A fluorescence and HPLC-based activity assay system for lipid modifying enzymes: Lipin/Pah phosphatidic acid phosphatase as an optimized example

Abstract

Quantitating the enzymatic activity of lipid modifying enzymes is a critical aspect of their study. Here, we describe a general approach to measure the activity of lipid modifying enzymes using fluorescently labeled lipid substrates. This approach uses high-performance liquid chromatography (HPLC) to separate the resulting fluorescent lipid product(s) of enzymatic catalysis from the fluorescent lipid substrates, and to concurrently quantify the products using a fluorescence detector. Here, we focus on a version of this assay that has been optimized for lipin/Pah phosphatidic acid phosphatases, a class of enzymes that catalyze the magnesium-dependent hydrolysis of phosphatidic acid into diacylglycerol. Details for delivery of a fluorescent lipid substrate to an enzyme in detergent mixed-micelles and liposomes are provided. We also describe methods to purify recombinant lipin/Pah phosphatidic acid phosphatases using E. coli as an overexpression system, which can present challenges given the role of phosphatidic acid as a central precursor in bacterial phospholipid synthesis. Overall, the fluorescence and HPLC-based activity assay described here is generally applicable to any lipid modifying enzyme and additionally allows for the detection of unsuspected products that would remain undetected by conventional methods.

1. INTRODUCTION

Lipins/Pahs are magnesium-dependent phosphatidic acid phosphatases (PAPs) that catalyze the hydrolysis of phosphatidic acid (PA) to generate diacylglycerol (DAG) and inorganic phosphate(Han et al., 2006). Lipins catalyze the penultimate step of triglyceride synthesis(Takeuchi & Reue, 2009), thus playing a major role in de novo phospholipid biosynthesis(Khalil et al., 2010; Reue & Wang, 2019), lipid metabolism(Coleman & Mashek, 2011; Siniossoglou, 2013), and energy storage(Reue & Donkor, 2006) (Fig. 1). Lipins are evolutionarily conserved from mammals(Péterfy et al., 2001; Reue & Wang, 2019) to plants(Chapman et al., 2019; Eastmond et al., 2010; Nakamura et al., 2009; Xu & Shanklin, 2016) and yeast(Pascual & Carman, 2013; Pillai et al., 2017), with the number of orthologs varying(Harris & Finck, 2011).

Figure 1: The glycerol-3-phosphate pathway.

Lipin catalyzes a key branch point in the glycerol-3-phosphate pathway. Lipin is the only peripheral membrane protein in this pathway. The other members, glycerol phosphate acyltransferase (GPAT), acylglycerol phosphate acyltransferase (AGPAT), and diacylglycerol acyltransferase (DGAT) are integral membrane proteins and use one molecule of fatty acyl coenzyme A (FA-CoA) to generate their product. Lipin is a magnesium-dependent enzyme that generates diacylglycerol (DAG) from the hydrolysis of phosphatidic acid (PA). DAG is a starting material for the de novo biosynthesis of major membrane phospholipids phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylserine (PS). PA is a precursor for other lipids such as phosphatidylinositol (PI) and cardiolipin (CL).

One of the first enzymatic assays to be developed that allowed the characterization of PAP activity involved monitoring the release of ³²P-labeled inorganic phosphate from phosphatidic acid(Carman & Lin, 1991; Martin et al., 1991; Walsh & Bell, 1986). This assay represented a critical breakthrough in the field and allowed several seminal discoveries and studies to be conducted(Donkor et al., 2007; Donkor et al., 2009; Eaton et al., 2013; Eaton et al., 2014; Granade & Harris, 2018; Han & Carman, 2010; Han et al., 2007; Han et al., 2006; Schweitzer et al., 2019; Stukey et al., 2023). This radioactive assay remains a common method to assess lipin/Pah PAP activity. In these assays, the PA substrate for PAP assays is prepared by reacting γ³²P-labeled adenosine triphosphate and unlabeled DAG using bacterial DAG kinase to generate ³²P-labelled PA that can be isolated using preparative TLC(Walsh & Bell, 1986). This ³²P-PA can be used as a substrate for lipin, which generates ³²P-inorganic phosphate that can be separated from the ³²P-labeled PA substrate and newly generated DAG using an organic extraction. The radioactive ³²P-inorganic phosphate generated from the reaction is present in the aqueous phase, while unreacted ³²P-labelled PA and the DAG generated remain in the organic phase. The radioactive ³²P-inorganic phosphate can then be quantified. This approach has many advantages including using a natural PA substrate that only differs from native PA by the isotope of phosphorous, but has the disadvantage that the generation of ³²P-labeled PA requires certain expertise and equipment and must be used within the relatively short half-life limits of ³²P(Dey et al., 2020).

