Synthesis of (S)-FTY720 vinylphosphonate analogues and evaluation of their potential as sphingosine kinase 1 inhibitors and activators

Article abstract of DOI:10.1016/j.bmc.2013.02.042

Sphingosine kinase 1 (SK1) is over-expressed in many cancers where it provides a selective growth and survival advantage to these cells. SK1 is thus a target for anti-cancer agents that can promote apoptosis of cancer cells. In previous work, we synthesized a novel allosteric SK1 inhibitor, (S)-FTY720 vinylphosphonate. We now report a more expeditious route to this inhibitor which features B-alkyl Suzuki coupling as a key step and show that replacement of the amino group in (S)-FTY720 vinylphosphonate with an azido group converts the vinylphosphonate from an allosteric inhibitor to an activator of SK1 at low micromolar concentrations. Our results demonstrate the feasibility of using the (S)-FTY720 vinylphosphonate scaffold to define structure-activity relationships in the allosteric site of SK1.

Full text of DOI:10.1016/j.bmc.2013.02.042

Bioorganic & Medicinal Chemistry 21 (2013) 2503–2510
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Synthesis of (S)-FTY720 vinylphosphonate analogues and evaluation
of their potential as sphingosine kinase 1 inhibitors and activators
Zheng Liu a, Neil MacRitchie b, Susan Pyne b, Nigel J. Pyne b, Robert Bittman a,
⇑
a Department of Chemistry and Biochemistry, Queens College, The City University of New York, Flushing, NY 11367-1597, USA
b Cell Biology Group, Strathclyde Institute of Pharmacy and Biomedical Science, University of Strathclyde, Glasgow G4 0RE, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
Sphingosine kinase 1 (SK1) is over-expressed in many cancers where it provides a selective growth and
survival advantage to these cells. SK1 is thus a target for anti-cancer agents that can promote apoptosis of
cancer cells. In previous work, we synthesized a novel allosteric SK1 inhibitor, (S)-FTY720 vinylphosph-
onate. We now report a more expeditious route to this inhibitor which features B-alkyl Suzuki coupling
as a key step and show that replacement of the amino group in (S)-FTY720 vinylphosphonate with an
azido group converts the vinylphosphonate from an allosteric inhibitor to an activator of SK1 at low
micromolar concentrations. Our results demonstrate the feasibility of using the (S)-FTY720 vinylphosph-
onate scaffold to define structure–activity relationships in the allosteric site of SK1.
Received 8 January 2013
Revised 19 February 2013
Accepted 27 February 2013
Available online 7 March 2013
Keywords:
Sphingosine 1-phosphate
B-alkyl Suzuki cross-coupling
FTY720
Ó 2013 Elsevier Ltd. All rights reserved.
Epoxide opening
Allosterism
1. Introduction
activity. We focused on modifying the polar head group while
retaining the lipid tail of 2 as follows: (a) changing the hydroxyl
Sphingosine kinase 1 (SK1) is over-expressed in various human
cancers and therefore inhibitors of SK1 are potential anticancer
agents.1 FTY720 (1, also known as fingolimod and by the trade-
mark name Gilenya, Fig. 1) is a structural analogue of sphingosine
that affects the activities of many sphingolipid-metabolizing en-
zymes in its non-phosphorylated form.2 On phosphorylation by
sphingosine kinase 2 (SK2), FTY720 is converted into the immuno-
modulator (S)-FTY720-phosphate, which is the first compound in a
new class of small molecule sphingosine 1-phosphate (S1P) recep-
tor agonists that alters lymphocyte trafficking.3 In 2010, 1 was ap-
proved as an orally deliverable drug for the treatment of multiple
sclerosis.4 The finding that non-phosphorylated 1 inhibited SK1
and SK2 led to interest in synthetic analogues of 1 as putative anti-
tumor agents.5 Indeed, we reported that (S)-FTY720 vinylphospho-
group to a fluorine atom, (b) converting the amino group to an
azide group, (c) replacing the vinyl-phosphonate moiety with a vi-
nyl-carboxylic acid, and (d) limiting the conformational freedom in
the polar head group by installing a lactone ring. The synthesis of
chiral analogues 3–7 involves a B-alkyl Suzuki cross-coupling reac-
tion as the key step. The route reported here is more efficient and
practical than the previous synthetic route to 2.5
2. Chemistry
The challenge in the synthesis of 2 and analogues 3–7 lies in the
construction of the sole quaternary stereocenter. Given the high
yield and stereoselectivity in our previous synthesis,5 we decided
to continue to employ Sharpless asymmetric epoxidation (SAE)9
of allylic alcohol 12 to introduce the stereocenter. Our new syn-
thetic route to allylic alcohol 12 started with Wittig methylenation
of the commercially available aldehyde 8 (Scheme 1). The resulting
terminal alkene 9 was subjected to a B-alkyl Suzuki cross-coupling
reaction10 with (2-iodoallyloxy)(tert-butyl)dimethylsilane (10),11
followed by removal of the TBS group in 11. Thus 12 was prepared
in 66% yield (over four steps), with only one purification step (by
column chromatography).
nate (2),6
a metabolically stabilized derivative of 1, is an
uncompetitive inhibitor (with sphingosine) and a mixed inhibitor
(with ATP) of SK1.7,8 This mode of inhibition is indicative of allo-
sterism contingent on formation of the sphingosine-SK1 complex,
which offers new drug discovery opportunities and also permits
further assessment of the role of SK1 in cancer.1
To gain further insight into the structure–activity relationships
in 2 that result in binding at allosteric site(s) in SK1 we have pre-
pared analogues 3–7 and evaluated their ability to modulate SK1
The scale used in this four-step procedure was up to 20 mmol. It
is noteworthy that after the hydroboration of 9 with 9-BBN, the
reaction mixture must be mixed with aqueous basic solution for
at least 30 min prior to the addition of the palladium catalyst.
⇑
Corresponding author. Tel.: +1 718 997 3279; fax: +1 718 997 3349.
E-mail address: robert.bittman@qc.cuny.edu (R. Bittman).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.02.042

2504
Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
OH
OH
OH
OH
CH2[P(O)(OMe)2]2
NaH, THF, 0 °C, 1 h
OMe
O
Ar
Ar
P
OMe
P
O
OH
O
NH2
NH2
O
O
15 (94%)
C8H17
C8H17
only E
Ar
(S)-FTY720 vinylphosphonate (2)
FTY720 (1)
(R'O)2P(O)CH2CO2Et
O
14
CO2Et
X
Y
X
i-PrOH/H2O, rt, overnight
OH
P
O
3
4
5
F
F
NH2
N3
OH N3
16 (94%)
OH
Ar =
C8H17
Y
R' = Et E/Z 3:1
R' = i-Pr E/Z 25:1
C8H17
O
O
OH
Scheme 2. Synthesis of 15 and 16 via HWE reaction.
CO2H
in the HWE reaction, as expected,18 provided excellent E selectivity
(16: E/Z 25:1).
Since opening of a vinyl epoxide with an azide generally gener-
ates an allylic azide when no additional stabilizing group is pres-
NH2
NH2
C8H17
C8H17
6
7
Figure 1. FTY720 (1) and analogues 2–7.
ent,
a [3,3]-sigmatropic equilibration process takes place to
create a mixture of regioisomers. Therefore, this reaction is of little
value,19 and only a few systematic studies on ring-opening reac-
tions of vinyl epoxides by azide ion have been published.19–23
The yield of 11 was markedly decreased when an aqueous NaOH
solution and Pd(dppf)Cl2ꢀCH2Cl2 were added simultaneously. In
addition, ethanolamine12 did not completely precipitate the boron
complex; we found that the rest of the complex was removed with
H2O2/NaOH solution.13
For SAE of 12 on a scale of 13 mmol, a substoichiometric amount
of catalyst [0.55 equiv of Ti(O-i-Pr)4] was used, leading to epoxy
alcohol 13 in 74% yield and 11% of a side product14 formed via epox-
ide opening by isopropoxide. It was reported by Sharpless and
co-workers15 that the catalytic procedure [10 mol % of Ti(O-i-Pr)4]
can prevent the epoxide-opening reaction for unsymmetrical-
disubstituted allylic alcohols. The er value of the resulting epoxide
was >20:1, as determined by analysis of the 1H NMR spectra of
the (R)- and (S)-MTPA derivatives of 13. Dess–Martin oxidation of
epoxy alcohol 13 proceeded in 90% yield.
