N WT cells there exists a natural variety of PI having C26:0 in the sn-1 position of glycerol [49], here called PI’, which order CCX282-B apparently cannot compensate for the complete loss of sphingolipids. PI’ ICG-001 web accounted for <1 of the total PI in WT cells, whereby no C26:0 was detected in other GPLs [43,49?1]. The genetic interaction of ELO2 and ELO3 with CST26 suggested that PI' helps cells to overcome a relative lack of sphingolipids and that Cst26 may be the still not identified acyltransferase making PI'. As shown in Fig 9A, blocking sphingolipid biosynthesis pharmacologically with myriocin (Myr) and Aureobasidin A (AbA) in BY4741 WT cells leads to the intensification of anPLOS Genetics | DOI:10.1371/journal.pgen.July 27,13 /Yeast E-MAP for Identification of Membrane Transporters Operating Lipid Flip FlopFig 9. Cst26 introduces VLCFAs into lyso-PI. Cells were grown to exponential phase in SC (2 mg/L of inositol). (A) 3 OD600 units of cells were washed and preincubated in inositol-free SC containing or not Myr (40 g/ml) and AbA (1 g/ml) for 10 min and then labeled with 10 Ci [2-3H]-myo-inositol for 4 h at 30 . Lipids were extracted, deacylated or not with NaOH, desalted and run on TLC in solvent 1. (B) zones in red rectangles in panel A were scraped off the plate, eluted with organic solvent, dried in a rotary evaporator and treated with PLA2. Red arrowheads point lyso-PI'. (C) radioscanning was used to quantify radioactivity in lanes containing lipids from Myr/AbA treated cells. Radioactivity was measured in the zone where lyso-PI' migrates and also in the rest of the lane (except for the origin); lyso-PI' was then calculated as a percentage of total radioactivity in the lane. In this way, results of the experiment of panel B and a second identical experiment were averaged. * P = 0.002. doi:10.1371/journal.pgen.1006160.gPLOS Genetics | DOI:10.1371/journal.pgen.July 27,14 /Yeast E-MAP for Identification of Membrane Transporters Operating Lipid Flip Flop[3H]-inositol labeled, mild base sensitive band that is less polar than PI, i.e. has a higher mobility on TLC than PI and that we tentatively considered as PI'. The increased synthesis of PI' can be attributed to the accumulation of C26:0-CoA, which in presence of Myr and AbA cannot flow towards its normal destination, ceramides. Biosynthesis of PI' is significantly weaker in cst26 (Fig 9A). Ordinary PI in WT cells most often contains a C16:0 in sn-1 and a C18:1 in sn-2 [43], so that phospholipase A2 (PLA2) reduces ordinary PI to a lyso-PI with a C16:0 in sn1 (Fig 9B, sample 1). When the [PI + PI'] of WT cells treated with Myr and AbA was subjected to PLA2 treatment, an additional lyso-PI with higher TLC mobility was generated (Fig 9B, sample 2). This additional lyso-PI is presumed to represent lyso-PI', retaining its C26:0 in sn-1; this species is also very abundant in lac1 lag1 cells (2.YDC1) that lack ceramide synthases [51]. PLA2 treatment of [PI + PI'] from cst26 cells treated with Myr/AbA, generated much less lyso-PI' than treatment of [PI + PI'] from Myr/AbA treated WT cells (Fig 9B, sample 5 vs. 2; Fig 9C). We interpret the data as to say that the bulk of PI carrying C26:0 in sn-1 in WT cells is generated by Cst26, and that in elo2 or elo3, Cst26 generates PIs having C18:0, C22:0 or C24:0 in sn-1 that may take over a function that is normally exerted by mature sphingolipids. Cst26 thus appears to be an acyl-CoA:lyso-PI acyltransferase that can transfer saturated FAs with 18 to 26 C atoms.Cwh.N WT cells there exists a natural variety of PI having C26:0 in the sn-1 position of glycerol [49], here called PI', which apparently cannot compensate for the complete loss of sphingolipids. PI' accounted for <1 of the total PI in WT cells, whereby no C26:0 was detected in other GPLs [43,49?1]. The genetic interaction of ELO2 and ELO3 with CST26 suggested that PI' helps cells to overcome a relative lack of sphingolipids and that Cst26 may be the still not identified acyltransferase making PI'. As shown in Fig 9A, blocking sphingolipid biosynthesis pharmacologically with myriocin (Myr) and Aureobasidin A (AbA) in BY4741 WT cells leads to the intensification of anPLOS Genetics | DOI:10.1371/journal.pgen.July 27,13 /Yeast E-MAP for Identification of Membrane Transporters Operating Lipid Flip FlopFig 9. Cst26 introduces VLCFAs into lyso-PI. Cells were grown to exponential phase in SC (2 mg/L of inositol). (A) 3 OD600 units of cells were washed and preincubated in inositol-free SC containing or not Myr (40 g/ml) and AbA (1 g/ml) for 10 min and then labeled with 10 Ci [2-3H]-myo-inositol for 4 h at 30 . Lipids were extracted, deacylated or not with NaOH, desalted and run on TLC in solvent 1. (B) zones in red rectangles in panel A were scraped off the plate, eluted with organic solvent, dried in a rotary evaporator and treated with PLA2. Red arrowheads point lyso-PI'. (C) radioscanning was used to quantify radioactivity in lanes containing lipids from Myr/AbA treated cells. Radioactivity was measured in the zone where lyso-PI' migrates and also in the rest of the lane (except for the origin); lyso-PI' was then calculated as a percentage of total radioactivity in the lane. In this way, results of the experiment of panel B and a second identical experiment were averaged. * P = 0.002. doi:10.1371/journal.pgen.1006160.gPLOS Genetics | DOI:10.1371/journal.pgen.July 27,14 /Yeast E-MAP for Identification of Membrane Transporters Operating Lipid Flip Flop[3H]-inositol labeled, mild base sensitive band that is less polar than PI, i.e. has a higher mobility on TLC than PI and that we tentatively considered as PI'. The increased synthesis of PI' can be attributed to the accumulation of C26:0-CoA, which in presence of Myr and AbA cannot flow towards its normal destination, ceramides. Biosynthesis of PI' is significantly weaker in cst26 (Fig 9A). Ordinary PI in WT cells most often contains a C16:0 in sn-1 and a C18:1 in sn-2 [43], so that phospholipase A2 (PLA2) reduces ordinary PI to a lyso-PI with a C16:0 in sn1 (Fig 9B, sample 1). When the [PI + PI'] of WT cells treated with Myr and AbA was subjected to PLA2 treatment, an additional lyso-PI with higher TLC mobility was generated (Fig 9B, sample 2). This additional lyso-PI is presumed to represent lyso-PI', retaining its C26:0 in sn-1; this species is also very abundant in lac1 lag1 cells (2.YDC1) that lack ceramide synthases [51]. PLA2 treatment of [PI + PI'] from cst26 cells treated with Myr/AbA, generated much less lyso-PI' than treatment of [PI + PI'] from Myr/AbA treated WT cells (Fig 9B, sample 5 vs. 2; Fig 9C). We interpret the data as to say that the bulk of PI carrying C26:0 in sn-1 in WT cells is generated by Cst26, and that in elo2 or elo3, Cst26 generates PIs having C18:0, C22:0 or C24:0 in sn-1 that may take over a function that is normally exerted by mature sphingolipids. Cst26 thus appears to be an acyl-CoA:lyso-PI acyltransferase that can transfer saturated FAs with 18 to 26 C atoms.Cwh.