Sphingosylphosphocholine is a naturally occurring lipid mediator in blood plasma: a possible role in regulating cardiac function via sphingolipid receptors.

Abstract

Blood plasma and serum contain factors that activate inwardly rectifying GIRK1/GIRK4 K+ channels in atrial myocytes via one or more non-atropine-sensitive receptors coupled to pertussis-toxin-sensitive G-proteins. This channel is also the target of muscarinic M(2) receptors activated by the physiological release of acetylcholine from parasympathetic nerve endings. By using a combination of HPLC and TLC techniques with matrix-assisted laser desorption ionization-time-of-flight MS, we purified and identified sphingosine 1-phosphate (SPP) and sphingosylphosphocholine (SPC) as the plasma and serum factors responsible for activating the inwardly rectifying K+ channel (I(K)). With the use of MS the concentration of SPC was estimated at 50 nM in plasma and 130 nM in serum; those concentrations exceeded the 1.5 nM EC(50) measured in guinea-pig atrial myocytes. With the use of reverse-transcriptase-mediated PCR and/or Western blot analysis, we detected Edg1, Edg3, Edg5 and Edg8 as well as OGR1 sphingolipid receptor transcripts and/or proteins. In perfused guinea-pig hearts, SPC exerted a negative chronotropic effect with a threshold concentration of 1 microM. SPC was completely removed after perfusion through the coronary circulation at a concentration of 10 microM. On the basis of their constitutive presence in plasma, the expression of specific receptors, and a mechanism of ligand inactivation, we propose that SPP and SPC might have a physiologically relevant role in the regulation of the heart.

Full text

Page 1

Biochem. J. (2001) 355, 189–197 (Printed in Great Britain)
189
Sphingosylphosphocholine is a naturally occurring lipid mediator in
blood plasma: a possible role in regulating cardiac function via
sphingolipid receptors
Ka?roly LILIOM*†1, Guoping SUN‡1, Moritz BU?NEMANN§1, Tama?s VIRA?G*, No?ra NUSSER*, Daniel L. BAKER*, De-an WANG*,
Matthew J. FABIAN*, Bodo BRANDTS§, Kirsten BENDER§, Andreas EICKEL§, Kafait U. MALIK?, Duane D. MILLER‡,
Dominic M. DESIDERIO¶, Ga?bor TIGYI* and Lutz POTT§2
*Department of Physiology, The University of Tennessee Health Sciences Center, Memphis, TN 38163, U.S.A., †Institute of Enzymology, Biological Research Center of
the Hungarian Academy of Sciences, H-1518 Budapest, P.O. Box 7, Hungary, ‡Department of Pharmaceutical Sciences, The University of Tennessee Health Sciences
Center, Memphis, TN 38163, U.S.A., §Institut fu ? r Physiologie, Ruhr-Universita?t, D-44780 Bochum, Germany, ?Department of Pharmacology, The University of Tennessee
Health Sciences Center, Memphis, TN 38163, U.S.A., and ¶The Charles B. Stout Neuroscience Mass Spectrometry Laboratory, Departments of Neurology and
Biochemistry, The University of Tennessee Health Sciences Center, Memphis, TN 38163, U.S.A.
Blood plasma and serum contain factors that activate inwardly
rectifying GIRK1?GIRK4 K+channels in atrial myocytes via
one or more non-atropine-sensitive receptors coupled to
pertussis-toxin-sensitive G-proteins. This channel is also the
target of muscarinic M?receptors activated by the physiological
release of acetylcholine from parasympathetic nerve endings.
By using a combination of HPLC and TLC techniques with
matrix-assisted laser desorption ionization–time-of-flight MS,
we purified and identified sphingosine 1-phosphate (SPP) and
sphingosylphosphocholine(SPC)astheplasmaandserumfactors
responsible for activating the inwardly rectifying K+ channel
(IK). With the use of MS the concentration of SPC was estimated
at 50 nM in plasma and 130 nM in serum; those concentrations
exceeded the 1.5 nM EC??measured in guinea-pig atrial myo-
INTRODUCTION
Sphingosine 1-phosphate (SPP) is a multifunctional lipid me-
diator that regulates a multitude of cellular functions in ?itro
[1–3]. SPP was first identified as a Ca?+-mobilizing second
messenger [4]. More recently, it has also been recognized as a
genuine mediator that activates specific G-protein-coupled
receptors of the Edg family [5]. In some cell types, sphingo-
sylphosphocholine (SPC) has been shown to elicit similar cellular
responses to those of SPP but in other cells the two sphingolipids
have different effects [1,6]. For example, distinct Gi-coupled
receptors have been identified for SPC in HL-60 leukaemia cells
and human neutrophils [6], whereas in guinea-pig atrial myocytes
both sphingolipids activate the same receptor(s) [7]. Our previous
work has established that in guinea-pig atrial myocytes SPC and
SPP are equipotent (EC??approx. 1.5 nM) activators of a K+
current, which can also be activated by acetylcholine (ACh)
(IK(ACh)[7–9]). The activation of IK(ACh)represents a major
mechanism of vagal slowing of the cardiac frequency that in-
cludes the M?muscarinic ACh receptor-mediated activation
of the GIRK1?GIRK4 (Kir3.1?Kir3.4) K+channel complex by
the βγ subunits of a pertussis-toxin-sensitive G-protein (reviewed
in [10]). However, these sphingolipids do not activate IKthrough
Abbreviations used: 2D, two-dimensional; ACh, acetylcholine; CHCA, α-cyano-4-hydroxycinnamic acid; ELSD, evaporative light-scattering detector;
IK, inwardly rectifying K+ channel; MALDI–TOF-MS, matrix-assisted laser desorption ionization–time-of-flight MS; RT–PCR, reverse-transcriptase-
mediated PCR; SPC, sphingosylphosphocholine; SPP, sphingosine 1-phosphate.
1These authors contributed equally to this work.
2To whom correspondence should be addressed (e-mail lutz.pott?ruhr-uni-bochum.de).
cytes. With the use of reverse-transcriptase-mediated PCR
and?or Western blot analysis, we detected Edg1, Edg3, Edg5 and
Edg8 as well as OGR1 sphingolipid receptor transcripts and?or
proteins. In perfused guinea-pig hearts, SPC exerted a negative
chronotropic effect with a threshold concentration of 1 µM. SPC
was completely removed after perfusion through the coronary
circulation at a concentration of 10 µM. On the basis of their
constitutive presence in plasma, the expression of specific
receptors, and a mechanism of ligand inactivation, we propose
that SPP and SPC might have a physiologically relevant role in
the regulation of the heart.
Key
OGR-1.
words: Edgreceptor,GIRK,lysosphingomyelin,
M?muscarinic ACh receptors because their effect is not inhibited
by atropine and is not cross-desensitized with ACh [7,11]. SPP
has been shown to be generated in platelets via the action of
sphingosine kinase, which is stored and released on stimulation
by thrombin [12]. The concentrations of SPP in human plasma
andserumhavebeenestimatedtobe200and500 nMrespectively
[13].
We have reported previously that serum and plasma contain
one or more albumin-associated lipid factors that activate IKin
a pertussis-toxin-sensitive fashion [11,14] and reverse the cAMP-
mediated increase in L-type Ca?+-current in atrial and ventricular
myocytes [15]. The objectives of the current investigation were
(1) to purify and identify these blood factors, (2) to assess their
action on the intact perfused heart and (3) to obtain further
informationon theidentity
receptor(s). Our results provide evidence that SPC is also present
in blood at a high nanomolar concentration and is responsible in
part for the activation of IKby plasma and serum. We also
describe a matrix-assisted laser desorption ionization–time-of-
flight MS (MALDI–TOF-MS) method for the identification of
SPC in biological fluids. The plasma concentration of SPC
increased more than 2.5-fold in serum, indicating that its
generation is linked to blood clotting. Moreover, circulating SPC
oftheirG-protein-coupled
? 2001 Biochemical Society

