Verification of the Necessity of the Tolyl Group of PF-543 for Sphingosine Kinase 1 Inhibitory Activity

PF-543, the most potent sphingosine kinase (SK) inhibitor, does not demonstrate effective anticancer activity in some cancer cells, unlike other known SK1 inhibitors. PF-543 has a non-lipid structure with a unique toluene backbone; however, the importance of this structure remains unclear. Therefore, the purpose of this study was to investigate changes in SK inhibitory and anticancer activities and to explore the role of the tolyl group structure of PF-543 through various modifications. We transformed the methyl group of PF-543 into hydrogen, fluorine, and hydroxy. PF-543 derivatives in which the methyl group was substituted by hydrogen and fluorine (compound 5) demonstrated SK1 inhibitory and anticancer activities similar to PF-543. Moreover, we performed molecular modeling studies of PF-543 and compound 5. To assess the metabolic stability of PF-543 and compound 5, we determined their degree of degradation using the liver microsomes of four different animal species (human, dog, rat, and mouse). However, both PF-543 and compound 5 showed poor microsomal stability. Therefore, for the medical applications of PF-543, the structural modifications of its other parts may be necessary. Our results provide important information for the design of additional PF-543 analogs.


Introduction
Sphingolipids are structurally modified by various enzymes and regulate various biological activities through the regulation of cancer cell signaling. Sphingosine and ceramide mediate apoptosis, whereas sphingosine-1-phosphate (S1P), sphingomyelin, and glucosylceramide have antiapoptotic and prosurvival roles. The majority of enzymes involved in sphingolipid metabolism have been identified through comprehensive research, and the regulation of these enzymes has the potential to be developed RB-005 is a sphingosine-based SK1-selective inhibitor (IC50 = 3.6 μM) and induced the proteasomal degradation of SK1 in human pulmonary arterial smooth muscle cells [9]. ROME, an SK2 inhibitor (Ki = 16 μM), can reduce MCF-7 breast cancer cell proliferation and prevent cell migration [9].
In 2010, FTY720 (Fingolimod, Gilenya, and Novartis) became the first oral treatment for multiple sclerosis to be approved by the US FDA ( Figure 1) [10]. FTY720 also activates protein phosphatase 2A to inhibit cell growth in various forms of cancers (hematologic malignancies and liver, bladder, and colorectal cancers). Additionally, FTY720 inhibits SK1 activity [11]. Diverse in vivo effects of these sphingolipids and FTY720 analogs have been a major obstacle in their development as drugs. Thus, the development of a sphingolipid regulator with a nonlipid structure and selective function is required.

Chemical Synthesis
PF-543 analogs containing fluorine and hydroxyl groups in their modified benzene backbones were synthesized using orcinol and 3-fluoro-5-methylphenol as starting materials, respectively (Scheme 1).

SK Activity Assay of PF-543 (1) and Derivatives (2-10)
Once synthesized, compounds 2-10 were treated with 20 µM using a sphingosine kinase assay kit to measure SK1/2 inhibition ( Figure 2). Compounds 8-10, which contained an introduced hydroxyl group, showed a weaker SK1 inhibitory effect. Among compounds 2-7 (which contained the introduced hydrogen and fluorine), the SK1 inhibitory effects reduced in compounds 4 and 7, which also contained an introduced 4-hydroxypiperidine. As a result, it was confirmed that the backbone structure of the polar phenol group replacing methylbenzene in the PF-543 structure reduced SK1 inhibition. Compounds 2 and 5, which had the same headgroups as PF-543, and compounds 3 and 6, which had the same headgroups as compound 22c (Figure 1), inhibited 100% of SK1 similar to PF-543 (at the same concentration). Compounds 4 and 7 had a 4-hydroxypiperidyl headgroup and displayed significantly reduced SK inhibition. This indicates that the pyrrolidine structure is an effective headgroup in PF-543 analogs compared with the piperidine structure. The SK inhibitory effects of compounds 2, 3, 5, and 6 indicated that the presence of the tolyl group in the PF-543 structure did not alter the inhibitory effects of SK.
which also contained an introduced 4-hydroxypiperidine. As a result, it was confirmed that the backbone structure of the polar phenol group replacing methylbenzene in the PF-543 structure reduced SK1 inhibition. Compounds 2 and 5, which had the same headgroups as PF-543, and compounds 3 and 6, which had the same headgroups as compound 22c (Figure 1), inhibited 100% of SK1 similar to PF-543 (at the same concentration). Compounds 4 and 7 had a 4-hydroxypiperidyl headgroup and displayed significantly reduced SK inhibition. This indicates that the pyrrolidine structure is an effective headgroup in PF-543 analogs compared with the piperidine structure. The SK inhibitory effects of compounds 2, 3, 5, and 6 indicated that the presence of the tolyl group in the PF-543 structure did not alter the inhibitory effects of SK.