We achieved regioselective epoxide ring openings of
,b-unsaturated esters 15 and 16 with Ti(O-i-Pr)2(N3)2 and NaN3/
NH4Cl, which permitted structural diversification of 2. With ,d-
position
c,d-epoxy-
a
c
epoxy-a,b-unsaturated esters, nucleophilic attack at the c
has been shown to predominate.20–24 As shown in Scheme 3, we
employed two different azide sources. Surprisingly, as shown in
Scheme 3, we found that azidolysis of 15 and 16 with NaN3/NH4Cl
OH
Ti(O-i-Pr)2(N3)2
PhMe, 85 °C
15 min
OMe
P OMe
O
(R)
(E)
Ar
N3
Horner–Wadsworth–Emmons (HWE) olefination of aldehyde
14 with tetramethyl methylenediphosphonate afforded exclusively
E-alkene 15 in 94% yield (Scheme 2). In contrast, poor E selectivity
(16a: E/Z 3:1)16 was observed when 14 was treated with triethyl
phosphonoacetate in H2O/iPrOH (1:1).17 However, the use of the
sterically larger diisopropyl (ethoxycarbonylmethyl)phosphonate
δ
γ
OMe
17 (68%)
O
18/17 1:17
Ar
P
OMe
O
N3
15
NaN3, NH4Cl,
EtOH
OMe
(S)
(E)
Ar
P OMe
O
microwave, 100 °C
20 min
OH
18 (72%)
n-BuLi, Ph3PCH3 Br
18/17 11:1
O
THF, -78 °C to rt
C8H17
C8H17
N3
8
9 (94%)
NaN3, NH4Cl
EtOH
Ar
CO2Et
microwave, 100 °C
1 h
OH
Ti(O-i-Pr)4
-(-)-DIPT
1. 9-BBN, THF
0 °C to rt, 3 h
D
19 (65%)
δ
γ
19/20 6:1
Sharpless
2. Pd(dppf)Cl2·CH2Cl2
RO
O
C8H17
epoxidation
(3.7 mol %)
Ar
CO2Et
3 M NaOH, 60 °C, 3 h
OH
11 R = TBS
6 M HCl
MeOH
1 h, rt
16
OTBS
I
12 R = H (70% from 9)
Ar
Ar
CO2Et
Ti(O-i-Pr)2(N3)2
PhMe, 85 °C
15 min
N3
10
20 (26%)
O
Dess-Martin
periodinane
O
O
O
O
Ar =
pyridine
CH2Cl2
2 h, rt
OH
C8H17
13 (er > 20:1, 74%)
C8H17
N3
21 (45%)
C8H17
14 (90%)
Scheme 1. Synthesis of 14.
Scheme 3. Opening of epoxides 15 and 16.

Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
2505
acids can be achieved via a pyridine-assisted trans–cis isomeriza-
tion,27–30 it appears that azide anion might play a similar role as
pyridine in our reaction.
OH
OMe
Ar
Ar
P
OMe
NH2
Reduction of an azide to an amine in the presence of a double
bond is not trivial. Both Staudinger reduction (Ph3P, THF/H2O)
and 1,3-propanedithiol/Et3N31 failed to produce satisfactory re-
sults. Reduction using Lindlar’s catalyst (H2, Pd/CaCO3, EtOH)32 re-
sulted in saturation of the double bond. Fortunately, as illustrated
in Scheme 4, we found that simultaneous reduction of the azide
and demethylation of methyl ester 17 was accomplished by using
SnCl2 in 95% MeOH,33 providing 2 in 69% yield, together with 22
(17% yield). Methyl ester 22 was transformed to 2 by treatment
with TMSBr in quantitative yield. Our new synthetic route to 2 con-
sists of nine steps from commercially available aldehyde 8 in 19%
overall yield. The azide analogue 5 was formed by demethylation
of 17 with TMSBr, followed by aqueous MeOH, in a quantitative
yield. The stereochemistry of 22 was confirmed by its specific rota-
OH
O
OMe
SnCl2
22 (17%)
TMSBr, CH2Cl2
Ar
Ar
P
OMe
MeHO
then
MeOH
rt, overnight
N3
O
OH
17
OH
P
OH
TMSBr, 6 h, rt
then MeOH, 1 h
NH2
O
2 (69%)
OH
OH
P
OH
N3
O
5 (quantitative)
tion: ½a 2D5
ꢁ
þ 20:0 (c 0.18, CHCl3) [lit.5
½
a 2D5
ꢁ
þ 18:8 (c 1.52, CHCl3)].
Scheme 4. Synthesis of 2 and 5.
Fluorination of 17 with DAST34 (ꢂ78 °C, overnight, and then at
rt for 3 h) produced 23 in 75% yield (Scheme 5). Termination of
the reaction at low temperature led to incomplete conversion. In
contrast to 17, reduction of 23 using Lindlar’s catalyst (H2, Pd/
CaCO3, EtOH)32 did not reduce the double bond, providing 24 in
51% yield. Demethylation of methyl esters 23 and 24 with TMSBr
followed by 95% MeOH afforded the target fluorine-containing
analogues 4 and 3, respectively, in quantitative yields. The unsatu-
rated carboxylic acid analogue 6 was prepared by reduction of 20
(SnCl2 in MeOH), followed by hydrolysis of ester 25 with LiOH in
THF/MeOH/H2O. Catalytic hydrogenation of 21 (H2, Pd/C) provided
lactone analogue 7 in 46% yield (Scheme 5).
in EtOH25 afforded the highly d-selective regioisomers 17 (68%
yield) and 19 (72% yield), respectively. The ratio of regioisomers
was determined by analysis of the 31P NMR spectra. On the other
hand, the desired and expected regioisomer 17 (c-attack) was ob-
tained by treatment of 15 with Ti(O-i-Pr)2(N3)2 (which was pre-
pared from Ti(O-i-Pr)4 and TMSN3).26
To our surprise, the reaction of 16 and Ti(O-i-Pr)2(N3)2
(3.9 equiv) in toluene afforded 20 (26% yield), accompanied by lac-
tone 21 (45% yield). The regioisomer 19 was not detected. The Z
configuration of alkene 21 was confirmed by the 1H NMR spec-
trum, which shows correlated two doublets (d 6.20 ppm,
J = 9.8 Hz, @CHCO2Et; d 6.75 ppm, J = 9.8 Hz, RCHCH@). Since d-lac-
3. Biological evaluation
tone formation from d-hydroxy-trans-a,b-unsaturated carboxylic
We have previously shown that (S)-vinylphosphonate 2 (50
inhibits SK1 activity by 62.7 0.9% (n = 3) using 3 M sphingosine
(Sph) as the substrate. (S)-FTY720 vinylphosphonate (2) is an
uncompetitive inhibitor (with sphingosine) of SK1 with
Kiu = 14.5 4.4
M.7 (S)-Vinylphosphonate 2 was not a substrate
lM)
l
F
a
OMe
DAST, CH2Cl2
H2/ Pd-CaCO3
EtOH, rt, 3 h
l
17
Ar
P
OMe
for SK1 (data not shown). We now show that analogues 3–7 have
markedly different effects on SK1 activity (Fig. 2). We found that
N3
-78 °C, overnight
rt, 3 h
O
23 [65%, 75% (brsm)]
F
F
TMSBr, 6 h, rt
then MeOH, 1 h
Ar
OMe
OH
P
Ar
P
OMe
OH
NH2
NH2
O
O
24 (51%)
3 (quantitative)
F
OH
P
TMSBr, 6 h, rt
then MeOH, 1 h
Ar
OH
23
N3
O
4 (quantitative)
OH
OH
LiOH
SnCl2
THF/MeOH/H2O
CO2Et
rt, overnight
6
Ar
Ar
CO2Et
N3
Ar
(96%)
MeOH
rt, 24 h
NH2
25 (80%)
20
O
O
O
O
H2, Pd/C (10%)
EtOH, rt, overnight
Figure 2. Effects of compounds 3–7 on SK1 activity. Substrates were
sphingosine (which corresponds to the Km of SK1)7 and 250
M ATP (n = 3, results
expressed as mean inhibition SE). Compounds 3–7 were not substrates for SK1
and were each used at 50 M)
M. The SK1-selective inhibitor BML-25835 (50
3 lM
Ar
l
N3
21
NH2
7 (46%)
l
l
inhibited SK1 activity by 74.5 3.3% (n = 3). Sph denotes sphingosine. The control is
Scheme 5. Synthesis of 3, 4, 6, and 7.
set at 100% and represents the SK1 activity against sphingosine alone.

