Research article

Sphingosine kinase type 2 inhibition elevates circulating sphingosine 1-phosphate

Yugesh Kharel, Mithun Raje, Ming Gao, Amanda M. Gellett, Jose L. Tomsig, Kevin R. Lynch, Webster L. Santos

Abstract

S1P (sphingosine 1-phosphate) is a pleiotropic lipid mediator involved in numerous cellular and physiological functions. Of note among these are cell survival and migration, as well as lymphocyte trafficking. S1P, which exerts its effects via five GPCRs (G-protein-coupled receptors) (S1P1–S1P5), is formed by the action of two SphKs (sphingosine kinases). Although SphK1 is the more intensively studied isotype, SphK2 is unique in it nuclear localization and has been reported to oppose some of the actions ascribed to SphK1. Although several scaffolds of SphK1 inhibitors have been described, there is a scarcity of selective SphK2 inhibitors that are necessary to evaluate the downstream effects of inhibition of this isotype. In the present paper we report a cationic amphiphilic small molecule that is a selective SphK2 inhibitor. In the course of characterizing this compound in wild-type and SphK-null mice, we discovered that administration of the inhibitor to wild-type mice resulted in a rapid increase in blood S1P, which is in contrast with our SphK1 inhibitor that drives circulating S1P levels down. Using a cohort of F2 hybrid mice, we confirmed, compared with wild-type mice, that circulating S1P levels were higher in SphK2-null mice and lower in SphK1-null mice. Thus both SphK1 and SphK2 inhibitors recapitulate the blood S1P levels observed in the corresponding null mice. Moreover, circulating S1P levels mirror SphK2 inhibitor levels, providing a convenient biomarker of target engagement.

  • sphingosine kinase (SphK)
  • sphingosine kinase inhibitor
  • sphingosine 1-phosphate (S1P)

INTRODUCTION

S1P [Sph (sphingosine) 1-phosphate] is a bioactive lipid involved in a host of cellular functions, including migration, differentiation, survival, angiogenesis and immune cell modulation. Extracellularly, S1P exerts its effects via five GPCRs (G-protein-coupled receptors), S1P1–S1P5, and, intracellularly, acts to modulate transcription complexes [1]. Owing, in part, to the remarkable clinical success of the S1P receptor agonist and immunomodulatory prodrug fingolimod (FTY720), S1P signalling is currently the subject of many investigations.

Two SphK (Sph kinase) isotypes (SphK1 [2] and SphK2 [3]) catalyse the phosphorylation of Sph and these enzymes are solely responsible for S1P synthesis [4]. Studies with Sphk1-null mice reveal that SphK1 is responsible for a substantial fraction of S1P production [5]. Not surprisingly, these kinases have come under increasing scrutiny as drug targets owing to their role in the production of S1P, which is implicated in a variety of pathological conditions, such as cancers, fibrosis, etc.

The relative importance of SphK1 compared with SphK2 as potential drug targets remains a topic of debate. Although SphK1 is reported to promote growth and survival [68], cell-based studies of SphK2 suggest that this enzyme is not protective, but rather it opposes proliferation while enhancing apoptosis [911]. Although several scaffolds of SphK1 inhibitors have been described [1215], SphK2 inhibitors are less common. One compound, ABC294640, has a Ki value of 8 μM against SphK2 [16]. This adamantyl compound has been reported to be efficacious in several disease models, including hepatic ischaemia/reperfusion injury [17], osteoarthritis [18], Crohn's disease [19], ulcerative colitis [20] and colon cancer [21], among others. However, ABC294640 was recently reported to bind to the oestrogen receptor where it has tamoxifen-like properties [22].

In the present paper, we report the results of our characterization of another SphK2 inhibitor, SLR080811 {(S)-2-[3(4-octylphenyl)-1,2,4-oxadiazol-5-yl]pyrrolidine-1-carboximidamide}. SLR080811 is a guanidine-based SphK2 inhibitor with a t1/2 of 4–5 h in mice. Although this molecule lowers S1P levels in cultured cells, it drives an SphK1-dependent increase in S1P in mice and thus mimics the elevated S1P levels observed in SphK2-null mice.

