Yeast phosphatidic acid phosphatase Pah1 hops and scoots along the membrane phospholipid bilayer

Phosphatidic acid (PA) phosphatase, encoded by PAH1 in the yeast Saccharomyces cerevisiae, catalyzes the Mg2+-dependent dephosphorylation of PA, producing diacylglycerol at the nuclear/endoplasmic reticulum (ER) membrane. This enzyme plays a major role in triacylglycerol synthesis and in the regulation of phospholipid synthesis. As an interfacial enzyme, PA phosphatase interacts with the membrane surface, binds its substrate, and catalyzes its reaction. The Triton X-100/PA-mixed micellar system has been utilized to examine the activity and regulation of yeast PA phosphatase. This system, however, does not resemble the in vivo environment of the membrane phospholipid bilayer. We developed an assay system that mimics the nuclear/ER membrane to assess PA phosphatase activity. PA was incorporated into unilamellar phospholipid vesicles (liposomes) composed of the major nuclear/ER membrane phospholipids phosphatidylcholine, phosphatidylethanolamine, phosphatidylinositol, and phosphatidylserine. We optimized this system to support enzyme-liposome interactions and to afford activity that is greater than that obtained with the aforementioned detergent system. Activity was regulated by phospholipid composition, whereas the enzyme's interaction with liposomes was insensitive to composition. Greater activity was attained with large (> 100 nm) versus small (50 nm) vesicles. The fatty-acyl moiety of PA had no effect on this activity. PA phosphatase activity was dependent on the bulk (hopping mode) and surface (scooting mode) concentrations of PA suggesting a mechanism by which the enzyme operates along the nuclear/ER membrane in vivo.

as a carbon source (1,42). Overall, the pah1 mutant exhibits a shortened chronological life span (40) with apoptotic cell death in the stationary phase (39). The mutant phenotypes that are ascribed to elevated PA content (6) can also be shown in wild-type cells by the overexpression of Dgk1 DAG kinase, which catalyzes the CTP-dependent conversion of DAG to PA ( Fig. 1) (51,52).
PA phosphatase is an interfacial enzyme; it interacts with the membrane surface, binds its substrate, and then catalyzes the production of DAG. The detergent/phospholipidmixed micelle system has facilitated the kinetic analyses of Model for the regulation of Pah1 PA phosphatase (PAP) localization and stability by phosphorylation and dephosphorylation, and the role of PA phosphatase in lipid synthesis. PA phosphatase is phosphorylated (small white circles) by multiple protein kinases in the cytoplasm; the phosphorylated enzyme translocates to the nuclear/ER membrane through its dephosphorylation by the Nem1-Spo7 complex. Dephosphorylated PA phosphatase that is associated with the nuclear/ER membrane catalyzes the conversion of PA to DAG, which is converted to TAG for storage in lipid droplets (LDs). The PA phosphatase molecule catalyzing the dephosphorylation of PA is denoted with blue color. The DAG produced in the PA phosphatase reaction may be converted to the phospholipids PC and/or PE via the Kennedy pathway if cells are supplemented with choline and/or ethanolamine. DAG may be converted to PA by the CTP-dependent DAG kinase (DGK) reaction. The PA phosphatase substrate PA is used for the synthesis of the major ER membrane phospholipids (PC, PE, PI, PS) via CDP-DAG. The unphosphorylated/dephosphorylated forms of PA phosphatase or protein kinase C (PKC)-phosphorylated PA phosphatase are degraded by the 20S proteasome (indicated by the dashed line arrows). More comprehensive descriptions of the lipid synthetic pathways and the regulation of PA phosphatase may be found elsewhere (5,6,11,12). several interfacial phospholipid synthetic (75)(76)(77)(78)(79)(80) and degrading (81)(82)(83)(84)(85)(86)(87)(88)(89)(90)(91)(92) enzymes. The in vitro assay to measure yeast PA phosphatase activity has been performed with PA solubilized in the detergent Triton X-100 (1,(93)(94)(95)(96). Triton X-100 forms a uniform mixed micelle with PA, providing a surface for catalysis (93). The detergent micelle system has permitted defined studies on the kinetics of the PA phosphatase reaction (1,93) as well as on the biochemical regulation of the enzyme by phospholipids (96), sphingolipids (94), nucleotides (95), and by phosphorylation (64)(65)(66)68). Although the Triton X-100/PA-mixed micelle assay allows for defined activity measurements, it does not resemble the in vivo environment of phospholipid bilayer membrane where PA is a component. Accordingly, we sought to establish an assay condition that mimics the in vivo environment of the nuclear/ER membrane in measuring PA phosphatase activity. PA was incorporated into unilamellar vesicles (liposomes) composed of the major ER membrane phospholipids, PC, PE, PI, and PS. This liposome system supported the interaction of PA phosphatase with the membrane and afforded the enzyme activity greater to that observed in the Triton X-100/PA-mixed micelle assay. Moreover, PA phosphatase catalyzed its reaction in the hopping and scooting modes, implicating how the enzyme operates along the nuclear/ER membrane in vivo.