2506
Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
activated SK1 activity. Third, titration of an inactive SK1 mutant
into cells expressing wild-type SK1 inhibited wild-type SK1 activ-
ity in a concentration-dependent manner. These findings are ex-
plained by a model in which SK1 is a minimal dimer, and that
formation of the wild-type SK1-inactive SK1 mutant dimer pro-
duces an enzyme in which the number of allosteric sites exceed
the competent catalytic site. Suppression of catalytic activity by
an excess of allosteric sites suggests that this site might constitute
an inhibitory domain. We proposed that SK1 contains an auto-
inhibitory domain which exists in an ‘on’ and ‘off’ state in equilib-
rium.7 The ‘on’ state is stabilized by allosteric inhibitors to further
suppress catalytic activity. Conversely, the ‘off’ state is stabilized
by allosteric activators to remove the effect of the inhibitory do-
main on catalytic activity.7 These findings can be adapted to an in-
duced-fit model. In the current study we show that the
replacement of the amino group with an azido group in (S)-
FTY720 vinylphosphonate (2) remarkably changed the compound
from an inhibitor to an activator of SK1. That such a relatively min-
or chemical modification in compound 2 affects the nature of allo-
sterism is entirely consistent with the notion that small changes in
the chemical nature of a compound can be translated into subtle
changes in the conformational state of the allosteric site. These
conformational changes result in profound changes, as evidenced
by turning an allosteric inhibitor into a compound that enhances
enzymatic activity at low micromolar concentrations of com-
pound. Our results also clearly demonstrate the feasibility of using
the (S)-FTY720 vinylphosphonate scaffold to define the structure–
activity relationships of the allosteric site in SK1.
Figure 3. Concentration response of compound 5 on SK1 activity. Substrates were
3 lM sphingosine and 250 lM ATP (n = 3, results expressed as mean inhibi-
tion SE). The control is set at 100% and represents the SK1 activity against
sphingosine alone.
compounds 3–7 are not substrates, since no phosphorylated prod-
uct was formed when Sph was omitted from the assay. In the pres-
ence of Sph, fluoride 3 is a weak inhibitor of SK1 (20% inhibition),
while vinyl-carboxylate 6 is ineffective (Fig. 2), indicating that
replacement of the –P(O)(OH)2 group with –CO2H (which retains a
negative charge at this position of the molecule) eliminates inhibi-
tion. Indeed, we have shown that (S)-vinylphosphonate 2 is a mixed
inhibitor (with ATP) of SK1.8 The enzyme kinetics data for binding of
2 to the allosteric site, after ATP binds first to the active site, are
uncompetitive with ATP, Kiu = 65.0 5.0
lM, and competitive with
5. Experimental section
5.1. Chemistry
ATP, Kic = 10.0 4.6 M. We proposed a ‘see-saw’ model for the
l
binding of 2 to the ATP binding site and to the allosteric site.8 In this
model, compound 2 competes with ATP for the ATP-binding site, but
flips to bind to an allosteric site when ATP binds to the catalytic site.8
Therefore, the –P(O)(OH)2 group appears to be essential for binding
of compound 2 to both the catalytic and allosteric sites. Amino
lactone 7 is also an inhibitor of SK1 (35% inhibition) (Fig. 2). Replace-
ment of the amino group with an azide group in compound 2
produced compounds 4 and 5; surprisingly, azido-fluoride 4 and
azido-alcohol 5 are activators of SK1 at low micromolar concentra-
tions (and they could not replace sphingosine as a substrate)
(Fig. 2). Compounds 4 and 5 both induced a 30–60% stimulation of
SK1 activity at low concentrations. The difference in these azide-
containing compounds concern fluorine in compound 4, which can
only accept hydrogen bonds, and a hydroxyl group in compound 5,
which can both accept and donate hydrogen bonds.
5.1.1. General experimental information
All reactions were carried out under a dry N2 atmosphere using
oven-dried glassware and magnetic stirring. The solvents were
dried before use as follows: THF and Et2O were heated at reflux
over sodium benzophenone ketyl; toluene was heated at reflux
over sodium; CH2Cl2 was dried over CaH2. Anhydrous N,N-diiso-
propylethylamine, acetonitrile, triethylamine, and benzene were
used directly as purchased. Aluminum TLC sheets (silica gel 60
F254) of 0.2-mm thickness were used to monitor the reactions.
The spots were visualized with short wavelength UV light or by
charring after spraying with a solution prepared from one of the
following solutions: phosphomolybdic acid (5.0 g) in 95% EtOH
(100 mL); p-anisaldehyde solution (2.5 mL of p-anisaldehyde,
2 mL of AcOH, and 3.5 mL of concd. H2SO4 in 100 mL of 95% EtOH);
or KMnO4 solution (1.5 g of KMnO4 and 1.5 g of K2CO3 in 100 mL of
0.1% NaOH). Flash chromatography was carried out with silica gel
60 (230–400 ASTM mesh). NMR spectra were obtained on a 400-
MHz or 500-MHz spectrometer. Chemical shifts were referenced
on residual solvent peaks: CDCl3 (d = 7.26 ppm for 1H NMR and
77.00 ppm for 13C NMR); external 85% H3PO4 (d 0.00) was used
for 31P NMR and CFCl3 (d 0.00) for 19F NMR. Optical rotations were
measured at rt in a 1.0-dm cell. High-resolution mass spectra were
acquired by electrospray ionization.
At concentrations below 50
activity with an EC50 of 8.3 3.0
binding to an allosteric site. At concentrations above 50
l
M, compound 5 stimulated SK1
M, n = 3 (Fig. 3), indicative of
M, the
l
l
% activation in response to compound 5 diminished markedly
(Fig. 3). This biphasic response suggests that compound 5 might
bind to the catalytic site (or alter its conformation) to inhibit SK1
activity at these higher concentrations. Similar results were ob-
tained with compound 4 (EC50 of 5.7 5.5 lM, n = 3, data not
shown). The biphasic curves found with both 4 and 5 indicate that
the azido group and not the fluoro or hydroxyl group is responsible
for this phenomenon.
5.1.2. Preparation of 1-octyl-4-vinylbenzene (9)
4. Conclusion
To a stirred suspension of Ph3PCH3Br (22.3 g, 62.4 mmol) in
anhydrous THF (350 mL) was added n-BuLi (25 mL, 2.5 M in hex-
ane, 62.5 mmol) dropwise at ꢂ78 °C. The mixture was stirred for
3 h at ꢂ78 °C. A solution of aldehyde 8 (6.80 g, 31.2 mmol) in
20 mL of THF was added slowly. The cooling bath was removed
after the addition, and the reaction mixture was stirred for 2 h,
during which time the temperature was increased slowly from
We have previously demonstrated that SK1 contains an alloste-
ric site which binds both inhibitors and activators. This is based on
three observations.7 First, compound 2 inhibited SK1 activity with
kinetics indicative of an allosteric interaction. Second, two novel
compounds, Bodipy-sphingosine and a regioisomer of FTY720,

Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
2507
ꢂ78 °C to rt. After 700 mL of hexane was added, the mixture was
5.1.5. Preparation of (E)-dimethyl 2-[(R)-2-(4-
octylphenethyl)oxiran-2-yl]vinylphosphonate (15)
passed through a pressed pad of silica gel, which was rinsed with
hexane/Et2O (10:1). The filtrate was concentrated under reduced
pressure to provide the pure alkene 9 (6.34 g, 94%). 1H NMR
(500 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.22–1.41 (m, 10H), 1.56–
1.67 (m, 2H), 2.55–2.65 (m, 2H), 5.18–5.23 (m, 1H), 5.69–5.76
(m, 1H), 6.67–6.76 (m, 1H), 7.14–7.19 (m, 2H), 7.32–7.38 (m,
2H); 13C NMR (125 MHz, CDCl3) d 14.1, 22.7, 29.26, 29.30, 29.5,
31.5, 31.9, 35.7, 112.7, 126.1, 128.5, 135.0, 136.7, 142.7.