MATERIALS AND METHODS

Materials

Sphk1−/− [5] and Sphk2−/− [4] mice were gifts from Dr Richard Proia [NIH (National Institutes of Health)/NIDDK (National Institute of Diabetes and Digestive and Kidney Disease), Bethesda, MD, U.S.A.]. C57BL/6J mice were from Jackson Laboratories. The plasmid encoding DGK (diacylglycerol kinase) α was a gift from Dr Kaoru Goto (Yamagata University School of Medicine, Yamagata, Japan). Adult mouse kidney fibroblast cultures were a gift from Dr Amandeep Bajwa (Center for Immunity, Inflammation and Regenerative Medicine, University of Virginia, Charlottesville, VA, U.S.A.). Deuterated (D7) S1P, S1P, Sph, C17-S1P and C17-Sph were purchased from Avanti Polar Lipids.

Synthesis of SLR080811

The details of the synthesis of SLR080811 and related compounds will be reported elsewhere (M. Raje, M. Gao and W.L. Santos, unpublished work).

Kinase assays

SphK activity was measured by a scintillation proximity assay as described previously [23]. Briefly, recombinant SphK1 or SphK2 were expressed in Sf9 insect cells, crude homogenates were prepared and incubated in 96-well FlashPlates (PerkinElmer) in a buffer containing D-erythro-Sph and [γ-33P]ATP. The [33P]S1P product, which adheres to the plate wall, was quantified by scintillation counting. To assay ceramide kinase or DGKs, the recombinant proteins were incubated with [γ-32P]ATP and substrate (C6 ceramide or 1-O-hexadecyl-2-acetyl-sn-glycerol respectively) and the lipid product, after recovery by organic extraction, was resolved by TLC, detected by autoradiography and quantified by liquid scintillation counting. These assays were performed with and without a fixed concentration of inhibitor and its effect on Km and Vmax was determined.

Sample preparation

Sample preparation protocols were as described in Shaner et al. [24] with minor modifications. Cell pellets (2–4×106 cells), whole blood (20 μl) or plasma (50 μl) was mixed with 2 ml of a methanol/chloroform solution (3:1) and transferred into a capped glass vial. Suspensions were supplemented with 10 μl of internal standard solution containing 10 pmoles each of C17-S1P or deuterated (D7) S1P, C17-Sph or deuterated (D7) Sph, and an undecyl analogue of compound 1a [25,26]. The mixture was placed in a bath sonicator for 10 min and incubated at 48°C for 16 h. The mixture was then cooled to ambient temperature (22°C) and mixed with 200 μl of 1M potassium hydroxide in methanol. The samples were sonicated again and incubated a further 2 h at 37°C. Samples were then neutralized by the addition of 20 μl of glacial acetic acid and transferred into 2 ml microcentrifuge tubes. Samples were then centrifuged at 12000 g for 12 min at 4°C. The supernatant fluid was collected in a separate glass vial and evaporated under a stream of nitrogen gas. Immediately prior to LC (liquid chromatography)-MS analysis, the dried material was dissolved in 0.3 ml of methanol and centrifuged at 12000 g for 12 min at 4°C. Then, 50 μl of the resulting supernatant fluid was analysed.

LC-MS protocol

Analyses were performed by LC-MS using a triple quadrupole mass spectrometer (AB-Sciex 4000 Q-Trap) coupled to a Shimadzu LC-20AD LC system. A binary solvent gradient with a flow rate of 1 ml/min was used to separate sphingolipids and drugs by reverse-phase chromatography using a Supelco Discovery C18 column (50 mm×2.1 mm, 5 μm bead size). Mobile phase A consisted of water/methanol/formic acid (79:20:1, by vol.), whereas mobile phase B was methanol/formic acid [99:1 (v/v)]. The run started with 100% A for 0.5 min. Solvent B was then increased linearly to 100% B in 5.1 min and held at 100% for 4.3 min. The column was finally re-equilibrated to 100% A for 1 min. Natural sphingolipids were detected using MRM (multiple reaction monitoring) protocols described previously [24] as follows: C17-S1P (366.4→250.4), S1P (380.4→264.4), dihydroS1P (382.4→266.4); deuterated (D7) C18-S1P (387.4→271.3), C17-Sph (286.4→250.3), Sph (300.5→264.4), dhSph (sphinganine) (302.5→260.0) and deuterated (D7) Sph (307.5→271.3). Fragmentation of compound SLR080811 was analysed by direct infusion of a 1 μM solution in methanol/formic acid [99:1 (v/v)] and it was found that the transition (371.1→140.1) in positive mode provided the most intense signal at the following voltages: DP (declustering potential)=76, EP (entrance potential)=10, CE (collision energy)=29 and CXP (collision cell exit potential)=10. All of the analytes were analysed simultaneously using the aforementioned MRMs. The retention times for all of the analytes under our experimental conditions were between 5.1 and 5.6 min. Ceramide (16:0) was measured in positive mode by monitoring the m/z 264.4 product ion and using a Supelco Supelcosil LC-NH2 column (50 mm×2.1 mm, 3 μm bead size) as described previously [24]. Quantification was carried out by measuring peak areas using commercially available software (Analyst 1.5.1).

