedlings total from four growth assays. D, Representative images of dark-grown hypocotyls and quantifications of length (E) and width (F); n = 39 hypocotyls total from three growth assays. Arrows indicate root-shoot transition. Different letters indicate statistically significant difference (P < 0.05; Student’s t test). G, Expression of ALA4 and ALA5 transcripts in wild type, ala4-2/5-3 KD, and ala4-3/5-1 KO mutants. ND, not detected. Error bars represent ± se; n = 3 expression replicates.

Plants with either ala4/5 -KD or -KO allele sets showed a similar dwarf phenotype. Compared to wild type, the rosette diameters were 4-fold and 6.5-fold smaller for KD and KO plants (P < 1E-11), respectively (Fig. 2, A and B). KO mutants were also 1.6-fold smaller than KD mutants (P < 1E-04). Primary root lengths were shorter, ranging from 3.1-fold (KD) to 4.3-fold (KO) less than that of wild type after 7 d (Fig. 2C). Both ala4/5 mutants also exhibited hypocotyls that were 2-fold shorter and 1.5-fold thicker than wild-type controls after 5 d of growth (Fig. 2, D–F). Additionally, the bolt length of mature ala4/5 plants was reduced 2.3-fold (KD) to 4.5-fold (KO) relative to wild type (Supplemental Fig. S1). Importantly, neither of the ala4 /5 mutants displayed reductions in total chlorophyll content, suggesting that the plants were otherwise healthy (Supplemental Fig. S2).

To confirm that the T-DNA disruptions were the cause of dwarfism in ala4/5 mutants, transgenes encoding either ALA4 or ALA5 under the control of a Cauliflower mosaic virus-35S promoter (35S promoter) were expressed in the mutants, which restored normal rosette sizes (Fig. 2, A and B). Thus, two independent sets of gene disruptions and transgene-rescue experiments establish that a complete (or even partial) loss of ALA4/5 function results in severe plant dwarfing.

ala4/5 Mutants Display Reductions in Cell Size

To determine if the dwarfism of ala4/5 plants was correlated with smaller cell size, cell borders were stained with propidium iodide. Images showed ala4/5 mutants had significant reductions in leaf pavement cell area, ranging from 1.5-fold (KD) to 2.7-fold (KO) less than wild type (Fig. 3, A–D). Hypocotyl cells from ala4/5 mutants grown in the dark were 1.7-fold shorter for both allele sets, and 1.5-fold (KD) to 2-fold (KO) wider relative to wild type (Fig. 3, F–J). Guard cells were reduced in size for both mutants (P < 1E-05; Fig. 3E), although this difference was determined to be <10%.

Figure 3.

Figure 3.

Open in a new tab

Loss of ALA4 and ALA5 results in impaired cellular expansion in vegetative tissues. Confocal images of propidium iodide-stained leaf epidermal cells (A–C) and dark-grown hypocotyl cells (F–H) for wild-type Col-0, ala4-2/5-3 KD, and ala4-3/5-1 KO, with a single cell shaded gray in each image. Scale bars = 50 μm. Images were analyzed with Fiji software (Schindelin et al., 2012) to quantify leaf cell area (D), guard cell length (E), hypocotyl cell length (I), and hypocotyl cell width (J). Different letters indicate statistically significant difference (P < 0.01; Student’s t test). Error bars represent ± se; n = 20 cells total from three plants.

Trichomes and root hairs were also examined. Root hairs were clearly visible, but were shorter by 2.5-fold (KD) to 2.6-fold (KO) compared to wild type (Supplemental Fig. S3, A–D). Trichomes were nearly absent in the mutants, and when detected showed either globular deformities or short pin-like structures (Supplemental Fig. S3, E–G). The reduced cell sizes indicate that ALA4/5 flippase activity is important for cellular expansion of multiple cell types in vegetative tissues, especially those highly dependent on anisotropic growth.

Ectopic Expression of ALA6 and ALA4 Rescues the Phenotypes of ala4/5 and ala6/7 Mutants, Respectively

To determine if there was any biochemical redundancy between cluster-3 ALAs expressed in vegetative tissues (ALA4 /5) and pollen (ALA6/7), we tested the ability of ALA6 and ALA4 transgenic expression to rescue ala4/5 and ala6/7 mutant phenotypes, respectively. ALA6 expression driven by a promoter derived from ALA4 provided a partial reversal of the ala4/5 dwarf phenotype in all five independent transgenic lines evaluated (Fig. 4).

Figure 4.

Figure 4.

Open in a new tab

Ectopic expression of ALA6 in vegetative tissues restores growth in an ala4/5 double mutant. A, Representative image of a growth comparison assay. Scale bar = 1 cm. B, Rosette diameter quantification of wild-type Col-0, ala4-3/5-1 KO mutant (ss1270), and Rescue 1–2 (ALA4p::ALA6 in KO, ss2504-5, ps2809) plants. Shown are two out of five independent transgenic rescue lines with similar results. Different letters indicate statistically significant difference (P < 1E-06; Student’s t test). Error bars represent ± sd; n = 5 growth assays.

Conversely, ALA4 was transgenically expressed in ala6/7 mutants using a promoter from a gene encoding the autoinhibited Ca2+-ATPase9 (ACA9) that drives expression primarily in pollen (Fig. 5). The ALA4 transgene was able to confer a full restoration of seed set in all independent transgenic lines tested (Fig. 5, A and B). Additionally, reciprocal crosses were used to evaluate pollen transmission efficiencies (Fig. 5C) using the same three transgenic lines shown in Figure 5B. To quantify differences in transmission efficiencies between mutant pollen with or without an ACA9p::ALA4 rescue construct, F1 seedlings were scored for a hygromycin-resistance selectable marker associated with the transgene construct. A transmission efficiency ratio (TEr) was calculated as the ratio of resistant to sensitive F1 offspring, with a TEr = 1 expected for normal Mendelian transmission in an outcross. For transmission of the rescue-transgene through the female gametes, none of the plants tested showed evidence of a significant difference from the expected TEr of 1, which confirmed that each transgenic plant was heterozygous for a single transgene insertion. By contrast, transmission of the transgene through pollen showed TEr values that were ∼88-fold higher. This demonstrated that the ala6/7 mutant pollen harboring ALA4 transgenes were rescued and able to easily outcompete mutant pollen without a transgene-rescue construct. Thus, both ectopic expression studies indicate that ALA6 and ALA4 are at least partially interchangeable when the corresponding transgene for each is expressed ectopically in vegetative tissues or pollen, respectively.

Figure 5.

Figure 5.

Open in a new tab

Ectopic expression of ALA4 in pollen restores seed set and normal pollen transmission in an ala6/7 double KO. A, De-stained silique images.Scale bar = 5 mm. B, Seeds per silique counts of wild-type Col-0, ala6/7 mutants (ss1351), and ala6/7 lines rescued by a pollen-specific ACA9p::ALA4 construct with a hygromycin-resistance selectable marker (Transgene rescues: TG-1, TG-2, TG-3, ss2149-2151, and ps1967). Silique orientation shown with an arrow pointing from stigma end to base. An asterisk indicates statistically significant difference compared to wild-type (P < 1E-11; Student’s t test). Error bars represent ± sd; n = 10–15 siliques total from at least two plants, specified in figure. C, Hygromycin resistance and sensitivity counts of male to female outcrosses performed between Col-0, ala6/7 mutants, and rescue lines were used to calculate TErs after selection of F1 offspring germinated on hygromycin (TEr = #HygR/#HygS). Observed (Obs) and Expected (Exp) TErs are shown. Statistical confidence P values were calculated from a Pearson’s chi-squared test (χ2).

ALA4-GFP Fusions Are Localized to Both the PM and Endomembrane

To determine the subcellular localization of ALA4/5, confocal imaging was used to evaluate vegetative cell types of ala4/5 mutants rescued by 35S promoter-driven expression of ALA4-GFP transgenes (Fig. 6). Comparisons were made to the following controls: cytosolic (GFP), PM (ACA8-GFP), endoplasmic reticulum (ER; ACA2-GFP), and chloroplast autofluorescence. In guard cells, ALA4-GFP showed accumulation at both the PM and internal endomembrane structures, including the ER that wraps around the nucleus.

