Potent and selective inhibition of acid sphingomyelinase by bisphosphonates

Full text of DOI:10.1002/anie.200903288

Communications
DOI: 10.1002/anie.200903288
Enzyme Inhibitors
Potent and Selective Inhibition of Acid Sphingomyelinase by
Bisphosphonates**
Anke G. Roth, Daniela Drescher, Yang Yang, Susanne Redmer, Stefan Uhlig, and
Christoph Arenz*
Dedicated to Professor Konrad Sandhoff on the occasion of his 70th birthday
The acid sphingomyelinase (aSMase) is emerging as an
putative sphingolipid- and cholesterol-rich membrane micro-
domains. These so-called “lipid rafts” have been suggested to
[
1–3]
important drug target for a variety of diseases.
Inhibition
[13]
of aSMase prevents bacterial infections in a rat model of
act as “signaling platforms”,
and there is significant
[4]
cystic fibrosis and formation of acute lung injury (ALI)
elicited by endotoxin, acid instillation, or platelet-activating
evidence that the cleavage of sphingomyelin to ceramide
can dramatically alter the biophysical properties of the
[
5]
[14]
factor (PAF). Moreover, aSMase is essential for infection of
putative rafts.
In addition, it is well established that
[6]
non-phagocytotic cells with Neisseria gonorrhoeae and
ceramide is a potent inductor of apoptosis, which is the
main reason for cell degeneration in many of the diseases
mentioned above. However, it is unknown whether ceramide
acts by remodeling the plasma membrane or by interacting
with proteins like cathepsin B, which is involved in cellular
signaling. Beside aSMase, two cytosolic, magnesium-depen-
dent and membrane-bound neutral sphingomyelinases
[
7]
formation of pulmonary emphysema. Pharmacological or
genetic inhibition of aSMase prevents apoptosis and degen-
[8]
eration of liver cells in a mouse model for Wilsonꢀs disease.
In addition, there are several reports that aSMase signifi-
[9]
cantly contributes to the formation of atherosclerotic plugs.
This promising progress in aSMase research, based on
sophisticated animal models and cultured cells from patients,
is thwarted, however, by the lack of potent and selective
inhibitors of this enzyme. Phosphatidylinositol-3,5-bisphos-
phate (PtdIns3,5P ), to date the most potent inhibitor (K =
[
15]
(nSMase1 and nSMase2)
and an alkaline sphingomyeli-
[
16]
nase are known, whose cellular function is rather unclear.
Recently nSMase has been shown to be essential for the
formation of exosomes, lipid vesicles that play a key role in
the infection by retroviruses. In contrast to aSMase, there
are some potent small-molecule inhibitors for nSMase.
[17]
2
M
[
10]
[18]
0
.53 mm), is not suitable for cell culture studies, because of
[
19]
its fivefold negative charge and its two long fatty acid chains
which cause it to stack in cellular membranes. Last but not
least, this inhibitor is labile towards phospholipases A , A , C,
Our attempts at synthesizing phosphonate analogues of
PtdIns3,5P as potential inhibitors of aSMase yielded only
1
2
2
[27]
and D and phosphoinositide phosphatases.
moderately active substances.
However, we also gained
The aSMase is a soluble lysosomal sphingolipid hydrolase,
which constitutively degrades sphingomyelin from internal-
access to a collection of (bis)phosphonates that had been
synthesized in the GDR Academy of Sciences and that
contained some compounds that are structurally related to
our phosphoinositide analogues. When we initially tested
these substances at a concentration of 20 mm, we were
surprised that some of them were potent inhibitors of
aSMase (Tables 1 and 2). Among these substances, a-amino-
[
11]
ized membrane fragments. Upon stimulation, however, a
portion of this enzyme can be found on the outer side of the
[12]
plasma membrane.
