Syntheses of fluorescence-labeled sphingosine 1-phosphate methylene and sulfur analogues as possible visible ligands to the receptor

Article abstract of DOI:10.1246/cl.2008.188

The methylene and sulfur analogues of sphingosine 1-phosphate labeled with a fluorescent group at the terminus of the backbone skeleton were stereoselectively synthesized as possible nonhydrolyzable visible ligands to the sphingosine 1-phosphate receptor,

Full text of DOI:10.1246/cl.2008.188

188
Chemistry Letters Vol.37, No.2 (2008)
Syntheses of Fluorescence-labeled Sphingosine 1-Phosphate Methylene
and Sulfur Analogues as Possible Visible Ligands to the Receptor
Tetsuya Yamamoto, Hiroko Hasegawa, Toshikazu Hakogi, and Shigeo Katsumuraꢀ
Department of Chemistry and Open Research Center on Organic Tool Molecules,
School of Science and Technology, Kwansei Gakuin University, 2-1 Gakuen, Sanda, Hyogo 669-1337
(Received October 15, 2007; CL-071137; E-mail: katsumura@ksc.kwansei.ac.jp)
The methylene and sulfur analogues of sphingosine 1-phos-
O
phate labeled with a fluorescent group at the terminus of the
backbone skeleton were stereoselectively synthesized as possi-
ble nonhydrolyzable visible ligands to the sphingosine 1-phos-
phate receptor, S1P1.
NHBoc
OTBS
NHBoc
OH
R
N
a
Ref. 7b
O
O
HO
OR'
8: R = H, R' = TBS
9: R = PMB, R' = Tf
NHBoc
OH
O
b-d
N-Boc-L-serine
7
R
N
O
CH2
OMe
e
h
P
CH2
OMe
OMe
P
O
OMe
OH
10: R = PMB
11: R = Boc
O
f,g
12
Sphingosine 1-phosphate (S1P) 1 is regarded as a remarka-
ble phospholipid1 because of its very attractive biological activ-
ities, such as the stimulation of DNA synthesis,2 cell growth,3
and platelet activation.4 During the course of our project to de-
velop effective tool molecules for the elucidation of the sphingo-
lipid behavior,5 we previously reported that the fluorescent-la-
beled sphingosine 1-phosphate (NBD-S1P) 2, which was recog-
nized as a ligand of its biological receptor (S1P1), could be
visualized it in cells under a fluorescent microscope.6 This
fluorescent molecule was, however, hydrolyzed at the moiety
of the phosphate ester function by S1P hydrolytic enzymes to
be metabolized into NBD-sphingosine and NBD-ceramide.
In this paper, we disclose the stereoselective syntheses of
both the fluorescence-labeled sphingosine 1-phosphonate 3
(NBD-C-S1P), the so-called methylene analogue, and sphingo-
sine 1-thiophosphate 4 (NBD-S-S1P), the so-called sulfur ana-
logue, as new second generation nonhydrolyzable tool molecules
to be useful for the elucidation of the S1P behavior in a cell.
In previous papers, we reported a versatile method for the
stereocontrolled synthesis of various kinds of sphingolipids,
such as natural sphingosine, ceramide, sphingosine 1-phosphate,
and sphingomyelin,7 by utilizing the olefin cross metathesis
protocol.8,9 As shown in Scheme 1, the syntheses of 3 and 4 were
then achieved by applying the cross-coupling method between
the fluorescence-labeled olefin 5 and amino alcohol parts 6.
The amino alcohol parts were stereoselectively synthesized
starting from the L-serine and from L-cysteine derivatives,
respectively.
O2N
NHBoc
k
i
(CH2)9
CH2
OMe
3
N
P
O
N
O
H
OMe
N
OR
13: R = H
14: R = TES
j
Scheme 2. Reagents and conditions: (a) NaH, THF, 50 ꢂC, 2 h.
(b) PMBCl, TBAI, NaH, THF, rt, 16 h. (c) 2 M HCl, MeOH, rt,
16 h. (d) Tf2O, 2,6-lutidine, CH2Cl2, ꢁ78 ꢂC, 94% (three steps).
(e) MePO(OMe)2, n-BuLi, THF, ꢁ78 ꢂC, 73%. (f) CAN, CH3CN,
H2O, 0 ꢂC, 80%. (g) Boc2O, Et3N, DMAP, CH2Cl2, 0 ꢂC, 94%.
