The Immune Modulator FTY720 Inhibits Sphingosine-1-phosphate Lyase Activity*

FTY720 is a novel immunomodulatory agent that inhibits lymphocyte trafficking and prevents allograft rejection. FTY720 is phosphorylated in vivo, and the phosphorylated drug acts as agonist for a family of G protein-coupled receptors that recognize sphingosine 1-phosphate. Evidence suggests that FTY720-phosphate-induced activation of S1P1 is responsible for its mechanism of action. FTY720 was rationally designed by modification of myriocin, a naturally occurring sphingoid base analog that causes immunosuppression by interrupting sphingolipid metabolism. In this study, we examined interactions between FTY720, FTY720-phosphate, and sphingosine-1-phosphate lyase, the enzyme responsible for irreversible sphingosine 1-phosphate degradation. FTY720-phosphate was stable in the presence of active sphingosine-1-phosphate lyase, demonstrating that the lyase does not contribute to FTY720 catabolism. Conversely, FTY720 inhibited sphingosine-1-phosphate lyase activity in vitro. Treatment of mice with FTY720 inhibited tissue sphingosine-1-phosphate lyase activity within 12 h, whereas lyase gene and protein expression were not significantly affected. Tissue sphingosine 1-phosphate levels remained stable or increased throughout treatment. These studies raise the possibility that disruption of sphingosine 1-phosphate metabolism may account for some effects of FTY720 on immune function and that sphingosine-1-phosphate lyase may be a potential target for immunomodulatory therapy.

FTY720 is a novel immunomodulatory agent that inhibits lymphocyte trafficking and prevents allograft rejection. FTY720 is phosphorylated in vivo, and the phosphorylated drug acts as agonist for a family of G protein-coupled receptors that recognize sphingosine 1-phosphate. Evidence suggests that FTY720-phosphate-induced activation of S1P 1 is responsible for its mechanism of action. FTY720 was rationally designed by modification of myriocin, a naturally occurring sphingoid base analog that causes immunosuppression by interrupting sphingolipid metabolism. In this study, we examined interactions between FTY720, FTY720-phosphate, and sphingosine-1-phosphate lyase, the enzyme responsible for irreversible sphingosine 1-phosphate degradation. FTY720-phosphate was stable in the presence of active sphingosine-1-phosphate lyase, demonstrating that the lyase does not contribute to FTY720 catabolism. Conversely, FTY720 inhibited sphingosine-1-phosphate lyase activity in vitro. Treatment of mice with FTY720 inhibited tissue sphingosine-1-phosphate lyase activity within 12 h, whereas lyase gene and protein expression were not significantly affected. Tissue sphingosine 1-phosphate levels remained stable or increased throughout treatment. These studies raise the possibility that disruption of sphingosine 1-phosphate metabolism may account for some effects of FTY720 on immune function and that sphingosine-1-phosphate lyase may be a potential target for immunomodulatory therapy.
FTY720 is a novel immunosuppressive agent that modulates lymphocyte trafficking and prevents allograft rejection (1,2). FTY720 treatment stimulates lymphocyte homing to peripheral lymph nodes and Peyer's patches, resulting in prolonged survival of allograft tissues and both prevention and treatment of various autoimmune diseases in animal models (3). FTY720 induces numerous effects on the immune system including inhibited egress of both naïve and activated CD4 ϩ , CD8 ϩ , and B lymphocytes from peripheral lymphoid organs and thymus, cysteinyl leukotriene-dependent T cell chemotaxis to lymph nodes, peripheral blood lymphopenia, egress of lymphocytes from the spleen, displacement of B cells from the marginal zone of the spleen, decreased ␤1 integrin expression on marginal zone B cells, homing of hematopoietic progenitor cells to the bone marrow, and decreased vascular permeability (4 -12). Importantly, while FTY720 prevents the migration of lymphocytes to allogeneic graft tissue and other sites of inflammation, it does not diminish the activation, proliferation, or effector functions of B and T lymphocytes in response to antigen stimulation (13). Toxicities associated with FTY720 are limited and distinct from those of other immunosuppressive drugs, which makes FTY720 an ideal candidate for combination immunosuppressive transplantation regimens.