To
a mixture of NaH (57–63% oil dispersion, 240 mg,
6.00 mmol) and 50 mL of THF was added a solution of 1.41 g
(6.08 mmol) of CH2[(P(O)(OMe)2] in 10 mL of THF at 0 °C. After
the mixture was stirred for 30 min, a solution of 14 (585 mg,
2.03 mmol) in 10 mL of THF was added. The reaction mixture
was stirred for 1 h at 0 °C, quenched with saturated aqueous NH4Cl
solution, and extracted with EtOAc. The organic layer was dried
(MgSO4) and concentrated under reduced pressure. The residue
was purified by chromatography (elution with EtOAc) to give
750 mg (94%) of 15. 1H NMR (400 MHz, CDCl3) d 0.83–0.92 (m,
3H), 1.20–1.38 (m, 10H), 1.53–1.64 (m, 2H), 1.90–2.01 (m, 1H),
2.08–2.18 (m, 1H), 2.53–2.59 (m, 2H), 2.62–2.77 (m, 3H), 2.87 (t,
J = 5.4 Hz, 1H), 3.72 (d, J = 5.5 Hz, 3H), 3.74 (d, J = 5.5 Hz, 3H),
5.95 (dd, J = 17.2, 19.4 Hz, 1H), 6.83 (dd, J = 17.2, 22.2 Hz, 1H),
7.05–7.13 (m, 4H); 13C NMR (100 MHz, CDCl3) d 14.1, 22.6, 29.2,
29.3, 29.4, 30.6, 31.5, 31.8, 35.2, 35.5, 52.38 (d, J = 5.4 Hz), 52.41
(d, J = 5.4 Hz), 55.9, 58.2 (d, J = 24.0 Hz), 116.5 (d, J = 189.6 Hz),
128.0, 128.5, 137.9, 140.8, 151.6 (d, J = 6.5 Hz); 31P NMR
(162 MHz, CDCl3) d 20.6; ESI-HRMS (M+H)+ calcd for C22H36O4P+
395.2346, found 395.2346.
5.1.3. Preparation of 2-methylene-4-(4-octylphenyl)butan-1-ol
(12)
A solution of alkene 9 (4.14 g, 19.1 mmol) in THF (15 mL) was
added dropwise to a solution of 9-BBN (200 mL, 0.5 M in THF,
100 mmol) at 0 °C. The reaction mixture was allowed to warm to
rt. After 1 h at rt, the mixture was cooled to 0 °C and 3 M aqueous
NaOH (210 mmol, 70 mL) was added. The mixture was stirred vigor-
ously for 1 h at rt, and then was added to a suspension of (2-iodoal-
lyloxy)(tert-butyl)dimethylsilane10 (6.35 g, 21.3 mmol) and
Pd(dppf)Cl2ꢀCH2Cl2 (570 mg, 0.70 mmol) in THF (50 mL). The reac-
tion mixture was stirred at 60 °C for 3 h. After cooling to rt, the mix-
ture was added to ethanolamine (12 mL) to precipitate the 9-BBN/
ethanolamine complex. The mixture was stirred overnight at rt,
poured into saturated aqueous NH4Cl solution, and extracted with
Et2O. The combined organic extracts were concentrated under re-
duced pressure and suspended in hexane/Et2O (100 mL, 10:1). The
mixture was filtered through a short pad of silica gel, which was
washed with hexane/Et2O (10:1). The filtrate was concentrated un-
der reduced pressure to provide the crude coupling product.
To a solution of crude coupling product in MeOH (200 mL) was
5.1.6. Preparation of (E)-ethyl 3-[(R)-2-(4-
octylphenethyl)oxiran-2-yl]acrylate (16)
K2CO3 (4.51 g, 32.6 mmol) was dissolved in water (10 mL), and
the solution was cooled to rt. This aqueous solution was then
added to a solution of aldehyde 14 (350 mg, 1.21 mmol) and (i-
PrO)2P(O)CH2CO2Et (866 lL, 3.64 mmol) in 2-propanol (10 mL) at
0 °C. The reaction mixture was allowed to warm to rt and was stir-
red overnight at rt. The reaction mixture was separated and the
lower aqueous layer was extracted with EtOAc. The combined or-
ganic layers were dried (Na2SO4), filtered, and concentrated under
reduced pressure. The residue was purified by flash chromatogra-
phy (hexane/EtOAc 3:1) to give 409 mg (94%) of 16. 1H NMR
(500 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 13H), 1.54–
1.63 (m, 2H), 1.93–2.01 (m, 1H), 2.09–2.16 (m, 1H), 2.57 (t,
J = 7.7 Hz, 2H), 2.65–2.75 (m, 3H), 2.88 (d, J = 5.4 Hz, 1H), 4.21 (q,
J = 7.1 Hz, 2H), 6.10 (d, J = 15.7 Hz, 1H), 6.91 (d, J = 15.7 Hz, 1H),
7.06–7.12 (m, 4H); 13C NMR (100 MHz, CDCl3) d 14.1, 14.2, 22.6,
29.2, 29.3, 29.5, 30.7, 31.5, 31.9, 35.45, 35.52, 55.8, 57.6, 60.6,
122.2, 128.1, 128.5, 138.1, 140.8, 146.6, 166.0; ESI-HRMS
added 6 M HCl (100 lL, 0.6 mmol). After being stirred at rt for 1 h,
the reaction mixture was quenched by slowly adding saturated
aqueous NaHCO3 solution. MeOH was removed under reduced
pressure, and the aqueous phase was extracted with Et2O. In order
to completely remove the boron complex, the combined organic
extracts were treated with 100 mL of 15% NaOH, followed by drop-
wise addition of 30% H2O2 (50 mL). The mixture was stirred at rt
for 0.5 h. The aqueous phase was extracted with Et2O, and the com-
bined organic phases were dried (MgSO4) and concentrated under
reduced pressure. The residue was purified by flash chromatogra-
phy (hexane/EtOAc 8:1, 6:1, to 4:1) to afford 3.67 g (70%) of 12.
The 1H and 13C NMR spectra of 12 matched the reported data.5
ESI-HRMS (M+Na)+ calcd for C19H30NaO+ 297.2189, found
297.2193.
(M+Na)+ calcd for C23H34NaOþ 381.2400, found 381.2401.
