LASS5 Is a Bona Fide Dihydroceramide Synthase That Selectively Utilizes Palmitoyl-CoA as Acyl Donor*

We demonstrated recently (Riebeling, C., Allegood, J.C., Wang, E., Merrill, A. H. Jr., and Futerman, A. H. (2003) J. Biol. Chem. 278, 43452–43459) that upon over-expression in human embryonic kidney cells, longevity assurance gene homolog 5 (LASS5, previously named TRH4) elevates the synthesis of (dihydro)ceramides selectively enriched in palmitic acid. To determine whether LASS5 is a bona fide dihydroceramide synthase or, alternatively, whether it modifies an endogenous dihydroceramide synthase, we over-expressed LASS5 with a hemagglutinin (HA) tag at the C terminus, solubilized it using digitonin, and purified it by immunoprecipitation. Solubilized LASS5-HA displays the same fatty acid selectivity as the membrane-bound enzyme. After elution from agarose beads, only one band could be detected by SDS-PAGE, and its identity was confirmed to be LASS5 by mass spectrometry. Dihydroceramide synthase activity of the eluted LASS5-HA protein was totally dependent on exogenously added phospholipids. Moreover, eluted LASS5-HA was highly selective toward palmitoyl-CoA as acyl donor and was inhibited by the (dihydro)ceramide synthase inhibitor, fumonisin B1. This study identifies LASS5 as a genuine dihydroceramide synthase and demonstrates that mammalian dihydroceramide synthases do not require additional subunits for their activity.

