Synthesis and Biological Evaluation of an Indomethacin Library Reveals a New Class of Angiogenesis-Related Kinase Inhibitors

Article abstract of DOI:10.1002/anie.200352582

Capture and release: Indomethacin has served as a lead compound for the synthesis of a library that revealed a new class of kinase inhibitors. A synthetic approach termed "resin-capture-release" was developed (see scheme) that allowed a large library to be synthesized easily. Six of the library compounds were found to be inhibitors of angiogenesis-related receptor tyrosine kinases.

Full text of DOI:10.1002/anie.200352582

Communications
Kinase Inhibitors
Synthesis and Biological Evaluation of an
Indomethacin Library Reveals a New Class of
Angiogenesis-Related Kinase Inhibitors**
Claudia Rosenbaum, Patrick Baumhof,
Ralf Mazitschek, Oliver Müller, Athanassios Giannis,*
and Herbert Waldmann*
The combinatorial synthesis of small-molecule libraries on
solid supports has emerged as a powerful and widely used
method for the discovery of new biologically active com-
pounds, in particular those for subsequent application in
chemical biology and medicinal chemistry research. The
choice of the underlying molecular framework of the
individual library members is vital to the success of this
approach. The hit rates in biochemical and cell biological
assays, as well as the quality of the hits and the lead
compounds derived from them are expected to be particularly
high, even with relatively small libraries, if the compound
class in question can be regarded as biologically prevali-
dated.[1] In this regard, indoles represent one of the most
relevant structural classes. The plethora of indole-based
biologically active natural products and indole-derived
drugs spans an enormous range of biological activity. Con-
sequently, the synthesis of indole-based compound libraries
has received a great deal of attention, in particular in drug
development.[2] Among the biologi-
cally active indole derivatives, indo-
methacin (1) is of particular interest.
This compound belongs to the non-
steroidal antiinflammatory drugs
(NSAIDs), which are widely applied
in the treatment of, for example, pain,
arthritis,[3] cardiovascular diseases,[4,5]
[*] Dipl.-Chem. P. Baumhof, Dr. R. Mazitschek, Prof. Dr. A. Giannis
Max-Planck-Institut für molekulare Physiologie
Abteilung Chemische Biologie
Otto-Hahn-Strasse 11, 44227 Dortmund (Germany)
and
Universität Dortmund, Fachbereich 3, Organische Chemie
Fax: ( +49)231-133-2499
E-mail: giannis@chemie.uni-leipzig.de
Dr. C. Rosenbaum, Prof Dr. H. Waldmann
Universität Leipzig
Institut für Organische Chemie
Johannisallee 29, 04103 Leipzig (Germany)
Fax: ( +49)341-9736-599
E-mail: herbert.waldmann@mpi-dortmund.mpg.de
PD Dr. O. Müller
Max-Planck-Institut für molekulare Physiologie
Abteilung Strukturelle Biologie
[**] This research was supported by the Max-Planck-Gesellschaft and
the Fonds der Chemischen Industrie.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
224
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200352582
Angew. Chem. Int. Ed. 2004, 43, 224 –228

Angewandte
Chemie
and Alzheimer's disease,[6,7] and in the treatment and
prevention of cancer.[8,9]
desired indole derivatives 10. In this reaction sequence, the
water hydrolyzes the hydrazones and thereby releases selec-
tively acylated hydrazines 7 into solution, where they
condense with ketones 8 to give hydrazone intermediates 9.
These compounds then undergo a [3,3] sigmatropic rear-
rangement under the reaction conditions, which shifts the
equilibrium between the hydrazones 6 and 9 irreversibly to
the desired side. This strategy employs ketones, acid chlorides,
and hydrazines as building blocks, all of which are either
commercially available in great variety or readily prepared by
numerous well-established synthesis methods. Use of these
building blocks therefore allows ready construction of a fairly
diverse compound collection.
It has been demonstrated that indomethacin induces G1
arrest and apoptosis of human colorectal cancer cells by
influencing the Wnt signaling pathway,[10] downregulates the
transcriptional activity of the peroxisome proliferation-acti-
vated receptor d,[11] and inhibits the formation of new blood
vessels from pre-existing ones (angiogenesis) through direct
effects on endothelial cells mediated by the mitogen-activated
protein (MAP) kinase signaling pathway.[9] Although the
precise molecular targets of indomethacin in these processes
have not been identified unambiguously, it is clear that
indomethacin fulfils the criterium of biological prevalidation.
