FTY720 (2-amino-[2-(4-n-octylphenyl)ethyl]-1,3-propanediol, Fingolimod™, 1, Chart 1) is a synthetic analogue of the chiral sphingolipid myriocin (2).1 As an analogue of sphingosine, FTY720 is phosphorylated in vivo by sphingosine kinases, affording (S)-FTY720-phosphate (3), which activates four of the five known sphingosine 1-phosphate (S1P, 2a) G protein-coupled receptors.2 Chart 1 Structures of FTY720 (1); myriocin (2); S1P (2a); and FTY720 phosphate, phosphonate, and (E)-vinylphosphonate analogues 3–5. Internalization and subsequent polyubiquitination of the S1P receptors leads to their proteasomal degradation and renders the cells unresponsive to S1P; therefore, lymphocytes are not capable of recirculation to peripheral inflammatory tissues.3 Thus, FTY720 has therapeutic potential, and in fact is the first S1P receptor modulator that has entered the stage of a phase-III clinical study.4 Several syntheses of 15 and of phosphate 3 have been accomplished.6 In contrast to phosphates such as 3, phosphonate analogues are resistant to the action of lipid phosphate phosphatases and may offer improved cellular stability. A racemic mixture of the nonhydrolyzable phosphonate analogue of FTY720 (4) was reported in which the C-O-P bond is replaced with a C-C-P bond;2b rac-4 was found to be a high-affinity agonist of the S1P-type 1 receptor (S1P1), with a similar potency as (S)-3.7 We report here the first asymmetric syntheses of the chiral phosphonate analogues of FTY720, (R)-4 and (S)-4. Oxazoline intermediate (S)-14 (Scheme 1), prepared by a modification of our previous route,6c was further elaborated to give the corresponding (E)-vinylphosphonate analogue (S)-5. We have included a preliminary pharmacological characterization of the effects of these analogues on the non-transformed rat intestinal epithelial cell line IEC-6. This study revealed that (S)-5, but not its (R) enantiomer, exerts a potent anti-apoptotic effect in a camptothecin (CPT)-induced apoptosis model.8 Unlike phosphate (S)-3, (S)-5 did not activate the S1P1 receptor of the Endothelial Differentiation Gene (EDG) family of G protein-coupled receptors, making it a novel enantioselective probe activating a cytoprotective mechanism against apoptosis induced by DNA damage. Scheme 1 Synthesis of (S)-14 from 4-bromobenzaldehyde Wittig reaction of 4-bromobenzaldehyde with the ylide of n-heptyltriphenylphosphonium bromide gave arylalkene 6 as an E,Z (1:3) mixture (Scheme 1). Sonogashira coupling between 6 and 4-(phenylmethoxy)-1-butyne delivered enyne 7 as a 1:3 E:Z mixture in 92% yield. Alcohol 8 was obtained on reduction of the unsaturated bonds and hydrogenolysis of the O-benzyl group in the presence of Pearlman’s catalyst. After Swern oxidation of 8 provided aldehyde 9, use of a Mannich reagent, Eschenmoser's salt,9 afforded α-methylene aldehyde 10. Reduction of 10 with NaBH4 in the presence of CeCl3 (to suppress conjugate reduction) gave allyl alcohol 11.10 CeCl3, which is a mild Lewis acid, is not required, since CsCl also provided 11 as the only product. Asymmetric Sharpless epoxidation11 of 11 with cumene hydroperoxide (CHP) in the presence of L-(+)-DIPT, Ti(OPr-i)4, and molecular sieves gave epoxide (S)-12.12 The synthesis of (S)-12 was accomplished in 7 steps from p-bromobenzaldehyde and in 46% overall yield. Reaction of alcohol (S)-12 with trichloroacetonitrile in the presence of DBU gave 2,3-epoxy-1-trichloroacetimidate (R)-13. The tetrasubstituted carbon in oxazoline 14 was set up bearing the desired nitrogen substituent by opening of epoxide (R)-13 with catalytic Et2AlCl,13 affording (S)-14 in 74% yield for the two steps. Swern oxidation of oxazoline (S)-14 gave oxazoline aldehyde 15 (Scheme 2), which on Horner-Wadsworth-Emmons reaction with tetramethyl methylenediphosphonate afforded ester (S)-16 in 87% yield and with an E/Z ratio of ~10:1. Simultaneous demethylation and release of the hydroxy and amino groups by treatment with trimethylsilyl bromide (TMSBr) provided (S)-5, but the yield was low. Therefore, the hydroxy and amino groups were first released by treatment with 1 M HCl. After the amine hydrochloride was neutralized (saturated aq Na2CO3), amino alcohol 17 was converted to (S)-5 with TMSBr followed by 95% methanol; 84% yield for the two steps. Reduction of (S)-5 using Pearlman’s catalyst gave (S)-4. Scheme 2 Synthesis of (S)-4 from (S)-14 Asymmetric epoxidation of 11 with D-(−)-DIPT gave epoxide (R)-12, which was converted via (R)-14 to (R)-5 in six steps (Scheme 2). Catalytic hydrogenation of (R)-5 afforded (R)-4 (Scheme 3). Scheme 3 Outline of the synthesis of (R)-5 and (R)-4 S1P promotes the survival of many cell types.14 Both 2a and (S)-3 protect oligodendrocyte progenitor cells from apoptotic cell death in response to growth factor withdrawal, and (S)-3 was also shown to be cytoprotective in response to pro-apoptotic cytokines and microglial activation.15 The ability of 2a, (S)-3, and phosphonate analogues 4 and 5 to protect IEC-6 cells from apoptotic cell death in response to the topoisomerase inhibitor CPT was assessed by DNA fragmentation. Pretreatment with 2a, (S)-3, (R)-4, or (R)-5 did not result in a significant reduction in DNA fragmentation in response to a 4-h treatment with 20 µM CPT. However, we found that the cytoprotective effect was enantioselective, since pretreatment with 1 µM of (S)-4 or (S)-5 showed a significant reduction (21 and 50%, respectively) in DNA fragmentation in response to CPT. In a preliminary study of the activity of the FTY720-phosphonate analogues on S1P receptors, we performed Ca2+ mobilization assays with HTC4 cells that were stably transfected with S1P1. As shown in Figure 1, the S1P1 transfectants were activated by (S)-3 to 76% of the maximal S1P-induced activation, and displayed a similar potency as S1P (13 ± 2 nM for S1P vs 9 ± 1 nM for (S)-3). (R)-5 and (R)-4 both showed a modest activity against S1P1 with Emax values that ranged from 73 to 93% of the maximal S1P-induced responses, and EC50 values that were increased by ~2- to 3-fold. (S)-4 activated S1P1 to 36% of the maximal S1P-induced response, and the EC50 value was increased by ~5-fold to 75 ± 21 nM. Since (S)-5 did not elicit a Ca2+ response from cells transfected with the S1P1 receptor, we conclude that the potent cytoprotective effect of (S)-5 is not mediated by S1P1. Figure 1 Ca2+ mobilization dose-response relationships for S1P (2a), (S)-3, and FTY720 analogues 4 and 5 in HTC4 cells expressing the S1P1 receptor. In conclusion, we have described the synthesis of the enantiomers of FTY720 phosphonate analogues 4 and 5. (S)-4 and (S)-5, but not 2a or (S)-3, all at 1 µM, protected IEC-6 cells from apoptosis. The extent of CPT-induced DNA fragmentation was reduced by 50% and 21% in the presence of 1 µM of (S)-5 and (S)-4, respectively. The potent cytoprotective activity of (S)-5 is not mediated by S1P1. Experiments are underway to characterize the cellular effects of these analogues.