Figure 6.

Figure 6.

Open in a new tab

Expression of ALA4-GFP in guard cells shows endomembrane and PM localization. Fluorescence confocal images of guard cells. Images taken from stable transgenic plants expressing transgenes under the control of a 35S promoter. An arrow marks the nucleus. Cells shown are expressing: cytosolic GFP (ss1811, ps346), PM-localized ACA8-GFP (ss248, ps396), ER-localized ACA2-GFP (ss2216, ps660), and ALA4-GFP (ss2145, ps1712). Scale bar = 5 µm. Similar results were observed in at least three independent plants. DIC, differential interference contrast.

Similar experiments were performed in pollen that compared the localization of an ACA9p-driven ALA4-GFP transgene. Comparisons were made to the following controls: cytosolic (yellow fluorescent protein [YFP]), PM (ACA9-YFP), ER (ACA2-GFP), and ALA6-GFP controls (Fig. 7). In support of the vegetative localization results, ALA4-GFP accumulated at both PM and endomembrane structures, which was similar to the previously characterized ALA6-GFP control.

Figure 7.

Figure 7.

Open in a new tab

Ectopic expression of ALA4-GFP in pollen shows endomembrane and PM localization. Fluorescence confocal images of growing pollen tubes 1 h after germination. Pollen were from stable transgenic plants expressing transgenes under the control of an ACA9 promoter. Pollen shown are expressing: cytosolic YFP (ss2228, ps532), PM-localized ACA9-YFP (ss2229, ps580), ER-localized ACA2-GFP (ss2254, ps585), ALA4-GFP (ss2161, ps2096), and ALA6-GFP (ss2399, ps1958). Scale bar = 5 µm. Similar results were observed in at least three independent plants.

ala4/5 Plants Exhibit Altered Membrane Lipid Composition

To evaluate whether the loss of ALA4/5 might cause differences in the relative abundance of specific cellular lipids, a lipidomic survey of glycerolipids was performed on total lipid extracts taken from rosette tissues of 14-d–old ala4/5 seedlings (Fig. 8, A and B, for profiles as percentage of total glycerolipid; Supplemental Fig. S4, for data normalized to dry weight of tissue). The profiling included quantification of the following major lipid components that comprise most of the membrane content of rosette tissue: digalactosyldiacylglycerol, monogalactosyldiacylglycerol, PG, PC, PE, and phosphatidylinositol (PI). The following minor membrane components were also analyzed: PS, PA, lysophosphatidylglycerol (LysoPG), LysoPC, and LysoPE. All glycerolipids monitored are detailed in Supplemental Table S1. The ala4/5 mutants displayed PA concentrations that were increased 3.4-fold (KD) to 4.2-fold (KO), and LysoPE levels that were increased 1.6-fold (KD) to 1.4-fold (KO; Fig. 8). Whereas other changes were detected when values were normalized to tissue dry weight, these changes were attributable to a 1.4-fold (KD) and 1.5-fold (KO) reduction in total glycerolipid content (Supplemental Fig. S4), and were not observed when values were normalized to percent total glycerolipid. Importantly, increases in PA were detected for both mutants regardless of the normalization parameter used.

Figure 8.

Figure 8.

Open in a new tab

Loss of ALA4 and ALA5 results in perturbations in the abundance of membrane glycerolipids and sphingolipids in rosettes. The relative abundance of glycerolipids (A and B) and sphingolipids (C) in rosettes from wild-type Col-0, ala4-2/5-3 KD, and ala4-3/5-1 KO. Analyzed glycerolipids include major components digalactosyldiacylglycerol (DGDG), monogalactosyldiacylglycerol (MGDG), PG, PC, PE, and PI, as well as minor components PS, PA, LysoPG, LysoPC, and LysoPE, presented as percentage of total glycerolipid signal; n = 5 independent seedling pools profiled for glycerolipids. Analyzed sphingolipids include LCBs, Cers, hCers, GlcCers, and GIPCs, presented as percentage of total sphingolipid; n = 3 independent seedling pools profiled for sphingolipids. Single asterisks indicates statistically significant difference for only KD or KO mutants and double asterisks indicates statistically significant difference for both ala4/5 mutants compared to wild type (P < 0.05; Student’s t test). Error bars represent ± sd.

The sphingolipid content of ala4/5 KO and KD rosettes was also evaluated to determine if the abundances of these structurally distinct lipids were perturbed. The sphingolipidomic study included quantification of long chain bases (LCBs), ceramides (Cers), hydroxyceramides (hCers), GlcCers, and GIPCs. All sphingolipids monitored are detailed in Supplemental Table S2. Both ala4/5 mutants displayed 1.4-fold increased GlcCers and 1.3-fold reduced GIPCs (Fig. 8C). GlcCer increases and GIPC decreases were broadly distributed across sphingolipids containing C16–C26 fatty acid moieties (Supplemental Fig. S5), and were not specific to any single GlcCer or GIPC species. In contrast to GlcCers and GIPCs, there were no significant changes in LCBs, Cers, or hCers (Fig. 8C). Together, these results provide evidence that a deficiency in ALA4/5 flippase activity impacts the relative abundance of both glycerolipid and sphingolipid membrane contents.

ALA4/5 Are Flippases that Transport PC, SM, PE, and PS

To test the lipid transport capabilities associated with cluster-3 ALAs, ALA5 was expressed in yeast either alone or together with three different ALIS β-subunits. The yeast host was a Saccharomyces cerevisiae mutant with negligible background flippase activity due to a triple mutation of endogenous flippases (drs2Δdnf1,2Δ). Yeast transformants were incubated with 7-nitrobenz-2-oxa-1,3-diazole (NBD)-labeled lipid analogs of PG, PS, PC, PE, LysoPC, and SM (Fig. 9A). Subsequent flow cytometry analyses of these yeast cells revealed that expression of ALA5/ALIS complexes, but not ALA5 alone, increased the uptake of exogenous PC, PE, and SM. By contrast, no increase in uptake was detected for PG, PS, or LysoPC.

Figure 9.

Figure 9.

Open in a new tab

ALA5 can transport PC, PE, SM, and PS membrane lipids. The uptake of NBD-labeled PG, PS, PE, PC, LysoPC, and SM (A), as well as SM, PC, Cer, GlcCer, GalCer, and LacCer (B) in drs2Δ dnf1,2Δ S. cerevisiae lines expressing ALA5 alone or in combination with β-subunits ALIS1, ALIS3, or ALIS5, normalized to empty vector (e.v.) values. Data are averages ± sd from at least three independent experiments. A two-factor ANOVA followed by a Tukey’s Honest Significant Difference test was used for statistical analysis. Asterisks indicate statistically significant difference compared to e.v. (P < 0.05). C, The same lines were dropped onto media containing Glc and onto Gal plates containing no added toxins (ø) or increasing concentrations of papuamide A, duramycin, or miltefosine (direction of gradient indicated by triangles). Error bars represent ± sd.

To evaluate whether the perturbations in GlcCer content observed in ala4/5 mutants might occur from a direct loss of GlcCer transport, an additional uptake assay was performed to determine if ALA5 could transport NBD-labeled Cers or glycosphingolipids such as GlcCers, lactosylCers (LacCers), or galactosylCers (GalCers; Fig. 9B). A parallel analysis of GIPCs was not performed because NBD-labeled GIPCs are not currently commercially available. Comparisons were made to a positive control of yeast expressing Dnf1p, which was recently described as showing GlcCer transport activity (Roland et al., 2019). Whereas the ALA5/ALIS-expressing lines failed to show any detectable transport of Cers, GlcCers, LacCers, or GalCers, the Dnf1p control facilitated the uptake of GlcCers and GalCers (but not Cers or LacCers).

