| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online November 9, 2000 as doi:10.1096/fj.00-0370fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Zentrum der Pharmakologie, Klinikum der Johann Wolfgang Goethe-Universität, D-60590, Frankfurt am Main, Germany; and
Department of Botany, UNIGEN Center for Molecular Biology, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
2Correspondence: Institute of Pharmacology and Toxicology, Klinikum der J.W. Goethe-Universität, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany. E-mail: Pfeilschifter{at}em.uni-frankfurt.de
SPECIFIC AIM
Our interest focuses on the sphingolipid-ceramide signaling cascade and on the identification of molecular targets of ceramide action, which has proved to be difficult and so far mainly indirect. We show that ceramide, which is an early messenger of inflammatory cytokine action exerts a dual effect on the cytosolic phospholipase A2 (cPLA2), the rate-limiting enzyme in arachidonic acid release and subsequent eicosanoid formation.
PRINCIPAL FINDINGS
1. Ceramide induces arachidonic acid release from mesangial cells
Stimulation of [3H]arachidonic acid-labeled
mesangial cells with short-chain (C6) or long-chain (C16) ceramides
reveals a dose-dependent increase in [3H]arachidonic acid
release after 60 min of stimulation. The long-chain C16-ceramide was
found to be more potent than the short-chain C6-ceramide to stimulate
arachidonic acid release.
Preincubation of mesangial cells with either an inhibitor of the classical MAPK cascade, PD 98059 (upto 20 µM), or an inhibitor of protein kinase C (PKC), Ro 31–8220 (at 1 µM), revealed no significant reduction of ceramide-induced arachidonic acid release, thus suggesting that the classical MAPK cascade and the conventional and novel PKCs are not involved in the ceramide-triggered response.
2. Direct binding of TID-ceramide to cPLA2
To evaluate whether cPLA2 is a direct target of
ceramide due to its amino-terminal CaLB domain, we used a photoaffinity
labeling analog of ceramide,
[125I]3-trifluoromethyl-3-(m-iodophenyl)diazirine-ceramide
(TID-ceramide), of high
[125I]-iodine-specific radioactivity (2000 Ci/mmol) and
incubated recombinant cPLA2 with this compound in a
cell-free system for 5 min, followed by photolysis at 364 nm. As seen
in Fig. 1A
, there is a strong labeling of cPLA2
by TID-ceramide. There was no difference in labeling between the D- and
L-erythro isomers of TID-ceramide, suggesting that the absolute
configuration is not critical for binding to
cPLA2.
|
Obviously, the calcium-dependent lipid binding (CaLB) domain of cPLA2 is an attractive candidate for ceramide binding, and we therefore generated mutants of cPLA2 that lack the CaLB domain. Using the baculovirus expression system, we produced 1) the catalytic domain of cPLA2 containing amino acid residues 132–749, which is devoid of the CaLB domain (cat), 2) as a control, the isolated CaLB domain containing the amino acids 1–131, and 3) a mutant of cPLA2 lacking the hinge region that links the CaLB domain to the catalytic domain (containing residues 1–147/459–749; Mut1).
As seen in Fig. 1A
, the catalytic domain of
cPLA2, which runs as an 
78 kDa
polypeptide on sodium dodecyl sulfate-polyacrylamide gel
electrophoresis (SDS-PAGE), did not bind
[125I]-TID-ceramide, whereas the mutant lacking
the hinge region in cPLA2 (Mut1) was still able
to bind to TID-ceramide (Fig. 1A
). Moreover, the insect cell
lysates containing recombinant CaLB domain showed a strong
[125I] labeling at 14.4 kDa, the predicted size
of the CaLB fragment (Fig. 1B
). The autoradiogramm of the
TID-ceramide binding (Fig. 1B
) revealed an additional
labeled band at a size of 29 kDa. Notably, this band was also
detected in the Western blot analysis with the
cPLA2-specific antibody in insect cell lysates
transfected with either empty baculovirus or CaLB-encoding virus. It
may be hypothesized that this 29 kDa cPLA2
immunoreactive band represents an endogenous baculovirus
cPLA2 that is also able to bind ceramide. Most
important, the TID-ceramide binding to cPLA2 was
completely Ca2+ dependent because in the absence
of Ca2+, TID-ceramide binding to
cPLA2 was completely abrogated, strongly arguing
for a specific binding of ceramide.
Moreover, another group of PLA2 was tested for TID-ceramide binding, the low molecular weight secreted PLA2 (sPLA2), which shows no sequence homology to cPLA2 and also lacks a CaLB domain, but still contains the hydrophobic substrate binding site. However, no binding of TID-ceramide to recombinant group IIA sPLA2 was detected.
3. Binding studies of cPLA2 to ceramide-containing
liposomes and cPLA2 activity in vitro
To examine whether ceramide is able to target cPLA2 to
the membrane, bringing the enzyme to its substrate and thus
facilitating cPLA2 activation, we performed in
vitro translocation studies using PC/ceramide liposomes.
Indeed, increasing mol% of ceramide in the PC-liposomes led to an
increased translocation of cPLA2 from the soluble fraction
to the membrane-like liposome fraction.