Page 2

190
K. Liliom and others
is effectively removed after a single passage through the coronary
vasculature. In support of a physiologically relevant role, we
show, with the use of an isolated perfused heart preparation, that
SPC in the high nanomolar to low micromolar concentration
range decreases the heart rate and synergizes with the effect of
muscarinicstimulation.Inagreementwiththesebiologicaleffects,
we show that rat atrial myocytes co-express mRNA and protein
for the Edg1, Edg3, Edg5 and Edg8 SPP receptors as well as
mRNA for the recently identified SPC receptor OGR1. A
preliminary report on this work has been published in an abstract
[16].
EXPERIMENTAL
Lipid extraction
Blood was obtained by cardiac puncture of anesthetized adult
New Zealand White rabbits or from the cubital vein of healthy
human volunteer donors. For the isolation of plasma, blood was
anticoagulated with 0.1 ml of 15% (w?v) EDTA. Blood was also
collected into siliconized vials and left to clot for 1 h at 37 ?C,
followed by a 24 h incubation at 4 ?C. Serum and plasma were
separated by centrifugation for 15 min at 2000 g, frozen in liquid
nitrogen and freeze-dried. The freeze-dried material was first
extracted three times with 5 ml of chloroform?methanol (2:1,
v?v) per g dry weight, and three more times with 5 ml of
methanol per g dry weight. Both organic solvents contained
0.01% butylated hydroxytoluene and were concentrated in ?acuo
at ambient temperature.
Fractionation of plasma and serum lipids by HPLC
Analytical HPLC fractionation of serum and plasma lipids was
performed by the method we developed earlier for the separation
ofSPCstereoisomers[7],usingasilica-packedMicrosorbcolumn
(250 mm?4.6 mm; 5 µm particle size; Rainin Instruments,
Woburn,MA,U.S.A.).Elutionwasmonitoredbyanevaporative
light-scatteringdetector(ELSD,ModelIIA;Varex,Burtonsville,
MD, U.S.A.) through a metering valve, splitting the eluate 1:4 at
a flow rate of 1 ml?min between the detector and a fraction
collector respectively. Fractions (1 min) were collected and
assayed for their ability to elicit IK(ACh)in guinea-pig atrial
myocytes.
A partial purification of the IK(ACh)-activating blood factors
was achieved by HPLC, using a Microsorb-Si semipreparative
column (250 mm?10 mm; 5 µm particle size; Rainin Instru-
ments). The column was eluted with a two-solvent gradient
consisting of solvents A [chloroform?methanol?water?30%
(v?v) NH?OH, 60:35:4.5:0.5 by vol.] and B [chloroform?
methanol?water?30% (v?v) NH?OH, 30:60:9.5:0.5 by vol.]
with a Waters HPLC system controlled by the Millennium 2010
version 2.1 software package (Waters Instruments, Milford,
MA, U.S.A.). Gradient elution with a 2 ml?min flow rate was
started with 100% solvent A for 10 min; the concentration of
solvent B was increased linearly to 100% over a 30 min period,
followed by a 60 min isocratic application of 100% solvent B.
All solvents were HPLC grade (Fischer Scientific, Pittsburgh,
PA, U.S.A.). Approximately 50 mg of lipid extract, dissolved in
2 ml of solvent A, was injected, and the eluate was monitored by
an ELSD.
Two-dimensional (2D) TLC
Separations by 2D-TLC on K6 silica-gel plates (20 cm?20 cm,
layer thickness 250 µm; Whatman, Fairfield, NJ, U.S.A.) were
performed as described previously [17]. Lipid spots were detected
after being stained with primulin (Aldrich Chemical, Milwaukee,
WI, U.S.A.) [18] and in some instances were also counterstained
with ninhydrin (Sigma). Each primulin-positive spot was scraped
off and eluted as described [17] for bioassay in guinea-pig or rat
atrial myocytes. For bioassay the lipids were dissolved in a
minimal volume of methanol and diluted with recording buffer.
MALDI–TOF-MS
MALDI–TOF-MS data were acquired with a Voyager-DE RP
Biospectrometry Workstation
systems, Farmingham, MA, U.S.A.) in the reflector mode. The
laser energy was set at 1700 arbitrary units. Samples (in 0.5 µl)
were mixed with 1.5 µl of α-cyano-4-hydroxycinnamic acid
(CHCA; Sigma) matrix dissolved in 3% (v?v) trifluoroacetic
acid?acetonitrile?methanol (1:5:4, by vol.) and left to dry at
room temperature. For estimation of the amount of SPC, known
concentrations of SPC standard ranging from 2 to 300 nM were
prepared and mixed with the matrix. Each concentration of
standard contained an equal amount of the CHCA matrix (50 µl)
and SPC in methanol to give a total volume of 65 µl. The laser
targets were loaded with 2 µl of each standard in duplicate; the
solutionwas dried at roomtemperature. Eachtarget was scanned
160 times, providing a total of 640 individual measurements for
every sample for averaging. The peak heights corresponding to
the matrix ([CHCA?H]+, m?z 190.5) and SPC ([SPC?H]+, m?z
465.3) were measured, and the logarithm of the ratio of the two
peak intensities (log[SPC?CHCA]) was plotted against the log-
arithm of the SPC concentration.
instrument (PerSeptiveBio-
Testing of generation of SPC and SPP from sphingomyelin ex
vivo
[??C]Sphingomyelin (1 µCi) (55 mCi?mmol; Amersham) was
incubated with 100 µl of human serum or plasma for up to 24 h.
The samples were freeze-dried and the sphingolipid fraction was
extracted as described above and run on TLC. The TLC plate
was exposed to ??C-sensitive Super Resolution screens of a
Cyclone imager (Packard Instruments). Authentic sphingo-
myelin, SPC and SPP standards were run alongside the samples
and stained with ninhydrin.
Isolation and culture of atrial myocytes
The method for the isolation of myocytes has been described in
detail elsewhere [11].
Patch clamp recording of IK(ACh)from atrial myocytes
Whole-cell voltage clamp recording [19] of the sphingolipid-
elicited currents was performed under previously described
experimental conditions [7]. Spingolipids were applied in
recording buffer and the currents were normalized to the current
induced by a saturating concentration (10 µM) of ACh in the
same cell.
Isolated perfused heart preparation
The animal protocols were reviewed and approved by the
Institutional Animal Care and Use Committee, and animals were
maintained in the AAALAC-accredited programme (where
AAALAC stands for American Association for Accreditation of
Laboratory Animal Care). Experiments in Germany were per-
formed with local ethics committee approval. The effects of SPC
were tested with an isolated retrogradely perfused heart prepa-
ration. Hearts from guinea-pigs (700–900 g) anaesthetized with
? 2001 Biochemical Society

