Switching head group selectivity in mammalian sphingolipid biosynthesis by active-site engineering of sphingomyelin synthases

Article abstract of DOI:10.1194/jlr.M068692

SM is a fundamental component of mammalian cell membranes that contributes to mechanical stability, signaling, and sorting. Its production involves the transfer of phosphocholine from phosphatidylcholine onto ceramide, a reaction catalyzed by SM synthase (SMS) 1 in the Golgi and SMS2 at the plasma membrane. Mammalian cells also synthesize trace amounts of the SM analog ceramide phosphoethanolamine (CPE), but the physiological relevance of CPE production is unclear. Previous work revealed that SMS2 is a bifunctional enzyme producing both SM and CPE, whereas a closely related enzyme, sphingomyelin synthase-related protein (SMSr)/SAMD8, acts as a monofunctional CPE synthase in the endoplasmatic reticulum. Using domain swapping and site-directed mutagenesis on enzymes expressed in defined lipid environments, we here identified structural determinants that mediate head group selectivity of SMS family members. Notably, a single residue adjacent to the catalytic histidine in the third exoplasmic loop profoundly influenced enzyme specificity, with glutamic acid permitting SMS-catalyzed CPE production and aspartic acid confining the enzyme to produce SM. An exchange of exoplasmic residues with SMSr proved sufficient to convert SMS1 into a bulk CPE synthase. This allowed us to establish mammalian cells that produce CPE rather than SM as the principal phosphosphingolipid and provide a model of the molecular interactions that impart catalytic specificity among SMS enzymes.

Full text of DOI:10.1194/jlr.M068692

Switching head group selectivity in mammalian
sphingolipid biosynthesis by active-site engineering of
sphingomyelin synthases
1
,
,†
1, ,†
†
Matthijs Kol, * Radhakrishnan Panatala, * Mirjana Nordmann,* Leoni Swart,
†
Leonie van Suijlekom, Birol Cabukusta,* Angelika Hilderink,* Tanja Grabietz,* John G. M. Mina,*
Pentti Somerharju, Sergei Korneev,* Fikadu G. Tafesse,** and Joost C. M. Holthuis *
§
2, ,†
Molecular C ell B iology D ivision, D epartment o f B iology/Chemistry,* U niversity o f O snabruck, 4 9076
Osnabruck, G ermany; M embrane B iochemistry & B iophysics, B ijvoet C enter a nd I nstitute o f
†
Biomembranes, U trecht U niversity, 3 584 C H U trecht, T he N etherlands; M edical B iochemistry, I nstitute o f
§
Biomedicine, U niversity o f H elsinki, H elsinki 0 0014, F inland; a nd M olecular M icrobiology &
Immunology,** O regon H ealth & S cience U niversity, P ortland, O R 9 7239
Abstract SM is a fundamental component of mammalian cell
Supplementary key words cell-free expression • ceramide phospho­
ethanolamine • c lick c hemistry • e nzyme m echanisms • G olgi a ppara­
tus • lipid biochemistry • lipidomics • model membranes • protein
engineering
membranes that contributes to mechanical stability, signaling,
and sorting. Its production involves the transfer of phospho-
choline from phosphatidylcholine onto ceramide, a reaction
catalyzed by SM synthase (SMS) 1 in the Golgi and SMS2 at the
plasma membrane. Mammalian cells also synthesize trace
amounts of the SM analog ceramide phosphoethanolamine
SM is a major structural component of mammalian cell
membranes and one of the end points in sphingolipid bio­
synthesis. T he b ulk o f S M i s p roduced i n t he G olgi l umen
and delivered by vesicular transport to the plasma mem­
brane, where it accumulates in the exoplasmic leaflet (1).
Owing to its unique ability to form extensive hydrogen
bonds with other membrane molecules, SM participates in
a multitude of cellular processes. SM is the preferred inter­
action partner of cholesterol, and this interaction has
important physiological consequences. Cell surface SM
degradation causes cholesterol to redistribute to the endo­
plasmatic reticulum (ER) (2), leading to downregulation
of HMG-CoA reductase, the rate-limiting step in choles­
terol biosynthesis (3). Besides directly influencing cellular
cholesterol h omeostasis, S M ( along w ith c holesterol) l ikely
contributes to the high packing density and thickening of
the lipid bilayer of the trans G olgi a nd p lasma m embrane,
which may influence protein sorting through hydrophobic
(CPE), but the physiological relevance of CPE production is
unclear. Previous work revealed that SMS2 is a bifunctional
enzyme producing both SM and CPE, whereas a closely related
enzyme, sphingomyelin synthase-related protein (SMSr)/
SAMD8, acts as a monofunctional CPE synthase in the endo-
plasmatic reticulum. Using domain swapping and site-directed
mutagenesis on enzymes expressed in defined lipid environ-
ments, we here identified structural determinants that mediate
head group selectivity of SMS family members. Notably, a sin-
gle residue adjacent to the catalytic histidine in the third exo-
plasmic loop profoundly influenced enzyme specificity, with
glutamic acid permitting SMS-catalyzed CPE production and
aspartic acid confining the enzyme to produce SM. An ex-
change of exoplasmic residues with SMSr proved sufficient to
convert SMS1 into a bulk CPE synthase. This allowed us to
establish mammalian cells that produce CPE rather than SM as
the principal phosphosphingolipid and provide a model of the
molecular interactions that impart catalytic specificity among
SMS enzymes.—Kol, M., R. Panatala, M. Nordmann, L. Swart,
L. v an S uijlekom, B . C abukusta, A . H ilderink, T . G rabietz, J . G . mismatching of membrane spans (4, 5). Specific inter­
M. M ina, P . S omerharju, S . K orneev, F . G . T afesse, a nd J . C . M .
Holthuis. Switching head group selectivity in mammalian
sphingolipid biosynthesis by active-site engineering of
sphingomyelin synthases. J. Lipid Res. 2016. 57: 1273–1285.
actions between SM and membrane spans have been re­
ported to influence the activity of various integral membrane
Abbreviations: Asp, a spartic a cid; C PE, c eramide p hosphoethanol­
amine; C PI, c eramide p hosphoinositol; E R, e ndoplasmatic r eticulum;
Glu, glutamic acid; HA, hemagglutinin; His, histidine; IMDM, Iscove
modified D ulbecco m edium; N BD, n itrobenzooxadiazole; N BD-Cer,
NBD-C6-ceramide; N BD-CPE, N BD-C6-CPE; O RF, o pen r eading f rame;
PC, phosphatidylcholine, PE, phosphatidylethanolamine; PI, phos­
phatidylinositiol; SMS, sphingomyelin synthase; SMSr, sphingomy­
elin synthase­related protein.
This work was supported by a Marie Curie Intra-European Fellowship
(
J.G.M.M.), the European Union Seventh Framework Programme Marie-Curie
ITN ‘Sphingonet’ Grant (289278), the Deutsche Forschungsgemeinschaft
Sonderforschungsbereich/SFB944-P14 (J.C.M.H.), and an Incentive Award of
the Faculty of Biology/Chemistry from the University of Osnabruck (M.K.). The
authors declare that they have no conflicts of interest with the contents of this
article.
Manuscript received 22 April 2016 and in revised form 9 May 2016.
1
M. Kol and R. Panatala contributed equally to this article.
To whom correspondence should be addressed.
2
Published, JLR Papers in Press, May 10, 2016
DOI 10.1194/jlr.M068692
e­mail: holthuis@uos.de
Copyright © 2016 by the American Society for Biochemistry and Molecular Biology, Inc.
This article is available online at http://www.jlr.org
Journal of Lipid Research Volume 57, 2016
1273