In fact, its core may be regarded as a privileged structure.[12]
We have developed a synthesis of an indomethacin library
and investigated the biological activity of the library com-
pounds with respect to possible inhibition of receptor tyrosine
kinases involved in angiogenesis. When planning the syn-
thesis, we envisioned that application of the Fischer indole
synthesis would allow a fairly diverse library to be built up
from three readily available building blocks in a very short
reaction sequence.[13] To establish a particularly practical and
efficient method, we implemented a “resin-capture-release”
strategy[14] that obviated the need for permanent attachment
to and final detachment from the solid support, which would
have required linker groups to be added and removed in two
additional steps. The chosen strategy is summarized in
Scheme 1. An aldehyde resin 2 is condensed with hydrazines
3 to yield polymer-bound hydrazones 4. The polymer in
intermediates 4 serves on the one hand as a reagent-capturing
auxiliary, which allows easy removal of surplus reagent, and
on the other hand as a temporary blocking function for the
terminal NH2 group of the hydrazine. Regioselective N-
acylation of the second nitrogen atom in the subsequent
step is thereby guaranteed. Surplus reagent is also readily
removed from the selectively N-acylated hydrazones 6. These
intermediates are then subjected to treatment with acid and
ketones 8 in the presence of traces of water to yield the
The building blocks shown in Scheme 2 were subjected to
this resin-capture-release indole synthesis reaction sequence
(for experimental details, see the Supporting Information).
The solid support was
a polystyrene aldehyde resin
(Advanced Chemtech) with a loading of 0.9 mmolgÀ1. A
library of 197 indomethacin analogues was synthesized (see
the Supporting Information) in overall yields ranging from
4% to quantitative (see Table 1). For example, indomethacin
1 was obtained by this four-step process in an overall yield of
63% after purification by chromatography.
These results demonstrate that the synthesis is compatible
with a variety of different functional groups in each building
block. Electron-withdrawing and electron-donating substitu-
ents can be incorporated in the hydrazine and the acid
chloride, and electron-poor and electron-rich heterocycles are
tolerated as well. The ketones may also incorporate different
heteroatom substituents. Not unexpectedly, the overall yield
was highest if activating electron-donating substituents were
present in the hydrazines. This result indicates the importance
of having two hydrazone-formation reactions in the whole
process. However, the synthesis of N-acylated indoles by the
resin-capture-release strategy is successful even in the pres-
ence of deactivating SO3H, COOH, and NO2 groups. The
crude reaction products were obtained with purities of
approximately 70 to over 99%, depending on the ketone
employed. All compounds were isolated in more
than 99% purity by simple chromatography.
Our initial biochemical investigation of the
indomethacin-derived compound library was
directed by the intriguing ability of the parent
compound to inhibit angiogenesis. Angiogenesis,
the development of new blood vessels from preexist-
ing ones, is central to wound repair, inflammation,
and embryonic development. Furthermore, aberrant
angiogenesis is considered to be a key step in tumor
growth, spread, and metastasis.[15,16] Vascular devel-
opment depends on endothelium-specific receptor
tyrosine kinases, in particular the vascular endothe-
lial growth factor receptors 1–3 (VEGFR1–3) and
the Tie-2 receptor.[17] All these receptors have been
implicated in tumor angiogenesis[18–22] and antagoni-
zation of Tie-2, VEGFR-2, or VEGF-D (a ligand of
VEGFR-3) inhibits tumor growth and tumor meta-
stasis in vivo.[21,23,24] The development of low-molec-
ular-weight inhibitors of these receptor tyrosine
kinases is among the most promising approaches to
Scheme 1. Concept of the “resin-capture-release” indole synthesis. DCE=1,2-
dichloroethane, TFA=trifluoroacetic acid.
Angew. Chem. Int. Ed. 2004, 43, 224 –228
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
225

Communications
kinase-activating protein upstream in the signal-
ing cascades that converge at MAP kinase.
Given the importance of the receptor tyrosine
kinases mentioned above and the fact that they
signal through MAP kinase, we speculated that
the observed antiangiogenic effect of indometha-
cin might be mediated at least in part by inhibition
of these kinases. This inhibiting effect appears to
be relatively weak, since 0.5 mm indomethacin
were required to inhibit angiogenesis in the
cellular test system.[9] Screening of an indometha-
cin-derived compound library could therefore
reveal analogues with higher biological activity.
To investigate this hypothesis, 134 compounds
considered to be representative of the entire
library were selected and assayed as possible
inhibitors of VEGFR-2, VEGFR-3, Tie-2, and
fibroblast growth factor receptor 1 (FGFR-1). In
addition, insulin-like growth factor 1 receptor
(IGF1R) was included in the screen. IGF1R
affects cell mitogenesis, survival, transformation,
and insulin-like activities by binding its ligands,
IGF1 and IGF2. This receptor influences postna-
tal growth physiology and its activity has been
associated with malignant disorders such as breast
cancer.[27] The antiapoptotic effect induced by the
IGF1/IGF1R system correlates with the induction
of chemoresistance in various tumors.[28]
Scheme 2. Building blocks employed in the “capture-and-release” Fischer indole synthesis.