This membrane-associated enzyme
shows biochemical activity in serum and urine and has been
termed secretory sphingomyelinase (sSMase), although it is
virtually identical to the lysosomal variant. Its activity is
elevated in several diseases. The secretory form of aSMase is
believed to play an important role in signal transduction, since
it alters the composition of the plasma membrane within
Table 1: Inhibition of aSMase by the initial phosphonate collection.
1
2
[a]
Cpd
R
R
Inhibition [%]
[
*] A. G. Roth, D. Drescher, S. Redmer, Prof. Dr. C. Arenz
Institut fꢀr Chemie, Humboldt Universitꢁt zu Berlin
Brook-Taylor-Strasse 2, 12489 Berlin (Germany)
Fax: (+49)30-2093-6947
1
2
3
H
H
16
2
NH2
47
E-mail: christoph.arenz@chemie.hu-berlin.de
Y. Yang, Prof. Dr. S. Uhlig
4
5
H
H
0
Insitut fꢀr Pharmakologie und Toxikologie
Medical Faculty, RWTH—Aachen University
Wendlingweg 2, 52075 Aachen (Germany)
[
**] The authors gratefully acknowledge funding by the Deutsche
Forschungsgemeinschaft (DFG AR376/2-1, DFG Uh 88/8-1) and the
Volkswagen-Stiftung.
À5
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200903288.
[a] The inhibition values were determined in a single experiment at
20 mm.
7560
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7560 –7563

Angewandte
Chemie
Table 2: Inhibition of aSMase by the initial bisphosphonate library.
1
2
[a]
Cpd
R
R
Inhibition [%]
6
OH
NH2
NH2
54
92
93
7
7
a
b
8
9
CH3
H
62
76
1
1
0
1
H
H
8
32
1
2
H
2
1
1
3
4
H
H
36
24
Scheme 1. Synthesis of bisphosphonates. Reagents and conditions:
a) NaH, toluene, 608C, 16 h; b) HCl, reflux, 16 h; c) P(OMe) , 08C,
3
2
9
h; d) HP(OMe) , nBu NH, 08C, 16 h; e) H PO , MsOH; then PCl ,
2 2 3 3 3
08C, 16 h; f) PCl /H PO , 708C, 12 h, then H O, 2 h.
3
3
3
2
[a] The inhibition values were determined in a single experiment at
2
0 mm.
aSMase-inhibiting bisphosphonates with regard to their
metal-binding properties, we tested 7c in the presence of
2
+
2+
2+
bisphosphonate 7b turned out to be about one order of
millimolar concentrations of Ca , Mg , and Zn ions,
respectively. The inhibitory activity was not significantly
diminished by the metal ions. However, the poor inhibition of
aSMase by N-phenylaminobisphosphonate 10 and by (w-
aminoalkyl)hydroxybisphosphonate 16e clearly shows that
complex-forming propensity and the presence of a hydro-
phobic moiety alone are not sufficient for aSMase inhibition.
In addition, we tested nearly all substances for inhibition
magnitude more potent than PtdIns3,5P . Furthermore, 7b
2
contains two more methylene units than 7a, which leads to a
dramatic increase in inhibitory potency.
To gain a deeper insight into the structure–activity
relationship, we decided to synthesize 15 additional
bisphosphonates harboring different functional groups at
the a position and displaying lipid tails of different lengths,
[
20]
2+
according to well-established protocols (Scheme 1). On the
basis of this new collection of compounds, we could show that
inhibition correlates with the length of the lipid tail (this
correlation is true as long as the substances are readily
soluble) and that a functional group with lone electron pairs
of the Mg -dependent nSMase, without observing any
substantial inhibition of this isoenzyme at concentrations up
to 100 mm (see the Supporting Information for details). This
clearly indicates that inhibition of aSMase is not only very
potent but highly selective for only this enzyme. Moreover,
we tested the best aSMase inhibitor, 7c, for inhibition of Ser/
Thr phosphatase 1 (PP1), which—like the phosphodiesterase
domain of aSMase—belongs to a family of dimetal-containing
phosphoesterases. The latter enzyme was not inhibited by
7c, even at a concentration of 2 mm (see the Supporting
Information for details), which shows that this aSMase
inhibitor is ineffective against PP1.
at the a position (-NH more strongly than -OH) increases the
2
inhibition of acid sphingomyelinase relative to the activity of
H-bisphosphonates 15a–d.