proteins, including ion channels, receptors, and mem­
enzyme. Next, the conserved His in the third exoplasmic
brane t rafficking m achinery ( 6, 7 ). S M a lso s erves a s r eceptor loop c arries o ut a n ucleophilic a ttack o n t he l ipid-phosphate
for viruses and pore-forming toxins (8, 9). Moreover, the ester bond, assisted by the conserved Asp. After formation
plasma membrane pool of SM acts as a reservoir of lipid
of a choline or ethanolamine phospho­His intermediate,
signaling m olecules, t he l iberation o f w hich i s c atalyzed b y DAG i s r eleased a nd r eplaced b y c eramide. T he c onserved
neutral, a lkaline, o r a cidic S Mases i n r esponse t o d iverse His near the second exoplasmic loop then acts as a nu­
stimuli (10). Ceramide generated by this pathway, along
with its downstream metabolites sphingosine and sphingo­
sine­1­phosphate, are critical regulators of cell survival,
proliferation, and migration (11).
cleophile to attack the primary hydroxyl group of ce­
ramide. This results in formation of SM or CPE, which are
released from the active site to allow a next round of ca­
talysis. While the importance of the catalytic triad residues
for SMS-catalyzed SM and CPE production has been con­
firmed by site­directed mutagenesis (18, 25), information
on the structural determinants that mediate head group
selectivity o f S MS e nzymes i s l acking. A s t hese e nzymes u se
SM synthesis is mediated by a phosphatidylcholine
(PC):ceramide choline­phosphotransferase or SM synthase
(SMS). This enzyme catalyzes the transfer of phosphocho­
line from PC onto ceramide, yielding SM and diacylglycerol
(
DAG). M ammals c ontain t wo S MS i soforms, n amely S MS1, the same reaction chemistry, their specificity for SM and/or
primarily responsible for de novo SM synthesis in the lumen
CPE p roduction i s m ost l ikely d efined b y d ifferences i n t he
of the trans Golgi, and SMS2, presumably serving a role in geometry of the phospholipid binding site. Identification of
regenerating SM from ceramides liberated by SMases on
the cell surface (12–15). Studies in mice revealed a role of
SMS1 and SMS2 in inflammation, atherosclerosis, and dia­
betes ( 15–17), s uggesting t hat t hese e nzymes r epresent r el­
residues involved in substrate recognition should not only
help to further elucidate the reaction mechanism of this
important class of enzymes, but also facilitate the develop­
ment of specific inhibitors to target the biological role of
evant p harmacological t argets. B esides b ulk a mounts o f S M, each e nzyme i ndividually.
mammals also produce small quantities of ceramide phos­
phoethanolamine (CPE), a widespread but poorly studied
SM analog. We and others previously showed that SMS2 is
In t his w ork, w e u se d omain s wapping a nd s ite-directed
mutagenesis o n e nzymes p roduced i n a l iposome-coupled
cell­free expression system to map structural elements re­
a b ifunctional e nzyme t hat p roduces b oth S M a nd C PE o n sponsible for head group selectivity of SMS family mem­
the c ell s urface, w hereas a c losely r elated e nzyme, s phingo­
myelin synthase­related protein (SMSr) SAMD8, acts as a
monofunctional CPE synthase in the lumen of the ER
bers. Our data reveal that enzyme specificity is strongly
influenced by single residues in close proximity of the
catalytic triad. We show that exchanging a limited number
of exoplasmic residues between SMS1 and SMSr is suffi­
cient to convert a conventional SMS into a monospecific
bulk producer of CPE. This enabled the establishment
of mammalian cells that produce CPE as their principal
phosphosphingolipid, thus creating an opportunity to elu­
cidate the biological significance of this enigmatic mem­
brane component.
(
18–21). Interestingly, SMSr is by far the best­conserved
SMS family member, with homologs in insects and various
marine o rganisms t hat l ack S M ( 18, 2 2). W hile t he b iologi­
cal relevance of SMS2­mediated CPE production remains to
be established, acute disruption of SMSr catalytic activity in
cultured mammalian and insect cells causes an accumula­
tion of ER ceramides, leading to a structural collapse of ER
exit sites and induction of mitochondrial apoptosis (18, 23).
These p henotypes a re s uppressed b y b locking d e n ovo c e­
ramide synthesis, stimulating ER export of ceramides or
targeting SMS1 to the ER, supporting a role of SMSr as
critical regulator of ER ceramide levels. However, ubiqui­
tous inactivation of SMSr catalytic activity in mice primarily
disrupted CPE biosynthesis in the brain without any obvious
impact on steady­state ceramide levels, cell integrity, or sur­
vival (20, 21). These contradictory outcomes could be due
to compensatory mechanisms that overcome a deregulation
of ER ceramides over time. The physiological relevance of
SMSr­mediated CPE production and the consequences of
its acute disruption in mammals remain to be established.
Analogous to members of the lipid phosphate phospha­
tase (LPP) superfamily, SMS enzymes likely contain a six-
times membrane­spanning core domain with an active site
comprising a conserved catalytic triad of two histidine (His)
residues and one aspartic acid (Asp) residue located in or
near the second and third exoplasmic loop (Fig. 1A) (12,
MATERIALS AND METHODS
Chemicals and reagents
Wheat germ phosphatidylinositol (PI), egg PC, and egg PE were
from Lipid Products UK. d­erythro­sphingosine and 1­palmitoyl­
2-{6-[(7-nitro-2-1,3-benzoxadiazol-4-yl)amino]hexanoyl}-sn­glycero­
3­phosphoethanolamine (NBD­PE) were from Avanti Polar Lipids.
NBD­C6­ceramide (NBD­Cer) and NBD­C6­SM (NBD­SM) were
from Molecular Probes. NBD­C6­CPE (NBD­CPE) was gener­
ously provided by Philippe Devaux (Institut de Biologie Physico­
Chimique, Paris). WEPRO2240 wheat germ extract was from
Cell-Free S ciences, A ccudenz f rom A ccurate C hemicals a nd [ 1,2-
14
C]ethanolamine hydrochloride (100 Ci/ml, 100–115 mCi/
mmol) from BioTrend. EndoH was from New England Biolabs.
Synthesis of clickable ceramide analog
As outlined in Fig. 3A, clickable ceramide (clickCer, 1) was
synthesized by the condensation of d­erythro­sphingosine and a
terminal acetylenic C15­fatty acid (7). The latter was prepared in
five steps starting from the condensation of Grignard reagent,
prepared from the commercially available chloride (2), and corre­
sponding d ecanoyl c hloride. T he k eto-function i n t he i ntermediate
24). T his s trongly s uggests t hat S MS-catalyzed S M a nd C PE
production follows an LPP­type reaction cycle, which starts
when a phospholipid head group donor, PC or phosphati­
dylethanolamine (PE), enters a single binding site in the
1
274
Journal of Lipid Research Volume 57, 2016

Fig. 1. The exoplasmic loops of SMSr and SMS1 harbor structural determinants of substrate selectivity. A:
Predicted m embrane t opology o f S MSr a nd S MS1. A ctive s ite r esidues a re m arked i n r ed. T he p ositions o f t wo
N-linked g lycosylation s ites i ntroduced b y s ite d irected m utagenesis i n h uman S MSr a re i ndicated. S AM, s terile

SMSr
motif. B: Anti­V5 immunoprecipitates prepared from HeLa cells transfected with V5­tagged human SMSr,
M180N G284N
, or SMSr
c onstructs w ere m ock t reated o r t reated w ith E ndoH a nd t hen s ubjected t o i mmunob­
lot a nalysis w ith a nti-V5 a ntibody. S MSr-G m arks t he m igration o f a n N ­glycosylated, EndoH­sensitive form of
M180N
G284N
SMSr that is produced exclusively in SMSr
­ and SMSr
­expressing cells. C: Sequence alignment of loop
c in SMS family members from vertebrates and fruit fly. The position of a residue critical for discriminating the
phospholipid head group donors PC and PE [Asp or glutamic acid (Glu)] is marked by an arrow. Database
accession numbers are as follows: human SMS1, BAD16809.1; chicken SMS1, ADY69193.1; frog SMS1,
NP_001008197.1; z ebrafish S MS1, N P_001071082.1; h uman S MS2, N P_689834.1; c hicken S MS2, X P_420492.1;
frog S MS2, A AH88568.1; z ebrafish S MS2, z gc:100911; h uman S MSr, Q 96LT4; c hicken S MSr, X P_426501.3; f rog
SMSr, Q28CJ3; zebrafish SMSr, zgc:162183; fruit fly SMSr, CG32380. D: TLC analysis of reaction products
E271D
formed when lysates of yeast strains expressing V5­tagged human SMS2 or SMS2
were incubated with
NBD­C6­ceramide (NBD­Cer). EV denotes yeast lysate from strain transfected with empty vector. SMS expres­
sion was verified by immunoblotting with anti­V5 antibody (bottom). Note that residue substitution E271D in
loop c converts SMS2 into a monofunctional SMS. E: TLC analysis of reaction products formed when lysates of
E343D
D327E
yeast strains expressing V5­tagged human SMSr, SMSr
, SMS1, SMS1
, or SMS1/r chimera in which loop
Lb1
Lbr
b was swapped (SMSr , SMS1 ) were incubated with NBD­Cer. SMS expression was verified by immunoblot­
ting with anti­V5 antibody (bottom). Note that residue substitution D327E combined with swapping loop b
D327E­Lbr
CPE
against that of SMSr converts SMS1 into a monofunctional CPE synthase (SMS1
shown in D and E are representative of two independent experiments.
or SMS1 ). Data
ketoester (3) was completely reduced using a sequence of three of the Me-ester in 6 gave click C15-fatty acid (7) with a yield of
steps, namely, reduction into the alcohol (4), tosylation, and fi­
nal reduction of the tosylate into the saturated fragment of 6. Si­
multaneous removal of the TMS­protecting group and hydrolysis
22% over five steps. Structure of clickCer (molecular weight
519.9) was confirmed by NMR spectroscopy ( H, C). The syn­
thesis of clickable analogs of SM (clickSM) and CPE (clickCPE)
1 13
Structural basis of sphingolipid synthase specificity
1275