3
5.1.7. Preparation of (S,E)-dimethyl 3-azido-3-(hydroxymethyl)-
5-(4-octylphenyl)pent-1-enylphosphonate (17)
5.1.4. Preparation of (S)-2-(4-octylphenethyl)oxirane-2-
carbaldehyde (14)
Ti(O-i-Pr)4 (2.4 mL, 8.02 mmol) and TMSN3 (2.2 mL, 16.6 mmol)
were added to anhydrous toluene (50 mL), and the mixture was
heated at reflux (85 °C) under N2 for at least 5 h. A solution of epox-
ide 15 (670 mg, 1.70 mmol) in anhydrous toluene (10 mL) was
added to the above solution in one portion. The mixture was stir-
red for 15 min at 85 °C and was then cooled to rt. The solvent
was removed under reduced pressure. Et2O (20 mL) was added, fol-
lowed by 10% HCl (40 mL). The solution was stirred at rt until two
clear phases appeared. The aqueous phase was extracted with
Et2O. The organic phase was dried over MgSO4, filtered, and con-
centrated under reduced pressure. Purification by flash chromatog-
raphy (elution with EtOAc) afforded 17 (500 mg, 68%) as a yellow
oil. 1H NMR (500 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m,
10H), 1.54–1.62 (m, 2H), 1.90–1.99 (m, 1H), 2.03–2.11 (m, 1H),
2.51–2.67 (m, 4H), 3.01 (s, 1H, OH), 3.70–3.80 (m, 8H), 6.03 (dd,
J = 17.1, 19.3 Hz, 1H), 6.72 (dd, J = 17.2, 22.7 Hz, 1H), 7.06–7.12
(m, 4H); 13C NMR (100 MHz, CDCl3) d 14.1, 22.6, 29.2, 29.3,
29.46, 29.50, 31.6, 31.9, 35.5, 36.0, 52.53 (d, J = 5.5 Hz), 52.55 (d,
J = 5.5 Hz), 67.4, 69.0 (d, J = 19.4 Hz), 118.1 (d, J = 186.9 Hz),
128.1, 128.6, 137.9, 140.9, 151.0 (d, J = 6.3 Hz); 31P NMR
Dess–Martin periodinane (3.18 g, 7.50 mmol) was added por-
tionwise to a solution of alcohol 13 (1.45 g, 85% purity, 4.24 mmol,)
in 4 mL of pyridine and 50 mL of CH2Cl2. After 2 h, the reaction
mixture was diluted with Et2O (100 mL), and saturated aqueous
NaHCO3 solution (20 mL) and saturated aqueous Na2S2O3 (20 mL)
were added to the reaction mixture. After vigorous stirring for
10 min, the mixture was extracted with Et2O. The combined organ-
ic layers were washed with saturated aqueous NaHCO3 solution,
dried (MgSO4), and filtered. The solvent was removed under re-
duced pressure, and the residue was purified by flash chromatog-
raphy on silica gel (EtOAc/hexane 1:10, 1:8, to 1:6) to afford
1.1 g (90%) of aldehyde 14. 1H NMR (500 MHz, CDCl3) d 0.83–
0.92 (m, 3H), 1.22–1.36 (m, 10H), 1.56–1.64 (m, 2H), 2.01–2.08
(m, 1H), 2.20–2.28 (m, 1H), 2.58 (t, J = 7.8 Hz, 2H), 2.72 (t,
J = 8.2 Hz, 2H), 2.99 (d, J = 4.6 Hz, 1H), 3.04 (d, J = 4.6 Hz, 1H),
7.10–7.13 (m, 4H), 8.89 (s, 1H); 13C NMR (125 MHz, CDCl3) d
14.1, 22.6, 29.2, 29.3, 29.5, 29.9, 30.2, 31.5, 31.9, 35.5, 49.8, 60.9,
128.1, 128.5, 138.0, 140.8, 198.8; ESI-HRMS (M+Na)+ calcd for
+
C
19H28NaO2 311.1982, found 311.1986.

2508
Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
(162 MHz, CDCl3) d 20.5; ESI-HRMS (M+H)+ calcd for C22H36N3O4P+
C
23H35N3NaO3 424.2571, found 424.2570. Data of 21: 1H NMR
+
438.2516, found 438.2519.
(400 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.53–
1.63 (m, 2H), 1.86–2.04 (m, 2H), 2.53–2.59 (m, 2H), 2.62–2.78
(m, 2H), 4.34–4.46 (m, 2H), 6.20 (d, J = 9.8 Hz, 1H), 6.75 (d,
J = 9.8 Hz, 1H), 7.04–7.13 (m, 4H); 13C NMR (100 MHz, CDCl3) d
14.1, 22.6, 29.0, 29.2, 29.3, 29.4, 31.5, 31.8, 35.5, 37.6, 58.7, 73.6,
122.6, 128.0, 128.7, 137.0, 141.2, 144.7, 161.6; ESI-HRMS
5.1.8. Preparation of (R,E)-dimethyl 3-(azidomethyl)-3-
hydroxy-5-(4-octylphenyl)pent-1-enylphosphonate (18)
A
mixture of 15 (45 mg, 0.114 mmol), NaN3 (37 mg,
0.569 mmol), and NH4Cl (49 mg, 0.916 mmol) in EtOH (2 mL)
was heated at 100 °C via microwave radiation for 20 min. The reac-
tion mixture was allowed to cool to rt and was then filtered
through a thin pad of Celite (eluting with EtOAc), and the filtrate
was concentrated under reduced pressure. The crude material
was purified by flash chromatography (elution with EtOAc) on sil-
ica gel to provide azide 18 (36 mg, 72%). 1H NMR (400 MHz, CDCl3)
d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.52–1.62 (m, 2H), 1.80–
1.90 (m, 1H), 1.92–2.02 (m, 1H), 2.48–2.59 (m, 3H), 2.62–2.72 (m,
1H), 3.13 (s, 1H, OH), 3.39 (s, 2H), 3.74 (dd, J = 2.4, 11.1 Hz, 6H),
6.10 (dd, J = 17.1, 20.3 Hz, 1H), 6.78 (dd, J = 17.1, 22.3 Hz, 1H),
7.04–7.12 (m, 4H); 13C NMR (100 MHz, CDCl3) d 14.1, 22.6, 29.0,
29.2, 29.3, 29.4, 31.5, 31.8, 35.5, 39.3, 52.46 (d, J = 5.6 Hz), 52.49
(d, J = 5.6 Hz), 59.1, 76.0 (d, J = 19.2 Hz), 116.6 (d, J = 187.8 Hz),
128.1, 128.5, 138.2, 140.8, 154.9 (d, J = 5.5 Hz); 31P NMR
(M+Na)+ calcd for C21H29N3NaO2 378.2152, found 378.2151.
+
5.1.11. Preparation of (S,E)-dimethyl 3-amino-3-
(hydroxymethyl)-5-(4-octylphenyl)pent-1-enylphosphonate
(22) and (S)-FTY720 vinylphosphonate (2)
To a solution of azide 17 (50 mg, 0.114 mmol) in 95% MeOH
(3 mL) was added SnCl2 (85 mmol, 0.448 mmol) at rt. After the
reaction mixture was stirred at rt overnight, the solvent was re-
moved under reduced pressure. The residue was diluted with
Et2O, and the mixture was neutralized with aqueous saturated
NaHCO3 solution, followed by aqueous NaOH solution (15%). The
aqueous layer was extracted with Et2O, and the combined organic
layers were dried (K2CO3), filtered, and concentrated under re-
duced pressure. The residue was purified by flash chromatography
(CHCl3/MeOH 20:1, CHCl3/MeOH/NH4OH 130:25:4, CHCl3/MeOH/
AcOH 65:25:4) to give 8 mg (17%) of 22 and 30 mg (69%) of (S)-
FTY720 vinylphosphonate (2). Data of 22: The 1H, 13C, and 31P
NMR spectra of 22 matched the reported data;5 see the Supple-
(162 MHz, CDCl3)
d
20.9; ESI-HRMS (M+Na)+ calcd for
C22H36N3NaO4P+ 460.2336, found 460.2334.
5.1.9. Preparation of (R,E)-ethyl 4-(azidomethyl)-4-hydroxy-6-
(4-octylphenyl)hex-2-enoate (19)
mentary data; ½a D25
ꢁ
þ 20:0 (c 0.18, CHCl3) [lit.5
½
a 2D5
ꢁ
þ 18:8 (c 1.52,
A
mixture of 16 (47 mg, 0.131 mmol), NaN3 (26 mg,
CHCl3)]; ESI-HRMS (M+H)+ calcd for C22H40NO4P+ 414.2758, found
414.2769. Data of 2: 1H, 13C, and 31P NMR spectra of 2 matched the
reported data;5 see the Supplementary data.