Materials
Avanti Polar Lipids was the source of soybean PI, dioleoyl derivatives of PC, PE, PS, and PA, other fatty-acyl derivatives of PA, and the polycarbonate filters used to prepare liposomes. Coomassie Blue R-250, molecular mass protein standards, and reagents for electrophoresis, immunoblotting, and protein assay were purchased from Bio-Rad. Malachite green was purchased from Fisher Scientific. GE Healthcare was the supplier of polyvinylidene difluoride paper and the enhanced chemifluorescence Western blotting detection kit. Millipore Sigma was the source of ammonium molybdate, BSA, and Triton X-100. The alkaline phosphatase-conjugated goat anti-rabbit IgG antibody was a product of Thermo Scientific. All other chemicals were reagent grade or better.

Preparation of purified Pah1
Escherichia coli-expressed His 6 -tagged yeast Pah1 was purified from bacterial cell extracts by affinity chromatography with nickel-nitrilotriacetic acid-agarose as described by Han, Wu, and Carman (1). The protein content of enzyme preparations was estimated by the method of Bradford (97) using BSA as a standard. SDS-PAGE (98) analysis indicated that the Pah1 preparation was highly purified.

Preparation of liposomes
Liposomes (unilamellar phospholipid vesicles) were prepared by the extrusion method of MacDonald et al. (99) using an Avanti miniextruder. Unless otherwise indicated, the dioleoyl derivatives of PC, PE, PS, PA, and soybean PI were used in this work. Chloroform was evaporated from the phospholipid mixtures under a stream of nitrogen to form a thin film, and residual solvent was removed in vacuo. Phospholipids were then resuspended in 20 mM Tris-HCl (pH 7.5) to a final concentration of 20 mM. After five cycles of freezing and thawing, the phospholipid suspensions were repeatedly extruded through a polycarbonate filter to produce vesicles with diameters of 51.3 ± 2.5 nm, 101.3 ± 2.5 nm, or 299.3 ± 3.7 nm. A Brookhaven Instruments particle size analyzer was used to confirm the size of liposomes. Under most conditions used in this study, the liposomes were made of PC/PE/PI/PS/PA (33. 75 Preparation of Triton X-100/PA-mixed micelles PA in chloroform was transferred to a test tube and solvent was removed with a stream of nitrogen; residue solvent was removed in vacuo. Uniform Triton X-100/PA-mixed micelles (90:10 mol%) were prepared by adding Triton X-100 to the dried PA (93). The molar percent of PA in the Triton X-100/PA-mixed micelles was calculated using the following formula, mol% PA = [PA (molar)]/ [PA (molar) + Triton X-100 (molar)] × 100.

Pah1-liposome interaction assay
Pah1 was incubated for 15 min with the indicated liposomes in 20 mM Tris-HCl (pH 7.5) buffer containing 150 mM NaCl in a total volume of 30 l at 30°C. Following incubation, the reaction mixture was subjected to centrifugation at 100,000 g for 1 h at 4°C, and the liposome pellet was resuspended in the same volume as the supernatant. Samples (20 l) of each fraction were separated by SDS-PAGE (98), followed by immunoblotting (100)(101)(102) with polyvinylidene difluoride membrane using rabbit anti-Pah1 antibody (63). Anti-Pah1 antibody was used at a final concentration of 2 g/ml. The goat anti-rabbit IgG antibody conjugated with alkaline phosphatase was used at a dilution of 1:4,000. Immune complexes were detected using the enhanced chemifluorescence immunoblotting substrate. Fluorimaging, using a Storm 865 molecular imager (GE Healthcare) was used to acquire fluorescence signals from immunoblots, and the intensities of the images were analyzed by ImageQuant TL software (GE Healthcare). A standard curve was used to ensure that the immunoblot signals were in the linear range of detection.