We demonstrated recently (Riebeling, C., Allegood, J.C., Wang, E., Merrill that upon over-expression in human embryonic kidney cells, longevity assurance gene homolog 5 (LASS5, previously named TRH4) elevates the synthesis of (dihydro)ceramides selectively enriched in palmitic acid. To determine whether LASS5 is a bona fide dihydroceramide synthase or, alternatively, whether it modifies an endogenous dihydroceramide synthase, we over-expressed LASS5 with a hemagglutinin (HA) tag at the C terminus, solubilized it using digitonin, and purified it by immunoprecipitation. Solubilized LASS5-HA displays the same fatty acid selectivity as the membrane-bound enzyme. After elution from agarose beads, only one band could be detected by SDS-PAGE, and its identity was confirmed to be LASS5 by mass spectrometry. Dihydroceramide synthase activity of the eluted LASS5-HA protein was totally dependent on exogenously added phospholipids. Moreover, eluted LASS5-HA was highly selective toward palmitoyl-CoA as acyl donor and was inhibited by the (dihydro)ceramide synthase inhibitor, fumonisin B1. This study identifies LASS5 as a genuine dihydroceramide synthase and demonstrates that mammalian dihydroceramide synthases do not require additional subunits for their activity.
Ceramide, an important lipid second messenger (1)(2)(3), consists of a sphingoid long chain base to which a fatty acid is attached at carbon-2 via an amide bond. Ceramide is also a key metabolite in the pathway of sphingolipid (SL) 2 biosynthesis (4), and within the past 2-3 years the molecular identities of most of the enzymes in this pathway have been identified (5). Among these, a family of mammalian genes that regulates ceramide synthesis has been discovered (6,7). Surprisingly, over-expression of each of these genes in various mammalian cells leads to an increase in ceramides containing different fatty acids (8 -10). Thus, over-expression of LASS1 leads to an increase in the synthesis of ceramides containing stearic acid (8), whereas over-expression of LASS5 leads to an increase of ceramide containing palmitic acid (9). However, it is not known whether these genes modify an endogenous ceramide synthase activity, and thereby confer fatty acid selectivity, or whether the LASS proteins themselves are bona fide dihydroceramide synthases.
We have now purified, by immunoprecipitation, LASS5 with an HA tag at the C terminus and demonstrate here that it is a genuine dihydroceramide synthase that displays the same fatty acid selectivity as seen upon its over-expression in mammalian cells. This is the first biochemical isolation of a mammalian dihydroceramide synthase and paves the way for studying the role of LASS proteins in regulating ceramide production both for SL biosynthesis and for the regulation of the defined aspects of cell physiology in which it plays vital roles.
Cell Culture and Transfection-Human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 100 IU/ml penicillin, and 100 g/ml streptomycin. Transfections using LASS5-HA (9) were performed by the calcium phosphate method. Transfections were also performed with pcDNA-HA as a control.
Immunoprecipitation of LASS5-HA-The digitonin-solubilized supernatant was concentrated using an iCON concentrator (Pierce) and then incubated with protein A-agarose for 1 h at 4°C to reduce nonspecific binding. After removal of protein A-agarose by centrifugation, the supernatant was incubated with an anti-HA agarose conjugate overnight at 4°C. The conjugate was then pelleted by centrifugation and washed with the same HEPES buffer used for homogenization followed by a wash with 1 M NaCl. LASS5-HA was eluted from the beads using 100 mM glycine, pH 2.5, and immediately neutralized using 1 M Trizma * This work was supported by Israel Science Foundation Grant 1047/03. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  (Tris base), pH 11. The eluate was concentrated using an iCON concentrator, and in some cases, DOPC or DOPS liposomes prepared as described (11), were added to the concentrated eluate prior to analysis of dihydroceramide synthase activity. The composition of the eluate was analyzed by SDS-polyacrylamide gel electrophoresis (10% polyacrylamide gel) and Western blotting using a rabbit polyclonal anti-HA antibody. Control immunoprecipitations were performed using cells that had been transfected with pcDNA-HA. Mass Spectrometry-Protein bands were excised from the SDS gel and subsequently reduced, alkylated, and in-gel-digested for 18 h with bovine trypsin (sequencing grade, Roche Diagnostics) at a concentration of 12.5 ng/l in 50 mM ammonium bicarbonate at 37°C. An extracted peptide solution was dried for subsequent matrix-assisted laser desorption/time of flight ionization (MALDI-TOF) and electrospray ionization mass spectrometric analyses.
Aliquots of the extracted peptide mixture and electroeluted proteins, dissolved in 0.1% trifluoroacetic acid or a mixture of formic acid/isopropanol/H 2 0 (1/3/2; v/v/v), were used for MALDI-TOF mass spectrometry. Peptide fingerprinting was performed on a Bruker Reflex III TM MALDI-TOF mass spectrometer (Bruker, Bremen, Germany) equipped with a delayed extraction ion source, a reflector, and a 337-nm nitrogen laser.
Dihydroceramide Synthase Assay-For cell fractions, 500 g of protein (determined using BCA reagent (Pierce)) was incubated with 15 M sphinganine and 20 M defatted bovine serum albumin (12) for 5 min at 37°C with or without fumonisin B1 (20 M), and the reaction was initiated by the addition of either 0.12 Ci of 1-[ 14 C]palmitoyl-CoA or 0.12 Ci of 1-[ 14 C]stearoyl-CoA for a further 20 min (9). When immunoprecipitated LASS5-HA was assayed, either sphinganine or sphingosine was used as long chain base, and the reaction time was 1 h. Lipids were extracted and levels of dihydroceramide synthesis analyzed by thin layer chromatography using chloroform/methanol/2 M ammonium hydroxide (40/10/1; v/v/v) as the developing solvent and using palmitoylsphingosine and stearoylsphingosine as standards. Lipids were visualized using a phosphorimaging screen (Fuji, Tokyo, Japan), recovered from TLC plates by scraping the silica directly into scintillation vials, and quantified by liquid scintillation counting.