0.400 mmol), and NH4Cl (21 mg, 0.40 mmol) in EtOH (2 mL) was
heated at 100 °C via microwave radiation for 1 h. The reaction mix-
ture was allowed to cool to rt and was then filtered through a thin
pad of Celite (eluting with EtOAc), and the filtrate was concen-
trated under reduced pressure. The crude material was purified
by flash chromatography (hexane/EtOAc 20:1, 10:1, to 8:1) on sil-
ica gel to provide azide 19 (34 mg, 65%). 1H NMR (400 MHz, CDCl3)
d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 13H), 1.52–1.63 (m, 2H), 1.82–
1.92 (m, 1H), 1.93–2.03 (m, 1H), 2.50–2.60 (m, 3H), 2.61–2.71 (m,
1H), 3.37–3.45 (m, 2H), 4.23 (q, J = 7.1 Hz, 2H), 6.19 (d, J = 15.6 Hz,
1H), 6.90 (d, J = 15.6 Hz, 1H), 7.04–7.12 (m, 4H); 13C NMR
(100 MHz, CDCl3) d 14.1, 14.2, 22.6, 29.1, 29.2, 29.3, 29.4, 31.5,
31.9, 35.5, 39.5, 59.5, 60.7, 75.4, 122.3, 128.1, 128.6, 138.2, 140.8,
5.1.12. Preparation of (S,E)-3-azido-3-(hydroxymethyl)-5-(4-
octylphenyl)pent-1-enylphosphonic acid (5)
To a solution of 17 (10 mg, 0.023 mmol) in 2 mL of dry CH2Cl2 at
rt was added 30 lL (0.23 mmol) of TMSBr. After the reaction mix-
ture had been stirred for 6 h, the solvent was removed, and the res-
idue was dried and dissolved in 2 mL of 95% MeOH with stirring for
1 h. Removal of the solvent afforded 9 mg (100%) of 5. 1H NMR
(500 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.52–
1.63 (m, 2H), 1.85–1.94 (m, 1H), 1.97–2.07 (m, 1H), 2.52–2.68
(m, 4H), 3.67–3.74 (m, 2H), 6.09–6.22 (m, 1H), 6.45–6.59 (m,
1H), 7.07–7.13 (m, 4H); 13C NMR (125 MHz, CDCl3) d 13.7, 22.4,
28.98, 29.04, 29.1, 29.2, 29.4, 31.3, 31.6, 35.2, 36.0, 68.4 (d,
J = 18.4 Hz), 127.8, 128.4, 138.1, 140.5, 146.7; 31P NMR (202 MHz,
149.1, 166.1; ESI-HRMS (M+Na)+ calcd for C23H35N3NaO3
424.2571, found 424.2573.
+
5.1.10. Preparation of (S,E)-ethyl 4-azido-4-(hydroxymethyl)-6-
(4-octylphenyl)hex-2-enoate (20) and (S)-5-(4-octylphenethyl)-
5-azido-5,6-dihydropyran-2-one (21)
CDCl3) d
15.5; ESI-HRMS (M+H)+ calcd for C22H33N3O4P+
410.2203, found 410.2206.
Ti(O-i-Pr)4 (400 mg, 1.40 mmol) and TMSN3 (394
lL,
5.1.13. Preparation of (S,E)-dimethyl 3-azido-3-(fluoromethyl)-
5-(4-octylphenyl)pent-1-enylphosphonate (23)
3.00 mmol) were added to anhydrous toluene (10 mL), and the
mixture was heated at reflux at 75 °C under N2 overnight. A solu-
tion of epoxide 16 (129 mg, 0.36 mmol) in anhydrous toluene
(5 mL) was added to the above solution in one portion. The mixture
was stirred for 15 min at 85 °C and was cooled to rt. The solvent
was removed under reduced pressure. Et2O (20 mL) was added, fol-
lowed by 10% HCl (10 mL). The solution was stirred at rt until two
clear phases appeared. The aqueous phase was extracted with
Et2O. The organic phase was dried over MgSO4, filtered, and con-
centrated under reduced pressure. Purification of the residue by
flash chromatography (elution with hexane/EtOAc, 10:1, 6:1, 5:1,
to 4:1) afforded 21 (58 mg, 45%) and 20 (37 mg, 26%). Data of 20:
1H NMR (400 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m,
13H), 1.52–1.63 (m, 2H), 1.92–2.01 (m, 1H), 2.03–2.13 (m, 1H),
2.53–2.70 (m, 4H), 3.69–3.78 (m, 2H), 4.22 (q, J = 7.1 Hz, 2H),
6.16 (d, J = 15.7 Hz, 1H), 6.84 (d, J = 15.7 Hz, 1H), 7.04–7.12 (m,
4H); 13C NMR (100 MHz, CDCl3) d 14.1, 14.2, 22.6, 29.2, 29.3,
29.4, 29.5, 31.5, 31.9, 35.5, 36.3, 60.9, 67.5, 68.0, 123.6, 128.1,
128.6, 137.9, 140.9, 145.1, 165.9; ESI-HRMS (M+Na)+ calcd for
To a solution of alcohol 17 (75 mg, 0.171 mmol) in anhydrous
CH2Cl2 (1.0 mL) was added diethylaminosulfur trifluoride (DAST)
(68
lL, 0.515 mmol) at ꢂ78 °C. The reaction mixture was stored
at ꢂ78 °C overnight, and then was stirred at rt for 3 h. The mixture
was poured into aqueous saturated NaHCO3 solution, the aqueous
phase was extracted with CH2Cl2, and the combined CH2Cl2 layers
were dried (Na2SO4), filtered, and concentrated under reduced
pressure. Purification by flash chromatography (elution with
PhMe/EtOAc, 1:1 to 100% EtOAc) afforded 23 (49 mg, 65%), along
with 10 mg (13%) of recovered alcohol 17. 1H NMR (500 MHz,
CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.53–1.63 (m,
2H), 1.88–2.02 (m, 1H), 2.12–2.22 (m, 1H), 2.51–2.60 (m, 3H),
2.64–2.72 (m, 1H), 3.34–3.53 (m, 2H), 3.77 (d, J = 11.1 Hz, 6H),
6.12 (dd, J = 17.1, 18.8, 1H), 6.73 (ddd, J = 17.1, 21.4, 22.8 Hz, 1H),
7.03–7.13 (m, 4H); 13C NMR (100 MHz, CDCl3) d 14.1, 22.6, 28.6,
28.7, 29.2, 29.3, 29.4, 31.5, 31.9, 35.5, 37.4 (d, J = 21.9 Hz), 52.5 (t,
J = 5.5 Hz), 56.5 (d, J = 23.9 Hz), 97.5 (dd, J = 19.6, 184.9 Hz), 118.0
(dd, J = 9.6, 187.8 Hz), 128.0, 128.6, 137.5, 141.0, 149.5 (dd,

Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
2509
J = 5.8, 20.4 Hz); 31P NMR (162 MHz, CDCl3) d 19.6; 19F NMR
and the mixture was neutralized with aqueous saturated NaHCO3
solution, followed by aqueous NaOH solution (15%). The aqueous
layer was extracted with Et2O, and the combined organic layers
were dried (MgSO4), filtered, and concentrated under reduced
pressure. The residue was purified by flash chromatography
(CHCl3/MeOH 40:1 to 20:1) to give 9 mg (80%) of 25. 1H NMR
(500 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 13H), 1.52–
1.63 (m, 2H), 1.73–1.89 (m, 2H), 2.47–2.64 (m, 4H), 3.48–3.54
(m, 2H), 4.22 (q, J = 7.1 Hz, 2H), 6.04 (d, J = 15.9 Hz, 1H), 6.98 (d,
J = 15.9 Hz, 1H), 7.05–7.11 (m, 4H); 13C NMR (100 MHz, CDCl3) d
14.1, 14.2, 21.9, 22.7, 29.2, 29.3, 29.4, 29.5, 29.7, 31.5, 31.9, 35.5,
39.7, 58.2, 60.6, 69.0, 121.3, 128.1, 128.5, 138.6, 140.7, 152.2,
(376 MHz, CDCl3)
C
d
ꢂ164.5; ESI-HRMS (M+H)+ calcd for
22H36FN3O3P+ 440.2473, found 440.2472.
5.1.14. Preparation of (S,E)-dimethyl 3-amino-3-(fluoromethyl)-
5-(4-octylphenyl)pent-1-enylphosphonate (24)
A solution of azide 23 (40 mg, 0.091 mmol) in ethanol (2 mL)
was treated with Lindlar’s catalyst (46 mg). The suspension was
purged with H2 for approximately 10 min and then was stirred
for 2 h under a balloon filled with H2. The crude reaction mixture
was filtered through a short pad of Celite, washed with EtOAc,
and concentrated. The residue was purified by flash chromatogra-
phy (CHCl3/MeOH 20:1) to give 19 mg (51%) of 24. 1H NMR
(500 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.53–
1.63 (m, 2H), 1.84–1.98 (m, 1H), 2.03–2.13 (m, 1H), 2.51–2.59
(m, 3H), 2.64–2.71 (m, 1H), 2.81–3.00 (m, 2H), 3.76 (dd, J = 3.5,
11.2 Hz, 6H), 6.05 (dd, J = 17.2, 19.9, 1H), 6.72 (ddd, J = 17.2, 22.0,
23.1 Hz, 1H), 7.04–7.12 (m, 4H); 13C NMR (100 MHz, CDCl3) d
14.1, 22.6, 28.8, 28.9, 29.2, 29.3, 29.4, 31.5, 31.8, 35.5, 37.6 (d,
J = 21.8 Hz), 48.9 (d, J = 25.0 Hz), 52.5 (d, J = 5.5 Hz), 98.6 (dd,
J = 18.8, 180.9 Hz), 117.3 (dd, J = 9.6, 188.2 Hz), 128.0, 128.5,
138.0, 140.8, 151.5 (dd, J = 5.7, 20.9 Hz); 31P NMR (162 MHz, CDCl3)
d 20.3; 19F NMR (376 MHz, CDCl3) d ꢂ170.3.