Cell culture

U937 cells were grown in RPMI 1640 medium supplemented with L-glutamate, 10% FBS (fetal bovine serum) and 1% penicillin/streptomycin at 37°C in an atmosphere containing 5% CO2 [12]. SKOV3 cells were grown in McCoy's 5a medium supplemented with 10% FBS and 1% penicillin/streptomycin at 37°C in an atmosphere containing 5% CO2. Mouse kidney fibroblasts were grown in low-glucose DMEM (Dulbecco's modified Eagle's medium) with 10% FBS. At 24 h before adding inhibitors, the growth medium was replaced with medium containing 0.5% FBS.

Cell viability assay

U937 cells were plated in 96-well plates at a density of 60000–80000 cells per well. Cells were treated with the indicated concentration of compounds for 24 h. Cell viability was assessed using the TACS® MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide] assay according to the manufacturer's protocol (R&D System). Briefly, 10 μl of MTT reagent was added to 100 μl of cell culture medium and the plate was incubated at 37°C for 4 h, followed by incubation with 100 μl of detergent reagent at room temperature (22°C) for 2 h. Absorbance was measured at a wavelength of 570 nm.

Generation of F2 hybrid mice

All of the animal protocols were approved prior to experimentation by the University of Virginia's School of Medicine Animal Care and Use Committee.

Sphk1−/− and Sphk2−/− mice were mated and 15 (5 male and 10 female) of the resulting F1 heterozygotes were intercrossed to yield a cohort of F2 hybrid mice. Mice were genotyped at weaning [P21 (postnatal day 21)] and the various genotypes were found at the expected Mendelian frequencies, except for the embryonic lethal double-null genotype [4]. After 12–14 h of fasting, blood was obtained from F2 mice from retro-orbital sinuses under light isofluorane anaesthesia. Aliquots of whole blood and plasma were processed for LC-MS analysis of S1P levels as described above.

Pharmacokinetic analysis

Groups of 8–12-week-old mice (strain C57BL/6J) were injected (intraperitoneally) with either SLR080811 (at a dose of 10 mg/kg) or an equal volume of vehicle [2% solution of hydroxypropyl-β-cyclodextrin (Cargill Cavitron 82004)]. After injection, animals were bled at the specified time points [ASAP (as soon as possible) time points were 1–2 min after dosing]. Whole blood was processed immediately for LC-MS analysis as described above.

RESULTS

Inhibitor design strategy

In the course of our previous studies of amidine-based SphK1 inhibitors [14,25] to identify compounds with longer half-lives, we reasoned that compounds containing an amidine or an amide might be rapidly hydrolysed in vivo. Therefore we synthesized compound SLR080811 (Figure 1), which retained the pyrrolidine ring of our SphK1 inhibitor compound 1a [26] while replacing the amidine with the more stable guanidine isostere. Both compounds feature the carboximidamide ‘warhead’ [25]. Furthermore, the amide in compound 1a was replaced by an oxadiazole group. SLR080811 was evaluated against recombinant SphKs, other lipid kinases, in whole cells and, finally, in wild-type and SphK-null mice.

