α/β Hydrolase Domain-containing 6 (ABHD6) Degrades the Late Endosomal/Lysosomal Lipid Bis(monoacylglycero)phosphate

Abstract

α/β Hydrolase domain-containing 6 (ABHD6) can act as monoacylglycerol hydrolase and is believed to play a role in endocannabinoid signaling as well as in the pathogenesis of obesity and liver steatosis. However, the mechanistic link between gene function and disease is incompletely understood. Here we aimed to further characterize the role of ABHD6 in lipid metabolism. We show that mouse and human ABHD6 degrade bis(monoacylglycero)phosphate (BMP) with high specific activity. BMP, also known as lysobisphosphatidic acid, is enriched in late endosomes/lysosomes, where it plays a key role in the formation of intraluminal vesicles and in lipid sorting. Up to now, little has been known about the catabolism of this lipid. Our data demonstrate that ABHD6 is responsible for ∼ 90% of the BMP hydrolase activity detected in the liver and that knockdown of ABHD6 increases hepatic BMP levels. Tissue fractionation and live-cell imaging experiments revealed that ABHD6 co-localizes with late endosomes/lysosomes. The enzyme is active at cytosolic pH and lacks acid hydrolase activity, implying that it degrades BMP exported from acidic organelles or de novo-formed BMP. In conclusion, our data suggest that ABHD6 controls BMP catabolism and is therefore part of the late endosomal/lysosomal lipid-sorting machinery.

Keywords: bis(monoacylglycero)phosphate; endocannabinoid; endosome; lipid metabolism; lysobisphosphatidic acid; lysosome; phospholipase.

FIGURE 1.

Substrate selectivity and pH dependence of ABHD6. A, the coding sequence of mouse ABHD6 was cloned into a mammalian expression vector (pcDNA4/HisMaxC) and overexpressed in COS-7 cells. Cell lysates (1000 × g supernatant) containing recombinant ABHD6 were incubated with various polar and neutral lipids. Lipid degradation was monitored by measuring the release of FFA. Cells expressing β-galactosidase (LacZ) were used as a negative control. Data are expressed as -fold increase in FFA release over LacZ. For more information about lipids, see “Experimental Procedures.” Inset, sn-2,2′-dioleoyl BMP. B and C, time- (B) and dose-dependent (C) hydrolysis of BMP by Cos-7 lysates overexpressing ABHD6. The activity detected in cells expressing LacZ was set as blank. D, degradation of sn-3,3′-BMP(S,S), sn-2,2′-BMP(S,S), sn-3,3′-BMP (R,R), and bis(diacylglycero)phosphate(S,S) by purified GST-tagged ABHD6. E, pH dependence of purified GST-ABHD6 using BMP(S,S) as substrate. Activity assays were performed in 50 mm acetate (pH 4.5–5.5), MES (pH 5.5–6.5), and bis-tris propane buffer (pH 6.5–9.0). F, substrate saturation kinetics of purified GST-ABHD6 using BMP(S,S) as substrate. G, K_m and V_max values using BMP(S,S), racemic MO, and LPG as substrate. Values were calculated by nonlinear regression analysis (SigmaPlot). H, TLC analysis showing that partial digestion of BMP with GST-ABHD6 results in the accumulation of LPG and FFA. After digestion, lipids were isolated by butanolic extraction and applied to TLC analysis using chloroform/methanol/water (60/25/4, v/v/v) as solvent. GST purified under identical conditions served as a negative control. I and F, FFA (I) and glycerol (F) release using BMP(S,S), racemic MO, or PG as substrate for GST-ABHD6. FFA and glycerol release were quantified using commercially available kits. Data are presented as mean ± S.D. (***, p < 0.001, Student's t test).

FIGURE 2.

ABHD6 co-localizes with late endosomes/lysosomes. A and B, lysosomal fractions of liver homogenates from overnight-fasted C57Bl/6 animals were isolated by (A) ultracentrifugation in a discontinuous OptiPrep™ gradient or (B) differential centrifugation. 10 μg of protein of various fractions was analyzed by Western blotting using an antibody specific for ABHD6 (14) and marker antibodies for LE/lysosomes (Rab-7 and LAMP-1), ER (PDI, protein disulfide isomerase), and mitochondria (COXIV). L, lysate (1000 × g supernatant); F1-F3, fractions 1–3 harvested from the top of the gradient at the interphase 0–10%, 10–15%, and 15–20% iodixanol, respectively; P, 12,000 × g pellet; CLF, crude lysosomal fraction; H, homogenate. C, protease protection assay performed with lysosomal liver fractions obtained using OptiPrep^TM gradient (F1). The F1 fraction was treated with 30 ng pf proteinase K at 37 °C for 5 min or carrier alone (control). Subsequently, proteins were precipitated with acetone and subjected to Western blot analysis. Proteins were detected using specific primary antibodies against Cathepsin D (Abcam) and ABHD6. D, live-cell imaging of COS-7 cells transfected with ABHD6-ECFP and co-stained with ER-Tracker DsRED2 (ER) and lipid droplet (LD) marker (LipidTOX^TM DeepRed, Invitrogen). Also shown is immunofluorescence of COS-7 transfected with tagless ABHD6. The enzyme was detected using ABHD6-specific antibody and a fluorescent labeled secondary antibody. Non-transfected cells were not stained by the antibody and served as a negative control. E and F, live-cell imaging of COS-7 cells co-transfected with ABHD6-ECFP and the LE marker RFP-Rab7. DIC, differential interference contrast. G, live-cell imaging of COS-7 cells co-transfected with ABHD6-CFP and the early endosome marker RFP-Rab5. (Scale bar = 10 μm).