will b e d escribed e lsewhere ( S. K orneev a nd J . H olthuis, u npub­
lished observations).
lipid mixture (egg PC/egg PE/wheat germ PI, 2:2:1 mol%) using
a miniextruder (Avanti Polar Lipids). In brief, 20 mol of total
lipid was dried under a flow of nitrogen to create a thin film. The
film was resuspended in 1 ml lipid rehydration buffer (25 mM
HEPES pH 7.5, 100 mM NaCl) by vigorous vortexing to create a
SMS expression constructs
For expression in yeast, open reading frames (ORFs) of hu­
man SMS enzymes were PCR amplified and cloned into the
pYES2.1/V5-His TOPO vector (Invitrogen) according to the
manufacturer's instructions. To create loop b swap constructs
20 m M l ipid s uspension. A fter s ix f reeze-thaw c ycles u sing l iquid
nitrogen and a 40°C water bath, liposomes with an average diam­
eter of 400 nm were obtained by extrusion of the lipid suspen­
sion through a 0.4 m track-etched polycarbonate membrane
Lb1
Lbr
SMSr and SMS1 , the sequences encoding Gly239-Gly278 in
SMS1 and Gly256-Gly294 in SMSr were exchanged by two con­
secutive fusion PCRs. Single amino acid substitutions were intro­
duced by site­directed mutagenesis using the megaprimer PCR
method (26). For expression in human HeLa and KBM7 cells, ORFs
of SMS enzymes were PCR amplified in-frame with a C-terminal
V5­His or hemagglutinin (HA) epitope and cloned into pcDNA3.1
or retroviral expression vector pLNX2 (Clontech). For cell-
free expression, ORFs of SMS enzymes were PCR amplified in-
frame with a C­terminal V5 epitope and cloned into the wheat
germ p EU-Flexi e xpression v ector ( kind g ift o f B rian G . F ox a nd
James D . B angs, U niversity o f W isconsin, M adison). A ll c onstructs
were sequence verified before use.
(
Whatman­Nuclepore). To obtain liposomes with an average di­
ameter of 100 nm, the 400 nm liposomes were extruded through
a 0.1 m membrane. Liposomes with a diameter of 30–50 nm
were obtained by tip sonication of the 100 nm liposomes on ice
until the suspension turned optically clear. The average diameter
of liposomes was confirmed by dynamic light scattering. Lipo­
some s uspensions w ere a liquoted, s nap f rozen i n l iquid n itrogen,
and stored at 80°C.
Cell-free expression of SMS enzymes
pEU­Flexi­SMS expression constructs were treated with Protein­
ase K to remove trace amounts of RNase, purified by phenol/chlo­
roform extraction and dissolved at 1 g/l in water. In vitro
transcription was carried out in 50 µl reaction volume containing
5 µg of DNA construct; 2 mM each of ATP, GTP, CTP, and UTP;
20 u nits o f S p6 R NA p olymerase; a nd 4 0 u nits o f R Nasin i n 1 00 m M
SMS glycosylation studies
M180N
HeLa cells were transfected with SMSr­V5­His, SMSr
­V5­
G284N
His, or SMSr
­V5­His constructs using Effectene Transfection
Reagent (Qiagen). After 24 h, cells were washed twice with PBS HEPES­KOH pH 7.8, 25 mM Mg­acetate, 2 mM spermidine, and
and l ysed i n L ysis B uffer ( 1% T riton X -100, 1 m M E DTA p H 8 .0, 10 mM DTT (28). After incubation at 37°C for 4 h, the reaction
150 mM NaCl, 20 mM Tris pH 7.5) containing 10 mM N­ethyl
mixture was centrifuged at 3400 g for 5 min at room temperature.
maleimide and protease inhibitor cocktail (1 µg/ml aprotinin, The supernatant, containing SMS mRNA, was collected to set up a
1
µg/ml leupeptin, 1 µg/ml pepstatin, 5 µg/ml antipain, 157 µg/ml
cell­free translation reaction. To this end, 20 µl of mRNA­containing
supernatant was used per 100 µl translation reaction containing
and the supernatants were incubated with Ni -NTA b eads ( Qiagen) 0.3 mM each of all 20 amino acids, 2 mM liposomes, 40 µg/ml cre­
for 2 h at 4°C while gently shaken. The beads were washed with atine kinase, 15 optical density (OD)260 WEPRO2240 wheat germ
Lysis Buffer, resuspended in 40 µl Glycoprotein Denaturing extract, 15 mM HEPES­KOH pH 7.8, 50 mM potassium acetate,
benzamidine). C ell l ysates w ere c entrifuged a t 7 ,000 g for 10 min,
2+
Buffer (0.5% SDS, 40 mM DTT) and incubated for 10 min at
1.25 mM Mg­acetate, 0.2 mM spermidine, 2 mM DTT, 0.6 mM ATP,
9
9°C. Next, 40 µl of 50 mM sodium acetate was added, and the 125 µ M G TP, 8 m M c reatine p hosphate, a nd 0 .0025% s odium a zide.
beads were split in two equal aliquots. After addition of 500 U of
EndoH to one aliquot, the beads were incubated at 37°C for 1 h
The reaction mixture was incubated for 4 h at 26°C (tube mode) or
transferred t o a 1 2 k Da m olecular w eight c ut-off d ialysis c up ( Nova­
and s ubjected t o S DS-PAGE a nd i mmunoblot a nalysis u sing m ouse gen) dipped in 5 ml reservoir buffer (same composition as transla­
anti­V5 antibody (Invitrogen) and HRP­conjugated goat anti­mouse
antibody (ThermoFischer).
tion reaction but omitting wheat germ extract, mRNA, creatine
kinase, and liposomes) and incubated for 16 h at 26°C (dialysis
mode). T ranslation r eactions w ere p rocessed f or S MS e nzyme a ctiv­
ity (see below) or subjected to quantitative immunoblotting using
mouse monoclonal anti­V5 antibody (Invitrogen Life Technologies)
and k nown a mounts o f a V 5-tagged 7 5 k Da r eference p rotein p uri­
fied from Escherichia coli (MBP­ceramide transfer protein­His6­V5).
Preparation of yeast lysates
Yeast strain IAY11 (MATa, ade2-1 trp1-1 can1-100 leu2-3,112
his3-11,15 ura3-52 ade3-853) was transformed with pYES2.1/
SMS­V5­His TOPO expression constructs and then grown in syn­
thetic medium containing 2% (w/v) galactose to early midloga­
rithmic phase. Cells were collected by centrifugation and washed
in ice­cold buffer R (15 mM KCl, 5 mM NaCl, 20 mM HEPES/
KOH pH 7.2). The wet cell pellet (2 g) was resuspended in a final
volume of 5 ml buffer R containing protease inhibitors (1 g/ml
aprotinin, 1 ug/ml leupeptin, 1 ug/ml pepstatin, 5 ug/ml anti­
Density flotation
Typically, 75 µl of translation reaction was mixed with 75 µl of
80% (w/v) Accudenz prepared in gradient buffer (25 mM
HEPES pH 7.5, 100 mM NaCl, 10% glycerol) and transferred to
the b ottom o f a 8 00 µ l 5 × 4 1 m m t hinwall U ltra-Clear t ube ( Beck­
pain, 1 m M b enzamidine, a nd 1 m M P MSF). C ells w ere l ysed b y man Coulter), carefully overlayed with 350 µl of 30% Accudenz
vigorous vortexing with 3 g glass beads at 4°C with intermittent
cooling on ice. A postnuclear supernatant was prepared by cen­
trifugation at 700 g for 10 min at 4°C. After addition of 0.11 vol of
and then with 100 µl of gradient buffer. Next, the samples were
centrifuged at 100,000 g in a MLS-50 rotor (Beckman Coulter)
for 4 h at 4°C. Fractions of 60 µl were collected from the top to
glycerol, lysates were aliquoted, snap-frozen in liquid nitrogen, the bottom of the gradient and processed for immunoblotting or
and stored at 80°C.
enzyme assays as indicated below. Gradient fractions were snap
frozen i n l iquid n itrogen a nd s tored a t  80°C.
Preparation of liposomes
Phospholipid stocks were prepared in chloroform-methanol
SMS activity assay with NBD-Cer
(9:1, v/v), briefly flushed with nitrogen gas, stored at 20°C in
Catalytic a ctivity o f S MS e nzymes e xpressed i n y east w ere a na­
brown glass vials, and periodically checked by TLC and iodine lyzed a s d escribed ( 19). T o a nalyze c ell-free p roduced S MS e nzymes
staining. Phospholipid concentrations were determined as de­
scribed (27). Unilamellar liposomes were prepared from a defined
for catalytic activity, 20–200 l of translation reaction mixture was
combined with buffer R containing protease inhibitors and
1
276
Journal of Lipid Research Volume 57, 2016