FTY720 was rationally designed based on chemical modifications of myriocin, a naturally occurring sphingoid base analog that causes immunosuppression by interrupting sphingolipid metabolism (14). FTY720 is phosphorylated in vivo (15), and the phosphorylated form of the drug acts as an agonist for a family of G protein-coupled receptors that recognize the sphingolipid metab-olite sphingosine 1-phosphate (S1P) 2 as the endogenous ligand (6,16). Genetic and biochemical evidence suggest that interactions between FTY720 and the S1P 1 receptor are primarily responsible for the effect of the drug on lymphocytes (5). While FTY720-P was first recognized as an S1P 1 receptor agonist, exposure to the drug also results in receptor internalization and down-regulation on the surface of lymphocytes. This latter effect is thought to be the major mechanism of action, since reconstitution of mice with hematopoietic progenitors or lymphocytes lacking S1P 1 expression emulates the effects of FTY720 treatment on lymphocyte egress and trafficking (17,18).
We considered the possibility that FTY720 might interact not only with sphingoid base phosphate receptors but also with enzymes involved in sphingoid base metabolism. Such enzymes might affect FTY720 or FTY720-P catabolism and pharmacokinetics. Conversely, FTY720 or FTY720-P might influence key enzymes of sphingolipid metabolism and thereby contribute to the mechanism of drug action and its pleiotropic effects on the immune system. Intracellular S1P levels are regulated by its synthesis from sphingosine by SK and its catabolism by lipid phosphatases and by SPL, a pyridoxal 5Ј-phosphate-dependent enzyme that resides in the endoplasmic reticulum and is responsible for the irreversible degradation of S1P to ethanolamine phosphate and hexadecenal (19). In this study, we explored interactions between FTY720, FTY720-P, and SPL. FTY720-P was stable for prolonged periods in the presence of highly active SPL, suggesting that SPL is not responsible for metabolism of the drug. Conversely, FTY720 inhibited SPL activity in vitro, and treatment of mice with clinically relevant doses of FTY720 inhibited SPL activity in thymic extracts within 12 h of drug administration, concomitant with the onset of lymphopenia. In contrast, no appreciable effect on SPL expression was observed. These studies raise the possibility that disruption of S1P metabolism may be partly responsible for the effects of FTY720 on immune function and that SPL may be a potential target for immunomodulatory therapy.

EXPERIMENTAL PROCEDURES
Materials-FTY720 and FTY720-P were kindly provided by Volker Brinkmann of Novartis (Basel, Switzerland). The human SPL adenoviral expression system was kindly provided by Timothy Hla (University of Connecticut Health Center). C17 and C18 sphingosines, Pso, and S1Ps were obtained from Biomol Research Laboratories, Inc. FTY720-P Metabolism-Whole cell extracts of HEK293 cells expressing human SPL via an adenoviral expression system were prepared as described previously (20). These extracts contain ϳ1430 pmol/mg of protein/min of SPL activity, based upon activity against DHS1-P substrate. HEK293 cells expressing GFP via the adenoviral expression system were employed as controls. Twenty nmol of FTY720-P were suspended in SPL reaction buffer (21), and 25 l of whole cell extract plus 25 l of extraction buffer (0.5 mM potassium phosphate buffer, pH 7.4, 2 mM EDTA, 2 mM 2-mercaptoethanol, 11% glycerol, 0.2 mM pyridoxal 5Ј-phosphate) were added to initiate the reaction. At various time points, the reaction was stopped by addition of 0.3 ml of concentrated HCl, 1 ml of 1 M KCl, and 1.5 ml of chloroform to the tube. The sample was mixed by vortexing and centrifuged at 1200 rpm for 5 min. Organic phase was collected, dried down using a speed vac and resuspended in methanol. After derivatization with OPA, the sample was injected onto the HPLC for quantification. The HPLC analysis was conducted using a Beckman system Gold 125 solvent module with a Spectra-Physics SP8410 fluorescence detector and a C18 (2) Luna 3-m, 75 ϫ 4.6-mm column from Phenomenex (CA). The mobile phase was methanol:10 mM potassium phosphate buffer, pH 7.2, 1 M TBAP in water ϭ 83:16:1, v/v/v, and the flow rate was 1 ml/min. The calculation of the FTY720-P in each sample was based on the integration of the peaks of interest. For comparison, 20 nmol of S1P were incubated with extracts from HEK293 cells overexpressing SPL or control HEK293 cells, and extraction of S1P was performed as for FTY720. Recovery of S1P was determined by HPLC quantitation, as described below for S1P measurements in mouse tissues. * This work was supported by National Institutes of Health Grant CA77528. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  Confirmation of FTY720-P by LCMS-Samples corresponding to FTY720-P peak identified on HPLC were evaluated by LCMS to confirm FTY720-P by mass analysis. Twenty microliters of sample were injected into the LCMS, and mass spectrum scans from 300 to 500 m/z and specific ions m/z 388 and 410 were quantified for FTY720-P. A reverse phase C18 column from Phenomenex (2.1 mm ϫ 15 cm) was used. Two mobile phases were used: solution A is water: methanol:acetic acid (69:30:1) containing 5 mM ammonium acetate, and solution B is methanol:acetic acid (99:1) containing 5 mM ammonium acetate. The program for those two mobile phases was started with 50% solutions A and B, continued for 2 min, and then solution B was increased to 100% in 6 min, continued for 5 min in 100% solution B, and finally solution B was decreased back to 50% in 2 min. The flow rate was 0.3 ml/min. The retention time is 11.2 min.