Figure 1 Chemical structure and inhibitory constants of compounds SLR080811 and 1a [26]

The chemical structure of compounds SLR080811 and 1a along with their inhibitory constants (Ki) for recombinant SphK1 and SphK2 are shown. Inhibitory constants were obtained by kinetic analysis of S1P production using variable concentrations of Sph and a fixed concentration of ATP in the presence and absence of compounds. These compounds exhibit a pattern of competitive inhibition, therefore Ki values were calculated as Ki=[I]/(Km′/Km−1), where [I] is the concentration of inhibitor, and Km′ and Km are the Michaelis constants obtained in the presence and absence of inhibitor. Measurements were carried out using [33P]ATP as a tracer and a microplate-based scintillation proximity assay for the detection of [33P]S1P, as described previously [23].

Evaluation of SLR080811 in vitro

SLR080811, prepared as the hydrochloride salt, was tested first against recombinant SphK1 and SphK2 using a broken cell assay (see the Materials and methods section). In these assays, SLR080811 was found to have inhibitory constants (Ki) of 1.3 and 12 μM for SphK2 and SphK1 respectively (Figure 1). Furthermore, SLR080811 was found to be competitive with Sph, but not with ATP. Because SLR080811 is a Sph analogue, we tested SLR080811 as an inhibitor of related lipid kinases, including ceramide kinase and DGKα. At a concentration of 3 μM, no inhibition of either enzyme was observed (results not shown).

We characterized SLR080811 in detail because of its SphK2-selectivity that, although modest (10-fold), is unusual in our carboximidamide series. We chose human leukaemia U937 cells for the evaluation of SphK2 inhibitors because they exhibit high SphK1 and SphK2 activities, can be cultured with ease, and, more importantly, these cells have been used in the past by us and other authors [12] to test SphK1 inhibitors, enabling comparisons of the effects of inhibitors.

We first treated U937 cultures with either vehicle or SLR080811 and measured the intracellular levels of S1P, dhS1P (dhSph 1-phosphate), Sph, dhSph and SLR080811. We observed that treatment of U937 cells with SLR080811, but not vehicle, resulted in decreased amounts of phosphorylated sphingolipids S1P and dhS1P (Figures 2a and 2c) and the concomitant increase of the corresponding non-phosphorylated precursors Sph and dhSph (Figures 2b and 2d). The data in Figures 2(a) and 2(c) indicate that the IC50 values of SLR080811 are less than the Ki (1.3 μM) determined against recombinant SphK2. In addition to sphingolipids, we also measured the intracellular concentration of SLR080811. As shown in Figure 2(e), SLR080811 accumulated inside U937 cells in a concentration-dependent manner, which might explain its potent effect on the levels of intracellular sphingolipids shown in Figures 2(a)–2(d).

Figure 2 Levels of sphingolipids and compound SLR080811 in U937 cells treated with various concentrations of compound SLR080811 as indicated

After a 2 h period of exposure, cells were harvested by centrifugation, lysed and the amounts of sphingolipids and SLR080811 in the lysates were measured by LC-MS as described in the Material and methods section. Amounts associated with cells are expressed as the number of pmoles per 106 cells. (a) S1P, (b) Sph, (c) dihydroS1P (dhS1P), (d) dhSph and (e) SLR080811. Values are means±S.D. for three independent experiments. *P<0.05, **P<0.01 and ***P<0.001 (one-way ANOVA, and Bonferroni's Multiple Comparison Test, compared with vehicle alone).

Treatment of human Jurkat T leukaemia cells or human SKOV3 ovarian cancer cells with SLR080811 for 2 h also resulted in decreased S1P (results not shown). In U937 cells, the effect of SLR080811 on intracellular sphingolipids was observed as early as 20 min after SLR080811 exposure (results not shown) and persisted for at least 72 h as shown in Figures 3(a) and 3(b). We also quantified one of the prominent ceramide species (C16:0) in these cells and found that this ceramide was significantly elevated, but only at the 48 and 72 h time points (Figure 3c).

Figure 3 Levels of sphingolipids and compound SLR080811 in cultured U937 cells treated with vehicle or SLR080811 (1 μM) for different times as indicated

Cells were harvested by centrifugation, lysed and the amounts of sphingolipids and SLR080811 in the lysates were measured by LC-MS as described in the Materials and methods section. Amounts are expressed as the number of pmoles per 106 cells. (a) S1P, (b) Sph and (c) C16:0 ceramide. Values are means±S.D. for three independent experiments. *P<0.05, **P<0.01 and ***P<0.001 [Student's t test, compared with the control (no compound treatment)]. SLR, SLR080811.