0
.5 mM N­ethylmaleinimide to a total volume of 400 l on ice.
cells (ATCC® CRL­3216™) grown in DMEM medium (Life Tech­
nologies) supplemented with 10% FBS (Sigma) were cotrans­
NBD-Cer w as a dded f rom a 2 m M e thanolic s tock t o a fi nal c oncen­
tration of 25 M. E nzyme r eactions w ere i ncubated a t 3 7°C f or 1 –2 h fected w ith p LNX2-SMS-HA expression c onstructs a nd p ackaging
with constant shaking. The reaction was stopped by adding 1 ml vectors (Clontech) using Lipofectamine 2000 (ThermoFisher).
methanol and 0.5 ml CHCl . Phase separation was induced by add­
The culture medium was changed 6 h post transfection. After 48 h,
supernatants were harvested, passed through a 0.45 m filter, and
the virus­containing filtrate was used to transduce KBM7/SMS1­
null cells. The cells were grown in Iscove modified Dulbecco me­
dium (IMDM; ThermoFisher) supplemented with 10% FBS and
3
ing 0.5 ml CHCl and 0.5 ml of 0.45% (w/v) NaCl. The lower
3
phase was evaporated under a stream of nitrogen, dissolved in
2
5 l CHCl /methanol (2:1), and then spotted at 120 nl/s on
3
NANO­ADAMANT HP­TLC plates (Macherey and Nagel) using a
CAMAG L inomat 5 T LC s ampler. T he T LC w as d eveloped fi rst i n 0.8 mg/ml geneticin (G418; ThermoFisher) for 10 days. Expres­
acetone, dried, and then in CHCl /methanol/25% NH OH
sion of HA-tagged SMS enzymes was confirmed by immunoblot­
50:25:6, v /v/v) u sing a C AMAG A DC2 T LC d eveloper. F luorescent ting using rat monoclonal anti­HA antibody.
3
4
(
lipids were visualized using a Typhoon FLA 9500 Biomolecular
Imager ( GE H ealthcare L ife S ciences) o perated i n C y2 fl uorescence Metabolic labeling
6
mode with 473 nm excitation laser, BPB1 filter, 50 m p ixel s ize,
and p hotomultiplier t ube ( PMT) v oltage s etting o f 2 90 V . N BD-
labeled SMS reaction products were quantified using known at 26°C for 24 h. KBM7 cells (2 × 10 ) were metabolically labeled
amounts of NBD­PE as reference.
Insect Sf21 cells (0.5 × 10 ) were metabolically labeled in 0.5 ml
insect X -press m edium ( Lonza) w ith 1  Ci of [ C]ethanolamine
6
14
in 5 m l I MDM m edium ( Gibco) c ontaining 1 0% F BS ( PAA L abo­
14
ratories) with 1 Ci of [ C]ethanolamine for 24 h at 37°C and
% CO . Lipids were extracted in CHCl /methanol/10 mM HAc
SMS activity assay with clickCer
5
2
3
Cell-free e xpression o f S MS e nzymes w as c arried o ut i n t he p res­
(5:22:1, v/v/v) and then processed according to Bligh and Dyer
ence of liposomes containing 2 mol% clickCer as above, except (31). H alf o f t he e xtract w as s ubjected t o m ild a lkaline h ydroly­
that D TT a nd s odium a zide w ere o mitted f rom t he t ranslation-
reaction buffer, as their presence was found to interfere with subse­
sis using sodium methoxide. Extracts were applied to TLC and
developed in CHCl /methanol/25% NH OH (50:25:6, v/v/v).
3
4
quent c lick r eactions. C ontrol e xperiments s howed t hat r emoval o f Radiolabeled lipids were detected by exposure to a Storage Phos­
these two compounds had no adverse effect on the yield or activity
of c ell-free p roduced S MS e nzymes. T ypically, 2 00  l of translation
reaction was combined with 200 l buffer R and incubated for 2 h
phor Screen (Fuji Photo Film), which was scanned using a Typhoon
FLA 9500 Biomolecular Imager operated at phosphoimaging
mode with a 635 nm laser, imaging processing filter, 50 m pixel
at 3 7°C w hile s haking. L ipids w ere e xtracted a s a bove, t ransferred size, a nd P MT s etting o f 6 00 V .
to Eppendorf Protein LoBind tubes, dried down in a speedvac,
and dissolved in 10 l CHCl . L ipid e xtracts w ere c lick r eacted w ith Lipid MS
3
the fl uorogenic d ye 3 -azido-7-hydroxycoumarin b y a ddition o f 6 4.5
l of a freshly prepared click-reaction mix containing 0.45 mM
-azido-7-hydroxycoumarin and 1.4 mM Cu(I)tetra(acetonitrile)
tetrafluoroborate in acetonitrile­ethanol (3:7, v/v) for 2.5 h at
5°C without shaking (29). The reaction was quenched by addi­
tion of 150 l methanol, dried down in a speedvac, dissolved in
CHCl /methanol (2:1, v/v), and applied on a TLC plate. The TLC
KBM7 cells grown for 48 h in IMDM medium supplemented
with 10% delipidated FBS (19) were subjected to total lipid ex­
traction according to Folch et al. (32) with 15:0 SM and 17:0 CPE
added as internal standards at the one­phase stage of extraction.
After evaporation of the solvent, the lipids were dissolved in

3
4
methanol/CHCl (2:1, v/v), the phosphate contents were deter­
3
3
mined (27), and the samples were stored at 20°C. To remove
glycerolipids, t he e xtract w as t aken t o d ryness u nder a s tream o f
nitrogen, NaOH (0.3 M final concentration) was added, and the
sample was incubated overnight at room temperature. After neu­
tralization w ith 0 .3 M H Cl, t he l ipids w ere e xtracted a s a bove a nd
dissolved in 40 l of LC/MS quality methanol. CPE and SM spe­
cies were separated on Acquity Ultra Performance LC system
equipped with an Acquity BEH­C18 1.0 × 150 mm column (Wa­
ters I nc.) a s i n B ickert e t a l. ( 20). T he c olumn e luent w as i nfused
to a Quattro Premier mass spectrometer (Waters) operated in
the positive ion mode, and the CPE and SM molecular species
were detected using selective reaction monitoring. The individ­
ual species were quantified from chromatograms using QuanLynx
was developed first in acetone and then in CHCl /methanol/
3
water/acetic acid (HAc) (65:25:4:1, v/v/v/v). Fluorescent lipids
were a nalyzed u sing a C hemiDoc X RS+ w ith U V-trans-illumination
(
detection settings for ethidium bromide, standard filter) and
Quantity O ne s oftware ( Bio-Rad).
Immunofluorescence microscopy
HeLa cells were seeded on glass coverslips and transfected with
pLNX2-SMS-HA constructs using Effectene transfection reagent
(Qiagen). A fter 2 4 h , c ells w ere fi xed w ith 4 % ( w/v) p araformalde­
hyde in PBS, quenched in 50 mM ammonium chloride, and per­
meabilized using PBS containing 0.1% (w/v) saponin and 0.2%
(w/v) BSA. Coverslips were incubated with rat anti­HA (Roche)
(
Waters Inc.) and Excel (Microsoft) software.
and mouse anti-GM130 (BD Bioscience) antibodies, washed, and
then incubated with donkey anti-rat Cy2 and donkey anti-mouse
Cy3 (Jackson ImmunoResearch). Coverslips were counterstained
with 4',6­diamidino­2­phenylindole (DAPI) and mounted using
Prolong G old A ntifade R eagent ( Molecular P robes). I mages w ere
captured using a DM5500 B epi­fluorescence microscope (Leica)
equipped with a PL­APO 63× (1.40 NA) oil immersion objective
and a SPOT Pursuit camera (Leica) with filter sets for differential
interference contrast, DAPI, green fluorescent protein, and
mCherry (Chroma Technology Corp.).
RESULTS AND DISCUSSION
Mapping structural determinants of SMS substrate
selectivity
While n o c rystal s tructure o f S MS e nzymes h as b een d e­
termined, protease protection experiments and a compar­
ative hydrophobicity analysis of SMS homologs predict a
six­times membrane­spanning core domain with a topol­
ogy as depicted in Fig. 1A (12, 24). This model positions
Retroviral transduction of KBM7 cells
Human myeloid leukemia KBM7-derived SMS1-null cells (30)
CPE
stably expressing HA­tagged SMS1 or SMS1 were created by ret­
roviral transduction. To this end, low-passage human HEK293T the catalytic triad of conserved His and Asp residues in or
Structural basis of sphingolipid synthase specificity
1277

