Dual roles of sphingolipids in signaling of the escape from and onset of apoptosis in a mouse cytotoxic T-cell line, CTLL-2.

In our previous study, the sphingosine-like immunosuppressant, ISP-1, was found to suppress the proliferation of an interleukin-2-dependent cytotoxic T cell line, CTLL-2, through the inhibition of serine palmitoyltransferase, which catalyzes the committed step of sphingolipid biosynthesis. Analysis of the effect of ISP-1 by flow cytometry revealed that the ISP-1-dependent decrease in cell number was not due to inhibition of the cell cycle progression of CTLL-2 cells but to the induction of apoptosis of the cells. The ISP-1-induced apoptosis was inhibited by the addition of sphingosine (2 μM), suggesting that this ISP-1-induced apoptosis is triggered by the decrease in the intracellular levels of sphingolipids caused by the inhibition of serine palmitoyltransferase. However, another interleukin-2-dependent cell line, F7, which was derived from a mouse pro-B cell line, did not show ISP-1-dependent apoptosis, indicating that the effect of ISP-1 may be specific for a certain type of T cell lineage such as CTLL-2. On the other hand, a high dose of sphingosine (5 μM) by itself induced the apoptosis of CTLL-2 cells. This sphingosine-dependent apoptosis was also observed with F7 cells. These results provide evidence that the intracellular levels of sphingolipids play an important role in the signaling of the escape from and onset of apoptosis of CTLL-2 cells.

Several lines of evidence have suggested that sphingosines are involved in apoptosis. Exogenously added cell-permeable ceramides induce the apoptosis of U937 human leukemia cells (1) and several other cell lines (2). Later, a natural ceramide was shown to induce apoptosis when added as a cell-permeable form in a mixture of dodecane and ethanol (3). Furthermore, the level of intracellular ceramide was shown to increase through the degradation of sphingomyelin under such conditions that apoptosis was induced by tumor necrosis factor-␣ (4 -6), ␥-irradiation, (7), and Fas ligation (8 -10). Ceramide is, therefore, proposed to be a second messenger for apoptosis. Recently, exogenously added sphingosine was also demonstrated to induce the apoptosis of human neutrophils (11) and HL-60 human leukemia cells (12), and the intracellular level of sphingosine increased in the cases of tumor necrosis factor-␣induced apoptosis of human neutrophils (11) and phorbol esterinduced apoptosis of HL-60 cells (12).
A new potent immunosuppressant, ISP-1, is a fungal product that has a structure very similar to that of sphingosine (13). ISP-1 inhibited the mouse allogenic mixed lymphocyte reaction and allo-reactive cytotoxic T lymphocyte generation in vivo, exhibiting a potency of 10 -100-fold greater than that of cyclosporin A, the most widely used immunosuppressant. Unlike cyclosporin A and FK506 (another popular immunosuppressant), ISP-1 did not interfere with IL-2 1 production in the mixed lymphocyte reaction (13), but instead it suppressed the IL-2-dependent growth of the mouse cytotoxic T cell line, CTLL-2 (14). This suppression is most probably triggered by a reduction of the intracellular levels of sphingolipids due to the inhibition of serine palmitoyltransferase, which catalyzes the first step of sphingolipid biosynthesis, i.e. the condensation of serine and palmitoyl-CoA into ketodihydrosphingosine (14).
In this paper, we report that the overall growth suppression of CTLL-2 cells by ISP-1 is not due to inhibition of cell proliferation but to the apoptosis of the cells caused by inhibition of the de novo synthesis of sphingosine. This is the first evidence that apoptosis is caused by a reduction in the intracellular levels of sphingolipids.

EXPERIMENTAL PROCEDURES
Materials-ISP-1 was obtained as described previously (14) and stored as a methanol solution at Ϫ20°C. Recombinant human IL-2 was kindly provided by Yoshitomi Pharmaceutical Industries, Tokyo. All other biological reagents were purchased from Sigma.
Cells and Cell Culture-The mouse cytotoxic T cell line, CTLL-2 (15), was obtained from the American Type Culture Collection. The F7 cell line (16) was kindly provided by Dr. T. Taniguchi, Tokyo University, Tokyo. These cells were maintained in RPMI 1640 containing 10% fetal calf serum and 50 units/ml IL-2. The administration of ISP-1 and sphingosine to the cells was performed as described previously (14).