The most obvious explanation for the decline in U937 cell-associated S1P and dhS1P in response to SLR080811 is decreased synthesis, but it is conceivable that the decline was somehow the result of increased metabolism via, for example, S1P phosphatase or S1P lyase, or increased S1P export. To discriminate between these possibilities, we used FTY720, an SphK2-selective substrate [27,28]. We observed that treatment of U937 cells with SLR080811 impaired their ability to convert FTY720 into FTY720-phosphate. As shown in Figure 4, we observed much lower levels of intracellular FTY720-phosphate in SLR080811-treated cells that, as expected, correlated with correspondingly higher levels of FTY720. This suggests that the reduction of intracellular S1P levels in U937 cells produced by SLR080811 is due to SphK2 inhibition rather than the alternative mechanisms mentioned above.

Figure 4 Levels of phospho-FTY720 and FTY720 in U397 cells treated with FTY720 and SLR080811

Cultured U937 cells were exposed to 1 μM FTY720 and two concentrations of SLR080811 as indicated in the Figure. After 2 h of exposure, cells were harvested by centrifugation, lysed and the amounts of FTY720 and phospho-FTY720 (FTY720-P) were measured by LC-MS as described in the Materials and methods section. (a) Accumulation of phospho-FTY720. (b) Accumulation of FTY720. Amounts are expressed as the number of pmoles per 106 cells. Drug and FTY720 concentrations on the x axis refer to the concentration of these molecules in the culture medium. Values are means±S.D. for three independent experiments. **P<0.01 and ***P<0.001 [one-way ANOVA and Bonferroni's multiple comparison test, compared with the control (no compound treatment)].

To evaluate further the selectivity of SLR080811 in vitro, we used SphK1-null and SphK2-null mouse kidney fibroblasts. Because the wild-type fibroblasts derived from adult mouse kidney have both SphK1 and SphK2 activity (results not shown), null cells are a useful model for testing compound selectivity. We found that SLR080811 reduces the levels of intracellular S1P in both wild-type and SphK1-null cells, but not in SphK2-null cells (Figure 5a). This suggests that SphK2, but not SphK1, is a target for SLR080811. In contrast with our expectations, however, the effect of SLR080811 on Sph levels was not selective for SphK2. As shown in Figure 5(b), both SphK1-null and SphK2-null fibroblasts exhibited increased concentrations of Sph on treatment with SLR080811 suggesting that other mechanisms may be at play in the regulation of these sphingolipids.

Figure 5 Selectivity of SLR080811 inhibition

Wild-type, Sphk1−/− and Sphk2−/− adult mouse kidney fibroblasts were exposed to 1 μM SLR080811 for 2 h. Cells were harvested by centrifugation, lysed and sphingolipids were measured by LC-MS as described in the Materials and methods section. (a) S1P and (b) Sph. Values are means±S.D. for three independent experiments. *P<0.05 and ***P<0.001 [Student's t test, compared with the control (no inhibitor)]. SLR, SLR080811; WT, wild-type.

Finally, we tested whether the effects of SLR080811 on U937 cells included cell toxicity. In general, we found that SLR080811 has no obvious cytotoxic effects on U397 cells. For example, cultures grew normally in medium containing up to 3 μM SLR080811 and there were no signs of cell growth inhibition (results not shown). Furthermore, we investigated the effect of SLR080811 on U397 cells using a standard assay that correlates cell viability with their redox potential (MTT assay, see the Materials and methods section). We found that SLR080811 had a slight cytotoxic effect that was apparent even at the lowest concentration tested, but was not concentration-dependent (Figure 6).

Figure 6 Viability of U937 cells treated with SLR080811

U937 cells were exposed to various concentrations of SLR080811 for 24 h as indicated. The viability of the cells was measured by an MTT assay as described in the Materials and methods section. Viability is directly proportional to the amount of formazan dye produced by live cells, as measured by absorbance at 570 nm. Values are means±S.D. for three independent experiments. *P<0.05 and **P<0.01 [one-way ANOVA, and Bonferroni's multiple comparison post-test, compared with the control (no inhibitor)].