Alternative approaches have more recently been described. One report used NBD-PA (1-hexanoyl-2-[6-[(7-nitro-2–1,3-benzoxadiazol-4-yl)amino]hexanoyl]-sn-glycero-3-phosphate) as a substrate and separated the enzymatic product, NBD-DAG, using TLC(Sasser et al., 2012). Another group utilized HPLC to quantify lipin activity from mammalian tissues using solubilized fluorescently labelled PA as a substrate(Burgdorf et al., 2009; Burgdorf et al., 2008). Our group later modified these approaches to use HPLC with a fluorescence detector to both separate NBD-PA from the NBD-DAG generated and quantitate lipin/Pah PAP activity from purified recombinant protein.

This approach can be applied to homogenized tissues(Burgdorf et al., 2009; Burgdorf et al., 2008; Csaki et al., 2014; Nadra et al., 2012), cell lysates(Han et al., 2006), and purified recombinant protein(Gao et al., 2024; Khayyo et al., 2020; Vitkovska et al., 2025). However, production of recombinant protein in E. coli requires careful preparation. This is because PA is the precursor for all phospholipid synthesis in E. coli. Thus, overexpression of active lipin/Pah PAPs in our hands has often resulted in several observations including difficultly in pelleting of the E. coli cells upon centrifugation due to accumulated DAG in cells. These artifacts can be avoided by expressing inactive lipin/Pah PAPs, which suggests overexpression of lipin/Pah PAPs can be somewhat toxic. Thus, as part of this chapter, we first describe approaches to produce active recombinant lipin/Pah PAP proteins using E. coli as an overexpression system. We then describe the incorporation of NBD-PA into Triton X-100 mixed micelles and liposomes as a method to deliver the substrate to the enzyme. Finally, we described the activity assay used in our lab which is broadly applicable to many other lipid modifying enzymes (Fig. 2) and was recently used by our lab to detect unanticipated products generated by the lipid modifying enzyme DDHD2, which possesses dual enzymatic functionality(Wu et al., 2025).

Figure 2: Fluorescence based phosphatidic acid phosphatase activity assay.

Either mixed micelles (top) or liposomes (bottom) are combined with protein in 1:1 ratio, then incubated at 30 °C for one hour. The reaction is quenched with the addition of chloroform (CHCl₃) and methanol (MeOH) in a 1:1 ratio, in three times excess of the reaction volume. The lower organic phase is collected and transferred to a new tube, where it is dried under nitrogen gas. The lipids are then resuspended in Mobile phase B, then spun at max speed. This mixture is then applied to HPLC to quantify and separate peaks corresponding to NBD-DAG and NBD-PA.

2. PURIFICATION

The protein preparation method described here has been successful for minimal lipin constructs that do not possess the long linker between the N-Lip and C-Lip regions, such as Tt Pah2(Khayyo et al., 2020), mouse lipin-2(Khayyo et al., 2020), and other minimal lipin/Pah PAPs from yeast and mammals (unpublished data). This method has also shown some success with full length lipins from yeast. Typically, these constructs are expressed in the ppSUMO vector with a SUMO tag, in addition to 6X-His tag, to increase protein solubility. This SUMO tag is cleaved with ULP1 protease after the nickel purification step. Expression of these active proteins in E. coli often results in the accumulation of DAG within cells, making the cells harder to pellet due to their accumulation of neutral lipids, even with prolonged and high force centrifugation. In fact, when overexpressing active lipin/Pah PAPs, it is normal for the pellets to have a slimy consistency. In this case, it is sufficient to pour off as much media as possible after centrifugation and flash freeze the cell pellet.