As a complementary approach to uptake assays, growth assays were employed to test survival of the previously mentioned yeast lines in the presence of drugs that are toxic when they bind specific surface-exposed lipids (Fig. 9C). The flippase deficiencies of drs2Δdnf1,2Δ yeast cells result in PS and PE accumulating on the outer leaflet of the PM, a trait that renders the cells sensitive to cytotoxic drugs papuamide A and duramycin that respectively bind to PS and PE. Only lines that expressed both ALA5 and an ALIS were able to survive high concentrations of papuamide A or duramycin, suggesting that ALA5 was capable of flipping both PS and PE from the exoplasmic to cytosolic membrane leaflet. It is noteworthy that despite the lack of NBD-PS uptake from exogenous sources (Fig. 9A), the ability to convey resistance to papuamide A provides evidence that ALA5 can mediate the flipping of PS already present in a membrane.

Additional yeast growth experiments were performed with miltefosine, an analog of LysoPC that is toxic when transported into a cell. In accordance to the previous uptake results (Fig. 9A), none of the ALA5/ALIS-expressing lines showed sensitivity to high concentrations of miltefosine, providing additional evidence that ALA5 is unable to transport LysoPC or its analogs.

DISCUSSION

ALA4 and ALA5 Are Important for Cellular Expansion and Vegetative Growth

In this study, we present genetic evidence that a double KO of ALA4 and ALA5 results in dwarfism, characterized by reduced growth in roots, hypocotyls, bolts, and rosettes (Fig. 2; Supplemental Fig. S1). This dwarfism was observed in two independent double-mutant lines, and was reversed through the expression of ALA4 or ALA5 transgenes driven by a 35S promoter (Fig. 2, A and B). Despite the reduced growth, ala4/5 plants were otherwise healthy with normal chlorophyll levels and no obvious chlorotic lesions indicative of cell death or poor health (Fig. 2A; Supplemental Fig. S2). To date, our observation of a 6.5-fold reduction in ala4/5 rosette diameter (Fig. 2B) is the most severe growth defect associated with a plant flippase deficiency, in comparison to an ∼1.2-fold reduction for an ala3 KO (McDowell et al., 2013), an ∼2.5-fold reduction for ALA1 RNAi lines grown at 8°C to 12°C (Gomès et al., 2000), and no significant changes reported for an ala2 KO (Zhu et al., 2017), an ala6/7 KO (McDowell et al., 2015), or an ala10 KO (Poulsen et al., 2015; Botella et al., 2016).

The loss of ALA4/5 also resulted in cell size reductions across multiple cell types, including leaf epidermal cells that showed a 2.7-fold reduction in surface area (Fig. 3). Whereas this indicates a deficiency in cell expansion, it does not by itself explain the more severe 6.5-fold reduction detected in overall rosette diameter. The disparity between organ and cell size reductions suggests that ala4/5 plants might also have a deficiency in generating or responding to growth signals, such as growth hormones.

ALA4/5 and ALA6/7 Are Functionally Interchangeable Flippases

ALA4–ALA7 all belong to the cluster-3 ALA subfamily (Fig. 1A) that is conserved between dicots and monocots (Baxter et al., 2003). Examination of expression profiles in public databases indicated that the genes encoding ALA4 and ALA5 are expressed primarily in vegetative tissues and ALA6/7 in pollen (Fig. 1B). Despite their expression in different cell types, ectopic expression of a pollen ALA6 transgene in vegetative tissues was capable of partially rescuing growth defects associated with ala4/5 mutants (Fig. 4). Additionally, a vegetative ALA4 ectopically expressed in pollen fully rescued the male-fertility defects associated with ala6 /7 mutants (Fig. 5). ALA4-GFP and ALA6-GFP also show similar endomembrane and PM localization patterns (Fig. 7). These results support a model whereby cluster-3 ALAs are at least partially redundant in biochemical and cellular functions. In support of this contention, the Oryza sativa genome contains only one cluster-3 ALA (Baxter et al., 2003), OsALA7, which presumably provides the equivalent ALA4, ALA5, ALA6, and ALA7–dependent functions to all tissue types.

Cluster-3 ALAs Are Flippases Capable of Transporting PC, SM, PE, and PS

Using a fluorescent lipid uptake assay in yeast, expression of ALA5 (with an ALIS β-subunit) was shown to mediate the transport of NBD-labeled PC > SM > PE, but not PS, PG, LysoPC, Cers, GlcCers, LacCers, or GalCers (Fig. 9, A and B). Whereas SM is not made in plants, the ability to transport SM indicates a potential for ALA4 and ALA5 to transport at least one type of sphingolipid. It is also noteworthy that ALA5 transported both NBD-PC and NBD-SM, which share an identical choline headgroup. This is consistent with reports that ALAs contain central cavities that accommodate the passage of lipids with specific headgroups (Jensen et al., 2017). However, no transport was detected for choline-containing LysoPC (i.e. PC lacking an sn-2 acyl group), indicating that a common choline headgroup is not the only feature of importance to ALA4 and ALA5 substrate specificity.

Whereas no uptake of exogenous NBD-PS was detected, a potentially more sensitive transport assay suggested that ALA5 can still flip endogenous PS already present within the membrane. In this alternative assay, ALA5/ALIS expression restored growth of a yeast flippase mutant in the presence of the toxin papuamide A (Fig. 9C), a depsipeptide thought to have a membrane pore-forming property that is potentiated when the toxin binds to surface-exposed PS (Parsons et al., 2006; Andjelic et al., 2008). Thus, ALA5 flippase activity was still able to remove endogenous PS from the outer leaflet of the PM and alleviate the toxicity of papuamide A, despite the lack of detectable uptake activity for an externally supplied NBD-labeled PS substrate. This suggests that the fluorescent NBD-lipid uptake assays might lack the sensitivity to detect very low levels of flippase activity, and leaves open the possibility that ALA4 and ALA5 provide functionally relevant levels of transport for some lipids that nevertheless fail to show a strong uptake in yeast assays.

Whereas ALA5 transport specificity (PC > SM > PE > PS; Fig. 9) was distinct from those reported for flippases from other ALA clusters (Gomès et al., 2000; López-Marqués et al., 2010; Poulsen et al., 2015; Jensen et al., 2017), the ability to flip PC and SM makes ALA5 activity most similar to that of ALA10 from cluster 2 (SM > PC > PE > PS > PG > LysoPC > GlcCers). Cluster-2 ALAs (ALA8–ALA12) are also the most closely related to cluster-3 ALAs in terms of overall protein sequence similarities, and the two groups might have conserved structural features related to substrate specificity. Nevertheless, the dwarfism observed in ala4/5 plants indicates that cluster-2 ALAs do not share functional redundancy with ALA4 and ALA5, despite their overlap in lipid transport preferences and similar levels of expression in all major vegetative tissues (Supplemental Fig. S6). Further research will be needed to identify and understand the unique biochemical or functional features associated with these two closely related ALA groups.

PA and LysoPE Accumulate in ala4/5 Mutants

Whereas glycerolipid profiling revealed that ala4/5 rosettes display increased levels of two low-abundance phospholipids, specifically PA (3.4- to 4-fold) and LysoPE (1.4- to 1.6-fold; Fig. 8B), it is not yet clear how these changes are related to the loss of ALA4 and ALA5 activity. PA is a diacyl lipid with a simple phosphate headgroup that is known to play important roles in signaling pathways that can impact plant growth (Hong et al., 2009). Increases in PA can result from multiple causes, including increased phospholipase D activity removing headgroups from more complex phospholipids (PC, PE, or PI), increased phosphorylation of diacylglycerol by diacylglycerol kinases, or decreases in the conversion of PA into other glycerolipids such as PC, PE, PI, PG, or PS (Athenstaedt and Daum, 1999; Testerink and Munnik, 2011). It is possible that the loss of ALA4 and ALA5 could directly or indirectly change one or more of these metabolic pathways. It is also noteworthy that the loss of ALA6 and ALA7 (from cluster 3) results in pollen grains with a similar ∼2-fold increase in PA (McDowell et al., 2015), indicating a consistent link among the ALA4–ALA7 clade of flippases and the regulation of PA homeostasis.