PA phosphatase assay
PA phosphatase activity was measured at 30°C for 15 min by following the release of water-soluble P i from chloroform-soluble PA; the P i produced in the reaction was measured with malachite green-molybdate reagent (55,103). The reaction mixture contained 50 mM Tris-HCl (pH 7.5), 1 mM MgCl 2 , enzyme protein, and the PA-containing liposomes in a final volume of 10 l. The molar ratio of PC/PE/PI/PS to PA was 9:1 (10 mol% PA). Alternatively, the PA phosphatase activity was measured with the Triton X-100/PA-mixed micelle as the substrate; the molar ratio of Triton X-100 to PA was 9:1 (10 mol% PA) (1). Enzyme assays were conducted in triplicate, and the average SD of the assays was ±5%. All enzyme reactions were linear with time and protein concentration. A unit of enzymatic activity was defined as the amount of enzyme that catalyzed the formation of 1 nmol of product per minute.

Data analysis
The enzyme kinetics module of SigmaPlot software was used to analyze kinetic data according to Michaelis-Menten and Hill equations. Microsoft Excel software was used for the statistical analysis of the data. The P values <0.05 were taken as a significant difference.

Rationale
In establishing a liposome model for PA phosphatase assay, we took advantage of the fact that the requirement of the Nem1-Spo7 complex for the dephosphorylation and membrane interaction of the enzyme is circumvented by its phosphorylation-deficient form (32,63,69,70,104). Moreover, phosphorylation decreases the interaction of PA phosphatase with simple PC/PA liposomes (104). Accordingly, we utilized the E. coli-expressed form of Pah1 (1), which is free of phosphorylation that occurs when the enzyme is expressed in yeast (69). Liposomes were prepared from PC, PE, PI, PS, and PA to mimic the phospholipid composition of the nuclear/ER membrane found in yeast (105)(106)(107). The amount of PA (10 mol%) used in the liposomes is a saturating surface concentration as determined from this work (see below, Fig. 9). The average diameter of the liposomes used in this study was 100 nm unless otherwise indicated, which is a typical size for large unilamellar vesicles (99). The fidelity of the Pah1 was assessed by the measurement of its PA phosphatase activity using the established Triton X-100/PA-mixed micellar assay (108).

Effect of phospholipid composition on the interaction of Pah1 with liposomes
We examined an optimal amount of Pah1 for its liposomal interaction using a saturating amount of liposomes (see below, Fig. 5) containing PC/PE/PI/PS/PA (Fig. 2). The divalent cation Mg 2+ , which is required for PA phosphatase activity (1, 3) but has no effect on the interaction of the enzyme with PC/PA liposomes (104), was omitted in the binding assay to prevent the dephosphorylation of PA; the presence of PA has an effect on the liposome interaction (104) (see below, Fig. 3). Most Pah1 bound to the liposomes when 37.5 ng of protein was used for the assay. The increasing amounts of Pah1 led to the saturation of the liposomes with a concomitant increase in the amount of unbound protein. Owing that 90% Pah1 was associated with the liposomes using 75 ng of protein and this amount was within the linear range for measuring PA phosphatase activity (see below), 75-80 ng protein was used in subsequent binding and PA phosphatase activity measurements.
In the next set of experiments, we questioned what effect the phospholipid composition would have on the interaction of Pah1 with liposomes (Fig. 3). Fifty-seven percent Pah1 associated with liposomes composed of PC/ PE/PI/PS. The presence of 10 mol% PA in the liposomes showed a 1.6-fold increase in the enzyme association. This result is consistent with a previous finding with liposomes composed of PC and PC/PA (104). The effect of phospholipid composition on the interaction of Pah1 with the PAcontaining liposomes was examined further. The different types of liposomes contained the major phospholipid PC and the substrate PA, but differed in the level and composition of the phospholipids PE, PI, and PS. In general, the phospholipid composition of PA-containing liposomes did not have a major effect on the interaction of Pah1 with liposomes (Fig. 3).