near the second and third predicted exoplasmic loops
loops b and c; F ig. 1 A), h ence a t t he s ide o f t he m embrane minants in loop b mediate phospholipid head group do­
where S MS-catalyzed S M a nd C PE i s t hought t o o ccur ( 12, nor s electivity o f S MS e nzymes. T he C PE-producing v ariant
Asp a nd G lu r esidues i n l oop c along with structural deter­
(
D327E­Lbr
1
9). T o v alidate t his m odel, w e i ntroduced N -linked g lyco­
of SMS1, SMS1
, will henceforth be referred to as
CPE
sylation sites in the first and second putative exoplasmic
SMS1
.
loops of V5­tagged human SMSr (loops a and b; Fig. 1A),
M180N
G284N
yielding SMSr
­V5 and SMSr
­V5. Heterologous
Functional analysis of SMS enzymes in defined lipid
environments
expression of these constructs in HeLa cells followed by
immunoblot analysis in each case yielded a gel­shifted,
EndoH­sensitive protein that was absent in SMSr­V5­
expressing cells (Fig. 1B). These results indicate that the
loops connecting the first with the second and third with
the fourth putative membrane span of SMSr face the ER
lumen, consistent with the model presented in Fig. 1A.
As S MS-catalyzed S M a nd C PE p roduction h as b een r e­
ported t o t ake p lace i n t he e xoplasmic l eaflet o f t he m em­
Eukaryotic cells, from mammals to yeast, contain ATP-
fueled flippases that translocate PE and PC from the exo­
plasmic to the cytosolic leaflet of late secretory organelles
(33–36). These activities may influence SMS­mediated SM
and CPE production by limiting the amount of PC and PE
available for consumption by SMS enzymes. For instance,
the preferential synthesis of SM over CPE by SMS2 in the
plasma membrane could be due to a high PC/PE ratio in
brane (19), substrate selectivity of SMS enzymes likely the exoplasmic leaflet implemented by aminophospho­
involves residues in the exoplasmic loops where interac­
tions with the polar head groups of the phospholipid do­
nors PC and PE are expected to occur. An alignment of
SMS sequences revealed an Asp residue immediately prox­
imal to the catalytic His in the third exoplasmic loop (loop
c) that is invariably present in monofunctional SMS, SMS1,
from diverse species (Fig. 1C). CPE­producing family
members SMS2 and SMSr, on the other hand, contain a
lipid­specific flippases rather than determined by an in­
trinsic p roperty o f t he e nzyme ( 19). T his l ed u s t o e valuate
the head group selectivity of native and engineered SMS
enzymes in a defined lipid environment. To this end, we
adopted a recently developed system for the cell­free pro­
duction of functional polytopic membrane proteins
(28, 37). In this system, SMS mRNA is synthesized by in
vitro transcription and used for cell­free translation in a
Glu r esidue a t t his p osition. T he i nfluence o f t his r esidue wheat germ extract (Fig. 2A). Translation was performed
on substrate selectivity was investigated by site­directed
in the presence of unilamellar liposomes containing equi­
mutagenesis o f V 5-tagged h uman S MS e nzymes e xpressed molar amounts of PC and PE. The liposomes are thought
in budding yeast, an organism lacking endogenous SM to capture and stabilize the nascent membrane spans of
and CPE synthase activities. When lysates of yeast cells ex­
pressing SMS2 were incubated with the fluorescent ce­
ramide analog NBD­Cer, both NBD­SM and NBD­CPE
were formed, in line with previous studies indicating that
cell­free produced proteins, but the mechanistic princi­
ples involved remain to be established. For cell­free ex­
pression of human SMSr, SMS1, SMS2, and SMS1 , the
CPE
corresponding cDNAs were cloned downstream of the Sp6
this enzyme possesses dual SM and CPE synthase activity promoter in pEU Flexi­vector pFlx (28). To facilitate de­
Fig. 1 D) ( 19). S ubstitution o f t he G lu r esidue i n l oop c to
tection of the cell­free produced SMS proteins, a V5 epit­
Asp s electively a bolished S MS2-catalyzed C PE p roduction, ope was introduced at their C termini. As shown in Fig. 2B,
thus converting SMS2 into a monofunctional SMS. Ex­
cell­free translation of SMS mRNA in the presence of lipo­
changing Glu for Asp in loop c of SMSr also blocked somes in each case yielded a V5­tagged protein of the ex­
(
SMSr­mediated CPE production (Fig. 1E). Conversely, ex­
changing Asp for Glu in loop c of SMS1 yielded a bifunc­
tional enzyme with both SM and CPE synthase activity
pected s ize. N o s uch p rotein w as d etected w hen t ranslation
was performed in the absence of mRNA.
To address whether cell-free-produced SMS enzymes
(
Fig. 1 E). T hus, a G lu r esidue i mmediately a djacent t o t he are incorporated into liposomes, the translation reactions
catalytic His in loop c permits SMS­mediated CPE produc­
were loaded at the bottom of an Accudenz step gradient
tion while an Asp at this position confines the enzyme to and subjected to high­speed centrifugation. As shown in
produce only SM.
Fig. 2C, the bulk of cell-free produced SMS2 was recov­
ered from the top fractions of the density gradient, which
also c ontained t he b ulk o f e xternally a dded l iposomes. I n
contrast, SMS2 produced in the absence of liposomes was
mainly found in the bottom fractions of the gradient. Add­
ing > 2 mM of liposomes to the translation reaction caused
a gradual loss of SMS2 flotation (data not shown). In addi­
tion, flotation of SMS2 was best supported by liposomes
with a diameter of 100 nm. Reducing the diameter of
The observation that exchanging the Glu and Asp resi­
dues in loop c did not cause a complete switch in head
group selectivity implied that additional residues in the
exoplasmic domains of SMS enzymes contribute to sub­
strate specificity. Sequence alignments of the first and sec­
ond exoplasmic loops did not reveal any obvious clues on
the identity of these residues. However, swapping the sec­
ond exoplasmic loop (loop b) of SMSr for that of SMS1
abolished SMSr­mediated CPE production (SMSr ; Fig. liposomes to 30 nm or increasing it to 400 nm in each
Lb1
1
E). S trikingly, a n S MS1 c himera c arrying l oop b of SMSr
case diminished the protein’s capacity to float (data not
shown). Thus, cell-free produced SMS enzymes appear
Lbr
(
SMS1 ) lost the ability to produce SM but gained CPE
synthase a ctivity. M oreover, a n e xchange o f G lu f or A sp i n most efficiently incorporated into 100 nm liposomes.
Lbr
loop c of the SMS1 chimera dramatically increased its
Cell-free-produced SMS2 displayed dual enzymatic ac­
CPE synthase activity (Fig. 1E). These data suggest that the
tivity analogous to that observed for SMS2 expressed in
1
278
Journal of Lipid Research Volume 57, 2016

Fig. 2. Cell­free expression and functional analysis of human SMS family members. A: Schematic outline
of the wheat germ­based dialysis mode for cell­free translation of SMS­V5 mRNA. Unless indicated otherwise,
translation reactions were supplemented with liposomes containing equal amounts of phospholipid head
group donors PC and PE. B: Translation reactions with or without SMS­V5 mRNA were subjected to immu­
noblot analysis using anti-V5 antibody. Known amounts of a 75 kDa V5-tagged reference protein, pRef-V5,
were included to allow quantification of cell­free produced SMS­V5 protein. Note that translation reactions
with S MS-V5 m RNA i n e ach c ase y ielded a n i mmunoreactive p rotein o f t he e xpected s ize. C : S MS2-V5 m RNA
was translated in the absence or presence of 2 mM liposomes. Translation reactions were subjected to density
gradient fractionation and immunoblotting using anti­V5 antibody. D: SMS2­V5 mRNA was translated in
wheat germ extract in the presence of the indicated amounts of liposomes. Cell­free produced SMS2­V5 was
incubated w ith N BD-Cer, a nd r eaction p roducts w ere a nalyzed b y T LC. E xpression o f S MS2-V5 w as v erified
by immunoblotting using anti­V5 antibody (bottom). E: TLC analysis of reaction products formed when
SMS2­V5 produced in the presence of 2 mM liposomes was incubated with NBD­Cer for the indicated period
of t ime. S ensitivity o f d etection w as i ncreased 1 0-fold c ompared w ith D t o v isualize S MS2-mediated p roduc­
tion of NBD-CPE. Migration of an unidentified fluorescent lipid that was also present in reactions lacking
SMS2 i s m arked b y a n a sterisk. F : Q uantification o f r eaction p roducts f ormed b y c ell-free p roduced S MS2-V5
when incubated with NBD-Cer for the indicated period of time. G: Functional analysis of cell-free produced
CPE
SMS1, SMS1 , a nd S MS2. T LC a nalysis o f r eaction p roducts f ormed w hen t he i ndicated S MS e nzymes p roduced
in the presence of 2 mM liposomes were incubated with NBD­Cer. SMS expression was verified by immunoblot­
ting using anti­V5 antibody (bottom). Data shown are representative of three independent experiments.
Structural basis of sphingolipid synthase specificity
1279

yeast or human cells, with a strong preference for SM syn­
thesis over CPE synthesis (Fig. 2D­F). Thus, even though
the enzyme was incorporated into liposomes containing
equal amounts of the head group donors PC and PE, its
specific activity as SMS was nearly 100­fold higher than
Switching head group selectivity in mammalian
sphingolipid biosynthesis
Having demonstrated that swapping exoplasmic resi­
dues with SMSr suffices to convert SMS1 into a CPE syn­
thase, we next set out to generate mammalian cells
producing CPE rather than SM as bulk phosphosphingo­
lipid. F irst, w e c hecked w hether t he e xchange o f r esidues
necessary for switching head group selectivity of SMS1 had

1
that as CPE synthase, namely, 2.0 ± 0.9 E-03 s (n = 5)

