A novel assay for the introduction of the vinyl ether double bond into plasmalogens using pyrene-labeled substrates

Abstract

Plasmanylethanolamine desaturase (PEDS) (EC 1.14.99.19) introduces the 1-prime double bond into plasmalogens, one of the most abundant phospholipids in the human body. This labile membrane enzyme has not been purified and its coding sequence is unknown. Previous assays for this enzyme used radiolabeled substrates followed by multistep processing. We describe here a straight-forward method for the quantification of PEDS in enzyme incubation mixtures using pyrene-labeled substrates and reversed-phase HPLC with fluorescence detection. After stopping the reaction with hydrochloric acid in acetonitrile, the mixture was directly injected into the HPLC system without the need of lipid extraction. The substrate, 1-O-pyrenedecyl-2-acyl-sn-glycero-3-phosphoethanolamine, and the lyso-substrate, 1-O-pyrenedecyl-sn-glycero-3-phosphoethanolamine, were prepared from RAW-12 cells deficient in PEDS activity and were compared for their performance in the assay. Plasmalogen levels in mouse tissues and in cultured cells did not correlate with PEDS levels, indicating that the desaturase might not be the rate limiting step for plasmalogen biosynthesis. Among selected mouse organs, the highest activities were found in kidney and in spleen. Incubation of intact cultivated mammalian cells with 1-O-pyrenedecyl-sn-glycerol, extraction of lipids, and treatment with hydrochloric or acetic acid in acetonitrile allowed sensitive monitoring of PEDS activity in intact cells.

Keywords: ether lipid; lyso-phosphatidyl ethanolamine; phosphatidyl ethanolamine; plasmanylethanolamine desaturase.

Fig. 1.

Reaction scheme of the PEDS assay carried out with lyso-substrate. The lyso-substrate, 1-O-pyrenedecyl-sn-glycero-3-phosphoethanolamine [I], is first acylated, presumably by CoA-independent transacylase, a reaction we find to be inhibited by PMSF and SKF 98625. The resulting 1-O-pyrenedecyl-2-acyl-sn-glycero-3-phosphoethanolamine [II] is heterogenous in that different residues (R) are attached. At least three different major peaks were separated by HPLC (see Fig. 3A). The PEDS reaction then introduces the vinyl ether double bond to yield 1-O-(1′Z)-pyrenedecenyl-2-acyl-sn-glycero-3-phosphoethanoalmines [III], i.e., pyrene-labeled plasmalogens. The amount of pyrene-labeled plasmalogens formed is then quantified by measuring pyrenedecanal released from the reaction products by HCl. To control for any nonplasmalogen-derived pyrenedecanal, the same enzyme incubation mixture is treated with acetic acid and analyzed in parallel. PE, phosphoethanolamine; R, fatty acid side chains.

Fig. 2.

Formation of pyrene-labeled plasmalogens from 1-O-pyrenedecyl-sn-glycerol by intact cells. A: Chromatograms of lipid extracts of RAW clones 12 and 108. Cells were cultivated for 24 h in the presence of 5 μM 1-O-pyrenedecyl-sn-glycerol, treated with HCl (to cleave the vinyl ether bond) or with acetic acid (which leaves the vinyl ether bond intact). RAW-12 has a defect in PEDS activity, which is intact in RAW-108 (8) (see text for details). The peak labeling corresponds to Fig. 1 ([II], 1-O-pyrenedecyl-2-acyl-sn-glycerophospholipids; [III], 1-O-(1′Z)-pyrenedecenyl-2-acyl-sn-glycerophospholipids; [IV], pyrenedecanal). X denotes peaks that originate from the reaction of pyrenedecanal with unknown compounds present in the lipid extract (see text for details). B: Time dependence of formation of pyrene-labeled phospholipids in RAW264.7 cells (filled circles, solid line) and RAW-12 cells (open circles, dashed line). Cells were cultivated for the indicated times in the presence of 5 μM 1-O-pyrenedecyl-sn-glycerol, harvested, lipids extracted, and the extracts treated with HCl (which cleaves the vinyl ether bond) or acetic acid (which leaves the vinyl ether bond intact), respectively. The amount of pyrene-labeled phospholipids was calculated from the area of retention times between 10 and 12 min in extracts treated with acetic acid. The mean ± SEM for four independent experiments is shown. C: Time dependence of formation of pyrene-labeled plasmalogen in RAW264.7 cells (filled circles, solid line) and RAW-12 cells (open circles, dashed line). The amount of pyrene-labeled plasmalogen was calculated from the amount of pyrenedecanal plus its derivative peaks (X) liberated upon HCl treatment and related to cellular protein. Free pyrenedecanal was not present, as judged by analysis of extracts treated with acetic acid only. The mean ± SEM for four independent experiments is shown. D: Percentage of pyrene-labeled 2-acyl-glycerophospholipids with vinyl ether double bond in RAW264.7 (filled circles, solid line) and RAW-12 (open circles, dashed line). Chromatograms of the extracts prepared for B and C were evaluated for the amount of aldehyde [including its derivatives (X)] formed in relation to the sum of aldehyde formed plus pyrene-labeled glycerophospholipids stable to HCl (retention times 10–12 min). The mean ± SEM for four independent experiments is shown.

