(Received for publication, December 5, 1996, and in revised form, March 19, 1997)
From the Institut für Organische Chemie und Biochemie der Universität Bonn, Gerhard-Domagk-Strasse 1, 53121 Bonn, Germany and the ¶ Fakultät für Chemie der Universität Konstanz, Postfach 5560, 78434 Konstanz, Germany
The effect of six different structurally modified sphingosine analogues on biosynthesis of sphingolipids was studied in primary cultured murine cerebellar neurons. Treatment of cells with cis-4-methylsphingosine at micromolar levels resulted in a markedly decreased sphingolipid biosynthesis, whereas the other compounds examined, trans-4-methylsphingosine, cis-5-methylsphingosine, trans-5-methylsphingosine, cis-sphingosine, and 1-deoxysphingosine, inhibited sphingolipid biosynthesis less efficiently. The inhibition of sphingolipid biosynthesis by the various compounds was paralleled by a decrease of serine palmitoyltransferase activity in situ. For cis-4-methylsphingosine the inhibitory effect on serine palmitoyltransferase activity was shown to be concentration- and time-dependent. Half-maximal reduction of enzyme activity occurred after 24 h of treatment with 10 µM of the compound. The activity of other enzymes of sphingolipid biosynthesis as well as phospholipid and protein biosynthesis was not affected.
Analysis of the sphingoid moiety of cellular sphingolipids suggests that the sphingosine analogues listed above were subject to degradation rather than being utilized as precursors for sphingolipid biosynthesis by cultured neurons. Except of 1-deoxysphingosine, the other five sphingosine analogues were shown to be substrates for sphingosine kinase in vitro. After 24 h of treatment of primary cerebellar neurons with the various sphingosine analogues the relative percentage of the respective intracellular 1-phosphate derivatives paralleled exactly the inhibitory effect on serine palmitoyltransferase activity observed when cells were treated with the unphosphorylated compounds. In contrast to the respective 1-phosphate derivatives of the other methyl-branched sphingosine analogues examined, cis-4-methylsphingosine 1-phosphate showed an intracellular accumulation suggesting a delayed turnover rate in cultured murine neurons for this compound. These results suggest that the inhibitory effect of the sphingosine analogues on serine palmitoyltransferase is mediated by their respective 1-phosphate derivatives and that the pronounced effect of cis-4-methylsphingosine is caused by a high intracellular concentration of cis-4-methylsphingosine 1-phosphate. cis-4-Methylsphingosine, in addition, caused drastic changes in cell morphology of primary cerebellar neurons, which were not observed when these cells were treated with one of the other sphingosine analogues examined.
Sphingolipids (glycosphingolipids and sphingomyelin) are primarily
plasma membrane molecules of eukaryotic cells. Glycosphingolipids consist of a hydrophobic part, ceramide, which is anchored in the outer
leaflet of the membrane bilayer, and a hydrophilic part composed of
carbohydrate chains extruding into the extracellular space. They are
thought to be involved in a variety of biological processes, such as
cell differentiation and morphogenesis (1); binding of epitopes of
viruses, bacteria, and toxins (2); and cell type-specific adhesion
processes (3), most of which are not clearly understood as of yet. In
addition, evidence emerged during the past years that sphingolipid
(SL)1 metabolites such as sphingosine (4),
ceramide (5), and sphingosine 1-phosphate (6) play an important role as
intracellular signaling molecules for a variety of different targets
(for review, see Ref. 7). Ceramide appears to be a mediator of both the inflammatory and the apoptotic response to tumor necrosis factor- (7) while sphingosine, besides its action as a potent inhibitor of
protein kinase C (8), as well as sphingosine 1-phosphate are thought to
act as mitogenic second messengers (9, 10). Therefore, the combined
action of different sphingolipid signaling molecules seems to regulate,
at least in part, the apoptotic and mitogenic cellular response.
The identification of factors interfering with SL synthesis and the examination of their mode of action is of great importance for studying the regulation of SL. Much of the current knowledge on biosynthesis of SL and regulation of their metabolism has been derived from studies with compounds inhibiting certain steps of the SL synthesis pathway (11-15). SL biosynthesis begins with the condensation of serine with palmitoyl-CoA to form 3-dehydrosphinganine and proceeds to the synthesis of ceramide which is then stepwise glycosylated in the Golgi compartment to form more complex glycosphingolipids (for review, see Ref. 16). The first step takes place in the cytosolic leaflet of the endoplasmic reticulum and is catalyzed by serine palmitoyltransferase (17). Studies from others and from our laboratory suggest that serine palmitoyltransferase might be a possible point of regulation of SL biosynthesis. It was shown that sphingosines with different chain lengths as well as azidosphingosine decreased the de novo SL biosynthesis in murine cerebellar neurons (18) and that treatment of cultured cerebellar cells with these compounds down-regulated serine palmitoyltransferase activity (19).
