The ORMDL/Orm-serine palmitoyltransferase (SPT) complex is directly regulated by ceramide: Reconstitution of SPT regulation in isolated membranes

Abstract

Sphingolipids compose a lipid family critical for membrane structure as well as intra- and intercellular signaling. De novo sphingolipid biosynthesis is initiated by the enzyme serine palmitoyltransferase (SPT), which resides in the endoplasmic reticulum (ER) membrane. In both yeast and mammalian species, SPT activity is homeostatically regulated through small ER membrane proteins, the Orms in yeast and the ORMDLs in mammalian cells. These proteins form stable complexes with SPT. In yeast, the homeostatic regulation of SPT relies, at least in part, on phosphorylation of the Orms. However, this does not appear to be the case for the mammalian ORMDLs. Here, we accomplished a cell-free reconstitution of the sphingolipid regulation of the ORMDL-SPT complex to probe the underlying regulatory mechanism. Sphingolipid and ORMDL-dependent regulation of SPT was demonstrated in isolated membranes, essentially free of cytosol. This suggests that this regulation does not require soluble cytosolic proteins or small molecules such as ATP. We found that this system is particularly responsive to the pro-apoptotic sphingolipid ceramide and that this response is strictly stereospecific, indicating that ceramide regulates the ORMDL-SPT complex via a specific binding interaction. Yeast membranes harboring the Orm-SPT system also directly responded to sphingolipid, suggesting that yeast cells have, in addition to Orm phosphorylation, an additional Orm-dependent SPT regulatory mechanism. Our results indicate that ORMDL/Orm-mediated regulation of SPT involves a direct interaction of sphingolipid with the membrane-bound components of the SPT-regulatory apparatus.

Keywords: cell signaling; endoplasmic reticulum (ER); homeostasis; lipid metabolism; lipid signaling; phytoceramide; sphingolipid; sphingomyelin; sphingosine.

Figure 1.

ORMDL-dependent inhibition of SPT by ceramide can be reconstituted with isolated membranes. A, SPT activity was measured in intact and permeabilized cells as described under “Experimental procedures.” B, lysates and total membranes were prepared and assay of SPT in response to 10 μm C8 ceramide was performed as described under “Experimental procedures” with incubations at 37 °C for 60 min. MeOH/BSA solutions were used as the control. Also shown is inhibition of SPT by 1 μm myriocin. Shown are the mean of nine technical replicates for intact and permeabilized cells and quadruplicate technical replicates for lysates and membranes, mean ± S.D. Shown is one representative of two duplicate experiments. C, samples assayed in A and B were assessed for ORMDL and subunit 1 of SPT (SPTLC1) by immunoblotting in the presence and absence of ORMDL knockdown by siRNA. Asterisks denote significance (p < 0.01) between control and C8 ceramide-treated samples by Student's two-tailed t test.

Figure 2.

Ceramides generated either in intact cells or in isolated membranes trigger ORMDL-dependent inhibition of SPT. A, cells preincubated either in the presence or absence of the ceramide synthase inhibitor Fumonisin B1 (FB1) overnight were incubated with or without 10 μm sphingosine or 10 μm C8 ceramide for 2 h then assayed for de novo sphingolipid biosynthesis. Shown are the mean ± S.D. of myriocin-inhibitable counts of six technical replicates. Asterisks denote significance (p < 0.001) between control and C8 ceramide or sphingosine-treated samples by Student's two-tailed t test. Shown is one of two virtually identical experiments. B, total cell membranes were preincubated with vehicle, 20 μm C8 ceramide, 20 μm sphingosine, 20 μm dihydrosphingosine, 50 μm 24:1 CoA, or sphingosines and 24:1 CoA for 40 min at 37 °C as described under “Experimental procedures.” The membranes were then assayed for SPT activity as described under “Experimental procedures.” Membranes were tested either in the absence or presence of the ceramide synthase inhibitor FB1 (20 μm). Also included was an incubation with 1 μm myriocin, a specific inhibitor of SPT, as a control. Asterisks denote significance (p < 0.05) between control and C8 ceramide or (dh)sphingosine and/or 24:1 CoA-treated samples by Student's two-tailed t test. C, membranes from cells transfected either with control siRNA or siRNA directed against all three ORMDL isoforms were preincubated with vehicle, 20 μm C8 ceramide, or 20 μm sphingosine and 50 μm 24:1 CoA, for 40 min as described under “Experimental procedures” and then assayed for SPT activity as described under “Experimental procedures.” Asterisks denote significance (p < 0.01) between control and ORMDL-depleted membranes with or without the indicated lipid treatments by Student's two-tailed t test. Shown is one of at least two (for intact cells) or three (for isolated membranes) independent experiments. Data are presented as the mean ± S.D. for 4 technical replicates.

Figure 3.