The purification buffers described below have been optimized by a thermal shift assay that demonstrated phosphate buffer at pH 7, specifically containing monobasic potassium phosphate and dibasic sodium phosphate, best stabilized a series of lipin/Pah PAP proteins. Additionally, the addition of the detergent Triton X-100 in the lysis buffer aids in solubilizing the protein from the membrane, but is no longer necessary for the soluble protein after lysis, and can be excluded from any subsequent buffers involved in washing and eluting the recombinant proteins from Ni-NTA (Nickel nitrilotriacetic acid) affinity columns.

2.1. Equipment:

• Incubator

• Incubated Shaker

• 2.5L Ultra Yield Flasks (Thompson Instrument Company #931136-B)

• Refrigerated Centrifuge

• Sonicator

• Fast Protein Liquid Chromatography (FPLC) Instrument

• HiLoad 26/600 Superdex 75 pg Gel Filtration Column (GE Healthcare)

• Ni-NTA Resin

• Gravity Filter Column

• Spectrophotometer

• Protein Concentration Spin Columns

2.2. Buffers and Reagents

• Luria Broth (LB) Media

• E. coli BL21 (DE3) RIPL cells

• Glycerol

• Terrific Broth (TB) Media

• Kanamycin sulfate

• Isopropyl β-D-1-thiogalactopyranoside (IPTG)

• β-Mercaptoethanol (βME)

• Phenylmethanesulfonyl fluoride (PMSF)

• SDS-PAGE tank setup

• Ubiquitin-like protease 1 (ULP1)

• Dithiothreitol (DTT)

• Precision Plus Protein Dual Color Standards (Bio-Rad Cat#: 1610374)

• Lysis Buffer: [100 mM KH₂PO₄ + Na₂HPO₄ buffer pH 7, 500 mM NaCl, 60 mM imidazole, 5% (v/v) glycerol, 1% Triton X-100, 2 mM βME, 0.5 mM PMSF]

• Wash Buffer: [100 mM KH₂PO₄ + Na₂HPO₄ buffer pH 7, 500 mM NaCl, 60 mM imidazole, 5% (v/v) glycerol, 2 mM βME, 0.5 mM PMSF]

• Elution Buffer: [100 mM KH₂PO₄ + Na₂HPO₄ buffer pH 7, 500 mM NaCl, 300 mM imidazole, 5% (v/v) glycerol, 2 mM βME, 0.5 mM PMSF]

• Gel Filtration Buffer: [20 mM Tris pH 8, 150 mM NaCl, 10 mM βME, 1 mM DTT]

2.3. Transformation

1. Thaw 50 μL of E. coli BL21 (DE3) RIPL cells on ice and add to 14 mL round bottom tube on ice.

2. Add 100 ng of lipin/Pah plasmid to round bottom tube with competent cells.

3. Incubate tube on ice for 15 minutes.

4. Heat shock tube at 42 °C for 45 seconds.

5. Return tube to ice for 2 minutes.

6. Add 200 μL of LB media to tube and transfer tube to incubator to shake at 37 °C for one hour at 250rpm.

7. Plate 150 μL of transformation reaction on LB agar plate with appropriate antibiotic.

8. Allow plate to dry then incubate overnight at 37 °C to allow for colony growth.

2.4. Expression

Day 1 – Starter Culture

1. Pick a single colony and use to inoculate 20 mL of LB supplemented with appropriate antibiotic in a 250 mL Erlenmeyer flask.

2. Grow this starter culture overnight with shaking at 37 °C and 250 rpm.

3. Prepare 1L of TB media supplemented with 4 mL glycerol in 2.5L Ultra Yield flask, seal top of flask with foil, and autoclave. Allow to cool overnight.

Day 2 – Expression

1. Pre-warm TB by placing in incubator shaker at 37 °C for 30 minutes.

2. Add appropriate antibiotic, then inoculate each liter with 15 mL of the overnight starter culture.

3. Shake at 250 rpm until an OD₆₀₀ of ~1.2 is reached.

4. Cool cultures at 10 °C with shaking for two hours.

5. Induce cultures with the addition of a final concentration of 30 μg/mL IPTG.

   Note: Increased concentrations of IPTG can be used to increase protein expression.