Page 3

191
Plasma lysosphingomyelin activates IK(ACh)in the heart
sodium pentobarbital were isolated and perfused with Krebs–
Henseleit buffer containing (in mM) 114 NaCl, 1.2 KH?PO?,
2.5 CaCl?, 1.2 MgSO?, 4.7 KCl, 25 NaHCO?and 5.5 glucose,
equilibrated with O??CO?(19:1), at 37 ?C. Hearts were perfused
at a constant rate of 18 ml?min with a peristaltic pump (Harvard
Apparatus, Millis, MA, U.S.A.). The perfusate and a water
jacket surrounding the heart were warmed to 37 ?C. Frequency
was measured by recording ventricular contractions or a ven-
tricular surface electrocardiogram. Activation of IK(ACh)strongly
affects sino-atrial automaticity and AV-nodal conduction. De-
creases in ventricular frequency are therefore related to both
parameters. Experiments were initiated after 30 min to allow
mechanical function of the heart to stabilize. Thereafter, doses
of ACh were injected in a 100 µl bolus. Each dose was repeated
twice at 3 min intervals before the next higher dose was injected.
SPC was infused at a constant rate of 0.34 ml?min for 20 min.
The SPC infusions were given consecutively in increasing order
of concentration, followed by a washout until recovery to pre-
exposure level occurred. To determine whether a subthreshold
infusion of SPC would modify the decrease in heart rate caused
by a subthreshold or larger dose of ACh, the following protocol
was used: after the stabilization period and before the SPC
infusion, baseline levels of ACh-induced decrease in heart rate
were determined by injecting 100 µl of 1–10 µM ACh twice
at 3 min intervals, followed by the infusion of 1 µM SPC at
0.34 ml?min. At 3 min intervals after approx. 10 min of 1 µM
SPC infusion, 100 µl boluses of 1–10 µM ACh were injected to
test for an ACh-induced decrease in heart rate.
Reverse-transcriptase-mediated PCR (RT–PCR) analysis of the
expression of sphingolipid receptor gene products
After isolation, rat atrial and ventricular myocytes (10? cells of
each) were plated in culture dishes in the absence of serum and
were incubated overnight to permit attachment. Under serum-
free conditions, fibroblasts and other non-myocyte cells did not
attach, providing a culture of virtually pure myocytes. The cells
were scraped from the dishes, pelleted by centrifugation and
placed into RNAlater (Ambion, Austin, TX, U.S.A.). Total
RNAwas isolatedwith TRIzol
Gaithersburg, MA, U.S.A.). To prevent DNA contamination,
1 µg of total RNA was digested with 1 µl of DNase1 for 15 min
at 37 ?C, followed by inactivation with 2.5 µl of 25 mM EDTA at
65 ?C for 10 min. cDNA was obtained by reverse transcription
with the use of the SUPERSCRIPT Preamplification System kit
(Gibco BRL). The oligonucleotide primers were based on the rat,
mouse and human receptor sequences as follows: Edg1 forward
5?-??tcatcgtccggcattacaacta-3?, reverse 5?-gagtgagcttgtaggtggtg-
???-3? (product size 273 bp); Edg3 forward 5?-???cttggtcatc-
tgcagcttcatc-3?, reverse 5?-tgctgatgcagaaggcaatgta???-3? (product
size 460); Edg5 forward 5?-??atgggcagcttgtactcggag-3?, reverse 5?-
cagccagcagacgataaagac???-3? (product size 720 bp); Edg6 for-
ward 5?-???tgaacatcacgctgagtgacct-3?, reverse 5?-gatcatcagcaccg-
tcttcagc???-3? (product size 510); Edg8 forward 5?-???atctgt-
gcgctctatgcaagga-3?, reverse 5?-ggtgtagatgataggattcagca????-3?
(product size 300); and OGR1 forward 5?-???ttcctgccctaccacgt-
gttgc-3?, reverse 5?-tggcgagttaggggtctggaag????-3? (product size
341 bp). After 2 min of predenaturation at 94 ?C, Taq DNA
polymerase (Promega) was added, and 30 amplification cycles
were run, each consisting of denaturation for 30 s at 94 ?C,
annealing for 1 min at 59 ?C and elongation for 1 min at 72 ?C.
The last elongation period was 5 min at 72 ?C. The products
were separated on 1.2% (w?v) DNA agarose gels and stained
with ethidium bromide.
reagent (Gibco BRL,
Immunoprecipitation and Western blot analysis of Edg3 and Edg5
receptor proteins
Rat atrial and ventricular tissues were carefully dissected and
homogenized with a Dounce homogenizer in ice-cold lysis buffer
[50 mM Hepes (pH 7.4)?50 mM NaCl?1 mM MgCl??2 mM
EDTA?1 mM vanadate containing Protease Inhibitor Cocktail
from Sigma]. The homogenate was centrifuged for 10 min at 4 ?C
and 1000 g to remove nuclei. To isolate the crude membrane
fraction, the supernatant was centrifuged for 1 h at 4 ?C and
100000 g. Protein concentration was determined with the Micro
BCA Protein Assay kit (Pierce, Rockford, IL, U.S.A.). For
immunoprecipitation, 150 µg of protein in 250 µl of sample
volume was preincubated with Protein G PLUS-Agarose (Santa
Cruz Biotechnology, Santa Cruz, CA, U.S.A.) for 5 h, followed
by incubation overnight with monoclonal antibodies specific for
a C-terminal epitope of Edg3 (2.5 µg per sample) and Edg5 (5 µg
per sample) (Antibody Solutions, Palo Alto, CA, U.S.A.) and
Protein G PLUS-Agarose. The immunocomplex was washed
three times, boiled for 3 min in sample buffer, and loaded for
SDS?PAGE[10%(w?v)gel].ForWesternblotanalysis,proteins
were transferred to PVDF membranes (Bio-Rad, Hercules, CA,
U.S.A.) with a semi-dry blotter (Bio-Rad). Non-specific binding
was blocked overnight in TTBS [10 mM Tris?HCl (pH 7.5)?
150 mM NaCl?0.05% (v?v) Tween 20] containing 5% (w?v)
non-fat dried milk (Bio-Rad). The membranes were incubated
with 1 µg?ml of either anti-Edg3 or anti-Edg5 N-terminal
monoclonal antibody in TTBS for 1 h, washed three times with
TTBSandincubatedwithahorseradish-peroxidase-labelledanti-
mouse secondary antibody (Sigma) for 30 min before being
developed with the SuperSignal Chemiluminescent Substrate
(Pierce).
Statistical analysis
Student’s t test was applied for the analysis of the results;
differences at P?0.05 were considered statistically significant.
RESULTS
We have shown previously that blood plasma or serum from
various mammalian species, when applied to atrial myocytes,
elicits an inwardly rectifying IK, indistinguishable from IK(ACh),
that is activated by ACh via M?receptors [7,14]. However, this
plasma- or serum-activated current was not blocked by atropine,
establishing that the ligand or ligands activating it were different
from ACh. We therefore first focused on the characterization of
this novel biological activity of serum and plasma. Independently
of the source or donor species, serum and plasma at a con-
centration of0.1% (v?v) elicited a saturating response of approx.
90% of the current that was activated by a saturating 10 µM
concentration of ACh. The EC??dilutions for human plasma and
serum were 10600 and 13200 respectively (results not shown).
Delipidated human albumin at a concentration as high as
11.6 µM was inactive. Serum albumin Cohn fraction V, at
5.8 µM (approx. 1% of its concentration in plasma), caused a
maximal activation of IK(ACh). The activity could be dissociated
from albumin by extraction with methanol. The current–voltage
relationship for the methanol extract of human serum albumin
was indistinguishable from that obtained for -erythro-SPC and
ACh [7], with a reversal potential at EK(?50 mV) and strong
inward rectification, typical of IK(ACh). Dialysis of plasma and
serum against aqueous buffers did not significantly alter their
threshold concentrations (results not shown). Because albumin
has been shown to bind tightly to a variety of bioactive molecules
? 2001 Biochemical Society