Cell Proliferation Assay-Cells were cultured at 5 ϫ 10 4 cells in 500 l of medium in the presence or absence of ISP-1. After incubation for 48 h, the viable cell number was determined by means of the 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay (17). The absorbance increased in proportion to the viable cell number, within the range examined in this experiment.
Flow Cytometry Analysis-Cells were cultured as indicated, harvested, stained with propidium iodide using a Cycle TEST TM PLUS DNA Reagent Kit (Becton Dickinson), and then subjected to flow cytometry with a FACScan TM flow cytometer (Becton Dickinson) for measurement of the cell cycle progression and apoptosis.
DNA Fragmentation Analysis-DNA fragmentation was analyzed by agarose gel electrophoresis as described by Selins and Cohen (18). Briefly, fragmented genomic DNA was extracted by incubating cells in 10 mM Tris-HCl buffer, pH 7.4, containing 10 mM EDTA and 0.5% Triton X-100 at 4°C for 10 min. Under these conditions, intact genomic DNA was not extracted. The cell suspension was centrifuged, and the resultant supernatant was incubated for 1 h at 37°C in the presence of RNase (400 g/ml) and then for an additional 1 h in the presence of proteinase K (400 g/ml). The fragmented DNA was precipitated with isopropyl alcohol and then analyzed by agarose gel electrophoresis.
Cell Morphology-For morphological analysis, cells were fixed with 1% glutaraldehyde overnight and then collected by centrifugation and resuspended in phosphate-buffered saline. The cell nuclei were stained with Hoechst 33342 (160 M) and then examined under a fluorescent microscope (Olympus BX 50).

RESULTS AND DISCUSSION
Although ISP-1 was demonstrated to be a potent inhibitor of serine palmitoyltransferase, which catalyzes the first step of sphingolipid biosynthesis, i.e. the condensation of serine and palmitoyl-CoA into ketodihydrosphingosine (14), the mechanism underlying the inhibition of cell growth by ISP-1 was not clear. In an attempt to elucidate the mechanism underlying the suppression of CTLL-2 cell growth by ISP-1, the effect of ISP-1 on cell cycle progression was investigated by the flow cytometric method. As shown in Fig. 1B, incubation of the cells in the presence of 47 nM ISP-1 for 48 h did not affect the cell cycle progression of CTLL-2 cells significantly but instead dramatically increased the hypodiploid DNA peaks, which are commonly found in apoptosis (19). In order to confirm that the cells were indeed in the process of apoptosis, the chromosomal DNA profiles were investigated. As shown in Fig. 2A, lane 2, genomic DNA obtained from ISP-1-treated cells gave a ladder on agarose gel electrophoresis, which is a characteristic of apoptosisspecific internucleosomal DNA fragmentation (20), while this fragmentation was not seen in mock-treated cells (lane 1). Further confirmation of apoptosis was the morphological changes of nuclei detected in ISP-1-treated cells. As shown in Fig. 3B, ISP-1-treated CTLL-2 cells showed morphological changes typical of apoptosis including condensed chromatin and fragmented nuclei revealed upon DNA staining with Hoechst 33342. All these results indicated that the ISP-1dependent decrease in the number of CTLL-2 cells was not due to inhibition of the proliferation of CTLL-2 cells but to the acceleration of cell death by apoptosis.
In order to determine whether or not inhibition of the condensation of serine and palmitoyl-CoA into ketodihydrosphingosine is the direct trigger of the apoptosis of CTLL-2 cells, the effect of sphingosine, a compound located downstream of ke-todihydrosphingosine in the sphingolipid biosynthesis pathway, on ISP-1-dependent apoptosis was studied. As shown in Fig. 1C, upon the addition of sphingosine (2 M) to the incubation medium containing 47 nM ISP-1, the hypodiploid DNA peaks disappeared almost completely, which is in good agreement with the almost 100% recovery of cell proliferation with the same treatment previously reported (14), suggesting again that the inhibition of cell growth by ISP-1 is caused by the progression of apoptosis due to the reagent.