Cytotoxic Effects of PF-543 (1) and Derivatives (2-10) in Cancer Cells
The cytotoxicity of compounds 2-10 and PF-543 at concentrations of 20 and 40 μM, respectively, was tested against HT29 cancer cells (Figure 3a). The cytotoxic effects of compounds 8-10 were weaker than those of PF-543 in colorectal cancer cells, similar to their decreased SK1 inhibitory effects. This indicates that polar groups in the SK1 inhibitor backbones may have reduced SK1 inhibitory and anticancer activities. In contrast, compounds 2-7 (including compounds 4 and 7) demonstrated anticancer activity similar to PF-543 against colorectal cancer cells. Additional studies were conducted on compound 5, which had the same headgroup structure as PF-543, to confirm the necessity of the tolyl group. Because compound 2, which also had the same headgroup as PF-543, is a known compound [15], anticancer activities were confirmed using compound 5 introduced with fluorine, which is a bioisostere of hydrogen. An evaluation of compound 5 revealed both SK1 inhibitory activity and cytotoxicity. The cytotoxic effect of compound 5 was similar to that of PF-543.

Cytotoxic Effects of PF-543 (1) and Derivatives (2-10) in Cancer Cells
The cytotoxicity of compounds 2-10 and PF-543 at concentrations of 20 and 40 µM, respectively, was tested against HT29 cancer cells (Figure 3a). The cytotoxic effects of compounds 8-10 were weaker than those of PF-543 in colorectal cancer cells, similar to their decreased SK1 inhibitory effects. This indicates that polar groups in the SK1 inhibitor backbones may have reduced SK1 inhibitory and anticancer activities. In contrast, compounds 2-7 (including compounds 4 and 7) demonstrated anticancer activity similar to PF-543 against colorectal cancer cells. Additional studies were conducted on compound 5, which had the same headgroup structure as PF-543, to confirm the necessity of the tolyl group. Because compound 2, which also had the same headgroup as PF-543, is a known compound [15], anticancer activities were confirmed using compound 5 introduced with fluorine, which is a bioisostere of hydrogen. An evaluation of compound 5 revealed both SK1 inhibitory activity and cytotoxicity. The cytotoxic effect of compound 5 was similar to that of PF-543. The apoptotic effects of compound 5 and PF-543 were further compared at concentrations of 20 and 40 µM, respectively. It was revealed that compound 5 had similar apoptotic effects. Therefore, the tolyl group in the PF-543 structure did not affect the anticancer activity in the same way as the SK1 inhibitory effects. The apoptotic effects of compound 5 and PF-543 were further compared at concentrations of 20 and 40 μM, respectively. It was revealed that compound 5 had similar apoptotic effects. Therefore, the tolyl group in the PF-543 structure did not affect the anticancer activity in the same way as the SK1 inhibitory effects.