Page 4

192
K. Liliom and others
Figure 1
lipids in rabbit plasma (A) and serum (B)
Semi-preparative HPLC separation of methanol-extractable polar
Approximately 50 mg of the methanol extract was loaded; the effluent was monitored with an
ELSD. Black bars mark the IK-positive fractions that were subjected to further analysis with 2D-
TLC. The retention times for the SPP and SPC standards were 45 and 75 min respectively. The
elution profile is highly representative of at least five separations.
and because delipidated albumin preparations were inactive, we
focused our purification strategy on plasma and serum lipids.
HPLC fractionation of polar serum lipids reveals two distinct
peaks of biological activity
Rabbit plasma and serum was first extracted with chloroform?
methanol to remove the less polar fraction of lipids, followed by
extraction with methanol. The methanol extraction yielded
2.3 mg of extract per ml of plasma and 1.9 mg per ml of serum.
This extraction procedure has been proved to be effective for the
isolation of albumin-associated bioactive glycerolipids [20,21]
from different biological fluids [17]. The methanol-extractable
polar lipid fraction prepared from serum or plasma was highly
active in eliciting IK, whereas the less polar fraction extracted
with chloroform?methanol was inactive (results not shown). We
therefore applied an analytical HPLC to fractionate the major
lipid classes present in the methanol extract. The eluate was
collected in 1 min fractions and dried under N?; the fractions
were reconstituted in 250 µl of methanol and applied to guinea-
pig atrial myocytes at a dilution of 1:250 in recording buffer.
Two completely separated peaks of activity were detected be-
tween 21 and 26 min and between 32 and 40 min. These retention
times were different from those of the bioactive glycerolipids
lysophosphatidic acid (16 min) and lysoplasmenyl glycero-
phosphate (12 min) but were similar to those of authentic
standards for SPP (22 min) and SPC (33 min) [22]. These peak
retention times were particularly revealing, because we have
Figure 2
semi-preparative HPLC of rabbit plasma and that of authentic SPP and SPC
standards
Two-dimensional TLC analysis of the active fractions from the
(A, B) HPLC fractions: 65–88 min (A); 35–58 min (B). (C) SPC standard; (D) SPP standard.
Separations were done first with basic (first dimension) and then with acidic (second dimension)
solvents. As detected by primulin, the spot with an RF1of 0.18 and an RF2of 0.05 in (A) co-
migrated with the SPC standard (RF10.19; RF20.06) and the spot with an RF1of 0.03 and an
RF2of 0.35 in (B) co-migrated with the SPP standard (RF10.03; RF20.34) [denoted by asterisks
in (A) and (B)]. Spots were scraped off the plates, lipids were recovered by extraction with
methanol and the spots were tested for the activation of IK(ACh). The active spots were those
denoted by asterisks in (A) and (B).
shown previously that SPP and SPC are potent activators of
atrial IK[7,9]. However, the active HPLC fractions contained a
mixture of different compounds when analysed by MS; the
amount of the individual components obtained from analytical
HPLC was below the detection limit for 2D-TLC and staining
with primulin.
A semipreparative HPLC procedure was developed for the
separation of IK-activating factors, which allowed the loading of
50 mg of methanol extract in each run (Figure 1). The eluates at
35–58 min and 65–88 min, which were co-eluted with the SPP
and SPC standards, were separately collected for plasma and
serum, and dried under N?. In the 35–58 min fraction, average
quantitiesof0.68 and0.25 mgofsolidmaterial wereobtained per
ml of plasma and serum respectively. In the 65–88 min fraction,
average quantities of 0.07 and 0.06 mg of dry material were
recovered per ml of plasma and serum respectively. The material
collected in the two fractions from plasma and serum was
dissolved in methanol?chloroform (2:1, v?v) and further
separated by 2D-TLC. Figure 2 shows representative 2D-TLC
results obtained for the plasma fractions and for authentic SPC
and SPP standards. In the 35–58 min fraction, 11 different
lipids stained with primulin, four of which also stained with
ninhydrin. In the 65–88 min fraction, eight individual spots were
resolved, four of which were ninhydrin-positive. Each of these
spots, detectable with primuline, was scraped off the duplicate
plates that were developed in the same run and processed for
bioassay at a 1:50 dilution in recording buffer containing a final
concentration of 0.01% methanol. Only a single spot, eluted
from a 2D-TLC separation of fraction 35–58 min, was bio-
logically active; it co-migrated with the SPP standard (RF?0.03;
RF?0.34). In the 65–88 min fraction another bioactive spot was
? 2001 Biochemical Society