(h) K2CO3, MeOH, rt, 74%. (i) 5, Grubbs cat. 2nd Generation,
CH2Cl2, reflux, 2 h, 85%. (j) TESCl, imidazole, DMF, 0 ꢂC. (k)
TMSBr, CH2Cl2, rt, then MeOH, rt, 68% (two steps).
which was synthesized from N-Boc-L-serine in 84% yield over
four steps as previously reported (Scheme 2).7b The introduction
of the dimethylphosphonylmethyl group to 8 was one of the key
steps of the synthesis of 3. After several trials, we found that the
electron-donating PMB group for protection of the oxazolidi-
none nitrogen and trifluoromethanesulfonyl group as a leaving
one were very important for the successful substitution reaction.
Thus, 8 was converted into 9 by the sequence of introducing a
PMB group, acid treatment, and formation of triflate in 94%
yield over three steps. The slow addition of the lithium anion
of dimethylphosphonylmethyl to the triflate at ꢁ78 ꢂC success-
fully produced the desired phosphonate 10 in 73% yield.10,11
The substitution reaction of the corresponding mesylate, bro-
mide, and iodide was unsuccessful. Attempts of a similar substi-
tution in the linear systems were also unsuccessful. Next, we ex-
changed the PMB group with a Boc group at this stage, because
the selective oxazolidinone ring opening of 10 without hydroly-
sis of the phosphonate ester moiety was not successful. Removal
of the PMB group of 10 by treatment with CAN followed by the
reaction with Boc2O produced 11 in 94% yield, which was treat-
ed with potassium carbonate in MeOH to afford the functional-
ized 12 in 74% yield. With the amino alcohol part in hand,
the olefin cross metathesis reaction was attempted under the
previously established conditions.7b Thus, 12 was stirred with
4 equiv. of the NBD-labeled olefin 5 and 0.03 equiv. of the
Grubbs catalyst 2nd generation in CH2Cl2 for 2 h at reflux to
produce the desired cross coupling product 13 in 85% yield,
whose geometry of the double bond was approximately 20:1
The oxazolidinone derivative 8, an intermediate of the
methylene analogue, was prepared by the NaH treatment of 7,
O2N
NH2
NH2
C13H27
O
OH
(CH2)9
O
OH
P
N
H
P
N
O
NBD Group
O
OH
OH
1 (S1P)
O
OH
N
OH
2 (NBD-S1P)
olefin cross
metathesis
NH2
NHBoc
XR
(CH2)9
NBD
X
OH
(CH2)9
NBD
N
P
N
H
H
O
OH
OH
OH
3 (NBD-C-S1P): X = CH2
4 (NBD-S-S1P): X = S
5
6: X = CH2, S
Amino Alcohol Part
Olefin Part
Scheme 1. Structure of sphingosine 1-phosphate and synthetic plan
of NBD-labeled analogues.
1
of the E:Z-based on the H NMR.
Copyright Ó 2008 The Chemical Society of Japan

Chemistry Letters Vol.37, No.2 (2008)
189
NHBoc
STBS
are expected to be useful nonhydrolyzable tool molecules to
investigate the behavior of sphingosine 1-phosphate in a cell.15
NHBoc
SBn
NHBoc
SR
a-c
e
HO
OTBS
18
O
OH
16: R = Bn
17: R = H
15
d
This work was partly supported by a Grant-in-Aid for Scien-
tific Research from the Ministry of Education, Culture, Sports,
Science and Technology, Japan. This work was also supported
by the Maching Fund Subsidy for Private University.
NHBoc
S
NHBoc
SH
NHBoc
S
h
f
OMe
OMe
P
O
OR
OTBS
19
OTBS
20
21: R = TBS
2
i
g
22: R = H
This paper is dedicated to the late Professor Yoshihiko Ito
for his outstanding contribution to synthetic organic chemistry.
O2N
N
NHBoc
S
j
k
(CH2)9
OMe
OMe
4
N
P
H
O
N
OH
O
References and Notes
23
1
For recent reviews on signal transduction mediated by sphingolipids, see:
a) J. Liao, J. Tao, G. Lin, D. Liu, Tetrahedron 2005, 61, 4715. b) A.
Delgado, J. Casas, A. Llebaria, J. L. Abad, ChemMedChem 2007, 2,
580. c) T. Baumruker, A. Billich, Curr. Immun. Rev. 2006, 2, 101, and
references cited therein.
.