Docking Study of PF-543 and Compound 5
Molecular modeling studies of PF-543 and compound 5 were conducted. The hydroxymethylpyrrolidine (OH and N of pyrrolidine) of compound 5 showed both hydrogen bonding and electrostatic interaction with Asp264 (protonated amine form). The phenyl linker of compound 5 showed hydrophobic interactions involving Ile260, Val263, Leu354 and Met358, while the benzene backbone of compound 5 showed hydrophobic interactions with the surrounding Phe389 and Met392. The terminal phenyl group of compound 5, similar to that of PF-543, formed hydrophobic interactions with aromatic rings and the surrounding Ala360 and Phe374. It was thus determined that compound 5 showed a binding mode similar to PF-543, indicating that its structure can replace the methyl group of PF-543 ( Figure 4).

Docking Study of PF-543 and Compound 5
Molecular modeling studies of PF-543 and compound 5 were conducted. The hydroxymethyl-pyrrolidine (OH and N of pyrrolidine) of compound 5 showed both hydrogen bonding and electrostatic interaction with Asp264 (protonated amine form). The phenyl linker of compound 5 showed hydrophobic interactions involving Ile260, Val263, Leu354 and Met358, while the benzene backbone of compound 5 showed hydrophobic interactions with the surrounding Phe389 and Met392. The terminal phenyl group of compound 5, similar to that of PF-543, formed hydrophobic interactions with aromatic rings and the surrounding Ala360 and Phe374. It was thus determined that compound 5 showed a binding mode similar to PF-543, indicating that its structure can replace the methyl group of PF-543 ( Figure 4). The apoptotic effects of compound 5 and PF-543 were further compared at concentrations of 20 and 40 μM, respectively. It was revealed that compound 5 had similar apoptotic effects. Therefore, the tolyl group in the PF-543 structure did not affect the anticancer activity in the same way as the SK1 inhibitory effects.

Docking Study of PF-543 and Compound 5
Molecular modeling studies of PF-543 and compound 5 were conducted. The hydroxymethylpyrrolidine (OH and N of pyrrolidine) of compound 5 showed both hydrogen bonding and electrostatic interaction with Asp264 (protonated amine form). The phenyl linker of compound 5 showed hydrophobic interactions involving Ile260, Val263, Leu354 and Met358, while the benzene backbone of compound 5 showed hydrophobic interactions with the surrounding Phe389 and Met392. The terminal phenyl group of compound 5, similar to that of PF-543, formed hydrophobic interactions with aromatic rings and the surrounding Ala360 and Phe374. It was thus determined that compound 5 showed a binding mode similar to PF-543, indicating that its structure can replace the methyl group of PF-543 ( Figure 4).

Metabolic Stability of PF-543 and Compound 5
To assess the metabolic stability of PF-543 and compound 5, we determined their degree of degradation using the liver microsomes of four different animal species (human, dog, rat, and mouse). PF-543 and compound 5 both demonstrated a low microsomal stability of <10% in all animal species (Table 1). These results show that the tolyl group of PF-543 did not affect the stability of PF-543.

5-(Bromomethyl)-1,3-phenylene diacetate (15)
Orcinol (2 g, 0.016 mol) was dissolved in pyridine (80 mL), acetic anhydride (4.57 mL, 0.048 mol) was added thereto, and the mixture was stirred at room temperature for 12 h. Water was added to stop the reaction, and it was concentrated under reduced pressure after EtOAc extraction and MgSO 4 drying. The resulting mixture 12 (1.7 g, 0.008 mol) was dissolved in EtOAc (50 mL) without purification, and N-bromosuccinimide (1.89 g, 0.01 mol) was added thereto and stirred at 60 • C for 2 days. The reaction was terminated with water and EtOAc, and the organic layer was dried over MgSO 4 and concentrated under reduced pressure. The resulting mixture was separated by column chromatography (n-hexane:EtOAc = 5:1) to give compound 15 ( (17) Compound 14 (1.1 g, 0.0045 mol) was placed in a sealed tube, dissolved in THF/DMF (2/1, 30 mL), and benzene sulfinic acid sodium salt (2.2 g, 0.013 mol) was added thereto. The reaction was stirred for 3 days while heating to 80 • C. After the reaction was cooled to room temperature, the reaction was terminated with water and EtOAc, dried over MgSO 4 and concentrated under reduced pressure.