Page 5

193
Plasma lysosphingomyelin activates IK(ACh)in the heart
Figure 3
fractions with retention times between 65 and 88 min
MALDI–TOF-MS determination of the lipids comprising the HPLC
(A) Mass spectrum of 2 nmol of D-erythro-SPC standard; (B, C) spectra of the 65–88 min
semipreparative HPLC fractions of methanol extracts of plasma (B) and serum (C). Labelled
peaks: [CHCA?H]+at m/z 190.5 is the protonated molecule ion of the CHCA matrix;
[2?CHCA?H]+at m/z 379.2 is the dimer of the CHCA matrix; [SPC?H]+at m/z 465.3
is the protonated molecule ion of SPC; and [SPC?Na]+at m/z 487.2 is the sodium adduct
ion of SPC. The logarithm of the peak intensity ratio of [SPC?H]+at m/z 465.3 and that of
[CHCA?H]+at m/z 190.5 gives a straight line if plotted against the logarithm of SPC
concentration, allowing the estimation of the concentration of SPC.
detected; this co-migrated with the SPC standard (RF?0.19; RF?
0.06). The corresponding spots in the duplicate plates were
ninhydrin-positive, lending strong support to the notion that
these spots did indeed correspond to SPP and SPC respectively.
Similarly, only two active lipids were detected in the methanol
extract from serum (results not shown).
Identification of SPC and SPP by MALDI–TOF-MS
The two eluted bioactive spots that co-migrated with the SPP
and SPC standards were analysed by MALDI–TOF-MS. The
spot in the plasma fraction that co-migrated with the SPC
standard was dominated by molecular ions at m?z 465.3 and
487.2, corresponding to [SPC?H]+and [SPC?Na]+. The active
spot recovered from the SPP region of the plasma chromatogram
showed three major molecular ions at m?z 380.3, 402.5
and 424.5, corresponding to [SPP?H]+, [SPP?Na]+ and
Figure 4
receptors
SPC and serum activate the same sphingolipid receptor or
Columns show the average currents measured in control guinea-pig atrial myocytes and in
myocytes desensitized with 100 nM SPC (mean?S.E.M. for 5–8 cells for each condition).
Desensitization led to a substantial decrease in responsiveness of myocytes to serum (filled
columns) and SPC (hatched columns) without a significant decrease in the response to ACh
(open columns).
[SPP?2Na]+respectively. Identical mass spectra were found for
the two corresponding spots isolated from serum (results not
shown). Figure 3 shows the mass spectrum of an SPC standard
and mass spectra for the 65–88 min HPLC fractions of methanol
extracts of plasma and serum. The two peaks at m?z 465.3
([SPC?H]+) and 487.2 ([SPC?Na]+) indicate the presence of
SPC in the plasma and serum samples. The intensity of the
[SPC?H]+ peak at m?z 465.3 showed a strict concentration
dependence compared with the m?z 190.5 peak of the
[CHCA?H]+matrix. The logarithm of the peak intensity ratio
of [SPC?H]+to [CHCA?H]+plotted against the logarithm of
SPC concentration yielded a straight line with a correlation
coefficient of 0.978. This tight relationship between the SPC
concentration and the normalized MS signal intensity over the
range 2–300 nM enabled us to estimate the plasma and serum
concentrations of SPC as 50?15 nM and 130?25 nM (n?4)
respectively (means?S.E.M.). Unfortunately, this method could
not be applied for the quantification of SPP because [SPP?H]+
at m?z 380.3 could not be clearly resolved from the CHCA
matrix dimer at m?z 379.2.
To exclude the possibility that SPC or SPP was generated ex
?i?o, [??C]sphingomyelin was incubated with plasma and serum
for up to 24 h. The lipid extracts prepared from these samples
with chloroform?methanol and with methanol showed no de-
tectable generation of SPP and SPC or breakdown of sphingo-
myelin (results not shown).
SPC and SPP desensitizes IKelicited by serum but not by ACh
We have previously shown that preincubation of guinea-pig
myocytes with SPC causes
desensitization of IKelicited by SPP, whereas the sensitivity of IK
to stimulation with M?muscarinic ACh receptors was not
affected [7]. This result indicated that SPP and SPC activate the
same population of receptors. To establish whether the IK-
activating serum factors activate the same subset of sphingolipid
receptors, we preincubated guinea-pig atrial myocytes with
100 nM SPC for ?10 h and tested their responsiveness to 1:100-
diluted rabbit serum and 10 µM ACh. Cells from untreated
(non-desensitized) cultures served as controls. Figure 4 shows
the completeheterologous
? 2001 Biochemical Society

Page 6

194
K. Liliom and others
Figure 5
manner in the isolated heart
SPC modifies cardiac function in a concentration-dependent
(A) The naturally occurring D-erythro-SPC (?) caused a concentration-dependent decrease in
heart rate, similar to that caused by ACh (results not shown). L-threo-SPC (?) was ineffective
up to 10 µM. *P?0.05 compared with control (by Student’s t test). (B) D-erythro-SPC at its
threshold concentration of 1 µM potentiates the effect of Ach, whereas L-threo-SPC at the same
concentration does not. Filled columns, ACh alone; hatched columns, 1 µM D-erythro-SPC plus
ACh; open columns, 1 µM L-threo-SPC plus ACh. Values are presented as means?S.E.M.
that the sensitivity of IKto 100 nM SPC and 1:100-diluted serum
was abolished by pretreatment with SPC; these saturating con-
centrations of SPC or serum evoked only a minimal current in
SPC-desensitized cells, whereas the response in the same cells to
10 µM ACh was unaffected. Comparable results were obtained
when 100 nM SPP was used for desensitization (results not
shown).
SPC decreases cardiac frequency in the isolated heart
ACh decreases spontaneous beating frequency in the isolated
perfused heart, which is due at least partly to the activation of
IK(ACh)[23,24]. In the isolated guinea-pig heart the naturally
occurring stereoisomer -erythro-SPC decreased the heart rate in
a concentration-dependent manner (Figure 5A). However, the
active concentration range in the intact heart was three orders of
magnitude larger than for the activation of IKin isolated
myocytes. The -threo-SPC stereoisomer, at concentrations up to
Figure 6Inactivation of SPC by the coronary vasculature
(A) The current trace shows that SPC was inactivated during the coronary passage, as the
10 µM SPC included in the perfusate exerted its maximal effect on myocytes, whereas
the effluate was devoid of the biological activity. Both the perfusate and the effluate were tested
at a 50-fold dilution in guinea-pig atrial myocytes for the activation of IK(ACh). (B) MALDI–
TOF spectrum for the methanol-soluble lipids prepared from the perfusate and the effluate. Note
the clear presence of the SPC molecular ions (m/z 465.5 and 487.5) in the perfusate and
its complete absence from the effluate. (C) IKactivated by SPC diluted in the effluate is not
attenuated compared with SPC diluted in recording buffer, suggesting that the effluate did
not contain inhibitory factors that diminish its effect.
10 µM, was ineffective in the perfused isolated heart (Figure 5A);
those results duplicate our earlier findings with this compound,
which was ineffective in activating IKin isolated myocytes [7].
Only -erythro-SPC caused an increase in coronary perfusion
pressure and a decrease in ventricular contractility parameters
(results not shown), properties that are likely to reflect actions
different from the activation of IK(ACh).
SPC and ACh elicited similar physiological responses in the
? 2001 Biochemical Society

