Discovery of novel sphingosine kinase 1 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.09.022

Sphingosine kinase 1 (SK1) is an important enzyme that regulates the balance between ceramide and sphingosine-1-phosphate (S1P). Potent and novel SK1 inhibitors (6ag, 9ab and 12aa) have been discovered through a series of modifications of sphingosine (1),

Full text of DOI:10.1016/j.bmcl.2009.09.022

Bioorganic & Medicinal Chemistry Letters 19 (2009) 6119–6121
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of novel sphingosine kinase 1 inhibitors
*
Yibin Xiang , Gary Asmussen, Michael Booker, Bradford Hirth, John L. Kane Jr., Junkai Liao,
Kevin D. Noson, Christopher Yee
Drug and Biomaterial R&D, Genzyme Corporation, 153 Second Avenue, Waltham, MA 02451, USA
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 an important enzyme that regulates the balance between ceramide and
sphingosine-1-phosphate (S1P). Potent and novel SK1 inhibitors (6ag, 9ab and 12aa) have been discov-
ered through a series of modifications of sphingosine (1), the substrate of this enzyme.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 18 August 2009
Revised 2 September 2009
Accepted 4 September 2009
Available online 10 September 2009
Keywords:
Sphingosine kinase
Enzyme
Inhibitor
SK1
3-Hydroxyproline
A large body of evidence from various academic and industrial
laboratories indicates that sphingolipid metabolites can play
important roles in biological processes far beyond the confines of
cellular membranes.1 The balance between cellular levels of sphin-
golipids is important for regulating cell function. For instance, it
has been reported that ceramide and sphingosine (1) induce apop-
tosis or growth arrest while sphingosine-1-phosphate (S1P, 2)
mediates proliferation and angiogenesis.2 Sphingosine kinases
(SKs), which mediate the conversion of 1 to 2 (Fig. 1), are important
enzymes in the sphingolipid metabolic pathway as they sit in a
crucial position to regulate the relative levels of S1P, sphingosine,
and ceramide.3
To date, two mammalian sphingosine kinases (SK1 and SK2)
have been identified. Although both enzymes are capable of phos-
phorylating sphingosine, there are reports indicating that they
have different cellular functions.4 SK1 promotes cell growth and
proliferation,5 whereas SK2 has the opposite effects.6 Much of the
research in this area has been directed towards understanding
the mechanism of the cellular and physiological functions of these
enzymes. Recently, SK1 has been implicated in several pathological
states, including various immune-mediated diseases, inflammation
and cancer.7
the sphingolipid metabolism pathway, and also lead to new poten-
tial approaches to the treatment of cancer and/or immune-medi-
ated diseases. Recently, Spiegel and co-workers revealed the
structure of a SK1-selective inhibitor (SK1-I, 11).9 The authors re-
port that this compound effectively reduced the growth and sur-
vival of human leukemia U937 cells in vitro. Furthermore, 11 has
been reported to suppress U937 growth in a murine xenograft
model.
In this Letter, we report our discovery of a novel class of SK1-
selective inhibitors based on a modified sphingosine scaffold. Our
original hypothesis for designing these inhibitors involved replac-
ing the aminodiol headpiece of sphingosine with a serine amide.
Given the complementary functionality, we believed that the
N,N-Dimethylsphingosine (DMS, 3) has been widely used for
modulating S1P biosynthesis, even though 3 is a relatively weak
and non-selective SK inhibitor.8 Identification of a more potent
and selective SK1 inhibitor could provide a useful tool for studying
* Corresponding author. Tel.: +1 781 434 3653; fax: +1 781 672 5823.
E-mail address: yibin.xiang@genzyme.com (Y. Xiang).
Figure 1.
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.09.022