1
versus 2.3 ± 1.5 E­05 s (n = 2). Omission of liposomes
from the translation reaction did not affect the overall
yield o f S MS2 b ut s trongly r educed i ts a bility t o s ynthesize any impact on its subcellular distribution. As shown in
sphingolipids (Fig. 2D). The finding that SMS2­mediated
sphingolipid production was not completely eliminated in
the absence of externally added liposomes suggested that
the wheat germ extract contained a residual pool of en­
dogenous phospholipid. Indeed, TLC analysis of the ex­
tract revealed the presence of trace amounts of PC (data
not shown). From these data, we conclude that the imbal­
ance in SM and CPE production by SMS2 is due to an
intrinsic property of the enzyme and that the liposome-
coupled wheat germ expression system offers a suitable
Fig. 4A, immunofluorescence microscopy of HeLa cells
transfected with HA­tagged versions of SMS1 and SMS1
CPE
revealed t hat b oth e nzymes l ocalize t o t he G olgi c omplex,
where bulk production of SM is known to occur (14). To
CPE
exchange Golgi-resident SMS1 for SMS1 , we used hu­
man myeloid leukemia KBM7-derived SMS1-null cells in
which the SMS1 gene was inactivated by insertional muta­
genesis (30). SMS1­null cells display only residual SMS ac­
tivity (15%) compared with control KBM7 cells (38).
This r esidual a ctivity i s l ikely d ue t o S MS2, w hich i s b arely
approach for dissecting the enzymatic characteristics of expressed in lymphoid cells (13, 20). SMS1­null cells stably
CPE
SMS family members in a controlled lipid environment.
expressing HA­tagged SMS1 or SMS1
were created by
retroviral t ransduction a nd e xpression o f t he e nzymes w as
verified by immunoblotting (Fig. 4B). As mammalian cells
normally produce only trace amounts of CPE (18), we ana­
lyzed KBM7 and SMS1-null cells expressing SMS1 or
CPE
SMS1
is a monofunctional CPE synthase
As shown in Fig. 2F, cell­free expression of SMS1 and
CPE
CPE
SMS1
in the presence of liposomes containing equal
SMS1 for their ability to synthesize CPE by metabolic
1
4
amounts of PC and PE yielded monofunctional enzymes labeling with [ C]ethanolamine. Labeling of Sf21 insect
with SM and CPE synthase activity, respectively. Moreover,
cells s erved a s c ontrol a s t hese c ells t ypically p roduce b ulk
these enzymes displayed similar specific activities (SMS1: amounts of CPE owing to an SMSr­unrelated CPE synthase
1
CPE
1
4
.2 ± 1.8 E­04 s , n = 3 ; S MS1 : 3.3 ± 1.7 E­04 s , n = 2).
These results corroborate the notion that phospholipid
unique to insects (22). Labeling KBM7, SMS1­null, or
14
SMS1­null cells expressing SMS1 with [ C]ethanolamine
head group selectivity of SMS family members is primarily
for 24 h did not yield any detectable levels of radioactive
controlled by residues in the enzyme’s second and third CPE. In contrast, labeled Sf21 and SMS1 ­expressing
CPE
exoplasmic loops (i.e., loops b and c, respectively). Cell­
free produced SMSr, on the other hand, displayed hardly
any CPE synthase activity. Whether this reflects a low in­
trinsic specific activity of the enzyme or failure of the
wheat germ expression system to generate catalytically ac­
tive SMSr remains to be established.
SMS1­null cells contained substantial amounts of radioac­
tive CPE (Fig. 4C), suggesting that SMS1 m ediates b ulk
CPE
production of CPE.
CPE
To assess the relative contribution of SMS1 and SMS1
to sphingolipid content at steady state, we next deter­
mined the lipid composition of KBM7 and SMS1­null cells
CPE
So far, catalytic activity of cell­free produced SMS family
expressing SMS1 or SMS1
using LC/MS/MS. As re­
members was determined by monitoring the enzymatic ported previously, removal of SMS1 resulted in a 6­fold
CPE
conversion of the fluorescent short­chain ceramide analog
NBD­Cer, which was added to SMS­containing proteolipo­
drop in SM levels (38). Expression of SMS1
caused a
40-fold increase in CPE levels (43 ± 7, n = 2), making
somes f rom a n e thanolic s tock. T o e xclude t he p ossibility CPE the most abundant phosphosphingolipid in SMS1­
that the bulky NBD moiety of this analog interferes with null cells (Fig. 4D). This increase concerned both short­
SMS substrate specificity and provide further proof that
the c ell-free p roduced e nzymes a re c orrectly e mbedded i n SM species synthesized by SMS1-expressing cells in terms
the lipid bilayer, we synthesized a ceramide analog carry­ of t heir c eramide b ackbone s tructures ( Fig. 4 E). N o s uch
ing a C 15-acyl c hain w ith a c lickable t erminal a lkyne g roup increase in CPE levels was observed in SMSr­overexpressing
and long­chain CPE species, which largely overlapped with
CPE
that closely mimics natural C16-ceramide (clickCer; Fig.
cells (18 and unpublished observations). Swapping SMS1
3
A). L iposomes c ontaining 2 m ol% c lickCer w ere u sed a s against SMS1 in KBM7 cells had no significant impact
membrane protein acceptor vesicles during cell­free pro­
on the levels of other sphingolipid (glucosylceramides,
duction of SMS enzymes. SMS-mediated conversion of ceramides) or phospholipid classes (PC, PE, PI, phosphati­
clickCer i nto c lickSM a nd c lickCPE w as m onitored b y T LC dylserine; u npublished o bservations). I mportantly, t hese r e­
CPE
analysis a fter c lick r eacting S MS p roteoliposomes w ith t he sults indicate that SMS1 , contrary to SMSr, functions
fluorogenic d ye 3 -azido-7-hydroxycoumarin [ Fig. 3 B; G ae­
as b ona fi de b ulk C PE s ynthase. A c omprehensive a nalysis
bler et al. 2013 (29)]. As shown in Fig. 3C, this approach of the consequences of substituting CPE for SM as bulk
CPE
confirmed that SMS1 and SMS1 act as monofunctional
SM and CPE synthases, respectively.
phosphosphingolipid on mammalian cell physiology will
be the subject of a separate study.
1
280
Journal of Lipid Research Volume 57, 2016

Fig. 3. Tracing c atalytic a ctivity o f c ell-free p roduced S MS e nzymes b y c lick c hemistry. A : A c lickable c eramide a nalog, c lickCer ( 1), w as
synthesized b y c ondensation o f d­erythro-sphingosine a nd a C 15-fatty a cid c arrying a t erminal a lkyne g roup ( 7). T he l atter w as s ynthesized
in five steps, as detailed in Materials and Methods. Reagents and conditions: a, Mg, Et O, C H O Cl, room temperature; b, NaBH ,
2
11 19
3
4
methanol-water, r oom t emperature; c , TsCl, Et N, 4 -dimethylaminopyridine, r oom t emperature; d , NaBH , D MSO, 7 5°C; e , KOH, metha­
3
4
nol, 5 0°C; f , s phingosine, 1 -ethyl-3-(3-dimethylaminopropyl)carbodiimide, h ydroxybenzotriazole, r oom t emperature. B : S chematic o utline
of t he c lick-chemistry-based S MS a ssay. C lickCer w as i ncorporated a s S MS s ubstrate i n l iposomes p resent d uring t he c ell-free t ranslation o f
SMS-V5 m RNA. A fter l ipid e xtraction, t he a lkyne m oiety w as c lick r eacted w ith t he fl uorogenic d ye 3 -azido-7-hydroxycoumarin t o y ield fl uo­
CPE
rescently l abeled l ipids. T he s cheme w as a dapted f rom G aebler e t a l. ( 28). C : V 5-tagged S MS1, S MS1 , SMS2, and SMSr were produced
cell free in the presence of liposomes containing 2 mol% clickCer. After lipid extraction, SMS reaction products were click reacted with
3-azido-7-hydroxycoumarin, separated by TLC, and analyzed by fluorescence detection. SMS expression was verified by immunoblotting
using anti­V5 antibody (bottom). Data shown are representative of two independent experiments.
Structural model of SMS substrate selectivity
amounts of the phospholipid head group donors PC and
PE. In contrast, we suggest that loop b of SMSr is more
tightly folded over the active site, thus excluding entrance
of a phosphocholine head group. This idea is supported by
our finding that exchanging loop b between SMS1 and
SMSr abolishes SMS1­mediated SM production (Fig. 1D).
In SMS-catalyzed SM/CPE production, transfer of the
PC/PE head group onto ceramide is thought to proceed via
formation of a choline/ethanolamine phospho­His inter­
mediate (24). Here, we defined two critical determinants of
head group selectivity among SMS family members, namely,
i) a n A sp/Glu r esidue a t t he - 1 p osition o f t he c atalytic H is Our m odel a lso p redicts t hat t he G lu r esidue j ust u pstream
that binds the phospholipid head group in the third exo­
plasmic loop, loop c; and ii) the second exoplasmic loop,
loop b, proximal to the other catalytic His (Fig. 5A). As out­
lined in the model depicted in Fig. 5B, we envision that
loop b in SMS1 and SMS2 has a spacious fold that allows
of the catalytic His in loop c of SMSr and SMS2 has a critical
role in stabilizing a phosphoethanolamine head group in
the active site, quite independently of the proposed loop b
interactions. Indeed, exchanging this Glu residue for Asp
blocked S MSr-mediated C PE p roduction w hile e xchanging
accommodation o f a b ulky p hosphocholine h ead g roup i n Asp f or G lu a t t he c orresponding p osition i n S MS1 y ielded
the active site, presumably via hydrophobic interactions
with the methyl groups. A phosphoethanolamine head
group l acks t his p ossibility o f i nteraction w ith l oop b , which
an S MS2-like e nzyme w ith b oth S M a nd C PE s ynthase a ctiv­
ity (Fig. 1D).
To our knowledge, the only proposed structure for an
would e xplain w hy S MS2 p referentially s ynthesizes S M o ver SMS e nzyme i s b ased o n a h uman S MS1 h omology m odel
CPE even when the enzyme has direct access to equal created by Zhang et al. (39). In their model, loop b is
Structural basis of sphingolipid synthase specificity
1281

Fig. 4. Switching head group selectivity in mammalian sphingolipid biosynthesis. A: HeLa cells were transfected with empty vector (EV),
CPE
SMS1­HA, or SMS1 -HA, fi xed a nd t hen d ouble l abeled w ith a ntibodies a gainst t he H A-tag ( green) a nd G olgi m arker G M130 ( red). N ote
that b oth S MS e nzymes l ocalize t o t he G olgi. S cale b ar, 1 0  m. B: Immunoblots of KBM7­derived WT and SMS1­null cells (SMS1KO) trans­
CPE
fected with EV, SMS1­HA, or SMS1 ­HA were stained with antibodies against the HA­tag and ­actin. C: KBM7­derived WT and SMS1KO
CPE
14
cells transfected with EV, SMS1­HA, or SMS1 ­HA were metabolically labeled with [ C]ethanolamine f or 2 4 h a nd t hen s ubjected t o l ipid
extraction, T LC a nalysis, a nd a utoradiography. I n s ome e xtracts, g lycerolipids w ere d eacylated b y m ild a lkaline h ydrolysis ( hydr: + ) p rior
CPE
to TLC analysis. Note that only SMS1 -expressing c ells p roduce a r adiolabeled l ipid r esistant t o a lkaline h ydrolysis a nd w ith a n R f o f C PE.
Metabolic labeling of insect Sf21 cells, which produce bulk amounts of CPE, served as control. D: SM and CPE levels in lipid extracts of
CPE
KBM7­derived WT and SMS1KO cells expressing SMS1­HA or SMS1 ­HA were determined by LC/MS/MS and expressed as mol% of total
CPE
phospholipid a nalyzed. E : L evels o f S M a nd C PE s pecies i n K BM7-SMS1KO c ells e xpressing S MS1-HA o r S MS1 ­HA were determined as
in D. Data shown in D and E are representative of two independent experiments.
folded on top of part of the active site, compatible with our
model. The model by Zhang et al. also shows additional
interactions of Phe173 and Phe177 from loop a with the
phosphocholine head group of an SM molecule, which
are suggested to contribute to SM binding. This leaves the
alternative possibility that loop b of SMSr in the context
not produce SM) while Phe177 is not strictly necessary for
binding of phosphocholine head group as SMS2 is a good
SM producer without it.
In our model, the point on the reaction coordinate at
which the head group “selectivity filter” is applied is not de­
fined. The way the model is drawn, for reasons of simplicity,
of an SMS1 enzyme blocks the interaction with the two assumes the coordination/exclusion of a choline/ethanol­
Phe residues, thereby abolishing SM synthesis. However,
Phe173 is conserved in SMS1, SMS2, and SMSr ,whereas
Phe177 only occurs in SMS1. Therefore, these Phe resi­
amine phospho­His intermediate. This would imply that PC
and PE are constantly probing the active site, but that the
intermediate is formed only when the appropriate phos­
dues unlikely contribute to the head group selectivity of pholipid head group donor has entered. Alternatively, ex­
SMS e nzymes a s P he173 i s c onserved i n S MSr ( which d oes clusion already takes place at the level of phospholipid
1
282
Journal of Lipid Research Volume 57, 2016