166.3; ESI-HRMS (M+H)+ calcd for C23H37NOþ 376.2846, found
3
376.2849.
5.1.18. Preparation of (S,E)-4-amino-4-(hydroxymethyl)-6-(4-
octylphenyl)hex-2-enoic acid (6)
To a solution of ester 25 (9 mg, 0.024 mmol) in THF/MeOH/H2O
(1 mL/1 mL/0.5 mL) was added LiOHꢀH2O (20 mg, 0.48 mmol) at rt.
The reaction mixture was stirred overnight at rt, and then was neu-
tralized with 1 M HCl (5 mL). The mixture was separated and the
aqueous layer was extracted with CH2Cl2. The combined organic
layers were dried (Na2SO4), filtered, and concentrated under re-
duced pressure to provide 6 (8 mg, 96%) as a white solid. 1H
NMR (500 MHz, CDCl3/CD3OD/CD3CO2D 80:20:1) d 0.83–0.92 (m,
3H), 1.20–1.36 (m, 10H), 1.54–1.63 (m, 2H), 2.01–2.18 (m, 2H),
2.53–2.69 (m, 4H), 3.76–3.82 (m, 2H), 6.15 (d, J = 16.3 Hz, 1H),
6.89 (d, J = 16.3 Hz, 1H), 7.09–7.12 (m, 4H); 13C NMR (100 MHz,
CDCl3/CD3OD/CD3CO2D 80:20:1) d 13.7, 22.4, 28.6, 29.0, 29.1,
29.2, 29.5, 31.3, 31.7, 35.3, 36.1, 61.1, 64.0, 124.0, 127.8, 128.4,
5.1.15. Preparation of (S,E)-3-amino-3-(fluoromethyl)-5-(4-
octylphenyl)pent-1-enylphosphonic acid (3)
To a solution of 24 (19 mg, 0.046 mmol) in 2 mL of dry CH2Cl2 at
rt was added 120 lL (0.92 mmol) of TMSBr. After the reaction mix-
ture had been stirred for 6 h, the solvent was removed, and the res-
idue was dried and dissolved in 2 mL of 95% MeOH with stirring for
1 h. Removal of the solvent afforded 18 mg (100%) of 3 as a white
solid. 1H NMR (400 MHz, CDCl3/CD3OD/CD3CO2D 80:20:1) d 0.83–
0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.50–1.63 (m, 2H), 1.97–2.16 (m,
2H), 2.50–2.73 (m, 4H), 3.25–3.48 (m, 2H), 6.26–6.43 (m, 1H),
6.46–6.68 (m, 1H), 7.03–7.12 (m, 4H); 13C NMR (125 MHz, CDCl3/
CD3OD/CD3CO2D 80:20:1) d 13.7, 22.3, 27.93, 27.96, 28.96, 29.03,
29.2, 31.3, 31.6, 35.2, 37.9 (d, J = 21.0 Hz), 45.1 (d, J = 22.4 Hz),
95.0 (dd, J = 183.4, 20.2 Hz),, 125.8 (d, J = 180.6 Hz), 127.7, 128.3,
137.1, 140.7, 142.4 (d, J = 20.5 Hz); 31P NMR (162 MHz, CDCl3/
CD3OD/CD3CO2D 80:20:1) d 13.3; 19F NMR (376 MHz, CDCl3/
CD3OD/CD3CO2D 80:20:1) d ꢂ167.7; ESI-HRMS (M+H)+ calcd for
136.9, 140.9, 143.4, 167.3; ESI-HRMS (M+H)+ calcd for C21H34NOþ
3
348.2533, found 348.2536.
5.1.19. Preparation of (S)-5-(4-octylphenethyl)-5-amino-
tetrahydropyran-2-one (7)
A solution of 21 (14 mg, 0.039 mmol) in EtOH (5 mL) was trea-
ted with Pd/C (14 mg, 10% w/w). The suspension was purged with
H2 for approximately 10 min and then was stirred overnight at rt
under a balloon filled with H2. The crude reaction mixture was fil-
tered through a short pad of Celite, washed with CH2Cl2, and con-
centrated. The residue was purified by flash chromatography
(CHCl3/MeOH 20:1 to 10:1) to give 6 mg (46%) of 7. 1H NMR
(400 MHz, CDCl3) d 0.83–0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.53–
1.63 (m, 2H), 1.80–1.90 (m, 2H), 1.90–2.09 (m, 2H), 2.34–2.51
(m, 2H), 2.51–2.64 (m, 4H), 3.50 (d, J = 11.3 Hz, 1H), 3.61 (d,
J = 11.3 Hz, 1H), 7.04–7.13 (m, 4H); 13C NMR (100 MHz, CDCl3) d
14.1, 22.6, 27.8, 29.3, 29.4, 29.5, 29.6, 30.8, 31.6, 31.9, 35.5, 38.8,
63.5, 68.3, 128.1, 128.5, 138.4, 140.7, 178.5; ESI-HRMS (M+Na)+
C
20H34FNO3P+ 386.2255, found 386.2256.
5.1.16. Preparation of (S,E)-3-azido-3-(fluoromethyl)-5-(4-
octylphenyl)pent-1-enylphosphonic acid (4)
To a solution of 23 (10 mg, 0.023 mmol) in 2 mL of dry CH2Cl2 at
rt was added 30 lL (0.23 mmol) of TMSBr. After the reaction mix-
ture had been stirred for 6 h, the solvent was removed, and the res-
idue was dried and dissolved in 2 mL of 95% MeOH with stirring for
1 h. Removal of the solvent afforded 10 mg (100%) of 4 as a white
solid. 1H NMR (400 MHz, CDCl3/CD3OD/CD3CO2D 80:20:1) d 0.83–
0.92 (m, 3H), 1.20–1.36 (m, 10H), 1.52–1.63 (m, 2H), 1.87–2.05 (m,
1H), 2.10–2.24 (m, 1H), 2.50–2.69 (m, 4H), 3.40–3.50 (m, 2H),
6.16–6.28 (m, 1H), 6.49–6.68 (m, 1H), 7.04–7.14 (m, 4H); 13C
NMR (125 MHz, CDCl3/CD3OD/CD3CO2D 80:20:1) d 13.7, 22.3,
28.3, 29.0, 29.2, 31.3, 31.6, 35.2, 37.1, 37.3, 56.3 (d, J = 24.4 Hz),
97.2 (d, J = 184.2 Hz), 127.8, 128.3, 137.7, 140.6, 145.2; 31P NMR
calcd for C21H31NNaOþ 354.2404, found 354.2404.
2
5.2. Biology
5.2.1. Cell culture
HEK 293 cells stably over-expressing GFP-SK1 (30-fold increase
in SK1 activity versus vector-transfected cells) were cultured in
DMEM supplemented with 10% European fetal calf serum, 100 U/
mL penicillin, 100 lg/mL streptomycin, 1% non-essential amino
(162 MHz, CDCl3/CD3OD/CD3CO2D 80:20:1)
d
15.2; ESI-HRMS
acids, and 0.8% geneticin at 37 °C in 5% CO2.
(M+H)+ calcd for C20H32FNO3P+ 386.2255, found 386.2256.