Table 1: Results of the library synthesis.
Of the 134 compounds investigated, 6 inhibit
the kinases with IC50 values in the low-micromolar
range (see Table 2).[29] Indomethacin itself does
not inhibit any of the kinases at concentrations of up to
100 mm. Although a clear structure–activity relationship
cannot be deduced from the limited number of compounds
investigated, several preliminary conclusions may be drawn.
A considerable degree of selectivity between the kinases can
be achieved, and both selectivity and potency can be
mediated by changing the substituents, in particular those
on the indole ring. Among the N-acyl groups, the p-
chlorobenzoic acid amide group appears to be particularly
favorable, but not limiting. The presence of a polar substitu-
ent at the 3-position of the indole core structure is not
beneficial for inhibitory activity. These initial findings indi-
cate that it should be possible to develop more potent and
selective inhibitors based on the molecular skeleton defined
by indomethacin-type compounds.
Our data do not clearly prove a direct link between the
antiangiogenic properties of indomethacin and inhibition of
angiogenesis-related receptor tyrosine kinases. However, the
finding that closely related analogues of the drug are active
inhibitors of these kinases suggests that such a link might
indeed exist. It is possible that, under the conditions of the
cellular assay mentioned above, the fairly hydrophobic
indomethacin concentrates in the plasma membrane and
thereby creates local concentrations that are substantially
higher than the overall concentration. These high local
concentrations might lead to inhibition of the receptors.
The discovery that indomethacin-derived indole deriva-
tives are inhibitors of angiogenesis-related receptor tyrosine
Building blocks
Yield[a] [%]
Building blocks
Yield[a] [%]
A-I
A-V
A-VII
B-I
C-I
45–quant.
9–26
C-IX
D-I
E-I
20–47
47
62–quant.
21–60
19–45
16–63
32–81
15–81
17–49
9–99
40–90
4–58
4–52
15–20
7–55
11–66
10–29
4–26
11–29
F-I
F-IV
F-VI
F-VII
F-VIII
F-IX
G-I
C-II
C-III
C-IV
C-VI
C-VII
C-VIII
21–46
H-I
[a] Yield ranges are given for the conversion of the respective N-acylated
hydrazones by treatment with ketones 1–16. All yields are based on the
initial loading of the aldehyde resin (0.9 mmolgÀ1) and refer to purified
products. The compounds were identified by LC-MS, GC-MS, NMR
spectroscopy, and high-resolution mass spectrometry.
the development of new, alternative antitumor drugs, and
several inhibitors of VEGFR-2 are in clinical trials.[25,26]
A high concentration (0.5 mm) of indomethacin inhibits
angiogenesis in vitro after stimulation of endothelial cells
with the angiogenic growth factors VEGFand basic fibroblast
growth factor.[9] Analysis of the cells indicated that treatment
with the drug led to reduced activity of MAP kinase, but it
remained unclear whether this low activity was a result of
direct inhibition of the enzyme or inhibition of a MAP-
226
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 224 –228

Angewandte
Chemie
Table 2: Inhibition of angiogenesis-related receptor tyrosine kinases by members of the compound
library.
Keywords: angiogenesis ·
combinatorial chemistry ·
indomethacin · kinase inhibitors ·
medicinal chemistry
.
IC50 for receptor tyrosine kinase [mm][a]
Compound
VEGFR-2
VEGFR-3
Tie-2
IGF1R
FGFR-1
[1] R. Breinbauer, I. R. Vetter, H.
Waldmann, Angew. Chem. 2002,
114, 3002 – 3015; Angew. Chem.
Int. Ed. 2002, 41, 2878 – 2890.
[2] For reviews on the indole ring as a
privileged structure and combina-
torial syntheses of indole-based
compound libraries, see: a) D. A.
Horton, G. T. Bourne, M. L.
Smythe, Chem. Rev. 2003, 103,
893 – 930; b) S. Bräse, C. Gil, K.
Knepper, Bioorg. Med. Chem. 2002,
10, 2415 – 2437.
21Æ8.8
n.d.
3Æ1.8
9Æ3.5
6Æ2.2
20Æ8.8
9Æ11.4
17Æ6.2
6Æ4.6
9Æ2.7
4Æ1.5
17Æ4.9
15Æ3.6
n.d.