Effect of phospholipid composition on the Pah1 PA phosphatase activity with liposomes
We confirmed the fidelity of the Pah1 enzyme preparation by the measurement of PA phosphatase activity using the established Triton X-100/PA-mixed micelle assay (108). To compare Triton X-100/PA-mixed micelles with the PA-containing liposomes in the enzyme assay, the level of PA was maintained at the saturating surface concentration of 10 mol% (93). As described previously (104), the PA phosphatase activity on PC/PA liposomes was 20% of the activity observed with Triton X-100/PA-mixed micelles (Fig. 4). However, when the liposomes contained the PC/PE/PI/PS/PA mixture resembling the phospholipid composition of the yeast ER membrane (105)(106)(107), the PA phosphatase activity was 1.4-fold higher than that measured with the Triton X-100/PA-mixed micelle (Fig. 4). To examine the dependence of PA phosphatase on liposome composition in more detail, the activity was measured with liposomes composed of the major phospholipids in different combinations (Fig. 5). For each combination, the PA surface concentration was maintained at 10 mol%. The PA phosphatase activity observed for each liposome composition was dependent on the PA molar concentration (e.g., liposome amount). Greatest activity was observed using liposomes composed of complex mixtures of the phospholipids. On the one hand, the V max with liposomes composed of PC/PE/PS/PA (2,300 nmol/min/mg) or PC/PE/PI/PA (2,500 nmol/min/mg) was 1. We considered that the fatty-acyl moiety of PA might affect the PA phosphatase activity on the liposomes composed of PC/PE/PI/PS/PA (Fig. 6). In wild-type yeast, about 50% of PA contains 16:0 at the sn-1 position and 18:1 at the sn-2 position (39). The activity observed with this fattyacyl composition was not majorly different from that observed with the dioleoyl derivative of PA routinely used in this study. Additionally, the variety of fatty-acyl compositions of   4. Pah1 PA phosphatase activity using Triton X-100/PAmixed micelles and phospholipid/PA liposomes. Pah1 (80 ng) was assayed (15 min) for PA phosphatase activity by following the release of P i from PA using saturating amounts of Triton X-100/PAmixed micelles (93) or phospholipid/PA liposomes. The surface concentration of PA within the detergent micelles or liposomes was 10 mol%. The activity using the Triton X-100/PA-mixed micelles (900 nmol/min/mg) was arbitrarily set at 100%. The data shown are the averages of the three experiments ± SD (error bars). The individual data points are also shown. *P < 0.05 versus the PA phosphatase activity with Triton X-100/PA-mixed micelles. the PA had little effect on the PA phosphatase activity using the liposomes made of PC/PE/PI/PS/PA (Fig. 6).
The liposomes used in this work were 100 nm in diameter, a size widely used for large unilamellar vesicles (99). We considered whether smaller (50 nm) or larger (300 nm) size liposomes would impact on the PA phosphatase activity (Fig. 7). At a saturating amount of liposomes at 1 mM PA, the activity observed with the 100 nm liposomes was 2-fold higher when compared with the activity of the 50 nm liposomes. However, the increase of liposome size to 300 nm had no effect on increasing the enzyme activity observed with the 100 nm liposomes.
Having characterized the liposome system in terms of phospholipid composition, we set forth to demonstrate that the PA phosphatase reaction is linear with time and the amount of Pah1 (Fig. 8). That the activity was linear with time and protein indicates that the reaction follows zero order kinetics with respect to PA within the liposome. Based on the results of these experiments, the routine assays for measuring PA phosphatase activity with the liposomes composed of PC/PE/PI/PS/PA were conducted for 15 min with 80 ng enzyme.

Dependence of Pah1 PA phosphatase activity on the surface concentration of PA in liposomes
Peripheral membrane enzymes such as Pah1 PA phosphatase (1,93) act in a cellular environment in which both three-dimensional bulk interactions (e.g., hopping mode)

1238
Journal of Lipid Research Volume 61, 2020 occur at an aqueous-membrane interface and two-dimensional surface interactions (e.g., scooting mode) occur at the membrane bilayer (110,111) (see below, Fig. 10). The data shown in Fig. 5 demonstrated that PA phosphatase activity was dependent on the molar concentration of PA (liposome amount), and thus functions in the hopping mode using the liposome system. We next set forth to demonstrate that the activity is dependent on the surface concentration of PA, and thus, functions in the scooting mode. For the experiment shown in Fig. 9, eight types of liposomes were prepared; in each liposome type, the molar ratio of PC/PE/PI/PS was maintained at 33.75:22.5:22.5:11.25 mol% to mimic the yeast nuclear/ER membrane, and the molar ratio of the combined phospholipid mixture to PA was varied to achieve the indicated surface concentrations of PA. Under the conditions of this experiment, the PA phosphatase activity was measured with a saturating amount of liposomes (1 mM PA). The kinetic data were analyzed according to the Michaelis-Menten and Hill equations using the Enzyme Kinetics Module of SigmaPlot. The data more closely followed positive cooperative kinetics (n = 2.1) as defined by the Hill equation (R 2 = 0.95) when compared with saturation kinetics as defined by the Michaelis-Menten equation (R 2 = 0.89). Maximum activity was observed at 4.9 mol%, and an analysis of the data according to the Hill equation yielded V max and K m values of 1,100 nmol/min/mg and 0.65 mol%, respectively. Overall, these data demonstrated that the enzyme operates in the scooting mode.