6120
Y. Xiang et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6119–6121
Table 1
new molecules generated by this modification would have affinity
to the SK. Additionally, the carboxylic acid of serine provides a con-
venient synthetic handle that would allow for installing a variety of
mimics for the lipophilic tail of sphingosine. A similar approach has
been successfully employed by Macdonald and co-workers in their
efforts to identify novel S1P receptor agonists.10 Thus, a standard
IC50s of sphingosine kinase 1 (SK1) inhibition11
*
Compd#
IC50 (lM)
3
5.7
5.0
4.3
>10
>10
2.1
2.8
0.65
2.1
1.6
6aa
6ab
6ac
6ad
6ae
6af
6ag
6ah
6ai
EDC-mediated amidation of N-t-Boc-L-serine with 4-octylaniline
followed by removal of the carbamate protecting group with TFA
afforded 6ab (Scheme 1). This compound and its enantiomer,
6aa, proved to be modest inhibitors of SK1, with potencies similar
to DMS (Table 1).11
Derivatives 6ac–bl were prepared with focusing on altering the
serine headpiece. The assay results revealed that the substitution
pattern had a significant impact on the potency of the compounds.
6aj
1.1
3.2
>10
5.0
6ak
6bl
7
Neither the O-methyl or a-methylated derivative (6ad or 6ac) had
9aa
9ab
9ac
9ad
9ae
9bc
10
12aa
12ab
12ac
12ba
12ca
13
2.2
activity against SK1, while the N-mono-methylated derivative 6af
demonstrated slightly improved potency relative to 6ab. A sub-
stantial increase in activity was observed when L-threonine was
incorporated as the polar headpiece. The resulting compound,
6ag, was nearly 10-fold more potent than 6ab (650 nM vs
4.3 lM). Similar levels of potency were also observed for the iso-
meric threonine analogs 6ah–6aj, although in all cases, activity
was reduced compared to 6ag.
Compounds 9aa–9bc were prepared to explore the impact of
altering the distance between the terminal alcohol functionality
and the amide group (Scheme 2). Interestingly, the homoserine
derivative 9ac was approximately 25 times more potent than the
serine derivative 6ab (180 nM vs 4.3 lM). In addition, the stereo-
chemistry of the homoserine analogs had a significant impact on
0.05
0.18
>10
4.0
3.5
>10
0.062
3.4
8.3
0.43
6.2
0.74
*
IC50: Concentration of the testing compounds to inhibit 50% activity of the
enzyme. The IC50s were determination based on a method described in Ref. 11. All
the values of IC50s are the average of at least two times determinations.
activity; the S-enantiomer 9ab was about 40 times more potent
than the R-enantiomer 9aa (50 nM vs 2.2
l
M). Further increases
the incorporation of an additional hydroxy substituent at the 4-po-
sition (12ac).
in the distance between the alcohol and amide groups, as demon-
strated in compounds 9ad and 9ae, significantly decreased activity
against SK1.
Encouraged by the results obtained with 6ag and 9ab, we made
further modifications to the polar headpiece by preparing a series
of 3-hydroxyproline analogs 12aa–ca (Scheme 3). In this series, the
most potent compound (12aa) was nearly 80-fold more active than
In all three series (6, 9, 12), the amide functionality was impor-
tant for activity. Amine-containing analogs (prepared by borane
reduction of the corresponding amides) had reduced activity
against SK1 (cf. 6ab vs 7, 9ab vs 10, 12aa vs 13). Furthermore,
the 4-octylanilide derivatives were more potent than the 4-octylb-
enzyl amide derivatives (cf. 6ab vs 6bl, 9ac vs 9bc and 12aa vs
12ba). None of the SK1 active compounds demonstrated any activ-
6ab (62 nM vs 4.3
lM). Inversion of the stereochemistry at the 3-
position (12ab) resulted in a marked decrease in potency as did
ity versus SK2 when screened at a concentration of 10 lM.
Scheme 1. Reagents: (a) EDC, DMAP, CH2Cl2; (b) TFA, CH2Cl2; (c) BH3–SMe2, THF.