Fig. 3.

PEDS assay with microsomal fractions. A: Chromatograms of typical enzyme incubations using microsomes from RAW-108 as the source for the enzymatic activity. The labeling of the peaks corresponds to Fig. 1 ([I], lyso substrate 1-O-pyrenedecyl-sn-glycero-3-phosphoethanolamine; [II], 1-O-pyrenedecyl-2-acyl-sn-glycerophospholipids; [III], 1-O-(1′Z)-pyrenedecenyl-2-acyl-sn-glycerophospholipids; [IV], pyrenedecanal) (see text for details). B: Inhibition of the enzymatic activity by PMSF. Assay incubations with substrate [II] (filled circles, solid line); assay incubations with lyso-substrate [I] (open circles, dashed line). The mean ± SEM for three independent experiments is shown. C: Inhibition of the enzymatic activity by SKF 96825. Assay incubations with substrate [II] (filled circles, solid line); assay incubations with lyso-substrate [I] (open circles, dashed line). The mean ± SEM for three independent experiments is shown. D: Inhibition of the enzymatic activity with thiol reagents. The enzymatic assay was performed with the lyso-substrate [I]. DTT (filled circles, solid line); DTE (open circles, dashed line); mercaptoethanol (open triangles, dashed line); TCEP (filled triangles, solid line). The mean ± SEM for three independent experiments is shown. E: Dependence of pyrenedecanal formed on the incubation time using the lyso-substrate [I] (mean ± SEM for three independent experiments). F: Dependence of the amount of pyrenedecanal formed on the amount of microsomal protein using the lyso-substrate [I] (mean ± S.E.M for three independent experiments). G: Dependence of the amount of pyrenedecanal formed on the concentration of lyso-substrate [I] employed (mean ± SEM for three independent experiments).

Fig. 4.

PEDS activities and plasmalogen levels in mouse tissues and cultured mammalian cells. A: PEDS levels in mouse tissues. Tissues from three male and three female 10-week-old C57BL/6N mice were homogenized in the presence of protease inhibitors. Microsomes were prepared and subjected to the PEDS assay using the lyso-substrate [I]. Because there was no gender difference, the mean ± SEM of six determinations is shown, with the exception of three for ovaries and testis, respectively. The dashed line indicates the detection limit of the method and gray bars indicate activities below the detection limit. B: Recovery of spiking of tissue microsomes with RAW-108 microsomes. Tissue microsomes (0.075 mg/ml) and RAW-108 microsomes (0.075 mg/ml) were incubated with the lyso-substrate [I] and compared with the PEDS activity with 0.075 mg/ml RAW-108 alone. The mean ± SEM for five to six independent measurements is shown, with the exception of three for ovary and testis, respectively. C: Plasmalogen levels in tissue homogenates. Plasmalogen levels were determined by HPLC quantification of dansyl-hydrazide derivatives as is detailed in the Materials and Methods section. The mean ± SEM for three independent measurements is shown D: PEDS activity in cell lines. Cell lines were cultivated under standard conditions, harvested, and microsomes prepared and subjected to the enzymatic assay using the lyso-substrate [I]. The mean ± SEM for three to eleven independent measurements is shown. E: Recovery of spiking of tissue microsomes with RAW-108 microsomes. A431 was used for spiking cell extracts instead of RAW-108 due to the one order of magnitude higher activity levels observed in cells as compared with mouse tissues. Microsomes (0.075 mg/ml) of the indicated cells and 0.075 mg/ml A431 microsomes were used with the lyso-substrate [I] and the recovery calculated. The mean ± SEM for two to eleven independent measurements is shown. F: Plasmalogen levels in cell homogenates. Plasmalogen levels were determined by HPLC quantification of dansyl-hydrazide derivatives as is detailed in the Materials and Methods section. The mean ± SEM for three independent measurements is shown.