Evaluation of SLR080811 in vivo: isotype selectivity and pharmacokinetics

As an extension of our experiments in vitro, we sought to evaluate the SphK isotype selectivity of SLR080811 in vivo. To this end, we injected groups of SphK1-null, SphK2-null and wild-type mice with a single intraperitoneal dose of SLR080811 and analysed the blood levels of S1P. We observed that SphK1-null mice exhibit reduced levels of blood S1P after injection, whereas in SphK2-null animals the blood levels of S1P did not change (Figure 7a). This result once again suggests that SLR080811 is a selective inhibitor of SphK2. Levels of S1P in SphK1-null mice reached the lowest point between 2 and 4 h after injection and slowly returned to pre-treatment levels at approximately 24 h. In addition to S1P, we also measured the blood levels of SLR080811. The kinetics of SLR080811 approximated that of S1P in the sense that we observed an early SLR080811 peak at 1 h, rather than 2 h (Figure 7b) as in the case of S1P, and then a slow disappearance of the compound over the subsequent 24 h.

Figure 7 S1P and SLR080811 levels in the blood of mice injected with SLR080811

Wild-type, or Sphk1- or Sphk2-null mice were administered SLR080811 (at a dose of 10 mg/kg, intraperitoneally). Blood samples were drawn at times 0, 1, 2, 4, 8 and 24 h post-injection. Levels of S1P (a) and SLR080811 (b) from blood samples of WT mice were measured by LC-MS. Values are means±S.D. for three to five mice per group. **P<0.01 [repeated measures two-way ANOVA, and Bonferroni's multiple comparison test compared with ASAP time point after injection of the compound (time 0)]. WT, wild-type.

The effect of a single dose of SLR080811 in wild-type mice was surprising, in that S1P levels increased rather than decreased (Figure 7a). To investigate this seemingly paradoxical result, we generated a cohort of age-matched F2 hybrid Sphk1 Sphk2 mice (see the Materials and methods section) and measured their blood and plasma S1P levels. These mice constituted a genetically homogeneous population whose levels of blood S1P should be less influenced by genetic variation than the parent strains. In this population we observed, as shown in Figure 8, the same pattern of blood S1P levels: high circulating S1P in SphK2-null animals, lower S1P levels in wild-type animals and low levels of S1P in SphK1-null animals. Thus it appears that SLR080811, by inhibiting SphK2, mimics the effect of lack of functional Sphk2 alleles. However, we note that a different SphK2 inhibitor, ABC294640, was reported to lower circulating S1P, albeit after 5 weeks of daily dosing into xenograft-bearing SCID (severe combined immunodeficiency) mice [29].

Figure 8 S1P levels in 8-week-old wild-type, Sphk1−/− and Sphk2−/− littermates

Blood was drawn from wild-type, Sphk1-null and SphK2-null mice and the S1P levels were measured by LC-MS as described in the Materials and methods section. Data shown are independent measurements of whole blood (a) or plasma (b) from four wild-type, nine Sphk1−/− and six Sphk2−/− mice. The Sphk-null mice were either heterozygous or wild-type at the other SphK locus. *P<0.05 and ***P<0.001 (one-way ANOVA, and Bonferroni's multiple comparison test, compared with wild-type). WT, wild-type.

Finally, in the course of the investigations described in the present paper we noticed that the recovery of the C17-S1P internal standard was greater than 100% when samples from SphK2−/− mice were analysed. This prompted us to investigate the possible presence of endogenous C17-S1P or an interfering analyte in those samples. We found that Sphk2−/− samples contained an analyte that is indistinguishable from authentic C17-S1P by LC/MS, i.e. their chromatographic retention times and transitions are indistinguishable. Because of these striking similarities we refer to this analyte as C17-S1P with the caveat that a definitive identification will require confirmation by additional analysis. Moreover we found that wild-type and Sphk1−/− mice samples also contained C17-S1P, albeit to a lesser extent. From the practical point of view, and regardless of the identity of this analyte, it was obvious to us that C17-S1P should not be used as an internal standard for our samples, rather we used deuterated (D7) S1P. A representative experiment illustrating this technical issue can be seen in Figure 9.