Scheme 3. Reagents and conditions: (a) Me(MeO)NH HCl, EDCI,
NMM, CH2Cl2, ꢁ15 ꢂC. (b) Vinyl bromide, Mg, THF, rt. (c)
LiAl(Ot-Bu)3H, EtOH, ꢁ78 ꢂC. (d) Li, liq. NH3, THF, reflux,
70% (four steps). (e) TBSCl, DMAP, Et3N, DMF, 0 ꢂC. (f) TBAF,
THF, ꢁ78 ꢂC. (g) PBu3, CH3CN, H2O, 52% (three steps). (h)
Trimethylphosphite, CBr4, 2,6-lutidine, CH2Cl2, 0 ꢂC, 56%. (i)
2 M HCl, THF, rt, 75%. (j) 5, Grubbs cat. 2nd Generation, CH2Cl2,
reflux, 2 h, 85%. (k) TMSBr, CH2Cl2, rt, then MeOH, rt, 80%.
2
3
H. Zhang, N. N. Desai, A. Olivera, T. Seki, G. Brooker, S. Spiegel, J. Cell
Biol. 1991, 114, 155.
M. J. Lee, S. Thangada, K. P. Claffey, N. Ancellin, C. H. Liu, M. Kluk,
M. Volpi, R. I. Sha’afi, T. Hla, Cell 1999, 99, 301.
4
5
Y. Yatomi, F. Ruan, S. Hakomori, Y. Igarashi, Blood 1995, 86, 193.
We reported the syntheses of the following various sphingomyelin
analogues: a) Carbon analogues: T. Hakogi, Y. Monden, M. Taichi,
S. Iwama, S. Fujii, K. Ikeda, S. Katsumura, J. Org. Chem. 2002, 67,
4839. b) Nitrogen analogue: T. Hakogi, M. Taichi, S. Katsumura, Org.
Lett. 2003, 5, 2801. c) Sulfur analogue: T. Hakogi, S. Fujii, M. Morita,
K. Ikeda, S. Katsumura, Bioorg. Med. Chem. Lett. 2005, 15, 2141. d)
Difluoromethylene analogue: T. Hakogi, T. Yamamoto, S. Fujii, K. Ikeda,
S. Katsumura, Tetrahedron Lett. 2006, 47, 2627.
The secondary hydroxy group in 13 was first protected with
a TES group to avoid the side reaction such as lactonization
resulting from nucleophilic attack of the hydroxy group to the
phosphonate moiety during the following demethylation. The
obtained TES derivative 14 was then treated with trimethylsilyl
bromide12 in CH2Cl2 followed by treatment with MeOH. The
desired 3 (NBD-C-S1P) was obtained in a pure form in 68%
yield over two steps.13
The synthesis of NBD-S-S1P 4 was also achieved from
N-Boc-S-benzyl-L-cysteine 15 by a procedure similar to that of
NBD-C-S1P 3 (Scheme 3). Thus, the sequence of the Weinreb
amide formation, introduction of a vinyl group, the highly
anti-selective reduction of the resulting ketone with lithium
tri- tert-butoxyaluminohydride, and then treatment with lithium
in liquid ammonia produced 17 in 70% yield over four steps
through 16. In order to introduce a phosphate group into the pri-
mary mercapto group, the protection of the secondary hydroxy
group was necessary. The regioselective desilylation with 0.95
6
7
T. Hakogi, T. Shigenari, S. Katsumura, T. Sano, T. Kohno, Y. Igarashi,
Bioorg. Med. Chem. Lett. 2003, 13, 661.
a) H. Hasegawa, T. Yamamoto, S. Hatano, T. Hakogi, S. Katsumura,
Chem. Lett. 2004, 33, 1592. b) T. Yamamoto, H. Hasegawa, T. Hakogi,
S. Katsumura, Org. Lett. 2006, 8, 5569.
8
9
a) R. H. Grubbs, Handbook of Metathesis, Wiley-VCH, Germany, 2003.
b) S. Connon, S. Blechert, Angew. Chem., Int. Ed. 2003, 42, 1900, and
references cited therein.
For other reports on functionalized sphingolipids: a) P. Nussbaumer, P.
Ettmayer, C. Peters, D. Rosenbeiger, K. Hoegenauer, Chem. Commun.
2005, 5086. b) C. Peters, A. Billich, M. Ghobrial, K. Hoegenauer, T.
Ullrich, P. Nussbaumer, J. Org. Chem. 2007, 72, 1842, and references
cited therein.
10 D. B. Berkowitz, G. Maiti, B. D. Charette, C. D. Dreis, R. G. MacDonald,
Org. Lett. 2004, 6, 4921.
11 The alkylation of the triflate 9 with a dimethyl phosphonate anion resulted
in a low yield when 9 was added to the lithium anion of a dimethyl
methylphosphonate solution at ꢁ78 ꢂC in Ref. 10.
12 C. E. McKenna, J. Schmidhauser, J. Chem. Soc., Chem. Commun. 1979,
739.