The small increase in LysoPE, a monoacyl PE molecule, can also be from multiple causes, including increased phospholipase A activity removing a fatty acid from the sn-2 position of the PE glycerol backbone (Bahn et al., 2003), or a decrease in the conversion of LysoPE to PE via LysoPE acyltransferases (Jasieniecka-Gazarkiewicz et al., 2017). LysoPE also has connections to growth signaling, as mutants deficient in LysoPE acyltransferases display LysoPE accumulation and growth inhibition. Interestingly, LysoPE is a known inhibitor of phospholipase D-dependent PA production in plants (Ryu et al., 1997), which raises the possibility that an increase in LysoPE might be part of a feed-back mechanism in ala4/5 mutants to try and renormalize PA concentrations.

Glycosphingolipid Metabolism Is Perturbed in ala4/5 Mutants

In addition to glycerolipid perturbations, sphingolipid profiling revealed that ala4/5 mutants had relative increases in GlcCers (1.4-fold) that were offset by decreases in GIPCs (1.3-fold; Fig. 8C). GlcCers are sphingolipids with a simple Glc headgroup, whereas GIPCs have a headgroup containing phosphate, inositol, and a variable chain of carbohydrates. Importantly, both glycosphingolipids are critical for plant growth and development (Wang et al., 2008; Mortimer et al., 2013; Msanne et al., 2015; Fang et al., 2016; Tartaglio et al., 2017). As with PA and LysoPE, it is not yet clear if changes in GlcCers or GIPCs result from changes in their biosynthesis or degradation. When considering an underlying cause, it is difficult to posit specific increases in the biosynthesis of GlcCers, as published examples show that increased production of GlcCers are accompanied by parallel increases in accumulation of GIPCs (Luttgeharm et al., 2015), which contrasts with the observed GIPC decreases in ala4/5 profiles (Fig. 8C). Reductions in GIPC biosynthesis also seems unlikely, as mutant lines defective in GIPC production did not show any accumulation of GlcCers (Tartaglio et al., 2017), which is also in contrast to what was observed in ala4/5 profiles. A simpler alternative is that the loss of ALA4 and ALA5 specifically decreases catabolism of GlcCers, or increases the relative catabolism of GIPCs.

There are multiple direct and indirect models to explain the mechanisms through which ala4/5-dependent changes in catabolism could impact GIPC homeostasis. For example, it is possible that ALA4 and ALA5 transport GIPCs across a membrane and away from degrading enzymes. An alternative indirect model is that ALA4 and ALA5 are involved in the bulk flow of GIPCs through a plant cell, perhaps by assisting in forming vesicles for removal away from a site of degradation. Regardless of the mechanism, the loss of ALA4 and ALA5 would then increase GIPCs exposure to their degrading enzymes, increase their rates of catabolism, and ultimately decrease their abundances.

The possibility for a specific reduction in GlcCer catabolism is especially intriguing, as lipid flippases have recently been implicated in GlcCer catabolism in mammals and fungi (Roland et al., 2019). The proposed model involves the lipid flippases Dnf1p/Dnf2p (Yeast) and ATP10D (Human) directly transporting GlcCers to cytosolic facing surfaces, thus exposing these lipids to degrading enzymes. As to whether ALA4 and ALA5 might play a similar direct role in GlcCer catabolism in plants, it is noteworthy that ALA5 failed to mediate the transport of NBD-labeled GlcCers in a fluorescent lipid uptake assay in yeast (Fig. 9B). However, we cannot rule out that ALA4 and ALA5 have the ability to transport low levels of endogenous GlcCers across a membrane bilayer, similar to the ability of ALA5 to remove endogenous PS from the PM surface and provide resistance to papuamide A without providing detectable uptake of externally supplied NBD-PS (Fig. 9, A and C). Similar to the above discussion concerning GIPCs, an alternative indirect model is that ALA4 and ALA5 are involved in the bulk flow of GlcCers through a plant cell, perhaps by assisting in forming vesicles for delivery to a site of degradation. After delivery, the GlcCers are possibly flipped for degradation by a non-ALA4 or non-ALA5 flippase, such as ALA1, which shows the highest levels of sequence similarity to GlcCer-transporting Dnf1p and ATP10D (Poulsen et al., 2015), or ALA10, which was previously shown to flip at least small amounts of GlcCers (Jensen et al., 2017). It is also possible that ALAs might form hetero-multimers in plants, with ALA4 and ALA5 potentially interacting with other ALAs or plant-specific proteins to form a more robust GlcCer flipping complex.

Modeling the Potential Mechanisms Behind Dwarfism in ala4/5 Mutants

Lipid flippases are involved in multiple cellular processes, which include maintaining the asymmetric composition of membrane leaflets, creating membrane curvature for vesicular trafficking, and facilitating lipid metabolism (Seigneuret and Devaux, 1984; Gall et al., 2002; Pomorski et al., 2003; Roland et al., 2019). Thus, dwarfism in ala4/5 mutants could arise from a deficiency in one or more of these important processes.

One possibility is that the accumulation of PA or LysoPE acts as a signal that inhibits cell expansion and plant growth. For example, PA is a dynamic secondary messenger that plays critical signaling roles in growth promotion (Hong et al., 2009). PA is also known to recruit and activate specific proteins to the surfaces of membranes, some of which can result in growth inhibition (Yao and Xue, 2018). Conversely, both PA and LysoPE have been shown to accumulate after the application of various biotic and abiotic stresses (Welti et al., 2002; Testerink and Munnik, 2005; Vu et al., 2015). Thus, PA or LysoPE accumulation in ala4/5 mutants could potentially disrupt growth signaling by instigating stress responses that inhibit growth, or by PA accumulation disrupting the normal recruitment, activation, and accumulation of important growth signaling proteins. However, whereas mutant lines that accumulate LysoPE have been shown to have growth reductions (Jasieniecka-Gazarkiewicz et al., 2017), these reductions were not as severe as those seen in ala4/5 mutants. Additionally, there are examples of plant lines that accumulate PA and display normal growth (Nakamura et al., 2009; Du et al., 2013). Thus, increases in PA and LysoPE alone cannot fully explain the growth defects of ala4/5 mutants.

In contrast to the rather minor growth deficiencies linked to PA or LysoPE perturbations, there are several examples of plant dwarfism correlated with sphingolipid accumulation (i.e. sphingolipidosis; Chen et al., 2008; Markham et al., 2011; Luttgeharm et al., 2015). Of the three previously characterized sphingolipidosis dwarfs, all displayed increases in Cers, hCers, GlcCers, and GIPCs containing C16 fatty acid moieties. Evidence suggests that these lines exhibited dwarfing due to the mislocalization of auxin transporters (Markham et al., 2011) and/or the accumulation of growth-inhibiting salicylic acid (Luttgeharm et al., 2015). As to whether ala4/5 mutants are also inhibited by a sphingolipid accumulation, it is worth noting that there are multiple features that make the sphingolipid profile of ala4/5 mutants distinct. First, the sphingolipid increases in ala4/5 plants were limited to GlcCers, and not other sphingolipid species (Fig. 8C). Additionally, the GlcCer increases in ala4/5 plants were more broadly distributed across fatty acid moieties from C16 to C26 (Supplemental Fig. S5), and were not restricted to C16 as reported for other mutants. Thus, it is not clear if ala4/5 mutants share a common underlying mechanism for a sphingolipidosis-induced growth defect, or if the growth defects are unrelated. Importantly, reductions in GIPC content are also associated with dwarfism (Fang et al., 2016; Tartaglio et al., 2017), and we cannot rule out that both GlcCer increases and GIPC reductions contribute independently to ala4/5 mutant growth impairments.

Regardless of whether ala4/5 mutant dwarfism is caused by an imbalance in the concentration of one or more specific lipid, or a pleiotropic defect related to a general impairment to vesicular trafficking, this study demonstrates that the ALA4 and ALA5 lipid flippases are of critical importance for cellular expansion and plant growth.