Figure 9 Levels of S1P and C17-S1P in whole blood of wild-type, SphK1-null and SphK2-null mice

Blood samples (20 μl) from a wild-type (a), an Sphk1−/− (b) and a Sphk2−/− (c) mouse (all F2 hybrids) were analysed by LC-MS for the presence of S1P and C17-S1P. Recovery was evaluated using deuterated (D7) S1P as an internal standard. The Figure shows the chromatograms produced by the Analyst software for these three analytes, which were obtained using the following transitions: S1P (380.4→264.4), C17-S1P (366.4→250.4) and D7-S1P (387.4→271.4). The intensity of the signals corresponds to the number of counts per second (cps) recorded by the ion detector of the mass spectrometer. Because the Analyst software uses the height of the tallest peak as a full scale, the y axis scale (cps) is different for each chromatogram. The numbers above the peaks correspond to the retention times in minutes. Retention time for authentic C17-S1P was 5.23±0.01 min, n=3 (results not shown). Amounts of analytes were calculated using the areas under the peaks and a standard curve. Areas in cps×min for the S1P, C17-S1P and D7-S1P peaks shown in the Figure were as follows: wild-type, 7.93×106, 1.37×105 and 2.36×106; Sphk1−/−, 3.67×106, 3.43×104 and 2.20×106; and Sphk2−/−, 4.23×107, 2.95×106 and 2.07×106. In these mice, which are representative of our F2 dihybrid cohort, C17-S1P is less than 2% of S1P in wild-type mice and less than 1% in SphK1-null mice, but is 7% in SphK2-null mice. In addition to revealing that the molar fraction of C17-S1P is higher in SphK2-null mouse blood, these chromatograms also reveal that the absolute concentration of S1P is 4–5-fold higher in SphK2-null mice than in wild-type mice.

DISCUSSION

In the present paper we describe SLR080811, a novel SphK2-selective inhibitor. SLR080811 was synthesized by modifying the chemical structure of compound 1a, an SphK1-selective inhibitor that we have previously reported [26]. The amidine and the amide group in compound 1a were replaced by guanidine and oxadiazole groups respectively in SLR080811. Although our original intention was to obtain a longer-lived version of compound 1a by switching to the more stable guanidine and oxadiazole groups, we unexpectedly found that these changes also brought about a reversal of selectivity for SphK isotypes. In the context of the present paper, ‘selectivity’ refers not to absolute selectivity for SphK2, i.e. no effect on SphK1, but rather to relative selectivity. As shown in Figure 1, SLR080811 is approximately one order of magnitude a more potent inhibitor of SphK2 as compared with SphK1 on the basis of the Ki values obtained in vitro with recombinant enzymes. In addition to the selectivity for SphK2, we found that SLR080811 does indeed have a longer half-life than compound 1a according to our original expectations. As shown in Figure 7 we estimate the half-life of SLR080811 to be 4–5 h, whereas the half-life for compound 1a is less than 1 h [26]. We are currently testing SLR080811 analogues to identify compounds with enhanced potency, half-life and/or isotype selectivity.

Our contention that SLR080811 is an SphK2-selective inhibitor is on the basis of observations carried out in different biological models both in vivo and in vitro. In agreement with our results with recombinant enzymes, the effects of SLR080811 in cultured cell lines, such as U937 cells, correspond to inhibition of S1P synthesis, namely reduction of S1P and dhS1P levels and concomitant elevation of their aminoalcohol precursors Sph and dhSph (Figure 2). That these effects are mediated, at least in part, by SphK2 blockade is demonstrated by our experiments with FTY720. This drug is a selective SphK2 substrate and, as shown in Figure 4, its phosphorylation is greatly diminished in the presence of SLR080811.

Consistent with our observations with U937 cells, SLR080811 had no effect on S1P levels in SphK2-null fibroblasts, whereas it reduced S1P levels in wild-type and SphK1-null fibroblasts, as shown in Figure 5(a). Taken together these results corroborate the SphK inhibition by SLR080811 and its SphK2 selectivity. Our observations on Sph levels in these cells, however, were not completely consistent with this picture. Although we were able to detect the anticipated accumulation of Sph in both wild-type and SphK1-null fibroblasts, we unexpectedly observed that Sph levels were higher in SphK2-null fibroblasts as well (Figure 5b). There is no simple explanation for this observation, other than to suggest that some other mechanism(s) not presently known may be involved in the regulation of the metabolism of sphingolipids. One possibility is that SLR080811 might have some inhibitory effect on ceramide synthase.