3-((Phenylsulfonyl)methyl)phenol (19)
m-Toly acetate (2 g, 0.013 mol) was dissolved in EtOAc (150 mL), and N-bromosuccinimide (2.37 g, 0.0133 mol) was added thereto, followed by stirring for 3 days while heating to 70 • C. After cooling to room temperature, water and EtOAc were added thereto and the reaction was completed. The organic layer was separated, dried over MgSO 4 , and concentrated to obtain a mixture of compound 13. Compound 13 was placed in a sealed tube and dissolved in 120 mL of THF/DMF (2/1). Benzenesulfinic acid sodium salt (4.27 g, 0.026 mol) was added thereto, followed by stirring for 3 days while heating to 80 o C. The reaction was cooled to room temperature, water and EtOAc were added and the reaction was completed. The organic layer was separated, dried over MgSO 4 , and concentrated to obtain mixture compound 18. The reaction mixture was dissolved in 150 mL of MeOH/H 2 O (1/1), NaHCO 3 (5 g) was added thereto, and the mixture was stirred for 12 h while heating at 40 • C. The reaction was cooled to room temperature, water and EtOAc were added thereto and the reaction was completed. The organic layer was separated, dried over MgSO 4 , and concentrated. The mixture was separated by column chromatogrphy (n-hexane:EtOAc = 2:1) to obtain compound 19. The chemical spectrum of compound 19 was consistent with that previously reported [15].

Cell Culture and Proliferation Assays
HT29 cells (Korean Cell Line Bank, Seoul, Korea) were maintained in RPMI (Biowest, Nuaillé, France) media with 10% fetal bovine serum (FBS), 100 U/mL penicillin and 100 µg/mL streptomycin at 37 • C in a humidified 5% CO 2 /95% air atmosphere. Cells were seeded in 96-well plates at a density of 3000 cells/100 µL/well and incubated for 24 h. The cells were then incubated in culture medium containing synthetic compounds. Following 24 h of incubation, the cell viability was determined using a EZ-CYTOX kit (DaeilLab Service, Seoul, Korea) according to the manufacturer's protocol (n = 12).

Annexin-V Staining of PF-543 and Derivative 5
The cell apoptosis assay was determined using annexin V-fluorescein isothiocyanate (FITC) (Biobud, Gyeonggi-do, Korea) and propedium iodide (PI) staining as described in the manufacturer's instructions (BioBud, Gyeonggi-do, Korea). Briefly, the cells were resuspended in binding buffer and then incubated with annexin-V-FITC and propidium iodide at room temperature. After incubation, stained cells were analyzed by MACSQuant Analyzer 10 Fow Cytometer (Miltenyi Biotec, Bergisch Gladbach, Germany). The experiment was conducted three times independently (n = 6).