[23]
Bisphosphonates are known to form bidentate complexes
2
+
2+
2+ [21]
with divalent metal ions like Ca , Zn , and Mg . If the
bisphosphonate contains an additional hydroxy or amino
group, even more stable tridentate complexes can be formed.
In fact, a-amino substitution leads to more stable complexes
than a-hydroxy substitution, suggesting that aSMase inhib-
ition also correlates with the tendency of the compounds to
Bisphosphonates are well-established drugs against osteo-
[
21]
porosis, bone cancer, and several other bone diseases. The
key for bisphosphonate specificity is their capacity to bind
rapidly and with high affinity to unknown structures on bone
surfaces. The high affinity to the bone surface leads to the fast
2
+
form complexes in which the Zn ion resides at the reactive
center of the aSMase. It is noteworthy that aSMase, both in its
2
+
[24]
lysosomal and its secreted form, is a Zn -dependent enzyme.
However, the lysosomal variant is not inhibited by EDTA and
clearance of bisphosphonates from blood and soft tissues.
Interestingly, zoledronic acid (18), a widely used drug against
osteoporosis, showed a marked inhibition of aSMase with an
IC50 value of approximately 5 mm (Table 3). This finding
emphasizes that the pharmacological relevance of aSMase
2
+
not stimulated by Zn ; this can be explained by abundance of
2
+
Zn ions in the lysosomes. In contrast, the secreted variant is
2
+
[22]
stimulated by Zn ions.
In order to characterize the
Angew. Chem. Int. Ed. 2009, 48, 7560 –7563
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7561

Communications
[
a]
Table 3: Inhibition of aSMase by the synthesized bisphosphonates.
Cpd
IC50 [mm]
Cpd
IC50 [mm]
Cpd
IC50 [mm]
7
7
7
7
1
1
a
b
c
d
5a
5b
4.66Æ1.07
0.04Æ0.01
0.02Æ0.00
0.29Æ0.09
0.35Æ0.08
0.31Æ0.12
15c
15d
16a
16b
16c
16d
0.30Æ0.05
0.17Æ0.04
0.16Æ0.04
0.08Æ0.01
0.07Æ0.01
6.80Æ2.40
16e
17a
17b
18
1.95Æ0.22
9.50Æ4.00
0.18Æ0.03
5.08Æ0.74
[a] The IC50 value for inhibition of nSMase was >100 mm for all
compounds.
inhibition by bisphosphonate drugs cannot completely be
ruled out.
Since aSMase produces the proapoptotic ceramide, inhib-
ition of this enzyme by RNAi or unspecific inhibitors
Figure 2. The aSMase inhibitor 7c reduces PAF-induced pulmonary
edema in isolated, ventilated, and perfused rat lungs (IPL). Weight
gain (DW) was measured 10 min after addition of PAF (5 nm).
[
25]
2+
reportedly protects cells from dexamethasone- and Cu -
[
8]
induced apoptosis. When we treated HepG2 liver cells with
À8
dexamethasone (10 m) in order to induce apoptosis, aSMase
inhibitor 7c at a concentration of 0.1 mm efficiently inhibited
apoptosis, as measured by DNA-fragmentation ELISA
phatase 1. The compound, which can be synthesized easily on
gram scale, is also active in cell culture and efficiently protects
HepG2 cells from dexamethasone-induced apoptosis. Thus,
we describe a powerful tool for aSMase research which is very
likely to replace the tricyclic antidepressants imipramine and
desipramine which have been widely used despite their
indirect and unspecific mode of action. Moreover, the fact
that bisphosphonates are potent aSMase inhibitors brings up
the exciting notion that this property may be relevant to their
well-established clinical usage in the treatment of osteopo-
rosis, a very common disease with a lifetime risk of 40–50%
(
Figure 1).