To investigate further the response of SL biosynthesis to various exogenous sphingoid bases we investigated in the present study the effect of four different methyl-branched sphingosine analogues as well as cis-sphingosine and 1-deoxysphingosine on SL biosynthesis in primary cultured neurons. We demonstrate that treatment of the cells with cis-4-methylsphingosine causes a strong inhibitory effect on serine palmitoyltransferase activity resulting in a decreased de novo synthesis of sphingolipids. In addition, drastic changes in cell morphology were observed using this compound. All four methyl-branched sphingosine analogues were found to be substrates of sphingosine kinase, however, only the 1-phosphate derivative of cis-4-methylsphingosine was shown to accumulate in the cell, suggesting that cis-4-methylsphingosine 1-phosphate is the active metabolite causing the physiological and morphological changes observed.
Six-day-old NMRI (Navy Marine Research Institute)
mice were obtained from Dr. Karzel of the Institute of Pharmacology of
the University of Bonn, Germany.
L-[3-14C]Serine (2.0 GBq/mmol),
UDP-[14C]glucose (10.5 GBq/mmol),
L-[U-14C]phenylalanine (16.6 GBq/mmol),
[1-14C]linoleic acid (2.03 GBq/mmol),
[9,10-3H]palmitic acid (2.0 TBq/mmol), and
[-32P]ATP (110 TBq/mmol) were purchased from Amersham
Corp. (Braunschweig, Germany).
NaH2[32P]O4 (3.7-37 GBq/mmol)
was obtained from ICN (Meckenheim, Germany).
The methyl-branched sphingosines cis-4-methylsphingosine, cis-5-methylsphingosine, trans-4-methylsphingosine, and trans-5-methylsphingosine as well as 1-deoxysphingosine were synthesized in the laboratory of Dr. R. R. Schmidt of the Faculty of Chemistry of the University of Konstanz, Germany, as described previously (20). cis-Sphingosine and sphingosine were synthesized according to Zimmermann and Schmidt (21). The purity of all sphingoid bases examined was at least 98%. C18-sphingosine (62.4 TBq/mol) was tritium-labeled on carbon 3 (22) and C18-sphinganine (12.5 TBq/mol) was tritium-labeled by reduction of the 4.5-double bond according to Schwarzmann and Sandhoff (23).
C12C12-ceramide was obtained by N-acylation of C12-sphingosine, as described (24). C12-sphingosine was synthesized in the laboratory of Dr. R. R. Schmidt, University of Konstanz, according to methods previously published (21).
Dulbecco's modified Eagle's medium, trypsin, deoxyribonuclease, bovine serum albumin, and horse serum were purchased from Life Technologies, Inc. (Karlsruhe, Germany). Thin layer Silica Gel 60 plates and LiChroprep RP18 were supplied by Merck (Darmstadt, Germany). Sephadex G-25 superfine was purchased from Pharmacia (Freiburg, Germany). The scintillation mixture Pico Fluor 40 was supplied by Packard (Frankfurt, Germany). All other chemicals were of analytical grade and obtained from Sigma (München, Germany) or Merck (Darmstadt, Germany).
Cell CultureGranule cells were cultured from cerebella of
6-day-old mice according to the method of Trenkner and Sidman (25).
Cells were isolated by mild trypsinization (0.05%, w/v) and
dissociated by repeated passage through a constricted Pasteur pipette
in a DNase solution (0.1%, w/v). The cells were then suspended in
Dulbecco's modified Eagle's medium containing 10% heat-inactivated
horse serum supplemented with 100 units/ml penicillin and 100 µg/ml streptomycin and plated onto poly-L-lysine-coated 35-mm
diameter Petri dishes (6 × 106 cells/dish). 24 h
after plating cytosine arabinoside was added to the medium (4 × 105 M) to arrest the division of non-neuronal
cells (26). Cell viability was determined by measuring trypan blue
exclusion or the release of cytosolic lactate dehydrogenase activity
using a commercially available test kit (Boehringer Mannheim,
Germany).
After 4 days in culture, murine cerebellar neurons were incubated with sphingoid bases added as complexes with bovine serum albumin to the culture medium containing 0.3% heat-inactivated horse serum.
Metabolic labeling of sphingolipids was performed as described previously (18). Newly synthesized sphingolipids were labeled by feeding cells with [14C]serine (2 µCi/ml), [3H]C18-sphingosine (2 µCi/ml), and [3H]C18-sphinganine (2 µCi/ml), respectively. After the appropriate labeling time as given under "Results" and in the legends of tables and figures, the cells were harvested and lipids were extracted from cell pellets by incubating the cells with 3 ml of chloroform/methanol/water (10:5:1, v/v) for 24 h at 50 °C. Phospholipids were degraded by mild alkaline hydrolysis with methanolic NaOH (100 mM) for 2 h at 37 °C. The lipid extracts were desalted by reversed-phase chromatography on LiChroprep RP18, applied to TLC plates, and chromatographed with chloroform, methanol, 0.22% CaCl2 (60:35:8, v/v). Sphingolipids were visualized by fluorography and identified by their RF values and enzymatic digestion (18). Radioactive bands were determined by radioscanning and scraped off the TLC plates for additional measurement by liquid scintillation counting.