MATERIALS AND METHODS

Plant Materials

ALA4 (At1g17500) and ALA5 (At1g72700) mutant alleles were introgressed to generate two independent double T-DNA insertion mutants: ala4-2 (SALK_129551)/5-3 (SALK_049232; seed stock ss948) and ala4-3 (WiscDsLox_435D3)/5-1 (SAIL_613_F05; ss1270; Sessions et al., 2002; Alonso et al., 2003; Woody et al., 2007). The ala6-1 (SALK_150173)/7-2 (SALK_125598) double mutant (ss1351) was generated as described in McDowell et al. (2015). Mutant lines were obtained from the Arabidopsis Biological Resource Facility at Ohio State University and were backcrossed to the ecotype Columbia-0. Double-mutant lines were PCR-genotyped to ensure homozygosity of both alleles using primers listed in Supplemental Table S3.

Transgenic events were produced by a floral dip method (Clough and Bent, 1998). Transgenic plants were selected for hygromycin resistance (25 mg/L) or kanamycin resistance (50 mg/L) provided by plant expression vectors used in this study. Rescues of ala4-3/5-1 were generated by transforming with constructs containing either 35S::ALA4 (Rescue 1, ss2145, transformed with plasmid stock ps1712), 35S::ALA5 (Rescue 2, ss2146, ps1713), or ALA4p::ALA6 (Rescue1-2, ss2504-5, ps2809). Rescues of ala6-1/7-2 contained ACA9p::ALA4 (TG1-3: ss2149-51, all transformed with ps1967). Imaging was performed on plant lines transformed with the following transgenes: 35S::GFP (ss1811, ps346), 35S::ACA8-GFP (ss248, ps396), 35S::ACA2-GFP (ss2216, ps660), 35S::ALA4-GFP (ss2145, ps1712), ACA9p::YFP (ss2228, ps532), ACA9p::ACA9-YFP (ss2229, ps580), ACA9p::ACA2-GFP (ss2254, ps585), ACA9p::ALA4-GFP (ss2161, ps2096), and ACA9p::ALA6-GFP (ss2399, ps1958).

Growth Conditions

Seeds were spread on plates containing 0.5× Murashige and Skoog medium (Phytotechnology Laboratories, pH 5.7) with 1% (w/v) agar and 0.05% (w/v) MES. Plates were stratified for 72 h at 4°C and then transferred to room temperature (∼23°C) under constant light conditions. Root growth measurements were made daily for 7 d. After 10 d of growth, seedlings were transplanted into soil (Sunshine SMB-238; SunGro Horticulture) supplemented with Marathon pesticide and Cleary Turf and Ornamental Systemic Fungicide according to manufacturer’s directed use. Plants were grown to maturity in growth chambers (Percival Scientific) under a long-day photoperiod (16-h light at 22°C/8 h dark at 22°C, 70% humidity, and ∼125 μmol m−2 s−1 light). Plate-grown seedlings used in root hair assays were grown as described above, except on sterile cellulose acetate sheets (Research Products International) and imaged after 7 d of growth. Seedlings used in dark-grown hypocotyl experiments were imaged after 6 d of growth at room temperature in the absence of light. Images were analyzed using Fiji software (Schindelin et al., 2012) for size measurements. Student’s t test was used to determine statistical significance for all assays unless specified otherwise. F1 progeny from male and female outcrosses were scored for transgenes by germinating seeds on growth media supplemented with 25 mg/L of hygromycin B (Gold Biotechnology). Pearson’s chi-squared test (χ2) was used to determine statistical significance for transmission assays. Selection for kanamycin-resistant seedlings was performed on growth media supplemented with 50 mg/L of kanamycin.

Chlorophyll Measurements

The rosettes of 10-d–old seedlings were dissected from root tissues and incubated in DMSO at 65°C for 1 h to extract total chlorophyll, as described in Hiscox and Israelstam (1979). The A645 and 663 nm was measured on an Ultrospec II spectrophotometer (LKB Instruments) and the chlorophyll content was subsequently calculated using Arnon’s equations and normalized to fresh weight.

Phylogenetic Tree and Expression Profiles

Protein sequence similarity comparisons were generated using the online tool Arabidopsis Heat Tree Viewer (http://arabidopsis-heat-tree.org), which utilizes the sequence alignment software ClustalW (Thompson et al., 1994). Phylogenetic bootstrap values were presented as thick lines for values > 0.98. Gene expression data for ALA4–12 was taken from the Arabidopsis eFP browser (Winter et al., 2007) for rosette, root, hypocotyl, and mature pollen. Root hair expression data are from Huang et al. (2017).

Plasmid Construction

All primer sequences and plant expression plasmid sequences are available in Supplemental Table S3 and Supplemental File S1, respectively. For expression in plants, ALA4, ALA5, and ALA6 genomic coding regions were PCR-amplified from wild-type Arabidopsis (Arabidopsis thaliana) genomic DNA using Phusion High-Fidelity DNA Polymerase (New England Biolabs) and moved into plant transformation vectors derived from pGreen II (Hellens et al., 2000) that contained hygromycin-resistance selectable markers. The 35S::ACA8-GFP (ps396) was the exception that contained a kanamycin-resistance selectable marker. PCR-derived regions were sequence-verified.

For expression in yeast, the full-length coding sequence of ALA5 was PCR-amplified from wild-type Arabidopsis complementary DNA (cDNA) and moved into plasmid pENTR/D-TOPO using the pENTR/D-TOPO Cloning Kit (Invitrogen, Life Technologies), rendering plasmid pMP2035. The ALA5 cDNA was subsequently cloned into a yeast expression plasmid by homologous recombination, achieved by cotransforming yeast with EcoRI- and SalI-digested pMP4075 (López-Marqués et al., 2012) and ALA5 PCR fragments flanked by the 25-bp region homologous to the opened vector. Yeast cells were lysed with acid-washed 0.5-mm glass beads and the recombinant plasmid product (pMP4842) was isolated using the GenElute Plasmid Miniprep Kit (Sigma-Aldrich). The plasmid was bulked in Escherichia coli and sequence-verified. The final construct contained an untagged version of ALA5 under the control of Gal-inducible GAL1 promoter linked to a HIS3 selection marker. Cloning of ALIS into yeast plasmids containing a URA3 marker was described in Poulsen et al. (2008).

Qualitative RT-PCR

Total RNA was extracted from 9-d–old seedlings using an RNeasy Plant Mini Kit (QIAGEN) followed by reverse transcription to cDNA using an iScript cDNA Synthesis Kit (Bio-Rad). Amplification for qualitative RT-PCR was performed using ExTaq Polymerase (Takara) and primer combinations listed in Supplemental Table S3. PCR products were separated on a 1.5% (w/v) agarose gel and stained with ethidium bromide before imaging. Gels were analyzed using the software package Fiji (Schindelin et al., 2012) for pixel density measurements. Mean pixel densities were normalized to actin7 control.

Fixation and Propidium Iodide Staining

Intact hypocotyls and leaf sections were fixed using methods described in Kalantidis et al. (2000). Fixed tissues were incubated in sterile water with 7.48 μm of propidium iodide (Sigma-Aldrich) at room temperature overnight before imaging. Images were analyzed with the software Fiji (Schindelin et al., 2012) for cell size measurements.

Confocal and Light Microscopy

Dark-grown hypocotyls, trichomes, and root hairs were imaged using a 4× objective lens on a S6D dissection microscope (Leica). For subcellular localization and cell size determination, images were acquired using a model no. IX81 FV1000 confocal microscope (Olympus) with the software package FluoView v1.07.03.00 (Olympus). A 40× objective lens (numerical aperture 1.30) was used for PI-stained leaf epidermal cell and hypocotyl cell images, and a 60× objective lens (numerical aperture 1.42) for guard cell and pollen tube images. An Argon-Ion laser was used for excitation at 488 nm (eGFP and propidium iodide) and 515 nm (eYFP), and a HeNe laser was used for excitation at 543 nm (chloroplasts). Spectral emission windows used for imaging vegetative cell types were 500–545 nm (eGFP), 687–787 nm (chloroplasts), and 597–637 nm (propidium iodide). Spectral emission windows for pollen imaging were 500–600 nm (eGFP) and 545–595 nm (eYFP).