[26]
for women in North America. And finally, the finding that
the aSMase inhibitor 7c significantly reduces edema forma-
tion in isolated rat lungs suggests that parenteral adminis-
tration of bisphosphonates may be useful for the treatment of
acute lung disorders and other inflammatory ailments.
Experimental Section
Enzyme assays: Crude preparations containing aSMase or nSMase
Figure 1. The aSMase inhibitor 7c (0.1 mm) inhibits dexamethasone-
Dex)-induced apoptosis in HepG2 cells. The absorbance in a DNA-
fragmentation ELISA is plotted.
[
19d]
were made from stripped rat brains.
The micellar nSMase assays
(
1
4
using C-labeled sphingomyelin as the substrate were performed as
[
19d]
described previously.
The fluorescent aSMase assay was per-
formed in a 384-well plate using HMU-PC (6-hexadecanoylamino-4-
methylumbelliferylphosphorylcholine) as the substrate. Reaction
mixtures consisted of 13.3 mL HMU-PC, 13.3 mL reaction buffer
(100 mm NaOAc, pH 5.2, 0.2% (w/v) Na-TC, 0.02% (w/v) NaN3,
0.2% (v/v) Triton X-100), and 13.3 mL enzyme preparation. Inhibitors
were added in various concentrations, and the reaction mixtures were
incubated for 3 h at 378C in a plate reader (FLUOstar OPTIMA,
BMG labtech). The fluorescence of 6-hexadecanoylamino-4-methyl-
Encouraged by the high biological activity in cultured
cells and because of the evident pharmacological interest in
potent aSMase inhibitors for the treatment of lung diseases,
we examined whether the inhibitor would also be able to
reduce PAF-induced pulmonary edema, similar to the unspe-
cific and indirect aSMase inhibitor imipramine. Indeed,
addition of 7c to the perfusate reduced edema formation in a
concentration-dependent fashion in isolated, ventilated, and
perfused rat lungs (IPL, Figure 2). Like imipramine (10 mm),
the inhibitor 7c attenuated but not completely prevented
edema formation in this model.
[5]
umbelliferone (HMU) was measured (excitation 380 nm, emission
14
4
60 nm) in real time. Assays using the C-labeled sphingomyelin gave
the same results.
Compound libraries and syntheses: All described compounds
1
13
31
were characterized by H, C, and P NMR spectroscopy and mass
spectrometry. The syntheses were performed as described previ-
[
20]
ously.
The simple bisphosphonate 7c is the most potent aSMase
inhibitor found to date. Its selectivity for aSMase is more than
Apoptosis assay: First, the kinetics of DNA fragmentation after
dexamethasone donation was measured in the lysate and in the
supernatant. Between 6 h and 8 h, there was a steep increase in
absorbance in the probes from the supernatant, which is typical for
apoptosis (data not shown). The apoptosis assay was performed
2
+
5
000 times greater than that for the Mg -dependent iso-
enzyme nSMase, and it is also nonselective for the dimetal-
containing remote aSMase homologue Ser/Thr protein phos-
7562
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7560 –7563

Angewandte
Chemie
according to the manufacturerꢀs protocol (Roche catalogue number
1585045). Briefly, cells were harvested and suspended in culture
medium (2 ꢁ 10 cellsmL ) containing BrdU labeling solution (10 mm
[8] P. A. Lang, M. Schenck, J. P. Nicolay, J. U. Becker, D. S. Kempe,
1
A. Lupescu, S. Koka, K. Eisele, B. A. Klarl, H. Rubben, K. W.
Schmid, K. Mann, S. Hildenbrand, H. Hefter, S. M. Huber, T.
Wieder, A. Erhardt, D. Haussinger, E. Gulbins, F. Lang, Nat.
Med. 2007, 13, 164.
