The role of lysosomal phospholipase A2 in the catabolism of bis(monoacylglycerol)phosphate and association with phospholipidosis

Abstract

Bis(monoacylglycerol)phosphate (BMP) is an acidic glycerophospholipid localized to late endosomes and lysosomes. However, the metabolism of BMP is poorly understood. Because many drugs that cause phospholipidosis inhibit lysosomal phospholipase A2 (LPLA2, PLA2G15, LYPLA3) activity, we investigated whether this enzyme has a role in BMPcatabolism. The incubation of recombinant human LPLA2 (hLPLA2) and liposomes containing the naturally occurring BMP (sn-(2-oleoyl-3-hydroxy)-glycerol-1-phospho-sn-1'-(2'-oleoyl-3'-hydroxy)-glycerol (S,S-(2,2',C_18:1)-BMP) resulted in the deacylation of this BMP isomer. The deacylation rate was 70 times lower than that of dioleoyl phosphatidylglycerol (DOPG), an isomer and precursor of BMP. The release rates of oleic acid from DOPG and four BMP stereoisomers by LPLA2 differed. The rank order of the rates of hydrolysis were DOPG>S,S-(3,3',C_18:1)-BMP>R,S-(3,1',C_18:1)-BMP>R,R-(1,1',C_18:1)>S,S-(2,2')-BMP. The cationic amphiphilic drug amiodarone (AMD) inhibited the deacylation of DOPG and BMP isomers by hLPLA2 in a concentration-dependent manner. Under these experimental conditions, the IC₅₀s of amiodarone-induced inhibition of the four BMP isomers and DOPG were less than 20 μM and approximately 30 μM, respectively. BMP accumulation was observed in AMD-treated RAW 264.7 cells. The accumulated BMP was significantly reduced by exogenous treatment of cells with active recombinant hLPLA2 but not with diisopropylfluorophosphate-inactivated recombinant hLPLA2. Finally, a series of cationic amphiphilic drugs known to cause phospholipidosis were screened for inhibition of LPLA2 activity as measured by either the transacylation or fatty acid hydrolysis of BMP or phosphatidylcholine as substrates. Fifteen compounds demonstrated significant inhibition with IC₅₀s ranging from 6.8 to 63.3 μM. These results indicate that LPLA2 degrades BMP isomers with different substrate specificities under acidic conditions and may be the key enzyme associated with BMP accumulation in drug-induced phospholipidosis.

Keywords: amiodarone; bis(monoacylglycerol)phosphate; cationic amphiphilic drug; lysosomal phospholipase; phospholipase A2 group 15; phospholipidosis.

Fig. 1

Structures of phosphatidylglycerol and the bis(monoacylglycerol)phosphate isomers used in this study.

Fig. 2

Degradation of 1,2-dioleoyl phosphatidylglycerol (DOPG) and sn-(2-oleoyl-3-hydroxy)-glycerol-1-phospho-sn-1'-(2′-oleoyl-3′-hydroxy)-glycerol (S,S-(2,2’,C_18:1)-BMP) by LPLA2. De-acylation activity by LPLA2 was determined as described in Materials and Methods. The reaction mixture contained 70 μM DODPC and 30 μM DOPG or S,S-(2,2′,C_18:1)-BMP as liposomes. The concentration of LPLA2 in the reaction mixture was 20 ng/ml and 200 ng/ml for DOPG assay and S,S-(2,2′,C_18:1)-BMP assay, respectively. The left and right panels show TLC plates and their kinetic curves, respectively.

Fig. 3

Degradation of (S,S-(2,2′,C_18:1)-BMP) by LPLA2 in the presence of N-acetyl-sphingosine (NAS). Deacylation and transacylation activities by LPLA2 were determined as described in Materials and methods. The reaction mixture contained 70 μM DODPC, 30 μM S,S-(2,2′,C_18:1)-BMP and 40 μM NAS as liposomes. The concentration of LPLA2 in the reaction mixture was 40 ng/ml. A: representative thin layer chromatogram demonstrating the formation of oleic acid and 1-O-oleoyl-N-acetylsphingosine in the presence of LPLA2. B: time dependent formation of oleic acid and 1-O-oleoyl-N-acetylsphingosine in the presence or absence of N-acetylesphingosine in the reaction mixture. Oleic acid and 1-O-oleoyl-NAS produced by LPLA2 in the presence of NAS were plotted by open circles (○) and open squares (□), respectively. The reaction rate curve indicated by closed circles (•) was same as that shown in the Fig. 1.

Fig. 4

Substrate specificity. LPLA2 was incubated with liposomes consisting of 70 μM DODPC and 30 μM DOPG or 30 μMBMP isomer such as R,S-(3,1′,C_18:1)-, R,R-(1,1′,C_18:1)-, S,S-(3,3′,C_18:1)- and S,S-(2,2′,C_18:1)-BMPs. The specific activities of LPLA2 for DOPG and BMP isomers were determined by the initial velocity using a linear kinetic curve for each reaction (A) as shown in supplemental Fig. S1. Panel B shows an expanded alignment of the specific activity for BMP isomers shown in Panel A. Asterisks (∗) indicate the distribution of LPLA2 activity on each substrate.