RESULTS
Initial experiments were performed to establish a means to solubilize membrane-bound LASS5-HA. Digitonin, which was previously used to solubilize Lag1p (13), a yeast LASS homolog (14,15), was able to solubilize LASS5-HA and preserve dihydroceramide synthase activity. Most of the dihydroceramide synthase activity was recovered in the digitoninsolubilized supernatant, which correlated with levels of LASS5-HA detected by Western blotting (Fig. 1A). Digitonin-solubilized LASS5-HA displayed the same fatty acid specificity as membranebound LASS5, showing an ϳ10-fold higher activity toward palmitoyl-CoA than toward stearoyl-CoA (Fig. 1B), and was inhibited by the (dihydro)ceramide synthase inhibitor fumonisin B1 (16) to a similar extent as observed in the homogenate (Fig. 1B).
Digitonin-solubilized LASS5-HA was subsequently immunoprecipitated using an anti-HA agarose conjugate and eluted using 100 mM glycine. A significant fraction of the immunoprecipitated LASS5-HA could be eluted from the beads (Fig. 2A), and analysis by SDS-polyacrylamide gel electrophoresis revealed only one band, with a molecular mass of ϳ48 kDa (Fig. 2B), similar to the predicted molecular mass of LASS5-HA. No other bands could be detected reproducibly either by silver staining of the gel (Fig. 2B) or by Coomassie staining (not shown). The identity of the band as LASS5-HA was confirmed by Western blotting (Fig. 2C) and by MALDI-TOF mass spectrometry (TABLE ONE).
In addition, no other peptides co-migrated with LASS5 (TABLE ONE), which was confirmed by nano-liquid chromatography electrospray ionization tandem mass spectrometry (not shown).
Eluted LASS5-HA did not show any dihydroceramide synthase activity unless phospholipids were added to the eluate. Of the two phospho-  A, LASS5-HA was solubilized using digitonin, immunoprecipitated, eluted from the beads, and detected by Western blotting using a rabbit polyclonal anti-HA antibody. Hom, homogenate; Sol, digitonin-solubilized supernatant; pcDNA, homogenate of pcDNA-HA-transfected cells; IP, immunoprecipitated LASS5-HA after elution from the beads using SDS elution buffer; Eluate, LASS5-HA eluted using 100 mM glycine, pH 2.5. B, SDS-polyacrylamide gel electrophoresis of LASS5-HA eluted using 100 mM glycine and detected by silver staining. The position of molecular size markers is shown. The pcDNA lane is taken from an identical experiment but transfected using pcDNA-HA rather than LASS5-HA. This experiment was repeated four times, and in all cases, the identity of the band was confirmed by MALDI-TOF mass spectrometry (TABLE ONE) or by Western blotting (C) using the same antibody as in A. lipids tested, DOPC was more efficient than DOPS at preserving dihydroceramide synthase activity (Fig. 3A). Moreover, eluted LASS5-HA was highly selective toward palmitoyl-CoA compared with stearoyl-CoA (Fig. 3B), could be inhibited by fumonisin B1, and was able to acylate both sphinganine and sphingosine (Fig. 3C). 3

DISCUSSION
In the current study we have demonstrated that LASS5 is a bona fide (dihydro)ceramide synthase. This is the first time that a mammalian dihydroceramide synthase has been purified, representing one of the last recalcitrant enzymes in the SL biosynthetic pathway (5).
In contrast to yeast, in which SLs contain only one kind of fatty acid, namely C26, mammalian (dihydro)ceramides contain a wide fatty acid spectrum (2,4). It was formerly assumed that this was due to a lack of specificity of dihydroceramide synthase with respect to the use of fattyacyl-CoAs. However, the discovery of a family of mammalian LASS genes that each synthesizes dihydroceramides containing different fatty acids (at least those characterized to date (8 -10)) demonstrates that this is not the case. In contrast, yeast have only two highly homologous ceramide synthase genes, LAG1 and LAC1, which together are responsible for the synthesis of C 26 -ceramides.
Evidence is accumulating that ceramides containing specific fatty acids are involved in defined cell functions in mammalian cells (2,17). This being the case, it might be expected that individual LASS genes would be expressed in either a tissue-specific (9,18) or temporal manner so as to supply specific ceramides for the distinct events in which they are involved. Our demonstration that LASS5 is a bona fide dihydroceramide synthase that selectively utilizes palmitoyl-CoA supports this likelihood and strengthens the hypothesis that specific LASS proteins play distinct roles either in ceramide signaling or in SL metabolism. Evidence for the latter has already been obtained by the observation that C 18 -ceramide formed by LASS1 is selectively channeled into neutral glyco-SLs but not gangliosides. The availability of purified LASS5 will permit systematic characterization of the reaction mechanism and modes of regulation of a mammalian dihydroceramide synthase (19).
The yeast ceramide synthase, Lag1p/Lac1p, was recently purified, and was shown to co-immunoprecipitate with an additional subunit, Lip1p (13), which was absolutely required for yeast ceramide synthesis, at least using C26-fatty acids. No mammalian homologs of Lip1p are found by data base searches, and our demonstration that no other protein is required for the dihydroceramide synthase activity of LASS5 indicates that LASS5 by itself is sufficient for catalytic activity in mammalian cells. However, the dihydroceramide synthase activity of LASS5 is totally dependent on exogenously added phospholipids, whereas no such requirement was observed for yeast Lag1p/Lac1p/Lip1p. Together, these data suggest a unique role for Lip1p in yeast that is not required in mammalian cells, and suggest that the regulation of mammalian (dihydro)ceramide synthesis may differ significantly from that in yeast.

MALDI-TOF MS identification of LASS5
Experimentally determined values and the corresponding theoretical masses of tryptic peptides are listed; data base analysis confirmed the protein as LASS5 (AAH46797). Mass accuracy (ppm) and amino acid sequences are shown for each peptide.