5
À1
final concentration) and plated in a 96-well cell-culture dish at ꢀ 1 ꢁ
4
1
0 cells per well. After 16 h, the cells were washed and new media
À8
was added. Then, the cells were treated with 10 m dexamethasone
and 0.1 mm 7. After the cells had been incubated for 7 h, 100 mL of the
supernatant was collected and added to a 96-well plate containing
immobilized anti-BrdU antibody. After incubation, removal of the
supernatant, and extensive washing, the secondary antibody and the
TMB substrate were added and absorbance was measured at 370 nm
[9] C. M. Devlin, A. R. Leventhal, G. Kuriakose, E. H. Schuchman,
K. J. Williams, I. Tabas, Arterioscler. Thromb. Vasc. Biol. 2008,
28, 1723.
[10] M. Kꢃlzer, C. Arenz, K. Ferlinz, N. Werth, H. Schulze, R.
Klingenstein, K. Sandhoff, Biol. Chem. 2003, 384, 1293.
[11] T. Kolter, K. Sandhoff, Angew. Chem. 1999, 111, 1632; Angew.
Chem. Int. Ed. 1999, 38, 1532.
(FLUOstar OPTIMA, BMG labtech). The experiment was per-
formed in five times.
[12] S. Marathe, S. L. Schissel, M. J. Yellin, N. Beatini, R. Mintzer,
K. J. Williams, I. Tabas, J. Biol. Chem. 1998, 273, 4081.
[13] K. Simons, E. Ikonen, Nature 1997, 387, 569.
[14] Megha, E. London, J. Biol. Chem. 2004, 279, 9997.
[15] a) B. Liu, D. F. Hassler, G. K. Smith, K. Weaver, Y. A. Hannun,
J. Biol. Chem. 1998, 273, 34472; b) A. A. Karakashian, N. V.
Giltiay, G. M. Smith, M. N. Nikolova-Karakashian, FASEB J.
2004, 18, 968.
PAF-induced pulmonary edema: Female Wistar rats (weight 220
to 250 g) were kept under standard laboratory conditions (rat food
and water ad libitum). Rat lungs were prepared, perfused, and
ventilated essentially as described. Briefly, lungs were perfused
through the pulmonary artery at a constant hydrostatic pressure
[
5]
(12 cm H
2
O) with Krebs–Henseleit buffer containing 2% albumin,
0
.1% glucose, and 0.3% HEPES. Edema formation was assessed by
continuously measuring the weight gain of the lung. In this model,
platelet-activating factor causes rapid edema formation that is in part
dependent on acid sphingomyelinase. Inhibitor 7c was dissolved in
buffer and added to the buffer reservoir 10 min prior to adminis-
tration of PAF (5 nmol). Isolated perfused rat lungs were perfused for
[16] L. Nyberg, R. D. Duan, J. Axelson, A. Nilsson, Biochim.
Biophys. Acta Lipids Lipid Metab. 1996, 1300, 42.
[17] K. Trajkovic, C. Hsu, S. Chiantia, L. Rajendran, D. Wenzel, F.
Wieland, P. Schwille, B. Brugger, M. Simons, Science 2008, 319,
1244.
3
0 min before 7c was added to the perfusate; 10 min later 5 nmol PAF
[18] M. Marsh, G. van Meer, Science 2008, 319, 1191.
[19] a) F. Nara, M. Tanaka, T. Hosoya, K. Suzuki-Konagai, T. Ogita,
J. Antibiot. 1999, 52, 525; b) C. Arenz, A. Giannis, Angew. Chem.
2000, 112, 1498; Angew. Chem. Int. Ed. 2000, 39, 1440; c) C.
Arenz, M. Thutewohl, O. Block, H. Waldmann, H. J. Altenbach,
A. Giannis, ChemBioChem 2001, 2, 141; d) V. Wascholowski, A.
Giannis, Angew. Chem. 2006, 118, 841; Angew. Chem. Int. Ed.