The response of membranes containing overexpressed ORMDL3 and a single-chain fusion of SPT subunits does not require either ATP or cytosol. ORMDL-dependent inhibition of SPT by ceramide in membranes overexpressing a single-chain fusion of SPT subunits and ORMDL3 is independent of ATP and cytosol. A, HeLa cells were transfected with either a control siRNA in combination with a construct expressing scSPT (22) and a plasmid directing overexpression of mouse ORMDL3 (labeled SPT + ORMDL) or siRNAs directed against all three ORMDL isoforms in combination with a plasmid only for scSPT (labeled SPT/siORMDL). After 24 h cells were broken by Dounce homogenization as described under “Experimental procedures.” Either total lysates (11 μg, unfractionated, containing both cytosol and membranes) or cytosol-free membranes (12 μg, isolated by ultracentrifugation) were assayed for SPT activity. Where indicated ATP (1 mm) or cytosol (8 μg) was added. Assays included BSA (control), BSA/C6 ceramide (10 μm), or myriocin (1 μm), a direct inhibitor of SPT. MeOH/BSA solutions were used as the Control. SPT activity was measured as described under “Experimental procedures” by the incorporation of [³H]serine into a total sphingolipid fraction. Incubations were performed for 90 min at 37 °C. Shown are the averages of triplicate samples, mean ± S.D. Asterisks denote significance (p < 0.02) between control and C6 ceramide-treated samples by Student's two-tailed t test. Shown is one representative of 3 similar experiments. B, acute ceramide-dependent inhibition of SPT does not involve changes in ORMDL levels. Membranes derived from HeLa cells overexpressing human ORMDL3 and scSPT were incubated with either BSA/MeOH or BSA/C8 ceramide (10 μm final concentration) under conditions identical to those required to trigger SPT inhibition (e.g. Fig. 1B). Membranes were collected by ultracentrifugation and subjected to SDS-PAGE electrophoresis, transferred to PVDF, and immunoblotting. Blots were cut just below the 47-kDa marker. The upper blot was probed with an antibody to the SPT subunit SPTLC1, which stains both endogenous SPTLC1 and the SPT fusion protein (scSPT). The lower blot was probed with an antibody to ORMDL. Shown are an equivalent amount of the membranes used in the incubation (Memb.) and triplicate incubations with either the BSA/MeOH control (MeOH) or BSA/C8 ceramide (C8-Cer).

Figure 4.

The ORMDL–SPT complex responds to ceramide and phytoceramide. The response to ceramide is more robust as the length of the N-acyl chain increases. 20 μm of the indicated sphingolipids, added as complexes with BSA, were tested for the ability to trigger inhibition of SPT activity as described under “Experimental procedures.” MeOH/BSA solutions were used as the control. The SPT inhibitor, myriocin (1 μm), was also tested to establish SPT-specific incorporation of [³H]serine into the sphingolipid fraction. Incubations were performed for 60 min at 37 °C. Shown are the averages of triplicate samples, mean ± S.D. Asterisks denote significance (p < 0.01) between control and lipid-treated samples by Student's two-tailed t test. Shown is one representative of 3 similar experiments. Con, BSA control; C2-C8 Cer, the fatty acid N-acylated to the sphingosine backbone; Phyto C6 Cer, C6 phytoceramide; C6 SM, C6 sphingomyelin; Sph, sphingosine; Myr, myriocin.

Figure 5.

SPT in isolated yeast membranes rapidly responds to phytoceramide in an Orm-dependent manner. A, Orm dependence of phytoceramide inhibition of SPT activity in isolated yeast microsomes. Microsomes isolated from WT yeast (WT) and yeast strains with deletions of Orm1 (orm1) Orm2 (orm2) or deleted of both Orm isoforms (orm1/2) were assayed for SPT activity in the absence (Control) or presence of 20 μm C8 phytoceramide (+ C8 phytoceramide) as described under “Experimental procedures.” Nonspecific background, as determined by assays in the absence of palmitoyl-CoA (less than 10% of signal), was subtracted. Asterisks denote significance (p < 0.001) between control and C8 phytoceramide-treated samples by Student's two-tailed t test. B, phytoceramide inhibition of SPT is rapid. Assays of SPT activity in isolated yeast microsomes was determined at various times as indicated in the absence (Control) or presence of 20 μm C8 phytoceramide (+ C8 phytoceramide) as described under “Experimental procedures.” The sample collected after 15 s, the shortest practical incubation time, is considered the background in this assay. Shown in A are the mean ± S.D. of 3 separate experiments with triplicate technical replicates in each. Shown for B is the mean ± S.D. of single points from 4 independent experiments. Asterisks denote significance (p < 0.005) between control and C8 phytoceramide-treated samples by Student's two-tailed t test.

Figure 6.

SPT in yeast microsomes is more responsive to phytoceramide than ceramide or dihydroceramide. The indicated lipids were added at 20 μm to yeast microsomes and assayed for SPT activity as described under “Experimental procedures.” Nonspecific background, as determined by assays in the absence of palmitoyl-CoA (less than 10% of signal), was subtracted. Shown are the mean ± S.D. of 3 separate experiments with triplicate technical replicates in each. Asterisks denote significance (p < 0.005) between control and sphingolipid-treated samples by Student's two-tailed t test.

Figure 7.

The ORMDL–SPT complex in isolated membranes is only responsive to the native stereoisomer of ceramide. The indicated stereoisomers of either (A) C8 ceramide (40 μm, tested in membranes from untransfected cells) or (B) C2 ceramide (20 μm, tested in membranes from cells overexpressing scSPT and hORMDL3), were added as complexes with BSA and tested for the ability to trigger inhibition of SPT as described under “Experimental procedures.” MeOH/BSA solutions were used as the Control. Note that d-erythro ceramide is the natural stereoisomer. The SPT inhibitor myriocin (1 μm) was also tested to establish SPT-specific incorporation of [³H]serine into the sphingolipid fraction. Incubations were performed for 60 min at 37 °C. Shown are the averages of quadruplicate samples, mean ± S.D. For A, single asterisks denote significance (p < 0.001) between control and ceramide-treated samples by Student's two-tailed t test, double asterisks denote significance (p < 0.02) between control and ceramide-treated samples by Student's two-tailed t test. For B, asterisks denote significance (p < 0.01) between control and ceramide-treated samples by Student's two-tailed t test. Shown is one representative of 3 similar experiments.