6. Increase incubator shaker temperature to 15 °C and allow expression for ~18 hours.

Day 3 – Harvesting Culture

1. Transfer cultures to centrifuge tubes and spin down in pre-cooled centrifuge for 15 minutes at 4,000 rpm (3,320 x g) and 4 °C to pellet cells.

2. Pour off media, then transfer cell pellet to 50 mL falcon tube. Store cell pellets at −80 °C.

   Note: Very active proteins will not form firm cell pellet due to accumulation of diacylglycerol. In this case, pour off as much media as possible and flash freeze ‘slimy’ pellet.

2.5. Purification

Nickel Affinity Chromatography

1. Remove cell pellet from −80 °C and allow to thaw at room temperature for one hour.

2. Resuspend the pellet in the least amount of lysis buffer possible to obtain a homogenous mixture. To mix, invert tube and vortex, ensuring as little bubbles as possible. Once resuspended, decant mixture into ice cold metal beaker.

   Note: Add PMSF and βME to lysis buffer directly before using. PMSF has a short half-life in water.

3. Sonicate this mixture at 85% power for 5 minutes total, in 1 minute intervals with bursts of 2 seconds on/ 2 seconds off on ice. Make sure mixture stays cold. Replace ice under metal beaker as necessary. Take an SDS-PAGE sample after sonication as ‘total lysate’.

   Note: At this step it is crucial to avoid foaming ang heating of the cell lysate, as this can denature proteins. Lysing of cells can be monitored by visual cues, such as the clarification of cell lysate during sonication, although this does not always occur.

4. Transfer cell lysate into precooled centrifuge tubes. Spin down at 26,000 rpm (81,770 x g) at 4 °C for 30 minutes to pellet cell debris. During this time, equilibrate 4 mL Ni-NTA resin with wash buffer.

   Note: The volume of Ni-NTA resin described here is a starting point and can be adjusted in subsequent purifications by monitoring the amount of protein lost during the wash step.

5. Incubate soluble fraction with equilibrated Ni-NTA beads and gently agitate this mixture at 4 °C for one hour to bind protein with beads. Take an SDS-PAGE sample at this time as ‘soluble fraction’.

6. Pour this mixture into gravity column with filter. Take first few drops from the gravity column as SDS-PAGE sample as ‘flow-through’ fraction.

7. Wash the nickel column with 5-15 column volumes of wash buffer. Take first few drops from the gravity column as SDS-PAGE sample as ‘wash’ fraction.

   Note: The wash volume described here is just a starting point and can be adjusted as necessary in subsequent purifications by monitoring the purity of the protein eluted from the nickel column. If there is a lot of contamination from proteins that are nonspecifically bound to the column, increase the wash volume. Add βME to buffers right before using.

8. Elute protein in 10 x 1 mL fractions with elution buffer. Keep fractions on ice and take a gel sample of each.

   Note: Gravity column can be kept at room temperature as long as buffers are kept ice cold.

9. Run each SDS-PAGE sample on a gel and visualize protein with Coomassie staining. Pool fractions containing protein for size exclusion chromatography.

10. Add ULP-1 protease to cleave SUMO tag (if needed). Store at 4 °C overnight.

Gel Filtration

1. Filter gel filtration buffer with 0.22 μm filter. Add βME and DTT directly before use and keep buffer cold.

2. Prepare Superdex 75 column by washing with 1 column volume of gel filtration buffer on FPLC.

3. Load pooled nickel elution fractions onto the column, follow with isocratic elution using gel filtration buffer. Collect fractions after the void volume.

4. Concentrate protein fractions using spin filters and refrigerated centrifuge. Use spectrophotometer to determine protein absorbance at 280nm and calculate protein concentration. Aliquot protein and flash freeze in PCR tubes in liquid nitrogen. Store at −80 °C.

3. ACTIVITY ASSAY

When generating Triton X-100 mixed micelles and liposomes with fluorescently labeled lipids, it is essential to keep them protected from light as much as possible, typically by using amber vials for storage and covering preparatory glass tubes with foil. Incorporating PA into liposomes and mixed micelles can require more time than other lipids, due to the small, negatively charged headgroup of PA. During the set up of a series of enzymatic reactions, it is imperative to keep all samples ice cold until they are placed in a water bath to prevent starting the reaction prematurely.