Fig. 5

Substrate preference with mixed isomers of BMP. Effect of tetrabutylammonium ion on S,S-(3,3′,C_18:1)-BMP degradation by hLPLA2. The reaction mixture contained 49 mM sodium acetate (pH 4.5), 10 μg/ml BSA, and 70 μM DODPC and 30 μM S,S-(3,3′,C_18:1)-BMP as liposomes, in the presence of or in the absence of 30 μM tetrabutylammonium hydroxide. The concentration of LPLA2 was 80 ng/ml. The reaction and lipid extraction were carried out as described in Materials and methods. A and B panels show TLC plate and their kinetics curves, respectively. The plate in panel A was developed in a solvent system consisting of C/M/pyridine (99:1:2, v/v). ACONT and TBAH indicate control (without tetrabutylammonium hydroxide) and tetrabutylammonium hydroxide (30 μM tetrabutylammonium hydroxide), respectively. C and D: Specificity of hLPLA2 for S,S-(2,2′,C_18:1)-BMP and S,S-(3,3′,C_18:1)-BMP. The reaction mixture contained 49 mM sodium acetate (pH 4.5), 10 μg/ml BSA, and 70 μM DODPC, 15 μM S,S-(2,2′,C_18:1)-BMP and 15 μM S,S-(3,3′,C_18:1)-BMP as liposomes. The reaction was initiated by adding hLPLA2 (the final concentration in the reaction mixture: 100 ng/ml), kept at 37°C, and terminated at the time points indicated in panel C as described in Materials and methods. The changes of both BMP peaks obtained by scanning of the plate in panel C are shown in panel D. 2,2′-BMP and 3,3′-BMP indicate S,S-(2,2′,C_18:1)-BMP and S,S-(3,3′,C_18:1)-BMP, respectively. Lipo denotes liposomes without LPLA2.

Fig. 6

Effect of amiodarone on the degradation of PODG and BMP isomers by LPLA2. LPLA2 was incubated with liposomes consisting of 70 μM DODPC and 30 μM DOPG or 30 μM BMP isomer such as S,R-(3,1′,C_18:1), R,R-(1,1′,C_18:1)-, S,S-(3,3′,C_18:1)- and S,S-(2,2′,C_18:1)-BMPs in the presence of different concentrations of amiodarone (1, 3.3, 10, 33.3 and 100 μM). LPLA2 activity was determined as described in the Methods section. The incubation times for each isomer were within the linear portion of the reactions as described in supplemental Fig. S1 and were respectively DOPG (2.5 min), R,S-BMP (12 min), R,R-(1,1′-diC_18:1)-BMP (15 min), S,S-(3,3′-diC_18:1)-BMP at (8 min), and S,S-(2,2′-diC_18:1)-BMP (32.5 min). LPLA2activity is expressed as percent of control activity as the average of duplicate assays. Control activity was LPLA2 activity in the absence of amiodarone. The control specific activities measured in duplicate for each substrate were DOPG 77.6 and 80.7 μmol/min/mg protein; S,R-(3,1′,C_18:1)-BMP 7.84 and 8.18 μmol/min/mg protein; R,R-(1,1′,C_18:1)-BMP 8.19 and 8.09 μmol/min/mg protein; S,S-(3,3′,C_18:1)-BMP 13.19 and 13.74 μmol/min/mg protein; and S,S-(2,2′,C_18:1)-BMP 0.977 and 1.05 μmol/min/mg protein.

Fig. 7

Effects of LPLA2 on cellular BMP accumulation by treatment with AMD. In panel A, recombinant hLPLA2 (85 ng/ml) treated with or without 20 mM DFP was incubated with liposomes (128 μM as phospholipid) consisting of DOPC/sulfatide/NAS (the molar ratio: 10:1:3) for 2.5 and 7.5 min at 37°C at pH 4.5. The reaction products were developed in a solvent system consisting of chloroform/acetic acid (9:1, v/v) (A). DFP-treated LPLA2 was used as inactive hLPLA2 in the cell culture system. See Materials and methods for details. RAW 264.7 cells were exposed with 0.312 μM AMD for 1, 3, 5 and 7 days or with 0.05% DMSO for 7 days. In panel B, cellular phospholipids extracted from AMD-treated or untreated cells were analyzed by TLC as described in Materials and methods. Panel D shows the cellular BMP level and error bars indicate standard deviation (n = 3). Also, the cells after 7-days AMD-treatment were replaced with fresh medium without AMD and incubated with hLPLA2 or inactive hLPLA2 (DFP-LPLA2) for 24 h (C and E). In panel E, the cellular BMP levels treated with these enzymes were compared to the BMP levels of 7-days AMD-treated cells (defined as 100%). Error bars indicate standard deviation (n = 4). The thin layer chromatograms displayed in panels B and C were obtained from the same plate but for clarity are displayed as two separate panels. The last lane in panel B is shown as spliced since it was not adjacent to the fifth lane in the original plate. The fifth lane in panel B and first lane in panel C represent the identical experimental condition and are duplicated in each panel. The raw data for these panels is accessible online at 10.6084/m9.figshare.25768965.

Supplemental Fig. 1

Initial deacylation curves for DOPG and BMP isomers. Ten ng of hLPLA2 was used in each reaction tube except for the S,S-(2,2′-diC18:1)-BMP assay, where 20 ng of LPLA2 was used. The net production of oleic acid by the enzyme (Y-axis) at each time point was determined by subtracting the oleic acid content in the absence of LPLA2 from the oleic acid content in the presence of LPLA2. For DOPG, the linear portion of the curve was observed between 0 and 2.5 min. The initial velocity of LPLA2 for DOPG were similarly determined. These were R,S-(3,1′- diC_18:1)-BMP, 0–16 min; R,R-(1,1′-diC_18:1)-BMP, 0–15 min; S,S-(3,3′-diC_18:1)-BMP, 0–10 min; and S,S-(2,2′-diC_18:1)-BMP, 0–60 min. Each panel represents one of three separate assays. The reaction curve in DOPG is the same as that shown in Fig. 2.