7Æ1.2
[3] A. A. Schuna, J. Am. Pharm. Assoc.
1998, 38, 728 – 735.
[4] S. H. Goodnight, Curr. Opin. Hem-
atol. 1996, 3, 355 – 360.
[5] L. A. Rodriguez, C. Varas, C.
Patrono, Epidemiology 2000, 11,
382 – 387.
[6] P. D. Sloane, Am. Fam. Physician
1998, 58, 1577 – 1586.
[7] B. L. Flynn, K. A. Theesen, Ann.
Pharmacother. 1999, 33, 840 – 849.
[8] S. J. Shiff, B. Rigas, Gastroenterol-
ogy 1997, 113, 1992 – 1998.
[9] M.K. Jones, H. Wang, B. M. Peskar,
E. Levin, R. M. Itani, I. J. Sarfels,
A. S. Tarnawski, Nat. Med. 1999, 5,
1418 – 1423.
[10] G. Hawcraft, M. D'Amico, C. Alba-
nese, A. F. Markhom, R. G. Pestell,
M. A. Kull, Carcinogenesis 2002, 23,
107 – 114.
[11] T.-C. He, T. A. Chan, B. Vogelstein,
K. W. Kinzler, Cell 1999, 99, 335 –
345.
n.a.
11Æ1.4
11Æ1.5
21Æ0.8
n.d.
n.d.
n.d.
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
19Æ2.8
10Æ0.3
[a] To assay the inhibitory activity the kinase-catalyzed phosphorylation of poly(Glu4-Tyr) was measured
in the presence of varying concentrations of inhibitor. The kinases were employed in the form of fusion
proteins of glutathione-S-transferase and the respective kinase domain. The relative amount of
phosphorylated substrate was quantified by means of an antiphosphortyrosine enzyme-linked
immunosorbent assay with an antiphosphotyrosine-antibody-conjugated horseradish peroxidase
(POD). The bound antibody was detected by the light emitted after addition of a chemiluminescence
substrate for POD. All IC50 values were calculated from at least four independent determinations. IC50
concentration required for 50% inhibition. n.d.: not determined. n.a.: remaining enzyme activity more
than 50% at 50 mm inhibitor concentration.
:
kinases, in particular Tie-2, VEGFR-3, and FGFR-1, as well
as the IGF1 receptor is of interest because only very few
classes of inhibitors are known for Tie-2, VEGFR-3, FGFR-1,
and IGF1R.[25,30,31] The structural framework defined by the
indomethacin core and the ease of its synthetic variation, for
example, by the “resin-capture-release” strategy described
herein, open up new opportunities for the development of
antiangiogenesis drugs and antagonists of the IGF1 receptor.
Compounds 11–13 are not particularly cytotoxic; exposure of
T98g brain cells to these compounds at a concentration of
10 mm for four days did not lead to an increase in cell
mortality.
[12] B. E. Evans, K. E. Rittle, M. G. Bock, R. M. DiPardo, R. M.
Freidinger, W. L. Whitter, G. F. Lundell, D. F. Veber, P. S.
Anderson, R. S. L. Chang, V. J. Lotti, D. J. Cerino, T. B. Chen,
P. J. Kling, K. A. Kunkel, J. P. Springer, J. Hirshfield, J. Med.
Chem. 1988, 31, 2235 – 2246.
[13] For Fischer indole syntheses on solid support, see: a) S. M.
Hutchins, K. T. Chapman, Tetrahedron Lett. 1996, 37, 4869 –
4872; b) L. Yang, Tetrahedron Lett. 2000, 41, 6981 – 6984;
c) L. C. Cooper, G. G. Chicchi, K. Dinnell, J. M. Elliott, G. J.
Hollingworth, M. M. Kurtz, K. L. Locker, D. Morrison, D. E.
Shaw, K.-L. Tsao, A. P. Watt, A. R. Williams, C. J. Swain, Bioorg.
Med. Chem. Lett. 2001, 11, 1233 – 1236; d) J. Tois, R. Franzꢀn, O.
Aitio, K. Huikko, J. Taskinen, Tetrahedron Lett. 2000, 41, 2443 –
2446.
In addition to the opportunity for the development of
antiangiogenesis drugs, the finding that the indomethacin
core structure defines a new class of kinase inhibitors is of
general relevance to medicinal chemistry and chemical
biology.
[14] A. Kirschning, H. Monenschein, R. Wittenberg, Chem. Eur. J.
2000, 6, 4445 – 4450.
[15] a) J. Folkman, Nat. Med. 1995, 1, 27 – 31; b) A. Giannis, F.