5.2.2. Sphingosine kinase 1 assay
5.1.17. Preparation of (S,E)-ethyl 4-amino-4-(hydroxymethyl)-
6-(4-octylphenyl)hex-2-enoate (25)
To a solution of azide 20 (12 mg, 0.03 mmol) in anhydrous
MeOH (1 mL) was added SnCl2 (27 mmol, 0.142 mmol) at rt. After
the reaction mixture was stirred at rt for 24 h, the solvent was re-
moved under reduced pressure. The residue was diluted with Et2O,
Sphingosine was solubilized in Triton X-100 (final concentra-
tion, 0.063% w/v) and combined with 20 mM Tris (pH 7.4), 1 mM
EDTA, 1 mM Na3VO4, 40 mM b-glycerophosphate, 1 mM NaF,
0.007% (v/v) b-mercaptoethanol, 20% (v/v) glycerol, 10
tinin, 10 g/mL soybean trypsin inhibitor, 1 mM PMSF, and 0.5 mM
4-deoxypyridoxine. SK1 activity was determined by incubating
lg/mL apro-
l

2510
Z. Liu et al. / Bioorg. Med. Chem. 21 (2013) 2503–2510
8. Lim, K. G.; Tonelli, F.; Berdyshev, E. V.; Gorshkova, I.; Leclercq, T.; Pitson, S. M.;
30 lg of recombinant SK1 in lysates from HEK 293 cells for 30 min
Bittman, R.; Pyne, S.; Pyne, N. J. Int. J. Biochem. Cell Biol. 2012, 44, 1457.
9. For a review of the Sharpless asymmetric epoxidation reaction, see: Katsuki, T.;
Martin, V. S. Org. React. 1996, 48, 1.
at 30 °C, in the presence of
3 lM sphingosine and 250 lM
[
c
-
32P]ATP (4.4 ꢃ 104 cpm/nmol) in 10 mM MgCl2 with or without
10. For a review of the B-alkyl Suzuki cross-coupling coupling reaction, see:
Chemler, S. R.; Trauner, D.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2001, 40,
4544.
11. Nicolaou, K. C.; Li, Y.; Uesaka, N.; Koftis, T. V.; Vyskocil, S.; Ling, T.;
Govindasamy, M.; Qian, W.; Bernal, F.; Chen, D. Y.-K. Angew. Chem., Int. Ed.
2003, 42, 3643.
12. Krishnamurthy, S.; Brown, H. C. J. Org. Chem. 1975, 40, 1864.
13. The crude product of allylic alcohol 12 contained a boron complex which had a
very similar Rf as 12 on TLC (hexane/EtOAc 3:1).
14. The yield of the side product of epoxide opening was calculated based on the
1H NMR spectrum of 13. The side product could not be separated at the stage of
SAE, but after Dess–Martin oxidation the side product was removed, and epoxy
aldehyde 14 was obtained in pure form.
inhibitor dissolved in DMSO or control (5% v/v DMSO). SK1 reac-
tions were terminated by the addition of 500 lL of 1-butanol and
mixed with 1 mL of 2 M KCl. The organic phase containing [32P]-
S1P was then extracted by washing twice with 1 mL of 2 M KCl be-
fore quantification by Cerenkov counting.
5.2.3. Testing compounds for ability to act as SK1 substrates in
assays in which sphingosine was omitted
The presence or absence of [32P]-labeled products in the butanol
phase was assessed by Cerenkov counting. No [32P]-labeled prod-
ucts were formed from compounds 3–7. We have previously used
this method to establish that other sphingosine-based compounds
may serve as substrates for SK1.36
15. Gao, Y.; Klunder, J. M.; Hanson, R. M.; Masamune, H.; Ko, S. Y.; Sharpless, K. B. J.
Am. Chem. Soc. 1987, 109, 5765.
16. The E/Z ratio was determined by the integration of the vinylic proton signals in
the 1H NMR spectrum.
17. Liu, Z.; Gong, Y.-Q.; Byun, H.-S.; Bittman, R. New J. Chem. 2010, 34, 470.
18. Nagaoka, H.; Kishi, Y. Tetrahedron 1981, 37, 3873.
19. Feldman, A. K.; Colasson, B.; Sharpless, K. B.; Fokin, V. V. J. Am. Chem. Soc. 2005,
127, 13444.
20. For a review of ring-opening reactions of allylic epoxides, see: Pineschi, M.;
Bertolini, F.; Di Bussolo, V.; Crotti, P. Curr. Org. Synth. 2009, 6, 290.
21. Righi, G.; Manni, L. S.; Bovicelli, P.; Pelagalli, R. Tetrahedron Lett. 2011, 52, 3895.
22. Marié, J.-C.; Courillon, C.; Malacria, M. Arkivoc 2007, 5, 277.
23. Ha, J. D.; Kim, S. Y.; Lee, S. J.; Kang, S. K.; Ahn, J. H.; Kim, S. S.; Choi, J.-K.
Tetrahedron Lett. 2004, 45, 5969.
Acknowledgment
This work was supported by NIH grant HL083187 (R.B.).
Supplementary data
Supplementary data (1H, 13C, 31P, and 19F NMR spectral data for
all compounds) associated with this article can be found, in the on-
line version, at http://dx.doi.org/10.1016/j.bmc.2013.02.042.
24. Miyashita, M.; Mizutani, T.; Tadano, G.; Iwata, Y.; Miyazawa, M.; Tanino, K.
Angew. Chem., Int. Ed. 2005, 44, 5094.
25. Wipf, P.; Fritch, P. C. J. Org. Chem. 1994, 59, 4875.
26. Caron, M.; Carlier, P. R.; Sharpless, K. B. J. Org. Chem. 1988, 53, 5185.
27. Ono, M.; Zhao, X. Y.; Shida, Y.; Akita, H. Tetrahedron 2007, 63, 10140.
28. Cai, C.; Liu, J.; Du, Y.; Linhardt, R. J. J. Org. Chem. 2010, 75, 5754.
29. Matsushima, Y.; Kino, J. Eur. J. Org. Chem. 2010, 2206.
30. Matsushima, Y.; Kino, J. Synthesis 2011, 1290.
References and notes
1. Pyne, S.; Bittman, R.; Pyne, N. J. Cancer Res. 2011, 71, 6576.
2. Pitman, M. R.; Woodcock, J. M.; Lopez, A. F.; Pitson, S. M. Curr. Mol. Med. 2012,
12, 1207.
31. Bayley, H.; Standring, D. N.; Knowles, J. R. Tetrahedron Lett. 1978, 39, 3633.
32. Corey, E. J.; Nicolaou, K. C.; Balanson, R. D.; Machida, Y. Synthesis 1975, 590.
33. (a) Evans, D. A.; Britton, T. C.; Ellman, J. A.; Dorow, R. L. J. Am. Chem. Soc. 1990,
112, 4011; (b) The reaction of 17 and SnCl2 in anhydrous MeOH produced 22
(ꢄ80% yield) only.
3. Brinkmann, V. Br. J. Pharm. 2009, 158, 1173.
4. Brinkmann, V.; Billich, A.; Baumruker, T.; Heining, P.; Schmouder, R.; Francis,
G.; Aradhye, S.; Burtin, P. Nat. Rev. Drug Disc. 2010, 9, 883.
5. Lu, X.; Sun, C.; Valentine, W. J.; E, S.; Liu, J.; Tigyi, G.; Bittman, R. J. Org. Chem.
2009, 74, 3192.
6. Tonelli, F.; Lim, K. G.; Loveridge, C.; Long, J.; Pitson, S. M.; Tigyi, G.; Bittman, R.;
Pyne, S.; Pyne, N. J. Cell. Signal. 2010, 22, 1536.
34. For a review of fluorination reactions using deoxo-fluor and DAST, see: Singh,
R. P.; Shreeve, J. M. Synthesis 2002, 2561.
35. Paugh, S. W.; Paugh, B. S.; Rahmani, M.; Kapitonov, D.; Kordula, T.; Milstien, S.;
Adams, J. K.; Zipkin, R. E.; Grant, S.; Spiegel, S. Blood 2008, 112, 1382.
36. Baek, D. J.; MacRitchie, N.; Pyne, N. J.; Pyne, S.; Bittman, R. Chem. Commun.
2013, 49, 2136.
7. Lim, K. G.; Tonelli, F.; Li, Z.; Lu, X.; Bittman, R.; Pyne, S.; Pyne, N. J. J. Biol. Chem.
2011, 286, 18633.