Our in vivo observations support the notion that SLR080811 is a SphK2-selective inhibitor in the sense that it lowers the levels of blood S1P in SphK1-null mice, but does not have an effect on blood S1P when administered to mice lacking SphK2 (Figure 7). It is of note that the basal levels of blood S1P in SphK2-null animals are much higher than those in SphK1-null or wild-type mice. This was reported previously by Zemann et al. [30] with their SphK2- null mice and then by Olivera et al. [31] and Sensken et al. [32]. Therefore it may be hypothesized that SphK1 is up-regulated when SphK2 activity is reduced. However, increased SphK1 activity by itself seems to be insufficient to raise S1P levels, as was demonstrated in transgenic mice with forced global expression of SphK1 [33]. These animals have high tissue SphK1 activity, but have levels of plasma S1P that are not different from wild-type animals. It appears that SphK2 suppression may have more complex consequences than just increasing SphK1 activity. A previous report showing that SphK2 may function as a transcriptional modulator [1] supports this contention, although this does not seem to be the case for the SphK2-null mice whose SphK1 mRNA levels were not different from wild-type mice [30]. According to all of these observations, it appears that there are factors at play in the regulation of S1P metabolism that we do not fully understand.

Our previous work with compound 1a, our SphK1-selective inhibitor, included the observation that this compound reduces the levels of blood S1P in wild-type mice [26]. This effect appears to be the logical consequence of SphK inhibition in the context of a simple model of S1P metabolism whereby reducing the synthesis of S1P results in a drop in S1P levels. We expected SLR080811 to exhibit a qualitatively similar effect because both compounds are SphK inhibitors and because they exhibit other similarities. For example, both compounds accumulate to high levels in U937 cells, possibly because they are recognized by the uptake system(s) that take up long-chain bases such as Sph and dhSph from the extracellular environment [34,35]. In addition, neither compound exerts major effects on cell viability at concentrations that result in extensive SphK blockade. However, we observed, in contrast with our expectations, that SLR080811 increases the levels of blood S1P in wild-type mice, i.e. the opposite effect of compound 1a. Obviously, this observation cannot be accommodated in a simple model of S1P metabolism. It seems that the most likely hypothesis that can be advanced to explain this effect is to postulate that SphK2 inactivation, whether genetic or pharmacological, leads to increased S1P levels. In this context, the effect of SLR080811 on wild-type animals can be related to the increased levels of S1P in SphK2-null mice. In any event, if the effects of compounds 1a and SLR080811 prove to be a general property of SphK inhibitors and persist with chronic dosing, then it should be possible to use SphK inhibitors to adjust the levels of S1P both positively and negatively. This may be useful not only in the investigation of S1P metabolism, but also in a clinical setting.

AUTHOR CONTRIBUTION

Mithun Raje, Ming Gao and Webster Santos designed, synthesized and purified SLR080811. Yugesh Kharel and Kevin Lynch designed the experiments to characterize SLR080811. Yugesh Kharel performed the experiments, except that Amanda Gellett did the DGK assays and statistical analyses and Jose Tomsig did the LC-MS sphingolipid analyses. Yugesh Kharel and Kevin Lynch wrote the paper and all of the authors participated in editing the paper prior to submission.

FUNDING

This work was supported by the National Institute of General Medical Sciences of the National Institutes of Health [grant number R01GM067958]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Acknowledgments

We thank Dr Richard Proia (National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Disease, Bethesda, MD, U.S.A.) for the gift of Sphk1−/− and Sphk2−/− mice, Dr Kaoru Goto (Yamagata University School of Medicine, Yamagata, Japan) and Dr Matthew Topham (University of Utah, Salt Lake City, UT, U.S.A.) for their gifts of DGK plasmids, and Dr Amandeep Bajwa (University of Virginia, Charlottesville, VA, U.S.A.) for her gift of mouse kidney fibroblasts. We acknowledge the technical help of Ms Devon McCurdy.

Abbreviations: ASAP, as soon as possible; DGK, diacylglycerol kinase; dhS1P, sphinganine 1-phosphate; dhSph, sphinganine; FBS, fetal bovine serum; LC, liquid chromatography; MRM, multiple reaction monitoring; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide; S1P, sphingosine 1-phosphate; Sph, sphingosine; SphK, Sph kinase

References

View Abstract