FIGURE 3.

BMP is resistant to lysosomal hydrolases and degraded efficiently in the slightly alkaline pH range in a process involving ABHD6. A and B, pH dependence of (A) BMP and (B) PG hydrolase activity of brain lysates. Degradation of BMP(S,S) and PG (2 mm each) was monitored between pH 4.5 and 9.0 as described in Fig. 1D. C, inhibition of BMPH activity of purified ABHD6 by KT182 using BMP(S,S) as substrate. D, inhibition of brain and liver BMPH activity by KT182. E and F, inhibition of BMPH activity by KT182 in LF of liver and brain. LF were purified as described in Fig. 2B. G and H, inhibition of MG hydrolase activity in LF of liver and brain by KT182. C—H, activity assays were performed at pH 8.0 using BMP(S,S) and 2-monoolein (2-MO) as substrate (2 mm each). Data are presented as mean ± S.D. and represent at least three independent experiments. *, p < 0.05; **, p < 0.01; ***, p < 0.001; analysis of variance.

FIGURE 4.

ABHD6 affects BMP catabolism. A, incorporation of ³H-labeled oleic acid into BMP of AML12 hepatocytes. Cells were cultivated for 24 h in DMEM/F-12 (1:1) containing 1% FCS. Subsequently, cells were exposed to DMEM/F-12 (1:1) containing 20% FCS and ³H-labeled oleic acid for 4 h either in the absence or presence of KT182. Inhibitor or carrier (dimethyl sulfoxide) was added 1 h before exposure to DMEM/F-12 (1:1) containing 20% FCS/[³H]oleic acid. Total lipids were extracted and separated by TLC analysis. The radioactivity co-migrating with BMP was determined by scintillation counting. B and C, effect of KT182 on BMP level and distribution in AML12 cells. Cells were cultivated under identical conditions as described in A but in the absence of ³H-labeled oleic acid. Total lipids were extracted in the presence of BMP 14:0 as internal standard and subjected to LC-MS analysis. For quantification of BMP 36:2 (B), a calibration curve was prepared. Other molecular species are shown in arbitrary units (AU)/mg of protein. D–G, liver BMP levels of male C57BL/6N mice treated with ASOs targeting ABHD6 (ASOα and ASOβ) or control ASO. Mice were maintained for 12 weeks on a normal chow diet or on a high-fat diet and injected biweekly with ASOs (25 mg/kg of body weight) as described previously (14). Subsequently, lipids were extracted from liver samples in the presence of BMP 14:0 (Avanti Polar Lipids) as internal standard and subjected to LC-MS analysis. D and E, relative total BMP levels (D) and BMP distribution (E) of mice maintained on a chow diet. F and G, relative total BMP levels (F) and BMP distribution (G) of mice maintained on a high-fat diet. The experiments shown in A–C are representative of two independent experiments performed in triplicate. Data are presented as mean ± S.D. *, p < 0.05; **, p < 0.01; ***, p < 0.001; analysis of variance).

FIGURE 5.

Activity of human ABHD6 and mutant variants of the enzyme. A, the coding sequence of hABHD6 was cloned into a mammalian expression vector (pcDNA4/HisMaxC), and the protein was expressed in COS-7 cells. Mutations were introduced by site-directed mutagenesis. For enzymatic assays, cell lysates (1000 × g supernatant) were incubated with BMP(S,S), and FFA release was determined. Lysates obtained from cells expressing β-galactosidase (LacZ) were used as a negative control. B, selected mutations of hABHD6 as reported by the NCBI dbSNP database (http://www.ncbi.nlm.nih.gov/projects/SNP). C, BMPH activity of wild-type hABHD6 and mutant variants of the enzyme. Data are expressed as percent activity of wild-type hABHD6 and were normalized to the expression levels of His-tagged proteins. The activity detected in lysates of LacZ-expressing cells was set as blank. Inset, Western blot of His-tagged variants using an anti-His antibody. Data are presented as mean ± S.D. ***, p < 0.001; analysis of variance.

FIGURE 6.

Proposed role of ABHD6 in the late endosomal/lysosomal pathway. Endocytic vesicles form early endosomes in the peripheral cytoplasm that move toward the perinuclear area along microtubuli. Their conversion into LE involves the formation of BMP-containing ILVs. LE finally fuse with lysosomes to form endolysosomes (51). BMP is resistant to acid hydrolysis and forms stable membrane structures within acidic organelles, promoting lipid digestion and sorting. BMP-rich ILVs of LE and endolysosomes can fuse with the limiting membrane, allowing the export of ILV contents. After back-fusion, BMP appears on the limiting membrane and is degraded by ABHD6 into LPG and FFA. Additionally, ABHD6 may counteract de novo BMP synthesis on the limiting membrane of LE.