Page 7

195
Plasma lysosphingomyelin activates IK(ACh)in the heart
Figure 7 RT–PCR analysis detects the presence of messages for Edg1, Edg3, Edg5, Edg8 and OGR1 receptors
The amplification was performed with both atrial and ventricular cultured myocytes separately; the products were analysed by agarose-gel electrophoresis. (A) Amplification for the Edg family of
sphingolipid receptors. Lane M, DNA standards; lane C, β-actin control (220 bp). For all Edg receptors, lanes were as follows: lane 1, positive control (plasmid); lane 2, sample; lane 3, negative
control (no transcription). (B) Amplification for the SPC-specific receptor OGR1. Lane M, DNA standards; lane C, positive control (plasmid); lane A, atrial myocytes; lane V, ventricular myocytes.
Figure 8
ventricular and atrial myocytes
Co-expression of the Edg3 and Edg5 receptor proteins in
Membrane preparations from carefully dissected rat atria and ventriculi were first immuno-
precipitated (IP) with anti-Edg3 and anti-Edg5 monoclonal antibodies recognizing the C-terminus
of the receptors (3CT and 5CT respectively), followed by Western blots (W) with monoclonal
antibodies against an N-terminal epitope (3NT and 5NT respectively). Western blots performed
without the primary antibodies served as negative controls (Ø). Identical results were obtained
when the N-terminal-specific antibodies were used for immunoprecipitation (results not shown).
isolated heart preparation and in atrial myocytes, raising the
possibility that they might jointly modulate the heart rate. To
test this hypothesis, SPC was perfused at a threshold con-
centration and ACh was applied at 1–10 µM, around its own
threshold concentration of 1 µM. As depicted in Figure 5(B), the
ACh-elicited decrease in heart rate was significantly augmented
inthepresenceof1 µM-erythro-SPC,whereasitwasunchanged
when the perfusate contained the inactive -threo-SPC stereo-
isomer.
The micromolar concentration required in the perfused heart
to decrease heart rate is difficult to reconcile with the 1.5 nM
EC??found for SPP and SPC in isolated myocytes [7,9]. One
explanation for this discrepancy could be a rapid inactivation by
degradation and?or cellular uptake during coronary perfusion.
To test this hypothesis, SPC was monitored in the coronary
effluate of the perfused heart preparation by using the atrial
myocyte IKas a highly sensitive bioassay and by MALDI–TOF-
MS.Intheseexperiments,10 µMSPCwasinfusedat0.34 ml?min
through the perfusion cannula; samples were taken from both
the coronary perfusate and the effluate. After 1:50 dilution
(approx. 200 nM) the aliquots were tested on atrial myocytes as
well as by MALDI–TOF-MS. As shown by the representative
trace in Figure 6(A), the perfusate induced an inward IK
corresponding to 90% of the maximum current that could be
activated by ACh. In contrast, the effluate from the coronary
sinus at the same dilution was completely inactive (Figure 6A).
Because the detection threshold for SPC in the guinea-pig atrial
myocyte assay is between 0.1 and 1 nM [7], this demonstrates
that more than 99% of the SPC had been removed during
coronary perfusion by a hitherto unknown mechanism. This
clearance mechanism was independent of the duration of per-
fusion of SPC between the durations of 30 s and 9 min tested
(results not shown). MALDI–TOF analysis of the lipid extract
prepared from the effluate showed no detectable amount of SPC
molecular ion, whereas the ion was clearly detectable in the
perfusate (Figure 6B). In addition, to exclude the generation of
inhibitory factors in the coronary vasculature, fresh SPC was
perfused in the effluate collected from the coronary sinus and
tested with SPC made up in buffer. These two samples showed
activity that was equally effective in the myocyte assay activation
of IK(Figure 6C).
Atrial myocytes express multiple subtypes of sphingolipid
receptors
Edg1, Edg3, Edg5, Edg6 and Edg8 are G-protein-coupled
receptors that are specifically activated by SPP and?or SPC
[25–30]. RT–PCR analysis detected the presence of Edg1, Edg3,
Edg5andEdg8mRNAinisolatedatrialandventricularmyocytes
(Figure 7A). Transcripts for Edg6 receptor could not be detected
? 2001 Biochemical Society

Page 8

196
K. Liliom and others
(Figure 7A). It has recently been shown that SPC, but not SPP,
is a high-affinity ligand for the ovarian cancer orphan receptor
OGR1 [31]. OGR1 transcripts were detectable in atrial and
ventricular myocytes (Figure 7B).
To ascertain the expression of Edg3 and Edg5 protein in the
atrium and ventricle, we applied monoclonal antibodies raised
against the N-terminal and C-terminal epitopes. The immuno-
precipitated protein was probed by using Western blots with a
second monoclonal antibody specific for N-terminal epitopes of
the receptors. Blots prepared from either tissue showed the
presence of a single band at approx. 42 kDa that was not present
when the blots were developed without the primary antibodies
(Figure 8). This 42 kDa antigen has been previously identified as
the Edg3 and Edg5 receptor protein [32].
DISCUSSION
The primary focus of our study was to identify blood-derived
mediatorsthataffecttheheartthroughaK+conductancechannel
in a membrane-delimited mechanism similar to the vagal neuro-
transmitter ACh. IK(ACh)is important for the physiological
regulation of cardiac parameters such as sino-atrial frequency,
duration of the atrial action potential and AV-nodal conduction
[10]. Using the myocyte IKas a bioassay, we found that the active
mediators were SPP and SPC on the basis of (1) co-migration
with authentic SPP and SPC standards in 2D-TLC and (2)
MALDI–TOF-MS analysis. In contrast with SPP, which has
been found in high nanomolar concentrations in blood plasma
[13], no results were previously available on the natural oc-
currence of SPC. Using MALDI–TOF-MS, we estimated the
SPC concentration in plasma and serum as 50?15 and
130?25 nM respectively. Thus our data establish that SPC is
also a naturally occurring lipid mediator in blood, with a slightly
lower concentration than SPP, which was previously estimated
as 200 and 500 nM in plasma and serum respectively [13].
Quantification of SPC should be treated as an estimate because
a quantitative determination could only be derived from experi-
ments that include a mass-tagged internal standard that was not
available to us. Our estimate for the concentration of SPC and
the previously reported concentration of SPP yield an approx.
250 nM combined plasma concentration for the two sphingo-
lipids. Given the 1:10000 dilution as the EC??plasma con-
centration for the activation of IKand that the EC??for SPP and
SPC was approx. 1.5 nM in this assay, one would expect a
combined plasma concentration of the two mediators of approx.
15 µM. The reason for the substantial difference between the
estimated and expected concentrations remains unclear and will
be further investigated. Serum or plasma might contain other
factors that enhance solubilization, potentiate and?or synergize
with the effect of the sphingolipids as well as other non-lipid
factors that have eluded our purification strategy. Nevertheless,
we tested each HPLC fraction and primulin-positive lipid spot in
the 2D-TLC separations for bioactivity in the IKassay. Thus, any
active lipid compound(s) other than SPP and SPC, if present,
was in a concentration that was below the threshold sensitivity of
our detection and bioassay systems.
Numerous studies in ?itro with established cell lines have
reported biological effects elicited by SPP and SPC [1,2,33].
Despite the fact that SPP is known as a platelet-derived sphingo-
lipid mediator [12,13], only a few recent studies have reported
on the cardiac effects of SPP in complex models such as the
canine isolated, blood-perfused sinoatrial node and papillary
muscle preparation [34] and a rat model in ?i?o [35]. In those
experiments SPP elicited species-dependent effects. In the canine
model, it evoked positive chronotropic and coronary vasocon-
strictor effects [34], whereas in the rat model circulating SPP
decreased the heart rate, ventricular contraction and blood
pressure [35]. However, there is a complete lack of information
on whether SPC has any acute effects in ?i?o at the organ level.
In the guinea-pig intact heart preparation we found that SPC
decreased cardiac frequency in a dose-dependent manner, with a
similar efficiency to that of ACh. Furthermore, as in isolated
atrial myocytes, so in the whole heart too was only the naturally
occurring -erythro stereoisomer of SPC active. In a limited
number of experiments, SPP showed similar effects in the guinea-
pig heart preparation (results not shown).
Because SPC and ACh both activate IK, and because both
compounds are capable of decreasing the cardiac frequency in
the intact heart preparation, we investigated whether their effect
is additive at the organ level. We found that when the perfusate
contained 1 µM -erythro-SPC, ACh was approximately an
order of magnitude more potent in decreasing the heart rate.
This modulation was specific for the -erythro-SPC stereoisomer
and was absent with -threo-SPC; such stereospecificity is the
same as that found previously for the activation of IK[7].
Together, these observations raise the possibility that SPC and
SPP,inconcertwithothermediators,contributetotheregulation
of heart rate and possibly to other cardiovascular parameters.
The genuine mediator role of SPC is supported not only by its
increased concentration after blood clotting but also by the
discovery of its inactivation during the coronary passage (Figure
6). The heart, and possibly other organs, might contain a
quantitative clearance mechanism for SPC. Because SPC is
present in blood at a high nanomolar steady-state concentration
and it is not generated ex ?i?o from sphingomyelin (results not
shown), we hypothesize that SPC, like SPP, could be released
into the bloodstream at a constant rate. Whereas plasma SPP has
been shown to originate from platelets, the source of SPC
remains to be identified.
Which receptor or receptors mediate the effect of SPC in the
heart? Recently, five G-protein-coupled receptors of the Edg
family, Edg1, Edg3, Edg5, Edg6 and Edg8, have been identified
that are selectively activated by SPP [25–30]. whereas these Edg
receptors bind SPC about two orders of magnitude less potently
than SPP, the ovarian cancer orphan receptor OGR1 has been
shown recently to be a high-affinity receptor for SPC [31].
RT–PCR analysis, with mRNA purified from isolated atrial and
ventricular rat myocyte cultures, confirmed the expression of
four Edg sphingolipid receptors with the exception of Edg6, as
well as OGR1 receptors. Further evidence was obtained by
Westernblotanalysis,showingthatbothEdg3andEdg5proteins
are expressed in atrial as well as ventricular membranes. The
antibodies currently available could not be used for immuno-
cytochemical localization of the antigen. Thus we could not
obtain direct evidence for the co-expression of these receptors in
the same cell. Nevertheless, because atrial myocytes form a
functionally coupled syncytium, the simultaneous detection of
Edg3 and Edg5 proteins lends strong support for a joint role
of these receptors in mediating responses to lysosphingolipids.
Because of the simultaneous presence of these receptors in
cardiac cells, we could not reach a definitive conclusion about
which (sub)types mediate each of the different sphingolipid
effects, which include activation of IK, inhibition of L-type Ca?+
current, enhancement of ACh response, effects on cardiac
contractility and coronary perfusion. The identification of the
receptor(s) and the signalling pathways involved in mediating
these effects must await a future study. Recently, Himmel et al.
[36] confirmed our previous results [7] reporting that SPC and
SPP are equipotent activators of IKin guinea-pig atrial myocytes,
whereas in mouse and human atrial myocytes SPC was less
? 2001 Biochemical Society