In Vitro Pollen Tube Growth

The pollen growth germination media was described in Boavida and McCormick (2007) and contained: 5 mm of CaCl2, 0.01% (w/v) H3 BO3, 5 mm of KCl, 10% (w/v) Suc, 1 mm of MgSO4 at pH 7.8, and 1.5% (w/v) low melting agarose (Nusieve). Pollen from stage 13–14 flowers was applied to 400 μL of solidified germination media on a microscope slide. Slides were incubated at room temperature (∼23°C) in a humidity chamber. Pollen tubes were grown for 1–2 h before imaging.

Glycerolipid Profiling

Plant tissue was obtained by dissecting 14-d–old rosettes and immediately performing lipid extractions as described in McDowell et al. (2015), with two exceptions, (1) the NaCl wash was excluded and (2) four additional 15-min extractions at 60°C using the lower phase of a solvent consisting of isopropanol/hexane/water (50:20:25, v/v/v; Markham and Jaworski, 2007). Extracts from five replicates were analyzed by the KS Lipidomics Research Center (https://www.k-state.edu/lipid/) using a routine polar lipid analysis in which 156 polar lipids were quantified using precursor and neutral loss electrospray ionization tandem mass spectrometry as described in Shiva et al. (2013). Electrospray ionization voltage was set at −17 V for LysoPG quantification, and 40 V for all other glycerolipids. Compound formula, mass transition, mode, scan function, and ions detected for all profiled glycerolipids are provided in Supplemental Table S1.

Sphingolipid Profiling

Rosettes were dissected from 14-d–old seedlings grown on an agar surface and flash-frozen in liquid nitrogen, followed by freeze-drying at −40°C and 133 E10−3 mBAR for 3 h in a Freezone 18 Freeze Dry System (Labconco). Sphingolipid extraction was performed as described in Markham and Jaworski (2007). Lyophilized root tissue was extracted in triplicate using 2–5 mg of tissue per sample. The subsequent liquid chromatography electrospray ionization tandem mass profiling of total sphingolipid was performed using a Prominence ultra-high performance liquid chromatograph (Shimadzu) feeding into a QTRAP4000 mass spectrometer (AB SCIEX). Electrospray ionization voltage was set at 5,000 V. Sphingolipids were separated on a Zorbax Eclipse Plus narrow bore RRHT C18 column, 2.1 × 100 mm, 1.8-μm particle size (Agilent) at 40°C and a flow rate of 0.2 mL/min using a binary gradient as described in Kimberlin et al. (2013). Sphingolipids from three replicates were profiled and quantified using the Multiple Reaction Monitoring method described in Markham and Jaworski (2007), with instrument-specific modifications as described by Kimberlin et al. (2013). Compound formulae, mass transitions (Q1/Q3), mode, scan function, and ions detected for all profiled sphingolipids are provided in Supplemental Table S2. Data analysis was performed using the softwares Analyst 1.5 and Multiquant 2.1 (AB SCIEX) as described in Markham and Jaworski (2007).

Yeast Strain and Culture

Saccharomyces cerevisiae mutant strain ZHY709 (MATα his3 leu2 ura3 met15 dnf1Δ dnf2Δ drs2::LEU2; Hua et al., 2002) was used as a host. For coexpression of ALA5 with ALIS genes, cells were simultaneously transformed by a lithium acetate method (Gietz and Woods, 2002) with two individual plasmids bearing the desired genes and His (ALA5) or uracil (ALIS) auxotrophic markers. For growth assays, transformants were grown in Synthetic Complete Gal (SG) medium (0.7% [w/v] Yeast Nitrogen Base, 2% [w/v] Gal,1.4 g/L yeast synthetic dropout medium lacking His and uracil; Sigma-Aldrich) at 30°C with 150-RPM shaking for 4 h to induce overexpression. Cultures were diluted with water to 0.1 OD600/mL, and either 5 or 3 μL was spotted onto solid SG (2% (w/v) agar added) or synthetic complete Glc (synthetic defined) medium plates (SG with 2% [w/v] Glc instead of Gal). SG solid media containing toxin gradients were prepared as described in Liu et al. (2011), except that they were stored at 4°C for 2 d before use to allow for diffusion of the toxins. Gradients contained the following maximum concentrations: 0.3 μg/mL of papuamide A (Flintbox; Lynsey Huxham), 1.6 μm of duramycin (Sigma-Aldrich), or 2.5 μg/mL of miltefosine (hexadecylphosphocholine; Calbiochem). Plates were incubated at 30°C for 4 d before imaging. For NBD-lipid uptake assays, transformants were grown as described in López-Marqués et al. (2010).

NBD-Lipid Uptake Assays

Fluorescent NBD lipids were purchased from Avanti Polar Lipids: NBD-PS, NBD-PE, NBD-PC, NBD-PG, NBD-LysoPC, NBD-SM, NBD-Cer, NBD-GlcCer, NBD-GalCer, and NBD-LacCer. All NBD lipid stocks (4 mm) were prepared in DMSO. Uptake experiments in yeast were performed as described in Poulsen et al. (2015). Cells were resuspended to 10 OD600/mL and incubated in selective SG medium supplemented with 32 μm of NBD lipids for 30 min at 30°C with periodic mixing. The cells were washed three times in ice-cold selection media that lacked Gal but contained 2% (w/v) sorbitol, 3% (w/v) bovine serum albumin, and 20 mm of NaN3. Flow cytometry was performed on a BD FACS (Becton Dickinson) equipped with an argon laser using the software CellQuest (Becton Dickinson). Before analysis, 107 cells were labeled with 1 μL of 1 mg/mL propidium iodide for staining of nonviable cells. Twenty-thousand cells were analyzed. Data were analyzed using the software Cyflogic (CyFlo). Viable yeast cells were selected based on forward/side-scatter gating and propidium iodide exclusion. NBD fluorescence of living cells was plotted on a histogram and the geometric-mean fluorescence intensity was used for further statistical analysis. Statistical significance was determined by a Tukey’s Honest Significant Difference test.

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL data libraries under accession numbers ALA4 (At1g17500), ALA5 (At1g72700), ALA6 (At1g54280), ALA7 (At3g13900). Arabidopsis Mutants: ala4-2 (SALK_129561), ala4-3 (WiscDsLox_435D3), ala5-1 (SAIL_613_F05), ala5-3 (SALK_049232), ala6-1 (SALK_150173), ala7-2 (SALK_125598).

Supplemental Data

The following supplemental materials are available.

Acknowledgments

The glycerolipid analyses described in this work were performed at the Kansas Lipidomics Research Center Analytical Laboratory.

Footnotes

1

This work was supported by the United States Department of Agriculture (HATCH grant no. NEV00384 to J.F.H.), the National Science Foundation (IOS grant no. 1656774 to J.F.H., MCB grant no. 1818297 to E.B.C., EPS grant no. 0236913, MCB grant nos. 1413036 and 0920663, and DBI grant nos. 0521587 and 1228622 for instrument acquisition and glycerolipidomics method development), the Villum Fonden (project no. 13234 to R.L.L.-M.), the Innovation Fund Denmark (project no. 8053-00035B to M.P.), the National Institute of General Medical Sciences of the National Institutes of Health (grant no. P20 GM103554 for microscopy), Kansas Technology Enterprise Corporation, K-IDeA Networks of Biomedical Research Excellence of the National Institutes of Health (grant no. P20GM103418), and Kansas State University.

[OPEN]

Articles can be viewed without a subscription.