Metabolic labeling of phospholipids was achieved by treatment of the cells with [14C]linoleic acid (1 µCi/ml) or [3H]palmitic acid (3.4 µCi/ml). The labeled fatty acids were added as complexes with bovine serum albumin. After extraction, the phospholipids were separated by TLC with chloroform/methanol/acetic acid/water (65:25:2:3, v/v).
Enzyme AssaysSerine palmitoyltransferase was measured as described previously (19) using [14C]serine and palmitoyl-CoA as substrates. In a total volume of 100 µl the assay mixture contained 0.1 M Hepes (pH 7.4), 5 mM dithiothreitol, 10 mM EDTA, 50 µM pyridoxal phosphate, 1.2 mM [14C]serine (1.6 µCi), 0.15 mM palmitoyl-CoA, and 120 µg of cell protein. After incubation for 10 min at 37 °C the reaction was terminated by addition of chloroform/methanol (5:3, v/v). The lipids were extracted by phase separation and applied to a TLC plate using chloroform, methanol, 2 M NH3 (40:10:1, v/v) as solvent system. The condensation product 3-dehydrosphinganine was detected by scanning the radioactivity, and the specific band was scraped off the TLC plate and determined by liquid scintillation counting.
Ceramide synthase was assayed using D[-erythro-4,5-3H]sphinganine and stearoyl-CoA as substrates. The reaction mixture in a total volume of 80 µl contained 0.1 M Tris buffered with sodium acetate at pH 7.4, 0.5 mM dithiothreitol, 100 µM stearoyl-CoA, 50 µM of the labeled sphinganine (0.5 µCi), which was previously sonicated for 2 min at 0 °C in the buffer solution, and 120 µg of cell protein. After incubation for 15 min at 37 °C, the lipids were extracted, separated by TLC (chloroform/methanol/water, 80:10:1, v/v), and the radiolabeled ceramide was determined by liquid scintillation counting as described above.
Glucosylceramide synthase was assayed in a total volume of 50 µl, containing 500 µM truncated ceramide (C12C12-ceramide), 125 µg of CHAPS, 50 mM MOPS buffer (pH 7.2), 5 mM Mn2+, 2.5 mM Mg2+, 5 mM 2-mercaptoethanol, 1 mM NADPH, 50 µM UDP-[14C]glucose (80,000 cpm/nmol), and 120 µg of cell protein. After incubation for 10 min at 37 °C the reaction was stopped by adding 1 ml of chloroform/methanol (2:1, v/v). The reaction mixture was desalted using a 1-ml Sephadex G-25 column. The column was washed twice with chloroform/methanol (2:1, v/v). The combined effluents were evaporated and the radioactivity was measured by liquid scintillation counting.
Sphingosine kinase was prepared and determined essentially as described
by Olivera et al. (27). Briefly, cells were washed with cold
phosphate-buffered saline and scraped in 0.1 M phosphate buffer (pH 7.4) containing 20% glycerol, 1 mM
mercaptoethanol, 1 mM EDTA, 1 mM sodium
orthovanadate, 15 mM sodium fluoride, 10 µg/ml leupeptin
and aprotinin, respectively, 1 mM phenylmethylsulfonyl fluoride, and 0.5 mM 4-deoxypyridoxine. Cells were then
lysed by freeze-thawing, centrifuged at 105,000 × g
for 90 min, and the supernatants (cytosol) were stored at 70 °C.
The protein concentration of supernatants was about 1 mg/ml.
For the in vitro sphingosine kinase assay 50 µg of cytosol
were used. 50 µM of each of the sphingosine analogues was
added as substrate, respectively, as a complex with bovine serum
albumin. The reaction mixtures contained 20 mM Tris buffer
(pH 7.4), 5% glycerol, 1 mM mercaptoethanol, 0.25 mM EDTA, 1 mM sodium orthovanadate, 15 mM sodium fluoride, 10 µg/ml leupeptin and aprotinin, 1 mM phenylmethylsulfonyl fluoride, and 0.5 mM
4-deoxypyridoxine in a final volume of 200 µl. The reaction was
started by addition of 10 µl of [-32P]ATP (1-2
µCi, 20 mM) and MgCl2 (100 mM)
and was allowed to continue for 30 min at 37 °C. Reaction
termination was achieved by addition of 20 µl of 1 N HCl
followed by 0.8 ml of chloroform/methanol/HCl (100:200:1, v/v). After
vigorous vortexing, 240 µl of 2 N KCl was added and
phases separated by centrifugation. The labeled lipids in the organic
phase were resolved by TLC on Silica Gel G-60 with 1-butanol, methanol,
acetic acid, water (80:20:10:2, v/v) and visualized by autoradiography.
20 nmol of sphingosine 1-phosphate (SPP) was added to each sample and
visualized by spraying with molybdenum blue as described (28). The
radioactive spots corresponding to authentic SPP were identified by
their RF value.