The ability of sphingosine to counteract the effect of ISP-1 is stereoselective and cannot be mimicked by dimethylsphingosine (14); therefore, it is unlikely to be due to inhibition of protein kinase C, which does not show this structural selectivity (21)(22)(23). These results suggested that exogenously added sphingosines did not function as inhibitors of protein kinase C and that ISP-1-induced apoptosis of CTLL-2 cells was triggered by the decrease in the intracellular levels of sphingolipids due to inhibition of serine palmitoyltransferase. The chemical nature of the downstream effector that is directly involved in apoptosis has not yet been identified. However, Fumonisin B1, which inhibits the conversion of sphingosine to ceramide (24), did not inhibit the effect of exogenously added sphingosine on the ISP-1 action (14), suggesting that ceramide is not the downstream effector and that sphingosine derivatives with a free amino group such as sphingosine and sphingosine 1-phosphate are possible candidates.
Incidentally, it was reported that exogenously added sphingosine itself induced apoptosis in human neutrophils (11) and HL-60 promyelocytic leukemia cells (12) when it was administered at a high concentration of 5-10 M. This was also true for CTLL-2 cells, as shown in Fig. 1D. Taken together, these results suggest that both hypo-and hyperintracellular levels of sphingolipids induced apoptosis of CTLL-2 cells and that the maintenance of the proper levels of these compounds is critically important for the cells to escape from apoptosis. It is also of interest to note that the sphingosine-induced DNA fragmentation was observed within 6 h after the addition of sphingosine (data not shown), while the ISP-1-induced DNA fragmentation was delayed requiring more than 32 h to occur (Fig. 2B). These results suggest that these two types of apoptosis caused by unusually low and high intracellular levels of sphingolipids took place through distinct biochemical pathways, respectively. The ISP-1-induced apoptosis seemed to occur following one or more cell divisions (data not shown). This type of apoptosis was also reported for hematopoietic cell lines that had been exposed to an equitoxic dose of ␥-irradiation and was designated as "delayed mitotic death" (25,26).
ISP-1 inhibits the proliferation of CTLL-2 cells and mouse splenic lymphocytes induced by allogenic antigen stimulation (13). In both cases IL-2 is responsible for their proliferation as the major growth factor. These results prompted us to suspect a direct linkage between the signal transduction through the IL-2 receptor and the function of ISP-1. In order to test this hypothesis, the effect of ISP-1 on the apoptosis of another IL-2-dependent cell line, F7, which was derived from BAF-B03 cells, and acquired IL-2 dependence through the introduction of mouse IL-2 receptor ␤ subunit cDNA (16) was studied. However, unlike CTLL-2 cells, F7 cells did not show apoptosis in response to ISP-1, although a high concentration of sphingosine induced the apoptosis of F7 cells (see Fig. 4C). Furthermore, when CTLL-2 cells were cultured in the presence IL-4 instead of IL-2, ISP-1 also induced the apoptosis of CTLL-2 cells (data not shown). These results indicated that there is no one-to-one direct relationship between the signal transduction through the IL-2 receptor and the function of ISP-1. However, the possibility that F7 cells have a specific ability to incorporate sphingolipids from the fetal calf serum in the culture medium could not be ruled out. Since the proliferation of stimulated T-lymphocytes was inhibited by ISP-1 (14), we then tried to identify the type of cells that is susceptible to ISP-1. When mouse splenic T cells were stimulated with concanavalin A in the presence or absence of ISP-1, CD4-negative and CD8-positive cells were much more sensitive to ISP-1 than CD4-positive and CD8-negative cells (data not shown). Taken together, ISP-1-induced apoptosis is not specific for IL-2-dependent cells but may be a characteristic of a CD8-positive and CD4-negative cytotoxic T cell lineage. In sharp contrast to the cell typespecific expression of ISP-1-induced apoptosis, sphingosine-induced apoptosis is observed in many cell types, including Chinese hamster ovary cells (27), human neutrophils (11), and HL-60 cells (12), as well as the CTLL-2 and F7 cells examined in this study. The effect of sphingosine may be associated with the protein kinase C inhibitory activity of sphingosine (27), because dimethylsphingosine, which is also a protein kinase C inhibitor, induced the apoptosis of CTLL-2 cells (data not shown), as reported in human neutrophils (11) and HL-60 cells (12). However, the detail mechanism of the involvement of protein kinase C in the sphingosine-induced apoptosis remains to be elucidated.