References

  1. Alonso JM, Stepanova AN, Leisse TJ, Kim CJ, Chen H, Shinn P, Stevenson DK, Zimmerman J, Barajas P, Cheuk R, et al. (2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 301: 653–657 [DOI] [PubMed] [Google Scholar]
  2. Andjelic CD, Planelles V, Barrows LR(2008) Characterizing the anti-HIV activity of papuamide A. Mar Drugs 6: 528–549 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Athenstaedt K, Daum G(1999) Phosphatidic acid, a key intermediate in lipid metabolism. Eur J Biochem 266: 1–16 [DOI] [PubMed] [Google Scholar]
  4. Bahn SC, Lee HY, Kim HJ, Ryu SB, Shin JS(2003) Characterization of Arabidopsis secretory phospholipase A2-γ cDNA and its enzymatic properties. FEBS Lett 553: 113–118 [DOI] [PubMed] [Google Scholar]
  5. Baxter I, Tchieu J, Sussman MR, Boutry M, Palmgren MG, Gribskov M, Harper JF, Axelsen KB(2003) Genomic comparison of P-type ATPase ion pumps in Arabidopsis and rice. Plant Physiol 132: 618–628 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Boavida LC, McCormick S(2007) Temperature as a determinant factor for increased and reproducible in vitro pollen germination in Arabidopsis thaliana. Plant J 52: 570–582 [DOI] [PubMed] [Google Scholar]
  7. Botella C, Sautron E, Boudiere L, Michaud M, Dubots E, Yamaryo-Botté Y, Albrieux C, Marechal E, Block MA, Jouhet J(2016) ALA10, a phospholipid flippase, controls FAD2/FAD3 desaturation of phosphatidylcholine in the ER and affects chloroplast lipid composition in Arabidopsis thaliana. Plant Physiol 170: 1300–1314 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chen M, Markham JE, Dietrich CR, Jaworski JG, Cahoon EB(2008) Sphingolipid long-chain base hydroxylation is important for growth and regulation of sphingolipid content and composition in Arabidopsis. Plant Cell 20: 1862–1878 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Clough SJ, Bent AF(1998) Floral dip: A simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant J 16: 735–743 [DOI] [PubMed] [Google Scholar]
  10. Du ZY, Chen MX, Chen QF, Xiao S, Chye ML(2013) Arabidopsis acyl-CoA-binding protein ACBP1 participates in the regulation of seed germination and seedling development. Plant J 74: 294–309 [DOI] [PubMed] [Google Scholar]
  11. Fang L, Ishikawa T, Rennie EA, Murawska GM, Lao J, Yan J, Tsai AY, Baidoo EE, Xu J, Keasling JD, et al. (2016) Loss of inositol phosphorylceramide sphingolipid mannosylation induces plant immune responses and reduces cellulose content in Arabidopsis. Plant Cell 28: 2991–3004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gall WE, Geething NC, Hua Z, Ingram MF, Liu K, Chen SI, Graham TR(2002) Drs2p-dependent formation of exocytic clathrin-coated vesicles in vivo. Curr Biol 12: 1623–1627 [DOI] [PubMed] [Google Scholar]
  13. Gietz RD, Woods RA(2002) Transformation of yeast by lithium acetate/single-stranded carrier DNA/polyethylene glycol method. Methods Enzymol 350: 87–96 [DOI] [PubMed] [Google Scholar]
  14. Gomès E, Jakobsen MK, Axelsen KB, Geisler M, Palmgren MG(2000) Chilling tolerance in Arabidopsis involves ALA1, a member of a new family of putative aminophospholipid translocases. Plant Cell 12: 2441–2454 [PMC free article] [PubMed] [Google Scholar]
  15. Guo Z, Lu J, Wang X, Zhan B, Li W, Ding S-W(2017) Lipid flippases promote antiviral silencing and the biogenesis of viral and host siRNAs in Arabidopsis. Proc Natl Acad Sci USA 114: 1377–1382 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hellens RP, Edwards EA, Leyland NR, Bean S, Mullineaux PM(2000) pGreen: A versatile and flexible binary Ti vector for Agrobacterium-mediated plant transformation. Plant Mol Biol 42: 819–832 [DOI] [PubMed] [Google Scholar]
  17. Hiscox JD, Israelstam GF(1979) A method for the extraction of chlorophyll from leaf tissue without maceration. Can J Bot 57: 1332–1334 [Google Scholar]
  18. Hong Y, Devaiah SP, Bahn SC, Thamasandra BN, Li M, Welti R, Wang X(2009) Phospholipase D epsilon and phosphatidic acid enhance Arabidopsis nitrogen signaling and growth. Plant J 58: 376–387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hua Z, Fatheddin P, Graham TR(2002) An essential subfamily of Drs2p-related P-type ATPases is required for protein trafficking between Golgi complex and endosomal/vacuolar system. Mol Biol Cell 13: 3162–3177 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Huang L, Shi X, Wang W, Ryu KH, Schiefelbein J(2017) Diversification of root hair development genes in vascular plants. Plant Physiol 174: 1697–1712 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jasieniecka-Gazarkiewicz K, Lager I, Carlsson AS, Gutbrod K, Peisker H, Dörmann P, Stymne S, Banaś A(2017) Acyl-CoA:Lysophosphatidylethanolamine acyltransferase activity regulates growth of Arabidopsis. Plant Physiol 174: 986–998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Jensen MS, Costa SR, Duelli AS, Andersen PA, Poulsen LR, Stanchev LD, Gourdon P, Palmgren M, Günther Pomorski T, et al. (2017) Phospholipid flipping involves a central cavity in P4 ATPases. Sci Rep 7: 17621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kalantidis K, Briarty LG, Wilson ZA(2000) Arabidopsis mutant characterization: Microscopy, mapping, and gene expression analysis In Wilson ZA, ed, Arabidopsis: A Practical Approach. Oxford University Press, Oxford, pp 93–94 [Google Scholar]
  24. Kimberlin AN, Majumder S, Han G, Chen M, Cahoon RE, Stone JM, Dunn TM, Cahoon EB(2013) Arabidopsis 56-amino acid serine palmitoyltransferase-interacting proteins stimulate sphingolipid synthesis, are essential, and affect mycotoxin sensitivity. Plant Cell 25: 4627–4639 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liu Y, Li J, Du J, Hu M, Bai H, Qi J, Gao C, Wei T, Su H, Jin J, et al. (2011) Accurate assessment of antibiotic susceptibility and screening resistant strains of a bacterial population by linear gradient plate. Sci China Life Sci 54: 953–960 [DOI] [PubMed] [Google Scholar]
  26. López-Marqués RL, Poulsen LR, Hanisch S, Meffert K, Buch-Pedersen MJ, Jakobsen MK, Pomorski TG, Palmgren MG(2010) Intracellular targeting signals and lipid specificity determinants of the ALA/ALIS P4-ATPase complex reside in the catalytic ALA α-subunit. Mol Biol Cell 21: 791–801 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. López-Marqués RL, Poulsen LR, Palmgren MG(2012) A putative plant aminophospholipid flippase, the Arabidopsis P4 ATPase ALA1, localizes to the plasma membrane following association with a β-subunit. PLoS One 7: e33042. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. López-Marqués RL, Theorin L, Palmgren MG, Pomorski TG(2014) P4-ATPases: Lipid flippases in cell membranes. Pflugers Arch 466: 1227–1240 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Luttgeharm KD, Chen M, Mehra A, Cahoon RE, Markham JE, Cahoon EB(2015) Overexpression of Arabidopsis ceramide synthases differentially affects growth, sphingolipid metabolism, programmed cell death, and mycotoxin resistance. Plant Physiol 169: 1108–1117 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Markham JE, Jaworski JG(2007) Rapid measurement of sphingolipids from Arabidopsis thaliana by reversed-phase high-performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry. Rapid Commun Mass Spectrom 21: 1304–1314 [DOI] [PubMed] [Google Scholar]