Primary cultured neurons were washed with phosphate-free Dulbecco's modified Eagle's medium (Life Technologies, Inc., Karlsruhe, Germany) and then incubated in the same basic medium supplemented with 40 µCi/ml 32Pi for 24 h as described (28). The cells were treated with 10 µM sphingosine, sphingosine 1-phosphate, or one of the sphingosine analogues, respectively, for different time intervals. The cells were then placed on ice, the medium was removed, and cells were harvested. Sphingosine phosphates were extracted as described by Yatomi et al. (29). Briefly, the cell pellet from each dish was suspended in 500 µl of 20 mM Hepes buffer (pH 7.4) containing 138 mM NaCl, 3.3 mM NaH2PO4, 2.9 mM KCl, 1 mM MgCl, and 1 mg/ml glucose. After addition of 3 ml of chloroform/methanol (1:2, v/v), samples were vigorously vortexed and sonicated for 30 min. Phases were separated by adding 2 ml of chloroform, 2 ml of 1 M KCl, and 100 µl of 7 N NH4OH. The alkaline upper phases containing 90% of sphingosine phosphates were transferred to new tubes, to which 3 ml of chloroform and 200 µl of concentrated HCl were added. Samples were vigorously vortexed and phases separated. The lower chloroform phases, formed under these new acidic conditions, were resolved by TLC in 1-butanol, methanol, acetic acid, water (80:20:10:20, v/v). Phosphorylated sphingosines were visualized by autoradiography and identified by their RF value.
Protein AssaysNewly synthesized total cellular proteins were estimated by measuring the incorporation of [14C]phenylalanine and [14C]serine into trichloroacetic acid-precipitable cell material. The radioactivity incorporated was measured by liquid scintillation counting. Total protein was quantified as described by Bradford using bovine serum albumin as standard (30).
Mass Measurements of Free Sphingoid Bases and SphingolipidsAfter treatment of cells with the respective sphingosine analogue, mass measurements of free sphingoid bases were conducted by HPLC as described previously (31) with C14-sphinganine as an internal standard. Briefly, lipids were extracted from cell homogenates and lyophilized cell culture media. Phospholipids were removed from the lipid extracts by hydrolysis in 50 mM NaOH (2 h, 37 °C). The free sphingoid bases were derivatized with o-phthaldialdehyde and determined by HPLC. cis-4-Methylsphingosine and trans-4-methylsphingosine were eluted with 5 mM phosphate buffer (pH 7.0)/methanol (1:9, v/v), whereas cis-sphingosine, cis- and trans-5-methylsphingosine, and 1-deoxysphingosine were eluted with 5 mM phosphate buffer (pH 7.0), methanol (1.7:9, v/v).
For mass determination of total sphingolipids the lipid extracts were subjected to acid hydrolysis according to Gaver and Sweeley (32) with 0.5 N argon-saturated methanolic HCl for 16 h at 50 °C. By this procedure sphingoid bases were released from glycosphingolipid, ceramides, and sphingomyelin, and then measured by HPLC as described above. The loss of compounds due to extraction or hydrolysis procedure was equilized by a defined amount of C12C12-ceramide which was added to each sample as an internal standard.
Primary cultured murine cerebellar neurons were
treated with 10 µM cis-4-methylsphingosine,
cis-5-methylsphingosine, cis-sphingosine, trans-4-methylsphingosine, and
trans-5-methylsphingosine as well as 1-deoxysphingosine,
respectively. The chemical structures of these compounds are given in
Fig. 1. After 24 h of incubation, [14C]serine, a precursor of sphingolipid biosynthesis,
was added to the culture medium and incorporation into newly
synthesized sphingolipids was measured.
In the presence of 10 µM cis-4-methylsphingosine the incorporation of [14C]serine into cellular sphingolipids was reduced to 40% of untreated controls (Table I, Fig. 2). In contrast, the structurally related compounds trans-4-methylsphingosine, cis-5-methylsphingosine, trans-5-methylsphingosine, cis-sphingosine, and 1-deoxysphingosine showed much less of an effect, if any, on [14C]serine incorporation. It is noteworthy, that trans-5-methylsphingosine treatment caused a band pattern comigrating with sphingomyelin which in this experiment was quite different compared with the other sphingosine analogues (lane 5 of Fig. 2). It is most unlikely that this difference in migration is due to an incorporation of trans-5-methylsphingosine into sphingomyelin as demonstrated in Fig. 8. At present, we cannot provide a conclusive explanation for this migration difference.
|
The decrease of [14C]serine incorporation into SL caused
by cis-4-methylsphingosine was
concentration-dependent, as shown in Fig. 3.
The decrease of [14C]serine incorporation after 24 h
preincubation was about 15% when cells were treated with 1 µM cis-4-methylsphingosine and more than 90%
when cells were treated with 100 µM of the compound. Treatment with 5, 10, 20, and 50 µM showed a decrease of
[14C]serine incorporation of about 25, 60, 80 and 90%,
respectively, as compared with control cells. The time dependence of
[14C]serine incorporation is depicted in Fig.