It is absolutely critical to ensure the reaction rate is linear with respect to both time and amount of enzyme added to the reaction. Typically, an ideal amount of enzyme for the linear range will turn over about 10-20% of substrate, but this may vary.

Quenching of the reaction involves the purification of lipids using the addition of a 1:1 ratio of chloroform and methanol, first described by Folch(Folch et al., 1957) and Bligh and Dyer(Bligh & Dyer, 1959). In this assay, the organic phase is collected since we monitor the amount of NBD-DAG generated. The aqueous phase contains the inorganic phosphate released as well as protein. Due to this, it is essential to collect as much of the organic layer as possible without collecting the protein at the interface between the organic and aqueous phases because large amounts of protein can clog the HPLC column. As an additional method to prevent this, there is also a high-speed spin at the final step of this assay, that should pellet any remaining denatured protein. When removing the sample to be directly applied to HPLC, one should avoid disturbing any potential pellet at the bottom of the tube.

Extensive controls have been done to ensure the method detailed below generates unilamellar vesicles. Dynamic light scattering was performed to confirm the size of the vesicles(Choi et al., 2023), as well as other characterization methods like dithionite quenching to determine unilamellarity.

3.1. Equipment

• High-Performance Liquid Chromatography (HPLC) instrument (Aligent)

• Spectra C8SR column (Peeke scientific, 3 μm particle, 3.0 × 150 mm)

• Tip sonicator

• Centrifuge

• Heated water bath

• Vortexer

• Water bath sonicator

• Nitrogen gas blowdown system

• Hamilton syringe

• Amber HPLC vials

3.2. Buffers and Reagents

• Triton X-100 detergent

• 14:12 NBD-PA (1-myristoyl-2-[12-[(7-nitro-2-1,3-benzoxadiazol-4-yl)amino]dodecanoyl]-sn-glycero-3-phosphate, Avanti)

• NBD-DAG (1-NBD-1,2-bis[O-decanoyl-sn-glycerol], Cayman)

• 2X Micelle Assay Buffer: [40 mM Tris pH 7.5, 20 mM MgCl₂, 300 mM NaCl, 4 mM βME]

• Liposome Buffer: [50 mM HEPES pH 8, 50 mM NaCl]

• HPLC grade Chloroform

• HPLC grade Methanol

• Mass spectrometry grade ammonium formate

• LC/MS grade Formic acid

• HPLC Mobile Phase A: [1 mM ammonium formate, 0.2% formic acid in HPLC grade H₂O]

• HPLC Mobile Phase B: [1 mM ammonium formate, 0.2% formic acid in HPLC grade MeOH]

3.3. Micelle Preparation

1. Dry the appropriate amount of NBD-PA under nitrogen gas to have the bulk PA concentration at 200 μM when resuspended in the volume of Triton X-100 that will allow 50 μL of micelles per reaction.

   Note: Using 100 μM NBD-PA = 3 mol% gives a good signal without maxing out the detector on HPLC.

2. Resuspend dried NBD-PA in the appropriate concentration of Triton X-100 to generate micelles with desired mol% of NBD PA in your final volume. Pipet up and down to mix.

3. Solubilize NBD-PA by placing tube in water bath sonicator in 15 minute bursts, pipetting up and down between, for up to an hour.

   Note: NBD-PA is properly solubilized when the solution is homogeneous and clear. Keep micelles protected from light.

3.4. Liposome Preparation

1. Dry appropriate amounts of DOPC, DOPE, and NBD-PA in 80:20:2 ratio under nitrogen gas to give final concentrations of 1 mM DOPC, 0.25 mM DOPE and 25 μM NBD-PA when resuspended.

2. Resuspend lipids in the appropriate volume of liposome buffer and pipet up and down.

3. To generate large unilamellar vesicles, freeze liposome solution in liquid nitrogen, then thaw in water. Repeat 6 times for a total of 7 freeze/thaw cycles.