Y. Xiang et al. / Bioorg. Med. Chem. Lett. 19 (2009) 6119–6121
6121
Scheme 2. Reagents: (a) EDC, DMAP, CH2Cl2; (b) H2, Pd/C, MeOH; (c) HCl, dioxane; (d) BH3–SMe2, THF.
Scheme 3. Reagents: (a) EDC, DMAP, CH2Cl2; (b) TFA, CH2Cl2; (c) BH3–SMe2, THF.
Khosrowbeygi, A.; Gao, S.; Miwa, N.; Jahangeer, S.; Nakamura, S. J. Biol. Chem.
2005, 280, 36318.
7. Tara, T. A.; Hannun, Y. A.; Obeid, L. M. J. Biochem. Mol. Biol. 2006, 39, 113.
8. Fujita, S.; Sugimoto, M.; Ogawa, T.; El-Ghendy, K.; Racker, E. Biochemistry 1989,
28, 6796.
9. Steven, W. P.; Paugh, B. S.; Rahman, M.; Kapitonov, D.; Almenera, J. A.; Kordula,
S. M.; Adams, J. K.; Zipkin, R. E.; Grant, S.; Spiegel, S. Blood 2008, 112, 1382.
10. Clemens, J. J.; Davis, M. D.; Lynch, K. R.; Macdonald, T. L. Bioorg. Med. Chem. Lett.
2003, 13, 3401.
11. The in vitro SK1 inhibition assays were conducted in the following manner:
In a 384 well polystyrene plate, human sphingosine kinase 1 (300 ng/ml,
from BPS Bioscience San Diego CA) was incubated in a 100 mM Hepes pH 7.5
In conclusion, we have identified novel potent SK1 inhibitors
through structural modification of sphingosine. Several of these
compounds, including 9ab and 12aa, are significantly more potent
than the previously reported SK1 inhibitor, N,N-dimethylsphingo-
sine. Further optimization of the 3-hydroxyproline series with
aims to improve in vitro activity and ADME properties will be re-
ported in a separate publication.
References and notes
buffer containing 2
Salt Lake City, UT), 25
l
M fluorescein labeled sphingosine (Echelon Biosciences,
1. Hannun, Y. A.; Luberto, C.; Argraves, K. M. Biochemistry 2001, 40, 4893.
2. (a) Maceyka, M.; Payne, S. G.; Milstien, S.; Spiegel, S. Biochim. Biophys. Acta
2002, 1585, 193; (b) Ogretmen, B.; Hannun, Y. A. Nat. Rev. Cancer 2004, 4, 604.
3. De Jonghe, S.; Van Overmeire, I.; Poulton, S.; Hendrix, C.; Busson, R.; Van
Calenbergh, S.; De Keukeleire, D.; Spiegel, S.; Herdewijn, P. Bioorg. Med. Chem.
Lett. 1999, 9, 3175.
4. (a) Kohama, T.; Olivera, A.; Edsall, L.; Nagiec, M. M.; Dickson, R.; Spiegel, S. J.
Biol. Chem. 1998, 273, 23722; (b) Melendez, A. J.; Carlos-Dias, E.; Gosink, M.;
Allen, J. M.; Takacs, L. Gene 2000, 251, 19.
5. (a) Xia, P.; Wang, L.; Moretti, P. A.; Albanese, N.; Chai, F.; Pitson, S. M.; D’Andrea,
R. J.; Gamble, J. R.; Vadas, M. A. J. Biol. Chem. 2002, 277, 7996; (b) Bonhuure, E.;
Pchejetski, D.; Aouali, N.; Morjani, H.; Levade, T.; Kohama, T.; Cuvillier, O.
Leukemia 2002, 20, 95.
6. (a) Maceyka, M.; Sankala, H.; Hait, N. C.; Stunff, H. L.; Liu, H.; Toman, R.; Collier,
C.; Zhang, M.; Satin, L. S.; Merrill, A. H.; Milstien, S., Jr.; Spiegel, S. J. Biol. Chem.
2005, 280, 37118; (b) Okada, T.; Ding, G.; Sonoda, H.; Kajimoto, T.; Haga, Y.;
lM ATP, 0.05% Triton X-100, 10% glycerol, 4 mM DTT,
20 mM MgCl2 and compound with a final concentration of 4.35% DMSO for
one hour at room temperature. The reaction was terminated by adding
40 mM EDTA in a 100 mM Hepes pH 7.5 buffer which contained 0.05% Triton
X-100, 2.5 % glycerol, 0.3% CR-3 (Caliper Life Sciences, Hopkinton, MA) and
0.6% DMSO. The plate was run for one cycle on a LabChip 3000 (Caliper Life
Sciences, Hopkinton, MA) in an off-chip mobility shift assay with an upstream
voltage of À500 V,
a
downstream voltage of À2400 volts and
a vacuum
pressure of À2.1 psi. The sample sip time was 0.2 s. The LabChip separates
and measures the amount of fluorescein labeled sphingosine and fluorescein
labeled sphingosine-1-phosphate present in each well. Results are expressed
as percent conversion of fluorescein labeled sphingosine by measuring peak
height for both the substrate and product. On every plate 100% inhibition
(substrate without enzyme) and 0% inhibition (substrate with enzyme and
DMSO) controls were used to calculate percent inhibition of tested
compounds.