4. Preincubation of the cells with 10 µM
cis-4-methylsphingosine for 24 h caused a decrease in
[14C]serine incorporation of about 60%, while the
decrease of label incorporation was more than 90% after 48 h
preincubation, compared with control cells. At concentrations above 30 µM for 24 h of preincubation or longer preincubation
times (10 µM for 48 h)
cis-4-methylsphingosine exhibited a cytotoxic effect onto
murine neuronal cells based on trypan blue exclusion assay and
morphology as assessed by light microscopy.
To differentiate whether the decreased [14C]serine labeling of SL was due to increased degradation or decreased biosynthesis of SL, pulse-chase experiments were performed. Cells were pulsed with [14C]serine for 24 h; cells were then chased in a medium containing an excess of unlabeled serine either in the absence or presence of cis-4-methylsphingosine. No changes in [14C]serine labeling were observed up to 48 h of chase in the presence of 10 µM cis-4-methylsphingosine (data not shown). These findings strongly suggest that the effect of cis-4-methylsphingosine on sphingolipid labeling is due to decreased de novo biosynthesis and not to increased degradation of SL.
The synthesis step affected by cis-4-methylsphingosine was
determined by incubating cells with [3H]sphingosine or
[3H]sphinganine instead of [14C]serine.
When labeled with these compounds, sphingolipid biosynthesis was not
affected by cis-4-methylsphingosine, since the incorporation of label was similar in cis-methylsphingosine-treated and
untreated cells (Fig. 5). The biosynthetic pathway of
sphingolipids begins with the serine palmitoyltransferase catalyzed
condensation of serine and palmitoyl-CoA to form 3-dehydrosphinganine,
which is then reduced by a NADPH-dependent reductase to
sphinganine. Since incorporation of labeled sphinganine did not show
any reproducible differences in cells with or without treatment with
cis-4-methylsphingosine, these results suggest that the
blocked step is prior to sphinganine formation.
cis-4-Methylsphingosine Specifically Interferes with Serine Palmitoyltransferase Activity
The above results indicate that cis-4-methylsphingosine blocks one of the initial steps of sphinganine synthesis. The activity of serine palmitoyltransferase, the enzyme which catalyzes the first synthetic step of SL, was therefore measured in the homogenate of cells treated with one of the six different sphingosine analogues for 24 h, respectively. The results are given in Table I. 10 µM cis-4-Methylsphingosine reduced the activity of serine palmitoyltransferase by about 50%, while its trans-analogue was less effective, causing a 30% reduction. The other four compounds tested did not significantly influence serine palmitoyltransferase activity. The decrease of enzymatic activity observed with the six different sphingosine analogues corresponded closely to their effect on incorporation of labeled serine into cellular sphingolipids, as demonstrated in Table I. These results indicate that the effect of cis-4-methylsphingosine treatment on SL biosynthesis is primarily caused by a decrease of serine palmitoyltransferase activity.
To ensure that only serine palmitoyltransferase and no other key enzymes of the sphingolipid pathway are affected by cis-4-methylsphingosine we measured the activities of ceramide synthase, which is also localized on the cytosolic face of the endoplasmic reticulum, and glucosylceramide synthase, an enzyme resident in the Golgi compartment. Neither ceramide synthase nor glucosylceramide synthase activities were affected after 24 h treatment of the cells with 10 µM cis-4-methylsphingosine. Moreover, treatment of cells with one of the six sphingosine analogues at a concentration of 10 µM for 48 h did not influence the incorporation of [14C]phenylalanine and [14C]serine into trichloroacetic acid-precipitable cell material, suggesting that neither of these compounds affect the overall protein biosynthesis. In addition, cis-4-methylsphingosine caused no detectable inhibition of the incorporation of [3H]palmitic acid and of [14C]linoleic acid into cellular phospholipids. Moreover, as demonstrated above (Fig. 5), cis-4-methylsphingosine treatment of the cells did not affect the metabolism of [3H]sphingosine and [3H]sphinganine significantly, suggesting that cis-4-methylsphingosine does not inhibit other sphingolipid synthesizing enzymes in vivo. Taken together these results strongly suggest that the inhibition of serine palmitoyltransferase activity by cis-4-methylsphingosine is highly specific.
The cis-4-Methylsphingosine Effect on Serine Palmitoyltransferase Requires Cellular IntegrityTo determine whether the mode of action of cis-4-methylsphingosine in decreasing serine palmitoyltransferase activity is performed by a direct interaction with the enzyme or requires cellular integrity, the enzymatic activity was measured in vitro in total cell homogenates of cultured murine cerebellar neurons in the absence and presence of this compound. No reduction of serine palmitoyltransferase activity was observed in vitro even at high concentrations of cis-4-methylsphingosine up to 1 mM. Thus, a decrease of enzymatic activity occurs only in intact cells which makes a direct inhibition of serine palmitoyltransferase by this compound unlikely.