SPL Enzyme Assay-Tissues were homogenized by sonication in 9 volumes of ice-cold homogenization buffer (5 mM MOPS, pH 7.5, 1 mM dithiothreitol, 1 mM EDTA, 0.25 M sucrose, and 5 g/ml each chymostatin, leupeptin, pepstatin A and antipain). SPL assays were conducted using 100 g of protein per assay and radioactive [4, H]DHS1-P substrate, as described previously (20). For in vitro enzyme assays, whole cell extracts were prepared from adenovirally infected HEK293 cells expressing SPL or GFP control and assayed as described above. For in vitro FTY720 and FTY720-P experiments, the drug was incubated with whole cell extracts prepared from HEK293 cells infected with an adenoviral vector expressing human SPL. The drug was incubated with extracts for 30 min, followed by addition of extracts to reaction buffer and substrate to initiate the reaction. Competition assays were performed using 100 g of protein from thymic extracts of mice treated with 1 mg/kg body weight of FTY720 and euthanized at 24 h after drug administration. Assays were conducted using increasing concentrations of cold DHS1-P substrate, from 10 -40 nmol.
Drug Treatments and Leukocyte/Lymphocyte Counts-Four-week-old FVB mice were obtained from Charles River Laboratories (Wilmington, MA). Animals were maintained in a pathogen-free facility in microisolator cages. A single dose of FTY720 at 1 mg/kg in sterile water was injected intraperitoneally at time 0. Mice were euthanized by CO 2 inhalation at various time points post-injection, followed by tissue harvest. Blood counts were determined by Coulter and manual differential. All animal studies were performed in accordance with the approved Children's Hospital Oakland Research Institute Institutional Animal Care and Use Committee protocols. S1P Measurements-S1P was isolated by two-phase lipid extraction, derivatized with OPA, and quantitated by HPLC essentially as described (22)(23)(24). All tissues were weighed prior to snap-freezing in liquid nitrogen and were stored in Ϫ100°C until processing. One nmol of PSo and PSo-P were added as internal standards before homogenization of tissues. Lipids were extracted with a 1-ml mixture of chloroform and methanol (1:2 ratio). Samples were incubated at room temperature with sonication for 3 h and then dried by SpeedVac. One ml of 0.1 N NaOH in methanol was added for 1 h at 37°C to achieve hydrolysis. After addition of 4 ml of chloroform and 4 ml of 1 M KCl, samples were mixed and phases separated by centrifugation. The aqueous phase was isolated and mixed with 0.3 ml of concentrated HCl and 6 ml of chloroform. The phases were separated by centrifugation, and the organic phase was collected and dried for S1P determination. All samples for S1P determination were re-dissolved in methanol. After derivation with OPA, samples were injected onto HPLC for quantification. The mobile phase was methanol:10 mM phosphate buffer, pH 7.2:1 M TBAP in water ϭ 83:16:1, v/v/v, and the flow rate was 1 ml/min. SPL Expression-To evaluate the effect of FTY720 and FTY720-P on SPL gene expression, HEK293 cells were transfected with a reporter construct containing 7.8 kb of sequence upstream of the SPL ATG start codon. The transfected cells were treated for 24 h with 3 M FTY720 or FTY720-P (the maximal dose at which no cytotoxicity was appreciated by 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyl-tetrazolium bromide (MTT) assay), at which time luciferase activity was determined, as described (25). Western blotting of mouse tissue extracts was performed using a peptide affinity-purified polyclonal sera raised against the murine SPL C-terminal peptide: NH 2 -V-T-Q-G-N-Q-M-N-G-S-P-K-P-R-COOH.