4. Sonicate in water bath sonicator for five minutes. Keep protected from light.

3.5. Assay Protocol

1. Dilute protein to desired concentrations in 2X activity assay buffer with βME added fresh prior to use. Keep on ice.

2. Mix protein and micelles/liposomes in 1:1 ratio, typically 50 μL of each is suitable. Keep tubes protected from light.

3. Incubate at 30 °C in water bath for one hour.

   Note: This temperature is just a suggestion and can be varied. However, 37 °C can denature PAP enzymes.

4. Quench reactions with the addition of 1:1 HPLC grade chloroform:methanol in 3 times excess of reaction volume.

5. Vortex each reaction for 10 seconds to mix. Then spin down for 2 minutes at 2000 rpm (771 x g) at room temperature to separate aqueous and organic phases.

6. Use Hamilton syringe to transfer organic phase (bottom) to fresh tube without collecting any of the aqueous phase. Dry organic phase under nitrogen gas.

   Note: Aqueous phase contains denatured protein and collecting it and applying to HPLC can clog column.

7. Resuspend organic phase in 100 μL of Mobile Phase B. Spin at max speed in tabletop centrifuge for five minutes.

8. Transfer 60 μL of reaction to amber HPLC vial without collecting from or disturbing the bottom of the tube.

9. Inject 10 μL of reaction onto HPLC column and run the following method with these parameters:

   Parameters and acquisition settings:

   Flow rate: 0.5mL/min

   Method Length: 16 minutes

   PMT Gain: 10

   Excitation wavelength: 470 nm

   Emission wavelength: 530 nm

Method Settings

Time (min)	% Mobile A	% Mobile B
1-8	20	80
8-12.5	2	98
12.5-16	20	80

4. QUANTIFICATION OF DAG PRODUCTION

4.1. Diacylglycerol Standard Curve

To best quantify the amount of DAG produced from reactions, a standard curve correlating the area of peaks generated by known quantities of NBD-DAG should be used (Fig. 3). A standard curve can be generated by drying down NBD-DAG under nitrogen gas, then resuspending in Mobile Phase B at varying concentrations and applying samples to the HPLC. The HPLC automatically integrates the area under each curve, then the area under each curve can be plotted against the pmol of DAG in each sample and fit to a simple linear regression to determine the equation for the line. The equation for this line can be used to quantitate how much NBD-DAG is produced by this assay.

Figure 3: Graph of NBD-DAG standards.

This standard curve was generated by drying down NBD-DAG under nitrogen gas, then resuspending in Mobile Phase B at varying concentrations and applying samples to HPLC. The area under each curve was plotted against the pmol of DAG in each sample and fit to a simple linear regression to determine the equation for the line.

4.2. Sample data and Quantification

When the samples from this activity assay are run on the HPLC, there should be two distinct peaks for product and substrate (Fig. 4). Running a blank sample of the substrate only with no protein added (Fig. 4A) is necessary to be able to distinguish the product peak (NBD-DAG) from the substrate peak (NBD-PA) (Fig. 4B). The HPLC mobile phase A and mobile phase B buffers described here are optimized for our specific HPLC column with hydrophobic resin (Spectra C8SR column, Peeke scientific). On our column, NBD-DAG has a shorter retention time, eluting first. The area corresponding to the peak for NBD-DAG can be correlated to the pmol of DAG produced using the equation derived from a standard curve. This pmol value is then multiplied by factor of 10 to quantitate the pmol of DAG in the entire 100 μL sample from the activity assay, since only 10 μL of this sample was injected into the HPLC. In cases where multiple, unsuspected peaks are present, these samples may be suitable for mass spectrometry to determine the identity of the unknown peak.

Figure 4: HPLC chromatograms of NBD-PA and NBD-DAG.

A: Chromatogram from blank sample containing 3 mol% NBD-PA Triton X-100 mixed micelles without protein added. Single peak with right shoulder has retention time of 11.34 minutes. Shoulder peak is present because the commercially available substrate contains slight impurity. B: Chromatogram from 3 mol% NBD-PA Triton X-100 mixed micelle reacted with Tt Pah2. Peak 1 corresponds to NBD-DAG with a retention time of 9.91 minutes and peak 2 corresponds to NBD-PA with a retention time of 11.24 minutes. The described method gives good separation of substrate and product.