Additionally, we investigated the effect of cycloheximide, an inhibitor
of protein biosynthesis, on serine palmitoyltransferase activity in the
presence or absence of cis-4-methylsphingosine as a function
of time (Fig. 6). The concentration used for
cycloheximide (0.5 mM) completely inhibited incorporation
of [14C]phenylalanine into trichloroacetic
acid-precipitable material after 3 h. Average half-life
(t1/2) of serine palmitoyltransferase under these
conditions was found to be 20 ± 3 h (mean value of five
separate experiments). In the presence of both drugs, the serine
palmitoyltransferase activity of murine cerebellar neurons decays with
first-order kinetics. The inhibitory effects of cycloheximide and
cis-4-methylsphingosine on serine palmitoyltransferase
activity were not found to be additive (Fig. 6), suggesting that both
drugs similarly interfere with serine palmitoyltransferase biosynthesis pathway rather than inhibiting serine palmitoyltransferase activity by
direct interaction with the enzyme.
The Six Sphingosine Analogues Tested Are Not Utilized in SL Biosynthesis
To determine whether the effects on sphingolipid
biosynthesis were due to cis-4-methylsphingosine itself or
possibly caused by potential biosynthetic metabolites instead, we
estimated its cellular content by HPLC measurement before and after
acid hydrolysis of the cellular lipid extracts. Acid hydrolysis causes
cleavage of the sugar chains and N-acyl groups of SL,
thereby allowing total sphingolipid mass to be measured by HPLC of the
o-phthalaldehyde derivatives of the long chain bases.
Determination of the mass of total cellular sphingolipids as assessed
by HPLC measurement of sphinganine and sphingosine after acid
hydrolysis did not change after treatment with
cis-4-methylsphingosine when compared with untreated
controls, although cis-4-methylsphingosine caused a 60%
reduction of de novo SL synthesis as assessed by
incorporation of labeled serine into cellular sphingolipids after
48 h (24 h of preincubation plus 24 h of labeling, Table I).
As demonstrated in Fig. 7, sphingosine content increased
from less than 1 nmol/mg protein to 14 nmol/mg protein and sphinganine
content increased from an almost undetectable mass up to 2 nmol/mg
protein. In contrast, no change was observed in the amount of
cis-4-methylsphingosine which remained at 5 nmol/mg protein
after acid hydrolysis. Similar results were obtained after treatment of
cells with the other five sphingosine analogues. As shown in Fig.
8, after 24 h of treatment with 10 µM
of the respective sphingosine analogue the mass of free and total (free
plus SL associated) of each sphingosine analogue did not differ
significantly, suggesting that none of the sphingoid bases were used as
precursor for SL biosynthesis.
The cellular uptake of all six sphingosine analogues was nearly quantitative as was determined by the measurement of their amount in culture media after 24 h of incubation time. After 24 h of incubation, except of 1-deoxysphingosine only trace amounts of the other five sphingosine analogues were detectable in the cellular lipid extracts, suggesting the compounds being subject to degradation (Fig. 8). In contrast, 1-deoxysphingosine was almost quantitatively recovered in the lipid extracts of the cells.
cis-4-Methylsphingosine Is Metabolized to Its 1-Phosphate Derivative Which Accumulates in the CellThe above results
indicate that degradation of sphingosine analogues appears likely
rather than utilization for SL biosynthesis. We therefore tested the
six sphingosine analogues for being substrates of sphingosine kinase,
which is the first enzyme in the degradation pathway of sphingosine
leading to the respective 1-phosphate derivative. As shown in Fig.
9, except of 1-deoxysphingosine, which was almost fully
recovered from cellular lipid extracts, the other five sphingosine analogues tested were efficiently phosphorylated in vitro.
However, relative to the trans-analogues the
cis-analogues were found to be less favored substrates for
sphingosine kinase in vitro, suggesting that the
cis/trans-configuration of the sphingosine is more important for this phosphorylation reaction than the presence and the position of
an additional methyl group.
Based on these results we determined the fate of the phosphorylated
analogues in cultured murine cerebellar neurons. As illustrated in Fig.
10, after 13 h of incubation the level of
cis-4-methylsphingosine phosphate was substantially elevated
compared with that of the other phosphorylated analogues. Relative to
cis-4-methylsphingosine-1-phosphate (100%) its
trans-analogue represented 20% and
cis-5-methylsphingosine-1-phosphate about 15%, while the
remaining sphingosine analogues did not exceed 5-10%. A similar
relation was observed after a 24-h incubation time. These results show
that the metabolic turnover of cis-4-methylsphingosine 1-phosphate was slow relative to the other sphingoid phosphates tested
and that this derivative accumulates in the cell. It is noteworthy that
the ranking of the sphingosine analogues in terms of the percentage of
the phosphorylated metabolites is in the same relative order and
magnitude as the effects of these compounds on serine
palmitoyltransferase activity, as given in Table I.
Treatment of Murine Cerebellar Neurons with cis-4-Methylsphingosine Is Accompanied with Changes in Cell Morphology
Light microscopy
of the cultured murine cerebellar neurons revealed dramatic changes in
cell morphology when cells were treated with
cis-4-methylsphingosine. Before treatment, granule neurons were fully differentiated and developed a rich network of fine fibers
as shown in Fig. 11a. Incubation with
trans-4-methylsphingosine, cis-5-methylsphingosine,
trans-5-methylsphingosine, cis-sphingosine, and
1-deoxysphingosine, respectively, did not cause any visible differences
in cell morphology between treated and untreated cells as examined by
light microscopy, even at higher concentrations (30 µM
for 24 h) or longer incubation times (10 µM for
72 h). In contrast, treatment of cerebellar neurons with 10 µM cis-4-methylsphingosine for 24 h
caused reaggregation of the former granule cells beginning 6 h
after addition of the compound to the medium (Fig. 11b).