RESULTS AND DISCUSSION
SPL Does Not Catabolize FTY720-P-FTY720 demonstrates a long elimination half-life after single dose administration (2). Known routes of drug metabolism include conversion to FTY720-P by SK, and oxidation of the octyl side chain, followed by ␤ oxidation. We were interested in determining whether FTY720 might function as a substrate for SPL, which recognizes phosphorylated long chain bases varying in chain length, hydroxylation, and saturation. To test this possibility, FTY720-P was incubated under standard SPL assay conditions with HEK293 cells demonstrating high levels of SPL activity by virtue of an adenoviral SPL expression system. Incubation proceeded for various lengths of time, after which FTY720-P was recovered by two-phase lipid extraction, derivatized with the fluorescent compound OPA, and quantitated by HPLC. As shown in Fig. 1a, FTY720-P remained stable for up to 24 h in the presence of SPL. In contrast, S1P levels diminished under the same conditions in cells expressing high levels of SPL but not in control cells expressing GFP (Fig. 1b). S1P degradation correlated with an increase in SPL reaction products, as determined using a standard SPL assay (Fig. 1c). Our results indicate that, under these conditions, S1P is degraded in an SPL-dependent fashion, whereas FTY720-P does not serve as a substrate for SPL.
Inhibition of SPL Activity by FTY720 in Vitro-We considered it likely that the structural features that allow FTY720 to be recognized by SK, FTY720-P to be recognized by S1P receptors, and both phosphorylated and unphosphorylated forms of the drug to be effectively translocated by long chain base transporters, as well as the access of FTY720 to intracellular compartments, would support interactions with other enzymes of long chain base metabolism (6,9,15,16). To eval-FIGURE 1. Stability of FTY720-P in the presence of SPL. a, 20 nmol of FTY720-P were incubated with extracts from cells containing high levels of SPL for 0 -24 h under standard SPL reaction conditions except for the absence of DHS1-P substrate. At the indicated times, the reaction was stopped, and FTY720-P was recovered and quantitated by HPLC as described under "Experimental Procedures." b, 20 nmol of S1P were incubated with extracts from cells containing high levels of SPL for 0 -24 h (SPL, gray bars) or alternatively with extracts from cells expressing GFP control and that have no appreciable SPL activity (GFP, black bars). At the indicated times, S1P was recovered and quantitated by HPLC as described under "Experimental Procedures." c, SPL reaction products were measured in extracts from cells overexpressing SPL (gray bars) or GFP control (black bars) and incubated with substrate in the standard SPL assay for the indicated time. uate whether either FTY720 or FTY720-P affect SPL activity, whole cell extracts were prepared from HEK293 cells expressing high levels of SPL and incubated with varying concentrations of FTY720 or FTY720-P. As shown in Fig. 2a, FTY720 inhibited SPL activity in vitro in a dose-dependent fashion, with effects occurring at concentrations as low as 300 nM with maximal effects observed at 30 M. In contrast, FTY720-P was inhibitory only at 30 M and was not as potent as FTY720 at this dose (Fig. 2b). Although treatment with FTY720 is associated with low (nanomolar) circulating drug levels, local concentrations within cells and subcellular compartments may be considerably higher, since FTY720 is actively transported into cells by transporters that recognize long chain bases (9).
Inhibition of SPL by FTY720 in Vivo-To determine whether SPL enzyme activity is influenced by FTY720 administration in vivo, 4-week-old female FVB mice were given 1 mg/kg FTY720 by intraperitoneal injection, and tissue SPL activity in thymus was assessed over time. As shown in Fig. 3a, inhibition of SPL was noted as early as 12 h after FTY720 treatment and peaked at 36 h, with a maximum inhibition of 40%. SPL inhibition occurred concomitant with lymphopenia (TABLE ONE) and was consistent with the timing of lymphopenia reported previously (4). To address whether the inhibition of SPL activity in homogenates of tissues isolated from FTY720 treated animals is a reversible phenomenon, tissue SPL activity was measured in the presence of increasing amounts of cold DHS1-P substrate. As shown in Fig. 3b, enzyme inhibition (21%) observed in tissues from animals treated with drug for 24 h was partially overcome (reduction to 14% inhibition) by a 2-fold increase in cold substrate, whereas the level of inhibition in the presence of 40 nmol of cold substrate (25%) was similar to that observed with 10 nmol of substrate (standard assay conditions). We conclude that the inhibition of SPL by FTY720 is at best partially reversed by competition with excess substrate under these conditions. FTY720 Does Not Significantly Affect SPL Gene or Protein Expression-SPL gene expression is inducible by GATA transcription factors (25) and other physiological conditions, including DNA damage, hypoxia, and other stress conditions. 3 Therefore, we examined SPL expression in the tissues of mice treated with FTY720, as well as in cell lines exposed to FTY720. Neither SPL gene expression as determined by an SPL-luciferase reporter system (Fig. 4a) nor tissue SPL protein expression as determined by immunoblotting (Fig. 4b) were significantly affected by FTY720 treatment at doses that inhibited enzyme activity. Although small changes in protein expression may not be detectible by immunoblotting, these studies suggest that the major effect of the drug on SPL activity is not by altering SPL expression.