To determine whether ceramide binding to cPLA2 is
accompanied by an activation of the enzyme, in vitro
activity assays were performed. As depicted in Fig. 2
, there is only a suboptimal activation of cPLA2
in the presence of [14C]stearoyl-arachidonyl-PC
(SAPC) liposomes alone. However, supplementation of diacylglycerol to
the substrate [14C]SAPC at a molar ratio of 2:1
led to a further sixfold activation of cPLA2.
|
When diacylglycerol was replaced by natural ceramide, an almost equally
potent activation of cPLA2 was observed (Fig. 2)
.
Combining SAPC with diacylglycerol and ceramide at a molar ratio of
2:0.5:0.5 revealed no further potentiation of
cPLA2 activation as compared to
SAPC/diacylglycerol or SAPC/ceramide, respectively. Moreover, liposomes
containing SAPC and sphingomyelin were prepared and tested for
cPLA2 activation in vitro. Under this
condition, a suboptimal activation was again achieved comparable to the
SAPC liposomes alone. However, when bacterial sphingomyelinase was
added to SAPC/sphingomyelin liposomes, a dose-dependent increase in
cPLA2 activity was observed, confirming that
ceramide, generated by the action of bacterial sphingomyelinase is
sufficient to stimulate cPLA2 activity. This
activation was Ca2+ dependent since, in the
absence of Ca2+, cPLA2
activity was completely blocked (Fig. 2)
.
CONCLUSIONS AND SIGNIFICANCE
A novel approach to identify downstream targets of ceramide is the use of a radioiodinated, photoaffinity-labeling analog of ceramide that has been successfully used to identify the kinase c-Raf and certain PKC isoenzymes as ceramide binding partners. In the present study we identified cPLA2 as a further direct target of the lipid signaling molecule ceramide.
The binding of TID-ceramide to the CaLB domain of cPLA2 seems to be specific and not simply due to an interaction with a hydrophobic sequence, since there is no binding to the catalytic domain of cPLA2, which lacks the CaLB domain but still contains another hydrophobic domain, the substrate binding site.
The ceramide binding to the CaLB domain is Ca2+ dependent, thus confirming other reports that Ca2+ ions are a prerequisite for lipid binding to the CaLB domain. It is worth noting that in the absence of Ca2+, neither a binding of cPLA2 to TID-ceramide, a translocation of cPLA2 to ceramide-containing liposomes, nor an activation of cPLA2 in a cell free system was observed. Thus, there is a close correlation between cPLA2 binding to ceramide and the increase of enzymatic activity.
When [3H]arachidonic acid-labeled renal mesangial cells were stimulated with exogenous ceramide of different chain lengths, we observed a dose-dependent release of [3H]arachidonic acid after 60 min of stimulation. A similar effect of ceramide was also reported for breast cancer epithelial cells, P388D1 macrophages, and platelets.
It has been reported that natural ceramide modulates cobra venom PLA2 activity in vitro by perturbing the structure of phospholipid bilayers. We show here for the first time that the group IV cPLA2, but not the group IIA sPLA2, is modulated directly by ceramide.
We observed that liposomes containing SAPC and sphingomyelin did not
show an increased activation of cPLA2 as compared
to SAPC liposomes alone. However, the simultaneous addition of
bacterial sphingomyelinase, which cleaves sphingomyelin to generate
ceramide, led to a significant increase in cPLA2
activation and SAPC hydrolysis. These data clearly indicate that
generation of ceramide does facilitate cPLA2
activation, probably by attracting cPLA2 to the
membrane and thereby providing improved access to its substrate SAPC.
This conclusion is stressed by the findings that 1)
synthetic TID-ceramide directly binds to cPLA2
in vitro (Fig. 1)
, and 2) ceramide-containing
liposomes cause an increased translocation of
cPLA2 to the liposomes.
A prerequisite for the demonstrated lipid–protein interaction to occur in vivo is that ceramide (generated by the action of a sphingomyelinase) and cPLA2 are located in the same cell compartment. Concerning the cellular localization of cPLA2 several conflicting reports have been provided localizing activated cPLA2 to the plasma membrane, to the nuclear envelope, or to the endoplasmatic reticulum.
Concerning ceramide generation, several studies have suggested that ceramide signaling is initiated in so-called caveolae, which are detergent-insoluble membrane domains especially enriched in sphingolipids. In other reports, neutral sphingomyelinase activity was localized in the plasma membrane as well as in microsomes, and ceramide can also be formed in the lysosomes by the action of an acidic sphingomyelinase.
We have provided evidence in this study for cPLA2 being a novel direct partner for ceramide binding. Ceramide, in turn, causes improved membrane docking of cPLA2, resulting in enhanced arachidonic acid mobilization and prostaglandin synthesis. Tracing ceramide binding partners is important in determining the diverse functional roles of this lipid molecule and may finally provide a means for novel antiinflammatory therapeutic strategies.
|
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.00-0370fje To cite
this article, use (November 9, 2000) FASEB J. 10.1096/fj.00-0370fje 
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