Fig. 5. Structural model of SMS head group selectivity. A: Membrane topology and structural elements
that contribute to substrate specificity of SMS enzymes. The invariant His and Asp residues that form the
catalytic triad are marked in red. Head group selectivity o f SMS enzymes is determined b y a single r esidue
adjacent to the catalytic His in exoplasmic loop c, t hat i s, A sp i n S MS1 ( marked i n g reen) o r G lu i n S MS2 a nd
SMSr ( marked i n y ellow), a long w ith s tructural i nformation i n l oop b ( marked i n b lue). S AM, s terile  motif.
B: Model explaining how structural elements in loops b and c o f S MS e nzymes c ooperate t o d etermine h ead
group selectivity. See text for details.
entry, that is, depending on the protein's selectivity, the ap­
propriate phospholipid gains access to the active site. The
difference between binding energies of SMS1 to PC and PE
calculated b y Z hang e t a l. ( 90 k J/mol) s uggests t hat t he l at­
ter mechanism applies (39). However, photoaffinity label­
It is of interest to note that in a study on the substrate
selectivity of an SMS­related ceramide phosphoinositol
(CPI) synthase family from trypanosomes (42), the resi­
due at the ­1 position of the catalytic His that binds the
phospholipid h ead g roup a lso h ad a m arked e ffect o n t he
ing experiments with diazirine-containing phospholipid various products formed by these enzymes (CPI, CPE,
analogs ( 40) c ombined w ith a t horough a nalysis o f e nzyme SM). The residues found at this position are more diver­
kinetics will be necessary to distinguish the above two gent ( Ser, P he) t han t hose i n S MS e nzymes ( Asp, G lu), i n
mechanisms. T he l atter w ill b e a c hallenging t ask [ see e .g., agreement with the increased divergence of the products
Mina et al. (41)], because for a bifunctional enzyme like (the head group of CPI has a net negative charge and is
SMS2, it involves systematic variation of the concentra­
tions of three different substrates (i.e., ceramide, PC, and
PE) when only the forward reactions are considered.
bulkier than that of SM and CPE). However, the authors
did not wish to speculate on the specific molecular inter­
actions that could cause the change in selectivity.
Structural basis of sphingolipid synthase specificity
1283

CONCLUDING R EMARKS
7. Contreras, F -X., A . M . E rnst, P . H aberkant, P . B jörkholm, E . L indahl,
B. Gönen, C. Tischer, A. Elofsson, G. von Heijne, C. Thiele, et al.
2
012. Molecular recognition of a single sphingolipid species by a
In t his w ork, w e d efined s tructural e lements n ear t he a c­
tive site of SMS enzymes that impart catalytic specificity.
Residues in the second and third exoplasmic loops proxi­
mal to the conserved catalytic triad turned out to be key
determinants of head group selectivity. Swapping these resi­
dues with SMSr allowed conversion of the Golgi-resident
SMS, S MS1, i nto a b ulk p roducer o f C PE a nd t he e stablish­
ment o f a m ammalian c ell l ine t hat s ynthesizes C PE r ather
than SM as the dominant phosphosphingolipid. CPE nor­
mally represents only a minor fraction of the mammalian
sphingolipid pool, with steady state levels 300­ to 1,500­fold
below those of SM (20). In contrast, CPE is the dominant
phosphosphingolipid in most insects, including Drosophila
protein’s transmembrane domain. Nature. 481: 5 25–529.
8. Miller, M. E., S. Adhikary, A. A. Kolokoltsov, and R. A. Davey.
2012. Ebolavirus requires acid sphingomyelinase activity and
plasma membrane sphingomyelin for infection. J. Virol. 86:
7
473–7483.
9. Yamaji-Hasegawa, A., F. Hullin-Matsuda, P. Greimel, and T.
Kobayashi. 2016. Pore­forming toxins: properties, diversity, and
uses as tools to image sphingomyelin and ceramide phosphoetha­
nolamine. Biochim. Biophys. Acta. 10.1016/j.bbamem.2015.10.012.
0. Adada, M., C. Luberto, and D. Canals. 2016. Inhibitors of the
sphingomyelin cycle: sphingomyelin synthases and sphingomyelin­
ases. Chem. Phys. Lipids. 10.1016/j.chemphyslip.2015.07.008.
1 1. Maceyka, M ., a nd S . S piegel. 2 014. S phingolipid m etabolites i n i n­
flammatory disease. Nature. 510: 58–67.
1 2. Huitema, K ., J . v an d en D ikkenberg, J . F . H . M . B rouwers, a nd J . C .
1
M. Holthuis. 2004. Identification of a family of animal sphingomy­
elin synthases. EMBO J. 23: 33–44.
(
22, 43). Unlike SM, CPE does not interact favorably with
cholesterol despite of its hydrogen­bonding properties
44). This may explain why insects are auxotroph for sterols
2
(
elin synthesis and cell growth in sphingomyelin synthase­defective
lymphoid cells. J. Biol. Chem. 279: 1 8688–18693.
and have a more relaxed structural requirement for sterols
in their membranes than mammals, which appear highly
adapted to the use of cholesterol (45, 46). Substitution of
SM for CPE may therefore have a profound impact on cho­
lesterol h omeostasis i n m ammalian c ells. O ur c urrent w ork
provides an ideal basis to test this prediction.
1 4. Tafesse, F . G ., K . H uitema, M . H ermansson, S . v an d er P oel, J . v an
den D ikkenberg, A . U phoff, P . S omerharju, a nd J . C . M . H olthuis.
2007. Both sphingomyelin synthases SMS1 and SMS2 are required
for sphingomyelin homeostasis and growth in human HeLa cells. J.
Biol. Chem. 282: 17537–17547.
1 5. Mitsutake, S ., K . Z ama, H . Y okota, T . Y oshida, M . T anaka, M . M itsui,
M. I kawa, M . O kabe, Y . T anaka, T . Y amashita, e t a l. 2 011. D ynamic
modification of sphingomyelin in lipid microdomains controls de­
velopment of obesity, fatty liver, and type 2 diabetes. J. Biol. Chem.
At present, the physiological relevance of CPE biosynthe­
sis in mammals is unclear (20, 21). We previously showed
that a cute d isruption o f S MSr-catalyzed C PE p roduction i n
cultured cells causes a deregulation of ER ceramides and
induction of mitochondrial apoptosis (18, 23). Whether
CPE acts as a signaling lipid in SMSr­mediated ceramide
homeostasis remains to be established. To address this
possibility, our current efforts focus on manipulating ER­
resident CPE pools by targeting active­site­engineered
2
86: 28544–28555.
16. Liu, J., C. Huan, M. Chakraborty, H. Zhang, D. Lu, M-S. Kuo, G.
Cao, and X-C. Jiang. 2009. Macrophage sphingomyelin synthase 2
deficiency decreases atherosclerosis in mice. Circ. Res. 105: 2 95–303.
1 7. Li, Z ., H . Z hang, J . L iu, C -P. L iang, Y . L i, G . T eitelman, T . B eyer, H .
H. B ui, D . A . P eake, Y . Z hang, e t a l. 2 011. R educing p lasma m em­
brane sphingomyelin increases insulin sensitivity. Mol. Cell. Biol. 31:
4
205–4218.
J. F. H. M. Brouwers, P. Somerharju, C. Rabouille, and J. C. M.
Holthuis. 2 009. S phingomyelin s ynthase-related p rotein S MSr c on­
trols ceramide homeostasis in the ER. J. Cell Biol. 185: 1013–1027.
CPE
SMS1 to the ER. The establishment of SMS variants with
altered head group selectivity not only provides fresh in­
sight into the reaction mechanism underlying SM and CPE
biosynthesis, but should also facilitate the development of
isoenzyme-specific i nhibitors.
1 9. Ternes, P., J. F. H. M. Brouwers, J. van den Dikkenberg, and J. C.
M. H olthuis. 2 009. S phingomyelin s ynthase S MS2 d isplays d ual a c­
tivity as ceramide phosphoethanolamine synthase. J. Lipid Res. 50:
2270–2277.
2 0. Bickert, A ., C . G inkel, M . K ol, K . v om D orp, H . J astrow, J . D egen,
R. L . J acobs, D . E . V ance, E . W interhager, X -C. J iang, e t a l. 2 015.
Functional c haracterization o f e nzymes c atalyzing c eramide p hos­
phoethanolamine biosynthesis in mice. J. Lipid Res. 56: 821–835.
The a uthors t hank T arja G rundström f or t echnical a ssistance.
2 1. Ding, T., I. Kabir, Y. Li, C. Lou, A. Yazdanyar, J. Xu, J. Dong, H.
Zhou, T . P ark, M . B outjdir, e t a l. 2 015. A ll m embers i n t he s phingo­
myelin synthase gene family have ceramide phosphoethanolamine
synthase activity. J. Lipid Res. 56: 537–545.
REFERENCES
2 2. Vacaru, A. M., J. van den Dikkenberg, P. Ternes, and J. C. M.
1. Slotte, J. P. 2013. Biological functions of sphingomyelins. Prog.
Lipid Res. 52: 424–437.
Holthuis. 2013. Ceramide phosphoethanolamine biosynthesis in
Drosophila is mediated by a unique ethanolamine phosphotrans­
ferase i n t he G olgi l umen. J . Biol. Chem. 288: 11520–11530.
2. Slotte, J . P ., a nd E . L . B ierman. 1 988. D epletion o f p lasma-membrane
sphingomyelin rapidly alters the distribution of cholesterol between
plasma membranes and intracellular cholesterol pools in cultured
fibroblasts. Biochem. J. 250: 653–658.
2 3. Tafesse, F. G., A. M. Vacaru, E. F. Bosma, M. Hermansson, A.
Jain, A. Hilderink, P. Somerharju, and J. C. M. Holthuis. 2014.
Sphingomyelin synthase­related protein SMSr is a suppressor
of ceramide­induced mitochondrial apoptosis. J. Cell Sci. 127:
445–454.
2 4. Tafesse, F . G ., P . T ernes, a nd J . C . M . H olthuis. 2 006. T he m ultigenic
sphingomyelin synthase family. J. Biol. Chem. 281: 2 9421–29425.
2 5. Yeang, C., S. Varshney, R. Wang, Y. Zhang, D. Ye, and X-C. Jiang.
2008. The domain responsible for sphingomyelin synthase (SMS)
activity. Biochim. Biophys. Acta. 1781: 610–617.
3. Gupta, A . K ., a nd H . R udney. 1 991. P lasma m embrane s phingomy­
elin a nd t he r egulation o f H MG-CoA r eductase a ctivity a nd c holes­
terol biosynthesis in cell cultures. J. Lipid Res. 32: 125–136.
4. Sharpe, H . J ., T . J . S tevens, a nd S . M unro. 2 010. A c omprehensive
comparison of transmembrane domains reveals organelle­specific
properties. Cell. 142: 1 58–169.
5. Quiroga, R., A. Trenchi, A. González Montoro, J. Valdez Taubas,
and H. J. F. Maccioni. 2013. Short transmembrane domains with
high­volume exoplasmic halves determine retention of type II mem­
brane p roteins i n t he G olgi c omplex. J . Cell Sci. 126: 5 344–5349.
2 6. Orr-Weaver, T . L ., J . W . S zostak, a nd R . J . R othstein. 1 983. G enetic
applications of yeast transformation with linear and gapped plas­
mids. Methods Enzymol. 101: 228–245.
6. Milescu, M ., F . B osmans, S . L ee, A . A . A labi, J . I . K im, a nd K . J . S wartz.
009. Interactions between lipids and voltage sensor paddles de­
tected with tarantula toxins. Nat. Struct. Mol. Biol. 16: 1080–1085.
2 7. Bartlett, E . M ., a nd D . H . L ewis. 1 970. S pectrophotometric d etermi­
nation of phosphate esters in the presence and absence of ortho­
phosphate. Anal. Biochem. 36: 1 59–167.
2
1
284
Journal of Lipid Research Volume 57, 2016