Rꢁbsam, Angew. Chem. 1997, 109, 606 – 609; Angew. Chem. Int.
Ed. Engl. 1997, 36, 588 – 590.
[16] a) J. Folkman, N. Engl. J. Med. 1971, 285, 1182 – 1186; b) P.
Carmeliet, R. K. Jain, Nature 2000, 407, 249 – 257.
[17] a) G. D. Yancopoulos, S. Davis, N. W. Gale, J. R. Rudge, S. J.
Wiegand, J. Holash, Nature 2000, 407, 242 – 248; b) P. C.
Maisonpierre, C. Suri, P. F. Jones, S. Bartunkova, S. J. Wiegand,
Received: August 7, 2003 [Z52582]
Angew. Chem. Int. Ed. 2004, 43, 224 –228
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
227

Communications
C. Radziejewski, D. Compton, J. McClain, T. H. Aldrich, N.
Papadopoulos, T. J. Daly, S. Davis, T. N. Sato, G. D. Yancopoulos,
Science 1997, 277, 55 – 60.
[18] S. Hiratsuka, Y. Maru, A. Okada, M. Seiki, T. Noda, M. Shibuya,
Cancer Res. 2001, 61, 1207 – 1213.
[19] H. Kubo, T. Fujiwara, L. Jussila, H. Hashi, M. Ogawa, K.
Shimizu, M. Awane, Y. Sakai, A. Takabayashi, K. Alitalo, Y.
Yamaoka, S. I. Nishikawa, Blood 2000, 96, 546 – 553.
[20] A. Stratmann, T. Acker, A. M. Burger, K. Amann, W. Risau,
K. H. Plate, Int. J. Cancer 2001, 91, 273 – 282.
[21] S. A. Stacker, C. Caesar, M. E. Baldwin, G. E. Thornton, R. A.
Williams, R. Prevo, D. G. Jackson, S. Nishikawa, H. Kubo, M. G.
Achen, Nat. Med. 2001, 7, 186 – 191.
[22] M. Skobe, T. Hawighorst, D. G. Jackson, R. Prevo, L. Janes, P.
Velasco, L. Riccardi, K. Alitalo, K. Claffey, M. Detmar, Nat.
Med. 2001, 7, 192 – 198.
[23] P. Lin, J. A. Buxton, A. Acheson, C. Radziejewski, P. C.
Maisonpierre, G. D. Yancopoulos, K. M. Channon, L. P. Hale,
M. W. Dewhirst, S. E. George, K. G. Peters, Proc. Natl. Acad.
Sci. USA 1998, 95, 8829 – 8834.
[24] J. Drevs, I. Hofmann, H. Hugenschmidt, C. Wittig, H. Madjar,
M. Muller, J. Wood, G. Martiny-Baron, C. Unger, D. Marme,
Cancer Res. 2000, 60, 4819 – 4824.
[25] D. H. Boschelli, Drugs Future 1999, 24, 515 – 537.
[26] See, for example: G. Bold, K.-H. Altmann, J. Frei, L. Marc, P. W.
Manley, P. Traxler, B. Wietfeld, J. Brꢁggen, E. Buchdunger, R.
Cozens, S. Ferrari, P. Furet, F. Hofmann, G. Martiny-Baron, J.
Mestan, J. Rꢂsel, M. Sills, D. Stover, F. Acemoglu, E. Boss, R.
Emmenegger, L. Lässer, E. Masso, R. Roth, C. Schlachter, W.
Vetterli, D. Wyss, J. M. Wood, J. Med. Chem. 2000, 43, 2310 –
2323, and references cited therein.
[27] M. J. Ellis, S. Jenkins, J. Hanfelt, M. E. Redington, M. Taylor, R.
Leek, K. Siddle, A. Harris, Breast Cancer Res. Treat. 1998, 52,
175.
[28] A. Grothey, W. Voigt, C. Schober, T. Muller, W. Dempke, H. J.
Schmoll, J. Cancer Res. Clin. Oncol. 1999, 125, 166 – 173.
[29] The structure–activity relationship apparent from the selected
134 compounds and comparison with the structures of the
remaining 63 compounds indicated that biochemical investiga-
tion of the latter would not reveal more potent inhibitors.
Therefore, 63 of the library compounds were not screened.
[30] P. Cohen, Nat. Rev. Drug Discovery 2002, 1, 309 – 315; A. J.
Bridges, Chem. Rev. 2001, 101, 2541 – 2571.
[31] L. Kissau, P. Stahl, R. Mazitschek, A. Giannis, H. Waldmann, J.
Med. Chem. 2003, 46, 2917 – 2931.
228
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 224 –228