These aggregates were connected by radial strongly fasciculated
neurites, which were not found in control cultures. Cell aggregates
remained stable for approximately 48 h, then neurites became
fragmented and finally cells died after about 72 h. When SL
biosynthesis was re-established by feeding radioactively labeled
sphingosine or sphinganine instead of serine, no reconstitution of the
morphology of control cells was observed. These results indicate that
decreased SL biosynthesis is most likely not responsible for the
morphological changes observed.
To exclude the possibility of an immediate cytotoxic effect of the sphingosine analogues under the conditions described above, parameters of cell viability were determined by measurement of lactate dehydrogenase release into the culture medium and trypan blue exclusion and were compared with the results obtained after cell lysis with 0.5% Triton X-100. As assessed by these parameters, the cells remained viable up to a concentration of 20 µM and 48 h of treatment even in case of treatment with cis-4-methylsphingosine, where drastic changes in cell morphology were observed.
Previous results from our laboratory have shown that exogenous sphingosine homologues of different chain length (12 and 18 carbon atoms) cause a decrease in serine palmitoyltransferase activity in primary cultured cerebellar neurons in a time- and concentration-dependent manner (19). Serine palmitoyltransferase catalyzes the first step in sphingolipid biosynthesis and is thought to catalyze the rate-limiting reaction in long chain base synthesis. A maximum decrease of approximately 80% in this enzyme's activity was achieved with doses as high as 50 µM sphingosine homologues. We also demonstrated that sphingosine homologues are utilized as precursors for the synthesis of complex sphingolipids (18). In the present study we now examined the effects of six structurally modified sphingosine analogues on sphingolipid metabolism. Four of these compounds were either cis- or trans-configurated sphingosines carrying a methyl group either at carbon atom 4 or 5. The two other analogues were cis-configurated sphingosine and 1-deoxysphingosine. cis-4-Methylsphingosine (10 µM) was found to exhibit a pronounced inhibitory effect on both de novo sphingolipid synthesis and serine palmitoyltransferase activity, the latter being decreased by about 50%. The inhibition of serine palmitoyltransferase activity by cis-4-methylsphingosine is highly specific as has been demonstrated for long chain sphingoid bases before (19). However, taken into account that the dose of cis-4-methylsphingosine was only 10 µM, its efficiency in inhibiting serine palmitoyltransferase activity is much greater compared with other long chain sphingoid bases lacking a methyl group at carbon atom 4.
cis-Configuration and/or introducing a methyl group on either carbon atom 4 or 5 of sphingosine seems to prevent the compound of being utilized for biosynthesis of more complex sphingolipids. After uptake by the cell none of the cis-configurated and methyl-branched sphingosine analogues were found to be channeled into the biosynthetic pathway. Rather, all of these compounds were subject to immediate degradation. Catabolism of sphingosine requires two independent and sequential steps. First, the hydroxyl on the first carbon of sphingosine is phosphorylated by sphingosine kinase (33, 34), followed by the cleavage of the resulting sphingosine 1-phosphate into ethanolamine phosphate and the corresponding aldehyde, the latter step mediated by sphingosine-phosphate lyase (35). Since sphingosine 1-phosphate turnover is extremely rapid and the compound is detectable in the cell only in a very low abundance, the action of sphingosine-kinase is considered as being the rate-limiting step of the degradation pathway (36, 37).
Surprisingly, the 1-phosphate derivative of cis-4-methylsphingosine was found in much higher abundance than the other methyl-branched sphingosine 1-phosphates. Since all of these compounds were efficiently phosphorylated by sphingosine kinase it appears most likely that cis-4-methylsphingosine 1-phosphate poorly qualifies as substrate for sphingosine-phosphate lyase, thereby accumulating in the cell. Data on sphingosine-phosphate lyase are poor due to its low activity and the difficulty of obtaining the phosphorylated sphingoid bases to study as substrates (35). Sphingosine-phosphate lyase in rat liver has been shown to act stereospecifically only on D-(+)-erythro (2D,3D)-isomers (37). Since all of the methyl-branched sphingosine analogues used in this study were in fact D-(+)-erythro-isomers and trans-4-methylsphingosine 1-phosphate as well as cis-sphingosine 1-phosphate did not accumulate in the cell, the cis-configuration in combination with the methyl group at position 4 seems to prevent the compound of being efficiently cleaved by sphingosine-phosphate lyase. In contrast, cis-sphingosine 1-phosphate methylated at position 5 did not show accumulation and was therefore apparently not sufficient to inhibit the enzyme reaction kinetics substantially. Another possible explanation for the accumulation of cis-4-methylsphingosine 1-phosphate might be this compound acting as an inhibitor of sphingosine-phosphate lyase. It has been proposed that N-methylated sphingosine phosphate could be a potent inhibitor of this enzyme (35). Further studies are to be performed to clarify the mode of action.