28. Goren, M . A ., A . N ozawa, S -I. M akino, R . L . W robel, a nd B . G . F ox.
009. C ell-free t ranslation o f i ntegral m embrane p roteins i nto u nil­
amelar liposomes. Methods Enzymol. 463: 647–673.
29. Gaebler, A ., R . M ilan, L . S traub, D . H oelper, L . K uerschner, a nd C .
3 8. Tafesse, F. G., S. Sanyal, J. Ashour, C. P. Guimaraes, M.
Hermansson, P. Somerharju, and H. L. Ploegh. 2013. Intact sphin­
gomyelin biosynthetic pathway is essential for intracellular trans­
port of influenza virus glycoproteins. Proc. Natl. Acad. Sci. USA.
110: 6406–6411.
2
Thiele. 2 013. A lkyne l ipids a s s ubstrates f or c lick c hemistry-based i n
vitro e nzymatic a ssays. J . Lipid Res. 54: 2 282–2290.
3 0. Carette, J. E., C. P. Guimaraes, M. Varadarajan, A. S. Park, I.
3 9. Zhang, Y., F. Lin, X. Deng, R. Wang, and D. Ye. 2011. Molecular
modeling of three­dimensional structure of human sphingomyelin
synthase. Chin. J. Chem. 29: 1567–1575.
4 0. Haberkant, P ., R . R aijmakers, M . W ildwater, T . S achsenheimer, B .
Brugger, K. Maeda, M. Houweling, A-C. Gavin, C. Schultz, G. van
Meer, e t a l. 2 013. I n v ivo p rofiling a nd v isualization o f c ellular p ro­
tein­lipid interactions using bifunctional fatty acids. Angew. Chem.
Int. Ed. Engl. 52: 4033–4038.
Wuethrich, A. Godarova, M. Kotecki, B. H. Cochran, E. Spooner,
H. L. Ploegh, et al. 2009. Haploid genetic screens in human cells
identify host factors used by pathogens. Science. 326: 1231–1235.
31. Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid
extraction and purification. Can. J. Biochem. Physiol. 37: 9 11–917.
32. Folch, J ., M . L ees, a nd G . H . S loane S tanley. 1 957. A s imple m ethod
for the isolation and purification of total lipides from animal tis­
sues. J. Biol. Chem. 226: 4 97–509.
3 3. Pomorski, T ., R . L ombardi, H . R iezman, P . F . D evaux, G . v an M eer,
4 1. Mina, J. G., J. A. Mosely, H. Z. Ali, H. Shams-Eldin, R. T.
Schwartz, P. G. Steel, and P. W. Denny. 2010. A plate-based as­
say system for analyses and screening of the Leishmania major
inositol phosphorylceramide synthase. Int. J. Biochem. Cell Biol.
42: 1553–1561.
4 2. Goren, M. A., B. G. Fox, and J. D. Bangs. 2011. Amino acid deter­
minants of substrate selectivity in the Trypanosoma brucei sphingo­
lipid synthase family. Biochemistry. 50: 8853–8861.
and J. C. M. Holthuis. 2003. Drs2p-related P-type ATPases Dnf1p
and Dnf2p are required for phospholipid translocation across the
yeast plasma membrane and serve a role in endocytosis. Mol. Biol.
Cell. 14: 1240–1254.
3 4. Alder-Baerens, N ., Q . L isman, L . L uong, T . P omorski, a nd J . C . M .
Holthuis. 2006. Loss of P4 ATPases Drs2p and Dnf3p disrupts ami­
nophospholipid t ransport a nd a symmetry i n y east p ost-Golgi s ecre­
tory vesicles. Mol. Biol. Cell. 17: 1632–1642.
4 3. Carvalho, M., J. L. Sampaio, W. Palm, M. Brankatschk, S. Eaton,
and A. Shevchenko. 2012. Effects of diet and development on the
Drosophila lipidome. Mol. Syst. Biol. 8: 600.
44. Térová, B ., R . H eczko, a nd J . P . S lotte. 2 005. O n t he i mportance o f
the phosphocholine methyl groups for sphingomyelin/cholesterol
interactions in membranes: a study with ceramide phosphoetha­
nolamine. Biophys. J. 88: 2 661–2669.
35. Takatsu, H., G. Tanaka, K. Segawa, J. Suzuki, S. Nagata, K.
Nakayama, and H-W. Shin. 2014. Phospholipid flippase activities
and substrate specificities of human type IV P­type ATPases local­
ized t o t he p lasma m embrane. J . Biol. Chem. 289: 33543–33556.
36. Panatala, R ., H . H ennrich, a nd J . C . M . H olthuis. 2 015. I nner w ork­
ings and biological impact of phospholipid flippases. J. Cell Sci. 128:
021–2032.
37. Sevova, E. S., M. A. Goren, K. J. Schwartz, F-F. Hsu, J. Turk, B. G.
4 5. Yu, L., K. von Bergmann, D. Lutjohann, H. H. Hobbs, and J. C.
Cohen. 2004. Selective sterol accumulation in ABCG5/ABCG8-
deficient mice. J. Lipid Res. 45: 301–307.
4 6. Carvalho, M., D. Schwudke, J. L. Sampaio, W. Palm, I. Riezman,
G. Dey, G. D. Gupta, S. Mayor, H. Riezman, A. Shevchenko, et al.
2010. Survival strategies of a sterol auxotroph. Development. 137:
3675–3685.
2
Fox, a nd J . D . B angs. 2 010. C ell-free s ynthesis a nd f unctional c har­
acterization o f s phingolipid s ynthases f rom p arasitic t rypanosoma­
tid p rotozoa. J . Biol. Chem. 285: 20580–20587.
Structural basis of sphingolipid synthase specificity
1285