Page 9

197
Plasma lysosphingomyelin activates IK(ACh)in the heart
potent, suggesting substantial interspecies differences in either
the ligand specificity of the sphingolipid receptor(s) or the
expression of a specific receptor subtype responsible for the high
potency of SPC in guinea-pigs. These authors proposed that the
non-specific pharmacological agent suramin potently and
selectively inhibits the activation of IK(ACh)by SPP in human
cardiomyocytes via the Edg3 receptor subtype. Their concept
was based on a report by Ancellin and Hla [37] illustrating a
differential inhibition of the SPP-mediated activation of Edg7,
but not Edg1 or Edg5, by suramin. However, Ancellin and Hla
used heterologous co-expression of Edg1 and Edg5 with the
chimaeric Gαqiprotein in Xenopus oocytes, whereas for Edg3
they expressed the receptor only. Furthermore, the IC??for
suramine in oocytes treated with 5 nM SPP was approx. 50 µM
[37], which is consistent with the concentration range at which
suramin usually exerts its non-specific pharmacological effects in
a variety of systems. In contrast, Himmel et al. [36] reported a
0.2 nM IC??for suramin against 1 µM SPP. These contradictions
must be resolved in further studies with knock-out animal
models and?or anti-sense oligonucleotide-mediated inhibition
studies.
We thank Anke Galhoff for excellent technical assistance. This work was supported
by grants from NATO Collaborative Research Programmes (950790), US PHS
HL61469 (toG.T.),NSFIBN9728147
Forschungsgemeinschaft Po 212/6-2 (to L.P.). Grants from NIH DRR 10522 (to
D.M.D.) and NSF 9604633 (to D.M.D.) supported the purchase of the MALDI–TOF
instrument. G.T. is an Established Investigator of the American Heart Association
(9640217N).
(toG.T.) andDeutsche
REFERENCES
1Meyer zu Heringdorf, D., Van Koppen, C. J. and Jakobs, K. H. (1997) Molecular
diversity of sphingolipid signalling. FEBS Lett. 410, 34–38
Spiegel, S. and Merrill, Jr, A. H. (1996) Sphingolipid metabolism and cell growth
regulation. FASEB J. 10, 1388–1397
Spiegel, S. and Milstien, S. (2000) Functions of a new family of sphingosine-1-
phosphate receptors. Biochim. Biophys. Acta 1484, 107–116
Olivera, A. and Spiegel, S. (1993) Sphingosine-1-phosphate as second messenger in
cell proliferation induced by PDGF and FCS mitogens. Nature (London) 365,
557–560
Pyne, S. and Pyne, N. J. (2000) Sphingosine 1-phosphate signalling in mammalian
cells. Biochem. J. 349, 385–402
Van Koppen, C. J., Meyer Zu Heringdorf, D., Zhang, C., Laser, K. T. and Jakobs,
K. H. (1996) A distinct Giprotein-coupled receptor for sphingosylphosphorylcholine in
human leukemia HL-60 cells and human neutrophils. Mol. Pharmacol. 49, 956–961
Bu ? nemann, M., Liliom, K., Brandts, B. K., Pott, L., Tseng, J. L., Desiderio, D. M.,
Sun, G., Miller, D. and Tigyi, G. (1996) A novel membrane receptor with high affinity
for lysosphingomyelin and sphingosine 1-phosphate in atrial myocytes. EMBO J. 15,
5527–5534
Bu ? nemann, M., Brandts, B., Meyer zu Heringdorf, D., van Koppen, C. J., Jakobs,
K. H. and Pott, L. (1995) Activation of muscarinic K+current in guinea-pig atrial
myocytes by sphingosine-1-phosphate. J. Physiol. (London) 489, 701–777
Van Koppen, C., Meyer zu Heringdorf, M., Laser, K. T., Zhang, C., Jakobs, K. H.,
Bunemann, M. and Pott, L. (1996) Activation of a high affinity Giprotein-coupled
plasma membrane receptor by sphingosine-1-phosphate. J. Biol. Chem. 271,
2082–2087
Yamada, M., Inanobe, A. and Kurachi, Y. (1998) G protein regulation of potassium
ion channels. Pharmacol. Rev. 50, 723–760
Banach, K., Hu ? ser, J., Lipp, P., Wellner, M. C. and Pott, L. (1993) Activation of
muscarinic K+current in guinea-pig atrial myocytes by a serum factor. J. Physiol.
(London) 461, 263–281
Yatomi, Y., Ruan, F., Hakomori, S. and Igarashi, Y. (1995) Sphingosine-1-phosphate:
a platelet-activating sphingolipid released from agonist-stimulated human platelets.
Blood 86, 193–202
Yatomi, Y., Igarashi, Y., Yang, L., Hisano, N., Qi, R., Asazuma, N., Satoh, K., Ozaki,
Y. and Kume, S. (1997) Sphingosine 1-phosphate, a bioactive sphingolipid
abundantly stored in platelets, is a normal constituent of human plasma and serum.
J. Biochem. (Tokyo) 121, 969–973
2
3
4
5
6
7
8
9
10
11
12
13
Received 21 August 2000/18 December 2000; accepted 24 January 2001
14Bu ? nemann, M. and Pott, L. (1993) Membrane-delimited activation of muscarinic K
current by an albumin-associated factor in guinea-pig atrial myocytes. Pflu ? gers Arch.
425, 329–334
Banach, K., Bu ? nemann, M., Hu ? ser, J. and Pott, L. (1993) Serum contains a potent
factor that decreases beta-adrenergic receptor-stimulated L-type Ca2+current in
cardiac myocytes. Pflu ? gers Arch. 423, 245–250