DISCUSSION
The PAH1-encoded Mg 2+ -dependent PA phosphatase has emerged as a key enzyme whose activity controls the metabolism of lipids emanating from PA ( Fig. 1) (5, 6, 11). The classic approach of measuring yeast PA phosphatase activity involves the incorporation of the substrate PA into a uniform Triton X-100/PA-mixed micelle (3,93,108). Although this detergent/phospholipid mixed micelle system has permitted defined activity measurements and studies to assess biochemical mechanisms of regulation (1, 64-66, 68, 93, 95, 96, 112), it lacks the mimicry of a biological membrane phospholipid bilayer. Here, we developed a liposome system, a widely accepted mimic of a biological membrane (99,113,114), to measure the activity of Pah1 PA phosphatase. To approximate the nuclear/ER membrane where the PA phosphatase catalyzes its reaction, the substrate PA was incorporated into liposomes comprised of PC/PE/PI/PS. These vesicles afforded a level of PA phosphatase activity greater to that obtained with the Triton X-100/PA-mixed micellar system (Fig. 4). As described previously (104), a simpler liposome composition (e.g., PC/ PA) supported PA phosphatase activity, but at a much reduced level when compared with that observed with Triton X-100/PA-mixed micelles or with PA incorporated into the complex mixture of phospholipids. The reduced activity observed with PC/PA liposomes relative to that with Triton X-100/PA-mixed micelles has been attributed to the inability of the enzyme to access substrate at the inner leaflet of the vesicle membrane (104). Whereas this explanation is reasonable, the work presented here demonstrates that PA phosphatase activity is dependent on the phospholipid composition of the liposomes; a complex composition approximating that found in vivo yielded a robust level of activity (Fig. 5). The complex mixture of PC/PE/PI/PS/ CDP-DAG, approximating the nuclear/ER membrane, also supports robust activity of the yeast phospholipid synthesis enzymes PI synthase (115) and PS synthase (116).
The dependence of PA phosphatase activity on the phospholipid composition of liposomes was not governed by differences in the interaction of the enzyme with the vesicles; enzyme interaction was largely unaffected by the type and complexity of the phospholipids in PA-containing liposomes (Fig. 3). Whereas PA was not required for enzyme interaction with the liposomes composed of PC/PE/PI/ PS, its presence afforded 60% greater interaction (Fig. 3). A similar situation has previously been shown with simple PC/PA liposomes (104) and with Triton X-100/PA-mixed micelles (93). While we did not examine the effect of PA fatty-acyl composition on enzyme-liposome interaction, the acyl composition of PA did not majorly affect the activity of the enzyme in liposomes composed of PC/PE/PI/PS/PA (Fig. 6). Whether the fatty-acyl content of PC, PE, PI, or PS would have an effect on PA phosphatase activity or enzyme-liposome interaction is unknown. The system developed here will permit additional studies to examine these questions, as well as whether other membrane lipids, such as CDP-DAG (96), and sphingoid bases (94), which have been shown to modulate PA phosphatase activity in Triton X-100/PA-mixed micelles, regulate the activity in liposomes. In our studies, we also observed that PA phosphatase activity depends on liposome size; greater activity was observed with large liposomes (e.g., 100 and 300 nm) when compared with small liposomes (50 nm). This raised the suggestion that the PA phosphatase activity is sensitive to membrane phospholipid curvature (Fig. 7) (117).
The mammalian lipin 1, 2, and 3 isoforms of PA phosphatase have also been measured with PA incorporated into liposomes (54,(118)(119)(120). For these enzymes, the activity is augmented by the di-anionic form of PA, which is favored by a much elevated PE content (e.g., 60 mol%) in PC/PA liposomes (119,120). Recognition of di-anionic PA is governed by the polybasic domains in the lipin proteins; phosphorylation prevents PA recognition for lipins 1 and 2, but has no effect for lipin 3 (119,120). The yeast PA phosphatase does not possess a polybasic domain, and its activity is not regulated by PE (at least in detergent micelles (96)) or the ionic nature of PA (G-S. Han and G. M. Carman, unpublished observations). Nonetheless, the liposomes used to measure the lipin PA phosphatase enzymes do not approximate the membrane phospholipid composition in mammalian cells.