FTY720 Administration Does Not Reduce Tissue S1P Levels-SK is responsible for S1P synthesis, whereas SPL and S1P phosphatases are responsible for S1P catabolism. Since FTY720 interacts with both SK and SPL and may potentially influence the efficiency of both S1P synthesis and degradation, we were interested to determine the sum effect of FTY720 administration on tissue S1P levels. Interestingly, FTY720 treatment was associated with a slight increase in thymic tissue S1P at 36 h, followed by a return to baseline levels by 48 h (TABLE TWO). S1P levels increased somewhat more dramatically in splenic tissue after FTY720 treatment. DHS1-P levels remained stable throughout 48 h of treatment (data not shown). These findings indicate that, despite potential competition with sphingosine by FTY720 as a substrate for SK, tissue S1P   . FTY720 inhibits SPL activity in vivo. a, one mg/kg FTY720 was administered by peritoneal injection to FVB mice at time 0. Mice were euthanized at the indicated times, and SPL activity in thymic tissue was determined as described under "Experimental Procedures." This experiment was performed in quadruplicate and is representative of two experiments. †, p Ͻ 0.02; † †, p Ͻ 0.005; † † †, p Ͻ 0.002. b, SPL activity in tissue homogenates of animals treated with FTY720 for 24 h (gray bars) and untreated controls (black bars) was measured in the presence of increasing concentrations of cold DHS1-P substrate. Actual SPL activity in each extract is shown. It should be noted that total activity decreases in both experimental and control tissue homogenates as the amount of cold substrate increases, due to reduction in specific activity of the substrate. FTY720mediated inhibition in extracts from treated animals was determined as percent of control and is 21% in the presence of 10 nmol of substrate, 14% in the presence of 20 nmol of substrate, and 25% in the presence of 40 nmol of substrate. levels do not decrease and may even increase transiently during FTY720 administration, consistent with FTY720-mediated inhibition of SPL.
In summary, these studies demonstrate that FTY720 inhibits the activity of SPL, the major enzyme responsible for irreversibly removing S1P from the sphingosine-S1P cycle. This effect is observed in vitro and in vivo at therapeutic doses of the drug and occurs concomitant with drug-induced lymphopenia. These studies suggest that FYT720, in addition to interacting with SK and S1P receptors, also interacts with SPL and that inhibition of SPL by FTY720 may potentially contribute to some aspects of drug function (Fig. 5). This notion is consistent with the observation that deoxypyridoxine, a competitive inhibitor of pyridoxal 5Ј-phosphate-dependent enzymes including SPL, is a potent immunosuppressant and prolongs allograft survival (26,27). Loss of SPL expression and concomitant changes in long chain base metabolism produce significant phenotypes in several developmental model systems (28 -30). However, these simple metazoan organisms do not express recognizable S1P receptors and lack well developed immune systems. Whether loss of SPL expression in mammals has any effect on lymphocyte trafficking or immunity should be clarified with the generation and characterization of mouse SPL knock-out models. FIGURE 5. Proposed mechanism of FTY720 effects on S1P metabolism. FTY720 enters the cell through facilitated transport, probably via ABC transporters at the plasma membrane. Upon entry into the cell, FTY720 interacts with and inhibits SPL and is converted to FTY720-P by the actions of SK. FTY720-P interacts with S1P receptors to inhibit lymphocyte trafficking.  . SPL expression is not significantly affected by FTY720. a, SPL gene expression. HEK293 cells were co-transfected with a human SPL luciferase reporter construct and the sea pansy luciferase reporter as a control for transfection efficiency. Twenty-four hours after transfection, cells were treated with FTY720 or vehicle and incubated for another 24 h at 37°C. Luciferase activity was then determined as a measure of expression, as described under "Experimental Procedures." Values are luciferase activity normalized to sea pansy luciferase. This experiment was performed in duplicate and is representative of two experiments. b, SPL protein expression. Thymic extracts from control mice (time 0) and mice treated with FTY720 were evaluated at different time points after drug administration for SPL protein expression by immunoblotting as described under "Experimental Procedures." Actin was used as a loading control. By densitometric analysis of SPL protein normalized to actin protein, the maximal decrease in SPL protein levels was 16% (data not shown).