Liliom, K., Bunemann, M., Sun, G., Miller, D., Desiderio, D. M., Brandts, B., Bender,
K., Pott., Nusser, N., Tigyi, G. (2000) Sphingosylphosphorylcholine is a bona fide
mediator regulating heart rate. Ann. Acad. N. Y. Acad. Sci. 905, 308–310
Liliom, K., Guan, Z., Tseng, J. L., Desiderio, D. M., Tigyi, G. and Watsky, M. A.
(1998) Growth factor-like phospholipids generated after corneal injury. Am. J. Physiol.
274, C1065–C1074
Wright, R. S. (1971) A reagent for the non-destructive location of steroids and some
other lipophilic materials on silica gel thin-layer chromatograms. J. Chromatogr. 59,
220–221
Hamill, O. P., Marty, A., Neher, E., Sakmann, B. and Sigworth, F. J. (1981) Improved
patch-clamp techniques for high-resolution current recording from cells and cell-free
membrane patches. Pflu ? gers Arch. 391, 85–100
Tigyi, G., Henschen, A. and Miledi, R. (1991) A factor that activates oscillatory
chloride currents in Xenopus oocytes copurifies with a subfraction of serum albumin.
J. Biol. Chem. 266, 20602–20609
Tigyi, G. and Miledi, R. (1992) Lysophosphatidates bound to serum albumin activate
membrane currents in Xenopus oocytes and neurite retraction in PC12
pheochromocytoma cells. J. Biol. Chem. 267, 21360–21367
Tigyi, G., Liliom, K., Fischer, D. J. and Guo, Z. (1999) Phospholipid growth factors:
identification and mechanism of action. In Lipid Second Messengers (Laychock, S. G.
and Rubin, R. P., eds.), pp. 51–81, CRC Press, Boca Raton, FL
Honjo, H., Kodama, I., Zang, W. J. and Boyett, M. R. (1992) Desensitization to
acetylcholine in single sinoatrial node cells isolated from rabbit hearts. Am.
J. Physiol. 263, H1779–H1789
Wickman, K., Nemec, J., Gendler, S. J. and Clapham, D. E. (1998) Abnormal heart
rate regulation in GIRK4 knockout mice. Neuron 20, 103–114
An, S., Bleu, T., Huang, W., Hallmark, O. G., Coughlin, S. R. and Goetzl, E. J. (1997)
Identification of cDNAs encoding two G protein-coupled receptors for
lysosphingolipids. FEBS Lett. 417, 279–282
Im, D.-S., Heise, C. E., Ancellin, N., O’Dowd, B. F., Shei, G., Heavens, R. P., Rigby,
M. R., Hla, T., Mandala, S., McAllister, G. et al. (2000) Characterization of a novel
sphingosine 1-phosphate receptor, Edg-8. J. Biol. Chem. 275, 14281–14286
Lee, M.-J., Van Brocklyn, J. R., Thangada, S., Liu, C. H., Hand, A. R., Menzeleev, R.,
Spiegel, S. and Hla, T. (1998) Sphingosine-1-phosphate as a ligand for the G protein-
coupled receptor EDG-1. Science 279, 1552–1555
Van Brocklyn, J. R., Gra?ler, M. H., Bernhardt, G., Hobson, J. P., Lipp, M. and
Spiegel, S. (2000) Sphingosine-1-phosphate is a ligand for the G protein-coupled
receptor EDG-6. Blood 95, 2624–2629
Yamazaki, Y., Kon, J., Sato, K., Tomura, H., Sato, M., Yoneya, T., Okazaki, H.,
Okajima, F. and Ohta, H. (2000) Edg-6 as a putative sphingosine 1-phosphate
receptor coupling to Ca2+signaling pathway. Biochem. Biophys. Res. Commun. 268,
583–589
Zondag, G. C. M., Postma, F. R., Van Etten, I., Verlaan, I. and Moolenaar, W. H.
(1998) Sphingosine 1-phosphate signalling through the G-protein-coupled receptor
Edg-1. Biochem. J. 330, 605–609
Xu, Y., Zhu, K., Hong, G., Wu, W., Baudhuin, L. M., Xiao, Y. and Damron, D. S.
(2000) Sphingosylphosphorylcholine is a ligand for ovarian cancer G-protein-coupled
receptor 1. Nature Cell Biol. 2, 261–267
Goetzl, E. J., Kong, Y. and Kenney, J. S. (1999) Lysophospholipid enhancement of
human T cell sensitivity to diphtheria toxin by increased expression of heparin-
binding epidermal growth factor. Proc. Assoc. Am. Physicians 111, 259–269
Goetzl, E. J. and An, S. (1998) Diversity of cellular receptors and functions for the
lysophospholipid growth factors lysophosphatidic acid and sphingosine 1-phosphate.
FASEB J. 12, 1589–1598
Sugiyama, A., Yatomi, Y., Ozaki, Y. and Hashimoto, K. (2000) Sphingosine
1-phosphate induces sinus tachycardia and coronary vasoconstriction in the canine
heart. Cardiovasc. Res. 46, 119–125
Sugiyama, A., Aye, N. N., Yatomi, Y., Ozaki, Y. and Hashimoto, K. (2000) Effects of
sphingosine 1-phosphate, a naturally occurring biologically active lysophospholipid, on
the rat cardiovascular system. Jpn. J. Pharmacol. 82, 338–342
Himmel, H. M., Meyer Zu Heringdorf, D., Graf, E., Dobrev, D., Kortner, A., Schu ? ler,
S., Jakobs, K. H. and Ravens, U. (2000) Evidence for Edg-3 receptor-mediated
activation of IK.AChby sphingosine-1-phosphate in human atrial cardiomyocytes. Mol.
Pharmacol. 58, 449–454
Ancellin, N. and Hla, T. (1999) Differential pharmacological properties and signal
transduction of the sphingosine-1-phosphate receptors EDG-1, EDG-3, and EDG-5.
J. Biol. Chem. 274, 18997–19002
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
? 2001 Biochemical Society