2006, 45, 827.
was added as a bolus and weight gain was monitored for 10 min. Data
are shown as mean (with standard deviation) from four independent
experiments in each group. Statistics: 0.1 mm 7c: p < 0.01 vs. PAF
alone; 1 mm 7c: p < 0.01 vs. PAFalone and vs. 0.1 mm 7c/PAF (Tukeyꢀs
test).
Received: June 17, 2009
Published online: September 10, 2009
[
20] a) L. M. Nguyen, E. Niesor, C. L. Bentzen, J. Med. Chem. 1987,
0, 1426; b) G. R. Kieczykowski, R. B. Jobson, D. G. Melillo,
3
D. F. Reinhold, V. J. Grenda, I. Shinkai, J. Org. Chem. 1995, 60,
8310; c) D. V. Griffiths, J. M. Hughes, J. W. Brown, J. C. Caesar,
S. P. Swetnam, S. A. Cumming, J. D. Kelly, Tetrahedron 1997, 53,
Keywords: bioorganic chemistry · bisphosphonates ·
drug targets · enzyme inhibitors · sphingolipids
.
17815; d) S. H. Szajnman, E. L. Ravaschino, R. Docampo, J. B.
Rodriguez, Bioorg. Med. Chem. Lett. 2005, 15, 4685.
21] R. G. Russell, Ann. N. Y. Acad. Sci. 2006, 1068, 367.
22] S. L. Schissel, G. A. Keesler, E. H. Schuchman, K. J. Williams, I.
Tabas, J. Biol. Chem. 1998, 273, 18250.
23] M. Seto, M. Whitlow, M. A. McCarrick, S. Srinivasan, Y. Zhu, R.
Pagila, R. Mintzer, D. Light, A. Johns, J. A. Meurer-Ogden,
Protein Sci. 2004, 13, 3172.
24] A. Schindeler, D. G. Little, J. Pharm. Sci. 2007, 96, 1872.
25] M. G. Cifone, G. Migliorati, R. Parroni, C. Marchetti, D.
Millimaggi, A. Santoni, C. Riccardi, Blood 1999, 93, 2282.
26] D. Elaine, M. Maysam Abdin, C. Cyrus, Rheum. Dis. Clin. North
Am. 2006, 32, 617.
[
1] S. Uhlig, E. Gulbins, Am. J. Respir. Crit. Care Med. 2008, 178,
100.
[
[
1
[
[
[
2] A. Billich, T. Baumruker, Subcell. Biochem. 2008, 49, 487.
3] E. L. Smith, E. H. Schuchman, FASEB J. 2008, 22, 3419.
4] V. Teichgrꢂber, M. Ulrich, N. Endlich, J. Riethmuller, B. Wilker,
C. C. De Oliveira-Munding, A. M. van Heeckeren, M. L. Barr,
G. von Kurthy, K. W. Schmid, M. Weller, B. Tummler, F. Lang,
H. Grassme, G. Doring, E. Gulbins, Nat. Med. 2008, 14, 382.
5] R. Gꢃggel, S. Winoto-Morbach, G. Vielhaber, Y. Imai, K.
Lindner, L. Brade, H. Brade, S. Ehlers, A. S. Slutsky, S. Schutze,
E. Gulbins, S. Uhlig, Nat. Med. 2004, 10, 155.
[
[
[
[
[
[
[
6] H. Grassmꢄ, E. Gulbins, B. Brenner, K. Ferlinz, K. Sandhoff, K.
Harzer, F. Lang, T. F. Meyer, Cell 1997, 91, 605.
27] C. Arenz, unpublished results.
[
7] I. Petrache, V. Natarajan, L. Zhen, T. R. Medler, A. T. Richter,
C. Cho, W. C. Hubbard, E. V. Berdyshev, R. M. Tuder, Nat. Med.
2005, 11, 491.
Angew. Chem. Int. Ed. 2009, 48, 7560 –7563
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7563