The finding of an accumulation of cis-4-methylsphingosine 1-phosphate in primary cultured cerebellar neurons raises the question, whether cis-4-methylsphingosine itself or its 1-phosphate derivative is responsible for the pronounced inhibitory effect on serine palmitoyltransferase activity. Differentiation of the contribution of either compound to enzyme activity inhibition would ask for specifically inhibiting sphingosine kinase to prevent formation of cis-4-methylsphingosine 1-phosphate. However, a specific sphingosine kinase inhibitor not structurally related to sphingoid bases and not exhibiting inhibition of other enzyme activities is, to our knowledge, not known. Nevertheless, our results revealed a striking correlation of the relative order and magnitude of the sphingosine analogues in terms of the percentage of the phosphorylated metabolites in vivo and the inhibitory effect of the respective compound on serine palmitoyltransferase activity, suggesting that the 1-phosphate derivative is the active metabolite mediating the inhibitory effect on SPT activity.
Our results clearly demonstrate that the inhibitory effect of cis-4-methylsphingosine and/or its 1-phosphate derivative on serine palmitoyltransferase is not being exhibited by a direct molecule to molecule interaction but appears to be mediated by a more complex mechanism instead. The exact mode of action remains unknown at present. In a previous study we demonstrated that the decrease of serine palmitoyltransferase activity observed after sphingosine treatment in cultured primary cerebellar neurons was not caused by a direct feedback inhibition (19). Other mechanisms might be involved. Sphingosine as well as sphingosine 1-phosphate have been shown to play an important role in cell growth regulation by a pathway that is independent of protein kinase C (9, 28). Sphingosine 1-phosphate, rather than sphingosine, has been proposed to act as second messenger in a mitogen-activated protein kinase-dependent signal transduction pathway (38), providing a link between the plasma membrane carrying the growth factor receptors, calcium mobilization from intracellular stores, and cellular proliferation (10, 28, 39). Treatment of Swiss 3T3 fibroblasts with sphingosine was shown to cause a mitogenic response which was primarily mediated by the conversion of sphingosine to its 1-phosphate derivative (40). Moreover, the mitogenic response of Swiss 3T3 fibroblasts to platelet-derived growth factor as well as fetal calf serum was demonstrated to be paralleled by a rapid and transient increase of sphingosine 1-phosphate (39). In the present study, the response of primary cultured mouse cerebellar neurons on cis-4-methylsphingosine treatment also demonstrated a marked effect on cell growth, which, however, was apoptotic rather than mitogenic, as primarily assessed by DNA fragmentation assays.2 This suggests that cis-4-methylsphingosine and its 1-phosphate derivative might act as substitutes in the same way as sphingosine and sphingosine 1-phosphate in mediating cellular responses in terms of cell growth regulation as well as inhibition of serine palmitoyltransferase activity. The unusual accumulation of cis-4-methylsphingosine 1-phosphate might mimic a persistent activation of the sphingolipid-based signal transduction pathway which could provide a plausible explanation for the pronounced effect observed upon the relatively small dose of 10 µM unphosphorylated cis-4-methylsphingosine.
Recent results have shown that the sphingosine-like immunosupressant ISP-1/myriocin induces apoptosis in the mouse cytotoxic T-cell line CTLL-2 and this was linked to the inhibitory effect on serine palmitoyltransferase activity by this compound (41). Since in that study the ISP-1/myriocin-induced apoptosis was attenuated by the addition of sphingosine, it was proposed that apoptosis was triggered by the decrease in the intracellular levels of sphingolipids in these cells (41). In our study the decreased sphingolipid biosynthesis is unlikely to have caused the changes in cellular morphology since biosynthesis was restored by feeding labeled sphinganine or sphingosine instead of serine, yet no restitution of or even changes toward a control cell morphology were observed. It is more likely that the marked inhibition of serine palmitoyltransferase activity as well as the apoptotic response of the cells were primarily triggered by the persistently increased level of cis-4-methylsphingosine 1-phosphate; however, from our studies we cannot provide a direct evidence for a linked mode of action between the inhibition of serine palmitoyltransferase activity and the alterations in cell morphology. Rather, both phenomenons might be mediated by independent mechanisms.
The opposite responses observed in Swiss 3T3 cells fed with sphingosine and primary cultured cerebellar neurons treated with cis-4-methylsphingosine might also indicate that the response to sphingolipid based signaling is cell-type specific, covering a wide range of possible cellular responses. In contrast to Swiss 3T3 fibroblasts treatment with sphingosine caused apoptosis in CTLL-2 cells as well as in human leukemic HL-60 cell lines (41, 42). More work is needed on different cell systems to determine the factors involved in mediating sphingolipid metabolism and in the sphingolipid based signaling pathway. For those studies we would like to propose cis-4-methylsphingosine as a promising pharmacological prodrug due to its unusual metabolism.
We thank Judith Weisgerber for help in preparing the figures.