Whereas the impetus for this work was to develop an assay for the measurement of PA phosphatase activity in an environment that mimics the phospholipid composition of the nuclear/ER membrane, it also provided information on how the enzyme operates at the membrane. That PA phosphatase activity was dependent on the bulk concentration of PA in the PC/PE/PI/PS/PA liposomes (Fig. 5) indicated that the enzyme operates in the hopping mode (Fig. 10) (110). The kinetic analysis with the PC/PE/PI/PS/PA liposomes demonstrated that PA phosphatase activity was dependent on the surface concentration of PA (Fig. 9), and thus the enzyme also operates in the scooting mode ( Fig. 10) (110). The surface K m value for PA of 0.65 mol% is within, if not below, the physiological concentration (105,106). Thus, small changes in the surface concentration of PA would have a major effect on the PA phosphatase activity at the nuclear/ER membrane. The enzyme activity exhibited cooperative kinetics with respect to the surface concentration of PA. Whether the Hill number of 2 reflects cooperative binding of two enzyme molecules or two substrate molecules is unclear and warrants additional studies. Owing that Pah1 appears to be monomeric (3), we favor a model by which the binding to one PA molecule facilitates interaction with a second substrate molecule. Precedence for such a mechanism (e.g., dual phospholipid model) comes from studies on the action of phospholipase A 2 toward phospholipid substrates (82,84,85,111).
According to the model shown in Fig. 10, following its dephosphorylation by the Nem1-Spo7 protein phosphatase complex, the Pah1 PA phosphatase enzyme hops onto the membrane surface. This process, which is dependent on the amphipathic helix found at the N terminus (70), is facilitated by the presence of PA in the nuclear/ER membrane (70). After membrane interaction and binding to PA, the enzyme catalyzes its reaction to produce DAG. Following the reaction, the enzyme remains on the membrane surface and scoots along to bind another molecule of PA and carry out another round of catalysis. One would expect that it is more efficient for PA phosphatase to operate in the scooting mode as opposed to hopping from membrane surface to membrane surface to bind to its substrate PA (110). Yet, PA phosphatase operating in the hopping mode might be physiologically important in the context that the enzyme is reported to associate with membranes other than the nuclear/ER membrane. For example, Pah1 is also found at the nuclear vacuolar junction (121), the inner nuclear membrane (122), and in the vicinity of lipid droplets (123,124). The phosphorylated form of Pah1 must be recruited to the nuclear/ER-associated Nem1-Spo7 complex for its dephosphorylation before membrane association and PA binding can occur (70). The ability of the unphosphorylated enzyme to operate in the hopping mode would permit its association with multiple membrane locations, especially if spatially close through organelle contact sites. In fact, a change in cell physiology (e.g., accelerated lipid droplet formation) might be a trigger for the enzyme to switch to its hopping mode of action. Testing this notion would benefit from the identification and mutation of specific residues required for the hopping and scooting modes of the enzyme.  10. Model for the action of Pah1 PA phosphatase (PAP) at the nuclear/ER membrane via the hopping and scooting modes. The phosphorylated form of PA phosphatase (small white circles) is dephosphorylated by the Nem1-Spo7 complex at the nuclear/ER membrane. The dephosphorylated PA phosphatase then hops onto the membrane surface. It then scoots along the membrane, binds to its substrate PA, and catalyzes the dephosphorylation of PA to produce DAG. Following the reaction, PA phosphatase scoots along the membrane until it binds another PA molecule. It then catalyzes the dephosphorylation of this PA to produce DAG. The PA phosphatase molecule catalyzing the dephosphorylation of PA is denoted with blue color.

1240
Journal of Lipid Research Volume 61, 2020 Pah1 PA phosphatase is known to associate with the membrane for its cellular function through the control of the nuclear/ER membrane-associated Nem1-Spo7 phosphatase (41); its dephosphorylation by the phosphatase complex is essential for its membrane localization and catalytic function (32,69,123). In future studies, we will reconstitute the protein phosphatase complex into liposomes to examine the regulation of recruitment and dephosphorylation of Pah1 followed by the enzymatic dephosphorylation of PA in this defined model system.

Data availability
All data are contained within the article.
The authors thank Gil-Soo Han for providing the Pah1 used in the enzyme-liposome interactions studies, helpful discussions during the course of this work, and for comments in the preparation of the manuscript. The authors also acknowledge Edward A. Dennis for comments during the initial stages of the work.