Molecular Modeling of PF-543 Derivative 5 against SK1
Molecular modeling study of 5 against the SphK1 was performed using Schrödinger Suite 2019-1 (Schrödinger, LLC, New York, NY, USA, 2019). The PF-543-bound crystal structure of SphK1 was obtained from the Protein Data Bank (http://www.rcsb.org/pdb), the PDB code for which was 4V24 [16]. The protein preparation was revised using Protein Preparation Wizard in Maestro v11.9 (Schrödinger, NYC, NY, USA), and the receptor grid box was generated using a 25 × 25 × 25 Å cubic size centered on complexed ligand. The ligand was minimized using an OPLS3e force field with a dielectric constant  Microsomal incubations were conducted in triplicate in 0.1 M potassium phosphate buffer (pH 7.4) in a clean 1.5 mL eppendorf tube. To evaluate reduced nicotinamide adenine dinucleotide phosphate (NADPH)-dependent metabolism, PF-543, derivative 5 or Verapamil (positive control) was incubated with pooled liver microsomes from human (HLM), dog (DLM), rat (RLM) and mouse (MLM) in the presence of an NADPH regenerating system in a final volume of 100 mL. The final incubation mixtures contained 0.5 mg/mL HLM, 0.1 M potassium phosphate buffer (pH 7.4), and an NADPH regenerating system (1 mM NADPH, 10 mM MgCl 2 ). HLM was added and the mixture was pre-incubated at 37 • C in a shaking incubator at approximately 350 rpm for 5 min under Thermo mixer (Eppendorf., Hamburg, Germany). Reactions were initiated by the addition of 1 mM NADPH and were then quenched by adding 40 µL of ice-cold acetonitrile containing 10 µM chlorproamide (CPP) as an internal standard at 0, 30 min. The incubation mixtures were then centrifuged at 15,000× g for 5 min at 4 • C. A 2 µL aliquot of the supernatant was directly injected into the LC-MS/MS system. The LC-MS/MS system consisted of the Nexera XR HPLC system (Shimadzu Co., Kyoto, Japan) coupled to the TSQ Vantage triple quadrupole mass spectrometer equipped with Xcalibur version 1.1.1 (Thermo Fisher Scientific Inc., Waltham, MA, USA).

Statistical Analysis
The data were presented as means ± SD. Non-parametric one-way ANOVA was applied to the data with heterogeneous variance. Turkey's post-procedures were conducted to determine intergroup variation. The difference was considered significant if p < 0.05. Graph Pad Prizm 7 (GraphPad Sofrware, Inc., La Jolla, CA, USA) was used to analyze all the data.

Conclusions
In this study, we synthesized the derivatives of the tolyl group of PF-543 with benzene, fluorobenzene, and phenol to improve the anticancer activity of PF-543. In addition, prolinol, hydroxypyrrolidine, and piperidine were introduced into the headgroups of each backbone derivative, and the SK1 inhibitory and anticancer activities of these synthesized compounds were measured. The inhibitory effects of SK in compounds 4, 7, and 10 reduced when a piperidine headgroup was introduced compared with those of PF-543 (1). All derivatives with a phenol backbone (compounds 8-10) displayed reduced SK1 inhibitory effects. Only compound 8 showed an inhibitory effect against SK2 alone. These results indicate that the tolyl group of PF-543 affects its SK inhibitory effect. In contrast, compounds with a benzene and fluorobenzene backbone (2-7) demonstrated anticancer activity similar to PF-543. In addition, the fluorine-substituted compound 5, with the same headgroup as PF-543, showed similar apoptotic effects. The molecular docking of PF-543 and compound 5 did not reveal any difference in binding between the two compounds. However, their microsomal stability was low in all tested animal species. Collectively, our results indicate that, although the fluorobenzene backbone can replace the PF-543 backbone, it does not improve its stability.
Despite the strong SK1 inhibitory effects of PF-543, various derivatives have not been developed due to its low anticancer activity and structural instability. The PF-543 analogs reported by Pfizer and other research groups provided only comparative information on the inhibitory effects of SK1/2; therefore, it remains unknown if the derivatives improved anticancer activity. The purpose of this study was to determine if the presence of the tolyl group in the PF-543 structure is required for its anticancer activity. Our results show that the tolyl group did not significantly affect the SK1 inhibitory effects and anticancer activities of PF-543. Furthermore, synthesized PF-543 derivative did not create additional structural stability. These results showed that it is possible to design a simplified PF-543 analog in which the tolyl group has been removed. The structural stability of PF-543 could be improved by designing derivatives in which a functional group other than the tolyl group is modified. Further structural modification of other parts of PF-543 may be warranted to make it suitable for medical use.