.
equiv. of Bu4NF nH2O at the sulfur of 18, which was prepared
from 17 by the usual silylation, gave O-silylated thiol 19 as
a mixture of disulfide 20, which was converted into 19 by a
tributylphosphine treatment. The phosphorylation of 19 with tri-
methylphosphite and carbon tetrabromide successfully produced
dimethylthiophosphate 21 in 56% yield with the aid of 2,6-luti-
dine. The desired phosphorylated amino alcohol 22 was obtained
in 75% yield by the hydrochloric acid treatment of 21 in THF,
although the treatment in MeOH gave a complex mixture. With
three types of thiol derivatives, benzyl thioether 16, thiol 17, and
thiophosphate 22 in hand, the olefin cross metathesis reactions
between these thiol derivatives and olefin 5 were examined. In
the case of 16, the reaction poorly proceeded to afford the corre-
sponding coupling product in only 22% yield. The reaction with
thiol 17 did not proceed at all. On the other hand, the reaction of
22 smoothly proceeded to produce the desired coupling product
23 in 85% yield. The objective NBD-S-S1P 4 was successfully
obtained by removal of both the methyl and Boc groups using
trimethylsilyl bromide and then MeOH.14
20:5
13 Spectra data of 3: ½ꢀꢃD +2.8 (c ¼ 0:10, CH3OH); 1H NMR (CD3OD,
400 MHz), ꢁ 8.42 (d, J ¼ 8:5 Hz, 1H), 6.26 (d, J ¼ 8:9 Hz, 1H), 5.86
(td, J ¼ 6:6, 15.3 Hz, 1H), 5.48 (dd, J ¼ 6:6, 15.4 Hz, 1H), 4.30 (dd,
J ¼ 4:3, 5.7 Hz, 1H), 3.48 (m, 2H), 3.30 (m, 1H), 2.07 (td, J ¼ 6:8,
6.8 Hz, 2H), 2.06–1.81 (m, 4H), 1.75 (tt, J ¼ 7:3, 7.3 Hz, 2H), 1.46–
1.28 (m, 12H); 13C NMR (CD3OD, 100 MHz), ꢁ 146.5, 145.6, 145.3,
138.5, 136.9, 122.5, 99.6, 72.3, 57.5 (d, JC{P ¼ 15:3 Hz), 44.8, 33.4,
30.52, 30.45, 30.33, 30.30, 30.1, 29.2, 28.0, 24.7 (d, JC{P ¼ 13:9 Hz), 23.2.
25:5
14 Spectra data of 4: ½ꢀꢃD ꢁ3:9 (c ¼ 1:06, CH3OH); 1H NMR (CD3OD,
400 MHz), ꢁ 8.44 (d, J ¼ 8:9 Hz, 1H), 6.28 (d, J ¼ 8:9 Hz, 1H), 5.83
(td, J ¼ 6:9, 15.6 Hz, 1H), 5.44 (dd, J ¼ 6:6, 15.3 Hz, 1H), 4.28 (dd,
J ¼ 5:7, 5.7 Hz, 2H), 3.48 (m, 2H) 3.29 (m, 1H), 3.01 (ddd, J ¼ 3:9,
15.9, 15.9 Hz, 1H), 2.85 (ddd, J ¼ 8:9, 14.9, 16.2 Hz, 1H), 2.15 (td,
J ¼ 6:9, 7.1 Hz, 2H), 17.3 (tt, J ¼ 7:1, 7.1 Hz, 2H), 1.45–1.23 (m, 12H);
13C NMR (CDCl3, 100 MHz), ꢁ 146.6, 145.8, 145.4, 138.6, 137.0,
127.9, 122.7, 99.6, 72.2, 58.8, 44.9, 33.4, 30.54, 30.45, 30.35, 30.26,
30.08, 29.3, 29.0 (d, JC{P ¼ 11:4 Hz), 28.0.
15 In the preliminary qualitative tests, the synthesized 3 and 4 showed a mod-
erate ability as ligands toward the sphingosine 1-phosphate receptor,
S1P1, based on the results that the synthesized 3 and 4 reasonably expelled
the radiolabeled S1P similarly to NBD-S1P in S1P1-expressing Chinese
hamster ovary cells. We are grateful to Prof. Igarashi and co-workers of
Hokkaido University for their biological testing.
In conclusion, we achieved the syntheses of new fluores-
cence-labeled methylene and sulfur analogues, 3 and 4, by our
olefin cross metathesis protocol as the key step. These analogues
Published on the web (Advance View) January 19, 2008; doi:10.1246/cl.2008.188