  31. Markham JE, Molino D, Gissot L, Bellec Y, Hématy K, Marion J, Belcram K, Palauqui J-C, Satiat-Jeunemaître B, Faure J-D(2011) Sphingolipids containing very-long-chain fatty acids define a secretory pathway for specific polar plasma membrane protein targeting in Arabidopsis. Plant Cell 23: 2362–2378 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. McDowell SC, López-Marqués RL, Cohen T, Brown E, Rosenberg A, Palmgren MG, Harper JF(2015) Loss of the Arabidopsis thaliana P4-ATPases ALA6 and ALA7 impairs pollen fitness and alters the pollen tube plasma membrane. Front Plant Sci 6: 197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. McDowell SC, López-Marqués RL, Poulsen LR, Palmgren MG, Harper JF(2013) Loss of the Arabidopsis thaliana P4-ATPase ALA3 reduces adaptability to temperature stresses and impairs vegetative, pollen, and ovule development. PLoS One 8: e62577. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mortimer JC, Yu X, Albrecht S, Sicilia F, Huichalaf M, Ampuero D, Michaelson LV, Murphy AM, Matsunaga T, Kurz S, et al. (2013) Abnormal glycosphingolipid mannosylation triggers salicylic acid-mediated responses in Arabidopsis. Plant Cell 25: 1881–1894 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Msanne J, Chen M, Luttgeharm KD, Bradley AM, Mays ES, Paper JM, Boyle DL, Cahoon RE, Schrick K, Cahoon EB(2015) Glucosylceramides are critical for cell-type differentiation and organogenesis, but not for cell viability in Arabidopsis. Plant J 84: 188–201 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Nakamura Y, Koizumi R, Shui G, Shimojima M, Wenk MR, Ito T, Ohta H(2009) Arabidopsis lipins mediate eukaryotic pathway of lipid metabolism and cope critically with phosphate starvation. Proc Natl Acad Sci USA 106: 20978–20983 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Niu Y, Qian D, Liu B, Ma J, Wan D, Wang X, He W, Xiang Y(2017) ALA6, a P4-type ATPase, is involved in heat stress responses in Arabidopsis thaliana. Front Plant Sci 8: 1732. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Parsons AB, Lopez A, Givoni IE, Williams DE, Gray CA, Porter J, Chua G, Sopko R, Brost RL, Ho C-H, et al. (2006) Exploring the mode-of-action of bioactive compounds by chemical-genetic profiling in yeast. Cell 126: 611–625 [DOI] [PubMed] [Google Scholar]
  39. Pomorski T, Lombardi R, Riezman H, Devaux PF, van Meer G, Holthuis JC(2003) Drs2p-related P-type ATPases Dnf1p and Dnf2p are required for phospholipid translocation across the yeast plasma membrane and serve a role in endocytosis. Mol Biol Cell 14: 1240–1254 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Poulsen LR, López-Marqués RL, McDowell SC, Okkeri J, Licht D, Schulz A, Pomorski T, Harper JF, Palmgren MG(2008) The Arabidopsis P4-ATPase ALA3 localizes to the Golgi and requires a beta-subunit to function in lipid translocation and secretory vesicle formation. Plant Cell 20: 658–676 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Poulsen LR, López-Marqués RL, Pedas PR, McDowell SC, Brown E, Kunze R, Harper JF, Pomorski TG, Palmgren M(2015) A phospholipid uptake system in the model plant Arabidopsis thaliana. Nat Commun 6: 7649. [DOI] [PubMed] [Google Scholar]
  42. Roland BP, Naito T, Best JT, Arnaiz-Yépez C, Takatsu H, Yu RJ, Shin H-W, Graham TR(2019) Yeast and human P4-ATPases transport glycosphingolipids using conserved structural motifs. J Biol Chem 294: 1794–1806 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Ryu SB, Karlsson BH, Ozgen M, Palta JP(1997) Inhibition of phospholipase D by lysophosphatidylethanolamine, a lipid-derived senescence retardant. Proc Natl Acad Sci USA 94: 12717–12721 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Saito K, Fujimura-Kamada K, Furuta N, Kato U, Umeda M, Tanaka K(2004) Cdc50p, a protein required for polarized growth, associates with the Drs2p P-type ATPase implicated in phospholipid translocation in Saccharomyces cerevisiae. Mol Biol Cell 15: 3418–3432 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sanz-Fernández M, Rodríguez-Serrano M, Sevilla-Perea A, Pena L, Mingorance MD, Sandalio LM, Romero-Puertas MC(2017) Screening Arabidopsis mutants in genes useful for phytoremediation. J Hazard Mater 335: 143–151 [DOI] [PubMed] [Google Scholar]
  46. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, et al. (2012) Fiji: An open-source platform for biological-image analysis. Nat Methods 9: 676–682 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Seigneuret M, Devaux PF(1984) ATP-dependent asymmetric distribution of spin-labeled phospholipids in the erythrocyte membrane: Relation to shape changes. Proc Natl Acad Sci USA 81: 3751–3755 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sessions A, Burke E, Presting G, Aux G, McElver J, Patton D, Dietrich B, Ho P, Bacwaden J, Ko C, et al. (2002) A high-throughput Arabidopsis reverse genetics system. Plant Cell 14: 2985–2994 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Shiva S, Vu HS, Roth MR, Zhou Z, Marepally SR, Nune DS, Lushington GH, Visvanathan M, Welti R(2013) Lipidomic analysis of plant membrane lipids by direct infusion tandem mass spectrometr In Munnik T, and Heilmann I, eds, Plant Lipid Signaling Protocols Methods in Molecular Biology (Methods and Protocols), Vol 1009 Humana Press, Totowa, NJ, pp 79–91 [DOI] [PubMed] [Google Scholar]
  50. Tartaglio V, Rennie EA, Cahoon R, Wang G, Baidoo E, Mortimer JC, Cahoon EB, Scheller HV(2017) Glycosylation of inositol phosphorylceramide sphingolipids is required for normal growth and reproduction in Arabidopsis. Plant J 89: 278–290 [DOI] [PubMed] [Google Scholar]
  51. Testerink C, Munnik T(2005) Phosphatidic acid: A multifunctional stress signaling lipid in plants. Trends Plant Sci 10: 368–375 [DOI] [PubMed] [Google Scholar]
  52. Testerink C, Munnik T(2011) Molecular, cellular, and physiological responses to phosphatidic acid formation in plants. J Exp Bot 62: 2349–2361 [DOI] [PubMed] [Google Scholar]
  53. Thompson JD, Higgins DG, Gibson TJ(1994) CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 4673–4680 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Underwood W, Ryan A, Somerville SC(2017) An Arabidopsis lipid flippase is required for timely recruitment of defenses to the host-pathogen interface at the plant cell surface. Mol Plant 10: 805–820 [DOI] [PubMed] [Google Scholar]
  55. Vu HS, Roston R, Shiva S, Hur M, Wurtele ES, Wang X, Shah J, Welti R(2015) Modifications of membrane lipids in response to wounding of Arabidopsis thaliana leaves. Plant Signal Behav 10: e1056422. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wang W, Yang X, Tangchaiburana S, Ndeh R, Markham JE, Tsegaye Y, Dunn TM, Wang GL, Bellizzi M, Parsons JF, et al. (2008) An inositolphosphorylceramide synthase is involved in regulation of plant programmed cell death associated with defense in Arabidopsis. Plant Cell 20: 3163–3179 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Welti R, Li W, Li M, Sang Y, Biesiada H, Zhou HE, Rajashekar CB, Williams TD, Wang X(2002) Profiling membrane lipids in plant stress responses: Role of phospholipase D alpha in freezing-induced lipid changes in Arabidopsis. J Biol Chem 277: 31994–32002 [DOI] [PubMed] [Google Scholar]
  58. Winter D, Vinegar B, Nahal H, Ammar R, Wilson GV, Provart NJ(2007) An “electronic fluorescent pictograph” browser for exploring and analyzing large-scale biological data sets. PLoS One 2: e718. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Woody ST, Austin-Phillips S, Amasino RM, Krysan PJ(2007) The WiscDsLox T-DNA collection: An Arabidopsis community resource generated by using an improved high-throughput T-DNA sequencing pipeline. J Plant Res 120: 157–165 [DOI] [PubMed] [Google Scholar]
  60. Yao HY, Xue HW(2018) Phosphatidic acid plays key roles regulating plant development and stress responses. J Integr Plant Biol 60: 851–863 [DOI] [PubMed] [Google Scholar]
  61. Zhang X, Oppenheimer DG(2009) IRREGULAR TRICHOME BRANCH 2 (ITB2) encodes a putative aminophospholipid translocase that regulates trichome branch elongation in Arabidopsis. Plant J 60: 195–206 [DOI] [PubMed] [Google Scholar]
  62. Zhu B, Gao H, Xu G, Wu D, Song S, Jiang H, Zhu S, Qi T, Xie D(2017) Arabidopsis ALA1 and ALA2 mediate RNAi-based antiviral immunity. Front Plant Sci 8: 422. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES